Description of problem: When selecting Other from Fast User Switcher then the X on tty7 failed to start: Xorg.1.log: adding fb map from cef1b000 for 6ec000 ret 0 cef1b000 front handle is 6 failed to get old fb, id 20 messages: Nov 11 11:38:36 dev-mk kernel: allocation failed: out of vmalloc space - use vmalloc=<size> to increase size. Nov 11 11:38:36 dev-mk kernel: [drm:drm_bo_evict] *ERROR* Buffer eviction failed Nov 11 11:38:36 dev-mk kernel: [drm:drm_buffer_object_validate] *ERROR* Failed moving buffer. Nov 11 11:38:36 dev-mk kernel: [drm:drm_buffer_object_validate] *ERROR* Out of aperture space or DRM memory quota. Nov 11 11:38:36 dev-mk kernel: mtrr: base(0xcef1b000) is not aligned on a size(0xb13000) boundary Nov 11 11:38:36 dev-mk kernel: [drm:drm_mode_getfb] *ERROR* invalid framebuffer id This might be related to my closed reports: Bug 468389 - Loading R500 Microcode nomodeset => BUG: unable to handle kernel NULL pointer Bug 468975 - System hanging on (II) RADEON(0): [RESUME] Attempting to re-init Radeon hardware. Version-Release number of selected component (if applicable): kernel-2.6.27.5-92.fc10.i686 modeset xorg-x11-drv-ati-6.9.0-44.fc10.i386 no xorg.conf
Created attachment 323167 [details] /var/log/Xorg.1.log Forgot to say: No compiz
Created attachment 323168 [details] /var/log/dmesg
airlied: oh I need to test fast-user-switch some more... airlied: yes starting the second X server might cause problems airlied: switching back to the first server is possibly broken. airlied: yup.. so I guess I'll play with switching once I get the current bits done. airlied: pretty much from what I can see. airlied: I'll play around today/tomorrow and see what I can fix.
This bug appears to have been reported against 'rawhide' during the Fedora 10 development cycle. Changing version to '10'. More information and reason for this action is here: http://fedoraproject.org/wiki/BugZappers/HouseKeeping
Is this still happening with -61 and -137 kernel?
ATI & KMS has improved a lot, thanks! But because of bug 466598 I can't use KMS on a daily basis and thus can't give it heavy testing right now. But I will leave it assigned and come back to it soon.
Some testing shows that it seems to work OK. But I couldn't reproduce it on demand before, so my testing doesn't prove much. If you think it is solved then please close it - I will reopen if I see it again. BUT then I switched to a user who had compiz. At first it worked fine, but when I switched back to non-compiz I got horrible corruption. Big tiles had been moved around. I don't know if this is related ...
That is long time ago, things have changed, and this report no longer has any value. Closing.