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. |
|||
Geth | MoarVM/main: 7c5be29d0a | (Geoffrey Broadwell)++ | tools/ucd2c.pl Make progress headers easier to see |
00:23 | |
MoarVM/main: eb066c7923 | (Geoffrey Broadwell)++ | tools/ucd2c.pl Minor tweaks to sequence code |
00:24 | ||
MoarVM/main: 2bee37a02d | (Geoffrey Broadwell)++ | tools/ucd2c.pl Refactor/clean up name aliases code As a side benefit, get rid of two globals that only passed info between two routines. |
00:26 | ||
MoarVM/main: e68d9893f8 | (Geoffrey Broadwell)++ | tools/ucd2c.pl Refactor/cleanup derived_property() and its callees |
02:34 | ||
MoarVM/main: e205f909ae | (Geoffrey Broadwell)++ | tools/ucd2c.pl Refactor/clean up UnicodeData and its callees |
05:06 | ||
MoarVM/main: 39d75a9eaa | (Geoffrey Broadwell)++ | tools/ucd2c.pl Move BidiMirroring next to other enumerated_property calls |
05:08 | ||
10:04
patrickb_ joined
10:08
Nicholas_ joined
10:14
camelia left,
Nicholas left,
patrickb left,
patrickb_ is now known as patrickb
10:25
camelia joined
|
|||
Geth | MoarVM/main: 4419c71114 | (Timo Paulssen)++ | 2 files Offer full file path including module name in debug server protocol. |
15:05 | |
MoarVM/main: bcfedbd0c1 | (Timo Paulssen)++ | src/core/validation.c In bytecode validation, output name of register/operand type instead of just the number |
|||
15:28
camelia left
|
|||
leont | I was looking at making improving the atomics, and frankly getting rid of libatomic_ops would make everything easier. | 15:41 | |
[Coke] | :) | ||
leont | C11 is lightyears ahead, and since 2022 MSVC also supports them so I think that'd be fine | ||
*C11 atomics | 15:42 | ||
15:59
camelia joined
16:42
camelia left
16:55
camelia joined
20:10
MasterDuke joined
|
|||
japhb | What makes the C11 atomics so much better than libatomic_ops? | 20:26 | |
Geth | MoarVM: Leont++ created pull request #1943: Remove libatomic_ops |
20:59 | |
MoarVM: Leont++ created pull request #1944: Stop using Windows 2019 runner |
21:19 | ||
leont | It would be helpful to merge the second one before the 10th, because of the brown-outs | 21:20 | |
japhb | leont: Will do as soon as we get CI on it. | 21:22 | |
My question above still stands. :-) | |||
(Don't get me wrong, I'm all for using standards when available. You just made it sound like it would be a massive improvement, and not just in reducing stuff we have to support.) | 21:23 | ||
leont | Several things really. | 21:25 | |
japhb | (I guess I should say "standards *that don't suck*", because of course some of them do. We're going to assume the C11 committee did better than that. ;-) ) | ||
leont | Firstly, it supports explicit memory ordering, which means we can choose the most performant option that is appropriate for the specific case. | 21:27 | |
One of the other nice things is that C11 atomics are type generic. So we don't need to cast from and to AO_t* all the time. This makes it much easier to use it with pointers in particular. | |||
Secondly, it supports more options. It seems libatomics lacks a plain exchange (or maybe I fail to find it), which is sometimes needed. | 21:28 | ||
japhb | Fair enough. Thanks for the summary! | 21:30 | |
dev.azure.com/MoarVM/MoarVM/_build...amp;l=1450 <-- Mac failure with this PR in place? | 21:43 | ||
leont | That is also failing on main | 21:44 | |
japhb | dev.azure.com/MoarVM/MoarVM/_build...6357f73cc3 <-- Ubuntu 20.04 retired | ||
dev.azure.com/MoarVM/MoarVM/_build...67f910b2b1 <-- same | 21:45 | ||
OK, so it looks like some CI cleanup is needed. But I'll merge #1944 now. | |||
leont | Yeah I was looking at the Ubuntu issue, but I have no idea what to do about Mac | 21:46 | |
japhb | Wait ... how do the clang/gcc _no_c11_atomics builds pass now that there is now libAO? | ||
21:48
MasterDuke left
|
|||
japhb | Oh duh, because 1943 hasn't been merged. I'm apparently an idiot. | 21:50 | |
Geth | MoarVM/main: 58eda115fd | (Leon Timmermans)++ | 2 files Stop using Windows 2019 runner They are no longer supported starting 2025-06-30. See also github.com/actions/runner-images/issues/12045 |
21:51 | |
MoarVM/main: 0c63932cf2 | (Geoffrey Broadwell)++ (committed using GitHub Web editor) | 2 files Merge pull request #1944 from Leont/windows-2022 Stop using Windows 2019 runner Notes: * There were two CI runs canceled because the Ubuntu 20.04 they specify is no longer supported in Azure. This will need a PR of its own for cleanup. * There is a Mac CI failing that is also failing on main (so no difference from this PR). |
|||
leont | I don't understand the Mac failure at all | 22:00 | |
Geth | MoarVM: Leont++ created pull request #1945: Upgrade to Ubuntu 24.04 |
22:08 | |
MoarVM/main: 78acd73d00 | timo++ (committed using GitHub Web editor) | 3 files Less memory waste binslurp and normalizecodes (#1938) * In read_fhb aka nqp::readfh don't keep too-large buffers around. We were allocating as much space as the caller was asking for, but in rakudo for slurp we are passing 1 megabytes as the size by default. ... (7 more lines) |
22:22 |