[Intel-gfx] [PATCH 0/7] gpu reset and page_flip removal, take 2

[Daniel Vetter]

Hi all,

I landed the locking change, which fixes the gpu reset deadlocks at least for
anything modern (and non-igt). Would be great if we can get the first patch
landed fast, just to address the regression in CI and get rid of the dmesg-warn.
Currently we're blind to dmesg noise in a lot of gpu hang tests and that's not
all that good (the timeout based wedging has a DRM_ERROR, which I think is the
right thing for unexpected delays).

The remaining patches, especially patch 3, probably need a bunch more soaking
time. I think 4-7 are ready for merging, not much pointing in holding them up
due to lack of boosting in atomic.

As usual, review&testing very much welcome.

Thanks,
Daniel

Daniel Vetter (7):
  drm/i915: Avoid the gpu reset vs. modeset deadlock
  drm/i915: Push i915_sw_fence_wait into the nonblocking atomic commit
  drm/i915: More surgically unbreak the modeset vs reset deadlock
  drm/i915: Rip out legacy page_flip completion/irq handling
  drm/i915: adjust has_pending_fb_unpin to atomic
  drm/i915: Remove intel_flip_work infrastructure
  drm/i915: Drop unpin stall in atomic_prepare_commit

 drivers/gpu/drm/i915/i915_debugfs.c  |  70 ------
 drivers/gpu/drm/i915/i915_drv.c      |   1 -
 drivers/gpu/drm/i915/i915_drv.h      |   7 +-
 drivers/gpu/drm/i915/i915_gem.c      |   2 -
 drivers/gpu/drm/i915/i915_irq.c      | 151 ++-----------
 drivers/gpu/drm/i915/intel_display.c | 424 +++++------------------------------
 drivers/gpu/drm/i915/intel_drv.h     |  26 +--
 drivers/gpu/drm/i915/intel_sprite.c  |   8 +-
 8 files changed, 88 insertions(+), 601 deletions(-)

-- 
2.13.2