ATI - No X without KMS

olvar · 2010-10-13 11:22:47

I have a laptop with an ATI Xpress 1100 card, which is recognized as an "ATI Radeon XPRESS 200M 5975 (PCIE)". I'm using the xf86-video-ati 6.13.2-2 driver.

Due to a bug in the driver, I cannot have KMS and suspend (to disk or ram) working toghether, and since I use suspend regularly I am forced to deactivate KMS.

Recently, upgrading after a week or so, I found that the Xserver will not start with the following message

...
[    26.773] (II) [KMS] drm report modesetting isn't supported.
[    26.773] (WW) Falling back to old probe method for vesa
...
[    29.611] (II) AIGLX: Screen 0 is not DRI2 capable
[    29.611] drmOpenDevice: node name is /dev/dri/card0
[    29.611] drmOpenDevice: open result is 11, (OK)
[    29.611] drmOpenByBusid: Searching for BusID pci:0000:01:05.0
[    29.611] drmOpenDevice: node name is /dev/dri/card0
[    29.611] drmOpenDevice: open result is 11, (OK)
[    29.611] drmOpenByBusid: drmOpenMinor returns 11
[    29.611] drmOpenByBusid: Interface 1.4 failed, trying 1.1
[    29.611] drmOpenByBusid: drmGetBusid reports pci:0000:01:05.0
[    29.759]
Backtrace:
[    29.759] 0: /usr/bin/X (xorg_backtrace+0x28) [0x46b318]
[    29.759] 1: /usr/bin/X (0x400000+0x5f299) [0x45f299]
[    29.759] 2: /lib/libpthread.so.0 (0x7ffe294f9000+0xf1c0) [0x7ffe295081c0]
[    29.759] Segmentation fault at address (nil)
[    29.759]
Fatal server error:
[    29.760] Caught signal 11 (Segmentation fault). Server aborting
[    29.760]
[    29.760]
Please consult the The X.Org Foundation support
         at http://wiki.x.org
 for help.
[    29.760] Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[    29.760]
[    30.260] (II) RADEON(0): RADEONRestoreMemMapRegisters() :
[    30.260] (II) RADEON(0):   MC_FB_LOCATION   : 0x7fff7e00 0x7fff7e00
[    30.260] (II) RADEON(0):   MC_AGP_LOCATION  : 0x003f0000
...

But if I enable KMS everithing works fine, but I keep loosing the ability to suspend.

Is anyone having the same problem, or could just be a misconfiguration from my part?
By the way, I'm using testing, by I have downgraded xf86-video-ati and xorg-server to extra with the same result.

Thanks!

Leonid.I · 2010-10-13 18:58:18

What are the problems with suspend -- any logs?

As far as I remember, when I used to suppress KMS, I saw no mention of it in Xorg.0.log. Do you disable KMS with the nomodeset kernel option?

EDIT: Wait, I saw the story at https://bbs.archlinux.org/viewtopic.php?id=101461... Is it really a bug in the driver?

Last edited by Leonid.I (2010-10-13 19:03:22)

olvar · 2010-10-13 20:43:09

It actually suspends, the problem is when it comes back, the screen gets "scrambled". I think the bug is this one https://bugzilla.redhat.com/show_bug.cgi?id=501380 although they are different cards.

I disable KMS by adding the line: "options radeon modeset=0" to my /etc/modprobe.d/radeon.conf. Giving it as a kernel parameter would have any aditional effects?

Last edited by olvar (2010-10-13 21:39:08)

Leonid.I · 2010-10-13 21:13:11

Well, I don't know actually. However, KMS is a kernel thing, right? So if you want to disable it, just append

nomodeset

to the appropriate entry in menu.lst.

PS: I have the same card as you on the laptop -- gonna check pm stuff.

olvar · 2010-10-14 11:43:45

Leonid.I wrote:

Well, I don't know actually. However, KMS is a kernel thing, right? So if you want to disable it, just append
nomodeset
to the appropriate entry in menu.lst.
PS: I have the same card as you on the laptop -- gonna check pm stuff.

Ty for your response!
I tried that with the same result. I guess the two methods are equivalent.
Please let me know how things go with your laptop!

Leonid.I · 2010-10-18 02:50:22

... and I'm back -- sorry for a delay

It turns out that pm-stuff was the least configured thing on my system. So, wireless and sound required tweaking, but everything else works. Here is the modesetting-related part of /var/log/pm-suspend.log:

Running hook /usr/lib/pm-utils/sleep.d/98video-quirk-db-handler suspend suspend:
Kernel modesetting video driver detected, not using quirks.

/usr/lib/pm-utils/sleep.d/98video-quirk-db-handler suspend suspend: success.
Running hook /usr/lib/pm-utils/sleep.d/99video suspend suspend:
kernel.acpi_video_flags = 0

/usr/lib/pm-utils/sleep.d/99video suspend suspend: success.

I haven't tried hibernation, but I guess from the video hardware point of view hibernation is the same as suspend...

EDIT: Oh, and I used HAL from within XFCE, which is pretty much the same as just calling pm-suspend as root.
EDIT2: And I assume that you checked pm-suspend without X, right?

Last edited by Leonid.I (2010-10-18 16:35:06)

Arch Linux

#1 2010-10-13 11:22:47

ATI - No X without KMS

#2 2010-10-13 18:58:18

Re: ATI - No X without KMS

#3 2010-10-13 20:43:09

Re: ATI - No X without KMS

#4 2010-10-13 21:13:11

Re: ATI - No X without KMS

#5 2010-10-14 11:43:45

Re: ATI - No X without KMS

#6 2010-10-18 02:50:22

Re: ATI - No X without KMS

Board footer