On Fri, Feb 14, 2014 at 02:06:07PM +0100, Daniel Vetter wrote: > Otherwise we end up tearing down fences when e.g. the client quits > way too early. Might or might not fix a fence pin_count BUG Ville has > reported. > > Cc: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx> > Signed-off-by: Daniel Vetter <daniel.vetter@xxxxxxxx> Good catch. I am slightly worried that our management of GGTT is a little haphazard, but given the slightly differing use-cases in i915_vma_unbind and set_cache_level, this looks cleaner than any alternative I quickly sketched. This and the preceding patches are Reviewed-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx
