Chris Mason <mason@suse.com> writes:
> >
> I think I see what you mean...perhaps we could do this by updating the
> stat data to reflect which blocks have actually been written. If we read
> a file that is incomplete, we have a choice of truncating off the bad
> data, or simply refusing to service any operations for that file. Or did
> you intend something else (my idea wouldn't be zero cost).
I guess he wants a truncation to zero bytes if the file was writen to,
but not closed yet when it crashed. For some apps like an ftp server
that makes sense, but I wouldn't want it on a general file system
(because I value my logfiles)
It also solves the problem of you seeing old data in a block after
a crash (this can happen with both reiserfs and ext2), so it may be
good for the paranoid.
Apparently AIX used to do or still does this for these security reasons.
-Andi
-
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/
This archive was generated by hypermail 2b29 : Fri Mar 31 2000 - 21:00:24 EST