Summary

This PR converts tests/extmod/machine_spi_rate.py to use the target wiring mechanism, moving the port-specefic definition of the SPI instances that are being tested out of the tests/extmod/machine_spi_rate.py file and into the target wiring code. As part of this, the test is changed to always use the default pins for the given SPI instance.

This simplifies the test, makes it easier to add support for new ports/boards, and allows reusing the spi_standalone_instances definition in other tests (if needed one day).

It also adds SPI test support for the mimxrt, nrf, renesas-ra and samd ports.

Testing

Tested by running tests/extmod/machine_spi_rate.py on (thanks to octoprobe hardware!):

• ARDUINO_NANO_33_BLE_SENSE
• ESP32_GENERIC_C3
• ESP32_GENERIC_P4
• ESP32_GENERIC_S3
• ESP8266_GENERIC
• PYBD_SF2
• PYBV10
• RPI_PICO
• RPI_PICO_W
• SEEED_XIAO_SAMD21
• TEENSY40

The test passes.

Trade-offs and Alternatives

There are few decisions made here:

1. This is not technically "target wiring" because nothing needs to be wired up. It just uses the SPI as-is to measure TX speed. But I think it's still fair enough to configure the SPI instances in the target wiring definitions. It keeps everything self contained.
2. Always use the default SPI pins for the given instance, instead of explicitly specifying them. Most ports already did this in the test, except for rp2 and esp32, so it's not a big difference in the way the test works. And this makes it much simpler to specify the SPI instances to test, you just specify the id (or None to use the default SPI instance).
3. The new spi_standalone_instances variable in the target wiring is a little different in name to existing variables used there, but I think it's acceptable.