Perl 6 - the future is here, just unevenly distributed

IRC log for #moarvm, 2015-11-19

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

All times shown according to UTC.

Time Nick Message
00:33 tokuhiro_ joined #moarvm
01:01 kjs_ joined #moarvm
02:24 ggoebel joined #moarvm
02:35 tokuhiro_ joined #moarvm
02:47 ilbot3 joined #moarvm
02:47 Topic for #moarvm is now https://github.com/moarvm/moarvm | IRC logs at  http://irclog.perlgeek.de/moarvm/today
03:05 tokuhiro_ joined #moarvm
03:39 vendethiel joined #moarvm
05:39 xiaomiao joined #moarvm
07:04 domidumont joined #moarvm
07:09 domidumont joined #moarvm
07:20 FROGGS joined #moarvm
08:14 brrt joined #moarvm
08:14 brrt good *
08:14 brrt btw, nwc10, if you keep on giving me articles like that, i won't ever finish my blog post :-P
08:37 FROGGS Arrrgh!
08:37 FROGGS bad nwc10! ò.ó
08:37 FROGGS :P
08:37 brrt i do wonder why the results have been so varying
08:38 FROGGS electromagnetic forces?
08:38 * FROGGS has no idea what the topic is about
08:41 brrt oh, the topic is 'retrofitting existing p-type languages on llvm'
08:41 brrt which has seen less-than-awesome results in various projects, despite llvm being without doubt an excellent quality compiler system
08:51 zakharyas joined #moarvm
09:00 zakharyas joined #moarvm
09:02 xiaomiao joined #moarvm
09:07 dalek Heuristic branch merge: pushed 58 commits to MoarVM/even-moar-jit by bdw
09:19 dalek MoarVM/even-moar-jit: ea87b0d | brrt++ | src/jit/graph.c:
09:19 dalek MoarVM/even-moar-jit: Fix references to jgb
09:19 dalek MoarVM/even-moar-jit:
09:19 dalek MoarVM/even-moar-jit: jgb has been killed in even-moar-jit in favor of the spesh iterator
09:19 dalek MoarVM/even-moar-jit: review: https://github.com/MoarVM/MoarVM/commit/ea87b0d5ec
09:26 diakopter brrt++ # rakudo builds with even-moar-jit branch
09:26 brrt \o/
09:36 JimmyZ I think dynasm is small and good enough, since luajit is very awesome, but i don't think llvm is small, and I din't see any popular dyn lang use llvm keep it  samll and fast enough. :P
09:38 JimmyZ and we don't have enough resource to make moar/dynasm good enough, so we don't have any time to do moar on llvm
09:39 brrt but dynasm and llvm are quite different things, i think?
09:39 brrt one is an assembler which has you specifiy all the instructions, and the other is a compiler which explicitly tells you not to care
09:40 JimmyZ yes. if someone want to do experimence moar on llvm, I don't think we will be mind :P
09:40 JimmyZ just fork moar
09:41 brrt hmm
09:43 JimmyZ or better, like Rperl, one can do  Perl 6 on llvm :P
09:45 jnthn Just because llvm is an excellent quality compiler system doesn't mean it'll naturally suit all languages.
09:46 jnthn I suspect part of the issue may be that dynamic language impls have different priorities.
09:48 jnthn For dynlangs, your performance comes greatly from spotting ways to remove indirections based on runtime data, rather than considering all the statically available data and throwing all you can at it.
09:50 jnthn So even-moar-jit ain't just exciting 'cus it will let us generate better code, it's also 'cus the expression trees will let us optimize away indirections in places we can't now.
09:51 JimmyZ I could see it :P
09:51 brrt aye, that is also what i want :-)
09:52 brrt (in a somewhat cross-platform way too :-))
09:53 JimmyZ spoke of luajit, I was not porting the code  from luajit to moar, I wanted to port the optimization algorithm :)
09:54 JimmyZ most is about AA and escape
09:54 brrt if you know how to do that... sure
09:54 brrt but there are quite a few moar things you want to take into account
09:54 JimmyZ yes
09:55 brrt i'm thinking of continuations, mostly
09:55 JimmyZ and I found  Trace spesh may make things   a bit easier.
09:56 brrt and dynamic lexicals
09:56 brrt i do think that is true yes
09:56 JimmyZ TraceJIT , that is it
09:56 brrt if nobody else has done it before i'm finished with my theses i'll make a stab at it
09:57 jnthn :)
09:57 JimmyZ well , if merge dynamic lex => local var in rakudo branch, dyn lex is not a a big problem
09:58 jnthn dynamic lexicals can never become locals
09:58 JimmyZ oh
09:58 jnthn Not in any useful case.
09:58 JimmyZ static lexical
09:58 jnthn They may be interesting in trace JIT though :)
09:59 JimmyZ anyway, I forgot the name, but there is a branch there :P
09:59 JimmyZ which will reudce getlex op
10:00 JimmyZ english still is hard to me .
10:00 nwc10 ASAN likes even-moar-jit. OK, ASAN *tolerates* it :-)
10:02 nwc10 JimmyZ: your English is still way better than any foreign language I can write
10:02 kjs_ joined #moarvm
10:33 brrt yeah, i know it leaks nwc10 :-)
10:33 brrt 'why haven't you fixed it then...'
10:34 brrt it wasn't a priority yet
10:43 kjs_ joined #moarvm
11:02 dalek MoarVM: 836f87e | (Pawel Murias)++ | / (18 files):
11:02 dalek MoarVM: Implement a istty_fh op.
11:02 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/836f87e0ad
11:46 timotimo o/
11:47 timotimo so ... are we really able to build rakudo with even-moar-jit?
12:02 timotimo damn you, Configure.pl
12:02 timotimo i can't run nqp, the serialization format in there is too high
12:02 timotimo with even-moar-jit :(
12:02 jnthn o.O
12:04 timotimo wait
12:05 timotimo the new commits weren't merged in %)
12:05 timotimo because git pull from a different branch doesn't destroy things
12:05 timotimo but isn't there still an env var that needs to be set so that the new jit activates?
12:08 timotimo 158:    jit_expr_enable = getenv("MVM_JIT_EXPR_ENABLE");
12:09 timotimo if people have been testing stuff without that set, they may have not actually tested the new jit at all ...
12:09 nwc10 er, OK, ASAN likes the code *without* that set...
12:11 timotimo hah
12:11 timotimo that's easy
12:15 timotimo the exprjit fragments are all so tiny :(
12:22 timotimo well, that's just because so many tiles were ripped out
12:42 timotimo .tell brrt if we can somehow know which nodes have a "size" parameter (and which one it is), we could put it into the same node in the tree so that things get a little less cluttered
12:43 timotimo and perhaps we could consider different node shaps for different things
12:50 kjs_ joined #moarvm
12:58 nwc10 .tell timotimo no bot here
13:02 timotimo ah
13:03 timotimo maybe we should finally invite yoleaux in here? or perhaps it's a good idea to only have it in #perl6 - unless yoleaux also does .tell across channels?
13:04 timotimo so you don't have to .botsnack in both channels to get every message
13:04 timotimo because you could forget that
13:10 nwc10 brrt: http://paste.scsys.co.uk/501659 -- MVM_JIT_EXPR_ENABLE + ASAN = Barf
13:11 jnthn ASAN++
13:11 nwc10 seems to be a fairly simple "use after free" caused by a realloc, but I didn't dig further than reading the stack trace
13:11 timotimo oh, realloc can be nasty, aye
13:13 timotimo in assign_tile there's a note about realloc right in front of MVM_DYNARR_ENSURE_SPACE, which probably does realloc
13:14 timotimo well, i don't know how memory management works for the whole thing, but brrt surely knows what the issue will be
13:31 camelia joined #moarvm
14:20 diakopter heh, barf
15:06 tokuhiro_ joined #moarvm
15:25 zakharyas1 joined #moarvm
15:30 zakharyas joined #moarvm
15:35 jnthn Time to cut the monthly, I guess :)
15:49 dalek MoarVM: 2b9df29 | jnthn++ | docs/ChangeLog:
15:49 dalek MoarVM: ChangeLog for 2015.11.
15:49 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/2b9df29b97
15:51 dalek MoarVM: c335cdd | jnthn++ | VERSION:
15:51 dalek MoarVM: Bump VERSION file.
15:51 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/c335cdd1e0
15:53 brrt joined #moarvm
15:53 brrt \o
15:53 brrt yes, i know what it is
15:53 brrt but, here is the confusing bit
15:54 brrt *i* would think that the lvalue of the assignment is evaluated *after* the computation of the rvalue
15:54 brrt however, such is clearly not the case
15:56 brrt i should probably fix that wholesale
15:59 colomon joined #moarvm
16:06 [Coke] jnthn++ # moarvm cut
16:08 dalek moarvm.org: 6388276 | jnthn++ | / (2 files):
16:08 dalek moarvm.org: Update site for 2015.11 release.
16:08 dalek moarvm.org: review: https://github.com/MoarVM/moarvm.org/commit/6388276811
16:08 tokuhiro_ joined #moarvm
16:14 timotimo that changelog looks good
16:14 jnthn :)
16:14 jnthn Yeah, busy month
16:14 jnthn > 600 commits across Moar, NQP, and Rakudo.
16:15 timotimo fitting for the months before The Big Release
16:16 hoelzro o/ #moarvm
16:16 hoelzro oh dang, I missed the release window =/
16:16 hoelzro in other news, I got a working MoarVM on Windows last night \o/
16:16 timotimo \o/
16:17 jnthn Thankfully, we do monthly releases, so you don't have that long to wait ;)
16:18 hoelzro =)
16:19 arnsholt There was a neat blogpost about the benefits of timed releases frmo the Postgres people a while back
16:19 arnsholt Basically saying that there were less rushed features when the releases were entirely predictable
17:12 ilmari perl5 sees that too
17:12 ilmari missing a release is less critical when you know there's going to be another next year
17:13 ilmari also, we've got better at marking new features experimental for the first few releases
17:13 ilmari or sometimes retroactively *cough* smartmatch *spit*
17:17 domidumont joined #moarvm
17:26 synbot6 joined #moarvm
17:53 Peter_R joined #moarvm
18:00 nwc10 seems that this month's moarvm release has completed before this month's parrot release
18:04 diakopter well there was only one commit since last month's release
18:08 xiaomiao joined #moarvm
18:13 tokuhiro_ joined #moarvm
19:13 colomon joined #moarvm
19:16 vendethiel joined #moarvm
19:28 diakopter___ joined #moarvm
19:34 kjs_ joined #moarvm
19:54 FROGGS joined #moarvm
19:55 dalek MoarVM/even-moar-jit: 836f87e | (Pawel Murias)++ | / (18 files):
19:55 dalek MoarVM/even-moar-jit: Implement a istty_fh op.
19:55 dalek MoarVM/even-moar-jit: review: https://github.com/MoarVM/MoarVM/commit/836f87e0ad
19:55 dalek MoarVM/even-moar-jit: 2b9df29 | jnthn++ | docs/ChangeLog:
19:55 dalek MoarVM/even-moar-jit: ChangeLog for 2015.11.
19:55 dalek MoarVM/even-moar-jit: review: https://github.com/MoarVM/MoarVM/commit/2b9df29b97
19:55 dalek MoarVM/even-moar-jit: c335cdd | jnthn++ | VERSION:
19:55 dalek MoarVM/even-moar-jit: Bump VERSION file.
19:55 dalek MoarVM/even-moar-jit: review: https://github.com/MoarVM/MoarVM/commit/c335cdd1e0
19:55 dalek MoarVM/even-moar-jit: c18a3d1 | (Matthew Wilson)++ | / (20 files):
19:55 dalek MoarVM/even-moar-jit: Merge pull request #302 from MoarVM/master
19:55 dalek MoarVM/even-moar-jit:
19:56 diakopter___ lazy diakopter--
20:01 vendethiel joined #moarvm
20:11 brrt joined #moarvm
20:14 tokuhiro_ joined #moarvm
20:21 lizmat joined #moarvm
20:33 tokuhiro_ joined #moarvm
20:45 kjs_ joined #moarvm
21:40 leont joined #moarvm
21:50 dalek MoarVM/even-moar-jit: 42c02d1 | brrt++ | src/jit/ (4 files):
21:50 dalek MoarVM/even-moar-jit: Fix use after free
21:50 dalek MoarVM/even-moar-jit:
21:50 dalek MoarVM/even-moar-jit: Apparantly the lvalue of tree->nodes[child] was evaluated *before*
21:50 dalek MoarVM/even-moar-jit: the rvalue of assign_tile, and assign_tile may realloc the nodes
21:50 dalek MoarVM/even-moar-jit: array. All this led to ASAN being angry (and rightly so!). This is
21:50 dalek MoarVM/even-moar-jit: fixed by providing a single macro to do it correctly.
21:50 dalek MoarVM/even-moar-jit:
21:50 dalek MoarVM/even-moar-jit: Also fix a couple of JIT-related memory leaks, including the
21:50 dalek MoarVM/even-moar-jit: nonclean destruction of graphs that had trees, and the branches
21:50 dalek MoarVM/even-moar-jit: stack that wasn't freed in the register allocator.
21:50 dalek MoarVM/even-moar-jit:
21:50 dalek MoarVM/even-moar-jit: nwc10++ for notifying this issue
21:50 dalek MoarVM/even-moar-jit: review: https://github.com/MoarVM/MoarVM/commit/42c02d15a3
21:57 brrt asan still falls over quite a few memory leaks, but, whatever
22:01 jnthn In the JIT, or generally?
22:08 brrt generally
22:09 brrt no more in the JIT, in fact
22:09 * brrt is going in the general direction of sleeping, so afk :-)
22:09 brrt see you tomorrow
22:12 * jnthn figures he'll do similar
22:12 jnthn 'night
22:35 tokuhiro_ joined #moarvm
23:11 camelia joined #moarvm

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