Summary: | after moving from dual head to single head during suspend/resume, X just shows mouse cursor | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | DRI | Reporter: | Frederik Himpe <fhimpe> | ||||||
Component: | DRM/Intel | Assignee: | Chris Wilson <chris> | ||||||
Status: | CLOSED FIXED | QA Contact: | |||||||
Severity: | normal | ||||||||
Priority: | medium | CC: | eugeni, jbarnes | ||||||
Version: | XOrg 6.7.0 | ||||||||
Hardware: | Other | ||||||||
OS: | All | ||||||||
Whiteboard: | |||||||||
i915 platform: | i915 features: | ||||||||
Attachments: |
|
Description
Frederik Himpe
2011-04-08 09:00:20 UTC
Created attachment 45425 [details]
intel_gpu_dump
output of intel_gpu_dump while this error happens
Created attachment 45426 [details]
vbios.dump
I should also have added that I'm using a Dell Latitude E6400 system with this graphics chipset:
00:02.1 Display controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07)
Subsystem: Dell Device 0233
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
Latency: 0
Region 0: Memory at f6b00000 (64-bit, non-prefetchable) [size=1M]
Capabilities: <access denied>
Frederik mentioned on irc that he thought it might be correlated with going from two output to a single output across suspend/resume. He included an trace-cmd record -e i915 that showed that the ddx was still rendering, so the question remains: where are we rendering to? [Hint for future self: I need to compare the target of the execbuffers with the current object attached to the scanout.] I have done some testing, and I'm now pretty sure that the move from dual head to single head across suspend/resume is indeed the trigger: every time I move from dual head to single head across suspend/resume on my system, this problem happens. This bug still happens with 2.6.39-rc6, however a quick test seems to indicate that this bug does not happen with drm-intel-staging branch. Bug still present in 3.0 How long until someone finally takes a look at this bug? Believed to be a GPU hang on WAIT_FOR_EVENT, fixed with kernel 3.4/3.5 and xf86-video-intel 2.19.0 |
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.