[PATCH] math: Refactor how to use libm-test-ulps

[Joseph Myers]

On Tue, 28 Jan 2025, Adhemerval Zanella Netto wrote:

> > I think it would be better for the keywords to be based on the format, not 
> > the type.  (In particular, if in future we have correctly rounded 
> > binary128 implementations, of rsqrt or cbrt or hypot, say, it would be 
> > desirable to have a single libm-test-ulps file describing those, rather 
> > than needing separate files for when those implementations are used for 
> > long double and when they are used for _Float128.  Though since those 
> > cases get the implementations via different sysdeps directories, our 
> > sysdeps system isn't really designed for this; the file would need to go 
> > somewhere more generic such as sysdeps/ieee754/.)
> 
> Are you suggesting changing to something line binary32, binary64, binary80,
> binary128, and ibmlonglong? If so it makes sense; although I think it would
> be a different and more extensive patch.

I suggest changing to the same names as used in auto-libm-test-*.  
There's no such thing as binary80 (binaryN is only defined for N = 16, 32, 
64, and N >= 128 a multiple of 32); the x86 long double format meets the 
IEEE 754 definition of a binary64-extended format (the m68k long double 
format doesn't).

-- 
Joseph S. Myers
josmyers@redhat.com