torture RAW-OPLOCK

[tridge at samba.org]

Peter,

 > unlink should generate a break
 > Acking in oplock handler
 > (158) wrong value for break_info.level  got 0x0 - should be 0x1

In this situation w2k3 does a "break to level II", whereas Samba3 does
a break to NONE. So as far as "matching the behaviour that windows
clients expect" Samba3 is doing the wrong thing.

I can't explain _why_ windows does a break to level II in this
situation, but if you test against w2k3 then you will see that it
definately does. I just checked a sniff carefully, and confirmed it.

So I am pretty sure the Samba4 RAW-OPLOCK test is correct (within the
definition of correct of "what w2k3 does").

Cheers, Tridge