[clang] [llvm] [ARM] Introduce intrinsics for MVE minnm/maxnm under strict-fp. (PR #169795)

[Simon Tatham via llvm-commits]

https://github.com/statham-arm approved this pull request.

I see – I suppose there are potential transformations along the lines of max(a,b)+c ↔ max(a+c,b+c) which might be invalidated by flushing denormals to zero.

(In that case I suppose the optimization might already be inhibited by the + operator being a strict-fp intrinsic, but there's probably some other case I haven't thought of!)

https://github.com/llvm/llvm-project/pull/169795