[AD] Strange timer behavior in 5.2.1.1

[Edgar Reynaldo <edgarreynaldo@xxxxxxxxxx>]

Hi guys.

I made some binaries for Allegro 5.2.1.1 using MinGW 5.3.0 and Neil Roy 
used them to recompile his Pacman game and he noticed some very bizarre 
behavior that is associated with Allegro timers.

It is detailed here :
https://www.allegro.cc/forums/thread/616446/1024510#target

Additionally, ex_timer_pause behaves strangely, detailed here :
https://www.allegro.cc/forums/thread/616446/1024512#target

There were some changes introduced in Allegro 5.2.0 which altered timer 
behavior. According to the changelog, it says :

"Don't stop and join the timer thread when stopping the last timer 
(prevents unnecessary delay in this situation on some platforms)."

Neil stops the timers in his game when waiting for a keypress. I suspect 
the change in 5.2.0 is affecting this. The timers in his game don't 
appear to be stopping, rather, the timer events seem to be piling up 
while the timers are stopped.

As for ex_timer_pause, I'm not sure what's going on. It's very strange 
that none of the printf output gets shown in any test run of the program.

Edgar