From 65f35aa76c0e21b0243fd734e513fd2263f22a18 Mon Sep 17 00:00:00 2001 From: Rahul Rameshbabu Date: Tue, 9 Apr 2024 16:25:16 -0700 Subject: [PATCH 01/16] ethtool: update tsinfo statistics attribute docs with correct type nla_put_uint can either write a u32 or u64 netlink attribute value. The size depends on whether the value can be represented with a u32 or requires a u64. Use a uint annotation in various documentation to represent this. Signed-off-by: Rahul Rameshbabu Link: https://lore.kernel.org/r/20240409232520.237613-2-rrameshbabu@nvidia.com Signed-off-by: Jakub Kicinski --- Documentation/networking/ethtool-netlink.rst | 6 +++--- include/uapi/linux/ethtool_netlink.h | 6 +++--- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/Documentation/networking/ethtool-netlink.rst b/Documentation/networking/ethtool-netlink.rst index 5dc42f7ce429..4e63d3708ed9 100644 --- a/Documentation/networking/ethtool-netlink.rst +++ b/Documentation/networking/ethtool-netlink.rst @@ -1254,9 +1254,9 @@ would be empty (no bit set). Additional hardware timestamping statistics response contents: ===================================== ====== =================================== - ``ETHTOOL_A_TS_STAT_TX_PKTS`` u64 Packets with Tx HW timestamps - ``ETHTOOL_A_TS_STAT_TX_LOST`` u64 Tx HW timestamp not arrived count - ``ETHTOOL_A_TS_STAT_TX_ERR`` u64 HW error request Tx timestamp count + ``ETHTOOL_A_TS_STAT_TX_PKTS`` uint Packets with Tx HW timestamps + ``ETHTOOL_A_TS_STAT_TX_LOST`` uint Tx HW timestamp not arrived count + ``ETHTOOL_A_TS_STAT_TX_ERR`` uint HW error request Tx timestamp count ===================================== ====== =================================== CABLE_TEST diff --git a/include/uapi/linux/ethtool_netlink.h b/include/uapi/linux/ethtool_netlink.h index 23e225f00fb0..b4f0d233d048 100644 --- a/include/uapi/linux/ethtool_netlink.h +++ b/include/uapi/linux/ethtool_netlink.h @@ -488,9 +488,9 @@ enum { enum { ETHTOOL_A_TS_STAT_UNSPEC, - ETHTOOL_A_TS_STAT_TX_PKTS, /* u64 */ - ETHTOOL_A_TS_STAT_TX_LOST, /* u64 */ - ETHTOOL_A_TS_STAT_TX_ERR, /* u64 */ + ETHTOOL_A_TS_STAT_TX_PKTS, /* uint */ + ETHTOOL_A_TS_STAT_TX_LOST, /* uint */ + ETHTOOL_A_TS_STAT_TX_ERR, /* uint */ /* add new constants above here */ __ETHTOOL_A_TS_STAT_CNT, From 4ede457542a615b08b1f25a25cb514402d033968 Mon Sep 17 00:00:00 2001 From: Hangbin Liu Date: Tue, 9 Apr 2024 16:35:04 +0800 Subject: [PATCH 02/16] doc/netlink/specs: Add bond support to rt_link.yaml Add bond support to rt_link.yaml. Here is an example output: $ ./tools/net/ynl/cli.py --spec Documentation/netlink/specs/rt_link.yaml \ --do getlink --json '{"ifname": "bond0"}' --output-json | jq '.linkinfo' { "kind": "bond", "data": { "mode": 4, "miimon": 100, ... "arp-interval": 0, "arp-ip-target": [ "192.168.1.1", "192.168.1.2" ], "arp-validate": 0, "arp-all-targets": 0, "ns-ip6-target": [ "2001::1", "2001::2" ], "primary-reselect": 0, ... "missed-max": 2, "ad-info": { "aggregator": 1, "num-ports": 1, "actor-key": 0, "partner-key": 1, "partner-mac": "00:00:00:00:00:00" } } } And here is the downlink info. $ ./tools/net/ynl/cli.py --spec Documentation/netlink/specs/rt_link.yaml \ --do getlink --json '{"ifname": "dummy0"}' --output-json | jq '.linkinfo' { "kind": "dummy", "slave-kind": "bond", "slave-data": { "state": 0, "mii-status": 0, "link-failure-count": 0, "perm-hwaddr": "f2:82:f7:cc:47:13", "queue-id": 0, "prio": 0 } } Signed-off-by: Hangbin Liu Link: https://lore.kernel.org/r/20240409083504.3900877-1-liuhangbin@gmail.com Signed-off-by: Jakub Kicinski --- Documentation/netlink/specs/rt_link.yaml | 163 +++++++++++++++++++++++ 1 file changed, 163 insertions(+) diff --git a/Documentation/netlink/specs/rt_link.yaml b/Documentation/netlink/specs/rt_link.yaml index e5dcb2cf1724..113ecd17c880 100644 --- a/Documentation/netlink/specs/rt_link.yaml +++ b/Documentation/netlink/specs/rt_link.yaml @@ -1035,6 +1035,165 @@ attribute-sets: type: sub-message sub-message: linkinfo-member-data-msg selector: slave-kind + - + name: linkinfo-bond-attrs + name-prefix: ifla-bond- + attributes: + - + name: mode + type: u8 + - + name: active-slave + type: u32 + - + name: miimon + type: u32 + - + name: updelay + type: u32 + - + name: downdelay + type: u32 + - + name: use-carrier + type: u8 + - + name: arp-interval + type: u32 + - + name: arp-ip-target + type: indexed-array + sub-type: u32 + byte-order: big-endian + display-hint: ipv4 + - + name: arp-validate + type: u32 + - + name: arp-all-targets + type: u32 + - + name: primary + type: u32 + - + name: primary-reselect + type: u8 + - + name: fail-over-mac + type: u8 + - + name: xmit-hash-policy + type: u8 + - + name: resend-igmp + type: u32 + - + name: num-peer-notif + type: u8 + - + name: all-slaves-active + type: u8 + - + name: min-links + type: u32 + - + name: lp-interval + type: u32 + - + name: packets-per-slave + type: u32 + - + name: ad-lacp-rate + type: u8 + - + name: ad-select + type: u8 + - + name: ad-info + type: nest + nested-attributes: bond-ad-info-attrs + - + name: ad-actor-sys-prio + type: u16 + - + name: ad-user-port-key + type: u16 + - + name: ad-actor-system + type: binary + display-hint: mac + - + name: tlb-dynamic-lb + type: u8 + - + name: peer-notif-delay + type: u32 + - + name: ad-lacp-active + type: u8 + - + name: missed-max + type: u8 + - + name: ns-ip6-target + type: indexed-array + sub-type: binary + display-hint: ipv6 + - + name: coupled-control + type: u8 + - + name: bond-ad-info-attrs + name-prefix: ifla-bond-ad-info- + attributes: + - + name: aggregator + type: u16 + - + name: num-ports + type: u16 + - + name: actor-key + type: u16 + - + name: partner-key + type: u16 + - + name: partner-mac + type: binary + display-hint: mac + - + name: bond-slave-attrs + name-prefix: ifla-bond-slave- + attributes: + - + name: state + type: u8 + - + name: mii-status + type: u8 + - + name: link-failure-count + type: u32 + - + name: perm-hwaddr + type: binary + display-hint: mac + - + name: queue-id + type: u16 + - + name: ad-aggregator-id + type: u16 + - + name: ad-actor-oper-port-state + type: u8 + - + name: ad-partner-oper-port-state + type: u16 + - + name: prio + type: u32 - name: linkinfo-bridge-attrs name-prefix: ifla-br- @@ -1716,6 +1875,9 @@ sub-messages: - name: linkinfo-data-msg formats: + - + value: bond + attribute-set: linkinfo-bond-attrs - value: bridge attribute-set: linkinfo-bridge-attrs @@ -1754,6 +1916,7 @@ sub-messages: attribute-set: linkinfo-brport-attrs - value: bond + attribute-set: bond-slave-attrs operations: enum-model: directional From 07409cf728448bf4b79a33b75f498cbff439354a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Tue, 9 Apr 2024 11:12:04 +0200 Subject: [PATCH 03/16] net: wan: fsl_qmc_hdlc: Convert to platform remove callback returning void MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The .remove() callback for a platform driver returns an int which makes many driver authors wrongly assume it's possible to do error handling by returning an error code. However the value returned is ignored (apart from emitting a warning) and this typically results in resource leaks. To improve here there is a quest to make the remove callback return void. In the first step of this quest all drivers are converted to .remove_new(), which already returns void. Eventually after all drivers are converted, .remove_new() will be renamed to .remove(). Trivially convert this driver from always returning zero in the remove callback to the void returning variant. Signed-off-by: Uwe Kleine-König Acked-by: Herve Codina Link: https://lore.kernel.org/r/20240409091203.39062-2-u.kleine-koenig@pengutronix.de Signed-off-by: Jakub Kicinski --- drivers/net/wan/fsl_qmc_hdlc.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/net/wan/fsl_qmc_hdlc.c b/drivers/net/wan/fsl_qmc_hdlc.c index f69b1f579a0c..c5e7ca793c43 100644 --- a/drivers/net/wan/fsl_qmc_hdlc.c +++ b/drivers/net/wan/fsl_qmc_hdlc.c @@ -765,15 +765,13 @@ framer_exit: return ret; } -static int qmc_hdlc_remove(struct platform_device *pdev) +static void qmc_hdlc_remove(struct platform_device *pdev) { struct qmc_hdlc *qmc_hdlc = platform_get_drvdata(pdev); unregister_hdlc_device(qmc_hdlc->netdev); free_netdev(qmc_hdlc->netdev); qmc_hdlc_framer_exit(qmc_hdlc); - - return 0; } static const struct of_device_id qmc_hdlc_id_table[] = { @@ -788,7 +786,7 @@ static struct platform_driver qmc_hdlc_driver = { .of_match_table = qmc_hdlc_id_table, }, .probe = qmc_hdlc_probe, - .remove = qmc_hdlc_remove, + .remove_new = qmc_hdlc_remove, }; module_platform_driver(qmc_hdlc_driver); From 8750539ba3178c2bb0d178a30ce57dae132cbbb8 Mon Sep 17 00:00:00 2001 From: Hangbin Liu Date: Tue, 9 Apr 2024 17:28:12 +0800 Subject: [PATCH 04/16] net: team: fix incorrect maxattr The maxattr should be the latest attr value, i.e. array size - 1, not total array size. Reported-by: syzbot+ecd7e07b4be038658c9f@syzkaller.appspotmail.com Fixes: 948dbafc15da ("net: team: use policy generated by YAML spec") Suggested-by: Jakub Kicinski Signed-off-by: Hangbin Liu Link: https://lore.kernel.org/r/20240409092812.3999785-1-liuhangbin@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/team/team_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/team/team_core.c b/drivers/net/team/team_core.c index 4e3c8d404957..8c7dbaf7c22e 100644 --- a/drivers/net/team/team_core.c +++ b/drivers/net/team/team_core.c @@ -2808,7 +2808,7 @@ static const struct genl_multicast_group team_nl_mcgrps[] = { static struct genl_family team_nl_family __ro_after_init = { .name = TEAM_GENL_NAME, .version = TEAM_GENL_VERSION, - .maxattr = ARRAY_SIZE(team_nl_policy), + .maxattr = ARRAY_SIZE(team_nl_policy) - 1, .policy = team_nl_policy, .netnsok = true, .module = THIS_MODULE, From 7e36c3372fd5a1e76e0f6028d7ef64a8ace5eb55 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Tue, 9 Apr 2024 14:08:16 +0300 Subject: [PATCH 05/16] selftests: fib_rule_tests: Add VRF tests After commit 40867d74c374 ("net: Add l3mdev index to flow struct and avoid oif reset for port devices") it is possible to configure FIB rules that match on iif / oif being a l3mdev port. It was not possible before as these parameters were reset to the ifindex of the l3mdev device itself prior to the FIB rules lookup. Add tests that cover this functionality as it does not seem to be covered by existing ones and I am aware of at least one user that needs this functionality in addition to the one mentioned in [1]. Reuse the existing FIB rules tests by simply configuring a VRF prior to the test and removing it afterwards. Differentiate the output of the non-VRF tests from the VRF tests by appending "(VRF)" to the test name if a l3mdev FIB rule is present. Verified that these tests do fail on kernel 5.15.y which does not include the previously mentioned commit: # ./fib_rule_tests.sh -t fib_rule6_vrf [...] TEST: rule6 check: oif redirect to table (VRF) [FAIL] [...] TEST: rule6 check: iif redirect to table (VRF) [FAIL] # ./fib_rule_tests.sh -t fib_rule4_vrf [...] TEST: rule4 check: oif redirect to table (VRF) [FAIL] [...] TEST: rule4 check: iif redirect to table (VRF) [FAIL] [1] https://lore.kernel.org/netdev/20200922131122.GB1601@ICIPI.localdomain/ Signed-off-by: Ido Schimmel Reviewed-by: David Ahern Link: https://lore.kernel.org/r/20240409110816.2508498-1-idosch@nvidia.com Signed-off-by: Jakub Kicinski --- tools/testing/selftests/net/fib_rule_tests.sh | 46 +++++++++++++++++-- 1 file changed, 43 insertions(+), 3 deletions(-) diff --git a/tools/testing/selftests/net/fib_rule_tests.sh b/tools/testing/selftests/net/fib_rule_tests.sh index 51157a5559b7..7c01f58a20de 100755 --- a/tools/testing/selftests/net/fib_rule_tests.sh +++ b/tools/testing/selftests/net/fib_rule_tests.sh @@ -9,6 +9,7 @@ PAUSE_ON_FAIL=${PAUSE_ON_FAIL:=no} RTABLE=100 RTABLE_PEER=101 +RTABLE_VRF=102 GW_IP4=192.51.100.2 SRC_IP=192.51.100.3 GW_IP6=2001:db8:1::2 @@ -17,7 +18,14 @@ SRC_IP6=2001:db8:1::3 DEV_ADDR=192.51.100.1 DEV_ADDR6=2001:db8:1::1 DEV=dummy0 -TESTS="fib_rule6 fib_rule4 fib_rule6_connect fib_rule4_connect" +TESTS=" + fib_rule6 + fib_rule4 + fib_rule6_connect + fib_rule4_connect + fib_rule6_vrf + fib_rule4_vrf +" SELFTEST_PATH="" @@ -27,13 +35,18 @@ log_test() local expected=$2 local msg="$3" + $IP rule show | grep -q l3mdev + if [ $? -eq 0 ]; then + msg="$msg (VRF)" + fi + if [ ${rc} -eq ${expected} ]; then nsuccess=$((nsuccess+1)) - printf "\n TEST: %-50s [ OK ]\n" "${msg}" + printf "\n TEST: %-60s [ OK ]\n" "${msg}" else ret=1 nfail=$((nfail+1)) - printf "\n TEST: %-50s [FAIL]\n" "${msg}" + printf "\n TEST: %-60s [FAIL]\n" "${msg}" if [ "${PAUSE_ON_FAIL}" = "yes" ]; then echo echo "hit enter to continue, 'q' to quit" @@ -130,6 +143,17 @@ cleanup_peer() ip netns del $peerns } +setup_vrf() +{ + $IP link add name vrf0 up type vrf table $RTABLE_VRF + $IP link set dev $DEV master vrf0 +} + +cleanup_vrf() +{ + $IP link del dev vrf0 +} + fib_check_iproute_support() { ip rule help 2>&1 | grep -q $1 @@ -248,6 +272,13 @@ fib_rule6_test() fi } +fib_rule6_vrf_test() +{ + setup_vrf + fib_rule6_test + cleanup_vrf +} + # Verify that the IPV6_TCLASS option of UDPv6 and TCPv6 sockets is properly # taken into account when connecting the socket and when sending packets. fib_rule6_connect_test() @@ -385,6 +416,13 @@ fib_rule4_test() fi } +fib_rule4_vrf_test() +{ + setup_vrf + fib_rule4_test + cleanup_vrf +} + # Verify that the IP_TOS option of UDPv4 and TCPv4 sockets is properly taken # into account when connecting the socket and when sending packets. fib_rule4_connect_test() @@ -467,6 +505,8 @@ do fib_rule4_test|fib_rule4) fib_rule4_test;; fib_rule6_connect_test|fib_rule6_connect) fib_rule6_connect_test;; fib_rule4_connect_test|fib_rule4_connect) fib_rule4_connect_test;; + fib_rule6_vrf_test|fib_rule6_vrf) fib_rule6_vrf_test;; + fib_rule4_vrf_test|fib_rule4_vrf) fib_rule4_vrf_test;; help) echo "Test names: $TESTS"; exit 0;; From 930fd7fe10d977ef880654e926b3a2c3dd52c657 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Tue, 9 Apr 2024 15:22:14 +0200 Subject: [PATCH 06/16] mlxsw: spectrum_ethtool: Add support for 100Gb/s per lane link modes The Spectrum-4 ASIC supports 100Gb/s per lane link modes, but the only one currently supported by the driver is 800Gb/s over eight lanes. Add support for 100Gb/s over one lane, 200Gb/s over two lanes and 400Gb/s over four lanes. Signed-off-by: Ido Schimmel Reviewed-by: Petr Machata Signed-off-by: Petr Machata Reviewed-by: Jacob Keller Link: https://lore.kernel.org/r/1d77830f6abcc4f0d57a7f845e5a6d97a75a434b.1712667750.git.petrm@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlxsw/reg.h | 3 + .../mellanox/mlxsw/spectrum_ethtool.c | 60 +++++++++++++++++++ 2 files changed, 63 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h index 8892654c685f..8adf86a6f5cc 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/reg.h +++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h @@ -4786,8 +4786,11 @@ MLXSW_ITEM32(reg, ptys, an_status, 0x04, 28, 4); #define MLXSW_REG_PTYS_EXT_ETH_SPEED_50GAUI_1_LAUI_1_50GBASE_CR_KR BIT(8) #define MLXSW_REG_PTYS_EXT_ETH_SPEED_CAUI_4_100GBASE_CR4_KR4 BIT(9) #define MLXSW_REG_PTYS_EXT_ETH_SPEED_100GAUI_2_100GBASE_CR2_KR2 BIT(10) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_100GAUI_1_100GBASE_CR_KR BIT(11) #define MLXSW_REG_PTYS_EXT_ETH_SPEED_200GAUI_4_200GBASE_CR4_KR4 BIT(12) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_200GAUI_2_200GBASE_CR2_KR2 BIT(13) #define MLXSW_REG_PTYS_EXT_ETH_SPEED_400GAUI_8 BIT(15) +#define MLXSW_REG_PTYS_EXT_ETH_SPEED_400GAUI_4_400GBASE_CR4_KR4 BIT(16) #define MLXSW_REG_PTYS_EXT_ETH_SPEED_800GAUI_8 BIT(19) /* reg_ptys_ext_eth_proto_cap diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c index 0f29e9c19411..a755b0a901d3 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c @@ -1648,6 +1648,18 @@ mlxsw_sp2_mask_ethtool_100gaui_2_100gbase_cr2_kr2[] = { #define MLXSW_SP2_MASK_ETHTOOL_100GAUI_2_100GBASE_CR2_KR2_LEN \ ARRAY_SIZE(mlxsw_sp2_mask_ethtool_100gaui_2_100gbase_cr2_kr2) +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_100gaui_1_100gbase_cr_kr[] = { + ETHTOOL_LINK_MODE_100000baseKR_Full_BIT, + ETHTOOL_LINK_MODE_100000baseSR_Full_BIT, + ETHTOOL_LINK_MODE_100000baseLR_ER_FR_Full_BIT, + ETHTOOL_LINK_MODE_100000baseCR_Full_BIT, + ETHTOOL_LINK_MODE_100000baseDR_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_100GAUI_1_100GBASE_CR_KR_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_100gaui_1_100gbase_cr_kr) + static const enum ethtool_link_mode_bit_indices mlxsw_sp2_mask_ethtool_200gaui_4_200gbase_cr4_kr4[] = { ETHTOOL_LINK_MODE_200000baseKR4_Full_BIT, @@ -1660,6 +1672,18 @@ mlxsw_sp2_mask_ethtool_200gaui_4_200gbase_cr4_kr4[] = { #define MLXSW_SP2_MASK_ETHTOOL_200GAUI_4_200GBASE_CR4_KR4_LEN \ ARRAY_SIZE(mlxsw_sp2_mask_ethtool_200gaui_4_200gbase_cr4_kr4) +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_200gaui_2_200gbase_cr2_kr2[] = { + ETHTOOL_LINK_MODE_200000baseKR2_Full_BIT, + ETHTOOL_LINK_MODE_200000baseSR2_Full_BIT, + ETHTOOL_LINK_MODE_200000baseLR2_ER2_FR2_Full_BIT, + ETHTOOL_LINK_MODE_200000baseDR2_Full_BIT, + ETHTOOL_LINK_MODE_200000baseCR2_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_200GAUI_2_200GBASE_CR2_KR2_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_200gaui_2_200gbase_cr2_kr2) + static const enum ethtool_link_mode_bit_indices mlxsw_sp2_mask_ethtool_400gaui_8[] = { ETHTOOL_LINK_MODE_400000baseKR8_Full_BIT, @@ -1672,6 +1696,18 @@ mlxsw_sp2_mask_ethtool_400gaui_8[] = { #define MLXSW_SP2_MASK_ETHTOOL_400GAUI_8_LEN \ ARRAY_SIZE(mlxsw_sp2_mask_ethtool_400gaui_8) +static const enum ethtool_link_mode_bit_indices +mlxsw_sp2_mask_ethtool_400gaui_4_400gbase_cr4_kr4[] = { + ETHTOOL_LINK_MODE_400000baseKR4_Full_BIT, + ETHTOOL_LINK_MODE_400000baseSR4_Full_BIT, + ETHTOOL_LINK_MODE_400000baseLR4_ER4_FR4_Full_BIT, + ETHTOOL_LINK_MODE_400000baseDR4_Full_BIT, + ETHTOOL_LINK_MODE_400000baseCR4_Full_BIT, +}; + +#define MLXSW_SP2_MASK_ETHTOOL_400GAUI_4_400GBASE_CR4_KR4_LEN \ + ARRAY_SIZE(mlxsw_sp2_mask_ethtool_400gaui_4_400gbase_cr4_kr4) + static const enum ethtool_link_mode_bit_indices mlxsw_sp2_mask_ethtool_800gaui_8[] = { ETHTOOL_LINK_MODE_800000baseCR8_Full_BIT, @@ -1816,6 +1852,14 @@ static const struct mlxsw_sp2_port_link_mode mlxsw_sp2_port_link_mode[] = { .speed = SPEED_100000, .width = 2, }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_100GAUI_1_100GBASE_CR_KR, + .mask_ethtool = mlxsw_sp2_mask_ethtool_100gaui_1_100gbase_cr_kr, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_100GAUI_1_100GBASE_CR_KR_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_1X, + .speed = SPEED_100000, + .width = 1, + }, { .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_200GAUI_4_200GBASE_CR4_KR4, .mask_ethtool = mlxsw_sp2_mask_ethtool_200gaui_4_200gbase_cr4_kr4, @@ -1825,6 +1869,14 @@ static const struct mlxsw_sp2_port_link_mode mlxsw_sp2_port_link_mode[] = { .speed = SPEED_200000, .width = 4, }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_200GAUI_2_200GBASE_CR2_KR2, + .mask_ethtool = mlxsw_sp2_mask_ethtool_200gaui_2_200gbase_cr2_kr2, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_200GAUI_2_200GBASE_CR2_KR2_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_2X, + .speed = SPEED_200000, + .width = 2, + }, { .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_400GAUI_8, .mask_ethtool = mlxsw_sp2_mask_ethtool_400gaui_8, @@ -1833,6 +1885,14 @@ static const struct mlxsw_sp2_port_link_mode mlxsw_sp2_port_link_mode[] = { .speed = SPEED_400000, .width = 8, }, + { + .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_400GAUI_4_400GBASE_CR4_KR4, + .mask_ethtool = mlxsw_sp2_mask_ethtool_400gaui_4_400gbase_cr4_kr4, + .m_ethtool_len = MLXSW_SP2_MASK_ETHTOOL_400GAUI_4_400GBASE_CR4_KR4_LEN, + .mask_sup_width = MLXSW_SP_PORT_MASK_WIDTH_4X, + .speed = SPEED_400000, + .width = 4, + }, { .mask = MLXSW_REG_PTYS_EXT_ETH_SPEED_800GAUI_8, .mask_ethtool = mlxsw_sp2_mask_ethtool_800gaui_8, From 17b0dfa1f35bf58c17ae75da4af99e6b2c51ed57 Mon Sep 17 00:00:00 2001 From: Pavan Chebbi Date: Tue, 9 Apr 2024 14:54:25 -0700 Subject: [PATCH 07/16] bnxt_en: Skip ethtool RSS context configuration in ifdown state The current implementation requires the ifstate to be up when configuring the RSS contexts. It will try to fetch the RX ring IDs and will crash if it is in ifdown state. Return error if !netif_running() to prevent the crash. An improved implementation is in the works to allow RSS contexts to be changed while in ifdown state. Fixes: b3d0083caf9a ("bnxt_en: Support RSS contexts in ethtool .{get|set}_rxfh()") Reviewed-by: Kalesh AP Reviewed-by: Somnath Kotur Signed-off-by: Pavan Chebbi Signed-off-by: Michael Chan Reviewed-by: Jacob Keller Link: https://lore.kernel.org/r/20240409215431.41424-2-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c index 9c49f629d565..68444234b268 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c @@ -1876,6 +1876,11 @@ static int bnxt_set_rxfh_context(struct bnxt *bp, return -EOPNOTSUPP; } + if (!netif_running(bp->dev)) { + NL_SET_ERR_MSG_MOD(extack, "Unable to set RSS contexts when interface is down"); + return -EAGAIN; + } + if (*rss_context != ETH_RXFH_CONTEXT_ALLOC) { rss_ctx = bnxt_get_rss_ctx_from_index(bp, *rss_context); if (!rss_ctx) { From 43226dccd1bd74c6bf97d8d357e782f4922d39e3 Mon Sep 17 00:00:00 2001 From: Kalesh AP Date: Tue, 9 Apr 2024 14:54:26 -0700 Subject: [PATCH 08/16] bnxt_en: Remove a redundant NULL check in bnxt_register_dev() The memory for "edev->ulp_tbl" is allocated inside the bnxt_rdma_aux_device_init() function. If it fails, the driver will not create the auxiliary device for RoCE. Hence the NULL check inside bnxt_register_dev() is unnecessary. Reviewed-by: Vikas Gupta Reviewed-by: Somnath Kotur Signed-off-by: Kalesh AP Signed-off-by: Michael Chan Reviewed-by: Jacob Keller Link: https://lore.kernel.org/r/20240409215431.41424-3-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c index 86dcd2c76587..fd890819d4bc 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c @@ -64,9 +64,6 @@ int bnxt_register_dev(struct bnxt_en_dev *edev, return -ENOMEM; ulp = edev->ulp_tbl; - if (!ulp) - return -ENOMEM; - ulp->handle = handle; rcu_assign_pointer(ulp->ulp_ops, ulp_ops); From b58f5a9c7034e02c391f4ae70fc62c927ce8f04f Mon Sep 17 00:00:00 2001 From: Vikas Gupta Date: Tue, 9 Apr 2024 14:54:27 -0700 Subject: [PATCH 09/16] bnxt_en: Remove unneeded MSIX base structure fields and code Ever since commit: 303432211324 ("bnxt_en: Remove runtime interrupt vector allocation") The MSIX base vector is effectively always 0. Remove all unneeded structure fields and code referencing the MSIX base. Reviewed-by: Somnath Kotur Reviewed-by: Andy Gospodarek Signed-off-by: Vikas Gupta Signed-off-by: Michael Chan Reviewed-by: Jacob Keller Link: https://lore.kernel.org/r/20240409215431.41424-4-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 31 +++---------------- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 20 +++--------- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 2 -- 3 files changed, 8 insertions(+), 45 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 795f3f957eb5..8213a37cba3e 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -3905,13 +3905,12 @@ static int bnxt_alloc_cp_sub_ring(struct bnxt *bp, static int bnxt_alloc_cp_rings(struct bnxt *bp) { bool sh = !!(bp->flags & BNXT_FLAG_SHARED_RINGS); - int i, j, rc, ulp_base_vec, ulp_msix; + int i, j, rc, ulp_msix; int tcs = bp->num_tc; if (!tcs) tcs = 1; ulp_msix = bnxt_get_ulp_msix_num(bp); - ulp_base_vec = bnxt_get_ulp_msix_base(bp); for (i = 0, j = 0; i < bp->cp_nr_rings; i++) { struct bnxt_napi *bnapi = bp->bnapi[i]; struct bnxt_cp_ring_info *cpr, *cpr2; @@ -3930,10 +3929,7 @@ static int bnxt_alloc_cp_rings(struct bnxt *bp) if (rc) return rc; - if (ulp_msix && i >= ulp_base_vec) - ring->map_idx = i + ulp_msix; - else - ring->map_idx = i; + ring->map_idx = ulp_msix + i; if (!(bp->flags & BNXT_FLAG_CHIP_P5_PLUS)) continue; @@ -7347,17 +7343,7 @@ static int bnxt_hwrm_reserve_rings(struct bnxt *bp, struct bnxt_hw_rings *hwr) int bnxt_nq_rings_in_use(struct bnxt *bp) { - int cp = bp->cp_nr_rings; - int ulp_msix, ulp_base; - - ulp_msix = bnxt_get_ulp_msix_num(bp); - if (ulp_msix) { - ulp_base = bnxt_get_ulp_msix_base(bp); - cp += ulp_msix; - if ((ulp_base + ulp_msix) > cp) - cp = ulp_base + ulp_msix; - } - return cp; + return bp->cp_nr_rings + bnxt_get_ulp_msix_num(bp); } static int bnxt_cp_rings_in_use(struct bnxt *bp) @@ -7373,16 +7359,7 @@ static int bnxt_cp_rings_in_use(struct bnxt *bp) static int bnxt_get_func_stat_ctxs(struct bnxt *bp) { - int ulp_stat = bnxt_get_ulp_stat_ctxs(bp); - int cp = bp->cp_nr_rings; - - if (!ulp_stat) - return cp; - - if (bnxt_nq_rings_in_use(bp) > cp + bnxt_get_ulp_msix_num(bp)) - return bnxt_get_ulp_msix_base(bp) + ulp_stat; - - return cp + ulp_stat; + return bp->cp_nr_rings + bnxt_get_ulp_stat_ctxs(bp); } static int bnxt_get_total_rss_ctxs(struct bnxt *bp, struct bnxt_hw_rings *hwr) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c index fd890819d4bc..cae88747b110 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c @@ -31,21 +31,20 @@ static DEFINE_IDA(bnxt_aux_dev_ids); static void bnxt_fill_msix_vecs(struct bnxt *bp, struct bnxt_msix_entry *ent) { struct bnxt_en_dev *edev = bp->edev; - int num_msix, idx, i; + int num_msix, i; if (!edev->ulp_tbl->msix_requested) { netdev_warn(bp->dev, "Requested MSI-X vectors insufficient\n"); return; } num_msix = edev->ulp_tbl->msix_requested; - idx = edev->ulp_tbl->msix_base; for (i = 0; i < num_msix; i++) { - ent[i].vector = bp->irq_tbl[idx + i].vector; - ent[i].ring_idx = idx + i; + ent[i].vector = bp->irq_tbl[i].vector; + ent[i].ring_idx = i; if (bp->flags & BNXT_FLAG_CHIP_P5_PLUS) ent[i].db_offset = bp->db_offset; else - ent[i].db_offset = (idx + i) * 0x80; + ent[i].db_offset = i * 0x80; } } @@ -111,17 +110,6 @@ int bnxt_get_ulp_msix_num(struct bnxt *bp) min_t(u32, roce_msix, num_online_cpus()) : 0); } -int bnxt_get_ulp_msix_base(struct bnxt *bp) -{ - if (bnxt_ulp_registered(bp->edev)) { - struct bnxt_en_dev *edev = bp->edev; - - if (edev->ulp_tbl->msix_requested) - return edev->ulp_tbl->msix_base; - } - return 0; -} - int bnxt_get_ulp_stat_ctxs(struct bnxt *bp) { if (bnxt_ulp_registered(bp->edev)) { diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h index b9e73de14b57..f8df4095399f 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h @@ -46,7 +46,6 @@ struct bnxt_ulp { unsigned long *async_events_bmap; u16 max_async_event_id; u16 msix_requested; - u16 msix_base; atomic_t ref_count; }; @@ -96,7 +95,6 @@ static inline bool bnxt_ulp_registered(struct bnxt_en_dev *edev) } int bnxt_get_ulp_msix_num(struct bnxt *bp); -int bnxt_get_ulp_msix_base(struct bnxt *bp); int bnxt_get_ulp_stat_ctxs(struct bnxt *bp); void bnxt_ulp_stop(struct bnxt *bp); void bnxt_ulp_start(struct bnxt *bp, int err); From 194fad5b27815ca80e832ac875c0026ff96fc243 Mon Sep 17 00:00:00 2001 From: Vikas Gupta Date: Tue, 9 Apr 2024 14:54:28 -0700 Subject: [PATCH 10/16] bnxt_en: Refactor bnxt_rdma_aux_device_init/uninit functions In its current form, bnxt_rdma_aux_device_init() not only initializes the necessary data structures of the newly created aux device but also adds the aux device into the aux bus subsytem. Refactor the logic into separate functions, first function to initialize the aux device along with the required resources and second, to actually add the device to the aux bus subsytem. This separation helps to create bnxt_en_dev much earlier and save its resources separately. Reviewed-by: Andy Gospodarek Signed-off-by: Vikas Gupta Signed-off-by: Michael Chan Reviewed-by: Jacob Keller Link: https://lore.kernel.org/r/20240409215431.41424-5-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 10 ++++-- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 33 ++++++++++++++----- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 2 ++ 3 files changed, 35 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 8213a37cba3e..8c24e83ba050 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -14786,10 +14786,13 @@ static void bnxt_remove_one(struct pci_dev *pdev) if (BNXT_PF(bp)) bnxt_sriov_disable(bp); - bnxt_rdma_aux_device_uninit(bp); + bnxt_rdma_aux_device_del(bp); bnxt_ptp_clear(bp); unregister_netdev(dev); + + bnxt_rdma_aux_device_uninit(bp); + bnxt_free_l2_filters(bp, true); bnxt_free_ntp_fltrs(bp, true); if (BNXT_SUPPORTS_MULTI_RSS_CTX(bp)) @@ -15408,13 +15411,15 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) if (BNXT_SUPPORTS_NTUPLE_VNIC(bp)) bnxt_init_multi_rss_ctx(bp); + bnxt_rdma_aux_device_init(bp); + rc = register_netdev(dev); if (rc) goto init_err_cleanup; bnxt_dl_fw_reporters_create(bp); - bnxt_rdma_aux_device_init(bp); + bnxt_rdma_aux_device_add(bp); bnxt_print_device_info(bp); @@ -15422,6 +15427,7 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) return 0; init_err_cleanup: + bnxt_rdma_aux_device_uninit(bp); bnxt_dl_unregister(bp); init_err_dl: bnxt_shutdown_tc(bp); diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c index cae88747b110..ae2b367b1226 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c @@ -291,7 +291,6 @@ void bnxt_rdma_aux_device_uninit(struct bnxt *bp) aux_priv = bp->aux_priv; adev = &aux_priv->aux_dev; - auxiliary_device_delete(adev); auxiliary_device_uninit(adev); } @@ -309,6 +308,14 @@ static void bnxt_aux_dev_release(struct device *dev) bp->aux_priv = NULL; } +void bnxt_rdma_aux_device_del(struct bnxt *bp) +{ + if (!bp->edev) + return; + + auxiliary_device_delete(&bp->aux_priv->aux_dev); +} + static void bnxt_set_edev_info(struct bnxt_en_dev *edev, struct bnxt *bp) { edev->net = bp->dev; @@ -332,6 +339,23 @@ static void bnxt_set_edev_info(struct bnxt_en_dev *edev, struct bnxt *bp) edev->ulp_tbl->msix_requested = bnxt_get_ulp_msix_num(bp); } +void bnxt_rdma_aux_device_add(struct bnxt *bp) +{ + struct auxiliary_device *aux_dev; + int rc; + + if (!bp->edev) + return; + + aux_dev = &bp->aux_priv->aux_dev; + rc = auxiliary_device_add(aux_dev); + if (rc) { + netdev_warn(bp->dev, "Failed to add auxiliary device for ROCE\n"); + auxiliary_device_uninit(aux_dev); + bp->flags &= ~BNXT_FLAG_ROCE_CAP; + } +} + void bnxt_rdma_aux_device_init(struct bnxt *bp) { struct auxiliary_device *aux_dev; @@ -386,13 +410,6 @@ void bnxt_rdma_aux_device_init(struct bnxt *bp) bp->edev = edev; bnxt_set_edev_info(edev, bp); - rc = auxiliary_device_add(aux_dev); - if (rc) { - netdev_warn(bp->dev, - "Failed to add auxiliary device for ROCE\n"); - goto aux_dev_uninit; - } - return; aux_dev_uninit: diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h index f8df4095399f..ae7266ddf167 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h @@ -103,6 +103,8 @@ void bnxt_ulp_irq_stop(struct bnxt *bp); void bnxt_ulp_irq_restart(struct bnxt *bp, int err); void bnxt_ulp_async_events(struct bnxt *bp, struct hwrm_async_event_cmpl *cmpl); void bnxt_rdma_aux_device_uninit(struct bnxt *bp); +void bnxt_rdma_aux_device_del(struct bnxt *bp); +void bnxt_rdma_aux_device_add(struct bnxt *bp); void bnxt_rdma_aux_device_init(struct bnxt *bp); int bnxt_register_dev(struct bnxt_en_dev *edev, struct bnxt_ulp_ops *ulp_ops, void *handle); From 2e4592dc9bee5ca4fd3da4c5451c7f97c76442bf Mon Sep 17 00:00:00 2001 From: Vikas Gupta Date: Tue, 9 Apr 2024 14:54:29 -0700 Subject: [PATCH 11/16] bnxt_en: Change MSIX/NQs allocation policy The existing scheme sets aside a number of MSIX/NQs for the RoCE driver whether the RoCE driver is registered or not. This scheme is not flexible and limits the resources available for the L2 rings if RoCE is never used. Modify the scheme so that the RoCE MSIX/NQs can be used by the L2 driver if they are not used for RoCE. The MSIX/NQs are now represented by 3 fields. bp->ulp_num_msix_want contains the desired default value, edev->ulp_num_msix_vec contains the available value (but not necessarily in use), and ulp_tbl->msix_requested contains the actual value in use by RoCE. The L2 driver can dip into edev->ulp_num_msix_vec if necessary. We need to add rtnl_lock() back in bnxt_register_dev() and bnxt_unregister_dev() to synchronize the MSIX usage between L2 and RoCE. Reviewed-by: Andy Gospodarek Signed-off-by: Vikas Gupta Signed-off-by: Michael Chan Reviewed-by: Jacob Keller Link: https://lore.kernel.org/r/20240409215431.41424-6-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 25 +++++- drivers/net/ethernet/broadcom/bnxt/bnxt.h | 1 + drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 77 +++++++++++++++---- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 8 +- 4 files changed, 92 insertions(+), 19 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 8c24e83ba050..88cf8f47e071 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -7549,6 +7549,20 @@ static int __bnxt_reserve_rings(struct bnxt *bp) if (!netif_is_rxfh_configured(bp->dev)) bnxt_set_dflt_rss_indir_tbl(bp, NULL); + if (!bnxt_ulp_registered(bp->edev) && BNXT_NEW_RM(bp)) { + int resv_msix, resv_ctx, ulp_msix, ulp_ctxs; + struct bnxt_hw_resc *hw_resc; + + hw_resc = &bp->hw_resc; + resv_msix = hw_resc->resv_irqs - bp->cp_nr_rings; + ulp_msix = bnxt_get_ulp_msix_num(bp); + ulp_msix = min_t(int, resv_msix, ulp_msix); + bnxt_set_ulp_msix_num(bp, ulp_msix); + resv_ctx = hw_resc->resv_stat_ctxs - bp->cp_nr_rings; + ulp_ctxs = min(resv_ctx, bnxt_get_ulp_stat_ctxs(bp)); + bnxt_set_ulp_stat_ctxs(bp, ulp_ctxs); + } + return rc; } @@ -14982,6 +14996,7 @@ static void bnxt_trim_dflt_sh_rings(struct bnxt *bp) static int bnxt_set_dflt_rings(struct bnxt *bp, bool sh) { int dflt_rings, max_rx_rings, max_tx_rings, rc; + int avail_msix; if (!bnxt_can_reserve_rings(bp)) return 0; @@ -15009,6 +15024,14 @@ static int bnxt_set_dflt_rings(struct bnxt *bp, bool sh) bp->cp_nr_rings = bp->tx_nr_rings_per_tc + bp->rx_nr_rings; bp->tx_nr_rings = bp->tx_nr_rings_per_tc; + avail_msix = bnxt_get_max_func_irqs(bp) - bp->cp_nr_rings; + if (avail_msix >= BNXT_MIN_ROCE_CP_RINGS) { + int ulp_num_msix = min(avail_msix, bp->ulp_num_msix_want); + + bnxt_set_ulp_msix_num(bp, ulp_num_msix); + bnxt_set_dflt_ulp_stat_ctxs(bp); + } + rc = __bnxt_reserve_rings(bp); if (rc && rc != -ENODEV) netdev_warn(bp->dev, "Unable to reserve tx rings\n"); @@ -15358,6 +15381,7 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) bnxt_set_rx_skb_mode(bp, false); bnxt_set_tpa_flags(bp); bnxt_set_ring_params(bp); + bnxt_rdma_aux_device_init(bp); rc = bnxt_set_dflt_rings(bp, true); if (rc) { if (BNXT_VF(bp) && rc == -ENODEV) { @@ -15411,7 +15435,6 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) if (BNXT_SUPPORTS_NTUPLE_VNIC(bp)) bnxt_init_multi_rss_ctx(bp); - bnxt_rdma_aux_device_init(bp); rc = register_netdev(dev); if (rc) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h index 0640fcb57ef8..ad57ef051798 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h @@ -2303,6 +2303,7 @@ struct bnxt { struct bnxt_irq *irq_tbl; int total_irqs; + int ulp_num_msix_want; u8 mac_addr[ETH_ALEN]; #ifdef CONFIG_BNXT_DCB diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c index ae2b367b1226..de2cb1d4cd98 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c @@ -48,6 +48,46 @@ static void bnxt_fill_msix_vecs(struct bnxt *bp, struct bnxt_msix_entry *ent) } } +int bnxt_get_ulp_msix_num(struct bnxt *bp) +{ + if (bp->edev) + return bp->edev->ulp_num_msix_vec; + return 0; +} + +void bnxt_set_ulp_msix_num(struct bnxt *bp, int num) +{ + if (bp->edev) + bp->edev->ulp_num_msix_vec = num; +} + +int bnxt_get_ulp_stat_ctxs(struct bnxt *bp) +{ + if (bp->edev) + return bp->edev->ulp_num_ctxs; + return 0; +} + +void bnxt_set_ulp_stat_ctxs(struct bnxt *bp, int num_ulp_ctx) +{ + if (bp->edev) + bp->edev->ulp_num_ctxs = num_ulp_ctx; +} + +void bnxt_set_dflt_ulp_stat_ctxs(struct bnxt *bp) +{ + if (bp->edev) { + bp->edev->ulp_num_ctxs = BNXT_MIN_ROCE_STAT_CTXS; + /* Reserve one additional stat_ctx for PF0 (except + * on 1-port NICs) as it also creates one stat_ctx + * for PF1 in case of RoCE bonding. + */ + if (BNXT_PF(bp) && !bp->pf.port_id && + bp->port_count > 1) + bp->edev->ulp_num_ctxs++; + } +} + int bnxt_register_dev(struct bnxt_en_dev *edev, struct bnxt_ulp_ops *ulp_ops, void *handle) @@ -56,11 +96,19 @@ int bnxt_register_dev(struct bnxt_en_dev *edev, struct bnxt *bp = netdev_priv(dev); unsigned int max_stat_ctxs; struct bnxt_ulp *ulp; + int rc = 0; + rtnl_lock(); + if (!bp->irq_tbl) { + rc = -ENODEV; + goto exit; + } max_stat_ctxs = bnxt_get_max_func_stat_ctxs(bp); if (max_stat_ctxs <= BNXT_MIN_ROCE_STAT_CTXS || - bp->cp_nr_rings == max_stat_ctxs) - return -ENOMEM; + bp->cp_nr_rings == max_stat_ctxs) { + rc = -ENOMEM; + goto exit; + } ulp = edev->ulp_tbl; ulp->handle = handle; @@ -69,9 +117,13 @@ int bnxt_register_dev(struct bnxt_en_dev *edev, if (test_bit(BNXT_STATE_OPEN, &bp->state)) bnxt_hwrm_vnic_cfg(bp, &bp->vnic_info[BNXT_VNIC_DEFAULT]); + edev->ulp_tbl->msix_requested = bnxt_get_ulp_msix_num(bp); + bnxt_fill_msix_vecs(bp, bp->edev->msix_entries); edev->flags |= BNXT_EN_FLAG_MSIX_REQUESTED; - return 0; +exit: + rtnl_unlock(); + return rc; } EXPORT_SYMBOL(bnxt_register_dev); @@ -83,8 +135,10 @@ void bnxt_unregister_dev(struct bnxt_en_dev *edev) int i = 0; ulp = edev->ulp_tbl; + rtnl_lock(); if (ulp->msix_requested) edev->flags &= ~BNXT_EN_FLAG_MSIX_REQUESTED; + edev->ulp_tbl->msix_requested = 0; if (ulp->max_async_event_id) bnxt_hwrm_func_drv_rgtr(bp, NULL, 0, true); @@ -97,11 +151,12 @@ void bnxt_unregister_dev(struct bnxt_en_dev *edev) msleep(100); i++; } + rtnl_unlock(); return; } EXPORT_SYMBOL(bnxt_unregister_dev); -int bnxt_get_ulp_msix_num(struct bnxt *bp) +static int bnxt_set_dflt_ulp_msix(struct bnxt *bp) { u32 roce_msix = BNXT_VF(bp) ? BNXT_MAX_VF_ROCE_MSIX : BNXT_MAX_ROCE_MSIX; @@ -110,18 +165,6 @@ int bnxt_get_ulp_msix_num(struct bnxt *bp) min_t(u32, roce_msix, num_online_cpus()) : 0); } -int bnxt_get_ulp_stat_ctxs(struct bnxt *bp) -{ - if (bnxt_ulp_registered(bp->edev)) { - struct bnxt_en_dev *edev = bp->edev; - - if (edev->ulp_tbl->msix_requested) - return BNXT_MIN_ROCE_STAT_CTXS; - } - - return 0; -} - int bnxt_send_msg(struct bnxt_en_dev *edev, struct bnxt_fw_msg *fw_msg) { @@ -336,7 +379,6 @@ static void bnxt_set_edev_info(struct bnxt_en_dev *edev, struct bnxt *bp) edev->pf_port_id = bp->pf.port_id; edev->en_state = bp->state; edev->bar0 = bp->bar0; - edev->ulp_tbl->msix_requested = bnxt_get_ulp_msix_num(bp); } void bnxt_rdma_aux_device_add(struct bnxt *bp) @@ -409,6 +451,7 @@ void bnxt_rdma_aux_device_init(struct bnxt *bp) aux_priv->edev = edev; bp->edev = edev; bnxt_set_edev_info(edev, bp); + bp->ulp_num_msix_want = bnxt_set_dflt_ulp_msix(bp); return; diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h index ae7266ddf167..04ce3328e66f 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h @@ -85,17 +85,23 @@ struct bnxt_en_dev { * updated in resume. */ void __iomem *bar0; + + u16 ulp_num_msix_vec; + u16 ulp_num_ctxs; }; static inline bool bnxt_ulp_registered(struct bnxt_en_dev *edev) { - if (edev && edev->ulp_tbl) + if (edev && rcu_access_pointer(edev->ulp_tbl->ulp_ops)) return true; return false; } int bnxt_get_ulp_msix_num(struct bnxt *bp); +void bnxt_set_ulp_msix_num(struct bnxt *bp, int num); int bnxt_get_ulp_stat_ctxs(struct bnxt *bp); +void bnxt_set_ulp_stat_ctxs(struct bnxt *bp, int num_ctxs); +void bnxt_set_dflt_ulp_stat_ctxs(struct bnxt *bp); void bnxt_ulp_stop(struct bnxt *bp); void bnxt_ulp_start(struct bnxt *bp, int err); void bnxt_ulp_sriov_cfg(struct bnxt *bp, int num_vfs); From d630624ebd708fb659b18a8d1423e8e26de075ab Mon Sep 17 00:00:00 2001 From: Vikas Gupta Date: Tue, 9 Apr 2024 14:54:30 -0700 Subject: [PATCH 12/16] bnxt_en: Utilize ulp client resources if RoCE is not registered If the RoCE driver is not registered for a RoCE capable device, add flexibility to use the RoCE resources (MSIX/NQs) for L2 purposes, such as additional rings configured by the user or for XDP. Reviewed-by: Andy Gospodarek Signed-off-by: Vikas Gupta Signed-off-by: Michael Chan Reviewed-by: Jacob Keller Link: https://lore.kernel.org/r/20240409215431.41424-7-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 41 +++++++++++++++---- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 14 +++++++ drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 2 + 3 files changed, 48 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 88cf8f47e071..a2e21fe64ab9 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -7470,14 +7470,27 @@ static bool bnxt_rings_ok(struct bnxt *bp, struct bnxt_hw_rings *hwr) static int __bnxt_reserve_rings(struct bnxt *bp) { struct bnxt_hw_rings hwr = {0}; + int cp = bp->cp_nr_rings; int rx_rings, rc; + int ulp_msix = 0; bool sh = false; int tx_cp; if (!bnxt_need_reserve_rings(bp)) return 0; - hwr.cp = bnxt_nq_rings_in_use(bp); + if (!bnxt_ulp_registered(bp->edev)) { + ulp_msix = bnxt_get_avail_msix(bp, bp->ulp_num_msix_want); + if (!ulp_msix) + bnxt_set_ulp_stat_ctxs(bp, 0); + + if (ulp_msix > bp->ulp_num_msix_want) + ulp_msix = bp->ulp_num_msix_want; + hwr.cp = cp + ulp_msix; + } else { + hwr.cp = bnxt_nq_rings_in_use(bp); + } + hwr.tx = bp->tx_nr_rings; hwr.rx = bp->rx_nr_rings; if (bp->flags & BNXT_FLAG_SHARED_RINGS) @@ -7550,12 +7563,11 @@ static int __bnxt_reserve_rings(struct bnxt *bp) bnxt_set_dflt_rss_indir_tbl(bp, NULL); if (!bnxt_ulp_registered(bp->edev) && BNXT_NEW_RM(bp)) { - int resv_msix, resv_ctx, ulp_msix, ulp_ctxs; + int resv_msix, resv_ctx, ulp_ctxs; struct bnxt_hw_resc *hw_resc; hw_resc = &bp->hw_resc; resv_msix = hw_resc->resv_irqs - bp->cp_nr_rings; - ulp_msix = bnxt_get_ulp_msix_num(bp); ulp_msix = min_t(int, resv_msix, ulp_msix); bnxt_set_ulp_msix_num(bp, ulp_msix); resv_ctx = hw_resc->resv_stat_ctxs - bp->cp_nr_rings; @@ -10609,13 +10621,23 @@ int bnxt_reserve_rings(struct bnxt *bp, bool irq_re_init) { bool irq_cleared = false; int tcs = bp->num_tc; + int irqs_required; int rc; if (!bnxt_need_reserve_rings(bp)) return 0; - if (irq_re_init && BNXT_NEW_RM(bp) && - bnxt_get_num_msix(bp) != bp->total_irqs) { + if (!bnxt_ulp_registered(bp->edev)) { + int ulp_msix = bnxt_get_avail_msix(bp, bp->ulp_num_msix_want); + + if (ulp_msix > bp->ulp_num_msix_want) + ulp_msix = bp->ulp_num_msix_want; + irqs_required = ulp_msix + bp->cp_nr_rings; + } else { + irqs_required = bnxt_get_num_msix(bp); + } + + if (irq_re_init && BNXT_NEW_RM(bp) && irqs_required != bp->total_irqs) { bnxt_ulp_irq_stop(bp); bnxt_clear_int_mode(bp); irq_cleared = true; @@ -13625,8 +13647,8 @@ int bnxt_check_rings(struct bnxt *bp, int tx, int rx, bool sh, int tcs, return -ENOMEM; hwr.stat = hwr.cp; if (BNXT_NEW_RM(bp)) { - hwr.cp += bnxt_get_ulp_msix_num(bp); - hwr.stat += bnxt_get_ulp_stat_ctxs(bp); + hwr.cp += bnxt_get_ulp_msix_num_in_use(bp); + hwr.stat += bnxt_get_ulp_stat_ctxs_in_use(bp); hwr.grp = rx; hwr.rss_ctx = bnxt_get_total_rss_ctxs(bp, &hwr); } @@ -14899,8 +14921,9 @@ static void _bnxt_get_max_rings(struct bnxt *bp, int *max_rx, int *max_tx, *max_rx = hw_resc->max_rx_rings; *max_cp = bnxt_get_max_func_cp_rings_for_en(bp); max_irq = min_t(int, bnxt_get_max_func_irqs(bp) - - bnxt_get_ulp_msix_num(bp), - hw_resc->max_stat_ctxs - bnxt_get_ulp_stat_ctxs(bp)); + bnxt_get_ulp_msix_num_in_use(bp), + hw_resc->max_stat_ctxs - + bnxt_get_ulp_stat_ctxs_in_use(bp)); if (!(bp->flags & BNXT_FLAG_CHIP_P5_PLUS)) *max_cp = min_t(int, *max_cp, max_irq); max_ring_grps = hw_resc->max_hw_ring_grps; diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c index de2cb1d4cd98..edb10aebd095 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c @@ -61,6 +61,13 @@ void bnxt_set_ulp_msix_num(struct bnxt *bp, int num) bp->edev->ulp_num_msix_vec = num; } +int bnxt_get_ulp_msix_num_in_use(struct bnxt *bp) +{ + if (bnxt_ulp_registered(bp->edev)) + return bp->edev->ulp_num_msix_vec; + return 0; +} + int bnxt_get_ulp_stat_ctxs(struct bnxt *bp) { if (bp->edev) @@ -74,6 +81,13 @@ void bnxt_set_ulp_stat_ctxs(struct bnxt *bp, int num_ulp_ctx) bp->edev->ulp_num_ctxs = num_ulp_ctx; } +int bnxt_get_ulp_stat_ctxs_in_use(struct bnxt *bp) +{ + if (bnxt_ulp_registered(bp->edev)) + return bp->edev->ulp_num_ctxs; + return 0; +} + void bnxt_set_dflt_ulp_stat_ctxs(struct bnxt *bp) { if (bp->edev) { diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h index 04ce3328e66f..b86baf901a5d 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h @@ -98,9 +98,11 @@ static inline bool bnxt_ulp_registered(struct bnxt_en_dev *edev) } int bnxt_get_ulp_msix_num(struct bnxt *bp); +int bnxt_get_ulp_msix_num_in_use(struct bnxt *bp); void bnxt_set_ulp_msix_num(struct bnxt *bp, int num); int bnxt_get_ulp_stat_ctxs(struct bnxt *bp); void bnxt_set_ulp_stat_ctxs(struct bnxt *bp, int num_ctxs); +int bnxt_get_ulp_stat_ctxs_in_use(struct bnxt *bp); void bnxt_set_dflt_ulp_stat_ctxs(struct bnxt *bp); void bnxt_ulp_stop(struct bnxt *bp); void bnxt_ulp_start(struct bnxt *bp, int err); From 008ce0fd39036fb077d6b24ffd146ae5ebe3bd77 Mon Sep 17 00:00:00 2001 From: Michael Chan Date: Tue, 9 Apr 2024 14:54:31 -0700 Subject: [PATCH 13/16] bnxt_en: Update MODULE_DESCRIPTION Update MODULE_DESCRIPTION to the more generic adapter family name. The old name only includes the first generation of supported adapters. Signed-off-by: Michael Chan Reviewed-by: Jacob Keller Link: https://lore.kernel.org/r/20240409215431.41424-8-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index a2e21fe64ab9..d728df139c9c 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -76,7 +76,7 @@ NETIF_MSG_TX_ERR) MODULE_LICENSE("GPL"); -MODULE_DESCRIPTION("Broadcom BCM573xx network driver"); +MODULE_DESCRIPTION("Broadcom NetXtreme network driver"); #define BNXT_RX_OFFSET (NET_SKB_PAD + NET_IP_ALIGN) #define BNXT_RX_DMA_OFFSET NET_SKB_PAD From bfec4e18f94351d2ff9073c4bffcc6f37bfa3e0b Mon Sep 17 00:00:00 2001 From: Erick Archer Date: Sat, 6 Apr 2024 16:23:35 +0200 Subject: [PATCH 14/16] net: mana: Add flex array to struct mana_cfg_rx_steer_req_v2 The "struct mana_cfg_rx_steer_req_v2" uses a dynamically sized set of trailing elements. Specifically, it uses a "mana_handle_t" array. So, use the preferred way in the kernel declaring a flexible array [1]. At the same time, prepare for the coming implementation by GCC and Clang of the __counted_by attribute. Flexible array members annotated with __counted_by can have their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family functions). This is a previous step to refactor the two consumers of this structure. drivers/infiniband/hw/mana/qp.c drivers/net/ethernet/microsoft/mana/mana_en.c The ultimate goal is to avoid the open-coded arithmetic in the memory allocator functions [2] using the "struct_size" macro. Link: https://www.kernel.org/doc/html/next/process/deprecated.html#zero-length-and-one-element-arrays [1] Link: https://www.kernel.org/doc/html/next/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments [2] Signed-off-by: Erick Archer Link: https://lore.kernel.org/r/AS8PR02MB7237E2900247571C9CB84C678B022@AS8PR02MB7237.eurprd02.prod.outlook.com Reviewed-by: Long Li Reviewed-by: Justin Stitt Signed-off-by: Leon Romanovsky --- include/net/mana/mana.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/net/mana/mana.h b/include/net/mana/mana.h index 76147feb0d10..46f741ebce21 100644 --- a/include/net/mana/mana.h +++ b/include/net/mana/mana.h @@ -671,6 +671,7 @@ struct mana_cfg_rx_steer_req_v2 { u8 hashkey[MANA_HASH_KEY_SIZE]; u8 cqe_coalescing_enable; u8 reserved2[7]; + mana_handle_t indir_tab[] __counted_by(num_indir_entries); }; /* HW DATA */ struct mana_cfg_rx_steer_resp { From 29b8e13a8b4c95ce629c8d4c84682f29af6f6bb5 Mon Sep 17 00:00:00 2001 From: Erick Archer Date: Sat, 6 Apr 2024 16:23:36 +0200 Subject: [PATCH 15/16] RDMA/mana_ib: Prefer struct_size over open coded arithmetic This is an effort to get rid of all multiplications from allocation functions in order to prevent integer overflows [1][2]. As the "req" variable is a pointer to "struct mana_cfg_rx_steer_req_v2" and this structure ends in a flexible array: struct mana_cfg_rx_steer_req_v2 { [...] mana_handle_t indir_tab[] __counted_by(num_indir_entries); }; the preferred way in the kernel is to use the struct_size() helper to do the arithmetic instead of the calculation "size + size * count" in the kzalloc() function. Moreover, use the "offsetof" helper to get the indirect table offset instead of the "sizeof" operator and avoid the open-coded arithmetic in pointers using the new flex member. This new structure member also allow us to remove the "req_indir_tab" variable since it is no longer needed. This way, the code is more readable and safer. This code was detected with the help of Coccinelle, and audited and modified manually. Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments [1] Link: https://github.com/KSPP/linux/issues/160 [2] Signed-off-by: Erick Archer Link: https://lore.kernel.org/r/AS8PR02MB72375EB06EE1A84A67BE722E8B022@AS8PR02MB7237.eurprd02.prod.outlook.com Reviewed-by: Long Li Reviewed-by: Justin Stitt Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/mana/qp.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) diff --git a/drivers/infiniband/hw/mana/qp.c b/drivers/infiniband/hw/mana/qp.c index 6e7627745c95..258f89464c10 100644 --- a/drivers/infiniband/hw/mana/qp.c +++ b/drivers/infiniband/hw/mana/qp.c @@ -15,15 +15,13 @@ static int mana_ib_cfg_vport_steering(struct mana_ib_dev *dev, struct mana_port_context *mpc = netdev_priv(ndev); struct mana_cfg_rx_steer_req_v2 *req; struct mana_cfg_rx_steer_resp resp = {}; - mana_handle_t *req_indir_tab; struct gdma_context *gc; u32 req_buf_size; int i, err; gc = mdev_to_gc(dev); - req_buf_size = - sizeof(*req) + sizeof(mana_handle_t) * MANA_INDIRECT_TABLE_SIZE; + req_buf_size = struct_size(req, indir_tab, MANA_INDIRECT_TABLE_SIZE); req = kzalloc(req_buf_size, GFP_KERNEL); if (!req) return -ENOMEM; @@ -44,20 +42,20 @@ static int mana_ib_cfg_vport_steering(struct mana_ib_dev *dev, req->rss_enable = true; req->num_indir_entries = MANA_INDIRECT_TABLE_SIZE; - req->indir_tab_offset = sizeof(*req); + req->indir_tab_offset = offsetof(struct mana_cfg_rx_steer_req_v2, + indir_tab); req->update_indir_tab = true; req->cqe_coalescing_enable = 1; - req_indir_tab = (mana_handle_t *)(req + 1); /* The ind table passed to the hardware must have * MANA_INDIRECT_TABLE_SIZE entries. Adjust the verb * ind_table to MANA_INDIRECT_TABLE_SIZE if required */ ibdev_dbg(&dev->ib_dev, "ind table size %u\n", 1 << log_ind_tbl_size); for (i = 0; i < MANA_INDIRECT_TABLE_SIZE; i++) { - req_indir_tab[i] = ind_table[i % (1 << log_ind_tbl_size)]; + req->indir_tab[i] = ind_table[i % (1 << log_ind_tbl_size)]; ibdev_dbg(&dev->ib_dev, "index %u handle 0x%llx\n", i, - req_indir_tab[i]); + req->indir_tab[i]); } req->update_hashkey = true; From a68292eb431619a5f8db9d4868346837c5606424 Mon Sep 17 00:00:00 2001 From: Erick Archer Date: Sat, 6 Apr 2024 16:23:37 +0200 Subject: [PATCH 16/16] net: mana: Avoid open coded arithmetic This is an effort to get rid of all multiplications from allocation functions in order to prevent integer overflows [1][2]. As the "req" variable is a pointer to "struct mana_cfg_rx_steer_req_v2" and this structure ends in a flexible array: struct mana_cfg_rx_steer_req_v2 { [...] mana_handle_t indir_tab[] __counted_by(num_indir_entries); }; the preferred way in the kernel is to use the struct_size() helper to do the arithmetic instead of the calculation "size + size * count" in the kzalloc() function. Moreover, use the "offsetof" helper to get the indirect table offset instead of the "sizeof" operator and avoid the open-coded arithmetic in pointers using the new flex member. This new structure member also allow us to remove the "req_indir_tab" variable since it is no longer needed. Now, it is also possible to use the "flex_array_size" helper to compute the size of these trailing elements in the "memcpy" function. This way, the code is more readable and safer. This code was detected with the help of Coccinelle, and audited and modified manually. Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments [1] Link: https://github.com/KSPP/linux/issues/160 [2] Signed-off-by: Erick Archer Link: https://lore.kernel.org/r/AS8PR02MB7237A21355C86EC0DCC0D83B8B022@AS8PR02MB7237.eurprd02.prod.outlook.com Reviewed-by: Justin Stitt Signed-off-by: Leon Romanovsky --- drivers/net/ethernet/microsoft/mana/mana_en.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/drivers/net/ethernet/microsoft/mana/mana_en.c b/drivers/net/ethernet/microsoft/mana/mana_en.c index 59287c6e6cee..62bf3e5661a6 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_en.c +++ b/drivers/net/ethernet/microsoft/mana/mana_en.c @@ -1058,11 +1058,10 @@ static int mana_cfg_vport_steering(struct mana_port_context *apc, struct mana_cfg_rx_steer_req_v2 *req; struct mana_cfg_rx_steer_resp resp = {}; struct net_device *ndev = apc->ndev; - mana_handle_t *req_indir_tab; u32 req_buf_size; int err; - req_buf_size = sizeof(*req) + sizeof(mana_handle_t) * num_entries; + req_buf_size = struct_size(req, indir_tab, num_entries); req = kzalloc(req_buf_size, GFP_KERNEL); if (!req) return -ENOMEM; @@ -1074,7 +1073,8 @@ static int mana_cfg_vport_steering(struct mana_port_context *apc, req->vport = apc->port_handle; req->num_indir_entries = num_entries; - req->indir_tab_offset = sizeof(*req); + req->indir_tab_offset = offsetof(struct mana_cfg_rx_steer_req_v2, + indir_tab); req->rx_enable = rx; req->rss_enable = apc->rss_state; req->update_default_rxobj = update_default_rxobj; @@ -1086,11 +1086,9 @@ static int mana_cfg_vport_steering(struct mana_port_context *apc, if (update_key) memcpy(&req->hashkey, apc->hashkey, MANA_HASH_KEY_SIZE); - if (update_tab) { - req_indir_tab = (mana_handle_t *)(req + 1); - memcpy(req_indir_tab, apc->rxobj_table, - req->num_indir_entries * sizeof(mana_handle_t)); - } + if (update_tab) + memcpy(req->indir_tab, apc->rxobj_table, + flex_array_size(req, indir_tab, req->num_indir_entries)); err = mana_send_request(apc->ac, req, req_buf_size, &resp, sizeof(resp));