[PATCH] Improve code in ctdb lock helper

[Martin Schwenke]

On Thu, 1 Dec 2016 21:26:58 +1100, Martin Schwenke <martin at meltin.net>
wrote:

> On Thu, 1 Dec 2016 10:48:38 +0100, Volker Lendecke <vl at samba.org> wrote:
> 
> > On Thu, Dec 01, 2016 at 04:42:02PM +1100, Martin Schwenke wrote:
> > > On Thu, 1 Dec 2016 15:33:09 +1100, Amitay Isaacs <amitay at gmail.com>
> > > wrote:
> > >   
> > > > Here are two patches that do the following:
> > > > 
> > > > 1. Drop the support for locking multiple databases since it's not required
> > > > any more.
> > > > 2. Add explicit unlocking of record/db when lock helper is exiting
> > > > 
> > > > Please review and push.  
> > > 
> > > Reviewed-by: Martin Schwenke <martin at meltin.net>  
> > 
> > NACK on the second one at least without confirmation that the
> > async signal-safety aspect has been validated :-)
> 
> OK.  I didn't think of that.  :-(
> 
> So this would be better if the signal handler set a flag and
> ctdb_wait_for_process_to_exit() was unrolled, with an additional check
> in the loop of the flag set by the signal handler?

I should always read all my unread mail before replying...  :-)

peace & happiness,
martin