Trying to track down my remaining lockup problem with Linux using my
first AMD AM2 motherboard. It's been a nightmare, but it's getting
better. but I think it might be a Linux bug. But I have some specific
info that might lead to something.
This motherboard use nVidia GeForce 6150 chipset
The computer seems to only freeze up during disk IO - specifically when
downloading using rsync.
It also seems to lock up when writing data to a RAID 0 EXT3 filesystem.
I'm using software raid.
I'm currently running the stock Fedora Core 6 kernel based on 1.6.18.1
After several lockups I tried reformatting the partition to see if
perhaps the data was so screwed up that it was the problem. The reformat
didn't fix it.
It also seems that when it does crash that recovery isn't as clean as it
usually is. It seems that I have to run e2fsck manually more often than
usual and that it find more things to fix. Other drives that are not
part of raid seem to not have this issue.
I hope this is enough information to help track this down.
On Mon, 2006-10-30 at 10:10 -0800, Marc Perkel wrote:
> I hope this is enough information to help track this down.
I doubt it - someone would need to replicate your exact hardware and
config to debug this further, and you don't provide the information
needed to do so.
Try to reproduce the lockups from a text console instead of X - you
might get an Oops on the screen.
Lee
On Mon, Oct 30, 2006 at 10:10:20AM -0800, Marc Perkel wrote:
> Trying to track down my remaining lockup problem with Linux using my
> first AMD AM2 motherboard. It's been a nightmare, but it's getting
> better. but I think it might be a Linux bug. But I have some specific
> info that might lead to something.
>
> This motherboard use nVidia GeForce 6150 chipset
> The computer seems to only freeze up during disk IO - specifically when
> downloading using rsync.
> It also seems to lock up when writing data to a RAID 0 EXT3 filesystem.
> I'm using software raid.
>
> I'm currently running the stock Fedora Core 6 kernel based on 1.6.18.1
A stock ftp.kernel.org kernel with as many debug options as possible
turned on might be a better choice.
> After several lockups I tried reformatting the partition to see if
> perhaps the data was so screwed up that it was the problem. The reformat
> didn't fix it.
>
> It also seems that when it does crash that recovery isn't as clean as it
> usually is. It seems that I have to run e2fsck manually more often than
> usual and that it find more things to fix. Other drives that are not
> part of raid seem to not have this issue.
>
> I hope this is enough information to help track this down.
Unfortunately, it doesn't help at all.
Please send a complete "dmesg -s 1000000".
Please follow Lee's suggestion and report any messages you see (a
photo with a digital camera is OK).
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
On Mon, 2006-10-30 at 10:10 -0800, Marc Perkel wrote:
> Trying to track down my remaining lockup problem with Linux using my
> first AMD AM2 motherboard. It's been a nightmare, but it's getting
> better. but I think it might be a Linux bug. But I have some specific
> info that might lead to something.
boot always with report_lost_ticks
try boot with notsc
--
S?rgio M.B.