Re: [PATCH] drm/amd: move DRM_ERROR log out of the mutex protect area

[Christian König]

Am 02.11.20 um 04:14 schrieb Bernard Zhao:
> In function amdgpu_register_gpu_instance, there is no need to
> protect DRM_ERROR in mutex mgpu_info.mutex.
> This change is to make the code to run a bit fast.

NAK, performance for an error path in module load is pretty much irrelevant.

This just looks like a random unnecessary code change to me.

Regards,
Christian.

> Signed-off-by: Bernard Zhao <bernard@xxxxxxxx>
> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
> index efda38349a03..cc61b0a5b8d1 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
> @@ -104,8 +104,8 @@ void amdgpu_register_gpu_instance(struct amdgpu_device *adev)
>   	mutex_lock(&mgpu_info.mutex);
>   
>   	if (mgpu_info.num_gpu >= MAX_GPU_INSTANCE) {
> -		DRM_ERROR("Cannot register more gpu instance\n");
>   		mutex_unlock(&mgpu_info.mutex);
> +		DRM_ERROR("Cannot register more gpu instance\n");
>   		return;
>   	}
>