<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">Am 08.08.2018 um 12:55 schrieb Nayan
Deshmukh:<br>
</div>
<blockquote type="cite"
cite="mid:CAFd4ddzXc7DaF_1Lz1ffpTngeAkHUX=pXnng0OagVqo95LBgGw@mail.gmail.com">
<meta http-equiv="content-type" content="text/html; charset=utf-8">
<div dir="ltr">Hi Christian,<br>
<div><br>
<div class="gmail_quote">
<div dir="ltr">On Wed, Aug 8, 2018 at 4:20 PM Christian
König <<a
href="mailto:ckoenig.leichtzumerken@gmail.com"
moz-do-not-send="true">ckoenig.leichtzumerken@gmail.com</a>>
wrote:<br>
</div>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">Ping,
Nayan any comments on that or can I commit it?<br>
<br>
This is just a stripped down version of my original
last_scheduled <br>
improvement patch.<br>
<br>
</blockquote>
<div>I had missed the mail yesterday. I was just looking at
the code right now. The code looks fine but I wanted to
run it once before giving it a Rb. <br>
<br>
</div>
<div>I was wondering why you stripped it down. Was there
some problem with the previous one?<br>
</div>
</div>
</div>
</div>
</blockquote>
<br>
No, not really. But this is a bug fix and bug fixes should be as
simple as possible.<br>
<br>
Improvements should come in separate patches. Didn't thought
initially that this is actually a fix.<br>
<br>
BTW: Another small fix is required, looks like we need the scheduler
correctly set for debugging even earlier.<br>
<br>
Christian.<br>
<br>
<blockquote type="cite"
cite="mid:CAFd4ddzXc7DaF_1Lz1ffpTngeAkHUX=pXnng0OagVqo95LBgGw@mail.gmail.com">
<div dir="ltr">
<div>
<div class="gmail_quote">
<div><br>
</div>
<div>Cheers,<br>
</div>
<div>Nayan <br>
</div>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
Christian.<br>
<br>
Am 07.08.2018 um 14:54 schrieb Christian König:<br>
> Make sure we access last_scheduled only after
checking that there are no<br>
> more jobs on the entity.<br>
><br>
> Signed-off-by: Christian König <<a
href="mailto:christian.koenig@amd.com" target="_blank"
moz-do-not-send="true">christian.koenig@amd.com</a>><br>
> ---<br>
> drivers/gpu/drm/scheduler/gpu_scheduler.c | 21
+++++++++++----------<br>
> 1 file changed, 11 insertions(+), 10 deletions(-)<br>
><br>
> diff --git
a/drivers/gpu/drm/scheduler/gpu_scheduler.c
b/drivers/gpu/drm/scheduler/gpu_scheduler.c<br>
> index 8ee249886473..bd7883d1b964 100644<br>
> --- a/drivers/gpu/drm/scheduler/gpu_scheduler.c<br>
> +++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c<br>
> @@ -568,19 +568,20 @@ void
drm_sched_entity_push_job(struct drm_sched_job *sched_job,<br>
> struct drm_sched_entity
*entity)<br>
> {<br>
> struct drm_sched_rq *rq = entity->rq;<br>
> - bool first, reschedule, idle;<br>
> + bool first;<br>
> <br>
> - idle = entity->last_scheduled == NULL ||<br>
> -
dma_fence_is_signaled(entity->last_scheduled);<br>
> first =
spsc_queue_count(&entity->job_queue) == 0;<br>
> - reschedule = idle && first &&
(entity->num_rq_list > 1);<br>
> + if (first && (entity->num_rq_list
> 1)) {<br>
> + struct dma_fence *fence;<br>
> <br>
> - if (reschedule) {<br>
> - rq =
drm_sched_entity_get_free_sched(entity);<br>
> - spin_lock(&entity->rq_lock);<br>
> -
drm_sched_rq_remove_entity(entity->rq, entity);<br>
> - entity->rq = rq;<br>
> - spin_unlock(&entity->rq_lock);<br>
> + fence =
READ_ONCE(entity->last_scheduled);<br>
> + if (fence == NULL ||
dma_fence_is_signaled(fence)) {<br>
> + rq =
drm_sched_entity_get_free_sched(entity);<br>
> +
spin_lock(&entity->rq_lock);<br>
> +
drm_sched_rq_remove_entity(entity->rq, entity);<br>
> + entity->rq = rq;<br>
> +
spin_unlock(&entity->rq_lock);<br>
> + }<br>
> }<br>
> <br>
> sched_job->sched = entity->rq->sched;<br>
<br>
</blockquote>
</div>
</div>
</div>
</blockquote>
<br>
</body>
</html>