Description of problem:
On a single-node x455, X will not start up post-install with the
default configuration written by anaconda. It appears to crash when
loading the xaa module:
(II) Module xaa: vendor="The XFree86 Project"
compiled for 4.1.0, module version 1.0.0
Fatal server error:
Caught signal 11. Server aborting
Disabling acceleration (with Option "noaccel") in
/etc/X11/XF86Config-4 allows X to start up and run without issues.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. run 'startx' with default config provided by anaconda
X aborts with signal 11
X starts up successfully
X runs without a problem during installation (I imagine it runs with
"noaccel" at that point).
Created attachment 96791 [details]
Created attachment 96792 [details]
Is this a regression of a previously working and supported system,
now not working? Or is this a system that is new, and has never
worked? Any details you could provide would be greatly appreciated.
Isn't this a duplicate of bug 112175? If it is I have been looking at
this. If it is a duplicate of bug 112175 then its not a regression. In
bug 112175 IBM has changed the system BIOS and reduced the IO port
space. I got an update on this yesterday from IBM but I have not had a
chance yet to digest the information and return to working on it. I'll
take ownership of both these, I hadn't noticed you were still assigned
I'm not sure that this is a duplicate of 112175, but I guess it's
possible that it is caused by the same IO port space issue. 112175
occurs with a 2-node system and results in a machine check, whereas
this occurs with a single node system and results in a seg fault
(signal 11). I'll try flashing back to BIOS 14 and see if that changes
anything; at least then we can say better whether these two problems
may be related.
The same thing happens with BIOS 14, so I don't think this is related
to the IO port problem.
Created attachment 96817 [details]
strace -i -v -o strace.log X
strace on X... I thought it might help if you can make more sense out of it
than I can.
I don't know what kind of weird state the machine was in when I found
this problem... I can no longer reproduce it, so I'll close this bug.