René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"ffi: Fast FFI bypasses `i32` range validation after V8 optimization","articleBody":"### Version\n\nmain\n\n### Platform\n\n```text\nmacOS 26.5.2\n```\n\n### Subsystem\n\nffi\n\n### What steps will reproduce the bug?\n\n```js\nimport { dlopen } from 'node:ffi';\n\nconst { lib, functions } = dlopen(null, {\n  abs: { arguments: ['i32'], return: 'i32' },\n});\n\nfunction call(value) {\n  return functions.abs(value);\n}\n\neval('%PrepareFunctionForOptimization(call)');\ncall(0);\ncall(0);\neval('%OptimizeFunctionOnNextCall(call)');\ncall(0);\n\nconsole.log(call(2 ** 32));\n\nlib.close();\n```\n\nRun with\n```console\n$ node --no-warnings --experimental-ffi --allow-natives-syntax repro.js\n```\n\n### How often does it reproduce? Is there a required condition?\n\nAlways\n\n### What is the expected behavior? Why is that the expected behavior?\n\nIt should throw `ERR_INVALID_ARG_VALUE` with `Argument 0 must be an int32`, as the generic FFI path does.\n\n### What do you see instead?\n\nThe optimized call accepts `2 ** 32`, truncates it to zero, and returns `0`\n\n### Additional information\n\n_No response_","author":{"url":"https://github.com/trivikr","@type":"Person","name":"trivikr"},"datePublished":"2026-07-23T06:30:16Z","url":"https://github.com/nodejs/node/issues/64690"}