01:29
FROGGS joined
01:48
ilbot3 joined
05:07
flaviusb joined
|
|||
FROGGS | o/ | 06:59 | |
07:12
domidumont joined
07:16
domidumont joined
|
|||
nwc10 | \o | 07:37 | |
07:53
geekosaur joined
|
|||
FROGGS | interesting... nqp::fromnum_I seems to be buggy on x32 | 08:05 | |
moritz | might very well be; i've only been able to test in on 62bit | 08:21 | |
FROGGS | the interesting bit is, so to say, that it works on i386, but not on x32 | 08:28 | |
10:09
Ven_ joined
|
|||
dalek | arVM: b654e7e | FROGGS++ | src/math/bigintops.c: use set_uint64 in from_num, otherwise we overflow on x32, see #425 |
10:31 | |
nebuchadnezzar | FROGGS: thanks a lot | 11:50 | |
FROGGS | nebuchadnezzar: I hope it helps :o) | 11:59 | |
nebuchadnezzar: what's the status of rakudo in debian ooc? is it still in unstable or in testing? or even not there yet? | 12:00 | ||
12:02
domidumont joined
12:18
zakharyas joined
|
|||
nebuchadnezzar | FROGGS: 2016.06 was automatically migrated in testing tracker.debian.org/pkg/rakudo | 12:38 | |
I hope we could do better now since you made many fixes | |||
domidumont: I think we could integrate the feature/switch-to-libffi branch and backport the upstream fixes to see how it does for all architectures. We may found other issues fixable before 2016.10 | 12:43 | ||
FROGGS | nebuchadnezzar: being able to fix more stuff would be awesome | 12:52 | |
13:25
Ven_ joined
|
|||
timotimo | The first line of the "Description:" should not start with the package name. For example, the package foo should not have a description like this: "foo is a program that...". | 13:47 | |
this is terrible! we have to fix it! | |||
:) | |||
dalek | arVM: 8eb9366 | (Jimmy Zhuo)++ | src/6model/reprs/C (2 files): removed unused code |
13:57 | |
arVM: 958a01f | (Jimmy Zhuo)++ | src/io/dirops.c: rewrite mkdir_p api, no function changes |
|||
14:06
Ven__ joined
14:24
Ven_ joined
|
|||
domidumont | nebuchadnezzar: agreed. We can't do worse than the actual situation where testing is stuck at 2016.06... | 14:31 | |
nebuchadnezzar: BTW, rakudo is now building and testing fine on armhf | |||
nebuchadnezzar: I've pushed all armhf patches on Debian repo. I'll let you add the x32 ones. | 14:49 | ||
nebuchadnezzar | ok | 15:14 | |
FROGGS | JimmyZ: why do you include math.h? are we using anything from it? | 15:16 | |
domidumont++, nebuchadnezzar++ | 15:17 | ||
JimmyZ: nvm, I didnt read the patch correctly | 15:18 | ||
JimmyZ++ | |||
15:39
Ven_ joined
|
|||
dalek | arVM: f0a70ee | (Jimmy Zhuo)++ | src/6model/reprs/C (2 files): remove unless header file |
16:15 | |
nebuchadnezzar | domidumont: I just push the x32 patches after testing them in an x32 schroot, nqp and rakudo build and their tests pass | 16:34 | |
FROGGS | nebuchadnezzar: is that 2016.09-2 then? | 16:47 | |
nebuchadnezzar: how long does it take until I see something happen here? buildd.debian.org/status/package.php?p=rakudo | 16:48 | ||
17:01
flaviusb joined
|
|||
nebuchadnezzar | FROGGS: we need to upload the new Moarvm, I don't have upload right so I ask domidumont | 17:14 | |
FROGGS | ahh :o) | 17:16 | |
18:16
Ven_ joined
18:21
Ven_ joined
18:27
dalek joined
18:36
Ven_ joined
18:58
Ven_ joined
|
|||
dalek | arVM: 688796b | FROGGS++ | src/6model/reprs/C (4 files): improve nativecall attribute error messages |
19:17 | |
20:57
Ven_ joined
21:03
geekosaur joined
21:11
Ven_ joined
|