|
01:48
ilbot3 joined
05:24
domidumont joined
05:54
domidumont joined
07:22
zakharyas joined
09:57
zakharyas joined
13:06
zakharyas joined
13:24
dalek joined
15:37
lizmat joined
17:27
zakharyas joined
18:02
FROGGS joined
19:52
Ven_ joined
|
|||
| dalek | arVM/even-moar-jit: ff5f25a | LemonBoy++ | src/io/syncfile.c: Fix mvm_tell for files that haven't been read from. |
20:16 | |
| arVM/even-moar-jit: 2793d53 | paultcochrane++ | src/profiler/heapsnapshot.c: Fix typo in comment |
|||
| arVM/even-moar-jit: b005d7e | paultcochrane++ | src/6model/ (4 files): Remove unused variables (src/6model) GCC noticed that these variables were unused; removing them did not cause any test failures in the nqp test suite. This commit is restricted to the `src/6model` directory in order to keep the patch size small. Fix mvm_tell for files that haven't been read from. |
|||
| arVM/even-moar-jit: 3d04391 | jnthn++ | src/io/syncsocket.c: Mark thread GC blocked while accepting a socket. This means that other threads doing work will not end up with their GC runs hanging until a connection comes in. |
|||
| arVM/even-moar-jit: 7b6c092 | moritz++ | Configure.pl: Configure: explicit use lib ".". Closes #403. |
|||
| arVM/even-moar-jit: ef129c4 | jnthn++ | src/core/threadcontext.c: Fix missing finalization queue cleanup. Only a small leak, but it adds to the clutter when looking at valgrind leak check output. |
|||
| arVM/even-moar-jit: 4938b65 | jnthn++ | src/core/bytecode.c: Mark SC used in bytecode loading as claimed. Otherwise, it might be considered unclaimed and so go uncollected, thus leading to things like EVAL leaking memory. |
|||
| arVM/even-moar-jit: 2eedba8 | jnthn++ | src/6model/ (2 files): Fix a-bit-too-few error in multi cache. The indexes we store for args are indexes into the args buffer (which includes names) rather than the callsite flags, meaning that they can be up to twice as long. Fixes a bug where we could fail to find some entries when caching callsites with many named args, resulting in the same entries being put into the cache again and again. |
|||
| arVM/even-moar-jit: b8a8e6b | brrt++ | / (20 files): Merge remote-tracking branch 'origin/master' into even-moar-jit |
|||
|
20:22
Ven__ joined
21:25
stmuk joined
21:48
hoelzro joined
22:02
lizmat joined
|
|||
| timotimo | huh. i've got MVM_JIT_EXPR_ENABLE=yes but i don't see the pretty outputs in the jitlog ?!? | 22:47 | |
| ah | 22:49 | ||
| it goes through itoa, so it has to be numeric | |||
| neat little trick of the day: select a digraph (% key helps to jump between braces for the selection) and do :w! dot -Tx11 | 22:56 | ||
| (when you have a selection in vim and hit :, it'll automatically insert the "use selection for this command" thing at the beginning) | 22:57 | ||
| i think i might turn that into a keybind for when i look at a jit log in vim | |||
| decont and wval are the top 2 "Cannot get template for" lines in my little test case | 23:07 | ||
| the next one after that causes only half as many graphs to be broken up: prepargs, very closely followed by sp_p6ogetvt_o | 23:08 | ||
| i imagine takedispatcher would be sort of easy to implement as a template. it's 6th with 291 break-offs (top spot has 935) | 23:10 | ||
| lastexpayload + return_o could be a nice combo to have; return_o is in the 5th spot (but potentially not easy to do?), and the 14th spot goes to lastexpayload (but pretty much every lastexpayload is immediately followed by return_o) | 23:13 | ||
| and with that i'll go to bed | |||
| a big chunk of break-offs during the beginning (i.e. inside the compiler) are from box_i (coming directly after hllboxtype_i, which *is* implemented) | 23:15 | ||