easy integration of SoftFloat functions with qfloat{80,128}
Note that [at least] on MSVS 2015, a number of obnoxious but innocuous warnings are produced when compiling qsoftfloat.lib, related to certain inline functions being already defined (the seconds ones ignored). This seems to be due to the way that SoftFloat determines whether its `inline' mechanisms have previously indicated that the function was indicated to be inline, which doesn't seem to work with MSVS 2015.
Furthermore, because those [should-be] inlined functions are the only things in certain source code files, such as s_mul128By32.c, the object file s_mul128By32.obj is excluded from the .lib.
With a certain amount of patching these warnings could be avoided.