Perl 6 - the future is here, just unevenly distributed

IRC log for #moarvm, 2016-09-13

| Channels | #moarvm index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
03:53 japhb joined #moarvm
04:20 japhb joined #moarvm
04:51 japhb joined #moarvm
05:18 japhb joined #moarvm
06:19 domidumont joined #moarvm
06:24 nebuchadnezzar joined #moarvm
07:23 zakharyas joined #moarvm
12:07 nebuchadnezzar hello
12:15 timotimo o/
12:22 TheLemonMan joined #moarvm
12:37 nebuchadnezzar if I rebuild moarvm from dyncall to libffi, should I rebuild nqp and rakudo ?
12:38 timotimo no need
12:38 nebuchadnezzar great, thanks
12:39 zakharyas joined #moarvm
12:43 TheLemonMan left #moarvm
15:12 jnthn joined #moarvm
16:26 domidumont joined #moarvm
17:24 dalek MoarVM/even-moar-jit: 7da475f | brrt++ | src/jit/linear_scan.c:
17:24 dalek MoarVM/even-moar-jit: Treat the different copy/phi nodes separately
17:24 dalek MoarVM/even-moar-jit:
17:24 dalek MoarVM/even-moar-jit: They were treated as the same, but because they are subtly
17:24 dalek MoarVM/even-moar-jit: different, they are best treated differently
17:24 dalek MoarVM/even-moar-jit: review: https://github.com/MoarVM/MoarVM/commit/7da475fb3c
17:24 dalek MoarVM/even-moar-jit: f24555a | brrt++ | src/jit/ (2 files):
17:24 dalek MoarVM/even-moar-jit: Change some types and format strings to make clang happy
17:24 dalek MoarVM/even-moar-jit:
17:24 dalek MoarVM/even-moar-jit: Contrary to my expectations, MVMint8 != char
17:24 dalek MoarVM/even-moar-jit: review: https://github.com/MoarVM/MoarVM/commit/f24555a7d2
17:24 dalek MoarVM/even-moar-jit: b308641 | brrt++ | / (3 files):
17:24 dalek MoarVM/even-moar-jit: Implement MVM_JIT_IF and MVM_JIT_EITHER via tiling
17:24 dalek MoarVM/even-moar-jit:
17:29 timotimo is that a rebase or new commits?
17:39 brrt joined #moarvm
17:39 brrt that's new commits
17:39 brrt i finally had a seat on the train :-)
17:40 timotimo cool!
17:40 timotimo yay for trains
17:40 timotimo with seats on them
17:40 timotimo but not actually *on*
17:43 brrt haha
17:45 * brrt is going to get icecream
17:45 timotimo yay!
18:22 FROGGS joined #moarvm
19:01 domidumont joined #moarvm
19:16 BinGOs_ joined #moarvm
19:16 TimToady joined #moarvm
19:23 moritz joined #moarvm
19:23 nebuchadnezzar joined #moarvm
19:23 btyler joined #moarvm
19:23 stmuk_ joined #moarvm
19:23 lizmat joined #moarvm
19:23 domidumont joined #moarvm
19:23 pyrimidine joined #moarvm
19:23 harrow joined #moarvm
19:23 FROGGS joined #moarvm
19:23 Util_ joined #moarvm
19:23 [Coke] joined #moarvm
19:23 krunen_ joined #moarvm
19:23 ggoebel joined #moarvm
19:23 arnsholt joined #moarvm
19:23 zostay joined #moarvm
19:23 JimmyZ joined #moarvm
19:23 avar joined #moarvm
19:23 geekosaur joined #moarvm
19:23 sivoais joined #moarvm
19:23 ilmari joined #moarvm
19:23 ChanServ joined #moarvm
19:23 synopsebot6 joined #moarvm
19:23 nwc10 joined #moarvm
19:23 jnthn joined #moarvm
19:23 dalek joined #moarvm
19:23 mst joined #moarvm
19:23 camelia joined #moarvm
19:23 konobi joined #moarvm
19:23 krunen joined #moarvm
19:23 _longines joined #moarvm
19:23 leego joined #moarvm
19:23 hoelzro joined #moarvm
19:23 timotimo joined #moarvm
19:23 japhb joined #moarvm
19:23 mtj_ joined #moarvm
19:23 nine joined #moarvm
19:25 BinGOs joined #moarvm
19:50 zakharyas joined #moarvm
21:14 Ven_ joined #moarvm
21:22 Ven_ joined #moarvm
21:29 ilbot3 joined #moarvm
21:29 Topic for #moarvm is now https://github.com/moarvm/moarvm | IRC logs at  http://irclog.perlgeek.de/moarvm/today
21:30 BinGOs joined #moarvm
21:33 TimToady joined #moarvm
21:36 Ven_ joined #moarvm
22:27 Zoffix joined #moarvm
22:49 Zoffix How would you go about this: we have a function that recurses and there's a case to trigger a segfault by recursing to deep?
22:49 Zoffix m: 7 ~ "\x[308]" x 110_000
22:49 camelia rakudo-moar 626a22: OUTPUT«(signal SEGV)»
22:49 Zoffix This is recursing in twiddle_trie_node() that eventually gets a stackoverflow
22:50 timotimo if we used the C stack, you could run your stack into your heap, but that's not a SEGV usually
22:50 timotimo oh
22:50 timotimo that is the C stack
22:50 timotimo so maybe that's it
22:51 Zoffix So to fix it we need to "run stack into heap"?
22:52 timotimo oh, no, that's how it explodes
22:52 Zoffix oh
22:52 timotimo there's one big area of memory. at one end lives the heap, at the other lives the stack
22:52 timotimo both grow towards each other
22:52 timotimo well, in the realm of virtual memory it's all a bit more interesting
22:52 timotimo but when your stack reaches your heap and you try to put more stuff on the stack ... boom
22:53 Zoffix K, I think I'll leave this for now and try something easier :)
22:53 Zoffix That was this ticket: https://rt.perl.org/Ticket/Display.html?id=129227#ticket-history
22:55 timotimo since the algorithm to change this trie is based on traversing it down and then up, and the nodes of the trie don't have an "up" pointer (i haven't looked, but i guess this is the case), we'd have to introduce our own data structure to facilitate "going up"
22:56 timotimo essentially, we'd make the same thing the stack was, but hopefully a bit more memory efficient, and swap-out-able
22:56 timotimo i need to go to bed
22:57 timotimo toodles
22:59 Zoffix night
23:04 konobi madvise?
23:05 konobi mmm... not what i was thinking of

| Channels | #moarvm index | Today | | Search | Google Search | Plain-Text | summary