On 29/07/15 21:25, h.b.furuseth@usit.uio.no wrote:

(...)And copythr should set mc_new=0 before exiting also when setting mc_status. so copyfd1 won't just sit and wait.

I forgot it clears mc_new before setting mc_status. It's if it didn't keep the mutex it would have to set mc_new afterwards (after locking the mutex).