We were already doing this for integer constants. This patch implements the same thing for floating point constants. Differential Revision: https://reviews.llvm.org/D88570
313 KiB
313 KiB