On Tue, Mar 08, 2016 at 11:08:18AM +0000, Tvrtko Ursulin wrote: > Well strictly speaking you don't need to do any platform matrix > testing for an r-b. :) But yes, the situation you are seeing on BSW > needs to be resolved. As discussed on IRC, it does seem to be unreproducible on !bsw (the presumption being that we cannot saturate the CPU with interrupt handler), but the fix may just be to remove the loop from cherryview_irq_handler. I'm kicking off a test run to see if that loop has any merit for GEM... -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
