mach_thread_self() needs to be balanced with mach_port_deallocate()

[Nikodemus Siivola]

To quote http://www.gnu.org/software/hurd/gnumach-doc/Thread-Information.html

"mach_thread_self has an effect equivalent to receiving a send right for the thread port. mach_thread_self returns the name of the send right. In particular, successive calls will increase the calling task's user-reference count for the send right."

I have no reason to believe this is a specialty of hurd -- according to this

  http://thread.gmane.org/gmane.lisp.openmcl.devel/6885

thread Darwin does exactly the same thing.

[Alastair Bridgewater (alastair-bridgewater)]

Is this fixed by 1.0.46.25 - 1.0.46.31, or is there still more to do?

[Cyrus Harmon (ch-launchpad)]

Yes, I think .25-.31 fix this problem. If we see new port leakage, let's start a new bug.