01:56 ilbot3 joined 03:57 AlexDaniel joined 05:50 live_the_dream joined, releasable6 joined 05:51 committable6 joined 05:52 undersightable6 joined 06:50 domidumont joined 06:56 domidumont joined 07:28 AlexDaniel joined
samcv why do i have all these .opt.yaml files? i am not seeing it on another platform. at least freebsd. though idk if it is something that we don't use anymore? 07:39
but i have a ton of them on my dev machine
08:12 zakharyas joined 08:17 zakharyas joined
lizmat m: use nqp; say nqp::istype(Mu:D, Mu:D) # is that to be expected, or a bug ? 08:22
camelia 0
lizmat m: use nqp; say nqp::istype(Mu:D, Mu) # seems ok
camelia 1
samcv anyone know why the appveyor builds don't run nqp test? 08:45
pretty sure that's why it's disabled. unless my changes caused it.
timotimo lizmat: Mu:D isn't definite, so it doesn't pass the definiteness check that Mu:D imposes on it 09:45
m: use nqp; subset Even of Int where * %% 2; nqp::istype(Even, Even) 09:46
camelia ( no output )
timotimo m: use nqp; subset Even of Int where * %% 2; say nqp::istype(Even, Even)
camelia 1
timotimo oh?
i guess in this case we assume something that is even already passed the check once?
lizmat m: use nqp; say nqp::istype(Int:D, Int:D) # should this be one then? 10:26
camelia 0
lizmat ah, I think I understand what you're saying 10:27
it can never pass the :D as a target because the source is a type object
hmmmm 10:28
11:05 dogbert17 joined 11:18 Zoffix joined
Zoffix Any tips for what this means/how to debug/fix it? "moar: src/gc/collect.c:218: process_worklist: Assertion `*item_ptr != item->sc_forward_u.forwarder' failed." 11:19
The same code first failed with "Bus error" before that error turned up
ZofBot: bus error? But I drive a Chevy Cavalier! 11:21
ZofBot Zoffix, list? pmurias: it does show up in a simple test: perl6 --ll-exception -e 'class Foo { method BUILD() { die } }
Zoffix Hm, it's garbage collection stuff. I think it's time for me to give up. I've not read the book yet 11:26
AlexDaniel Zoffix: how did you get this? 11:27
Zoffix AlexDaniel: github.com/rakudo/rakudo/issues/17...-380066336 11:30
R#1709 11:31
synopsebot R#1709 [open]: github.com/rakudo/rakudo/issues/1709 [ASYNC][āš  blocker āš ] A race in HyperRace causing rare crash/hang
Zoffix With the golf it now takes ~3m to repro a problem instead of original ~15-20m, so I'd call it a win on my part :)
11:33 Zoffix left
nwc10 good, er, wait, come back! 11:39
nwc10 has t/spec/S17-supply/syntax.t ........................................
hung currently
it does it every so often 11:40
I wonder if it's the same one
11:49 domidumont joined
dogbert17 Zoffix: your example, i.e. ./perl6 -e 'loop { my @ = (^20000).race.grep(*.self); print "." } hung for me as well 12:37
12:55 zakharyas joined 13:16 lizmat joined 13:53 lizmat joined 14:05 lizmat joined
Geth MoarVM/spesh-limit-debug-log: 4fcf839ace | (Bart Wiegmans)++ | 6 files
[Spesh] Wrap debug printing

The intention is that MVM_spesh_debug_enabled(tc) can be made more specific than just having a spesh log file handle or not.
Specifically, setting MVM_SPESH_LIMIT in conjunction with MVM_SPESH_LOG will restrict the log to that particular frame.		 14:15
14:20 tbrowder_ joined 14:23 zakharyas joined 14:35 lizmat joined
dogbert17 .seen timotimo 14:42
yoleaux I saw timotimo 12:20Z in #perl6: <timotimo> yes, it's a beautiful thing, DDT
dogbert17 .tell timotimo added MVM_dump_backtrace output to github.com/MoarVM/MoarVM/issues/837. Perhaps useful, perhaps not :) 14:44
yoleaux dogbert17: I'll pass your message to timotimo.
14:50 lizmat joined 15:04 lizmat joined 15:20 lizmat joined 15:46 lizmat joined 15:52 AlexDaniel joined 15:58 brrt joined
brrt good * #moarvm 15:59
16:29 lizmat joined 16:31 zakharyas joined 16:37 domidumont joined 17:13 lizmat joined 17:38 btyler joined 18:13 lizmat joined
samcv .till Zoffix compiling is failing on solaris due to INFINITY 18:54
i'll try and fix it
19:13 lizmat joined 19:14 Zoffix joined
Zoffix samcv: damn :( I already had to add a define for MVC: github.com/MoarVM/MoarVM/commit/fa...abad4b62c9 19:14
Maybe solaris will like `#define INFINITY 1.0/0.0` 19:15
samcv though to be fair it doesn't compile period
:P
Zoffix :)
samcv getting some weird error compiling dyncall
Zoffix `#define INFINITY (1.0/0.0)` rather
samcv /usr/include/libelf.h: error # error "large files are not supported by libelf"
say what now 19:16
Zoffix oh, right, #include float and then (DBL_MAX + DBL_MAX) 19:17
Maybe I should just do that for all of 'em
samcv not sure. are you sure it's not platform specific? 19:18
i tried adding an assert(isinf(0)); into my code. but that didn't trigger. how do i get an assert to trigger? or does it not let me do function calls?
Zoffix DBL_MAX is the maximum number that can fit into a double, and double of that is infinity
samcv maybe we need a test when you do Configure.pl to make sure that whatever we define infinity as, it's actually infinity on that platform
timotimo oh, btw, we have defines for infinity and friends in moarvm already 19:19
samcv so it's true on EVERY platform?
timotimo: yeah i saw that
timotimo i thought i had already told y'all, but i must have forgotten?
samcv except AIX?
Zoffix timotimo: what are they?
samcv grep INFINITY
Zoffix That only gives what I had 19:20
Ah MVM_NUM_POSINF
samcv i see MVM_NUM_POSINF
Zoffix "static const" that'll limit it to just that file, won't it? 19:21
will fix it later
19:22 Zoffix left
samcv wow solaris is so weird 19:26
`ping localhost` #> "localhost is alive" that's literally what it shows 19:27
and nothing else
some things do weird stuff..
geekosaur there's an option to make it do what you expect 19:35
solaris has more than a bit of historical baggage it can't shed without breaking enterprise uses
often makes sense to stick whichever /usr/xpg*/bin exists in $PATH before the standard stuff 19:36
samcv well i have a manpage for fbconfig but i don't seem to have a binary 19:40
20:06 lizmat joined
samcv libtommath doesn't compile on solaris 20:30
20:30 lizmat joined
samcv i was able to fix it. though i may need to submit a PR to them if they haven't fixed it 20:32
geekosaur: it seems #include <atomic_ops.h> is giving solaris some issues 20:43
saying it can't find it
hmm i may have broke that 20:44
gonna do a clean clone just to check 20:50
20:52 lizmat joined 21:02 lizmat joined 21:13 lizmat joined
samcv geekosaur: i can't install gcc-5 it seems. and i want to see if the issues i'm getting are just a bug with gcc 4 21:24
pkg install: No matching version of developer/gcc-5 can be installed: 21:25
Reject: pkg://solaris/developer/gcc-5@5.4.0-5.12.0.0.0.115.0
pkg://solaris/developer/gcc-5@5.4.0-5.12.0.0.0.122.0
Reason: No version for 'require' dependency on release/evaluation can be found
I get this
any clue
it seems like the packages may be masked or something? since they seem to exist but i'm not sure how to install them
geekosaur looks like you have to tell it to permit evaluation / testing packages?
pkg install --accept, apparently 21:29
samcv yeah i tried that 21:30
geekosaur no idea then. it's beena while since I actually worked with solaris 21:32
and, well, with discontinuation they may have nuked it or something
Geth MoarVM: 96e432da09 | (Zoffix Znet)++ | 3 files
Share INFINITY def between files that use it
 22:16
22:40 Kaiepi joined
Geth MoarVM: 957bb4bd8d | (Timo Paulssen)++ | 2 files
remove wval-deserialization-retry mechanism for now

it's causing crashes.		 23:21
MoarVM: 45a3e8e6b2 | (Samantha McVey)++ | build/setup.pm
Fix solaris 32 bit build

32 bit solaris doesn't support _FILE_OFFSET_BITS=64 and on 64 bit it is enabled by default.		 23:23
MoarVM: 1504a27bb4 | (Samantha McVey)++ | build/setup.pm
Don't build specify -D_FILE_OFFSET_BITS for dyncall

This fails on 32-bit Solaris and on 64-bit Solaris is not needed. Use
  -U_FILE_OFFSET_BITS so we just negate our other CFLAGS, so we can
compile everything except dyncall with -D_FILE_OFFSET_BITS so we don't lose the ability to work with large files.
MoarVM: 22c941f891 | (Samantha McVey)++ | 3rdparty/libtommath
Bump libtommath with fix for building on Solaris

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!