rawhide report: 20080802 changes (X server dies)

Michal Jaegermann michal at harddata.com
Sun Aug 3 01:21:38 UTC 2008


On Sat, Aug 02, 2008 at 04:19:21PM -0700, Tom London wrote:
> 0.211 generates a "black screen of death" for me with gdm boot
> (runlevel 5).  Checking the logs, there is a repeatable kernel Oops in
> the drm code that seems to savage X
> (https://bugzilla.redhat.com/show_bug.cgi?id=457598, kernel BUG at
> drivers/gpu/drm/drm_fops.c:144! invalid opcode: 0000).

Thanks for a bug number although I see
BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
instead of "invalid opcode", followed by further recursive oopses
and memory obviously messed up beyond recognition.  I attached
what I got to bz 457598.

> However, booting 0.211 to runlevel 3 and then issuing a 'startx' works
> for me (Thinkpad X60, Intel 945 graphics) without the Oops.

That really does not matter in my case.  In both cases the same
problems show up.

   Michal

p.s. A "new-and-improved" bugzilla seems to be quite messed up.
It forgets logins, location of attachements and who knows what
else.  A good question if it "remembered" to add CC to me even
if this was clearly selected.  So far nothing showed so possibly
this was forgotten too and I had to login four times in a row
to add attachments/comments.  Did anybody test that thingy before
deploying?




More information about the fedora-test-list mailing list