> On Jun 18, Linus Torvalds wrote:
>
> > >[eastep@eastep-3 eastep]$ cat /proc/version
> > >Linux version 2.2.10 (eastep@eastep-3.mis.tandem.com) (gcc version
> > >egcs-2.91.66
> >
> > Does the corruption go away if you use a compiler that is not known to
> > miscompile the kernel, eg 2.7.2?
>
> I had at least two ext2fs corruptions plus two SIGSEGs due to bad bits
> in binaries in buffer cache (emacs/mutt) and one full lockup,
> all with 2.2.9/2.2.10 compiled with 2.7.2.3 (SuSE 6.1).
more data from this morning:
tried 2.2.10 again, using two simultaneous `make -j10 bzImage' build sessions.
the first run aborted with solid lock (again no kernel message, no more
VC switching or Alt+SysReq) after 2*6 kernel builds (1.5 hours).
there have been no compile errors for these 2*6 builds though.
the next run was more `successfull': 2*12 builds did not lockup, but
I got 7 aborts with SIGSEGV, gcc internal error etc., 4/3 per session.
compared with 2*27 perfect kernel builds with 2.2.5, now I'm pretty sure
it's not a hardware bug on my side.
Harald
PS: I'm even monitoring CPU temperature and CPU fan rotation, no single
problem or difference between bad 2.2.10 and good 2.2.5 runs showed up.
--
All SCSI disks will from now on ___ _____
be required to send an email notice 0--,| /OOOOOOO\
24 hours prior to complete hardware failure! <_/ / /OOOOOOOOOOO\
\ \/OOOOOOOOOOOOOOO\
\ OOOOOOOOOOOOOOOOO|//
Harald Koenig, \/\/\/\/\/\/\/\/\/
Inst.f.Theoret.Astrophysik // / \\ \
koenig@tat.physik.uni-tuebingen.de ^^^^^ ^^^^^
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/