On Thu, 15 Dec 2011 10:38:16 +0000, Chris Wilson <chris at chris-wilson.co.uk> wrote: > Reverting de842eff4101 (drm/i915: Wait for LVDS panel power sequence) > should get another 0.4s back if intel_lvds_enable() is still in the > critical path. After that the focus looks to be upon speeding up > modeset. *not* doing modeset for LVDS/eDP is the key here -- the panel is probably already running, and all we need to do is change the scaling parameters, which (on new hardware) doesn't require a full mode set. I worked with vorlon to investigate what this would take and got pretty far down the path; the big missing piece is that we have to have a valid scanout buffer for the whole setup sequence. -- keith.packard at intel.com -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 827 bytes Desc: not available URL: <http://lists.freedesktop.org/archives/intel-gfx/attachments/20111215/7e5a8b58/attachment.pgp>
