On Wed, Sep 12, 2012 at 04:31:50PM +0200, Daniel Vetter wrote: > On Wed, Sep 12, 2012 at 10:06:34AM -0300, Paulo Zanoni wrote: > > From: Paulo Zanoni <paulo.r.zanoni at intel.com> > > > > Signed-off-by: Paulo Zanoni <paulo.r.zanoni at intel.com> > Queued for -next, thanks for the patch. I've needed to resolve some > conflicts since I didn't pick up all previous patches, but nothing to > onerous. Wanted to add: I think the clock computation is one of the things most responsible for tons of encoder type checks in the crtc mode_set code. My idea is to move the clock computation out into an enhanced adjust_mode stage, and deal with the encoder special-cases in the encoder->adjust_mode callback. That way we can also do (eventually) pll allocation and bandwidth calculations/link allocations _before_ we start with the actual modeset sequence. Which is obviously required for decent global modeset support. But for now this is a good start to group the clock cruft together and hide it someplace cuddly. -Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch
