07:49 <re_irc> <@luojia65:matrix.org> Is fu740-hal project in active development? I need to use it in RustSBI firmware implementation on Unmatched motherboard

07:56 <re_irc> <@luojia65:matrix.org> I have a simple question: how to use 'svd patch' command? :)

14:34 <re_irc> <@nikomatsakis:matrix.org> Hey folks! I remember seeing some kind of cool embedded debugger project, but I can't find it now. Anybody have any idea what i'm talking about? Was it all a dream?? 💭

14:36 <re_irc> <@dirbaio:matrix.org> https://probe.rs/ probably

15:11 <re_irc> <@jamesmunns:beeper.com> Another one that comes to mind (not embedded related) is headcrab: https://github.com/headcrab-rs/headcrab

15:11 <re_irc> <@jamesmunns:beeper.com> There's also Ferrous' Knurling toolkit, which is generally based on the probe-rs library: https://knurling.ferrous-systems.com/tools/

15:41 <re_irc> <@jamesmunns:beeper.com> ubik:matrix.org: Sometimes you will also need to enable certain flags in the RCC to set alternate clock sources. For example in the STM32F3xx-HAL, we had to manually change the clock source for USART2 in order to hit certain baudrates.

15:42 <re_irc> <@jamesmunns:beeper.com> I find the BEST way to do it is to do your configuration through stm32cubemx, then work backwards from there to figure out how to hit those multipliers with the HAL.

16:09 <re_irc> <@firefrommoonlight:matrix.org> In this case, there's a flag that chooses the I2S source as either PLL12SCLK, or external clock

16:11 <re_irc> <@firefrommoonlight:matrix.org> *Maybe not. There's a selector in Cube, but I can't find it in the RM

17:04 <re_irc> <@ubik:matrix.org> Yeah, I've installed cube. It sure does help.

17:05 <re_irc> <@ubik:matrix.org> Thanks. Now it's fine, but I had to double the PLL output.

18:22 <re_irc> <@chmanie:matrix.org> I just upgraded some of the dependencies in my project and now I'm getting an error message that I can't really make anything of: `error: failed to load bc of "cortex_m_rt-93b4ca2522753c0e.cortex_m_rt.ab7089e1-cgu.0.rcgu.o":[201~': symbol multiply defined! error: failed to load bc of "cortex_m_rt-93b4ca2522753c0e.cortex_m_rt.ab7089e1-cgu.0.rcgu.o"` I found this issue about similar conflicts but I can't really..

18:22 <re_irc> ... figure out a course of action here: https://github.com/stm32-rs/stm32f4xx-hal/issues/369.

18:23 <re_irc> <@chmanie:matrix.org> I'm using `cortex-m = "0.7.2"`, `cortex-m-rt = "0.7"`, `stm32f4xx-hal = "0.10.1"`, with `rt` feature

18:25 <re_irc> <@dirbaio:matrix.org> you can't have multiple cortex-m-rt versions

18:26 <re_irc> <@dirbaio:matrix.org> stm32f4xx-hal 0.10.1 uses stm32f4 0.13 which uses cortex-m-rt 0.6

18:26 <re_irc> <@dirbaio:matrix.org> either stick to cortex-m-rt 0.6, or patch stuff so everything is 0.7

18:28 <re_irc> <@chmanie:matrix.org> Aaah, yes, I see

18:28 <re_irc> <@chmanie:matrix.org> Thanks!

20:33 <re_irc> <@jacobrosenthal:matrix.org> ryankurte: Does the rust-iot org / radiohal groups have a discussion location? A room on matrix? :)

21:09 <re_irc> <@ubik:matrix.org> I fell into the same problem as chmanie recently.

21:36 <re_irc> <@hanno:braun-odw.eu> jacobrosenthal:matrix.org: https://matrix.to/#/#rust-iot:matrix.org

21:54 <re_irc> <@ubik:matrix.org> Could someone point me to an example using DMA with stm32f4xx-hal on a buffer which is not pre-computed (i.e. you can't get away with just pre-filling the DMA buffer and have to mutate it somehow)? Bonus points if it's a double buffer.

21:55 <re_irc> <@ubik:matrix.org> This example is nice, but I'd like to be able to edit the buffer and I'm not sure how to implement it. https://github.com/stm32-rs/stm32f4xx-hal/blob/7305ef4cd9b45a74d5102c3c9f4cda49b96498f0/examples/i2s-audio-out-dma.rs

21:55 <re_irc> <@ubik:matrix.org> I imagine I'd have to implement `ReadBuffer`?

21:56 <re_irc> <@jamesmunns:beeper.com> ub|k `heapless::Pool` also implements the embedded-dma traits

21:56 <re_irc> <@jamesmunns:beeper.com> it's what I typically use for dma stuff

21:57 <re_irc> <@jamesmunns:beeper.com> I have a branch of bbqueue that also does for its grants, but I never published that (oops)

21:58 <re_irc> <@ubik:matrix.org> Oh cool

21:58 <re_irc> <@ubik:matrix.org> Thanks, that's good news

21:58 <re_irc> <@jamesmunns:beeper.com> (if anyone needs the bbqueue branch, lemme know, I can try to get that released this weekend)

21:58 <re_irc> <@jamesmunns:beeper.com> if not it'll remain on the stack until I need it :D

22:19 <re_irc> <@ubik:matrix.org> James Munns: are you sure?

22:19 <re_irc> <@ubik:matrix.org> = note: the following trait bounds were not satisfied:

22:19 <re_irc> <@ubik:matrix.org> `DMA_BUFFER: embedded_dma::StaticReadBuffer`

22:19 <re_irc> <@ubik:matrix.org> `<DMA_BUFFER as embedded_dma::StaticReadBuffer>::Word = u16`

22:20 <re_irc> <@ubik:matrix.org> (DMA_BUFFER is my pool's "name")

22:22 <re_irc> <@jamesmunns:beeper.com> Yes, should be?

22:22 <re_irc> <@jamesmunns:beeper.com> as long as your pool is for something like `[u16; 32]`

22:22 <re_irc> <@jamesmunns:beeper.com> (or whatever size)

22:22 <re_irc> <@jamesmunns:beeper.com> what does your `pool!` invocation look like?

22:22 <re_irc> <@henrik_alser:matrix.org> ubik:matrix.org: https://github.com/kalkyl/f411-rtic/blob/main/src/bin/serial-dma-tx.rs

22:23 <re_irc> <@ubik:matrix.org> pool!(

22:23 <re_irc> <@ubik:matrix.org> DMA_BUFFER: [u16; BUFFER_SIZE * 2]

22:23 <re_irc> <@ubik:matrix.org> );

22:24 <re_irc> <@jamesmunns:beeper.com> hmmmmmmmmmm

22:24 <re_irc> <@jamesmunns:beeper.com> I've used it with `u8`s and it works, not sure about u16s...

23:10 <cr1901> How does cortex-m-rt prevent multiple versions again? Is it the non-mangled symbols?

23:10 <re_irc> <@adamgreig:matrix.org> all the modern versions of c-m-rt use the `links` field on Cargo.toml

23:10 <re_irc> <@adamgreig:matrix.org> so Cargo itself prevents it

23:11 <cr1901> _or_ are the non-mangled symbols a side effect of another way to prevent multiple versions?

23:11 <cr1901> ahhh

23:11 <re_irc> <@adamgreig:matrix.org> but anyone with an outdated dependency on an older 0.6 release before we added `links` will instead run into a less-clear linker error about multiple symbols, yea

23:11 <re_irc> <@adamgreig:matrix.org> or missing symbols, depending on the nature of what went wrong

23:11 <cr1901> missing symbols?

23:11 <re_irc> <@adamgreig:matrix.org> cortex-m-rt does indeed also define a bunch of no_mangle symbols so linking would fail on those

23:12 <re_irc> <@adamgreig:matrix.org> the missing symbols is confusing but if you have a PAC using one version of c-m-rt and your code uses a newer version you can end up with the new version not finding the interrupt names that should have been exported by the PAC?

23:12 <re_irc> <@adamgreig:matrix.org> I haven't fully riddled it through

23:13 <cr1901> hmmm, doesn't make sense to me (since the PAC will be the one depending on c-m-rt, not the other way around).

23:13 <cr1901> But I'll take your word for t

23:13 <cr1901> it* :D

23:14 <re_irc> <@adamgreig:matrix.org> yea it doesn't make sense to me either and yet somehow a bunch of people were getting missing symbols not duplicate ones

23:15 <re_irc> <@adamgreig:matrix.org> hmm... well, a lot of people also got `__INTERRUPTS` being duplicate

23:15 <re_irc> <@adamgreig:matrix.org> which does make sense

23:19 <re_irc> <@ubik:matrix.org> henrik_alser: thanks, but I'm not using rtic