Created attachment 115791 [details] dmesg info with drm.debug=0x1e log_buf_len=16M ==System Environment== ----------------------------------------------------- Regression: not sure Non-working platforms: BDW ==Kernel== -------------------------------------------------- commit 6cf0fff478fa64991c2b7f2a44f2815d7a43a0b5 Author: Daniel Vetter <daniel.vetter@ffwll.ch> Date: Wed May 13 11:29:25 2015 +0200 drm-intel-nightly: 2015y-05m-13d-09h-28m-49s UTC integration manifest ==Bug detailed description== -------------------------------------------------- Dpy_misc_s3d fails on the latest nightly kernel, It does not exists on the latest fixes branch kernel. I will try to find good commit on next-queued branch. ==Reproduce steps== ---------------------------- 1, ./testdisplay -3
on the next-quened branch. Bisect pending... commit[1] good commit[2] bad [1]commit c7240c3bc5d6610b42dbb10fda71bbbf1dad5515 Author: Rodrigo Vivi <rodrigo.vivi@intel.com> Date: Fri Apr 10 11:15:10 2015 -0700 drm/i915: PSR VLV: Add single frame update. According to spec: "In PSR HW or SW mode, SW set this bit before writing registers for a flip. It will be self-clear when it gets to the PSR active state." [2]commit 097f8261ddda4b1896dd335dec95dedeecfeaa1b Author: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Date: Thu Apr 16 21:11:08 2015 +0800 drm/i915/audio: remove duplicated include from intel_audio.c Remove duplicated include.
Please bisect.
By bisected, shows the first bad commit is d328c9d. commit d328c9d78d64ca11e744fe227096990430a88477 Author: Daniel Vetter <daniel.vetter@ffwll.ch> AuthorDate: Fri Apr 10 16:22:37 2015 +0200 Commit: Daniel Vetter <daniel.vetter@ffwll.ch> CommitDate: Thu Apr 16 11:20:28 2015 +0200 drm/i915: Select starting pipe bpp irrespective or the primary plane Since universal planes the primary plane might not be around, and it's kinda silly to restrict the pipe bpp to the primary plane if we might end up displaying a 10bpc video overlay. And with atomic we might very well enable a pipe without a primary plane. So just use the platform max as a starting point and then restrict appropriately. Of course this is all still a bit moot as long as we artificially compress everything to max 8bpc because we don't use the hi-bpc gamma tables.
Created attachment 115878 [details] [review] Patch that limits bpc to 10 Could you test if the problem still happens with the attached patch?
(In reply to Ander Conselvan de Oliveira from comment #4) > Created attachment 115878 [details] [review] [review] > Patch that limits bpc to 10 > > Could you test if the problem still happens with the attached patch? This problem still happens with this patch.
Could you attach the dmesg with with drm.debug=0x1e log_buf_len=16M while running with that patch applied?
Created attachment 115892 [details] dmesg info with drm.debug=0x1e log_buf_len=16M
Created attachment 115896 [details] [review] Second attempt I didn't notice the HDMI code sneaks behind our backs and increases pipe bpp from 10 to 12 under some conditions. Could you test with the attached patch?
Test it on the latest nightly kernel with this patch, this problem does not exists.
(In reply to ye.tian from comment #9) > Test it on the latest nightly kernel with this patch, this problem does not > exists. Ok, so the problem is i915 is now trying to use 12 bpc HDMI which is broken. Could you check if the bug exists if you apply the following patch series? That is supposed to actually fix the 12 bpc HDMI support instead of disabling it. http://lists.freedesktop.org/archives/intel-gfx/2015-May/065857.html
(In reply to Ander Conselvan de Oliveira from comment #10) > (In reply to ye.tian from comment #9) > > Test it on the latest nightly kernel with this patch, this problem does not > > exists. > > Ok, so the problem is i915 is now trying to use 12 bpc HDMI which is broken. > > Could you check if the bug exists if you apply the following patch series? > That is supposed to actually fix the 12 bpc HDMI support instead of > disabling it. > > http://lists.freedesktop.org/archives/intel-gfx/2015-May/065857.html I did not find the following patch, do you re-send it? drm/i915: Send GCP infoframes for deep color HDMI sinks The following two patches are repeated. http://lists.freedesktop.org/archives/intel-gfx/2015-May/065858.html http://lists.freedesktop.org/archives/intel-gfx/2015-May/065881.html
Could you try the branch ville-hdmi-12bpc-fixes from https://github.com/anderco/linux.git ? I applied those patches there on top of -nightly.
(In reply to Ander Conselvan de Oliveira from comment #12) > Could you try the branch ville-hdmi-12bpc-fixes from > https://github.com/anderco/linux.git ? I applied those patches there on top > of -nightly. Test it on kernel (d417fc), this problem does not exists.
*** Bug 90675 has been marked as a duplicate of this bug. ***
commit 5e3daaca09f5158eff9c92290faa1d2001ecc6e4 Author: Daniel Vetter <daniel.vetter@ffwll.ch> Date: Thu May 28 09:38:45 2015 +0200 drm/i915: Disable 12bpc hdmi for now in -nightly.
Test it on the latest kernel (drm-intel-nightly_b44f67_20150529), this problem does not exists. So verified.
Closing verified+fixed.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.