| IRC logs at
Set by AlexDaniel on 12 June 2018.
00:54 Altai-man_ joined 00:57 sena_kun left 01:11 leont left 02:47 chansen_ left 02:50 chansen_ joined 02:55 sena_kun joined 02:57 Altai-man_ left 04:54 Altai-man_ joined 04:57 sena_kun left 06:55 sena_kun joined 06:57 Altai-man_ left 08:55 Altai-man_ joined 08:57 sena_kun left 09:10 MasterDuke left 09:27 leont joined 09:51 MasterDuke joined
nwc10 good *, #moarvm 09:53
Altai-man_ o/ 09:54
nwc10 \o 09:56
Altai-man_ prepares moarvm release
10:56 sena_kun joined 10:57 Altai-man_ left
Geth MoarVM: 29cc53bfed | Altai-man++ | docs/ChangeLog
Update ChangeLog for 2020.06 release
sena_kun Can anyone please check ^ ?
MasterDuke hm. where was i with that Proc::Async sometimes slowdown weirdness? 11:06
lots of time spent in MVM_gc_root_add_frame_roots_to_worklist and MVM_gc_root_add_frame_registers_to_worklist, but hadn't figured out why i think
sena_kun prepares a build to upload 11:34
Geth MoarVM/release-2020.06: dd39f4a5e2 | Altai-man++ | VERSION
Bump VERSION for release
MoarVM: Altai-man++ created pull request #1316:
Bump VERSION for release
MoarVM: dd39f4a5e2 | Altai-man++ | VERSION
Bump VERSION for release
MoarVM: 1a45d28a43 | Altai-man++ (committed using GitHub Web editor) | VERSION
Merge pull request #1316 from MoarVM/2020.06

Bump VERSION for release
sena_kun And we have a MoarVM 2020.06 release. \o/ 12:15
dogbert17 sena_kun++ 12:36
sena_kun hopes the changelog is ok 12:37
dogbert17 thinks so
sena_kun wait
gah 12:38
of course it is wrong
Geth MoarVM: 7d5cc1f3a4 | Altai-man++ | docs/ChangeLog
Fix changelog
12:55 Altai-man_ joined 12:57 sena_kun left
lizmat Altai-man_ I think a whitespace only change is not really a change :-) 13:01
Altai-man_ lizmat, rendering on the website was broken, sweeping all the categories under one, which is a failure. I did not rebuild the tarball, since that's relatively ok, I guess. 13:03
lizmat yeah... that's what I meant :-)
MasterDuke i enabled the GC_DEBUG_LOG and get at ton of output. one line is `Thread 1 run 238 : processing 318048 items from in tray`. is that a lot of items? 13:35
hm. the largest "processing" value when compiling CORE.c is only 20k 13:58
ah. and if the perl being executed is is `for(1..100_000)` instead of `while(1)` then it's only 36k 14:00
looking at a profile of the `for(1..100_000)` version, 200k BOOTCodes are allocated at `has @.command is List;`? 14:03
14:56 sena_kun joined 14:57 Altai-man_ left
timotimo MasterDuke: can you get a speshlog and find the same filename / line number to see what code that corresponds to? 15:43
MasterDuke `Latest statistics for ''` 15:45
`Total hits: 1`
... `1 x type List (Conc)`; `1 x static frame 'new' (5414) (caller is outer: 0, multi 0)`; `1 x type tuple: Type 0: List (TypeObj)` 15:46
15:46 zakharyas joined
MasterDuke the cuid also shows up in the Proc::Async.BUILDALL 15:47
i don't think i've ever seen a `has` show up in a profile/spesh log before 15:49
timotimo i'd expect it to show up when you have a = in there that does some kind of calculation 15:50
15:50 zakharyas left
MasterDuke the code being run is just: `my $proc = "perl", "-E", q|for(1..100_000) { say "y" x 100 }|; react { whenever $proc.stdout.lines { say "stdout: $_"; }; whenever $proc.stderr.lines { say "stderr: $_"; }; whenever $proc.start { say "Exit code: {.exitcode}"; done; } }` 15:52
timotimo well, how is "has foo is bar" implemented?
that could have something to do with it
MasterDuke dunno. but `class A { has @.c is List; }; my $a; $a =[$_,])) for ^100_000; say $a` doesn't show the same thing in a profile 15:53
timotimo interesting
MasterDuke afk for a bit, but let me know if you have any suggestions 15:58
16:55 Altai-man_ joined 16:57 sena_kun left 18:06 patrickb joined 18:55 sena_kun joined 18:57 Altai-man_ left 20:24 zakharyas joined 20:55 Altai-man_ joined 20:58 sena_kun left 21:47 zakharyas left 22:56 sena_kun joined 22:57 Altai-man_ left 23:52 leont left, patrickz joined 23:56 patrickb left