Memory management bug in threaded runtime.

[Thomas Davie]

Hey,

I seem to have come across an interesting memory leak in GNUstep, I've attached 
a program to demonstrate the issue.  As you can see, as soon as you create a 
second thread, things go awry:

Here's my output on ubuntu:
2011-05-25 09:39:04.824 BDS[23379] Testing outside thread
2011-05-25 09:39:04.825 BDS[23379] Allocing A 0xb90b00
2011-05-25 09:39:04.825 BDS[23379] Releasing A 0xb90b00 (count before 1)
2011-05-25 09:39:04.825 BDS[23379] Dealloc A 0xb90b00
2011-05-25 09:39:04.825 BDS[23379] 
2011-05-25 09:39:04.825 BDS[23379] Testing inside thread
2011-05-25 09:39:04.826 BDS[23379] Allocing A 0xb60310
2011-05-25 09:39:04.826 BDS[23379] Releasing A 0xb60310 (count before 1)
2011-05-25 09:39:04.826 BDS[23379] 
2011-05-25 09:39:04.826 BDS[23379] Testing outside thread again
2011-05-25 09:39:04.826 BDS[23379] Allocing A 0xb87d60
2011-05-25 09:39:04.826 BDS[23379] Releasing A 0xb87d60 (count before 1)

test.m
Description: Binary data