On Fri, 30 Sep 2011 19:09:46 +0200, Daniel Vetter <daniel at ffwll.ch> wrote:
> grep shows that we also write to PCH_PP_CONTROL in intel_lvds.c in the > dpms functions - any reasons these two writes are left out? Upon a bit of review: The bspec makes it clear that this write protect key only needs to be written for eDP on DPA -- it's a work-around for a bug where panel power sequencing wouldn't work right. The LVDS code does disable write protect in the _init function, which seems global enough, but misses the resume case. We shouldn't ever need to set this field though; it write protects registers only when the panel is running. We could presumably remove the write protect disable entirely in the LVDS code. So, I think the patch as written is correct. -- keith.packard at intel.com -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: <http://lists.freedesktop.org/archives/dri-devel/attachments/20110930/0630fac5/attachment.pgp>
