[v2] Enable inlining issignalingf within glibc
Commit Message
From: "Paul A. Clarke" <pc@us.ibm.com>
issignalingf is a very small function used in some areas where
better performance (and smaller code) might be helpful.
Establish a means to inline issignalingf, and use that for powerpc.
2019-09-25 Paul A. Clarke <pc@us.ibm.com>
* math/s_canonicalize_template.c: Include math_private.h
* math/s_fmin_template.c: Likewise.
* math/s_fmax_template.c: Likewise.
* math/s_fmaxmag_template.c: Likewise.
* math/s_fminmag_template.c: Likewise.
* sysdeps/ieee754/flt-32/s_issignalingf.c: Move code body
to __issignalingf_inline. Undefine __issignalingf.
Include s_issignalingf_inline.h.
* sysdeps/ieee754/flt-32/s_issignalingf_inline.h: New.
* sysdeps/powerpc/fpu/math_private.h: (__issignalingf): New.
Include s_issignalingf_inline.h.
---
v2: Oops. Forgot to include <s_issignalingf_inline.h> in s_signalingf.c
so other arches will continue to work as before.
math/s_canonicalize_template.c | 1 +
math/s_fmax_template.c | 1 +
math/s_fmaxmag_template.c | 1 +
math/s_fmin_template.c | 2 +-
math/s_fminmag_template.c | 1 +
sysdeps/ieee754/flt-32/s_issignalingf.c | 20 +++--------
sysdeps/ieee754/flt-32/s_issignalingf_inline.h | 47 ++++++++++++++++++++++++++
sysdeps/powerpc/fpu/math_private.h | 3 ++
8 files changed, 59 insertions(+), 17 deletions(-)
create mode 100644 sysdeps/ieee754/flt-32/s_issignalingf_inline.h
Comments
On Fri, 27 Sep 2019, Paul A. Clarke wrote:
> From: "Paul A. Clarke" <pc@us.ibm.com>
>
> issignalingf is a very small function used in some areas where
> better performance (and smaller code) might be helpful.
>
> Establish a means to inline issignalingf, and use that for powerpc.
Please don't add any math_private.h includes for optimizing function
calls, or any optimized definitions of math.h functions that only live in
math_private.h. (Optimized definitions of functions that are only
declared in math_private.h, such as __ieee754_* functions, are OK.)
See my commit c52944e8ccb15158b7e44cbb75fb46d81400d75c. Where code uses
anything from math_private.h, we want it to be something *only* in
math_private.h, so that a build failure results if a math_private.h
include is missing, rather than silently having less optimized code.
I.e. any such optimized inlines belong in include/math.h like the one for
__isinff128 for GCC before GCC 7.
@@ -17,6 +17,7 @@
<https://www.gnu.org/licenses/>. */
#include <math.h>
+#include <math_private.h>
int
M_DECL_FUNC (__canonicalize) (FLOAT *cx, const FLOAT *x)
@@ -18,6 +18,7 @@
<https://www.gnu.org/licenses/>. */
#include <math.h>
+#include <math_private.h>
FLOAT
M_DECL_FUNC (__fmax) (FLOAT x, FLOAT y)
@@ -17,6 +17,7 @@
<https://www.gnu.org/licenses/>. */
#include <math.h>
+#include <math_private.h>
FLOAT
M_DECL_FUNC (__fmaxmag) (FLOAT x, FLOAT y)
@@ -18,7 +18,7 @@
<https://www.gnu.org/licenses/>. */
#include <math.h>
-
+#include <math_private.h>
FLOAT
M_DECL_FUNC (__fmin) (FLOAT x, FLOAT y)
@@ -17,6 +17,7 @@
<https://www.gnu.org/licenses/>. */
#include <math.h>
+#include <math_private.h>
FLOAT
M_DECL_FUNC (__fminmag) (FLOAT x, FLOAT y)
@@ -19,25 +19,13 @@
#include <math.h>
#include <math_private.h>
#include <nan-high-order-bit.h>
+#include <s_issignalingf_inline.h>
+
+#undef __issignalingf
int
__issignalingf (float x)
{
- uint32_t xi;
- GET_FLOAT_WORD (xi, x);
-#if HIGH_ORDER_BIT_IS_SET_FOR_SNAN
- /* We only have to care about the high-order bit of x's significand, because
- having it set (sNaN) already makes the significand different from that
- used to designate infinity. */
- return (xi & 0x7fc00000) == 0x7fc00000;
-#else
- /* To keep the following comparison simple, toggle the quiet/signaling bit,
- so that it is set for sNaNs. This is inverse to IEEE 754-2008 (as well as
- common practice for IEEE 754-1985). */
- xi ^= 0x00400000;
- /* We have to compare for greater (instead of greater or equal), because x's
- significand being all-zero designates infinity not NaN. */
- return (xi & 0x7fffffff) > 0x7fc00000;
-#endif
+ return __issignalingf_inline (x);
}
libm_hidden_def (__issignalingf)
new file mode 100644
@@ -0,0 +1,47 @@
+/* Test for signaling NaN.
+ Copyright (C) 2013-2019 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
+ <https://www.gnu.org/licenses/>. */
+
+#ifndef _S_ISSIGNALING_INLINE_H
+#define _S_ISSIGNALING_INLINE_H
+
+#include <math.h>
+#include <math_private.h>
+#include <nan-high-order-bit.h>
+
+extern __always_inline int
+__issignalingf_inline (float x)
+{
+ uint32_t xi;
+ GET_FLOAT_WORD (xi, x);
+#if HIGH_ORDER_BIT_IS_SET_FOR_SNAN
+ /* We only have to care about the high-order bit of x's significand, because
+ having it set (sNaN) already makes the significand different from that
+ used to designate infinity. */
+ return (xi & 0x7fc00000) == 0x7fc00000;
+#else
+ /* To keep the following comparison simple, toggle the quiet/signaling bit,
+ so that it is set for sNaNs. This is inverse to IEEE 754-2008 (as well as
+ common practice for IEEE 754-1985). */
+ xi ^= 0x00400000;
+ /* We have to compare for greater (instead of greater or equal), because x's
+ significand being all-zero designates infinity not NaN. */
+ return (xi & 0x7fffffff) > 0x7fc00000;
+#endif
+}
+
+#endif
@@ -35,4 +35,7 @@ __ieee754_sqrtf128 (_Float128 __x)
}
#endif
+#define __issignalingf __issignalingf_inline
+#include <s_issignalingf_inline.h>
+
#endif /* _PPC_MATH_PRIVATE_H_ */