chore: interpret all execution_success programs#8887
Conversation
There was a problem hiding this comment.
This is very nice to have. I only wish for the tests that are currently failing we could assert what we expect the error to be but I don't think we need that in this PR (or at all if we can fix these issues quickly enough). It'd just help to remind us when these issues are fixed to un-ignore the relevant test.
I was thinking about this some minutes ago. Instead of ignoring these tests, it would be nice to still run them and have them pass if they fail, but have them fail if they pass. I'll try to do it in a follow-up PR for all ignored tests (at least in nargo_cli). |
Automated pull of nightly from the [noir](https://github.com/noir-lang/noir) programming language, a dependency of Aztec. BEGIN_COMMIT_OVERRIDE chore: interpret all execution_success programs (noir-lang/noir#8887) fix: always error if integer literal doesn't fit its type on the fron… (noir-lang/noir#8885) fix: Count array copies for slice functions (noir-lang/noir#8867) chore(ssa): Simplify truncate when the bit size we are truncating to is greater than the value's max bit size (noir-lang/noir#8875) fix(ssa): Validate field to integer cast (noir-lang/noir#8799) feat: SSA interpreter checks return value (noir-lang/noir#8883) chore: add return value to every execution_success program that produces an output (noir-lang/noir#8882) chore(fuzz): Run fuzzing deterministically on CI (noir-lang/noir#8868) fix: (SSA interpreter) check requires_acir_gen_predicate for enable_side_effects (noir-lang/noir#8869) feat: allow printing each SSA interpreter definition (noir-lang/noir#8865) fix: handle unconditional break during SSA codegen in all cases (noir-lang/noir#8855) fix: update external lib commit tdd.nr (noir-lang/noir#8823) chore: bump dependencies (noir-lang/noir#8838) chore(fuzz): Enable SSA Interpreter fuzzing on all passes (noir-lang/noir#8610) chore: Prune changelog older than ~1 year (<0.32.0) (noir-lang/noir#8856) chore(docs): Add links to awesome-noir in sidebar (noir-lang/noir#8854) chore: add a regression test for #8727 (noir-lang/noir#8851) chore(fuzz): Handle overflows in comptime fuzzing (noir-lang/noir#8847) END_COMMIT_OVERRIDE --------- Co-authored-by: AztecBot <tech@aztecprotocol.com>
Automated pull of nightly from the [noir](https://github.com/noir-lang/noir) programming language, a dependency of Aztec. BEGIN_COMMIT_OVERRIDE chore: interpret all execution_success programs (noir-lang/noir#8887) fix: always error if integer literal doesn't fit its type on the fron… (noir-lang/noir#8885) fix: Count array copies for slice functions (noir-lang/noir#8867) chore(ssa): Simplify truncate when the bit size we are truncating to is greater than the value's max bit size (noir-lang/noir#8875) fix(ssa): Validate field to integer cast (noir-lang/noir#8799) feat: SSA interpreter checks return value (noir-lang/noir#8883) chore: add return value to every execution_success program that produces an output (noir-lang/noir#8882) chore(fuzz): Run fuzzing deterministically on CI (noir-lang/noir#8868) fix: (SSA interpreter) check requires_acir_gen_predicate for enable_side_effects (noir-lang/noir#8869) feat: allow printing each SSA interpreter definition (noir-lang/noir#8865) fix: handle unconditional break during SSA codegen in all cases (noir-lang/noir#8855) fix: update external lib commit tdd.nr (noir-lang/noir#8823) chore: bump dependencies (noir-lang/noir#8838) chore(fuzz): Enable SSA Interpreter fuzzing on all passes (noir-lang/noir#8610) chore: Prune changelog older than ~1 year (<0.32.0) (noir-lang/noir#8856) chore(docs): Add links to awesome-noir in sidebar (noir-lang/noir#8854) chore: add a regression test for #8727 (noir-lang/noir#8851) chore(fuzz): Handle overflows in comptime fuzzing (noir-lang/noir#8847) END_COMMIT_OVERRIDE --------- Co-authored-by: AztecBot <tech@aztecprotocol.com>
I just tried implementing this but it's tricky because for interpret tests, some pass normally but fail with For now I'll abandon this, though I'll try to fix make the interpreter tests pass. |
Automated pull of nightly from the [noir](https://github.com/noir-lang/noir) programming language, a dependency of Aztec. BEGIN_COMMIT_OVERRIDE chore: interpret all execution_success programs (noir-lang/noir#8887) fix: always error if integer literal doesn't fit its type on the fron… (noir-lang/noir#8885) fix: Count array copies for slice functions (noir-lang/noir#8867) chore(ssa): Simplify truncate when the bit size we are truncating to is greater than the value's max bit size (noir-lang/noir#8875) fix(ssa): Validate field to integer cast (noir-lang/noir#8799) feat: SSA interpreter checks return value (noir-lang/noir#8883) chore: add return value to every execution_success program that produces an output (noir-lang/noir#8882) chore(fuzz): Run fuzzing deterministically on CI (noir-lang/noir#8868) fix: (SSA interpreter) check requires_acir_gen_predicate for enable_side_effects (noir-lang/noir#8869) feat: allow printing each SSA interpreter definition (noir-lang/noir#8865) fix: handle unconditional break during SSA codegen in all cases (noir-lang/noir#8855) fix: update external lib commit tdd.nr (noir-lang/noir#8823) chore: bump dependencies (noir-lang/noir#8838) chore(fuzz): Enable SSA Interpreter fuzzing on all passes (noir-lang/noir#8610) chore: Prune changelog older than ~1 year (<0.32.0) (noir-lang/noir#8856) chore(docs): Add links to awesome-noir in sidebar (noir-lang/noir#8854) chore: add a regression test for AztecProtocol#8727 (noir-lang/noir#8851) chore(fuzz): Handle overflows in comptime fuzzing (noir-lang/noir#8847) END_COMMIT_OVERRIDE --------- Co-authored-by: AztecBot <tech@aztecprotocol.com>
Description
Problem
No issue, but will help us improve the interpreter.
Summary
Runs
nargo interpretagainst all execution_success test programs, except a few that currently don't work well.Also fixes one issue where the ABI was built in a way that doesn't match how it's built everywhere else (this made some tests fail, which were around 43, and then the number got down to 35).
I decided to not keep fixing issues here as there seem to be a few of them. Some of them might be easier to fix (like the "IntrinsicArgumentCountMismatch" ones), but in any case we can probably fix them in follow-up PRs.
Additional Context
Documentation
Check one:
PR Checklist
cargo fmton default settings.