From 31057979cdadfee9f934746fd84046b43506ba61 Mon Sep 17 00:00:00 2001 From: Patrisious Haddad Date: Thu, 25 Dec 2025 15:27:13 +0200 Subject: [PATCH 1/5] net/mlx5: Lag, multipath, give priority for routes with smaller network prefix Today multipath offload is controlled by a single route and the route controlling is selected if it meets one of the following criteria: 1. No controlling route is set. 2. New route destination is the same as old one. 3. New route metric is lower than old route metric. This can cause unwanted behaviour in case a new route is added with a smaller network prefix which should get the priority. Fix this by adding a new criteria to give priority to new route with a smaller network prefix. Fixes: ad11c4f1d8fd ("net/mlx5e: Lag, Only handle events from highest priority multipath entry") Signed-off-by: Patrisious Haddad Signed-off-by: Mark Bloch Link: https://patch.msgid.link/20251225132717.358820-2-mbloch@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/lag/mp.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag/mp.c b/drivers/net/ethernet/mellanox/mlx5/core/lag/mp.c index aee17fcf3b36..cdc99fe5c956 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lag/mp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lag/mp.c @@ -173,10 +173,15 @@ static void mlx5_lag_fib_route_event(struct mlx5_lag *ldev, unsigned long event, } /* Handle multipath entry with lower priority value */ - if (mp->fib.mfi && mp->fib.mfi != fi && + if (mp->fib.mfi && (mp->fib.dst != fen_info->dst || mp->fib.dst_len != fen_info->dst_len) && - fi->fib_priority >= mp->fib.priority) + mp->fib.dst_len <= fen_info->dst_len && + !(mp->fib.dst_len == fen_info->dst_len && + fi->fib_priority < mp->fib.priority)) { + mlx5_core_dbg(ldev->pf[idx].dev, + "Multipath entry with lower priority was rejected\n"); return; + } nh_dev0 = mlx5_lag_get_next_fib_dev(ldev, fi, NULL); nh_dev1 = mlx5_lag_get_next_fib_dev(ldev, fi, nh_dev0); From 6c75dc9de40ff91ec2b621b78f6cd9031762067c Mon Sep 17 00:00:00 2001 From: Alexei Lazar Date: Thu, 25 Dec 2025 15:27:14 +0200 Subject: [PATCH 2/5] net/mlx5e: Don't gate FEC histograms on ppcnt_statistical_group Currently, the ppcnt_statistical_group capability check incorrectly gates access to FEC histogram statistics. This capability applies only to statistical and physical counter groups, not for histogram data. Restrict the ppcnt_statistical_group check to the Physical_Layer_Counters and Physical_Layer_Statistical_Counters groups. Histogram statistics access remains gated by the pphcr capability. The issue is harmless as of today, as it happens that ppcnt_statistical_group is set on all existing devices that have pphcr set. Fixes: 6b81b8a0b197 ("net/mlx5e: Don't query FEC statistics when FEC is disabled") Signed-off-by: Alexei Lazar Reviewed-by: Tariq Toukan Signed-off-by: Mark Bloch Link: https://patch.msgid.link/20251225132717.358820-3-mbloch@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en_stats.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c index a2802cfc9b98..a8af84fc9763 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c @@ -1608,12 +1608,13 @@ void mlx5e_stats_fec_get(struct mlx5e_priv *priv, { int mode = fec_active_mode(priv->mdev); - if (mode == MLX5E_FEC_NOFEC || - !MLX5_CAP_PCAM_FEATURE(priv->mdev, ppcnt_statistical_group)) + if (mode == MLX5E_FEC_NOFEC) return; - fec_set_corrected_bits_total(priv, fec_stats); - fec_set_block_stats(priv, mode, fec_stats); + if (MLX5_CAP_PCAM_FEATURE(priv->mdev, ppcnt_statistical_group)) { + fec_set_corrected_bits_total(priv, fec_stats); + fec_set_block_stats(priv, mode, fec_stats); + } if (MLX5_CAP_PCAM_REG(priv->mdev, pphcr)) fec_set_histograms_stats(priv, mode, hist); From 7d36a4a8bf62dc508bc6bb4b59727aec25064ca5 Mon Sep 17 00:00:00 2001 From: Gal Pressman Date: Thu, 25 Dec 2025 15:27:15 +0200 Subject: [PATCH 3/5] net/mlx5e: Fix NULL pointer dereference in ioctl module EEPROM query The mlx5_query_mcia() function unconditionally dereferences the status pointer to store the MCIA register status value. However, mlx5e_get_module_id() passes NULL since it doesn't need the status value. Add a NULL check before dereferencing the status pointer to prevent a NULL pointer dereference. Fixes: 2e4c44b12f4d ("net/mlx5: Refactor EEPROM query error handling to return status separately") Signed-off-by: Gal Pressman Reviewed-by: Tariq Toukan Reviewed-by: Dragos Tatulea Signed-off-by: Mark Bloch Link: https://patch.msgid.link/20251225132717.358820-4-mbloch@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/port.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/port.c b/drivers/net/ethernet/mellanox/mlx5/core/port.c index 85a9e534f442..8f36454dd196 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/port.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/port.c @@ -393,9 +393,11 @@ static int mlx5_query_mcia(struct mlx5_core_dev *dev, if (err) return err; - *status = MLX5_GET(mcia_reg, out, status); - if (*status) + if (MLX5_GET(mcia_reg, out, status)) { + if (status) + *status = MLX5_GET(mcia_reg, out, status); return -EIO; + } ptr = MLX5_ADDR_OF(mcia_reg, out, dword_0); memcpy(data, ptr, size); From 144297e2a24e3e54aee1180ec21120ea38822b97 Mon Sep 17 00:00:00 2001 From: Gal Pressman Date: Thu, 25 Dec 2025 15:27:16 +0200 Subject: [PATCH 4/5] net/mlx5e: Don't print error message due to invalid module Dumping module EEPROM on newer modules is supported through the netlink interface only. Querying with old userspace ethtool (or other tools, such as 'lshw') which still uses the ioctl interface results in an error message that could flood dmesg (in addition to the expected error return value). The original message was added under the assumption that the driver should be able to handle all module types, but now that such flows are easily triggered from userspace, it doesn't serve its purpose. Change the log level of the print in mlx5_query_module_eeprom() to debug. Fixes: bb64143eee8c ("net/mlx5e: Add ethtool support for dump module EEPROM") Signed-off-by: Gal Pressman Reviewed-by: Tariq Toukan Signed-off-by: Mark Bloch Link: https://patch.msgid.link/20251225132717.358820-5-mbloch@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/port.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/port.c b/drivers/net/ethernet/mellanox/mlx5/core/port.c index 8f36454dd196..7f8bed353e67 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/port.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/port.c @@ -431,7 +431,8 @@ int mlx5_query_module_eeprom(struct mlx5_core_dev *dev, mlx5_qsfp_eeprom_params_set(&query.i2c_address, &query.page, &offset); break; default: - mlx5_core_err(dev, "Module ID not recognized: 0x%x\n", module_id); + mlx5_core_dbg(dev, "Module ID not recognized: 0x%x\n", + module_id); return -EINVAL; } From 0462a15d2d1fafd3d48cf3c7c67393e42d03908c Mon Sep 17 00:00:00 2001 From: Cosmin Ratiu Date: Thu, 25 Dec 2025 15:27:17 +0200 Subject: [PATCH 5/5] net/mlx5e: Dealloc forgotten PSP RX modify header The commit which added RX steering rules for PSP forgot to free a modify header HW object on the cleanup path, which lead to health errors when reloading the driver and uninitializing the device: mlx5_core 0000:08:00.0: poll_health:803:(pid 3021): Fatal error 3 detected Fix that by saving the modify header pointer in the PSP steering struct and deallocating it after freeing the rule which references it. Fixes: 9536fbe10c9d ("net/mlx5e: Add PSP steering in local NIC RX") Signed-off-by: Cosmin Ratiu Reviewed-by: Dragos Tatulea Reviewed-by: Tariq Toukan Signed-off-by: Mark Bloch Link: https://patch.msgid.link/20251225132717.358820-6-mbloch@nvidia.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/mellanox/mlx5/core/en_accel/psp.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c index 38e7c77cc851..9a74438ce10a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c @@ -44,6 +44,7 @@ struct mlx5e_accel_fs_psp_prot { struct mlx5_flow_table *ft; struct mlx5_flow_group *miss_group; struct mlx5_flow_handle *miss_rule; + struct mlx5_modify_hdr *rx_modify_hdr; struct mlx5_flow_destination default_dest; struct mlx5e_psp_rx_err rx_err; u32 refcnt; @@ -286,13 +287,19 @@ static int accel_psp_fs_rx_err_create_ft(struct mlx5e_psp_fs *fs, return err; } -static void accel_psp_fs_rx_fs_destroy(struct mlx5e_accel_fs_psp_prot *fs_prot) +static void accel_psp_fs_rx_fs_destroy(struct mlx5e_psp_fs *fs, + struct mlx5e_accel_fs_psp_prot *fs_prot) { if (fs_prot->def_rule) { mlx5_del_flow_rules(fs_prot->def_rule); fs_prot->def_rule = NULL; } + if (fs_prot->rx_modify_hdr) { + mlx5_modify_header_dealloc(fs->mdev, fs_prot->rx_modify_hdr); + fs_prot->rx_modify_hdr = NULL; + } + if (fs_prot->miss_rule) { mlx5_del_flow_rules(fs_prot->miss_rule); fs_prot->miss_rule = NULL; @@ -396,6 +403,7 @@ static int accel_psp_fs_rx_create_ft(struct mlx5e_psp_fs *fs, modify_hdr = NULL; goto out_err; } + fs_prot->rx_modify_hdr = modify_hdr; flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST | MLX5_FLOW_CONTEXT_ACTION_CRYPTO_DECRYPT | @@ -416,7 +424,7 @@ static int accel_psp_fs_rx_create_ft(struct mlx5e_psp_fs *fs, goto out; out_err: - accel_psp_fs_rx_fs_destroy(fs_prot); + accel_psp_fs_rx_fs_destroy(fs, fs_prot); out: kvfree(flow_group_in); kvfree(spec); @@ -433,7 +441,7 @@ static int accel_psp_fs_rx_destroy(struct mlx5e_psp_fs *fs, enum accel_fs_psp_ty /* The netdev unreg already happened, so all offloaded rule are already removed */ fs_prot = &accel_psp->fs_prot[type]; - accel_psp_fs_rx_fs_destroy(fs_prot); + accel_psp_fs_rx_fs_destroy(fs, fs_prot); accel_psp_fs_rx_err_destroy_ft(fs, &fs_prot->rx_err);