[compiler-rt] [builtins] Support building the 128-bit float functions on ld80 platforms (PR #68132)

[Sean Perry via llvm-commits]

perry-ca wrote:

Richard, I've discovered that this change regresses a build we have for the z/os platform.  It might impact others too.  The problem we've discovered is that __multc3 is no longer being defined for our builds.  We have QUAD_PRECISION but don't have TF_MODE.  This change is making multc3.cc empty unless CRT_HAS_TF_MODE is defined.   That breaks z/os and I suspect some other platforms.

This same problem impacts __divtc3.  I'm not sure about the entire scope yet.

https://github.com/llvm/llvm-project/pull/68132