jnthn There be new 6guts: 6guts.wordpress.com/2016/12/09/com...ing-crash/ 00:03
'night, folks 00:04
timotimo <3 00:08
gnite jnthn
oh those headings 00:09
we're hearing of pulao (apparently another name for pilaw) for the first time, and it appears on first glance to just be risotto, but not using "risotto rice" 00:30
and apparently you put in like 20 different kinds of spice 00:34
00:35 pyrimidine joined 01:38 pyrimidine joined
TimToady has to be careful because sometimes those have wheat noodley things in 'em 02:07
02:26 pyrimidine joined 02:49 ilbot3 joined
JimmyZ jnthn: re 6guts new post, MVM_OSR_INLINE_DISABLE should be MVM_SPESH_INLINE_DISABLE? 03:03
03:59 pyrimidine joined 06:40 FROGGS joined
brrt oooh 06:42
new 6guts
06:47 domidumont joined 06:53 domidumont joined
FROGGS wow, what a journey :o) 07:51
brrt indeed 07:54
08:07 domidumont joined
brrt i'm wondering how you actually can reduce the exposed surface of an object like MVMFrame 08:10
08:46 zakharyas joined
jnthn JimmyZ: Correct that thinko; thanks :) 10:53
10:55 japhb joined
jnthn Uh, *corrected :) 11:07
heh :)
11:21 japhb joined 12:48 pyrimidine joined 12:55 japhb joined 13:22 japhb joined 13:28 pyrimidine joined 14:11 pyrimidine joined 14:13 TimToady joined
dogbert17 jnthn: are you still around? 14:44
.tell jnthn It looks as if github.com/rakudo/rakudo/commit/58...0f449f8a9e breaks t/spec/S17-procasync/no-runaway-file-limit.t. Is it ok with you if we revert that fix? 14:49
yoleaux2 dogbert17: I'll pass your message to jnthn.
15:02 pyrimidine joined 15:07 pyrimidine joined 15:09 pyrimidine joined
dalek arVM/even-moar-jit: 43723b4 | brrt++ | src/jit/linear_scan.c:
Synthetic tiles also take part in ordering

If they don't, a live range may become live later than it really is or be rejected earlier than it should, and we shouldn't have that.
15:23
17:49 domidumont joined 18:44 leedo_ joined
lizmat dogbert17 jnthn I've just reverted 58cdfd80b30444ba4755bb3 19:07
19:09 pyrimidine joined
dogbert17 lizmat++ 19:31
19:38 pyrimidine joined 19:54 zakharyas joined
nine timotimo: your commit fea28ed247c7756d4fe4313f329af91681b4a3a6 "implement throwpayload* in the JIT" breaks Inline::Perl5 :/ 20:03
babydrop saw that referenced elsewhere too 20:04
somethign with namespacing I think
babydrop rakes recent tickets
20:21 pyrimidine joined 21:03 pyrimidine joined
timotimo oh? crap! 21:47
nine: is that "just" bisected or also with jit disable and such? :S
nine timotimo: MVM_JIT_DISABLE=1 fixes the issue 21:50
21:51 pyrimidine joined
timotimo ;_; 21:52
and reverting the commit on top of the latest master also fixes it, i expect?
nine timotimo: confirmed 21:57
timotimo ;_; 21:58
feel free to push the revert, then
dalek arVM: 8627dd1 | niner++ | src/jit/graph.c:
Revert "implement throwpayload* in the JIT"

This reverts commit fea28ed247c7756d4fe4313f329af91681b4a3a6 which breaks Inline::Perl5's tests and apparently causes other issues, too.
22:00
timotimo what, other issues?
22:04 domidumont joined
babydrop oh, I totally forgot I went to rake for tickets..... 22:18
I think ugexe found some sort of namespace bug that bisected to that commit.
Oh, never mind. Ugexe's thing was this one: github.com/rakudo/rakudo/commit/ab...8acd5bfc71 22:21
And "the other" that I recall was pyrimidine bisecting the bug with Inline::Perl5 doh >:)
timotimo ok 22:37
23:03 pyrimidine joined 23:56 pyrimidine joined