Hi all, I've dug around in Bspec a bit to check out whether we're missing any workarounds or have anything ordered the wrong way around in our setup sequence for the fdi links. And indeed, there are a few small things. I have the illusion that this patch series fixes fdi b/c issues for me, that will probably last until the first test report. There's still a funny w/a left that I'm working on - but it looks like it's a voltage setting in a link somewhere that the bios should fix. So I'll likely only add a check plus a debug printk. The 2nd last patch introduces a new modeset hook and explains the big plans I have for it. For this patch series it's overkill, but I'm already working on a small haswell feature that also plugs into the same hook. Comments, review and testing highly welcome. Cheers, Daniel [Now actually the right pile of patches ... ] Daniel Vetter (9): drm/i915: shut up spurious message in intel_dp_get_hw_state drm/i915: Write the FDI RX TU size reg at the right time drm/i915: set FDI_RX_MISC to recommended values on CPT/PPT drm/i915: add comment about pch pll enabling rules drm/i915: CPT/PPT pch dp transcoder workaround drm/i915: BUG on impossible pch dp port drm/i915: drop unnecessary check from fdi_link_train code drm/i915: add ->display.modeset_global_resources callback drm/i915: check fdi B/C lane sharing constraint drivers/gpu/drm/i915/i915_drv.h | 1 + drivers/gpu/drm/i915/i915_reg.h | 13 ++- drivers/gpu/drm/i915/intel_display.c | 175 +++++++++++++++++++++++++++++++---- drivers/gpu/drm/i915/intel_dp.c | 5 +- 4 files changed, 169 insertions(+), 25 deletions(-) -- 1.7.11.4
