since .128 and up to .129-5 I habe been able to reliable reproduce the
what looks like a problem with teh scheduler by doing this:
In VC1 I start "dd if=/dev/hdc of=dev/null".
Then I switch to VC2 and start "find /usr/src/linux -type f -exec cp
{} /dev/null \;".
The process in VC1 always run fine but the process running in VC2
suddenly freezes and all disk activity stops. A top running on VC3
shows that system load goes way up (it was at 20+ after I left running
for ~9 hours two find's like the above).
Even after the dd in VC1 finishes the find keeps frozen.
One thing that I noticed is that switching VC's unfreezes the find for
a little while: it does something and then freezes again until the next
VC switch1.
I ran a strace on the find and it looks like it gets frozen at
wait4().
Anyone has seen this ugly behavior I am seeing here?
I am available as a guinea pig to track this down.
Hardware info.:
UP kernel (pre-2.1.129-5)
Pentium 75MHz
CD-ROM on /dev/hdc and hd on /dev/hdb.
Thanks,
peloy.-
-
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/