> > > - something clears the locked state without waking people up. Do you
> > > use "md" or anything else that plays around with buffers?
> Oh, I agree. I wasn't really serious, because the K5 bug would have to be
> obvious under other circumstances too (ie an out-of-order bug that does
> the wrong thing when interrupts happen).
I don't know if my experiences shed any light on this but I thought
that an additional datapoint might be helpful.
I setup a raid0 stripe on my machine at the university. INTEL
motherboard with dual PII-300's and 256Mbyte of RAM. Disk system is
SCSI with a Buslogic 958 driving a pair of 4Gbyte IBM drives. I
used two 300Mbyte partitions, on on each drive, to create the
composite filesystem, chunk size was 8K.
I NFS-mounted the second machine in my office to the SMP box. I am
using an SMC Etherpower II (EPIC) board in the SMP machine. I
basically did the following to replicate the /u filesystem.
cd /mnt/u # NFS volume
find ./dir1 ./dir2 | cpio -o | (cd /u; cpio -ivd)
The copy proceeded up to about the 43Mbyte point when I noticed the
machine slowing down significantly. Shortly thereafter the copy
halted. I was able to switch virtual consoles but that was about it.
The kernel was 2.0.33 compiled with SMP support. I finally had to hit
the reset button to get the machine back. I know that this isn't very
much data to go on and I don't even know if it is the same problem but
I thought that the data might be useful.
The machine isn't production yet so we could repeat the process
without damaging anything.
> Linus
Greg
}-- End of excerpt from Linus Torvalds
As always,
Dr. G.W. Wettstein Enjellic Systems Development - Specialists in
4206 N. 19th Ave. intranet based enterprise information solutions.
Fargo, ND 58102 WWW: http://www.enjellic.com
Phone: 701-281-1686 EMAIL: greg@wind.enjellic.com
------------------------------------------------------------------------------
"The software said it required Windows 3.1 or better so I installed Linux."
-- Mark MaClark