leanprover / leanprover/lean4

`partial` defeats `#eval` guardrails

Open
#10,790 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

bug P-medium
Dominant language
Lean
Stars
9.2k
Forks
990
Avg merge
1d 17h
Merged PRs (30d)
175

Description

Prerequisites

Please put an X between the brackets as you perform the following steps:

Description
partial def f (n : Nat) : Nat :=
  if n = 0 then
    #[1, 2, 3][4]'sorry
  else
    f (n - 1)

#eval f 1

causes an OOB access resulting in nonsense or crashes, even though #eval is supposed to prevent such things from happening by checking for sorrys.

Steps to Reproduce
  1. Copy the code above into live.lean-lang.org

Expected behavior: Error message aborting evaluation since the expression depends on the 'sorry' axiom, which can lead to runtime instability and crashes.

Actual behavior: nonsense output or crashes

Versions

Lean 4.25.0-nightly-2025-10-13

Impact

Add 👍 to issues you consider important. If others are impacted by this issue, please ask them to add 👍 to it.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by running the minimal reproducer from the issue in live.lean-lang.org and examine how #eval checks for sorry dependencies when the function is marked partial. Done means the example reports the expected sorry-abort error instead of producing nonsense output or crashing.

Written by the indexing model from the issue text.

Assessment

Domain
compilers
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Clearly specified
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.