🚀 go-pugleaf

     I'm in the process of building gcc2 on a Sun SPARCstation, and I decided
to look at the output of enquire after building the first stage.  According to
enquire, gcc2 does not support the SPARC 128-bit floating point format.  (Long
doubles are the same as the 64-bit doubles.)  Sun's unbundled ANSI C compiler
implements long double using the 128-bit format, at least to the extent that
sizeof(long double) == 16, and you can do add, subtract, multiply and divide
to 128-bit precision.  (But there's still no long double math library; that is
supposed to be part of SunOS 5.0.)
     Is there something I've done wrong in compiling gcc2?  Is there some
switch I need to set, or is this work not done yet?

Brian