On Thu, Jun 27, 2019 at 06:00:21PM +0300, Mika Kuoppala wrote:
> Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> writes:
>
> > Quoting Mika Kuoppala (2019-06-27 15:44:38)
> >> If we untrack wakerefs, the actual count may reach zero.
> >> However the krealloced owners array is still there and
> >> needs to be taken care of. Free the owners unconditionally
> >> to fix the leak.
> >
> > That is true.
> >
> >> Fixes: dbf99c1f8c7e ("drm/i915: Force printing wakeref tacking during pm_cleanup")
> >
> > However, the bug is mine
> > Fixes: bd780f37a361 ("drm/i915: Track all held rpm wakerefs")
>
> Indeed. Apologies Imre!
>
> I personally know the reviewer of the original patch
> and will have a discussion.
Good catch, I've been staring at the kmemleak logs J-P provided, but
couldn't see how it can happen.
> -Mika
>
> >
> >> Reported-by: Juha-Pekka Heikkila <juhapekka.heikkila@xxxxxxxxx>
> >> Cc: Juha-Pekka Heikkila <juhapekka.heikkila@xxxxxxxxx>
> >> Cc: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
> >> Cc: Imre Deak <imre.deak@xxxxxxxxx>
> >> Signed-off-by: Mika Kuoppala <mika.kuoppala@xxxxxxxxxxxxxxx>
> > Reviewed-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
> > -Chris
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
