Occurance: Iīve copied batches of big files (3-100MB each, 1-1.5GB
together) over several file systems, the processes accessing the harddisk
at one time hung up -> means they didnīt return from the kernel call, so
I wasnīt able to kill them (even with -9!!), the load average kept at a
niveau about 3.1 - 3.7 and I wasnīt able to umount the file systems
(because these processes had open files).
A similiar error occured while grabbing a CD-Audio with cdda2wav (the
grabbing stopped for a split of a second, because I had high load on
other programs). It only seems to occur on high load produced by few tasks.
Filesystems:
NFS (on eth0, ne2xxx, module) -> ext2 (EIDE, static)
iso9660 (on ATAPI-CDRom) -> ext2 (same as above)
audio-grab (cooked_ioctl, ATAPI-CDR) -> ext2 ("")
I am no VFS expert but I guess itīs a race condition in the generic level
(VFS).
If you need more info or want to check patches: send me email directly
(currently Iīm not on the list, my mailbox canīt handle that high traffic).
If there is a patch: please mail me too!
happy hacking,
Konrad Rosenbaum
-
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/