[PATCH] drm/amdgpu: fix deadlock of reservation between cs and gpu reset

[zhoucm1]

pls ignore this one, will send another new one.

On 2017年04月27日 15:20, Chunming Zhou wrote:
> the case could happen when gpu reset:
> 1. when gpu reset, cs can be continue until sw queue is full, then push job will wait with holding pd reservation.
> 2. gpu_reset routine will also need pd reservation to restore page table from their shadow.
> 3. cs is waiting for gpu_reset complete, but gpu reset is waiting for cs releases reservation.
>
> Change-Id: I0f66d04b2bef3433035109623c8a5c5992c84202
> Signed-off-by: Chunming Zhou <David1.Zhou at amd.com>
> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> index 9edb1a4..8c6d036 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> @@ -1076,7 +1076,6 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
>   	amdgpu_job_free_resources(job);
>   
>   	trace_amdgpu_cs_ioctl(job);
> -	amd_sched_entity_push_job(&job->base);
>   
>   	return 0;
>   }
> @@ -1132,6 +1131,8 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
>   
>   out:
>   	amdgpu_cs_parser_fini(&parser, r, reserved_buffers);
> +	if (!r)
> +		amd_sched_entity_push_job(&parser.job->base);
>   	return r;
>   }
>