|
Welcome to the main channel on the development of MoarVM, a virtual machine for NQP and Rakudo (moarvm.org). This channel is being logged for historical purposes. Set by lizmat on 24 May 2021. |
|||
|
00:49
vrurg_ joined,
vrurg left,
vrurg_ is now known as vrurg
02:03
epony joined
05:01
epony left
05:02
epony joined
11:13
sena_kun joined
16:18
epony left
16:20
epony joined
|
|||
| Geth | MoarVM/2024.010: a7a2916b50 | (Justin DeVuyst)++ | 2 files Update changelog and version |
19:21 | |
| MoarVM: jdv++ created pull request #1787: Update changelog and version |
19:27 | ||
| MoarVM/main: a7a2916b50 | (Justin DeVuyst)++ | 2 files Update changelog and version |
|||
| MoarVM/main: 7c58f389d6 | (Justin DeVuyst)++ (committed using GitHub Web editor) | 2 files Merge pull request #1787 from MoarVM/2024.010 Update changelog and version |
|||
|
22:24
kjp left
22:25
japhb_ is now known as japhb
22:26
kjp joined
|
|||
| patrickb | Is there a way to GC unroot an address I've added earlier? | 23:00 | |
| I want to put an object into a frame further down the call stack. IIUC I need to root the address the moment I set it. But then how to I unroot it once the frame is popped some time later? | 23:03 | ||
|
23:13
sena_kun left
|
|||
| patrickb | Technically I could search through the entire call stack on every GC for those objects, but I guess that's slower than rooting when set and unrooting on frame unwind. | 23:19 | |
| I have just found MVM_callstack_mark_current_thread. I guess that answers that question. | 23:29 | ||
| I think it's not technically wrong to GC mark random addresses. But is it bad performance wise? I just noticed that frame->extra->exit_handler_result is unconditionally marked. | 23:39 | ||