On Wed, 2010-01-06 at 13:37 -0500, Dale Worley wrote:
> As you all know, an extremely convenient way to seize a semaphore is to
> use an OsLock object:
>
> ...
> {
> OsLock lock(semaphore);
>
> ...do stuff, protected by 'semaphore'...
> }
> ...
>
> But sometimes this paradigm is awkward because at the end of "do stuff",
> you want to call another function, and you need to release 'semaphore'
> before you call that function. Of course, you can have the protected
> code set a flag variable, and then do the call outside of the block with
> the OsLock, but that is hard to read, and the coding gets awkward. (It
> gets even messier when the arguments to the function "should" be
> temporary variables declared in nested blocks.)
>
> I've coded an addition to OsLock that allows the code to "prematurely"
> release the lock, before the OsLock object is destroyed. The paradigm
> is:
>
> ...
> {
> OsLock lock(semaphore);
>
> ...do stuff, protected by 'semaphore'...
> lock.release();
> // 'semaphore' is free now.
> external_method();
> }
> ...
>
> Of course, you have to check that flow of control from the call
> of ::release() to the end of the block declaring the OsLock does not
> require 'semaphore' be held.
>
> The current code for this change is available as:
> svn diff -c 17614
> http://sipxecs.sipfoundry.org/rep/sipXecs/branches/user_dworley_subscribe/sipXportLib
>
> Comments?
Why not just use the underlying OsBSem in the first place? It's exactly
what you just did - the only advantage of OsLock is that it releases
based on the scope.
_______________________________________________
sipx-dev mailing list [email protected]
List Archive: http://list.sipfoundry.org/archive/sipx-dev
Unsubscribe: http://list.sipfoundry.org/mailman/listinfo/sipx-dev
sipXecs IP PBX -- http://www.sipfoundry.org/
