[gentoo-user] X Freezes With Firefox on Many Post 2.6.38 Kernels

On Wednesday 27 July 2011 17:18:19 James Wall wrote:

On Wed, Jul 27, 2011 at 4:03 PM, Dale wrote:
> Here is a update. Let's see what folks think about this situation. I
> mentioned in another thread that I did a from scratch kernel. It was a
> .35 version. It seemed to work fine, for a while. When I tell
> Seamonkey to download to my desktop, it works fine. The minute I tell
> it to save it to my large 750Gb drive, I get a kernel panic. Keep in
> mind, there is nothing OS related on that drive. Nothing OS at all.
> It is videos, CD ISO's and such as that.
>
> Here is another thing I just found out. I did download a few videos I
> wanted to save. They were on my desktop and who likes desktop clutter.
> So, I dragged them over to the large data drive. I did this by
> dragging from the desktop to a open Konqueror window. This was not
> downloading or anything, just a straight move operation. It copied a
> few Mbs and panic. This had nothing to do with Seamonkey either.

This looks like a drive/cable issue, since it only occurs on the one
drive. If both drives are SATA, I would try swapping the cables to
rule out a bad cable. If the problem stays with the drive I would
first try a different SATA port to see if that clears up the issue.

I would also check that all the cables are plugged in properly and that there
is nothing conductive (like metal) touching the drive where it really
shouldn't.
Maybe open the case, take the drive out and put it on a big sturdy cardboard
box to avoid possible shorts.

> So, did this issue just move from a Seamonkey sort of problem to
> completely something else? Hmmmmm. After the crash, I boot to single
> user mode. I ran resierfsck --fix-fixable on the drive. Not one
> error.

Did you do that on a mounted drive? I would first try a filesystem check before
using that command.

> I ran the smart thingy and not one error there either.

Did you force the short and long tests to be run and waited for them to be
finished? On a large drive, the long test can easily take several hours
(without any indication of how far it actually is)

> Thinking file system is bad in the kernel, well my /home directory is
> on reiserfs too. It is the one that works.

If it were the reiserfs implementation, the issue would be more common.

> Now, what the heck is this about? Does this make sense to anyone?

It does, there is something wrong with that drive.

Another thing you could try is to plug that drive into a different machine (I
believe you still have your old one?) and see if the same issue occurs there.
Also, now would be a good time to have backups of the data on that drive :)

Joost

Joost Roeleveld wrote:

On Wednesday 27 July 2011 17:18:19 James Wall wrote:

On Wed, Jul 27, 2011 at 4:03 PM, Dale wrote:

Here is a update. Let's see what folks think about this situation. I
mentioned in another thread that I did a from scratch kernel. It was a
.35 version. It seemed to work fine, for a while. When I tell
Seamonkey to download to my desktop, it works fine. The minute I tell
it to save it to my large 750Gb drive, I get a kernel panic. Keep in
mind, there is nothing OS related on that drive. Nothing OS at all.
It is videos, CD ISO's and such as that.

Here is another thing I just found out. I did download a few videos I
wanted to save. They were on my desktop and who likes desktop clutter.
So, I dragged them over to the large data drive. I did this by
dragging from the desktop to a open Konqueror window. This was not
downloading or anything, just a straight move operation. It copied a
few Mbs and panic. This had nothing to do with Seamonkey either.

This looks like a drive/cable issue, since it only occurs on the one
drive. If both drives are SATA, I would try swapping the cables to
rule out a bad cable. If the problem stays with the drive I would
first try a different SATA port to see if that clears up the issue.

I would also check that all the cables are plugged in properly and that there
is nothing conductive (like metal) touching the drive where it really
shouldn't.
Maybe open the case, take the drive out and put it on a big sturdy cardboard
box to avoid possible shorts.

I did check all the connections. I unplugged all the things drive
related, power and data, and everything looked fine. No dust, no
corrosion or anything that I could see. I did use a flashlight and a
magnifying glass to check. It could still be something I didn't see but
I looked.

So, did this issue just move from a Seamonkey sort of problem to
completely something else? Hmmmmm. After the crash, I boot to single
user mode. I ran resierfsck --fix-fixable on the drive. Not one
error.

Did you do that on a mounted drive? I would first try a filesystem check before
using that command.

I went to single user to run that. It didn't report fixing anything or
any other problems.

I ran the smart thingy and not one error there either.

Did you force the short and long tests to be run and waited for them to be
finished? On a large drive, the long test can easily take several hours
(without any indication of how far it actually is)

I did the long one. It ran while I took a nap. It does take a good
while to run. You nailed that one for sure. I just wonder how long a
3Tb drive would take. o_O I'm going to run it again tho. See if it
picks up on anything now.

Thinking file system is bad in the kernel, well my /home directory is
on reiserfs too. It is the one that works.

If it were the reiserfs implementation, the issue would be more common.

That's what I think too. It's also not the only partition that I use
reiserfs on either. I would think they would all have some sort of
weirdness if it was that. Then again, things tend to pick on me a LOT. :/

Now, what the heck is this about? Does this make sense to anyone?

It does, there is something wrong with that drive.

Another thing you could try is to plug that drive into a different machine (I
believe you still have your old one?) and see if the same issue occurs there.
Also, now would be a good time to have backups of the data on that drive :)

I may test that drive in my old rig. I have a SATA card in there.
Actually, it was originally in that rig. I did make backups of the
stuff I have room for. I just can't back up my video/audio files tho.

I'll post if anything changes or I get around to testing the drive in
the old rig. My garden and stuff sort of has me running. I picked
three 5 gallon buckets of okra yesterday. O_O

Dale

:-) :-)

Pardon me. My brain passed gas here. lol Could it be that my drives
file system has ran out of inodes or whatever they are called? That may
explain why I can't copy anything to it but it works fine as far as
reading goes.

Thoughts? How do I check/change it? Headed to some man pages too.

Dale

:-) :-)

On Thu, Jul 28, 2011 at 06:37:24PM -0500, Dale wrote:

Pardon me. My brain passed gas here. lol Could it be that my
drives file system has ran out of inodes or whatever they are
called?

That usually won't throw a kernel panic. That usually just gives an
error.

W
Willie W. Wong
Data aequatione quotcunque fluentes quantitae involvente fluxiones invenire
et vice versa ~~~ I. Newton

On Thu, Jul 28, 2011 at 7:37 PM, Dale wrote:

Pardon me.  My brain passed gas here.  lol  Could it be that my drives file
system has ran out of inodes or whatever they are called?  That may explain
why I can't copy anything to it but it works fine as far as reading goes.

Thoughts?  How do I check/change it?  Headed to some man pages too.

Stupid questions:
1) Is the filesystem mounted read-only?
2) Could the hard drive be in read-only mode? (Used to be there was
some flag you could trigger with hdparm to write-protect a hard drive.
Never poked that flag myself, I just remember that it was there.)

:wq