Of course for 2.1, but also for 2.0 as well... Right now, SSLMutex is pretty binary :)
+1 (looks reasonable, I assume it works :) )
if you think any of it needs more testing, just say so and maybe I can have time to play
slight cleanup that I should do in my own *&^$ code instead of bothering you about: replace %lu and (unsigned long) cast of getpid() with the use of APR_PID_T_FMT
this should clean up PR 8122
