From c83172b0639c8a005c0dd3b36252dc22ddd9f19c Mon Sep 17 00:00:00 2001 From: Gavin Li Date: Fri, 25 Nov 2022 04:15:40 +0200 Subject: [PATCH 1/7] net/mlx5e: Set uplink rep as NETNS_LOCAL Previously, NETNS_LOCAL was not set for uplink representors, inconsistent with VF representors, and allowed the uplink representor to be moved between net namespaces and separated from the VF representors it shares the core device with. Such usage would break the isolation model of namespaces, as devices in different namespaces would have access to shared memory. To solve this issue, set NETNS_LOCAL for uplink representors if eswitch is in switchdev mode. Fixes: 7a9fb35e8c3a ("net/mlx5e: Do not reload ethernet ports when changing eswitch mode") Signed-off-by: Gavin Li Reviewed-by: Gavi Teitz Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index a7f2ab22cc40..7ca7e9b57607 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -4150,8 +4150,12 @@ static netdev_features_t mlx5e_fix_features(struct net_device *netdev, } } - if (mlx5e_is_uplink_rep(priv)) + if (mlx5e_is_uplink_rep(priv)) { features = mlx5e_fix_uplink_rep_features(netdev, features); + features |= NETIF_F_NETNS_LOCAL; + } else { + features &= ~NETIF_F_NETNS_LOCAL; + } mutex_unlock(&priv->state_lock); From 662404b24a4c4d839839ed25e3097571f5938b9b Mon Sep 17 00:00:00 2001 From: Gavin Li Date: Thu, 9 Feb 2023 12:48:52 +0200 Subject: [PATCH 2/7] net/mlx5e: Block entering switchdev mode with ns inconsistency Upon entering switchdev mode, VF/SF representors are spawned in the devlink instance's net namespace, whereas the PF net device transforms into the uplink representor, remaining in the net namespace the PF net device was in. Therefore, if a PF net device's namespace is different from its parent devlink net namespace, entering switchdev mode can create an illegal situation where all representors sharing the same core device are NOT in the same net namespace. To avoid this issue, block entering switchdev mode for devices whose child netdev net namespace has diverged from the parent devlink's. Fixes: 7768d1971de6 ("net/mlx5: E-Switch, Add control for encapsulation") Signed-off-by: Gavin Li Reviewed-by: Gavi Teitz Signed-off-by: Saeed Mahameed --- .../mellanox/mlx5/core/eswitch_offloads.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 22075943bb58..25a8076a77bf 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -3405,6 +3405,18 @@ static int esw_inline_mode_to_devlink(u8 mlx5_mode, u8 *mode) return 0; } +static bool esw_offloads_devlink_ns_eq_netdev_ns(struct devlink *devlink) +{ + struct net *devl_net, *netdev_net; + struct mlx5_eswitch *esw; + + esw = mlx5_devlink_eswitch_get(devlink); + netdev_net = dev_net(esw->dev->mlx5e_res.uplink_netdev); + devl_net = devlink_net(devlink); + + return net_eq(devl_net, netdev_net); +} + int mlx5_devlink_eswitch_mode_set(struct devlink *devlink, u16 mode, struct netlink_ext_ack *extack) { @@ -3419,6 +3431,13 @@ int mlx5_devlink_eswitch_mode_set(struct devlink *devlink, u16 mode, if (esw_mode_from_devlink(mode, &mlx5_mode)) return -EINVAL; + if (mode == DEVLINK_ESWITCH_MODE_SWITCHDEV && + !esw_offloads_devlink_ns_eq_netdev_ns(devlink)) { + NL_SET_ERR_MSG_MOD(extack, + "Can't change E-Switch mode to switchdev when netdev net namespace has diverged from the devlink's."); + return -EPERM; + } + mlx5_lag_disable_change(esw->dev); err = mlx5_esw_try_lock(esw); if (err < 0) { From 922f56e9a795d6f3dd72d3428ebdd7ee040fa855 Mon Sep 17 00:00:00 2001 From: Lama Kayal Date: Tue, 31 Jan 2023 14:07:03 +0200 Subject: [PATCH 3/7] net/mlx5: Fix steering rules cleanup vport's mc, uc and multicast rules are not deleted in teardown path when EEH happens. Since the vport's promisc settings(uc, mc and all) in firmware are reset after EEH, mlx5 driver will try to delete the above rules in the initialization path. This cause kernel crash because these software rules are no longer valid. Fix by nullifying these rules right after delete to avoid accessing any dangling pointers. Call Trace: __list_del_entry_valid+0xcc/0x100 (unreliable) tree_put_node+0xf4/0x1b0 [mlx5_core] tree_remove_node+0x30/0x70 [mlx5_core] mlx5_del_flow_rules+0x14c/0x1f0 [mlx5_core] esw_apply_vport_rx_mode+0x10c/0x200 [mlx5_core] esw_update_vport_rx_mode+0xb4/0x180 [mlx5_core] esw_vport_change_handle_locked+0x1ec/0x230 [mlx5_core] esw_enable_vport+0x130/0x260 [mlx5_core] mlx5_eswitch_enable_sriov+0x2a0/0x2f0 [mlx5_core] mlx5_device_enable_sriov+0x74/0x440 [mlx5_core] mlx5_load_one+0x114c/0x1550 [mlx5_core] mlx5_pci_resume+0x68/0xf0 [mlx5_core] eeh_report_resume+0x1a4/0x230 eeh_pe_dev_traverse+0x98/0x170 eeh_handle_normal_event+0x3e4/0x640 eeh_handle_event+0x4c/0x370 eeh_event_handler+0x14c/0x210 kthread+0x168/0x1b0 ret_from_kernel_thread+0x5c/0x84 Fixes: a35f71f27a61 ("net/mlx5: E-Switch, Implement promiscuous rx modes vf request handling") Signed-off-by: Huy Nguyen Signed-off-by: Lama Kayal Reviewed-by: Tariq Toukan Reviewed-by: Maor Dickman Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index 0f052513fefa..8bdf28762f41 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -959,6 +959,7 @@ void mlx5_esw_vport_disable(struct mlx5_eswitch *esw, u16 vport_num) */ esw_vport_change_handle_locked(vport); vport->enabled_events = 0; + esw_apply_vport_rx_mode(esw, vport, false, false); esw_vport_cleanup(esw, vport); esw->enabled_vports--; From 6e9d51b1a5cb8d750c3daf89db4f4cdfd1051819 Mon Sep 17 00:00:00 2001 From: Roy Novich Date: Wed, 1 Mar 2023 15:47:11 +0200 Subject: [PATCH 4/7] net/mlx5e: Initialize link speed to zero mlx5e_port_max_linkspeed does not guarantee value assignment for speed. Avoid cases where link_speed might be used uninitialized. In case mlx5e_port_max_linkspeed fails, a default link speed of 50000 will be used for the calculations. Fixes: 3f6d08d196b2 ("net/mlx5e: Add RSS support for hairpin") Signed-off-by: Roy Novich Reviewed-by: Tariq Toukan Reviewed-by: Aya Levin Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 6bfed633343a..87a2850b32d0 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -1103,8 +1103,8 @@ static void mlx5e_hairpin_params_init(struct mlx5e_hairpin_params *hairpin_params, struct mlx5_core_dev *mdev) { + u32 link_speed = 0; u64 link_speed64; - u32 link_speed; hairpin_params->mdev = mdev; /* set hairpin pair per each 50Gbs share of the link */ From 7e3fce82d945cf6e7f99034b113ff2d250d7524d Mon Sep 17 00:00:00 2001 From: Emeel Hakim Date: Mon, 20 Mar 2023 13:13:55 +0200 Subject: [PATCH 5/7] net/mlx5e: Overcome slow response for first macsec ASO WQE MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit First ASO WQE poll causes a cache miss in hardware hence the resut is delayed. It causes to the situation where such WQE is polled earlier than it is needed. Add logic to retry ASO CQ polling operation. Fixes: 739cfa34518e ("net/mlx5: Make ASO poll CQ usable in atomic context")  Signed-off-by: Emeel Hakim Reviewed-by: Leon Romanovsky Reviewed-by: Raed Salem Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/en_accel/macsec.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c index 8af53178e40d..33b3620ea45c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c @@ -1412,6 +1412,7 @@ static int macsec_aso_query(struct mlx5_core_dev *mdev, struct mlx5e_macsec *mac struct mlx5e_macsec_aso *aso; struct mlx5_aso_wqe *aso_wqe; struct mlx5_aso *maso; + unsigned long expires; int err; aso = &macsec->aso; @@ -1425,7 +1426,13 @@ static int macsec_aso_query(struct mlx5_core_dev *mdev, struct mlx5e_macsec *mac macsec_aso_build_wqe_ctrl_seg(aso, &aso_wqe->aso_ctrl, NULL); mlx5_aso_post_wqe(maso, false, &aso_wqe->ctrl); - err = mlx5_aso_poll_cq(maso, false); + expires = jiffies + msecs_to_jiffies(10); + do { + err = mlx5_aso_poll_cq(maso, false); + if (err) + usleep_range(2, 10); + } while (err && time_is_after_jiffies(expires)); + if (err) goto err_out; From 44d553188c38ac74b799dfdcebafef2f7bb70942 Mon Sep 17 00:00:00 2001 From: Maher Sanalla Date: Wed, 15 Mar 2023 11:04:38 +0200 Subject: [PATCH 6/7] net/mlx5: Read the TC mapping of all priorities on ETS query When ETS configurations are queried by the user to get the mapping assignment between packet priority and traffic class, only priorities up to maximum TCs are queried from QTCT register in FW to retrieve their assigned TC, leaving the rest of the priorities mapped to the default TC #0 which might be misleading. Fix by querying the TC mapping of all priorities on each ETS query, regardless of the maximum number of TCs configured in FW. Fixes: 820c2c5e773d ("net/mlx5e: Read ETS settings directly from firmware") Signed-off-by: Maher Sanalla Reviewed-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c b/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c index 2449731b7d79..89de92d06483 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c @@ -117,12 +117,14 @@ static int mlx5e_dcbnl_ieee_getets(struct net_device *netdev, if (!MLX5_CAP_GEN(priv->mdev, ets)) return -EOPNOTSUPP; - ets->ets_cap = mlx5_max_tc(priv->mdev) + 1; - for (i = 0; i < ets->ets_cap; i++) { + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { err = mlx5_query_port_prio_tc(mdev, i, &ets->prio_tc[i]); if (err) return err; + } + ets->ets_cap = mlx5_max_tc(priv->mdev) + 1; + for (i = 0; i < ets->ets_cap; i++) { err = mlx5_query_port_tc_group(mdev, i, &tc_group[i]); if (err) return err; From 640fcdbcf27fc62de9223f958ceb4e897a00e791 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Mon, 27 Feb 2023 14:16:10 +0300 Subject: [PATCH 7/7] net/mlx5: E-Switch, Fix an Oops in error handling code The error handling dereferences "vport". There is nothing we can do if it is an error pointer except returning the error code. Fixes: 133dcfc577ea ("net/mlx5: E-Switch, Alloc and free unique metadata for match") Signed-off-by: Dan Carpenter Reviewed-by: Roi Dayan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/esw/acl/ingress_ofld.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/acl/ingress_ofld.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/acl/ingress_ofld.c index d55775627a47..50d2ea323979 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/acl/ingress_ofld.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/acl/ingress_ofld.c @@ -364,8 +364,7 @@ int mlx5_esw_acl_ingress_vport_metadata_update(struct mlx5_eswitch *esw, u16 vpo if (WARN_ON_ONCE(IS_ERR(vport))) { esw_warn(esw->dev, "vport(%d) invalid!\n", vport_num); - err = PTR_ERR(vport); - goto out; + return PTR_ERR(vport); } esw_acl_ingress_ofld_rules_destroy(esw, vport);