Most of the test cases for folding fail to compile if we force it to be compiled into brillig with an ICE. See below:
tom at tom-desktop in ~/Programming/aztec/noir/test_programs/execution_success/fold_fibonacci on tf/enforce-testing-in-brillig [$]
$ nargo compile --force-brillig
error: Internal Consistency Evaluators Errors:
This is likely a bug. Consider opening an issue at https://github.com/noir-lang/noir/issues
┌─ std/aes128.nr:1:1
│
1 │ #[foreign(aes128_encrypt)]
│ - ICE: "Cannot find linked fn f1"
│
Aborting due to 1 previous error
If we're compiling in brillig can make a call to a #[fold] function we should be able to just ignore that and treat it as a regular brillig function.
Also this diagnostic is quite odd, not sure why this is pointing at aes128.
See PR https://github.com/noir-lang/noir/pull/5128
Most of the test cases for folding fail to compile if we force it to be compiled into brillig with an ICE. See below:
If we're compiling in brillig can make a call to a
#[fold]
function we should be able to just ignore that and treat it as a regular brillig function.Also this diagnostic is quite odd, not sure why this is pointing at aes128.