This adds a pattern for i64 zext_inreg(i32 extract_vector_elt X), producing a single UMOVvi16 instruction that is already expected to clear the top bits. The exact pattern that this matches is and(anyext(vector_extract X, lane), 0xff), similar to the sext patterns higher up in the same file.
Details
Details
Diff Detail
Diff Detail
- Repository
- rG LLVM Github Monorepo