[Lucid] Radeon Xpress 200M needs PCI quirk to fix or disable MSI

I get the third "possibility" (grub -> blackscreen) with an Xpress 1250. Not sure how to try nonmodeset

I have this problem with a Mobility Radeon X1600. I also get screen corruption with KWIN with the nomodeset solution.

Unless I add radeon.modeset=0 (i.e. disable KMS) I invariably get a screen corruption after log-on (using kdm). That is, I can log on but at that point the screen will ‘black out’ and the computer/X will lock up.

I do not see any screen corruption with kwin using radeon.modeset=0, though. lshw reports:

lshw -C display
WARNING: you should run this program as super-user.
  *-display UNCLAIMED
       description: VGA compatible controller
       product: RC410 [Radeon Xpress 200M]
       vendor: ATI Technologies Inc
       physical id: 5
       bus info: pci@0000:01:05.0
       version: 00
       width: 32 bits
       clock: 66MHz
       capabilities: bus_master cap_list
       configuration: latency=64 mingnt=8
       resources: memory:90000000-9fffffff(prefetchable) ioport:7800(size=256) memory:f88f0000-f88fffff memory:f88c0000-f88dffff(prefetchable)

It would help if there would be kernel message log from original problem. Setting up netconsole to stream the kernel messages to 2nd machine is probably the easiest solution.

Screen corruption is different bug and most likely caused by problematic AGP transfers.

You can either set your card to PCI mode or disable download from screen accerlation from exa.

Both will affect performance quite a lot.

I have the same result on my HP6735b with a AMD RS780M/RS780MN (Radeon HD 3200) using ati or radeonhd as driver.
On boot up I see non sense on the screen. The I can see the log on screen and the computer freeze. Sometimes just on begin of showing the screen, sometimes after selecting one user...
Starting with vesa a default works fine, just with the wrong resolution.

Erni35

additional comment:

starting a rescue console and then as root startx works fine with the radeon driver and vesa too

Erni35: "starting a rescue console"
How?

Today I tried to boot up the LiveCD again,the desktop showed then I went to the Apperance menu and disabled Compiz. Then I was browsing the Administration menu the whole screen turned white. Ctrl+Alt+F1 and Alt+SysRq+R+S+E+I+U+B didn't work --> hard reset. Everything worked fine for 20 seconds. YEAH!! :D Thats a progress.

I have the following code in my grub.cfg

menuentry "Ubuntu, with Linux 2.6.33-rc7-dietmar3 (recovery mode)" {
        recordfail
        insmod ext2
        set root=(hd0,7)
        search --no-floppy --fs-uuid --set 464f0d6e-adda-415d-97b9-d606e1b836b5
        echo Loading Linux 2.6.33-rc7-dietmar3 ...
        linux /boot/vmlinuz-2.6.33-rc7-dietmar3 root=/dev/sda7 ro single
        echo Loading initial ramdisk ...
}

I found one solution. Please verify this. I found many files from my AMD driver installation just after deinstalling them.
I deleted all *fglrx* files and reinstalled the kernel (2.3.33 too) and hav now just problems after waking up from screen suspend.

Erni35

I've got the syslog with a persistent USB daily image. It says the problem is in the kernel DRM.

Having trouble with KMS too, starting about 4 weeks ago. First blamed Plymouth, but although there are issues there as well, I'm now sure that this is a separate problem. (see also bug 519891).

Problem comes and goes, about half the time I can boot normally and get into my system fine. The other times the system gives me a pattern of vertical coloured lines and is unusable. Hard reboot and adding 'nomodeset' to the grub linux-line does give me a usable system.

This is a Dell Latitude 131L, AMD64, Lucid with all updates applied at least once a day (usually twice a day).
From lspci:
01:05.0 VGA compatible controller: ATI Technologies Inc RS482 [Radeon Xpress 200M]
From grub.cfg:
menuentry "Ubuntu, with Linux 2.6.32-13-generic" {
        recordfail
 insmod ext2
 set root=(hd0,5)
 search --no-floppy --fs-uuid --set 46062d1f-f00c-4276-9ff1-541279640a93
 linux /boot/vmlinuz-2.6.32-13-generic root=UUID=46062d1f-f00c-4276-9ff1-541279640a93 ro quiet splash
 initrd /boot/initrd.img-2.6.32-13-generic
}

I insert the 'nomodeset' right before the 'quiet splash'

There are plenty of systems out there with this chip, and installing Lucid will break them as it is now. This should be concerned important, probably critical.

If I can do more to help get this fixed, please ask.

cheers
Tom

Arch Linux and Fedora has the same problem
http://bbs.archlinux.org/viewtopic.php?id=86211&p=3
https://bugzilla.redhat.com/show_bug.cgi?id=531825

One interesting post:

"I upgraded today to

kernel-2.6.31.6-142.fc12.i686
xorg-x11-drv-ati-6.13.0-0.12.20091119git437113124.fc12.i686

The kernel update did not seem to cause any change, but the radeon driver
update did caused one change, instead of getting a black background with
working mouse cursor, I instead get the fedora wallpaper with working mouse
cursor.

syslog is still getting filled with the same errors as before. "

From 2.6.33-rc7 changelog:

drm/radeon/kms: fix r300 vram width calculations

Maybe that will fix this. I will check it out If I will have time.

https://bugzilla.redhat.com/show_bug.cgi?id=532308
"I installed almost all recent kernel packages in koji. 2.6.31.5-91.rc1 is the
lastest stable X200M/KMS kernel. 2.6.31.5-96 is the first buggy kernel."

Any decision about this serious regression/bug? Will UMS be used for X200M in Lucid?

With 'nomodeset' I still get the garbled screen (and indeed tiny pixels
whenever I type something) due an issue in Plymouth, but can fix it by
using alt-sysreq-k.

Without 'nomodeset' I'm getting all vertical lines and cannot use my
system. Can not get past that with alt-sysreq-k or ctrl-alt-backspace.
Every once in a while it actually boots ok without the 'nomodeset' and
then Plymouth works ok too, but at the moment I feel this is getting
less frequent.
Because I can't reproduce the behaviour without 'nomodeset' right or
wrong, I feel there is some issue to do with timing or a race condition
somewhere between the kernel and the radeon driver.
But I'm not a dev.

Updating Lucid twice a day. I'm not trying daily cd's unless I can't get
into my system even by booting into rescue mode.

cheers
Tom

Lollerke wrote:
> https://bugzilla.redhat.com/show_bug.cgi?id=532308
> "I installed almost all recent kernel packages in koji. 2.6.31.5-91.rc1 is the
> lastest stable X200M/KMS kernel. 2.6.31.5-96 is the first buggy kernel."
>
> Any decision about this serious regression/bug? Will UMS be used for
> X200M in Lucid?
>
> ** Bug watch added: Red Hat Bugzilla #532308
> https://bugzilla.redhat.com/show_bug.cgi?id=532308
>
>

Still no change here.

Just loaded today's updates and was asked to reboot.
Was able to boot *once* without 'nomodeset' and everything worked just fine. Tried it again and got the vertical lines again. And again. And again.
So I booted with 'nomodeset' and got the -well known- garbled screen, so I gave that the alt-sysrq-k routine. After that I could log on and use my pc ok.

cheers
Tom

From the Fedora bug report:

"No change with 2.6.33 or drm-radeon-testing. Rest is git master.

Is there some more investigation that can be done? This is really a showstopper
for switching to KMS with RV250 AGP, and as far as I know KMS left staging in
2.6.33. "

Why this bug isn't marked as Confirmed and not linked to a milestone/tracked in Lucid? In Red Hat Bugzilla it's confirmed and the importance is set to high. This is a show stopper for old Radeon card users and Lucid will ship KMS enabled. What other information I need to provide to this bug?

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

I tried today the mainline kernel and have no problems till now.
If I found some of these, i report this.

Dietmar

What kernel version exactly?
I've got a new error in the log with the -15.22 kernel:
"Mar 3 10:36:39 toshibal30 kernel: [ 29.883648] [drm:drm_mode_getfb] *ERROR* invalid framebuffer id"

update:
Got a lot of upgraded packages over the last days, including X and a new kernel. Result is still the same, and still disappointing.
Loaded the new kernel and X yesterday, rebooted and everything worked fine. No screen corruption, vertical lines etc, Plymouth also doing exactly what it should.
Logged in, logged out, rebooted. No Plymouth, just vertical lines and unusable system. Killed X, but that would not help either.
So now I'm back at booting without KMS (using the 'nomodeset' option) which gives me a garbled screen. Luckily I can get past that by killing X and then everything works pretty well.

I really think this should be picked up and fixed, since it is the sort of 'does *not* work out of the box' issue that drives people away from Linux back to Winblows.

Importance = high
Serious regression

cheers
Tom

This problem is somewhat related to this bug:
https://bugs.launchpad.net/bugs/305301

Seriously, this is an almost 2 year old bug, and just got any worse with kms... :(

There may be some relation to 305301, but on my system suspend / hibernate and resume worked perfectly in Jaunty. On moving to Karmic both got broken and are still not fixed completely (see 466868-susp/res and 447192-hib/res)
The problem here has no much obvious to do with that, because it appears right after (cold) booting and leaves you with an unusable system.

cheers
Tom

I tried 2.6.33-020633-generic #020633 SMP Thu Feb 25 10:10:03 UTC 2010 x86_64 GNU/Linux
from mainline just wirh quiet and slash as kernel command line and the radeon driver, not radeonhd

kr,
erni35

With 2.6.33-020633-generic and KMS,everything was fine for 5 minutes then I got a white screen and nothing worked.

Yup, the same here... :( I've also tried the same kernel with karmic, needless to say what happened...

@#37:
That bug also mentions that black/white screen can (and often does) happen after cold boot, so imho these to problems are from the same origin.
I'd love to help bisecting this problem, but don't have a clue, how. Any suggestion, where and how to start?

Cold booting works fine for me. Rarely I have got a black screen but after pressing CTRL+ALT+F1 then CTRL+ALT+F7 everything is back to normal. That is another problem which has a bug report.

New information from the Red Hat bugzilla report (the problem may be in radeon_fence.c - line 159):

Mar 4 15:41:04 t41 kernel: WARNING: at
drivers/gpu/drm/radeon/radeon_fence.c:159 radeon_fence_signaled+0x56/0x83
[radeon]()
Mar 4 15:41:04 t41 kernel: Hardware name: 2373TG5
Mar 4 15:41:04 t41 kernel: Querying an unemited fence : f4fa7ba0 !
Mar 4 15:41:04 t41 kernel: Modules linked in: tun sunrpc cpufreq_ondemand
acpi_cpufreq ip6t_REJECT nf_conntrack_ipv6 ip6table_filter ip6_tables ipv6
uinput hdaps input_polldev arc4 ecb ath5k mac80211 ath ppdev cfg80211 nsc_ircc
snd_intel8x0 snd_intel8x0m parport_pc irda snd_ac97_codec snd_seq crc_ccitt
parport ac97_bus thinkpad_acpi rfkill iTCO_wdt iTCO_vendor_support
snd_seq_device snd_pcm e1000 snd_timer snd soundcore snd_page_alloc i2c_i801
joydev dm_multipath video output yenta_socket rsrc_nonstatic radeon ttm
drm_kms_helper drm i2c_algo_bit i2c_core [last unloaded: scsi_wait_scan]
Mar 4 15:41:04 t41 kernel: Pid: 1512, comm: Xorg Not tainted
2.6.32.9-64.fc12.i686 #1
Mar 4 15:41:04 t41 kernel: Call Trace:
Mar 4 15:41:04 t41 kernel: [<c043a2fd>] warn_slowpath_common+0x6a/0x81
Mar 4 15:41:04 t41 kernel: [<f7daf28c>] ? radeon_fence_signaled+0x56/0x83
[radeon]
Mar 4 15:41:04 t41 kernel: [<c043a352>] warn_slowpath_fmt+0x29/0x2c
Mar 4 15:41:04 t41 kernel: [<f7daf28c>] radeon_fence_signaled+0x56/0x83
[radeon]
Mar 4 15:41:04 t41 kernel: [<f7daf2f7>] radeon_fence_wait+0x3e/0x2a1 [radeon]
Mar 4 15:41:04 t41 kernel: [<f7daf66f>] ? radeon_fence_create+0x21/0xda
[radeon]
Mar 4 15:41:04 t41 kernel: [<f7dbda04>] radeon_ib_get+0xf4/0x19e [radeon]
Mar 4 15:41:04 t41 kernel: [<f7dbe87a>] radeon_cs_ioctl+0x80/0x162 [radeon]
Mar 4 15:41:04 t41 kernel: [<f7c9194f>] drm_ioctl+0x251/0x2fa [drm]
Mar 4 15:41:04 t41 kernel: [<f7dbe7fa>] ? radeon_cs_ioctl+0x0/0x162 [radeon]
Mar 4 15:41:04 t41 kernel: [<c04541f5>] ? autoremove_wake_function+0x0/0x34
Mar 4 15:41:04 t41 kernel: [<c0586d10>] ? file_has_perm+0x89/0xa3
Mar 4 15:41:04 t41 kernel: [<f7c916fe>] ? drm_ioctl+0x0/0x2fa [drm]
Mar 4 15:41:04 t41 kernel: [<c04e5826>] vfs_ioctl+0x1d/0x76
Mar 4 15:41:04 t41 kernel: [<c04e5dc0>] do_vfs_ioctl+0x493/0x4d1
Mar 4 15:41:04 t41 kernel: [<c0586fb4>] ? selinux_file_ioctl+0x43/0x46
Mar 4 15:41:04 t41 kernel: [<c04e5e44>] sys_ioctl+0x46/0x66
Mar 4 15:41:04 t41 kernel: [<c040365c>] syscall_call+0x7/0xb
Mar 4 15:41:04 t41 kernel: ---[ end trace d14a0fcc30e3dc30 ]---
Mar 4 15:41:04 t41 kernel: [drm:radeon_ib_schedule] *ERROR* radeon: couldn't
schedule IB(4).
Mar 4 15:41:04 t41 kernel: [drm:radeon_cs_ioctl] *ERROR* Faild to schedule IB
!

since 4-5 days I have no problems with
menuentry "Ubuntu, with Linux 2.6.33-020633-generic" {
        recordfail
        insmod ext2
        set root=(hd0,7)
        search --no-floppy --fs-uuid --set 02d66724-aea6-4406-a59c-fc2edce4bb4c
        linux /boot/vmlinuz-2.6.33-020633-generic root=UUID=02d66724-aea6-4406-a59c-fc2edce4bb4c ro quiet splash
        initrd /boot/initrd.img-2.6.33-020633-generic
}

with my
01:05.0 VGA compatible controller: ATI Technologies Inc RS780M/RS780MN [Radeon HD 3200 Graphics]

graphics and kernel 2.6.33-020633-generic #020633 SMP Thu Feb 25 10:10:03 UTC 2010 x86_64 GNU/Linux

kr,
erni35

today in the morning i run in the same trouble. I started with the 2.6.33 as above and after login the notebook freeze.
But then I realised the change to the situation before. I turned on visuel effects. After switch them off, everything is fine again.

I work with 2.6.33 and radeon driver.

Erni35

one comment, I work with a gnome desktop.

Erni35

Erni35: My desktop PC has the same graphics card (ATI Radeon HD3200 - RS780) card and KMS worked perfectly with 2.6.32.

I think the ATI KMS code is not ready for stable use. Sometimes it works,but serious problems appear on all kinds of GPUs (R100,R200,R300,R400,R500 and R600),but I never heard an R700 problem.

Kernel 2.6.32-16.23 is out: *update DRM to mainline v2.6.33 (from changelog).

It seems that KMS is causing problems on Compaq Evo N1015v as well. Screen freezes after some time. Compiz was extremely slow so I disabled it.

Running sudo lshw -c video with KMS enabled froze the computer as well.

$ sudo lshw -c video
[sudo] password for matej:
  *-display UNCLAIMED
       description: VGA compatible controller
       product: Radeon Mobility U1
       vendor: ATI Technologies Inc
       physical id: 5
       bus info: pci@0000:01:05.0
       version: 00
       width: 32 bits
       clock: 66MHz
       capabilities: agp agp-2.0 pm bus_master cap_list
       configuration: latency=66 mingnt=8
       resources: memory:f0000000-f7ffffff(prefetchable) ioport:9000(size=256) memory:e8100000-e810ffff memory:e8120000-e813ffff(prefetchable)

$ lspci | grep ATI
00:00.0 Host bridge: ATI Technologies Inc AGP Bridge [IGP 320M] (rev 13)
00:01.0 PCI bridge: ATI Technologies Inc PCI Bridge [IGP 320M] (rev 01)
01:05.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility U1

Alt+SysRq + R + S + E + I + U + B worked today so I have got log file. Same as before...
I think that the X200M needs to be KMS blacklisted because the newest DRM didn't help.

I have a Radeon Xpress 200M in a Dell Inspiron 1501 working quite well with 2.6.32-16-generic

VGA compatible controller [0300]: ATI Technologies Inc RS482 [Radeon Xpress 200M] [1002:5975]
Subsystem: Dell Device [1028:01f5]

http://en.wikipedia.org/wiki/Xpress_200
I have the Intel version (RS480).

I also have the Intel chipset version, running on a Amilo LI1718 Laptop. Still having problems.

Need to set radeon.modeset=0 to even be able to boot to a usuable desktop, but suspend still doesn't work, and Hibernate doesn't seem too reliable either.

This is such a crazy graphics chip: some people having severe problems with it, others running flawless. Could it be due to the AMD/Intel chipset versions being slightly different (apart from just an increase of 100 MHz in clock-speed)?

I also have a FuSi Amilo Li1718 notebook with Ati X200m GPU. Suspend/resume does not work on Karmic. Some guyes promised that the 2.6.33 version of Linux kernel will solve the problem but it doen't.

If you want to use accelerated graphics without any problem and have working suspend/resume I suggest you to downgrade to Jaunty and after that downgrade your xorg packages from Intrepid's repos and also install the legacy proprietary fglrx driver. (After xorg downgrade you need to lock the version of related packages and you can use the Jaunty repos again) It works fine for me. I think the xorg downgrade doen't work from Karmic.

@Tim (#50):
Mine's a Dell Latitude 131L, which is basically the same board but in a business casing.
Having nothing but trouble since 2 months and can only use my laptop booting with 'nomodeset'.

Seems we should find out why it does work for you. Must be a difference between the pc's. Any info I should collect / output I should post here?

cheers
Tom

Is the apport-collect command limited to one time use?
Because if I run "apport-collect 509273" the questions are coming in infinite loop and I can't update the bug with the newest 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

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

I tried to run with kernel 2.6.33 from http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.33/.

KMS issues don't show up. Graphics, however, are very, very slow. Much slower than in Karmic.

Read Matej's comment, so I thought to try something funny: got the daily kernel and headers from kernel.ubuntu.com (after all this aint M$, we can boot more than one kernel).Anyway, same result as before:
- Installation, reboot, all works fine.
- Log in, log out, reboot to double check.
- Vertical coloured bars and unusable system. Hard reset.
- Booting with 'nomodeset'. system comes up, but logging on crashes X.
- X comes back and I can log on and use my pc.

Linux tom-laptop 2.6.34-999-generic #201003101003 SMP Wed Mar 10 10:10:19 UTC 2010 x86_64 GNU/Linux

I keep wondering why everything seems to work after a kernel upgrade, but just once. After that it's pretty much hopeless.
What changes between 2 consecutive boots? Is readahead involved somehow? Is it a race condition tickled by the way, order or speed that some parts are loaded??

cheers, Tom

I don't know if it could help,
I have an Amilo Li1718 with ATI X200M.
I just tried the alpha release from Wednesday, 10 March 2010, in the start up page of the live CD(USB), I set the options (F6) to "nomodeset" and now I can boot normaly, they was one issue when I press "enter" and went back to GDM, all others work better since I can write this text now and explore the live USB now. I hope it could hep

hi,
 just to confirm that, still with "nomodeset", i can boot properly to the desktop, at hitting Entry key brings me back to GDM,after that I have no problem. One point to mention: I don't have any more plymooth when I choose "nomodeset" option. may be Plymooth has something to do with this issue. I'm still writing from Live USB.

Also on an AMilo LI1718.

I think it may be GDM that is causing the screen corruption when booting with radeon.modeset=0, as if I boot Kubuntu (and therefore KDM instead of GDM) with the same option I don't get a corrupted log in screen where I have to ALT+SysRq+K to reset it.

Regardless of that I _still_ can't suspend and resume, and Hibernate doesn't always function. I need at least one of the to work as using a laptop is unusable without them.

Incidentally, after trying and failing to suspend and resume, networking in Kubuntu (and probably Gnome (haven't tested yet) seems to die, being "unmanaged". Having the Wicd network application installed gets around this, as that still works.

This is getting interesting :-)

Yesterday there were updates to Xserver-core and the ATI/Radeon driver in the same batch that contained Plymouth 0.8.0~-14.
To my surprise I can now start up without the dreaded 'nomodeset' option. This works fine, Plymouth and X all doing what they should first time around. So far, so good.

In order to find out what Plymouth would do when booting under 'nomodeset' (triaging another bug) I rebooted and tried that. Guess what..... the situation has reversed. Using 'nomodeset' now gives me a screen with vertical coloured bars and lines, and there it all ends.

I guess I can't complain about my system finally booting without setting extra options, but it sort of bothers me that it now bugs the other way around.
Still, I'm happily using my system at the moment.

cheers
Tom

Just tried the latest (14th) Ubuntu Daily LiveCD and using "nomodeset" doesn't even get me to the desktop now, the display doesn't kick in after the splash. And without nomodeset gets me there only for the display to switch itself off after a few minutes, which is different than the screen freezing and becoming garbled, but the end result is the same.

So whatever the most recent changes were, they've been a regression for me, as at least with nomodeset I could have a fully working desktop experience (no resume from Suspend is another bug).

Wish I could edit previous comments...

Wanted to add that I've given up bothering to test linux on this Amilo LI1718 laptop, as thanks to the myriad of hard power-offs and ons (a hell of a lot of them) when the display dies, it's now sounding none too healthy. What life it has left will now be spent in either windows XP, Vista or 7 for which feature-freezed working drivers are available.

Spoke too soon (#89).
Tried hibernating (worked) and resuming (crashed). Did a hard reboot and behaviour is back to where it was; 'nomodeset' when booting else vertical lines y nada mas.

(sigh!)
I do hate it when errors come and go at -seemingly- random moments. There must be a way to reliably reproduce this.

cheers Tom

Will this card (ATI X200M Intel version) be KMS blacklisted in Beta1?

Did some googling on ATI and KMS and found the following:
Over at freedesktop.org there are multiple bugs linked to KMS and ATI cards. Most of it *should* be fixed in kernel 2.6.33 or above, which is not the one that Lucid runs now (2.6.32).
I linked this bug to the upstream one most likely to be related on bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=25662

This information is confirmed by https://wiki.ubuntu.com/X/KernelModeSetting. Here it is also stated that the latest kernel will be better when using KMS and Radeon cards. Also this page mentions that there is indeed a race condition possible with KMS and ATI:

=======
Module loading issues on ATI

If you get "RADEONDRIGetVersion failed" in Xorg.0.log it is because of a race issue. The radeon module is not initialized in time before X is starting, and X will falsely believe that there is no KMS support and do its own modesetting...
To make sure the radeon modules is initialized before gdm (and X) is starting, insert "modprobe radeon" into your /etc/init/gdm.conf just before the gdm-binary is executed:
    ...
    initctl emit starting-dm DM=gdm

    modprobe radeon
    exec gdm-binary $CONFIG_FILE
end script
======

So I installed the daily from http://kernel.ubuntu.com/~kernel-ppa/mainline/daily/current/ dated 20100315.
I also edited /etc/init/gdm.con as described above. (Yes, that is 2 changes in 1 go, but I felt if one would fix it, the other wouldn't hurt.)
Results: this seems to help a bit. I've had a number of boots successfully without 'nomodeset'. One time it did fark up but rebooting gave me a good system. So not solved yet, but probably on the right track.

I think I see a pattern that things may be stable for a while (for better or for worse), but when something makes ureadahead regenerate its data things work OK once, then break. We might conclude that ureadahead speeds up booting so much, that it triggers the existing race condition between the kernel and X both trying to load the ATI driver .
In short: it trips over its own feet and falls flat on its face....

Will report back later.

cheers Tom

I tried to add "modprobe radeon", but I still get a screen freeze. I couldn't reach the link you post here(@Tom Louwrier), it saids this is a forbiden page. I still have to add at the grub editing screen "nomodeset" to be able to work with Lucid.
I'm booting without any trouble without "nomodeset", but the screen get frozen at random time.
With "nomodeset" I don't have any trouble.

I downloaded the new Beta1, I have done the update too and I still the same trouble. I still have to put "nomodeset" in grub at boot time to work properly. Without "nomodeset" I still get the screen frozen.

Exact information on which commit causes this problem (from the the freedesktop bug):

Starting with the 2.6.31 kernel, KMS works fine until this commit:
3e5cb98dfe87cc61d0a1119dd8aa2b1e4cfab424
After this commit, xdm freezes at the login screen, but I can still switch to
another virtual terminal. My dmesg log has the message described in comment #3
repeated several times. This behavior continues until commit:
6b46362c0ea472b174c336786fd406c504326ad4
After this commit, I experience the behavior described in comment #1.

I am now using X server 1.7.5 and the latest git versions of mesa, libdrm, and
ddx.

You are right, the link I used to get the daily kernel build is now invalid. Try this one:
http://kernel.ubuntu.com/~kernel-ppa/mainline/daily/?C=M;O=D

You will want to try the most recent edition of Lucid or Karmic, depending on your current distro.

gr
Tom

I had this problem with lucid and had to put radeon.modeset=0 into grub to boot. Later, due to lucid breaking after an update (PS disabling ctrl-alt-backspace in lucid is a bad idea, and even with the radeon.modeset=0, ctrl-alt-f1 showed vertical bars, so I couldn't get to a terminal from there to restart the gnome-panels or nautilus), I did a fresh install without importing any of my old settings and the problem seemed to go away, even though I still had the vertical bar problem with the live cd itself. I can only hope that the problem is solved, because the vertical bars do not always appear on every boot and mostly occur randomly, and I have not had much of a chance to thoroughly test my laptop with many boots.

Using the 2.6.33 kernel completely fixed the problem, but wireless was nonfunctional in that kernel build so I couldn't stick to that.

Okay, it's still happening. It just happened not to on the first 3 or 4 boots for whatever reason.

I tried v2.6.34-rc1 from PPA and still the same...

There's a fix in the freedesktop bug: http://bugs.freedesktop.org/show_bug.cgi?id=25662#c12

Will this patch be included in the beta 2 release?

Well, i hope it will!
Btw, using this patch, do you have to use the nomodeset option?

2010/3/24 Lollerke <email address hidden>

> Will this patch be included in the beta 2 release?
>
> --
> [Lucid] ATI KMS causes DRM error on X200M
> https://bugs.launchpad.net/bugs/509273
> You received this bug notification because you are a direct subscriber
> of the bug.
>

"Btw, using this patch, do you have to use the nomodeset option?"

No, everything is gonna work as it should (including the flicker free boot ,because the boot screen will use the same resolution as the dekstop).

Until the patch arrives here's the solution: instead of nomodeset, boot with pci=nomsi command. This is doing exactly the same as the patch.

With pci=nomsi I have KMS working and no crashing so far.

danwood76: Yeah, it works (my uptime is over 1 hour now with KMS)

Tried that, got vertical lines immediately.
Back to using 'nomodeset' :-(

Tom

Also tried it, works fine, only if the graphics wasn't that damn slow...
Graphical performance is simply 2-3 times worse, than in karmic.

2010/3/25 Tom Louwrier <email address hidden>

> Tried that, got vertical lines immediately.
> Back to using 'nomodeset' :-(
>
> Tom
>
> --
> [Lucid] ATI KMS causes DRM error on X200M
> https://bugs.launchpad.net/bugs/509273
> You received this bug notification because you are a direct subscriber
> of the bug.
>

Tom: Have you tried it with today's daily live CD? It must work.
Karesz: Thats another bug.

@Lollerke: Nope, I'm running Lucid on my everyday laptop as working installation. Lucid amd64 with updates loaded at least twice a day.
I know it's not really production grade yet, but it's the only way for me to test it and take it seriously.

Tom

Tom: I have Lucid installed on my laptop too. The only difference is that I have the i386 version. Maybe the amd64 version is causing your problems.

Adding that line did not fix it for me at all.
Tristan

Tristan: amd64 or i386? Do you have the Intel version of X200M?

i386, It's an Asus F5RL laptop, ATI Radeon XPress 1100,
lspci:
00:00.0 Host bridge: ATI Technologies Inc Device 5a31 (rev 01)
00:01.0 PCI bridge: ATI Technologies Inc RS480 PCI Bridge
00:04.0 PCI bridge: ATI Technologies Inc RS480 PCI Bridge
00:05.0 PCI bridge: ATI Technologies Inc RS480 PCI Bridge
00:06.0 PCI bridge: ATI Technologies Inc RS480 PCI Bridge
00:07.0 PCI bridge: ATI Technologies Inc RS480 PCI Bridge
00:12.0 SATA controller: ATI Technologies Inc SB600 Non-Raid-5 SATA
00:13.0 USB Controller: ATI Technologies Inc SB600 USB (OHCI0)
00:13.1 USB Controller: ATI Technologies Inc SB600 USB (OHCI1)
00:13.2 USB Controller: ATI Technologies Inc SB600 USB (OHCI2)
00:13.3 USB Controller: ATI Technologies Inc SB600 USB (OHCI3)
00:13.4 USB Controller: ATI Technologies Inc SB600 USB (OHCI4)
00:13.5 USB Controller: ATI Technologies Inc SB600 USB Controller (EHCI)
00:14.0 SMBus: ATI Technologies Inc SBx00 SMBus Controller (rev 13)
00:14.1 IDE interface: ATI Technologies Inc SB600 IDE
00:14.2 Audio device: ATI Technologies Inc SBx00 Azalia (Intel HDA)
00:14.3 ISA bridge: ATI Technologies Inc SB600 PCI to LPC Bridge
00:14.4 PCI bridge: ATI Technologies Inc SBx00 PCI to PCI Bridge
01:05.0 VGA compatible controller: ATI Technologies Inc RC410 [Radeon Xpress 200M]
02:00.0 Ethernet controller: Atheros Communications Inc. AR5001 Wireless Network Adapter (rev 01)
06:00.0 Ethernet controller: Atheros Communications L2 100 Mbit Ethernet Adapter (rev a0)

My main issue is the same as the original bug, usually occurs as soon as a window is opened, and Ctrl-Alt-F1 does nothing.
Tristan

Here's the proper bug report with a modified patch (should replace the freedesktop-bug #25662): https://bugzilla.kernel.org/show_bug.cgi?id=15626

mistake in the bug text: (should replace the freedesktop-bugs #25662)

Other KMS bugs get linked to the beta 2 milestone,why this bug isn't?

Booting with pci=nomsi KMS works, but sometimes graphics is very very slow. It is random:( (same as @Karesz?)

@Petr: That's the whole point: this issue comes and goes at random. Booting with 'pci=nomsi' has failed several times for me too, booting with 'nomodeset' works nearly always and booting without extra options (as it should be) works every now and then.
You are right that the performance of graphics is sometimes very much degraded.

For the moment I can't test this very well because my graphical login is broken (bug ), but when I try a normal boot, I'm still stuck with 'nomodeset' 9 out of 10 times.
Status should still be 'confirmed', maybe 'triaged' (any devs out there who can comment??)

cheers
Tom

I think we are moving in circles, it leads nowhere:( bugzillas from freedesktop, ubuntu, fedora, still fulfilling the same bug every new release... it looks to very complicated bug, race condition...

Should we contact developers directly? I am from Czech Republic and I would take my notebook and go to every part of my country to any developer who is able to solve this bug. And buy food&drink for him!

Sorry for this OT but I am desperate. I will provide any logs which are needed.

"Booting with pci=nomsi KMS works, but sometimes graphics is very very slow."

You mean 3D graphics? I only tested KMS with in 2D and the performance was good.

I only tested KMS in 2D and the performance was good.

A developer should change the bug report because the problem isn't in the xserver-xorg-driver-ati package.

Well, I've tested 2D, and it was anything but good. Also tried karmic with
pci=nomsi with the 2.6.32 kernel, also pretty slow, both 2 and 3D. :(

2010/3/26 Lollerke <email address hidden>

> A developer should change the bug report because the problem isn't in
> the xserver-xorg-driver-ati package.
>
> --
> [Lucid] ATI KMS causes DRM error on X200M
> https://bugs.launchpad.net/bugs/509273
> You received this bug notification because you are a direct subscriber
> of the bug.
>

But the pci=nomsi seem to work for me, all of the time. (had about 10-15
boots)

2010/3/26 Lengyel Karesz <email address hidden>

> Well, I've tested 2D, and it was anything but good. Also tried karmic with
> pci=nomsi with the 2.6.32 kernel, also pretty slow, both 2 and 3D. :(
>
> 2010/3/26 Lollerke <email address hidden>
>
> A developer should change the bug report because the problem isn't in
>> the xserver-xorg-driver-ati package.
>>
>> --
>> [Lucid] ATI KMS causes DRM error on X200M
>> https://bugs.launchpad.net/bugs/509273
>> You received this bug notification because you are a direct subscriber
>> of the bug.
>>
>
>

@Lollerke: (sorry for vague info), yes 3D graphics, and with desktop-effects on (of course) whole desktop...

Petr: That needs another bug report (like: KMS causes major 3D performance regression on X200M). I have desktop effects turned off on my system and it was totally usable and fast with KMS.

Ok I just had crash after my system had been on for a couple of hours. (first one since adding pci=nomsi)
Below is the kernel trace:

I notice radeon and drm in the output, is this related to this bug or should I open a new one?

Mar 26 21:20:02 danny-laptop kernel: [15086.551703] BUG: unable to handle kernel paging request at 00001187
Mar 26 21:20:02 danny-laptop kernel: [15086.551716] IP: [<f822bc68>] drm_mm_search_free_in_range+0x28/0xe0 [drm]
Mar 26 21:20:02 danny-laptop kernel: [15086.551749] *pde = 760b5067
Mar 26 21:20:02 danny-laptop kernel: [15086.551753] Oops: 0000 [#1] SMP
Mar 26 21:20:02 danny-laptop kernel: [15086.551758] last sysfs file: /sys/devices/LNXSYSTM:00/LNXSYBUS:00/PNP0C0A:00/power_supply/BAT1/charge_full
Mar 26 21:20:02 danny-laptop kernel: [15086.551763] Modules linked in: wlan_tkip binfmt_misc ppdev snd_hda_codec_realtek fbcon tileblit font bitblit snd_hda_intel softcursor snd_hda_codec snd_hwdep wlan_scan_sta snd_pcm_oss snd_mixer_oss snd_pcm ath_rate_sample snd_seq_dummy snd_seq_oss snd_seq_midi snd_rawmidi pcmcia joydev ath_pci snd_seq_midi_event radeon ttm drm_kms_helper snd_seq drm i2c_algo_bit yenta_socket snd_timer snd_seq_device snd rsrc_nonstatic soundcore shpchp pcmcia_core snd_page_alloc ati_agp i2c_piix4 agpgart video wlan ath_hal(P) psmouse output serio_raw lp parport usbhid 8139too hid 8139cp mii pata_atiixp sata_sil
Mar 26 21:20:02 danny-laptop kernel: [15086.551828]
Mar 26 21:20:02 danny-laptop kernel: [15086.551833] Pid: 919, comm: Xorg Tainted: P (2.6.32-17-generic #26-Ubuntu) Satellite L30
Mar 26 21:20:02 danny-laptop kernel: [15086.551837] EIP: 0060:[<f822bc68>] EFLAGS: 00213293 CPU: 1
Mar 26 21:20:02 danny-laptop kernel: [15086.551849] EIP is at drm_mm_search_free_in_range+0x28/0xe0 [drm]
Mar 26 21:20:02 danny-laptop kernel: [15086.551852] EAX: 00000001 EBX: 00000001 ECX: 00001187 EDX: 00000000
Mar 26 21:20:02 danny-laptop kernel: [15086.551855] ESI: 00001187 EDI: 00000001 EBP: e6a79ca8 ESP: e6a79c80
Mar 26 21:20:02 danny-laptop kernel: [15086.551859] DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
Mar 26 21:20:02 danny-laptop kernel: [15086.551862] Process Xorg (pid: 919, ti=e6a78000 task=e4cf19a0 task.ti=e6a78000)
Mar 26 21:20:02 danny-laptop kernel: [15086.551865] Stack:
Mar 26 21:20:02 danny-laptop kernel: [15086.551867] db490a80 00000000 00000000 e3b58500 e6b5c42c 00000001 00000000 e6a79d58
Mar 26 21:20:02 danny-laptop kernel: [15086.551876] <0> f6809d14 e6a79d28 e6a79cdc f8387579 00000000 00008000 00000001 00285906
Mar 26 21:20:02 danny-laptop kernel: [15086.551885] <0> e6b5c42c 00000000 00008000 f6a9324c 00000002 00000001 00000000 e6a79d38
Mar 26 21:20:02 danny-laptop kernel: [15086.551895] Call Trace:
Mar 26 21:20:02 danny-laptop kernel: [15086.551910] [<f8387579>] ? ttm_bo_man_get_node+0x79/0xd0 [ttm]
Mar 26 21:20:02 danny-laptop kernel: [15086.551917] [<f83877c4>] ? ttm_bo_mem_space+0x1f4/0x4d0 [ttm]
Mar 26 21:20:02 danny-laptop kernel: [15086.551925] [<f83865cb>] ? ttm_bo_unmap_virtual+0x4b/0x60 [ttm]
Mar 26 21:20:02 danny-laptop kernel: [15086.551933] [<f838832e>] ? ttm_bo_move_buffer+0xae/0x120 [ttm]
Mar 26 21:20:02 danny-laptop kernel: [15086.551941] [<f8388436>] ? ...

I think its a new one. You should report it with apport.

My laptop is running fine with pci=nomsi with no errors.

danwood76: What type of Toshiba L30 do you have? Mine is an L30-134.

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

=== system now useless ===

Loaded a lot of updates today: a new kernel, gdm, gnome-session-manager and others.
- Booting without 'nomodeset' does not work because KMS is broken, no change here (bug 509273)
- Have to boot into recovery mode and then resume normal boot because gdm is broken (bug 516520)
- Could not use 'resume normal boot' anymore because Plymouth seems broken (no bugreport yet but removing Plymouth solved this. will fime one in a minute)
- Can not get into my GUI anymore by manually starting X, because now X/Gnome is broken (no bugreport yet but will do so in a minute)

My Ubuntu installation is now effectively FOOBAR, so I'm back to booting Windows XP..... hip hip hooray! (not)
I must say that Alpha was a lot more stable than beta is now :-((

gr
Tom

Tom: and if you boot with pci=nomsi? Maybe it works this time for you.

Tom: can you modify the bug report? xserver-xorg-driver-ati is not affected,it should be changed to linux (I changed the bug watch already to the Linux Bugzilla Tracker)

hi Lollerke

Sorry, I was a bit frustrated when I posted #150.
@151: no, didn't work. I tried both the normal boot and the recovery mode twice. 'pci=nomsi' gives me the vertical lines, just like a boot without any extra options. 'nomodeset' works (but other bugs still bite me after that).
@152: OK, done that. Though I still feel the problem lies between the kernel and the ati driver, I will follow you in this.

Any other things I can test / information I can collect?

cheers
Tom

I uploaded the proper patch and patch fix by Alex Deucher.

Tom: Can you test 'pci=nomsi' with todays lucid daily live CD? Because it's working for everybody except for you. What kind of X200 do you have,Intel or AMD version?

Will this be resolved for beta2 or we have to wait for the final release? I provided all the information for this bug and the patches are working.

Will try a daily live cd asap (like tomorrow) using the amd64 build, but
so far only nomodeset has worked for me so I'm sceptical.

ATI or Intel? No idea really, but my laptop has an AMD cpu and chipset so
I'd think ATI.
Anyway; this is info on the card as reported by Winblows XP. Will try to
poat something like output from lspci tomorrow.

Will report back soon.

Tom

Back in my desktop again... (forced a reinstall of
gnome-desktop-environment with all dependencies)
Attached the output of lspci., hope that gives more information.

gr
Tom

You have the AMD version: ATI Technologies Inc RS482 [Radeon Xpress 200M]
I have the Intel version: ATI Technologies Inc RC410 [Radeon Xpress 200M]

Have you tried the daily live yet?

Just downloaded and tried a daily live CD, booted from it and chose 'Try Ubuntu without installing'. Here are the results:
- straight boot (no extra options): black screen, vertical coloured lines, hangs
- with 'nomodeset': black screen, vertical coloured lines, hangs
- with 'pci=nomsi': black screen, vertical coloured lines, hangs

There were loads of updates over the last 24hrs, including several to gnome/gdm components and to Xserver-core and the ATI and Radeon driver. Before those I could at least get into my system via 'recovery mode (with nomodeset of course) > resume boot > login > startx'. That is now broken. Are you still telling me that Xserver and ATI / Radeon driver are not involved? Kernel didn't change since I last used the desktop fine yesterday (though getting there through the backdoor)
As of today I can not use that workaround anymore, neither a 'normal' boot (gdm busted, won't let me login), neither the live CD.

Funny, I've been running this laptop with XP (works), Jaunty (worked great) and Karmic (worked). Since last week Lucid has become one big regression. KMS, devicekit, gdm, plymouth all cause headaches. Not a good start for a LTS release!

I will try the alternative CD to perform a complete reinstall, and see if that fixes anything. If not, I'm going to look for another OS. Something reliable. I expect many newcomers to do the same: "does not work out of the box".

Tom

Tom: You can try the i386 version if the amd64 is not working for you,because there I see amd64 specific KMS bugs out there.

--there

I used to test Lucid since Aplha3 with "nomodeset", and now since the last update yesterday (31-03-2010), I'm not that able to have the desktop working. Yet with or without "nomodeset" the desktop keeps freezing every time I boot. on the update list, there were xserver-ati-radeon and other xserver-ati ... I think this bug is related with Xserver. Now I'm on Karmic.
I'll download the today's daily build and make a fresh install of this Lucid Beta1 and see what comes.

Ngassam Nkwenga: Try "pci=nomsi" instead of "nomodeset".

ok, I have booted with "pci=nomsi" instead of "nomodeset" and it seems to be working since I have no freeze.
was not the patch suppose to save this issue? cause since last update, I have not try to boot without any options

The patch does the same as the "pci=nomsi" command.

then I think I have something wrong cause I have all update and I can
not test Lucid without boot options(nomodeset or pci=nomsi), the
system keep freezing, but with pci=nomsi I can test the beta again.

do I need to add the patch manually ? if yes, then how could I make it ?
thanks

2010/4/1 Lollerke <email address hidden>:
> The patch does the same as the "pci=nomsi" command.
>
> --
> [Lucid] ATI KMS causes DRM error on X200M
> https://bugs.launchpad.net/bugs/509273
> You received this bug notification because you are a direct subscriber
> of the bug.
>

Ngassam:
Are you booting a liveCD/live usb stick or from the hard drive?
If you're booting a removable media, then try to press F2 when you see the
two little icons on the lower part of the screen, there you'll be able to
edit the boot options!

2010/4/1 Ngassam Nkwenga <email address hidden>

> then I think I have something wrong cause I have all update and I can
> not test Lucid without boot options(nomodeset or pci=nomsi), the
> system keep freezing, but with pci=nomsi I can test the beta again.
>
> do I need to add the patch manually ? if yes, then how could I make it ?
> thanks
>
>
> 2010/4/1 Lollerke <email address hidden>:
> > The patch does the same as the "pci=nomsi" command.
> >
> > --
> > [Lucid] ATI KMS causes DRM error on X200M
> > https://bugs.launchpad.net/bugs/509273
> > You received this bug notification because you are a direct subscriber
> > of the bug.
> >
>
> --
> [Lucid] ATI KMS causes DRM error on X200M
> https://bugs.launchpad.net/bugs/509273
> You received this bug notification because you are a direct subscriber
> of the bug.
>

Just tried the i386 daily (as downloaded and burned yesterday), booted from it, hit F2 and F6 to modify the boot options. Then chose 'Try Ubuntu without installing'. Here are the results:
- straight boot (no extra options): black screen, vertical coloured lines, hangs
- with 'nomodeset': black screen, vertical coloured lines, hangs
- with 'pci=nomsi': black screen, vertical coloured lines, hangs

So no change at all compared with the amd64 edition.

sigh..
Tom

Tom: I'm out of ideas..

@Karesz:
No, I've installed it on an usb drive.

Le jeudi 01 avril 2010 à 13:44 +0000, Karesz a écrit :
> Ngassam:
> Are you booting a liveCD/live usb stick or from the hard drive?
> If you're booting a removable media, then try to press F2 when you see the
> two little icons on the lower part of the screen, there you'll be able to
> edit the boot options!
>
> 2010/4/1 Ngassam Nkwenga <email address hidden>
>
> > then I think I have something wrong cause I have all update and I can
> > not test Lucid without boot options(nomodeset or pci=nomsi), the
> > system keep freezing, but with pci=nomsi I can test the beta again.
> >
> > do I need to add the patch manually ? if yes, then how could I make it ?
> > thanks
> >
> >
> > 2010/4/1 Lollerke <email address hidden>:
> > > The patch does the same as the "pci=nomsi" command.
> > >
> > > --
> > > [Lucid] ATI KMS causes DRM error on X200M
> > > https://bugs.launchpad.net/bugs/509273
> > > You received this bug notification because you are a direct subscriber
> > > of the bug.
> > >
> >
> > --
> > [Lucid] ATI KMS causes DRM error on X200M
> > https://bugs.launchpad.net/bugs/509273
> > You received this bug notification because you are a direct subscriber
> > of the bug.
> >
>

I've tried to boot the i386 version from a USB drive (daily build of Apr 1st) and I still have the
same problem. I reach the screen where the options "try without install" or "install" should
be available, but I get only the background with an empty grey square.

@Lollerke:

Where did you get 0001-PCI-quirks-disable-msi-on-AMD-rs4xx-internal-gfx-br.patch?

@Tom Louwrier:

Since you seem to be having a different issue compared to others in this bug, I suggest opening a new bug for it. It's easier to track issues when we work on only one issue per bug report.

Thanks

Hi guys,

Had to do without you for nearly 48 hours and was getting withdrawal symptoms ;-)

@ Chase: OK, I will have to think about how to write that bug report without repeating all my posts here.
@ Lollerke: I'm out of ideas too, but the good thing about errors that come and go is that tonight it went.
Somehow my system got past the boot issues and let me boot into recovery mode (with 'nomodeset'). I got to load like 174 updates, including a new kernel, Xcore, ATI and Radeon and quite some Gnome elements. After that a normal boot and logging on without issues.
In short: I'm back (for now).

cheers
Tom

Chase Douglas: I've got the patches from here: https://bugzilla.kernel.org/show_bug.cgi?id=15626
It's actually one patch,because the bigger is a patch and the smaller is a fix for the patch.

Hi, with the new update, I still need to add "pci=nomsi" on grub to have
no issue on the desktop when Im in.
remember: I have an amilo Li1718 with an ATI X200M serie card. without
any options at boot, I have to in console mode to avoid any freeze, once
I change back to GUI mode (on the same boot) the freeze come back again.

Ngassam Nkwenga: You need to add "pci=nomsi" until the bug status changed to fix released (its just confirmed yet).

If you find that booting with 'pci=nomsi' fixes the issue, please can a few people post the output of exactly 'sudo lspci -vvd1002:'?
This will show us the Message Signal Interrupt address and data fields of AMD PCI devices.

Hi Daniel j.
hier is an output for your request:

 Daniel J Blueman a écrit :
> If you find that booting with 'pci=nomsi' fixes the issue, please can a few people post the output of exactly 'sudo lspci -vvd1002:'?
> This will show us the Message Signal Interrupt address and data fields of AMD PCI devices.
>

Thanks Ngassam! I forgot to ask if you can boot without 'pci=nomsi' to capture this information, as MSI is disabled [1]. If need be, after X is started (eg at the login screen), login at a virtual terminal (eg alt-F1), and capture the information to a file, then reboot to add to this report.

--- [1]
...
Capabilities: [70] Message Signalled Interrupts: Mask- 64bit- Queue=0/0 Enable-
  Address: 00000000 Data: 0000
...

@Daniel:
hier is the log without boot options:
if you need something just tell me

Le dimanche 04 avril 2010 à 15:04 +0000, Daniel J Blueman a écrit :
> Thanks Ngassam! I forgot to ask if you can boot without 'pci=nomsi' to
> capture this information, as MSI is disabled [1]. If need be, after X is
> started (eg at the login screen), login at a virtual terminal (eg
> alt-F1), and capture the information to a file, then reboot to add to
> this report.
>
> --- [1]
> ...
> Capabilities: [70] Message Signalled Interrupts: Mask- 64bit- Queue=0/0 Enable-
> Address: 00000000 Data: 0000
> ...
>

Thanks Ngassam. Decoding the MSI address field, the interrupt will occur on the lowest priority APICs with logical ID 3 (usually core 3 in non-large systems) [1].

Can you see if rewriting the address with a simpler one prevents the crashing? Boot without 'pci=nomsi' and run:

$ sudo setpci -s 01:05.0 84.l=fee00000

(note, this won't work on other systems)

If this prevents the crashing, it's possible that I could cook up a MSI-quirk patch to write the simpler address vector only. If it fails, we need to not enable MSI for this board via adding a no-MSI quirk for this PCI device.

Thanks, Daniel

--- [1]

01:05.0 VGA compatible controller: ATI Technologies Inc RC410 [Radeon Xpress 200M]
...
 Capabilities: [80] Message Signalled Interrupts: Mask- 64bit- Queue=0/0 Enable+
  Address: fee0300c Data: 4181

I can provide the lspci output if you can provide the method for saving it to a log file? (yes I should now really...) :)

@Daniel , I tried with your suggestion but it keep on freezing, so i'm
stick with the "pci=nomsi" option.

@Mark : just do it this way:
sudo lspci -vvd1002: > filename.extension

in my case I choose "lspci" as filename and "log" as extension. the ">"
is the redirection command. the result is to be found in the current
directory( default is your /home)
ie. sudo lspci -vvd1002: > lspci.log

Le dimanche 04 avril 2010 à 18:59 +0000, Mark a écrit :
> I can provide the lspci output if you can provide the method for saving
> it to a log file? (yes I should now really...) :)
>

Ngassam, thanks for the tip.

Here's mine. Rebooted and removed pci=nomsi from boot options at the grub screen, then grabbed it.

I've uploaded a test kernel to http://people.canonical.com/~cndougla/509273. It includes the patch Lollerke attached, which disables pci msi interrupts for specific systems. Anyone who had their issues fixed with the pci=nomsi boot option should try this kernel to see if it works without the boot option. If so, we will look into including the patch in the Lucid kernel.

Thanks

Chase Douglas: It works

Checking the upstream bug [https://bugzilla.kernel.org/show_bug.cgi?id=15626], Alex's patch (which has been accepted) disables MSI for the whole system probably for good reason, not just for that device or the child buses - thus what I was going to suggest [1] probably won't work.

--- [1]

diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index 81d19d5..603cd97 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -2215,6 +2215,16 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_NVIDIA,
    PCI_DEVICE_ID_NVIDIA_NVENET_15,
    nvenet_msi_disable);

+/* The Radeon Xpress 200M/RS480 has problems with delivering MSI interrupts
+ * correctly, causing crashing; disable MSI for this device.
+ */
+static void __devinit radeon_msi_disable(struct pci_dev *dev)
+{
+ dev_info(&dev->dev, "Disabling MSI for ATI Radeon\n");
+ dev->no_msi = 1;
+}
+DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_ATI, 0x5a3f, radeon_msi_disable);
+
 static int __devinit ht_check_msi_mapping(struct pci_dev *dev)
 {
  int pos, ttl = 48;

Chase Douglas: it works for me too.
I tried it yesterday, and this morning still working.

 Chase Douglas a écrit :
> I've uploaded a test kernel to
> http://people.canonical.com/~cndougla/509273. It includes the patch
> Lollerke attached, which disables pci msi interrupts for specific
> systems. Anyone who had their issues fixed with the pci=nomsi boot
> option should try this kernel to see if it works without the boot
> option. If so, we will look into including the patch in the Lucid
> kernel.
>
> Thanks
>

@Chase Douglas:

The patch works for me too! , great , i've disable the pci=nomsi option, i've downloaded and installed the patch , and for now it seems to work perfectly!

Noo freeze , no crash , great! , i hope this patch will be insert permanently in the ubuntu kernel

Cloudany: and the 3D acceleration is working?

@Petr: If a give in the terminal " glxinfo | grep "direct rendering" " , the answer is YES

             If i activate compiz effect: Expose' works but is not so fluid in the movements but is ok it works!

So 3D acceleration work perfectly!

(Where's my prize for helping you? ahah i'm joking of course , i'm glad to make ubuntu better than ever)

Today when i've booted Lucid Lynx it gives me a problem , the 3d acceleration is deactivated, if i give glxinfo | grep "direct rendering, it say nothing.

Sorry for the double post , but i've noticed that ubuntu have installed the ati catalyst control center, uninstalling that everything is going fine. I beg your pardon.

Same problem in ubuntu 10.04 beta 2 , is that normal?

@Cloudany:

Yes, the beta 2 freeze was in place one week ago. No changes were allowed until the release of beta 2 unless they were serious (like the installer dying half-way through). Now that it has been released, the freeze is lifted and we can add in more bug fixes that are less serious. The last freeze will be one week from today. After that, only the most serious bugs will be addressed before release of the final image.

Of course, we'll still be working on bug fixes, but after next week they will be deferred to stable release updates and will not make it onto the release CDs.

This patch has been sent to the kernel-team mailing list and will likely be committed to the lucid kernel shortly. When that occurs, the status of this bug will be updated to 'Fix Committed'. As long as that occurs before next thursday, you can rest assured the fix will be in the release CDs.

Thank you a lot, your words makes me happy (really) , because i use ubuntu for work and i need ubuntu 10.04 on my Pc, thank you a lot , so i will wait for the 29th April, thank you a lot for your pacience to answer my questions.

I have some days no problems with KMS, ATI and 2.6.32 or 2.6.34 (from mainline).
But today I recognise one problem with my notebook. If I start the notebook unplugged from power chord I couldn't start up to gnome. The screen freeze before the passward line and also starting xorg from rescue console results in a frozen screen.

The system: ubuntu lucid, HP 6735b running 64bit ubuntu.

What information can help?

Erni35

@Erni35:

Your new issue is likely unrelated to this bug. Please open a new bug so we can track it more easily. The best way is to run 'ubuntu-bug linux'.

Thanks

Erni35: I have the same problem,very strange.

This bug was fixed in the package linux - 2.6.32-20.29

---------------
linux (2.6.32-20.29) lucid; urgency=low

  [ Andy Whitcroft ]

  * Revert "SAUCE: Use MODULE_IMPORT macro to tie intel_agp to i915"
    - LP: #542251
  * add Breaks: against hardy lvm2
    - LP: #528155

  [ Colin Watson ]

  * d-i -- enable udebs for generic-pae
    - LP: #160366

  [ Stefan Bader ]

  * [Config] Add xen netboot support
    - LP: #160366

  [ Takashi Iwai ]

  * (pre-stable): input: Support Clickpad devices in ClickZone mode
    - LP: #516329

  [ Upstream Kernel Changes ]

  * Revert "(pre-stable) Bluetooth: Fix sleeping function in RFCOMM within
    invalid context"
    - LP: #553837
  * Revert "(pre-stable) USB: fix usbfs regression"
    - LP: #553837
  * Revert "(pre-stable) softlockup: Stop spurious softlockup messages due
    to overflow"
    - LP: #553837
  * Revert "(pre-stable) drm/nouveau: report unknown connector state if lid
    closed"
    - LP: #553837
  * drivers/scsi/ses.c: eliminate double free
    - LP: #553837
  * decompress: fix new decompressor for PIC
    - LP: #553837
  * ARM: Fix decompressor's kernel size estimation for ROM=y
    - LP: #553837
  * MIPS: Cleanup forgotten label_module_alloc in tlbex.c
    - LP: #553837
  * tg3: Fix tg3_poll_controller() passing wrong pointer to tg3_interrupt()
    - LP: #553837
  * tg3: Fix 5906 transmit hangs
    - LP: #553837
  * ALSA: hda - Fix input source elements of secondary ADCs on Realtek
    - LP: #553837
  * ALSA: hda: enable MSI for Gateway M-6866
    - LP: #538918, #553837
  * timekeeping: Prevent oops when GENERIC_TIME=n
    - LP: #553837
  * Input: alps - add support for the touchpad on Toshiba Tecra A11-11L
    - LP: #553837
  * Input: i8042 - add ALDI/MEDION netbook E1222 to qurik reset table
    - LP: #553837
  * i2c-i801: Don't use the block buffer for I2C block writes
    - LP: #553837
  * ath5k: dont use external sleep clock in AP mode
    - LP: #553837
  * ath5k: fix setup for CAB queue
    - LP: #553837
  * ring-buffer: Move disabled check into preempt disable section
    - LP: #553837
  * function-graph: Init curr_ret_stack with ret_stack
    - LP: #553837
  * Bluetooth: Fix sleeping function in RFCOMM within invalid context
    - LP: #553837
  * tracing: Use same local variable when resetting the ring buffer
    - LP: #553837
  * tracing: Disable buffer switching when starting or stopping trace
    - LP: #553837
  * tracing: Do not record user stack trace from NMI context
    - LP: #553837
  * PCI: unconditionally clear AER uncorr status register during cleanup
    - LP: #553837
  * efifb: fix framebuffer handoff
    - LP: #553837
  * coredump: suppress uid comparison test if core output files are pipes
    - LP: #553837
  * V4L/DVB (13961): em28xx-dvb: fix memleak in dvb_fini()
    - LP: #553837
  * hrtimer: Tune hrtimer_interrupt hang logic
    - LP: #553837
  * x86, apic: Don't use logical-flat mode when CPU hotplug may exceed 8
    CPUs
    - LP: #553837
  * mvsas: add support for Adaptec ASC-1045/1405 SAS/SATA HBA
    - LP: #553837
  * pci: add support for 82576NS serdes to existing SR-IOV quirk
    - LP: #553837
  * sched: Mark boot-cpu active before smp_init()
    -...

Has this fix been incorporated in the daily build from 10/04/2010? I'm still getting similar error messages when booting from a USB made from today's build. Needless to say I can't boot from it.

Screen shot of my boot attempt

I also got a frozen screnn on boot up if I started disconnected from power chord. The screen is dimmed at start up and I never reach the log in screen of gdm. Plugged in I have no problems but I couldn't go to a console for logs.

Erni35

BTW, I also tried the "pci=nomsi" option but all it does in my case is turn the monitor off or, to be more precise, put it into power save mode and the only way out is a hard reboot.

@Sennaista and @Erni35:

Both of you have separate bugs from this one. Please open new bugs for these issues.

Thanks

FYI:
"pci=nomsi" solved my problem with Ubuntu 10.04 Lucid Lynx on notebook Asus A9RP (ATI Technologies Inc RC410 [Radeon Xpress 200M])

Thank you very much.

Linuxfan: You don't need "pci=nomsi" with the 2.6.32-20 kernel because its fixed.

I still see this issue on my acer travelmate 8000 with it's ati radeon mobility 9700 (rv350) using the -20 kernel.

When I boot with kms enabled, I get to see plymouth nicely, but as soon as kdm starts, the screen is corrupted. swicthing back to the console shows the same errors printed out continously as in #61 https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-ati/+bug/507148/comments/61

OK well, I have not idea what else was changed in the latest Kernel, but when I ran system update this morning and installed it, on reboot I get no graphical "Ubuntu", just a stalled wall of console text about EISA this and isapnp that . This is with "pci=nomsi" both enabled and disabled (to eliminate it). Powering off and using the previous .19 Kernel (with "pci=nomsi" enabled), and all is well.

Just grabbed the latest, 12th, Daily Live CD and get the same result as above, a wall of console text and no boot.

Anyone else using an Amilo LI1718 confirm this?

@Yves, @Mark:

Both of your issues are separate from this bug. Please open new bugs for them.

Thanks

I did so:

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/561789

Maybe someone with the right karma can make sure the bug get's tagged to be part of lucid.