# Proposal: Make integer overflow a logic error

**URL:** <https://forum.modular.com/t/proposal-make-integer-overflow-a-logic-error/2529>\
**Category:** Language Design\
**Created:** [December 16, 2025, 9:13am UTC](https://forum.modular.com/t/proposal-make-integer-overflow-a-logic-error/2529 "2025-12-16T09:13:16Z")\
**Posts on this page:** 1\
**Page:** 2

<div class="post-metadata">

**Author:** ![owenhilyard](https://sea1.discourse-cdn.com/flex001/user_avatar/forum.modular.com/owenhilyard/32/41_2.png) [@owenhilyard](https://forum.modular.com/u/owenhilyard)\
**Post date:** [December 25, 2025, 7:40pm UTC](https://forum.modular.com/t/proposal-make-integer-overflow-a-logic-error/2529/21 "2025-12-25T19:40:42Z")

</div>

I agree, LLVM’s `libfuzzer` is very, very good at rapidly converging on “interesting” values. Also, with values like `INT_MAX`, an `0` being important values to test, I’d argue that test code should expose places where overflow can happen if you aren’t taking a `Range[dtype, min, max, width]` type of some sort that compute where overflow can occur.

[Previous page](https://forum.modular.com/t/proposal-make-integer-overflow-a-logic-error/2529.md?page=1)
