jnthn If we iterated up then the bug would probably not have happened :P 00:00
Geth arVM: fc5fa05636 | (Samantha McVey)++ | 2 files
Fix ASAN errors in generate_codepoints_by_name

Off by one error.
00:01
arVM: ed2df9ac90 | (Jonathan Worthington)++ | 2 files
Merge pull request #499 from samcv/asan

Fix ASAN errors in generate_codepoints_by_name
jnthn jnthn-- for not spotting that in review
00:18 vendethiel joined 00:50 travis-ci joined
travis-ci MoarVM build passed. Jonathan Worthington 'Merge pull request #499 from samcv/asan 00:50
travis-ci.org/MoarVM/MoarVM/builds/191805589 github.com/MoarVM/MoarVM/compare/d...2df9ac90d5
00:50 travis-ci left 02:02 FROGGS__ joined 02:48 ilbot3 joined 03:00 FROGGS_ joined 04:15 vendethiel joined 04:33 vendethiel joined
samcv .tell jnthn rewrite is gonna be great. to access the binary properties all you will need mybitfield[10].NFG_QC 04:37
yoleaux2 samcv: I'll pass your message to jnthn.
samcv :)
there is going to be compression for some ranges where there are no characters, like an offset. but it will be way faster 04:38
and determining an offset is not too bad
truncated version of the skeleton of what will hopefully be the rewrite gist.github.com/0692493144a5bb6892...8887c0176f 04:47
05:26 ggoebel joined
nwc10 samcv++ 08:11
ASAN is happy
samcv :)
08:58 domidumont joined 09:04 domidumont joined 11:06 domidumont joined 13:06 brrt joined
brrt .tell samcv keep in mind that bitfield structs will still be 4 bytes wide at least 13:06
yoleaux2 brrt: I'll pass your message to samcv.
Geth arVM/even-moar-jit: 56 commits pushed by bdw++
review: github.com/MoarVM/MoarVM/compare/4...1818179181
13:13
brrt merge commit
13:16 domidumont joined 13:34 Geth joined 13:42 Ven joined 14:36 Geth joined
nwc10 jnthn: ASAN found something new paste.scsys.co.uk/540705 16:24
jnthn nwc10: Ooooh. Thank you! 16:28
yoleaux2 04:37Z <samcv> jnthn: rewrite is gonna be great. to access the binary properties all you will need mybitfield[10].NFG_QC
jnthn nwc10: That first backtrace is exactly what I occasionally got under GDB. On its own it's really not enlightening.
16:29 stmuk joined
jnthn The freed/allocated are...weird too but provide a much better idea of where to look 16:29
samcv: Sounds rather nice. :-) 16:30
nwc10 I'm glad that it makes some sense to you 16:32
jnthn Well, it doesn't make sense yet, but it gives some leads I was missing :) 16:33
17:00 TimToady joined 17:43 zakharyas joined
samcv brokenchicken, yep i know that is true. have no implemented all the packing 19:27
yoleaux2 13:06Z <brrt> samcv: keep in mind that bitfield structs will still be 4 bytes wide at least
samcv 4bytes on 32bit 8 on 64
brokenchicken what? 20:17
Ah.
samcv the uh aligning of struct data structures and things 22:16
jnthn, what do you mean by leads
oh what nwc10 said? sorry i just woke up 22:17
timotimo right, the asan output ought to give jnthn an idea where to look 22:18
jnthn samcv: Yes, was talking abou tthe ASAN output :) 23:36
jnthn heads sleepwards 23:48