[mythtv] [PATCH] process cutlist without transcoding (final)

Isaac Richards ijr at po.cwru.edu
Wed Jul 9 11:33:13 EDT 2003

On Wednesday 09 July 2003 08:01 am, Geoffrey Hausheer wrote:
> On Wed, 9 Jul 2003 03:37:18 -0400, "Isaac Richards ijr-at-po.cwru.edu
> |mythtv/1.0-Allow|" <e6kbhbk35r0t at sneakemail.com> said:
> |
> > I'm applying this, but I made some changes to the transcoder thread
> > (database
> > connection locking related) that I'd like you to test before I enable the
> > thread.  Could you please make sure things work normally with the
> > additional
> > locking?
> Sure, I'll test it, but I don't exactly see what you are trying to
> achieve with the locking. Apparently, you are trying to prevent
> simultaneous db access, but with the way you implemented it, only access
> between a customEvent and the transcode thread are protected.  Wouldn't
> it be better to pass in a global mutex, and protect all of mythbackends
> threads from each other?  Or is that the master plan, and this is just
> one step on the way?

Every distinct database connection needs to protected against simultaneous 
access from different threads -- the transcode thread has its own connection, 
but that same connection was being used in that customEvent.  So, it doesn't 
need to be protected outside of that file, but just inside, so the 
customEvent doesn't do something with that db connection while the transcode 
thread is.


More information about the mythtv-dev mailing list