Search
22:06:14
stassats
you can unschedule it, except that unschedule-timer won't return
22:08:41
stassats
pfdietz: that's a lot of progv
22:15:09
stassats
pfdietz: 235 nested progvs now => 8 seconds 1,049,533,616 bytes, 1.3.14 => uhm
22:15:15
stassats
spinning for over a minute
22:15:41
stassats
102.201 seconds 8,954,912,224 bytes consed
22:25:11
stassats
it actually looks like my ir1-phases thingy actually improves this (but not as dramatically as compared to 1.3.14)
22:28:20
stassats
HEAD 3.937 / 1,041,508,528 bytes, before ir1-phases 14.636 / 8,088,860,176
22:28:39
stassats
(a different machine from the previous timings)
22:31:50
stassats
and it generates the same code (as far as i can tell)
22:32:50
stassats
(that was my motivation, not only fixing a bug but reducing redundant optimization passes)
22:37:43
stassats
AeroNotix: i don't really see why it should behave like this
22:37:48
stassats
but i'm not into timers
23:28:41
AeroNotix
stassats: it's pretty much how all cancellable timers I've used before work