v1.7.4-rc.1 : Use internal bn_gf2_n_mul() function instead of cx_bn_gf2_n_mul() syscall

[aido]

Checklist

• App update process has been followed
• Target branch is develop
• Application version has been bumped

Description

This pull request is a pre-release of v1.7.4

See discussion at aido#32 for some background.

Reliance on the cx_bn_gf2_n_mul() syscall has been removed in favour of the internal bn_gf2_n_mul() function for all devices. If this version does not have the same issues as described in aido#32 then we will know for sure that cx_bn_gf2_n_mul() is causing issue.

If cx_bn_gf2_n_mul() is shown to be the issue we can then release a v1.7.4-rc.2 that implements the suggestion from @srasoamiaramanana-ledger i.e.

Could you try to call cx_bn_gf2_n_mul() without overlapping the variable for the result and the variables for the operand ?

[aido]

As an aside, I notice that the "Compilation & test" action consistently fails on every pull request due to issues with a token when uploading to codecov.io.

There may be two ways to fix this:

1. Install the GitHub Codecov app for this repo
   or
2. Change the following to false
   

   app-seed-tool/.github/workflows/ci-workflow.yml

   Line 63 in f188c9f

   fail_ci_if_error: true

Which would you prefer?

[lpascal-ledger]

I've configured the repo with the correct Codecov token, yet it fails still Rate limit reached. Please upload with the Codecov repository upload token to resolve issue. Srly annoying. Maybe it'll apply once it's merged?

[lpascal-ledger]

@aido 1.7.4-rc1 was deployed on NanoS, S+ and X

[aido]

@lpascal-ledger

Srly annoying. Maybe it'll apply once it's merged?

The next PR will bump the Codecov GitHub action from v3 to v4. Let's see if that helps.

[lpascal-ledger]

@aido also the workflows triggered by the merge and the tag were successful, that's promising.