Add an undocumented option for forcing busy waits.

Some people complained about the usage of non busy waits:

  * I was told that there's an input lag with nanosleep(). I still doubt
    that, but since the problem is easy to solve...
  * Some Intel CPUs throttle the GPU if the selected CPU pstate is too
    low. This is especially a problem on Haswell mobile CPUs. Keeping
    a core busy works around that.

src/backends/unix/main.c

@@ -39,7 +39,10 @@ main(int argc, char **argv)
 	int time, oldtime, newtime;
 	int verLen, i;
 	const char* versionString;
+
+#ifndef BUSY_WAIT
 	struct timespec t;
+#endif
 
 	/* register signal handler */
 	registerHandler();
@@ -131,9 +134,11 @@ main(int argc, char **argv)
 		/* find time spent rendering last frame */
 		do
 		{
+#ifndef BUSY_WAIT
 			/* Sleep 10 microseconds */
 			t.tv_nsec = 10000;
 			nanosleep(&t, NULL);
+#endif
 
 			newtime = Sys_Milliseconds();
 			time = newtime - oldtime;