|
00:27
SmokeMachine left,
SmokeMachine joined
|
|||
| timo | that kind of thing should be fixed by my moarvm commit from ... uhhkonsole | 03:24 | |
| konsole | |||
| that was me trying unsuccessfully to operate my desktop without a screen. please disregard | 03:36 | ||
| Geth | MoarVM/spesh_unbox_i_cpointer: 2822ef1db1 | (Timo Paulssen)++ | src/6model/reprs/CPointer.c Optimize passing CPointer to nativecall ever so slightly We emit an unbox_i operation in a compiled nativecall body in order to get the pointer value to pass to the native function. Without a spesh method on the CPointer repr, this would interpret as, and jit into, a call to CPointer's get_int. Instead of a call, we just emit a spesh op to do the memory offset and read for us, which the JIT also likes very much. |
03:58 | |
| MoarVM/spesh_unbox_i_cpointer: 6f2dec3b67 | (Timo Paulssen)++ | src/6model/reprs/CPointer.c Optimize passing CPointer to nativecall ever so slightly We emit an unbox_i operation in a compiled nativecall body in order to get the pointer value to pass to the native function. Without a spesh method on the CPointer repr, this would interpret as, and jit into, a call to CPointer's get_int. Instead of a call, we just emit a spesh op to do the memory offset and read for us, which the JIT also likes very much. |
03:59 | ||
| timo | rebase + typo fix | ||
| Geth | MoarVM/main: 6f2dec3b67 | (Timo Paulssen)++ | src/6model/reprs/CPointer.c Optimize passing CPointer to nativecall ever so slightly We emit an unbox_i operation in a compiled nativecall body in order to get the pointer value to pass to the native function. Without a spesh method on the CPointer repr, this would interpret as, and jit into, a call to CPointer's get_int. Instead of a call, we just emit a spesh op to do the memory offset and read for us, which the JIT also likes very much. |
08:07 | |
| MoarVM/main: f6f0bd256a | niner++ (committed using GitHub Web editor) | src/6model/reprs/CPointer.c Merge pull request #1604 from MoarVM/spesh_unbox_i_cpointer Optimize passing CPointer to nativecall ever so slightly |
|||
| lizmat | timo: is that safe for the 2025.01 release ? | 09:26 | |
|
09:44
sena_kun joined
|
|||
| patrickb | Can we get the Windows long path fix in? | 09:49 | |
| lizmat | it couldn't get any worse on Windows, or could it ? | 09:50 | |
| patrickb | did you look at the diff? | ||
| It's a single char change with no possibility for regressions. | 09:51 | ||
| lizmat | hadn't looked yet, yeah, that looks ok | ||
| patrickb | In my naive world view at least. :-P | 09:52 | |
| Geth | MoarVM/main: 7f9662c092 | (Patrick Böker)++ (committed using GitHub Web editor) | src/io/io.c Fix corner case of Windows long path handling |
||
| patrickb | Cool. thanks! | 09:54 | |
| lizmat | bumping.... | ||
| patrickb | That change hopefully finally fixes module installation on Windows. | 09:55 | |
|
13:35
kjp left
15:02
guifa joined
15:11
sena_kun left,
sena_kun joined
15:15
sena_kun left
15:16
sena_kun joined
15:47
guifa left,
guifa joined
|
|||
| jdv | [Coke]: there is no need, per se, for the delete/ignore in the moarvm changelog | 16:00 | |
| timo | it's probably safe | 16:41 | |
| still a bit bold to merge it during release time | 16:43 | ||
|
16:55
guifa left
17:32
guifa joined
17:57
guifa left
18:04
guifa joined
|
|||
| [Coke] | jdv: yup, just in case. Figured the rakudo log was more important/required more fiddling in general. | 18:20 | |
| I am not really happy about the DAY OF set of commits to main on MoarVM | |||
| Who is vouching for that? | |||
| I will plan on rolling that back before the release and someone can re-apply it after unless we get concensus | 18:21 | ||
| timo | yeah feel free to do that if it's not too much hassle | 18:24 | |
|
18:25
guifa left
|
|||
| jdv | the what now? | 18:37 | |
| ah, i see now | 18:38 | ||
| Geth | MoarVM/revert-1604-spesh_unbox_i_cpointer: c1763d868f | (Will Coleda)++ (committed using GitHub Web editor) | src/6model/reprs/CPointer.c Revert "Optimize passing CPointer to nativecall ever so slightly" |
18:56 | |
| MoarVM: coke++ created pull request #1899: Revert "Optimize passing CPointer to nativecall ever so slightly" |
|||
| MoarVM/main: c1763d868f | (Will Coleda)++ (committed using GitHub Web editor) | src/6model/reprs/CPointer.c Revert "Optimize passing CPointer to nativecall ever so slightly" |
18:57 | ||
| MoarVM/main: d8353be724 | (Will Coleda)++ (committed using GitHub Web editor) | src/6model/reprs/CPointer.c Merge pull request #1899 from MoarVM/revert-1604-spesh_unbox_i_cpointer Revert "Optimize passing CPointer to nativecall ever so slightly" |
|||
| [Coke] | This was a PR from 2021 that just happened to get a commit 15 hours ago. | 18:58 | |
| .seen niner | 18:59 | ||
| tellable6 | [Coke], I haven't seen niner around, did you mean nine? | ||
| [Coke] | .seen nine | ||
| tellable6 | [Coke], I saw nine 2024-11-27T13:36:16Z in #raku-dev: <nine> I would just get used to that :D | ||
| [Coke] | we also have the stray commit 7f9662c09291fc3b1b46d6c609694af755944be5 | 19:00 | |
| linkable6 | (2025-01-25) github.com/MoarVM/MoarVM/commit/7f9662c092 Fix corner case of Windows long path handling | ||
| [Coke] | This looks like a safer commit, it's changing a > to a >= in a win specific set of code for paths that are already dealing with max-path issues. | 19:01 | |
| If you had asked me, I would have said "don't commit anything to main". :| Anyone else have thoughts on this one? | 19:02 | ||
| patrickb | Sorry to have caused you a hassle. Didn't intend to... | 19:09 | |
| [Coke] | Yours seems much safer than the other one; but any changes this close to release are a risk | 19:10 | |
| Did you get a CI run with it? | |||
| patrickb | Incidentally that change fixes the install issues you reported a few days back. But yes, merges right before a release - meh. | 19:12 | |
| CI was green. | 19:13 | ||
| [Coke] | ok. We'll roll the dice on that one. :) | 19:25 | |
|
21:04
kjp joined
21:56
guifa joined
22:05
kjp left,
kjp joined
|
|||
| Geth | MoarVM/2025.010: e37b568420 | (Will Coleda)++ | 2 files Update changelog and version |
22:11 | |
|
22:44
sena_kun left
|
|||
| Geth | MoarVM: coke++ created pull request #1900: Update changelog and version |
23:06 | |
| jdv | anyone in here have moarvm.org repo admin? | 23:13 | |
| need [Coke] added | |||
| [Coke] | short term, we have github.com/MoarVM/moarvm.org/pull/16 | 23:15 | |
| timo | yeah i'm getting all these "implausibly old timestamp" outputs | 23:23 | |
| jdv | ditto | ||
| timo | can i see your log output or something? | 23:24 | |
| oh i guess it was the github runner that created these? | 23:27 | ||
| [Coke] | the other tar files? I think so | 23:28 | |
| timo | hm those are tar not zip and the artifacts that github actions is creating are all zip files | ||
| oh | 23:29 | ||
| i guess there's a .tar.gz in the .zip file | |||
| Geth | MoarVM/main: e37b568420 | (Will Coleda)++ | 2 files Update changelog and version |
23:31 | |
| MoarVM/main: 1d088dfb1a | (Will Coleda)++ (committed using GitHub Web editor) | 2 files Merge pull request #1900 from MoarVM/2025.010 Update changelog and version |
|||
| [Coke] | merging that now in case it's required for the Ake stuff - it hadn't finished CI, but previous commits with actual code were ok | 23:32 | |
| even with the hiccups, This is so much less stressful than I remember Christmas being. :) | 23:36 | ||
| timo | seeing loads of projects that experienced the implausibly old timestamps at some point | 23:45 | |
| [Coke] | ok, good | ||
| "see, it's not just our very out of practice rel mgr" | |||