On Mon, Apr 16, 2007 at 09:19:43AM +0200, Aurelien Jarno wrote: > Florian Lohoff a écrit : > > Hi, > > today rem.rfc822.org once again crashed as it did on the 11th. This time > > i captured the oops on the serial console. > > This could be related to the glibc 2.5. We have uploaded it on the 11th > and on the 14th, but the package is still marked as building. We got no > news from [EMAIL PROTECTED] > > That seems to be coherent with the fact ld.so is mentioned in dmesg. > Could you please verify that glibc was building at that time it crashes? > Also if you have access at ball could you please verify the problem is > the same on mips?
I guess the glibc build was ongoing:
drwxr-xr-x 5 buildd buildd 4096 Apr 14 21:32 glibc-2.5/
I guess we have discovered some kind of kernel exploit or at least DoS
as a kernel crash based on whatever the userspace does is unacceptable.
I was just observing the crashes and as i got more than "machine is
dead" i felt like better send it people. My hope was that Thiemo or
someone else knows about this and can tell whether this is fixed in
the current etch kernel.
> Note that glibc 2.5-1 builds and work correctly here on both my local
> mips and mipsel machines, it builds well also on casals.debian.org.
> Maybe a kernel upgrade could fix the problem.
Selfmade or Debian kernel ?
Flo
--
Florian Lohoff [EMAIL PROTECTED] +49-171-2280134
Those who would give up a little freedom to get a little
security shall soon have neither - Benjamin Franklin
signature.asc
Description: Digital signature
