AuroraMiddleware/glibc
1
0
Fork 0
mirror of https://sourceware.org/git/glibc.git synced 2024-12-25 20:21:07 +00:00
Code Issues Packages Projects Releases Wiki Activity
9c21573c02
glibc/sysdeps/generic/math_private.h
Siddhesh Poyarekar 2506109403 Set/restore rounding mode only when needed 
The most common use case of math functions is with default rounding
mode, i.e. rounding to nearest.  Setting and restoring rounding mode
is an unnecessary overhead for this, so I've added support for a
context, which does the set/restore only if the FP status needs a
change.  The code is written such that only x86 uses these.  Other
architectures should be unaffected by it, but would definitely benefit
if the set/restore has as much overhead relative to the rest of the
code, as the x86 bits do.

Here's a summary of the performance improvement due to these
improvements; I've only mentioned functions that use the set/restore
and have benchmark inputs for x86_64:

Before:

cos(): ITERS:4.69335e+08: TOTAL:28884.6Mcy, MAX:4080.28cy, MIN:57.562cy, 16248.6 calls/Mcy
exp(): ITERS:4.47604e+08: TOTAL:28796.2Mcy, MAX:207.721cy, MIN:62.385cy, 15543.9 calls/Mcy
pow(): ITERS:1.63485e+08: TOTAL:28879.9Mcy, MAX:362.255cy, MIN:172.469cy, 5660.86 calls/Mcy
sin(): ITERS:3.89578e+08: TOTAL:28900Mcy, MAX:704.859cy, MIN:47.583cy, 13480.2 calls/Mcy
tan(): ITERS:7.0971e+07: TOTAL:28902.2Mcy, MAX:1357.79cy, MIN:388.58cy, 2455.55 calls/Mcy

After:

cos(): ITERS:6.0014e+08: TOTAL:28875.9Mcy, MAX:364.283cy, MIN:45.716cy, 20783.4 calls/Mcy
exp(): ITERS:5.48578e+08: TOTAL:28764.9Mcy, MAX:191.617cy, MIN:51.011cy, 19071.1 calls/Mcy
pow(): ITERS:1.70013e+08: TOTAL:28873.6Mcy, MAX:689.522cy, MIN:163.989cy, 5888.18 calls/Mcy
sin(): ITERS:4.64079e+08: TOTAL:28891.5Mcy, MAX:6959.3cy, MIN:36.189cy, 16062.8 calls/Mcy
tan(): ITERS:7.2354e+07: TOTAL:28898.9Mcy, MAX:1295.57cy, MIN:380.698cy, 2503.7 calls/Mcy

So the improvements are:

cos: 27.9089%
exp: 22.6919%
pow: 4.01564%
sin: 19.1585%
tan: 1.96086%

The downside of the change is that it will have an adverse performance
impact on non-default rounding modes, but I think the tradeoff is
justified.
2013-06-12 10:36:48 +05:30

621 lines
20 KiB
C
Raw Blame History

/*
* ====================================================
* Copyright (C) 1993 by Sun Microsystems, Inc. All rights reserved.
*
* Developed at SunPro, a Sun Microsystems, Inc. business.
* Permission to use, copy, modify, and distribute this
* software is freely granted, provided that this notice
* is preserved.
* ====================================================
*/
/*
* from: @(#)fdlibm.h 5.1 93/09/24
*/
#ifndef _MATH_PRIVATE_H_
#define _MATH_PRIVATE_H_
#include <endian.h>
#include <stdint.h>
#include <sys/types.h>
#include <fenv.h>
/* The original fdlibm code used statements like:
n0 = ((*(int*)&one)>>29)^1; * index of high word *
ix0 = *(n0+(int*)&x); * high word of x *
ix1 = *((1-n0)+(int*)&x); * low word of x *
to dig two 32 bit words out of the 64 bit IEEE floating point
value. That is non-ANSI, and, moreover, the gcc instruction
scheduler gets it wrong. We instead use the following macros.
Unlike the original code, we determine the endianness at compile
time, not at run time; I don't see much benefit to selecting
endianness at run time. */
/* A union which permits us to convert between a double and two 32 bit
ints. */
#if __FLOAT_WORD_ORDER == BIG_ENDIAN
typedef union
{
double value;
struct
{
u_int32_t msw;
u_int32_t lsw;
} parts;
uint64_t word;
} ieee_double_shape_type;
#endif
#if __FLOAT_WORD_ORDER == LITTLE_ENDIAN
typedef union
{
double value;
struct
{
u_int32_t lsw;
u_int32_t msw;
} parts;
uint64_t word;
} ieee_double_shape_type;
#endif
/* Get two 32 bit ints from a double. */
#define EXTRACT_WORDS(ix0,ix1,d) \
do { \
ieee_double_shape_type ew_u; \
ew_u.value = (d); \
(ix0) = ew_u.parts.msw; \
(ix1) = ew_u.parts.lsw; \
} while (0)
/* Get the more significant 32 bit int from a double. */
#ifndef GET_HIGH_WORD
# define GET_HIGH_WORD(i,d) \
do { \
ieee_double_shape_type gh_u; \
gh_u.value = (d); \
(i) = gh_u.parts.msw; \
} while (0)
#endif
/* Get the less significant 32 bit int from a double. */
#ifndef GET_LOW_WORD
# define GET_LOW_WORD(i,d) \
do { \
ieee_double_shape_type gl_u; \
gl_u.value = (d); \
(i) = gl_u.parts.lsw; \
} while (0)
#endif
/* Get all in one, efficient on 64-bit machines. */
#ifndef EXTRACT_WORDS64
# define EXTRACT_WORDS64(i,d) \
do { \
ieee_double_shape_type gh_u; \
gh_u.value = (d); \
(i) = gh_u.word; \
} while (0)
#endif
/* Set a double from two 32 bit ints. */
#ifndef INSERT_WORDS
# define INSERT_WORDS(d,ix0,ix1) \
do { \
ieee_double_shape_type iw_u; \
iw_u.parts.msw = (ix0); \
iw_u.parts.lsw = (ix1); \
(d) = iw_u.value; \
} while (0)
#endif
/* Get all in one, efficient on 64-bit machines. */
#ifndef INSERT_WORDS64
# define INSERT_WORDS64(d,i) \
do { \
ieee_double_shape_type iw_u; \
iw_u.word = (i); \
(d) = iw_u.value; \
} while (0)
#endif
/* Set the more significant 32 bits of a double from an int. */
#ifndef SET_HIGH_WORD
#define SET_HIGH_WORD(d,v) \
do { \
ieee_double_shape_type sh_u; \
sh_u.value = (d); \
sh_u.parts.msw = (v); \
(d) = sh_u.value; \
} while (0)
#endif
/* Set the less significant 32 bits of a double from an int. */
#ifndef SET_LOW_WORD
# define SET_LOW_WORD(d,v) \
do { \
ieee_double_shape_type sl_u; \
sl_u.value = (d); \
sl_u.parts.lsw = (v); \
(d) = sl_u.value; \
} while (0)
#endif
/* A union which permits us to convert between a float and a 32 bit
int. */
typedef union
{
float value;
u_int32_t word;
} ieee_float_shape_type;
/* Get a 32 bit int from a float. */
#ifndef GET_FLOAT_WORD
# define GET_FLOAT_WORD(i,d) \
do { \
ieee_float_shape_type gf_u; \
gf_u.value = (d); \
(i) = gf_u.word; \
} while (0)
#endif
/* Set a float from a 32 bit int. */
#ifndef SET_FLOAT_WORD
# define SET_FLOAT_WORD(d,i) \
do { \
ieee_float_shape_type sf_u; \
sf_u.word = (i); \
(d) = sf_u.value; \
} while (0)
#endif
/* Get long double macros from a separate header. */
#include <math_ldbl.h>
/* ieee style elementary functions */
extern double __ieee754_sqrt (double);
extern double __ieee754_acos (double);
extern double __ieee754_acosh (double);
extern double __ieee754_log (double);
extern double __ieee754_atanh (double);
extern double __ieee754_asin (double);
extern double __ieee754_atan2 (double,double);
extern double __ieee754_exp (double);
extern double __ieee754_exp2 (double);
extern double __ieee754_exp10 (double);
extern double __ieee754_cosh (double);
extern double __ieee754_fmod (double,double);
extern double __ieee754_pow (double,double);
extern double __ieee754_lgamma_r (double,int *);
extern double __ieee754_gamma_r (double,int *);
extern double __ieee754_lgamma (double);
extern double __ieee754_gamma (double);
extern double __ieee754_log10 (double);
extern double __ieee754_log2 (double);
extern double __ieee754_sinh (double);
extern double __ieee754_hypot (double,double);
extern double __ieee754_j0 (double);
extern double __ieee754_j1 (double);
extern double __ieee754_y0 (double);
extern double __ieee754_y1 (double);
extern double __ieee754_jn (int,double);
extern double __ieee754_yn (int,double);
extern double __ieee754_remainder (double,double);
extern int32_t __ieee754_rem_pio2 (double,double*);
extern double __ieee754_scalb (double,double);
extern int __ieee754_ilogb (double);
/* fdlibm kernel function */
extern double __kernel_standard (double,double,int);
extern float __kernel_standard_f (float,float,int);
extern long double __kernel_standard_l (long double,long double,int);
extern double __kernel_sin (double,double,int);
extern double __kernel_cos (double,double);
extern double __kernel_tan (double,double,int);
extern int __kernel_rem_pio2 (double*,double*,int,int,int, const int32_t*);
/* internal functions. */
extern double __copysign (double x, double __y);
extern inline double __copysign (double x, double y)
{ return __builtin_copysign (x, y); }
/* ieee style elementary float functions */
extern float __ieee754_sqrtf (float);
extern float __ieee754_acosf (float);
extern float __ieee754_acoshf (float);
extern float __ieee754_logf (float);
extern float __ieee754_atanhf (float);
extern float __ieee754_asinf (float);
extern float __ieee754_atan2f (float,float);
extern float __ieee754_expf (float);
extern float __ieee754_exp2f (float);
extern float __ieee754_exp10f (float);
extern float __ieee754_coshf (float);
extern float __ieee754_fmodf (float,float);
extern float __ieee754_powf (float,float);
extern float __ieee754_lgammaf_r (float,int *);
extern float __ieee754_gammaf_r (float,int *);
extern float __ieee754_lgammaf (float);
extern float __ieee754_gammaf (float);
extern float __ieee754_log10f (float);
extern float __ieee754_log2f (float);
extern float __ieee754_sinhf (float);
extern float __ieee754_hypotf (float,float);
extern float __ieee754_j0f (float);
extern float __ieee754_j1f (float);
extern float __ieee754_y0f (float);
extern float __ieee754_y1f (float);
extern float __ieee754_jnf (int,float);
extern float __ieee754_ynf (int,float);
extern float __ieee754_remainderf (float,float);
extern int32_t __ieee754_rem_pio2f (float,float*);
extern float __ieee754_scalbf (float,float);
extern int __ieee754_ilogbf (float);
/* float versions of fdlibm kernel functions */
extern float __kernel_sinf (float,float,int);
extern float __kernel_cosf (float,float);
extern float __kernel_tanf (float,float,int);
extern int __kernel_rem_pio2f (float*,float*,int,int,int, const int32_t*);
/* internal functions. */
extern float __copysignf (float x, float __y);
extern inline float __copysignf (float x, float y)
{ return __builtin_copysignf (x, y); }
/* ieee style elementary long double functions */
extern long double __ieee754_sqrtl (long double);
extern long double __ieee754_acosl (long double);
extern long double __ieee754_acoshl (long double);
extern long double __ieee754_logl (long double);
extern long double __ieee754_atanhl (long double);
extern long double __ieee754_asinl (long double);
extern long double __ieee754_atan2l (long double,long double);
extern long double __ieee754_expl (long double);
extern long double __ieee754_exp2l (long double);
extern long double __ieee754_exp10l (long double);
extern long double __ieee754_coshl (long double);
extern long double __ieee754_fmodl (long double,long double);
extern long double __ieee754_powl (long double,long double);
extern long double __ieee754_lgammal_r (long double,int *);
extern long double __ieee754_gammal_r (long double,int *);
extern long double __ieee754_lgammal (long double);
extern long double __ieee754_gammal (long double);
extern long double __ieee754_log10l (long double);
extern long double __ieee754_log2l (long double);
extern long double __ieee754_sinhl (long double);
extern long double __ieee754_hypotl (long double,long double);
extern long double __ieee754_j0l (long double);
extern long double __ieee754_j1l (long double);
extern long double __ieee754_y0l (long double);
extern long double __ieee754_y1l (long double);
extern long double __ieee754_jnl (int,long double);
extern long double __ieee754_ynl (int,long double);
extern long double __ieee754_remainderl (long double,long double);
extern int __ieee754_rem_pio2l (long double,long double*);
extern long double __ieee754_scalbl (long double,long double);
extern int __ieee754_ilogbl (long double);
/* long double versions of fdlibm kernel functions */
extern long double __kernel_sinl (long double,long double,int);
extern long double __kernel_cosl (long double,long double);
extern long double __kernel_tanl (long double,long double,int);
extern void __kernel_sincosl (long double,long double,
long double *,long double *, int);
extern int __kernel_rem_pio2l (long double*,long double*,int,int,
int,const int*);
#ifndef NO_LONG_DOUBLE
/* prototypes required to compile the ldbl-96 support without warnings */
extern int __finitel (long double);
extern int __ilogbl (long double);
extern int __isinfl (long double);
extern int __isnanl (long double);
extern long double __atanl (long double);
extern long double __copysignl (long double, long double);
extern long double __expm1l (long double);
extern long double __floorl (long double);
extern long double __frexpl (long double, int *);
extern long double __ldexpl (long double, int);
extern long double __log1pl (long double);
extern long double __nanl (const char *);
extern long double __rintl (long double);
extern long double __scalbnl (long double, int);
extern long double __sqrtl (long double x);
extern long double fabsl (long double x);
extern void __sincosl (long double, long double *, long double *);
extern long double __logbl (long double x);
extern long double __significandl (long double x);
extern inline long double __copysignl (long double x, long double y)
{ return __builtin_copysignl (x, y); }
#endif
/* Prototypes for functions of the IBM Accurate Mathematical Library. */
extern double __exp1 (double __x, double __xx, double __error);
extern double __sin (double __x);
extern double __cos (double __x);
extern int __branred (double __x, double *__a, double *__aa);
extern void __doasin (double __x, double __dx, double __v[]);
extern void __dubsin (double __x, double __dx, double __v[]);
extern void __dubcos (double __x, double __dx, double __v[]);
extern double __halfulp (double __x, double __y);
extern double __sin32 (double __x, double __res, double __res1);
extern double __cos32 (double __x, double __res, double __res1);
extern double __mpsin (double __x, double __dx);
extern double __mpcos (double __x, double __dx);
extern double __mpsin1 (double __x);
extern double __mpcos1 (double __x);
extern double __slowexp (double __x);
extern double __slowpow (double __x, double __y, double __z);
extern void __docos (double __x, double __dx, double __v[]);
/* Return X^2 + Y^2 - 1, computed without large cancellation error.
It is given that 1 > X >= Y >= epsilon / 2, and that either X >=
0.75 or Y >= 0.5. */
extern float __x2y2m1f (float x, float y);
extern double __x2y2m1 (double x, double y);
extern long double __x2y2m1l (long double x, long double y);
/* Compute the product of X + X_EPS, X + X_EPS + 1, ..., X + X_EPS + N
- 1, in the form R * (1 + *EPS) where the return value R is an
approximation to the product and *EPS is set to indicate the
approximate error in the return value. X is such that all the
values X + 1, ..., X + N - 1 are exactly representable, and X_EPS /
X is small enough that factors quadratic in it can be
neglected. */
extern float __gamma_productf (float x, float x_eps, int n, float *eps);
extern double __gamma_product (double x, double x_eps, int n, double *eps);
extern long double __gamma_productl (long double x, long double x_eps,
int n, long double *eps);
#ifndef math_opt_barrier
# define math_opt_barrier(x) \
({ __typeof (x) __x = (x); __asm ("" : "+m" (__x)); __x; })
# define math_force_eval(x) \
({ __typeof (x) __x = (x); __asm __volatile__ ("" : : "m" (__x)); })
#endif
/* The standards only specify one variant of the fenv.h interfaces.
But at least for some architectures we can be more efficient if we
know what operations are going to be performed. Therefore we
define additional interfaces. By default they refer to the normal
interfaces. */
static __always_inline void
default_libc_feholdexcept (fenv_t *e)
{
(void) feholdexcept (e);
}
#ifndef libc_feholdexcept
# define libc_feholdexcept default_libc_feholdexcept
#endif
#ifndef libc_feholdexceptf
# define libc_feholdexceptf default_libc_feholdexcept
#endif
#ifndef libc_feholdexceptl
# define libc_feholdexceptl default_libc_feholdexcept
#endif
static __always_inline void
default_libc_fesetround (int r)
{
(void) fesetround (r);
}
#ifndef libc_fesetround
# define libc_fesetround default_libc_fesetround
#endif
#ifndef libc_fesetroundf
# define libc_fesetroundf default_libc_fesetround
#endif
#ifndef libc_fesetroundl
# define libc_fesetroundl default_libc_fesetround
#endif
static __always_inline void
default_libc_feholdexcept_setround (fenv_t *e, int r)
{
feholdexcept (e);
fesetround (r);
}
#ifndef libc_feholdexcept_setround
# define libc_feholdexcept_setround default_libc_feholdexcept_setround
#endif
#ifndef libc_feholdexcept_setroundf
# define libc_feholdexcept_setroundf default_libc_feholdexcept_setround
#endif
#ifndef libc_feholdexcept_setroundl
# define libc_feholdexcept_setroundl default_libc_feholdexcept_setround
#endif
#ifndef libc_feholdsetround_53bit
# define libc_feholdsetround_53bit libc_feholdsetround
#endif
#ifndef libc_fetestexcept
# define libc_fetestexcept fetestexcept
#endif
#ifndef libc_fetestexceptf
# define libc_fetestexceptf fetestexcept
#endif
#ifndef libc_fetestexceptl
# define libc_fetestexceptl fetestexcept
#endif
static __always_inline void
default_libc_fesetenv (fenv_t *e)
{
(void) fesetenv (e);
}
#ifndef libc_fesetenv
# define libc_fesetenv default_libc_fesetenv
#endif
#ifndef libc_fesetenvf
# define libc_fesetenvf default_libc_fesetenv
#endif
#ifndef libc_fesetenvl
# define libc_fesetenvl default_libc_fesetenv
#endif
static __always_inline void
default_libc_feupdateenv (fenv_t *e)
{
(void) feupdateenv (e);
}
#ifndef libc_feupdateenv
# define libc_feupdateenv default_libc_feupdateenv
#endif
#ifndef libc_feupdateenvf
# define libc_feupdateenvf default_libc_feupdateenv
#endif
#ifndef libc_feupdateenvl
# define libc_feupdateenvl default_libc_feupdateenv
#endif
#ifndef libc_feresetround_53bit
# define libc_feresetround_53bit libc_feresetround
#endif
static __always_inline int
default_libc_feupdateenv_test (fenv_t *e, int ex)
{
int ret = fetestexcept (ex);
feupdateenv (e);
return ret;
}
#ifndef libc_feupdateenv_test
# define libc_feupdateenv_test default_libc_feupdateenv_test
#endif
#ifndef libc_feupdateenv_testf
# define libc_feupdateenv_testf default_libc_feupdateenv_test
#endif
#ifndef libc_feupdateenv_testl
# define libc_feupdateenv_testl default_libc_feupdateenv_test
#endif
/* Save and set the rounding mode. The use of fenv_t to store the old mode
allows a target-specific version of this function to avoid converting the
rounding mode from the fpu format. By default we have no choice but to
manipulate the entire env. */
#ifndef libc_feholdsetround
# define libc_feholdsetround libc_feholdexcept_setround
#endif
#ifndef libc_feholdsetroundf
# define libc_feholdsetroundf libc_feholdexcept_setroundf
#endif
#ifndef libc_feholdsetroundl
# define libc_feholdsetroundl libc_feholdexcept_setroundl
#endif
/* ... and the reverse. */
#ifndef libc_feresetround
# define libc_feresetround libc_feupdateenv
#endif
#ifndef libc_feresetroundf
# define libc_feresetroundf libc_feupdateenvf
#endif
#ifndef libc_feresetroundl
# define libc_feresetroundl libc_feupdateenvl
#endif
/* ... and a version that may also discard exceptions. */
#ifndef libc_feresetround_noex
# define libc_feresetround_noex libc_fesetenv
#endif
#ifndef libc_feresetround_noexf
# define libc_feresetround_noexf libc_fesetenvf
#endif
#ifndef libc_feresetround_noexl
# define libc_feresetround_noexl libc_fesetenvl
#endif
#if HAVE_RM_CTX
/* Set/Restore Rounding Modes only when necessary. If defined, these functions
set/restore floating point state only if the state needed within the lexical
block is different from the current state. This saves a lot of time when
the floating point unit is much slower than the fixed point units. */
# ifndef libc_feresetround_noex_ctx
# define libc_feresetround_noex_ctx libc_fesetenv_ctx
# endif
# ifndef libc_feresetround_noexf_ctx
# define libc_feresetround_noexf_ctx libc_fesetenvf_ctx
# endif
# ifndef libc_feresetround_noexl_ctx
# define libc_feresetround_noexl_ctx libc_fesetenvl_ctx
# endif
# ifndef libc_feholdsetround_53bit_ctx
# define libc_feholdsetround_53bit_ctx libc_feholdsetround_ctx
# endif
# ifndef libc_feresetround_53bit_ctx
# define libc_feresetround_53bit_ctx libc_feresetround_ctx
# endif
# define SET_RESTORE_ROUND_GENERIC(RM,ROUNDFUNC,CLEANUPFUNC) \
struct rm_ctx ctx __attribute__((cleanup(CLEANUPFUNC ## _ctx))); \
ROUNDFUNC ## _ctx (&ctx, (RM))
#else
# define SET_RESTORE_ROUND_GENERIC(RM, ROUNDFUNC, CLEANUPFUNC) \
fenv_t __libc_save_rm __attribute__((cleanup(CLEANUPFUNC))); \
ROUNDFUNC (&__libc_save_rm, (RM))
#endif
/* Save and restore the rounding mode within a lexical block. */
#define SET_RESTORE_ROUND(RM) \
SET_RESTORE_ROUND_GENERIC (RM, libc_feholdsetround, libc_feresetround)
#define SET_RESTORE_ROUNDF(RM) \
SET_RESTORE_ROUND_GENERIC (RM, libc_feholdsetroundf, libc_feresetroundf)
#define SET_RESTORE_ROUNDL(RM) \
SET_RESTORE_ROUND_GENERIC (RM, libc_feholdsetroundl, libc_feresetroundl)
/* Save and restore the rounding mode within a lexical block, and also
the set of exceptions raised within the block may be discarded. */
#define SET_RESTORE_ROUND_NOEX(RM) \
SET_RESTORE_ROUND_GENERIC (RM, libc_feholdsetround, libc_feresetround_noex)
#define SET_RESTORE_ROUND_NOEXF(RM) \
SET_RESTORE_ROUND_GENERIC (RM, libc_feholdsetroundf, libc_feresetround_noexf)
#define SET_RESTORE_ROUND_NOEXL(RM) \
SET_RESTORE_ROUND_GENERIC (RM, libc_feholdsetroundl, libc_feresetround_noexl)
/* Like SET_RESTORE_ROUND, but also set rounding precision to 53 bits. */
#define SET_RESTORE_ROUND_53BIT(RM) \
SET_RESTORE_ROUND_GENERIC (RM, libc_feholdsetround_53bit, \
libc_feresetround_53bit)
#define __nan(str) \
(__builtin_constant_p (str) && str[0] == '\0' ? NAN : __nan (str))
#define __nanf(str) \
(__builtin_constant_p (str) && str[0] == '\0' ? NAN : __nan (str))
#define __nanl(str) \
(__builtin_constant_p (str) && str[0] == '\0' ? NAN : __nan (str))
#endif /* _MATH_PRIVATE_H_ */
Reference in New Issue View Git Blame Copy Permalink