Rethink STM32 I2C v2 HAL (#78) ...

* Initial attempt at rethinking the STM32 I2C v2 HAL.  Makes single-byte work properly and adds a new 'state' variable to track what the hardware is doing.

* Fix some initial test failures

* Fix incorrect logic

* Fix more incorrect logic

* Tabs to spaces

* Fix repeated starts with single-byte API

* Fix race condition causing stop() after nacked address to sometimes break things

* Fix missed i2c structs that should have been removed

* Fix doing a repeated start from single-byte to transaction API causing I2C peripheral to lock up

* Fix xferOperation being set wrong for repeated starts, causing the peripheral to hang

* Fix race condition with repeated start after single-byte operation

* Fix compilation for targets that use I2C IP v1

* Fix initialization of XferOperation for API v1, optimize stop()

* Remove unneeded line

* Add docs for I2C events

STM32: increase i2c slave rx limit. ...

Use uint16_t variables for i2c slave_rx_buffer_size and slave_rx_count
variables. This allows to receive more than 255 bytes in slave mode. The
bytes are received one by one in slave mode so there are no hardware
limitations forcing a 1 byte rx count limit.

B_U585I_IOT02A supports OSPI

STM32L5/STM32U5 : CAN suport

STM32U5 : B_U585I_IOT02A introduction

STM32U5 : NUCLEO_U575ZI_Q introduction

STM32U5 : STM32U5xxxxx sub-families

STM32U5 specific driver files

STM32U5 CMSIS update

STM32U5 / GPIO HAL : GPIO_AF3_TIM1 is missing

STM32U5 ST_HAL_SPI workaround ...

Internal ticket 112451

STM32U5 / CMSIS : USB_OTG patch

STM32U5: STM32Cube_FW_U5_V1.0.0