On Sat, Jul 27, 2013 at 11:30:16PM +0100, Chris Wilson wrote: > On Sat, Jul 27, 2013 at 03:13:16PM -0700, Ben Widawsky wrote: > > On Sat, Jul 27, 2013 at 05:23:55PM +0100, Chris Wilson wrote: > > > > This should only return true if no_fbc_reason != FBC_OK, right? > > We want the message on transition from FBC_OK -> error as well so that > we log when we go from enabling FBC to having to disable it for some > reason. That reason is likely to be the same over many off -> on -> off > loops, but useful info I think. You could argue that it would be useful > to refresh the no_fbc_reason message on every modeset for debugging > convenience - but I'm biased as I've found this message to always be > noise. > -Chris What I meant is, you don't want it to return true if you go from: FBC_OFF_FOR_REASON_A->FBC_OFF_FOR_REASON_B Only when you go from: FBC_OK -> !FBC_OK -- Ben Widawsky, Intel Open Source Technology Center _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx
