Created attachment 115470 [details] dmesg System Environment -------------------------- Regression: Yes. Good commit: d600654ab94b325f253e267422dcf60302120ea0 Non-working platforms: SandyBridge Bug detailed description -------------------------- 1.echo mem > /sys/power/state 2.resume the machine 3.repeat the above steps for 5 times => The system fail to suspend from second time. Dmesg shows: [ 101.825551] Call Trace: [ 101.825556] <IRQ> [<ffffffff81795847>] ? dump_stack+0x40/0x50 [ 101.825559] [<ffffffff8103bd5a>] ? warn_slowpath_common+0x98/0xb0 [ 101.825561] [<ffffffff81700aef>] ? dev_watchdog+0x171/0x1d7 [ 101.825563] [<ffffffff8103bdb7>] ? warn_slowpath_fmt+0x45/0x4a [ 101.825564] [<ffffffff81700aef>] ? dev_watchdog+0x171/0x1d7 [ 101.825566] [<ffffffff8170097e>] ? dev_graft_qdisc+0x66/0x66 [ 101.825568] [<ffffffff8107bbff>] ? call_timer_fn+0x30/0xd0 [ 101.825570] [<ffffffff8170097e>] ? dev_graft_qdisc+0x66/0x66 [ 101.825571] [<ffffffff8107c39d>] ? run_timer_softirq+0x1c0/0x21b [ 101.825574] [<ffffffff81087bb4>] ? tick_sched_timer+0x2f/0x55 [ 101.825576] [<ffffffff8103e9d9>] ? __do_softirq+0xd6/0x22d [ 101.825577] [<ffffffff8103ec64>] ? irq_exit+0x34/0x78 [ 101.825580] [<ffffffff81026bb6>] ? smp_apic_timer_interrupt+0x39/0x43 [ 101.825582] [<ffffffff8179beea>] ? apic_timer_interrupt+0x6a/0x70 [ 101.825585] <EOI> [<ffffffff816af580>] ? ladder_select_state+0x161/0x161 [ 101.825586] [<ffffffff816af580>] ? ladder_select_state+0x161/0x161 [ 101.825589] [<ffffffff8106541e>] ? cpu_startup_entry+0x262/0x350 [ 101.825591] ---[ end trace 8f778530c6e3a543 ]--- [ 101.833536] [drm] stuck on render ring [ 101.833919] [drm] GPU HANG: ecode 6:0:0x00d3ffe0, reason: Ring hung, action: reset [ 101.833941] [drm:i915_reset_and_wakeup] resetting chip [ 101.833963] i915 0000:00:02.0: GEM idle failed, resume might fail [ 101.833975] pci_pm_suspend(): i915_pm_suspend+0x0/0x3f [i915] returns -11 [ 101.833978] dpm_run_callback(): pci_pm_suspend+0x0/0xf6 returns -11 [ 101.833981] PM: Device 0000:00:02.0 failed to suspend async: error -11 [ 101.834020] PM: Some devices failed to suspend, or early wake event detected [ 101.834197] rtc_cmos 00:02: System wakeup disabled by ACPI [ 101.834779] serial 00:04: activated [ 101.835979] drm/i915: Resetting chip after gpu hang [ 101.855657] sd 0:0:0:0: [sda] Starting disk [ 101.858584] tpm_tis 00:05: TPM is disabled/deactivated (0x7) [ 101.994590] PM: resume of devices complete after 160.383 msecs [ 102.029304] PM: Finishing wakeup. [ 102.029305] Restarting tasks ... done. ---------------------------------------------------------- ==Reproduce steps== ---------------------------- 1.echo mem > /sys/power/state 2.resume the machine 3.repeat step 1-2 5 times. Result: The system fail to suspend from second time.
Could you please bisect?
*** Bug 90252 has been marked as a duplicate of this bug. ***
Test with the latest drm-intel-nightly kernel(commit 2a36ab374449e3794b4be6be01f1a47cb6ee5b83)on SNB platform,the issue still exist. Bisect result: Bisect shows 5c5f645773b6d147bf68c350674dc3ef4f8de83d is the first bad commit. commit 5c5f645773b6d147bf68c350674dc3ef4f8de83d Author: Daniel Vetter <daniel.vetter@ffwll.ch> Date: Tue Apr 14 17:35:14 2015 +0200 drm/i915: Unify aliasing ppgtt handling With the dynamic pagetable alloc code aliasing ppgtt special-cases where again mixed in all over the place with the low-level init code. Extract the va preallocation and clearing again into the common code where aliasing ppgtt gets set up. Note that with this we don't set the size of the aliasing ppgtt to the size of the parent ggtt address space. Which isn't required at all since except for the ppgtt setup/cleanup code no one ever looks at this. Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Could you try the drm-intel-nightly branch from: git://anongit.freedesktop.org/drm-intel ?
Closing for inactivity from submitter. If problem arise, please file new bug with all information needed. Thank you.
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.