Retry failed pages after verify

[ygoe]

I've recently updated my pyupdi copy and see that the tx echo check was removed. This was the major source of failures for me in the past.

Now it still happens from time to time that the read verification reports incorrect bytes. Often just one or two, sometimes a lot more. I've hacked the code in a way that it just tries reading it all a second time if something was incorrect. And then, it often works without any errors, or sometimes with another number of wrong bytes at other positions.

So the read verification doesn't seem to be very reliable either. I'm thinking about changing the code so that it re-reads the failed pages once again, and also re-writes these pages if the error persists. The idea is to make the flash process reliable with minimal NVM write wear.

Is this something you'd be interested in? Could you give me some hints about how to do that? I've never written anything in Python before so I'll have to figure out the syntax and language basics on the way. (I already started this for my first pull request here today.)