[needs linux-2.6.36.2] [Radeon kernel module] drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command stream

Note: if you are getting "drmRadeonCmdBuffer: -12" this is another problem see bug #533784.

I am getting this:
drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command stream. See dmesg for more info.
when trying to start the game 0ad or douze map with sauerbraten (with shaders enabled).

dmesg shows this after 0ad crash:
[ 204.087282] [drm:r100_cs_track_check] *ERROR* [drm] Buffer too small for color buffer 0 (need 4190208 have 3145728) !
[ 204.087294] [drm:r100_cs_track_check] *ERROR* [drm] color buffer 0 (1024 4 0 1023)
[ 204.087300] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !

and this with sauerbraten:
[ 1144.213905] radeon 0000:01:00.0: (PW 1) Vertex array 0 need 83826965 dwords have 112640 dwords
[ 1144.213915] [drm:r100_cs_track_check] *ERROR* Max indices 16765393
[ 1144.213921] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !

Card:
GL_RENDERER = Mesa DRI R300 (RV530 71C5) 20090101 x86/MMX/SSE2 TCL DRI2

Running the eternallands game I get this to standard out:
Allocating 1014 x 713 radeon RBO (pitch 1024)
Allocating 1014 x 713 radeon RBO (pitch 1024)
Allocating 1014 x 713 radeon RBO (pitch 1024)
drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command stream. See dmesg for more info.

And this in dmesg:
[ 90.853952] [drm:r100_cs_track_texture_check] *ERROR* No texture bound to unit 0
[ 90.853967] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !

This is for a Thinkpad X31 using "ATI Technologies Inc Radeon Mobility M6 LY"

Hi Fabio,

Please be sure to confirm this issue exists with the latest development release of Ubuntu. ISO CD images are available from http://cdimage.ubuntu.com/releases/ . If the issue remains, please run the following command from a Terminal (Applications->Accessories->Terminal). It will automatically gather and attach updated debug information to this report.

apport-collect -p linux 557266

Also, if you could test the latest upstream kernel available that would be great. It will allow additional upstream developers to examine the issue. Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . Once you've tested the upstream kernel, please remove the 'needs-upstream-testing' tag. This can be done by clicking on the yellow pencil icon next to the tag located at the bottom of the bug description and deleting the 'needs-upstream-testing' text. Please let us know your results.

Thanks in advance.

    [This is an automated message. Apologies if it has reached you inappropriately; please just reply to this message indicating so.]

AlsaVersion: Advanced Linux Sound Architecture Driver Version 1.0.21.
Architecture: i386
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: paul 1349 F.... pulseaudio
CRDA: Error: [Errno 2] No such file or directory
Card0.Amixer.info:
 Card hw:0 'I82801DBICH4'/'Intel 82801DB-ICH4 with AD1981B at irq 5'
   Mixer name : 'Analog Devices AD1981B'
   Components : 'AC97a:41445374'
   Controls : 26
   Simple ctrls : 18
Card29.Amixer.info:
 Card hw:29 'ThinkPadEC'/'ThinkPad Console Audio Control at EC reg 0x30, fw 1QHT22WW-1.07b'
   Mixer name : 'ThinkPad EC 1QHT22WW-1.07b'
   Components : ''
   Controls : 2
   Simple ctrls : 1
Card29.Amixer.values:
 Simple mixer control 'Console',0
   Capabilities: pvolume pvolume-joined pswitch pswitch-joined penum
   Playback channels: Mono
   Limits: Playback 0 - 14
   Mono: Playback 6 [43%] [on]
DistroRelease: Ubuntu 10.04
HibernationDevice: RESUME=UUID=bae2b247-206b-4a57-a7b3-b5d68c7e963e
InstallationMedia: Ubuntu 10.04 "Lucid Lynx" - Beta i386 (20100318)
Lsusb:
 Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
 Bus 003 Device 002: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse
 Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
 Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
 Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
MachineType: IBM 26722G3
Package: linux (not installed)
PccardctlIdent:
 Socket 0:
   no product info available
 Socket 1:
   no product info available
PccardctlStatus:
 Socket 0:
   no card
 Socket 1:
   no card
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.32-19-generic root=UUID=6f8536c6-ed76-44d1-8321-d86639de2c1c ro quiet splash
ProcEnviron:
 LANG=en_GB.utf8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.32-19.28-generic 2.6.32.10+drm33.1
Regression: Yes
RelatedPackageVersions: linux-firmware 1.34
Reproducible: Yes
RfKill:

Tags: lucid graphics regression-release needs-upstream-testing
Uname: Linux 2.6.32-19-generic i686
UserGroups: adm admin cdrom dialout lpadmin plugdev sambashare
dmi.bios.date: 01/23/2005
dmi.bios.vendor: IBM
dmi.bios.version: 1QET94WW (3.00d)
dmi.board.name: 26722G3
dmi.board.vendor: IBM
dmi.board.version: Not Available
dmi.chassis.asset.tag: No Asset Information
dmi.chassis.type: 10
dmi.chassis.vendor: IBM
dmi.chassis.version: Not Available
dmi.modalias: dmi:bvnIBM:bvr1QET94WW(3.00d):bd01/23/2005:svnIBM:pn26722G3:pvrThinkPadX31:rvnIBM:rn26722G3:rvrNotAvailable:cvnIBM:ct10:cvrNotAvailable:
dmi.product.name: 26722G3
dmi.product.version: ThinkPad X31
dmi.sys.vendor: IBM

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

The bug is still reproducible with current r300 driver from mesa git and 2.6.34-rc6 from kernel mainline build. I suspect, however, that this is a mesa bug and not a kernel bug. See also this comment from https://bugs.freedesktop.org/show_bug.cgi?id=27507#c1 :
"0ad problem is that mesa sets wrong height for rendering area (1023 instead 768)."

I'm seeing this as well. I have an ATI FireGL 8800 (r200), and the error I see in dmesg is

[drm:r100_cs_track_texture_check] *ERROR* No texture bound to unit 1
[drm:radeon_cs_ioctl] *ERROR* Invalid command stream !

glxgears works, but openarena fails, gltron fails....

I see the same message for my ATI Technologies Inc Radeon R200 QL [Radeon 8500 LE] on Ubuntu 10.04 while playing World of Padman. The game exits with the message:

drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command stream. See dmesg for more info.

Also dmesg shows this:

[135106.802528] ------------[ cut here ]------------
[135106.802571] WARNING: at /build/buildd/linux-2.6.32/arch/x86/kernel/apic/ipi.c:109 default_send_IPI_mask_logical+0xb2/0xf0()
[135106.802586] Hardware name: -[622140U]-
[135106.802591] empty IPI mask
[135106.802600] Modules linked in: autofs4 binfmt_misc nfsd exportfs nfs lockd nfs_acl auth_rpcgss cx88_blackbird cx2341x tda827x tda8290 fbcon tileblit font bitblit softcursor snd_intel8x0 vga16fb vgastate tuner snd_ac97_codec cx88_alsa ac97_bus snd_pcm_oss snd_mixer_oss snd_pcm cx8802 cx8800 snd_seq_dummy arc4 cx88xx radeon pc87360 v4l2_common rt2500pci ttm snd_seq_oss rt2x00pci videodev ir_common snd_seq_midi rt2x00lib snd_rawmidi drm_kms_helper snd_seq_midi_event led_class tveeprom v4l1_compat snd_seq videobuf_dma_sg mac80211 drm btcx_risc snd_timer i2c_algo_bit snd_seq_device cfg80211 snd e7xxx_edac eeprom_93cx6 w83791d hwmon_vid videobuf_core soundcore snd_page_alloc edac_core i2c_i801 sunrpc ppdev intel_agp psmouse shpchp serio_raw agpgart parport_pc lp parport usbhid hid usb_storage floppy tg3
[135106.802866] Pid: 17, comm: events/2 Not tainted 2.6.32-22-generic #33-Ubuntu
[135106.802871] Call Trace:
[135106.802895] [<c014c3d2>] warn_slowpath_common+0x72/0xa0
[135106.802905] [<c0120952>] ? default_send_IPI_mask_logical+0xb2/0xf0
[135106.802913] [<c0120952>] ? default_send_IPI_mask_logical+0xb2/0xf0
[135106.802925] [<c014c44b>] warn_slowpath_fmt+0x2b/0x30
[135106.802934] [<c0120952>] default_send_IPI_mask_logical+0xb2/0xf0
[135106.802956] [<c011edb1>] native_send_call_func_ipi+0x41/0x60
[135106.802969] [<c017c298>] smp_call_function_many+0x148/0x1c0
[135106.802983] [<c012c730>] ? __change_page_attr_set_clr+0x60/0x1c0
[135106.802999] [<c012bf90>] ? __cpa_flush_range+0x0/0x30
[135106.803007] [<c012bf90>] ? __cpa_flush_range+0x0/0x30
[135106.803024] [<c017c334>] smp_call_function+0x24/0x30
[135106.803035] [<c0152cff>] on_each_cpu+0x1f/0x50
[135106.803046] [<c012cb81>] change_page_attr_set_clr+0x2f1/0x3b0
[135106.803053] [<c012cd77>] _set_memory_wb+0x37/0x40
[135106.803070] [<c012cd9d>] set_memory_wb+0x1d/0x50
[135106.803077] [<c012cde7>] set_pages_wb+0x17/0x20
[135106.803111] [<f85e4221>] ttm_tt_set_caching+0x61/0x1e0 [ttm]
[135106.803142] [<f803406e>] ? agp_free_memory+0x9e/0x120 [agpgart]
[135106.803164] [<f85e43c9>] ttm_tt_free_alloced_pages+0x29/0xb0 [ttm]
[135106.803187] [<f85e4917>] ttm_tt_destroy+0x37/0x90 [ttm]
[135106.803203] [<f85e4fe9>] ttm_bo_release_list+0x59/0xb0 [ttm]
[135106.803223] [<f85e4f90>] ? ttm_bo_release_list+0x0/0xb0 [ttm]
[135106.803242] [<c034cb3d>] kref_put+0x2d/0x60
[135106.803260] [<f85e5f0e>] ttm_bo_delayed_delete+0x8e/0xe0 [ttm]
[135106.803274] [<f85e5f78>] ttm_bo_delayed_workqueue+0x18/0x30 [ttm]
[135106.803291] [<c016369e>] run_workqueue+0x8e/0x150
[135106.803310] [<f85e5f60>] ? ttm_bo_delayed_workqueue+0x0/0x30 [ttm]
[135106.803323] [<c01637e...

I have the same issue with running wow under wine (radeon 8500 le, ubuntu 10.04)

I read somewhere that adding "radeon.modeset = 0" param to kernel boot options should work, although the error does go away the game still doesn't work (looks like a paging error) and it does have some disadvantages, the desktop doesn't show a wallpaper and other rather annoying things

Any new info ?

Some of you might be affected by http://bugs.freedesktop.org/show_bug.cgi?id=28459

This has been fixed upstream in kernel 2.6.35-rc4. From the changelog:

commit f9da52d54eb0e8822b5e7f32ab1cfa6522533d6e
Author: Roland Scheidegger
Date: Sat Jun 12 12:12:37 2010 -0400

    drm/radeon/kms: CS checker texture fixes for r1xx/r2xx/r3xx

    fixes:
    https://bugs.freedesktop.org/show_bug.cgi?id=28459

    agd5f: apply to r1xx/r2xx as well.

    Signed-off-by: Roland Scheidegger
    Cc: <email address hidden>
    Signed-off-by: Alex Deucher
    Signed-off-by: Dave Airlie

madbiologist, I believe that is a different, possibly unrelated bug.

For my part, I am running the linux-image-2.6.35-020635rc4-generic Ubuntu mainline kernel right now, and I still get the "drmRadeonCmdBuffer: -22" error.

Downgrading my kernel to linux-image-2.6.31-02063113-generic makes this problem go away. (This is with X11 packages from the xorg-edgers PPA, but those shouldn't be necessary, since they never seemed to make a difference anyway.)

I believe this is technically a Karmic kernel---it's the most recent available mainline build of 2.6.31, from June 4---but have not noticed any anomalies from my Lucid system running on it. For those interested in this workaround, kernel packages are available here:

    http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.13-karmic/

I've done some further testing, and have two more observations to report:

1. Using the aforementioned 2.6.31 kernel with stock Ubuntu X11 packages works beautifully---no PPA packages are needed. (In fact, with the xorg-edgers packages, I was getting weird core/XI-event-translation errors that prevented the mouse from working in at least two GL apps.)

2. This bug is still reproducible in the Maverick Alpha 2 live CD environment.

Confirmed fixed on maverick.

I no longer get this bug in the Maverick RC live-CD environment, but OpenGL programs now fail in a different manner. See bug #656100.

Maybe in another two years' time, the Radeon driver will work as well as it did two years ago.

I still experience the bug in maverick with all updates

all opengl apps crash with
"drmRadeonCmdBuffer: -22 .."

and these dmesg lines:
[ 210.307120] [drm:r100_cs_track_texture_check] *ERROR* No texture bound to unit 1
[ 210.307129] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !

@Alexandru:
You'll probably need also this patch:
http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=commitdiff;h=a41ceb1c17af06a17c0d88e987215ef20b93c471
that was committed in 2.6.36-rc5 and should hopefully backported to 2.6.35.x.

Fabio, I'm still getting this very bug in an up-to-date Maverick/amd64 install. I am running with

* Radeon packages from the xorg-edgers PPA at https://launchpad.net/~xorg-edgers/+archive/radeon

* The 2.6.36-020636-generic mainline kernel from http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.36-maverick/

I am attaching a RADEON_DEBUG=all log from gltron(1), which crashes quickly and doesn't produce much output. Please let me know what additional information may be useful.

@Daniel:
Probably you are hitting a different bug then or this bug was fixed on ours cards but not on your r200. I'd suggest you to open a new bug report at:
https://bugs.freedesktop.org/enter_bug.cgi?product=Mesa
so the problem can be further analyzed by mesa developers.

https://bugs.freedesktop.org/show_bug.cgi?id=31138

Also still getting this or a similar issue with Ubuntu 10.10 and Radeon X1650 even after switching to a 2.6 mainline kernel build.

$ lspci -nn | grep VGA
01:00.0 VGA compatible controller [0300]: ATI Technologies Inc RV535 [Radeon X1650 Series] [1002:71c7] (rev 9e)

$ glxgears

Running synchronized to the vertical refresh. The framerate should be
approximately the same as the monitor refresh rate.
302 frames in 5.0 seconds = 60.274 FPS

That FPS is way too low for the card. Attempting to resize the window instantly produces a crash and the following message:

drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command stream. See dmesg for more info.

$ dmesg | tail
[ 431.214850] [drm:r100_cs_track_check] *ERROR* [drm] Buffer too small for color buffer 0 (need 400640 have 389120) !
[ 431.214860] [drm:r100_cs_track_check] *ERROR* [drm] color buffer 0 (320 4 0 313)
[ 431.214863] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !

@Alexei: Have a look at

    https://bugs.freedesktop.org/show_bug.cgi?id=30036

It looks like that may address the issue with your card, though the only way to try it is to wait for a new kernel with the fix to come along, or patch and build your own.

@Fabio:

    https://bugs.freedesktop.org/show_bug.cgi?id=25544

*thumbs up*

@Alexei: your problem should be fixed with kernel 2.6.36, it's still not fixed in 2.6.35.x.

@Daniel: great!

Fabio, I meant to say "even after switching to a 2.6.36 mainline kernel build". Maybe I installed one without these patches. Do you know if they made it into any .36 builds yet?

It should be fixed with every 2.6.36 mainline kernel, no need to add other patches.

It doesn't seem http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=commitdiff;h=a41ceb1c17af06a17c0d88e987215ef20b93c471 fixed it. Installed http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.36-rc6-maverick/. I think that's the first .36 release candidate with the patch in it, but I am not sure I am connecting the patch to the mainline release it got merged into correctly. But I previously posted about still having the problem with a more recent kernel from http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.36-maverick/. What should I do?

$ glxgears

Running synchronized to the vertical refresh. The framerate should be
approximately the same as the monitor refresh rate.
302 frames in 5.0 seconds = 60.378 FPS

Instantly after resizing:

drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command stream. See dmesg for more info.

$ dmesg | tail
[ 255.016584] [drm:r100_cs_track_check] *ERROR* [drm] Buffer too small for z buffer (need 612864 have 516096) !
[ 255.016594] [drm:r100_cs_track_check] *ERROR* [drm] zbuffer (384 4 0 399)
[ 255.016597] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !

$ uname -a
Linux 2.6.36-020636rc6-generic #201009291126 SMP Wed Sep 29 12:42:36 UTC 2010 i686 GNU/Linux

Could this be another issue? The error is about "z buffer" this time instead of "color buffer".

@Alexei: for the "z buffer" error look here:
https://bugs.freedesktop.org/show_bug.cgi?id=31152

Thanks, Fabio! I would have to rebuild the entire computer at this point to upgrade the video card since AGP is now PCI, hard disks are SATA, etc., and why should I when everything still works. Or at least worked, up until upgrading to the latest Ubuntu. Do you think older Radeon cards like X1650 will get back to stable operation in 10.10 at some point in the near future?

@Alexei: it should be more stable with the gallium r300 driver (r300g).

Alexei: The R300 Gallium3D driver (R300g) will be the default in Ubuntu 11.04 - see http://www.phoronix.com/scan.php?page=news_item&px=ODcyMg

Sonal Santan and Alexandru Campeanu: You are experiencing the same bug as Daniel Richard G. - see comment #45.

This bug has been fixed upstream in kernel 2.6.37-rc1. A PPA of this kernel should soon be available at http://kernel.ubuntu.com/~kernel-ppa/mainline/

thanks madbiologist, it works fine with the new kernel

Lenovo T60 w. X1400. Kubuntu 10.10 64-bit, installed 2.6.37-RC2, still getting the same messages as people above.

anne@Spock3:~/ripa$ glxgears
Running synchronized to the vertical refresh. The framerate should be
approximately the same as the monitor refresh rate.
drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command stream. See dmesg for more info.

anne@Spock3:~/ripa$ uname -a
Linux Spock3 2.6.37-020637rc2-generic #201011160905 SMP Tue Nov 16 09:08:47 UTC 2010 x86_64 GNU/Linux

anne@Spock3:~/ripa$ dmesg|tail
[ 361.957180] [drm:r100_cs_track_check] *ERROR* [drm] Buffer too small for z buffer (need 467200 have 389120) !
[ 361.957186] [drm:r100_cs_track_check] *ERROR* [drm] zbuffer (320 4 0 365)
[ 361.957189] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !
anne@Spock3:~/ripa$

Has anyone tried Natty, anything in the kernel there that'd help?

@Risto, I think you're seeing the same bug as Alexei in comment #44. Have a look at

    https://bugs.freedesktop.org/show_bug.cgi?id=31152

Please give the packages in the xorg-edgers/radeon PPA a try, as those should have the fix:

    https://launchpad.net/~xorg-edgers/+archive/radeon

Hi All,

I too have got the 'drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command stream. See dmesg for more info.' bug. I have very little Ubuntu/Linux experience would someone be able to guide me step by step on how to fix this.

Reading the above posts I'm sure I'm running 'xorg-edgers/radeon PPA' as someone has in the post suggested that it might of fixed my problem running trying to get Mixxx to run.

Any help will be greatly appreciated.

Regards

Gary

@Gary,

If you're already using packages from the xorg-edgers/radeon PPA, then the only thing left to do is install a mainline kernel, as the kernel-side fixes to this issue have not yet been backported to the normal kernel shipping with Ubuntu Maverick. Have a look at this page:

    https://wiki.ubuntu.com/Kernel/MainlineBuilds

Kernel version 2.6.37 was released recently, so that's the one you'll likely want to use:

    http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.37-natty/

(it says "natty," but it will work fine on Maverick. This is what I'm running myself.)

@Daniel

Thanks, I've downloaded:

linux-image-2.6.37-020637-generic_2.6.37-020637.201101050908_i386.deb
linux-headers-2.6.37-020637-generic_2.6.37-020637.201101050908_i386.deb

Is it just a case of installing the .deb files? If not, how do I install 2.6.37?

@Gary, don't forget linux-headers-*_all.deb.

Yes, all you would need to do is install the packages, and then reboot to actually start using the new kernel. You can run "uname -r" at a terminal prompt to double-check that it is active; the command should print out

    2.6.37-020637rc3-generic

in response.

Fixed on natty.

I have the same issue with Ubuntu 11.10 Oneiric Ocelot, kernel 3.0.0, ATI Radeon Mobility M6 LY chipset

*********************************WARN_ONCE*********************************
File radeon_swtcl.c function r100_swtcl_flush line 325
Rendering was 9 commands larger than predicted size. We might overflow command buffer.
***************************************************************************
drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command stream. See dmesg for more info.

dmesg says:
[ 7238.042912] [drm:r100_cs_track_texture_check] *ERROR* No texture bound to unit 0
[ 7238.042925] [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !