Created attachment 113388 [details]
dmesg info

It also exist after disable ppgtt.

Please include the part between ------------[ cut here ]------------ and the call trace when reporting bugs with backtraces, i.e.

[   69.351228] WARNING: CPU: 2 PID: 0 at drivers/gpu/drm/i915/i915_irq.c:1750 gen6_rps_irq_handler+0x38/0xfb [i915]()
[   69.351231] GEN9+: unexpected RPS IRQ
[   69.351234] Modules linked in: dm_mod snd_hda_codec_realtek snd_hda_codec_generic ppdev snd_hda_intel pcspkr snd_hda_controller snd_hda_codec snd_hwdep snd_pcm i2c_i801 snd_timer snd soundcore wmi battery parport_pc parport ac acpi_cpufreq i915 button video drm_kms_helper drm cfbfillrect cfbimgblt cfbcopyarea
[   69.351270] CPU: 2 PID: 0 Comm: swapper/2 Tainted: G     U         3.19.0-rc7_drm-intel-nightly_b4442e_20150208+ #198
[   69.351274] Hardware name: Intel Corporation Skylake Client platform/Skylake Y LPDDR3 RVP3, BIOS SKLSE2R1.86C.B069.R00.1501192136 01/19/2015
[   69.351277]  0000000000000000 0000000000000009 ffffffff8179a69b ffff88014e483de8
[   69.351283]  ffffffff8103bdec 000000024e492480 ffffffffa00b4976 ffff880149088068
[   69.351289]  ffff880144380000 0000000000000001 0000000000000010 0000000000000010
[   69.351295] Call Trace:

etc.

Also, please attach dmesg with drm.debug, not i915.debug which does not exist.

/* TODO: RPS on GEN9+ is not supported yet. */

Created attachment 113396 [details]
dmesg with drm.debug info

[  349.308315] ------------[ cut here ]------------
[  349.308362] WARNING: CPU: 0 PID: 5614 at drivers/gpu/drm/i915/i915_irq.c:1750 gen6_rps_irq_handler+0x38/0xfb [i915]()
[  349.308366] GEN9+: unexpected RPS IRQ
[  349.308369] Modules linked in: dm_mod ppdev snd_hda_codec_realtek snd_hda_codec_generic snd_hda_intel snd_hda_controller serio_raw pcspkr snd_hda_codec snd_hwdep snd_pcm snd_timer snd soundcore i2c_i801 wmi battery parport_pc parport ac acpi_cpufreq i915 button video drm_kms_helper drm cfbfillrect cfbimgblt cfbcopyarea
[  349.308411] CPU: 0 PID: 5614 Comm: nautilus Tainted: G        W      3.19.0-rc7_drm-intel-nightly_9583cb_20150205+ #78
[  349.308416] Hardware name: Intel Corporation Skylake Client platform/Skylake Y LPDDR3 RVP3, BIOS SKLSE2R1.86C.B068.R00.1501090908 01/09/2015
[  349.308420]  0000000000000000 0000000000000009 ffffffff8179a69b ffff88014e403de8
[  349.308425]  ffffffff8103bdec ffff8801441bf000 ffffffffa00b4a06 0000000000000246
[  349.308430]  ffff880004d10000 0000000000000001 0000000000000010 0000000000000010
[  349.308435] Call Trace:
[  349.308437]  <IRQ>  [<ffffffff8179a69b>] ? dump_stack+0x40/0x50
[  349.308453]  [<ffffffff8103bdec>] ? warn_slowpath_common+0x98/0xb0
[  349.308475]  [<ffffffffa00b4a06>] ? gen6_rps_irq_handler+0x38/0xfb [i915]
[  349.308481]  [<ffffffff8103be9c>] ? warn_slowpath_fmt+0x45/0x4a
[  349.308509]  [<ffffffffa00c61d2>] ? gen9_write32+0x79/0x29a [i915]
[  349.308529]  [<ffffffffa00b4a06>] ? gen6_rps_irq_handler+0x38/0xfb [i915]
[  349.308550]  [<ffffffffa00b5909>] ? gen8_gt_irq_handler.isra.15+0x192/0x21d [i915]
[  349.308568]  [<ffffffffa00b5a15>] ? gen8_irq_handler+0x81/0x334 [i915]
[  349.308576]  [<ffffffff81072153>] ? handle_irq_event_percpu+0x4f/0x179
[  349.308582]  [<ffffffff810722ab>] ? handle_irq_event+0x2e/0x4f
[  349.308587]  [<ffffffff810746e1>] ? handle_edge_irq+0xbc/0xd1
[  349.308592]  [<ffffffff81004295>] ? handle_irq+0x15/0x20
[  349.308596]  [<ffffffff81003d5a>] ? do_IRQ+0x41/0xc0
[  349.308601]  [<ffffffff817a0baa>] ? common_interrupt+0x6a/0x6a
[  349.308603]  <EOI>  [<ffffffff817a0092>] ? system_call_fastpath+0x12/0x17
[  349.308610] ---[ end trace a217c161d9041962 ]---

(In reply to ye.tian from comment #5)
> Created attachment 113396 [details]
> dmesg with drm.debug info

Please do the same with dmesg starting from early boot. Please add log_buf_len=4M or something to get more info if necessary.

Created attachment 113397 [details]
dmesg info

This is solved by the series from Akash that is on the mailing list. Will review.

It also exists on latest kernel(4.0.0-rc1).

(In reply to ye.tian from comment #10)
> It also exists on latest kernel(4.0.0-rc1).

kernel(4.0.0-rc1)is -nightly: 376ebc108.

Last version of Akash's series is:

  http://lists.freedesktop.org/archives/intel-gfx/2015-March/061479.html

There's also a deadlock with that series, hopefully fixed by that last instalment (Jira: VIZ-5105)

*** Bug 89119 has been marked as a duplicate of this bug. ***

commit 59fb255e234cc8840dcb2a4ac99b698c3ff9f4cd
Author: Akash Goel <akash.goel@intel.com>
Date:   Fri Mar 6 11:07:23 2015 +0530

    drm/i915/skl: Enabling processing of Turbo interrupts

commit aa448626d81d18e74bd24f22eba9a0b0822d3879
Author: Akash Goel <akash.goel@intel.com>
Date:   Fri Mar 6 11:07:23 2015 +0530

    drm/i915/skl: Enabling processing of Turbo interrupts

Verified it on kernel(aa448626d), It does not exists on the latest kernel.