Welcome to the main channel on the development of MoarVM, a virtual machine for NQP and Rakudo (moarvm.org). This channel is being logged for historical purposes.
Set by lizmat on 24 May 2021.
00:05 [Coke]_ left 00:07 [Coke] joined 01:28 ShaneC left
Nicholas good *. * 06:01
MasterDuke getting back into project euler and rakudo is significantly faster than when i was last doing PE, very nice 08:10
sort of related, but why would MVM_callstack_unwind_frame be near the top of a perf report? 08:13
lizmat I've got good news regarding github.com/rakudo/rakudo/issues/1920 08:39
the problem *is* caused by DYNAMIC being called from multiple threads at the same time 08:40
and an interaction between $*CWD and $*REPO
the test script now always succeeds
the only downside is that it broke pre-compiling modules 08:41
nine Well that _is_ a bit of a downer :D 09:05
lizmat yeah, but at least I got a nice stacktrace to work on fixing that -) 09:12
rather than just segfaults
jnthnwrthngtn MasterDuke: Assuming you mean inclusive time, then that's where compiling dispatch programs shows up 09:15
MasterDuke: Take a look at its callees to see if it's that or something else 09:16
timo who's ready for the next iteration of critical remote code execution vulnerabilities in Spring? 09:33
10:01 sena_kun left 10:02 sena_kun joined
jnthnwrthngtn Ready to patch them or ready to exploit them? :) 10:05
timo porque no los dos 10:10
jnthnwrthngtn :D 10:12
10:57 Altai-man joined
lizmat meh, very useful error message while installing core modules: 11:05
Use of uninitialized value of type Any in string context.
Methods .^name, .raku, .gist, or .say can be used to stringify it to something meaningful.
in any statement_control at /Users/liz/Github/rakudo.moar/blib/Perl6/Grammar.moarvm line 1
14:37 rypervenche joined 14:55 Altai-man left
nine jnthnwrthngtn: I think your "Improve error when `make` encounters non-Match $/" commit broke Inline::Perl5: build.opensuse.org/package/live_bu...5.4/x86_64 15:11
15:28 vrurg_ joined 15:31 vrurg left
jnthnwrthngtn How is that even possible? 18:55
If it ain't a `Match` then the bind must fail
So I don't see how it could have done anything except turned one exception into another 18:57
Looking at the stack trace, maybe there's some NQP involved and it gets an NQPMatch? But then I still don't understand how the attribute bind would succeed. 18:58
lizmat Match does NQPMatchRole, that would make an NQPMatch work, no ? 19:04
jnthnwrthngtn Well, it'd avoid the exception but it still makes no sense at all 19:08
Because the class should match
Oh, I guess there's that "slot index" optimization (that I'm kinda pondering getting rid of) that might be making it blindly access a particular slot of the object without checking 19:09
And then it's mostly luck that an NQPMatch and a Match put $!make in the same place
(Said opt pre-dates spesh, and isn't worth so much with it) 19:10
nine Such a thing exists? Oh wow 19:11
jnthnwrthngtn Yeah, back to the Parrot days I think, when I was scratching around for any technique for speed-ups 19:12
Ah well, there's another cleanup that we can expect to end with yak shaves... 19:13
japhb At what point have we sheared the entire flock and need to wait for them to regrow their wool? 19:27
.oO( Imagine an infinite flock of yaks, from which you shear every other yak. You have now sheared half of the yaks, but there still are an infinite number left, and it has taken you an infinite amount of time to get even this far ... )
nine Is it true that English doesn't have a collective noun for a group of yaks? I'm disappointed :D 19:31
lizmat a gaze of yaks ? 19:36
ah, no, that's for raccoons :-)
Nicholas Wikipedia is claiming "herd" en.wikipedia.org/wiki/List_of_animal_names#Y 19:37
20:05 sena_kun left 20:06 sena_kun joined
[Coke] (collective nouns)-- 20:58
21:20 discord-raku-bot left 21:21 discord-raku-bot joined 23:23 discord-raku-bot left, discord-raku-bot joined