[clang] [llvm] [clang][X86] Return Win64 __int128 indirectly (PR #210358)

[via cfe-commits]

Andarwinux wrote:

It's really terrible to deliberately choose a slower method. Unlike f128, int128 operations aren't necessarily very slow: https://godbolt.org/z/34jxfjx4E

> If that's the case then I don't think the performance difference is especially concerning. The type already needs to be passed in memory and even a simple f128 addition is ~300 instructions (~50 for the shortest nontrivial fast path), an extra movups won't matter much.

https://github.com/llvm/llvm-project/pull/210358