[PATCH net 0/4] mlx5 misc fixes 2025-10-22

[PATCH net 0/4] mlx5 misc fixes 2025-10-22

[Tariq Toukan]

Hi,

This patchset provides misc bug fixes from the team to the mlx5 core and
Eth drivers.

Thanks,
Tariq.


Alexei Lazar (2):
  net/mlx5: Add PPHCR to PCAM supported registers mask
  net/mlx5e: Skip PPHCR register query if not supported by the device

Patrisious Haddad (2):
  net/mlx5: Refactor devcom to return NULL on failure
  net/mlx5: Fix IPsec cleanup over MPV device

 .../mellanox/mlx5/core/en_accel/ipsec.h       |  5 ++
 .../mellanox/mlx5/core/en_accel/ipsec_fs.c    | 25 ++++++++-
 .../net/ethernet/mellanox/mlx5/core/en_main.c |  8 +--
 .../ethernet/mellanox/mlx5/core/en_stats.c    |  4 +-
 .../mellanox/mlx5/core/eswitch_offloads.c     |  4 +-
 .../net/ethernet/mellanox/mlx5/core/lag/lag.c |  7 ++-
 .../ethernet/mellanox/mlx5/core/lib/clock.c   |  2 +-
 .../ethernet/mellanox/mlx5/core/lib/devcom.c  | 53 +++++++++----------
 .../net/ethernet/mellanox/mlx5/core/lib/sd.c  |  4 +-
 .../net/ethernet/mellanox/mlx5/core/main.c    |  5 +-
 include/linux/mlx5/mlx5_ifc.h                 |  4 +-
 11 files changed, 75 insertions(+), 46 deletions(-)


base-commit: d63f0391d6c7b75e1a847e1a26349fa8cad0004d
-- 
2.31.1

[PATCH net 1/4] net/mlx5: Add PPHCR to PCAM supported registers mask

[Tariq Toukan]

From: Alexei Lazar <alazar@nvidia.com>

Add the PPHCR bit to the port_access_reg_cap_mask field of PCAM
register to indicate that the device supports the PPHCR register
and the RS-FEC histogram feature.

Signed-off-by: Alexei Lazar <alazar@nvidia.com>
Reviewed-by: Yael Chemla <ychemla@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
---
 include/linux/mlx5/mlx5_ifc.h | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 07614cd95bed..1b0b36aa2a76 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -10833,7 +10833,9 @@ struct mlx5_ifc_pcam_regs_5000_to_507f_bits {
 	u8         port_access_reg_cap_mask_127_to_96[0x20];
 	u8         port_access_reg_cap_mask_95_to_64[0x20];
 
-	u8         port_access_reg_cap_mask_63_to_36[0x1c];
+	u8         port_access_reg_cap_mask_63[0x1];
+	u8         pphcr[0x1];
+	u8         port_access_reg_cap_mask_61_to_36[0x1a];
 	u8         pplm[0x1];
 	u8         port_access_reg_cap_mask_34_to_32[0x3];
 
-- 
2.31.1

[PATCH net 2/4] net/mlx5e: Skip PPHCR register query if not supported by the device

[Tariq Toukan]

From: Alexei Lazar <alazar@nvidia.com>

Check the PCAM supported registers mask before querying the PPHCR
register, as it is not supported in older devices.

Fixes: 44907e7c8fd0 ("net/mlx5e: Add logic to read RS-FEC histogram bin ranges from PPHCR")
Signed-off-by: Alexei Lazar <alazar@nvidia.com>
Reviewed-by: Yael Chemla <ychemla@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_stats.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
index 7c029a7d0fd7..a2802cfc9b98 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
@@ -1614,7 +1614,9 @@ void mlx5e_stats_fec_get(struct mlx5e_priv *priv,
 
 	fec_set_corrected_bits_total(priv, fec_stats);
 	fec_set_block_stats(priv, mode, fec_stats);
-	fec_set_histograms_stats(priv, mode, hist);
+
+	if (MLX5_CAP_PCAM_REG(priv->mdev, pphcr))
+		fec_set_histograms_stats(priv, mode, hist);
 }
 
 #define PPORT_ETH_EXT_OFF(c) \
-- 
2.31.1

[PATCH net 3/4] net/mlx5: Refactor devcom to return NULL on failure

[Tariq Toukan]

From: Patrisious Haddad <phaddad@nvidia.com>

Devcom device and component registration isn't always critical to the
functionality of the caller, hence the registration can fail and we can
continue working with an ERR_PTR value saved inside a variable.

In order to avoid that make sure all devcom failures return NULL.

Signed-off-by: Patrisious Haddad <phaddad@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
---
 .../net/ethernet/mellanox/mlx5/core/en_main.c |  6 +--
 .../mellanox/mlx5/core/eswitch_offloads.c     |  4 +-
 .../net/ethernet/mellanox/mlx5/core/lag/lag.c |  7 ++-
 .../ethernet/mellanox/mlx5/core/lib/clock.c   |  2 +-
 .../ethernet/mellanox/mlx5/core/lib/devcom.c  | 53 +++++++++----------
 .../net/ethernet/mellanox/mlx5/core/lib/sd.c  |  4 +-
 .../net/ethernet/mellanox/mlx5/core/main.c    |  5 +-
 7 files changed, 39 insertions(+), 42 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index a56825921c23..41fd5eee6306 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -242,8 +242,8 @@ static int mlx5e_devcom_init_mpv(struct mlx5e_priv *priv, u64 *data)
 						      &attr,
 						      mlx5e_devcom_event_mpv,
 						      priv);
-	if (IS_ERR(priv->devcom))
-		return PTR_ERR(priv->devcom);
+	if (!priv->devcom)
+		return -EINVAL;
 
 	if (mlx5_core_is_mp_master(priv->mdev)) {
 		mlx5_devcom_send_event(priv->devcom, MPV_DEVCOM_MASTER_UP,
@@ -256,7 +256,7 @@ static int mlx5e_devcom_init_mpv(struct mlx5e_priv *priv, u64 *data)
 
 static void mlx5e_devcom_cleanup_mpv(struct mlx5e_priv *priv)
 {
-	if (IS_ERR_OR_NULL(priv->devcom))
+	if (!priv->devcom)
 		return;
 
 	if (mlx5_core_is_mp_master(priv->mdev)) {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index 4cf995be127d..34749814f19b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -3129,7 +3129,7 @@ void mlx5_esw_offloads_devcom_init(struct mlx5_eswitch *esw,
 						     attr,
 						     mlx5_esw_offloads_devcom_event,
 						     esw);
-	if (IS_ERR(esw->devcom))
+	if (!esw->devcom)
 		return;
 
 	mlx5_devcom_send_event(esw->devcom,
@@ -3140,7 +3140,7 @@ void mlx5_esw_offloads_devcom_init(struct mlx5_eswitch *esw,
 
 void mlx5_esw_offloads_devcom_cleanup(struct mlx5_eswitch *esw)
 {
-	if (IS_ERR_OR_NULL(esw->devcom))
+	if (!esw->devcom)
 		return;
 
 	mlx5_devcom_send_event(esw->devcom,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c b/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c
index 59c00c911275..3db0387bf6dc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c
@@ -1430,11 +1430,10 @@ static int mlx5_lag_register_hca_devcom_comp(struct mlx5_core_dev *dev)
 		mlx5_devcom_register_component(dev->priv.devc,
 					       MLX5_DEVCOM_HCA_PORTS,
 					       &attr, NULL, dev);
-	if (IS_ERR(dev->priv.hca_devcom_comp)) {
+	if (!dev->priv.hca_devcom_comp) {
 		mlx5_core_err(dev,
-			      "Failed to register devcom HCA component, err: %ld\n",
-			      PTR_ERR(dev->priv.hca_devcom_comp));
-		return PTR_ERR(dev->priv.hca_devcom_comp);
+			      "Failed to register devcom HCA component.");
+		return -EINVAL;
 	}
 
 	return 0;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c
index d0ba83d77cd1..29e7fa09c32c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c
@@ -1444,7 +1444,7 @@ static void mlx5_shared_clock_register(struct mlx5_core_dev *mdev, u64 key)
 	compd = mlx5_devcom_register_component(mdev->priv.devc,
 					       MLX5_DEVCOM_SHARED_CLOCK,
 					       &attr, NULL, mdev);
-	if (IS_ERR(compd))
+	if (!compd)
 		return;
 
 	mdev->clock_state->compdev = compd;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c
index faa2833602c8..e749618229bc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c
@@ -76,20 +76,18 @@ mlx5_devcom_dev_alloc(struct mlx5_core_dev *dev)
 struct mlx5_devcom_dev *
 mlx5_devcom_register_device(struct mlx5_core_dev *dev)
 {
-	struct mlx5_devcom_dev *devc;
+	struct mlx5_devcom_dev *devc = NULL;
 
 	mutex_lock(&dev_list_lock);
 
 	if (devcom_dev_exists(dev)) {
-		devc = ERR_PTR(-EEXIST);
+		mlx5_core_err(dev, "devcom device already exists");
 		goto out;
 	}
 
 	devc = mlx5_devcom_dev_alloc(dev);
-	if (!devc) {
-		devc = ERR_PTR(-ENOMEM);
+	if (!devc)
 		goto out;
-	}
 
 	list_add_tail(&devc->list, &devcom_dev_list);
 out:
@@ -110,8 +108,10 @@ mlx5_devcom_dev_release(struct kref *ref)
 
 void mlx5_devcom_unregister_device(struct mlx5_devcom_dev *devc)
 {
-	if (!IS_ERR_OR_NULL(devc))
-		kref_put(&devc->ref, mlx5_devcom_dev_release);
+	if (!devc)
+		return;
+
+	kref_put(&devc->ref, mlx5_devcom_dev_release);
 }
 
 static struct mlx5_devcom_comp *
@@ -122,7 +122,7 @@ mlx5_devcom_comp_alloc(u64 id, const struct mlx5_devcom_match_attr *attr,
 
 	comp = kzalloc(sizeof(*comp), GFP_KERNEL);
 	if (!comp)
-		return ERR_PTR(-ENOMEM);
+		return NULL;
 
 	comp->id = id;
 	comp->key.key = attr->key;
@@ -160,7 +160,7 @@ devcom_alloc_comp_dev(struct mlx5_devcom_dev *devc,
 
 	devcom = kzalloc(sizeof(*devcom), GFP_KERNEL);
 	if (!devcom)
-		return ERR_PTR(-ENOMEM);
+		return NULL;
 
 	kref_get(&devc->ref);
 	devcom->devc = devc;
@@ -240,31 +240,28 @@ mlx5_devcom_register_component(struct mlx5_devcom_dev *devc,
 			       mlx5_devcom_event_handler_t handler,
 			       void *data)
 {
-	struct mlx5_devcom_comp_dev *devcom;
+	struct mlx5_devcom_comp_dev *devcom = NULL;
 	struct mlx5_devcom_comp *comp;
 
-	if (IS_ERR_OR_NULL(devc))
-		return ERR_PTR(-EINVAL);
+	if (!devc)
+		return NULL;
 
 	mutex_lock(&comp_list_lock);
 	comp = devcom_component_get(devc, id, attr, handler);
-	if (IS_ERR(comp)) {
-		devcom = ERR_PTR(-EINVAL);
+	if (IS_ERR(comp))
 		goto out_unlock;
-	}
 
 	if (!comp) {
 		comp = mlx5_devcom_comp_alloc(id, attr, handler);
-		if (IS_ERR(comp)) {
-			devcom = ERR_CAST(comp);
+		if (!comp)
 			goto out_unlock;
-		}
+
 		list_add_tail(&comp->comp_list, &devcom_comp_list);
 	}
 	mutex_unlock(&comp_list_lock);
 
 	devcom = devcom_alloc_comp_dev(devc, comp, data);
-	if (IS_ERR(devcom))
+	if (!devcom)
 		kref_put(&comp->ref, mlx5_devcom_comp_release);
 
 	return devcom;
@@ -276,8 +273,10 @@ mlx5_devcom_register_component(struct mlx5_devcom_dev *devc,
 
 void mlx5_devcom_unregister_component(struct mlx5_devcom_comp_dev *devcom)
 {
-	if (!IS_ERR_OR_NULL(devcom))
-		devcom_free_comp_dev(devcom);
+	if (!devcom)
+		return;
+
+	devcom_free_comp_dev(devcom);
 }
 
 int mlx5_devcom_comp_get_size(struct mlx5_devcom_comp_dev *devcom)
@@ -296,7 +295,7 @@ int mlx5_devcom_send_event(struct mlx5_devcom_comp_dev *devcom,
 	int err = 0;
 	void *data;
 
-	if (IS_ERR_OR_NULL(devcom))
+	if (!devcom)
 		return -ENODEV;
 
 	comp = devcom->comp;
@@ -338,7 +337,7 @@ void mlx5_devcom_comp_set_ready(struct mlx5_devcom_comp_dev *devcom, bool ready)
 
 bool mlx5_devcom_comp_is_ready(struct mlx5_devcom_comp_dev *devcom)
 {
-	if (IS_ERR_OR_NULL(devcom))
+	if (!devcom)
 		return false;
 
 	return READ_ONCE(devcom->comp->ready);
@@ -348,7 +347,7 @@ bool mlx5_devcom_for_each_peer_begin(struct mlx5_devcom_comp_dev *devcom)
 {
 	struct mlx5_devcom_comp *comp;
 
-	if (IS_ERR_OR_NULL(devcom))
+	if (!devcom)
 		return false;
 
 	comp = devcom->comp;
@@ -421,21 +420,21 @@ void *mlx5_devcom_get_next_peer_data_rcu(struct mlx5_devcom_comp_dev *devcom,
 
 void mlx5_devcom_comp_lock(struct mlx5_devcom_comp_dev *devcom)
 {
-	if (IS_ERR_OR_NULL(devcom))
+	if (!devcom)
 		return;
 	down_write(&devcom->comp->sem);
 }
 
 void mlx5_devcom_comp_unlock(struct mlx5_devcom_comp_dev *devcom)
 {
-	if (IS_ERR_OR_NULL(devcom))
+	if (!devcom)
 		return;
 	up_write(&devcom->comp->sem);
 }
 
 int mlx5_devcom_comp_trylock(struct mlx5_devcom_comp_dev *devcom)
 {
-	if (IS_ERR_OR_NULL(devcom))
+	if (!devcom)
 		return 0;
 	return down_write_trylock(&devcom->comp->sem);
 }
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c
index f5c2701f6e87..8e17daae48af 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c
@@ -221,8 +221,8 @@ static int sd_register(struct mlx5_core_dev *dev)
 	attr.net = mlx5_core_net(dev);
 	devcom = mlx5_devcom_register_component(dev->priv.devc, MLX5_DEVCOM_SD_GROUP,
 						&attr, NULL, dev);
-	if (IS_ERR(devcom))
-		return PTR_ERR(devcom);
+	if (!devcom)
+		return -EINVAL;
 
 	sd->devcom = devcom;
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index df93625c9dfa..70c156591b0b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -978,9 +978,8 @@ static int mlx5_init_once(struct mlx5_core_dev *dev)
 	int err;
 
 	dev->priv.devc = mlx5_devcom_register_device(dev);
-	if (IS_ERR(dev->priv.devc))
-		mlx5_core_warn(dev, "failed to register devcom device %pe\n",
-			       dev->priv.devc);
+	if (!dev->priv.devc)
+		mlx5_core_warn(dev, "failed to register devcom device\n");
 
 	err = mlx5_query_board_id(dev);
 	if (err) {
-- 
2.31.1

[PATCH net 4/4] net/mlx5: Fix IPsec cleanup over MPV device

[Tariq Toukan]

From: Patrisious Haddad <phaddad@nvidia.com>

When we do mlx5e_detach_netdev() we eventually disable blocking events
notifier, among those events are IPsec MPV events from IB to core.

So before disabling those blocking events, make sure to also unregister
the devcom device and mark all this device operations as complete,
in order to prevent the other device from using invalid netdev
during future devcom events which could cause the trace below.

BUG: kernel NULL pointer dereference, address: 0000000000000010
PGD 146427067 P4D 146427067 PUD 146488067 PMD 0
Oops: Oops: 0000 [#1] SMP
CPU: 1 UID: 0 PID: 7735 Comm: devlink Tainted: GW 6.12.0-rc6_for_upstream_min_debug_2024_11_08_00_46 #1
Tainted: [W]=WARN
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
RIP: 0010:mlx5_devcom_comp_set_ready+0x5/0x40 [mlx5_core]
Code: 00 01 48 83 05 23 32 1e 00 01 41 b8 ed ff ff ff e9 60 ff ff ff 48 83 05 00 32 1e 00 01 eb e3 66 0f 1f 44 00 00 0f 1f 44 00 00 <48> 8b 47 10 48 83 05 5f 32 1e 00 01 48 8b 50 40 48 85 d2 74 05 40
RSP: 0018:ffff88811a5c35f8 EFLAGS: 00010206
RAX: ffff888106e8ab80 RBX: ffff888107d7e200 RCX: ffff88810d6f0a00
RDX: ffff88810d6f0a00 RSI: 0000000000000001 RDI: 0000000000000000
RBP: ffff88811a17e620 R08: 0000000000000040 R09: 0000000000000000
R10: ffff88811a5c3618 R11: 0000000de85d51bd R12: ffff88811a17e600
R13: ffff88810d6f0a00 R14: 0000000000000000 R15: ffff8881034bda80
FS:  00007f27bdf89180(0000) GS:ffff88852c880000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000010 CR3: 000000010f159005 CR4: 0000000000372eb0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
 ? __die+0x20/0x60
 ? page_fault_oops+0x150/0x3e0
 ? exc_page_fault+0x74/0x130
 ? asm_exc_page_fault+0x22/0x30
 ? mlx5_devcom_comp_set_ready+0x5/0x40 [mlx5_core]
 mlx5e_devcom_event_mpv+0x42/0x60 [mlx5_core]
 mlx5_devcom_send_event+0x8c/0x170 [mlx5_core]
 blocking_event+0x17b/0x230 [mlx5_core]
 notifier_call_chain+0x35/0xa0
 blocking_notifier_call_chain+0x3d/0x60
 mlx5_blocking_notifier_call_chain+0x22/0x30 [mlx5_core]
 mlx5_core_mp_event_replay+0x12/0x20 [mlx5_core]
 mlx5_ib_bind_slave_port+0x228/0x2c0 [mlx5_ib]
 mlx5_ib_stage_init_init+0x664/0x9d0 [mlx5_ib]
 ? idr_alloc_cyclic+0x50/0xb0
 ? __kmalloc_cache_noprof+0x167/0x340
 ? __kmalloc_noprof+0x1a7/0x430
 __mlx5_ib_add+0x34/0xd0 [mlx5_ib]
 mlx5r_probe+0xe9/0x310 [mlx5_ib]
 ? kernfs_add_one+0x107/0x150
 ? __mlx5_ib_add+0xd0/0xd0 [mlx5_ib]
 auxiliary_bus_probe+0x3e/0x90
 really_probe+0xc5/0x3a0
 ? driver_probe_device+0x90/0x90
 __driver_probe_device+0x80/0x160
 driver_probe_device+0x1e/0x90
 __device_attach_driver+0x7d/0x100
 bus_for_each_drv+0x80/0xd0
 __device_attach+0xbc/0x1f0
 bus_probe_device+0x86/0xa0
 device_add+0x62d/0x830
 __auxiliary_device_add+0x3b/0xa0
 ? auxiliary_device_init+0x41/0x90
 add_adev+0xd1/0x150 [mlx5_core]
 mlx5_rescan_drivers_locked+0x21c/0x300 [mlx5_core]
 esw_mode_change+0x6c/0xc0 [mlx5_core]
 mlx5_devlink_eswitch_mode_set+0x21e/0x640 [mlx5_core]
 devlink_nl_eswitch_set_doit+0x60/0xe0
 genl_family_rcv_msg_doit+0xd0/0x120
 genl_rcv_msg+0x180/0x2b0
 ? devlink_get_from_attrs_lock+0x170/0x170
 ? devlink_nl_eswitch_get_doit+0x290/0x290
 ? devlink_nl_pre_doit_port_optional+0x50/0x50
 ? genl_family_rcv_msg_dumpit+0xf0/0xf0
 netlink_rcv_skb+0x54/0x100
 genl_rcv+0x24/0x40
 netlink_unicast+0x1fc/0x2d0
 netlink_sendmsg+0x1e4/0x410
 __sock_sendmsg+0x38/0x60
 ? sockfd_lookup_light+0x12/0x60
 __sys_sendto+0x105/0x160
 ? __sys_recvmsg+0x4e/0x90
 __x64_sys_sendto+0x20/0x30
 do_syscall_64+0x4c/0x100
 entry_SYSCALL_64_after_hwframe+0x4b/0x53
RIP: 0033:0x7f27bc91b13a
Code: bb 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 8b 05 fa 96 2c 00 45 89 c9 4c 63 d1 48 63 ff 85 c0 75 15 b8 2c 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 76 f3 c3 0f 1f 40 00 41 55 41 54 4d 89 c5 55
RSP: 002b:00007fff369557e8 EFLAGS: 00000246 ORIG_RAX: 000000000000002c
RAX: ffffffffffffffda RBX: 0000000009c54b10 RCX: 00007f27bc91b13a
RDX: 0000000000000038 RSI: 0000000009c54b10 RDI: 0000000000000006
RBP: 0000000009c54920 R08: 00007f27bd0030e0 R09: 000000000000000c
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000001
 </TASK>
Modules linked in: mlx5_vdpa vringh vhost_iotlb vdpa xt_MASQUERADE nf_conntrack_netlink nfnetlink iptable_nat xt_addrtype xt_conntrack nf_nat br_netfilter rpcsec_gss_krb5 auth_rpcgss oid_registry overlay rpcrdma rdma_ucm ib_iser libiscsi ib_umad scsi_transport_iscsi ib_ipoib rdma_cm iw_cm ib_cm mlx5_fwctl mlx5_ib ib_uverbs ib_core mlx5_core
CR2: 0000000000000010

Fixes: 82f9378c443c ("net/mlx5: Handle IPsec steering upon master unbind/bind")
Signed-off-by: Patrisious Haddad <phaddad@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
---
 .../mellanox/mlx5/core/en_accel/ipsec.h       |  5 ++++
 .../mellanox/mlx5/core/en_accel/ipsec_fs.c    | 25 +++++++++++++++++--
 .../net/ethernet/mellanox/mlx5/core/en_main.c |  2 ++
 3 files changed, 30 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec.h b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec.h
index 5d7c15abfcaf..f8eaaf37963b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec.h
@@ -342,6 +342,7 @@ void mlx5e_ipsec_build_accel_xfrm_attrs(struct mlx5e_ipsec_sa_entry *sa_entry,
 void mlx5e_ipsec_handle_mpv_event(int event, struct mlx5e_priv *slave_priv,
 				  struct mlx5e_priv *master_priv);
 void mlx5e_ipsec_send_event(struct mlx5e_priv *priv, int event);
+void mlx5e_ipsec_disable_events(struct mlx5e_priv *priv);
 
 static inline struct mlx5_core_dev *
 mlx5e_ipsec_sa2dev(struct mlx5e_ipsec_sa_entry *sa_entry)
@@ -387,6 +388,10 @@ static inline void mlx5e_ipsec_handle_mpv_event(int event, struct mlx5e_priv *sl
 static inline void mlx5e_ipsec_send_event(struct mlx5e_priv *priv, int event)
 {
 }
+
+static inline void mlx5e_ipsec_disable_events(struct mlx5e_priv *priv)
+{
+}
 #endif
 
 #endif	/* __MLX5E_IPSEC_H__ */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
index bf1d2769d4f1..feef86fff4bf 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c
@@ -2893,9 +2893,30 @@ void mlx5e_ipsec_handle_mpv_event(int event, struct mlx5e_priv *slave_priv,
 
 void mlx5e_ipsec_send_event(struct mlx5e_priv *priv, int event)
 {
-	if (!priv->ipsec)
-		return; /* IPsec not supported */
+	if (!priv->ipsec || mlx5_devcom_comp_get_size(priv->devcom) < 2)
+		return; /* IPsec not supported or no peers */
 
 	mlx5_devcom_send_event(priv->devcom, event, event, priv);
 	wait_for_completion(&priv->ipsec->comp);
 }
+
+void mlx5e_ipsec_disable_events(struct mlx5e_priv *priv)
+{
+	struct mlx5_devcom_comp_dev *tmp = NULL;
+	struct mlx5e_priv *peer_priv;
+
+	if (!priv->devcom)
+		return;
+
+	if (!mlx5_devcom_for_each_peer_begin(priv->devcom))
+		goto out;
+
+	peer_priv = mlx5_devcom_get_next_peer_data(priv->devcom, &tmp);
+	if (peer_priv)
+		complete_all(&peer_priv->ipsec->comp);
+
+	mlx5_devcom_for_each_peer_end(priv->devcom);
+out:
+	mlx5_devcom_unregister_component(priv->devcom);
+	priv->devcom = NULL;
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 41fd5eee6306..9c46511e7b43 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -266,6 +266,7 @@ static void mlx5e_devcom_cleanup_mpv(struct mlx5e_priv *priv)
 	}
 
 	mlx5_devcom_unregister_component(priv->devcom);
+	priv->devcom = NULL;
 }
 
 static int blocking_event(struct notifier_block *nb, unsigned long event, void *data)
@@ -6120,6 +6121,7 @@ static void mlx5e_nic_disable(struct mlx5e_priv *priv)
 	if (mlx5e_monitor_counter_supported(priv))
 		mlx5e_monitor_counter_cleanup(priv);
 
+	mlx5e_ipsec_disable_events(priv);
 	mlx5e_disable_blocking_events(priv);
 	mlx5e_disable_async_events(priv);
 	mlx5_lag_remove_netdev(mdev, priv->netdev);
-- 
2.31.1

Re: [PATCH net 4/4] net/mlx5: Fix IPsec cleanup over MPV device

[Jakub Kicinski]

On Wed, 22 Oct 2025 15:29:42 +0300 Tariq Toukan wrote:
> BUG: kernel NULL pointer dereference, address: 0000000000000010
> PGD 146427067 P4D 146427067 PUD 146488067 PMD 0
> Oops: Oops: 0000 [#1] SMP
> CPU: 1 UID: 0 PID: 7735 Comm: devlink Tainted: GW 6.12.0-rc6_for_upstream_min_debug_2024_11_08_00_46 #1
> Tainted: [W]=WARN
> Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
> RIP: 0010:mlx5_devcom_comp_set_ready+0x5/0x40 [mlx5_core]
> Code: 00 01 48 83 05 23 32 1e 00 01 41 b8 ed ff ff ff e9 60 ff ff ff 48 83 05 00 32 1e 00 01 eb e3 66 0f 1f 44 00 00 0f 1f 44 00 00 <48> 8b 47 10 48 83 05 5f 32 1e 00 01 48 8b 50 40 48 85 d2 74 05 40
> RSP: 0018:ffff88811a5c35f8 EFLAGS: 00010206
> RAX: ffff888106e8ab80 RBX: ffff888107d7e200 RCX: ffff88810d6f0a00
> RDX: ffff88810d6f0a00 RSI: 0000000000000001 RDI: 0000000000000000
> RBP: ffff88811a17e620 R08: 0000000000000040 R09: 0000000000000000
> R10: ffff88811a5c3618 R11: 0000000de85d51bd R12: ffff88811a17e600
> R13: ffff88810d6f0a00 R14: 0000000000000000 R15: ffff8881034bda80
> FS:  00007f27bdf89180(0000) GS:ffff88852c880000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 0000000000000010 CR3: 000000010f159005 CR4: 0000000000372eb0
> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> Call Trace:
>  <TASK>
>  ? __die+0x20/0x60
>  ? page_fault_oops+0x150/0x3e0
>  ? exc_page_fault+0x74/0x130
>  ? asm_exc_page_fault+0x22/0x30
>  ? mlx5_devcom_comp_set_ready+0x5/0x40 [mlx5_core]
>  mlx5e_devcom_event_mpv+0x42/0x60 [mlx5_core]
>  mlx5_devcom_send_event+0x8c/0x170 [mlx5_core]
>  blocking_event+0x17b/0x230 [mlx5_core]
>  notifier_call_chain+0x35/0xa0
>  blocking_notifier_call_chain+0x3d/0x60
>  mlx5_blocking_notifier_call_chain+0x22/0x30 [mlx5_core]
>  mlx5_core_mp_event_replay+0x12/0x20 [mlx5_core]
>  mlx5_ib_bind_slave_port+0x228/0x2c0 [mlx5_ib]
>  mlx5_ib_stage_init_init+0x664/0x9d0 [mlx5_ib]
>  ? idr_alloc_cyclic+0x50/0xb0
>  ? __kmalloc_cache_noprof+0x167/0x340
>  ? __kmalloc_noprof+0x1a7/0x430
>  __mlx5_ib_add+0x34/0xd0 [mlx5_ib]
>  mlx5r_probe+0xe9/0x310 [mlx5_ib]
>  ? kernfs_add_one+0x107/0x150
>  ? __mlx5_ib_add+0xd0/0xd0 [mlx5_ib]
>  auxiliary_bus_probe+0x3e/0x90
>  really_probe+0xc5/0x3a0
>  ? driver_probe_device+0x90/0x90
>  __driver_probe_device+0x80/0x160
>  driver_probe_device+0x1e/0x90
>  __device_attach_driver+0x7d/0x100
>  bus_for_each_drv+0x80/0xd0
>  __device_attach+0xbc/0x1f0
>  bus_probe_device+0x86/0xa0
>  device_add+0x62d/0x830
>  __auxiliary_device_add+0x3b/0xa0
>  ? auxiliary_device_init+0x41/0x90
>  add_adev+0xd1/0x150 [mlx5_core]
>  mlx5_rescan_drivers_locked+0x21c/0x300 [mlx5_core]
>  esw_mode_change+0x6c/0xc0 [mlx5_core]
>  mlx5_devlink_eswitch_mode_set+0x21e/0x640 [mlx5_core]
>  devlink_nl_eswitch_set_doit+0x60/0xe0
>  genl_family_rcv_msg_doit+0xd0/0x120
>  genl_rcv_msg+0x180/0x2b0
>  ? devlink_get_from_attrs_lock+0x170/0x170
>  ? devlink_nl_eswitch_get_doit+0x290/0x290
>  ? devlink_nl_pre_doit_port_optional+0x50/0x50
>  ? genl_family_rcv_msg_dumpit+0xf0/0xf0
>  netlink_rcv_skb+0x54/0x100
>  genl_rcv+0x24/0x40
>  netlink_unicast+0x1fc/0x2d0
>  netlink_sendmsg+0x1e4/0x410
>  __sock_sendmsg+0x38/0x60
>  ? sockfd_lookup_light+0x12/0x60
>  __sys_sendto+0x105/0x160
>  ? __sys_recvmsg+0x4e/0x90
>  __x64_sys_sendto+0x20/0x30
>  do_syscall_64+0x4c/0x100
>  entry_SYSCALL_64_after_hwframe+0x4b/0x53
> RIP: 0033:0x7f27bc91b13a
> Code: bb 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 8b 05 fa 96 2c 00 45 89 c9 4c 63 d1 48 63 ff 85 c0 75 15 b8 2c 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 76 f3 c3 0f 1f 40 00 41 55 41 54 4d 89 c5 55
> RSP: 002b:00007fff369557e8 EFLAGS: 00000246 ORIG_RAX: 000000000000002c
> RAX: ffffffffffffffda RBX: 0000000009c54b10 RCX: 00007f27bc91b13a
> RDX: 0000000000000038 RSI: 0000000009c54b10 RDI: 0000000000000006
> RBP: 0000000009c54920 R08: 00007f27bd0030e0 R09: 000000000000000c
> R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
> R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000001
>  </TASK>
> Modules linked in: mlx5_vdpa vringh vhost_iotlb vdpa xt_MASQUERADE nf_conntrack_netlink nfnetlink iptable_nat xt_addrtype xt_conntrack nf_nat br_netfilter rpcsec_gss_krb5 auth_rpcgss oid_registry overlay rpcrdma rdma_ucm ib_iser libiscsi ib_umad scsi_transport_iscsi ib_ipoib rdma_cm iw_cm ib_cm mlx5_fwctl mlx5_ib ib_uverbs ib_core mlx5_core
> CR2: 0000000000000010

Please trim the crashes in the future.

> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
> index 41fd5eee6306..9c46511e7b43 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
> @@ -266,6 +266,7 @@ static void mlx5e_devcom_cleanup_mpv(struct mlx5e_priv *priv)
>  	}
>  
>  	mlx5_devcom_unregister_component(priv->devcom);
> +	priv->devcom = NULL;
>  }

This feels a little like it should be in patch 3. 
But I guess the two are inextricably linked anyway :\

Re: [PATCH net 0/4] mlx5 misc fixes 2025-10-22

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Wed, 22 Oct 2025 15:29:38 +0300 you wrote:
> Hi,
> 
> This patchset provides misc bug fixes from the team to the mlx5 core and
> Eth drivers.
> 
> Thanks,
> Tariq.
> 
> [...]

Here is the summary with links:
  - [net,1/4] net/mlx5: Add PPHCR to PCAM supported registers mask
    https://git.kernel.org/netdev/net/c/bb65e0c141f8
  - [net,2/4] net/mlx5e: Skip PPHCR register query if not supported by the device
    https://git.kernel.org/netdev/net/c/d58a9a917aa3
  - [net,3/4] net/mlx5: Refactor devcom to return NULL on failure
    https://git.kernel.org/netdev/net/c/8f82f89550da
  - [net,4/4] net/mlx5: Fix IPsec cleanup over MPV device
    https://git.kernel.org/netdev/net/c/664f76be38a1

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html