00:38 vendethiel joined 01:07 vendethiel joined 01:35 vendethiel joined
diakopter timotimo: yeah, seems like those would be optimizations done by a decent compiler 01:42
01:47 ilbot3 joined 04:32 vendethiel joined 04:51 geekosaur joined 06:24 domidumont joined 06:45 domidumont joined 06:49 domidumont joined 07:09 zakharyas joined 07:38 orbus joined 07:39 ggoebel16 joined 08:36 vendethiel joined 09:06 vendethiel joined
timotimo ah? well, that's nice :) 10:35
10:37 vendethiel joined 10:57 leont joined 11:30 vendethiel joined 11:36 brrt joined 11:59 vendethiel joined 12:51 vendethiel joined 13:43 dalek joined 14:13 vendethiel joined
diakopter ahoy 14:36
jnthn dobry den :) 14:40
diakopter jnthn: any suggestions where to debug the precomp test? 14:53
14:53 vendethiel joined
jnthn diakopter: Beyond doing a debug, non-optimized build and attaching VS to the hung procs? 14:53
diakopter hm, yeah 14:56
jnthn The same but on another platform, hopefully to avoid the hosed stack trace? 15:02
diakopter it seems fine on linux 15:14
I could try --no-jit, but I don't know if the precomp thing passes it through to the subprocess
timotimo i think it does copy over at least some arguments, like --stagestats i've seen show a bunch of results when something got precompiled, if i'm not mistaken 15:21
15:28 vendethiel joined 16:38 vendethiel joined 16:41 geekosaur joined 17:34 vendethiel joined 17:36 domidumont joined 18:25 Ven joined 18:43 geekosaur joined 21:59 lizmat joined 22:34 vendethiel joined 23:08 vendethiel joined 23:20 TimToady joined 23:23 cognominal joined 23:52 vendethiel joined
dalek arVM: 9e5b79f | coke++ | src/strings/decode_stream.c:
fix typo in comment
23:55