[PATCH] Fix the C++ version of issignaling when __NO_LONG_DOUBLE_MATH is defined

["Gabriel F. T. Gomes" <gftg at linux dot vnet dot ibm dot com>]

On Wed, 23 Aug 2017 00:41:23 +0000
Joseph Myers <joseph@codesourcery.com> wrote:

>This patch breaks building tests for long double = double configurations.
>__issignalingl does not exist in that case; the long double version of
>issignaling has to call __issignaling if __NO_LONG_DOUBLE_MATH is defined.

The following patch should fix this, however I do not have access to a
system where __NO_LONG_DOUBLE_MATH is defined.  I'm building this with
build-many-glibcs.py, but that will take some time.

OK for master?

-- 8< --
When __NO_LONG_DOUBLE_MATH is defined, __issignalingl is not available,
thus issignaling with long double argument should call __issignaling,
instead.

Tested for powerpc64le.

	* math/math.h [defined __cplusplus] (issignaling): In the long
	double case, call __issignalingl only if __NO_LONG_DOUBLE_MATH
	is not defined.  Call __issignaling, otherwise.
---
 math/math.h | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/math/math.h b/math/math.h
index b5d6c43fcf..5acbe88906 100644
--- a/math/math.h
+++ b/math/math.h
@@ -486,7 +486,15 @@ enum
 extern "C++" {
 inline int issignaling (float __val) { return __issignalingf (__val); }
 inline int issignaling (double __val) { return __issignaling (__val); }
-inline int issignaling (long double __val) { return __issignalingl (__val); }
+inline int
+issignaling (long double __val)
+{
+#  ifdef __NO_LONG_DOUBLE_MATH
+  return __issignaling (__val);
+#  else
+  return __issignalingl (__val);
+#  endif
+}
 #  if __HAVE_DISTINCT_FLOAT128
 inline int issignaling (_Float128 __val) { return __issignalingf128 (__val); }
 #  endif
-- 
2.13.5