Arch Linux

xorg log file is on /var/log/Xorg.*

I decided to reinstall archlinux.

I used an ftp install when installed the latest kernel i.e. kernel26-2.6.22.2-1

When I started the X.org Server it froze the computer again.
I needed to do a restart the computer by pressing the power button.

I reinstalled the older kernel i.e. kernel26-2.6.21.6-1 and that works.

I use the video driver xf86-video-i810


This is the Xorg error
----------------------------------------------------
Fatal server error:
lockup

(II) AIGLX: Suspending AIGLX clients for VT switch
Error in I810WaitLpRing(), now is 63899, start is 61898
pgetbl_ctl: 0x1f200001 pgetbl_err: 0xb078000
ipeir: 0 iphdr: ffffffff
LP ring tail: 48 head: 2c len: f001 start 504000
eir: 0 esr: 1 emr: 3d
instdone: ff7a instpm: 0
memmode: 4 instps: 10
hwstam: 9ac7 ier: 0 imr: 9ac7 iir: 0
space: 65500 wanted 65528

FatalError re-entered, aborting
lockup

I fixed the problem. The module "dri" in the xorg.conf file was some how causing X.org to lockup.

Section "Module"
Load "ddc" #ddc probing of monitor
Load "dbe"
Load "extmod"
Load "glx"
Load "bitmap" #bitmap-fonts
Load "type1"
Load "freetype"
Load "record"
#Load "i2c"
#Load "vbe"
#Load "int10"
#Load "dri" #This module was causing X to crash
EndSection

dri is installed by the package libgl-dri.
the dri module did not cause X.org to lockup before kernel26-2.6.22.1-3

Should I close this bug?

I forgot to say that the window manager I was using was icewm. The latest version from ftp, 1.2.30-1

status with latest kernel?

Using the latest kernel26 version 2.6.23.1-5 xorg freezes if I don't uncomment the glx module in xorg.conf
Without this module some games run very slow (3D acceleration?) but other than that everything else is fine.

I think this is the error's log, from /var/log/Xorg.0.log.old

Error in I810WaitLpRing(), now is 14013737, start is 14011736
pgetbl_ctl: 0x1e6a0001 pgetbl_err: 0x0
ipeir: 0 iphdr: ffffffff
LP ring tail: 88 head: 2c len: f001 start 504000
eir: 0 esr: 1 emr: 3d
instdone: ff7a instpm: 0
memmode: 4 instps: 10
hwstam: 9ac7 ier: 0 imr: 9ac7 iir: 0
space: 65436 wanted 65528
(II) I810(0): [drm] removed 1 reserved context for kernel
(II) I810(0): [drm] unmapping 8192 bytes of SAREA 0xe09d0000 at 0xb7a74000

Fatal server error:
lockup

(II) AIGLX: Suspending AIGLX clients for VT switch
Error in I810WaitLpRing(), now is 14018743, start is 14016742
pgetbl_ctl: 0x1e6a0001 pgetbl_err: 0x0
ipeir: 0 iphdr: ffffffff
LP ring tail: 90 head: 2c len: f001 start 504000
eir: 0 esr: 1 emr: 3d
instdone: ff7a instpm: 0
memmode: 4 instps: 10
hwstam: 9ac7 ier: 0 imr: 9ac7 iir: 0
space: 65428 wanted 65528

FatalError re-entered, aborting
lockup

Status with the latest kernel is fine. kernel 2.6.23.1-6

However if I use 16 bit colour in the xorg.conf file X crashes.

Section "Screen"
Identifier "Screen0"
Device "Card0"
Monitor "Monitor0"
DefaultDepth 16
SubSection "Display"
Viewport 0 0
Depth 1
EndSubSection
SubSection "Display"
Viewport 0 0
Depth 4
EndSubSection
SubSection "Display"
Viewport 0 0
Depth 8
EndSubSection
SubSection "Display"
Viewport 0 0
Depth 15
EndSubSection
SubSection "Display"
Viewport 0 0
Depth 16
EndSubSection
SubSection "Display"
Viewport 0 0
Depth 24
EndSubSection
EndSection

Are you guys getting anywhere with this ? I run i810 driver and have to swict off DRI by setting color depth to smoething than than 16. I would like to have DRI enabled. Any joy ??

status on .24 kernel from testing?

still an issue?

Yes, 2.6.24.

Running 2.6.24 too and it's still an issue.
Has anyone found any workaround to successfully run xorg on 16-bit color depth other than downgrading the kernel?

Did anybody try 2.6.25 from testing?

Hey guys,
I think I know what the problem is -

the xf86-video-i810 driver has been deprecated; and I've seen a comment by intel developers confirming this.

The xf86-video-intel is what I use, and it works fine.

Oh, and you might try to add

"-br"

to your defaultserverargs in the startx script.

-- vsk

Is this ok to close now?

Unfortunately the problem isn't solved by the new xf86-video-intel driver (this has been knowned for sometime for people who desperately can't make this old piece of hardware work under 16 bit color depth to enable acceleration).

This issue doesn't seem to affect all the ships though according to the post from Vedant Kumar (vsk).

So could everyone still experiencing this problem post their lspci output for the video here? Here's mine:

00:02.0 VGA compatible controller: Intel Corporation 82815 Chipset Graphics Controller (CGC) (rev 02)

I have two machines with this problem - here is the lspci from both:

1) 00:02.0 VGA compatible controller: Intel Corporation 82815 Chipset Graphics Controller (CGC) (rev 02)

2) 00:01.0 VGA compatible controller: Intel Corporation 82810 (CGC) Chipset Graphics Controller (rev 03)

As you can see both have different chipsets but both have this problem. I have this problem with both the intel and the i810 drivers.

I've been following this bug report for a while, as I had similar problems. It seemed that i managed to solve it, somehow, and i was happy :P . So, a few days ago, i modified my xorg.conf to delete "unuseful" lines.
When i booted again, I had the screen with a mix of colours and the mouse pointer (impossible to move it).
Trying to restore the previous situation, it seems i found my problem: i can't delete the modeline about my resolution. If i do so, the problem is there again. And i don't know why it works like that (with the same configuration, on a gentoo box, i have no problems). But what i understood, is that my problem was not a bug, it's just in my configuration (human mistake as most of times).
This is just to say: try to look in your xorg.conf, try to make your own modeline (you can use http://xtiming.sourceforge.net/ for example, that worked for me), before assuming it's a bug.
In two words: don't desperate :D

Hi pikiweb.

No luck with custom modeline, the exact same freeze happens. Did you actually revert back to 16 bit color depth after your changes? (the problem only happens in 16 bit color depth)

Did you use any special value for dot clock frequency?

Yes, i'm using 16 bit color now, without evident problems

I didn't understand your last question about dot clock frequency..my english is too bad!

There's a field called "Dot Clock Frequency" in http://xtiming.sourceforge.net/, did you fill in some special value here or in any other field before generating your modeline?

Tried the custom Modeline - same problem still here..

no, i didn't fill it at all (as you maybe understood, i don't know what "Dot clock frequency" is). I only put my values for the Horizontal and Vertical frequencies, the resolution, the refresh rate, the aspect ratio. I think i both tried with and without Interlace.
But it seems that it worked for me..If you want, i can post my xorg.conf as soon as possible, so you can compare with yours...

Please post your xorg.conf when possible.

here it is. It's not "tidy" but it works. I hope it helps! :)

Thanks for posting your xorg.conf - unfortunately it doesn't solve my problem - X locks up the same as with my original xorg.conf.

Same here...

Pikiweb could you please post your lspci output for the video?

sure, here is the important part:

00:01.0 VGA compatible controller: Intel Corporation 82810E DC-133 (CGC) Chipset Graphics Controller (rev 03)

status with 2.6.26 ?

anything new?

I have given up on this and settled for no DRI on this machine.