Re: Most programs crash under openmosix

On Thu, Mar 16, 2006 at 08:26:04AM -0500, CSights wrote:
> Hi Gert,
> 	A problem simular to this has been reported before on the openmosix list:
> 	A summary:  Someone using (non-official gentoo) kernel version 2.4.27 was 
> getting virtual pointer problems using perl.
> 	Tab (of 2.6 fame) responded 'that was supposed to be fixed in 2.4.26'.
> 
> http://sourceforge.net/mailarchive/message.php?msg_id=10329688
> http://sourceforge.net/mailarchive/message.php?msg_id=10329687
> 
> 	I also use gentoo and am having "kernel NULL pointer dereference at virtual 
> address" problems with the old official 2.4.24 kernel, but it only happens 
> with one version of a C++ program.  The same program works when I use 
> different parameters, and other people are using the cluster with no problem 
> so I have been putting off trying to track down what the exact problem is.

 Do you have libc6-i686 installed (or whatever gentoo calls it)?  On a
Mandrake cluster node, I had to move aside /lib/tls to stop perl from
segfaulting with oM.  You should have:
clusternode:~$ ldd /bin/cat
        libc.so.6 => /lib/libc.so.6 (0x40028000)
        /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)

Not (as on my non-OM desktop):
peter@frisbee:~$ ldd /bin/cat
        linux-gate.so.1 =>  (0xffffe000)
        libc.so.6 => /lib/tls/i686/cmov/libc.so.6 (0xb7e32000)
        /lib/ld-linux.so.2 (0xb7f78000)

 And I could never get 2.4.26 to be stable.  I think I remember compiling it
up and PXE booting two machines with that kernel and a CHAOS ramdisk, and
running the infinite-loop testapp on one crashed the other one.

 The only oM kernel that's ever been stable on my dual Athlon and dual
Opteron machines is Tab's 2.4.22-om20031215 (which has the pipe-bug fix).
The official 2.4.24 is running stably on a non-SMP cluster of 10 cheap P4 
machines, though.  Maybe I've always had some other thing going wrong that
caused problems whenever I've tried other kernels.  (maybe filesystems like
jfs, or reiserfs?)  I'm probably getting overly pessimistic about oM...

-- 
#define X(x,y) x##y
Peter Cordes ;  e-mail: X(peter@cor , des.ca)

"The gods confound the man who first found out how to distinguish the hours!
 Confound him, too, who in this place set up a sundial, to cut and hack
 my day so wretchedly into small pieces!" -- Plautus, 200 BC


-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642