[Solved] Kernel 3.16.0 and i915 causes Kwin crashes

antoineaquilina · 2014-08-18 09:30:32

Had a similar problem this morning after update to kernel liniux-3.16.1-1 from linux-3.15.4-1.
Gnome just gave me the message "there is a problem...."

Did a manual roll-back to kernel linux-3.15.4-1 and system seems to be ok for now.

Still looking for a proper fix to this.

Corey Goettsch · 2014-08-18 20:16:52

Hey Marioarun,

Thanks for the LTS kernel idea. I updated the kernel last night, and found that I couldn't get into cinnamon this morning. I spent most of the day working in LXDE, but now that I've put it on a long-term support kernel, I'm back in Cinnamon, and my OpenGL works again.

I'll use this kernel until they fix the problem.

spacepluk · 2014-08-20 13:56:17

Hi, I patched 3.16.1 with this and it seems to be working.

diegoviola · 2014-08-20 15:37:10

Deleted.

Last edited by diegoviola (2023-04-17 22:30:00)

orschiro · 2014-08-20 21:43:47

Good to hear that the issue is resolved upstream.

I assume the following error message and respective crashes on boot/standby/hibernation are related to it.

Aug 19 23:07:28 thinkpad kernel: [drm:stop_ring] *ERROR* render ring :timed out trying to stop ring
Aug 19 23:07:28 thinkpad kernel: [drm:stop_ring] *ERROR* render ring :timed out trying to stop ring
Aug 19 23:07:28 thinkpad kernel: [drm:init_ring_common] *ERROR* failed to set render ring head to zero ctl 0001f401 head 0000f17c tail 00000000 start 003eb000
Aug 19 23:07:28 thinkpad kernel: [drm:__i915_drm_thaw] *ERROR* failed to re-initialize GPU, declaring wedged!

Blask · 2014-08-21 07:22:26

I have the same problem, I just install xf86-video-intel-devel in the AUR, and it works fine for me until restart again, then I reinstall xf86-video-intel from offical repo, It works again until I restart again and again

Last edited by Blask (2014-08-21 10:49:04)

ganlu · 2014-08-21 08:07:53

Have been marked as resolved by upstream, will Arch patch related package for quick fix?
https://bugzilla.kernel.org/show_bug.cgi?id=82481

graysky · 2014-08-21 09:01:48

ganlu wrote:

Have been marked as resolved by upstream, will Arch patch related package for quick fix?
https://bugzilla.kernel.org/show_bug.cgi?id=82481

Only if you ask them to do it. Open a fly spray or check to see if there is one already and post the suggestion with link to patch.

kinleyd · 2014-08-22 03:49:16

Having the same problem - intel video and gnome setup here. Any chance of getting the fix in 3.16.1-2 or 3.16.1-3?

Added: I've reverted the kernel so my system is working. Only problem is my last kernel version was 3.14.x which is pretty old. Running that version against all the updates that came with 3.16.1-1 has made my system pretty unstable. So hoping that what's been fixed upstream will come along soon. Hope someone's made that flyspray request to the Arch maintainers!

No problems on my other system with the nvidia gpu.

Last edited by kinleyd (2014-08-22 04:34:15)

butler360 · 2014-08-22 04:11:38

graysky wrote:

ganlu wrote:
Have been marked as resolved by upstream, will Arch patch related package for quick fix?
https://bugzilla.kernel.org/show_bug.cgi?id=82481
Only if you ask them to do it. Open a fly spray or check to see if there is one already and post the suggestion with link to patch.

Just curious, Graysky, since you are aware of this issue is it in your CK kernel? Just wondering how you work with these kinds of problems.

graysky · 2014-08-22 18:52:01

@butler- I usually follow Arch's lead. Can you confirm that this patch works?

fqqbar · 2014-08-23 17:16:06

mainline kernel 3.17-rc1 from AUR fixed it for me.

butler360 · 2014-08-24 00:57:47

graysky wrote:

@butler- I usually follow Arch's lead. Can you confirm that this patch works?

Thanks for the answer.

I don't have access to that computer at the moment. I will try it if I do.

graysky · 2014-08-24 06:44:41

As pointed out by others, this is known to fix the issue upstream. Can anyone confirm that the patch indeed solves the issue?

Patch: https://bugs.freedesktop.org/attachment.cgi?id=104224

If so, and given that the developer himself mentioned on Aug 7th, that it will get backported to stable kernels eventually, I or someone else should open flyspray asking for its inclusion in the ARCH kernel. I can't verify the fix on my hardware due to my inability to take down the box right now. Can someone else?

graysky · 2014-08-24 21:45:52

OK... I managed to bring the box down for the update. I have rebooted it 4 times now after patching the 3.16.1-1 kernel with the aforementioned patch and have no experienced the errors.

EDIT: https://bugs.archlinux.org/task/41683

Last edited by graysky (2014-08-24 21:54:39)

kinleyd · 2014-08-25 04:21:32

Thank you graysky!

graysky · 2014-08-25 07:06:32

Ack... this one was reported by others around Aug 1st. Anyway, the Arch devs are aware of it.

https://bugs.archlinux.org/task/41045
https://bugs.archlinux.org/task/41573

orschiro · 2014-08-27 05:47:09

Just a quick question to clarify whether this issue is related:

If you put your machine into suspend/hibernation, does Xorg crash on resume?

For me it does, putting me back to tty, but I am not sure if it is related to the error message:

Aug 27 07:43:45 thinkpad kernel: [drm:stop_ring] *ERROR* render ring :timed out trying to stop ring
Aug 27 07:43:45 thinkpad kernel: [drm:stop_ring] *ERROR* render ring :timed out trying to stop ring
Aug 27 07:43:45 thinkpad kernel: [drm:init_ring_common] *ERROR* failed to set render ring head to zero ctl 0001f401 head 0000f17c tail 00000000 start 003eb000
Aug 27 07:43:45 thinkpad kernel: [drm:__i915_drm_thaw] *ERROR* failed to re-initialize GPU, declaring wedged!

Last edited by orschiro (2014-08-27 05:48:25)

somehost · 2014-09-02 12:52:34

This is my first tour on Arch and I happened to install it on a Dell Latitude E6400 that is running the Intel GM45 chipset when these issues cropped up so I have been pulling my hair out and RTFM'ing like there is no tomorrow.

It has been an awesome learning experience which has pushed the boundaries of what I know about Linux which is why I switched to Arch so I can't complain.

Having said that, I am frustrated as hell with the i915 regression issue and have bitten the bullet and built the linux-mainline kernel from the AUR.

3.17.0-1-mainline #1 SMP PREEMPT Mon Sep 1 23:09:02 SAST 2014 x86_64 GNU/Linux

I am no longer experiencing wedges on boot, hung display with active mouse on resume or crashes.

This is the only change that I have made outside of the Arch repo's which seems to have sorted out the problems that I have had since my first install.

graysky · 2014-09-02 14:07:55

@somehost - can u confirm that the upstream patch against 3.16.x fixes your problem? It could be that there is not a need to run the unstable 3.17 code base.

lachokds · 2014-09-02 17:17:31

graysky wrote:

@butler- I usually follow Arch's lead. Can you confirm that this patch works?

I've also had this problem on my computer, which is an Acer eMachies D728-4693 laptop. I applied the patch and rebuilt my kernel (originally 3.16.1) with ABS, but it hasn't solved the problem yet.

After patching and recompiling, dmesg says this

$ dmesg | grep acer
[    0.000000] ACPI: SLIC 0x0000000077BFEE8A 000176 (v01 ACRSYS ACRPRDCT 06040000 acer 00000000)
[    8.736144] acer_wmi: Acer Laptop ACPI-WMI Extras
[    8.736226] acer_wmi: Function bitmap for Communication Button: 0x1
[    8.736322] acer_wmi: Disabling ACPI video driver

Also, here's a few things I found on my journal

$ journalctl -xn 1000 | grep acpi
sep 02 11:04:02 arch kernel: Command line: BOOT_IMAGE=/boot/vmlinuz-linux root=UUID=1dda5d71-fbde-4e86-b34c-6c0793214041 rw quiet quiet splash acpi_osi=Linux acpi_backlight=vendor
sep 02 11:04:02 arch kernel: ACPI: LAPIC (acpi_id[0x00] lapic_id[0x00] enabled)
sep 02 11:04:02 arch kernel: ACPI: LAPIC (acpi_id[0x01] lapic_id[0x01] enabled)
sep 02 11:04:02 arch kernel: ACPI: LAPIC_NMI (acpi_id[0x00] high edge lint[0x1])
sep 02 11:04:02 arch kernel: ACPI: LAPIC_NMI (acpi_id[0x01] high edge lint[0x1])
sep 02 11:04:02 arch kernel: Kernel command line: BOOT_IMAGE=/boot/vmlinuz-linux root=UUID=1dda5d71-fbde-4e86-b34c-6c0793214041 rw quiet quiet splash acpi_osi=Linux acpi_backlight=vendor
sep 02 11:04:02 arch kernel: acpiphp: ACPI Hot Plug PCI Controller Driver version: 0.5
sep 02 11:04:02 arch kernel: acpi PNP0A08:00: _OSC: OS supports [ExtendedConfig ASPM ClockPM Segments MSI]
sep 02 11:04:02 arch kernel: acpi PNP0A08:00: _OSC failed (AE_NOT_FOUND); disabling ASPM
sep 02 11:04:03 arch kernel: ACPI: acpi_idle registered with cpuidle

$ journalctl -xn 1000 | grep acer
sep 01 11:26:09 arch kernel: acer_wmi: Acer Laptop ACPI-WMI Extras
sep 01 11:26:09 arch kernel: acer_wmi: Function bitmap for Communication Button: 0x1
sep 01 11:26:09 arch kernel: acer_wmi: Disabling ACPI video driver
sep 02 11:04:02 arch kernel: ACPI: SLIC 0x0000000077BFEE8A 000176 (v01 ACRSYS ACRPRDCT 06040000 acer 00000000)
sep 02 11:04:05 arch kernel: acer_wmi: Acer Laptop ACPI-WMI Extras
sep 02 11:04:05 arch kernel: acer_wmi: Function bitmap for Communication Button: 0x1
sep 02 11:04:05 arch kernel: acer_wmi: Disabling ACPI video driver

$ journalctl -xn 1000 | grep vga
sep 01 11:26:08 arch kernel: vgaarb: device changed decodes: PCI:0000:00:02.0,olddecodes=io+mem,decodes=io+mem:owns=io+mem
sep 02 11:04:02 arch kernel: vgaarb: device added: PCI:0000:00:02.0,decodes=io+mem,owns=io+mem,locks=none
sep 02 11:04:02 arch kernel: vgaarb: loaded
sep 02 11:04:02 arch kernel: vgaarb: bridge control possible 0000:00:02.0
sep 02 11:04:04 arch kernel: vgaarb: device changed decodes: PCI:0000:00:02.0,olddecodes=io+mem,decodes=io+mem:owns=io+mem

$ journalctl -xn 1000 | grep intel
sep 01 11:26:08 arch kernel: fb: switching to inteldrmfb from VESA VGA
sep 01 11:26:09 arch kernel: fbcon: inteldrmfb (fb0) is primary device
sep 01 11:26:09 arch kernel: i915 0000:00:02.0: fb0: inteldrmfb frame buffer device
sep 01 11:26:09 arch kernel: snd_hda_intel 0000:00:1b.0: irq 45 for MSI/MSI-X
sep 02 11:04:02 arch kernel: intel_idle: does not run on family 6 model 23
sep 02 11:04:03 arch kernel: agpgart-intel 0000:00:00.0: Intel GM45 Chipset
sep 02 11:04:03 arch kernel: agpgart-intel 0000:00:00.0: detected gtt size: 2097152K total, 262144K mappable
sep 02 11:04:03 arch kernel: agpgart-intel 0000:00:00.0: detected 131072K stolen memory
sep 02 11:04:03 arch kernel: agpgart-intel 0000:00:00.0: AGP aperture is 256M @ 0xd0000000
sep 02 11:04:04 arch kernel: fb: switching to inteldrmfb from VESA VGA
sep 02 11:04:05 arch kernel: fbcon: inteldrmfb (fb0) is primary device
sep 02 11:04:05 arch kernel: i915 0000:00:02.0: fb0: inteldrmfb frame buffer device
sep 02 11:04:05 arch kernel: snd_hda_intel 0000:00:1b.0: irq 45 for MSI/MSI-X

So, it seems that it's somehow disabling the acer_wmi module and insisting on switching to the Intel one, which disables the Fn keys' functionality regarding screen brightness. Does anyone have any other ideas/patches for solving this?

diegoviola · 2014-09-04 09:56:36

Deleted.

Last edited by diegoviola (2023-04-17 22:30:06)

marioraun · 2014-09-09 19:38:15

Fixed in 3.16.2-1 just installed

Frontier · 2014-09-10 08:05:09

marioraun wrote:

Fixed in 3.16.2-1 just installed

Here on my Packard Bell vr46 (Intel GMA4500) Kwin still crashes; it's impossible to enable OpenGL 3.1 (as used previously).
I had to change the acceleration method from SNA to glamor and switch to OpenGL 2.0 in order for KDE/Kwin to start without crashing.

The situation improved - I could boot only after disabling desktop effects - but the issue more or less still exists; with Kernel 3.15.x I could boot with SNA/OpenGL 3.1 without issues, now I can't.

orschiro · 2014-09-10 10:16:43

Same for me. Kernel 3.16.2-1 still crashes.

Fortunately, user C. Simon (mentalis) provides a pre-compiled kernel package.

Source: https://bugs.archlinux.org/themes/ArchLinux/comment.png

Download: https://www.bloodycloud.net/public.php? … f1703d8e5c

Arch Linux

#26 2014-08-18 09:30:32

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#27 2014-08-18 20:16:52

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#28 2014-08-20 13:56:17

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#29 2014-08-20 15:37:10

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#30 2014-08-20 21:43:47

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#31 2014-08-21 07:22:26

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#32 2014-08-21 08:07:53

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#33 2014-08-21 09:01:48

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#34 2014-08-22 03:49:16

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#35 2014-08-22 04:11:38

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#36 2014-08-22 18:52:01

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#37 2014-08-23 17:16:06

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#38 2014-08-24 00:57:47

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#39 2014-08-24 06:44:41

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#40 2014-08-24 21:45:52

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#41 2014-08-25 04:21:32

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#42 2014-08-25 07:06:32

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#43 2014-08-27 05:47:09

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#44 2014-09-02 12:52:34

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#45 2014-09-02 14:07:55

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#46 2014-09-02 17:17:31

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#47 2014-09-04 09:56:36

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#48 2014-09-09 19:38:15

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#49 2014-09-10 08:05:09

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

#50 2014-09-10 10:16:43

Re: [Solved] Kernel 3.16.0 and i915 causes Kwin crashes

Board footer