On 13 Aug 2000, Kai Henningsen wrote:
>
> Incidentally, is there any *real* problem with the current VFS and user
> space filesystems, or is that a SMOP?
The current VFS should in many ways be a lot _easier_ for user-space
filesystems, simply because the VFS layer does almost all of the
synchronization for you.
However, I would expect there to be deadlocks. In particular, I would be
very surprised indeed if you couldn't deadlock the machine after having
done a shared writable mapping on a user-level filesystem. Add some memory
pressure, cause the user-level filesystem to want to allocate memory,
cause that memory allocation to want to write back a dirty page on the
user-level filesystem...
There might be other cases like that. I suspect that we might have a
read-only user-level filesystem without any real pain - and that would be
sufficient for the regular "show the contents of a tar-file" kind of
filesystem.
Linus
-
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 : Tue Aug 15 2000 - 21:00:30 EST