On 7 Sep 2006, at 16:55, Wim Oudshoorn wrote:
Can you see any problems with this?
I don't think -retain needs to lock the gate, and as long as we are
only locking in the -release method as above, I think we can avoid
using a recursive lock (as I don't think we call -release inside any
region where we already lock the gate), but I may have missed
something.

I think you are right that the retain does not need the lock.
However, there needs to be a recursive lock because in the method:

- initWithReceivePort:sendPort:

the connection_table_gate locks the region:

  /* Here is the GNUstep version, which allows the delegate to specify
     a substitute.  Note: The delegate is responsible for freeing
     newConn if it returns something different. */
  if ([del respondsToSelector: @selector(connection:didConnect:)])
    {
      self = [del connection: parent didConnect: self];
    }

The comment suggest that usercode could well try to release the connection.

Thanks ....I agree that it seems unsafe to use a simple lock, so we do need to have a recursive one. Checking the source code, it seems I forgot to change back from recursive to non-recursive anyway, so the current code in svn should be OK even in the case that an application invalidates a connection during its creation.



_______________________________________________
Bug-gnustep mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/bug-gnustep

Reply via email to