Add MiscellaneousSelectVerifier for ReportBody verification

[nick-mobilecoin]

No description provided.

[nick-mobilecoin]

Current dependencies on/for this PR:

• main
  • PR Add workspace elements #1
    • PR Generic And and Or verification steps #2
      • PR Port Attributes verification from MobileCoin repo #3
        • PR Port in ConfigId verifier for ReportBody #9
          • PR Add no alloc build to CI #10
            • PR Rework Verifiers to be passed in the evidence #11
              • PR Update verifiers to rely on subset of evidence #12
              • PR Add ConfigSvn verifier #18
                • PR Macroize the common ReportBody member Accessors #20
                  • PR Add IsvSvnVerifier #21
                  • PR Add MiscellaneousSelectVerifier for ReportBody verification #22 👈
                  • PR Add Verifier for MrEnclave #23
                  • PR Add Verifier for MrSigner #24
                  • PR Add Verifier for ReportData #25
                  • PR Use generic impl instead of self_accessor macro #26
                  • PR Consolidate equality logic behind Generic interface #30
                  • PR Add display capability to Verifier output #31
                  • PR Add Not verifier for negating verifier results #33
                  • PR Update to mc-sgx-*** 0.5.0 #34
                  • PR Update ReportDataVerifier to use a mask #35
                  • PR Add CpuSvnVerifier #38
    • PR Remove specific CODEOWNERS for metadata files #4

This comment was auto-generated by Graphite.

[codecov]

Codecov Report

Merging #22 (2a492d8) into nick/isv_version (f430127) will increase coverage by 0.58%.
The diff coverage is 97.22%.

@@                 Coverage Diff                  @@
##           nick/isv_version      #22      +/-   ##
====================================================
+ Coverage             91.86%   92.44%   +0.58%     
====================================================
  Files                     2        2              
  Lines                   295      331      +36     
====================================================
+ Hits                    271      306      +35     
- Misses                   24       25       +1     

Impacted Files	Coverage Δ
verifier/src/lib.rs	85.29% <ø> (ø)
verifier/src/report_body.rs	97.43% <97.22%> (-0.05%)	⬇️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[jcape]

This LGTM, but reading through all these, can they be collapsed into either a macro or a blanket impl over generics?

e.g. the blanket impls version would look something like:

/// A marker trait supporting objects which should be verified with ct_eq()

trait IntoVerificationError {
    fn into_verification_error(expected: Self, actual: Self) -> VerificationError;
}

trait VerifiedWithCtEq: ConstantTimeEq + IntoVerificationError {}

// this can be macro-ized.
impl IntoVerificationError for MiscellaneousSelect {
    fn into_verification_error(expected, Self, actual: Self) -> VerificationError {
        VerificationError::MiscellaneousSelectMismatch {
            expected,
            actual
        }
    }
}
impl VerifiedWithCtEq for MiscellaneousSelect {}

// etc.
impl VerifiedWithCtEq for ConfigId {}
impl VerifiedWithCtEq for ExtProdId {}

/// Equality verifier
pub struct EqualityVerifier<T: VerifiedWithEq + Clone> {
    expected: T
}

impl<T: VerifiedWithCtEq, E: Accessor<T>> Verifier<E> for EqualityVerifier<T> {
    type Error = VerificationError;

    fn verify(&self, evidence: &E) -> CtOption<Self::Error> {
        let expected = self.expected.clone();
        let actual = evidence.get();

        let is_eq = expected.ct_eq(actual);
        CtOption::new(
            T::into_verification_error(expected, actual),
            !is_eq
        )
    }
}

Whereas a macro would just macroize the Verifier impl at the end, but would require having everything be visible from the macro.

[nick-mobilecoin]

bump @varsha888 due to restack

[nick-mobilecoin]

This LGTM, but reading through all these, can they be collapsed into either a macro or a blanket impl over generics?

Will work on this as a follow on PR on top of #26 or after all of them are merged, added task to #7