[flang-commits] [flang] [flang] optimize `tand` precision (PR #155629)

[Tom Eccles via flang-commits]

tblah wrote:

We support f128 if either libquadmath is available or if libm defines macros indicating that it supports f128: `LDBL_MANT_DIG == 113`. See https://github.com/llvm/llvm-project/pull/106957

https://github.com/llvm/llvm-project/pull/155629