github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
Geth MoarVM: e783f56759 | (Samantha McVey)++ | tools/update-changelog.p6
Unbreak update-changelog.p6 and use new Git::Log module

Use my new module, Git::Log instead of a super hackish parsing of git log's output. The new version should be less error prone and simpler.		 00:39
samcv the changelog is so long
samcv got the script working again though, here is the output atm gist.github.com/90fb70be9a536518d1...fd584d91ba
AlexDaniel samcv: if it is generated like that, maybe it should also include commit shas at the end like rakudo changelog does 03:07
or maybe not, you decide :)
lizmat .tell brrt nice blog post, s/langauge/language/ 09:12
yoleaux lizmat: I'll pass your message to brrt.
brrt \o 09:35
yoleaux 09:12Z <lizmat> brrt: nice blog post, s/langauge/language/
brrt thank you lizmat :-) changed
TimToady brrt: I think I saw a s/Truffe/Truffle/ while you're at it 15:23
TimToady though given the extra images in my "good" eye, counting ascenders is a bit problematic right now :) 15:24
hmm, not on, actually... 15:26
.tell brrt I think I saw a s/Truffe/Truffle/ while you're at it
yoleaux TimToady: I'll pass your message to brrt.
timotimo hm, do we ever inline something into a P6opaque that has its own unmanaged_size? because in that case we'd be missing calling that on our inlined objects 17:34
i want to record the maxrss in the profiler, though it'd be much nicer if i could get a "current rss" value instead :| 17:53
jnthn timotimo: I guess a P6bigint with an allocated bigint part would be such a case 18:35
timotimo that's what i thought; we inline P6str sometimes, but that's "just" a pointer to a VMString iirc 18:37
the allocation tracker may want to learn a bit more about some ops, like ones that create a VMString + a P6str
imgur.com/Lh7Ye6s - until then it'll look a bit like this; i'll also add a column with bytes * count 18:39
*sigh* i wonder if it's time to burn this system to the ground and start anew 18:58
brrt \o 19:32
yoleaux 15:26Z <TimToady> brrt: I think I saw a s/Truffe/Truffle/ while you're at it
brrt thanks, updated
lizmat .tell brrt www.reddit.com/r/perl6/comments/9a...e/e4vw9br/ 22:59
yoleaux lizmat: I'll pass your message to brrt.
jnthn I suspect he really meant bytes, and that he's also talking precisely about the size of the machine code produced for that function 23:03
But it's also a bit misleading, in that JIT will have been triggered by OSR
And that means it includes the bunch of boilerplate that we spit out at the start of every program unit, such as the stuff that deals with CTXSAVE 23:04
Of course, that isn't part of the interesting loop under consideration
So it's probably not as bad as those numbers make it look
lizmat jnthn: thanks! 23:12
afk again&&
timotimo i'll write something about the OSR thing 23:20
timotimo oooh, why don't i take the rusage of cpu time before and after every gc run and use that to figure out how much the program has been doing between GC runs (and i guess also during the GC run) 23:44

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!