Created attachment 101210 [details] dmesg ==System Environment== -------------------------- Regression: Yes. This bug only find on -nightly branch. Non-working platforms: HSW ==kernel== -------------------------- origin/drm-intel-nightly: e9ce9f961282205b5954a0da4e48e96884190a77(fails) drm-intel-nightly: 2014y-06m-16d-21h-22m-19s integration manifest origin/drm-intel-next-queued: 8c0f92e1f87149bd2429015c9d0600a13a34d607(tried 5 times, didn't find this issue) drm/i915: Fix comment about our plane remapping on gen2/3 origin/drm-intel-fixes: 223a6f2b975ab35d93270ea1d4fb6e0ac6b27fe6(tried 5 times, didn't find this issue) drm/i915/bdw: remove erroneous chv specific workarounds from bdw code ==Bug detailed description== ----------------------------- igt/kms_flip causes "WARNING: CPU: 5 PID: 4368 at drivers/gpu/drm/i915/intel_pm.c:6174 check_power_well_state+0x69/0x6f [i915]()" Case list: igt/kms_flip/vblank-vs-dpms-suspend igt/kms_flip/vblank-vs-dpms-suspend-interruptible igt/kms_flip/vblank-vs-modeset-suspend igt/kms_flip/vblank-vs-modeset-suspend-interruptible igt/kms_flip/vblank-vs-suspend igt/kms_flip/vblank-vs-suspend-interruptible Output: [root@x-hsw26 tests]# ./kms_flip --run-subtest vblank-vs-dpms-suspend IGT-Version: 1.7-gd123de6 (x86_64) (Linux: 3.15.0-rc8_drm-intel-nightly_e9ce9f_20140617+ x86_64) Using monotonic timestamps Beginning vblank-vs-dpms-suspend on crtc 6, connector 25 1920x1080 60 1920 2008 2052 2200 1080 1082 1087 1125 0x5 0x48 148500 .rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Jun 17 21:28:21 2014 .rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Jun 17 21:28:54 2014 vblank-vs-dpms-suspend on crtc 6, connector 25: PASSED Beginning vblank-vs-dpms-suspend on crtc 10, connector 25 1920x1080 60 1920 2008 2052 2200 1080 1082 1087 1125 0x5 0x48 148500 .rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Jun 17 21:29:27 2014 .rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Jun 17 21:30:00 2014 vblank-vs-dpms-suspend on crtc 10, connector 25: PASSED Beginning vblank-vs-dpms-suspend on crtc 14, connector 25 1920x1080 60 1920 2008 2052 2200 1080 1082 1087 1125 0x5 0x48 148500 .rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Jun 17 21:30:33 2014 .rtcwake: wakeup from "mem" using /dev/rtc0 at Tue Jun 17 21:31:06 2014 vblank-vs-dpms-suspend on crtc 14, connector 25: PASSED Subtest vblank-vs-dpms-suspend: SUCCESS ==Reproduce steps== ---------------------------- 1. ./kms_flip --run-subtest vblank-vs-dpms-suspend
DMESG shows: [ 100.529130] WARNING: CPU: 5 PID: 4368 at drivers/gpu/drm/i915/intel_pm.c:6174 check_power_well_state+0x69/0x6f [i915]() [ 100.529131] state mismatch for 'display' (always_on 0 hw state 0 use-count 3 disable_power_well 1 [ 100.529138] Modules linked in: dm_mod snd_hda_codec_realtek snd_hda_codec_generic snd_hda_codec_hdmi iTCO_wdt iTCO_vendor_support pcspkr serio_raw i2c_i801 firewire_ohci lpc_ich mfd_core snd_hda_intel snd_hda_controller firewire_core crc_itu_t snd_hda_codec snd_hwdep snd_pcm snd_timer snd soundcore battery wmi acpi_cpufreq i915 video button drm_kms_helper drm [ 100.529140] CPU: 5 PID: 4368 Comm: kworker/u16:148 Tainted: G W 3.15.0-rc8_drm-intel-nightly_e9ce9f_20140617+ #3634 [ 100.529140] Hardware name: /DZ87KLT75K, BIOS KLZ8711D.86A.0336.2013.0516.1957 05/16/2013 [ 100.529143] Workqueue: events_unbound async_run_entry_fn [ 100.529144] 0000000000000000 0000000000000009 ffffffff81772bca ffff880255dafc68 [ 100.529145] ffffffff81036e54 0000000000000286 ffffffffa0069e8b ffff88025661c000 [ 100.529146] ffffffffa00fe338 ffffffffa00fe338 0000000000000001 0000000000800000 [ 100.529147] Call Trace: [ 100.529150] [<ffffffff81772bca>] ? dump_stack+0x41/0x51 [ 100.529153] [<ffffffff81036e54>] ? warn_slowpath_common+0x78/0x90 [ 100.529156] [<ffffffffa0069e8b>] ? check_power_well_state+0x69/0x6f [i915] [ 100.529158] [<ffffffff81036f04>] ? warn_slowpath_fmt+0x45/0x4a [ 100.529162] [<ffffffffa0069e8b>] ? check_power_well_state+0x69/0x6f [i915] [ 100.529166] [<ffffffffa006df31>] ? intel_display_power_get+0x82/0xb9 [i915] [ 100.529173] [<ffffffffa00aad1c>] ? intel_display_set_init_power+0x22/0x34 [i915] [ 100.529177] [<ffffffffa006ddff>] ? intel_power_domains_init_hw+0x21/0x6b [i915] [ 100.529180] [<ffffffffa0064607>] ? i915_resume_early+0x3d/0x3d [i915] [ 100.529183] [<ffffffffa0064135>] ? i915_drm_thaw_early+0x40/0x46 [i915] [ 100.529186] [<ffffffff813d129b>] ? dpm_run_callback+0x30/0x6b [ 100.529187] [<ffffffff813d1950>] ? device_resume_early+0x130/0x172 [ 100.529189] [<ffffffff813d1b78>] ? async_resume_early+0x14/0x38 [ 100.529189] [<ffffffff81053d9f>] ? async_run_entry_fn+0x55/0x10b [ 100.529192] [<ffffffff8104aade>] ? process_one_work+0x1e2/0x363 [ 100.529193] [<ffffffff8104b385>] ? worker_thread+0x1c7/0x2bc [ 100.529194] [<ffffffff8104b1be>] ? manage_workers.isra.23+0x1b5/0x1b5 [ 100.529196] [<ffffffff8104fb32>] ? kthread+0xc5/0xcd [ 100.529197] [<ffffffff8104fa6d>] ? kthread_freezable_should_stop+0x40/0x40 [ 100.529199] [<ffffffff8177e9bc>] ? ret_from_fork+0x7c/0xb0 [ 100.529200] [<ffffffff8104fa6d>] ? kthread_freezable_should_stop+0x40/0x40
I have tested it on the latest nightly branch(drm-intel-nightly_eb2af9_20140730)and I didn't find the Call Trace.
Tossing the bug to Imre to check. I guess we can close it if it no longer happens.
I can't reproduce it either. I assume it was fixed by: commit b04c5bd6fda54703e56f29569e4bca489d6c5a5c Author: Borun Fu <borun.fu@intel.com> Date: Sat Jul 12 10:02:27 2014 +0530 drm/i915: Power gating display wells during i915_pm_suspend Setting to fixed based on this.
The warning unable to reproduce on latest -nightly(723dbbfb6ce02b5662c0c4cac9be3bcdfbb11d35). But the test was failed, which tracked by bug 79052 [root@x-hsw24 tests]# ./kms_flip --run-subtest vblank-vs-dpms-suspend IGT-Version: 1.7-geda904c (x86_64) (Linux: 3.16.0_drm-intel-nightly_723dbb_20140813+ x86_64) Using monotonic timestamps Beginning vblank-vs-dpms-suspend on crtc 7, connector 18 1920x1080 60 1920 2008 2052 2200 1080 1084 1089 1125 0x5 0x48 148500 .rtcwake: wakeup from "mem" using /dev/rtc0 at Wed Aug 13 06:44:25 2014 .rtcwake: wakeup from "mem" using /dev/rtc0 at Wed Aug 13 06:44:57 2014 Test assertion failure function check_state, file kms_flip.c:557: Failed assertion: timercmp(&es->last_received_ts, &es->current_ts, <) vblank ts before the vblank was issued! timerdiff -1s, 987323us Subtest vblank-vs-dpms-suspend: FAIL [root@x-hsw24 tests]# dmesg -r|egrep "<[1-4]>"|grep drm [root@x-hsw24 tests]#
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.