If they are ordered v0, v2, v1, v3, the compiler can find just a few simd optimizations itself. The new optimization I could observe on x86-64 was using 128 bit registers for the v = key ^ constant operations in new / reset. |
||
|---|---|---|
| .. | ||
| mod.rs | ||
| sip.rs | ||
If they are ordered v0, v2, v1, v3, the compiler can find just a few simd optimizations itself. The new optimization I could observe on x86-64 was using 128 bit registers for the v = key ^ constant operations in new / reset. |
||
|---|---|---|
| .. | ||
| mod.rs | ||
| sip.rs | ||