[Intel-gfx] [RFC PATCH 0/2] atomic dpms support

Daniel Vetter daniel.vetter at ffwll.ch
Thu Dec 18 08:08:50 PST 2014


Hi all,

Here's the atomic dpms I've promised. Still a few things open that I need to
look at once Rob's series has landed and I can rebase:
- Not sure whether we should expose crtc_state->active or not as atomic prop. I
  guess it won't hurt to do so, so I'll do that.
- Some of the sanity checks need to be moved from helpers to the core state
  checks Rob's patches add. There's a FIXME in the first patch for this.
- I haven't yet added sanity checks for pageflips to reject any event generation
  when the pipe is off (since really that doesn't make a lot of sense) to align
  everyone with what i915 does already.
- We can't yet do the same thing for vblank events from the vblank wait ioctl
  since that doesn't yet deal in kms objects at all. So that check needs to
  write for the big drm_irq.c rewrite.
- A bit more testing - my development vehicle is still a lot of hacks and cargo
  culting here ;-)

Anyway just wanted to get this out for feedback and review before I head off for
2 weeks of vacation.

Cheers, Daniel

Daniel Vetter (2):
  drm/atomic: Add drm_crtc_state->active
  drm/atomic-helper: add connector->dpms() implementation

 drivers/gpu/drm/drm_atomic_helper.c | 136 ++++++++++++++++++++++++++++++++++--
 include/drm/drm_atomic_helper.h     |   2 +
 include/drm/drm_crtc.h              |  10 +++
 3 files changed, 143 insertions(+), 5 deletions(-)

-- 
2.1.3



More information about the Intel-gfx mailing list