Avoid localplt issues from x86 fereaiseexcept inline
Commit Message
Building for x86_64 with float128 support, I get a localplt test
failure from lrintf128 calling feraiseexcept.
The problem is that an inline optimized version of feraiseexcept calls
__feraiseexcept_renamed in cases where it doesn't completely expand
inline, and that in turn is redirected to feraiseexcept for a library
call, so meaning the redirection of feraiseexcept to
__GI_feraiseexcept inside libm is lost for that call.
This patch fixes the problem by moving the redirect to an internal
header in the _LIBC case, with the internal header using
__GI_feraiseexcept where appropriate. I'd welcome any comments on
whether this is the best approach for fixing this or if another way is
better.
Tested for x86_64 (in conjunction with float128 patches). Not yet
committed.
2017-06-22 Joseph Myers <joseph@codesourcery.com>
* sysdeps/x86/fpu/bits/fenv.h [_LIBC] (__feraiseexcept_renamed):
Do not declare.
* sysdeps/x86/fpu/include/bits/fenv.h [_LIBC &&
__USE_EXTERN_INLINES] (__feraiseexcept_renamed): Declare here,
redirected to __GI_feraiseexcept if [SHARED && IS_IN (libm)].
Comments
@@ -120,7 +120,9 @@ femode_t;
__BEGIN_DECLS
/* Optimized versions. */
+#ifndef _LIBC
extern int __REDIRECT_NTH (__feraiseexcept_renamed, (int), feraiseexcept);
+#endif
__extern_always_inline void
__NTH (__feraiseexcept_invalid_divbyzero (int __excepts))
{
@@ -17,6 +17,15 @@
<http://www.gnu.org/licenses/>. */
#ifndef _BITS_FENV_H
+
+#if defined _LIBC && defined __USE_EXTERN_INLINES
+# if defined SHARED && IS_IN (libm)
+extern int __REDIRECT_NTH (__feraiseexcept_renamed, (int), __GI_feraiseexcept);
+# else
+extern int __REDIRECT_NTH (__feraiseexcept_renamed, (int), feraiseexcept);
+# endif
+#endif
+
#include_next <bits/fenv.h>
# ifndef _ISOMAC