[PATCH] drm/etnaviv: don't block scheduler when GPU is still active

[Lucas Stach]

Hi,

does anyone have some bandwidth to review this?

Regards,
Lucas

Am Freitag, dem 31.03.2023 um 13:00 +0200 schrieb Lucas Stach:
> Since 45ecaea73883 ("drm/sched: Partial revert of 'drm/sched: Keep
> s_fence->parent pointer'") still active jobs aren't put back in the
> pending list on drm_sched_start(), as they don't have a active
> parent fence anymore, so if the GPU is still working and the timeout
> is extended, all currently active jobs will be freed.
> 
> To avoid prematurely freeing jobs that are still active on the GPU,
> don't block the scheduler until we are fully committed to actually
> reset the GPU.
> 
> Cc: stable at vger.kernel.org #6.0
> Signed-off-by: Lucas Stach <l.stach at pengutronix.de>
> ---
> The behavior change in the scheduler is unfortunate and at least
> deserves some updated documentation. This change aligns etnaviv with
> the behavior of other drivers and avoids the issue.
> ---
>  drivers/gpu/drm/etnaviv/etnaviv_sched.c | 15 +++++----------
>  1 file changed, 5 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/gpu/drm/etnaviv/etnaviv_sched.c b/drivers/gpu/drm/etnaviv/etnaviv_sched.c
> index 1ae87dfd19c4..35d7c2ef7a57 100644
> --- a/drivers/gpu/drm/etnaviv/etnaviv_sched.c
> +++ b/drivers/gpu/drm/etnaviv/etnaviv_sched.c
> @@ -38,15 +38,12 @@ static enum drm_gpu_sched_stat etnaviv_sched_timedout_job(struct drm_sched_job
>  	u32 dma_addr;
>  	int change;
>  
> -	/* block scheduler */
> -	drm_sched_stop(&gpu->sched, sched_job);
> -
>  	/*
>  	 * If the GPU managed to complete this jobs fence, the timout is
>  	 * spurious. Bail out.
>  	 */
>  	if (dma_fence_is_signaled(submit->out_fence))
> -		goto out_no_timeout;
> +		return DRM_GPU_SCHED_STAT_NOMINAL;
>  
>  	/*
>  	 * If the GPU is still making forward progress on the front-end (which
> @@ -59,9 +56,12 @@ static enum drm_gpu_sched_stat etnaviv_sched_timedout_job(struct drm_sched_job
>  	    change < 0 || change > 16) {
>  		gpu->hangcheck_dma_addr = dma_addr;
>  		gpu->hangcheck_fence = gpu->completed_fence;
> -		goto out_no_timeout;
> +		return DRM_GPU_SCHED_STAT_NOMINAL;
>  	}
>  
> +	/* block scheduler */
> +	drm_sched_stop(&gpu->sched, sched_job);
> +
>  	if(sched_job)
>  		drm_sched_increase_karma(sched_job);
>  
> @@ -73,11 +73,6 @@ static enum drm_gpu_sched_stat etnaviv_sched_timedout_job(struct drm_sched_job
>  
>  	drm_sched_start(&gpu->sched, true);
>  	return DRM_GPU_SCHED_STAT_NOMINAL;
> -
> -out_no_timeout:
> -	/* restart scheduler after GPU is usable again */
> -	drm_sched_start(&gpu->sched, true);
> -	return DRM_GPU_SCHED_STAT_NOMINAL;
>  }
>  
>  static void etnaviv_sched_free_job(struct drm_sched_job *sched_job)