Summary: | [ivb rcs flips 3.14] GPU sporadically hangs during normal desktop usage | ||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Product: | DRI | Reporter: | Andreas Hermann <anhermann> | ||||||||||||||
Component: | DRM/Intel | Assignee: | Chris Wilson <chris> | ||||||||||||||
Status: | CLOSED FIXED | QA Contact: | Intel GFX Bugs mailing list <intel-gfx-bugs> | ||||||||||||||
Severity: | normal | ||||||||||||||||
Priority: | medium | CC: | adrin.jalali, anarsoul, andreas.kurz, andrejsstepanovs, athikom.v, brstrand, bugs.freedesktop.org, bugzilla, chrassig, cristiklein, dgilbert, diniz.bortolotto, django.janny, freedesktop, freeniki, fuujuhi-ml1, fvisin, gary.c.wang, glennpowers, guillecorrea, intel-gfx-bugs, jason, jean-paul, jeff34tx, juliansuhr, k0l0b0k.void, kamuz2000, kees.bakker, lapenna.anthony, libbkmz.dev, lkml, mails.bugs.freedesktop.org-2009, mar.kolya, matijs, max.amram, metan, milan, nemesis, patrik.lundquist, pierg75, pukayana, rashkur, rbbeltran.09, seo.boon.ng, simon, sven.koehler, tgrabiec, the.warl0ck.1989 | ||||||||||||||
Version: | XOrg git | ||||||||||||||||
Hardware: | All | ||||||||||||||||
OS: | Linux (All) | ||||||||||||||||
Whiteboard: | |||||||||||||||||
i915 platform: | ALL | i915 features: | |||||||||||||||
Attachments: |
|
It's the rcs flip dragon. This should be improved in recent kernels. *** Bug 78120 has been marked as a duplicate of this bug. *** *** Bug 78195 has been marked as a duplicate of this bug. *** *** Bug 78243 has been marked as a duplicate of this bug. *** (In reply to comment #1) > It's the rcs flip dragon. This should be improved in recent kernels. And in which kernel should this be fixed so I can download that one? "rcs flip dragon" on google didnt result in any kernel change log hit... *** Bug 78286 has been marked as a duplicate of this bug. *** Created attachment 98467 [details] Another crash dump Attaching my crash dump in case this is of interest. (initially attached to https://bugs.freedesktop.org/show_bug.cgi?id=76956) If this problem is fixed in recent kernels, it would be interesting to get a clear indication where (which package/version, at least which commit) it is fixed. This way people at Canonical can back-port this fix to one of the LTS supported kernel. Same for other distros I guess. *** Bug 78551 has been marked as a duplicate of this bug. *** Created attachment 98837 [details]
Ubuntu 14.04 GPU dump
Seen on Ubuntu 14.04:
Linux moon 3.13.0-24-generic #47-Ubuntu SMP Fri May 2 23:30:00 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux
*** Bug 78445 has been marked as a duplicate of this bug. *** Created attachment 99397 [details]
GPU (Intel) crash dump
Hi,
The same thing is happening to me, also. Ubuntu 14.04, uname -a output:
Linux misel-pc 3.13.0-24-generic #47-Ubuntu SMP Fri May 2 23:30:00 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux
(In reply to comment #11) > Created attachment 99397 [details] > GPU (Intel) crash dump > > Hi, > > The same thing is happening to me, also. Ubuntu 14.04, uname -a output: > Linux misel-pc 3.13.0-24-generic #47-Ubuntu SMP Fri May 2 23:30:00 UTC 2014 > x86_64 x86_64 x86_64 GNU/Linux No, it isn't. You want bug 54226. *** Bug 79296 has been marked as a duplicate of this bug. *** *** Bug 79297 has been marked as a duplicate of this bug. *** Created attachment 99940 [details] [review] [PATCH] drm/i915: Don't use RCS flips on gen7+ when multiple pipes are enabled This patch might fix the problem. It simply goes back to using BCS for all flips when multiple pipes are active. *** Bug 79390 has been marked as a duplicate of this bug. *** Ville, could you backport the patch to 3.14? Created attachment 100213 [details] [review] i915-dont-use-rcs-on-gen7-for-3.14.patch Here's backport of Ville's patch for 3.14 *** Bug 79626 has been marked as a duplicate of this bug. *** *** Bug 79705 has been marked as a duplicate of this bug. *** *** Bug 79890 has been marked as a duplicate of this bug. *** *** Bug 79954 has been marked as a duplicate of this bug. *** *** Bug 80080 has been marked as a duplicate of this bug. *** *** Bug 80746 has been marked as a duplicate of this bug. *** *** Bug 80754 has been marked as a duplicate of this bug. *** 3.15.2 is also affected. *** Bug 80802 has been marked as a duplicate of this bug. *** commit 2a92d5bca1999b69c78f3c3e97b5484985b094b9 Author: Chris Wilson <chris@chris-wilson.co.uk> Date: Tue Jul 8 10:40:29 2014 +0100 drm/i915: Disable RCS flips on Ivybridge We currently see random GPU hangs when using RCS flips with multiple pipes on Ivybridge. Now that we have mmio flips, we can fairly cheaply fallback to using CPU driven flips instead. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=77104 Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> *** Bug 81539 has been marked as a duplicate of this bug. *** *** Bug 82129 has been marked as a duplicate of this bug. *** *** Bug 82470 has been marked as a duplicate of this bug. *** *** Bug 82651 has been marked as a duplicate of this bug. *** *** Bug 82729 has been marked as a duplicate of this bug. *** *** Bug 82836 has been marked as a duplicate of this bug. *** *** Bug 82839 has been marked as a duplicate of this bug. *** *** Bug 83056 has been marked as a duplicate of this bug. *** *** Bug 83074 has been marked as a duplicate of this bug. *** *** Bug 83343 has been marked as a duplicate of this bug. *** *** Bug 83345 has been marked as a duplicate of this bug. *** *** Bug 82601 has been marked as a duplicate of this bug. *** *** Bug 83528 has been marked as a duplicate of this bug. *** *** Bug 82850 has been marked as a duplicate of this bug. *** *** Bug 83683 has been marked as a duplicate of this bug. *** *** Bug 83812 has been marked as a duplicate of this bug. *** *** Bug 84150 has been marked as a duplicate of this bug. *** *** Bug 84433 has been marked as a duplicate of this bug. *** *** Bug 84464 has been marked as a duplicate of this bug. *** *** Bug 84520 has been marked as a duplicate of this bug. *** Hi, Just got again the same bug report in dmesg, so for me the bug is NOT FIXED. This is on Ubuntu 14.04.1 LTS. Can someone here explain why this bug is marked FIXED, although there are still many reports showing it is not fixed? If this because of Ubuntu not being up-to-date, could you give some info about which package must be updated, to what version, etc? Maybe it's fixed in upstream dev branch? I have Archlinux and I get the same thing on a daily basis. (In reply to comment #50) > Maybe it's fixed in upstream dev branch? > > I have Archlinux and I get the same thing on a daily basis. Vote for this task: https://bugs.archlinux.org/task/41979 (In reply to comment #49) > Just got again the same bug report in dmesg, so for me the bug is NOT FIXED. > This is on Ubuntu 14.04.1 LTS. > > Can someone here explain why this bug is marked FIXED, although there are > still many reports showing it is not fixed? This is an upstream drm/i915 bug tracker, and the bug is fixed upstream. > If this because of Ubuntu not being up-to-date, could you give some info > about which package must be updated, to what version, etc? Please file bugs to your distros. Thanks. *** Bug 84716 has been marked as a duplicate of this bug. *** *** Bug 85069 has been marked as a duplicate of this bug. *** I filed a bug report for Ubuntu: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1384597 Upgrade kernel on 3.17 solve problem. THX. can confirm upgrade kernel 3.17 fixes this issue. (In reply to redwoz from comment #57) > can confirm upgrade kernel 3.17 fixes this issue. Not for me, though. I am running 3.17.1 + kde-5 and it complains when it starts OTOH this might never get fixed since this[1] guy don't want to work on intel bugs anymore :-( [1] http://mjg59.dreamwidth.org/32778.html (In reply to Milan Hodoscek from comment #58) > OTOH this might never get fixed since this[1] guy don't want to work on > intel bugs anymore :-( All the credit to Matthew for everything he's done for Linux, but what you're saying is not just a little bit unfair in the drm/i915 context. He's got 10 commits in i915, last one from 2011. Not a single comment on i915 bugs in fdo bugzilla. *** Bug 86308 has been marked as a duplicate of this bug. *** *** Bug 86892 has been marked as a duplicate of this bug. *** I'm experiencing 2 to 3 hangs per week on Linux 3.18.1. That's pretty terrible if you ask me. Why is this bug marked as closed? *** Bug 88244 has been marked as a duplicate of this bug. *** The(In reply to Sven from comment #62) > I'm experiencing 2 to 3 hangs per week on Linux 3.18.1. That's pretty > terrible if you ask me. Why is this bug marked as closed? Same here on Arch with kernel 3.18.2 Where is the workaround/patch/anything? It is fixed. If you still have a bug, it is not this. (In reply to Chris Wilson from comment #65) > It is fixed. If you still have a bug, it is not this. Indeed, the bug is fixed in rc2 of kernel 3.19! I found it here: https://wiki.archlinux.org/index.php/Intel_graphics#GPU_hangs_with_Haswell.27s_and_Ivy_Bridge.27s_HD_Graphics Now the screen only flickering black but not freeze after installed kernel from http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.19-rc4-vivid/ I hope it helps better others than "it is not this". (In reply to zamiere from comment #66) > I hope it helps better others than "it is not this". Because that is a different bug (bug 87759)... And had you opened a bug report and shown us the error state along with the usual log files I could have helped you. (In reply to Chris Wilson from comment #67) > (In reply to zamiere from comment #66) > > I hope it helps better others than "it is not this". > > Because that is a different bug (bug 87759)... And had you opened a bug > report and shown us the error state along with the usual log files I could > have helped you. No. It was 83677. Nevermind. Intel driver is currently unusable slow and unstable, it was a waste of time. *** Bug 89618 has been marked as a duplicate of this bug. *** *** Bug 91601 has been marked as a duplicate of this bug. *** *** Bug 92539 has been marked as a duplicate of this bug. *** |
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.
Created attachment 96978 [details] GPU Crash Dump GPU sporadically hangs during normal desktop usage, but recovers and a crash dump is created (see attached file). I'm using kernel 3.14.0 on Gentoo Linux with KDE Desktop. In my xorg.conf I have the option 'TearFree' enabled and disabled any tearing prevention mechanisms in the window manager (KWin).