Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
X.org Lockups (part 2)
View unanswered posts
View posts from last 24 hours
View posts from last 7 days

Goto page Previous  1, 2, 3, ... 35, 36, 37  Next  
This topic is locked: you cannot edit posts or make replies.    Gentoo Forums Forum Index Desktop Environments
View previous topic :: View next topic  
Author Message
chonger
n00b
n00b


Joined: 01 Aug 2004
Posts: 27

PostPosted: Wed May 11, 2005 10:10 pm    Post subject: Same problem Reply with quote

I have been having the same problem recently. My machine would lock up while I'm at work (or overnight) and I'd have to reboot. Attempts to ssh into my machine wouldn't work. Would not respond to pings. Just plain dead.

I can't pin-point what I changed recently to cause this, but I'll throw out some of the changes I've made around the time this problem started.

1. Upgraded to kde 3.4
- No locks for a LONG LONG time.
2. Installed Zope.
3. Installed Plone.
4. Installed various products for Zope.
- Fine for a couple of days, no lock ups.
5. Installed mod_ssl for Apache.
- The lock ups started happening the next day.
6. Stopped running Apache.
- Still locking up
7. Stopped running Zope.
- Still locking up
8. emerge sync && emerge --deep --world
- Still locking up

I would think it was mod_ssl in Apache that was causing problems, but the lock ups persisted even when Apache is not running. I may have installed some other packages around this time that caused this problem, I do not know.

I hope this helps for now. I will be switching to gnome (instead of the unstable kde 3.4) for a while and see if that helps with stability.

I will report back later and keep yall up-2-date on what's going on. Hopefully I can provide some insight into what's been causing this problem.
Back to top
View user's profile Send private message
MalachiX
Tux's lil' helper
Tux's lil' helper


Joined: 10 Apr 2005
Posts: 101
Location: Miami, Fl

PostPosted: Thu May 12, 2005 1:21 am    Post subject: Reply with quote

This isn't a kde bug..... I've noticed a lot of people are saying it is a kde bug. The bug affects all distro's and window managers. We are trying to solve a problem with X freezing/hanging/malfuctioning in general. We have to stop pretending that this is only distro or window manager related. It isn't even Xorg specific, Xfree freezes also.
Back to top
View user's profile Send private message
jdgill0
Veteran
Veteran


Joined: 25 Mar 2003
Posts: 1366
Location: Lexington, Ky -- USA

PostPosted: Thu May 12, 2005 2:21 am    Post subject: Reply with quote

OK, everyone ... got a new "experience" to share.

I upgraded to the following today:
xorg-x11-6.8.2-r2
gentoo-2.6.11-r8
still running ati-drivers-8.12.10 with DRI enabled again.

So, I was playing a DVD movie, was moving around from virt. desktop to virt. desktop, when I got stuck on "one" virt. desktop. I run KDE, so I have a desktop pager in kicker. Although it still showed all my virtual desktops, I could not switch to another desktop -- seemingly condensing all of them to a single desktop, sort of. I could not move any windows around, the title bars of the windows were gone.

HOWEVER, I did not loose all control!!!! ... that is, I could still interact with the applications. In fact I closed everything (or most everything), and it seemed I was closing applications that should have been on different desktops -- the reason I said it sort of condensed my desktops to one desktop ... I then logged out of KDE back to the console!!! ... then restarted X without any problems.
_________________
Vim has excellent syntax highlighting for configuration files: emerge gentoo-syntax
Learn how to use Vim: vimtutor
Back to top
View user's profile Send private message
MalachiX
Tux's lil' helper
Tux's lil' helper


Joined: 10 Apr 2005
Posts: 101
Location: Miami, Fl

PostPosted: Thu May 12, 2005 5:36 am    Post subject: Reply with quote

I noticed odd behavior once upgrading/downgrading also. When I used a former version of glibc, fluxbox just became REALLY sluggish instead of actually freezing. I'm not sure what the meaning of the different behavior is implying though.
Back to top
View user's profile Send private message
Gergan Penkov
Veteran
Veteran


Joined: 17 Jul 2004
Posts: 1464
Location: das kleinste Kuhdorf Deutschlands :)

PostPosted: Thu May 12, 2005 10:18 am    Post subject: Reply with quote

I think that this is a threading problem.
Are there some of you, which are NOT using nptl, there were some posts about switching to 2.4.x kernel helped in stopping the lockups, which means that not using the nptl must be fine. All the freezes have sth in common, they freeze the current screen, but all the demons and the programs continue to run, which is to say that this is userspace problem. The thing is that this comes up quickly when we are using proprietary drivers, which stresses the system in some way and accellerates the freezes, I expect that there are lockups with the opensource drivers, but they are so rare, that no one seems to notice them.
I don't have lockups for 2-3 weeks now with the nvidia-driver (RenderAccel not set), but somehow all the consoles are gone dead before a week or so ::)), which is not a serious problem, as I could work with X and it doesn't freeze::)), but on the other hand shows that the problem is still here and the next victim could be X :::)))
_________________
"I knew when an angel whispered into my ear,
You gotta get him away, yeah
Hey little bitch!
Be glad you finally walked away or you may have not lived another day."
Godsmack
Back to top
View user's profile Send private message
Nerdanel
Apprentice
Apprentice


Joined: 27 Apr 2003
Posts: 161
Location: Finland

PostPosted: Thu May 12, 2005 12:23 pm    Post subject: Reply with quote

Gergan Penkov wrote:
I expect that there are lockups with the opensource drivers, but they are so rare, that no one seems to notice them.


I'm having lockups with opensource nv. They used to happen once a day on average, which is very noticeable as I used to have my machine on for months on end, but my increasing of Konqueror's list of addresses to remember from 500 to 5000 has made them happen once in a few days.
Back to top
View user's profile Send private message
jdgill0
Veteran
Veteran


Joined: 25 Mar 2003
Posts: 1366
Location: Lexington, Ky -- USA

PostPosted: Thu May 12, 2005 3:37 pm    Post subject: Reply with quote

Gergan Penkov,

You brought up the idea of maybe a threading problem specifically with nptl. I have read elsewhere that USE="nptl nptlonly" should be used together, else it can cause problems. What is your all's take on that? Are you all already using both? I am currently not using nptlonly.
_________________
Vim has excellent syntax highlighting for configuration files: emerge gentoo-syntax
Learn how to use Vim: vimtutor
Back to top
View user's profile Send private message
MalachiX
Tux's lil' helper
Tux's lil' helper


Joined: 10 Apr 2005
Posts: 101
Location: Miami, Fl

PostPosted: Thu May 12, 2005 3:42 pm    Post subject: Reply with quote

I don't have either of those keywords in my USE flags.
Back to top
View user's profile Send private message
drphibes
Guru
Guru


Joined: 30 Nov 2004
Posts: 432

PostPosted: Thu May 12, 2005 4:25 pm    Post subject: Reply with quote

jdgill0 wrote:
Gergan Penkov,

You brought up the idea of maybe a threading problem specifically with nptl. I have read elsewhere that USE="nptl nptlonly" should be used together, else it can cause problems. What is your all's take on that? Are you all already using both? I am currently not using nptlonly.

I use nptl in /etc/make.conf and nptlonly in /etc/portage/package.use (glibc entry). The nptl flag adds support for nptl, but also builds the linuxthreads libs. nptl + nptlonly builds only the nptl libs. If you want nptlonly, I believe you should use both flags since the flag nptl is what _includes_ nptl, whereas nptlonly is what _excludes_ linuxthreads.

For me the lockups were really bad with nvidia-kernel 7174. I am back to 6629-r4 and am stable again. My symptoms were the usual X frozen but mouse works, able to login via ssh, top shows X at 99% cpu usage. The only GL app I use regularly is xscreensaver w/pipes. I use xv exclusively for video apps like mplayer and xine and tvtime.

Asus P4PE / P4 2.4 GHz
NV20 chipset (Geforce3) graphics card
NvAGP 1
Xorg 6.8.2-r1
Xfce 4.2.1.1
kernel 2.6.11-gentoo-r8
gcc 3.3.5-20050130 (toolchain is all x86)
glibc 2.3.4.20041102-r1 (nptlonly)

ACCEPT_KEYWORDS=x86 with quite a few ~x86 packages installed

doc


Last edited by drphibes on Thu May 12, 2005 4:36 pm; edited 1 time in total
Back to top
View user's profile Send private message
Gergan Penkov
Veteran
Veteran


Joined: 17 Jul 2004
Posts: 1464
Location: das kleinste Kuhdorf Deutschlands :)

PostPosted: Thu May 12, 2005 4:31 pm    Post subject: Reply with quote

I have wrote earlier that my problem is that in the moment in which I changed my profile to 2005.0, ufed and profuse stopped working and have cleared all my use flags, so I could not say, which flag or which update has led to this specific problem, but I think that earlier I have used nptlonly and I haven't had any ipv6 things on. But the problem seems irreversible in some sense, which is doubtfull, because this could only mean, that it is in the hardware::)) I think that it is possible to revert all of this if we downgrade kernel-headers together with glibc, but I don't have the time to do this. I don't know how gentoo installs the headers, there seems to be some configuration but I don't see the config file, used in the ebuild and no one says what are the options, which are enabled in the kernel-headers-ebuild, which after all is sth very disturbing:( The other option is to build glibc against the 2.4 headers and try if this helps (of course without nptl).
_________________
"I knew when an angel whispered into my ear,
You gotta get him away, yeah
Hey little bitch!
Be glad you finally walked away or you may have not lived another day."
Godsmack
Back to top
View user's profile Send private message
drphibes
Guru
Guru


Joined: 30 Nov 2004
Posts: 432

PostPosted: Thu May 12, 2005 4:49 pm    Post subject: Reply with quote

I think until we can reliably reproduce the lockup, we'll never diagnose it. All we really have are anecdotal reports on a huge variety of platforms and, unfortunately, we can only disprove by counterexample. Someone reports a certain combination works -- someone else knocks it down, etc. Reporting that a combination "works" solves, proves and demonstrates very little. Sadly.
Back to top
View user's profile Send private message
MalachiX
Tux's lil' helper
Tux's lil' helper


Joined: 10 Apr 2005
Posts: 101
Location: Miami, Fl

PostPosted: Thu May 12, 2005 5:29 pm    Post subject: Reply with quote

While salvaging whatever information I could find I found a forum where a guy reported that he had this same problem and after upgrading his BIOS the latest version everything was working correctly again. Can anyone here report for sure that they are using the latest BIOS? I know for sure I'm a whole version behind in mine.
Back to top
View user's profile Send private message
Gergan Penkov
Veteran
Veteran


Joined: 17 Jul 2004
Posts: 1464
Location: das kleinste Kuhdorf Deutschlands :)

PostPosted: Thu May 12, 2005 5:32 pm    Post subject: Reply with quote

I use epox 8rda nforce2 mainboard and have its bios updated after the first lockups and it didn't helped me ::((
_________________
"I knew when an angel whispered into my ear,
You gotta get him away, yeah
Hey little bitch!
Be glad you finally walked away or you may have not lived another day."
Godsmack
Back to top
View user's profile Send private message
Gergan Penkov
Veteran
Veteran


Joined: 17 Jul 2004
Posts: 1464
Location: das kleinste Kuhdorf Deutschlands :)

PostPosted: Thu May 12, 2005 9:41 pm    Post subject: Reply with quote

It seems to me that changing from linux26-headers-2.6.8.1-r2 to linux-headers-2.6.8.1-r4 has created all this problems and I am not sure if the version change or the ebuild change has triggered all this. I will try and downgrade to 2.6.8.1-r2 and see if this helps.
_________________
"I knew when an angel whispered into my ear,
You gotta get him away, yeah
Hey little bitch!
Be glad you finally walked away or you may have not lived another day."
Godsmack
Back to top
View user's profile Send private message
jdgill0
Veteran
Veteran


Joined: 25 Mar 2003
Posts: 1366
Location: Lexington, Ky -- USA

PostPosted: Fri May 13, 2005 2:12 am    Post subject: Reply with quote

All of my problems have been occuring under linux-headers-2.6.8.1-r2 ... I have never upgraded to 2.6.8.1-r4. In fact the headers I have back go back to when they were still called linux26-headers , i.e. the last emerge I have is linux26-headers-2.6.8.1-r2, on Feb. 12th, 2005.
_________________
Vim has excellent syntax highlighting for configuration files: emerge gentoo-syntax
Learn how to use Vim: vimtutor
Back to top
View user's profile Send private message
Nerdanel
Apprentice
Apprentice


Joined: 27 Apr 2003
Posts: 161
Location: Finland

PostPosted: Fri May 13, 2005 6:05 am    Post subject: Reply with quote

My first lockups occurred under linux-headers-2.4.22-r1. My current linux-headers are 2.6.8.-r1.
Back to top
View user's profile Send private message
vatachino
n00b
n00b


Joined: 14 Oct 2004
Posts: 29

PostPosted: Fri May 13, 2005 8:02 am    Post subject: Reply with quote

I post this in the hope that it might be helpful to somebody. X used to lock up for me. I haven't had any lockups since I made the following alterations a few weeks ago (I did them to get suspend-to-disk2 working from http://www.suspend2.net):

    Disable AGP support in the kernel (CONFIG_AGP is not set) and recompile.

    Remove all references to DRI in xorg.conf

    Use NVAGP "1" (0 works too)

This still gives me 1500+ fps from glxgears, which is a bit down from the 1800+ which I've got previously, but I'd rather have a stable X session than highly optimised OpenGl performance!
Setup details:
Sony Vaio PCG-GRT996ZP
i686 Mobile Intel(R) Pentium(R) 4 CPU 3.20GHz
Vanilla Kernel 2.6.11.7 (SMP/HT, 4k stacks) with suspend2 patches
Nvidia GeForce FX Go5600 rev 161
nvidia-kernel-1.0.7174-r1
Back to top
View user's profile Send private message
Gergan Penkov
Veteran
Veteran


Joined: 17 Jul 2004
Posts: 1464
Location: das kleinste Kuhdorf Deutschlands :)

PostPosted: Fri May 13, 2005 8:11 am    Post subject: Reply with quote

Nerdanel, jdgill0, thanks for the reply, you save me from a new total recompile:::)))But it is somehow strange the problem must be somewhere, i swear i didn't have such lockups earlier... this is simply a disaster. I have no more ideas:(
_________________
"I knew when an angel whispered into my ear,
You gotta get him away, yeah
Hey little bitch!
Be glad you finally walked away or you may have not lived another day."
Godsmack
Back to top
View user's profile Send private message
bravecobra
Tux's lil' helper
Tux's lil' helper


Joined: 26 Dec 2002
Posts: 130
Location: Planet Earth (sometimes)

PostPosted: Fri May 13, 2005 8:31 am    Post subject: Reply with quote

I'm havin lockups too but that with the ati-drivers, so this fix doesn't apply to me. I've been follwing this thread for some time now, still no solution.
_________________
Brave Cobra
http://www.bravecobra.com
Back to top
View user's profile Send private message
DrKayBee
Tux's lil' helper
Tux's lil' helper


Joined: 11 May 2004
Posts: 143
Location: Columbus, OH

PostPosted: Fri May 13, 2005 12:36 pm    Post subject: I can reliably crash X Reply with quote

I've gone through the thread so far. I have been sufferring from random X lockups while watching MythTV / reading the web using firefox / watching movies etc. I am running 2.6.10-r6, KDE 3.3, Nvidia MX4400 PCI (not AGP) (nvidia's binary driver) on a sempron 2600 with 512 ram on a soyo dragon (Via kt800) mobo. The X version is 6.8.2-r2.

Here is my tested and true way to crash X:
1: Open Konsole
2: Do something that produces a large text output: e.g. "emerge -uDpv world"
3: Start scrolling up with the scroll wheel
4: X will crash, but mouse will continue to respond. It's not possible to go to a virtual terminal using Ctrl-Alt-F1 as well.

No other programs are loaded. The computer responds to SSH - I can log in and look at the processes using top. Here I see that X is consuming 99.7% of the CPU - essentially, as much as it can get. All the other processes are still running.

I've not had this problem running games (Neverwinter natively / Counterstrike on Cedega) so far. I think I can rule out flaky hardware after hours of successful emerging / playing 3D games and the like.

Well - that's my one additional data point to a saga that's going on far too long.
_________________
Cheers,
KB
<-----------
Yes - that's Herman Toothrot
Back to top
View user's profile Send private message
vatachino
n00b
n00b


Joined: 14 Oct 2004
Posts: 29

PostPosted: Fri May 13, 2005 1:59 pm    Post subject: Reply with quote

In order to isolate whether the hangs are due to hardware-specific issues, or whether they are xorg related, it would be a useful test if those experiencing hangs could try using the VESA xorg driver. Using the VESA driver with the vesafb-tng option in standard Gentoo kernels (as described in http://gentoo-wiki.com/HOWTO_fbsplash), I can get my full 1600x1200 screen resolution - albeit at only 16bit.

The key bits of my xorg.conf look like this:

Code:
Section "Device"
    Identifier  "vesa"
    Driver      "vesa"
EndSection


Section "Screen"
    Identifier  "Screen 1"
    Device       "vesa"
    Monitor     "laptop panel"
    DefaultDepth 16

    Subsection "Display"
        Depth       16
        Modes       "1600x1200" "1280x1024" "1024x768" "800x600" "640x480"
    EndSubsection
EndSection


I also have to do

Code:
opengl-update xorg-x11


and I get a working system.

It would be very interesting to find if this change in the graphics config cures the hang/lockup problem people are having - because in this case it would confirm that it is not xorg in general which is the problem but specific hardware acceleration code.
Back to top
View user's profile Send private message
Nerdanel
Apprentice
Apprentice


Joined: 27 Apr 2003
Posts: 161
Location: Finland

PostPosted: Fri May 13, 2005 2:27 pm    Post subject: Reply with quote

I'm trying to make a list of vaguely suspicious low-level programs I emerged before the freezes appeared. I'm not counting Gentoo's r-versions.

glib 2.4.8 or 2.6.2
util-linux 2.12i
procps 3.2.4
reiserfsprogs 3.6.19
binutils 2.15.92
gcc 3.3.5
bin86 0.16.16 or 0.16.17
autoconf and automake (muttiple versions with wrappers)
libtool 1.5.10
psmisc 21.5
libpcre 5.0

I can't quite recall how long my box had been up, so this is inexact... Note also that I can't shut out a hardware problem since my problems first appeared when my motherboard broke and had to be replaced.
Back to top
View user's profile Send private message
lizardcry
Tux's lil' helper
Tux's lil' helper


Joined: 13 May 2005
Posts: 87
Location: Iceland

PostPosted: Fri May 13, 2005 2:42 pm    Post subject: Reply with quote

Try out adding these params to the kernel:

idle=poll noapic nolapic

also, try hdparm -X udma3




Nerdanel wrote:
I'm trying to make a list of vaguely suspicious low-level programs I emerged before the freezes appeared. I'm not counting Gentoo's r-versions.

glib 2.4.8 or 2.6.2
util-linux 2.12i
procps 3.2.4
reiserfsprogs 3.6.19
binutils 2.15.92
gcc 3.3.5
bin86 0.16.16 or 0.16.17
autoconf and automake (muttiple versions with wrappers)
libtool 1.5.10
psmisc 21.5
libpcre 5.0

I can't quite recall how long my box had been up, so this is inexact... Note also that I can't shut out a hardware problem since my problems first appeared when my motherboard broke and had to be replaced.
Back to top
View user's profile Send private message
alaa
n00b
n00b


Joined: 23 Feb 2004
Posts: 57

PostPosted: Fri May 13, 2005 2:51 pm    Post subject: Reply with quote

I would just like to share something with everyone here. I'm suffering from this X lockup problem on only one of my two machines. There is really no major difference between the two boxes in terms of running software except that my emerge history on both is quite different. The one that locks up is the only machine on which I used to use the -D option while emerging.

Another interesting thing to note is that my roommate's work PC (which is exactly the same as my machine which locks up) is solid as a rock, and he says that he doesn't have "any" USE flags set.

I'm thinking the problem might be due to the way we've been emerging packages plus the USE flags.

The good news is that I've recently purchased a new PC for my home (Athlon 64 3500+) and I'm installing gentoo on it right now (with no USE flags, i.e. whatever the defaults are), and I'll set it up exactly the same way I setup my work machine (you know, the one that locks up).

So expect an update from me in a couple of days.
Back to top
View user's profile Send private message
chonger
n00b
n00b


Joined: 01 Aug 2004
Posts: 27

PostPosted: Fri May 13, 2005 4:22 pm    Post subject: Reply with quote

Reporting back:

For what its worth, I haven't had any more problems after switching to gnome. This is not the solution to my problem, but at least its buying my system some stability so it doesn't freeze my server everytime I go to work.

To provide some background on why I decided to switch away from KDE. I've noticed that I can force a lockup on my system when running the unstable KDE (emerge with ~x86 unmasked) AND pasting a huge amount (text, 362kb) of contents into the klipper AND double clicking on the klipper icon in the system tray. I thought to myself that if interaction between KDE's klipper and the clipboard can cause the system to freeze, then I should try using gnome for a while.

I will report back again and update on the stability of my system.
Back to top
View user's profile Send private message
Display posts from previous:   
This topic is locked: you cannot edit posts or make replies.    Gentoo Forums Forum Index Desktop Environments All times are GMT
Goto page Previous  1, 2, 3, ... 35, 36, 37  Next
Page 2 of 37

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum