Hanno Braun

Which optimization level is this happening with? I believe `--release` uses `opt-level=3` by default, which optimizes for speed. `opt-level=s` or `z` should optimize for size. (not saying that these call...

That looks useful, thanks! I'll try it tomorrow.

@kjetilkjeka Thanks for looking into it. I dug around your branch for a bit and found the culprit! svd2rust expects the register size to be in bits, while the value...

> This makes sense, it would be weird to have 28 bytes only using 1 bit of each of them. That's exactly what's happening here, though. The registers are 8...

@jamesmunns The issue is no longer relevant to me, since it was a bug in the SVD file that caused the bug in svd2rust, and I've since fixed the SVD...

@jamesmunns This issue persists. I can still reproduce. I'll summarize what's wrong. Here's the SVD snippet from my original comment: ``` xml 29 0x1 0-28 B%s Byte pin registers port...

I prefer model 1, because believe that HALs should provide a single `Peripherals` struct with all peripheral APIs in it, like the PAC does (see [explanation](https://github.com/nrf-rs/nrf-hal/issues/7) and [practical example](https://github.com/lpc-rs/lpc8xx-hal/blob/85923c568c07298e3b2b17bbbe81dd13f7f7837c/src/lib.rs#L179-L553)). If...

Before I answer the comments, I'd like to clarify my role here: Yes, I'm a member of this org, yes I'm officially a maintainer here, but I haven't really done...

@Dirbaio > What _is_ missing is having `free` properly deinitialize and disable the peripherals, which is something we should do anyway. I always imagined that one use case for `free`-type...

I think we'd need a separate tri-state mode that is distinct from input or output mode. So you'd call `into_tri_state_pin`, which gets you a `Pin`, which would then make the...