float128: Add signbit alternative for old compilers
Commit Message
In math/math.h, __MATH_TG will expand signbit to __builtin_signbit*,
e.g.: __builtin_signbitf128, before GCC 6. However, there has never
been a __builtin_signbitf128 in GCC and the type-generic builtin is
only available since GCC 6. For older GCC, this patch defines
__builtin_signbitf128 to __signbitf128, so that the internal function
is used instead of the non-existent builtin.
This patch also changes the implementation of __signbitf128, because
it was reusing the implementation of __signbitl from ldbl-128, which
calls __builtin_signbitl. Using the long double version of the
builtin is not correct on machines where _Float128 is ABI-distinct
from long double (i.e.: ia64, powerpc64le, x86, x86_84). The new
implementation does not rely on builtins, yet it doesn't harm builds
with newer compilers, because __MATH_TG selects the builtin in those
cases.
The new code does not currently affect powerpc64le builds, because
only GCC 6.2 fulfills the requirements from configure. It might
affect powerpc64le builds if those requirements are backported to
older versions of the compiler. The new code affects x86_64 builds,
since glibc is supposed to build correctly with older versions of GCC.
Tested for powerpc64le and x86_64.
* include/math.h (__signbitf128): Define as hidden.
* sysdeps/ieee754/float128/s_signbitf128.c (__signbit):
Reimplement without builtins.
* sysdeps/ia64/bits/floatn.h [!__GNUC_PREREQ (6, 0)]
(__builtin_signbitf128): Define to __signbitf128.
* sysdeps/powerpc/bits/floatn.h: Likewise.
* sysdeps/x86/bits/floatn.h: Likewise.
---
include/math.h | 1 +
sysdeps/ia64/bits/floatn.h | 8 ++++++++
sysdeps/ieee754/float128/s_signbitf128.c | 31 ++++++++++++++++++++++++++++++-
sysdeps/powerpc/bits/floatn.h | 8 ++++++++
sysdeps/x86/bits/floatn.h | 8 ++++++++
5 files changed, 55 insertions(+), 1 deletion(-)
Comments
On Fri, 30 Jun 2017, Gabriel F. T. Gomes wrote:
> +int
> +__signbitf128 (_Float128 x)
> +{
> + int64_t e;
> +
> + GET_FLOAT128_MSW64 (e, x);
> + return e < 0;
I think you should use __builtin_signbit for GCC >= 6, with a comment that
once GCC >= 6 is required for building glibc, this file can go back to
wrapping the float128 version. OK with that change.
On Fri, 30 Jun 2017 20:40:25 +0000
Joseph Myers <joseph@codesourcery.com> wrote:
> On Fri, 30 Jun 2017, Gabriel F. T. Gomes wrote:
>
> > +int
> > +__signbitf128 (_Float128 x)
> > +{
> > + int64_t e;
> > +
> > + GET_FLOAT128_MSW64 (e, x);
> > + return e < 0;
>
> I think you should use __builtin_signbit for GCC >= 6 with a comment that
> once GCC >= 6 is required for building glibc, this file can go back to
> wrapping the float128 version.
OK.
> OK with that change.
Is this OK to go in now, before the freeze?
On Fri, 30 Jun 2017, Gabriel F. T. Gomes wrote:
> On Fri, 30 Jun 2017 20:40:25 +0000
> Joseph Myers <joseph@codesourcery.com> wrote:
>
> > On Fri, 30 Jun 2017, Gabriel F. T. Gomes wrote:
> >
> > > +int
> > > +__signbitf128 (_Float128 x)
> > > +{
> > > + int64_t e;
> > > +
> > > + GET_FLOAT128_MSW64 (e, x);
> > > + return e < 0;
> >
> > I think you should use __builtin_signbit for GCC >= 6 with a comment that
> > once GCC >= 6 is required for building glibc, this file can go back to
> > wrapping the float128 version.
>
> OK.
>
> > OK with that change.
>
> Is this OK to go in now, before the freeze?
Yes.
@@ -31,6 +31,7 @@ hidden_proto (__isnanl)
hidden_proto (__finitef128)
hidden_proto (__isinff128)
hidden_proto (__isnanf128)
+hidden_proto (__signbitf128)
# endif
# endif
@@ -89,6 +89,14 @@ typedef __float128 _Float128;
# define __builtin_nansf128(x) ((_Float128) __builtin_nans (x))
# endif
+/* In math/math.h, __MATH_TG will expand signbit to __builtin_signbit*,
+ e.g.: __builtin_signbitf128, before GCC 6. However, there has never
+ been a __builtin_signbitf128 in GCC and the type-generic builtin is
+ only available since GCC 6. */
+# if !__GNUC_PREREQ (6, 0)
+# define __builtin_signbitf128 __signbitf128
+# endif
+
#endif
#endif /* _BITS_FLOATN_H */
@@ -1,2 +1,31 @@
+/* Return nonzero value if number is negative.
+ Copyright (C) 2017 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, see
+ <http://www.gnu.org/licenses/>. */
+
#include <float128_private.h>
-#include "../ldbl-128/s_signbitl.c"
+#include <math.h>
+#include <math_private.h>
+
+int
+__signbitf128 (_Float128 x)
+{
+ int64_t e;
+
+ GET_FLOAT128_MSW64 (e, x);
+ return e < 0;
+}
+hidden_def (__signbitf128)
@@ -87,6 +87,14 @@ typedef __float128 _Float128;
# define __builtin_nansf128 __builtin_nansq
# endif
+/* In math/math.h, __MATH_TG will expand signbit to __builtin_signbit*,
+ e.g.: __builtin_signbitf128, before GCC 6. However, there has never
+ been a __builtin_signbitf128 in GCC and the type-generic builtin is
+ only available since GCC 6. */
+# if !__GNUC_PREREQ (6, 0)
+# define __builtin_signbitf128 __signbitf128
+# endif
+
#endif
#endif /* _BITS_FLOATN_H */
@@ -91,6 +91,14 @@ typedef __float128 _Float128;
# define __builtin_nansf128(x) ((_Float128) __builtin_nans (x))
# endif
+/* In math/math.h, __MATH_TG will expand signbit to __builtin_signbit*,
+ e.g.: __builtin_signbitf128, before GCC 6. However, there has never
+ been a __builtin_signbitf128 in GCC and the type-generic builtin is
+ only available since GCC 6. */
+# if !__GNUC_PREREQ (6, 0)
+# define __builtin_signbitf128 __signbitf128
+# endif
+
#endif
#endif /* _BITS_FLOATN_H */