Bug 182260

Created attachment 124952 [details]
fc5t3 xorg.conf

Created attachment 124953 [details]
Log of failing X server

Created attachment 124954 [details]
working xorg.conf, for reference

Created attachment 124955 [details]
Log of working X server

I don't think this is related exactly, but on fc5 test3 , on x86 hardware, X
segfaults on me every now and again.  Here's what the log says:

Backtrace:
0: /usr/bin/Xorg(xf86SigHandler+0x87) [0x80b85e7]
1: [0x7f8420]
2: /usr/bin/Xorg(CopyGC+0x1d4) [0x80982e4]
3: /usr/bin/Xorg(ProcCopyGC+0x105) [0x8086d85]
4: /usr/bin/Xorg(Dispatch+0x19b) [0x80888eb]
5: /usr/bin/Xorg(main+0x487) [0x80701d7]
6: /lib/i686/nosegneg/libc.so.6(__libc_start_main+0xdc) [0x68b7a4]
7: /usr/bin/Xorg(FontFileCompleteXLFD+0xb9) [0x806f511]

Fatal server error:
Caught signal 11.  Server aborting

Oops, guess I should mention, xorg-x11-server-Xorg-1.0.1-7 on my end.

On my end, this only seems to happen under KDE for some reason.

Looks like I spoke too soon.  Just crashed on me when Gnome was running:

Backtrace:
0: /usr/bin/Xorg(xf86SigHandler+0x87) [0x80b85e7]
1: [0x4a9420]
2: /usr/lib/xorg/modules/libfb.so(fbCopyAreammx+0x192) [0x20b3f2]
3: /usr/lib/xorg/modules/libfb.so(fbCopyNtoN+0x20c) [0x1f80ac]
4: /usr/lib/xorg/modules/libfb.so(fbCopyRegion+0xbb) [0x1f6f9b]
5: /usr/lib/xorg/modules/libfb.so(fbDoCopy+0x46f) [0x1f75ff]
6: /usr/lib/xorg/modules/libfb.so(fbCopyArea+0x78) [0x1f77c8]
7: /usr/lib/xorg/modules/libxaa.so [0xca84bd]
8: /usr/lib/xorg/modules/libxaa.so(XAACopyArea+0xe7) [0xca9737]
9: /usr/bin/Xorg [0x8152bb9]
10: /usr/bin/Xorg [0x812ddc4]
11: /usr/bin/Xorg [0x812ec32]
12: /usr/bin/Xorg [0x812f9c6]
13: /usr/bin/Xorg(Dispatch+0x19b) [0x80888eb]
14: /usr/bin/Xorg(main+0x487) [0x80701d7]
15: /lib/i686/nosegneg/libc.so.6(__libc_start_main+0xdc) [0x68b7a4]
16: /usr/bin/Xorg(FontFileCompleteXLFD+0xb9) [0x806f511]

Fatal server error:
Caught signal 11.  Server aborting


Course now I'm kinda worried about my system.  I'm gonna to run memtest and see
if that still comes up clean.

well memtest ran a few hours with no errors.  And I did have mandriva on this
thing previously without much problems, so now I'm feeling pretty confident it's
not a hardware problem.

Glen:  None of the backtraces you've attached are related to this bug report.
The only thing in common is that you're experiencing a SEGV.  Please report
those unrelated SEGV's to X.Org at http://bugs.freedesktop.org bugzilla in
the "xorg" component.


This bug is about an X server startup failure on PowerPC hardware.

(In reply to comment #0)
> On my dual G5 the X server segfaults during install, and then after falling back
> to a text installation it also segfaults in the real system. It worked for
> firstboot, but after selecting the 1920x1200 mode I desired for my flat panel,
> it started to die.
> 
> Program received signal SIGSEGV, Segmentation fault.
> xf86InitViewport (pScr=0x10252de8) at xf86Cursor.c:121
> 121       pScr->frameX1 = pScr->frameX0 + pScr->modes->HDisplay - 1;
> (gdb) p pScr->modes
> $1 = 0x0
> (gdb) bt
> #0  xf86InitViewport (pScr=0x10252de8) at xf86Cursor.c:121
> #1  0x10067910 in InitOutput (pScreenInfo=0x10224614, argc=6, argv=0xffabaa24)
>     at xf86Init.c:693
> #2  0x24000482 in ?? ()
> #3  0x1002702c in main (argc=6, argv=0xffabaa24, envp=Variable "envp" is not
> available.
> ) at main.c:372


David, does this still occur with FC5 current, and is this backtrace
consistent?

Yeah, I believe so. I think it's the result of two bugs. First we end up with an
empty mode list because something went wrong. Then we SEGV because we assume the
mode list isn't empty. I thought both of those got fixed before Xorg 6.9 though.

Is this still a problem in rawhide?

Closing per lack of response to previous request for information.
This bug was originally filed against a much earlier version of Fedora
Core, and significant changes have taken place since the last version
for which this bug is confirmed.

Note that FC3 and FC4 are supported by Fedora Legacy for security
fixes only.  Please install a still supported version and retest.  If
it still occurs on FC5 or FC6, please reopen and assign to the correct
version.  Otherwise, if this a security issue, please change the
product to Fedora Legacy.  Thanks, and we are sorry that we did not
get to this bug earlier.