[Bug libc/6981] __STDC_IEC_559__ should not be defined unconditionally

["joseph at codesourcery dot com" <sourceware-bugzilla at sourceware dot org>]

http://sourceware.org/bugzilla/show_bug.cgi?id=6981

--- Comment #14 from joseph at codesourcery dot com <joseph at codesourcery dot com> ---
On Sat, 8 Feb 2014, vincent-srcware at vinc17 dot net wrote:

> The bindings are complete when FLT_EVAL_METHOD is 0, because in such a case an
> operation on float numbers is rounded to the float precision, and an operation
> on double numbers is rounded to the double precision. That's why I suggested to
> define __STDC_IEC_559__ only when FLT_EVAL_METHOD is 0, until the new TS is
> implemented.

(a) __STDC_IEC_559__ is purely about support for Annex F, not about 
whether Annex F provides complete support for another standard in the case 
of the implementation choices made.

(b) As with __STDC__, __STDC_IEC_559__ is being defined in terms of intent 
(that the compiler options used indicate an intent consistent with Annex 
F, rather than specifying something incompatible like -ffinite-math-only) 
rather than completeness.

(c) Implementation of the new standard would trigger defining 
__STDC_IEC_60559_BFP__, rather than any change to __STDC_IEC_559__.

-- 
You are receiving this mail because:
You are on the CC list for the bug.