00:30 camelia joined 01:55 FROGGS_ joined 01:57 ilbot3 joined 05:38 Util joined 06:01 AlexDaniel joined 06:31 robertle joined 07:04 domidumont joined 07:12 domidumont joined 07:38 lizmat joined 07:49 Kaiepi joined 07:52 Kaiepi joined 09:45 domidumont joined 09:47 lizmat joined
samcv dogbert17: just arrived in NL and got your message 11:48
--4048-- WARNING: unhandled amd64-linux syscall: 1000318 hmm 11:49
jnthn: hmm actually i think valgrind is right 11:51
jnthn samcv: Ah, in that the syscall isn't actually available? 11:54
We seem to be coping with that, though...
samcv jnthn: no, i accidently left in debug code to make sure the syscall was handled gracefully
jnthn oh.
samcv i am reading the valgrind document though, since i may want to implement that anyway 11:55
Geth MoarVM: e5d028ce2f | (Samantha McVey)++ | 2 files
Use all uppercase for Hangul Syllable names

U+AC00 from Hangul Syllable GA -> HANGUL SYLLABLE GA
MoarVM: 28f7fe711b | (Samantha McVey)++ | src/platform/random.c
Fix getrandom Linux syscall by removing debug code

This code was accidently left in to make sure the syscall failing was handled gracefully. Now that it's removed it should work properly.
samcv jnthn: valgrind warning is fixed now 12:07
since now it recognizes the syscall number
12:08 lizmat joined
jnthn ooh, nice :) 12:09
samcv luckily they have it included, because their description seemed pretty annoying to do if it wasn't one of the already programmed in sys call's 12:11
dogbert17 samcv++ 12:14
timotimo oh, your nl visit is now! 12:23
on friday i came back from a short visit to my parent's which is near the nl border :| 12:24
samcv timotimo: too bad i missed you 12:26
pretty jetlagged atm
timotimo i can only imagine 12:27
if we meet in person you can bonk me on the head for not doing anything for the unicode database compression thing :|
12:38 timotimo joined
samcv hah, don't worry about it 12:39
chansen_ compressing the character names? 12:44
dogbert17 samcv: you now have a filing test in t/spec/S15-unicode-information/uniname 12:59
13:09 MasterDuke joined 14:25 robertle joined 14:47 lizmat joined 15:15 domidumont joined 15:18 zakharyas joined 15:21 zakharyas joined 15:27 lizmat joined 15:41 Kaiepi joined 15:50 AlexDaniel joined 16:02 committable6 joined 16:03 bisectable6 joined
samcv dogbert17: should be fixed now 16:27
17:29 lizmat joined 18:15 Kaiepi joined 18:33 brrt joined 18:48 AlexDaniel joined
Geth MoarVM/jit-stack-walker: 9 commits pushed by (Bart Wiegmans)++ 19:15
brrt ml64 syntax is quite different from gnu assembler syntax, even with intel syntax 19:18
Geth MoarVM/2018.04.1: 5d36d30a0b | (Zoffix Znet)++ (committed by Samantha McVey) | src/strings/parse_num.c
Fix denormals and precision issues in num parser

With the original method we had problems with handling denomals
  (so e.g. 5e-324 ended up as 0e0) and had precision issues where
NOT the closest representable num was generated for a given string, which causes value drift in repeated num->str roundtripping.
... (8 more lines)
MoarVM/2018.04.1: 43f9dcfe83 | (Zoffix Znet)++ (committed by Samantha McVey) | src/strings/parse_num.c
Remove trailing whitespace
lizmat 250bpm.com/blog:124 # maybe of interest to jnthn :-) 20:00
brrt: ERROR: Cannot update submodule because directory exists and is not empty. 20:29
>>> Please delete the following folder and try again:
ah, that's old
brrt ah, ok. that's not my doing either, though :-)
lizmat make runs to completion :-) 20:31
brrt that's something 20:32
then i presume my double-labelling workarround works
on windows it is complete pain though 20:33
then again, i may be able to use that to my advantage 20:38
21:26 Kaiepi joined
jnthn lizmat: Looks like a good read; thanks :) 21:34
Geth MoarVM/pluggable-spesh: 6b121e81ef | (Jonathan Worthington)++ | 10 files
Stub the spesh plugin ops
MoarVM/pluggable-spesh: 4e454f48d4 | (Jonathan Worthington)++ | src/mast/compiler.c
Implement bytecode assembly for speshresolve
jnthn (just a rebase)
Geth MoarVM/pluggable-spesh: b12fc2df3b | (Jonathan Worthington)++ | 3 files
Implement spesh plugin registration/resolution

Since the lookup only happens on a miss, the lock here is probably not going to be too highly contended over. We can revisit that if profiling suggests otherwise (with some kind of read-only lookup).
jnthn Though that one is new :)
22:44 Kaiepi joined 23:01 Kaiepi joined