|
00:02
FROGGS joined
00:22
tokuhiro_ joined
00:34
tokuhiro_ joined
00:53
JimmyZ joined
02:27
FROGGS joined
05:07
FROGGS joined
05:33
ingy joined
06:43
FROGGS joined
07:07
brrt joined
07:19
Ven joined
|
|||
| brrt | \o | 07:38 | |
| nwc10 | o/ | 07:39 | |
| brrt | i have good news! dynasm is broken | 07:46 | |
| more importantly | |||
| i have golfed down where it is broken | |||
| FROGGS | brrt++ | 07:47 | |
| brrt | it's kind of weird, actually | 07:49 | |
| but, i'm sure i'll figure it out | |||
| FROGGS | well, otherwise it would be boring | 07:50 | |
| brrt | that's true, i guess | 07:52 | |
| there's supposed to be an extra extra byte in movs? it's weird, really | |||
| FROGGS | I have no knowledge whatsoever about asm | 08:01 | |
| psch | i have enough asm knowledge to be dangerous! | ||
| ...not enough to be useful, though | 08:02 | ||
| FROGGS | *g* | 08:03 | |
| psch | reminds me of my neglected H6809 emulator, too | 08:05 | |
| i had to drop out of uni though, so i don't think i'll continue it | |||
| brrt | asm is fun | 08:18 | |
| FROGGS | so is pir | ||
| JimmyZ | pasm :) | 08:19 | |
| jnthn | Don't give me nightmares :P | 08:20 | |
| timotimo | brrt: ah, good news indeed. good news that you found it, in any case | 08:21 | |
| brrt | the difference between PIR and real asm is that asm has viable implementations :-P | 08:22 | |
| well, now i just have to figure out why, you know? | |||
| timotimo | yeah | ||
| brrt | and i'm under no illusion that this is the last one | ||
| timotimo | but it seems like you have somewhat of a golf | ||
| i'm just not so happy dynasm keeps stealing so much of your time | 08:23 | ||
| jnthn | www.moarvm.org/releases/MoarVM-2015.08.tar.gz now in place | 08:32 | |
|
08:42
Ven joined
08:54
zakharyas joined
|
|||
| dalek | arVM: c14a568 | jnthn++ | src/core/frame.c: Remove debugging info from error. |
08:58 | |
|
09:05
lizmat joined
|
|||
| dalek | arVM: 4b427ed | jnthn++ | src/core/threads.c: Make thread join force a GC run. Avoids sitting in a loop starting/joining threads gobbling resources due to a GC never being triggered. |
09:26 | |
| timotimo | oh, that's a good point | 09:27 | |
|
09:45
Ven joined
10:28
Ven joined
11:23
virtualsue joined
11:43
virtualsue left
12:36
brrt joined
|
|||
| brrt | \o | 12:36 | |
| nwc10 | o/ | ||
| brrt | timotimo: don't worry about that, it's what i do :-) | ||
| i'm wondering, it's quite probably worth my while to enumerate all permissisble bytecode segments in dynasm | 12:37 | ||
| then compile it, make a dump, extract the information from objdump, and diff them | 12:38 | ||
| seems like a job for.... | |||
| THE DUCT TAPE CALLED PERL (c) | |||
| timotimo | <3 | 12:39 | |
| (yes, i am forbidding you from using perl4, perl5 or perl6. deal with it.) | |||
| brrt | why perl6? | 12:40 | |
| then it's NQP for me :-P | |||
| timotimo | .o( just kidding ) | 12:41 | |
| psch | perl3 shoulda been forbidden too, no? | 12:44 | |
| timotimo | indeed | ||
| FROGGS .oO( perl<3 ) | 13:39 | ||
|
13:45
FROGGS[mobile] joined
14:07
brrt joined
14:11
tokuhiro_ joined
14:12
dalek joined
14:25
FROGGS joined
14:44
dalek joined
14:45
pyrimidine joined
14:58
camelia joined
15:11
tokuhiro_ joined
|
|||
| jnthn | Interesting stat: inlining currently eliminates 10% of call frames in Text::CSV | 16:21 | |
| FROGGS | that's... a lot, isnt it? | 16:23 | |
| jnthn | FROGGS: Aye | 16:24 | |
|
16:45
brrt joined
|
|||
| JimmyZ | hmm, I got a segfault from t/spec/S17-lowlevel/thread.t | 16:46 | |
| Program received signal SIGSEGV, Segmentation fault. | 16:48 | ||
| 0x00007ffff78b4037 in MVM_thread_id (tc=0x603770, thread_obj=0x2fae4d0) at src/core/threads.c:171 | |||
| 171 return ((MVMThread *)thread_obj)->body.tc->thread_id; | |||
| p ((MVMThread *)thread_obj)->body | 16:49 | ||
| $3 = {invokee = 0x0, thread = 140737231140608, tc = 0x0, next = 0x2fae430, stage = 6, native_thread_id = 140737231140608, app_lifetime = 0} | |||
|
17:14
vendethiel joined
|
|||
| JimmyZ | jnthn: github.com/MoarVM/MoarVM/commit/4b...e80677544b # this one cause the segfault | 17:15 | |
|
18:21
ShimmerFairy joined
|
|||
| jnthn | Why on earth does method new in MapIterCommon not get spesh'd at all despite being hot?! | 19:14 | |
|
19:20
orbusbak joined
19:21
orbusbak joined
|
|||
| [Coke] | ./perl6 --profile #hotornot | 19:30 | |
|
19:44
vendethiel joined
20:21
tokuhiro_ joined
20:35
colomon joined
21:03
zakharyas joined
21:05
brrt joined
21:34
colomon joined
21:55
lizmat joined
22:40
FROGGS joined
|
|||
| japhb has a sudden desire to make a hot or not spoof containing profile segments | 23:33 | ||