|
00:28
MasterDuke joined
|
|||
| MasterDuke | timo: this looks like something you'd be interested in pypy.org/posts/2024/10/jit-peephole-dsl.html | 00:29 | |
| spidermonkey.dev/blog/2024/10/16/7...ckend.html mentions the dominance-calculating algorithm they were previously using is i believe the same one moarvm uses github.com/MoarVM/MoarVM/blob/main...ph.c#L3-L7 | 00:32 | ||
| i don't remember any dominance-related functions showing up very high in perf profiles, but hey, here's a potential research area for improvement | 00:35 | ||
|
09:14
sena_kun joined
09:24
sena_kun left
|
|||
| timo | i've thought about a DSL to use in our different optimizers for a long long time. mainly the rakudo qast optimizer, then later when spesh arrived for the spesh optimizer as well | 10:27 | |
| nothing ever came of those thoughts so far | |||
|
10:37
MasterDuke left
|
|||
| Geth | MoarVM/more_telemetry: 99cec5d1dc | (Timo Paulssen)++ | src/core/exceptions.c telemetry: output something when oops or panic |
12:04 | |
| MoarVM/more_telemetry: 52badcdfea | (Timo Paulssen)++ | src/profiler/telemeh.c telemetry: ensure at least one batch is serialized |
|||
| MoarVM/spesh_max_inline_depth: 8dc4afa0dc | (Timo Paulssen)++ | 2 files spesh: don't deeply recurse into inline attempts fixes #1845 |
13:16 | ||
| lizmat | whee! | 13:17 | |
| timo | probably shouldn't put this in the current release, or at least not without another blin run? | 13:19 | |
| [Coke] | correct. | 13:45 | |
| jdv | well. a blin run takes ~12h and the release is in less than that:) | 14:06 | |
| Geth | MoarVM/jit_unwind_info_for_windows: f255a28e0a | (Timo Paulssen)++ | src/jit/compile.c WIP: RtlAddFunctionTable for jitted code on windows |
17:01 | |
| MoarVM: timo++ created pull request #1852: WIP: RtlAddFunctionTable for jitted code on windows |
17:03 | ||
| jdv | timo: can you please bump nqp and rakudo if and when you commit to main on moarvm | 17:11 | |
| timo | you mean i should always add a bump when i put something into moarvm? | ||
| lizmat | I usually do that when I see there are commits in Mpar | ||
| *Moar | 17:12 | ||
| jdv | looks like i missed the top commit for the last blin run | ||
| probably cause i was having trouble with blin:( | |||
| timo | the vmeventsubscribe one? | ||
| jdv | yes. how confident are you with that? we either revert and put back later or bump it in and release without blin'ing it | 17:13 | |
| timo | if so, don't worry about it, that API currently has no users, it's totally fine to release with the commit one before it | ||
| lizmat | jdv: I concur | ||
| jdv | ok, i'll just release HEAD~1 then | 17:14 | |
| timo: yeah, that's the convention that preceded me | |||
| Geth | MoarVM/2024.100: ee329994f9 | (Justin DeVuyst)++ | 2 files Update changelog and version |
17:28 | |
| MoarVM: jdv++ created pull request #1853: Update changelog and version |
17:33 | ||
| MoarVM/jit_unwind_info_for_windows: eac290d675 | (Timo Paulssen)++ | 2 files mark platform/io functions on windows MVM_PUBLIC |
17:34 | ||
| MoarVM/main: ee329994f9 | (Justin DeVuyst)++ | 2 files Update changelog and version |
17:36 | ||
| MoarVM/main: 4e73df509f | (Justin DeVuyst)++ (committed using GitHub Web editor) | 2 files Merge pull request #1853 from MoarVM/2024.100 Update changelog and version |
|||
| jdv | done here | 17:37 | |
| timo | *big sigh*, looks like compiling with telemeh on windows is going to fail for this release | 17:41 | |
| not sure many people did that anyway? | 17:42 | ||
| is there an errata spot for that kind of thing? | 17:43 | ||
| jdv | don't think so | 17:44 | |
| timo | i seem to recall we have a section in the release notes or announcement where we put stuff like that | 17:59 | |
| Geth | MoarVM/jit_unwind_info_for_windows: 2e58e355d4 | (Timo Paulssen)++ | src/jit/compile.c set correct offset values, plus debugspam |
18:24 | |
| MoarVM/jit_unwind_info_for_windows: 9febceb2b3 | (Timo Paulssen)++ | src/jit/compile.c debugspam |
18:47 | ||
|
18:49
sena_kun joined
20:41
notable6__ left,
bloatable6__ left,
coverable6 left,
shareable6 left,
sourceable6 left,
releasable6 left,
evalable6 left,
unicodable6__ left,
greppable6 left,
quotable6 left,
committable6 left,
nativecallable6 left,
tellable6 left,
bisectable6 left,
linkable6 left,
benchable6 left
20:43
bloatable6 joined,
releasable6 joined
20:44
benchable6 joined,
sourceable6 joined,
evalable6 joined,
shareable6 joined
20:45
bisectable6 joined,
committable6 joined,
notable6 joined,
linkable6 joined,
quotable6 joined
20:46
unicodable6 joined,
coverable6 joined,
nativecallable6 joined,
greppable6 joined
20:47
tellable6 joined
22:13
sena_kun left
|
|||