From fd86344823b521149bb31d91eba900ba3525efa6 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Fri, 12 Apr 2024 19:19:28 -0700 Subject: [PATCH 01/14] af_unix: Try not to hold unix_gc_lock during accept(). Commit dcf70df2048d ("af_unix: Fix up unix_edge.successor for embryo socket.") added spin_lock(&unix_gc_lock) in accept() path, and it caused regression in a stress test as reported by kernel test robot. If the embryo socket is not part of the inflight graph, we need not hold the lock. To decide that in O(1) time and avoid the regression in the normal use case, 1. add a new stat unix_sk(sk)->scm_stat.nr_unix_fds 2. count the number of inflight AF_UNIX sockets in the receive queue under unix_state_lock() 3. move unix_update_edges() call under unix_state_lock() 4. avoid locking if nr_unix_fds is 0 in unix_update_edges() Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-lkp/202404101427.92a08551-oliver.sang@intel.com Signed-off-by: Kuniyuki Iwashima Link: https://lore.kernel.org/r/20240413021928.20946-1-kuniyu@amazon.com Signed-off-by: Paolo Abeni --- include/net/af_unix.h | 1 + net/unix/af_unix.c | 2 +- net/unix/garbage.c | 20 ++++++++++++++++---- 3 files changed, 18 insertions(+), 5 deletions(-) diff --git a/include/net/af_unix.h b/include/net/af_unix.h index 7311b77edfc7..872ff2a50372 100644 --- a/include/net/af_unix.h +++ b/include/net/af_unix.h @@ -67,6 +67,7 @@ struct unix_skb_parms { struct scm_stat { atomic_t nr_fds; + unsigned long nr_unix_fds; }; #define UNIXCB(skb) (*(struct unix_skb_parms *)&((skb)->cb)) diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index 142d210b5b03..ed16d5f66df8 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -1719,12 +1719,12 @@ static int unix_accept(struct socket *sock, struct socket *newsock, int flags, } tsk = skb->sk; - unix_update_edges(unix_sk(tsk)); skb_free_datagram(sk, skb); wake_up_interruptible(&unix_sk(sk)->peer_wait); /* attach accepted sock to socket */ unix_state_lock(tsk); + unix_update_edges(unix_sk(tsk)); newsock->state = SS_CONNECTED; unix_sock_inherit_flags(sock, newsock); sock_graft(tsk, newsock); diff --git a/net/unix/garbage.c b/net/unix/garbage.c index 12a4ec27e0d4..95240a59808f 100644 --- a/net/unix/garbage.c +++ b/net/unix/garbage.c @@ -209,6 +209,7 @@ void unix_add_edges(struct scm_fp_list *fpl, struct unix_sock *receiver) unix_add_edge(fpl, edge); } while (i < fpl->count_unix); + receiver->scm_stat.nr_unix_fds += fpl->count_unix; WRITE_ONCE(unix_tot_inflight, unix_tot_inflight + fpl->count_unix); out: WRITE_ONCE(fpl->user->unix_inflight, fpl->user->unix_inflight + fpl->count); @@ -222,6 +223,7 @@ out: void unix_del_edges(struct scm_fp_list *fpl) { + struct unix_sock *receiver; int i = 0; spin_lock(&unix_gc_lock); @@ -235,6 +237,8 @@ void unix_del_edges(struct scm_fp_list *fpl) unix_del_edge(fpl, edge); } while (i < fpl->count_unix); + receiver = fpl->edges[0].successor; + receiver->scm_stat.nr_unix_fds -= fpl->count_unix; WRITE_ONCE(unix_tot_inflight, unix_tot_inflight - fpl->count_unix); out: WRITE_ONCE(fpl->user->unix_inflight, fpl->user->unix_inflight - fpl->count); @@ -246,10 +250,18 @@ out: void unix_update_edges(struct unix_sock *receiver) { - spin_lock(&unix_gc_lock); - unix_update_graph(unix_sk(receiver->listener)->vertex); - receiver->listener = NULL; - spin_unlock(&unix_gc_lock); + /* nr_unix_fds is only updated under unix_state_lock(). + * If it's 0 here, the embryo socket is not part of the + * inflight graph, and GC will not see it, so no lock needed. + */ + if (!receiver->scm_stat.nr_unix_fds) { + receiver->listener = NULL; + } else { + spin_lock(&unix_gc_lock); + unix_update_graph(unix_sk(receiver->listener)->vertex); + receiver->listener = NULL; + spin_unlock(&unix_gc_lock); + } } int unix_prepare_fpl(struct scm_fp_list *fpl) From 9b07508845f9fff9519efabb18ad1d5ffdc214a5 Mon Sep 17 00:00:00 2001 From: Krzysztof Kozlowski Date: Sun, 14 Apr 2024 17:49:29 +0200 Subject: [PATCH 02/14] net: dsa: microchip: drop unneeded MODULE_ALIAS The ID table already has respective entry and MODULE_DEVICE_TABLE and creates proper alias for SPI driver. Having another MODULE_ALIAS causes the alias to be duplicated. Signed-off-by: Krzysztof Kozlowski Link: https://lore.kernel.org/r/20240414154929.127045-1-krzk@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/dsa/microchip/ksz_spi.c | 7 ------- 1 file changed, 7 deletions(-) diff --git a/drivers/net/dsa/microchip/ksz_spi.c b/drivers/net/dsa/microchip/ksz_spi.c index 79ae01507a8a..8e8d83213b04 100644 --- a/drivers/net/dsa/microchip/ksz_spi.c +++ b/drivers/net/dsa/microchip/ksz_spi.c @@ -232,13 +232,6 @@ static struct spi_driver ksz_spi_driver = { module_spi_driver(ksz_spi_driver); -MODULE_ALIAS("spi:ksz9477"); -MODULE_ALIAS("spi:ksz9896"); -MODULE_ALIAS("spi:ksz9897"); -MODULE_ALIAS("spi:ksz9893"); -MODULE_ALIAS("spi:ksz9563"); -MODULE_ALIAS("spi:ksz8563"); -MODULE_ALIAS("spi:ksz9567"); MODULE_ALIAS("spi:lan937x"); MODULE_AUTHOR("Tristram Ha "); MODULE_DESCRIPTION("Microchip ksz Series Switch SPI Driver"); From c2b640529ec7595b4f82f1630da1ac9697a0c927 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Mon, 15 Apr 2024 11:07:13 +0100 Subject: [PATCH 03/14] net/handshake: remove redundant assignment to variable ret The variable is being assigned an value and then is being re-assigned a new value in the next statement. The assignment is redundant and can be removed. Cleans up clang scan build warning: net/handshake/tlshd.c:216:2: warning: Value stored to 'ret' is never read [deadcode.DeadStores] Signed-off-by: Colin Ian King Reviewed-by: Jason Xing Acked-by: Chuck Lever Link: https://lore.kernel.org/r/20240415100713.483399-1-colin.i.king@gmail.com Signed-off-by: Jakub Kicinski --- net/handshake/tlshd.c | 1 - 1 file changed, 1 deletion(-) diff --git a/net/handshake/tlshd.c b/net/handshake/tlshd.c index d697f68c598c..d6f52839827e 100644 --- a/net/handshake/tlshd.c +++ b/net/handshake/tlshd.c @@ -213,7 +213,6 @@ static int tls_handshake_accept(struct handshake_req *req, if (!hdr) goto out_cancel; - ret = -EMSGSIZE; ret = nla_put_s32(msg, HANDSHAKE_A_ACCEPT_SOCKFD, fd); if (ret < 0) goto out_cancel; From 19ad0f3a16d2c48da995ad481366b9ecb5b5aeb5 Mon Sep 17 00:00:00 2001 From: Jiapeng Chong Date: Mon, 15 Apr 2024 11:14:56 +0800 Subject: [PATCH 04/14] net: ipa: Remove unnecessary print function dev_err() The print function dev_err() is redundant because platform_get_irq_byname() already prints an error. ./drivers/net/ipa/ipa_interrupt.c:300:2-9: line 300 is redundant because platform_get_irq() already prints an error. Reported-by: Abaci Robot Closes: https://bugzilla.openanolis.cn/show_bug.cgi?id=8756 Signed-off-by: Jiapeng Chong Reviewed-by: Simon Horman Link: https://lore.kernel.org/r/20240415031456.10805-1-jiapeng.chong@linux.alibaba.com Signed-off-by: Jakub Kicinski --- drivers/net/ipa/ipa_interrupt.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/drivers/net/ipa/ipa_interrupt.c b/drivers/net/ipa/ipa_interrupt.c index c3e8784d51d9..c44ec05f71e6 100644 --- a/drivers/net/ipa/ipa_interrupt.c +++ b/drivers/net/ipa/ipa_interrupt.c @@ -291,16 +291,12 @@ void ipa_interrupt_deconfig(struct ipa *ipa) /* Initialize the IPA interrupt structure */ struct ipa_interrupt *ipa_interrupt_init(struct platform_device *pdev) { - struct device *dev = &pdev->dev; struct ipa_interrupt *interrupt; int irq; irq = platform_get_irq_byname(pdev, "ipa"); - if (irq <= 0) { - dev_err(dev, "DT error %d getting \"ipa\" IRQ property\n", irq); - + if (irq <= 0) return ERR_PTR(irq ? : -EINVAL); - } interrupt = kzalloc(sizeof(*interrupt), GFP_KERNEL); if (!interrupt) From 427f6acbc88154f40f68391dcfea36b821fcd06a Mon Sep 17 00:00:00 2001 From: Peng Fan Date: Mon, 15 Apr 2024 18:36:21 +0800 Subject: [PATCH 05/14] dt-bindings: net: nxp,dwmac-imx: allow nvmem cells property Allow nvmem-cells and nvmem-cell-names to get mac_address from onchip fuse. Signed-off-by: Peng Fan Reviewed-by: Frank Li Acked-by: Conor Dooley Link: https://lore.kernel.org/r/20240415103621.1644735-1-peng.fan@oss.nxp.com Signed-off-by: Jakub Kicinski --- Documentation/devicetree/bindings/net/nxp,dwmac-imx.yaml | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/Documentation/devicetree/bindings/net/nxp,dwmac-imx.yaml b/Documentation/devicetree/bindings/net/nxp,dwmac-imx.yaml index 4c01cae7c93a..87bc4416eadf 100644 --- a/Documentation/devicetree/bindings/net/nxp,dwmac-imx.yaml +++ b/Documentation/devicetree/bindings/net/nxp,dwmac-imx.yaml @@ -66,6 +66,10 @@ properties: Should be phandle/offset pair. The phandle to the syscon node which encompases the GPR register, and the offset of the GPR register. + nvmem-cells: true + + nvmem-cell-names: true + snps,rmii_refclk_ext: $ref: /schemas/types.yaml#/definitions/flag description: From 86600ea11dc18ecd10110c13148d0fb04a80ceea Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Mon, 15 Apr 2024 14:23:44 +0200 Subject: [PATCH 06/14] ip6_vti: fix memleak on netns dismantle kmemleak reports net_device resources are no longer released, restore needs_free_netdev toggle. Sample backtrace: unreferenced object 0xffff88810874f000 (size 4096): [..] [<00000000a2b8af8b>] __kmalloc_node+0x209/0x290 [<0000000040b0a1a9>] alloc_netdev_mqs+0x58/0x470 [<00000000b4be1e78>] vti6_init_net+0x94/0x230 [<000000008830c1ea>] ops_init+0x32/0xc0 [<000000006a26fa8f>] setup_net+0x134/0x2e0 [..] Cc: Breno Leitao Fixes: a9b2d55a8f1e ("ip6_vti: Do not use custom stat allocator") Signed-off-by: Florian Westphal Reviewed-by: Simon Horman Link: https://lore.kernel.org/r/20240415122346.26503-1-fw@strlen.de Signed-off-by: Jakub Kicinski --- net/ipv6/ip6_vti.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv6/ip6_vti.c b/net/ipv6/ip6_vti.c index 4d68a0777b0c..78344cf3867e 100644 --- a/net/ipv6/ip6_vti.c +++ b/net/ipv6/ip6_vti.c @@ -901,6 +901,7 @@ static void vti6_dev_setup(struct net_device *dev) { dev->netdev_ops = &vti6_netdev_ops; dev->header_ops = &ip_tunnel_header_ops; + dev->needs_free_netdev = true; dev->pcpu_stat_type = NETDEV_PCPU_STAT_TSTATS; dev->type = ARPHRD_TUNNEL6; From 232d79aaa7816aa09766962616c768d349947f17 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Mon, 15 Apr 2024 17:45:51 -0700 Subject: [PATCH 07/14] selftests: drv-net: add stdout to the command failed exception ping prints all the info to stdout. To make debug easier capture stdout in the Exception raised when command unexpectedly fails. Reviewed-by: Willem de Bruijn Link: https://lore.kernel.org/r/20240416004556.1618804-2-kuba@kernel.org Signed-off-by: Jakub Kicinski --- tools/testing/selftests/net/lib/py/utils.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/net/lib/py/utils.py b/tools/testing/selftests/net/lib/py/utils.py index f0d425731fd4..19612348c30d 100644 --- a/tools/testing/selftests/net/lib/py/utils.py +++ b/tools/testing/selftests/net/lib/py/utils.py @@ -33,7 +33,8 @@ class cmd: if self.proc.returncode != 0 and fail: if len(stderr) > 0 and stderr[-1] == "\n": stderr = stderr[:-1] - raise Exception("Command failed: %s\n%s" % (self.proc.args, stderr)) + raise Exception("Command failed: %s\nSTDOUT: %s\nSTDERR: %s" % + (self.proc.args, stdout, stderr)) def ip(args, json=None, ns=None): From 438ce84bae90b197314e781a20d8a043187b6e59 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Mon, 15 Apr 2024 17:45:52 -0700 Subject: [PATCH 08/14] selftests: drv-net: add config for netdevsim Real driver testing will obviously require enabling more options, but will require more manual setup in the first place. For CIs running purely software tests we need to enable netdevsim. Reviewed-by: Willem de Bruijn Link: https://lore.kernel.org/r/20240416004556.1618804-3-kuba@kernel.org Signed-off-by: Jakub Kicinski --- tools/testing/selftests/drivers/net/config | 2 ++ 1 file changed, 2 insertions(+) create mode 100644 tools/testing/selftests/drivers/net/config diff --git a/tools/testing/selftests/drivers/net/config b/tools/testing/selftests/drivers/net/config new file mode 100644 index 000000000000..f6a58ce8a230 --- /dev/null +++ b/tools/testing/selftests/drivers/net/config @@ -0,0 +1,2 @@ +CONFIG_IPV6=y +CONFIG_NETDEVSIM=m From 57f73805b99d3b454be87d405cc473b8feb84a1a Mon Sep 17 00:00:00 2001 From: "Kory Maincent (Dent Project)" Date: Sun, 14 Apr 2024 16:21:50 +0200 Subject: [PATCH 09/14] MAINTAINERS: net: Add Oleksij to pse-pd maintainers Oleksij was the first to add support for pse-pd net subsystem. Add himself to the maintainers seems logical. Acked-by: Oleksij Rempel Signed-off-by: Kory Maincent Reviewed-by: Andrew Lunn Link: https://lore.kernel.org/r/20240414-feature_poe-v8-1-e4bf1e860da5@bootlin.com Signed-off-by: Jakub Kicinski --- MAINTAINERS | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 5ba3fe6ac09c..f22698a7859f 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17780,6 +17780,13 @@ F: include/net/psample.h F: include/uapi/linux/psample.h F: net/psample +PSE NETWORK DRIVER +M: Oleksij Rempel +L: netdev@vger.kernel.org +S: Maintained +F: Documentation/devicetree/bindings/net/pse-pd/ +F: drivers/net/pse-pd/ + PSTORE FILESYSTEM M: Kees Cook R: Tony Luck From 93c0d8c0ac309cc927c0b5e9a52af4c954d66da5 Mon Sep 17 00:00:00 2001 From: "Kory Maincent (Dent Project)" Date: Sun, 14 Apr 2024 16:21:51 +0200 Subject: [PATCH 10/14] of: property: Add fw_devlink support for pse parent This allows fw_devlink to create device links between consumers of a PSE and the supplier of the PSE. Reviewed-by: Andrew Lunn Signed-off-by: Kory Maincent Link: https://lore.kernel.org/r/20240414-feature_poe-v8-2-e4bf1e860da5@bootlin.com Signed-off-by: Jakub Kicinski --- drivers/of/property.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/of/property.c b/drivers/of/property.c index a6358ee99b74..cc064151853e 100644 --- a/drivers/of/property.c +++ b/drivers/of/property.c @@ -1252,6 +1252,7 @@ DEFINE_SIMPLE_PROP(backlight, "backlight", NULL) DEFINE_SIMPLE_PROP(panel, "panel", NULL) DEFINE_SIMPLE_PROP(msi_parent, "msi-parent", "#msi-cells") DEFINE_SIMPLE_PROP(post_init_providers, "post-init-providers", NULL) +DEFINE_SIMPLE_PROP(pses, "pses", "#pse-cells") DEFINE_SUFFIX_PROP(regulators, "-supply", NULL) DEFINE_SUFFIX_PROP(gpio, "-gpio", "#gpio-cells") @@ -1357,6 +1358,7 @@ static const struct supplier_bindings of_supplier_bindings[] = { { .parse_prop = parse_backlight, }, { .parse_prop = parse_panel, }, { .parse_prop = parse_msi_parent, }, + { .parse_prop = parse_pses, }, { .parse_prop = parse_gpio_compat, }, { .parse_prop = parse_interrupts, }, { .parse_prop = parse_regulators, }, From b010bf72da5ada39f4e722fa28d4275df8673229 Mon Sep 17 00:00:00 2001 From: "Kory Maincent (Dent Project)" Date: Sun, 14 Apr 2024 16:21:52 +0200 Subject: [PATCH 11/14] net: pse-pd: Rectify and adapt the naming of admin_cotrol member of struct pse_control_config In commit 18ff0bcda6d1 ("ethtool: add interface to interact with Ethernet Power Equipment"), the 'pse_control_config' structure was introduced, housing a single member labeled 'admin_cotrol' responsible for maintaining the operational state of the PoDL PSE functions. A noticeable typographical error exists in the naming of this field ('cotrol' should be corrected to 'control'), which this commit aims to rectify. Furthermore, with upcoming extensions of this structure to encompass PoE functionalities, the field is being renamed to 'podl_admin_state' to distinctly indicate that this state is tailored specifically for PoDL." Reviewed-by: Oleksij Rempel Reviewed-by: Andrew Lunn Signed-off-by: Kory Maincent Link: https://lore.kernel.org/r/20240414-feature_poe-v8-3-e4bf1e860da5@bootlin.com Signed-off-by: Jakub Kicinski --- drivers/net/pse-pd/pse_regulator.c | 8 ++++---- include/linux/pse-pd/pse.h | 4 ++-- net/ethtool/pse-pd.c | 2 +- 3 files changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/net/pse-pd/pse_regulator.c b/drivers/net/pse-pd/pse_regulator.c index e2bf8306ca90..1dedf4de296e 100644 --- a/drivers/net/pse-pd/pse_regulator.c +++ b/drivers/net/pse-pd/pse_regulator.c @@ -31,10 +31,10 @@ pse_reg_ethtool_set_config(struct pse_controller_dev *pcdev, unsigned long id, struct pse_reg_priv *priv = to_pse_reg(pcdev); int ret; - if (priv->admin_state == config->admin_cotrol) + if (priv->admin_state == config->podl_admin_control) return 0; - switch (config->admin_cotrol) { + switch (config->podl_admin_control) { case ETHTOOL_PODL_PSE_ADMIN_STATE_ENABLED: ret = regulator_enable(priv->ps); break; @@ -43,14 +43,14 @@ pse_reg_ethtool_set_config(struct pse_controller_dev *pcdev, unsigned long id, break; default: dev_err(pcdev->dev, "Unknown admin state %i\n", - config->admin_cotrol); + config->podl_admin_control); ret = -ENOTSUPP; } if (ret) return ret; - priv->admin_state = config->admin_cotrol; + priv->admin_state = config->podl_admin_control; return 0; } diff --git a/include/linux/pse-pd/pse.h b/include/linux/pse-pd/pse.h index fb724c65c77b..199cf4ae3cf2 100644 --- a/include/linux/pse-pd/pse.h +++ b/include/linux/pse-pd/pse.h @@ -15,11 +15,11 @@ struct pse_controller_dev; /** * struct pse_control_config - PSE control/channel configuration. * - * @admin_cotrol: set PoDL PSE admin control as described in + * @podl_admin_control: set PoDL PSE admin control as described in * IEEE 802.3-2018 30.15.1.2.1 acPoDLPSEAdminControl */ struct pse_control_config { - enum ethtool_podl_pse_admin_state admin_cotrol; + enum ethtool_podl_pse_admin_state podl_admin_control; }; /** diff --git a/net/ethtool/pse-pd.c b/net/ethtool/pse-pd.c index cc478af77111..aef57a058f0d 100644 --- a/net/ethtool/pse-pd.c +++ b/net/ethtool/pse-pd.c @@ -130,7 +130,7 @@ ethnl_set_pse(struct ethnl_req_info *req_info, struct genl_info *info) struct phy_device *phydev; /* this values are already validated by the ethnl_pse_set_policy */ - config.admin_cotrol = nla_get_u32(tb[ETHTOOL_A_PODL_PSE_ADMIN_CONTROL]); + config.podl_admin_control = nla_get_u32(tb[ETHTOOL_A_PODL_PSE_ADMIN_CONTROL]); phydev = dev->phydev; if (!phydev) { From af3b4b0e59decfc4db2eafbb4e0d0658529019f6 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ar=C4=B1n=C3=A7=20=C3=9CNAL?= Date: Sun, 14 Apr 2024 00:08:13 +0300 Subject: [PATCH 12/14] net: phy: mediatek-ge: do not disable EEE advertisement MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The mediatek-ge PHY driver already disables EEE advertisement on the switch PHYs but my testing [1] shows that it is somehow enabled afterwards. Disabling EEE advertisement before the PHY driver initialises keeps it off. Therefore, remove disabling EEE advertisement here as it's useless. Link: https://lore.kernel.org/netdev/d286ea27-e911-4dcb-9037-b75f22b437b8@arinc9.com/ [1] Signed-off-by: Arınç ÜNAL Signed-off-by: David S. Miller --- drivers/net/phy/mediatek-ge.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/net/phy/mediatek-ge.c b/drivers/net/phy/mediatek-ge.c index a493ae01b267..54ea64a37ab3 100644 --- a/drivers/net/phy/mediatek-ge.c +++ b/drivers/net/phy/mediatek-ge.c @@ -23,9 +23,6 @@ static int mtk_gephy_write_page(struct phy_device *phydev, int page) static void mtk_gephy_config_init(struct phy_device *phydev) { - /* Disable EEE */ - phy_write_mmd(phydev, MDIO_MMD_AN, MDIO_AN_EEE_ADV, 0); - /* Enable HW auto downshift */ phy_modify_paged(phydev, MTK_PHY_PAGE_EXTENDED, 0x14, 0, BIT(4)); From 9ef1ed26a67b817fd08faf851468ef040db7d280 Mon Sep 17 00:00:00 2001 From: Yujie Liu Date: Tue, 16 Apr 2024 13:32:10 +0800 Subject: [PATCH 13/14] selftests: fix netfilter path in Makefile Netfilter tests have been moved to a subdir under selftests/net by patch series [1]. Fix the path in selftests/Makefile accordingly. This helps fix the following error: tools/testing/selftests$ make ... make[1]: Entering directory 'tools/testing/selftests' make[1]: *** netfilter: No such file or directory. Stop. make[1]: Leaving directory 'tools/testing/selftests' [1] https://lore.kernel.org/all/20240411233624.8129-1-fw@strlen.de/ Reported-by: kernel test robot Signed-off-by: Yujie Liu Signed-off-by: David S. Miller --- tools/testing/selftests/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/Makefile b/tools/testing/selftests/Makefile index 6dab886d6f7a..c785b6256a45 100644 --- a/tools/testing/selftests/Makefile +++ b/tools/testing/selftests/Makefile @@ -64,7 +64,7 @@ TARGETS += net/hsr TARGETS += net/mptcp TARGETS += net/openvswitch TARGETS += net/tcp_ao -TARGETS += netfilter +TARGETS += net/netfilter TARGETS += nsfs TARGETS += perf_events TARGETS += pidfd From 2bd99aef1b19e6da09eff692bc0a09d61d785782 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 16 Apr 2024 09:50:54 +0000 Subject: [PATCH 14/14] tcp: accept bare FIN packets under memory pressure Andrew Oates reported that some macOS hosts could repeatedly send FIN packets even if the remote peer drops them and send back DUP ACK RWIN 0 packets. 20:27:16.968254 gif0 In IP macos > victim: Flags [SEW], seq 1950399762, win 65535, options [mss 1460,nop,wscale 6,nop,nop,TS val 501897188 ecr 0,sackOK,eol], length 0 20:27:16.968339 gif0 Out IP victim > macos: Flags [S.E], seq 2995489058, ack 1950399763, win 1448, options [mss 1460,sackOK,TS val 3829877593 ecr 501897188,nop,wscale 0], length 0 20:27:16.968833 gif0 In IP macos > victim: Flags [.], ack 1, win 2058, options [nop,nop,TS val 501897188 ecr 3829877593], length 0 20:27:16.968885 gif0 In IP macos > victim: Flags [P.], seq 1:1449, ack 1, win 2058, options [nop,nop,TS val 501897188 ecr 3829877593], length 1448 20:27:16.968896 gif0 Out IP victim > macos: Flags [.], ack 1449, win 0, options [nop,nop,TS val 3829877593 ecr 501897188], length 0 20:27:19.454593 gif0 In IP macos > victim: Flags [F.], seq 1449, ack 1, win 2058, options [nop,nop,TS val 501899674 ecr 3829877593], length 0 20:27:19.454675 gif0 Out IP victim > macos: Flags [.], ack 1449, win 0, options [nop,nop,TS val 3829880079 ecr 501899674], length 0 20:27:19.455116 gif0 In IP macos > victim: Flags [F.], seq 1449, ack 1, win 2058, options [nop,nop,TS val 501899674 ecr 3829880079], length 0 The retransmits/dup-ACKs then repeat in a tight loop. RFC 9293 3.4. Sequence Numbers states : Note that when the receive window is zero no segments should be acceptable except ACK segments. Thus, it is be possible for a TCP to maintain a zero receive window while transmitting data and receiving ACKs. However, even when the receive window is zero, a TCP must process the RST and URG fields of all incoming segments. Even if we could consider a bare FIN.ACK packet to be an ACK in RFC terms, the retransmits should use exponential backoff. Accepting the FIN in linux does not add extra memory costs, because the FIN flag will simply be merged to the tail skb in the receive queue, and incoming packet is freed. Reported-by: Andrew Oates Signed-off-by: Eric Dumazet Cc: Neal Cardwell Cc: Christoph Paasch Cc: Vidhi Goel Acked-by: Neal Cardwell Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 5a45a0923a1f..384fa5e2f065 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -5174,6 +5174,16 @@ static void tcp_data_queue(struct sock *sk, struct sk_buff *skb) */ if (TCP_SKB_CB(skb)->seq == tp->rcv_nxt) { if (tcp_receive_window(tp) == 0) { + /* Some stacks are known to send bare FIN packets + * in a loop even if we send RWIN 0 in our ACK. + * Accepting this FIN does not hurt memory pressure + * because the FIN flag will simply be merged to the + * receive queue tail skb in most cases. + */ + if (!skb->len && + (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)) + goto queue_and_out; + reason = SKB_DROP_REASON_TCP_ZEROWINDOW; NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPZEROWINDOWDROP); goto out_of_window; @@ -5188,7 +5198,7 @@ queue_and_out: inet_csk_schedule_ack(sk); sk->sk_data_ready(sk); - if (skb_queue_len(&sk->sk_receive_queue)) { + if (skb_queue_len(&sk->sk_receive_queue) && skb->len) { reason = SKB_DROP_REASON_PROTO_MEM; NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPRCVQDROP); goto drop;