This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH v2] Provide a C++ version of issignaling that does not use __MATH_TG
- From: Joseph Myers <joseph at codesourcery dot com>
- To: "Gabriel F. T. Gomes" <gftg at linux dot vnet dot ibm dot com>
- Cc: <libc-alpha at sourceware dot org>
- Date: Fri, 18 Aug 2017 13:55:18 +0000
- Subject: Re: [PATCH v2] Provide a C++ version of issignaling that does not use __MATH_TG
- Authentication-results: sourceware.org; auth=none
- References: <20170818122438.18055-1-gftg@linux.vnet.ibm.com>
On Fri, 18 Aug 2017, Gabriel F. T. Gomes wrote:
> +/* There is no signaling_NaN for _Float128 in std::numeric_limits.
> + Include ieee754_float128.h and use the bitfields in the union
> + ieee854_float128.ieee_nan to build a signaling NaN. */
> +#if __HAVE_FLOAT128
> +#include <ieee754_float128.h>
> +#endif
That needs to be __HAVE_DISTINCT_FLOAT128 (and so the other conditionals
in this test need to be as well), since this internal header is in a
sysdeps directory that won't be used for float128-as-ABI-alias. Also note
preprocessor indentation, "# include".
(In the C++ case, float128-as-ABI-alias will be a typedef for long double
even with GCC 7 and later, so there's less need for C++ tests for it in
that case.)
--
Joseph S. Myers
joseph@codesourcery.com