|
01:48
ilbot3 joined
03:45
mst joined
09:19
brrt joined
09:41
Ven joined
09:53
Ven_ joined
10:12
colomon joined
10:26
Ven joined
10:40
Ven_ joined
10:51
BinGOs_ joined
10:52
hoelzro_ joined
11:00
brrt joined
11:11
BinGOs joined
11:53
colomon joined
12:11
Ven joined
12:27
colomon joined
12:28
Ven joined
12:35
FROGGS joined
12:45
Ven joined
12:51
colomon joined
13:03
brrt joined
13:05
Ven joined
|
|||
| dalek | arVM: 72a4c1a | timotimo++ | tools/moar-gdb.py: make moar-gdb.py source-able, instead of only autoloadable when not autoloading, "current_objfile" has no meaning. |
13:50 | |
|
13:56
colomon joined
13:59
colomon joined
14:03
colomon joined
|
|||
| timotimo | gdb is annoying to work with sometimes | 14:07 | |
| set logging file /tmp/foobar.txt, break in MVM_frame_find_contextual_by_name, commands "print name"; c; end | |||
| run it for a few minutes | 14:08 | ||
| turns out, i forgot to "set logging on" | |||
| fair enough. start over, run it. | |||
| turns out "start" will clear out plugins, so the names were just pointer values | |||
| there doesn't seem to be a simple way to make the big line about breakpoints being hit from the output so i could just have the results for names | 14:09 | ||
| result: table flipped. the results would probably not have been terribly interesting anyway | 14:11 | ||
| $*W shows up a bunch of times, $*ACTIONS, too. then a bunch of others. | 14:12 | ||
| we wanted to have the ACTIONS linked to from ParseShared, didn't we? i wonder if that's still an interesting optimization right now | |||
|
14:12
colomon joined
14:19
colomon joined
14:29
colomon joined
14:34
colomon joined
|
|||
| nine | timotimo: if you find gdb annoying try working with jdb for some time :) | 14:38 | |
| timotimo | ugh | ||
| jdb is probably just an afterthought | 14:39 | ||
| brrt | seconded | ||
| timotimo | "we should probably also let people who don't have an IDE debug their shit" | ||
|
14:40
colomon joined
|
|||
| jnthn | printf for the win! :P | 14:44 | |
| timotimo | what attracted me to the idea of using gdb is you don't have to rebuild, and you have sort of a repl-like experimentation setup for your stuff | 14:45 | |
|
14:46
colomon joined
|
|||
| nine | jnthn: printf debugging is much more fun when the time between adding the printf and looking at the result is significantly < 10 minutes ;) | 14:47 | |
| timotimo | ah, yeah | 14:48 | |
| well, on moarvm when you change a .c file you only need to rebuild that single file and re-link, plus the relatively quick install step | |||
| brrt | gdb is awesome | 14:52 | |
| or at least the principle is | 14:53 | ||
| lldb is on the other hand clearly not meant for human use | |||
| timotimo | beh. and that's all people have on osx? | ||
|
14:56
zakharyas joined
|
|||
| dalek | arVM/even-moar-jit: ac8b1e1 | brrt++ | src/jit/ (5 files): Simplify the tile function signature Tile objects already contain all the necessary values, so we can just pass it rather than push all its elements. |
14:58 | |
| arVM/even-moar-jit: ba7f715 | brrt++ | src/jit/ (2 files): Use static register buffers No need to allocate fixed-size buffers |
|||
| brrt | no, you can use gdb, but you have to jump through hoops | ||
| timotimo | OK :( | 14:59 | |
| brrt | well, fortunately, i don't have to deal with that anymore (in the recent future at least) | ||
| near future of course | 15:02 | ||
|
15:02
colomon joined
|
|||
| brrt | i'm watching too much doctor who | 15:02 | |
| timotimo | :D | ||
| standing with your back to the future? | 15:03 | ||
| jnthn | nine: Yes, very true... | ||
| jnthn does use gdb, the VS debugger, etc. also :) | |||
| Though in terms of tooling, the valgrind suite is perhaps what I reach for most these days | 15:04 | ||
| brrt | :-) | 15:05 | |
| i'm going to go and ehm, have dinner or something | |||
| jnthn | om nom :) | 15:06 | |
| jnthn is feeling super lazy and will order dinner to home today :) | |||
| Got some nice cooking planned for next week though :) | |||
| timotimo | the usual time/money trade-off ... :) | 15:07 | |
| brrt is that lazy more often than he would like | |||
| aye, and energy | |||
| jnthn | Yeah, energy is my problem. | ||
| timotimo | right, makes sense | ||
| jnthn | Slept really well last night, better than for a while. And my body is doign the "oooh, rest is good, let's keep doing that" thing :) | ||
| *doing | |||
| timotimo | i have friends who live right in my neighbourhood. they have cats, and we cook & eat dinner together almost every day now | ||
| jnthn parses the sentence the wrong way, and has a mental image of timotimo cooking together with a bunch of cats | 15:08 | ||
| timotimo | well, the cats surely try to "help" with the ingredients | ||
| jnthn | ;-) | ||
| timotimo | "oh, your meat isn't chewed up yet? let me help with that" | 15:09 | |
| "oops, it's gone" | |||
| jnthn | ;) | 15:10 | |
|
16:34
Ven joined
16:43
Ven joined
17:03
Ven joined
17:21
vendethiel joined
17:22
Ven joined
17:42
Ven joined
17:46
brrt joined
|
|||
| brrt | rats are much like that, really | 17:55 | |
|
18:12
Ven joined
18:32
Ven joined
|
|||
| timotimo | i suppose many animals are | 18:35 | |
| except, some can be taught to not be as forward about their urges | |||
|
18:56
zakharyas joined
19:00
colomon joined
19:33
mst joined
|
|||
| nine | even some humans | 19:36 | |
| geekosaur | executive function is always tricky and cranky | 19:44 | |
|
22:45
colomon joined
|
|||