| Summary: | ThinkPad x201 laggy after resume, i915 errors on hangcheck timer | ||
|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Paul W. Frields <stickster> |
| Component: | kernel | Assignee: | Kernel Maintainer List <kernel-maint> |
| Status: | CLOSED ERRATA | QA Contact: | Fedora Extras Quality Assurance <extras-qa> |
| Severity: | unspecified | Docs Contact: | |
| Priority: | unspecified | ||
| Version: | 15 | CC: | airlied, el, gansalmon, itamar, jerome, jonathan, kernel-maint, madhu.chinakonda, martin, pfrields, rod, simon, sven, tsl |
| Target Milestone: | --- | Keywords: | Reopened |
| Target Release: | --- | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2012-06-04 18:57:26 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
|
Description
Paul W. Frields
2011-04-23 12:25:44 UTC
Hm, seems to be fixed with kernel 2.6.38.3, at least in: https://admin.fedoraproject.org/updates/kernel-2.6.38.3-18.fc15 Spoke too soon -- problem reasserted itself after a longer period of use. I confirm the same behaviour on a Lenovo T410s equipped with intel core i5 M520 integrated graphics (ironlake mobile). I've been using fc15 for weeks and didn't notice it until a few days, which I think correspond to the upgrade to 2.6.38.3-18.fc15.i686.PAE on my side. But as Paul mentions above, it's not easy to reproduce until some time of use, so I might have missed it. I've the same issue on a dell latitude e6500 with kernel 2.6.38.6-26.rc1.fc15.x86_64. Resumed from suspend with lots of unresponsiveness and the error message. Noticing the comment about powertop - I have also recently enabled runtime-pm options across the various pci bus devices as suggested by powertop which I had not done on previous kernel versions. I've tested this a little further and it doesn't appear that my power tuning settings make a difference. The problem asserts itself after an extended period of use. This is still a problem with kernel-2.6.38.6-27.fc15.x86_64. I experience this behavior and corresponding log entries as well, though irregularly. Acer 4820TG i3 CPU ironlake GPU kernel 2.6.38.7-30.fc15.x86_64. I have been using powertop suggested power tunings similar to other reports. Same problem, powertop fixes it magically. I have *no* powertop tunables in place that cover graphics, just all usb devices power managed, VM writeback increased, NMI watchdog off, audio codec power management on (snd-hda-intel), wake on lan off (both wlan and lan) and cpu scaler set to ondemand. All other tunables are "Bad"/untouched. Kernel is 2.6.38.7-30.fc15.i686 Graphics card is 945GME on an Asus Eee 100H, xorg-x11-drv-intel is 2.15.0-3.fc15 Particularly with GNOME 3's prominent suspend usage, this problem is annoying (when not knowing about the work-around). Some dmesg output: [68631.792072] [drm:i915_hangcheck_ring_idle] *ERROR* Hangcheck timer elapsed... render ring idle [waiting on 9876, at 9876], missed IRQ? [68633.496104] [drm:i915_hangcheck_ring_idle] *ERROR* Hangcheck timer elapsed... render ring idle [waiting on 9892, at 9894], missed IRQ? [68635.020083] [drm:i915_hangcheck_ring_idle] *ERROR* Hangcheck timer elapsed... render ring idle [waiting on 9898, at 9898], missed IRQ? [68636.536051] [drm:i915_hangcheck_ring_idle] *ERROR* Hangcheck timer elapsed... render ring idle [waiting on 9904, at 9904], missed IRQ? [68638.064085] [drm:i915_hangcheck_ring_idle] *ERROR* Hangcheck timer elapsed... render ring idle [waiting on 9913, at 9913], missed IRQ? Hitting the same bug on a Lenovo x220 -- kernel 2.6.38.8-32.fc15.x86_64 Linus hit it http://lists.freedesktop.org/archives/dri-devel/2011-January/006987.html did some bisection that seems to have clarified the situation some - http://lists.freedesktop.org/archives/dri-devel/2011-January/007030.html and led to a fix that seems to be working http://lists.freedesktop.org/archives/dri-devel/2011-January/007041.html Hopefully the fixup patch can be isolated and applied to 2.6.38 series on F15... Looks like we have a dupe at #706293 - reported against 2.6.39 kernels - comments there hint that i915.semaphores=1 reduces the frequency. From a comment: Appears to be fixed with 3.0-0.rc4.git0.2.fc16.x86_64, presumably by commit 498e720b96379d8ee9c294950a01534a73defcf3 "drm/i915: Fix gen6 (SNB) missed BLT ring interrupts." Same issue with f15 on a Toshiba NB505 laptop. (This is the top selling netbook on Amazon.com). 00:02.1 Display controller: Intel Corporation N10 Family Integrated Graphics Controller Subsystem: Toshiba America Info Systems Device fdc0 Flags: bus master, fast devsel, latency 0 Memory at f0280000 (32-bit, non-prefetchable) [size=512K] Capabilities: [d0] Power Management version 2 I'm seeing this on my Dell e6410 with Core i5 M540 CPU with IronLake GPU. Looks like 684097 might be the same bug. Upstream bug that looks similar https://bugs.freedesktop.org/show_bug.cgi?id=38529 This should have been fixed in the 3.x kernels. If you are still seeing this issue with the 3.3 F16 or newer kernel, please reopen. |