[PATCH] drm/amd/gfx11: move the gfx mutex into the caller

SRINIVASAN SHANMUGAM srinivasan.shanmugam at amd.com
Tue Aug 20 15:59:11 UTC 2024


On 8/20/2024 8:09 PM, Alex Deucher wrote:
> Otherwise we can fail to drop the software mutex when
> we fail to take the hardware mutex.
>
> Fixes: 76acba7b7f12 ("drm/amdgpu/gfx11: add a mutex for the gfx semaphore")
> Reported-by: Dan Carpenter <dan.carpenter at linaro.org>
> Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
> ---
>   drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c | 7 +++----
>   1 file changed, 3 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
> index 5704ad25a49d6..aa7fdece8ad42 100644
> --- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
> +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
> @@ -4747,8 +4747,6 @@ int gfx_v11_0_request_gfx_index_mutex(struct amdgpu_device *adev,
>   {
>   	u32 i, tmp, val;
>   
> -	if (req)
> -		mutex_lock(&adev->gfx.reset_sem_mutex);
>   	for (i = 0; i < adev->usec_timeout; i++) {
>   		/* Request with MeId=2, PipeId=0 */
>   		tmp = REG_SET_FIELD(0, CP_GFX_INDEX_MUTEX, REQUEST, req);
> @@ -4769,8 +4767,6 @@ int gfx_v11_0_request_gfx_index_mutex(struct amdgpu_device *adev,
>   		}
>   		udelay(1);
>   	}
> -	if (!req)
> -		mutex_unlock(&adev->gfx.reset_sem_mutex);
>   
>   	if (i >= adev->usec_timeout)
>   		return -EINVAL;
> @@ -4818,8 +4814,10 @@ static int gfx_v11_0_soft_reset(void *handle)
>   	mutex_unlock(&adev->srbm_mutex);
>   
>   	/* Try to acquire the gfx mutex before access to CP_VMID_RESET */
> +	mutex_lock(&adev->gfx.reset_sem_mutex);
>   	r = gfx_v11_0_request_gfx_index_mutex(adev, true);
>   	if (r) {
> +		mutex_unlock(&adev->gfx.reset_sem_mutex);
>   		DRM_ERROR("Failed to acquire the gfx mutex during soft reset\n");
>   		return r;
>   	}
> @@ -4834,6 +4832,7 @@ static int gfx_v11_0_soft_reset(void *handle)
>   
>   	/* release the gfx mutex */
>   	r = gfx_v11_0_request_gfx_index_mutex(adev, false);
> +	mutex_unlock(&adev->gfx.reset_sem_mutex);
>   	if (r) {
>   		DRM_ERROR("Failed to release the gfx mutex during soft reset\n");
>   		return r;
By moving locking and unlocking of the mutex to the gfx_v11_0_soft_reset 
function (the caller). This ensures that the mutex is always unlocked, 
regardless of whether (indicated by if (i >= adev->usec_timeout)) 
succeeds or fails.

Acked-by: Srinivasan Shanmugam <srinivasan.shanmugam at amd.com>


More information about the dri-devel mailing list