diff --git a/.clang-format b/.clang-format index f49620f506f1..f3923a1f9858 100644 --- a/.clang-format +++ b/.clang-format @@ -78,6 +78,8 @@ ForEachMacros: - 'ata_qc_for_each_with_internal' - 'ax25_for_each' - 'ax25_uid_for_each' + - '__bio_for_each_bvec' + - 'bio_for_each_bvec' - 'bio_for_each_integrity_vec' - '__bio_for_each_segment' - 'bio_for_each_segment' @@ -118,10 +120,12 @@ ForEachMacros: - 'drm_for_each_legacy_plane' - 'drm_for_each_plane' - 'drm_for_each_plane_mask' + - 'drm_for_each_privobj' - 'drm_mm_for_each_hole' - 'drm_mm_for_each_node' - 'drm_mm_for_each_node_in_range' - 'drm_mm_for_each_node_safe' + - 'flow_action_for_each' - 'for_each_active_drhd_unit' - 'for_each_active_iommu' - 'for_each_available_child_of_node' @@ -158,6 +162,9 @@ ForEachMacros: - 'for_each_dss_dev' - 'for_each_efi_memory_desc' - 'for_each_efi_memory_desc_in_map' + - 'for_each_element' + - 'for_each_element_extid' + - 'for_each_element_id' - 'for_each_endpoint_of_node' - 'for_each_evictable_lru' - 'for_each_fib6_node_rt_rcu' @@ -195,6 +202,7 @@ ForEachMacros: - 'for_each_net_rcu' - 'for_each_new_connector_in_state' - 'for_each_new_crtc_in_state' + - 'for_each_new_mst_mgr_in_state' - 'for_each_new_plane_in_state' - 'for_each_new_private_obj_in_state' - 'for_each_node' @@ -210,8 +218,10 @@ ForEachMacros: - 'for_each_of_pci_range' - 'for_each_old_connector_in_state' - 'for_each_old_crtc_in_state' + - 'for_each_old_mst_mgr_in_state' - 'for_each_oldnew_connector_in_state' - 'for_each_oldnew_crtc_in_state' + - 'for_each_oldnew_mst_mgr_in_state' - 'for_each_oldnew_plane_in_state' - 'for_each_oldnew_plane_in_state_reverse' - 'for_each_oldnew_private_obj_in_state' @@ -243,6 +253,9 @@ ForEachMacros: - 'for_each_sg_dma_page' - 'for_each_sg_page' - 'for_each_sibling_event' + - 'for_each_subelement' + - 'for_each_subelement_extid' + - 'for_each_subelement_id' - '__for_each_thread' - 'for_each_thread' - 'for_each_zone' @@ -252,6 +265,8 @@ ForEachMacros: - 'fwnode_for_each_child_node' - 'fwnode_graph_for_each_endpoint' - 'gadget_for_each_ep' + - 'genradix_for_each' + - 'genradix_for_each_from' - 'hash_for_each' - 'hash_for_each_possible' - 'hash_for_each_possible_rcu' @@ -293,7 +308,11 @@ ForEachMacros: - 'key_for_each' - 'key_for_each_safe' - 'klp_for_each_func' + - 'klp_for_each_func_safe' + - 'klp_for_each_func_static' - 'klp_for_each_object' + - 'klp_for_each_object_safe' + - 'klp_for_each_object_static' - 'kvm_for_each_memslot' - 'kvm_for_each_vcpu' - 'list_for_each' @@ -324,6 +343,8 @@ ForEachMacros: - 'media_device_for_each_intf' - 'media_device_for_each_link' - 'media_device_for_each_pad' + - 'mp_bvec_for_each_page' + - 'mp_bvec_for_each_segment' - 'nanddev_io_for_each_page' - 'netdev_for_each_lower_dev' - 'netdev_for_each_lower_private' @@ -375,6 +396,7 @@ ForEachMacros: - 'rht_for_each_rcu' - 'rht_for_each_rcu_continue' - '__rq_for_each_bio' + - 'rq_for_each_bvec' - 'rq_for_each_segment' - 'scsi_for_each_prot_sg' - 'scsi_for_each_sg' @@ -410,6 +432,8 @@ ForEachMacros: - 'v4l2_m2m_for_each_src_buf_safe' - 'virtio_device_for_each_vq' - 'xa_for_each' + - 'xa_for_each_marked' + - 'xa_for_each_start' - 'xas_for_each' - 'xas_for_each_conflict' - 'xas_for_each_marked' diff --git a/.mailmap b/.mailmap index 37e1847c7988..ae2bcad06f4b 100644 --- a/.mailmap +++ b/.mailmap @@ -156,6 +156,8 @@ Morten Welinder Morten Welinder Mythri P K Nguyen Anh Quynh +Nicolas Pitre +Nicolas Pitre Paolo 'Blaisorblade' Giarrusso Patrick Mochel Paul Burton @@ -224,3 +226,5 @@ Yakir Yang Yusuke Goda Gustavo Padovan Gustavo Padovan +Changbin Du +Changbin Du diff --git a/Documentation/accounting/psi.txt b/Documentation/accounting/psi.txt index b8ca28b60215..7e71c9c1d8e9 100644 --- a/Documentation/accounting/psi.txt +++ b/Documentation/accounting/psi.txt @@ -56,12 +56,12 @@ situation from a state where some tasks are stalled but the CPU is still doing productive work. As such, time spent in this subset of the stall state is tracked separately and exported in the "full" averages. -The ratios are tracked as recent trends over ten, sixty, and three -hundred second windows, which gives insight into short term events as -well as medium and long term trends. The total absolute stall time is -tracked and exported as well, to allow detection of latency spikes -which wouldn't necessarily make a dent in the time averages, or to -average trends over custom time frames. +The ratios (in %) are tracked as recent trends over ten, sixty, and +three hundred second windows, which gives insight into short term events +as well as medium and long term trends. The total absolute stall time +(in us) is tracked and exported as well, to allow detection of latency +spikes which wouldn't necessarily make a dent in the time averages, +or to average trends over custom time frames. Cgroup2 interface ================= diff --git a/Documentation/bpf/btf.rst b/Documentation/bpf/btf.rst index 9a60a5d60e38..7313d354f20e 100644 --- a/Documentation/bpf/btf.rst +++ b/Documentation/bpf/btf.rst @@ -148,16 +148,16 @@ The ``btf_type.size * 8`` must be equal to or greater than ``BTF_INT_BITS()`` for the type. The maximum value of ``BTF_INT_BITS()`` is 128. The ``BTF_INT_OFFSET()`` specifies the starting bit offset to calculate values -for this int. For example, a bitfield struct member has: * btf member bit -offset 100 from the start of the structure, * btf member pointing to an int -type, * the int type has ``BTF_INT_OFFSET() = 2`` and ``BTF_INT_BITS() = 4`` +for this int. For example, a bitfield struct member has: + * btf member bit offset 100 from the start of the structure, + * btf member pointing to an int type, + * the int type has ``BTF_INT_OFFSET() = 2`` and ``BTF_INT_BITS() = 4`` Then in the struct memory layout, this member will occupy ``4`` bits starting from bits ``100 + 2 = 102``. Alternatively, the bitfield struct member can be the following to access the same bits as the above: - * btf member bit offset 102, * btf member pointing to an int type, * the int type has ``BTF_INT_OFFSET() = 0`` and ``BTF_INT_BITS() = 4`` diff --git a/Documentation/devicetree/bindings/arm/cpus.yaml b/Documentation/devicetree/bindings/arm/cpus.yaml index 365dcf384d73..82dd7582e945 100644 --- a/Documentation/devicetree/bindings/arm/cpus.yaml +++ b/Documentation/devicetree/bindings/arm/cpus.yaml @@ -228,7 +228,7 @@ patternProperties: - renesas,r9a06g032-smp - rockchip,rk3036-smp - rockchip,rk3066-smp - - socionext,milbeaut-m10v-smp + - socionext,milbeaut-m10v-smp - ste,dbx500-smp cpu-release-addr: diff --git a/Documentation/devicetree/bindings/display/amlogic,meson-dw-hdmi.txt b/Documentation/devicetree/bindings/display/amlogic,meson-dw-hdmi.txt index bf4a18047309..3a50a7862cf3 100644 --- a/Documentation/devicetree/bindings/display/amlogic,meson-dw-hdmi.txt +++ b/Documentation/devicetree/bindings/display/amlogic,meson-dw-hdmi.txt @@ -37,6 +37,7 @@ Required properties: - GXL (S905X, S905D) : "amlogic,meson-gxl-dw-hdmi" - GXM (S912) : "amlogic,meson-gxm-dw-hdmi" followed by the common "amlogic,meson-gx-dw-hdmi" + - G12A (S905X2, S905Y2, S905D2) : "amlogic,meson-g12a-dw-hdmi" - reg: Physical base address and length of the controller's registers. - interrupts: The HDMI interrupt number - clocks, clock-names : must have the phandles to the HDMI iahb and isfr clocks, @@ -66,6 +67,9 @@ corresponding to each HDMI output and input. S905X (GXL) VENC Input TMDS Output S905D (GXL) VENC Input TMDS Output S912 (GXM) VENC Input TMDS Output + S905X2 (G12A) VENC Input TMDS Output + S905Y2 (G12A) VENC Input TMDS Output + S905D2 (G12A) VENC Input TMDS Output Example: diff --git a/Documentation/devicetree/bindings/display/amlogic,meson-vpu.txt b/Documentation/devicetree/bindings/display/amlogic,meson-vpu.txt index 419444e2b4d3..be40a780501c 100644 --- a/Documentation/devicetree/bindings/display/amlogic,meson-vpu.txt +++ b/Documentation/devicetree/bindings/display/amlogic,meson-vpu.txt @@ -57,6 +57,7 @@ Required properties: - GXL (S905X, S905D) : "amlogic,meson-gxl-vpu" - GXM (S912) : "amlogic,meson-gxm-vpu" followed by the common "amlogic,meson-gx-vpu" + - G12A (S905X2, S905Y2, S905D2) : "amlogic,meson-g12a-vpu" - reg: base address and size of he following memory-mapped regions : - vpu - hhi @@ -83,6 +84,9 @@ corresponding to each VPU output. S905X (GXL) CVBS VDAC HDMI-TX S905D (GXL) CVBS VDAC HDMI-TX S912 (GXM) CVBS VDAC HDMI-TX + S905X2 (G12A) CVBS VDAC HDMI-TX + S905Y2 (G12A) CVBS VDAC HDMI-TX + S905D2 (G12A) CVBS VDAC HDMI-TX Example: diff --git a/Documentation/devicetree/bindings/display/bridge/ti,tfp410.txt b/Documentation/devicetree/bindings/display/bridge/ti,tfp410.txt index 3f903af93949..5ff4f64ef8e8 100644 --- a/Documentation/devicetree/bindings/display/bridge/ti,tfp410.txt +++ b/Documentation/devicetree/bindings/display/bridge/ti,tfp410.txt @@ -18,7 +18,14 @@ This device has two video ports. Their connections are modeled using the OF graph bindings specified in [1]. Each port node shall have a single endpoint. - Port 0 is the DPI input port. Its endpoint subnode shall contain a - pclk-sample property and a remote-endpoint property as specified in [1]. + pclk-sample and bus-width property and a remote-endpoint property as specified + in [1]. + - If pclk-sample is not defined, pclk-sample = 0 should be assumed for + backward compatibility. + - If bus-width is not defined then bus-width = 24 should be assumed for + backward compatibility. + bus-width = 24: 24 data lines are connected and single-edge mode + bus-width = 12: 12 data lines are connected and dual-edge mode - Port 1 is the DVI output port. Its endpoint subnode shall contain a remote-endpoint property is specified in [1]. @@ -43,6 +50,7 @@ tfp410: encoder@0 { tfp410_in: endpoint@0 { pclk-sample = <1>; + bus-width = <24>; remote-endpoint = <&dpi_out>; }; }; diff --git a/Documentation/devicetree/bindings/display/msm/gmu.txt b/Documentation/devicetree/bindings/display/msm/gmu.txt index 3439b38e60f2..90af5b0a56a9 100644 --- a/Documentation/devicetree/bindings/display/msm/gmu.txt +++ b/Documentation/devicetree/bindings/display/msm/gmu.txt @@ -24,7 +24,10 @@ Required properties: * "cxo" * "axi" * "mnoc" -- power-domains: should be <&clock_gpucc GPU_CX_GDSC> +- power-domains: should be: + <&clock_gpucc GPU_CX_GDSC> + <&clock_gpucc GPU_GX_GDSC> +- power-domain-names: Matching names for the power domains - iommus: phandle to the adreno iommu - operating-points-v2: phandle to the OPP operating points @@ -51,7 +54,10 @@ Example: <&gcc GCC_GPU_MEMNOC_GFX_CLK>; clock-names = "gmu", "cxo", "axi", "memnoc"; - power-domains = <&gpucc GPU_CX_GDSC>; + power-domains = <&gpucc GPU_CX_GDSC>, + <&gpucc GPU_GX_GDSC>; + power-domain-names = "cx", "gx"; + iommus = <&adreno_smmu 5>; operating-points-v2 = <&gmu_opp_table>; diff --git a/Documentation/devicetree/bindings/display/msm/gpu.txt b/Documentation/devicetree/bindings/display/msm/gpu.txt index aad1aef682f7..2b8fd26c43b0 100644 --- a/Documentation/devicetree/bindings/display/msm/gpu.txt +++ b/Documentation/devicetree/bindings/display/msm/gpu.txt @@ -22,9 +22,14 @@ Required properties: - qcom,adreno-630.2 - iommus: optional phandle to an adreno iommu instance - operating-points-v2: optional phandle to the OPP operating points +- interconnects: optional phandle to an interconnect provider. See + ../interconnect/interconnect.txt for details. - qcom,gmu: For GMU attached devices a phandle to the GMU device that will control the power for the GPU. Applicable targets: - qcom,adreno-630.2 +- zap-shader: For a5xx and a6xx devices this node contains a memory-region that + points to reserved memory to store the zap shader that can be used to help + bring the GPU out of secure mode. Example 3xx/4xx/a5xx: @@ -70,6 +75,12 @@ Example a6xx (with GMU): operating-points-v2 = <&gpu_opp_table>; + interconnects = <&rsc_hlos MASTER_GFX3D &rsc_hlos SLAVE_EBI1>; + qcom,gmu = <&gmu>; + + zap-shader { + memory-region = <&zap_shader_region>; + }; }; }; diff --git a/Documentation/devicetree/bindings/display/panel/feiyang,fy07024di26a30d.txt b/Documentation/devicetree/bindings/display/panel/feiyang,fy07024di26a30d.txt new file mode 100644 index 000000000000..82caa7b65ae8 --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/feiyang,fy07024di26a30d.txt @@ -0,0 +1,20 @@ +Feiyang FY07024DI26A30-D 7" MIPI-DSI LCD Panel + +Required properties: +- compatible: must be "feiyang,fy07024di26a30d" +- reg: DSI virtual channel used by that screen +- avdd-supply: analog regulator dc1 switch +- dvdd-supply: 3v3 digital regulator +- reset-gpios: a GPIO phandle for the reset pin + +Optional properties: +- backlight: phandle for the backlight control. + +panel@0 { + compatible = "feiyang,fy07024di26a30d"; + reg = <0>; + avdd-supply = <®_dc1sw>; + dvdd-supply = <®_dldo2>; + reset-gpios = <&pio 3 24 GPIO_ACTIVE_HIGH>; /* LCD-RST: PD24 */ + backlight = <&backlight>; +}; diff --git a/Documentation/devicetree/bindings/display/panel/innolux,p079zca.txt b/Documentation/devicetree/bindings/display/panel/innolux,p079zca.txt index d0f55161579a..3ab8c7412cf6 100644 --- a/Documentation/devicetree/bindings/display/panel/innolux,p079zca.txt +++ b/Documentation/devicetree/bindings/display/panel/innolux,p079zca.txt @@ -12,7 +12,7 @@ Optional properties: Example: &mipi_dsi { - panel { + panel@0 { compatible = "innolux,p079zca"; reg = <0>; power-supply = <...>; diff --git a/Documentation/devicetree/bindings/display/panel/innolux,p097pfg.txt b/Documentation/devicetree/bindings/display/panel/innolux,p097pfg.txt index 595d9dfeffd3..d1cab3a8f0fb 100644 --- a/Documentation/devicetree/bindings/display/panel/innolux,p097pfg.txt +++ b/Documentation/devicetree/bindings/display/panel/innolux,p097pfg.txt @@ -13,7 +13,7 @@ Optional properties: Example: &mipi_dsi { - panel { + panel@0 { compatible = "innolux,p079zca"; reg = <0>; avdd-supply = <...>; diff --git a/Documentation/devicetree/bindings/display/panel/kingdisplay,kd097d04.txt b/Documentation/devicetree/bindings/display/panel/kingdisplay,kd097d04.txt index 164a5fa236da..cfefff688614 100644 --- a/Documentation/devicetree/bindings/display/panel/kingdisplay,kd097d04.txt +++ b/Documentation/devicetree/bindings/display/panel/kingdisplay,kd097d04.txt @@ -12,7 +12,7 @@ Optional properties: Example: &mipi_dsi { - panel { + panel@0 { compatible = "kingdisplay,kd097d04"; reg = <0>; power-supply = <...>; diff --git a/Documentation/devicetree/bindings/display/panel/lg,acx467akm-7.txt b/Documentation/devicetree/bindings/display/panel/lg,acx467akm-7.txt new file mode 100644 index 000000000000..fc1e1b325e49 --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/lg,acx467akm-7.txt @@ -0,0 +1,7 @@ +LG ACX467AKM-7 4.95" 1080×1920 LCD Panel + +Required properties: +- compatible: must be "lg,acx467akm-7" + +This binding is compatible with the simple-panel binding, which is specified +in simple-panel.txt in this directory. diff --git a/Documentation/devicetree/bindings/display/panel/rocktech,jh057n00900.txt b/Documentation/devicetree/bindings/display/panel/rocktech,jh057n00900.txt new file mode 100644 index 000000000000..1b5763200cf6 --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/rocktech,jh057n00900.txt @@ -0,0 +1,18 @@ +Rocktech jh057n00900 5.5" 720x1440 TFT LCD panel + +Required properties: +- compatible: should be "rocktech,jh057n00900" +- reg: DSI virtual channel of the peripheral +- reset-gpios: panel reset gpio +- backlight: phandle of the backlight device attached to the panel + +Example: + + &mipi_dsi { + panel@0 { + compatible = "rocktech,jh057n00900"; + reg = <0>; + backlight = <&backlight>; + reset-gpios = <&gpio3 13 GPIO_ACTIVE_LOW>; + }; + }; diff --git a/Documentation/devicetree/bindings/display/rockchip/rockchip,rk3066-hdmi.txt b/Documentation/devicetree/bindings/display/rockchip/rockchip,rk3066-hdmi.txt new file mode 100644 index 000000000000..d1ad31bca8d9 --- /dev/null +++ b/Documentation/devicetree/bindings/display/rockchip/rockchip,rk3066-hdmi.txt @@ -0,0 +1,72 @@ +Rockchip specific extensions for rk3066 HDMI +============================================ + +Required properties: +- compatible: + "rockchip,rk3066-hdmi"; +- reg: + Physical base address and length of the controller's registers. +- clocks, clock-names: + Phandle to HDMI controller clock, name should be "hclk". +- interrupts: + HDMI interrupt number. +- power-domains: + Phandle to the RK3066_PD_VIO power domain. +- rockchip,grf: + This soc uses GRF regs to switch the HDMI TX input between vop0 and vop1. +- ports: + Contains one port node with two endpoints, numbered 0 and 1, + connected respectively to vop0 and vop1. + Contains one port node with one endpoint + connected to a hdmi-connector node. +- pinctrl-0, pinctrl-name: + Switch the iomux for the HPD/I2C pins to HDMI function. + +Example: + hdmi: hdmi@10116000 { + compatible = "rockchip,rk3066-hdmi"; + reg = <0x10116000 0x2000>; + interrupts = ; + clocks = <&cru HCLK_HDMI>; + clock-names = "hclk"; + power-domains = <&power RK3066_PD_VIO>; + rockchip,grf = <&grf>; + pinctrl-names = "default"; + pinctrl-0 = <&hdmii2c_xfer>, <&hdmi_hpd>; + + ports { + #address-cells = <1>; + #size-cells = <0>; + hdmi_in: port@0 { + reg = <0>; + #address-cells = <1>; + #size-cells = <0>; + hdmi_in_vop0: endpoint@0 { + reg = <0>; + remote-endpoint = <&vop0_out_hdmi>; + }; + hdmi_in_vop1: endpoint@1 { + reg = <1>; + remote-endpoint = <&vop1_out_hdmi>; + }; + }; + hdmi_out: port@1 { + reg = <1>; + hdmi_out_con: endpoint { + remote-endpoint = <&hdmi_con_in>; + }; + }; + }; + }; + +&pinctrl { + hdmi { + hdmi_hpd: hdmi-hpd { + rockchip,pins = <0 RK_PA0 1 &pcfg_pull_default>; + }; + hdmii2c_xfer: hdmii2c-xfer { + rockchip,pins = <0 RK_PA1 1 &pcfg_pull_none>, + <0 RK_PA2 1 &pcfg_pull_none>; + }; + }; +}; diff --git a/Documentation/devicetree/bindings/display/ste,mcde.txt b/Documentation/devicetree/bindings/display/ste,mcde.txt new file mode 100644 index 000000000000..4c33c692bd5f --- /dev/null +++ b/Documentation/devicetree/bindings/display/ste,mcde.txt @@ -0,0 +1,104 @@ +ST-Ericsson Multi Channel Display Engine MCDE + +The ST-Ericsson MCDE is a display controller with support for compositing +and displaying several channels memory resident graphics data on DSI or +LCD displays or bridges. It is used in the ST-Ericsson U8500 platform. + +Required properties: + +- compatible: must be: + "ste,mcde" +- reg: register base for the main MCDE control registers, should be + 0x1000 in size +- interrupts: the interrupt line for the MCDE +- epod-supply: a phandle to the EPOD regulator +- vana-supply: a phandle to the analog voltage regulator +- clocks: an array of the MCDE clocks in this strict order: + MCDECLK (main MCDE clock), LCDCLK (LCD clock), PLLDSI + (HDMI clock), DSI0ESCLK (DSI0 energy save clock), + DSI1ESCLK (DSI1 energy save clock), DSI2ESCLK (DSI2 energy + save clock) +- clock-names: must be the following array: + "mcde", "lcd", "hdmi" + to match the required clock inputs above. +- #address-cells: should be <1> (for the DSI hosts that will be children) +- #size-cells: should be <1> (for the DSI hosts that will be children) +- ranges: this should always be stated + +Required subnodes: + +The devicetree must specify subnodes for the DSI host adapters. +These must have the following characteristics: + +- compatible: must be: + "ste,mcde-dsi" +- reg: must specify the register range for the DSI host +- vana-supply: phandle to the VANA voltage regulator +- clocks: phandles to the high speed and low power (energy save) clocks + the high speed clock is not present on the third (dsi2) block, so it + should only have the "lp" clock +- clock-names: "hs" for the high speed clock and "lp" for the low power + (energy save) clock +- #address-cells: should be <1> +- #size-cells: should be <0> + +Display panels and bridges will appear as children on the DSI hosts, and +the displays are connected to the DSI hosts using the common binding +for video transmitter interfaces; see +Documentation/devicetree/bindings/media/video-interfaces.txt + +If a DSI host is unused (not connected) it will have no children defined. + +Example: + +mcde@a0350000 { + compatible = "ste,mcde"; + reg = <0xa0350000 0x1000>; + interrupts = ; + epod-supply = <&db8500_b2r2_mcde_reg>; + vana-supply = <&ab8500_ldo_ana_reg>; + clocks = <&prcmu_clk PRCMU_MCDECLK>, /* Main MCDE clock */ + <&prcmu_clk PRCMU_LCDCLK>, /* LCD clock */ + <&prcmu_clk PRCMU_PLLDSI>; /* HDMI clock */ + clock-names = "mcde", "lcd", "hdmi"; + #address-cells = <1>; + #size-cells = <1>; + ranges; + + dsi0: dsi@a0351000 { + compatible = "ste,mcde-dsi"; + reg = <0xa0351000 0x1000>; + vana-supply = <&ab8500_ldo_ana_reg>; + clocks = <&prcmu_clk PRCMU_DSI0CLK>, <&prcmu_clk PRCMU_DSI0ESCCLK>; + clock-names = "hs", "lp"; + #address-cells = <1>; + #size-cells = <0>; + + panel { + compatible = "samsung,s6d16d0"; + reg = <0>; + vdd1-supply = <&ab8500_ldo_aux1_reg>; + reset-gpios = <&gpio2 1 GPIO_ACTIVE_LOW>; + }; + + }; + dsi1: dsi@a0352000 { + compatible = "ste,mcde-dsi"; + reg = <0xa0352000 0x1000>; + vana-supply = <&ab8500_ldo_ana_reg>; + clocks = <&prcmu_clk PRCMU_DSI1CLK>, <&prcmu_clk PRCMU_DSI1ESCCLK>; + clock-names = "hs", "lp"; + #address-cells = <1>; + #size-cells = <0>; + }; + dsi2: dsi@a0353000 { + compatible = "ste,mcde-dsi"; + reg = <0xa0353000 0x1000>; + vana-supply = <&ab8500_ldo_ana_reg>; + /* This DSI port only has the Low Power / Energy Save clock */ + clocks = <&prcmu_clk PRCMU_DSI2ESCCLK>; + clock-names = "lp"; + #address-cells = <1>; + #size-cells = <0>; + }; +}; diff --git a/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.txt b/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.txt new file mode 100644 index 000000000000..b8be9dbc68b4 --- /dev/null +++ b/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.txt @@ -0,0 +1,92 @@ +ARM Mali Bifrost GPU +==================== + +Required properties: + +- compatible : + * Since Mali Bifrost GPU model/revision is fully discoverable by reading + some determined registers, must contain the following: + + "arm,mali-bifrost" + * which must be preceded by one of the following vendor specifics: + + "amlogic,meson-g12a-mali" + +- reg : Physical base address of the device and length of the register area. + +- interrupts : Contains the three IRQ lines required by Mali Bifrost devices, + in the following defined order. + +- interrupt-names : Contains the names of IRQ resources in this exact defined + order: "job", "mmu", "gpu". + +Optional properties: + +- clocks : Phandle to clock for the Mali Bifrost device. + +- mali-supply : Phandle to regulator for the Mali device. Refer to + Documentation/devicetree/bindings/regulator/regulator.txt for details. + +- operating-points-v2 : Refer to Documentation/devicetree/bindings/opp/opp.txt + for details. + +- resets : Phandle of the GPU reset line. + +Vendor-specific bindings +------------------------ + +The Mali GPU is integrated very differently from one SoC to +another. In order to accommodate those differences, you have the option +to specify one more vendor-specific compatible, among: + +- "amlogic,meson-g12a-mali" + Required properties: + - resets : Should contain phandles of : + + GPU reset line + + GPU APB glue reset line + +Example for a Mali-G31: + +gpu@ffa30000 { + compatible = "amlogic,meson-g12a-mali", "arm,mali-bifrost"; + reg = <0xffe40000 0x10000>; + interrupts = , + , + ; + interrupt-names = "job", "mmu", "gpu"; + clocks = <&clk CLKID_MALI>; + mali-supply = <&vdd_gpu>; + operating-points-v2 = <&gpu_opp_table>; + resets = <&reset RESET_DVALIN_CAPB3>, <&reset RESET_DVALIN>; +}; + +gpu_opp_table: opp_table0 { + compatible = "operating-points-v2"; + + opp@533000000 { + opp-hz = /bits/ 64 <533000000>; + opp-microvolt = <1250000>; + }; + opp@450000000 { + opp-hz = /bits/ 64 <450000000>; + opp-microvolt = <1150000>; + }; + opp@400000000 { + opp-hz = /bits/ 64 <400000000>; + opp-microvolt = <1125000>; + }; + opp@350000000 { + opp-hz = /bits/ 64 <350000000>; + opp-microvolt = <1075000>; + }; + opp@266000000 { + opp-hz = /bits/ 64 <266000000>; + opp-microvolt = <1025000>; + }; + opp@160000000 { + opp-hz = /bits/ 64 <160000000>; + opp-microvolt = <925000>; + }; + opp@100000000 { + opp-hz = /bits/ 64 <100000000>; + opp-microvolt = <912500>; + }; +}; diff --git a/Documentation/devicetree/bindings/gpu/aspeed-gfx.txt b/Documentation/devicetree/bindings/gpu/aspeed-gfx.txt new file mode 100644 index 000000000000..958bdf962339 --- /dev/null +++ b/Documentation/devicetree/bindings/gpu/aspeed-gfx.txt @@ -0,0 +1,41 @@ +Device tree configuration for the GFX display device on the ASPEED SoCs + +Required properties: + - compatible + * Must be one of the following: + + aspeed,ast2500-gfx + + aspeed,ast2400-gfx + * In addition, the ASPEED pinctrl bindings require the 'syscon' property to + be present + + - reg: Physical base address and length of the GFX registers + + - interrupts: interrupt number for the GFX device + + - clocks: clock number used to generate the pixel clock + + - resets: reset line that must be released to use the GFX device + + - memory-region: + Phandle to a memory region to allocate from, as defined in + Documentation/devicetree/bindings/reserved-memory/reserved-memory.txt + + +Example: + +gfx: display@1e6e6000 { + compatible = "aspeed,ast2500-gfx", "syscon"; + reg = <0x1e6e6000 0x1000>; + reg-io-width = <4>; + clocks = <&syscon ASPEED_CLK_GATE_D1CLK>; + resets = <&syscon ASPEED_RESET_CRT1>; + interrupts = <0x19>; + memory-region = <&gfx_memory>; +}; + +gfx_memory: framebuffer { + size = <0x01000000>; + alignment = <0x01000000>; + compatible = "shared-dma-pool"; + reusable; +}; diff --git a/Documentation/devicetree/bindings/hwmon/adc128d818.txt b/Documentation/devicetree/bindings/hwmon/adc128d818.txt index 08bab0e94d25..d0ae46d7bac3 100644 --- a/Documentation/devicetree/bindings/hwmon/adc128d818.txt +++ b/Documentation/devicetree/bindings/hwmon/adc128d818.txt @@ -26,7 +26,7 @@ Required node properties: Optional node properties: - - ti,mode: Operation mode (see above). + - ti,mode: Operation mode (u8) (see above). Example (operation mode 2): @@ -34,5 +34,5 @@ Example (operation mode 2): adc128d818@1d { compatible = "ti,adc128d818"; reg = <0x1d>; - ti,mode = <2>; + ti,mode = /bits/ 8 <2>; }; diff --git a/Documentation/devicetree/bindings/i2c/i2c-xscale.txt b/Documentation/devicetree/bindings/i2c/i2c-iop3xx.txt similarity index 100% rename from Documentation/devicetree/bindings/i2c/i2c-xscale.txt rename to Documentation/devicetree/bindings/i2c/i2c-iop3xx.txt diff --git a/Documentation/devicetree/bindings/i2c/i2c-mtk.txt b/Documentation/devicetree/bindings/i2c/i2c-mt65xx.txt similarity index 100% rename from Documentation/devicetree/bindings/i2c/i2c-mtk.txt rename to Documentation/devicetree/bindings/i2c/i2c-mt65xx.txt diff --git a/Documentation/devicetree/bindings/i2c/i2c-st-ddci2c.txt b/Documentation/devicetree/bindings/i2c/i2c-stu300.txt similarity index 100% rename from Documentation/devicetree/bindings/i2c/i2c-st-ddci2c.txt rename to Documentation/devicetree/bindings/i2c/i2c-stu300.txt diff --git a/Documentation/devicetree/bindings/i2c/i2c-sunxi-p2wi.txt b/Documentation/devicetree/bindings/i2c/i2c-sun6i-p2wi.txt similarity index 100% rename from Documentation/devicetree/bindings/i2c/i2c-sunxi-p2wi.txt rename to Documentation/devicetree/bindings/i2c/i2c-sun6i-p2wi.txt diff --git a/Documentation/devicetree/bindings/i2c/i2c-vt8500.txt b/Documentation/devicetree/bindings/i2c/i2c-wmt.txt similarity index 100% rename from Documentation/devicetree/bindings/i2c/i2c-vt8500.txt rename to Documentation/devicetree/bindings/i2c/i2c-wmt.txt diff --git a/Documentation/devicetree/bindings/net/dsa/qca8k.txt b/Documentation/devicetree/bindings/net/dsa/qca8k.txt index bbcb255c3150..93a7469e70d4 100644 --- a/Documentation/devicetree/bindings/net/dsa/qca8k.txt +++ b/Documentation/devicetree/bindings/net/dsa/qca8k.txt @@ -12,10 +12,15 @@ Required properties: Subnodes: The integrated switch subnode should be specified according to the binding -described in dsa/dsa.txt. As the QCA8K switches do not have a N:N mapping of -port and PHY id, each subnode describing a port needs to have a valid phandle -referencing the internal PHY connected to it. The CPU port of this switch is -always port 0. +described in dsa/dsa.txt. If the QCA8K switch is connect to a SoC's external +mdio-bus each subnode describing a port needs to have a valid phandle +referencing the internal PHY it is connected to. This is because there's no +N:N mapping of port and PHY id. + +Don't use mixed external and internal mdio-bus configurations, as this is +not supported by the hardware. + +The CPU port of this switch is always port 0. A CPU port node has the following optional node: @@ -31,8 +36,9 @@ For QCA8K the 'fixed-link' sub-node supports only the following properties: - 'full-duplex' (boolean, optional), to indicate that full duplex is used. When absent, half duplex is assumed. -Example: +Examples: +for the external mdio-bus configuration: &mdio0 { phy_port1: phy@0 { @@ -55,12 +61,12 @@ Example: reg = <4>; }; - switch0@0 { + switch@10 { compatible = "qca,qca8337"; #address-cells = <1>; #size-cells = <0>; - reg = <0>; + reg = <0x10>; ports { #address-cells = <1>; @@ -108,3 +114,56 @@ Example: }; }; }; + +for the internal master mdio-bus configuration: + + &mdio0 { + switch@10 { + compatible = "qca,qca8337"; + #address-cells = <1>; + #size-cells = <0>; + + reg = <0x10>; + + ports { + #address-cells = <1>; + #size-cells = <0>; + + port@0 { + reg = <0>; + label = "cpu"; + ethernet = <&gmac1>; + phy-mode = "rgmii"; + fixed-link { + speed = 1000; + full-duplex; + }; + }; + + port@1 { + reg = <1>; + label = "lan1"; + }; + + port@2 { + reg = <2>; + label = "lan2"; + }; + + port@3 { + reg = <3>; + label = "lan3"; + }; + + port@4 { + reg = <4>; + label = "lan4"; + }; + + port@5 { + reg = <5>; + label = "wan"; + }; + }; + }; + }; diff --git a/Documentation/devicetree/bindings/serial/mtk-uart.txt b/Documentation/devicetree/bindings/serial/mtk-uart.txt index 742cb470595b..bcfb13194f16 100644 --- a/Documentation/devicetree/bindings/serial/mtk-uart.txt +++ b/Documentation/devicetree/bindings/serial/mtk-uart.txt @@ -16,6 +16,7 @@ Required properties: * "mediatek,mt8127-uart" for MT8127 compatible UARTS * "mediatek,mt8135-uart" for MT8135 compatible UARTS * "mediatek,mt8173-uart" for MT8173 compatible UARTS + * "mediatek,mt8183-uart", "mediatek,mt6577-uart" for MT8183 compatible UARTS * "mediatek,mt6577-uart" for MT6577 and all of the above - reg: The base address of the UART register bank. diff --git a/Documentation/devicetree/bindings/vendor-prefixes.txt b/Documentation/devicetree/bindings/vendor-prefixes.txt index 5e6ea1f991d8..9506140167d6 100644 --- a/Documentation/devicetree/bindings/vendor-prefixes.txt +++ b/Documentation/devicetree/bindings/vendor-prefixes.txt @@ -345,6 +345,7 @@ ricoh Ricoh Co. Ltd. rikomagic Rikomagic Tech Corp. Ltd riscv RISC-V Foundation rockchip Fuzhou Rockchip Electronics Co., Ltd +rocktech ROCKTECH DISPLAYS LIMITED rohm ROHM Semiconductor Co., Ltd ronbo Ronbo Electronics roofull Shenzhen Roofull Technology Co, Ltd diff --git a/Documentation/filesystems/mount_api.txt b/Documentation/filesystems/mount_api.txt index 944d1965e917..00ff0cfccfa7 100644 --- a/Documentation/filesystems/mount_api.txt +++ b/Documentation/filesystems/mount_api.txt @@ -12,11 +12,13 @@ CONTENTS (4) Filesystem context security. - (5) VFS filesystem context operations. + (5) VFS filesystem context API. - (6) Parameter description. + (6) Superblock creation helpers. - (7) Parameter helper functions. + (7) Parameter description. + + (8) Parameter helper functions. ======== @@ -41,12 +43,15 @@ The creation of new mounts is now to be done in a multistep process: (7) Destroy the context. -To support this, the file_system_type struct gains a new field: +To support this, the file_system_type struct gains two new fields: int (*init_fs_context)(struct fs_context *fc); + const struct fs_parameter_description *parameters; -which is invoked to set up the filesystem-specific parts of a filesystem -context, including the additional space. +The first is invoked to set up the filesystem-specific parts of a filesystem +context, including the additional space, and the second points to the +parameter description for validation at registration time and querying by a +future system call. Note that security initialisation is done *after* the filesystem is called so that the namespaces may be adjusted first. @@ -73,9 +78,9 @@ context. This is represented by the fs_context structure: void *s_fs_info; unsigned int sb_flags; unsigned int sb_flags_mask; + unsigned int s_iflags; + unsigned int lsm_flags; enum fs_context_purpose purpose:8; - bool sloppy:1; - bool silent:1; ... }; @@ -141,6 +146,10 @@ The fs_context fields are as follows: Which bits SB_* flags are to be set/cleared in super_block::s_flags. + (*) unsigned int s_iflags + + These will be bitwise-OR'd with s->s_iflags when a superblock is created. + (*) enum fs_context_purpose This indicates the purpose for which the context is intended. The @@ -150,17 +159,6 @@ The fs_context fields are as follows: FS_CONTEXT_FOR_SUBMOUNT -- New automatic submount of extant mount FS_CONTEXT_FOR_RECONFIGURE -- Change an existing mount - (*) bool sloppy - (*) bool silent - - These are set if the sloppy or silent mount options are given. - - [NOTE] sloppy is probably unnecessary when userspace passes over one - option at a time since the error can just be ignored if userspace deems it - to be unimportant. - - [NOTE] silent is probably redundant with sb_flags & SB_SILENT. - The mount context is created by calling vfs_new_fs_context() or vfs_dup_fs_context() and is destroyed with put_fs_context(). Note that the structure is not refcounted. @@ -342,28 +340,47 @@ number of operations used by the new mount code for this purpose: It should return 0 on success or a negative error code on failure. -================================= -VFS FILESYSTEM CONTEXT OPERATIONS -================================= +========================== +VFS FILESYSTEM CONTEXT API +========================== -There are four operations for creating a filesystem context and -one for destroying a context: +There are four operations for creating a filesystem context and one for +destroying a context: - (*) struct fs_context *vfs_new_fs_context(struct file_system_type *fs_type, - struct dentry *reference, - unsigned int sb_flags, - unsigned int sb_flags_mask, - enum fs_context_purpose purpose); + (*) struct fs_context *fs_context_for_mount( + struct file_system_type *fs_type, + unsigned int sb_flags); - Create a filesystem context for a given filesystem type and purpose. This - allocates the filesystem context, sets the superblock flags, initialises - the security and calls fs_type->init_fs_context() to initialise the - filesystem private data. + Allocate a filesystem context for the purpose of setting up a new mount, + whether that be with a new superblock or sharing an existing one. This + sets the superblock flags, initialises the security and calls + fs_type->init_fs_context() to initialise the filesystem private data. - reference can be NULL or it may indicate the root dentry of a superblock - that is going to be reconfigured (FS_CONTEXT_FOR_RECONFIGURE) or - the automount point that triggered a submount (FS_CONTEXT_FOR_SUBMOUNT). - This is provided as a source of namespace information. + fs_type specifies the filesystem type that will manage the context and + sb_flags presets the superblock flags stored therein. + + (*) struct fs_context *fs_context_for_reconfigure( + struct dentry *dentry, + unsigned int sb_flags, + unsigned int sb_flags_mask); + + Allocate a filesystem context for the purpose of reconfiguring an + existing superblock. dentry provides a reference to the superblock to be + configured. sb_flags and sb_flags_mask indicate which superblock flags + need changing and to what. + + (*) struct fs_context *fs_context_for_submount( + struct file_system_type *fs_type, + struct dentry *reference); + + Allocate a filesystem context for the purpose of creating a new mount for + an automount point or other derived superblock. fs_type specifies the + filesystem type that will manage the context and the reference dentry + supplies the parameters. Namespaces are propagated from the reference + dentry's superblock also. + + Note that it's not a requirement that the reference dentry be of the same + filesystem type as fs_type. (*) struct fs_context *vfs_dup_fs_context(struct fs_context *src_fc); @@ -390,20 +407,6 @@ context pointer or a negative error code. For the remaining operations, if an error occurs, a negative error code will be returned. - (*) int vfs_get_tree(struct fs_context *fc); - - Get or create the mountable root and superblock, using the parameters in - the filesystem context to select/configure the superblock. This invokes - the ->validate() op and then the ->get_tree() op. - - [NOTE] ->validate() could perhaps be rolled into ->get_tree() and - ->reconfigure(). - - (*) struct vfsmount *vfs_create_mount(struct fs_context *fc); - - Create a mount given the parameters in the specified filesystem context. - Note that this does not attach the mount to anything. - (*) int vfs_parse_fs_param(struct fs_context *fc, struct fs_parameter *param); @@ -432,17 +435,80 @@ returned. clear the pointer, but then becomes responsible for disposing of the object. - (*) int vfs_parse_fs_string(struct fs_context *fc, char *key, + (*) int vfs_parse_fs_string(struct fs_context *fc, const char *key, const char *value, size_t v_size); - A wrapper around vfs_parse_fs_param() that just passes a constant string. + A wrapper around vfs_parse_fs_param() that copies the value string it is + passed. (*) int generic_parse_monolithic(struct fs_context *fc, void *data); Parse a sys_mount() data page, assuming the form to be a text list consisting of key[=val] options separated by commas. Each item in the list is passed to vfs_mount_option(). This is the default when the - ->parse_monolithic() operation is NULL. + ->parse_monolithic() method is NULL. + + (*) int vfs_get_tree(struct fs_context *fc); + + Get or create the mountable root and superblock, using the parameters in + the filesystem context to select/configure the superblock. This invokes + the ->get_tree() method. + + (*) struct vfsmount *vfs_create_mount(struct fs_context *fc); + + Create a mount given the parameters in the specified filesystem context. + Note that this does not attach the mount to anything. + + +=========================== +SUPERBLOCK CREATION HELPERS +=========================== + +A number of VFS helpers are available for use by filesystems for the creation +or looking up of superblocks. + + (*) struct super_block * + sget_fc(struct fs_context *fc, + int (*test)(struct super_block *sb, struct fs_context *fc), + int (*set)(struct super_block *sb, struct fs_context *fc)); + + This is the core routine. If test is non-NULL, it searches for an + existing superblock matching the criteria held in the fs_context, using + the test function to match them. If no match is found, a new superblock + is created and the set function is called to set it up. + + Prior to the set function being called, fc->s_fs_info will be transferred + to sb->s_fs_info - and fc->s_fs_info will be cleared if set returns + success (ie. 0). + +The following helpers all wrap sget_fc(): + + (*) int vfs_get_super(struct fs_context *fc, + enum vfs_get_super_keying keying, + int (*fill_super)(struct super_block *sb, + struct fs_context *fc)) + + This creates/looks up a deviceless superblock. The keying indicates how + many superblocks of this type may exist and in what manner they may be + shared: + + (1) vfs_get_single_super + + Only one such superblock may exist in the system. Any further + attempt to get a new superblock gets this one (and any parameter + differences are ignored). + + (2) vfs_get_keyed_super + + Multiple superblocks of this type may exist and they're keyed on + their s_fs_info pointer (for example this may refer to a + namespace). + + (3) vfs_get_independent_super + + Multiple independent superblocks of this type may exist. This + function never matches an existing one and always creates a new + one. ===================== @@ -454,35 +520,22 @@ There's a core description struct that links everything together: struct fs_parameter_description { const char name[16]; - u8 nr_params; - u8 nr_alt_keys; - u8 nr_enums; - bool ignore_unknown; - bool no_source; - const char *const *keys; - const struct constant_table *alt_keys; const struct fs_parameter_spec *specs; const struct fs_parameter_enum *enums; }; For example: - enum afs_param { + enum { Opt_autocell, Opt_bar, Opt_dyn, Opt_foo, Opt_source, - nr__afs_params }; static const struct fs_parameter_description afs_fs_parameters = { .name = "kAFS", - .nr_params = nr__afs_params, - .nr_alt_keys = ARRAY_SIZE(afs_param_alt_keys), - .nr_enums = ARRAY_SIZE(afs_param_enums), - .keys = afs_param_keys, - .alt_keys = afs_param_alt_keys, .specs = afs_param_specs, .enums = afs_param_enums, }; @@ -494,28 +547,24 @@ The members are as follows: The name to be used in error messages generated by the parse helper functions. - (2) u8 nr_params; + (2) const struct fs_parameter_specification *specs; - The number of discrete parameter identifiers. This indicates the number - of elements in the ->types[] array and also limits the values that may be - used in the values that the ->keys[] array maps to. + Table of parameter specifications, terminated with a null entry, where the + entries are of type: - It is expected that, for example, two parameters that are related, say - "acl" and "noacl" with have the same ID, but will be flagged to indicate - that one is the inverse of the other. The value can then be picked out - from the parse result. - - (3) const struct fs_parameter_specification *specs; - - Table of parameter specifications, where the entries are of type: - - struct fs_parameter_type { - enum fs_parameter_spec type:8; - u8 flags; + struct fs_parameter_spec { + const char *name; + u8 opt; + enum fs_parameter_type type:8; + unsigned short flags; }; - and the parameter identifier is the index to the array. 'type' indicates - the desired value type and must be one of: + The 'name' field is a string to match exactly to the parameter key (no + wildcards, patterns and no case-independence) and 'opt' is the value that + will be returned by the fs_parser() function in the case of a successful + match. + + The 'type' field indicates the desired value type and must be one of: TYPE NAME EXPECTED VALUE RESULT IN ======================= ======================= ===================== @@ -525,85 +574,65 @@ The members are as follows: fs_param_is_u32_octal 32-bit octal int result->uint_32 fs_param_is_u32_hex 32-bit hex int result->uint_32 fs_param_is_s32 32-bit signed int result->int_32 + fs_param_is_u64 64-bit unsigned int result->uint_64 fs_param_is_enum Enum value name result->uint_32 fs_param_is_string Arbitrary string param->string fs_param_is_blob Binary blob param->blob fs_param_is_blockdev Blockdev path * Needs lookup fs_param_is_path Path * Needs lookup - fs_param_is_fd File descriptor param->file - - And each parameter can be qualified with 'flags': - - fs_param_v_optional The value is optional - fs_param_neg_with_no If key name is prefixed with "no", it is false - fs_param_neg_with_empty If value is "", it is false - fs_param_deprecated The parameter is deprecated. - - For example: - - static const struct fs_parameter_spec afs_param_specs[nr__afs_params] = { - [Opt_autocell] = { fs_param_is flag }, - [Opt_bar] = { fs_param_is_enum }, - [Opt_dyn] = { fs_param_is flag }, - [Opt_foo] = { fs_param_is_bool, fs_param_neg_with_no }, - [Opt_source] = { fs_param_is_string }, - }; + fs_param_is_fd File descriptor result->int_32 Note that if the value is of fs_param_is_bool type, fs_parse() will try to match any string value against "0", "1", "no", "yes", "false", "true". - [!] NOTE that the table must be sorted according to primary key name so - that ->keys[] is also sorted. + Each parameter can also be qualified with 'flags': - (4) const char *const *keys; + fs_param_v_optional The value is optional + fs_param_neg_with_no result->negated set if key is prefixed with "no" + fs_param_neg_with_empty result->negated set if value is "" + fs_param_deprecated The parameter is deprecated. - Table of primary key names for the parameters. There must be one entry - per defined parameter. The table is optional if ->nr_params is 0. The - table is just an array of names e.g.: + These are wrapped with a number of convenience wrappers: - static const char *const afs_param_keys[nr__afs_params] = { - [Opt_autocell] = "autocell", - [Opt_bar] = "bar", - [Opt_dyn] = "dyn", - [Opt_foo] = "foo", - [Opt_source] = "source", + MACRO SPECIFIES + ======================= =============================================== + fsparam_flag() fs_param_is_flag + fsparam_flag_no() fs_param_is_flag, fs_param_neg_with_no + fsparam_bool() fs_param_is_bool + fsparam_u32() fs_param_is_u32 + fsparam_u32oct() fs_param_is_u32_octal + fsparam_u32hex() fs_param_is_u32_hex + fsparam_s32() fs_param_is_s32 + fsparam_u64() fs_param_is_u64 + fsparam_enum() fs_param_is_enum + fsparam_string() fs_param_is_string + fsparam_blob() fs_param_is_blob + fsparam_bdev() fs_param_is_blockdev + fsparam_path() fs_param_is_path + fsparam_fd() fs_param_is_fd + + all of which take two arguments, name string and option number - for + example: + + static const struct fs_parameter_spec afs_param_specs[] = { + fsparam_flag ("autocell", Opt_autocell), + fsparam_flag ("dyn", Opt_dyn), + fsparam_string ("source", Opt_source), + fsparam_flag_no ("foo", Opt_foo), + {} }; - [!] NOTE that the table must be sorted such that the table can be searched - with bsearch() using strcmp(). This means that the Opt_* values must - correspond to the entries in this table. - - (5) const struct constant_table *alt_keys; - u8 nr_alt_keys; - - Table of additional key names and their mappings to parameter ID plus the - number of elements in the table. This is optional. The table is just an - array of { name, integer } pairs, e.g.: - - static const struct constant_table afs_param_keys[] = { - { "baz", Opt_bar }, - { "dynamic", Opt_dyn }, - }; - - [!] NOTE that the table must be sorted such that strcmp() can be used with - bsearch() to search the entries. - - The parameter ID can also be fs_param_key_removed to indicate that a - deprecated parameter has been removed and that an error will be given. - This differs from fs_param_deprecated where the parameter may still have - an effect. - - Further, the behaviour of the parameter may differ when an alternate name - is used (for instance with NFS, "v3", "v4.2", etc. are alternate names). + An addition macro, __fsparam() is provided that takes an additional pair + of arguments to specify the type and the flags for anything that doesn't + match one of the above macros. (6) const struct fs_parameter_enum *enums; - u8 nr_enums; - Table of enum value names to integer mappings and the number of elements - stored therein. This is of type: + Table of enum value names to integer mappings, terminated with a null + entry. This is of type: struct fs_parameter_enum { - u8 param_id; + u8 opt; char name[14]; u8 value; }; @@ -621,11 +650,6 @@ The members are as follows: try to look the value up in the enum table and the result will be stored in the parse result. - (7) bool no_source; - - If this is set, fs_parse() will ignore any "source" parameter and not - pass it to the filesystem. - The parser should be pointed to by the parser pointer in the file_system_type struct as this will provide validation on registration (if CONFIG_VALIDATE_FS_PARSER=y) and will allow the description to be queried from @@ -650,9 +674,8 @@ process the parameters it is given. int value; }; - and it must be sorted such that it can be searched using bsearch() using - strcmp(). If a match is found, the corresponding value is returned. If a - match isn't found, the not_found value is returned instead. + If a match is found, the corresponding value is returned. If a match + isn't found, the not_found value is returned instead. (*) bool validate_constant_table(const struct constant_table *tbl, size_t tbl_size, @@ -665,36 +688,36 @@ process the parameters it is given. should just be set to lie inside the low-to-high range. If all is good, true is returned. If the table is invalid, errors are - logged to dmesg, the stack is dumped and false is returned. + logged to dmesg and false is returned. + + (*) bool fs_validate_description(const struct fs_parameter_description *desc); + + This performs some validation checks on a parameter description. It + returns true if the description is good and false if it is not. It will + log errors to dmesg if validation fails. (*) int fs_parse(struct fs_context *fc, - const struct fs_param_parser *parser, + const struct fs_parameter_description *desc, struct fs_parameter *param, - struct fs_param_parse_result *result); + struct fs_parse_result *result); This is the main interpreter of parameters. It uses the parameter - description (parser) to look up the name of the parameter to use and to - convert that to a parameter ID (stored in result->key). + description to look up a parameter by key name and to convert that to an + option number (which it returns). If successful, and if the parameter type indicates the result is a boolean, integer or enum type, the value is converted by this function and - the result stored in result->{boolean,int_32,uint_32}. + the result stored in result->{boolean,int_32,uint_32,uint_64}. If a match isn't initially made, the key is prefixed with "no" and no value is present then an attempt will be made to look up the key with the prefix removed. If this matches a parameter for which the type has flag - fs_param_neg_with_no set, then a match will be made and the value will be - set to false/0/NULL. + fs_param_neg_with_no set, then a match will be made and result->negated + will be set to true. - If the parameter is successfully matched and, optionally, parsed - correctly, 1 is returned. If the parameter isn't matched and - parser->ignore_unknown is set, then 0 is returned. Otherwise -EINVAL is - returned. - - (*) bool fs_validate_description(const struct fs_parameter_description *desc); - - This is validates the parameter description. It returns true if the - description is good and false if it is not. + If the parameter isn't matched, -ENOPARAM will be returned; if the + parameter is matched, but the value is erroneous, -EINVAL will be + returned; otherwise the parameter's option number will be returned. (*) int fs_lookup_param(struct fs_context *fc, struct fs_parameter *value, diff --git a/Documentation/gpu/drm-kms-helpers.rst b/Documentation/gpu/drm-kms-helpers.rst index 58b375e47615..14102ae035dc 100644 --- a/Documentation/gpu/drm-kms-helpers.rst +++ b/Documentation/gpu/drm-kms-helpers.rst @@ -107,6 +107,12 @@ fbdev Helper Functions Reference .. kernel-doc:: drivers/gpu/drm/drm_fb_helper.c :export: +format Helper Functions Reference +================================= + +.. kernel-doc:: drivers/gpu/drm/drm_format_helper.c + :export: + Framebuffer CMA Helper Functions Reference ========================================== diff --git a/Documentation/gpu/kms-properties.csv b/Documentation/gpu/kms-properties.csv index bfde04eddd14..07ed22ea3bd6 100644 --- a/Documentation/gpu/kms-properties.csv +++ b/Documentation/gpu/kms-properties.csv @@ -17,7 +17,6 @@ Owner Module/Drivers,Group,Property Name,Type,Property Values,Object attached,De ,Virtual GPU,“suggested X”,RANGE,"Min=0, Max=0xffffffff",Connector,property to suggest an X offset for a connector ,,“suggested Y”,RANGE,"Min=0, Max=0xffffffff",Connector,property to suggest an Y offset for a connector ,Optional,"""aspect ratio""",ENUM,"{ ""None"", ""4:3"", ""16:9"" }",Connector,TDB -,Optional,"""content type""",ENUM,"{ ""No Data"", ""Graphics"", ""Photo"", ""Cinema"", ""Game"" }",Connector,TBD i915,Generic,"""Broadcast RGB""",ENUM,"{ ""Automatic"", ""Full"", ""Limited 16:235"" }",Connector,"When this property is set to Limited 16:235 and CTM is set, the hardware will be programmed with the result of the multiplication of CTM by the limited range matrix to ensure the pixels normaly in the range 0..1.0 are remapped to the range 16/255..235/255." ,,“audio”,ENUM,"{ ""force-dvi"", ""off"", ""auto"", ""on"" }",Connector,TBD ,SDVO-TV,“mode”,ENUM,"{ ""NTSC_M"", ""NTSC_J"", ""NTSC_443"", ""PAL_B"" } etc.",Connector,TBD diff --git a/Documentation/gpu/meson.rst b/Documentation/gpu/meson.rst index 479f6f51a13b..b9e2f9aa3bd8 100644 --- a/Documentation/gpu/meson.rst +++ b/Documentation/gpu/meson.rst @@ -42,12 +42,6 @@ Video Encoder .. kernel-doc:: drivers/gpu/drm/meson/meson_venc.c :doc: Video Encoder -Video Canvas Management -======================= - -.. kernel-doc:: drivers/gpu/drm/meson/meson_canvas.c - :doc: Canvas - Video Clocks ============ diff --git a/Documentation/i2c/busses/i2c-i801 b/Documentation/i2c/busses/i2c-i801 index d1ee484a787d..ee9984f35868 100644 --- a/Documentation/i2c/busses/i2c-i801 +++ b/Documentation/i2c/busses/i2c-i801 @@ -36,6 +36,7 @@ Supported adapters: * Intel Cannon Lake (PCH) * Intel Cedar Fork (PCH) * Intel Ice Lake (PCH) + * Intel Comet Lake (PCH) Datasheets: Publicly available at the Intel website On Intel Patsburg and later chipsets, both the normal host SMBus controller diff --git a/Documentation/lzo.txt b/Documentation/lzo.txt index f79934225d8d..ca983328976b 100644 --- a/Documentation/lzo.txt +++ b/Documentation/lzo.txt @@ -102,9 +102,11 @@ Byte sequences dictionary which is empty, and that it will always be invalid at this place. - 17 : bitstream version. If the first byte is 17, the next byte - gives the bitstream version (version 1 only). If the first byte - is not 17, the bitstream version is 0. + 17 : bitstream version. If the first byte is 17, and compressed + stream length is at least 5 bytes (length of shortest possible + versioned bitstream), the next byte gives the bitstream version + (version 1 only). + Otherwise, the bitstream version is 0. 18..21 : copy 0..3 literals state = (byte - 17) = 0..3 [ copy literals ] diff --git a/Documentation/networking/bpf_flow_dissector.rst b/Documentation/networking/bpf_flow_dissector.rst new file mode 100644 index 000000000000..b375ae2ec2c4 --- /dev/null +++ b/Documentation/networking/bpf_flow_dissector.rst @@ -0,0 +1,126 @@ +.. SPDX-License-Identifier: GPL-2.0 + +================== +BPF Flow Dissector +================== + +Overview +======== + +Flow dissector is a routine that parses metadata out of the packets. It's +used in the various places in the networking subsystem (RFS, flow hash, etc). + +BPF flow dissector is an attempt to reimplement C-based flow dissector logic +in BPF to gain all the benefits of BPF verifier (namely, limits on the +number of instructions and tail calls). + +API +=== + +BPF flow dissector programs operate on an ``__sk_buff``. However, only the +limited set of fields is allowed: ``data``, ``data_end`` and ``flow_keys``. +``flow_keys`` is ``struct bpf_flow_keys`` and contains flow dissector input +and output arguments. + +The inputs are: + * ``nhoff`` - initial offset of the networking header + * ``thoff`` - initial offset of the transport header, initialized to nhoff + * ``n_proto`` - L3 protocol type, parsed out of L2 header + +Flow dissector BPF program should fill out the rest of the ``struct +bpf_flow_keys`` fields. Input arguments ``nhoff/thoff/n_proto`` should be +also adjusted accordingly. + +The return code of the BPF program is either BPF_OK to indicate successful +dissection, or BPF_DROP to indicate parsing error. + +__sk_buff->data +=============== + +In the VLAN-less case, this is what the initial state of the BPF flow +dissector looks like:: + + +------+------+------------+-----------+ + | DMAC | SMAC | ETHER_TYPE | L3_HEADER | + +------+------+------------+-----------+ + ^ + | + +-- flow dissector starts here + + +.. code:: c + + skb->data + flow_keys->nhoff point to the first byte of L3_HEADER + flow_keys->thoff = nhoff + flow_keys->n_proto = ETHER_TYPE + +In case of VLAN, flow dissector can be called with the two different states. + +Pre-VLAN parsing:: + + +------+------+------+-----+-----------+-----------+ + | DMAC | SMAC | TPID | TCI |ETHER_TYPE | L3_HEADER | + +------+------+------+-----+-----------+-----------+ + ^ + | + +-- flow dissector starts here + +.. code:: c + + skb->data + flow_keys->nhoff point the to first byte of TCI + flow_keys->thoff = nhoff + flow_keys->n_proto = TPID + +Please note that TPID can be 802.1AD and, hence, BPF program would +have to parse VLAN information twice for double tagged packets. + + +Post-VLAN parsing:: + + +------+------+------+-----+-----------+-----------+ + | DMAC | SMAC | TPID | TCI |ETHER_TYPE | L3_HEADER | + +------+------+------+-----+-----------+-----------+ + ^ + | + +-- flow dissector starts here + +.. code:: c + + skb->data + flow_keys->nhoff point the to first byte of L3_HEADER + flow_keys->thoff = nhoff + flow_keys->n_proto = ETHER_TYPE + +In this case VLAN information has been processed before the flow dissector +and BPF flow dissector is not required to handle it. + + +The takeaway here is as follows: BPF flow dissector program can be called with +the optional VLAN header and should gracefully handle both cases: when single +or double VLAN is present and when it is not present. The same program +can be called for both cases and would have to be written carefully to +handle both cases. + + +Reference Implementation +======================== + +See ``tools/testing/selftests/bpf/progs/bpf_flow.c`` for the reference +implementation and ``tools/testing/selftests/bpf/flow_dissector_load.[hc]`` +for the loader. bpftool can be used to load BPF flow dissector program as well. + +The reference implementation is organized as follows: + * ``jmp_table`` map that contains sub-programs for each supported L3 protocol + * ``_dissect`` routine - entry point; it does input ``n_proto`` parsing and + does ``bpf_tail_call`` to the appropriate L3 handler + +Since BPF at this point doesn't support looping (or any jumping back), +jmp_table is used instead to handle multiple levels of encapsulation (and +IPv6 options). + + +Current Limitations +=================== +BPF flow dissector doesn't support exporting all the metadata that in-kernel +C-based implementation can export. Notable example is single VLAN (802.1Q) +and double VLAN (802.1AD) tags. Please refer to the ``struct bpf_flow_keys`` +for a set of information that's currently can be exported from the BPF context. diff --git a/Documentation/networking/index.rst b/Documentation/networking/index.rst index 5449149be496..984e68f9e026 100644 --- a/Documentation/networking/index.rst +++ b/Documentation/networking/index.rst @@ -9,6 +9,7 @@ Contents: netdev-FAQ af_xdp batman-adv + bpf_flow_dissector can can_ucan_protocol device_drivers/freescale/dpaa2/index diff --git a/Documentation/networking/msg_zerocopy.rst b/Documentation/networking/msg_zerocopy.rst index 18c1415e7bfa..ace56204dd03 100644 --- a/Documentation/networking/msg_zerocopy.rst +++ b/Documentation/networking/msg_zerocopy.rst @@ -50,7 +50,7 @@ the excellent reporting over at LWN.net or read the original code. patchset [PATCH net-next v4 0/9] socket sendmsg MSG_ZEROCOPY - http://lkml.kernel.org/r/20170803202945.70750-1-willemdebruijn.kernel@gmail.com + https://lkml.kernel.org/netdev/20170803202945.70750-1-willemdebruijn.kernel@gmail.com Interface diff --git a/Documentation/networking/netdev-FAQ.rst b/Documentation/networking/netdev-FAQ.rst index 0ac5fa77f501..8c7a713cf657 100644 --- a/Documentation/networking/netdev-FAQ.rst +++ b/Documentation/networking/netdev-FAQ.rst @@ -131,6 +131,19 @@ it to the maintainer to figure out what is the most recent and current version that should be applied. If there is any doubt, the maintainer will reply and ask what should be done. +Q: I made changes to only a few patches in a patch series should I resend only those changed? +-------------------------------------------------------------------------------------------- +A: No, please resend the entire patch series and make sure you do number your +patches such that it is clear this is the latest and greatest set of patches +that can be applied. + +Q: I submitted multiple versions of a patch series and it looks like a version other than the last one has been accepted, what should I do? +------------------------------------------------------------------------------------------------------------------------------------------- +A: There is no revert possible, once it is pushed out, it stays like that. +Please send incremental versions on top of what has been merged in order to fix +the patches the way they would look like if your latest patch series was to be +merged. + Q: How can I tell what patches are queued up for backporting to the various stable releases? -------------------------------------------------------------------------------------------- A: Normally Greg Kroah-Hartman collects stable commits himself, but for diff --git a/Documentation/networking/nf_flowtable.txt b/Documentation/networking/nf_flowtable.txt index 54128c50d508..ca2136c76042 100644 --- a/Documentation/networking/nf_flowtable.txt +++ b/Documentation/networking/nf_flowtable.txt @@ -44,10 +44,10 @@ including the Netfilter hooks and the flowtable fastpath bypass. / \ / \ |Routing | / \ --> ingress ---> prerouting ---> |decision| | postrouting |--> neigh_xmit \_________/ \__________/ ---------- \____________/ ^ - | ^ | | ^ | - flowtable | | ____\/___ | | - | | | / \ | | - __\/___ | --------->| forward |------------ | + | ^ | ^ | + flowtable | ____\/___ | | + | | / \ | | + __\/___ | | forward |------------ | |-----| | \_________/ | |-----| | 'flow offload' rule | |-----| | adds entry to | diff --git a/Documentation/networking/snmp_counter.rst b/Documentation/networking/snmp_counter.rst index 52b026be028f..38a4edc4522b 100644 --- a/Documentation/networking/snmp_counter.rst +++ b/Documentation/networking/snmp_counter.rst @@ -413,7 +413,7 @@ algorithm. .. _F-RTO: https://tools.ietf.org/html/rfc5682 TCP Fast Path -============ +============= When kernel receives a TCP packet, it has two paths to handler the packet, one is fast path, another is slow path. The comment in kernel code provides a good explanation of them, I pasted them below:: @@ -681,6 +681,7 @@ The TCP stack receives an out of order duplicate packet, so it sends a DSACK to the sender. * TcpExtTCPDSACKRecv + The TCP stack receives a DSACK, which indicates an acknowledged duplicate packet is received. @@ -690,7 +691,7 @@ The TCP stack receives a DSACK, which indicate an out of order duplicate packet is received. invalid SACK and DSACK -==================== +====================== When a SACK (or DSACK) block is invalid, a corresponding counter would be updated. The validation method is base on the start/end sequence number of the SACK block. For more details, please refer the comment @@ -704,11 +705,13 @@ explaination: .. _Add counters for discarded SACK blocks: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=18f02545a9a16c9a89778b91a162ad16d510bb32 * TcpExtTCPSACKDiscard + This counter indicates how many SACK blocks are invalid. If the invalid SACK block is caused by ACK recording, the TCP stack will only ignore it and won't update this counter. * TcpExtTCPDSACKIgnoredOld and TcpExtTCPDSACKIgnoredNoUndo + When a DSACK block is invalid, one of these two counters would be updated. Which counter will be updated depends on the undo_marker flag of the TCP socket. If the undo_marker is not set, the TCP stack isn't @@ -719,7 +722,7 @@ will be updated. If the undo_marker is set, TcpExtTCPDSACKIgnoredOld will be updated. As implied in its name, it might be an old packet. SACK shift -========= +========== The linux networking stack stores data in sk_buff struct (skb for short). If a SACK block acrosses multiple skb, the TCP stack will try to re-arrange data in these skb. E.g. if a SACK block acknowledges seq @@ -730,12 +733,15 @@ seq 14 to 20. All data in skb2 will be moved to skb1, and skb2 will be discard, this operation is 'merge'. * TcpExtTCPSackShifted + A skb is shifted * TcpExtTCPSackMerged + A skb is merged * TcpExtTCPSackShiftFallback + A skb should be shifted or merged, but the TCP stack doesn't do it for some reasons. diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt index 7de9eee73fcd..67068c47c591 100644 --- a/Documentation/virtual/kvm/api.txt +++ b/Documentation/virtual/kvm/api.txt @@ -5,25 +5,32 @@ The Definitive KVM (Kernel-based Virtual Machine) API Documentation ---------------------- The kvm API is a set of ioctls that are issued to control various aspects -of a virtual machine. The ioctls belong to three classes +of a virtual machine. The ioctls belong to three classes: - System ioctls: These query and set global attributes which affect the whole kvm subsystem. In addition a system ioctl is used to create - virtual machines + virtual machines. - VM ioctls: These query and set attributes that affect an entire virtual machine, for example memory layout. In addition a VM ioctl is used to - create virtual cpus (vcpus). + create virtual cpus (vcpus) and devices. - Only run VM ioctls from the same process (address space) that was used - to create the VM. + VM ioctls must be issued from the same process (address space) that was + used to create the VM. - vcpu ioctls: These query and set attributes that control the operation of a single virtual cpu. - Only run vcpu ioctls from the same thread that was used to create the - vcpu. + vcpu ioctls should be issued from the same thread that was used to create + the vcpu, except for asynchronous vcpu ioctl that are marked as such in + the documentation. Otherwise, the first ioctl after switching threads + could see a performance impact. + - device ioctls: These query and set attributes that control the operation + of a single device. + + device ioctls must be issued from the same process (address space) that + was used to create the VM. 2. File descriptors ------------------- @@ -32,17 +39,34 @@ The kvm API is centered around file descriptors. An initial open("/dev/kvm") obtains a handle to the kvm subsystem; this handle can be used to issue system ioctls. A KVM_CREATE_VM ioctl on this handle will create a VM file descriptor which can be used to issue VM -ioctls. A KVM_CREATE_VCPU ioctl on a VM fd will create a virtual cpu -and return a file descriptor pointing to it. Finally, ioctls on a vcpu -fd can be used to control the vcpu, including the important task of -actually running guest code. +ioctls. A KVM_CREATE_VCPU or KVM_CREATE_DEVICE ioctl on a VM fd will +create a virtual cpu or device and return a file descriptor pointing to +the new resource. Finally, ioctls on a vcpu or device fd can be used +to control the vcpu or device. For vcpus, this includes the important +task of actually running guest code. In general file descriptors can be migrated among processes by means of fork() and the SCM_RIGHTS facility of unix domain socket. These kinds of tricks are explicitly not supported by kvm. While they will not cause harm to the host, their actual behavior is not guaranteed by -the API. The only supported use is one virtual machine per process, -and one vcpu per thread. +the API. See "General description" for details on the ioctl usage +model that is supported by KVM. + +It is important to note that althought VM ioctls may only be issued from +the process that created the VM, a VM's lifecycle is associated with its +file descriptor, not its creator (process). In other words, the VM and +its resources, *including the associated address space*, are not freed +until the last reference to the VM's file descriptor has been released. +For example, if fork() is issued after ioctl(KVM_CREATE_VM), the VM will +not be freed until both the parent (original) process and its child have +put their references to the VM's file descriptor. + +Because a VM's resources are not freed until the last reference to its +file descriptor is released, creating additional references to a VM via +via fork(), dup(), etc... without careful consideration is strongly +discouraged and may have unwanted side effects, e.g. memory allocated +by and on behalf of the VM's process may not be freed/unaccounted when +the VM is shut down. It is important to note that althought VM ioctls may only be issued from @@ -515,11 +539,15 @@ c) KVM_INTERRUPT_SET_LEVEL Note that any value for 'irq' other than the ones stated above is invalid and incurs unexpected behavior. +This is an asynchronous vcpu ioctl and can be invoked from any thread. + MIPS: Queues an external interrupt to be injected into the virtual CPU. A negative interrupt number dequeues the interrupt. +This is an asynchronous vcpu ioctl and can be invoked from any thread. + 4.17 KVM_DEBUG_GUEST @@ -1086,14 +1114,12 @@ struct kvm_userspace_memory_region { #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0) #define KVM_MEM_READONLY (1UL << 1) -This ioctl allows the user to create or modify a guest physical memory -slot. When changing an existing slot, it may be moved in the guest -physical memory space, or its flags may be modified. It may not be -resized. Slots may not overlap in guest physical address space. -Bits 0-15 of "slot" specifies the slot id and this value should be -less than the maximum number of user memory slots supported per VM. -The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS, -if this capability is supported by the architecture. +This ioctl allows the user to create, modify or delete a guest physical +memory slot. Bits 0-15 of "slot" specify the slot id and this value +should be less than the maximum number of user memory slots supported per +VM. The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS, +if this capability is supported by the architecture. Slots may not +overlap in guest physical address space. If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of "slot" specifies the address space which is being modified. They must be @@ -1102,6 +1128,10 @@ KVM_CAP_MULTI_ADDRESS_SPACE capability. Slots in separate address spaces are unrelated; the restriction on overlapping slots only applies within each address space. +Deleting a slot is done by passing zero for memory_size. When changing +an existing slot, it may be moved in the guest physical memory space, +or its flags may be modified, but it may not be resized. + Memory for the region is taken starting at the address denoted by the field userspace_addr, which must point at user addressable memory for the entire memory slot size. Any object may back this memory, including @@ -2493,7 +2523,7 @@ KVM_S390_MCHK (vm, vcpu) - machine check interrupt; cr 14 bits in parm, machine checks needing further payload are not supported by this ioctl) -Note that the vcpu ioctl is asynchronous to vcpu execution. +This is an asynchronous vcpu ioctl and can be invoked from any thread. 4.78 KVM_PPC_GET_HTAB_FD @@ -3042,8 +3072,7 @@ KVM_S390_INT_EMERGENCY - sigp emergency; parameters in .emerg KVM_S390_INT_EXTERNAL_CALL - sigp external call; parameters in .extcall KVM_S390_MCHK - machine check interrupt; parameters in .mchk - -Note that the vcpu ioctl is asynchronous to vcpu execution. +This is an asynchronous vcpu ioctl and can be invoked from any thread. 4.94 KVM_S390_GET_IRQ_STATE diff --git a/Documentation/virtual/kvm/mmu.txt b/Documentation/virtual/kvm/mmu.txt index f365102c80f5..2efe0efc516e 100644 --- a/Documentation/virtual/kvm/mmu.txt +++ b/Documentation/virtual/kvm/mmu.txt @@ -142,7 +142,7 @@ Shadow pages contain the following information: If clear, this page corresponds to a guest page table denoted by the gfn field. role.quadrant: - When role.cr4_pae=0, the guest uses 32-bit gptes while the host uses 64-bit + When role.gpte_is_8_bytes=0, the guest uses 32-bit gptes while the host uses 64-bit sptes. That means a guest page table contains more ptes than the host, so multiple shadow pages are needed to shadow one guest page. For first-level shadow pages, role.quadrant can be 0 or 1 and denotes the @@ -158,9 +158,9 @@ Shadow pages contain the following information: The page is invalid and should not be used. It is a root page that is currently pinned (by a cpu hardware register pointing to it); once it is unpinned it will be destroyed. - role.cr4_pae: - Contains the value of cr4.pae for which the page is valid (e.g. whether - 32-bit or 64-bit gptes are in use). + role.gpte_is_8_bytes: + Reflects the size of the guest PTE for which the page is valid, i.e. '1' + if 64-bit gptes are in use, '0' if 32-bit gptes are in use. role.nxe: Contains the value of efer.nxe for which the page is valid. role.cr0_wp: @@ -173,6 +173,9 @@ Shadow pages contain the following information: Contains the value of cr4.smap && !cr0.wp for which the page is valid (pages for which this is true are different from other pages; see the treatment of cr0.wp=0 below). + role.ept_sp: + This is a virtual flag to denote a shadowed nested EPT page. ept_sp + is true if "cr0_wp && smap_andnot_wp", an otherwise invalid combination. role.smm: Is 1 if the page is valid in system management mode. This field determines which of the kvm_memslots array was used to build this diff --git a/MAINTAINERS b/MAINTAINERS index dcf0b4f65fbe..e233b3c48546 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1167,7 +1167,7 @@ S: Supported T: git git://linux-arm.org/linux-ld.git for-upstream/mali-dp F: drivers/gpu/drm/arm/display/include/ F: drivers/gpu/drm/arm/display/komeda/ -F: Documentation/devicetree/bindings/display/arm/arm,komeda.txt +F: Documentation/devicetree/bindings/display/arm,komeda.txt F: Documentation/gpu/komeda-kms.rst ARM MALI-DP DRM DRIVER @@ -1180,6 +1180,15 @@ F: drivers/gpu/drm/arm/ F: Documentation/devicetree/bindings/display/arm,malidp.txt F: Documentation/gpu/afbc.rst +ARM MALI PANFROST DRM DRIVER +M: Rob Herring +M: Tomeu Vizoso +L: dri-devel@lists.freedesktop.org +S: Supported +T: git git://anongit.freedesktop.org/drm/drm-misc +F: drivers/gpu/drm/panfrost/ +F: include/uapi/drm/panfrost_drm.h + ARM MFM AND FLOPPY DRIVERS M: Ian Molton S: Maintained @@ -1893,14 +1902,15 @@ T: git git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-nomadik.git ARM/NUVOTON NPCM ARCHITECTURE M: Avi Fishman M: Tomer Maimon +M: Tali Perry R: Patrick Venture R: Nancy Yuen -R: Brendan Higgins +R: Benjamin Fair L: openbmc@lists.ozlabs.org (moderated for non-subscribers) S: Supported F: arch/arm/mach-npcm/ F: arch/arm/boot/dts/nuvoton-npcm* -F: include/dt-bindings/clock/nuvoton,npcm7xx-clks.h +F: include/dt-bindings/clock/nuvoton,npcm7xx-clock.h F: drivers/*/*npcm* F: Documentation/devicetree/bindings/*/*npcm* F: Documentation/devicetree/bindings/*/*/*npcm* @@ -2356,7 +2366,7 @@ F: arch/arm/mm/cache-uniphier.c F: arch/arm64/boot/dts/socionext/uniphier* F: drivers/bus/uniphier-system-bus.c F: drivers/clk/uniphier/ -F: drivers/dmaengine/uniphier-mdmac.c +F: drivers/dma/uniphier-mdmac.c F: drivers/gpio/gpio-uniphier.c F: drivers/i2c/busses/i2c-uniphier* F: drivers/irqchip/irq-uniphier-aidet.c @@ -4129,7 +4139,7 @@ F: drivers/cpuidle/* F: include/linux/cpuidle.h CRAMFS FILESYSTEM -M: Nicolas Pitre +M: Nicolas Pitre S: Maintained F: Documentation/filesystems/cramfs.txt F: fs/cramfs/ @@ -4894,6 +4904,14 @@ M: Dave Airlie S: Odd Fixes F: drivers/gpu/drm/ast/ +DRM DRIVER FOR ASPEED BMC GFX +M: Joel Stanley +L: linux-aspeed@lists.ozlabs.org +T: git git://anongit.freedesktop.org/drm/drm-misc +S: Supported +F: drivers/gpu/drm/aspeed/ +F: Documentation/devicetree/bindings/gpu/aspeed-gfx.txt + DRM DRIVER FOR BOCHS VIRTUAL GPU M: Gerd Hoffmann L: virtualization@lists.linux-foundation.org @@ -4907,6 +4925,12 @@ T: git git://anongit.freedesktop.org/drm/drm-misc S: Maintained F: drivers/gpu/drm/tve200/ +DRM DRIVER FOR FEIYANG FY07024DI26A30-D MIPI-DSI LCD PANELS +M: Jagan Teki +S: Maintained +F: drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c +F: Documentation/devicetree/bindings/display/panel/feiyang,fy07024di26a30d.txt + DRM DRIVER FOR ILITEK ILI9225 PANELS M: David Lechner S: Maintained @@ -4998,6 +5022,12 @@ S: Orphan / Obsolete F: drivers/gpu/drm/r128/ F: include/uapi/drm/r128_drm.h +DRM DRIVER FOR ROCKTECH JH057N00900 PANELS +M: Guido Günther +S: Maintained +F: drivers/gpu/drm/panel/panel-rocktech-jh057n00900.c +F: Documentation/devicetree/bindings/display/panel/rocktech,jh057n00900.txt + DRM DRIVER FOR SAVAGE VIDEO CARDS S: Orphan / Obsolete F: drivers/gpu/drm/savage/ @@ -5186,6 +5216,15 @@ S: Maintained F: drivers/gpu/drm/hisilicon/ F: Documentation/devicetree/bindings/display/hisilicon/ +DRM DRIVERS FOR LIMA +M: Qiang Yu +L: dri-devel@lists.freedesktop.org +L: lima@lists.freedesktop.org (moderated for non-subscribers) +S: Maintained +F: drivers/gpu/drm/lima/ +F: include/uapi/drm/lima_drm.h +T: git git://anongit.freedesktop.org/drm/drm-misc + DRM DRIVERS FOR MEDIATEK M: CK Hu M: Philipp Zabel @@ -5840,7 +5879,7 @@ L: netdev@vger.kernel.org S: Maintained F: Documentation/ABI/testing/sysfs-bus-mdio F: Documentation/devicetree/bindings/net/mdio* -F: Documentation/networking/phy.txt +F: Documentation/networking/phy.rst F: drivers/net/phy/ F: drivers/of/of_mdio.c F: drivers/of/of_net.c @@ -6415,7 +6454,6 @@ L: linux-kernel@vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git locking/core S: Maintained F: kernel/futex.c -F: kernel/futex_compat.c F: include/asm-generic/futex.h F: include/linux/futex.h F: include/uapi/linux/futex.h @@ -7523,7 +7561,7 @@ F: include/net/mac802154.h F: include/net/af_ieee802154.h F: include/net/cfg802154.h F: include/net/ieee802154_netdev.h -F: Documentation/networking/ieee802154.txt +F: Documentation/networking/ieee802154.rst IFE PROTOCOL M: Yotam Gigi @@ -13989,7 +14027,7 @@ F: drivers/media/rc/serial_ir.c SFC NETWORK DRIVER M: Solarflare linux maintainers M: Edward Cree -M: Bert Kenward +M: Martin Habets L: netdev@vger.kernel.org S: Supported F: drivers/net/ethernet/sfc/ @@ -16516,7 +16554,7 @@ F: drivers/char/virtio_console.c F: include/linux/virtio_console.h F: include/uapi/linux/virtio_console.h -VIRTIO CORE, NET AND BLOCK DRIVERS +VIRTIO CORE AND NET DRIVERS M: "Michael S. Tsirkin" M: Jason Wang L: virtualization@lists.linux-foundation.org @@ -16531,6 +16569,19 @@ F: include/uapi/linux/virtio_*.h F: drivers/crypto/virtio/ F: mm/balloon_compaction.c +VIRTIO BLOCK AND SCSI DRIVERS +M: "Michael S. Tsirkin" +M: Jason Wang +R: Paolo Bonzini +R: Stefan Hajnoczi +L: virtualization@lists.linux-foundation.org +S: Maintained +F: drivers/block/virtio_blk.c +F: drivers/scsi/virtio_scsi.c +F: include/uapi/linux/virtio_blk.h +F: include/uapi/linux/virtio_scsi.h +F: drivers/vhost/scsi.c + VIRTIO CRYPTO DRIVER M: Gonglei L: virtualization@lists.linux-foundation.org diff --git a/Makefile b/Makefile index c0a34064c574..71fd5c2ce067 100644 --- a/Makefile +++ b/Makefile @@ -2,7 +2,7 @@ VERSION = 5 PATCHLEVEL = 1 SUBLEVEL = 0 -EXTRAVERSION = -rc2 +EXTRAVERSION = -rc5 NAME = Shy Crocodile # *DOCUMENTATION* @@ -31,26 +31,12 @@ _all: # descending is started. They are now explicitly listed as the # prepare rule. -# Ugly workaround for Debian make-kpkg: -# make-kpkg directly includes the top Makefile of Linux kernel. In such a case, -# skip sub-make to support debian_* targets in ruleset/kernel_version.mk, but -# displays warning to discourage such abusage. -ifneq ($(word 2, $(MAKEFILE_LIST)),) -$(warning Do not include top Makefile of Linux Kernel) -sub-make-done := 1 -MAKEFLAGS += -rR -endif - -ifneq ($(sub-make-done),1) +ifneq ($(sub_make_done),1) # Do not use make's built-in rules and variables # (this increases performance and avoids hard-to-debug behaviour) MAKEFLAGS += -rR -# 'MAKEFLAGS += -rR' does not become immediately effective for old -# GNU Make versions. Cancel implicit rules for this Makefile. -$(lastword $(MAKEFILE_LIST)): ; - # Avoid funny character set dependencies unexport LC_ALL LC_COLLATE=C @@ -153,6 +139,7 @@ $(if $(KBUILD_OUTPUT),, \ # 'sub-make' below. MAKEFLAGS += --include-dir=$(CURDIR) +need-sub-make := 1 else # Do not print "Entering directory ..." at all for in-tree build. @@ -160,6 +147,18 @@ MAKEFLAGS += --no-print-directory endif # ifneq ($(KBUILD_OUTPUT),) +ifneq ($(filter 3.%,$(MAKE_VERSION)),) +# 'MAKEFLAGS += -rR' does not immediately become effective for GNU Make 3.x +# We need to invoke sub-make to avoid implicit rules in the top Makefile. +need-sub-make := 1 +# Cancel implicit rules for this Makefile. +$(lastword $(MAKEFILE_LIST)): ; +endif + +export sub_make_done := 1 + +ifeq ($(need-sub-make),1) + PHONY += $(MAKECMDGOALS) sub-make $(filter-out _all sub-make $(CURDIR)/Makefile, $(MAKECMDGOALS)) _all: sub-make @@ -167,12 +166,15 @@ $(filter-out _all sub-make $(CURDIR)/Makefile, $(MAKECMDGOALS)) _all: sub-make # Invoke a second make in the output directory, passing relevant variables sub-make: - $(Q)$(MAKE) sub-make-done=1 \ + $(Q)$(MAKE) \ $(if $(KBUILD_OUTPUT),-C $(KBUILD_OUTPUT) KBUILD_SRC=$(CURDIR)) \ -f $(CURDIR)/Makefile $(filter-out _all sub-make,$(MAKECMDGOALS)) -else # sub-make-done +endif # need-sub-make +endif # sub_make_done + # We process the rest of the Makefile if this is the final invocation of make +ifeq ($(need-sub-make),) # Do not print "Entering directory ...", # but we want to display it when entering to the output directory @@ -497,7 +499,8 @@ outputmakefile: ifneq ($(KBUILD_SRC),) $(Q)ln -fsn $(srctree) source $(Q)$(CONFIG_SHELL) $(srctree)/scripts/mkmakefile $(srctree) - $(Q){ echo "# this is build directory, ignore it"; echo "*"; } > .gitignore + $(Q)test -e .gitignore || \ + { echo "# this is build directory, ignore it"; echo "*"; } > .gitignore endif ifneq ($(shell $(CC) --version 2>&1 | head -n 1 | grep clang),) @@ -677,7 +680,7 @@ KBUILD_CFLAGS += $(call cc-disable-warning, format-overflow) KBUILD_CFLAGS += $(call cc-disable-warning, int-in-bool-context) ifdef CONFIG_CC_OPTIMIZE_FOR_SIZE -KBUILD_CFLAGS += $(call cc-option,-Oz,-Os) +KBUILD_CFLAGS += -Os else KBUILD_CFLAGS += -O2 endif @@ -950,9 +953,11 @@ mod_sign_cmd = true endif export mod_sign_cmd +HOST_LIBELF_LIBS = $(shell pkg-config libelf --libs 2>/dev/null || echo -lelf) + ifdef CONFIG_STACK_VALIDATION has_libelf := $(call try-run,\ - echo "int main() {}" | $(HOSTCC) -xc -o /dev/null -lelf -,1,0) + echo "int main() {}" | $(HOSTCC) -xc -o /dev/null $(HOST_LIBELF_LIBS) -,1,0) ifeq ($(has_libelf),1) objtool_target := tools/objtool FORCE else @@ -1757,7 +1762,7 @@ existing-targets := $(wildcard $(sort $(targets))) endif # ifeq ($(config-targets),1) endif # ifeq ($(mixed-targets),1) -endif # sub-make-done +endif # need-sub-make PHONY += FORCE FORCE: diff --git a/arch/alpha/include/asm/Kbuild b/arch/alpha/include/asm/Kbuild index dc0ab28baca1..70b783333965 100644 --- a/arch/alpha/include/asm/Kbuild +++ b/arch/alpha/include/asm/Kbuild @@ -6,6 +6,7 @@ generic-y += exec.h generic-y += export.h generic-y += fb.h generic-y += irq_work.h +generic-y += kvm_para.h generic-y += mcs_spinlock.h generic-y += mm-arch-hooks.h generic-y += preempt.h diff --git a/arch/alpha/include/uapi/asm/kvm_para.h b/arch/alpha/include/uapi/asm/kvm_para.h deleted file mode 100644 index baacc4996d18..000000000000 --- a/arch/alpha/include/uapi/asm/kvm_para.h +++ /dev/null @@ -1,2 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */ -#include diff --git a/arch/arc/include/asm/Kbuild b/arch/arc/include/asm/Kbuild index b41f8881ecc8..decc306a3b52 100644 --- a/arch/arc/include/asm/Kbuild +++ b/arch/arc/include/asm/Kbuild @@ -11,6 +11,7 @@ generic-y += hardirq.h generic-y += hw_irq.h generic-y += irq_regs.h generic-y += irq_work.h +generic-y += kvm_para.h generic-y += local.h generic-y += local64.h generic-y += mcs_spinlock.h diff --git a/arch/arc/include/asm/syscall.h b/arch/arc/include/asm/syscall.h index 29de09804306..c7a4201ed62b 100644 --- a/arch/arc/include/asm/syscall.h +++ b/arch/arc/include/asm/syscall.h @@ -55,12 +55,11 @@ syscall_set_return_value(struct task_struct *task, struct pt_regs *regs, */ static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) + unsigned long *args) { unsigned long *inside_ptregs = &(regs->r0); - inside_ptregs -= i; - - BUG_ON((i + n) > 6); + unsigned int n = 6; + unsigned int i = 0; while (n--) { args[i++] = (*inside_ptregs); diff --git a/arch/arc/include/uapi/asm/Kbuild b/arch/arc/include/uapi/asm/Kbuild index 755bb11323d8..1c72f04ff75d 100644 --- a/arch/arc/include/uapi/asm/Kbuild +++ b/arch/arc/include/uapi/asm/Kbuild @@ -1,2 +1 @@ -generic-y += kvm_para.h generic-y += ucontext.h diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 054ead960f98..850b4805e2d1 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -596,6 +596,7 @@ config ARCH_DAVINCI select HAVE_IDE select PM_GENERIC_DOMAINS if PM select PM_GENERIC_DOMAINS_OF if PM && OF + select REGMAP_MMIO select RESET_CONTROLLER select SPARSE_IRQ select USE_OF diff --git a/arch/arm/boot/dts/am335x-evm.dts b/arch/arm/boot/dts/am335x-evm.dts index dce5be5df97b..edcff79879e7 100644 --- a/arch/arm/boot/dts/am335x-evm.dts +++ b/arch/arm/boot/dts/am335x-evm.dts @@ -57,6 +57,24 @@ enable-active-high; }; + /* TPS79501 */ + v1_8d_reg: fixedregulator-v1_8d { + compatible = "regulator-fixed"; + regulator-name = "v1_8d"; + vin-supply = <&vbat>; + regulator-min-microvolt = <1800000>; + regulator-max-microvolt = <1800000>; + }; + + /* TPS79501 */ + v3_3d_reg: fixedregulator-v3_3d { + compatible = "regulator-fixed"; + regulator-name = "v3_3d"; + vin-supply = <&vbat>; + regulator-min-microvolt = <3300000>; + regulator-max-microvolt = <3300000>; + }; + matrix_keypad: matrix_keypad0 { compatible = "gpio-matrix-keypad"; debounce-delay-ms = <5>; @@ -499,10 +517,10 @@ status = "okay"; /* Regulators */ - AVDD-supply = <&vaux2_reg>; - IOVDD-supply = <&vaux2_reg>; - DRVDD-supply = <&vaux2_reg>; - DVDD-supply = <&vbat>; + AVDD-supply = <&v3_3d_reg>; + IOVDD-supply = <&v3_3d_reg>; + DRVDD-supply = <&v3_3d_reg>; + DVDD-supply = <&v1_8d_reg>; }; }; diff --git a/arch/arm/boot/dts/am335x-evmsk.dts b/arch/arm/boot/dts/am335x-evmsk.dts index b128998097ce..2c2d8b5b8cf5 100644 --- a/arch/arm/boot/dts/am335x-evmsk.dts +++ b/arch/arm/boot/dts/am335x-evmsk.dts @@ -73,6 +73,24 @@ enable-active-high; }; + /* TPS79518 */ + v1_8d_reg: fixedregulator-v1_8d { + compatible = "regulator-fixed"; + regulator-name = "v1_8d"; + vin-supply = <&vbat>; + regulator-min-microvolt = <1800000>; + regulator-max-microvolt = <1800000>; + }; + + /* TPS78633 */ + v3_3d_reg: fixedregulator-v3_3d { + compatible = "regulator-fixed"; + regulator-name = "v3_3d"; + vin-supply = <&vbat>; + regulator-min-microvolt = <3300000>; + regulator-max-microvolt = <3300000>; + }; + leds { pinctrl-names = "default"; pinctrl-0 = <&user_leds_s0>; @@ -501,10 +519,10 @@ status = "okay"; /* Regulators */ - AVDD-supply = <&vaux2_reg>; - IOVDD-supply = <&vaux2_reg>; - DRVDD-supply = <&vaux2_reg>; - DVDD-supply = <&vbat>; + AVDD-supply = <&v3_3d_reg>; + IOVDD-supply = <&v3_3d_reg>; + DRVDD-supply = <&v3_3d_reg>; + DVDD-supply = <&v1_8d_reg>; }; }; diff --git a/arch/arm/boot/dts/am33xx-l4.dtsi b/arch/arm/boot/dts/am33xx-l4.dtsi index f459ec316a22..ca6d9f02a800 100644 --- a/arch/arm/boot/dts/am33xx-l4.dtsi +++ b/arch/arm/boot/dts/am33xx-l4.dtsi @@ -1762,7 +1762,7 @@ reg = <0xcc000 0x4>; reg-names = "rev"; /* Domains (P, C): per_pwrdm, l4ls_clkdm */ - clocks = <&l4ls_clkctrl AM3_D_CAN0_CLKCTRL 0>; + clocks = <&l4ls_clkctrl AM3_L4LS_D_CAN0_CLKCTRL 0>; clock-names = "fck"; #address-cells = <1>; #size-cells = <1>; @@ -1785,7 +1785,7 @@ reg = <0xd0000 0x4>; reg-names = "rev"; /* Domains (P, C): per_pwrdm, l4ls_clkdm */ - clocks = <&l4ls_clkctrl AM3_D_CAN1_CLKCTRL 0>; + clocks = <&l4ls_clkctrl AM3_L4LS_D_CAN1_CLKCTRL 0>; clock-names = "fck"; #address-cells = <1>; #size-cells = <1>; diff --git a/arch/arm/boot/dts/bcm2835-rpi-b-rev2.dts b/arch/arm/boot/dts/bcm2835-rpi-b-rev2.dts index 5641d162dfdb..28e7513ce617 100644 --- a/arch/arm/boot/dts/bcm2835-rpi-b-rev2.dts +++ b/arch/arm/boot/dts/bcm2835-rpi-b-rev2.dts @@ -93,7 +93,7 @@ }; &hdmi { - hpd-gpios = <&gpio 46 GPIO_ACTIVE_LOW>; + hpd-gpios = <&gpio 46 GPIO_ACTIVE_HIGH>; }; &pwm { diff --git a/arch/arm/boot/dts/imx6dl-yapp4-common.dtsi b/arch/arm/boot/dts/imx6dl-yapp4-common.dtsi index b715ab0fa1ff..e8d800fec637 100644 --- a/arch/arm/boot/dts/imx6dl-yapp4-common.dtsi +++ b/arch/arm/boot/dts/imx6dl-yapp4-common.dtsi @@ -114,9 +114,9 @@ reg = <2>; }; - switch@0 { + switch@10 { compatible = "qca,qca8334"; - reg = <0>; + reg = <10>; switch_ports: ports { #address-cells = <1>; @@ -125,7 +125,7 @@ ethphy0: port@0 { reg = <0>; label = "cpu"; - phy-mode = "rgmii"; + phy-mode = "rgmii-id"; ethernet = <&fec>; fixed-link { diff --git a/arch/arm/boot/dts/imx6qdl-icore-rqs.dtsi b/arch/arm/boot/dts/imx6qdl-icore-rqs.dtsi index 1d1b4bd0670f..a4217f564a53 100644 --- a/arch/arm/boot/dts/imx6qdl-icore-rqs.dtsi +++ b/arch/arm/boot/dts/imx6qdl-icore-rqs.dtsi @@ -264,7 +264,7 @@ pinctrl-2 = <&pinctrl_usdhc3_200mhz>; vmcc-supply = <®_sd3_vmmc>; cd-gpios = <&gpio1 1 GPIO_ACTIVE_LOW>; - bus-witdh = <4>; + bus-width = <4>; no-1-8-v; status = "okay"; }; @@ -275,7 +275,7 @@ pinctrl-1 = <&pinctrl_usdhc4_100mhz>; pinctrl-2 = <&pinctrl_usdhc4_200mhz>; vmcc-supply = <®_sd4_vmmc>; - bus-witdh = <8>; + bus-width = <8>; no-1-8-v; non-removable; status = "okay"; diff --git a/arch/arm/boot/dts/imx6qdl-phytec-pfla02.dtsi b/arch/arm/boot/dts/imx6qdl-phytec-pfla02.dtsi index 433bf09a1954..027df06c5dc7 100644 --- a/arch/arm/boot/dts/imx6qdl-phytec-pfla02.dtsi +++ b/arch/arm/boot/dts/imx6qdl-phytec-pfla02.dtsi @@ -91,6 +91,7 @@ pinctrl-0 = <&pinctrl_enet>; phy-handle = <ðphy>; phy-mode = "rgmii"; + phy-reset-duration = <10>; /* in msecs */ phy-reset-gpios = <&gpio3 23 GPIO_ACTIVE_LOW>; phy-supply = <&vdd_eth_io_reg>; status = "disabled"; diff --git a/arch/arm/boot/dts/imx6ull-pinfunc-snvs.h b/arch/arm/boot/dts/imx6ull-pinfunc-snvs.h index f6fb6783c193..54cfe72295aa 100644 --- a/arch/arm/boot/dts/imx6ull-pinfunc-snvs.h +++ b/arch/arm/boot/dts/imx6ull-pinfunc-snvs.h @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 +/* SPDX-License-Identifier: GPL-2.0 */ /* * Copyright (C) 2016 Freescale Semiconductor, Inc. * Copyright (C) 2017 NXP diff --git a/arch/arm/boot/dts/rk3288-tinker.dtsi b/arch/arm/boot/dts/rk3288-tinker.dtsi index aa107ee41b8b..ef653c3209bc 100644 --- a/arch/arm/boot/dts/rk3288-tinker.dtsi +++ b/arch/arm/boot/dts/rk3288-tinker.dtsi @@ -254,6 +254,7 @@ }; vccio_sd: LDO_REG5 { + regulator-boot-on; regulator-min-microvolt = <1800000>; regulator-max-microvolt = <3300000>; regulator-name = "vccio_sd"; @@ -430,7 +431,7 @@ bus-width = <4>; cap-mmc-highspeed; cap-sd-highspeed; - card-detect-delay = <200>; + broken-cd; disable-wp; /* wp not hooked up */ pinctrl-names = "default"; pinctrl-0 = <&sdmmc_clk &sdmmc_cmd &sdmmc_cd &sdmmc_bus4>; diff --git a/arch/arm/boot/dts/rk3288-veyron.dtsi b/arch/arm/boot/dts/rk3288-veyron.dtsi index 0bc2409f6903..192dbc089ade 100644 --- a/arch/arm/boot/dts/rk3288-veyron.dtsi +++ b/arch/arm/boot/dts/rk3288-veyron.dtsi @@ -25,8 +25,6 @@ gpio_keys: gpio-keys { compatible = "gpio-keys"; - #address-cells = <1>; - #size-cells = <0>; pinctrl-names = "default"; pinctrl-0 = <&pwr_key_l>; diff --git a/arch/arm/boot/dts/rk3288.dtsi b/arch/arm/boot/dts/rk3288.dtsi index ca7d52daa8fb..a024d1e7e74c 100644 --- a/arch/arm/boot/dts/rk3288.dtsi +++ b/arch/arm/boot/dts/rk3288.dtsi @@ -70,7 +70,7 @@ compatible = "arm,cortex-a12"; reg = <0x501>; resets = <&cru SRST_CORE1>; - operating-points = <&cpu_opp_table>; + operating-points-v2 = <&cpu_opp_table>; #cooling-cells = <2>; /* min followed by max */ clock-latency = <40000>; clocks = <&cru ARMCLK>; @@ -80,7 +80,7 @@ compatible = "arm,cortex-a12"; reg = <0x502>; resets = <&cru SRST_CORE2>; - operating-points = <&cpu_opp_table>; + operating-points-v2 = <&cpu_opp_table>; #cooling-cells = <2>; /* min followed by max */ clock-latency = <40000>; clocks = <&cru ARMCLK>; @@ -90,7 +90,7 @@ compatible = "arm,cortex-a12"; reg = <0x503>; resets = <&cru SRST_CORE3>; - operating-points = <&cpu_opp_table>; + operating-points-v2 = <&cpu_opp_table>; #cooling-cells = <2>; /* min followed by max */ clock-latency = <40000>; clocks = <&cru ARMCLK>; @@ -1119,8 +1119,6 @@ clock-names = "ref", "pclk"; power-domains = <&power RK3288_PD_VIO>; rockchip,grf = <&grf>; - #address-cells = <1>; - #size-cells = <0>; status = "disabled"; ports { @@ -1282,27 +1280,27 @@ gpu_opp_table: gpu-opp-table { compatible = "operating-points-v2"; - opp@100000000 { + opp-100000000 { opp-hz = /bits/ 64 <100000000>; opp-microvolt = <950000>; }; - opp@200000000 { + opp-200000000 { opp-hz = /bits/ 64 <200000000>; opp-microvolt = <950000>; }; - opp@300000000 { + opp-300000000 { opp-hz = /bits/ 64 <300000000>; opp-microvolt = <1000000>; }; - opp@400000000 { + opp-400000000 { opp-hz = /bits/ 64 <400000000>; opp-microvolt = <1100000>; }; - opp@500000000 { + opp-500000000 { opp-hz = /bits/ 64 <500000000>; opp-microvolt = <1200000>; }; - opp@600000000 { + opp-600000000 { opp-hz = /bits/ 64 <600000000>; opp-microvolt = <1250000>; }; diff --git a/arch/arm/boot/dts/sama5d2-pinfunc.h b/arch/arm/boot/dts/sama5d2-pinfunc.h index 1c01a6f843d8..28a2e45752fe 100644 --- a/arch/arm/boot/dts/sama5d2-pinfunc.h +++ b/arch/arm/boot/dts/sama5d2-pinfunc.h @@ -518,7 +518,7 @@ #define PIN_PC9__GPIO PINMUX_PIN(PIN_PC9, 0, 0) #define PIN_PC9__FIQ PINMUX_PIN(PIN_PC9, 1, 3) #define PIN_PC9__GTSUCOMP PINMUX_PIN(PIN_PC9, 2, 1) -#define PIN_PC9__ISC_D0 PINMUX_PIN(PIN_PC9, 2, 1) +#define PIN_PC9__ISC_D0 PINMUX_PIN(PIN_PC9, 3, 1) #define PIN_PC9__TIOA4 PINMUX_PIN(PIN_PC9, 4, 2) #define PIN_PC10 74 #define PIN_PC10__GPIO PINMUX_PIN(PIN_PC10, 0, 0) diff --git a/arch/arm/configs/imx_v4_v5_defconfig b/arch/arm/configs/imx_v4_v5_defconfig index 8661dd9b064a..b37f8e675e40 100644 --- a/arch/arm/configs/imx_v4_v5_defconfig +++ b/arch/arm/configs/imx_v4_v5_defconfig @@ -170,6 +170,9 @@ CONFIG_IMX_SDMA=y # CONFIG_IOMMU_SUPPORT is not set CONFIG_IIO=y CONFIG_FSL_MX25_ADC=y +CONFIG_PWM=y +CONFIG_PWM_IMX1=y +CONFIG_PWM_IMX27=y CONFIG_EXT4_FS=y # CONFIG_DNOTIFY is not set CONFIG_VFAT_FS=y diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig index 5586a5074a96..50fb01d70b10 100644 --- a/arch/arm/configs/imx_v6_v7_defconfig +++ b/arch/arm/configs/imx_v6_v7_defconfig @@ -398,7 +398,7 @@ CONFIG_MAG3110=y CONFIG_MPL3115=y CONFIG_PWM=y CONFIG_PWM_FSL_FTM=y -CONFIG_PWM_IMX=y +CONFIG_PWM_IMX27=y CONFIG_NVMEM_IMX_OCOTP=y CONFIG_NVMEM_VF610_OCOTP=y CONFIG_TEE=y diff --git a/arch/arm/include/asm/kvm_mmu.h b/arch/arm/include/asm/kvm_mmu.h index 2de96a180166..31de4ab93005 100644 --- a/arch/arm/include/asm/kvm_mmu.h +++ b/arch/arm/include/asm/kvm_mmu.h @@ -381,6 +381,17 @@ static inline int kvm_read_guest_lock(struct kvm *kvm, return ret; } +static inline int kvm_write_guest_lock(struct kvm *kvm, gpa_t gpa, + const void *data, unsigned long len) +{ + int srcu_idx = srcu_read_lock(&kvm->srcu); + int ret = kvm_write_guest(kvm, gpa, data, len); + + srcu_read_unlock(&kvm->srcu, srcu_idx); + + return ret; +} + static inline void *kvm_get_hyp_vector(void) { switch(read_cpuid_part()) { diff --git a/arch/arm/include/asm/stage2_pgtable.h b/arch/arm/include/asm/stage2_pgtable.h index de2089501b8b..9e11dce55e06 100644 --- a/arch/arm/include/asm/stage2_pgtable.h +++ b/arch/arm/include/asm/stage2_pgtable.h @@ -75,6 +75,8 @@ static inline bool kvm_stage2_has_pud(struct kvm *kvm) #define S2_PMD_MASK PMD_MASK #define S2_PMD_SIZE PMD_SIZE +#define S2_PUD_MASK PUD_MASK +#define S2_PUD_SIZE PUD_SIZE static inline bool kvm_stage2_has_pmd(struct kvm *kvm) { diff --git a/arch/arm/include/asm/syscall.h b/arch/arm/include/asm/syscall.h index 06dea6bce293..080ce70cab12 100644 --- a/arch/arm/include/asm/syscall.h +++ b/arch/arm/include/asm/syscall.h @@ -55,53 +55,22 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { - if (n == 0) - return; + args[0] = regs->ARM_ORIG_r0; + args++; - if (i + n > SYSCALL_MAX_ARGS) { - unsigned long *args_bad = args + SYSCALL_MAX_ARGS - i; - unsigned int n_bad = n + i - SYSCALL_MAX_ARGS; - pr_warn("%s called with max args %d, handling only %d\n", - __func__, i + n, SYSCALL_MAX_ARGS); - memset(args_bad, 0, n_bad * sizeof(args[0])); - n = SYSCALL_MAX_ARGS - i; - } - - if (i == 0) { - args[0] = regs->ARM_ORIG_r0; - args++; - i++; - n--; - } - - memcpy(args, ®s->ARM_r0 + i, n * sizeof(args[0])); + memcpy(args, ®s->ARM_r0 + 1, 5 * sizeof(args[0])); } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { - if (n == 0) - return; + regs->ARM_ORIG_r0 = args[0]; + args++; - if (i + n > SYSCALL_MAX_ARGS) { - pr_warn("%s called with max args %d, handling only %d\n", - __func__, i + n, SYSCALL_MAX_ARGS); - n = SYSCALL_MAX_ARGS - i; - } - - if (i == 0) { - regs->ARM_ORIG_r0 = args[0]; - args++; - i++; - n--; - } - - memcpy(®s->ARM_r0 + i, args, n * sizeof(args[0])); + memcpy(®s->ARM_r0 + 1, args, 5 * sizeof(args[0])); } static inline int syscall_get_arch(void) diff --git a/arch/arm/include/uapi/asm/Kbuild b/arch/arm/include/uapi/asm/Kbuild index 23b4464c0995..ce8573157774 100644 --- a/arch/arm/include/uapi/asm/Kbuild +++ b/arch/arm/include/uapi/asm/Kbuild @@ -3,3 +3,4 @@ generated-y += unistd-common.h generated-y += unistd-oabi.h generated-y += unistd-eabi.h +generic-y += kvm_para.h diff --git a/arch/arm/include/uapi/asm/kvm_para.h b/arch/arm/include/uapi/asm/kvm_para.h deleted file mode 100644 index baacc4996d18..000000000000 --- a/arch/arm/include/uapi/asm/kvm_para.h +++ /dev/null @@ -1,2 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */ -#include diff --git a/arch/arm/mach-at91/pm.c b/arch/arm/mach-at91/pm.c index 51e808adb00c..2a757dcaa1a5 100644 --- a/arch/arm/mach-at91/pm.c +++ b/arch/arm/mach-at91/pm.c @@ -591,13 +591,13 @@ static int __init at91_pm_backup_init(void) np = of_find_compatible_node(NULL, NULL, "atmel,sama5d2-securam"); if (!np) - goto securam_fail; + goto securam_fail_no_ref_dev; pdev = of_find_device_by_node(np); of_node_put(np); if (!pdev) { pr_warn("%s: failed to find securam device!\n", __func__); - goto securam_fail; + goto securam_fail_no_ref_dev; } sram_pool = gen_pool_get(&pdev->dev, NULL); @@ -620,6 +620,8 @@ static int __init at91_pm_backup_init(void) return 0; securam_fail: + put_device(&pdev->dev); +securam_fail_no_ref_dev: iounmap(pm_data.sfrbu); pm_data.sfrbu = NULL; return ret; diff --git a/arch/arm/mach-imx/cpuidle-imx6q.c b/arch/arm/mach-imx/cpuidle-imx6q.c index bfeb25aaf9a2..326e870d7123 100644 --- a/arch/arm/mach-imx/cpuidle-imx6q.c +++ b/arch/arm/mach-imx/cpuidle-imx6q.c @@ -16,30 +16,23 @@ #include "cpuidle.h" #include "hardware.h" -static atomic_t master = ATOMIC_INIT(0); -static DEFINE_SPINLOCK(master_lock); +static int num_idle_cpus = 0; +static DEFINE_SPINLOCK(cpuidle_lock); static int imx6q_enter_wait(struct cpuidle_device *dev, struct cpuidle_driver *drv, int index) { - if (atomic_inc_return(&master) == num_online_cpus()) { - /* - * With this lock, we prevent other cpu to exit and enter - * this function again and become the master. - */ - if (!spin_trylock(&master_lock)) - goto idle; + spin_lock(&cpuidle_lock); + if (++num_idle_cpus == num_online_cpus()) imx6_set_lpm(WAIT_UNCLOCKED); - cpu_do_idle(); - imx6_set_lpm(WAIT_CLOCKED); - spin_unlock(&master_lock); - goto done; - } + spin_unlock(&cpuidle_lock); -idle: cpu_do_idle(); -done: - atomic_dec(&master); + + spin_lock(&cpuidle_lock); + if (num_idle_cpus-- == num_online_cpus()) + imx6_set_lpm(WAIT_CLOCKED); + spin_unlock(&cpuidle_lock); return index; } diff --git a/arch/arm/mach-imx/mach-imx51.c b/arch/arm/mach-imx/mach-imx51.c index c7169c2f94c4..08c7892866c2 100644 --- a/arch/arm/mach-imx/mach-imx51.c +++ b/arch/arm/mach-imx/mach-imx51.c @@ -59,6 +59,7 @@ static void __init imx51_m4if_setup(void) return; m4if_base = of_iomap(np, 0); + of_node_put(np); if (!m4if_base) { pr_err("Unable to map M4IF registers\n"); return; diff --git a/arch/arm/mach-iop13xx/setup.c b/arch/arm/mach-iop13xx/setup.c index 53c316f7301e..fe4932fda01d 100644 --- a/arch/arm/mach-iop13xx/setup.c +++ b/arch/arm/mach-iop13xx/setup.c @@ -300,7 +300,7 @@ static struct resource iop13xx_adma_2_resources[] = { } }; -static u64 iop13xx_adma_dmamask = DMA_BIT_MASK(64); +static u64 iop13xx_adma_dmamask = DMA_BIT_MASK(32); static struct iop_adma_platform_data iop13xx_adma_0_data = { .hw_id = 0, .pool_size = PAGE_SIZE, @@ -324,7 +324,7 @@ static struct platform_device iop13xx_adma_0_channel = { .resource = iop13xx_adma_0_resources, .dev = { .dma_mask = &iop13xx_adma_dmamask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), .platform_data = (void *) &iop13xx_adma_0_data, }, }; @@ -336,7 +336,7 @@ static struct platform_device iop13xx_adma_1_channel = { .resource = iop13xx_adma_1_resources, .dev = { .dma_mask = &iop13xx_adma_dmamask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), .platform_data = (void *) &iop13xx_adma_1_data, }, }; @@ -348,7 +348,7 @@ static struct platform_device iop13xx_adma_2_channel = { .resource = iop13xx_adma_2_resources, .dev = { .dma_mask = &iop13xx_adma_dmamask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), .platform_data = (void *) &iop13xx_adma_2_data, }, }; diff --git a/arch/arm/mach-iop13xx/tpmi.c b/arch/arm/mach-iop13xx/tpmi.c index db511ec2b1df..116feb6b261e 100644 --- a/arch/arm/mach-iop13xx/tpmi.c +++ b/arch/arm/mach-iop13xx/tpmi.c @@ -152,7 +152,7 @@ static struct resource iop13xx_tpmi_3_resources[] = { } }; -u64 iop13xx_tpmi_mask = DMA_BIT_MASK(64); +u64 iop13xx_tpmi_mask = DMA_BIT_MASK(32); static struct platform_device iop13xx_tpmi_0_device = { .name = "iop-tpmi", .id = 0, @@ -160,7 +160,7 @@ static struct platform_device iop13xx_tpmi_0_device = { .resource = iop13xx_tpmi_0_resources, .dev = { .dma_mask = &iop13xx_tpmi_mask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), }, }; @@ -171,7 +171,7 @@ static struct platform_device iop13xx_tpmi_1_device = { .resource = iop13xx_tpmi_1_resources, .dev = { .dma_mask = &iop13xx_tpmi_mask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), }, }; @@ -182,7 +182,7 @@ static struct platform_device iop13xx_tpmi_2_device = { .resource = iop13xx_tpmi_2_resources, .dev = { .dma_mask = &iop13xx_tpmi_mask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), }, }; @@ -193,7 +193,7 @@ static struct platform_device iop13xx_tpmi_3_device = { .resource = iop13xx_tpmi_3_resources, .dev = { .dma_mask = &iop13xx_tpmi_mask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), }, }; diff --git a/arch/arm/mach-milbeaut/platsmp.c b/arch/arm/mach-milbeaut/platsmp.c index 591543c81399..3ea880f5fcb7 100644 --- a/arch/arm/mach-milbeaut/platsmp.c +++ b/arch/arm/mach-milbeaut/platsmp.c @@ -65,6 +65,7 @@ static void m10v_smp_init(unsigned int max_cpus) writel(KERNEL_UNBOOT_FLAG, m10v_smp_base + cpu * 4); } +#ifdef CONFIG_HOTPLUG_CPU static void m10v_cpu_die(unsigned int l_cpu) { gic_cpu_if_down(0); @@ -83,12 +84,15 @@ static int m10v_cpu_kill(unsigned int l_cpu) return 1; } +#endif static struct smp_operations m10v_smp_ops __initdata = { .smp_prepare_cpus = m10v_smp_init, .smp_boot_secondary = m10v_boot_secondary, +#ifdef CONFIG_HOTPLUG_CPU .cpu_die = m10v_cpu_die, .cpu_kill = m10v_cpu_kill, +#endif }; CPU_METHOD_OF_DECLARE(m10v_smp, "socionext,milbeaut-m10v-smp", &m10v_smp_ops); diff --git a/arch/arm/mach-omap1/board-ams-delta.c b/arch/arm/mach-omap1/board-ams-delta.c index be30c3c061b4..1b15d593837e 100644 --- a/arch/arm/mach-omap1/board-ams-delta.c +++ b/arch/arm/mach-omap1/board-ams-delta.c @@ -182,6 +182,7 @@ static struct resource latch1_resources[] = { static struct bgpio_pdata latch1_pdata = { .label = LATCH1_LABEL, + .base = -1, .ngpio = LATCH1_NGPIO, }; @@ -219,6 +220,7 @@ static struct resource latch2_resources[] = { static struct bgpio_pdata latch2_pdata = { .label = LATCH2_LABEL, + .base = -1, .ngpio = LATCH2_NGPIO, }; diff --git a/arch/arm/mach-omap2/display.c b/arch/arm/mach-omap2/display.c index 1444b4b4bd9f..439e143cad7b 100644 --- a/arch/arm/mach-omap2/display.c +++ b/arch/arm/mach-omap2/display.c @@ -250,8 +250,10 @@ static int __init omapdss_init_of(void) if (!node) return 0; - if (!of_device_is_available(node)) + if (!of_device_is_available(node)) { + of_node_put(node); return 0; + } pdev = of_find_device_by_node(node); diff --git a/arch/arm/plat-iop/adma.c b/arch/arm/plat-iop/adma.c index a4d1f8de3b5b..d9612221e484 100644 --- a/arch/arm/plat-iop/adma.c +++ b/arch/arm/plat-iop/adma.c @@ -143,7 +143,7 @@ struct platform_device iop3xx_dma_0_channel = { .resource = iop3xx_dma_0_resources, .dev = { .dma_mask = &iop3xx_adma_dmamask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), .platform_data = (void *) &iop3xx_dma_0_data, }, }; @@ -155,7 +155,7 @@ struct platform_device iop3xx_dma_1_channel = { .resource = iop3xx_dma_1_resources, .dev = { .dma_mask = &iop3xx_adma_dmamask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), .platform_data = (void *) &iop3xx_dma_1_data, }, }; @@ -167,7 +167,7 @@ struct platform_device iop3xx_aau_channel = { .resource = iop3xx_aau_resources, .dev = { .dma_mask = &iop3xx_adma_dmamask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), .platform_data = (void *) &iop3xx_aau_data, }, }; diff --git a/arch/arm/plat-orion/common.c b/arch/arm/plat-orion/common.c index a6c81ce00f52..8647cb80a93b 100644 --- a/arch/arm/plat-orion/common.c +++ b/arch/arm/plat-orion/common.c @@ -622,7 +622,7 @@ static struct platform_device orion_xor0_shared = { .resource = orion_xor0_shared_resources, .dev = { .dma_mask = &orion_xor_dmamask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), .platform_data = &orion_xor0_pdata, }, }; @@ -683,7 +683,7 @@ static struct platform_device orion_xor1_shared = { .resource = orion_xor1_shared_resources, .dev = { .dma_mask = &orion_xor_dmamask, - .coherent_dma_mask = DMA_BIT_MASK(64), + .coherent_dma_mask = DMA_BIT_MASK(32), .platform_data = &orion_xor1_pdata, }, }; diff --git a/arch/arm64/Kconfig.platforms b/arch/arm64/Kconfig.platforms index 70498a033cf5..b5ca9c50876d 100644 --- a/arch/arm64/Kconfig.platforms +++ b/arch/arm64/Kconfig.platforms @@ -27,6 +27,7 @@ config ARCH_BCM2835 bool "Broadcom BCM2835 family" select TIMER_OF select GPIOLIB + select MFD_CORE select PINCTRL select PINCTRL_BCM2835 select ARM_AMBA diff --git a/arch/arm64/boot/dts/altera/socfpga_stratix10.dtsi b/arch/arm64/boot/dts/altera/socfpga_stratix10.dtsi index 7c649f6b14cb..cd7c76e58b09 100644 --- a/arch/arm64/boot/dts/altera/socfpga_stratix10.dtsi +++ b/arch/arm64/boot/dts/altera/socfpga_stratix10.dtsi @@ -162,6 +162,7 @@ rx-fifo-depth = <16384>; snps,multicast-filter-bins = <256>; iommus = <&smmu 1>; + altr,sysmgr-syscon = <&sysmgr 0x44 0>; status = "disabled"; }; @@ -179,6 +180,7 @@ rx-fifo-depth = <16384>; snps,multicast-filter-bins = <256>; iommus = <&smmu 2>; + altr,sysmgr-syscon = <&sysmgr 0x48 0>; status = "disabled"; }; @@ -196,6 +198,7 @@ rx-fifo-depth = <16384>; snps,multicast-filter-bins = <256>; iommus = <&smmu 3>; + altr,sysmgr-syscon = <&sysmgr 0x4c 0>; status = "disabled"; }; diff --git a/arch/arm64/boot/dts/nvidia/tegra186.dtsi b/arch/arm64/boot/dts/nvidia/tegra186.dtsi index bb2045be8814..97aeb946ed5e 100644 --- a/arch/arm64/boot/dts/nvidia/tegra186.dtsi +++ b/arch/arm64/boot/dts/nvidia/tegra186.dtsi @@ -321,7 +321,6 @@ nvidia,default-trim = <0x9>; nvidia,dqs-trim = <63>; mmc-hs400-1_8v; - supports-cqe; status = "disabled"; }; diff --git a/arch/arm64/boot/dts/renesas/r8a774c0.dtsi b/arch/arm64/boot/dts/renesas/r8a774c0.dtsi index 61a0afb74e63..1ea684af99c4 100644 --- a/arch/arm64/boot/dts/renesas/r8a774c0.dtsi +++ b/arch/arm64/boot/dts/renesas/r8a774c0.dtsi @@ -2,7 +2,7 @@ /* * Device Tree Source for the RZ/G2E (R8A774C0) SoC * - * Copyright (C) 2018 Renesas Electronics Corp. + * Copyright (C) 2018-2019 Renesas Electronics Corp. */ #include @@ -1150,9 +1150,8 @@ <&cpg CPG_CORE R8A774C0_CLK_S3D1C>, <&scif_clk>; clock-names = "fck", "brg_int", "scif_clk"; - dmas = <&dmac1 0x5b>, <&dmac1 0x5a>, - <&dmac2 0x5b>, <&dmac2 0x5a>; - dma-names = "tx", "rx", "tx", "rx"; + dmas = <&dmac0 0x5b>, <&dmac0 0x5a>; + dma-names = "tx", "rx"; power-domains = <&sysc R8A774C0_PD_ALWAYS_ON>; resets = <&cpg 202>; status = "disabled"; diff --git a/arch/arm64/boot/dts/renesas/r8a77990.dtsi b/arch/arm64/boot/dts/renesas/r8a77990.dtsi index a69faa60ea4d..d2ad665fe2d9 100644 --- a/arch/arm64/boot/dts/renesas/r8a77990.dtsi +++ b/arch/arm64/boot/dts/renesas/r8a77990.dtsi @@ -2,7 +2,7 @@ /* * Device Tree Source for the R-Car E3 (R8A77990) SoC * - * Copyright (C) 2018 Renesas Electronics Corp. + * Copyright (C) 2018-2019 Renesas Electronics Corp. */ #include @@ -1067,9 +1067,8 @@ <&cpg CPG_CORE R8A77990_CLK_S3D1C>, <&scif_clk>; clock-names = "fck", "brg_int", "scif_clk"; - dmas = <&dmac1 0x5b>, <&dmac1 0x5a>, - <&dmac2 0x5b>, <&dmac2 0x5a>; - dma-names = "tx", "rx", "tx", "rx"; + dmas = <&dmac0 0x5b>, <&dmac0 0x5a>; + dma-names = "tx", "rx"; power-domains = <&sysc R8A77990_PD_ALWAYS_ON>; resets = <&cpg 202>; status = "disabled"; diff --git a/arch/arm64/boot/dts/rockchip/rk3328-roc-cc.dts b/arch/arm64/boot/dts/rockchip/rk3328-roc-cc.dts index 33c44e857247..0e34354b2092 100644 --- a/arch/arm64/boot/dts/rockchip/rk3328-roc-cc.dts +++ b/arch/arm64/boot/dts/rockchip/rk3328-roc-cc.dts @@ -108,8 +108,8 @@ snps,reset-gpio = <&gpio1 RK_PC2 GPIO_ACTIVE_LOW>; snps,reset-active-low; snps,reset-delays-us = <0 10000 50000>; - tx_delay = <0x25>; - rx_delay = <0x11>; + tx_delay = <0x24>; + rx_delay = <0x18>; status = "okay"; }; diff --git a/arch/arm64/boot/dts/rockchip/rk3328-rock64.dts b/arch/arm64/boot/dts/rockchip/rk3328-rock64.dts index 2157a528276b..79b4d1d4b5d6 100644 --- a/arch/arm64/boot/dts/rockchip/rk3328-rock64.dts +++ b/arch/arm64/boot/dts/rockchip/rk3328-rock64.dts @@ -46,8 +46,7 @@ vcc_host1_5v: vcc_otg_5v: vcc-host1-5v-regulator { compatible = "regulator-fixed"; - enable-active-high; - gpio = <&gpio0 RK_PA2 GPIO_ACTIVE_HIGH>; + gpio = <&gpio0 RK_PA2 GPIO_ACTIVE_LOW>; pinctrl-names = "default"; pinctrl-0 = <&usb20_host_drv>; regulator-name = "vcc_host1_5v"; diff --git a/arch/arm64/boot/dts/rockchip/rk3328.dtsi b/arch/arm64/boot/dts/rockchip/rk3328.dtsi index 84f14b132e8f..dabef1a21649 100644 --- a/arch/arm64/boot/dts/rockchip/rk3328.dtsi +++ b/arch/arm64/boot/dts/rockchip/rk3328.dtsi @@ -1445,11 +1445,11 @@ sdmmc0 { sdmmc0_clk: sdmmc0-clk { - rockchip,pins = <1 RK_PA6 1 &pcfg_pull_none_4ma>; + rockchip,pins = <1 RK_PA6 1 &pcfg_pull_none_8ma>; }; sdmmc0_cmd: sdmmc0-cmd { - rockchip,pins = <1 RK_PA4 1 &pcfg_pull_up_4ma>; + rockchip,pins = <1 RK_PA4 1 &pcfg_pull_up_8ma>; }; sdmmc0_dectn: sdmmc0-dectn { @@ -1461,14 +1461,14 @@ }; sdmmc0_bus1: sdmmc0-bus1 { - rockchip,pins = <1 RK_PA0 1 &pcfg_pull_up_4ma>; + rockchip,pins = <1 RK_PA0 1 &pcfg_pull_up_8ma>; }; sdmmc0_bus4: sdmmc0-bus4 { - rockchip,pins = <1 RK_PA0 1 &pcfg_pull_up_4ma>, - <1 RK_PA1 1 &pcfg_pull_up_4ma>, - <1 RK_PA2 1 &pcfg_pull_up_4ma>, - <1 RK_PA3 1 &pcfg_pull_up_4ma>; + rockchip,pins = <1 RK_PA0 1 &pcfg_pull_up_8ma>, + <1 RK_PA1 1 &pcfg_pull_up_8ma>, + <1 RK_PA2 1 &pcfg_pull_up_8ma>, + <1 RK_PA3 1 &pcfg_pull_up_8ma>; }; sdmmc0_gpio: sdmmc0-gpio { @@ -1642,50 +1642,50 @@ rgmiim1_pins: rgmiim1-pins { rockchip,pins = /* mac_txclk */ - <1 RK_PB4 2 &pcfg_pull_none_12ma>, + <1 RK_PB4 2 &pcfg_pull_none_8ma>, /* mac_rxclk */ - <1 RK_PB5 2 &pcfg_pull_none_2ma>, + <1 RK_PB5 2 &pcfg_pull_none_4ma>, /* mac_mdio */ - <1 RK_PC3 2 &pcfg_pull_none_2ma>, + <1 RK_PC3 2 &pcfg_pull_none_4ma>, /* mac_txen */ - <1 RK_PD1 2 &pcfg_pull_none_12ma>, + <1 RK_PD1 2 &pcfg_pull_none_8ma>, /* mac_clk */ - <1 RK_PC5 2 &pcfg_pull_none_2ma>, + <1 RK_PC5 2 &pcfg_pull_none_4ma>, /* mac_rxdv */ - <1 RK_PC6 2 &pcfg_pull_none_2ma>, + <1 RK_PC6 2 &pcfg_pull_none_4ma>, /* mac_mdc */ - <1 RK_PC7 2 &pcfg_pull_none_2ma>, + <1 RK_PC7 2 &pcfg_pull_none_4ma>, /* mac_rxd1 */ - <1 RK_PB2 2 &pcfg_pull_none_2ma>, + <1 RK_PB2 2 &pcfg_pull_none_4ma>, /* mac_rxd0 */ - <1 RK_PB3 2 &pcfg_pull_none_2ma>, + <1 RK_PB3 2 &pcfg_pull_none_4ma>, /* mac_txd1 */ - <1 RK_PB0 2 &pcfg_pull_none_12ma>, + <1 RK_PB0 2 &pcfg_pull_none_8ma>, /* mac_txd0 */ - <1 RK_PB1 2 &pcfg_pull_none_12ma>, + <1 RK_PB1 2 &pcfg_pull_none_8ma>, /* mac_rxd3 */ - <1 RK_PB6 2 &pcfg_pull_none_2ma>, + <1 RK_PB6 2 &pcfg_pull_none_4ma>, /* mac_rxd2 */ - <1 RK_PB7 2 &pcfg_pull_none_2ma>, + <1 RK_PB7 2 &pcfg_pull_none_4ma>, /* mac_txd3 */ - <1 RK_PC0 2 &pcfg_pull_none_12ma>, + <1 RK_PC0 2 &pcfg_pull_none_8ma>, /* mac_txd2 */ - <1 RK_PC1 2 &pcfg_pull_none_12ma>, + <1 RK_PC1 2 &pcfg_pull_none_8ma>, /* mac_txclk */ - <0 RK_PB0 1 &pcfg_pull_none>, + <0 RK_PB0 1 &pcfg_pull_none_8ma>, /* mac_txen */ - <0 RK_PB4 1 &pcfg_pull_none>, + <0 RK_PB4 1 &pcfg_pull_none_8ma>, /* mac_clk */ - <0 RK_PD0 1 &pcfg_pull_none>, + <0 RK_PD0 1 &pcfg_pull_none_4ma>, /* mac_txd1 */ - <0 RK_PC0 1 &pcfg_pull_none>, + <0 RK_PC0 1 &pcfg_pull_none_8ma>, /* mac_txd0 */ - <0 RK_PC1 1 &pcfg_pull_none>, + <0 RK_PC1 1 &pcfg_pull_none_8ma>, /* mac_txd3 */ - <0 RK_PC7 1 &pcfg_pull_none>, + <0 RK_PC7 1 &pcfg_pull_none_8ma>, /* mac_txd2 */ - <0 RK_PC6 1 &pcfg_pull_none>; + <0 RK_PC6 1 &pcfg_pull_none_8ma>; }; rmiim1_pins: rmiim1-pins { diff --git a/arch/arm64/boot/dts/rockchip/rk3399-rock-pi-4.dts b/arch/arm64/boot/dts/rockchip/rk3399-rock-pi-4.dts index 4a543f2117d4..844eac939a97 100644 --- a/arch/arm64/boot/dts/rockchip/rk3399-rock-pi-4.dts +++ b/arch/arm64/boot/dts/rockchip/rk3399-rock-pi-4.dts @@ -158,6 +158,7 @@ }; &hdmi { + ddc-i2c-bus = <&i2c3>; pinctrl-names = "default"; pinctrl-0 = <&hdmi_cec>; status = "okay"; diff --git a/arch/arm64/include/asm/futex.h b/arch/arm64/include/asm/futex.h index cccb83ad7fa8..e1d95f08f8e1 100644 --- a/arch/arm64/include/asm/futex.h +++ b/arch/arm64/include/asm/futex.h @@ -30,8 +30,8 @@ do { \ " prfm pstl1strm, %2\n" \ "1: ldxr %w1, %2\n" \ insn "\n" \ -"2: stlxr %w3, %w0, %2\n" \ -" cbnz %w3, 1b\n" \ +"2: stlxr %w0, %w3, %2\n" \ +" cbnz %w0, 1b\n" \ " dmb ish\n" \ "3:\n" \ " .pushsection .fixup,\"ax\"\n" \ @@ -50,30 +50,30 @@ do { \ static inline int arch_futex_atomic_op_inuser(int op, int oparg, int *oval, u32 __user *_uaddr) { - int oldval = 0, ret, tmp; + int oldval, ret, tmp; u32 __user *uaddr = __uaccess_mask_ptr(_uaddr); pagefault_disable(); switch (op) { case FUTEX_OP_SET: - __futex_atomic_op("mov %w0, %w4", + __futex_atomic_op("mov %w3, %w4", ret, oldval, uaddr, tmp, oparg); break; case FUTEX_OP_ADD: - __futex_atomic_op("add %w0, %w1, %w4", + __futex_atomic_op("add %w3, %w1, %w4", ret, oldval, uaddr, tmp, oparg); break; case FUTEX_OP_OR: - __futex_atomic_op("orr %w0, %w1, %w4", + __futex_atomic_op("orr %w3, %w1, %w4", ret, oldval, uaddr, tmp, oparg); break; case FUTEX_OP_ANDN: - __futex_atomic_op("and %w0, %w1, %w4", + __futex_atomic_op("and %w3, %w1, %w4", ret, oldval, uaddr, tmp, ~oparg); break; case FUTEX_OP_XOR: - __futex_atomic_op("eor %w0, %w1, %w4", + __futex_atomic_op("eor %w3, %w1, %w4", ret, oldval, uaddr, tmp, oparg); break; default: diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h index b0742a16c6c9..ebeefcf835e8 100644 --- a/arch/arm64/include/asm/kvm_mmu.h +++ b/arch/arm64/include/asm/kvm_mmu.h @@ -445,6 +445,17 @@ static inline int kvm_read_guest_lock(struct kvm *kvm, return ret; } +static inline int kvm_write_guest_lock(struct kvm *kvm, gpa_t gpa, + const void *data, unsigned long len) +{ + int srcu_idx = srcu_read_lock(&kvm->srcu); + int ret = kvm_write_guest(kvm, gpa, data, len); + + srcu_read_unlock(&kvm->srcu, srcu_idx); + + return ret; +} + #ifdef CONFIG_KVM_INDIRECT_VECTORS /* * EL2 vectors can be mapped and rerouted in a number of ways, diff --git a/arch/arm64/include/asm/module.h b/arch/arm64/include/asm/module.h index 905e1bb0e7bd..cd9f4e9d04d3 100644 --- a/arch/arm64/include/asm/module.h +++ b/arch/arm64/include/asm/module.h @@ -73,4 +73,9 @@ static inline bool is_forbidden_offset_for_adrp(void *place) struct plt_entry get_plt_entry(u64 dst, void *pc); bool plt_entries_equal(const struct plt_entry *a, const struct plt_entry *b); +static inline bool plt_entry_is_initialized(const struct plt_entry *e) +{ + return e->adrp || e->add || e->br; +} + #endif /* __ASM_MODULE_H */ diff --git a/arch/arm64/include/asm/syscall.h b/arch/arm64/include/asm/syscall.h index ad8be16a39c9..a179df3674a1 100644 --- a/arch/arm64/include/asm/syscall.h +++ b/arch/arm64/include/asm/syscall.h @@ -65,52 +65,22 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { - if (n == 0) - return; + args[0] = regs->orig_x0; + args++; - if (i + n > SYSCALL_MAX_ARGS) { - unsigned long *args_bad = args + SYSCALL_MAX_ARGS - i; - unsigned int n_bad = n + i - SYSCALL_MAX_ARGS; - pr_warning("%s called with max args %d, handling only %d\n", - __func__, i + n, SYSCALL_MAX_ARGS); - memset(args_bad, 0, n_bad * sizeof(args[0])); - } - - if (i == 0) { - args[0] = regs->orig_x0; - args++; - i++; - n--; - } - - memcpy(args, ®s->regs[i], n * sizeof(args[0])); + memcpy(args, ®s->regs[1], 5 * sizeof(args[0])); } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { - if (n == 0) - return; + regs->orig_x0 = args[0]; + args++; - if (i + n > SYSCALL_MAX_ARGS) { - pr_warning("%s called with max args %d, handling only %d\n", - __func__, i + n, SYSCALL_MAX_ARGS); - n = SYSCALL_MAX_ARGS - i; - } - - if (i == 0) { - regs->orig_x0 = args[0]; - args++; - i++; - n--; - } - - memcpy(®s->regs[i], args, n * sizeof(args[0])); + memcpy(®s->regs[1], args, 5 * sizeof(args[0])); } /* diff --git a/arch/arm64/kernel/ftrace.c b/arch/arm64/kernel/ftrace.c index 8e4431a8821f..07b298120182 100644 --- a/arch/arm64/kernel/ftrace.c +++ b/arch/arm64/kernel/ftrace.c @@ -107,8 +107,7 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) trampoline = get_plt_entry(addr, mod->arch.ftrace_trampoline); if (!plt_entries_equal(mod->arch.ftrace_trampoline, &trampoline)) { - if (!plt_entries_equal(mod->arch.ftrace_trampoline, - &(struct plt_entry){})) { + if (plt_entry_is_initialized(mod->arch.ftrace_trampoline)) { pr_err("ftrace: far branches to multiple entry points unsupported inside a single module\n"); return -EINVAL; } diff --git a/arch/arm64/kernel/sdei.c b/arch/arm64/kernel/sdei.c index 5ba4465e44f0..ea94cf8f9dc6 100644 --- a/arch/arm64/kernel/sdei.c +++ b/arch/arm64/kernel/sdei.c @@ -94,6 +94,9 @@ static bool on_sdei_normal_stack(unsigned long sp, struct stack_info *info) unsigned long low = (unsigned long)raw_cpu_read(sdei_stack_normal_ptr); unsigned long high = low + SDEI_STACK_SIZE; + if (!low) + return false; + if (sp < low || sp >= high) return false; @@ -111,6 +114,9 @@ static bool on_sdei_critical_stack(unsigned long sp, struct stack_info *info) unsigned long low = (unsigned long)raw_cpu_read(sdei_stack_critical_ptr); unsigned long high = low + SDEI_STACK_SIZE; + if (!low) + return false; + if (sp < low || sp >= high) return false; diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c index f8482fe5a190..413d566405d1 100644 --- a/arch/arm64/kernel/setup.c +++ b/arch/arm64/kernel/setup.c @@ -217,7 +217,7 @@ static void __init request_standard_resources(void) num_standard_resources = memblock.memory.cnt; res_size = num_standard_resources * sizeof(*standard_resources); - standard_resources = memblock_alloc_low(res_size, SMP_CACHE_BYTES); + standard_resources = memblock_alloc(res_size, SMP_CACHE_BYTES); if (!standard_resources) panic("%s: Failed to allocate %zu bytes\n", __func__, res_size); diff --git a/arch/arm64/kernel/traps.c b/arch/arm64/kernel/traps.c index 8ad119c3f665..29755989f616 100644 --- a/arch/arm64/kernel/traps.c +++ b/arch/arm64/kernel/traps.c @@ -102,10 +102,16 @@ static void dump_instr(const char *lvl, struct pt_regs *regs) void dump_backtrace(struct pt_regs *regs, struct task_struct *tsk) { struct stackframe frame; - int skip; + int skip = 0; pr_debug("%s(regs = %p tsk = %p)\n", __func__, regs, tsk); + if (regs) { + if (user_mode(regs)) + return; + skip = 1; + } + if (!tsk) tsk = current; @@ -126,7 +132,6 @@ void dump_backtrace(struct pt_regs *regs, struct task_struct *tsk) frame.graph = 0; #endif - skip = !!regs; printk("Call trace:\n"); do { /* skip until specified stack frame */ @@ -176,15 +181,13 @@ static int __die(const char *str, int err, struct pt_regs *regs) return ret; print_modules(); - __show_regs(regs); pr_emerg("Process %.*s (pid: %d, stack limit = 0x%p)\n", TASK_COMM_LEN, tsk->comm, task_pid_nr(tsk), end_of_stack(tsk)); + show_regs(regs); - if (!user_mode(regs)) { - dump_backtrace(regs, tsk); + if (!user_mode(regs)) dump_instr(KERN_EMERG, regs); - } return ret; } diff --git a/arch/arm64/kvm/reset.c b/arch/arm64/kvm/reset.c index f16a5f8ff2b4..e2a0500cd7a2 100644 --- a/arch/arm64/kvm/reset.c +++ b/arch/arm64/kvm/reset.c @@ -123,6 +123,9 @@ int kvm_reset_vcpu(struct kvm_vcpu *vcpu) int ret = -EINVAL; bool loaded; + /* Reset PMU outside of the non-preemptible section */ + kvm_pmu_vcpu_reset(vcpu); + preempt_disable(); loaded = (vcpu->cpu != -1); if (loaded) @@ -170,9 +173,6 @@ int kvm_reset_vcpu(struct kvm_vcpu *vcpu) vcpu->arch.reset_state.reset = false; } - /* Reset PMU */ - kvm_pmu_vcpu_reset(vcpu); - /* Default workaround setup is enabled (if supported) */ if (kvm_arm_have_ssbd() == KVM_SSBD_KERNEL) vcpu->arch.workaround_flags |= VCPU_WORKAROUND_2_FLAG; diff --git a/arch/c6x/include/asm/Kbuild b/arch/c6x/include/asm/Kbuild index 63b4a1705182..249c9f6f26dc 100644 --- a/arch/c6x/include/asm/Kbuild +++ b/arch/c6x/include/asm/Kbuild @@ -19,6 +19,7 @@ generic-y += irq_work.h generic-y += kdebug.h generic-y += kmap_types.h generic-y += kprobes.h +generic-y += kvm_para.h generic-y += local.h generic-y += mcs_spinlock.h generic-y += mm-arch-hooks.h diff --git a/arch/c6x/include/asm/syscall.h b/arch/c6x/include/asm/syscall.h index ae2be315ee9c..15ba8599858e 100644 --- a/arch/c6x/include/asm/syscall.h +++ b/arch/c6x/include/asm/syscall.h @@ -46,78 +46,27 @@ static inline void syscall_set_return_value(struct task_struct *task, } static inline void syscall_get_arguments(struct task_struct *task, - struct pt_regs *regs, unsigned int i, - unsigned int n, unsigned long *args) + struct pt_regs *regs, + unsigned long *args) { - switch (i) { - case 0: - if (!n--) - break; - *args++ = regs->a4; - case 1: - if (!n--) - break; - *args++ = regs->b4; - case 2: - if (!n--) - break; - *args++ = regs->a6; - case 3: - if (!n--) - break; - *args++ = regs->b6; - case 4: - if (!n--) - break; - *args++ = regs->a8; - case 5: - if (!n--) - break; - *args++ = regs->b8; - case 6: - if (!n--) - break; - default: - BUG(); - } + *args++ = regs->a4; + *args++ = regs->b4; + *args++ = regs->a6; + *args++ = regs->b6; + *args++ = regs->a8; + *args = regs->b8; } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { - switch (i) { - case 0: - if (!n--) - break; - regs->a4 = *args++; - case 1: - if (!n--) - break; - regs->b4 = *args++; - case 2: - if (!n--) - break; - regs->a6 = *args++; - case 3: - if (!n--) - break; - regs->b6 = *args++; - case 4: - if (!n--) - break; - regs->a8 = *args++; - case 5: - if (!n--) - break; - regs->a9 = *args++; - case 6: - if (!n) - break; - default: - BUG(); - } + regs->a4 = *args++; + regs->b4 = *args++; + regs->a6 = *args++; + regs->b6 = *args++; + regs->a8 = *args++; + regs->a9 = *args; } #endif /* __ASM_C6X_SYSCALLS_H */ diff --git a/arch/c6x/include/uapi/asm/Kbuild b/arch/c6x/include/uapi/asm/Kbuild index 755bb11323d8..1c72f04ff75d 100644 --- a/arch/c6x/include/uapi/asm/Kbuild +++ b/arch/c6x/include/uapi/asm/Kbuild @@ -1,2 +1 @@ -generic-y += kvm_para.h generic-y += ucontext.h diff --git a/arch/csky/include/asm/syscall.h b/arch/csky/include/asm/syscall.h index d637445737b7..bda0a446c63e 100644 --- a/arch/csky/include/asm/syscall.h +++ b/arch/csky/include/asm/syscall.h @@ -43,30 +43,20 @@ syscall_set_return_value(struct task_struct *task, struct pt_regs *regs, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) + unsigned long *args) { - BUG_ON(i + n > 6); - if (i == 0) { - args[0] = regs->orig_a0; - args++; - i++; - n--; - } - memcpy(args, ®s->a1 + i * sizeof(regs->a1), n * sizeof(args[0])); + args[0] = regs->orig_a0; + args++; + memcpy(args, ®s->a1, 5 * sizeof(args[0])); } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) + const unsigned long *args) { - BUG_ON(i + n > 6); - if (i == 0) { - regs->orig_a0 = args[0]; - args++; - i++; - n--; - } - memcpy(®s->a1 + i * sizeof(regs->a1), args, n * sizeof(regs->a0)); + regs->orig_a0 = args[0]; + args++; + memcpy(®s->a1, args, 5 * sizeof(regs->a1)); } static inline int diff --git a/arch/h8300/include/asm/Kbuild b/arch/h8300/include/asm/Kbuild index 3e7c8ecf151e..e3dead402e5f 100644 --- a/arch/h8300/include/asm/Kbuild +++ b/arch/h8300/include/asm/Kbuild @@ -23,6 +23,7 @@ generic-y += irq_work.h generic-y += kdebug.h generic-y += kmap_types.h generic-y += kprobes.h +generic-y += kvm_para.h generic-y += linkage.h generic-y += local.h generic-y += local64.h diff --git a/arch/h8300/include/asm/syscall.h b/arch/h8300/include/asm/syscall.h index 924990401237..ddd483c6ca95 100644 --- a/arch/h8300/include/asm/syscall.h +++ b/arch/h8300/include/asm/syscall.h @@ -17,34 +17,14 @@ syscall_get_nr(struct task_struct *task, struct pt_regs *regs) static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) + unsigned long *args) { - BUG_ON(i + n > 6); - - while (n > 0) { - switch (i) { - case 0: - *args++ = regs->er1; - break; - case 1: - *args++ = regs->er2; - break; - case 2: - *args++ = regs->er3; - break; - case 3: - *args++ = regs->er4; - break; - case 4: - *args++ = regs->er5; - break; - case 5: - *args++ = regs->er6; - break; - } - i++; - n--; - } + *args++ = regs->er1; + *args++ = regs->er2; + *args++ = regs->er3; + *args++ = regs->er4; + *args++ = regs->er5; + *args = regs->er6; } diff --git a/arch/h8300/include/uapi/asm/Kbuild b/arch/h8300/include/uapi/asm/Kbuild index 755bb11323d8..1c72f04ff75d 100644 --- a/arch/h8300/include/uapi/asm/Kbuild +++ b/arch/h8300/include/uapi/asm/Kbuild @@ -1,2 +1 @@ -generic-y += kvm_para.h generic-y += ucontext.h diff --git a/arch/hexagon/include/asm/Kbuild b/arch/hexagon/include/asm/Kbuild index b25fd42aa0f4..d046e8ccdf78 100644 --- a/arch/hexagon/include/asm/Kbuild +++ b/arch/hexagon/include/asm/Kbuild @@ -19,6 +19,7 @@ generic-y += irq_work.h generic-y += kdebug.h generic-y += kmap_types.h generic-y += kprobes.h +generic-y += kvm_para.h generic-y += local.h generic-y += local64.h generic-y += mcs_spinlock.h diff --git a/arch/hexagon/include/asm/syscall.h b/arch/hexagon/include/asm/syscall.h index 4af9c7b6f13a..ae3a1e24fabd 100644 --- a/arch/hexagon/include/asm/syscall.h +++ b/arch/hexagon/include/asm/syscall.h @@ -37,10 +37,8 @@ static inline long syscall_get_nr(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { - BUG_ON(i + n > 6); - memcpy(args, &(®s->r00)[i], n * sizeof(args[0])); + memcpy(args, &(®s->r00)[0], 6 * sizeof(args[0])); } #endif diff --git a/arch/hexagon/include/uapi/asm/kvm_para.h b/arch/hexagon/include/uapi/asm/kvm_para.h deleted file mode 100644 index baacc4996d18..000000000000 --- a/arch/hexagon/include/uapi/asm/kvm_para.h +++ /dev/null @@ -1,2 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */ -#include diff --git a/arch/ia64/include/asm/Kbuild b/arch/ia64/include/asm/Kbuild index 43e21fe3499c..11f191689c9e 100644 --- a/arch/ia64/include/asm/Kbuild +++ b/arch/ia64/include/asm/Kbuild @@ -2,6 +2,7 @@ generated-y += syscall_table.h generic-y += compat.h generic-y += exec.h generic-y += irq_work.h +generic-y += kvm_para.h generic-y += mcs_spinlock.h generic-y += mm-arch-hooks.h generic-y += preempt.h diff --git a/arch/ia64/include/asm/syscall.h b/arch/ia64/include/asm/syscall.h index 1d0b875fec44..0d9e7fab4a79 100644 --- a/arch/ia64/include/asm/syscall.h +++ b/arch/ia64/include/asm/syscall.h @@ -59,26 +59,19 @@ static inline void syscall_set_return_value(struct task_struct *task, } extern void ia64_syscall_get_set_arguments(struct task_struct *task, - struct pt_regs *regs, unsigned int i, unsigned int n, - unsigned long *args, int rw); + struct pt_regs *regs, unsigned long *args, int rw); static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { - BUG_ON(i + n > 6); - - ia64_syscall_get_set_arguments(task, regs, i, n, args, 0); + ia64_syscall_get_set_arguments(task, regs, args, 0); } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { - BUG_ON(i + n > 6); - - ia64_syscall_get_set_arguments(task, regs, i, n, args, 1); + ia64_syscall_get_set_arguments(task, regs, args, 1); } static inline int syscall_get_arch(void) diff --git a/arch/ia64/include/uapi/asm/Kbuild b/arch/ia64/include/uapi/asm/Kbuild index 20018cb883a9..62a9522af51e 100644 --- a/arch/ia64/include/uapi/asm/Kbuild +++ b/arch/ia64/include/uapi/asm/Kbuild @@ -1,2 +1 @@ generated-y += unistd_64.h -generic-y += kvm_para.h diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c index 6d50ede0ed69..bf9c24d9ce84 100644 --- a/arch/ia64/kernel/ptrace.c +++ b/arch/ia64/kernel/ptrace.c @@ -2179,12 +2179,11 @@ static void syscall_get_set_args_cb(struct unw_frame_info *info, void *data) } void ia64_syscall_get_set_arguments(struct task_struct *task, - struct pt_regs *regs, unsigned int i, unsigned int n, - unsigned long *args, int rw) + struct pt_regs *regs, unsigned long *args, int rw) { struct syscall_get_set_args data = { - .i = i, - .n = n, + .i = 0, + .n = 6, .args = args, .regs = regs, .rw = rw, diff --git a/arch/m68k/include/asm/Kbuild b/arch/m68k/include/asm/Kbuild index 95f8f631c4df..2c359d9e80f6 100644 --- a/arch/m68k/include/asm/Kbuild +++ b/arch/m68k/include/asm/Kbuild @@ -13,6 +13,7 @@ generic-y += irq_work.h generic-y += kdebug.h generic-y += kmap_types.h generic-y += kprobes.h +generic-y += kvm_para.h generic-y += local.h generic-y += local64.h generic-y += mcs_spinlock.h diff --git a/arch/m68k/include/uapi/asm/Kbuild b/arch/m68k/include/uapi/asm/Kbuild index 8a7ad40be463..7417847dc438 100644 --- a/arch/m68k/include/uapi/asm/Kbuild +++ b/arch/m68k/include/uapi/asm/Kbuild @@ -1,2 +1 @@ generated-y += unistd_32.h -generic-y += kvm_para.h diff --git a/arch/microblaze/include/asm/Kbuild b/arch/microblaze/include/asm/Kbuild index 791cc8d54d0a..1a8285c3f693 100644 --- a/arch/microblaze/include/asm/Kbuild +++ b/arch/microblaze/include/asm/Kbuild @@ -17,6 +17,7 @@ generic-y += irq_work.h generic-y += kdebug.h generic-y += kmap_types.h generic-y += kprobes.h +generic-y += kvm_para.h generic-y += linkage.h generic-y += local.h generic-y += local64.h diff --git a/arch/microblaze/include/asm/syscall.h b/arch/microblaze/include/asm/syscall.h index 220decd605a4..833d3a53dab3 100644 --- a/arch/microblaze/include/asm/syscall.h +++ b/arch/microblaze/include/asm/syscall.h @@ -82,18 +82,22 @@ static inline void microblaze_set_syscall_arg(struct pt_regs *regs, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { + unsigned int i = 0; + unsigned int n = 6; + while (n--) *args++ = microblaze_get_syscall_arg(regs, i++); } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { + unsigned int i = 0; + unsigned int n = 6; + while (n--) microblaze_set_syscall_arg(regs, i++, *args++); } diff --git a/arch/microblaze/include/uapi/asm/Kbuild b/arch/microblaze/include/uapi/asm/Kbuild index 3ce84fbb2678..13f59631c576 100644 --- a/arch/microblaze/include/uapi/asm/Kbuild +++ b/arch/microblaze/include/uapi/asm/Kbuild @@ -1,3 +1,2 @@ generated-y += unistd_32.h -generic-y += kvm_para.h generic-y += ucontext.h diff --git a/arch/mips/configs/generic/board-ocelot.config b/arch/mips/configs/generic/board-ocelot.config index f607888d2483..184eb65a6ba7 100644 --- a/arch/mips/configs/generic/board-ocelot.config +++ b/arch/mips/configs/generic/board-ocelot.config @@ -1,6 +1,10 @@ # require CONFIG_CPU_MIPS32_R2=y CONFIG_LEGACY_BOARD_OCELOT=y +CONFIG_FIT_IMAGE_FDT_OCELOT=y + +CONFIG_BRIDGE=y +CONFIG_GENERIC_PHY=y CONFIG_MTD=y CONFIG_MTD_CMDLINE_PARTS=y @@ -19,6 +23,8 @@ CONFIG_SERIAL_8250_CONSOLE=y CONFIG_SERIAL_OF_PLATFORM=y CONFIG_NETDEVICES=y +CONFIG_NET_SWITCHDEV=y +CONFIG_NET_DSA=y CONFIG_MSCC_OCELOT_SWITCH=y CONFIG_MSCC_OCELOT_SWITCH_OCELOT=y CONFIG_MDIO_MSCC_MIIM=y @@ -35,6 +41,8 @@ CONFIG_SPI_DESIGNWARE=y CONFIG_SPI_DW_MMIO=y CONFIG_SPI_SPIDEV=y +CONFIG_PINCTRL_OCELOT=y + CONFIG_GPIO_SYSFS=y CONFIG_POWER_RESET=y diff --git a/arch/mips/include/asm/syscall.h b/arch/mips/include/asm/syscall.h index 6cf8ffb5367e..a2b4748655df 100644 --- a/arch/mips/include/asm/syscall.h +++ b/arch/mips/include/asm/syscall.h @@ -116,9 +116,10 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { + unsigned int i = 0; + unsigned int n = 6; int ret; /* O32 ABI syscall() */ diff --git a/arch/mips/kernel/kgdb.c b/arch/mips/kernel/kgdb.c index 6e574c02e4c3..ea781b29f7f1 100644 --- a/arch/mips/kernel/kgdb.c +++ b/arch/mips/kernel/kgdb.c @@ -33,6 +33,7 @@ #include #include #include +#include static struct hard_trap_info { unsigned char tt; /* Trap type code for MIPS R3xxx and R4xxx */ @@ -214,7 +215,7 @@ void kgdb_call_nmi_hook(void *ignored) old_fs = get_fs(); set_fs(KERNEL_DS); - kgdb_nmicallback(raw_smp_processor_id(), NULL); + kgdb_nmicallback(raw_smp_processor_id(), get_irq_regs()); set_fs(old_fs); } diff --git a/arch/mips/kernel/ptrace.c b/arch/mips/kernel/ptrace.c index 0057c910bc2f..3a62f80958e1 100644 --- a/arch/mips/kernel/ptrace.c +++ b/arch/mips/kernel/ptrace.c @@ -1419,7 +1419,7 @@ asmlinkage long syscall_trace_enter(struct pt_regs *regs, long syscall) sd.nr = syscall; sd.arch = syscall_get_arch(); - syscall_get_arguments(current, regs, 0, 6, args); + syscall_get_arguments(current, regs, args); for (i = 0; i < 6; i++) sd.args[i] = args[i]; sd.instruction_pointer = KSTK_EIP(current); diff --git a/arch/mips/sgi-ip27/ip27-irq.c b/arch/mips/sgi-ip27/ip27-irq.c index 710a59764b01..a32f843cdbe0 100644 --- a/arch/mips/sgi-ip27/ip27-irq.c +++ b/arch/mips/sgi-ip27/ip27-irq.c @@ -118,7 +118,6 @@ static void shutdown_bridge_irq(struct irq_data *d) { struct hub_irq_data *hd = irq_data_get_irq_chip_data(d); struct bridge_controller *bc; - int pin = hd->pin; if (!hd) return; @@ -126,7 +125,7 @@ static void shutdown_bridge_irq(struct irq_data *d) disable_hub_irq(d); bc = hd->bc; - bridge_clr(bc, b_int_enable, (1 << pin)); + bridge_clr(bc, b_int_enable, (1 << hd->pin)); bridge_read(bc, b_wid_tflush); } diff --git a/arch/nds32/include/asm/syscall.h b/arch/nds32/include/asm/syscall.h index f7e5e86765fe..671ebd357496 100644 --- a/arch/nds32/include/asm/syscall.h +++ b/arch/nds32/include/asm/syscall.h @@ -108,81 +108,41 @@ void syscall_set_return_value(struct task_struct *task, struct pt_regs *regs, * syscall_get_arguments - extract system call parameter values * @task: task of interest, must be blocked * @regs: task_pt_regs() of @task - * @i: argument index [0,5] - * @n: number of arguments; n+i must be [1,6]. * @args: array filled with argument values * - * Fetches @n arguments to the system call starting with the @i'th argument - * (from 0 through 5). Argument @i is stored in @args[0], and so on. - * An arch inline version is probably optimal when @i and @n are constants. + * Fetches 6 arguments to the system call (from 0 through 5). The first + * argument is stored in @args[0], and so on. * * It's only valid to call this when @task is stopped for tracing on * entry to a system call, due to %TIF_SYSCALL_TRACE or %TIF_SYSCALL_AUDIT. - * It's invalid to call this with @i + @n > 6; we only support system calls - * taking up to 6 arguments. */ #define SYSCALL_MAX_ARGS 6 void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) + unsigned long *args) { - if (n == 0) - return; - if (i + n > SYSCALL_MAX_ARGS) { - unsigned long *args_bad = args + SYSCALL_MAX_ARGS - i; - unsigned int n_bad = n + i - SYSCALL_MAX_ARGS; - pr_warning("%s called with max args %d, handling only %d\n", - __func__, i + n, SYSCALL_MAX_ARGS); - memset(args_bad, 0, n_bad * sizeof(args[0])); - memset(args_bad, 0, n_bad * sizeof(args[0])); - } - - if (i == 0) { - args[0] = regs->orig_r0; - args++; - i++; - n--; - } - - memcpy(args, ®s->uregs[0] + i, n * sizeof(args[0])); + args[0] = regs->orig_r0; + args++; + memcpy(args, ®s->uregs[0] + 1, 5 * sizeof(args[0])); } /** * syscall_set_arguments - change system call parameter value * @task: task of interest, must be in system call entry tracing * @regs: task_pt_regs() of @task - * @i: argument index [0,5] - * @n: number of arguments; n+i must be [1,6]. * @args: array of argument values to store * - * Changes @n arguments to the system call starting with the @i'th argument. - * Argument @i gets value @args[0], and so on. - * An arch inline version is probably optimal when @i and @n are constants. + * Changes 6 arguments to the system call. The first argument gets value + * @args[0], and so on. * * It's only valid to call this when @task is stopped for tracing on * entry to a system call, due to %TIF_SYSCALL_TRACE or %TIF_SYSCALL_AUDIT. - * It's invalid to call this with @i + @n > 6; we only support system calls - * taking up to 6 arguments. */ void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { - if (n == 0) - return; + regs->orig_r0 = args[0]; + args++; - if (i + n > SYSCALL_MAX_ARGS) { - pr_warn("%s called with max args %d, handling only %d\n", - __func__, i + n, SYSCALL_MAX_ARGS); - n = SYSCALL_MAX_ARGS - i; - } - - if (i == 0) { - regs->orig_r0 = args[0]; - args++; - i++; - n--; - } - - memcpy(®s->uregs[0] + i, args, n * sizeof(args[0])); + memcpy(®s->uregs[0] + 1, args, 5 * sizeof(args[0])); } #endif /* _ASM_NDS32_SYSCALL_H */ diff --git a/arch/nios2/include/asm/Kbuild b/arch/nios2/include/asm/Kbuild index 8fde4fa2c34f..88a667d12aaa 100644 --- a/arch/nios2/include/asm/Kbuild +++ b/arch/nios2/include/asm/Kbuild @@ -23,6 +23,7 @@ generic-y += irq_work.h generic-y += kdebug.h generic-y += kmap_types.h generic-y += kprobes.h +generic-y += kvm_para.h generic-y += local.h generic-y += mcs_spinlock.h generic-y += mm-arch-hooks.h diff --git a/arch/nios2/include/asm/syscall.h b/arch/nios2/include/asm/syscall.h index 9de220854c4a..d7624ed06efb 100644 --- a/arch/nios2/include/asm/syscall.h +++ b/arch/nios2/include/asm/syscall.h @@ -58,81 +58,25 @@ static inline void syscall_set_return_value(struct task_struct *task, } static inline void syscall_get_arguments(struct task_struct *task, - struct pt_regs *regs, unsigned int i, unsigned int n, - unsigned long *args) + struct pt_regs *regs, unsigned long *args) { - BUG_ON(i + n > 6); - - switch (i) { - case 0: - if (!n--) - break; - *args++ = regs->r4; - case 1: - if (!n--) - break; - *args++ = regs->r5; - case 2: - if (!n--) - break; - *args++ = regs->r6; - case 3: - if (!n--) - break; - *args++ = regs->r7; - case 4: - if (!n--) - break; - *args++ = regs->r8; - case 5: - if (!n--) - break; - *args++ = regs->r9; - case 6: - if (!n--) - break; - default: - BUG(); - } + *args++ = regs->r4; + *args++ = regs->r5; + *args++ = regs->r6; + *args++ = regs->r7; + *args++ = regs->r8; + *args = regs->r9; } static inline void syscall_set_arguments(struct task_struct *task, - struct pt_regs *regs, unsigned int i, unsigned int n, - const unsigned long *args) + struct pt_regs *regs, const unsigned long *args) { - BUG_ON(i + n > 6); - - switch (i) { - case 0: - if (!n--) - break; - regs->r4 = *args++; - case 1: - if (!n--) - break; - regs->r5 = *args++; - case 2: - if (!n--) - break; - regs->r6 = *args++; - case 3: - if (!n--) - break; - regs->r7 = *args++; - case 4: - if (!n--) - break; - regs->r8 = *args++; - case 5: - if (!n--) - break; - regs->r9 = *args++; - case 6: - if (!n) - break; - default: - BUG(); - } + regs->r4 = *args++; + regs->r5 = *args++; + regs->r6 = *args++; + regs->r7 = *args++; + regs->r8 = *args++; + regs->r9 = *args; } #endif diff --git a/arch/nios2/include/uapi/asm/Kbuild b/arch/nios2/include/uapi/asm/Kbuild index 755bb11323d8..1c72f04ff75d 100644 --- a/arch/nios2/include/uapi/asm/Kbuild +++ b/arch/nios2/include/uapi/asm/Kbuild @@ -1,2 +1 @@ -generic-y += kvm_para.h generic-y += ucontext.h diff --git a/arch/openrisc/include/asm/Kbuild b/arch/openrisc/include/asm/Kbuild index 5a73e2956ac4..22aa97136c01 100644 --- a/arch/openrisc/include/asm/Kbuild +++ b/arch/openrisc/include/asm/Kbuild @@ -20,6 +20,7 @@ generic-y += irq_work.h generic-y += kdebug.h generic-y += kmap_types.h generic-y += kprobes.h +generic-y += kvm_para.h generic-y += local.h generic-y += mcs_spinlock.h generic-y += mm-arch-hooks.h diff --git a/arch/openrisc/include/asm/syscall.h b/arch/openrisc/include/asm/syscall.h index 2db9f1cf0694..b4ff07c1baed 100644 --- a/arch/openrisc/include/asm/syscall.h +++ b/arch/openrisc/include/asm/syscall.h @@ -56,20 +56,16 @@ syscall_set_return_value(struct task_struct *task, struct pt_regs *regs, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) + unsigned long *args) { - BUG_ON(i + n > 6); - - memcpy(args, ®s->gpr[3 + i], n * sizeof(args[0])); + memcpy(args, ®s->gpr[3], 6 * sizeof(args[0])); } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) + const unsigned long *args) { - BUG_ON(i + n > 6); - - memcpy(®s->gpr[3 + i], args, n * sizeof(args[0])); + memcpy(®s->gpr[3], args, 6 * sizeof(args[0])); } static inline int syscall_get_arch(void) diff --git a/arch/openrisc/include/uapi/asm/Kbuild b/arch/openrisc/include/uapi/asm/Kbuild index 755bb11323d8..1c72f04ff75d 100644 --- a/arch/openrisc/include/uapi/asm/Kbuild +++ b/arch/openrisc/include/uapi/asm/Kbuild @@ -1,2 +1 @@ -generic-y += kvm_para.h generic-y += ucontext.h diff --git a/arch/parisc/include/asm/Kbuild b/arch/parisc/include/asm/Kbuild index 6f49e77d82a2..9bcd0c903dbb 100644 --- a/arch/parisc/include/asm/Kbuild +++ b/arch/parisc/include/asm/Kbuild @@ -11,6 +11,7 @@ generic-y += irq_regs.h generic-y += irq_work.h generic-y += kdebug.h generic-y += kprobes.h +generic-y += kvm_para.h generic-y += local.h generic-y += local64.h generic-y += mcs_spinlock.h diff --git a/arch/parisc/include/asm/ptrace.h b/arch/parisc/include/asm/ptrace.h index 2a27b275ab09..9ff033d261ab 100644 --- a/arch/parisc/include/asm/ptrace.h +++ b/arch/parisc/include/asm/ptrace.h @@ -22,13 +22,14 @@ unsigned long profile_pc(struct pt_regs *); static inline unsigned long regs_return_value(struct pt_regs *regs) { - return regs->gr[20]; + return regs->gr[28]; } static inline void instruction_pointer_set(struct pt_regs *regs, unsigned long val) { - regs->iaoq[0] = val; + regs->iaoq[0] = val; + regs->iaoq[1] = val + 4; } /* Query offset/name of register from its name/offset */ diff --git a/arch/parisc/include/asm/syscall.h b/arch/parisc/include/asm/syscall.h index 8bff1a58c97f..62a6d477fae0 100644 --- a/arch/parisc/include/asm/syscall.h +++ b/arch/parisc/include/asm/syscall.h @@ -18,29 +18,15 @@ static inline long syscall_get_nr(struct task_struct *tsk, } static inline void syscall_get_arguments(struct task_struct *tsk, - struct pt_regs *regs, unsigned int i, - unsigned int n, unsigned long *args) + struct pt_regs *regs, + unsigned long *args) { - BUG_ON(i); - - switch (n) { - case 6: - args[5] = regs->gr[21]; - case 5: - args[4] = regs->gr[22]; - case 4: - args[3] = regs->gr[23]; - case 3: - args[2] = regs->gr[24]; - case 2: - args[1] = regs->gr[25]; - case 1: - args[0] = regs->gr[26]; - case 0: - break; - default: - BUG(); - } + args[5] = regs->gr[21]; + args[4] = regs->gr[22]; + args[3] = regs->gr[23]; + args[2] = regs->gr[24]; + args[1] = regs->gr[25]; + args[0] = regs->gr[26]; } static inline long syscall_get_return_value(struct task_struct *task, diff --git a/arch/parisc/include/uapi/asm/Kbuild b/arch/parisc/include/uapi/asm/Kbuild index 22fdbd08cdc8..2bd5b392277c 100644 --- a/arch/parisc/include/uapi/asm/Kbuild +++ b/arch/parisc/include/uapi/asm/Kbuild @@ -1,3 +1,2 @@ generated-y += unistd_32.h generated-y += unistd_64.h -generic-y += kvm_para.h diff --git a/arch/parisc/kernel/process.c b/arch/parisc/kernel/process.c index eb39e7e380d7..841db71958cd 100644 --- a/arch/parisc/kernel/process.c +++ b/arch/parisc/kernel/process.c @@ -210,12 +210,6 @@ void __cpuidle arch_cpu_idle(void) static int __init parisc_idle_init(void) { - const char *marker; - - /* check QEMU/SeaBIOS marker in PAGE0 */ - marker = (char *) &PAGE0->pad0; - running_on_qemu = (memcmp(marker, "SeaBIOS", 8) == 0); - if (!running_on_qemu) cpu_idle_poll_ctrl(1); diff --git a/arch/parisc/kernel/setup.c b/arch/parisc/kernel/setup.c index 15dd9e21be7e..d908058d05c1 100644 --- a/arch/parisc/kernel/setup.c +++ b/arch/parisc/kernel/setup.c @@ -397,6 +397,9 @@ void __init start_parisc(void) int ret, cpunum; struct pdc_coproc_cfg coproc_cfg; + /* check QEMU/SeaBIOS marker in PAGE0 */ + running_on_qemu = (memcmp(&PAGE0->pad0, "SeaBIOS", 8) == 0); + cpunum = smp_processor_id(); init_cpu_topology(); diff --git a/arch/powerpc/include/asm/mmu.h b/arch/powerpc/include/asm/mmu.h index 598cdcdd1355..8ddd4a91bdc1 100644 --- a/arch/powerpc/include/asm/mmu.h +++ b/arch/powerpc/include/asm/mmu.h @@ -352,7 +352,7 @@ static inline bool strict_kernel_rwx_enabled(void) #if defined(CONFIG_SPARSEMEM_VMEMMAP) && defined(CONFIG_SPARSEMEM_EXTREME) && \ defined (CONFIG_PPC_64K_PAGES) #define MAX_PHYSMEM_BITS 51 -#elif defined(CONFIG_SPARSEMEM) +#elif defined(CONFIG_PPC64) #define MAX_PHYSMEM_BITS 46 #endif diff --git a/arch/powerpc/include/asm/ppc-opcode.h b/arch/powerpc/include/asm/ppc-opcode.h index c5698a523bb1..23f7ed796f38 100644 --- a/arch/powerpc/include/asm/ppc-opcode.h +++ b/arch/powerpc/include/asm/ppc-opcode.h @@ -302,6 +302,7 @@ /* Misc instructions for BPF compiler */ #define PPC_INST_LBZ 0x88000000 #define PPC_INST_LD 0xe8000000 +#define PPC_INST_LDX 0x7c00002a #define PPC_INST_LHZ 0xa0000000 #define PPC_INST_LWZ 0x80000000 #define PPC_INST_LHBRX 0x7c00062c @@ -309,6 +310,7 @@ #define PPC_INST_STB 0x98000000 #define PPC_INST_STH 0xb0000000 #define PPC_INST_STD 0xf8000000 +#define PPC_INST_STDX 0x7c00012a #define PPC_INST_STDU 0xf8000001 #define PPC_INST_STW 0x90000000 #define PPC_INST_STWU 0x94000000 diff --git a/arch/powerpc/include/asm/syscall.h b/arch/powerpc/include/asm/syscall.h index 1a0e7a8b1c81..1243045bad2d 100644 --- a/arch/powerpc/include/asm/syscall.h +++ b/arch/powerpc/include/asm/syscall.h @@ -65,22 +65,20 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { unsigned long val, mask = -1UL; - - BUG_ON(i + n > 6); + unsigned int n = 6; #ifdef CONFIG_COMPAT if (test_tsk_thread_flag(task, TIF_32BIT)) mask = 0xffffffff; #endif while (n--) { - if (n == 0 && i == 0) + if (n == 0) val = regs->orig_gpr3; else - val = regs->gpr[3 + i + n]; + val = regs->gpr[3 + n]; args[n] = val & mask; } @@ -88,15 +86,12 @@ static inline void syscall_get_arguments(struct task_struct *task, static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { - BUG_ON(i + n > 6); - memcpy(®s->gpr[3 + i], args, n * sizeof(args[0])); + memcpy(®s->gpr[3], args, 6 * sizeof(args[0])); /* Also copy the first argument into orig_gpr3 */ - if (i == 0 && n > 0) - regs->orig_gpr3 = args[0]; + regs->orig_gpr3 = args[0]; } static inline int syscall_get_arch(void) diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S index a5b8fbae56a0..9481a117e242 100644 --- a/arch/powerpc/kernel/exceptions-64s.S +++ b/arch/powerpc/kernel/exceptions-64s.S @@ -656,11 +656,17 @@ EXC_COMMON_BEGIN(data_access_slb_common) ld r4,PACA_EXSLB+EX_DAR(r13) std r4,_DAR(r1) addi r3,r1,STACK_FRAME_OVERHEAD +BEGIN_MMU_FTR_SECTION + /* HPT case, do SLB fault */ bl do_slb_fault cmpdi r3,0 bne- 1f b fast_exception_return 1: /* Error case */ +MMU_FTR_SECTION_ELSE + /* Radix case, access is outside page table range */ + li r3,-EFAULT +ALT_MMU_FTR_SECTION_END_IFCLR(MMU_FTR_TYPE_RADIX) std r3,RESULT(r1) bl save_nvgprs RECONCILE_IRQ_STATE(r10, r11) @@ -705,11 +711,17 @@ EXC_COMMON_BEGIN(instruction_access_slb_common) EXCEPTION_PROLOG_COMMON(0x480, PACA_EXSLB) ld r4,_NIP(r1) addi r3,r1,STACK_FRAME_OVERHEAD +BEGIN_MMU_FTR_SECTION + /* HPT case, do SLB fault */ bl do_slb_fault cmpdi r3,0 bne- 1f b fast_exception_return 1: /* Error case */ +MMU_FTR_SECTION_ELSE + /* Radix case, access is outside page table range */ + li r3,-EFAULT +ALT_MMU_FTR_SECTION_END_IFCLR(MMU_FTR_TYPE_RADIX) std r3,RESULT(r1) bl save_nvgprs RECONCILE_IRQ_STATE(r10, r11) diff --git a/arch/powerpc/kernel/head_32.S b/arch/powerpc/kernel/head_32.S index 48051c8977c5..e25b615e9f9e 100644 --- a/arch/powerpc/kernel/head_32.S +++ b/arch/powerpc/kernel/head_32.S @@ -851,10 +851,6 @@ __secondary_start: tophys(r4,r2) addi r4,r4,THREAD /* phys address of our thread_struct */ mtspr SPRN_SPRG_THREAD,r4 -#ifdef CONFIG_PPC_RTAS - li r3,0 - stw r3, RTAS_SP(r4) /* 0 => not in RTAS */ -#endif lis r4, (swapper_pg_dir - PAGE_OFFSET)@h ori r4, r4, (swapper_pg_dir - PAGE_OFFSET)@l mtspr SPRN_SPRG_PGDIR, r4 @@ -941,10 +937,6 @@ start_here: tophys(r4,r2) addi r4,r4,THREAD /* init task's THREAD */ mtspr SPRN_SPRG_THREAD,r4 -#ifdef CONFIG_PPC_RTAS - li r3,0 - stw r3, RTAS_SP(r4) /* 0 => not in RTAS */ -#endif lis r4, (swapper_pg_dir - PAGE_OFFSET)@h ori r4, r4, (swapper_pg_dir - PAGE_OFFSET)@l mtspr SPRN_SPRG_PGDIR, r4 diff --git a/arch/powerpc/kernel/kvm.c b/arch/powerpc/kernel/kvm.c index 683b5b3805bd..cd381e2291df 100644 --- a/arch/powerpc/kernel/kvm.c +++ b/arch/powerpc/kernel/kvm.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include @@ -712,6 +713,12 @@ static void kvm_use_magic_page(void) static __init void kvm_free_tmp(void) { + /* + * Inform kmemleak about the hole in the .bss section since the + * corresponding pages will be unmapped with DEBUG_PAGEALLOC=y. + */ + kmemleak_free_part(&kvm_tmp[kvm_tmp_index], + ARRAY_SIZE(kvm_tmp) - kvm_tmp_index); free_reserved_area(&kvm_tmp[kvm_tmp_index], &kvm_tmp[ARRAY_SIZE(kvm_tmp)], -1, NULL); } diff --git a/arch/powerpc/kernel/vdso32/gettimeofday.S b/arch/powerpc/kernel/vdso32/gettimeofday.S index 1e0bc5955a40..afd516b572f8 100644 --- a/arch/powerpc/kernel/vdso32/gettimeofday.S +++ b/arch/powerpc/kernel/vdso32/gettimeofday.S @@ -98,7 +98,7 @@ V_FUNCTION_BEGIN(__kernel_clock_gettime) * can be used, r7 contains NSEC_PER_SEC. */ - lwz r5,WTOM_CLOCK_SEC(r9) + lwz r5,(WTOM_CLOCK_SEC+LOPART)(r9) lwz r6,WTOM_CLOCK_NSEC(r9) /* We now have our offset in r5,r6. We create a fake dependency diff --git a/arch/powerpc/lib/memcmp_64.S b/arch/powerpc/lib/memcmp_64.S index 844d8e774492..b7f6f6e0b6e8 100644 --- a/arch/powerpc/lib/memcmp_64.S +++ b/arch/powerpc/lib/memcmp_64.S @@ -215,11 +215,20 @@ _GLOBAL_TOC(memcmp) beq .Lzero .Lcmp_rest_lt8bytes: - /* Here we have only less than 8 bytes to compare with. at least s1 - * Address is aligned with 8 bytes. - * The next double words are load and shift right with appropriate - * bits. + /* + * Here we have less than 8 bytes to compare. At least s1 is aligned to + * 8 bytes, but s2 may not be. We must make sure s2 + 7 doesn't cross a + * page boundary, otherwise we might read past the end of the buffer and + * trigger a page fault. We use 4K as the conservative minimum page + * size. If we detect that case we go to the byte-by-byte loop. + * + * Otherwise the next double word is loaded from s1 and s2, and shifted + * right to compare the appropriate bits. */ + clrldi r6,r4,(64-12) // r6 = r4 & 0xfff + cmpdi r6,0xff8 + bgt .Lshort + subfic r6,r5,8 slwi r6,r6,3 LD rA,0,r3 diff --git a/arch/powerpc/net/bpf_jit.h b/arch/powerpc/net/bpf_jit.h index 549e9490ff2a..dcac37745b05 100644 --- a/arch/powerpc/net/bpf_jit.h +++ b/arch/powerpc/net/bpf_jit.h @@ -51,6 +51,8 @@ #define PPC_LIS(r, i) PPC_ADDIS(r, 0, i) #define PPC_STD(r, base, i) EMIT(PPC_INST_STD | ___PPC_RS(r) | \ ___PPC_RA(base) | ((i) & 0xfffc)) +#define PPC_STDX(r, base, b) EMIT(PPC_INST_STDX | ___PPC_RS(r) | \ + ___PPC_RA(base) | ___PPC_RB(b)) #define PPC_STDU(r, base, i) EMIT(PPC_INST_STDU | ___PPC_RS(r) | \ ___PPC_RA(base) | ((i) & 0xfffc)) #define PPC_STW(r, base, i) EMIT(PPC_INST_STW | ___PPC_RS(r) | \ @@ -65,7 +67,9 @@ #define PPC_LBZ(r, base, i) EMIT(PPC_INST_LBZ | ___PPC_RT(r) | \ ___PPC_RA(base) | IMM_L(i)) #define PPC_LD(r, base, i) EMIT(PPC_INST_LD | ___PPC_RT(r) | \ - ___PPC_RA(base) | IMM_L(i)) + ___PPC_RA(base) | ((i) & 0xfffc)) +#define PPC_LDX(r, base, b) EMIT(PPC_INST_LDX | ___PPC_RT(r) | \ + ___PPC_RA(base) | ___PPC_RB(b)) #define PPC_LWZ(r, base, i) EMIT(PPC_INST_LWZ | ___PPC_RT(r) | \ ___PPC_RA(base) | IMM_L(i)) #define PPC_LHZ(r, base, i) EMIT(PPC_INST_LHZ | ___PPC_RT(r) | \ @@ -85,17 +89,6 @@ ___PPC_RA(a) | ___PPC_RB(b)) #define PPC_BPF_STDCX(s, a, b) EMIT(PPC_INST_STDCX | ___PPC_RS(s) | \ ___PPC_RA(a) | ___PPC_RB(b)) - -#ifdef CONFIG_PPC64 -#define PPC_BPF_LL(r, base, i) do { PPC_LD(r, base, i); } while(0) -#define PPC_BPF_STL(r, base, i) do { PPC_STD(r, base, i); } while(0) -#define PPC_BPF_STLU(r, base, i) do { PPC_STDU(r, base, i); } while(0) -#else -#define PPC_BPF_LL(r, base, i) do { PPC_LWZ(r, base, i); } while(0) -#define PPC_BPF_STL(r, base, i) do { PPC_STW(r, base, i); } while(0) -#define PPC_BPF_STLU(r, base, i) do { PPC_STWU(r, base, i); } while(0) -#endif - #define PPC_CMPWI(a, i) EMIT(PPC_INST_CMPWI | ___PPC_RA(a) | IMM_L(i)) #define PPC_CMPDI(a, i) EMIT(PPC_INST_CMPDI | ___PPC_RA(a) | IMM_L(i)) #define PPC_CMPW(a, b) EMIT(PPC_INST_CMPW | ___PPC_RA(a) | \ diff --git a/arch/powerpc/net/bpf_jit32.h b/arch/powerpc/net/bpf_jit32.h index dc50a8d4b3b9..21744d8aa053 100644 --- a/arch/powerpc/net/bpf_jit32.h +++ b/arch/powerpc/net/bpf_jit32.h @@ -122,6 +122,10 @@ DECLARE_LOAD_FUNC(sk_load_byte_msh); #define PPC_NTOHS_OFFS(r, base, i) PPC_LHZ_OFFS(r, base, i) #endif +#define PPC_BPF_LL(r, base, i) do { PPC_LWZ(r, base, i); } while(0) +#define PPC_BPF_STL(r, base, i) do { PPC_STW(r, base, i); } while(0) +#define PPC_BPF_STLU(r, base, i) do { PPC_STWU(r, base, i); } while(0) + #define SEEN_DATAREF 0x10000 /* might call external helpers */ #define SEEN_XREG 0x20000 /* X reg is used */ #define SEEN_MEM 0x40000 /* SEEN_MEM+(1< MAX_TAIL_CALL_CNT) * goto out; */ - PPC_LD(b2p[TMP_REG_1], 1, bpf_jit_stack_tailcallcnt(ctx)); + PPC_BPF_LL(b2p[TMP_REG_1], 1, bpf_jit_stack_tailcallcnt(ctx)); PPC_CMPLWI(b2p[TMP_REG_1], MAX_TAIL_CALL_CNT); PPC_BCC(COND_GT, out); @@ -265,7 +265,7 @@ static void bpf_jit_emit_tail_call(u32 *image, struct codegen_context *ctx, u32 /* prog = array->ptrs[index]; */ PPC_MULI(b2p[TMP_REG_1], b2p_index, 8); PPC_ADD(b2p[TMP_REG_1], b2p[TMP_REG_1], b2p_bpf_array); - PPC_LD(b2p[TMP_REG_1], b2p[TMP_REG_1], offsetof(struct bpf_array, ptrs)); + PPC_BPF_LL(b2p[TMP_REG_1], b2p[TMP_REG_1], offsetof(struct bpf_array, ptrs)); /* * if (prog == NULL) @@ -275,7 +275,7 @@ static void bpf_jit_emit_tail_call(u32 *image, struct codegen_context *ctx, u32 PPC_BCC(COND_EQ, out); /* goto *(prog->bpf_func + prologue_size); */ - PPC_LD(b2p[TMP_REG_1], b2p[TMP_REG_1], offsetof(struct bpf_prog, bpf_func)); + PPC_BPF_LL(b2p[TMP_REG_1], b2p[TMP_REG_1], offsetof(struct bpf_prog, bpf_func)); #ifdef PPC64_ELF_ABI_v1 /* skip past the function descriptor */ PPC_ADDI(b2p[TMP_REG_1], b2p[TMP_REG_1], @@ -606,7 +606,7 @@ bpf_alu32_trunc: * the instructions generated will remain the * same across all passes */ - PPC_STD(dst_reg, 1, bpf_jit_stack_local(ctx)); + PPC_BPF_STL(dst_reg, 1, bpf_jit_stack_local(ctx)); PPC_ADDI(b2p[TMP_REG_1], 1, bpf_jit_stack_local(ctx)); PPC_LDBRX(dst_reg, 0, b2p[TMP_REG_1]); break; @@ -662,7 +662,7 @@ emit_clear: PPC_LI32(b2p[TMP_REG_1], imm); src_reg = b2p[TMP_REG_1]; } - PPC_STD(src_reg, dst_reg, off); + PPC_BPF_STL(src_reg, dst_reg, off); break; /* @@ -709,7 +709,7 @@ emit_clear: break; /* dst = *(u64 *)(ul) (src + off) */ case BPF_LDX | BPF_MEM | BPF_DW: - PPC_LD(dst_reg, src_reg, off); + PPC_BPF_LL(dst_reg, src_reg, off); break; /* diff --git a/arch/powerpc/platforms/pseries/pseries_energy.c b/arch/powerpc/platforms/pseries/pseries_energy.c index 6ed22127391b..921f12182f3e 100644 --- a/arch/powerpc/platforms/pseries/pseries_energy.c +++ b/arch/powerpc/platforms/pseries/pseries_energy.c @@ -77,18 +77,27 @@ static u32 cpu_to_drc_index(int cpu) ret = drc.drc_index_start + (thread_index * drc.sequential_inc); } else { - const __be32 *indexes; - - indexes = of_get_property(dn, "ibm,drc-indexes", NULL); - if (indexes == NULL) - goto err_of_node_put; + u32 nr_drc_indexes, thread_drc_index; /* - * The first element indexes[0] is the number of drc_indexes - * returned in the list. Hence thread_index+1 will get the - * drc_index corresponding to core number thread_index. + * The first element of ibm,drc-indexes array is the + * number of drc_indexes returned in the list. Hence + * thread_index+1 will get the drc_index corresponding + * to core number thread_index. */ - ret = indexes[thread_index + 1]; + rc = of_property_read_u32_index(dn, "ibm,drc-indexes", + 0, &nr_drc_indexes); + if (rc) + goto err_of_node_put; + + WARN_ON_ONCE(thread_index > nr_drc_indexes); + rc = of_property_read_u32_index(dn, "ibm,drc-indexes", + thread_index + 1, + &thread_drc_index); + if (rc) + goto err_of_node_put; + + ret = thread_drc_index; } rc = 0; diff --git a/arch/powerpc/platforms/pseries/ras.c b/arch/powerpc/platforms/pseries/ras.c index d97d52772789..452dcfd7e5dd 100644 --- a/arch/powerpc/platforms/pseries/ras.c +++ b/arch/powerpc/platforms/pseries/ras.c @@ -550,6 +550,7 @@ static void pseries_print_mce_info(struct pt_regs *regs, "UE", "SLB", "ERAT", + "Unknown", "TLB", "D-Cache", "Unknown", diff --git a/arch/riscv/include/asm/fixmap.h b/arch/riscv/include/asm/fixmap.h index 57afe604b495..c207f6634b91 100644 --- a/arch/riscv/include/asm/fixmap.h +++ b/arch/riscv/include/asm/fixmap.h @@ -26,7 +26,7 @@ enum fixed_addresses { }; #define FIXADDR_SIZE (__end_of_fixed_addresses * PAGE_SIZE) -#define FIXADDR_TOP (PAGE_OFFSET) +#define FIXADDR_TOP (VMALLOC_START) #define FIXADDR_START (FIXADDR_TOP - FIXADDR_SIZE) #define FIXMAP_PAGE_IO PAGE_KERNEL diff --git a/arch/riscv/include/asm/syscall.h b/arch/riscv/include/asm/syscall.h index bba3da6ef157..a3d5273ded7c 100644 --- a/arch/riscv/include/asm/syscall.h +++ b/arch/riscv/include/asm/syscall.h @@ -72,32 +72,20 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { - BUG_ON(i + n > 6); - if (i == 0) { - args[0] = regs->orig_a0; - args++; - i++; - n--; - } - memcpy(args, ®s->a1 + i * sizeof(regs->a1), n * sizeof(args[0])); + args[0] = regs->orig_a0; + args++; + memcpy(args, ®s->a1, 5 * sizeof(args[0])); } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { - BUG_ON(i + n > 6); - if (i == 0) { - regs->orig_a0 = args[0]; - args++; - i++; - n--; - } - memcpy(®s->a1 + i * sizeof(regs->a1), args, n * sizeof(regs->a0)); + regs->orig_a0 = args[0]; + args++; + memcpy(®s->a1, args, 5 * sizeof(regs->a1)); } static inline int syscall_get_arch(void) diff --git a/arch/riscv/include/asm/uaccess.h b/arch/riscv/include/asm/uaccess.h index a00168b980d2..fb53a8089e76 100644 --- a/arch/riscv/include/asm/uaccess.h +++ b/arch/riscv/include/asm/uaccess.h @@ -300,7 +300,7 @@ do { \ " .balign 4\n" \ "4:\n" \ " li %0, %6\n" \ - " jump 2b, %1\n" \ + " jump 3b, %1\n" \ " .previous\n" \ " .section __ex_table,\"a\"\n" \ " .balign " RISCV_SZPTR "\n" \ diff --git a/arch/riscv/kernel/Makefile b/arch/riscv/kernel/Makefile index f13f7f276639..598568168d35 100644 --- a/arch/riscv/kernel/Makefile +++ b/arch/riscv/kernel/Makefile @@ -4,7 +4,6 @@ ifdef CONFIG_FTRACE CFLAGS_REMOVE_ftrace.o = -pg -CFLAGS_REMOVE_setup.o = -pg endif extra-y += head.o @@ -29,8 +28,6 @@ obj-y += vdso.o obj-y += cacheinfo.o obj-y += vdso/ -CFLAGS_setup.o := -mcmodel=medany - obj-$(CONFIG_FPU) += fpu.o obj-$(CONFIG_SMP) += smpboot.o obj-$(CONFIG_SMP) += smp.o diff --git a/arch/riscv/kernel/module.c b/arch/riscv/kernel/module.c index 7dd308129b40..2872edce894d 100644 --- a/arch/riscv/kernel/module.c +++ b/arch/riscv/kernel/module.c @@ -141,7 +141,7 @@ static int apply_r_riscv_hi20_rela(struct module *me, u32 *location, { s32 hi20; - if (IS_ENABLED(CMODEL_MEDLOW)) { + if (IS_ENABLED(CONFIG_CMODEL_MEDLOW)) { pr_err( "%s: target %016llx can not be addressed by the 32-bit offset from PC = %p\n", me->name, (long long)v, location); diff --git a/arch/riscv/kernel/setup.c b/arch/riscv/kernel/setup.c index ecb654f6a79e..540a331d1376 100644 --- a/arch/riscv/kernel/setup.c +++ b/arch/riscv/kernel/setup.c @@ -48,14 +48,6 @@ struct screen_info screen_info = { }; #endif -unsigned long va_pa_offset; -EXPORT_SYMBOL(va_pa_offset); -unsigned long pfn_base; -EXPORT_SYMBOL(pfn_base); - -unsigned long empty_zero_page[PAGE_SIZE / sizeof(unsigned long)] __page_aligned_bss; -EXPORT_SYMBOL(empty_zero_page); - /* The lucky hart to first increment this variable will boot the other cores */ atomic_t hart_lottery; unsigned long boot_cpu_hartid; diff --git a/arch/riscv/mm/Makefile b/arch/riscv/mm/Makefile index eb22ab49b3e0..b68aac701803 100644 --- a/arch/riscv/mm/Makefile +++ b/arch/riscv/mm/Makefile @@ -1,3 +1,9 @@ + +CFLAGS_init.o := -mcmodel=medany +ifdef CONFIG_FTRACE +CFLAGS_REMOVE_init.o = -pg +endif + obj-y += init.o obj-y += fault.o obj-y += extable.o diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c index b379a75ac6a6..5fd8c922e1c2 100644 --- a/arch/riscv/mm/init.c +++ b/arch/riscv/mm/init.c @@ -25,6 +25,10 @@ #include #include +unsigned long empty_zero_page[PAGE_SIZE / sizeof(unsigned long)] + __page_aligned_bss; +EXPORT_SYMBOL(empty_zero_page); + static void __init zone_sizes_init(void) { unsigned long max_zone_pfns[MAX_NR_ZONES] = { 0, }; @@ -143,6 +147,11 @@ void __init setup_bootmem(void) } } +unsigned long va_pa_offset; +EXPORT_SYMBOL(va_pa_offset); +unsigned long pfn_base; +EXPORT_SYMBOL(pfn_base); + pgd_t swapper_pg_dir[PTRS_PER_PGD] __page_aligned_bss; pgd_t trampoline_pg_dir[PTRS_PER_PGD] __initdata __aligned(PAGE_SIZE); @@ -172,6 +181,25 @@ void __set_fixmap(enum fixed_addresses idx, phys_addr_t phys, pgprot_t prot) } } +/* + * setup_vm() is called from head.S with MMU-off. + * + * Following requirements should be honoured for setup_vm() to work + * correctly: + * 1) It should use PC-relative addressing for accessing kernel symbols. + * To achieve this we always use GCC cmodel=medany. + * 2) The compiler instrumentation for FTRACE will not work for setup_vm() + * so disable compiler instrumentation when FTRACE is enabled. + * + * Currently, the above requirements are honoured by using custom CFLAGS + * for init.o in mm/Makefile. + */ + +#ifndef __riscv_cmodel_medany +#error "setup_vm() is called from head.S before relocate so it should " + "not use absolute addressing." +#endif + asmlinkage void __init setup_vm(void) { extern char _start; diff --git a/arch/s390/include/asm/ap.h b/arch/s390/include/asm/ap.h index 1a6a7092d942..e94a0a28b5eb 100644 --- a/arch/s390/include/asm/ap.h +++ b/arch/s390/include/asm/ap.h @@ -360,4 +360,15 @@ static inline struct ap_queue_status ap_dqap(ap_qid_t qid, return reg1; } +/* + * Interface to tell the AP bus code that a configuration + * change has happened. The bus code should at least do + * an ap bus resource rescan. + */ +#if IS_ENABLED(CONFIG_ZCRYPT) +void ap_bus_cfg_chg(void); +#else +static inline void ap_bus_cfg_chg(void){}; +#endif + #endif /* _ASM_S390_AP_H_ */ diff --git a/arch/s390/include/asm/elf.h b/arch/s390/include/asm/elf.h index 7d22a474a040..f74639a05f0f 100644 --- a/arch/s390/include/asm/elf.h +++ b/arch/s390/include/asm/elf.h @@ -252,11 +252,14 @@ do { \ /* * Cache aliasing on the latest machines calls for a mapping granularity - * of 512KB. For 64-bit processes use a 512KB alignment and a randomization - * of up to 1GB. For 31-bit processes the virtual address space is limited, - * use no alignment and limit the randomization to 8MB. + * of 512KB for the anonymous mapping base. For 64-bit processes use a + * 512KB alignment and a randomization of up to 1GB. For 31-bit processes + * the virtual address space is limited, use no alignment and limit the + * randomization to 8MB. + * For the additional randomization of the program break use 32MB for + * 64-bit and 8MB for 31-bit. */ -#define BRK_RND_MASK (is_compat_task() ? 0x7ffUL : 0x3ffffUL) +#define BRK_RND_MASK (is_compat_task() ? 0x7ffUL : 0x1fffUL) #define MMAP_RND_MASK (is_compat_task() ? 0x7ffUL : 0x3ff80UL) #define MMAP_ALIGN_MASK (is_compat_task() ? 0 : 0x7fUL) #define STACK_RND_MASK MMAP_RND_MASK diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h index cc0947e08b6f..5b9f10b1e55d 100644 --- a/arch/s390/include/asm/lowcore.h +++ b/arch/s390/include/asm/lowcore.h @@ -91,52 +91,53 @@ struct lowcore { __u64 hardirq_timer; /* 0x02e8 */ __u64 softirq_timer; /* 0x02f0 */ __u64 steal_timer; /* 0x02f8 */ - __u64 last_update_timer; /* 0x0300 */ - __u64 last_update_clock; /* 0x0308 */ - __u64 int_clock; /* 0x0310 */ - __u64 mcck_clock; /* 0x0318 */ - __u64 clock_comparator; /* 0x0320 */ - __u64 boot_clock[2]; /* 0x0328 */ + __u64 avg_steal_timer; /* 0x0300 */ + __u64 last_update_timer; /* 0x0308 */ + __u64 last_update_clock; /* 0x0310 */ + __u64 int_clock; /* 0x0318*/ + __u64 mcck_clock; /* 0x0320 */ + __u64 clock_comparator; /* 0x0328 */ + __u64 boot_clock[2]; /* 0x0330 */ /* Current process. */ - __u64 current_task; /* 0x0338 */ - __u64 kernel_stack; /* 0x0340 */ + __u64 current_task; /* 0x0340 */ + __u64 kernel_stack; /* 0x0348 */ /* Interrupt, DAT-off and restartstack. */ - __u64 async_stack; /* 0x0348 */ - __u64 nodat_stack; /* 0x0350 */ - __u64 restart_stack; /* 0x0358 */ + __u64 async_stack; /* 0x0350 */ + __u64 nodat_stack; /* 0x0358 */ + __u64 restart_stack; /* 0x0360 */ /* Restart function and parameter. */ - __u64 restart_fn; /* 0x0360 */ - __u64 restart_data; /* 0x0368 */ - __u64 restart_source; /* 0x0370 */ + __u64 restart_fn; /* 0x0368 */ + __u64 restart_data; /* 0x0370 */ + __u64 restart_source; /* 0x0378 */ /* Address space pointer. */ - __u64 kernel_asce; /* 0x0378 */ - __u64 user_asce; /* 0x0380 */ - __u64 vdso_asce; /* 0x0388 */ + __u64 kernel_asce; /* 0x0380 */ + __u64 user_asce; /* 0x0388 */ + __u64 vdso_asce; /* 0x0390 */ /* * The lpp and current_pid fields form a * 64-bit value that is set as program * parameter with the LPP instruction. */ - __u32 lpp; /* 0x0390 */ - __u32 current_pid; /* 0x0394 */ + __u32 lpp; /* 0x0398 */ + __u32 current_pid; /* 0x039c */ /* SMP info area */ - __u32 cpu_nr; /* 0x0398 */ - __u32 softirq_pending; /* 0x039c */ - __u32 preempt_count; /* 0x03a0 */ - __u32 spinlock_lockval; /* 0x03a4 */ - __u32 spinlock_index; /* 0x03a8 */ - __u32 fpu_flags; /* 0x03ac */ - __u64 percpu_offset; /* 0x03b0 */ - __u64 vdso_per_cpu_data; /* 0x03b8 */ - __u64 machine_flags; /* 0x03c0 */ - __u64 gmap; /* 0x03c8 */ - __u8 pad_0x03d0[0x0400-0x03d0]; /* 0x03d0 */ + __u32 cpu_nr; /* 0x03a0 */ + __u32 softirq_pending; /* 0x03a4 */ + __u32 preempt_count; /* 0x03a8 */ + __u32 spinlock_lockval; /* 0x03ac */ + __u32 spinlock_index; /* 0x03b0 */ + __u32 fpu_flags; /* 0x03b4 */ + __u64 percpu_offset; /* 0x03b8 */ + __u64 vdso_per_cpu_data; /* 0x03c0 */ + __u64 machine_flags; /* 0x03c8 */ + __u64 gmap; /* 0x03d0 */ + __u8 pad_0x03d8[0x0400-0x03d8]; /* 0x03d8 */ /* br %r1 trampoline */ __u16 br_r1_trampoline; /* 0x0400 */ diff --git a/arch/s390/include/asm/syscall.h b/arch/s390/include/asm/syscall.h index 96f9a9151fde..59c3e91f2cdb 100644 --- a/arch/s390/include/asm/syscall.h +++ b/arch/s390/include/asm/syscall.h @@ -56,40 +56,32 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { unsigned long mask = -1UL; + unsigned int n = 6; - /* - * No arguments for this syscall, there's nothing to do. - */ - if (!n) - return; - - BUG_ON(i + n > 6); #ifdef CONFIG_COMPAT if (test_tsk_thread_flag(task, TIF_31BIT)) mask = 0xffffffff; #endif while (n-- > 0) - if (i + n > 0) - args[n] = regs->gprs[2 + i + n] & mask; - if (i == 0) - args[0] = regs->orig_gpr2 & mask; + if (n > 0) + args[n] = regs->gprs[2 + n] & mask; + + args[0] = regs->orig_gpr2 & mask; } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { - BUG_ON(i + n > 6); + unsigned int n = 6; + while (n-- > 0) - if (i + n > 0) - regs->gprs[2 + i + n] = args[n]; - if (i == 0) - regs->orig_gpr2 = args[0]; + if (n > 0) + regs->gprs[2 + n] = args[n]; + regs->orig_gpr2 = args[0]; } static inline int syscall_get_arch(void) diff --git a/arch/s390/kernel/perf_cpum_cf_diag.c b/arch/s390/kernel/perf_cpum_cf_diag.c index c6fad208c2fa..b6854812d2ed 100644 --- a/arch/s390/kernel/perf_cpum_cf_diag.c +++ b/arch/s390/kernel/perf_cpum_cf_diag.c @@ -196,23 +196,30 @@ static void cf_diag_perf_event_destroy(struct perf_event *event) */ static int __hw_perf_event_init(struct perf_event *event) { - struct cpu_cf_events *cpuhw = this_cpu_ptr(&cpu_cf_events); struct perf_event_attr *attr = &event->attr; + struct cpu_cf_events *cpuhw; enum cpumf_ctr_set i; int err = 0; - debug_sprintf_event(cf_diag_dbg, 5, - "%s event %p cpu %d authorized %#x\n", __func__, - event, event->cpu, cpuhw->info.auth_ctl); + debug_sprintf_event(cf_diag_dbg, 5, "%s event %p cpu %d\n", __func__, + event, event->cpu); event->hw.config = attr->config; event->hw.config_base = 0; - local64_set(&event->count, 0); - /* Add all authorized counter sets to config_base */ + /* Add all authorized counter sets to config_base. The + * the hardware init function is either called per-cpu or just once + * for all CPUS (event->cpu == -1). This depends on the whether + * counting is started for all CPUs or on a per workload base where + * the perf event moves from one CPU to another CPU. + * Checking the authorization on any CPU is fine as the hardware + * applies the same authorization settings to all CPUs. + */ + cpuhw = &get_cpu_var(cpu_cf_events); for (i = CPUMF_CTR_SET_BASIC; i < CPUMF_CTR_SET_MAX; ++i) if (cpuhw->info.auth_ctl & cpumf_ctr_ctl[i]) event->hw.config_base |= cpumf_ctr_ctl[i]; + put_cpu_var(cpu_cf_events); /* No authorized counter sets, nothing to count/sample */ if (!event->hw.config_base) { diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index 3fe1c77c361b..bd197baf1dc3 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -266,7 +266,8 @@ static void pcpu_prepare_secondary(struct pcpu *pcpu, int cpu) lc->percpu_offset = __per_cpu_offset[cpu]; lc->kernel_asce = S390_lowcore.kernel_asce; lc->machine_flags = S390_lowcore.machine_flags; - lc->user_timer = lc->system_timer = lc->steal_timer = 0; + lc->user_timer = lc->system_timer = + lc->steal_timer = lc->avg_steal_timer = 0; __ctl_store(lc->cregs_save_area, 0, 15); save_access_regs((unsigned int *) lc->access_regs_save_area); memcpy(lc->stfle_fac_list, S390_lowcore.stfle_fac_list, diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c index 98f850e00008..a69a0911ed0e 100644 --- a/arch/s390/kernel/vtime.c +++ b/arch/s390/kernel/vtime.c @@ -124,7 +124,7 @@ static void account_system_index_scaled(struct task_struct *p, u64 cputime, */ static int do_account_vtime(struct task_struct *tsk) { - u64 timer, clock, user, guest, system, hardirq, softirq, steal; + u64 timer, clock, user, guest, system, hardirq, softirq; timer = S390_lowcore.last_update_timer; clock = S390_lowcore.last_update_clock; @@ -182,12 +182,6 @@ static int do_account_vtime(struct task_struct *tsk) if (softirq) account_system_index_scaled(tsk, softirq, CPUTIME_SOFTIRQ); - steal = S390_lowcore.steal_timer; - if ((s64) steal > 0) { - S390_lowcore.steal_timer = 0; - account_steal_time(cputime_to_nsecs(steal)); - } - return virt_timer_forward(user + guest + system + hardirq + softirq); } @@ -213,8 +207,19 @@ void vtime_task_switch(struct task_struct *prev) */ void vtime_flush(struct task_struct *tsk) { + u64 steal, avg_steal; + if (do_account_vtime(tsk)) virt_timer_expire(); + + steal = S390_lowcore.steal_timer; + avg_steal = S390_lowcore.avg_steal_timer / 2; + if ((s64) steal > 0) { + S390_lowcore.steal_timer = 0; + account_steal_time(steal); + avg_steal += steal; + } + S390_lowcore.avg_steal_timer = avg_steal; } /* diff --git a/arch/sh/boards/of-generic.c b/arch/sh/boards/of-generic.c index 958f46da3a79..d91065e81a4e 100644 --- a/arch/sh/boards/of-generic.c +++ b/arch/sh/boards/of-generic.c @@ -164,10 +164,10 @@ static struct sh_machine_vector __initmv sh_of_generic_mv = { struct sh_clk_ops; -void __init arch_init_clk_ops(struct sh_clk_ops **ops, int idx) +void __init __weak arch_init_clk_ops(struct sh_clk_ops **ops, int idx) { } -void __init plat_irq_setup(void) +void __init __weak plat_irq_setup(void) { } diff --git a/arch/sh/include/asm/Kbuild b/arch/sh/include/asm/Kbuild index a6ef3fee5f85..7bf2cb680d32 100644 --- a/arch/sh/include/asm/Kbuild +++ b/arch/sh/include/asm/Kbuild @@ -9,6 +9,7 @@ generic-y += emergency-restart.h generic-y += exec.h generic-y += irq_regs.h generic-y += irq_work.h +generic-y += kvm_para.h generic-y += local.h generic-y += local64.h generic-y += mcs_spinlock.h diff --git a/arch/sh/include/asm/syscall_32.h b/arch/sh/include/asm/syscall_32.h index 6e118799831c..8c9d7e5e5dcc 100644 --- a/arch/sh/include/asm/syscall_32.h +++ b/arch/sh/include/asm/syscall_32.h @@ -48,51 +48,28 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { - /* - * Do this simply for now. If we need to start supporting - * fetching arguments from arbitrary indices, this will need some - * extra logic. Presently there are no in-tree users that depend - * on this behaviour. - */ - BUG_ON(i); /* Argument pattern is: R4, R5, R6, R7, R0, R1 */ - switch (n) { - case 6: args[5] = regs->regs[1]; - case 5: args[4] = regs->regs[0]; - case 4: args[3] = regs->regs[7]; - case 3: args[2] = regs->regs[6]; - case 2: args[1] = regs->regs[5]; - case 1: args[0] = regs->regs[4]; - case 0: - break; - default: - BUG(); - } + args[5] = regs->regs[1]; + args[4] = regs->regs[0]; + args[3] = regs->regs[7]; + args[2] = regs->regs[6]; + args[1] = regs->regs[5]; + args[0] = regs->regs[4]; } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { - /* Same note as above applies */ - BUG_ON(i); - - switch (n) { - case 6: regs->regs[1] = args[5]; - case 5: regs->regs[0] = args[4]; - case 4: regs->regs[7] = args[3]; - case 3: regs->regs[6] = args[2]; - case 2: regs->regs[5] = args[1]; - case 1: regs->regs[4] = args[0]; - break; - default: - BUG(); - } + regs->regs[1] = args[5]; + regs->regs[0] = args[4]; + regs->regs[7] = args[3]; + regs->regs[6] = args[2]; + regs->regs[5] = args[1]; + regs->regs[4] = args[0]; } static inline int syscall_get_arch(void) diff --git a/arch/sh/include/asm/syscall_64.h b/arch/sh/include/asm/syscall_64.h index 43882580c7f9..22fad97da066 100644 --- a/arch/sh/include/asm/syscall_64.h +++ b/arch/sh/include/asm/syscall_64.h @@ -47,20 +47,16 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { - BUG_ON(i + n > 6); - memcpy(args, ®s->regs[2 + i], n * sizeof(args[0])); + memcpy(args, ®s->regs[2], 6 * sizeof(args[0])); } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { - BUG_ON(i + n > 6); - memcpy(®s->regs[2 + i], args, n * sizeof(args[0])); + memcpy(®s->regs[2], args, 6 * sizeof(args[0])); } static inline int syscall_get_arch(void) diff --git a/arch/sh/include/uapi/asm/Kbuild b/arch/sh/include/uapi/asm/Kbuild index ecfbd40924dd..b8812c74c1de 100644 --- a/arch/sh/include/uapi/asm/Kbuild +++ b/arch/sh/include/uapi/asm/Kbuild @@ -1,5 +1,4 @@ # SPDX-License-Identifier: GPL-2.0 generated-y += unistd_32.h -generic-y += kvm_para.h generic-y += ucontext.h diff --git a/arch/sparc/include/asm/Kbuild b/arch/sparc/include/asm/Kbuild index b82f64e28f55..a22cfd5c0ee8 100644 --- a/arch/sparc/include/asm/Kbuild +++ b/arch/sparc/include/asm/Kbuild @@ -9,6 +9,7 @@ generic-y += exec.h generic-y += export.h generic-y += irq_regs.h generic-y += irq_work.h +generic-y += kvm_para.h generic-y += linkage.h generic-y += local.h generic-y += local64.h diff --git a/arch/sparc/include/asm/syscall.h b/arch/sparc/include/asm/syscall.h index 053989e3f6a6..4d075434e816 100644 --- a/arch/sparc/include/asm/syscall.h +++ b/arch/sparc/include/asm/syscall.h @@ -96,11 +96,11 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { int zero_extend = 0; unsigned int j; + unsigned int n = 6; #ifdef CONFIG_SPARC64 if (test_tsk_thread_flag(task, TIF_32BIT)) @@ -108,7 +108,7 @@ static inline void syscall_get_arguments(struct task_struct *task, #endif for (j = 0; j < n; j++) { - unsigned long val = regs->u_regs[UREG_I0 + i + j]; + unsigned long val = regs->u_regs[UREG_I0 + j]; if (zero_extend) args[j] = (u32) val; @@ -119,13 +119,12 @@ static inline void syscall_get_arguments(struct task_struct *task, static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { - unsigned int j; + unsigned int i; - for (j = 0; j < n; j++) - regs->u_regs[UREG_I0 + i + j] = args[j]; + for (i = 0; i < 6; i++) + regs->u_regs[UREG_I0 + i] = args[i]; } static inline int syscall_get_arch(void) diff --git a/arch/sparc/include/uapi/asm/kvm_para.h b/arch/sparc/include/uapi/asm/kvm_para.h deleted file mode 100644 index baacc4996d18..000000000000 --- a/arch/sparc/include/uapi/asm/kvm_para.h +++ /dev/null @@ -1,2 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */ -#include diff --git a/arch/sparc/kernel/pci_sun4v.c b/arch/sparc/kernel/pci_sun4v.c index a8af6023c126..14b93c5564e3 100644 --- a/arch/sparc/kernel/pci_sun4v.c +++ b/arch/sparc/kernel/pci_sun4v.c @@ -73,6 +73,11 @@ static inline void iommu_batch_start(struct device *dev, unsigned long prot, uns p->npages = 0; } +static inline bool iommu_use_atu(struct iommu *iommu, u64 mask) +{ + return iommu->atu && mask > DMA_BIT_MASK(32); +} + /* Interrupts must be disabled. */ static long iommu_batch_flush(struct iommu_batch *p, u64 mask) { @@ -92,7 +97,7 @@ static long iommu_batch_flush(struct iommu_batch *p, u64 mask) prot &= (HV_PCI_MAP_ATTR_READ | HV_PCI_MAP_ATTR_WRITE); while (npages != 0) { - if (mask <= DMA_BIT_MASK(32) || !pbm->iommu->atu) { + if (!iommu_use_atu(pbm->iommu, mask)) { num = pci_sun4v_iommu_map(devhandle, HV_PCI_TSBID(0, entry), npages, @@ -179,7 +184,6 @@ static void *dma_4v_alloc_coherent(struct device *dev, size_t size, unsigned long flags, order, first_page, npages, n; unsigned long prot = 0; struct iommu *iommu; - struct atu *atu; struct iommu_map_table *tbl; struct page *page; void *ret; @@ -205,13 +209,11 @@ static void *dma_4v_alloc_coherent(struct device *dev, size_t size, memset((char *)first_page, 0, PAGE_SIZE << order); iommu = dev->archdata.iommu; - atu = iommu->atu; - mask = dev->coherent_dma_mask; - if (mask <= DMA_BIT_MASK(32) || !atu) + if (!iommu_use_atu(iommu, mask)) tbl = &iommu->tbl; else - tbl = &atu->tbl; + tbl = &iommu->atu->tbl; entry = iommu_tbl_range_alloc(dev, tbl, npages, NULL, (unsigned long)(-1), 0); @@ -333,7 +335,7 @@ static void dma_4v_free_coherent(struct device *dev, size_t size, void *cpu, atu = iommu->atu; devhandle = pbm->devhandle; - if (dvma <= DMA_BIT_MASK(32)) { + if (!iommu_use_atu(iommu, dvma)) { tbl = &iommu->tbl; iotsb_num = 0; /* we don't care for legacy iommu */ } else { @@ -374,7 +376,7 @@ static dma_addr_t dma_4v_map_page(struct device *dev, struct page *page, npages >>= IO_PAGE_SHIFT; mask = *dev->dma_mask; - if (mask <= DMA_BIT_MASK(32)) + if (!iommu_use_atu(iommu, mask)) tbl = &iommu->tbl; else tbl = &atu->tbl; @@ -510,7 +512,7 @@ static int dma_4v_map_sg(struct device *dev, struct scatterlist *sglist, IO_PAGE_SIZE) >> IO_PAGE_SHIFT; mask = *dev->dma_mask; - if (mask <= DMA_BIT_MASK(32)) + if (!iommu_use_atu(iommu, mask)) tbl = &iommu->tbl; else tbl = &atu->tbl; diff --git a/arch/um/include/asm/syscall-generic.h b/arch/um/include/asm/syscall-generic.h index 9fb9cf8cd39a..98e50c50c12e 100644 --- a/arch/um/include/asm/syscall-generic.h +++ b/arch/um/include/asm/syscall-generic.h @@ -53,84 +53,30 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { const struct uml_pt_regs *r = ®s->regs; - switch (i) { - case 0: - if (!n--) - break; - *args++ = UPT_SYSCALL_ARG1(r); - case 1: - if (!n--) - break; - *args++ = UPT_SYSCALL_ARG2(r); - case 2: - if (!n--) - break; - *args++ = UPT_SYSCALL_ARG3(r); - case 3: - if (!n--) - break; - *args++ = UPT_SYSCALL_ARG4(r); - case 4: - if (!n--) - break; - *args++ = UPT_SYSCALL_ARG5(r); - case 5: - if (!n--) - break; - *args++ = UPT_SYSCALL_ARG6(r); - case 6: - if (!n--) - break; - default: - BUG(); - break; - } + *args++ = UPT_SYSCALL_ARG1(r); + *args++ = UPT_SYSCALL_ARG2(r); + *args++ = UPT_SYSCALL_ARG3(r); + *args++ = UPT_SYSCALL_ARG4(r); + *args++ = UPT_SYSCALL_ARG5(r); + *args = UPT_SYSCALL_ARG6(r); } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { struct uml_pt_regs *r = ®s->regs; - switch (i) { - case 0: - if (!n--) - break; - UPT_SYSCALL_ARG1(r) = *args++; - case 1: - if (!n--) - break; - UPT_SYSCALL_ARG2(r) = *args++; - case 2: - if (!n--) - break; - UPT_SYSCALL_ARG3(r) = *args++; - case 3: - if (!n--) - break; - UPT_SYSCALL_ARG4(r) = *args++; - case 4: - if (!n--) - break; - UPT_SYSCALL_ARG5(r) = *args++; - case 5: - if (!n--) - break; - UPT_SYSCALL_ARG6(r) = *args++; - case 6: - if (!n--) - break; - default: - BUG(); - break; - } + UPT_SYSCALL_ARG1(r) = *args++; + UPT_SYSCALL_ARG2(r) = *args++; + UPT_SYSCALL_ARG3(r) = *args++; + UPT_SYSCALL_ARG4(r) = *args++; + UPT_SYSCALL_ARG5(r) = *args++; + UPT_SYSCALL_ARG6(r) = *args; } /* See arch/x86/um/asm/syscall.h for syscall_get_arch() definition. */ diff --git a/arch/unicore32/include/asm/Kbuild b/arch/unicore32/include/asm/Kbuild index 1d1544b6ca74..d77d953c04c1 100644 --- a/arch/unicore32/include/asm/Kbuild +++ b/arch/unicore32/include/asm/Kbuild @@ -18,6 +18,7 @@ generic-y += irq_work.h generic-y += kdebug.h generic-y += kmap_types.h generic-y += kprobes.h +generic-y += kvm_para.h generic-y += local.h generic-y += mcs_spinlock.h generic-y += mm-arch-hooks.h diff --git a/arch/unicore32/include/uapi/asm/Kbuild b/arch/unicore32/include/uapi/asm/Kbuild index 755bb11323d8..1c72f04ff75d 100644 --- a/arch/unicore32/include/uapi/asm/Kbuild +++ b/arch/unicore32/include/uapi/asm/Kbuild @@ -1,2 +1 @@ -generic-y += kvm_para.h generic-y += ucontext.h diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index c1f9b3cf437c..5ad92419be19 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -2217,14 +2217,8 @@ config RANDOMIZE_MEMORY_PHYSICAL_PADDING If unsure, leave at the default value. config HOTPLUG_CPU - bool "Support for hot-pluggable CPUs" + def_bool y depends on SMP - ---help--- - Say Y here to allow turning CPUs off and on. CPUs can be - controlled through /sys/devices/system/cpu. - ( Note: power management support will enable this option - automatically on SMP systems. ) - Say N if you want to disable CPU hotplug. config BOOTPARAM_HOTPLUG_CPU0 bool "Set default setting of cpu0_hotpluggable" diff --git a/arch/x86/Makefile b/arch/x86/Makefile index 2d8b9d8ca4f8..a587805c6687 100644 --- a/arch/x86/Makefile +++ b/arch/x86/Makefile @@ -219,8 +219,12 @@ ifdef CONFIG_RETPOLINE # Additionally, avoid generating expensive indirect jumps which # are subject to retpolines for small number of switch cases. # clang turns off jump table generation by default when under - # retpoline builds, however, gcc does not for x86. - KBUILD_CFLAGS += $(call cc-option,--param=case-values-threshold=20) + # retpoline builds, however, gcc does not for x86. This has + # only been fixed starting from gcc stable version 8.4.0 and + # onwards, but not for older ones. See gcc bug #86952. + ifndef CONFIG_CC_IS_CLANG + KBUILD_CFLAGS += $(call cc-option,-fno-jump-tables) + endif endif archscripts: scripts_basic diff --git a/arch/x86/boot/compressed/misc.h b/arch/x86/boot/compressed/misc.h index fd13655e0f9b..d2f184165934 100644 --- a/arch/x86/boot/compressed/misc.h +++ b/arch/x86/boot/compressed/misc.h @@ -120,8 +120,6 @@ static inline void console_init(void) void set_sev_encryption_mask(void); -#endif - /* acpi.c */ #ifdef CONFIG_ACPI acpi_physical_address get_rsdp_addr(void); @@ -135,3 +133,5 @@ int count_immovable_mem_regions(void); #else static inline int count_immovable_mem_regions(void) { return 0; } #endif + +#endif /* BOOT_COMPRESSED_MISC_H */ diff --git a/arch/x86/events/amd/core.c b/arch/x86/events/amd/core.c index 7d2d7c801dba..0ecfac84ba91 100644 --- a/arch/x86/events/amd/core.c +++ b/arch/x86/events/amd/core.c @@ -3,10 +3,14 @@ #include #include #include +#include #include +#include #include "../perf_event.h" +static DEFINE_PER_CPU(unsigned int, perf_nmi_counter); + static __initconst const u64 amd_hw_cache_event_ids [PERF_COUNT_HW_CACHE_MAX] [PERF_COUNT_HW_CACHE_OP_MAX] @@ -429,6 +433,132 @@ static void amd_pmu_cpu_dead(int cpu) } } +/* + * When a PMC counter overflows, an NMI is used to process the event and + * reset the counter. NMI latency can result in the counter being updated + * before the NMI can run, which can result in what appear to be spurious + * NMIs. This function is intended to wait for the NMI to run and reset + * the counter to avoid possible unhandled NMI messages. + */ +#define OVERFLOW_WAIT_COUNT 50 + +static void amd_pmu_wait_on_overflow(int idx) +{ + unsigned int i; + u64 counter; + + /* + * Wait for the counter to be reset if it has overflowed. This loop + * should exit very, very quickly, but just in case, don't wait + * forever... + */ + for (i = 0; i < OVERFLOW_WAIT_COUNT; i++) { + rdmsrl(x86_pmu_event_addr(idx), counter); + if (counter & (1ULL << (x86_pmu.cntval_bits - 1))) + break; + + /* Might be in IRQ context, so can't sleep */ + udelay(1); + } +} + +static void amd_pmu_disable_all(void) +{ + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); + int idx; + + x86_pmu_disable_all(); + + /* + * This shouldn't be called from NMI context, but add a safeguard here + * to return, since if we're in NMI context we can't wait for an NMI + * to reset an overflowed counter value. + */ + if (in_nmi()) + return; + + /* + * Check each counter for overflow and wait for it to be reset by the + * NMI if it has overflowed. This relies on the fact that all active + * counters are always enabled when this function is caled and + * ARCH_PERFMON_EVENTSEL_INT is always set. + */ + for (idx = 0; idx < x86_pmu.num_counters; idx++) { + if (!test_bit(idx, cpuc->active_mask)) + continue; + + amd_pmu_wait_on_overflow(idx); + } +} + +static void amd_pmu_disable_event(struct perf_event *event) +{ + x86_pmu_disable_event(event); + + /* + * This can be called from NMI context (via x86_pmu_stop). The counter + * may have overflowed, but either way, we'll never see it get reset + * by the NMI if we're already in the NMI. And the NMI latency support + * below will take care of any pending NMI that might have been + * generated by the overflow. + */ + if (in_nmi()) + return; + + amd_pmu_wait_on_overflow(event->hw.idx); +} + +/* + * Because of NMI latency, if multiple PMC counters are active or other sources + * of NMIs are received, the perf NMI handler can handle one or more overflowed + * PMC counters outside of the NMI associated with the PMC overflow. If the NMI + * doesn't arrive at the LAPIC in time to become a pending NMI, then the kernel + * back-to-back NMI support won't be active. This PMC handler needs to take into + * account that this can occur, otherwise this could result in unknown NMI + * messages being issued. Examples of this is PMC overflow while in the NMI + * handler when multiple PMCs are active or PMC overflow while handling some + * other source of an NMI. + * + * Attempt to mitigate this by using the number of active PMCs to determine + * whether to return NMI_HANDLED if the perf NMI handler did not handle/reset + * any PMCs. The per-CPU perf_nmi_counter variable is set to a minimum of the + * number of active PMCs or 2. The value of 2 is used in case an NMI does not + * arrive at the LAPIC in time to be collapsed into an already pending NMI. + */ +static int amd_pmu_handle_irq(struct pt_regs *regs) +{ + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); + int active, handled; + + /* + * Obtain the active count before calling x86_pmu_handle_irq() since + * it is possible that x86_pmu_handle_irq() may make a counter + * inactive (through x86_pmu_stop). + */ + active = __bitmap_weight(cpuc->active_mask, X86_PMC_IDX_MAX); + + /* Process any counter overflows */ + handled = x86_pmu_handle_irq(regs); + + /* + * If a counter was handled, record the number of possible remaining + * NMIs that can occur. + */ + if (handled) { + this_cpu_write(perf_nmi_counter, + min_t(unsigned int, 2, active)); + + return handled; + } + + if (!this_cpu_read(perf_nmi_counter)) + return NMI_DONE; + + this_cpu_dec(perf_nmi_counter); + + return NMI_HANDLED; +} + static struct event_constraint * amd_get_event_constraints(struct cpu_hw_events *cpuc, int idx, struct perf_event *event) @@ -621,11 +751,11 @@ static ssize_t amd_event_sysfs_show(char *page, u64 config) static __initconst const struct x86_pmu amd_pmu = { .name = "AMD", - .handle_irq = x86_pmu_handle_irq, - .disable_all = x86_pmu_disable_all, + .handle_irq = amd_pmu_handle_irq, + .disable_all = amd_pmu_disable_all, .enable_all = x86_pmu_enable_all, .enable = x86_pmu_enable_event, - .disable = x86_pmu_disable_event, + .disable = amd_pmu_disable_event, .hw_config = amd_pmu_hw_config, .schedule_events = x86_schedule_events, .eventsel = MSR_K7_EVNTSEL0, @@ -732,7 +862,7 @@ void amd_pmu_enable_virt(void) cpuc->perf_ctr_virt_mask = 0; /* Reload all events */ - x86_pmu_disable_all(); + amd_pmu_disable_all(); x86_pmu_enable_all(0); } EXPORT_SYMBOL_GPL(amd_pmu_enable_virt); @@ -750,7 +880,7 @@ void amd_pmu_disable_virt(void) cpuc->perf_ctr_virt_mask = AMD64_EVENTSEL_HOSTONLY; /* Reload all events */ - x86_pmu_disable_all(); + amd_pmu_disable_all(); x86_pmu_enable_all(0); } EXPORT_SYMBOL_GPL(amd_pmu_disable_virt); diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c index e2b1447192a8..81911e11a15d 100644 --- a/arch/x86/events/core.c +++ b/arch/x86/events/core.c @@ -1349,8 +1349,9 @@ void x86_pmu_stop(struct perf_event *event, int flags) struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); struct hw_perf_event *hwc = &event->hw; - if (__test_and_clear_bit(hwc->idx, cpuc->active_mask)) { + if (test_bit(hwc->idx, cpuc->active_mask)) { x86_pmu.disable(event); + __clear_bit(hwc->idx, cpuc->active_mask); cpuc->events[hwc->idx] = NULL; WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED); hwc->state |= PERF_HES_STOPPED; @@ -1447,16 +1448,8 @@ int x86_pmu_handle_irq(struct pt_regs *regs) apic_write(APIC_LVTPC, APIC_DM_NMI); for (idx = 0; idx < x86_pmu.num_counters; idx++) { - if (!test_bit(idx, cpuc->active_mask)) { - /* - * Though we deactivated the counter some cpus - * might still deliver spurious interrupts still - * in flight. Catch them: - */ - if (__test_and_clear_bit(idx, cpuc->running)) - handled++; + if (!test_bit(idx, cpuc->active_mask)) continue; - } event = cpuc->events[idx]; diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 8baa441d8000..f61dcbef20ff 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -3185,7 +3185,7 @@ static int intel_pmu_hw_config(struct perf_event *event) return ret; if (event->attr.precise_ip) { - if (!event->attr.freq) { + if (!(event->attr.freq || event->attr.wakeup_events)) { event->hw.flags |= PERF_X86_EVENT_AUTO_RELOAD; if (!(event->attr.sample_type & ~intel_pmu_large_pebs_flags(event))) @@ -3575,6 +3575,12 @@ static void intel_pmu_cpu_starting(int cpu) cpuc->lbr_sel = NULL; + if (x86_pmu.flags & PMU_FL_TFA) { + WARN_ON_ONCE(cpuc->tfa_shadow); + cpuc->tfa_shadow = ~0ULL; + intel_set_tfa(cpuc, false); + } + if (x86_pmu.version > 1) flip_smm_bit(&x86_pmu.attr_freeze_on_smi); diff --git a/arch/x86/include/asm/bitops.h b/arch/x86/include/asm/bitops.h index d153d570bb04..8e790ec219a5 100644 --- a/arch/x86/include/asm/bitops.h +++ b/arch/x86/include/asm/bitops.h @@ -36,16 +36,17 @@ * bit 0 is the LSB of addr; bit 32 is the LSB of (addr+1). */ -#define BITOP_ADDR(x) "+m" (*(volatile long *) (x)) +#define RLONG_ADDR(x) "m" (*(volatile long *) (x)) +#define WBYTE_ADDR(x) "+m" (*(volatile char *) (x)) -#define ADDR BITOP_ADDR(addr) +#define ADDR RLONG_ADDR(addr) /* * We do the locked ops that don't return the old value as * a mask operation on a byte. */ #define IS_IMMEDIATE(nr) (__builtin_constant_p(nr)) -#define CONST_MASK_ADDR(nr, addr) BITOP_ADDR((void *)(addr) + ((nr)>>3)) +#define CONST_MASK_ADDR(nr, addr) WBYTE_ADDR((void *)(addr) + ((nr)>>3)) #define CONST_MASK(nr) (1 << ((nr) & 7)) /** @@ -73,7 +74,7 @@ set_bit(long nr, volatile unsigned long *addr) : "memory"); } else { asm volatile(LOCK_PREFIX __ASM_SIZE(bts) " %1,%0" - : BITOP_ADDR(addr) : "Ir" (nr) : "memory"); + : : RLONG_ADDR(addr), "Ir" (nr) : "memory"); } } @@ -88,7 +89,7 @@ set_bit(long nr, volatile unsigned long *addr) */ static __always_inline void __set_bit(long nr, volatile unsigned long *addr) { - asm volatile(__ASM_SIZE(bts) " %1,%0" : ADDR : "Ir" (nr) : "memory"); + asm volatile(__ASM_SIZE(bts) " %1,%0" : : ADDR, "Ir" (nr) : "memory"); } /** @@ -110,8 +111,7 @@ clear_bit(long nr, volatile unsigned long *addr) : "iq" ((u8)~CONST_MASK(nr))); } else { asm volatile(LOCK_PREFIX __ASM_SIZE(btr) " %1,%0" - : BITOP_ADDR(addr) - : "Ir" (nr)); + : : RLONG_ADDR(addr), "Ir" (nr) : "memory"); } } @@ -131,7 +131,7 @@ static __always_inline void clear_bit_unlock(long nr, volatile unsigned long *ad static __always_inline void __clear_bit(long nr, volatile unsigned long *addr) { - asm volatile(__ASM_SIZE(btr) " %1,%0" : ADDR : "Ir" (nr)); + asm volatile(__ASM_SIZE(btr) " %1,%0" : : ADDR, "Ir" (nr) : "memory"); } static __always_inline bool clear_bit_unlock_is_negative_byte(long nr, volatile unsigned long *addr) @@ -139,7 +139,7 @@ static __always_inline bool clear_bit_unlock_is_negative_byte(long nr, volatile bool negative; asm volatile(LOCK_PREFIX "andb %2,%1" CC_SET(s) - : CC_OUT(s) (negative), ADDR + : CC_OUT(s) (negative), WBYTE_ADDR(addr) : "ir" ((char) ~(1 << nr)) : "memory"); return negative; } @@ -155,13 +155,9 @@ static __always_inline bool clear_bit_unlock_is_negative_byte(long nr, volatile * __clear_bit() is non-atomic and implies release semantics before the memory * operation. It can be used for an unlock if no other CPUs can concurrently * modify other bits in the word. - * - * No memory barrier is required here, because x86 cannot reorder stores past - * older loads. Same principle as spin_unlock. */ static __always_inline void __clear_bit_unlock(long nr, volatile unsigned long *addr) { - barrier(); __clear_bit(nr, addr); } @@ -176,7 +172,7 @@ static __always_inline void __clear_bit_unlock(long nr, volatile unsigned long * */ static __always_inline void __change_bit(long nr, volatile unsigned long *addr) { - asm volatile(__ASM_SIZE(btc) " %1,%0" : ADDR : "Ir" (nr)); + asm volatile(__ASM_SIZE(btc) " %1,%0" : : ADDR, "Ir" (nr) : "memory"); } /** @@ -196,8 +192,7 @@ static __always_inline void change_bit(long nr, volatile unsigned long *addr) : "iq" ((u8)CONST_MASK(nr))); } else { asm volatile(LOCK_PREFIX __ASM_SIZE(btc) " %1,%0" - : BITOP_ADDR(addr) - : "Ir" (nr)); + : : RLONG_ADDR(addr), "Ir" (nr) : "memory"); } } @@ -242,8 +237,8 @@ static __always_inline bool __test_and_set_bit(long nr, volatile unsigned long * asm(__ASM_SIZE(bts) " %2,%1" CC_SET(c) - : CC_OUT(c) (oldbit), ADDR - : "Ir" (nr)); + : CC_OUT(c) (oldbit) + : ADDR, "Ir" (nr) : "memory"); return oldbit; } @@ -282,8 +277,8 @@ static __always_inline bool __test_and_clear_bit(long nr, volatile unsigned long asm volatile(__ASM_SIZE(btr) " %2,%1" CC_SET(c) - : CC_OUT(c) (oldbit), ADDR - : "Ir" (nr)); + : CC_OUT(c) (oldbit) + : ADDR, "Ir" (nr) : "memory"); return oldbit; } @@ -294,8 +289,8 @@ static __always_inline bool __test_and_change_bit(long nr, volatile unsigned lon asm volatile(__ASM_SIZE(btc) " %2,%1" CC_SET(c) - : CC_OUT(c) (oldbit), ADDR - : "Ir" (nr) : "memory"); + : CC_OUT(c) (oldbit) + : ADDR, "Ir" (nr) : "memory"); return oldbit; } @@ -326,7 +321,7 @@ static __always_inline bool variable_test_bit(long nr, volatile const unsigned l asm volatile(__ASM_SIZE(bt) " %2,%1" CC_SET(c) : CC_OUT(c) (oldbit) - : "m" (*(unsigned long *)addr), "Ir" (nr)); + : "m" (*(unsigned long *)addr), "Ir" (nr) : "memory"); return oldbit; } diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h index ce95b8cbd229..0e56ff7e4848 100644 --- a/arch/x86/include/asm/cpufeature.h +++ b/arch/x86/include/asm/cpufeature.h @@ -112,8 +112,9 @@ extern const char * const x86_bug_flags[NBUGINTS*32]; test_cpu_cap(c, bit)) #define this_cpu_has(bit) \ - (__builtin_constant_p(bit) && REQUIRED_MASK_BIT_SET(bit) ? 1 : \ - x86_this_cpu_test_bit(bit, (unsigned long *)&cpu_info.x86_capability)) + (__builtin_constant_p(bit) && REQUIRED_MASK_BIT_SET(bit) ? 1 : \ + x86_this_cpu_test_bit(bit, \ + (unsigned long __percpu *)&cpu_info.x86_capability)) /* * This macro is for detection of features which need kernel diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index a5db4475e72d..159b5988292f 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -253,14 +253,14 @@ struct kvm_mmu_memory_cache { * kvm_memory_slot.arch.gfn_track which is 16 bits, so the role bits used * by indirect shadow page can not be more than 15 bits. * - * Currently, we used 14 bits that are @level, @cr4_pae, @quadrant, @access, + * Currently, we used 14 bits that are @level, @gpte_is_8_bytes, @quadrant, @access, * @nxe, @cr0_wp, @smep_andnot_wp and @smap_andnot_wp. */ union kvm_mmu_page_role { u32 word; struct { unsigned level:4; - unsigned cr4_pae:1; + unsigned gpte_is_8_bytes:1; unsigned quadrant:2; unsigned direct:1; unsigned access:3; @@ -350,6 +350,7 @@ struct kvm_mmu_page { }; struct kvm_pio_request { + unsigned long linear_rip; unsigned long count; int in; int port; @@ -568,6 +569,7 @@ struct kvm_vcpu_arch { bool tpr_access_reporting; u64 ia32_xss; u64 microcode_version; + u64 arch_capabilities; /* * Paging state of the vcpu @@ -1192,6 +1194,8 @@ struct kvm_x86_ops { int (*nested_enable_evmcs)(struct kvm_vcpu *vcpu, uint16_t *vmcs_version); uint16_t (*nested_get_evmcs_version)(struct kvm_vcpu *vcpu); + + bool (*need_emulation_on_page_fault)(struct kvm_vcpu *vcpu); }; struct kvm_arch_async_pf { @@ -1252,7 +1256,7 @@ void kvm_mmu_clear_dirty_pt_masked(struct kvm *kvm, gfn_t gfn_offset, unsigned long mask); void kvm_mmu_zap_all(struct kvm *kvm); void kvm_mmu_invalidate_mmio_sptes(struct kvm *kvm, u64 gen); -unsigned int kvm_mmu_calculate_mmu_pages(struct kvm *kvm); +unsigned int kvm_mmu_calculate_default_mmu_pages(struct kvm *kvm); void kvm_mmu_change_mmu_pages(struct kvm *kvm, unsigned int kvm_nr_mmu_pages); int load_pdptrs(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu, unsigned long cr3); diff --git a/arch/x86/include/asm/realmode.h b/arch/x86/include/asm/realmode.h index 63b3393bd98e..c53682303c9c 100644 --- a/arch/x86/include/asm/realmode.h +++ b/arch/x86/include/asm/realmode.h @@ -77,7 +77,11 @@ static inline size_t real_mode_size_needed(void) return ALIGN(real_mode_blob_end - real_mode_blob, PAGE_SIZE); } -void set_real_mode_mem(phys_addr_t mem, size_t size); +static inline void set_real_mode_mem(phys_addr_t mem) +{ + real_mode_header = (struct real_mode_header *) __va(mem); +} + void reserve_real_mode(void); #endif /* __ASSEMBLY__ */ diff --git a/arch/x86/include/asm/syscall.h b/arch/x86/include/asm/syscall.h index d653139857af..4c305471ec33 100644 --- a/arch/x86/include/asm/syscall.h +++ b/arch/x86/include/asm/syscall.h @@ -91,11 +91,9 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { - BUG_ON(i + n > 6); - memcpy(args, ®s->bx + i, n * sizeof(args[0])); + memcpy(args, ®s->bx, 6 * sizeof(args[0])); } static inline void syscall_set_arguments(struct task_struct *task, @@ -116,124 +114,50 @@ static inline int syscall_get_arch(void) static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { # ifdef CONFIG_IA32_EMULATION - if (task->thread_info.status & TS_COMPAT) - switch (i) { - case 0: - if (!n--) break; - *args++ = regs->bx; - case 1: - if (!n--) break; - *args++ = regs->cx; - case 2: - if (!n--) break; - *args++ = regs->dx; - case 3: - if (!n--) break; - *args++ = regs->si; - case 4: - if (!n--) break; - *args++ = regs->di; - case 5: - if (!n--) break; - *args++ = regs->bp; - case 6: - if (!n--) break; - default: - BUG(); - break; - } - else + if (task->thread_info.status & TS_COMPAT) { + *args++ = regs->bx; + *args++ = regs->cx; + *args++ = regs->dx; + *args++ = regs->si; + *args++ = regs->di; + *args = regs->bp; + } else # endif - switch (i) { - case 0: - if (!n--) break; - *args++ = regs->di; - case 1: - if (!n--) break; - *args++ = regs->si; - case 2: - if (!n--) break; - *args++ = regs->dx; - case 3: - if (!n--) break; - *args++ = regs->r10; - case 4: - if (!n--) break; - *args++ = regs->r8; - case 5: - if (!n--) break; - *args++ = regs->r9; - case 6: - if (!n--) break; - default: - BUG(); - break; - } + { + *args++ = regs->di; + *args++ = regs->si; + *args++ = regs->dx; + *args++ = regs->r10; + *args++ = regs->r8; + *args = regs->r9; + } } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { # ifdef CONFIG_IA32_EMULATION - if (task->thread_info.status & TS_COMPAT) - switch (i) { - case 0: - if (!n--) break; - regs->bx = *args++; - case 1: - if (!n--) break; - regs->cx = *args++; - case 2: - if (!n--) break; - regs->dx = *args++; - case 3: - if (!n--) break; - regs->si = *args++; - case 4: - if (!n--) break; - regs->di = *args++; - case 5: - if (!n--) break; - regs->bp = *args++; - case 6: - if (!n--) break; - default: - BUG(); - break; - } - else + if (task->thread_info.status & TS_COMPAT) { + regs->bx = *args++; + regs->cx = *args++; + regs->dx = *args++; + regs->si = *args++; + regs->di = *args++; + regs->bp = *args; + } else # endif - switch (i) { - case 0: - if (!n--) break; - regs->di = *args++; - case 1: - if (!n--) break; - regs->si = *args++; - case 2: - if (!n--) break; - regs->dx = *args++; - case 3: - if (!n--) break; - regs->r10 = *args++; - case 4: - if (!n--) break; - regs->r8 = *args++; - case 5: - if (!n--) break; - regs->r9 = *args++; - case 6: - if (!n--) break; - default: - BUG(); - break; - } + { + regs->di = *args++; + regs->si = *args++; + regs->dx = *args++; + regs->r10 = *args++; + regs->r8 = *args++; + regs->r9 = *args; + } } static inline int syscall_get_arch(void) diff --git a/arch/x86/include/asm/xen/hypercall.h b/arch/x86/include/asm/xen/hypercall.h index de6f0d59a24f..2863c2026655 100644 --- a/arch/x86/include/asm/xen/hypercall.h +++ b/arch/x86/include/asm/xen/hypercall.h @@ -206,6 +206,9 @@ xen_single_call(unsigned int call, __HYPERCALL_DECLS; __HYPERCALL_5ARG(a1, a2, a3, a4, a5); + if (call >= PAGE_SIZE / sizeof(hypercall_page[0])) + return -EINVAL; + asm volatile(CALL_NOSPEC : __HYPERCALL_5PARAM : [thunk_target] "a" (&hypercall_page[call]) diff --git a/arch/x86/kernel/cpu/resctrl/monitor.c b/arch/x86/kernel/cpu/resctrl/monitor.c index f33f11f69078..1573a0a6b525 100644 --- a/arch/x86/kernel/cpu/resctrl/monitor.c +++ b/arch/x86/kernel/cpu/resctrl/monitor.c @@ -501,11 +501,8 @@ out_unlock: void cqm_setup_limbo_handler(struct rdt_domain *dom, unsigned long delay_ms) { unsigned long delay = msecs_to_jiffies(delay_ms); - struct rdt_resource *r; int cpu; - r = &rdt_resources_all[RDT_RESOURCE_L3]; - cpu = cpumask_any(&dom->cpu_mask); dom->cqm_work_cpu = cpu; diff --git a/arch/x86/kernel/cpu/resctrl/rdtgroup.c b/arch/x86/kernel/cpu/resctrl/rdtgroup.c index 399601eda8e4..54b9eef3eea9 100644 --- a/arch/x86/kernel/cpu/resctrl/rdtgroup.c +++ b/arch/x86/kernel/cpu/resctrl/rdtgroup.c @@ -2039,14 +2039,14 @@ out: enum rdt_param { Opt_cdp, Opt_cdpl2, - Opt_mba_mpbs, + Opt_mba_mbps, nr__rdt_params }; static const struct fs_parameter_spec rdt_param_specs[] = { fsparam_flag("cdp", Opt_cdp), fsparam_flag("cdpl2", Opt_cdpl2), - fsparam_flag("mba_mpbs", Opt_mba_mpbs), + fsparam_flag("mba_MBps", Opt_mba_mbps), {} }; @@ -2072,7 +2072,7 @@ static int rdt_parse_param(struct fs_context *fc, struct fs_parameter *param) case Opt_cdpl2: ctx->enable_cdpl2 = true; return 0; - case Opt_mba_mpbs: + case Opt_mba_mbps: if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) return -EINVAL; ctx->enable_mba_mbps = true; diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c index 27c43525a05f..421899f6ad7b 100644 --- a/arch/x86/kvm/hyperv.c +++ b/arch/x86/kvm/hyperv.c @@ -526,7 +526,9 @@ static int stimer_set_config(struct kvm_vcpu_hv_stimer *stimer, u64 config, new_config.enable = 0; stimer->config.as_uint64 = new_config.as_uint64; - stimer_mark_pending(stimer, false); + if (stimer->config.enable) + stimer_mark_pending(stimer, false); + return 0; } @@ -542,7 +544,10 @@ static int stimer_set_count(struct kvm_vcpu_hv_stimer *stimer, u64 count, stimer->config.enable = 0; else if (stimer->config.auto_enable) stimer->config.enable = 1; - stimer_mark_pending(stimer, false); + + if (stimer->config.enable) + stimer_mark_pending(stimer, false); + return 0; } diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 7837ab001d80..eee455a8a612 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -182,7 +182,7 @@ struct kvm_shadow_walk_iterator { static const union kvm_mmu_page_role mmu_base_role_mask = { .cr0_wp = 1, - .cr4_pae = 1, + .gpte_is_8_bytes = 1, .nxe = 1, .smep_andnot_wp = 1, .smap_andnot_wp = 1, @@ -2205,6 +2205,7 @@ static bool kvm_mmu_prepare_zap_page(struct kvm *kvm, struct kvm_mmu_page *sp, static void kvm_mmu_commit_zap_page(struct kvm *kvm, struct list_head *invalid_list); + #define for_each_valid_sp(_kvm, _sp, _gfn) \ hlist_for_each_entry(_sp, \ &(_kvm)->arch.mmu_page_hash[kvm_page_table_hashfn(_gfn)], hash_link) \ @@ -2215,12 +2216,17 @@ static void kvm_mmu_commit_zap_page(struct kvm *kvm, for_each_valid_sp(_kvm, _sp, _gfn) \ if ((_sp)->gfn != (_gfn) || (_sp)->role.direct) {} else +static inline bool is_ept_sp(struct kvm_mmu_page *sp) +{ + return sp->role.cr0_wp && sp->role.smap_andnot_wp; +} + /* @sp->gfn should be write-protected at the call site */ static bool __kvm_sync_page(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp, struct list_head *invalid_list) { - if (sp->role.cr4_pae != !!is_pae(vcpu) - || vcpu->arch.mmu->sync_page(vcpu, sp) == 0) { + if ((!is_ept_sp(sp) && sp->role.gpte_is_8_bytes != !!is_pae(vcpu)) || + vcpu->arch.mmu->sync_page(vcpu, sp) == 0) { kvm_mmu_prepare_zap_page(vcpu->kvm, sp, invalid_list); return false; } @@ -2423,7 +2429,7 @@ static struct kvm_mmu_page *kvm_mmu_get_page(struct kvm_vcpu *vcpu, role.level = level; role.direct = direct; if (role.direct) - role.cr4_pae = 0; + role.gpte_is_8_bytes = true; role.access = access; if (!vcpu->arch.mmu->direct_map && vcpu->arch.mmu->root_level <= PT32_ROOT_LEVEL) { @@ -4794,7 +4800,6 @@ static union kvm_mmu_role kvm_calc_mmu_role_common(struct kvm_vcpu *vcpu, role.base.access = ACC_ALL; role.base.nxe = !!is_nx(vcpu); - role.base.cr4_pae = !!is_pae(vcpu); role.base.cr0_wp = is_write_protection(vcpu); role.base.smm = is_smm(vcpu); role.base.guest_mode = is_guest_mode(vcpu); @@ -4815,6 +4820,7 @@ kvm_calc_tdp_mmu_root_page_role(struct kvm_vcpu *vcpu, bool base_only) role.base.ad_disabled = (shadow_accessed_mask == 0); role.base.level = kvm_x86_ops->get_tdp_level(vcpu); role.base.direct = true; + role.base.gpte_is_8_bytes = true; return role; } @@ -4879,6 +4885,7 @@ kvm_calc_shadow_mmu_root_page_role(struct kvm_vcpu *vcpu, bool base_only) role.base.smap_andnot_wp = role.ext.cr4_smap && !is_write_protection(vcpu); role.base.direct = !is_paging(vcpu); + role.base.gpte_is_8_bytes = !!is_pae(vcpu); if (!is_long_mode(vcpu)) role.base.level = PT32E_ROOT_LEVEL; @@ -4918,18 +4925,26 @@ static union kvm_mmu_role kvm_calc_shadow_ept_root_page_role(struct kvm_vcpu *vcpu, bool accessed_dirty, bool execonly) { - union kvm_mmu_role role; + union kvm_mmu_role role = {0}; - /* Base role is inherited from root_mmu */ - role.base.word = vcpu->arch.root_mmu.mmu_role.base.word; - role.ext = kvm_calc_mmu_role_ext(vcpu); + /* SMM flag is inherited from root_mmu */ + role.base.smm = vcpu->arch.root_mmu.mmu_role.base.smm; role.base.level = PT64_ROOT_4LEVEL; + role.base.gpte_is_8_bytes = true; role.base.direct = false; role.base.ad_disabled = !accessed_dirty; role.base.guest_mode = true; role.base.access = ACC_ALL; + /* + * WP=1 and NOT_WP=1 is an impossible combination, use WP and the + * SMAP variation to denote shadow EPT entries. + */ + role.base.cr0_wp = true; + role.base.smap_andnot_wp = true; + + role.ext = kvm_calc_mmu_role_ext(vcpu); role.ext.execonly = execonly; return role; @@ -5179,7 +5194,7 @@ static bool detect_write_misaligned(struct kvm_mmu_page *sp, gpa_t gpa, gpa, bytes, sp->role.word); offset = offset_in_page(gpa); - pte_size = sp->role.cr4_pae ? 8 : 4; + pte_size = sp->role.gpte_is_8_bytes ? 8 : 4; /* * Sometimes, the OS only writes the last one bytes to update status @@ -5203,7 +5218,7 @@ static u64 *get_written_sptes(struct kvm_mmu_page *sp, gpa_t gpa, int *nspte) page_offset = offset_in_page(gpa); level = sp->role.level; *nspte = 1; - if (!sp->role.cr4_pae) { + if (!sp->role.gpte_is_8_bytes) { page_offset <<= 1; /* 32->64 */ /* * A 32-bit pde maps 4MB while the shadow pdes map @@ -5393,10 +5408,12 @@ emulate: * This can happen if a guest gets a page-fault on data access but the HW * table walker is not able to read the instruction page (e.g instruction * page is not present in memory). In those cases we simply restart the - * guest. + * guest, with the exception of AMD Erratum 1096 which is unrecoverable. */ - if (unlikely(insn && !insn_len)) - return 1; + if (unlikely(insn && !insn_len)) { + if (!kvm_x86_ops->need_emulation_on_page_fault(vcpu)) + return 1; + } er = x86_emulate_instruction(vcpu, cr2, emulation_type, insn, insn_len); @@ -5509,7 +5526,9 @@ slot_handle_level_range(struct kvm *kvm, struct kvm_memory_slot *memslot, if (need_resched() || spin_needbreak(&kvm->mmu_lock)) { if (flush && lock_flush_tlb) { - kvm_flush_remote_tlbs(kvm); + kvm_flush_remote_tlbs_with_address(kvm, + start_gfn, + iterator.gfn - start_gfn + 1); flush = false; } cond_resched_lock(&kvm->mmu_lock); @@ -5517,7 +5536,8 @@ slot_handle_level_range(struct kvm *kvm, struct kvm_memory_slot *memslot, } if (flush && lock_flush_tlb) { - kvm_flush_remote_tlbs(kvm); + kvm_flush_remote_tlbs_with_address(kvm, start_gfn, + end_gfn - start_gfn + 1); flush = false; } @@ -6011,7 +6031,7 @@ out: /* * Calculate mmu pages needed for kvm. */ -unsigned int kvm_mmu_calculate_mmu_pages(struct kvm *kvm) +unsigned int kvm_mmu_calculate_default_mmu_pages(struct kvm *kvm) { unsigned int nr_mmu_pages; unsigned int nr_pages = 0; diff --git a/arch/x86/kvm/mmutrace.h b/arch/x86/kvm/mmutrace.h index 9f6c855a0043..dd30dccd2ad5 100644 --- a/arch/x86/kvm/mmutrace.h +++ b/arch/x86/kvm/mmutrace.h @@ -29,10 +29,10 @@ \ role.word = __entry->role; \ \ - trace_seq_printf(p, "sp gfn %llx l%u%s q%u%s %s%s" \ + trace_seq_printf(p, "sp gfn %llx l%u %u-byte q%u%s %s%s" \ " %snxe %sad root %u %s%c", \ __entry->gfn, role.level, \ - role.cr4_pae ? " pae" : "", \ + role.gpte_is_8_bytes ? 8 : 4, \ role.quadrant, \ role.direct ? " direct" : "", \ access_str[role.access], \ diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index b5b128a0a051..e0a791c3d4fc 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -6422,11 +6422,11 @@ e_free: return ret; } -static int get_num_contig_pages(int idx, struct page **inpages, - unsigned long npages) +static unsigned long get_num_contig_pages(unsigned long idx, + struct page **inpages, unsigned long npages) { unsigned long paddr, next_paddr; - int i = idx + 1, pages = 1; + unsigned long i = idx + 1, pages = 1; /* find the number of contiguous pages starting from idx */ paddr = __sme_page_pa(inpages[idx]); @@ -6445,12 +6445,12 @@ static int get_num_contig_pages(int idx, struct page **inpages, static int sev_launch_update_data(struct kvm *kvm, struct kvm_sev_cmd *argp) { - unsigned long vaddr, vaddr_end, next_vaddr, npages, size; + unsigned long vaddr, vaddr_end, next_vaddr, npages, pages, size, i; struct kvm_sev_info *sev = &to_kvm_svm(kvm)->sev_info; struct kvm_sev_launch_update_data params; struct sev_data_launch_update_data *data; struct page **inpages; - int i, ret, pages; + int ret; if (!sev_guest(kvm)) return -ENOTTY; @@ -6799,7 +6799,8 @@ static int sev_dbg_crypt(struct kvm *kvm, struct kvm_sev_cmd *argp, bool dec) struct page **src_p, **dst_p; struct kvm_sev_dbg debug; unsigned long n; - int ret, size; + unsigned int size; + int ret; if (!sev_guest(kvm)) return -ENOTTY; @@ -6807,6 +6808,11 @@ static int sev_dbg_crypt(struct kvm *kvm, struct kvm_sev_cmd *argp, bool dec) if (copy_from_user(&debug, (void __user *)(uintptr_t)argp->data, sizeof(debug))) return -EFAULT; + if (!debug.len || debug.src_uaddr + debug.len < debug.src_uaddr) + return -EINVAL; + if (!debug.dst_uaddr) + return -EINVAL; + vaddr = debug.src_uaddr; size = debug.len; vaddr_end = vaddr + size; @@ -6857,8 +6863,8 @@ static int sev_dbg_crypt(struct kvm *kvm, struct kvm_sev_cmd *argp, bool dec) dst_vaddr, len, &argp->error); - sev_unpin_memory(kvm, src_p, 1); - sev_unpin_memory(kvm, dst_p, 1); + sev_unpin_memory(kvm, src_p, n); + sev_unpin_memory(kvm, dst_p, n); if (ret) goto err; @@ -7098,6 +7104,36 @@ static int nested_enable_evmcs(struct kvm_vcpu *vcpu, return -ENODEV; } +static bool svm_need_emulation_on_page_fault(struct kvm_vcpu *vcpu) +{ + bool is_user, smap; + + is_user = svm_get_cpl(vcpu) == 3; + smap = !kvm_read_cr4_bits(vcpu, X86_CR4_SMAP); + + /* + * Detect and workaround Errata 1096 Fam_17h_00_0Fh + * + * In non SEV guest, hypervisor will be able to read the guest + * memory to decode the instruction pointer when insn_len is zero + * so we return true to indicate that decoding is possible. + * + * But in the SEV guest, the guest memory is encrypted with the + * guest specific key and hypervisor will not be able to decode the + * instruction pointer so we will not able to workaround it. Lets + * print the error and request to kill the guest. + */ + if (is_user && smap) { + if (!sev_guest(vcpu->kvm)) + return true; + + pr_err_ratelimited("KVM: Guest triggered AMD Erratum 1096\n"); + kvm_make_request(KVM_REQ_TRIPLE_FAULT, vcpu); + } + + return false; +} + static struct kvm_x86_ops svm_x86_ops __ro_after_init = { .cpu_has_kvm_support = has_svm, .disabled_by_bios = is_disabled, @@ -7231,6 +7267,8 @@ static struct kvm_x86_ops svm_x86_ops __ro_after_init = { .nested_enable_evmcs = nested_enable_evmcs, .nested_get_evmcs_version = nested_get_evmcs_version, + + .need_emulation_on_page_fault = svm_need_emulation_on_page_fault, }; static int __init svm_init(void) diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c index f24a2c225070..7ec9bb1dd723 100644 --- a/arch/x86/kvm/vmx/nested.c +++ b/arch/x86/kvm/vmx/nested.c @@ -500,6 +500,17 @@ static void nested_vmx_disable_intercept_for_msr(unsigned long *msr_bitmap_l1, } } +static inline void enable_x2apic_msr_intercepts(unsigned long *msr_bitmap) { + int msr; + + for (msr = 0x800; msr <= 0x8ff; msr += BITS_PER_LONG) { + unsigned word = msr / BITS_PER_LONG; + + msr_bitmap[word] = ~0; + msr_bitmap[word + (0x800 / sizeof(long))] = ~0; + } +} + /* * Merge L0's and L1's MSR bitmap, return false to indicate that * we do not use the hardware. @@ -541,39 +552,44 @@ static inline bool nested_vmx_prepare_msr_bitmap(struct kvm_vcpu *vcpu, return false; msr_bitmap_l1 = (unsigned long *)kmap(page); - if (nested_cpu_has_apic_reg_virt(vmcs12)) { - /* - * L0 need not intercept reads for MSRs between 0x800 and 0x8ff, it - * just lets the processor take the value from the virtual-APIC page; - * take those 256 bits directly from the L1 bitmap. - */ - for (msr = 0x800; msr <= 0x8ff; msr += BITS_PER_LONG) { - unsigned word = msr / BITS_PER_LONG; - msr_bitmap_l0[word] = msr_bitmap_l1[word]; - msr_bitmap_l0[word + (0x800 / sizeof(long))] = ~0; - } - } else { - for (msr = 0x800; msr <= 0x8ff; msr += BITS_PER_LONG) { - unsigned word = msr / BITS_PER_LONG; - msr_bitmap_l0[word] = ~0; - msr_bitmap_l0[word + (0x800 / sizeof(long))] = ~0; - } - } - nested_vmx_disable_intercept_for_msr( - msr_bitmap_l1, msr_bitmap_l0, - X2APIC_MSR(APIC_TASKPRI), - MSR_TYPE_W); + /* + * To keep the control flow simple, pay eight 8-byte writes (sixteen + * 4-byte writes on 32-bit systems) up front to enable intercepts for + * the x2APIC MSR range and selectively disable them below. + */ + enable_x2apic_msr_intercepts(msr_bitmap_l0); + + if (nested_cpu_has_virt_x2apic_mode(vmcs12)) { + if (nested_cpu_has_apic_reg_virt(vmcs12)) { + /* + * L0 need not intercept reads for MSRs between 0x800 + * and 0x8ff, it just lets the processor take the value + * from the virtual-APIC page; take those 256 bits + * directly from the L1 bitmap. + */ + for (msr = 0x800; msr <= 0x8ff; msr += BITS_PER_LONG) { + unsigned word = msr / BITS_PER_LONG; + + msr_bitmap_l0[word] = msr_bitmap_l1[word]; + } + } - if (nested_cpu_has_vid(vmcs12)) { nested_vmx_disable_intercept_for_msr( msr_bitmap_l1, msr_bitmap_l0, - X2APIC_MSR(APIC_EOI), - MSR_TYPE_W); - nested_vmx_disable_intercept_for_msr( - msr_bitmap_l1, msr_bitmap_l0, - X2APIC_MSR(APIC_SELF_IPI), - MSR_TYPE_W); + X2APIC_MSR(APIC_TASKPRI), + MSR_TYPE_R | MSR_TYPE_W); + + if (nested_cpu_has_vid(vmcs12)) { + nested_vmx_disable_intercept_for_msr( + msr_bitmap_l1, msr_bitmap_l0, + X2APIC_MSR(APIC_EOI), + MSR_TYPE_W); + nested_vmx_disable_intercept_for_msr( + msr_bitmap_l1, msr_bitmap_l0, + X2APIC_MSR(APIC_SELF_IPI), + MSR_TYPE_W); + } } if (spec_ctrl) @@ -2585,6 +2601,11 @@ static int nested_check_host_control_regs(struct kvm_vcpu *vcpu, !nested_host_cr4_valid(vcpu, vmcs12->host_cr4) || !nested_cr3_valid(vcpu, vmcs12->host_cr3)) return -EINVAL; + + if (is_noncanonical_address(vmcs12->host_ia32_sysenter_esp, vcpu) || + is_noncanonical_address(vmcs12->host_ia32_sysenter_eip, vcpu)) + return -EINVAL; + /* * If the load IA32_EFER VM-exit control is 1, bits reserved in the * IA32_EFER MSR must be 0 in the field for that register. In addition, diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index c73375e01ab8..ab432a930ae8 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -1683,12 +1683,6 @@ static int vmx_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) msr_info->data = to_vmx(vcpu)->spec_ctrl; break; - case MSR_IA32_ARCH_CAPABILITIES: - if (!msr_info->host_initiated && - !guest_cpuid_has(vcpu, X86_FEATURE_ARCH_CAPABILITIES)) - return 1; - msr_info->data = to_vmx(vcpu)->arch_capabilities; - break; case MSR_IA32_SYSENTER_CS: msr_info->data = vmcs_read32(GUEST_SYSENTER_CS); break; @@ -1895,11 +1889,6 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) vmx_disable_intercept_for_msr(vmx->vmcs01.msr_bitmap, MSR_IA32_PRED_CMD, MSR_TYPE_W); break; - case MSR_IA32_ARCH_CAPABILITIES: - if (!msr_info->host_initiated) - return 1; - vmx->arch_capabilities = data; - break; case MSR_IA32_CR_PAT: if (vmcs_config.vmentry_ctrl & VM_ENTRY_LOAD_IA32_PAT) { if (!kvm_mtrr_valid(vcpu, MSR_IA32_CR_PAT, data)) @@ -4088,8 +4077,6 @@ static void vmx_vcpu_setup(struct vcpu_vmx *vmx) ++vmx->nmsrs; } - vmx->arch_capabilities = kvm_get_arch_capabilities(); - vm_exit_controls_init(vmx, vmx_vmexit_ctrl()); /* 22.2.1, 20.8.1 */ @@ -7409,6 +7396,11 @@ static int enable_smi_window(struct kvm_vcpu *vcpu) return 0; } +static bool vmx_need_emulation_on_page_fault(struct kvm_vcpu *vcpu) +{ + return 0; +} + static __init int hardware_setup(void) { unsigned long host_bndcfgs; @@ -7711,6 +7703,7 @@ static struct kvm_x86_ops vmx_x86_ops __ro_after_init = { .set_nested_state = NULL, .get_vmcs12_pages = NULL, .nested_enable_evmcs = NULL, + .need_emulation_on_page_fault = vmx_need_emulation_on_page_fault, }; static void vmx_cleanup_l1d_flush(void) diff --git a/arch/x86/kvm/vmx/vmx.h b/arch/x86/kvm/vmx/vmx.h index 1554cb45b393..a1e00d0a2482 100644 --- a/arch/x86/kvm/vmx/vmx.h +++ b/arch/x86/kvm/vmx/vmx.h @@ -190,7 +190,6 @@ struct vcpu_vmx { u64 msr_guest_kernel_gs_base; #endif - u64 arch_capabilities; u64 spec_ctrl; u32 vm_entry_controls_shadow; diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 65e4559eef2f..099b851dabaf 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1125,7 +1125,7 @@ static u32 msrs_to_save[] = { #endif MSR_IA32_TSC, MSR_IA32_CR_PAT, MSR_VM_HSAVE_PA, MSR_IA32_FEATURE_CONTROL, MSR_IA32_BNDCFGS, MSR_TSC_AUX, - MSR_IA32_SPEC_CTRL, MSR_IA32_ARCH_CAPABILITIES, + MSR_IA32_SPEC_CTRL, MSR_IA32_RTIT_CTL, MSR_IA32_RTIT_STATUS, MSR_IA32_RTIT_CR3_MATCH, MSR_IA32_RTIT_OUTPUT_BASE, MSR_IA32_RTIT_OUTPUT_MASK, MSR_IA32_RTIT_ADDR0_A, MSR_IA32_RTIT_ADDR0_B, @@ -1158,6 +1158,7 @@ static u32 emulated_msrs[] = { MSR_IA32_TSC_ADJUST, MSR_IA32_TSCDEADLINE, + MSR_IA32_ARCH_CAPABILITIES, MSR_IA32_MISC_ENABLE, MSR_IA32_MCG_STATUS, MSR_IA32_MCG_CTL, @@ -2443,6 +2444,11 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info) if (msr_info->host_initiated) vcpu->arch.microcode_version = data; break; + case MSR_IA32_ARCH_CAPABILITIES: + if (!msr_info->host_initiated) + return 1; + vcpu->arch.arch_capabilities = data; + break; case MSR_EFER: return set_efer(vcpu, data); case MSR_K7_HWCR: @@ -2747,6 +2753,12 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info) case MSR_IA32_UCODE_REV: msr_info->data = vcpu->arch.microcode_version; break; + case MSR_IA32_ARCH_CAPABILITIES: + if (!msr_info->host_initiated && + !guest_cpuid_has(vcpu, X86_FEATURE_ARCH_CAPABILITIES)) + return 1; + msr_info->data = vcpu->arch.arch_capabilities; + break; case MSR_IA32_TSC: msr_info->data = kvm_scale_tsc(vcpu, rdtsc()) + vcpu->arch.tsc_offset; break; @@ -6523,14 +6535,27 @@ int kvm_emulate_instruction_from_buffer(struct kvm_vcpu *vcpu, } EXPORT_SYMBOL_GPL(kvm_emulate_instruction_from_buffer); +static int complete_fast_pio_out(struct kvm_vcpu *vcpu) +{ + vcpu->arch.pio.count = 0; + + if (unlikely(!kvm_is_linear_rip(vcpu, vcpu->arch.pio.linear_rip))) + return 1; + + return kvm_skip_emulated_instruction(vcpu); +} + static int kvm_fast_pio_out(struct kvm_vcpu *vcpu, int size, unsigned short port) { unsigned long val = kvm_register_read(vcpu, VCPU_REGS_RAX); int ret = emulator_pio_out_emulated(&vcpu->arch.emulate_ctxt, size, port, &val, 1); - /* do not return to emulator after return from userspace */ - vcpu->arch.pio.count = 0; + + if (!ret) { + vcpu->arch.pio.linear_rip = kvm_get_linear_rip(vcpu); + vcpu->arch.complete_userspace_io = complete_fast_pio_out; + } return ret; } @@ -6541,6 +6566,11 @@ static int complete_fast_pio_in(struct kvm_vcpu *vcpu) /* We should only ever be called with arch.pio.count equal to 1 */ BUG_ON(vcpu->arch.pio.count != 1); + if (unlikely(!kvm_is_linear_rip(vcpu, vcpu->arch.pio.linear_rip))) { + vcpu->arch.pio.count = 0; + return 1; + } + /* For size less than 4 we merge, else we zero extend */ val = (vcpu->arch.pio.size < 4) ? kvm_register_read(vcpu, VCPU_REGS_RAX) : 0; @@ -6553,7 +6583,7 @@ static int complete_fast_pio_in(struct kvm_vcpu *vcpu) vcpu->arch.pio.port, &val, 1); kvm_register_write(vcpu, VCPU_REGS_RAX, val); - return 1; + return kvm_skip_emulated_instruction(vcpu); } static int kvm_fast_pio_in(struct kvm_vcpu *vcpu, int size, @@ -6572,6 +6602,7 @@ static int kvm_fast_pio_in(struct kvm_vcpu *vcpu, int size, return ret; } + vcpu->arch.pio.linear_rip = kvm_get_linear_rip(vcpu); vcpu->arch.complete_userspace_io = complete_fast_pio_in; return 0; @@ -6579,16 +6610,13 @@ static int kvm_fast_pio_in(struct kvm_vcpu *vcpu, int size, int kvm_fast_pio(struct kvm_vcpu *vcpu, int size, unsigned short port, int in) { - int ret = kvm_skip_emulated_instruction(vcpu); + int ret; - /* - * TODO: we might be squashing a KVM_GUESTDBG_SINGLESTEP-triggered - * KVM_EXIT_DEBUG here. - */ if (in) - return kvm_fast_pio_in(vcpu, size, port) && ret; + ret = kvm_fast_pio_in(vcpu, size, port); else - return kvm_fast_pio_out(vcpu, size, port) && ret; + ret = kvm_fast_pio_out(vcpu, size, port); + return ret && kvm_skip_emulated_instruction(vcpu); } EXPORT_SYMBOL_GPL(kvm_fast_pio); @@ -8733,6 +8761,7 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu) { + vcpu->arch.arch_capabilities = kvm_get_arch_capabilities(); vcpu->arch.msr_platform_info = MSR_PLATFORM_INFO_CPUID_FAULT; kvm_vcpu_mtrr_init(vcpu); vcpu_load(vcpu); @@ -9429,13 +9458,9 @@ void kvm_arch_commit_memory_region(struct kvm *kvm, const struct kvm_memory_slot *new, enum kvm_mr_change change) { - int nr_mmu_pages = 0; - if (!kvm->arch.n_requested_mmu_pages) - nr_mmu_pages = kvm_mmu_calculate_mmu_pages(kvm); - - if (nr_mmu_pages) - kvm_mmu_change_mmu_pages(kvm, nr_mmu_pages); + kvm_mmu_change_mmu_pages(kvm, + kvm_mmu_calculate_default_mmu_pages(kvm)); /* * Dirty logging tracks sptes in 4k granularity, meaning that large diff --git a/arch/x86/mm/mmap.c b/arch/x86/mm/mmap.c index db3165714521..dc726e07d8ba 100644 --- a/arch/x86/mm/mmap.c +++ b/arch/x86/mm/mmap.c @@ -230,7 +230,7 @@ bool mmap_address_hint_valid(unsigned long addr, unsigned long len) /* Can we access it for direct reading/writing? Must be RAM: */ int valid_phys_addr_range(phys_addr_t addr, size_t count) { - return addr + count <= __pa(high_memory); + return addr + count - 1 <= __pa(high_memory - 1); } /* Can we access it through mmap? Must be a valid physical address: */ diff --git a/arch/x86/platform/efi/quirks.c b/arch/x86/platform/efi/quirks.c index 458a0e2bcc57..a25a9fd987a9 100644 --- a/arch/x86/platform/efi/quirks.c +++ b/arch/x86/platform/efi/quirks.c @@ -449,7 +449,7 @@ void __init efi_free_boot_services(void) */ rm_size = real_mode_size_needed(); if (rm_size && (start + rm_size) < (1<<20) && size >= rm_size) { - set_real_mode_mem(start, rm_size); + set_real_mode_mem(start); start += rm_size; size -= rm_size; } diff --git a/arch/x86/realmode/init.c b/arch/x86/realmode/init.c index d10105825d57..7dce39c8c034 100644 --- a/arch/x86/realmode/init.c +++ b/arch/x86/realmode/init.c @@ -15,15 +15,6 @@ u32 *trampoline_cr4_features; /* Hold the pgd entry used on booting additional CPUs */ pgd_t trampoline_pgd_entry; -void __init set_real_mode_mem(phys_addr_t mem, size_t size) -{ - void *base = __va(mem); - - real_mode_header = (struct real_mode_header *) base; - printk(KERN_DEBUG "Base memory trampoline at [%p] %llx size %zu\n", - base, (unsigned long long)mem, size); -} - void __init reserve_real_mode(void) { phys_addr_t mem; @@ -42,7 +33,7 @@ void __init reserve_real_mode(void) } memblock_reserve(mem, size); - set_real_mode_mem(mem, size); + set_real_mode_mem(mem); } static void __init setup_real_mode(void) diff --git a/arch/xtensa/include/asm/Kbuild b/arch/xtensa/include/asm/Kbuild index 42b6cb3d16f7..3843198e03d4 100644 --- a/arch/xtensa/include/asm/Kbuild +++ b/arch/xtensa/include/asm/Kbuild @@ -15,6 +15,7 @@ generic-y += irq_work.h generic-y += kdebug.h generic-y += kmap_types.h generic-y += kprobes.h +generic-y += kvm_para.h generic-y += local.h generic-y += local64.h generic-y += mcs_spinlock.h diff --git a/arch/xtensa/include/asm/processor.h b/arch/xtensa/include/asm/processor.h index f7dd895b2353..0c14018d1c26 100644 --- a/arch/xtensa/include/asm/processor.h +++ b/arch/xtensa/include/asm/processor.h @@ -187,15 +187,18 @@ struct thread_struct { /* Clearing a0 terminates the backtrace. */ #define start_thread(regs, new_pc, new_sp) \ - memset(regs, 0, sizeof(*regs)); \ - regs->pc = new_pc; \ - regs->ps = USER_PS_VALUE; \ - regs->areg[1] = new_sp; \ - regs->areg[0] = 0; \ - regs->wmask = 1; \ - regs->depc = 0; \ - regs->windowbase = 0; \ - regs->windowstart = 1; + do { \ + memset((regs), 0, sizeof(*(regs))); \ + (regs)->pc = (new_pc); \ + (regs)->ps = USER_PS_VALUE; \ + (regs)->areg[1] = (new_sp); \ + (regs)->areg[0] = 0; \ + (regs)->wmask = 1; \ + (regs)->depc = 0; \ + (regs)->windowbase = 0; \ + (regs)->windowstart = 1; \ + (regs)->syscall = NO_SYSCALL; \ + } while (0) /* Forward declaration */ struct task_struct; diff --git a/arch/xtensa/include/asm/syscall.h b/arch/xtensa/include/asm/syscall.h index a168bf81c7f4..91dc06d58060 100644 --- a/arch/xtensa/include/asm/syscall.h +++ b/arch/xtensa/include/asm/syscall.h @@ -59,45 +59,24 @@ static inline void syscall_set_return_value(struct task_struct *task, static inline void syscall_get_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, unsigned long *args) { static const unsigned int reg[] = XTENSA_SYSCALL_ARGUMENT_REGS; - unsigned int j; + unsigned int i; - if (n == 0) - return; - - WARN_ON_ONCE(i + n > SYSCALL_MAX_ARGS); - - for (j = 0; j < n; ++j) { - if (i + j < SYSCALL_MAX_ARGS) - args[j] = regs->areg[reg[i + j]]; - else - args[j] = 0; - } + for (i = 0; i < 6; ++i) + args[i] = regs->areg[reg[i]]; } static inline void syscall_set_arguments(struct task_struct *task, struct pt_regs *regs, - unsigned int i, unsigned int n, const unsigned long *args) { static const unsigned int reg[] = XTENSA_SYSCALL_ARGUMENT_REGS; - unsigned int j; + unsigned int i; - if (n == 0) - return; - - if (WARN_ON_ONCE(i + n > SYSCALL_MAX_ARGS)) { - if (i < SYSCALL_MAX_ARGS) - n = SYSCALL_MAX_ARGS - i; - else - return; - } - - for (j = 0; j < n; ++j) - regs->areg[reg[i + j]] = args[j]; + for (i = 0; i < 6; ++i) + regs->areg[reg[i]] = args[i]; } asmlinkage long xtensa_rt_sigreturn(struct pt_regs*); diff --git a/arch/xtensa/include/uapi/asm/Kbuild b/arch/xtensa/include/uapi/asm/Kbuild index 8a7ad40be463..7417847dc438 100644 --- a/arch/xtensa/include/uapi/asm/Kbuild +++ b/arch/xtensa/include/uapi/asm/Kbuild @@ -1,2 +1 @@ generated-y += unistd_32.h -generic-y += kvm_para.h diff --git a/arch/xtensa/kernel/entry.S b/arch/xtensa/kernel/entry.S index e50f5124dc6f..e54af8b7e0f8 100644 --- a/arch/xtensa/kernel/entry.S +++ b/arch/xtensa/kernel/entry.S @@ -1860,6 +1860,8 @@ ENTRY(system_call) l32i a7, a2, PT_SYSCALL 1: + s32i a7, a1, 4 + /* syscall = sys_call_table[syscall_nr] */ movi a4, sys_call_table @@ -1893,8 +1895,12 @@ ENTRY(system_call) retw 1: + l32i a4, a1, 4 + l32i a3, a2, PT_SYSCALL + s32i a4, a2, PT_SYSCALL mov a6, a2 call4 do_syscall_trace_leave + s32i a3, a2, PT_SYSCALL retw ENDPROC(system_call) diff --git a/arch/xtensa/kernel/stacktrace.c b/arch/xtensa/kernel/stacktrace.c index 174c11f13bba..b9f82510c650 100644 --- a/arch/xtensa/kernel/stacktrace.c +++ b/arch/xtensa/kernel/stacktrace.c @@ -253,10 +253,14 @@ static int return_address_cb(struct stackframe *frame, void *data) return 1; } +/* + * level == 0 is for the return address from the caller of this function, + * not from this function itself. + */ unsigned long return_address(unsigned level) { struct return_addr_data r = { - .skip = level + 1, + .skip = level, }; walk_stackframe(stack_pointer(NULL), return_address_cb, &r); return r.addr; diff --git a/arch/xtensa/mm/mmu.c b/arch/xtensa/mm/mmu.c index 2fb7d1172228..03678c4afc39 100644 --- a/arch/xtensa/mm/mmu.c +++ b/arch/xtensa/mm/mmu.c @@ -33,7 +33,7 @@ static void * __init init_pmd(unsigned long vaddr, unsigned long n_pages) pte = memblock_alloc_low(n_pages * sizeof(pte_t), PAGE_SIZE); if (!pte) - panic("%s: Failed to allocate %zu bytes align=%lx\n", + panic("%s: Failed to allocate %lu bytes align=%lx\n", __func__, n_pages * sizeof(pte_t), PAGE_SIZE); for (i = 0; i < n_pages; ++i) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index 4c592496a16a..dfb8cb0af13a 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -674,7 +674,7 @@ static bool bfq_symmetric_scenario(struct bfq_data *bfqd) * at least two nodes. */ return !(varied_queue_weights || multiple_classes_busy -#ifdef BFQ_GROUP_IOSCHED_ENABLED +#ifdef CONFIG_BFQ_GROUP_IOSCHED || bfqd->num_groups_with_pending_reqs > 0 #endif ); @@ -2822,7 +2822,7 @@ static void bfq_dispatch_remove(struct request_queue *q, struct request *rq) bfq_remove_request(q, rq); } -static void __bfq_bfqq_expire(struct bfq_data *bfqd, struct bfq_queue *bfqq) +static bool __bfq_bfqq_expire(struct bfq_data *bfqd, struct bfq_queue *bfqq) { /* * If this bfqq is shared between multiple processes, check @@ -2855,9 +2855,11 @@ static void __bfq_bfqq_expire(struct bfq_data *bfqd, struct bfq_queue *bfqq) /* * All in-service entities must have been properly deactivated * or requeued before executing the next function, which - * resets all in-service entites as no more in service. + * resets all in-service entities as no more in service. This + * may cause bfqq to be freed. If this happens, the next + * function returns true. */ - __bfq_bfqd_reset_in_service(bfqd); + return __bfq_bfqd_reset_in_service(bfqd); } /** @@ -3262,7 +3264,6 @@ void bfq_bfqq_expire(struct bfq_data *bfqd, bool slow; unsigned long delta = 0; struct bfq_entity *entity = &bfqq->entity; - int ref; /* * Check whether the process is slow (see bfq_bfqq_is_slow). @@ -3347,10 +3348,8 @@ void bfq_bfqq_expire(struct bfq_data *bfqd, * reason. */ __bfq_bfqq_recalc_budget(bfqd, bfqq, reason); - ref = bfqq->ref; - __bfq_bfqq_expire(bfqd, bfqq); - - if (ref == 1) /* bfqq is gone, no more actions on it */ + if (__bfq_bfqq_expire(bfqd, bfqq)) + /* bfqq is gone, no more actions on it */ return; bfqq->injected_service = 0; diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h index 062e1c4787f4..86394e503ca9 100644 --- a/block/bfq-iosched.h +++ b/block/bfq-iosched.h @@ -995,7 +995,7 @@ bool __bfq_deactivate_entity(struct bfq_entity *entity, bool ins_into_idle_tree); bool next_queue_may_preempt(struct bfq_data *bfqd); struct bfq_queue *bfq_get_next_queue(struct bfq_data *bfqd); -void __bfq_bfqd_reset_in_service(struct bfq_data *bfqd); +bool __bfq_bfqd_reset_in_service(struct bfq_data *bfqd); void bfq_deactivate_bfqq(struct bfq_data *bfqd, struct bfq_queue *bfqq, bool ins_into_idle_tree, bool expiration); void bfq_activate_bfqq(struct bfq_data *bfqd, struct bfq_queue *bfqq); diff --git a/block/bfq-wf2q.c b/block/bfq-wf2q.c index 63311d1ff1ed..ae4d000ac0af 100644 --- a/block/bfq-wf2q.c +++ b/block/bfq-wf2q.c @@ -1012,7 +1012,7 @@ static void __bfq_activate_entity(struct bfq_entity *entity, entity->on_st = true; } -#ifdef BFQ_GROUP_IOSCHED_ENABLED +#ifdef CONFIG_BFQ_GROUP_IOSCHED if (!bfq_entity_to_bfqq(entity)) { /* bfq_group */ struct bfq_group *bfqg = container_of(entity, struct bfq_group, entity); @@ -1605,7 +1605,8 @@ struct bfq_queue *bfq_get_next_queue(struct bfq_data *bfqd) return bfqq; } -void __bfq_bfqd_reset_in_service(struct bfq_data *bfqd) +/* returns true if the in-service queue gets freed */ +bool __bfq_bfqd_reset_in_service(struct bfq_data *bfqd) { struct bfq_queue *in_serv_bfqq = bfqd->in_service_queue; struct bfq_entity *in_serv_entity = &in_serv_bfqq->entity; @@ -1629,8 +1630,20 @@ void __bfq_bfqd_reset_in_service(struct bfq_data *bfqd) * service tree either, then release the service reference to * the queue it represents (taken with bfq_get_entity). */ - if (!in_serv_entity->on_st) + if (!in_serv_entity->on_st) { + /* + * If no process is referencing in_serv_bfqq any + * longer, then the service reference may be the only + * reference to the queue. If this is the case, then + * bfqq gets freed here. + */ + int ref = in_serv_bfqq->ref; bfq_put_queue(in_serv_bfqq); + if (ref == 1) + return true; + } + + return false; } void bfq_deactivate_bfqq(struct bfq_data *bfqd, struct bfq_queue *bfqq, diff --git a/block/bio.c b/block/bio.c index b64cedc7f87c..716510ecd7ff 100644 --- a/block/bio.c +++ b/block/bio.c @@ -1298,8 +1298,11 @@ struct bio *bio_copy_user_iov(struct request_queue *q, } } - if (bio_add_pc_page(q, bio, page, bytes, offset) < bytes) + if (bio_add_pc_page(q, bio, page, bytes, offset) < bytes) { + if (!map_data) + __free_page(page); break; + } len -= bytes; offset = 0; diff --git a/block/blk-core.c b/block/blk-core.c index 4673ebe42255..a55389ba8779 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -1245,8 +1245,6 @@ static int blk_cloned_rq_check_limits(struct request_queue *q, */ blk_status_t blk_insert_cloned_request(struct request_queue *q, struct request *rq) { - blk_qc_t unused; - if (blk_cloned_rq_check_limits(q, rq)) return BLK_STS_IOERR; @@ -1262,7 +1260,7 @@ blk_status_t blk_insert_cloned_request(struct request_queue *q, struct request * * bypass a potential scheduler on the bottom device for * insert. */ - return blk_mq_try_issue_directly(rq->mq_hctx, rq, &unused, true, true); + return blk_mq_request_issue_directly(rq, true); } EXPORT_SYMBOL_GPL(blk_insert_cloned_request); diff --git a/block/blk-flush.c b/block/blk-flush.c index 6e0f2d97fc6d..d95f94892015 100644 --- a/block/blk-flush.c +++ b/block/blk-flush.c @@ -220,7 +220,7 @@ static void flush_end_io(struct request *flush_rq, blk_status_t error) blk_mq_tag_set_rq(hctx, flush_rq->tag, fq->orig_rq); flush_rq->tag = -1; } else { - blk_mq_put_driver_tag_hctx(hctx, flush_rq); + blk_mq_put_driver_tag(flush_rq); flush_rq->internal_tag = -1; } @@ -324,7 +324,7 @@ static void mq_flush_data_end_io(struct request *rq, blk_status_t error) if (q->elevator) { WARN_ON(rq->tag < 0); - blk_mq_put_driver_tag_hctx(hctx, rq); + blk_mq_put_driver_tag(rq); } /* diff --git a/block/blk-mq-sched.c b/block/blk-mq-sched.c index 40905539afed..aa6bc5c02643 100644 --- a/block/blk-mq-sched.c +++ b/block/blk-mq-sched.c @@ -423,10 +423,12 @@ void blk_mq_sched_insert_requests(struct blk_mq_hw_ctx *hctx, * busy in case of 'none' scheduler, and this way may save * us one extra enqueue & dequeue to sw queue. */ - if (!hctx->dispatch_busy && !e && !run_queue_async) + if (!hctx->dispatch_busy && !e && !run_queue_async) { blk_mq_try_issue_list_directly(hctx, list); - else - blk_mq_insert_requests(hctx, ctx, list); + if (list_empty(list)) + return; + } + blk_mq_insert_requests(hctx, ctx, list); } blk_mq_run_hw_queue(hctx, run_queue_async); diff --git a/block/blk-mq.c b/block/blk-mq.c index 70b210a308c4..9516304a38ee 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -59,7 +59,8 @@ static int blk_mq_poll_stats_bkt(const struct request *rq) } /* - * Check if any of the ctx's have pending work in this hardware queue + * Check if any of the ctx, dispatch list or elevator + * have pending work in this hardware queue. */ static bool blk_mq_hctx_has_pending(struct blk_mq_hw_ctx *hctx) { @@ -653,6 +654,13 @@ bool blk_mq_complete_request(struct request *rq) } EXPORT_SYMBOL(blk_mq_complete_request); +void blk_mq_complete_request_sync(struct request *rq) +{ + WRITE_ONCE(rq->state, MQ_RQ_COMPLETE); + rq->q->mq_ops->complete(rq); +} +EXPORT_SYMBOL_GPL(blk_mq_complete_request_sync); + int blk_mq_request_started(struct request *rq) { return blk_mq_rq_state(rq) != MQ_RQ_IDLE; @@ -1071,7 +1079,13 @@ static int blk_mq_dispatch_wake(wait_queue_entry_t *wait, unsigned mode, hctx = container_of(wait, struct blk_mq_hw_ctx, dispatch_wait); spin_lock(&hctx->dispatch_wait_lock); - list_del_init(&wait->entry); + if (!list_empty(&wait->entry)) { + struct sbitmap_queue *sbq; + + list_del_init(&wait->entry); + sbq = &hctx->tags->bitmap_tags; + atomic_dec(&sbq->ws_active); + } spin_unlock(&hctx->dispatch_wait_lock); blk_mq_run_hw_queue(hctx, true); @@ -1087,6 +1101,7 @@ static int blk_mq_dispatch_wake(wait_queue_entry_t *wait, unsigned mode, static bool blk_mq_mark_tag_wait(struct blk_mq_hw_ctx *hctx, struct request *rq) { + struct sbitmap_queue *sbq = &hctx->tags->bitmap_tags; struct wait_queue_head *wq; wait_queue_entry_t *wait; bool ret; @@ -1109,7 +1124,7 @@ static bool blk_mq_mark_tag_wait(struct blk_mq_hw_ctx *hctx, if (!list_empty_careful(&wait->entry)) return false; - wq = &bt_wait_ptr(&hctx->tags->bitmap_tags, hctx)->wait; + wq = &bt_wait_ptr(sbq, hctx)->wait; spin_lock_irq(&wq->lock); spin_lock(&hctx->dispatch_wait_lock); @@ -1119,6 +1134,7 @@ static bool blk_mq_mark_tag_wait(struct blk_mq_hw_ctx *hctx, return false; } + atomic_inc(&sbq->ws_active); wait->flags &= ~WQ_FLAG_EXCLUSIVE; __add_wait_queue(wq, wait); @@ -1139,6 +1155,7 @@ static bool blk_mq_mark_tag_wait(struct blk_mq_hw_ctx *hctx, * someone else gets the wakeup. */ list_del_init(&wait->entry); + atomic_dec(&sbq->ws_active); spin_unlock(&hctx->dispatch_wait_lock); spin_unlock_irq(&wq->lock); @@ -1701,11 +1718,12 @@ void blk_mq_flush_plug_list(struct blk_plug *plug, bool from_schedule) unsigned int depth; list_splice_init(&plug->mq_list, &list); - plug->rq_count = 0; if (plug->rq_count > 2 && plug->multiple_queues) list_sort(NULL, &list, plug_rq_cmp); + plug->rq_count = 0; + this_q = NULL; this_hctx = NULL; this_ctx = NULL; @@ -1790,74 +1808,76 @@ static blk_status_t __blk_mq_issue_directly(struct blk_mq_hw_ctx *hctx, return ret; } -blk_status_t blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx, +static blk_status_t __blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx, struct request *rq, blk_qc_t *cookie, - bool bypass, bool last) + bool bypass_insert, bool last) { struct request_queue *q = rq->q; bool run_queue = true; - blk_status_t ret = BLK_STS_RESOURCE; - int srcu_idx; - bool force = false; - hctx_lock(hctx, &srcu_idx); /* - * hctx_lock is needed before checking quiesced flag. + * RCU or SRCU read lock is needed before checking quiesced flag. * - * When queue is stopped or quiesced, ignore 'bypass', insert - * and return BLK_STS_OK to caller, and avoid driver to try to - * dispatch again. + * When queue is stopped or quiesced, ignore 'bypass_insert' from + * blk_mq_request_issue_directly(), and return BLK_STS_OK to caller, + * and avoid driver to try to dispatch again. */ - if (unlikely(blk_mq_hctx_stopped(hctx) || blk_queue_quiesced(q))) { + if (blk_mq_hctx_stopped(hctx) || blk_queue_quiesced(q)) { run_queue = false; - bypass = false; - goto out_unlock; + bypass_insert = false; + goto insert; } - if (unlikely(q->elevator && !bypass)) - goto out_unlock; + if (q->elevator && !bypass_insert) + goto insert; if (!blk_mq_get_dispatch_budget(hctx)) - goto out_unlock; + goto insert; if (!blk_mq_get_driver_tag(rq)) { blk_mq_put_dispatch_budget(hctx); - goto out_unlock; + goto insert; } - /* - * Always add a request that has been through - *.queue_rq() to the hardware dispatch list. - */ - force = true; - ret = __blk_mq_issue_directly(hctx, rq, cookie, last); -out_unlock: + return __blk_mq_issue_directly(hctx, rq, cookie, last); +insert: + if (bypass_insert) + return BLK_STS_RESOURCE; + + blk_mq_request_bypass_insert(rq, run_queue); + return BLK_STS_OK; +} + +static void blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx, + struct request *rq, blk_qc_t *cookie) +{ + blk_status_t ret; + int srcu_idx; + + might_sleep_if(hctx->flags & BLK_MQ_F_BLOCKING); + + hctx_lock(hctx, &srcu_idx); + + ret = __blk_mq_try_issue_directly(hctx, rq, cookie, false, true); + if (ret == BLK_STS_RESOURCE || ret == BLK_STS_DEV_RESOURCE) + blk_mq_request_bypass_insert(rq, true); + else if (ret != BLK_STS_OK) + blk_mq_end_request(rq, ret); + + hctx_unlock(hctx, srcu_idx); +} + +blk_status_t blk_mq_request_issue_directly(struct request *rq, bool last) +{ + blk_status_t ret; + int srcu_idx; + blk_qc_t unused_cookie; + struct blk_mq_hw_ctx *hctx = rq->mq_hctx; + + hctx_lock(hctx, &srcu_idx); + ret = __blk_mq_try_issue_directly(hctx, rq, &unused_cookie, true, last); hctx_unlock(hctx, srcu_idx); - switch (ret) { - case BLK_STS_OK: - break; - case BLK_STS_DEV_RESOURCE: - case BLK_STS_RESOURCE: - if (force) { - blk_mq_request_bypass_insert(rq, run_queue); - /* - * We have to return BLK_STS_OK for the DM - * to avoid livelock. Otherwise, we return - * the real result to indicate whether the - * request is direct-issued successfully. - */ - ret = bypass ? BLK_STS_OK : ret; - } else if (!bypass) { - blk_mq_sched_insert_request(rq, false, - run_queue, false); - } - break; - default: - if (!bypass) - blk_mq_end_request(rq, ret); - break; - } return ret; } @@ -1865,20 +1885,22 @@ out_unlock: void blk_mq_try_issue_list_directly(struct blk_mq_hw_ctx *hctx, struct list_head *list) { - blk_qc_t unused; - blk_status_t ret = BLK_STS_OK; - while (!list_empty(list)) { + blk_status_t ret; struct request *rq = list_first_entry(list, struct request, queuelist); list_del_init(&rq->queuelist); - if (ret == BLK_STS_OK) - ret = blk_mq_try_issue_directly(hctx, rq, &unused, - false, + ret = blk_mq_request_issue_directly(rq, list_empty(list)); + if (ret != BLK_STS_OK) { + if (ret == BLK_STS_RESOURCE || + ret == BLK_STS_DEV_RESOURCE) { + blk_mq_request_bypass_insert(rq, list_empty(list)); - else - blk_mq_sched_insert_request(rq, false, true, false); + break; + } + blk_mq_end_request(rq, ret); + } } /* @@ -1886,7 +1908,7 @@ void blk_mq_try_issue_list_directly(struct blk_mq_hw_ctx *hctx, * the driver there was more coming, but that turned out to * be a lie. */ - if (ret != BLK_STS_OK && hctx->queue->mq_ops->commit_rqs) + if (!list_empty(list) && hctx->queue->mq_ops->commit_rqs) hctx->queue->mq_ops->commit_rqs(hctx); } @@ -1993,19 +2015,21 @@ static blk_qc_t blk_mq_make_request(struct request_queue *q, struct bio *bio) plug->rq_count--; } blk_add_rq_to_plug(plug, rq); + trace_block_plug(q); blk_mq_put_ctx(data.ctx); if (same_queue_rq) { data.hctx = same_queue_rq->mq_hctx; + trace_block_unplug(q, 1, true); blk_mq_try_issue_directly(data.hctx, same_queue_rq, - &cookie, false, true); + &cookie); } } else if ((q->nr_hw_queues > 1 && is_sync) || (!q->elevator && !data.hctx->dispatch_busy)) { blk_mq_put_ctx(data.ctx); blk_mq_bio_to_request(rq, bio); - blk_mq_try_issue_directly(data.hctx, rq, &cookie, false, true); + blk_mq_try_issue_directly(data.hctx, rq, &cookie); } else { blk_mq_put_ctx(data.ctx); blk_mq_bio_to_request(rq, bio); @@ -2322,7 +2346,7 @@ static int blk_mq_init_hctx(struct request_queue *q, return 0; free_fq: - kfree(hctx->fq); + blk_free_flush_queue(hctx->fq); exit_hctx: if (set->ops->exit_hctx) set->ops->exit_hctx(hctx, hctx_idx); diff --git a/block/blk-mq.h b/block/blk-mq.h index 0ed8e5a8729f..423ea88ab6fb 100644 --- a/block/blk-mq.h +++ b/block/blk-mq.h @@ -70,10 +70,8 @@ void blk_mq_request_bypass_insert(struct request *rq, bool run_queue); void blk_mq_insert_requests(struct blk_mq_hw_ctx *hctx, struct blk_mq_ctx *ctx, struct list_head *list); -blk_status_t blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx, - struct request *rq, - blk_qc_t *cookie, - bool bypass, bool last); +/* Used by blk_insert_cloned_request() to issue request directly */ +blk_status_t blk_mq_request_issue_directly(struct request *rq, bool last); void blk_mq_try_issue_list_directly(struct blk_mq_hw_ctx *hctx, struct list_head *list); @@ -224,15 +222,6 @@ static inline void __blk_mq_put_driver_tag(struct blk_mq_hw_ctx *hctx, } } -static inline void blk_mq_put_driver_tag_hctx(struct blk_mq_hw_ctx *hctx, - struct request *rq) -{ - if (rq->tag == -1 || rq->internal_tag == -1) - return; - - __blk_mq_put_driver_tag(hctx, rq); -} - static inline void blk_mq_put_driver_tag(struct request *rq) { if (rq->tag == -1 || rq->internal_tag == -1) diff --git a/drivers/acpi/acpica/evgpe.c b/drivers/acpi/acpica/evgpe.c index 62d3aa74277b..5e9d7348c16f 100644 --- a/drivers/acpi/acpica/evgpe.c +++ b/drivers/acpi/acpica/evgpe.c @@ -81,8 +81,12 @@ acpi_status acpi_ev_enable_gpe(struct acpi_gpe_event_info *gpe_event_info) ACPI_FUNCTION_TRACE(ev_enable_gpe); - /* Enable the requested GPE */ + /* Clear the GPE status */ + status = acpi_hw_clear_gpe(gpe_event_info); + if (ACPI_FAILURE(status)) + return_ACPI_STATUS(status); + /* Enable the requested GPE */ status = acpi_hw_low_set_gpe(gpe_event_info, ACPI_GPE_ENABLE); return_ACPI_STATUS(status); } diff --git a/drivers/acpi/acpica/nsobject.c b/drivers/acpi/acpica/nsobject.c index 8638f43cfc3d..79d86da1c892 100644 --- a/drivers/acpi/acpica/nsobject.c +++ b/drivers/acpi/acpica/nsobject.c @@ -186,6 +186,10 @@ void acpi_ns_detach_object(struct acpi_namespace_node *node) } } + if (obj_desc->common.type == ACPI_TYPE_REGION) { + acpi_ut_remove_address_range(obj_desc->region.space_id, node); + } + /* Clear the Node entry in all cases */ node->object = NULL; diff --git a/drivers/acpi/bus.c b/drivers/acpi/bus.c index 6ecbbabf1233..eec263c9019e 100644 --- a/drivers/acpi/bus.c +++ b/drivers/acpi/bus.c @@ -1043,9 +1043,6 @@ void __init acpi_early_init(void) acpi_permanent_mmap = true; - /* Initialize debug output. Linux does not use ACPICA defaults */ - acpi_dbg_level = ACPI_LV_INFO | ACPI_LV_REPAIR; - #ifdef CONFIG_X86 /* * If the machine falls into the DMI check table, diff --git a/drivers/acpi/cppc_acpi.c b/drivers/acpi/cppc_acpi.c index 1b207fca1420..d4244e7d0e38 100644 --- a/drivers/acpi/cppc_acpi.c +++ b/drivers/acpi/cppc_acpi.c @@ -1150,8 +1150,13 @@ int cppc_get_perf_caps(int cpunum, struct cppc_perf_caps *perf_caps) cpc_read(cpunum, nominal_reg, &nom); perf_caps->nominal_perf = nom; - cpc_read(cpunum, guaranteed_reg, &guaranteed); - perf_caps->guaranteed_perf = guaranteed; + if (guaranteed_reg->type != ACPI_TYPE_BUFFER || + IS_NULL_REG(&guaranteed_reg->cpc_entry.reg)) { + perf_caps->guaranteed_perf = 0; + } else { + cpc_read(cpunum, guaranteed_reg, &guaranteed); + perf_caps->guaranteed_perf = guaranteed; + } cpc_read(cpunum, lowest_non_linear_reg, &min_nonlinear); perf_caps->lowest_nonlinear_perf = min_nonlinear; diff --git a/drivers/android/binder.c b/drivers/android/binder.c index 8685882da64c..4b9c7ca492e6 100644 --- a/drivers/android/binder.c +++ b/drivers/android/binder.c @@ -2057,7 +2057,8 @@ static size_t binder_get_object(struct binder_proc *proc, size_t object_size = 0; read_size = min_t(size_t, sizeof(*object), buffer->data_size - offset); - if (read_size < sizeof(*hdr) || !IS_ALIGNED(offset, sizeof(u32))) + if (offset > buffer->data_size || read_size < sizeof(*hdr) || + !IS_ALIGNED(offset, sizeof(u32))) return 0; binder_alloc_copy_from_buffer(&proc->alloc, object, buffer, offset, read_size); diff --git a/drivers/android/binder_alloc.c b/drivers/android/binder_alloc.c index 6389467670a0..195f120c4e8c 100644 --- a/drivers/android/binder_alloc.c +++ b/drivers/android/binder_alloc.c @@ -927,14 +927,13 @@ enum lru_status binder_alloc_free_page(struct list_head *item, index = page - alloc->pages; page_addr = (uintptr_t)alloc->buffer + index * PAGE_SIZE; + + mm = alloc->vma_vm_mm; + if (!mmget_not_zero(mm)) + goto err_mmget; + if (!down_write_trylock(&mm->mmap_sem)) + goto err_down_write_mmap_sem_failed; vma = binder_alloc_get_vma(alloc); - if (vma) { - if (!mmget_not_zero(alloc->vma_vm_mm)) - goto err_mmget; - mm = alloc->vma_vm_mm; - if (!down_read_trylock(&mm->mmap_sem)) - goto err_down_write_mmap_sem_failed; - } list_lru_isolate(lru, item); spin_unlock(lock); @@ -945,10 +944,9 @@ enum lru_status binder_alloc_free_page(struct list_head *item, zap_page_range(vma, page_addr, PAGE_SIZE); trace_binder_unmap_user_end(alloc, index); - - up_read(&mm->mmap_sem); - mmput(mm); } + up_write(&mm->mmap_sem); + mmput(mm); trace_binder_unmap_kernel_start(alloc, index); diff --git a/drivers/ata/libata-zpodd.c b/drivers/ata/libata-zpodd.c index b3ed8f9953a8..173e6f2dd9af 100644 --- a/drivers/ata/libata-zpodd.c +++ b/drivers/ata/libata-zpodd.c @@ -52,38 +52,52 @@ static int eject_tray(struct ata_device *dev) /* Per the spec, only slot type and drawer type ODD can be supported */ static enum odd_mech_type zpodd_get_mech_type(struct ata_device *dev) { - char buf[16]; + char *buf; unsigned int ret; - struct rm_feature_desc *desc = (void *)(buf + 8); + struct rm_feature_desc *desc; struct ata_taskfile tf; static const char cdb[] = { GPCMD_GET_CONFIGURATION, 2, /* only 1 feature descriptor requested */ 0, 3, /* 3, removable medium feature */ 0, 0, 0,/* reserved */ - 0, sizeof(buf), + 0, 16, 0, 0, 0, }; + buf = kzalloc(16, GFP_KERNEL); + if (!buf) + return ODD_MECH_TYPE_UNSUPPORTED; + desc = (void *)(buf + 8); + ata_tf_init(dev, &tf); tf.flags = ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE; tf.command = ATA_CMD_PACKET; tf.protocol = ATAPI_PROT_PIO; - tf.lbam = sizeof(buf); + tf.lbam = 16; ret = ata_exec_internal(dev, &tf, cdb, DMA_FROM_DEVICE, - buf, sizeof(buf), 0); - if (ret) + buf, 16, 0); + if (ret) { + kfree(buf); return ODD_MECH_TYPE_UNSUPPORTED; + } - if (be16_to_cpu(desc->feature_code) != 3) + if (be16_to_cpu(desc->feature_code) != 3) { + kfree(buf); return ODD_MECH_TYPE_UNSUPPORTED; + } - if (desc->mech_type == 0 && desc->load == 0 && desc->eject == 1) + if (desc->mech_type == 0 && desc->load == 0 && desc->eject == 1) { + kfree(buf); return ODD_MECH_TYPE_SLOT; - else if (desc->mech_type == 1 && desc->load == 0 && desc->eject == 1) + } else if (desc->mech_type == 1 && desc->load == 0 && + desc->eject == 1) { + kfree(buf); return ODD_MECH_TYPE_DRAWER; - else + } else { + kfree(buf); return ODD_MECH_TYPE_UNSUPPORTED; + } } /* Test if ODD is zero power ready by sense code */ diff --git a/drivers/block/null_blk_main.c b/drivers/block/null_blk_main.c index 417a9f15c116..d7ac09c092f2 100644 --- a/drivers/block/null_blk_main.c +++ b/drivers/block/null_blk_main.c @@ -1748,6 +1748,11 @@ static int __init null_init(void) return -EINVAL; } + if (g_home_node != NUMA_NO_NODE && g_home_node >= nr_online_nodes) { + pr_err("null_blk: invalid home_node value\n"); + g_home_node = NUMA_NO_NODE; + } + if (g_queue_mode == NULL_Q_RQ) { pr_err("null_blk: legacy IO path no longer available\n"); return -EINVAL; diff --git a/drivers/block/paride/pcd.c b/drivers/block/paride/pcd.c index 377a694dc228..6d415b20fb70 100644 --- a/drivers/block/paride/pcd.c +++ b/drivers/block/paride/pcd.c @@ -314,6 +314,7 @@ static void pcd_init_units(void) disk->queue = blk_mq_init_sq_queue(&cd->tag_set, &pcd_mq_ops, 1, BLK_MQ_F_SHOULD_MERGE); if (IS_ERR(disk->queue)) { + put_disk(disk); disk->queue = NULL; continue; } @@ -750,6 +751,8 @@ static int pcd_detect(void) printk("%s: No CD-ROM drive found\n", name); for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++) { + if (!cd->disk) + continue; blk_cleanup_queue(cd->disk->queue); cd->disk->queue = NULL; blk_mq_free_tag_set(&cd->tag_set); @@ -1010,8 +1013,14 @@ static int __init pcd_init(void) pcd_probe_capabilities(); if (register_blkdev(major, name)) { - for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++) + for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++) { + if (!cd->disk) + continue; + + blk_cleanup_queue(cd->disk->queue); + blk_mq_free_tag_set(&cd->tag_set); put_disk(cd->disk); + } return -EBUSY; } @@ -1032,6 +1041,9 @@ static void __exit pcd_exit(void) int unit; for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++) { + if (!cd->disk) + continue; + if (cd->present) { del_gendisk(cd->disk); pi_release(cd->pi); diff --git a/drivers/block/paride/pf.c b/drivers/block/paride/pf.c index 103b617cdc31..35e6e271b219 100644 --- a/drivers/block/paride/pf.c +++ b/drivers/block/paride/pf.c @@ -762,6 +762,8 @@ static int pf_detect(void) printk("%s: No ATAPI disk detected\n", name); for (pf = units, unit = 0; unit < PF_UNITS; pf++, unit++) { + if (!pf->disk) + continue; blk_cleanup_queue(pf->disk->queue); pf->disk->queue = NULL; blk_mq_free_tag_set(&pf->tag_set); @@ -1029,8 +1031,13 @@ static int __init pf_init(void) pf_busy = 0; if (register_blkdev(major, name)) { - for (pf = units, unit = 0; unit < PF_UNITS; pf++, unit++) + for (pf = units, unit = 0; unit < PF_UNITS; pf++, unit++) { + if (!pf->disk) + continue; + blk_cleanup_queue(pf->disk->queue); + blk_mq_free_tag_set(&pf->tag_set); put_disk(pf->disk); + } return -EBUSY; } @@ -1051,6 +1058,9 @@ static void __exit pf_exit(void) int unit; unregister_blkdev(major, name); for (pf = units, unit = 0; unit < PF_UNITS; pf++, unit++) { + if (!pf->disk) + continue; + if (pf->present) del_gendisk(pf->disk); diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c index 4bc083b7c9b5..2a7ca4a1e6f7 100644 --- a/drivers/block/virtio_blk.c +++ b/drivers/block/virtio_blk.c @@ -513,6 +513,8 @@ static int init_vq(struct virtio_blk *vblk) if (err) num_vqs = 1; + num_vqs = min_t(unsigned int, nr_cpu_ids, num_vqs); + vblk->vqs = kmalloc_array(num_vqs, sizeof(*vblk->vqs), GFP_KERNEL); if (!vblk->vqs) return -ENOMEM; diff --git a/drivers/block/xsysace.c b/drivers/block/xsysace.c index 87ccef4bd69e..32a21b8d1d85 100644 --- a/drivers/block/xsysace.c +++ b/drivers/block/xsysace.c @@ -1090,6 +1090,8 @@ static int ace_setup(struct ace_device *ace) return 0; err_read: + /* prevent double queue cleanup */ + ace->gd->queue = NULL; put_disk(ace->gd); err_alloc_disk: blk_cleanup_queue(ace->queue); diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c index e7a5f1d1c314..399cad7daae7 100644 --- a/drivers/block/zram/zram_drv.c +++ b/drivers/block/zram/zram_drv.c @@ -290,18 +290,8 @@ static ssize_t idle_store(struct device *dev, struct zram *zram = dev_to_zram(dev); unsigned long nr_pages = zram->disksize >> PAGE_SHIFT; int index; - char mode_buf[8]; - ssize_t sz; - sz = strscpy(mode_buf, buf, sizeof(mode_buf)); - if (sz <= 0) - return -EINVAL; - - /* ignore trailing new line */ - if (mode_buf[sz - 1] == '\n') - mode_buf[sz - 1] = 0x00; - - if (strcmp(mode_buf, "all")) + if (!sysfs_streq(buf, "all")) return -EINVAL; down_read(&zram->init_lock); @@ -635,25 +625,15 @@ static ssize_t writeback_store(struct device *dev, struct bio bio; struct bio_vec bio_vec; struct page *page; - ssize_t ret, sz; - char mode_buf[8]; - int mode = -1; + ssize_t ret; + int mode; unsigned long blk_idx = 0; - sz = strscpy(mode_buf, buf, sizeof(mode_buf)); - if (sz <= 0) - return -EINVAL; - - /* ignore trailing newline */ - if (mode_buf[sz - 1] == '\n') - mode_buf[sz - 1] = 0x00; - - if (!strcmp(mode_buf, "idle")) + if (sysfs_streq(buf, "idle")) mode = IDLE_WRITEBACK; - else if (!strcmp(mode_buf, "huge")) + else if (sysfs_streq(buf, "huge")) mode = HUGE_WRITEBACK; - - if (mode == -1) + else return -EINVAL; down_read(&zram->init_lock); diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c index ded198328f21..7db48ae65cd2 100644 --- a/drivers/bluetooth/btusb.c +++ b/drivers/bluetooth/btusb.c @@ -2942,6 +2942,7 @@ static int btusb_config_oob_wake(struct hci_dev *hdev) return 0; } + irq_set_status_flags(irq, IRQ_NOAUTOEN); ret = devm_request_irq(&hdev->dev, irq, btusb_oob_wake_handler, 0, "OOB Wake-on-BT", data); if (ret) { @@ -2956,7 +2957,6 @@ static int btusb_config_oob_wake(struct hci_dev *hdev) } data->oob_wake_irq = irq; - disable_irq(irq); bt_dev_info(hdev, "OOB Wake-on-BT configured at IRQ %u", irq); return 0; } diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index 72866a004f07..466ebd84ad17 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -348,7 +348,7 @@ config XILINX_HWICAP config R3964 tristate "Siemens R3964 line discipline" - depends on TTY + depends on TTY && BROKEN ---help--- This driver allows synchronous communication with devices using the Siemens R3964 packet protocol. Unless you are dealing with special diff --git a/drivers/char/tpm/eventlog/tpm2.c b/drivers/char/tpm/eventlog/tpm2.c index d8b77133a83a..f824563fc28d 100644 --- a/drivers/char/tpm/eventlog/tpm2.c +++ b/drivers/char/tpm/eventlog/tpm2.c @@ -37,8 +37,8 @@ * * Returns size of the event. If it is an invalid event, returns 0. */ -static int calc_tpm2_event_size(struct tcg_pcr_event2_head *event, - struct tcg_pcr_event *event_header) +static size_t calc_tpm2_event_size(struct tcg_pcr_event2_head *event, + struct tcg_pcr_event *event_header) { struct tcg_efi_specid_event_head *efispecid; struct tcg_event_field *event_field; diff --git a/drivers/char/tpm/tpm-dev-common.c b/drivers/char/tpm/tpm-dev-common.c index 8856cce5a23b..817ae09a369e 100644 --- a/drivers/char/tpm/tpm-dev-common.c +++ b/drivers/char/tpm/tpm-dev-common.c @@ -233,12 +233,19 @@ __poll_t tpm_common_poll(struct file *file, poll_table *wait) __poll_t mask = 0; poll_wait(file, &priv->async_wait, wait); + mutex_lock(&priv->buffer_mutex); - if (!priv->response_read || priv->response_length) + /* + * The response_length indicates if there is still response + * (or part of it) to be consumed. Partial reads decrease it + * by the number of bytes read, and write resets it the zero. + */ + if (priv->response_length) mask = EPOLLIN | EPOLLRDNORM; else mask = EPOLLOUT | EPOLLWRNORM; + mutex_unlock(&priv->buffer_mutex); return mask; } diff --git a/drivers/char/tpm/tpm-interface.c b/drivers/char/tpm/tpm-interface.c index 83ece5639f86..ae1030c9b086 100644 --- a/drivers/char/tpm/tpm-interface.c +++ b/drivers/char/tpm/tpm-interface.c @@ -402,15 +402,13 @@ int tpm_pm_suspend(struct device *dev) if (chip->flags & TPM_CHIP_FLAG_ALWAYS_POWERED) return 0; - if (chip->flags & TPM_CHIP_FLAG_TPM2) { - mutex_lock(&chip->tpm_mutex); - if (!tpm_chip_start(chip)) { + if (!tpm_chip_start(chip)) { + if (chip->flags & TPM_CHIP_FLAG_TPM2) tpm2_shutdown(chip, TPM2_SU_STATE); - tpm_chip_stop(chip); - } - mutex_unlock(&chip->tpm_mutex); - } else { - rc = tpm1_pm_suspend(chip, tpm_suspend_pcr); + else + rc = tpm1_pm_suspend(chip, tpm_suspend_pcr); + + tpm_chip_stop(chip); } return rc; diff --git a/drivers/clk/at91/clk-programmable.c b/drivers/clk/at91/clk-programmable.c index 89d6f3736dbf..f8edbb65eda3 100644 --- a/drivers/clk/at91/clk-programmable.c +++ b/drivers/clk/at91/clk-programmable.c @@ -20,8 +20,7 @@ #define PROG_ID_MAX 7 #define PROG_STATUS_MASK(id) (1 << ((id) + 8)) -#define PROG_PRES_MASK 0x7 -#define PROG_PRES(layout, pckr) ((pckr >> layout->pres_shift) & PROG_PRES_MASK) +#define PROG_PRES(layout, pckr) ((pckr >> layout->pres_shift) & layout->pres_mask) #define PROG_MAX_RM9200_CSS 3 struct clk_programmable { @@ -37,20 +36,29 @@ static unsigned long clk_programmable_recalc_rate(struct clk_hw *hw, unsigned long parent_rate) { struct clk_programmable *prog = to_clk_programmable(hw); + const struct clk_programmable_layout *layout = prog->layout; unsigned int pckr; + unsigned long rate; regmap_read(prog->regmap, AT91_PMC_PCKR(prog->id), &pckr); - return parent_rate >> PROG_PRES(prog->layout, pckr); + if (layout->is_pres_direct) + rate = parent_rate / (PROG_PRES(layout, pckr) + 1); + else + rate = parent_rate >> PROG_PRES(layout, pckr); + + return rate; } static int clk_programmable_determine_rate(struct clk_hw *hw, struct clk_rate_request *req) { + struct clk_programmable *prog = to_clk_programmable(hw); + const struct clk_programmable_layout *layout = prog->layout; struct clk_hw *parent; long best_rate = -EINVAL; unsigned long parent_rate; - unsigned long tmp_rate; + unsigned long tmp_rate = 0; int shift; int i; @@ -60,10 +68,18 @@ static int clk_programmable_determine_rate(struct clk_hw *hw, continue; parent_rate = clk_hw_get_rate(parent); - for (shift = 0; shift < PROG_PRES_MASK; shift++) { - tmp_rate = parent_rate >> shift; - if (tmp_rate <= req->rate) - break; + if (layout->is_pres_direct) { + for (shift = 0; shift <= layout->pres_mask; shift++) { + tmp_rate = parent_rate / (shift + 1); + if (tmp_rate <= req->rate) + break; + } + } else { + for (shift = 0; shift < layout->pres_mask; shift++) { + tmp_rate = parent_rate >> shift; + if (tmp_rate <= req->rate) + break; + } } if (tmp_rate > req->rate) @@ -137,16 +153,23 @@ static int clk_programmable_set_rate(struct clk_hw *hw, unsigned long rate, if (!div) return -EINVAL; - shift = fls(div) - 1; + if (layout->is_pres_direct) { + shift = div - 1; - if (div != (1 << shift)) - return -EINVAL; + if (shift > layout->pres_mask) + return -EINVAL; + } else { + shift = fls(div) - 1; - if (shift >= PROG_PRES_MASK) - return -EINVAL; + if (div != (1 << shift)) + return -EINVAL; + + if (shift >= layout->pres_mask) + return -EINVAL; + } regmap_update_bits(prog->regmap, AT91_PMC_PCKR(prog->id), - PROG_PRES_MASK << layout->pres_shift, + layout->pres_mask << layout->pres_shift, shift << layout->pres_shift); return 0; @@ -202,19 +225,25 @@ at91_clk_register_programmable(struct regmap *regmap, } const struct clk_programmable_layout at91rm9200_programmable_layout = { + .pres_mask = 0x7, .pres_shift = 2, .css_mask = 0x3, .have_slck_mck = 0, + .is_pres_direct = 0, }; const struct clk_programmable_layout at91sam9g45_programmable_layout = { + .pres_mask = 0x7, .pres_shift = 2, .css_mask = 0x3, .have_slck_mck = 1, + .is_pres_direct = 0, }; const struct clk_programmable_layout at91sam9x5_programmable_layout = { + .pres_mask = 0x7, .pres_shift = 4, .css_mask = 0x7, .have_slck_mck = 0, + .is_pres_direct = 0, }; diff --git a/drivers/clk/at91/pmc.h b/drivers/clk/at91/pmc.h index 672a79bda88c..a0e5ce9c9b9e 100644 --- a/drivers/clk/at91/pmc.h +++ b/drivers/clk/at91/pmc.h @@ -71,9 +71,11 @@ struct clk_pll_characteristics { }; struct clk_programmable_layout { + u8 pres_mask; u8 pres_shift; u8 css_mask; u8 have_slck_mck; + u8 is_pres_direct; }; extern const struct clk_programmable_layout at91rm9200_programmable_layout; diff --git a/drivers/clk/at91/sama5d2.c b/drivers/clk/at91/sama5d2.c index 1f70cb164b06..81943fac4537 100644 --- a/drivers/clk/at91/sama5d2.c +++ b/drivers/clk/at91/sama5d2.c @@ -125,6 +125,14 @@ static const struct { .pll = true }, }; +static const struct clk_programmable_layout sama5d2_programmable_layout = { + .pres_mask = 0xff, + .pres_shift = 4, + .css_mask = 0x7, + .have_slck_mck = 0, + .is_pres_direct = 1, +}; + static void __init sama5d2_pmc_setup(struct device_node *np) { struct clk_range range = CLK_RANGE(0, 0); @@ -249,7 +257,7 @@ static void __init sama5d2_pmc_setup(struct device_node *np) hw = at91_clk_register_programmable(regmap, name, parent_names, 6, i, - &at91sam9x5_programmable_layout); + &sama5d2_programmable_layout); if (IS_ERR(hw)) goto err_free; } diff --git a/drivers/clk/imx/clk-pll14xx.c b/drivers/clk/imx/clk-pll14xx.c index 1acfa3e3cfb4..113d71042199 100644 --- a/drivers/clk/imx/clk-pll14xx.c +++ b/drivers/clk/imx/clk-pll14xx.c @@ -362,7 +362,7 @@ struct clk *imx_clk_pll14xx(const char *name, const char *parent_name, switch (pll_clk->type) { case PLL_1416X: - if (!pll->rate_table) + if (!pll_clk->rate_table) init.ops = &clk_pll1416x_min_ops; else init.ops = &clk_pll1416x_ops; diff --git a/drivers/clk/mediatek/clk-gate.c b/drivers/clk/mediatek/clk-gate.c index 9628d4e7690b..85daf826619a 100644 --- a/drivers/clk/mediatek/clk-gate.c +++ b/drivers/clk/mediatek/clk-gate.c @@ -169,11 +169,10 @@ struct clk *mtk_clk_register_gate( return ERR_PTR(-ENOMEM); init.name = name; - init.flags = CLK_SET_RATE_PARENT; + init.flags = flags | CLK_SET_RATE_PARENT; init.parent_names = parent_name ? &parent_name : NULL; init.num_parents = parent_name ? 1 : 0; init.ops = ops; - init.flags = flags; cg->regmap = regmap; cg->set_ofs = set_ofs; diff --git a/drivers/clk/meson/clk-pll.c b/drivers/clk/meson/clk-pll.c index 41e16dd7272a..7a14ac9b2fec 100644 --- a/drivers/clk/meson/clk-pll.c +++ b/drivers/clk/meson/clk-pll.c @@ -120,7 +120,7 @@ static bool meson_clk_pll_is_better(unsigned long rate, return true; } else { /* Round down */ - if (now < rate && best < now) + if (now <= rate && best < now) return true; } diff --git a/drivers/clk/meson/g12a.c b/drivers/clk/meson/g12a.c index 0e1ce8c03259..f7b11e1eeebe 100644 --- a/drivers/clk/meson/g12a.c +++ b/drivers/clk/meson/g12a.c @@ -960,14 +960,14 @@ static struct clk_regmap g12a_sd_emmc_c_clk0 = { /* VPU Clock */ static const char * const g12a_vpu_parent_names[] = { - "fclk_div4", "fclk_div3", "fclk_div5", "fclk_div7", + "fclk_div3", "fclk_div4", "fclk_div5", "fclk_div7", "mpll1", "vid_pll", "hifi_pll", "gp0_pll", }; static struct clk_regmap g12a_vpu_0_sel = { .data = &(struct clk_regmap_mux_data){ .offset = HHI_VPU_CLK_CNTL, - .mask = 0x3, + .mask = 0x7, .shift = 9, }, .hw.init = &(struct clk_init_data){ @@ -1011,7 +1011,7 @@ static struct clk_regmap g12a_vpu_0 = { static struct clk_regmap g12a_vpu_1_sel = { .data = &(struct clk_regmap_mux_data){ .offset = HHI_VPU_CLK_CNTL, - .mask = 0x3, + .mask = 0x7, .shift = 25, }, .hw.init = &(struct clk_init_data){ diff --git a/drivers/clk/meson/gxbb.c b/drivers/clk/meson/gxbb.c index 04df2e208ed6..29ffb4fde714 100644 --- a/drivers/clk/meson/gxbb.c +++ b/drivers/clk/meson/gxbb.c @@ -2216,6 +2216,7 @@ static struct clk_regmap gxbb_vdec_1_div = { .offset = HHI_VDEC_CLK_CNTL, .shift = 0, .width = 7, + .flags = CLK_DIVIDER_ROUND_CLOSEST, }, .hw.init = &(struct clk_init_data){ .name = "vdec_1_div", @@ -2261,6 +2262,7 @@ static struct clk_regmap gxbb_vdec_hevc_div = { .offset = HHI_VDEC2_CLK_CNTL, .shift = 16, .width = 7, + .flags = CLK_DIVIDER_ROUND_CLOSEST, }, .hw.init = &(struct clk_init_data){ .name = "vdec_hevc_div", diff --git a/drivers/clk/meson/vid-pll-div.c b/drivers/clk/meson/vid-pll-div.c index 08bcc01c0923..daff235bc763 100644 --- a/drivers/clk/meson/vid-pll-div.c +++ b/drivers/clk/meson/vid-pll-div.c @@ -82,8 +82,8 @@ static unsigned long meson_vid_pll_div_recalc_rate(struct clk_hw *hw, div = _get_table_val(meson_parm_read(clk->map, &pll_div->val), meson_parm_read(clk->map, &pll_div->sel)); if (!div || !div->divider) { - pr_info("%s: Invalid config value for vid_pll_div\n", __func__); - return parent_rate; + pr_debug("%s: Invalid config value for vid_pll_div\n", __func__); + return 0; } return DIV_ROUND_UP_ULL(parent_rate * div->multiplier, div->divider); diff --git a/drivers/clk/x86/clk-pmc-atom.c b/drivers/clk/x86/clk-pmc-atom.c index d977193842df..19174835693b 100644 --- a/drivers/clk/x86/clk-pmc-atom.c +++ b/drivers/clk/x86/clk-pmc-atom.c @@ -165,7 +165,7 @@ static const struct clk_ops plt_clk_ops = { }; static struct clk_plt *plt_clk_register(struct platform_device *pdev, int id, - void __iomem *base, + const struct pmc_clk_data *pmc_data, const char **parent_names, int num_parents) { @@ -184,9 +184,17 @@ static struct clk_plt *plt_clk_register(struct platform_device *pdev, int id, init.num_parents = num_parents; pclk->hw.init = &init; - pclk->reg = base + PMC_CLK_CTL_OFFSET + id * PMC_CLK_CTL_SIZE; + pclk->reg = pmc_data->base + PMC_CLK_CTL_OFFSET + id * PMC_CLK_CTL_SIZE; spin_lock_init(&pclk->lock); + /* + * On some systems, the pmc_plt_clocks already enabled by the + * firmware are being marked as critical to avoid them being + * gated by the clock framework. + */ + if (pmc_data->critical && plt_clk_is_enabled(&pclk->hw)) + init.flags |= CLK_IS_CRITICAL; + ret = devm_clk_hw_register(&pdev->dev, &pclk->hw); if (ret) { pclk = ERR_PTR(ret); @@ -332,7 +340,7 @@ static int plt_clk_probe(struct platform_device *pdev) return PTR_ERR(parent_names); for (i = 0; i < PMC_CLK_NUM; i++) { - data->clks[i] = plt_clk_register(pdev, i, pmc_data->base, + data->clks[i] = plt_clk_register(pdev, i, pmc_data, parent_names, data->nparents); if (IS_ERR(data->clks[i])) { err = PTR_ERR(data->clks[i]); diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index e22f0dbaebb1..2986119dd31f 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -385,7 +385,10 @@ static int intel_pstate_get_cppc_guranteed(int cpu) if (ret) return ret; - return cppc_perf.guaranteed_perf; + if (cppc_perf.guaranteed_perf) + return cppc_perf.guaranteed_perf; + + return cppc_perf.nominal_perf; } #else /* CONFIG_ACPI_CPPC_LIB */ @@ -2593,6 +2596,9 @@ static int __init intel_pstate_init(void) const struct x86_cpu_id *id; int rc; + if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) + return -ENODEV; + if (no_load) return -ENODEV; @@ -2608,7 +2614,7 @@ static int __init intel_pstate_init(void) } else { id = x86_match_cpu(intel_pstate_cpu_ids); if (!id) { - pr_info("CPU ID not supported\n"); + pr_info("CPU model not supported\n"); return -ENODEV; } diff --git a/drivers/cpufreq/scpi-cpufreq.c b/drivers/cpufreq/scpi-cpufreq.c index 3f49427766b8..2b51e0718c9f 100644 --- a/drivers/cpufreq/scpi-cpufreq.c +++ b/drivers/cpufreq/scpi-cpufreq.c @@ -189,8 +189,8 @@ static int scpi_cpufreq_exit(struct cpufreq_policy *policy) clk_put(priv->clk); dev_pm_opp_free_cpufreq_table(priv->cpu_dev, &policy->freq_table); - kfree(priv); dev_pm_opp_remove_all_dynamic(priv->cpu_dev); + kfree(priv); return 0; } diff --git a/drivers/crypto/caam/caamhash.c b/drivers/crypto/caam/caamhash.c index b1eadc6652b5..7205d9f4029e 100644 --- a/drivers/crypto/caam/caamhash.c +++ b/drivers/crypto/caam/caamhash.c @@ -865,19 +865,18 @@ static int ahash_update_ctx(struct ahash_request *req) if (ret) goto unmap_ctx; - if (mapped_nents) { + if (mapped_nents) sg_to_sec4_sg_last(req->src, mapped_nents, edesc->sec4_sg + sec4_sg_src_index, 0); - if (*next_buflen) - scatterwalk_map_and_copy(next_buf, req->src, - to_hash - *buflen, - *next_buflen, 0); - } else { + else sg_to_sec4_set_last(edesc->sec4_sg + sec4_sg_src_index - 1); - } + if (*next_buflen) + scatterwalk_map_and_copy(next_buf, req->src, + to_hash - *buflen, + *next_buflen, 0); desc = edesc->hw_desc; edesc->sec4_sg_dma = dma_map_single(jrdev, edesc->sec4_sg, diff --git a/drivers/dma-buf/Makefile b/drivers/dma-buf/Makefile index 0913a6ccab5a..1f006e083eb9 100644 --- a/drivers/dma-buf/Makefile +++ b/drivers/dma-buf/Makefile @@ -1,4 +1,5 @@ -obj-y := dma-buf.o dma-fence.o dma-fence-array.o reservation.o seqno-fence.o +obj-y := dma-buf.o dma-fence.o dma-fence-array.o dma-fence-chain.o \ + reservation.o seqno-fence.o obj-$(CONFIG_SYNC_FILE) += sync_file.o obj-$(CONFIG_SW_SYNC) += sw_sync.o sync_debug.o obj-$(CONFIG_UDMABUF) += udmabuf.o diff --git a/drivers/dma-buf/dma-fence-chain.c b/drivers/dma-buf/dma-fence-chain.c new file mode 100644 index 000000000000..93c42078cb57 --- /dev/null +++ b/drivers/dma-buf/dma-fence-chain.c @@ -0,0 +1,242 @@ +/* + * fence-chain: chain fences together in a timeline + * + * Copyright (C) 2018 Advanced Micro Devices, Inc. + * Authors: + * Christian König + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + */ + +#include + +static bool dma_fence_chain_enable_signaling(struct dma_fence *fence); + +/** + * dma_fence_chain_get_prev - use RCU to get a reference to the previous fence + * @chain: chain node to get the previous node from + * + * Use dma_fence_get_rcu_safe to get a reference to the previous fence of the + * chain node. + */ +static struct dma_fence *dma_fence_chain_get_prev(struct dma_fence_chain *chain) +{ + struct dma_fence *prev; + + rcu_read_lock(); + prev = dma_fence_get_rcu_safe(&chain->prev); + rcu_read_unlock(); + return prev; +} + +/** + * dma_fence_chain_walk - chain walking function + * @fence: current chain node + * + * Walk the chain to the next node. Returns the next fence or NULL if we are at + * the end of the chain. Garbage collects chain nodes which are already + * signaled. + */ +struct dma_fence *dma_fence_chain_walk(struct dma_fence *fence) +{ + struct dma_fence_chain *chain, *prev_chain; + struct dma_fence *prev, *replacement, *tmp; + + chain = to_dma_fence_chain(fence); + if (!chain) { + dma_fence_put(fence); + return NULL; + } + + while ((prev = dma_fence_chain_get_prev(chain))) { + + prev_chain = to_dma_fence_chain(prev); + if (prev_chain) { + if (!dma_fence_is_signaled(prev_chain->fence)) + break; + + replacement = dma_fence_chain_get_prev(prev_chain); + } else { + if (!dma_fence_is_signaled(prev)) + break; + + replacement = NULL; + } + + tmp = cmpxchg((void **)&chain->prev, (void *)prev, (void *)replacement); + if (tmp == prev) + dma_fence_put(tmp); + else + dma_fence_put(replacement); + dma_fence_put(prev); + } + + dma_fence_put(fence); + return prev; +} +EXPORT_SYMBOL(dma_fence_chain_walk); + +/** + * dma_fence_chain_find_seqno - find fence chain node by seqno + * @pfence: pointer to the chain node where to start + * @seqno: the sequence number to search for + * + * Advance the fence pointer to the chain node which will signal this sequence + * number. If no sequence number is provided then this is a no-op. + * + * Returns EINVAL if the fence is not a chain node or the sequence number has + * not yet advanced far enough. + */ +int dma_fence_chain_find_seqno(struct dma_fence **pfence, uint64_t seqno) +{ + struct dma_fence_chain *chain; + + if (!seqno) + return 0; + + chain = to_dma_fence_chain(*pfence); + if (!chain || chain->base.seqno < seqno) + return -EINVAL; + + dma_fence_chain_for_each(*pfence, &chain->base) { + if ((*pfence)->context != chain->base.context || + to_dma_fence_chain(*pfence)->prev_seqno < seqno) + break; + } + dma_fence_put(&chain->base); + + return 0; +} +EXPORT_SYMBOL(dma_fence_chain_find_seqno); + +static const char *dma_fence_chain_get_driver_name(struct dma_fence *fence) +{ + return "dma_fence_chain"; +} + +static const char *dma_fence_chain_get_timeline_name(struct dma_fence *fence) +{ + return "unbound"; +} + +static void dma_fence_chain_irq_work(struct irq_work *work) +{ + struct dma_fence_chain *chain; + + chain = container_of(work, typeof(*chain), work); + + /* Try to rearm the callback */ + if (!dma_fence_chain_enable_signaling(&chain->base)) + /* Ok, we are done. No more unsignaled fences left */ + dma_fence_signal(&chain->base); + dma_fence_put(&chain->base); +} + +static void dma_fence_chain_cb(struct dma_fence *f, struct dma_fence_cb *cb) +{ + struct dma_fence_chain *chain; + + chain = container_of(cb, typeof(*chain), cb); + irq_work_queue(&chain->work); + dma_fence_put(f); +} + +static bool dma_fence_chain_enable_signaling(struct dma_fence *fence) +{ + struct dma_fence_chain *head = to_dma_fence_chain(fence); + + dma_fence_get(&head->base); + dma_fence_chain_for_each(fence, &head->base) { + struct dma_fence_chain *chain = to_dma_fence_chain(fence); + struct dma_fence *f = chain ? chain->fence : fence; + + dma_fence_get(f); + if (!dma_fence_add_callback(f, &head->cb, dma_fence_chain_cb)) { + dma_fence_put(fence); + return true; + } + dma_fence_put(f); + } + dma_fence_put(&head->base); + return false; +} + +static bool dma_fence_chain_signaled(struct dma_fence *fence) +{ + dma_fence_chain_for_each(fence, fence) { + struct dma_fence_chain *chain = to_dma_fence_chain(fence); + struct dma_fence *f = chain ? chain->fence : fence; + + if (!dma_fence_is_signaled(f)) { + dma_fence_put(fence); + return false; + } + } + + return true; +} + +static void dma_fence_chain_release(struct dma_fence *fence) +{ + struct dma_fence_chain *chain = to_dma_fence_chain(fence); + + dma_fence_put(rcu_dereference_protected(chain->prev, true)); + dma_fence_put(chain->fence); + dma_fence_free(fence); +} + +const struct dma_fence_ops dma_fence_chain_ops = { + .use_64bit_seqno = true, + .get_driver_name = dma_fence_chain_get_driver_name, + .get_timeline_name = dma_fence_chain_get_timeline_name, + .enable_signaling = dma_fence_chain_enable_signaling, + .signaled = dma_fence_chain_signaled, + .release = dma_fence_chain_release, +}; +EXPORT_SYMBOL(dma_fence_chain_ops); + +/** + * dma_fence_chain_init - initialize a fence chain + * @chain: the chain node to initialize + * @prev: the previous fence + * @fence: the current fence + * + * Initialize a new chain node and either start a new chain or add the node to + * the existing chain of the previous fence. + */ +void dma_fence_chain_init(struct dma_fence_chain *chain, + struct dma_fence *prev, + struct dma_fence *fence, + uint64_t seqno) +{ + struct dma_fence_chain *prev_chain = to_dma_fence_chain(prev); + uint64_t context; + + spin_lock_init(&chain->lock); + rcu_assign_pointer(chain->prev, prev); + chain->fence = fence; + chain->prev_seqno = 0; + init_irq_work(&chain->work, dma_fence_chain_irq_work); + + /* Try to reuse the context of the previous chain node. */ + if (prev_chain && __dma_fence_is_later(seqno, prev->seqno, prev->ops)) { + context = prev->context; + chain->prev_seqno = prev->seqno; + } else { + context = dma_fence_context_alloc(1); + /* Make sure that we always have a valid sequence number. */ + if (prev_chain) + seqno = max(prev->seqno, seqno); + } + + dma_fence_init(&chain->base, &dma_fence_chain_ops, + &chain->lock, context, seqno); +} +EXPORT_SYMBOL(dma_fence_chain_init); diff --git a/drivers/dma-buf/sw_sync.c b/drivers/dma-buf/sw_sync.c index 32dcf7b4c935..119b2ffbc2c9 100644 --- a/drivers/dma-buf/sw_sync.c +++ b/drivers/dma-buf/sw_sync.c @@ -161,7 +161,7 @@ static bool timeline_fence_signaled(struct dma_fence *fence) { struct sync_timeline *parent = dma_fence_parent(fence); - return !__dma_fence_is_later(fence->seqno, parent->value); + return !__dma_fence_is_later(fence->seqno, parent->value, fence->ops); } static bool timeline_fence_enable_signaling(struct dma_fence *fence) diff --git a/drivers/dma-buf/sync_file.c b/drivers/dma-buf/sync_file.c index 4f6305ca52c8..ed3fb6e5224c 100644 --- a/drivers/dma-buf/sync_file.c +++ b/drivers/dma-buf/sync_file.c @@ -258,7 +258,8 @@ static struct sync_file *sync_file_merge(const char *name, struct sync_file *a, i_b++; } else { - if (__dma_fence_is_later(pt_a->seqno, pt_b->seqno)) + if (__dma_fence_is_later(pt_a->seqno, pt_b->seqno, + pt_a->ops)) add_fence(fences, &i, pt_a); else add_fence(fences, &i, pt_b); diff --git a/drivers/dma/stm32-mdma.c b/drivers/dma/stm32-mdma.c index 4e0eede599a8..ac0301b69593 100644 --- a/drivers/dma/stm32-mdma.c +++ b/drivers/dma/stm32-mdma.c @@ -1578,11 +1578,9 @@ static int stm32_mdma_probe(struct platform_device *pdev) dmadev->nr_channels = nr_channels; dmadev->nr_requests = nr_requests; - ret = device_property_read_u32_array(&pdev->dev, "st,ahb-addr-masks", + device_property_read_u32_array(&pdev->dev, "st,ahb-addr-masks", dmadev->ahb_addr_masks, count); - if (ret) - return ret; dmadev->nr_ahb_addr_masks = count; res = platform_get_resource(pdev, IORESOURCE_MEM, 0); diff --git a/drivers/gpio/gpio-adnp.c b/drivers/gpio/gpio-adnp.c index 91b90c0cea73..12acdac85820 100644 --- a/drivers/gpio/gpio-adnp.c +++ b/drivers/gpio/gpio-adnp.c @@ -132,8 +132,10 @@ static int adnp_gpio_direction_input(struct gpio_chip *chip, unsigned offset) if (err < 0) goto out; - if (err & BIT(pos)) - err = -EACCES; + if (value & BIT(pos)) { + err = -EPERM; + goto out; + } err = 0; diff --git a/drivers/gpio/gpio-aspeed.c b/drivers/gpio/gpio-aspeed.c index 854bce4fb9e7..217507002dbc 100644 --- a/drivers/gpio/gpio-aspeed.c +++ b/drivers/gpio/gpio-aspeed.c @@ -1224,6 +1224,8 @@ static int __init aspeed_gpio_probe(struct platform_device *pdev) gpio->offset_timer = devm_kzalloc(&pdev->dev, gpio->chip.ngpio, GFP_KERNEL); + if (!gpio->offset_timer) + return -ENOMEM; return aspeed_gpio_setup_irqs(gpio, pdev); } diff --git a/drivers/gpio/gpio-exar.c b/drivers/gpio/gpio-exar.c index 0ecd2369c2ca..a09d2f9ebacc 100644 --- a/drivers/gpio/gpio-exar.c +++ b/drivers/gpio/gpio-exar.c @@ -148,6 +148,8 @@ static int gpio_exar_probe(struct platform_device *pdev) mutex_init(&exar_gpio->lock); index = ida_simple_get(&ida_index, 0, 0, GFP_KERNEL); + if (index < 0) + goto err_destroy; sprintf(exar_gpio->name, "exar_gpio%d", index); exar_gpio->gpio_chip.label = exar_gpio->name; diff --git a/drivers/gpio/gpio-mockup.c b/drivers/gpio/gpio-mockup.c index 154d959e8993..b6a4efce7c92 100644 --- a/drivers/gpio/gpio-mockup.c +++ b/drivers/gpio/gpio-mockup.c @@ -204,8 +204,8 @@ static ssize_t gpio_mockup_debugfs_read(struct file *file, struct gpio_mockup_chip *chip; struct seq_file *sfile; struct gpio_chip *gc; + int val, cnt; char buf[3]; - int val, rv; if (*ppos != 0) return 0; @@ -216,13 +216,9 @@ static ssize_t gpio_mockup_debugfs_read(struct file *file, gc = &chip->gc; val = gpio_mockup_get(gc, priv->offset); - snprintf(buf, sizeof(buf), "%d\n", val); + cnt = snprintf(buf, sizeof(buf), "%d\n", val); - rv = copy_to_user(usr_buf, buf, sizeof(buf)); - if (rv) - return rv; - - return sizeof(buf) - 1; + return simple_read_from_buffer(usr_buf, size, ppos, buf, cnt); } static ssize_t gpio_mockup_debugfs_write(struct file *file, diff --git a/drivers/gpio/gpiolib-of.c b/drivers/gpio/gpiolib-of.c index 8b9c3ab70f6e..6a3ec575a404 100644 --- a/drivers/gpio/gpiolib-of.c +++ b/drivers/gpio/gpiolib-of.c @@ -120,7 +120,8 @@ static void of_gpio_flags_quirks(struct device_node *np, * to determine if the flags should have inverted semantics. */ if (IS_ENABLED(CONFIG_SPI_MASTER) && - of_property_read_bool(np, "cs-gpios")) { + of_property_read_bool(np, "cs-gpios") && + !strcmp(propname, "cs-gpios")) { struct device_node *child; u32 cs; int ret; @@ -142,16 +143,16 @@ static void of_gpio_flags_quirks(struct device_node *np, * conflict and the "spi-cs-high" flag will * take precedence. */ - if (of_property_read_bool(np, "spi-cs-high")) { + if (of_property_read_bool(child, "spi-cs-high")) { if (*flags & OF_GPIO_ACTIVE_LOW) { pr_warn("%s GPIO handle specifies active low - ignored\n", - of_node_full_name(np)); + of_node_full_name(child)); *flags &= ~OF_GPIO_ACTIVE_LOW; } } else { if (!(*flags & OF_GPIO_ACTIVE_LOW)) pr_info("%s enforce active low on chipselect handle\n", - of_node_full_name(np)); + of_node_full_name(child)); *flags |= OF_GPIO_ACTIVE_LOW; } break; @@ -717,7 +718,13 @@ int of_gpiochip_add(struct gpio_chip *chip) of_node_get(chip->of_node); - return of_gpiochip_scan_gpios(chip); + status = of_gpiochip_scan_gpios(chip); + if (status) { + of_node_put(chip->of_node); + gpiochip_remove_pin_ranges(chip); + } + + return status; } void of_gpiochip_remove(struct gpio_chip *chip) diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c index 144af0733581..0495bf1d480a 100644 --- a/drivers/gpio/gpiolib.c +++ b/drivers/gpio/gpiolib.c @@ -2776,7 +2776,7 @@ int gpiod_set_debounce(struct gpio_desc *desc, unsigned debounce) } config = pinconf_to_config_packed(PIN_CONFIG_INPUT_DEBOUNCE, debounce); - return gpio_set_config(chip, gpio_chip_hwgpio(desc), config); + return chip->set_config(chip, gpio_chip_hwgpio(desc), config); } EXPORT_SYMBOL_GPL(gpiod_set_debounce); @@ -2813,7 +2813,7 @@ int gpiod_set_transitory(struct gpio_desc *desc, bool transitory) packed = pinconf_to_config_packed(PIN_CONFIG_PERSIST_STATE, !transitory); gpio = gpio_chip_hwgpio(desc); - rc = gpio_set_config(chip, gpio, packed); + rc = chip->set_config(chip, gpio, packed); if (rc == -ENOTSUPP) { dev_dbg(&desc->gdev->dev, "Persistence not supported for GPIO %d\n", gpio); diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig index 5e1bc630b885..2267e84d5cb4 100644 --- a/drivers/gpu/drm/Kconfig +++ b/drivers/gpu/drm/Kconfig @@ -231,8 +231,6 @@ config DRM_AMDGPU source "drivers/gpu/drm/amd/amdgpu/Kconfig" -source "drivers/gpu/drm/amd/lib/Kconfig" - source "drivers/gpu/drm/nouveau/Kconfig" source "drivers/gpu/drm/i915/Kconfig" @@ -257,6 +255,9 @@ config DRM_VKMS If M is selected the module will be called vkms. +config DRM_ATI_PCIGART + bool + source "drivers/gpu/drm/exynos/Kconfig" source "drivers/gpu/drm/rockchip/Kconfig" @@ -337,12 +338,19 @@ source "drivers/gpu/drm/xen/Kconfig" source "drivers/gpu/drm/vboxvideo/Kconfig" +source "drivers/gpu/drm/lima/Kconfig" + +source "drivers/gpu/drm/panfrost/Kconfig" + +source "drivers/gpu/drm/aspeed/Kconfig" + # Keep legacy drivers last menuconfig DRM_LEGACY bool "Enable legacy drivers (DANGEROUS)" depends on DRM && MMU select DRM_VM + select DRM_ATI_PCIGART if PCI help Enable legacy DRI1 drivers. Those drivers expose unsafe and dangerous APIs to user-space, which can be used to circumvent access diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile index e630eccb951c..72f5036d9bfa 100644 --- a/drivers/gpu/drm/Makefile +++ b/drivers/gpu/drm/Makefile @@ -3,11 +3,9 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -drm-y := drm_auth.o drm_bufs.o drm_cache.o \ - drm_context.o drm_dma.o \ +drm-y := drm_auth.o drm_cache.o \ drm_file.o drm_gem.o drm_ioctl.o drm_irq.o \ - drm_lock.o drm_memory.o drm_drv.o \ - drm_scatter.o drm_pci.o \ + drm_memory.o drm_drv.o drm_pci.o \ drm_sysfs.o drm_hashtab.o drm_mm.o \ drm_crtc.o drm_fourcc.o drm_modes.o drm_edid.o \ drm_encoder_slave.o \ @@ -21,12 +19,13 @@ drm-y := drm_auth.o drm_bufs.o drm_cache.o \ drm_syncobj.o drm_lease.o drm_writeback.o drm_client.o \ drm_atomic_uapi.o +drm-$(CONFIG_DRM_LEGACY) += drm_legacy_misc.o drm_bufs.o drm_context.o drm_dma.o drm_scatter.o drm_lock.o drm-$(CONFIG_DRM_LIB_RANDOM) += lib/drm_random.o drm-$(CONFIG_DRM_VM) += drm_vm.o drm-$(CONFIG_COMPAT) += drm_ioc32.o drm-$(CONFIG_DRM_GEM_CMA_HELPER) += drm_gem_cma_helper.o drm-$(CONFIG_DRM_GEM_SHMEM_HELPER) += drm_gem_shmem_helper.o -drm-$(CONFIG_PCI) += ati_pcigart.o +drm-$(CONFIG_DRM_ATI_PCIGART) += ati_pcigart.o drm-$(CONFIG_DRM_PANEL) += drm_panel.o drm-$(CONFIG_OF) += drm_of.o drm-$(CONFIG_AGP) += drm_agpsupport.o @@ -38,7 +37,8 @@ drm_kms_helper-y := drm_crtc_helper.o drm_dp_helper.o drm_dsc.o drm_probe_helper drm_kms_helper_common.o drm_dp_dual_mode_helper.o \ drm_simple_kms_helper.o drm_modeset_helper.o \ drm_scdc_helper.o drm_gem_framebuffer_helper.o \ - drm_atomic_state_helper.o drm_damage_helper.o + drm_atomic_state_helper.o drm_damage_helper.o \ + drm_format_helper.o drm_kms_helper-$(CONFIG_DRM_PANEL_BRIDGE) += bridge/panel.o drm_kms_helper-$(CONFIG_DRM_FBDEV_EMULATION) += drm_fb_helper.o @@ -57,7 +57,6 @@ obj-$(CONFIG_DRM_TTM) += ttm/ obj-$(CONFIG_DRM_SCHED) += scheduler/ obj-$(CONFIG_DRM_TDFX) += tdfx/ obj-$(CONFIG_DRM_R128) += r128/ -obj-y += amd/lib/ obj-$(CONFIG_HSA_AMD) += amd/amdkfd/ obj-$(CONFIG_DRM_RADEON)+= radeon/ obj-$(CONFIG_DRM_AMDGPU)+= amd/amdgpu/ @@ -111,3 +110,6 @@ obj-$(CONFIG_DRM_PL111) += pl111/ obj-$(CONFIG_DRM_TVE200) += tve200/ obj-$(CONFIG_DRM_XEN) += xen/ obj-$(CONFIG_DRM_VBOXVIDEO) += vboxvideo/ +obj-$(CONFIG_DRM_LIMA) += lima/ +obj-$(CONFIG_DRM_PANFROST) += panfrost/ +obj-$(CONFIG_DRM_ASPEED_GFX) += aspeed/ diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index 466da5954a68..f8c58c425eb9 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile @@ -53,7 +53,8 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \ amdgpu_ucode.o amdgpu_bo_list.o amdgpu_ctx.o amdgpu_sync.o \ amdgpu_gtt_mgr.o amdgpu_vram_mgr.o amdgpu_virt.o amdgpu_atomfirmware.o \ amdgpu_vf_error.o amdgpu_sched.o amdgpu_debugfs.o amdgpu_ids.o \ - amdgpu_gmc.o amdgpu_xgmi.o amdgpu_csa.o + amdgpu_gmc.o amdgpu_xgmi.o amdgpu_csa.o amdgpu_ras.o amdgpu_vm_cpu.o \ + amdgpu_vm_sdma.o # add asic specific block amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 8d0d7f3dd5fb..14398f55f602 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -83,6 +83,7 @@ #include "amdgpu_gem.h" #include "amdgpu_doorbell.h" #include "amdgpu_amdkfd.h" +#include "amdgpu_smu.h" #define MAX_GPU_INSTANCE 16 @@ -156,6 +157,8 @@ extern int amdgpu_emu_mode; extern uint amdgpu_smu_memory_pool_size; extern uint amdgpu_dc_feature_mask; extern struct amdgpu_mgpu_info mgpu_info; +extern int amdgpu_ras_enable; +extern uint amdgpu_ras_mask; #ifdef CONFIG_DRM_AMDGPU_SI extern int amdgpu_si_support; @@ -433,6 +436,12 @@ struct amdgpu_cs_chunk { void *kdata; }; +struct amdgpu_cs_post_dep { + struct drm_syncobj *syncobj; + struct dma_fence_chain *chain; + u64 point; +}; + struct amdgpu_cs_parser { struct amdgpu_device *adev; struct drm_file *filp; @@ -462,8 +471,8 @@ struct amdgpu_cs_parser { /* user fence */ struct amdgpu_bo_list_entry uf_entry; - unsigned num_post_dep_syncobjs; - struct drm_syncobj **post_dep_syncobjs; + unsigned num_post_deps; + struct amdgpu_cs_post_dep *post_deps; }; static inline u32 amdgpu_get_ib_value(struct amdgpu_cs_parser *p, @@ -702,7 +711,6 @@ enum amd_hw_ip_block_type { struct amd_powerplay { void *pp_handle; const struct amd_pm_funcs *pp_funcs; - uint32_t pp_feature; }; #define AMDGPU_RESET_MAGIC_NUM 64 @@ -825,6 +833,7 @@ struct amdgpu_device { /* For pre-DCE11. DCE11 and later are in "struct amdgpu_device->dm" */ struct work_struct hotplug_work; struct amdgpu_irq_src crtc_irq; + struct amdgpu_irq_src vupdate_irq; struct amdgpu_irq_src pageflip_irq; struct amdgpu_irq_src hpd_irq; @@ -842,6 +851,9 @@ struct amdgpu_device { struct amd_powerplay powerplay; bool pp_force_state_enabled; + /* smu */ + struct smu_context smu; + /* dpm */ struct amdgpu_pm pm; u32 cg_flags; @@ -922,6 +934,8 @@ struct amdgpu_device { int asic_reset_res; struct work_struct xgmi_reset_work; + + bool in_baco_reset; }; static inline struct amdgpu_device *amdgpu_ttm_adev(struct ttm_bo_device *bdev) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index fe1d7368c1e6..acf8ae0cee9a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -640,4 +640,8 @@ int kgd2kfd_post_reset(struct kfd_dev *kfd) void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry) { } + +void kgd2kfd_set_sram_ecc_flag(struct kfd_dev *kfd) +{ +} #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index 0b31a1859023..775f815f9521 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -230,5 +230,6 @@ int kgd2kfd_quiesce_mm(struct mm_struct *mm); int kgd2kfd_resume_mm(struct mm_struct *mm); int kgd2kfd_schedule_evict_and_restore_process(struct mm_struct *mm, struct dma_fence *fence); +void kgd2kfd_set_sram_ecc_flag(struct kfd_dev *kfd); #endif /* AMDGPU_AMDKFD_H_INCLUDED */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 1921dec3df7a..a6e5184d436c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -410,15 +410,7 @@ static int add_bo_to_vm(struct amdgpu_device *adev, struct kgd_mem *mem, if (p_bo_va_entry) *p_bo_va_entry = bo_va_entry; - /* Allocate new page tables if needed and validate - * them. - */ - ret = amdgpu_vm_alloc_pts(adev, vm, va, amdgpu_bo_size(bo)); - if (ret) { - pr_err("Failed to allocate pts, err=%d\n", ret); - goto err_alloc_pts; - } - + /* Allocate validate page tables if needed */ ret = vm_validate_pt_pd_bos(vm); if (ret) { pr_err("validate_pt_pd_bos() failed\n"); @@ -741,13 +733,7 @@ static int update_gpuvm_pte(struct amdgpu_device *adev, struct amdgpu_sync *sync) { int ret; - struct amdgpu_vm *vm; - struct amdgpu_bo_va *bo_va; - struct amdgpu_bo *bo; - - bo_va = entry->bo_va; - vm = bo_va->base.vm; - bo = bo_va->base.bo; + struct amdgpu_bo_va *bo_va = entry->bo_va; /* Update the page tables */ ret = amdgpu_vm_bo_update(adev, bo_va, false); @@ -906,7 +892,8 @@ static int init_kfd_vm(struct amdgpu_vm *vm, void **process_info, pr_err("validate_pt_pd_bos() failed\n"); goto validate_pd_fail; } - amdgpu_bo_sync_wait(vm->root.base.bo, AMDGPU_FENCE_OWNER_KFD, false); + ret = amdgpu_bo_sync_wait(vm->root.base.bo, + AMDGPU_FENCE_OWNER_KFD, false); if (ret) goto wait_pd_fail; amdgpu_bo_fence(vm->root.base.bo, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c index b61e1dc61b4c..f96d75c6e099 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c @@ -28,8 +28,6 @@ #include "atom.h" #include "atombios.h" -#define get_index_into_master_table(master_table, table_name) (offsetof(struct master_table, table_name) / sizeof(uint16_t)) - bool amdgpu_atomfirmware_gpu_supports_virtualization(struct amdgpu_device *adev) { int index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, @@ -238,10 +236,71 @@ int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev) return 0; } +/* + * Return true if vbios enabled ecc by default, if umc info table is available + * or false if ecc is not enabled or umc info table is not available + */ +bool amdgpu_atomfirmware_mem_ecc_supported(struct amdgpu_device *adev) +{ + struct amdgpu_mode_info *mode_info = &adev->mode_info; + int index; + u16 data_offset, size; + union umc_info *umc_info; + u8 frev, crev; + bool ecc_default_enabled = false; + + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + umc_info); + + if (amdgpu_atom_parse_data_header(mode_info->atom_context, + index, &size, &frev, &crev, &data_offset)) { + /* support umc_info 3.1+ */ + if ((frev == 3 && crev >= 1) || (frev > 3)) { + umc_info = (union umc_info *) + (mode_info->atom_context->bios + data_offset); + ecc_default_enabled = + (le32_to_cpu(umc_info->v31.umc_config) & + UMC_CONFIG__DEFAULT_MEM_ECC_ENABLE) ? true : false; + } + } + + return ecc_default_enabled; +} + union firmware_info { struct atom_firmware_info_v3_1 v31; }; +/* + * Return true if vbios supports sram ecc or false if not + */ +bool amdgpu_atomfirmware_sram_ecc_supported(struct amdgpu_device *adev) +{ + struct amdgpu_mode_info *mode_info = &adev->mode_info; + int index; + u16 data_offset, size; + union firmware_info *firmware_info; + u8 frev, crev; + bool sram_ecc_supported = false; + + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + firmwareinfo); + + if (amdgpu_atom_parse_data_header(adev->mode_info.atom_context, + index, &size, &frev, &crev, &data_offset)) { + /* support firmware_info 3.1 + */ + if ((frev == 3 && crev >=1) || (frev > 3)) { + firmware_info = (union firmware_info *) + (mode_info->atom_context->bios + data_offset); + sram_ecc_supported = + (le32_to_cpu(firmware_info->v31.firmware_capability) & + ATOM_FIRMWARE_CAP_SRAM_ECC) ? true : false; + } + } + + return sram_ecc_supported; +} + union smu_info { struct atom_smu_info_v3_1 v31; }; @@ -346,11 +405,11 @@ int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev) (mode_info->atom_context->bios + data_offset); switch (crev) { case 4: - adev->gfx.config.max_shader_engines = gfx_info->v24.gc_num_se; - adev->gfx.config.max_cu_per_sh = gfx_info->v24.gc_num_cu_per_sh; - adev->gfx.config.max_sh_per_se = gfx_info->v24.gc_num_sh_per_se; - adev->gfx.config.max_backends_per_se = gfx_info->v24.gc_num_rb_per_se; - adev->gfx.config.max_texture_channel_caches = gfx_info->v24.gc_num_tccs; + adev->gfx.config.max_shader_engines = gfx_info->v24.max_shader_engines; + adev->gfx.config.max_cu_per_sh = gfx_info->v24.max_cu_per_sh; + adev->gfx.config.max_sh_per_se = gfx_info->v24.max_sh_per_se; + adev->gfx.config.max_backends_per_se = gfx_info->v24.max_backends_per_se; + adev->gfx.config.max_texture_channel_caches = gfx_info->v24.max_texture_channel_caches; adev->gfx.config.max_gprs = le16_to_cpu(gfx_info->v24.gc_num_gprs); adev->gfx.config.max_gs_threads = gfx_info->v24.gc_num_max_gs_thds; adev->gfx.config.gs_vgt_table_depth = gfx_info->v24.gc_gs_table_depth; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h index 20f158fd3b76..5ec6f92f353c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h @@ -24,6 +24,8 @@ #ifndef __AMDGPU_ATOMFIRMWARE_H__ #define __AMDGPU_ATOMFIRMWARE_H__ +#define get_index_into_master_table(master_table, table_name) (offsetof(struct master_table, table_name) / sizeof(uint16_t)) + bool amdgpu_atomfirmware_gpu_supports_virtualization(struct amdgpu_device *adev); void amdgpu_atomfirmware_scratch_regs_init(struct amdgpu_device *adev); int amdgpu_atomfirmware_allocate_fb_scratch(struct amdgpu_device *adev); @@ -31,5 +33,7 @@ int amdgpu_atomfirmware_get_vram_width(struct amdgpu_device *adev); int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev); int amdgpu_atomfirmware_get_clock_info(struct amdgpu_device *adev); int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev); +bool amdgpu_atomfirmware_mem_ecc_supported(struct amdgpu_device *adev); +bool amdgpu_atomfirmware_sram_ecc_supported(struct amdgpu_device *adev); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 52a5e4fdc95b..2f6239b6be6f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -215,6 +215,8 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, union drm_amdgpu_cs case AMDGPU_CHUNK_ID_SYNCOBJ_IN: case AMDGPU_CHUNK_ID_SYNCOBJ_OUT: case AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES: + case AMDGPU_CHUNK_ID_SYNCOBJ_TIMELINE_WAIT: + case AMDGPU_CHUNK_ID_SYNCOBJ_TIMELINE_SIGNAL: break; default: @@ -804,9 +806,11 @@ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, ttm_eu_backoff_reservation(&parser->ticket, &parser->validated); - for (i = 0; i < parser->num_post_dep_syncobjs; i++) - drm_syncobj_put(parser->post_dep_syncobjs[i]); - kfree(parser->post_dep_syncobjs); + for (i = 0; i < parser->num_post_deps; i++) { + drm_syncobj_put(parser->post_deps[i].syncobj); + kfree(parser->post_deps[i].chain); + } + kfree(parser->post_deps); dma_fence_put(parser->fence); @@ -1117,13 +1121,18 @@ static int amdgpu_cs_process_fence_dep(struct amdgpu_cs_parser *p, } static int amdgpu_syncobj_lookup_and_add_to_sync(struct amdgpu_cs_parser *p, - uint32_t handle) + uint32_t handle, u64 point, + u64 flags) { - int r; struct dma_fence *fence; - r = drm_syncobj_find_fence(p->filp, handle, 0, 0, &fence); - if (r) + int r; + + r = drm_syncobj_find_fence(p->filp, handle, point, flags, &fence); + if (r) { + DRM_ERROR("syncobj %u failed to find fence @ %llu (%d)!\n", + handle, point, r); return r; + } r = amdgpu_sync_fence(p->adev, &p->job->sync, fence, true); dma_fence_put(fence); @@ -1134,46 +1143,118 @@ static int amdgpu_syncobj_lookup_and_add_to_sync(struct amdgpu_cs_parser *p, static int amdgpu_cs_process_syncobj_in_dep(struct amdgpu_cs_parser *p, struct amdgpu_cs_chunk *chunk) { + struct drm_amdgpu_cs_chunk_sem *deps; unsigned num_deps; int i, r; - struct drm_amdgpu_cs_chunk_sem *deps; deps = (struct drm_amdgpu_cs_chunk_sem *)chunk->kdata; num_deps = chunk->length_dw * 4 / sizeof(struct drm_amdgpu_cs_chunk_sem); - for (i = 0; i < num_deps; ++i) { - r = amdgpu_syncobj_lookup_and_add_to_sync(p, deps[i].handle); + r = amdgpu_syncobj_lookup_and_add_to_sync(p, deps[i].handle, + 0, 0); if (r) return r; } + + return 0; +} + + +static int amdgpu_cs_process_syncobj_timeline_in_dep(struct amdgpu_cs_parser *p, + struct amdgpu_cs_chunk *chunk) +{ + struct drm_amdgpu_cs_chunk_syncobj *syncobj_deps; + unsigned num_deps; + int i, r; + + syncobj_deps = (struct drm_amdgpu_cs_chunk_syncobj *)chunk->kdata; + num_deps = chunk->length_dw * 4 / + sizeof(struct drm_amdgpu_cs_chunk_syncobj); + for (i = 0; i < num_deps; ++i) { + r = amdgpu_syncobj_lookup_and_add_to_sync(p, + syncobj_deps[i].handle, + syncobj_deps[i].point, + syncobj_deps[i].flags); + if (r) + return r; + } + return 0; } static int amdgpu_cs_process_syncobj_out_dep(struct amdgpu_cs_parser *p, struct amdgpu_cs_chunk *chunk) { + struct drm_amdgpu_cs_chunk_sem *deps; unsigned num_deps; int i; - struct drm_amdgpu_cs_chunk_sem *deps; + deps = (struct drm_amdgpu_cs_chunk_sem *)chunk->kdata; num_deps = chunk->length_dw * 4 / sizeof(struct drm_amdgpu_cs_chunk_sem); - p->post_dep_syncobjs = kmalloc_array(num_deps, - sizeof(struct drm_syncobj *), - GFP_KERNEL); - p->num_post_dep_syncobjs = 0; + p->post_deps = kmalloc_array(num_deps, sizeof(*p->post_deps), + GFP_KERNEL); + p->num_post_deps = 0; - if (!p->post_dep_syncobjs) + if (!p->post_deps) + return -ENOMEM; + + + for (i = 0; i < num_deps; ++i) { + p->post_deps[i].syncobj = + drm_syncobj_find(p->filp, deps[i].handle); + if (!p->post_deps[i].syncobj) + return -EINVAL; + p->post_deps[i].chain = NULL; + p->post_deps[i].point = 0; + p->num_post_deps++; + } + + return 0; +} + + +static int amdgpu_cs_process_syncobj_timeline_out_dep(struct amdgpu_cs_parser *p, + struct amdgpu_cs_chunk + *chunk) +{ + struct drm_amdgpu_cs_chunk_syncobj *syncobj_deps; + unsigned num_deps; + int i; + + syncobj_deps = (struct drm_amdgpu_cs_chunk_syncobj *)chunk->kdata; + num_deps = chunk->length_dw * 4 / + sizeof(struct drm_amdgpu_cs_chunk_syncobj); + + p->post_deps = kmalloc_array(num_deps, sizeof(*p->post_deps), + GFP_KERNEL); + p->num_post_deps = 0; + + if (!p->post_deps) return -ENOMEM; for (i = 0; i < num_deps; ++i) { - p->post_dep_syncobjs[i] = drm_syncobj_find(p->filp, deps[i].handle); - if (!p->post_dep_syncobjs[i]) + struct amdgpu_cs_post_dep *dep = &p->post_deps[i]; + + dep->chain = NULL; + if (syncobj_deps[i].point) { + dep->chain = kmalloc(sizeof(*dep->chain), GFP_KERNEL); + if (!dep->chain) + return -ENOMEM; + } + + dep->syncobj = drm_syncobj_find(p->filp, + syncobj_deps[i].handle); + if (!dep->syncobj) { + kfree(dep->chain); return -EINVAL; - p->num_post_dep_syncobjs++; + } + dep->point = syncobj_deps[i].point; + p->num_post_deps++; } + return 0; } @@ -1187,19 +1268,33 @@ static int amdgpu_cs_dependencies(struct amdgpu_device *adev, chunk = &p->chunks[i]; - if (chunk->chunk_id == AMDGPU_CHUNK_ID_DEPENDENCIES || - chunk->chunk_id == AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES) { + switch (chunk->chunk_id) { + case AMDGPU_CHUNK_ID_DEPENDENCIES: + case AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES: r = amdgpu_cs_process_fence_dep(p, chunk); if (r) return r; - } else if (chunk->chunk_id == AMDGPU_CHUNK_ID_SYNCOBJ_IN) { + break; + case AMDGPU_CHUNK_ID_SYNCOBJ_IN: r = amdgpu_cs_process_syncobj_in_dep(p, chunk); if (r) return r; - } else if (chunk->chunk_id == AMDGPU_CHUNK_ID_SYNCOBJ_OUT) { + break; + case AMDGPU_CHUNK_ID_SYNCOBJ_OUT: r = amdgpu_cs_process_syncobj_out_dep(p, chunk); if (r) return r; + break; + case AMDGPU_CHUNK_ID_SYNCOBJ_TIMELINE_WAIT: + r = amdgpu_cs_process_syncobj_timeline_in_dep(p, chunk); + if (r) + return r; + break; + case AMDGPU_CHUNK_ID_SYNCOBJ_TIMELINE_SIGNAL: + r = amdgpu_cs_process_syncobj_timeline_out_dep(p, chunk); + if (r) + return r; + break; } } @@ -1210,8 +1305,17 @@ static void amdgpu_cs_post_dependencies(struct amdgpu_cs_parser *p) { int i; - for (i = 0; i < p->num_post_dep_syncobjs; ++i) - drm_syncobj_replace_fence(p->post_dep_syncobjs[i], p->fence); + for (i = 0; i < p->num_post_deps; ++i) { + if (p->post_deps[i].chain && p->post_deps[i].point) { + drm_syncobj_add_point(p->post_deps[i].syncobj, + p->post_deps[i].chain, + p->fence, p->post_deps[i].point); + p->post_deps[i].chain = NULL; + } else { + drm_syncobj_replace_fence(p->post_deps[i].syncobj, + p->fence); + } + } } static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c index 7e22be7ca68a..54dd02a898b9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c @@ -92,15 +92,6 @@ int amdgpu_map_static_csa(struct amdgpu_device *adev, struct amdgpu_vm *vm, return -ENOMEM; } - r = amdgpu_vm_alloc_pts(adev, (*bo_va)->base.vm, csa_addr, - size); - if (r) { - DRM_ERROR("failed to allocate pts for static CSA, err=%d\n", r); - amdgpu_vm_bo_rmv(adev, *bo_va); - ttm_eu_backoff_reservation(&ticket, &list); - return r; - } - r = amdgpu_vm_bo_map(adev, *bo_va, csa_addr, 0, size, AMDGPU_PTE_READABLE | AMDGPU_PTE_WRITEABLE | AMDGPU_PTE_EXECUTABLE); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c index 7b526593eb77..a28a3d722ba2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c @@ -26,6 +26,7 @@ #include #include "amdgpu.h" #include "amdgpu_sched.h" +#include "amdgpu_ras.h" #define to_amdgpu_ctx_entity(e) \ container_of((e), struct amdgpu_ctx_entity, entity) @@ -344,6 +345,7 @@ static int amdgpu_ctx_query2(struct amdgpu_device *adev, { struct amdgpu_ctx *ctx; struct amdgpu_ctx_mgr *mgr; + uint32_t ras_counter; if (!fpriv) return -EINVAL; @@ -368,6 +370,21 @@ static int amdgpu_ctx_query2(struct amdgpu_device *adev, if (atomic_read(&ctx->guilty)) out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_GUILTY; + /*query ue count*/ + ras_counter = amdgpu_ras_query_error_count(adev, false); + /*ras counter is monotonic increasing*/ + if (ras_counter != ctx->ras_counter_ue) { + out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_RAS_UE; + ctx->ras_counter_ue = ras_counter; + } + + /*query ce count*/ + ras_counter = amdgpu_ras_query_error_count(adev, true); + if (ras_counter != ctx->ras_counter_ce) { + out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_RAS_CE; + ctx->ras_counter_ce = ras_counter; + } + mutex_unlock(&mgr->lock); return 0; } @@ -541,32 +558,26 @@ void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr) idr_init(&mgr->ctx_handles); } -void amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr) +long amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr, long timeout) { unsigned num_entities = amdgput_ctx_total_num_entities(); struct amdgpu_ctx *ctx; struct idr *idp; uint32_t id, i; - long max_wait = MAX_WAIT_SCHED_ENTITY_Q_EMPTY; idp = &mgr->ctx_handles; mutex_lock(&mgr->lock); idr_for_each_entry(idp, ctx, id) { - - if (!ctx->adev) { - mutex_unlock(&mgr->lock); - return; - } - for (i = 0; i < num_entities; i++) { struct drm_sched_entity *entity; entity = &ctx->entities[0][i].entity; - max_wait = drm_sched_entity_flush(entity, max_wait); + timeout = drm_sched_entity_flush(entity, timeout); } } mutex_unlock(&mgr->lock); + return timeout; } void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr) @@ -579,10 +590,6 @@ void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr) idp = &mgr->ctx_handles; idr_for_each_entry(idp, ctx, id) { - - if (!ctx->adev) - return; - if (kref_read(&ctx->refcount) != 1) { DRM_ERROR("ctx %p is still alive\n", ctx); continue; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h index b3b012c0a7da..5f1b54c9bcdb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h @@ -49,6 +49,8 @@ struct amdgpu_ctx { enum drm_sched_priority override_priority; struct mutex lock; atomic_t guilty; + uint32_t ras_counter_ce; + uint32_t ras_counter_ue; }; struct amdgpu_ctx_mgr { @@ -82,7 +84,7 @@ int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr); void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr); -void amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr); +long amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr, long timeout); void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index 4ae3ff9a1d4c..8930d66f2204 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c @@ -568,10 +568,9 @@ static ssize_t amdgpu_debugfs_sensor_read(struct file *f, char __user *buf, idx = *pos >> 2; valuesize = sizeof(values); - if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->read_sensor) - r = amdgpu_dpm_read_sensor(adev, idx, &values[0], &valuesize); - else - return -EINVAL; + r = amdgpu_dpm_read_sensor(adev, idx, &values[0], &valuesize); + if (r) + return r; if (size > valuesize) return -EINVAL; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 4f8fb4ecde34..d47354997e3c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -60,6 +60,7 @@ #include "amdgpu_pm.h" #include "amdgpu_xgmi.h" +#include "amdgpu_ras.h" MODULE_FIRMWARE("amdgpu/vega10_gpu_info.bin"); MODULE_FIRMWARE("amdgpu/vega12_gpu_info.bin"); @@ -1506,7 +1507,9 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) return -EAGAIN; } - adev->powerplay.pp_feature = amdgpu_pp_feature_mask; + adev->pm.pp_feature = amdgpu_pp_feature_mask; + if (amdgpu_sriov_vf(adev)) + adev->pm.pp_feature &= ~PP_GFXOFF_MASK; for (i = 0; i < adev->num_ip_blocks; i++) { if ((amdgpu_ip_block_mask & (1 << i)) == 0) { @@ -1638,6 +1641,10 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev) { int i, r; + r = amdgpu_ras_init(adev); + if (r) + return r; + for (i = 0; i < adev->num_ip_blocks; i++) { if (!adev->ip_blocks[i].status.valid) continue; @@ -1681,6 +1688,13 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev) } } + r = amdgpu_ib_pool_init(adev); + if (r) { + dev_err(adev->dev, "IB initialization failed (%d).\n", r); + amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_IB_INIT_FAIL, 0, r); + goto init_failed; + } + r = amdgpu_ucode_create_bo(adev); /* create ucode bo when sw_init complete*/ if (r) goto init_failed; @@ -1869,6 +1883,8 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev) { int i, r; + amdgpu_ras_pre_fini(adev); + if (adev->gmc.xgmi.num_physical_nodes > 1) amdgpu_xgmi_remove_device(adev); @@ -1917,6 +1933,7 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev) amdgpu_free_static_csa(&adev->virt.csa_obj); amdgpu_device_wb_fini(adev); amdgpu_device_vram_scratch_fini(adev); + amdgpu_ib_pool_fini(adev); } r = adev->ip_blocks[i].version->funcs->sw_fini((void *)adev); @@ -1937,6 +1954,8 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev) adev->ip_blocks[i].status.late_initialized = false; } + amdgpu_ras_fini(adev); + if (amdgpu_sriov_vf(adev)) if (amdgpu_virt_release_full_gpu(adev, false)) DRM_ERROR("failed to release exclusive mode on fini\n"); @@ -1999,6 +2018,10 @@ static void amdgpu_device_ip_late_init_func_handler(struct work_struct *work) r = amdgpu_device_enable_mgpu_fan_boost(); if (r) DRM_ERROR("enable mgpu fan boost failed (%d).\n", r); + + /*set to low pstate by default */ + amdgpu_xgmi_set_pstate(adev, 0); + } static void amdgpu_device_delay_enable_gfx_off(struct work_struct *work) @@ -2369,7 +2392,7 @@ static void amdgpu_device_xgmi_reset_func(struct work_struct *__work) adev->asic_reset_res = amdgpu_asic_reset(adev); if (adev->asic_reset_res) - DRM_WARN("ASIC reset failed with err r, %d for drm dev, %s", + DRM_WARN("ASIC reset failed with error, %d for drm dev, %s", adev->asic_reset_res, adev->ddev->unique); } @@ -2448,6 +2471,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, mutex_init(&adev->virt.vf_errors.lock); hash_init(adev->mn_hash); mutex_init(&adev->lock_reset); + mutex_init(&adev->virt.dpm_mutex); amdgpu_device_check_arguments(adev); @@ -2642,13 +2666,6 @@ fence_driver_init: /* Get a log2 for easy divisions. */ adev->mm_stats.log2_max_MBps = ilog2(max(1u, max_MBps)); - r = amdgpu_ib_pool_init(adev); - if (r) { - dev_err(adev->dev, "IB initialization failed (%d).\n", r); - amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_IB_INIT_FAIL, 0, r); - goto failed; - } - amdgpu_fbdev_init(adev); r = amdgpu_pm_sysfs_init(adev); @@ -2694,6 +2711,9 @@ fence_driver_init: goto failed; } + /* must succeed. */ + amdgpu_ras_post_init(adev); + return 0; failed: @@ -2726,7 +2746,6 @@ void amdgpu_device_fini(struct amdgpu_device *adev) else drm_atomic_helper_shutdown(adev->ddev); } - amdgpu_ib_pool_fini(adev); amdgpu_fence_driver_fini(adev); amdgpu_pm_sysfs_fini(adev); amdgpu_fbdev_fini(adev); @@ -3173,11 +3192,16 @@ static int amdgpu_device_recover_vram(struct amdgpu_device *adev) break; if (fence) { - r = dma_fence_wait_timeout(fence, false, tmo); + tmo = dma_fence_wait_timeout(fence, false, tmo); dma_fence_put(fence); fence = next; - if (r <= 0) + if (tmo == 0) { + r = -ETIMEDOUT; break; + } else if (tmo < 0) { + r = tmo; + break; + } } else { fence = next; } @@ -3188,8 +3212,8 @@ static int amdgpu_device_recover_vram(struct amdgpu_device *adev) tmo = dma_fence_wait_timeout(fence, false, tmo); dma_fence_put(fence); - if (r <= 0 || tmo <= 0) { - DRM_ERROR("recover vram bo from shadow failed\n"); + if (r < 0 || tmo <= 0) { + DRM_ERROR("recover vram bo from shadow failed, r is %ld, tmo is %ld\n", r, tmo); return -EIO; } @@ -3219,6 +3243,8 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, if (r) return r; + amdgpu_amdkfd_pre_reset(adev); + /* Resume IP prior to SMC */ r = amdgpu_device_ip_reinit_early_sriov(adev); if (r) @@ -3238,6 +3264,7 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, amdgpu_irq_gpu_reset_resume_helper(adev); r = amdgpu_ib_ring_tests(adev); + amdgpu_amdkfd_post_reset(adev); error: amdgpu_virt_init_data_exchange(adev); @@ -3370,7 +3397,7 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive, r = amdgpu_asic_reset(tmp_adev); if (r) { - DRM_ERROR("ASIC reset failed with err r, %d for drm dev, %s", + DRM_ERROR("ASIC reset failed with error, %d for drm dev, %s", r, tmp_adev->ddev->unique); break; } @@ -3387,6 +3414,11 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive, break; } } + + list_for_each_entry(tmp_adev, device_list_handle, + gmc.xgmi.head) { + amdgpu_ras_reserve_bad_pages(tmp_adev); + } } } @@ -3625,6 +3657,7 @@ static void amdgpu_device_get_min_pci_speed_width(struct amdgpu_device *adev, struct pci_dev *pdev = adev->pdev; enum pci_bus_speed cur_speed; enum pcie_link_width cur_width; + u32 ret = 1; *speed = PCI_SPEED_UNKNOWN; *width = PCIE_LNK_WIDTH_UNKNOWN; @@ -3632,6 +3665,10 @@ static void amdgpu_device_get_min_pci_speed_width(struct amdgpu_device *adev, while (pdev) { cur_speed = pcie_get_speed_cap(pdev); cur_width = pcie_get_width_cap(pdev); + ret = pcie_bandwidth_available(adev->pdev, NULL, + NULL, &cur_width); + if (!ret) + cur_width = PCIE_LNK_WIDTH_RESRV; if (cur_speed != PCI_SPEED_UNKNOWN) { if (*speed == PCI_SPEED_UNKNOWN) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c index 344967df3137..523b8ab6b04e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c @@ -904,3 +904,19 @@ amdgpu_get_vce_clock_state(void *handle, u32 idx) return NULL; } + +int amdgpu_dpm_get_sclk(struct amdgpu_device *adev, bool low) +{ + if (is_support_sw_smu(adev)) + return smu_get_sclk(&adev->smu, low); + else + return (adev)->powerplay.pp_funcs->get_sclk((adev)->powerplay.pp_handle, (low)); +} + +int amdgpu_dpm_get_mclk(struct amdgpu_device *adev, bool low) +{ + if (is_support_sw_smu(adev)) + return smu_get_mclk(&adev->smu, low); + else + return (adev)->powerplay.pp_funcs->get_mclk((adev)->powerplay.pp_handle, (low)); +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h index e871e022c129..dca35407879d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h @@ -260,9 +260,6 @@ enum amdgpu_pcie_gen { #define amdgpu_dpm_enable_bapm(adev, e) \ ((adev)->powerplay.pp_funcs->enable_bapm((adev)->powerplay.pp_handle, (e))) -#define amdgpu_dpm_read_sensor(adev, idx, value, size) \ - ((adev)->powerplay.pp_funcs->read_sensor((adev)->powerplay.pp_handle, (idx), (value), (size))) - #define amdgpu_dpm_set_fan_control_mode(adev, m) \ ((adev)->powerplay.pp_funcs->set_fan_control_mode((adev)->powerplay.pp_handle, (m))) @@ -281,18 +278,18 @@ enum amdgpu_pcie_gen { #define amdgpu_dpm_set_fan_speed_rpm(adev, s) \ ((adev)->powerplay.pp_funcs->set_fan_speed_rpm)((adev)->powerplay.pp_handle, (s)) -#define amdgpu_dpm_get_sclk(adev, l) \ - ((adev)->powerplay.pp_funcs->get_sclk((adev)->powerplay.pp_handle, (l))) - -#define amdgpu_dpm_get_mclk(adev, l) \ - ((adev)->powerplay.pp_funcs->get_mclk((adev)->powerplay.pp_handle, (l))) - #define amdgpu_dpm_force_performance_level(adev, l) \ ((adev)->powerplay.pp_funcs->force_performance_level((adev)->powerplay.pp_handle, (l))) #define amdgpu_dpm_get_current_power_state(adev) \ ((adev)->powerplay.pp_funcs->get_current_power_state((adev)->powerplay.pp_handle)) +#define amdgpu_smu_get_current_power_state(adev) \ + ((adev)->smu.ppt_funcs->get_current_power_state(&((adev)->smu))) + +#define amdgpu_smu_set_power_state(adev) \ + ((adev)->smu.ppt_funcs->set_power_state(&((adev)->smu))) + #define amdgpu_dpm_get_pp_num_states(adev, data) \ ((adev)->powerplay.pp_funcs->get_pp_num_states((adev)->powerplay.pp_handle, data)) @@ -448,6 +445,9 @@ struct amdgpu_pm { uint32_t smu_prv_buffer_size; struct amdgpu_bo *smu_prv_buffer; bool ac_power; + /* powerplay feature */ + uint32_t pp_feature; + }; #define R600_SSTU_DFLT 0 @@ -486,6 +486,8 @@ void amdgpu_dpm_print_ps_status(struct amdgpu_device *adev, u32 amdgpu_dpm_get_vblank_time(struct amdgpu_device *adev); u32 amdgpu_dpm_get_vrefresh(struct amdgpu_device *adev); void amdgpu_dpm_get_active_displays(struct amdgpu_device *adev); +int amdgpu_dpm_read_sensor(struct amdgpu_device *adev, enum amd_pp_sensors sensor, + void *data, uint32_t *size); bool amdgpu_is_internal_thermal_sensor(enum amdgpu_int_thermal_type sensor); @@ -504,4 +506,8 @@ enum amdgpu_pcie_gen amdgpu_get_pcie_gen_support(struct amdgpu_device *adev, struct amd_vce_state* amdgpu_get_vce_clock_state(void *handle, u32 idx); +extern int amdgpu_dpm_get_sclk(struct amdgpu_device *adev, bool low); + +extern int amdgpu_dpm_get_mclk(struct amdgpu_device *adev, bool low); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 8a0732088640..1e2cc9d68a05 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -74,9 +74,11 @@ * - 3.28.0 - Add AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES * - 3.29.0 - Add AMDGPU_IB_FLAG_RESET_GDS_MAX_WAVE_ID * - 3.30.0 - Add AMDGPU_SCHED_OP_CONTEXT_PRIORITY_OVERRIDE. + * - 3.31.0 - Add support for per-flip tiling attribute changes with DC + * - 3.32.0 - Add syncobj timeline support to AMDGPU_CS. */ #define KMS_DRIVER_MAJOR 3 -#define KMS_DRIVER_MINOR 30 +#define KMS_DRIVER_MINOR 32 #define KMS_DRIVER_PATCHLEVEL 0 int amdgpu_vram_limit = 0; @@ -117,8 +119,8 @@ uint amdgpu_pg_mask = 0xffffffff; uint amdgpu_sdma_phase_quantum = 32; char *amdgpu_disable_cu = NULL; char *amdgpu_virtual_display = NULL; -/* OverDrive(bit 14),gfxoff(bit 15),stutter mode(bit 17) disabled by default*/ -uint amdgpu_pp_feature_mask = 0xfffd3fff; +/* OverDrive(bit 14) disabled by default*/ +uint amdgpu_pp_feature_mask = 0xffffbfff; int amdgpu_ngg = 0; int amdgpu_prim_buf_per_se = 0; int amdgpu_pos_buf_per_se = 0; @@ -136,6 +138,8 @@ uint amdgpu_dc_feature_mask = 0; struct amdgpu_mgpu_info mgpu_info = { .mutex = __MUTEX_INITIALIZER(mgpu_info.mutex), }; +int amdgpu_ras_enable = -1; +uint amdgpu_ras_mask = 0xffffffff; /** * DOC: vramlimit (int) @@ -494,6 +498,21 @@ module_param_named(gpu_recovery, amdgpu_gpu_recovery, int, 0444); MODULE_PARM_DESC(emu_mode, "Emulation mode, (1 = enable, 0 = disable)"); module_param_named(emu_mode, amdgpu_emu_mode, int, 0444); +/** + * DOC: ras_enable (int) + * Enable RAS features on the GPU (0 = disable, 1 = enable, -1 = auto (default)) + */ +MODULE_PARM_DESC(ras_enable, "Enable RAS features on the GPU (0 = disable, 1 = enable, -1 = auto (default))"); +module_param_named(ras_enable, amdgpu_ras_enable, int, 0444); + +/** + * DOC: ras_mask (uint) + * Mask of RAS features to enable (default 0xffffffff), only valid when ras_enable == 1 + * See the flags in drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h + */ +MODULE_PARM_DESC(ras_mask, "Mask of RAS features to enable (default 0xffffffff), only valid when ras_enable == 1"); +module_param_named(ras_mask, amdgpu_ras_mask, uint, 0444); + /** * DOC: si_support (int) * Set SI support driver. This parameter works after set config CONFIG_DRM_AMDGPU_SI. For SI asic, when radeon driver is enabled, @@ -1159,13 +1178,14 @@ static int amdgpu_flush(struct file *f, fl_owner_t id) { struct drm_file *file_priv = f->private_data; struct amdgpu_fpriv *fpriv = file_priv->driver_priv; + long timeout = MAX_WAIT_SCHED_ENTITY_Q_EMPTY; - amdgpu_ctx_mgr_entity_flush(&fpriv->ctx_mgr); + timeout = amdgpu_ctx_mgr_entity_flush(&fpriv->ctx_mgr, timeout); + timeout = amdgpu_vm_wait_idle(&fpriv->vm, timeout); - return 0; + return timeout >= 0 ? 0 : timeout; } - static const struct file_operations amdgpu_driver_kms_fops = { .owner = THIS_MODULE, .open = drm_open, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c index 5cbde74b97dd..e47609218839 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c @@ -49,12 +49,11 @@ static int amdgpufb_open(struct fb_info *info, int user) { - struct amdgpu_fbdev *rfbdev = info->par; - struct amdgpu_device *adev = rfbdev->adev; - int ret = pm_runtime_get_sync(adev->ddev->dev); + struct drm_fb_helper *fb_helper = info->par; + int ret = pm_runtime_get_sync(fb_helper->dev->dev); if (ret < 0 && ret != -EACCES) { - pm_runtime_mark_last_busy(adev->ddev->dev); - pm_runtime_put_autosuspend(adev->ddev->dev); + pm_runtime_mark_last_busy(fb_helper->dev->dev); + pm_runtime_put_autosuspend(fb_helper->dev->dev); return ret; } return 0; @@ -63,11 +62,10 @@ amdgpufb_open(struct fb_info *info, int user) static int amdgpufb_release(struct fb_info *info, int user) { - struct amdgpu_fbdev *rfbdev = info->par; - struct amdgpu_device *adev = rfbdev->adev; + struct drm_fb_helper *fb_helper = info->par; - pm_runtime_mark_last_busy(adev->ddev->dev); - pm_runtime_put_autosuspend(adev->ddev->dev); + pm_runtime_mark_last_busy(fb_helper->dev->dev); + pm_runtime_put_autosuspend(fb_helper->dev->dev); return 0; } @@ -233,9 +231,6 @@ static int amdgpufb_create(struct drm_fb_helper *helper, goto out; } - info->par = rfbdev; - info->skip_vt_switch = true; - ret = amdgpu_display_framebuffer_init(adev->ddev, &rfbdev->rfb, &mode_cmd, gobj); if (ret) { @@ -248,10 +243,6 @@ static int amdgpufb_create(struct drm_fb_helper *helper, /* setup helper */ rfbdev->helper.fb = fb; - strcpy(info->fix.id, "amdgpudrmfb"); - - drm_fb_helper_fill_fix(info, fb->pitches[0], fb->format->depth); - info->fbops = &amdgpufb_ops; tmp = amdgpu_bo_gpu_offset(abo) - adev->gmc.vram_start; @@ -260,7 +251,7 @@ static int amdgpufb_create(struct drm_fb_helper *helper, info->screen_base = amdgpu_bo_kptr(abo); info->screen_size = amdgpu_bo_size(abo); - drm_fb_helper_fill_var(info, &rfbdev->helper, sizes->fb_width, sizes->fb_height); + drm_fb_helper_fill_info(info, &rfbdev->helper, sizes); /* setup aperture base/size for vesafb takeover */ info->apertures->ranges[0].base = adev->ddev->mode_config.fb_base; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index ee47c11e92ce..4dee2326b29c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -136,8 +136,9 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f, { struct amdgpu_device *adev = ring->adev; struct amdgpu_fence *fence; - struct dma_fence *old, **ptr; + struct dma_fence __rcu **ptr; uint32_t seq; + int r; fence = kmem_cache_alloc(amdgpu_fence_slab, GFP_KERNEL); if (fence == NULL) @@ -153,15 +154,24 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f, seq, flags | AMDGPU_FENCE_FLAG_INT); ptr = &ring->fence_drv.fences[seq & ring->fence_drv.num_fences_mask]; + if (unlikely(rcu_dereference_protected(*ptr, 1))) { + struct dma_fence *old; + + rcu_read_lock(); + old = dma_fence_get_rcu_safe(ptr); + rcu_read_unlock(); + + if (old) { + r = dma_fence_wait(old, false); + dma_fence_put(old); + if (r) + return r; + } + } + /* This function can't be called concurrently anyway, otherwise * emitting the fence would mess up the hardware ring buffer. */ - old = rcu_dereference_protected(*ptr, 1); - if (old && !dma_fence_is_signaled(old)) { - DRM_INFO("rcu slot is busy\n"); - dma_fence_wait(old, false); - } - rcu_assign_pointer(*ptr, dma_fence_get(&fence->base)); *f = &fence->base; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index d21dd2f369da..d4fcf5475464 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -31,6 +31,7 @@ #include #include "amdgpu.h" #include "amdgpu_display.h" +#include "amdgpu_xgmi.h" void amdgpu_gem_object_free(struct drm_gem_object *gobj) { @@ -627,11 +628,6 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data, switch (args->operation) { case AMDGPU_VA_OP_MAP: - r = amdgpu_vm_alloc_pts(adev, bo_va->base.vm, args->va_address, - args->map_size); - if (r) - goto error_backoff; - va_flags = amdgpu_gmc_get_pte_flags(adev, args->flags); r = amdgpu_vm_bo_map(adev, bo_va, args->va_address, args->offset_in_bo, args->map_size, @@ -647,11 +643,6 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data, args->map_size); break; case AMDGPU_VA_OP_REPLACE: - r = amdgpu_vm_alloc_pts(adev, bo_va->base.vm, args->va_address, - args->map_size); - if (r) - goto error_backoff; - va_flags = amdgpu_gmc_get_pte_flags(adev, args->flags); r = amdgpu_vm_bo_replace_map(adev, bo_va, args->va_address, args->offset_in_bo, args->map_size, @@ -678,6 +669,7 @@ int amdgpu_gem_op_ioctl(struct drm_device *dev, void *data, struct amdgpu_device *adev = dev->dev_private; struct drm_amdgpu_gem_op *args = data; struct drm_gem_object *gobj; + struct amdgpu_vm_bo_base *base; struct amdgpu_bo *robj; int r; @@ -716,6 +708,15 @@ int amdgpu_gem_op_ioctl(struct drm_device *dev, void *data, amdgpu_bo_unreserve(robj); break; } + for (base = robj->vm_bo; base; base = base->next) + if (amdgpu_xgmi_same_hive(amdgpu_ttm_adev(robj->tbo.bdev), + amdgpu_ttm_adev(base->vm->root.base.bo->tbo.bdev))) { + r = -EINVAL; + amdgpu_bo_unreserve(robj); + goto out; + } + + robj->preferred_domains = args->value & (AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT | AMDGPU_GEM_DOMAIN_CPU); @@ -745,17 +746,25 @@ int amdgpu_mode_dumb_create(struct drm_file *file_priv, struct amdgpu_device *adev = dev->dev_private; struct drm_gem_object *gobj; uint32_t handle; + u64 flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; u32 domain; int r; + /* + * The buffer returned from this function should be cleared, but + * it can only be done if the ring is enabled or we'll fail to + * create the buffer. + */ + if (adev->mman.buffer_funcs_enabled) + flags |= AMDGPU_GEM_CREATE_VRAM_CLEARED; + args->pitch = amdgpu_align_pitch(adev, args->width, DIV_ROUND_UP(args->bpp, 8), 0); args->size = (u64)args->pitch * args->height; args->size = ALIGN(args->size, PAGE_SIZE); domain = amdgpu_bo_get_preferred_pin_domain(adev, amdgpu_display_supported_domains(adev)); - r = amdgpu_gem_object_create(adev, args->size, 0, domain, - AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED, + r = amdgpu_gem_object_create(adev, args->size, 0, domain, flags, ttm_bo_type_device, NULL, &gobj); if (r) return -ENOMEM; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c index 97a60da62004..997932ebbb83 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c @@ -390,7 +390,7 @@ void amdgpu_gfx_compute_mqd_sw_fini(struct amdgpu_device *adev) void amdgpu_gfx_off_ctrl(struct amdgpu_device *adev, bool enable) { - if (!(adev->powerplay.pp_feature & PP_GFXOFF_MASK)) + if (!(adev->pm.pp_feature & PP_GFXOFF_MASK)) return; if (!adev->powerplay.pp_funcs || !adev->powerplay.pp_funcs->set_powergating_by_smu) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h index f790e15bcd08..09fc53af3d35 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h @@ -258,6 +258,9 @@ struct amdgpu_gfx { /* pipe reservation */ struct mutex pipe_reserve_mutex; DECLARE_BITMAP (pipe_reserve_bitmap, AMDGPU_MAX_COMPUTE_QUEUES); + + /*ras */ + struct ras_common_if *ras_if; }; #define amdgpu_gfx_get_gpu_clock_counter(adev) (adev)->gfx.funcs->get_gpu_clock_counter((adev)) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c index d73367cab4f3..250d9212cc38 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c @@ -79,6 +79,33 @@ uint64_t amdgpu_gmc_pd_addr(struct amdgpu_bo *bo) return pd_addr; } +/** + * amdgpu_gmc_set_pte_pde - update the page tables using CPU + * + * @adev: amdgpu_device pointer + * @cpu_pt_addr: cpu address of the page table + * @gpu_page_idx: entry in the page table to update + * @addr: dst addr to write into pte/pde + * @flags: access flags + * + * Update the page tables using CPU. + */ +int amdgpu_gmc_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr, + uint32_t gpu_page_idx, uint64_t addr, + uint64_t flags) +{ + void __iomem *ptr = (void *)cpu_pt_addr; + uint64_t value; + + /* + * The following is for PTE only. GART does not have PDEs. + */ + value = addr & 0x0000FFFFFFFFF000ULL; + value |= flags; + writeq(value, ptr + (gpu_page_idx * 8)); + return 0; +} + /** * amdgpu_gmc_agp_addr - return the address in the AGP address space * @@ -213,3 +240,58 @@ void amdgpu_gmc_agp_location(struct amdgpu_device *adev, struct amdgpu_gmc *mc) dev_info(adev->dev, "AGP: %lluM 0x%016llX - 0x%016llX\n", mc->agp_size >> 20, mc->agp_start, mc->agp_end); } + +/** + * amdgpu_gmc_filter_faults - filter VM faults + * + * @adev: amdgpu device structure + * @addr: address of the VM fault + * @pasid: PASID of the process causing the fault + * @timestamp: timestamp of the fault + * + * Returns: + * True if the fault was filtered and should not be processed further. + * False if the fault is a new one and needs to be handled. + */ +bool amdgpu_gmc_filter_faults(struct amdgpu_device *adev, uint64_t addr, + uint16_t pasid, uint64_t timestamp) +{ + struct amdgpu_gmc *gmc = &adev->gmc; + + uint64_t stamp, key = addr << 4 | pasid; + struct amdgpu_gmc_fault *fault; + uint32_t hash; + + /* If we don't have space left in the ring buffer return immediately */ + stamp = max(timestamp, AMDGPU_GMC_FAULT_TIMEOUT + 1) - + AMDGPU_GMC_FAULT_TIMEOUT; + if (gmc->fault_ring[gmc->last_fault].timestamp >= stamp) + return true; + + /* Try to find the fault in the hash */ + hash = hash_64(key, AMDGPU_GMC_FAULT_HASH_ORDER); + fault = &gmc->fault_ring[gmc->fault_hash[hash].idx]; + while (fault->timestamp >= stamp) { + uint64_t tmp; + + if (fault->key == key) + return true; + + tmp = fault->timestamp; + fault = &gmc->fault_ring[fault->next]; + + /* Check if the entry was reused */ + if (fault->timestamp >= tmp) + break; + } + + /* Add the fault to the ring */ + fault = &gmc->fault_ring[gmc->last_fault]; + fault->key = key; + fault->timestamp = timestamp; + + /* And update the hash */ + fault->next = gmc->fault_hash[hash].idx; + gmc->fault_hash[hash].idx = gmc->last_fault++; + return false; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h index 81e6070d255b..071145ac67b5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h @@ -43,8 +43,34 @@ */ #define AMDGPU_GMC_HOLE_MASK 0x0000ffffffffffffULL +/* + * Ring size as power of two for the log of recent faults. + */ +#define AMDGPU_GMC_FAULT_RING_ORDER 8 +#define AMDGPU_GMC_FAULT_RING_SIZE (1 << AMDGPU_GMC_FAULT_RING_ORDER) + +/* + * Hash size as power of two for the log of recent faults + */ +#define AMDGPU_GMC_FAULT_HASH_ORDER 8 +#define AMDGPU_GMC_FAULT_HASH_SIZE (1 << AMDGPU_GMC_FAULT_HASH_ORDER) + +/* + * Number of IH timestamp ticks until a fault is considered handled + */ +#define AMDGPU_GMC_FAULT_TIMEOUT 5000ULL + struct firmware; +/* + * GMC page fault information + */ +struct amdgpu_gmc_fault { + uint64_t timestamp; + uint64_t next:AMDGPU_GMC_FAULT_RING_ORDER; + uint64_t key:52; +}; + /* * VMHUB structures, functions & helpers */ @@ -71,12 +97,6 @@ struct amdgpu_gmc_funcs { /* Change the VMID -> PASID mapping */ void (*emit_pasid_mapping)(struct amdgpu_ring *ring, unsigned vmid, unsigned pasid); - /* write pte/pde updates using the cpu */ - int (*set_pte_pde)(struct amdgpu_device *adev, - void *cpu_pt_addr, /* cpu addr of page table */ - uint32_t gpu_page_idx, /* pte/pde to update */ - uint64_t addr, /* addr to write into pte/pde */ - uint64_t flags); /* access flags */ /* enable/disable PRT support */ void (*set_prt)(struct amdgpu_device *adev, bool enable); /* set pte flags based per asic */ @@ -147,15 +167,22 @@ struct amdgpu_gmc { struct kfd_vm_fault_info *vm_fault_info; atomic_t vm_fault_info_updated; + struct amdgpu_gmc_fault fault_ring[AMDGPU_GMC_FAULT_RING_SIZE]; + struct { + uint64_t idx:AMDGPU_GMC_FAULT_RING_ORDER; + } fault_hash[AMDGPU_GMC_FAULT_HASH_SIZE]; + uint64_t last_fault:AMDGPU_GMC_FAULT_RING_ORDER; + const struct amdgpu_gmc_funcs *gmc_funcs; struct amdgpu_xgmi xgmi; + struct amdgpu_irq_src ecc_irq; + struct ras_common_if *ras_if; }; #define amdgpu_gmc_flush_gpu_tlb(adev, vmid, type) (adev)->gmc.gmc_funcs->flush_gpu_tlb((adev), (vmid), (type)) #define amdgpu_gmc_emit_flush_gpu_tlb(r, vmid, addr) (r)->adev->gmc.gmc_funcs->emit_flush_gpu_tlb((r), (vmid), (addr)) #define amdgpu_gmc_emit_pasid_mapping(r, vmid, pasid) (r)->adev->gmc.gmc_funcs->emit_pasid_mapping((r), (vmid), (pasid)) -#define amdgpu_gmc_set_pte_pde(adev, pt, idx, addr, flags) (adev)->gmc.gmc_funcs->set_pte_pde((adev), (pt), (idx), (addr), (flags)) #define amdgpu_gmc_get_vm_pde(adev, level, dst, flags) (adev)->gmc.gmc_funcs->get_vm_pde((adev), (level), (dst), (flags)) #define amdgpu_gmc_get_pte_flags(adev, flags) (adev)->gmc.gmc_funcs->get_vm_pte_flags((adev),(flags)) @@ -189,6 +216,9 @@ static inline uint64_t amdgpu_gmc_sign_extend(uint64_t addr) void amdgpu_gmc_get_pde_for_bo(struct amdgpu_bo *bo, int level, uint64_t *addr, uint64_t *flags); +int amdgpu_gmc_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr, + uint32_t gpu_page_idx, uint64_t addr, + uint64_t flags); uint64_t amdgpu_gmc_pd_addr(struct amdgpu_bo *bo); uint64_t amdgpu_gmc_agp_addr(struct ttm_buffer_object *bo); void amdgpu_gmc_vram_location(struct amdgpu_device *adev, struct amdgpu_gmc *mc, @@ -197,5 +227,7 @@ void amdgpu_gmc_gart_location(struct amdgpu_device *adev, struct amdgpu_gmc *mc); void amdgpu_gmc_agp_location(struct amdgpu_device *adev, struct amdgpu_gmc *mc); +bool amdgpu_gmc_filter_faults(struct amdgpu_device *adev, uint64_t addr, + uint16_t pasid, uint64_t timestamp); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c index da7b1b92d9cf..62591d081856 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c @@ -36,6 +36,47 @@ struct amdgpu_gtt_node { struct ttm_buffer_object *tbo; }; +/** + * DOC: mem_info_gtt_total + * + * The amdgpu driver provides a sysfs API for reporting current total size of + * the GTT. + * The file mem_info_gtt_total is used for this, and returns the total size of + * the GTT block, in bytes + */ +static ssize_t amdgpu_mem_info_gtt_total_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + + return snprintf(buf, PAGE_SIZE, "%llu\n", + (adev->mman.bdev.man[TTM_PL_TT].size) * PAGE_SIZE); +} + +/** + * DOC: mem_info_gtt_used + * + * The amdgpu driver provides a sysfs API for reporting current total amount of + * used GTT. + * The file mem_info_gtt_used is used for this, and returns the current used + * size of the GTT block, in bytes + */ +static ssize_t amdgpu_mem_info_gtt_used_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + + return snprintf(buf, PAGE_SIZE, "%llu\n", + amdgpu_gtt_mgr_usage(&adev->mman.bdev.man[TTM_PL_TT])); +} + +static DEVICE_ATTR(mem_info_gtt_total, S_IRUGO, + amdgpu_mem_info_gtt_total_show, NULL); +static DEVICE_ATTR(mem_info_gtt_used, S_IRUGO, + amdgpu_mem_info_gtt_used_show, NULL); + /** * amdgpu_gtt_mgr_init - init GTT manager and DRM MM * @@ -50,6 +91,7 @@ static int amdgpu_gtt_mgr_init(struct ttm_mem_type_manager *man, struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev); struct amdgpu_gtt_mgr *mgr; uint64_t start, size; + int ret; mgr = kzalloc(sizeof(*mgr), GFP_KERNEL); if (!mgr) @@ -61,6 +103,18 @@ static int amdgpu_gtt_mgr_init(struct ttm_mem_type_manager *man, spin_lock_init(&mgr->lock); atomic64_set(&mgr->available, p_size); man->priv = mgr; + + ret = device_create_file(adev->dev, &dev_attr_mem_info_gtt_total); + if (ret) { + DRM_ERROR("Failed to create device file mem_info_gtt_total\n"); + return ret; + } + ret = device_create_file(adev->dev, &dev_attr_mem_info_gtt_used); + if (ret) { + DRM_ERROR("Failed to create device file mem_info_gtt_used\n"); + return ret; + } + return 0; } @@ -74,12 +128,17 @@ static int amdgpu_gtt_mgr_init(struct ttm_mem_type_manager *man, */ static int amdgpu_gtt_mgr_fini(struct ttm_mem_type_manager *man) { + struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev); struct amdgpu_gtt_mgr *mgr = man->priv; spin_lock(&mgr->lock); drm_mm_takedown(&mgr->mm); spin_unlock(&mgr->lock); kfree(mgr); man->priv = NULL; + + device_remove_file(adev->dev, &dev_attr_mem_info_gtt_total); + device_remove_file(adev->dev, &dev_attr_mem_info_gtt_used); + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index 0b8ef2d27d6b..fe393a46f881 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -35,6 +35,7 @@ #include "amdgpu_trace.h" #define AMDGPU_IB_TEST_TIMEOUT msecs_to_jiffies(1000) +#define AMDGPU_IB_TEST_GFX_XGMI_TIMEOUT msecs_to_jiffies(2000) /* * IB @@ -344,6 +345,8 @@ int amdgpu_ib_ring_tests(struct amdgpu_device *adev) * cost waiting for it coming back under RUNTIME only */ tmo_gfx = 8 * AMDGPU_IB_TEST_TIMEOUT; + } else if (adev->gmc.xgmi.hive_id) { + tmo_gfx = AMDGPU_IB_TEST_GFX_XGMI_TIMEOUT; } for (i = 0; i < adev->num_rings; ++i) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c index 1c50be3ab8a9..934dfdcb4e73 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c @@ -142,6 +142,7 @@ void amdgpu_ih_ring_fini(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih) */ int amdgpu_ih_process(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih) { + unsigned int count = AMDGPU_IH_MAX_NUM_IVS; u32 wptr; if (!ih->enabled || adev->shutdown) @@ -159,7 +160,7 @@ restart_ih: /* Order reading of wptr vs. reading of IH ring data */ rmb(); - while (ih->rptr != wptr) { + while (ih->rptr != wptr && --count) { amdgpu_irq_dispatch(adev, ih); ih->rptr &= ih->ptr_mask; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h index 113a1ba13d4a..4e0bb645176d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h @@ -24,6 +24,9 @@ #ifndef __AMDGPU_IH_H__ #define __AMDGPU_IH_H__ +/* Maximum number of IVs processed at once */ +#define AMDGPU_IH_MAX_NUM_IVS 32 + struct amdgpu_device; struct amdgpu_iv_entry; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index e860412043bb..b17d0545728e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -39,6 +39,7 @@ #include "amdgpu_amdkfd.h" #include "amdgpu_gem.h" #include "amdgpu_display.h" +#include "amdgpu_ras.h" static void amdgpu_unregister_gpu_instance(struct amdgpu_device *adev) { @@ -296,6 +297,17 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info, fw_info->ver = adev->pm.fw_version; fw_info->feature = 0; break; + case AMDGPU_INFO_FW_TA: + if (query_fw->index > 1) + return -EINVAL; + if (query_fw->index == 0) { + fw_info->ver = adev->psp.ta_fw_version; + fw_info->feature = adev->psp.ta_xgmi_ucode_version; + } else { + fw_info->ver = adev->psp.ta_fw_version; + fw_info->feature = adev->psp.ta_ras_ucode_version; + } + break; case AMDGPU_INFO_FW_SDMA: if (query_fw->index >= adev->sdma.num_instances) return -EINVAL; @@ -684,6 +696,10 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file if (adev->pm.dpm_enabled) { dev_info.max_engine_clock = amdgpu_dpm_get_sclk(adev, false) * 10; dev_info.max_memory_clock = amdgpu_dpm_get_mclk(adev, false) * 10; + } else if (amdgpu_sriov_vf(adev) && amdgim_is_hwperf(adev) && + adev->virt.ops->get_pp_clk) { + dev_info.max_engine_clock = amdgpu_virt_get_sclk(adev, false) * 10; + dev_info.max_memory_clock = amdgpu_virt_get_mclk(adev, false) * 10; } else { dev_info.max_engine_clock = adev->clock.default_sclk * 10; dev_info.max_memory_clock = adev->clock.default_mclk * 10; @@ -909,6 +925,18 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file case AMDGPU_INFO_VRAM_LOST_COUNTER: ui32 = atomic_read(&adev->vram_lost_counter); return copy_to_user(out, &ui32, min(size, 4u)) ? -EFAULT : 0; + case AMDGPU_INFO_RAS_ENABLED_FEATURES: { + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + uint64_t ras_mask; + + if (!ras) + return -EINVAL; + ras_mask = (uint64_t)ras->supported << 32 | ras->features; + + return copy_to_user(out, &ras_mask, + min_t(u64, size, sizeof(ras_mask))) ? + -EFAULT : 0; + } default: DRM_DEBUG_KMS("Invalid request %d\n", info->query); return -EINVAL; @@ -1328,6 +1356,16 @@ static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data) seq_printf(m, "ASD feature version: %u, firmware version: 0x%08x\n", fw_info.feature, fw_info.ver); + query_fw.fw_type = AMDGPU_INFO_FW_TA; + for (i = 0; i < 2; i++) { + query_fw.index = i; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + continue; + seq_printf(m, "TA %s feature version: %u, firmware version: 0x%08x\n", + i ? "RAS" : "XGMI", fw_info.feature, fw_info.ver); + } + /* SMC */ query_fw.fw_type = AMDGPU_INFO_FW_SMC; ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index 889e443eeee7..2e9e3db778c6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -58,7 +58,7 @@ struct amdgpu_hpd; #define to_amdgpu_encoder(x) container_of(x, struct amdgpu_encoder, base) #define to_amdgpu_framebuffer(x) container_of(x, struct amdgpu_framebuffer, base) -#define to_dm_plane_state(x) container_of(x, struct dm_plane_state, base); +#define to_dm_plane_state(x) container_of(x, struct dm_plane_state, base) #define AMDGPU_MAX_HPD_PINS 6 #define AMDGPU_MAX_CRTCS 6 @@ -406,7 +406,7 @@ struct amdgpu_crtc { struct amdgpu_flip_work *pflip_works; enum amdgpu_flip_status pflip_status; int deferred_flip_completion; - u64 last_flip_vblank; + u32 last_flip_vblank; /* pll sharing */ struct amdgpu_atom_ss ss; bool ss_enabled; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index 220a6a7b1bc1..c430e8259038 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h @@ -72,6 +72,8 @@ struct amdgpu_bo_va { /* If the mappings are cleared or filled */ bool cleared; + + bool is_xgmi; }; struct amdgpu_bo { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c index a7adb7b6bd98..4b7a076eea9c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c @@ -28,6 +28,7 @@ #include "amdgpu_pm.h" #include "amdgpu_dpm.h" #include "amdgpu_display.h" +#include "amdgpu_smu.h" #include "atom.h" #include #include @@ -80,6 +81,27 @@ void amdgpu_pm_acpi_event_handler(struct amdgpu_device *adev) } } +int amdgpu_dpm_read_sensor(struct amdgpu_device *adev, enum amd_pp_sensors sensor, + void *data, uint32_t *size) +{ + int ret = 0; + + if (!data || !size) + return -EINVAL; + + if (is_support_sw_smu(adev)) + ret = smu_read_sensor(&adev->smu, sensor, data, size); + else { + if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->read_sensor) + ret = adev->powerplay.pp_funcs->read_sensor((adev)->powerplay.pp_handle, + sensor, data, size); + else + ret = -EINVAL; + } + + return ret; +} + /** * DOC: power_dpm_state * @@ -122,7 +144,9 @@ static ssize_t amdgpu_get_dpm_state(struct device *dev, struct amdgpu_device *adev = ddev->dev_private; enum amd_pm_state_type pm; - if (adev->powerplay.pp_funcs->get_current_power_state) + if (adev->smu.ppt_funcs->get_current_power_state) + pm = amdgpu_smu_get_current_power_state(adev); + else if (adev->powerplay.pp_funcs->get_current_power_state) pm = amdgpu_dpm_get_current_power_state(adev); else pm = adev->pm.dpm.user_state; @@ -240,7 +264,9 @@ static ssize_t amdgpu_get_dpm_forced_performance_level(struct device *dev, (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return snprintf(buf, PAGE_SIZE, "off\n"); - if (adev->powerplay.pp_funcs->get_performance_level) + if (is_support_sw_smu(adev)) + level = smu_get_performance_level(&adev->smu); + else if (adev->powerplay.pp_funcs->get_performance_level) level = amdgpu_dpm_get_performance_level(adev); else level = adev->pm.dpm.forced_level; @@ -273,7 +299,9 @@ static ssize_t amdgpu_set_dpm_forced_performance_level(struct device *dev, (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - if (adev->powerplay.pp_funcs->get_performance_level) + if (is_support_sw_smu(adev)) + current_level = smu_get_performance_level(&adev->smu); + else if (adev->powerplay.pp_funcs->get_performance_level) current_level = amdgpu_dpm_get_performance_level(adev); if (strncmp("low", buf, strlen("low")) == 0) { @@ -299,10 +327,35 @@ static ssize_t amdgpu_set_dpm_forced_performance_level(struct device *dev, goto fail; } + if (amdgpu_sriov_vf(adev)) { + if (amdgim_is_hwperf(adev) && + adev->virt.ops->force_dpm_level) { + mutex_lock(&adev->pm.mutex); + adev->virt.ops->force_dpm_level(adev, level); + mutex_unlock(&adev->pm.mutex); + return count; + } else { + return -EINVAL; + } + } + if (current_level == level) return count; - if (adev->powerplay.pp_funcs->force_performance_level) { + if (is_support_sw_smu(adev)) { + mutex_lock(&adev->pm.mutex); + if (adev->pm.dpm.thermal_active) { + count = -EINVAL; + mutex_unlock(&adev->pm.mutex); + goto fail; + } + ret = smu_force_performance_level(&adev->smu, level); + if (ret) + count = -EINVAL; + else + adev->pm.dpm.forced_level = level; + mutex_unlock(&adev->pm.mutex); + } else if (adev->powerplay.pp_funcs->force_performance_level) { mutex_lock(&adev->pm.mutex); if (adev->pm.dpm.thermal_active) { count = -EINVAL; @@ -328,9 +381,13 @@ static ssize_t amdgpu_get_pp_num_states(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; struct pp_states_info data; - int i, buf_len; + int i, buf_len, ret; - if (adev->powerplay.pp_funcs->get_pp_num_states) + if (is_support_sw_smu(adev)) { + ret = smu_get_power_num_states(&adev->smu, &data); + if (ret) + return ret; + } else if (adev->powerplay.pp_funcs->get_pp_num_states) amdgpu_dpm_get_pp_num_states(adev, &data); buf_len = snprintf(buf, PAGE_SIZE, "states: %d\n", data.nums); @@ -351,23 +408,29 @@ static ssize_t amdgpu_get_pp_cur_state(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; struct pp_states_info data; + struct smu_context *smu = &adev->smu; enum amd_pm_state_type pm = 0; - int i = 0; + int i = 0, ret = 0; - if (adev->powerplay.pp_funcs->get_current_power_state + if (is_support_sw_smu(adev)) { + pm = smu_get_current_power_state(smu); + ret = smu_get_power_num_states(smu, &data); + if (ret) + return ret; + } else if (adev->powerplay.pp_funcs->get_current_power_state && adev->powerplay.pp_funcs->get_pp_num_states) { pm = amdgpu_dpm_get_current_power_state(adev); amdgpu_dpm_get_pp_num_states(adev, &data); - - for (i = 0; i < data.nums; i++) { - if (pm == data.states[i]) - break; - } - - if (i == data.nums) - i = -EINVAL; } + for (i = 0; i < data.nums; i++) { + if (pm == data.states[i]) + break; + } + + if (i == data.nums) + i = -EINVAL; + return snprintf(buf, PAGE_SIZE, "%d\n", i); } @@ -397,6 +460,8 @@ static ssize_t amdgpu_set_pp_force_state(struct device *dev, if (strlen(buf) == 1) adev->pp_force_state_enabled = false; + else if (is_support_sw_smu(adev)) + adev->pp_force_state_enabled = false; else if (adev->powerplay.pp_funcs->dispatch_tasks && adev->powerplay.pp_funcs->get_pp_num_states) { struct pp_states_info data; @@ -442,7 +507,12 @@ static ssize_t amdgpu_get_pp_table(struct device *dev, char *table = NULL; int size; - if (adev->powerplay.pp_funcs->get_pp_table) + if (is_support_sw_smu(adev)) { + size = smu_sys_get_pp_table(&adev->smu, (void **)&table); + if (size < 0) + return size; + } + else if (adev->powerplay.pp_funcs->get_pp_table) size = amdgpu_dpm_get_pp_table(adev, &table); else return 0; @@ -462,8 +532,13 @@ static ssize_t amdgpu_set_pp_table(struct device *dev, { struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; + int ret = 0; - if (adev->powerplay.pp_funcs->set_pp_table) + if (is_support_sw_smu(adev)) { + ret = smu_sys_set_pp_table(&adev->smu, (void *)buf, count); + if (ret) + return ret; + } else if (adev->powerplay.pp_funcs->set_pp_table) amdgpu_dpm_set_pp_table(adev, buf, count); return count; @@ -586,19 +661,29 @@ static ssize_t amdgpu_set_pp_od_clk_voltage(struct device *dev, tmp_str++; } - if (adev->powerplay.pp_funcs->odn_edit_dpm_table) - ret = amdgpu_dpm_odn_edit_dpm_table(adev, type, + if (is_support_sw_smu(adev)) { + ret = smu_od_edit_dpm_table(&adev->smu, type, + parameter, parameter_size); + + if (ret) + return -EINVAL; + } else { + if (adev->powerplay.pp_funcs->odn_edit_dpm_table) + ret = amdgpu_dpm_odn_edit_dpm_table(adev, type, parameter, parameter_size); - if (ret) - return -EINVAL; - - if (type == PP_OD_COMMIT_DPM_TABLE) { - if (adev->powerplay.pp_funcs->dispatch_tasks) { - amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL); - return count; - } else { + if (ret) return -EINVAL; + + if (type == PP_OD_COMMIT_DPM_TABLE) { + if (adev->powerplay.pp_funcs->dispatch_tasks) { + amdgpu_dpm_dispatch_task(adev, + AMD_PP_TASK_READJUST_POWER_STATE, + NULL); + return count; + } else { + return -EINVAL; + } } } @@ -613,7 +698,13 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev, struct amdgpu_device *adev = ddev->dev_private; uint32_t size = 0; - if (adev->powerplay.pp_funcs->print_clock_levels) { + if (is_support_sw_smu(adev)) { + size = smu_print_clk_levels(&adev->smu, OD_SCLK, buf); + size += smu_print_clk_levels(&adev->smu, OD_MCLK, buf+size); + size += smu_print_clk_levels(&adev->smu, OD_VDDC_CURVE, buf+size); + size += smu_print_clk_levels(&adev->smu, OD_RANGE, buf+size); + return size; + } else if (adev->powerplay.pp_funcs->print_clock_levels) { size = amdgpu_dpm_print_clock_levels(adev, OD_SCLK, buf); size += amdgpu_dpm_print_clock_levels(adev, OD_MCLK, buf+size); size += amdgpu_dpm_print_clock_levels(adev, OD_VDDC_CURVE, buf+size); @@ -711,7 +802,13 @@ static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; - if (adev->powerplay.pp_funcs->print_clock_levels) + if (amdgpu_sriov_vf(adev) && amdgim_is_hwperf(adev) && + adev->virt.ops->get_pp_clk) + return adev->virt.ops->get_pp_clk(adev, PP_SCLK, buf); + + if (is_support_sw_smu(adev)) + return smu_print_clk_levels(&adev->smu, PP_SCLK, buf); + else if (adev->powerplay.pp_funcs->print_clock_levels) return amdgpu_dpm_print_clock_levels(adev, PP_SCLK, buf); else return snprintf(buf, PAGE_SIZE, "\n"); @@ -767,7 +864,9 @@ static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev, if (ret) return ret; - if (adev->powerplay.pp_funcs->force_clock_level) + if (is_support_sw_smu(adev)) + ret = smu_force_clk_levels(&adev->smu, PP_SCLK, mask); + else if (adev->powerplay.pp_funcs->force_clock_level) ret = amdgpu_dpm_force_clock_level(adev, PP_SCLK, mask); if (ret) @@ -783,7 +882,9 @@ static ssize_t amdgpu_get_pp_dpm_mclk(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; - if (adev->powerplay.pp_funcs->print_clock_levels) + if (is_support_sw_smu(adev)) + return smu_print_clk_levels(&adev->smu, PP_MCLK, buf); + else if (adev->powerplay.pp_funcs->print_clock_levels) return amdgpu_dpm_print_clock_levels(adev, PP_MCLK, buf); else return snprintf(buf, PAGE_SIZE, "\n"); @@ -803,7 +904,9 @@ static ssize_t amdgpu_set_pp_dpm_mclk(struct device *dev, if (ret) return ret; - if (adev->powerplay.pp_funcs->force_clock_level) + if (is_support_sw_smu(adev)) + ret = smu_force_clk_levels(&adev->smu, PP_MCLK, mask); + else if (adev->powerplay.pp_funcs->force_clock_level) ret = amdgpu_dpm_force_clock_level(adev, PP_MCLK, mask); if (ret) @@ -819,7 +922,9 @@ static ssize_t amdgpu_get_pp_dpm_socclk(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; - if (adev->powerplay.pp_funcs->print_clock_levels) + if (is_support_sw_smu(adev)) + return smu_print_clk_levels(&adev->smu, PP_SOCCLK, buf); + else if (adev->powerplay.pp_funcs->print_clock_levels) return amdgpu_dpm_print_clock_levels(adev, PP_SOCCLK, buf); else return snprintf(buf, PAGE_SIZE, "\n"); @@ -839,7 +944,9 @@ static ssize_t amdgpu_set_pp_dpm_socclk(struct device *dev, if (ret) return ret; - if (adev->powerplay.pp_funcs->force_clock_level) + if (is_support_sw_smu(adev)) + ret = smu_force_clk_levels(&adev->smu, PP_SOCCLK, mask); + else if (adev->powerplay.pp_funcs->force_clock_level) ret = amdgpu_dpm_force_clock_level(adev, PP_SOCCLK, mask); if (ret) @@ -855,7 +962,9 @@ static ssize_t amdgpu_get_pp_dpm_fclk(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; - if (adev->powerplay.pp_funcs->print_clock_levels) + if (is_support_sw_smu(adev)) + return smu_print_clk_levels(&adev->smu, PP_FCLK, buf); + else if (adev->powerplay.pp_funcs->print_clock_levels) return amdgpu_dpm_print_clock_levels(adev, PP_FCLK, buf); else return snprintf(buf, PAGE_SIZE, "\n"); @@ -875,7 +984,9 @@ static ssize_t amdgpu_set_pp_dpm_fclk(struct device *dev, if (ret) return ret; - if (adev->powerplay.pp_funcs->force_clock_level) + if (is_support_sw_smu(adev)) + ret = smu_force_clk_levels(&adev->smu, PP_FCLK, mask); + else if (adev->powerplay.pp_funcs->force_clock_level) ret = amdgpu_dpm_force_clock_level(adev, PP_FCLK, mask); if (ret) @@ -891,7 +1002,9 @@ static ssize_t amdgpu_get_pp_dpm_dcefclk(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; - if (adev->powerplay.pp_funcs->print_clock_levels) + if (is_support_sw_smu(adev)) + return smu_print_clk_levels(&adev->smu, PP_DCEFCLK, buf); + else if (adev->powerplay.pp_funcs->print_clock_levels) return amdgpu_dpm_print_clock_levels(adev, PP_DCEFCLK, buf); else return snprintf(buf, PAGE_SIZE, "\n"); @@ -911,7 +1024,9 @@ static ssize_t amdgpu_set_pp_dpm_dcefclk(struct device *dev, if (ret) return ret; - if (adev->powerplay.pp_funcs->force_clock_level) + if (is_support_sw_smu(adev)) + ret = smu_force_clk_levels(&adev->smu, PP_DCEFCLK, mask); + else if (adev->powerplay.pp_funcs->force_clock_level) ret = amdgpu_dpm_force_clock_level(adev, PP_DCEFCLK, mask); if (ret) @@ -927,7 +1042,9 @@ static ssize_t amdgpu_get_pp_dpm_pcie(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; - if (adev->powerplay.pp_funcs->print_clock_levels) + if (is_support_sw_smu(adev)) + return smu_print_clk_levels(&adev->smu, PP_PCIE, buf); + else if (adev->powerplay.pp_funcs->print_clock_levels) return amdgpu_dpm_print_clock_levels(adev, PP_PCIE, buf); else return snprintf(buf, PAGE_SIZE, "\n"); @@ -947,7 +1064,9 @@ static ssize_t amdgpu_set_pp_dpm_pcie(struct device *dev, if (ret) return ret; - if (adev->powerplay.pp_funcs->force_clock_level) + if (is_support_sw_smu(adev)) + ret = smu_force_clk_levels(&adev->smu, PP_PCIE, mask); + else if (adev->powerplay.pp_funcs->force_clock_level) ret = amdgpu_dpm_force_clock_level(adev, PP_PCIE, mask); if (ret) @@ -964,7 +1083,9 @@ static ssize_t amdgpu_get_pp_sclk_od(struct device *dev, struct amdgpu_device *adev = ddev->dev_private; uint32_t value = 0; - if (adev->powerplay.pp_funcs->get_sclk_od) + if (is_support_sw_smu(adev)) + value = smu_get_od_percentage(&(adev->smu), OD_SCLK); + else if (adev->powerplay.pp_funcs->get_sclk_od) value = amdgpu_dpm_get_sclk_od(adev); return snprintf(buf, PAGE_SIZE, "%d\n", value); @@ -986,14 +1107,19 @@ static ssize_t amdgpu_set_pp_sclk_od(struct device *dev, count = -EINVAL; goto fail; } - if (adev->powerplay.pp_funcs->set_sclk_od) - amdgpu_dpm_set_sclk_od(adev, (uint32_t)value); - if (adev->powerplay.pp_funcs->dispatch_tasks) { - amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL); + if (is_support_sw_smu(adev)) { + value = smu_set_od_percentage(&(adev->smu), OD_SCLK, (uint32_t)value); } else { - adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps; - amdgpu_pm_compute_clocks(adev); + if (adev->powerplay.pp_funcs->set_sclk_od) + amdgpu_dpm_set_sclk_od(adev, (uint32_t)value); + + if (adev->powerplay.pp_funcs->dispatch_tasks) { + amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL); + } else { + adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps; + amdgpu_pm_compute_clocks(adev); + } } fail: @@ -1008,7 +1134,9 @@ static ssize_t amdgpu_get_pp_mclk_od(struct device *dev, struct amdgpu_device *adev = ddev->dev_private; uint32_t value = 0; - if (adev->powerplay.pp_funcs->get_mclk_od) + if (is_support_sw_smu(adev)) + value = smu_get_od_percentage(&(adev->smu), OD_MCLK); + else if (adev->powerplay.pp_funcs->get_mclk_od) value = amdgpu_dpm_get_mclk_od(adev); return snprintf(buf, PAGE_SIZE, "%d\n", value); @@ -1030,14 +1158,19 @@ static ssize_t amdgpu_set_pp_mclk_od(struct device *dev, count = -EINVAL; goto fail; } - if (adev->powerplay.pp_funcs->set_mclk_od) - amdgpu_dpm_set_mclk_od(adev, (uint32_t)value); - if (adev->powerplay.pp_funcs->dispatch_tasks) { - amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL); + if (is_support_sw_smu(adev)) { + value = smu_set_od_percentage(&(adev->smu), OD_MCLK, (uint32_t)value); } else { - adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps; - amdgpu_pm_compute_clocks(adev); + if (adev->powerplay.pp_funcs->set_mclk_od) + amdgpu_dpm_set_mclk_od(adev, (uint32_t)value); + + if (adev->powerplay.pp_funcs->dispatch_tasks) { + amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL); + } else { + adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps; + amdgpu_pm_compute_clocks(adev); + } } fail: @@ -1071,7 +1204,9 @@ static ssize_t amdgpu_get_pp_power_profile_mode(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; - if (adev->powerplay.pp_funcs->get_power_profile_mode) + if (is_support_sw_smu(adev)) + return smu_get_power_profile_mode(&adev->smu, buf); + else if (adev->powerplay.pp_funcs->get_power_profile_mode) return amdgpu_dpm_get_power_profile_mode(adev, buf); return snprintf(buf, PAGE_SIZE, "\n"); @@ -1121,9 +1256,10 @@ static ssize_t amdgpu_set_pp_power_profile_mode(struct device *dev, } } parameter[parameter_size] = profile_mode; - if (adev->powerplay.pp_funcs->set_power_profile_mode) + if (is_support_sw_smu(adev)) + ret = smu_set_power_profile_mode(&adev->smu, parameter, parameter_size); + else if (adev->powerplay.pp_funcs->set_power_profile_mode) ret = amdgpu_dpm_set_power_profile_mode(adev, parameter, parameter_size); - if (!ret) return count; fail: @@ -1146,14 +1282,10 @@ static ssize_t amdgpu_get_busy_percent(struct device *dev, struct amdgpu_device *adev = ddev->dev_private; int r, value, size = sizeof(value); - /* sanity check PP is enabled */ - if (!(adev->powerplay.pp_funcs && - adev->powerplay.pp_funcs->read_sensor)) - return -EINVAL; - /* read the IP busy sensor */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_LOAD, (void *)&value, &size); + if (r) return r; @@ -1247,11 +1379,6 @@ static ssize_t amdgpu_hwmon_show_temp(struct device *dev, (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - /* sanity check PP is enabled */ - if (!(adev->powerplay.pp_funcs && - adev->powerplay.pp_funcs->read_sensor)) - return -EINVAL; - /* get the temperature */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_TEMP, (void *)&temp, &size); @@ -1283,11 +1410,14 @@ static ssize_t amdgpu_hwmon_get_pwm1_enable(struct device *dev, { struct amdgpu_device *adev = dev_get_drvdata(dev); u32 pwm_mode = 0; + if (is_support_sw_smu(adev)) { + pwm_mode = smu_get_fan_control_mode(&adev->smu); + } else { + if (!adev->powerplay.pp_funcs->get_fan_control_mode) + return -EINVAL; - if (!adev->powerplay.pp_funcs->get_fan_control_mode) - return -EINVAL; - - pwm_mode = amdgpu_dpm_get_fan_control_mode(adev); + pwm_mode = amdgpu_dpm_get_fan_control_mode(adev); + } return sprintf(buf, "%i\n", pwm_mode); } @@ -1306,14 +1436,22 @@ static ssize_t amdgpu_hwmon_set_pwm1_enable(struct device *dev, (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - if (!adev->powerplay.pp_funcs->set_fan_control_mode) - return -EINVAL; + if (is_support_sw_smu(adev)) { + err = kstrtoint(buf, 10, &value); + if (err) + return err; - err = kstrtoint(buf, 10, &value); - if (err) - return err; + smu_set_fan_control_mode(&adev->smu, value); + } else { + if (!adev->powerplay.pp_funcs->set_fan_control_mode) + return -EINVAL; - amdgpu_dpm_set_fan_control_mode(adev, value); + err = kstrtoint(buf, 10, &value); + if (err) + return err; + + amdgpu_dpm_set_fan_control_mode(adev, value); + } return count; } @@ -1345,8 +1483,10 @@ static ssize_t amdgpu_hwmon_set_pwm1(struct device *dev, if ((adev->flags & AMD_IS_PX) && (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - - pwm_mode = amdgpu_dpm_get_fan_control_mode(adev); + if (is_support_sw_smu(adev)) + pwm_mode = smu_get_fan_control_mode(&adev->smu); + else + pwm_mode = amdgpu_dpm_get_fan_control_mode(adev); if (pwm_mode != AMD_FAN_CTRL_MANUAL) { pr_info("manual fan speed control should be enabled first\n"); return -EINVAL; @@ -1358,7 +1498,11 @@ static ssize_t amdgpu_hwmon_set_pwm1(struct device *dev, value = (value * 100) / 255; - if (adev->powerplay.pp_funcs->set_fan_speed_percent) { + if (is_support_sw_smu(adev)) { + err = smu_set_fan_speed_percent(&adev->smu, value); + if (err) + return err; + } else if (adev->powerplay.pp_funcs->set_fan_speed_percent) { err = amdgpu_dpm_set_fan_speed_percent(adev, value); if (err) return err; @@ -1380,7 +1524,11 @@ static ssize_t amdgpu_hwmon_get_pwm1(struct device *dev, (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - if (adev->powerplay.pp_funcs->get_fan_speed_percent) { + if (is_support_sw_smu(adev)) { + err = smu_get_fan_speed_percent(&adev->smu, &speed); + if (err) + return err; + } else if (adev->powerplay.pp_funcs->get_fan_speed_percent) { err = amdgpu_dpm_get_fan_speed_percent(adev, &speed); if (err) return err; @@ -1404,7 +1552,11 @@ static ssize_t amdgpu_hwmon_get_fan1_input(struct device *dev, (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - if (adev->powerplay.pp_funcs->get_fan_speed_rpm) { + if (is_support_sw_smu(adev)) { + err = smu_get_current_rpm(&adev->smu, &speed); + if (err) + return err; + } else if (adev->powerplay.pp_funcs->get_fan_speed_rpm) { err = amdgpu_dpm_get_fan_speed_rpm(adev, &speed); if (err) return err; @@ -1422,9 +1574,6 @@ static ssize_t amdgpu_hwmon_get_fan1_min(struct device *dev, u32 size = sizeof(min_rpm); int r; - if (!adev->powerplay.pp_funcs->read_sensor) - return -EINVAL; - r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_MIN_FAN_RPM, (void *)&min_rpm, &size); if (r) @@ -1442,9 +1591,6 @@ static ssize_t amdgpu_hwmon_get_fan1_max(struct device *dev, u32 size = sizeof(max_rpm); int r; - if (!adev->powerplay.pp_funcs->read_sensor) - return -EINVAL; - r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_MAX_FAN_RPM, (void *)&max_rpm, &size); if (r) @@ -1466,7 +1612,11 @@ static ssize_t amdgpu_hwmon_get_fan1_target(struct device *dev, (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - if (adev->powerplay.pp_funcs->get_fan_speed_rpm) { + if (is_support_sw_smu(adev)) { + err = smu_get_current_rpm(&adev->smu, &rpm); + if (err) + return err; + } else if (adev->powerplay.pp_funcs->get_fan_speed_rpm) { err = amdgpu_dpm_get_fan_speed_rpm(adev, &rpm); if (err) return err; @@ -1484,7 +1634,11 @@ static ssize_t amdgpu_hwmon_set_fan1_target(struct device *dev, u32 value; u32 pwm_mode; - pwm_mode = amdgpu_dpm_get_fan_control_mode(adev); + if (is_support_sw_smu(adev)) + pwm_mode = smu_get_fan_control_mode(&adev->smu); + else + pwm_mode = amdgpu_dpm_get_fan_control_mode(adev); + if (pwm_mode != AMD_FAN_CTRL_MANUAL) return -ENODATA; @@ -1497,7 +1651,11 @@ static ssize_t amdgpu_hwmon_set_fan1_target(struct device *dev, if (err) return err; - if (adev->powerplay.pp_funcs->set_fan_speed_rpm) { + if (is_support_sw_smu(adev)) { + err = smu_set_fan_speed_rpm(&adev->smu, value); + if (err) + return err; + } else if (adev->powerplay.pp_funcs->set_fan_speed_rpm) { err = amdgpu_dpm_set_fan_speed_rpm(adev, value); if (err) return err; @@ -1513,11 +1671,14 @@ static ssize_t amdgpu_hwmon_get_fan1_enable(struct device *dev, struct amdgpu_device *adev = dev_get_drvdata(dev); u32 pwm_mode = 0; - if (!adev->powerplay.pp_funcs->get_fan_control_mode) - return -EINVAL; - - pwm_mode = amdgpu_dpm_get_fan_control_mode(adev); + if (is_support_sw_smu(adev)) { + pwm_mode = smu_get_fan_control_mode(&adev->smu); + } else { + if (!adev->powerplay.pp_funcs->get_fan_control_mode) + return -EINVAL; + pwm_mode = amdgpu_dpm_get_fan_control_mode(adev); + } return sprintf(buf, "%i\n", pwm_mode == AMD_FAN_CTRL_AUTO ? 0 : 1); } @@ -1536,8 +1697,6 @@ static ssize_t amdgpu_hwmon_set_fan1_enable(struct device *dev, (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - if (!adev->powerplay.pp_funcs->set_fan_control_mode) - return -EINVAL; err = kstrtoint(buf, 10, &value); if (err) @@ -1550,7 +1709,13 @@ static ssize_t amdgpu_hwmon_set_fan1_enable(struct device *dev, else return -EINVAL; - amdgpu_dpm_set_fan_control_mode(adev, pwm_mode); + if (is_support_sw_smu(adev)) { + smu_set_fan_control_mode(&adev->smu, pwm_mode); + } else { + if (!adev->powerplay.pp_funcs->set_fan_control_mode) + return -EINVAL; + amdgpu_dpm_set_fan_control_mode(adev, pwm_mode); + } return count; } @@ -1569,11 +1734,6 @@ static ssize_t amdgpu_hwmon_show_vddgfx(struct device *dev, (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - /* sanity check PP is enabled */ - if (!(adev->powerplay.pp_funcs && - adev->powerplay.pp_funcs->read_sensor)) - return -EINVAL; - /* get the voltage */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDGFX, (void *)&vddgfx, &size); @@ -1608,11 +1768,6 @@ static ssize_t amdgpu_hwmon_show_vddnb(struct device *dev, (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - /* sanity check PP is enabled */ - if (!(adev->powerplay.pp_funcs && - adev->powerplay.pp_funcs->read_sensor)) - return -EINVAL; - /* get the voltage */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDNB, (void *)&vddnb, &size); @@ -1644,11 +1799,6 @@ static ssize_t amdgpu_hwmon_show_power_avg(struct device *dev, (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - /* sanity check PP is enabled */ - if (!(adev->powerplay.pp_funcs && - adev->powerplay.pp_funcs->read_sensor)) - return -EINVAL; - /* get the voltage */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_POWER, (void *)&query, &size); @@ -1675,7 +1825,10 @@ static ssize_t amdgpu_hwmon_show_power_cap_max(struct device *dev, struct amdgpu_device *adev = dev_get_drvdata(dev); uint32_t limit = 0; - if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) { + if (is_support_sw_smu(adev)) { + smu_get_power_limit(&adev->smu, &limit, true); + return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000); + } else if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) { adev->powerplay.pp_funcs->get_power_limit(adev->powerplay.pp_handle, &limit, true); return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000); } else { @@ -1690,7 +1843,10 @@ static ssize_t amdgpu_hwmon_show_power_cap(struct device *dev, struct amdgpu_device *adev = dev_get_drvdata(dev); uint32_t limit = 0; - if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) { + if (is_support_sw_smu(adev)) { + smu_get_power_limit(&adev->smu, &limit, false); + return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000); + } else if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) { adev->powerplay.pp_funcs->get_power_limit(adev->powerplay.pp_handle, &limit, false); return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000); } else { @@ -1713,7 +1869,9 @@ static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev, return err; value = value / 1000000; /* convert to Watt */ - if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->set_power_limit) { + if (is_support_sw_smu(adev)) { + adev->smu.funcs->set_power_limit(&adev->smu, value); + } else if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->set_power_limit) { err = adev->powerplay.pp_funcs->set_power_limit(adev->powerplay.pp_handle, value); if (err) return err; @@ -1967,18 +2125,20 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj, attr == &sensor_dev_attr_fan1_enable.dev_attr.attr)) return 0; - /* mask fan attributes if we have no bindings for this asic to expose */ - if ((!adev->powerplay.pp_funcs->get_fan_speed_percent && - attr == &sensor_dev_attr_pwm1.dev_attr.attr) || /* can't query fan */ - (!adev->powerplay.pp_funcs->get_fan_control_mode && - attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr)) /* can't query state */ - effective_mode &= ~S_IRUGO; + if (!is_support_sw_smu(adev)) { + /* mask fan attributes if we have no bindings for this asic to expose */ + if ((!adev->powerplay.pp_funcs->get_fan_speed_percent && + attr == &sensor_dev_attr_pwm1.dev_attr.attr) || /* can't query fan */ + (!adev->powerplay.pp_funcs->get_fan_control_mode && + attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr)) /* can't query state */ + effective_mode &= ~S_IRUGO; - if ((!adev->powerplay.pp_funcs->set_fan_speed_percent && - attr == &sensor_dev_attr_pwm1.dev_attr.attr) || /* can't manage fan */ - (!adev->powerplay.pp_funcs->set_fan_control_mode && - attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr)) /* can't manage state */ - effective_mode &= ~S_IWUSR; + if ((!adev->powerplay.pp_funcs->set_fan_speed_percent && + attr == &sensor_dev_attr_pwm1.dev_attr.attr) || /* can't manage fan */ + (!adev->powerplay.pp_funcs->set_fan_control_mode && + attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr)) /* can't manage state */ + effective_mode &= ~S_IWUSR; + } if ((adev->flags & AMD_IS_APU) && (attr == &sensor_dev_attr_power1_average.dev_attr.attr || @@ -1987,20 +2147,22 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj, attr == &sensor_dev_attr_power1_cap.dev_attr.attr)) return 0; - /* hide max/min values if we can't both query and manage the fan */ - if ((!adev->powerplay.pp_funcs->set_fan_speed_percent && - !adev->powerplay.pp_funcs->get_fan_speed_percent) && - (!adev->powerplay.pp_funcs->set_fan_speed_rpm && - !adev->powerplay.pp_funcs->get_fan_speed_rpm) && - (attr == &sensor_dev_attr_pwm1_max.dev_attr.attr || - attr == &sensor_dev_attr_pwm1_min.dev_attr.attr)) - return 0; + if (!is_support_sw_smu(adev)) { + /* hide max/min values if we can't both query and manage the fan */ + if ((!adev->powerplay.pp_funcs->set_fan_speed_percent && + !adev->powerplay.pp_funcs->get_fan_speed_percent) && + (!adev->powerplay.pp_funcs->set_fan_speed_rpm && + !adev->powerplay.pp_funcs->get_fan_speed_rpm) && + (attr == &sensor_dev_attr_pwm1_max.dev_attr.attr || + attr == &sensor_dev_attr_pwm1_min.dev_attr.attr)) + return 0; - if ((!adev->powerplay.pp_funcs->set_fan_speed_rpm && - !adev->powerplay.pp_funcs->get_fan_speed_rpm) && - (attr == &sensor_dev_attr_fan1_max.dev_attr.attr || - attr == &sensor_dev_attr_fan1_min.dev_attr.attr)) - return 0; + if ((!adev->powerplay.pp_funcs->set_fan_speed_rpm && + !adev->powerplay.pp_funcs->get_fan_speed_rpm) && + (attr == &sensor_dev_attr_fan1_max.dev_attr.attr || + attr == &sensor_dev_attr_fan1_min.dev_attr.attr)) + return 0; + } /* only APUs have vddnb */ if (!(adev->flags & AMD_IS_APU) && @@ -2039,9 +2201,7 @@ void amdgpu_dpm_thermal_work_handler(struct work_struct *work) if (!adev->pm.dpm_enabled) return; - if (adev->powerplay.pp_funcs && - adev->powerplay.pp_funcs->read_sensor && - !amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_TEMP, + if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_TEMP, (void *)&temp, &size)) { if (temp < adev->pm.dpm.thermal.min_temp) /* switch back the user state */ @@ -2267,7 +2427,13 @@ static void amdgpu_dpm_change_power_state_locked(struct amdgpu_device *adev) void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable) { - if (adev->powerplay.pp_funcs->set_powergating_by_smu) { + int ret = 0; + if (is_support_sw_smu(adev)) { + ret = smu_dpm_set_power_gate(&adev->smu, AMD_IP_BLOCK_TYPE_UVD, enable); + if (ret) + DRM_ERROR("[SW SMU]: dpm enable uvd failed, state = %s, ret = %d. \n", + enable ? "true" : "false", ret); + } else if (adev->powerplay.pp_funcs->set_powergating_by_smu) { /* enable/disable UVD */ mutex_lock(&adev->pm.mutex); amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_UVD, !enable); @@ -2288,7 +2454,13 @@ void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable) void amdgpu_dpm_enable_vce(struct amdgpu_device *adev, bool enable) { - if (adev->powerplay.pp_funcs->set_powergating_by_smu) { + int ret = 0; + if (is_support_sw_smu(adev)) { + ret = smu_dpm_set_power_gate(&adev->smu, AMD_IP_BLOCK_TYPE_VCE, enable); + if (ret) + DRM_ERROR("[SW SMU]: dpm enable vce failed, state = %s, ret = %d. \n", + enable ? "true" : "false", ret); + } else if (adev->powerplay.pp_funcs->set_powergating_by_smu) { /* enable/disable VCE */ mutex_lock(&adev->pm.mutex); amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_VCE, !enable); @@ -2413,7 +2585,8 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev) "pp_power_profile_mode\n"); return ret; } - if (hwmgr->od_enabled) { + if ((is_support_sw_smu(adev) && adev->smu.od_enabled) || + (!is_support_sw_smu(adev) && hwmgr->od_enabled)) { ret = device_create_file(adev->dev, &dev_attr_pp_od_clk_voltage); if (ret) { @@ -2489,7 +2662,8 @@ void amdgpu_pm_sysfs_fini(struct amdgpu_device *adev) device_remove_file(adev->dev, &dev_attr_pp_mclk_od); device_remove_file(adev->dev, &dev_attr_pp_power_profile_mode); - if (hwmgr->od_enabled) + if ((is_support_sw_smu(adev) && adev->smu.od_enabled) || + (!is_support_sw_smu(adev) && hwmgr->od_enabled)) device_remove_file(adev->dev, &dev_attr_pp_od_clk_voltage); device_remove_file(adev->dev, &dev_attr_gpu_busy_percent); @@ -2516,28 +2690,38 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev) amdgpu_fence_wait_empty(ring); } - if (adev->powerplay.pp_funcs->dispatch_tasks) { - if (!amdgpu_device_has_dc_support(adev)) { + if (is_support_sw_smu(adev)) { + struct smu_context *smu = &adev->smu; + struct smu_dpm_context *smu_dpm = &adev->smu.smu_dpm; + mutex_lock(&(smu->mutex)); + smu_handle_task(&adev->smu, + smu_dpm->dpm_level, + AMD_PP_TASK_DISPLAY_CONFIG_CHANGE); + mutex_unlock(&(smu->mutex)); + } else { + if (adev->powerplay.pp_funcs->dispatch_tasks) { + if (!amdgpu_device_has_dc_support(adev)) { + mutex_lock(&adev->pm.mutex); + amdgpu_dpm_get_active_displays(adev); + adev->pm.pm_display_cfg.num_display = adev->pm.dpm.new_active_crtc_count; + adev->pm.pm_display_cfg.vrefresh = amdgpu_dpm_get_vrefresh(adev); + adev->pm.pm_display_cfg.min_vblank_time = amdgpu_dpm_get_vblank_time(adev); + /* we have issues with mclk switching with refresh rates over 120 hz on the non-DC code. */ + if (adev->pm.pm_display_cfg.vrefresh > 120) + adev->pm.pm_display_cfg.min_vblank_time = 0; + if (adev->powerplay.pp_funcs->display_configuration_change) + adev->powerplay.pp_funcs->display_configuration_change( + adev->powerplay.pp_handle, + &adev->pm.pm_display_cfg); + mutex_unlock(&adev->pm.mutex); + } + amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_DISPLAY_CONFIG_CHANGE, NULL); + } else { mutex_lock(&adev->pm.mutex); amdgpu_dpm_get_active_displays(adev); - adev->pm.pm_display_cfg.num_display = adev->pm.dpm.new_active_crtc_count; - adev->pm.pm_display_cfg.vrefresh = amdgpu_dpm_get_vrefresh(adev); - adev->pm.pm_display_cfg.min_vblank_time = amdgpu_dpm_get_vblank_time(adev); - /* we have issues with mclk switching with refresh rates over 120 hz on the non-DC code. */ - if (adev->pm.pm_display_cfg.vrefresh > 120) - adev->pm.pm_display_cfg.min_vblank_time = 0; - if (adev->powerplay.pp_funcs->display_configuration_change) - adev->powerplay.pp_funcs->display_configuration_change( - adev->powerplay.pp_handle, - &adev->pm.pm_display_cfg); + amdgpu_dpm_change_power_state_locked(adev); mutex_unlock(&adev->pm.mutex); } - amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_DISPLAY_CONFIG_CHANGE, NULL); - } else { - mutex_lock(&adev->pm.mutex); - amdgpu_dpm_get_active_displays(adev); - amdgpu_dpm_change_power_state_locked(adev); - mutex_unlock(&adev->pm.mutex); } } @@ -2553,11 +2737,6 @@ static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *a uint32_t query = 0; int size; - /* sanity check PP is enabled */ - if (!(adev->powerplay.pp_funcs && - adev->powerplay.pp_funcs->read_sensor)) - return -EINVAL; - /* GPU Clocks */ size = sizeof(value); seq_printf(m, "GFX Clocks and Power:\n"); @@ -2649,7 +2828,7 @@ static int amdgpu_debugfs_pm_info(struct seq_file *m, void *data) if ((adev->flags & AMD_IS_PX) && (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) { seq_printf(m, "PX asic powered off\n"); - } else if (adev->powerplay.pp_funcs->debugfs_print_current_performance_level) { + } else if (!is_support_sw_smu(adev) && adev->powerplay.pp_funcs->debugfs_print_current_performance_level) { mutex_lock(&adev->pm.mutex); if (adev->powerplay.pp_funcs->debugfs_print_current_performance_level) adev->powerplay.pp_funcs->debugfs_print_current_performance_level(adev, m); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 3091488cd8cc..905cce1814f3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -120,6 +120,7 @@ psp_cmd_submit_buf(struct psp_context *psp, { int ret; int index; + int timeout = 2000; memset(psp->cmd_buf_mem, 0, PSP_CMD_BUFFER_SIZE); @@ -133,8 +134,11 @@ psp_cmd_submit_buf(struct psp_context *psp, return ret; } - while (*((unsigned int *)psp->fence_buf) != index) + while (*((unsigned int *)psp->fence_buf) != index) { + if (--timeout == 0) + break; msleep(1); + } /* In some cases, psp response status is not 0 even there is no * problem while the command is submitted. Some version of PSP FW @@ -143,12 +147,14 @@ psp_cmd_submit_buf(struct psp_context *psp, * during psp initialization to avoid breaking hw_init and it doesn't * return -EINVAL. */ - if (psp->cmd_buf_mem->resp.status) { + if (psp->cmd_buf_mem->resp.status || !timeout) { if (ucode) DRM_WARN("failed to load ucode id (%d) ", ucode->ucode_id); DRM_WARN("psp command failed and response status is (%d)\n", psp->cmd_buf_mem->resp.status); + if (!timeout) + return -EINVAL; } /* get xGMI session id from response buffer */ @@ -181,13 +187,13 @@ static int psp_tmr_init(struct psp_context *psp) int ret; /* - * Allocate 3M memory aligned to 1M from Frame Buffer (local - * physical). + * According to HW engineer, they prefer the TMR address be "naturally + * aligned" , e.g. the start address be an integer divide of TMR size. * * Note: this memory need be reserved till the driver * uninitializes. */ - ret = amdgpu_bo_create_kernel(psp->adev, PSP_TMR_SIZE, 0x100000, + ret = amdgpu_bo_create_kernel(psp->adev, PSP_TMR_SIZE, PSP_TMR_SIZE, AMDGPU_GEM_DOMAIN_VRAM, &psp->tmr_bo, &psp->tmr_mc_addr, &psp->tmr_buf); @@ -466,6 +472,206 @@ static int psp_xgmi_initialize(struct psp_context *psp) return ret; } +// ras begin +static void psp_prep_ras_ta_load_cmd_buf(struct psp_gfx_cmd_resp *cmd, + uint64_t ras_ta_mc, uint64_t ras_mc_shared, + uint32_t ras_ta_size, uint32_t shared_size) +{ + cmd->cmd_id = GFX_CMD_ID_LOAD_TA; + cmd->cmd.cmd_load_ta.app_phy_addr_lo = lower_32_bits(ras_ta_mc); + cmd->cmd.cmd_load_ta.app_phy_addr_hi = upper_32_bits(ras_ta_mc); + cmd->cmd.cmd_load_ta.app_len = ras_ta_size; + + cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = lower_32_bits(ras_mc_shared); + cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = upper_32_bits(ras_mc_shared); + cmd->cmd.cmd_load_ta.cmd_buf_len = shared_size; +} + +static int psp_ras_init_shared_buf(struct psp_context *psp) +{ + int ret; + + /* + * Allocate 16k memory aligned to 4k from Frame Buffer (local + * physical) for ras ta <-> Driver + */ + ret = amdgpu_bo_create_kernel(psp->adev, PSP_RAS_SHARED_MEM_SIZE, + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, + &psp->ras.ras_shared_bo, + &psp->ras.ras_shared_mc_addr, + &psp->ras.ras_shared_buf); + + return ret; +} + +static int psp_ras_load(struct psp_context *psp) +{ + int ret; + struct psp_gfx_cmd_resp *cmd; + + /* + * TODO: bypass the loading in sriov for now + */ + if (amdgpu_sriov_vf(psp->adev)) + return 0; + + cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); + if (!cmd) + return -ENOMEM; + + memset(psp->fw_pri_buf, 0, PSP_1_MEG); + memcpy(psp->fw_pri_buf, psp->ta_ras_start_addr, psp->ta_ras_ucode_size); + + psp_prep_ras_ta_load_cmd_buf(cmd, psp->fw_pri_mc_addr, + psp->ras.ras_shared_mc_addr, + psp->ta_ras_ucode_size, PSP_RAS_SHARED_MEM_SIZE); + + ret = psp_cmd_submit_buf(psp, NULL, cmd, + psp->fence_buf_mc_addr); + + if (!ret) { + psp->ras.ras_initialized = 1; + psp->ras.session_id = cmd->resp.session_id; + } + + kfree(cmd); + + return ret; +} + +static void psp_prep_ras_ta_unload_cmd_buf(struct psp_gfx_cmd_resp *cmd, + uint32_t ras_session_id) +{ + cmd->cmd_id = GFX_CMD_ID_UNLOAD_TA; + cmd->cmd.cmd_unload_ta.session_id = ras_session_id; +} + +static int psp_ras_unload(struct psp_context *psp) +{ + int ret; + struct psp_gfx_cmd_resp *cmd; + + /* + * TODO: bypass the unloading in sriov for now + */ + if (amdgpu_sriov_vf(psp->adev)) + return 0; + + cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); + if (!cmd) + return -ENOMEM; + + psp_prep_ras_ta_unload_cmd_buf(cmd, psp->ras.session_id); + + ret = psp_cmd_submit_buf(psp, NULL, cmd, + psp->fence_buf_mc_addr); + + kfree(cmd); + + return ret; +} + +static void psp_prep_ras_ta_invoke_cmd_buf(struct psp_gfx_cmd_resp *cmd, + uint32_t ta_cmd_id, + uint32_t ras_session_id) +{ + cmd->cmd_id = GFX_CMD_ID_INVOKE_CMD; + cmd->cmd.cmd_invoke_cmd.session_id = ras_session_id; + cmd->cmd.cmd_invoke_cmd.ta_cmd_id = ta_cmd_id; + /* Note: cmd_invoke_cmd.buf is not used for now */ +} + +int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id) +{ + int ret; + struct psp_gfx_cmd_resp *cmd; + + /* + * TODO: bypass the loading in sriov for now + */ + if (amdgpu_sriov_vf(psp->adev)) + return 0; + + cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); + if (!cmd) + return -ENOMEM; + + psp_prep_ras_ta_invoke_cmd_buf(cmd, ta_cmd_id, + psp->ras.session_id); + + ret = psp_cmd_submit_buf(psp, NULL, cmd, + psp->fence_buf_mc_addr); + + kfree(cmd); + + return ret; +} + +int psp_ras_enable_features(struct psp_context *psp, + union ta_ras_cmd_input *info, bool enable) +{ + struct ta_ras_shared_memory *ras_cmd; + int ret; + + if (!psp->ras.ras_initialized) + return -EINVAL; + + ras_cmd = (struct ta_ras_shared_memory *)psp->ras.ras_shared_buf; + memset(ras_cmd, 0, sizeof(struct ta_ras_shared_memory)); + + if (enable) + ras_cmd->cmd_id = TA_RAS_COMMAND__ENABLE_FEATURES; + else + ras_cmd->cmd_id = TA_RAS_COMMAND__DISABLE_FEATURES; + + ras_cmd->ras_in_message = *info; + + ret = psp_ras_invoke(psp, ras_cmd->cmd_id); + if (ret) + return -EINVAL; + + return ras_cmd->ras_status; +} + +static int psp_ras_terminate(struct psp_context *psp) +{ + int ret; + + if (!psp->ras.ras_initialized) + return 0; + + ret = psp_ras_unload(psp); + if (ret) + return ret; + + psp->ras.ras_initialized = 0; + + /* free ras shared memory */ + amdgpu_bo_free_kernel(&psp->ras.ras_shared_bo, + &psp->ras.ras_shared_mc_addr, + &psp->ras.ras_shared_buf); + + return 0; +} + +static int psp_ras_initialize(struct psp_context *psp) +{ + int ret; + + if (!psp->ras.ras_initialized) { + ret = psp_ras_init_shared_buf(psp); + if (ret) + return ret; + } + + ret = psp_ras_load(psp); + if (ret) + return ret; + + return 0; +} +// ras end + static int psp_hw_start(struct psp_context *psp) { struct amdgpu_device *adev = psp->adev; @@ -473,25 +679,35 @@ static int psp_hw_start(struct psp_context *psp) if (!amdgpu_sriov_vf(adev) || !adev->in_gpu_reset) { ret = psp_bootloader_load_sysdrv(psp); - if (ret) + if (ret) { + DRM_ERROR("PSP load sysdrv failed!\n"); return ret; + } ret = psp_bootloader_load_sos(psp); - if (ret) + if (ret) { + DRM_ERROR("PSP load sos failed!\n"); return ret; + } } ret = psp_ring_create(psp, PSP_RING_TYPE__KM); - if (ret) + if (ret) { + DRM_ERROR("PSP create ring failed!\n"); return ret; + } ret = psp_tmr_load(psp); - if (ret) + if (ret) { + DRM_ERROR("PSP load tmr failed!\n"); return ret; + } ret = psp_asd_load(psp); - if (ret) + if (ret) { + DRM_ERROR("PSP load asd failed!\n"); return ret; + } if (adev->gmc.xgmi.num_physical_nodes > 1) { ret = psp_xgmi_initialize(psp); @@ -502,6 +718,15 @@ static int psp_hw_start(struct psp_context *psp) dev_err(psp->adev->dev, "XGMI: Failed to initialize XGMI session\n"); } + + + if (psp->adev->psp.ta_fw) { + ret = psp_ras_initialize(psp); + if (ret) + dev_err(psp->adev->dev, + "RAS: Failed to initialize RAS\n"); + } + return 0; } @@ -665,53 +890,52 @@ static int psp_load_fw(struct amdgpu_device *adev) &psp->fence_buf_mc_addr, &psp->fence_buf); if (ret) - goto failed_mem2; + goto failed; ret = amdgpu_bo_create_kernel(adev, PSP_CMD_BUFFER_SIZE, PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, &psp->cmd_buf_bo, &psp->cmd_buf_mc_addr, (void **)&psp->cmd_buf_mem); if (ret) - goto failed_mem1; + goto failed; memset(psp->fence_buf, 0, PSP_FENCE_BUFFER_SIZE); ret = psp_ring_init(psp, PSP_RING_TYPE__KM); - if (ret) - goto failed_mem; + if (ret) { + DRM_ERROR("PSP ring init failed!\n"); + goto failed; + } ret = psp_tmr_init(psp); - if (ret) - goto failed_mem; + if (ret) { + DRM_ERROR("PSP tmr init failed!\n"); + goto failed; + } ret = psp_asd_init(psp); - if (ret) - goto failed_mem; + if (ret) { + DRM_ERROR("PSP asd init failed!\n"); + goto failed; + } skip_memalloc: ret = psp_hw_start(psp); if (ret) - goto failed_mem; + goto failed; ret = psp_np_fw_load(psp); if (ret) - goto failed_mem; + goto failed; return 0; -failed_mem: - amdgpu_bo_free_kernel(&psp->cmd_buf_bo, - &psp->cmd_buf_mc_addr, - (void **)&psp->cmd_buf_mem); -failed_mem1: - amdgpu_bo_free_kernel(&psp->fence_buf_bo, - &psp->fence_buf_mc_addr, &psp->fence_buf); -failed_mem2: - amdgpu_bo_free_kernel(&psp->fw_pri_bo, - &psp->fw_pri_mc_addr, &psp->fw_pri_buf); failed: - kfree(psp->cmd); - psp->cmd = NULL; + /* + * all cleanup jobs (xgmi terminate, ras terminate, + * ring destroy, cmd/fence/fw buffers destory, + * psp->cmd destory) are delayed to psp_hw_fini + */ return ret; } @@ -753,6 +977,9 @@ static int psp_hw_fini(void *handle) psp->xgmi_context.initialized == 1) psp_xgmi_terminate(psp); + if (psp->adev->psp.ta_fw) + psp_ras_terminate(psp); + psp_ring_destroy(psp, PSP_RING_TYPE__KM); amdgpu_bo_free_kernel(&psp->tmr_bo, &psp->tmr_mc_addr, &psp->tmr_buf); @@ -786,6 +1013,14 @@ static int psp_suspend(void *handle) } } + if (psp->adev->psp.ta_fw) { + ret = psp_ras_terminate(psp); + if (ret) { + DRM_ERROR("Failed to terminate ras ta\n"); + return ret; + } + } + ret = psp_ring_stop(psp, PSP_RING_TYPE__KM); if (ret) { DRM_ERROR("PSP ring stop failed\n"); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h index 2ef98cc755d6..cde113f07c96 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h @@ -28,11 +28,13 @@ #include "amdgpu.h" #include "psp_gfx_if.h" #include "ta_xgmi_if.h" +#include "ta_ras_if.h" #define PSP_FENCE_BUFFER_SIZE 0x1000 #define PSP_CMD_BUFFER_SIZE 0x1000 #define PSP_ASD_SHARED_MEM_SIZE 0x4000 #define PSP_XGMI_SHARED_MEM_SIZE 0x4000 +#define PSP_RAS_SHARED_MEM_SIZE 0x4000 #define PSP_1_MEG 0x100000 #define PSP_TMR_SIZE 0x400000 @@ -88,6 +90,9 @@ struct psp_funcs int (*xgmi_set_topology_info)(struct psp_context *psp, int number_devices, struct psp_xgmi_topology_info *topology); bool (*support_vmr_ring)(struct psp_context *psp); + int (*ras_trigger_error)(struct psp_context *psp, + struct ta_ras_trigger_error_input *info); + int (*ras_cure_posion)(struct psp_context *psp, uint64_t *mode_ptr); }; struct psp_xgmi_context { @@ -98,6 +103,16 @@ struct psp_xgmi_context { void *xgmi_shared_buf; }; +struct psp_ras_context { + /*ras fw*/ + bool ras_initialized; + uint32_t session_id; + struct amdgpu_bo *ras_shared_bo; + uint64_t ras_shared_mc_addr; + void *ras_shared_buf; + struct amdgpu_ras *ras; +}; + struct psp_context { struct amdgpu_device *adev; @@ -150,10 +165,15 @@ struct psp_context /* xgmi ta firmware and buffer */ const struct firmware *ta_fw; + uint32_t ta_fw_version; uint32_t ta_xgmi_ucode_version; uint32_t ta_xgmi_ucode_size; uint8_t *ta_xgmi_start_addr; + uint32_t ta_ras_ucode_version; + uint32_t ta_ras_ucode_size; + uint8_t *ta_ras_start_addr; struct psp_xgmi_context xgmi_context; + struct psp_ras_context ras; }; struct amdgpu_psp_funcs { @@ -207,6 +227,13 @@ struct psp_xgmi_topology_info { #define amdgpu_psp_check_fw_loading_status(adev, i) (adev)->firmware.funcs->check_fw_loading_status((adev), (i)) +#define psp_ras_trigger_error(psp, info) \ + ((psp)->funcs->ras_trigger_error ? \ + (psp)->funcs->ras_trigger_error((psp), (info)) : -EINVAL) +#define psp_ras_cure_posion(psp, addr) \ + ((psp)->funcs->ras_cure_posion ? \ + (psp)->funcs->ras_cure_posion(psp, (addr)) : -EINVAL) + extern const struct amd_ip_funcs psp_ip_funcs; extern const struct amdgpu_ip_block_version psp_v3_1_ip_block; @@ -217,6 +244,11 @@ extern const struct amdgpu_ip_block_version psp_v10_0_ip_block; int psp_gpu_reset(struct amdgpu_device *adev); int psp_xgmi_invoke(struct psp_context *psp, uint32_t ta_cmd_id); + +int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id); +int psp_ras_enable_features(struct psp_context *psp, + union ta_ras_cmd_input *info, bool enable); + extern const struct amdgpu_ip_block_version psp_v11_0_ip_block; #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c new file mode 100644 index 000000000000..22bd21efe6b1 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -0,0 +1,1482 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * + */ +#include +#include +#include +#include "amdgpu.h" +#include "amdgpu_ras.h" +#include "amdgpu_atomfirmware.h" + +struct ras_ih_data { + /* interrupt bottom half */ + struct work_struct ih_work; + int inuse; + /* IP callback */ + ras_ih_cb cb; + /* full of entries */ + unsigned char *ring; + unsigned int ring_size; + unsigned int element_size; + unsigned int aligned_element_size; + unsigned int rptr; + unsigned int wptr; +}; + +struct ras_fs_data { + char sysfs_name[32]; + char debugfs_name[32]; +}; + +struct ras_err_data { + unsigned long ue_count; + unsigned long ce_count; +}; + +struct ras_err_handler_data { + /* point to bad pages array */ + struct { + unsigned long bp; + struct amdgpu_bo *bo; + } *bps; + /* the count of entries */ + int count; + /* the space can place new entries */ + int space_left; + /* last reserved entry's index + 1 */ + int last_reserved; +}; + +struct ras_manager { + struct ras_common_if head; + /* reference count */ + int use; + /* ras block link */ + struct list_head node; + /* the device */ + struct amdgpu_device *adev; + /* debugfs */ + struct dentry *ent; + /* sysfs */ + struct device_attribute sysfs_attr; + int attr_inuse; + + /* fs node name */ + struct ras_fs_data fs_data; + + /* IH data */ + struct ras_ih_data ih_data; + + struct ras_err_data err_data; +}; + +const char *ras_error_string[] = { + "none", + "parity", + "single_correctable", + "multi_uncorrectable", + "poison", +}; + +const char *ras_block_string[] = { + "umc", + "sdma", + "gfx", + "mmhub", + "athub", + "pcie_bif", + "hdp", + "xgmi_wafl", + "df", + "smn", + "sem", + "mp0", + "mp1", + "fuse", +}; + +#define ras_err_str(i) (ras_error_string[ffs(i)]) +#define ras_block_str(i) (ras_block_string[i]) + +#define AMDGPU_RAS_FLAG_INIT_BY_VBIOS 1 +#define RAS_DEFAULT_FLAGS (AMDGPU_RAS_FLAG_INIT_BY_VBIOS) + +static void amdgpu_ras_self_test(struct amdgpu_device *adev) +{ + /* TODO */ +} + +static ssize_t amdgpu_ras_debugfs_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) +{ + struct ras_manager *obj = (struct ras_manager *)file_inode(f)->i_private; + struct ras_query_if info = { + .head = obj->head, + }; + ssize_t s; + char val[128]; + + if (amdgpu_ras_error_query(obj->adev, &info)) + return -EINVAL; + + s = snprintf(val, sizeof(val), "%s: %lu\n%s: %lu\n", + "ue", info.ue_count, + "ce", info.ce_count); + if (*pos >= s) + return 0; + + s -= *pos; + s = min_t(u64, s, size); + + + if (copy_to_user(buf, &val[*pos], s)) + return -EINVAL; + + *pos += s; + + return s; +} + +static const struct file_operations amdgpu_ras_debugfs_ops = { + .owner = THIS_MODULE, + .read = amdgpu_ras_debugfs_read, + .write = NULL, + .llseek = default_llseek +}; + +static int amdgpu_ras_find_block_id_by_name(const char *name, int *block_id) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(ras_block_string); i++) { + *block_id = i; + if (strcmp(name, ras_block_str(i)) == 0) + return 0; + } + return -EINVAL; +} + +static int amdgpu_ras_debugfs_ctrl_parse_data(struct file *f, + const char __user *buf, size_t size, + loff_t *pos, struct ras_debug_if *data) +{ + ssize_t s = min_t(u64, 64, size); + char str[65]; + char block_name[33]; + char err[9] = "ue"; + int op = -1; + int block_id; + u64 address, value; + + if (*pos) + return -EINVAL; + *pos = size; + + memset(str, 0, sizeof(str)); + memset(data, 0, sizeof(*data)); + + if (copy_from_user(str, buf, s)) + return -EINVAL; + + if (sscanf(str, "disable %32s", block_name) == 1) + op = 0; + else if (sscanf(str, "enable %32s %8s", block_name, err) == 2) + op = 1; + else if (sscanf(str, "inject %32s %8s", block_name, err) == 2) + op = 2; + else if (str[0] && str[1] && str[2] && str[3]) + /* ascii string, but commands are not matched. */ + return -EINVAL; + + if (op != -1) { + if (amdgpu_ras_find_block_id_by_name(block_name, &block_id)) + return -EINVAL; + + data->head.block = block_id; + data->head.type = memcmp("ue", err, 2) == 0 ? + AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE : + AMDGPU_RAS_ERROR__SINGLE_CORRECTABLE; + data->op = op; + + if (op == 2) { + if (sscanf(str, "%*s %*s %*s %llu %llu", + &address, &value) != 2) + if (sscanf(str, "%*s %*s %*s 0x%llx 0x%llx", + &address, &value) != 2) + return -EINVAL; + data->inject.address = address; + data->inject.value = value; + } + } else { + if (size < sizeof(*data)) + return -EINVAL; + + if (copy_from_user(data, buf, sizeof(*data))) + return -EINVAL; + } + + return 0; +} +/* + * DOC: ras debugfs control interface + * + * It accepts struct ras_debug_if who has two members. + * + * First member: ras_debug_if::head or ras_debug_if::inject. + * + * head is used to indicate which IP block will be under control. + * + * head has four members, they are block, type, sub_block_index, name. + * block: which IP will be under control. + * type: what kind of error will be enabled/disabled/injected. + * sub_block_index: some IPs have subcomponets. say, GFX, sDMA. + * name: the name of IP. + * + * inject has two more members than head, they are address, value. + * As their names indicate, inject operation will write the + * value to the address. + * + * Second member: struct ras_debug_if::op. + * It has three kinds of operations. + * 0: disable RAS on the block. Take ::head as its data. + * 1: enable RAS on the block. Take ::head as its data. + * 2: inject errors on the block. Take ::inject as its data. + * + * How to use the interface? + * programs: + * copy the struct ras_debug_if in your codes and initialize it. + * write the struct to the control node. + * + * bash: + * echo op block [error [address value]] > .../ras/ras_ctrl + * op: disable, enable, inject + * disable: only block is needed + * enable: block and error are needed + * inject: error, address, value are needed + * block: umc, smda, gfx, ......... + * see ras_block_string[] for details + * error: ue, ce + * ue: multi_uncorrectable + * ce: single_correctable + * + * here are some examples for bash commands, + * echo inject umc ue 0x0 0x0 > /sys/kernel/debug/dri/0/ras/ras_ctrl + * echo inject umc ce 0 0 > /sys/kernel/debug/dri/0/ras/ras_ctrl + * echo disable umc > /sys/kernel/debug/dri/0/ras/ras_ctrl + * + * How to check the result? + * + * For disable/enable, please check ras features at + * /sys/class/drm/card[0/1/2...]/device/ras/features + * + * For inject, please check corresponding err count at + * /sys/class/drm/card[0/1/2...]/device/ras/[gfx/sdma/...]_err_count + * + * NOTE: operation is only allowed on blocks which are supported. + * Please check ras mask at /sys/module/amdgpu/parameters/ras_mask + */ +static ssize_t amdgpu_ras_debugfs_ctrl_write(struct file *f, const char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)file_inode(f)->i_private; + struct ras_debug_if data; + int ret = 0; + + ret = amdgpu_ras_debugfs_ctrl_parse_data(f, buf, size, pos, &data); + if (ret) + return -EINVAL; + + if (!amdgpu_ras_is_supported(adev, data.head.block)) + return -EINVAL; + + switch (data.op) { + case 0: + ret = amdgpu_ras_feature_enable(adev, &data.head, 0); + break; + case 1: + ret = amdgpu_ras_feature_enable(adev, &data.head, 1); + break; + case 2: + ret = amdgpu_ras_error_inject(adev, &data.inject); + break; + default: + ret = -EINVAL; + break; + }; + + if (ret) + return -EINVAL; + + return size; +} + +static const struct file_operations amdgpu_ras_debugfs_ctrl_ops = { + .owner = THIS_MODULE, + .read = NULL, + .write = amdgpu_ras_debugfs_ctrl_write, + .llseek = default_llseek +}; + +static ssize_t amdgpu_ras_sysfs_read(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct ras_manager *obj = container_of(attr, struct ras_manager, sysfs_attr); + struct ras_query_if info = { + .head = obj->head, + }; + + if (amdgpu_ras_error_query(obj->adev, &info)) + return -EINVAL; + + return snprintf(buf, PAGE_SIZE, "%s: %lu\n%s: %lu\n", + "ue", info.ue_count, + "ce", info.ce_count); +} + +/* obj begin */ + +#define get_obj(obj) do { (obj)->use++; } while (0) +#define alive_obj(obj) ((obj)->use) + +static inline void put_obj(struct ras_manager *obj) +{ + if (obj && --obj->use == 0) + list_del(&obj->node); + if (obj && obj->use < 0) { + DRM_ERROR("RAS ERROR: Unbalance obj(%s) use\n", obj->head.name); + } +} + +/* make one obj and return it. */ +static struct ras_manager *amdgpu_ras_create_obj(struct amdgpu_device *adev, + struct ras_common_if *head) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_manager *obj; + + if (!con) + return NULL; + + if (head->block >= AMDGPU_RAS_BLOCK_COUNT) + return NULL; + + obj = &con->objs[head->block]; + /* already exist. return obj? */ + if (alive_obj(obj)) + return NULL; + + obj->head = *head; + obj->adev = adev; + list_add(&obj->node, &con->head); + get_obj(obj); + + return obj; +} + +/* return an obj equal to head, or the first when head is NULL */ +static struct ras_manager *amdgpu_ras_find_obj(struct amdgpu_device *adev, + struct ras_common_if *head) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_manager *obj; + int i; + + if (!con) + return NULL; + + if (head) { + if (head->block >= AMDGPU_RAS_BLOCK_COUNT) + return NULL; + + obj = &con->objs[head->block]; + + if (alive_obj(obj)) { + WARN_ON(head->block != obj->head.block); + return obj; + } + } else { + for (i = 0; i < AMDGPU_RAS_BLOCK_COUNT; i++) { + obj = &con->objs[i]; + if (alive_obj(obj)) { + WARN_ON(i != obj->head.block); + return obj; + } + } + } + + return NULL; +} +/* obj end */ + +/* feature ctl begin */ +static int amdgpu_ras_is_feature_allowed(struct amdgpu_device *adev, + struct ras_common_if *head) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + + return con->hw_supported & BIT(head->block); +} + +static int amdgpu_ras_is_feature_enabled(struct amdgpu_device *adev, + struct ras_common_if *head) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + + return con->features & BIT(head->block); +} + +/* + * if obj is not created, then create one. + * set feature enable flag. + */ +static int __amdgpu_ras_feature_enable(struct amdgpu_device *adev, + struct ras_common_if *head, int enable) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_manager *obj = amdgpu_ras_find_obj(adev, head); + + /* If hardware does not support ras, then do not create obj. + * But if hardware support ras, we can create the obj. + * Ras framework checks con->hw_supported to see if it need do + * corresponding initialization. + * IP checks con->support to see if it need disable ras. + */ + if (!amdgpu_ras_is_feature_allowed(adev, head)) + return 0; + if (!(!!enable ^ !!amdgpu_ras_is_feature_enabled(adev, head))) + return 0; + + if (enable) { + if (!obj) { + obj = amdgpu_ras_create_obj(adev, head); + if (!obj) + return -EINVAL; + } else { + /* In case we create obj somewhere else */ + get_obj(obj); + } + con->features |= BIT(head->block); + } else { + if (obj && amdgpu_ras_is_feature_enabled(adev, head)) { + con->features &= ~BIT(head->block); + put_obj(obj); + } + } + + return 0; +} + +/* wrapper of psp_ras_enable_features */ +int amdgpu_ras_feature_enable(struct amdgpu_device *adev, + struct ras_common_if *head, bool enable) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + union ta_ras_cmd_input info; + int ret; + + if (!con) + return -EINVAL; + + if (!enable) { + info.disable_features = (struct ta_ras_disable_features_input) { + .block_id = amdgpu_ras_block_to_ta(head->block), + .error_type = amdgpu_ras_error_to_ta(head->type), + }; + } else { + info.enable_features = (struct ta_ras_enable_features_input) { + .block_id = amdgpu_ras_block_to_ta(head->block), + .error_type = amdgpu_ras_error_to_ta(head->type), + }; + } + + /* Do not enable if it is not allowed. */ + WARN_ON(enable && !amdgpu_ras_is_feature_allowed(adev, head)); + /* Are we alerady in that state we are going to set? */ + if (!(!!enable ^ !!amdgpu_ras_is_feature_enabled(adev, head))) + return 0; + + ret = psp_ras_enable_features(&adev->psp, &info, enable); + if (ret) { + DRM_ERROR("RAS ERROR: %s %s feature failed ret %d\n", + enable ? "enable":"disable", + ras_block_str(head->block), + ret); + return -EINVAL; + } + + /* setup the obj */ + __amdgpu_ras_feature_enable(adev, head, enable); + + return 0; +} + +/* Only used in device probe stage and called only once. */ +int amdgpu_ras_feature_enable_on_boot(struct amdgpu_device *adev, + struct ras_common_if *head, bool enable) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + int ret; + + if (!con) + return -EINVAL; + + if (con->flags & AMDGPU_RAS_FLAG_INIT_BY_VBIOS) { + /* If ras is enabled by vbios, we set up ras object first in + * both case. For enable, that is all what we need do. For + * disable, we need perform a ras TA disable cmd after that. + */ + ret = __amdgpu_ras_feature_enable(adev, head, 1); + if (ret) + return ret; + + if (!enable) + ret = amdgpu_ras_feature_enable(adev, head, 0); + } else + ret = amdgpu_ras_feature_enable(adev, head, enable); + + return ret; +} + +static int amdgpu_ras_disable_all_features(struct amdgpu_device *adev, + bool bypass) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_manager *obj, *tmp; + + list_for_each_entry_safe(obj, tmp, &con->head, node) { + /* bypass psp. + * aka just release the obj and corresponding flags + */ + if (bypass) { + if (__amdgpu_ras_feature_enable(adev, &obj->head, 0)) + break; + } else { + if (amdgpu_ras_feature_enable(adev, &obj->head, 0)) + break; + } + } + + return con->features; +} + +static int amdgpu_ras_enable_all_features(struct amdgpu_device *adev, + bool bypass) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + int ras_block_count = AMDGPU_RAS_BLOCK_COUNT; + int i; + const enum amdgpu_ras_error_type default_ras_type = + AMDGPU_RAS_ERROR__NONE; + + for (i = 0; i < ras_block_count; i++) { + struct ras_common_if head = { + .block = i, + .type = default_ras_type, + .sub_block_index = 0, + }; + strcpy(head.name, ras_block_str(i)); + if (bypass) { + /* + * bypass psp. vbios enable ras for us. + * so just create the obj + */ + if (__amdgpu_ras_feature_enable(adev, &head, 1)) + break; + } else { + if (amdgpu_ras_feature_enable(adev, &head, 1)) + break; + } + } + + return con->features; +} +/* feature ctl end */ + +/* query/inject/cure begin */ +int amdgpu_ras_error_query(struct amdgpu_device *adev, + struct ras_query_if *info) +{ + struct ras_manager *obj = amdgpu_ras_find_obj(adev, &info->head); + + if (!obj) + return -EINVAL; + /* TODO might read the register to read the count */ + + info->ue_count = obj->err_data.ue_count; + info->ce_count = obj->err_data.ce_count; + + return 0; +} + +/* wrapper of psp_ras_trigger_error */ +int amdgpu_ras_error_inject(struct amdgpu_device *adev, + struct ras_inject_if *info) +{ + struct ras_manager *obj = amdgpu_ras_find_obj(adev, &info->head); + struct ta_ras_trigger_error_input block_info = { + .block_id = amdgpu_ras_block_to_ta(info->head.block), + .inject_error_type = amdgpu_ras_error_to_ta(info->head.type), + .sub_block_index = info->head.sub_block_index, + .address = info->address, + .value = info->value, + }; + int ret = 0; + + if (!obj) + return -EINVAL; + + ret = psp_ras_trigger_error(&adev->psp, &block_info); + if (ret) + DRM_ERROR("RAS ERROR: inject %s error failed ret %d\n", + ras_block_str(info->head.block), + ret); + + return ret; +} + +int amdgpu_ras_error_cure(struct amdgpu_device *adev, + struct ras_cure_if *info) +{ + /* psp fw has no cure interface for now. */ + return 0; +} + +/* get the total error counts on all IPs */ +int amdgpu_ras_query_error_count(struct amdgpu_device *adev, + bool is_ce) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_manager *obj; + struct ras_err_data data = {0, 0}; + + if (!con) + return -EINVAL; + + list_for_each_entry(obj, &con->head, node) { + struct ras_query_if info = { + .head = obj->head, + }; + + if (amdgpu_ras_error_query(adev, &info)) + return -EINVAL; + + data.ce_count += info.ce_count; + data.ue_count += info.ue_count; + } + + return is_ce ? data.ce_count : data.ue_count; +} +/* query/inject/cure end */ + + +/* sysfs begin */ + +static ssize_t amdgpu_ras_sysfs_features_read(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct amdgpu_ras *con = + container_of(attr, struct amdgpu_ras, features_attr); + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + struct ras_common_if head; + int ras_block_count = AMDGPU_RAS_BLOCK_COUNT; + int i; + ssize_t s; + struct ras_manager *obj; + + s = scnprintf(buf, PAGE_SIZE, "feature mask: 0x%x\n", con->features); + + for (i = 0; i < ras_block_count; i++) { + head.block = i; + + if (amdgpu_ras_is_feature_enabled(adev, &head)) { + obj = amdgpu_ras_find_obj(adev, &head); + s += scnprintf(&buf[s], PAGE_SIZE - s, + "%s: %s\n", + ras_block_str(i), + ras_err_str(obj->head.type)); + } else + s += scnprintf(&buf[s], PAGE_SIZE - s, + "%s: disabled\n", + ras_block_str(i)); + } + + return s; +} + +static int amdgpu_ras_sysfs_create_feature_node(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct attribute *attrs[] = { + &con->features_attr.attr, + NULL + }; + struct attribute_group group = { + .name = "ras", + .attrs = attrs, + }; + + con->features_attr = (struct device_attribute) { + .attr = { + .name = "features", + .mode = S_IRUGO, + }, + .show = amdgpu_ras_sysfs_features_read, + }; + sysfs_attr_init(attrs[0]); + + return sysfs_create_group(&adev->dev->kobj, &group); +} + +static int amdgpu_ras_sysfs_remove_feature_node(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct attribute *attrs[] = { + &con->features_attr.attr, + NULL + }; + struct attribute_group group = { + .name = "ras", + .attrs = attrs, + }; + + sysfs_remove_group(&adev->dev->kobj, &group); + + return 0; +} + +int amdgpu_ras_sysfs_create(struct amdgpu_device *adev, + struct ras_fs_if *head) +{ + struct ras_manager *obj = amdgpu_ras_find_obj(adev, &head->head); + + if (!obj || obj->attr_inuse) + return -EINVAL; + + get_obj(obj); + + memcpy(obj->fs_data.sysfs_name, + head->sysfs_name, + sizeof(obj->fs_data.sysfs_name)); + + obj->sysfs_attr = (struct device_attribute){ + .attr = { + .name = obj->fs_data.sysfs_name, + .mode = S_IRUGO, + }, + .show = amdgpu_ras_sysfs_read, + }; + sysfs_attr_init(&obj->sysfs_attr.attr); + + if (sysfs_add_file_to_group(&adev->dev->kobj, + &obj->sysfs_attr.attr, + "ras")) { + put_obj(obj); + return -EINVAL; + } + + obj->attr_inuse = 1; + + return 0; +} + +int amdgpu_ras_sysfs_remove(struct amdgpu_device *adev, + struct ras_common_if *head) +{ + struct ras_manager *obj = amdgpu_ras_find_obj(adev, head); + + if (!obj || !obj->attr_inuse) + return -EINVAL; + + sysfs_remove_file_from_group(&adev->dev->kobj, + &obj->sysfs_attr.attr, + "ras"); + obj->attr_inuse = 0; + put_obj(obj); + + return 0; +} + +static int amdgpu_ras_sysfs_remove_all(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_manager *obj, *tmp; + + list_for_each_entry_safe(obj, tmp, &con->head, node) { + amdgpu_ras_sysfs_remove(adev, &obj->head); + } + + amdgpu_ras_sysfs_remove_feature_node(adev); + + return 0; +} +/* sysfs end */ + +/* debugfs begin */ +static int amdgpu_ras_debugfs_create_ctrl_node(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct drm_minor *minor = adev->ddev->primary; + struct dentry *root = minor->debugfs_root, *dir; + struct dentry *ent; + + dir = debugfs_create_dir("ras", root); + if (IS_ERR(dir)) + return -EINVAL; + + con->dir = dir; + + ent = debugfs_create_file("ras_ctrl", + S_IWUGO | S_IRUGO, con->dir, + adev, &amdgpu_ras_debugfs_ctrl_ops); + if (IS_ERR(ent)) { + debugfs_remove(con->dir); + return -EINVAL; + } + + con->ent = ent; + return 0; +} + +int amdgpu_ras_debugfs_create(struct amdgpu_device *adev, + struct ras_fs_if *head) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_manager *obj = amdgpu_ras_find_obj(adev, &head->head); + struct dentry *ent; + + if (!obj || obj->ent) + return -EINVAL; + + get_obj(obj); + + memcpy(obj->fs_data.debugfs_name, + head->debugfs_name, + sizeof(obj->fs_data.debugfs_name)); + + ent = debugfs_create_file(obj->fs_data.debugfs_name, + S_IWUGO | S_IRUGO, con->dir, + obj, &amdgpu_ras_debugfs_ops); + + if (IS_ERR(ent)) + return -EINVAL; + + obj->ent = ent; + + return 0; +} + +int amdgpu_ras_debugfs_remove(struct amdgpu_device *adev, + struct ras_common_if *head) +{ + struct ras_manager *obj = amdgpu_ras_find_obj(adev, head); + + if (!obj || !obj->ent) + return 0; + + debugfs_remove(obj->ent); + obj->ent = NULL; + put_obj(obj); + + return 0; +} + +static int amdgpu_ras_debugfs_remove_all(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_manager *obj, *tmp; + + list_for_each_entry_safe(obj, tmp, &con->head, node) { + amdgpu_ras_debugfs_remove(adev, &obj->head); + } + + debugfs_remove(con->ent); + debugfs_remove(con->dir); + con->dir = NULL; + con->ent = NULL; + + return 0; +} +/* debugfs end */ + +/* ras fs */ + +static int amdgpu_ras_fs_init(struct amdgpu_device *adev) +{ + amdgpu_ras_sysfs_create_feature_node(adev); + amdgpu_ras_debugfs_create_ctrl_node(adev); + + return 0; +} + +static int amdgpu_ras_fs_fini(struct amdgpu_device *adev) +{ + amdgpu_ras_debugfs_remove_all(adev); + amdgpu_ras_sysfs_remove_all(adev); + return 0; +} +/* ras fs end */ + +/* ih begin */ +static void amdgpu_ras_interrupt_handler(struct ras_manager *obj) +{ + struct ras_ih_data *data = &obj->ih_data; + struct amdgpu_iv_entry entry; + int ret; + + while (data->rptr != data->wptr) { + rmb(); + memcpy(&entry, &data->ring[data->rptr], + data->element_size); + + wmb(); + data->rptr = (data->aligned_element_size + + data->rptr) % data->ring_size; + + /* Let IP handle its data, maybe we need get the output + * from the callback to udpate the error type/count, etc + */ + if (data->cb) { + ret = data->cb(obj->adev, &entry); + /* ue will trigger an interrupt, and in that case + * we need do a reset to recovery the whole system. + * But leave IP do that recovery, here we just dispatch + * the error. + */ + if (ret == AMDGPU_RAS_UE) { + obj->err_data.ue_count++; + } + /* Might need get ce count by register, but not all IP + * saves ce count, some IP just use one bit or two bits + * to indicate ce happened. + */ + } + } +} + +static void amdgpu_ras_interrupt_process_handler(struct work_struct *work) +{ + struct ras_ih_data *data = + container_of(work, struct ras_ih_data, ih_work); + struct ras_manager *obj = + container_of(data, struct ras_manager, ih_data); + + amdgpu_ras_interrupt_handler(obj); +} + +int amdgpu_ras_interrupt_dispatch(struct amdgpu_device *adev, + struct ras_dispatch_if *info) +{ + struct ras_manager *obj = amdgpu_ras_find_obj(adev, &info->head); + struct ras_ih_data *data = &obj->ih_data; + + if (!obj) + return -EINVAL; + + if (data->inuse == 0) + return 0; + + /* Might be overflow... */ + memcpy(&data->ring[data->wptr], info->entry, + data->element_size); + + wmb(); + data->wptr = (data->aligned_element_size + + data->wptr) % data->ring_size; + + schedule_work(&data->ih_work); + + return 0; +} + +int amdgpu_ras_interrupt_remove_handler(struct amdgpu_device *adev, + struct ras_ih_if *info) +{ + struct ras_manager *obj = amdgpu_ras_find_obj(adev, &info->head); + struct ras_ih_data *data; + + if (!obj) + return -EINVAL; + + data = &obj->ih_data; + if (data->inuse == 0) + return 0; + + cancel_work_sync(&data->ih_work); + + kfree(data->ring); + memset(data, 0, sizeof(*data)); + put_obj(obj); + + return 0; +} + +int amdgpu_ras_interrupt_add_handler(struct amdgpu_device *adev, + struct ras_ih_if *info) +{ + struct ras_manager *obj = amdgpu_ras_find_obj(adev, &info->head); + struct ras_ih_data *data; + + if (!obj) { + /* in case we registe the IH before enable ras feature */ + obj = amdgpu_ras_create_obj(adev, &info->head); + if (!obj) + return -EINVAL; + } else + get_obj(obj); + + data = &obj->ih_data; + /* add the callback.etc */ + *data = (struct ras_ih_data) { + .inuse = 0, + .cb = info->cb, + .element_size = sizeof(struct amdgpu_iv_entry), + .rptr = 0, + .wptr = 0, + }; + + INIT_WORK(&data->ih_work, amdgpu_ras_interrupt_process_handler); + + data->aligned_element_size = ALIGN(data->element_size, 8); + /* the ring can store 64 iv entries. */ + data->ring_size = 64 * data->aligned_element_size; + data->ring = kmalloc(data->ring_size, GFP_KERNEL); + if (!data->ring) { + put_obj(obj); + return -ENOMEM; + } + + /* IH is ready */ + data->inuse = 1; + + return 0; +} + +static int amdgpu_ras_interrupt_remove_all(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_manager *obj, *tmp; + + list_for_each_entry_safe(obj, tmp, &con->head, node) { + struct ras_ih_if info = { + .head = obj->head, + }; + amdgpu_ras_interrupt_remove_handler(adev, &info); + } + + return 0; +} +/* ih end */ + +/* recovery begin */ +static void amdgpu_ras_do_recovery(struct work_struct *work) +{ + struct amdgpu_ras *ras = + container_of(work, struct amdgpu_ras, recovery_work); + + amdgpu_device_gpu_recover(ras->adev, 0); + atomic_set(&ras->in_recovery, 0); +} + +static int amdgpu_ras_release_vram(struct amdgpu_device *adev, + struct amdgpu_bo **bo_ptr) +{ + /* no need to free it actually. */ + amdgpu_bo_free_kernel(bo_ptr, NULL, NULL); + return 0; +} + +/* reserve vram with size@offset */ +static int amdgpu_ras_reserve_vram(struct amdgpu_device *adev, + uint64_t offset, uint64_t size, + struct amdgpu_bo **bo_ptr) +{ + struct ttm_operation_ctx ctx = { false, false }; + struct amdgpu_bo_param bp; + int r = 0; + int i; + struct amdgpu_bo *bo; + + if (bo_ptr) + *bo_ptr = NULL; + memset(&bp, 0, sizeof(bp)); + bp.size = size; + bp.byte_align = PAGE_SIZE; + bp.domain = AMDGPU_GEM_DOMAIN_VRAM; + bp.flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS | + AMDGPU_GEM_CREATE_NO_CPU_ACCESS; + bp.type = ttm_bo_type_kernel; + bp.resv = NULL; + + r = amdgpu_bo_create(adev, &bp, &bo); + if (r) + return -EINVAL; + + r = amdgpu_bo_reserve(bo, false); + if (r) + goto error_reserve; + + offset = ALIGN(offset, PAGE_SIZE); + for (i = 0; i < bo->placement.num_placement; ++i) { + bo->placements[i].fpfn = offset >> PAGE_SHIFT; + bo->placements[i].lpfn = (offset + size) >> PAGE_SHIFT; + } + + ttm_bo_mem_put(&bo->tbo, &bo->tbo.mem); + r = ttm_bo_mem_space(&bo->tbo, &bo->placement, &bo->tbo.mem, &ctx); + if (r) + goto error_pin; + + r = amdgpu_bo_pin_restricted(bo, + AMDGPU_GEM_DOMAIN_VRAM, + offset, + offset + size); + if (r) + goto error_pin; + + if (bo_ptr) + *bo_ptr = bo; + + amdgpu_bo_unreserve(bo); + return r; + +error_pin: + amdgpu_bo_unreserve(bo); +error_reserve: + amdgpu_bo_unref(&bo); + return r; +} + +/* alloc/realloc bps array */ +static int amdgpu_ras_realloc_eh_data_space(struct amdgpu_device *adev, + struct ras_err_handler_data *data, int pages) +{ + unsigned int old_space = data->count + data->space_left; + unsigned int new_space = old_space + pages; + unsigned int align_space = ALIGN(new_space, 1024); + void *tmp = kmalloc(align_space * sizeof(*data->bps), GFP_KERNEL); + + if (!tmp) + return -ENOMEM; + + if (data->bps) { + memcpy(tmp, data->bps, + data->count * sizeof(*data->bps)); + kfree(data->bps); + } + + data->bps = tmp; + data->space_left += align_space - old_space; + return 0; +} + +/* it deal with vram only. */ +int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev, + unsigned long *bps, int pages) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_err_handler_data *data; + int i = pages; + int ret = 0; + + if (!con || !con->eh_data || !bps || pages <= 0) + return 0; + + mutex_lock(&con->recovery_lock); + data = con->eh_data; + if (!data) + goto out; + + if (data->space_left <= pages) + if (amdgpu_ras_realloc_eh_data_space(adev, data, pages)) { + ret = -ENOMEM; + goto out; + } + + while (i--) + data->bps[data->count++].bp = bps[i]; + + data->space_left -= pages; +out: + mutex_unlock(&con->recovery_lock); + + return ret; +} + +/* called in gpu recovery/init */ +int amdgpu_ras_reserve_bad_pages(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_err_handler_data *data; + uint64_t bp; + struct amdgpu_bo *bo; + int i; + + if (!con || !con->eh_data) + return 0; + + mutex_lock(&con->recovery_lock); + data = con->eh_data; + if (!data) + goto out; + /* reserve vram at driver post stage. */ + for (i = data->last_reserved; i < data->count; i++) { + bp = data->bps[i].bp; + + if (amdgpu_ras_reserve_vram(adev, bp << PAGE_SHIFT, + PAGE_SIZE, &bo)) + DRM_ERROR("RAS ERROR: reserve vram %llx fail\n", bp); + + data->bps[i].bo = bo; + data->last_reserved = i + 1; + } +out: + mutex_unlock(&con->recovery_lock); + return 0; +} + +/* called when driver unload */ +static int amdgpu_ras_release_bad_pages(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_err_handler_data *data; + struct amdgpu_bo *bo; + int i; + + if (!con || !con->eh_data) + return 0; + + mutex_lock(&con->recovery_lock); + data = con->eh_data; + if (!data) + goto out; + + for (i = data->last_reserved - 1; i >= 0; i--) { + bo = data->bps[i].bo; + + amdgpu_ras_release_vram(adev, &bo); + + data->bps[i].bo = bo; + data->last_reserved = i; + } +out: + mutex_unlock(&con->recovery_lock); + return 0; +} + +static int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev) +{ + /* TODO + * write the array to eeprom when SMU disabled. + */ + return 0; +} + +static int amdgpu_ras_load_bad_pages(struct amdgpu_device *adev) +{ + /* TODO + * read the array to eeprom when SMU disabled. + */ + return 0; +} + +static int amdgpu_ras_recovery_init(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_err_handler_data **data = &con->eh_data; + + *data = kmalloc(sizeof(**data), + GFP_KERNEL|__GFP_ZERO); + if (!*data) + return -ENOMEM; + + mutex_init(&con->recovery_lock); + INIT_WORK(&con->recovery_work, amdgpu_ras_do_recovery); + atomic_set(&con->in_recovery, 0); + con->adev = adev; + + amdgpu_ras_load_bad_pages(adev); + amdgpu_ras_reserve_bad_pages(adev); + + return 0; +} + +static int amdgpu_ras_recovery_fini(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_err_handler_data *data = con->eh_data; + + cancel_work_sync(&con->recovery_work); + amdgpu_ras_save_bad_pages(adev); + amdgpu_ras_release_bad_pages(adev); + + mutex_lock(&con->recovery_lock); + con->eh_data = NULL; + kfree(data->bps); + kfree(data); + mutex_unlock(&con->recovery_lock); + + return 0; +} +/* recovery end */ + +/* + * check hardware's ras ability which will be saved in hw_supported. + * if hardware does not support ras, we can skip some ras initializtion and + * forbid some ras operations from IP. + * if software itself, say boot parameter, limit the ras ability. We still + * need allow IP do some limited operations, like disable. In such case, + * we have to initialize ras as normal. but need check if operation is + * allowed or not in each function. + */ +static void amdgpu_ras_check_supported(struct amdgpu_device *adev, + uint32_t *hw_supported, uint32_t *supported) +{ + *hw_supported = 0; + *supported = 0; + + if (amdgpu_sriov_vf(adev) || + adev->asic_type != CHIP_VEGA20) + return; + + if (adev->is_atom_fw && + (amdgpu_atomfirmware_mem_ecc_supported(adev) || + amdgpu_atomfirmware_sram_ecc_supported(adev))) + *hw_supported = AMDGPU_RAS_BLOCK_MASK; + + *supported = amdgpu_ras_enable == 0 ? + 0 : *hw_supported & amdgpu_ras_mask; +} + +int amdgpu_ras_init(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + + if (con) + return 0; + + con = kmalloc(sizeof(struct amdgpu_ras) + + sizeof(struct ras_manager) * AMDGPU_RAS_BLOCK_COUNT, + GFP_KERNEL|__GFP_ZERO); + if (!con) + return -ENOMEM; + + con->objs = (struct ras_manager *)(con + 1); + + amdgpu_ras_set_context(adev, con); + + amdgpu_ras_check_supported(adev, &con->hw_supported, + &con->supported); + con->features = 0; + INIT_LIST_HEAD(&con->head); + /* Might need get this flag from vbios. */ + con->flags = RAS_DEFAULT_FLAGS; + + if (amdgpu_ras_recovery_init(adev)) + goto recovery_out; + + amdgpu_ras_mask &= AMDGPU_RAS_BLOCK_MASK; + + if (amdgpu_ras_fs_init(adev)) + goto fs_out; + + amdgpu_ras_self_test(adev); + + DRM_INFO("RAS INFO: ras initialized successfully, " + "hardware ability[%x] ras_mask[%x]\n", + con->hw_supported, con->supported); + return 0; +fs_out: + amdgpu_ras_recovery_fini(adev); +recovery_out: + amdgpu_ras_set_context(adev, NULL); + kfree(con); + + return -EINVAL; +} + +/* do some init work after IP late init as dependence */ +void amdgpu_ras_post_init(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_manager *obj, *tmp; + + if (!con) + return; + + if (con->flags & AMDGPU_RAS_FLAG_INIT_BY_VBIOS) { + /* Set up all other IPs which are not implemented. There is a + * tricky thing that IP's actual ras error type should be + * MULTI_UNCORRECTABLE, but as driver does not handle it, so + * ERROR_NONE make sense anyway. + */ + amdgpu_ras_enable_all_features(adev, 1); + + /* We enable ras on all hw_supported block, but as boot + * parameter might disable some of them and one or more IP has + * not implemented yet. So we disable them on behalf. + */ + list_for_each_entry_safe(obj, tmp, &con->head, node) { + if (!amdgpu_ras_is_supported(adev, obj->head.block)) { + amdgpu_ras_feature_enable(adev, &obj->head, 0); + /* there should be no any reference. */ + WARN_ON(alive_obj(obj)); + } + } + } +} + +/* do some fini work before IP fini as dependence */ +int amdgpu_ras_pre_fini(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + + if (!con) + return 0; + + /* Need disable ras on all IPs here before ip [hw/sw]fini */ + amdgpu_ras_disable_all_features(adev, 0); + amdgpu_ras_recovery_fini(adev); + return 0; +} + +int amdgpu_ras_fini(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + + if (!con) + return 0; + + amdgpu_ras_fs_fini(adev); + amdgpu_ras_interrupt_remove_all(adev); + + WARN(con->features, "Feature mask is not cleared"); + + if (con->features) + amdgpu_ras_disable_all_features(adev, 1); + + amdgpu_ras_set_context(adev, NULL); + kfree(con); + + return 0; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h new file mode 100644 index 000000000000..eaef5edefc34 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h @@ -0,0 +1,294 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * + */ +#ifndef _AMDGPU_RAS_H +#define _AMDGPU_RAS_H + +#include +#include +#include "amdgpu.h" +#include "amdgpu_psp.h" +#include "ta_ras_if.h" + +enum amdgpu_ras_block { + AMDGPU_RAS_BLOCK__UMC = 0, + AMDGPU_RAS_BLOCK__SDMA, + AMDGPU_RAS_BLOCK__GFX, + AMDGPU_RAS_BLOCK__MMHUB, + AMDGPU_RAS_BLOCK__ATHUB, + AMDGPU_RAS_BLOCK__PCIE_BIF, + AMDGPU_RAS_BLOCK__HDP, + AMDGPU_RAS_BLOCK__XGMI_WAFL, + AMDGPU_RAS_BLOCK__DF, + AMDGPU_RAS_BLOCK__SMN, + AMDGPU_RAS_BLOCK__SEM, + AMDGPU_RAS_BLOCK__MP0, + AMDGPU_RAS_BLOCK__MP1, + AMDGPU_RAS_BLOCK__FUSE, + + AMDGPU_RAS_BLOCK__LAST +}; + +#define AMDGPU_RAS_BLOCK_COUNT AMDGPU_RAS_BLOCK__LAST +#define AMDGPU_RAS_BLOCK_MASK ((1ULL << AMDGPU_RAS_BLOCK_COUNT) - 1) + +enum amdgpu_ras_error_type { + AMDGPU_RAS_ERROR__NONE = 0, + AMDGPU_RAS_ERROR__PARITY = 1, + AMDGPU_RAS_ERROR__SINGLE_CORRECTABLE = 2, + AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE = 4, + AMDGPU_RAS_ERROR__POISON = 8, +}; + +enum amdgpu_ras_ret { + AMDGPU_RAS_SUCCESS = 0, + AMDGPU_RAS_FAIL, + AMDGPU_RAS_UE, + AMDGPU_RAS_CE, + AMDGPU_RAS_PT, +}; + +struct ras_common_if { + enum amdgpu_ras_block block; + enum amdgpu_ras_error_type type; + uint32_t sub_block_index; + /* block name */ + char name[32]; +}; + +typedef int (*ras_ih_cb)(struct amdgpu_device *adev, + struct amdgpu_iv_entry *entry); + +struct amdgpu_ras { + /* ras infrastructure */ + /* for ras itself. */ + uint32_t hw_supported; + /* for IP to check its ras ability. */ + uint32_t supported; + uint32_t features; + struct list_head head; + /* debugfs */ + struct dentry *dir; + /* debugfs ctrl */ + struct dentry *ent; + /* sysfs */ + struct device_attribute features_attr; + /* block array */ + struct ras_manager *objs; + + /* gpu recovery */ + struct work_struct recovery_work; + atomic_t in_recovery; + struct amdgpu_device *adev; + /* error handler data */ + struct ras_err_handler_data *eh_data; + struct mutex recovery_lock; + + uint32_t flags; +}; + +/* interfaces for IP */ + +struct ras_fs_if { + struct ras_common_if head; + char sysfs_name[32]; + char debugfs_name[32]; +}; + +struct ras_query_if { + struct ras_common_if head; + unsigned long ue_count; + unsigned long ce_count; +}; + +struct ras_inject_if { + struct ras_common_if head; + uint64_t address; + uint64_t value; +}; + +struct ras_cure_if { + struct ras_common_if head; + uint64_t address; +}; + +struct ras_ih_if { + struct ras_common_if head; + ras_ih_cb cb; +}; + +struct ras_dispatch_if { + struct ras_common_if head; + struct amdgpu_iv_entry *entry; +}; + +struct ras_debug_if { + union { + struct ras_common_if head; + struct ras_inject_if inject; + }; + int op; +}; +/* work flow + * vbios + * 1: ras feature enable (enabled by default) + * psp + * 2: ras framework init (in ip_init) + * IP + * 3: IH add + * 4: debugfs/sysfs create + * 5: query/inject + * 6: debugfs/sysfs remove + * 7: IH remove + * 8: feature disable + */ + +#define amdgpu_ras_get_context(adev) ((adev)->psp.ras.ras) +#define amdgpu_ras_set_context(adev, ras_con) ((adev)->psp.ras.ras = (ras_con)) + +/* check if ras is supported on block, say, sdma, gfx */ +static inline int amdgpu_ras_is_supported(struct amdgpu_device *adev, + unsigned int block) +{ + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + + return ras && (ras->supported & (1 << block)); +} + +int amdgpu_ras_query_error_count(struct amdgpu_device *adev, + bool is_ce); + +/* error handling functions */ +int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev, + unsigned long *bps, int pages); + +int amdgpu_ras_reserve_bad_pages(struct amdgpu_device *adev); + +static inline int amdgpu_ras_reset_gpu(struct amdgpu_device *adev, + bool is_baco) +{ + /* remove me when gpu reset works on vega20 A1. */ +#if 0 + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + + if (atomic_cmpxchg(&ras->in_recovery, 0, 1) == 0) + schedule_work(&ras->recovery_work); +#endif + return 0; +} + +static inline enum ta_ras_block +amdgpu_ras_block_to_ta(enum amdgpu_ras_block block) { + switch (block) { + case AMDGPU_RAS_BLOCK__UMC: + return TA_RAS_BLOCK__UMC; + case AMDGPU_RAS_BLOCK__SDMA: + return TA_RAS_BLOCK__SDMA; + case AMDGPU_RAS_BLOCK__GFX: + return TA_RAS_BLOCK__GFX; + case AMDGPU_RAS_BLOCK__MMHUB: + return TA_RAS_BLOCK__MMHUB; + case AMDGPU_RAS_BLOCK__ATHUB: + return TA_RAS_BLOCK__ATHUB; + case AMDGPU_RAS_BLOCK__PCIE_BIF: + return TA_RAS_BLOCK__PCIE_BIF; + case AMDGPU_RAS_BLOCK__HDP: + return TA_RAS_BLOCK__HDP; + case AMDGPU_RAS_BLOCK__XGMI_WAFL: + return TA_RAS_BLOCK__XGMI_WAFL; + case AMDGPU_RAS_BLOCK__DF: + return TA_RAS_BLOCK__DF; + case AMDGPU_RAS_BLOCK__SMN: + return TA_RAS_BLOCK__SMN; + case AMDGPU_RAS_BLOCK__SEM: + return TA_RAS_BLOCK__SEM; + case AMDGPU_RAS_BLOCK__MP0: + return TA_RAS_BLOCK__MP0; + case AMDGPU_RAS_BLOCK__MP1: + return TA_RAS_BLOCK__MP1; + case AMDGPU_RAS_BLOCK__FUSE: + return TA_RAS_BLOCK__FUSE; + default: + WARN_ONCE(1, "RAS ERROR: unexpected block id %d\n", block); + return TA_RAS_BLOCK__UMC; + } +} + +static inline enum ta_ras_error_type +amdgpu_ras_error_to_ta(enum amdgpu_ras_error_type error) { + switch (error) { + case AMDGPU_RAS_ERROR__NONE: + return TA_RAS_ERROR__NONE; + case AMDGPU_RAS_ERROR__PARITY: + return TA_RAS_ERROR__PARITY; + case AMDGPU_RAS_ERROR__SINGLE_CORRECTABLE: + return TA_RAS_ERROR__SINGLE_CORRECTABLE; + case AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE: + return TA_RAS_ERROR__MULTI_UNCORRECTABLE; + case AMDGPU_RAS_ERROR__POISON: + return TA_RAS_ERROR__POISON; + default: + WARN_ONCE(1, "RAS ERROR: unexpected error type %d\n", error); + return TA_RAS_ERROR__NONE; + } +} + +/* called in ip_init and ip_fini */ +int amdgpu_ras_init(struct amdgpu_device *adev); +void amdgpu_ras_post_init(struct amdgpu_device *adev); +int amdgpu_ras_fini(struct amdgpu_device *adev); +int amdgpu_ras_pre_fini(struct amdgpu_device *adev); + +int amdgpu_ras_feature_enable(struct amdgpu_device *adev, + struct ras_common_if *head, bool enable); + +int amdgpu_ras_feature_enable_on_boot(struct amdgpu_device *adev, + struct ras_common_if *head, bool enable); + +int amdgpu_ras_sysfs_create(struct amdgpu_device *adev, + struct ras_fs_if *head); + +int amdgpu_ras_sysfs_remove(struct amdgpu_device *adev, + struct ras_common_if *head); + +int amdgpu_ras_debugfs_create(struct amdgpu_device *adev, + struct ras_fs_if *head); + +int amdgpu_ras_debugfs_remove(struct amdgpu_device *adev, + struct ras_common_if *head); + +int amdgpu_ras_error_query(struct amdgpu_device *adev, + struct ras_query_if *info); + +int amdgpu_ras_error_inject(struct amdgpu_device *adev, + struct ras_inject_if *info); + +int amdgpu_ras_interrupt_add_handler(struct amdgpu_device *adev, + struct ras_ih_if *info); + +int amdgpu_ras_interrupt_remove_handler(struct amdgpu_device *adev, + struct ras_ih_if *info); + +int amdgpu_ras_interrupt_dispatch(struct amdgpu_device *adev, + struct ras_dispatch_if *info); +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index 335a0edf114b..8f5026c123ef 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -248,6 +248,8 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring, */ if (ring->funcs->type == AMDGPU_RING_TYPE_KIQ) sched_hw_submission = max(sched_hw_submission, 256); + else if (ring == &adev->sdma.instance[0].page) + sched_hw_submission = 256; if (ring->adev == NULL) { if (adev->num_rings >= AMDGPU_MAX_RINGS) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h index 16b1a6ae5ba6..1ba9ba3b54f7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h @@ -28,9 +28,8 @@ #define AMDGPU_MAX_SDMA_INSTANCES 2 enum amdgpu_sdma_irq { - AMDGPU_SDMA_IRQ_TRAP0 = 0, - AMDGPU_SDMA_IRQ_TRAP1, - + AMDGPU_SDMA_IRQ_INSTANCE0 = 0, + AMDGPU_SDMA_IRQ_INSTANCE1, AMDGPU_SDMA_IRQ_LAST }; @@ -49,9 +48,11 @@ struct amdgpu_sdma { struct amdgpu_sdma_instance instance[AMDGPU_MAX_SDMA_INSTANCES]; struct amdgpu_irq_src trap_irq; struct amdgpu_irq_src illegal_inst_irq; + struct amdgpu_irq_src ecc_irq; int num_instances; uint32_t srbm_soft_reset; bool has_page_queue; + struct ras_common_if *ras_if; }; /* diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 73e71e61dc99..0c52d1f9fe0f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -50,8 +50,6 @@ #include "amdgpu_sdma.h" #include "bif/bif_4_1_d.h" -#define DRM_FILE_PAGE_OFFSET (0x100000000ULL >> PAGE_SHIFT) - static int amdgpu_map_buffer(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem, unsigned num_pages, uint64_t offset, unsigned window, @@ -1424,6 +1422,13 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo, struct dma_fence *f; int i; + /* Don't evict VM page tables while they are busy, otherwise we can't + * cleanly handle page faults. + */ + if (bo->type == ttm_bo_type_kernel && + !reservation_object_test_signaled_rcu(bo->resv, true)) + return false; + /* If bo is a KFD BO, check if the bo belongs to the current process. * If true, then return false as any KFD process needs all its BOs to * be resident to run successfully @@ -1671,7 +1676,6 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) r = ttm_bo_device_init(&adev->mman.bdev, &amdgpu_bo_driver, adev->ddev->anon_inode->i_mapping, - DRM_FILE_PAGE_OFFSET, adev->need_dma32); if (r) { DRM_ERROR("failed initializing buffer object driver(%d).\n", r); @@ -1877,14 +1881,9 @@ void amdgpu_ttm_set_buffer_funcs_status(struct amdgpu_device *adev, bool enable) int amdgpu_mmap(struct file *filp, struct vm_area_struct *vma) { - struct drm_file *file_priv; - struct amdgpu_device *adev; + struct drm_file *file_priv = filp->private_data; + struct amdgpu_device *adev = file_priv->minor->dev->dev_private; - if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET)) - return -EINVAL; - - file_priv = filp->private_data; - adev = file_priv->minor->dev->dev_private; if (adev == NULL) return -EINVAL; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index 462a04e0f5e6..7e7f9ed89ee1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -375,4 +375,53 @@ void amdgpu_virt_init_data_exchange(struct amdgpu_device *adev) } } +static uint32_t parse_clk(char *buf, bool min) +{ + char *ptr = buf; + uint32_t clk = 0; + + do { + ptr = strchr(ptr, ':'); + if (!ptr) + break; + ptr+=2; + clk = simple_strtoul(ptr, NULL, 10); + } while (!min); + + return clk * 100; +} + +uint32_t amdgpu_virt_get_sclk(struct amdgpu_device *adev, bool lowest) +{ + char *buf = NULL; + uint32_t clk = 0; + + buf = kzalloc(PAGE_SIZE, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + adev->virt.ops->get_pp_clk(adev, PP_SCLK, buf); + clk = parse_clk(buf, lowest); + + kfree(buf); + + return clk; +} + +uint32_t amdgpu_virt_get_mclk(struct amdgpu_device *adev, bool lowest) +{ + char *buf = NULL; + uint32_t clk = 0; + + buf = kzalloc(PAGE_SIZE, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + adev->virt.ops->get_pp_clk(adev, PP_MCLK, buf); + clk = parse_clk(buf, lowest); + + kfree(buf); + + return clk; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h index 722deefc0a7e..584947b7ccf3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h @@ -57,6 +57,8 @@ struct amdgpu_virt_ops { int (*reset_gpu)(struct amdgpu_device *adev); int (*wait_reset)(struct amdgpu_device *adev); void (*trans_msg)(struct amdgpu_device *adev, u32 req, u32 data1, u32 data2, u32 data3); + int (*get_pp_clk)(struct amdgpu_device *adev, u32 type, char *buf); + int (*force_dpm_level)(struct amdgpu_device *adev, u32 level); }; /* @@ -83,6 +85,8 @@ enum AMDGIM_FEATURE_FLAG { AMDGIM_FEATURE_GIM_LOAD_UCODES = 0x2, /* VRAM LOST by GIM */ AMDGIM_FEATURE_GIM_FLR_VRAMLOST = 0x4, + /* HW PERF SIM in GIM */ + AMDGIM_FEATURE_HW_PERF_SIMULATION = (1 << 3), }; struct amd_sriov_msg_pf2vf_info_header { @@ -252,6 +256,8 @@ struct amdgpu_virt { struct amdgpu_vf_error_buffer vf_errors; struct amdgpu_virt_fw_reserve fw_reserve; uint32_t gim_feature; + /* protect DPM events to GIM */ + struct mutex dpm_mutex; }; #define amdgpu_sriov_enabled(adev) \ @@ -278,6 +284,9 @@ static inline bool is_virtual_machine(void) #endif } +#define amdgim_is_hwperf(adev) \ + ((adev)->virt.gim_feature & AMDGIM_FEATURE_HW_PERF_SIMULATION) + bool amdgpu_virt_mmio_blocked(struct amdgpu_device *adev); void amdgpu_virt_init_setting(struct amdgpu_device *adev); uint32_t amdgpu_virt_kiq_rreg(struct amdgpu_device *adev, uint32_t reg); @@ -295,5 +304,7 @@ int amdgpu_virt_fw_reserve_get_checksum(void *obj, unsigned long obj_size, unsigned int key, unsigned int chksum); void amdgpu_virt_init_data_exchange(struct amdgpu_device *adev); +uint32_t amdgpu_virt_get_sclk(struct amdgpu_device *adev, bool lowest); +uint32_t amdgpu_virt_get_mclk(struct amdgpu_device *adev, bool lowest); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 16fcb56c232b..a07c85815b7a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -34,6 +34,7 @@ #include "amdgpu_trace.h" #include "amdgpu_amdkfd.h" #include "amdgpu_gmc.h" +#include "amdgpu_xgmi.h" /** * DOC: GPUVM @@ -65,50 +66,6 @@ INTERVAL_TREE_DEFINE(struct amdgpu_bo_va_mapping, rb, uint64_t, __subtree_last, #undef START #undef LAST -/** - * struct amdgpu_pte_update_params - Local structure - * - * Encapsulate some VM table update parameters to reduce - * the number of function parameters - * - */ -struct amdgpu_pte_update_params { - - /** - * @adev: amdgpu device we do this update for - */ - struct amdgpu_device *adev; - - /** - * @vm: optional amdgpu_vm we do this update for - */ - struct amdgpu_vm *vm; - - /** - * @src: address where to copy page table entries from - */ - uint64_t src; - - /** - * @ib: indirect buffer to fill with commands - */ - struct amdgpu_ib *ib; - - /** - * @func: Function which actually does the update - */ - void (*func)(struct amdgpu_pte_update_params *params, - struct amdgpu_bo *bo, uint64_t pe, - uint64_t addr, unsigned count, uint32_t incr, - uint64_t flags); - /** - * @pages_addr: - * - * DMA addresses to use for mapping, used during VM update by CPU - */ - dma_addr_t *pages_addr; -}; - /** * struct amdgpu_prt_cb - Helper to disable partial resident texture feature from a fence callback */ @@ -182,6 +139,22 @@ static unsigned amdgpu_vm_num_entries(struct amdgpu_device *adev, return AMDGPU_VM_PTE_COUNT(adev); } +/** + * amdgpu_vm_num_ats_entries - return the number of ATS entries in the root PD + * + * @adev: amdgpu_device pointer + * + * Returns: + * The number of entries in the root page directory which needs the ATS setting. + */ +static unsigned amdgpu_vm_num_ats_entries(struct amdgpu_device *adev) +{ + unsigned shift; + + shift = amdgpu_vm_level_shift(adev, adev->vm_manager.root_level); + return AMDGPU_GMC_HOLE_START >> (shift + AMDGPU_GPU_PAGE_SHIFT); +} + /** * amdgpu_vm_entries_mask - the mask to get the entry number of a PD/PT * @@ -333,7 +306,7 @@ static void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base, return; vm->bulk_moveable = false; - if (bo->tbo.type == ttm_bo_type_kernel) + if (bo->tbo.type == ttm_bo_type_kernel && bo->parent) amdgpu_vm_bo_relocated(base); else amdgpu_vm_bo_idle(base); @@ -504,47 +477,6 @@ static void amdgpu_vm_pt_next(struct amdgpu_device *adev, } } -/** - * amdgpu_vm_pt_first_leaf - get first leaf PD/PT - * - * @adev: amdgpu_device pointer - * @vm: amdgpu_vm structure - * @start: start addr of the walk - * @cursor: state to initialize - * - * Start a walk and go directly to the leaf node. - */ -static void amdgpu_vm_pt_first_leaf(struct amdgpu_device *adev, - struct amdgpu_vm *vm, uint64_t start, - struct amdgpu_vm_pt_cursor *cursor) -{ - amdgpu_vm_pt_start(adev, vm, start, cursor); - while (amdgpu_vm_pt_descendant(adev, cursor)); -} - -/** - * amdgpu_vm_pt_next_leaf - get next leaf PD/PT - * - * @adev: amdgpu_device pointer - * @cursor: current state - * - * Walk the PD/PT tree to the next leaf node. - */ -static void amdgpu_vm_pt_next_leaf(struct amdgpu_device *adev, - struct amdgpu_vm_pt_cursor *cursor) -{ - amdgpu_vm_pt_next(adev, cursor); - if (cursor->pfn != ~0ll) - while (amdgpu_vm_pt_descendant(adev, cursor)); -} - -/** - * for_each_amdgpu_vm_pt_leaf - walk over all leaf PDs/PTs in the hierarchy - */ -#define for_each_amdgpu_vm_pt_leaf(adev, vm, start, end, cursor) \ - for (amdgpu_vm_pt_first_leaf((adev), (vm), (start), &(cursor)); \ - (cursor).pfn <= end; amdgpu_vm_pt_next_leaf((adev), &(cursor))) - /** * amdgpu_vm_pt_first_dfs - start a deep first search * @@ -556,12 +488,31 @@ static void amdgpu_vm_pt_next_leaf(struct amdgpu_device *adev, */ static void amdgpu_vm_pt_first_dfs(struct amdgpu_device *adev, struct amdgpu_vm *vm, + struct amdgpu_vm_pt_cursor *start, struct amdgpu_vm_pt_cursor *cursor) { - amdgpu_vm_pt_start(adev, vm, 0, cursor); + if (start) + *cursor = *start; + else + amdgpu_vm_pt_start(adev, vm, 0, cursor); while (amdgpu_vm_pt_descendant(adev, cursor)); } +/** + * amdgpu_vm_pt_continue_dfs - check if the deep first search should continue + * + * @start: starting point for the search + * @entry: current entry + * + * Returns: + * True when the search should continue, false otherwise. + */ +static bool amdgpu_vm_pt_continue_dfs(struct amdgpu_vm_pt_cursor *start, + struct amdgpu_vm_pt *entry) +{ + return entry && (!start || entry != start->entry); +} + /** * amdgpu_vm_pt_next_dfs - get the next node for a deep first search * @@ -587,11 +538,11 @@ static void amdgpu_vm_pt_next_dfs(struct amdgpu_device *adev, /** * for_each_amdgpu_vm_pt_dfs_safe - safe deep first search of all PDs/PTs */ -#define for_each_amdgpu_vm_pt_dfs_safe(adev, vm, cursor, entry) \ - for (amdgpu_vm_pt_first_dfs((adev), (vm), &(cursor)), \ +#define for_each_amdgpu_vm_pt_dfs_safe(adev, vm, start, cursor, entry) \ + for (amdgpu_vm_pt_first_dfs((adev), (vm), (start), &(cursor)), \ (entry) = (cursor).entry, amdgpu_vm_pt_next_dfs((adev), &(cursor));\ - (entry); (entry) = (cursor).entry, \ - amdgpu_vm_pt_next_dfs((adev), &(cursor))) + amdgpu_vm_pt_continue_dfs((start), (entry)); \ + (entry) = (cursor).entry, amdgpu_vm_pt_next_dfs((adev), &(cursor))) /** * amdgpu_vm_get_pd_bo - add the VM PD to a validation list @@ -712,18 +663,11 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm, if (bo->tbo.type != ttm_bo_type_kernel) { amdgpu_vm_bo_moved(bo_base); } else { - if (vm->use_cpu_for_update) - r = amdgpu_bo_kmap(bo, NULL); + vm->update_funcs->map_table(bo); + if (bo->parent) + amdgpu_vm_bo_relocated(bo_base); else - r = amdgpu_ttm_alloc_gart(&bo->tbo); - if (r) - break; - if (bo->shadow) { - r = amdgpu_ttm_alloc_gart(&bo->shadow->tbo); - if (r) - break; - } - amdgpu_vm_bo_relocated(bo_base); + amdgpu_vm_bo_idle(bo_base); } } @@ -751,8 +695,6 @@ bool amdgpu_vm_ready(struct amdgpu_vm *vm) * @adev: amdgpu_device pointer * @vm: VM to clear BO from * @bo: BO to clear - * @level: level this BO is at - * @pte_support_ats: indicate ATS support from PTE * * Root PD needs to be reserved when calling this. * @@ -760,99 +702,112 @@ bool amdgpu_vm_ready(struct amdgpu_vm *vm) * 0 on success, errno otherwise. */ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, - struct amdgpu_vm *vm, struct amdgpu_bo *bo, - unsigned level, bool pte_support_ats) + struct amdgpu_vm *vm, + struct amdgpu_bo *bo) { struct ttm_operation_ctx ctx = { true, false }; - struct dma_fence *fence = NULL; + unsigned level = adev->vm_manager.root_level; + struct amdgpu_vm_update_params params; + struct amdgpu_bo *ancestor = bo; unsigned entries, ats_entries; - struct amdgpu_ring *ring; - struct amdgpu_job *job; uint64_t addr; int r; - entries = amdgpu_bo_size(bo) / 8; + /* Figure out our place in the hierarchy */ + if (ancestor->parent) { + ++level; + while (ancestor->parent->parent) { + ++level; + ancestor = ancestor->parent; + } + } - if (pte_support_ats) { - if (level == adev->vm_manager.root_level) { - ats_entries = amdgpu_vm_level_shift(adev, level); - ats_entries += AMDGPU_GPU_PAGE_SHIFT; - ats_entries = AMDGPU_GMC_HOLE_START >> ats_entries; - ats_entries = min(ats_entries, entries); - entries -= ats_entries; + entries = amdgpu_bo_size(bo) / 8; + if (!vm->pte_support_ats) { + ats_entries = 0; + + } else if (!bo->parent) { + ats_entries = amdgpu_vm_num_ats_entries(adev); + ats_entries = min(ats_entries, entries); + entries -= ats_entries; + + } else { + struct amdgpu_vm_pt *pt; + + pt = container_of(ancestor->vm_bo, struct amdgpu_vm_pt, base); + ats_entries = amdgpu_vm_num_ats_entries(adev); + if ((pt - vm->root.entries) >= ats_entries) { + ats_entries = 0; } else { ats_entries = entries; entries = 0; } - } else { - ats_entries = 0; } - ring = container_of(vm->entity.rq->sched, struct amdgpu_ring, sched); - r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); if (r) - goto error; + return r; - r = amdgpu_ttm_alloc_gart(&bo->tbo); + if (bo->shadow) { + r = ttm_bo_validate(&bo->shadow->tbo, &bo->shadow->placement, + &ctx); + if (r) + return r; + } + + r = vm->update_funcs->map_table(bo); if (r) return r; - r = amdgpu_job_alloc_with_ib(adev, 64, &job); + memset(¶ms, 0, sizeof(params)); + params.adev = adev; + params.vm = vm; + + r = vm->update_funcs->prepare(¶ms, AMDGPU_FENCE_OWNER_KFD, NULL); if (r) - goto error; + return r; - addr = amdgpu_bo_gpu_offset(bo); + addr = 0; if (ats_entries) { - uint64_t ats_value; + uint64_t value = 0, flags; - ats_value = AMDGPU_PTE_DEFAULT_ATC; - if (level != AMDGPU_VM_PTB) - ats_value |= AMDGPU_PDE_PTE; + flags = AMDGPU_PTE_DEFAULT_ATC; + if (level != AMDGPU_VM_PTB) { + /* Handle leaf PDEs as PTEs */ + flags |= AMDGPU_PDE_PTE; + amdgpu_gmc_get_vm_pde(adev, level, &value, &flags); + } + + r = vm->update_funcs->update(¶ms, bo, addr, 0, ats_entries, + value, flags); + if (r) + return r; - amdgpu_vm_set_pte_pde(adev, &job->ibs[0], addr, 0, - ats_entries, 0, ats_value); addr += ats_entries * 8; } if (entries) { - uint64_t value = 0; + uint64_t value = 0, flags = 0; - /* Workaround for fault priority problem on GMC9 */ - if (level == AMDGPU_VM_PTB && adev->asic_type >= CHIP_VEGA10) - value = AMDGPU_PTE_EXECUTABLE; + if (adev->asic_type >= CHIP_VEGA10) { + if (level != AMDGPU_VM_PTB) { + /* Handle leaf PDEs as PTEs */ + flags |= AMDGPU_PDE_PTE; + amdgpu_gmc_get_vm_pde(adev, level, + &value, &flags); + } else { + /* Workaround for fault priority problem on GMC9 */ + flags = AMDGPU_PTE_EXECUTABLE; + } + } - amdgpu_vm_set_pte_pde(adev, &job->ibs[0], addr, 0, - entries, 0, value); + r = vm->update_funcs->update(¶ms, bo, addr, 0, entries, + value, flags); + if (r) + return r; } - amdgpu_ring_pad_ib(ring, &job->ibs[0]); - - WARN_ON(job->ibs[0].length_dw > 64); - r = amdgpu_sync_resv(adev, &job->sync, bo->tbo.resv, - AMDGPU_FENCE_OWNER_KFD, false); - if (r) - goto error_free; - - r = amdgpu_job_submit(job, &vm->entity, AMDGPU_FENCE_OWNER_UNDEFINED, - &fence); - if (r) - goto error_free; - - amdgpu_bo_fence(bo, fence, true); - dma_fence_put(fence); - - if (bo->shadow) - return amdgpu_vm_clear_bo(adev, vm, bo->shadow, - level, pte_support_ats); - - return 0; - -error_free: - amdgpu_job_free(job); - -error: - return r; + return vm->update_funcs->commit(¶ms, NULL); } /** @@ -883,89 +838,56 @@ static void amdgpu_vm_bo_param(struct amdgpu_device *adev, struct amdgpu_vm *vm, } /** - * amdgpu_vm_alloc_pts - Allocate page tables. + * amdgpu_vm_alloc_pts - Allocate a specific page table * * @adev: amdgpu_device pointer * @vm: VM to allocate page tables for - * @saddr: Start address which needs to be allocated - * @size: Size from start address we need. + * @cursor: Which page table to allocate * - * Make sure the page directories and page tables are allocated + * Make sure a specific page table or directory is allocated. * * Returns: - * 0 on success, errno otherwise. + * 1 if page table needed to be allocated, 0 if page table was already + * allocated, negative errno if an error occurred. */ -int amdgpu_vm_alloc_pts(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - uint64_t saddr, uint64_t size) +static int amdgpu_vm_alloc_pts(struct amdgpu_device *adev, + struct amdgpu_vm *vm, + struct amdgpu_vm_pt_cursor *cursor) { - struct amdgpu_vm_pt_cursor cursor; + struct amdgpu_vm_pt *entry = cursor->entry; + struct amdgpu_bo_param bp; struct amdgpu_bo *pt; - bool ats = false; - uint64_t eaddr; int r; - /* validate the parameters */ - if (saddr & AMDGPU_GPU_PAGE_MASK || size & AMDGPU_GPU_PAGE_MASK) - return -EINVAL; + if (cursor->level < AMDGPU_VM_PTB && !entry->entries) { + unsigned num_entries; - eaddr = saddr + size - 1; - - if (vm->pte_support_ats) - ats = saddr < AMDGPU_GMC_HOLE_START; - - saddr /= AMDGPU_GPU_PAGE_SIZE; - eaddr /= AMDGPU_GPU_PAGE_SIZE; - - if (eaddr >= adev->vm_manager.max_pfn) { - dev_err(adev->dev, "va above limit (0x%08llX >= 0x%08llX)\n", - eaddr, adev->vm_manager.max_pfn); - return -EINVAL; + num_entries = amdgpu_vm_num_entries(adev, cursor->level); + entry->entries = kvmalloc_array(num_entries, + sizeof(*entry->entries), + GFP_KERNEL | __GFP_ZERO); + if (!entry->entries) + return -ENOMEM; } - for_each_amdgpu_vm_pt_leaf(adev, vm, saddr, eaddr, cursor) { - struct amdgpu_vm_pt *entry = cursor.entry; - struct amdgpu_bo_param bp; + if (entry->base.bo) + return 0; - if (cursor.level < AMDGPU_VM_PTB) { - unsigned num_entries; + amdgpu_vm_bo_param(adev, vm, cursor->level, &bp); - num_entries = amdgpu_vm_num_entries(adev, cursor.level); - entry->entries = kvmalloc_array(num_entries, - sizeof(*entry->entries), - GFP_KERNEL | - __GFP_ZERO); - if (!entry->entries) - return -ENOMEM; - } + r = amdgpu_bo_create(adev, &bp, &pt); + if (r) + return r; + /* Keep a reference to the root directory to avoid + * freeing them up in the wrong order. + */ + pt->parent = amdgpu_bo_ref(cursor->parent->base.bo); + amdgpu_vm_bo_base_init(&entry->base, vm, pt); - if (entry->base.bo) - continue; - - amdgpu_vm_bo_param(adev, vm, cursor.level, &bp); - - r = amdgpu_bo_create(adev, &bp, &pt); - if (r) - return r; - - if (vm->use_cpu_for_update) { - r = amdgpu_bo_kmap(pt, NULL); - if (r) - goto error_free_pt; - } - - /* Keep a reference to the root directory to avoid - * freeing them up in the wrong order. - */ - pt->parent = amdgpu_bo_ref(cursor.parent->base.bo); - - amdgpu_vm_bo_base_init(&entry->base, vm, pt); - - r = amdgpu_vm_clear_bo(adev, vm, pt, cursor.level, ats); - if (r) - goto error_free_pt; - } + r = amdgpu_vm_clear_bo(adev, vm, pt); + if (r) + goto error_free_pt; return 0; @@ -975,32 +897,46 @@ error_free_pt: return r; } +/** + * amdgpu_vm_free_table - fre one PD/PT + * + * @entry: PDE to free + */ +static void amdgpu_vm_free_table(struct amdgpu_vm_pt *entry) +{ + if (entry->base.bo) { + entry->base.bo->vm_bo = NULL; + list_del(&entry->base.vm_status); + amdgpu_bo_unref(&entry->base.bo->shadow); + amdgpu_bo_unref(&entry->base.bo); + } + kvfree(entry->entries); + entry->entries = NULL; +} + /** * amdgpu_vm_free_pts - free PD/PT levels * * @adev: amdgpu device structure * @vm: amdgpu vm structure + * @start: optional cursor where to start freeing PDs/PTs * * Free the page directory or page table level and all sub levels. */ static void amdgpu_vm_free_pts(struct amdgpu_device *adev, - struct amdgpu_vm *vm) + struct amdgpu_vm *vm, + struct amdgpu_vm_pt_cursor *start) { struct amdgpu_vm_pt_cursor cursor; struct amdgpu_vm_pt *entry; - for_each_amdgpu_vm_pt_dfs_safe(adev, vm, cursor, entry) { + vm->bulk_moveable = false; - if (entry->base.bo) { - entry->base.bo->vm_bo = NULL; - list_del(&entry->base.vm_status); - amdgpu_bo_unref(&entry->base.bo->shadow); - amdgpu_bo_unref(&entry->base.bo); - } - kvfree(entry->entries); - } + for_each_amdgpu_vm_pt_dfs_safe(adev, vm, start, cursor, entry) + amdgpu_vm_free_table(entry); - BUG_ON(vm->root.base.bo); + if (start) + amdgpu_vm_free_table(start->entry); } /** @@ -1211,66 +1147,6 @@ struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm, return NULL; } -/** - * amdgpu_vm_do_set_ptes - helper to call the right asic function - * - * @params: see amdgpu_pte_update_params definition - * @bo: PD/PT to update - * @pe: addr of the page entry - * @addr: dst addr to write into pe - * @count: number of page entries to update - * @incr: increase next addr by incr bytes - * @flags: hw access flags - * - * Traces the parameters and calls the right asic functions - * to setup the page table using the DMA. - */ -static void amdgpu_vm_do_set_ptes(struct amdgpu_pte_update_params *params, - struct amdgpu_bo *bo, - uint64_t pe, uint64_t addr, - unsigned count, uint32_t incr, - uint64_t flags) -{ - pe += amdgpu_bo_gpu_offset(bo); - trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags); - - if (count < 3) { - amdgpu_vm_write_pte(params->adev, params->ib, pe, - addr | flags, count, incr); - - } else { - amdgpu_vm_set_pte_pde(params->adev, params->ib, pe, addr, - count, incr, flags); - } -} - -/** - * amdgpu_vm_do_copy_ptes - copy the PTEs from the GART - * - * @params: see amdgpu_pte_update_params definition - * @bo: PD/PT to update - * @pe: addr of the page entry - * @addr: dst addr to write into pe - * @count: number of page entries to update - * @incr: increase next addr by incr bytes - * @flags: hw access flags - * - * Traces the parameters and calls the DMA function to copy the PTEs. - */ -static void amdgpu_vm_do_copy_ptes(struct amdgpu_pte_update_params *params, - struct amdgpu_bo *bo, - uint64_t pe, uint64_t addr, - unsigned count, uint32_t incr, - uint64_t flags) -{ - uint64_t src = (params->src + (addr >> 12) * 8); - - pe += amdgpu_bo_gpu_offset(bo); - trace_amdgpu_vm_copy_ptes(pe, src, count); - - amdgpu_vm_copy_pte(params->adev, params->ib, pe, src, count); -} - /** * amdgpu_vm_map_gart - Resolve gart mapping of addr * @@ -1283,7 +1159,7 @@ static void amdgpu_vm_do_copy_ptes(struct amdgpu_pte_update_params *params, * Returns: * The pointer for the page table entry. */ -static uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr) +uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr) { uint64_t result; @@ -1298,88 +1174,31 @@ static uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr) return result; } -/** - * amdgpu_vm_cpu_set_ptes - helper to update page tables via CPU - * - * @params: see amdgpu_pte_update_params definition - * @bo: PD/PT to update - * @pe: kmap addr of the page entry - * @addr: dst addr to write into pe - * @count: number of page entries to update - * @incr: increase next addr by incr bytes - * @flags: hw access flags - * - * Write count number of PT/PD entries directly. - */ -static void amdgpu_vm_cpu_set_ptes(struct amdgpu_pte_update_params *params, - struct amdgpu_bo *bo, - uint64_t pe, uint64_t addr, - unsigned count, uint32_t incr, - uint64_t flags) -{ - unsigned int i; - uint64_t value; - - pe += (unsigned long)amdgpu_bo_kptr(bo); - - trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags); - - for (i = 0; i < count; i++) { - value = params->pages_addr ? - amdgpu_vm_map_gart(params->pages_addr, addr) : - addr; - amdgpu_gmc_set_pte_pde(params->adev, (void *)(uintptr_t)pe, - i, value, flags); - addr += incr; - } -} - -/** - * amdgpu_vm_update_func - helper to call update function - * - * Calls the update function for both the given BO as well as its shadow. - */ -static void amdgpu_vm_update_func(struct amdgpu_pte_update_params *params, - struct amdgpu_bo *bo, - uint64_t pe, uint64_t addr, - unsigned count, uint32_t incr, - uint64_t flags) -{ - if (bo->shadow) - params->func(params, bo->shadow, pe, addr, count, incr, flags); - params->func(params, bo, pe, addr, count, incr, flags); -} - /* * amdgpu_vm_update_pde - update a single level in the hierarchy * * @param: parameters for the update * @vm: requested vm - * @parent: parent directory * @entry: entry to update * * Makes sure the requested entry in parent is up to date. */ -static void amdgpu_vm_update_pde(struct amdgpu_pte_update_params *params, - struct amdgpu_vm *vm, - struct amdgpu_vm_pt *parent, - struct amdgpu_vm_pt *entry) +static int amdgpu_vm_update_pde(struct amdgpu_vm_update_params *params, + struct amdgpu_vm *vm, + struct amdgpu_vm_pt *entry) { + struct amdgpu_vm_pt *parent = amdgpu_vm_pt_parent(entry); struct amdgpu_bo *bo = parent->base.bo, *pbo; uint64_t pde, pt, flags; unsigned level; - /* Don't update huge pages here */ - if (entry->huge) - return; - for (level = 0, pbo = bo->parent; pbo; ++level) pbo = pbo->parent; level += params->adev->vm_manager.root_level; amdgpu_gmc_get_pde_for_bo(entry->base.bo, level, &pt, &flags); pde = (entry - parent->entries) * 8; - amdgpu_vm_update_func(params, bo, pde, pt, 1, 0, flags); + return vm->update_funcs->update(params, bo, pde, pt, 1, 0, flags); } /* @@ -1396,7 +1215,7 @@ static void amdgpu_vm_invalidate_pds(struct amdgpu_device *adev, struct amdgpu_vm_pt_cursor cursor; struct amdgpu_vm_pt *entry; - for_each_amdgpu_vm_pt_dfs_safe(adev, vm, cursor, entry) + for_each_amdgpu_vm_pt_dfs_safe(adev, vm, NULL, cursor, entry) if (entry->base.bo && !entry->base.moved) amdgpu_vm_bo_relocated(&entry->base); } @@ -1415,89 +1234,39 @@ static void amdgpu_vm_invalidate_pds(struct amdgpu_device *adev, int amdgpu_vm_update_directories(struct amdgpu_device *adev, struct amdgpu_vm *vm) { - struct amdgpu_pte_update_params params; - struct amdgpu_job *job; - unsigned ndw = 0; - int r = 0; + struct amdgpu_vm_update_params params; + int r; if (list_empty(&vm->relocated)) return 0; -restart: memset(¶ms, 0, sizeof(params)); params.adev = adev; + params.vm = vm; - if (vm->use_cpu_for_update) { - r = amdgpu_bo_sync_wait(vm->root.base.bo, - AMDGPU_FENCE_OWNER_VM, true); - if (unlikely(r)) - return r; - - params.func = amdgpu_vm_cpu_set_ptes; - } else { - ndw = 512 * 8; - r = amdgpu_job_alloc_with_ib(adev, ndw * 4, &job); - if (r) - return r; - - params.ib = &job->ibs[0]; - params.func = amdgpu_vm_do_set_ptes; - } + r = vm->update_funcs->prepare(¶ms, AMDGPU_FENCE_OWNER_VM, NULL); + if (r) + return r; while (!list_empty(&vm->relocated)) { - struct amdgpu_vm_pt *pt, *entry; + struct amdgpu_vm_pt *entry; entry = list_first_entry(&vm->relocated, struct amdgpu_vm_pt, base.vm_status); amdgpu_vm_bo_idle(&entry->base); - pt = amdgpu_vm_pt_parent(entry); - if (!pt) - continue; - - amdgpu_vm_update_pde(¶ms, vm, pt, entry); - - if (!vm->use_cpu_for_update && - (ndw - params.ib->length_dw) < 32) - break; - } - - if (vm->use_cpu_for_update) { - /* Flush HDP */ - mb(); - amdgpu_asic_flush_hdp(adev, NULL); - } else if (params.ib->length_dw == 0) { - amdgpu_job_free(job); - } else { - struct amdgpu_bo *root = vm->root.base.bo; - struct amdgpu_ring *ring; - struct dma_fence *fence; - - ring = container_of(vm->entity.rq->sched, struct amdgpu_ring, - sched); - - amdgpu_ring_pad_ib(ring, params.ib); - amdgpu_sync_resv(adev, &job->sync, root->tbo.resv, - AMDGPU_FENCE_OWNER_VM, false); - WARN_ON(params.ib->length_dw > ndw); - r = amdgpu_job_submit(job, &vm->entity, AMDGPU_FENCE_OWNER_VM, - &fence); + r = amdgpu_vm_update_pde(¶ms, vm, entry); if (r) goto error; - - amdgpu_bo_fence(root, fence, true); - dma_fence_put(vm->last_update); - vm->last_update = fence; } - if (!list_empty(&vm->relocated)) - goto restart; - + r = vm->update_funcs->commit(¶ms, &vm->last_update); + if (r) + goto error; return 0; error: amdgpu_vm_invalidate_pds(adev, vm); - amdgpu_job_free(job); return r; } @@ -1506,7 +1275,7 @@ error: * * Make sure to set the right flags for the PTEs at the desired level. */ -static void amdgpu_vm_update_flags(struct amdgpu_pte_update_params *params, +static void amdgpu_vm_update_flags(struct amdgpu_vm_update_params *params, struct amdgpu_bo *bo, unsigned level, uint64_t pe, uint64_t addr, unsigned count, uint32_t incr, @@ -1525,13 +1294,14 @@ static void amdgpu_vm_update_flags(struct amdgpu_pte_update_params *params, flags |= AMDGPU_PTE_EXECUTABLE; } - amdgpu_vm_update_func(params, bo, pe, addr, count, incr, flags); + params->vm->update_funcs->update(params, bo, pe, addr, count, incr, + flags); } /** * amdgpu_vm_fragment - get fragment for PTEs * - * @params: see amdgpu_pte_update_params definition + * @params: see amdgpu_vm_update_params definition * @start: first PTE to handle * @end: last PTE to handle * @flags: hw mapping flags @@ -1540,7 +1310,7 @@ static void amdgpu_vm_update_flags(struct amdgpu_pte_update_params *params, * * Returns the first possible fragment for the start and end address. */ -static void amdgpu_vm_fragment(struct amdgpu_pte_update_params *params, +static void amdgpu_vm_fragment(struct amdgpu_vm_update_params *params, uint64_t start, uint64_t end, uint64_t flags, unsigned int *frag, uint64_t *frag_end) { @@ -1573,7 +1343,7 @@ static void amdgpu_vm_fragment(struct amdgpu_pte_update_params *params, max_frag = 31; /* system pages are non continuously */ - if (params->src) { + if (params->pages_addr) { *frag = 0; *frag_end = end; return; @@ -1592,7 +1362,7 @@ static void amdgpu_vm_fragment(struct amdgpu_pte_update_params *params, /** * amdgpu_vm_update_ptes - make sure that page tables are valid * - * @params: see amdgpu_pte_update_params definition + * @params: see amdgpu_vm_update_params definition * @start: start of GPU address range * @end: end of GPU address range * @dst: destination address to map to, the next dst inside the function @@ -1603,7 +1373,7 @@ static void amdgpu_vm_fragment(struct amdgpu_pte_update_params *params, * Returns: * 0 for success, -EINVAL for failure. */ -static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, +static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params, uint64_t start, uint64_t end, uint64_t dst, uint64_t flags) { @@ -1611,6 +1381,7 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, struct amdgpu_vm_pt_cursor cursor; uint64_t frag_start = start, frag_end; unsigned int frag; + int r; /* figure out the initial fragment */ amdgpu_vm_fragment(params, frag_start, end, flags, &frag, &frag_end); @@ -1618,12 +1389,15 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, /* walk over the address space and update the PTs */ amdgpu_vm_pt_start(adev, params->vm, start, &cursor); while (cursor.pfn < end) { - struct amdgpu_bo *pt = cursor.entry->base.bo; unsigned shift, parent_shift, mask; uint64_t incr, entry_end, pe_start; + struct amdgpu_bo *pt; - if (!pt) - return -ENOENT; + r = amdgpu_vm_alloc_pts(params->adev, params->vm, &cursor); + if (r) + return r; + + pt = cursor.entry->base.bo; /* The root level can't be a huge page */ if (cursor.level == adev->vm_manager.root_level) { @@ -1632,16 +1406,10 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, continue; } - /* If it isn't already handled it can't be a huge page */ - if (cursor.entry->huge) { - /* Add the entry to the relocated list to update it. */ - cursor.entry->huge = false; - amdgpu_vm_bo_relocated(&cursor.entry->base); - } - shift = amdgpu_vm_level_shift(adev, cursor.level); parent_shift = amdgpu_vm_level_shift(adev, cursor.level - 1); - if (adev->asic_type < CHIP_VEGA10) { + if (adev->asic_type < CHIP_VEGA10 && + (flags & AMDGPU_PTE_VALID)) { /* No huge page support before GMC v9 */ if (cursor.level != AMDGPU_VM_PTB) { if (!amdgpu_vm_pt_descendant(adev, &cursor)) @@ -1697,9 +1465,9 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, } while (frag_start < entry_end); if (amdgpu_vm_pt_descendant(adev, &cursor)) { - /* Mark all child entries as huge */ + /* Free all child entries */ while (cursor.pfn < frag_start) { - cursor.entry->huge = true; + amdgpu_vm_free_pts(adev, params->vm, &cursor); amdgpu_vm_pt_next(adev, &cursor); } @@ -1738,137 +1506,28 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, uint64_t flags, uint64_t addr, struct dma_fence **fence) { - struct amdgpu_ring *ring; + struct amdgpu_vm_update_params params; void *owner = AMDGPU_FENCE_OWNER_VM; - unsigned nptes, ncmds, ndw; - struct amdgpu_job *job; - struct amdgpu_pte_update_params params; - struct dma_fence *f = NULL; int r; memset(¶ms, 0, sizeof(params)); params.adev = adev; params.vm = vm; + params.pages_addr = pages_addr; /* sync to everything except eviction fences on unmapping */ if (!(flags & AMDGPU_PTE_VALID)) owner = AMDGPU_FENCE_OWNER_KFD; - if (vm->use_cpu_for_update) { - /* params.src is used as flag to indicate system Memory */ - if (pages_addr) - params.src = ~0; - - /* Wait for PT BOs to be idle. PTs share the same resv. object - * as the root PD BO - */ - r = amdgpu_bo_sync_wait(vm->root.base.bo, owner, true); - if (unlikely(r)) - return r; - - /* Wait for any BO move to be completed */ - if (exclusive) { - r = dma_fence_wait(exclusive, true); - if (unlikely(r)) - return r; - } - - params.func = amdgpu_vm_cpu_set_ptes; - params.pages_addr = pages_addr; - return amdgpu_vm_update_ptes(¶ms, start, last + 1, - addr, flags); - } - - ring = container_of(vm->entity.rq->sched, struct amdgpu_ring, sched); - - nptes = last - start + 1; - - /* - * reserve space for two commands every (1 << BLOCK_SIZE) - * entries or 2k dwords (whatever is smaller) - */ - ncmds = ((nptes >> min(adev->vm_manager.block_size, 11u)) + 1); - - /* The second command is for the shadow pagetables. */ - if (vm->root.base.bo->shadow) - ncmds *= 2; - - /* padding, etc. */ - ndw = 64; - - if (pages_addr) { - /* copy commands needed */ - ndw += ncmds * adev->vm_manager.vm_pte_funcs->copy_pte_num_dw; - - /* and also PTEs */ - ndw += nptes * 2; - - params.func = amdgpu_vm_do_copy_ptes; - - } else { - /* set page commands needed */ - ndw += ncmds * 10; - - /* extra commands for begin/end fragments */ - ncmds = 2 * adev->vm_manager.fragment_size; - if (vm->root.base.bo->shadow) - ncmds *= 2; - - ndw += 10 * ncmds; - - params.func = amdgpu_vm_do_set_ptes; - } - - r = amdgpu_job_alloc_with_ib(adev, ndw * 4, &job); + r = vm->update_funcs->prepare(¶ms, owner, exclusive); if (r) return r; - params.ib = &job->ibs[0]; - - if (pages_addr) { - uint64_t *pte; - unsigned i; - - /* Put the PTEs at the end of the IB. */ - i = ndw - nptes * 2; - pte= (uint64_t *)&(job->ibs->ptr[i]); - params.src = job->ibs->gpu_addr + i * 4; - - for (i = 0; i < nptes; ++i) { - pte[i] = amdgpu_vm_map_gart(pages_addr, addr + i * - AMDGPU_GPU_PAGE_SIZE); - pte[i] |= flags; - } - addr = 0; - } - - r = amdgpu_sync_fence(adev, &job->sync, exclusive, false); - if (r) - goto error_free; - - r = amdgpu_sync_resv(adev, &job->sync, vm->root.base.bo->tbo.resv, - owner, false); - if (r) - goto error_free; - r = amdgpu_vm_update_ptes(¶ms, start, last + 1, addr, flags); if (r) - goto error_free; + return r; - amdgpu_ring_pad_ib(ring, params.ib); - WARN_ON(params.ib->length_dw > ndw); - r = amdgpu_job_submit(job, &vm->entity, AMDGPU_FENCE_OWNER_VM, &f); - if (r) - goto error_free; - - amdgpu_bo_fence(vm->root.base.bo, f, true); - dma_fence_put(*fence); - *fence = f; - return 0; - -error_free: - amdgpu_job_free(job); - return r; + return vm->update_funcs->commit(¶ms, fence); } /** @@ -1880,6 +1539,7 @@ error_free: * @vm: requested vm * @mapping: mapped range and flags to use for the update * @flags: HW flags for the mapping + * @bo_adev: amdgpu_device pointer that bo actually been allocated * @nodes: array of drm_mm_nodes with the MC addresses * @fence: optional resulting fence * @@ -1895,6 +1555,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, struct amdgpu_vm *vm, struct amdgpu_bo_va_mapping *mapping, uint64_t flags, + struct amdgpu_device *bo_adev, struct drm_mm_node *nodes, struct dma_fence **fence) { @@ -1949,7 +1610,6 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, if (pages_addr) { uint64_t count; - max_entries = min(max_entries, 16ull * 1024ull); for (count = 1; count < max_entries / AMDGPU_GPU_PAGES_IN_CPU_PAGE; ++count) { @@ -1969,7 +1629,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, } } else if (flags & AMDGPU_PTE_VALID) { - addr += adev->vm_manager.vram_base_offset; + addr += bo_adev->vm_manager.vram_base_offset; addr += pfn << PAGE_SHIFT; } @@ -2016,6 +1676,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct drm_mm_node *nodes; struct dma_fence *exclusive, **last_update; uint64_t flags; + struct amdgpu_device *bo_adev = adev; int r; if (clear || !bo) { @@ -2034,10 +1695,12 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, exclusive = reservation_object_get_excl(bo->tbo.resv); } - if (bo) + if (bo) { flags = amdgpu_ttm_tt_pte_flags(adev, bo->tbo.ttm, mem); - else + bo_adev = amdgpu_ttm_adev(bo->tbo.bdev); + } else { flags = 0x0; + } if (clear || (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv)) last_update = &vm->last_update; @@ -2054,7 +1717,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, list_for_each_entry(mapping, &bo_va->invalids, list) { r = amdgpu_vm_bo_split_mapping(adev, exclusive, pages_addr, vm, - mapping, flags, nodes, + mapping, flags, bo_adev, nodes, last_update); if (r) return r; @@ -2374,6 +2037,16 @@ struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev, INIT_LIST_HEAD(&bo_va->valids); INIT_LIST_HEAD(&bo_va->invalids); + if (bo && amdgpu_xgmi_same_hive(adev, amdgpu_ttm_adev(bo->tbo.bdev)) && + (bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM)) { + bo_va->is_xgmi = true; + mutex_lock(&adev->vm_manager.lock_pstate); + /* Power up XGMI if it can be potentially used */ + if (++adev->vm_manager.xgmi_map_counter == 1) + amdgpu_xgmi_set_pstate(adev, 1); + mutex_unlock(&adev->vm_manager.lock_pstate); + } + return bo_va; } @@ -2792,6 +2465,14 @@ void amdgpu_vm_bo_rmv(struct amdgpu_device *adev, } dma_fence_put(bo_va->last_pt_update); + + if (bo && bo_va->is_xgmi) { + mutex_lock(&adev->vm_manager.lock_pstate); + if (--adev->vm_manager.xgmi_map_counter == 0) + amdgpu_xgmi_set_pstate(adev, 0); + mutex_unlock(&adev->vm_manager.lock_pstate); + } + kfree(bo_va); } @@ -2949,20 +2630,16 @@ void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t min_vm_size, adev->vm_manager.fragment_size); } -static struct amdgpu_retryfault_hashtable *init_fault_hash(void) +/** + * amdgpu_vm_wait_idle - wait for the VM to become idle + * + * @vm: VM object to wait for + * @timeout: timeout to wait for VM to become idle + */ +long amdgpu_vm_wait_idle(struct amdgpu_vm *vm, long timeout) { - struct amdgpu_retryfault_hashtable *fault_hash; - - fault_hash = kmalloc(sizeof(*fault_hash), GFP_KERNEL); - if (!fault_hash) - return fault_hash; - - INIT_CHASH_TABLE(fault_hash->hash, - AMDGPU_PAGEFAULT_HASH_BITS, 8, 0); - spin_lock_init(&fault_hash->lock); - fault_hash->count = 0; - - return fault_hash; + return reservation_object_wait_timeout_rcu(vm->root.base.bo->tbo.resv, + true, true, timeout); } /** @@ -3018,6 +2695,11 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, vm->use_cpu_for_update ? "CPU" : "SDMA"); WARN_ONCE((vm->use_cpu_for_update && !amdgpu_gmc_vram_full_visible(&adev->gmc)), "CPU update of VM recommended only for large BAR system\n"); + + if (vm->use_cpu_for_update) + vm->update_funcs = &amdgpu_vm_cpu_funcs; + else + vm->update_funcs = &amdgpu_vm_sdma_funcs; vm->last_update = NULL; amdgpu_vm_bo_param(adev, vm, adev->vm_manager.root_level, &bp); @@ -3037,9 +2719,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, amdgpu_vm_bo_base_init(&vm->root.base, vm, root); - r = amdgpu_vm_clear_bo(adev, vm, root, - adev->vm_manager.root_level, - vm->pte_support_ats); + r = amdgpu_vm_clear_bo(adev, vm, root); if (r) goto error_unreserve; @@ -3058,12 +2738,6 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, vm->pasid = pasid; } - vm->fault_hash = init_fault_hash(); - if (!vm->fault_hash) { - r = -ENOMEM; - goto error_free_root; - } - INIT_KFIFO(vm->faults); return 0; @@ -3134,9 +2808,8 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, uns * changing any other state, in case it fails. */ if (pte_support_ats != vm->pte_support_ats) { - r = amdgpu_vm_clear_bo(adev, vm, vm->root.base.bo, - adev->vm_manager.root_level, - pte_support_ats); + vm->pte_support_ats = pte_support_ats; + r = amdgpu_vm_clear_bo(adev, vm, vm->root.base.bo); if (r) goto free_idr; } @@ -3144,7 +2817,6 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, uns /* Update VM state */ vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode & AMDGPU_VM_USE_CPU_FOR_COMPUTE); - vm->pte_support_ats = pte_support_ats; DRM_DEBUG_DRIVER("VM update mode is %s\n", vm->use_cpu_for_update ? "CPU" : "SDMA"); WARN_ONCE((vm->use_cpu_for_update && !amdgpu_gmc_vram_full_visible(&adev->gmc)), @@ -3219,15 +2891,10 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) struct amdgpu_bo_va_mapping *mapping, *tmp; bool prt_fini_needed = !!adev->gmc.gmc_funcs->set_prt; struct amdgpu_bo *root; - u64 fault; int i, r; amdgpu_amdkfd_gpuvm_destroy_cb(adev, vm); - /* Clear pending page faults from IH when the VM is destroyed */ - while (kfifo_get(&vm->faults, &fault)) - amdgpu_vm_clear_fault(vm->fault_hash, fault); - if (vm->pasid) { unsigned long flags; @@ -3236,9 +2903,6 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags); } - kfree(vm->fault_hash); - vm->fault_hash = NULL; - drm_sched_entity_destroy(&vm->entity); if (!RB_EMPTY_ROOT(&vm->va.rb_root)) { @@ -3267,10 +2931,11 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) if (r) { dev_err(adev->dev, "Leaking page tables because BO reservation failed\n"); } else { - amdgpu_vm_free_pts(adev, vm); + amdgpu_vm_free_pts(adev, vm, NULL); amdgpu_bo_unreserve(root); } amdgpu_bo_unref(&root); + WARN_ON(vm->root.base.bo); dma_fence_put(vm->last_update); for (i = 0; i < AMDGPU_MAX_VMHUBS; i++) amdgpu_vmid_free_reserved(adev, vm, i); @@ -3315,6 +2980,9 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev) idr_init(&adev->vm_manager.pasid_idr); spin_lock_init(&adev->vm_manager.pasid_lock); + + adev->vm_manager.xgmi_map_counter = 0; + mutex_init(&adev->vm_manager.lock_pstate); } /** @@ -3405,78 +3073,3 @@ void amdgpu_vm_set_task_info(struct amdgpu_vm *vm) } } } - -/** - * amdgpu_vm_add_fault - Add a page fault record to fault hash table - * - * @fault_hash: fault hash table - * @key: 64-bit encoding of PASID and address - * - * This should be called when a retry page fault interrupt is - * received. If this is a new page fault, it will be added to a hash - * table. The return value indicates whether this is a new fault, or - * a fault that was already known and is already being handled. - * - * If there are too many pending page faults, this will fail. Retry - * interrupts should be ignored in this case until there is enough - * free space. - * - * Returns 0 if the fault was added, 1 if the fault was already known, - * -ENOSPC if there are too many pending faults. - */ -int amdgpu_vm_add_fault(struct amdgpu_retryfault_hashtable *fault_hash, u64 key) -{ - unsigned long flags; - int r = -ENOSPC; - - if (WARN_ON_ONCE(!fault_hash)) - /* Should be allocated in amdgpu_vm_init - */ - return r; - - spin_lock_irqsave(&fault_hash->lock, flags); - - /* Only let the hash table fill up to 50% for best performance */ - if (fault_hash->count >= (1 << (AMDGPU_PAGEFAULT_HASH_BITS-1))) - goto unlock_out; - - r = chash_table_copy_in(&fault_hash->hash, key, NULL); - if (!r) - fault_hash->count++; - - /* chash_table_copy_in should never fail unless we're losing count */ - WARN_ON_ONCE(r < 0); - -unlock_out: - spin_unlock_irqrestore(&fault_hash->lock, flags); - return r; -} - -/** - * amdgpu_vm_clear_fault - Remove a page fault record - * - * @fault_hash: fault hash table - * @key: 64-bit encoding of PASID and address - * - * This should be called when a page fault has been handled. Any - * future interrupt with this key will be processed as a new - * page fault. - */ -void amdgpu_vm_clear_fault(struct amdgpu_retryfault_hashtable *fault_hash, u64 key) -{ - unsigned long flags; - int r; - - if (!fault_hash) - return; - - spin_lock_irqsave(&fault_hash->lock, flags); - - r = chash_table_remove(&fault_hash->hash, key, NULL); - if (!WARN_ON_ONCE(r < 0)) { - fault_hash->count--; - WARN_ON_ONCE(fault_hash->count < 0); - } - - spin_unlock_irqrestore(&fault_hash->lock, flags); -} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 81ff8177f092..91baf95212a6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -30,7 +30,6 @@ #include #include #include -#include #include "amdgpu_sync.h" #include "amdgpu_ring.h" @@ -140,7 +139,6 @@ struct amdgpu_vm_bo_base { struct amdgpu_vm_pt { struct amdgpu_vm_bo_base base; - bool huge; /* array of page tables, one for each directory entry */ struct amdgpu_vm_pt *entries; @@ -167,11 +165,6 @@ struct amdgpu_vm_pte_funcs { uint32_t incr, uint64_t flags); }; -#define AMDGPU_VM_FAULT(pasid, addr) (((u64)(pasid) << 48) | (addr)) -#define AMDGPU_VM_FAULT_PASID(fault) ((u64)(fault) >> 48) -#define AMDGPU_VM_FAULT_ADDR(fault) ((u64)(fault) & 0xfffffffff000ULL) - - struct amdgpu_task_info { char process_name[TASK_COMM_LEN]; char task_name[TASK_COMM_LEN]; @@ -179,11 +172,52 @@ struct amdgpu_task_info { pid_t tgid; }; -#define AMDGPU_PAGEFAULT_HASH_BITS 8 -struct amdgpu_retryfault_hashtable { - DECLARE_CHASH_TABLE(hash, AMDGPU_PAGEFAULT_HASH_BITS, 8, 0); - spinlock_t lock; - int count; +/** + * struct amdgpu_vm_update_params + * + * Encapsulate some VM table update parameters to reduce + * the number of function parameters + * + */ +struct amdgpu_vm_update_params { + + /** + * @adev: amdgpu device we do this update for + */ + struct amdgpu_device *adev; + + /** + * @vm: optional amdgpu_vm we do this update for + */ + struct amdgpu_vm *vm; + + /** + * @pages_addr: + * + * DMA addresses to use for mapping + */ + dma_addr_t *pages_addr; + + /** + * @job: job to used for hw submission + */ + struct amdgpu_job *job; + + /** + * @num_dw_left: number of dw left for the IB + */ + unsigned int num_dw_left; +}; + +struct amdgpu_vm_update_funcs { + int (*map_table)(struct amdgpu_bo *bo); + int (*prepare)(struct amdgpu_vm_update_params *p, void * owner, + struct dma_fence *exclusive); + int (*update)(struct amdgpu_vm_update_params *p, + struct amdgpu_bo *bo, uint64_t pe, uint64_t addr, + unsigned count, uint32_t incr, uint64_t flags); + int (*commit)(struct amdgpu_vm_update_params *p, + struct dma_fence **fence); }; struct amdgpu_vm { @@ -221,7 +255,10 @@ struct amdgpu_vm { struct amdgpu_vmid *reserved_vmid[AMDGPU_MAX_VMHUBS]; /* Flag to indicate if VM tables are updated by CPU or GPU (SDMA) */ - bool use_cpu_for_update; + bool use_cpu_for_update; + + /* Functions to use for VM table updates */ + const struct amdgpu_vm_update_funcs *update_funcs; /* Flag to indicate ATS support from PTE for GFX9 */ bool pte_support_ats; @@ -245,7 +282,6 @@ struct amdgpu_vm { struct ttm_lru_bulk_move lru_bulk_move; /* mark whether can do the bulk move */ bool bulk_moveable; - struct amdgpu_retryfault_hashtable *fault_hash; }; struct amdgpu_vm_manager { @@ -267,6 +303,7 @@ struct amdgpu_vm_manager { const struct amdgpu_vm_pte_funcs *vm_pte_funcs; struct drm_sched_rq *vm_pte_rqs[AMDGPU_MAX_RINGS]; unsigned vm_pte_num_rqs; + struct amdgpu_ring *page_fault; /* partial resident texture handling */ spinlock_t prt_lock; @@ -283,14 +320,23 @@ struct amdgpu_vm_manager { */ struct idr pasid_idr; spinlock_t pasid_lock; + + /* counter of mapped memory through xgmi */ + uint32_t xgmi_map_counter; + struct mutex lock_pstate; }; #define amdgpu_vm_copy_pte(adev, ib, pe, src, count) ((adev)->vm_manager.vm_pte_funcs->copy_pte((ib), (pe), (src), (count))) #define amdgpu_vm_write_pte(adev, ib, pe, value, count, incr) ((adev)->vm_manager.vm_pte_funcs->write_pte((ib), (pe), (value), (count), (incr))) #define amdgpu_vm_set_pte_pde(adev, ib, pe, addr, count, incr, flags) ((adev)->vm_manager.vm_pte_funcs->set_pte_pde((ib), (pe), (addr), (count), (incr), (flags))) +extern const struct amdgpu_vm_update_funcs amdgpu_vm_cpu_funcs; +extern const struct amdgpu_vm_update_funcs amdgpu_vm_sdma_funcs; + void amdgpu_vm_manager_init(struct amdgpu_device *adev); void amdgpu_vm_manager_fini(struct amdgpu_device *adev); + +long amdgpu_vm_wait_idle(struct amdgpu_vm *vm, long timeout); int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, int vm_context, unsigned int pasid); int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, unsigned int pasid); @@ -303,9 +349,6 @@ bool amdgpu_vm_ready(struct amdgpu_vm *vm); int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm, int (*callback)(void *p, struct amdgpu_bo *bo), void *param); -int amdgpu_vm_alloc_pts(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - uint64_t saddr, uint64_t size); int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_pipe_sync); int amdgpu_vm_update_directories(struct amdgpu_device *adev, struct amdgpu_vm *vm); @@ -319,6 +362,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, bool clear); void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev, struct amdgpu_bo *bo, bool evicted); +uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr); struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm, struct amdgpu_bo *bo); struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev, @@ -358,11 +402,6 @@ void amdgpu_vm_set_task_info(struct amdgpu_vm *vm); void amdgpu_vm_move_to_lru_tail(struct amdgpu_device *adev, struct amdgpu_vm *vm); - -int amdgpu_vm_add_fault(struct amdgpu_retryfault_hashtable *fault_hash, u64 key); - -void amdgpu_vm_clear_fault(struct amdgpu_retryfault_hashtable *fault_hash, u64 key); - void amdgpu_vm_del_from_lru_notify(struct ttm_buffer_object *bo); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c new file mode 100644 index 000000000000..5222d165abfc --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c @@ -0,0 +1,127 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#include "amdgpu_vm.h" +#include "amdgpu_object.h" +#include "amdgpu_trace.h" + +/** + * amdgpu_vm_cpu_map_table - make sure new PDs/PTs are kmapped + * + * @table: newly allocated or validated PD/PT + */ +static int amdgpu_vm_cpu_map_table(struct amdgpu_bo *table) +{ + return amdgpu_bo_kmap(table, NULL); +} + +/** + * amdgpu_vm_cpu_prepare - prepare page table update with the CPU + * + * @p: see amdgpu_vm_update_params definition + * @owner: owner we need to sync to + * @exclusive: exclusive move fence we need to sync to + * + * Returns: + * Negativ errno, 0 for success. + */ +static int amdgpu_vm_cpu_prepare(struct amdgpu_vm_update_params *p, void *owner, + struct dma_fence *exclusive) +{ + int r; + + /* Wait for PT BOs to be idle. PTs share the same resv. object + * as the root PD BO + */ + r = amdgpu_bo_sync_wait(p->vm->root.base.bo, owner, true); + if (unlikely(r)) + return r; + + /* Wait for any BO move to be completed */ + if (exclusive) { + r = dma_fence_wait(exclusive, true); + if (unlikely(r)) + return r; + } + + return 0; +} + +/** + * amdgpu_vm_cpu_update - helper to update page tables via CPU + * + * @p: see amdgpu_vm_update_params definition + * @bo: PD/PT to update + * @pe: kmap addr of the page entry + * @addr: dst addr to write into pe + * @count: number of page entries to update + * @incr: increase next addr by incr bytes + * @flags: hw access flags + * + * Write count number of PT/PD entries directly. + */ +static int amdgpu_vm_cpu_update(struct amdgpu_vm_update_params *p, + struct amdgpu_bo *bo, uint64_t pe, + uint64_t addr, unsigned count, uint32_t incr, + uint64_t flags) +{ + unsigned int i; + uint64_t value; + + pe += (unsigned long)amdgpu_bo_kptr(bo); + + trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags); + + for (i = 0; i < count; i++) { + value = p->pages_addr ? + amdgpu_vm_map_gart(p->pages_addr, addr) : + addr; + amdgpu_gmc_set_pte_pde(p->adev, (void *)(uintptr_t)pe, + i, value, flags); + addr += incr; + } + return 0; +} + +/** + * amdgpu_vm_cpu_commit - commit page table update to the HW + * + * @p: see amdgpu_vm_update_params definition + * @fence: unused + * + * Make sure that the hardware sees the page table updates. + */ +static int amdgpu_vm_cpu_commit(struct amdgpu_vm_update_params *p, + struct dma_fence **fence) +{ + /* Flush HDP */ + mb(); + amdgpu_asic_flush_hdp(p->adev, NULL); + return 0; +} + +const struct amdgpu_vm_update_funcs amdgpu_vm_cpu_funcs = { + .map_table = amdgpu_vm_cpu_map_table, + .prepare = amdgpu_vm_cpu_prepare, + .update = amdgpu_vm_cpu_update, + .commit = amdgpu_vm_cpu_commit +}; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c new file mode 100644 index 000000000000..ddd181f5ed37 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c @@ -0,0 +1,270 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#include "amdgpu_vm.h" +#include "amdgpu_job.h" +#include "amdgpu_object.h" +#include "amdgpu_trace.h" + +#define AMDGPU_VM_SDMA_MIN_NUM_DW 256u +#define AMDGPU_VM_SDMA_MAX_NUM_DW (16u * 1024u) + +/** + * amdgpu_vm_sdma_map_table - make sure new PDs/PTs are GTT mapped + * + * @table: newly allocated or validated PD/PT + */ +static int amdgpu_vm_sdma_map_table(struct amdgpu_bo *table) +{ + int r; + + r = amdgpu_ttm_alloc_gart(&table->tbo); + if (r) + return r; + + if (table->shadow) + r = amdgpu_ttm_alloc_gart(&table->shadow->tbo); + + return r; +} + +/** + * amdgpu_vm_sdma_prepare - prepare SDMA command submission + * + * @p: see amdgpu_vm_update_params definition + * @owner: owner we need to sync to + * @exclusive: exclusive move fence we need to sync to + * + * Returns: + * Negativ errno, 0 for success. + */ +static int amdgpu_vm_sdma_prepare(struct amdgpu_vm_update_params *p, + void *owner, struct dma_fence *exclusive) +{ + struct amdgpu_bo *root = p->vm->root.base.bo; + unsigned int ndw = AMDGPU_VM_SDMA_MIN_NUM_DW; + int r; + + r = amdgpu_job_alloc_with_ib(p->adev, ndw * 4, &p->job); + if (r) + return r; + + r = amdgpu_sync_fence(p->adev, &p->job->sync, exclusive, false); + if (r) + return r; + + r = amdgpu_sync_resv(p->adev, &p->job->sync, root->tbo.resv, + owner, false); + if (r) + return r; + + p->num_dw_left = ndw; + return 0; +} + +/** + * amdgpu_vm_sdma_commit - commit SDMA command submission + * + * @p: see amdgpu_vm_update_params definition + * @fence: resulting fence + * + * Returns: + * Negativ errno, 0 for success. + */ +static int amdgpu_vm_sdma_commit(struct amdgpu_vm_update_params *p, + struct dma_fence **fence) +{ + struct amdgpu_bo *root = p->vm->root.base.bo; + struct amdgpu_ib *ib = p->job->ibs; + struct amdgpu_ring *ring; + struct dma_fence *f; + int r; + + ring = container_of(p->vm->entity.rq->sched, struct amdgpu_ring, sched); + + WARN_ON(ib->length_dw == 0); + amdgpu_ring_pad_ib(ring, ib); + WARN_ON(ib->length_dw > p->num_dw_left); + r = amdgpu_job_submit(p->job, &p->vm->entity, + AMDGPU_FENCE_OWNER_VM, &f); + if (r) + goto error; + + amdgpu_bo_fence(root, f, true); + if (fence) + swap(*fence, f); + dma_fence_put(f); + return 0; + +error: + amdgpu_job_free(p->job); + return r; +} + + +/** + * amdgpu_vm_sdma_copy_ptes - copy the PTEs from mapping + * + * @p: see amdgpu_vm_update_params definition + * @bo: PD/PT to update + * @pe: addr of the page entry + * @count: number of page entries to copy + * + * Traces the parameters and calls the DMA function to copy the PTEs. + */ +static void amdgpu_vm_sdma_copy_ptes(struct amdgpu_vm_update_params *p, + struct amdgpu_bo *bo, uint64_t pe, + unsigned count) +{ + struct amdgpu_ib *ib = p->job->ibs; + uint64_t src = ib->gpu_addr; + + src += p->num_dw_left * 4; + + pe += amdgpu_bo_gpu_offset(bo); + trace_amdgpu_vm_copy_ptes(pe, src, count); + + amdgpu_vm_copy_pte(p->adev, ib, pe, src, count); +} + +/** + * amdgpu_vm_sdma_set_ptes - helper to call the right asic function + * + * @p: see amdgpu_vm_update_params definition + * @bo: PD/PT to update + * @pe: addr of the page entry + * @addr: dst addr to write into pe + * @count: number of page entries to update + * @incr: increase next addr by incr bytes + * @flags: hw access flags + * + * Traces the parameters and calls the right asic functions + * to setup the page table using the DMA. + */ +static void amdgpu_vm_sdma_set_ptes(struct amdgpu_vm_update_params *p, + struct amdgpu_bo *bo, uint64_t pe, + uint64_t addr, unsigned count, + uint32_t incr, uint64_t flags) +{ + struct amdgpu_ib *ib = p->job->ibs; + + pe += amdgpu_bo_gpu_offset(bo); + trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags); + if (count < 3) { + amdgpu_vm_write_pte(p->adev, ib, pe, addr | flags, + count, incr); + } else { + amdgpu_vm_set_pte_pde(p->adev, ib, pe, addr, + count, incr, flags); + } +} + +/** + * amdgpu_vm_sdma_update - execute VM update + * + * @p: see amdgpu_vm_update_params definition + * @bo: PD/PT to update + * @pe: addr of the page entry + * @addr: dst addr to write into pe + * @count: number of page entries to update + * @incr: increase next addr by incr bytes + * @flags: hw access flags + * + * Reserve space in the IB, setup mapping buffer on demand and write commands to + * the IB. + */ +static int amdgpu_vm_sdma_update(struct amdgpu_vm_update_params *p, + struct amdgpu_bo *bo, uint64_t pe, + uint64_t addr, unsigned count, uint32_t incr, + uint64_t flags) +{ + unsigned int i, ndw, nptes; + uint64_t *pte; + int r; + + do { + ndw = p->num_dw_left; + ndw -= p->job->ibs->length_dw; + + if (ndw < 32) { + r = amdgpu_vm_sdma_commit(p, NULL); + if (r) + return r; + + /* estimate how many dw we need */ + ndw = 32; + if (p->pages_addr) + ndw += count * 2; + ndw = max(ndw, AMDGPU_VM_SDMA_MIN_NUM_DW); + ndw = min(ndw, AMDGPU_VM_SDMA_MAX_NUM_DW); + + r = amdgpu_job_alloc_with_ib(p->adev, ndw * 4, &p->job); + if (r) + return r; + + p->num_dw_left = ndw; + } + + if (!p->pages_addr) { + /* set page commands needed */ + if (bo->shadow) + amdgpu_vm_sdma_set_ptes(p, bo->shadow, pe, addr, + count, incr, flags); + amdgpu_vm_sdma_set_ptes(p, bo, pe, addr, count, + incr, flags); + return 0; + } + + /* copy commands needed */ + ndw -= p->adev->vm_manager.vm_pte_funcs->copy_pte_num_dw * + (bo->shadow ? 2 : 1); + + /* for padding */ + ndw -= 7; + + nptes = min(count, ndw / 2); + + /* Put the PTEs at the end of the IB. */ + p->num_dw_left -= nptes * 2; + pte = (uint64_t *)&(p->job->ibs->ptr[p->num_dw_left]); + for (i = 0; i < nptes; ++i, addr += incr) { + pte[i] = amdgpu_vm_map_gart(p->pages_addr, addr); + pte[i] |= flags; + } + + if (bo->shadow) + amdgpu_vm_sdma_copy_ptes(p, bo->shadow, pe, nptes); + amdgpu_vm_sdma_copy_ptes(p, bo, pe, nptes); + + pe += nptes * 8; + count -= nptes; + } while (count); + + return 0; +} + +const struct amdgpu_vm_update_funcs amdgpu_vm_sdma_funcs = { + .map_table = amdgpu_vm_sdma_map_table, + .prepare = amdgpu_vm_sdma_prepare, + .update = amdgpu_vm_sdma_update, + .commit = amdgpu_vm_sdma_commit +}; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c index 3f9d5d00c9b3..ec9ea3fdbb4a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c @@ -32,6 +32,85 @@ struct amdgpu_vram_mgr { atomic64_t vis_usage; }; +/** + * DOC: mem_info_vram_total + * + * The amdgpu driver provides a sysfs API for reporting current total VRAM + * available on the device + * The file mem_info_vram_total is used for this and returns the total + * amount of VRAM in bytes + */ +static ssize_t amdgpu_mem_info_vram_total_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + + return snprintf(buf, PAGE_SIZE, "%llu\n", adev->gmc.real_vram_size); +} + +/** + * DOC: mem_info_vis_vram_total + * + * The amdgpu driver provides a sysfs API for reporting current total + * visible VRAM available on the device + * The file mem_info_vis_vram_total is used for this and returns the total + * amount of visible VRAM in bytes + */ +static ssize_t amdgpu_mem_info_vis_vram_total_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + + return snprintf(buf, PAGE_SIZE, "%llu\n", adev->gmc.visible_vram_size); +} + +/** + * DOC: mem_info_vram_used + * + * The amdgpu driver provides a sysfs API for reporting current total VRAM + * available on the device + * The file mem_info_vram_used is used for this and returns the total + * amount of currently used VRAM in bytes + */ +static ssize_t amdgpu_mem_info_vram_used_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + + return snprintf(buf, PAGE_SIZE, "%llu\n", + amdgpu_vram_mgr_usage(&adev->mman.bdev.man[TTM_PL_VRAM])); +} + +/** + * DOC: mem_info_vis_vram_used + * + * The amdgpu driver provides a sysfs API for reporting current total of + * used visible VRAM + * The file mem_info_vis_vram_used is used for this and returns the total + * amount of currently used visible VRAM in bytes + */ +static ssize_t amdgpu_mem_info_vis_vram_used_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + + return snprintf(buf, PAGE_SIZE, "%llu\n", + amdgpu_vram_mgr_vis_usage(&adev->mman.bdev.man[TTM_PL_VRAM])); +} + +static DEVICE_ATTR(mem_info_vram_total, S_IRUGO, + amdgpu_mem_info_vram_total_show, NULL); +static DEVICE_ATTR(mem_info_vis_vram_total, S_IRUGO, + amdgpu_mem_info_vis_vram_total_show,NULL); +static DEVICE_ATTR(mem_info_vram_used, S_IRUGO, + amdgpu_mem_info_vram_used_show, NULL); +static DEVICE_ATTR(mem_info_vis_vram_used, S_IRUGO, + amdgpu_mem_info_vis_vram_used_show, NULL); + /** * amdgpu_vram_mgr_init - init VRAM manager and DRM MM * @@ -43,7 +122,9 @@ struct amdgpu_vram_mgr { static int amdgpu_vram_mgr_init(struct ttm_mem_type_manager *man, unsigned long p_size) { + struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev); struct amdgpu_vram_mgr *mgr; + int ret; mgr = kzalloc(sizeof(*mgr), GFP_KERNEL); if (!mgr) @@ -52,6 +133,29 @@ static int amdgpu_vram_mgr_init(struct ttm_mem_type_manager *man, drm_mm_init(&mgr->mm, 0, p_size); spin_lock_init(&mgr->lock); man->priv = mgr; + + /* Add the two VRAM-related sysfs files */ + ret = device_create_file(adev->dev, &dev_attr_mem_info_vram_total); + if (ret) { + DRM_ERROR("Failed to create device file mem_info_vram_total\n"); + return ret; + } + ret = device_create_file(adev->dev, &dev_attr_mem_info_vis_vram_total); + if (ret) { + DRM_ERROR("Failed to create device file mem_info_vis_vram_total\n"); + return ret; + } + ret = device_create_file(adev->dev, &dev_attr_mem_info_vram_used); + if (ret) { + DRM_ERROR("Failed to create device file mem_info_vram_used\n"); + return ret; + } + ret = device_create_file(adev->dev, &dev_attr_mem_info_vis_vram_used); + if (ret) { + DRM_ERROR("Failed to create device file mem_info_vis_vram_used\n"); + return ret; + } + return 0; } @@ -65,6 +169,7 @@ static int amdgpu_vram_mgr_init(struct ttm_mem_type_manager *man, */ static int amdgpu_vram_mgr_fini(struct ttm_mem_type_manager *man) { + struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev); struct amdgpu_vram_mgr *mgr = man->priv; spin_lock(&mgr->lock); @@ -72,6 +177,10 @@ static int amdgpu_vram_mgr_fini(struct ttm_mem_type_manager *man) spin_unlock(&mgr->lock); kfree(mgr); man->priv = NULL; + device_remove_file(adev->dev, &dev_attr_mem_info_vram_total); + device_remove_file(adev->dev, &dev_attr_mem_info_vis_vram_total); + device_remove_file(adev->dev, &dev_attr_mem_info_vram_used); + device_remove_file(adev->dev, &dev_attr_mem_info_vis_vram_used); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c index 407dd16cc35c..a48c84c51775 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c @@ -24,6 +24,7 @@ #include #include "amdgpu.h" #include "amdgpu_xgmi.h" +#include "amdgpu_smu.h" static DEFINE_MUTEX(xgmi_mutex); @@ -34,12 +35,132 @@ static DEFINE_MUTEX(xgmi_mutex); static struct amdgpu_hive_info xgmi_hives[AMDGPU_MAX_XGMI_HIVE]; static unsigned hive_count = 0; - void *amdgpu_xgmi_hive_try_lock(struct amdgpu_hive_info *hive) { return &hive->device_list; } +static ssize_t amdgpu_xgmi_show_hive_id(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct amdgpu_hive_info *hive = + container_of(attr, struct amdgpu_hive_info, dev_attr); + + return snprintf(buf, PAGE_SIZE, "%llu\n", hive->hive_id); +} + +static int amdgpu_xgmi_sysfs_create(struct amdgpu_device *adev, + struct amdgpu_hive_info *hive) +{ + int ret = 0; + + if (WARN_ON(hive->kobj)) + return -EINVAL; + + hive->kobj = kobject_create_and_add("xgmi_hive_info", &adev->dev->kobj); + if (!hive->kobj) { + dev_err(adev->dev, "XGMI: Failed to allocate sysfs entry!\n"); + return -EINVAL; + } + + hive->dev_attr = (struct device_attribute) { + .attr = { + .name = "xgmi_hive_id", + .mode = S_IRUGO, + + }, + .show = amdgpu_xgmi_show_hive_id, + }; + + ret = sysfs_create_file(hive->kobj, &hive->dev_attr.attr); + if (ret) { + dev_err(adev->dev, "XGMI: Failed to create device file xgmi_hive_id\n"); + kobject_del(hive->kobj); + kobject_put(hive->kobj); + hive->kobj = NULL; + } + + return ret; +} + +static void amdgpu_xgmi_sysfs_destroy(struct amdgpu_device *adev, + struct amdgpu_hive_info *hive) +{ + sysfs_remove_file(hive->kobj, &hive->dev_attr.attr); + kobject_del(hive->kobj); + kobject_put(hive->kobj); + hive->kobj = NULL; +} + +static ssize_t amdgpu_xgmi_show_device_id(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + + return snprintf(buf, PAGE_SIZE, "%llu\n", adev->gmc.xgmi.node_id); + +} + + +static DEVICE_ATTR(xgmi_device_id, S_IRUGO, amdgpu_xgmi_show_device_id, NULL); + + +static int amdgpu_xgmi_sysfs_add_dev_info(struct amdgpu_device *adev, + struct amdgpu_hive_info *hive) +{ + int ret = 0; + char node[10] = { 0 }; + + /* Create xgmi device id file */ + ret = device_create_file(adev->dev, &dev_attr_xgmi_device_id); + if (ret) { + dev_err(adev->dev, "XGMI: Failed to create device file xgmi_device_id\n"); + return ret; + } + + /* Create sysfs link to hive info folder on the first device */ + if (adev != hive->adev) { + ret = sysfs_create_link(&adev->dev->kobj, hive->kobj, + "xgmi_hive_info"); + if (ret) { + dev_err(adev->dev, "XGMI: Failed to create link to hive info"); + goto remove_file; + } + } + + sprintf(node, "node%d", hive->number_devices); + /* Create sysfs link form the hive folder to yourself */ + ret = sysfs_create_link(hive->kobj, &adev->dev->kobj, node); + if (ret) { + dev_err(adev->dev, "XGMI: Failed to create link from hive info"); + goto remove_link; + } + + goto success; + + +remove_link: + sysfs_remove_link(&adev->dev->kobj, adev->ddev->unique); + +remove_file: + device_remove_file(adev->dev, &dev_attr_xgmi_device_id); + +success: + return ret; +} + +static void amdgpu_xgmi_sysfs_rem_dev_info(struct amdgpu_device *adev, + struct amdgpu_hive_info *hive) +{ + device_remove_file(adev->dev, &dev_attr_xgmi_device_id); + sysfs_remove_link(&adev->dev->kobj, adev->ddev->unique); + sysfs_remove_link(hive->kobj, adev->ddev->unique); +} + + + struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev, int lock) { int i; @@ -66,18 +187,50 @@ struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev, int lo /* initialize new hive if not exist */ tmp = &xgmi_hives[hive_count++]; + + if (amdgpu_xgmi_sysfs_create(adev, tmp)) { + mutex_unlock(&xgmi_mutex); + return NULL; + } + + tmp->adev = adev; tmp->hive_id = adev->gmc.xgmi.hive_id; INIT_LIST_HEAD(&tmp->device_list); mutex_init(&tmp->hive_lock); mutex_init(&tmp->reset_lock); + if (lock) mutex_lock(&tmp->hive_lock); - + tmp->pstate = -1; mutex_unlock(&xgmi_mutex); return tmp; } +int amdgpu_xgmi_set_pstate(struct amdgpu_device *adev, int pstate) +{ + int ret = 0; + struct amdgpu_hive_info *hive = amdgpu_get_xgmi_hive(adev, 0); + + if (!hive) + return 0; + + if (hive->pstate == pstate) + return 0; + + dev_dbg(adev->dev, "Set xgmi pstate %d.\n", pstate); + + if (is_support_sw_smu(adev)) + ret = smu_set_xgmi_pstate(&adev->smu, pstate); + if (ret) + dev_err(adev->dev, + "XGMI: Set pstate failure on device %llx, hive %llx, ret %d", + adev->gmc.xgmi.node_id, + adev->gmc.xgmi.hive_id, ret); + + return ret; +} + int amdgpu_xgmi_update_topology(struct amdgpu_hive_info *hive, struct amdgpu_device *adev) { int ret = -EINVAL; @@ -156,8 +309,17 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev) break; } - dev_info(adev->dev, "XGMI: Add node %d, hive 0x%llx.\n", - adev->gmc.xgmi.physical_node_id, adev->gmc.xgmi.hive_id); + if (!ret) + ret = amdgpu_xgmi_sysfs_add_dev_info(adev, hive); + + if (!ret) + dev_info(adev->dev, "XGMI: Add node %d, hive 0x%llx.\n", + adev->gmc.xgmi.physical_node_id, adev->gmc.xgmi.hive_id); + else + dev_err(adev->dev, "XGMI: Failed to add node %d, hive 0x%llx ret: %d\n", + adev->gmc.xgmi.physical_node_id, adev->gmc.xgmi.hive_id, + ret); + mutex_unlock(&hive->hive_lock); exit: @@ -176,9 +338,11 @@ void amdgpu_xgmi_remove_device(struct amdgpu_device *adev) return; if (!(hive->number_devices--)) { + amdgpu_xgmi_sysfs_destroy(adev, hive); mutex_destroy(&hive->hive_lock); mutex_destroy(&hive->reset_lock); } else { + amdgpu_xgmi_sysfs_rem_dev_info(adev, hive); mutex_unlock(&hive->hive_lock); } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h index 14bc60664159..3e9c91e9a4bf 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h @@ -29,13 +29,25 @@ struct amdgpu_hive_info { struct list_head device_list; struct psp_xgmi_topology_info topology_info; int number_devices; - struct mutex hive_lock, - reset_lock; + struct mutex hive_lock, reset_lock; + struct kobject *kobj; + struct device_attribute dev_attr; + struct amdgpu_device *adev; + int pstate; /*0 -- low , 1 -- high , -1 unknown*/ }; struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev, int lock); int amdgpu_xgmi_update_topology(struct amdgpu_hive_info *hive, struct amdgpu_device *adev); int amdgpu_xgmi_add_device(struct amdgpu_device *adev); void amdgpu_xgmi_remove_device(struct amdgpu_device *adev); +int amdgpu_xgmi_set_pstate(struct amdgpu_device *adev, int pstate); + +static inline bool amdgpu_xgmi_same_hive(struct amdgpu_device *adev, + struct amdgpu_device *bo_adev) +{ + return (adev != bo_adev && + adev->gmc.xgmi.hive_id && + adev->gmc.xgmi.hive_id == bo_adev->gmc.xgmi.hive_id); +} #endif diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c index 189599b694e8..d42808b05971 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c @@ -977,8 +977,8 @@ static int cik_sdma_sw_init(void *handle) r = amdgpu_ring_init(adev, ring, 1024, &adev->sdma.trap_irq, (i == 0) ? - AMDGPU_SDMA_IRQ_TRAP0 : - AMDGPU_SDMA_IRQ_TRAP1); + AMDGPU_SDMA_IRQ_INSTANCE0 : + AMDGPU_SDMA_IRQ_INSTANCE1); if (r) return r; } @@ -1114,7 +1114,7 @@ static int cik_sdma_set_trap_irq_state(struct amdgpu_device *adev, u32 sdma_cntl; switch (type) { - case AMDGPU_SDMA_IRQ_TRAP0: + case AMDGPU_SDMA_IRQ_INSTANCE0: switch (state) { case AMDGPU_IRQ_STATE_DISABLE: sdma_cntl = RREG32(mmSDMA0_CNTL + SDMA0_REGISTER_OFFSET); @@ -1130,7 +1130,7 @@ static int cik_sdma_set_trap_irq_state(struct amdgpu_device *adev, break; } break; - case AMDGPU_SDMA_IRQ_TRAP1: + case AMDGPU_SDMA_IRQ_INSTANCE1: switch (state) { case AMDGPU_IRQ_STATE_DISABLE: sdma_cntl = RREG32(mmSDMA0_CNTL + SDMA1_REGISTER_OFFSET); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c index 305276c7e4bf..c0cb244f58cd 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c @@ -782,6 +782,25 @@ static void gfx_v6_0_tiling_mode_table_init(struct amdgpu_device *adev) BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2); + tilemode[18] = MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + ARRAY_MODE(ARRAY_1D_TILED_THICK) | + PIPE_CONFIG(ADDR_SURF_P4_8x16); + tilemode[19] = MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + ARRAY_MODE(ARRAY_2D_TILED_XTHICK) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) | + NUM_BANKS(ADDR_SURF_16_BANK) | + TILE_SPLIT(split_equal_to_row_size); + tilemode[20] = MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + ARRAY_MODE(ARRAY_2D_TILED_THICK) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) | + NUM_BANKS(ADDR_SURF_16_BANK) | + TILE_SPLIT(split_equal_to_row_size); tilemode[21] = MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | ARRAY_MODE(ARRAY_2D_TILED_THIN1) | PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index b8e50a34bdb3..02955e6e9dd9 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -3236,6 +3236,7 @@ static void gfx_v8_0_tiling_mode_table_init(struct amdgpu_device *adev) dev_warn(adev->dev, "Unknown chip type (%d) in function gfx_v8_0_tiling_mode_table_init() falling through to CHIP_CARRIZO\n", adev->asic_type); + /* fall through */ case CHIP_CARRIZO: modearray[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index d0309e8c9d12..ba67d1023264 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -40,6 +40,8 @@ #include "ivsrcid/gfx/irqsrcs_gfx_9_0.h" +#include "amdgpu_ras.h" + #define GFX9_NUM_GFX_RINGS 1 #define GFX9_MEC_HPD_SIZE 4096 #define RLCG_UCODE_LOADING_START_ADDRESS 0x00002000L @@ -576,6 +578,27 @@ static void gfx_v9_0_check_fw_write_wait(struct amdgpu_device *adev) } } +static void gfx_v9_0_check_if_need_gfxoff(struct amdgpu_device *adev) +{ + switch (adev->asic_type) { + case CHIP_VEGA10: + case CHIP_VEGA12: + case CHIP_VEGA20: + break; + case CHIP_RAVEN: + if (adev->rev_id >= 0x8 || adev->pdev->device == 0x15d8) + break; + if ((adev->gfx.rlc_fw_version < 531) || + (adev->gfx.rlc_fw_version == 53815) || + (adev->gfx.rlc_feature_version < 1) || + !adev->gfx.rlc.is_rlc_v2_1) + adev->pm.pp_feature &= ~PP_GFXOFF_MASK; + break; + default: + break; + } +} + static int gfx_v9_0_init_microcode(struct amdgpu_device *adev) { const char *chip_name; @@ -828,6 +851,7 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev) } out: + gfx_v9_0_check_if_need_gfxoff(adev); gfx_v9_0_check_fw_write_wait(adev); if (err) { dev_err(adev->dev, @@ -1639,6 +1663,18 @@ static int gfx_v9_0_sw_init(void *handle) if (r) return r; + /* ECC error */ + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_ECC_ERROR, + &adev->gfx.cp_ecc_error_irq); + if (r) + return r; + + /* FUE error */ + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_FUE_ERROR, + &adev->gfx.cp_ecc_error_irq); + if (r) + return r; + adev->gfx.gfx_current_status = AMDGPU_GFX_NORMAL_MODE; gfx_v9_0_scratch_init(adev); @@ -1731,6 +1767,20 @@ static int gfx_v9_0_sw_fini(void *handle) int i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX) && + adev->gfx.ras_if) { + struct ras_common_if *ras_if = adev->gfx.ras_if; + struct ras_ih_if ih_info = { + .head = *ras_if, + }; + + amdgpu_ras_debugfs_remove(adev, ras_if); + amdgpu_ras_sysfs_remove(adev, ras_if); + amdgpu_ras_interrupt_remove_handler(adev, &ih_info); + amdgpu_ras_feature_enable(adev, ras_if, 0); + kfree(ras_if); + } + amdgpu_bo_free_kernel(&adev->gds.oa_gfx_bo, NULL, NULL); amdgpu_bo_free_kernel(&adev->gds.gws_gfx_bo, NULL, NULL); amdgpu_bo_free_kernel(&adev->gds.gds_gfx_bo, NULL, NULL); @@ -2405,8 +2455,6 @@ static int gfx_v9_0_rlc_resume(struct amdgpu_device *adev) /* disable CG */ WREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL, 0); - adev->gfx.rlc.funcs->reset(adev); - gfx_v9_0_init_pg(adev); if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) { @@ -3305,6 +3353,7 @@ static int gfx_v9_0_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + amdgpu_irq_put(adev, &adev->gfx.cp_ecc_error_irq, 0); amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0); amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0); @@ -3494,6 +3543,80 @@ static int gfx_v9_0_early_init(void *handle) return 0; } +static int gfx_v9_0_process_ras_data_cb(struct amdgpu_device *adev, + struct amdgpu_iv_entry *entry); + +static int gfx_v9_0_ecc_late_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct ras_common_if **ras_if = &adev->gfx.ras_if; + struct ras_ih_if ih_info = { + .cb = gfx_v9_0_process_ras_data_cb, + }; + struct ras_fs_if fs_info = { + .sysfs_name = "gfx_err_count", + .debugfs_name = "gfx_err_inject", + }; + struct ras_common_if ras_block = { + .block = AMDGPU_RAS_BLOCK__GFX, + .type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE, + .sub_block_index = 0, + .name = "gfx", + }; + int r; + + if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX)) { + amdgpu_ras_feature_enable_on_boot(adev, &ras_block, 0); + return 0; + } + + if (*ras_if) + goto resume; + + *ras_if = kmalloc(sizeof(**ras_if), GFP_KERNEL); + if (!*ras_if) + return -ENOMEM; + + **ras_if = ras_block; + + r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1); + if (r) + goto feature; + + ih_info.head = **ras_if; + fs_info.head = **ras_if; + + r = amdgpu_ras_interrupt_add_handler(adev, &ih_info); + if (r) + goto interrupt; + + r = amdgpu_ras_debugfs_create(adev, &fs_info); + if (r) + goto debugfs; + + r = amdgpu_ras_sysfs_create(adev, &fs_info); + if (r) + goto sysfs; +resume: + r = amdgpu_irq_get(adev, &adev->gfx.cp_ecc_error_irq, 0); + if (r) + goto irq; + + return 0; +irq: + amdgpu_ras_sysfs_remove(adev, *ras_if); +sysfs: + amdgpu_ras_debugfs_remove(adev, *ras_if); +debugfs: + amdgpu_ras_interrupt_remove_handler(adev, &ih_info); +interrupt: + amdgpu_ras_feature_enable(adev, *ras_if, 0); +feature: + kfree(*ras_if); + *ras_if = NULL; + return -EINVAL; +} + static int gfx_v9_0_late_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -3507,6 +3630,10 @@ static int gfx_v9_0_late_init(void *handle) if (r) return r; + r = gfx_v9_0_ecc_late_init(handle); + if (r) + return r; + return 0; } @@ -4543,6 +4670,45 @@ static int gfx_v9_0_set_priv_inst_fault_state(struct amdgpu_device *adev, return 0; } +#define ENABLE_ECC_ON_ME_PIPE(me, pipe) \ + WREG32_FIELD15(GC, 0, CP_ME##me##_PIPE##pipe##_INT_CNTL,\ + CP_ECC_ERROR_INT_ENABLE, 1) + +#define DISABLE_ECC_ON_ME_PIPE(me, pipe) \ + WREG32_FIELD15(GC, 0, CP_ME##me##_PIPE##pipe##_INT_CNTL,\ + CP_ECC_ERROR_INT_ENABLE, 0) + +static int gfx_v9_0_set_cp_ecc_error_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + unsigned type, + enum amdgpu_interrupt_state state) +{ + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + WREG32_FIELD15(GC, 0, CP_INT_CNTL_RING0, + CP_ECC_ERROR_INT_ENABLE, 0); + DISABLE_ECC_ON_ME_PIPE(1, 0); + DISABLE_ECC_ON_ME_PIPE(1, 1); + DISABLE_ECC_ON_ME_PIPE(1, 2); + DISABLE_ECC_ON_ME_PIPE(1, 3); + break; + + case AMDGPU_IRQ_STATE_ENABLE: + WREG32_FIELD15(GC, 0, CP_INT_CNTL_RING0, + CP_ECC_ERROR_INT_ENABLE, 1); + ENABLE_ECC_ON_ME_PIPE(1, 0); + ENABLE_ECC_ON_ME_PIPE(1, 1); + ENABLE_ECC_ON_ME_PIPE(1, 2); + ENABLE_ECC_ON_ME_PIPE(1, 3); + break; + default: + break; + } + + return 0; +} + + static int gfx_v9_0_set_eop_interrupt_state(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned type, @@ -4659,6 +4825,34 @@ static int gfx_v9_0_priv_inst_irq(struct amdgpu_device *adev, return 0; } +static int gfx_v9_0_process_ras_data_cb(struct amdgpu_device *adev, + struct amdgpu_iv_entry *entry) +{ + /* TODO ue will trigger an interrupt. */ + kgd2kfd_set_sram_ecc_flag(adev->kfd.dev); + amdgpu_ras_reset_gpu(adev, 0); + return AMDGPU_RAS_UE; +} + +static int gfx_v9_0_cp_ecc_error_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + struct ras_common_if *ras_if = adev->gfx.ras_if; + struct ras_dispatch_if ih_data = { + .entry = entry, + }; + + if (!ras_if) + return 0; + + ih_data.head = *ras_if; + + DRM_ERROR("CP ECC ERROR IRQ\n"); + amdgpu_ras_interrupt_dispatch(adev, &ih_data); + return 0; +} + static const struct amd_ip_funcs gfx_v9_0_ip_funcs = { .name = "gfx_v9_0", .early_init = gfx_v9_0_early_init, @@ -4820,6 +5014,12 @@ static const struct amdgpu_irq_src_funcs gfx_v9_0_priv_inst_irq_funcs = { .process = gfx_v9_0_priv_inst_irq, }; +static const struct amdgpu_irq_src_funcs gfx_v9_0_cp_ecc_error_irq_funcs = { + .set = gfx_v9_0_set_cp_ecc_error_state, + .process = gfx_v9_0_cp_ecc_error_irq, +}; + + static void gfx_v9_0_set_irq_funcs(struct amdgpu_device *adev) { adev->gfx.eop_irq.num_types = AMDGPU_CP_IRQ_LAST; @@ -4830,6 +5030,9 @@ static void gfx_v9_0_set_irq_funcs(struct amdgpu_device *adev) adev->gfx.priv_inst_irq.num_types = 1; adev->gfx.priv_inst_irq.funcs = &gfx_v9_0_priv_inst_irq_funcs; + + adev->gfx.cp_ecc_error_irq.num_types = 2; /*C5 ECC error and C9 FUE error*/ + adev->gfx.cp_ecc_error_irq.funcs = &gfx_v9_0_cp_ecc_error_irq_funcs; } static void gfx_v9_0_set_rlc_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c index f5edddf3b29d..7bb5359d0bbd 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c @@ -143,7 +143,7 @@ static void gfxhub_v1_0_init_cache_regs(struct amdgpu_device *adev) /* XXX for emulation, Refer to closed source code.*/ tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, L2_PDE0_CACHE_TAG_GENERATION_MODE, 0); - tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, PDE_FAULT_CLASSIFICATION, 1); + tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, PDE_FAULT_CLASSIFICATION, 0); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, CONTEXT1_IDENTITY_ACCESS_MODE, 1); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, IDENTITY_MODE_FRAGMENT_SIZE, 0); WREG32_SOC15(GC, 0, mmVM_L2_CNTL, tmp); @@ -236,7 +236,7 @@ static void gfxhub_v1_0_setup_vmid_config(struct amdgpu_device *adev) block_size); /* Send no-retry XNACK on fault to suppress VM fault storm. */ tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, - RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0); + RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 1); WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_CNTL, i, tmp); WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32, i*2, 0); WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32, i*2, 0); diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c index 98fd9208877f..b06d876da2d9 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -225,7 +225,7 @@ static void gmc_v6_0_vram_gtt_location(struct amdgpu_device *adev, u64 base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF; base <<= 24; - amdgpu_gmc_vram_location(adev, &adev->gmc, base); + amdgpu_gmc_vram_location(adev, mc, base); amdgpu_gmc_gart_location(adev, mc); } @@ -383,20 +383,6 @@ static uint64_t gmc_v6_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring, return pd_addr; } -static int gmc_v6_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr, - uint32_t gpu_page_idx, uint64_t addr, - uint64_t flags) -{ - void __iomem *ptr = (void *)cpu_pt_addr; - uint64_t value; - - value = addr & 0xFFFFFFFFFFFFF000ULL; - value |= flags; - writeq(value, ptr + (gpu_page_idx * 8)); - - return 0; -} - static uint64_t gmc_v6_0_get_vm_pte_flags(struct amdgpu_device *adev, uint32_t flags) { @@ -1169,7 +1155,6 @@ static const struct amd_ip_funcs gmc_v6_0_ip_funcs = { static const struct amdgpu_gmc_funcs gmc_v6_0_gmc_funcs = { .flush_gpu_tlb = gmc_v6_0_flush_gpu_tlb, .emit_flush_gpu_tlb = gmc_v6_0_emit_flush_gpu_tlb, - .set_pte_pde = gmc_v6_0_set_pte_pde, .set_prt = gmc_v6_0_set_prt, .get_vm_pde = gmc_v6_0_get_vm_pde, .get_vm_pte_flags = gmc_v6_0_get_vm_pte_flags diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 3e9c5034febe..75aa3332aee2 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -242,7 +242,7 @@ static void gmc_v7_0_vram_gtt_location(struct amdgpu_device *adev, u64 base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF; base <<= 24; - amdgpu_gmc_vram_location(adev, &adev->gmc, base); + amdgpu_gmc_vram_location(adev, mc, base); amdgpu_gmc_gart_location(adev, mc); } @@ -460,31 +460,6 @@ static void gmc_v7_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid, amdgpu_ring_emit_wreg(ring, mmIH_VMID_0_LUT + vmid, pasid); } -/** - * gmc_v7_0_set_pte_pde - update the page tables using MMIO - * - * @adev: amdgpu_device pointer - * @cpu_pt_addr: cpu address of the page table - * @gpu_page_idx: entry in the page table to update - * @addr: dst addr to write into pte/pde - * @flags: access flags - * - * Update the page tables using the CPU. - */ -static int gmc_v7_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr, - uint32_t gpu_page_idx, uint64_t addr, - uint64_t flags) -{ - void __iomem *ptr = (void *)cpu_pt_addr; - uint64_t value; - - value = addr & 0xFFFFFFFFFFFFF000ULL; - value |= flags; - writeq(value, ptr + (gpu_page_idx * 8)); - - return 0; -} - static uint64_t gmc_v7_0_get_vm_pte_flags(struct amdgpu_device *adev, uint32_t flags) { @@ -1376,7 +1351,6 @@ static const struct amdgpu_gmc_funcs gmc_v7_0_gmc_funcs = { .flush_gpu_tlb = gmc_v7_0_flush_gpu_tlb, .emit_flush_gpu_tlb = gmc_v7_0_emit_flush_gpu_tlb, .emit_pasid_mapping = gmc_v7_0_emit_pasid_mapping, - .set_pte_pde = gmc_v7_0_set_pte_pde, .set_prt = gmc_v7_0_set_prt, .get_vm_pte_flags = gmc_v7_0_get_vm_pte_flags, .get_vm_pde = gmc_v7_0_get_vm_pde diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index 29dde64bf2e7..8a3b5e6fc6c9 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -433,7 +433,7 @@ static void gmc_v8_0_vram_gtt_location(struct amdgpu_device *adev, base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF; base <<= 24; - amdgpu_gmc_vram_location(adev, &adev->gmc, base); + amdgpu_gmc_vram_location(adev, mc, base); amdgpu_gmc_gart_location(adev, mc); } @@ -662,50 +662,26 @@ static void gmc_v8_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid, amdgpu_ring_emit_wreg(ring, mmIH_VMID_0_LUT + vmid, pasid); } -/** - * gmc_v8_0_set_pte_pde - update the page tables using MMIO +/* + * PTE format on VI: + * 63:40 reserved + * 39:12 4k physical page base address + * 11:7 fragment + * 6 write + * 5 read + * 4 exe + * 3 reserved + * 2 snooped + * 1 system + * 0 valid * - * @adev: amdgpu_device pointer - * @cpu_pt_addr: cpu address of the page table - * @gpu_page_idx: entry in the page table to update - * @addr: dst addr to write into pte/pde - * @flags: access flags - * - * Update the page tables using the CPU. + * PDE format on VI: + * 63:59 block fragment size + * 58:40 reserved + * 39:1 physical base address of PTE + * bits 5:1 must be 0. + * 0 valid */ -static int gmc_v8_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr, - uint32_t gpu_page_idx, uint64_t addr, - uint64_t flags) -{ - void __iomem *ptr = (void *)cpu_pt_addr; - uint64_t value; - - /* - * PTE format on VI: - * 63:40 reserved - * 39:12 4k physical page base address - * 11:7 fragment - * 6 write - * 5 read - * 4 exe - * 3 reserved - * 2 snooped - * 1 system - * 0 valid - * - * PDE format on VI: - * 63:59 block fragment size - * 58:40 reserved - * 39:1 physical base address of PTE - * bits 5:1 must be 0. - * 0 valid - */ - value = addr & 0x000000FFFFFFF000ULL; - value |= flags; - writeq(value, ptr + (gpu_page_idx * 8)); - - return 0; -} static uint64_t gmc_v8_0_get_vm_pte_flags(struct amdgpu_device *adev, uint32_t flags) @@ -1743,7 +1719,6 @@ static const struct amdgpu_gmc_funcs gmc_v8_0_gmc_funcs = { .flush_gpu_tlb = gmc_v8_0_flush_gpu_tlb, .emit_flush_gpu_tlb = gmc_v8_0_emit_flush_gpu_tlb, .emit_pasid_mapping = gmc_v8_0_emit_pasid_mapping, - .set_pte_pde = gmc_v8_0_set_pte_pde, .set_prt = gmc_v8_0_set_prt, .get_vm_pte_flags = gmc_v8_0_get_vm_pte_flags, .get_vm_pde = gmc_v8_0_get_vm_pde diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 53327498efbf..3fd79e07944d 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -47,6 +47,8 @@ #include "ivsrcid/vmc/irqsrcs_vmc_1_0.h" +#include "amdgpu_ras.h" + /* add these here since we already include dce12 headers and these are for DCN */ #define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION 0x055d #define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION_BASE_IDX 2 @@ -84,121 +86,182 @@ static const struct soc15_reg_golden golden_settings_athub_1_0_0[] = SOC15_REG_GOLDEN_VALUE(ATHUB, 0, mmRPB_ARB_CNTL2, 0x00ff00ff, 0x00080008) }; -/* Ecc related register addresses, (BASE + reg offset) */ -/* Universal Memory Controller caps (may be fused). */ -/* UMCCH:UmcLocalCap */ -#define UMCLOCALCAPS_ADDR0 (0x00014306 + 0x00000000) -#define UMCLOCALCAPS_ADDR1 (0x00014306 + 0x00000800) -#define UMCLOCALCAPS_ADDR2 (0x00014306 + 0x00001000) -#define UMCLOCALCAPS_ADDR3 (0x00014306 + 0x00001800) -#define UMCLOCALCAPS_ADDR4 (0x00054306 + 0x00000000) -#define UMCLOCALCAPS_ADDR5 (0x00054306 + 0x00000800) -#define UMCLOCALCAPS_ADDR6 (0x00054306 + 0x00001000) -#define UMCLOCALCAPS_ADDR7 (0x00054306 + 0x00001800) -#define UMCLOCALCAPS_ADDR8 (0x00094306 + 0x00000000) -#define UMCLOCALCAPS_ADDR9 (0x00094306 + 0x00000800) -#define UMCLOCALCAPS_ADDR10 (0x00094306 + 0x00001000) -#define UMCLOCALCAPS_ADDR11 (0x00094306 + 0x00001800) -#define UMCLOCALCAPS_ADDR12 (0x000d4306 + 0x00000000) -#define UMCLOCALCAPS_ADDR13 (0x000d4306 + 0x00000800) -#define UMCLOCALCAPS_ADDR14 (0x000d4306 + 0x00001000) -#define UMCLOCALCAPS_ADDR15 (0x000d4306 + 0x00001800) - -/* Universal Memory Controller Channel config. */ -/* UMCCH:UMC_CONFIG */ -#define UMCCH_UMC_CONFIG_ADDR0 (0x00014040 + 0x00000000) -#define UMCCH_UMC_CONFIG_ADDR1 (0x00014040 + 0x00000800) -#define UMCCH_UMC_CONFIG_ADDR2 (0x00014040 + 0x00001000) -#define UMCCH_UMC_CONFIG_ADDR3 (0x00014040 + 0x00001800) -#define UMCCH_UMC_CONFIG_ADDR4 (0x00054040 + 0x00000000) -#define UMCCH_UMC_CONFIG_ADDR5 (0x00054040 + 0x00000800) -#define UMCCH_UMC_CONFIG_ADDR6 (0x00054040 + 0x00001000) -#define UMCCH_UMC_CONFIG_ADDR7 (0x00054040 + 0x00001800) -#define UMCCH_UMC_CONFIG_ADDR8 (0x00094040 + 0x00000000) -#define UMCCH_UMC_CONFIG_ADDR9 (0x00094040 + 0x00000800) -#define UMCCH_UMC_CONFIG_ADDR10 (0x00094040 + 0x00001000) -#define UMCCH_UMC_CONFIG_ADDR11 (0x00094040 + 0x00001800) -#define UMCCH_UMC_CONFIG_ADDR12 (0x000d4040 + 0x00000000) -#define UMCCH_UMC_CONFIG_ADDR13 (0x000d4040 + 0x00000800) -#define UMCCH_UMC_CONFIG_ADDR14 (0x000d4040 + 0x00001000) -#define UMCCH_UMC_CONFIG_ADDR15 (0x000d4040 + 0x00001800) - -/* Universal Memory Controller Channel Ecc config. */ -/* UMCCH:EccCtrl */ -#define UMCCH_ECCCTRL_ADDR0 (0x00014053 + 0x00000000) -#define UMCCH_ECCCTRL_ADDR1 (0x00014053 + 0x00000800) -#define UMCCH_ECCCTRL_ADDR2 (0x00014053 + 0x00001000) -#define UMCCH_ECCCTRL_ADDR3 (0x00014053 + 0x00001800) -#define UMCCH_ECCCTRL_ADDR4 (0x00054053 + 0x00000000) -#define UMCCH_ECCCTRL_ADDR5 (0x00054053 + 0x00000800) -#define UMCCH_ECCCTRL_ADDR6 (0x00054053 + 0x00001000) -#define UMCCH_ECCCTRL_ADDR7 (0x00054053 + 0x00001800) -#define UMCCH_ECCCTRL_ADDR8 (0x00094053 + 0x00000000) -#define UMCCH_ECCCTRL_ADDR9 (0x00094053 + 0x00000800) -#define UMCCH_ECCCTRL_ADDR10 (0x00094053 + 0x00001000) -#define UMCCH_ECCCTRL_ADDR11 (0x00094053 + 0x00001800) -#define UMCCH_ECCCTRL_ADDR12 (0x000d4053 + 0x00000000) -#define UMCCH_ECCCTRL_ADDR13 (0x000d4053 + 0x00000800) -#define UMCCH_ECCCTRL_ADDR14 (0x000d4053 + 0x00001000) -#define UMCCH_ECCCTRL_ADDR15 (0x000d4053 + 0x00001800) - -static const uint32_t ecc_umclocalcap_addrs[] = { - UMCLOCALCAPS_ADDR0, - UMCLOCALCAPS_ADDR1, - UMCLOCALCAPS_ADDR2, - UMCLOCALCAPS_ADDR3, - UMCLOCALCAPS_ADDR4, - UMCLOCALCAPS_ADDR5, - UMCLOCALCAPS_ADDR6, - UMCLOCALCAPS_ADDR7, - UMCLOCALCAPS_ADDR8, - UMCLOCALCAPS_ADDR9, - UMCLOCALCAPS_ADDR10, - UMCLOCALCAPS_ADDR11, - UMCLOCALCAPS_ADDR12, - UMCLOCALCAPS_ADDR13, - UMCLOCALCAPS_ADDR14, - UMCLOCALCAPS_ADDR15, +static const uint32_t ecc_umc_mcumc_ctrl_addrs[] = { + (0x000143c0 + 0x00000000), + (0x000143c0 + 0x00000800), + (0x000143c0 + 0x00001000), + (0x000143c0 + 0x00001800), + (0x000543c0 + 0x00000000), + (0x000543c0 + 0x00000800), + (0x000543c0 + 0x00001000), + (0x000543c0 + 0x00001800), + (0x000943c0 + 0x00000000), + (0x000943c0 + 0x00000800), + (0x000943c0 + 0x00001000), + (0x000943c0 + 0x00001800), + (0x000d43c0 + 0x00000000), + (0x000d43c0 + 0x00000800), + (0x000d43c0 + 0x00001000), + (0x000d43c0 + 0x00001800), + (0x001143c0 + 0x00000000), + (0x001143c0 + 0x00000800), + (0x001143c0 + 0x00001000), + (0x001143c0 + 0x00001800), + (0x001543c0 + 0x00000000), + (0x001543c0 + 0x00000800), + (0x001543c0 + 0x00001000), + (0x001543c0 + 0x00001800), + (0x001943c0 + 0x00000000), + (0x001943c0 + 0x00000800), + (0x001943c0 + 0x00001000), + (0x001943c0 + 0x00001800), + (0x001d43c0 + 0x00000000), + (0x001d43c0 + 0x00000800), + (0x001d43c0 + 0x00001000), + (0x001d43c0 + 0x00001800), }; -static const uint32_t ecc_umcch_umc_config_addrs[] = { - UMCCH_UMC_CONFIG_ADDR0, - UMCCH_UMC_CONFIG_ADDR1, - UMCCH_UMC_CONFIG_ADDR2, - UMCCH_UMC_CONFIG_ADDR3, - UMCCH_UMC_CONFIG_ADDR4, - UMCCH_UMC_CONFIG_ADDR5, - UMCCH_UMC_CONFIG_ADDR6, - UMCCH_UMC_CONFIG_ADDR7, - UMCCH_UMC_CONFIG_ADDR8, - UMCCH_UMC_CONFIG_ADDR9, - UMCCH_UMC_CONFIG_ADDR10, - UMCCH_UMC_CONFIG_ADDR11, - UMCCH_UMC_CONFIG_ADDR12, - UMCCH_UMC_CONFIG_ADDR13, - UMCCH_UMC_CONFIG_ADDR14, - UMCCH_UMC_CONFIG_ADDR15, +static const uint32_t ecc_umc_mcumc_ctrl_mask_addrs[] = { + (0x000143e0 + 0x00000000), + (0x000143e0 + 0x00000800), + (0x000143e0 + 0x00001000), + (0x000143e0 + 0x00001800), + (0x000543e0 + 0x00000000), + (0x000543e0 + 0x00000800), + (0x000543e0 + 0x00001000), + (0x000543e0 + 0x00001800), + (0x000943e0 + 0x00000000), + (0x000943e0 + 0x00000800), + (0x000943e0 + 0x00001000), + (0x000943e0 + 0x00001800), + (0x000d43e0 + 0x00000000), + (0x000d43e0 + 0x00000800), + (0x000d43e0 + 0x00001000), + (0x000d43e0 + 0x00001800), + (0x001143e0 + 0x00000000), + (0x001143e0 + 0x00000800), + (0x001143e0 + 0x00001000), + (0x001143e0 + 0x00001800), + (0x001543e0 + 0x00000000), + (0x001543e0 + 0x00000800), + (0x001543e0 + 0x00001000), + (0x001543e0 + 0x00001800), + (0x001943e0 + 0x00000000), + (0x001943e0 + 0x00000800), + (0x001943e0 + 0x00001000), + (0x001943e0 + 0x00001800), + (0x001d43e0 + 0x00000000), + (0x001d43e0 + 0x00000800), + (0x001d43e0 + 0x00001000), + (0x001d43e0 + 0x00001800), }; -static const uint32_t ecc_umcch_eccctrl_addrs[] = { - UMCCH_ECCCTRL_ADDR0, - UMCCH_ECCCTRL_ADDR1, - UMCCH_ECCCTRL_ADDR2, - UMCCH_ECCCTRL_ADDR3, - UMCCH_ECCCTRL_ADDR4, - UMCCH_ECCCTRL_ADDR5, - UMCCH_ECCCTRL_ADDR6, - UMCCH_ECCCTRL_ADDR7, - UMCCH_ECCCTRL_ADDR8, - UMCCH_ECCCTRL_ADDR9, - UMCCH_ECCCTRL_ADDR10, - UMCCH_ECCCTRL_ADDR11, - UMCCH_ECCCTRL_ADDR12, - UMCCH_ECCCTRL_ADDR13, - UMCCH_ECCCTRL_ADDR14, - UMCCH_ECCCTRL_ADDR15, +static const uint32_t ecc_umc_mcumc_status_addrs[] = { + (0x000143c2 + 0x00000000), + (0x000143c2 + 0x00000800), + (0x000143c2 + 0x00001000), + (0x000143c2 + 0x00001800), + (0x000543c2 + 0x00000000), + (0x000543c2 + 0x00000800), + (0x000543c2 + 0x00001000), + (0x000543c2 + 0x00001800), + (0x000943c2 + 0x00000000), + (0x000943c2 + 0x00000800), + (0x000943c2 + 0x00001000), + (0x000943c2 + 0x00001800), + (0x000d43c2 + 0x00000000), + (0x000d43c2 + 0x00000800), + (0x000d43c2 + 0x00001000), + (0x000d43c2 + 0x00001800), + (0x001143c2 + 0x00000000), + (0x001143c2 + 0x00000800), + (0x001143c2 + 0x00001000), + (0x001143c2 + 0x00001800), + (0x001543c2 + 0x00000000), + (0x001543c2 + 0x00000800), + (0x001543c2 + 0x00001000), + (0x001543c2 + 0x00001800), + (0x001943c2 + 0x00000000), + (0x001943c2 + 0x00000800), + (0x001943c2 + 0x00001000), + (0x001943c2 + 0x00001800), + (0x001d43c2 + 0x00000000), + (0x001d43c2 + 0x00000800), + (0x001d43c2 + 0x00001000), + (0x001d43c2 + 0x00001800), }; +static int gmc_v9_0_ecc_interrupt_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 bits, i, tmp, reg; + + bits = 0x7f; + + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + for (i = 0; i < ARRAY_SIZE(ecc_umc_mcumc_ctrl_addrs); i++) { + reg = ecc_umc_mcumc_ctrl_addrs[i]; + tmp = RREG32(reg); + tmp &= ~bits; + WREG32(reg, tmp); + } + for (i = 0; i < ARRAY_SIZE(ecc_umc_mcumc_ctrl_mask_addrs); i++) { + reg = ecc_umc_mcumc_ctrl_mask_addrs[i]; + tmp = RREG32(reg); + tmp &= ~bits; + WREG32(reg, tmp); + } + break; + case AMDGPU_IRQ_STATE_ENABLE: + for (i = 0; i < ARRAY_SIZE(ecc_umc_mcumc_ctrl_addrs); i++) { + reg = ecc_umc_mcumc_ctrl_addrs[i]; + tmp = RREG32(reg); + tmp |= bits; + WREG32(reg, tmp); + } + for (i = 0; i < ARRAY_SIZE(ecc_umc_mcumc_ctrl_mask_addrs); i++) { + reg = ecc_umc_mcumc_ctrl_mask_addrs[i]; + tmp = RREG32(reg); + tmp |= bits; + WREG32(reg, tmp); + } + break; + default: + break; + } + + return 0; +} + +static int gmc_v9_0_process_ras_data_cb(struct amdgpu_device *adev, + struct amdgpu_iv_entry *entry) +{ + kgd2kfd_set_sram_ecc_flag(adev->kfd.dev); + amdgpu_ras_reset_gpu(adev, 0); + return AMDGPU_RAS_UE; +} + +static int gmc_v9_0_process_ecc_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + struct ras_common_if *ras_if = adev->gmc.ras_if; + struct ras_dispatch_if ih_data = { + .entry = entry, + }; + + if (!ras_if) + return 0; + + ih_data.head = *ras_if; + + amdgpu_ras_interrupt_dispatch(adev, &ih_data); + return 0; +} + static int gmc_v9_0_vm_fault_interrupt_state(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned type, @@ -244,62 +307,6 @@ static int gmc_v9_0_vm_fault_interrupt_state(struct amdgpu_device *adev, return 0; } -/** - * vega10_ih_prescreen_iv - prescreen an interrupt vector - * - * @adev: amdgpu_device pointer - * - * Returns true if the interrupt vector should be further processed. - */ -static bool gmc_v9_0_prescreen_iv(struct amdgpu_device *adev, - struct amdgpu_iv_entry *entry, - uint64_t addr) -{ - struct amdgpu_vm *vm; - u64 key; - int r; - - /* No PASID, can't identify faulting process */ - if (!entry->pasid) - return true; - - /* Not a retry fault */ - if (!(entry->src_data[1] & 0x80)) - return true; - - /* Track retry faults in per-VM fault FIFO. */ - spin_lock(&adev->vm_manager.pasid_lock); - vm = idr_find(&adev->vm_manager.pasid_idr, entry->pasid); - if (!vm) { - /* VM not found, process it normally */ - spin_unlock(&adev->vm_manager.pasid_lock); - return true; - } - - key = AMDGPU_VM_FAULT(entry->pasid, addr); - r = amdgpu_vm_add_fault(vm->fault_hash, key); - - /* Hash table is full or the fault is already being processed, - * ignore further page faults - */ - if (r != 0) { - spin_unlock(&adev->vm_manager.pasid_lock); - return false; - } - /* No locking required with single writer and single reader */ - r = kfifo_put(&vm->faults, key); - if (!r) { - /* FIFO is full. Ignore it until there is space */ - amdgpu_vm_clear_fault(vm->fault_hash, key); - spin_unlock(&adev->vm_manager.pasid_lock); - return false; - } - - spin_unlock(&adev->vm_manager.pasid_lock); - /* It's the first fault for this address, process it normally */ - return true; -} - static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) @@ -312,9 +319,11 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev, addr = (u64)entry->src_data[0] << 12; addr |= ((u64)entry->src_data[1] & 0xf) << 44; - if (!gmc_v9_0_prescreen_iv(adev, entry, addr)) + if (retry_fault && amdgpu_gmc_filter_faults(adev, addr, entry->pasid, + entry->timestamp)) return 1; /* This also prevents sending it to KFD */ + /* If it's the first fault for this address, process it normally */ if (!amdgpu_sriov_vf(adev)) { status = RREG32(hub->vm_l2_pro_fault_status); WREG32_P(hub->vm_l2_pro_fault_cntl, 1, ~1); @@ -350,10 +359,19 @@ static const struct amdgpu_irq_src_funcs gmc_v9_0_irq_funcs = { .process = gmc_v9_0_process_interrupt, }; + +static const struct amdgpu_irq_src_funcs gmc_v9_0_ecc_funcs = { + .set = gmc_v9_0_ecc_interrupt_state, + .process = gmc_v9_0_process_ecc_irq, +}; + static void gmc_v9_0_set_irq_funcs(struct amdgpu_device *adev) { adev->gmc.vm_fault.num_types = 1; adev->gmc.vm_fault.funcs = &gmc_v9_0_irq_funcs; + + adev->gmc.ecc_irq.num_types = 1; + adev->gmc.ecc_irq.funcs = &gmc_v9_0_ecc_funcs; } static uint32_t gmc_v9_0_get_invalidate_req(unsigned int vmid, @@ -466,64 +484,37 @@ static void gmc_v9_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid, amdgpu_ring_emit_wreg(ring, reg, pasid); } -/** - * gmc_v9_0_set_pte_pde - update the page tables using MMIO +/* + * PTE format on VEGA 10: + * 63:59 reserved + * 58:57 mtype + * 56 F + * 55 L + * 54 P + * 53 SW + * 52 T + * 50:48 reserved + * 47:12 4k physical page base address + * 11:7 fragment + * 6 write + * 5 read + * 4 exe + * 3 Z + * 2 snooped + * 1 system + * 0 valid * - * @adev: amdgpu_device pointer - * @cpu_pt_addr: cpu address of the page table - * @gpu_page_idx: entry in the page table to update - * @addr: dst addr to write into pte/pde - * @flags: access flags - * - * Update the page tables using the CPU. + * PDE format on VEGA 10: + * 63:59 block fragment size + * 58:55 reserved + * 54 P + * 53:48 reserved + * 47:6 physical base address of PD or PTE + * 5:3 reserved + * 2 C + * 1 system + * 0 valid */ -static int gmc_v9_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr, - uint32_t gpu_page_idx, uint64_t addr, - uint64_t flags) -{ - void __iomem *ptr = (void *)cpu_pt_addr; - uint64_t value; - - /* - * PTE format on VEGA 10: - * 63:59 reserved - * 58:57 mtype - * 56 F - * 55 L - * 54 P - * 53 SW - * 52 T - * 50:48 reserved - * 47:12 4k physical page base address - * 11:7 fragment - * 6 write - * 5 read - * 4 exe - * 3 Z - * 2 snooped - * 1 system - * 0 valid - * - * PDE format on VEGA 10: - * 63:59 block fragment size - * 58:55 reserved - * 54 P - * 53:48 reserved - * 47:6 physical base address of PD or PTE - * 5:3 reserved - * 2 C - * 1 system - * 0 valid - */ - - /* - * The following is for PTE only. GART does not have PDEs. - */ - value = addr & 0x0000FFFFFFFFF000ULL; - value |= flags; - writeq(value, ptr + (gpu_page_idx * 8)); - return 0; -} static uint64_t gmc_v9_0_get_vm_pte_flags(struct amdgpu_device *adev, uint32_t flags) @@ -593,7 +584,6 @@ static const struct amdgpu_gmc_funcs gmc_v9_0_gmc_funcs = { .flush_gpu_tlb = gmc_v9_0_flush_gpu_tlb, .emit_flush_gpu_tlb = gmc_v9_0_emit_flush_gpu_tlb, .emit_pasid_mapping = gmc_v9_0_emit_pasid_mapping, - .set_pte_pde = gmc_v9_0_set_pte_pde, .get_vm_pte_flags = gmc_v9_0_get_vm_pte_flags, .get_vm_pde = gmc_v9_0_get_vm_pde }; @@ -620,85 +610,6 @@ static int gmc_v9_0_early_init(void *handle) return 0; } -static int gmc_v9_0_ecc_available(struct amdgpu_device *adev) -{ - uint32_t reg_val; - uint32_t reg_addr; - uint32_t field_val; - size_t i; - uint32_t fv2; - size_t lost_sheep; - - DRM_DEBUG("ecc: gmc_v9_0_ecc_available()\n"); - - lost_sheep = 0; - for (i = 0; i < ARRAY_SIZE(ecc_umclocalcap_addrs); ++i) { - reg_addr = ecc_umclocalcap_addrs[i]; - DRM_DEBUG("ecc: " - "UMCCH_UmcLocalCap[%zu]: reg_addr: 0x%08x\n", - i, reg_addr); - reg_val = RREG32(reg_addr); - field_val = REG_GET_FIELD(reg_val, UMCCH0_0_UmcLocalCap, - EccDis); - DRM_DEBUG("ecc: " - "reg_val: 0x%08x, " - "EccDis: 0x%08x, ", - reg_val, field_val); - if (field_val) { - DRM_ERROR("ecc: UmcLocalCap:EccDis is set.\n"); - ++lost_sheep; - } - } - - for (i = 0; i < ARRAY_SIZE(ecc_umcch_umc_config_addrs); ++i) { - reg_addr = ecc_umcch_umc_config_addrs[i]; - DRM_DEBUG("ecc: " - "UMCCH0_0_UMC_CONFIG[%zu]: reg_addr: 0x%08x", - i, reg_addr); - reg_val = RREG32(reg_addr); - field_val = REG_GET_FIELD(reg_val, UMCCH0_0_UMC_CONFIG, - DramReady); - DRM_DEBUG("ecc: " - "reg_val: 0x%08x, " - "DramReady: 0x%08x\n", - reg_val, field_val); - - if (!field_val) { - DRM_ERROR("ecc: UMC_CONFIG:DramReady is not set.\n"); - ++lost_sheep; - } - } - - for (i = 0; i < ARRAY_SIZE(ecc_umcch_eccctrl_addrs); ++i) { - reg_addr = ecc_umcch_eccctrl_addrs[i]; - DRM_DEBUG("ecc: " - "UMCCH_EccCtrl[%zu]: reg_addr: 0x%08x, ", - i, reg_addr); - reg_val = RREG32(reg_addr); - field_val = REG_GET_FIELD(reg_val, UMCCH0_0_EccCtrl, - WrEccEn); - fv2 = REG_GET_FIELD(reg_val, UMCCH0_0_EccCtrl, - RdEccEn); - DRM_DEBUG("ecc: " - "reg_val: 0x%08x, " - "WrEccEn: 0x%08x, " - "RdEccEn: 0x%08x\n", - reg_val, field_val, fv2); - - if (!field_val) { - DRM_DEBUG("ecc: WrEccEn is not set\n"); - ++lost_sheep; - } - if (!fv2) { - DRM_DEBUG("ecc: RdEccEn is not set\n"); - ++lost_sheep; - } - } - - DRM_DEBUG("ecc: lost_sheep: %zu\n", lost_sheep); - return lost_sheep == 0; -} - static bool gmc_v9_0_keep_stolen_memory(struct amdgpu_device *adev) { @@ -751,10 +662,82 @@ static int gmc_v9_0_allocate_vm_inv_eng(struct amdgpu_device *adev) return 0; } +static int gmc_v9_0_ecc_late_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct ras_common_if **ras_if = &adev->gmc.ras_if; + struct ras_ih_if ih_info = { + .cb = gmc_v9_0_process_ras_data_cb, + }; + struct ras_fs_if fs_info = { + .sysfs_name = "umc_err_count", + .debugfs_name = "umc_err_inject", + }; + struct ras_common_if ras_block = { + .block = AMDGPU_RAS_BLOCK__UMC, + .type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE, + .sub_block_index = 0, + .name = "umc", + }; + int r; + + if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC)) { + amdgpu_ras_feature_enable_on_boot(adev, &ras_block, 0); + return 0; + } + /* handle resume path. */ + if (*ras_if) + goto resume; + + *ras_if = kmalloc(sizeof(**ras_if), GFP_KERNEL); + if (!*ras_if) + return -ENOMEM; + + **ras_if = ras_block; + + r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1); + if (r) + goto feature; + + ih_info.head = **ras_if; + fs_info.head = **ras_if; + + r = amdgpu_ras_interrupt_add_handler(adev, &ih_info); + if (r) + goto interrupt; + + r = amdgpu_ras_debugfs_create(adev, &fs_info); + if (r) + goto debugfs; + + r = amdgpu_ras_sysfs_create(adev, &fs_info); + if (r) + goto sysfs; +resume: + r = amdgpu_irq_get(adev, &adev->gmc.ecc_irq, 0); + if (r) + goto irq; + + return 0; +irq: + amdgpu_ras_sysfs_remove(adev, *ras_if); +sysfs: + amdgpu_ras_debugfs_remove(adev, *ras_if); +debugfs: + amdgpu_ras_interrupt_remove_handler(adev, &ih_info); +interrupt: + amdgpu_ras_feature_enable(adev, *ras_if, 0); +feature: + kfree(*ras_if); + *ras_if = NULL; + return -EINVAL; +} + + static int gmc_v9_0_late_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - int r; + bool r; if (!gmc_v9_0_keep_stolen_memory(adev)) amdgpu_bo_late_init(adev); @@ -762,20 +745,36 @@ static int gmc_v9_0_late_init(void *handle) r = gmc_v9_0_allocate_vm_inv_eng(adev); if (r) return r; + /* Check if ecc is available */ + if (!amdgpu_sriov_vf(adev)) { + switch (adev->asic_type) { + case CHIP_VEGA10: + case CHIP_VEGA20: + r = amdgpu_atomfirmware_mem_ecc_supported(adev); + if (!r) { + DRM_INFO("ECC is not present.\n"); + if (adev->df_funcs->enable_ecc_force_par_wr_rmw) + adev->df_funcs->enable_ecc_force_par_wr_rmw(adev, false); + } else { + DRM_INFO("ECC is active.\n"); + } - if (adev->asic_type == CHIP_VEGA10 && !amdgpu_sriov_vf(adev)) { - r = gmc_v9_0_ecc_available(adev); - if (r == 1) { - DRM_INFO("ECC is active.\n"); - } else if (r == 0) { - DRM_INFO("ECC is not present.\n"); - adev->df_funcs->enable_ecc_force_par_wr_rmw(adev, false); - } else { - DRM_ERROR("gmc_v9_0_ecc_available() failed. r: %d\n", r); - return r; + r = amdgpu_atomfirmware_sram_ecc_supported(adev); + if (!r) { + DRM_INFO("SRAM ECC is not present.\n"); + } else { + DRM_INFO("SRAM ECC is active.\n"); + } + break; + default: + break; } } + r = gmc_v9_0_ecc_late_init(handle); + if (r) + return r; + return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0); } @@ -787,7 +786,7 @@ static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev, base = mmhub_v1_0_get_fb_location(adev); /* add the xgmi offset of the physical node */ base += adev->gmc.xgmi.physical_node_id * adev->gmc.xgmi.node_segment_size; - amdgpu_gmc_vram_location(adev, &adev->gmc, base); + amdgpu_gmc_vram_location(adev, mc, base); amdgpu_gmc_gart_location(adev, mc); if (!amdgpu_sriov_vf(adev)) amdgpu_gmc_agp_location(adev, mc); @@ -987,6 +986,12 @@ static int gmc_v9_0_sw_init(void *handle) if (r) return r; + /* interrupt sent to DF. */ + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DF, 0, + &adev->gmc.ecc_irq); + if (r) + return r; + /* Set the internal MC address mask * This is the max address of the GPU's * internal address space. @@ -1052,6 +1057,22 @@ static int gmc_v9_0_sw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC) && + adev->gmc.ras_if) { + struct ras_common_if *ras_if = adev->gmc.ras_if; + struct ras_ih_if ih_info = { + .head = *ras_if, + }; + + /*remove fs first*/ + amdgpu_ras_debugfs_remove(adev, ras_if); + amdgpu_ras_sysfs_remove(adev, ras_if); + /*remove the IH*/ + amdgpu_ras_interrupt_remove_handler(adev, &ih_info); + amdgpu_ras_feature_enable(adev, ras_if, 0); + kfree(ras_if); + } + amdgpu_gem_force_release(adev); amdgpu_vm_manager_fini(adev); @@ -1198,6 +1219,7 @@ static int gmc_v9_0_hw_fini(void *handle) return 0; } + amdgpu_irq_put(adev, &adev->gmc.ecc_irq, 0); amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); gmc_v9_0_gart_disable(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c index 0c9a2c03504e..f2e6b148ccad 100644 --- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c @@ -2824,7 +2824,7 @@ static int kv_dpm_init(struct amdgpu_device *adev) pi->caps_tcp_ramping = true; } - if (adev->powerplay.pp_feature & PP_SCLK_DEEP_SLEEP_MASK) + if (adev->pm.pp_feature & PP_SCLK_DEEP_SLEEP_MASK) pi->caps_sclk_ds = true; else pi->caps_sclk_ds = false; diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c index d0d966d6080a..1741056e6af6 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c @@ -163,7 +163,7 @@ static void mmhub_v1_0_init_cache_regs(struct amdgpu_device *adev) /* XXX for emulation, Refer to closed source code.*/ tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, L2_PDE0_CACHE_TAG_GENERATION_MODE, 0); - tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, PDE_FAULT_CLASSIFICATION, 1); + tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, PDE_FAULT_CLASSIFICATION, 0); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, CONTEXT1_IDENTITY_ACCESS_MODE, 1); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, IDENTITY_MODE_FRAGMENT_SIZE, 0); WREG32_SOC15(MMHUB, 0, mmVM_L2_CNTL, tmp); @@ -255,7 +255,7 @@ static void mmhub_v1_0_setup_vmid_config(struct amdgpu_device *adev) block_size); /* Send no-retry XNACK on fault to suppress VM fault storm. */ tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, - RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0); + RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 1); WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_CNTL, i, tmp); WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32, i*2, 0); WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32, i*2, 0); diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c index 73851ebb3833..8dbad496b29f 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c @@ -157,6 +157,82 @@ static void xgpu_ai_mailbox_trans_msg (struct amdgpu_device *adev, xgpu_ai_mailbox_set_valid(adev, false); } +static int xgpu_ai_get_pp_clk(struct amdgpu_device *adev, u32 type, char *buf) +{ + int r = 0; + u32 req, val, size; + + if (!amdgim_is_hwperf(adev) || buf == NULL) + return -EBADRQC; + + switch(type) { + case PP_SCLK: + req = IDH_IRQ_GET_PP_SCLK; + break; + case PP_MCLK: + req = IDH_IRQ_GET_PP_MCLK; + break; + default: + return -EBADRQC; + } + + mutex_lock(&adev->virt.dpm_mutex); + + xgpu_ai_mailbox_trans_msg(adev, req, 0, 0, 0); + + r = xgpu_ai_poll_msg(adev, IDH_SUCCESS); + if (!r && adev->fw_vram_usage.va != NULL) { + val = RREG32_NO_KIQ( + SOC15_REG_OFFSET(NBIO, 0, + mmBIF_BX_PF0_MAILBOX_MSGBUF_RCV_DW1)); + size = strnlen((((char *)adev->virt.fw_reserve.p_pf2vf) + + val), PAGE_SIZE); + + if (size < PAGE_SIZE) + strcpy(buf,((char *)adev->virt.fw_reserve.p_pf2vf + val)); + else + size = 0; + + r = size; + goto out; + } + + r = xgpu_ai_poll_msg(adev, IDH_FAIL); + if(r) + pr_info("%s DPM request failed", + (type == PP_SCLK)? "SCLK" : "MCLK"); + +out: + mutex_unlock(&adev->virt.dpm_mutex); + return r; +} + +static int xgpu_ai_force_dpm_level(struct amdgpu_device *adev, u32 level) +{ + int r = 0; + u32 req = IDH_IRQ_FORCE_DPM_LEVEL; + + if (!amdgim_is_hwperf(adev)) + return -EBADRQC; + + mutex_lock(&adev->virt.dpm_mutex); + xgpu_ai_mailbox_trans_msg(adev, req, level, 0, 0); + + r = xgpu_ai_poll_msg(adev, IDH_SUCCESS); + if (!r) + goto out; + + r = xgpu_ai_poll_msg(adev, IDH_FAIL); + if (!r) + pr_info("DPM request failed"); + else + pr_info("Mailbox is broken"); + +out: + mutex_unlock(&adev->virt.dpm_mutex); + return r; +} + static int xgpu_ai_send_access_requests(struct amdgpu_device *adev, enum idh_request req) { @@ -375,4 +451,6 @@ const struct amdgpu_virt_ops xgpu_ai_virt_ops = { .reset_gpu = xgpu_ai_request_reset, .wait_reset = NULL, .trans_msg = xgpu_ai_mailbox_trans_msg, + .get_pp_clk = xgpu_ai_get_pp_clk, + .force_dpm_level = xgpu_ai_force_dpm_level, }; diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h index b4a9ceea334b..39d151b79153 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h @@ -35,6 +35,10 @@ enum idh_request { IDH_REL_GPU_FINI_ACCESS, IDH_REQ_GPU_RESET_ACCESS, + IDH_IRQ_FORCE_DPM_LEVEL = 10, + IDH_IRQ_GET_PP_SCLK, + IDH_IRQ_GET_PP_MCLK, + IDH_LOG_VF_ERROR = 200, }; @@ -43,6 +47,8 @@ enum idh_event { IDH_READY_TO_ACCESS_GPU, IDH_FLR_NOTIFICATION, IDH_FLR_NOTIFICATION_CMPL, + IDH_SUCCESS, + IDH_FAIL, IDH_EVENT_MAX }; diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c index cc967dbfd631..6590143c3f75 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c @@ -118,7 +118,8 @@ static void nbio_v6_1_ih_doorbell_range(struct amdgpu_device *adev, if (use_doorbell) { ih_doorbell_range = REG_SET_FIELD(ih_doorbell_range, BIF_IH_DOORBELL_RANGE, OFFSET, doorbell_index); - ih_doorbell_range = REG_SET_FIELD(ih_doorbell_range, BIF_IH_DOORBELL_RANGE, SIZE, 2); + ih_doorbell_range = REG_SET_FIELD(ih_doorbell_range, + BIF_IH_DOORBELL_RANGE, SIZE, 6); } else ih_doorbell_range = REG_SET_FIELD(ih_doorbell_range, BIF_IH_DOORBELL_RANGE, SIZE, 0); diff --git a/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h b/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h index f3a7d207af07..2f79765b4bdb 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h +++ b/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h @@ -43,6 +43,7 @@ enum psp_gfx_crtl_cmd_id GFX_CTRL_CMD_ID_ENABLE_INT = 0x00050000, /* enable PSP-to-Gfx interrupt */ GFX_CTRL_CMD_ID_DISABLE_INT = 0x00060000, /* disable PSP-to-Gfx interrupt */ GFX_CTRL_CMD_ID_MODE1_RST = 0x00070000, /* trigger the Mode 1 reset */ + GFX_CTRL_CMD_ID_GBR_IH_SET = 0x00080000, /* set Gbr IH_RB_CNTL registers */ GFX_CTRL_CMD_ID_CONSUME_CMD = 0x000A0000, /* send interrupt to psp for updating write pointer of vf */ GFX_CTRL_CMD_ID_DESTROY_GPCOM_RING = 0x000C0000, /* destroy GPCOM ring */ diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c index 860b70d80d3c..b91df7bd1d98 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c @@ -33,6 +33,9 @@ #include "sdma0/sdma0_4_0_offset.h" #include "nbio/nbio_7_4_offset.h" +#include "oss/osssys_4_0_offset.h" +#include "oss/osssys_4_0_sh_mask.h" + MODULE_FIRMWARE("amdgpu/vega20_sos.bin"); MODULE_FIRMWARE("amdgpu/vega20_asd.bin"); MODULE_FIRMWARE("amdgpu/vega20_ta.bin"); @@ -113,6 +116,13 @@ static int psp_v11_0_init_microcode(struct psp_context *psp) adev->psp.ta_xgmi_ucode_size = le32_to_cpu(ta_hdr->ta_xgmi_size_bytes); adev->psp.ta_xgmi_start_addr = (uint8_t *)ta_hdr + le32_to_cpu(ta_hdr->header.ucode_array_offset_bytes); + + adev->psp.ta_fw_version = le32_to_cpu(ta_hdr->header.ucode_version); + + adev->psp.ta_ras_ucode_version = le32_to_cpu(ta_hdr->ta_ras_ucode_version); + adev->psp.ta_ras_ucode_size = le32_to_cpu(ta_hdr->ta_ras_size_bytes); + adev->psp.ta_ras_start_addr = (uint8_t *)adev->psp.ta_xgmi_start_addr + + le32_to_cpu(ta_hdr->ta_ras_offset_bytes); } return 0; @@ -217,6 +227,37 @@ static int psp_v11_0_bootloader_load_sos(struct psp_context *psp) return ret; } +static void psp_v11_0_reroute_ih(struct psp_context *psp) +{ + struct amdgpu_device *adev = psp->adev; + uint32_t tmp; + + /* Change IH ring for VMC */ + tmp = REG_SET_FIELD(0, IH_CLIENT_CFG_DATA, CREDIT_RETURN_ADDR, 0x1244b); + tmp = REG_SET_FIELD(tmp, IH_CLIENT_CFG_DATA, CLIENT_TYPE, 1); + tmp = REG_SET_FIELD(tmp, IH_CLIENT_CFG_DATA, RING_ID, 1); + + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_69, 3); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_70, tmp); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_64, GFX_CTRL_CMD_ID_GBR_IH_SET); + + mdelay(20); + psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64), + 0x80000000, 0x8000FFFF, false); + + /* Change IH ring for UMC */ + tmp = REG_SET_FIELD(0, IH_CLIENT_CFG_DATA, CREDIT_RETURN_ADDR, 0x1216b); + tmp = REG_SET_FIELD(tmp, IH_CLIENT_CFG_DATA, RING_ID, 1); + + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_69, 4); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_70, tmp); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_64, GFX_CTRL_CMD_ID_GBR_IH_SET); + + mdelay(20); + psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64), + 0x80000000, 0x8000FFFF, false); +} + static int psp_v11_0_ring_init(struct psp_context *psp, enum psp_ring_type ring_type) { @@ -224,6 +265,8 @@ static int psp_v11_0_ring_init(struct psp_context *psp, struct psp_ring *ring; struct amdgpu_device *adev = psp->adev; + psp_v11_0_reroute_ih(psp); + ring = &psp->km_ring; ring->ring_type = ring_type; @@ -631,7 +674,7 @@ static int psp_v11_0_xgmi_set_topology_info(struct psp_context *psp, for (i = 0; i < topology_info_input->num_nodes; i++) { topology_info_input->nodes[i].node_id = topology->nodes[i].node_id; topology_info_input->nodes[i].num_hops = topology->nodes[i].num_hops; - topology_info_input->nodes[i].is_sharing_enabled = topology->nodes[i].is_sharing_enabled; + topology_info_input->nodes[i].is_sharing_enabled = 1; topology_info_input->nodes[i].sdma_engine = topology->nodes[i].sdma_engine; } @@ -679,6 +722,54 @@ static int psp_v11_0_xgmi_get_node_id(struct psp_context *psp, uint64_t *node_id return 0; } +static int psp_v11_0_ras_trigger_error(struct psp_context *psp, + struct ta_ras_trigger_error_input *info) +{ + struct ta_ras_shared_memory *ras_cmd; + int ret; + + if (!psp->ras.ras_initialized) + return -EINVAL; + + ras_cmd = (struct ta_ras_shared_memory *)psp->ras.ras_shared_buf; + memset(ras_cmd, 0, sizeof(struct ta_ras_shared_memory)); + + ras_cmd->cmd_id = TA_RAS_COMMAND__TRIGGER_ERROR; + ras_cmd->ras_in_message.trigger_error = *info; + + ret = psp_ras_invoke(psp, ras_cmd->cmd_id); + if (ret) + return -EINVAL; + + return ras_cmd->ras_status; +} + +static int psp_v11_0_ras_cure_posion(struct psp_context *psp, uint64_t *mode_ptr) +{ +#if 0 + // not support yet. + struct ta_ras_shared_memory *ras_cmd; + int ret; + + if (!psp->ras.ras_initialized) + return -EINVAL; + + ras_cmd = (struct ta_ras_shared_memory *)psp->ras.ras_shared_buf; + memset(ras_cmd, 0, sizeof(struct ta_ras_shared_memory)); + + ras_cmd->cmd_id = TA_RAS_COMMAND__CURE_POISON; + ras_cmd->ras_in_message.cure_poison.mode_ptr = mode_ptr; + + ret = psp_ras_invoke(psp, ras_cmd->cmd_id); + if (ret) + return -EINVAL; + + return ras_cmd->ras_status; +#else + return -EINVAL; +#endif +} + static const struct psp_funcs psp_v11_0_funcs = { .init_microcode = psp_v11_0_init_microcode, .bootloader_load_sysdrv = psp_v11_0_bootloader_load_sysdrv, @@ -695,6 +786,8 @@ static const struct psp_funcs psp_v11_0_funcs = { .xgmi_get_hive_id = psp_v11_0_xgmi_get_hive_id, .xgmi_get_node_id = psp_v11_0_xgmi_get_node_id, .support_vmr_ring = psp_v11_0_support_vmr_ring, + .ras_trigger_error = psp_v11_0_ras_trigger_error, + .ras_cure_posion = psp_v11_0_ras_cure_posion, }; void psp_v11_0_set_psp_funcs(struct psp_context *psp) diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c index 0487e3a4e9e7..143f0fae69d5 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c @@ -37,6 +37,9 @@ #include "sdma0/sdma0_4_0_offset.h" #include "nbio/nbio_6_1_offset.h" +#include "oss/osssys_4_0_offset.h" +#include "oss/osssys_4_0_sh_mask.h" + MODULE_FIRMWARE("amdgpu/vega10_sos.bin"); MODULE_FIRMWARE("amdgpu/vega10_asd.bin"); MODULE_FIRMWARE("amdgpu/vega12_sos.bin"); @@ -252,6 +255,37 @@ static int psp_v3_1_ring_init(struct psp_context *psp, return 0; } +static void psp_v3_1_reroute_ih(struct psp_context *psp) +{ + struct amdgpu_device *adev = psp->adev; + uint32_t tmp; + + /* Change IH ring for VMC */ + tmp = REG_SET_FIELD(0, IH_CLIENT_CFG_DATA, CREDIT_RETURN_ADDR, 0x1244b); + tmp = REG_SET_FIELD(tmp, IH_CLIENT_CFG_DATA, CLIENT_TYPE, 1); + tmp = REG_SET_FIELD(tmp, IH_CLIENT_CFG_DATA, RING_ID, 1); + + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_69, 3); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_70, tmp); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_64, GFX_CTRL_CMD_ID_GBR_IH_SET); + + mdelay(20); + psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64), + 0x80000000, 0x8000FFFF, false); + + /* Change IH ring for UMC */ + tmp = REG_SET_FIELD(0, IH_CLIENT_CFG_DATA, CREDIT_RETURN_ADDR, 0x1216b); + tmp = REG_SET_FIELD(tmp, IH_CLIENT_CFG_DATA, RING_ID, 1); + + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_69, 4); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_70, tmp); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_64, GFX_CTRL_CMD_ID_GBR_IH_SET); + + mdelay(20); + psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64), + 0x80000000, 0x8000FFFF, false); +} + static int psp_v3_1_ring_create(struct psp_context *psp, enum psp_ring_type ring_type) { @@ -260,6 +294,8 @@ static int psp_v3_1_ring_create(struct psp_context *psp, struct psp_ring *ring = &psp->km_ring; struct amdgpu_device *adev = psp->adev; + psp_v3_1_reroute_ih(psp); + /* Write low address of the ring to C2PMSG_69 */ psp_ring_reg = lower_32_bits(ring->ring_mem_mc_addr); WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_69, psp_ring_reg); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c index cca3552b36ed..36196372e8db 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c @@ -870,8 +870,8 @@ static int sdma_v2_4_sw_init(void *handle) r = amdgpu_ring_init(adev, ring, 1024, &adev->sdma.trap_irq, (i == 0) ? - AMDGPU_SDMA_IRQ_TRAP0 : - AMDGPU_SDMA_IRQ_TRAP1); + AMDGPU_SDMA_IRQ_INSTANCE0 : + AMDGPU_SDMA_IRQ_INSTANCE1); if (r) return r; } @@ -1006,7 +1006,7 @@ static int sdma_v2_4_set_trap_irq_state(struct amdgpu_device *adev, u32 sdma_cntl; switch (type) { - case AMDGPU_SDMA_IRQ_TRAP0: + case AMDGPU_SDMA_IRQ_INSTANCE0: switch (state) { case AMDGPU_IRQ_STATE_DISABLE: sdma_cntl = RREG32(mmSDMA0_CNTL + SDMA0_REGISTER_OFFSET); @@ -1022,7 +1022,7 @@ static int sdma_v2_4_set_trap_irq_state(struct amdgpu_device *adev, break; } break; - case AMDGPU_SDMA_IRQ_TRAP1: + case AMDGPU_SDMA_IRQ_INSTANCE1: switch (state) { case AMDGPU_IRQ_STATE_DISABLE: sdma_cntl = RREG32(mmSDMA0_CNTL + SDMA1_REGISTER_OFFSET); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c index 0ce8331baeb2..6d39544e7829 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c @@ -1154,8 +1154,8 @@ static int sdma_v3_0_sw_init(void *handle) r = amdgpu_ring_init(adev, ring, 1024, &adev->sdma.trap_irq, (i == 0) ? - AMDGPU_SDMA_IRQ_TRAP0 : - AMDGPU_SDMA_IRQ_TRAP1); + AMDGPU_SDMA_IRQ_INSTANCE0 : + AMDGPU_SDMA_IRQ_INSTANCE1); if (r) return r; } @@ -1340,7 +1340,7 @@ static int sdma_v3_0_set_trap_irq_state(struct amdgpu_device *adev, u32 sdma_cntl; switch (type) { - case AMDGPU_SDMA_IRQ_TRAP0: + case AMDGPU_SDMA_IRQ_INSTANCE0: switch (state) { case AMDGPU_IRQ_STATE_DISABLE: sdma_cntl = RREG32(mmSDMA0_CNTL + SDMA0_REGISTER_OFFSET); @@ -1356,7 +1356,7 @@ static int sdma_v3_0_set_trap_irq_state(struct amdgpu_device *adev, break; } break; - case AMDGPU_SDMA_IRQ_TRAP1: + case AMDGPU_SDMA_IRQ_INSTANCE1: switch (state) { case AMDGPU_IRQ_STATE_DISABLE: sdma_cntl = RREG32(mmSDMA0_CNTL + SDMA1_REGISTER_OFFSET); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index c816e55d43a9..1ec60f54b992 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -41,6 +41,8 @@ #include "ivsrcid/sdma0/irqsrcs_sdma0_4_0.h" #include "ivsrcid/sdma1/irqsrcs_sdma1_4_0.h" +#include "amdgpu_ras.h" + MODULE_FIRMWARE("amdgpu/vega10_sdma.bin"); MODULE_FIRMWARE("amdgpu/vega10_sdma1.bin"); MODULE_FIRMWARE("amdgpu/vega12_sdma.bin"); @@ -1493,6 +1495,87 @@ static int sdma_v4_0_early_init(void *handle) return 0; } +static int sdma_v4_0_process_ras_data_cb(struct amdgpu_device *adev, + struct amdgpu_iv_entry *entry); + +static int sdma_v4_0_late_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct ras_common_if **ras_if = &adev->sdma.ras_if; + struct ras_ih_if ih_info = { + .cb = sdma_v4_0_process_ras_data_cb, + }; + struct ras_fs_if fs_info = { + .sysfs_name = "sdma_err_count", + .debugfs_name = "sdma_err_inject", + }; + struct ras_common_if ras_block = { + .block = AMDGPU_RAS_BLOCK__SDMA, + .type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE, + .sub_block_index = 0, + .name = "sdma", + }; + int r; + + if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__SDMA)) { + amdgpu_ras_feature_enable_on_boot(adev, &ras_block, 0); + return 0; + } + + /* handle resume path. */ + if (*ras_if) + goto resume; + + *ras_if = kmalloc(sizeof(**ras_if), GFP_KERNEL); + if (!*ras_if) + return -ENOMEM; + + **ras_if = ras_block; + + r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1); + if (r) + goto feature; + + ih_info.head = **ras_if; + fs_info.head = **ras_if; + + r = amdgpu_ras_interrupt_add_handler(adev, &ih_info); + if (r) + goto interrupt; + + r = amdgpu_ras_debugfs_create(adev, &fs_info); + if (r) + goto debugfs; + + r = amdgpu_ras_sysfs_create(adev, &fs_info); + if (r) + goto sysfs; +resume: + r = amdgpu_irq_get(adev, &adev->sdma.ecc_irq, AMDGPU_SDMA_IRQ_INSTANCE0); + if (r) + goto irq; + + r = amdgpu_irq_get(adev, &adev->sdma.ecc_irq, AMDGPU_SDMA_IRQ_INSTANCE1); + if (r) { + amdgpu_irq_put(adev, &adev->sdma.ecc_irq, AMDGPU_SDMA_IRQ_INSTANCE0); + goto irq; + } + + return 0; +irq: + amdgpu_ras_sysfs_remove(adev, *ras_if); +sysfs: + amdgpu_ras_debugfs_remove(adev, *ras_if); +debugfs: + amdgpu_ras_interrupt_remove_handler(adev, &ih_info); +interrupt: + amdgpu_ras_feature_enable(adev, *ras_if, 0); +feature: + kfree(*ras_if); + *ras_if = NULL; + return -EINVAL; +} + static int sdma_v4_0_sw_init(void *handle) { struct amdgpu_ring *ring; @@ -1511,6 +1594,18 @@ static int sdma_v4_0_sw_init(void *handle) if (r) return r; + /* SDMA SRAM ECC event */ + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA0, SDMA0_4_0__SRCID__SDMA_SRAM_ECC, + &adev->sdma.ecc_irq); + if (r) + return r; + + /* SDMA SRAM ECC event */ + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA1, SDMA1_4_0__SRCID__SDMA_SRAM_ECC, + &adev->sdma.ecc_irq); + if (r) + return r; + for (i = 0; i < adev->sdma.num_instances; i++) { ring = &adev->sdma.instance[i].ring; ring->ring_obj = NULL; @@ -1526,8 +1621,8 @@ static int sdma_v4_0_sw_init(void *handle) r = amdgpu_ring_init(adev, ring, 1024, &adev->sdma.trap_irq, (i == 0) ? - AMDGPU_SDMA_IRQ_TRAP0 : - AMDGPU_SDMA_IRQ_TRAP1); + AMDGPU_SDMA_IRQ_INSTANCE0 : + AMDGPU_SDMA_IRQ_INSTANCE1); if (r) return r; @@ -1546,8 +1641,8 @@ static int sdma_v4_0_sw_init(void *handle) r = amdgpu_ring_init(adev, ring, 1024, &adev->sdma.trap_irq, (i == 0) ? - AMDGPU_SDMA_IRQ_TRAP0 : - AMDGPU_SDMA_IRQ_TRAP1); + AMDGPU_SDMA_IRQ_INSTANCE0 : + AMDGPU_SDMA_IRQ_INSTANCE1); if (r) return r; } @@ -1561,6 +1656,22 @@ static int sdma_v4_0_sw_fini(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; int i; + if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__SDMA) && + adev->sdma.ras_if) { + struct ras_common_if *ras_if = adev->sdma.ras_if; + struct ras_ih_if ih_info = { + .head = *ras_if, + }; + + /*remove fs first*/ + amdgpu_ras_debugfs_remove(adev, ras_if); + amdgpu_ras_sysfs_remove(adev, ras_if); + /*remove the IH*/ + amdgpu_ras_interrupt_remove_handler(adev, &ih_info); + amdgpu_ras_feature_enable(adev, ras_if, 0); + kfree(ras_if); + } + for (i = 0; i < adev->sdma.num_instances; i++) { amdgpu_ring_fini(&adev->sdma.instance[i].ring); if (adev->sdma.has_page_queue) @@ -1598,6 +1709,9 @@ static int sdma_v4_0_hw_fini(void *handle) if (amdgpu_sriov_vf(adev)) return 0; + amdgpu_irq_put(adev, &adev->sdma.ecc_irq, AMDGPU_SDMA_IRQ_INSTANCE0); + amdgpu_irq_put(adev, &adev->sdma.ecc_irq, AMDGPU_SDMA_IRQ_INSTANCE1); + sdma_v4_0_ctx_switch_enable(adev, false); sdma_v4_0_enable(adev, false); @@ -1666,13 +1780,12 @@ static int sdma_v4_0_set_trap_irq_state(struct amdgpu_device *adev, unsigned type, enum amdgpu_interrupt_state state) { - unsigned int instance = (type == AMDGPU_SDMA_IRQ_TRAP0) ? 0 : 1; u32 sdma_cntl; - sdma_cntl = RREG32_SDMA(instance, mmSDMA0_CNTL); + sdma_cntl = RREG32_SDMA(type, mmSDMA0_CNTL); sdma_cntl = REG_SET_FIELD(sdma_cntl, SDMA0_CNTL, TRAP_ENABLE, state == AMDGPU_IRQ_STATE_ENABLE ? 1 : 0); - WREG32_SDMA(instance, mmSDMA0_CNTL, sdma_cntl); + WREG32_SDMA(type, mmSDMA0_CNTL, sdma_cntl); return 0; } @@ -1714,6 +1827,58 @@ static int sdma_v4_0_process_trap_irq(struct amdgpu_device *adev, return 0; } +static int sdma_v4_0_process_ras_data_cb(struct amdgpu_device *adev, + struct amdgpu_iv_entry *entry) +{ + uint32_t instance, err_source; + + switch (entry->client_id) { + case SOC15_IH_CLIENTID_SDMA0: + instance = 0; + break; + case SOC15_IH_CLIENTID_SDMA1: + instance = 1; + break; + default: + return 0; + } + + switch (entry->src_id) { + case SDMA0_4_0__SRCID__SDMA_SRAM_ECC: + err_source = 0; + break; + case SDMA0_4_0__SRCID__SDMA_ECC: + err_source = 1; + break; + default: + return 0; + } + + kgd2kfd_set_sram_ecc_flag(adev->kfd.dev); + + amdgpu_ras_reset_gpu(adev, 0); + + return AMDGPU_RAS_UE; +} + +static int sdma_v4_0_process_ecc_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + struct ras_common_if *ras_if = adev->sdma.ras_if; + struct ras_dispatch_if ih_data = { + .entry = entry, + }; + + if (!ras_if) + return 0; + + ih_data.head = *ras_if; + + amdgpu_ras_interrupt_dispatch(adev, &ih_data); + return 0; +} + static int sdma_v4_0_process_illegal_inst_irq(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) @@ -1741,6 +1906,25 @@ static int sdma_v4_0_process_illegal_inst_irq(struct amdgpu_device *adev, return 0; } +static int sdma_v4_0_set_ecc_irq_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 sdma_edc_config; + + u32 reg_offset = (type == AMDGPU_SDMA_IRQ_INSTANCE0) ? + sdma_v4_0_get_reg_offset(adev, 0, mmSDMA0_EDC_CONFIG) : + sdma_v4_0_get_reg_offset(adev, 1, mmSDMA0_EDC_CONFIG); + + sdma_edc_config = RREG32(reg_offset); + sdma_edc_config = REG_SET_FIELD(sdma_edc_config, SDMA0_EDC_CONFIG, ECC_INT_ENABLE, + state == AMDGPU_IRQ_STATE_ENABLE ? 1 : 0); + WREG32(reg_offset, sdma_edc_config); + + return 0; +} + static void sdma_v4_0_update_medium_grain_clock_gating( struct amdgpu_device *adev, bool enable) @@ -1906,7 +2090,7 @@ static void sdma_v4_0_get_clockgating_state(void *handle, u32 *flags) const struct amd_ip_funcs sdma_v4_0_ip_funcs = { .name = "sdma_v4_0", .early_init = sdma_v4_0_early_init, - .late_init = NULL, + .late_init = sdma_v4_0_late_init, .sw_init = sdma_v4_0_sw_init, .sw_fini = sdma_v4_0_sw_fini, .hw_init = sdma_v4_0_hw_init, @@ -2008,11 +2192,20 @@ static const struct amdgpu_irq_src_funcs sdma_v4_0_illegal_inst_irq_funcs = { .process = sdma_v4_0_process_illegal_inst_irq, }; +static const struct amdgpu_irq_src_funcs sdma_v4_0_ecc_irq_funcs = { + .set = sdma_v4_0_set_ecc_irq_state, + .process = sdma_v4_0_process_ecc_irq, +}; + + + static void sdma_v4_0_set_irq_funcs(struct amdgpu_device *adev) { adev->sdma.trap_irq.num_types = AMDGPU_SDMA_IRQ_LAST; adev->sdma.trap_irq.funcs = &sdma_v4_0_trap_irq_funcs; adev->sdma.illegal_inst_irq.funcs = &sdma_v4_0_illegal_inst_irq_funcs; + adev->sdma.ecc_irq.num_types = AMDGPU_SDMA_IRQ_LAST; + adev->sdma.ecc_irq.funcs = &sdma_v4_0_ecc_irq_funcs; } /** @@ -2077,8 +2270,8 @@ static const struct amdgpu_buffer_funcs sdma_v4_0_buffer_funcs = { static void sdma_v4_0_set_buffer_funcs(struct amdgpu_device *adev) { adev->mman.buffer_funcs = &sdma_v4_0_buffer_funcs; - if (adev->sdma.has_page_queue) - adev->mman.buffer_funcs_ring = &adev->sdma.instance[0].page; + if (adev->sdma.has_page_queue && adev->sdma.num_instances > 1) + adev->mman.buffer_funcs_ring = &adev->sdma.instance[1].page; else adev->mman.buffer_funcs_ring = &adev->sdma.instance[0].ring; } @@ -2097,15 +2290,22 @@ static void sdma_v4_0_set_vm_pte_funcs(struct amdgpu_device *adev) unsigned i; adev->vm_manager.vm_pte_funcs = &sdma_v4_0_vm_pte_funcs; - for (i = 0; i < adev->sdma.num_instances; i++) { - if (adev->sdma.has_page_queue) + if (adev->sdma.has_page_queue && adev->sdma.num_instances > 1) { + for (i = 1; i < adev->sdma.num_instances; i++) { sched = &adev->sdma.instance[i].page.sched; - else + adev->vm_manager.vm_pte_rqs[i - 1] = + &sched->sched_rq[DRM_SCHED_PRIORITY_KERNEL]; + } + adev->vm_manager.vm_pte_num_rqs = adev->sdma.num_instances - 1; + adev->vm_manager.page_fault = &adev->sdma.instance[0].page; + } else { + for (i = 0; i < adev->sdma.num_instances; i++) { sched = &adev->sdma.instance[i].ring.sched; - adev->vm_manager.vm_pte_rqs[i] = - &sched->sched_rq[DRM_SCHED_PRIORITY_KERNEL]; + adev->vm_manager.vm_pte_rqs[i] = + &sched->sched_rq[DRM_SCHED_PRIORITY_KERNEL]; + } + adev->vm_manager.vm_pte_num_rqs = adev->sdma.num_instances; } - adev->vm_manager.vm_pte_num_rqs = adev->sdma.num_instances; } const struct amdgpu_ip_block_version sdma_v4_0_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/si_dma.c b/drivers/gpu/drm/amd/amdgpu/si_dma.c index f15f196684ba..3eeefd40dae0 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dma.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dma.c @@ -503,8 +503,8 @@ static int si_dma_sw_init(void *handle) r = amdgpu_ring_init(adev, ring, 1024, &adev->sdma.trap_irq, (i == 0) ? - AMDGPU_SDMA_IRQ_TRAP0 : - AMDGPU_SDMA_IRQ_TRAP1); + AMDGPU_SDMA_IRQ_INSTANCE0 : + AMDGPU_SDMA_IRQ_INSTANCE1); if (r) return r; } @@ -591,7 +591,7 @@ static int si_dma_set_trap_irq_state(struct amdgpu_device *adev, u32 sdma_cntl; switch (type) { - case AMDGPU_SDMA_IRQ_TRAP0: + case AMDGPU_SDMA_IRQ_INSTANCE0: switch (state) { case AMDGPU_IRQ_STATE_DISABLE: sdma_cntl = RREG32(DMA_CNTL + DMA0_REGISTER_OFFSET); @@ -607,7 +607,7 @@ static int si_dma_set_trap_irq_state(struct amdgpu_device *adev, break; } break; - case AMDGPU_SDMA_IRQ_TRAP1: + case AMDGPU_SDMA_IRQ_INSTANCE1: switch (state) { case AMDGPU_IRQ_STATE_DISABLE: sdma_cntl = RREG32(DMA_CNTL + DMA1_REGISTER_OFFSET); diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c index 41e01a7f57a4..d57e75e5c71f 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c @@ -4098,14 +4098,13 @@ static int si_notify_smc_display_change(struct amdgpu_device *adev, static void si_program_response_times(struct amdgpu_device *adev) { - u32 voltage_response_time, backbias_response_time, acpi_delay_time, vbi_time_out; + u32 voltage_response_time, acpi_delay_time, vbi_time_out; u32 vddc_dly, acpi_dly, vbi_dly; u32 reference_clock; si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_mvdd_chg_time, 1); voltage_response_time = (u32)adev->pm.dpm.voltage_response_time; - backbias_response_time = (u32)adev->pm.dpm.backbias_response_time; if (voltage_response_time == 0) voltage_response_time = 1000; diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index ed89a101f73f..bdb5ad93990d 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -63,6 +63,7 @@ #include "vcn_v1_0.h" #include "dce_virtual.h" #include "mxgpu_ai.h" +#include "amdgpu_smu.h" #define mmMP0_MISC_CGTT_CTRL0 0x01b9 #define mmMP0_MISC_CGTT_CTRL0_BASE_IDX 0 @@ -392,6 +393,7 @@ void soc15_program_register_sequence(struct amdgpu_device *adev, static int soc15_asic_mode1_reset(struct amdgpu_device *adev) { u32 i; + int ret = 0; amdgpu_atombios_scratch_regs_engine_hung(adev, true); @@ -402,7 +404,9 @@ static int soc15_asic_mode1_reset(struct amdgpu_device *adev) pci_save_state(adev->pdev); - psp_gpu_reset(adev); + ret = psp_gpu_reset(adev); + if (ret) + dev_err(adev->dev, "GPU mode1 reset failed\n"); pci_restore_state(adev->pdev); @@ -417,7 +421,7 @@ static int soc15_asic_mode1_reset(struct amdgpu_device *adev) amdgpu_atombios_scratch_regs_engine_hung(adev, false); - return 0; + return ret; } static int soc15_asic_get_baco_capability(struct amdgpu_device *adev, bool *cap) @@ -451,6 +455,8 @@ static int soc15_asic_baco_reset(struct amdgpu_device *adev) dev_info(adev->dev, "GPU BACO reset\n"); + adev->in_baco_reset = 1; + return 0; } @@ -461,6 +467,7 @@ static int soc15_asic_reset(struct amdgpu_device *adev) switch (adev->asic_type) { case CHIP_VEGA10: + case CHIP_VEGA12: soc15_asic_get_baco_capability(adev, &baco_reset); break; default: @@ -602,8 +609,12 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev) } amdgpu_device_ip_block_add(adev, &gfx_v9_0_ip_block); amdgpu_device_ip_block_add(adev, &sdma_v4_0_ip_block); - if (!amdgpu_sriov_vf(adev)) - amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); + if (!amdgpu_sriov_vf(adev)) { + if (is_support_sw_smu(adev)) + amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block); + else + amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); + } if (adev->enable_virtual_display || amdgpu_sriov_vf(adev)) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -927,7 +938,7 @@ static int soc15_common_early_init(void *handle) adev->pg_flags = AMD_PG_SUPPORT_SDMA | AMD_PG_SUPPORT_VCN; } - if (adev->powerplay.pp_feature & PP_GFXOFF_MASK) + if (adev->pm.pp_feature & PP_GFXOFF_MASK) adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG | AMD_PG_SUPPORT_CP | AMD_PG_SUPPORT_RLC_SMU_HS; diff --git a/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h b/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h new file mode 100644 index 000000000000..0b4e7b55595a --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h @@ -0,0 +1,108 @@ +/****************************************************************************\ +* +* File Name ta_ras_if.h +* Project AMD PSP SW IP Module +* +* Description Interface to the RAS Trusted Application +* +* Copyright 2019 Advanced Micro Devices, Inc. +* +* Permission is hereby granted, free of charge, to any person obtaining a copy of this software +* and associated documentation files (the "Software"), to deal in the Software without restriction, +* including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, +* and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, +* subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in all copies or substantial +* portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL +* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR +* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, +* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR +* OTHER DEALINGS IN THE SOFTWARE. +*/ +#ifndef _TA_RAS_IF_H +#define _TA_RAS_IF_H + +/* Responses have bit 31 set */ +#define RSP_ID_MASK (1U << 31) +#define RSP_ID(cmdId) (((uint32_t)(cmdId)) | RSP_ID_MASK) + +#define TA_NUM_BLOCK_MAX 14 + +enum ras_command { + TA_RAS_COMMAND__ENABLE_FEATURES = 0, + TA_RAS_COMMAND__DISABLE_FEATURES, + TA_RAS_COMMAND__TRIGGER_ERROR, +}; + +enum ta_ras_status { + TA_RAS_STATUS__SUCCESS = 0x00, + TA_RAS_STATUS__RESET_NEEDED = 0x01, + TA_RAS_STATUS__ERROR_INVALID_PARAMETER = 0x02, + TA_RAS_STATUS__ERROR_RAS_NOT_AVAILABLE = 0x03, + TA_RAS_STATUS__ERROR_RAS_DUPLICATE_CMD = 0x04, + TA_RAS_STATUS__ERROR_INJECTION_FAILED = 0x05 +}; + +enum ta_ras_block { + TA_RAS_BLOCK__UMC = 0, + TA_RAS_BLOCK__SDMA, + TA_RAS_BLOCK__GFX, + TA_RAS_BLOCK__MMHUB, + TA_RAS_BLOCK__ATHUB, + TA_RAS_BLOCK__PCIE_BIF, + TA_RAS_BLOCK__HDP, + TA_RAS_BLOCK__XGMI_WAFL, + TA_RAS_BLOCK__DF, + TA_RAS_BLOCK__SMN, + TA_RAS_BLOCK__SEM, + TA_RAS_BLOCK__MP0, + TA_RAS_BLOCK__MP1, + TA_RAS_BLOCK__FUSE = (TA_NUM_BLOCK_MAX - 1), +}; + +enum ta_ras_error_type { + TA_RAS_ERROR__NONE = 0, + TA_RAS_ERROR__PARITY = 1, + TA_RAS_ERROR__SINGLE_CORRECTABLE = 2, + TA_RAS_ERROR__MULTI_UNCORRECTABLE = 4, + TA_RAS_ERROR__POISON = 8 +}; + +struct ta_ras_enable_features_input { + enum ta_ras_block block_id; + enum ta_ras_error_type error_type; +}; + +struct ta_ras_disable_features_input { + enum ta_ras_block block_id; + enum ta_ras_error_type error_type; +}; + +struct ta_ras_trigger_error_input { + enum ta_ras_block block_id; + enum ta_ras_error_type inject_error_type; + uint32_t sub_block_index; + uint64_t address; + uint64_t value; +}; + +union ta_ras_cmd_input { + struct ta_ras_enable_features_input enable_features; + struct ta_ras_disable_features_input disable_features; + struct ta_ras_trigger_error_input trigger_error; +}; + +struct ta_ras_shared_memory { + uint32_t cmd_id; + uint32_t resp_id; + enum ta_ras_status ras_status; + uint32_t reserved; + union ta_ras_cmd_input ras_in_message; +}; + +#endif // TL_RAS_IF_H_ diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c index 6d1f804277f8..1b2f69a9a24e 100644 --- a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c @@ -136,6 +136,25 @@ static uint32_t vega10_ih_rb_cntl(struct amdgpu_ih_ring *ih, uint32_t ih_rb_cntl return ih_rb_cntl; } +static uint32_t vega10_ih_doorbell_rptr(struct amdgpu_ih_ring *ih) +{ + u32 ih_doorbell_rtpr = 0; + + if (ih->use_doorbell) { + ih_doorbell_rtpr = REG_SET_FIELD(ih_doorbell_rtpr, + IH_DOORBELL_RPTR, OFFSET, + ih->doorbell_index); + ih_doorbell_rtpr = REG_SET_FIELD(ih_doorbell_rtpr, + IH_DOORBELL_RPTR, + ENABLE, 1); + } else { + ih_doorbell_rtpr = REG_SET_FIELD(ih_doorbell_rtpr, + IH_DOORBELL_RPTR, + ENABLE, 0); + } + return ih_doorbell_rtpr; +} + /** * vega10_ih_irq_init - init and enable the interrupt ring * @@ -150,8 +169,8 @@ static uint32_t vega10_ih_rb_cntl(struct amdgpu_ih_ring *ih, uint32_t ih_rb_cntl static int vega10_ih_irq_init(struct amdgpu_device *adev) { struct amdgpu_ih_ring *ih; + u32 ih_rb_cntl; int ret = 0; - u32 ih_rb_cntl, ih_doorbell_rtpr; u32 tmp; /* disable irqs */ @@ -177,23 +196,11 @@ static int vega10_ih_irq_init(struct amdgpu_device *adev) upper_32_bits(ih->wptr_addr) & 0xFFFF); /* set rptr, wptr to 0 */ - WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR, 0); WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR, 0); + WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR, 0); - ih_doorbell_rtpr = RREG32_SOC15(OSSSYS, 0, mmIH_DOORBELL_RPTR); - if (adev->irq.ih.use_doorbell) { - ih_doorbell_rtpr = REG_SET_FIELD(ih_doorbell_rtpr, - IH_DOORBELL_RPTR, OFFSET, - adev->irq.ih.doorbell_index); - ih_doorbell_rtpr = REG_SET_FIELD(ih_doorbell_rtpr, - IH_DOORBELL_RPTR, - ENABLE, 1); - } else { - ih_doorbell_rtpr = REG_SET_FIELD(ih_doorbell_rtpr, - IH_DOORBELL_RPTR, - ENABLE, 0); - } - WREG32_SOC15(OSSSYS, 0, mmIH_DOORBELL_RPTR, ih_doorbell_rtpr); + WREG32_SOC15(OSSSYS, 0, mmIH_DOORBELL_RPTR, + vega10_ih_doorbell_rptr(ih)); ih = &adev->irq.ih1; if (ih->ring_size) { @@ -203,11 +210,18 @@ static int vega10_ih_irq_init(struct amdgpu_device *adev) ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1); ih_rb_cntl = vega10_ih_rb_cntl(ih, ih_rb_cntl); + ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, + WPTR_OVERFLOW_ENABLE, 0); + ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, + RB_FULL_DRAIN_ENABLE, 1); WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1, ih_rb_cntl); /* set rptr, wptr to 0 */ - WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR_RING1, 0); WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR_RING1, 0); + WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR_RING1, 0); + + WREG32_SOC15(OSSSYS, 0, mmIH_DOORBELL_RPTR_RING1, + vega10_ih_doorbell_rptr(ih)); } ih = &adev->irq.ih2; @@ -216,13 +230,16 @@ static int vega10_ih_irq_init(struct amdgpu_device *adev) WREG32_SOC15(OSSSYS, 0, mmIH_RB_BASE_HI_RING2, (ih->gpu_addr >> 40) & 0xff); - ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1); + ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2); ih_rb_cntl = vega10_ih_rb_cntl(ih, ih_rb_cntl); WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2, ih_rb_cntl); /* set rptr, wptr to 0 */ - WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR_RING2, 0); WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR_RING2, 0); + WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR_RING2, 0); + + WREG32_SOC15(OSSSYS, 0, mmIH_DOORBELL_RPTR_RING2, + vega10_ih_doorbell_rptr(ih)); } tmp = RREG32_SOC15(OSSSYS, 0, mmIH_STORM_CLIENT_LIST_CNTL); @@ -449,20 +466,23 @@ static int vega10_ih_sw_init(void *handle) if (r) return r; - if (adev->asic_type == CHIP_VEGA10) { - r = amdgpu_ih_ring_init(adev, &adev->irq.ih1, PAGE_SIZE, true); - if (r) - return r; - - r = amdgpu_ih_ring_init(adev, &adev->irq.ih2, PAGE_SIZE, true); - if (r) - return r; - } - - /* TODO add doorbell for IH1 & IH2 as well */ adev->irq.ih.use_doorbell = true; adev->irq.ih.doorbell_index = adev->doorbell_index.ih << 1; + r = amdgpu_ih_ring_init(adev, &adev->irq.ih1, PAGE_SIZE, true); + if (r) + return r; + + adev->irq.ih1.use_doorbell = true; + adev->irq.ih1.doorbell_index = (adev->doorbell_index.ih + 1) << 1; + + r = amdgpu_ih_ring_init(adev, &adev->irq.ih2, PAGE_SIZE, true); + if (r) + return r; + + adev->irq.ih2.use_doorbell = true; + adev->irq.ih2.doorbell_index = (adev->doorbell_index.ih + 2) << 1; + r = amdgpu_irq_init(adev); return r; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index 8be9677c0c07..2fee3063a0d6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c @@ -320,6 +320,7 @@ static const struct kfd_deviceid supported_devices[] = { { 0x9876, &carrizo_device_info }, /* Carrizo */ { 0x9877, &carrizo_device_info }, /* Carrizo */ { 0x15DD, &raven_device_info }, /* Raven */ + { 0x15D8, &raven_device_info }, /* Raven */ #endif { 0x67A0, &hawaii_device_info }, /* Hawaii */ { 0x67A1, &hawaii_device_info }, /* Hawaii */ @@ -466,6 +467,8 @@ struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd, memset(&kfd->doorbell_available_index, 0, sizeof(kfd->doorbell_available_index)); + atomic_set(&kfd->sram_ecc_flag, 0); + return kfd; } @@ -661,6 +664,9 @@ int kgd2kfd_post_reset(struct kfd_dev *kfd) return ret; count = atomic_dec_return(&kfd_locked); WARN_ONCE(count != 0, "KFD reset ref. error"); + + atomic_set(&kfd->sram_ecc_flag, 0); + return 0; } @@ -1024,6 +1030,12 @@ int kfd_gtt_sa_free(struct kfd_dev *kfd, struct kfd_mem_obj *mem_obj) return 0; } +void kgd2kfd_set_sram_ecc_flag(struct kfd_dev *kfd) +{ + if (kfd) + atomic_inc(&kfd->sram_ecc_flag); +} + #if defined(CONFIG_DEBUG_FS) /* This function will send a package to HIQ to hang the HWS diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c index e9f0e0a1b41c..6e1d41c5bf86 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c @@ -1011,25 +1011,41 @@ void kfd_signal_vm_fault_event(struct kfd_dev *dev, unsigned int pasid, void kfd_signal_reset_event(struct kfd_dev *dev) { struct kfd_hsa_hw_exception_data hw_exception_data; + struct kfd_hsa_memory_exception_data memory_exception_data; struct kfd_process *p; struct kfd_event *ev; unsigned int temp; uint32_t id, idx; + int reset_cause = atomic_read(&dev->sram_ecc_flag) ? + KFD_HW_EXCEPTION_ECC : + KFD_HW_EXCEPTION_GPU_HANG; /* Whole gpu reset caused by GPU hang and memory is lost */ memset(&hw_exception_data, 0, sizeof(hw_exception_data)); hw_exception_data.gpu_id = dev->id; hw_exception_data.memory_lost = 1; + hw_exception_data.reset_cause = reset_cause; + + memset(&memory_exception_data, 0, sizeof(memory_exception_data)); + memory_exception_data.ErrorType = KFD_MEM_ERR_SRAM_ECC; + memory_exception_data.gpu_id = dev->id; + memory_exception_data.failure.imprecise = true; idx = srcu_read_lock(&kfd_processes_srcu); hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) { mutex_lock(&p->event_mutex); id = KFD_FIRST_NONSIGNAL_EVENT_ID; - idr_for_each_entry_continue(&p->event_idr, ev, id) + idr_for_each_entry_continue(&p->event_idr, ev, id) { if (ev->type == KFD_EVENT_TYPE_HW_EXCEPTION) { ev->hw_exception_data = hw_exception_data; set_event(ev); } + if (ev->type == KFD_EVENT_TYPE_MEMORY && + reset_cause == KFD_HW_EXCEPTION_ECC) { + ev->memory_exception_data = memory_exception_data; + set_event(ev); + } + } mutex_unlock(&p->event_mutex); } srcu_read_unlock(&kfd_processes_srcu, idx); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index 0eeee3c6d6dc..9e0230965675 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -276,6 +276,9 @@ struct kfd_dev { uint64_t hive_id; bool pci_atomic_requested; + + /* SRAM ECC flag */ + atomic_t sram_ecc_flag; }; enum kfd_mempool { diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index 09da91644f9f..2cb09e088dce 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -37,6 +37,7 @@ #include "kfd_device_queue_manager.h" #include "kfd_iommu.h" #include "amdgpu_amdkfd.h" +#include "amdgpu_ras.h" /* topology_device_list - Master list of all topology devices */ static struct list_head topology_device_list; @@ -1197,6 +1198,7 @@ int kfd_topology_add_device(struct kfd_dev *gpu) void *crat_image = NULL; size_t image_size = 0; int proximity_domain; + struct amdgpu_ras *ctx; INIT_LIST_HEAD(&temp_topology_device_list); @@ -1328,6 +1330,20 @@ int kfd_topology_add_device(struct kfd_dev *gpu) dev->node_props.capability |= HSA_CAP_ATS_PRESENT; } + ctx = amdgpu_ras_get_context((struct amdgpu_device *)(dev->gpu->kgd)); + if (ctx) { + /* kfd only concerns sram ecc on GFX/SDMA and HBM ecc on UMC */ + dev->node_props.capability |= + (((ctx->features & BIT(AMDGPU_RAS_BLOCK__SDMA)) != 0) || + ((ctx->features & BIT(AMDGPU_RAS_BLOCK__GFX)) != 0)) ? + HSA_CAP_SRAM_EDCSUPPORTED : 0; + dev->node_props.capability |= ((ctx->features & BIT(AMDGPU_RAS_BLOCK__UMC)) != 0) ? + HSA_CAP_MEM_EDCSUPPORTED : 0; + + dev->node_props.capability |= (ctx->features != 0) ? + HSA_CAP_RASEVENTNOTIFY : 0; + } + kfd_debug_print_topology(); if (!res) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h index 92a19be07344..84710cfd23c2 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h @@ -48,6 +48,10 @@ #define HSA_CAP_DOORBELL_TYPE_2_0 0x2 #define HSA_CAP_AQL_QUEUE_DOUBLE_MAP 0x00004000 +#define HSA_CAP_SRAM_EDCSUPPORTED 0x00080000 +#define HSA_CAP_MEM_EDCSUPPORTED 0x00100000 +#define HSA_CAP_RASEVENTNOTIFY 0x00200000 + struct kfd_node_properties { uint64_t hive_id; uint32_t cpu_cores_count; diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index fb27783d7a54..923cb5ca5a57 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -111,7 +111,8 @@ amdgpu_dm_update_connector_after_detect(struct amdgpu_dm_connector *aconnector); static int amdgpu_dm_plane_init(struct amdgpu_display_manager *dm, struct drm_plane *plane, - unsigned long possible_crtcs); + unsigned long possible_crtcs, + const struct dc_plane_cap *plane_cap); static int amdgpu_dm_crtc_init(struct amdgpu_display_manager *dm, struct drm_plane *plane, uint32_t link_index); @@ -137,30 +138,6 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev, static void handle_cursor_update(struct drm_plane *plane, struct drm_plane_state *old_plane_state); - - -static const enum drm_plane_type dm_plane_type_default[AMDGPU_MAX_PLANES] = { - DRM_PLANE_TYPE_PRIMARY, - DRM_PLANE_TYPE_PRIMARY, - DRM_PLANE_TYPE_PRIMARY, - DRM_PLANE_TYPE_PRIMARY, - DRM_PLANE_TYPE_PRIMARY, - DRM_PLANE_TYPE_PRIMARY, -}; - -static const enum drm_plane_type dm_plane_type_carizzo[AMDGPU_MAX_PLANES] = { - DRM_PLANE_TYPE_PRIMARY, - DRM_PLANE_TYPE_PRIMARY, - DRM_PLANE_TYPE_PRIMARY, - DRM_PLANE_TYPE_OVERLAY,/* YUV Capable Underlay */ -}; - -static const enum drm_plane_type dm_plane_type_stoney[AMDGPU_MAX_PLANES] = { - DRM_PLANE_TYPE_PRIMARY, - DRM_PLANE_TYPE_PRIMARY, - DRM_PLANE_TYPE_OVERLAY, /* YUV Capable Underlay */ -}; - /* * dm_vblank_get_counter * @@ -275,12 +252,22 @@ get_crtc_by_otg_inst(struct amdgpu_device *adev, return NULL; } +static inline bool amdgpu_dm_vrr_active(struct dm_crtc_state *dm_state) +{ + return dm_state->freesync_config.state == VRR_STATE_ACTIVE_VARIABLE || + dm_state->freesync_config.state == VRR_STATE_ACTIVE_FIXED; +} + static void dm_pflip_high_irq(void *interrupt_params) { struct amdgpu_crtc *amdgpu_crtc; struct common_irq_params *irq_params = interrupt_params; struct amdgpu_device *adev = irq_params->adev; unsigned long flags; + struct drm_pending_vblank_event *e; + struct dm_crtc_state *acrtc_state; + uint32_t vpos, hpos, v_blank_start, v_blank_end; + bool vrr_active; amdgpu_crtc = get_crtc_by_otg_inst(adev, irq_params->irq_src - IRQ_TYPE_PFLIP); @@ -303,26 +290,98 @@ static void dm_pflip_high_irq(void *interrupt_params) return; } - /* Update to correct count(s) if racing with vblank irq */ - amdgpu_crtc->last_flip_vblank = drm_crtc_accurate_vblank_count(&amdgpu_crtc->base); + /* page flip completed. */ + e = amdgpu_crtc->event; + amdgpu_crtc->event = NULL; - /* wake up userspace */ - if (amdgpu_crtc->event) { - drm_crtc_send_vblank_event(&amdgpu_crtc->base, amdgpu_crtc->event); - - /* page flip completed. clean up */ - amdgpu_crtc->event = NULL; - - } else + if (!e) WARN_ON(1); + acrtc_state = to_dm_crtc_state(amdgpu_crtc->base.state); + vrr_active = amdgpu_dm_vrr_active(acrtc_state); + + /* Fixed refresh rate, or VRR scanout position outside front-porch? */ + if (!vrr_active || + !dc_stream_get_scanoutpos(acrtc_state->stream, &v_blank_start, + &v_blank_end, &hpos, &vpos) || + (vpos < v_blank_start)) { + /* Update to correct count and vblank timestamp if racing with + * vblank irq. This also updates to the correct vblank timestamp + * even in VRR mode, as scanout is past the front-porch atm. + */ + drm_crtc_accurate_vblank_count(&amdgpu_crtc->base); + + /* Wake up userspace by sending the pageflip event with proper + * count and timestamp of vblank of flip completion. + */ + if (e) { + drm_crtc_send_vblank_event(&amdgpu_crtc->base, e); + + /* Event sent, so done with vblank for this flip */ + drm_crtc_vblank_put(&amdgpu_crtc->base); + } + } else if (e) { + /* VRR active and inside front-porch: vblank count and + * timestamp for pageflip event will only be up to date after + * drm_crtc_handle_vblank() has been executed from late vblank + * irq handler after start of back-porch (vline 0). We queue the + * pageflip event for send-out by drm_crtc_handle_vblank() with + * updated timestamp and count, once it runs after us. + * + * We need to open-code this instead of using the helper + * drm_crtc_arm_vblank_event(), as that helper would + * call drm_crtc_accurate_vblank_count(), which we must + * not call in VRR mode while we are in front-porch! + */ + + /* sequence will be replaced by real count during send-out. */ + e->sequence = drm_crtc_vblank_count(&amdgpu_crtc->base); + e->pipe = amdgpu_crtc->crtc_id; + + list_add_tail(&e->base.link, &adev->ddev->vblank_event_list); + e = NULL; + } + + /* Keep track of vblank of this flip for flip throttling. We use the + * cooked hw counter, as that one incremented at start of this vblank + * of pageflip completion, so last_flip_vblank is the forbidden count + * for queueing new pageflips if vsync + VRR is enabled. + */ + amdgpu_crtc->last_flip_vblank = amdgpu_get_vblank_counter_kms(adev->ddev, + amdgpu_crtc->crtc_id); + amdgpu_crtc->pflip_status = AMDGPU_FLIP_NONE; spin_unlock_irqrestore(&adev->ddev->event_lock, flags); - DRM_DEBUG_DRIVER("%s - crtc :%d[%p], pflip_stat:AMDGPU_FLIP_NONE\n", - __func__, amdgpu_crtc->crtc_id, amdgpu_crtc); + DRM_DEBUG_DRIVER("crtc:%d[%p], pflip_stat:AMDGPU_FLIP_NONE, vrr[%d]-fp %d\n", + amdgpu_crtc->crtc_id, amdgpu_crtc, + vrr_active, (int) !e); +} - drm_crtc_vblank_put(&amdgpu_crtc->base); +static void dm_vupdate_high_irq(void *interrupt_params) +{ + struct common_irq_params *irq_params = interrupt_params; + struct amdgpu_device *adev = irq_params->adev; + struct amdgpu_crtc *acrtc; + struct dm_crtc_state *acrtc_state; + + acrtc = get_crtc_by_otg_inst(adev, irq_params->irq_src - IRQ_TYPE_VUPDATE); + + if (acrtc) { + acrtc_state = to_dm_crtc_state(acrtc->base.state); + + DRM_DEBUG_DRIVER("crtc:%d, vupdate-vrr:%d\n", acrtc->crtc_id, + amdgpu_dm_vrr_active(acrtc_state)); + + /* Core vblank handling is done here after end of front-porch in + * vrr mode, as vblank timestamping will give valid results + * while now done after front-porch. This will also deliver + * page-flip completion events that have been queued to us + * if a pageflip happened inside front-porch. + */ + if (amdgpu_dm_vrr_active(acrtc_state)) + drm_crtc_handle_vblank(&acrtc->base); + } } static void dm_crtc_high_irq(void *interrupt_params) @@ -335,11 +394,24 @@ static void dm_crtc_high_irq(void *interrupt_params) acrtc = get_crtc_by_otg_inst(adev, irq_params->irq_src - IRQ_TYPE_VBLANK); if (acrtc) { - drm_crtc_handle_vblank(&acrtc->base); - amdgpu_dm_crtc_handle_crc_irq(&acrtc->base); - acrtc_state = to_dm_crtc_state(acrtc->base.state); + DRM_DEBUG_DRIVER("crtc:%d, vupdate-vrr:%d\n", acrtc->crtc_id, + amdgpu_dm_vrr_active(acrtc_state)); + + /* Core vblank handling at start of front-porch is only possible + * in non-vrr mode, as only there vblank timestamping will give + * valid results while done in front-porch. Otherwise defer it + * to dm_vupdate_high_irq after end of front-porch. + */ + if (!amdgpu_dm_vrr_active(acrtc_state)) + drm_crtc_handle_vblank(&acrtc->base); + + /* Following stuff must happen at start of vblank, for crc + * computation and below-the-range btr support in vrr mode. + */ + amdgpu_dm_crtc_handle_crc_irq(&acrtc->base); + if (acrtc_state->stream && acrtc_state->vrr_params.supported && acrtc_state->freesync_config.state == VRR_STATE_ACTIVE_VARIABLE) { @@ -912,9 +984,16 @@ static int dm_resume(void *handle) struct drm_plane *plane; struct drm_plane_state *new_plane_state; struct dm_plane_state *dm_new_plane_state; + struct dm_atomic_state *dm_state = to_dm_atomic_state(dm->atomic_obj.state); enum dc_connection_type new_connection_type = dc_connection_none; int i; + /* Recreate dc_state - DC invalidates it when setting power state to S3. */ + dc_release_state(dm_state->context); + dm_state->context = dc_create_state(dm->dc); + /* TODO: Remove dc_state->dccg, use dc->dccg directly. */ + dc_resource_state_construct(dm->dc, dm_state->context); + /* power on hardware */ dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D0); @@ -1457,6 +1536,27 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev) dm_crtc_high_irq, c_irq_params); } + /* Use VUPDATE interrupt */ + for (i = VISLANDS30_IV_SRCID_D1_V_UPDATE_INT; i <= VISLANDS30_IV_SRCID_D6_V_UPDATE_INT; i += 2) { + r = amdgpu_irq_add_id(adev, client_id, i, &adev->vupdate_irq); + if (r) { + DRM_ERROR("Failed to add vupdate irq id!\n"); + return r; + } + + int_params.int_context = INTERRUPT_HIGH_IRQ_CONTEXT; + int_params.irq_source = + dc_interrupt_to_irq_source(dc, i, 0); + + c_irq_params = &adev->dm.vupdate_params[int_params.irq_source - DC_IRQ_SOURCE_VUPDATE1]; + + c_irq_params->adev = adev; + c_irq_params->irq_src = int_params.irq_source; + + amdgpu_dm_irq_register_interrupt(adev, &int_params, + dm_vupdate_high_irq, c_irq_params); + } + /* Use GRPH_PFLIP interrupt */ for (i = VISLANDS30_IV_SRCID_D1_GRPH_PFLIP; i <= VISLANDS30_IV_SRCID_D6_GRPH_PFLIP; i += 2) { @@ -1542,6 +1642,34 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev) dm_crtc_high_irq, c_irq_params); } + /* Use VUPDATE_NO_LOCK interrupt on DCN, which seems to correspond to + * the regular VUPDATE interrupt on DCE. We want DC_IRQ_SOURCE_VUPDATEx + * to trigger at end of each vblank, regardless of state of the lock, + * matching DCE behaviour. + */ + for (i = DCN_1_0__SRCID__OTG0_IHC_V_UPDATE_NO_LOCK_INTERRUPT; + i <= DCN_1_0__SRCID__OTG0_IHC_V_UPDATE_NO_LOCK_INTERRUPT + adev->mode_info.num_crtc - 1; + i++) { + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, i, &adev->vupdate_irq); + + if (r) { + DRM_ERROR("Failed to add vupdate irq id!\n"); + return r; + } + + int_params.int_context = INTERRUPT_HIGH_IRQ_CONTEXT; + int_params.irq_source = + dc_interrupt_to_irq_source(dc, i, 0); + + c_irq_params = &adev->dm.vupdate_params[int_params.irq_source - DC_IRQ_SOURCE_VUPDATE1]; + + c_irq_params->adev = adev; + c_irq_params->irq_src = int_params.irq_source; + + amdgpu_dm_irq_register_interrupt(adev, &int_params, + dm_vupdate_high_irq, c_irq_params); + } + /* Use GRPH_PFLIP interrupt */ for (i = DCN_1_0__SRCID__HUBP0_FLIP_INTERRUPT; i <= DCN_1_0__SRCID__HUBP0_FLIP_INTERRUPT + adev->mode_info.num_crtc - 1; @@ -1593,15 +1721,10 @@ static int dm_atomic_get_state(struct drm_atomic_state *state, struct amdgpu_device *adev = dev->dev_private; struct amdgpu_display_manager *dm = &adev->dm; struct drm_private_state *priv_state; - int ret; if (*dm_state) return 0; - ret = drm_modeset_lock(&dm->atomic_obj_lock, state->acquire_ctx); - if (ret) - return ret; - priv_state = drm_atomic_get_private_obj_state(state, &dm->atomic_obj); if (IS_ERR(priv_state)) return PTR_ERR(priv_state); @@ -1658,17 +1781,16 @@ dm_atomic_duplicate_state(struct drm_private_obj *obj) __drm_atomic_helper_private_obj_duplicate_state(obj, &new_state->base); - new_state->context = dc_create_state(); + old_state = to_dm_atomic_state(obj->state); + + if (old_state && old_state->context) + new_state->context = dc_copy_state(old_state->context); + if (!new_state->context) { kfree(new_state); return NULL; } - old_state = to_dm_atomic_state(obj->state); - if (old_state && old_state->context) - dc_resource_state_copy_construct(old_state->context, - new_state->context); - return &new_state->base; } @@ -1708,13 +1830,11 @@ static int amdgpu_dm_mode_config_init(struct amdgpu_device *adev) adev->ddev->mode_config.fb_base = adev->gmc.aper_base; - drm_modeset_lock_init(&adev->dm.atomic_obj_lock); - state = kzalloc(sizeof(*state), GFP_KERNEL); if (!state) return -ENOMEM; - state->context = dc_create_state(); + state->context = dc_create_state(adev->dm.dc); if (!state->context) { kfree(state); return -ENOMEM; @@ -1841,39 +1961,42 @@ amdgpu_dm_register_backlight_device(struct amdgpu_display_manager *dm) #endif static int initialize_plane(struct amdgpu_display_manager *dm, - struct amdgpu_mode_info *mode_info, - int plane_id) + struct amdgpu_mode_info *mode_info, int plane_id, + enum drm_plane_type plane_type, + const struct dc_plane_cap *plane_cap) { struct drm_plane *plane; unsigned long possible_crtcs; int ret = 0; plane = kzalloc(sizeof(struct drm_plane), GFP_KERNEL); - mode_info->planes[plane_id] = plane; - if (!plane) { DRM_ERROR("KMS: Failed to allocate plane\n"); return -ENOMEM; } - plane->type = mode_info->plane_type[plane_id]; + plane->type = plane_type; /* - * HACK: IGT tests expect that each plane can only have - * one possible CRTC. For now, set one CRTC for each - * plane that is not an underlay, but still allow multiple - * CRTCs for underlay planes. + * HACK: IGT tests expect that the primary plane for a CRTC + * can only have one possible CRTC. Only expose support for + * any CRTC if they're not going to be used as a primary plane + * for a CRTC - like overlay or underlay planes. */ possible_crtcs = 1 << plane_id; if (plane_id >= dm->dc->caps.max_streams) possible_crtcs = 0xff; - ret = amdgpu_dm_plane_init(dm, mode_info->planes[plane_id], possible_crtcs); + ret = amdgpu_dm_plane_init(dm, plane, possible_crtcs, plane_cap); if (ret) { DRM_ERROR("KMS: Failed to initialize plane\n"); + kfree(plane); return ret; } + if (mode_info) + mode_info->planes[plane_id] = plane; + return ret; } @@ -1916,8 +2039,9 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev) struct amdgpu_encoder *aencoder = NULL; struct amdgpu_mode_info *mode_info = &adev->mode_info; uint32_t link_cnt; - int32_t total_overlay_planes, total_primary_planes; + int32_t primary_planes; enum dc_connection_type new_connection_type = dc_connection_none; + const struct dc_plane_cap *plane; link_cnt = dm->dc->caps.max_links; if (amdgpu_dm_mode_config_init(dm->adev)) { @@ -1925,24 +2049,53 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev) return -EINVAL; } - /* Identify the number of planes to be initialized */ - total_overlay_planes = dm->dc->caps.max_slave_planes; - total_primary_planes = dm->dc->caps.max_planes - dm->dc->caps.max_slave_planes; + /* There is one primary plane per CRTC */ + primary_planes = dm->dc->caps.max_streams; + ASSERT(primary_planes <= AMDGPU_MAX_PLANES); - /* First initialize overlay planes, index starting after primary planes */ - for (i = (total_overlay_planes - 1); i >= 0; i--) { - if (initialize_plane(dm, mode_info, (total_primary_planes + i))) { - DRM_ERROR("KMS: Failed to initialize overlay plane\n"); + /* + * Initialize primary planes, implicit planes for legacy IOCTLS. + * Order is reversed to match iteration order in atomic check. + */ + for (i = (primary_planes - 1); i >= 0; i--) { + plane = &dm->dc->caps.planes[i]; + + if (initialize_plane(dm, mode_info, i, + DRM_PLANE_TYPE_PRIMARY, plane)) { + DRM_ERROR("KMS: Failed to initialize primary plane\n"); goto fail; } } - /* Initialize primary planes */ - for (i = (total_primary_planes - 1); i >= 0; i--) { - if (initialize_plane(dm, mode_info, i)) { - DRM_ERROR("KMS: Failed to initialize primary plane\n"); + /* + * Initialize overlay planes, index starting after primary planes. + * These planes have a higher DRM index than the primary planes since + * they should be considered as having a higher z-order. + * Order is reversed to match iteration order in atomic check. + * + * Only support DCN for now, and only expose one so we don't encourage + * userspace to use up all the pipes. + */ + for (i = 0; i < dm->dc->caps.max_planes; ++i) { + struct dc_plane_cap *plane = &dm->dc->caps.planes[i]; + + if (plane->type != DC_PLANE_TYPE_DCN_UNIVERSAL) + continue; + + if (!plane->blends_with_above || !plane->blends_with_below) + continue; + + if (!plane->pixel_format_support.argb8888) + continue; + + if (initialize_plane(dm, NULL, primary_planes + i, + DRM_PLANE_TYPE_OVERLAY, plane)) { + DRM_ERROR("KMS: Failed to initialize overlay plane\n"); goto fail; } + + /* Only create one overlay plane. */ + break; } for (i = 0; i < dm->dc->caps.max_streams; i++) @@ -2042,8 +2195,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev) fail: kfree(aencoder); kfree(aconnector); - for (i = 0; i < dm->dc->caps.max_planes; i++) - kfree(mode_info->planes[i]); + return -EINVAL; } @@ -2124,53 +2276,45 @@ static int dm_early_init(void *handle) adev->mode_info.num_crtc = 6; adev->mode_info.num_hpd = 6; adev->mode_info.num_dig = 6; - adev->mode_info.plane_type = dm_plane_type_default; break; case CHIP_KAVERI: adev->mode_info.num_crtc = 4; adev->mode_info.num_hpd = 6; adev->mode_info.num_dig = 7; - adev->mode_info.plane_type = dm_plane_type_default; break; case CHIP_KABINI: case CHIP_MULLINS: adev->mode_info.num_crtc = 2; adev->mode_info.num_hpd = 6; adev->mode_info.num_dig = 6; - adev->mode_info.plane_type = dm_plane_type_default; break; case CHIP_FIJI: case CHIP_TONGA: adev->mode_info.num_crtc = 6; adev->mode_info.num_hpd = 6; adev->mode_info.num_dig = 7; - adev->mode_info.plane_type = dm_plane_type_default; break; case CHIP_CARRIZO: adev->mode_info.num_crtc = 3; adev->mode_info.num_hpd = 6; adev->mode_info.num_dig = 9; - adev->mode_info.plane_type = dm_plane_type_carizzo; break; case CHIP_STONEY: adev->mode_info.num_crtc = 2; adev->mode_info.num_hpd = 6; adev->mode_info.num_dig = 9; - adev->mode_info.plane_type = dm_plane_type_stoney; break; case CHIP_POLARIS11: case CHIP_POLARIS12: adev->mode_info.num_crtc = 5; adev->mode_info.num_hpd = 5; adev->mode_info.num_dig = 5; - adev->mode_info.plane_type = dm_plane_type_default; break; case CHIP_POLARIS10: case CHIP_VEGAM: adev->mode_info.num_crtc = 6; adev->mode_info.num_hpd = 6; adev->mode_info.num_dig = 6; - adev->mode_info.plane_type = dm_plane_type_default; break; case CHIP_VEGA10: case CHIP_VEGA12: @@ -2178,14 +2322,12 @@ static int dm_early_init(void *handle) adev->mode_info.num_crtc = 6; adev->mode_info.num_hpd = 6; adev->mode_info.num_dig = 6; - adev->mode_info.plane_type = dm_plane_type_default; break; #if defined(CONFIG_DRM_AMD_DC_DCN1_0) case CHIP_RAVEN: adev->mode_info.num_crtc = 4; adev->mode_info.num_hpd = 4; adev->mode_info.num_dig = 4; - adev->mode_info.plane_type = dm_plane_type_default; break; #endif default: @@ -2243,56 +2385,63 @@ static const struct drm_encoder_funcs amdgpu_dm_encoder_funcs = { .destroy = amdgpu_dm_encoder_destroy, }; -static bool fill_rects_from_plane_state(const struct drm_plane_state *state, - struct dc_plane_state *plane_state) + +static int fill_dc_scaling_info(const struct drm_plane_state *state, + struct dc_scaling_info *scaling_info) { - plane_state->src_rect.x = state->src_x >> 16; - plane_state->src_rect.y = state->src_y >> 16; - /* we ignore the mantissa for now and do not deal with floating pixels :( */ - plane_state->src_rect.width = state->src_w >> 16; + int scale_w, scale_h; - if (plane_state->src_rect.width == 0) - return false; + memset(scaling_info, 0, sizeof(*scaling_info)); - plane_state->src_rect.height = state->src_h >> 16; - if (plane_state->src_rect.height == 0) - return false; + /* Source is fixed 16.16 but we ignore mantissa for now... */ + scaling_info->src_rect.x = state->src_x >> 16; + scaling_info->src_rect.y = state->src_y >> 16; - plane_state->dst_rect.x = state->crtc_x; - plane_state->dst_rect.y = state->crtc_y; + scaling_info->src_rect.width = state->src_w >> 16; + if (scaling_info->src_rect.width == 0) + return -EINVAL; + + scaling_info->src_rect.height = state->src_h >> 16; + if (scaling_info->src_rect.height == 0) + return -EINVAL; + + scaling_info->dst_rect.x = state->crtc_x; + scaling_info->dst_rect.y = state->crtc_y; if (state->crtc_w == 0) - return false; + return -EINVAL; - plane_state->dst_rect.width = state->crtc_w; + scaling_info->dst_rect.width = state->crtc_w; if (state->crtc_h == 0) - return false; + return -EINVAL; - plane_state->dst_rect.height = state->crtc_h; + scaling_info->dst_rect.height = state->crtc_h; - plane_state->clip_rect = plane_state->dst_rect; + /* DRM doesn't specify clipping on destination output. */ + scaling_info->clip_rect = scaling_info->dst_rect; - switch (state->rotation & DRM_MODE_ROTATE_MASK) { - case DRM_MODE_ROTATE_0: - plane_state->rotation = ROTATION_ANGLE_0; - break; - case DRM_MODE_ROTATE_90: - plane_state->rotation = ROTATION_ANGLE_90; - break; - case DRM_MODE_ROTATE_180: - plane_state->rotation = ROTATION_ANGLE_180; - break; - case DRM_MODE_ROTATE_270: - plane_state->rotation = ROTATION_ANGLE_270; - break; - default: - plane_state->rotation = ROTATION_ANGLE_0; - break; - } + /* TODO: Validate scaling per-format with DC plane caps */ + scale_w = scaling_info->dst_rect.width * 1000 / + scaling_info->src_rect.width; - return true; + if (scale_w < 250 || scale_w > 16000) + return -EINVAL; + + scale_h = scaling_info->dst_rect.height * 1000 / + scaling_info->src_rect.height; + + if (scale_h < 250 || scale_h > 16000) + return -EINVAL; + + /* + * The "scaling_quality" can be ignored for now, quality = 0 has DC + * assume reasonable defaults based on the format. + */ + + return 0; } + static int get_fb_info(const struct amdgpu_framebuffer *amdgpu_fb, uint64_t *tiling_flags) { @@ -2321,10 +2470,16 @@ static inline uint64_t get_dcc_address(uint64_t address, uint64_t tiling_flags) return offset ? (address + offset * 256) : 0; } -static bool fill_plane_dcc_attributes(struct amdgpu_device *adev, - const struct amdgpu_framebuffer *afb, - struct dc_plane_state *plane_state, - uint64_t info) +static int +fill_plane_dcc_attributes(struct amdgpu_device *adev, + const struct amdgpu_framebuffer *afb, + const enum surface_pixel_format format, + const enum dc_rotation_angle rotation, + const union plane_size *plane_size, + const union dc_tiling_info *tiling_info, + const uint64_t info, + struct dc_plane_dcc_param *dcc, + struct dc_plane_address *address) { struct dc *dc = adev->dm.dc; struct dc_dcc_surface_param input; @@ -2337,133 +2492,103 @@ static bool fill_plane_dcc_attributes(struct amdgpu_device *adev, memset(&output, 0, sizeof(output)); if (!offset) - return false; + return 0; + + if (format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN) + return 0; if (!dc->cap_funcs.get_dcc_compression_cap) - return false; + return -EINVAL; - input.format = plane_state->format; - input.surface_size.width = - plane_state->plane_size.grph.surface_size.width; - input.surface_size.height = - plane_state->plane_size.grph.surface_size.height; - input.swizzle_mode = plane_state->tiling_info.gfx9.swizzle; + input.format = format; + input.surface_size.width = plane_size->grph.surface_size.width; + input.surface_size.height = plane_size->grph.surface_size.height; + input.swizzle_mode = tiling_info->gfx9.swizzle; - if (plane_state->rotation == ROTATION_ANGLE_0 || - plane_state->rotation == ROTATION_ANGLE_180) + if (rotation == ROTATION_ANGLE_0 || rotation == ROTATION_ANGLE_180) input.scan = SCAN_DIRECTION_HORIZONTAL; - else if (plane_state->rotation == ROTATION_ANGLE_90 || - plane_state->rotation == ROTATION_ANGLE_270) + else if (rotation == ROTATION_ANGLE_90 || rotation == ROTATION_ANGLE_270) input.scan = SCAN_DIRECTION_VERTICAL; if (!dc->cap_funcs.get_dcc_compression_cap(dc, &input, &output)) - return false; + return -EINVAL; if (!output.capable) - return false; + return -EINVAL; if (i64b == 0 && output.grph.rgb.independent_64b_blks != 0) - return false; + return -EINVAL; - plane_state->dcc.enable = 1; - plane_state->dcc.grph.meta_pitch = + dcc->enable = 1; + dcc->grph.meta_pitch = AMDGPU_TILING_GET(info, DCC_PITCH_MAX) + 1; - plane_state->dcc.grph.independent_64b_blks = i64b; + dcc->grph.independent_64b_blks = i64b; dcc_address = get_dcc_address(afb->address, info); - plane_state->address.grph.meta_addr.low_part = - lower_32_bits(dcc_address); - plane_state->address.grph.meta_addr.high_part = - upper_32_bits(dcc_address); + address->grph.meta_addr.low_part = lower_32_bits(dcc_address); + address->grph.meta_addr.high_part = upper_32_bits(dcc_address); - return true; + return 0; } -static int fill_plane_attributes_from_fb(struct amdgpu_device *adev, - struct dc_plane_state *plane_state, - const struct amdgpu_framebuffer *amdgpu_fb) +static int +fill_plane_buffer_attributes(struct amdgpu_device *adev, + const struct amdgpu_framebuffer *afb, + const enum surface_pixel_format format, + const enum dc_rotation_angle rotation, + const uint64_t tiling_flags, + union dc_tiling_info *tiling_info, + union plane_size *plane_size, + struct dc_plane_dcc_param *dcc, + struct dc_plane_address *address) { - uint64_t tiling_flags; - unsigned int awidth; - const struct drm_framebuffer *fb = &amdgpu_fb->base; - int ret = 0; - struct drm_format_name_buf format_name; + const struct drm_framebuffer *fb = &afb->base; + int ret; - ret = get_fb_info( - amdgpu_fb, - &tiling_flags); + memset(tiling_info, 0, sizeof(*tiling_info)); + memset(plane_size, 0, sizeof(*plane_size)); + memset(dcc, 0, sizeof(*dcc)); + memset(address, 0, sizeof(*address)); - if (ret) - return ret; - - switch (fb->format->format) { - case DRM_FORMAT_C8: - plane_state->format = SURFACE_PIXEL_FORMAT_GRPH_PALETA_256_COLORS; - break; - case DRM_FORMAT_RGB565: - plane_state->format = SURFACE_PIXEL_FORMAT_GRPH_RGB565; - break; - case DRM_FORMAT_XRGB8888: - case DRM_FORMAT_ARGB8888: - plane_state->format = SURFACE_PIXEL_FORMAT_GRPH_ARGB8888; - break; - case DRM_FORMAT_XRGB2101010: - case DRM_FORMAT_ARGB2101010: - plane_state->format = SURFACE_PIXEL_FORMAT_GRPH_ARGB2101010; - break; - case DRM_FORMAT_XBGR2101010: - case DRM_FORMAT_ABGR2101010: - plane_state->format = SURFACE_PIXEL_FORMAT_GRPH_ABGR2101010; - break; - case DRM_FORMAT_XBGR8888: - case DRM_FORMAT_ABGR8888: - plane_state->format = SURFACE_PIXEL_FORMAT_GRPH_ABGR8888; - break; - case DRM_FORMAT_NV21: - plane_state->format = SURFACE_PIXEL_FORMAT_VIDEO_420_YCbCr; - break; - case DRM_FORMAT_NV12: - plane_state->format = SURFACE_PIXEL_FORMAT_VIDEO_420_YCrCb; - break; - default: - DRM_ERROR("Unsupported screen format %s\n", - drm_get_format_name(fb->format->format, &format_name)); - return -EINVAL; - } - - memset(&plane_state->address, 0, sizeof(plane_state->address)); - memset(&plane_state->tiling_info, 0, sizeof(plane_state->tiling_info)); - memset(&plane_state->dcc, 0, sizeof(plane_state->dcc)); - - if (plane_state->format < SURFACE_PIXEL_FORMAT_VIDEO_BEGIN) { - plane_state->address.type = PLN_ADDR_TYPE_GRAPHICS; - plane_state->plane_size.grph.surface_size.x = 0; - plane_state->plane_size.grph.surface_size.y = 0; - plane_state->plane_size.grph.surface_size.width = fb->width; - plane_state->plane_size.grph.surface_size.height = fb->height; - plane_state->plane_size.grph.surface_pitch = - fb->pitches[0] / fb->format->cpp[0]; - /* TODO: unhardcode */ - plane_state->color_space = COLOR_SPACE_SRGB; + if (format < SURFACE_PIXEL_FORMAT_VIDEO_BEGIN) { + plane_size->grph.surface_size.x = 0; + plane_size->grph.surface_size.y = 0; + plane_size->grph.surface_size.width = fb->width; + plane_size->grph.surface_size.height = fb->height; + plane_size->grph.surface_pitch = + fb->pitches[0] / fb->format->cpp[0]; + address->type = PLN_ADDR_TYPE_GRAPHICS; + address->grph.addr.low_part = lower_32_bits(afb->address); + address->grph.addr.high_part = upper_32_bits(afb->address); } else { - awidth = ALIGN(fb->width, 64); - plane_state->address.type = PLN_ADDR_TYPE_VIDEO_PROGRESSIVE; - plane_state->plane_size.video.luma_size.x = 0; - plane_state->plane_size.video.luma_size.y = 0; - plane_state->plane_size.video.luma_size.width = awidth; - plane_state->plane_size.video.luma_size.height = fb->height; - /* TODO: unhardcode */ - plane_state->plane_size.video.luma_pitch = awidth; + uint64_t chroma_addr = afb->address + fb->offsets[1]; - plane_state->plane_size.video.chroma_size.x = 0; - plane_state->plane_size.video.chroma_size.y = 0; - plane_state->plane_size.video.chroma_size.width = awidth; - plane_state->plane_size.video.chroma_size.height = fb->height; - plane_state->plane_size.video.chroma_pitch = awidth / 2; + plane_size->video.luma_size.x = 0; + plane_size->video.luma_size.y = 0; + plane_size->video.luma_size.width = fb->width; + plane_size->video.luma_size.height = fb->height; + plane_size->video.luma_pitch = + fb->pitches[0] / fb->format->cpp[0]; - /* TODO: unhardcode */ - plane_state->color_space = COLOR_SPACE_YCBCR709; + plane_size->video.chroma_size.x = 0; + plane_size->video.chroma_size.y = 0; + /* TODO: set these based on surface format */ + plane_size->video.chroma_size.width = fb->width / 2; + plane_size->video.chroma_size.height = fb->height / 2; + + plane_size->video.chroma_pitch = + fb->pitches[1] / fb->format->cpp[1]; + + address->type = PLN_ADDR_TYPE_VIDEO_PROGRESSIVE; + address->video_progressive.luma_addr.low_part = + lower_32_bits(afb->address); + address->video_progressive.luma_addr.high_part = + upper_32_bits(afb->address); + address->video_progressive.chroma_addr.low_part = + lower_32_bits(chroma_addr); + address->video_progressive.chroma_addr.high_part = + upper_32_bits(chroma_addr); } /* Fill GFX8 params */ @@ -2477,21 +2602,21 @@ static int fill_plane_attributes_from_fb(struct amdgpu_device *adev, num_banks = AMDGPU_TILING_GET(tiling_flags, NUM_BANKS); /* XXX fix me for VI */ - plane_state->tiling_info.gfx8.num_banks = num_banks; - plane_state->tiling_info.gfx8.array_mode = + tiling_info->gfx8.num_banks = num_banks; + tiling_info->gfx8.array_mode = DC_ARRAY_2D_TILED_THIN1; - plane_state->tiling_info.gfx8.tile_split = tile_split; - plane_state->tiling_info.gfx8.bank_width = bankw; - plane_state->tiling_info.gfx8.bank_height = bankh; - plane_state->tiling_info.gfx8.tile_aspect = mtaspect; - plane_state->tiling_info.gfx8.tile_mode = + tiling_info->gfx8.tile_split = tile_split; + tiling_info->gfx8.bank_width = bankw; + tiling_info->gfx8.bank_height = bankh; + tiling_info->gfx8.tile_aspect = mtaspect; + tiling_info->gfx8.tile_mode = DC_ADDR_SURF_MICRO_TILING_DISPLAY; } else if (AMDGPU_TILING_GET(tiling_flags, ARRAY_MODE) == DC_ARRAY_1D_TILED_THIN1) { - plane_state->tiling_info.gfx8.array_mode = DC_ARRAY_1D_TILED_THIN1; + tiling_info->gfx8.array_mode = DC_ARRAY_1D_TILED_THIN1; } - plane_state->tiling_info.gfx8.pipe_config = + tiling_info->gfx8.pipe_config = AMDGPU_TILING_GET(tiling_flags, PIPE_CONFIG); if (adev->asic_type == CHIP_VEGA10 || @@ -2499,60 +2624,249 @@ static int fill_plane_attributes_from_fb(struct amdgpu_device *adev, adev->asic_type == CHIP_VEGA20 || adev->asic_type == CHIP_RAVEN) { /* Fill GFX9 params */ - plane_state->tiling_info.gfx9.num_pipes = + tiling_info->gfx9.num_pipes = adev->gfx.config.gb_addr_config_fields.num_pipes; - plane_state->tiling_info.gfx9.num_banks = + tiling_info->gfx9.num_banks = adev->gfx.config.gb_addr_config_fields.num_banks; - plane_state->tiling_info.gfx9.pipe_interleave = + tiling_info->gfx9.pipe_interleave = adev->gfx.config.gb_addr_config_fields.pipe_interleave_size; - plane_state->tiling_info.gfx9.num_shader_engines = + tiling_info->gfx9.num_shader_engines = adev->gfx.config.gb_addr_config_fields.num_se; - plane_state->tiling_info.gfx9.max_compressed_frags = + tiling_info->gfx9.max_compressed_frags = adev->gfx.config.gb_addr_config_fields.max_compress_frags; - plane_state->tiling_info.gfx9.num_rb_per_se = + tiling_info->gfx9.num_rb_per_se = adev->gfx.config.gb_addr_config_fields.num_rb_per_se; - plane_state->tiling_info.gfx9.swizzle = + tiling_info->gfx9.swizzle = AMDGPU_TILING_GET(tiling_flags, SWIZZLE_MODE); - plane_state->tiling_info.gfx9.shaderEnable = 1; + tiling_info->gfx9.shaderEnable = 1; - fill_plane_dcc_attributes(adev, amdgpu_fb, plane_state, - tiling_flags); + ret = fill_plane_dcc_attributes(adev, afb, format, rotation, + plane_size, tiling_info, + tiling_flags, dcc, address); + if (ret) + return ret; } - plane_state->visible = true; - plane_state->scaling_quality.h_taps_c = 0; - plane_state->scaling_quality.v_taps_c = 0; - - /* is this needed? is plane_state zeroed at allocation? */ - plane_state->scaling_quality.h_taps = 0; - plane_state->scaling_quality.v_taps = 0; - plane_state->stereo_format = PLANE_STEREO_FORMAT_NONE; - - return ret; - + return 0; } -static int fill_plane_attributes(struct amdgpu_device *adev, - struct dc_plane_state *dc_plane_state, - struct drm_plane_state *plane_state, - struct drm_crtc_state *crtc_state) +static void +fill_blending_from_plane_state(const struct drm_plane_state *plane_state, + bool *per_pixel_alpha, bool *global_alpha, + int *global_alpha_value) +{ + *per_pixel_alpha = false; + *global_alpha = false; + *global_alpha_value = 0xff; + + if (plane_state->plane->type != DRM_PLANE_TYPE_OVERLAY) + return; + + if (plane_state->pixel_blend_mode == DRM_MODE_BLEND_PREMULTI) { + static const uint32_t alpha_formats[] = { + DRM_FORMAT_ARGB8888, + DRM_FORMAT_RGBA8888, + DRM_FORMAT_ABGR8888, + }; + uint32_t format = plane_state->fb->format->format; + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(alpha_formats); ++i) { + if (format == alpha_formats[i]) { + *per_pixel_alpha = true; + break; + } + } + } + + if (plane_state->alpha < 0xffff) { + *global_alpha = true; + *global_alpha_value = plane_state->alpha >> 8; + } +} + +static int +fill_plane_color_attributes(const struct drm_plane_state *plane_state, + const enum surface_pixel_format format, + enum dc_color_space *color_space) +{ + bool full_range; + + *color_space = COLOR_SPACE_SRGB; + + /* DRM color properties only affect non-RGB formats. */ + if (format < SURFACE_PIXEL_FORMAT_VIDEO_BEGIN) + return 0; + + full_range = (plane_state->color_range == DRM_COLOR_YCBCR_FULL_RANGE); + + switch (plane_state->color_encoding) { + case DRM_COLOR_YCBCR_BT601: + if (full_range) + *color_space = COLOR_SPACE_YCBCR601; + else + *color_space = COLOR_SPACE_YCBCR601_LIMITED; + break; + + case DRM_COLOR_YCBCR_BT709: + if (full_range) + *color_space = COLOR_SPACE_YCBCR709; + else + *color_space = COLOR_SPACE_YCBCR709_LIMITED; + break; + + case DRM_COLOR_YCBCR_BT2020: + if (full_range) + *color_space = COLOR_SPACE_2020_YCBCR; + else + return -EINVAL; + break; + + default: + return -EINVAL; + } + + return 0; +} + +static int +fill_dc_plane_info_and_addr(struct amdgpu_device *adev, + const struct drm_plane_state *plane_state, + const uint64_t tiling_flags, + struct dc_plane_info *plane_info, + struct dc_plane_address *address) +{ + const struct drm_framebuffer *fb = plane_state->fb; + const struct amdgpu_framebuffer *afb = + to_amdgpu_framebuffer(plane_state->fb); + struct drm_format_name_buf format_name; + int ret; + + memset(plane_info, 0, sizeof(*plane_info)); + + switch (fb->format->format) { + case DRM_FORMAT_C8: + plane_info->format = + SURFACE_PIXEL_FORMAT_GRPH_PALETA_256_COLORS; + break; + case DRM_FORMAT_RGB565: + plane_info->format = SURFACE_PIXEL_FORMAT_GRPH_RGB565; + break; + case DRM_FORMAT_XRGB8888: + case DRM_FORMAT_ARGB8888: + plane_info->format = SURFACE_PIXEL_FORMAT_GRPH_ARGB8888; + break; + case DRM_FORMAT_XRGB2101010: + case DRM_FORMAT_ARGB2101010: + plane_info->format = SURFACE_PIXEL_FORMAT_GRPH_ARGB2101010; + break; + case DRM_FORMAT_XBGR2101010: + case DRM_FORMAT_ABGR2101010: + plane_info->format = SURFACE_PIXEL_FORMAT_GRPH_ABGR2101010; + break; + case DRM_FORMAT_XBGR8888: + case DRM_FORMAT_ABGR8888: + plane_info->format = SURFACE_PIXEL_FORMAT_GRPH_ABGR8888; + break; + case DRM_FORMAT_NV21: + plane_info->format = SURFACE_PIXEL_FORMAT_VIDEO_420_YCbCr; + break; + case DRM_FORMAT_NV12: + plane_info->format = SURFACE_PIXEL_FORMAT_VIDEO_420_YCrCb; + break; + default: + DRM_ERROR( + "Unsupported screen format %s\n", + drm_get_format_name(fb->format->format, &format_name)); + return -EINVAL; + } + + switch (plane_state->rotation & DRM_MODE_ROTATE_MASK) { + case DRM_MODE_ROTATE_0: + plane_info->rotation = ROTATION_ANGLE_0; + break; + case DRM_MODE_ROTATE_90: + plane_info->rotation = ROTATION_ANGLE_90; + break; + case DRM_MODE_ROTATE_180: + plane_info->rotation = ROTATION_ANGLE_180; + break; + case DRM_MODE_ROTATE_270: + plane_info->rotation = ROTATION_ANGLE_270; + break; + default: + plane_info->rotation = ROTATION_ANGLE_0; + break; + } + + plane_info->visible = true; + plane_info->stereo_format = PLANE_STEREO_FORMAT_NONE; + + ret = fill_plane_color_attributes(plane_state, plane_info->format, + &plane_info->color_space); + if (ret) + return ret; + + ret = fill_plane_buffer_attributes(adev, afb, plane_info->format, + plane_info->rotation, tiling_flags, + &plane_info->tiling_info, + &plane_info->plane_size, + &plane_info->dcc, address); + if (ret) + return ret; + + fill_blending_from_plane_state( + plane_state, &plane_info->per_pixel_alpha, + &plane_info->global_alpha, &plane_info->global_alpha_value); + + return 0; +} + +static int fill_dc_plane_attributes(struct amdgpu_device *adev, + struct dc_plane_state *dc_plane_state, + struct drm_plane_state *plane_state, + struct drm_crtc_state *crtc_state) { const struct amdgpu_framebuffer *amdgpu_fb = to_amdgpu_framebuffer(plane_state->fb); - const struct drm_crtc *crtc = plane_state->crtc; - int ret = 0; - - if (!fill_rects_from_plane_state(plane_state, dc_plane_state)) - return -EINVAL; - - ret = fill_plane_attributes_from_fb( - crtc->dev->dev_private, - dc_plane_state, - amdgpu_fb); + struct dc_scaling_info scaling_info; + struct dc_plane_info plane_info; + uint64_t tiling_flags; + int ret; + ret = fill_dc_scaling_info(plane_state, &scaling_info); if (ret) return ret; + dc_plane_state->src_rect = scaling_info.src_rect; + dc_plane_state->dst_rect = scaling_info.dst_rect; + dc_plane_state->clip_rect = scaling_info.clip_rect; + dc_plane_state->scaling_quality = scaling_info.scaling_quality; + + ret = get_fb_info(amdgpu_fb, &tiling_flags); + if (ret) + return ret; + + ret = fill_dc_plane_info_and_addr(adev, plane_state, tiling_flags, + &plane_info, + &dc_plane_state->address); + if (ret) + return ret; + + dc_plane_state->format = plane_info.format; + dc_plane_state->color_space = plane_info.color_space; + dc_plane_state->format = plane_info.format; + dc_plane_state->plane_size = plane_info.plane_size; + dc_plane_state->rotation = plane_info.rotation; + dc_plane_state->horizontal_mirror = plane_info.horizontal_mirror; + dc_plane_state->stereo_format = plane_info.stereo_format; + dc_plane_state->tiling_info = plane_info.tiling_info; + dc_plane_state->visible = plane_info.visible; + dc_plane_state->per_pixel_alpha = plane_info.per_pixel_alpha; + dc_plane_state->global_alpha = plane_info.global_alpha; + dc_plane_state->global_alpha_value = plane_info.global_alpha_value; + dc_plane_state->dcc = plane_info.dcc; + /* * Always set input transfer function, since plane state is refreshed * every time. @@ -3136,12 +3450,41 @@ dm_crtc_duplicate_state(struct drm_crtc *crtc) return &state->base; } +static inline int dm_set_vupdate_irq(struct drm_crtc *crtc, bool enable) +{ + enum dc_irq_source irq_source; + struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc); + struct amdgpu_device *adev = crtc->dev->dev_private; + int rc; + + irq_source = IRQ_TYPE_VUPDATE + acrtc->otg_inst; + + rc = dc_interrupt_set(adev->dm.dc, irq_source, enable) ? 0 : -EBUSY; + + DRM_DEBUG_DRIVER("crtc %d - vupdate irq %sabling: r=%d\n", + acrtc->crtc_id, enable ? "en" : "dis", rc); + return rc; +} static inline int dm_set_vblank(struct drm_crtc *crtc, bool enable) { enum dc_irq_source irq_source; struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc); struct amdgpu_device *adev = crtc->dev->dev_private; + struct dm_crtc_state *acrtc_state = to_dm_crtc_state(crtc->state); + int rc = 0; + + if (enable) { + /* vblank irq on -> Only need vupdate irq in vrr mode */ + if (amdgpu_dm_vrr_active(acrtc_state)) + rc = dm_set_vupdate_irq(crtc, true); + } else { + /* vblank irq off -> vupdate irq off */ + rc = dm_set_vupdate_irq(crtc, false); + } + + if (rc) + return rc; irq_source = IRQ_TYPE_VBLANK + acrtc->otg_inst; return dc_interrupt_set(adev->dm.dc, irq_source, enable) ? 0 : -EBUSY; @@ -3519,6 +3862,38 @@ static void dm_crtc_helper_disable(struct drm_crtc *crtc) { } +static bool does_crtc_have_active_plane(struct drm_crtc_state *new_crtc_state) +{ + struct drm_atomic_state *state = new_crtc_state->state; + struct drm_plane *plane; + int num_active = 0; + + drm_for_each_plane_mask(plane, state->dev, new_crtc_state->plane_mask) { + struct drm_plane_state *new_plane_state; + + /* Cursor planes are "fake". */ + if (plane->type == DRM_PLANE_TYPE_CURSOR) + continue; + + new_plane_state = drm_atomic_get_new_plane_state(state, plane); + + if (!new_plane_state) { + /* + * The plane is enable on the CRTC and hasn't changed + * state. This means that it previously passed + * validation and is therefore enabled. + */ + num_active += 1; + continue; + } + + /* We need a framebuffer to be considered enabled. */ + num_active += (new_plane_state->fb != NULL); + } + + return num_active > 0; +} + static int dm_crtc_helper_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state) { @@ -3537,6 +3912,11 @@ static int dm_crtc_helper_atomic_check(struct drm_crtc *crtc, if (!dm_crtc_state->stream) return 0; + /* We want at least one hardware plane enabled to use the stream. */ + if (state->enable && state->active && + !does_crtc_have_active_plane(state)) + return -EINVAL; + if (dc_validate_stream(dc, dm_crtc_state->stream) == DC_OK) return 0; @@ -3583,11 +3963,8 @@ static void dm_drm_plane_reset(struct drm_plane *plane) amdgpu_state = kzalloc(sizeof(*amdgpu_state), GFP_KERNEL); WARN_ON(amdgpu_state == NULL); - if (amdgpu_state) { - plane->state = &amdgpu_state->base; - plane->state->plane = plane; - plane->state->rotation = DRM_MODE_ROTATE_0; - } + if (amdgpu_state) + __drm_atomic_helper_plane_reset(plane, &amdgpu_state->base); } static struct drm_plane_state * @@ -3637,10 +4014,8 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane, struct drm_gem_object *obj; struct amdgpu_device *adev; struct amdgpu_bo *rbo; - uint64_t chroma_addr = 0; struct dm_plane_state *dm_plane_state_new, *dm_plane_state_old; - uint64_t tiling_flags, dcc_address; - unsigned int awidth; + uint64_t tiling_flags; uint32_t domain; int r; @@ -3693,29 +4068,11 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane, dm_plane_state_old->dc_state != dm_plane_state_new->dc_state) { struct dc_plane_state *plane_state = dm_plane_state_new->dc_state; - if (plane_state->format < SURFACE_PIXEL_FORMAT_VIDEO_BEGIN) { - plane_state->address.grph.addr.low_part = lower_32_bits(afb->address); - plane_state->address.grph.addr.high_part = upper_32_bits(afb->address); - - dcc_address = - get_dcc_address(afb->address, tiling_flags); - plane_state->address.grph.meta_addr.low_part = - lower_32_bits(dcc_address); - plane_state->address.grph.meta_addr.high_part = - upper_32_bits(dcc_address); - } else { - awidth = ALIGN(new_state->fb->width, 64); - plane_state->address.type = PLN_ADDR_TYPE_VIDEO_PROGRESSIVE; - plane_state->address.video_progressive.luma_addr.low_part - = lower_32_bits(afb->address); - plane_state->address.video_progressive.luma_addr.high_part - = upper_32_bits(afb->address); - chroma_addr = afb->address + (u64)awidth * new_state->fb->height; - plane_state->address.video_progressive.chroma_addr.low_part - = lower_32_bits(chroma_addr); - plane_state->address.video_progressive.chroma_addr.high_part - = upper_32_bits(chroma_addr); - } + fill_plane_buffer_attributes( + adev, afb, plane_state->format, plane_state->rotation, + tiling_flags, &plane_state->tiling_info, + &plane_state->plane_size, &plane_state->dcc, + &plane_state->address); } return 0; @@ -3747,13 +4104,18 @@ static int dm_plane_atomic_check(struct drm_plane *plane, { struct amdgpu_device *adev = plane->dev->dev_private; struct dc *dc = adev->dm.dc; - struct dm_plane_state *dm_plane_state = to_dm_plane_state(state); + struct dm_plane_state *dm_plane_state; + struct dc_scaling_info scaling_info; + int ret; + + dm_plane_state = to_dm_plane_state(state); if (!dm_plane_state->dc_state) return 0; - if (!fill_rects_from_plane_state(state, dm_plane_state->dc_state)) - return -EINVAL; + ret = fill_dc_scaling_info(state, &scaling_info); + if (ret) + return ret; if (dc_validate_plane(dc, dm_plane_state->dc_state) == DC_OK) return 0; @@ -3828,62 +4190,111 @@ static const uint32_t rgb_formats[] = { DRM_FORMAT_ABGR8888, }; -static const uint32_t yuv_formats[] = { - DRM_FORMAT_NV12, - DRM_FORMAT_NV21, +static const uint32_t overlay_formats[] = { + DRM_FORMAT_XRGB8888, + DRM_FORMAT_ARGB8888, + DRM_FORMAT_RGBA8888, + DRM_FORMAT_XBGR8888, + DRM_FORMAT_ABGR8888, }; static const u32 cursor_formats[] = { DRM_FORMAT_ARGB8888 }; -static int amdgpu_dm_plane_init(struct amdgpu_display_manager *dm, - struct drm_plane *plane, - unsigned long possible_crtcs) +static int get_plane_formats(const struct drm_plane *plane, + const struct dc_plane_cap *plane_cap, + uint32_t *formats, int max_formats) { - int res = -EPERM; + int i, num_formats = 0; + + /* + * TODO: Query support for each group of formats directly from + * DC plane caps. This will require adding more formats to the + * caps list. + */ switch (plane->type) { case DRM_PLANE_TYPE_PRIMARY: - res = drm_universal_plane_init( - dm->adev->ddev, - plane, - possible_crtcs, - &dm_plane_funcs, - rgb_formats, - ARRAY_SIZE(rgb_formats), - NULL, plane->type, NULL); + for (i = 0; i < ARRAY_SIZE(rgb_formats); ++i) { + if (num_formats >= max_formats) + break; + + formats[num_formats++] = rgb_formats[i]; + } + + if (plane_cap && plane_cap->pixel_format_support.nv12) + formats[num_formats++] = DRM_FORMAT_NV12; break; + case DRM_PLANE_TYPE_OVERLAY: - res = drm_universal_plane_init( - dm->adev->ddev, - plane, - possible_crtcs, - &dm_plane_funcs, - yuv_formats, - ARRAY_SIZE(yuv_formats), - NULL, plane->type, NULL); + for (i = 0; i < ARRAY_SIZE(overlay_formats); ++i) { + if (num_formats >= max_formats) + break; + + formats[num_formats++] = overlay_formats[i]; + } break; + case DRM_PLANE_TYPE_CURSOR: - res = drm_universal_plane_init( - dm->adev->ddev, - plane, - possible_crtcs, - &dm_plane_funcs, - cursor_formats, - ARRAY_SIZE(cursor_formats), - NULL, plane->type, NULL); + for (i = 0; i < ARRAY_SIZE(cursor_formats); ++i) { + if (num_formats >= max_formats) + break; + + formats[num_formats++] = cursor_formats[i]; + } break; } + return num_formats; +} + +static int amdgpu_dm_plane_init(struct amdgpu_display_manager *dm, + struct drm_plane *plane, + unsigned long possible_crtcs, + const struct dc_plane_cap *plane_cap) +{ + uint32_t formats[32]; + int num_formats; + int res = -EPERM; + + num_formats = get_plane_formats(plane, plane_cap, formats, + ARRAY_SIZE(formats)); + + res = drm_universal_plane_init(dm->adev->ddev, plane, possible_crtcs, + &dm_plane_funcs, formats, num_formats, + NULL, plane->type, NULL); + if (res) + return res; + + if (plane->type == DRM_PLANE_TYPE_OVERLAY && + plane_cap && plane_cap->per_pixel_alpha) { + unsigned int blend_caps = BIT(DRM_MODE_BLEND_PIXEL_NONE) | + BIT(DRM_MODE_BLEND_PREMULTI); + + drm_plane_create_alpha_property(plane); + drm_plane_create_blend_mode_property(plane, blend_caps); + } + + if (plane->type == DRM_PLANE_TYPE_PRIMARY && + plane_cap && plane_cap->pixel_format_support.nv12) { + /* This only affects YUV formats. */ + drm_plane_create_color_properties( + plane, + BIT(DRM_COLOR_YCBCR_BT601) | + BIT(DRM_COLOR_YCBCR_BT709), + BIT(DRM_COLOR_YCBCR_LIMITED_RANGE) | + BIT(DRM_COLOR_YCBCR_FULL_RANGE), + DRM_COLOR_YCBCR_BT709, DRM_COLOR_YCBCR_LIMITED_RANGE); + } + drm_plane_helper_add(plane, &dm_plane_helper_funcs); /* Create (reset) the plane state */ if (plane->funcs->reset) plane->funcs->reset(plane); - - return res; + return 0; } static int amdgpu_dm_crtc_init(struct amdgpu_display_manager *dm, @@ -3900,7 +4311,7 @@ static int amdgpu_dm_crtc_init(struct amdgpu_display_manager *dm, goto fail; cursor_plane->type = DRM_PLANE_TYPE_CURSOR; - res = amdgpu_dm_plane_init(dm, cursor_plane, 0); + res = amdgpu_dm_plane_init(dm, cursor_plane, 0, NULL); acrtc = kzalloc(sizeof(struct amdgpu_crtc), GFP_KERNEL); if (!acrtc) @@ -4334,6 +4745,8 @@ static int amdgpu_dm_connector_init(struct amdgpu_display_manager *dm, DRM_ERROR("Failed to create debugfs for connector"); goto out_free; } + aconnector->debugfs_dpcd_address = 0; + aconnector->debugfs_dpcd_size = 0; #endif if (connector_type == DRM_MODE_CONNECTOR_DisplayPort @@ -4473,9 +4886,13 @@ static int get_cursor_position(struct drm_plane *plane, struct drm_crtc *crtc, x = plane->state->crtc_x; y = plane->state->crtc_y; - /* avivo cursor are offset into the total surface */ - x += crtc->primary->state->src_x >> 16; - y += crtc->primary->state->src_y >> 16; + + if (crtc->primary->state) { + /* avivo cursor are offset into the total surface */ + x += crtc->primary->state->src_x >> 16; + y += crtc->primary->state->src_y >> 16; + } + if (x < 0) { xorigin = min(-x, amdgpu_crtc->max_cursor_width - 1); x = 0; @@ -4533,6 +4950,7 @@ static void handle_cursor_update(struct drm_plane *plane, amdgpu_crtc->cursor_width = plane->state->crtc_w; amdgpu_crtc->cursor_height = plane->state->crtc_h; + memset(&attributes, 0, sizeof(attributes)); attributes.address.high_part = upper_32_bits(address); attributes.address.low_part = lower_32_bits(address); attributes.width = plane->state->crtc_w; @@ -4583,7 +5001,6 @@ static void update_freesync_state_on_stream( { struct mod_vrr_params vrr_params = new_crtc_state->vrr_params; struct dc_info_packet vrr_infopacket = {0}; - struct mod_freesync_config config = new_crtc_state->freesync_config; if (!new_stream) return; @@ -4596,20 +5013,6 @@ static void update_freesync_state_on_stream( if (!new_stream->timing.h_total || !new_stream->timing.v_total) return; - if (new_crtc_state->vrr_supported && - config.min_refresh_in_uhz && - config.max_refresh_in_uhz) { - config.state = new_crtc_state->base.vrr_enabled ? - VRR_STATE_ACTIVE_VARIABLE : - VRR_STATE_INACTIVE; - } else { - config.state = VRR_STATE_UNSUPPORTED; - } - - mod_freesync_build_vrr_params(dm->freesync_module, - new_stream, - &config, &vrr_params); - if (surface) { mod_freesync_handle_preflip( dm->freesync_module, @@ -4650,12 +5053,82 @@ static void update_freesync_state_on_stream( (int)vrr_params.state); } +static void pre_update_freesync_state_on_stream( + struct amdgpu_display_manager *dm, + struct dm_crtc_state *new_crtc_state) +{ + struct dc_stream_state *new_stream = new_crtc_state->stream; + struct mod_vrr_params vrr_params = new_crtc_state->vrr_params; + struct mod_freesync_config config = new_crtc_state->freesync_config; + + if (!new_stream) + return; + + /* + * TODO: Determine why min/max totals and vrefresh can be 0 here. + * For now it's sufficient to just guard against these conditions. + */ + if (!new_stream->timing.h_total || !new_stream->timing.v_total) + return; + + if (new_crtc_state->vrr_supported && + config.min_refresh_in_uhz && + config.max_refresh_in_uhz) { + config.state = new_crtc_state->base.vrr_enabled ? + VRR_STATE_ACTIVE_VARIABLE : + VRR_STATE_INACTIVE; + } else { + config.state = VRR_STATE_UNSUPPORTED; + } + + mod_freesync_build_vrr_params(dm->freesync_module, + new_stream, + &config, &vrr_params); + + new_crtc_state->freesync_timing_changed |= + (memcmp(&new_crtc_state->vrr_params.adjust, + &vrr_params.adjust, + sizeof(vrr_params.adjust)) != 0); + + new_crtc_state->vrr_params = vrr_params; +} + +static void amdgpu_dm_handle_vrr_transition(struct dm_crtc_state *old_state, + struct dm_crtc_state *new_state) +{ + bool old_vrr_active = amdgpu_dm_vrr_active(old_state); + bool new_vrr_active = amdgpu_dm_vrr_active(new_state); + + if (!old_vrr_active && new_vrr_active) { + /* Transition VRR inactive -> active: + * While VRR is active, we must not disable vblank irq, as a + * reenable after disable would compute bogus vblank/pflip + * timestamps if it likely happened inside display front-porch. + * + * We also need vupdate irq for the actual core vblank handling + * at end of vblank. + */ + dm_set_vupdate_irq(new_state->base.crtc, true); + drm_crtc_vblank_get(new_state->base.crtc); + DRM_DEBUG_DRIVER("%s: crtc=%u VRR off->on: Get vblank ref\n", + __func__, new_state->base.crtc->base.id); + } else if (old_vrr_active && !new_vrr_active) { + /* Transition VRR active -> inactive: + * Allow vblank irq disable again for fixed refresh rate. + */ + dm_set_vupdate_irq(new_state->base.crtc, false); + drm_crtc_vblank_put(new_state->base.crtc); + DRM_DEBUG_DRIVER("%s: crtc=%u VRR on->off: Drop vblank ref\n", + __func__, new_state->base.crtc->base.id); + } +} + static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, struct dc_state *dc_state, struct drm_device *dev, struct amdgpu_display_manager *dm, struct drm_crtc *pcrtc, - bool *wait_for_vblank) + bool wait_for_vblank) { uint32_t i, r; uint64_t timestamp_ns; @@ -4667,32 +5140,25 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, struct dm_crtc_state *acrtc_state = to_dm_crtc_state(new_pcrtc_state); struct dm_crtc_state *dm_old_crtc_state = to_dm_crtc_state(drm_atomic_get_old_crtc_state(state, pcrtc)); - int flip_count = 0, planes_count = 0, vpos, hpos; + int planes_count = 0, vpos, hpos; unsigned long flags; struct amdgpu_bo *abo; - uint64_t tiling_flags, dcc_address; - uint32_t target, target_vblank; - uint64_t last_flip_vblank; - bool vrr_active = acrtc_state->freesync_config.state == VRR_STATE_ACTIVE_VARIABLE; - - struct { - struct dc_surface_update surface_updates[MAX_SURFACES]; - struct dc_flip_addrs flip_addrs[MAX_SURFACES]; - struct dc_stream_update stream_update; - } *flip; - + uint64_t tiling_flags; + uint32_t target_vblank, last_flip_vblank; + bool vrr_active = amdgpu_dm_vrr_active(acrtc_state); + bool pflip_present = false; struct { struct dc_surface_update surface_updates[MAX_SURFACES]; struct dc_plane_info plane_infos[MAX_SURFACES]; struct dc_scaling_info scaling_infos[MAX_SURFACES]; + struct dc_flip_addrs flip_addrs[MAX_SURFACES]; struct dc_stream_update stream_update; - } *full; + } *bundle; - flip = kzalloc(sizeof(*flip), GFP_KERNEL); - full = kzalloc(sizeof(*full), GFP_KERNEL); + bundle = kzalloc(sizeof(*bundle), GFP_KERNEL); - if (!flip || !full) { - dm_error("Failed to allocate update bundles\n"); + if (!bundle) { + dm_error("Failed to allocate update bundle\n"); goto cleanup; } @@ -4701,8 +5167,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, struct drm_crtc *crtc = new_plane_state->crtc; struct drm_crtc_state *new_crtc_state; struct drm_framebuffer *fb = new_plane_state->fb; - struct amdgpu_framebuffer *afb = to_amdgpu_framebuffer(fb); - bool pflip_needed; + bool plane_needs_flip; struct dc_plane_state *dc_plane; struct dm_plane_state *dm_new_plane_state = to_dm_plane_state(new_plane_state); @@ -4717,122 +5182,90 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, if (!new_crtc_state->active) continue; - pflip_needed = old_plane_state->fb && - old_plane_state->fb != new_plane_state->fb; - dc_plane = dm_new_plane_state->dc_state; - if (pflip_needed) { - /* - * Assume even ONE crtc with immediate flip means - * entire can't wait for VBLANK - * TODO Check if it's correct - */ - if (new_pcrtc_state->pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC) - *wait_for_vblank = false; - - /* - * TODO This might fail and hence better not used, wait - * explicitly on fences instead - * and in general should be called for - * blocking commit to as per framework helpers - */ - abo = gem_to_amdgpu_bo(fb->obj[0]); - r = amdgpu_bo_reserve(abo, true); - if (unlikely(r != 0)) - DRM_ERROR("failed to reserve buffer before flip\n"); - - /* - * Wait for all fences on this FB. Do limited wait to avoid - * deadlock during GPU reset when this fence will not signal - * but we hold reservation lock for the BO. - */ - r = reservation_object_wait_timeout_rcu(abo->tbo.resv, - true, false, - msecs_to_jiffies(5000)); - if (unlikely(r == 0)) - DRM_ERROR("Waiting for fences timed out."); - - - - amdgpu_bo_get_tiling_flags(abo, &tiling_flags); - - amdgpu_bo_unreserve(abo); - - flip->flip_addrs[flip_count].address.grph.addr.low_part = lower_32_bits(afb->address); - flip->flip_addrs[flip_count].address.grph.addr.high_part = upper_32_bits(afb->address); - - dcc_address = get_dcc_address(afb->address, tiling_flags); - flip->flip_addrs[flip_count].address.grph.meta_addr.low_part = lower_32_bits(dcc_address); - flip->flip_addrs[flip_count].address.grph.meta_addr.high_part = upper_32_bits(dcc_address); - - flip->flip_addrs[flip_count].flip_immediate = - (crtc->state->pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC) != 0; - - timestamp_ns = ktime_get_ns(); - flip->flip_addrs[flip_count].flip_timestamp_in_us = div_u64(timestamp_ns, 1000); - flip->surface_updates[flip_count].flip_addr = &flip->flip_addrs[flip_count]; - flip->surface_updates[flip_count].surface = dc_plane; - - if (!flip->surface_updates[flip_count].surface) { - DRM_ERROR("No surface for CRTC: id=%d\n", - acrtc_attach->crtc_id); - continue; - } - - if (plane == pcrtc->primary) - update_freesync_state_on_stream( - dm, - acrtc_state, - acrtc_state->stream, - dc_plane, - flip->flip_addrs[flip_count].flip_timestamp_in_us); - - DRM_DEBUG_DRIVER("%s Flipping to hi: 0x%x, low: 0x%x\n", - __func__, - flip->flip_addrs[flip_count].address.grph.addr.high_part, - flip->flip_addrs[flip_count].address.grph.addr.low_part); - - flip_count += 1; - } - - full->surface_updates[planes_count].surface = dc_plane; + bundle->surface_updates[planes_count].surface = dc_plane; if (new_pcrtc_state->color_mgmt_changed) { - full->surface_updates[planes_count].gamma = dc_plane->gamma_correction; - full->surface_updates[planes_count].in_transfer_func = dc_plane->in_transfer_func; + bundle->surface_updates[planes_count].gamma = dc_plane->gamma_correction; + bundle->surface_updates[planes_count].in_transfer_func = dc_plane->in_transfer_func; } + fill_dc_scaling_info(new_plane_state, + &bundle->scaling_infos[planes_count]); - full->scaling_infos[planes_count].scaling_quality = dc_plane->scaling_quality; - full->scaling_infos[planes_count].src_rect = dc_plane->src_rect; - full->scaling_infos[planes_count].dst_rect = dc_plane->dst_rect; - full->scaling_infos[planes_count].clip_rect = dc_plane->clip_rect; - full->surface_updates[planes_count].scaling_info = &full->scaling_infos[planes_count]; + bundle->surface_updates[planes_count].scaling_info = + &bundle->scaling_infos[planes_count]; + plane_needs_flip = old_plane_state->fb && new_plane_state->fb; - full->plane_infos[planes_count].color_space = dc_plane->color_space; - full->plane_infos[planes_count].format = dc_plane->format; - full->plane_infos[planes_count].plane_size = dc_plane->plane_size; - full->plane_infos[planes_count].rotation = dc_plane->rotation; - full->plane_infos[planes_count].horizontal_mirror = dc_plane->horizontal_mirror; - full->plane_infos[planes_count].stereo_format = dc_plane->stereo_format; - full->plane_infos[planes_count].tiling_info = dc_plane->tiling_info; - full->plane_infos[planes_count].visible = dc_plane->visible; - full->plane_infos[planes_count].per_pixel_alpha = dc_plane->per_pixel_alpha; - full->plane_infos[planes_count].dcc = dc_plane->dcc; - full->surface_updates[planes_count].plane_info = &full->plane_infos[planes_count]; + pflip_present = pflip_present || plane_needs_flip; + + if (!plane_needs_flip) { + planes_count += 1; + continue; + } + + /* + * TODO This might fail and hence better not used, wait + * explicitly on fences instead + * and in general should be called for + * blocking commit to as per framework helpers + */ + abo = gem_to_amdgpu_bo(fb->obj[0]); + r = amdgpu_bo_reserve(abo, true); + if (unlikely(r != 0)) { + DRM_ERROR("failed to reserve buffer before flip\n"); + WARN_ON(1); + } + + /* Wait for all fences on this FB */ + WARN_ON(reservation_object_wait_timeout_rcu(abo->tbo.resv, true, false, + MAX_SCHEDULE_TIMEOUT) < 0); + + amdgpu_bo_get_tiling_flags(abo, &tiling_flags); + + amdgpu_bo_unreserve(abo); + + fill_dc_plane_info_and_addr( + dm->adev, new_plane_state, tiling_flags, + &bundle->plane_infos[planes_count], + &bundle->flip_addrs[planes_count].address); + + bundle->surface_updates[planes_count].plane_info = + &bundle->plane_infos[planes_count]; + + bundle->flip_addrs[planes_count].flip_immediate = + (crtc->state->pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC) != 0; + + timestamp_ns = ktime_get_ns(); + bundle->flip_addrs[planes_count].flip_timestamp_in_us = div_u64(timestamp_ns, 1000); + bundle->surface_updates[planes_count].flip_addr = &bundle->flip_addrs[planes_count]; + bundle->surface_updates[planes_count].surface = dc_plane; + + if (!bundle->surface_updates[planes_count].surface) { + DRM_ERROR("No surface for CRTC: id=%d\n", + acrtc_attach->crtc_id); + continue; + } + + if (plane == pcrtc->primary) + update_freesync_state_on_stream( + dm, + acrtc_state, + acrtc_state->stream, + dc_plane, + bundle->flip_addrs[planes_count].flip_timestamp_in_us); + + DRM_DEBUG_DRIVER("%s Flipping to hi: 0x%x, low: 0x%x\n", + __func__, + bundle->flip_addrs[planes_count].address.grph.addr.high_part, + bundle->flip_addrs[planes_count].address.grph.addr.low_part); planes_count += 1; } - /* - * TODO: For proper atomic behaviour, we should be calling into DC once with - * all the changes. However, DC refuses to do pageflips and non-pageflip - * changes in the same call. Change DC to respect atomic behaviour, - * hopefully eliminating dc_*_update structs in their entirety. - */ - if (flip_count) { + if (pflip_present) { if (!vrr_active) { /* Use old throttling in non-vrr fixed refresh rate mode * to keep flip scheduling based on target vblank counts @@ -4840,7 +5273,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, * clients using the GLX_OML_sync_control extension or * DRI3/Present extension with defined target_msc. */ - last_flip_vblank = drm_crtc_vblank_count(pcrtc); + last_flip_vblank = amdgpu_get_vblank_counter_kms(dm->ddev, acrtc_attach->crtc_id); } else { /* For variable refresh rate mode only: @@ -4856,11 +5289,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, spin_unlock_irqrestore(&pcrtc->dev->event_lock, flags); } - target = (uint32_t)last_flip_vblank + *wait_for_vblank; - - /* Prepare wait for target vblank early - before the fence-waits */ - target_vblank = target - (uint32_t)drm_crtc_vblank_count(pcrtc) + - amdgpu_get_vblank_counter_kms(pcrtc->dev, acrtc_attach->crtc_id); + target_vblank = last_flip_vblank + wait_for_vblank; /* * Wait until we're out of the vertical blank period before the one @@ -4891,43 +5320,34 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, if (acrtc_state->stream) { if (acrtc_state->freesync_timing_changed) - flip->stream_update.adjust = + bundle->stream_update.adjust = &acrtc_state->stream->adjust; if (acrtc_state->freesync_vrr_info_changed) - flip->stream_update.vrr_infopacket = + bundle->stream_update.vrr_infopacket = &acrtc_state->stream->vrr_infopacket; } - - mutex_lock(&dm->dc_lock); - dc_commit_updates_for_stream(dm->dc, - flip->surface_updates, - flip_count, - acrtc_state->stream, - &flip->stream_update, - dc_state); - mutex_unlock(&dm->dc_lock); } if (planes_count) { if (new_pcrtc_state->mode_changed) { - full->stream_update.src = acrtc_state->stream->src; - full->stream_update.dst = acrtc_state->stream->dst; + bundle->stream_update.src = acrtc_state->stream->src; + bundle->stream_update.dst = acrtc_state->stream->dst; } if (new_pcrtc_state->color_mgmt_changed) - full->stream_update.out_transfer_func = acrtc_state->stream->out_transfer_func; + bundle->stream_update.out_transfer_func = acrtc_state->stream->out_transfer_func; acrtc_state->stream->abm_level = acrtc_state->abm_level; if (acrtc_state->abm_level != dm_old_crtc_state->abm_level) - full->stream_update.abm_level = &acrtc_state->abm_level; + bundle->stream_update.abm_level = &acrtc_state->abm_level; mutex_lock(&dm->dc_lock); dc_commit_updates_for_stream(dm->dc, - full->surface_updates, + bundle->surface_updates, planes_count, acrtc_state->stream, - &full->stream_update, + &bundle->stream_update, dc_state); mutex_unlock(&dm->dc_lock); } @@ -4937,8 +5357,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, handle_cursor_update(plane, old_plane_state); cleanup: - kfree(flip); - kfree(full); + kfree(bundle); } /* @@ -4952,8 +5371,7 @@ cleanup: static void amdgpu_dm_crtc_copy_transient_flags(struct drm_crtc_state *crtc_state, struct dc_stream_state *stream_state) { - stream_state->mode_changed = - crtc_state->mode_changed || crtc_state->active_changed; + stream_state->mode_changed = drm_atomic_crtc_needs_modeset(crtc_state); } static int amdgpu_dm_atomic_commit(struct drm_device *dev, @@ -5036,7 +5454,7 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) dc_state = dm_state->context; } else { /* No state changes, retain current state. */ - dc_state_temp = dc_create_state(); + dc_state_temp = dc_create_state(dm->dc); ASSERT(dc_state_temp); dc_state = dc_state_temp; dc_resource_state_copy_construct_current(dm->dc, dc_state); @@ -5205,6 +5623,12 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) mutex_unlock(&dm->dc_lock); } + /* Update freesync state before amdgpu_dm_handle_vrr_transition(). */ + for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) { + dm_new_crtc_state = to_dm_crtc_state(new_crtc_state); + pre_update_freesync_state_on_stream(dm, dm_new_crtc_state); + } + for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) { /* @@ -5218,6 +5642,11 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) dm_new_crtc_state = to_dm_crtc_state(new_crtc_state); dm_old_crtc_state = to_dm_crtc_state(old_crtc_state); + + /* Handle vrr on->off / off->on transitions */ + amdgpu_dm_handle_vrr_transition(dm_old_crtc_state, + dm_new_crtc_state); + modeset_needed = modeset_required( new_crtc_state, dm_new_crtc_state->stream, @@ -5235,13 +5664,17 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) #endif } + for_each_new_crtc_in_state(state, crtc, new_crtc_state, j) + if (new_crtc_state->pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC) + wait_for_vblank = false; + /* update planes when needed per crtc*/ for_each_new_crtc_in_state(state, crtc, new_crtc_state, j) { dm_new_crtc_state = to_dm_crtc_state(new_crtc_state); if (dm_new_crtc_state->stream) amdgpu_dm_commit_planes(state, dc_state, dev, - dm, crtc, &wait_for_vblank); + dm, crtc, wait_for_vblank); } @@ -5429,9 +5862,11 @@ static void get_freesync_config_for_crtc( struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(new_con_state->base.connector); struct drm_display_mode *mode = &new_crtc_state->base.mode; + int vrefresh = drm_mode_vrefresh(mode); new_crtc_state->vrr_supported = new_con_state->freesync_capable && - aconnector->min_vfreq <= drm_mode_vrefresh(mode); + vrefresh >= aconnector->min_vfreq && + vrefresh <= aconnector->max_vfreq; if (new_crtc_state->vrr_supported) { new_crtc_state->stream->ignore_msa_timing_param = true; @@ -5481,21 +5916,12 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm, struct amdgpu_dm_connector *aconnector = NULL; struct drm_connector_state *drm_new_conn_state = NULL, *drm_old_conn_state = NULL; struct dm_connector_state *dm_new_conn_state = NULL, *dm_old_conn_state = NULL; - struct drm_plane_state *new_plane_state = NULL; new_stream = NULL; dm_old_crtc_state = to_dm_crtc_state(old_crtc_state); dm_new_crtc_state = to_dm_crtc_state(new_crtc_state); acrtc = to_amdgpu_crtc(crtc); - - new_plane_state = drm_atomic_get_new_plane_state(state, new_crtc_state->crtc->primary); - - if (new_crtc_state->enable && new_plane_state && !new_plane_state->fb) { - ret = -EINVAL; - goto fail; - } - aconnector = amdgpu_dm_find_first_crtc_matching_connector(state, crtc); /* TODO This hack should go away */ @@ -5658,6 +6084,9 @@ skip_modeset: update_stream_scaling_settings( &new_crtc_state->mode, dm_new_conn_state, dm_new_crtc_state->stream); + /* ABM settings */ + dm_new_crtc_state->abm_level = dm_new_conn_state->abm_level; + /* * Color management settings. We also update color properties * when a modeset is needed, to ensure it gets reprogrammed. @@ -5682,6 +6111,69 @@ fail: return ret; } +static bool should_reset_plane(struct drm_atomic_state *state, + struct drm_plane *plane, + struct drm_plane_state *old_plane_state, + struct drm_plane_state *new_plane_state) +{ + struct drm_plane *other; + struct drm_plane_state *old_other_state, *new_other_state; + struct drm_crtc_state *new_crtc_state; + int i; + + /* + * TODO: Remove this hack once the checks below are sufficient + * enough to determine when we need to reset all the planes on + * the stream. + */ + if (state->allow_modeset) + return true; + + /* Exit early if we know that we're adding or removing the plane. */ + if (old_plane_state->crtc != new_plane_state->crtc) + return true; + + /* old crtc == new_crtc == NULL, plane not in context. */ + if (!new_plane_state->crtc) + return false; + + new_crtc_state = + drm_atomic_get_new_crtc_state(state, new_plane_state->crtc); + + if (!new_crtc_state) + return true; + + if (drm_atomic_crtc_needs_modeset(new_crtc_state)) + return true; + + /* + * If there are any new primary or overlay planes being added or + * removed then the z-order can potentially change. To ensure + * correct z-order and pipe acquisition the current DC architecture + * requires us to remove and recreate all existing planes. + * + * TODO: Come up with a more elegant solution for this. + */ + for_each_oldnew_plane_in_state(state, other, old_other_state, new_other_state, i) { + if (other->type == DRM_PLANE_TYPE_CURSOR) + continue; + + if (old_other_state->crtc != new_plane_state->crtc && + new_other_state->crtc != new_plane_state->crtc) + continue; + + if (old_other_state->crtc != new_other_state->crtc) + return true; + + /* TODO: Remove this once we can handle fast format changes. */ + if (old_other_state->fb && new_other_state->fb && + old_other_state->fb->format != new_other_state->fb->format) + return true; + } + + return false; +} + static int dm_update_plane_state(struct dc *dc, struct drm_atomic_state *state, struct drm_plane *plane, @@ -5696,8 +6188,7 @@ static int dm_update_plane_state(struct dc *dc, struct drm_crtc_state *old_crtc_state, *new_crtc_state; struct dm_crtc_state *dm_new_crtc_state, *dm_old_crtc_state; struct dm_plane_state *dm_new_plane_state, *dm_old_plane_state; - /* TODO return page_flip_needed() function */ - bool pflip_needed = !state->allow_modeset; + bool needs_reset; int ret = 0; @@ -5710,10 +6201,12 @@ static int dm_update_plane_state(struct dc *dc, if (plane->type == DRM_PLANE_TYPE_CURSOR) return 0; + needs_reset = should_reset_plane(state, plane, old_plane_state, + new_plane_state); + /* Remove any changed/removed planes */ if (!enable) { - if (pflip_needed && - plane->type != DRM_PLANE_TYPE_OVERLAY) + if (!needs_reset) return 0; if (!old_plane_crtc) @@ -5764,7 +6257,7 @@ static int dm_update_plane_state(struct dc *dc, if (!dm_new_crtc_state->stream) return 0; - if (pflip_needed && plane->type != DRM_PLANE_TYPE_OVERLAY) + if (!needs_reset) return 0; WARN_ON(dm_new_plane_state->dc_state); @@ -5776,7 +6269,7 @@ static int dm_update_plane_state(struct dc *dc, DRM_DEBUG_DRIVER("Enabling DRM plane: %d on DRM crtc %d\n", plane->base.id, new_plane_crtc->base.id); - ret = fill_plane_attributes( + ret = fill_dc_plane_attributes( new_plane_crtc->dev->dev_private, dc_new_plane_state, new_plane_state, @@ -5824,10 +6317,11 @@ static int dm_update_plane_state(struct dc *dc, } static int -dm_determine_update_type_for_commit(struct dc *dc, +dm_determine_update_type_for_commit(struct amdgpu_display_manager *dm, struct drm_atomic_state *state, enum surface_update_type *out_type) { + struct dc *dc = dm->dc; struct dm_atomic_state *dm_state = NULL, *old_dm_state = NULL; int i, j, num_plane, ret = 0; struct drm_plane_state *old_plane_state, *new_plane_state; @@ -5841,21 +6335,22 @@ dm_determine_update_type_for_commit(struct dc *dc, struct dc_stream_status *status = NULL; struct dc_surface_update *updates; - struct dc_plane_state *surface; enum surface_update_type update_type = UPDATE_TYPE_FAST; updates = kcalloc(MAX_SURFACES, sizeof(*updates), GFP_KERNEL); - surface = kcalloc(MAX_SURFACES, sizeof(*surface), GFP_KERNEL); - if (!updates || !surface) { - DRM_ERROR("Plane or surface update failed to allocate"); + if (!updates) { + DRM_ERROR("Failed to allocate plane updates\n"); /* Set type to FULL to avoid crashing in DC*/ update_type = UPDATE_TYPE_FULL; goto cleanup; } for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) { - struct dc_stream_update stream_update = { 0 }; + struct dc_scaling_info scaling_info; + struct dc_stream_update stream_update; + + memset(&stream_update, 0, sizeof(stream_update)); new_dm_crtc_state = to_dm_crtc_state(new_crtc_state); old_dm_crtc_state = to_dm_crtc_state(old_crtc_state); @@ -5883,23 +6378,12 @@ dm_determine_update_type_for_commit(struct dc *dc, goto cleanup; } - if (!state->allow_modeset) - continue; - if (crtc != new_plane_crtc) continue; - updates[num_plane].surface = &surface[num_plane]; + updates[num_plane].surface = new_dm_plane_state->dc_state; if (new_crtc_state->mode_changed) { - updates[num_plane].surface->src_rect = - new_dm_plane_state->dc_state->src_rect; - updates[num_plane].surface->dst_rect = - new_dm_plane_state->dc_state->dst_rect; - updates[num_plane].surface->rotation = - new_dm_plane_state->dc_state->rotation; - updates[num_plane].surface->in_transfer_func = - new_dm_plane_state->dc_state->in_transfer_func; stream_update.dst = new_dm_crtc_state->stream->dst; stream_update.src = new_dm_crtc_state->stream->src; } @@ -5915,6 +6399,13 @@ dm_determine_update_type_for_commit(struct dc *dc, new_dm_crtc_state->stream->out_transfer_func; } + ret = fill_dc_scaling_info(new_plane_state, + &scaling_info); + if (ret) + goto cleanup; + + updates[num_plane].scaling_info = &scaling_info; + num_plane++; } @@ -5934,8 +6425,14 @@ dm_determine_update_type_for_commit(struct dc *dc, status = dc_stream_get_status_from_state(old_dm_state->context, new_dm_crtc_state->stream); + /* + * TODO: DC modifies the surface during this call so we need + * to lock here - find a way to do this without locking. + */ + mutex_lock(&dm->dc_lock); update_type = dc_check_update_surfaces_for_stream(dc, updates, num_plane, &stream_update, status); + mutex_unlock(&dm->dc_lock); if (update_type > UPDATE_TYPE_MED) { update_type = UPDATE_TYPE_FULL; @@ -5945,7 +6442,6 @@ dm_determine_update_type_for_commit(struct dc *dc, cleanup: kfree(updates); - kfree(surface); *out_type = update_type; return ret; @@ -6129,7 +6625,7 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev, lock_and_validation_needed = true; } - ret = dm_determine_update_type_for_commit(dc, state, &update_type); + ret = dm_determine_update_type_for_commit(&adev->dm, state, &update_type); if (ret) goto fail; @@ -6144,9 +6640,6 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev, */ if (lock_and_validation_needed && overall_update_type <= UPDATE_TYPE_FAST) WARN(1, "Global lock should be Set, overall_update_type should be UPDATE_TYPE_MED or UPDATE_TYPE_FULL"); - else if (!lock_and_validation_needed && overall_update_type > UPDATE_TYPE_FAST) - WARN(1, "Global lock should NOT be set, overall_update_type should be UPDATE_TYPE_FAST"); - if (overall_update_type > UPDATE_TYPE_FAST) { ret = dm_atomic_get_state(state, &dm_state); @@ -6157,7 +6650,7 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev, if (ret) goto fail; - if (dc_validate_global_state(dc, dm_state->context) != DC_OK) { + if (dc_validate_global_state(dc, dm_state->context, false) != DC_OK) { ret = -EINVAL; goto fail; } diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h index fbd161ddc3f4..3a0b6164c755 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -132,8 +132,6 @@ struct amdgpu_display_manager { */ struct drm_private_obj atomic_obj; - struct drm_modeset_lock atomic_obj_lock; - /** * @dc_lock: * @@ -184,6 +182,15 @@ struct amdgpu_display_manager { struct common_irq_params vblank_params[DC_IRQ_SOURCE_VBLANK6 - DC_IRQ_SOURCE_VBLANK1 + 1]; + /** + * @vupdate_params: + * + * Vertical update IRQ parameters, passed to registered handlers when + * triggered. + */ + struct common_irq_params + vupdate_params[DC_IRQ_SOURCE_VUPDATE6 - DC_IRQ_SOURCE_VUPDATE1 + 1]; + spinlock_t irq_handler_list_table_lock; struct backlight_device *backlight_dev; @@ -240,6 +247,10 @@ struct amdgpu_dm_connector { struct mutex hpd_lock; bool fake_enable; +#ifdef CONFIG_DEBUG_FS + uint32_t debugfs_dpcd_address; + uint32_t debugfs_dpcd_size; +#endif }; #define to_amdgpu_dm_connector(x) container_of(x, struct amdgpu_dm_connector, base) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index 216e48cec716..7258c992a2bf 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -126,46 +126,51 @@ int amdgpu_dm_set_regamma_lut(struct dm_crtc_state *crtc) crtc->base.state->dev->dev_private; struct drm_color_lut *lut; uint32_t lut_size; - struct dc_gamma *gamma; + struct dc_gamma *gamma = NULL; enum dc_transfer_func_type old_type = stream->out_transfer_func->type; bool ret; - if (!blob) { + if (!blob && adev->asic_type <= CHIP_RAVEN) { /* By default, use the SRGB predefined curve.*/ stream->out_transfer_func->type = TF_TYPE_PREDEFINED; stream->out_transfer_func->tf = TRANSFER_FUNCTION_SRGB; return 0; } - lut = (struct drm_color_lut *)blob->data; - lut_size = blob->length / sizeof(struct drm_color_lut); + if (blob) { + lut = (struct drm_color_lut *)blob->data; + lut_size = blob->length / sizeof(struct drm_color_lut); - gamma = dc_create_gamma(); - if (!gamma) - return -ENOMEM; + gamma = dc_create_gamma(); + if (!gamma) + return -ENOMEM; - gamma->num_entries = lut_size; - if (gamma->num_entries == MAX_COLOR_LEGACY_LUT_ENTRIES) - gamma->type = GAMMA_RGB_256; - else if (gamma->num_entries == MAX_COLOR_LUT_ENTRIES) - gamma->type = GAMMA_CS_TFM_1D; - else { - /* Invalid lut size */ - dc_gamma_release(&gamma); - return -EINVAL; + gamma->num_entries = lut_size; + if (gamma->num_entries == MAX_COLOR_LEGACY_LUT_ENTRIES) + gamma->type = GAMMA_RGB_256; + else if (gamma->num_entries == MAX_COLOR_LUT_ENTRIES) + gamma->type = GAMMA_CS_TFM_1D; + else { + /* Invalid lut size */ + dc_gamma_release(&gamma); + return -EINVAL; + } + + /* Convert drm_lut into dc_gamma */ + __drm_lut_to_dc_gamma(lut, gamma, gamma->type == GAMMA_RGB_256); } - /* Convert drm_lut into dc_gamma */ - __drm_lut_to_dc_gamma(lut, gamma, gamma->type == GAMMA_RGB_256); - - /* Call color module to translate into something DC understands. Namely - * a transfer function. + /* predefined gamma ROM only exist for RAVEN and pre-RAVEN ASIC, + * set canRomBeUsed accordingly */ stream->out_transfer_func->type = TF_TYPE_DISTRIBUTED_POINTS; ret = mod_color_calculate_regamma_params(stream->out_transfer_func, - gamma, true, adev->asic_type <= CHIP_RAVEN, NULL); - dc_gamma_release(&gamma); + gamma, true, adev->asic_type <= CHIP_RAVEN, NULL); + + if (gamma) + dc_gamma_release(&gamma); + if (!ret) { stream->out_transfer_func->type = old_type; DRM_ERROR("Out of memory when calculating regamma params\n"); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c index 4a55cde027cf..1d5fc5ad3bee 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c @@ -29,6 +29,7 @@ #include "amdgpu.h" #include "amdgpu_dm.h" #include "amdgpu_dm_debugfs.h" +#include "dm_helpers.h" /* function description * get/ set DP configuration: lane_count, link_rate, spread_spectrum @@ -688,8 +689,131 @@ static int vrr_range_show(struct seq_file *m, void *data) return 0; } + +/* function description + * + * generic SDP message access for testing + * + * debugfs sdp_message is located at /syskernel/debug/dri/0/DP-x + * + * SDP header + * Hb0 : Secondary-Data Packet ID + * Hb1 : Secondary-Data Packet type + * Hb2 : Secondary-Data-packet-specific header, Byte 0 + * Hb3 : Secondary-Data-packet-specific header, Byte 1 + * + * for using custom sdp message: input 4 bytes SDP header and 32 bytes raw data + */ +static ssize_t dp_sdp_message_debugfs_write(struct file *f, const char __user *buf, + size_t size, loff_t *pos) +{ + int r; + uint8_t data[36]; + struct amdgpu_dm_connector *connector = file_inode(f)->i_private; + struct dm_crtc_state *acrtc_state; + uint32_t write_size = 36; + + if (connector->base.status != connector_status_connected) + return -ENODEV; + + if (size == 0) + return 0; + + acrtc_state = to_dm_crtc_state(connector->base.state->crtc->state); + + r = copy_from_user(data, buf, write_size); + + write_size -= r; + + dc_stream_send_dp_sdp(acrtc_state->stream, data, write_size); + + return write_size; +} + DEFINE_SHOW_ATTRIBUTE(vrr_range); +static ssize_t dp_dpcd_address_write(struct file *f, const char __user *buf, + size_t size, loff_t *pos) +{ + int r; + struct amdgpu_dm_connector *connector = file_inode(f)->i_private; + + if (size < sizeof(connector->debugfs_dpcd_address)) + return 0; + + r = copy_from_user(&connector->debugfs_dpcd_address, + buf, sizeof(connector->debugfs_dpcd_address)); + + return size - r; +} + +static ssize_t dp_dpcd_size_write(struct file *f, const char __user *buf, + size_t size, loff_t *pos) +{ + int r; + struct amdgpu_dm_connector *connector = file_inode(f)->i_private; + + if (size < sizeof(connector->debugfs_dpcd_size)) + return 0; + + r = copy_from_user(&connector->debugfs_dpcd_size, + buf, sizeof(connector->debugfs_dpcd_size)); + + if (connector->debugfs_dpcd_size > 256) + connector->debugfs_dpcd_size = 0; + + return size - r; +} + +static ssize_t dp_dpcd_data_write(struct file *f, const char __user *buf, + size_t size, loff_t *pos) +{ + int r; + char *data; + struct amdgpu_dm_connector *connector = file_inode(f)->i_private; + struct dc_link *link = connector->dc_link; + uint32_t write_size = connector->debugfs_dpcd_size; + + if (size < write_size) + return 0; + + data = kzalloc(write_size, GFP_KERNEL); + if (!data) + return 0; + + r = copy_from_user(data, buf, write_size); + + dm_helpers_dp_write_dpcd(link->ctx, link, + connector->debugfs_dpcd_address, data, write_size - r); + kfree(data); + return write_size - r; +} + +static ssize_t dp_dpcd_data_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) +{ + int r; + char *data; + struct amdgpu_dm_connector *connector = file_inode(f)->i_private; + struct dc_link *link = connector->dc_link; + uint32_t read_size = connector->debugfs_dpcd_size; + + if (size < read_size) + return 0; + + data = kzalloc(read_size, GFP_KERNEL); + if (!data) + return 0; + + dm_helpers_dp_read_dpcd(link->ctx, link, + connector->debugfs_dpcd_address, data, read_size); + + r = copy_to_user(buf, data, read_size); + + kfree(data); + return read_size - r; +} + static const struct file_operations dp_link_settings_debugfs_fops = { .owner = THIS_MODULE, .read = dp_link_settings_read, @@ -710,6 +834,31 @@ static const struct file_operations dp_phy_test_pattern_fops = { .llseek = default_llseek }; +static const struct file_operations sdp_message_fops = { + .owner = THIS_MODULE, + .write = dp_sdp_message_debugfs_write, + .llseek = default_llseek +}; + +static const struct file_operations dp_dpcd_address_debugfs_fops = { + .owner = THIS_MODULE, + .write = dp_dpcd_address_write, + .llseek = default_llseek +}; + +static const struct file_operations dp_dpcd_size_debugfs_fops = { + .owner = THIS_MODULE, + .write = dp_dpcd_size_write, + .llseek = default_llseek +}; + +static const struct file_operations dp_dpcd_data_debugfs_fops = { + .owner = THIS_MODULE, + .read = dp_dpcd_data_read, + .write = dp_dpcd_data_write, + .llseek = default_llseek +}; + static const struct { char *name; const struct file_operations *fops; @@ -717,7 +866,11 @@ static const struct { {"link_settings", &dp_link_settings_debugfs_fops}, {"phy_settings", &dp_phy_settings_debugfs_fop}, {"test_pattern", &dp_phy_test_pattern_fops}, - {"vrr_range", &vrr_range_fops} + {"vrr_range", &vrr_range_fops}, + {"sdp_message", &sdp_message_fops}, + {"aux_dpcd_address", &dp_dpcd_address_debugfs_fops}, + {"aux_dpcd_size", &dp_dpcd_size_debugfs_fops}, + {"aux_dpcd_data", &dp_dpcd_data_debugfs_fops} }; int connector_debugfs_init(struct amdgpu_dm_connector *connector) @@ -842,6 +995,35 @@ static const struct drm_info_list amdgpu_dm_debugfs_list[] = { {"amdgpu_target_backlight_pwm", &target_backlight_read}, }; +/* + * Sets the DC visual confirm debug option from the given string. + * Example usage: echo 1 > /sys/kernel/debug/dri/0/amdgpu_visual_confirm + */ +static int visual_confirm_set(void *data, u64 val) +{ + struct amdgpu_device *adev = data; + + adev->dm.dc->debug.visual_confirm = (enum visual_confirm)val; + + return 0; +} + +/* + * Reads the DC visual confirm debug option value into the given buffer. + * Example usage: cat /sys/kernel/debug/dri/0/amdgpu_dm_visual_confirm + */ +static int visual_confirm_get(void *data, u64 *val) +{ + struct amdgpu_device *adev = data; + + *val = adev->dm.dc->debug.visual_confirm; + + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(visual_confirm_fops, visual_confirm_get, + visual_confirm_set, "%llu\n"); + int dtn_debugfs_init(struct amdgpu_device *adev) { static const struct file_operations dtn_log_fops = { @@ -867,5 +1049,13 @@ int dtn_debugfs_init(struct amdgpu_device *adev) adev, &dtn_log_fops); - return PTR_ERR_OR_ZERO(ent); + if (IS_ERR(ent)) + return PTR_ERR(ent); + + ent = debugfs_create_file_unsafe("amdgpu_dm_visual_confirm", 0644, root, + adev, &visual_confirm_fops); + if (IS_ERR(ent)) + return PTR_ERR(ent); + + return 0; } diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c index b39766bd2840..e6cd67342df8 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c @@ -264,7 +264,7 @@ bool dm_helpers_dp_mst_write_payload_allocation_table( } /* - * poll pending down reply before clear payload allocation table + * poll pending down reply */ void dm_helpers_dp_mst_poll_pending_down_reply( struct dc_context *ctx, diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c index cd10f77cdeb0..fd22b4474dbf 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c @@ -674,11 +674,30 @@ static int amdgpu_dm_set_crtc_irq_state(struct amdgpu_device *adev, __func__); } +static int amdgpu_dm_set_vupdate_irq_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + unsigned int crtc_id, + enum amdgpu_interrupt_state state) +{ + return dm_irq_state( + adev, + source, + crtc_id, + state, + IRQ_TYPE_VUPDATE, + __func__); +} + static const struct amdgpu_irq_src_funcs dm_crtc_irq_funcs = { .set = amdgpu_dm_set_crtc_irq_state, .process = amdgpu_dm_irq_handler, }; +static const struct amdgpu_irq_src_funcs dm_vupdate_irq_funcs = { + .set = amdgpu_dm_set_vupdate_irq_state, + .process = amdgpu_dm_irq_handler, +}; + static const struct amdgpu_irq_src_funcs dm_pageflip_irq_funcs = { .set = amdgpu_dm_set_pflip_irq_state, .process = amdgpu_dm_irq_handler, @@ -695,6 +714,9 @@ void amdgpu_dm_set_irq_funcs(struct amdgpu_device *adev) adev->crtc_irq.num_types = adev->mode_info.num_crtc; adev->crtc_irq.funcs = &dm_crtc_irq_funcs; + adev->vupdate_irq.num_types = adev->mode_info.num_crtc; + adev->vupdate_irq.funcs = &dm_vupdate_irq_funcs; + adev->pageflip_irq.num_types = adev->mode_info.num_crtc; adev->pageflip_irq.funcs = &dm_pageflip_irq_funcs; diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c index c4ea3a91f17a..6e205ee36ac3 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c @@ -84,6 +84,7 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux, { ssize_t result = 0; struct aux_payload payload; + enum aux_channel_operation_result operation_result; if (WARN_ON(msg->size > 16)) return -E2BIG; @@ -97,13 +98,27 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux, payload.mot = (msg->request & DP_AUX_I2C_MOT) != 0; payload.defer_delay = 0; - result = dc_link_aux_transfer(TO_DM_AUX(aux)->ddc_service, &payload); + result = dc_link_aux_transfer_raw(TO_DM_AUX(aux)->ddc_service, &payload, + &operation_result); if (payload.write) result = msg->size; - if (result < 0) /* DC doesn't know about kernel error codes */ - result = -EIO; + if (result < 0) + switch (operation_result) { + case AUX_CHANNEL_OPERATION_SUCCEEDED: + break; + case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON: + case AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN: + result = -EIO; + break; + case AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY: + result = -EBUSY; + break; + case AUX_CHANNEL_OPERATION_FAILED_TIMEOUT: + result = -ETIMEDOUT; + break; + } return result; } diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c index a114954d6a5b..350e7a620d45 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c @@ -33,6 +33,7 @@ #include "amdgpu_dm_irq.h" #include "amdgpu_pm.h" #include "dm_pp_smu.h" +#include "amdgpu_smu.h" bool dm_pp_apply_display_requirements( @@ -40,6 +41,7 @@ bool dm_pp_apply_display_requirements( const struct dm_pp_display_configuration *pp_display_cfg) { struct amdgpu_device *adev = ctx->driver_context; + struct smu_context *smu = &adev->smu; int i; if (adev->pm.dpm_enabled) { @@ -105,6 +107,9 @@ bool dm_pp_apply_display_requirements( adev->powerplay.pp_funcs->display_configuration_change( adev->powerplay.pp_handle, &adev->pm.pm_display_cfg); + else + smu_display_configuration_change(smu, + &adev->pm.pm_display_cfg); amdgpu_pm_compute_clocks(adev); } @@ -308,6 +313,12 @@ bool dm_pp_get_clock_levels_by_type( if (adev->powerplay.pp_funcs->get_clock_by_type(pp_handle, dc_to_pp_clock_type(clk_type), &pp_clks)) { /* Error in pplib. Provide default values. */ + return true; + } + } else if (adev->smu.funcs && adev->smu.funcs->get_clock_by_type) { + if (smu_get_clock_by_type(&adev->smu, + dc_to_pp_clock_type(clk_type), + &pp_clks)) { get_default_clock_levels(clk_type, dc_clks); return true; } @@ -324,6 +335,13 @@ bool dm_pp_get_clock_levels_by_type( validation_clks.memory_max_clock = 80000; validation_clks.level = 0; } + } else if (adev->smu.funcs && adev->smu.funcs->get_max_high_clocks) { + if (smu_get_max_high_clocks(&adev->smu, &validation_clks)) { + DRM_INFO("DM_PPLIB: Warning: using default validation clocks!\n"); + validation_clks.engine_max_clock = 72000; + validation_clks.memory_max_clock = 80000; + validation_clks.level = 0; + } } DRM_INFO("DM_PPLIB: Validation clocks:\n"); @@ -374,14 +392,21 @@ bool dm_pp_get_clock_levels_by_type_with_latency( void *pp_handle = adev->powerplay.pp_handle; struct pp_clock_levels_with_latency pp_clks = { 0 }; const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; + int ret; - if (!pp_funcs || !pp_funcs->get_clock_by_type_with_latency) - return false; + if (pp_funcs && pp_funcs->get_clock_by_type_with_latency) { + ret = pp_funcs->get_clock_by_type_with_latency(pp_handle, + dc_to_pp_clock_type(clk_type), + &pp_clks); + if (ret) + return false; + } else if (adev->smu.ppt_funcs && adev->smu.ppt_funcs->get_clock_by_type_with_latency) { + if (smu_get_clock_by_type_with_latency(&adev->smu, + dc_to_pp_clock_type(clk_type), + &pp_clks)) + return false; + } - if (pp_funcs->get_clock_by_type_with_latency(pp_handle, - dc_to_pp_clock_type(clk_type), - &pp_clks)) - return false; pp_to_dc_clock_levels_with_latency(&pp_clks, clk_level_info, clk_type); @@ -397,14 +422,20 @@ bool dm_pp_get_clock_levels_by_type_with_voltage( void *pp_handle = adev->powerplay.pp_handle; struct pp_clock_levels_with_voltage pp_clk_info = {0}; const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; + int ret; - if (!pp_funcs || !pp_funcs->get_clock_by_type_with_voltage) - return false; - - if (pp_funcs->get_clock_by_type_with_voltage(pp_handle, - dc_to_pp_clock_type(clk_type), - &pp_clk_info)) - return false; + if (pp_funcs && pp_funcs->get_clock_by_type_with_voltage) { + ret = pp_funcs->get_clock_by_type_with_voltage(pp_handle, + dc_to_pp_clock_type(clk_type), + &pp_clk_info); + if (ret) + return false; + } else if (adev->smu.ppt_funcs && adev->smu.ppt_funcs->get_clock_by_type_with_voltage) { + if (smu_get_clock_by_type_with_voltage(&adev->smu, + dc_to_pp_clock_type(clk_type), + &pp_clk_info)) + return false; + } pp_to_dc_clock_levels_with_voltage(&pp_clk_info, clk_level_info, clk_type); @@ -445,6 +476,10 @@ bool dm_pp_apply_clock_for_voltage_request( ret = adev->powerplay.pp_funcs->display_clock_voltage_request( adev->powerplay.pp_handle, &pp_clock_request); + else if (adev->smu.funcs && + adev->smu.funcs->display_clock_voltage_request) + ret = smu_display_clock_voltage_request(&adev->smu, + &pp_clock_request); if (ret) return false; return true; @@ -462,6 +497,8 @@ bool dm_pp_get_static_clocks( ret = adev->powerplay.pp_funcs->get_current_clocks( adev->powerplay.pp_handle, &pp_clk_info); + else if (adev->smu.funcs) + ret = smu_get_current_clocks(&adev->smu, &pp_clk_info); if (ret) return false; @@ -472,27 +509,6 @@ bool dm_pp_get_static_clocks( return true; } -void pp_rv_set_display_requirement(struct pp_smu *pp, - struct pp_smu_display_requirement_rv *req) -{ - const struct dc_context *ctx = pp->dm; - struct amdgpu_device *adev = ctx->driver_context; - void *pp_handle = adev->powerplay.pp_handle; - const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; - struct pp_display_clock_request clock = {0}; - - if (!pp_funcs || !pp_funcs->display_clock_voltage_request) - return; - - clock.clock_type = amd_pp_dcf_clock; - clock.clock_freq_in_khz = req->hard_min_dcefclk_mhz * 1000; - pp_funcs->display_clock_voltage_request(pp_handle, &clock); - - clock.clock_type = amd_pp_f_clock; - clock.clock_freq_in_khz = req->hard_min_fclk_mhz * 1000; - pp_funcs->display_clock_voltage_request(pp_handle, &clock); -} - void pp_rv_set_wm_ranges(struct pp_smu *pp, struct pp_smu_wm_range_sets *ranges) { @@ -508,9 +524,6 @@ void pp_rv_set_wm_ranges(struct pp_smu *pp, wm_with_clock_ranges.num_wm_dmif_sets = ranges->num_reader_wm_sets; wm_with_clock_ranges.num_wm_mcif_sets = ranges->num_writer_wm_sets; - if (!pp_funcs || !pp_funcs->set_watermarks_for_clocks_ranges) - return; - for (i = 0; i < wm_with_clock_ranges.num_wm_dmif_sets; i++) { if (ranges->reader_wm_sets[i].wm_inst > 3) wm_dce_clocks[i].wm_set_id = WM_SET_A; @@ -543,7 +556,13 @@ void pp_rv_set_wm_ranges(struct pp_smu *pp, ranges->writer_wm_sets[i].min_drain_clk_mhz * 1000; } - pp_funcs->set_watermarks_for_clocks_ranges(pp_handle, &wm_with_clock_ranges); + if (pp_funcs && pp_funcs->set_watermarks_for_clocks_ranges) + pp_funcs->set_watermarks_for_clocks_ranges(pp_handle, + &wm_with_clock_ranges); + else if (adev->smu.funcs && + adev->smu.funcs->set_watermarks_for_clock_ranges) + smu_set_watermarks_for_clock_ranges(&adev->smu, + &wm_with_clock_ranges); } void pp_rv_set_pme_wa_enable(struct pp_smu *pp) @@ -553,10 +572,10 @@ void pp_rv_set_pme_wa_enable(struct pp_smu *pp) void *pp_handle = adev->powerplay.pp_handle; const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; - if (!pp_funcs || !pp_funcs->notify_smu_enable_pwe) - return; - - pp_funcs->notify_smu_enable_pwe(pp_handle); + if (pp_funcs && pp_funcs->notify_smu_enable_pwe) + pp_funcs->notify_smu_enable_pwe(pp_handle); + else if (adev->smu.funcs) + smu_notify_smu_enable_pwe(&adev->smu); } void pp_rv_set_active_display_count(struct pp_smu *pp, int count) @@ -611,17 +630,16 @@ void pp_rv_set_hard_min_fclk_by_freq(struct pp_smu *pp, int mhz) pp_funcs->set_hard_min_fclk_by_freq(pp_handle, mhz); } -void dm_pp_get_funcs_rv( +void dm_pp_get_funcs( struct dc_context *ctx, - struct pp_smu_funcs_rv *funcs) + struct pp_smu_funcs *funcs) { - funcs->pp_smu.dm = ctx; - funcs->set_display_requirement = pp_rv_set_display_requirement; - funcs->set_wm_ranges = pp_rv_set_wm_ranges; - funcs->set_pme_wa_enable = pp_rv_set_pme_wa_enable; - funcs->set_display_count = pp_rv_set_active_display_count; - funcs->set_min_deep_sleep_dcfclk = pp_rv_set_min_deep_sleep_dcfclk; - funcs->set_hard_min_dcfclk_by_freq = pp_rv_set_hard_min_dcefclk_by_freq; - funcs->set_hard_min_fclk_by_freq = pp_rv_set_hard_min_fclk_by_freq; + funcs->rv_funcs.pp_smu.dm = ctx; + funcs->rv_funcs.set_wm_ranges = pp_rv_set_wm_ranges; + funcs->rv_funcs.set_pme_wa_enable = pp_rv_set_pme_wa_enable; + funcs->rv_funcs.set_display_count = pp_rv_set_active_display_count; + funcs->rv_funcs.set_min_deep_sleep_dcfclk = pp_rv_set_min_deep_sleep_dcfclk; + funcs->rv_funcs.set_hard_min_dcfclk_by_freq = pp_rv_set_hard_min_dcefclk_by_freq; + funcs->rv_funcs.set_hard_min_fclk_by_freq = pp_rv_set_hard_min_fclk_by_freq; } diff --git a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c index f28989860fd8..1e9a2d352068 100644 --- a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c +++ b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c @@ -449,6 +449,11 @@ static inline unsigned int clamp_ux_dy( return min_clamp; } +unsigned int dc_fixpt_u4d19(struct fixed31_32 arg) +{ + return ux_dy(arg.value, 4, 19); +} + unsigned int dc_fixpt_u3d19(struct fixed31_32 arg) { return ux_dy(arg.value, 3, 19); diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c index eb62d10bb65c..1b4b51657f5e 100644 --- a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c +++ b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c @@ -247,6 +247,53 @@ static enum dcn_bw_defs tl_pixel_format_to_bw_defs(enum surface_pixel_format for } } +enum source_macro_tile_size swizzle_mode_to_macro_tile_size(enum swizzle_mode_values sw_mode) +{ + switch (sw_mode) { + /* for 4/8/16 high tiles */ + case DC_SW_LINEAR: + return dm_4k_tile; + case DC_SW_4KB_S: + case DC_SW_4KB_S_X: + return dm_4k_tile; + case DC_SW_64KB_S: + case DC_SW_64KB_S_X: + case DC_SW_64KB_S_T: + return dm_64k_tile; + case DC_SW_VAR_S: + case DC_SW_VAR_S_X: + return dm_256k_tile; + + /* For 64bpp 2 high tiles */ + case DC_SW_4KB_D: + case DC_SW_4KB_D_X: + return dm_4k_tile; + case DC_SW_64KB_D: + case DC_SW_64KB_D_X: + case DC_SW_64KB_D_T: + return dm_64k_tile; + case DC_SW_VAR_D: + case DC_SW_VAR_D_X: + return dm_256k_tile; + + case DC_SW_4KB_R: + case DC_SW_4KB_R_X: + return dm_4k_tile; + case DC_SW_64KB_R: + case DC_SW_64KB_R_X: + return dm_64k_tile; + case DC_SW_VAR_R: + case DC_SW_VAR_R_X: + return dm_256k_tile; + + /* Unsupported swizzle modes for dcn */ + case DC_SW_256B_S: + default: + ASSERT(0); /* Not supported */ + return 0; + } +} + static void pipe_ctx_to_e2e_pipe_params ( const struct pipe_ctx *pipe, struct _vcs_dpi_display_pipe_params_st *input) @@ -287,46 +334,7 @@ static void pipe_ctx_to_e2e_pipe_params ( input->src.cur0_src_width = 128; /* TODO: Cursor calcs, not curently stored */ input->src.cur0_bpp = 32; - switch (pipe->plane_state->tiling_info.gfx9.swizzle) { - /* for 4/8/16 high tiles */ - case DC_SW_LINEAR: - input->src.macro_tile_size = dm_4k_tile; - break; - case DC_SW_4KB_S: - case DC_SW_4KB_S_X: - input->src.macro_tile_size = dm_4k_tile; - break; - case DC_SW_64KB_S: - case DC_SW_64KB_S_X: - case DC_SW_64KB_S_T: - input->src.macro_tile_size = dm_64k_tile; - break; - case DC_SW_VAR_S: - case DC_SW_VAR_S_X: - input->src.macro_tile_size = dm_256k_tile; - break; - - /* For 64bpp 2 high tiles */ - case DC_SW_4KB_D: - case DC_SW_4KB_D_X: - input->src.macro_tile_size = dm_4k_tile; - break; - case DC_SW_64KB_D: - case DC_SW_64KB_D_X: - case DC_SW_64KB_D_T: - input->src.macro_tile_size = dm_64k_tile; - break; - case DC_SW_VAR_D: - case DC_SW_VAR_D_X: - input->src.macro_tile_size = dm_256k_tile; - break; - - /* Unsupported swizzle modes for dcn */ - case DC_SW_256B_S: - default: - ASSERT(0); /* Not supported */ - break; - } + input->src.macro_tile_size = swizzle_mode_to_macro_tile_size(pipe->plane_state->tiling_info.gfx9.swizzle); switch (pipe->plane_state->rotation) { case ROTATION_ANGLE_0: @@ -466,7 +474,7 @@ static void dcn_bw_calc_rq_dlg_ttu( input.clks_cfg.dcfclk_mhz = v->dcfclk; input.clks_cfg.dispclk_mhz = v->dispclk; input.clks_cfg.dppclk_mhz = v->dppclk; - input.clks_cfg.refclk_mhz = dc->res_pool->ref_clock_inKhz / 1000.0; + input.clks_cfg.refclk_mhz = dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000.0; input.clks_cfg.socclk_mhz = v->socclk; input.clks_cfg.voltage = v->voltage_level; // dc->dml.logger = pool->base.logger; @@ -536,28 +544,28 @@ static void calc_wm_sets_and_perf_params( v->fabric_and_dram_bandwidth = v->fabric_and_dram_bandwidth_vnom0p8; dispclkdppclkdcfclk_deep_sleep_prefetch_parameters_watermarks_and_performance_calculation(v); - context->bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns = + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns = v->stutter_exit_watermark * 1000; - context->bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns = + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns = v->stutter_enter_plus_exit_watermark * 1000; - context->bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns = + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns = v->dram_clock_change_watermark * 1000; - context->bw.dcn.watermarks.b.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; - context->bw.dcn.watermarks.b.urgent_ns = v->urgent_watermark * 1000; + context->bw_ctx.bw.dcn.watermarks.b.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; + context->bw_ctx.bw.dcn.watermarks.b.urgent_ns = v->urgent_watermark * 1000; v->dcfclk_per_state[1] = v->dcfclkv_nom0p8; v->dcfclk_per_state[0] = v->dcfclkv_nom0p8; v->dcfclk = v->dcfclkv_nom0p8; dispclkdppclkdcfclk_deep_sleep_prefetch_parameters_watermarks_and_performance_calculation(v); - context->bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns = + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns = v->stutter_exit_watermark * 1000; - context->bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns = + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns = v->stutter_enter_plus_exit_watermark * 1000; - context->bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns = + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns = v->dram_clock_change_watermark * 1000; - context->bw.dcn.watermarks.c.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; - context->bw.dcn.watermarks.c.urgent_ns = v->urgent_watermark * 1000; + context->bw_ctx.bw.dcn.watermarks.c.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; + context->bw_ctx.bw.dcn.watermarks.c.urgent_ns = v->urgent_watermark * 1000; } if (v->voltage_level < 3) { @@ -571,14 +579,14 @@ static void calc_wm_sets_and_perf_params( v->dcfclk = v->dcfclkv_max0p9; dispclkdppclkdcfclk_deep_sleep_prefetch_parameters_watermarks_and_performance_calculation(v); - context->bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns = + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns = v->stutter_exit_watermark * 1000; - context->bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns = + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns = v->stutter_enter_plus_exit_watermark * 1000; - context->bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns = + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns = v->dram_clock_change_watermark * 1000; - context->bw.dcn.watermarks.d.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; - context->bw.dcn.watermarks.d.urgent_ns = v->urgent_watermark * 1000; + context->bw_ctx.bw.dcn.watermarks.d.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; + context->bw_ctx.bw.dcn.watermarks.d.urgent_ns = v->urgent_watermark * 1000; } v->fabric_and_dram_bandwidth_per_state[2] = v->fabric_and_dram_bandwidth_vnom0p8; @@ -591,20 +599,20 @@ static void calc_wm_sets_and_perf_params( v->dcfclk = v->dcfclk_per_state[v->voltage_level]; dispclkdppclkdcfclk_deep_sleep_prefetch_parameters_watermarks_and_performance_calculation(v); - context->bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns = + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns = v->stutter_exit_watermark * 1000; - context->bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns = + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns = v->stutter_enter_plus_exit_watermark * 1000; - context->bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = v->dram_clock_change_watermark * 1000; - context->bw.dcn.watermarks.a.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; - context->bw.dcn.watermarks.a.urgent_ns = v->urgent_watermark * 1000; + context->bw_ctx.bw.dcn.watermarks.a.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; + context->bw_ctx.bw.dcn.watermarks.a.urgent_ns = v->urgent_watermark * 1000; if (v->voltage_level >= 2) { - context->bw.dcn.watermarks.b = context->bw.dcn.watermarks.a; - context->bw.dcn.watermarks.c = context->bw.dcn.watermarks.a; + context->bw_ctx.bw.dcn.watermarks.b = context->bw_ctx.bw.dcn.watermarks.a; + context->bw_ctx.bw.dcn.watermarks.c = context->bw_ctx.bw.dcn.watermarks.a; } if (v->voltage_level >= 3) - context->bw.dcn.watermarks.d = context->bw.dcn.watermarks.a; + context->bw_ctx.bw.dcn.watermarks.d = context->bw_ctx.bw.dcn.watermarks.a; } #endif @@ -693,8 +701,15 @@ static void hack_bounding_box(struct dcn_bw_internal_vars *v, bool dcn_validate_bandwidth( struct dc *dc, - struct dc_state *context) + struct dc_state *context, + bool fast_validate) { + /* + * we want a breakdown of the various stages of validation, which the + * perf_trace macro doesn't support + */ + BW_VAL_TRACE_SETUP(); + const struct resource_pool *pool = dc->res_pool; struct dcn_bw_internal_vars *v = &context->dcn_bw_vars; int i, input_idx; @@ -703,6 +718,9 @@ bool dcn_validate_bandwidth( float bw_limit; PERFORMANCE_TRACE_START(); + + BW_VAL_TRACE_COUNT(); + if (dcn_bw_apply_registry_override(dc)) dcn_bw_sync_calcs_and_dml(dc); @@ -1000,13 +1018,16 @@ bool dcn_validate_bandwidth( dc->debug.sr_enter_plus_exit_time_dpm0_ns / 1000.0f; if (dc->debug.sr_exit_time_dpm0_ns) v->sr_exit_time = dc->debug.sr_exit_time_dpm0_ns / 1000.0f; - dc->dml.soc.sr_enter_plus_exit_time_us = v->sr_enter_plus_exit_time; - dc->dml.soc.sr_exit_time_us = v->sr_exit_time; + context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = v->sr_enter_plus_exit_time; + context->bw_ctx.dml.soc.sr_exit_time_us = v->sr_exit_time; mode_support_and_system_configuration(v); } - if (v->voltage_level != 5) { + BW_VAL_TRACE_END_VOLTAGE_LEVEL(); + + if (v->voltage_level != number_of_states_plus_one && !fast_validate) { float bw_consumed = v->total_bandwidth_consumed_gbyte_per_second; + if (bw_consumed < v->fabric_and_dram_bandwidth_vmin0p65) bw_consumed = v->fabric_and_dram_bandwidth_vmin0p65; else if (bw_consumed < v->fabric_and_dram_bandwidth_vmid0p72) @@ -1027,58 +1048,60 @@ bool dcn_validate_bandwidth( */ dispclkdppclkdcfclk_deep_sleep_prefetch_parameters_watermarks_and_performance_calculation(v); - context->bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns = + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns = v->stutter_exit_watermark * 1000; - context->bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns = + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns = v->stutter_enter_plus_exit_watermark * 1000; - context->bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = v->dram_clock_change_watermark * 1000; - context->bw.dcn.watermarks.a.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; - context->bw.dcn.watermarks.a.urgent_ns = v->urgent_watermark * 1000; - context->bw.dcn.watermarks.b = context->bw.dcn.watermarks.a; - context->bw.dcn.watermarks.c = context->bw.dcn.watermarks.a; - context->bw.dcn.watermarks.d = context->bw.dcn.watermarks.a; + context->bw_ctx.bw.dcn.watermarks.a.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; + context->bw_ctx.bw.dcn.watermarks.a.urgent_ns = v->urgent_watermark * 1000; + context->bw_ctx.bw.dcn.watermarks.b = context->bw_ctx.bw.dcn.watermarks.a; + context->bw_ctx.bw.dcn.watermarks.c = context->bw_ctx.bw.dcn.watermarks.a; + context->bw_ctx.bw.dcn.watermarks.d = context->bw_ctx.bw.dcn.watermarks.a; - context->bw.dcn.clk.fclk_khz = (int)(bw_consumed * 1000000 / + context->bw_ctx.bw.dcn.clk.fclk_khz = (int)(bw_consumed * 1000000 / (ddr4_dram_factor_single_Channel * v->number_of_channels)); if (bw_consumed == v->fabric_and_dram_bandwidth_vmin0p65) { - context->bw.dcn.clk.fclk_khz = (int)(bw_consumed * 1000000 / 32); + context->bw_ctx.bw.dcn.clk.fclk_khz = (int)(bw_consumed * 1000000 / 32); } - context->bw.dcn.clk.dcfclk_deep_sleep_khz = (int)(v->dcf_clk_deep_sleep * 1000); - context->bw.dcn.clk.dcfclk_khz = (int)(v->dcfclk * 1000); + context->bw_ctx.bw.dcn.clk.dcfclk_deep_sleep_khz = (int)(v->dcf_clk_deep_sleep * 1000); + context->bw_ctx.bw.dcn.clk.dcfclk_khz = (int)(v->dcfclk * 1000); - context->bw.dcn.clk.dispclk_khz = (int)(v->dispclk * 1000); + context->bw_ctx.bw.dcn.clk.dispclk_khz = (int)(v->dispclk * 1000); if (dc->debug.max_disp_clk == true) - context->bw.dcn.clk.dispclk_khz = (int)(dc->dcn_soc->max_dispclk_vmax0p9 * 1000); + context->bw_ctx.bw.dcn.clk.dispclk_khz = (int)(dc->dcn_soc->max_dispclk_vmax0p9 * 1000); - if (context->bw.dcn.clk.dispclk_khz < + if (context->bw_ctx.bw.dcn.clk.dispclk_khz < dc->debug.min_disp_clk_khz) { - context->bw.dcn.clk.dispclk_khz = + context->bw_ctx.bw.dcn.clk.dispclk_khz = dc->debug.min_disp_clk_khz; } - context->bw.dcn.clk.dppclk_khz = context->bw.dcn.clk.dispclk_khz / v->dispclk_dppclk_ratio; - context->bw.dcn.clk.phyclk_khz = v->phyclk_per_state[v->voltage_level]; + context->bw_ctx.bw.dcn.clk.dppclk_khz = context->bw_ctx.bw.dcn.clk.dispclk_khz / v->dispclk_dppclk_ratio; + context->bw_ctx.bw.dcn.clk.phyclk_khz = v->phyclk_per_state[v->voltage_level]; switch (v->voltage_level) { case 0: - context->bw.dcn.clk.max_supported_dppclk_khz = + context->bw_ctx.bw.dcn.clk.max_supported_dppclk_khz = (int)(dc->dcn_soc->max_dppclk_vmin0p65 * 1000); break; case 1: - context->bw.dcn.clk.max_supported_dppclk_khz = + context->bw_ctx.bw.dcn.clk.max_supported_dppclk_khz = (int)(dc->dcn_soc->max_dppclk_vmid0p72 * 1000); break; case 2: - context->bw.dcn.clk.max_supported_dppclk_khz = + context->bw_ctx.bw.dcn.clk.max_supported_dppclk_khz = (int)(dc->dcn_soc->max_dppclk_vnom0p8 * 1000); break; default: - context->bw.dcn.clk.max_supported_dppclk_khz = + context->bw_ctx.bw.dcn.clk.max_supported_dppclk_khz = (int)(dc->dcn_soc->max_dppclk_vmax0p9 * 1000); break; } + BW_VAL_TRACE_END_WATERMARKS(); + for (i = 0, input_idx = 0; i < pool->pipe_count; i++) { struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; @@ -1141,7 +1164,7 @@ bool dcn_validate_bandwidth( hsplit_pipe->pipe_dlg_param.vblank_end = pipe->pipe_dlg_param.vblank_end; } else { /* pipe not split previously needs split */ - hsplit_pipe = find_idle_secondary_pipe(&context->res_ctx, pool); + hsplit_pipe = find_idle_secondary_pipe(&context->res_ctx, pool, pipe); ASSERT(hsplit_pipe); split_stream_across_pipes( &context->res_ctx, pool, @@ -1169,13 +1192,17 @@ bool dcn_validate_bandwidth( input_idx++; } + } else if (v->voltage_level == number_of_states_plus_one) { + BW_VAL_TRACE_SKIP(fail); + } else if (fast_validate) { + BW_VAL_TRACE_SKIP(fast); } if (v->voltage_level == 0) { - dc->dml.soc.sr_enter_plus_exit_time_us = + context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->dcn_soc->sr_enter_plus_exit_time; - dc->dml.soc.sr_exit_time_us = dc->dcn_soc->sr_exit_time; + context->bw_ctx.dml.soc.sr_exit_time_us = dc->dcn_soc->sr_exit_time; } /* @@ -1188,6 +1215,7 @@ bool dcn_validate_bandwidth( kernel_fpu_end(); PERFORMANCE_TRACE_END(); + BW_VAL_TRACE_FINISH(); if (bw_limit_pass && v->voltage_level != 5) return true; @@ -1395,12 +1423,14 @@ void dcn_bw_update_from_pplib(struct dc *dc) void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc) { - struct pp_smu_funcs_rv *pp = dc->res_pool->pp_smu; + struct pp_smu_funcs_rv *pp = NULL; struct pp_smu_wm_range_sets ranges = {0}; int min_fclk_khz, min_dcfclk_khz, socclk_khz; const int overdrive = 5000000; /* 5 GHz to cover Overdrive */ - if (!pp->set_wm_ranges) + if (dc->res_pool->pp_smu) + pp = &dc->res_pool->pp_smu->rv_funcs; + if (!pp || !pp->set_wm_ranges) return; kernel_fpu_begin(); diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index c68fbd55db3c..dda10b1f8574 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -524,6 +524,14 @@ void dc_link_set_preferred_link_settings(struct dc *dc, struct dc_stream_state *link_stream; struct dc_link_settings store_settings = *link_setting; + link->preferred_link_setting = store_settings; + + /* Retrain with preferred link settings only relevant for + * DP signal type + */ + if (!dc_is_dp_signal(link->connector_signal)) + return; + for (i = 0; i < MAX_PIPES; i++) { pipe = &dc->current_state->res_ctx.pipe_ctx[i]; if (pipe->stream && pipe->stream->link) { @@ -538,7 +546,10 @@ void dc_link_set_preferred_link_settings(struct dc *dc, link_stream = link->dc->current_state->res_ctx.pipe_ctx[i].stream; - link->preferred_link_setting = store_settings; + /* Cannot retrain link if backend is off */ + if (link_stream->dpms_off) + return; + if (link_stream) decide_link_settings(link_stream, &store_settings); @@ -573,6 +584,28 @@ void dc_link_set_test_pattern(struct dc_link *link, cust_pattern_size); } +uint32_t dc_link_bandwidth_kbps( + const struct dc_link *link, + const struct dc_link_settings *link_setting) +{ + uint32_t link_bw_kbps = link_setting->link_rate * LINK_RATE_REF_FREQ_IN_KHZ; /* bytes per sec */ + + link_bw_kbps *= 8; /* 8 bits per byte*/ + link_bw_kbps *= link_setting->lane_count; + + return link_bw_kbps; + +} + +const struct dc_link_settings *dc_link_get_link_cap( + const struct dc_link *link) +{ + if (link->preferred_link_setting.lane_count != LANE_COUNT_UNKNOWN && + link->preferred_link_setting.link_rate != LINK_RATE_UNKNOWN) + return &link->preferred_link_setting; + return &link->verified_link_cap; +} + static void destruct(struct dc *dc) { dc_release_state(dc->current_state); @@ -621,6 +654,10 @@ static bool construct(struct dc *dc, #endif enum dce_version dc_version = DCE_VERSION_UNKNOWN; + dc->config = init_params->flags; + + memcpy(&dc->bb_overrides, &init_params->bb_overrides, sizeof(dc->bb_overrides)); + dc_dceip = kzalloc(sizeof(*dc_dceip), GFP_KERNEL); if (!dc_dceip) { dm_error("%s: failed to create dceip\n", __func__); @@ -668,13 +705,6 @@ static bool construct(struct dc *dc, dc_ctx->dc_stream_id_count = 0; dc->ctx = dc_ctx; - dc->current_state = dc_create_state(); - - if (!dc->current_state) { - dm_error("%s: failed to create validate ctx\n", __func__); - goto fail; - } - /* Create logger */ dc_ctx->dce_environment = init_params->dce_environment; @@ -722,14 +752,22 @@ static bool construct(struct dc *dc, goto fail; } - dc->res_pool = dc_create_resource_pool( - dc, - init_params->num_virtual_links, - dc_version, - init_params->asic_id); + dc->res_pool = dc_create_resource_pool(dc, init_params, dc_version); if (!dc->res_pool) goto fail; + /* Creation of current_state must occur after dc->dml + * is initialized in dc_create_resource_pool because + * on creation it copies the contents of dc->dml + */ + + dc->current_state = dc_create_state(dc); + + if (!dc->current_state) { + dm_error("%s: failed to create validate ctx\n", __func__); + goto fail; + } + dc_resource_state_construct(dc, dc->current_state); if (!create_links(dc, init_params->num_virtual_links)) @@ -746,7 +784,7 @@ fail: static void disable_dangling_plane(struct dc *dc, struct dc_state *context) { int i, j; - struct dc_state *dangling_context = dc_create_state(); + struct dc_state *dangling_context = dc_create_state(dc); struct dc_state *current_ctx; if (dangling_context == NULL) @@ -811,8 +849,6 @@ struct dc *dc_create(const struct dc_init_data *init_params) if (dc->res_pool->dmcu != NULL) dc->versions.dmcu_version = dc->res_pool->dmcu->dmcu_version; - dc->config = init_params->flags; - dc->build_id = DC_BUILD_ID; DC_LOG_DC("Display Core initialized\n"); @@ -969,7 +1005,7 @@ static bool context_changed( return false; } -bool dc_validate_seamless_boot_timing(struct dc *dc, +bool dc_validate_seamless_boot_timing(const struct dc *dc, const struct dc_sink *sink, struct dc_crtc_timing *crtc_timing) { @@ -1060,7 +1096,13 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c if (!dcb->funcs->is_accelerated_mode(dcb)) dc->hwss.enable_accelerated_mode(dc, context); - dc->hwss.prepare_bandwidth(dc, context); + for (i = 0; i < context->stream_count; i++) { + if (context->streams[i]->apply_seamless_boot_optimization) + dc->optimize_seamless_boot = true; + } + + if (!dc->optimize_seamless_boot) + dc->hwss.prepare_bandwidth(dc, context); /* re-program planes for existing stream, in case we need to * free up plane resource for later use @@ -1135,12 +1177,15 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c dc_enable_stereo(dc, context, dc_streams, context->stream_count); - /* pplib is notified if disp_num changed */ - dc->hwss.optimize_bandwidth(dc, context); + if (!dc->optimize_seamless_boot) + /* pplib is notified if disp_num changed */ + dc->hwss.optimize_bandwidth(dc, context); for (i = 0; i < context->stream_count; i++) context->streams[i]->mode_changed = false; + memset(&context->commit_hints, 0, sizeof(context->commit_hints)); + dc_release_state(dc->current_state); dc->current_state = context; @@ -1177,7 +1222,7 @@ bool dc_post_update_surfaces_to_stream(struct dc *dc) int i; struct dc_state *context = dc->current_state; - if (dc->optimized_required == false) + if (!dc->optimized_required || dc->optimize_seamless_boot) return true; post_surface_trace(dc); @@ -1195,18 +1240,60 @@ bool dc_post_update_surfaces_to_stream(struct dc *dc) return true; } -struct dc_state *dc_create_state(void) +struct dc_state *dc_create_state(struct dc *dc) { struct dc_state *context = kzalloc(sizeof(struct dc_state), GFP_KERNEL); if (!context) return NULL; + /* Each context must have their own instance of VBA and in order to + * initialize and obtain IP and SOC the base DML instance from DC is + * initially copied into every context + */ +#ifdef CONFIG_DRM_AMD_DC_DCN1_0 + memcpy(&context->bw_ctx.dml, &dc->dml, sizeof(struct display_mode_lib)); +#endif kref_init(&context->refcount); + return context; } +struct dc_state *dc_copy_state(struct dc_state *src_ctx) +{ + int i, j; + struct dc_state *new_ctx = kzalloc(sizeof(struct dc_state), + GFP_KERNEL); + + if (!new_ctx) + return NULL; + + memcpy(new_ctx, src_ctx, sizeof(struct dc_state)); + + for (i = 0; i < MAX_PIPES; i++) { + struct pipe_ctx *cur_pipe = &new_ctx->res_ctx.pipe_ctx[i]; + + if (cur_pipe->top_pipe) + cur_pipe->top_pipe = &new_ctx->res_ctx.pipe_ctx[cur_pipe->top_pipe->pipe_idx]; + + if (cur_pipe->bottom_pipe) + cur_pipe->bottom_pipe = &new_ctx->res_ctx.pipe_ctx[cur_pipe->bottom_pipe->pipe_idx]; + + } + + for (i = 0; i < new_ctx->stream_count; i++) { + dc_stream_retain(new_ctx->streams[i]); + for (j = 0; j < new_ctx->stream_status[i].plane_count; j++) + dc_plane_state_retain( + new_ctx->stream_status[i].plane_states[j]); + } + + kref_init(&new_ctx->refcount); + + return new_ctx; +} + void dc_retain_state(struct dc_state *context) { kref_get(&context->refcount); @@ -1661,6 +1748,7 @@ static void commit_planes_do_stream_update(struct dc *dc, continue; if (stream_update->dpms_off) { + dc->hwss.pipe_control_lock(dc, pipe_ctx, true); if (*stream_update->dpms_off) { core_link_disable_stream(pipe_ctx, KEEP_ACQUIRED_RESOURCE); dc->hwss.optimize_bandwidth(dc, dc->current_state); @@ -1668,6 +1756,7 @@ static void commit_planes_do_stream_update(struct dc *dc, dc->hwss.prepare_bandwidth(dc, dc->current_state); core_link_enable_stream(dc->current_state, pipe_ctx); } + dc->hwss.pipe_control_lock(dc, pipe_ctx, false); } if (stream_update->abm_level && pipe_ctx->stream_res.abm) { @@ -1695,7 +1784,16 @@ static void commit_planes_for_stream(struct dc *dc, int i, j; struct pipe_ctx *top_pipe_to_program = NULL; - if (update_type == UPDATE_TYPE_FULL) { + if (dc->optimize_seamless_boot && surface_count > 0) { + /* Optimize seamless boot flag keeps clocks and watermarks high until + * first flip. After first flip, optimization is required to lower + * bandwidth. + */ + dc->optimize_seamless_boot = false; + dc->optimized_required = true; + } + + if (update_type == UPDATE_TYPE_FULL && !dc->optimize_seamless_boot) { dc->hwss.prepare_bandwidth(dc, context); context_clock_trace(dc, context); } @@ -1795,7 +1893,7 @@ void dc_commit_updates_for_stream(struct dc *dc, if (update_type >= UPDATE_TYPE_FULL) { /* initialize scratch memory for building context */ - context = dc_create_state(); + context = dc_create_state(dc); if (context == NULL) { DC_ERROR("Failed to allocate new validate context!\n"); return; @@ -2080,13 +2178,13 @@ void dc_link_remove_remote_sink(struct dc_link *link, struct dc_sink *sink) void get_clock_requirements_for_state(struct dc_state *state, struct AsicStateEx *info) { - info->displayClock = (unsigned int)state->bw.dcn.clk.dispclk_khz; - info->engineClock = (unsigned int)state->bw.dcn.clk.dcfclk_khz; - info->memoryClock = (unsigned int)state->bw.dcn.clk.dramclk_khz; - info->maxSupportedDppClock = (unsigned int)state->bw.dcn.clk.max_supported_dppclk_khz; - info->dppClock = (unsigned int)state->bw.dcn.clk.dppclk_khz; - info->socClock = (unsigned int)state->bw.dcn.clk.socclk_khz; - info->dcfClockDeepSleep = (unsigned int)state->bw.dcn.clk.dcfclk_deep_sleep_khz; - info->fClock = (unsigned int)state->bw.dcn.clk.fclk_khz; - info->phyClock = (unsigned int)state->bw.dcn.clk.phyclk_khz; + info->displayClock = (unsigned int)state->bw_ctx.bw.dcn.clk.dispclk_khz; + info->engineClock = (unsigned int)state->bw_ctx.bw.dcn.clk.dcfclk_khz; + info->memoryClock = (unsigned int)state->bw_ctx.bw.dcn.clk.dramclk_khz; + info->maxSupportedDppClock = (unsigned int)state->bw_ctx.bw.dcn.clk.max_supported_dppclk_khz; + info->dppClock = (unsigned int)state->bw_ctx.bw.dcn.clk.dppclk_khz; + info->socClock = (unsigned int)state->bw_ctx.bw.dcn.clk.socclk_khz; + info->dcfClockDeepSleep = (unsigned int)state->bw_ctx.bw.dcn.clk.dcfclk_deep_sleep_khz; + info->fClock = (unsigned int)state->bw_ctx.bw.dcn.clk.fclk_khz; + info->phyClock = (unsigned int)state->bw_ctx.bw.dcn.clk.phyclk_khz; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c index 73d049506618..5903e7822f98 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c @@ -351,19 +351,19 @@ void context_clock_trace( DC_LOGGER_INIT(dc->ctx->logger); CLOCK_TRACE("Current: dispclk_khz:%d max_dppclk_khz:%d dcfclk_khz:%d\n" "dcfclk_deep_sleep_khz:%d fclk_khz:%d socclk_khz:%d\n", - context->bw.dcn.clk.dispclk_khz, - context->bw.dcn.clk.dppclk_khz, - context->bw.dcn.clk.dcfclk_khz, - context->bw.dcn.clk.dcfclk_deep_sleep_khz, - context->bw.dcn.clk.fclk_khz, - context->bw.dcn.clk.socclk_khz); + context->bw_ctx.bw.dcn.clk.dispclk_khz, + context->bw_ctx.bw.dcn.clk.dppclk_khz, + context->bw_ctx.bw.dcn.clk.dcfclk_khz, + context->bw_ctx.bw.dcn.clk.dcfclk_deep_sleep_khz, + context->bw_ctx.bw.dcn.clk.fclk_khz, + context->bw_ctx.bw.dcn.clk.socclk_khz); CLOCK_TRACE("Calculated: dispclk_khz:%d max_dppclk_khz:%d dcfclk_khz:%d\n" "dcfclk_deep_sleep_khz:%d fclk_khz:%d socclk_khz:%d\n", - context->bw.dcn.clk.dispclk_khz, - context->bw.dcn.clk.dppclk_khz, - context->bw.dcn.clk.dcfclk_khz, - context->bw.dcn.clk.dcfclk_deep_sleep_khz, - context->bw.dcn.clk.fclk_khz, - context->bw.dcn.clk.socclk_khz); + context->bw_ctx.bw.dcn.clk.dispclk_khz, + context->bw_ctx.bw.dcn.clk.dppclk_khz, + context->bw_ctx.bw.dcn.clk.dcfclk_khz, + context->bw_ctx.bw.dcn.clk.dcfclk_deep_sleep_khz, + context->bw_ctx.bw.dcn.clk.fclk_khz, + context->bw_ctx.bw.dcn.clk.socclk_khz); #endif } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index 4eba3c4800b6..3ef68a249f4d 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -58,7 +58,6 @@ ******************************************************************************/ enum { - LINK_RATE_REF_FREQ_IN_MHZ = 27, PEAK_FACTOR_X1000 = 1006, /* * Some receivers fail to train on first try and are good @@ -640,7 +639,8 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason) bool same_dpcd = true; enum dc_connection_type new_connection_type = dc_connection_none; DC_LOGGER_INIT(link->ctx->logger); - if (link->connector_signal == SIGNAL_TYPE_VIRTUAL) + + if (dc_is_virtual_signal(link->connector_signal)) return false; if (false == dc_link_detect_sink(link, &new_connection_type)) { @@ -720,9 +720,8 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason) same_dpcd = false; } /* Active dongle plug in without display or downstream unplug*/ - if (link->type == dc_connection_active_dongle - && link->dpcd_caps.sink_count. - bits.SINK_COUNT == 0) { + if (link->type == dc_connection_active_dongle && + link->dpcd_caps.sink_count.bits.SINK_COUNT == 0) { if (prev_sink != NULL) { /* Downstream unplug */ dc_sink_release(prev_sink); @@ -1172,8 +1171,6 @@ static bool construct( goto create_fail; } - - /* TODO: #DAL3 Implement id to str function.*/ LINK_INFO("Connector[%d] description:" "signal %d\n", @@ -1207,7 +1204,7 @@ static bool construct( link->link_enc = link->dc->res_pool->funcs->link_enc_create( &enc_init_data); - if( link->link_enc == NULL) { + if (link->link_enc == NULL) { DC_ERROR("Failed to create link encoder!\n"); goto link_enc_create_fail; } @@ -1399,9 +1396,18 @@ static enum dc_status enable_link_dp( /* get link settings for video mode timing */ decide_link_settings(stream, &link_settings); + /* If link settings are different than current and link already enabled + * then need to disable before programming to new rate. + */ + if (link->link_status.link_active && + (link->cur_link_settings.lane_count != link_settings.lane_count || + link->cur_link_settings.link_rate != link_settings.link_rate)) { + dp_disable_link_phy(link, pipe_ctx->stream->signal); + } + pipe_ctx->stream_res.pix_clk_params.requested_sym_clk = link_settings.link_rate * LINK_RATE_REF_FREQ_IN_KHZ; - state->dccg->funcs->update_clocks(state->dccg, state, false); + state->clk_mgr->funcs->update_clocks(state->clk_mgr, state, false); dp_enable_link_phy( link, @@ -1466,14 +1472,14 @@ static enum dc_status enable_link_dp_mst( if (link->cur_link_settings.lane_count != LANE_COUNT_UNKNOWN) return DC_OK; - /* to make sure the pending down rep can be processed - * before clear payload table - */ - dm_helpers_dp_mst_poll_pending_down_reply(link->ctx, link); - /* clear payload table */ dm_helpers_dp_mst_clear_payload_allocation_table(link->ctx, link); + /* to make sure the pending down rep can be processed + * before enabling the link + */ + dm_helpers_dp_mst_poll_pending_down_reply(link->ctx, link); + /* set the sink to MST mode before enabling the link */ dp_enable_mst_on_sink(link, true); @@ -1982,7 +1988,7 @@ static void enable_link_hdmi(struct pipe_ctx *pipe_ctx) pipe_ctx->stream->signal, stream->phy_pix_clk); - if (pipe_ctx->stream->signal == SIGNAL_TYPE_HDMI_TYPE_A) + if (dc_is_hdmi_signal(pipe_ctx->stream->signal)) dal_ddc_service_read_scdc_data(link->ddc); } @@ -2074,11 +2080,28 @@ static void disable_link(struct dc_link *link, enum signal_type signal) } } +static uint32_t get_timing_pixel_clock_100hz(const struct dc_crtc_timing *timing) +{ + + uint32_t pxl_clk = timing->pix_clk_100hz; + + if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) + pxl_clk /= 2; + else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) + pxl_clk = pxl_clk * 2 / 3; + + if (timing->display_color_depth == COLOR_DEPTH_101010) + pxl_clk = pxl_clk * 10 / 8; + else if (timing->display_color_depth == COLOR_DEPTH_121212) + pxl_clk = pxl_clk * 12 / 8; + + return pxl_clk; +} + static bool dp_active_dongle_validate_timing( const struct dc_crtc_timing *timing, const struct dpcd_caps *dpcd_caps) { - unsigned int required_pix_clk_100hz = timing->pix_clk_100hz; const struct dc_dongle_caps *dongle_caps = &dpcd_caps->dongle_caps; switch (dpcd_caps->dongle_type) { @@ -2115,13 +2138,6 @@ static bool dp_active_dongle_validate_timing( return false; } - - /* Check Color Depth and Pixel Clock */ - if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) - required_pix_clk_100hz /= 2; - else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) - required_pix_clk_100hz = required_pix_clk_100hz * 2 / 3; - switch (timing->display_color_depth) { case COLOR_DEPTH_666: case COLOR_DEPTH_888: @@ -2130,14 +2146,11 @@ static bool dp_active_dongle_validate_timing( case COLOR_DEPTH_101010: if (dongle_caps->dp_hdmi_max_bpc < 10) return false; - required_pix_clk_100hz = required_pix_clk_100hz * 10 / 8; break; case COLOR_DEPTH_121212: if (dongle_caps->dp_hdmi_max_bpc < 12) return false; - required_pix_clk_100hz = required_pix_clk_100hz * 12 / 8; break; - case COLOR_DEPTH_141414: case COLOR_DEPTH_161616: default: @@ -2145,7 +2158,7 @@ static bool dp_active_dongle_validate_timing( return false; } - if (required_pix_clk_100hz > (dongle_caps->dp_hdmi_max_pixel_clk * 10)) + if (get_timing_pixel_clock_100hz(timing) > (dongle_caps->dp_hdmi_max_pixel_clk_in_khz * 10)) return false; return true; @@ -2166,7 +2179,7 @@ enum dc_status dc_link_validate_mode_timing( return DC_OK; /* Passive Dongle */ - if (0 != max_pix_clk && timing->pix_clk_100hz > max_pix_clk) + if (max_pix_clk != 0 && get_timing_pixel_clock_100hz(timing) > max_pix_clk) return DC_EXCEED_DONGLE_CAP; /* Active Dongle*/ @@ -2284,14 +2297,13 @@ void core_link_resume(struct dc_link *link) static struct fixed31_32 get_pbn_per_slot(struct dc_stream_state *stream) { - struct dc_link_settings *link_settings = - &stream->link->cur_link_settings; - uint32_t link_rate_in_mbps = - link_settings->link_rate * LINK_RATE_REF_FREQ_IN_MHZ; - struct fixed31_32 mbps = dc_fixpt_from_int( - link_rate_in_mbps * link_settings->lane_count); + struct fixed31_32 mbytes_per_sec; + uint32_t link_rate_in_mbytes_per_sec = dc_link_bandwidth_kbps(stream->link, &stream->link->cur_link_settings); + link_rate_in_mbytes_per_sec /= 8000; /* Kbits to MBytes */ - return dc_fixpt_div_int(mbps, 54); + mbytes_per_sec = dc_fixpt_from_int(link_rate_in_mbytes_per_sec); + + return dc_fixpt_div_int(mbytes_per_sec, 54); } static int get_color_depth(enum dc_color_depth color_depth) @@ -2316,7 +2328,7 @@ static struct fixed31_32 get_pbn_from_timing(struct pipe_ctx *pipe_ctx) uint32_t denominator; bpc = get_color_depth(pipe_ctx->stream_res.pix_clk_params.color_depth); - kbps = pipe_ctx->stream_res.pix_clk_params.requested_pix_clk_100hz / 10 * bpc * 3; + kbps = dc_bandwidth_in_kbps_from_timing(&pipe_ctx->stream->timing); /* * margin 5300ppm + 300ppm ~ 0.6% as per spec, factor is 1.006 @@ -2551,12 +2563,12 @@ void core_link_enable_stream( struct dc_state *state, struct pipe_ctx *pipe_ctx) { - struct dc *core_dc = pipe_ctx->stream->ctx->dc; + struct dc *core_dc = pipe_ctx->stream->ctx->dc; struct dc_stream_state *stream = pipe_ctx->stream; enum dc_status status; DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger); - if (pipe_ctx->stream->signal != SIGNAL_TYPE_VIRTUAL) { + if (!dc_is_virtual_signal(pipe_ctx->stream->signal)) { stream->link->link_enc->funcs->setup( stream->link->link_enc, pipe_ctx->stream->signal); @@ -2570,9 +2582,10 @@ void core_link_enable_stream( pipe_ctx->stream_res.stream_enc->funcs->dp_set_stream_attribute( pipe_ctx->stream_res.stream_enc, &stream->timing, - stream->output_color_space); + stream->output_color_space, + stream->link->dpcd_caps.dprx_feature.bits.SST_SPLIT_SDP_CAP); - if (dc_is_hdmi_signal(pipe_ctx->stream->signal)) + if (dc_is_hdmi_tmds_signal(pipe_ctx->stream->signal)) pipe_ctx->stream_res.stream_enc->funcs->hdmi_set_stream_attribute( pipe_ctx->stream_res.stream_enc, &stream->timing, @@ -2660,12 +2673,18 @@ void core_link_enable_stream( void core_link_disable_stream(struct pipe_ctx *pipe_ctx, int option) { struct dc *core_dc = pipe_ctx->stream->ctx->dc; + struct dc_stream_state *stream = pipe_ctx->stream; core_dc->hwss.blank_stream(pipe_ctx); if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST) deallocate_mst_payload(pipe_ctx); + if (dc_is_hdmi_signal(pipe_ctx->stream->signal)) + dal_ddc_service_write_scdc_data( + stream->link->ddc, 0, + stream->timing.flags.LTE_340MCSC_SCRAMBLE); + core_dc->hwss.disable_stream(pipe_ctx, option); disable_link(pipe_ctx->stream->link, pipe_ctx->stream->signal); @@ -2730,3 +2749,49 @@ void dc_link_enable_hpd_filter(struct dc_link *link, bool enable) } } +uint32_t dc_bandwidth_in_kbps_from_timing( + const struct dc_crtc_timing *timing) +{ + uint32_t bits_per_channel = 0; + uint32_t kbps; + + switch (timing->display_color_depth) { + case COLOR_DEPTH_666: + bits_per_channel = 6; + break; + case COLOR_DEPTH_888: + bits_per_channel = 8; + break; + case COLOR_DEPTH_101010: + bits_per_channel = 10; + break; + case COLOR_DEPTH_121212: + bits_per_channel = 12; + break; + case COLOR_DEPTH_141414: + bits_per_channel = 14; + break; + case COLOR_DEPTH_161616: + bits_per_channel = 16; + break; + default: + break; + } + + ASSERT(bits_per_channel != 0); + + kbps = timing->pix_clk_100hz / 10; + kbps *= bits_per_channel; + + if (timing->flags.Y_ONLY != 1) { + /*Only YOnly make reduce bandwidth by 1/3 compares to RGB*/ + kbps *= 3; + if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) + kbps /= 2; + else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) + kbps = kbps * 2 / 3; + } + + return kbps; + +} diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c index b7ee63cd8dc7..f02092a0dc76 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c @@ -573,12 +573,28 @@ bool dal_ddc_service_query_ddc_data( return ret; } -int dc_link_aux_transfer(struct ddc_service *ddc, - struct aux_payload *payload) +/* dc_link_aux_transfer_raw() - Attempt to transfer + * the given aux payload. This function does not perform + * retries or handle error states. The reply is returned + * in the payload->reply and the result through + * *operation_result. Returns the number of bytes transferred, + * or -1 on a failure. + */ +int dc_link_aux_transfer_raw(struct ddc_service *ddc, + struct aux_payload *payload, + enum aux_channel_operation_result *operation_result) { - return dce_aux_transfer(ddc, payload); + return dce_aux_transfer_raw(ddc, payload, operation_result); } +/* dc_link_aux_transfer_with_retries() - Attempt to submit an + * aux payload, retrying on timeouts, defers, and busy states + * as outlined in the DP spec. Returns true if the request + * was successful. + * + * Unless you want to implement your own retry semantics, this + * is probably the one you want. + */ bool dc_link_aux_transfer_with_retries(struct ddc_service *ddc, struct aux_payload *payload) { diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c index 09d301216076..a6424c70f4c5 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c @@ -93,12 +93,10 @@ static void dpcd_set_link_settings( struct dc_link *link, const struct link_training_settings *lt_settings) { - uint8_t rate = (uint8_t) - (lt_settings->link_settings.link_rate); + uint8_t rate; union down_spread_ctrl downspread = { {0} }; union lane_count_set lane_count_set = { {0} }; - uint8_t link_set_buffer[2]; downspread.raw = (uint8_t) (lt_settings->link_settings.link_spread); @@ -111,29 +109,42 @@ static void dpcd_set_link_settings( lane_count_set.bits.POST_LT_ADJ_REQ_GRANTED = link->dpcd_caps.max_ln_count.bits.POST_LT_ADJ_REQ_SUPPORTED; - link_set_buffer[0] = rate; - link_set_buffer[1] = lane_count_set.raw; - - core_link_write_dpcd(link, DP_LINK_BW_SET, - link_set_buffer, 2); core_link_write_dpcd(link, DP_DOWNSPREAD_CTRL, &downspread.raw, sizeof(downspread)); + core_link_write_dpcd(link, DP_LANE_COUNT_SET, + &lane_count_set.raw, 1); + if (link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_14 && - (link->dpcd_caps.link_rate_set >= 1 && - link->dpcd_caps.link_rate_set <= 8)) { + lt_settings->link_settings.use_link_rate_set == true) { + rate = 0; + core_link_write_dpcd(link, DP_LINK_BW_SET, &rate, 1); core_link_write_dpcd(link, DP_LINK_RATE_SET, - &link->dpcd_caps.link_rate_set, 1); + <_settings->link_settings.link_rate_set, 1); + } else { + rate = (uint8_t) (lt_settings->link_settings.link_rate); + core_link_write_dpcd(link, DP_LINK_BW_SET, &rate, 1); } - DC_LOG_HW_LINK_TRAINING("%s\n %x rate = %x\n %x lane = %x\n %x spread = %x\n", - __func__, - DP_LINK_BW_SET, - lt_settings->link_settings.link_rate, - DP_LANE_COUNT_SET, - lt_settings->link_settings.lane_count, - DP_DOWNSPREAD_CTRL, - lt_settings->link_settings.link_spread); + if (rate) { + DC_LOG_HW_LINK_TRAINING("%s\n %x rate = %x\n %x lane = %x\n %x spread = %x\n", + __func__, + DP_LINK_BW_SET, + lt_settings->link_settings.link_rate, + DP_LANE_COUNT_SET, + lt_settings->link_settings.lane_count, + DP_DOWNSPREAD_CTRL, + lt_settings->link_settings.link_spread); + } else { + DC_LOG_HW_LINK_TRAINING("%s\n %x rate set = %x\n %x lane = %x\n %x spread = %x\n", + __func__, + DP_LINK_RATE_SET, + lt_settings->link_settings.link_rate_set, + DP_LANE_COUNT_SET, + lt_settings->link_settings.lane_count, + DP_DOWNSPREAD_CTRL, + lt_settings->link_settings.link_spread); + } } @@ -952,6 +963,8 @@ enum link_training_result dc_link_dp_perform_link_training( lt_settings.link_settings.link_rate = link_setting->link_rate; lt_settings.link_settings.lane_count = link_setting->lane_count; + lt_settings.link_settings.use_link_rate_set = link_setting->use_link_rate_set; + lt_settings.link_settings.link_rate_set = link_setting->link_rate_set; /*@todo[vdevulap] move SS to LS, should not be handled by displaypath*/ @@ -1075,7 +1088,7 @@ static struct dc_link_settings get_max_link_cap(struct dc_link *link) { /* Set Default link settings */ struct dc_link_settings max_link_cap = {LANE_COUNT_FOUR, LINK_RATE_HIGH, - LINK_SPREAD_05_DOWNSPREAD_30KHZ}; + LINK_SPREAD_05_DOWNSPREAD_30KHZ, false, 0}; /* Higher link settings based on feature supported */ if (link->link_enc->features.flags.bits.IS_HBR2_CAPABLE) @@ -1520,69 +1533,6 @@ static bool decide_fallback_link_setting( return true; } -static uint32_t bandwidth_in_kbps_from_timing( - const struct dc_crtc_timing *timing) -{ - uint32_t bits_per_channel = 0; - uint32_t kbps; - - switch (timing->display_color_depth) { - case COLOR_DEPTH_666: - bits_per_channel = 6; - break; - case COLOR_DEPTH_888: - bits_per_channel = 8; - break; - case COLOR_DEPTH_101010: - bits_per_channel = 10; - break; - case COLOR_DEPTH_121212: - bits_per_channel = 12; - break; - case COLOR_DEPTH_141414: - bits_per_channel = 14; - break; - case COLOR_DEPTH_161616: - bits_per_channel = 16; - break; - default: - break; - } - - ASSERT(bits_per_channel != 0); - - kbps = timing->pix_clk_100hz / 10; - kbps *= bits_per_channel; - - if (timing->flags.Y_ONLY != 1) { - /*Only YOnly make reduce bandwidth by 1/3 compares to RGB*/ - kbps *= 3; - if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) - kbps /= 2; - else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) - kbps = kbps * 2 / 3; - } - - return kbps; - -} - -static uint32_t bandwidth_in_kbps_from_link_settings( - const struct dc_link_settings *link_setting) -{ - uint32_t link_rate_in_kbps = link_setting->link_rate * - LINK_RATE_REF_FREQ_IN_KHZ; - - uint32_t lane_count = link_setting->lane_count; - uint32_t kbps = link_rate_in_kbps; - - kbps *= lane_count; - kbps *= 8; /* 8 bits per byte*/ - - return kbps; - -} - bool dp_validate_mode_timing( struct dc_link *link, const struct dc_crtc_timing *timing) @@ -1598,8 +1548,7 @@ bool dp_validate_mode_timing( timing->v_addressable == (uint32_t) 480) return true; - /* We always use verified link settings */ - link_setting = &link->verified_link_cap; + link_setting = dc_link_get_link_cap(link); /* TODO: DYNAMIC_VALIDATION needs to be implemented */ /*if (flags.DYNAMIC_VALIDATION == 1 && @@ -1607,8 +1556,8 @@ bool dp_validate_mode_timing( link_setting = &link->verified_link_cap; */ - req_bw = bandwidth_in_kbps_from_timing(timing); - max_bw = bandwidth_in_kbps_from_link_settings(link_setting); + req_bw = dc_bandwidth_in_kbps_from_timing(timing); + max_bw = dc_link_bandwidth_kbps(link, link_setting); if (req_bw <= max_bw) { /* remember the biggest mode here, during @@ -1629,19 +1578,106 @@ bool dp_validate_mode_timing( return false; } +static bool decide_dp_link_settings(struct dc_link *link, struct dc_link_settings *link_setting, uint32_t req_bw) +{ + struct dc_link_settings initial_link_setting = { + LANE_COUNT_ONE, LINK_RATE_LOW, LINK_SPREAD_DISABLED, false, 0}; + struct dc_link_settings current_link_setting = + initial_link_setting; + uint32_t link_bw; + + /* search for the minimum link setting that: + * 1. is supported according to the link training result + * 2. could support the b/w requested by the timing + */ + while (current_link_setting.link_rate <= + link->verified_link_cap.link_rate) { + link_bw = dc_link_bandwidth_kbps( + link, + ¤t_link_setting); + if (req_bw <= link_bw) { + *link_setting = current_link_setting; + return true; + } + + if (current_link_setting.lane_count < + link->verified_link_cap.lane_count) { + current_link_setting.lane_count = + increase_lane_count( + current_link_setting.lane_count); + } else { + current_link_setting.link_rate = + increase_link_rate( + current_link_setting.link_rate); + current_link_setting.lane_count = + initial_link_setting.lane_count; + } + } + + return false; +} + +static bool decide_edp_link_settings(struct dc_link *link, struct dc_link_settings *link_setting, uint32_t req_bw) +{ + struct dc_link_settings initial_link_setting; + struct dc_link_settings current_link_setting; + uint32_t link_bw; + + if (link->dpcd_caps.dpcd_rev.raw < DPCD_REV_14 || + link->dpcd_caps.edp_supported_link_rates_count == 0 || + link->dc->config.optimize_edp_link_rate == false) { + *link_setting = link->verified_link_cap; + return true; + } + + memset(&initial_link_setting, 0, sizeof(initial_link_setting)); + initial_link_setting.lane_count = LANE_COUNT_ONE; + initial_link_setting.link_rate = link->dpcd_caps.edp_supported_link_rates[0]; + initial_link_setting.link_spread = LINK_SPREAD_DISABLED; + initial_link_setting.use_link_rate_set = true; + initial_link_setting.link_rate_set = 0; + current_link_setting = initial_link_setting; + + /* search for the minimum link setting that: + * 1. is supported according to the link training result + * 2. could support the b/w requested by the timing + */ + while (current_link_setting.link_rate <= + link->verified_link_cap.link_rate) { + link_bw = dc_link_bandwidth_kbps( + link, + ¤t_link_setting); + if (req_bw <= link_bw) { + *link_setting = current_link_setting; + return true; + } + + if (current_link_setting.lane_count < + link->verified_link_cap.lane_count) { + current_link_setting.lane_count = + increase_lane_count( + current_link_setting.lane_count); + } else { + if (current_link_setting.link_rate_set < link->dpcd_caps.edp_supported_link_rates_count) { + current_link_setting.link_rate_set++; + current_link_setting.link_rate = + link->dpcd_caps.edp_supported_link_rates[current_link_setting.link_rate_set]; + current_link_setting.lane_count = + initial_link_setting.lane_count; + } else + break; + } + } + return false; +} + void decide_link_settings(struct dc_stream_state *stream, struct dc_link_settings *link_setting) { - - struct dc_link_settings initial_link_setting = { - LANE_COUNT_ONE, LINK_RATE_LOW, LINK_SPREAD_DISABLED}; - struct dc_link_settings current_link_setting = - initial_link_setting; struct dc_link *link; uint32_t req_bw; - uint32_t link_bw; - req_bw = bandwidth_in_kbps_from_timing(&stream->timing); + req_bw = dc_bandwidth_in_kbps_from_timing(&stream->timing); link = stream->link; @@ -1664,38 +1700,11 @@ void decide_link_settings(struct dc_stream_state *stream, return; } - /* EDP use the link cap setting */ if (link->connector_signal == SIGNAL_TYPE_EDP) { - *link_setting = link->verified_link_cap; - return; - } - - /* search for the minimum link setting that: - * 1. is supported according to the link training result - * 2. could support the b/w requested by the timing - */ - while (current_link_setting.link_rate <= - link->verified_link_cap.link_rate) { - link_bw = bandwidth_in_kbps_from_link_settings( - ¤t_link_setting); - if (req_bw <= link_bw) { - *link_setting = current_link_setting; + if (decide_edp_link_settings(link, link_setting, req_bw)) return; - } - - if (current_link_setting.lane_count < - link->verified_link_cap.lane_count) { - current_link_setting.lane_count = - increase_lane_count( - current_link_setting.lane_count); - } else { - current_link_setting.link_rate = - increase_link_rate( - current_link_setting.link_rate); - current_link_setting.lane_count = - initial_link_setting.lane_count; - } - } + } else if (decide_dp_link_settings(link, link_setting, req_bw)) + return; BREAK_TO_DEBUGGER(); ASSERT(link->verified_link_cap.lane_count != LANE_COUNT_UNKNOWN); @@ -2155,11 +2164,7 @@ bool is_mst_supported(struct dc_link *link) bool is_dp_active_dongle(const struct dc_link *link) { - enum display_dongle_type dongle_type = link->dpcd_caps.dongle_type; - - return (dongle_type == DISPLAY_DONGLE_DP_VGA_CONVERTER) || - (dongle_type == DISPLAY_DONGLE_DP_DVI_CONVERTER) || - (dongle_type == DISPLAY_DONGLE_DP_HDMI_CONVERTER); + return link->dpcd_caps.is_branch_dev; } static int translate_dpcd_max_bpc(enum dpcd_downstream_port_max_bpc bpc) @@ -2193,6 +2198,9 @@ static void get_active_converter_info( return; } + /* DPCD 0x5 bit 0 = 1, it indicate it's branch device */ + link->dpcd_caps.is_branch_dev = ds_port.fields.PORT_PRESENT; + switch (ds_port.fields.PORT_TYPE) { case DOWNSTREAM_VGA: link->dpcd_caps.dongle_type = DISPLAY_DONGLE_DP_VGA_CONVERTER; @@ -2234,8 +2242,8 @@ static void get_active_converter_info( hdmi_caps = {.raw = det_caps[3] }; union dwnstream_port_caps_byte2 hdmi_color_caps = {.raw = det_caps[2] }; - link->dpcd_caps.dongle_caps.dp_hdmi_max_pixel_clk = - det_caps[1] * 25000; + link->dpcd_caps.dongle_caps.dp_hdmi_max_pixel_clk_in_khz = + det_caps[1] * 2500; link->dpcd_caps.dongle_caps.is_dp_hdmi_s3d_converter = hdmi_caps.bits.FRAME_SEQ_TO_FRAME_PACK; @@ -2252,7 +2260,7 @@ static void get_active_converter_info( translate_dpcd_max_bpc( hdmi_color_caps.bits.MAX_BITS_PER_COLOR_COMPONENT); - if (link->dpcd_caps.dongle_caps.dp_hdmi_max_pixel_clk != 0) + if (link->dpcd_caps.dongle_caps.dp_hdmi_max_pixel_clk_in_khz != 0) link->dpcd_caps.dongle_caps.extendedCapValid = true; } @@ -2347,6 +2355,10 @@ static bool retrieve_link_cap(struct dc_link *link) { uint8_t dpcd_data[DP_ADAPTER_CAP - DP_DPCD_REV + 1]; + /*Only need to read 1 byte starting from DP_DPRX_FEATURE_ENUMERATION_LIST. + */ + uint8_t dpcd_dprx_data = '\0'; + struct dp_device_vendor_id sink_id; union down_stream_port_count down_strm_port_count; union edp_configuration_cap edp_config_cap; @@ -2383,7 +2395,10 @@ static bool retrieve_link_cap(struct dc_link *link) aux_rd_interval.raw = dpcd_data[DP_TRAINING_AUX_RD_INTERVAL]; - if (aux_rd_interval.bits.EXT_RECIEVER_CAP_FIELD_PRESENT == 1) { + link->dpcd_caps.ext_receiver_cap_field_present = + aux_rd_interval.bits.EXT_RECEIVER_CAP_FIELD_PRESENT == 1 ? true:false; + + if (aux_rd_interval.bits.EXT_RECEIVER_CAP_FIELD_PRESENT == 1) { uint8_t ext_cap_data[16]; memset(ext_cap_data, '\0', sizeof(ext_cap_data)); @@ -2404,7 +2419,38 @@ static bool retrieve_link_cap(struct dc_link *link) } link->dpcd_caps.dpcd_rev.raw = - dpcd_data[DP_DPCD_REV - DP_DPCD_REV]; + dpcd_data[DP_DPCD_REV - DP_DPCD_REV]; + + if (link->dpcd_caps.dpcd_rev.raw >= 0x14) { + for (i = 0; i < read_dpcd_retry_cnt; i++) { + status = core_link_read_dpcd( + link, + DP_DPRX_FEATURE_ENUMERATION_LIST, + &dpcd_dprx_data, + sizeof(dpcd_dprx_data)); + if (status == DC_OK) + break; + } + + link->dpcd_caps.dprx_feature.raw = dpcd_dprx_data; + + if (status != DC_OK) + dm_error("%s: Read DPRX caps data failed.\n", __func__); + } + + else { + link->dpcd_caps.dprx_feature.raw = 0; + } + + + /* Error condition checking... + * It is impossible for Sink to report Max Lane Count = 0. + * It is possible for Sink to report Max Link Rate = 0, if it is + * an eDP device that is reporting specialized link rates in the + * SUPPORTED_LINK_RATE table. + */ + if (dpcd_data[DP_MAX_LANE_COUNT - DP_DPCD_REV] == 0) + return false; ds_port.byte = dpcd_data[DP_DOWNSTREAMPORT_PRESENT - DP_DPCD_REV]; @@ -2536,35 +2582,65 @@ enum dc_link_rate linkRateInKHzToLinkRateMultiplier(uint32_t link_rate_in_khz) void detect_edp_sink_caps(struct dc_link *link) { - uint8_t supported_link_rates[16] = {0}; + uint8_t supported_link_rates[16]; uint32_t entry; uint32_t link_rate_in_khz; enum dc_link_rate link_rate = LINK_RATE_UNKNOWN; + union lane_count_set lane_count_set = { {0} }; + uint8_t link_bw_set; + uint8_t link_rate_set; retrieve_link_cap(link); + link->dpcd_caps.edp_supported_link_rates_count = 0; + memset(supported_link_rates, 0, sizeof(supported_link_rates)); - if (link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_14) { + if (link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_14 && + link->dc->config.optimize_edp_link_rate) { // Read DPCD 00010h - 0001Fh 16 bytes at one shot core_link_read_dpcd(link, DP_SUPPORTED_LINK_RATES, supported_link_rates, sizeof(supported_link_rates)); - link->dpcd_caps.link_rate_set = 0; for (entry = 0; entry < 16; entry += 2) { // DPCD register reports per-lane link rate = 16-bit link rate capability - // value X 200 kHz. Need multipler to find link rate in kHz. + // value X 200 kHz. Need multiplier to find link rate in kHz. link_rate_in_khz = (supported_link_rates[entry+1] * 0x100 + supported_link_rates[entry]) * 200; if (link_rate_in_khz != 0) { link_rate = linkRateInKHzToLinkRateMultiplier(link_rate_in_khz); - if (link->reported_link_cap.link_rate < link_rate) { - link->reported_link_cap.link_rate = link_rate; - link->dpcd_caps.link_rate_set = entry; - } + link->dpcd_caps.edp_supported_link_rates[link->dpcd_caps.edp_supported_link_rates_count] = link_rate; + link->dpcd_caps.edp_supported_link_rates_count++; } } } link->verified_link_cap = link->reported_link_cap; + + // Read DPCD 00101h to find out the number of lanes currently set + core_link_read_dpcd(link, DP_LANE_COUNT_SET, + &lane_count_set.raw, sizeof(lane_count_set)); + link->cur_link_settings.lane_count = lane_count_set.bits.LANE_COUNT_SET; + + // Read DPCD 00100h to find if standard link rates are set + core_link_read_dpcd(link, DP_LINK_BW_SET, + &link_bw_set, sizeof(link_bw_set)); + + if (link_bw_set == 0) { + /* If standard link rates are not being used, + * Read DPCD 00115h to find the link rate set used + */ + core_link_read_dpcd(link, DP_LINK_RATE_SET, + &link_rate_set, sizeof(link_rate_set)); + + if (link_rate_set < link->dpcd_caps.edp_supported_link_rates_count) { + link->cur_link_settings.link_rate = + link->dpcd_caps.edp_supported_link_rates[link_rate_set]; + link->cur_link_settings.link_rate_set = link_rate_set; + link->cur_link_settings.use_link_rate_set = true; + } + } else { + link->cur_link_settings.link_rate = link_bw_set; + link->cur_link_settings.use_link_rate_set = false; + } } void dc_link_dp_enable_hpd(const struct dc_link *link) @@ -2601,6 +2677,7 @@ static void set_crtc_test_pattern(struct dc_link *link, enum dc_color_depth color_depth = pipe_ctx-> stream->timing.display_color_depth; struct bit_depth_reduction_params params; + struct output_pixel_processor *opp = pipe_ctx->stream_res.opp; memset(¶ms, 0, sizeof(params)); @@ -2640,8 +2717,7 @@ static void set_crtc_test_pattern(struct dc_link *link, { /* disable bit depth reduction */ pipe_ctx->stream->bit_depth_params = params; - pipe_ctx->stream_res.opp->funcs-> - opp_program_bit_depth_reduction(pipe_ctx->stream_res.opp, ¶ms); + opp->funcs->opp_program_bit_depth_reduction(opp, ¶ms); if (pipe_ctx->stream_res.tg->funcs->set_test_pattern) pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg, controller_test_pattern, color_depth); @@ -2650,11 +2726,9 @@ static void set_crtc_test_pattern(struct dc_link *link, case DP_TEST_PATTERN_VIDEO_MODE: { /* restore bitdepth reduction */ - resource_build_bit_depth_reduction_params(pipe_ctx->stream, - ¶ms); + resource_build_bit_depth_reduction_params(pipe_ctx->stream, ¶ms); pipe_ctx->stream->bit_depth_params = params; - pipe_ctx->stream_res.opp->funcs-> - opp_program_bit_depth_reduction(pipe_ctx->stream_res.opp, ¶ms); + opp->funcs->opp_program_bit_depth_reduction(opp, ¶ms); if (pipe_ctx->stream_res.tg->funcs->set_test_pattern) pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg, CONTROLLER_DP_TEST_PATTERN_VIDEOMODE, diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c index 349ab8017776..eac7186e4f08 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c @@ -31,6 +31,8 @@ #include "opp.h" #include "timing_generator.h" #include "transform.h" +#include "dccg.h" +#include "dchubbub.h" #include "dpp.h" #include "core_types.h" #include "set_mode_types.h" @@ -104,44 +106,43 @@ enum dce_version resource_parse_asic_id(struct hw_asic_id asic_id) return dc_version; } -struct resource_pool *dc_create_resource_pool( - struct dc *dc, - int num_virtual_links, - enum dce_version dc_version, - struct hw_asic_id asic_id) +struct resource_pool *dc_create_resource_pool(struct dc *dc, + const struct dc_init_data *init_data, + enum dce_version dc_version) { struct resource_pool *res_pool = NULL; switch (dc_version) { case DCE_VERSION_8_0: res_pool = dce80_create_resource_pool( - num_virtual_links, dc); + init_data->num_virtual_links, dc); break; case DCE_VERSION_8_1: res_pool = dce81_create_resource_pool( - num_virtual_links, dc); + init_data->num_virtual_links, dc); break; case DCE_VERSION_8_3: res_pool = dce83_create_resource_pool( - num_virtual_links, dc); + init_data->num_virtual_links, dc); break; case DCE_VERSION_10_0: res_pool = dce100_create_resource_pool( - num_virtual_links, dc); + init_data->num_virtual_links, dc); break; case DCE_VERSION_11_0: res_pool = dce110_create_resource_pool( - num_virtual_links, dc, asic_id); + init_data->num_virtual_links, dc, + init_data->asic_id); break; case DCE_VERSION_11_2: case DCE_VERSION_11_22: res_pool = dce112_create_resource_pool( - num_virtual_links, dc); + init_data->num_virtual_links, dc); break; case DCE_VERSION_12_0: case DCE_VERSION_12_1: res_pool = dce120_create_resource_pool( - num_virtual_links, dc); + init_data->num_virtual_links, dc); break; #if defined(CONFIG_DRM_AMD_DC_DCN1_0) @@ -149,8 +150,7 @@ struct resource_pool *dc_create_resource_pool( #if defined(CONFIG_DRM_AMD_DC_DCN1_01) case DCN_VERSION_1_01: #endif - res_pool = dcn10_create_resource_pool( - num_virtual_links, dc); + res_pool = dcn10_create_resource_pool(init_data, dc); break; #endif @@ -163,7 +163,28 @@ struct resource_pool *dc_create_resource_pool( if (dc->ctx->dc_bios->funcs->get_firmware_info( dc->ctx->dc_bios, &fw_info) == BP_RESULT_OK) { - res_pool->ref_clock_inKhz = fw_info.pll_info.crystal_frequency; + res_pool->ref_clocks.xtalin_clock_inKhz = fw_info.pll_info.crystal_frequency; + + if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { + // On FPGA these dividers are currently not configured by GDB + res_pool->ref_clocks.dccg_ref_clock_inKhz = res_pool->ref_clocks.xtalin_clock_inKhz; + res_pool->ref_clocks.dchub_ref_clock_inKhz = res_pool->ref_clocks.xtalin_clock_inKhz; + } else if (res_pool->dccg && res_pool->hubbub) { + // If DCCG reference frequency cannot be determined (usually means not set to xtalin) then this is a critical error + // as this value must be known for DCHUB programming + (res_pool->dccg->funcs->get_dccg_ref_freq)(res_pool->dccg, + fw_info.pll_info.crystal_frequency, + &res_pool->ref_clocks.dccg_ref_clock_inKhz); + + // Similarly, if DCHUB reference frequency cannot be determined, then it is also a critical error + (res_pool->hubbub->funcs->get_dchub_ref_freq)(res_pool->hubbub, + res_pool->ref_clocks.dccg_ref_clock_inKhz, + &res_pool->ref_clocks.dchub_ref_clock_inKhz); + } else { + // Not all ASICs have DCCG sw component + res_pool->ref_clocks.dccg_ref_clock_inKhz = res_pool->ref_clocks.xtalin_clock_inKhz; + res_pool->ref_clocks.dchub_ref_clock_inKhz = res_pool->ref_clocks.xtalin_clock_inKhz; + } } else ASSERT_CRITICAL(false); } @@ -260,6 +281,7 @@ bool resource_construct( pool->stream_enc_count++; } } + dc->caps.dynamic_audio = false; if (pool->audio_count < pool->stream_enc_count) { dc->caps.dynamic_audio = true; @@ -1014,24 +1036,60 @@ enum dc_status resource_build_scaling_params_for_context( struct pipe_ctx *find_idle_secondary_pipe( struct resource_context *res_ctx, - const struct resource_pool *pool) + const struct resource_pool *pool, + const struct pipe_ctx *primary_pipe) { int i; struct pipe_ctx *secondary_pipe = NULL; /* - * search backwards for the second pipe to keep pipe - * assignment more consistent + * We add a preferred pipe mapping to avoid the chance that + * MPCCs already in use will need to be reassigned to other trees. + * For example, if we went with the strict, assign backwards logic: + * + * (State 1) + * Display A on, no surface, top pipe = 0 + * Display B on, no surface, top pipe = 1 + * + * (State 2) + * Display A on, no surface, top pipe = 0 + * Display B on, surface enable, top pipe = 1, bottom pipe = 5 + * + * (State 3) + * Display A on, surface enable, top pipe = 0, bottom pipe = 5 + * Display B on, surface enable, top pipe = 1, bottom pipe = 4 + * + * The state 2->3 transition requires remapping MPCC 5 from display B + * to display A. + * + * However, with the preferred pipe logic, state 2 would look like: + * + * (State 2) + * Display A on, no surface, top pipe = 0 + * Display B on, surface enable, top pipe = 1, bottom pipe = 4 + * + * This would then cause 2->3 to not require remapping any MPCCs. */ - - for (i = pool->pipe_count - 1; i >= 0; i--) { - if (res_ctx->pipe_ctx[i].stream == NULL) { - secondary_pipe = &res_ctx->pipe_ctx[i]; - secondary_pipe->pipe_idx = i; - break; + if (primary_pipe) { + int preferred_pipe_idx = (pool->pipe_count - 1) - primary_pipe->pipe_idx; + if (res_ctx->pipe_ctx[preferred_pipe_idx].stream == NULL) { + secondary_pipe = &res_ctx->pipe_ctx[preferred_pipe_idx]; + secondary_pipe->pipe_idx = preferred_pipe_idx; } } + /* + * search backwards for the second pipe to keep pipe + * assignment more consistent + */ + if (!secondary_pipe) + for (i = pool->pipe_count - 1; i >= 0; i--) { + if (res_ctx->pipe_ctx[i].stream == NULL) { + secondary_pipe = &res_ctx->pipe_ctx[i]; + secondary_pipe->pipe_idx = i; + break; + } + } return secondary_pipe; } @@ -1214,6 +1272,9 @@ bool dc_add_plane_to_context( free_pipe->clock_source = tail_pipe->clock_source; free_pipe->top_pipe = tail_pipe; tail_pipe->bottom_pipe = free_pipe; + } else if (free_pipe->bottom_pipe && free_pipe->bottom_pipe->plane_state == NULL) { + ASSERT(free_pipe->bottom_pipe->stream_res.opp != free_pipe->stream_res.opp); + free_pipe->bottom_pipe->plane_state = plane_state; } /* assign new surfaces*/ @@ -1224,6 +1285,35 @@ bool dc_add_plane_to_context( return true; } +struct pipe_ctx *dc_res_get_odm_bottom_pipe(struct pipe_ctx *pipe_ctx) +{ + struct pipe_ctx *bottom_pipe = pipe_ctx->bottom_pipe; + + /* ODM should only be updated once per otg */ + if (pipe_ctx->top_pipe) + return NULL; + + while (bottom_pipe) { + if (bottom_pipe->stream_res.opp != pipe_ctx->stream_res.opp) + break; + bottom_pipe = bottom_pipe->bottom_pipe; + } + + return bottom_pipe; +} + +bool dc_res_is_odm_head_pipe(struct pipe_ctx *pipe_ctx) +{ + struct pipe_ctx *top_pipe = pipe_ctx->top_pipe; + + if (!top_pipe) + return false; + if (top_pipe && top_pipe->stream_res.opp == pipe_ctx->stream_res.opp) + return false; + + return true; +} + bool dc_remove_plane_from_context( const struct dc *dc, struct dc_stream_state *stream, @@ -1247,10 +1337,14 @@ bool dc_remove_plane_from_context( /* release pipe for plane*/ for (i = pool->pipe_count - 1; i >= 0; i--) { - struct pipe_ctx *pipe_ctx; + struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; - if (context->res_ctx.pipe_ctx[i].plane_state == plane_state) { - pipe_ctx = &context->res_ctx.pipe_ctx[i]; + if (pipe_ctx->plane_state == plane_state) { + if (dc_res_is_odm_head_pipe(pipe_ctx)) { + pipe_ctx->plane_state = NULL; + pipe_ctx->bottom_pipe = NULL; + continue; + } if (pipe_ctx->top_pipe) pipe_ctx->top_pipe->bottom_pipe = pipe_ctx->bottom_pipe; @@ -1268,8 +1362,9 @@ bool dc_remove_plane_from_context( */ if (!pipe_ctx->top_pipe) { pipe_ctx->plane_state = NULL; - pipe_ctx->bottom_pipe = NULL; - } else { + if (!dc_res_get_odm_bottom_pipe(pipe_ctx)) + pipe_ctx->bottom_pipe = NULL; + } else { memset(pipe_ctx, 0, sizeof(*pipe_ctx)); } } @@ -1674,6 +1769,9 @@ enum dc_status dc_remove_stream_from_ctx( for (i = 0; i < MAX_PIPES; i++) { if (new_ctx->res_ctx.pipe_ctx[i].stream == stream && !new_ctx->res_ctx.pipe_ctx[i].top_pipe) { + struct pipe_ctx *odm_pipe = + dc_res_get_odm_bottom_pipe(&new_ctx->res_ctx.pipe_ctx[i]); + del_pipe = &new_ctx->res_ctx.pipe_ctx[i]; ASSERT(del_pipe->stream_res.stream_enc); @@ -1698,6 +1796,8 @@ enum dc_status dc_remove_stream_from_ctx( dc->res_pool->funcs->remove_stream_from_ctx(dc, new_ctx, stream); memset(del_pipe, 0, sizeof(*del_pipe)); + if (odm_pipe) + memset(odm_pipe, 0, sizeof(*odm_pipe)); break; } @@ -1855,6 +1955,7 @@ enum dc_status resource_map_pool_resources( struct dc_context *dc_ctx = dc->ctx; struct pipe_ctx *pipe_ctx = NULL; int pipe_idx = -1; + struct dc_bios *dcb = dc->ctx->dc_bios; /* TODO Check if this is needed */ /*if (!resource_is_stream_unchanged(old_context, stream)) { @@ -1869,6 +1970,13 @@ enum dc_status resource_map_pool_resources( calculate_phy_pix_clks(stream); + /* TODO: Check Linux */ + if (dc->config.allow_seamless_boot_optimization && + !dcb->funcs->is_accelerated_mode(dcb)) { + if (dc_validate_seamless_boot_timing(dc, stream->sink, &stream->timing)) + stream->apply_seamless_boot_optimization = true; + } + if (stream->apply_seamless_boot_optimization) pipe_idx = acquire_resource_from_hw_enabled_state( &context->res_ctx, @@ -1951,7 +2059,7 @@ void dc_resource_state_construct( const struct dc *dc, struct dc_state *dst_ctx) { - dst_ctx->dccg = dc->res_pool->clk_mgr; + dst_ctx->clk_mgr = dc->res_pool->clk_mgr; } /** @@ -1959,12 +2067,14 @@ void dc_resource_state_construct( * Checks HW resource availability and bandwidth requirement. * @dc: dc struct for this driver * @new_ctx: state to be validated + * @fast_validate: set to true if only yes/no to support matters * * Return: DC_OK if the result can be programmed. Otherwise, an error code. */ enum dc_status dc_validate_global_state( struct dc *dc, - struct dc_state *new_ctx) + struct dc_state *new_ctx, + bool fast_validate) { enum dc_status result = DC_ERROR_UNEXPECTED; int i, j; @@ -2019,7 +2129,7 @@ enum dc_status dc_validate_global_state( result = resource_build_scaling_params_for_context(dc, new_ctx); if (result == DC_OK) - if (!dc->res_pool->funcs->validate_bandwidth(dc, new_ctx)) + if (!dc->res_pool->funcs->validate_bandwidth(dc, new_ctx, fast_validate)) result = DC_FAIL_BANDWIDTH_VALIDATE; return result; @@ -2315,6 +2425,21 @@ static void set_spd_info_packet( *info_packet = stream->vrr_infopacket; } +static void set_dp_sdp_info_packet( + struct dc_info_packet *info_packet, + struct dc_stream_state *stream) +{ + /* SPD info packet for custom sdp message */ + + /* Return if false. If true, + * set the corresponding bit in the info packet + */ + if (!stream->dpsdp_infopacket.valid) + return; + + *info_packet = stream->dpsdp_infopacket; +} + static void set_hdr_static_info_packet( struct dc_info_packet *info_packet, struct dc_stream_state *stream) @@ -2411,6 +2536,7 @@ void resource_build_info_frame(struct pipe_ctx *pipe_ctx) info->spd.valid = false; info->hdrsmd.valid = false; info->vsc.valid = false; + info->dpsdp.valid = false; signal = pipe_ctx->stream->signal; @@ -2430,6 +2556,8 @@ void resource_build_info_frame(struct pipe_ctx *pipe_ctx) set_spd_info_packet(&info->spd, pipe_ctx->stream); set_hdr_static_info_packet(&info->hdrsmd, pipe_ctx->stream); + + set_dp_sdp_info_packet(&info->dpsdp, pipe_ctx->stream); } patch_gamut_packet_checksum(&info->gamut); @@ -2657,10 +2785,11 @@ enum dc_status dc_validate_stream(struct dc *dc, struct dc_stream_state *stream) if (!tg->funcs->validate_timing(tg, &stream->timing)) res = DC_FAIL_CONTROLLER_VALIDATE; - if (res == DC_OK) + if (res == DC_OK) { if (!link->link_enc->funcs->validate_output_with_stream( link->link_enc, stream)) res = DC_FAIL_ENC_VALIDATE; + } /* TODO: validate audio ASIC caps, encoder */ diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c index 996298c35f42..e10479d58c11 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c @@ -29,6 +29,9 @@ #include "resource.h" #include "ipp.h" #include "timing_generator.h" +#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#include "dcn10/dcn10_hw_sequencer.h" +#endif #define DC_LOGGER dc->ctx->logger @@ -196,6 +199,35 @@ struct dc_stream_status *dc_stream_get_status( return dc_stream_get_status_from_state(dc->current_state, stream); } +static void delay_cursor_until_vupdate(struct pipe_ctx *pipe_ctx, struct dc *dc) +{ +#if defined(CONFIG_DRM_AMD_DC_DCN1_0) + unsigned int vupdate_line; + unsigned int lines_to_vupdate, us_to_vupdate, vpos, nvpos; + struct dc_stream_state *stream = pipe_ctx->stream; + unsigned int us_per_line; + + if (stream->ctx->asic_id.chip_family == FAMILY_RV && + ASIC_REV_IS_RAVEN(stream->ctx->asic_id.hw_internal_rev)) { + + vupdate_line = get_vupdate_offset_from_vsync(pipe_ctx); + if (!dc_stream_get_crtc_position(dc, &stream, 1, &vpos, &nvpos)) + return; + + if (vpos >= vupdate_line) + return; + + us_per_line = stream->timing.h_total * 10000 / stream->timing.pix_clk_100hz; + lines_to_vupdate = vupdate_line - vpos; + us_to_vupdate = lines_to_vupdate * us_per_line; + + /* 70 us is a conservative estimate of cursor update time*/ + if (us_to_vupdate < 70) + udelay(us_to_vupdate); + } +#endif +} + /** * dc_stream_set_cursor_attributes() - Update cursor attributes and set cursor surface address */ @@ -234,6 +266,8 @@ bool dc_stream_set_cursor_attributes( if (!pipe_to_program) { pipe_to_program = pipe_ctx; + + delay_cursor_until_vupdate(pipe_ctx, core_dc); core_dc->hwss.pipe_control_lock(core_dc, pipe_to_program, true); } @@ -283,6 +317,8 @@ bool dc_stream_set_cursor_position( if (!pipe_to_program) { pipe_to_program = pipe_ctx; + + delay_cursor_until_vupdate(pipe_ctx, core_dc); core_dc->hwss.pipe_control_lock(core_dc, pipe_to_program, true); } @@ -314,6 +350,68 @@ uint32_t dc_stream_get_vblank_counter(const struct dc_stream_state *stream) return 0; } +static void build_dp_sdp_info_frame(struct pipe_ctx *pipe_ctx, + const uint8_t *custom_sdp_message, + unsigned int sdp_message_size) +{ + uint8_t i; + struct encoder_info_frame *info = &pipe_ctx->stream_res.encoder_info_frame; + + /* set valid info */ + info->dpsdp.valid = true; + + /* set sdp message header */ + info->dpsdp.hb0 = custom_sdp_message[0]; /* package id */ + info->dpsdp.hb1 = custom_sdp_message[1]; /* package type */ + info->dpsdp.hb2 = custom_sdp_message[2]; /* package specific byte 0 any data */ + info->dpsdp.hb3 = custom_sdp_message[3]; /* package specific byte 0 any data */ + + /* set sdp message data */ + for (i = 0; i < 32; i++) + info->dpsdp.sb[i] = (custom_sdp_message[i+4]); + +} + +static void invalid_dp_sdp_info_frame(struct pipe_ctx *pipe_ctx) +{ + struct encoder_info_frame *info = &pipe_ctx->stream_res.encoder_info_frame; + + /* in-valid info */ + info->dpsdp.valid = false; +} + +bool dc_stream_send_dp_sdp(const struct dc_stream_state *stream, + const uint8_t *custom_sdp_message, + unsigned int sdp_message_size) +{ + int i; + struct dc *core_dc; + struct resource_context *res_ctx; + + if (stream == NULL) { + dm_error("DC: dc_stream is NULL!\n"); + return false; + } + + core_dc = stream->ctx->dc; + res_ctx = &core_dc->current_state->res_ctx; + + for (i = 0; i < MAX_PIPES; i++) { + struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i]; + + if (pipe_ctx->stream != stream) + continue; + + build_dp_sdp_info_frame(pipe_ctx, custom_sdp_message, sdp_message_size); + + core_dc->hwss.update_info_frame(pipe_ctx); + + invalid_dp_sdp_info_frame(pipe_ctx); + } + + return true; +} + bool dc_stream_get_scanoutpos(const struct dc_stream_state *stream, uint32_t *v_blank_start, uint32_t *v_blank_end, diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c index ee6bd50f60b8..a5e86f9b148f 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c @@ -119,6 +119,19 @@ const struct dc_plane_status *dc_plane_get_status( if (core_dc->current_state == NULL) return NULL; + /* Find the current plane state and set its pending bit to false */ + for (i = 0; i < core_dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe_ctx = + &core_dc->current_state->res_ctx.pipe_ctx[i]; + + if (pipe_ctx->plane_state != plane_state) + continue; + + pipe_ctx->plane_state->status.is_flip_pending = false; + + break; + } + for (i = 0; i < core_dc->res_pool->pipe_count; i++) { struct pipe_ctx *pipe_ctx = &core_dc->current_state->res_ctx.pipe_ctx[i]; diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 1a7fd6aa77eb..3459e39714bc 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -39,9 +39,10 @@ #include "inc/hw/dmcu.h" #include "dml/display_mode_lib.h" -#define DC_VER "3.2.17" +#define DC_VER "3.2.26" #define MAX_SURFACES 3 +#define MAX_PLANES 6 #define MAX_STREAMS 6 #define MAX_SINKS_PER_LINK 4 @@ -53,6 +54,41 @@ struct dc_versions { struct dmcu_version dmcu_version; }; +enum dc_plane_type { + DC_PLANE_TYPE_INVALID, + DC_PLANE_TYPE_DCE_RGB, + DC_PLANE_TYPE_DCE_UNDERLAY, + DC_PLANE_TYPE_DCN_UNIVERSAL, +}; + +struct dc_plane_cap { + enum dc_plane_type type; + uint32_t blends_with_above : 1; + uint32_t blends_with_below : 1; + uint32_t per_pixel_alpha : 1; + struct { + uint32_t argb8888 : 1; + uint32_t nv12 : 1; + uint32_t fp16 : 1; + } pixel_format_support; + // max upscaling factor x1000 + // upscaling factors are always >= 1 + // for example, 1080p -> 8K is 4.0, or 4000 raw value + struct { + uint32_t argb8888; + uint32_t nv12; + uint32_t fp16; + } max_upscale_factor; + // max downscale factor x1000 + // downscale factors are always <= 1 + // for example, 8K -> 1080p is 0.25, or 250 raw value + struct { + uint32_t argb8888; + uint32_t nv12; + uint32_t fp16; + } max_downscale_factor; +}; + struct dc_caps { uint32_t max_streams; uint32_t max_links; @@ -73,6 +109,7 @@ struct dc_caps { bool force_dp_tps4_for_cp2520; bool disable_dp_clk_share; bool psp_setup_panel_mode; + struct dc_plane_cap planes[MAX_PLANES]; }; struct dc_dcc_surface_param { @@ -164,6 +201,9 @@ struct dc_config { bool gpu_vm_support; bool disable_disp_pll_sharing; bool fbc_support; + bool optimize_edp_link_rate; + bool disable_fractional_pwm; + bool allow_seamless_boot_optimization; }; enum visual_confirm { @@ -203,8 +243,60 @@ struct dc_clocks { int fclk_khz; int phyclk_khz; int dramclk_khz; + bool p_state_change_support; }; +struct dc_bw_validation_profile { + bool enable; + + unsigned long long total_ticks; + unsigned long long voltage_level_ticks; + unsigned long long watermark_ticks; + unsigned long long rq_dlg_ticks; + + unsigned long long total_count; + unsigned long long skip_fast_count; + unsigned long long skip_pass_count; + unsigned long long skip_fail_count; +}; + +#define BW_VAL_TRACE_SETUP() \ + unsigned long long end_tick = 0; \ + unsigned long long voltage_level_tick = 0; \ + unsigned long long watermark_tick = 0; \ + unsigned long long start_tick = dc->debug.bw_val_profile.enable ? \ + dm_get_timestamp(dc->ctx) : 0 + +#define BW_VAL_TRACE_COUNT() \ + if (dc->debug.bw_val_profile.enable) \ + dc->debug.bw_val_profile.total_count++ + +#define BW_VAL_TRACE_SKIP(status) \ + if (dc->debug.bw_val_profile.enable) { \ + if (!voltage_level_tick) \ + voltage_level_tick = dm_get_timestamp(dc->ctx); \ + dc->debug.bw_val_profile.skip_ ## status ## _count++; \ + } + +#define BW_VAL_TRACE_END_VOLTAGE_LEVEL() \ + if (dc->debug.bw_val_profile.enable) \ + voltage_level_tick = dm_get_timestamp(dc->ctx) + +#define BW_VAL_TRACE_END_WATERMARKS() \ + if (dc->debug.bw_val_profile.enable) \ + watermark_tick = dm_get_timestamp(dc->ctx) + +#define BW_VAL_TRACE_FINISH() \ + if (dc->debug.bw_val_profile.enable) { \ + end_tick = dm_get_timestamp(dc->ctx); \ + dc->debug.bw_val_profile.total_ticks += end_tick - start_tick; \ + dc->debug.bw_val_profile.voltage_level_ticks += voltage_level_tick - start_tick; \ + if (watermark_tick) { \ + dc->debug.bw_val_profile.watermark_ticks += watermark_tick - voltage_level_tick; \ + dc->debug.bw_val_profile.rq_dlg_ticks += end_tick - watermark_tick; \ + } \ + } + struct dc_debug_options { enum visual_confirm visual_confirm; bool sanity_checks; @@ -257,6 +349,8 @@ struct dc_debug_options { bool skip_detection_link_training; unsigned int force_odm_combine; //bit vector based on otg inst unsigned int force_fclk_khz; + bool disable_tri_buf; + struct dc_bw_validation_profile bw_val_profile; }; struct dc_debug_data { @@ -265,6 +359,14 @@ struct dc_debug_data { uint32_t auxErrorCount; }; +struct dc_bounding_box_overrides { + int sr_exit_time_ns; + int sr_enter_plus_exit_time_ns; + int urgent_latency_ns; + int percent_of_ideal_drambw; + int dram_clock_change_latency_ns; +}; + struct dc_state; struct resource_pool; struct dce_hwseq; @@ -274,6 +376,7 @@ struct dc { struct dc_cap_funcs cap_funcs; struct dc_config config; struct dc_debug_options debug; + struct dc_bounding_box_overrides bb_overrides; struct dc_context *ctx; uint8_t link_count; @@ -298,8 +401,12 @@ struct dc { struct hw_sequencer_funcs hwss; struct dce_hwseq *hwseq; + /* Require to optimize clocks and bandwidth for added/removed planes */ bool optimized_required; + /* Require to maintain clocks and bandwidth for UEFI enabled HW */ + bool optimize_seamless_boot; + /* FBC compressor */ struct compressor *fbc_compressor; @@ -327,6 +434,7 @@ struct dc_init_data { struct hw_asic_id asic_id; void *driver; /* ctx */ struct cgs_device *cgs_device; + struct dc_bounding_box_overrides bb_overrides; int num_virtual_links; /* @@ -594,7 +702,7 @@ struct dc_validation_set { uint8_t plane_count; }; -bool dc_validate_seamless_boot_timing(struct dc *dc, +bool dc_validate_seamless_boot_timing(const struct dc *dc, const struct dc_sink *sink, struct dc_crtc_timing *crtc_timing); @@ -602,9 +710,14 @@ enum dc_status dc_validate_plane(struct dc *dc, const struct dc_plane_state *pla void get_clock_requirements_for_state(struct dc_state *state, struct AsicStateEx *info); +/* + * fast_validate: we return after determining if we can support the new state, + * but before we populate the programming info + */ enum dc_status dc_validate_global_state( struct dc *dc, - struct dc_state *new_ctx); + struct dc_state *new_ctx, + bool fast_validate); void dc_resource_state_construct( @@ -633,7 +746,8 @@ void dc_resource_state_destruct(struct dc_state *context); bool dc_commit_state(struct dc *dc, struct dc_state *context); -struct dc_state *dc_create_state(void); +struct dc_state *dc_create_state(struct dc *dc); +struct dc_state *dc_copy_state(struct dc_state *src_ctx); void dc_retain_state(struct dc_state *context); void dc_release_state(struct dc_state *context); @@ -645,9 +759,16 @@ struct dpcd_caps { union dpcd_rev dpcd_rev; union max_lane_count max_ln_count; union max_down_spread max_down_spread; + union dprx_feature dprx_feature; + + /* valid only for eDP v1.4 or higher*/ + uint8_t edp_supported_link_rates_count; + enum dc_link_rate edp_supported_link_rates[8]; /* dongle type (DP converter, CV smart dongle) */ enum display_dongle_type dongle_type; + /* branch device or sink device */ + bool is_branch_dev; /* Dongle's downstream count. */ union sink_count sink_count; /* If dongle_type == DISPLAY_DONGLE_DP_HDMI_CONVERTER, @@ -663,11 +784,11 @@ struct dpcd_caps { int8_t branch_dev_name[6]; int8_t branch_hw_revision; int8_t branch_fw_revision[2]; - uint8_t link_rate_set; bool allow_invalid_MSA_timing_param; bool panel_mode_edp; bool dpcd_display_control_capable; + bool ext_receiver_cap_field_present; }; #include "dc_link.h" diff --git a/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h b/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h index 05c8c31d8b31..4ef97f65e55d 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h @@ -68,6 +68,8 @@ enum aux_transaction_reply { AUX_TRANSACTION_REPLY_AUX_ACK = 0x00, AUX_TRANSACTION_REPLY_AUX_NACK = 0x01, AUX_TRANSACTION_REPLY_AUX_DEFER = 0x02, + AUX_TRANSACTION_REPLY_I2C_OVER_AUX_NACK = 0x04, + AUX_TRANSACTION_REPLY_I2C_OVER_AUX_DEFER = 0x08, AUX_TRANSACTION_REPLY_I2C_ACK = 0x00, AUX_TRANSACTION_REPLY_I2C_NACK = 0x10, diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h index d4eab33c453b..11c68a399267 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h @@ -94,6 +94,8 @@ struct dc_link_settings { enum dc_lane_count lane_count; enum dc_link_rate link_rate; enum dc_link_spread link_spread; + bool use_link_rate_set; + uint8_t link_rate_set; }; struct dc_lane_settings { @@ -420,10 +422,24 @@ union edp_configuration_cap { uint8_t raw; }; +union dprx_feature { + struct { + uint8_t GTC_CAP:1; // bit 0: DP 1.3+ + uint8_t SST_SPLIT_SDP_CAP:1; // bit 1: DP 1.4 + uint8_t AV_SYNC_CAP:1; // bit 2: DP 1.3+ + uint8_t VSC_SDP_COLORIMETRY_SUPPORTED:1; // bit 3: DP 1.3+ + uint8_t VSC_EXT_VESA_SDP_SUPPORTED:1; // bit 4: DP 1.4 + uint8_t VSC_EXT_VESA_SDP_CHAINING_SUPPORTED:1; // bit 5: DP 1.4 + uint8_t VSC_EXT_CEA_SDP_SUPPORTED:1; // bit 6: DP 1.4 + uint8_t VSC_EXT_CEA_SDP_CHAINING_SUPPORTED:1; // bit 7: DP 1.4 + } bits; + uint8_t raw; +}; + union training_aux_rd_interval { struct { uint8_t TRAINIG_AUX_RD_INTERVAL:7; - uint8_t EXT_RECIEVER_CAP_FIELD_PRESENT:1; + uint8_t EXT_RECEIVER_CAP_FIELD_PRESENT:1; } bits; uint8_t raw; }; diff --git a/drivers/gpu/drm/amd/display/dc/dc_helper.c b/drivers/gpu/drm/amd/display/dc/dc_helper.c index 597d38393379..5e6c5eff49cf 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_helper.c +++ b/drivers/gpu/drm/amd/display/dc/dc_helper.c @@ -51,20 +51,16 @@ static inline void set_reg_field_value_masks( field_value_mask->mask = field_value_mask->mask | mask; } -uint32_t generic_reg_update_ex(const struct dc_context *ctx, - uint32_t addr, uint32_t reg_val, int n, +static void set_reg_field_values(struct dc_reg_value_masks *field_value_mask, + uint32_t addr, int n, uint8_t shift1, uint32_t mask1, uint32_t field_value1, - ...) + va_list ap) { - struct dc_reg_value_masks field_value_mask = {0}; uint32_t shift, mask, field_value; int i = 1; - va_list ap; - va_start(ap, field_value1); - /* gather all bits value/mask getting updated in this register */ - set_reg_field_value_masks(&field_value_mask, + set_reg_field_value_masks(field_value_mask, field_value1, mask1, shift1); while (i < n) { @@ -72,10 +68,48 @@ uint32_t generic_reg_update_ex(const struct dc_context *ctx, mask = va_arg(ap, uint32_t); field_value = va_arg(ap, uint32_t); - set_reg_field_value_masks(&field_value_mask, + set_reg_field_value_masks(field_value_mask, field_value, mask, shift); i++; } +} + +uint32_t generic_reg_update_ex(const struct dc_context *ctx, + uint32_t addr, int n, + uint8_t shift1, uint32_t mask1, uint32_t field_value1, + ...) +{ + struct dc_reg_value_masks field_value_mask = {0}; + uint32_t reg_val; + va_list ap; + + va_start(ap, field_value1); + + set_reg_field_values(&field_value_mask, addr, n, shift1, mask1, + field_value1, ap); + + va_end(ap); + + /* mmio write directly */ + reg_val = dm_read_reg(ctx, addr); + reg_val = (reg_val & ~field_value_mask.mask) | field_value_mask.value; + dm_write_reg(ctx, addr, reg_val); + return reg_val; +} + +uint32_t generic_reg_set_ex(const struct dc_context *ctx, + uint32_t addr, uint32_t reg_val, int n, + uint8_t shift1, uint32_t mask1, uint32_t field_value1, + ...) +{ + struct dc_reg_value_masks field_value_mask = {0}; + va_list ap; + + va_start(ap, field_value1); + + set_reg_field_values(&field_value_mask, addr, n, shift1, mask1, + field_value1, ap); + va_end(ap); @@ -85,6 +119,24 @@ uint32_t generic_reg_update_ex(const struct dc_context *ctx, return reg_val; } +uint32_t dm_read_reg_func( + const struct dc_context *ctx, + uint32_t address, + const char *func_name) +{ + uint32_t value; +#ifdef DM_CHECK_ADDR_0 + if (address == 0) { + DC_ERR("invalid register read; address = 0\n"); + return 0; + } +#endif + value = cgs_read_register(ctx->cgs_device, address); + trace_amdgpu_dc_rreg(&ctx->perf_trace->read_count, address, value); + + return value; +} + uint32_t generic_reg_get(const struct dc_context *ctx, uint32_t addr, uint8_t shift, uint32_t mask, uint32_t *field_value) { @@ -235,7 +287,7 @@ uint32_t generic_reg_get(const struct dc_context *ctx, } */ -uint32_t generic_reg_wait(const struct dc_context *ctx, +void generic_reg_wait(const struct dc_context *ctx, uint32_t addr, uint32_t shift, uint32_t mask, uint32_t condition_value, unsigned int delay_between_poll_us, unsigned int time_out_num_tries, const char *func_name, int line) @@ -265,7 +317,7 @@ uint32_t generic_reg_wait(const struct dc_context *ctx, DC_LOG_DC("REG_WAIT taking a while: %dms in %s line:%d\n", delay_between_poll_us * i / 1000, func_name, line); - return reg_val; + return; } } @@ -275,8 +327,6 @@ uint32_t generic_reg_wait(const struct dc_context *ctx, if (!IS_FPGA_MAXIMUS_DC(ctx->dce_environment)) BREAK_TO_DEBUGGER(); - - return reg_val; } void generic_write_indirect_reg(const struct dc_context *ctx, diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h index 8fc223defed4..cc7ffac64c96 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_link.h +++ b/drivers/gpu/drm/amd/display/dc/dc_link.h @@ -246,10 +246,18 @@ void dc_link_set_test_pattern(struct dc_link *link, const struct link_training_settings *p_link_settings, const unsigned char *p_custom_pattern, unsigned int cust_pattern_size); +uint32_t dc_link_bandwidth_kbps( + const struct dc_link *link, + const struct dc_link_settings *link_setting); + +const struct dc_link_settings *dc_link_get_link_cap( + const struct dc_link *link); bool dc_submit_i2c( struct dc *dc, uint32_t link_index, struct i2c_command *cmd); +uint32_t dc_bandwidth_in_kbps_from_timing( + const struct dc_crtc_timing *timing); #endif /* DC_LINK_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h index 5657cb3a2ad3..17fa3bf6cf7b 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_stream.h +++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h @@ -80,6 +80,7 @@ struct dc_stream_state { struct dc_info_packet vrr_infopacket; struct dc_info_packet vsc_infopacket; struct dc_info_packet vsp_infopacket; + struct dc_info_packet dpsdp_infopacket; struct rect src; /* composition area */ struct rect dst; /* stream addressable area */ @@ -221,6 +222,13 @@ struct dc_stream_state *dc_get_stream_at_index(struct dc *dc, uint8_t i); */ uint32_t dc_stream_get_vblank_counter(const struct dc_stream_state *stream); +/* + * Send dp sdp message. + */ +bool dc_stream_send_dp_sdp(const struct dc_stream_state *stream, + const uint8_t *custom_sdp_message, + unsigned int sdp_message_size); + /* TODO: Return parsed values rather than direct register read * This has a dependency on the caller (amdgpu_display_get_crtc_scanoutpos) * being refactored properly to be dce-specific diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h index da2009a108cf..6c2a3d9a4c2e 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_types.h @@ -103,7 +103,7 @@ struct dc_context { }; -#define DC_MAX_EDID_BUFFER_SIZE 512 +#define DC_MAX_EDID_BUFFER_SIZE 1024 #define EDID_BLOCK_SIZE 128 #define MAX_SURFACE_NUM 4 #define NUM_PIXEL_FORMATS 10 @@ -395,7 +395,7 @@ struct dc_dongle_caps { bool is_dp_hdmi_ycbcr422_converter; bool is_dp_hdmi_ycbcr420_converter; uint32_t dp_hdmi_max_bpc; - uint32_t dp_hdmi_max_pixel_clk; + uint32_t dp_hdmi_max_pixel_clk_in_khz; }; /* Scaling format */ enum scaling_transformation { @@ -550,9 +550,9 @@ struct psr_config { unsigned char psr_version; unsigned int psr_rfb_setup_time; bool psr_exit_link_training_required; - bool psr_frame_capture_indication_req; unsigned int psr_sdp_transmit_line_num_deadline; + bool allow_smu_optimizations; }; union dmcu_psr_level { @@ -654,6 +654,7 @@ struct psr_context { * continuing powerd own */ unsigned int frame_delay; + bool allow_smu_optimizations; }; struct colorspace_transform { diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c index da96229db53a..855360b1414f 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c @@ -50,6 +50,7 @@ #define MCP_ABM_LEVEL_SET 0x65 #define MCP_ABM_PIPE_SET 0x66 #define MCP_BL_SET 0x67 +#define MCP_BL_SET_PWM_FRAC 0x6A /* Enable or disable Fractional PWM */ #define MCP_DISABLE_ABM_IMMEDIATELY 255 @@ -390,6 +391,23 @@ static bool dce_abm_init_backlight(struct abm *abm) REG_UPDATE(BL_PWM_GRP1_REG_LOCK, BL_PWM_GRP1_REG_LOCK, 0); + /* Wait until microcontroller is ready to process interrupt */ + REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0, 100, 800); + + /* Set PWM fractional enable/disable */ + value = (abm->ctx->dc->config.disable_fractional_pwm == false) ? 1 : 0; + REG_WRITE(MASTER_COMM_DATA_REG1, value); + + /* Set command to enable or disable fractional PWM microcontroller */ + REG_UPDATE(MASTER_COMM_CMD_REG, MASTER_COMM_CMD_REG_BYTE0, + MCP_BL_SET_PWM_FRAC); + + /* Notify microcontroller of new command */ + REG_UPDATE(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 1); + + /* Ensure command has been executed before continuing */ + REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0, 100, 800); + return true; } diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c index 4febf4ef7240..937b5cffd7ef 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c @@ -171,24 +171,24 @@ static void submit_channel_request( (request->action == I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT))); if (REG(AUXN_IMPCAL)) { /* clear_aux_error */ - REG_UPDATE_SEQ(AUXN_IMPCAL, AUXN_CALOUT_ERROR_AK, - 1, - 0); + REG_UPDATE_SEQ_2(AUXN_IMPCAL, + AUXN_CALOUT_ERROR_AK, 1, + AUXN_CALOUT_ERROR_AK, 0); - REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_CALOUT_ERROR_AK, - 1, - 0); + REG_UPDATE_SEQ_2(AUXP_IMPCAL, + AUXP_CALOUT_ERROR_AK, 1, + AUXP_CALOUT_ERROR_AK, 0); /* force_default_calibrate */ - REG_UPDATE_1BY1_2(AUXN_IMPCAL, + REG_UPDATE_SEQ_2(AUXN_IMPCAL, AUXN_IMPCAL_ENABLE, 1, AUXN_IMPCAL_OVERRIDE_ENABLE, 0); /* bug? why AUXN update EN and OVERRIDE_EN 1 by 1 while AUX P toggles OVERRIDE? */ - REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_IMPCAL_OVERRIDE_ENABLE, - 1, - 0); + REG_UPDATE_SEQ_2(AUXP_IMPCAL, + AUXP_IMPCAL_OVERRIDE_ENABLE, 1, + AUXP_IMPCAL_OVERRIDE_ENABLE, 0); } /* set the delay and the number of bytes to write */ @@ -267,7 +267,7 @@ static int read_channel_reply(struct dce_aux *engine, uint32_t size, if (!bytes_replied) return -1; - REG_UPDATE_1BY1_3(AUX_SW_DATA, + REG_UPDATE_SEQ_3(AUX_SW_DATA, AUX_SW_INDEX, 0, AUX_SW_AUTOINCREMENT_DISABLE, 1, AUX_SW_DATA_RW, 1); @@ -317,9 +317,10 @@ static enum aux_channel_operation_result get_channel_status( *returned_bytes = 0; /* poll to make sure that SW_DONE is asserted */ - value = REG_WAIT(AUX_SW_STATUS, AUX_SW_DONE, 1, + REG_WAIT(AUX_SW_STATUS, AUX_SW_DONE, 1, 10, aux110->timeout_period/10); + value = REG_READ(AUX_SW_STATUS); /* in case HPD is LOW, exit AUX transaction */ if ((value & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) return AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON; @@ -374,7 +375,6 @@ static bool acquire( struct dce_aux *engine, struct ddc *ddc) { - enum gpio_result result; if (!is_engine_available(engine)) @@ -439,12 +439,12 @@ static enum i2caux_transaction_action i2caux_action_from_payload(struct aux_payl return I2CAUX_TRANSACTION_ACTION_DP_READ; } -int dce_aux_transfer(struct ddc_service *ddc, - struct aux_payload *payload) +int dce_aux_transfer_raw(struct ddc_service *ddc, + struct aux_payload *payload, + enum aux_channel_operation_result *operation_result) { struct ddc *ddc_pin = ddc->ddc_pin; struct dce_aux *aux_engine; - enum aux_channel_operation_result operation_result; struct aux_request_transaction_data aux_req; struct aux_reply_transaction_data aux_rep; uint8_t returned_bytes = 0; @@ -455,7 +455,8 @@ int dce_aux_transfer(struct ddc_service *ddc, memset(&aux_rep, 0, sizeof(aux_rep)); aux_engine = ddc->ctx->dc->res_pool->engines[ddc_pin->pin_data->en]; - acquire(aux_engine, ddc_pin); + if (!acquire(aux_engine, ddc_pin)) + return -1; if (payload->i2c_over_aux) aux_req.type = AUX_TRANSACTION_TYPE_I2C; @@ -470,28 +471,26 @@ int dce_aux_transfer(struct ddc_service *ddc, aux_req.data = payload->data; submit_channel_request(aux_engine, &aux_req); - operation_result = get_channel_status(aux_engine, &returned_bytes); + *operation_result = get_channel_status(aux_engine, &returned_bytes); - switch (operation_result) { - case AUX_CHANNEL_OPERATION_SUCCEEDED: - res = read_channel_reply(aux_engine, payload->length, - payload->data, payload->reply, - &status); - break; - case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON: - res = 0; - break; - case AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN: - case AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY: - case AUX_CHANNEL_OPERATION_FAILED_TIMEOUT: + if (*operation_result == AUX_CHANNEL_OPERATION_SUCCEEDED) { + read_channel_reply(aux_engine, payload->length, + payload->data, payload->reply, + &status); + res = returned_bytes; + } else { res = -1; - break; } + release_engine(aux_engine); return res; } -#define AUX_RETRY_MAX 7 +#define AUX_MAX_RETRIES 7 +#define AUX_MAX_DEFER_RETRIES 7 +#define AUX_MAX_I2C_DEFER_RETRIES 7 +#define AUX_MAX_INVALID_REPLY_RETRIES 2 +#define AUX_MAX_TIMEOUT_RETRIES 3 bool dce_aux_transfer_with_retries(struct ddc_service *ddc, struct aux_payload *payload) @@ -499,24 +498,85 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc, int i, ret = 0; uint8_t reply; bool payload_reply = true; + enum aux_channel_operation_result operation_result; + int aux_ack_retries = 0, + aux_defer_retries = 0, + aux_i2c_defer_retries = 0, + aux_timeout_retries = 0, + aux_invalid_reply_retries = 0; if (!payload->reply) { payload_reply = false; payload->reply = &reply; } - for (i = 0; i < AUX_RETRY_MAX; i++) { - ret = dce_aux_transfer(ddc, payload); + for (i = 0; i < AUX_MAX_RETRIES; i++) { + ret = dce_aux_transfer_raw(ddc, payload, &operation_result); + switch (operation_result) { + case AUX_CHANNEL_OPERATION_SUCCEEDED: + aux_timeout_retries = 0; + aux_invalid_reply_retries = 0; - if (ret >= 0) { - if (*payload->reply == 0) { - if (!payload_reply) - payload->reply = NULL; - return true; + switch (*payload->reply) { + case AUX_TRANSACTION_REPLY_AUX_ACK: + if (!payload->write && payload->length != ret) { + if (++aux_ack_retries >= AUX_MAX_RETRIES) + goto fail; + else + udelay(300); + } else + return true; + break; + + case AUX_TRANSACTION_REPLY_AUX_DEFER: + case AUX_TRANSACTION_REPLY_I2C_OVER_AUX_NACK: + case AUX_TRANSACTION_REPLY_I2C_OVER_AUX_DEFER: + if (++aux_defer_retries >= AUX_MAX_DEFER_RETRIES) + goto fail; + break; + + case AUX_TRANSACTION_REPLY_I2C_DEFER: + aux_defer_retries = 0; + if (++aux_i2c_defer_retries >= AUX_MAX_I2C_DEFER_RETRIES) + goto fail; + break; + + case AUX_TRANSACTION_REPLY_AUX_NACK: + case AUX_TRANSACTION_REPLY_HPD_DISCON: + default: + goto fail; } - } + break; - udelay(1000); + case AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY: + if (++aux_invalid_reply_retries >= AUX_MAX_INVALID_REPLY_RETRIES) + goto fail; + else + udelay(400); + break; + + case AUX_CHANNEL_OPERATION_FAILED_TIMEOUT: + if (++aux_timeout_retries >= AUX_MAX_TIMEOUT_RETRIES) + goto fail; + else { + /* + * DP 1.4, 2.8.2: AUX Transaction Response/Reply Timeouts + * According to the DP spec there should be 3 retries total + * with a 400us wait inbetween each. Hardware already waits + * for 550us therefore no wait is required here. + */ + } + break; + + case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON: + case AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN: + default: + goto fail; + } } + +fail: + if (!payload_reply) + payload->reply = NULL; return false; } diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h index d27f22c05e4b..aab5f0c34584 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h @@ -123,8 +123,9 @@ bool dce110_aux_engine_acquire( struct dce_aux *aux_engine, struct ddc *ddc); -int dce_aux_transfer(struct ddc_service *ddc, - struct aux_payload *cmd); +int dce_aux_transfer_raw(struct ddc_service *ddc, + struct aux_payload *cmd, + enum aux_channel_operation_result *operation_result); bool dce_aux_transfer_with_retries(struct ddc_service *ddc, struct aux_payload *cmd); diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clk_mgr.c index 6e142c2db986..963686380738 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clk_mgr.c @@ -222,7 +222,7 @@ static enum dm_pp_clocks_state dce_get_required_clocks_state( * all required clocks */ for (i = clk_mgr_dce->max_clks_state; i >= DM_PP_CLOCKS_STATE_ULTRA_LOW; i--) - if (context->bw.dce.dispclk_khz > + if (context->bw_ctx.bw.dce.dispclk_khz > clk_mgr_dce->max_clks_by_state[i].display_clk_khz || max_pix_clk > clk_mgr_dce->max_clks_by_state[i].pixel_clk_khz) @@ -232,7 +232,7 @@ static enum dm_pp_clocks_state dce_get_required_clocks_state( if (low_req_clk > clk_mgr_dce->max_clks_state) { /* set max clock state for high phyclock, invalid on exceeding display clock */ if (clk_mgr_dce->max_clks_by_state[clk_mgr_dce->max_clks_state].display_clk_khz - < context->bw.dce.dispclk_khz) + < context->bw_ctx.bw.dce.dispclk_khz) low_req_clk = DM_PP_CLOCKS_STATE_INVALID; else low_req_clk = clk_mgr_dce->max_clks_state; @@ -610,22 +610,22 @@ static void dce11_pplib_apply_display_requirements( struct dm_pp_display_configuration *pp_display_cfg = &context->pp_display_cfg; pp_display_cfg->all_displays_in_sync = - context->bw.dce.all_displays_in_sync; + context->bw_ctx.bw.dce.all_displays_in_sync; pp_display_cfg->nb_pstate_switch_disable = - context->bw.dce.nbp_state_change_enable == false; + context->bw_ctx.bw.dce.nbp_state_change_enable == false; pp_display_cfg->cpu_cc6_disable = - context->bw.dce.cpuc_state_change_enable == false; + context->bw_ctx.bw.dce.cpuc_state_change_enable == false; pp_display_cfg->cpu_pstate_disable = - context->bw.dce.cpup_state_change_enable == false; + context->bw_ctx.bw.dce.cpup_state_change_enable == false; pp_display_cfg->cpu_pstate_separation_time = - context->bw.dce.blackout_recovery_time_us; + context->bw_ctx.bw.dce.blackout_recovery_time_us; - pp_display_cfg->min_memory_clock_khz = context->bw.dce.yclk_khz + pp_display_cfg->min_memory_clock_khz = context->bw_ctx.bw.dce.yclk_khz / MEMORY_TYPE_MULTIPLIER_CZ; pp_display_cfg->min_engine_clock_khz = determine_sclk_from_bounding_box( dc, - context->bw.dce.sclk_khz); + context->bw_ctx.bw.dce.sclk_khz); /* * As workaround for >4x4K lightup set dcfclock to min_engine_clock value. @@ -638,7 +638,7 @@ static void dce11_pplib_apply_display_requirements( pp_display_cfg->min_engine_clock_khz : 0; pp_display_cfg->min_engine_clock_deep_sleep_khz - = context->bw.dce.sclk_deep_sleep_khz; + = context->bw_ctx.bw.dce.sclk_deep_sleep_khz; pp_display_cfg->avail_mclk_switch_time_us = dce110_get_min_vblank_time_us(context); @@ -669,7 +669,7 @@ static void dce_update_clocks(struct clk_mgr *clk_mgr, { struct dce_clk_mgr *clk_mgr_dce = TO_DCE_CLK_MGR(clk_mgr); struct dm_pp_power_level_change_request level_change_req; - int patched_disp_clk = context->bw.dce.dispclk_khz; + int patched_disp_clk = context->bw_ctx.bw.dce.dispclk_khz; /*TODO: W/A for dal3 linux, investigate why this works */ if (!clk_mgr_dce->dfs_bypass_active) @@ -696,7 +696,7 @@ static void dce11_update_clocks(struct clk_mgr *clk_mgr, { struct dce_clk_mgr *clk_mgr_dce = TO_DCE_CLK_MGR(clk_mgr); struct dm_pp_power_level_change_request level_change_req; - int patched_disp_clk = context->bw.dce.dispclk_khz; + int patched_disp_clk = context->bw_ctx.bw.dce.dispclk_khz; /*TODO: W/A for dal3 linux, investigate why this works */ if (!clk_mgr_dce->dfs_bypass_active) @@ -711,7 +711,7 @@ static void dce11_update_clocks(struct clk_mgr *clk_mgr, } if (should_set_clock(safe_to_lower, patched_disp_clk, clk_mgr->clks.dispclk_khz)) { - context->bw.dce.dispclk_khz = dce_set_clock(clk_mgr, patched_disp_clk); + context->bw_ctx.bw.dce.dispclk_khz = dce_set_clock(clk_mgr, patched_disp_clk); clk_mgr->clks.dispclk_khz = patched_disp_clk; } dce11_pplib_apply_display_requirements(clk_mgr->ctx->dc, context); @@ -723,7 +723,7 @@ static void dce112_update_clocks(struct clk_mgr *clk_mgr, { struct dce_clk_mgr *clk_mgr_dce = TO_DCE_CLK_MGR(clk_mgr); struct dm_pp_power_level_change_request level_change_req; - int patched_disp_clk = context->bw.dce.dispclk_khz; + int patched_disp_clk = context->bw_ctx.bw.dce.dispclk_khz; /*TODO: W/A for dal3 linux, investigate why this works */ if (!clk_mgr_dce->dfs_bypass_active) @@ -751,7 +751,7 @@ static void dce12_update_clocks(struct clk_mgr *clk_mgr, struct dce_clk_mgr *clk_mgr_dce = TO_DCE_CLK_MGR(clk_mgr); struct dm_pp_clock_for_voltage_req clock_voltage_req = {0}; int max_pix_clk = get_max_pixel_clock_for_all_paths(context); - int patched_disp_clk = context->bw.dce.dispclk_khz; + int patched_disp_clk = context->bw_ctx.bw.dce.dispclk_khz; /*TODO: W/A for dal3 linux, investigate why this works */ if (!clk_mgr_dce->dfs_bypass_active) diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c index 71d5777de961..f70437aae8e0 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c @@ -978,7 +978,7 @@ static bool dce110_clock_source_power_down( } static bool get_pixel_clk_frequency_100hz( - struct clock_source *clock_source, + const struct clock_source *clock_source, unsigned int inst, unsigned int *pixel_clk_khz) { diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c index c2926cf19dee..aa586672e8cd 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c @@ -213,9 +213,6 @@ static bool dce_dmcu_setup_psr(struct dmcu *dmcu, link->link_enc->funcs->psr_program_secondary_packet(link->link_enc, psr_context->sdpTransmitLineNumDeadline); - if (psr_context->psr_level.bits.SKIP_SMU_NOTIFICATION) - REG_UPDATE(SMU_INTERRUPT_CONTROL, DC_SMU_INT_ENABLE, 1); - /* waitDMCUReadyForCmd */ REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0, dmcu_wait_reg_ready_interval, @@ -594,7 +591,7 @@ static bool dcn10_dmcu_setup_psr(struct dmcu *dmcu, link->link_enc->funcs->psr_program_secondary_packet(link->link_enc, psr_context->sdpTransmitLineNumDeadline); - if (psr_context->psr_level.bits.SKIP_SMU_NOTIFICATION) + if (psr_context->allow_smu_optimizations) REG_UPDATE(SMU_INTERRUPT_CONTROL, DC_SMU_INT_ENABLE, 1); /* waitDMCUReadyForCmd */ @@ -615,6 +612,7 @@ static bool dcn10_dmcu_setup_psr(struct dmcu *dmcu, psr_context->psrFrameCaptureIndicationReq; masterCmdData1.bits.aux_chan = psr_context->channel; masterCmdData1.bits.aux_repeat = psr_context->aux_repeats; + masterCmdData1.bits.allow_smu_optimizations = psr_context->allow_smu_optimizations; dm_write_reg(dmcu->ctx, REG(MASTER_COMM_DATA_REG1), masterCmdData1.u32All); @@ -635,6 +633,7 @@ static bool dcn10_dmcu_setup_psr(struct dmcu *dmcu, dm_write_reg(dmcu->ctx, REG(MASTER_COMM_DATA_REG3), masterCmdData3.u32All); + /* setDMCUParam_Cmd */ REG_UPDATE(MASTER_COMM_CMD_REG, MASTER_COMM_CMD_REG_BYTE0, PSR_SET); diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.h b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.h index c24c0e5ea44e..60ce56f60ae3 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.h @@ -199,16 +199,16 @@ struct dce_dmcu { ******************************************************************/ union dce_dmcu_psr_config_data_reg1 { struct { - unsigned int timehyst_frames:8; /*[7:0]*/ - unsigned int hyst_lines:7; /*[14:8]*/ - unsigned int rfb_update_auto_en:1; /*[15:15]*/ - unsigned int dp_port_num:3; /*[18:16]*/ - unsigned int dcp_sel:3; /*[21:19]*/ - unsigned int phy_type:1; /*[22:22]*/ - unsigned int frame_cap_ind:1; /*[23:23]*/ - unsigned int aux_chan:3; /*[26:24]*/ - unsigned int aux_repeat:4; /*[30:27]*/ - unsigned int reserved:1; /*[31:31]*/ + unsigned int timehyst_frames:8; /*[7:0]*/ + unsigned int hyst_lines:7; /*[14:8]*/ + unsigned int rfb_update_auto_en:1; /*[15:15]*/ + unsigned int dp_port_num:3; /*[18:16]*/ + unsigned int dcp_sel:3; /*[21:19]*/ + unsigned int phy_type:1; /*[22:22]*/ + unsigned int frame_cap_ind:1; /*[23:23]*/ + unsigned int aux_chan:3; /*[26:24]*/ + unsigned int aux_repeat:4; /*[30:27]*/ + unsigned int allow_smu_optimizations:1; /*[31:31]*/ } bits; unsigned int u32All; }; @@ -236,7 +236,7 @@ union dce_dmcu_psr_config_data_reg3 { struct { unsigned int psr_level:16; /*[15:0]*/ unsigned int link_rate:4; /*[19:16]*/ - unsigned int reserved:12; /*[31:20]*/ + unsigned int reserved:12; /*[31:20]*/ } bits; unsigned int u32All; }; diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c b/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c index 40f2d6e0b122..cd26161bcc4d 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c @@ -346,6 +346,16 @@ static void release_engine( } +static bool is_engine_available(struct dce_i2c_hw *dce_i2c_hw) +{ + unsigned int arbitrate; + + REG_GET(DC_I2C_ARBITRATION, DC_I2C_REG_RW_CNTL_STATUS, &arbitrate); + if (arbitrate == DC_I2C_REG_RW_CNTL_STATUS_DMCU_ONLY) + return false; + return true; +} + struct dce_i2c_hw *acquire_i2c_hw_engine( struct resource_pool *pool, struct ddc *ddc) @@ -368,7 +378,7 @@ struct dce_i2c_hw *acquire_i2c_hw_engine( if (!dce_i2c_hw) return NULL; - if (pool->i2c_hw_buffer_in_use) + if (pool->i2c_hw_buffer_in_use || !is_engine_available(dce_i2c_hw)) return NULL; do { diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.h b/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.h index 7f19bb439665..575500755b2e 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.h @@ -29,7 +29,8 @@ enum dc_i2c_status { DC_I2C_STATUS__DC_I2C_STATUS_IDLE, DC_I2C_STATUS__DC_I2C_STATUS_USED_BY_SW, - DC_I2C_STATUS__DC_I2C_STATUS_USED_BY_HW + DC_I2C_STATUS__DC_I2C_STATUS_USED_BY_HW, + DC_I2C_REG_RW_CNTL_STATUS_DMCU_ONLY = 2, }; enum dc_i2c_arbitration { @@ -129,7 +130,8 @@ enum { I2C_SF(DC_I2C_DATA, DC_I2C_DATA, mask_sh),\ I2C_SF(DC_I2C_DATA, DC_I2C_INDEX, mask_sh),\ I2C_SF(DC_I2C_DATA, DC_I2C_INDEX_WRITE, mask_sh),\ - I2C_SF(MICROSECOND_TIME_BASE_DIV, XTAL_REF_DIV, mask_sh) + I2C_SF(MICROSECOND_TIME_BASE_DIV, XTAL_REF_DIV, mask_sh),\ + I2C_SF(DC_I2C_ARBITRATION, DC_I2C_REG_RW_CNTL_STATUS, mask_sh) #define I2C_COMMON_MASK_SH_LIST_DCE110(mask_sh)\ I2C_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(mask_sh),\ @@ -170,6 +172,7 @@ struct dce_i2c_shift { uint8_t DC_I2C_INDEX; uint8_t DC_I2C_INDEX_WRITE; uint8_t XTAL_REF_DIV; + uint8_t DC_I2C_REG_RW_CNTL_STATUS; }; struct dce_i2c_mask { @@ -207,6 +210,7 @@ struct dce_i2c_mask { uint32_t DC_I2C_INDEX; uint32_t DC_I2C_INDEX_WRITE; uint32_t XTAL_REF_DIV; + uint32_t DC_I2C_REG_RW_CNTL_STATUS; }; struct dce_i2c_registers { diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c index 1fa2d4fd7a35..14309fe6f2e6 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c @@ -272,7 +272,8 @@ static void dce110_update_hdmi_info_packet( static void dce110_stream_encoder_dp_set_stream_attribute( struct stream_encoder *enc, struct dc_crtc_timing *crtc_timing, - enum dc_color_space output_color_space) + enum dc_color_space output_color_space, + uint32_t enable_sdp_splitting) { #if defined(CONFIG_DRM_AMD_DC_DCN1_0) uint32_t h_active_start; @@ -977,7 +978,7 @@ static void dce110_stream_encoder_dp_unblank( uint64_t m_vid_l = n_vid; - m_vid_l *= param->pixel_clk_khz; + m_vid_l *= param->timing.pix_clk_100hz / 10; m_vid_l = div_u64(m_vid_l, param->link_settings.link_rate * LINK_RATE_REF_FREQ_IN_KHZ); diff --git a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c index 23044e6723e8..e938bf9986d3 100644 --- a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c @@ -378,6 +378,28 @@ static const struct resource_caps res_cap = { .num_ddc = 6, }; +static const struct dc_plane_cap plane_cap = { + .type = DC_PLANE_TYPE_DCE_RGB, + + .pixel_format_support = { + .argb8888 = true, + .nv12 = false, + .fp16 = false + }, + + .max_upscale_factor = { + .argb8888 = 16000, + .nv12 = 1, + .fp16 = 1 + }, + + .max_downscale_factor = { + .argb8888 = 250, + .nv12 = 1, + .fp16 = 1 + } +}; + #define CTX ctx #define REG(reg) mm ## reg @@ -756,7 +778,8 @@ static enum dc_status build_mapped_resource( bool dce100_validate_bandwidth( struct dc *dc, - struct dc_state *context) + struct dc_state *context, + bool fast_validate) { int i; bool at_least_one_pipe = false; @@ -768,11 +791,11 @@ bool dce100_validate_bandwidth( if (at_least_one_pipe) { /* TODO implement when needed but for now hardcode max value*/ - context->bw.dce.dispclk_khz = 681000; - context->bw.dce.yclk_khz = 250000 * MEMORY_TYPE_MULTIPLIER_CZ; + context->bw_ctx.bw.dce.dispclk_khz = 681000; + context->bw_ctx.bw.dce.yclk_khz = 250000 * MEMORY_TYPE_MULTIPLIER_CZ; } else { - context->bw.dce.dispclk_khz = 0; - context->bw.dce.yclk_khz = 0; + context->bw_ctx.bw.dce.dispclk_khz = 0; + context->bw_ctx.bw.dce.yclk_khz = 0; } return true; @@ -1023,6 +1046,9 @@ static bool construct( dc->caps.max_planes = pool->base.pipe_count; + for (i = 0; i < dc->caps.max_planes; ++i) + dc->caps.planes[i] = plane_cap; + if (!resource_construct(num_virtual_links, dc, &pool->base, &res_create_funcs)) goto res_create_fail; diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c index 5e4db3712eef..7ac50ab1b762 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c @@ -616,7 +616,7 @@ dce110_set_output_transfer_func(struct pipe_ctx *pipe_ctx, void dce110_update_info_frame(struct pipe_ctx *pipe_ctx) { - bool is_hdmi; + bool is_hdmi_tmds; bool is_dp; ASSERT(pipe_ctx->stream); @@ -624,13 +624,13 @@ void dce110_update_info_frame(struct pipe_ctx *pipe_ctx) if (pipe_ctx->stream_res.stream_enc == NULL) return; /* this is not root pipe */ - is_hdmi = dc_is_hdmi_signal(pipe_ctx->stream->signal); + is_hdmi_tmds = dc_is_hdmi_tmds_signal(pipe_ctx->stream->signal); is_dp = dc_is_dp_signal(pipe_ctx->stream->signal); - if (!is_hdmi && !is_dp) + if (!is_hdmi_tmds && !is_dp) return; - if (is_hdmi) + if (is_hdmi_tmds) pipe_ctx->stream_res.stream_enc->funcs->update_hdmi_info_packets( pipe_ctx->stream_res.stream_enc, &pipe_ctx->stream_res.encoder_info_frame); @@ -935,13 +935,31 @@ void hwss_edp_backlight_control( edp_receiver_ready_T9(link); } +// Static helper function which calls the correct function +// based on pp_smu version +static void set_pme_wa_enable_by_version(struct dc *dc) +{ + struct pp_smu_funcs *pp_smu = NULL; + + if (dc->res_pool->pp_smu) + pp_smu = dc->res_pool->pp_smu; + + if (pp_smu) { + if (pp_smu->ctx.ver == PP_SMU_VER_RV && pp_smu->rv_funcs.set_pme_wa_enable) + pp_smu->rv_funcs.set_pme_wa_enable(&(pp_smu->ctx)); + } +} + void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx) { - struct dc *core_dc = pipe_ctx->stream->ctx->dc; /* notify audio driver for audio modes of monitor */ - struct pp_smu_funcs_rv *pp_smu = core_dc->res_pool->pp_smu; + struct dc *core_dc = pipe_ctx->stream->ctx->dc; + struct pp_smu_funcs *pp_smu = NULL; unsigned int i, num_audio = 1; + if (core_dc->res_pool->pp_smu) + pp_smu = core_dc->res_pool->pp_smu; + if (pipe_ctx->stream_res.audio) { for (i = 0; i < MAX_PIPES; i++) { /*current_state not updated yet*/ @@ -951,30 +969,31 @@ void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx) pipe_ctx->stream_res.audio->funcs->az_enable(pipe_ctx->stream_res.audio); - if (num_audio >= 1 && pp_smu != NULL && pp_smu->set_pme_wa_enable != NULL) + if (num_audio >= 1 && pp_smu != NULL) /*this is the first audio. apply the PME w/a in order to wake AZ from D3*/ - pp_smu->set_pme_wa_enable(&pp_smu->pp_smu); + set_pme_wa_enable_by_version(core_dc); /* un-mute audio */ /* TODO: audio should be per stream rather than per link */ pipe_ctx->stream_res.stream_enc->funcs->audio_mute_control( - pipe_ctx->stream_res.stream_enc, false); + pipe_ctx->stream_res.stream_enc, false); } } void dce110_disable_audio_stream(struct pipe_ctx *pipe_ctx, int option) { struct dc *dc = pipe_ctx->stream->ctx->dc; + struct pp_smu_funcs *pp_smu = NULL; pipe_ctx->stream_res.stream_enc->funcs->audio_mute_control( pipe_ctx->stream_res.stream_enc, true); if (pipe_ctx->stream_res.audio) { - struct pp_smu_funcs_rv *pp_smu = dc->res_pool->pp_smu; + if (dc->res_pool->pp_smu) + pp_smu = dc->res_pool->pp_smu; if (option != KEEP_ACQUIRED_RESOURCE || - !dc->debug.az_endpoint_mute_only) { + !dc->debug.az_endpoint_mute_only) /*only disalbe az_endpoint if power down or free*/ pipe_ctx->stream_res.audio->funcs->az_disable(pipe_ctx->stream_res.audio); - } if (dc_is_dp_signal(pipe_ctx->stream->signal)) pipe_ctx->stream_res.stream_enc->funcs->dp_audio_disable( @@ -989,9 +1008,9 @@ void dce110_disable_audio_stream(struct pipe_ctx *pipe_ctx, int option) update_audio_usage(&dc->current_state->res_ctx, dc->res_pool, pipe_ctx->stream_res.audio, false); pipe_ctx->stream_res.audio = NULL; } - if (pp_smu != NULL && pp_smu->set_pme_wa_enable != NULL) + if (pp_smu != NULL) /*this is the first audio. apply the PME w/a in order to wake AZ from D3*/ - pp_smu->set_pme_wa_enable(&pp_smu->pp_smu); + set_pme_wa_enable_by_version(dc); /* TODO: notify audio driver for if audio modes list changed * add audio mode list change flag */ @@ -1007,7 +1026,7 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option) struct dc_link *link = stream->link; struct dc *dc = pipe_ctx->stream->ctx->dc; - if (dc_is_hdmi_signal(pipe_ctx->stream->signal)) + if (dc_is_hdmi_tmds_signal(pipe_ctx->stream->signal)) pipe_ctx->stream_res.stream_enc->funcs->stop_hdmi_info_packets( pipe_ctx->stream_res.stream_enc); @@ -1032,7 +1051,7 @@ void dce110_unblank_stream(struct pipe_ctx *pipe_ctx, struct dc_link *link = stream->link; /* only 3 items below are used by unblank */ - params.pixel_clk_khz = pipe_ctx->stream->timing.pix_clk_100hz / 10; + params.timing = pipe_ctx->stream->timing; params.link_settings.link_rate = link_settings->link_rate; if (dc_is_dp_signal(pipe_ctx->stream->signal)) @@ -1147,8 +1166,8 @@ static void build_audio_output( if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT || pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST) { audio_output->pll_info.dp_dto_source_clock_in_khz = - state->dccg->funcs->get_dp_ref_clk_frequency( - state->dccg); + state->clk_mgr->funcs->get_dp_ref_clk_frequency( + state->clk_mgr); } audio_output->pll_info.feed_back_divider = @@ -1349,7 +1368,7 @@ static enum dc_status apply_single_controller_ctx_to_hw( pipe_ctx->stream_res.tg->funcs->set_static_screen_control( pipe_ctx->stream_res.tg, event_triggers); - if (pipe_ctx->stream->signal != SIGNAL_TYPE_VIRTUAL) + if (!dc_is_virtual_signal(pipe_ctx->stream->signal)) pipe_ctx->stream_res.stream_enc->funcs->dig_connect_to_otg( pipe_ctx->stream_res.stream_enc, pipe_ctx->stream_res.tg->inst); @@ -1358,7 +1377,7 @@ static enum dc_status apply_single_controller_ctx_to_hw( pipe_ctx->stream_res.opp, COLOR_SPACE_YCBCR601, stream->timing.display_color_depth, - pipe_ctx->stream->signal); + stream->signal); pipe_ctx->stream_res.opp->funcs->opp_program_fmt( pipe_ctx->stream_res.opp, @@ -1532,6 +1551,9 @@ void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context) } } + if (dc->hwss.init_pipes) + dc->hwss.init_pipes(dc, context); + if (edp_link) { /* this seems to cause blank screens on DCE8 */ if ((dc->ctx->dce_version == DCE_VERSION_8_0) || @@ -1608,18 +1630,18 @@ static void dce110_set_displaymarks( dc->bw_vbios->blackout_duration, pipe_ctx->stream); pipe_ctx->plane_res.mi->funcs->mem_input_program_display_marks( pipe_ctx->plane_res.mi, - context->bw.dce.nbp_state_change_wm_ns[num_pipes], - context->bw.dce.stutter_exit_wm_ns[num_pipes], - context->bw.dce.stutter_entry_wm_ns[num_pipes], - context->bw.dce.urgent_wm_ns[num_pipes], + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[num_pipes], + context->bw_ctx.bw.dce.stutter_exit_wm_ns[num_pipes], + context->bw_ctx.bw.dce.stutter_entry_wm_ns[num_pipes], + context->bw_ctx.bw.dce.urgent_wm_ns[num_pipes], total_dest_line_time_ns); if (i == underlay_idx) { num_pipes++; pipe_ctx->plane_res.mi->funcs->mem_input_program_chroma_display_marks( pipe_ctx->plane_res.mi, - context->bw.dce.nbp_state_change_wm_ns[num_pipes], - context->bw.dce.stutter_exit_wm_ns[num_pipes], - context->bw.dce.urgent_wm_ns[num_pipes], + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[num_pipes], + context->bw_ctx.bw.dce.stutter_exit_wm_ns[num_pipes], + context->bw_ctx.bw.dce.urgent_wm_ns[num_pipes], total_dest_line_time_ns); } num_pipes++; @@ -2612,7 +2634,7 @@ void dce110_set_cursor_position(struct pipe_ctx *pipe_ctx) struct mem_input *mi = pipe_ctx->plane_res.mi; struct dc_cursor_mi_param param = { .pixel_clk_khz = pipe_ctx->stream->timing.pix_clk_100hz / 10, - .ref_clk_khz = pipe_ctx->stream->ctx->dc->res_pool->ref_clock_inKhz, + .ref_clk_khz = pipe_ctx->stream->ctx->dc->res_pool->ref_clocks.xtalin_clock_inKhz, .viewport = pipe_ctx->plane_res.scl_data.viewport, .h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz, .v_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.vert, diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c index 7549adaa1542..dcd04e9ea76b 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c @@ -392,6 +392,55 @@ static const struct resource_caps stoney_resource_cap = { .num_ddc = 3, }; +static const struct dc_plane_cap plane_cap = { + .type = DC_PLANE_TYPE_DCE_RGB, + .blends_with_below = true, + .blends_with_above = true, + .per_pixel_alpha = 1, + + .pixel_format_support = { + .argb8888 = true, + .nv12 = false, + .fp16 = false + }, + + .max_upscale_factor = { + .argb8888 = 16000, + .nv12 = 1, + .fp16 = 1 + }, + + .max_downscale_factor = { + .argb8888 = 250, + .nv12 = 1, + .fp16 = 1 + } +}; + +static const struct dc_plane_cap underlay_plane_cap = { + .type = DC_PLANE_TYPE_DCE_UNDERLAY, + .blends_with_above = true, + .per_pixel_alpha = 1, + + .pixel_format_support = { + .argb8888 = false, + .nv12 = true, + .fp16 = false + }, + + .max_upscale_factor = { + .argb8888 = 1, + .nv12 = 16000, + .fp16 = 1 + }, + + .max_downscale_factor = { + .argb8888 = 1, + .nv12 = 250, + .fp16 = 1 + } +}; + #define CTX ctx #define REG(reg) mm ## reg @@ -854,7 +903,8 @@ static enum dc_status build_mapped_resource( static bool dce110_validate_bandwidth( struct dc *dc, - struct dc_state *context) + struct dc_state *context, + bool fast_validate) { bool result = false; @@ -868,7 +918,7 @@ static bool dce110_validate_bandwidth( dc->bw_vbios, context->res_ctx.pipe_ctx, dc->res_pool->pipe_count, - &context->bw.dce)) + &context->bw_ctx.bw.dce)) result = true; if (!result) @@ -878,8 +928,8 @@ static bool dce110_validate_bandwidth( context->streams[0]->timing.v_addressable, context->streams[0]->timing.pix_clk_100hz / 10); - if (memcmp(&dc->current_state->bw.dce, - &context->bw.dce, sizeof(context->bw.dce))) { + if (memcmp(&dc->current_state->bw_ctx.bw.dce, + &context->bw_ctx.bw.dce, sizeof(context->bw_ctx.bw.dce))) { DC_LOG_BANDWIDTH_CALCS( "%s: finish,\n" @@ -893,34 +943,34 @@ static bool dce110_validate_bandwidth( "sclk: %d sclk_sleep: %d yclk: %d blackout_recovery_time_us: %d\n" , __func__, - context->bw.dce.nbp_state_change_wm_ns[0].b_mark, - context->bw.dce.nbp_state_change_wm_ns[0].a_mark, - context->bw.dce.urgent_wm_ns[0].b_mark, - context->bw.dce.urgent_wm_ns[0].a_mark, - context->bw.dce.stutter_exit_wm_ns[0].b_mark, - context->bw.dce.stutter_exit_wm_ns[0].a_mark, - context->bw.dce.nbp_state_change_wm_ns[1].b_mark, - context->bw.dce.nbp_state_change_wm_ns[1].a_mark, - context->bw.dce.urgent_wm_ns[1].b_mark, - context->bw.dce.urgent_wm_ns[1].a_mark, - context->bw.dce.stutter_exit_wm_ns[1].b_mark, - context->bw.dce.stutter_exit_wm_ns[1].a_mark, - context->bw.dce.nbp_state_change_wm_ns[2].b_mark, - context->bw.dce.nbp_state_change_wm_ns[2].a_mark, - context->bw.dce.urgent_wm_ns[2].b_mark, - context->bw.dce.urgent_wm_ns[2].a_mark, - context->bw.dce.stutter_exit_wm_ns[2].b_mark, - context->bw.dce.stutter_exit_wm_ns[2].a_mark, - context->bw.dce.stutter_mode_enable, - context->bw.dce.cpuc_state_change_enable, - context->bw.dce.cpup_state_change_enable, - context->bw.dce.nbp_state_change_enable, - context->bw.dce.all_displays_in_sync, - context->bw.dce.dispclk_khz, - context->bw.dce.sclk_khz, - context->bw.dce.sclk_deep_sleep_khz, - context->bw.dce.yclk_khz, - context->bw.dce.blackout_recovery_time_us); + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[0].b_mark, + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[0].a_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[0].b_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[0].a_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[0].b_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[0].a_mark, + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[1].b_mark, + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[1].a_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[1].b_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[1].a_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[1].b_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[1].a_mark, + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[2].b_mark, + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[2].a_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[2].b_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[2].a_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[2].b_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[2].a_mark, + context->bw_ctx.bw.dce.stutter_mode_enable, + context->bw_ctx.bw.dce.cpuc_state_change_enable, + context->bw_ctx.bw.dce.cpup_state_change_enable, + context->bw_ctx.bw.dce.nbp_state_change_enable, + context->bw_ctx.bw.dce.all_displays_in_sync, + context->bw_ctx.bw.dce.dispclk_khz, + context->bw_ctx.bw.dce.sclk_khz, + context->bw_ctx.bw.dce.sclk_deep_sleep_khz, + context->bw_ctx.bw.dce.yclk_khz, + context->bw_ctx.bw.dce.blackout_recovery_time_us); } return result; } @@ -1371,6 +1421,11 @@ static bool construct( dc->caps.max_planes = pool->base.pipe_count; + for (i = 0; i < pool->base.underlay_pipe_index; ++i) + dc->caps.planes[i] = plane_cap; + + dc->caps.planes[pool->base.underlay_pipe_index] = underlay_plane_cap; + bw_calcs_init(dc->bw_dceip, dc->bw_vbios, dc->ctx->asic_id); bw_calcs_data_update_from_pplib(dc); diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c index ea3065d63372..a480b15f6885 100644 --- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c @@ -397,6 +397,28 @@ static const struct resource_caps polaris_11_resource_cap = { .num_ddc = 5, }; +static const struct dc_plane_cap plane_cap = { + .type = DC_PLANE_TYPE_DCE_RGB, + + .pixel_format_support = { + .argb8888 = true, + .nv12 = false, + .fp16 = false + }, + + .max_upscale_factor = { + .argb8888 = 16000, + .nv12 = 1, + .fp16 = 1 + }, + + .max_downscale_factor = { + .argb8888 = 250, + .nv12 = 1, + .fp16 = 1 + } +}; + #define CTX ctx #define REG(reg) mm ## reg @@ -804,7 +826,8 @@ static enum dc_status build_mapped_resource( bool dce112_validate_bandwidth( struct dc *dc, - struct dc_state *context) + struct dc_state *context, + bool fast_validate) { bool result = false; @@ -818,7 +841,7 @@ bool dce112_validate_bandwidth( dc->bw_vbios, context->res_ctx.pipe_ctx, dc->res_pool->pipe_count, - &context->bw.dce)) + &context->bw_ctx.bw.dce)) result = true; if (!result) @@ -826,8 +849,8 @@ bool dce112_validate_bandwidth( "%s: Bandwidth validation failed!", __func__); - if (memcmp(&dc->current_state->bw.dce, - &context->bw.dce, sizeof(context->bw.dce))) { + if (memcmp(&dc->current_state->bw_ctx.bw.dce, + &context->bw_ctx.bw.dce, sizeof(context->bw_ctx.bw.dce))) { DC_LOG_BANDWIDTH_CALCS( "%s: finish,\n" @@ -841,34 +864,34 @@ bool dce112_validate_bandwidth( "sclk: %d sclk_sleep: %d yclk: %d blackout_recovery_time_us: %d\n" , __func__, - context->bw.dce.nbp_state_change_wm_ns[0].b_mark, - context->bw.dce.nbp_state_change_wm_ns[0].a_mark, - context->bw.dce.urgent_wm_ns[0].b_mark, - context->bw.dce.urgent_wm_ns[0].a_mark, - context->bw.dce.stutter_exit_wm_ns[0].b_mark, - context->bw.dce.stutter_exit_wm_ns[0].a_mark, - context->bw.dce.nbp_state_change_wm_ns[1].b_mark, - context->bw.dce.nbp_state_change_wm_ns[1].a_mark, - context->bw.dce.urgent_wm_ns[1].b_mark, - context->bw.dce.urgent_wm_ns[1].a_mark, - context->bw.dce.stutter_exit_wm_ns[1].b_mark, - context->bw.dce.stutter_exit_wm_ns[1].a_mark, - context->bw.dce.nbp_state_change_wm_ns[2].b_mark, - context->bw.dce.nbp_state_change_wm_ns[2].a_mark, - context->bw.dce.urgent_wm_ns[2].b_mark, - context->bw.dce.urgent_wm_ns[2].a_mark, - context->bw.dce.stutter_exit_wm_ns[2].b_mark, - context->bw.dce.stutter_exit_wm_ns[2].a_mark, - context->bw.dce.stutter_mode_enable, - context->bw.dce.cpuc_state_change_enable, - context->bw.dce.cpup_state_change_enable, - context->bw.dce.nbp_state_change_enable, - context->bw.dce.all_displays_in_sync, - context->bw.dce.dispclk_khz, - context->bw.dce.sclk_khz, - context->bw.dce.sclk_deep_sleep_khz, - context->bw.dce.yclk_khz, - context->bw.dce.blackout_recovery_time_us); + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[0].b_mark, + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[0].a_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[0].b_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[0].a_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[0].b_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[0].a_mark, + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[1].b_mark, + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[1].a_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[1].b_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[1].a_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[1].b_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[1].a_mark, + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[2].b_mark, + context->bw_ctx.bw.dce.nbp_state_change_wm_ns[2].a_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[2].b_mark, + context->bw_ctx.bw.dce.urgent_wm_ns[2].a_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[2].b_mark, + context->bw_ctx.bw.dce.stutter_exit_wm_ns[2].a_mark, + context->bw_ctx.bw.dce.stutter_mode_enable, + context->bw_ctx.bw.dce.cpuc_state_change_enable, + context->bw_ctx.bw.dce.cpup_state_change_enable, + context->bw_ctx.bw.dce.nbp_state_change_enable, + context->bw_ctx.bw.dce.all_displays_in_sync, + context->bw_ctx.bw.dce.dispclk_khz, + context->bw_ctx.bw.dce.sclk_khz, + context->bw_ctx.bw.dce.sclk_deep_sleep_khz, + context->bw_ctx.bw.dce.yclk_khz, + context->bw_ctx.bw.dce.blackout_recovery_time_us); } return result; } @@ -887,7 +910,7 @@ enum dc_status resource_map_phy_clock_resources( return DC_ERROR_UNEXPECTED; if (dc_is_dp_signal(pipe_ctx->stream->signal) - || pipe_ctx->stream->signal == SIGNAL_TYPE_VIRTUAL) + || dc_is_virtual_signal(pipe_ctx->stream->signal)) pipe_ctx->clock_source = dc->res_pool->dp_clock_source; else @@ -1310,6 +1333,9 @@ static bool construct( dc->caps.max_planes = pool->base.pipe_count; + for (i = 0; i < dc->caps.max_planes; ++i) + dc->caps.planes[i] = plane_cap; + /* Create hardware sequencer */ dce112_hw_sequencer_construct(dc); diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h index 95a403396219..1f57ebc6f9b4 100644 --- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h +++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h @@ -44,7 +44,8 @@ enum dc_status dce112_validate_with_context( bool dce112_validate_bandwidth( struct dc *dc, - struct dc_state *context); + struct dc_state *context, + bool fast_validate); enum dc_status dce112_add_stream_to_ctx( struct dc *dc, diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c index 312a0aebf91f..6d49c7143c67 100644 --- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c @@ -454,6 +454,28 @@ static const struct resource_caps res_cap = { .num_ddc = 6, }; +static const struct dc_plane_cap plane_cap = { + .type = DC_PLANE_TYPE_DCE_RGB, + + .pixel_format_support = { + .argb8888 = true, + .nv12 = false, + .fp16 = false + }, + + .max_upscale_factor = { + .argb8888 = 16000, + .nv12 = 1, + .fp16 = 1 + }, + + .max_downscale_factor = { + .argb8888 = 250, + .nv12 = 1, + .fp16 = 1 + } +}; + static const struct dc_debug_options debug_defaults = { .disable_clock_gate = true, }; @@ -1171,6 +1193,9 @@ static bool construct( dc->caps.max_planes = pool->base.pipe_count; + for (i = 0; i < dc->caps.max_planes; ++i) + dc->caps.planes[i] = plane_cap; + bw_calcs_init(dc->bw_dceip, dc->bw_vbios, dc->ctx->asic_id); bw_calcs_data_update_from_pplib(dc); diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c index c109ace96be9..27d0cc394963 100644 --- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c @@ -387,6 +387,28 @@ static const struct resource_caps res_cap_83 = { .num_ddc = 2, }; +static const struct dc_plane_cap plane_cap = { + .type = DC_PLANE_TYPE_DCE_RGB, + + .pixel_format_support = { + .argb8888 = true, + .nv12 = false, + .fp16 = false + }, + + .max_upscale_factor = { + .argb8888 = 16000, + .nv12 = 1, + .fp16 = 1 + }, + + .max_downscale_factor = { + .argb8888 = 250, + .nv12 = 1, + .fp16 = 1 + } +}; + static const struct dce_dmcu_registers dmcu_regs = { DMCU_DCE80_REG_LIST() }; @@ -790,7 +812,8 @@ static void destruct(struct dce110_resource_pool *pool) bool dce80_validate_bandwidth( struct dc *dc, - struct dc_state *context) + struct dc_state *context, + bool fast_validate) { int i; bool at_least_one_pipe = false; @@ -802,11 +825,11 @@ bool dce80_validate_bandwidth( if (at_least_one_pipe) { /* TODO implement when needed but for now hardcode max value*/ - context->bw.dce.dispclk_khz = 681000; - context->bw.dce.yclk_khz = 250000 * MEMORY_TYPE_MULTIPLIER_CZ; + context->bw_ctx.bw.dce.dispclk_khz = 681000; + context->bw_ctx.bw.dce.yclk_khz = 250000 * MEMORY_TYPE_MULTIPLIER_CZ; } else { - context->bw.dce.dispclk_khz = 0; - context->bw.dce.yclk_khz = 0; + context->bw_ctx.bw.dce.dispclk_khz = 0; + context->bw_ctx.bw.dce.yclk_khz = 0; } return true; @@ -1032,6 +1055,10 @@ static bool dce80_construct( } dc->caps.max_planes = pool->base.pipe_count; + + for (i = 0; i < dc->caps.max_planes; ++i) + dc->caps.planes[i] = plane_cap; + dc->caps.disable_dp_clk_share = true; if (!resource_construct(num_virtual_links, dc, &pool->base, @@ -1237,6 +1264,10 @@ static bool dce81_construct( } dc->caps.max_planes = pool->base.pipe_count; + + for (i = 0; i < dc->caps.max_planes; ++i) + dc->caps.planes[i] = plane_cap; + dc->caps.disable_dp_clk_share = true; if (!resource_construct(num_virtual_links, dc, &pool->base, @@ -1438,6 +1469,10 @@ static bool dce83_construct( } dc->caps.max_planes = pool->base.pipe_count; + + for (i = 0; i < dc->caps.max_planes; ++i) + dc->caps.planes[i] = plane_cap; + dc->caps.disable_dp_clk_share = true; if (!resource_construct(num_virtual_links, dc, &pool->base, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_clk_mgr.c index afe8c42211cd..0d9bee8d5ab9 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_clk_mgr.c @@ -43,23 +43,6 @@ #define DC_LOGGER \ clk_mgr->ctx->logger -void dcn1_pplib_apply_display_requirements( - struct dc *dc, - struct dc_state *context) -{ - struct dm_pp_display_configuration *pp_display_cfg = &context->pp_display_cfg; - - pp_display_cfg->min_engine_clock_khz = dc->res_pool->clk_mgr->clks.dcfclk_khz; - pp_display_cfg->min_memory_clock_khz = dc->res_pool->clk_mgr->clks.fclk_khz; - pp_display_cfg->min_engine_clock_deep_sleep_khz = dc->res_pool->clk_mgr->clks.dcfclk_deep_sleep_khz; - pp_display_cfg->min_dcfc_deep_sleep_clock_khz = dc->res_pool->clk_mgr->clks.dcfclk_deep_sleep_khz; - pp_display_cfg->min_dcfclock_khz = dc->res_pool->clk_mgr->clks.dcfclk_khz; - pp_display_cfg->disp_clk_khz = dc->res_pool->clk_mgr->clks.dispclk_khz; - dce110_fill_display_configs(context, pp_display_cfg); - - dm_pp_apply_display_requirements(dc->ctx, pp_display_cfg); -} - static int dcn1_determine_dppclk_threshold(struct clk_mgr *clk_mgr, struct dc_clocks *new_clocks) { bool request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz; @@ -167,11 +150,11 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr, { struct dc *dc = clk_mgr->ctx->dc; struct dc_debug_options *debug = &dc->debug; - struct dc_clocks *new_clocks = &context->bw.dcn.clk; + struct dc_clocks *new_clocks = &context->bw_ctx.bw.dcn.clk; struct pp_smu_display_requirement_rv *smu_req_cur = &dc->res_pool->pp_smu_req; struct pp_smu_display_requirement_rv smu_req = *smu_req_cur; - struct pp_smu_funcs_rv *pp_smu = dc->res_pool->pp_smu; + struct pp_smu_funcs_rv *pp_smu = NULL; bool send_request_to_increase = false; bool send_request_to_lower = false; int display_count; @@ -179,7 +162,8 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr, bool enter_display_off = false; display_count = get_active_display_cnt(dc, context); - + if (dc->res_pool->pp_smu) + pp_smu = &dc->res_pool->pp_smu->rv_funcs; if (display_count == 0) enter_display_off = true; @@ -189,7 +173,7 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr, * if function pointer not set up, this message is * sent as part of pplib_apply_display_requirements. */ - if (pp_smu->set_display_count) + if (pp_smu && pp_smu->set_display_count) pp_smu->set_display_count(&pp_smu->pp_smu, display_count); smu_req.display_count = display_count; @@ -239,17 +223,13 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr, */ if (send_request_to_increase) { /*use dcfclk to request voltage*/ - if (pp_smu->set_hard_min_fclk_by_freq && + if (pp_smu && pp_smu->set_hard_min_fclk_by_freq && pp_smu->set_hard_min_dcfclk_by_freq && pp_smu->set_min_deep_sleep_dcfclk) { pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, smu_req.hard_min_fclk_mhz); pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, smu_req.hard_min_dcefclk_mhz); pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, smu_req.min_deep_sleep_dcefclk_mhz); - } else { - if (pp_smu->set_display_requirement) - pp_smu->set_display_requirement(&pp_smu->pp_smu, &smu_req); - dcn1_pplib_apply_display_requirements(dc, context); } } @@ -265,17 +245,13 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr, if (!send_request_to_increase && send_request_to_lower) { /*use dcfclk to request voltage*/ - if (pp_smu->set_hard_min_fclk_by_freq && + if (pp_smu && pp_smu->set_hard_min_fclk_by_freq && pp_smu->set_hard_min_dcfclk_by_freq && pp_smu->set_min_deep_sleep_dcfclk) { pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, smu_req.hard_min_fclk_mhz); pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, smu_req.hard_min_dcefclk_mhz); pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, smu_req.min_deep_sleep_dcefclk_mhz); - } else { - if (pp_smu->set_display_requirement) - pp_smu->set_display_requirement(&pp_smu->pp_smu, &smu_req); - dcn1_pplib_apply_display_requirements(dc, context); } } diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_clk_mgr.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_clk_mgr.h index a995eda443a3..97007cf33665 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_clk_mgr.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_clk_mgr.h @@ -34,10 +34,6 @@ struct clk_bypass { uint32_t dprefclk_bypass; }; -void dcn1_pplib_apply_display_requirements( - struct dc *dc, - struct dc_state *context); - struct clk_mgr *dcn1_clk_mgr_create(struct dc_context *ctx); #endif //__DCN10_CLK_MGR_H__ diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c index cd1ebe57ed59..6f4b24756323 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c @@ -91,13 +91,6 @@ enum dscl_mode_sel { DSCL_MODE_DSCL_BYPASS = 6 }; -enum gamut_remap_select { - GAMUT_REMAP_BYPASS = 0, - GAMUT_REMAP_COEFF, - GAMUT_REMAP_COMA_COEFF, - GAMUT_REMAP_COMB_COEFF -}; - void dpp_read_state(struct dpp *dpp_base, struct dcn_dpp_state *s) { @@ -392,6 +385,10 @@ void dpp1_cnv_setup ( default: break; } + + /* Set default color space based on format if none is given. */ + color_space = input_color_space ? input_color_space : color_space; + REG_SET(CNVC_SURFACE_PIXEL_FORMAT, 0, CNVC_SURFACE_PIXEL_FORMAT, pixel_format); REG_UPDATE(FORMAT_CONTROL, FORMAT_CONTROL__ALPHA_EN, alpha_en); @@ -403,7 +400,7 @@ void dpp1_cnv_setup ( for (i = 0; i < 12; i++) tbl_entry.regval[i] = input_csc_color_matrix.matrix[i]; - tbl_entry.color_space = input_color_space; + tbl_entry.color_space = color_space; if (color_space >= COLOR_SPACE_YCBCR601) select = INPUT_CSC_SELECT_ICSC; diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c index 41f0f4c912e7..882bcc5a40f6 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c @@ -88,13 +88,6 @@ enum dscl_mode_sel { DSCL_MODE_DSCL_BYPASS = 6 }; -enum gamut_remap_select { - GAMUT_REMAP_BYPASS = 0, - GAMUT_REMAP_COEFF, - GAMUT_REMAP_COMA_COEFF, - GAMUT_REMAP_COMB_COEFF -}; - static const struct dpp_input_csc_matrix dpp_input_csc_matrix[] = { {COLOR_SPACE_SRGB, {0x2000, 0, 0, 0, 0, 0x2000, 0, 0, 0, 0, 0x2000, 0} }, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c index c7642e748297..ce21a290bf3e 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c @@ -406,15 +406,25 @@ void dpp1_dscl_calc_lb_num_partitions( int *num_part_y, int *num_part_c) { + int lb_memory_size, lb_memory_size_c, lb_memory_size_a, num_partitions_a, + lb_bpc, memory_line_size_y, memory_line_size_c, memory_line_size_a; + int line_size = scl_data->viewport.width < scl_data->recout.width ? scl_data->viewport.width : scl_data->recout.width; int line_size_c = scl_data->viewport_c.width < scl_data->recout.width ? scl_data->viewport_c.width : scl_data->recout.width; - int lb_bpc = dpp1_dscl_get_lb_depth_bpc(scl_data->lb_params.depth); - int memory_line_size_y = (line_size * lb_bpc + 71) / 72; /* +71 to ceil */ - int memory_line_size_c = (line_size_c * lb_bpc + 71) / 72; /* +71 to ceil */ - int memory_line_size_a = (line_size + 5) / 6; /* +5 to ceil */ - int lb_memory_size, lb_memory_size_c, lb_memory_size_a, num_partitions_a; + + if (line_size == 0) + line_size = 1; + + if (line_size_c == 0) + line_size_c = 1; + + + lb_bpc = dpp1_dscl_get_lb_depth_bpc(scl_data->lb_params.depth); + memory_line_size_y = (line_size * lb_bpc + 71) / 72; /* +71 to ceil */ + memory_line_size_c = (line_size_c * lb_bpc + 71) / 72; /* +71 to ceil */ + memory_line_size_a = (line_size + 5) / 6; /* +5 to ceil */ if (lb_config == LB_MEMORY_CONFIG_1) { lb_memory_size = 816; diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c index e161ad836812..295cbd5b843f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c @@ -258,8 +258,9 @@ void hubbub1_wm_change_req_wa(struct hubbub *hubbub) { struct dcn10_hubbub *hubbub1 = TO_DCN10_HUBBUB(hubbub); - REG_UPDATE_SEQ(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, - DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 0, 1); + REG_UPDATE_SEQ_2(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, + DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 0, + DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 1); } void hubbub1_program_watermarks( diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c index 683829466a44..0ba68d41b9c3 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c @@ -1150,28 +1150,9 @@ void hubp1_cursor_set_position( REG_UPDATE(CURSOR_CONTROL, CURSOR_ENABLE, cur_en); - //account for cases where we see negative offset relative to overlay plane - if (src_x_offset < 0 && src_y_offset < 0) { - REG_SET_2(CURSOR_POSITION, 0, - CURSOR_X_POSITION, 0, - CURSOR_Y_POSITION, 0); - x_hotspot -= src_x_offset; - y_hotspot -= src_y_offset; - } else if (src_x_offset < 0) { - REG_SET_2(CURSOR_POSITION, 0, - CURSOR_X_POSITION, 0, - CURSOR_Y_POSITION, pos->y); - x_hotspot -= src_x_offset; - } else if (src_y_offset < 0) { - REG_SET_2(CURSOR_POSITION, 0, + REG_SET_2(CURSOR_POSITION, 0, CURSOR_X_POSITION, pos->x, - CURSOR_Y_POSITION, 0); - y_hotspot -= src_y_offset; - } else { - REG_SET_2(CURSOR_POSITION, 0, - CURSOR_X_POSITION, pos->x, - CURSOR_Y_POSITION, pos->y); - } + CURSOR_Y_POSITION, pos->y); REG_SET_2(CURSOR_HOT_SPOT, 0, CURSOR_HOT_SPOT_X, x_hotspot, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h index a6d6dfe00617..db98ba361686 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h @@ -247,7 +247,7 @@ .field_name = reg_name ## __ ## field_name ## post_fix /* Mask/shift struct generation macro for all ASICs (including those with reduced functionality) */ -#define HUBP_MASK_SH_LIST_DCN(mask_sh)\ +#define HUBP_MASK_SH_LIST_DCN_COMMON(mask_sh)\ HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_BLANK_EN, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_TTU_DISABLE, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_UNDERFLOW_STATUS, mask_sh),\ @@ -331,7 +331,6 @@ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG, META_CHUNK_SIZE, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG, MIN_META_CHUNK_SIZE, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG, DPTE_GROUP_SIZE, mask_sh),\ - HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG, MPTE_GROUP_SIZE, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG, SWATH_HEIGHT, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG, PTE_ROW_HEIGHT_LINEAR, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG_C, CHUNK_SIZE_C, mask_sh),\ @@ -339,7 +338,6 @@ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG_C, META_CHUNK_SIZE_C, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG_C, MIN_META_CHUNK_SIZE_C, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG_C, DPTE_GROUP_SIZE_C, mask_sh),\ - HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG_C, MPTE_GROUP_SIZE_C, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG_C, SWATH_HEIGHT_C, mask_sh),\ HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG_C, PTE_ROW_HEIGHT_LINEAR_C, mask_sh),\ HUBP_SF(HUBPREQ0_BLANK_OFFSET_0, REFCYC_H_BLANK_END, mask_sh),\ @@ -373,6 +371,11 @@ HUBP_SF(HUBPREQ0_DCN_SURF0_TTU_CNTL1, REFCYC_PER_REQ_DELIVERY_PRE, mask_sh),\ HUBP_SF(HUBP0_HUBP_CLK_CNTL, HUBP_CLOCK_ENABLE, mask_sh) +#define HUBP_MASK_SH_LIST_DCN(mask_sh)\ + HUBP_MASK_SH_LIST_DCN_COMMON(mask_sh),\ + HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG, MPTE_GROUP_SIZE, mask_sh),\ + HUBP_SF(HUBP0_DCHUBP_REQ_SIZE_CONFIG_C, MPTE_GROUP_SIZE_C, mask_sh) + /* Mask/shift struct generation macro for ASICs with VM */ #define HUBP_MASK_SH_LIST_DCN_VM(mask_sh)\ HUBP_SF(HUBPREQ0_NOM_PARAMETERS_0, DST_Y_PER_PTE_ROW_NOM_L, mask_sh),\ @@ -595,6 +598,9 @@ type AGP_BASE;\ type AGP_BOT;\ type AGP_TOP;\ + type DCN_VM_SYSTEM_APERTURE_DEFAULT_SYSTEM;\ + type DCN_VM_SYSTEM_APERTURE_DEFAULT_ADDR_MSB;\ + type DCN_VM_SYSTEM_APERTURE_DEFAULT_ADDR_LSB;\ /* todo: get these from GVM instead of reading registers ourselves */\ type PAGE_DIRECTORY_ENTRY_HI32;\ type PAGE_DIRECTORY_ENTRY_LO32;\ diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index d1a8f1c302a9..dab370676bfa 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -65,7 +65,7 @@ void print_microsec(struct dc_context *dc_ctx, struct dc_log_buffer_ctx *log_ctx, uint32_t ref_cycle) { - const uint32_t ref_clk_mhz = dc_ctx->dc->res_pool->ref_clock_inKhz / 1000; + const uint32_t ref_clk_mhz = dc_ctx->dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000; static const unsigned int frac = 1000; uint32_t us_x10 = (ref_cycle * frac) / ref_clk_mhz; @@ -345,13 +345,13 @@ void dcn10_log_hw_state(struct dc *dc, DTN_INFO("\nCALCULATED Clocks: dcfclk_khz:%d dcfclk_deep_sleep_khz:%d dispclk_khz:%d\n" "dppclk_khz:%d max_supported_dppclk_khz:%d fclk_khz:%d socclk_khz:%d\n\n", - dc->current_state->bw.dcn.clk.dcfclk_khz, - dc->current_state->bw.dcn.clk.dcfclk_deep_sleep_khz, - dc->current_state->bw.dcn.clk.dispclk_khz, - dc->current_state->bw.dcn.clk.dppclk_khz, - dc->current_state->bw.dcn.clk.max_supported_dppclk_khz, - dc->current_state->bw.dcn.clk.fclk_khz, - dc->current_state->bw.dcn.clk.socclk_khz); + dc->current_state->bw_ctx.bw.dcn.clk.dcfclk_khz, + dc->current_state->bw_ctx.bw.dcn.clk.dcfclk_deep_sleep_khz, + dc->current_state->bw_ctx.bw.dcn.clk.dispclk_khz, + dc->current_state->bw_ctx.bw.dcn.clk.dppclk_khz, + dc->current_state->bw_ctx.bw.dcn.clk.max_supported_dppclk_khz, + dc->current_state->bw_ctx.bw.dcn.clk.fclk_khz, + dc->current_state->bw_ctx.bw.dcn.clk.socclk_khz); log_mpc_crc(dc, log_ctx); @@ -714,7 +714,7 @@ static enum dc_status dcn10_enable_stream_timing( return DC_OK; } -static void reset_back_end_for_pipe( +static void dcn10_reset_back_end_for_pipe( struct dc *dc, struct pipe_ctx *pipe_ctx, struct dc_state *context) @@ -889,22 +889,23 @@ void hwss1_plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx) dcn10_verify_allow_pstate_change_high(dc); } -static void plane_atomic_power_down(struct dc *dc, struct pipe_ctx *pipe_ctx) +static void plane_atomic_power_down(struct dc *dc, + struct dpp *dpp, + struct hubp *hubp) { struct dce_hwseq *hws = dc->hwseq; - struct dpp *dpp = pipe_ctx->plane_res.dpp; DC_LOGGER_INIT(dc->ctx->logger); if (REG(DC_IP_REQUEST_CNTL)) { REG_SET(DC_IP_REQUEST_CNTL, 0, IP_REQUEST_EN, 1); dpp_pg_control(hws, dpp->inst, false); - hubp_pg_control(hws, pipe_ctx->plane_res.hubp->inst, false); + hubp_pg_control(hws, hubp->inst, false); dpp->funcs->dpp_reset(dpp); REG_SET(DC_IP_REQUEST_CNTL, 0, IP_REQUEST_EN, 0); DC_LOG_DEBUG( - "Power gated front end %d\n", pipe_ctx->pipe_idx); + "Power gated front end %d\n", hubp->inst); } } @@ -931,7 +932,9 @@ static void plane_atomic_disable(struct dc *dc, struct pipe_ctx *pipe_ctx) hubp->power_gated = true; dc->optimized_required = false; /* We're powering off, no need to optimize */ - plane_atomic_power_down(dc, pipe_ctx); + plane_atomic_power_down(dc, + pipe_ctx->plane_res.dpp, + pipe_ctx->plane_res.hubp); pipe_ctx->stream = NULL; memset(&pipe_ctx->stream_res, 0, sizeof(pipe_ctx->stream_res)); @@ -976,16 +979,14 @@ static void dcn10_init_pipes(struct dc *dc, struct dc_state *context) * to non-preferred front end. If pipe_ctx->stream is not NULL, * we will use the pipe, so don't disable */ - if (pipe_ctx->stream != NULL) + if (pipe_ctx->stream != NULL && can_apply_seamless_boot) continue; - if (tg->funcs->is_tg_enabled(tg)) - tg->funcs->lock(tg); - /* Blank controller using driver code instead of * command table. */ if (tg->funcs->is_tg_enabled(tg)) { + tg->funcs->lock(tg); tg->funcs->set_blank(tg, true); hwss_wait_for_blank_complete(tg); } @@ -1001,16 +1002,19 @@ static void dcn10_init_pipes(struct dc *dc, struct dc_state *context) struct dpp *dpp = dc->res_pool->dpps[i]; struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; - // W/A for issue with dc_post_update_surfaces_to_stream - hubp->power_gated = true; - /* There is assumption that pipe_ctx is not mapping irregularly * to non-preferred front end. If pipe_ctx->stream is not NULL, * we will use the pipe, so don't disable */ - if (pipe_ctx->stream != NULL) + if (can_apply_seamless_boot && + pipe_ctx->stream != NULL && + pipe_ctx->stream_res.tg->funcs->is_tg_enabled( + pipe_ctx->stream_res.tg)) continue; + /* Disable on the current state so the new one isn't cleared. */ + pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i]; + dpp->funcs->dpp_reset(dpp); pipe_ctx->stream_res.tg = tg; @@ -1108,6 +1112,22 @@ static void dcn10_init_hw(struct dc *dc) link->link_status.link_active = true; } + /* If taking control over from VBIOS, we may want to optimize our first + * mode set, so we need to skip powering down pipes until we know which + * pipes we want to use. + * Otherwise, if taking control is not possible, we need to power + * everything down. + */ + if (dcb->funcs->is_accelerated_mode(dcb)) { + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct hubp *hubp = dc->res_pool->hubps[i]; + struct dpp *dpp = dc->res_pool->dpps[i]; + + dc->res_pool->opps[i]->mpc_tree_params.opp_id = dc->res_pool->opps[i]->inst; + plane_atomic_power_down(dc, dpp, hubp); + } + } + for (i = 0; i < dc->res_pool->audio_count; i++) { struct audio *audio = dc->res_pool->audios[i]; @@ -1137,12 +1157,9 @@ static void dcn10_init_hw(struct dc *dc) enable_power_gating_plane(dc->hwseq, true); memset(&dc->res_pool->clk_mgr->clks, 0, sizeof(dc->res_pool->clk_mgr->clks)); - - if (dc->hwss.init_pipes) - dc->hwss.init_pipes(dc, dc->current_state); } -static void reset_hw_ctx_wrap( +static void dcn10_reset_hw_ctx_wrap( struct dc *dc, struct dc_state *context) { @@ -1164,10 +1181,9 @@ static void reset_hw_ctx_wrap( pipe_need_reprogram(pipe_ctx_old, pipe_ctx)) { struct clock_source *old_clk = pipe_ctx_old->clock_source; - reset_back_end_for_pipe(dc, pipe_ctx_old, dc->current_state); - if (dc->hwss.enable_stream_gating) { + dcn10_reset_back_end_for_pipe(dc, pipe_ctx_old, dc->current_state); + if (dc->hwss.enable_stream_gating) dc->hwss.enable_stream_gating(dc, pipe_ctx); - } if (old_clk) old_clk->funcs->cs_power_down(old_clk); } @@ -1836,7 +1852,7 @@ void dcn10_get_hdr_visual_confirm_color( switch (top_pipe_ctx->plane_res.scl_data.format) { case PIXEL_FORMAT_ARGB2101010: - if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_UNITY) { + if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_PQ) { /* HDR10, ARGB2101010 - set boarder color to red */ color->color_r_cr = color_value; } @@ -1931,7 +1947,7 @@ static void update_dpp(struct dpp *dpp, struct dc_plane_state *plane_state) plane_state->format, EXPANSION_MODE_ZERO, plane_state->input_csc_color_matrix, - COLOR_SPACE_YCBCR601_LIMITED); + plane_state->color_space); //set scale and bias registers build_prescale_params(&bns_params, plane_state); @@ -2051,7 +2067,7 @@ void update_dchubp_dpp( * divided by 2 */ if (plane_state->update_flags.bits.full_update) { - bool should_divided_by_2 = context->bw.dcn.clk.dppclk_khz <= + bool should_divided_by_2 = context->bw_ctx.bw.dcn.clk.dppclk_khz <= dc->res_pool->clk_mgr->clks.dispclk_khz / 2; dpp->funcs->dpp_dppclk_control( @@ -2120,6 +2136,9 @@ void update_dchubp_dpp( if (pipe_ctx->stream->cursor_attributes.address.quad_part != 0) { dc->hwss.set_cursor_position(pipe_ctx); dc->hwss.set_cursor_attribute(pipe_ctx); + + if (dc->hwss.set_cursor_sdr_white_level) + dc->hwss.set_cursor_sdr_white_level(pipe_ctx); } if (plane_state->update_flags.bits.full_update) { @@ -2310,6 +2329,7 @@ static void dcn10_apply_ctx_for_surface( int i; struct timing_generator *tg; bool removed_pipe[4] = { false }; + bool interdependent_update = false; struct pipe_ctx *top_pipe_to_program = find_top_pipe_for_stream(dc, context, stream); DC_LOGGER_INIT(dc->ctx->logger); @@ -2319,7 +2339,13 @@ static void dcn10_apply_ctx_for_surface( tg = top_pipe_to_program->stream_res.tg; - dcn10_pipe_control_lock(dc, top_pipe_to_program, true); + interdependent_update = top_pipe_to_program->plane_state && + top_pipe_to_program->plane_state->update_flags.bits.full_update; + + if (interdependent_update) + lock_all_pipes(dc, context, true); + else + dcn10_pipe_control_lock(dc, top_pipe_to_program, true); if (num_planes == 0) { /* OTG blank before remove all front end */ @@ -2339,15 +2365,9 @@ static void dcn10_apply_ctx_for_surface( */ if (pipe_ctx->plane_state && !old_pipe_ctx->plane_state) { if (old_pipe_ctx->stream_res.tg == tg && - old_pipe_ctx->plane_res.hubp && - old_pipe_ctx->plane_res.hubp->opp_id != 0xf) { + old_pipe_ctx->plane_res.hubp && + old_pipe_ctx->plane_res.hubp->opp_id != 0xf) dcn10_disable_plane(dc, old_pipe_ctx); - /* - * power down fe will unlock when calling reset, need - * to lock it back here. Messy, need rework. - */ - pipe_ctx->stream_res.tg->funcs->lock(pipe_ctx->stream_res.tg); - } } if ((!pipe_ctx->plane_state || @@ -2366,29 +2386,25 @@ static void dcn10_apply_ctx_for_surface( if (num_planes > 0) program_all_pipe_in_tree(dc, top_pipe_to_program, context); - dcn10_pipe_control_lock(dc, top_pipe_to_program, false); - - if (top_pipe_to_program->plane_state && - top_pipe_to_program->plane_state->update_flags.bits.full_update) + if (interdependent_update) for (i = 0; i < dc->res_pool->pipe_count; i++) { struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; - tg = pipe_ctx->stream_res.tg; /* Skip inactive pipes and ones already updated */ - if (!pipe_ctx->stream || pipe_ctx->stream == stream - || !pipe_ctx->plane_state - || !tg->funcs->is_tg_enabled(tg)) + if (!pipe_ctx->stream || pipe_ctx->stream == stream || + !pipe_ctx->plane_state || !tg->funcs->is_tg_enabled(tg)) continue; - tg->funcs->lock(tg); - pipe_ctx->plane_res.hubp->funcs->hubp_setup_interdependent( pipe_ctx->plane_res.hubp, &pipe_ctx->dlg_regs, &pipe_ctx->ttu_regs); - - tg->funcs->unlock(tg); } + if (interdependent_update) + lock_all_pipes(dc, context, false); + else + dcn10_pipe_control_lock(dc, top_pipe_to_program, false); + if (num_planes == 0) false_optc_underflow_wa(dc, stream, tg); @@ -2425,7 +2441,7 @@ static void dcn10_prepare_bandwidth( if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { if (context->stream_count == 0) - context->bw.dcn.clk.phyclk_khz = 0; + context->bw_ctx.bw.dcn.clk.phyclk_khz = 0; dc->res_pool->clk_mgr->funcs->update_clocks( dc->res_pool->clk_mgr, @@ -2434,8 +2450,8 @@ static void dcn10_prepare_bandwidth( } hubbub1_program_watermarks(dc->res_pool->hubbub, - &context->bw.dcn.watermarks, - dc->res_pool->ref_clock_inKhz / 1000, + &context->bw_ctx.bw.dcn.watermarks, + dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000, true); dcn10_stereo_hw_frame_pack_wa(dc, context); @@ -2455,7 +2471,7 @@ static void dcn10_optimize_bandwidth( if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { if (context->stream_count == 0) - context->bw.dcn.clk.phyclk_khz = 0; + context->bw_ctx.bw.dcn.clk.phyclk_khz = 0; dc->res_pool->clk_mgr->funcs->update_clocks( dc->res_pool->clk_mgr, @@ -2464,8 +2480,8 @@ static void dcn10_optimize_bandwidth( } hubbub1_program_watermarks(dc->res_pool->hubbub, - &context->bw.dcn.watermarks, - dc->res_pool->ref_clock_inKhz / 1000, + &context->bw_ctx.bw.dcn.watermarks, + dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000, true); dcn10_stereo_hw_frame_pack_wa(dc, context); @@ -2654,7 +2670,7 @@ static void dcn10_update_pending_status(struct pipe_ctx *pipe_ctx) flip_pending = pipe_ctx->plane_res.hubp->funcs->hubp_is_flip_pending( pipe_ctx->plane_res.hubp); - plane_state->status.is_flip_pending = flip_pending; + plane_state->status.is_flip_pending = plane_state->status.is_flip_pending || flip_pending; if (!flip_pending) plane_state->status.current_address = plane_state->status.requested_address; @@ -2685,16 +2701,22 @@ static void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx) struct dpp *dpp = pipe_ctx->plane_res.dpp; struct dc_cursor_mi_param param = { .pixel_clk_khz = pipe_ctx->stream->timing.pix_clk_100hz / 10, - .ref_clk_khz = pipe_ctx->stream->ctx->dc->res_pool->ref_clock_inKhz, + .ref_clk_khz = pipe_ctx->stream->ctx->dc->res_pool->ref_clocks.dchub_ref_clock_inKhz, .viewport = pipe_ctx->plane_res.scl_data.viewport, .h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz, .v_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.vert, .rotation = pipe_ctx->plane_state->rotation, .mirror = pipe_ctx->plane_state->horizontal_mirror }; + uint32_t x_plane = pipe_ctx->plane_state->dst_rect.x; + uint32_t y_plane = pipe_ctx->plane_state->dst_rect.y; + uint32_t x_offset = min(x_plane, pos_cpy.x); + uint32_t y_offset = min(y_plane, pos_cpy.y); - pos_cpy.x_hotspot += pipe_ctx->plane_state->dst_rect.x; - pos_cpy.y_hotspot += pipe_ctx->plane_state->dst_rect.y; + pos_cpy.x -= x_offset; + pos_cpy.y -= y_offset; + pos_cpy.x_hotspot += (x_plane - x_offset); + pos_cpy.y_hotspot += (y_plane - y_offset); if (pipe_ctx->plane_state->address.type == PLN_ADDR_TYPE_VIDEO_PROGRESSIVE) @@ -2789,6 +2811,33 @@ int get_vupdate_offset_from_vsync(struct pipe_ctx *pipe_ctx) return vertical_line_start; } +void lock_all_pipes(struct dc *dc, + struct dc_state *context, + bool lock) +{ + struct pipe_ctx *pipe_ctx; + struct timing_generator *tg; + int i; + + for (i = 0; i < dc->res_pool->pipe_count; i++) { + pipe_ctx = &context->res_ctx.pipe_ctx[i]; + tg = pipe_ctx->stream_res.tg; + /* + * Only lock the top pipe's tg to prevent redundant + * (un)locking. Also skip if pipe is disabled. + */ + if (pipe_ctx->top_pipe || + !pipe_ctx->stream || !pipe_ctx->plane_state || + !tg->funcs->is_tg_enabled(tg)) + continue; + + if (lock) + tg->funcs->lock(tg); + else + tg->funcs->unlock(tg); + } +} + static void calc_vupdate_position( struct pipe_ctx *pipe_ctx, uint32_t *start_line, @@ -2882,6 +2931,29 @@ static void dcn10_setup_vupdate_interrupt(struct pipe_ctx *pipe_ctx) tg->funcs->setup_vertical_interrupt2(tg, start_line); } +static void dcn10_unblank_stream(struct pipe_ctx *pipe_ctx, + struct dc_link_settings *link_settings) +{ + struct encoder_unblank_param params = { { 0 } }; + struct dc_stream_state *stream = pipe_ctx->stream; + struct dc_link *link = stream->link; + + /* only 3 items below are used by unblank */ + params.timing = pipe_ctx->stream->timing; + + params.link_settings.link_rate = link_settings->link_rate; + + if (dc_is_dp_signal(pipe_ctx->stream->signal)) { + if (params.timing.pixel_encoding == PIXEL_ENCODING_YCBCR420) + params.timing.pix_clk_100hz /= 2; + pipe_ctx->stream_res.stream_enc->funcs->dp_unblank(pipe_ctx->stream_res.stream_enc, ¶ms); + } + + if (link->local_sink && link->local_sink->sink_signal == SIGNAL_TYPE_EDP) { + link->dc->hwss.edp_backlight_control(link, true); + } +} + static const struct hw_sequencer_funcs dcn10_funcs = { .program_gamut_remap = program_gamut_remap, .init_hw = dcn10_init_hw, @@ -2903,7 +2975,7 @@ static const struct hw_sequencer_funcs dcn10_funcs = { .update_info_frame = dce110_update_info_frame, .enable_stream = dce110_enable_stream, .disable_stream = dce110_disable_stream, - .unblank_stream = dce110_unblank_stream, + .unblank_stream = dcn10_unblank_stream, .blank_stream = dce110_blank_stream, .enable_audio_stream = dce110_enable_audio_stream, .disable_audio_stream = dce110_disable_audio_stream, @@ -2913,7 +2985,7 @@ static const struct hw_sequencer_funcs dcn10_funcs = { .pipe_control_lock = dcn10_pipe_control_lock, .prepare_bandwidth = dcn10_prepare_bandwidth, .optimize_bandwidth = dcn10_optimize_bandwidth, - .reset_hw_ctx_wrap = reset_hw_ctx_wrap, + .reset_hw_ctx_wrap = dcn10_reset_hw_ctx_wrap, .enable_stream_timing = dcn10_enable_stream_timing, .set_drr = set_drr, .get_position = get_position, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h index 6d66084df55f..4b3b27a5d23b 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h @@ -83,4 +83,8 @@ struct pipe_ctx *find_top_pipe_for_stream( int get_vupdate_offset_from_vsync(struct pipe_ctx *pipe_ctx); +void lock_all_pipes(struct dc *dc, + struct dc_state *context, + bool lock); + #endif /* __DC_HWSS_DCN10_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c index 98f41d250978..991622da9ed5 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c @@ -77,7 +77,7 @@ static unsigned int dcn10_get_hubbub_state(struct dc *dc, char *pBuf, unsigned i unsigned int chars_printed = 0; unsigned int remaining_buffer = bufSize; - const uint32_t ref_clk_mhz = dc_ctx->dc->res_pool->ref_clock_inKhz / 1000; + const uint32_t ref_clk_mhz = dc_ctx->dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000; static const unsigned int frac = 1000; memset(&wm, 0, sizeof(struct dcn_hubbub_wm)); @@ -115,7 +115,7 @@ static unsigned int dcn10_get_hubp_states(struct dc *dc, char *pBuf, unsigned in unsigned int chars_printed = 0; unsigned int remaining_buffer = bufSize; - const uint32_t ref_clk_mhz = dc_ctx->dc->res_pool->ref_clock_inKhz / 1000; + const uint32_t ref_clk_mhz = dc_ctx->dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000; static const unsigned int frac = 1000; if (invarOnly) @@ -472,12 +472,12 @@ static unsigned int dcn10_get_clock_states(struct dc *dc, char *pBuf, unsigned i chars_printed = snprintf_count(pBuf, bufSize, "dcfclk,dcfclk_deep_sleep,dispclk," "dppclk,fclk,socclk\n" "%d,%d,%d,%d,%d,%d\n", - dc->current_state->bw.dcn.clk.dcfclk_khz, - dc->current_state->bw.dcn.clk.dcfclk_deep_sleep_khz, - dc->current_state->bw.dcn.clk.dispclk_khz, - dc->current_state->bw.dcn.clk.dppclk_khz, - dc->current_state->bw.dcn.clk.fclk_khz, - dc->current_state->bw.dcn.clk.socclk_khz); + dc->current_state->bw_ctx.bw.dcn.clk.dcfclk_khz, + dc->current_state->bw_ctx.bw.dcn.clk.dcfclk_deep_sleep_khz, + dc->current_state->bw_ctx.bw.dcn.clk.dispclk_khz, + dc->current_state->bw_ctx.bw.dcn.clk.dppclk_khz, + dc->current_state->bw_ctx.bw.dcn.clk.fclk_khz, + dc->current_state->bw_ctx.bw.dcn.clk.socclk_khz); remaining_buffer -= chars_printed; pBuf += chars_printed; diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c index a9db372688ff..0126a44ba012 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c @@ -1304,7 +1304,6 @@ void dcn10_link_encoder_connect_dig_be_to_fe( #define HPD_REG_UPDATE_N(reg_name, n, ...) \ generic_reg_update_ex(CTX, \ HPD_REG(reg_name), \ - HPD_REG_READ(reg_name), \ n, __VA_ARGS__) #define HPD_REG_UPDATE(reg_name, field, val) \ @@ -1337,7 +1336,6 @@ void dcn10_link_encoder_disable_hpd(struct link_encoder *enc) #define AUX_REG_UPDATE_N(reg_name, n, ...) \ generic_reg_update_ex(CTX, \ AUX_REG(reg_name), \ - AUX_REG_READ(reg_name), \ n, __VA_ARGS__) #define AUX_REG_UPDATE(reg_name, field, val) \ diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c index 09d74070a49b..7eccb54c421d 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c @@ -516,6 +516,31 @@ static const struct resource_caps rv2_res_cap = { }; #endif +static const struct dc_plane_cap plane_cap = { + .type = DC_PLANE_TYPE_DCN_UNIVERSAL, + .blends_with_above = true, + .blends_with_below = true, + .per_pixel_alpha = true, + + .pixel_format_support = { + .argb8888 = true, + .nv12 = true, + .fp16 = true + }, + + .max_upscale_factor = { + .argb8888 = 16000, + .nv12 = 16000, + .fp16 = 1 + }, + + .max_downscale_factor = { + .argb8888 = 250, + .nv12 = 250, + .fp16 = 1 + } +}; + static const struct dc_debug_options debug_defaults_drv = { .sanity_checks = true, .disable_dmcu = true, @@ -848,14 +873,14 @@ void dcn10_clock_source_destroy(struct clock_source **clk_src) *clk_src = NULL; } -static struct pp_smu_funcs_rv *dcn10_pp_smu_create(struct dc_context *ctx) +static struct pp_smu_funcs *dcn10_pp_smu_create(struct dc_context *ctx) { - struct pp_smu_funcs_rv *pp_smu = kzalloc(sizeof(*pp_smu), GFP_KERNEL); + struct pp_smu_funcs *pp_smu = kzalloc(sizeof(*pp_smu), GFP_KERNEL); if (!pp_smu) return pp_smu; - dm_pp_get_funcs_rv(ctx, pp_smu); + dm_pp_get_funcs(ctx, pp_smu); return pp_smu; } @@ -865,10 +890,7 @@ static void destruct(struct dcn10_resource_pool *pool) for (i = 0; i < pool->base.stream_enc_count; i++) { if (pool->base.stream_enc[i] != NULL) { - /* TODO: free dcn version of stream encoder once implemented - * rather than using virtual stream encoder - */ - kfree(pool->base.stream_enc[i]); + kfree(DCN10STRENC_FROM_STRENC(pool->base.stream_enc[i])); pool->base.stream_enc[i] = NULL; } } @@ -921,9 +943,6 @@ static void destruct(struct dcn10_resource_pool *pool) } } - for (i = 0; i < pool->base.stream_enc_count; i++) - kfree(pool->base.stream_enc[i]); - for (i = 0; i < pool->base.audio_count; i++) { if (pool->base.audios[i]) dce_aud_destroy(&pool->base.audios[i]); @@ -1078,7 +1097,7 @@ static struct pipe_ctx *dcn10_acquire_idle_pipe_for_layer( { struct resource_context *res_ctx = &context->res_ctx; struct pipe_ctx *head_pipe = resource_get_head_pipe_for_stream(res_ctx, stream); - struct pipe_ctx *idle_pipe = find_idle_secondary_pipe(res_ctx, pool); + struct pipe_ctx *idle_pipe = find_idle_secondary_pipe(res_ctx, pool, head_pipe); if (!head_pipe) { ASSERT(0); @@ -1143,7 +1162,7 @@ static enum dc_status dcn10_validate_global(struct dc *dc, struct dc_state *cont continue; if (context->stream_status[i].plane_count > 2) - return false; + return DC_FAIL_UNSUPPORTED_1; for (j = 0; j < context->stream_status[i].plane_count; j++) { struct dc_plane_state *plane = @@ -1351,7 +1370,7 @@ static bool construct( goto fail; } - dml_init_instance(&dc->dml, DML_PROJECT_RAVEN1); + dml_init_instance(&dc->dml, &dcn1_0_soc, &dcn1_0_ip, DML_PROJECT_RAVEN1); memcpy(dc->dcn_ip, &dcn10_ip_defaults, sizeof(dcn10_ip_defaults)); memcpy(dc->dcn_soc, &dcn10_soc_defaults, sizeof(dcn10_soc_defaults)); @@ -1510,6 +1529,9 @@ static bool construct( dcn10_hw_sequencer_construct(dc); dc->caps.max_planes = pool->base.pipe_count; + for (i = 0; i < dc->caps.max_planes; ++i) + dc->caps.planes[i] = plane_cap; + dc->cap_funcs = cap_funcs; return true; @@ -1522,7 +1544,7 @@ fail: } struct resource_pool *dcn10_create_resource_pool( - uint8_t num_virtual_links, + const struct dc_init_data *init_data, struct dc *dc) { struct dcn10_resource_pool *pool = @@ -1531,7 +1553,7 @@ struct resource_pool *dcn10_create_resource_pool( if (!pool) return NULL; - if (construct(num_virtual_links, dc, pool)) + if (construct(init_data->num_virtual_links, dc, pool)) return &pool->base; BREAK_TO_DEBUGGER(); diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.h index 8f71225bc61b..999c684a0b36 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.h @@ -39,7 +39,7 @@ struct dcn10_resource_pool { struct resource_pool base; }; struct resource_pool *dcn10_create_resource_pool( - uint8_t num_virtual_links, + const struct dc_init_data *init_data, struct dc *dc); diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c index b08254121251..8ee9f6dc1d62 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c @@ -245,7 +245,8 @@ static void enc1_update_hdmi_info_packet( void enc1_stream_encoder_dp_set_stream_attribute( struct stream_encoder *enc, struct dc_crtc_timing *crtc_timing, - enum dc_color_space output_color_space) + enum dc_color_space output_color_space, + uint32_t enable_sdp_splitting) { uint32_t h_active_start; uint32_t v_active_start; @@ -298,7 +299,6 @@ void enc1_stream_encoder_dp_set_stream_attribute( break; case PIXEL_ENCODING_YCBCR420: dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_YCBCR420; - REG_UPDATE(DP_VID_TIMING, DP_VID_N_MUL, 1); break; default: dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_RGB444; @@ -726,13 +726,19 @@ void enc1_stream_encoder_update_dp_info_packets( 3, /* packetIndex */ &info_frame->hdrsmd); + if (info_frame->dpsdp.valid) + enc1_update_generic_info_packet( + enc1, + 4,/* packetIndex */ + &info_frame->dpsdp); + /* enable/disable transmission of packet(s). * If enabled, packet transmission begins on the next frame */ REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP0_ENABLE, info_frame->vsc.valid); REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP2_ENABLE, info_frame->spd.valid); REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP3_ENABLE, info_frame->hdrsmd.valid); - + REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP4_ENABLE, info_frame->dpsdp.valid); /* This bit is the master enable bit. * When enabling secondary stream engine, @@ -797,10 +803,10 @@ void enc1_stream_encoder_dp_blank( */ REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_DIS_DEFER, 2); /* Larger delay to wait until VBLANK - use max retry of - * 10us*3000=30ms. This covers 16.6ms of typical 60 Hz mode + + * 10us*5000=50ms. This covers 41.7ms of minimum 24 Hz mode + * a little more because we may not trust delay accuracy. */ - max_retries = DP_BLANK_MAX_RETRY * 150; + max_retries = DP_BLANK_MAX_RETRY * 250; /* disable DP stream */ REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, 0); @@ -833,14 +839,19 @@ void enc1_stream_encoder_dp_unblank( if (param->link_settings.link_rate != LINK_RATE_UNKNOWN) { uint32_t n_vid = 0x8000; uint32_t m_vid; + uint32_t n_multiply = 0; + uint64_t m_vid_l = n_vid; + /* YCbCr 4:2:0 : Computed VID_M will be 2X the input rate */ + if (param->timing.pixel_encoding == PIXEL_ENCODING_YCBCR420) { + /*this param->pixel_clk_khz is half of 444 rate for 420 already*/ + n_multiply = 1; + } /* M / N = Fstream / Flink * m_vid / n_vid = pixel rate / link rate */ - uint64_t m_vid_l = n_vid; - - m_vid_l *= param->pixel_clk_khz; + m_vid_l *= param->timing.pix_clk_100hz / 10; m_vid_l = div_u64(m_vid_l, param->link_settings.link_rate * LINK_RATE_REF_FREQ_IN_KHZ); @@ -859,7 +870,9 @@ void enc1_stream_encoder_dp_unblank( REG_UPDATE(DP_VID_M, DP_VID_M, m_vid); - REG_UPDATE(DP_VID_TIMING, DP_VID_M_N_GEN_EN, 1); + REG_UPDATE_2(DP_VID_TIMING, + DP_VID_M_N_GEN_EN, 1, + DP_VID_N_MUL, n_multiply); } /* set DIG_START to 0x1 to resync FIFO */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h index b7c800e10a32..e654c2f55971 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h @@ -462,7 +462,8 @@ void enc1_update_generic_info_packet( void enc1_stream_encoder_dp_set_stream_attribute( struct stream_encoder *enc, struct dc_crtc_timing *crtc_timing, - enum dc_color_space output_color_space); + enum dc_color_space output_color_space, + uint32_t enable_sdp_splitting); void enc1_stream_encoder_hdmi_set_stream_attribute( struct stream_encoder *enc, diff --git a/drivers/gpu/drm/amd/display/dc/dm_helpers.h b/drivers/gpu/drm/amd/display/dc/dm_helpers.h index e81b24374bcb..ccbfe9680d27 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_helpers.h +++ b/drivers/gpu/drm/amd/display/dc/dm_helpers.h @@ -58,7 +58,7 @@ bool dm_helpers_dp_mst_write_payload_allocation_table( bool enable); /* - * poll pending down reply before clear payload allocation table + * poll pending down reply */ void dm_helpers_dp_mst_poll_pending_down_reply( struct dc_context *ctx, diff --git a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h index 14bed5b1fa97..cc6891b8ea69 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h +++ b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h @@ -30,6 +30,8 @@ * interface to PPLIB/SMU to setup clocks and pstate requirements on SoC */ +typedef bool BOOLEAN; + enum pp_smu_ver { /* * PP_SMU_INTERFACE_X should be interpreted as the interface defined @@ -137,12 +139,6 @@ struct pp_smu_funcs_rv { /* PME w/a */ void (*set_pme_wa_enable)(struct pp_smu *pp); - /* - * Legacy functions. Used for backwards comp. with existing - * PPlib code. - */ - void (*set_display_requirement)(struct pp_smu *pp, - struct pp_smu_display_requirement_rv *req); }; struct pp_smu_funcs { diff --git a/drivers/gpu/drm/amd/display/dc/dm_services.h b/drivers/gpu/drm/amd/display/dc/dm_services.h index 1961cc6d9143..b426ba02b793 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_services.h +++ b/drivers/gpu/drm/amd/display/dc/dm_services.h @@ -52,30 +52,17 @@ irq_handler_idx dm_register_interrupt( * GPU registers access * */ - +uint32_t dm_read_reg_func( + const struct dc_context *ctx, + uint32_t address, + const char *func_name); /* enable for debugging new code, this adds 50k to the driver size. */ /* #define DM_CHECK_ADDR_0 */ #define dm_read_reg(ctx, address) \ dm_read_reg_func(ctx, address, __func__) -static inline uint32_t dm_read_reg_func( - const struct dc_context *ctx, - uint32_t address, - const char *func_name) -{ - uint32_t value; -#ifdef DM_CHECK_ADDR_0 - if (address == 0) { - DC_ERR("invalid register read; address = 0\n"); - return 0; - } -#endif - value = cgs_read_register(ctx->cgs_device, address); - trace_amdgpu_dc_rreg(&ctx->perf_trace->read_count, address, value); - return value; -} #define dm_write_reg(ctx, address, value) \ dm_write_reg_func(ctx, address, value, __func__) @@ -144,10 +131,14 @@ static inline uint32_t set_reg_field_value_ex( reg_name ## __ ## reg_field ## _MASK,\ reg_name ## __ ## reg_field ## __SHIFT) -uint32_t generic_reg_update_ex(const struct dc_context *ctx, +uint32_t generic_reg_set_ex(const struct dc_context *ctx, uint32_t addr, uint32_t reg_val, int n, uint8_t shift1, uint32_t mask1, uint32_t field_value1, ...); +uint32_t generic_reg_update_ex(const struct dc_context *ctx, + uint32_t addr, int n, + uint8_t shift1, uint32_t mask1, uint32_t field_value1, ...); + #define FD(reg_field) reg_field ## __SHIFT, \ reg_field ## _MASK @@ -155,7 +146,7 @@ uint32_t generic_reg_update_ex(const struct dc_context *ctx, * return number of poll before condition is met * return 0 if condition is not meet after specified time out tries */ -unsigned int generic_reg_wait(const struct dc_context *ctx, +void generic_reg_wait(const struct dc_context *ctx, uint32_t addr, uint32_t mask, uint32_t shift, uint32_t condition_value, unsigned int delay_between_poll_us, unsigned int time_out_num_tries, const char *func_name, int line); @@ -172,11 +163,10 @@ unsigned int generic_reg_wait(const struct dc_context *ctx, #define generic_reg_update_soc15(ctx, inst_offset, reg_name, n, ...)\ generic_reg_update_ex(ctx, DCE_BASE.instance[0].segment[mm##reg_name##_BASE_IDX] + mm##reg_name + inst_offset, \ - dm_read_reg_func(ctx, mm##reg_name + DCE_BASE.instance[0].segment[mm##reg_name##_BASE_IDX] + inst_offset, __func__), \ n, __VA_ARGS__) #define generic_reg_set_soc15(ctx, inst_offset, reg_name, n, ...)\ - generic_reg_update_ex(ctx, DCE_BASE.instance[0].segment[mm##reg_name##_BASE_IDX] + mm##reg_name + inst_offset, 0, \ + generic_reg_set_ex(ctx, DCE_BASE.instance[0].segment[mm##reg_name##_BASE_IDX] + mm##reg_name + inst_offset, 0, \ n, __VA_ARGS__) #define get_reg_field_value_soc15(reg_value, block, reg_num, reg_name, reg_field)\ @@ -223,8 +213,8 @@ bool dm_pp_notify_wm_clock_changes( const struct dc_context *ctx, struct dm_pp_wm_sets_with_clock_ranges *wm_with_clock_ranges); -void dm_pp_get_funcs_rv(struct dc_context *ctx, - struct pp_smu_funcs_rv *funcs); +void dm_pp_get_funcs(struct dc_context *ctx, + struct pp_smu_funcs *funcs); /* DAL calls this function to notify PP about completion of Mode Set. * For PP it means that current DCE clocks are those which were returned diff --git a/drivers/gpu/drm/amd/display/dc/dm_services_types.h b/drivers/gpu/drm/amd/display/dc/dm_services_types.h index 77200711abbe..a3d1be20dd9d 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_services_types.h +++ b/drivers/gpu/drm/amd/display/dc/dm_services_types.h @@ -29,7 +29,7 @@ #include "os_types.h" #include "dc_types.h" -struct pp_smu_funcs_rv; +struct pp_smu_funcs; struct dm_pp_clock_range { int min_khz; diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c index d303b789adfe..80ffd7d958b2 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c @@ -26,40 +26,14 @@ #include "display_mode_lib.h" #include "dc_features.h" -extern const struct _vcs_dpi_ip_params_st dcn1_0_ip; -extern const struct _vcs_dpi_soc_bounding_box_st dcn1_0_soc; - -static void set_soc_bounding_box(struct _vcs_dpi_soc_bounding_box_st *soc, enum dml_project project) +void dml_init_instance(struct display_mode_lib *lib, + const struct _vcs_dpi_soc_bounding_box_st *soc_bb, + const struct _vcs_dpi_ip_params_st *ip_params, + enum dml_project project) { - switch (project) { - case DML_PROJECT_RAVEN1: - *soc = dcn1_0_soc; - break; - default: - ASSERT(0); - break; - } -} - -static void set_ip_params(struct _vcs_dpi_ip_params_st *ip, enum dml_project project) -{ - switch (project) { - case DML_PROJECT_RAVEN1: - *ip = dcn1_0_ip; - break; - default: - ASSERT(0); - break; - } -} - -void dml_init_instance(struct display_mode_lib *lib, enum dml_project project) -{ - if (lib->project != project) { - set_soc_bounding_box(&lib->soc, project); - set_ip_params(&lib->ip, project); - lib->project = project; - } + lib->soc = *soc_bb; + lib->ip = *ip_params; + lib->project = project; } const char *dml_get_status_message(enum dm_validation_status status) diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h index a730e0209c05..1b546dba34bd 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h @@ -41,7 +41,10 @@ struct display_mode_lib { struct dal_logger *logger; }; -void dml_init_instance(struct display_mode_lib *lib, enum dml_project project); +void dml_init_instance(struct display_mode_lib *lib, + const struct _vcs_dpi_soc_bounding_box_st *soc_bb, + const struct _vcs_dpi_ip_params_st *ip_params, + enum dml_project project); const char *dml_get_status_message(enum dm_validation_status status); diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h index 391183e3428f..c5b791d158a7 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h @@ -25,6 +25,8 @@ #ifndef __DISPLAY_MODE_STRUCTS_H__ #define __DISPLAY_MODE_STRUCTS_H__ +#define MAX_CLOCK_LIMIT_STATES 8 + typedef struct _vcs_dpi_voltage_scaling_st voltage_scaling_st; typedef struct _vcs_dpi_soc_bounding_box_st soc_bounding_box_st; typedef struct _vcs_dpi_ip_params_st ip_params_st; @@ -103,7 +105,7 @@ struct _vcs_dpi_soc_bounding_box_st { double xfc_xbuf_latency_tolerance_us; int use_urgent_burst_bw; unsigned int num_states; - struct _vcs_dpi_voltage_scaling_st clock_limits[8]; + struct _vcs_dpi_voltage_scaling_st clock_limits[MAX_CLOCK_LIMIT_STATES]; }; struct _vcs_dpi_ip_params_st { @@ -416,6 +418,7 @@ struct _vcs_dpi_display_dlg_regs_st { unsigned int refcyc_per_vm_group_flip; unsigned int refcyc_per_vm_req_vblank; unsigned int refcyc_per_vm_req_flip; + unsigned int refcyc_per_vm_dmdata; }; struct _vcs_dpi_display_ttu_regs_st { diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c b/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c index 48400d642610..e2d82aacd3bc 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c +++ b/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c @@ -321,6 +321,9 @@ void print__dlg_regs_st(struct display_mode_lib *mode_lib, display_dlg_regs_st d dml_print( "DML_RQ_DLG_CALC: xfc_reg_remote_surface_flip_latency = 0x%0x\n", dlg_regs.xfc_reg_remote_surface_flip_latency); + dml_print( + "DML_RQ_DLG_CALC: refcyc_per_vm_dmdata = 0x%0x\n", + dlg_regs.refcyc_per_vm_dmdata); dml_print("DML_RQ_DLG_CALC: =====================================\n"); } diff --git a/drivers/gpu/drm/amd/display/dc/inc/clock_source.h b/drivers/gpu/drm/amd/display/dc/inc/clock_source.h index fe6301cb8681..1b01a9a58d14 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/clock_source.h +++ b/drivers/gpu/drm/amd/display/dc/inc/clock_source.h @@ -167,7 +167,7 @@ struct clock_source_funcs { struct pixel_clk_params *, struct pll_settings *); bool (*get_pixel_clk_frequency_100hz)( - struct clock_source *clock_source, + const struct clock_source *clock_source, unsigned int inst, unsigned int *pixel_clk_khz); }; diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_status.h b/drivers/gpu/drm/amd/display/dc/inc/core_status.h index 2e61a22ef4b2..8dca3b7700e5 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/core_status.h +++ b/drivers/gpu/drm/amd/display/dc/inc/core_status.h @@ -43,7 +43,7 @@ enum dc_status { DC_FAIL_BANDWIDTH_VALIDATE = 13, /* BW and Watermark validation */ DC_FAIL_SCALING = 14, DC_FAIL_DP_LINK_TRAINING = 15, - + DC_FAIL_UNSUPPORTED_1 = 18, DC_ERROR_UNEXPECTED = -1 }; diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h index 986ed1728644..88a82a23d259 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h +++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h @@ -95,10 +95,10 @@ struct resource_funcs { void (*link_init)(struct dc_link *link); struct link_encoder *(*link_enc_create)( const struct encoder_init_data *init); - bool (*validate_bandwidth)( struct dc *dc, - struct dc_state *context); + struct dc_state *context, + bool fast_validate); enum dc_status (*validate_global)( struct dc *dc, @@ -144,7 +144,7 @@ struct resource_pool { struct stream_encoder *stream_enc[MAX_PIPES * 2]; struct hubbub *hubbub; struct mpc *mpc; - struct pp_smu_funcs_rv *pp_smu; + struct pp_smu_funcs *pp_smu; struct pp_smu_display_requirement_rv pp_smu_req; struct dce_aux *engines[MAX_PIPES]; struct dce_i2c_hw *hw_i2cs[MAX_PIPES]; @@ -154,7 +154,12 @@ struct resource_pool { unsigned int pipe_count; unsigned int underlay_pipe_index; unsigned int stream_enc_count; - unsigned int ref_clock_inKhz; + + struct { + unsigned int xtalin_clock_inKhz; + unsigned int dccg_ref_clock_inKhz; + unsigned int dchub_ref_clock_inKhz; + } ref_clocks; unsigned int timing_generator_count; /* @@ -262,18 +267,22 @@ struct dcn_bw_output { struct dcn_watermark_set watermarks; }; -union bw_context { +union bw_output { struct dcn_bw_output dcn; struct dce_bw_output dce; }; +struct bw_context { + union bw_output bw; + struct display_mode_lib dml; +}; /** * struct dc_state - The full description of a state requested by a user * * @streams: Stream properties * @stream_status: The planes on a given stream * @res_ctx: Persistent state of resources - * @bw: The output from bandwidth and watermark calculations + * @bw_ctx: The output from bandwidth and watermark calculations and the DML * @pp_display_cfg: PowerPlay clocks and settings * @dcn_bw_vars: non-stack memory to support bandwidth calculations * @@ -285,7 +294,7 @@ struct dc_state { struct resource_context res_ctx; - union bw_context bw; + struct bw_context bw_ctx; /* Note: these are big structures, do *not* put on stack! */ struct dm_pp_display_configuration pp_display_cfg; @@ -293,7 +302,11 @@ struct dc_state { struct dcn_bw_internal_vars dcn_bw_vars; #endif - struct clk_mgr *dccg; + struct clk_mgr *clk_mgr; + + struct { + bool full_update_needed : 1; + } commit_hints; struct kref refcount; }; diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h index 16fd4dc6c4dd..b1fab251c09b 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h @@ -95,8 +95,9 @@ bool dal_ddc_service_query_ddc_data( uint8_t *read_buf, uint32_t read_size); -int dc_link_aux_transfer(struct ddc_service *ddc, - struct aux_payload *payload); +int dc_link_aux_transfer_raw(struct ddc_service *ddc, + struct aux_payload *payload, + enum aux_channel_operation_result *operation_result); bool dc_link_aux_transfer_with_retries(struct ddc_service *ddc, struct aux_payload *payload); diff --git a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h index ece954a40a8e..263c09630c06 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h @@ -621,7 +621,8 @@ extern const struct dcn_ip_params dcn10_ip_defaults; bool dcn_validate_bandwidth( struct dc *dc, - struct dc_state *context); + struct dc_state *context, + bool fast_validate); unsigned int dcn_find_dcfclk_suits_all( const struct dc *dc, @@ -631,5 +632,7 @@ void dcn_bw_update_from_pplib(struct dc *dc); void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc); void dcn_bw_sync_calcs_and_dml(struct dc *dc); +enum source_macro_tile_size swizzle_mode_to_macro_tile_size(enum swizzle_mode_values sw_mode); + #endif /* __DCN_CALCS_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h index 23a4b18e5fee..31bd6d5183ab 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h @@ -42,6 +42,8 @@ struct clk_mgr_funcs { bool safe_to_lower); int (*get_dp_ref_clk_frequency)(struct clk_mgr *clk_mgr); + + void (*init_clocks)(struct clk_mgr *clk_mgr); }; #endif /* __DAL_CLK_MGR_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h index 95a56d012626..05ee5295d2c1 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h @@ -39,6 +39,10 @@ struct dccg_funcs { void (*update_dpp_dto)(struct dccg *dccg, int dpp_inst, int req_dppclk); + void (*get_dccg_ref_freq)(struct dccg *dccg, + unsigned int xtalin_freq_inKhz, + unsigned int *dccg_ref_freq_inKhz); + void (*dccg_init)(struct dccg *dccg); }; #endif //__DAL_DCCG_H__ diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h index 9d2d8e51306c..5e8fead3c09a 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h @@ -73,6 +73,10 @@ struct hubbub_funcs { void (*wm_read_state)(struct hubbub *hubbub, struct dcn_hubbub_wm *wm); + + void (*get_dchub_ref_freq)(struct hubbub *hubbub, + unsigned int dccg_ref_freq_inKhz, + unsigned int *dchub_ref_freq_inKhz); }; struct hubbub { diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h index da85537a4488..4c8e2c6fb6db 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h @@ -146,6 +146,12 @@ struct out_csc_color_matrix { uint16_t regval[12]; }; +enum gamut_remap_select { + GAMUT_REMAP_BYPASS = 0, + GAMUT_REMAP_COEFF, + GAMUT_REMAP_COMA_COEFF, + GAMUT_REMAP_COMB_COEFF +}; enum opp_regamma { OPP_REGAMMA_BYPASS = 0, diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h index 4051493557bc..49854eb73d1d 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h @@ -63,11 +63,13 @@ struct encoder_info_frame { struct dc_info_packet vsc; /* HDR Static MetaData */ struct dc_info_packet hdrsmd; + /* custom sdp message */ + struct dc_info_packet dpsdp; }; struct encoder_unblank_param { struct dc_link_settings link_settings; - unsigned int pixel_clk_khz; + struct dc_crtc_timing timing; }; struct encoder_set_dp_phy_pattern_param { @@ -88,7 +90,8 @@ struct stream_encoder_funcs { void (*dp_set_stream_attribute)( struct stream_encoder *enc, struct dc_crtc_timing *crtc_timing, - enum dc_color_space output_color_space); + enum dc_color_space output_color_space, + uint32_t enable_sdp_splitting); void (*hdmi_set_stream_attribute)( struct stream_encoder *enc, diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h index c25f7df7b5e3..067d53caf28a 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h @@ -187,8 +187,10 @@ struct timing_generator_funcs { bool (*did_triggered_reset_occur)(struct timing_generator *tg); void (*setup_global_swap_lock)(struct timing_generator *tg, const struct dcp_gsl_params *gsl_params); + void (*setup_global_lock)(struct timing_generator *tg); void (*unlock)(struct timing_generator *tg); void (*lock)(struct timing_generator *tg); + void (*lock_global)(struct timing_generator *tg); void (*enable_reset_trigger)(struct timing_generator *tg, int source_tg_inst); void (*enable_crtc_reset)(struct timing_generator *tg, diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h index 7676f25216b1..33905468e2b9 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h @@ -176,6 +176,10 @@ struct hw_sequencer_funcs { struct dc *dc, struct pipe_ctx *pipe, bool lock); + void (*pipe_control_lock_global)( + struct dc *dc, + struct pipe_ctx *pipe, + bool lock); void (*blank_pixel_data)( struct dc *dc, struct pipe_ctx *pipe_ctx, diff --git a/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h b/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h index cf5a84b9e27c..8503d9cc4763 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h +++ b/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h @@ -52,7 +52,7 @@ /* macro to set register fields. */ #define REG_SET_N(reg_name, n, initial_val, ...) \ - generic_reg_update_ex(CTX, \ + generic_reg_set_ex(CTX, \ REG(reg_name), \ initial_val, \ n, __VA_ARGS__) @@ -225,7 +225,6 @@ #define REG_UPDATE_N(reg_name, n, ...) \ generic_reg_update_ex(CTX, \ REG(reg_name), \ - REG_READ(reg_name), \ n, __VA_ARGS__) #define REG_UPDATE(reg_name, field, val) \ @@ -380,16 +379,11 @@ /* macro to update a register field to specified values in given sequences. * useful when toggling bits */ -#define REG_UPDATE_SEQ(reg, field, value1, value2) \ -{ uint32_t val = REG_UPDATE(reg, field, value1); \ - REG_SET(reg, val, field, value2); } - -/* macro to update fields in register 1 field at a time in given order */ -#define REG_UPDATE_1BY1_2(reg, f1, v1, f2, v2) \ +#define REG_UPDATE_SEQ_2(reg, f1, v1, f2, v2) \ { uint32_t val = REG_UPDATE(reg, f1, v1); \ REG_SET(reg, val, f2, v2); } -#define REG_UPDATE_1BY1_3(reg, f1, v1, f2, v2, f3, v3) \ +#define REG_UPDATE_SEQ_3(reg, f1, v1, f2, v2, f3, v3) \ { uint32_t val = REG_UPDATE(reg, f1, v1); \ val = REG_SET(reg, val, f2, v2); \ REG_SET(reg, val, f3, v3); } diff --git a/drivers/gpu/drm/amd/display/dc/inc/resource.h b/drivers/gpu/drm/amd/display/dc/inc/resource.h index 0086a2f1d21a..3ce0a4fc5822 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/resource.h +++ b/drivers/gpu/drm/amd/display/dc/inc/resource.h @@ -70,11 +70,9 @@ bool resource_construct( struct resource_pool *pool, const struct resource_create_funcs *create_funcs); -struct resource_pool *dc_create_resource_pool( - struct dc *dc, - int num_virtual_links, - enum dce_version dc_version, - struct hw_asic_id asic_id); +struct resource_pool *dc_create_resource_pool(struct dc *dc, + const struct dc_init_data *init_data, + enum dce_version dc_version); void dc_destroy_resource_pool(struct dc *dc); @@ -131,7 +129,8 @@ bool resource_attach_surfaces_to_context( struct pipe_ctx *find_idle_secondary_pipe( struct resource_context *res_ctx, - const struct resource_pool *pool); + const struct resource_pool *pool, + const struct pipe_ctx *primary_pipe); bool resource_is_stream_unchanged( struct dc_state *old_context, struct dc_stream_state *stream); @@ -172,4 +171,7 @@ void update_audio_usage( unsigned int resource_pixel_format_to_bpp(enum surface_pixel_format format); +struct pipe_ctx *dc_res_get_odm_bottom_pipe(struct pipe_ctx *pipe_ctx); +bool dc_res_is_odm_head_pipe(struct pipe_ctx *pipe_ctx); + #endif /* DRIVERS_GPU_DRM_AMD_DC_DEV_DC_INC_RESOURCE_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c b/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c index afe0876fe6f8..86987f5e8bd5 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c +++ b/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c @@ -81,6 +81,11 @@ static const struct irq_source_info_funcs vblank_irq_info_funcs = { .ack = NULL }; +static const struct irq_source_info_funcs vupdate_irq_info_funcs = { + .set = NULL, + .ack = NULL +}; + #define hpd_int_entry(reg_num)\ [DC_IRQ_SOURCE_HPD1 + reg_num] = {\ .enable_reg = mmHPD ## reg_num ## _DC_HPD_INT_CONTROL,\ @@ -137,7 +142,7 @@ static const struct irq_source_info_funcs vblank_irq_info_funcs = { CRTC_V_UPDATE_INT_STATUS__CRTC_V_UPDATE_INT_CLEAR_MASK,\ .ack_value =\ CRTC_V_UPDATE_INT_STATUS__CRTC_V_UPDATE_INT_CLEAR_MASK,\ - .funcs = &vblank_irq_info_funcs\ + .funcs = &vupdate_irq_info_funcs\ } #define vblank_int_entry(reg_num)\ diff --git a/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c b/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c index 1ea7256ec89b..750ba0ab4106 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c +++ b/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c @@ -84,6 +84,11 @@ static const struct irq_source_info_funcs vblank_irq_info_funcs = { .ack = NULL }; +static const struct irq_source_info_funcs vupdate_irq_info_funcs = { + .set = NULL, + .ack = NULL +}; + #define BASE_INNER(seg) \ DCE_BASE__INST0_SEG ## seg @@ -140,7 +145,7 @@ static const struct irq_source_info_funcs vblank_irq_info_funcs = { IRQ_REG_ENTRY(CRTC, reg_num,\ CRTC_INTERRUPT_CONTROL, CRTC_V_UPDATE_INT_MSK,\ CRTC_V_UPDATE_INT_STATUS, CRTC_V_UPDATE_INT_CLEAR),\ - .funcs = &vblank_irq_info_funcs\ + .funcs = &vupdate_irq_info_funcs\ } #define vblank_int_entry(reg_num)\ diff --git a/drivers/gpu/drm/amd/display/dc/irq/dce80/irq_service_dce80.c b/drivers/gpu/drm/amd/display/dc/irq/dce80/irq_service_dce80.c index 8a2066c313fe..de218fe84a43 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/dce80/irq_service_dce80.c +++ b/drivers/gpu/drm/amd/display/dc/irq/dce80/irq_service_dce80.c @@ -84,6 +84,10 @@ static const struct irq_source_info_funcs vblank_irq_info_funcs = { .ack = NULL }; +static const struct irq_source_info_funcs vupdate_irq_info_funcs = { + .set = NULL, + .ack = NULL +}; #define hpd_int_entry(reg_num)\ [DC_IRQ_SOURCE_INVALID + reg_num] = {\ @@ -142,7 +146,7 @@ static const struct irq_source_info_funcs vblank_irq_info_funcs = { CRTC_V_UPDATE_INT_STATUS__CRTC_V_UPDATE_INT_CLEAR_MASK,\ .ack_value =\ CRTC_V_UPDATE_INT_STATUS__CRTC_V_UPDATE_INT_CLEAR_MASK,\ - .funcs = &vblank_irq_info_funcs\ + .funcs = &vupdate_irq_info_funcs\ } #define vblank_int_entry(reg_num)\ diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c b/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c index e04ae49243f6..10ac6deff5ff 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c +++ b/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c @@ -56,6 +56,18 @@ enum dc_irq_source to_dal_irq_source_dcn10( return DC_IRQ_SOURCE_VBLANK5; case DCN_1_0__SRCID__DC_D6_OTG_VSTARTUP: return DC_IRQ_SOURCE_VBLANK6; + case DCN_1_0__SRCID__OTG0_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE1; + case DCN_1_0__SRCID__OTG1_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE2; + case DCN_1_0__SRCID__OTG2_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE3; + case DCN_1_0__SRCID__OTG3_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE4; + case DCN_1_0__SRCID__OTG4_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE5; + case DCN_1_0__SRCID__OTG5_IHC_V_UPDATE_NO_LOCK_INTERRUPT: + return DC_IRQ_SOURCE_VUPDATE6; case DCN_1_0__SRCID__HUBP0_FLIP_INTERRUPT: return DC_IRQ_SOURCE_PFLIP1; case DCN_1_0__SRCID__HUBP1_FLIP_INTERRUPT: @@ -153,6 +165,11 @@ static const struct irq_source_info_funcs vblank_irq_info_funcs = { .ack = NULL }; +static const struct irq_source_info_funcs vupdate_no_lock_irq_info_funcs = { + .set = NULL, + .ack = NULL +}; + #define BASE_INNER(seg) \ DCE_BASE__INST0_SEG ## seg @@ -203,12 +220,15 @@ static const struct irq_source_info_funcs vblank_irq_info_funcs = { .funcs = &pflip_irq_info_funcs\ } -#define vupdate_int_entry(reg_num)\ +/* vupdate_no_lock_int_entry maps to DC_IRQ_SOURCE_VUPDATEx, to match semantic + * of DCE's DC_IRQ_SOURCE_VUPDATEx. + */ +#define vupdate_no_lock_int_entry(reg_num)\ [DC_IRQ_SOURCE_VUPDATE1 + reg_num] = {\ IRQ_REG_ENTRY(OTG, reg_num,\ - OTG_GLOBAL_SYNC_STATUS, VUPDATE_INT_EN,\ - OTG_GLOBAL_SYNC_STATUS, VUPDATE_EVENT_CLEAR),\ - .funcs = &vblank_irq_info_funcs\ + OTG_GLOBAL_SYNC_STATUS, VUPDATE_NO_LOCK_INT_EN,\ + OTG_GLOBAL_SYNC_STATUS, VUPDATE_NO_LOCK_EVENT_CLEAR),\ + .funcs = &vupdate_no_lock_irq_info_funcs\ } #define vblank_int_entry(reg_num)\ @@ -315,12 +335,12 @@ irq_source_info_dcn10[DAL_IRQ_SOURCES_NUMBER] = { dc_underflow_int_entry(6), [DC_IRQ_SOURCE_DMCU_SCP] = dummy_irq_entry(), [DC_IRQ_SOURCE_VBIOS_SW] = dummy_irq_entry(), - vupdate_int_entry(0), - vupdate_int_entry(1), - vupdate_int_entry(2), - vupdate_int_entry(3), - vupdate_int_entry(4), - vupdate_int_entry(5), + vupdate_no_lock_int_entry(0), + vupdate_no_lock_int_entry(1), + vupdate_no_lock_int_entry(2), + vupdate_no_lock_int_entry(3), + vupdate_no_lock_int_entry(4), + vupdate_no_lock_int_entry(5), vblank_int_entry(0), vblank_int_entry(1), vblank_int_entry(2), diff --git a/drivers/gpu/drm/amd/display/dc/virtual/virtual_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/virtual/virtual_stream_encoder.c index 3dc1733eea20..fdcf9e66d852 100644 --- a/drivers/gpu/drm/amd/display/dc/virtual/virtual_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/virtual/virtual_stream_encoder.c @@ -29,7 +29,8 @@ static void virtual_stream_encoder_dp_set_stream_attribute( struct stream_encoder *enc, struct dc_crtc_timing *crtc_timing, - enum dc_color_space output_color_space) {} + enum dc_color_space output_color_space, + uint32_t enable_sdp_splitting) {} static void virtual_stream_encoder_hdmi_set_stream_attribute( struct stream_encoder *enc, diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h index 52a73332befb..89ef9f6860e5 100644 --- a/drivers/gpu/drm/amd/display/include/fixed31_32.h +++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h @@ -503,6 +503,8 @@ static inline int dc_fixpt_ceil(struct fixed31_32 arg) * fractional */ +unsigned int dc_fixpt_u4d19(struct fixed31_32 arg); + unsigned int dc_fixpt_u3d19(struct fixed31_32 arg); unsigned int dc_fixpt_u2d19(struct fixed31_32 arg); diff --git a/drivers/gpu/drm/amd/display/include/signal_types.h b/drivers/gpu/drm/amd/display/include/signal_types.h index f56d2891475f..beed70179bb5 100644 --- a/drivers/gpu/drm/amd/display/include/signal_types.h +++ b/drivers/gpu/drm/amd/display/include/signal_types.h @@ -45,6 +45,11 @@ enum signal_type { }; /* help functions for signal types manipulation */ +static inline bool dc_is_hdmi_tmds_signal(enum signal_type signal) +{ + return (signal == SIGNAL_TYPE_HDMI_TYPE_A); +} + static inline bool dc_is_hdmi_signal(enum signal_type signal) { return (signal == SIGNAL_TYPE_HDMI_TYPE_A); diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c index 0fbc8fbc3541..a1055413bade 100644 --- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c +++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c @@ -1854,6 +1854,8 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf, coordinates_x, axis_x, curve, MAX_HW_POINTS, tf_pts, mapUserRamp && ramp && ramp->type == GAMMA_RGB_256); + if (ramp->type == GAMMA_CUSTOM) + apply_lut_1d(ramp, MAX_HW_POINTS, tf_pts); ret = true; diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c index bfd27f10879e..3d867e34f8b3 100644 --- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c +++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c @@ -37,6 +37,8 @@ #define RENDER_TIMES_MAX_COUNT 10 /* Threshold to exit BTR (to avoid frequent enter-exits at the lower limit) */ #define BTR_EXIT_MARGIN 2000 +/* Threshold to change BTR multiplier (to avoid frequent changes) */ +#define BTR_DRIFT_MARGIN 2000 /*Threshold to exit fixed refresh rate*/ #define FIXED_REFRESH_EXIT_MARGIN_IN_HZ 4 /* Number of consecutive frames to check before entering/exiting fixed refresh*/ @@ -48,6 +50,93 @@ struct core_freesync { struct dc *dc; }; +void setFieldWithMask(unsigned char *dest, unsigned int mask, unsigned int value) +{ + unsigned int shift = 0; + + if (!mask || !dest) + return; + + while (!((mask >> shift) & 1)) + shift++; + + //reset + *dest = *dest & ~mask; + //set + //dont let value span past mask + value = value & (mask >> shift); + //insert value + *dest = *dest | (value << shift); +} + +// VTEM Byte Offset +#define VRR_VTEM_PB0 0 +#define VRR_VTEM_PB1 1 +#define VRR_VTEM_PB2 2 +#define VRR_VTEM_PB3 3 +#define VRR_VTEM_PB4 4 +#define VRR_VTEM_PB5 5 +#define VRR_VTEM_PB6 6 + +#define VRR_VTEM_MD0 7 +#define VRR_VTEM_MD1 8 +#define VRR_VTEM_MD2 9 +#define VRR_VTEM_MD3 10 + + +// VTEM Byte Masks +//PB0 +#define MASK__VRR_VTEM_PB0__RESERVED0 0x01 +#define MASK__VRR_VTEM_PB0__SYNC 0x02 +#define MASK__VRR_VTEM_PB0__VFR 0x04 +#define MASK__VRR_VTEM_PB0__AFR 0x08 +#define MASK__VRR_VTEM_PB0__DS_TYPE 0x30 + //0: Periodic pseudo-static EM Data Set + //1: Periodic dynamic EM Data Set + //2: Unique EM Data Set + //3: Reserved +#define MASK__VRR_VTEM_PB0__END 0x40 +#define MASK__VRR_VTEM_PB0__NEW 0x80 + +//PB1 +#define MASK__VRR_VTEM_PB1__RESERVED1 0xFF + +//PB2 +#define MASK__VRR_VTEM_PB2__ORGANIZATION_ID 0xFF + //0: This is a Vendor Specific EM Data Set + //1: This EM Data Set is defined by This Specification (HDMI 2.1 r102.clean) + //2: This EM Data Set is defined by CTA-861-G + //3: This EM Data Set is defined by VESA +//PB3 +#define MASK__VRR_VTEM_PB3__DATA_SET_TAG_MSB 0xFF +//PB4 +#define MASK__VRR_VTEM_PB4__DATA_SET_TAG_LSB 0xFF +//PB5 +#define MASK__VRR_VTEM_PB5__DATA_SET_LENGTH_MSB 0xFF +//PB6 +#define MASK__VRR_VTEM_PB6__DATA_SET_LENGTH_LSB 0xFF + + + +//PB7-27 (20 bytes): +//PB7 = MD0 +#define MASK__VRR_VTEM_MD0__VRR_EN 0x01 +#define MASK__VRR_VTEM_MD0__M_CONST 0x02 +#define MASK__VRR_VTEM_MD0__RESERVED2 0x0C +#define MASK__VRR_VTEM_MD0__FVA_FACTOR_M1 0xF0 + +//MD1 +#define MASK__VRR_VTEM_MD1__BASE_VFRONT 0xFF + +//MD2 +#define MASK__VRR_VTEM_MD2__BASE_REFRESH_RATE_98 0x03 +#define MASK__VRR_VTEM_MD2__RB 0x04 +#define MASK__VRR_VTEM_MD2__RESERVED3 0xF8 + +//MD3 +#define MASK__VRR_VTEM_MD3__BASE_REFRESH_RATE_07 0xFF + + #define MOD_FREESYNC_TO_CORE(mod_freesync)\ container_of(mod_freesync, struct core_freesync, public) @@ -248,6 +337,7 @@ static void apply_below_the_range(struct core_freesync *core_freesync, unsigned int frames_to_insert = 0; unsigned int min_frame_duration_in_ns = 0; unsigned int max_render_time_in_us = in_out_vrr->max_duration_in_us; + unsigned int delta_from_mid_point_delta_in_us; min_frame_duration_in_ns = ((unsigned int) (div64_u64( (1000000000ULL * 1000000), @@ -318,10 +408,27 @@ static void apply_below_the_range(struct core_freesync *core_freesync, /* Choose number of frames to insert based on how close it * can get to the mid point of the variable range. */ - if (delta_from_mid_point_in_us_1 < delta_from_mid_point_in_us_2) + if (delta_from_mid_point_in_us_1 < delta_from_mid_point_in_us_2) { frames_to_insert = mid_point_frames_ceil; - else + delta_from_mid_point_delta_in_us = delta_from_mid_point_in_us_2 - + delta_from_mid_point_in_us_1; + } else { frames_to_insert = mid_point_frames_floor; + delta_from_mid_point_delta_in_us = delta_from_mid_point_in_us_1 - + delta_from_mid_point_in_us_2; + } + + /* Prefer current frame multiplier when BTR is enabled unless it drifts + * too far from the midpoint + */ + if (in_out_vrr->btr.frames_to_insert != 0 && + delta_from_mid_point_delta_in_us < BTR_DRIFT_MARGIN) { + if (((last_render_time_in_us / in_out_vrr->btr.frames_to_insert) < + in_out_vrr->max_duration_in_us) && + ((last_render_time_in_us / in_out_vrr->btr.frames_to_insert) > + in_out_vrr->min_duration_in_us)) + frames_to_insert = in_out_vrr->btr.frames_to_insert; + } /* Either we've calculated the number of frames to insert, * or we need to insert min duration frames @@ -469,16 +576,14 @@ static void build_vrr_infopacket_header_vtem(enum signal_type signal, // HB0, HB1, HB2 indicates PacketType VTEMPacket infopacket->hb0 = 0x7F; infopacket->hb1 = 0xC0; - infopacket->hb2 = 0x00; - /* HB3 Bit Fields - * Reserved :1 = 0 - * Sync :1 = 0 - * VFR :1 = 1 - * Ds_Type :2 = 0 - * End :1 = 0 - * New :1 = 0 - */ - infopacket->hb3 = 0x20; + infopacket->hb2 = 0x00; //sequence_index + + setFieldWithMask(&infopacket->sb[VRR_VTEM_PB0], MASK__VRR_VTEM_PB0__VFR, 1); + setFieldWithMask(&infopacket->sb[VRR_VTEM_PB2], MASK__VRR_VTEM_PB2__ORGANIZATION_ID, 1); + setFieldWithMask(&infopacket->sb[VRR_VTEM_PB3], MASK__VRR_VTEM_PB3__DATA_SET_TAG_MSB, 0); + setFieldWithMask(&infopacket->sb[VRR_VTEM_PB4], MASK__VRR_VTEM_PB4__DATA_SET_TAG_LSB, 1); + setFieldWithMask(&infopacket->sb[VRR_VTEM_PB5], MASK__VRR_VTEM_PB5__DATA_SET_LENGTH_MSB, 0); + setFieldWithMask(&infopacket->sb[VRR_VTEM_PB6], MASK__VRR_VTEM_PB6__DATA_SET_LENGTH_LSB, 4); } static void build_vrr_infopacket_header_v1(enum signal_type signal, @@ -583,45 +688,36 @@ static void build_vrr_vtem_infopacket_data(const struct dc_stream_state *stream, const struct mod_vrr_params *vrr, struct dc_info_packet *infopacket) { - /* dc_info_packet to VtemPacket Translation of Bit-fields, - * SB[6] - * unsigned char VRR_EN :1 - * unsigned char M_CONST :1 - * unsigned char Reserved2 :2 - * unsigned char FVA_Factor_M1 :4 - * SB[7] - * unsigned char Base_Vfront :8 - * SB[8] - * unsigned char Base_Refresh_Rate_98 :2 - * unsigned char RB :1 - * unsigned char Reserved3 :5 - * SB[9] - * unsigned char Base_RefreshRate_07 :8 - */ unsigned int fieldRateInHz; if (vrr->state == VRR_STATE_ACTIVE_VARIABLE || - vrr->state == VRR_STATE_ACTIVE_FIXED){ - infopacket->sb[6] |= 0x80; //VRR_EN Bit = 1 + vrr->state == VRR_STATE_ACTIVE_FIXED) { + setFieldWithMask(&infopacket->sb[VRR_VTEM_MD0], MASK__VRR_VTEM_MD0__VRR_EN, 1); } else { - infopacket->sb[6] &= 0x7F; //VRR_EN Bit = 0 + setFieldWithMask(&infopacket->sb[VRR_VTEM_MD0], MASK__VRR_VTEM_MD0__VRR_EN, 0); } if (!stream->timing.vic) { - infopacket->sb[7] = stream->timing.v_front_porch; + setFieldWithMask(&infopacket->sb[VRR_VTEM_MD1], MASK__VRR_VTEM_MD1__BASE_VFRONT, + stream->timing.v_front_porch); + /* TODO: In dal2, we check mode flags for a reduced blanking timing. * Need a way to relay that information to this function. * if("ReducedBlanking") * { - * infopacket->sb[8] |= 0x20; //Set 3rd bit to 1 + * setFieldWithMask(&infopacket->sb[VRR_VTEM_MD2], MASK__VRR_VTEM_MD2__RB, 1; * } */ - fieldRateInHz = (stream->timing.pix_clk_100hz * 100)/ - (stream->timing.h_total * stream->timing.v_total); - infopacket->sb[8] |= ((fieldRateInHz & 0x300) >> 2); - infopacket->sb[9] |= fieldRateInHz & 0xFF; + //TODO: DAL2 does FixPoint and rounding. Here we might need to account for that + fieldRateInHz = (stream->timing.pix_clk_100hz * 100)/ + (stream->timing.h_total * stream->timing.v_total); + + setFieldWithMask(&infopacket->sb[VRR_VTEM_MD2], MASK__VRR_VTEM_MD2__BASE_REFRESH_RATE_98, + fieldRateInHz >> 8); + setFieldWithMask(&infopacket->sb[VRR_VTEM_MD3], MASK__VRR_VTEM_MD3__BASE_REFRESH_RATE_07, + fieldRateInHz); } infopacket->valid = true; @@ -745,6 +841,8 @@ static void build_vrr_infopacket_vtem(const struct dc_stream_state *stream, { //VTEM info packet for HdmiVrr + memset(infopacket, 0, sizeof(struct dc_info_packet)); + //VTEM Packet is structured differently build_vrr_infopacket_header_vtem(stream->signal, infopacket); build_vrr_vtem_infopacket_data(stream, vrr, infopacket); diff --git a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c index 038b88221c5f..b3810b864676 100644 --- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c +++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c @@ -41,9 +41,12 @@ static const unsigned char min_reduction_table[13] = { static const unsigned char max_reduction_table[13] = { 0xf5, 0xe5, 0xd9, 0xcd, 0xb1, 0xa5, 0xa5, 0x80, 0x65, 0x4d, 0x4d, 0x4d, 0x32}; -/* ABM 2.2 Min Reduction effectively disabled (100% for all configs)*/ +/* Possible ABM 2.2 Min Reduction configs from least aggressive to most aggressive + * 0 1 2 3 4 5 6 7 8 9 10 11 12 + * 100 100 100 100 100 100 100 100 100 92.2 83.1 75.3 75.3 % + */ static const unsigned char min_reduction_table_v_2_2[13] = { -0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff}; +0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xeb, 0xd4, 0xc0, 0xc0}; /* Possible ABM 2.2 Max Reduction configs from least aggressive to most aggressive * 0 1 2 3 4 5 6 7 8 9 10 11 12 @@ -408,9 +411,9 @@ void fill_iram_v_2_2(struct iram_table_v_2_2 *ram_table, struct dmcu_iram_parame ram_table->flags = 0x0; ram_table->deviation_gain[0] = 0xb3; - ram_table->deviation_gain[1] = 0xb3; - ram_table->deviation_gain[2] = 0xb3; - ram_table->deviation_gain[3] = 0xb3; + ram_table->deviation_gain[1] = 0xa8; + ram_table->deviation_gain[2] = 0x98; + ram_table->deviation_gain[3] = 0x68; ram_table->min_reduction[0][0] = min_reduction_table_v_2_2[abm_config[set][0]]; ram_table->min_reduction[1][0] = min_reduction_table_v_2_2[abm_config[set][0]]; @@ -505,7 +508,7 @@ void fill_iram_v_2_2(struct iram_table_v_2_2 *ram_table, struct dmcu_iram_parame ram_table->contrastFactor[0] = 0x99; ram_table->contrastFactor[1] = 0x99; - ram_table->contrastFactor[2] = 0x99; + ram_table->contrastFactor[2] = 0x90; ram_table->contrastFactor[3] = 0x80; ram_table->iir_curve[0] = 0x65; diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index 470d7b89071a..574bf6e70763 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -137,6 +137,7 @@ enum DC_FEATURE_MASK { DC_FBC_MASK = 0x1, }; +enum amd_dpm_forced_level; /** * struct amd_ip_funcs - general hooks for managing amdgpu IP Blocks */ @@ -186,6 +187,8 @@ struct amd_ip_funcs { enum amd_powergating_state state); /** @get_clockgating_state: get current clockgating status */ void (*get_clockgating_state)(void *handle, u32 *flags); + /** @enable_umd_pstate: enable UMD powerstate */ + int (*enable_umd_pstate)(void *handle, enum amd_dpm_forced_level *level); }; diff --git a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h index 442ca7c471a5..6109f5ad25ad 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h @@ -141,6 +141,8 @@ #define mmUVD_GPCOM_VCPU_DATA0_BASE_IDX 1 #define mmUVD_GPCOM_VCPU_DATA1 0x03c5 #define mmUVD_GPCOM_VCPU_DATA1_BASE_IDX 1 +#define mmUVD_ENGINE_CNTL 0x03c6 +#define mmUVD_ENGINE_CNTL_BASE_IDX 1 #define mmUVD_UDEC_DBW_UV_ADDR_CONFIG 0x03d2 #define mmUVD_UDEC_DBW_UV_ADDR_CONFIG_BASE_IDX 1 #define mmUVD_UDEC_ADDR_CONFIG 0x03d3 diff --git a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_sh_mask.h index 63457f9df4c5..f84bed6eecb9 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_sh_mask.h @@ -312,6 +312,11 @@ //UVD_GPCOM_VCPU_DATA1 #define UVD_GPCOM_VCPU_DATA1__DATA1__SHIFT 0x0 #define UVD_GPCOM_VCPU_DATA1__DATA1_MASK 0xFFFFFFFFL +//UVD_ENGINE_CNTL +#define UVD_ENGINE_CNTL__ENGINE_START_MASK 0x1 +#define UVD_ENGINE_CNTL__ENGINE_START__SHIFT 0x0 +#define UVD_ENGINE_CNTL__ENGINE_START_MODE_MASK 0x2 +#define UVD_ENGINE_CNTL__ENGINE_START_MODE__SHIFT 0x1 //UVD_UDEC_DBW_UV_ADDR_CONFIG #define UVD_UDEC_DBW_UV_ADDR_CONFIG__NUM_PIPES__SHIFT 0x0 #define UVD_UDEC_DBW_UV_ADDR_CONFIG__PIPE_INTERLEAVE_SIZE__SHIFT 0x3 diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h index 8eb0bb241210..08769b4b7a74 100644 --- a/drivers/gpu/drm/amd/include/atomfirmware.h +++ b/drivers/gpu/drm/amd/include/atomfirmware.h @@ -494,6 +494,9 @@ enum atombios_firmware_capability ATOM_FIRMWARE_CAP_FIRMWARE_POSTED = 0x00000001, ATOM_FIRMWARE_CAP_GPU_VIRTUALIZATION = 0x00000002, ATOM_FIRMWARE_CAP_WMI_SUPPORT = 0x00000040, + ATOM_FIRMWARE_CAP_HWEMU_ENABLE = 0x00000080, + ATOM_FIRMWARE_CAP_HWEMU_UMC_CFG = 0x00000100, + ATOM_FIRMWARE_CAP_SRAM_ECC = 0x00000200, }; enum atom_cooling_solution_id{ @@ -528,6 +531,35 @@ struct atom_firmware_info_v3_2 { uint32_t reserved2[3]; }; +struct atom_firmware_info_v3_3 +{ + struct atom_common_table_header table_header; + uint32_t firmware_revision; + uint32_t bootup_sclk_in10khz; + uint32_t bootup_mclk_in10khz; + uint32_t firmware_capability; // enum atombios_firmware_capability + uint32_t main_call_parser_entry; /* direct address of main parser call in VBIOS binary. */ + uint32_t bios_scratch_reg_startaddr; // 1st bios scratch register dword address + uint16_t bootup_vddc_mv; + uint16_t bootup_vddci_mv; + uint16_t bootup_mvddc_mv; + uint16_t bootup_vddgfx_mv; + uint8_t mem_module_id; + uint8_t coolingsolution_id; /*0: Air cooling; 1: Liquid cooling ... */ + uint8_t reserved1[2]; + uint32_t mc_baseaddr_high; + uint32_t mc_baseaddr_low; + uint8_t board_i2c_feature_id; // enum of atom_board_i2c_feature_id_def + uint8_t board_i2c_feature_gpio_id; // i2c id find in gpio_lut data table gpio_id + uint8_t board_i2c_feature_slave_addr; + uint8_t reserved3; + uint16_t bootup_mvddq_mv; + uint16_t bootup_mvpp_mv; + uint32_t zfbstartaddrin16mb; + uint32_t pplib_pptable_id; // if pplib_pptable_id!=0, pplib get powerplay table inside driver instead of from VBIOS + uint32_t reserved2[2]; +}; + /* *************************************************************************** Data Table lcd_info structure @@ -1226,16 +1258,17 @@ struct atom_gfx_info_v2_3 { uint32_t rm21_sram_vmin_value; }; -struct atom_gfx_info_v2_4 { +struct atom_gfx_info_v2_4 +{ struct atom_common_table_header table_header; uint8_t gfxip_min_ver; uint8_t gfxip_max_ver; - uint8_t gc_num_se; - uint8_t max_tile_pipes; - uint8_t gc_num_cu_per_sh; - uint8_t gc_num_sh_per_se; - uint8_t gc_num_rb_per_se; - uint8_t gc_num_tccs; + uint8_t max_shader_engines; + uint8_t reserved; + uint8_t max_cu_per_sh; + uint8_t max_sh_per_se; + uint8_t max_backends_per_se; + uint8_t max_texture_channel_caches; uint32_t regaddr_cp_dma_src_addr; uint32_t regaddr_cp_dma_src_addr_hi; uint32_t regaddr_cp_dma_dst_addr; @@ -1780,6 +1813,56 @@ struct atom_umc_info_v3_1 uint32_t mem_refclk_10khz; }; +// umc_info.umc_config +enum atom_umc_config_def { + UMC_CONFIG__ENABLE_1KB_INTERLEAVE_MODE = 0x00000001, + UMC_CONFIG__DEFAULT_MEM_ECC_ENABLE = 0x00000002, + UMC_CONFIG__ENABLE_HBM_LANE_REPAIR = 0x00000004, + UMC_CONFIG__ENABLE_BANK_HARVESTING = 0x00000008, + UMC_CONFIG__ENABLE_PHY_REINIT = 0x00000010, + UMC_CONFIG__DISABLE_UCODE_CHKSTATUS = 0x00000020, +}; + +struct atom_umc_info_v3_2 +{ + struct atom_common_table_header table_header; + uint32_t ucode_version; + uint32_t ucode_rom_startaddr; + uint32_t ucode_length; + uint16_t umc_reg_init_offset; + uint16_t customer_ucode_name_offset; + uint16_t mclk_ss_percentage; + uint16_t mclk_ss_rate_10hz; + uint8_t umcip_min_ver; + uint8_t umcip_max_ver; + uint8_t vram_type; //enum of atom_dgpu_vram_type + uint8_t umc_config; + uint32_t mem_refclk_10khz; + uint32_t pstate_uclk_10khz[4]; + uint16_t umcgoldenoffset; + uint16_t densitygoldenoffset; +}; + +struct atom_umc_info_v3_3 +{ + struct atom_common_table_header table_header; + uint32_t ucode_reserved; + uint32_t ucode_rom_startaddr; + uint32_t ucode_length; + uint16_t umc_reg_init_offset; + uint16_t customer_ucode_name_offset; + uint16_t mclk_ss_percentage; + uint16_t mclk_ss_rate_10hz; + uint8_t umcip_min_ver; + uint8_t umcip_max_ver; + uint8_t vram_type; //enum of atom_dgpu_vram_type + uint8_t umc_config; + uint32_t mem_refclk_10khz; + uint32_t pstate_uclk_10khz[4]; + uint16_t umcgoldenoffset; + uint16_t densitygoldenoffset; + uint32_t reserved[4]; +}; /* *************************************************************************** diff --git a/drivers/gpu/drm/amd/include/linux/chash.h b/drivers/gpu/drm/amd/include/linux/chash.h deleted file mode 100644 index 6dc159924ed1..000000000000 --- a/drivers/gpu/drm/amd/include/linux/chash.h +++ /dev/null @@ -1,366 +0,0 @@ -/* - * Copyright 2017 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - */ - -#ifndef _LINUX_CHASH_H -#define _LINUX_CHASH_H - -#include -#include -#include -#include - -#if BITS_PER_LONG == 32 -# define _CHASH_LONG_SHIFT 5 -#elif BITS_PER_LONG == 64 -# define _CHASH_LONG_SHIFT 6 -#else -# error "Unexpected BITS_PER_LONG" -#endif - -struct __chash_table { - u8 bits; - u8 key_size; - unsigned int value_size; - u32 size_mask; - unsigned long *occup_bitmap, *valid_bitmap; - union { - u32 *keys32; - u64 *keys64; - }; - u8 *values; - -#ifdef CONFIG_CHASH_STATS - u64 hits, hits_steps, hits_time_ns; - u64 miss, miss_steps, miss_time_ns; - u64 relocs, reloc_dist; -#endif -}; - -#define __CHASH_BITMAP_SIZE(bits) \ - (((1 << (bits)) + BITS_PER_LONG - 1) / BITS_PER_LONG) -#define __CHASH_ARRAY_SIZE(bits, size) \ - ((((size) << (bits)) + sizeof(long) - 1) / sizeof(long)) - -#define __CHASH_DATA_SIZE(bits, key_size, value_size) \ - (__CHASH_BITMAP_SIZE(bits) * 2 + \ - __CHASH_ARRAY_SIZE(bits, key_size) + \ - __CHASH_ARRAY_SIZE(bits, value_size)) - -#define STRUCT_CHASH_TABLE(bits, key_size, value_size) \ - struct { \ - struct __chash_table table; \ - unsigned long data \ - [__CHASH_DATA_SIZE(bits, key_size, value_size)];\ - } - -/** - * struct chash_table - Dynamically allocated closed hash table - * - * Use this struct for dynamically allocated hash tables (using - * chash_table_alloc and chash_table_free), where the size is - * determined at runtime. - */ -struct chash_table { - struct __chash_table table; - unsigned long *data; -}; - -/** - * DECLARE_CHASH_TABLE - macro to declare a closed hash table - * @table: name of the declared hash table - * @bts: Table size will be 2^bits entries - * @key_sz: Size of hash keys in bytes, 4 or 8 - * @val_sz: Size of data values in bytes, can be 0 - * - * This declares the hash table variable with a static size. - * - * The closed hash table stores key-value pairs with low memory and - * lookup overhead. In operation it performs no dynamic memory - * management. The data being stored does not require any - * list_heads. The hash table performs best with small @val_sz and as - * long as some space (about 50%) is left free in the table. But the - * table can still work reasonably efficiently even when filled up to - * about 90%. If bigger data items need to be stored and looked up, - * store the pointer to it as value in the hash table. - * - * @val_sz may be 0. This can be useful when all the stored - * information is contained in the key itself and the fact that it is - * in the hash table (or not). - */ -#define DECLARE_CHASH_TABLE(table, bts, key_sz, val_sz) \ - STRUCT_CHASH_TABLE(bts, key_sz, val_sz) table - -#ifdef CONFIG_CHASH_STATS -#define __CHASH_STATS_INIT(prefix), \ - prefix.hits = 0, \ - prefix.hits_steps = 0, \ - prefix.hits_time_ns = 0, \ - prefix.miss = 0, \ - prefix.miss_steps = 0, \ - prefix.miss_time_ns = 0, \ - prefix.relocs = 0, \ - prefix.reloc_dist = 0 -#else -#define __CHASH_STATS_INIT(prefix) -#endif - -#define __CHASH_TABLE_INIT(prefix, data, bts, key_sz, val_sz) \ - prefix.bits = (bts), \ - prefix.key_size = (key_sz), \ - prefix.value_size = (val_sz), \ - prefix.size_mask = ((1 << bts) - 1), \ - prefix.occup_bitmap = &data[0], \ - prefix.valid_bitmap = &data \ - [__CHASH_BITMAP_SIZE(bts)], \ - prefix.keys64 = (u64 *)&data \ - [__CHASH_BITMAP_SIZE(bts) * 2], \ - prefix.values = (u8 *)&data \ - [__CHASH_BITMAP_SIZE(bts) * 2 + \ - __CHASH_ARRAY_SIZE(bts, key_sz)] \ - __CHASH_STATS_INIT(prefix) - -/** - * DEFINE_CHASH_TABLE - macro to define and initialize a closed hash table - * @tbl: name of the declared hash table - * @bts: Table size will be 2^bits entries - * @key_sz: Size of hash keys in bytes, 4 or 8 - * @val_sz: Size of data values in bytes, can be 0 - * - * Note: the macro can be used for global and local hash table variables. - */ -#define DEFINE_CHASH_TABLE(tbl, bts, key_sz, val_sz) \ - DECLARE_CHASH_TABLE(tbl, bts, key_sz, val_sz) = { \ - .table = { \ - __CHASH_TABLE_INIT(, (tbl).data, bts, key_sz, val_sz) \ - }, \ - .data = {0} \ - } - -/** - * INIT_CHASH_TABLE - Initialize a hash table declared by DECLARE_CHASH_TABLE - * @tbl: name of the declared hash table - * @bts: Table size will be 2^bits entries - * @key_sz: Size of hash keys in bytes, 4 or 8 - * @val_sz: Size of data values in bytes, can be 0 - */ -#define INIT_CHASH_TABLE(tbl, bts, key_sz, val_sz) \ - __CHASH_TABLE_INIT(((tbl).table), (tbl).data, bts, key_sz, val_sz) - -int chash_table_alloc(struct chash_table *table, u8 bits, u8 key_size, - unsigned int value_size, gfp_t gfp_mask); -void chash_table_free(struct chash_table *table); - -/** - * chash_table_dump_stats - Dump statistics of a closed hash table - * @tbl: Pointer to the table structure - * - * Dumps some performance statistics of the table gathered in operation - * in the kernel log using pr_debug. If CONFIG_DYNAMIC_DEBUG is enabled, - * user must turn on messages for chash.c (file chash.c +p). - */ -#ifdef CONFIG_CHASH_STATS -#define chash_table_dump_stats(tbl) __chash_table_dump_stats(&(*tbl).table) - -void __chash_table_dump_stats(struct __chash_table *table); -#else -#define chash_table_dump_stats(tbl) -#endif - -/** - * chash_table_reset_stats - Reset statistics of a closed hash table - * @tbl: Pointer to the table structure - */ -#ifdef CONFIG_CHASH_STATS -#define chash_table_reset_stats(tbl) __chash_table_reset_stats(&(*tbl).table) - -static inline void __chash_table_reset_stats(struct __chash_table *table) -{ - (void)table __CHASH_STATS_INIT((*table)); -} -#else -#define chash_table_reset_stats(tbl) -#endif - -/** - * chash_table_copy_in - Copy a new value into the hash table - * @tbl: Pointer to the table structure - * @key: Key of the entry to add or update - * @value: Pointer to value to copy, may be NULL - * - * If @key already has an entry, its value is replaced. Otherwise a - * new entry is added. If @value is NULL, the value is left unchanged - * or uninitialized. Returns 1 if an entry already existed, 0 if a new - * entry was added or %-ENOMEM if there was no free space in the - * table. - */ -#define chash_table_copy_in(tbl, key, value) \ - __chash_table_copy_in(&(*tbl).table, key, value) - -int __chash_table_copy_in(struct __chash_table *table, u64 key, - const void *value); - -/** - * chash_table_copy_out - Copy a value out of the hash table - * @tbl: Pointer to the table structure - * @key: Key of the entry to find - * @value: Pointer to value to copy, may be NULL - * - * If @value is not NULL and the table has a non-0 value_size, the - * value at @key is copied to @value. Returns the slot index of the - * entry or %-EINVAL if @key was not found. - */ -#define chash_table_copy_out(tbl, key, value) \ - __chash_table_copy_out(&(*tbl).table, key, value, false) - -int __chash_table_copy_out(struct __chash_table *table, u64 key, - void *value, bool remove); - -/** - * chash_table_remove - Remove an entry from the hash table - * @tbl: Pointer to the table structure - * @key: Key of the entry to find - * @value: Pointer to value to copy, may be NULL - * - * If @value is not NULL and the table has a non-0 value_size, the - * value at @key is copied to @value. The entry is removed from the - * table. Returns the slot index of the removed entry or %-EINVAL if - * @key was not found. - */ -#define chash_table_remove(tbl, key, value) \ - __chash_table_copy_out(&(*tbl).table, key, value, true) - -/* - * Low level iterator API used internally by the above functions. - */ -struct chash_iter { - struct __chash_table *table; - unsigned long mask; - int slot; -}; - -/** - * CHASH_ITER_INIT - Initialize a hash table iterator - * @tbl: Pointer to hash table to iterate over - * @s: Initial slot number - */ -#define CHASH_ITER_INIT(table, s) { \ - table, \ - 1UL << ((s) & (BITS_PER_LONG - 1)), \ - s \ - } -/** - * CHASH_ITER_SET - Set hash table iterator to new slot - * @iter: Iterator - * @s: Slot number - */ -#define CHASH_ITER_SET(iter, s) \ - (iter).mask = 1UL << ((s) & (BITS_PER_LONG - 1)), \ - (iter).slot = (s) -/** - * CHASH_ITER_INC - Increment hash table iterator - * @table: Hash table to iterate over - * - * Wraps around at the end. - */ -#define CHASH_ITER_INC(iter) do { \ - (iter).mask = (iter).mask << 1 | \ - (iter).mask >> (BITS_PER_LONG - 1); \ - (iter).slot = ((iter).slot + 1) & (iter).table->size_mask; \ - } while (0) - -static inline bool chash_iter_is_valid(const struct chash_iter iter) -{ - BUG_ON((unsigned)iter.slot >= (1 << iter.table->bits)); - return !!(iter.table->valid_bitmap[iter.slot >> _CHASH_LONG_SHIFT] & - iter.mask); -} -static inline bool chash_iter_is_empty(const struct chash_iter iter) -{ - BUG_ON((unsigned)iter.slot >= (1 << iter.table->bits)); - return !(iter.table->occup_bitmap[iter.slot >> _CHASH_LONG_SHIFT] & - iter.mask); -} - -static inline void chash_iter_set_valid(const struct chash_iter iter) -{ - BUG_ON((unsigned)iter.slot >= (1 << iter.table->bits)); - iter.table->valid_bitmap[iter.slot >> _CHASH_LONG_SHIFT] |= iter.mask; - iter.table->occup_bitmap[iter.slot >> _CHASH_LONG_SHIFT] |= iter.mask; -} -static inline void chash_iter_set_invalid(const struct chash_iter iter) -{ - BUG_ON((unsigned)iter.slot >= (1 << iter.table->bits)); - iter.table->valid_bitmap[iter.slot >> _CHASH_LONG_SHIFT] &= ~iter.mask; -} -static inline void chash_iter_set_empty(const struct chash_iter iter) -{ - BUG_ON((unsigned)iter.slot >= (1 << iter.table->bits)); - iter.table->occup_bitmap[iter.slot >> _CHASH_LONG_SHIFT] &= ~iter.mask; -} - -static inline u32 chash_iter_key32(const struct chash_iter iter) -{ - BUG_ON(iter.table->key_size != 4); - BUG_ON((unsigned)iter.slot >= (1 << iter.table->bits)); - return iter.table->keys32[iter.slot]; -} -static inline u64 chash_iter_key64(const struct chash_iter iter) -{ - BUG_ON(iter.table->key_size != 8); - BUG_ON((unsigned)iter.slot >= (1 << iter.table->bits)); - return iter.table->keys64[iter.slot]; -} -static inline u64 chash_iter_key(const struct chash_iter iter) -{ - BUG_ON((unsigned)iter.slot >= (1 << iter.table->bits)); - return (iter.table->key_size == 4) ? - iter.table->keys32[iter.slot] : iter.table->keys64[iter.slot]; -} - -static inline u32 chash_iter_hash32(const struct chash_iter iter) -{ - BUG_ON(iter.table->key_size != 4); - return hash_32(chash_iter_key32(iter), iter.table->bits); -} - -static inline u32 chash_iter_hash64(const struct chash_iter iter) -{ - BUG_ON(iter.table->key_size != 8); - return hash_64(chash_iter_key64(iter), iter.table->bits); -} - -static inline u32 chash_iter_hash(const struct chash_iter iter) -{ - return (iter.table->key_size == 4) ? - hash_32(chash_iter_key32(iter), iter.table->bits) : - hash_64(chash_iter_key64(iter), iter.table->bits); -} - -static inline void *chash_iter_value(const struct chash_iter iter) -{ - BUG_ON((unsigned)iter.slot >= (1 << iter.table->bits)); - return iter.table->values + - ((unsigned long)iter.slot * iter.table->value_size); -} - -#endif /* _LINUX_CHASH_H */ diff --git a/drivers/gpu/drm/amd/lib/Kconfig b/drivers/gpu/drm/amd/lib/Kconfig deleted file mode 100644 index 776ef3434c10..000000000000 --- a/drivers/gpu/drm/amd/lib/Kconfig +++ /dev/null @@ -1,28 +0,0 @@ -menu "AMD Library routines" - -# -# Closed hash table -# -config CHASH - tristate - default DRM_AMDGPU - help - Statically sized closed hash table implementation with low - memory and CPU overhead. - -config CHASH_STATS - bool "Closed hash table performance statistics" - depends on CHASH - default n - help - Enable collection of performance statistics for closed hash tables. - -config CHASH_SELFTEST - bool "Closed hash table self test" - depends on CHASH - default n - help - Runs a selftest during module load. Several module parameters - are available to modify the behaviour of the test. - -endmenu diff --git a/drivers/gpu/drm/amd/lib/Makefile b/drivers/gpu/drm/amd/lib/Makefile deleted file mode 100644 index 690243001e1a..000000000000 --- a/drivers/gpu/drm/amd/lib/Makefile +++ /dev/null @@ -1,32 +0,0 @@ -# -# Copyright 2017 Advanced Micro Devices, Inc. -# -# Permission is hereby granted, free of charge, to any person obtaining a -# copy of this software and associated documentation files (the "Software"), -# to deal in the Software without restriction, including without limitation -# the rights to use, copy, modify, merge, publish, distribute, sublicense, -# and/or sell copies of the Software, and to permit persons to whom the -# Software is furnished to do so, subject to the following conditions: -# -# The above copyright notice and this permission notice shall be included in -# all copies or substantial portions of the Software. -# -# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR -# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, -# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL -# THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR -# OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, -# ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR -# OTHER DEALINGS IN THE SOFTWARE. -# -# -# Makefile for AMD library routines, which are used by AMD driver -# components. -# -# This is for common library routines that can be shared between AMD -# driver components or later moved to kernel/lib for sharing with -# other drivers. - -ccflags-y := -I$(src)/../include - -obj-$(CONFIG_CHASH) += chash.o diff --git a/drivers/gpu/drm/amd/lib/chash.c b/drivers/gpu/drm/amd/lib/chash.c deleted file mode 100644 index b8e45f356a1c..000000000000 --- a/drivers/gpu/drm/amd/lib/chash.c +++ /dev/null @@ -1,638 +0,0 @@ -/* - * Copyright 2017 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - */ - -#include -#include -#include -#include -#include -#include -#include -#include - -/** - * chash_table_alloc - Allocate closed hash table - * @table: Pointer to the table structure - * @bits: Table size will be 2^bits entries - * @key_size: Size of hash keys in bytes, 4 or 8 - * @value_size: Size of data values in bytes, can be 0 - */ -int chash_table_alloc(struct chash_table *table, u8 bits, u8 key_size, - unsigned int value_size, gfp_t gfp_mask) -{ - if (bits > 31) - return -EINVAL; - - if (key_size != 4 && key_size != 8) - return -EINVAL; - - table->data = kcalloc(__CHASH_DATA_SIZE(bits, key_size, value_size), - sizeof(long), gfp_mask); - if (!table->data) - return -ENOMEM; - - __CHASH_TABLE_INIT(table->table, table->data, - bits, key_size, value_size); - - return 0; -} -EXPORT_SYMBOL(chash_table_alloc); - -/** - * chash_table_free - Free closed hash table - * @table: Pointer to the table structure - */ -void chash_table_free(struct chash_table *table) -{ - kfree(table->data); -} -EXPORT_SYMBOL(chash_table_free); - -#ifdef CONFIG_CHASH_STATS - -#define DIV_FRAC(nom, denom, quot, frac, frac_digits) do { \ - u64 __nom = (nom); \ - u64 __denom = (denom); \ - u64 __quot, __frac; \ - u32 __rem; \ - \ - while (__denom >> 32) { \ - __nom >>= 1; \ - __denom >>= 1; \ - } \ - __quot = __nom; \ - __rem = do_div(__quot, __denom); \ - __frac = __rem * (frac_digits) + (__denom >> 1); \ - do_div(__frac, __denom); \ - (quot) = __quot; \ - (frac) = __frac; \ - } while (0) - -void __chash_table_dump_stats(struct __chash_table *table) -{ - struct chash_iter iter = CHASH_ITER_INIT(table, 0); - u32 filled = 0, empty = 0, tombstones = 0; - u64 quot1, quot2; - u32 frac1, frac2; - - do { - if (chash_iter_is_valid(iter)) - filled++; - else if (chash_iter_is_empty(iter)) - empty++; - else - tombstones++; - CHASH_ITER_INC(iter); - } while (iter.slot); - - pr_debug("chash: key size %u, value size %u\n", - table->key_size, table->value_size); - pr_debug(" Slots total/filled/empty/tombstones: %u / %u / %u / %u\n", - 1 << table->bits, filled, empty, tombstones); - if (table->hits > 0) { - DIV_FRAC(table->hits_steps, table->hits, quot1, frac1, 1000); - DIV_FRAC(table->hits * 1000, table->hits_time_ns, - quot2, frac2, 1000); - } else { - quot1 = quot2 = 0; - frac1 = frac2 = 0; - } - pr_debug(" Hits (avg.cost, rate): %llu (%llu.%03u, %llu.%03u M/s)\n", - table->hits, quot1, frac1, quot2, frac2); - if (table->miss > 0) { - DIV_FRAC(table->miss_steps, table->miss, quot1, frac1, 1000); - DIV_FRAC(table->miss * 1000, table->miss_time_ns, - quot2, frac2, 1000); - } else { - quot1 = quot2 = 0; - frac1 = frac2 = 0; - } - pr_debug(" Misses (avg.cost, rate): %llu (%llu.%03u, %llu.%03u M/s)\n", - table->miss, quot1, frac1, quot2, frac2); - if (table->hits + table->miss > 0) { - DIV_FRAC(table->hits_steps + table->miss_steps, - table->hits + table->miss, quot1, frac1, 1000); - DIV_FRAC((table->hits + table->miss) * 1000, - (table->hits_time_ns + table->miss_time_ns), - quot2, frac2, 1000); - } else { - quot1 = quot2 = 0; - frac1 = frac2 = 0; - } - pr_debug(" Total (avg.cost, rate): %llu (%llu.%03u, %llu.%03u M/s)\n", - table->hits + table->miss, quot1, frac1, quot2, frac2); - if (table->relocs > 0) { - DIV_FRAC(table->hits + table->miss, table->relocs, - quot1, frac1, 1000); - DIV_FRAC(table->reloc_dist, table->relocs, quot2, frac2, 1000); - pr_debug(" Relocations (freq, avg.dist): %llu (1:%llu.%03u, %llu.%03u)\n", - table->relocs, quot1, frac1, quot2, frac2); - } else { - pr_debug(" No relocations\n"); - } -} -EXPORT_SYMBOL(__chash_table_dump_stats); - -#undef DIV_FRAC -#endif - -#define CHASH_INC(table, a) ((a) = ((a) + 1) & (table)->size_mask) -#define CHASH_ADD(table, a, b) (((a) + (b)) & (table)->size_mask) -#define CHASH_SUB(table, a, b) (((a) - (b)) & (table)->size_mask) -#define CHASH_IN_RANGE(table, slot, first, last) \ - (CHASH_SUB(table, slot, first) <= CHASH_SUB(table, last, first)) - -/*#define CHASH_DEBUG Uncomment this to enable verbose debug output*/ -#ifdef CHASH_DEBUG -static void chash_table_dump(struct __chash_table *table) -{ - struct chash_iter iter = CHASH_ITER_INIT(table, 0); - - do { - if ((iter.slot & 3) == 0) - pr_debug("%04x: ", iter.slot); - - if (chash_iter_is_valid(iter)) - pr_debug("[%016llx] ", chash_iter_key(iter)); - else if (chash_iter_is_empty(iter)) - pr_debug("[ ] "); - else - pr_debug("[ ] "); - - if ((iter.slot & 3) == 3) - pr_debug("\n"); - - CHASH_ITER_INC(iter); - } while (iter.slot); - - if ((iter.slot & 3) != 0) - pr_debug("\n"); -} - -static int chash_table_check(struct __chash_table *table) -{ - u32 hash; - struct chash_iter iter = CHASH_ITER_INIT(table, 0); - struct chash_iter cur = CHASH_ITER_INIT(table, 0); - - do { - if (!chash_iter_is_valid(iter)) { - CHASH_ITER_INC(iter); - continue; - } - - hash = chash_iter_hash(iter); - CHASH_ITER_SET(cur, hash); - while (cur.slot != iter.slot) { - if (chash_iter_is_empty(cur)) { - pr_err("Path to element at %x with hash %x broken at slot %x\n", - iter.slot, hash, cur.slot); - chash_table_dump(table); - return -EINVAL; - } - CHASH_ITER_INC(cur); - } - - CHASH_ITER_INC(iter); - } while (iter.slot); - - return 0; -} -#endif - -static void chash_iter_relocate(struct chash_iter dst, struct chash_iter src) -{ - BUG_ON(src.table == dst.table && src.slot == dst.slot); - BUG_ON(src.table->key_size != dst.table->key_size); - BUG_ON(src.table->value_size != dst.table->value_size); - - if (dst.table->key_size == 4) - dst.table->keys32[dst.slot] = src.table->keys32[src.slot]; - else - dst.table->keys64[dst.slot] = src.table->keys64[src.slot]; - - if (dst.table->value_size) - memcpy(chash_iter_value(dst), chash_iter_value(src), - dst.table->value_size); - - chash_iter_set_valid(dst); - chash_iter_set_invalid(src); - -#ifdef CONFIG_CHASH_STATS - if (src.table == dst.table) { - dst.table->relocs++; - dst.table->reloc_dist += - CHASH_SUB(dst.table, src.slot, dst.slot); - } -#endif -} - -/** - * __chash_table_find - Helper for looking up a hash table entry - * @iter: Pointer to hash table iterator - * @key: Key of the entry to find - * @for_removal: set to true if the element will be removed soon - * - * Searches for an entry in the hash table with a given key. iter must - * be initialized by the caller to point to the home position of the - * hypothetical entry, i.e. it must be initialized with the hash table - * and the key's hash as the initial slot for the search. - * - * This function also does some local clean-up to speed up future - * look-ups by relocating entries to better slots and removing - * tombstones that are no longer needed. - * - * If @for_removal is true, the function avoids relocating the entry - * that is being returned. - * - * Returns 0 if the search is successful. In this case iter is updated - * to point to the found entry. Otherwise %-EINVAL is returned and the - * iter is updated to point to the first available slot for the given - * key. If the table is full, the slot is set to -1. - */ -static int chash_table_find(struct chash_iter *iter, u64 key, - bool for_removal) -{ -#ifdef CONFIG_CHASH_STATS - u64 ts1 = local_clock(); -#endif - u32 hash = iter->slot; - struct chash_iter first_redundant = CHASH_ITER_INIT(iter->table, -1); - int first_avail = (for_removal ? -2 : -1); - - while (!chash_iter_is_valid(*iter) || chash_iter_key(*iter) != key) { - if (chash_iter_is_empty(*iter)) { - /* Found an empty slot, which ends the - * search. Clean up any preceding tombstones - * that are no longer needed because they lead - * to no-where - */ - if ((int)first_redundant.slot < 0) - goto not_found; - while (first_redundant.slot != iter->slot) { - if (!chash_iter_is_valid(first_redundant)) - chash_iter_set_empty(first_redundant); - CHASH_ITER_INC(first_redundant); - } -#ifdef CHASH_DEBUG - chash_table_check(iter->table); -#endif - goto not_found; - } else if (!chash_iter_is_valid(*iter)) { - /* Found a tombstone. Remember it as candidate - * for relocating the entry we're looking for - * or for adding a new entry with the given key - */ - if (first_avail == -1) - first_avail = iter->slot; - /* Or mark it as the start of a series of - * potentially redundant tombstones - */ - else if (first_redundant.slot == -1) - CHASH_ITER_SET(first_redundant, iter->slot); - } else if (first_redundant.slot >= 0) { - /* Found a valid, occupied slot with a - * preceding series of tombstones. Relocate it - * to a better position that no longer depends - * on those tombstones - */ - u32 cur_hash = chash_iter_hash(*iter); - - if (!CHASH_IN_RANGE(iter->table, cur_hash, - first_redundant.slot + 1, - iter->slot)) { - /* This entry has a hash at or before - * the first tombstone we found. We - * can relocate it to that tombstone - * and advance to the next tombstone - */ - chash_iter_relocate(first_redundant, *iter); - do { - CHASH_ITER_INC(first_redundant); - } while (chash_iter_is_valid(first_redundant)); - } else if (cur_hash != iter->slot) { - /* Relocate entry to its home position - * or as close as possible so it no - * longer depends on any preceding - * tombstones - */ - struct chash_iter new_iter = - CHASH_ITER_INIT(iter->table, cur_hash); - - while (new_iter.slot != iter->slot && - chash_iter_is_valid(new_iter)) - CHASH_ITER_INC(new_iter); - - if (new_iter.slot != iter->slot) - chash_iter_relocate(new_iter, *iter); - } - } - - CHASH_ITER_INC(*iter); - if (iter->slot == hash) { - iter->slot = -1; - goto not_found; - } - } - -#ifdef CONFIG_CHASH_STATS - iter->table->hits++; - iter->table->hits_steps += CHASH_SUB(iter->table, iter->slot, hash) + 1; -#endif - - if (first_avail >= 0) { - CHASH_ITER_SET(first_redundant, first_avail); - chash_iter_relocate(first_redundant, *iter); - iter->slot = first_redundant.slot; - iter->mask = first_redundant.mask; - } - -#ifdef CONFIG_CHASH_STATS - iter->table->hits_time_ns += local_clock() - ts1; -#endif - - return 0; - -not_found: -#ifdef CONFIG_CHASH_STATS - iter->table->miss++; - iter->table->miss_steps += (iter->slot < 0) ? - (1 << iter->table->bits) : - CHASH_SUB(iter->table, iter->slot, hash) + 1; -#endif - - if (first_avail >= 0) - CHASH_ITER_SET(*iter, first_avail); - -#ifdef CONFIG_CHASH_STATS - iter->table->miss_time_ns += local_clock() - ts1; -#endif - - return -EINVAL; -} - -int __chash_table_copy_in(struct __chash_table *table, u64 key, - const void *value) -{ - u32 hash = (table->key_size == 4) ? - hash_32(key, table->bits) : hash_64(key, table->bits); - struct chash_iter iter = CHASH_ITER_INIT(table, hash); - int r = chash_table_find(&iter, key, false); - - /* Found an existing entry */ - if (!r) { - if (value && table->value_size) - memcpy(chash_iter_value(iter), value, - table->value_size); - return 1; - } - - /* Is there a place to add a new entry? */ - if (iter.slot < 0) { - pr_err("Hash table overflow\n"); - return -ENOMEM; - } - - chash_iter_set_valid(iter); - - if (table->key_size == 4) - table->keys32[iter.slot] = key; - else - table->keys64[iter.slot] = key; - if (value && table->value_size) - memcpy(chash_iter_value(iter), value, table->value_size); - - return 0; -} -EXPORT_SYMBOL(__chash_table_copy_in); - -int __chash_table_copy_out(struct __chash_table *table, u64 key, - void *value, bool remove) -{ - u32 hash = (table->key_size == 4) ? - hash_32(key, table->bits) : hash_64(key, table->bits); - struct chash_iter iter = CHASH_ITER_INIT(table, hash); - int r = chash_table_find(&iter, key, remove); - - if (r < 0) - return r; - - if (value && table->value_size) - memcpy(value, chash_iter_value(iter), table->value_size); - - if (remove) - chash_iter_set_invalid(iter); - - return iter.slot; -} -EXPORT_SYMBOL(__chash_table_copy_out); - -#ifdef CONFIG_CHASH_SELFTEST -/** - * chash_self_test - Run a self-test of the hash table implementation - * @bits: Table size will be 2^bits entries - * @key_size: Size of hash keys in bytes, 4 or 8 - * @min_fill: Minimum fill level during the test - * @max_fill: Maximum fill level during the test - * @iterations: Number of test iterations - * - * The test adds and removes entries from a hash table, cycling the - * fill level between min_fill and max_fill entries. Also tests lookup - * and value retrieval. - */ -static int __init chash_self_test(u8 bits, u8 key_size, - int min_fill, int max_fill, - u64 iterations) -{ - struct chash_table table; - int ret; - u64 add_count, rmv_count; - u64 value; - - if (key_size == 4 && iterations > 0xffffffff) - return -EINVAL; - if (min_fill >= max_fill) - return -EINVAL; - - ret = chash_table_alloc(&table, bits, key_size, sizeof(u64), - GFP_KERNEL); - if (ret) { - pr_err("chash_table_alloc failed: %d\n", ret); - return ret; - } - - for (add_count = 0, rmv_count = 0; add_count < iterations; - add_count++) { - /* When we hit the max_fill level, remove entries down - * to min_fill - */ - if (add_count - rmv_count == max_fill) { - u64 find_count = rmv_count; - - /* First try to find all entries that we're - * about to remove, confirm their value, test - * writing them back a second time. - */ - for (; add_count - find_count > min_fill; - find_count++) { - ret = chash_table_copy_out(&table, find_count, - &value); - if (ret < 0) { - pr_err("chash_table_copy_out failed: %d\n", - ret); - goto out; - } - if (value != ~find_count) { - pr_err("Wrong value retrieved for key 0x%llx, expected 0x%llx got 0x%llx\n", - find_count, ~find_count, value); -#ifdef CHASH_DEBUG - chash_table_dump(&table.table); -#endif - ret = -EFAULT; - goto out; - } - ret = chash_table_copy_in(&table, find_count, - &value); - if (ret != 1) { - pr_err("copy_in second time returned %d, expected 1\n", - ret); - ret = -EFAULT; - goto out; - } - } - /* Remove them until we hit min_fill level */ - for (; add_count - rmv_count > min_fill; rmv_count++) { - ret = chash_table_remove(&table, rmv_count, - NULL); - if (ret < 0) { - pr_err("chash_table_remove failed: %d\n", - ret); - goto out; - } - } - } - - /* Add a new value */ - value = ~add_count; - ret = chash_table_copy_in(&table, add_count, &value); - if (ret != 0) { - pr_err("copy_in first time returned %d, expected 0\n", - ret); - ret = -EFAULT; - goto out; - } - } - - chash_table_dump_stats(&table); - chash_table_reset_stats(&table); - -out: - chash_table_free(&table); - return ret; -} - -static unsigned int chash_test_bits = 10; -MODULE_PARM_DESC(test_bits, - "Selftest number of hash bits ([4..20], default=10)"); -module_param_named(test_bits, chash_test_bits, uint, 0444); - -static unsigned int chash_test_keysize = 8; -MODULE_PARM_DESC(test_keysize, "Selftest keysize (4 or 8, default=8)"); -module_param_named(test_keysize, chash_test_keysize, uint, 0444); - -static unsigned int chash_test_minfill; -MODULE_PARM_DESC(test_minfill, "Selftest minimum #entries (default=50%)"); -module_param_named(test_minfill, chash_test_minfill, uint, 0444); - -static unsigned int chash_test_maxfill; -MODULE_PARM_DESC(test_maxfill, "Selftest maximum #entries (default=80%)"); -module_param_named(test_maxfill, chash_test_maxfill, uint, 0444); - -static unsigned long chash_test_iters; -MODULE_PARM_DESC(test_iters, "Selftest iterations (default=1000 x #entries)"); -module_param_named(test_iters, chash_test_iters, ulong, 0444); - -static int __init chash_init(void) -{ - int ret; - u64 ts1_ns; - - /* Skip self test on user errors */ - if (chash_test_bits < 4 || chash_test_bits > 20) { - pr_err("chash: test_bits out of range [4..20].\n"); - return 0; - } - if (chash_test_keysize != 4 && chash_test_keysize != 8) { - pr_err("chash: test_keysize invalid. Must be 4 or 8.\n"); - return 0; - } - - if (!chash_test_minfill) - chash_test_minfill = (1 << chash_test_bits) / 2; - if (!chash_test_maxfill) - chash_test_maxfill = (1 << chash_test_bits) * 4 / 5; - if (!chash_test_iters) - chash_test_iters = (1 << chash_test_bits) * 1000; - - if (chash_test_minfill >= (1 << chash_test_bits)) { - pr_err("chash: test_minfill too big. Must be < table size.\n"); - return 0; - } - if (chash_test_maxfill >= (1 << chash_test_bits)) { - pr_err("chash: test_maxfill too big. Must be < table size.\n"); - return 0; - } - if (chash_test_minfill >= chash_test_maxfill) { - pr_err("chash: test_minfill must be < test_maxfill.\n"); - return 0; - } - if (chash_test_keysize == 4 && chash_test_iters > 0xffffffff) { - pr_err("chash: test_iters must be < 4G for 4 byte keys.\n"); - return 0; - } - - ts1_ns = local_clock(); - ret = chash_self_test(chash_test_bits, chash_test_keysize, - chash_test_minfill, chash_test_maxfill, - chash_test_iters); - if (!ret) { - u64 ts_delta_us = local_clock() - ts1_ns; - u64 iters_per_second = (u64)chash_test_iters * 1000000; - - do_div(ts_delta_us, 1000); - do_div(iters_per_second, ts_delta_us); - pr_info("chash: self test took %llu us, %llu iterations/s\n", - ts_delta_us, iters_per_second); - } else { - pr_err("chash: self test failed: %d\n", ret); - } - - return ret; -} - -module_init(chash_init); - -#endif /* CONFIG_CHASH_SELFTEST */ - -MODULE_DESCRIPTION("Closed hash table"); -MODULE_LICENSE("GPL and additional rights"); diff --git a/drivers/gpu/drm/amd/powerplay/Makefile b/drivers/gpu/drm/amd/powerplay/Makefile index 231785a9e24c..ec87b3430d12 100644 --- a/drivers/gpu/drm/amd/powerplay/Makefile +++ b/drivers/gpu/drm/amd/powerplay/Makefile @@ -35,7 +35,7 @@ AMD_POWERPLAY = $(addsuffix /Makefile,$(addprefix $(FULL_AMD_PATH)/powerplay/,$( include $(AMD_POWERPLAY) -POWER_MGR = amd_powerplay.o +POWER_MGR = amd_powerplay.o amdgpu_smu.o smu_v11_0.o vega20_ppt.o AMD_PP_POWER = $(addprefix $(AMD_PP_PATH)/,$(POWER_MGR)) diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c index 3f73f7cd18b9..bea1587d352d 100644 --- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c @@ -53,7 +53,7 @@ static int amd_powerplay_create(struct amdgpu_device *adev) mutex_init(&hwmgr->smu_lock); hwmgr->chip_family = adev->family; hwmgr->chip_id = adev->asic_type; - hwmgr->feature_mask = adev->powerplay.pp_feature; + hwmgr->feature_mask = adev->pm.pp_feature; hwmgr->display_config = &adev->pm.pm_display_cfg; adev->powerplay.pp_handle = hwmgr; adev->powerplay.pp_funcs = &pp_dpm_funcs; @@ -1304,7 +1304,7 @@ static int pp_notify_smu_enable_pwe(void *handle) if (hwmgr->hwmgr_func->smus_notify_pwe == NULL) { pr_info_ratelimited("%s was not implemented.\n", __func__); - return -EINVAL;; + return -EINVAL; } mutex_lock(&hwmgr->smu_lock); @@ -1341,7 +1341,7 @@ static int pp_set_min_deep_sleep_dcefclk(void *handle, uint32_t clock) if (hwmgr->hwmgr_func->set_min_deep_sleep_dcefclk == NULL) { pr_debug("%s was not implemented.\n", __func__); - return -EINVAL;; + return -EINVAL; } mutex_lock(&hwmgr->smu_lock); @@ -1360,7 +1360,7 @@ static int pp_set_hard_min_dcefclk_by_freq(void *handle, uint32_t clock) if (hwmgr->hwmgr_func->set_hard_min_dcefclk_by_freq == NULL) { pr_debug("%s was not implemented.\n", __func__); - return -EINVAL;; + return -EINVAL; } mutex_lock(&hwmgr->smu_lock); @@ -1379,7 +1379,7 @@ static int pp_set_hard_min_fclk_by_freq(void *handle, uint32_t clock) if (hwmgr->hwmgr_func->set_hard_min_fclk_by_freq == NULL) { pr_debug("%s was not implemented.\n", __func__); - return -EINVAL;; + return -EINVAL; } mutex_lock(&hwmgr->smu_lock); diff --git a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c new file mode 100644 index 000000000000..c058c784180e --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c @@ -0,0 +1,1253 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#include "pp_debug.h" +#include +#include +#include "amdgpu.h" +#include "amdgpu_smu.h" +#include "soc15_common.h" +#include "smu_v11_0.h" +#include "atom.h" +#include "amd_pcie.h" + +int smu_dpm_set_power_gate(struct smu_context *smu, uint32_t block_type, + bool gate) +{ + int ret = 0; + + switch (block_type) { + case AMD_IP_BLOCK_TYPE_UVD: + ret = smu_dpm_set_uvd_enable(smu, gate); + break; + case AMD_IP_BLOCK_TYPE_VCE: + ret = smu_dpm_set_vce_enable(smu, gate); + break; + default: + break; + } + + return ret; +} + +enum amd_pm_state_type smu_get_current_power_state(struct smu_context *smu) +{ + /* not support power state */ + return POWER_STATE_TYPE_DEFAULT; +} + +int smu_get_power_num_states(struct smu_context *smu, + struct pp_states_info *state_info) +{ + if (!state_info) + return -EINVAL; + + /* not support power state */ + memset(state_info, 0, sizeof(struct pp_states_info)); + state_info->nums = 0; + + return 0; +} + +int smu_common_read_sensor(struct smu_context *smu, enum amd_pp_sensors sensor, + void *data, uint32_t *size) +{ + int ret = 0; + + switch (sensor) { + case AMDGPU_PP_SENSOR_STABLE_PSTATE_SCLK: + *((uint32_t *)data) = smu->pstate_sclk; + *size = 4; + break; + case AMDGPU_PP_SENSOR_STABLE_PSTATE_MCLK: + *((uint32_t *)data) = smu->pstate_mclk; + *size = 4; + break; + case AMDGPU_PP_SENSOR_ENABLED_SMC_FEATURES_MASK: + ret = smu_feature_get_enabled_mask(smu, (uint32_t *)data, 2); + *size = 8; + break; + default: + ret = -EINVAL; + break; + } + + if (ret) + *size = 0; + + return ret; +} + +int smu_update_table_with_arg(struct smu_context *smu, uint16_t table_id, uint16_t exarg, + void *table_data, bool drv2smu) +{ + struct smu_table_context *smu_table = &smu->smu_table; + struct smu_table *table = NULL; + int ret = 0; + uint32_t table_index; + + if (!table_data || table_id >= smu_table->table_count) + return -EINVAL; + + table_index = (exarg << 16) | table_id; + + table = &smu_table->tables[table_id]; + + if (drv2smu) + memcpy(table->cpu_addr, table_data, table->size); + + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetDriverDramAddrHigh, + upper_32_bits(table->mc_address)); + if (ret) + return ret; + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetDriverDramAddrLow, + lower_32_bits(table->mc_address)); + if (ret) + return ret; + ret = smu_send_smc_msg_with_param(smu, drv2smu ? + SMU_MSG_TransferTableDram2Smu : + SMU_MSG_TransferTableSmu2Dram, + table_index); + if (ret) + return ret; + + if (!drv2smu) + memcpy(table_data, table->cpu_addr, table->size); + + return ret; +} + +bool is_support_sw_smu(struct amdgpu_device *adev) +{ + if (amdgpu_dpm != 1) + return false; + + if (adev->asic_type >= CHIP_VEGA20 && adev->asic_type != CHIP_RAVEN) + return true; + + return false; +} + +int smu_sys_get_pp_table(struct smu_context *smu, void **table) +{ + struct smu_table_context *smu_table = &smu->smu_table; + + if (!smu_table->power_play_table && !smu_table->hardcode_pptable) + return -EINVAL; + + if (smu_table->hardcode_pptable) + *table = smu_table->hardcode_pptable; + else + *table = smu_table->power_play_table; + + return smu_table->power_play_table_size; +} + +int smu_sys_set_pp_table(struct smu_context *smu, void *buf, size_t size) +{ + struct smu_table_context *smu_table = &smu->smu_table; + ATOM_COMMON_TABLE_HEADER *header = (ATOM_COMMON_TABLE_HEADER *)buf; + int ret = 0; + + if (header->usStructureSize != size) { + pr_err("pp table size not matched !\n"); + return -EIO; + } + + mutex_lock(&smu->mutex); + if (!smu_table->hardcode_pptable) + smu_table->hardcode_pptable = kzalloc(size, GFP_KERNEL); + if (!smu_table->hardcode_pptable) { + ret = -ENOMEM; + goto failed; + } + + memcpy(smu_table->hardcode_pptable, buf, size); + smu_table->power_play_table = smu_table->hardcode_pptable; + smu_table->power_play_table_size = size; + mutex_unlock(&smu->mutex); + + ret = smu_reset(smu); + if (ret) + pr_info("smu reset failed, ret = %d\n", ret); + + return ret; + +failed: + mutex_unlock(&smu->mutex); + return ret; +} + +int smu_feature_init_dpm(struct smu_context *smu) +{ + struct smu_feature *feature = &smu->smu_feature; + int ret = 0; + uint32_t unallowed_feature_mask[SMU_FEATURE_MAX/32]; + + mutex_lock(&feature->mutex); + bitmap_fill(feature->allowed, SMU_FEATURE_MAX); + mutex_unlock(&feature->mutex); + + ret = smu_get_unallowed_feature_mask(smu, unallowed_feature_mask, + SMU_FEATURE_MAX/32); + if (ret) + return ret; + + mutex_lock(&feature->mutex); + bitmap_andnot(feature->allowed, feature->allowed, + (unsigned long *)unallowed_feature_mask, + feature->feature_num); + mutex_unlock(&feature->mutex); + + return ret; +} + +int smu_feature_is_enabled(struct smu_context *smu, int feature_id) +{ + struct smu_feature *feature = &smu->smu_feature; + int ret = 0; + + WARN_ON(feature_id > feature->feature_num); + + mutex_lock(&feature->mutex); + ret = test_bit(feature_id, feature->enabled); + mutex_unlock(&feature->mutex); + + return ret; +} + +int smu_feature_set_enabled(struct smu_context *smu, int feature_id, bool enable) +{ + struct smu_feature *feature = &smu->smu_feature; + int ret = 0; + + WARN_ON(feature_id > feature->feature_num); + + mutex_lock(&feature->mutex); + ret = smu_feature_update_enable_state(smu, feature_id, enable); + if (ret) + goto failed; + + if (enable) + test_and_set_bit(feature_id, feature->enabled); + else + test_and_clear_bit(feature_id, feature->enabled); + +failed: + mutex_unlock(&feature->mutex); + + return ret; +} + +int smu_feature_is_supported(struct smu_context *smu, int feature_id) +{ + struct smu_feature *feature = &smu->smu_feature; + int ret = 0; + + WARN_ON(feature_id > feature->feature_num); + + mutex_lock(&feature->mutex); + ret = test_bit(feature_id, feature->supported); + mutex_unlock(&feature->mutex); + + return ret; +} + +int smu_feature_set_supported(struct smu_context *smu, int feature_id, + bool enable) +{ + struct smu_feature *feature = &smu->smu_feature; + int ret = 0; + + WARN_ON(feature_id > feature->feature_num); + + mutex_unlock(&feature->mutex); + if (enable) + test_and_set_bit(feature_id, feature->supported); + else + test_and_clear_bit(feature_id, feature->supported); + mutex_unlock(&feature->mutex); + + return ret; +} + +static int smu_set_funcs(struct amdgpu_device *adev) +{ + struct smu_context *smu = &adev->smu; + + switch (adev->asic_type) { + case CHIP_VEGA20: + adev->pm.pp_feature &= ~PP_GFXOFF_MASK; + if (adev->pm.pp_feature & PP_OVERDRIVE_MASK) + smu->od_enabled = true; + smu_v11_0_set_smu_funcs(smu); + break; + default: + return -EINVAL; + } + + return 0; +} + +static int smu_early_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct smu_context *smu = &adev->smu; + + smu->adev = adev; + mutex_init(&smu->mutex); + + return smu_set_funcs(adev); +} + +static int smu_late_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct smu_context *smu = &adev->smu; + mutex_lock(&smu->mutex); + smu_handle_task(&adev->smu, + smu->smu_dpm.dpm_level, + AMD_PP_TASK_COMPLETE_INIT); + mutex_unlock(&smu->mutex); + + return 0; +} + +int smu_get_atom_data_table(struct smu_context *smu, uint32_t table, + uint16_t *size, uint8_t *frev, uint8_t *crev, + uint8_t **addr) +{ + struct amdgpu_device *adev = smu->adev; + uint16_t data_start; + + if (!amdgpu_atom_parse_data_header(adev->mode_info.atom_context, table, + size, frev, crev, &data_start)) + return -EINVAL; + + *addr = (uint8_t *)adev->mode_info.atom_context->bios + data_start; + + return 0; +} + +static int smu_initialize_pptable(struct smu_context *smu) +{ + /* TODO */ + return 0; +} + +static int smu_smc_table_sw_init(struct smu_context *smu) +{ + int ret; + + ret = smu_initialize_pptable(smu); + if (ret) { + pr_err("Failed to init smu_initialize_pptable!\n"); + return ret; + } + + /** + * Create smu_table structure, and init smc tables such as + * TABLE_PPTABLE, TABLE_WATERMARKS, TABLE_SMU_METRICS, and etc. + */ + ret = smu_init_smc_tables(smu); + if (ret) { + pr_err("Failed to init smc tables!\n"); + return ret; + } + + /** + * Create smu_power_context structure, and allocate smu_dpm_context and + * context size to fill the smu_power_context data. + */ + ret = smu_init_power(smu); + if (ret) { + pr_err("Failed to init smu_init_power!\n"); + return ret; + } + + return 0; +} + +static int smu_smc_table_sw_fini(struct smu_context *smu) +{ + int ret; + + ret = smu_fini_smc_tables(smu); + if (ret) { + pr_err("Failed to smu_fini_smc_tables!\n"); + return ret; + } + + return 0; +} + +static int smu_sw_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct smu_context *smu = &adev->smu; + int ret; + + if (!is_support_sw_smu(adev)) + return -EINVAL; + + smu->pool_size = adev->pm.smu_prv_buffer_size; + smu->smu_feature.feature_num = SMU_FEATURE_MAX; + mutex_init(&smu->smu_feature.mutex); + bitmap_zero(smu->smu_feature.supported, SMU_FEATURE_MAX); + bitmap_zero(smu->smu_feature.enabled, SMU_FEATURE_MAX); + bitmap_zero(smu->smu_feature.allowed, SMU_FEATURE_MAX); + smu->watermarks_bitmap = 0; + smu->power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; + smu->default_power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; + + smu->workload_mask = 1 << smu->workload_prority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT]; + smu->workload_prority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT] = 0; + smu->workload_prority[PP_SMC_POWER_PROFILE_FULLSCREEN3D] = 1; + smu->workload_prority[PP_SMC_POWER_PROFILE_POWERSAVING] = 2; + smu->workload_prority[PP_SMC_POWER_PROFILE_VIDEO] = 3; + smu->workload_prority[PP_SMC_POWER_PROFILE_VR] = 4; + smu->workload_prority[PP_SMC_POWER_PROFILE_COMPUTE] = 5; + smu->workload_prority[PP_SMC_POWER_PROFILE_CUSTOM] = 6; + + smu->workload_setting[0] = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; + smu->workload_setting[1] = PP_SMC_POWER_PROFILE_FULLSCREEN3D; + smu->workload_setting[2] = PP_SMC_POWER_PROFILE_POWERSAVING; + smu->workload_setting[3] = PP_SMC_POWER_PROFILE_VIDEO; + smu->workload_setting[4] = PP_SMC_POWER_PROFILE_VR; + smu->workload_setting[5] = PP_SMC_POWER_PROFILE_COMPUTE; + smu->workload_setting[6] = PP_SMC_POWER_PROFILE_CUSTOM; + smu->display_config = &adev->pm.pm_display_cfg; + + smu->smu_dpm.dpm_level = AMD_DPM_FORCED_LEVEL_AUTO; + smu->smu_dpm.requested_dpm_level = AMD_DPM_FORCED_LEVEL_AUTO; + ret = smu_init_microcode(smu); + if (ret) { + pr_err("Failed to load smu firmware!\n"); + return ret; + } + + ret = smu_smc_table_sw_init(smu); + if (ret) { + pr_err("Failed to sw init smc table!\n"); + return ret; + } + + return 0; +} + +static int smu_sw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct smu_context *smu = &adev->smu; + int ret; + + if (!is_support_sw_smu(adev)) + return -EINVAL; + + ret = smu_smc_table_sw_fini(smu); + if (ret) { + pr_err("Failed to sw fini smc table!\n"); + return ret; + } + + ret = smu_fini_power(smu); + if (ret) { + pr_err("Failed to init smu_fini_power!\n"); + return ret; + } + + return 0; +} + +static int smu_init_fb_allocations(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + struct smu_table_context *smu_table = &smu->smu_table; + struct smu_table *tables = smu_table->tables; + uint32_t table_count = smu_table->table_count; + uint32_t i = 0; + int32_t ret = 0; + + if (table_count <= 0) + return -EINVAL; + + for (i = 0 ; i < table_count; i++) { + if (tables[i].size == 0) + continue; + ret = amdgpu_bo_create_kernel(adev, + tables[i].size, + tables[i].align, + tables[i].domain, + &tables[i].bo, + &tables[i].mc_address, + &tables[i].cpu_addr); + if (ret) + goto failed; + } + + return 0; +failed: + for (; i > 0; i--) { + if (tables[i].size == 0) + continue; + amdgpu_bo_free_kernel(&tables[i].bo, + &tables[i].mc_address, + &tables[i].cpu_addr); + + } + return ret; +} + +static int smu_fini_fb_allocations(struct smu_context *smu) +{ + struct smu_table_context *smu_table = &smu->smu_table; + struct smu_table *tables = smu_table->tables; + uint32_t table_count = smu_table->table_count; + uint32_t i = 0; + + if (table_count == 0 || tables == NULL) + return 0; + + for (i = 0 ; i < table_count; i++) { + if (tables[i].size == 0) + continue; + amdgpu_bo_free_kernel(&tables[i].bo, + &tables[i].mc_address, + &tables[i].cpu_addr); + } + + return 0; +} + +static int smu_override_pcie_parameters(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + uint32_t pcie_gen = 0, pcie_width = 0, smu_pcie_arg; + int ret; + + if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4) + pcie_gen = 3; + else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) + pcie_gen = 2; + else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2) + pcie_gen = 1; + else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1) + pcie_gen = 0; + + /* Bit 31:16: LCLK DPM level. 0 is DPM0, and 1 is DPM1 + * Bit 15:8: PCIE GEN, 0 to 3 corresponds to GEN1 to GEN4 + * Bit 7:0: PCIE lane width, 1 to 7 corresponds is x1 to x32 + */ + if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X16) + pcie_width = 6; + else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X12) + pcie_width = 5; + else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X8) + pcie_width = 4; + else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X4) + pcie_width = 3; + else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X2) + pcie_width = 2; + else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X1) + pcie_width = 1; + + smu_pcie_arg = (1 << 16) | (pcie_gen << 8) | pcie_width; + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_OverridePcieParameters, + smu_pcie_arg); + if (ret) + pr_err("[%s] Attempt to override pcie params failed!\n", __func__); + return ret; +} + +static int smu_smc_table_hw_init(struct smu_context *smu, + bool initialize) +{ + struct amdgpu_device *adev = smu->adev; + int ret; + + if (smu_is_dpm_running(smu) && adev->in_suspend) { + pr_info("dpm has been enabled\n"); + return 0; + } + + ret = smu_init_display(smu); + if (ret) + return ret; + + if (initialize) { + ret = smu_read_pptable_from_vbios(smu); + if (ret) + return ret; + + /* get boot_values from vbios to set revision, gfxclk, and etc. */ + ret = smu_get_vbios_bootup_values(smu); + if (ret) + return ret; + + ret = smu_get_clk_info_from_vbios(smu); + if (ret) + return ret; + + /* + * check if the format_revision in vbios is up to pptable header + * version, and the structure size is not 0. + */ + ret = smu_get_clk_info_from_vbios(smu); + if (ret) + return ret; + + ret = smu_check_pptable(smu); + if (ret) + return ret; + + /* + * allocate vram bos to store smc table contents. + */ + ret = smu_init_fb_allocations(smu); + if (ret) + return ret; + + /* + * Parse pptable format and fill PPTable_t smc_pptable to + * smu_table_context structure. And read the smc_dpm_table from vbios, + * then fill it into smc_pptable. + */ + ret = smu_parse_pptable(smu); + if (ret) + return ret; + + /* + * Send msg GetDriverIfVersion to check if the return value is equal + * with DRIVER_IF_VERSION of smc header. + */ + ret = smu_check_fw_version(smu); + if (ret) + return ret; + } + + /* + * Copy pptable bo in the vram to smc with SMU MSGs such as + * SetDriverDramAddr and TransferTableDram2Smu. + */ + ret = smu_write_pptable(smu); + if (ret) + return ret; + + /* issue RunAfllBtc msg */ + ret = smu_run_afll_btc(smu); + if (ret) + return ret; + + ret = smu_feature_set_allowed_mask(smu); + if (ret) + return ret; + + ret = smu_system_features_control(smu, true); + if (ret) + return ret; + + ret = smu_override_pcie_parameters(smu); + if (ret) + return ret; + + ret = smu_notify_display_change(smu); + if (ret) + return ret; + + /* + * Set min deep sleep dce fclk with bootup value from vbios via + * SetMinDeepSleepDcefclk MSG. + */ + ret = smu_set_min_dcef_deep_sleep(smu); + if (ret) + return ret; + + /* + * Set initialized values (get from vbios) to dpm tables context such as + * gfxclk, memclk, dcefclk, and etc. And enable the DPM feature for each + * type of clks. + */ + if (initialize) { + ret = smu_populate_smc_pptable(smu); + if (ret) + return ret; + + ret = smu_init_max_sustainable_clocks(smu); + if (ret) + return ret; + } + + ret = smu_set_od8_default_settings(smu, initialize); + if (ret) + return ret; + + if (initialize) { + ret = smu_populate_umd_state_clk(smu); + if (ret) + return ret; + + ret = smu_get_power_limit(smu, &smu->default_power_limit, false); + if (ret) + return ret; + } + + /* + * Set PMSTATUSLOG table bo address with SetToolsDramAddr MSG for tools. + */ + ret = smu_set_tool_table_location(smu); + + return ret; +} + +/** + * smu_alloc_memory_pool - allocate memory pool in the system memory + * + * @smu: amdgpu_device pointer + * + * This memory pool will be used for SMC use and msg SetSystemVirtualDramAddr + * and DramLogSetDramAddr can notify it changed. + * + * Returns 0 on success, error on failure. + */ +static int smu_alloc_memory_pool(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + struct smu_table_context *smu_table = &smu->smu_table; + struct smu_table *memory_pool = &smu_table->memory_pool; + uint64_t pool_size = smu->pool_size; + int ret = 0; + + if (pool_size == SMU_MEMORY_POOL_SIZE_ZERO) + return ret; + + memory_pool->size = pool_size; + memory_pool->align = PAGE_SIZE; + memory_pool->domain = AMDGPU_GEM_DOMAIN_GTT; + + switch (pool_size) { + case SMU_MEMORY_POOL_SIZE_256_MB: + case SMU_MEMORY_POOL_SIZE_512_MB: + case SMU_MEMORY_POOL_SIZE_1_GB: + case SMU_MEMORY_POOL_SIZE_2_GB: + ret = amdgpu_bo_create_kernel(adev, + memory_pool->size, + memory_pool->align, + memory_pool->domain, + &memory_pool->bo, + &memory_pool->mc_address, + &memory_pool->cpu_addr); + break; + default: + break; + } + + return ret; +} + +static int smu_free_memory_pool(struct smu_context *smu) +{ + struct smu_table_context *smu_table = &smu->smu_table; + struct smu_table *memory_pool = &smu_table->memory_pool; + int ret = 0; + + if (memory_pool->size == SMU_MEMORY_POOL_SIZE_ZERO) + return ret; + + amdgpu_bo_free_kernel(&memory_pool->bo, + &memory_pool->mc_address, + &memory_pool->cpu_addr); + + memset(memory_pool, 0, sizeof(struct smu_table)); + + return ret; +} + +static int smu_hw_init(void *handle) +{ + int ret; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct smu_context *smu = &adev->smu; + + if (!is_support_sw_smu(adev)) + return -EINVAL; + + if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) { + ret = smu_load_microcode(smu); + if (ret) + return ret; + } + + ret = smu_check_fw_status(smu); + if (ret) { + pr_err("SMC firmware status is not correct\n"); + return ret; + } + + mutex_lock(&smu->mutex); + + ret = smu_feature_init_dpm(smu); + if (ret) + goto failed; + + ret = smu_smc_table_hw_init(smu, true); + if (ret) + goto failed; + + ret = smu_alloc_memory_pool(smu); + if (ret) + goto failed; + + /* + * Use msg SetSystemVirtualDramAddr and DramLogSetDramAddr can notify + * pool location. + */ + ret = smu_notify_memory_pool_location(smu); + if (ret) + goto failed; + + ret = smu_start_thermal_control(smu); + if (ret) + goto failed; + + mutex_unlock(&smu->mutex); + + adev->pm.dpm_enabled = true; + + pr_info("SMU is initialized successfully!\n"); + + return 0; + +failed: + mutex_unlock(&smu->mutex); + return ret; +} + +static int smu_hw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct smu_context *smu = &adev->smu; + struct smu_table_context *table_context = &smu->smu_table; + int ret = 0; + + if (!is_support_sw_smu(adev)) + return -EINVAL; + + kfree(table_context->driver_pptable); + table_context->driver_pptable = NULL; + + kfree(table_context->max_sustainable_clocks); + table_context->max_sustainable_clocks = NULL; + + kfree(table_context->od_feature_capabilities); + table_context->od_feature_capabilities = NULL; + + kfree(table_context->od_settings_max); + table_context->od_settings_max = NULL; + + kfree(table_context->od_settings_min); + table_context->od_settings_min = NULL; + + kfree(table_context->overdrive_table); + table_context->overdrive_table = NULL; + + kfree(table_context->od8_settings); + table_context->od8_settings = NULL; + + ret = smu_fini_fb_allocations(smu); + if (ret) + return ret; + + ret = smu_free_memory_pool(smu); + if (ret) + return ret; + + return 0; +} + +int smu_reset(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + int ret = 0; + + ret = smu_hw_fini(adev); + if (ret) + return ret; + + ret = smu_hw_init(adev); + if (ret) + return ret; + + return ret; +} + +static int smu_suspend(void *handle) +{ + int ret; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct smu_context *smu = &adev->smu; + + if (!is_support_sw_smu(adev)) + return -EINVAL; + + ret = smu_system_features_control(smu, false); + if (ret) + return ret; + + smu->watermarks_bitmap &= ~(WATERMARKS_LOADED); + + return 0; +} + +static int smu_resume(void *handle) +{ + int ret; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct smu_context *smu = &adev->smu; + + if (!is_support_sw_smu(adev)) + return -EINVAL; + + pr_info("SMU is resuming...\n"); + + mutex_lock(&smu->mutex); + + ret = smu_smc_table_hw_init(smu, false); + if (ret) + goto failed; + + ret = smu_start_thermal_control(smu); + if (ret) + goto failed; + + mutex_unlock(&smu->mutex); + + pr_info("SMU is resumed successfully!\n"); + + return 0; +failed: + mutex_unlock(&smu->mutex); + return ret; +} + +int smu_display_configuration_change(struct smu_context *smu, + const struct amd_pp_display_configuration *display_config) +{ + int index = 0; + int num_of_active_display = 0; + + if (!is_support_sw_smu(smu->adev)) + return -EINVAL; + + if (!display_config) + return -EINVAL; + + mutex_lock(&smu->mutex); + + smu_set_deep_sleep_dcefclk(smu, + display_config->min_dcef_deep_sleep_set_clk / 100); + + for (index = 0; index < display_config->num_path_including_non_display; index++) { + if (display_config->displays[index].controller_id != 0) + num_of_active_display++; + } + + smu_set_active_display_count(smu, num_of_active_display); + + smu_store_cc6_data(smu, display_config->cpu_pstate_separation_time, + display_config->cpu_cc6_disable, + display_config->cpu_pstate_disable, + display_config->nb_pstate_switch_disable); + + mutex_unlock(&smu->mutex); + + return 0; +} + +static int smu_get_clock_info(struct smu_context *smu, + struct smu_clock_info *clk_info, + enum smu_perf_level_designation designation) +{ + int ret; + struct smu_performance_level level = {0}; + + if (!clk_info) + return -EINVAL; + + ret = smu_get_perf_level(smu, PERF_LEVEL_ACTIVITY, &level); + if (ret) + return -EINVAL; + + clk_info->min_mem_clk = level.memory_clock; + clk_info->min_eng_clk = level.core_clock; + clk_info->min_bus_bandwidth = level.non_local_mem_freq * level.non_local_mem_width; + + ret = smu_get_perf_level(smu, designation, &level); + if (ret) + return -EINVAL; + + clk_info->min_mem_clk = level.memory_clock; + clk_info->min_eng_clk = level.core_clock; + clk_info->min_bus_bandwidth = level.non_local_mem_freq * level.non_local_mem_width; + + return 0; +} + +int smu_get_current_clocks(struct smu_context *smu, + struct amd_pp_clock_info *clocks) +{ + struct amd_pp_simple_clock_info simple_clocks = {0}; + struct smu_clock_info hw_clocks; + int ret = 0; + + if (!is_support_sw_smu(smu->adev)) + return -EINVAL; + + mutex_lock(&smu->mutex); + + smu_get_dal_power_level(smu, &simple_clocks); + + if (smu->support_power_containment) + ret = smu_get_clock_info(smu, &hw_clocks, + PERF_LEVEL_POWER_CONTAINMENT); + else + ret = smu_get_clock_info(smu, &hw_clocks, PERF_LEVEL_ACTIVITY); + + if (ret) { + pr_err("Error in smu_get_clock_info\n"); + goto failed; + } + + clocks->min_engine_clock = hw_clocks.min_eng_clk; + clocks->max_engine_clock = hw_clocks.max_eng_clk; + clocks->min_memory_clock = hw_clocks.min_mem_clk; + clocks->max_memory_clock = hw_clocks.max_mem_clk; + clocks->min_bus_bandwidth = hw_clocks.min_bus_bandwidth; + clocks->max_bus_bandwidth = hw_clocks.max_bus_bandwidth; + clocks->max_engine_clock_in_sr = hw_clocks.max_eng_clk; + clocks->min_engine_clock_in_sr = hw_clocks.min_eng_clk; + + if (simple_clocks.level == 0) + clocks->max_clocks_state = PP_DAL_POWERLEVEL_7; + else + clocks->max_clocks_state = simple_clocks.level; + + if (!smu_get_current_shallow_sleep_clocks(smu, &hw_clocks)) { + clocks->max_engine_clock_in_sr = hw_clocks.max_eng_clk; + clocks->min_engine_clock_in_sr = hw_clocks.min_eng_clk; + } + +failed: + mutex_unlock(&smu->mutex); + return ret; +} + +static int smu_set_clockgating_state(void *handle, + enum amd_clockgating_state state) +{ + return 0; +} + +static int smu_set_powergating_state(void *handle, + enum amd_powergating_state state) +{ + return 0; +} + +static int smu_enable_umd_pstate(void *handle, + enum amd_dpm_forced_level *level) +{ + uint32_t profile_mode_mask = AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD | + AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK | + AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK | + AMD_DPM_FORCED_LEVEL_PROFILE_PEAK; + + struct smu_context *smu = (struct smu_context*)(handle); + struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm); + if (!smu_dpm_ctx->dpm_context) + return -EINVAL; + + if (!(smu_dpm_ctx->dpm_level & profile_mode_mask)) { + /* enter umd pstate, save current level, disable gfx cg*/ + if (*level & profile_mode_mask) { + smu_dpm_ctx->saved_dpm_level = smu_dpm_ctx->dpm_level; + smu_dpm_ctx->enable_umd_pstate = true; + amdgpu_device_ip_set_clockgating_state(smu->adev, + AMD_IP_BLOCK_TYPE_GFX, + AMD_CG_STATE_UNGATE); + amdgpu_device_ip_set_powergating_state(smu->adev, + AMD_IP_BLOCK_TYPE_GFX, + AMD_PG_STATE_UNGATE); + } + } else { + /* exit umd pstate, restore level, enable gfx cg*/ + if (!(*level & profile_mode_mask)) { + if (*level == AMD_DPM_FORCED_LEVEL_PROFILE_EXIT) + *level = smu_dpm_ctx->saved_dpm_level; + smu_dpm_ctx->enable_umd_pstate = false; + amdgpu_device_ip_set_clockgating_state(smu->adev, + AMD_IP_BLOCK_TYPE_GFX, + AMD_CG_STATE_GATE); + amdgpu_device_ip_set_powergating_state(smu->adev, + AMD_IP_BLOCK_TYPE_GFX, + AMD_PG_STATE_GATE); + } + } + + return 0; +} + +int smu_adjust_power_state_dynamic(struct smu_context *smu, + enum amd_dpm_forced_level level, + bool skip_display_settings) +{ + int ret = 0; + int index = 0; + uint32_t sclk_mask, mclk_mask, soc_mask; + long workload; + struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm); + + if (!skip_display_settings) { + ret = smu_display_config_changed(smu); + if (ret) { + pr_err("Failed to change display config!"); + return ret; + } + } + + ret = smu_apply_clocks_adjust_rules(smu); + if (ret) { + pr_err("Failed to apply clocks adjust rules!"); + return ret; + } + + if (!skip_display_settings) { + ret = smu_notify_smc_dispaly_config(smu); + if (ret) { + pr_err("Failed to notify smc display config!"); + return ret; + } + } + + if (smu_dpm_ctx->dpm_level != level) { + switch (level) { + case AMD_DPM_FORCED_LEVEL_HIGH: + ret = smu_force_dpm_limit_value(smu, true); + break; + case AMD_DPM_FORCED_LEVEL_LOW: + ret = smu_force_dpm_limit_value(smu, false); + break; + + case AMD_DPM_FORCED_LEVEL_AUTO: + ret = smu_unforce_dpm_levels(smu); + break; + + case AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD: + case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK: + case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK: + case AMD_DPM_FORCED_LEVEL_PROFILE_PEAK: + ret = smu_get_profiling_clk_mask(smu, level, + &sclk_mask, + &mclk_mask, + &soc_mask); + if (ret) + return ret; + smu_force_clk_levels(smu, PP_SCLK, 1 << sclk_mask); + smu_force_clk_levels(smu, PP_MCLK, 1 << mclk_mask); + break; + + case AMD_DPM_FORCED_LEVEL_MANUAL: + case AMD_DPM_FORCED_LEVEL_PROFILE_EXIT: + default: + break; + } + + if (!ret) + smu_dpm_ctx->dpm_level = level; + } + + if (smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL) { + index = fls(smu->workload_mask); + index = index > 0 && index <= WORKLOAD_POLICY_MAX ? index - 1 : 0; + workload = smu->workload_setting[index]; + + if (smu->power_profile_mode != workload) + smu_set_power_profile_mode(smu, &workload, 0); + } + + return ret; +} + +int smu_handle_task(struct smu_context *smu, + enum amd_dpm_forced_level level, + enum amd_pp_task task_id) +{ + int ret = 0; + + switch (task_id) { + case AMD_PP_TASK_DISPLAY_CONFIG_CHANGE: + ret = smu_pre_display_config_changed(smu); + if (ret) + return ret; + ret = smu_set_cpu_power_state(smu); + if (ret) + return ret; + ret = smu_adjust_power_state_dynamic(smu, level, false); + break; + case AMD_PP_TASK_COMPLETE_INIT: + case AMD_PP_TASK_READJUST_POWER_STATE: + ret = smu_adjust_power_state_dynamic(smu, level, true); + break; + default: + break; + } + + return ret; +} + +const struct amd_ip_funcs smu_ip_funcs = { + .name = "smu", + .early_init = smu_early_init, + .late_init = smu_late_init, + .sw_init = smu_sw_init, + .sw_fini = smu_sw_fini, + .hw_init = smu_hw_init, + .hw_fini = smu_hw_fini, + .suspend = smu_suspend, + .resume = smu_resume, + .is_idle = NULL, + .check_soft_reset = NULL, + .wait_for_idle = NULL, + .soft_reset = NULL, + .set_clockgating_state = smu_set_clockgating_state, + .set_powergating_state = smu_set_powergating_state, + .enable_umd_pstate = smu_enable_umd_pstate, +}; + +const struct amdgpu_ip_block_version smu_v11_0_ip_block = +{ + .type = AMD_IP_BLOCK_TYPE_SMC, + .major = 11, + .minor = 0, + .rev = 0, + .funcs = &smu_ip_funcs, +}; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile b/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile index 0b3c6d1d52e4..cc63705920dc 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile @@ -35,7 +35,8 @@ HARDWARE_MGR = hwmgr.o processpptables.o \ vega12_thermal.o \ pp_overdriver.o smu_helper.o \ vega20_processpptables.o vega20_hwmgr.o vega20_powertune.o \ - vega20_thermal.o common_baco.o vega10_baco.o vega20_baco.o + vega20_thermal.o common_baco.o vega10_baco.o vega20_baco.o \ + vega12_baco.o smu9_baco.o AMD_PP_HWMGR = $(addprefix $(AMD_PP_PATH)/hwmgr/,$(HARDWARE_MGR)) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c index c1c51c115e57..70f7f47a2fcf 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c @@ -76,7 +76,7 @@ int phm_set_power_state(struct pp_hwmgr *hwmgr, int phm_enable_dynamic_state_management(struct pp_hwmgr *hwmgr) { struct amdgpu_device *adev = NULL; - int ret = -EINVAL;; + int ret = -EINVAL; PHM_FUNC_CHECK(hwmgr); adev = hwmgr->adev; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c index 0ad8fe4a6277..f32e3d0aaea6 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c @@ -114,11 +114,6 @@ static int smu10_initialize_dpm_defaults(struct pp_hwmgr *hwmgr) smu10_data->num_active_display = 0; smu10_data->deep_sleep_dcefclk = 0; - if (hwmgr->feature_mask & PP_GFXOFF_MASK) - smu10_data->gfx_off_controled_by_driver = true; - else - smu10_data->gfx_off_controled_by_driver = false; - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SclkDeepSleep); @@ -330,9 +325,9 @@ static bool smu10_is_gfx_on(struct pp_hwmgr *hwmgr) static int smu10_disable_gfx_off(struct pp_hwmgr *hwmgr) { - struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend); + struct amdgpu_device *adev = hwmgr->adev; - if (smu10_data->gfx_off_controled_by_driver) { + if (adev->pm.pp_feature & PP_GFXOFF_MASK) { smum_send_msg_to_smc(hwmgr, PPSMC_MSG_DisableGfxOff); /* confirm gfx is back to "on" state */ @@ -350,9 +345,9 @@ static int smu10_disable_dpm_tasks(struct pp_hwmgr *hwmgr) static int smu10_enable_gfx_off(struct pp_hwmgr *hwmgr) { - struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend); + struct amdgpu_device *adev = hwmgr->adev; - if (smu10_data->gfx_off_controled_by_driver) + if (adev->pm.pp_feature & PP_GFXOFF_MASK) smum_send_msg_to_smc(hwmgr, PPSMC_MSG_EnableGfxOff); return 0; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c index 83d3d935f3ac..048757e8f494 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c @@ -77,7 +77,7 @@ #define PCIE_BUS_CLK 10000 #define TCLK (PCIE_BUS_CLK / 10) -static const struct profile_mode_setting smu7_profiling[7] = +static struct profile_mode_setting smu7_profiling[7] = {{0, 0, 0, 0, 0, 0, 0, 0}, {1, 0, 100, 30, 1, 0, 100, 10}, {1, 10, 0, 30, 0, 0, 0, 0}, @@ -4984,17 +4984,27 @@ static int smu7_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, uint mode = input[size]; switch (mode) { case PP_SMC_POWER_PROFILE_CUSTOM: - if (size < 8) + if (size < 8 && size != 0) return -EINVAL; - - tmp.bupdate_sclk = input[0]; - tmp.sclk_up_hyst = input[1]; - tmp.sclk_down_hyst = input[2]; - tmp.sclk_activity = input[3]; - tmp.bupdate_mclk = input[4]; - tmp.mclk_up_hyst = input[5]; - tmp.mclk_down_hyst = input[6]; - tmp.mclk_activity = input[7]; + /* If only CUSTOM is passed in, use the saved values. Check + * that we actually have a CUSTOM profile by ensuring that + * the "use sclk" or the "use mclk" bits are set + */ + tmp = smu7_profiling[PP_SMC_POWER_PROFILE_CUSTOM]; + if (size == 0) { + if (tmp.bupdate_sclk == 0 && tmp.bupdate_mclk == 0) + return -EINVAL; + } else { + tmp.bupdate_sclk = input[0]; + tmp.sclk_up_hyst = input[1]; + tmp.sclk_down_hyst = input[2]; + tmp.sclk_activity = input[3]; + tmp.bupdate_mclk = input[4]; + tmp.mclk_up_hyst = input[5]; + tmp.mclk_down_hyst = input[6]; + tmp.mclk_activity = input[7]; + smu7_profiling[PP_SMC_POWER_PROFILE_CUSTOM] = tmp; + } if (!smum_update_dpm_settings(hwmgr, &tmp)) { memcpy(&data->current_profile_setting, &tmp, sizeof(struct profile_mode_setting)); hwmgr->power_profile_mode = mode; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu9_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu9_baco.c new file mode 100644 index 000000000000..de0a37f7c632 --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu9_baco.c @@ -0,0 +1,66 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include "amdgpu.h" +#include "soc15.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" +#include "soc15_common.h" +#include "vega10_inc.h" +#include "smu9_baco.h" + +int smu9_baco_get_capability(struct pp_hwmgr *hwmgr, bool *cap) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev); + uint32_t reg, data; + + *cap = false; + if (!phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_BACO)) + return 0; + + WREG32(0x12074, 0xFFF0003B); + data = RREG32(0x12075); + + if (data == 0x1) { + reg = RREG32_SOC15(NBIF, 0, mmRCC_BIF_STRAP0); + + if (reg & RCC_BIF_STRAP0__STRAP_PX_CAPABLE_MASK) + *cap = true; + } + + return 0; +} + +int smu9_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev); + uint32_t reg; + + reg = RREG32_SOC15(NBIF, 0, mmBACO_CNTL); + + if (reg & BACO_CNTL__BACO_MODE_MASK) + /* gfx has already entered BACO state */ + *state = BACO_STATE_IN; + else + *state = BACO_STATE_OUT; + return 0; +} diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu9_baco.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu9_baco.h new file mode 100644 index 000000000000..84e90f801ac3 --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu9_baco.h @@ -0,0 +1,31 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __SMU9_BACO_H__ +#define __SMU9_BACO_H__ +#include "hwmgr.h" +#include "common_baco.h" + +extern int smu9_baco_get_capability(struct pp_hwmgr *hwmgr, bool *cap); +extern int smu9_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state); + +#endif diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_baco.c index 7337be5602e4..d168af4a4d78 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_baco.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_baco.c @@ -85,48 +85,11 @@ static const struct soc15_baco_cmd_entry clean_baco_tbl[] = {CMD_WRITE, SOC15_REG_ENTRY(NBIF, 0, mmBIOS_SCRATCH_7), 0, 0, 0, 0}, }; -int vega10_baco_get_capability(struct pp_hwmgr *hwmgr, bool *cap) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev); - uint32_t reg, data; - - *cap = false; - if (!phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_BACO)) - return 0; - - WREG32(0x12074, 0xFFF0003B); - data = RREG32(0x12075); - - if (data == 0x1) { - reg = RREG32_SOC15(NBIF, 0, mmRCC_BIF_STRAP0); - - if (reg & RCC_BIF_STRAP0__STRAP_PX_CAPABLE_MASK) - *cap = true; - } - - return 0; -} - -int vega10_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev); - uint32_t reg; - - reg = RREG32_SOC15(NBIF, 0, mmBACO_CNTL); - - if (reg & BACO_CNTL__BACO_MODE_MASK) - /* gfx has already entered BACO state */ - *state = BACO_STATE_IN; - else - *state = BACO_STATE_OUT; - return 0; -} - int vega10_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state) { enum BACO_STATE cur_state; - vega10_baco_get_state(hwmgr, &cur_state); + smu9_baco_get_state(hwmgr, &cur_state); if (cur_state == state) /* aisc already in the target state */ diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_baco.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_baco.h index f7a3ffa744b3..96d793f026a5 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_baco.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_baco.h @@ -22,11 +22,8 @@ */ #ifndef __VEGA10_BACO_H__ #define __VEGA10_BACO_H__ -#include "hwmgr.h" -#include "common_baco.h" +#include "smu9_baco.h" -extern int vega10_baco_get_capability(struct pp_hwmgr *hwmgr, bool *cap); -extern int vega10_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state); extern int vega10_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state); #endif diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c index 5c4f701939ea..384c37875cd0 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c @@ -1427,6 +1427,15 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) vega10_setup_default_pcie_table(hwmgr); + /* Zero out the saved copy of the CUSTOM profile + * This will be checked when trying to set the profile + * and will require that new values be passed in + */ + data->custom_profile_mode[0] = 0; + data->custom_profile_mode[1] = 0; + data->custom_profile_mode[2] = 0; + data->custom_profile_mode[3] = 0; + /* save a copy of the default DPM table */ memcpy(&(data->golden_dpm_table), &(data->dpm_table), sizeof(struct vega10_dpm_table)); @@ -4904,16 +4913,23 @@ static int vega10_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, ui uint8_t FPS; uint8_t use_rlc_busy; uint8_t min_active_level; + uint32_t power_profile_mode = input[size]; - hwmgr->power_profile_mode = input[size]; - - smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetWorkloadMask, - 1<power_profile_mode); - - if (hwmgr->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) { - if (size == 0 || size > 4) + if (power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) { + if (size != 0 && size != 4) return -EINVAL; + /* If size = 0 and the CUSTOM profile has been set already + * then just apply the profile. The copy stored in the hwmgr + * is zeroed out on init + */ + if (size == 0) { + if (data->custom_profile_mode[0] != 0) + goto out; + else + return -EINVAL; + } + data->custom_profile_mode[0] = busy_set_point = input[0]; data->custom_profile_mode[1] = FPS = input[1]; data->custom_profile_mode[2] = use_rlc_busy = input[2]; @@ -4924,6 +4940,11 @@ static int vega10_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, ui use_rlc_busy << 16 | min_active_level<<24); } +out: + smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetWorkloadMask, + 1 << power_profile_mode); + hwmgr->power_profile_mode = power_profile_mode; + return 0; } @@ -5170,8 +5191,8 @@ static const struct pp_hwmgr_func vega10_hwmgr_funcs = { .set_power_limit = vega10_set_power_limit, .odn_edit_dpm_table = vega10_odn_edit_dpm_table, .get_performance_level = vega10_get_performance_level, - .get_asic_baco_capability = vega10_baco_get_capability, - .get_asic_baco_state = vega10_baco_get_state, + .get_asic_baco_capability = smu9_baco_get_capability, + .get_asic_baco_state = smu9_baco_get_state, .set_asic_baco_state = vega10_baco_set_state, .enable_mgpu_fan_boost = vega10_enable_mgpu_fan_boost, .get_ppfeature_status = vega10_get_ppfeature_status, diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_baco.c new file mode 100644 index 000000000000..9d8ca94a8f0c --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_baco.c @@ -0,0 +1,119 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include "amdgpu.h" +#include "soc15.h" +#include "soc15_hw_ip.h" +#include "vega10_ip_offset.h" +#include "soc15_common.h" +#include "vega12_inc.h" +#include "vega12_ppsmc.h" +#include "vega12_baco.h" + +static const struct soc15_baco_cmd_entry pre_baco_tbl[] = +{ + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmBIF_DOORBELL_CNTL_BASE_IDX, mmBIF_DOORBELL_CNTL, BIF_DOORBELL_CNTL__DOORBELL_MONITOR_EN_MASK, BIF_DOORBELL_CNTL__DOORBELL_MONITOR_EN__SHIFT, 0, 0 }, + { CMD_WRITE, NBIF_HWID, 0, mmBIF_FB_EN_BASE_IDX, mmBIF_FB_EN, 0, 0, 0, 0 }, + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_DSTATE_BYPASS_MASK, BACO_CNTL__BACO_DSTATE_BYPASS__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_RST_INTR_MASK_MASK, BACO_CNTL__BACO_RST_INTR_MASK__SHIFT, 0, 1 } +}; + +static const struct soc15_baco_cmd_entry enter_baco_tbl[] = +{ + { CMD_WAITFOR, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__SOC_DOMAIN_IDLE_MASK, THM_BACO_CNTL__SOC_DOMAIN_IDLE__SHIFT, 0xffffffff, 0x80000000 }, + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_BIF_LCLK_SWITCH_MASK, BACO_CNTL__BACO_BIF_LCLK_SWITCH__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_DUMMY_EN_MASK, BACO_CNTL__BACO_DUMMY_EN__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_SOC_VDCI_RESET_MASK, THM_BACO_CNTL__BACO_SOC_VDCI_RESET__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_SMNCLK_MUX_MASK, THM_BACO_CNTL__BACO_SMNCLK_MUX__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_ISO_EN_MASK, THM_BACO_CNTL__BACO_ISO_EN__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_AEB_ISO_EN_MASK, THM_BACO_CNTL__BACO_AEB_ISO_EN__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_ANA_ISO_EN_MASK, THM_BACO_CNTL__BACO_ANA_ISO_EN__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_SOC_REFCLK_OFF_MASK, THM_BACO_CNTL__BACO_SOC_REFCLK_OFF__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 1 }, + { CMD_DELAY_MS, 0, 0, 0, 5, 0 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_RESET_EN_MASK, THM_BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_PWROKRAW_CNTL_MASK, THM_BACO_CNTL__BACO_PWROKRAW_CNTL__SHIFT, 0, 0 }, + { CMD_WAITFOR, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, BACO_CNTL__BACO_MODE__SHIFT, 0xffffffff, 0x100 } +}; + +static const struct soc15_baco_cmd_entry exit_baco_tbl[] = +{ + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 0 }, + { CMD_DELAY_MS, 0, 0, 0, 0, 0, 0, 10, 0 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_SOC_REFCLK_OFF_MASK, THM_BACO_CNTL__BACO_SOC_REFCLK_OFF__SHIFT, 0, 0 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_ANA_ISO_EN_MASK, THM_BACO_CNTL__BACO_ANA_ISO_EN__SHIFT, 0, 0 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_AEB_ISO_EN_MASK, THM_BACO_CNTL__BACO_AEB_ISO_EN__SHIFT, 0, 0 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_ISO_EN_MASK, THM_BACO_CNTL__BACO_ISO_EN__SHIFT, 0, 0 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_PWROKRAW_CNTL_MASK, THM_BACO_CNTL__BACO_PWROKRAW_CNTL__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_SMNCLK_MUX_MASK, THM_BACO_CNTL__BACO_SMNCLK_MUX__SHIFT, 0, 0 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_SOC_VDCI_RESET_MASK, THM_BACO_CNTL__BACO_SOC_VDCI_RESET__SHIFT, 0, 0 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_EXIT_MASK, THM_BACO_CNTL__BACO_EXIT__SHIFT, 0, 1 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_RESET_EN_MASK, THM_BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0 }, + { CMD_WAITFOR, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_EXIT_MASK, 0, 0xffffffff, 0 }, + { CMD_READMODIFYWRITE, THM_HWID, 0, mmTHM_BACO_CNTL_BASE_IDX, mmTHM_BACO_CNTL, THM_BACO_CNTL__BACO_SB_AXI_FENCE_MASK, THM_BACO_CNTL__BACO_SB_AXI_FENCE__SHIFT, 0, 0 }, + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_DUMMY_EN_MASK, BACO_CNTL__BACO_DUMMY_EN__SHIFT, 0, 0 }, + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_BIF_LCLK_SWITCH_MASK, BACO_CNTL__BACO_BIF_LCLK_SWITCH__SHIFT, 0, 0 }, + { CMD_READMODIFYWRITE, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 0 }, + { CMD_WAITFOR, NBIF_HWID, 0, mmRCC_BACO_CNTL_MISC_BASE_IDX, mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, 0, 0xffffffff, 0 } +}; + +static const struct soc15_baco_cmd_entry clean_baco_tbl[] = +{ + { CMD_WRITE, NBIF_HWID, 0, mmBIOS_SCRATCH_6_BASE_IDX, mmBIOS_SCRATCH_6, 0, 0, 0, 0 }, + { CMD_WRITE, NBIF_HWID, 0, mmBIOS_SCRATCH_7_BASE_IDX, mmBIOS_SCRATCH_7, 0, 0, 0, 0 } +}; + +int vega12_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state) +{ + enum BACO_STATE cur_state; + + smu9_baco_get_state(hwmgr, &cur_state); + + if (cur_state == state) + /* aisc already in the target state */ + return 0; + + if (state == BACO_STATE_IN) { + if (soc15_baco_program_registers(hwmgr, pre_baco_tbl, + ARRAY_SIZE(pre_baco_tbl))) { + if (smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_EnterBaco, 0)) + return -EINVAL; + + if (soc15_baco_program_registers(hwmgr, enter_baco_tbl, + ARRAY_SIZE(enter_baco_tbl))) + return 0; + } + } else if (state == BACO_STATE_OUT) { + /* HW requires at least 20ms between regulator off and on */ + msleep(20); + /* Execute Hardware BACO exit sequence */ + if (soc15_baco_program_registers(hwmgr, exit_baco_tbl, + ARRAY_SIZE(exit_baco_tbl))) { + if (soc15_baco_program_registers(hwmgr, clean_baco_tbl, + ARRAY_SIZE(clean_baco_tbl))) + return 0; + } + } + + return -EINVAL; +} diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_baco.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_baco.h new file mode 100644 index 000000000000..57b72e5a95ae --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_baco.h @@ -0,0 +1,29 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __VEGA12_BACO_H__ +#define __VEGA12_BACO_H__ +#include "smu9_baco.h" + +extern int vega12_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state); + +#endif diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c index bdb48e94eff6..707cd4b0357f 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c @@ -45,6 +45,7 @@ #include "ppinterrupt.h" #include "pp_overdriver.h" #include "pp_thermal.h" +#include "vega12_baco.h" static int vega12_force_clock_level(struct pp_hwmgr *hwmgr, @@ -2626,8 +2627,12 @@ static const struct pp_hwmgr_func vega12_hwmgr_funcs = { .start_thermal_controller = vega12_start_thermal_controller, .powergate_gfx = vega12_gfx_off_control, .get_performance_level = vega12_get_performance_level, + .get_asic_baco_capability = smu9_baco_get_capability, + .get_asic_baco_state = smu9_baco_get_state, + .set_asic_baco_state = vega12_baco_set_state, .get_ppfeature_status = vega12_get_ppfeature_status, .set_ppfeature_status = vega12_set_ppfeature_status, + }; int vega12_hwmgr_init(struct pp_hwmgr *hwmgr) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_inc.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_inc.h index 30b278c50222..e6d9e84059e1 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_inc.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_inc.h @@ -35,5 +35,7 @@ #include "asic_reg/gc/gc_9_2_1_sh_mask.h" #include "asic_reg/nbio/nbio_6_1_offset.h" +#include "asic_reg/nbio/nbio_6_1_offset.h" +#include "asic_reg/nbio/nbio_6_1_sh_mask.h" #endif diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.c index 5e8602a79b1c..df6ff9252401 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.c @@ -27,6 +27,7 @@ #include "vega20_inc.h" #include "vega20_ppsmc.h" #include "vega20_baco.h" +#include "vega20_smumgr.h" @@ -101,3 +102,14 @@ int vega20_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state) return 0; } + +int vega20_baco_apply_vdci_flush_workaround(struct pp_hwmgr *hwmgr) +{ + int ret = 0; + + ret = vega20_set_pptable_driver_address(hwmgr); + if (ret) + return ret; + + return smum_send_msg_to_smc(hwmgr, PPSMC_MSG_BacoWorkAroundFlushVDCI); +} diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.h index 51c7f8392925..f06471e712dc 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.h @@ -28,5 +28,6 @@ extern int vega20_baco_get_capability(struct pp_hwmgr *hwmgr, bool *cap); extern int vega20_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state); extern int vega20_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state); +extern int vega20_baco_apply_vdci_flush_workaround(struct pp_hwmgr *hwmgr); #endif diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c index 9aa7bec1b5fe..39a547084e90 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c @@ -91,6 +91,12 @@ static void vega20_set_default_registry_data(struct pp_hwmgr *hwmgr) * MP0CLK DS */ data->registry_data.disallowed_features = 0xE0041C00; + /* ECC feature should be disabled on old SMUs */ + smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetSmuVersion); + hwmgr->smu_version = smum_get_argument(hwmgr); + if (hwmgr->smu_version < 0x282100) + data->registry_data.disallowed_features |= FEATURE_ECC_MASK; + data->registry_data.od_state_in_dc_support = 0; data->registry_data.thermal_support = 1; data->registry_data.skip_baco_hardware = 0; @@ -357,6 +363,7 @@ static void vega20_init_dpm_defaults(struct pp_hwmgr *hwmgr) data->smu_features[GNLD_DS_MP1CLK].smu_feature_id = FEATURE_DS_MP1CLK_BIT; data->smu_features[GNLD_DS_MP0CLK].smu_feature_id = FEATURE_DS_MP0CLK_BIT; data->smu_features[GNLD_XGMI].smu_feature_id = FEATURE_XGMI_BIT; + data->smu_features[GNLD_ECC].smu_feature_id = FEATURE_ECC_BIT; for (i = 0; i < GNLD_FEATURES_MAX; i++) { data->smu_features[i].smu_feature_bitmap = @@ -427,6 +434,7 @@ static int vega20_hwmgr_backend_init(struct pp_hwmgr *hwmgr) hwmgr->platform_descriptor.clockStep.memoryClock = 500; data->total_active_cus = adev->gfx.cu_info.number; + data->is_custom_profile_set = false; return 0; } @@ -443,6 +451,7 @@ static int vega20_init_sclk_threshold(struct pp_hwmgr *hwmgr) static int vega20_setup_asic_task(struct pp_hwmgr *hwmgr) { + struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev); int ret = 0; ret = vega20_init_sclk_threshold(hwmgr); @@ -450,7 +459,15 @@ static int vega20_setup_asic_task(struct pp_hwmgr *hwmgr) "Failed to init sclk threshold!", return ret); - return 0; + if (adev->in_baco_reset) { + adev->in_baco_reset = 0; + + ret = vega20_baco_apply_vdci_flush_workaround(hwmgr); + if (ret) + pr_err("Failed to apply vega20 baco workaround!\n"); + } + + return ret; } /* @@ -3020,7 +3037,8 @@ static int vega20_get_ppfeature_status(struct pp_hwmgr *hwmgr, char *buf) "FCLK_DS", "MP1CLK_DS", "MP0CLK_DS", - "XGMI"}; + "XGMI", + "ECC"}; static const char *output_title[] = { "FEATURES", "BITMASK", @@ -3462,6 +3480,7 @@ static int vega20_apply_clocks_adjust_rules(struct pp_hwmgr *hwmgr) struct vega20_single_dpm_table *dpm_table; bool vblank_too_short = false; bool disable_mclk_switching; + bool disable_fclk_switching; uint32_t i, latency; disable_mclk_switching = ((1 < hwmgr->display_config->num_display) && @@ -3537,13 +3556,20 @@ static int vega20_apply_clocks_adjust_rules(struct pp_hwmgr *hwmgr) if (hwmgr->display_config->nb_pstate_switch_disable) dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + if ((disable_mclk_switching && + (dpm_table->dpm_state.hard_min_level == dpm_table->dpm_levels[dpm_table->count - 1].value)) || + hwmgr->display_config->min_mem_set_clock / 100 >= dpm_table->dpm_levels[dpm_table->count - 1].value) + disable_fclk_switching = true; + else + disable_fclk_switching = false; + /* fclk */ dpm_table = &(data->dpm_table.fclk_table); dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; dpm_table->dpm_state.soft_max_level = VG20_CLOCK_MAX_DEFAULT; dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; dpm_table->dpm_state.hard_max_level = VG20_CLOCK_MAX_DEFAULT; - if (hwmgr->display_config->nb_pstate_switch_disable) + if (hwmgr->display_config->nb_pstate_switch_disable || disable_fclk_switching) dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; /* vclk */ @@ -3810,16 +3836,19 @@ static int vega20_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, ui { DpmActivityMonitorCoeffInt_t activity_monitor; int workload_type, result = 0; + uint32_t power_profile_mode = input[size]; - hwmgr->power_profile_mode = input[size]; - - if (hwmgr->power_profile_mode > PP_SMC_POWER_PROFILE_CUSTOM) { - pr_err("Invalid power profile mode %d\n", hwmgr->power_profile_mode); + if (power_profile_mode > PP_SMC_POWER_PROFILE_CUSTOM) { + pr_err("Invalid power profile mode %d\n", power_profile_mode); return -EINVAL; } - if (hwmgr->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) { - if (size < 10) + if (power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) { + struct vega20_hwmgr *data = + (struct vega20_hwmgr *)(hwmgr->backend); + if (size == 0 && !data->is_custom_profile_set) + return -EINVAL; + if (size < 10 && size != 0) return -EINVAL; result = vega20_get_activity_monitor_coeff(hwmgr, @@ -3829,6 +3858,13 @@ static int vega20_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, ui "[SetPowerProfile] Failed to get activity monitor!", return result); + /* If size==0, then we want to apply the already-configured + * CUSTOM profile again. Just apply it, since we checked its + * validity above + */ + if (size == 0) + goto out; + switch (input[0]) { case 0: /* Gfxclk */ activity_monitor.Gfx_FPS = input[1]; @@ -3879,17 +3915,21 @@ static int vega20_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, ui result = vega20_set_activity_monitor_coeff(hwmgr, (uint8_t *)(&activity_monitor), WORKLOAD_PPLIB_CUSTOM_BIT); + data->is_custom_profile_set = true; PP_ASSERT_WITH_CODE(!result, "[SetPowerProfile] Failed to set activity monitor!", return result); } +out: /* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */ workload_type = - conv_power_profile_to_pplib_workload(hwmgr->power_profile_mode); + conv_power_profile_to_pplib_workload(power_profile_mode); smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetWorkloadMask, 1 << workload_type); + hwmgr->power_profile_mode = power_profile_mode; + return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.h index a5bc758ae097..2c3125f82b24 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.h @@ -80,6 +80,7 @@ enum { GNLD_DS_MP1CLK, GNLD_DS_MP0CLK, GNLD_XGMI, + GNLD_ECC, GNLD_FEATURES_MAX }; @@ -530,6 +531,8 @@ struct vega20_hwmgr { bool pcie_parameters_override; uint32_t pcie_gen_level1; uint32_t pcie_width_level1; + + bool is_custom_profile_set; }; #define VEGA20_DPM2_NEAR_TDP_DEC 10 diff --git a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h new file mode 100644 index 000000000000..c8b168b3413b --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h @@ -0,0 +1,770 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ +#ifndef __AMDGPU_SMU_H__ +#define __AMDGPU_SMU_H__ + +#include "amdgpu.h" +#include "kgd_pp_interface.h" +#include "dm_pp_interface.h" + +struct smu_hw_power_state { + unsigned int magic; +}; + +struct smu_power_state; + +enum smu_state_ui_label { + SMU_STATE_UI_LABEL_NONE, + SMU_STATE_UI_LABEL_BATTERY, + SMU_STATE_UI_TABEL_MIDDLE_LOW, + SMU_STATE_UI_LABEL_BALLANCED, + SMU_STATE_UI_LABEL_MIDDLE_HIGHT, + SMU_STATE_UI_LABEL_PERFORMANCE, + SMU_STATE_UI_LABEL_BACO, +}; + +enum smu_state_classification_flag { + SMU_STATE_CLASSIFICATION_FLAG_BOOT = 0x0001, + SMU_STATE_CLASSIFICATION_FLAG_THERMAL = 0x0002, + SMU_STATE_CLASSIFICATIN_FLAG_LIMITED_POWER_SOURCE = 0x0004, + SMU_STATE_CLASSIFICATION_FLAG_RESET = 0x0008, + SMU_STATE_CLASSIFICATION_FLAG_FORCED = 0x0010, + SMU_STATE_CLASSIFICATION_FLAG_USER_3D_PERFORMANCE = 0x0020, + SMU_STATE_CLASSIFICATION_FLAG_USER_2D_PERFORMANCE = 0x0040, + SMU_STATE_CLASSIFICATION_FLAG_3D_PERFORMANCE = 0x0080, + SMU_STATE_CLASSIFICATION_FLAG_AC_OVERDIRVER_TEMPLATE = 0x0100, + SMU_STATE_CLASSIFICATION_FLAG_UVD = 0x0200, + SMU_STATE_CLASSIFICATION_FLAG_3D_PERFORMANCE_LOW = 0x0400, + SMU_STATE_CLASSIFICATION_FLAG_ACPI = 0x0800, + SMU_STATE_CLASSIFICATION_FLAG_HD2 = 0x1000, + SMU_STATE_CLASSIFICATION_FLAG_UVD_HD = 0x2000, + SMU_STATE_CLASSIFICATION_FLAG_UVD_SD = 0x4000, + SMU_STATE_CLASSIFICATION_FLAG_USER_DC_PERFORMANCE = 0x8000, + SMU_STATE_CLASSIFICATION_FLAG_DC_OVERDIRVER_TEMPLATE = 0x10000, + SMU_STATE_CLASSIFICATION_FLAG_BACO = 0x20000, + SMU_STATE_CLASSIFICATIN_FLAG_LIMITED_POWER_SOURCE2 = 0x40000, + SMU_STATE_CLASSIFICATION_FLAG_ULV = 0x80000, + SMU_STATE_CLASSIFICATION_FLAG_UVD_MVC = 0x100000, +}; + +struct smu_state_classification_block { + enum smu_state_ui_label ui_label; + enum smu_state_classification_flag flags; + int bios_index; + bool temporary_state; + bool to_be_deleted; +}; + +struct smu_state_pcie_block { + unsigned int lanes; +}; + +enum smu_refreshrate_source { + SMU_REFRESHRATE_SOURCE_EDID, + SMU_REFRESHRATE_SOURCE_EXPLICIT +}; + +struct smu_state_display_block { + bool disable_frame_modulation; + bool limit_refreshrate; + enum smu_refreshrate_source refreshrate_source; + int explicit_refreshrate; + int edid_refreshrate_index; + bool enable_vari_bright; +}; + +struct smu_state_memroy_block { + bool dll_off; + uint8_t m3arb; + uint8_t unused[3]; +}; + +struct smu_state_software_algorithm_block { + bool disable_load_balancing; + bool enable_sleep_for_timestamps; +}; + +struct smu_temperature_range { + int min; + int max; +}; + +struct smu_state_validation_block { + bool single_display_only; + bool disallow_on_dc; + uint8_t supported_power_levels; +}; + +struct smu_uvd_clocks { + uint32_t vclk; + uint32_t dclk; +}; + +/** +* Structure to hold a SMU Power State. +*/ +struct smu_power_state { + uint32_t id; + struct list_head ordered_list; + struct list_head all_states_list; + + struct smu_state_classification_block classification; + struct smu_state_validation_block validation; + struct smu_state_pcie_block pcie; + struct smu_state_display_block display; + struct smu_state_memroy_block memory; + struct smu_temperature_range temperatures; + struct smu_state_software_algorithm_block software; + struct smu_uvd_clocks uvd_clocks; + struct smu_hw_power_state hardware; +}; + +enum smu_message_type +{ + SMU_MSG_TestMessage = 0, + SMU_MSG_GetSmuVersion, + SMU_MSG_GetDriverIfVersion, + SMU_MSG_SetAllowedFeaturesMaskLow, + SMU_MSG_SetAllowedFeaturesMaskHigh, + SMU_MSG_EnableAllSmuFeatures, + SMU_MSG_DisableAllSmuFeatures, + SMU_MSG_EnableSmuFeaturesLow, + SMU_MSG_EnableSmuFeaturesHigh, + SMU_MSG_DisableSmuFeaturesLow, + SMU_MSG_DisableSmuFeaturesHigh, + SMU_MSG_GetEnabledSmuFeaturesLow, + SMU_MSG_GetEnabledSmuFeaturesHigh, + SMU_MSG_SetWorkloadMask, + SMU_MSG_SetPptLimit, + SMU_MSG_SetDriverDramAddrHigh, + SMU_MSG_SetDriverDramAddrLow, + SMU_MSG_SetToolsDramAddrHigh, + SMU_MSG_SetToolsDramAddrLow, + SMU_MSG_TransferTableSmu2Dram, + SMU_MSG_TransferTableDram2Smu, + SMU_MSG_UseDefaultPPTable, + SMU_MSG_UseBackupPPTable, + SMU_MSG_RunBtc, + SMU_MSG_RequestI2CBus, + SMU_MSG_ReleaseI2CBus, + SMU_MSG_SetFloorSocVoltage, + SMU_MSG_SoftReset, + SMU_MSG_StartBacoMonitor, + SMU_MSG_CancelBacoMonitor, + SMU_MSG_EnterBaco, + SMU_MSG_SetSoftMinByFreq, + SMU_MSG_SetSoftMaxByFreq, + SMU_MSG_SetHardMinByFreq, + SMU_MSG_SetHardMaxByFreq, + SMU_MSG_GetMinDpmFreq, + SMU_MSG_GetMaxDpmFreq, + SMU_MSG_GetDpmFreqByIndex, + SMU_MSG_GetDpmClockFreq, + SMU_MSG_GetSsVoltageByDpm, + SMU_MSG_SetMemoryChannelConfig, + SMU_MSG_SetGeminiMode, + SMU_MSG_SetGeminiApertureHigh, + SMU_MSG_SetGeminiApertureLow, + SMU_MSG_SetMinLinkDpmByIndex, + SMU_MSG_OverridePcieParameters, + SMU_MSG_OverDriveSetPercentage, + SMU_MSG_SetMinDeepSleepDcefclk, + SMU_MSG_ReenableAcDcInterrupt, + SMU_MSG_NotifyPowerSource, + SMU_MSG_SetUclkFastSwitch, + SMU_MSG_SetUclkDownHyst, + SMU_MSG_GfxDeviceDriverReset, + SMU_MSG_GetCurrentRpm, + SMU_MSG_SetVideoFps, + SMU_MSG_SetTjMax, + SMU_MSG_SetFanTemperatureTarget, + SMU_MSG_PrepareMp1ForUnload, + SMU_MSG_DramLogSetDramAddrHigh, + SMU_MSG_DramLogSetDramAddrLow, + SMU_MSG_DramLogSetDramSize, + SMU_MSG_SetFanMaxRpm, + SMU_MSG_SetFanMinPwm, + SMU_MSG_ConfigureGfxDidt, + SMU_MSG_NumOfDisplays, + SMU_MSG_RemoveMargins, + SMU_MSG_ReadSerialNumTop32, + SMU_MSG_ReadSerialNumBottom32, + SMU_MSG_SetSystemVirtualDramAddrHigh, + SMU_MSG_SetSystemVirtualDramAddrLow, + SMU_MSG_WaflTest, + SMU_MSG_SetFclkGfxClkRatio, + SMU_MSG_AllowGfxOff, + SMU_MSG_DisallowGfxOff, + SMU_MSG_GetPptLimit, + SMU_MSG_GetDcModeMaxDpmFreq, + SMU_MSG_GetDebugData, + SMU_MSG_SetXgmiMode, + SMU_MSG_RunAfllBtc, + SMU_MSG_ExitBaco, + SMU_MSG_PrepareMp1ForReset, + SMU_MSG_PrepareMp1ForShutdown, + SMU_MSG_SetMGpuFanBoostLimitRpm, + SMU_MSG_GetAVFSVoltageByDpm, + SMU_MSG_MAX_COUNT, +}; + +enum smu_memory_pool_size +{ + SMU_MEMORY_POOL_SIZE_ZERO = 0, + SMU_MEMORY_POOL_SIZE_256_MB = 0x10000000, + SMU_MEMORY_POOL_SIZE_512_MB = 0x20000000, + SMU_MEMORY_POOL_SIZE_1_GB = 0x40000000, + SMU_MEMORY_POOL_SIZE_2_GB = 0x80000000, +}; + +#define SMU_TABLE_INIT(tables, table_id, s, a, d) \ + do { \ + tables[table_id].size = s; \ + tables[table_id].align = a; \ + tables[table_id].domain = d; \ + } while (0) + +struct smu_table { + uint64_t size; + uint32_t align; + uint8_t domain; + uint64_t mc_address; + void *cpu_addr; + struct amdgpu_bo *bo; +}; + +enum smu_perf_level_designation { + PERF_LEVEL_ACTIVITY, + PERF_LEVEL_POWER_CONTAINMENT, +}; + +struct smu_performance_level { + uint32_t core_clock; + uint32_t memory_clock; + uint32_t vddc; + uint32_t vddci; + uint32_t non_local_mem_freq; + uint32_t non_local_mem_width; +}; + +struct smu_clock_info { + uint32_t min_mem_clk; + uint32_t max_mem_clk; + uint32_t min_eng_clk; + uint32_t max_eng_clk; + uint32_t min_bus_bandwidth; + uint32_t max_bus_bandwidth; +}; + +struct smu_bios_boot_up_values +{ + uint32_t revision; + uint32_t gfxclk; + uint32_t uclk; + uint32_t socclk; + uint32_t dcefclk; + uint32_t eclk; + uint32_t vclk; + uint32_t dclk; + uint16_t vddc; + uint16_t vddci; + uint16_t mvddc; + uint16_t vdd_gfx; + uint8_t cooling_id; + uint32_t pp_table_id; +}; + +struct smu_table_context +{ + void *power_play_table; + uint32_t power_play_table_size; + void *hardcode_pptable; + + void *max_sustainable_clocks; + struct smu_bios_boot_up_values boot_values; + void *driver_pptable; + struct smu_table *tables; + uint32_t table_count; + struct smu_table memory_pool; + uint16_t software_shutdown_temp; + uint8_t thermal_controller_type; + uint16_t TDPODLimit; + + uint8_t *od_feature_capabilities; + uint32_t *od_settings_max; + uint32_t *od_settings_min; + void *overdrive_table; + void *od8_settings; + bool od_gfxclk_update; + bool od_memclk_update; +}; + +struct smu_dpm_context { + uint32_t dpm_context_size; + void *dpm_context; + void *golden_dpm_context; + bool enable_umd_pstate; + enum amd_dpm_forced_level dpm_level; + enum amd_dpm_forced_level saved_dpm_level; + enum amd_dpm_forced_level requested_dpm_level; + struct smu_power_state *dpm_request_power_state; + struct smu_power_state *dpm_current_power_state; + struct mclock_latency_table *mclk_latency_table; +}; + +struct smu_power_context { + void *power_context; + uint32_t power_context_size; +}; + + +#define SMU_FEATURE_MAX (64) +struct smu_feature +{ + uint32_t feature_num; + DECLARE_BITMAP(supported, SMU_FEATURE_MAX); + DECLARE_BITMAP(allowed, SMU_FEATURE_MAX); + DECLARE_BITMAP(enabled, SMU_FEATURE_MAX); + struct mutex mutex; +}; + +struct smu_clocks { + uint32_t engine_clock; + uint32_t memory_clock; + uint32_t bus_bandwidth; + uint32_t engine_clock_in_sr; + uint32_t dcef_clock; + uint32_t dcef_clock_in_sr; +}; + +#define MAX_REGULAR_DPM_NUM 16 +struct mclk_latency_entries { + uint32_t frequency; + uint32_t latency; +}; +struct mclock_latency_table { + uint32_t count; + struct mclk_latency_entries entries[MAX_REGULAR_DPM_NUM]; +}; + +#define WORKLOAD_POLICY_MAX 7 +struct smu_context +{ + struct amdgpu_device *adev; + + const struct smu_funcs *funcs; + const struct pptable_funcs *ppt_funcs; + struct mutex mutex; + uint64_t pool_size; + + struct smu_table_context smu_table; + struct smu_dpm_context smu_dpm; + struct smu_power_context smu_power; + struct smu_feature smu_feature; + struct amd_pp_display_configuration *display_config; + + uint32_t pstate_sclk; + uint32_t pstate_mclk; + + bool od_enabled; + uint32_t power_limit; + uint32_t default_power_limit; + + bool support_power_containment; + bool disable_watermark; + +#define WATERMARKS_EXIST (1 << 0) +#define WATERMARKS_LOADED (1 << 1) + uint32_t watermarks_bitmap; + + uint32_t workload_mask; + uint32_t workload_prority[WORKLOAD_POLICY_MAX]; + uint32_t workload_setting[WORKLOAD_POLICY_MAX]; + uint32_t power_profile_mode; + uint32_t default_power_profile_mode; + + uint32_t smc_if_version; +}; + +struct pptable_funcs { + int (*alloc_dpm_context)(struct smu_context *smu); + int (*store_powerplay_table)(struct smu_context *smu); + int (*check_powerplay_table)(struct smu_context *smu); + int (*append_powerplay_table)(struct smu_context *smu); + int (*get_smu_msg_index)(struct smu_context *smu, uint32_t index); + int (*run_afll_btc)(struct smu_context *smu); + int (*get_unallowed_feature_mask)(struct smu_context *smu, uint32_t *feature_mask, uint32_t num); + enum amd_pm_state_type (*get_current_power_state)(struct smu_context *smu); + int (*set_default_dpm_table)(struct smu_context *smu); + int (*set_power_state)(struct smu_context *smu); + int (*populate_umd_state_clk)(struct smu_context *smu); + int (*print_clk_levels)(struct smu_context *smu, enum pp_clock_type type, char *buf); + int (*force_clk_levels)(struct smu_context *smu, enum pp_clock_type type, uint32_t mask); + int (*set_default_od8_settings)(struct smu_context *smu); + int (*update_specified_od8_value)(struct smu_context *smu, + uint32_t index, + uint32_t value); + int (*get_od_percentage)(struct smu_context *smu, enum pp_clock_type type); + int (*set_od_percentage)(struct smu_context *smu, + enum pp_clock_type type, + uint32_t value); + int (*od_edit_dpm_table)(struct smu_context *smu, + enum PP_OD_DPM_TABLE_COMMAND type, + long *input, uint32_t size); + int (*get_clock_by_type_with_latency)(struct smu_context *smu, + enum amd_pp_clock_type type, + struct + pp_clock_levels_with_latency + *clocks); + int (*get_clock_by_type_with_voltage)(struct smu_context *smu, + enum amd_pp_clock_type type, + struct + pp_clock_levels_with_voltage + *clocks); + int (*get_power_profile_mode)(struct smu_context *smu, char *buf); + int (*set_power_profile_mode)(struct smu_context *smu, long *input, uint32_t size); + enum amd_dpm_forced_level (*get_performance_level)(struct smu_context *smu); + int (*force_performance_level)(struct smu_context *smu, enum amd_dpm_forced_level level); + int (*pre_display_config_changed)(struct smu_context *smu); + int (*display_config_changed)(struct smu_context *smu); + int (*apply_clocks_adjust_rules)(struct smu_context *smu); + int (*notify_smc_dispaly_config)(struct smu_context *smu); + int (*force_dpm_limit_value)(struct smu_context *smu, bool highest); + int (*unforce_dpm_levels)(struct smu_context *smu); + int (*upload_dpm_level)(struct smu_context *smu, bool max, + uint32_t feature_mask); + int (*get_profiling_clk_mask)(struct smu_context *smu, + enum amd_dpm_forced_level level, + uint32_t *sclk_mask, + uint32_t *mclk_mask, + uint32_t *soc_mask); + int (*set_cpu_power_state)(struct smu_context *smu); +}; + +struct smu_funcs +{ + int (*init_microcode)(struct smu_context *smu); + int (*init_smc_tables)(struct smu_context *smu); + int (*fini_smc_tables)(struct smu_context *smu); + int (*init_power)(struct smu_context *smu); + int (*fini_power)(struct smu_context *smu); + int (*load_microcode)(struct smu_context *smu); + int (*check_fw_status)(struct smu_context *smu); + int (*read_pptable_from_vbios)(struct smu_context *smu); + int (*get_vbios_bootup_values)(struct smu_context *smu); + int (*get_clk_info_from_vbios)(struct smu_context *smu); + int (*check_pptable)(struct smu_context *smu); + int (*parse_pptable)(struct smu_context *smu); + int (*populate_smc_pptable)(struct smu_context *smu); + int (*check_fw_version)(struct smu_context *smu); + int (*write_pptable)(struct smu_context *smu); + int (*set_min_dcef_deep_sleep)(struct smu_context *smu); + int (*set_tool_table_location)(struct smu_context *smu); + int (*notify_memory_pool_location)(struct smu_context *smu); + int (*write_watermarks_table)(struct smu_context *smu); + int (*set_last_dcef_min_deep_sleep_clk)(struct smu_context *smu); + int (*system_features_control)(struct smu_context *smu, bool en); + int (*send_smc_msg)(struct smu_context *smu, uint16_t msg); + int (*send_smc_msg_with_param)(struct smu_context *smu, uint16_t msg, uint32_t param); + int (*read_smc_arg)(struct smu_context *smu, uint32_t *arg); + int (*init_display)(struct smu_context *smu); + int (*set_allowed_mask)(struct smu_context *smu); + int (*get_enabled_mask)(struct smu_context *smu, uint32_t *feature_mask, uint32_t num); + bool (*is_dpm_running)(struct smu_context *smu); + int (*update_feature_enable_state)(struct smu_context *smu, uint32_t feature_id, bool enabled); + int (*notify_display_change)(struct smu_context *smu); + int (*get_power_limit)(struct smu_context *smu, uint32_t *limit, bool def); + int (*set_power_limit)(struct smu_context *smu, uint32_t n); + int (*get_current_clk_freq)(struct smu_context *smu, uint32_t clk_id, uint32_t *value); + int (*init_max_sustainable_clocks)(struct smu_context *smu); + int (*start_thermal_control)(struct smu_context *smu); + int (*read_sensor)(struct smu_context *smu, enum amd_pp_sensors sensor, + void *data, uint32_t *size); + int (*set_deep_sleep_dcefclk)(struct smu_context *smu, uint32_t clk); + int (*set_active_display_count)(struct smu_context *smu, uint32_t count); + int (*store_cc6_data)(struct smu_context *smu, uint32_t separation_time, + bool cc6_disable, bool pstate_disable, + bool pstate_switch_disable); + int (*get_clock_by_type)(struct smu_context *smu, + enum amd_pp_clock_type type, + struct amd_pp_clocks *clocks); + int (*get_max_high_clocks)(struct smu_context *smu, + struct amd_pp_simple_clock_info *clocks); + int (*display_clock_voltage_request)(struct smu_context *smu, struct + pp_display_clock_request + *clock_req); + int (*get_dal_power_level)(struct smu_context *smu, + struct amd_pp_simple_clock_info *clocks); + int (*get_perf_level)(struct smu_context *smu, + enum smu_perf_level_designation designation, + struct smu_performance_level *level); + int (*get_current_shallow_sleep_clocks)(struct smu_context *smu, + struct smu_clock_info *clocks); + int (*notify_smu_enable_pwe)(struct smu_context *smu); + int (*set_watermarks_for_clock_ranges)(struct smu_context *smu, + struct dm_pp_wm_sets_with_clock_ranges_soc15 *clock_ranges); + int (*set_od8_default_settings)(struct smu_context *smu, + bool initialize); + int (*conv_power_profile_to_pplib_workload)(int power_profile); + int (*get_power_profile_mode)(struct smu_context *smu, char *buf); + int (*set_power_profile_mode)(struct smu_context *smu, long *input, uint32_t size); + int (*update_od8_settings)(struct smu_context *smu, + uint32_t index, + uint32_t value); + int (*dpm_set_uvd_enable)(struct smu_context *smu, bool enable); + int (*dpm_set_vce_enable)(struct smu_context *smu, bool enable); + uint32_t (*get_sclk)(struct smu_context *smu, bool low); + uint32_t (*get_mclk)(struct smu_context *smu, bool low); + int (*get_current_rpm)(struct smu_context *smu, uint32_t *speed); + uint32_t (*get_fan_control_mode)(struct smu_context *smu); + int (*set_fan_control_mode)(struct smu_context *smu, uint32_t mode); + int (*get_fan_speed_percent)(struct smu_context *smu, uint32_t *speed); + int (*set_fan_speed_percent)(struct smu_context *smu, uint32_t speed); + int (*set_fan_speed_rpm)(struct smu_context *smu, uint32_t speed); + int (*set_xgmi_pstate)(struct smu_context *smu, uint32_t pstate); + +}; + +#define smu_init_microcode(smu) \ + ((smu)->funcs->init_microcode ? (smu)->funcs->init_microcode((smu)) : 0) +#define smu_init_smc_tables(smu) \ + ((smu)->funcs->init_smc_tables ? (smu)->funcs->init_smc_tables((smu)) : 0) +#define smu_fini_smc_tables(smu) \ + ((smu)->funcs->fini_smc_tables ? (smu)->funcs->fini_smc_tables((smu)) : 0) +#define smu_init_power(smu) \ + ((smu)->funcs->init_power ? (smu)->funcs->init_power((smu)) : 0) +#define smu_fini_power(smu) \ + ((smu)->funcs->fini_power ? (smu)->funcs->fini_power((smu)) : 0) +#define smu_load_microcode(smu) \ + ((smu)->funcs->load_microcode ? (smu)->funcs->load_microcode((smu)) : 0) +#define smu_check_fw_status(smu) \ + ((smu)->funcs->check_fw_status ? (smu)->funcs->check_fw_status((smu)) : 0) +#define smu_read_pptable_from_vbios(smu) \ + ((smu)->funcs->read_pptable_from_vbios ? (smu)->funcs->read_pptable_from_vbios((smu)) : 0) +#define smu_get_vbios_bootup_values(smu) \ + ((smu)->funcs->get_vbios_bootup_values ? (smu)->funcs->get_vbios_bootup_values((smu)) : 0) +#define smu_get_clk_info_from_vbios(smu) \ + ((smu)->funcs->get_clk_info_from_vbios ? (smu)->funcs->get_clk_info_from_vbios((smu)) : 0) +#define smu_check_pptable(smu) \ + ((smu)->funcs->check_pptable ? (smu)->funcs->check_pptable((smu)) : 0) +#define smu_parse_pptable(smu) \ + ((smu)->funcs->parse_pptable ? (smu)->funcs->parse_pptable((smu)) : 0) +#define smu_populate_smc_pptable(smu) \ + ((smu)->funcs->populate_smc_pptable ? (smu)->funcs->populate_smc_pptable((smu)) : 0) +#define smu_check_fw_version(smu) \ + ((smu)->funcs->check_fw_version ? (smu)->funcs->check_fw_version((smu)) : 0) +#define smu_write_pptable(smu) \ + ((smu)->funcs->write_pptable ? (smu)->funcs->write_pptable((smu)) : 0) +#define smu_set_min_dcef_deep_sleep(smu) \ + ((smu)->funcs->set_min_dcef_deep_sleep ? (smu)->funcs->set_min_dcef_deep_sleep((smu)) : 0) +#define smu_set_tool_table_location(smu) \ + ((smu)->funcs->set_tool_table_location ? (smu)->funcs->set_tool_table_location((smu)) : 0) +#define smu_notify_memory_pool_location(smu) \ + ((smu)->funcs->notify_memory_pool_location ? (smu)->funcs->notify_memory_pool_location((smu)) : 0) +#define smu_write_watermarks_table(smu) \ + ((smu)->funcs->write_watermarks_table ? (smu)->funcs->write_watermarks_table((smu)) : 0) +#define smu_set_last_dcef_min_deep_sleep_clk(smu) \ + ((smu)->funcs->set_last_dcef_min_deep_sleep_clk ? (smu)->funcs->set_last_dcef_min_deep_sleep_clk((smu)) : 0) +#define smu_system_features_control(smu, en) \ + ((smu)->funcs->system_features_control ? (smu)->funcs->system_features_control((smu), (en)) : 0) +#define smu_init_max_sustainable_clocks(smu) \ + ((smu)->funcs->init_max_sustainable_clocks ? (smu)->funcs->init_max_sustainable_clocks((smu)) : 0) +#define smu_set_od8_default_settings(smu, initialize) \ + ((smu)->funcs->set_od8_default_settings ? (smu)->funcs->set_od8_default_settings((smu), (initialize)) : 0) +#define smu_update_od8_settings(smu, index, value) \ + ((smu)->funcs->update_od8_settings ? (smu)->funcs->update_od8_settings((smu), (index), (value)) : 0) +#define smu_get_current_rpm(smu, speed) \ + ((smu)->funcs->get_current_rpm ? (smu)->funcs->get_current_rpm((smu), (speed)) : 0) +#define smu_set_fan_speed_rpm(smu, speed) \ + ((smu)->funcs->set_fan_speed_rpm ? (smu)->funcs->set_fan_speed_rpm((smu), (speed)) : 0) +#define smu_send_smc_msg(smu, msg) \ + ((smu)->funcs->send_smc_msg? (smu)->funcs->send_smc_msg((smu), (msg)) : 0) +#define smu_send_smc_msg_with_param(smu, msg, param) \ + ((smu)->funcs->send_smc_msg_with_param? (smu)->funcs->send_smc_msg_with_param((smu), (msg), (param)) : 0) +#define smu_read_smc_arg(smu, arg) \ + ((smu)->funcs->read_smc_arg? (smu)->funcs->read_smc_arg((smu), (arg)) : 0) +#define smu_alloc_dpm_context(smu) \ + ((smu)->ppt_funcs->alloc_dpm_context ? (smu)->ppt_funcs->alloc_dpm_context((smu)) : 0) +#define smu_init_display(smu) \ + ((smu)->funcs->init_display ? (smu)->funcs->init_display((smu)) : 0) +#define smu_feature_set_allowed_mask(smu) \ + ((smu)->funcs->set_allowed_mask? (smu)->funcs->set_allowed_mask((smu)) : 0) +#define smu_feature_get_enabled_mask(smu, mask, num) \ + ((smu)->funcs->get_enabled_mask? (smu)->funcs->get_enabled_mask((smu), (mask), (num)) : 0) +#define smu_is_dpm_running(smu) \ + ((smu)->funcs->is_dpm_running ? (smu)->funcs->is_dpm_running((smu)) : 0) +#define smu_feature_update_enable_state(smu, feature_id, enabled) \ + ((smu)->funcs->update_feature_enable_state? (smu)->funcs->update_feature_enable_state((smu), (feature_id), (enabled)) : 0) +#define smu_notify_display_change(smu) \ + ((smu)->funcs->notify_display_change? (smu)->funcs->notify_display_change((smu)) : 0) +#define smu_store_powerplay_table(smu) \ + ((smu)->ppt_funcs->store_powerplay_table ? (smu)->ppt_funcs->store_powerplay_table((smu)) : 0) +#define smu_check_powerplay_table(smu) \ + ((smu)->ppt_funcs->check_powerplay_table ? (smu)->ppt_funcs->check_powerplay_table((smu)) : 0) +#define smu_append_powerplay_table(smu) \ + ((smu)->ppt_funcs->append_powerplay_table ? (smu)->ppt_funcs->append_powerplay_table((smu)) : 0) +#define smu_set_default_dpm_table(smu) \ + ((smu)->ppt_funcs->set_default_dpm_table ? (smu)->ppt_funcs->set_default_dpm_table((smu)) : 0) +#define smu_populate_umd_state_clk(smu) \ + ((smu)->ppt_funcs->populate_umd_state_clk ? (smu)->ppt_funcs->populate_umd_state_clk((smu)) : 0) +#define smu_set_default_od8_settings(smu) \ + ((smu)->ppt_funcs->set_default_od8_settings ? (smu)->ppt_funcs->set_default_od8_settings((smu)) : 0) +#define smu_update_specified_od8_value(smu, index, value) \ + ((smu)->ppt_funcs->update_specified_od8_value ? (smu)->ppt_funcs->update_specified_od8_value((smu), (index), (value)) : 0) +#define smu_get_power_limit(smu, limit, def) \ + ((smu)->funcs->get_power_limit ? (smu)->funcs->get_power_limit((smu), (limit), (def)) : 0) +#define smu_set_power_limit(smu, limit) \ + ((smu)->funcs->set_power_limit ? (smu)->funcs->set_power_limit((smu), (limit)) : 0) +#define smu_get_current_clk_freq(smu, clk_id, value) \ + ((smu)->funcs->get_current_clk_freq? (smu)->funcs->get_current_clk_freq((smu), (clk_id), (value)) : 0) +#define smu_print_clk_levels(smu, type, buf) \ + ((smu)->ppt_funcs->print_clk_levels ? (smu)->ppt_funcs->print_clk_levels((smu), (type), (buf)) : 0) +#define smu_force_clk_levels(smu, type, level) \ + ((smu)->ppt_funcs->force_clk_levels ? (smu)->ppt_funcs->force_clk_levels((smu), (type), (level)) : 0) +#define smu_get_od_percentage(smu, type) \ + ((smu)->ppt_funcs->get_od_percentage ? (smu)->ppt_funcs->get_od_percentage((smu), (type)) : 0) +#define smu_set_od_percentage(smu, type, value) \ + ((smu)->ppt_funcs->set_od_percentage ? (smu)->ppt_funcs->set_od_percentage((smu), (type), (value)) : 0) +#define smu_od_edit_dpm_table(smu, type, input, size) \ + ((smu)->ppt_funcs->od_edit_dpm_table ? (smu)->ppt_funcs->od_edit_dpm_table((smu), (type), (input), (size)) : 0) +#define smu_start_thermal_control(smu) \ + ((smu)->funcs->start_thermal_control? (smu)->funcs->start_thermal_control((smu)) : 0) +#define smu_read_sensor(smu, sensor, data, size) \ + ((smu)->funcs->read_sensor? (smu)->funcs->read_sensor((smu), (sensor), (data), (size)) : 0) +#define smu_get_power_profile_mode(smu, buf) \ + ((smu)->funcs->get_power_profile_mode ? (smu)->funcs->get_power_profile_mode((smu), buf) : 0) +#define smu_set_power_profile_mode(smu, param, param_size) \ + ((smu)->funcs->set_power_profile_mode ? (smu)->funcs->set_power_profile_mode((smu), (param), (param_size)) : 0) +#define smu_get_performance_level(smu) \ + ((smu)->ppt_funcs->get_performance_level ? (smu)->ppt_funcs->get_performance_level((smu)) : 0) +#define smu_force_performance_level(smu, level) \ + ((smu)->ppt_funcs->force_performance_level ? (smu)->ppt_funcs->force_performance_level((smu), (level)) : 0) +#define smu_pre_display_config_changed(smu) \ + ((smu)->ppt_funcs->pre_display_config_changed ? (smu)->ppt_funcs->pre_display_config_changed((smu)) : 0) +#define smu_display_config_changed(smu) \ + ((smu)->ppt_funcs->display_config_changed ? (smu)->ppt_funcs->display_config_changed((smu)) : 0) +#define smu_apply_clocks_adjust_rules(smu) \ + ((smu)->ppt_funcs->apply_clocks_adjust_rules ? (smu)->ppt_funcs->apply_clocks_adjust_rules((smu)) : 0) +#define smu_notify_smc_dispaly_config(smu) \ + ((smu)->ppt_funcs->notify_smc_dispaly_config ? (smu)->ppt_funcs->notify_smc_dispaly_config((smu)) : 0) +#define smu_force_dpm_limit_value(smu, highest) \ + ((smu)->ppt_funcs->force_dpm_limit_value ? (smu)->ppt_funcs->force_dpm_limit_value((smu), (highest)) : 0) +#define smu_unforce_dpm_levels(smu) \ + ((smu)->ppt_funcs->unforce_dpm_levels ? (smu)->ppt_funcs->unforce_dpm_levels((smu)) : 0) +#define smu_upload_dpm_level(smu, max, feature_mask) \ + ((smu)->ppt_funcs->upload_dpm_level ? (smu)->ppt_funcs->upload_dpm_level((smu), (max), (feature_mask)) : 0) +#define smu_get_profiling_clk_mask(smu, level, sclk_mask, mclk_mask, soc_mask) \ + ((smu)->ppt_funcs->get_profiling_clk_mask ? (smu)->ppt_funcs->get_profiling_clk_mask((smu), (level), (sclk_mask), (mclk_mask), (soc_mask)) : 0) +#define smu_set_cpu_power_state(smu) \ + ((smu)->ppt_funcs->set_cpu_power_state ? (smu)->ppt_funcs->set_cpu_power_state((smu)) : 0) +#define smu_get_fan_control_mode(smu) \ + ((smu)->funcs->get_fan_control_mode ? (smu)->funcs->get_fan_control_mode((smu)) : 0) +#define smu_set_fan_control_mode(smu, value) \ + ((smu)->funcs->set_fan_control_mode ? (smu)->funcs->set_fan_control_mode((smu), (value)) : 0) +#define smu_get_fan_speed_percent(smu, speed) \ + ((smu)->funcs->get_fan_speed_percent ? (smu)->funcs->get_fan_speed_percent((smu), (speed)) : 0) +#define smu_set_fan_speed_percent(smu, speed) \ + ((smu)->funcs->set_fan_speed_percent ? (smu)->funcs->set_fan_speed_percent((smu), (speed)) : 0) + +#define smu_msg_get_index(smu, msg) \ + ((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_msg_index? (smu)->ppt_funcs->get_smu_msg_index((smu), (msg)) : -EINVAL) : -EINVAL) +#define smu_run_afll_btc(smu) \ + ((smu)->ppt_funcs? ((smu)->ppt_funcs->run_afll_btc? (smu)->ppt_funcs->run_afll_btc((smu)) : 0) : 0) +#define smu_get_unallowed_feature_mask(smu, feature_mask, num) \ + ((smu)->ppt_funcs? ((smu)->ppt_funcs->get_unallowed_feature_mask? (smu)->ppt_funcs->get_unallowed_feature_mask((smu), (feature_mask), (num)) : 0) : 0) +#define smu_set_deep_sleep_dcefclk(smu, clk) \ + ((smu)->funcs->set_deep_sleep_dcefclk ? (smu)->funcs->set_deep_sleep_dcefclk((smu), (clk)) : 0) +#define smu_set_active_display_count(smu, count) \ + ((smu)->funcs->set_active_display_count ? (smu)->funcs->set_active_display_count((smu), (count)) : 0) +#define smu_store_cc6_data(smu, st, cc6_dis, pst_dis, pst_sw_dis) \ + ((smu)->funcs->store_cc6_data ? (smu)->funcs->store_cc6_data((smu), (st), (cc6_dis), (pst_dis), (pst_sw_dis)) : 0) +#define smu_get_clock_by_type(smu, type, clocks) \ + ((smu)->funcs->get_clock_by_type ? (smu)->funcs->get_clock_by_type((smu), (type), (clocks)) : 0) +#define smu_get_max_high_clocks(smu, clocks) \ + ((smu)->funcs->get_max_high_clocks ? (smu)->funcs->get_max_high_clocks((smu), (clocks)) : 0) +#define smu_get_clock_by_type_with_latency(smu, type, clocks) \ + ((smu)->ppt_funcs->get_clock_by_type_with_latency ? (smu)->ppt_funcs->get_clock_by_type_with_latency((smu), (type), (clocks)) : 0) +#define smu_get_clock_by_type_with_voltage(smu, type, clocks) \ + ((smu)->ppt_funcs->get_clock_by_type_with_voltage ? (smu)->ppt_funcs->get_clock_by_type_with_voltage((smu), (type), (clocks)) : 0) +#define smu_display_clock_voltage_request(smu, clock_req) \ + ((smu)->funcs->display_clock_voltage_request ? (smu)->funcs->display_clock_voltage_request((smu), (clock_req)) : 0) +#define smu_get_dal_power_level(smu, clocks) \ + ((smu)->funcs->get_dal_power_level ? (smu)->funcs->get_dal_power_level((smu), (clocks)) : 0) +#define smu_get_perf_level(smu, designation, level) \ + ((smu)->funcs->get_perf_level ? (smu)->funcs->get_perf_level((smu), (designation), (level)) : 0) +#define smu_get_current_shallow_sleep_clocks(smu, clocks) \ + ((smu)->funcs->get_current_shallow_sleep_clocks ? (smu)->funcs->get_current_shallow_sleep_clocks((smu), (clocks)) : 0) +#define smu_notify_smu_enable_pwe(smu) \ + ((smu)->funcs->notify_smu_enable_pwe ? (smu)->funcs->notify_smu_enable_pwe((smu)) : 0) +#define smu_set_watermarks_for_clock_ranges(smu, clock_ranges) \ + ((smu)->funcs->set_watermarks_for_clock_ranges ? (smu)->funcs->set_watermarks_for_clock_ranges((smu), (clock_ranges)) : 0) +#define smu_dpm_set_uvd_enable(smu, enable) \ + ((smu)->funcs->dpm_set_uvd_enable ? (smu)->funcs->dpm_set_uvd_enable((smu), (enable)) : 0) +#define smu_dpm_set_vce_enable(smu, enable) \ + ((smu)->funcs->dpm_set_vce_enable ? (smu)->funcs->dpm_set_vce_enable((smu), (enable)) : 0) +#define smu_get_sclk(smu, low) \ + ((smu)->funcs->get_sclk ? (smu)->funcs->get_sclk((smu), (low)) : 0) +#define smu_get_mclk(smu, low) \ + ((smu)->funcs->get_mclk ? (smu)->funcs->get_mclk((smu), (low)) : 0) +#define smu_set_xgmi_pstate(smu, pstate) \ + ((smu)->funcs->set_xgmi_pstate ? (smu)->funcs->set_xgmi_pstate((smu), (pstate)) : 0) + + +extern int smu_get_atom_data_table(struct smu_context *smu, uint32_t table, + uint16_t *size, uint8_t *frev, uint8_t *crev, + uint8_t **addr); + +extern const struct amd_ip_funcs smu_ip_funcs; + +extern const struct amdgpu_ip_block_version smu_v11_0_ip_block; +extern int smu_feature_init_dpm(struct smu_context *smu); + +extern int smu_feature_is_enabled(struct smu_context *smu, int feature_id); +extern int smu_feature_set_enabled(struct smu_context *smu, int feature_id, bool enable); +extern int smu_feature_is_supported(struct smu_context *smu, int feature_id); +extern int smu_feature_set_supported(struct smu_context *smu, int feature_id, bool enable); + +int smu_update_table_with_arg(struct smu_context *smu, uint16_t table_id, uint16_t exarg, + void *table_data, bool drv2smu); +#define smu_update_table(smu, table_id, table_data, drv2smu) \ + smu_update_table_with_arg((smu), (table_id), 0, (table_data), (drv2smu)) + +bool is_support_sw_smu(struct amdgpu_device *adev); +int smu_reset(struct smu_context *smu); +int smu_common_read_sensor(struct smu_context *smu, enum amd_pp_sensors sensor, + void *data, uint32_t *size); +int smu_sys_get_pp_table(struct smu_context *smu, void **table); +int smu_sys_set_pp_table(struct smu_context *smu, void *buf, size_t size); +int smu_get_power_num_states(struct smu_context *smu, struct pp_states_info *state_info); +enum amd_pm_state_type smu_get_current_power_state(struct smu_context *smu); + +/* smu to display interface */ +extern int smu_display_configuration_change(struct smu_context *smu, const + struct amd_pp_display_configuration + *display_config); +extern int smu_get_current_clocks(struct smu_context *smu, + struct amd_pp_clock_info *clocks); +extern int smu_dpm_set_power_gate(struct smu_context *smu,uint32_t block_type, bool gate); +extern int smu_handle_task(struct smu_context *smu, + enum amd_dpm_forced_level level, + enum amd_pp_task task_id); +#endif diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if.h b/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if.h index 63d5cf691549..195c4ae67058 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if.h @@ -99,7 +99,7 @@ #define FEATURE_DS_MP1CLK_BIT 30 #define FEATURE_DS_MP0CLK_BIT 31 #define FEATURE_XGMI_BIT 32 -#define FEATURE_SPARE_33_BIT 33 +#define FEATURE_ECC_BIT 33 #define FEATURE_SPARE_34_BIT 34 #define FEATURE_SPARE_35_BIT 35 #define FEATURE_SPARE_36_BIT 36 @@ -165,7 +165,8 @@ #define FEATURE_DS_FCLK_MASK (1 << FEATURE_DS_FCLK_BIT ) #define FEATURE_DS_MP1CLK_MASK (1 << FEATURE_DS_MP1CLK_BIT ) #define FEATURE_DS_MP0CLK_MASK (1 << FEATURE_DS_MP0CLK_BIT ) -#define FEATURE_XGMI_MASK (1 << FEATURE_XGMI_BIT ) +#define FEATURE_XGMI_MASK (1ULL << FEATURE_XGMI_BIT ) +#define FEATURE_ECC_MASK (1ULL << FEATURE_ECC_BIT ) #define DPM_OVERRIDE_DISABLE_SOCCLK_PID 0x00000001 #define DPM_OVERRIDE_DISABLE_UCLK_PID 0x00000002 diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h new file mode 100644 index 000000000000..aa8d81f4111e --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h @@ -0,0 +1,89 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __SMU_V11_0_H__ +#define __SMU_V11_0_H__ + +#include "amdgpu_smu.h" + +/* MP Apertures */ +#define MP0_Public 0x03800000 +#define MP0_SRAM 0x03900000 +#define MP1_Public 0x03b00000 +#define MP1_SRAM 0x03c00004 + +/* address block */ +#define smnMP1_FIRMWARE_FLAGS 0x3010024 +#define smnMP0_FW_INTF 0x30101c0 +#define smnMP1_PUB_CTRL 0x3010b14 + +struct smu_11_0_max_sustainable_clocks { + uint32_t display_clock; + uint32_t phy_clock; + uint32_t pixel_clock; + uint32_t uclock; + uint32_t dcef_clock; + uint32_t soc_clock; +}; + +struct smu_11_0_dpm_table { + uint32_t min; /* MHz */ + uint32_t max; /* MHz */ +}; + +struct smu_11_0_dpm_tables { + struct smu_11_0_dpm_table soc_table; + struct smu_11_0_dpm_table gfx_table; + struct smu_11_0_dpm_table uclk_table; + struct smu_11_0_dpm_table eclk_table; + struct smu_11_0_dpm_table vclk_table; + struct smu_11_0_dpm_table dclk_table; + struct smu_11_0_dpm_table dcef_table; + struct smu_11_0_dpm_table pixel_table; + struct smu_11_0_dpm_table display_table; + struct smu_11_0_dpm_table phy_table; + struct smu_11_0_dpm_table fclk_table; +}; + +struct smu_11_0_dpm_context { + struct smu_11_0_dpm_tables dpm_tables; + uint32_t workload_policy_mask; + uint32_t dcef_min_ds_clk; +}; + +enum smu_11_0_power_state { + SMU_11_0_POWER_STATE__D0 = 0, + SMU_11_0_POWER_STATE__D1, + SMU_11_0_POWER_STATE__D3, /* Sleep*/ + SMU_11_0_POWER_STATE__D4, /* Hibernate*/ + SMU_11_0_POWER_STATE__D5, /* Power off*/ +}; + +struct smu_11_0_power_context { + uint32_t power_source; + uint8_t in_power_limit_boost_mode; + enum smu_11_0_power_state power_state; +}; + +void smu_v11_0_set_smu_funcs(struct smu_context *smu); + +#endif diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_ppsmc.h new file mode 100644 index 000000000000..f466f624ad32 --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_ppsmc.h @@ -0,0 +1,128 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef SMU_V11_0_PPSMC_H +#define SMU_V11_0_PPSMC_H + +// SMU Response Codes: +#define PPSMC_Result_OK 0x1 +#define PPSMC_Result_Failed 0xFF +#define PPSMC_Result_UnknownCmd 0xFE +#define PPSMC_Result_CmdRejectedPrereq 0xFD +#define PPSMC_Result_CmdRejectedBusy 0xFC + +// Message Definitions: +// BASIC +#define PPSMC_MSG_TestMessage 0x1 +#define PPSMC_MSG_GetSmuVersion 0x2 +#define PPSMC_MSG_GetDriverIfVersion 0x3 +#define PPSMC_MSG_SetAllowedFeaturesMaskLow 0x4 +#define PPSMC_MSG_SetAllowedFeaturesMaskHigh 0x5 +#define PPSMC_MSG_EnableAllSmuFeatures 0x6 +#define PPSMC_MSG_DisableAllSmuFeatures 0x7 +#define PPSMC_MSG_EnableSmuFeaturesLow 0x8 +#define PPSMC_MSG_EnableSmuFeaturesHigh 0x9 +#define PPSMC_MSG_DisableSmuFeaturesLow 0xA +#define PPSMC_MSG_DisableSmuFeaturesHigh 0xB +#define PPSMC_MSG_GetEnabledSmuFeaturesLow 0xC +#define PPSMC_MSG_GetEnabledSmuFeaturesHigh 0xD +#define PPSMC_MSG_SetDriverDramAddrHigh 0xE +#define PPSMC_MSG_SetDriverDramAddrLow 0xF +#define PPSMC_MSG_SetToolsDramAddrHigh 0x10 +#define PPSMC_MSG_SetToolsDramAddrLow 0x11 +#define PPSMC_MSG_TransferTableSmu2Dram 0x12 +#define PPSMC_MSG_TransferTableDram2Smu 0x13 +#define PPSMC_MSG_UseDefaultPPTable 0x14 +#define PPSMC_MSG_UseBackupPPTable 0x15 +#define PPSMC_MSG_SetSystemVirtualDramAddrHigh 0x16 +#define PPSMC_MSG_SetSystemVirtualDramAddrLow 0x17 + +//BACO/BAMACO/BOMACO +#define PPSMC_MSG_EnterBaco 0x18 +#define PPSMC_MSG_ExitBaco 0x19 + +//DPM +#define PPSMC_MSG_SetSoftMinByFreq 0x1A +#define PPSMC_MSG_SetSoftMaxByFreq 0x1B +#define PPSMC_MSG_SetHardMinByFreq 0x1C +#define PPSMC_MSG_SetHardMaxByFreq 0x1D +#define PPSMC_MSG_GetMinDpmFreq 0x1E +#define PPSMC_MSG_GetMaxDpmFreq 0x1F +#define PPSMC_MSG_GetDpmFreqByIndex 0x20 +#define PPSMC_MSG_OverridePcieParameters 0x21 +#define PPSMC_MSG_SetMinDeepSleepDcefclk 0x22 +#define PPSMC_MSG_SetWorkloadMask 0x23 +#define PPSMC_MSG_SetUclkFastSwitch 0x24 +#define PPSMC_MSG_GetAvfsVoltageByDpm 0x25 +#define PPSMC_MSG_SetVideoFps 0x26 +#define PPSMC_MSG_GetDcModeMaxDpmFreq 0x27 + +//Power Gating +#define PPSMC_MSG_AllowGfxOff 0x28 +#define PPSMC_MSG_DisallowGfxOff 0x29 +#define PPSMC_MSG_PowerUpVcn 0x2A +#define PPSMC_MSG_PowerDownVcn 0x2B +#define PPSMC_MSG_PowerUpJpeg 0x2C +#define PPSMC_MSG_PowerDownJpeg 0x2D +//reserve 0x2A to 0x2F for PG harvesting TBD + +//I2C Interface +#define PPSMC_RequestI2cTransaction 0x30 + +//Resets +#define PPSMC_MSG_SoftReset 0x31 //FIXME Need confirmation from driver +#define PPSMC_MSG_PrepareMp1ForUnload 0x32 +#define PPSMC_MSG_PrepareMp1ForReset 0x33 +#define PPSMC_MSG_PrepareMp1ForShutdown 0x34 + +//ACDC Power Source +#define PPSMC_MSG_SetPptLimit 0x35 +#define PPSMC_MSG_GetPptLimit 0x36 +#define PPSMC_MSG_ReenableAcDcInterrupt 0x37 +#define PPSMC_MSG_NotifyPowerSource 0x38 +//#define PPSMC_MSG_GfxDeviceDriverReset 0x39 //FIXME mode1 and 2 resets will go directly go PSP + +//BTC +#define PPSMC_MSG_RunBtc 0x3A + +//Debug +#define PPSMC_MSG_DramLogSetDramAddrHigh 0x3B +#define PPSMC_MSG_DramLogSetDramAddrLow 0x3C +#define PPSMC_MSG_DramLogSetDramSize 0x3D +#define PPSMC_MSG_GetDebugData 0x3E + +//Others +#define PPSMC_MSG_ConfigureGfxDidt 0x3F +#define PPSMC_MSG_NumOfDisplays 0x40 + +#define PPSMC_MSG_SetMemoryChannelConfig 0x41 +#define PPSMC_MSG_SetGeminiMode 0x42 +#define PPSMC_MSG_SetGeminiApertureHigh 0x43 +#define PPSMC_MSG_SetGeminiApertureLow 0x44 + +#define PPSMC_Message_Count 0x45 + +typedef uint32_t PPSMC_Result; +typedef uint32_t PPSMC_Msg; + +#endif diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h new file mode 100644 index 000000000000..92c65b80bde2 --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h @@ -0,0 +1,147 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ +#ifndef SMU_11_0_PPTABLE_H +#define SMU_11_0_PPTABLE_H + + +#define SMU_11_0_TABLE_FORMAT_REVISION 12 + +//// POWERPLAYTABLE::ulPlatformCaps +#define SMU_11_0_PP_PLATFORM_CAP_POWERPLAY 0x1 +#define SMU_11_0_PP_PLATFORM_CAP_SBIOSPOWERSOURCE 0x2 +#define SMU_11_0_PP_PLATFORM_CAP_HARDWAREDC 0x4 +#define SMU_11_0_PP_PLATFORM_CAP_BACO 0x8 +#define SMU_11_0_PP_PLATFORM_CAP_MACO 0x10 +#define SMU_11_0_PP_PLATFORM_CAP_SHADOWPSTATE 0x20 + +// SMU_11_0_PP_THERMALCONTROLLER - Thermal Controller Type +#define SMU_11_0_PP_THERMALCONTROLLER_NONE 0 + +#define SMU_11_0_PP_OVERDRIVE_VERSION 0x0800 +#define SMU_11_0_PP_POWERSAVINGCLOCK_VERSION 0x0100 + +enum SMU_11_0_ODFEATURE_ID { + SMU_11_0_ODFEATURE_GFXCLK_LIMITS = 1 << 0, //GFXCLK Limit feature + SMU_11_0_ODFEATURE_GFXCLK_CURVE = 1 << 1, //GFXCLK Curve feature + SMU_11_0_ODFEATURE_UCLK_MAX = 1 << 2, //UCLK Limit feature + SMU_11_0_ODFEATURE_POWER_LIMIT = 1 << 3, //Power Limit feature + SMU_11_0_ODFEATURE_FAN_ACOUSTIC_LIMIT = 1 << 4, //Fan Acoustic RPM feature + SMU_11_0_ODFEATURE_FAN_SPEED_MIN = 1 << 5, //Minimum Fan Speed feature + SMU_11_0_ODFEATURE_TEMPERATURE_FAN = 1 << 6, //Fan Target Temperature Limit feature + SMU_11_0_ODFEATURE_TEMPERATURE_SYSTEM = 1 << 7, //Operating Temperature Limit feature + SMU_11_0_ODFEATURE_MEMORY_TIMING_TUNE = 1 << 8, //AC Timing Tuning feature + SMU_11_0_ODFEATURE_FAN_ZERO_RPM_CONTROL = 1 << 9, //Zero RPM feature + SMU_11_0_ODFEATURE_AUTO_UV_ENGINE = 1 << 10, //Auto Under Volt GFXCLK feature + SMU_11_0_ODFEATURE_AUTO_OC_ENGINE = 1 << 11, //Auto Over Clock GFXCLK feature + SMU_11_0_ODFEATURE_AUTO_OC_MEMORY = 1 << 12, //Auto Over Clock MCLK feature + SMU_11_0_ODFEATURE_FAN_CURVE = 1 << 13, //VICTOR TODO + SMU_11_0_ODFEATURE_COUNT = 14, +}; +#define SMU_11_0_MAX_ODFEATURE 32 //Maximum Number of OD Features + +enum SMU_11_0_ODSETTING_ID { + SMU_11_0_ODSETTING_GFXCLKFMAX = 0, + SMU_11_0_ODSETTING_GFXCLKFMIN, + SMU_11_0_ODSETTING_VDDGFXCURVEFREQ_P1, + SMU_11_0_ODSETTING_VDDGFXCURVEVOLTAGE_P1, + SMU_11_0_ODSETTING_VDDGFXCURVEFREQ_P2, + SMU_11_0_ODSETTING_VDDGFXCURVEVOLTAGE_P2, + SMU_11_0_ODSETTING_VDDGFXCURVEFREQ_P3, + SMU_11_0_ODSETTING_VDDGFXCURVEVOLTAGE_P3, + SMU_11_0_ODSETTING_UCLKFMAX, + SMU_11_0_ODSETTING_POWERPERCENTAGE, + SMU_11_0_ODSETTING_FANRPMMIN, + SMU_11_0_ODSETTING_FANRPMACOUSTICLIMIT, + SMU_11_0_ODSETTING_FANTARGETTEMPERATURE, + SMU_11_0_ODSETTING_OPERATINGTEMPMAX, + SMU_11_0_ODSETTING_ACTIMING, + SMU_11_0_ODSETTING_FAN_ZERO_RPM_CONTROL, + SMU_11_0_ODSETTING_AUTOUVENGINE, + SMU_11_0_ODSETTING_AUTOOCENGINE, + SMU_11_0_ODSETTING_AUTOOCMEMORY, + SMU_11_0_ODSETTING_COUNT, +}; +#define SMU_11_0_MAX_ODSETTING 32 //Maximum Number of ODSettings + +struct smu_11_0_overdrive_table +{ + uint8_t revision; //Revision = SMU_11_0_PP_OVERDRIVE_VERSION + uint8_t reserve[3]; //Zero filled field reserved for future use + uint32_t feature_count; //Total number of supported features + uint32_t setting_count; //Total number of supported settings + uint8_t cap[SMU_11_0_MAX_ODFEATURE]; //OD feature support flags + uint32_t max[SMU_11_0_MAX_ODSETTING]; //default maximum settings + uint32_t min[SMU_11_0_MAX_ODSETTING]; //default minimum settings +} __attribute__((packed)); + +enum SMU_11_0_PPCLOCK_ID { + SMU_11_0_PPCLOCK_GFXCLK = 0, + SMU_11_0_PPCLOCK_VCLK, + SMU_11_0_PPCLOCK_DCLK, + SMU_11_0_PPCLOCK_ECLK, + SMU_11_0_PPCLOCK_SOCCLK, + SMU_11_0_PPCLOCK_UCLK, + SMU_11_0_PPCLOCK_DCEFCLK, + SMU_11_0_PPCLOCK_DISPCLK, + SMU_11_0_PPCLOCK_PIXCLK, + SMU_11_0_PPCLOCK_PHYCLK, + SMU_11_0_PPCLOCK_COUNT, +}; +#define SMU_11_0_MAX_PPCLOCK 16 //Maximum Number of PP Clocks + +struct smu_11_0_power_saving_clock_table +{ + uint8_t revision; //Revision = SMU_11_0_PP_POWERSAVINGCLOCK_VERSION + uint8_t reserve[3]; //Zero filled field reserved for future use + uint32_t count; //power_saving_clock_count = SMU_11_0_PPCLOCK_COUNT + uint32_t max[SMU_11_0_MAX_PPCLOCK]; //PowerSavingClock Mode Clock Maximum array In MHz + uint32_t min[SMU_11_0_MAX_PPCLOCK]; //PowerSavingClock Mode Clock Minimum array In MHz +} __attribute__((packed)); + +struct smu_11_0_powerplay_table +{ + struct atom_common_table_header header; + uint8_t table_revision; + uint32_t table_size; //Driver portion table size. The offset to smc_pptable including header size + uint32_t golden_pp_id; + uint32_t golden_revision; + uint16_t format_id; + uint32_t platform_caps; //POWERPLAYABLE::ulPlatformCaps + + uint8_t thermal_controller_type; //one of SMU_11_0_PP_THERMALCONTROLLER + + uint16_t small_power_limit1; + uint16_t small_power_limit2; + uint16_t boost_power_limit; + uint16_t od_turbo_power_limit; //Power limit setting for Turbo mode in Performance UI Tuning. + uint16_t od_power_save_power_limit; //Power limit setting for PowerSave/Optimal mode in Performance UI Tuning. + uint16_t software_shutdown_temp; + + uint16_t reserve[6]; //Zero filled field reserved for future use + + struct smu_11_0_power_saving_clock_table power_saving_clock; + struct smu_11_0_overdrive_table overdrive_table; + + PPTable_t smc_pptable; //PPTable_t in smu11_driver_if.h +} __attribute__((packed)); + +#endif diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega20_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/vega20_ppsmc.h index 4f63a736ea0e..a0883038f3c3 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/vega20_ppsmc.h +++ b/drivers/gpu/drm/amd/powerplay/inc/vega20_ppsmc.h @@ -119,7 +119,8 @@ #define PPSMC_MSG_PrepareMp1ForShutdown 0x5A #define PPSMC_MSG_SetMGpuFanBoostLimitRpm 0x5D #define PPSMC_MSG_GetAVFSVoltageByDpm 0x5F -#define PPSMC_Message_Count 0x60 +#define PPSMC_MSG_BacoWorkAroundFlushVDCI 0x60 +#define PPSMC_Message_Count 0x61 typedef uint32_t PPSMC_Result; typedef uint32_t PPSMC_Msg; diff --git a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c new file mode 100644 index 000000000000..c478b38662d0 --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c @@ -0,0 +1,1972 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#include "pp_debug.h" +#include +#include "amdgpu.h" +#include "amdgpu_smu.h" +#include "atomfirmware.h" +#include "amdgpu_atomfirmware.h" +#include "smu_v11_0.h" +#include "smu11_driver_if.h" +#include "soc15_common.h" +#include "atom.h" +#include "vega20_ppt.h" +#include "pp_thermal.h" + +#include "asic_reg/thm/thm_11_0_2_offset.h" +#include "asic_reg/thm/thm_11_0_2_sh_mask.h" +#include "asic_reg/mp/mp_9_0_offset.h" +#include "asic_reg/mp/mp_9_0_sh_mask.h" +#include "asic_reg/nbio/nbio_7_4_offset.h" +#include "asic_reg/smuio/smuio_9_0_offset.h" +#include "asic_reg/smuio/smuio_9_0_sh_mask.h" + +MODULE_FIRMWARE("amdgpu/vega20_smc.bin"); + +#define SMU11_TOOL_SIZE 0x19000 +#define SMU11_THERMAL_MINIMUM_ALERT_TEMP 0 +#define SMU11_THERMAL_MAXIMUM_ALERT_TEMP 255 + +#define SMU11_TEMPERATURE_UNITS_PER_CENTIGRADES 1000 +#define SMU11_VOLTAGE_SCALE 4 + +#define SMC_DPM_FEATURE (FEATURE_DPM_PREFETCHER_MASK | \ + FEATURE_DPM_GFXCLK_MASK | \ + FEATURE_DPM_UCLK_MASK | \ + FEATURE_DPM_SOCCLK_MASK | \ + FEATURE_DPM_UVD_MASK | \ + FEATURE_DPM_VCE_MASK | \ + FEATURE_DPM_MP0CLK_MASK | \ + FEATURE_DPM_LINK_MASK | \ + FEATURE_DPM_DCEFCLK_MASK) + +static int smu_v11_0_send_msg_without_waiting(struct smu_context *smu, + uint16_t msg) +{ + struct amdgpu_device *adev = smu->adev; + WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, msg); + return 0; +} + +static int smu_v11_0_read_arg(struct smu_context *smu, uint32_t *arg) +{ + struct amdgpu_device *adev = smu->adev; + + *arg = RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82); + return 0; +} + +static int smu_v11_0_wait_for_response(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + uint32_t cur_value, i; + + for (i = 0; i < adev->usec_timeout; i++) { + cur_value = RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90); + if ((cur_value & MP1_C2PMSG_90__CONTENT_MASK) != 0) + break; + udelay(1); + } + + /* timeout means wrong logic */ + if (i == adev->usec_timeout) + return -ETIME; + + return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90) == 0x1 ? 0 : -EIO; +} + +static int smu_v11_0_send_msg(struct smu_context *smu, uint16_t msg) +{ + struct amdgpu_device *adev = smu->adev; + int ret = 0, index = 0; + + index = smu_msg_get_index(smu, msg); + if (index < 0) + return index; + + smu_v11_0_wait_for_response(smu); + + WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0); + + smu_v11_0_send_msg_without_waiting(smu, (uint16_t)index); + + ret = smu_v11_0_wait_for_response(smu); + + if (ret) + pr_err("Failed to send message 0x%x, response 0x%x\n", index, + ret); + + return ret; + +} + +static int +smu_v11_0_send_msg_with_param(struct smu_context *smu, uint16_t msg, + uint32_t param) +{ + + struct amdgpu_device *adev = smu->adev; + int ret = 0, index = 0; + + index = smu_msg_get_index(smu, msg); + if (index < 0) + return index; + + ret = smu_v11_0_wait_for_response(smu); + if (ret) + pr_err("Failed to send message 0x%x, response 0x%x, param 0x%x\n", + index, ret, param); + + WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0); + + WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82, param); + + smu_v11_0_send_msg_without_waiting(smu, (uint16_t)index); + + ret = smu_v11_0_wait_for_response(smu); + if (ret) + pr_err("Failed to send message 0x%x, response 0x%x param 0x%x\n", + index, ret, param); + + return ret; +} + +static int smu_v11_0_init_microcode(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + const char *chip_name; + char fw_name[30]; + int err = 0; + const struct smc_firmware_header_v1_0 *hdr; + const struct common_firmware_header *header; + struct amdgpu_firmware_info *ucode = NULL; + + switch (adev->asic_type) { + case CHIP_VEGA20: + chip_name = "vega20"; + break; + default: + BUG(); + } + + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_smc.bin", chip_name); + + err = request_firmware(&adev->pm.fw, fw_name, adev->dev); + if (err) + goto out; + err = amdgpu_ucode_validate(adev->pm.fw); + if (err) + goto out; + + hdr = (const struct smc_firmware_header_v1_0 *) adev->pm.fw->data; + amdgpu_ucode_print_smc_hdr(&hdr->header); + adev->pm.fw_version = le32_to_cpu(hdr->header.ucode_version); + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + ucode = &adev->firmware.ucode[AMDGPU_UCODE_ID_SMC]; + ucode->ucode_id = AMDGPU_UCODE_ID_SMC; + ucode->fw = adev->pm.fw; + header = (const struct common_firmware_header *)ucode->fw->data; + adev->firmware.fw_size += + ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); + } + +out: + if (err) { + DRM_ERROR("smu_v11_0: Failed to load firmware \"%s\"\n", + fw_name); + release_firmware(adev->pm.fw); + adev->pm.fw = NULL; + } + return err; +} + +static int smu_v11_0_load_microcode(struct smu_context *smu) +{ + return 0; +} + +static int smu_v11_0_check_fw_status(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + uint32_t mp1_fw_flags; + + mp1_fw_flags = RREG32_PCIE(MP1_Public | + (smnMP1_FIRMWARE_FLAGS & 0xffffffff)); + + if ((mp1_fw_flags & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) >> + MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED__SHIFT) + return 0; + + return -EIO; +} + +static int smu_v11_0_check_fw_version(struct smu_context *smu) +{ + uint32_t smu_version = 0xff; + int ret = 0; + + ret = smu_send_smc_msg(smu, SMU_MSG_GetDriverIfVersion); + if (ret) + goto err; + + ret = smu_read_smc_arg(smu, &smu_version); + if (ret) + goto err; + + if (smu_version != smu->smc_if_version) + ret = -EINVAL; +err: + return ret; +} + +static int smu_v11_0_read_pptable_from_vbios(struct smu_context *smu) +{ + int ret, index; + uint16_t size; + uint8_t frev, crev; + void *table; + + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + powerplayinfo); + + ret = smu_get_atom_data_table(smu, index, &size, &frev, &crev, + (uint8_t **)&table); + if (ret) + return ret; + + if (!smu->smu_table.power_play_table) + smu->smu_table.power_play_table = table; + if (!smu->smu_table.power_play_table_size) + smu->smu_table.power_play_table_size = size; + + return 0; +} + +static int smu_v11_0_init_dpm_context(struct smu_context *smu) +{ + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + + if (smu_dpm->dpm_context || smu_dpm->dpm_context_size != 0) + return -EINVAL; + + return smu_alloc_dpm_context(smu); +} + +static int smu_v11_0_fini_dpm_context(struct smu_context *smu) +{ + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + + if (!smu_dpm->dpm_context || smu_dpm->dpm_context_size == 0) + return -EINVAL; + + kfree(smu_dpm->dpm_context); + kfree(smu_dpm->golden_dpm_context); + kfree(smu_dpm->dpm_current_power_state); + kfree(smu_dpm->dpm_request_power_state); + smu_dpm->dpm_context = NULL; + smu_dpm->golden_dpm_context = NULL; + smu_dpm->dpm_context_size = 0; + smu_dpm->dpm_current_power_state = NULL; + smu_dpm->dpm_request_power_state = NULL; + + return 0; +} + +static int smu_v11_0_init_smc_tables(struct smu_context *smu) +{ + struct smu_table_context *smu_table = &smu->smu_table; + struct smu_table *tables = NULL; + int ret = 0; + + if (smu_table->tables || smu_table->table_count != 0) + return -EINVAL; + + tables = kcalloc(TABLE_COUNT, sizeof(struct smu_table), GFP_KERNEL); + if (!tables) + return -ENOMEM; + + smu_table->tables = tables; + smu_table->table_count = TABLE_COUNT; + + SMU_TABLE_INIT(tables, TABLE_PPTABLE, sizeof(PPTable_t), + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, TABLE_WATERMARKS, sizeof(Watermarks_t), + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, TABLE_SMU_METRICS, sizeof(SmuMetrics_t), + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, TABLE_OVERDRIVE, sizeof(OverDriveTable_t), + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, TABLE_PMSTATUSLOG, SMU11_TOOL_SIZE, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, TABLE_ACTIVITY_MONITOR_COEFF, + sizeof(DpmActivityMonitorCoeffInt_t), + PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM); + + ret = smu_v11_0_init_dpm_context(smu); + if (ret) + return ret; + + return 0; +} + +static int smu_v11_0_fini_smc_tables(struct smu_context *smu) +{ + struct smu_table_context *smu_table = &smu->smu_table; + int ret = 0; + + if (!smu_table->tables || smu_table->table_count == 0) + return -EINVAL; + + kfree(smu_table->tables); + smu_table->tables = NULL; + smu_table->table_count = 0; + + ret = smu_v11_0_fini_dpm_context(smu); + if (ret) + return ret; + return 0; +} + +static int smu_v11_0_init_power(struct smu_context *smu) +{ + struct smu_power_context *smu_power = &smu->smu_power; + + if (smu_power->power_context || smu_power->power_context_size != 0) + return -EINVAL; + + smu_power->power_context = kzalloc(sizeof(struct smu_11_0_dpm_context), + GFP_KERNEL); + if (!smu_power->power_context) + return -ENOMEM; + smu_power->power_context_size = sizeof(struct smu_11_0_dpm_context); + + return 0; +} + +static int smu_v11_0_fini_power(struct smu_context *smu) +{ + struct smu_power_context *smu_power = &smu->smu_power; + + if (!smu_power->power_context || smu_power->power_context_size == 0) + return -EINVAL; + + kfree(smu_power->power_context); + smu_power->power_context = NULL; + smu_power->power_context_size = 0; + + return 0; +} + +int smu_v11_0_get_vbios_bootup_values(struct smu_context *smu) +{ + int ret, index; + uint16_t size; + uint8_t frev, crev; + struct atom_common_table_header *header; + struct atom_firmware_info_v3_3 *v_3_3; + struct atom_firmware_info_v3_1 *v_3_1; + + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + firmwareinfo); + + ret = smu_get_atom_data_table(smu, index, &size, &frev, &crev, + (uint8_t **)&header); + if (ret) + return ret; + + if (header->format_revision != 3) { + pr_err("unknown atom_firmware_info version! for smu11\n"); + return -EINVAL; + } + + switch (header->content_revision) { + case 0: + case 1: + case 2: + v_3_1 = (struct atom_firmware_info_v3_1 *)header; + smu->smu_table.boot_values.revision = v_3_1->firmware_revision; + smu->smu_table.boot_values.gfxclk = v_3_1->bootup_sclk_in10khz; + smu->smu_table.boot_values.uclk = v_3_1->bootup_mclk_in10khz; + smu->smu_table.boot_values.socclk = 0; + smu->smu_table.boot_values.dcefclk = 0; + smu->smu_table.boot_values.vddc = v_3_1->bootup_vddc_mv; + smu->smu_table.boot_values.vddci = v_3_1->bootup_vddci_mv; + smu->smu_table.boot_values.mvddc = v_3_1->bootup_mvddc_mv; + smu->smu_table.boot_values.vdd_gfx = v_3_1->bootup_vddgfx_mv; + smu->smu_table.boot_values.cooling_id = v_3_1->coolingsolution_id; + smu->smu_table.boot_values.pp_table_id = 0; + break; + case 3: + default: + v_3_3 = (struct atom_firmware_info_v3_3 *)header; + smu->smu_table.boot_values.revision = v_3_3->firmware_revision; + smu->smu_table.boot_values.gfxclk = v_3_3->bootup_sclk_in10khz; + smu->smu_table.boot_values.uclk = v_3_3->bootup_mclk_in10khz; + smu->smu_table.boot_values.socclk = 0; + smu->smu_table.boot_values.dcefclk = 0; + smu->smu_table.boot_values.vddc = v_3_3->bootup_vddc_mv; + smu->smu_table.boot_values.vddci = v_3_3->bootup_vddci_mv; + smu->smu_table.boot_values.mvddc = v_3_3->bootup_mvddc_mv; + smu->smu_table.boot_values.vdd_gfx = v_3_3->bootup_vddgfx_mv; + smu->smu_table.boot_values.cooling_id = v_3_3->coolingsolution_id; + smu->smu_table.boot_values.pp_table_id = v_3_3->pplib_pptable_id; + } + + return 0; +} + +static int smu_v11_0_get_clk_info_from_vbios(struct smu_context *smu) +{ + int ret, index; + struct amdgpu_device *adev = smu->adev; + struct atom_get_smu_clock_info_parameters_v3_1 input = {0}; + struct atom_get_smu_clock_info_output_parameters_v3_1 *output; + + input.clk_id = SMU11_SYSPLL0_SOCCLK_ID; + input.command = GET_SMU_CLOCK_INFO_V3_1_GET_CLOCK_FREQ; + index = get_index_into_master_table(atom_master_list_of_command_functions_v2_1, + getsmuclockinfo); + + ret = amdgpu_atom_execute_table(adev->mode_info.atom_context, index, + (uint32_t *)&input); + if (ret) + return -EINVAL; + + output = (struct atom_get_smu_clock_info_output_parameters_v3_1 *)&input; + smu->smu_table.boot_values.socclk = le32_to_cpu(output->atom_smu_outputclkfreq.smu_clock_freq_hz) / 10000; + + memset(&input, 0, sizeof(input)); + input.clk_id = SMU11_SYSPLL0_DCEFCLK_ID; + input.command = GET_SMU_CLOCK_INFO_V3_1_GET_CLOCK_FREQ; + index = get_index_into_master_table(atom_master_list_of_command_functions_v2_1, + getsmuclockinfo); + + ret = amdgpu_atom_execute_table(adev->mode_info.atom_context, index, + (uint32_t *)&input); + if (ret) + return -EINVAL; + + output = (struct atom_get_smu_clock_info_output_parameters_v3_1 *)&input; + smu->smu_table.boot_values.dcefclk = le32_to_cpu(output->atom_smu_outputclkfreq.smu_clock_freq_hz) / 10000; + + memset(&input, 0, sizeof(input)); + input.clk_id = SMU11_SYSPLL0_ECLK_ID; + input.command = GET_SMU_CLOCK_INFO_V3_1_GET_CLOCK_FREQ; + index = get_index_into_master_table(atom_master_list_of_command_functions_v2_1, + getsmuclockinfo); + + ret = amdgpu_atom_execute_table(adev->mode_info.atom_context, index, + (uint32_t *)&input); + if (ret) + return -EINVAL; + + output = (struct atom_get_smu_clock_info_output_parameters_v3_1 *)&input; + smu->smu_table.boot_values.eclk = le32_to_cpu(output->atom_smu_outputclkfreq.smu_clock_freq_hz) / 10000; + + memset(&input, 0, sizeof(input)); + input.clk_id = SMU11_SYSPLL0_VCLK_ID; + input.command = GET_SMU_CLOCK_INFO_V3_1_GET_CLOCK_FREQ; + index = get_index_into_master_table(atom_master_list_of_command_functions_v2_1, + getsmuclockinfo); + + ret = amdgpu_atom_execute_table(adev->mode_info.atom_context, index, + (uint32_t *)&input); + if (ret) + return -EINVAL; + + output = (struct atom_get_smu_clock_info_output_parameters_v3_1 *)&input; + smu->smu_table.boot_values.vclk = le32_to_cpu(output->atom_smu_outputclkfreq.smu_clock_freq_hz) / 10000; + + memset(&input, 0, sizeof(input)); + input.clk_id = SMU11_SYSPLL0_DCLK_ID; + input.command = GET_SMU_CLOCK_INFO_V3_1_GET_CLOCK_FREQ; + index = get_index_into_master_table(atom_master_list_of_command_functions_v2_1, + getsmuclockinfo); + + ret = amdgpu_atom_execute_table(adev->mode_info.atom_context, index, + (uint32_t *)&input); + if (ret) + return -EINVAL; + + output = (struct atom_get_smu_clock_info_output_parameters_v3_1 *)&input; + smu->smu_table.boot_values.dclk = le32_to_cpu(output->atom_smu_outputclkfreq.smu_clock_freq_hz) / 10000; + + return 0; +} + +static int smu_v11_0_notify_memory_pool_location(struct smu_context *smu) +{ + struct smu_table_context *smu_table = &smu->smu_table; + struct smu_table *memory_pool = &smu_table->memory_pool; + int ret = 0; + uint64_t address; + uint32_t address_low, address_high; + + if (memory_pool->size == 0 || memory_pool->cpu_addr == NULL) + return ret; + + address = (uintptr_t)memory_pool->cpu_addr; + address_high = (uint32_t)upper_32_bits(address); + address_low = (uint32_t)lower_32_bits(address); + + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_SetSystemVirtualDramAddrHigh, + address_high); + if (ret) + return ret; + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_SetSystemVirtualDramAddrLow, + address_low); + if (ret) + return ret; + + address = memory_pool->mc_address; + address_high = (uint32_t)upper_32_bits(address); + address_low = (uint32_t)lower_32_bits(address); + + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_DramLogSetDramAddrHigh, + address_high); + if (ret) + return ret; + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_DramLogSetDramAddrLow, + address_low); + if (ret) + return ret; + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_DramLogSetDramSize, + (uint32_t)memory_pool->size); + if (ret) + return ret; + + return ret; +} + +static int smu_v11_0_check_pptable(struct smu_context *smu) +{ + int ret; + + ret = smu_check_powerplay_table(smu); + return ret; +} + +static int smu_v11_0_parse_pptable(struct smu_context *smu) +{ + int ret; + + struct smu_table_context *table_context = &smu->smu_table; + + if (table_context->driver_pptable) + return -EINVAL; + + table_context->driver_pptable = kzalloc(sizeof(PPTable_t), GFP_KERNEL); + + if (!table_context->driver_pptable) + return -ENOMEM; + + ret = smu_store_powerplay_table(smu); + if (ret) + return -EINVAL; + + ret = smu_append_powerplay_table(smu); + + return ret; +} + +static int smu_v11_0_populate_smc_pptable(struct smu_context *smu) +{ + int ret; + + ret = smu_set_default_dpm_table(smu); + + return ret; +} + +static int smu_v11_0_write_pptable(struct smu_context *smu) +{ + struct smu_table_context *table_context = &smu->smu_table; + int ret = 0; + + ret = smu_update_table(smu, TABLE_PPTABLE, table_context->driver_pptable, true); + + return ret; +} + +static int smu_v11_0_write_watermarks_table(struct smu_context *smu) +{ + return smu_update_table(smu, TABLE_WATERMARKS, + smu->smu_table.tables[TABLE_WATERMARKS].cpu_addr, true); +} + +static int smu_v11_0_set_deep_sleep_dcefclk(struct smu_context *smu, uint32_t clk) +{ + int ret; + + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_SetMinDeepSleepDcefclk, clk); + if (ret) + pr_err("SMU11 attempt to set divider for DCEFCLK Failed!"); + + return ret; +} + +static int smu_v11_0_set_min_dcef_deep_sleep(struct smu_context *smu) +{ + struct smu_table_context *table_context = &smu->smu_table; + + if (!table_context) + return -EINVAL; + + return smu_set_deep_sleep_dcefclk(smu, + table_context->boot_values.dcefclk / 100); +} + +static int smu_v11_0_set_tool_table_location(struct smu_context *smu) +{ + int ret = 0; + struct smu_table *tool_table = &smu->smu_table.tables[TABLE_PMSTATUSLOG]; + + if (tool_table->mc_address) { + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_SetToolsDramAddrHigh, + upper_32_bits(tool_table->mc_address)); + if (!ret) + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_SetToolsDramAddrLow, + lower_32_bits(tool_table->mc_address)); + } + + return ret; +} + +static int smu_v11_0_init_display(struct smu_context *smu) +{ + int ret = 0; + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_NumOfDisplays, 0); + return ret; +} + +static int smu_v11_0_update_feature_enable_state(struct smu_context *smu, uint32_t feature_id, bool enabled) +{ + uint32_t feature_low = 0, feature_high = 0; + int ret = 0; + + if (feature_id >= 0 && feature_id < 31) + feature_low = (1 << feature_id); + else if (feature_id > 31 && feature_id < 63) + feature_high = (1 << feature_id); + else + return -EINVAL; + + if (enabled) { + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_EnableSmuFeaturesLow, + feature_low); + if (ret) + return ret; + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_EnableSmuFeaturesHigh, + feature_high); + if (ret) + return ret; + + } else { + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_DisableSmuFeaturesLow, + feature_low); + if (ret) + return ret; + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_DisableSmuFeaturesHigh, + feature_high); + if (ret) + return ret; + + } + + return ret; +} + +static int smu_v11_0_set_allowed_mask(struct smu_context *smu) +{ + struct smu_feature *feature = &smu->smu_feature; + int ret = 0; + uint32_t feature_mask[2]; + + mutex_lock(&feature->mutex); + if (bitmap_empty(feature->allowed, SMU_FEATURE_MAX) || feature->feature_num < 64) + goto failed; + + bitmap_copy((unsigned long *)feature_mask, feature->allowed, 64); + + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetAllowedFeaturesMaskHigh, + feature_mask[1]); + if (ret) + goto failed; + + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetAllowedFeaturesMaskLow, + feature_mask[0]); + if (ret) + goto failed; + +failed: + mutex_unlock(&feature->mutex); + return ret; +} + +static int smu_v11_0_get_enabled_mask(struct smu_context *smu, + uint32_t *feature_mask, uint32_t num) +{ + uint32_t feature_mask_high = 0, feature_mask_low = 0; + int ret = 0; + + if (!feature_mask || num < 2) + return -EINVAL; + + ret = smu_send_smc_msg(smu, SMU_MSG_GetEnabledSmuFeaturesHigh); + if (ret) + return ret; + ret = smu_read_smc_arg(smu, &feature_mask_high); + if (ret) + return ret; + + ret = smu_send_smc_msg(smu, SMU_MSG_GetEnabledSmuFeaturesLow); + if (ret) + return ret; + ret = smu_read_smc_arg(smu, &feature_mask_low); + if (ret) + return ret; + + feature_mask[0] = feature_mask_low; + feature_mask[1] = feature_mask_high; + + return ret; +} + +static bool smu_v11_0_is_dpm_running(struct smu_context *smu) +{ + int ret = 0; + uint32_t feature_mask[2]; + unsigned long feature_enabled; + ret = smu_v11_0_get_enabled_mask(smu, feature_mask, 2); + feature_enabled = (unsigned long)((uint64_t)feature_mask[0] | + ((uint64_t)feature_mask[1] << 32)); + return !!(feature_enabled & SMC_DPM_FEATURE); +} + +static int smu_v11_0_system_features_control(struct smu_context *smu, + bool en) +{ + struct smu_feature *feature = &smu->smu_feature; + uint32_t feature_mask[2]; + int ret = 0; + + ret = smu_send_smc_msg(smu, (en ? SMU_MSG_EnableAllSmuFeatures : + SMU_MSG_DisableAllSmuFeatures)); + if (ret) + return ret; + ret = smu_feature_get_enabled_mask(smu, feature_mask, 2); + if (ret) + return ret; + + bitmap_copy(feature->enabled, (unsigned long *)&feature_mask, + feature->feature_num); + bitmap_copy(feature->supported, (unsigned long *)&feature_mask, + feature->feature_num); + + return ret; +} + +static int smu_v11_0_notify_display_change(struct smu_context *smu) +{ + int ret = 0; + + if (smu_feature_is_enabled(smu, FEATURE_DPM_UCLK_BIT)) + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetUclkFastSwitch, 1); + + return ret; +} + +static int +smu_v11_0_get_max_sustainable_clock(struct smu_context *smu, uint32_t *clock, + PPCLK_e clock_select) +{ + int ret = 0; + + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_GetDcModeMaxDpmFreq, + clock_select << 16); + if (ret) { + pr_err("[GetMaxSustainableClock] Failed to get max DC clock from SMC!"); + return ret; + } + + ret = smu_read_smc_arg(smu, clock); + if (ret) + return ret; + + if (*clock != 0) + return 0; + + /* if DC limit is zero, return AC limit */ + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_GetMaxDpmFreq, + clock_select << 16); + if (ret) { + pr_err("[GetMaxSustainableClock] failed to get max AC clock from SMC!"); + return ret; + } + + ret = smu_read_smc_arg(smu, clock); + + return ret; +} + +static int smu_v11_0_init_max_sustainable_clocks(struct smu_context *smu) +{ + struct smu_11_0_max_sustainable_clocks *max_sustainable_clocks; + int ret = 0; + + max_sustainable_clocks = kzalloc(sizeof(struct smu_11_0_max_sustainable_clocks), + GFP_KERNEL); + smu->smu_table.max_sustainable_clocks = (void *)max_sustainable_clocks; + + max_sustainable_clocks->uclock = smu->smu_table.boot_values.uclk / 100; + max_sustainable_clocks->soc_clock = smu->smu_table.boot_values.socclk / 100; + max_sustainable_clocks->dcef_clock = smu->smu_table.boot_values.dcefclk / 100; + max_sustainable_clocks->display_clock = 0xFFFFFFFF; + max_sustainable_clocks->phy_clock = 0xFFFFFFFF; + max_sustainable_clocks->pixel_clock = 0xFFFFFFFF; + + if (smu_feature_is_enabled(smu, FEATURE_DPM_UCLK_BIT)) { + ret = smu_v11_0_get_max_sustainable_clock(smu, + &(max_sustainable_clocks->uclock), + PPCLK_UCLK); + if (ret) { + pr_err("[%s] failed to get max UCLK from SMC!", + __func__); + return ret; + } + } + + if (smu_feature_is_enabled(smu, FEATURE_DPM_SOCCLK_BIT)) { + ret = smu_v11_0_get_max_sustainable_clock(smu, + &(max_sustainable_clocks->soc_clock), + PPCLK_SOCCLK); + if (ret) { + pr_err("[%s] failed to get max SOCCLK from SMC!", + __func__); + return ret; + } + } + + if (smu_feature_is_enabled(smu, FEATURE_DPM_DCEFCLK_BIT)) { + ret = smu_v11_0_get_max_sustainable_clock(smu, + &(max_sustainable_clocks->dcef_clock), + PPCLK_DCEFCLK); + if (ret) { + pr_err("[%s] failed to get max DCEFCLK from SMC!", + __func__); + return ret; + } + + ret = smu_v11_0_get_max_sustainable_clock(smu, + &(max_sustainable_clocks->display_clock), + PPCLK_DISPCLK); + if (ret) { + pr_err("[%s] failed to get max DISPCLK from SMC!", + __func__); + return ret; + } + ret = smu_v11_0_get_max_sustainable_clock(smu, + &(max_sustainable_clocks->phy_clock), + PPCLK_PHYCLK); + if (ret) { + pr_err("[%s] failed to get max PHYCLK from SMC!", + __func__); + return ret; + } + ret = smu_v11_0_get_max_sustainable_clock(smu, + &(max_sustainable_clocks->pixel_clock), + PPCLK_PIXCLK); + if (ret) { + pr_err("[%s] failed to get max PIXCLK from SMC!", + __func__); + return ret; + } + } + + if (max_sustainable_clocks->soc_clock < max_sustainable_clocks->uclock) + max_sustainable_clocks->uclock = max_sustainable_clocks->soc_clock; + + return 0; +} + +static int smu_v11_0_get_power_limit(struct smu_context *smu, + uint32_t *limit, + bool get_default) +{ + int ret = 0; + + if (get_default) { + mutex_lock(&smu->mutex); + *limit = smu->default_power_limit; + if (smu->od_enabled) { + *limit *= (100 + smu->smu_table.TDPODLimit); + *limit /= 100; + } + mutex_unlock(&smu->mutex); + } else { + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_GetPptLimit, + POWER_SOURCE_AC << 16); + if (ret) { + pr_err("[%s] get PPT limit failed!", __func__); + return ret; + } + smu_read_smc_arg(smu, limit); + smu->power_limit = *limit; + } + + return ret; +} + +static int smu_v11_0_set_power_limit(struct smu_context *smu, uint32_t n) +{ + uint32_t max_power_limit; + int ret = 0; + + if (n == 0) + n = smu->default_power_limit; + + max_power_limit = smu->default_power_limit; + + if (smu->od_enabled) { + max_power_limit *= (100 + smu->smu_table.TDPODLimit); + max_power_limit /= 100; + } + + if (smu_feature_is_enabled(smu, FEATURE_PPT_BIT)) + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetPptLimit, n); + if (ret) { + pr_err("[%s] Set power limit Failed!", __func__); + return ret; + } + + return ret; +} + +static int smu_v11_0_get_current_clk_freq(struct smu_context *smu, uint32_t clk_id, uint32_t *value) +{ + int ret = 0; + uint32_t freq; + + if (clk_id >= PPCLK_COUNT || !value) + return -EINVAL; + + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_GetDpmClockFreq, (clk_id << 16)); + if (ret) + return ret; + + ret = smu_read_smc_arg(smu, &freq); + if (ret) + return ret; + + freq *= 100; + *value = freq; + + return ret; +} + +static int smu_v11_0_get_thermal_range(struct smu_context *smu, + struct PP_TemperatureRange *range) +{ + memcpy(range, &SMU7ThermalWithDelayPolicy[0], sizeof(struct PP_TemperatureRange)); + + range->max = smu->smu_table.software_shutdown_temp * + PP_TEMPERATURE_UNITS_PER_CENTIGRADES; + + return 0; +} + +static int smu_v11_0_set_thermal_range(struct smu_context *smu, + struct PP_TemperatureRange *range) +{ + struct amdgpu_device *adev = smu->adev; + int low = SMU11_THERMAL_MINIMUM_ALERT_TEMP * + PP_TEMPERATURE_UNITS_PER_CENTIGRADES; + int high = SMU11_THERMAL_MAXIMUM_ALERT_TEMP * + PP_TEMPERATURE_UNITS_PER_CENTIGRADES; + uint32_t val; + + if (low < range->min) + low = range->min; + if (high > range->max) + high = range->max; + + if (low > high) + return -EINVAL; + + val = RREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_CTRL); + val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, MAX_IH_CREDIT, 5); + val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, THERM_IH_HW_ENA, 1); + val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTH, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES)); + val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTL, (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES)); + val = val & (~THM_THERMAL_INT_CTRL__THERM_TRIGGER_MASK_MASK); + + WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_CTRL, val); + + return 0; +} + +static int smu_v11_0_enable_thermal_alert(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + uint32_t val = 0; + + val |= (1 << THM_THERMAL_INT_ENA__THERM_INTH_CLR__SHIFT); + val |= (1 << THM_THERMAL_INT_ENA__THERM_INTL_CLR__SHIFT); + val |= (1 << THM_THERMAL_INT_ENA__THERM_TRIGGER_CLR__SHIFT); + + WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_ENA, val); + + return 0; +} + +static int smu_v11_0_set_thermal_fan_table(struct smu_context *smu) +{ + int ret; + struct smu_table_context *table_context = &smu->smu_table; + PPTable_t *pptable = table_context->driver_pptable; + + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetFanTemperatureTarget, + (uint32_t)pptable->FanTargetTemperature); + + return ret; +} + +static int smu_v11_0_start_thermal_control(struct smu_context *smu) +{ + int ret = 0; + struct PP_TemperatureRange range; + struct amdgpu_device *adev = smu->adev; + + smu_v11_0_get_thermal_range(smu, &range); + + if (smu->smu_table.thermal_controller_type) { + ret = smu_v11_0_set_thermal_range(smu, &range); + if (ret) + return ret; + + ret = smu_v11_0_enable_thermal_alert(smu); + if (ret) + return ret; + ret = smu_v11_0_set_thermal_fan_table(smu); + if (ret) + return ret; + } + + adev->pm.dpm.thermal.min_temp = range.min; + adev->pm.dpm.thermal.max_temp = range.max; + + return ret; +} + +static int smu_v11_0_get_current_activity_percent(struct smu_context *smu, + uint32_t *value) +{ + int ret = 0; + SmuMetrics_t metrics; + + if (!value) + return -EINVAL; + + ret = smu_update_table(smu, TABLE_SMU_METRICS, (void *)&metrics, false); + if (ret) + return ret; + + *value = metrics.AverageGfxActivity; + + return 0; +} + +static int smu_v11_0_thermal_get_temperature(struct smu_context *smu, uint32_t *value) +{ + struct amdgpu_device *adev = smu->adev; + uint32_t temp = 0; + + if (!value) + return -EINVAL; + + temp = RREG32_SOC15(THM, 0, mmCG_MULT_THERMAL_STATUS); + temp = (temp & CG_MULT_THERMAL_STATUS__CTF_TEMP_MASK) >> + CG_MULT_THERMAL_STATUS__CTF_TEMP__SHIFT; + + temp = temp & 0x1ff; + temp *= SMU11_TEMPERATURE_UNITS_PER_CENTIGRADES; + + *value = temp; + + return 0; +} + +static int smu_v11_0_get_gpu_power(struct smu_context *smu, uint32_t *value) +{ + int ret = 0; + SmuMetrics_t metrics; + + if (!value) + return -EINVAL; + + ret = smu_update_table(smu, TABLE_SMU_METRICS, (void *)&metrics, false); + if (ret) + return ret; + + *value = metrics.CurrSocketPower << 8; + + return 0; +} + +static uint16_t convert_to_vddc(uint8_t vid) +{ + return (uint16_t) ((6200 - (vid * 25)) / SMU11_VOLTAGE_SCALE); +} + +static int smu_v11_0_get_gfx_vdd(struct smu_context *smu, uint32_t *value) +{ + struct amdgpu_device *adev = smu->adev; + uint32_t vdd = 0, val_vid = 0; + + if (!value) + return -EINVAL; + val_vid = (RREG32_SOC15(SMUIO, 0, mmSMUSVI0_TEL_PLANE0) & + SMUSVI0_TEL_PLANE0__SVI0_PLANE0_VDDCOR_MASK) >> + SMUSVI0_TEL_PLANE0__SVI0_PLANE0_VDDCOR__SHIFT; + + vdd = (uint32_t)convert_to_vddc((uint8_t)val_vid); + + *value = vdd; + + return 0; + +} + +static int smu_v11_0_read_sensor(struct smu_context *smu, + enum amd_pp_sensors sensor, + void *data, uint32_t *size) +{ + struct smu_table_context *table_context = &smu->smu_table; + PPTable_t *pptable = table_context->driver_pptable; + int ret = 0; + switch (sensor) { + case AMDGPU_PP_SENSOR_GPU_LOAD: + ret = smu_v11_0_get_current_activity_percent(smu, + (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_GFX_MCLK: + ret = smu_get_current_clk_freq(smu, PPCLK_UCLK, (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_GFX_SCLK: + ret = smu_get_current_clk_freq(smu, PPCLK_GFXCLK, (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_GPU_TEMP: + ret = smu_v11_0_thermal_get_temperature(smu, (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_GPU_POWER: + ret = smu_v11_0_get_gpu_power(smu, (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_VDDGFX: + ret = smu_v11_0_get_gfx_vdd(smu, (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_UVD_POWER: + *(uint32_t *)data = smu_feature_is_enabled(smu, FEATURE_DPM_UVD_BIT) ? 1 : 0; + *size = 4; + break; + case AMDGPU_PP_SENSOR_VCE_POWER: + *(uint32_t *)data = smu_feature_is_enabled(smu, FEATURE_DPM_VCE_BIT) ? 1 : 0; + *size = 4; + break; + case AMDGPU_PP_SENSOR_MIN_FAN_RPM: + *(uint32_t *)data = 0; + *size = 4; + break; + case AMDGPU_PP_SENSOR_MAX_FAN_RPM: + *(uint32_t *)data = pptable->FanMaximumRpm; + *size = 4; + break; + default: + ret = smu_common_read_sensor(smu, sensor, data, size); + break; + } + + if (ret) + *size = 0; + + return ret; +} + +static int +smu_v11_0_display_clock_voltage_request(struct smu_context *smu, + struct pp_display_clock_request + *clock_req) +{ + enum amd_pp_clock_type clk_type = clock_req->clock_type; + int ret = 0; + PPCLK_e clk_select = 0; + uint32_t clk_freq = clock_req->clock_freq_in_khz / 1000; + + if (smu_feature_is_enabled(smu, FEATURE_DPM_DCEFCLK_BIT)) { + switch (clk_type) { + case amd_pp_dcef_clock: + clk_select = PPCLK_DCEFCLK; + break; + case amd_pp_disp_clock: + clk_select = PPCLK_DISPCLK; + break; + case amd_pp_pixel_clock: + clk_select = PPCLK_PIXCLK; + break; + case amd_pp_phy_clock: + clk_select = PPCLK_PHYCLK; + break; + default: + pr_info("[%s] Invalid Clock Type!", __func__); + ret = -EINVAL; + break; + } + + if (ret) + goto failed; + + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetHardMinByFreq, + (clk_select << 16) | clk_freq); + } + +failed: + return ret; +} + +static int smu_v11_0_set_watermarks_table(struct smu_context *smu, + Watermarks_t *table, struct + dm_pp_wm_sets_with_clock_ranges_soc15 + *clock_ranges) +{ + int i; + + if (!table || !clock_ranges) + return -EINVAL; + + if (clock_ranges->num_wm_dmif_sets > 4 || + clock_ranges->num_wm_mcif_sets > 4) + return -EINVAL; + + for (i = 0; i < clock_ranges->num_wm_dmif_sets; i++) { + table->WatermarkRow[1][i].MinClock = + cpu_to_le16((uint16_t) + (clock_ranges->wm_dmif_clocks_ranges[i].wm_min_dcfclk_clk_in_khz / + 1000)); + table->WatermarkRow[1][i].MaxClock = + cpu_to_le16((uint16_t) + (clock_ranges->wm_dmif_clocks_ranges[i].wm_max_dcfclk_clk_in_khz / + 1000)); + table->WatermarkRow[1][i].MinUclk = + cpu_to_le16((uint16_t) + (clock_ranges->wm_dmif_clocks_ranges[i].wm_min_mem_clk_in_khz / + 1000)); + table->WatermarkRow[1][i].MaxUclk = + cpu_to_le16((uint16_t) + (clock_ranges->wm_dmif_clocks_ranges[i].wm_max_mem_clk_in_khz / + 1000)); + table->WatermarkRow[1][i].WmSetting = (uint8_t) + clock_ranges->wm_dmif_clocks_ranges[i].wm_set_id; + } + + for (i = 0; i < clock_ranges->num_wm_mcif_sets; i++) { + table->WatermarkRow[0][i].MinClock = + cpu_to_le16((uint16_t) + (clock_ranges->wm_mcif_clocks_ranges[i].wm_min_socclk_clk_in_khz / + 1000)); + table->WatermarkRow[0][i].MaxClock = + cpu_to_le16((uint16_t) + (clock_ranges->wm_mcif_clocks_ranges[i].wm_max_socclk_clk_in_khz / + 1000)); + table->WatermarkRow[0][i].MinUclk = + cpu_to_le16((uint16_t) + (clock_ranges->wm_mcif_clocks_ranges[i].wm_min_mem_clk_in_khz / + 1000)); + table->WatermarkRow[0][i].MaxUclk = + cpu_to_le16((uint16_t) + (clock_ranges->wm_mcif_clocks_ranges[i].wm_max_mem_clk_in_khz / + 1000)); + table->WatermarkRow[0][i].WmSetting = (uint8_t) + clock_ranges->wm_mcif_clocks_ranges[i].wm_set_id; + } + + return 0; +} + +static int +smu_v11_0_set_watermarks_for_clock_ranges(struct smu_context *smu, struct + dm_pp_wm_sets_with_clock_ranges_soc15 + *clock_ranges) +{ + int ret = 0; + struct smu_table *watermarks = &smu->smu_table.tables[TABLE_WATERMARKS]; + Watermarks_t *table = watermarks->cpu_addr; + + if (!smu->disable_watermark && + smu_feature_is_enabled(smu, FEATURE_DPM_DCEFCLK_BIT) && + smu_feature_is_enabled(smu, FEATURE_DPM_SOCCLK_BIT)) { + smu_v11_0_set_watermarks_table(smu, table, clock_ranges); + smu->watermarks_bitmap |= WATERMARKS_EXIST; + smu->watermarks_bitmap &= ~WATERMARKS_LOADED; + } + + return ret; +} + +static int smu_v11_0_get_clock_ranges(struct smu_context *smu, + uint32_t *clock, + PPCLK_e clock_select, + bool max) +{ + int ret; + *clock = 0; + if (max) { + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_GetMaxDpmFreq, + (clock_select << 16)); + if (ret) { + pr_err("[GetClockRanges] Failed to get max clock from SMC!\n"); + return ret; + } + smu_read_smc_arg(smu, clock); + } else { + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_GetMinDpmFreq, + (clock_select << 16)); + if (ret) { + pr_err("[GetClockRanges] Failed to get min clock from SMC!\n"); + return ret; + } + smu_read_smc_arg(smu, clock); + } + + return 0; +} + +static uint32_t smu_v11_0_dpm_get_sclk(struct smu_context *smu, bool low) +{ + uint32_t gfx_clk; + int ret; + + if (!smu_feature_is_enabled(smu, FEATURE_DPM_GFXCLK_BIT)) { + pr_err("[GetSclks]: gfxclk dpm not enabled!\n"); + return -EPERM; + } + + if (low) { + ret = smu_v11_0_get_clock_ranges(smu, &gfx_clk, PPCLK_GFXCLK, false); + if (ret) { + pr_err("[GetSclks]: fail to get min PPCLK_GFXCLK\n"); + return ret; + } + } else { + ret = smu_v11_0_get_clock_ranges(smu, &gfx_clk, PPCLK_GFXCLK, true); + if (ret) { + pr_err("[GetSclks]: fail to get max PPCLK_GFXCLK\n"); + return ret; + } + } + + return (gfx_clk * 100); +} + +static uint32_t smu_v11_0_dpm_get_mclk(struct smu_context *smu, bool low) +{ + uint32_t mem_clk; + int ret; + + if (!smu_feature_is_enabled(smu, FEATURE_DPM_UCLK_BIT)) { + pr_err("[GetMclks]: memclk dpm not enabled!\n"); + return -EPERM; + } + + if (low) { + ret = smu_v11_0_get_clock_ranges(smu, &mem_clk, PPCLK_UCLK, false); + if (ret) { + pr_err("[GetMclks]: fail to get min PPCLK_UCLK\n"); + return ret; + } + } else { + ret = smu_v11_0_get_clock_ranges(smu, &mem_clk, PPCLK_GFXCLK, true); + if (ret) { + pr_err("[GetMclks]: fail to get max PPCLK_UCLK\n"); + return ret; + } + } + + return (mem_clk * 100); +} + +static int smu_v11_0_set_od8_default_settings(struct smu_context *smu, + bool initialize) +{ + struct smu_table_context *table_context = &smu->smu_table; + int ret; + + if (initialize) { + if (table_context->overdrive_table) + return -EINVAL; + + table_context->overdrive_table = kzalloc(sizeof(OverDriveTable_t), GFP_KERNEL); + + if (!table_context->overdrive_table) + return -ENOMEM; + + ret = smu_update_table(smu, TABLE_OVERDRIVE, table_context->overdrive_table, false); + if (ret) { + pr_err("Failed to export over drive table!\n"); + return ret; + } + + smu_set_default_od8_settings(smu); + } + + ret = smu_update_table(smu, TABLE_OVERDRIVE, table_context->overdrive_table, true); + if (ret) { + pr_err("Failed to import over drive table!\n"); + return ret; + } + + return 0; +} + +static int smu_v11_0_conv_power_profile_to_pplib_workload(int power_profile) +{ + int pplib_workload = 0; + + switch (power_profile) { + case PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT: + pplib_workload = WORKLOAD_DEFAULT_BIT; + break; + case PP_SMC_POWER_PROFILE_FULLSCREEN3D: + pplib_workload = WORKLOAD_PPLIB_FULL_SCREEN_3D_BIT; + break; + case PP_SMC_POWER_PROFILE_POWERSAVING: + pplib_workload = WORKLOAD_PPLIB_POWER_SAVING_BIT; + break; + case PP_SMC_POWER_PROFILE_VIDEO: + pplib_workload = WORKLOAD_PPLIB_VIDEO_BIT; + break; + case PP_SMC_POWER_PROFILE_VR: + pplib_workload = WORKLOAD_PPLIB_VR_BIT; + break; + case PP_SMC_POWER_PROFILE_COMPUTE: + pplib_workload = WORKLOAD_PPLIB_COMPUTE_BIT; + break; + case PP_SMC_POWER_PROFILE_CUSTOM: + pplib_workload = WORKLOAD_PPLIB_CUSTOM_BIT; + break; + } + + return pplib_workload; +} + +static int smu_v11_0_get_power_profile_mode(struct smu_context *smu, char *buf) +{ + DpmActivityMonitorCoeffInt_t activity_monitor; + uint32_t i, size = 0; + uint16_t workload_type = 0; + static const char *profile_name[] = { + "BOOTUP_DEFAULT", + "3D_FULL_SCREEN", + "POWER_SAVING", + "VIDEO", + "VR", + "COMPUTE", + "CUSTOM"}; + static const char *title[] = { + "PROFILE_INDEX(NAME)", + "CLOCK_TYPE(NAME)", + "FPS", + "UseRlcBusy", + "MinActiveFreqType", + "MinActiveFreq", + "BoosterFreqType", + "BoosterFreq", + "PD_Data_limit_c", + "PD_Data_error_coeff", + "PD_Data_error_rate_coeff"}; + int result = 0; + + if (!buf) + return -EINVAL; + + size += sprintf(buf + size, "%16s %s %s %s %s %s %s %s %s %s %s\n", + title[0], title[1], title[2], title[3], title[4], title[5], + title[6], title[7], title[8], title[9], title[10]); + + for (i = 0; i <= PP_SMC_POWER_PROFILE_CUSTOM; i++) { + /* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */ + workload_type = smu_v11_0_conv_power_profile_to_pplib_workload(i); + result = smu_update_table_with_arg(smu, TABLE_ACTIVITY_MONITOR_COEFF, + workload_type, &activity_monitor, false); + if (result) { + pr_err("[%s] Failed to get activity monitor!", __func__); + return result; + } + + size += sprintf(buf + size, "%2d %14s%s:\n", + i, profile_name[i], (i == smu->power_profile_mode) ? "*" : " "); + + size += sprintf(buf + size, "%19s %d(%13s) %7d %7d %7d %7d %7d %7d %7d %7d %7d\n", + " ", + 0, + "GFXCLK", + activity_monitor.Gfx_FPS, + activity_monitor.Gfx_UseRlcBusy, + activity_monitor.Gfx_MinActiveFreqType, + activity_monitor.Gfx_MinActiveFreq, + activity_monitor.Gfx_BoosterFreqType, + activity_monitor.Gfx_BoosterFreq, + activity_monitor.Gfx_PD_Data_limit_c, + activity_monitor.Gfx_PD_Data_error_coeff, + activity_monitor.Gfx_PD_Data_error_rate_coeff); + + size += sprintf(buf + size, "%19s %d(%13s) %7d %7d %7d %7d %7d %7d %7d %7d %7d\n", + " ", + 1, + "SOCCLK", + activity_monitor.Soc_FPS, + activity_monitor.Soc_UseRlcBusy, + activity_monitor.Soc_MinActiveFreqType, + activity_monitor.Soc_MinActiveFreq, + activity_monitor.Soc_BoosterFreqType, + activity_monitor.Soc_BoosterFreq, + activity_monitor.Soc_PD_Data_limit_c, + activity_monitor.Soc_PD_Data_error_coeff, + activity_monitor.Soc_PD_Data_error_rate_coeff); + + size += sprintf(buf + size, "%19s %d(%13s) %7d %7d %7d %7d %7d %7d %7d %7d %7d\n", + " ", + 2, + "UCLK", + activity_monitor.Mem_FPS, + activity_monitor.Mem_UseRlcBusy, + activity_monitor.Mem_MinActiveFreqType, + activity_monitor.Mem_MinActiveFreq, + activity_monitor.Mem_BoosterFreqType, + activity_monitor.Mem_BoosterFreq, + activity_monitor.Mem_PD_Data_limit_c, + activity_monitor.Mem_PD_Data_error_coeff, + activity_monitor.Mem_PD_Data_error_rate_coeff); + + size += sprintf(buf + size, "%19s %d(%13s) %7d %7d %7d %7d %7d %7d %7d %7d %7d\n", + " ", + 3, + "FCLK", + activity_monitor.Fclk_FPS, + activity_monitor.Fclk_UseRlcBusy, + activity_monitor.Fclk_MinActiveFreqType, + activity_monitor.Fclk_MinActiveFreq, + activity_monitor.Fclk_BoosterFreqType, + activity_monitor.Fclk_BoosterFreq, + activity_monitor.Fclk_PD_Data_limit_c, + activity_monitor.Fclk_PD_Data_error_coeff, + activity_monitor.Fclk_PD_Data_error_rate_coeff); + } + + return size; +} + +static int smu_v11_0_set_power_profile_mode(struct smu_context *smu, long *input, uint32_t size) +{ + DpmActivityMonitorCoeffInt_t activity_monitor; + int workload_type = 0, ret = 0; + + smu->power_profile_mode = input[size]; + + if (smu->power_profile_mode > PP_SMC_POWER_PROFILE_CUSTOM) { + pr_err("Invalid power profile mode %d\n", smu->power_profile_mode); + return -EINVAL; + } + + if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) { + ret = smu_update_table_with_arg(smu, TABLE_ACTIVITY_MONITOR_COEFF, + WORKLOAD_PPLIB_CUSTOM_BIT, &activity_monitor, false); + if (ret) { + pr_err("[%s] Failed to get activity monitor!", __func__); + return ret; + } + + switch (input[0]) { + case 0: /* Gfxclk */ + activity_monitor.Gfx_FPS = input[1]; + activity_monitor.Gfx_UseRlcBusy = input[2]; + activity_monitor.Gfx_MinActiveFreqType = input[3]; + activity_monitor.Gfx_MinActiveFreq = input[4]; + activity_monitor.Gfx_BoosterFreqType = input[5]; + activity_monitor.Gfx_BoosterFreq = input[6]; + activity_monitor.Gfx_PD_Data_limit_c = input[7]; + activity_monitor.Gfx_PD_Data_error_coeff = input[8]; + activity_monitor.Gfx_PD_Data_error_rate_coeff = input[9]; + break; + case 1: /* Socclk */ + activity_monitor.Soc_FPS = input[1]; + activity_monitor.Soc_UseRlcBusy = input[2]; + activity_monitor.Soc_MinActiveFreqType = input[3]; + activity_monitor.Soc_MinActiveFreq = input[4]; + activity_monitor.Soc_BoosterFreqType = input[5]; + activity_monitor.Soc_BoosterFreq = input[6]; + activity_monitor.Soc_PD_Data_limit_c = input[7]; + activity_monitor.Soc_PD_Data_error_coeff = input[8]; + activity_monitor.Soc_PD_Data_error_rate_coeff = input[9]; + break; + case 2: /* Uclk */ + activity_monitor.Mem_FPS = input[1]; + activity_monitor.Mem_UseRlcBusy = input[2]; + activity_monitor.Mem_MinActiveFreqType = input[3]; + activity_monitor.Mem_MinActiveFreq = input[4]; + activity_monitor.Mem_BoosterFreqType = input[5]; + activity_monitor.Mem_BoosterFreq = input[6]; + activity_monitor.Mem_PD_Data_limit_c = input[7]; + activity_monitor.Mem_PD_Data_error_coeff = input[8]; + activity_monitor.Mem_PD_Data_error_rate_coeff = input[9]; + break; + case 3: /* Fclk */ + activity_monitor.Fclk_FPS = input[1]; + activity_monitor.Fclk_UseRlcBusy = input[2]; + activity_monitor.Fclk_MinActiveFreqType = input[3]; + activity_monitor.Fclk_MinActiveFreq = input[4]; + activity_monitor.Fclk_BoosterFreqType = input[5]; + activity_monitor.Fclk_BoosterFreq = input[6]; + activity_monitor.Fclk_PD_Data_limit_c = input[7]; + activity_monitor.Fclk_PD_Data_error_coeff = input[8]; + activity_monitor.Fclk_PD_Data_error_rate_coeff = input[9]; + break; + } + + ret = smu_update_table_with_arg(smu, TABLE_ACTIVITY_MONITOR_COEFF, + WORKLOAD_PPLIB_COMPUTE_BIT, &activity_monitor, true); + if (ret) { + pr_err("[%s] Failed to set activity monitor!", __func__); + return ret; + } + } + + /* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */ + workload_type = + smu_v11_0_conv_power_profile_to_pplib_workload(smu->power_profile_mode); + smu_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, + 1 << workload_type); + + return ret; +} + +static int smu_v11_0_update_od8_settings(struct smu_context *smu, + uint32_t index, + uint32_t value) +{ + struct smu_table_context *table_context = &smu->smu_table; + int ret; + + ret = smu_update_table(smu, TABLE_OVERDRIVE, + table_context->overdrive_table, false); + if (ret) { + pr_err("Failed to export over drive table!\n"); + return ret; + } + + smu_update_specified_od8_value(smu, index, value); + + ret = smu_update_table(smu, TABLE_OVERDRIVE, + table_context->overdrive_table, true); + if (ret) { + pr_err("Failed to import over drive table!\n"); + return ret; + } + + return 0; +} + +static int smu_v11_0_dpm_set_uvd_enable(struct smu_context *smu, bool enable) +{ + if (!smu_feature_is_supported(smu, FEATURE_DPM_VCE_BIT)) + return 0; + + if (enable == smu_feature_is_enabled(smu, FEATURE_DPM_VCE_BIT)) + return 0; + + return smu_feature_set_enabled(smu, FEATURE_DPM_VCE_BIT, enable); +} + +static int smu_v11_0_dpm_set_vce_enable(struct smu_context *smu, bool enable) +{ + if (!smu_feature_is_supported(smu, FEATURE_DPM_UVD_BIT)) + return 0; + + if (enable == smu_feature_is_enabled(smu, FEATURE_DPM_UVD_BIT)) + return 0; + + return smu_feature_set_enabled(smu, FEATURE_DPM_UVD_BIT, enable); +} + +static int smu_v11_0_get_current_rpm(struct smu_context *smu, + uint32_t *current_rpm) +{ + int ret; + + ret = smu_send_smc_msg(smu, SMU_MSG_GetCurrentRpm); + + if (ret) { + pr_err("Attempt to get current RPM from SMC Failed!\n"); + return ret; + } + + smu_read_smc_arg(smu, current_rpm); + + return 0; +} + +static uint32_t +smu_v11_0_get_fan_control_mode(struct smu_context *smu) +{ + if (!smu_feature_is_enabled(smu, FEATURE_FAN_CONTROL_BIT)) + return AMD_FAN_CTRL_MANUAL; + else + return AMD_FAN_CTRL_AUTO; +} + +static int +smu_v11_0_get_fan_speed_percent(struct smu_context *smu, + uint32_t *speed) +{ + int ret = 0; + uint32_t percent = 0; + uint32_t current_rpm; + PPTable_t *pptable = smu->smu_table.driver_pptable; + + ret = smu_v11_0_get_current_rpm(smu, ¤t_rpm); + percent = current_rpm * 100 / pptable->FanMaximumRpm; + *speed = percent > 100 ? 100 : percent; + + return ret; +} + +static int +smu_v11_0_smc_fan_control(struct smu_context *smu, bool start) +{ + int ret = 0; + + if (smu_feature_is_supported(smu, FEATURE_FAN_CONTROL_BIT)) + return 0; + + ret = smu_feature_set_enabled(smu, FEATURE_FAN_CONTROL_BIT, start); + if (ret) + pr_err("[%s]%s smc FAN CONTROL feature failed!", + __func__, (start ? "Start" : "Stop")); + + return ret; +} + +static int +smu_v11_0_set_fan_static_mode(struct smu_context *smu, uint32_t mode) +{ + struct amdgpu_device *adev = smu->adev; + + WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2, + REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2), + CG_FDO_CTRL2, TMIN, 0)); + WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2, + REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2), + CG_FDO_CTRL2, FDO_PWM_MODE, mode)); + + return 0; +} + +static int +smu_v11_0_set_fan_speed_percent(struct smu_context *smu, uint32_t speed) +{ + struct amdgpu_device *adev = smu->adev; + uint32_t duty100; + uint32_t duty; + uint64_t tmp64; + bool stop = 0; + + if (speed > 100) + speed = 100; + + if (smu_v11_0_smc_fan_control(smu, stop)) + return -EINVAL; + duty100 = REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL1), + CG_FDO_CTRL1, FMAX_DUTY100); + if (!duty100) + return -EINVAL; + + tmp64 = (uint64_t)speed * duty100; + do_div(tmp64, 100); + duty = (uint32_t)tmp64; + + WREG32_SOC15(THM, 0, mmCG_FDO_CTRL0, + REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL0), + CG_FDO_CTRL0, FDO_STATIC_DUTY, duty)); + + return smu_v11_0_set_fan_static_mode(smu, FDO_PWM_MODE_STATIC); +} + +static int +smu_v11_0_set_fan_control_mode(struct smu_context *smu, + uint32_t mode) +{ + int ret = 0; + bool start = 1; + bool stop = 0; + + switch (mode) { + case AMD_FAN_CTRL_NONE: + ret = smu_v11_0_set_fan_speed_percent(smu, 100); + break; + case AMD_FAN_CTRL_MANUAL: + ret = smu_v11_0_smc_fan_control(smu, stop); + break; + case AMD_FAN_CTRL_AUTO: + ret = smu_v11_0_smc_fan_control(smu, start); + break; + default: + break; + } + + if (ret) { + pr_err("[%s]Set fan control mode failed!", __func__); + return -EINVAL; + } + + return ret; +} + +static int smu_v11_0_set_fan_speed_rpm(struct smu_context *smu, + uint32_t speed) +{ + struct amdgpu_device *adev = smu->adev; + int ret; + uint32_t tach_period, crystal_clock_freq; + bool stop = 0; + + if (!speed) + return -EINVAL; + + mutex_lock(&(smu->mutex)); + ret = smu_v11_0_smc_fan_control(smu, stop); + if (ret) + goto set_fan_speed_rpm_failed; + + crystal_clock_freq = amdgpu_asic_get_xclk(adev); + tach_period = 60 * crystal_clock_freq * 10000 / (8 * speed); + WREG32_SOC15(THM, 0, mmCG_TACH_CTRL, + REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_TACH_CTRL), + CG_TACH_CTRL, TARGET_PERIOD, + tach_period)); + + ret = smu_v11_0_set_fan_static_mode(smu, FDO_PWM_MODE_STATIC_RPM); + +set_fan_speed_rpm_failed: + mutex_unlock(&(smu->mutex)); + return ret; +} + +static int smu_v11_0_set_xgmi_pstate(struct smu_context *smu, + uint32_t pstate) +{ + /* send msg to SMU to set pstate */ + return 0; +} + +static const struct smu_funcs smu_v11_0_funcs = { + .init_microcode = smu_v11_0_init_microcode, + .load_microcode = smu_v11_0_load_microcode, + .check_fw_status = smu_v11_0_check_fw_status, + .check_fw_version = smu_v11_0_check_fw_version, + .send_smc_msg = smu_v11_0_send_msg, + .send_smc_msg_with_param = smu_v11_0_send_msg_with_param, + .read_smc_arg = smu_v11_0_read_arg, + .read_pptable_from_vbios = smu_v11_0_read_pptable_from_vbios, + .init_smc_tables = smu_v11_0_init_smc_tables, + .fini_smc_tables = smu_v11_0_fini_smc_tables, + .init_power = smu_v11_0_init_power, + .fini_power = smu_v11_0_fini_power, + .get_vbios_bootup_values = smu_v11_0_get_vbios_bootup_values, + .get_clk_info_from_vbios = smu_v11_0_get_clk_info_from_vbios, + .notify_memory_pool_location = smu_v11_0_notify_memory_pool_location, + .check_pptable = smu_v11_0_check_pptable, + .parse_pptable = smu_v11_0_parse_pptable, + .populate_smc_pptable = smu_v11_0_populate_smc_pptable, + .write_pptable = smu_v11_0_write_pptable, + .write_watermarks_table = smu_v11_0_write_watermarks_table, + .set_min_dcef_deep_sleep = smu_v11_0_set_min_dcef_deep_sleep, + .set_tool_table_location = smu_v11_0_set_tool_table_location, + .init_display = smu_v11_0_init_display, + .set_allowed_mask = smu_v11_0_set_allowed_mask, + .get_enabled_mask = smu_v11_0_get_enabled_mask, + .is_dpm_running = smu_v11_0_is_dpm_running, + .system_features_control = smu_v11_0_system_features_control, + .update_feature_enable_state = smu_v11_0_update_feature_enable_state, + .notify_display_change = smu_v11_0_notify_display_change, + .get_power_limit = smu_v11_0_get_power_limit, + .set_power_limit = smu_v11_0_set_power_limit, + .get_current_clk_freq = smu_v11_0_get_current_clk_freq, + .init_max_sustainable_clocks = smu_v11_0_init_max_sustainable_clocks, + .start_thermal_control = smu_v11_0_start_thermal_control, + .read_sensor = smu_v11_0_read_sensor, + .set_deep_sleep_dcefclk = smu_v11_0_set_deep_sleep_dcefclk, + .display_clock_voltage_request = smu_v11_0_display_clock_voltage_request, + .set_watermarks_for_clock_ranges = smu_v11_0_set_watermarks_for_clock_ranges, + .get_sclk = smu_v11_0_dpm_get_sclk, + .get_mclk = smu_v11_0_dpm_get_mclk, + .set_od8_default_settings = smu_v11_0_set_od8_default_settings, + .conv_power_profile_to_pplib_workload = smu_v11_0_conv_power_profile_to_pplib_workload, + .get_power_profile_mode = smu_v11_0_get_power_profile_mode, + .set_power_profile_mode = smu_v11_0_set_power_profile_mode, + .update_od8_settings = smu_v11_0_update_od8_settings, + .dpm_set_uvd_enable = smu_v11_0_dpm_set_uvd_enable, + .dpm_set_vce_enable = smu_v11_0_dpm_set_vce_enable, + .get_current_rpm = smu_v11_0_get_current_rpm, + .get_fan_control_mode = smu_v11_0_get_fan_control_mode, + .set_fan_control_mode = smu_v11_0_set_fan_control_mode, + .get_fan_speed_percent = smu_v11_0_get_fan_speed_percent, + .set_fan_speed_percent = smu_v11_0_set_fan_speed_percent, + .set_fan_speed_rpm = smu_v11_0_set_fan_speed_rpm, + .set_xgmi_pstate = smu_v11_0_set_xgmi_pstate, +}; + +void smu_v11_0_set_smu_funcs(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + + smu->funcs = &smu_v11_0_funcs; + switch (adev->asic_type) { + case CHIP_VEGA20: + vega20_set_ppt_funcs(smu); + break; + default: + pr_warn("Unknown asic for smu11\n"); + } +} diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c index d111dd4e03d7..6d11076a79ba 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c @@ -212,6 +212,10 @@ static int smu10_start_smu(struct pp_hwmgr *hwmgr) hwmgr->smu_version = smu10_read_arg_from_smc(hwmgr); adev->pm.fw_version = hwmgr->smu_version >> 8; + if (adev->rev_id < 0x8 && adev->pdev->device != 0x15d8 && + adev->pm.fw_version < 0x1e45) + adev->pm.pp_feature &= ~PP_GFXOFF_MASK; + if (smu10_verify_smc_interface(hwmgr)) return -EINVAL; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.c index ba00744c3413..f301a73f6df1 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.c @@ -367,6 +367,26 @@ static int vega20_set_tools_address(struct pp_hwmgr *hwmgr) return ret; } +int vega20_set_pptable_driver_address(struct pp_hwmgr *hwmgr) +{ + struct vega20_smumgr *priv = + (struct vega20_smumgr *)(hwmgr->smu_backend); + int ret = 0; + + PP_ASSERT_WITH_CODE((ret = vega20_send_msg_to_smc_with_parameter(hwmgr, + PPSMC_MSG_SetDriverDramAddrHigh, + upper_32_bits(priv->smu_tables.entry[TABLE_PPTABLE].mc_addr))) == 0, + "[SetPPtabeDriverAddress] Attempt to Set Dram Addr High Failed!", + return ret); + PP_ASSERT_WITH_CODE((ret = vega20_send_msg_to_smc_with_parameter(hwmgr, + PPSMC_MSG_SetDriverDramAddrLow, + lower_32_bits(priv->smu_tables.entry[TABLE_PPTABLE].mc_addr))) == 0, + "[SetPPtabeDriverAddress] Attempt to Set Dram Addr Low Failed!", + return ret); + + return ret; +} + static int vega20_smu_init(struct pp_hwmgr *hwmgr) { struct vega20_smumgr *priv; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.h index 77349c3f0162..ec953ab13e87 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.h @@ -55,6 +55,7 @@ int vega20_set_activity_monitor_coeff(struct pp_hwmgr *hwmgr, uint8_t *table, uint16_t workload_type); int vega20_get_activity_monitor_coeff(struct pp_hwmgr *hwmgr, uint8_t *table, uint16_t workload_type); +int vega20_set_pptable_driver_address(struct pp_hwmgr *hwmgr); #endif diff --git a/drivers/gpu/drm/amd/powerplay/vega20_ppt.c b/drivers/gpu/drm/amd/powerplay/vega20_ppt.c new file mode 100644 index 000000000000..8fafcbdb1dfd --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/vega20_ppt.c @@ -0,0 +1,2413 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#include "pp_debug.h" +#include +#include "amdgpu.h" +#include "amdgpu_smu.h" +#include "atomfirmware.h" +#include "amdgpu_atomfirmware.h" +#include "smu_v11_0.h" +#include "smu11_driver_if.h" +#include "soc15_common.h" +#include "atom.h" +#include "power_state.h" +#include "vega20_ppt.h" +#include "vega20_pptable.h" +#include "vega20_ppsmc.h" +#include "nbio/nbio_7_4_sh_mask.h" + +#define smnPCIE_LC_SPEED_CNTL 0x11140290 +#define smnPCIE_LC_LINK_WIDTH_CNTL 0x11140288 + +#define MSG_MAP(msg) \ + [SMU_MSG_##msg] = PPSMC_MSG_##msg + +static int vega20_message_map[SMU_MSG_MAX_COUNT] = { + MSG_MAP(TestMessage), + MSG_MAP(GetSmuVersion), + MSG_MAP(GetDriverIfVersion), + MSG_MAP(SetAllowedFeaturesMaskLow), + MSG_MAP(SetAllowedFeaturesMaskHigh), + MSG_MAP(EnableAllSmuFeatures), + MSG_MAP(DisableAllSmuFeatures), + MSG_MAP(EnableSmuFeaturesLow), + MSG_MAP(EnableSmuFeaturesHigh), + MSG_MAP(DisableSmuFeaturesLow), + MSG_MAP(DisableSmuFeaturesHigh), + MSG_MAP(GetEnabledSmuFeaturesLow), + MSG_MAP(GetEnabledSmuFeaturesHigh), + MSG_MAP(SetWorkloadMask), + MSG_MAP(SetPptLimit), + MSG_MAP(SetDriverDramAddrHigh), + MSG_MAP(SetDriverDramAddrLow), + MSG_MAP(SetToolsDramAddrHigh), + MSG_MAP(SetToolsDramAddrLow), + MSG_MAP(TransferTableSmu2Dram), + MSG_MAP(TransferTableDram2Smu), + MSG_MAP(UseDefaultPPTable), + MSG_MAP(UseBackupPPTable), + MSG_MAP(RunBtc), + MSG_MAP(RequestI2CBus), + MSG_MAP(ReleaseI2CBus), + MSG_MAP(SetFloorSocVoltage), + MSG_MAP(SoftReset), + MSG_MAP(StartBacoMonitor), + MSG_MAP(CancelBacoMonitor), + MSG_MAP(EnterBaco), + MSG_MAP(SetSoftMinByFreq), + MSG_MAP(SetSoftMaxByFreq), + MSG_MAP(SetHardMinByFreq), + MSG_MAP(SetHardMaxByFreq), + MSG_MAP(GetMinDpmFreq), + MSG_MAP(GetMaxDpmFreq), + MSG_MAP(GetDpmFreqByIndex), + MSG_MAP(GetDpmClockFreq), + MSG_MAP(GetSsVoltageByDpm), + MSG_MAP(SetMemoryChannelConfig), + MSG_MAP(SetGeminiMode), + MSG_MAP(SetGeminiApertureHigh), + MSG_MAP(SetGeminiApertureLow), + MSG_MAP(SetMinLinkDpmByIndex), + MSG_MAP(OverridePcieParameters), + MSG_MAP(OverDriveSetPercentage), + MSG_MAP(SetMinDeepSleepDcefclk), + MSG_MAP(ReenableAcDcInterrupt), + MSG_MAP(NotifyPowerSource), + MSG_MAP(SetUclkFastSwitch), + MSG_MAP(SetUclkDownHyst), + MSG_MAP(GetCurrentRpm), + MSG_MAP(SetVideoFps), + MSG_MAP(SetTjMax), + MSG_MAP(SetFanTemperatureTarget), + MSG_MAP(PrepareMp1ForUnload), + MSG_MAP(DramLogSetDramAddrHigh), + MSG_MAP(DramLogSetDramAddrLow), + MSG_MAP(DramLogSetDramSize), + MSG_MAP(SetFanMaxRpm), + MSG_MAP(SetFanMinPwm), + MSG_MAP(ConfigureGfxDidt), + MSG_MAP(NumOfDisplays), + MSG_MAP(RemoveMargins), + MSG_MAP(ReadSerialNumTop32), + MSG_MAP(ReadSerialNumBottom32), + MSG_MAP(SetSystemVirtualDramAddrHigh), + MSG_MAP(SetSystemVirtualDramAddrLow), + MSG_MAP(WaflTest), + MSG_MAP(SetFclkGfxClkRatio), + MSG_MAP(AllowGfxOff), + MSG_MAP(DisallowGfxOff), + MSG_MAP(GetPptLimit), + MSG_MAP(GetDcModeMaxDpmFreq), + MSG_MAP(GetDebugData), + MSG_MAP(SetXgmiMode), + MSG_MAP(RunAfllBtc), + MSG_MAP(ExitBaco), + MSG_MAP(PrepareMp1ForReset), + MSG_MAP(PrepareMp1ForShutdown), + MSG_MAP(SetMGpuFanBoostLimitRpm), + MSG_MAP(GetAVFSVoltageByDpm), +}; + +static int vega20_get_smu_msg_index(struct smu_context *smc, uint32_t index) +{ + int val; + + if (index >= SMU_MSG_MAX_COUNT) + return -EINVAL; + + val = vega20_message_map[index]; + if (val > PPSMC_Message_Count) + return -EINVAL; + + return val; +} + +static int vega20_allocate_dpm_context(struct smu_context *smu) +{ + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + + if (smu_dpm->dpm_context) + return -EINVAL; + + smu_dpm->dpm_context = kzalloc(sizeof(struct vega20_dpm_table), + GFP_KERNEL); + if (!smu_dpm->dpm_context) + return -ENOMEM; + + if (smu_dpm->golden_dpm_context) + return -EINVAL; + + smu_dpm->golden_dpm_context = kzalloc(sizeof(struct vega20_dpm_table), + GFP_KERNEL); + if (!smu_dpm->golden_dpm_context) + return -ENOMEM; + + smu_dpm->dpm_context_size = sizeof(struct vega20_dpm_table); + + smu_dpm->dpm_current_power_state = kzalloc(sizeof(struct smu_power_state), + GFP_KERNEL); + if (!smu_dpm->dpm_current_power_state) + return -ENOMEM; + + smu_dpm->dpm_request_power_state = kzalloc(sizeof(struct smu_power_state), + GFP_KERNEL); + if (!smu_dpm->dpm_request_power_state) + return -ENOMEM; + + return 0; +} + +static int vega20_setup_od8_information(struct smu_context *smu) +{ + ATOM_Vega20_POWERPLAYTABLE *powerplay_table = NULL; + struct smu_table_context *table_context = &smu->smu_table; + + uint32_t od_feature_count, od_feature_array_size, + od_setting_count, od_setting_array_size; + + if (!table_context->power_play_table) + return -EINVAL; + + powerplay_table = table_context->power_play_table; + + if (powerplay_table->OverDrive8Table.ucODTableRevision == 1) { + /* Setup correct ODFeatureCount, and store ODFeatureArray from + * powerplay table to od_feature_capabilities */ + od_feature_count = + (le32_to_cpu(powerplay_table->OverDrive8Table.ODFeatureCount) > + ATOM_VEGA20_ODFEATURE_COUNT) ? + ATOM_VEGA20_ODFEATURE_COUNT : + le32_to_cpu(powerplay_table->OverDrive8Table.ODFeatureCount); + + od_feature_array_size = sizeof(uint8_t) * od_feature_count; + + if (table_context->od_feature_capabilities) + return -EINVAL; + + table_context->od_feature_capabilities = kmemdup(&powerplay_table->OverDrive8Table.ODFeatureCapabilities, + od_feature_array_size, + GFP_KERNEL); + if (!table_context->od_feature_capabilities) + return -ENOMEM; + + /* Setup correct ODSettingCount, and store ODSettingArray from + * powerplay table to od_settings_max and od_setting_min */ + od_setting_count = + (le32_to_cpu(powerplay_table->OverDrive8Table.ODSettingCount) > + ATOM_VEGA20_ODSETTING_COUNT) ? + ATOM_VEGA20_ODSETTING_COUNT : + le32_to_cpu(powerplay_table->OverDrive8Table.ODSettingCount); + + od_setting_array_size = sizeof(uint32_t) * od_setting_count; + + if (table_context->od_settings_max) + return -EINVAL; + + table_context->od_settings_max = kmemdup(&powerplay_table->OverDrive8Table.ODSettingsMax, + od_setting_array_size, + GFP_KERNEL); + + if (!table_context->od_settings_max) { + kfree(table_context->od_feature_capabilities); + table_context->od_feature_capabilities = NULL; + return -ENOMEM; + } + + if (table_context->od_settings_min) + return -EINVAL; + + table_context->od_settings_min = kmemdup(&powerplay_table->OverDrive8Table.ODSettingsMin, + od_setting_array_size, + GFP_KERNEL); + + if (!table_context->od_settings_min) { + kfree(table_context->od_feature_capabilities); + table_context->od_feature_capabilities = NULL; + kfree(table_context->od_settings_max); + table_context->od_settings_max = NULL; + return -ENOMEM; + } + } + + return 0; +} + +static int vega20_store_powerplay_table(struct smu_context *smu) +{ + ATOM_Vega20_POWERPLAYTABLE *powerplay_table = NULL; + struct smu_table_context *table_context = &smu->smu_table; + int ret; + + if (!table_context->power_play_table) + return -EINVAL; + + powerplay_table = table_context->power_play_table; + + memcpy(table_context->driver_pptable, &powerplay_table->smcPPTable, + sizeof(PPTable_t)); + + table_context->software_shutdown_temp = powerplay_table->usSoftwareShutdownTemp; + table_context->thermal_controller_type = powerplay_table->ucThermalControllerType; + table_context->TDPODLimit = le32_to_cpu(powerplay_table->OverDrive8Table.ODSettingsMax[ATOM_VEGA20_ODSETTING_POWERPERCENTAGE]); + + ret = vega20_setup_od8_information(smu); + + return ret; +} + +static int vega20_append_powerplay_table(struct smu_context *smu) +{ + struct smu_table_context *table_context = &smu->smu_table; + PPTable_t *smc_pptable = table_context->driver_pptable; + struct atom_smc_dpm_info_v4_4 *smc_dpm_table; + int index, i, ret; + + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + smc_dpm_info); + + ret = smu_get_atom_data_table(smu, index, NULL, NULL, NULL, + (uint8_t **)&smc_dpm_table); + if (ret) + return ret; + + smc_pptable->MaxVoltageStepGfx = smc_dpm_table->maxvoltagestepgfx; + smc_pptable->MaxVoltageStepSoc = smc_dpm_table->maxvoltagestepsoc; + + smc_pptable->VddGfxVrMapping = smc_dpm_table->vddgfxvrmapping; + smc_pptable->VddSocVrMapping = smc_dpm_table->vddsocvrmapping; + smc_pptable->VddMem0VrMapping = smc_dpm_table->vddmem0vrmapping; + smc_pptable->VddMem1VrMapping = smc_dpm_table->vddmem1vrmapping; + + smc_pptable->GfxUlvPhaseSheddingMask = smc_dpm_table->gfxulvphasesheddingmask; + smc_pptable->SocUlvPhaseSheddingMask = smc_dpm_table->soculvphasesheddingmask; + smc_pptable->ExternalSensorPresent = smc_dpm_table->externalsensorpresent; + + smc_pptable->GfxMaxCurrent = smc_dpm_table->gfxmaxcurrent; + smc_pptable->GfxOffset = smc_dpm_table->gfxoffset; + smc_pptable->Padding_TelemetryGfx = smc_dpm_table->padding_telemetrygfx; + + smc_pptable->SocMaxCurrent = smc_dpm_table->socmaxcurrent; + smc_pptable->SocOffset = smc_dpm_table->socoffset; + smc_pptable->Padding_TelemetrySoc = smc_dpm_table->padding_telemetrysoc; + + smc_pptable->Mem0MaxCurrent = smc_dpm_table->mem0maxcurrent; + smc_pptable->Mem0Offset = smc_dpm_table->mem0offset; + smc_pptable->Padding_TelemetryMem0 = smc_dpm_table->padding_telemetrymem0; + + smc_pptable->Mem1MaxCurrent = smc_dpm_table->mem1maxcurrent; + smc_pptable->Mem1Offset = smc_dpm_table->mem1offset; + smc_pptable->Padding_TelemetryMem1 = smc_dpm_table->padding_telemetrymem1; + + smc_pptable->AcDcGpio = smc_dpm_table->acdcgpio; + smc_pptable->AcDcPolarity = smc_dpm_table->acdcpolarity; + smc_pptable->VR0HotGpio = smc_dpm_table->vr0hotgpio; + smc_pptable->VR0HotPolarity = smc_dpm_table->vr0hotpolarity; + + smc_pptable->VR1HotGpio = smc_dpm_table->vr1hotgpio; + smc_pptable->VR1HotPolarity = smc_dpm_table->vr1hotpolarity; + smc_pptable->Padding1 = smc_dpm_table->padding1; + smc_pptable->Padding2 = smc_dpm_table->padding2; + + smc_pptable->LedPin0 = smc_dpm_table->ledpin0; + smc_pptable->LedPin1 = smc_dpm_table->ledpin1; + smc_pptable->LedPin2 = smc_dpm_table->ledpin2; + + smc_pptable->PllGfxclkSpreadEnabled = smc_dpm_table->pllgfxclkspreadenabled; + smc_pptable->PllGfxclkSpreadPercent = smc_dpm_table->pllgfxclkspreadpercent; + smc_pptable->PllGfxclkSpreadFreq = smc_dpm_table->pllgfxclkspreadfreq; + + smc_pptable->UclkSpreadEnabled = 0; + smc_pptable->UclkSpreadPercent = smc_dpm_table->uclkspreadpercent; + smc_pptable->UclkSpreadFreq = smc_dpm_table->uclkspreadfreq; + + smc_pptable->FclkSpreadEnabled = smc_dpm_table->fclkspreadenabled; + smc_pptable->FclkSpreadPercent = smc_dpm_table->fclkspreadpercent; + smc_pptable->FclkSpreadFreq = smc_dpm_table->fclkspreadfreq; + + smc_pptable->FllGfxclkSpreadEnabled = smc_dpm_table->fllgfxclkspreadenabled; + smc_pptable->FllGfxclkSpreadPercent = smc_dpm_table->fllgfxclkspreadpercent; + smc_pptable->FllGfxclkSpreadFreq = smc_dpm_table->fllgfxclkspreadfreq; + + for (i = 0; i < I2C_CONTROLLER_NAME_COUNT; i++) { + smc_pptable->I2cControllers[i].Enabled = + smc_dpm_table->i2ccontrollers[i].enabled; + smc_pptable->I2cControllers[i].SlaveAddress = + smc_dpm_table->i2ccontrollers[i].slaveaddress; + smc_pptable->I2cControllers[i].ControllerPort = + smc_dpm_table->i2ccontrollers[i].controllerport; + smc_pptable->I2cControllers[i].ThermalThrottler = + smc_dpm_table->i2ccontrollers[i].thermalthrottler; + smc_pptable->I2cControllers[i].I2cProtocol = + smc_dpm_table->i2ccontrollers[i].i2cprotocol; + smc_pptable->I2cControllers[i].I2cSpeed = + smc_dpm_table->i2ccontrollers[i].i2cspeed; + } + + return 0; +} + +static int vega20_check_powerplay_table(struct smu_context *smu) +{ + ATOM_Vega20_POWERPLAYTABLE *powerplay_table = NULL; + struct smu_table_context *table_context = &smu->smu_table; + + powerplay_table = table_context->power_play_table; + + if (powerplay_table->sHeader.format_revision < ATOM_VEGA20_TABLE_REVISION_VEGA20) { + pr_err("Unsupported PPTable format!"); + return -EINVAL; + } + + if (!powerplay_table->sHeader.structuresize) { + pr_err("Invalid PowerPlay Table!"); + return -EINVAL; + } + + return 0; +} + +static int vega20_run_btc_afll(struct smu_context *smu) +{ + return smu_send_smc_msg(smu, SMU_MSG_RunAfllBtc); +} + +static int +vega20_get_unallowed_feature_mask(struct smu_context *smu, + uint32_t *feature_mask, uint32_t num) +{ + if (num > 2) + return -EINVAL; + + feature_mask[0] = 0xE0041C00; + feature_mask[1] = 0xFFFFFFFE; /* bit32~bit63 is Unsupported */ + + return 0; +} + +static enum +amd_pm_state_type vega20_get_current_power_state(struct smu_context *smu) +{ + enum amd_pm_state_type pm_type; + struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm); + + if (!smu_dpm_ctx->dpm_context || + !smu_dpm_ctx->dpm_current_power_state) + return -EINVAL; + + mutex_lock(&(smu->mutex)); + switch (smu_dpm_ctx->dpm_current_power_state->classification.ui_label) { + case SMU_STATE_UI_LABEL_BATTERY: + pm_type = POWER_STATE_TYPE_BATTERY; + break; + case SMU_STATE_UI_LABEL_BALLANCED: + pm_type = POWER_STATE_TYPE_BALANCED; + break; + case SMU_STATE_UI_LABEL_PERFORMANCE: + pm_type = POWER_STATE_TYPE_PERFORMANCE; + break; + default: + if (smu_dpm_ctx->dpm_current_power_state->classification.flags & SMU_STATE_CLASSIFICATION_FLAG_BOOT) + pm_type = POWER_STATE_TYPE_INTERNAL_BOOT; + else + pm_type = POWER_STATE_TYPE_DEFAULT; + break; + } + mutex_unlock(&(smu->mutex)); + + return pm_type; +} + +static int +vega20_set_single_dpm_table(struct smu_context *smu, + struct vega20_single_dpm_table *single_dpm_table, + PPCLK_e clk_id) +{ + int ret = 0; + uint32_t i, num_of_levels = 0, clk; + + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_GetDpmFreqByIndex, + (clk_id << 16 | 0xFF)); + if (ret) { + pr_err("[GetNumOfDpmLevel] failed to get dpm levels!"); + return ret; + } + + smu_read_smc_arg(smu, &num_of_levels); + if (!num_of_levels) { + pr_err("[GetNumOfDpmLevel] number of clk levels is invalid!"); + return -EINVAL; + } + + single_dpm_table->count = num_of_levels; + + for (i = 0; i < num_of_levels; i++) { + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_GetDpmFreqByIndex, + (clk_id << 16 | i)); + if (ret) { + pr_err("[GetDpmFreqByIndex] failed to get dpm freq by index!"); + return ret; + } + smu_read_smc_arg(smu, &clk); + if (!clk) { + pr_err("[GetDpmFreqByIndex] clk value is invalid!"); + return -EINVAL; + } + single_dpm_table->dpm_levels[i].value = clk; + single_dpm_table->dpm_levels[i].enabled = true; + } + return 0; +} + +static void vega20_init_single_dpm_state(struct vega20_dpm_state *dpm_state) +{ + dpm_state->soft_min_level = 0x0; + dpm_state->soft_max_level = 0xffff; + dpm_state->hard_min_level = 0x0; + dpm_state->hard_max_level = 0xffff; +} + +static int vega20_set_default_dpm_table(struct smu_context *smu) +{ + int ret; + + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + struct vega20_dpm_table *dpm_table = NULL; + struct vega20_single_dpm_table *single_dpm_table; + + dpm_table = smu_dpm->dpm_context; + + /* socclk */ + single_dpm_table = &(dpm_table->soc_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_SOCCLK_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, + PPCLK_SOCCLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get socclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 1; + single_dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.socclk / 100; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + /* gfxclk */ + single_dpm_table = &(dpm_table->gfx_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_GFXCLK_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, + PPCLK_GFXCLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get gfxclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 1; + single_dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.gfxclk / 100; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + /* memclk */ + single_dpm_table = &(dpm_table->mem_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_UCLK_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, + PPCLK_UCLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get memclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 1; + single_dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.uclk / 100; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + /* eclk */ + single_dpm_table = &(dpm_table->eclk_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_VCE_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, PPCLK_ECLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get eclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 1; + single_dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.eclk / 100; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + /* vclk */ + single_dpm_table = &(dpm_table->vclk_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_UVD_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, PPCLK_VCLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get vclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 1; + single_dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.vclk / 100; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + /* dclk */ + single_dpm_table = &(dpm_table->dclk_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_UVD_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, PPCLK_DCLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get dclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 1; + single_dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.dclk / 100; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + /* dcefclk */ + single_dpm_table = &(dpm_table->dcef_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_DCEFCLK_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, + PPCLK_DCEFCLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get dcefclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 1; + single_dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.dcefclk / 100; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + /* pixclk */ + single_dpm_table = &(dpm_table->pixel_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_DCEFCLK_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, + PPCLK_PIXCLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get pixclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 0; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + /* dispclk */ + single_dpm_table = &(dpm_table->display_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_DCEFCLK_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, + PPCLK_DISPCLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get dispclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 0; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + /* phyclk */ + single_dpm_table = &(dpm_table->phy_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_DCEFCLK_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, + PPCLK_PHYCLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get phyclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 0; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + /* fclk */ + single_dpm_table = &(dpm_table->fclk_table); + + if (smu_feature_is_enabled(smu,FEATURE_DPM_FCLK_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, + PPCLK_FCLK); + if (ret) { + pr_err("[SetupDefaultDpmTable] failed to get fclk dpm levels!"); + return ret; + } + } else { + single_dpm_table->count = 0; + } + vega20_init_single_dpm_state(&(single_dpm_table->dpm_state)); + + memcpy(smu_dpm->golden_dpm_context, dpm_table, + sizeof(struct vega20_dpm_table)); + + return 0; +} + +static int vega20_populate_umd_state_clk(struct smu_context *smu) +{ + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + struct vega20_dpm_table *dpm_table = NULL; + struct vega20_single_dpm_table *gfx_table = NULL; + struct vega20_single_dpm_table *mem_table = NULL; + + dpm_table = smu_dpm->dpm_context; + gfx_table = &(dpm_table->gfx_table); + mem_table = &(dpm_table->mem_table); + + smu->pstate_sclk = gfx_table->dpm_levels[0].value; + smu->pstate_mclk = mem_table->dpm_levels[0].value; + + if (gfx_table->count > VEGA20_UMD_PSTATE_GFXCLK_LEVEL && + mem_table->count > VEGA20_UMD_PSTATE_MCLK_LEVEL) { + smu->pstate_sclk = gfx_table->dpm_levels[VEGA20_UMD_PSTATE_GFXCLK_LEVEL].value; + smu->pstate_mclk = mem_table->dpm_levels[VEGA20_UMD_PSTATE_MCLK_LEVEL].value; + } + + smu->pstate_sclk = smu->pstate_sclk * 100; + smu->pstate_mclk = smu->pstate_mclk * 100; + + return 0; +} + +static int vega20_get_clk_table(struct smu_context *smu, + struct pp_clock_levels_with_latency *clocks, + struct vega20_single_dpm_table *dpm_table) +{ + int i, count; + + count = (dpm_table->count > MAX_NUM_CLOCKS) ? MAX_NUM_CLOCKS : dpm_table->count; + clocks->num_levels = count; + + for (i = 0; i < count; i++) { + clocks->data[i].clocks_in_khz = + dpm_table->dpm_levels[i].value * 1000; + clocks->data[i].latency_in_us = 0; + } + + return 0; +} + +static int vega20_print_clk_levels(struct smu_context *smu, + enum pp_clock_type type, char *buf) +{ + int i, now, size = 0; + int ret = 0; + uint32_t gen_speed, lane_width; + struct amdgpu_device *adev = smu->adev; + struct pp_clock_levels_with_latency clocks; + struct vega20_single_dpm_table *single_dpm_table; + struct smu_table_context *table_context = &smu->smu_table; + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + struct vega20_dpm_table *dpm_table = NULL; + struct vega20_od8_settings *od8_settings = + (struct vega20_od8_settings *)table_context->od8_settings; + OverDriveTable_t *od_table = + (OverDriveTable_t *)(table_context->overdrive_table); + PPTable_t *pptable = (PPTable_t *)table_context->driver_pptable; + + dpm_table = smu_dpm->dpm_context; + + switch (type) { + case PP_SCLK: + ret = smu_get_current_clk_freq(smu, PPCLK_GFXCLK, &now); + if (ret) { + pr_err("Attempt to get current gfx clk Failed!"); + return ret; + } + + single_dpm_table = &(dpm_table->gfx_table); + ret = vega20_get_clk_table(smu, &clocks, single_dpm_table); + if (ret) { + pr_err("Attempt to get gfx clk levels Failed!"); + return ret; + } + + for (i = 0; i < clocks.num_levels; i++) + size += sprintf(buf + size, "%d: %uMhz %s\n", i, + clocks.data[i].clocks_in_khz / 1000, + (clocks.data[i].clocks_in_khz == now * 10) + ? "*" : ""); + break; + + case PP_MCLK: + ret = smu_get_current_clk_freq(smu, PPCLK_UCLK, &now); + if (ret) { + pr_err("Attempt to get current mclk Failed!"); + return ret; + } + + single_dpm_table = &(dpm_table->mem_table); + ret = vega20_get_clk_table(smu, &clocks, single_dpm_table); + if (ret) { + pr_err("Attempt to get memory clk levels Failed!"); + return ret; + } + + for (i = 0; i < clocks.num_levels; i++) + size += sprintf(buf + size, "%d: %uMhz %s\n", + i, clocks.data[i].clocks_in_khz / 1000, + (clocks.data[i].clocks_in_khz == now * 10) + ? "*" : ""); + break; + + case PP_SOCCLK: + ret = smu_get_current_clk_freq(smu, PPCLK_SOCCLK, &now); + if (ret) { + pr_err("Attempt to get current socclk Failed!"); + return ret; + } + + single_dpm_table = &(dpm_table->soc_table); + ret = vega20_get_clk_table(smu, &clocks, single_dpm_table); + if (ret) { + pr_err("Attempt to get socclk levels Failed!"); + return ret; + } + + for (i = 0; i < clocks.num_levels; i++) + size += sprintf(buf + size, "%d: %uMhz %s\n", + i, clocks.data[i].clocks_in_khz / 1000, + (clocks.data[i].clocks_in_khz == now * 10) + ? "*" : ""); + break; + + case PP_FCLK: + ret = smu_get_current_clk_freq(smu, PPCLK_FCLK, &now); + if (ret) { + pr_err("Attempt to get current fclk Failed!"); + return ret; + } + + single_dpm_table = &(dpm_table->fclk_table); + for (i = 0; i < single_dpm_table->count; i++) + size += sprintf(buf + size, "%d: %uMhz %s\n", + i, single_dpm_table->dpm_levels[i].value, + (single_dpm_table->dpm_levels[i].value == now / 100) + ? "*" : ""); + break; + + case PP_DCEFCLK: + ret = smu_get_current_clk_freq(smu, PPCLK_DCEFCLK, &now); + if (ret) { + pr_err("Attempt to get current dcefclk Failed!"); + return ret; + } + + single_dpm_table = &(dpm_table->dcef_table); + ret = vega20_get_clk_table(smu, &clocks, single_dpm_table); + if (ret) { + pr_err("Attempt to get dcefclk levels Failed!"); + return ret; + } + + for (i = 0; i < clocks.num_levels; i++) + size += sprintf(buf + size, "%d: %uMhz %s\n", + i, clocks.data[i].clocks_in_khz / 1000, + (clocks.data[i].clocks_in_khz == now * 10) ? "*" : ""); + break; + + case PP_PCIE: + gen_speed = (RREG32_PCIE(smnPCIE_LC_SPEED_CNTL) & + PSWUSP0_PCIE_LC_SPEED_CNTL__LC_CURRENT_DATA_RATE_MASK) + >> PSWUSP0_PCIE_LC_SPEED_CNTL__LC_CURRENT_DATA_RATE__SHIFT; + lane_width = (RREG32_PCIE(smnPCIE_LC_LINK_WIDTH_CNTL) & + PCIE_LC_LINK_WIDTH_CNTL__LC_LINK_WIDTH_RD_MASK) + >> PCIE_LC_LINK_WIDTH_CNTL__LC_LINK_WIDTH_RD__SHIFT; + for (i = 0; i < NUM_LINK_LEVELS; i++) + size += sprintf(buf + size, "%d: %s %s %dMhz %s\n", i, + (pptable->PcieGenSpeed[i] == 0) ? "2.5GT/s," : + (pptable->PcieGenSpeed[i] == 1) ? "5.0GT/s," : + (pptable->PcieGenSpeed[i] == 2) ? "8.0GT/s," : + (pptable->PcieGenSpeed[i] == 3) ? "16.0GT/s," : "", + (pptable->PcieLaneCount[i] == 1) ? "x1" : + (pptable->PcieLaneCount[i] == 2) ? "x2" : + (pptable->PcieLaneCount[i] == 3) ? "x4" : + (pptable->PcieLaneCount[i] == 4) ? "x8" : + (pptable->PcieLaneCount[i] == 5) ? "x12" : + (pptable->PcieLaneCount[i] == 6) ? "x16" : "", + pptable->LclkFreq[i], + (gen_speed == pptable->PcieGenSpeed[i]) && + (lane_width == pptable->PcieLaneCount[i]) ? + "*" : ""); + break; + + case OD_SCLK: + if (od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMIN].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMAX].feature_id) { + size = sprintf(buf, "%s:\n", "OD_SCLK"); + size += sprintf(buf + size, "0: %10uMhz\n", + od_table->GfxclkFmin); + size += sprintf(buf + size, "1: %10uMhz\n", + od_table->GfxclkFmax); + } + + break; + + case OD_MCLK: + if (od8_settings->od8_settings_array[OD8_SETTING_UCLK_FMAX].feature_id) { + size = sprintf(buf, "%s:\n", "OD_MCLK"); + size += sprintf(buf + size, "1: %10uMhz\n", + od_table->UclkFmax); + } + + break; + + case OD_VDDC_CURVE: + if (od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ1].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ2].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ3].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE1].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE2].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE3].feature_id) { + size = sprintf(buf, "%s:\n", "OD_VDDC_CURVE"); + size += sprintf(buf + size, "0: %10uMhz %10dmV\n", + od_table->GfxclkFreq1, + od_table->GfxclkVolt1 / VOLTAGE_SCALE); + size += sprintf(buf + size, "1: %10uMhz %10dmV\n", + od_table->GfxclkFreq2, + od_table->GfxclkVolt2 / VOLTAGE_SCALE); + size += sprintf(buf + size, "2: %10uMhz %10dmV\n", + od_table->GfxclkFreq3, + od_table->GfxclkVolt3 / VOLTAGE_SCALE); + } + + break; + + case OD_RANGE: + size = sprintf(buf, "%s:\n", "OD_RANGE"); + + if (od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMIN].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMAX].feature_id) { + size += sprintf(buf + size, "SCLK: %7uMhz %10uMhz\n", + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMIN].min_value, + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMAX].max_value); + } + + if (od8_settings->od8_settings_array[OD8_SETTING_UCLK_FMAX].feature_id) { + single_dpm_table = &(dpm_table->mem_table); + ret = vega20_get_clk_table(smu, &clocks, single_dpm_table); + if (ret) { + pr_err("Attempt to get memory clk levels Failed!"); + return ret; + } + + size += sprintf(buf + size, "MCLK: %7uMhz %10uMhz\n", + clocks.data[0].clocks_in_khz / 1000, + od8_settings->od8_settings_array[OD8_SETTING_UCLK_FMAX].max_value); + } + + if (od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ1].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ2].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ3].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE1].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE2].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE3].feature_id) { + size += sprintf(buf + size, "VDDC_CURVE_SCLK[0]: %7uMhz %10uMhz\n", + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ1].min_value, + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ1].max_value); + size += sprintf(buf + size, "VDDC_CURVE_VOLT[0]: %7dmV %11dmV\n", + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE1].min_value, + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE1].max_value); + size += sprintf(buf + size, "VDDC_CURVE_SCLK[1]: %7uMhz %10uMhz\n", + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ2].min_value, + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ2].max_value); + size += sprintf(buf + size, "VDDC_CURVE_VOLT[1]: %7dmV %11dmV\n", + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE2].min_value, + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE2].max_value); + size += sprintf(buf + size, "VDDC_CURVE_SCLK[2]: %7uMhz %10uMhz\n", + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ3].min_value, + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ3].max_value); + size += sprintf(buf + size, "VDDC_CURVE_VOLT[2]: %7dmV %11dmV\n", + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE3].min_value, + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE3].max_value); + } + + break; + + default: + break; + } + return size; +} + +static int vega20_upload_dpm_level(struct smu_context *smu, bool max, + uint32_t feature_mask) +{ + struct vega20_dpm_table *dpm_table; + struct vega20_single_dpm_table *single_dpm_table; + uint32_t freq; + int ret = 0; + + dpm_table = smu->smu_dpm.dpm_context; + + if (smu_feature_is_enabled(smu, FEATURE_DPM_GFXCLK_BIT) && + (feature_mask & FEATURE_DPM_GFXCLK_MASK)) { + single_dpm_table = &(dpm_table->gfx_table); + freq = max ? single_dpm_table->dpm_state.soft_max_level : + single_dpm_table->dpm_state.soft_min_level; + ret = smu_send_smc_msg_with_param(smu, + (max ? SMU_MSG_SetSoftMaxByFreq : SMU_MSG_SetSoftMinByFreq), + (PPCLK_GFXCLK << 16) | (freq & 0xffff)); + if (ret) { + pr_err("Failed to set soft %s gfxclk !\n", + max ? "max" : "min"); + return ret; + } + } + + if (smu_feature_is_enabled(smu, FEATURE_DPM_UCLK_BIT) && + (feature_mask & FEATURE_DPM_UCLK_MASK)) { + single_dpm_table = &(dpm_table->mem_table); + freq = max ? single_dpm_table->dpm_state.soft_max_level : + single_dpm_table->dpm_state.soft_min_level; + ret = smu_send_smc_msg_with_param(smu, + (max ? SMU_MSG_SetSoftMaxByFreq : SMU_MSG_SetSoftMinByFreq), + (PPCLK_UCLK << 16) | (freq & 0xffff)); + if (ret) { + pr_err("Failed to set soft %s memclk !\n", + max ? "max" : "min"); + return ret; + } + } + + if (smu_feature_is_enabled(smu, FEATURE_DPM_SOCCLK_BIT) && + (feature_mask & FEATURE_DPM_SOCCLK_MASK)) { + single_dpm_table = &(dpm_table->soc_table); + freq = max ? single_dpm_table->dpm_state.soft_max_level : + single_dpm_table->dpm_state.soft_min_level; + ret = smu_send_smc_msg_with_param(smu, + (max ? SMU_MSG_SetSoftMaxByFreq : SMU_MSG_SetSoftMinByFreq), + (PPCLK_SOCCLK << 16) | (freq & 0xffff)); + if (ret) { + pr_err("Failed to set soft %s socclk !\n", + max ? "max" : "min"); + return ret; + } + } + + if (smu_feature_is_enabled(smu, FEATURE_DPM_FCLK_BIT) && + (feature_mask & FEATURE_DPM_FCLK_MASK)) { + single_dpm_table = &(dpm_table->fclk_table); + freq = max ? single_dpm_table->dpm_state.soft_max_level : + single_dpm_table->dpm_state.soft_min_level; + ret = smu_send_smc_msg_with_param(smu, + (max ? SMU_MSG_SetSoftMaxByFreq : SMU_MSG_SetSoftMinByFreq), + (PPCLK_FCLK << 16) | (freq & 0xffff)); + if (ret) { + pr_err("Failed to set soft %s fclk !\n", + max ? "max" : "min"); + return ret; + } + } + + if (smu_feature_is_enabled(smu, FEATURE_DPM_DCEFCLK_BIT) && + (feature_mask & FEATURE_DPM_DCEFCLK_MASK)) { + single_dpm_table = &(dpm_table->dcef_table); + freq = single_dpm_table->dpm_state.hard_min_level; + if (!max) { + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_SetHardMinByFreq, + (PPCLK_DCEFCLK << 16) | (freq & 0xffff)); + if (ret) { + pr_err("Failed to set hard min dcefclk !\n"); + return ret; + } + } + } + + return ret; +} + +static int vega20_force_clk_levels(struct smu_context *smu, + enum pp_clock_type type, uint32_t mask) +{ + struct vega20_dpm_table *dpm_table; + struct vega20_single_dpm_table *single_dpm_table; + uint32_t soft_min_level, soft_max_level, hard_min_level; + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + int ret = 0; + + if (smu_dpm->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL) { + pr_info("force clock level is for dpm manual mode only.\n"); + return -EINVAL; + } + + mutex_lock(&(smu->mutex)); + + soft_min_level = mask ? (ffs(mask) - 1) : 0; + soft_max_level = mask ? (fls(mask) - 1) : 0; + + dpm_table = smu->smu_dpm.dpm_context; + + switch (type) { + case PP_SCLK: + single_dpm_table = &(dpm_table->gfx_table); + + if (soft_max_level >= single_dpm_table->count) { + pr_err("Clock level specified %d is over max allowed %d\n", + soft_max_level, single_dpm_table->count - 1); + ret = -EINVAL; + break; + } + + single_dpm_table->dpm_state.soft_min_level = + single_dpm_table->dpm_levels[soft_min_level].value; + single_dpm_table->dpm_state.soft_max_level = + single_dpm_table->dpm_levels[soft_max_level].value; + + ret = vega20_upload_dpm_level(smu, false, FEATURE_DPM_GFXCLK_MASK); + if (ret) { + pr_err("Failed to upload boot level to lowest!\n"); + break; + } + + ret = vega20_upload_dpm_level(smu, true, FEATURE_DPM_GFXCLK_MASK); + if (ret) + pr_err("Failed to upload dpm max level to highest!\n"); + + break; + + case PP_MCLK: + single_dpm_table = &(dpm_table->mem_table); + + if (soft_max_level >= single_dpm_table->count) { + pr_err("Clock level specified %d is over max allowed %d\n", + soft_max_level, single_dpm_table->count - 1); + ret = -EINVAL; + break; + } + + single_dpm_table->dpm_state.soft_min_level = + single_dpm_table->dpm_levels[soft_min_level].value; + single_dpm_table->dpm_state.soft_max_level = + single_dpm_table->dpm_levels[soft_max_level].value; + + ret = vega20_upload_dpm_level(smu, false, FEATURE_DPM_UCLK_MASK); + if (ret) { + pr_err("Failed to upload boot level to lowest!\n"); + break; + } + + ret = vega20_upload_dpm_level(smu, true, FEATURE_DPM_UCLK_MASK); + if (ret) + pr_err("Failed to upload dpm max level to highest!\n"); + + break; + + case PP_SOCCLK: + single_dpm_table = &(dpm_table->soc_table); + + if (soft_max_level >= single_dpm_table->count) { + pr_err("Clock level specified %d is over max allowed %d\n", + soft_max_level, single_dpm_table->count - 1); + ret = -EINVAL; + break; + } + + single_dpm_table->dpm_state.soft_min_level = + single_dpm_table->dpm_levels[soft_min_level].value; + single_dpm_table->dpm_state.soft_max_level = + single_dpm_table->dpm_levels[soft_max_level].value; + + ret = vega20_upload_dpm_level(smu, false, FEATURE_DPM_SOCCLK_MASK); + if (ret) { + pr_err("Failed to upload boot level to lowest!\n"); + break; + } + + ret = vega20_upload_dpm_level(smu, true, FEATURE_DPM_SOCCLK_MASK); + if (ret) + pr_err("Failed to upload dpm max level to highest!\n"); + + break; + + case PP_FCLK: + single_dpm_table = &(dpm_table->fclk_table); + + if (soft_max_level >= single_dpm_table->count) { + pr_err("Clock level specified %d is over max allowed %d\n", + soft_max_level, single_dpm_table->count - 1); + ret = -EINVAL; + break; + } + + single_dpm_table->dpm_state.soft_min_level = + single_dpm_table->dpm_levels[soft_min_level].value; + single_dpm_table->dpm_state.soft_max_level = + single_dpm_table->dpm_levels[soft_max_level].value; + + ret = vega20_upload_dpm_level(smu, false, FEATURE_DPM_FCLK_MASK); + if (ret) { + pr_err("Failed to upload boot level to lowest!\n"); + break; + } + + ret = vega20_upload_dpm_level(smu, true, FEATURE_DPM_FCLK_MASK); + if (ret) + pr_err("Failed to upload dpm max level to highest!\n"); + + break; + + case PP_DCEFCLK: + hard_min_level = soft_min_level; + single_dpm_table = &(dpm_table->dcef_table); + + if (hard_min_level >= single_dpm_table->count) { + pr_err("Clock level specified %d is over max allowed %d\n", + hard_min_level, single_dpm_table->count - 1); + ret = -EINVAL; + break; + } + + single_dpm_table->dpm_state.hard_min_level = + single_dpm_table->dpm_levels[hard_min_level].value; + + ret = vega20_upload_dpm_level(smu, false, FEATURE_DPM_DCEFCLK_MASK); + if (ret) + pr_err("Failed to upload boot level to lowest!\n"); + + break; + + case PP_PCIE: + if (soft_min_level >= NUM_LINK_LEVELS || + soft_max_level >= NUM_LINK_LEVELS) { + ret = -EINVAL; + break; + } + + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_SetMinLinkDpmByIndex, soft_min_level); + if (ret) + pr_err("Failed to set min link dpm level!\n"); + + break; + + default: + break; + } + + mutex_unlock(&(smu->mutex)); + return ret; +} + +static int vega20_get_clock_by_type_with_latency(struct smu_context *smu, + enum amd_pp_clock_type type, + struct pp_clock_levels_with_latency *clocks) +{ + int ret; + struct vega20_single_dpm_table *single_dpm_table; + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + struct vega20_dpm_table *dpm_table = NULL; + + dpm_table = smu_dpm->dpm_context; + + mutex_lock(&smu->mutex); + + switch (type) { + case amd_pp_sys_clock: + single_dpm_table = &(dpm_table->gfx_table); + ret = vega20_get_clk_table(smu, clocks, single_dpm_table); + break; + case amd_pp_mem_clock: + single_dpm_table = &(dpm_table->mem_table); + ret = vega20_get_clk_table(smu, clocks, single_dpm_table); + break; + case amd_pp_dcef_clock: + single_dpm_table = &(dpm_table->dcef_table); + ret = vega20_get_clk_table(smu, clocks, single_dpm_table); + break; + case amd_pp_soc_clock: + single_dpm_table = &(dpm_table->soc_table); + ret = vega20_get_clk_table(smu, clocks, single_dpm_table); + break; + default: + ret = -EINVAL; + } + + mutex_unlock(&smu->mutex); + return ret; +} + +static int vega20_overdrive_get_gfx_clk_base_voltage(struct smu_context *smu, + uint32_t *voltage, + uint32_t freq) +{ + int ret; + + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_GetAVFSVoltageByDpm, + ((AVFS_CURVE << 24) | (OD8_HOTCURVE_TEMPERATURE << 16) | freq)); + if (ret) { + pr_err("[GetBaseVoltage] failed to get GFXCLK AVFS voltage from SMU!"); + return ret; + } + + smu_read_smc_arg(smu, voltage); + *voltage = *voltage / VOLTAGE_SCALE; + + return 0; +} + +static int vega20_set_default_od8_setttings(struct smu_context *smu) +{ + struct smu_table_context *table_context = &smu->smu_table; + OverDriveTable_t *od_table = (OverDriveTable_t *)(table_context->overdrive_table); + struct vega20_od8_settings *od8_settings = NULL; + PPTable_t *smc_pptable = table_context->driver_pptable; + int i, ret; + + if (table_context->od8_settings) + return -EINVAL; + + table_context->od8_settings = kzalloc(sizeof(struct vega20_od8_settings), GFP_KERNEL); + + if (!table_context->od8_settings) + return -ENOMEM; + + memset(table_context->od8_settings, 0, sizeof(struct vega20_od8_settings)); + od8_settings = (struct vega20_od8_settings *)table_context->od8_settings; + + if (smu_feature_is_enabled(smu, FEATURE_DPM_SOCCLK_BIT)) { + if (table_context->od_feature_capabilities[ATOM_VEGA20_ODFEATURE_GFXCLK_LIMITS] && + table_context->od_settings_max[OD8_SETTING_GFXCLK_FMAX] > 0 && + table_context->od_settings_min[OD8_SETTING_GFXCLK_FMIN] > 0 && + (table_context->od_settings_max[OD8_SETTING_GFXCLK_FMAX] >= + table_context->od_settings_min[OD8_SETTING_GFXCLK_FMIN])) { + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMIN].feature_id = + OD8_GFXCLK_LIMITS; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMAX].feature_id = + OD8_GFXCLK_LIMITS; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMIN].default_value = + od_table->GfxclkFmin; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMAX].default_value = + od_table->GfxclkFmax; + } + + if (table_context->od_feature_capabilities[ATOM_VEGA20_ODFEATURE_GFXCLK_CURVE] && + (table_context->od_settings_min[OD8_SETTING_GFXCLK_VOLTAGE1] >= + smc_pptable->MinVoltageGfx / VOLTAGE_SCALE) && + (table_context->od_settings_max[OD8_SETTING_GFXCLK_VOLTAGE3] <= + smc_pptable->MaxVoltageGfx / VOLTAGE_SCALE) && + (table_context->od_settings_min[OD8_SETTING_GFXCLK_VOLTAGE1] <= + table_context->od_settings_max[OD8_SETTING_GFXCLK_VOLTAGE3])) { + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ1].feature_id = + OD8_GFXCLK_CURVE; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE1].feature_id = + OD8_GFXCLK_CURVE; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ2].feature_id = + OD8_GFXCLK_CURVE; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE2].feature_id = + OD8_GFXCLK_CURVE; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ3].feature_id = + OD8_GFXCLK_CURVE; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE3].feature_id = + OD8_GFXCLK_CURVE; + + od_table->GfxclkFreq1 = od_table->GfxclkFmin; + od_table->GfxclkFreq2 = (od_table->GfxclkFmin + od_table->GfxclkFmax) / 2; + od_table->GfxclkFreq3 = od_table->GfxclkFmax; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ1].default_value = + od_table->GfxclkFreq1; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ2].default_value = + od_table->GfxclkFreq2; + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ3].default_value = + od_table->GfxclkFreq3; + + ret = vega20_overdrive_get_gfx_clk_base_voltage(smu, + &od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE1].default_value, + od_table->GfxclkFreq1); + if (ret) + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE1].default_value = 0; + od_table->GfxclkVolt1 = + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE1].default_value + * VOLTAGE_SCALE; + ret = vega20_overdrive_get_gfx_clk_base_voltage(smu, + &od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE2].default_value, + od_table->GfxclkFreq2); + if (ret) + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE2].default_value = 0; + od_table->GfxclkVolt2 = + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE2].default_value + * VOLTAGE_SCALE; + ret = vega20_overdrive_get_gfx_clk_base_voltage(smu, + &od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE3].default_value, + od_table->GfxclkFreq3); + if (ret) + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE3].default_value = 0; + od_table->GfxclkVolt3 = + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE3].default_value + * VOLTAGE_SCALE; + } + } + + if (smu_feature_is_enabled(smu, FEATURE_DPM_UCLK_BIT)) { + if (table_context->od_feature_capabilities[ATOM_VEGA20_ODFEATURE_UCLK_MAX] && + table_context->od_settings_min[OD8_SETTING_UCLK_FMAX] > 0 && + table_context->od_settings_max[OD8_SETTING_UCLK_FMAX] > 0 && + (table_context->od_settings_max[OD8_SETTING_UCLK_FMAX] >= + table_context->od_settings_min[OD8_SETTING_UCLK_FMAX])) { + od8_settings->od8_settings_array[OD8_SETTING_UCLK_FMAX].feature_id = + OD8_UCLK_MAX; + od8_settings->od8_settings_array[OD8_SETTING_UCLK_FMAX].default_value = + od_table->UclkFmax; + } + } + + if (table_context->od_feature_capabilities[ATOM_VEGA20_ODFEATURE_POWER_LIMIT] && + table_context->od_settings_min[OD8_SETTING_POWER_PERCENTAGE] > 0 && + table_context->od_settings_min[OD8_SETTING_POWER_PERCENTAGE] <= 100 && + table_context->od_settings_max[OD8_SETTING_POWER_PERCENTAGE] > 0 && + table_context->od_settings_max[OD8_SETTING_POWER_PERCENTAGE] <= 100) { + od8_settings->od8_settings_array[OD8_SETTING_POWER_PERCENTAGE].feature_id = + OD8_POWER_LIMIT; + od8_settings->od8_settings_array[OD8_SETTING_POWER_PERCENTAGE].default_value = + od_table->OverDrivePct; + } + + if (smu_feature_is_enabled(smu, FEATURE_FAN_CONTROL_BIT)) { + if (table_context->od_feature_capabilities[ATOM_VEGA20_ODFEATURE_FAN_ACOUSTIC_LIMIT] && + table_context->od_settings_min[OD8_SETTING_FAN_ACOUSTIC_LIMIT] > 0 && + table_context->od_settings_max[OD8_SETTING_FAN_ACOUSTIC_LIMIT] > 0 && + (table_context->od_settings_max[OD8_SETTING_FAN_ACOUSTIC_LIMIT] >= + table_context->od_settings_min[OD8_SETTING_FAN_ACOUSTIC_LIMIT])) { + od8_settings->od8_settings_array[OD8_SETTING_FAN_ACOUSTIC_LIMIT].feature_id = + OD8_ACOUSTIC_LIMIT_SCLK; + od8_settings->od8_settings_array[OD8_SETTING_FAN_ACOUSTIC_LIMIT].default_value = + od_table->FanMaximumRpm; + } + + if (table_context->od_feature_capabilities[ATOM_VEGA20_ODFEATURE_FAN_SPEED_MIN] && + table_context->od_settings_min[OD8_SETTING_FAN_MIN_SPEED] > 0 && + table_context->od_settings_max[OD8_SETTING_FAN_MIN_SPEED] > 0 && + (table_context->od_settings_max[OD8_SETTING_FAN_MIN_SPEED] >= + table_context->od_settings_min[OD8_SETTING_FAN_MIN_SPEED])) { + od8_settings->od8_settings_array[OD8_SETTING_FAN_MIN_SPEED].feature_id = + OD8_FAN_SPEED_MIN; + od8_settings->od8_settings_array[OD8_SETTING_FAN_MIN_SPEED].default_value = + od_table->FanMinimumPwm * smc_pptable->FanMaximumRpm / 100; + } + } + + if (smu_feature_is_enabled(smu, FEATURE_THERMAL_BIT)) { + if (table_context->od_feature_capabilities[ATOM_VEGA20_ODFEATURE_TEMPERATURE_FAN] && + table_context->od_settings_min[OD8_SETTING_FAN_TARGET_TEMP] > 0 && + table_context->od_settings_max[OD8_SETTING_FAN_TARGET_TEMP] > 0 && + (table_context->od_settings_max[OD8_SETTING_FAN_TARGET_TEMP] >= + table_context->od_settings_min[OD8_SETTING_FAN_TARGET_TEMP])) { + od8_settings->od8_settings_array[OD8_SETTING_FAN_TARGET_TEMP].feature_id = + OD8_TEMPERATURE_FAN; + od8_settings->od8_settings_array[OD8_SETTING_FAN_TARGET_TEMP].default_value = + od_table->FanTargetTemperature; + } + + if (table_context->od_feature_capabilities[ATOM_VEGA20_ODFEATURE_TEMPERATURE_SYSTEM] && + table_context->od_settings_min[OD8_SETTING_OPERATING_TEMP_MAX] > 0 && + table_context->od_settings_max[OD8_SETTING_OPERATING_TEMP_MAX] > 0 && + (table_context->od_settings_max[OD8_SETTING_OPERATING_TEMP_MAX] >= + table_context->od_settings_min[OD8_SETTING_OPERATING_TEMP_MAX])) { + od8_settings->od8_settings_array[OD8_SETTING_OPERATING_TEMP_MAX].feature_id = + OD8_TEMPERATURE_SYSTEM; + od8_settings->od8_settings_array[OD8_SETTING_OPERATING_TEMP_MAX].default_value = + od_table->MaxOpTemp; + } + } + + for (i = 0; i < OD8_SETTING_COUNT; i++) { + if (od8_settings->od8_settings_array[i].feature_id) { + od8_settings->od8_settings_array[i].min_value = + table_context->od_settings_min[i]; + od8_settings->od8_settings_array[i].max_value = + table_context->od_settings_max[i]; + od8_settings->od8_settings_array[i].current_value = + od8_settings->od8_settings_array[i].default_value; + } else { + od8_settings->od8_settings_array[i].min_value = 0; + od8_settings->od8_settings_array[i].max_value = 0; + od8_settings->od8_settings_array[i].current_value = 0; + } + } + + return 0; +} + +static int vega20_get_od_percentage(struct smu_context *smu, + enum pp_clock_type type) +{ + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + struct vega20_dpm_table *dpm_table = NULL; + struct vega20_dpm_table *golden_table = NULL; + struct vega20_single_dpm_table *single_dpm_table; + struct vega20_single_dpm_table *golden_dpm_table; + int value, golden_value; + + dpm_table = smu_dpm->dpm_context; + golden_table = smu_dpm->golden_dpm_context; + + switch (type) { + case OD_SCLK: + single_dpm_table = &(dpm_table->gfx_table); + golden_dpm_table = &(golden_table->gfx_table); + break; + case OD_MCLK: + single_dpm_table = &(dpm_table->mem_table); + golden_dpm_table = &(golden_table->mem_table); + break; + default: + return -EINVAL; + break; + } + + value = single_dpm_table->dpm_levels[single_dpm_table->count - 1].value; + golden_value = golden_dpm_table->dpm_levels[golden_dpm_table->count - 1].value; + + value -= golden_value; + value = DIV_ROUND_UP(value * 100, golden_value); + + return value; +} + +static int +vega20_get_profiling_clk_mask(struct smu_context *smu, + enum amd_dpm_forced_level level, + uint32_t *sclk_mask, + uint32_t *mclk_mask, + uint32_t *soc_mask) +{ + struct vega20_dpm_table *dpm_table = (struct vega20_dpm_table *)smu->smu_dpm.dpm_context; + struct vega20_single_dpm_table *gfx_dpm_table; + struct vega20_single_dpm_table *mem_dpm_table; + struct vega20_single_dpm_table *soc_dpm_table; + + if (!smu->smu_dpm.dpm_context) + return -EINVAL; + + gfx_dpm_table = &dpm_table->gfx_table; + mem_dpm_table = &dpm_table->mem_table; + soc_dpm_table = &dpm_table->soc_table; + + *sclk_mask = 0; + *mclk_mask = 0; + *soc_mask = 0; + + if (gfx_dpm_table->count > VEGA20_UMD_PSTATE_GFXCLK_LEVEL && + mem_dpm_table->count > VEGA20_UMD_PSTATE_MCLK_LEVEL && + soc_dpm_table->count > VEGA20_UMD_PSTATE_SOCCLK_LEVEL) { + *sclk_mask = VEGA20_UMD_PSTATE_GFXCLK_LEVEL; + *mclk_mask = VEGA20_UMD_PSTATE_MCLK_LEVEL; + *soc_mask = VEGA20_UMD_PSTATE_SOCCLK_LEVEL; + } + + if (level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK) { + *sclk_mask = 0; + } else if (level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK) { + *mclk_mask = 0; + } else if (level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + *sclk_mask = gfx_dpm_table->count - 1; + *mclk_mask = mem_dpm_table->count - 1; + *soc_mask = soc_dpm_table->count - 1; + } + + return 0; +} + +static int +vega20_set_uclk_to_highest_dpm_level(struct smu_context *smu, + struct vega20_single_dpm_table *dpm_table) +{ + int ret = 0; + struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm); + if (!smu_dpm_ctx->dpm_context) + return -EINVAL; + + if (smu_feature_is_enabled(smu, FEATURE_DPM_UCLK_BIT)) { + if (dpm_table->count <= 0) { + pr_err("[%s] Dpm table has no entry!", __func__); + return -EINVAL; + } + + if (dpm_table->count > NUM_UCLK_DPM_LEVELS) { + pr_err("[%s] Dpm table has too many entries!", __func__); + return -EINVAL; + } + + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_SetHardMinByFreq, + (PPCLK_UCLK << 16) | dpm_table->dpm_state.hard_min_level); + if (ret) { + pr_err("[%s] Set hard min uclk failed!", __func__); + return ret; + } + } + + return ret; +} + +static int vega20_pre_display_config_changed(struct smu_context *smu) +{ + int ret = 0; + struct vega20_dpm_table *dpm_table = smu->smu_dpm.dpm_context; + + if (!smu->smu_dpm.dpm_context) + return -EINVAL; + + smu_send_smc_msg_with_param(smu, SMU_MSG_NumOfDisplays, 0); + ret = vega20_set_uclk_to_highest_dpm_level(smu, + &dpm_table->mem_table); + if (ret) + pr_err("Failed to set uclk to highest dpm level"); + return ret; +} + +static int vega20_display_config_changed(struct smu_context *smu) +{ + int ret = 0; + + if (!smu->funcs) + return -EINVAL; + + if (!smu->smu_dpm.dpm_context || + !smu->smu_table.tables || + !smu->smu_table.tables[TABLE_WATERMARKS].cpu_addr) + return -EINVAL; + + if ((smu->watermarks_bitmap & WATERMARKS_EXIST) && + !(smu->watermarks_bitmap & WATERMARKS_LOADED)) { + ret = smu->funcs->write_watermarks_table(smu); + if (ret) { + pr_err("Failed to update WMTABLE!"); + return ret; + } + smu->watermarks_bitmap |= WATERMARKS_LOADED; + } + + if ((smu->watermarks_bitmap & WATERMARKS_EXIST) && + smu_feature_is_supported(smu, FEATURE_DPM_DCEFCLK_BIT) && + smu_feature_is_supported(smu, FEATURE_DPM_SOCCLK_BIT)) { + smu_send_smc_msg_with_param(smu, + SMU_MSG_NumOfDisplays, + smu->display_config->num_display); + } + + return ret; +} + +static int vega20_apply_clocks_adjust_rules(struct smu_context *smu) +{ + struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm); + struct vega20_dpm_table *dpm_ctx = (struct vega20_dpm_table *)(smu_dpm_ctx->dpm_context); + struct vega20_single_dpm_table *dpm_table; + bool vblank_too_short = false; + bool disable_mclk_switching; + uint32_t i, latency; + + disable_mclk_switching = ((1 < smu->display_config->num_display) && + !smu->display_config->multi_monitor_in_sync) || vblank_too_short; + latency = smu->display_config->dce_tolerable_mclk_in_active_latency; + + /* gfxclk */ + dpm_table = &(dpm_ctx->gfx_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (VEGA20_UMD_PSTATE_GFXCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_GFXCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_GFXCLK_LEVEL].value; + } + + if (smu_dpm_ctx->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[0].value; + } + + if (smu_dpm_ctx->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + + /* memclk */ + dpm_table = &(dpm_ctx->mem_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (VEGA20_UMD_PSTATE_MCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_MCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_MCLK_LEVEL].value; + } + + if (smu_dpm_ctx->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[0].value; + } + + if (smu_dpm_ctx->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + + /* honour DAL's UCLK Hardmin */ + if (dpm_table->dpm_state.hard_min_level < (smu->display_config->min_mem_set_clock / 100)) + dpm_table->dpm_state.hard_min_level = smu->display_config->min_mem_set_clock / 100; + + /* Hardmin is dependent on displayconfig */ + if (disable_mclk_switching) { + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + for (i = 0; i < smu_dpm_ctx->mclk_latency_table->count - 1; i++) { + if (smu_dpm_ctx->mclk_latency_table->entries[i].latency <= latency) { + if (dpm_table->dpm_levels[i].value >= (smu->display_config->min_mem_set_clock / 100)) { + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[i].value; + break; + } + } + } + } + + if (smu->display_config->nb_pstate_switch_disable) + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + /* vclk */ + dpm_table = &(dpm_ctx->vclk_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (VEGA20_UMD_PSTATE_UVDCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_UVDCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_UVDCLK_LEVEL].value; + } + + if (smu_dpm_ctx->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + + /* dclk */ + dpm_table = &(dpm_ctx->dclk_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (VEGA20_UMD_PSTATE_UVDCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_UVDCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_UVDCLK_LEVEL].value; + } + + if (smu_dpm_ctx->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + + /* socclk */ + dpm_table = &(dpm_ctx->soc_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (VEGA20_UMD_PSTATE_SOCCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_SOCCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_SOCCLK_LEVEL].value; + } + + if (smu_dpm_ctx->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + + /* eclk */ + dpm_table = &(dpm_ctx->eclk_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (VEGA20_UMD_PSTATE_VCEMCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_VCEMCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA20_UMD_PSTATE_VCEMCLK_LEVEL].value; + } + + if (smu_dpm_ctx->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + return 0; +} + +static int +vega20_notify_smc_dispaly_config(struct smu_context *smu) +{ + struct vega20_dpm_table *dpm_table = smu->smu_dpm.dpm_context; + struct vega20_single_dpm_table *memtable = &dpm_table->mem_table; + struct smu_clocks min_clocks = {0}; + struct pp_display_clock_request clock_req; + int ret = 0; + + min_clocks.dcef_clock = smu->display_config->min_dcef_set_clk; + min_clocks.dcef_clock_in_sr = smu->display_config->min_dcef_deep_sleep_set_clk; + min_clocks.memory_clock = smu->display_config->min_mem_set_clock; + + if (smu_feature_is_supported(smu, FEATURE_DPM_DCEFCLK_BIT)) { + clock_req.clock_type = amd_pp_dcef_clock; + clock_req.clock_freq_in_khz = min_clocks.dcef_clock * 10; + if (!smu->funcs->display_clock_voltage_request(smu, &clock_req)) { + if (smu_feature_is_supported(smu, FEATURE_DS_DCEFCLK_BIT)) { + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_SetMinDeepSleepDcefclk, + min_clocks.dcef_clock_in_sr/100); + if (ret) { + pr_err("Attempt to set divider for DCEFCLK Failed!"); + return ret; + } + } + } else { + pr_info("Attempt to set Hard Min for DCEFCLK Failed!"); + } + } + + if (smu_feature_is_enabled(smu, FEATURE_DPM_UCLK_BIT)) { + memtable->dpm_state.hard_min_level = min_clocks.memory_clock/100; + ret = smu_send_smc_msg_with_param(smu, + SMU_MSG_SetHardMinByFreq, + (PPCLK_UCLK << 16) | memtable->dpm_state.hard_min_level); + if (ret) { + pr_err("[%s] Set hard min uclk failed!", __func__); + return ret; + } + } + + return 0; +} + +static uint32_t vega20_find_lowest_dpm_level(struct vega20_single_dpm_table *table) +{ + uint32_t i; + + for (i = 0; i < table->count; i++) { + if (table->dpm_levels[i].enabled) + break; + } + if (i >= table->count) { + i = 0; + table->dpm_levels[i].enabled = true; + } + + return i; +} + +static uint32_t vega20_find_highest_dpm_level(struct vega20_single_dpm_table *table) +{ + int i = 0; + + if (!table) { + pr_err("[%s] DPM Table does not exist!", __func__); + return 0; + } + if (table->count <= 0) { + pr_err("[%s] DPM Table has no entry!", __func__); + return 0; + } + if (table->count > MAX_REGULAR_DPM_NUMBER) { + pr_err("[%s] DPM Table has too many entries!", __func__); + return MAX_REGULAR_DPM_NUMBER - 1; + } + + for (i = table->count - 1; i >= 0; i--) { + if (table->dpm_levels[i].enabled) + break; + } + if (i < 0) { + i = 0; + table->dpm_levels[i].enabled = true; + } + + return i; +} + +static int vega20_force_dpm_limit_value(struct smu_context *smu, bool highest) +{ + uint32_t soft_level; + int ret = 0; + struct vega20_dpm_table *dpm_table = + (struct vega20_dpm_table *)smu->smu_dpm.dpm_context; + + if (highest) + soft_level = vega20_find_highest_dpm_level(&(dpm_table->gfx_table)); + else + soft_level = vega20_find_lowest_dpm_level(&(dpm_table->gfx_table)); + + dpm_table->gfx_table.dpm_state.soft_min_level = + dpm_table->gfx_table.dpm_state.soft_max_level = + dpm_table->gfx_table.dpm_levels[soft_level].value; + + if (highest) + soft_level = vega20_find_highest_dpm_level(&(dpm_table->mem_table)); + else + soft_level = vega20_find_lowest_dpm_level(&(dpm_table->mem_table)); + + dpm_table->mem_table.dpm_state.soft_min_level = + dpm_table->mem_table.dpm_state.soft_max_level = + dpm_table->mem_table.dpm_levels[soft_level].value; + + if (highest) + soft_level = vega20_find_highest_dpm_level(&(dpm_table->soc_table)); + else + soft_level = vega20_find_lowest_dpm_level(&(dpm_table->soc_table)); + + dpm_table->soc_table.dpm_state.soft_min_level = + dpm_table->soc_table.dpm_state.soft_max_level = + dpm_table->soc_table.dpm_levels[soft_level].value; + + ret = vega20_upload_dpm_level(smu, false, 0xFFFFFFFF); + if (ret) { + pr_err("Failed to upload boot level to %s!\n", + highest ? "highest" : "lowest"); + return ret; + } + + ret = vega20_upload_dpm_level(smu, true, 0xFFFFFFFF); + if (ret) { + pr_err("Failed to upload dpm max level to %s!\n!", + highest ? "highest" : "lowest"); + return ret; + } + + return ret; +} + +static int vega20_unforce_dpm_levels(struct smu_context *smu) +{ + uint32_t soft_min_level, soft_max_level; + int ret = 0; + struct vega20_dpm_table *dpm_table = + (struct vega20_dpm_table *)smu->smu_dpm.dpm_context; + + soft_min_level = vega20_find_lowest_dpm_level(&(dpm_table->gfx_table)); + soft_max_level = vega20_find_highest_dpm_level(&(dpm_table->gfx_table)); + dpm_table->gfx_table.dpm_state.soft_min_level = + dpm_table->gfx_table.dpm_levels[soft_min_level].value; + dpm_table->gfx_table.dpm_state.soft_max_level = + dpm_table->gfx_table.dpm_levels[soft_max_level].value; + + soft_min_level = vega20_find_lowest_dpm_level(&(dpm_table->mem_table)); + soft_max_level = vega20_find_highest_dpm_level(&(dpm_table->mem_table)); + dpm_table->mem_table.dpm_state.soft_min_level = + dpm_table->gfx_table.dpm_levels[soft_min_level].value; + dpm_table->mem_table.dpm_state.soft_max_level = + dpm_table->gfx_table.dpm_levels[soft_max_level].value; + + soft_min_level = vega20_find_lowest_dpm_level(&(dpm_table->soc_table)); + soft_max_level = vega20_find_highest_dpm_level(&(dpm_table->soc_table)); + dpm_table->soc_table.dpm_state.soft_min_level = + dpm_table->soc_table.dpm_levels[soft_min_level].value; + dpm_table->soc_table.dpm_state.soft_max_level = + dpm_table->soc_table.dpm_levels[soft_max_level].value; + + ret = smu_upload_dpm_level(smu, false, 0xFFFFFFFF); + if (ret) { + pr_err("Failed to upload DPM Bootup Levels!"); + return ret; + } + + ret = smu_upload_dpm_level(smu, true, 0xFFFFFFFF); + if (ret) { + pr_err("Failed to upload DPM Max Levels!"); + return ret; + } + + return ret; +} + +static enum amd_dpm_forced_level vega20_get_performance_level(struct smu_context *smu) +{ + struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm); + if (!smu_dpm_ctx->dpm_context) + return -EINVAL; + + if (smu_dpm_ctx->dpm_level != smu_dpm_ctx->saved_dpm_level) { + mutex_lock(&(smu->mutex)); + smu_dpm_ctx->saved_dpm_level = smu_dpm_ctx->dpm_level; + mutex_unlock(&(smu->mutex)); + } + return smu_dpm_ctx->dpm_level; +} + +static int +vega20_force_performance_level(struct smu_context *smu, enum amd_dpm_forced_level level) +{ + int ret = 0; + int i; + struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm); + + if (!smu_dpm_ctx->dpm_context) + return -EINVAL; + + for (i = 0; i < smu->adev->num_ip_blocks; i++) { + if (smu->adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SMC) + break; + } + + mutex_lock(&smu->mutex); + + smu->adev->ip_blocks[i].version->funcs->enable_umd_pstate(smu, &level); + ret = smu_handle_task(smu, level, + AMD_PP_TASK_READJUST_POWER_STATE); + + mutex_unlock(&smu->mutex); + + return ret; +} + +static int vega20_update_specified_od8_value(struct smu_context *smu, + uint32_t index, + uint32_t value) +{ + struct smu_table_context *table_context = &smu->smu_table; + OverDriveTable_t *od_table = + (OverDriveTable_t *)(table_context->overdrive_table); + struct vega20_od8_settings *od8_settings = + (struct vega20_od8_settings *)table_context->od8_settings; + + switch (index) { + case OD8_SETTING_GFXCLK_FMIN: + od_table->GfxclkFmin = (uint16_t)value; + break; + + case OD8_SETTING_GFXCLK_FMAX: + if (value < od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMAX].min_value || + value > od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMAX].max_value) + return -EINVAL; + od_table->GfxclkFmax = (uint16_t)value; + break; + + case OD8_SETTING_GFXCLK_FREQ1: + od_table->GfxclkFreq1 = (uint16_t)value; + break; + + case OD8_SETTING_GFXCLK_VOLTAGE1: + od_table->GfxclkVolt1 = (uint16_t)value; + break; + + case OD8_SETTING_GFXCLK_FREQ2: + od_table->GfxclkFreq2 = (uint16_t)value; + break; + + case OD8_SETTING_GFXCLK_VOLTAGE2: + od_table->GfxclkVolt2 = (uint16_t)value; + break; + + case OD8_SETTING_GFXCLK_FREQ3: + od_table->GfxclkFreq3 = (uint16_t)value; + break; + + case OD8_SETTING_GFXCLK_VOLTAGE3: + od_table->GfxclkVolt3 = (uint16_t)value; + break; + + case OD8_SETTING_UCLK_FMAX: + if (value < od8_settings->od8_settings_array[OD8_SETTING_UCLK_FMAX].min_value || + value > od8_settings->od8_settings_array[OD8_SETTING_UCLK_FMAX].max_value) + return -EINVAL; + od_table->UclkFmax = (uint16_t)value; + break; + + case OD8_SETTING_POWER_PERCENTAGE: + od_table->OverDrivePct = (int16_t)value; + break; + + case OD8_SETTING_FAN_ACOUSTIC_LIMIT: + od_table->FanMaximumRpm = (uint16_t)value; + break; + + case OD8_SETTING_FAN_MIN_SPEED: + od_table->FanMinimumPwm = (uint16_t)value; + break; + + case OD8_SETTING_FAN_TARGET_TEMP: + od_table->FanTargetTemperature = (uint16_t)value; + break; + + case OD8_SETTING_OPERATING_TEMP_MAX: + od_table->MaxOpTemp = (uint16_t)value; + break; + } + + return 0; +} + +static int vega20_set_od_percentage(struct smu_context *smu, + enum pp_clock_type type, + uint32_t value) +{ + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + struct vega20_dpm_table *dpm_table = NULL; + struct vega20_dpm_table *golden_table = NULL; + struct vega20_single_dpm_table *single_dpm_table; + struct vega20_single_dpm_table *golden_dpm_table; + uint32_t od_clk, index; + int ret = 0; + int feature_enabled; + PPCLK_e clk_id; + + mutex_lock(&(smu->mutex)); + + dpm_table = smu_dpm->dpm_context; + golden_table = smu_dpm->golden_dpm_context; + + switch (type) { + case OD_SCLK: + single_dpm_table = &(dpm_table->gfx_table); + golden_dpm_table = &(golden_table->gfx_table); + feature_enabled = smu_feature_is_enabled(smu, FEATURE_DPM_GFXCLK_BIT); + clk_id = PPCLK_GFXCLK; + index = OD8_SETTING_GFXCLK_FMAX; + break; + case OD_MCLK: + single_dpm_table = &(dpm_table->mem_table); + golden_dpm_table = &(golden_table->mem_table); + feature_enabled = smu_feature_is_enabled(smu, FEATURE_DPM_UCLK_BIT); + clk_id = PPCLK_UCLK; + index = OD8_SETTING_UCLK_FMAX; + break; + default: + ret = -EINVAL; + break; + } + + if (ret) + goto set_od_failed; + + od_clk = golden_dpm_table->dpm_levels[golden_dpm_table->count - 1].value * value; + od_clk /= 100; + od_clk += golden_dpm_table->dpm_levels[golden_dpm_table->count - 1].value; + + ret = smu_update_od8_settings(smu, index, od_clk); + if (ret) { + pr_err("[Setoverdrive] failed to set od clk!\n"); + goto set_od_failed; + } + + if (feature_enabled) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, + clk_id); + if (ret) { + pr_err("[Setoverdrive] failed to refresh dpm table!\n"); + goto set_od_failed; + } + } else { + single_dpm_table->count = 1; + single_dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.gfxclk / 100; + } + + ret = smu_handle_task(smu, smu_dpm->dpm_level, + AMD_PP_TASK_READJUST_POWER_STATE); + +set_od_failed: + mutex_unlock(&(smu->mutex)); + + return ret; +} + +static int vega20_odn_edit_dpm_table(struct smu_context *smu, + enum PP_OD_DPM_TABLE_COMMAND type, + long *input, uint32_t size) +{ + struct smu_table_context *table_context = &smu->smu_table; + OverDriveTable_t *od_table = + (OverDriveTable_t *)(table_context->overdrive_table); + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + struct vega20_dpm_table *dpm_table = NULL; + struct vega20_single_dpm_table *single_dpm_table; + struct vega20_od8_settings *od8_settings = + (struct vega20_od8_settings *)table_context->od8_settings; + struct pp_clock_levels_with_latency clocks; + int32_t input_index, input_clk, input_vol, i; + int od8_id; + int ret = 0; + + dpm_table = smu_dpm->dpm_context; + + if (!input) { + pr_warn("NULL user input for clock and voltage\n"); + return -EINVAL; + } + + switch (type) { + case PP_OD_EDIT_SCLK_VDDC_TABLE: + if (!(od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMIN].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMAX].feature_id)) { + pr_info("Sclk min/max frequency overdrive not supported\n"); + return -EOPNOTSUPP; + } + + for (i = 0; i < size; i += 2) { + if (i + 2 > size) { + pr_info("invalid number of input parameters %d\n", size); + return -EINVAL; + } + + input_index = input[i]; + input_clk = input[i + 1]; + + if (input_index != 0 && input_index != 1) { + pr_info("Invalid index %d\n", input_index); + pr_info("Support min/max sclk frequency settingonly which index by 0/1\n"); + return -EINVAL; + } + + if (input_clk < od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMIN].min_value || + input_clk > od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMAX].max_value) { + pr_info("clock freq %d is not within allowed range [%d - %d]\n", + input_clk, + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMIN].min_value, + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FMAX].max_value); + return -EINVAL; + } + + if (input_index == 0 && od_table->GfxclkFmin != input_clk) { + od_table->GfxclkFmin = input_clk; + table_context->od_gfxclk_update = true; + } else if (input_index == 1 && od_table->GfxclkFmax != input_clk) { + od_table->GfxclkFmax = input_clk; + table_context->od_gfxclk_update = true; + } + } + + break; + + case PP_OD_EDIT_MCLK_VDDC_TABLE: + if (!od8_settings->od8_settings_array[OD8_SETTING_UCLK_FMAX].feature_id) { + pr_info("Mclk max frequency overdrive not supported\n"); + return -EOPNOTSUPP; + } + + single_dpm_table = &(dpm_table->mem_table); + ret = vega20_get_clk_table(smu, &clocks, single_dpm_table); + if (ret) { + pr_err("Attempt to get memory clk levels Failed!"); + return ret; + } + + for (i = 0; i < size; i += 2) { + if (i + 2 > size) { + pr_info("invalid number of input parameters %d\n", + size); + return -EINVAL; + } + + input_index = input[i]; + input_clk = input[i + 1]; + + if (input_index != 1) { + pr_info("Invalid index %d\n", input_index); + pr_info("Support max Mclk frequency setting only which index by 1\n"); + return -EINVAL; + } + + if (input_clk < clocks.data[0].clocks_in_khz / 1000 || + input_clk > od8_settings->od8_settings_array[OD8_SETTING_UCLK_FMAX].max_value) { + pr_info("clock freq %d is not within allowed range [%d - %d]\n", + input_clk, + clocks.data[0].clocks_in_khz / 1000, + od8_settings->od8_settings_array[OD8_SETTING_UCLK_FMAX].max_value); + return -EINVAL; + } + + if (input_index == 1 && od_table->UclkFmax != input_clk) { + table_context->od_gfxclk_update = true; + od_table->UclkFmax = input_clk; + } + } + + break; + + case PP_OD_EDIT_VDDC_CURVE: + if (!(od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ1].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ2].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_FREQ3].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE1].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE2].feature_id && + od8_settings->od8_settings_array[OD8_SETTING_GFXCLK_VOLTAGE3].feature_id)) { + pr_info("Voltage curve calibrate not supported\n"); + return -EOPNOTSUPP; + } + + for (i = 0; i < size; i += 3) { + if (i + 3 > size) { + pr_info("invalid number of input parameters %d\n", + size); + return -EINVAL; + } + + input_index = input[i]; + input_clk = input[i + 1]; + input_vol = input[i + 2]; + + if (input_index > 2) { + pr_info("Setting for point %d is not supported\n", + input_index + 1); + pr_info("Three supported points index by 0, 1, 2\n"); + return -EINVAL; + } + + od8_id = OD8_SETTING_GFXCLK_FREQ1 + 2 * input_index; + if (input_clk < od8_settings->od8_settings_array[od8_id].min_value || + input_clk > od8_settings->od8_settings_array[od8_id].max_value) { + pr_info("clock freq %d is not within allowed range [%d - %d]\n", + input_clk, + od8_settings->od8_settings_array[od8_id].min_value, + od8_settings->od8_settings_array[od8_id].max_value); + return -EINVAL; + } + + od8_id = OD8_SETTING_GFXCLK_VOLTAGE1 + 2 * input_index; + if (input_vol < od8_settings->od8_settings_array[od8_id].min_value || + input_vol > od8_settings->od8_settings_array[od8_id].max_value) { + pr_info("clock voltage %d is not within allowed range [%d- %d]\n", + input_vol, + od8_settings->od8_settings_array[od8_id].min_value, + od8_settings->od8_settings_array[od8_id].max_value); + return -EINVAL; + } + + switch (input_index) { + case 0: + od_table->GfxclkFreq1 = input_clk; + od_table->GfxclkVolt1 = input_vol * VOLTAGE_SCALE; + break; + case 1: + od_table->GfxclkFreq2 = input_clk; + od_table->GfxclkVolt2 = input_vol * VOLTAGE_SCALE; + break; + case 2: + od_table->GfxclkFreq3 = input_clk; + od_table->GfxclkVolt3 = input_vol * VOLTAGE_SCALE; + break; + } + } + + break; + + case PP_OD_RESTORE_DEFAULT_TABLE: + ret = smu_update_table(smu, TABLE_OVERDRIVE, table_context->overdrive_table, false); + if (ret) { + pr_err("Failed to export over drive table!\n"); + return ret; + } + + break; + + case PP_OD_COMMIT_DPM_TABLE: + ret = smu_update_table(smu, TABLE_OVERDRIVE, table_context->overdrive_table, true); + if (ret) { + pr_err("Failed to import over drive table!\n"); + return ret; + } + + /* retrieve updated gfxclk table */ + if (table_context->od_gfxclk_update) { + table_context->od_gfxclk_update = false; + single_dpm_table = &(dpm_table->gfx_table); + + if (smu_feature_is_enabled(smu, FEATURE_DPM_GFXCLK_BIT)) { + ret = vega20_set_single_dpm_table(smu, single_dpm_table, + PPCLK_GFXCLK); + if (ret) { + pr_err("[Setoverdrive] failed to refresh dpm table!\n"); + return ret; + } + } else { + single_dpm_table->count = 1; + single_dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.gfxclk / 100; + } + } + + break; + + default: + return -EINVAL; + } + + if (type == PP_OD_COMMIT_DPM_TABLE) { + mutex_lock(&(smu->mutex)); + ret = smu_handle_task(smu, smu_dpm->dpm_level, + AMD_PP_TASK_READJUST_POWER_STATE); + mutex_unlock(&(smu->mutex)); + } + + return ret; +} + +static const struct pptable_funcs vega20_ppt_funcs = { + .alloc_dpm_context = vega20_allocate_dpm_context, + .store_powerplay_table = vega20_store_powerplay_table, + .check_powerplay_table = vega20_check_powerplay_table, + .append_powerplay_table = vega20_append_powerplay_table, + .get_smu_msg_index = vega20_get_smu_msg_index, + .run_afll_btc = vega20_run_btc_afll, + .get_unallowed_feature_mask = vega20_get_unallowed_feature_mask, + .get_current_power_state = vega20_get_current_power_state, + .set_default_dpm_table = vega20_set_default_dpm_table, + .set_power_state = NULL, + .populate_umd_state_clk = vega20_populate_umd_state_clk, + .print_clk_levels = vega20_print_clk_levels, + .force_clk_levels = vega20_force_clk_levels, + .get_clock_by_type_with_latency = vega20_get_clock_by_type_with_latency, + .set_default_od8_settings = vega20_set_default_od8_setttings, + .get_od_percentage = vega20_get_od_percentage, + .get_performance_level = vega20_get_performance_level, + .force_performance_level = vega20_force_performance_level, + .update_specified_od8_value = vega20_update_specified_od8_value, + .set_od_percentage = vega20_set_od_percentage, + .od_edit_dpm_table = vega20_odn_edit_dpm_table, + .pre_display_config_changed = vega20_pre_display_config_changed, + .display_config_changed = vega20_display_config_changed, + .apply_clocks_adjust_rules = vega20_apply_clocks_adjust_rules, + .notify_smc_dispaly_config = vega20_notify_smc_dispaly_config, + .force_dpm_limit_value = vega20_force_dpm_limit_value, + .unforce_dpm_levels = vega20_unforce_dpm_levels, + .upload_dpm_level = vega20_upload_dpm_level, + .get_profiling_clk_mask = vega20_get_profiling_clk_mask, +}; + +void vega20_set_ppt_funcs(struct smu_context *smu) +{ + smu->ppt_funcs = &vega20_ppt_funcs; + smu->smc_if_version = SMU11_DRIVER_IF_VERSION; +} diff --git a/drivers/gpu/drm/amd/powerplay/vega20_ppt.h b/drivers/gpu/drm/amd/powerplay/vega20_ppt.h new file mode 100644 index 000000000000..5a0d2af63173 --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/vega20_ppt.h @@ -0,0 +1,129 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __VEGA20_PPT_H__ +#define __VEGA20_PPT_H__ + +#define VEGA20_UMD_PSTATE_GFXCLK_LEVEL 0x3 +#define VEGA20_UMD_PSTATE_SOCCLK_LEVEL 0x3 +#define VEGA20_UMD_PSTATE_MCLK_LEVEL 0x2 +#define VEGA20_UMD_PSTATE_UVDCLK_LEVEL 0x3 +#define VEGA20_UMD_PSTATE_VCEMCLK_LEVEL 0x3 + +#define MAX_REGULAR_DPM_NUMBER 16 +#define MAX_PCIE_CONF 2 + +#define VOLTAGE_SCALE 4 +#define AVFS_CURVE 0 +#define OD8_HOTCURVE_TEMPERATURE 85 + +struct vega20_dpm_level { + bool enabled; + uint32_t value; + uint32_t param1; +}; + +struct vega20_dpm_state { + uint32_t soft_min_level; + uint32_t soft_max_level; + uint32_t hard_min_level; + uint32_t hard_max_level; +}; + +struct vega20_single_dpm_table { + uint32_t count; + struct vega20_dpm_state dpm_state; + struct vega20_dpm_level dpm_levels[MAX_REGULAR_DPM_NUMBER]; +}; + +struct vega20_pcie_table { + uint16_t count; + uint8_t pcie_gen[MAX_PCIE_CONF]; + uint8_t pcie_lane[MAX_PCIE_CONF]; + uint32_t lclk[MAX_PCIE_CONF]; +}; + +struct vega20_dpm_table { + struct vega20_single_dpm_table soc_table; + struct vega20_single_dpm_table gfx_table; + struct vega20_single_dpm_table mem_table; + struct vega20_single_dpm_table eclk_table; + struct vega20_single_dpm_table vclk_table; + struct vega20_single_dpm_table dclk_table; + struct vega20_single_dpm_table dcef_table; + struct vega20_single_dpm_table pixel_table; + struct vega20_single_dpm_table display_table; + struct vega20_single_dpm_table phy_table; + struct vega20_single_dpm_table fclk_table; + struct vega20_pcie_table pcie_table; +}; + +enum OD8_FEATURE_ID +{ + OD8_GFXCLK_LIMITS = 1 << 0, + OD8_GFXCLK_CURVE = 1 << 1, + OD8_UCLK_MAX = 1 << 2, + OD8_POWER_LIMIT = 1 << 3, + OD8_ACOUSTIC_LIMIT_SCLK = 1 << 4, //FanMaximumRpm + OD8_FAN_SPEED_MIN = 1 << 5, //FanMinimumPwm + OD8_TEMPERATURE_FAN = 1 << 6, //FanTargetTemperature + OD8_TEMPERATURE_SYSTEM = 1 << 7, //MaxOpTemp + OD8_MEMORY_TIMING_TUNE = 1 << 8, + OD8_FAN_ZERO_RPM_CONTROL = 1 << 9 +}; + +enum OD8_SETTING_ID +{ + OD8_SETTING_GFXCLK_FMIN = 0, + OD8_SETTING_GFXCLK_FMAX, + OD8_SETTING_GFXCLK_FREQ1, + OD8_SETTING_GFXCLK_VOLTAGE1, + OD8_SETTING_GFXCLK_FREQ2, + OD8_SETTING_GFXCLK_VOLTAGE2, + OD8_SETTING_GFXCLK_FREQ3, + OD8_SETTING_GFXCLK_VOLTAGE3, + OD8_SETTING_UCLK_FMAX, + OD8_SETTING_POWER_PERCENTAGE, + OD8_SETTING_FAN_ACOUSTIC_LIMIT, + OD8_SETTING_FAN_MIN_SPEED, + OD8_SETTING_FAN_TARGET_TEMP, + OD8_SETTING_OPERATING_TEMP_MAX, + OD8_SETTING_AC_TIMING, + OD8_SETTING_FAN_ZERO_RPM_CONTROL, + OD8_SETTING_COUNT +}; + +struct vega20_od8_single_setting { + uint32_t feature_id; + int32_t min_value; + int32_t max_value; + int32_t current_value; + int32_t default_value; +}; + +struct vega20_od8_settings { + struct vega20_od8_single_setting od8_settings_array[OD8_SETTING_COUNT]; +}; + +extern void vega20_set_ppt_funcs(struct smu_context *smu); + +#endif diff --git a/drivers/gpu/drm/arm/display/include/malidp_utils.h b/drivers/gpu/drm/arm/display/include/malidp_utils.h index 63cc47cefcf8..8cfd91196e15 100644 --- a/drivers/gpu/drm/arm/display/include/malidp_utils.h +++ b/drivers/gpu/drm/arm/display/include/malidp_utils.h @@ -7,10 +7,41 @@ #ifndef _MALIDP_UTILS_ #define _MALIDP_UTILS_ +#include + #define has_bit(nr, mask) (BIT(nr) & (mask)) #define has_bits(bits, mask) (((bits) & (mask)) == (bits)) #define dp_for_each_set_bit(bit, mask) \ for_each_set_bit((bit), ((unsigned long *)&(mask)), sizeof(mask) * 8) +#define dp_wait_cond(__cond, __tries, __min_range, __max_range) \ +({ \ + int num_tries = __tries; \ + while (!__cond && (num_tries > 0)) { \ + usleep_range(__min_range, __max_range); \ + if (__cond) \ + break; \ + num_tries--; \ + } \ + num_tries; \ +}) + +/* the restriction of range is [start, end] */ +struct malidp_range { + u32 start; + u32 end; +}; + +static inline void set_range(struct malidp_range *rg, u32 start, u32 end) +{ + rg->start = start; + rg->end = end; +} + +static inline bool in_range(struct malidp_range *rg, u32 v) +{ + return (v >= rg->start) && (v <= rg->end); +} + #endif /* _MALIDP_UTILS_ */ diff --git a/drivers/gpu/drm/arm/display/komeda/Makefile b/drivers/gpu/drm/arm/display/komeda/Makefile index 1b875e5dc0f6..d593125236ae 100644 --- a/drivers/gpu/drm/arm/display/komeda/Makefile +++ b/drivers/gpu/drm/arm/display/komeda/Makefile @@ -16,6 +16,7 @@ komeda-y := \ komeda_private_obj.o komeda-y += \ - d71/d71_dev.o + d71/d71_dev.o \ + d71/d71_component.o obj-$(CONFIG_DRM_KOMEDA) += komeda.o diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c b/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c new file mode 100644 index 000000000000..c56cfc2de147 --- /dev/null +++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c @@ -0,0 +1,684 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * (C) COPYRIGHT 2018 ARM Limited. All rights reserved. + * Author: James.Qian.Wang + * + */ + +#include +#include "d71_dev.h" +#include "komeda_kms.h" +#include "malidp_io.h" +#include "komeda_framebuffer.h" + +static void get_resources_id(u32 hw_id, u32 *pipe_id, u32 *comp_id) +{ + u32 id = BLOCK_INFO_BLK_ID(hw_id); + u32 pipe = id; + + switch (BLOCK_INFO_BLK_TYPE(hw_id)) { + case D71_BLK_TYPE_LPU_WB_LAYER: + id = KOMEDA_COMPONENT_WB_LAYER; + break; + case D71_BLK_TYPE_CU_SPLITTER: + id = KOMEDA_COMPONENT_SPLITTER; + break; + case D71_BLK_TYPE_CU_SCALER: + pipe = id / D71_PIPELINE_MAX_SCALERS; + id %= D71_PIPELINE_MAX_SCALERS; + id += KOMEDA_COMPONENT_SCALER0; + break; + case D71_BLK_TYPE_CU: + id += KOMEDA_COMPONENT_COMPIZ0; + break; + case D71_BLK_TYPE_LPU_LAYER: + pipe = id / D71_PIPELINE_MAX_LAYERS; + id %= D71_PIPELINE_MAX_LAYERS; + id += KOMEDA_COMPONENT_LAYER0; + break; + case D71_BLK_TYPE_DOU_IPS: + id += KOMEDA_COMPONENT_IPS0; + break; + case D71_BLK_TYPE_CU_MERGER: + id = KOMEDA_COMPONENT_MERGER; + break; + case D71_BLK_TYPE_DOU: + id = KOMEDA_COMPONENT_TIMING_CTRLR; + break; + default: + id = 0xFFFFFFFF; + } + + if (comp_id) + *comp_id = id; + + if (pipe_id) + *pipe_id = pipe; +} + +static u32 get_valid_inputs(struct block_header *blk) +{ + u32 valid_inputs = 0, comp_id; + int i; + + for (i = 0; i < PIPELINE_INFO_N_VALID_INPUTS(blk->pipeline_info); i++) { + get_resources_id(blk->input_ids[i], NULL, &comp_id); + if (comp_id == 0xFFFFFFFF) + continue; + valid_inputs |= BIT(comp_id); + } + + return valid_inputs; +} + +static void get_values_from_reg(void __iomem *reg, u32 offset, + u32 count, u32 *val) +{ + u32 i, addr; + + for (i = 0; i < count; i++) { + addr = offset + (i << 2); + /* 0xA4 is WO register */ + if (addr != 0xA4) + val[i] = malidp_read32(reg, addr); + else + val[i] = 0xDEADDEAD; + } +} + +static void dump_block_header(struct seq_file *sf, void __iomem *reg) +{ + struct block_header hdr; + u32 i, n_input, n_output; + + d71_read_block_header(reg, &hdr); + seq_printf(sf, "BLOCK_INFO:\t\t0x%X\n", hdr.block_info); + seq_printf(sf, "PIPELINE_INFO:\t\t0x%X\n", hdr.pipeline_info); + + n_output = PIPELINE_INFO_N_OUTPUTS(hdr.pipeline_info); + n_input = PIPELINE_INFO_N_VALID_INPUTS(hdr.pipeline_info); + + for (i = 0; i < n_input; i++) + seq_printf(sf, "VALID_INPUT_ID%u:\t0x%X\n", + i, hdr.input_ids[i]); + + for (i = 0; i < n_output; i++) + seq_printf(sf, "OUTPUT_ID%u:\t\t0x%X\n", + i, hdr.output_ids[i]); +} + +static u32 to_rot_ctrl(u32 rot) +{ + u32 lr_ctrl = 0; + + switch (rot & DRM_MODE_ROTATE_MASK) { + case DRM_MODE_ROTATE_0: + lr_ctrl |= L_ROT(L_ROT_R0); + break; + case DRM_MODE_ROTATE_90: + lr_ctrl |= L_ROT(L_ROT_R90); + break; + case DRM_MODE_ROTATE_180: + lr_ctrl |= L_ROT(L_ROT_R180); + break; + case DRM_MODE_ROTATE_270: + lr_ctrl |= L_ROT(L_ROT_R270); + break; + } + + if (rot & DRM_MODE_REFLECT_X) + lr_ctrl |= L_HFLIP; + if (rot & DRM_MODE_REFLECT_Y) + lr_ctrl |= L_VFLIP; + + return lr_ctrl; +} + +static inline u32 to_d71_input_id(struct komeda_component_output *output) +{ + struct komeda_component *comp = output->component; + + return comp ? (comp->hw_id + output->output_port) : 0; +} + +static void d71_layer_disable(struct komeda_component *c) +{ + malidp_write32_mask(c->reg, BLK_CONTROL, L_EN, 0); +} + +static void d71_layer_update(struct komeda_component *c, + struct komeda_component_state *state) +{ + struct komeda_layer_state *st = to_layer_st(state); + struct drm_plane_state *plane_st = state->plane->state; + struct drm_framebuffer *fb = plane_st->fb; + struct komeda_fb *kfb = to_kfb(fb); + u32 __iomem *reg = c->reg; + u32 ctrl_mask = L_EN | L_ROT(L_ROT_R270) | L_HFLIP | L_VFLIP | L_TBU_EN; + u32 ctrl = L_EN | to_rot_ctrl(st->rot); + int i; + + for (i = 0; i < fb->format->num_planes; i++) { + malidp_write32(reg, + BLK_P0_PTR_LOW + i * LAYER_PER_PLANE_REGS * 4, + lower_32_bits(st->addr[i])); + malidp_write32(reg, + BLK_P0_PTR_HIGH + i * LAYER_PER_PLANE_REGS * 4, + upper_32_bits(st->addr[i])); + if (i >= 2) + break; + + malidp_write32(reg, + BLK_P0_STRIDE + i * LAYER_PER_PLANE_REGS * 4, + fb->pitches[i] & 0xFFFF); + } + + malidp_write32(reg, LAYER_FMT, kfb->format_caps->hw_id); + malidp_write32(reg, BLK_IN_SIZE, HV_SIZE(st->hsize, st->vsize)); + + malidp_write32_mask(reg, BLK_CONTROL, ctrl_mask, ctrl); +} + +static void d71_layer_dump(struct komeda_component *c, struct seq_file *sf) +{ + u32 v[15], i; + bool rich, rgb2rgb; + char *prefix; + + get_values_from_reg(c->reg, LAYER_INFO, 1, &v[14]); + if (v[14] & 0x1) { + rich = true; + prefix = "LR_"; + } else { + rich = false; + prefix = "LS_"; + } + + rgb2rgb = !!(v[14] & L_INFO_CM); + + dump_block_header(sf, c->reg); + + seq_printf(sf, "%sLAYER_INFO:\t\t0x%X\n", prefix, v[14]); + + get_values_from_reg(c->reg, 0xD0, 1, v); + seq_printf(sf, "%sCONTROL:\t\t0x%X\n", prefix, v[0]); + if (rich) { + get_values_from_reg(c->reg, 0xD4, 1, v); + seq_printf(sf, "LR_RICH_CONTROL:\t0x%X\n", v[0]); + } + get_values_from_reg(c->reg, 0xD8, 4, v); + seq_printf(sf, "%sFORMAT:\t\t0x%X\n", prefix, v[0]); + seq_printf(sf, "%sIT_COEFFTAB:\t\t0x%X\n", prefix, v[1]); + seq_printf(sf, "%sIN_SIZE:\t\t0x%X\n", prefix, v[2]); + seq_printf(sf, "%sPALPHA:\t\t0x%X\n", prefix, v[3]); + + get_values_from_reg(c->reg, 0x100, 3, v); + seq_printf(sf, "%sP0_PTR_LOW:\t\t0x%X\n", prefix, v[0]); + seq_printf(sf, "%sP0_PTR_HIGH:\t\t0x%X\n", prefix, v[1]); + seq_printf(sf, "%sP0_STRIDE:\t\t0x%X\n", prefix, v[2]); + + get_values_from_reg(c->reg, 0x110, 2, v); + seq_printf(sf, "%sP1_PTR_LOW:\t\t0x%X\n", prefix, v[0]); + seq_printf(sf, "%sP1_PTR_HIGH:\t\t0x%X\n", prefix, v[1]); + if (rich) { + get_values_from_reg(c->reg, 0x118, 1, v); + seq_printf(sf, "LR_P1_STRIDE:\t\t0x%X\n", v[0]); + + get_values_from_reg(c->reg, 0x120, 2, v); + seq_printf(sf, "LR_P2_PTR_LOW:\t\t0x%X\n", v[0]); + seq_printf(sf, "LR_P2_PTR_HIGH:\t\t0x%X\n", v[1]); + + get_values_from_reg(c->reg, 0x130, 12, v); + for (i = 0; i < 12; i++) + seq_printf(sf, "LR_YUV_RGB_COEFF%u:\t0x%X\n", i, v[i]); + } + + if (rgb2rgb) { + get_values_from_reg(c->reg, LAYER_RGB_RGB_COEFF0, 12, v); + for (i = 0; i < 12; i++) + seq_printf(sf, "LS_RGB_RGB_COEFF%u:\t0x%X\n", i, v[i]); + } + + get_values_from_reg(c->reg, 0x160, 3, v); + seq_printf(sf, "%sAD_CONTROL:\t\t0x%X\n", prefix, v[0]); + seq_printf(sf, "%sAD_H_CROP:\t\t0x%X\n", prefix, v[1]); + seq_printf(sf, "%sAD_V_CROP:\t\t0x%X\n", prefix, v[2]); +} + +static struct komeda_component_funcs d71_layer_funcs = { + .update = d71_layer_update, + .disable = d71_layer_disable, + .dump_register = d71_layer_dump, +}; + +static int d71_layer_init(struct d71_dev *d71, + struct block_header *blk, u32 __iomem *reg) +{ + struct komeda_component *c; + struct komeda_layer *layer; + u32 pipe_id, layer_id, layer_info; + + get_resources_id(blk->block_info, &pipe_id, &layer_id); + c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*layer), + layer_id, + BLOCK_INFO_INPUT_ID(blk->block_info), + &d71_layer_funcs, 0, + get_valid_inputs(blk), + 1, reg, "LPU%d_LAYER%d", pipe_id, layer_id); + if (IS_ERR(c)) { + DRM_ERROR("Failed to add layer component\n"); + return PTR_ERR(c); + } + + layer = to_layer(c); + layer_info = malidp_read32(reg, LAYER_INFO); + + if (layer_info & L_INFO_RF) + layer->layer_type = KOMEDA_FMT_RICH_LAYER; + else + layer->layer_type = KOMEDA_FMT_SIMPLE_LAYER; + + set_range(&layer->hsize_in, 4, d71->max_line_size); + set_range(&layer->vsize_in, 4, d71->max_vsize); + + malidp_write32(reg, LAYER_PALPHA, D71_PALPHA_DEF_MAP); + + layer->supported_rots = DRM_MODE_ROTATE_MASK | DRM_MODE_REFLECT_MASK; + + return 0; +} + +static int d71_wb_layer_init(struct d71_dev *d71, + struct block_header *blk, u32 __iomem *reg) +{ + DRM_DEBUG("Detect D71_Wb_Layer.\n"); + + return 0; +} + +static void d71_component_disable(struct komeda_component *c) +{ + u32 __iomem *reg = c->reg; + u32 i; + + malidp_write32(reg, BLK_CONTROL, 0); + + for (i = 0; i < c->max_active_inputs; i++) + malidp_write32(reg, BLK_INPUT_ID0 + (i << 2), 0); +} + +static void compiz_enable_input(u32 __iomem *id_reg, + u32 __iomem *cfg_reg, + u32 input_hw_id, + struct komeda_compiz_input_cfg *cin) +{ + u32 ctrl = CU_INPUT_CTRL_EN; + u8 blend = cin->pixel_blend_mode; + + if (blend == DRM_MODE_BLEND_PIXEL_NONE) + ctrl |= CU_INPUT_CTRL_PAD; + else if (blend == DRM_MODE_BLEND_PREMULTI) + ctrl |= CU_INPUT_CTRL_PMUL; + + ctrl |= CU_INPUT_CTRL_ALPHA(cin->layer_alpha); + + malidp_write32(id_reg, BLK_INPUT_ID0, input_hw_id); + + malidp_write32(cfg_reg, CU_INPUT0_SIZE, + HV_SIZE(cin->hsize, cin->vsize)); + malidp_write32(cfg_reg, CU_INPUT0_OFFSET, + HV_OFFSET(cin->hoffset, cin->voffset)); + malidp_write32(cfg_reg, CU_INPUT0_CONTROL, ctrl); +} + +static void d71_compiz_update(struct komeda_component *c, + struct komeda_component_state *state) +{ + struct komeda_compiz_state *st = to_compiz_st(state); + u32 __iomem *reg = c->reg; + u32 __iomem *id_reg, *cfg_reg; + u32 index, input_hw_id; + + for_each_changed_input(state, index) { + id_reg = reg + index; + cfg_reg = reg + index * CU_PER_INPUT_REGS; + input_hw_id = to_d71_input_id(&state->inputs[index]); + if (state->active_inputs & BIT(index)) { + compiz_enable_input(id_reg, cfg_reg, + input_hw_id, &st->cins[index]); + } else { + malidp_write32(id_reg, BLK_INPUT_ID0, 0); + malidp_write32(cfg_reg, CU_INPUT0_CONTROL, 0); + } + } + + malidp_write32(reg, BLK_SIZE, HV_SIZE(st->hsize, st->vsize)); +} + +static void d71_compiz_dump(struct komeda_component *c, struct seq_file *sf) +{ + u32 v[8], i; + + dump_block_header(sf, c->reg); + + get_values_from_reg(c->reg, 0x80, 5, v); + for (i = 0; i < 5; i++) + seq_printf(sf, "CU_INPUT_ID%u:\t\t0x%X\n", i, v[i]); + + get_values_from_reg(c->reg, 0xA0, 5, v); + seq_printf(sf, "CU_IRQ_RAW_STATUS:\t0x%X\n", v[0]); + seq_printf(sf, "CU_IRQ_CLEAR:\t\t0x%X\n", v[1]); + seq_printf(sf, "CU_IRQ_MASK:\t\t0x%X\n", v[2]); + seq_printf(sf, "CU_IRQ_STATUS:\t\t0x%X\n", v[3]); + seq_printf(sf, "CU_STATUS:\t\t0x%X\n", v[4]); + + get_values_from_reg(c->reg, 0xD0, 2, v); + seq_printf(sf, "CU_CONTROL:\t\t0x%X\n", v[0]); + seq_printf(sf, "CU_SIZE:\t\t0x%X\n", v[1]); + + get_values_from_reg(c->reg, 0xDC, 1, v); + seq_printf(sf, "CU_BG_COLOR:\t\t0x%X\n", v[0]); + + for (i = 0, v[4] = 0xE0; i < 5; i++, v[4] += 0x10) { + get_values_from_reg(c->reg, v[4], 3, v); + seq_printf(sf, "CU_INPUT%u_SIZE:\t\t0x%X\n", i, v[0]); + seq_printf(sf, "CU_INPUT%u_OFFSET:\t0x%X\n", i, v[1]); + seq_printf(sf, "CU_INPUT%u_CONTROL:\t0x%X\n", i, v[2]); + } + + get_values_from_reg(c->reg, 0x130, 2, v); + seq_printf(sf, "CU_USER_LOW:\t\t0x%X\n", v[0]); + seq_printf(sf, "CU_USER_HIGH:\t\t0x%X\n", v[1]); +} + +struct komeda_component_funcs d71_compiz_funcs = { + .update = d71_compiz_update, + .disable = d71_component_disable, + .dump_register = d71_compiz_dump, +}; + +static int d71_compiz_init(struct d71_dev *d71, + struct block_header *blk, u32 __iomem *reg) +{ + struct komeda_component *c; + struct komeda_compiz *compiz; + u32 pipe_id, comp_id; + + get_resources_id(blk->block_info, &pipe_id, &comp_id); + + c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*compiz), + comp_id, + BLOCK_INFO_INPUT_ID(blk->block_info), + &d71_compiz_funcs, + CU_NUM_INPUT_IDS, get_valid_inputs(blk), + CU_NUM_OUTPUT_IDS, reg, + "CU%d", pipe_id); + if (IS_ERR(c)) + return PTR_ERR(c); + + compiz = to_compiz(c); + + set_range(&compiz->hsize, D71_MIN_LINE_SIZE, d71->max_line_size); + set_range(&compiz->vsize, D71_MIN_VERTICAL_SIZE, d71->max_vsize); + + return 0; +} + +static void d71_improc_update(struct komeda_component *c, + struct komeda_component_state *state) +{ + struct komeda_improc_state *st = to_improc_st(state); + u32 __iomem *reg = c->reg; + u32 index, input_hw_id; + + for_each_changed_input(state, index) { + input_hw_id = state->active_inputs & BIT(index) ? + to_d71_input_id(&state->inputs[index]) : 0; + malidp_write32(reg, BLK_INPUT_ID0 + index * 4, input_hw_id); + } + + malidp_write32(reg, BLK_SIZE, HV_SIZE(st->hsize, st->vsize)); +} + +static void d71_improc_dump(struct komeda_component *c, struct seq_file *sf) +{ + u32 v[12], i; + + dump_block_header(sf, c->reg); + + get_values_from_reg(c->reg, 0x80, 2, v); + seq_printf(sf, "IPS_INPUT_ID0:\t\t0x%X\n", v[0]); + seq_printf(sf, "IPS_INPUT_ID1:\t\t0x%X\n", v[1]); + + get_values_from_reg(c->reg, 0xC0, 1, v); + seq_printf(sf, "IPS_INFO:\t\t0x%X\n", v[0]); + + get_values_from_reg(c->reg, 0xD0, 3, v); + seq_printf(sf, "IPS_CONTROL:\t\t0x%X\n", v[0]); + seq_printf(sf, "IPS_SIZE:\t\t0x%X\n", v[1]); + seq_printf(sf, "IPS_DEPTH:\t\t0x%X\n", v[2]); + + get_values_from_reg(c->reg, 0x130, 12, v); + for (i = 0; i < 12; i++) + seq_printf(sf, "IPS_RGB_RGB_COEFF%u:\t0x%X\n", i, v[i]); + + get_values_from_reg(c->reg, 0x170, 12, v); + for (i = 0; i < 12; i++) + seq_printf(sf, "IPS_RGB_YUV_COEFF%u:\t0x%X\n", i, v[i]); +} + +struct komeda_component_funcs d71_improc_funcs = { + .update = d71_improc_update, + .disable = d71_component_disable, + .dump_register = d71_improc_dump, +}; + +static int d71_improc_init(struct d71_dev *d71, + struct block_header *blk, u32 __iomem *reg) +{ + struct komeda_component *c; + struct komeda_improc *improc; + u32 pipe_id, comp_id, value; + + get_resources_id(blk->block_info, &pipe_id, &comp_id); + + c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*improc), + comp_id, + BLOCK_INFO_INPUT_ID(blk->block_info), + &d71_improc_funcs, IPS_NUM_INPUT_IDS, + get_valid_inputs(blk), + IPS_NUM_OUTPUT_IDS, reg, "DOU%d_IPS", pipe_id); + if (IS_ERR(c)) { + DRM_ERROR("Failed to add improc component\n"); + return PTR_ERR(c); + } + + improc = to_improc(c); + improc->supported_color_depths = BIT(8) | BIT(10); + improc->supported_color_formats = DRM_COLOR_FORMAT_RGB444 | + DRM_COLOR_FORMAT_YCRCB444 | + DRM_COLOR_FORMAT_YCRCB422; + value = malidp_read32(reg, BLK_INFO); + if (value & IPS_INFO_CHD420) + improc->supported_color_formats |= DRM_COLOR_FORMAT_YCRCB420; + + improc->supports_csc = true; + improc->supports_gamma = true; + + return 0; +} + +static void d71_timing_ctrlr_disable(struct komeda_component *c) +{ + malidp_write32_mask(c->reg, BLK_CONTROL, BS_CTRL_EN, 0); +} + +static void d71_timing_ctrlr_update(struct komeda_component *c, + struct komeda_component_state *state) +{ + struct drm_crtc_state *crtc_st = state->crtc->state; + u32 __iomem *reg = c->reg; + struct videomode vm; + u32 value; + + drm_display_mode_to_videomode(&crtc_st->adjusted_mode, &vm); + + malidp_write32(reg, BS_ACTIVESIZE, HV_SIZE(vm.hactive, vm.vactive)); + malidp_write32(reg, BS_HINTERVALS, BS_H_INTVALS(vm.hfront_porch, + vm.hback_porch)); + malidp_write32(reg, BS_VINTERVALS, BS_V_INTVALS(vm.vfront_porch, + vm.vback_porch)); + + value = BS_SYNC_VSW(vm.vsync_len) | BS_SYNC_HSW(vm.hsync_len); + value |= vm.flags & DISPLAY_FLAGS_VSYNC_HIGH ? BS_SYNC_VSP : 0; + value |= vm.flags & DISPLAY_FLAGS_HSYNC_HIGH ? BS_SYNC_HSP : 0; + malidp_write32(reg, BS_SYNC, value); + + malidp_write32(reg, BS_PROG_LINE, D71_DEFAULT_PREPRETCH_LINE - 1); + malidp_write32(reg, BS_PREFETCH_LINE, D71_DEFAULT_PREPRETCH_LINE); + + /* configure bs control register */ + value = BS_CTRL_EN | BS_CTRL_VM; + + malidp_write32(reg, BLK_CONTROL, value); +} + +void d71_timing_ctrlr_dump(struct komeda_component *c, struct seq_file *sf) +{ + u32 v[8], i; + + dump_block_header(sf, c->reg); + + get_values_from_reg(c->reg, 0xC0, 1, v); + seq_printf(sf, "BS_INFO:\t\t0x%X\n", v[0]); + + get_values_from_reg(c->reg, 0xD0, 8, v); + seq_printf(sf, "BS_CONTROL:\t\t0x%X\n", v[0]); + seq_printf(sf, "BS_PROG_LINE:\t\t0x%X\n", v[1]); + seq_printf(sf, "BS_PREFETCH_LINE:\t0x%X\n", v[2]); + seq_printf(sf, "BS_BG_COLOR:\t\t0x%X\n", v[3]); + seq_printf(sf, "BS_ACTIVESIZE:\t\t0x%X\n", v[4]); + seq_printf(sf, "BS_HINTERVALS:\t\t0x%X\n", v[5]); + seq_printf(sf, "BS_VINTERVALS:\t\t0x%X\n", v[6]); + seq_printf(sf, "BS_SYNC:\t\t0x%X\n", v[7]); + + get_values_from_reg(c->reg, 0x100, 3, v); + seq_printf(sf, "BS_DRIFT_TO:\t\t0x%X\n", v[0]); + seq_printf(sf, "BS_FRAME_TO:\t\t0x%X\n", v[1]); + seq_printf(sf, "BS_TE_TO:\t\t0x%X\n", v[2]); + + get_values_from_reg(c->reg, 0x110, 3, v); + for (i = 0; i < 3; i++) + seq_printf(sf, "BS_T%u_INTERVAL:\t\t0x%X\n", i, v[i]); + + get_values_from_reg(c->reg, 0x120, 5, v); + for (i = 0; i < 2; i++) { + seq_printf(sf, "BS_CRC%u_LOW:\t\t0x%X\n", i, v[i << 1]); + seq_printf(sf, "BS_CRC%u_HIGH:\t\t0x%X\n", i, v[(i << 1) + 1]); + } + seq_printf(sf, "BS_USER:\t\t0x%X\n", v[4]); +} + +struct komeda_component_funcs d71_timing_ctrlr_funcs = { + .update = d71_timing_ctrlr_update, + .disable = d71_timing_ctrlr_disable, + .dump_register = d71_timing_ctrlr_dump, +}; + +static int d71_timing_ctrlr_init(struct d71_dev *d71, + struct block_header *blk, u32 __iomem *reg) +{ + struct komeda_component *c; + struct komeda_timing_ctrlr *ctrlr; + u32 pipe_id, comp_id; + + get_resources_id(blk->block_info, &pipe_id, &comp_id); + + c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*ctrlr), + KOMEDA_COMPONENT_TIMING_CTRLR, + BLOCK_INFO_INPUT_ID(blk->block_info), + &d71_timing_ctrlr_funcs, + 1, BIT(KOMEDA_COMPONENT_IPS0 + pipe_id), + BS_NUM_OUTPUT_IDS, reg, "DOU%d_BS", pipe_id); + if (IS_ERR(c)) { + DRM_ERROR("Failed to add display_ctrl component\n"); + return PTR_ERR(c); + } + + ctrlr = to_ctrlr(c); + + ctrlr->supports_dual_link = true; + + return 0; +} + +int d71_probe_block(struct d71_dev *d71, + struct block_header *blk, u32 __iomem *reg) +{ + struct d71_pipeline *pipe; + int blk_id = BLOCK_INFO_BLK_ID(blk->block_info); + + int err = 0; + + switch (BLOCK_INFO_BLK_TYPE(blk->block_info)) { + case D71_BLK_TYPE_GCU: + break; + + case D71_BLK_TYPE_LPU: + pipe = d71->pipes[blk_id]; + pipe->lpu_addr = reg; + break; + + case D71_BLK_TYPE_LPU_LAYER: + err = d71_layer_init(d71, blk, reg); + break; + + case D71_BLK_TYPE_LPU_WB_LAYER: + err = d71_wb_layer_init(d71, blk, reg); + break; + + case D71_BLK_TYPE_CU: + pipe = d71->pipes[blk_id]; + pipe->cu_addr = reg; + err = d71_compiz_init(d71, blk, reg); + break; + + case D71_BLK_TYPE_CU_SPLITTER: + case D71_BLK_TYPE_CU_SCALER: + case D71_BLK_TYPE_CU_MERGER: + break; + + case D71_BLK_TYPE_DOU: + pipe = d71->pipes[blk_id]; + pipe->dou_addr = reg; + break; + + case D71_BLK_TYPE_DOU_IPS: + err = d71_improc_init(d71, blk, reg); + break; + + case D71_BLK_TYPE_DOU_FT_COEFF: + pipe = d71->pipes[blk_id]; + pipe->dou_ft_coeff_addr = reg; + break; + + case D71_BLK_TYPE_DOU_BS: + err = d71_timing_ctrlr_init(d71, blk, reg); + break; + + case D71_BLK_TYPE_GLB_LT_COEFF: + break; + + case D71_BLK_TYPE_GLB_SCL_COEFF: + d71->glb_scl_coeff_addr[blk_id] = reg; + break; + + default: + DRM_ERROR("Unknown block (block_info: 0x%x) is found\n", + blk->block_info); + err = -EINVAL; + break; + } + + return err; +} diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c index edbf9daa1545..72631d673f85 100644 --- a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c +++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c @@ -4,13 +4,375 @@ * Author: James.Qian.Wang * */ + +#include +#include "d71_dev.h" #include "malidp_io.h" -#include "komeda_dev.h" + +static u64 get_lpu_event(struct d71_pipeline *d71_pipeline) +{ + u32 __iomem *reg = d71_pipeline->lpu_addr; + u32 status, raw_status; + u64 evts = 0ULL; + + raw_status = malidp_read32(reg, BLK_IRQ_RAW_STATUS); + if (raw_status & LPU_IRQ_IBSY) + evts |= KOMEDA_EVENT_IBSY; + if (raw_status & LPU_IRQ_EOW) + evts |= KOMEDA_EVENT_EOW; + + if (raw_status & (LPU_IRQ_ERR | LPU_IRQ_IBSY)) { + u32 restore = 0, tbu_status; + /* Check error of LPU status */ + status = malidp_read32(reg, BLK_STATUS); + if (status & LPU_STATUS_AXIE) { + restore |= LPU_STATUS_AXIE; + evts |= KOMEDA_ERR_AXIE; + } + if (status & LPU_STATUS_ACE0) { + restore |= LPU_STATUS_ACE0; + evts |= KOMEDA_ERR_ACE0; + } + if (status & LPU_STATUS_ACE1) { + restore |= LPU_STATUS_ACE1; + evts |= KOMEDA_ERR_ACE1; + } + if (status & LPU_STATUS_ACE2) { + restore |= LPU_STATUS_ACE2; + evts |= KOMEDA_ERR_ACE2; + } + if (status & LPU_STATUS_ACE3) { + restore |= LPU_STATUS_ACE3; + evts |= KOMEDA_ERR_ACE3; + } + if (restore != 0) + malidp_write32_mask(reg, BLK_STATUS, restore, 0); + + restore = 0; + /* Check errors of TBU status */ + tbu_status = malidp_read32(reg, LPU_TBU_STATUS); + if (tbu_status & LPU_TBU_STATUS_TCF) { + restore |= LPU_TBU_STATUS_TCF; + evts |= KOMEDA_ERR_TCF; + } + if (tbu_status & LPU_TBU_STATUS_TTNG) { + restore |= LPU_TBU_STATUS_TTNG; + evts |= KOMEDA_ERR_TTNG; + } + if (tbu_status & LPU_TBU_STATUS_TITR) { + restore |= LPU_TBU_STATUS_TITR; + evts |= KOMEDA_ERR_TITR; + } + if (tbu_status & LPU_TBU_STATUS_TEMR) { + restore |= LPU_TBU_STATUS_TEMR; + evts |= KOMEDA_ERR_TEMR; + } + if (tbu_status & LPU_TBU_STATUS_TTF) { + restore |= LPU_TBU_STATUS_TTF; + evts |= KOMEDA_ERR_TTF; + } + if (restore != 0) + malidp_write32_mask(reg, LPU_TBU_STATUS, restore, 0); + } + + malidp_write32(reg, BLK_IRQ_CLEAR, raw_status); + return evts; +} + +static u64 get_cu_event(struct d71_pipeline *d71_pipeline) +{ + u32 __iomem *reg = d71_pipeline->cu_addr; + u32 status, raw_status; + u64 evts = 0ULL; + + raw_status = malidp_read32(reg, BLK_IRQ_RAW_STATUS); + if (raw_status & CU_IRQ_OVR) + evts |= KOMEDA_EVENT_OVR; + + if (raw_status & (CU_IRQ_ERR | CU_IRQ_OVR)) { + status = malidp_read32(reg, BLK_STATUS) & 0x7FFFFFFF; + if (status & CU_STATUS_CPE) + evts |= KOMEDA_ERR_CPE; + if (status & CU_STATUS_ZME) + evts |= KOMEDA_ERR_ZME; + if (status & CU_STATUS_CFGE) + evts |= KOMEDA_ERR_CFGE; + if (status) + malidp_write32_mask(reg, BLK_STATUS, status, 0); + } + + malidp_write32(reg, BLK_IRQ_CLEAR, raw_status); + + return evts; +} + +static u64 get_dou_event(struct d71_pipeline *d71_pipeline) +{ + u32 __iomem *reg = d71_pipeline->dou_addr; + u32 status, raw_status; + u64 evts = 0ULL; + + raw_status = malidp_read32(reg, BLK_IRQ_RAW_STATUS); + if (raw_status & DOU_IRQ_PL0) + evts |= KOMEDA_EVENT_VSYNC; + if (raw_status & DOU_IRQ_UND) + evts |= KOMEDA_EVENT_URUN; + + if (raw_status & (DOU_IRQ_ERR | DOU_IRQ_UND)) { + u32 restore = 0; + + status = malidp_read32(reg, BLK_STATUS); + if (status & DOU_STATUS_DRIFTTO) { + restore |= DOU_STATUS_DRIFTTO; + evts |= KOMEDA_ERR_DRIFTTO; + } + if (status & DOU_STATUS_FRAMETO) { + restore |= DOU_STATUS_FRAMETO; + evts |= KOMEDA_ERR_FRAMETO; + } + if (status & DOU_STATUS_TETO) { + restore |= DOU_STATUS_TETO; + evts |= KOMEDA_ERR_TETO; + } + if (status & DOU_STATUS_CSCE) { + restore |= DOU_STATUS_CSCE; + evts |= KOMEDA_ERR_CSCE; + } + + if (restore != 0) + malidp_write32_mask(reg, BLK_STATUS, restore, 0); + } + + malidp_write32(reg, BLK_IRQ_CLEAR, raw_status); + return evts; +} + +static u64 get_pipeline_event(struct d71_pipeline *d71_pipeline, u32 gcu_status) +{ + u32 evts = 0ULL; + + if (gcu_status & (GLB_IRQ_STATUS_LPU0 | GLB_IRQ_STATUS_LPU1)) + evts |= get_lpu_event(d71_pipeline); + + if (gcu_status & (GLB_IRQ_STATUS_CU0 | GLB_IRQ_STATUS_CU1)) + evts |= get_cu_event(d71_pipeline); + + if (gcu_status & (GLB_IRQ_STATUS_DOU0 | GLB_IRQ_STATUS_DOU1)) + evts |= get_dou_event(d71_pipeline); + + return evts; +} + +static irqreturn_t +d71_irq_handler(struct komeda_dev *mdev, struct komeda_events *evts) +{ + struct d71_dev *d71 = mdev->chip_data; + u32 status, gcu_status, raw_status; + + gcu_status = malidp_read32(d71->gcu_addr, GLB_IRQ_STATUS); + + if (gcu_status & GLB_IRQ_STATUS_GCU) { + raw_status = malidp_read32(d71->gcu_addr, BLK_IRQ_RAW_STATUS); + if (raw_status & GCU_IRQ_CVAL0) + evts->pipes[0] |= KOMEDA_EVENT_FLIP; + if (raw_status & GCU_IRQ_CVAL1) + evts->pipes[1] |= KOMEDA_EVENT_FLIP; + if (raw_status & GCU_IRQ_ERR) { + status = malidp_read32(d71->gcu_addr, BLK_STATUS); + if (status & GCU_STATUS_MERR) { + evts->global |= KOMEDA_ERR_MERR; + malidp_write32_mask(d71->gcu_addr, BLK_STATUS, + GCU_STATUS_MERR, 0); + } + } + + malidp_write32(d71->gcu_addr, BLK_IRQ_CLEAR, raw_status); + } + + if (gcu_status & GLB_IRQ_STATUS_PIPE0) + evts->pipes[0] |= get_pipeline_event(d71->pipes[0], gcu_status); + + if (gcu_status & GLB_IRQ_STATUS_PIPE1) + evts->pipes[1] |= get_pipeline_event(d71->pipes[1], gcu_status); + + return gcu_status ? IRQ_HANDLED : IRQ_NONE; +} + +#define ENABLED_GCU_IRQS (GCU_IRQ_CVAL0 | GCU_IRQ_CVAL1 | \ + GCU_IRQ_MODE | GCU_IRQ_ERR) +#define ENABLED_LPU_IRQS (LPU_IRQ_IBSY | LPU_IRQ_ERR | LPU_IRQ_EOW) +#define ENABLED_CU_IRQS (CU_IRQ_OVR | CU_IRQ_ERR) +#define ENABLED_DOU_IRQS (DOU_IRQ_UND | DOU_IRQ_ERR) + +static int d71_enable_irq(struct komeda_dev *mdev) +{ + struct d71_dev *d71 = mdev->chip_data; + struct d71_pipeline *pipe; + u32 i; + + malidp_write32_mask(d71->gcu_addr, BLK_IRQ_MASK, + ENABLED_GCU_IRQS, ENABLED_GCU_IRQS); + for (i = 0; i < d71->num_pipelines; i++) { + pipe = d71->pipes[i]; + malidp_write32_mask(pipe->cu_addr, BLK_IRQ_MASK, + ENABLED_CU_IRQS, ENABLED_CU_IRQS); + malidp_write32_mask(pipe->lpu_addr, BLK_IRQ_MASK, + ENABLED_LPU_IRQS, ENABLED_LPU_IRQS); + malidp_write32_mask(pipe->dou_addr, BLK_IRQ_MASK, + ENABLED_DOU_IRQS, ENABLED_DOU_IRQS); + } + return 0; +} + +static int d71_disable_irq(struct komeda_dev *mdev) +{ + struct d71_dev *d71 = mdev->chip_data; + struct d71_pipeline *pipe; + u32 i; + + malidp_write32_mask(d71->gcu_addr, BLK_IRQ_MASK, ENABLED_GCU_IRQS, 0); + for (i = 0; i < d71->num_pipelines; i++) { + pipe = d71->pipes[i]; + malidp_write32_mask(pipe->cu_addr, BLK_IRQ_MASK, + ENABLED_CU_IRQS, 0); + malidp_write32_mask(pipe->lpu_addr, BLK_IRQ_MASK, + ENABLED_LPU_IRQS, 0); + malidp_write32_mask(pipe->dou_addr, BLK_IRQ_MASK, + ENABLED_DOU_IRQS, 0); + } + return 0; +} + +static int d71_reset(struct d71_dev *d71) +{ + u32 __iomem *gcu = d71->gcu_addr; + int ret; + + malidp_write32_mask(gcu, BLK_CONTROL, + GCU_CONTROL_SRST, GCU_CONTROL_SRST); + + ret = dp_wait_cond(!(malidp_read32(gcu, BLK_CONTROL) & GCU_CONTROL_SRST), + 100, 1000, 10000); + + return ret > 0 ? 0 : -ETIMEDOUT; +} + +void d71_read_block_header(u32 __iomem *reg, struct block_header *blk) +{ + int i; + + blk->block_info = malidp_read32(reg, BLK_BLOCK_INFO); + if (BLOCK_INFO_BLK_TYPE(blk->block_info) == D71_BLK_TYPE_RESERVED) + return; + + blk->pipeline_info = malidp_read32(reg, BLK_PIPELINE_INFO); + + /* get valid input and output ids */ + for (i = 0; i < PIPELINE_INFO_N_VALID_INPUTS(blk->pipeline_info); i++) + blk->input_ids[i] = malidp_read32(reg + i, BLK_VALID_INPUT_ID0); + for (i = 0; i < PIPELINE_INFO_N_OUTPUTS(blk->pipeline_info); i++) + blk->output_ids[i] = malidp_read32(reg + i, BLK_OUTPUT_ID0); +} + +static void d71_cleanup(struct komeda_dev *mdev) +{ + struct d71_dev *d71 = mdev->chip_data; + + if (!d71) + return; + + devm_kfree(mdev->dev, d71); + mdev->chip_data = NULL; +} static int d71_enum_resources(struct komeda_dev *mdev) { - /* TODO add enum resources */ - return -1; + struct d71_dev *d71; + struct komeda_pipeline *pipe; + struct block_header blk; + u32 __iomem *blk_base; + u32 i, value, offset; + int err; + + d71 = devm_kzalloc(mdev->dev, sizeof(*d71), GFP_KERNEL); + if (!d71) + return -ENOMEM; + + mdev->chip_data = d71; + d71->mdev = mdev; + d71->gcu_addr = mdev->reg_base; + d71->periph_addr = mdev->reg_base + (D71_BLOCK_OFFSET_PERIPH >> 2); + + err = d71_reset(d71); + if (err) { + DRM_ERROR("Fail to reset d71 device.\n"); + goto err_cleanup; + } + + /* probe GCU */ + value = malidp_read32(d71->gcu_addr, GLB_CORE_INFO); + d71->num_blocks = value & 0xFF; + d71->num_pipelines = (value >> 8) & 0x7; + + if (d71->num_pipelines > D71_MAX_PIPELINE) { + DRM_ERROR("d71 supports %d pipelines, but got: %d.\n", + D71_MAX_PIPELINE, d71->num_pipelines); + err = -EINVAL; + goto err_cleanup; + } + + /* probe PERIPH */ + value = malidp_read32(d71->periph_addr, BLK_BLOCK_INFO); + if (BLOCK_INFO_BLK_TYPE(value) != D71_BLK_TYPE_PERIPH) { + DRM_ERROR("access blk periph but got blk: %d.\n", + BLOCK_INFO_BLK_TYPE(value)); + err = -EINVAL; + goto err_cleanup; + } + + value = malidp_read32(d71->periph_addr, PERIPH_CONFIGURATION_ID); + + d71->max_line_size = value & PERIPH_MAX_LINE_SIZE ? 4096 : 2048; + d71->max_vsize = 4096; + d71->num_rich_layers = value & PERIPH_NUM_RICH_LAYERS ? 2 : 1; + d71->supports_dual_link = value & PERIPH_SPLIT_EN ? true : false; + d71->integrates_tbu = value & PERIPH_TBU_EN ? true : false; + + for (i = 0; i < d71->num_pipelines; i++) { + pipe = komeda_pipeline_add(mdev, sizeof(struct d71_pipeline), + NULL); + if (IS_ERR(pipe)) { + err = PTR_ERR(pipe); + goto err_cleanup; + } + d71->pipes[i] = to_d71_pipeline(pipe); + } + + /* loop the register blks and probe */ + i = 2; /* exclude GCU and PERIPH */ + offset = D71_BLOCK_SIZE; /* skip GCU */ + while (i < d71->num_blocks) { + blk_base = mdev->reg_base + (offset >> 2); + + d71_read_block_header(blk_base, &blk); + if (BLOCK_INFO_BLK_TYPE(blk.block_info) != D71_BLK_TYPE_RESERVED) { + err = d71_probe_block(d71, &blk, blk_base); + if (err) + goto err_cleanup; + i++; + } + + offset += D71_BLOCK_SIZE; + } + + DRM_DEBUG("total %d (out of %d) blocks are found.\n", + i, d71->num_blocks); + + return 0; + +err_cleanup: + d71_cleanup(mdev); + return err; } #define __HW_ID(__group, __format) \ @@ -93,13 +455,12 @@ static void d71_init_fmt_tbl(struct komeda_dev *mdev) static struct komeda_dev_funcs d71_chip_funcs = { .init_format_table = d71_init_fmt_tbl, .enum_resources = d71_enum_resources, - .cleanup = NULL, + .cleanup = d71_cleanup, + .irq_handler = d71_irq_handler, + .enable_irq = d71_enable_irq, + .disable_irq = d71_disable_irq, }; -#define GLB_ARCH_ID 0x000 -#define GLB_CORE_ID 0x004 -#define GLB_CORE_INFO 0x008 - struct komeda_dev_funcs * d71_identify(u32 __iomem *reg_base, struct komeda_chip_info *chip) { diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.h b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.h new file mode 100644 index 000000000000..7465c57d9774 --- /dev/null +++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.h @@ -0,0 +1,50 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * (C) COPYRIGHT 2018 ARM Limited. All rights reserved. + * Author: James.Qian.Wang + * + */ +#ifndef _D71_DEV_H_ +#define _D71_DEV_H_ + +#include "komeda_dev.h" +#include "komeda_pipeline.h" +#include "d71_regs.h" + +struct d71_pipeline { + struct komeda_pipeline base; + + /* d71 private pipeline blocks */ + u32 __iomem *lpu_addr; + u32 __iomem *cu_addr; + u32 __iomem *dou_addr; + u32 __iomem *dou_ft_coeff_addr; /* forward transform coeffs table */ +}; + +struct d71_dev { + struct komeda_dev *mdev; + + int num_blocks; + int num_pipelines; + int num_rich_layers; + u32 max_line_size; + u32 max_vsize; + u32 supports_dual_link : 1; + u32 integrates_tbu : 1; + + /* global register blocks */ + u32 __iomem *gcu_addr; + /* scaling coeffs table */ + u32 __iomem *glb_scl_coeff_addr[D71_MAX_GLB_SCL_COEFF]; + u32 __iomem *periph_addr; + + struct d71_pipeline *pipes[D71_MAX_PIPELINE]; +}; + +#define to_d71_pipeline(x) container_of(x, struct d71_pipeline, base) + +int d71_probe_block(struct d71_dev *d71, + struct block_header *blk, u32 __iomem *reg); +void d71_read_block_header(u32 __iomem *reg, struct block_header *blk); + +#endif /* !_D71_DEV_H_ */ diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h b/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h new file mode 100644 index 000000000000..2d5e6d00b42c --- /dev/null +++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h @@ -0,0 +1,530 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * (C) COPYRIGHT 2018 ARM Limited. All rights reserved. + * Author: James.Qian.Wang + * + */ +#ifndef _D71_REG_H_ +#define _D71_REG_H_ + +/* Common block registers offset */ +#define BLK_BLOCK_INFO 0x000 +#define BLK_PIPELINE_INFO 0x004 +#define BLK_VALID_INPUT_ID0 0x020 +#define BLK_OUTPUT_ID0 0x060 +#define BLK_INPUT_ID0 0x080 +#define BLK_IRQ_RAW_STATUS 0x0A0 +#define BLK_IRQ_CLEAR 0x0A4 +#define BLK_IRQ_MASK 0x0A8 +#define BLK_IRQ_STATUS 0x0AC +#define BLK_STATUS 0x0B0 +#define BLK_INFO 0x0C0 +#define BLK_CONTROL 0x0D0 +#define BLK_SIZE 0x0D4 +#define BLK_IN_SIZE 0x0E0 + +#define BLK_P0_PTR_LOW 0x100 +#define BLK_P0_PTR_HIGH 0x104 +#define BLK_P0_STRIDE 0x108 +#define BLK_P1_PTR_LOW 0x110 +#define BLK_P1_PTR_HIGH 0x114 +#define BLK_P1_STRIDE 0x118 +#define BLK_P2_PTR_LOW 0x120 +#define BLK_P2_PTR_HIGH 0x124 + +#define BLOCK_INFO_N_SUBBLKS(x) ((x) & 0x000F) +#define BLOCK_INFO_BLK_ID(x) (((x) & 0x00F0) >> 4) +#define BLOCK_INFO_BLK_TYPE(x) (((x) & 0xFF00) >> 8) +#define BLOCK_INFO_INPUT_ID(x) ((x) & 0xFFF0) +#define BLOCK_INFO_TYPE_ID(x) (((x) & 0x0FF0) >> 4) + +#define PIPELINE_INFO_N_OUTPUTS(x) ((x) & 0x000F) +#define PIPELINE_INFO_N_VALID_INPUTS(x) (((x) & 0x0F00) >> 8) + +/* Common block control register bits */ +#define BLK_CTRL_EN BIT(0) +/* Common size macro */ +#define HV_SIZE(h, v) (((h) & 0x1FFF) + (((v) & 0x1FFF) << 16)) +#define HV_OFFSET(h, v) (((h) & 0xFFF) + (((v) & 0xFFF) << 16)) +#define HV_CROP(h, v) (((h) & 0xFFF) + (((v) & 0xFFF) << 16)) + +/* AD_CONTROL register */ +#define AD_CONTROL 0x160 + +/* AD_CONTROL register bits */ +#define AD_AEN BIT(0) +#define AD_YT BIT(1) +#define AD_BS BIT(2) +#define AD_WB BIT(3) +#define AD_TH BIT(4) + +/* Global Control Unit */ +#define GLB_ARCH_ID 0x000 +#define GLB_CORE_ID 0x004 +#define GLB_CORE_INFO 0x008 +#define GLB_IRQ_STATUS 0x010 + +#define GCU_CONFIG_VALID0 0x0D4 +#define GCU_CONFIG_VALID1 0x0D8 + +/* GCU_CONTROL_BITS */ +#define GCU_CONTROL_MODE(x) ((x) & 0x7) +#define GCU_CONTROL_SRST BIT(16) + +/* GCU opmode */ +#define INACTIVE_MODE 0 +#define TBU_CONNECT_MODE 1 +#define TBU_DISCONNECT_MODE 2 +#define DO0_ACTIVE_MODE 3 +#define DO1_ACTIVE_MODE 4 +#define DO01_ACTIVE_MODE 5 + +/* GLB_IRQ_STATUS bits */ +#define GLB_IRQ_STATUS_GCU BIT(0) +#define GLB_IRQ_STATUS_LPU0 BIT(8) +#define GLB_IRQ_STATUS_LPU1 BIT(9) +#define GLB_IRQ_STATUS_ATU0 BIT(10) +#define GLB_IRQ_STATUS_ATU1 BIT(11) +#define GLB_IRQ_STATUS_ATU2 BIT(12) +#define GLB_IRQ_STATUS_ATU3 BIT(13) +#define GLB_IRQ_STATUS_CU0 BIT(16) +#define GLB_IRQ_STATUS_CU1 BIT(17) +#define GLB_IRQ_STATUS_DOU0 BIT(24) +#define GLB_IRQ_STATUS_DOU1 BIT(25) + +#define GLB_IRQ_STATUS_PIPE0 (GLB_IRQ_STATUS_LPU0 |\ + GLB_IRQ_STATUS_ATU0 |\ + GLB_IRQ_STATUS_ATU1 |\ + GLB_IRQ_STATUS_CU0 |\ + GLB_IRQ_STATUS_DOU0) + +#define GLB_IRQ_STATUS_PIPE1 (GLB_IRQ_STATUS_LPU1 |\ + GLB_IRQ_STATUS_ATU2 |\ + GLB_IRQ_STATUS_ATU3 |\ + GLB_IRQ_STATUS_CU1 |\ + GLB_IRQ_STATUS_DOU1) + +#define GLB_IRQ_STATUS_ATU (GLB_IRQ_STATUS_ATU0 |\ + GLB_IRQ_STATUS_ATU1 |\ + GLB_IRQ_STATUS_ATU2 |\ + GLB_IRQ_STATUS_ATU3) + +/* GCU_IRQ_BITS */ +#define GCU_IRQ_CVAL0 BIT(0) +#define GCU_IRQ_CVAL1 BIT(1) +#define GCU_IRQ_MODE BIT(4) +#define GCU_IRQ_ERR BIT(11) + +/* GCU_STATUS_BITS */ +#define GCU_STATUS_MODE(x) ((x) & 0x7) +#define GCU_STATUS_MERR BIT(4) +#define GCU_STATUS_TCS0 BIT(8) +#define GCU_STATUS_TCS1 BIT(9) +#define GCU_STATUS_ACTIVE BIT(31) + +/* GCU_CONFIG_VALIDx BITS */ +#define GCU_CONFIG_CVAL BIT(0) + +/* PERIPHERAL registers */ +#define PERIPH_MAX_LINE_SIZE BIT(0) +#define PERIPH_NUM_RICH_LAYERS BIT(4) +#define PERIPH_SPLIT_EN BIT(8) +#define PERIPH_TBU_EN BIT(12) +#define PERIPH_AFBC_DMA_EN BIT(16) +#define PERIPH_CONFIGURATION_ID 0x1D4 + +/* LPU register */ +#define LPU_TBU_STATUS 0x0B4 +#define LPU_RAXI_CONTROL 0x0D0 +#define LPU_WAXI_CONTROL 0x0D4 +#define LPU_TBU_CONTROL 0x0D8 + +/* LPU_xAXI_CONTROL_BITS */ +#define TO_RAXI_AOUTSTDCAPB(x) (x) +#define TO_RAXI_BOUTSTDCAPB(x) ((x) << 8) +#define TO_RAXI_BEN(x) ((x) << 15) +#define TO_xAXI_BURSTLEN(x) ((x) << 16) +#define TO_xAXI_AxQOS(x) ((x) << 24) +#define TO_xAXI_ORD(x) ((x) << 31) +#define TO_WAXI_OUTSTDCAPB(x) (x) + +#define RAXI_AOUTSTDCAPB_MASK 0x7F +#define RAXI_BOUTSTDCAPB_MASK 0x7F00 +#define RAXI_BEN_MASK BIT(15) +#define xAXI_BURSTLEN_MASK 0x3F0000 +#define xAXI_AxQOS_MASK 0xF000000 +#define xAXI_ORD_MASK BIT(31) +#define WAXI_OUTSTDCAPB_MASK 0x3F + +/* LPU_TBU_CONTROL BITS */ +#define TO_TBU_DOUTSTDCAPB(x) (x) +#define TBU_DOUTSTDCAPB_MASK 0x3F + +/* LPU_IRQ_BITS */ +#define LPU_IRQ_IBSY BIT(10) +#define LPU_IRQ_ERR BIT(11) +#define LPU_IRQ_EOW BIT(12) +#define LPU_IRQ_PL0 BIT(13) + +/* LPU_STATUS_BITS */ +#define LPU_STATUS_AXIED(x) ((x) & 0xF) +#define LPU_STATUS_AXIE BIT(4) +#define LPU_STATUS_AXIRP BIT(5) +#define LPU_STATUS_AXIWP BIT(6) +#define LPU_STATUS_ACE0 BIT(16) +#define LPU_STATUS_ACE1 BIT(17) +#define LPU_STATUS_ACE2 BIT(18) +#define LPU_STATUS_ACE3 BIT(19) +#define LPU_STATUS_ACTIVE BIT(31) + +#define AXIEID_MASK 0xF +#define AXIE_MASK LPU_STATUS_AXIE +#define AXIRP_MASK LPU_STATUS_AXIRP +#define AXIWP_MASK LPU_STATUS_AXIWP + +#define FROM_AXIEID(reg) ((reg) & AXIEID_MASK) +#define TO_AXIE(x) ((x) << 4) +#define FROM_AXIRP(reg) (((reg) & AXIRP_MASK) >> 5) +#define FROM_AXIWP(reg) (((reg) & AXIWP_MASK) >> 6) + +/* LPU_TBU_STATUS_BITS */ +#define LPU_TBU_STATUS_TCF BIT(1) +#define LPU_TBU_STATUS_TTNG BIT(2) +#define LPU_TBU_STATUS_TITR BIT(8) +#define LPU_TBU_STATUS_TEMR BIT(16) +#define LPU_TBU_STATUS_TTF BIT(31) + +/* LPU_TBU_CONTROL BITS */ +#define LPU_TBU_CTRL_TLBPEN BIT(16) + +/* CROSSBAR CONTROL BITS */ +#define CBU_INPUT_CTRL_EN BIT(0) +#define CBU_NUM_INPUT_IDS 5 +#define CBU_NUM_OUTPUT_IDS 5 + +/* CU register */ +#define CU_BG_COLOR 0x0DC +#define CU_INPUT0_SIZE 0x0E0 +#define CU_INPUT0_OFFSET 0x0E4 +#define CU_INPUT0_CONTROL 0x0E8 +#define CU_INPUT1_SIZE 0x0F0 +#define CU_INPUT1_OFFSET 0x0F4 +#define CU_INPUT1_CONTROL 0x0F8 +#define CU_INPUT2_SIZE 0x100 +#define CU_INPUT2_OFFSET 0x104 +#define CU_INPUT2_CONTROL 0x108 +#define CU_INPUT3_SIZE 0x110 +#define CU_INPUT3_OFFSET 0x114 +#define CU_INPUT3_CONTROL 0x118 +#define CU_INPUT4_SIZE 0x120 +#define CU_INPUT4_OFFSET 0x124 +#define CU_INPUT4_CONTROL 0x128 + +#define CU_PER_INPUT_REGS 4 + +#define CU_NUM_INPUT_IDS 5 +#define CU_NUM_OUTPUT_IDS 1 + +/* CU control register bits */ +#define CU_CTRL_COPROC BIT(0) + +/* CU_IRQ_BITS */ +#define CU_IRQ_OVR BIT(9) +#define CU_IRQ_ERR BIT(11) + +/* CU_STATUS_BITS */ +#define CU_STATUS_CPE BIT(0) +#define CU_STATUS_ZME BIT(1) +#define CU_STATUS_CFGE BIT(2) +#define CU_STATUS_ACTIVE BIT(31) + +/* CU input control register bits */ +#define CU_INPUT_CTRL_EN BIT(0) +#define CU_INPUT_CTRL_PAD BIT(1) +#define CU_INPUT_CTRL_PMUL BIT(2) +#define CU_INPUT_CTRL_ALPHA(x) (((x) & 0xFF) << 8) + +/* DOU register */ + +/* DOU_IRQ_BITS */ +#define DOU_IRQ_UND BIT(8) +#define DOU_IRQ_ERR BIT(11) +#define DOU_IRQ_PL0 BIT(13) +#define DOU_IRQ_PL1 BIT(14) + +/* DOU_STATUS_BITS */ +#define DOU_STATUS_DRIFTTO BIT(0) +#define DOU_STATUS_FRAMETO BIT(1) +#define DOU_STATUS_TETO BIT(2) +#define DOU_STATUS_CSCE BIT(8) +#define DOU_STATUS_ACTIVE BIT(31) + +/* Layer registers */ +#define LAYER_INFO 0x0C0 +#define LAYER_R_CONTROL 0x0D4 +#define LAYER_FMT 0x0D8 +#define LAYER_LT_COEFFTAB 0x0DC +#define LAYER_PALPHA 0x0E4 + +#define LAYER_YUV_RGB_COEFF0 0x130 + +#define LAYER_AD_H_CROP 0x164 +#define LAYER_AD_V_CROP 0x168 + +#define LAYER_RGB_RGB_COEFF0 0x170 + +/* L_CONTROL_BITS */ +#define L_EN BIT(0) +#define L_IT BIT(4) +#define L_R2R BIT(5) +#define L_FT BIT(6) +#define L_ROT(x) (((x) & 3) << 8) +#define L_HFLIP BIT(10) +#define L_VFLIP BIT(11) +#define L_TBU_EN BIT(16) +#define L_A_RCACHE(x) (((x) & 0xF) << 28) +#define L_ROT_R0 0 +#define L_ROT_R90 1 +#define L_ROT_R180 2 +#define L_ROT_R270 3 + +/* LAYER_R_CONTROL BITS */ +#define LR_CHI422_BILINEAR 0 +#define LR_CHI422_REPLICATION 1 +#define LR_CHI420_JPEG (0 << 2) +#define LR_CHI420_MPEG (1 << 2) + +#define L_ITSEL(x) ((x) & 0xFFF) +#define L_FTSEL(x) (((x) & 0xFFF) << 16) + +#define LAYER_PER_PLANE_REGS 4 + +/* Layer_WR registers */ +#define LAYER_WR_PROG_LINE 0x0D4 +#define LAYER_WR_FORMAT 0x0D8 + +/* Layer_WR control bits */ +#define LW_OFM BIT(4) +#define LW_LALPHA(x) (((x) & 0xFF) << 8) +#define LW_A_WCACHE(x) (((x) & 0xF) << 28) +#define LW_TBU_EN BIT(16) + +#define AxCACHE_MASK 0xF0000000 + +/* Layer AXI R/W cache setting */ +#define AxCACHE_B BIT(0) /* Bufferable */ +#define AxCACHE_M BIT(1) /* Modifiable */ +#define AxCACHE_RA BIT(2) /* Read-Allocate */ +#define AxCACHE_WA BIT(3) /* Write-Allocate */ + +/* Layer info bits */ +#define L_INFO_RF BIT(0) +#define L_INFO_CM BIT(1) +#define L_INFO_ABUF_SIZE(x) (((x) >> 4) & 0x7) + +/* Scaler registers */ +#define SC_COEFFTAB 0x0DC +#define SC_OUT_SIZE 0x0E4 +#define SC_H_CROP 0x0E8 +#define SC_V_CROP 0x0EC +#define SC_H_INIT_PH 0x0F0 +#define SC_H_DELTA_PH 0x0F4 +#define SC_V_INIT_PH 0x0F8 +#define SC_V_DELTA_PH 0x0FC +#define SC_ENH_LIMITS 0x130 +#define SC_ENH_COEFF0 0x134 + +#define SC_MAX_ENH_COEFF 9 + +/* SC_CTRL_BITS */ +#define SC_CTRL_SCL BIT(0) +#define SC_CTRL_LS BIT(1) +#define SC_CTRL_AP BIT(4) +#define SC_CTRL_IENH BIT(8) +#define SC_CTRL_RGBSM BIT(16) +#define SC_CTRL_ASM BIT(17) + +#define SC_VTSEL(vtal) ((vtal) << 16) + +#define SC_NUM_INPUTS_IDS 1 +#define SC_NUM_OUTPUTS_IDS 1 + +#define MG_NUM_INPUTS_IDS 2 +#define MG_NUM_OUTPUTS_IDS 1 + +/* Merger registers */ +#define MG_INPUT_ID0 BLK_INPUT_ID0 +#define MG_INPUT_ID1 (MG_INPUT_ID0 + 4) +#define MG_SIZE BLK_SIZE + +/* Splitter registers */ +#define SP_OVERLAP_SIZE 0xD8 + +/* Backend registers */ +#define BS_INFO 0x0C0 +#define BS_PROG_LINE 0x0D4 +#define BS_PREFETCH_LINE 0x0D8 +#define BS_BG_COLOR 0x0DC +#define BS_ACTIVESIZE 0x0E0 +#define BS_HINTERVALS 0x0E4 +#define BS_VINTERVALS 0x0E8 +#define BS_SYNC 0x0EC +#define BS_DRIFT_TO 0x100 +#define BS_FRAME_TO 0x104 +#define BS_TE_TO 0x108 +#define BS_T0_INTERVAL 0x110 +#define BS_T1_INTERVAL 0x114 +#define BS_T2_INTERVAL 0x118 +#define BS_CRC0_LOW 0x120 +#define BS_CRC0_HIGH 0x124 +#define BS_CRC1_LOW 0x128 +#define BS_CRC1_HIGH 0x12C +#define BS_USER 0x130 + +/* BS control register bits */ +#define BS_CTRL_EN BIT(0) +#define BS_CTRL_VM BIT(1) +#define BS_CTRL_BM BIT(2) +#define BS_CTRL_HMASK BIT(4) +#define BS_CTRL_VD BIT(5) +#define BS_CTRL_TE BIT(8) +#define BS_CTRL_TS BIT(9) +#define BS_CTRL_TM BIT(12) +#define BS_CTRL_DL BIT(16) +#define BS_CTRL_SBS BIT(17) +#define BS_CTRL_CRC BIT(18) +#define BS_CTRL_PM BIT(20) + +/* BS active size/intervals */ +#define BS_H_INTVALS(hfp, hbp) (((hfp) & 0xFFF) + (((hbp) & 0x3FF) << 16)) +#define BS_V_INTVALS(vfp, vbp) (((vfp) & 0x3FFF) + (((vbp) & 0xFF) << 16)) + +/* BS_SYNC bits */ +#define BS_SYNC_HSW(x) ((x) & 0x3FF) +#define BS_SYNC_HSP BIT(12) +#define BS_SYNC_VSW(x) (((x) & 0xFF) << 16) +#define BS_SYNC_VSP BIT(28) + +#define BS_NUM_INPUT_IDS 0 +#define BS_NUM_OUTPUT_IDS 0 + +/* Image process registers */ +#define IPS_DEPTH 0x0D8 +#define IPS_RGB_RGB_COEFF0 0x130 +#define IPS_RGB_YUV_COEFF0 0x170 + +#define IPS_DEPTH_MARK 0xF + +/* IPS control register bits */ +#define IPS_CTRL_RGB BIT(0) +#define IPS_CTRL_FT BIT(4) +#define IPS_CTRL_YUV BIT(8) +#define IPS_CTRL_CHD422 BIT(9) +#define IPS_CTRL_CHD420 BIT(10) +#define IPS_CTRL_LPF BIT(11) +#define IPS_CTRL_DITH BIT(12) +#define IPS_CTRL_CLAMP BIT(16) +#define IPS_CTRL_SBS BIT(17) + +/* IPS info register bits */ +#define IPS_INFO_CHD420 BIT(10) + +#define IPS_NUM_INPUT_IDS 2 +#define IPS_NUM_OUTPUT_IDS 1 + +/* FT_COEFF block registers */ +#define FT_COEFF0 0x80 +#define GLB_IT_COEFF 0x80 + +/* GLB_SC_COEFF registers */ +#define GLB_SC_COEFF_ADDR 0x0080 +#define GLB_SC_COEFF_DATA 0x0084 +#define GLB_LT_COEFF_DATA 0x0080 + +#define GLB_SC_COEFF_MAX_NUM 1024 +#define GLB_LT_COEFF_NUM 65 +/* GLB_SC_ADDR */ +#define SC_COEFF_R_ADDR BIT(18) +#define SC_COEFF_G_ADDR BIT(17) +#define SC_COEFF_B_ADDR BIT(16) + +#define SC_COEFF_DATA(x, y) (((y) & 0xFFFF) | (((x) & 0xFFFF) << 16)) + +enum d71_blk_type { + D71_BLK_TYPE_GCU = 0x00, + D71_BLK_TYPE_LPU = 0x01, + D71_BLK_TYPE_CU = 0x02, + D71_BLK_TYPE_DOU = 0x03, + D71_BLK_TYPE_AEU = 0x04, + D71_BLK_TYPE_GLB_LT_COEFF = 0x05, + D71_BLK_TYPE_GLB_SCL_COEFF = 0x06, /* SH/SV scaler coeff */ + D71_BLK_TYPE_GLB_SC_COEFF = 0x07, + D71_BLK_TYPE_PERIPH = 0x08, + D71_BLK_TYPE_LPU_TRUSTED = 0x09, + D71_BLK_TYPE_AEU_TRUSTED = 0x0A, + D71_BLK_TYPE_LPU_LAYER = 0x10, + D71_BLK_TYPE_LPU_WB_LAYER = 0x11, + D71_BLK_TYPE_CU_SPLITTER = 0x20, + D71_BLK_TYPE_CU_SCALER = 0x21, + D71_BLK_TYPE_CU_MERGER = 0x22, + D71_BLK_TYPE_DOU_IPS = 0x30, + D71_BLK_TYPE_DOU_BS = 0x31, + D71_BLK_TYPE_DOU_FT_COEFF = 0x32, + D71_BLK_TYPE_AEU_DS = 0x40, + D71_BLK_TYPE_AEU_AES = 0x41, + D71_BLK_TYPE_RESERVED = 0xFF +}; + +/* Constant of components */ +#define D71_MAX_PIPELINE 2 +#define D71_PIPELINE_MAX_SCALERS 2 +#define D71_PIPELINE_MAX_LAYERS 4 + +#define D71_MAX_GLB_IT_COEFF 3 +#define D71_MAX_GLB_SCL_COEFF 4 + +#define D71_MAX_LAYERS_PER_LPU 4 +#define D71_BLOCK_MAX_INPUT 9 +#define D71_BLOCK_MAX_OUTPUT 5 +#define D71_MAX_SC_PER_CU 2 + +#define D71_BLOCK_OFFSET_PERIPH 0xFE00 +#define D71_BLOCK_SIZE 0x0200 + +#define D71_DEFAULT_PREPRETCH_LINE 5 +#define D71_BUS_WIDTH_16_BYTES 16 + +#define D71_MIN_LINE_SIZE 64 +#define D71_MIN_VERTICAL_SIZE 64 +#define D71_SC_MIN_LIN_SIZE 4 +#define D71_SC_MIN_VERTICAL_SIZE 4 +#define D71_SC_MAX_LIN_SIZE 2048 +#define D71_SC_MAX_VERTICAL_SIZE 4096 + +#define D71_SC_MAX_UPSCALING 64 +#define D71_SC_MAX_DOWNSCALING 6 +#define D71_SC_SPLIT_OVERLAP 8 +#define D71_SC_ENH_SPLIT_OVERLAP 1 + +#define D71_MG_MIN_MERGED_SIZE 4 +#define D71_MG_MAX_MERGED_HSIZE 4032 +#define D71_MG_MAX_MERGED_VSIZE 4096 + +#define D71_PALPHA_DEF_MAP 0xFFAA5500 +#define D71_LAYER_CONTROL_DEFAULT 0x30000000 +#define D71_WB_LAYER_CONTROL_DEFAULT 0x3000FF00 +#define D71_BS_CONTROL_DEFAULT 0x00000002 + +struct block_header { + u32 block_info; + u32 pipeline_info; + u32 input_ids[D71_BLOCK_MAX_INPUT]; + u32 output_ids[D71_BLOCK_MAX_OUTPUT]; +}; + +static inline u32 get_block_type(struct block_header *blk) +{ + return BLOCK_INFO_BLK_TYPE(blk->block_info); +} + +#endif /* !_D71_REG_H_ */ diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c b/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c index 3ca5718aa0c2..f88a14927be9 100644 --- a/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c +++ b/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c @@ -18,6 +18,24 @@ #include "komeda_dev.h" #include "komeda_kms.h" +void komeda_crtc_handle_event(struct komeda_crtc *kcrtc, + struct komeda_events *evts) +{ + struct drm_crtc *crtc = &kcrtc->base; + u32 events = evts->pipes[kcrtc->master->id]; + + if (events & KOMEDA_EVENT_VSYNC) + drm_crtc_handle_vblank(crtc); + + /* will handle it together with the write back support */ + if (events & KOMEDA_EVENT_EOW) + DRM_DEBUG("EOW.\n"); + + /* will handle it with crtc->flush */ + if (events & KOMEDA_EVENT_FLIP) + DRM_DEBUG("FLIP Done.\n"); +} + struct drm_crtc_helper_funcs komeda_crtc_helper_funcs = { }; diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_dev.c b/drivers/gpu/drm/arm/display/komeda/komeda_dev.c index 70e9bb7fa30c..24548b87e182 100644 --- a/drivers/gpu/drm/arm/display/komeda/komeda_dev.c +++ b/drivers/gpu/drm/arm/display/komeda/komeda_dev.c @@ -8,11 +8,57 @@ #include #include #include +#ifdef CONFIG_DEBUG_FS +#include +#include +#endif #include #include "komeda_dev.h" +static int komeda_register_show(struct seq_file *sf, void *x) +{ + struct komeda_dev *mdev = sf->private; + int i; + + if (mdev->funcs->dump_register) + mdev->funcs->dump_register(mdev, sf); + + for (i = 0; i < mdev->n_pipelines; i++) + komeda_pipeline_dump_register(mdev->pipelines[i], sf); + + return 0; +} + +static int komeda_register_open(struct inode *inode, struct file *filp) +{ + return single_open(filp, komeda_register_show, inode->i_private); +} + +static const struct file_operations komeda_register_fops = { + .owner = THIS_MODULE, + .open = komeda_register_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +#ifdef CONFIG_DEBUG_FS +static void komeda_debugfs_init(struct komeda_dev *mdev) +{ + if (!debugfs_initialized()) + return; + + mdev->debugfs_root = debugfs_create_dir("komeda", NULL); + if (IS_ERR_OR_NULL(mdev->debugfs_root)) + return; + + debugfs_create_file("register", 0444, mdev->debugfs_root, + mdev, &komeda_register_fops); +} +#endif + static int komeda_parse_pipe_dt(struct komeda_dev *mdev, struct device_node *np) { struct komeda_pipeline *pipe; @@ -53,6 +99,7 @@ static int komeda_parse_pipe_dt(struct komeda_dev *mdev, struct device_node *np) static int komeda_parse_dt(struct device *dev, struct komeda_dev *mdev) { + struct platform_device *pdev = to_platform_device(dev); struct device_node *child, *np = dev->of_node; struct clk *clk; int ret; @@ -62,6 +109,11 @@ static int komeda_parse_dt(struct device *dev, struct komeda_dev *mdev) return PTR_ERR(clk); mdev->mclk = clk; + mdev->irq = platform_get_irq(pdev, 0); + if (mdev->irq < 0) { + DRM_ERROR("could not get IRQ number.\n"); + return mdev->irq; + } for_each_available_child_of_node(np, child) { if (of_node_cmp(child->name, "pipeline") == 0) { @@ -147,6 +199,16 @@ struct komeda_dev *komeda_dev_create(struct device *dev) goto err_cleanup; } + err = komeda_assemble_pipelines(mdev); + if (err) { + DRM_ERROR("assemble display pipelines failed.\n"); + goto err_cleanup; + } + +#ifdef CONFIG_DEBUG_FS + komeda_debugfs_init(mdev); +#endif + return mdev; err_cleanup: @@ -160,6 +222,10 @@ void komeda_dev_destroy(struct komeda_dev *mdev) struct komeda_dev_funcs *funcs = mdev->funcs; int i; +#ifdef CONFIG_DEBUG_FS + debugfs_remove_recursive(mdev->debugfs_root); +#endif + for (i = 0; i < mdev->n_pipelines; i++) { komeda_pipeline_destroy(mdev, mdev->pipelines[i]); mdev->pipelines[i] = NULL; diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_dev.h b/drivers/gpu/drm/arm/display/komeda/komeda_dev.h index 0f77dead6a23..8eae2620ce77 100644 --- a/drivers/gpu/drm/arm/display/komeda/komeda_dev.h +++ b/drivers/gpu/drm/arm/display/komeda/komeda_dev.h @@ -13,6 +13,33 @@ #include "malidp_product.h" #include "komeda_format_caps.h" +#define KOMEDA_EVENT_VSYNC BIT_ULL(0) +#define KOMEDA_EVENT_FLIP BIT_ULL(1) +#define KOMEDA_EVENT_URUN BIT_ULL(2) +#define KOMEDA_EVENT_IBSY BIT_ULL(3) +#define KOMEDA_EVENT_OVR BIT_ULL(4) +#define KOMEDA_EVENT_EOW BIT_ULL(5) +#define KOMEDA_EVENT_MODE BIT_ULL(6) + +#define KOMEDA_ERR_TETO BIT_ULL(14) +#define KOMEDA_ERR_TEMR BIT_ULL(15) +#define KOMEDA_ERR_TITR BIT_ULL(16) +#define KOMEDA_ERR_CPE BIT_ULL(17) +#define KOMEDA_ERR_CFGE BIT_ULL(18) +#define KOMEDA_ERR_AXIE BIT_ULL(19) +#define KOMEDA_ERR_ACE0 BIT_ULL(20) +#define KOMEDA_ERR_ACE1 BIT_ULL(21) +#define KOMEDA_ERR_ACE2 BIT_ULL(22) +#define KOMEDA_ERR_ACE3 BIT_ULL(23) +#define KOMEDA_ERR_DRIFTTO BIT_ULL(24) +#define KOMEDA_ERR_FRAMETO BIT_ULL(25) +#define KOMEDA_ERR_CSCE BIT_ULL(26) +#define KOMEDA_ERR_ZME BIT_ULL(27) +#define KOMEDA_ERR_MERR BIT_ULL(28) +#define KOMEDA_ERR_TCF BIT_ULL(29) +#define KOMEDA_ERR_TTNG BIT_ULL(30) +#define KOMEDA_ERR_TTF BIT_ULL(31) + /* malidp device id */ enum { MALI_D71 = 0, @@ -39,6 +66,11 @@ struct komeda_product_data { struct komeda_dev; +struct komeda_events { + u64 global; + u64 pipes[KOMEDA_MAX_PIPELINES]; +}; + /** * struct komeda_dev_funcs * @@ -60,6 +92,20 @@ struct komeda_dev_funcs { int (*enum_resources)(struct komeda_dev *mdev); /** @cleanup: call to chip to cleanup komeda_dev->chip data */ void (*cleanup)(struct komeda_dev *mdev); + /** + * @irq_handler: + * + * for CORE to get the HW event from the CHIP when interrupt happened. + */ + irqreturn_t (*irq_handler)(struct komeda_dev *mdev, + struct komeda_events *events); + /** @enable_irq: enable irq */ + int (*enable_irq)(struct komeda_dev *mdev); + /** @disable_irq: disable irq */ + int (*disable_irq)(struct komeda_dev *mdev); + + /** @dump_register: Optional, dump registers to seq_file */ + void (*dump_register)(struct komeda_dev *mdev, struct seq_file *seq); }; /** @@ -81,6 +127,9 @@ struct komeda_dev { /** @mck: HW main engine clk */ struct clk *mclk; + /** @irq: irq number */ + int irq; + int n_pipelines; struct komeda_pipeline *pipelines[KOMEDA_MAX_PIPELINES]; @@ -93,6 +142,8 @@ struct komeda_dev { * destroyed by &komeda_dev_funcs.cleanup() */ void *chip_data; + + struct dentry *debugfs_root; }; static inline bool diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_kms.c b/drivers/gpu/drm/arm/display/komeda/komeda_kms.c index 47a58ab20434..b214edbfbbc6 100644 --- a/drivers/gpu/drm/arm/display/komeda/komeda_kms.c +++ b/drivers/gpu/drm/arm/display/komeda/komeda_kms.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include "komeda_dev.h" @@ -33,10 +34,31 @@ static int komeda_gem_cma_dumb_create(struct drm_file *file, return drm_gem_cma_dumb_create_internal(file, dev, args); } +static irqreturn_t komeda_kms_irq_handler(int irq, void *data) +{ + struct drm_device *drm = data; + struct komeda_dev *mdev = drm->dev_private; + struct komeda_kms_dev *kms = to_kdev(drm); + struct komeda_events evts; + irqreturn_t status; + u32 i; + + /* Call into the CHIP to recognize events */ + memset(&evts, 0, sizeof(evts)); + status = mdev->funcs->irq_handler(mdev, &evts); + + /* Notify the crtc to handle the events */ + for (i = 0; i < kms->n_crtcs; i++) + komeda_crtc_handle_event(&kms->crtcs[i], &evts); + + return status; +} + static struct drm_driver komeda_kms_driver = { .driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC | - DRIVER_PRIME, + DRIVER_PRIME | DRIVER_HAVE_IRQ, .lastclose = drm_fb_helper_lastclose, + .irq_handler = komeda_kms_irq_handler, .gem_free_object_unlocked = drm_gem_cma_free_object, .gem_vm_ops = &drm_gem_cma_vm_ops, .dumb_create = komeda_gem_cma_dumb_create, @@ -144,12 +166,22 @@ struct komeda_kms_dev *komeda_kms_attach(struct komeda_dev *mdev) drm_mode_config_reset(drm); - err = drm_dev_register(drm, 0); + err = drm_irq_install(drm, mdev->irq); if (err) goto cleanup_mode_config; + err = mdev->funcs->enable_irq(mdev); + if (err) + goto uninstall_irq; + + err = drm_dev_register(drm, 0); + if (err) + goto uninstall_irq; + return kms; +uninstall_irq: + drm_irq_uninstall(drm); cleanup_mode_config: drm_mode_config_cleanup(drm); free_kms: @@ -162,7 +194,9 @@ void komeda_kms_detach(struct komeda_kms_dev *kms) struct drm_device *drm = &kms->base; struct komeda_dev *mdev = drm->dev_private; + mdev->funcs->disable_irq(mdev); drm_dev_unregister(drm); + drm_irq_uninstall(drm); component_unbind_all(mdev->dev, drm); komeda_kms_cleanup_private_objs(mdev); drm_mode_config_cleanup(drm); diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_kms.h b/drivers/gpu/drm/arm/display/komeda/komeda_kms.h index 874e9c9f0749..15ac8b85506c 100644 --- a/drivers/gpu/drm/arm/display/komeda/komeda_kms.h +++ b/drivers/gpu/drm/arm/display/komeda/komeda_kms.h @@ -12,6 +12,8 @@ #include #include #include +#include