[fglrx]Computer freeze when switching user

Can you reproduce the freeze?
Do you have compiz enable? If yes, can you reproduce it without compiz?
Can you look into kern.log to see if there is some important information such as segfaut.

I mark this report until we have more information.

I am having this issue with compis enabled as well as with compiz disabled.

In my kern.log, there's absolutely nothing about the freeze.

I have ATI Radeon HD2400 PRO AGP, with fglrx (8.54.3) drivers installed.

Now, with Intrepid final, I am sure that this happens only with fglrx drivers installed.
Without them, I could normally switch user.

I can confirm this bug on Intrepid final. I have a Radeon HD3400 with the Intrepid fglrx drivers installed. When changing to guest session, screen goes black and system hangs. Happens with or without Compiz enabled.

Hmm. I am having same bug, and also some confusion:

1. I am running fglrx on Intrepid release 8.10 on an amd64 SMP box.

2. Switching to "guest" or "guest session" produces the black screen and system hang/crash
    described by Harm Jans.

3. I do not see how to switch users any more - shouldn't there be a "switch users"
    line down around the restart/shutdown/hiberate/suspend/logout" lines?

4. I am not running compiz, but am running dual-head monitors.

5. The failure is 100% reproducible.

I don't think, this bug is a duplicate of bug #266911.
It does nothing with the fglrx issue and the fix, described in the bug #266911 didn't work for me.

I would also notice, that if I move mouse immediately after clicking to "guest session" (before the system halt), I get back to the previous login - in my case, I have to login the same way as after I abort screensaver (I have automatic screen locking set)

unmarking as duplicate of bug #266911 as this particularly is about an issue with fglrx driver

I disabled the "third-party and proprietary hardware drivers" (System/Administration/Hardware Drivers), but
the system still crashes on. I think the "guest session" is not relevant, inasmuch as the system
crashes instantly if I attempt to "switch users", too.

Please let me know if I can be any use in fault isolation here.

Same here: amd64 HP EliteBook 6930p with ATI Mobility Radeon HD 3400.

This problem comes from FLGRX proprietary driver from AMD/ATI. Please complain to them about this issue...
There's no such thing with opensource driver radeonhd (though there are other issues ...)
According to this post : http://www.phoronix.com/forums/showthread.php?t=13671&highlight=switch+user
[...]
However, this issue has been mentioned repeatedly over the last 6 months (at least), and the ATI devs response was along the lines of "switch user? who uses that?", so don't hold your breath waiting for a fix because they don't really seem to care.
[...]

I can confirm this bug in a fully updated Ubuntu 8.04 system. I am running 64-bit Intel Q6600 and a Radeon HD4670.
It happens on both the latest fglrx 8.12 driver and the radeonhd driver. I am beginning to think this may not be a ATI Radeon issue. Can someone test this with the 'radeon' driver in 8.04?

I should add that I tried on the latest radeonhd driver fetched from GIT on 12/23/08, and still experience this bug!

Please ignore my last two comments. I realized that radeonhd was not enabled after all. After enabling the 'radeonhd' driver, the problem went away. So the problem is truly with the fglrx driver. Sorry for the confusion.

I'm having troubles with the same bug i think.

I was trying this command:
xinit xterm -- :1 vt8

Shouldn't this simply open an xterm applet (without window manager) on display 1, virtual terminal 8 ?

It just makes my screen go black and my keyboard unresponsive :-(

Is this related ?

abatcher: Yes, it is probably related to this bug, if you use the fglrx driver. On my machine, the issue is reproducable as well.

 - me too - (with an Ati Radeon HD 2400 pro)

I'm affected by this Problem since last Summers Intrepid Update. (switched from x86 to AMD64)
First with Catalyst 8.10 prerelease then Catalyst 8.10 stable.

I tested 8.12 (stable), 9.1 beta and 9.2 ces09-preview. To no avail (except: 9.2 has additional problems, but its not even beta so its ok)

Like Anakin Starkiller wrote, presumably this won't get fixed in the near future :-(
(but using xvideo output with compiz is fine now :-) (Nevertheless I would prefere a working second X server))

I even switched from a 2-head configuration to one single Monitor. No luck.

(btw: I'm using this feature not only to switch users. Very often I just start a game at another X Server to be able to switch between the full screen game and the rest of my desktop without closing the game. Thats also impossible for now. Like abacher wrote, it's not even possible to start a second X screen from a text tty. (btw gaming... thats the reason I prefere fglrx for now...))

I am using an AMD64, dual-head, with fglrx. It dies on switch users, of course.

From what hede says, dual-head is not the problem, and it just seems to
be fglrx itself. [I do not understand the implications of abatcher's

        xinit xterm -- :1 vt8

comment in this regard.

Perhaps we can convince the MyLightWorksFine people at AMD that there are virtues to
having multiple users signed on at once. It doesn't look like we'll get it fixed, otherwise.

1. I use multiple users for security reasons - I have development accounts that have access
    to source code, and I try to keep those accounts distinct from my generic email/web account, etc.
    Part of this has to do with keeping my data out of client areas, and vice versa. It's not
    security in the secrecy sense, but in the safety of different accounts when my fumblefingers err,
    just as we avoid running as root.

2. Hede uses multiple users to get full-screen access and access to normal desktop without
    logging in/out. Very convenient.

I would appreciate it if the AMD people responsible for fglrx can give us some statement as
to why these are not virtuous activities. If they cannot do so, then can they please give us
some estimate of a repair date for their broken software?

As far as I understand, fast-user-switch uses startx/xinit in the background to start a new X session on another display ? Or is this wrong and my comment irrelevant ?

If I'm on the right track it means the problem is not only with fast-user-switching, but with the underlying command xinit ?

If this is a fglrx problem, other distro's should experience the same thing. Searching in the fedora forums, all I could find was a small thread: http://fedoraforum.org/forum/showthread.php?p=1148421 But they are talking about compiz being the problem. This is not related, because we have the problem too with compiz disabled ?

I wont call it "broken". It's more like a car not having a trailer hitch.

If you really need it, you can have it by either buy a new car (nvidia) or add one to your car (you cannot do this with a ferrari i.e. you'll lose 3D with high-end-cards using xorgs radeon or radeonhd driver).

And at least most people simply don't need it...

Funny analogy. I feel it's more like having a trailer hitch (the command is available), but using it breaks your car (computer freezes).

If there is really no fix (in the near future), maybe the fast-user-switch should be disabled when fglrx is enabled ?

I am running w/compiz disabled, as far as I can tell.

This is a duplicate of Bug#325133, which includes all the necessary information generated by the ubuntu-bug command line tool.

I have 2 user on the system, if I have Visual effects on I don't see the bug.
But if I have visual Effects set to normal I get a white screen when switching between users.
I'm using ubuntu jaunty 9.04 latest updates with nvidia drivers 180.35 that come with jaunty.

I have 2 user on the system, if I have Visual effects off I don't see the bug.

But if I have visual Effects set to normal, I get a white screen when switching between users.
I'm using ubuntu jaunty 9.04 latest updates with nvidia drivers 180.35 that come with jaunty.

This is about ati/fglrx, not nvidia.

Lafa - you just need to enable "Unredirect fullscreen windows" in ccsm and your problem will go away.

Same here, when i choose "Guest session" the screen goes black and computer is not responding.

Computer: HP elitebook 8530p
Kernel: 2.6.27-11-generic #1 SMP Thu Jan 29 19:24:39 UTC 2009 i686 GNU/Linux
VGA: 01:00.0 VGA compatible controller: ATI Technologies Inc Mobility Radeon HD 3650
Driver: fglrx
Compiz: Disabled

As said earlier the system also hangs after
xinit xterm -- :1 vt8

ctrl+alt+F8 works OK.

If I disable proprietary driver it all works fine.
Come on AMD! Where are your GNU/Linux support?

Bug still presents in ubuntu 9.04 with ati-9.4 drivers
With open drivers fast user switching works fine.

My card is HD2600 PRO

Bug stills present in Jaunty with both Proprietary and lastest driver from ATI. When I removed ATI driver this function works fine.

I'm using 2600XT.

I can confirm this bug with ATI HD 2400 PRO and K/Ubuntu 9.04 Jaunty Jackalope

I can confirm it as well and I find it as very important for me. We share this PC home and we all are often login and just switch users. I would appreciate any other solution than turning back to windows. If this the aim of ATI, they will probably succeed until I will buy new PC with NVIDIA :-)
I have HD 2600 XT with 9.04 Jaunty amd64 and this problem is with latest ATI drivers as well as with the fglrx driver offered by Ubuntu.
If I could be helpful with this bug, I will.

I too have an ATI Radeon HD 2400 Pro and suffer this problem both on Linux Mint 5 (Hardy-based) and 7 (Jaunty-based). This bug is an absolute killer for me because the rest of my family uses my computer too, and trying to tell them not to use switch user when my account is locked because it will break the system is most likely a nonstarter! Using an open-source driver does fix the problem, but removes 3D support which is marginally more important to me then switch-user support. But why are we forced to choose? Is it too much to ask for ATI to give us drivers that let us use ALL the functions of our operating systems? Come on, sort out fglrx! This should definitely be a high-priority bug, as it's ruinous for anyone who uses switch-user, and unfortunately the open source equivalent doesn't yet utilise the full potential of the card (though I live in hope).

The fact that you can crash your computer with 2 mouse clicks alone should make this a high-priorty bug.

Similar problem when using ATI driver, version 9.5 and Radeon 2600HD. Compiz is enabled.

$ lspci | grep VGA
01:00.0 VGA compatible controller: ATI Technologies Inc RV630 [Radeon HD 2600XT]

Computer blanks out the screen and freezes. I cannot connect via SSH even.

On a related note, can anyone recommend a good nvidia card for Linux with full 3D support? :-)

I've posted a new version of the -fglrx driver to our xorg-edgers PPA,
would you mind testing it either on Jaunty or Karmic and see if it
resolves this bug?

Get fglrx 8.620 here:

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

Thank you for bringing this bug to our attention. Unfortunately a paper cut should be a small usability issue that affects many people and is quick and easy to fix. I'm afraid this bug can't be addressed as part of this project.

This is a hardware specific issue, Not a papercut

A paper cut is a minor usability annoyance that an average user would encounter on his/her first day of using a new installation of Ubuntu 9.10.

For further info about papercuts criteria , pls read > https://wiki.ubuntu.com/PaperCut

Don't worry though, This bug has been marked as "invalid" ONLY in the papercuts project.

For resolution of the bug, kindly identify the projects affected and assign the bug to that project, otherwise the devs of the concerned project might not be notified of this problem.

Bryce:

I can try the new driver over the weekend, but I'm running intrepid. (I'm not intrepid
enough to upgrade until I get some major development work out the door).

If you think that will be (a) of use to you and (b) work, I'll try it.
Please let me know.

Bryce Harrington píše v Pá 26. 06. 2009 v 03:46 +0000:
> I've posted a new version of the -fglrx driver to our xorg-edgers PPA,
> would you mind testing it either on Jaunty or Karmic and see if it
> resolves this bug?
>
> Get fglrx 8.620 here:
>
> https://edge.launchpad.net/~xorg-edgers/+archive/ppa
>
>
> ** Changed in: fglrx-installer (Ubuntu)
> Status: Confirmed => New
>
> ** Changed in: fglrx-installer (Ubuntu)
> Status: New => Incomplete
>

I have Jaunty and eventually I can test the new driver you posted.
However, I am already running the 8.620 fglrx driver from AMD website
with no change regarding to this bug.

--
Lukáš Chmela <email address hidden>
IRC: <email address hidden>, ICQ: 202077459
514D 3C69 8498 E400 0ACE 90F2 00E3 CACA A5BA ECC7

Maybe i can test it with Kubuntu Jaunty 9.04
Just need to know how to let this bug occur/happen with KDE...

Hi All,
I have just installed the drivers, You mentioned.
FYI: The previous version of driver, I have used, was the latest version
from AMD web.
To be sure, here is the list:
fglrx-modaliases_8.620-0ubuntu1~jaunty_amd64.deb
libamdxvba1_8.620-0ubuntu1~jaunty_amd64.deb
xorg-driver-fglrx_8.620-0ubuntu1~jaunty_amd64.deb
fglrx-kernel-source_8.620-0ubuntu1~jaunty_amd64.deb
fglrx-amdcccle_8.620-0ubuntu1~jaunty_amd64.deb

At the first attempt, I was asked to install newer version of libdrm2,
so I used this one
libdrm2_2.4.11+git20090606.3d4bfe8c-0ubuntu0sarvatt_amd64.deb

after that:
sudo aticonfig --initial -f
restarted PC
checked fgl_glxgears and everything worked fine until trying to change
user. After that the PC froze immediately, just like before with any
other fglrx drive in jaunty or interpid.

If you have any other idea, how to solve the problem and be able to
change Win to Ubuntu, I will really appreciate it and try to help you if
my poor knowledge of linux will not be a problem.

Thanks
                Ivo

Bryce Harrington píše v Pá 26. 06. 2009 v 03:46 +0000:

> I've posted a new version of the -fglrx driver to our xorg-edgers PPA,
> would you mind testing it either on Jaunty or Karmic and see if it
> resolves this bug?
>
> Get fglrx 8.620 here:
>
> https://edge.launchpad.net/~xorg-edgers/+archive/ppa
>
>
> ** Changed in: fglrx-installer (Ubuntu)
> Status: Confirmed => New
>
> ** Changed in: fglrx-installer (Ubuntu)
> Status: New => Incomplete
>

Thanks for testing.

On Fri, Jun 26, 2009 at 12:32:54PM -0000, Robert Bernecky wrote:
> Bryce:
>
> I can try the new driver over the weekend, but I'm running intrepid. (I'm not intrepid
> enough to upgrade until I get some major development work out the door).
>
> If you think that will be (a) of use to you and (b) work, I'll try it.
> Please let me know.

Hi Robert, don't worry about it - the testing that others have done
indicate this bug still seems to be a problem with the newer driver.

Bryce Harrington >> are you guys from xorg-edgers from ppa stil working on this issue ?

is anyone working on a way to only use certain features of a proprietary driver, when necessary?
obviously i don't need faulty AMD/ATI hardware accelleration fuctions for a simple user switch..
that would be great.

or perhaps they could go open source on the configuration tool for the proprietary driver!?

i cannot believe, that a professional hardware manufacturer makes his drivers faulty like that, if the open source amateur project builds them reliably.
for the future of free software, we need to increase the independence of free software against proprietary material.

did AMD ever announce going open source with their drivers?
http://itknowledgeexchange.techtarget.com/enterprise-linux/amd-will-deliver-open-graphics-drivers/
obviously they did already 2 years ago..
whatever happened to that?

how very sorry..

frederik.nnaji >> ... open source drivers from AMD/ATI actually does exist... just try to google radeonhd and/or radeon drivers ;)
See http://www.phoronix.com/forums/forumdisplay.php?f=43 and http://www.x.org/wiki/RadeonFeature (to see the implemented features)

Open drivers - suxx(slow).
Prop drivers - suxx(buggy).

AMD/ATI don't care about users - then AMD/ATI go to hell.

I sold my ATI HD 2600 to windows user. I bought NVIDIA GeForce 9600GT. And I am happy! Games work very fast, HD video work very fast and drivers on my Ubuntu 9.10 installing in just 2 clicks.

I boycott AMD/ATI and I advice the same to you. They can't fix this bug almost year! My nerves not sustained any more.

There has been a new release of catalyst driver last week.... (9.9) Maybe someone could try it out?. Though, I don't think the problem will ever be solved unless to use open source driver, which is what you have to do, even if it still missing some features, once we'll reach karmic+1 or karmic+2, it will be very interesting)

When I use the Switch User applet, the screen goes completely blank. I must power down and start-up again. As my wife and I share the same machine, and have set up 2 users, this is an annoyance.

PC is a Dell 530S, with an ATI Radeon HD 2400 card, VGA-compatible, RV610, 64-bit, 33Mhz. Using the fglrx driver. Not sure which version of the driver.

I believe I'm still seeing this on a fully updated (as of today) Karmic
01:05.0 VGA compatible controller: ATI Technologies Inc RS482 [Radeon Xpress 200M]
Linux inspiron-1501 2.6.31-14-generic #48-Ubuntu SMP Fri Oct 16 14:04:26 UTC 2009 i686 GNU/Linux

I can confirm that this bug remains with the latest karmic and fglrx.
IS ANYBODY READING THIS THREAD ??

I am wholehearted by ubuntu and Karmic is perfect but I can't understand, that so much effort of the community is sent to hell just because of such stupid bug.
>>>>This is probably the most important reason, why Microsoft is still winning so much. <<<<

What about to celebrate THE FIRST ANNIVERSARY OF THIS BUG ?

Ivos,

I am reading the thread. I'm in the same situation you are, and appreciate
your comments. Now, when I upgrade to Karmic, I will know what to expect. I
am just another user, not a developer.

I share your frustration, and I very much appreciate knowing what's
happening with other users in the same situation.

Regards,

Stephen

On Sun, Oct 25, 2009 at 6:48 AM, Ivos <email address hidden> wrote:

> I can confirm that this bug remains with the latest karmic and fglrx.
> IS ANYBODY READING THIS THREAD ??
>
> I am wholehearted by ubuntu and Karmic is perfect but I can't understand,
> that so much effort of the community is sent to hell just because of such
> stupid bug.
> >>>>This is probably the most important reason, why Microsoft is still
> winning so much. <<<<
>
> --
> [fglrx]Computer freeze when switching user
> https://bugs.launchpad.net/bugs/290704
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in NULL Project: Invalid
> Status in Ubuntu: Invalid
> Status in “fglrx-installer” package in Ubuntu: Confirmed
>
> Bug description:
> When I try to switch user by the panel applet or new login in system tools
> or to use guest session, my screen gets dark and computer freezes. It
> happens also in Hardy for me. I found only this line in syslog:
>
> Oct 29 13:47:55 DESKTOP-Linux acpid: client connected from 5435[0:0]
>
>
> and this line in auth.log:
>
> Oct 29 13:48:19 DESKTOP-Linux gnome-screensaver-dialog: gkr-pam: unlocked
> 'login' keyring
>

I also share this frustration. Perhaps someone from ATI can give some indication of when/if they plan
to repair this problem, If they don't, then we can all remove ATI (and AMD) from our vendor lists.

I am going to try to point this out to the folks at Theinquirer and Theregister and see if their news coverage
can increase the interest level at ATI.

Just stop using those fraking fglrx drivers. AMD/ATI won't fix them...and ubuntu devs can't do anything for you in this case !
You can complain against AMD/ATI directly ;)
Fortunately, AMD/ATI has recently released some documentation of their cards, so the open source drivers (radeon and radeonhd) are pretty well advanced now...and there should be completely usable (3D, hd video decoding) with the future version of ubuntu (the one that'll be shipped with kernel 2.6.32). Watch phoronix.com for more information about the status of radeon/radeonhd drivers ;)

Thanks and understand.
Actually that is the problem with Linux generally and the point I believe Ubuntu can push the opensource world a big step ahead. The manufacturers need a strong partner speaking with clear words and having some reasonable number of customers behind it.
What I would expect from Ubuntu and that is actually the only one but very big blame. There is written in hardware driver part of menu, that the fglrx driver ais tested by Ubuntu developers and when I use it, it crashes computer to total death (anything except hard reset doesn't work). Actually I am not developer, but I would expected from a good system, that there could be black screen but I will be still able to control the system by keyboard and maybe get the command line with the open-source driver or anything but not lose the data. ;-)

If Ubuntu should be for a normal people (manufacturers will have to take claims more seriously) there should be tested with at least some HW and ther should be a database, where I would be able to find HW with easy getting knowledge if this HW is tested (It will work without any problems) or at least which problems are facing the users. So I stupid user can look in this DB and see, that ATI will not work with 3D, fglrx driver has so many bugs (list) and look that these bugs were reported to AMD months or years ago and response is fuck-off. Getting access to this DB even from HW driver menu instead of note, that i was tested by Ubuntu developers, AMD will get publicity, that already will wake them up. - Anb I as a stupid user will look into the HW DB and buy card, which is tested by Ubuntu. I as a user will be happy, because of my PC will not crash and will tell everywhere, that Ubuntu is perfectly correct. With correct I mean, that it not work perfect with any HW but will know, that if I use the tested one, I will have NO problem. And I will buy it.

SORRY FO BEING SO LONG BUT WITHOUT DOING THIS, UBUNTU WILL NEVER BE FOR NORMAL USERS AND hw MANUFACTURER WILL NOT TAKE IT SERIOUSLY.

Hi Ivos,

I totally share your excitement and of course I agree with you.
The hardware should be as functional as possible, however it will always
have a small number of bugs which has to be fixed. Nothing is bug-free,
neither in the Windows' world. And that's the reason to make new drivers
and solve old problems. The only confusing thing on this all is, that
this bug has been really let forgotten and nobody cares about it (that's
my view). The only thing we can do is to remind it to the original
developers.

Ivos píše v Po 26. 10. 2009 v 14:27 +0000:
> Thanks and understand.
> Actually that is the problem with Linux generally and the point I believe Ubuntu can push the opensource world a big step ahead. The manufacturers need a strong partner speaking with clear words and having some reasonable number of customers behind it.
> What I would expect from Ubuntu and that is actually the only one but very big blame. There is written in hardware driver part of menu, that the fglrx driver ais tested by Ubuntu developers and when I use it, it crashes computer to total death (anything except hard reset doesn't work). Actually I am not developer, but I would expected from a good system, that there could be black screen but I will be still able to control the system by keyboard and maybe get the command line with the open-source driver or anything but not lose the data. ;-)
>
> If Ubuntu should be for a normal people (manufacturers will have to take
> claims more seriously) there should be tested with at least some HW and
> ther should be a database, where I would be able to find HW with easy
> getting knowledge if this HW is tested (It will work without any
> problems) or at least which problems are facing the users. So I stupid
> user can look in this DB and see, that ATI will not work with 3D, fglrx
> driver has so many bugs (list) and look that these bugs were reported to
> AMD months or years ago and response is fuck-off. Getting access to this
> DB even from HW driver menu instead of note, that i was tested by Ubuntu
> developers, AMD will get publicity, that already will wake them up. -
> Anb I as a stupid user will look into the HW DB and buy card, which is
> tested by Ubuntu. I as a user will be happy, because of my PC will not
> crash and will tell everywhere, that Ubuntu is perfectly correct. With
> correct I mean, that it not work perfect with any HW but will know, that
> if I use the tested one, I will have NO problem. And I will buy it.
>
> SORRY FO BEING SO LONG BUT WITHOUT DOING THIS, UBUNTU WILL NEVER BE FOR
> NORMAL USERS AND hw MANUFACTURER WILL NOT TAKE IT SERIOUSLY.
>
--
Lukáš Chmela <email address hidden>
IRC: <email address hidden>, ICQ: 202077459
514D 3C69 8498 E400 0ACE 90F2 00E3 CACA A5BA ECC7

Hi folks,

I am the original reporter of this bug and of course, I'm still reading
this thread. I also sent a "bug report" on the official AMD site but I
didn't get any response from then, neither it was added into official
release notes as known bug. It is even more frustrating, that it remains
opened for almost a year (believing that the bug was here even earlier).

I finally realized, that it was just a "crew feedback", not a bugreport
site.. well, I'm not very sure if I also filled a bugreport on the
unofficial AMD's bugzilla. I'll make sure and let you know then

Thanks a lot and hope for better results.
--
Lukáš Chmela <email address hidden>
IRC: <email address hidden>, ICQ: 202077459
514D 3C69 8498 E400 0ACE 90F2 00E3 CACA A5BA ECC7

Hi Lukas and all,
please don't understand my comment as complain to developers. There are bugs and will be forever. But the situation with HW is really bad, bad, bad. And the biggest thing Ubuntu can do is to test and make the situation clear. It looks like less important, but I am convinced that actually the Ubuntu NOTE "Tested" instead of "Not recommended, can crash your PC ..." is much more important and should not happen. I personally think, that the historical role of Ubuntu is not in developing new and new SW, but in making the "force", testing and giving stamps "tested" and presenting that e.g. ATI is not Ubuntu friendly etc. ....

From my particular card (ATI Technologies Inc RS482 [Radeon Xpress 200M]) I've found this "You need to have xorg-server 1.5.3 install that is the last version who support fglrx for Radeon Xpress 200M" on the Gentoo wiki at http://en.gentoo-wiki.com/wiki/Dell_Inspiron_1501
Might that be related to this problem?

Same problem running Karmic Koala with fglrx 8.66.....
Hope some day this bug is fixed!

Dream on. Not in the next few releases ....

I can confirm this bug still exists in Ubuntu 9.10 with the ATI Mobility Radeon HD 3400 Series. Unfortunately I am stuck with this card in my laptop, but there's no way I'd ever buy an ATI card for a desktop machine running Linux...

[This bug was nominated for lucid, but no evidence exists in bug report that it still affects lucid, so the nomination is not acceptable at this time. A lucid compatible version of -fglrx was added just recently, so this needs retested.]

This is fixed for me in the latest version of the fglrx driver on lucid.

I just tried the new (10.2) fglrx driver from ati website in my 9.10 Ubuntu and it works well. To be more specific, I can't say, if it was fixed due to Ubuntu updates or the driver, because of I used the open source driver last months.

I will test it in Lucid with ubuntu fglrx driver later on.

>>> For me it's fixed.

This is the first good news I've heard on this topic in many moons!

I'm running Intrepid 8.10, with local kernel mods (PAPI) that I don't want to mess with
at present, so upgrading the kernel at this juncture is not something I want to do,
at least in the near-term.

If someone can give me some assurance that:

1. The new fglrx driver is likely to work with 8.10, and

2. The crash is likely due to the old fglrx driver, rather than to the linux kernel.

then, I'll give the new fglrx code a try.

Otherwise, it'll have to wait until I finish my current project and benchmarks...

Or, if someone can assure me that the kernel/X upgrades are a definite
prerequisite for the fix, then I'll at least know where I'm at with this.

Thanks.
Robert

For me it's fixed in 10.04.

Confirmed, working for me with fglrx installed, latest lucid.

I can now switch to the guest session, but when I logged out of there I had to do a ctl+alt+Backspace to get to the login window again.

[This is an automatic notification.]

Hi Lukáš,

This bug was reported against an earlier version of Ubuntu, can you
test if it still occurs on Lucid?

Please note we also provide technical support for older versions of
Ubuntu, but not in the bug tracker. Instead, to raise the issue through
normal support channels, please see:

    http://www.ubuntu.com/support

If you are the original reporter and can still reproduce the issue on
Lucid, please run the following command to refresh the report:

  apport-collect 290704

If you are not the original reporter, please file a new bug report, so
we can work with you as the original reporter instead (you can reference
bug 290704 in your report if you think it may be related):

  ubuntu-bug xorg

If by chance you can no longer reproduce the issue on Lucid or if you
feel it is no longer relevant, please mark the bug report 'Fix Released'
or 'Invalid' as appropriate, at the following URL:

  https://bugs.launchpad.net/ubuntu/+bug/290704

Hi,
This problem is happening for me on Ubuntu 10.04 LTS.
I have the black screen when switching between users.
The problem doesn't happen immediately, but only after a few switches.
Interesting it only started happening when I upgraded from Hardy to Lucid (and I haven't changed my graphics card between upgrades).
I have NVIDIA Accelerated graphics card (version 173)
If you require further information, please don't hesitate to ask (you may need to give instructions as to how to obtain this info as I'm a relative newbie)

Simon.

Architecture: i386
DistroRelease: Ubuntu 10.04
DkmsStatus:
 nvidia-173, 173.14.22, 2.6.24-27-generic, i686: installed
 nvidia-173, 173.14.22, 2.6.32-22-generic, i686: installed
MachineType: Dell Computer Corporation Dimension 8300
NonfreeKernelModules: nvidia
Package: fglrx-installer (not installed)
ProcCmdLine: root=UUID=87fd2342-0f15-4fcb-aa25-72e306fc182d ro quiet splash
ProcEnviron:
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.32-22.33-generic 2.6.32.11+drm33.2
Tags: lucid lucid
Uname: Linux 2.6.32-22-generic i686
UserGroups: adm admin cdrom dialout dip fax floppy fuse lpadmin plugdev tape video
dmi.bios.date: 02/23/2004
dmi.bios.vendor: Dell Computer Corporation
dmi.bios.version: A05
dmi.board.name: 0W2562
dmi.board.vendor: Dell Computer Corp.
dmi.chassis.type: 6
dmi.chassis.vendor: Dell Computer Corporation
dmi.modalias: dmi:bvnDellComputerCorporation:bvrA05:bd02/23/2004:svnDellComputerCorporation:pnDimension8300:pvr:rvnDellComputerCorp.:rn0W2562:rvr:cvnDellComputerCorporation:ct6:cvr:
dmi.product.name: Dimension 8300
dmi.sys.vendor: Dell Computer Corporation
system:
 distro: Ubuntu
 codename: lucid
 architecture: i686
 kernel: 2.6.32-22-generic
xkbcomp:
 Error: command ['xkbcomp', ':0', '-w0', '-'] failed with exit code 1: No protocol specified
 Error: Cannot open display ":0"
                   Exiting

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

Simon Hill píše v St 26. 05. 2010 v 22:28 +0000:
> Hi,
> This problem is happening for me on Ubuntu 10.04 LTS.
> I have the black screen when switching between users.
> The problem doesn't happen immediately, but only after a few switches.
> Interesting it only started happening when I upgraded from Hardy to Lucid (and I haven't changed my graphics card between upgrades).
> I have NVIDIA Accelerated graphics card (version 173)
> If you require further information, please don't hesitate to ask (you may need to give instructions as to how to obtain this info as I'm a relative newbie)
>
> Simon.
>
> ** Changed in: fglrx-installer (Ubuntu)
> Status: Incomplete => Confirmed
>

Hi Simon,
This bug describes a problem with fglrx drivers which are proprietary
amd drivers. It is not related to nvidia, so you should create a new bug
report.

Bryce,
This bug has been partially solved in lucid. It now doesn't make the pc
to stop responding, it logs in correctly, but the screen is corrupted a
bit (in right bottom corner and cursor as well). I'll post here a photo
of that, it's not capturable by the PrintScreen button.

Anyway, I will do apport-collect, if it may help in any way..

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

For anyone who has nvidia graphics card as is affected by this bug, please report it here:
https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-173/+bug/587069

I get exactly the same bug as Lukas described, starting a second session works, but the cursor and a block of the lower right corner of the screen is corrupted.
This is on Kubuntu 10.04 with the fglrx driver from the repositories, my card is (as listed in lspci) an "ATI Technologies Inc RV730 Pro AGP [Radeon HD 4600 Series]"

@ Heru Herdianto :

Please, don't play with the form options. Thanks in advance.