On Mon, 2 Aug 2004, Tupshin Harper wrote:
> First, of all, it turns out that I can mount the loopback copy of this
> filesystem, though I'm not sure if the difference between that version
> and the original disk version is due to additional runs of xfs_repair
> or something else. Trying to mount the original still gives the
> generic "wrong fs type, bad superblock" message.
dmesg will tell you the real error.
> When I mount the loopback version there are lots of directories with
> leading slashes and if I try to copy almost anything off of it (even
> things without leading slashes), I get "Unknown error 990"
This is xfs shutting down; again, dmesg will tell you more.
> and repair gives thousands and thousands of errors like:
> entry at block 1078 offset 1792 in directory inode 58812988 has
> illegal name "/27523.": entry "/27524." at block 1078 offset 1816 in
> directory inode 58812988 references invalid inode 18374686479671623679
> clearing inode number in entry at offset 1816...
How did the filesystem go bad in the first place? What prompted
the repair?
> If at all useful, here's the very end of a full strace output:
Hm, lots of seemingly successful directory reads, followed by a failure
message...
Not sure on this one yet...
-Eric
|