[PATCH 1/5] drm/sched: Optimise drm_sched_entity_push_job

[Philipp Stanner]

On Wed, 2024-10-16 at 08:41 +0100, Tvrtko Ursulin wrote:
> 
> On 15/10/2024 15:00, Philipp Stanner wrote:
> > > [...]
> > > How about this:
> > > 
> > > """
> > > In FIFO mode (which is the default), both
> > > drm_sched_entity_push_job()
> > > and drm_sched_rq_update_fifo(), where the latter calls the
> > > former,
> > > are
> > > currently taking and releasing the same entity->rq_lock.
> > > 
> > > We can avoid that design inelegance, and also have a miniscule
> > > efficiency improvement on the idle submit path, by introducing a
> > > new
> > > drm_sched_rq_update_fifo_locked() helper and pulling up the lock
> > > taking
> > > to its callers.
> > > """
> > 
> > That looks good to me
> 
> Cool. So this for 1/5, your text and some tweaks for 4/5, anything
> else 
> or I can respin with that?

Just went through all of them and looks good to me, I think we can go
with that.

Note that below your signature in the latest answer to patch 4 I
mentioned a second position where the old name was forgotten; was not
sure if you saw that.

Regards,
P.



> 
> Regards,
> 
> Tvrtko
> 
> > 
> > P.
> > 
> > 
> > > 
> > > > (Let me read through the other patches briefly. Then we should
> > > > be
> > > > good
> > > > with v2 of this series.. or would it be v3 then?)
> > > 
> > > Depends how you count. By unique series titles or by fundamental
> > > content.
> > > 
> > > Regards,
> > > 
> > > Tvrtko
> > > 
> > 
>