2025-07-17 12:03 CEST
  

View Issue Details [ Jump to Notes ] [ Issue History ] [ Print ]
IDProjectCategoryView StatusDate SubmittedLast Update
0002277Spring engineGfx driver issuespublic2011-01-01 22:072011-05-03 21:28
Reporterjandd 
Assigned Toabma 
PrioritynormalSeveritycrashReproducibilityhave not tried
StatusresolvedResolutionunable to reproduce 
Product Version0.82.6.1 
Target VersionFixed in Version 
Summary0002277: spring exits with error "drmRadeonCmdBuffer: -22"
DescriptionA user of my Debian package of spring reported a crash on his machine in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=608594 I copy the bug report below.
Additional InformationWhen trying to start a game with "Test the game" some status messages
appear (loading this, initialising that) and then spring exits with an
error message:
drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command
stream. See dmesg for more info.

dmesg says:
[ 234.282985] [drm:r100_cs_track_check] *ERROR* [drm] No buffer for z
buffer !
[ 234.282992] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !
[ 419.540963] [drm:r100_cs_track_check] *ERROR* [drm] No buffer for z
buffer !
[ 419.540970] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !
[ 467.032293] [drm:r100_cs_track_check] *ERROR* [drm] No buffer for z
buffer !
[ 467.032300] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !

Googling doesn't provide further insights for me, but I suspect some
issue with the video driver and OpenGL. Please reassign if appropriate.

Regards,
Felix
TagsNo tags attached.
Checked infolog.txt for Errors
Attached Files
  • txt file icon infolog.txt (1,538 bytes) 2011-01-04 21:33 - 
    LogOutput initialized.
Spring 0.82.6.1
Build date/time: Nov  7 2010 09:11:35
Available log subsystems: CollisionVolume, unit, ArchiveScanner, VFS-detail, VFS, Sound
Enabled log subsystems: Sound
Enable or disable log subsystems using the LogSubsystems configuration key
  or the SPRING_LOG_SUBSYSTEMS environment variable (both comma separated).
using default configuration source "/home/flixh/.springrc"
[CMyMath::Init] CPU SSE mask: 112, flags:
	SSE 1.0:  1,  SSE 2.0:  1
	SSE 3.0:  0, SSSE 3.0:  0
	SSE 4.1:  0,  SSE 4.2:  0
	SSE 4.0A: 0,  SSE 5.0A: 0
	using streflop SSE FP-math mode, CPU supports SSE instructions
OS: Linux
OS: 32bit native mode
Using read-write data directory: /home/flixh/.spring/
Using read-only data directory: /usr/share/games/spring/
Using read-only data directory: /usr/lib/spring/
Scanning: /usr/share/games/spring/base
Scanning: /home/flixh/.spring/maps
Scanning: /home/flixh/.spring/mods
Supported Video modes: 1600x1024, 1440x900, 1400x1050, 1360x768, 1280x1024, 1280x960, 1280x854, 1280x800, 1280x720, 1152x864, 1152x768, 1024x768, 848x480, 800x600, 800x512, 720x480, 720x450, 680x384, 640x512, 640x480, 576x432, 512x384, 400x300, 320x240
Video mode set to 1400x1050/32bit
[      0] SDL:  1.2.14
[      0] GL:   1.5 Mesa 7.7.1
[      0] GL:   DRI R300 Project
[      0] GL:   Mesa DRI R300 (RV380 3150) 20090101 x86/MMX/SSE2 TCL DRI2
[      0] GLSL: (null)
[      0] GLEW: 1.5.4
[      0] Joysticks found: 0
[      0] Joystick 0 not found
[      0] Video mode set to 1400x982/32bit
[      0] User exited
    txt file icon infolog.txt (1,538 bytes) 2011-01-04 21:33 +

-Relationships
+Relationships

-Notes

~0006184

abma (administrator)

2011-01-02 17:45

 this heavily looks like a driver issue... do other 3d-applications work?

my suggestion is to install fglrx...

to get more help, please attach infolog.txt from such a crash, but i think, this crash isn't spring's fault.

~0006192

flixh (reporter)

2011-01-04 21:35

 infolog.txt attached...
Other 3d-applications work, e.g. glxgears or tuxracer run.
fglrx is unfortunately not an option as this breaks a
couple of other things.

~0006195

hoijui (reporter)

2011-01-05 10:08

 as this is an open source driver, we actually have a chance to find out whats going on!!!! :D ;-)

look what i found:
http://www.cs.fsu.edu/~baker/devices/lxr/http/source/linux/drivers/gpu/drm/radeon/r300.c#L848
(source code line of the error message, i think)

if we now knew where in spring code this happens, we might have a realistic chance to fix it, whether it is a bug there or here!

if you did not compile spring yourself, but used the debian package, you would have to get the debug symbols package for spring, run spring in gdb, and issue a "bt" when it crashes, and then please upload the output.

~0006198

flixh (reporter)

2011-01-05 21:06

 I tried to get a backtrace but have not been successful. Loading the debugging symbols works but gdb tells me that there is no stack. Is that because it does not crash but exists gracefully?
Otherwise I would need some more guidance to set a proper breakpoint or how to get a backtrace.

Thanks for investigating,
Felix

~0006595

flixh (reporter)

2011-05-03 21:23

 I cannot reproduce with debian's version 0.82.7.1+dfsg1-2 anymore. Unfortunately I cannot confirm if this is because of the new spring version or some kernel/driver update.

So this bug should probably be closed.
Felix

~0006597

abma (administrator)

2011-05-03 21:28

 thanks for reporting back!

i think this is because of the driver...
+Notes

-Issue History
Date Modified Username Field Change
2011-01-01 22:07 jandd New Issue
2011-01-02 17:45 abma Note Added: 0006184
2011-01-02 17:45 abma Status new => feedback
2011-01-02 20:29 abma Severity minor => crash
2011-01-04 21:33 flixh File Added: infolog.txt
2011-01-04 21:35 flixh Note Added: 0006192
2011-01-05 10:08 hoijui Note Added: 0006195
2011-01-05 21:06 flixh Note Added: 0006198
2011-05-03 21:23 flixh Note Added: 0006595
2011-05-03 21:28 abma Note Added: 0006597
2011-05-03 21:28 abma Status feedback => resolved
2011-05-03 21:28 abma Resolution open => unable to reproduce
2011-05-03 21:28 abma Assigned To => abma
+Issue History