Re: [RESEND PATCH v4] drm: Don't free jobs in wait_event_interruptible()

[Koenig, Christian]

Am 25.10.19 um 12:26 schrieb Steven Price:
> On 25/10/2019 10:49, Koenig, Christian wrote:
>> Am 24.10.19 um 18:24 schrieb Steven Price:
>>> drm_sched_cleanup_jobs() attempts to free finished jobs, however because
>>> it is called as the condition of wait_event_interruptible() it must not
>>> sleep. Unfortuantly some free callbacks (notibly for Panfrost) do sleep.
>>>
>>> Instead let's rename drm_sched_cleanup_jobs() to
>>> drm_sched_get_cleanup_job() and simply return a job for processing if
>>> there is one. The caller can then call the free_job() callback outside
>>> the wait_event_interruptible() where sleeping is possible before
>>> re-checking and returning to sleep if necessary.
>>>
>>> Signed-off-by: Steven Price <steven.price@xxxxxxx>
>>> ---
>>> Previous posting: https://lore.kernel.org/lkml/20190926141630.14258-1-steven.price@xxxxxxx/
>>>
>>>    drivers/gpu/drm/scheduler/sched_main.c | 45 +++++++++++++++-----------
>>>    1 file changed, 26 insertions(+), 19 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c
>>> index 9a0ee74d82dc..148468447ba9 100644
>>> --- a/drivers/gpu/drm/scheduler/sched_main.c
>>> +++ b/drivers/gpu/drm/scheduler/sched_main.c
>>> @@ -622,43 +622,41 @@ static void drm_sched_process_job(struct dma_fence *f, struct dma_fence_cb *cb)
>>>    }
>>>    
>>>    /**
>>> - * drm_sched_cleanup_jobs - destroy finished jobs
>>> + * drm_sched_get_cleanup_job - fetch the next finished job to be destroyed
>>>     *
>>>     * @sched: scheduler instance
>>>     *
>>> - * Remove all finished jobs from the mirror list and destroy them.
>>> + * Returns the next finished job from the mirror list (if there is one)
>>> + * ready for it to be destroyed.
>>>     */
>>> -static void drm_sched_cleanup_jobs(struct drm_gpu_scheduler *sched)
>>> +static struct drm_sched_job *
>>> +drm_sched_get_cleanup_job(struct drm_gpu_scheduler *sched)
>>>    {
>>> +	struct drm_sched_job *job = NULL;
>> Please don't initialize job here.
> Good spot, will fix.
>
>>>    	unsigned long flags;
>>>    
>>>    	/* Don't destroy jobs while the timeout worker is running */
>>>    	if (sched->timeout != MAX_SCHEDULE_TIMEOUT &&
>>>    	    !cancel_delayed_work(&sched->work_tdr))
>>> -		return;
>>> -
>>> +		return NULL;
>>>    
>>> -	while (!list_empty(&sched->ring_mirror_list)) {
>>> -		struct drm_sched_job *job;
>>> +	spin_lock_irqsave(&sched->job_list_lock, flags);
>>>    
>>> -		job = list_first_entry(&sched->ring_mirror_list,
>>> +	job = list_first_entry_or_null(&sched->ring_mirror_list,
>>>    				       struct drm_sched_job, node);
>>> -		if (!dma_fence_is_signaled(&job->s_fence->finished))
>>> -			break;
>>>    
>>> -		spin_lock_irqsave(&sched->job_list_lock, flags);
>>> +	if (job && dma_fence_is_signaled(&job->s_fence->finished)) {
>>>    		/* remove job from ring_mirror_list */
>>>    		list_del_init(&job->node);
>>> -		spin_unlock_irqrestore(&sched->job_list_lock, flags);
>>> -
>>> -		sched->ops->free_job(job);
>>> +	} else {
>>> +		job = NULL;
>>> +		/* queue timeout for next job */
>>> +		drm_sched_start_timeout(sched);
>>>    	}
>>>    
>>> -	/* queue timeout for next job */
>>> -	spin_lock_irqsave(&sched->job_list_lock, flags);
>>> -	drm_sched_start_timeout(sched);
>>>    	spin_unlock_irqrestore(&sched->job_list_lock, flags);
>>>    
>>> +	return job;
>>>    }
>>>    
>>>    /**
>>> @@ -698,12 +696,21 @@ static int drm_sched_main(void *param)
>>>    		struct drm_sched_fence *s_fence;
>>>    		struct drm_sched_job *sched_job;
>>>    		struct dma_fence *fence;
>>> +		struct drm_sched_job *cleanup_job = NULL;
>>>    
>>>    		wait_event_interruptible(sched->wake_up_worker,
>>> -					 (drm_sched_cleanup_jobs(sched),
>>> +					 (cleanup_job = drm_sched_get_cleanup_job(sched)) ||
>>>    					 (!drm_sched_blocked(sched) &&
>>>    					  (entity = drm_sched_select_entity(sched))) ||
>>> -					 kthread_should_stop()));
>>> +					 kthread_should_stop());
>>> +
>>> +		while (cleanup_job) {
>> Better make this just "if (cleanup_job)"... to make sure that we stop
>> immediately when the thread should stop.
> Ok, no problem. Note that this is a change in behaviour (previously
> drm_sched_cleanup_jobs() was called *before* checking
> kthread_should_stop()). But I can't see the harm.

Yeah, but this is actually a rather nice improvement.

When we say that the thread should stop then that should happen 
immediately and not cleanup the jobs first.

Christian.

>
> Steve
>
>> Apart from that looks good to me now,
>> Christian.
>>
>>> +			sched->ops->free_job(cleanup_job);
>>> +			/* queue timeout for next job */
>>> +			drm_sched_start_timeout(sched);
>>> +
>>> +			cleanup_job = drm_sched_get_cleanup_job(sched);
>>> +		}
>>>    
>>>    		if (!entity)
>>>    			continue;
>> _______________________________________________
>> dri-devel mailing list
>> dri-devel@xxxxxxxxxxxxxxxxxxxxx
>> https://lists.freedesktop.org/mailman/listinfo/dri-devel
>>