github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm Set by AlexDaniel on 12 June 2018. |
|||
MasterDuke | jnthn, timotimo: ah ha! the SEGV when profiling does happen even without my patch! | 00:21 | |
ugexe | unix,win: add uv_gettimeofday() #2221 github.com/libuv/libuv/pull/2221 | 01:26 | |
seems like that aught to be useful for something | 01:27 | ||
01:56
zakharyas joined
06:19
robertle left
07:28
domidumont joined
07:50
robertle joined
08:52
zakharyas left
09:10
zakharyas joined
09:46
zakharyas left
|
|||
lizmat | jnthn: seems like Range.clamp could be a useful addition: stackoverflow.com/questions/552507...c-in-perl6 | 09:50 | |
11:13
domidumont left
|
|||
AlexDaniel | lizmat: perhaps start an issue on github.com/perl6/problem-solving | 12:10 | |
lizmat: IIRC it's not just clamp, there are several mathy things missing | |||
so it would be great to have a bunch of them added in v6.e in a coherent way | 12:11 | ||
lizmat | seems like you have a better idea as to what needs to be added :-) | ||
AlexDaniel | not from the top of my head | 12:12 | |
but we should start a list | |||
I can file an issue though… | |||
lizmat | ++AlexDaniel :-) | 12:13 | |
AlexDaniel | I have to run now, but that's a start: github.com/perl6/problem-solving/issues/4 | 12:17 | |
12:28
AlexDaniel left
12:36
zakharyas joined
13:22
lucasb joined,
zakharyas left
13:26
AlexDaniel joined
14:36
mgoebel joined
14:46
brrt joined
|
|||
brrt | ohai #moarvm | 14:47 | |
mgoebel: I think you reached out to me on email, correct? | 14:49 | ||
mgoebel | brrt: Yes! I'm the student that reached out with interest about the GSoC project. | 14:53 | |
brrt | cool | 14:56 | |
so, are there any questions you have so far, anything I can clarify? | 14:57 | ||
have you gotten so far as to get rakudo perl6 working on your machine? | |||
mgoebel | Yes, I downloaded and installed it a couple months ago, but don't have the latest version and am working on updating that now. | 15:00 | |
15:00
MasterDuke left
|
|||
mgoebel | I'm an undergraduate computer science senior at Georgia Tech. I did an Advanced OS course and a Compilers & Interpreters course last spring which were super hard but a ton of fun, but it was definitely just the tip of the iceberg and there's a lot I don't know. I'm most comfortable in C and I don't have a lot of experience with Perl, though I'm working on picking it up. | 15:03 | |
brrt | Cool. | 15:04 | |
mgoebel | Are there any write-ups (maybe on the compiler or related things) that might help me tackle the initial learning curve that you could point me towards? | ||
brrt | Well, not everybody here has a lot of experience with perl5, so that's hardly a requirement | ||
On the rakudo compiler, there's been a bunch of slides | 15:05 | ||
I can find them | |||
but to be honest jnthn knows a lot more about those | |||
I'm mostly the JIT guy around here | |||
mgoebel: github.com/edumentab/rakudo-and-nq...als-course | 15:06 | ||
this should give you some insight into how things work together | |||
but it is kind of old and doesn't talk a lot about MoarVM internals | |||
we have a docs folder, and istr that samcv was at some point busy making online documentation | |||
ggoebel | jnthn: is there a link to an index of your papers? | ||
brrt | ggoebel: www.jnthn.net/slides.html there's that | 15:07 | |
mgoebel | brrt: Thanks! I'll start with that and work from there. | 15:08 | |
brrt | great | ||
so I should stress, regarding the linking idea; I looked into this and it seemed doable, but it's not a predetermined path | 15:09 | ||
and there is at least one complexity still, which is that we use dynamic linking to load some extra, rakudo specific code in the VM (so called extops) | |||
now.... extops are somewhat deprecated, but they're not gone yet | 15:10 | ||
samcv | brrt, istr? | ||
brrt | 'I seem to recall' | ||
I recall seeing a documentation website that you built ... times ago | |||
samcv | yeah | 15:12 | |
well. that i got like part way done. i thought istr was some op :) | |||
brrt | hehe | ||
samcv | and i didn't remember one called istr() :-) | ||
brrt | could've been | ||
mgoebel: obviously we have the docs/ directory as well, but I don't think there's much relevant to this case | 15:13 | ||
samcv | brrt, i found malloc_trim() interesting. i had no clue that existed | ||
brrt | samcv: I think that was actually MasterDuke's suggestion; neither did I, btw | ||
samcv | This function is automatically called by free(3) in certain | 15:14 | |
circumstances | |||
M_TRIM_THRESHOLD default is 128*1024. but somehow | 15:16 | ||
maybe it just doesn't always get triggered | |||
15:18
patrickb joined
|
|||
samcv | interestingly there's the MALLOC_TRIM_THRESHOLD_ ENV var which can change the default M_TRIM_THRESHOLD | 15:19 | |
i go from 8.7 max to 6.3 with my script that cats a linux iso 5 times doing malloc_trim(128*1024); | 15:30 | ||
mgoebel | brrt: Good to know, I'll keep an eye out for other ways of doing it as well | 15:32 | |
I'll work through the rakudo and nqp course and look through the slides and will likely have a lot of questions after that | |||
bbiab | |||
brrt | later | 15:35 | |
samcv: that's a fairly big difference | 15:36 | ||
15:42
kawaii joined
16:26
patrickb left
16:42
mgoebel left
16:48
robertle left
|
|||
lizmat wonders if badootech.badoo.com/when-pigs-fly-...4fb6bfa20f is something for brrt or jnthn | 16:58 | ||
17:05
brrt left
|
|||
Geth | MoarVM: samcv++ created pull request #1072: Ask malloc to free kernel pages on GC collection |
17:08 | |
17:45
domidumont joined,
brrt joined
18:04
brrt left
18:30
lucasb left
18:31
patrickb joined
|
|||
Geth | MoarVM/master: 9 commits pushed by (Patrick Sebastian Zimmermann)++, (Patrick Böker)++, (Aleks-Daniel Jakimenko-Aleksejev)++
|
19:36 | |
19:47
Geth left,
Geth joined
|
|||
timotimo | cool! | 20:01 | |
20:06
robertle joined
20:27
ggoebel left
21:39
domidumont left
22:22
km1 joined
22:23
km1 left
23:05
patrickb left
|