|
github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm Set by AlexDaniel on 12 June 2018. |
|||
|
00:14
brrt left
|
|||
| Geth | MoarVM: 729303de7c | (Timo Paulssen)++ | src/profiler/heapsnapshot.c remove left-over debug prints from heap snapshot/confprog |
01:08 | |
|
04:41
brett-soric joined
04:56
MasterDuke joined
05:30
brett-soric left
05:40
MasterDuke left
06:29
avar left,
avar joined,
avar left,
avar joined
06:45
patrickb joined
06:51
domidumont joined
07:46
zakharyas joined
08:39
sena_kun joined
10:06
zakharyas left
10:28
patrickb left
10:55
domidumont left
11:54
patrickb joined
12:28
zakharyas joined
12:33
domidumont joined
13:03
domidumont1 joined
13:05
domidumont left
13:19
brrt joined
13:33
lucasb joined
|
|||
| brrt | \o | 13:35 | |
|
13:59
brrt left
|
|||
| nwc10 | o/ | 14:00 | |
| jnthn | o/ | ||
|
14:12
pamplemousse joined
|
|||
| pamplemousse | \o | 14:13 | |
|
14:23
MasterDuke joined
14:30
domidumont joined
14:33
domidumont1 left
|
|||
| timotimo | oh zstd, your error messages are to die for … | 14:39 | |
| sorry, of course i meant "to die from" | 14:46 | ||
| the assertion i get with DEBUGLEVEL cranked up to 3 isn't a whole lot better, but i guess it's a start ... :| | 14:57 | ||
|
15:01
domidumont1 joined
15:04
domidumont left
|
|||
| timotimo | pushing bytes around, what fun | 15:08 | |
| making every single off-by-one in the book possible | 15:09 | ||
| bleehhh this buffer shuffling | 15:17 | ||
| jnthn | Could be worse. Could be using the Java ByteBuffer API :P | 15:19 | |
| timotimo | i've kind of made my own, and yeah, it's kind of bad :) | 15:20 | |
| jnthn | I dunno if it's just me, but that one seems optimized for me being very confused and getting stuff wrong | 15:21 | |
| timotimo | this problem space is able to get me very confused as well | ||
| that's the api that has .flip and such, right? | 15:22 | ||
| i just invented the acronym FBB, it's like BFF, but i haven't decided what it means yet | 15:23 | ||
| Cannot resolve caller splice(NativeCall::Types::CArray[uint8]:D, Int:D, Int:D); Routine does not have any candidates. Is only the proto defined? | 15:25 | ||
| aaw | |||
| is this ... working?!? | 15:32 | ||
| slow, tho | 15:33 | ||
| and of course it runs into another error :< | 15:38 | ||
|
15:40
AlexDaniel left,
AlexDaniel joined
15:42
dogbert11 joined
15:44
dogbert17 left
|
|||
| pamplemousse | I'm running into a bit of an issue building MoarVM, NQP, and Rakudo. I pulled all the recent changes, submodule update-ed, etc and tried to make each of them and none of them would build. I re-ran the configure script for MoarVM and it builds now but NQP and Rakudo are still not building. | 15:44 | |
| NQP complains that "MoarVM op 'freemem' is unknown as a core or extension op" and Rakudo complains that moar_core_sources does not exist. | |||
| Any thoughts? | |||
| I have no local changes | 15:45 | ||
| timotimo | that sounds like the moarvm is too old for that nqp | ||
| what does 'git show" show as the latest commit in the three repos? | 15:46 | ||
| pamplemousse | MoarVM's repo has a commit from today, NQP's from June 27th, and Rakudo's from July 1st | 15:48 | |
| timotimo | OK, perhaps the "moar" that last successfully installed is an older one | 15:49 | |
|
15:56
patrickb left
|
|||
| pamplemousse | I've installed the newly built one that I built with the changes I pulled today and I'm still running into the same problem | 16:02 | |
| timotimo | ah dang | ||
| Kaiepi | might be an issue in nqp | ||
| timotimo | perhaps something related to the freemem op actually did go wrong | ||
| Kaiepi | i was able to install moar/nqp/rakudo a few days ago so maybe something went wrong super recently | 16:03 | |
| pamplemousse | It's also possible I'm overlooking something on my end | 16:04 | |
| timotimo | you could give completely nuking whatever your --prefix is set to to the ground a try | ||
| if you do that, you can probably paste the entire console history and maybe we spot something | 16:05 | ||
| pamplemousse | I'll give it a shot | ||
| Kaiepi | i'm not sure if nqp was rebootstrapped or not when those ops were added, i think that was around the time i was told i didn't need to do that with every new op added | 16:11 | |
| timotimo | right, only if nqp itself needs the op in its compiler | 16:12 | |
| Kaiepi | should probably rebase my getsockopt/setsockopt grant pullreq that's been open forever so it doesn't rebootstrap at some point | 16:13 | |
| don't really need to worry about it until i'm towards the end of when i finish my work on IO::Socket::INET since the last changes don't make any sense without socket option support existing though | 16:14 | ||
| timotimo | mhm | ||
| dogbert11 is trying to build Rakudo on his shiny new Raspberry Pi 4 | 16:20 | ||
| pamplemousse | I'm getting the feeling it is a problem on my end. I think it probably was an issue with it being an old version of moar. I noticed that the version of moar being built inside the MoarVM repo wasn't the same as the one in my prefix/bin directory. Tried manually copying that over, forgetting that installing copies other files into other places that also need to be referenced by nqp and rakudo. | 16:26 | |
| timotimo | ooooh | 16:27 | |
| "killall moar" can fix that | |||
| of course any still-running perl6 processes can die from that | |||
| so maybe first "pidof moar" to see if there's anything, and if so, what | |||
|
16:28
squashable6 left
|
|||
| Kaiepi | never heard of either of those commands, i use pgrep and pkill | 16:29 | |
| pamplemousse | Now that I've nuked everything and trying to set it up again, I'm noticing that everything run by calling "make install" in the moar dirrectory does nothing. Any call to /usr/bin/perl -MExtUtils::Command does nothing | ||
| timotimo | o_O | 16:31 | |
| "hard drive full"? *shruuug* that's very weird | |||
|
16:32
squashable6 joined
|
|||
| pamplemousse | Nope, I definitely have space. And if I try to make the directories it's trying to make, but using mkdir instead of /usr/bin/perl -MExtUtils::Command -e mkpath, it makes the directories just fine. | 16:32 | |
|
16:33
AlexDaniel left
|
|||
| timotimo | hum. "strace -f make install" will spit out a crapton of text, but maybe something obvious pops out | 16:33 | |
|
16:34
squashable6 left
|
|||
| ugexe | pamplemousse: what if you try to create directories using that perl command, but somewhere else? | 16:34 | |
|
16:34
squashable6 joined
|
|||
| pamplemousse | I think I've figured it out! Anything that uses a ~ in the path to the directory being made fails to make the directory. So ugexe, I can make directories elsewhere and inside the MoarVM repo (thank you for asking, that's how I was able to figure out what was going wrong) using relative and absolute addressing, but nothing with a ~ | 16:44 | |
| Kaiepi | ah | 16:46 | |
| use $HOME instead, that works for me | |||
| timotimo | oh damn | ||
| i wonder if your moarvm folder now has a ~ folder in it that has everything there | 16:47 | ||
| pamplemousse | It sure does | 16:48 | |
| Thank you all, sorry I didn't catch that sooner | 16:51 | ||
| timotimo | ehh, that would't make my top 10 embarassing things i needed help with :P | 16:53 | |
| Kaiepi | ditto | 16:54 | |
|
18:01
domidumont1 left
18:02
brrt joined
18:23
zakharyas left
18:26
brrt left
18:27
squashable6 left
18:29
squashable6 joined
18:40
AlexDaniel joined
18:42
MasterDuke left
19:02
MasterDuke joined
19:43
brrt joined
20:04
Altai-man_ joined
20:06
sena_kun left
20:10
Altai-man_ left
20:42
squashable6 left
20:44
squashable6 joined
20:52
brrt left
20:53
brrt joined
20:58
lizmat joined
21:16
brrt left
21:28
AlexDaniel left
22:21
pamplemousse left,
pamplemousse joined
22:30
pamplemousse left
23:53
AlexDaniel joined
|
|||