Chris Mason wrote:
On Sat, 2008-06-21 at 11:35 +0200, Thomas Lindroth wrote:
Chris Mason wrote:
So, is the bittorrent program run every time you see the memory
corruption?
Could you please send along a few of the most common errors you see?
It's possible to recreate it with mv, cp and other tools without having
the torrent program running.
I've found a way to trigger the problem every time. If I copy a folder
with music videos to a freshly created FS it will always fail on the
same file when about 5.6G of data has been copied. The FS is 30G and
there is nothing special with that file. If I only copy that file to a
fresh FS everything works. I've also tried to run
dd if=/dev/zero of=test to a fresh FS and I was able to write 15G, but I
was not able to remove that file.
Something in your config is much less stable than my test boxes here.
Focusing on the dd test, it looks like random corruption in the FS
metadata. I see you've got gentoo i386 and a 2.6.24 kernel, are there
any other patches in that kernel?
-chris
The only thing that differs from a vanilla 2.6.24 kernel is a different
drivers/video/logo/logo_linux_clut224.ppm file. That should really not
affect anything.
Here is the .config I use: http://www.cyd.liu.se/~tholi945/config
I followed lavajoe's advice and tried the latest portage ebuild. It did
contain the hotfixes but it had the same problem as my own checkout.
I've tried running in UP mode by issuing
"echo 0 > /sys/devices/system/cpu/cpu1/online" but it made no difference.
This is probably not compiler related but to make sure I'm using gcc 4.1.2
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
