3 ms·
3% code size reduction while changing semantics is borderline worthless*. 3% code size reduction without changing semantics would be more interesting but might
by JonChesterfield 3y ago
3% code size reduction while changing semantics is borderline worthless*.
3% code size reduction without changing semantics would be more interesting but might still be a bad thing for performance.
*fast-math etc is a thing, where similar-enough output is fine
- cec 3y agoHey Jon, we don't change semantics. We just choose optimization passes to run to match a particular input code. Agreed that code size wins could regress performance. We don't measure that yet, but will be looking into it next. There are some applications where codesize above-all-else is what to optimize for.