Bug 23622
| Summary: | null-pointer deref in ttm_bo_reserve() during using KDE 4.5 compositing session with radeon kms | ||
|---|---|---|---|
| Product: | Drivers | Reporter: | Martin Steigerwald (Martin) |
| Component: | Video(DRI - non Intel) | Assignee: | drivers_video-dri |
| Status: | RESOLVED UNREPRODUCIBLE | ||
| Severity: | normal | CC: | akpm, alexdeucher |
| Priority: | P1 | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Kernel Version: | 2.6.36-tp42-gtt-vram-apart-v2-dirty | Subsystem: | |
| Regression: | Yes | Bisected commit-id: | |
| Attachments: |
syslog with all the backtraces until I rebooted the machine
dmesg of 2.6.37-rc3 syslog with more backtraces |
||
|
Description
Martin Steigerwald
2010-11-23 11:06:30 UTC
Created attachment 37952 [details]
syslog with all the backtraces until I rebooted the machine
packed with bzip2, seems to be to big otherwise.
The title is inaccurate. This is actually a null-pointer deref in ttm_bo_reserve(). Looks like you ran out of graphics memory. What size is your AGP aperture set to? Please attach your dmesg output. Created attachment 37982 [details]
dmesg of 2.6.37-rc3
Andrew, thanks for hint regarding the title. I changed it as per your suggestion.
Alex, dmesg of freshly booted 2.6.37-rc3 (upto KDE Display Manager). Tell me if you need 2.6.36 as well.
Created attachment 37992 [details]
syslog with more backtraces
this syslog contains some additional backtraces. Including another occurence of this with 2.6.36 where kwin got stuck in a 95-100% CPU loop and the desktop was basically frozen, no mouse movement, no switching to tty. And some allocation features with previously booted 2.6.37-rc3 after playing around with desktop cube and effects which feel a bit faster than with 2.6.36.
I will test the two desktop gfx corruption case tomorrow during holding my training.
agpgart-intel 0000:00:00.0: AGP 2.0 bridge agpgart-intel 0000:00:00.0: putting AGP V2 device into 4x mode radeon 0000:01:00.0: putting AGP V2 device into 4x mode radeon 0000:01:00.0: GTT: 256M 0xD0000000 - 0xDFFFFFFF [drm] Generation 2 PCI interface, using max accessible memory radeon 0000:01:00.0: VRAM: 128M 0xE0000000 - 0xE7FFFFFF (64M used) [drm] radeon: irq initialized. [drm] Detected VRAM RAM=128M, BAR=128M [drm] RAM width 128bits DDR [TTM] Zone kernel: Available graphics memory: 437562 kiB. [TTM] Zone highmem: Available graphics memory: 1031134 kiB. [TTM] Initializing pool allocator. [drm] radeon: 64M of VRAM memory ready [drm] radeon: 256M of GTT memory ready. Is 256 MB the AGP aperture size? You can try increasing your AGP aperture size in the bios, but there's not a lot you can do if you run out of graphics memory other than trying to handle things more gracefully. I did not found any option for this in the ThinkPad BIOS. But shouldn't the driver just get more slow, i.e. increase swapping of textures between gfx card and main memory, when gfx memory is low? Well, maybe thats what you mean by "more gracefully". Anyway, will run without compositing when using two sessions simultaneously. Seems fair enough with that old gfx card ;). Will leave this report open as reminder for the "more gracefully" thing, i.e. to fix that null pointer deference for example ;). Thanks. Not my main notebook aynmore and did not have any visible issues anymore. Thus closing as unreproducible. |