Formalize the IFR

[diondokter]

Closes #43

This PR gets rid of a lot of code and replaces it with:

• A typed model of the IFR using bitbybit
• A sort of builder for IFR provisioning

[diondokter]

Oh, are we ok with increasing the MSRV to 1.95?

[diondokter]

Ok, status:

• The bootloader can set the initial config
• The NXP tooling is fighting me very hard, so I haven't had it boot to a signed image, but I think that's because my image is not signed correctly
• I can't really test the rest of the provisioning since I don't have that code. But the provisioning now is a builder pattern. All functionality should be the same, you just have to migrate to the builder

[diondokter]

Some pointers to help with review

[felipebalbi]

Can we get miri in CI at the same time?

[alamfarjadf]

You can have the CFPA provisioned but CMPA left alone. On a dev unit, you will skip the initial provision step if anyone ever configures ACL settings (or any other CFPA bits) but leaves CMPA alone.

I think we should also check that CMPA header is 0xFFFF_FFFF along with CFPA page version 0xFFFF_FFFF.

[alamfarjadf]

Will the basic fusing /secure boot status helper functions make a comeback? Those would be helpful to expose to the application side (e.g. Is MCU fused? Hybrid SB tuned on? etc.)

[diondokter]

I did add check_production_settings in the provisioning builder that does those checks. The application can call that during the provisioning process

[diondokter]

TODO:

Only go into dev mode when the development lifecycle is active AND secure boot is not enabled.
This allows people to turn on secure boot and then find out if their hashes are wrong

[diondokter]

Ok, as discussed, the last extra check got in

[alamfarjadf]

You can have the CFPA provisioned but CMPA left alone. On a dev unit, you will skip the initial provision step if anyone ever configures ACL settings (or any other CFPA bits) but leaves CMPA alone.

I think we should also check that CMPA header is 0xFFFF_FFFF along with CFPA page version 0xFFFF_FFFF.

@diondokter last open item I believe.

[diondokter]

CI seems to be failing due to a rust bug: rust-lang/rust#164093

[felipebalbi]

@jerrysxie, this PR is ready. We need one more approval and a bypass because the "doc" failure is a known rustc regression.

[alamfarjadf]

#49 is all set! Thank you!