On Mon, Nov 14, 2016 at 09:01:00AM +0000, Tvrtko Ursulin wrote: > > On 14/11/2016 08:56, Chris Wilson wrote: > >Localise the static struct lock_class_key to the caller of > >i915_sw_fence_init() so that we create a lock_class instance for each > >unique sw_fence rather than all sw_fences sharing the same > >lock_class. This eliminate some lockdep false positive when using fences > >from within fence callbacks. > > > >Signed-off-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> > > This one had Joonas' r-b. No, this is a new patch. Tackling the same issue of nested annotations but this time adding the classes to the fence not to the timeline (which is the next patch). This is the difference between v2 and v3, and kills that remining nasty spin_lock_irqsave_nested() from v2. -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
