libera/#sbcl - IRC Chatlog

9:09:39 karlosz looks like constraint propagation needs some more hacking around local call arguments

9:09:52 karlosz it doesn't handle array relation constraints

9:10:11 karlosz so recursive functions with array accesses can't eliminate bounds reliably

9:11:05 karlosz does anyone have a good harness for benchmarking compile times and stuff like CP?

9:11:12 karlosz where do people get their numbers from

9:39:20 aeth quicklisp or ASDF says how long a compilation takes if you ql:quickload with :verbose t so that's where I've personally always seen compilation time timed.

9:40:24 aeth e.g. "compilation finished in 0:00:02.961" for a file that takes almost 3 seconds

10:08:10 Josh_2 I am getting very odd behaviour using condition-wait. https://plaster.tymoon.eu/v/7TQCU1KY2 I am not passing in a timeout, it is :timeout nil and the mutex above is also :timeout nil :wait-p t

10:12:52 Josh_2 I can try and adjust it so that the timeout key is not passed at all

10:21:28 aeth Oh, oops, what I said is probably a side effect of using package-inferred-system in ASDF.

11:36:50 Josh_2 Reverted my code back to bt2 and now I dont have this mystery timeout

11:54:26 glenneth_ ** NICK glenneth

12:22:50 Josh_2 Shot in the dark but bt2 manually reacquires the mutex if condition-wait returns nil

12:23:56 Josh_2 In my stacktrace it looks like the lock is not held

13:00:36 Josh_2 It was not that. Oh well