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.
Geth rakudo: a5b9fc2de5 | (Nick Logan)++ | 2 files
Fix thread safety of some state variables
 00:07
03:06 AlexDani` joined 03:07 p6bannerbot sets mode: +v AlexDani` 03:10 AlexDaniel left 05:48 hankache joined 05:49 p6bannerbot sets mode: +v hankache, lizmat left 06:00 lizmat joined, p6bannerbot sets mode: +v lizmat 06:04 lizmat left 06:20 hankache left 06:50 AlexDani` is now known as AlexDaniel 10:03 lizmat joined, p6bannerbot sets mode: +v lizmat
lizmat Files=1264, Tests=88011, 377 wallclock secs (20.31 usr 6.16 sys + 2667.83 cusr 230.70 csys = 2925.00 CPU) 10:04
lizmat is glad he doesn't have to do -> int $_ in benchmarks anymore to prevent a Scalar allocation each iteration 10:06
nine progress :) 10:08
10:56 ilmari_ left, tyil[m] left, AlexDaniel` left 11:08 AlexDaniel` joined, p6bannerbot sets mode: +v AlexDaniel` 11:29 tyil[m] joined, ilmari_ joined, p6bannerbot sets mode: +v ilmari_, p6bannerbot sets mode: +v tyil[m] 12:51 MasterDuke joined 12:52 p6bannerbot sets mode: +v MasterDuke 13:00 MasterDuke left 13:10 lucasb joined, p6bannerbot sets mode: +v lucasb
jnthn lizmat: It wasn't a Scalar allocation, it was a boxing of the native Int :) 13:50
14:20 [Tux] joined 14:21 p6bannerbot sets mode: +v [Tux]
[Tux] Rakudo version 2018.12-157-ga5b9fc2de - MoarVM version 2018.12-13-g473324ee3
csv-ip5xs1.002 - 1.032
csv-ip5xs-208.177 - 8.489
csv-parser23.615 - 25.448
csv-test-xs-200.443 - 0.453
test7.578 - 8.447
test-t1.965 - 1.966
test-t --race0.875 - 0.938
test-t-2031.843 - 35.607
test-t-20 --race10.047 - 10.102
 14:57
15:12 Ven`` joined 15:13 p6bannerbot sets mode: +v Ven`` 15:16 lizmat left 15:49 lucasb left 16:04 pmurias joined, p6bannerbot sets mode: +v pmurias
pmurias when precompiling scripts in rakudo.moarvm where do we want to store the precompiled result? 16:05
nine: ping 16:25
16:50 Ven`` left
nine pmurias: I'd imagine the same precomp store as modules go? 17:41
pmurias nine: even in the perl6-m foo.p6 case? 17:45
nine: for modules the store depends where they are installed to, and use lib has it's own .precomp store right? 17:48
nine We always store into $*REPO's precomp store, i.e. the head of the chain. Regardless where the actual source is coming from. 18:05
18:06 lizmat joined, p6bannerbot sets mode: +v lizmat
nine This effectively prevents us from missing some dependency just because it was precompiled into a repo that's not in the current repo chain 18:06
pmurias nine: so with scripts you think we should store in the $*REPO? 18:10
nine: btw what has been stopping us from having precompilation of scripts? 18:11
nine tuits ;)
yes, store it in $*REPO's store 18:12
pmurias getting moarvm compiled_script.moarvm needs more work arounds but loading from the main.nqp should avoid a lot of problem 18:14
won't we get into problems with scripts filling up the store over time? 18:16
18:19 patrickb joined 18:20 p6bannerbot sets mode: +v patrickb
pmurias nine: also why is CompUnit::PrecompilationRepository a repository? it doesn't keep track where things are store just takes care of precompilation right? 18:42
robertle regarding precomp stores filling up over time: can't we use atime/mtime and have some sort of periodic cleanup? 18:45
pmurias we can, but it's something that likely should be kept in mind 18:48
lizmat is there a reason why we don't / cannot have a Int.unival ? 18:50
or a Str.unival ? 18:51
ah, I guess I was thrown by the specificity of the error
m: dd unival("a") 18:52
camelia NaN
pmurias nine: I guess it's a repository for precompiled things 19:05
nine: I think it would help readability of the CompUnit::* if the main roles had a textual description of what they do 19:06
19:38 patrickb left
Geth rakudo: 08ae9bfcff | (Elizabeth Mattijsen)++ | 3 files
Streamline unival

  - move logic to Int, as the basic logic operates on an Int
  - rewrite logic using state to logic using nqp::ifnull
  - add Str.unival method
  - adapt Cool.unival method accordingly
  - makes unival() more than 2x as fast		 20:07
roast: 205a6fb543 | (Aleks-Daniel Jakimenko-Aleksejev)++ | integration/weird-errors.t
Add a test for chain op miscompilation issue

Resolves rakudo/rakudo#1886.		 20:10
synopsebot RAKUDO#1886 [open]: github.com/rakudo/rakudo/issues/1886 [testneeded] Miscompilation (optimization) of chain ops
Geth roast: 330ef1c34a | (Aleks-Daniel Jakimenko-Aleksejev)++ | integration/weird-errors.t
Give the test its own block
 20:13
roast: Kaiepi++ created pull request #508:
Add tests for R#2143		 20:17
synopsebot R#2143 [open]: github.com/rakudo/rakudo/issues/2143 [Hacktoberfest][easy to resolve][good first issue][testneeded] Can't .gist/.Str/.perl Distribution::Resource:U
20:19 lizmat left, lizmat joined, p6bannerbot sets mode: +v lizmat
Geth roast: 9137a7dbd5 | (Elizabeth Mattijsen)++ | S15-unicode-information/unival.t
Make unival tests less picky about how they die
 20:21
roast: 1c121e979b | (Aleks-Daniel Jakimenko-Aleksejev)++ | integration/weird-errors.t
Add test for segfaulting coercer

Resolves rakudo/rakudo#1374.
synopsebot RAKUDO#1374 [open]: github.com/rakudo/rakudo/issues/1374 [testneeded] SEGV with QAST tree in core Actions
Geth rakudo: Kaiepi++ created pull request #2585:
Add tests for R#2143		 20:37
synopsebot R#2143 [open]: github.com/rakudo/rakudo/issues/2143 [Hacktoberfest][easy to resolve][good first issue][testneeded] Can't .gist/.Str/.perl Distribution::Resource:U
Geth rakudo: f39be10cea | (Ben Davies)++ | t/02-rakudo/99-misc.t
Add tests for R#2143

Fixes #2143		 20:40
rakudo: b20ff803d1 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | t/02-rakudo/99-misc.t
Merge pull request #2585 from Kaiepi/squashathon

Add tests for R#2143
synopsebot R#2143 [closed]: github.com/rakudo/rakudo/issues/2143 [Hacktoberfest][easy to resolve][good first issue][testneeded] Can't .gist/.Str/.perl Distribution::Resource:U
rakudo: 27332f0441 | (Elizabeth Mattijsen)++ | src/core/Cool.pm6
unival() is the same for all backends but the jvm
 20:45
rakudo: 485dec4df7 | (Elizabeth Mattijsen)++ | src/core/Cool.pm6
univals() is the same on all backends but the jvm
 21:13
21:15 chansen___ is now known as chansen 21:16 pmurias left, chansen is now known as chansen_
Geth rakudo: 64c88f9198 | (Elizabeth Mattijsen)++ | 2 files
univals() should be handled by Str.univals

  - so add that
  - refer all others to Str.univals		 21:46
21:53 travis-ci joined, p6bannerbot sets mode: +v travis-ci
travis-ci Rakudo build passed. Elizabeth Mattijsen 'univals() is the same on all backends but the jvm' 21:53
travis-ci.org/rakudo/rakudo/builds/475807842 github.com/rakudo/rakudo/compare/2...5dec4df7bb
21:53 travis-ci left 23:39 evalable6 left 23:41 evalable6 joined, p6bannerbot sets mode: +v evalable6

Powered by Cro and the Raku® Programming Language.

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

Thank you!