On Mon, Apr 09, 2007 at 08:31:37AM -0400, Trond Myklebust wrote:
> On Mon, 2007-04-09 at 13:09 +0200, Jörn Engel wrote:
> > That surely doesn't make life any easier for filesystem developers, I
> > agree. From that point of view, all telldir cookies should end their
> > life at closedir time. For "rm -r" it would be sufficient if the nfs
> > client simply didn't seekdir at all. For "ls -lR", this would return
> > duplicate dentries.
>
> Please go read the NFS spec. The only thing an NFS client has in order
> to read a directory is a READDIR operation that in essence takes a
> filehandle and a cookie as its arguments. Unless the server is able to
> return the entire rest of the directory in one RPC reply, the client
> needs to send a second READDIR operation with a cookie from the previous
> READDIR operation. The server is expected to return cookies for _each_
> entry in the directory.
>
> That is a protocol limitation, not a client limitation.
<Groan>
And after quickly checking RFC 3010, I see this limitation hasn't been
lifted in NFSv4.
Speaking of which, right now ext3 doesn't know whether it's talking to
an NFSv2 or NFS v3/v4 server, so it's always passing a 32-bit cookie.
If NFSv3/v4 could use an explicit interface to request a 64-bit
cookie, instead of just relying on the f_pos field in the file handle,
we can reduce the chance of hash collisions when reading an ext3
directory significantly.
If there are 2 or 3 directory entries that have a hash collision,
would the NFS protocol allow the server to juggle things so that those
2-3 directory entries with the hash collision are sent back in a
single readdir RPC reply? Is it aceptable/legal to have multiple
entries in the same READDIR reply packet have the same cookie value?
- Ted
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
