PIN verification using Optiga

[andrewkozlik]

Creating this draft PR, so we can start discussion about the cryptographic aspects of the PIN verification process using Optiga. @onvej-sl

I did some refactoring around the ui_progress, which as it turns out is not really necessary, since we decided to set PIN_STRETCH_ITERATIONS to 1, so it takes no noticeable time (unless the security monitor kicks in). But I think the refactor still has some value, so I decided to keep it. It might be used in optiga.c in the future if we have our own configuration of the security monitor which will allow us to increase PIN_STRETCH_ITERATIONS. For now I removed some of the unused code in a separate commit, so that it can be reverted when the time comes.

[andrewkozlik]

@onvej-sl has reviewed the cryptographic solution and will do a final check tomorrow. I would like another set of eyes to go over the code, @matejcik please. I am hoping to get this merged tomorrow. The reason why it is in draft form is because it's branched off of the hashing-to-curve PR, which is not merged yet.

[andrewkozlik]

The last two commits 4419899 4ed316b that I added today reintroduce ui_progress() into the Optiga code. They don't change anything about the functionality. The execution of the Optiga part takes about 1 second to complete, so this smooths out the progress a bit better.

[matejcik]

LGTM code-wise, except the progess(0) thing you discovered

[andrewkozlik]

LGTM code-wise, except the progess(0) thing you discovered

Should be fixed by 38b331d