Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by Zoffix on 27 July 2018.
Kaiepi AlexDaniel, i'm pretty sure what you were asking about is a leak, but i won't be able to confirm that it is or not until tomorrow since i found some race conditions while testing that i haven't finished fixing yet 00:51
AlexDaniel Kaiepi: alright, thanks for checking it 00:52
I left a comment on a PR so that nobody hits merge accidentally 00:53
I mean, on your PR
03:48 irced left 03:58 gugod joined 06:43 robertle left 08:09 MasterDuke left 08:17 commavir left 08:19 commavir joined 08:28 BeastieBot joined 08:33 chloekek joined 08:40 Kaiepi left 08:56 MasterDuke joined, chloekek left 09:01 robertle joined
|Tux| Rakudo version 2019.07.1-260-g9dd67cb37 - MoarVM version 2019.07.1-180-gcf3b545e5
csv-ip5xs0.689 - 0.690
csv-ip5xs-204.936 - 5.274
csv-parser21.401 - 21.509
csv-test-xs-200.441 - 0.445
test6.929 - 7.481
test-t1.751 - 1.766
test-t --race0.774 - 0.853
test-t-2029.797 - 30.301
test-t-20 --race9.088 - 9.362
10:29
10:37 Kaiepi joined 10:46 Voldenet left 12:23 lucasb joined
kawaii usercontent.irccloud-cdn.com/file/...-06.tar.gz 13:07
AlexDaniel: ^
timotimo cdn.discordapp.com/attachments/538...nknown.png - neato makes the overview svg look much more readable! 13:12
13:26 chloekek joined
AlexDaniel it's readable enough when there are just a few regressions 13:40
tellable6 2019-09-06T06:05:47Z #perl6 <jmerelo> AlexDaniel are hooks ready and everything?
13:43 j3nnn1 joined
Kaiepi bastille% perl6 -e 'say [+] "malloc_dump_old".IO.slurp.lines[2..*].map(*.split(/\s+/)[2])' 13:54
51117774
bastille% perl6 -e 'say [+] "malloc_dump_new".IO.slurp.lines[2..*].map(*.split(/\s+/)[2])'
50639454
managed to cut down the memory leaked when just binding an async tcp socket and connecting to it by 50KB on my branch 13:56
or at least that's what openbsd's libc claims
*500KB 13:57
14:30 Xliff joined
vrurg Hm, nmake cannot do ?=. wtf... :( 14:34
14:57 |Tux| left 15:01 chloekek left
AlexDaniel kawaii: github.com/rakudo/rakudo/issues/31...-528886751 15:18
kawaii AlexDaniel: HEAD from yesterday, if that makes a difference 15:19
AlexDaniel kawaii: so you just used --new=HEAD right? 15:20
kawaii Yep
15:26 |Tux| joined 15:39 |Tux| left 15:56 |Tux| joined 15:59 robertle left 16:21 AlexDaniel` left, rba[m] left, Guest999 left
vrurg kawaii: would it be possible for you to somehow test Cro::Core manually? I'm positive that something went wrong about this Blin run. 16:21
kawaii I can do so later, sure
vrurg kawaii: just for the record: my Rakudo from HEAD is 2019.07.1-180-gcf3b545e5
kawaii: thanks! I'm gonna be afk for a couple of hours anyway. 16:22
16:34 AlexDaniel` joined 17:02 Kaiepi left 17:09 rba[m] joined, Guest48976 joined 17:38 BeastieBot left 17:47 Ven`` joined 17:58 Ven`` left 18:08 Voldenet joined, Voldenet left, Voldenet joined
Xliff Does anyone know if there are any problems with CStructs if the last attribute is another CStruct included by "HAS"? 18:24
I keep getting a Type object when it shouldn't be/
18:28 Xliff left 18:54 Ven`` joined
nine What the? In this code the Foo object gets destroyed and its DESTROY called: my $foo = Foo.new; is $foo.test(), 1, 'obj survives'; is $foo.test(), 1, 'obj survives'; 18:54
In this code it's not: my $foo = Foo.new; is $foo.test(), 1, 'obj survives' for ^2;
In this not either: my $foo = Foo.new; -> { is $foo.test(), 1, 'obj survives'; }() 18:55
But in this it works: my $foo = Foo.new; -> $foo { is $foo.test(), 1, 'obj survives'; }($foo)
Oh, actually it does! It just takes many more iterations of the test. So I guess that some code will cause $foo to become referenced by the old generation and it's much harder to trigger a full collection 19:08
19:15 Kaiepi joined 20:19 MasterDuke75 joined 20:22 MasterDuke left 20:39 go|dfish joined 23:24 Ven`` left