Hi,
As if to say, "F"-you, the screen changed from my normal desktop, with Firefox, Azureus, a few terminals, and a few VMs running, to a solid blue screen with just the Fedora squiggly 'F' in the middle of the screen, and the whole computer locked up.
Completely dead. Catatonic. Unresponsive. Nothing in the logs.
What the hell happened?
This is really aggravating, because it has been one problem after another, and I'm beginning to rethink my decision on this desktop.
Ideas greatly appreciated. Thanks, Alex
On 27 July 2010 16:44, Alex mysqlstudent@gmail.com wrote:
Hi,
As if to say, "F"-you, the screen changed from my normal desktop, with Firefox, Azureus, a few terminals, and a few VMs running, to a solid blue screen with just the Fedora squiggly 'F' in the middle of the screen, and the whole computer locked up.
Completely dead. Catatonic. Unresponsive. Nothing in the logs.
you could reboot, switch to a text console (e.g ctrl f2) have a look in $HOME/.xsession-errors, though I've you've logged in again already it's probably been overwritten. my guess would be that something killed your X session. If you're running f13 and have an nvidia card, then I've seen this too. The only way i got a working environment was to switch to nouveau and disable 3d eye candy. Anything else caused problems. (actually the f12 kernel + f13 nvidia rpms also worked on f13, but I decided that was a backwards step) if you have other hardware or are running f12 then this is something else (though looking in .xsession-errors is still valid).
What the hell happened?
This is really aggravating, because it has been one problem after another,
that's kind of the point of fedora. if you want something stable, use CentOS (or RHEL) or maybe an ubuntu variant. I think the benefits of a cutting edge distro outweigh the drawbacks, but I agree that it is very frustrating when something fundamental stops working.
and I'm beginning to rethink my decision on this desktop.
Ideas greatly appreciated. Thanks, Alex
Regards,
Chris
Hi,
Completely dead. Catatonic. Unresponsive. Nothing in the logs.
you could reboot, switch to a text console (e.g ctrl f2) have a look in $HOME/.xsession-errors, though I've you've logged in again already it's probably been overwritten. my guess would be that something killed your X session. If you're running f13 and have an nvidia card,
It's a Sapphire HD5770 (Radeon). Here's the .xsession-errors.old, which I believe is the one from the crash:
There are quite a few errors in there, but nothing that I think is directly related to it going catatonic. Bunch of network errors from azureus, which I assume happened after whatever killed the network and desktop.
that's kind of the point of fedora. if you want something stable, use CentOS (or RHEL) or maybe an ubuntu variant. I think the benefits of a cutting edge distro outweigh the drawbacks, but I agree that it is very frustrating when something fundamental stops working.
Yes, frustrating, particularly when it's one fundamental thing after another, like ethernet detection, printing, and basic networking.
Thanks, Alex
On 07/27/2010 04:16 PM, Alex wrote:
that's kind of the point of fedora. if you want something stable, use CentOS (or RHEL) or maybe an ubuntu variant. I think the benefits of a cutting edge distro outweigh the drawbacks, but I agree that it is very frustrating when something fundamental stops working.
Yes, frustrating, particularly when it's one fundamental thing after another, like ethernet detection, printing, and basic networking.
Sure, but it works both ways: sometimes getting up-to-date software from upstream fixes bugs that in a more "stable" distro you'd have to wait much longer for. For example, I used to have to install the proprietary nvidia driver, but one day I upgraded Fedora and the latest nouveau driver just worked. There's a lot to be said for that.
Andrew.
Hi,
Yes, frustrating, particularly when it's one fundamental thing after another, like ethernet detection, printing, and basic networking.
Sure, but it works both ways: sometimes getting up-to-date software from upstream fixes bugs that in a more "stable" distro you'd have to wait much longer for. For example, I used to have to install the
Yes, for the most part, I agree. However, I've set up FC13 on at least three desktops with varying processors, motherboards, ethernet, disks, etc, and have had problems that would have prevented someone with less experience than myself from being able to install it. On one of them, I had to rmmod the forcedeth and r8169 drivers, then modprobe the r8169 driver in order for it to properly link to the switch, all while the installer is running. On another, the installer would crash when making changes to the disk layout.
I've used Fedora since it was Red Hat, and started with before 4.1, probably some twelve years ago. The first install was Yggdrasil Linux on 60 floppy disks on my 80386. I pretty much dropped Linux on the desktop entirely a number of years ago because I got frustrated with the lack of progress.
By mentioning that I've been using Linux a really long time, I don't mean to imply I'm an expert, but I have quite a bit of experience, and don't plan on giving up on it ever. I mentioned the forcedeth example because the board was at least three years old, so it's not like it was some experimental driver where I'm the first person to use it.
My staff uses Ubuntu on their desktop because "it just works", but I like RPM better, have been using Red Hat for ever, and believe that they have better availability to development, new applications, and wider support for new technologies.
I'm sure that if I spent the time I could figure out where the problem is with ghostscript that is causing my printing problems, but that's not my job any longer. I'm just often very surprised when I see these "RHCE" guys that can't troubleshoot these types of problems, and wonder if they're the same clueless ones that are testing these before RH releases it.
Thanks for listing to me rant for a bit. Regards, Alex
technologies. I just wish
On 07/28/2010 12:12 AM, Alex wrote:
Yes, frustrating, particularly when it's one fundamental thing after another, like ethernet detection, printing, and basic networking.
Sure, but it works both ways: sometimes getting up-to-date software from upstream fixes bugs that in a more "stable" distro you'd have to wait much longer for. For example, I used to have to install the
Yes, for the most part, I agree. However, I've set up FC13 on at least three desktops with varying processors, motherboards, ethernet, disks, etc, and have had problems that would have prevented someone with less experience than myself from being able to install it. On one of them, I had to rmmod the forcedeth and r8169 drivers, then modprobe the r8169 driver in order for it to properly link to the switch, all while the installer is running. On another, the installer would crash when making changes to the disk layout.
I've used Fedora since it was Red Hat, and started with before 4.1, probably some twelve years ago. The first install was Yggdrasil Linux on 60 floppy disks on my 80386. I pretty much dropped Linux on the desktop entirely a number of years ago because I got frustrated with the lack of progress.
By mentioning that I've been using Linux a really long time, I don't mean to imply I'm an expert, but I have quite a bit of experience, and don't plan on giving up on it ever. I mentioned the forcedeth example because the board was at least three years old, so it's not like it was some experimental driver where I'm the first person to use it.
OK.
My staff uses Ubuntu on their desktop because "it just works", but I like RPM better, have been using Red Hat for ever, and believe that they have better availability to development, new applications, and wider support for new technologies.
I'm sure that if I spent the time I could figure out where the problem is with ghostscript that is causing my printing problems, but that's not my job any longer. I'm just often very surprised when I see these "RHCE" guys that can't troubleshoot these types of problems, and wonder if they're the same clueless ones that are testing these before RH releases it.
Thanks for listing to me rant for a bit.
The crack about "these RHCE guys" is rather out of order.
However, fair enough: everyone's experience is valuable. But while I'm not going to deny your problems, I just haven't had an experience anything like that with any of my boxes. It's inevitable that, with improvements, some things will break with a new release, given the huge range of PC hardware.
But I'm not being complacent. Any installation problems like the one you describe are serious.
Andrew.
Hi,
Thanks for listing to me rant for a bit.
The crack about "these RHCE guys" is rather out of order.
However, fair enough: everyone's experience is valuable. But while I'm not going to deny your problems, I just haven't had an experience anything like that with any of my boxes. It's inevitable that, with improvements, some things will break with a new release, given the huge range of PC hardware.
Yes, I agree, and appreciate your attention. I don't mean to sound disgruntled or defamatory, and also don't mean to imply they are categorically clueless. Probably should have left that comment out; it just stands to my frustration, particularly after posting questions repeatedly without solutions.
I also realize it's not "shovelware"; there are always tons of new features, and development is rapid, and usually well-tested. It's certainly much more complex than it was many years ago. Maybe I just wish there was a better balance between being able to rely on it for desktop productivity, with more of the development left for rawhide.
Thanks, Alex
On Tue, 2010-07-27 at 10:44 -0400, Alex wrote:
Hi,
As if to say, "F"-you, the screen changed from my normal desktop, with Firefox, Azureus, a few terminals, and a few VMs running, to a solid blue screen with just the Fedora squiggly 'F' in the middle of the screen, and the whole computer locked up.
Completely dead. Catatonic. Unresponsive. Nothing in the logs.
What the hell happened?
This is really aggravating, because it has been one problem after another, and I'm beginning to rethink my decision on this desktop.
Sounds like your machine started going into suspend / hibernate / shutdown.
Did you close the lid? Did you set it to suspend on lid close?
~m
Hi,
Sounds like your machine started going into suspend / hibernate / shutdown.
Did you close the lid? Did you set it to suspend on lid close?
It's my desktop. I have the display set to turn off after 10 minutes, but I don't think it was even idle for ten minutes. The rest of the power management doesn't work.
Oh yeah, forgot to add that to the list of fundamental problems -- power management doesn't work. Here's a few messages relating to that:
powernow-k8: Found 1 AMD Phenom(tm) II X6 1090T Processor processors (5 cpu cores) (version 2.20.00) [Firmware Bug]: powernow-k8: No compatible ACPI _PSS objects found. [Firmware Bug]: powernow-k8: Try again with latest BIOS.
The sixth core isn't enabled for some reason.
It's a very recent board (Asus Crosshair IV) with the latest BIOS. Where is the best place to go for power management docs?
Thanks, Alex
On Tue, Jul 27, 2010 at 11:25 AM, Alex mysqlstudent@gmail.com wrote:
Hi,
Sounds like your machine started going into suspend / hibernate / shutdown.
Did you close the lid? Did you set it to suspend on lid close?
It's my desktop. I have the display set to turn off after 10 minutes, but I don't think it was even idle for ten minutes. The rest of the power management doesn't work.
Oh yeah, forgot to add that to the list of fundamental problems -- power management doesn't work. Here's a few messages relating to that:
powernow-k8: Found 1 AMD Phenom(tm) II X6 1090T Processor processors (5 cpu cores) (version 2.20.00) [Firmware Bug]: powernow-k8: No compatible ACPI _PSS objects found. [Firmware Bug]: powernow-k8: Try again with latest BIOS.
The sixth core isn't enabled for some reason.
It's a very recent board (Asus Crosshair IV) with the latest BIOS. Where is the best place to go for power management docs?
Well, let's see.
There are potentially so many different things going on here that, were it me, I'd want to make sure that the machine does something sane under some circumstance, even if it meant installing a different operating system on a different hard drive.
"Sane," as far as I'm concerned, would mean recognizing all six cores. Maybe all the problems aren't Fedora.
Robert.
Hi,
(5 cpu cores) (version 2.20.00) [Firmware Bug]: powernow-k8: No compatible ACPI _PSS objects found. [Firmware Bug]: powernow-k8: Try again with latest BIOS.
The sixth core isn't enabled for some reason.
It's a very recent board (Asus Crosshair IV) with the latest BIOS. Where is the best place to go for power management docs?
There are potentially so many different things going on here that, were it me, I'd want to make sure that the machine does something sane
There isn't anything wrong with the computer. I went into the BIOS and enabled the sixth core. For some reason it was disabled, perhaps during the BIOS update, and I just didn't go into the BIOS to check. I've put the board, chip, memory, and all other components through pretty aggressive tests prior to the install.
Maybe all the problems aren't Fedora.
I hacked up the installer during installation to get the ethernet controller module to load properly. The printer works fine on Ubuntu and Windows and a much older version of Linux. I learned what not to do during the install to prevent the disk partitioning from crashing. Even six months after it was released some java(script?) applications would crash FF. That's now apparently fixed.
At this time I think the power management problems on my AMD X6 processor and Asus Crosshair MB are related to the upstream power management developers, so it's perhaps only indirectly related to Fedora. However, there is still a "Suspend" button on my desktop that is ineffective, and Fedora should know that it doesn't work and either disable it or produce some type of information/error message that lets me know it's not working.
The VM and bridging problems I have with Fedora, including the inability for the network-manager to support bridging configurations, is a feature drawback which I don't blame on them, but I still wonder how anyone could be productive with the virt-manager application without having a bridge, which has to be done by hand. The two don't seem to work together, which makes it pretty impossible to use in any effective way.
Thanks, Alex
Alex <mysqlstudent <at> gmail.com> writes:
Hi,
As if to say, "F"-you, the screen changed from my normal desktop, with Firefox, Azureus, a few terminals, and a few VMs running, to a solid blue screen with just the Fedora squiggly 'F' in the middle of the screen, and the whole computer locked up. ...
Hi, I have had a couple of similar experiences with repeating pattern: - F13 with Gnome - system with what I consider acceptable CPU (Intel(R) Celeron(R) M processor), Intel graphics (well supported), RAM and swap for my average daily tasks Mem: 760192k total, 710384k used, 49808k free, 50372k buffers Swap: 1171760k total, 6080k used, 1165680k free, 479000k cached - always Firefox open with at least 5-8 tabs I suspect it to be the main source of instability: - directly I happened to visit some sites that were so "intensely heavy" in content that they could literally lock the FF (and the machine) for good. Perhaps the cause was HTML plus JavaScript plus intentionally "cleverly" coded pages (I even had Java Runtime turned off) that apparently FF could not process smoothly. I think FF should quickly move to tabs and other processes separation as Google Chrome did. - indirectly I visited a very popular site with a small Java applet that on a first time startup almost regularly knocked the FF down; I passed its URL to FF devs). Java environment is a hog and a cause of instability, still. - complete lockup, could not open xterm to kill some process, no reaction to CTL-Alt-Backspace, could move cursor only. For that reason I added to my panel the "bomb" app (Force a misbehaving application to quit) to get access to my desktop system (I hope to be effective with it; waiting to try it out ...).
I could only add that, because I prefer a text login, I capture my X and desktop apps workings with ~/.startx.log file: $ alias startx='/usr/bin/startx -- -nolisten tcp >& .startx.log' This may give some clue about the nature of any problems.
I find F13 quite stable (despite its dev nature); I prefer Gnome as well over KDE, by any means. So I do not think you should drop either of them. JB
Hi,
I suspect it to be the main source of instability: - directly I happened to visit some sites that were so "intensely heavy" in content that they could literally lock the FF (and the machine) for good. Perhaps the cause was HTML plus JavaScript plus intentionally "cleverly" coded pages (I even had Java Runtime turned off) that apparently FF could not process smoothly.
Had it consumed all your available swap, too? Was the disk thrashing while this was happening? It just locks the machine, and doesn't crash so you can create a bug report and send the FF people?
I think FF should quickly move to tabs and other processes separation as Google Chrome did.
I'm sure there's a reason they haven't done this. Perhaps we'll see it in the v4.0 version, which should be available shortly.
- indirectly I visited a very popular site with a small Java applet that on a first time startup almost regularly knocked the FF down; I passed its URL to FF devs).
Often times these problems actually result from various add-ons which are poorly written, leak memory, leave files open, locking problems, etc.
Java environment is a hog and a cause of instability, still.
I found Azureus on FC13 consuming 5GB of RAM! Damn. Five friggin gigs!
- complete lockup, could not open xterm to kill some process, no reaction to
CTL-Alt-Backspace, could move cursor only. For that reason I added to my panel the "bomb" app (Force a misbehaving application to quit) to get access to my desktop system (I hope to be effective with it; waiting to try it out ...).
I don't understand how that would work. If you're able to click on an application to kill something, why can't you just manually kill it?
$ alias startx='/usr/bin/startx -- -nolisten tcp >& .startx.log' This may give some clue about the nature of any problems.
There are a few errors. I'll post a separate thread to see if someone can help me troubleshoot them.
I find F13 quite stable (despite its dev nature); I prefer Gnome as well over KDE, by any means. So I do not think you should drop either of them.
It's gotten better over the past few months.
Thanks, Alex
On 27/07/10 15:44, Alex wrote:
Hi,
As if to say, "F"-you, the screen changed from my normal desktop, with Firefox, Azureus, a few terminals, and a few VMs running, to a solid blue screen with just the Fedora squiggly 'F' in the middle of the screen, and the whole computer locked up.
Having a quick look. What is the host box? cpu\ram Is NetworkManager runnng? What do you use for the VM's? Vbox xen kvm etc.
Hi,
As if to say, "F"-you, the screen changed from my normal desktop, with Firefox, Azureus, a few terminals, and a few VMs running, to a solid blue screen with just the Fedora squiggly 'F' in the middle of the screen, and the whole computer locked up.
Having a quick look. What is the host box? cpu\ram Is NetworkManager runnng? What do you use for the VM's? Vbox xen kvm etc.
It's an Asus Crosshair IV with 8GB and an AMD X6 1090T, Sapphire HD5770 and a WD 6G/s SATA3 disk.
I'm using KVM, but it just happened again without any VMs running. I'm now not using NetworkManager, and instead just 'network', since I wanted to use bridging, which NetworkManager doesn't seem to support.
There's nothing in the logs to indicate what may be causing this. I didn't see it happen the last time, since the monitor was turned off, but the previously it presented a blue screen with the F in the center, and no mouse cursor.
Thanks, Alex
On Wed, Jul 28, 2010 at 3:43 PM, Alex mysqlstudent@gmail.com wrote:
There's nothing in the logs to indicate what may be causing this. I didn't see it happen the last time, since the monitor was turned off, but the previously it presented a blue screen with the F in the center, and no mouse cursor.
That's not really a BSOD, it's the Fedora boot splash screen. You can hit ESC and have the system print the text boot messages (uncovering them by clearing the blue screen).
Hi,
That's not really a BSOD, it's the Fedora boot splash screen. You can hit ESC and have the system print the text boot messages (uncovering them by clearing the blue screen).
Nope, it was completely unresponsive. No network, keyboard, or mouse. No ctrl-alt-bs.
The one time I saw it happen, the blue splash screen appeared and I believe nearly immediately it became unresponsive.
Under what circumstances would that splash screen occur, when the system had already booted and I had logged in? I don't believe the system was idle for longer than a few minutes, and the screensaver/power management turns off the monitor after ten minutes, so I don't believe that was the problem either.
Thanks, Alex
On Wed, Jul 28, 2010 at 4:06 PM, Alex mysqlstudent@gmail.com wrote:
Hi,
That's not really a BSOD, it's the Fedora boot splash screen. You can hit ESC and have the system print the text boot messages (uncovering them by clearing the blue screen).
Nope, it was completely unresponsive. No network, keyboard, or mouse. No ctrl-alt-bs.
The one time I saw it happen, the blue splash screen appeared and I believe nearly immediately it became unresponsive.
You may consider disabling the splash screen in /boot/grub/grub.conf by finding your kernel entry:
title Fedora (2.6.33.6-147.fc13.i686.PAE) root (hd0,0) kernel /vmlinuz-2.6.33.6-147.fc13.i686.PAE ro root=/dev/mapper/vg_circe-lv_root rd_LVM_LV=vg_circe/lv_root rd_LVM_LV=vg_circe/lv_swap rd_NO_LUKS rd_NO_MD rd_NO_DM LANG=en_US.UTF-8 SYSFONT=latarcyrheb-sun16 KEYTABLE=us rhgb quiet initrd /initramfs-2.6.33.6-147.fc13.i686.PAE.img
Edit this entry to remove the "rhgb quiet" argument. Then, if this happens again, and the system is rebooting, it will freeze in a state where you can (hopefully) see where it's frozen right there in the boot messages.
Under what circumstances would that splash screen occur, when the system had already booted and I had logged in? I don't believe the system was idle for longer than a few minutes, and the screensaver/power management turns off the monitor after ten minutes, so I don't believe that was the problem either.
I can't think of any circumstance where it would appear on an already booted system that is actively being used so unfortunately, I can't answer this one.
On Wednesday, July 28, 2010 22:06:22 Alex wrote:
That's not really a BSOD, it's the Fedora boot splash screen. You can hit ESC and have the system print the text boot messages (uncovering them by clearing the blue screen).
Nope, it was completely unresponsive. No network, keyboard, or mouse. No ctrl-alt-bs.
Can you verify that the kernel is alive? Can you login from another machine using ssh? Does the machine respond to a ping?
If you use GDM/Gnome, ctrl-alt-bs is disabled, IIRC.
The one time I saw it happen, the blue splash screen appeared and I believe nearly immediately it became unresponsive.
Under what circumstances would that splash screen occur, when the system had already booted and I had logged in? I don't believe the system was idle for longer than a few minutes, and the screensaver/power management turns off the monitor after ten minutes, so I don't believe that was the problem either.
I would guess this is an X issue. And typically due to a bug in graphics drivers. What video hardware do you have? What drivers do you use?
When it happens again, I would take a look at /var/log/Xorg.0.log (if you can ssh from a remote machine) or /var/log/Xorg.0.log.old (if you cannot and have rebooted into runlevel 5 again). If X got hung, something there should give you a clue what happened. If you post the log here, maybe we can help more.
HTH, :-) Marko
Hi Alex,
I talked to the plymouth maintainer and he said this is a known bug; I don't know if the fix is in F13 yet but let me explain what's going on:
You've got two bugs. One is a kernel panic. The other is that when the kernel is panicking, rather than showing useful debug information, it is dumping whatever was last on the framebuffer (the plymouth splash screen) to the screen. This is a mistake that was since found in the relevant code.
I'm no expert on getting useful kernel panic debug info but I bet if you check /var/log/messages you might see some clues (usually a good bet when the system suddenly stops working like that; I'm sorry I didn't suggest it before.) I'm sure someone else on the list can also help you trap the kernel panic to figure out what is causing the kernel panic - that is beyond my skill set.
I don't have the bug number for the plymouth issue handy I'm afraid; but hopefully there's enough to go on that you or someone else here could find it. I don't know if there is a workaround or not.
I hope this helps, good luck
~m
Hi,
I talked to the plymouth maintainer and he said this is a known bug; I don't know if the fix is in F13 yet but let me explain what's going on:
You've got two bugs. One is a kernel panic. The other is that when the kernel is panicking, rather than showing useful debug information, it is
Good to know, thanks. It's just good to know there is a confirmed bug, so I'm confident it is being worked on, and a fix will be available in the near future.
I checked the logs after the reboot, and there wasn't anything useful from syslog or the X11 logs.
Thanks again, Alex
On 30 July 2010 14:31, Alex mysqlstudent@gmail.com wrote:
Hi,
I talked to the plymouth maintainer and he said this is a known bug; I don't know if the fix is in F13 yet but let me explain what's going on:
You've got two bugs. One is a kernel panic. The other is that when the kernel is panicking, rather than showing useful debug information, it is
Good to know, thanks. It's just good to know there is a confirmed bug, so I'm confident it is being worked on, and a fix will be available in the near future.
From what I understood from Máirín's response, the bug in plymouth is
worked upon. But there is no way to tell the reason for the kernel panic. I think you should still try to investigate the reason for that. I am sorry I don't have any suggestions for you to try. Luckily I have never faced a kernel panic myself. :(
I checked the logs after the reboot, and there wasn't anything useful from syslog or the X11 logs.
Thanks again, Alex
On Fri, 2010-07-30 at 14:48 -0700, suvayu ali wrote:
On 30 July 2010 14:31, Alex mysqlstudent@gmail.com wrote:
Hi,
I talked to the plymouth maintainer and he said this is a known bug; I don't know if the fix is in F13 yet but let me explain what's going on:
You've got two bugs. One is a kernel panic. The other is that when the kernel is panicking, rather than showing useful debug information, it is
Good to know, thanks. It's just good to know there is a confirmed bug, so I'm confident it is being worked on, and a fix will be available in the near future.
From what I understood from Máirín's response, the bug in plymouth is
worked upon. But there is no way to tell the reason for the kernel panic. I think you should still try to investigate the reason for that. I am sorry I don't have any suggestions for you to try. Luckily I have never faced a kernel panic myself. :(
You're right, Suyayu - the bug that's being worked on is the plymouth bug (well actually, I think it's a kernel bug, but it displays plymouth so plymouth is the visible part.) But that won't fix the core problem - the kernel panic - because we don't yet know what's causing Alex's system to panic in the first place.
I think kernel panics are usually because of hardware issues, device driver issues, or a genuine kernel bug (but I think the latter is by far more rare.) Do you have any hardware that might be suspect, Alex? Maybe a RAM module you recently swapped into the machine?
One tool that might be useful is called kdump. It's part of Fedora and it loads a second uncrashed kernel to capture data from the crashed kernel to help figure out what's going on. There's some instructions on how to use kdump in the Fedora wiki:
http://fedoraproject.org/wiki/Kernel/kdump
I would imagine if the output of running kdump after triggering a crash doesn't indicate the problem in an obvious way (e.g., might be obvious-ish if there was mention of the brand name of a piece of hw you're aware of in the output), I think the output would certainly make for a good bug report.
Hope this helps, ~m
On Tue, 2010-07-27 at 10:44 -0400, Alex wrote:
As if to say, "F"-you, the screen changed from my normal desktop, with Firefox, Azureus, a few terminals, and a few VMs running, to a solid blue screen with just the Fedora squiggly 'F' in the middle of the screen, and the whole computer locked up.
Completely dead. Catatonic. Unresponsive. Nothing in the logs.
What the hell happened?
This is really aggravating, because it has been one problem after another, and I'm beginning to rethink my decision on this desktop.
Lately I've been having random desktop lockups (i.e. I can't do anything), but I can still hear my mp3s playing. My mouse stops working on a whim, forcing me to use my keyboard to close apps and then switch to a pty to kill my messed up Gnome session. Sound stops working for no apparent, although I can bring it back easily enough by restarting pulseaudio.
Just a little while ago my desktop froze again and I had to do a hard power off to get it back.
I'm not sure if it's something flaky with my hardware or if it's Fedora 12 that's somehow hosed. My PC at work runs Fedora 12 flawlessly, so I'm starting to think my hardware is wonky, but I'm not yet convinced.
Regards,
Ranbir
On 08/05/2010 01:56 PM, Kanwar Ranbir Sandhu wrote:
On Tue, 2010-07-27 at 10:44 -0400, Alex wrote:
As if to say, "F"-you, the screen changed from my normal desktop, with Firefox, Azureus, a few terminals, and a few VMs running, to a solid blue screen with just the Fedora squiggly 'F' in the middle of the screen, and the whole computer locked up.
Completely dead. Catatonic. Unresponsive. Nothing in the logs.
What the hell happened?
This is really aggravating, because it has been one problem after another, and I'm beginning to rethink my decision on this desktop.
Lately I've been having random desktop lockups (i.e. I can't do anything), but I can still hear my mp3s playing. My mouse stops working on a whim, forcing me to use my keyboard to close apps and then switch to a pty to kill my messed up Gnome session. Sound stops working for no apparent, although I can bring it back easily enough by restarting pulseaudio.
Just a little while ago my desktop froze again and I had to do a hard power off to get it back.
I'm not sure if it's something flaky with my hardware or if it's Fedora 12 that's somehow hosed. My PC at work runs Fedora 12 flawlessly, so I'm starting to think my hardware is wonky, but I'm not yet convinced.
Are you using an Asus motherboard such as the M3N78-VM? There are known problems with the nVidia MCP78S-based USB ports on the Asus boards (not just the model above) that can cause random lock ups. This isn't a Linux-only thing--there are a number of reports that it does it with Winblows also. The dead giveaway is the mouse and/or keyboard freezing with X starting to suck up 100% of one of the CPU cores. I know--I've been bitten by it quite a lot recently.
There's no fix that I know of. It appears to be a hardware design issue. I have purchased a new motherboard (can't remember the make) that specifically does NOT use nVidia. It uses an AMD chipset instead of nVidia and an ATI video card. I haven't installed it yet, but will this weekend. Also will give me a chance to poke around with the ATI video drivers (fun, fun, fun!). ---------------------------------------------------------------------- - Rick Stevens, Systems Engineer, C2 Hosting ricks@nerd.com - - AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 - - - - "Doctor! My brain hurts!" "It will have to come out!" - ----------------------------------------------------------------------
-
Alex -
Andrew Haley -
Chris Rouch -
Christofer C. Bell -
Frank Murphy -
JB -
Marko Vojinovic -
Máirín Duffy -
Ranbir -
Rick Stevens -
Robert Myers -
Suvayu Ali