Bug 60534
| Summary: | i915 deadlock related to lock(&dev->struct_mutex); | ||
|---|---|---|---|
| Product: | Drivers | Reporter: | Alexander E. Patrakov (patrakov) |
| Component: | Video(DRI - Intel) | Assignee: | Daniel Vetter (daniel) |
| Status: | RESOLVED CODE_FIX | ||
| Severity: | normal | CC: | alan, daniel, intel-gfx-bugs, stumpish |
| Priority: | P3 | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Kernel Version: | 3.10-rc7 plus patches | Subsystem: | |
| Regression: | No | Bisected commit-id: | |
| Attachments: | Dmesg illustrating the problem | ||
Known issue with the current prime code, we need w/w mutexes (now mergd into latest git) and a completely reworking locking in drm/i915 (not even started yet). This will take a while to fix. Daniel, have we finally fixed this? No, still need to rewrite half the driver ... The nouveau/radeon side of this little project from Maarten is merged into 3.18 though. If the problem persists with latest kernels, please file a bug at the freedesktop.org bugzilla [1], referencing this bug. Thank you. [1] https://bugs.freedesktop.org/enter_bug.cgi?product=DRI&component=DRM/Intel |
Created attachment 106845 [details] Dmesg illustrating the problem I have a 3.10-rc7 kernel plus radeon-related patches from branch 'drm-next-3.11-wip-4' of git://people.freedesktop.org/~agd5f/linux My laptop has both an Intel graphics chip in itself and a Radeon chip in a hot-pluggable dock station. Both graphics cards have usable outputs. So I was playing with xrandr --setprovideroffloadsink and --setprovideroutputsource. By doing so, I was able to cause a deadlock in the i915 driver. I am going to attach a dmesg that illustrates it.