net/mlx5e: Fix netif state handling

BugLink: https://bugs.launchpad.net/bugs/2106632

mlx5e_suspend cleans resources only if netif_device_present() returns
true. However, mlx5e_resume changes the state of netif, via
mlx5e_nic_enable, only if reg_state == NETREG_REGISTERED.
In the below case, the above leads to NULL-ptr Oops[1] and memory
leaks:

mlx5e_probe
 _mlx5e_resume
  mlx5e_attach_netdev
   mlx5e_nic_enable  <-- netdev not reg, not calling netif_device_attach()
  register_netdev <-- failed for some reason.
ERROR_FLOW:
 _mlx5e_suspend <-- netif_device_present return false, resources aren't freed :(

Hence, clean resources in this case as well.

[1]
BUG: kernel NULL pointer dereference, address: 0000000000000000
PGD 0 P4D 0
Oops: 0010 [#1] SMP
CPU: 2 PID: 9345 Comm: test-ovs-ct-gen Not tainted 6.5.0_for_upstream_min_debug_2023_09_05_16_01 #1
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
RIP: 0010:0x0
Code: Unable to access opcode bytes at0xffffffffffffffd6.
RSP: 0018:ffff888178aaf758 EFLAGS: 00010246
Call Trace:
 <TASK>
 ? __die+0x20/0x60
 ? page_fault_oops+0x14c/0x3c0
 ? exc_page_fault+0x75/0x140
 ? asm_exc_page_fault+0x22/0x30
 notifier_call_chain+0x35/0xb0
 blocking_notifier_call_chain+0x3d/0x60
 mlx5_blocking_notifier_call_chain+0x22/0x30 [mlx5_core]
 mlx5_core_uplink_netdev_event_replay+0x3e/0x60 [mlx5_core]
 mlx5_mdev_netdev_track+0x53/0x60 [mlx5_ib]
 mlx5_ib_roce_init+0xc3/0x340 [mlx5_ib]
 __mlx5_ib_add+0x34/0xd0 [mlx5_ib]
 mlx5r_probe+0xe1/0x210 [mlx5_ib]
 ? auxiliary_match_id+0x6a/0x90
 auxiliary_bus_probe+0x38/0x80
 ? driver_sysfs_add+0x51/0x80
 really_probe+0xc9/0x3e0
 ? 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+0x637/0x840
 __auxiliary_device_add+0x3b/0xa0
 add_adev+0xc9/0x140 [mlx5_core]
 mlx5_rescan_drivers_locked+0x22a/0x310 [mlx5_core]
 mlx5_register_device+0x53/0xa0 [mlx5_core]
 mlx5_init_one_devl_locked+0x5c4/0x9c0 [mlx5_core]
 mlx5_init_one+0x3b/0x60 [mlx5_core]
 probe_one+0x44c/0x730 [mlx5_core]
 local_pci_probe+0x3e/0x90
 pci_device_probe+0xbf/0x210
 ? kernfs_create_link+0x5d/0xa0
 ? sysfs_do_create_link_sd+0x60/0xc0
 really_probe+0xc9/0x3e0
 ? 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
 pci_bus_add_device+0x54/0x80
 pci_iov_add_virtfn+0x2e6/0x320
 sriov_enable+0x208/0x420
 mlx5_core_sriov_configure+0x9e/0x200 [mlx5_core]
 sriov_numvfs_store+0xae/0x1a0
 kernfs_fop_write_iter+0x10c/0x1a0
 vfs_write+0x291/0x3c0
 ksys_write+0x5f/0xe0
 do_syscall_64+0x3d/0x90
 entry_SYSCALL_64_after_hwframe+0x46/0xb0
 CR2: 0000000000000000
 ---[ end trace 0000000000000000  ]---

Fixes: 2c3b5be ("net/mlx5e: More generic netdev management API")
Signed-off-by: Shay Drory <[email protected]>
Signed-off-by: Tariq Toukan <[email protected]>
Reviewed-by: Simon Horman <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
(cherry picked from commit 3d5918477f94e4c2f064567875c475468e264644)
[diewald: dependency commit for 2a4f56fbcc47 net/mlx5e: Keep netdev when
leave switchdev for devlink set legacy only]
Signed-off-by: Manuel Diewald <[email protected]>
Signed-off-by: Mehmet Basaran <[email protected]>

Author: shayshyi

drivers/net/ethernet/mellanox/mlx5/core/en_main.c

@@ -6011,14 +6011,14 @@ static int mlx5e_resume(struct auxiliary_device *adev)
 	return 0;
 }
 
-static int _mlx5e_suspend(struct auxiliary_device *adev)
+static int _mlx5e_suspend(struct auxiliary_device *adev, bool pre_netdev_reg)
 {
 	struct mlx5e_dev *mlx5e_dev = auxiliary_get_drvdata(adev);
 	struct mlx5e_priv *priv = mlx5e_dev->priv;
 	struct net_device *netdev = priv->netdev;
 	struct mlx5_core_dev *mdev = priv->mdev;
 
-	if (!netif_device_present(netdev)) {
+	if (!pre_netdev_reg && !netif_device_present(netdev)) {
 		if (test_bit(MLX5E_STATE_DESTROYING, &priv->state))
 			mlx5e_destroy_mdev_resources(mdev);
 		return -ENODEV;
@@ -6031,7 +6031,7 @@ static int _mlx5e_suspend(struct auxiliary_device *adev)
 
 static int mlx5e_suspend(struct auxiliary_device *adev, pm_message_t state)
 {
-	return _mlx5e_suspend(adev);
+	return _mlx5e_suspend(adev, false);
 }
 
 static int _mlx5e_probe(struct auxiliary_device *adev)
@@ -6095,7 +6095,7 @@ static int _mlx5e_probe(struct auxiliary_device *adev)
 	return 0;
 
 err_resume:
-	_mlx5e_suspend(adev);
+	_mlx5e_suspend(adev, true);
 err_profile_cleanup:
 	profile->cleanup(priv);
 err_destroy_netdev:
@@ -6121,7 +6121,7 @@ static void mlx5e_remove(struct auxiliary_device *adev)
 	mlx5_core_uplink_netdev_set(priv->mdev, NULL);
 	mlx5e_dcbnl_delete_app(priv);
 	unregister_netdev(priv->netdev);
-	_mlx5e_suspend(adev);
+	_mlx5e_suspend(adev, false);
 	/* Avoid cleanup if profile rollback failed. */
 	if (priv->profile)
 		priv->profile->cleanup(priv);