Lines Matching full:scalars
4773 // To use VPERMILPS to splat scalars, the second half of indicies must
4938 // Recurse into ISD::VECTOR_SHUFFLE node to find scalars.
4951 // Recurse into target specific vector shuffles to find scalars.
5722 // and 2) ensure that i64 scalars are eliminated on x86-32 hosts.
17511 // Optimize trunc store (of multiple scalars) to shuffle and store.