Feature/digital input stamp

[ryan-summers]

This PR implements support for asynchronous collection of timestamps on the DI0 input.

TODO:

• [DMA] Next transfer returns number of remaining data stm32-rs/stm32h7xx-hal#173 should get merged into the stm32-rs/dma branch
• Add manual configuration of sampling + timestamping timer prescalers + periods as per Timestamping #170
• Verify the accurate collection of timestamp values

[ryan-summers]

As a test case, I input a 10Hz square wave (3V amplitude) to DI0 and collected 32 timestamps. I then calculated the delta between the timestamps and printed it using GDB:

892                     info!("All stamps gathered: {:?}", deltas);
(gdb) p/x deltas
$1 = [0x9894d3, 0x9894d2, 0x9894d2, 0x9894d2, 0x989dc7, 0x9894d2, 0x9894d2, 0x9894d1, 0x9894d1, 
  0x9894d1, 0x9894d0, 0x9894cf, 0x9894d0, 0x9894d0, 0x989dc5, 0x9894d0, 0x9894d1, 0x9894d1, 0x9894d3, 
  0x9894d2, 0x9894d2, 0x9894d2, 0x9894d1, 0x9894d1, 0x989dc6, 0x9894d0, 0x9894d1, 0x9894d2, 0x9894d2, 
  0x9894d2, 0x9894d3]
(gdb) p deltas
$2 = [9999571, 9999570, 9999570, 9999570, 10001863, 9999570, 9999570, 9999569, 9999569, 9999569, 
  9999568, 9999567, 9999568, 9999568, 10001861, 9999568, 9999569, 9999569, 9999571, 9999570, 9999570, 
  9999570, 9999569, 9999569, 10001862, 9999568, 9999569, 9999570, 9999570, 9999570, 9999571]
(gdb) 

As can be seen, the deltas are all approximately near the count of 10,000,000, which corresponds with 100ms (the period of a 10Hz wave) with the timestamp timer running at 100MHz. The variations here are likely due to inaccuracies in my clock source as opposed to fluctuations in the timer timestamp captures.

[jordens]

As a test case, I input a 10Hz square wave (3V amplitude) to DI0 and collected 32 timestamps. I then calculated the delta between the timestamps and printed it using GDB:

Nice!
But do check the max input range of the DIO (from schematics). I don't know whether it deals with negative voltages.

[jordens]

Good from my end.