Created attachment 91185 [details] dmesg System Environment: -------------------------- Platform: Ironlake Kernel: (drm-intel-fixes)a885b3ccc74d8e38074e1c43a47c354c5ea0b01e Bug detailed description: ------------------------- It fails on Ironlake with -fixes kernel. It works well on -nightly branch. Fellowing cases also fail: igt/kms_flip/flip-vs-panning-interruptible igt/kms_flip/rcs-wf_vblank-vs-modeset The latest known good commit:acc240d41ea1ab9c488a79219fb313b5b46265ae The latest known bad commit: a885b3ccc74d8e38074e1c43a47c354c5ea0b01e output: IGT-Version: 1.5-g4f28614 (i686) (Linux: 3.13.0-rc3_drm-intel-fixes_a885b3_20131224+ i686) Using monotonic timestamps Beginning flip-vs-panning on crtc 3, connector 10 1366x768 60 1366 1398 1422 1426 768 771 775 806 0x9 0x48 69000 .....unexpected flip seq 27465, should be >= 27466 Subtest flip-vs-panning: FAIL Reproduce steps: ---------------- 1. ./kms_flip --run-subtest flip-vs-panning
It has another Bug 65558.
Hi Can you please finish bisecting so we can discover which commit exactly broke the feature? The information about latest known good/bad commit looks confusing. Also, the dmesg file seems too short, perhaps it's incomplete. Thanks, Paulo
Ping again about the bisect result ... And like Paulo noticed the dmesg doesn't contain any i915 message, so seems cut short. Please attach a new one.
Created attachment 91735 [details] full dmesg Bisect is in process. The latest known bad commit: 2dc1733fd4ebd4a67c9312d85f1dbc2ac2ffeb26 The latest known good commit: 05454c26eb3587b56abc5eb139797ac5afb6d77a
Try to bisect it. It randomly fails on many commits.
So even the old kernels randomly fail? Just wondering whether we should take out the regression tag ... Or could this perhaps be a regression in i-g-t?
(In reply to comment #6) > So even the old kernels randomly fail? Just wondering whether we should take > out the regression tag ... > > Or could this perhaps be a regression in i-g-t? Test on latest kernel and old igt commit, It also fails. It's not igt regression. Run on earlier kernel commit(05454c26eb3587b56abc5eb139797ac5afb6d77a), It fails 2 in 10 runs. Run on latest kernel commit, It fails 20 in 20 runs.
Hua, can you help to retest this issue.
Test it on latest -nightly kernel, it still fails. output: IGT-Version: 1.6-g252d115 (x86_64) (Linux: 3.15.0-rc3_drm-intel-nightly_f79ba7_20140519+ x86_64) Using monotonic timestamps Beginning flip-vs-panning on crtc 5, connector 10 1280x1024 60 1280 1328 1440 1688 1024 1025 1028 1066 0x5 0x48 108000 ...................................................Test assertion failure function check_state, file kms_flip.c:564: Last errno: 0, Success Failed assertion: es->current_seq - (es->last_seq + o->seq_step) <= 1UL << 23 unexpected flip seq 4330, should be >= 4331 Subtest flip-vs-panning: FAIL
commit de3f30c53eff5ce58c04e8491bbae545aa4e6188 Author: Ville Syrjälä <ville.syrjala@linux.intel.com> Date: Tue Apr 15 21:41:34 2014 +0300 drm/i915: Fix mmio vs. CS flip race on ILK+
It still fails on latest -fixes kernel. output: IGT-Version: 1.6-gff3c122 (x86_64) (Linux: 3.15.0-rc5_drm-intel-fixes_f93e94_20140523+ x86_64) Using monotonic timestamps Beginning flip-vs-panning on crtc 5, connector 10 1280x1024 60 1280 1328 1440 1688 1024 1025 1028 1066 0x5 0x48 108000 .......Test assertion failure function check_state, file kms_flip.c:565: Last errno: 0, Success Failed assertion: es->current_seq - (es->last_seq + o->seq_step) <= 1UL << 23 unexpected flip seq 5317, should be >= 5318 Subtest flip-vs-panning: FAIL
Verified.Fixed.
Closing old verified.
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.