diff mbox series

[net] net: hns3: fix kernel crash when uninstalling driver

Message ID 20241101091507.3644584-1-shaojijie@huawei.com (mailing list archive)
State New
Delegated to: Netdev Maintainers
Headers show
Series [net] net: hns3: fix kernel crash when uninstalling driver | expand

Checks

Context Check Description
netdev/series_format success Single patches do not need cover letters
netdev/tree_selection success Clearly marked for net
netdev/ynl success Generated files up to date; no warnings/errors; no diff in generated;
netdev/fixes_present success Fixes tag present in non-next series
netdev/header_inline success No static functions without inline keyword in header files
netdev/build_32bit success Errors and warnings before: 3 this patch: 3
netdev/build_tools success No tools touched, skip
netdev/cc_maintainers fail 2 blamed authors not CCed: huangguangbin2@huawei.com lipeng321@huawei.com; 2 maintainers not CCed: lipeng321@huawei.com huangguangbin2@huawei.com
netdev/build_clang success Errors and warnings before: 3 this patch: 3
netdev/verify_signedoff success Signed-off-by tag matches author and committer
netdev/deprecated_api success None detected
netdev/check_selftest success No net selftest shell script
netdev/verify_fixes success Fixes tag looks correct
netdev/build_allmodconfig_warn success Errors and warnings before: 4 this patch: 4
netdev/checkpatch success total: 0 errors, 0 warnings, 0 checks, 12 lines checked
netdev/build_clang_rust success No Rust files in patch. Skipping build
netdev/kdoc success Errors and warnings before: 0 this patch: 0
netdev/source_inline success Was 0 now: 0
netdev/contest success net-next-2024-11-03--21-00 (tests: 781)

Commit Message

Jijie Shao Nov. 1, 2024, 9:15 a.m. UTC
From: Peiyang Wang <wangpeiyang1@huawei.com>

When the driver is uninstalled and the VF is disabled concurrently, a
kernel crash occurs. The reason is that the two actions call function
pci_disable_sriov(). The num_VFs is checked to determine whether to
release the corresponding resources. During the second calling, num_VFs
is not 0 and the resource release function is called. However, the
corresponding resource has been released during the first invoking.
Therefore, the problem occurs:

[15277.839633][T50670] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000020
...
[15278.131557][T50670] Call trace:
[15278.134686][T50670]  klist_put+0x28/0x12c
[15278.138682][T50670]  klist_del+0x14/0x20
[15278.142592][T50670]  device_del+0xbc/0x3c0
[15278.146676][T50670]  pci_remove_bus_device+0x84/0x120
[15278.151714][T50670]  pci_stop_and_remove_bus_device+0x6c/0x80
[15278.157447][T50670]  pci_iov_remove_virtfn+0xb4/0x12c
[15278.162485][T50670]  sriov_disable+0x50/0x11c
[15278.166829][T50670]  pci_disable_sriov+0x24/0x30
[15278.171433][T50670]  hnae3_unregister_ae_algo_prepare+0x60/0x90 [hnae3]
[15278.178039][T50670]  hclge_exit+0x28/0xd0 [hclge]
[15278.182730][T50670]  __se_sys_delete_module.isra.0+0x164/0x230
[15278.188550][T50670]  __arm64_sys_delete_module+0x1c/0x30
[15278.193848][T50670]  invoke_syscall+0x50/0x11c
[15278.198278][T50670]  el0_svc_common.constprop.0+0x158/0x164
[15278.203837][T50670]  do_el0_svc+0x34/0xcc
[15278.207834][T50670]  el0_svc+0x20/0x30

For details, see the following figure.

     rmmod hclge              disable VFs
----------------------------------------------------
hclge_exit()            sriov_numvfs_store()
  ...                     device_lock()
  pci_disable_sriov()     hns3_pci_sriov_configure()
                            pci_disable_sriov()
                              sriov_disable()
    sriov_disable()             if !num_VFs :
      if !num_VFs :               return;
        return;                 sriov_del_vfs()
      sriov_del_vfs()             ...
        ...                       klist_put()
        klist_put()               ...
        ...                     num_VFs = 0;
      num_VFs = 0;        device_unlock();

In this patch, when driver is removing, we get the device_lock()
to protect num_VFs, just like sriov_numvfs_store().

Fixes: 0dd8a25f355b ("net: hns3: disable sriov before unload hclge layer")
Signed-off-by: Peiyang Wang <wangpeiyang1@huawei.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hnae3.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

Comments

Simon Horman Nov. 4, 2024, 5:43 p.m. UTC | #1
On Fri, Nov 01, 2024 at 05:15:07PM +0800, Jijie Shao wrote:
> From: Peiyang Wang <wangpeiyang1@huawei.com>
> 
> When the driver is uninstalled and the VF is disabled concurrently, a
> kernel crash occurs. The reason is that the two actions call function
> pci_disable_sriov(). The num_VFs is checked to determine whether to
> release the corresponding resources. During the second calling, num_VFs
> is not 0 and the resource release function is called. However, the
> corresponding resource has been released during the first invoking.
> Therefore, the problem occurs:
> 
> [15277.839633][T50670] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000020
> ...
> [15278.131557][T50670] Call trace:
> [15278.134686][T50670]  klist_put+0x28/0x12c
> [15278.138682][T50670]  klist_del+0x14/0x20
> [15278.142592][T50670]  device_del+0xbc/0x3c0
> [15278.146676][T50670]  pci_remove_bus_device+0x84/0x120
> [15278.151714][T50670]  pci_stop_and_remove_bus_device+0x6c/0x80
> [15278.157447][T50670]  pci_iov_remove_virtfn+0xb4/0x12c
> [15278.162485][T50670]  sriov_disable+0x50/0x11c
> [15278.166829][T50670]  pci_disable_sriov+0x24/0x30
> [15278.171433][T50670]  hnae3_unregister_ae_algo_prepare+0x60/0x90 [hnae3]
> [15278.178039][T50670]  hclge_exit+0x28/0xd0 [hclge]
> [15278.182730][T50670]  __se_sys_delete_module.isra.0+0x164/0x230
> [15278.188550][T50670]  __arm64_sys_delete_module+0x1c/0x30
> [15278.193848][T50670]  invoke_syscall+0x50/0x11c
> [15278.198278][T50670]  el0_svc_common.constprop.0+0x158/0x164
> [15278.203837][T50670]  do_el0_svc+0x34/0xcc
> [15278.207834][T50670]  el0_svc+0x20/0x30
> 
> For details, see the following figure.
> 
>      rmmod hclge              disable VFs
> ----------------------------------------------------
> hclge_exit()            sriov_numvfs_store()
>   ...                     device_lock()
>   pci_disable_sriov()     hns3_pci_sriov_configure()
>                             pci_disable_sriov()
>                               sriov_disable()
>     sriov_disable()             if !num_VFs :
>       if !num_VFs :               return;
>         return;                 sriov_del_vfs()
>       sriov_del_vfs()             ...
>         ...                       klist_put()
>         klist_put()               ...
>         ...                     num_VFs = 0;
>       num_VFs = 0;        device_unlock();
> 
> In this patch, when driver is removing, we get the device_lock()
> to protect num_VFs, just like sriov_numvfs_store().
> 
> Fixes: 0dd8a25f355b ("net: hns3: disable sriov before unload hclge layer")
> Signed-off-by: Peiyang Wang <wangpeiyang1@huawei.com>
> Signed-off-by: Jijie Shao <shaojijie@huawei.com>

Reviewed-by: Simon Horman <horms@kernel.org>
diff mbox series

Patch

diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.c b/drivers/net/ethernet/hisilicon/hns3/hnae3.c
index 67b0bf310daa..9a63fbc69408 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.c
@@ -25,8 +25,11 @@  void hnae3_unregister_ae_algo_prepare(struct hnae3_ae_algo *ae_algo)
 		pci_id = pci_match_id(ae_algo->pdev_id_table, ae_dev->pdev);
 		if (!pci_id)
 			continue;
-		if (IS_ENABLED(CONFIG_PCI_IOV))
+		if (IS_ENABLED(CONFIG_PCI_IOV)) {
+			device_lock(&ae_dev->pdev->dev);
 			pci_disable_sriov(ae_dev->pdev);
+			device_unlock(&ae_dev->pdev->dev);
+		}
 	}
 }
 EXPORT_SYMBOL(hnae3_unregister_ae_algo_prepare);