When I switch to VT, I get a system freeze. The hardware is Intel Corporation 82852/855GM, more details at the URL. I'll attach a backtrace.
Created attachment 11597 [details] Backtrace of the segfault Backtrace + the value of xkbi->desc, which is NULL. I forgot to mention that I use the debian package of xorg server, of experimental, which is version 1.4,recompiled with -O0 Note that the only 100% way to reproduce this bug is to play a video (with xv) before switching to VT, but it sometimes occurs without playing a video before (Maybe that's a different bug then, as I have not taken stacktraces of that faults). This bug might be related to Bug 12437.
As the bug occurs in some keyboard-related code: No matter whether NUM is enabled or disabled under X, the LED on my keyboard is off. It is on as soon as I switch to VT (unless it crashes). This also applies to CAPS-LOCK. However, the internal state is preserved, as THESE THREE WORDS are written with CAPS-LOCK enabled.
(In reply to comment #2) > As the bug occurs in some keyboard-related code: No matter whether NUM is > enabled or disabled under X, the LED on my keyboard is off. It is on as soon as > I switch to VT (unless it crashes). This also applies to CAPS-LOCK. > > However, the internal state is preserved, as THESE THREE WORDS are written with > CAPS-LOCK enabled. > Willi: Can you please try with a recent X server. The xkb code has improved since and I can't reproduce your problem with the current version. If it still occurs, please attach your log file.
I can't reproduce it anymore, as all the xv related issues have been resolved for me. I have another crash bug (on context switch) which occurs very seldomly that might be related to this bug, but I haven't had yet time to really investigate it.
(In reply to comment #4) > I can't reproduce it anymore, as all the xv related issues have been resolved > for me. > > I have another crash bug (on context switch) which occurs very seldomly that > might be related to this bug, but I haven't had yet time to really investigate > it. > actually, since this bug is marked as a blocker for server 1.5 - are you able to test the server-1.5-branch as well? If you can't reproduce it, I guess we can mark it as closed.
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.