Messages in this thread | | | From | "Stephen C. Tweedie" <> | Date | Thu, 27 Jan 2000 20:20:21 +0000 (GMT) | Subject | Re: i-node corruption -- who to contact? |
| |
Hi,
On Wed, 26 Jan 2000 15:52:03 -0700, Fred Christiansen <fredch@boi.hp.com> said:
> The tool suite creates two partitions on a disk, creates an ext2 file system > in the first partition, and then does file system testing (i.e., the 1st > partition) and raw device testing (to the 2nd partition). In the course > of file system testing, explicit fsck -y -f's are occasionally > performed.
On the live filesystem? There are no cache coherency guarantees for fsck on mounted filesystems.
Other than that, there were a couple of races fixed in the final 2.2.14 kernel release concerning interactions between block device access and filesystem access to the same device, so even a readonly "fsck -n" on a mounted device could cause the filesystem to get confused in certain cases. Can you reproduce the problems on 2.2.14 kernels?
--Stephen
- 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/
| |