|
github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm Set by AlexDaniel on 12 June 2018. |
|||
|
01:11
tellable6 left,
evalable6 left
01:14
tellable6 joined,
evalable6 joined
01:22
sena_kun left
01:37
sena_kun joined
02:47
vrurg_ joined
02:48
vrurg left
03:23
sena_kun left
03:38
sena_kun joined
04:53
ZzZombo_ joined
04:56
ZzZombo left
04:57
ZzZombo__ joined,
ZzZombo__ is now known as ZzZombo
04:59
ZzZombo_ left
05:22
sena_kun left
05:38
sena_kun joined
05:58
Kaiepi left
05:59
Kaiepi joined
06:04
Kaiepi left
06:05
Kaiepi joined
06:18
Kaiepi left,
Kaiepi joined
06:20
Kaiepi left
06:21
Kaiepi joined
07:23
sena_kun left
07:39
sena_kun joined
09:18
sena_kun left
09:33
sena_kun joined
11:23
sena_kun left
11:38
sena_kun joined
|
|||
| Geth_ | MoarVM: 41708648cf | (Stefan Seifert)++ | src/core/interp.c Revert "Workaround for segfault somehow caused by NativeCall callbacks" This reverts commit 7ab23a1b89d4a191fc0f9547ba2a0b6866ce3d95 as a proper fix has been found. |
11:59 | |
| MoarVM: ae66545c7b | (Stefan Seifert)++ | 2 files Fix segfaults after callbacks originating in JIT compiled calls to native code Before running a non-toplevel frame, we need to make sure, the caller's frame has a valid return_address. While MVM_nativecall_invoke_jit and MVM_nativecall_invoke did so, the JIT implementation of calls to JIT compiled native functions did not. This could lead to segfaults when native code called from JIT compiled code ran callbacks which then tried to return to the NULL address. Fix by initializing tc->cur_frame->return_address in the callback_handler, so it doesn't matter how we ended up there. |
|||
| nine | sena_kun: this is the real fix for the release ^^^ | 12:02 | |
|
12:45
sena_kun left
13:11
sena_kun joined
|
|||
| tadzik | sena_kun: pretty much all of my stuff is abandoned :) | 13:14 | |
| tellable6 | 2019-12-21T14:12:40Z #raku-dev <patrickb> tadzik Thanks for the reply and heads up! | ||
| 2019-12-21T14:12:53Z #raku-dev <patrickb> tadzik I'll go forward in that direction then. | |||
| 2019-12-25T22:54:06Z #moarvm <sena_kun> tadzik is github.com/tadzik/Module-Toolkit abandoned? | |||
|
13:22
sena_kun left
13:37
sena_kun joined
14:19
Kaeipi joined
14:20
Kaiepi left
14:23
Kaeipi left
14:24
Kaiepi joined
14:29
lucasb joined
14:44
vrurg joined
14:46
vrurg_ left
14:50
AlexDaniel left,
vrurg_ joined
14:52
vrurg left
15:22
sena_kun left
15:38
sena_kun joined
16:00
domidumont joined
16:54
Kaiepi left
16:58
Kaiepi joined
17:24
sena_kun left
17:39
sena_kun joined
18:28
lucasb left
19:23
sena_kun left
19:38
sena_kun joined
19:43
domidumont left
|
|||
| sena_kun | o/ | 19:49 | |
| tadzik | \o | 19:50 | |
| sena_kun: in case you missed it: pretty much all of my stuff is abandoned :) | |||
| sena_kun | so I checked, well, everything from our ecosystem. I think, a moarvm release can be shipped from github.com/MoarVM/MoarVM/commit/7a...6866ce3d95 (as I checked this one) | ||
| tadzik: yeah, needed to make sure. :) | 19:51 | ||
| I am not sure about including github.com/MoarVM/MoarVM/commit/ae...076d38c7a2 and don't think I'll test another 400 modules from our ecosystem by hands, at least not this release. :S | 19:52 | ||
| but if it looks sane to core devs and passes tests, I think it is pretty much safe anyway | |||
| (does it count as a green light?) | 19:53 | ||
| lizmat | sena_kun: if nine is ok with it, I'm ok with it :-) | 19:54 | |
| sena_kun hopes we'll be able to do the release this month | 19:55 | ||
| lizmat too :-) | 20:09 | ||
|
21:22
sena_kun left
21:38
sena_kun joined
23:04
Kaeipi joined,
Kaiepi left
23:23
sena_kun left
23:36
sena_kun joined
23:44
sena_kun left
|
|||