Dan Whitman

@rnd-ash Gotcha, it seems to me like not taking full ownership of the `Pclk` structs is a mistake. Can anyone (@jbeaurivage @ianrrees @bradleyharden ) confirm this?

@ianrrees Gotcha. Do you have an opinion when it comes to the approach taken by PRs? (i.e. multiple things in one PR or one PR peripheral)

@rnd-ash The GCLK cannot be disabled as long as the `Pclk` exists. The reason is that enabling the `Pclk` will increment the compile-time consumer counter (the `N` generic type) for...

A couple of larger issues with this effort need to be discussed: 1. I tried to migrate the RTC and its RTIC monotonic, and ran into the same issue that...

@ianrrees The thumbv7 v2 API seems to handle this by using, for example, `clock::v2::types::Tc0Tc1` to represent their `Pclk`. However, this presents difficulties in using both TC0 and TC1 at the...

@rnd-ash It is done. Good idea with the separate branch so that everything can be fixed and done here before merging into `master` down the road.

@rnd-ash That sounds good. I have not had a lot of time to work on this stuff lately, unfortunately, but I am still working on an easy-to-use USB serial crate....

@rnd-ash I see that you've now taken full ownership of the `Pclk`, which is great. Can you change the name of the `release` method to `free`? I think this name...

@rnd-ash I don't know much about the QSPI, but evidently it has two AHB clock buses, `CLK_QSPI_AHB` and `CLK_QSPI2X_AHB`, represented in the v2 API by `clock::v2::types::Qspi` and `clock::v2::types::Qspi2x`, respectively. Your...

@rnd-ash Gotcha, it's fine to not offer DDR support at this point. I just wanted the rationale to be documented here.