On 08/06/2012 11:14 PM, bugzilla-daemon@xxxxxxxxxxxxxxxxxxx wrote:
https://bugzilla.kernel.org/show_bug.cgi?id=45621
--- Comment #6 from markus.doits@xxxxxxxxxxxxxx 2012-08-06 22:14:42 ---
Just ran memtest for 3+ hours, 3 passes finished successfully. I know, longer
is better, but cannot run it longer atm (since it's a server with crucial
infrastructure). System is not overclocked or so, everything default.
memtest86 is completely useless at finding marginal memory. It'll only
typically find a problem if there is a properly, permanently dead or
stuck bit in the RAM.
If Windows is an option temporarily for testing use OCCT. If it isn't,
I've found something like this to work pretty well for finding CPU and
memory stability issues:
#####
# Assuming your /dev/shm is mounted on tmpfs - make sure it's big
# enough. Adjust count= accordingly
dd if=/dev/urandom bs=1M count=1024 of=/dev/shm/test
# Run one instance of this per CPU thread
#!/bin/bash
sum=`sha512sum /tmp/test`
while (/bin/true); do
echo -n .
sum2=`sha512sum /tmp/test`
if [ "$sum" != "$sum2" ]; then
echo “”
echo “Error detected. $sum != $sum2″
exit
fi
done
#####
That should be a lot better at finding CPU/cache/MCH/RAM issues.
Gordan
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
