is_locked optimization

[Jeremy Allison]

On Sun, Dec 12, 2004 at 09:56:50AM +1100, tridge at samba.org wrote:
> 
> the flush happens with the oplock still held. The client logically
> holds the oplock until it either acknowledges the oplock break from
> the server, or it closes the file. I'm not sure if this harms your
> planned optimisation, but it is at least worth thinking about.

The optimisation is only on read/write, not lock - lock requests still
get set on the file - so it shouldn't affect this. Until the client
releases the oplock it's still logically the only thing that can affect
the file state. To be completely sure and as we know when the oplock
break request has been sent we can turn this off once the break request
is sent.

> I'll be very interested to see if it helps. Maybe you could extend the
> loadfile in current versions of NBENCH to add more byte range locking?

Hmmm. Ok, once I've finished the cancelation tests and the large directory
code change I'll try and get to it !

Jeremy.