Kernel corrupts my reiserfs partition

Message

Skydive · Post by **Skydive** » Sun Feb 06, 2005 3:55 pm

Hi,

I have recently been trying to update from devfs to udev. I had to recompile my kernel with some other options (no devfs and stuff like that). I couldn't find my old kernel config so I started from scratch going through all the kernel options. Once my kernel and modules were compiled I rebooted and got the following kernel output:

Code: Select all

VFS: Mounted root (reiserfs filesystem) readonly.

and a bit further I got the following error:

Code: Select all

Warning: Unable to open initial console
Unable to find swap-space signature

I thought I had picked a few options that were incompatible with my hardware so I rebooted with my old (working) kernel image, but while booting I got this error:

Code: Select all

VFS: Cannot open root device "hdb1" or hdb1
Please append a correct "root=" boot option
Kernel panic: VFS: Unable to mount root fs on hdb1

So somehow my new kernel image had done something to my partition so that my old working kernel image couldn't boot anymore. There was no way for me to boot my system from the hard disk so I used the gentoo install CD to see what was wrong with my disk.
I entered the following command:

Code: Select all

reiserfsck /dev/hdb1

The command didn't complete, it said something about a problem at a certain node and that it was going to skip checking the entire subtree, which seemed to be almost my entire partition. I also got a message that said that the only way to solve the problem was to use the option --rebuild-tree, so I entered:

Code: Select all

reiserfsck /dev/hdb1 --rebuild-tree

This essentially solved the problem, but I went through this entire process two times and one time all my directories in "/" were moved over to lost+found and got some weird name made up of numbers.

Anyway, my question is how can something like this happen, what could be the cause of this? The new kernel image that I had made seemed to mount my partition in readonly mode and yet my partition had been modified while booting from that image.

Thanks in advance and let me know if I should post my kernel config file.

GungHo · Post by **GungHo** » Mon Feb 07, 2005 8:56 am

Hi Skydive,

I had nearly the same problem. A boot of a freshly compiled 2.6.10-r? completely blasted my / and /boot reiserfs partitions

.

As far as I can remember the kernel started booting, also the sysv init began running.

I needed some hours to repair Gentoo, had also to do a

reiserfsck --rebuild-tree

for this 2 partitions, got such #123456 style named directories and files, had to reinstall grub, even some binaries in /bin had been lost (the binary of mount, for instance), and I had to reemerge some corrupted packages.

A lot of nerving trouble

. Happy me I had a recent backup

With 2.6.9-r? some weeks ago I had some other probs, so I'm staying with 2.4.27.
Next time I'll give 2.6.x a chance, I'll boot it from diskette. And reserve some hours if the same trouble occurs again.

The first time I had such severe damage since I'm using Linux

Or do you think we should file a bugreport ?

Skydive · Post by **Skydive** » Tue Feb 08, 2005 5:47 pm

This was also the first time for me that I had such problems.
I recognize the problem you had with a corrupted 'mount' binary. I had the same problem with 'bash' so I copied it from the livecd and reemerged bash as soon as my system was up and running.

I'm running kernel 2.6.5 for quite a while now and I've never had any problems. I only had this problem with 2.6.9-gentoo-r13, but I didn't use the same config file.
I think any 2.6 kernel should work fine on my system as long as I pick the correct kernel correct options.

If the fact that the kernel is writing to disk (instead of only reading) while booting is not considered normal behavior then we should definately report a bug.

GungHo · Post by **GungHo** » Wed Feb 09, 2005 7:01 am

Hi Skydive,

you also had a corrupted mount binary ? Me too ! I think this could not be coincidence

. I also had corrupted binaries in the fileutils, procps, an other packages but that could also be related to the fsck, which was not able to fully recover all content of the corrupted filesystems.

For me it's also the first time that I had such an experience, and I surely have built kernels several hundred times in the past

.

My /boot and / partitions are both reiserfs, and /boot is far away from any limits which could cause trouble (e.g. 1024 cylinders, 8GB and the like), this kind of disklayout has worked for me for a long time, not only with grub, but also with lilo. It has always been a good idea to have a working backup kernel, but I have never had such severe corruptions of filesystems.

I think the kernel is indeed writing to disk, I cannot imagine another reason whether the partitions get corrupted. Maybe the mount binary or the whole process of the fsck'ing/mounting of all filesystems is related with this prob, I don't know.

When I'm browsing this forum, I'm unsure if I should give 2.6.10 another chance, there seems to be other trouble related to the 2.6er line of the kernel. Maybe it would be better to wait, until the kernel developers spawn a 2.7er line, and 2.6 further stabilizes. I'm not interested spending hours recovering heavily corrupted filesystems

I had the idea to boot a completely rebuilt kernel from floppy, but to be honest, it would not surprise me if this measure would not anticipate the situation which led to the corruption

DarkStorm_Inc · Post by **DarkStorm_Inc** » Wed Feb 09, 2005 1:33 pm

an isolated bug?? Ive been running 2.6.10 gentoo-dev-sources on a reiser mount with no problems since day one. hmmm... wonder how many people are having this problem..

MagnusBerg · Post by **MagnusBerg** » Wed Feb 09, 2005 3:13 pm

READ HOLE THE FUCKING MANUAL

I did not until yesterday but now i'm running Udev. I distroyed my reiserfs twice until I read ALL and understand that to do to make Udev work. Take a special look at point 3.1 before rebooting. And the textblock in point 2.5 before compiling the kernel. You must compile the kernel with udevsupport in the initrd.

Be aware that after reboot things in /dev change name or/and place so you need to change your mouseconfig in Xorg.conf and possible other things to. In my case radio and video.
Don't be afraid to dissable the tarball in /etc/conf.d/rc as said in point 2.5 but you can wait with this until you have rebooted and seen that Udev works. I had 63 radionodes and 63 videonodes in /dev before disabling the tarball.

(I shout to myself to yesterday becuase I was to lacy reading hole the manual and wasted many hours instead of reading it. And yesterday wasn't my first attempt to run Udew, I had distroyed my fs before

)

GungHo · Post by **GungHo** » Thu Feb 10, 2005 7:04 am

@DarkStorm_Inc,

I see the point, chances are indeed low, that there is a bug in the kernel. OTOH, chances are also low, that 2 people make the same mistake indepentdent from each other, and get nearly the same misbehaviour.

@MagnusBerg,

I know this document, I also know http://www.gentoo.org/doc/en/migration-to-2.6.xml which has been updated just a few days ago. End of last year I had a 2.6.9 kernel nearly running (APIC did not work, problem with NIC), I did the migration following the instructions of this guide. I have built the 2.6.10 kernel based on the config file of this nearly working 2.6.9 kernel, which used udev and not devfs, and not an initrd. And without filesystem corruption

Sigh

, think I'll have to redo the whole procedure again

GungHo · Post by **GungHo** » Thu Feb 10, 2005 8:35 pm

ok, I have again built a new 2.6..10-r4 kernel.

Good news is, my / an d/boot filesystems are alive

2.6er kernels could not be booted of floppy anymore ! Only harddisk is supported.

In the moment I'm dealing with the same prob as in http://forums.gentoo.org/viewtopic.php?t=290820. Lets see whether deactivating the "Advanced Partioning" helps

You can call me odysseus

GungHo · Post by **GungHo** » Thu Feb 10, 2005 9:22 pm

another bump !

The "Advanced Partitioning" definitively was the prob, that the kernel mentioned in the last message did not come up. Simply deactivate this option.

X11 works (but change the mouse device !)
Ethernet works (but not with the tulip driver, that has worked for me for ages !). But with the de4x5 driver.
Alsa works (but I'm a long time alsa user

)
My USB ports are recognized, do not know whether they work
XFS works, as does VFAT and ISO9660
3d acceleration of my old Voodoo 3/3000 works

Traffic shaping does not work (in this first attempt), as iptables also does not work. The modules can not be found. Hmmm.
My APIC (which has worked flawless for ages) is not activated, but could be manually (this is different than 2.6.9)
xawtv does not work, but that could be the tuner type or the like.
Lmsensors does not work. Maybe I have to emerge 2.9.0

My first impressions of 2.6.10-r4

So far, so good

.

Meanwhile I think that the kernel in my first 2.6.10 attempt was misconfigured, but I don't know exactly what I have done wrong. For this new kernel I have followed the instructions (of the 2.6 migration guide) very tightly. Now I have to get the missing features.

Skydive · Post by **Skydive** » Tue Feb 15, 2005 4:59 pm

Ok, I've reconfigured the kernel (version 2.6.10 this time) and it works fine now

I still thinks it's odd that the kernel can corrupt my partition in read-only mode. Even if I make a mistake configuring my kernel, it shouldn't touch my disk.

GungHo · Post by **GungHo** » Tue Feb 15, 2005 6:11 pm

Hi Skydive,

100% agree !!!

As told, I have built a lot of kernels over my time using linux, but I have never had such a severe corruption of my box !!

Happy Trails

Thom N2h · Post by **Thom N2h** » Wed Feb 16, 2005 10:10 pm

I think i have the same problem. Yet before doing something i need to be sure.
I did :

Code: Select all

reiserfsck --check /dev/hda5

No errors were return.
So Did you remember having trouble or corrupted files when using --check ?

GungHo · Post by **GungHo** » Thu Feb 17, 2005 6:20 am

Hi,

I believe I got an explicit message to do a

Code: Select all

reiserfsck --rebuild_tree

But if you have this prob, your filesystem is already corrupted, and you have no other chance than to do the rebuild_tree, or do a mkreiserfs on the corrupted filesystem. But then the content of the filesystem is away ! With rebuild_tree there is already a chance to get the data back

Thom N2h · Post by **Thom N2h** » Thu Feb 17, 2005 9:48 am

If i boot on a live CD They is no probem i can copy or delete data on my drive. So i don't know if it come from fs :-/
I do a bck for the moment tell me what you think about that

GungHo · Post by **GungHo** » Thu Feb 17, 2005 10:09 am

But when the filesystem is accessible, when you have booted from a CD, it cannot be corrupted !

Do you really see the normal content of the filesystem, when you boot from CD ?

Doing a backup is really a good idea

.

What have you done, when the problem came up the first time ? Has the filesystem ever been mountable ?

Thom N2h · Post by **Thom N2h** » Thu Feb 17, 2005 10:42 am

I did a an emerge world with xorg-6.8.2 and maybe udev 0.53. I need to look at emerge.log
I also used rfstool under windows to read my partitions. It was the first time.

Then i reboot. My file system is on read only (just reseirfs my ext3 partition is ok), so i can't lauch any init script or dhcpcd. anything.
THe fstab is not change. Nothing spécial in the log .

GungHo · Post by **GungHo** » Thu Feb 17, 2005 12:44 pm

Can you boot from the installed Linux (not from a CD !)

Can you then post the output of "mount" ?

You should describe the setup of your system, it's all a bit unclear

Thom N2h · Post by **Thom N2h** » Thu Feb 17, 2005 7:23 pm

GungHo wrote:Can you boot from the installed Linux (not from a CD !)

yes but I have a lot of errors cause to the ro fs.

Can you then post the output of "mount" ?

You should describe the setup of your system, it's all a bit unclear

I think i have udev 0.52

here my fstab :

Code: Select all

/dev/hda2		/boot		reiserfs	noauto,noatime,notail	1 2
/dev/hda5		/		reiserfs	noatime			0 1
/dev/hda6		none		swap		sw			0 0
/dev/hda7 		/home		ext3		noatime			0 0
/dev/hda8		/usr		reiserfs	noatime			0 0
/dev/cdroms/cdrom0	/mnt/dvdrom	iso9660		noauto,ro		0 0
/dev/cdroms/cdrom1	/mnt/cdrom	iso9660		noauto,ro		0 0
/dev/hda1		/mnt/hda1	ntfs		umask=2 		0 0
/dev/hda9		/mnt/hda9	ntfs		umask=2 		0 0
/dev/hda10              /mnt/hda10      ntfs            umask=2                 0 0
/dev/hda11              /mnt/hda11      ntfs            umask=2                 0 0
/dev/hda12            /var/tmp/portage  reiserfs        noatime                 0 0
/dev/sda1               /mnt/stick      vfat            umask=0,noauto          0 0

Here the lastest emerge :

Code: Select all

Started emerge on: Feb 12, 2005 15:58:23
1108223903:  *** emerge --verbose world
1108223904:  >>> emerge (1 of 18) rox-base/rox-2.2.0 to /
1108223904:  === (1 of 18) Cleaning (rox-base/rox-2.2.0::/usr/portage/rox-base/rox/rox-2.2.0.ebuild)
1108223905:  === (1 of 18) Compiling/Merging (rox-base/rox-2.2.0::/usr/portage/rox-base/rox/rox-2.2.0.ebuild)
1108223946: Started emerge on: Feb 12, 2005 15:59:06
1108223946:  *** emerge --fetchonly world
1108223948:  >>> emerge (1 of 18) rox-base/rox-2.2.0 to /
1108223948:  >>> emerge (2 of 18) sys-apps/gawk-3.1.4-r2 to /
1108223956:  >>> emerge (3 of 18) sys-devel/bc-1.06-r6 to /
1108223957:  >>> emerge (4 of 18) media-sound/xmms-1.2.10-r13 to /
1108223959:  >>> emerge (5 of 18) media-plugins/xmms-mikmod-1.2.10 to /
1108223960:  >>> emerge (6 of 18) media-plugins/xmms-alsa-1.2.10-r1 to /
1108223961:  >>> emerge (7 of 18) media-plugins/xmms-esd-1.2.10-r1 to /
1108223962:  >>> emerge (8 of 18) media-plugins/xmms-vorbis-1.2.10-r1 to /
1108223963:  >>> emerge (9 of 18) media-plugins/xmms-oss-1.2.10 to /
1108223964:  >>> emerge (10 of 18) x11-base/opengl-update-2.1.1-r1 to /
1108223964:  >>> emerge (11 of 18) sys-libs/readline-5.0-r2 to /
1108223976:  >>> emerge (12 of 18) x11-base/xorg-x11-6.8.2 to /
1108224010:  === (1 of 18) Post-Build Cleaning (rox-base/rox-2.2.0::/usr/portage/rox-base/rox/rox-2.2.0.ebuild)
1108224010:  >>> AUTOCLEAN: rox-base/rox
1108224016: === Unmerging... (rox-base/rox-2.1.5)
1108224017:  >>> unmerge success: rox-base/rox-2.1.5
1108224017:  ::: completed emerge (1 of 18) rox-base/rox-2.2.0 to /
1108224017:  >>> emerge (2 of 18) sys-apps/gawk-3.1.4-r2 to /
1108224017:  === (2 of 18) Cleaning (sys-apps/gawk-3.1.4-r2::/usr/portage/sys-apps/gawk/gawk-3.1.4-r2.ebuild)
1108224018:  === (2 of 18) Compiling/Merging (sys-apps/gawk-3.1.4-r2::/usr/portage/sys-apps/gawk/gawk-3.1.4-r2.ebuild)
1108224070:  === (2 of 18) Post-Build Cleaning (sys-apps/gawk-3.1.4-r2::/usr/portage/sys-apps/gawk/gawk-3.1.4-r2.ebuild)
1108224071:  >>> AUTOCLEAN: sys-apps/gawk
1108224076: === Unmerging... (sys-apps/gawk-3.1.4-r1)
1108224078:  >>> unmerge success: sys-apps/gawk-3.1.4-r1
1108224079:  ::: completed emerge (2 of 18) sys-apps/gawk-3.1.4-r2 to /
1108224079:  >>> emerge (3 of 18) sys-devel/bc-1.06-r6 to /
1108224079:  === (3 of 18) Cleaning (sys-devel/bc-1.06-r6::/usr/portage/sys-devel/bc/bc-1.06-r6.ebuild)
1108224079:  === (3 of 18) Compiling/Merging (sys-devel/bc-1.06-r6::/usr/portage/sys-devel/bc/bc-1.06-r6.ebuild)
1108224086:  >>> emerge (13 of 18) net-ftp/ftp-0.17-r4 to /
1108224087:  >>> emerge (14 of 18) sys-apps/net-tools-1.60-r11 to /
1108224090:  >>> emerge (15 of 18) media-tv/mythtv-0.17 to /
1108224095:  === (3 of 18) Post-Build Cleaning (sys-devel/bc-1.06-r6::/usr/portage/sys-devel/bc/bc-1.06-r6.ebuild)
1108224096:  >>> AUTOCLEAN: sys-devel/bc
1108224101: === Unmerging... (sys-devel/bc-1.06-r5)
1108224103:  >>> unmerge success: sys-devel/bc-1.06-r5
1108224103:  ::: completed emerge (3 of 18) sys-devel/bc-1.06-r6 to /
1108224103:  >>> emerge (4 of 18) media-sound/xmms-1.2.10-r13 to /
1108224103:  === (4 of 18) Cleaning (media-sound/xmms-1.2.10-r13::/usr/portage/media-sound/xmms/xmms-1.2.10-r13.ebuild)
1108224104:  === (4 of 18) Compiling/Merging (media-sound/xmms-1.2.10-r13::/usr/portage/media-sound/xmms/xmms-1.2.10-r13.ebuild)
1108224105:  *** terminating.
1108224126:  >>> emerge (16 of 18) sys-devel/prelink-20041123 to /
1108224222:  >>> emerge (17 of 18) media-libs/gstreamer-0.8.9-r1 to /
1108224222:  >>> emerge (18 of 18) dev-lang/perl-5.8.6-r3 to /
1108224222:  *** Finished. Cleaning up...
1108224222:  *** terminating.
1108225123: Started emerge on: Feb 12, 2005 16:18:43
1108225123:  *** emerge --verbose world
1108225124:  >>> emerge (1 of 15) media-sound/xmms-1.2.10-r13 to /
1108225124:  === (1 of 15) Cleaning (media-sound/xmms-1.2.10-r13::/usr/portage/media-sound/xmms/xmms-1.2.10-r13.ebuild)
1108225125:  === (1 of 15) Compiling/Merging (media-sound/xmms-1.2.10-r13::/usr/portage/media-sound/xmms/xmms-1.2.10-r13.ebuild)
1108225126:  *** terminating.
1108225140: Started emerge on: Feb 12, 2005 16:19:00
1108225140:  *** emerge --fetchonly world
1108225141:  >>> emerge (1 of 15) media-sound/xmms-1.2.10-r13 to /
1108225142:  >>> emerge (2 of 15) media-plugins/xmms-mikmod-1.2.10 to /
1108225143:  >>> emerge (3 of 15) media-plugins/xmms-alsa-1.2.10-r1 to /
1108225145:  >>> emerge (4 of 15) media-plugins/xmms-esd-1.2.10-r1 to /
1108225146:  >>> emerge (5 of 15) media-plugins/xmms-vorbis-1.2.10-r1 to /
1108225147:  >>> emerge (6 of 15) media-plugins/xmms-oss-1.2.10 to /
1108225148:  >>> emerge (7 of 15) x11-base/opengl-update-2.1.1-r1 to /
1108225148:  >>> emerge (8 of 15) sys-libs/readline-5.0-r2 to /
1108225148:  >>> emerge (9 of 15) x11-base/xorg-x11-6.8.2 to /
1108225149:  >>> emerge (10 of 15) net-ftp/ftp-0.17-r4 to /
1108225149:  >>> emerge (11 of 15) sys-apps/net-tools-1.60-r11 to /
1108225149:  >>> emerge (12 of 15) media-tv/mythtv-0.17 to /
1108225149:  >>> emerge (13 of 15) sys-devel/prelink-20041123 to /
1108225240:  >>> emerge (14 of 15) media-libs/gstreamer-0.8.9-r1 to /
1108225240:  >>> emerge (15 of 15) dev-lang/perl-5.8.6-r3 to /
1108225240:  *** Finished. Cleaning up...
1108225240:  *** terminating.
1108225480: Started emerge on: Feb 12, 2005 16:24:40
1108225480:  *** emerge --fetchonly world
1108225481:  >>> emerge (1 of 9) x11-base/opengl-update-2.1.1-r1 to /
1108225481:  >>> emerge (2 of 9) sys-libs/readline-5.0-r2 to /
1108225481:  >>> emerge (3 of 9) x11-base/xorg-x11-6.8.2 to /
1108225482:  >>> emerge (4 of 9) net-ftp/ftp-0.17-r4 to /
1108225482:  >>> emerge (5 of 9) sys-apps/net-tools-1.60-r11 to /
1108225482:  >>> emerge (6 of 9) media-tv/mythtv-0.17 to /
1108225482:  >>> emerge (7 of 9) sys-devel/prelink-20041123 to /
1108225490:  *** terminating.
1108225513: Started emerge on: Feb 12, 2005 16:25:13
1108225513:  *** emerge --fetchonly world
1108225514:  >>> emerge (1 of 8) x11-base/opengl-update-2.1.1-r1 to /
1108225514:  >>> emerge (2 of 8) sys-libs/readline-5.0-r2 to /
1108225514:  >>> emerge (3 of 8) x11-base/xorg-x11-6.8.2 to /
1108225515:  >>> emerge (4 of 8) net-ftp/ftp-0.17-r4 to /
1108225515:  >>> emerge (5 of 8) sys-apps/net-tools-1.60-r11 to /
1108225515:  >>> emerge (6 of 8) media-tv/mythtv-0.17 to /
1108225515:  >>> emerge (7 of 8) media-libs/gstreamer-0.8.9-r1 to /
1108225515:  >>> emerge (8 of 8) dev-lang/perl-5.8.6-r3 to /
1108225515:  *** Finished. Cleaning up...
1108225515:  *** terminating.
1108225523: Started emerge on: Feb 12, 2005 16:25:23
1108225523:  *** emerge --verbose world
1108225524:  >>> emerge (1 of 8) x11-base/opengl-update-2.1.1-r1 to /
1108225524:  === (1 of 8) Cleaning (x11-base/opengl-update-2.1.1-r1::/usr/local/portage/x11-base/opengl-update/opengl-update-2.1.1-r1.ebuild)
1108225524:  === (1 of 8) Compiling/Merging (x11-base/opengl-update-2.1.1-r1::/usr/local/portage/x11-base/opengl-update/opengl-update-2.1.1-r1.ebuild)
1108225526:  *** terminating.
1108225583: Started emerge on: Feb 12, 2005 16:26:23
1108225583:  *** emerge --verbose world
1108225584:  >>> emerge (1 of 8) x11-base/opengl-update-2.1.1-r1 to /
1108225584:  === (1 of 8) Cleaning (x11-base/opengl-update-2.1.1-r1::/usr/local/portage/x11-base/opengl-update/opengl-update-2.1.1-r1.ebuild)
1108225584:  === (1 of 8) Compiling/Merging (x11-base/opengl-update-2.1.1-r1::/usr/local/portage/x11-base/opengl-update/opengl-update-2.1.1-r1.ebuild)
1108225589:  *** terminating.
1108225655: Started emerge on: Feb 12, 2005 16:27:35
1108225655:  *** emerge --verbose world
1108225656:  >>> emerge (1 of 8) x11-base/opengl-update-2.1.1-r1 to /
1108225656:  === (1 of 8) Cleaning (x11-base/opengl-update-2.1.1-r1::/usr/local/portage/x11-base/opengl-update/opengl-update-2.1.1-r1.ebuild)
1108225657:  === (1 of 8) Compiling/Merging (x11-base/opengl-update-2.1.1-r1::/usr/local/portage/x11-base/opengl-update/opengl-update-2.1.1-r1.ebuild)
1108225675:  === (1 of 8) Post-Build Cleaning (x11-base/opengl-update-2.1.1-r1::/usr/local/portage/x11-base/opengl-update/opengl-update-2.1.1-r1.ebuild)
1108225676:  >>> AUTOCLEAN: x11-base/opengl-update
1108225681: === Unmerging... (x11-base/opengl-update-2.1_pre4)
1108225682:  >>> unmerge success: x11-base/opengl-update-2.1_pre4
1108225686:  ::: completed emerge (1 of 8) x11-base/opengl-update-2.1.1-r1 to /
1108225686:  >>> emerge (2 of 8) sys-libs/readline-5.0-r2 to /
1108225686:  === (2 of 8) Cleaning (sys-libs/readline-5.0-r2::/usr/portage/sys-libs/readline/readline-5.0-r2.ebuild)
1108225687:  === (2 of 8) Compiling/Merging (sys-libs/readline-5.0-r2::/usr/portage/sys-libs/readline/readline-5.0-r2.ebuild)
1108225719:  === (2 of 8) Post-Build Cleaning (sys-libs/readline-5.0-r2::/usr/portage/sys-libs/readline/readline-5.0-r2.ebuild)
1108225720:  >>> AUTOCLEAN: sys-libs/readline
1108225725: === Unmerging... (sys-libs/readline-5.0-r1)
1108225736:  >>> unmerge success: sys-libs/readline-5.0-r1
1108225737:  ::: completed emerge (2 of 8) sys-libs/readline-5.0-r2 to /
1108225737:  >>> emerge (3 of 8) x11-base/xorg-x11-6.8.2 to /
1108225737:  === (3 of 8) Cleaning (x11-base/xorg-x11-6.8.2::/usr/portage/x11-base/xorg-x11/xorg-x11-6.8.2.ebuild)
1108225737:  === (3 of 8) Compiling/Merging (x11-base/xorg-x11-6.8.2::/usr/portage/x11-base/xorg-x11/xorg-x11-6.8.2.ebuild)
1108226526: Started emerge on: Feb 12, 2005 16:42:06
1108226526:  *** emerge  search traceroute
1108226530:  *** terminating.
1108226567: Started emerge on: Feb 12, 2005 16:42:47
1108226567:  *** emerge --verbose xtraceroute
1108226567:  >>> emerge (1 of 4) net-analyzer/traceroute-1.4_p12-r2 to /
1108226567:  === (1 of 4) Cleaning (net-analyzer/traceroute-1.4_p12-r2::/usr/portage/net-analyzer/traceroute/traceroute-1.4_p12-r2.ebuild)
1108226569:  === (1 of 4) Compiling/Merging (net-analyzer/traceroute-1.4_p12-r2::/usr/portage/net-analyzer/traceroute/traceroute-1.4_p12-r2.ebuild)
1108226613:  === (1 of 4) Post-Build Cleaning (net-analyzer/traceroute-1.4_p12-r2::/usr/portage/net-analyzer/traceroute/traceroute-1.4_p12-r2.ebuild)
1108226614:  >>> AUTOCLEAN: net-analyzer/traceroute
1108226615:  --- AUTOCLEAN: Nothing unmerged.
1108226615:  ::: completed emerge (1 of 4) net-analyzer/traceroute-1.4_p12-r2 to /
1108226615:  >>> emerge (2 of 4) net-misc/host-991529 to /
1108226615:  === (2 of 4) Cleaning (net-misc/host-991529::/usr/portage/net-misc/host/host-991529.ebuild)
1108226616:  === (2 of 4) Compiling/Merging (net-misc/host-991529::/usr/portage/net-misc/host/host-991529.ebuild)
1108226640:  === (2 of 4) Post-Build Cleaning (net-misc/host-991529::/usr/portage/net-misc/host/host-991529.ebuild)
1108226641:  >>> AUTOCLEAN: net-misc/host
1108226641:  --- AUTOCLEAN: Nothing unmerged.
1108226641:  ::: completed emerge (2 of 4) net-misc/host-991529 to /
1108226641:  >>> emerge (3 of 4) x11-libs/gtkglarea-1.2.3-r1 to /
1108226641:  === (3 of 4) Cleaning (x11-libs/gtkglarea-1.2.3-r1::/usr/portage/x11-libs/gtkglarea/gtkglarea-1.2.3-r1.ebuild)
1108226642:  === (3 of 4) Compiling/Merging (x11-libs/gtkglarea-1.2.3-r1::/usr/portage/x11-libs/gtkglarea/gtkglarea-1.2.3-r1.ebuild)
1108226724:  === (3 of 4) Post-Build Cleaning (x11-libs/gtkglarea-1.2.3-r1::/usr/portage/x11-libs/gtkglarea/gtkglarea-1.2.3-r1.ebuild)
1108226725:  >>> AUTOCLEAN: x11-libs/gtkglarea
1108226725:  --- AUTOCLEAN: Nothing unmerged.
1108226725:  ::: completed emerge (3 of 4) x11-libs/gtkglarea-1.2.3-r1 to /
1108226725:  >>> emerge (4 of 4) net-analyzer/xtraceroute-0.9.1 to /
1108226725:  === (4 of 4) Cleaning (net-analyzer/xtraceroute-0.9.1::/usr/portage/net-analyzer/xtraceroute/xtraceroute-0.9.1.ebuild)
1108226727:  === (4 of 4) Compiling/Merging (net-analyzer/xtraceroute-0.9.1::/usr/portage/net-analyzer/xtraceroute/xtraceroute-0.9.1.ebuild)
1108226787:  === (4 of 4) Updating world file (net-analyzer/xtraceroute-0.9.1)
1108226787:  === (4 of 4) Post-Build Cleaning (net-analyzer/xtraceroute-0.9.1::/usr/portage/net-analyzer/xtraceroute/xtraceroute-0.9.1.ebuild)
1108226788:  >>> AUTOCLEAN: net-analyzer/xtraceroute
1108226788:  --- AUTOCLEAN: Nothing unmerged.
1108226788:  ::: completed emerge (4 of 4) net-analyzer/xtraceroute-0.9.1 to /
1108226788:  *** Finished. Cleaning up...
1108226791:  *** exiting successfully.
1108226804:  *** terminating.
1108228208:  === (3 of 8) Post-Build Cleaning (x11-base/xorg-x11-6.8.2::/usr/portage/x11-base/xorg-x11/xorg-x11-6.8.2.ebuild)
1108228211:  >>> AUTOCLEAN: x11-base/xorg-x11
1108228216: === Unmerging... (x11-base/xorg-x11-6.8.1.904)
1108228226:  >>> unmerge success: x11-base/xorg-x11-6.8.1.904
1108228227:  ::: completed emerge (3 of 8) x11-base/xorg-x11-6.8.2 to /
1108228227:  >>> emerge (4 of 8) net-ftp/ftp-0.17-r4 to /
1108228227:  === (4 of 8) Cleaning (net-ftp/ftp-0.17-r4::/usr/portage/net-ftp/ftp/ftp-0.17-r4.ebuild)
1108228227:  === (4 of 8) Compiling/Merging (net-ftp/ftp-0.17-r4::/usr/portage/net-ftp/ftp/ftp-0.17-r4.ebuild)
1108228245:  === (4 of 8) Post-Build Cleaning (net-ftp/ftp-0.17-r4::/usr/portage/net-ftp/ftp/ftp-0.17-r4.ebuild)
1108228245:  >>> AUTOCLEAN: net-ftp/ftp
1108228250: === Unmerging... (net-ftp/ftp-0.17-r3)
1108228252:  >>> unmerge success: net-ftp/ftp-0.17-r3
1108228252:  ::: completed emerge (4 of 8) net-ftp/ftp-0.17-r4 to /
1108228252:  >>> emerge (5 of 8) sys-apps/net-tools-1.60-r11 to /
1108228252:  === (5 of 8) Cleaning (sys-apps/net-tools-1.60-r11::/usr/portage/sys-apps/net-tools/net-tools-1.60-r11.ebuild)
1108228253:  === (5 of 8) Compiling/Merging (sys-apps/net-tools-1.60-r11::/usr/portage/sys-apps/net-tools/net-tools-1.60-r11.ebuild)
1108228277:  === (5 of 8) Post-Build Cleaning (sys-apps/net-tools-1.60-r11::/usr/portage/sys-apps/net-tools/net-tools-1.60-r11.ebuild)
1108228278:  >>> AUTOCLEAN: sys-apps/net-tools
1108228283: === Unmerging... (sys-apps/net-tools-1.60-r10)
1108228286:  >>> unmerge success: sys-apps/net-tools-1.60-r10
1108228286:  ::: completed emerge (5 of 8) sys-apps/net-tools-1.60-r11 to /
1108228286:  >>> emerge (6 of 8) media-tv/mythtv-0.17 to /
1108228286:  === (6 of 8) Cleaning (media-tv/mythtv-0.17::/usr/portage/media-tv/mythtv/mythtv-0.17.ebuild)
1108228287:  === (6 of 8) Compiling/Merging (media-tv/mythtv-0.17::/usr/portage/media-tv/mythtv/mythtv-0.17.ebuild)
1108228633:  *** terminating.
1108228944: Started emerge on: Feb 12, 2005 17:22:24
1108228944:  *** emerge --verbose unmerge mythtv
1108228949: === Unmerging... (media-tv/mythtv-0.16.20050115-r1)
1108228961:  >>> unmerge success: media-tv/mythtv-0.16.20050115-r1
1108228977:  *** exiting successfully.
1108228979:  *** terminating.
1108229187: Started emerge on: Feb 12, 2005 17:26:27
1108229187:  *** emerge --verbose world
1108229188:  >>> emerge (1 of 2) media-libs/gstreamer-0.8.9-r1 to /
1108229188:  === (1 of 2) Cleaning (media-libs/gstreamer-0.8.9-r1::/usr/portage/media-libs/gstreamer/gstreamer-0.8.9-r1.ebuild)
1108229188:  === (1 of 2) Compiling/Merging (media-libs/gstreamer-0.8.9-r1::/usr/portage/media-libs/gstreamer/gstreamer-0.8.9-r1.ebuild)
1108229384:  === (1 of 2) Post-Build Cleaning (media-libs/gstreamer-0.8.9-r1::/usr/portage/media-libs/gstreamer/gstreamer-0.8.9-r1.ebuild)
1108229385:  >>> AUTOCLEAN: media-libs/gstreamer
1108229390: === Unmerging... (media-libs/gstreamer-0.8.9)
1108229395:  >>> unmerge success: media-libs/gstreamer-0.8.9
1108229395:  ::: completed emerge (1 of 2) media-libs/gstreamer-0.8.9-r1 to /
1108229395:  >>> emerge (2 of 2) dev-lang/perl-5.8.6-r3 to /
1108229395:  === (2 of 2) Cleaning (dev-lang/perl-5.8.6-r3::/usr/portage/dev-lang/perl/perl-5.8.6-r3.ebuild)
1108229396:  === (2 of 2) Compiling/Merging (dev-lang/perl-5.8.6-r3::/usr/portage/dev-lang/perl/perl-5.8.6-r3.ebuild)
1108229794:  === (2 of 2) Post-Build Cleaning (dev-lang/perl-5.8.6-r3::/usr/portage/dev-lang/perl/perl-5.8.6-r3.ebuild)
1108229794:  >>> AUTOCLEAN: dev-lang/perl
1108229800: === Unmerging... (dev-lang/perl-5.8.6-r2)
1108229803:  >>> unmerge success: dev-lang/perl-5.8.6-r2
1108229803:  ::: completed emerge (2 of 2) dev-lang/perl-5.8.6-r3 to /
1108229803:  *** Finished. Cleaning up...
1108229805:  *** exiting successfully.
1108229805:  *** terminating.
1108246657: Started emerge on: Feb 12, 2005 22:17:37
1108246657:  *** emerge --verbose vlc
1108246657:  >>> emerge (1 of 1) media-video/vlc-0.8.1 to /
1108246657:  === (1 of 1) Cleaning (media-video/vlc-0.8.1::/usr/portage/media-video/vlc/vlc-0.8.1.ebuild)
1108246658:  === (1 of 1) Compiling/Merging (media-video/vlc-0.8.1::/usr/portage/media-video/vlc/vlc-0.8.1.ebuild)
1108246877:  *** terminating.

Some lines in dmesg :

Code: Select all


Linux version 2.6.10-gentoo-r6 (root@tpc) (version gcc 3.4.3 20050110 (Gentoo Linux 3.4.3.20050110, ssp-3.4.3.20050110


ReiserFS: hda5: found reiserfs format "3.6" with standard journal
ReiserFS: hda5: using ordered data mode
ReiserFS: hda5: journal params: device hda5, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: hda5: checking transaction log (hda5)
ReiserFS: hda5: Using r5 hash to sort names
VFS: Mounted root (reiserfs filesystem) readonly.
Freeing unused kernel memory: 172k freed


EXT3-fs warning: maximal mount count reached, running e2fsck is recommended
kjournald starting.  Commit interval 5 seconds
EXT3 FS on hda7, internal journal
EXT3-fs: mounted filesystem with ordered data mode.
ReiserFS: hda8: found reiserfs format "3.6" with standard journal
ReiserFS: hda8: using ordered data mode
ReiserFS: hda8: journal params: device hda8, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: hda8: checking transaction log (hda8)
ReiserFS: hda8: Using r5 hash to sort names
NTFS volume version 3.0.
NTFS volume version 3.0.
NTFS volume version 3.0.
NTFS volume version 3.0.
ReiserFS: hda12: found reiserfs format "3.6" with standard journal
ReiserFS: hda12: using ordered data mode
ReiserFS: hda12: journal params: device hda12, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: hda12: checking transaction log (hda12)
ReiserFS: hda12: Using r5 hash to sort names

GungHo · Post by **GungHo** » Thu Feb 17, 2005 9:30 pm

hmmmm, I do not see anything suspicious.... but that you have used an unstable release of gcc for building the kernel of your box. gcc-3.3.5-r1 is (in the moment) the newest stable one. I would never do such hazardous things. Maybe this has corrupted your box, maybe not. Who knows

Is your whole system built using this release of gcc ? You have also emerged an unstable release of udev (latest stable is 045) !

xorg 6.8.2.is also unstable, the latest stable is 6.8.0-r4. I think you are living at a knife's edge

. Do you overclock your box ?

cheers

Thom N2h · Post by **Thom N2h** » Fri Feb 18, 2005 3:32 am

hum I am on ~x86 so most of my software are unstable.
My system is overclock too

And some packages are compile with the lastest version of gcc. some old packages are compile with 3.3.5.

I need my system operational in 6 hours. I don't want to break it now
For the moment i think i will run my system with chroot during the next week. I need a blender render.

Then i will try to recompile my kernel with 3.3.5
Recompile all the system with 3.3.5 or with 3.4.3

Then if it does not work i will touch to the file system.
Whatever bck is ok

GungHo · Post by **GungHo** » Fri Feb 18, 2005 6:31 am

Hi Thom N2h,

well, I have seen that even your perl or your xmms are unstable. My personal opinion: maybe it's ok to use unstable releases for such nonrelevant components, it doesn't much matter when xmms dies, or produces distortion or whatever.

But for such vital components as the compiler (where the codequality of the whole system depends on) or the kernel or say glibc I would never use unstable releases, only if I'm a developer of such a component.

The same with overclocking. I know the argument of the overclockers "there are reserves in the system I want to use", but in todays CPU's are roughly about 50 million transistors, and noone can predict what happens when by overclocking such a highly integrated device a hotspot of say 100 overheated transistors results. What will then happen to the stability of your box ???? I can not tell. I think nobody can tell. But one thing is sure, if this happens for a longer period the overheated transistors will be permanently damaged and destroyed.

Going beyond the specs of a technical devices may result in permanent damage of this device. Doesn't matter whether the device is a tire on a car, or a screw, or an engine in a car or a CPU.

But thats my personal opinion, as I do not want to live on the bleeding edge

. I like to have a box as stable as possible, and do not want to spend time and/or money repairing it over and over again

Enough said

sawatts · Post by **sawatts** » Mon Feb 21, 2005 12:31 pm

I had similar issue yesterday, which is still unresolved.

Gentoo 2.6.10-r6 has been running happily on my system for nearly a week (after m/b replacement). I had got nearly everything working smoothly (notably mga_vid, and pwc -- webcam better than I ever got it to work on XP!). I even had my Logitech wireless desktop kdb/mouse ("Logitech Optical Desktop Delux SE" afair), with its various media keys -- though I did this by swapping the PS/2 leads on a running system [more...]

The last thing I was trying to sort out was 'autofs'. Somewhere on the forum it was mentioned that support for this had to be in the kernel (autofs 4?) rather than as a module (which I had), so recompiled the kernel to include this, and installed, shifting the previous module out of the way. I have rebuild the kernel a number of times to take particular support in/out/modularise as I have been building the system.

Rebooted, and first noticed that the Logitech keyboard was *really* slow to respond (X) -- press, wait, wait, try again... -- though it was working after a fashion. The mouse wasnt working at all. Swapping cables back to the old wired variety and it worked normally (no reboot or restart).

Then things started to really go wrong...

Logging into X, lm_sensors (via gkrellm), which had been reporting CPU temps accurately now showed >60C for both. Reboot and BIOS check reported the expected 35-40C, which I have been seeing consistently previously (system has been powered up for several days).

Then I started to get kernel panics... Possibly after a freeze/hardreboot from X (kernal panic report not visible due to X?).

Sometimes during boot, sometimes after (sticking to console for now).

I had to reiserfsck --rebuild-tree on "/usr" and "/opt" (afair). "/usr" was particularly bad -- I think about three cycles of '--check' and '--rebuild-tree' (including one time where the latter *failed* to complete!). I still have some corruption on "/var", but couldnt remount,ro at the time.

Strangely, rebooting from liveCD also panic'ed... I really hope that this is down to a dodgy CDR media (I have had problems with this batch in the past), and I will burn a new copy [2004.3] today on better media.

I am not sure if there is anything anyone can help with here yet, but I thought I'd just post my woes

PS: Logitech wireless keyboard/mouse seems not to initialise correctly on reboot. This uses PS/2 connections (or USB), so I would expect it behave like a normal kb/mouse (heck, you have to be able to use the k/b in BIOS for col). But while it works if I hot-replace my old wired ones, and sync the new kb/mouse to the wireless hub, on reboot the mouse doesnt work, the kb is virtually unusable. Previously, on Mandrake, when I tried to use them initially they seemed to behave randomly -- though the motherboard was on its way out at the time (as it turns out). ...but other problems take higher priority right now...

Ho hum...
Si.
[/list]

Machine:
MSI K7D Master-L
2 x Athlon MP 1900+
1 GB RAM
2 x 80 GB HDD
DVD, CDRW
Matrox G400 DH 32MB & Hercules PRO920 monitor [matroxfb, mga_vid (patched)]
Logitech Webcam 4000 (?) [pwc (usb-pwc-re) (excellent!)]
Regular KB/Mouse or Logitech wireless desktop w. optical mouse [PS/2]

GungHo · Post by **GungHo** » Mon Feb 21, 2005 3:57 pm

Hi sawatts,

I have seen your other post.

Do you overclock ?
Kernel compiled with an unstable compiler (

to last post) ?
Your cooler is in good contact with the surface of the CPU ?

Have you tried to go back to vanilla-sources, or a 2.4 kernel ?

I have used autofs for a long time (with 2.4), but never had to link it into the kernel, it always worked as a module. I prefer modularized kernels

. But I have not tried autofs with 2.6

In my case I never had kernel panics other than when rebooting (when the filesystems had been corrupted) or a overheated CPU. Meanwhile all works well (knock on wood) but not the fan rotations of lm-sensors. But thats all.

But I have not converted my box to NPTL. I do not fully trust the 2.6er kernels, and want to have an easy fallback to 2.4.

sawatts · Post by **sawatts** » Mon Feb 21, 2005 5:36 pm

GungHo wrote: Do you overclock ?
Kernel compiled with an unstable compiler ( to last post) ?
Your cooler is in good contact with the surface of the CPU ?
Have you tried to go back to vanilla-sources, or a 2.4 kernel ?

No overclocking.

I haven't changed the compiler from the default Gentoo one.

I believe that the sink & fans are in good contact -- but I will remount them tonight just to be sure. The machine was running (powered up) continually since about Wednesday (5 days), though obviously the load would be greater while I was working on it over the weekend. BIOS temperatures seemed ok, but they wouldn't reflect a bad CPU mount.

Only used the 2.6 kernels on this box. Will have to consider a fall-back kernel, but really want to build on the newer features (esp NPTL for SMP).

Kernel panics where mostly during boot, or shortly afterwards. Otherwise, some apps failed with errors such as "symbol ferr?r not found in libc" (note corrupted symbol name! Well scr?wed...).

Sadely enough, before the panic started lm-sensors where reporting correct fan and temperatures! (as per BIOS readings). I was briefly impressed...

Guess I know what I will be doing most evenings this week

GungHo · Post by **GungHo** » Mon Feb 21, 2005 6:32 pm

No overclocking.

I haven't changed the compiler from the default Gentoo one.

very good

!

I believe that the sink & fans are in good contact -- but I will remount them tonight just to be sure. The machine was running (powered up) continually since about Wednesday (5 days), though obviously the load would be greater while I was working on it over the weekend. BIOS temperatures seemed ok, but they wouldn't reflect a bad CPU mount.

checking this could not be wrong, but ... it may not be the cause of the prob

Only used the 2.6 kernels on this box. Will have to consider a fall-back kernel, but really want to build on the newer features (esp NPTL for SMP).

ok, for SMP boxes NPTL would be worth using.

Kernel panics where mostly during boot, or shortly afterwards. Otherwise, some apps failed with errors such as "symbol ferr?r not found in libc" (note corrupted symbol name! Well scr?wed...).

have also not seen kernel panics other than on boot

Sadely enough, before the panic started lm-sensors where reporting correct fan and temperatures! (as per BIOS readings). I was briefly impressed...

in the moment there are 26°C in the case and 37°C CPU (Athlon XP-2400+). When doing a kernel compile or an emerge or the like, it raises to over 40, sometimes 45°C, but rarely more in this season (but in summer !)

Guess I know what I will be doing most evenings this week

Good Luck !