[PATCH] drm/amdkfd: Cleanup workqueue during module unload

[Chen, Xiaogang]

On 3/20/2024 5:52 PM, Mukul Joshi wrote:
> Caution: This message originated from an External Source. Use proper caution when opening attachments, clicking links, or responding.
>
>
> Destroy the high priority workqueue that handles interrupts
> during KFD node cleanup.
>
> Signed-off-by: Mukul Joshi<mukul.joshi at amd.com>
> ---
>   drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c | 2 ++
>   1 file changed, 2 insertions(+)
>
> diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
> index dd3c43c1ad70..9b6b6e882593 100644
> --- a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
> +++ b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
> @@ -104,6 +104,8 @@ void kfd_interrupt_exit(struct kfd_node *node)
>           */
>          flush_workqueue(node->ih_wq);
>
> +       destroy_workqueue(node->ih_wq);
> +

Here I think we should cancel work items that are still in the work 
queue, not flush workqueue node->ih_wq. In this case the kfd functions 
have been terminated, there is no way to handle the left work items. 
That would make work queue flush never finish. I think it is the reason 
there are orphan kernel tasks.

After cancel left work items we can call destroy_workqueue.

Regards

Xiaogang

>          kfifo_free(&node->ih_fifo);
>   }
>
> --
> 2.35.1
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/amd-gfx/attachments/20240326/d94c5687/attachment.htm>