re: bin/53891: sandy bridge i965 SNA bus error

[matthew green <mrg%eterna.com.au@localhost>]

>  As mrg enabled backtrace -> log, here is another example - 0x12 is an
>  interesting address...

pity that libbacktrace's unwinder can't cope beyond signals:

>  [   240.916] (**) Option "XkbRules" "base"
>  [   240.916] (**) Option "XkbModel" "pc105"
>  [   240.916] (**) Option "XkbLayout" "us"
>  [   240.916] (II) XINPUT: Adding extended input device "<default keyboard>" (type: KEYBOARD, id 7)
>  [   245.830] (EE) 
>  [   245.830] (EE) Backtrace:
>  [   245.836] (EE) 0: /usr/X11R7/bin/X (xorg_backtrace+0x3d) [0x3b37d3]
>  [   245.836] (EE) 1: /usr/X11R7/bin/X (os_move_fd+0x67) [0x3af9f9]
>  [   245.836] (EE) 2: /usr/lib/libc.so.12 (__sigtramp_siginfo_2+0x0) [0x7f7ff32a9ac0]
>  [   245.836] (EE) 
>  [   245.836] (EE) Segmentation fault at address 0x12
>  [   245.836] (EE) 
>  Fatal server error:
>  [   245.836] (EE) Caught signal 11 (Segmentation fault). Server aborting

so all really 0x12 means is that someone tried to access a struct member
that is at offset 0x12 but using a NULL pointer.  however, as we don't
have the stack trace beyond the signal, don't know what it was.  ie, my
enabling of backtrace doesn't help the main problem i hoped for.