Welcome to the main channel on the development of MoarVM, a virtual machine for NQP and Rakudo (moarvm.org). This channel is being logged for historical purposes.
Set by lizmat on 24 May 2021.
00:06 reportable6 left 00:10 linkable6 joined 03:13 frost joined 04:08 reportable6 joined 04:24 [Coke] left, [Coke] joined 06:08 reportable6 left 06:10 reportable6 joined
nine_ I wonder for which problem would bitwise string ops be the best solution. 09:01
Nicholas "bitwise operations on byte buffers" I think 09:03
I can't remember if I ever used them, or was just considering them as a solution to a problem, but using (equal length) "strings" of 0-255 gets you various efficient bitwise represtations, which was a potential solution for at least one problem 09:05
oh, that sentance needed better editing.
#moarvm, * good, *, *. ish. 09:06
nine_ Yeah, that sounds more like byte buffers, for which I can totally see the need for bitwise operations 09:11
Nicholas but in perl 5 (and perl 4) all you have are hammers 09:12
or something like that. Desert, floor polish, characters string and byte buffer. 09:13
09:55 evalable6 left, linkable6 left, evalable6 joined 09:56 linkable6 joined
Nicholas ooh, fruitcake level shiny - lists.tetaneutral.net/pipermail/cf...00013.html 10:47
I bet we can't build on them... 10:48
Geth MoarVM: nwc10++ created pull request #1638:
Add mimalloc submodule, and default to mimalloc where possible
10:50
10:56 evalable6 left, linkable6 left, linkable6 joined 10:58 evalable6 joined 11:30 Altai-man joined 11:41 Altai-man left 11:42 Altai-man joined
dogbert11 does anyone dare to merge github.com/MoarVM/MoarVM/pull/1636 ? 12:07
12:07 reportable6 left
Nicholas Nope, can't build: 12:17
3rdparty/libatomicops/src/atomic_ops.h:430:5: error: #error Cannot implement AO_compare_and_swap_full on this architecture.
430 | # error Cannot implement AO_compare_and_swap_full on this architecture.
| ^~~~~
and oh my, updating submodules took a long time (and had to be edited to use ssh) 12:18
oh, mmm, I wonder if we could optionally use C11 atomics instead of libatomics API if we have the former... 12:19
nine_ That would be nice 12:48
Does MSVC support C11 atomics?
Nicholas "Not enough" for the probe I created to pass. Turns out that mimalloc has #elif defined(_MSC_VER) 12:51
and a special-snowflake implementation 12:52
13:10 reportable6 joined
Geth MoarVM/mimalloc-optional: 8 commits pushed by (Nicholas Clark)++, (Daniel Green)++ 14:04
14:29 frost left 15:53 shinto joined
[Coke] good moarning. 16:04
[Coke] wonders how hard it is to make this work in anything other than English 16:05
lizmat goede moargen :-) 16:07
Gute Moargen
Bon Joar (meh)
[Coke] Danke! 16:11
moarshi moarshi? . o O (Nope) 16:12
gfldex In some parts of DE "Morgen" is acutally pronounced "Moargen". The "r" is almost lost tho. 16:51
lizmat yeah. "mogge" is also quite common in NL :-) which would be close to "moagge" :-) 16:57
Geth MoarVM: 9c2df858c3 | (Stefan Seifert)++ | 3 files
Treat MVM_CALLSITE_ARG_(FLAT | NAMED) like FLAT_NAMED

If both the FLAT and NAMED flags are set, this means exactly the same as the FLAT_NAMED flag, so we treat them the same. Later on we can free up the FLAT_NAMED flag. Doing this in two steps to avoid bootstrapping issues.
16:59
MoarVM: b1069f7c6c | (Stefan Seifert)++ | 6 files
Free up MVM_CALLSITE_ARG_FLAT_NAMED flag

The flag is superseeded by the combination of MVM_CALLSITE_ARG_FLAT and MVM_CALLSITE_ARG_NAMED. Freeing it up, so we have a flag available for unsigned integers.
lizmat feels a bump coming :-) 17:00
nine_ lizmat: the NQP commit already contains a bump 17:01
lizmat aha! ok
nine_ This is still preparatory work for the actual unsigned fixes
lizmat yeah, got that :-)
nine_ I'm glad that the branches are shrinking :) 17:02
17:08 Core5549 joined 17:11 shinto left 17:16 shinto joined, Core5549 left 17:18 shinto left 17:25 shinto joined
lizmat nine_ : fwiw, HEAD on nqp tests clean for me on MacOS 17:33
18:08 reportable6 left 18:17 nine_ is now known as nine
nine Good to know :) 18:17
19:09 reportable6 joined 19:11 Core6356 joined, shinto left 19:12 shinto joined 19:14 Core2487 joined 19:15 Core4335 joined, Core6356 left 19:16 Core9796 joined 19:17 shinto left 19:18 Core2487 left 19:19 Core4335 left 19:29 Altai-man left
MasterDuke Nicholas: many thanks for the mimalloc work. did you notice my concerns/questions about the telemeh-related changes? 19:44
Nicholas yes, sorry, hadn't got to a reply on that one yet 19:46
MasterDuke no worries, i just realized i made them on the old pr
hm. i did just compile CORE.c with telemetry logging and didn't get any errors... 19:49
which i honestly find surprising, i thought those particular changes were in fact incorrect. maybe not... 19:51
Nicholas++ 19:58
Nicholas thank you for the sanity check 20:01
20:29 evalable6 left, linkable6 left 20:31 evalable6 joined 20:45 shinto joined 20:46 Core3143 joined 20:49 Core9796 left 20:50 shinto left 23:30 linkable6 joined