Andrew Leech

@dpgeorge there's something strange going on with running the finalizer, or not. This PR adds the print statement in `__del__` which is reflected in the unit test exp. This test...

> Could try to `gc.collect` a couple of times, that usually does the trick Yeah I did try that earlier, tried to trim the repros down to the minimum. Just...

That's really interesting, I got the impression that the number of allocations before and after was key here. More importantly, I need to test whether it's always reliability run on...

Yeah that sounds like what I was doing with these examples, adding more/less to them until the worked/didn't work. As a more important test I just added the example C...

> Do any of you have time to take a look at this proposed API? Unfortunately I've got very little actual knowledge about can usage, my interest is more in...

From what you've shown it can probably be tested on a standalone pybd with a miso pin simply tied high, should return 1's when working, 0's on failure. It does...

Assuming it could be a dma config issue; the dma configs are different for each SPI peripheral index (eg. `SPI(0)` vs `SPI(1)`). Is your failing test on a particular SPI...

I was asking about the index just to assist in reproducing and/or looking for code changes related to that index, I'll have a quick play with SPI(2) here and see...

Maybe re-run the failing code against the eeprom, ie. reproduce the failure again, then print out the pins to ensure they're still configured for spi correctly? eg. ``` >>> Pin('Y6')...

@projectgus There's more context to this change in the original PR it was included in: https://github.com/micropython/micropython/pull/13253#issuecomment-1893326692 https://github.com/pyserial/pyserial/issues/729