Bug 6591 - xserver crash with AMD64 (with vesa, fbdev, or radeon), at miFillPolyHelper.
Summary: xserver crash with AMD64 (with vesa, fbdev, or radeon), at miFillPolyHelper.
Status: RESOLVED WORKSFORME
Alias: None
Product: xorg
Classification: Unclassified
Component: Server/General (show other bugs)
Version: 6.9.0
Hardware: x86 (IA32) Linux (All)
: highest critical
Assignee: Maciej Cencora
QA Contact: Xorg Project Team
URL:
Whiteboard:
Keywords: have-backtrace
Depends on:
Blocks:
 
Reported: 2006-04-14 01:09 UTC by Maciej Cencora
Modified: 2011-10-15 11:43 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
xorg.conf (13.50 KB, text/plain)
2006-04-16 18:04 UTC, Maciej Cencora
no flags Details
Xorg.log - crash log (50.28 KB, text/plain)
2006-04-16 18:21 UTC, Maciej Cencora
no flags Details
lspci -v (6.92 KB, text/plain)
2006-05-23 23:22 UTC, Maciej Cencora
no flags Details
full backtrace (5.02 KB, text/plain)
2006-07-11 07:45 UTC, Maciej Cencora
no flags Details

Description Maciej Cencora 2006-04-14 01:09:47 UTC
Slamd64 with linux 2.6.16.1 on AMD64
X server crashes during normal work randomly (sometimes without backtrace in
logs). Sometimes I can work for couple of hours without a crash, but sometimes X
crashes very often (more than 2 times per 5 minutes). I realized that it crashes
more often when I want to select group of icons with mouse (creating a
rectagonal selection - or whatever it is called :P)

I don't think it is related to hardware or some drivers cause I got the same
with linux-2.4.32 kernel, and other drivers(vesa, fbdev, currently radeon).

Backtrace:
0: /usr/X11R6/bin/X(xf86SigHandler+0x92) [0x448902]
1: /lib64/tls/libc.so.6 [0x2b7a29982340]
2: /usr/X11R6/bin/X(miFillPolyHelper+0x2c4) [0x5439e4]
3: /usr/X11R6/bin/X(miWideDash+0x77e) [0x5477be]
4: /usr/X11R6/bin/X(miPolyRectangle+0x1b3) [0x542503]
5: /usr/X11R6/bin/X [0x525f22]
6: /usr/X11R6/bin/X(ProcPolyRectangle+0x18f) [0x483b8f]
7: /usr/X11R6/bin/X(Dispatch+0x196) [0x487b26]
8: /usr/X11R6/bin/X(main+0x409) [0x494619]
9: /lib64/tls/libc.so.6(__libc_start_main+0xdb) [0x2b7a2996f4cb]
10: /usr/X11R6/bin/X [0x430b9a]

Sorry for my english and any mistakes, that's my first bugreport :P
Comment 1 Erik Andren 2006-04-16 16:48:05 UTC
Please post your xorg.conf and your xorg.log after the crash.
Comment 2 Maciej Cencora 2006-04-16 18:04:08 UTC
Created attachment 5330 [details]
xorg.conf
Comment 3 Maciej Cencora 2006-04-16 18:21:16 UTC
Created attachment 5331 [details]
Xorg.log - crash log
Comment 4 Maciej Cencora 2006-04-16 18:28:03 UTC
Also I realized that when I am in 4th runlevel (in other distros it is 5), X
server crashes and restarts (because of kdm). But when I run X server in 3rd
runlevel it hangs whole computer(I must reset comp manually or with Magic SysRq)
without backtrace in Xorg.log (there is only first line: "Backtrace:" and
nothing more).
Comment 5 Erik Andren 2006-04-16 18:59:43 UTC
There's a trick around it: Log in remotely via SSH and attach gdb to the X
process to get the backtrace.
Comment 6 Michel Dänzer 2006-04-17 00:44:43 UTC
Does it also happen if you don't override the amount of video RAM using the
VideoRam directive? The drivers usually know exactly how much they can use.
Comment 7 Maciej Cencora 2006-04-17 00:54:07 UTC
Yes, it also crashes if I don't override VideoRAM.
Comment 8 Maciej Cencora 2006-05-23 23:20:38 UTC
I tested it on other graphics(S3 Virge PCI) and got the same.
Maybe some hardware information will be useful:
MSI K8N Neo3-F (with the newest BIOS)
AMD Sempron 64bit(s.754)
2x512MB RAM OEM
ABIT Radeon X600Pro HDTV 256MB
SB Live!
WiFi Dlink DWL-510
Maxtor 60GB PATA
DVD-RW LG GSA-4167

I don't know if it's related but sometimes when doing this selection, some parts
of that dotted rectagon stays on desktop, and if I want to get rid of it I have
to refresh desktop.
Comment 9 Maciej Cencora 2006-05-23 23:22:39 UTC
Created attachment 5720 [details]
lspci -v
Comment 10 Maciej Cencora 2006-07-11 07:45:03 UTC
Created attachment 6189 [details]
full backtrace

Sorry for the delay, but I don't have two PC's to get backtrace using ssh (but
I managed using multi-seat configuration. It doesn't work - at all - on R6.9
version, but it works - with some hangs, crashes etc - on R7.1)
I compiled R7.1 from sources, and found that it crashes too. So I compiled it
again with -g -O0 flags, applied multiseat configuration and got full
backtrace.

I don't know what do miWideDash/miWideDashSegment/miFillPolyHelper functions
do, so please send me link to a doc or small description and maybe I'll be able
to find the bug and write a patch.

I've forgotten to write that it must be 64-bit arch related bug because I have
no crashes when using slackware.
Comment 11 Daniel Stone 2007-02-27 01:31:29 UTC
Sorry about the phenomenal bug spam, guys.  Adding xorg-team@ to the QA contact so bugs don't get lost in future.
Comment 12 Jeremy Huddleston Sequoia 2011-10-15 11:43:37 UTC
Sorry for the slow feedback here.  This bug is really old, so I'm going to 
close it.  If you are still experiencing a problem, please reopen with a newer 
backtrace, and I will take a look.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.