On Wed, 04 May 2011 14:20:27 -0700, Keith Packard <kei...@keithp.com> wrote: > On Wed, 04 May 2011 20:40:37 +0100, Chris Wilson <ch...@chris-wilson.co.uk> > wrote: > > > We can remove dpms_mode in favour of solely using active. > > That sounds like a good plan. > > And what about the intel_dp->dpms_mode value? Should it be switched to a > simple 'active' boolean as well? Or is the existing crtc->active value > sufficient?
The sticky point is the usage within release-load-detect-pipe where we restore the "dpms-mode". As it stands today, we treat dpms-mode as if it were just a boolean value (ignoring the unknown value). But we should keep the code generic, if it doesn't incur any additional burden, so I'm favouring a multi-valued active or actual-dpms-mode. -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list Intel-gfx@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/intel-gfx
