[Bug math/22235] New: iscanonical in C++ and float128

["jsm28 at gcc dot gnu.org" <sourceware-bugzilla at sourceware dot org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=22235

            Bug ID: 22235
           Summary: iscanonical in C++ and float128
           Product: glibc
           Version: 2.26
            Status: NEW
          Severity: normal
          Priority: P2
         Component: math
          Assignee: unassigned at sourceware dot org
          Reporter: jsm28 at gcc dot gnu.org
                CC: gabriel at inconstante dot eti.br
  Target Milestone: ---

Similar to the other such issues with __MATH_TG not working in C++ in
configurations with float128 support, the iscanonical macro also uses __MATH_TG
in certain configurations, and, not being a C99 macro, is not part of the
standard C++ library, so we need to handle it like issignaling and iszero
(providing C++ overloads).

The default definition in the top-level bits/iscanonical.h is fine (it doesn't
use __MATH_TG, just evaluates its argument and returns 1).  It's
sysdeps/ieee754/ldbl-96/bits/iscanonical.h and
sysdeps/ieee754/ldbl-128ibm/bits/iscanonical.h that need fixing to provide
overloads when they would overwise use __MATH_TG.

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