[PATCH 1/8] drm/amdgpu: grab an additional reference on the gang fence v2

SRINIVASAN SHANMUGAM srinivasan.shanmugam at amd.com
Fri Mar 14 04:09:03 UTC 2025


On 3/7/2025 7:18 PM, Christian König wrote:
> We keep the gang submission fence around in adev, make sure that it
> stays alive.
>
> v2: fix memory leak on retry
>
> Signed-off-by: Christian König <christian.koenig at amd.com>
> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 10 +++++++++-
>   1 file changed, 9 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> index 198d29faa754..337543ec615c 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> @@ -6889,18 +6889,26 @@ struct dma_fence *amdgpu_device_switch_gang(struct amdgpu_device *adev,
>   {
>   	struct dma_fence *old = NULL;
>   
> +	dma_fence_get(gang);
>   	do {
>   		dma_fence_put(old);
>   		old = amdgpu_device_get_gang(adev);
>   		if (old == gang)
>   			break;
>   
> -		if (!dma_fence_is_signaled(old))
> +		if (!dma_fence_is_signaled(old)) {

Here, should we need to check ?

     // Check if old fence isn't signaled
     if (old && !dma_fence_is_signaled(old)) {

> +			dma_fence_put(gang);
>   			return old;
> +		}
>   
>   	} while (cmpxchg((struct dma_fence __force **)&adev->gang_submit,
>   			 old, gang) != old);
>   
> +	/*
> +	 * Drop it once for the exchanged reference in adev and once for the
> +	 * thread local reference acquired in amdgpu_device_get_gang().
> +	 */
> +	dma_fence_put(old);

if (old)
     dma_fence_put(old); // Ensure to release old reference  only if it 
is valid?


>   	dma_fence_put(old);
>   	return NULL;
>   }


More information about the amd-gfx mailing list