00:12 <cr1901> Amanieu: https://github.com/rust-lang/rust/pull/93219#issuecomment-1019379023 Oh dear... was this supposed to happen?

00:12 <Amanieu> Yea, you need to update the GCC and cranelift backends too.

00:12 <Amanieu> I think ./x.py check will make sure they compile.

00:13 <cr1901> Oops

00:18 <agg> Amanieu: could you try leaving and re-joining this room to see if it helps the irc bridge? sorry for the trouble...

00:19 <Amanieu> test

00:19 <agg> it seems to have made a puppet for you on matrix but the puppet isn't in the matrix room...

00:19 <agg> no joy :/

00:20 <agg> hifi: any idea what could be going on with heisenbridge? @irc_libera_amanieu:psion.agg.io seems to exist on matrix and the name is pilled when mentioned here, but they're not in #rust-embedded:matrix.org user list and none of the messages make it through

00:20 <agg> but it works OK for everyone else it seems

00:21 <agg> restarting heisenbridge hasn't helped, and I can't kick that user since they're already not in the room

00:21 <cr1901> Why won't Libera add this channel to their official bridge?

00:21 <agg> it's a mess

00:22 <agg> you can join it through the official bridge at #rust-embedded:libera.chat from matrix actually

00:22 <Amanieu> Should I try reconnecting to libera?

00:22 <agg> but, all the matrix users are in #rust-embedded:matrix.org, which is a different room; the new official libera bridge can't bring an existing matrix room into irc afaik

00:22 <agg> Amanieu: you could try but I don't think the bridge would see that any different to you leaving this channel so I don't think it would make a difference...

00:24 <agg> cr1901: and then the second problem: we still don't own this channel on the libera side because it needs someone from rust core to register #rust namespace with them and delegate #rust-embedded to us, as we had on freenode

00:24 <cr1901> Ahhh, and presumably Freenode was before everyone on upstream Rust became allergic to IRC?

00:25 <agg> well we asked them to do it then too and they did

00:25 <agg> I'll try pinging core again

00:40 <agg> ping sent

00:44 <cr1901> tyvm

03:57 <re_irc> <@hifi:vi.fi> agg I think you've not updated the bridge for like 6 months, hard to say anymore as a lot has changed

03:57 <re_irc> <@hifi:vi.fi> also there were issues with Dendrite originally

03:58 <re_irc> <@adamgreig:matrix.org> Heh. I updated to the latest version with python 3.7 support a few days ago to see if it would help but no difference

03:58 <re_irc> <@adamgreig:matrix.org> Anything I can do to purge the current matrix user and force it to remake it?

03:58 <re_irc> <@hifi:vi.fi> the major change that could make a difference was the next version

03:59 <re_irc> <@adamgreig:matrix.org> 1.7.1 iirc

03:59 <re_irc> <@hifi:vi.fi> there's a good chance it works with 3.7 anyway

03:59 <re_irc> <@adamgreig:matrix.org> Pip won't let me install it from pypi but I will try and force its hand and if not I'll install a newer python

04:00 <re_irc> <@adamgreig:matrix.org> Thanks for the tip

04:00 <re_irc> <@hifi:vi.fi> but I think the real issue is with Dendrite as if restarting doesn't help its lying to the bridge 😅

04:02 <re_irc> <@adamgreig:matrix.org> I could believe it... I'll give that a kick too.

04:04 <re_irc> <@hifi:vi.fi> since the latest version is using mautrix it's more likely to work with Dendrite as I don't remember anyone complaining any mautrix bridge but working with it

06:03 <re_irc> <@yuhanliin:matrix.org> cr1901: Does your PR support doing arithmetic/bitwise instructions directly on memory locations (like in `msp430-atomic`)?

08:27 <cr1901> yuhanliin: No, that's beyond the scope of rust inline asm

08:28 <cr1901> Meaning until something that supports directly operating on memory locations is supported, there's gonna be a small size penalty for atomics

08:28 <cr1901> Still far superior to refcell/cell dance tho

08:32 <cr1901> Yuhanliin: If you're asking "can rust inline asm, especially msp430, access memory locations?", then: yes, yes it can. And I already adapted msp430-atomic to the new syntax, and tested a deployed firmware

08:32 <cr1901> Everything's working fine aside from the 50 byte or so size penalty

08:32 <cr1901> https://github.com/cr1901/msp430-atomic/tree/new-asm

09:30 <re_irc> <@allexoll:matrix.org> Is it possible to feature gate a defmt test with a feature of the hal dependency? I tried something with

09:30 <re_irc> <@allexoll:matrix.org> ```rust

09:30 <re_irc> <@allexoll:matrix.org> #[test]

09:30 <re_irc> ... long message truncated: https://psion.agg.io/_matrix/media/r0/download/psion.agg.io/37db06ccd642560bf3b584faf5016f9a7843c14513fe85c223cf718f782af3dc (12 lines)

09:30 <re_irc> <@allexoll:matrix.org> #[cfg(feature = "lpc546xx-hal/flexcomm-10")]

13:27 <re_irc> <@jamesmunns:beeper.com> You probably want to add a "local" feature

13:29 <re_irc> <@jamesmunns:beeper.com> like

13:29 <re_irc> <@jamesmunns:beeper.com> ```toml

13:29 <re_irc> ... long message truncated: https://psion.agg.io/_matrix/media/r0/download/psion.agg.io/e814e074405e1a6a8eea7f378f1de31423aa842e67b764eaf4ccf41732d8f739 (9 lines)

13:29 <re_irc> <@jamesmunns:beeper.com> test001 = ["lpc546xx-hal/flexcomm-10"]

13:29 <re_irc> <@jamesmunns:beeper.com> [features]

14:57 <re_irc> <@allexoll:matrix.org> its kind of sketchy but that works, thanks!

15:03 <agg> (testing bridge, ignore)

21:51 <re_irc> <Lachlan Sneff> Does anyone know of a crate that has a "Vec<T, N>", where the backing memory is a "&'static mut", instead of an inline array?

21:52 <Lumpio-> Does it need an N type parameter if you have a slice

21:53 <re_irc> <Lachlan Sneff> I would prefer the memory to actually be "&mut static [T; N]"

21:53 <re_irc> <Lachlan Sneff> So, ideally yes

21:57 <re_irc> <dirbaio> just use "&'static mut heapless::Vec<T, N>"

21:58 <re_irc> <dirbaio> it doesn't directly exist because it'd be super awkward to use: you wouldn't be able to (safely) create it, nor clone it..