Originally Posted by curaga
I'm not at all familiar with Wine details, but here's how it usually is.
Big X11 lock:
Every time they call any X function, lock a mutex (or other scalable construct) so that only one thread can do those at a time.
Replaced with a set of smaller locks, which scales better.
The same thing was done with the big kernel lock.
Thanks for the clarification....BTW, in what Kernel version was that also solved ?
I'm curious about that...