Re: [PATCH v1 2/2] drm: Clear the fence pointer when writeback job signaled

[Liviu Dudau <Liviu.Dudau@arm.com>]

On Thu, Aug 01, 2019 at 06:31:13AM +0000, Lowry Li (Arm Technology China) wrote:
> Hi Liviu,
> 
> On Wed, Jul 31, 2019 at 01:15:25PM +0000, Liviu Dudau wrote:
> > Hi Lowry,
> > 
> > On Wed, Jul 31, 2019 at 11:04:45AM +0000, Lowry Li (Arm Technology China) wrote:
> > > During it signals the completion of a writeback job, after releasing
> > > the out_fence, we'd clear the pointer.
> > > 
> > > Check if fence left over in drm_writeback_cleanup_job(), release it.
> > > 
> > > Signed-off-by: Lowry Li (Arm Technology China) <lowry.li@arm.com>
> > > ---
> > >  drivers/gpu/drm/drm_writeback.c | 23 +++++++++++++++--------
> > >  1 file changed, 15 insertions(+), 8 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/drm_writeback.c b/drivers/gpu/drm/drm_writeback.c
> > > index ff138b6..43d9e3b 100644
> > > --- a/drivers/gpu/drm/drm_writeback.c
> > > +++ b/drivers/gpu/drm/drm_writeback.c
> > > @@ -324,6 +324,9 @@ void drm_writeback_cleanup_job(struct drm_writeback_job *job)
> > >  	if (job->fb)
> > >  		drm_framebuffer_put(job->fb);
> > >  
> > > +	if (job->out_fence)
> > > +		dma_fence_put(job->out_fence);
> > > +
> > >  	kfree(job);
> > >  }
> > 
> > This change looks good.
> > 
> > >  EXPORT_SYMBOL(drm_writeback_cleanup_job);
> > > @@ -366,25 +369,29 @@ static void cleanup_work(struct work_struct *work)
> > >  {
> > >  	unsigned long flags;
> > >  	struct drm_writeback_job *job;
> > > +	struct dma_fence *out_fence;
> > >  
> > >  	spin_lock_irqsave(&wb_connector->job_lock, flags);
> > >  	job = list_first_entry_or_null(&wb_connector->job_queue,
> > >  				       struct drm_writeback_job,
> > >  				       list_entry);
> > > -	if (job) {
> > > +	if (job)
> > >  		list_del(&job->list_entry);
> > > -		if (job->out_fence) {
> > > -			if (status)
> > > -				dma_fence_set_error(job->out_fence, status);
> > > -			dma_fence_signal(job->out_fence);
> > > -			dma_fence_put(job->out_fence);
> > 
> > *Here*
> > 
> > > -		}
> > > -	}
> > > +
> > >  	spin_unlock_irqrestore(&wb_connector->job_lock, flags);
> > >  
> > >  	if (WARN_ON(!job))
> > >  		return;
> > >  
> > > +	out_fence = job->out_fence;
> > > +	if (out_fence) {
> > > +		if (status)
> > > +			dma_fence_set_error(out_fence, status);
> > > +		dma_fence_signal(out_fence);
> > > +		dma_fence_put(out_fence);
> > > +		job->out_fence = NULL;
> > > +	}
> > > +
> > 
> > I don't get the point of this change. Why not just add job->out_fence = NULL
> > where *Here* is?
> >
> > Best regards,
> > Liviu 
> Besides setting NULL, also did a refine by moving the fence operation
> out of the lock block.

OK, now it makes sense. May I suggest you add that to the commit message?

Otherwise, Acked-by: Liviu Dudau <liviu.dudau@arm.com>

Best regards,
Liviu

> 
> Best regards,
> Lowry 
> > >  	INIT_WORK(&job->cleanup_work, cleanup_work);
> > >  	queue_work(system_long_wq, &job->cleanup_work);
> > >  }
> > > -- 
> > > 1.9.1
> > > 
> > 
> > -- 
> > ====================
> > | I would like to |
> > | fix the world,  |
> > | but they're not |
> > | giving me the   |
> >  \ source code!  /
> >   ---------------
> >     ¯\_(ツ)_/¯
> 
> -- 
> Regards,
> Lowry

-- 
====================
| I would like to |
| fix the world,  |
| but they're not |
| giving me the   |
 \ source code!  /
  ---------------
    ¯\_(ツ)_/¯