PR #665 made stm32-metapac rebuild when the chip definition was changed.
Though it used the lowercase version of the chip name as a filename
which probably worked fine on windows with its case-independent
filesystem, but was causing constant rebuilds on linux
667: Remove duplicate stm32-metapac/src/common.rs with chiptool r=Dirbaio a=nviennot
There's a duplicate file common.rs with the chiptool crate. This PR makes the source of truth the one in chiptool.
This PR is a good pair with https://github.com/embassy-rs/chiptool/pull/4
Co-authored-by: Nicolas Viennot <nicolas@viennot.biz>
The source files are unreadable because they're not fmt'd, and
they take up a LOT of space when generating docs for all 1200 chips
because they don't deduplicate.
* Chips that have multiple cores will be exposed as chipname_corename,
i.e. stm32wl55jc_cm4
* Chips that have single cores will use the chip family as feature name
and pick the first and only core from the list
* Add support for stm32wl55 chip family
Currently this looks up the frequency in the global singleton that must
be initialized by the per-chip RCC implementation. At present, this is
only done for the L0 family of chips.
* Moves the tim2-specific code into macro which always uses TIM2
* For peripherals without clock specified, attempt to locate enable and
reset registers in the RCC block matching the peripheral name. This
could be useful for peripherals where deducing the clock name might
not be feasible, but it remains to be tested with more chip families
to see if it is sufficiently accurate.
Adds RccPeripheral trait for peripherals implementing clock enable and reset for a given peripheral.
Add macro table generting implementations of RccPeripheral for peripherals with clock set, currently restricted to SPI.
