On 03/06/16 15:09, George Dunlap wrote:
Ah -- right; it turns out that I wasn't actually using nanosleep()
inside that "nsleep() call, I was using pselect. (I think that's
because I had determined that would be the most portable, but I can't
remember why now.)
Ah, I thought you were just too lazy to type "nanosleep".
If you're using select for sleeps because portability, you're probably
living the 80's ;)
In any case, with nanosleep() and this patch, then I'm getting nice
small deltas:
Good to hear.
