|
01:12
btyler joined
01:21
FROGGS joined
02:23
jnap joined
02:26
donaldh joined
02:32
ashleydev joined
03:55
jnap joined
04:56
jnap joined
05:57
jnap joined
06:55
zakharyas joined,
FROGGS joined
06:57
jnap joined
07:58
jnap joined
08:17
cognominal__ joined
08:18
ggoebel111113 joined,
zakharyas joined
08:21
hoelzro_ joined
08:53
lue joined,
hoelzro_ joined,
zakharyas joined,
ggoebel111113 joined,
cognominal__ joined,
FROGGS joined,
ashleydev joined,
donaldh joined,
colomon joined,
daxim joined,
dalek joined,
cxreg joined,
odc joined,
woolfy joined,
harrow joined,
flussence joined,
_sri joined,
timotimo joined,
brother joined,
bcode joined,
tadzik joined,
[Coke] joined,
Util joined,
synopsebot joined,
camelia joined,
sergot joined,
japhb joined,
jnthn joined,
tokuhirom joined,
xiaomiao joined,
retupmoca joined,
TimToady joined,
dagurval_ joined,
avar joined,
ingy joined,
moritz joined,
BinGOs joined,
JimmyZ joined,
lee_ joined,
masak joined,
nwc10 joined,
ChanServ joined
08:59
jnap joined
09:01
lizmat joined
09:10
lue joined
10:00
jnap joined
11:00
jnap joined
11:01
jnap1 joined
11:03
jnap joined
12:04
jnap joined
12:45
jnap joined
|
|||
| dalek | arVM/loop_labels: 0612081 | (Tobias Leich)++ | src/core/exceptions.c: make rules for searching for handlers clear That is, we either need to hit the correct label to continue, or otherwise need to have an exception that does not require a label. |
13:20 | |
| arVM/loop_labels: e4b6d79 | (Tobias Leich)++ | src/core/exceptions. (2 files): add const for handlers that check for responsibility |
13:24 | ||
| arVM/loop_labels: ba299b2 | (Tobias Leich)++ | lib/MAST/Nodes.nqp: add const for handlers that check for responsibility |
13:27 | ||
| Heuristic branch merge: pushed 18 commits to MoarVM/loop_labels by FROGGS | 13:37 | ||
|
13:58
btyler joined
14:13
jnap joined
14:25
donaldh joined,
donaldh_ joined
14:47
LLamaRider joined
15:37
dalek joined
15:58
[1]Ryan joined
16:01
btyler_ joined
16:03
[1]Ryan left
18:39
colomon joined
|
|||
| nwc10 | jnthn: OK, so it seems that Rakudo does not like it if MoarVM was compiled with char being unsigned | 18:55 | |
| Stage parse : Could not locate compile-time value for symbol Whatever | |||
| verified on x86 and x86_64 | 18:56 | ||
| when the thunderstorm is over, I'll power up the Pi, and see if -fsigned-char there gets us a build | |||
| jnthn | Hmmm | ||
| nwc10 | Yes, mmmm. | ||
| I don't know how to figure that one out | |||
| not expecting you to say "do this" | |||
| jnthn | I thought we had MVMuint8 and MVMint8 to try and make sure stuff got what it wanted. | ||
| nwc10 | what I mean is, I'm not sure how to even try to bisect it | 18:57 | |
| jnthn | Maybe check that things are being set up properly in that regard. | ||
| nwc10 | might be it - treat 'char' as a 4 letter word | ||
| and remove the expletives :-) | |||
| ASAN doesn't throw a wobbly, so it won't help | 19:02 | ||
| is there a way to get gdb to print every file/line as its executed? | |||
| ie a bit like hammering 's' (for step) | 19:03 | ||
| because the diff for two execution traces is one way to zoom in on Perl code bugs | |||
|
19:04
LLamaRider joined
19:14
FROGGS[mobile] joined
20:29
LLamaRider joined
21:39
LLamaRider joined
21:45
btyler joined
|
|||
| dalek | arVM/spesh_trace: 19dbdf4 | jnthn++ | / (6 files): Add lexical lookup lifetime hinting ops. Allow us to convey that a given lexical lookup can be relied on to always resolve to the same thing, or can be relied on to resolve to the same thing per invocant type. The former is useful for setting sub lookups, the latter for type variables. For now don't do anything extra, but spesh will be able to improve on this. |
22:55 | |
| arVM/spesh_trace: 87a4c4c | jnthn++ | / (5 files): Speculate some speculative optimization spesh ops. |
22:56 | ||