[lkml]   [1996]   [Nov]   [17]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread

From: Trevor Johnson <>
>> tmp is a character device owned by user 34000 something in group
>> 48000 something, and the mode flags are essentially random. Being
> I've experienced this on a system which was shut down by using the power
> switch (people were used to DOS). The mv utility might be usable, to move
> the corrupt file out of the way so it can be replaced.

Hmm.. I always shut down my system properly, and due to the one hour GPF
I never leave it online (hence I would know if the power went out on me).
I did not think of using mv at the time... Perhaps if it happens again I
can move it out of the way as suggested, but I will still be unable to
reboot, since e2fsck doesnt quite know how to fix the problem.

>> to know, but I am left locked out of my system with the grim
>> possibility that I will have to reinstall and wax everything on my
>> hd. Please dont take this email the wrong way... I appreciate
> I would boot from a floppy, run fsck, then mount the damaged partition(s)
> and take a look. If they were seriously damaged, I'd tar up their
> contents (or at least /etc/ and /home/) to the DOS partition, or tape, or
> whatever's available.
e2fsck was unable to fix the problem, and just asked me if I wanted tp
relocate the offending nodes. Without being able to repair the filesystem
I was also unable to mount it, and could not access my data... For the record,
e2fsck was telling me something to the effect that a node was out of it's
group and that relocating it was desirable but may result in "SEVERE DATA
LOSS". It was right, because after relocating it once, it relocated to the
same destination position over and over again in subsequent fs chk's. I
ended up taking the hit and reinstalled a fresh root partition.

>> motivated by pure frustration, but why is so much time being put
>> into 2.1.x when 2.0.x is not even stable? From what I hear,
> The 1.0 and 1.2 series stopped at 1.0.9 and 1.2.13, but we have 2.0.25 and
> Linus is continuing to maintain it this time. Lots of people are running
> 2..0 and whatever fixes they find will be folded in, but it's important to
> add new features too, thus 2.1.
Hopefully there will soon be a 2.0.26 to fix the ret_from_sys_call GPF too...

From: Albert Cahalan <>
>> I then proceeded to call a friend who, before my mention of my
>> problem, told me that he needed to reboot due to a seg faulting ps.
> Tell him (and yourself) to upgrade ps as described in
> /usr/src/linux/Documentation/Changes. Always RTFM first.
dont get me started. 'RTFM' implies idiocracy. I am an engineer. The FIRST
thing I do if something doesnt go right is read the instructions.

> Cool. Check your logs, then boot from a floppy and run fsck.
> Your logs may contain information about a dying hard disk.
Good idea if I was able to access the drive...

> The directory flag may be set. Try fsck first, then use the ext2
> filesystem editor debugfs found in the ext2tools package.
I did not have access to the debugfs prog at the time of repair, and would
probably not have enough knowledge of ext2 internals to use it correctly
if I had.

>> stupid thing I could have done, because I am now in Windows thanks
>> to a kernel panic that is keeping me out of my system. I went to
> You need to report that. Of course, a proper report would mean
> you look up the 8-digit hex numbers in I hope you
> have a copy of the file on your DOS disk.
Like I stated earlier, I could not access the drive and therefore could also
not access /var/log/*

>> go use my boot disk, and found out that Redhat no longer has
>> mount on their boot disk. How they pulled that off I would like
>> to know, but I am left locked out of my system with the grim
> Complain. They will get the idea eventually. I even provided a
I plan on it. I also plan on downloading slackware and using it for a while
until RedHat gets their quality back up. They used to have a good

>> quite a few people are having my same GPF problem when away from
>> the keyboard, ps dies on every system I know that runs 2.0.x,
> Every sys admin you know failed to read
> /usr/src/linux/Documentation/Changes before the 2.0 upgrade.
> Either that, or RedHat is more broken then I thought it was.
The problem wasnt with redhat, and since everyone I checked with installed
a fresh copy of RedHat 4.0, which complies with all the minimum versions
as stated in Documentation/Changes, you have no basis for your "RTFM"

 \ /
  Last update: 2005-03-22 13:37    [W:0.017 / U:0.532 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site