typo in tdb_firstkey?

[Jeremy Allison]

On Thu, Apr 19, 2007 at 01:30:04PM -0700, Taj Khattra wrote:
> hi,
> 
> re: samba-4_0 revision r22117
> 
> should the F_RDLCK in tdb_firstkey():
> 
>    tdb->travlocks.lock_rw = F_RDLCK;
> 
> be F_WRLCK - the subsequent unlock in firstkey, and the lock/unlock in
> nextkey all use F_WRLCK?
> 
> in any case, i was wondering why we take a write-lock on a hash chain
> in firstkey/nextkey instead of a read-lock?  (it's only held briefly,
> so i don't know if it matters in practice)

Taj,

	I think you are correct on this. I've just fixed traverse
to use travlocks.lock_rw consistently. Please svn checkout and
review the change.

Thanks a *LOT* for noticing this,

Jeremy.