Once this bug occurs, it is readily reproducible (or at least I think this is a reproduction case).
On my system, once X server wedges I can login to the machine via ssh and "kill -9" the X server process. A new process starts, and immediately throws the same stack traceback.
So there's the potential of intervening with /usr/bin/X, or /usr/lib/gdm/gdm-simple-slave (it's parent) to collect additional strace or debug information.
Once this bug occurs, it is readily reproducible (or at least I think this is a reproduction case).
On my system, once X server wedges I can login to the machine via ssh and "kill -9" the X server process. A new process starts, and immediately throws the same stack traceback.
So there's the potential of intervening with /usr/bin/X, or /usr/lib/ gdm/gdm- simple- slave (it's parent) to collect additional strace or debug information.