diff options
author | Roland McGrath <roland@gnu.org> | 1996-03-17 01:58:17 +0000 |
---|---|---|
committer | Roland McGrath <roland@gnu.org> | 1996-03-17 01:58:17 +0000 |
commit | 6025c399e9fd73efeb5b93db78ec021d3fce2b1c (patch) | |
tree | e213f46f05f3cb8638d22a81d437a32754062173 /sysdeps/m68k/fpu/__math.h | |
parent | d3b7d2ac93ed6fce3f7cb8e567d2b96719dde01a (diff) | |
download | glibc-6025c399e9fd73efeb5b93db78ec021d3fce2b1c.tar.gz |
Sat Mar 16 20:08:22 1996 David Mosberger-Tang <davidm@azstarnet.com>
* sysdeps/alpha/memchr.S: new file.
* sysdeps/alpha/memchr.c: obsolete file removed.
Sat Mar 16 16:26:09 1996 Roland McGrath <roland@charlie-brown.gnu.ai.mit.edu>
* misc/Makefile (headers): Add sysexits.h.
* misc/sysexits.h: New file.
Thu Mar 14 15:20:45 1996 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>
* sysdeps/libm-ieee754/e_atan2.c (__ieee754_atan2): Change atan
call to __atan.
* sysdeps/libm-ieee754/e_atan2f.c (__ieee754_atan2f): Change atanf
call to __atanf.
* sysdeps/m68k/fpu/e_acos.c, sysdeps/m68k/fpu/e_acosf.c,
sysdeps/m68k/fpu/e_asin.c, sysdeps/m68k/fpu/e_asinf.c,
sysdeps/m68k/fpu/e_atanh.c, sysdeps/m68k/fpu/e_atanhf.c,
sysdeps/m68k/fpu/e_cosh.c, sysdeps/m68k/fpu/e_coshf.c,
sysdeps/m68k/fpu/e_exp.c, sysdeps/m68k/fpu/e_expf.c,
sysdeps/m68k/fpu/e_fmod.c, sysdeps/m68k/fpu/e_fmodf.c,
sysdeps/m68k/fpu/e_log.c, sysdeps/m68k/fpu/e_log10.c,
sysdeps/m68k/fpu/e_log10f.c, sysdeps/m68k/fpu/e_logf.c,
sysdeps/m68k/fpu/e_pow.c, sysdeps/m68k/fpu/e_powf.c,
sysdeps/m68k/fpu/e_remainder.c, sysdeps/m68k/fpu/e_remainderf.c,
sysdeps/m68k/fpu/e_scalb.c, sysdeps/m68k/fpu/e_scalbf.c,
sysdeps/m68k/fpu/e_sinh.c, sysdeps/m68k/fpu/e_sinhf.c,
sysdeps/m68k/fpu/e_sqrt.c, sysdeps/m68k/fpu/e_sqrtf.c,
sysdeps/m68k/fpu/k_cos.c, sysdeps/m68k/fpu/k_cosf.c,
sysdeps/m68k/fpu/k_sin.c, sysdeps/m68k/fpu/k_sinf.c,
sysdeps/m68k/fpu/k_tan.c, sysdeps/m68k/fpu/k_tanf.c,
sysdeps/m68k/fpu/s_atan.c, sysdeps/m68k/fpu/s_atanf.c,
sysdeps/m68k/fpu/s_ceil.c, sysdeps/m68k/fpu/s_ceilf.c,
sysdeps/m68k/fpu/s_cos.c, sysdeps/m68k/fpu/s_cosf.c,
sysdeps/m68k/fpu/s_expm1.c, sysdeps/m68k/fpu/s_expm1f.c,
sysdeps/m68k/fpu/s_fabs.c, sysdeps/m68k/fpu/s_fabsf.c,
sysdeps/m68k/fpu/s_finite.c, sysdeps/m68k/fpu/s_finitef.c,
sysdeps/m68k/fpu/s_floor.c, sysdeps/m68k/fpu/s_floorf.c,
sysdeps/m68k/fpu/s_frexp.c, sysdeps/m68k/fpu/s_frexpf.c,
sysdeps/m68k/fpu/s_ilogb.c, sysdeps/m68k/fpu/s_ilogbf.c,
sysdeps/m68k/fpu/s_isinf.c, sysdeps/m68k/fpu/s_isinff.c,
sysdeps/m68k/fpu/s_isnan.c, sysdeps/m68k/fpu/s_isnanf.c,
sysdeps/m68k/fpu/s_ldexp.c, sysdeps/m68k/fpu/s_ldexpf.c,
sysdeps/m68k/fpu/s_log1p.c, sysdeps/m68k/fpu/s_log1pf.c,
sysdeps/m68k/fpu/s_logb.c, sysdeps/m68k/fpu/s_logbf.c,
sysdeps/m68k/fpu/s_modf.c, sysdeps/m68k/fpu/s_modff.c,
sysdeps/m68k/fpu/s_rint.c, sysdeps/m68k/fpu/s_rintf.c,
sysdeps/m68k/fpu/s_scalbn.c, sysdeps/m68k/fpu/s_scalbnf.c,
sysdeps/m68k/fpu/s_significand.c,
sysdeps/m68k/fpu/s_significandf.c, sysdeps/m68k/fpu/s_sin.c,
sysdeps/m68k/fpu/s_sinf.c, sysdeps/m68k/fpu/s_tan.c,
sysdeps/m68k/fpu/s_tanf.c, sysdeps/m68k/fpu/s_tanh.c,
sysdeps/m68k/fpu/s_tanhf.c: New files, for m68881 port of fdlibm.
* sysdeps/m68k/fpu/__math.h: Rewritten for fdlibm.
* sysdeps/m68k/fpu/isinfl.c: Rewritten to get argument type right.
* sysdeps/m68k/fpu/isnanl.c: Likewise.
Thu Mar 14 06:01:07 1996 Roland McGrath <roland@charlie-brown.gnu.ai.mit.edu>
* posix/glob.c (glob): In GLOB_BRACE brace expansion, fix buffer size
calculation to include trailing invariant portion. Don't use alloca;
instead use a dynamic auto array for GCC, malloc for non-GCC.
Handle nested braces properly.
* elf/elf.h (Elf32_auxv_t): Specify prototype (void) for `a_un.a_fcn'.
* libc-symbols.h (lint): New macro.
Fri Mar 15 01:18:00 1996 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>
* libio/iosetvbuf.c: Add weak alias setvbuf.
Diffstat (limited to 'sysdeps/m68k/fpu/__math.h')
-rw-r--r-- | sysdeps/m68k/fpu/__math.h | 364 |
1 files changed, 280 insertions, 84 deletions
diff --git a/sysdeps/m68k/fpu/__math.h b/sysdeps/m68k/fpu/__math.h index e357364fb7..f59c168e79 100644 --- a/sysdeps/m68k/fpu/__math.h +++ b/sysdeps/m68k/fpu/__math.h @@ -19,8 +19,6 @@ Cambridge, MA 02139, USA. */ #ifdef __GNUC__ #include <sys/cdefs.h> -#define __need_Emath -#include <errno.h> #ifdef __NO_MATH_INLINES /* This is used when defining the functions themselves. Define them with @@ -34,59 +32,87 @@ Cambridge, MA 02139, USA. */ #define __MATH_INLINES 1 #endif -#define __inline_mathop2(func, op) \ - __m81_inline double \ - __m81_u(func)(double __mathop_x) __attribute__((__const__)); \ - __m81_inline double \ - __m81_u(func)(double __mathop_x) \ +/* Define a const math function. */ +#define __m81_defun(rettype, func, args) \ + __m81_inline rettype \ + __m81_u(func) args __attribute__((__const__)); \ + __m81_inline rettype \ + __m81_u(func) args + +#define __inline_mathop(func, op) \ + __m81_defun (double, func, (double __mathop_x)) \ { \ double __result; \ __asm("f" __STRING(op) "%.x %1, %0" : "=f" (__result) : "f" (__mathop_x));\ return __result; \ } -#define __inline_mathop(op) __inline_mathop2(op, op) - -__inline_mathop(acos) -__inline_mathop(asin) -__inline_mathop(atan) -__inline_mathop(cos) -__inline_mathop(sin) -__inline_mathop(tan) -__inline_mathop(cosh) -__inline_mathop(sinh) -__inline_mathop(tanh) -__inline_mathop2(exp, etox) -__inline_mathop2(fabs, abs) -__inline_mathop(log10) -__inline_mathop2(log, logn) -__inline_mathop(sqrt) - -__inline_mathop2(__rint, int) -__inline_mathop2(__expm1, etoxm1) - -#ifdef __USE_MISC -#ifndef __NO_MATH_INLINES -__inline_mathop2(rint, int) -__inline_mathop2(expm1, etoxm1) -#endif -__inline_mathop2(log1p, lognp1) -__inline_mathop(atanh) -#endif -__m81_inline double -__m81_u(__drem)(double __x, double __y) __attribute__ ((__const__)); -__m81_inline double -__m81_u(__drem)(double __x, double __y) +#define __inline_mathopf(func, op) \ + __m81_defun (float, func, (float __mathop_x)) \ + { \ + float __result; \ + __asm("f" __STRING(op) "%.x %1, %0" : "=f" (__result) : "f" (__mathop_x));\ + return __result; \ + } + +/* ieee style elementary functions */ +__inline_mathop(__ieee754_acos, acos) +__inline_mathop(__ieee754_asin, asin) +__inline_mathop(__ieee754_cosh, cosh) +__inline_mathop(__ieee754_sinh, sinh) +__inline_mathop(__ieee754_exp, etox) +__inline_mathop(__ieee754_log10, log10) +__inline_mathop(__ieee754_log, logn) +__inline_mathop(__ieee754_sqrt, sqrt) +__inline_mathop(__ieee754_atanh, atanh) + +/* ieee style elementary float functions */ +__inline_mathopf(__ieee754_acosf, acos) +__inline_mathopf(__ieee754_asinf, asin) +__inline_mathopf(__ieee754_coshf, cosh) +__inline_mathopf(__ieee754_sinhf, sinh) +__inline_mathopf(__ieee754_expf, etox) +__inline_mathopf(__ieee754_log10f, log10) +__inline_mathopf(__ieee754_logf, logn) +__inline_mathopf(__ieee754_sqrtf, sqrt) +__inline_mathopf(__ieee754_atanhf, atan) + +__inline_mathop(__atan, atan) +__inline_mathop(__cos, cos) +__inline_mathop(__sin, sin) +__inline_mathop(__tan, tan) +__inline_mathop(__tanh, tanh) +__inline_mathop(__fabs, abs) +__inline_mathop(__sqrt, sqrt) + +__inline_mathop(__rint, int) +__inline_mathop(__expm1, etoxm1) +__inline_mathop(__log1p, lognp1) +__inline_mathop(__logb, log2) +__inline_mathop(__significand, getman) + +__inline_mathopf(__atanf, atan) +__inline_mathopf(__cosf, cos) +__inline_mathopf(__sinf, sin) +__inline_mathopf(__tanf, tan) +__inline_mathopf(__tanhf, tanh) +__inline_mathopf(__fabsf, abs) +__inline_mathopf(__sqrtf, sqrt) + +__inline_mathopf(__rintf, int) +__inline_mathopf(__expm1f, etoxm1) +__inline_mathopf(__log1pf, lognp1) +__inline_mathopf(__logbf, log2) +__inline_mathopf(__significandf, getman) + +__m81_defun (double, __ieee754_remainder, (double __x, double __y)) { double __result; __asm("frem%.x %1, %0" : "=f" (__result) : "f" (__y), "0" (__x)); return __result; } -__m81_inline double -__m81_u(ldexp)(double __x, int __e) __attribute__ ((__const__)); -__m81_inline double -__m81_u(ldexp)(double __x, int __e) +__m81_defun (double, __ldexp, (double __x, int __e)) { double __result; double __double_e = (double) __e; @@ -94,10 +120,7 @@ __m81_u(ldexp)(double __x, int __e) return __result; } -__m81_inline double -__m81_u(fmod)(double __x, double __y) __attribute__ ((__const__)); -__m81_inline double -__m81_u(fmod)(double __x, double __y) +__m81_defun (double, __ieee754_fmod, (double __x, double __y)) { double __result; __asm("fmod%.x %1, %0" : "=f" (__result) : "f" (__y), "0" (__x)); @@ -105,7 +128,7 @@ __m81_u(fmod)(double __x, double __y) } __m81_inline double -__m81_u(frexp)(double __value, int *__expptr) +__m81_u(__frexp)(double __value, int *__expptr) { double __mantissa, __exponent; __asm("fgetexp%.x %1, %0" : "=f" (__exponent) : "f" (__value)); @@ -114,10 +137,7 @@ __m81_u(frexp)(double __value, int *__expptr) return __mantissa; } -__m81_inline double -__m81_u(floor)(double __x) __attribute__ ((__const__)); -__m81_inline double -__m81_u(floor)(double __x) +__m81_defun (double, __floor, (double __x)) { double __result; unsigned long int __ctrl_reg; @@ -133,16 +153,13 @@ __m81_u(floor)(double __x) return __result; } -__m81_inline double -__m81_u(pow)(double __x, double __y) __attribute__ ((__const__)); -__m81_inline double -__m81_u(pow)(double __x, double __y) +__m81_defun (double, __ieee754_pow, (double __x, double __y)) { double __result; if (__x == 0.0) { if (__y <= 0.0) - __result = __infnan (EDOM); + __result = 0.0 / 0.0; else __result = 0.0; } @@ -162,22 +179,19 @@ __m81_u(pow)(double __x, double __y) if (__y == __temp) { int i = (int) __y; - __result = __m81_u (exp) (__y * __m81_u (log) (-__x)); + __result = __m81_u(__ieee754_exp)(__y * __m81_u(__ieee754_log)(-__x)); if (i & 1) __result = -__result; } else - __result = __infnan (EDOM); + __result = 0.0 / 0.0; } else - __result = __m81_u(exp)(__y * __m81_u(log)(__x)); + __result = __m81_u(__ieee754_exp)(__y * __m81_u(__ieee754_log)(__x)); return __result; } -__m81_inline double -__m81_u(ceil)(double __x) __attribute__ ((__const__)); -__m81_inline double -__m81_u(ceil)(double __x) +__m81_defun (double, __ceil, (double __x)) { double __result; unsigned long int __ctrl_reg; @@ -194,30 +208,25 @@ __m81_u(ceil)(double __x) } __m81_inline double -__m81_u(modf)(double __value, double *__iptr) +__m81_u(__modf)(double __value, double *__iptr) { - double __modf_int = __m81_u(floor)(__value); + double __modf_int; + __asm ("fintrz%.x %1, %0" : "=f" (__modf_int) : "f" (__value)); *__iptr = __modf_int; return __value - __modf_int; } -__m81_inline int -__m81_u(__isinf)(double __value) __attribute__ ((__const__)); -__m81_inline int -__m81_u(__isinf)(double __value) +__m81_defun (int, __isinf, (double __value)) { /* There is no branch-condition for infinity, so we must extract and examine the condition codes manually. */ unsigned long int __fpsr; __asm("ftst%.x %1\n" "fmove%.l %/fpsr, %0" : "=dm" (__fpsr) : "f" (__value)); - return (__fpsr & (2 << (3 * 8))) ? (__value < 0 ? -1 : 1) : 0; + return (__fpsr & (2 << 24)) ? (__fpsr & (8 << 24) ? -1 : 1) : 0; } -__m81_inline int -__m81_u(__isnan)(double __value) __attribute__ ((__const__)); -__m81_inline int -__m81_u(__isnan)(double __value) +__m81_defun (int, __isnan, (double __value)) { char __result; __asm("ftst%.x %1\n" @@ -225,23 +234,210 @@ __m81_u(__isnan)(double __value) return __result; } -__m81_inline int -__m81_u(__isinfl)(long double __value) __attribute__ ((__const__)); -__m81_inline int -__m81_u(__isinfl)(long double __value) +__m81_defun (int, __finite, (double __value)) +{ + /* There is no branch-condition for infinity, so we must extract and + examine the condition codes manually. */ + unsigned long int __fpsr; + __asm ("ftst%.x %1\n" + "fmove%.l %/fpsr, %0" : "=dm" (__fpsr) : "f" (__value)); + return (__fpsr & (3 << 24)) == 0; +} + +__m81_defun (int, __ilogb, (double __x)) +{ + double __result; + __asm("fgetexp%.x %1, %0" : "=f" (__result) : "f" (__x)); + return (int) __result; +} + +__m81_defun (double, __ieee754_scalb, (double __x, double __n)) +{ + double __result; + __asm ("fscale%.x %1, %0" : "=f" (__result) : "f" (__n), "0" (__x)); + return __result; +} + +__m81_defun (double, __scalbn, (double __x, int __n)) +{ + double __result; + double __double_n = (double) __n; + __asm ("fscale%.x %1, %0" : "=f" (__result) : "f" (__double_n), "0" (__x)); + return __result; +} + +__m81_defun (float, __ieee754_remainderf, (float __x, float __y)) +{ + float __result; + __asm("frem%.x %1, %0" : "=f" (__result) : "f" (__y), "0" (__x)); + return __result; +} + +__m81_defun (float, __ldexpf, (float __x, int __e)) +{ + float __result; + float __float_e = (float) __e; + __asm("fscale%.x %1, %0" : "=f" (__result) : "f" (__float_e), "0" (__x)); + return __result; +} + +__m81_defun (float, __ieee754_fmodf, (float __x, float __y)) +{ + float __result; + __asm("fmod%.x %1, %0" : "=f" (__result) : "f" (__y), "0" (__x)); + return __result; +} + +__m81_inline float +__m81_u(__frexpf)(float __value, int *__expptr) +{ + float __mantissa, __exponent; + __asm("fgetexp%.x %1, %0" : "=f" (__exponent) : "f" (__value)); + __asm("fgetman%.x %1, %0" : "=f" (__mantissa) : "f" (__value)); + *__expptr = (int) __exponent; + return __mantissa; +} + +__m81_defun (float, __floorf, (float __x)) +{ + float __result; + unsigned long int __ctrl_reg; + __asm __volatile__ ("fmove%.l %!, %0" : "=dm" (__ctrl_reg)); + /* Set rounding towards negative infinity. */ + __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ + : "dmi" ((__ctrl_reg & ~0x10) | 0x20)); + /* Convert X to an integer, using -Inf rounding. */ + __asm __volatile__ ("fint%.x %1, %0" : "=f" (__result) : "f" (__x)); + /* Restore the previous rounding mode. */ + __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ + : "dmi" (__ctrl_reg)); + return __result; +} + +__m81_defun (float, __ieee754_powf, (float __x, float __y)) +{ + float __result; + if (__x == 0.0f) + { + if (__y <= 0.0f) + __result = 0.0f / 0.0f; + else + __result = 0.0f; + } + else if (__y == 0.0f || __x == 1.0f) + __result = 1.0; + else if (__y == 1.0f) + __result = __x; + else if (__y == 2.0f) + __result = __x * __x; + else if (__x == 10.0f) + __asm("ftentox%.x %1, %0" : "=f" (__result) : "f" (__y)); + else if (__x == 2.0f) + __asm("ftwotox%.x %1, %0" : "=f" (__result) : "f" (__y)); + else if (__x < 0.0f) + { + float __temp = __m81_u(__rintf)(__y); + if (__y == __temp) + { + int i = (int) __y; + __result = __m81_u(__ieee754_expf)(__y * __m81_u(__ieee754_logf)(-__x)); + if (i & 1) + __result = -__result; + } + else + __result = 0.0f / 0.0f; + } + else + __result = __m81_u(__ieee754_expf)(__y * __m81_u(__ieee754_logf)(__x)); + return __result; +} + +__m81_defun (float, __ceilf, (float __x)) +{ + float __result; + unsigned long int __ctrl_reg; + __asm __volatile__ ("fmove%.l %!, %0" : "=dm" (__ctrl_reg)); + /* Set rounding towards positive infinity. */ + __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ + : "dmi" (__ctrl_reg | 0x30)); + /* Convert X to an integer, using +Inf rounding. */ + __asm __volatile__ ("fint%.x %1, %0" : "=f" (__result) : "f" (__x)); + /* Restore the previous rounding mode. */ + __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ + : "dmi" (__ctrl_reg)); + return __result; +} + +__m81_inline float +__m81_u(__modff)(float __value, float *__iptr) +{ + float __modf_int; + __asm ("fintrz%.x %1, %0" : "=f" (__modf_int) : "f" (__value)); + *__iptr = __modf_int; + return __value - __modf_int; +} + +__m81_defun (int, __isinff, (float __value)) +{ + /* There is no branch-condition for infinity, + so we must extract and examine the condition codes manually. */ + unsigned long int __fpsr; + __asm("ftst%.x %1\n" + "fmove%.l %/fpsr, %0" : "=dm" (__fpsr) : "f" (__value)); + return (__fpsr & (2 << 24)) ? (__fpsr & (8 << 24) ? -1 : 1) : 0; +} + +__m81_defun (int, __isnanf, (float __value)) +{ + char __result; + __asm("ftst%.x %1\n" + "fsun %0" : "=dm" (__result) : "f" (__value)); + return __result; +} + +__m81_defun (int, __finitef, (float __value)) +{ + /* There is no branch-condition for infinity, so we must extract and + examine the condition codes manually. */ + unsigned long int __fpsr; + __asm ("ftst%.x %1\n" + "fmove%.l %/fpsr, %0" : "=dm" (__fpsr) : "f" (__value)); + return (__fpsr & (3 << 24)) == 0; +} + +__m81_defun (int, __ilogbf, (float __x)) +{ + float __result; + __asm("fgetexp%.x %1, %0" : "=f" (__result) : "f" (__x)); + return (int) __result; +} + +__m81_defun (float, __ieee754_scalbf, (float __x, float __n)) +{ + float __result; + __asm ("fscale%.x %1, %0" : "=f" (__result) : "f" (__n), "0" (__x)); + return __result; +} + +__m81_defun (float, __scalbnf, (float __x, int __n)) +{ + float __result; + float __float_n = (float) __n; + __asm ("fscale%.x %1, %0" : "=f" (__result) : "f" (__float_n), "0" (__x)); + return __result; +} + +__m81_defun (int, __isinfl, (long double __value)) { /* There is no branch-condition for infinity, so we must extract and examine the condition codes manually. */ unsigned long int __fpsr; __asm("ftst%.x %1\n" "fmove%.l %/fpsr, %0" : "=dm" (__fpsr) : "f" (__value)); - return (__fpsr & (2 << (3 * 8))) ? (__value < 0 ? -1 : 1) : 0; + return (__fpsr & (2 << 24)) ? (__fpsr & (8 << 24) ? -1 : 1) : 0; } -__m81_inline int -__m81_u(__isnanl)(long double __value) __attribute__ ((__const__)); -__m81_inline int -__m81_u(__isnanl)(long double __value) +__m81_defun (int, __isnanl, (long double __value)) { char __result; __asm("ftst%.x %1\n" |