In the Intel Itanium Architecture Software Developer's Manual, Volume 1:Application Architecture, Rev 2.1 Dated October 2002, in section 5.1, Data Types and Formats of the Floating-point Programming Model chapter, it states:
A seventh data type, IEEE-style quad precision, is supported by software routines. A future architecture extension may include additional support for the quad-precision real type.
Question: Are these software routines available for the Itanium 2? If so, where does one gain access to them? Will these future architecture extensions be included in Madison?
The C compiler has a data type _Quad, as you can see by running strings on it.
Likewise, youcould try todecipher the function of the long_double_size_128 option. As these things aren't documented, one thing you can be sure of is they haven't been tested as thoroughly as more standard features. Fortranquad format is documented.
I don't know what distinction you could make between Itanium2 and Madison. As far as I know, there aren't any significant differences at the floating point instruction set level all the way from Madison to Montecito.