fix[venom]: reject statically undersized abi_decode input#5094
Open
banteg wants to merge 4 commits into
Open
Conversation
| assert_compile_failed(lambda: get_contract(bad_code), exception) | ||
|
|
||
|
|
||
| def test_abi_decode_undersized_buffer_venom(): |
Member
There was a problem hiding this comment.
do we not already have a test for this which is just not tripped on venom pipeline?
Contributor
Author
There was a problem hiding this comment.
Yes, there was already a length-mismatch compile-failure table. I folded the public abi_decode(data, (uint256, uint256)) undersized-buffer case into test_abi_decode_length_mismatch and removed the separate Venom-named test in b0de8d8. Validation: uv run --python 3.12 -m pytest tests/functional/builtins/codegen/test_abi_decode.py -k length_mismatch -q and the focused black check.
charles-cooper
approved these changes
Jun 15, 2026
charles-cooper
left a comment
charles-cooper
left a comment
Member
There was a problem hiding this comment.
ok in principle; see my testing comment. paging @HodanPlodky for review
Gas ChangesNo changes detected. Summary
|
📊 Bytecode Size Changes (venom)No changes detected. Full bytecode sizes
|
HodanPlodky
reviewed
Jun 18, 2026
| ( | ||
| """ | ||
| @external | ||
| def foo(data: Bytes[4]) -> (uint256, uint256): |
Collaborator
There was a problem hiding this comment.
this test passes also in the master why was the change in logic necessary?
Contributor
Author
There was a problem hiding this comment.
You're right - the folded table row passes on master in a default local run because that path uses legacy codegen. The regression is only in Venom.
I added test_abi_decode_undersized_buffer_venom, which forces Settings(experimental_codegen=True): on origin/master the reproducer compiles, while this branch raises StructureException. I also updated the PR body to make that distinction explicit.
Validation:
uv run --python 3.12 -m pytest tests/functional/builtins/codegen/test_abi_decode.py -k "length_mismatch or undersized_buffer_venom" -q
uv run --python 3.12 -m pytest tests/functional/builtins/codegen/test_abi_decode.py -k "length_mismatch or undersized_buffer_venom" --experimental-codegen -q
uv run --python 3.12 black --check tests/functional/builtins/codegen/test_abi_decode.py vyper/codegen_venom/builtins/abi.py
git diff --check
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What I did
Fixes #5055.
The Venom
abi_decodelowering was missing legacy's compile-time check that the input bytestring's max length can fit the ABI size bound of the requested output type.abi_decode(data, (uint256, uint256))withdata: Bytes[4]compiled under--experimental-codegeninto code that can only revert at runtime, while legacy rejects it at compile time.How I did it
Ported the check from legacy
ABIDecode.build_IRintolower_abi_decode, afterwrapped_typis computed and before any IR is emitted. The condition is the same as legacy: reject wheninput_max_len < wrapped_typ.abi_type.size_bound(), usingStructureExceptionat the input argument source location.How to verify it
The shared length-mismatch table now includes the public
abi_decode(data, (uint256, uint256))shape. That row can pass on master in a default local run because it is exercising legacy codegen there; the differentiating regression is the explicittest_abi_decode_undersized_buffer_venom, which forcesSettings(experimental_codegen=True).I checked that the direct experimental-codegen reproducer compiles on
origin/masterand raisesStructureExceptionon this branch.Focused validation:
Commit message
Description for the changelog
Fix: experimental codegen rejects
abi_decodewith statically undersized input buffers at compile time, matching legacy.Cute Animal Picture