[PATCH net-next] devlink: Remove extra device_lock assert checks

[PATCH net-next] devlink: Remove extra device_lock assert checks

[Leon Romanovsky]

From: Leon Romanovsky <leonro@nvidia.com>

PCI core code in the pci_call_probe() has a path that doesn't hold
device_lock. It happens because the ->probe() is called through the
workqueue mechanism.

   349 static int pci_call_probe(struct pci_driver *drv, struct pci_dev *dev,
   350                           const struct pci_device_id *id)
   351 {
   352
....
   377         if (cpu < nr_cpu_ids)
   378                 error = work_on_cpu(cpu, local_pci_probe, &ddi);

Luckily enough, the core still ensures that only single flow is executed,
so it safe to remove the assert checks that anyway were added for annotations
purposes.

Fixes: b88f7b1203bf ("devlink: Annotate devlink API calls")
Reported-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
---
 net/core/devlink.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/net/core/devlink.c b/net/core/devlink.c
index 3ce6147a2fe8..3464854015a2 100644
--- a/net/core/devlink.c
+++ b/net/core/devlink.c
@@ -9147,7 +9147,6 @@ void devlink_register(struct devlink *devlink)
 {
 	ASSERT_DEVLINK_NOT_REGISTERED(devlink);
 	/* Make sure that we are in .probe() routine */
-	device_lock_assert(devlink->dev);
 
 	mutex_lock(&devlink_mutex);
 	xa_set_mark(&devlinks, devlink->index, DEVLINK_REGISTERED);
@@ -9165,7 +9164,6 @@ void devlink_unregister(struct devlink *devlink)
 {
 	ASSERT_DEVLINK_REGISTERED(devlink);
 	/* Make sure that we are in .remove() routine */
-	device_lock_assert(devlink->dev);
 
 	devlink_put(devlink);
 	wait_for_completion(&devlink->comp);
-- 
2.31.1

Re: [PATCH net-next] devlink: Remove extra device_lock assert checks

[patchwork-bot+netdevbpf]

Hello:

This patch was applied to netdev/net-next.git (master)
by David S. Miller <davem@davemloft.net>:

On Tue, 19 Oct 2021 10:49:54 +0300 you wrote:
> From: Leon Romanovsky <leonro@nvidia.com>
> 
> PCI core code in the pci_call_probe() has a path that doesn't hold
> device_lock. It happens because the ->probe() is called through the
> workqueue mechanism.
> 
>    349 static int pci_call_probe(struct pci_driver *drv, struct pci_dev *dev,
>    350                           const struct pci_device_id *id)
>    351 {
>    352
> ....
>    377         if (cpu < nr_cpu_ids)
>    378                 error = work_on_cpu(cpu, local_pci_probe, &ddi);
> 
> [...]

Here is the summary with links:
  - [net-next] devlink: Remove extra device_lock assert checks
    https://git.kernel.org/netdev/net-next/c/cb3dc8901ba4

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