| IRC logs at
Set by AlexDaniel on 12 June 2018.
00:28 dogbert17 left, dogbert17 joined 01:34 sena_kun joined 01:35 Altai-man_ left 01:43 unicodable6 joined, tellable6 joined 01:44 evalable6 joined, linkable6 left 01:46 linkable6 joined 03:33 Altai-man_ joined 03:36 sena_kun left 05:34 sena_kun joined 05:35 Altai-man_ left 07:33 Altai-man_ joined 07:36 sena_kun left 08:35 kawaii left 08:37 kawaii joined 08:52 zakharyas joined 08:54 domidumont joined 09:02 domidumont left 09:06 domidumont joined 09:10 domidumont1 joined 09:14 domidumont left 09:34 sena_kun joined 09:35 Altai-man_ left 09:40 domidumont1 left, domidumont joined 09:42 domidumont left 09:43 domidumont joined
nwc10 good *, #moarvm 10:28
11:02 domidumont left 11:30 yego joined 11:33 yego left, Altai-man_ joined 11:35 sena_kun left 11:58 zakharyas left
Guest1277 Hello, nwc10. It's a bit empty here atm, no trace of jnthn, nine, timotimo, MasterDuke or brrt. Perhaps they're having lunch. 12:15
jnthn Lunch was nice. :) 12:49
nine Guest1277: well, $dayjob is kinda demanding :) 12:55
Guest1277 nine: indeed 13:04
13:29 lucasb joined 13:34 sena_kun joined 13:36 Altai-man_ left 13:50 zakharyas joined 14:14 MasterDuke left 14:32 brrt joined
brrt \o 14:32
nwc10 o/ 14:33
Guest1277 Hello brrt 14:36
14:56 brrt left
Guest1277 on master I get the following test failures when using MVM_SPESH_NODELAY=1 and MVM_SPESH_BLOCKING=1 14:58
t/spec/S02-types/isDEPRECATED.rakudo.moar and t/spec/S05-grammar/parse_and_parsefile-6e.t 14:59
jnthn Guest1277: I think the S02 one is longstanding and about the backtrace being less accurate, the other one looks like a real spesh bug though 15:00
Guest1277 jnthn: do you want the spesh bug reported in the issue tracker 15:03
jnthn Can do 15:06
15:12 brrt joined 15:33 Altai-man_ joined 15:35 sena_kun left
Guest1277 heh, github gives me a 500 when I try to report an issue 15:39
.oO( perhaps you should make an issue for that )
Guest1277 aargh 15:42
something's seriously messed up atm 15:44
"Investigating - We are investigating reports of increased error rates and webhook delivery times. "
M#1242 15:59
linkable6 M#1242 [open]: Failure when running S05-grammar/parse_and_parsefile-6e.t under MVM_SPESH_BLOCKING and MVM_SPESH_NODELAY
brrt Guest1277: does it go away with MVM_JIT_EXRP_DISABLE=1 ? 16:03
Altai-man_ a regression or? 16:04
Guest1277 brrt: nope 16:09
Altai-man_: if AlexDaniel is around we can find out 16:10
or someone who knows bisectable well 16:11
Altai-man_ Guest1277, can't you just test it on 2020.01? it'll be enough for deciding if we want a blocker tag or not
Guest1277 Altai-man_: sure, will do 16:13
brrt Guest1277: I mistyped.... MVM_JIT_EXPR_DISABLE=1
Guest1277 brrt: no change :) 16:14
AlexDaniel hello? 16:15
e: shell ‘cd sandbox; git clone’ 16:16
evalable6 Cloning into 'roast'...
AlexDaniel e: run <perl6 t/spec/S05-grammar/parse_and_parsefile-6e.t>; 16:17
evalable6 (exit code 1) Could not open t/spec/S05-grammar…
AlexDaniel, Full output:
AlexDaniel e: run <perl6 sandbox/t/spec/S05-grammar/parse_and_parsefile-6e.t>;
evalable6 (exit code 1) Could not open sandbox/t/spec/S05…
AlexDaniel, Full output:
AlexDaniel e: run <perl6 sandbox/roast/S05-grammar/parse_and_parsefile-6e.t>; 16:18
evalable6 (exit code 1) 1..24
ok 1 - grammar is created u…
AlexDaniel, Full output:
AlexDaniel e: %*ENV<MVM_SPESH_BLOCKING> = 1; %*ENV<MVM_SPESH_NODELAY> = 1; run <perl6 sandbox/roast/S05-grammar/parse_and_parsefile-6e.t>;
evalable6 (exit code 1) 1..24
ok 1 - grammar is created u…
AlexDaniel, Full output:
AlexDaniel e: chdir ‘sandbox/roast’; %*ENV<MVM_SPESH_BLOCKING> = 1; %*ENV<MVM_SPESH_NODELAY> = 1; run <perl6 S05-grammar/parse_and_parsefile-6e.t>; 16:19
evalable6 (exit code 1) 1..24
ok 1 - grammar is created u…
AlexDaniel, Full output:
AlexDaniel that?
e: chdir ‘sandbox/roast’; run <perl6 S05-grammar/parse_and_parsefile-6e.t>;
evalable6 1..24
ok 1 - grammar is created using 6.e versi…
AlexDaniel, Full output:
AlexDaniel yep 16:20
6c: chdir ‘sandbox/roast’; %*ENV<MVM_SPESH_BLOCKING> = 1; %*ENV<MVM_SPESH_NODELAY> = 1; run <perl6 S05-grammar/parse_and_parsefile-6e.t>;
committable6 AlexDaniel,
AlexDaniel so it was there since at least 2019.11 16:21
and that's very likely related to the .raku change? 16:22
if you want we can checkout an older version of roast and see what's there 16:24
16:32 zakharyas left
Guest1277 AlexDaniel++ 16:34
the difference compared to the derived-specializations branch are t/spec/S11-modules/require.t and t/spec/S17-procasync/nonexistent.t 16:39
lizmat so is there something that I should / could fix ? 16:41
Guest1277 bindkey_o requires a concrete object (got a Nil type object instead) 16:44
in block <unit> at t/spec/S11-modules/require.t line 108
test 21 in t/spec/S11-modules/require.t
jnthn Guest1277: Yeah, that's pretty much the difference I found too
lizmat so when does that test fail exactly ? 16:46
Guest1277 lizmat: when run under jnthn's derived-specializations branch and with MVM_SPESH_BLOCKING and MVM_SPESH_NODELAY set 16:47
jnthn And S05-grammar/parse_and_parsefile-6e.t fails on master with those flags also 16:48
16:48 patrickb joined
Guest1277 brrt: does this mean anything to you - moar: src/jit/x64/tiles.dasc:675: MVM_jit_tile_sub_load_idx: Assertion `out != in1' failed. 16:50
jnthn Guest1277: brrt already reproduced that one yesterday I think :) 16:51
Or at least, is aware of it 16:52
Guest1277 ah, cool
lizmat Unexpected named argument 'cache' passed 16:53
at gen/moar/stage2/NQPHLL.nqp:2463
but that line *does* have a :$cache
ok, I moved that test to be the first in the file, and then it just passed 16:54
so it's something that gets borked *before* that
hmm... should smart-matching a Failure perhaps disable the Failure ? 17:04
m: my $a :=; dd $a ~~ Failure; dd $a.Int # still throws
camelia Bool::True
in block <unit> at <tmp> line 1
lizmat m: my $a :=; dd $a.Bool; dd $a.Int # still throws 17:05
camelia Bool::False
lizmat *does not throw
brrt Guest1277: I heard about that yes, but I haven't reproduced it yet 17:06
17:21 Kaiepi left, Kaiepi joined
brrt I seem to get failures when running 'make spectest' though 17:23
not make test
patrickb o/ 17:27
Can someone have a look at ? 17:28
17:31 Kaiepi left 17:32 Kaiepi joined 17:34 sena_kun joined 17:35 Altai-man_ left 17:36 brrt left, domidumont joined
Geth MoarVM: 2cc4d30271 | (Patrick Böker)++ | 2 files
Revert "Dynamic library loading works with unicody paths on Windows"

This reverts commit efeec455e025f273eede110e099a8f8b21b5e108. It's not needed anymore, because the fix has been pushed upstream and been released with Dyncall 1.1 which we just bumped to.
MoarVM: b0604e575b | (Jonathan Worthington)++ (committed using GitHub Web editor) | 2 files
Merge pull request #1241 from patrickbkr/win-unicode-workaround-revert

Revert "Dynamic library loading works with unicody paths on Windows"
patrickb Thanks!
lizmat m: my $a =; $a.not; dd $a # doesn't this look, weird ??? 17:39
camelia Failure $a = &CORE::infix:<orelse>( => => "Failed"), backtrace =>, *.self)
lizmat where does the infix<orelse> come from ?
dogbert17 .tell brrt compile moarvm with --no-optimize and then run: MVM_SPESH_NODELAY=1 MVM_SPESH_BLOCKING=1 ./perl6-m -Ilib t/spec/S05-grammar/parse_and_parsefile-6e.t 17:55
tellable6 dogbert17, I'll pass your message to brrt
18:08 domidumont left 18:48 MasterDuke joined 19:32 brrt joined 19:33 Altai-man_ joined 19:35 sena_kun left 19:51 brrt` joined 19:52 brrt left 20:01 Kaiepi left, Kaiepi joined 20:30 brrt`` joined 20:32 brrt` left 21:34 sena_kun joined 21:35 Altai-man_ left 22:19 lucasb left 22:21 patrickb left 22:31 brrt`` left 23:21 MasterDuke left 23:33 Altai-man_ joined 23:35 sena_kun left