03:20 camelia joined 07:25 colomon joined 07:27 lizmat joined 07:42 FROGGS joined, Ven joined
dalek arVM: 6fec043 | paultcochrane++ | src/io/procops.c:
Fix copy-and-paste error

The attribute si->ds_stderr is actually being destroyed in this part of the code (similarly to si->ds_stdout in th block above this change), thus the call to `MVM_string_decodestream_destory` should really use si->ds_stderr instead of si->ds_stdout.
07:58
arVM: 7fdde0b | FROGGS++ | src/io/procops.c:
Merge pull request #238 from paultcochrane/pr/fix-copy-paste-error

Fix copy-and-paste error
07:58 Ven joined 08:15 Ven joined 08:19 _itz_ joined 08:24 camelia joined 08:39 Ven joined 09:07 virtualsue joined 09:09 colomon joined
virtualsue can haz PR: github.com/MoarVM/MoarVM/pull/239 09:09
dalek arVM: 48f5a84 | (Sue Spence)++ | src/mast/driver.c:
Replace use of comma op inside an init of an array of pointers as this wasn't doing what was intended. Also gets rid of the last warning in this codebase (as of now).
09:17
arVM: e187d89 | (Jimmy Zhuo)++ | src/mast/driver.c:
Merge pull request #239 from virtualsue/master

Replace use of comma op inside an init of an array of pointers as thi…
[Coke] is just waiting for the travisci to finish.
JimmyZ virtualsue: thank you :)
[Coke] ... ah, JimmyZ++ got it. virtualsue++
JimmyZ well, clang is slow ...
from the travis ci .. 09:18
_itz_ clang36 is a lot faster compiling moar than earlier versions 09:24
JimmyZ yeah
09:29 lizmat joined 09:38 Ven joined
hoelzro m: gist.github.com/hoelzro/73404929704270e26891 09:41
camelia rakudo-moar 5fb81f: OUTPUT«before waithere␤␤Internal error: Unwound entire stack and missed handler␤»
hoelzro weeeeee
FROGGS lol 09:42
(actually not so loud to not to disturb people)
hoelzro hehe 09:44
we probably shouldn't even compile that program
JimmyZ p: gist.github.com/hoelzro/73404929704270e26891 09:48
r: gist.github.com/hoelzro/73404929704270e26891
camelia rakudo-jvm 5fb81f: OUTPUT«before wait␤here␤Exception in thread "Thread-4" org.perl6.nqp.runtime.LexoticException␤ at org.perl6.nqp.runtime.ThreadContext.<init>(ThreadContext.java:119)␤ at org.perl6.nqp.runtime.GlobalContext.getCurrentThreadContext(GlobalContext.java:340)␤ …» 09:49
..rakudo-moar 5fb81f: OUTPUT«before wait␤here␤Internal error: Unwound entire stack and missed handler␤»
JimmyZ looks like return cause it 09:51
timotimo yeah
hoelzro yeah, it's return 09:52
JimmyZ without start, it is also wrong 09:54
so, it is a rakudo bug, not moarvm 09:55
[Coke] ugh. I feel like we've closed more tickets than this. only at 1,039 10:09
ww. 10:10
hoelzro JimmyZ: I agree; I don't think we should be compiling that
10:22 Ven joined
dalek arVM: 601561e | FROGGS++ | src/io/procops.c:
move declaration before code
10:34
11:20 brrt joined 11:29 virtualsue joined
brrt good * 11:40
dalek arVM: 4c2c3cb | paultcochrane++ | src/core/dll.c:
Unlock mutex before returning

The locked mutex could return without being unlocked beforehand. This change adds the missing unlock and avoids a potential thread blockage.
11:42
arVM: f69883d | lizmat++ | src/core/dll.c:
Merge pull request #240 from paultcochrane/pr/unlock-mutex

Unlock mutex before returning
lizmat brrt o/ 11:44
12:10 Ven joined 12:15 virtualsue joined 12:23 rarara joined 12:40 virtualsue joined
dalek arVM: 45cae9b | paultcochrane++ | src/6model/reprs/MultiDimArray.c:
Initialise all spec attributes to default values

Not all of the `spec` struct attributes are assigned within
  `get_elem_storage_spec()`. This change sets all attributes so that none
will be potentially used uninitialised.
12:44
arVM: 10dce90 | (Jimmy Zhuo)++ | src/6model/reprs/MultiDimArray.c:
Merge pull request #237 from paultcochrane/pr/initialise-storage-spec

Initialise all spec attributes to default values
arVM: 1a4e7d8 | paultcochrane++ | src/core/exceptions.c:
Plug resource leaks in exceptions.c

Since `MVM_string_utf8_encode_C_string()` was being called inside another expression/function, it was implicitly allocating memory which wasn't being released. This change refactors the code allocate the memory explicitly so that it can be freed, while also avoiding potentially dereferencing NULL.
12:45
arVM: b8ee3c5 | (Jimmy Zhuo)++ | src/core/exceptions.c:
Merge pull request #241 from paultcochrane/pr/fix-exceptions-resource-leak

Plug resource leaks in exceptions.c 10e8783 | (Jimmy Zhuo)++ | src/6model/reprs/MultiDimArray.c: cleans up
12:48 Ven joined
dalek arVM: e87e8cc | hoelzro++ | build/setup.pm:
Make mixed declarations and statements fatal on GCC

Prevent hoelzro (me) from breaking the Windows build!
12:54
arVM: 409d30e | timotimo++ | src/core/exceptions.c:
move declaration to beginning of function
12:57
FROGGS hoelzro++
hoelzro timotimo beat me to the fix =/ 12:58
timotimo :-) 13:02
13:24 mtj_ joined 13:25 nebuchadnezzar joined, _itz joined, ShimmerFairy joined, tadzik joined 13:26 camelia joined 13:27 virtualsue joined 13:40 travis-ci joined
travis-ci MoarVM build failed. Rob Hoelz 'Make mixed declarations and statements fatal on GCC 13:40
travis-ci.org/MoarVM/MoarVM/builds/77812544 github.com/MoarVM/MoarVM/compare/1...7e8cce4e70
13:40 travis-ci left, virtualsue joined 13:44 Ven joined, colomon joined
dalek arVM: b5db76e | FROGGS++ | src/core/nativecall_libffi.c:
move declaration before code
13:49
timotimo thank ci for running libffi, too
13:51 travis-ci joined
travis-ci MoarVM build failed. Timo Paulssen 'move declaration to beginning of function' 13:51
travis-ci.org/MoarVM/MoarVM/builds/77812865 github.com/MoarVM/MoarVM/compare/e...9d30e97393
13:51 travis-ci left
FROGGS my commit should pass the thang 13:53
13:55 brrt joined
dalek arVM: be09cbe | paultcochrane++ | src/spesh/graph.c:
Remove resource leaks in spesh/graph.c

  `MVM_spesh_dump` was being used inside a `printf`, thus the implicitly
allocated memory can't be released. This refactors the code to make the allocation explicit so that the memory can be freed.
13:55
arVM: 4b284dd | FROGGS++ | src/spesh/graph.c:
Merge pull request #242 from paultcochrane/pr/plug-resource-leak-in-graph

Remove resource leaks in spesh/graph.c
13:58 colomon joined
brrt lizmat o/ 14:07
brrt has been picking low-hanging fruit today
14:08 dalek joined
brrt what :-o i stop at enter-conditional? 14:10
awesome 14:11
14:13 travis-ci joined
travis-ci MoarVM build passed. Tobias Leich 'move declaration before code' 14:13
travis-ci.org/MoarVM/MoarVM/builds/77816654 github.com/MoarVM/MoarVM/compare/4...db76e5c7a3
14:13 travis-ci left
timotimo brrt: does that mean we're going even further forward towards running jitted code? 14:13
brrt yes 14:14
onward, comrades
14:24 travis-ci joined
travis-ci MoarVM build passed. Tobias Leich 'Merge pull request #242 from paultcochrane/pr/plug-resource-leak-in-graph 14:24
travis-ci.org/MoarVM/MoarVM/builds/77817076 github.com/MoarVM/MoarVM/compare/b...284dd9bd3e
14:24 travis-ci left 14:56 Ven joined
dalek arVM: d4385c9 | hoelzro++ | src/ (3 files):
Add ready callback for async procs

This way the HLL can find out when a process is ready
15:02
timotimo brrt: looking forward to seeing the next commit :) 15:03
brrt next commit removes the 'mem' thingie
timotimo ah 15:04
15:05 dalek joined 15:16 travis-ci joined
travis-ci MoarVM build failed. Rob Hoelz 'Add ready callback for async procs 15:16
travis-ci.org/MoarVM/MoarVM/builds/77822270 github.com/MoarVM/MoarVM/compare/4...385c98dd81
15:16 travis-ci left
dalek arVM: 80db16c | hoelzro++ | src/io/procops.c:
Fix mixed declaration failure
15:19
timotimo do you never compile before you push? :) 15:26
JimmyZ I guess he use clang 15:27
timotimo poor he 15:28
he has a thinkpad, why would he use clang?
hoelzro timotimo: no, I forgot to re Configure.pl, so my Makefile was out of date =/
timotimo that makes sense 15:29
hoelzro hmm...should Makefile check that its mtime > mtime(Configure.pl)?
15:39 travis-ci joined
travis-ci MoarVM build passed. Rob Hoelz 'Fix mixed declaration failure' 15:39
travis-ci.org/MoarVM/MoarVM/builds/77823656 github.com/MoarVM/MoarVM/compare/d...db16ce910b
15:39 travis-ci left 16:11 colomon joined 16:22 virtualsue joined 16:40 colomon joined 16:47 lizmat joined 16:48 FROGGS joined
dalek arVM: 37300a9 | paultcochrane++ | src/spesh/candidate.c:
Fix resource leak in spesh/candidate.c

The `before` and `after` variables were not being freed before returning from the function and were thus leaking resources. This change plugs the leak.
16:52
arVM: 1ee1e91 | FROGGS++ | src/spesh/candidate.c:
Merge pull request #243 from paultcochrane/pr/fix-resource-leak-candidate

Fix resource leak in spesh/candidate.c
17:56 ShimmerFairy joined 18:42 colomon joined 18:47 lizmat joined 18:50 lizmat_ joined 19:36 colomon joined 19:57 colomon joined 20:04 colomon joined 20:19 colomon joined 20:23 FROGGS joined 20:25 lizmat joined
dalek arVM: bd65022 | paultcochrane++ | src/core/dll.c:
Add missing curlies to if statement

In my previous fix of the mutex unlock issue, the code flow had changed. To maintain the old behaviour curlies are needed in the if statement.
20:51
arVM: e8f7c01 | lizmat++ | src/core/dll.c:
Merge pull request #244 from paultcochrane/pr/add-missing-curlies

Add missing curlies to if statement
20:56 colomon joined 21:03 vendethiel joined 21:05 virtualsue joined 22:32 ggoebel joined, _itz_ joined 22:40 _itz joined 22:48 _itz joined 23:03 _itz_ joined