Created attachment 81305 [details] dmesg System Environment: -------------------------- Arch: x86_64 Platform: Ivybridge Kernel: (drm-intel-fixes) 19b2dbde5732170a03bd82cc8bd442cf88d856f7 Bug detailed description: ----------------------------- When run kms_flip subcase with VGA/DP, failed to set mode. It happens on ivybridge with drm-intel-fixes kernel, works well on drm-intel-next-queued kernel. It also works well with HDMI. I test it on igt commit：43a086269795c5b41bdab0dc2c456be0de97e762， It also has this issue. Test it on drm-intel-fixes kernel commit:8abbbaf6adb46157b6bd416f7616b555cc6a332f, It also happens. output： Using monotonic timestamps running testcase: absolute-wf_vblank Beginning absolute-wf_vblank on crtc 3, connector 9 2048x1152 60 2048 2184 2400 2752 1152 1155 1160 1195 0x6 0x48 197000 ............................................................. absolute-wf_vblank on crtc 3, connector 9: PASSED Beginning absolute-wf_vblank on crtc 5, connector 9 2048x1152 60 2048 2184 2400 2752 1152 1155 1160 1195 0x6 0x48 197000 ............................................................. absolute-wf_vblank on crtc 5, connector 9: PASSED Beginning absolute-wf_vblank on crtc 7, connector 9 2048x1152 60 2048 2184 2400 2752 1152 1155 1160 1195 0x6 0x48 197000 failed to set mode (2048x1152@60Hz): Invalid argument Reproduce steps: ---------------------------- 1. ./kms_flip --run-subtest absolute-wf_vblank
Another expected failure for overdriving the third pipe.
commit 6de9736394242f207a43e828a8236e4f63c16e03 Author: Chris Wilson <chris@chris-wilson.co.uk> Date: Mon Jun 24 12:03:54 2013 +0100 tests/kms_flip: Suppress failure from setting mode We may fail to set a mode if it fails some hidden constraints, such as bandwidth on the third pipe. This is expected, so skip testing such modes. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=66111 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
The output shows pass and return value is 0. But "<3>[ 230.406125] [drm:intel_crtc_set_config] *ERROR* failed to set mode on [CRTC:7], err = -22" still exists in dmesg. output Using monotonic timestamps running testcase: absolute-wf_vblank Beginning absolute-wf_vblank on crtc 3, connector 22 2560x1600 60 2560 2608 2640 2720 1600 1603 1609 1646 0x9 0x48 268500 ............................................................. absolute-wf_vblank on crtc 3, connector 22: PASSED Beginning absolute-wf_vblank on crtc 5, connector 22 2560x1600 60 2560 2608 2640 2720 1600 1603 1609 1646 0x9 0x48 268500 ............................................................. absolute-wf_vblank on crtc 5, connector 22: PASSED Beginning absolute-wf_vblank on crtc 7, connector 22 2560x1600 60 2560 2608 2640 2720 1600 1603 1609 1646 0x9 0x48 268500 run: dmesg -r | egrep "<[1-3]>" |grep drm <3>[ 230.406125] [drm:intel_crtc_set_config] *ERROR* failed to set mode on [CRTC:7], err = -22
Created attachment 81381 [details] dmesg
The kernel is reporting that userspace is doing something broken. Userspace sees the error and skips the test.
Patch to tune down the dmesg output: https://patchwork.kernel.org/patch/2775471/
(In reply to comment #6) > Patch to tune down the dmesg output: > > https://patchwork.kernel.org/patch/2775471/ Test with this patch.It has another issue. dmesg -r | egrep "<[1-3]>" |grep drm <3>[ 65.605165] [drm:intel_pipe_config_compare] *ERROR* mismatch in adjusted_mode.flags (expected 1, found 0) <3>[ 70.975957] [drm:intel_pipe_config_compare] *ERROR* mismatch in adjusted_mode.flags (expected 1, found 0)
Created attachment 81436 [details] dmesg with patch
(In reply to comment #7) > (In reply to comment #6) > > Patch to tune down the dmesg output: > > > > https://patchwork.kernel.org/patch/2775471/ > > Test with this patch.It has another issue. > dmesg -r | egrep "<[1-3]>" |grep drm > <3>[ 65.605165] [drm:intel_pipe_config_compare] *ERROR* mismatch in > adjusted_mode.flags (expected 1, found 0) > <3>[ 70.975957] [drm:intel_pipe_config_compare] *ERROR* mismatch in > adjusted_mode.flags (expected 1, found 0) That's just the adjuste_mode flag mismatch bug which we're already tracking in bug #65287
(In reply to comment #3) > The output shows pass and return value is 0. But "<3>[ 230.406125] > [drm:intel_crtc_set_config] *ERROR* failed to set mode on [CRTC:7], err = > -22" still exists in dmesg. > output > Using monotonic timestamps > running testcase: absolute-wf_vblank > Beginning absolute-wf_vblank on crtc 3, connector 22 > 2560x1600 60 2560 2608 2640 2720 1600 1603 1609 1646 0x9 0x48 268500 > ............................................................. > absolute-wf_vblank on crtc 3, connector 22: PASSED > > Beginning absolute-wf_vblank on crtc 5, connector 22 > 2560x1600 60 2560 2608 2640 2720 1600 1603 1609 1646 0x9 0x48 268500 > ............................................................. > absolute-wf_vblank on crtc 5, connector 22: PASSED > > Beginning absolute-wf_vblank on crtc 7, connector 22 > 2560x1600 60 2560 2608 2640 2720 1600 1603 1609 1646 0x9 0x48 268500 > > run: dmesg -r | egrep "<[1-3]>" |grep drm > <3>[ 230.406125] [drm:intel_crtc_set_config] *ERROR* failed to set mode on > [CRTC:7], err = -22 It happens with DP, If attach HDMI, it works well.
Are you sure you've tested -nightly? That ERROR message has really been changed to a debug output, so if you still see it you're booting the wrong kernel ...
(In reply to comment #11) > Are you sure you've tested -nightly? That ERROR message has really been > changed to a debug output, so if you still see it you're booting the wrong > kernel ... I tested it on drm-intel-fixes kernel.It works well on -nightly kernel.
We are around the merge window again, which means -fixes doesn't contain all the fixes ;-) The new -fixes should have the fix.
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.