mirror of
https://github.com/holub/mame
synced 2025-04-23 00:39:36 +03:00
6957c46998
- Update to Softloat 3 from 2 - FREM and FMOD now generate the quotient bits in FPSR, required by Apple's SANE to do sin/cos/tan properly. - FMOVE of a float to a Dx integer register generates the exception status bits, fixing square roots in SANE - Rewrote how FMOVEM instructions are decoded and executed, fixing issues including skipping too few or too many opcode bytes and causing serious weird behavior. - FPU instructions all now have more realistic cycle timings for a 68881. - All FPU instructions now generate exception bits in FPSR. 3rdparty/softfloat3: Updates [R. Belmont] - Softfloat3 was always being built for a big-endian host, causing incorrect math on LE x64 and AArch64 machines. - Fixed up Softfloat3 to build properly as part of MAME and up-ported the Bochs extensions. In latest Bochs, they were only partially up-ported and Softfloat3 had been hacked up to be more like 2; here they're fixed to work with stock Softfloat3.
410 lines
13 KiB
C
410 lines
13 KiB
C
/*============================================================================
|
|
This source file is an extension to the SoftFloat IEC/IEEE Floating-point
|
|
Arithmetic Package, Release 2b, written for Bochs (x86 achitecture simulator)
|
|
floating point emulation.
|
|
|
|
THIS SOFTWARE IS DISTRIBUTED AS IS, FOR FREE. Although reasonable effort has
|
|
been made to avoid it, THIS SOFTWARE MAY CONTAIN FAULTS THAT WILL AT TIMES
|
|
RESULT IN INCORRECT BEHAVIOR. USE OF THIS SOFTWARE IS RESTRICTED TO PERSONS
|
|
AND ORGANIZATIONS WHO CAN AND WILL TAKE FULL RESPONSIBILITY FOR ALL LOSSES,
|
|
COSTS, OR OTHER PROBLEMS THEY INCUR DUE TO THE SOFTWARE, AND WHO FURTHERMORE
|
|
EFFECTIVELY INDEMNIFY JOHN HAUSER AND THE INTERNATIONAL COMPUTER SCIENCE
|
|
INSTITUTE (possibly via similar legal warning) AGAINST ALL LOSSES, COSTS, OR
|
|
OTHER PROBLEMS INCURRED BY THEIR CUSTOMERS AND CLIENTS DUE TO THE SOFTWARE.
|
|
|
|
Derivative works are acceptable, even for commercial purposes, so long as
|
|
(1) the source code for the derivative work includes prominent notice that
|
|
the work is derivative, and (2) the source code includes prominent notice with
|
|
these four paragraphs for those parts of this code that are retained.
|
|
=============================================================================*/
|
|
|
|
/*============================================================================
|
|
* Written for Bochs (x86 achitecture simulator) by
|
|
* Stanislav Shwartsman [sshwarts at sourceforge net]
|
|
* ==========================================================================*/
|
|
|
|
#define FLOAT128
|
|
|
|
#include "../build/MAME/platform.h"
|
|
#include "../source/include/internals.h"
|
|
#include "../source/include/softfloat.h"
|
|
|
|
#include "fpu_constant.h"
|
|
#include "softfloat-extra.h"
|
|
#include "softfloat-helpers.h"
|
|
#include "softfloat-specialize.h"
|
|
|
|
static const extFloat80_t floatx80_one = packFloatx80(0, 0x3fff, uint64_t(0x8000000000000000));
|
|
static const extFloat80_t floatx80_ln_2 = packFloatx80(0, 0x3ffe, 0xb17217f7d1cf79acU);
|
|
static const extFloat80_t floatx80_log10_2 = packFloatx80(0, 0x3ffd, 0x9a209a84fbcff798U);
|
|
|
|
static const float128_t float128_one =
|
|
packFloat128(uint64_t(0x3fff000000000000), uint64_t(0x0000000000000000));
|
|
static const float128_t float128_two =
|
|
packFloat128(uint64_t(0x4000000000000000), uint64_t(0x0000000000000000));
|
|
|
|
static const float128_t float128_ln2inv2 =
|
|
packFloat128(uint64_t(0x400071547652b82f), uint64_t(0xe1777d0ffda0d23a));
|
|
|
|
#define SQRT2_HALF_SIG uint64_t(0xb504f333f9de6484)
|
|
|
|
extern float128_t OddPoly(float128_t x, const float128_t *arr, int n);
|
|
|
|
#define L2_ARR_SIZE 9
|
|
|
|
static float128_t ln_arr[L2_ARR_SIZE] =
|
|
{
|
|
PACK_FLOAT_128(0x3fff000000000000, 0x0000000000000000), /* 1 */
|
|
PACK_FLOAT_128(0x3ffd555555555555, 0x5555555555555555), /* 3 */
|
|
PACK_FLOAT_128(0x3ffc999999999999, 0x999999999999999a), /* 5 */
|
|
PACK_FLOAT_128(0x3ffc249249249249, 0x2492492492492492), /* 7 */
|
|
PACK_FLOAT_128(0x3ffbc71c71c71c71, 0xc71c71c71c71c71c), /* 9 */
|
|
PACK_FLOAT_128(0x3ffb745d1745d174, 0x5d1745d1745d1746), /* 11 */
|
|
PACK_FLOAT_128(0x3ffb3b13b13b13b1, 0x3b13b13b13b13b14), /* 13 */
|
|
PACK_FLOAT_128(0x3ffb111111111111, 0x1111111111111111), /* 15 */
|
|
PACK_FLOAT_128(0x3ffae1e1e1e1e1e1, 0xe1e1e1e1e1e1e1e2) /* 17 */
|
|
};
|
|
|
|
static float128_t poly_ln(float128_t x1)
|
|
{
|
|
/*
|
|
//
|
|
// 3 5 7 9 11 13 15
|
|
// 1+u u u u u u u u
|
|
// 1/2 ln --- ~ u + --- + --- + --- + --- + ---- + ---- + ---- =
|
|
// 1-u 3 5 7 9 11 13 15
|
|
//
|
|
// 2 4 6 8 10 12 14
|
|
// u u u u u u u
|
|
// = u * [ 1 + --- + --- + --- + --- + ---- + ---- + ---- ] =
|
|
// 3 5 7 9 11 13 15
|
|
//
|
|
// 3 3
|
|
// -- 4k -- 4k+2
|
|
// p(u) = > C * u q(u) = > C * u
|
|
// -- 2k -- 2k+1
|
|
// k=0 k=0
|
|
//
|
|
// 1+u 2
|
|
// 1/2 ln --- ~ u * [ p(u) + u * q(u) ]
|
|
// 1-u
|
|
//
|
|
*/
|
|
return OddPoly(x1, (const float128_t*) ln_arr, L2_ARR_SIZE);
|
|
}
|
|
|
|
/* required sqrt(2)/2 < x < sqrt(2) */
|
|
static float128_t poly_l2(float128_t x)
|
|
{
|
|
/* using float128 for approximation */
|
|
float128_t x_p1 = f128_add(x, float128_one);
|
|
float128_t x_m1 = f128_sub(x, float128_one);
|
|
x = f128_div(x_m1, x_p1);
|
|
x = poly_ln(x);
|
|
x = f128_mul(x, float128_ln2inv2);
|
|
return x;
|
|
}
|
|
|
|
static float128_t poly_l2p1(float128_t x)
|
|
{
|
|
/* using float128 for approximation */
|
|
float128_t x_plus2 = f128_add(x, float128_two);
|
|
x = f128_div(x, x_plus2);
|
|
x = poly_ln(x);
|
|
x = f128_mul(x, float128_ln2inv2);
|
|
return x;
|
|
}
|
|
|
|
// =================================================
|
|
// FYL2X Compute y * log (x)
|
|
// 2
|
|
// =================================================
|
|
|
|
//
|
|
// Uses the following identities:
|
|
//
|
|
// 1. ----------------------------------------------------------
|
|
// ln(x)
|
|
// log (x) = -------, ln (x*y) = ln(x) + ln(y)
|
|
// 2 ln(2)
|
|
//
|
|
// 2. ----------------------------------------------------------
|
|
// 1+u x-1
|
|
// ln (x) = ln -----, when u = -----
|
|
// 1-u x+1
|
|
//
|
|
// 3. ----------------------------------------------------------
|
|
// 3 5 7 2n+1
|
|
// 1+u u u u u
|
|
// ln ----- = 2 [ u + --- + --- + --- + ... + ------ + ... ]
|
|
// 1-u 3 5 7 2n+1
|
|
//
|
|
|
|
extFloat80_t extFloat80_fyl2x(extFloat80_t a, extFloat80_t b)
|
|
{
|
|
// handle unsupported extended double-precision floating encodings
|
|
if (extF80_isUnsupported(a) || extF80_isUnsupported(b)) {
|
|
invalid:
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid;
|
|
return floatx80_default_nan;
|
|
}
|
|
|
|
uint64_t aSig = extF80_fraction(a);
|
|
int32_t aExp = extF80_exp(a);
|
|
int aSign = extF80_sign(a);
|
|
uint64_t bSig = extF80_fraction(b);
|
|
int32_t bExp = extF80_exp(b);
|
|
int bSign = extF80_sign(b);
|
|
|
|
int zSign = bSign ^ 1;
|
|
|
|
if (aExp == 0x7FFF) {
|
|
if ((aSig<<1) || ((bExp == 0x7FFF) && (bSig<<1))) {
|
|
const uint128 nan = softfloat_propagateNaNExtF80UI(a.signExp, aSig, b.signExp, bSig);
|
|
extFloat80_t rv;
|
|
rv.signExp = nan.v64;
|
|
rv.signif = nan.v0;
|
|
return rv;
|
|
}
|
|
if (aSign) goto invalid;
|
|
else {
|
|
if (! bExp) {
|
|
if (! bSig) goto invalid;
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid; // actually denormal
|
|
}
|
|
return packFloatx80(bSign, 0x7FFF, uint64_t(0x8000000000000000));
|
|
}
|
|
}
|
|
if (bExp == 0x7FFF) {
|
|
if (bSig << 1) {
|
|
const uint128 nan = softfloat_propagateNaNExtF80UI(a.signExp, aSig, b.signExp, bSig);
|
|
extFloat80_t rv;
|
|
rv.signExp = nan.v64;
|
|
rv.signif = nan.v0;
|
|
return rv;
|
|
}
|
|
if (aSign && (uint64_t)(aExp | aSig)) goto invalid;
|
|
if (aSig && ! aExp)
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid; // actually denormal
|
|
if (aExp < 0x3FFF)
|
|
{
|
|
return packFloatx80(zSign, 0x7FFF, uint64_t(0x8000000000000000));
|
|
}
|
|
if (aExp == 0x3FFF && ! (aSig<<1)) goto invalid;
|
|
return packFloatx80(bSign, 0x7FFF, uint64_t(0x8000000000000000));
|
|
}
|
|
if (! aExp) {
|
|
if (! aSig) {
|
|
if ((bExp | bSig) == 0) goto invalid;
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid; // divide by zero
|
|
return packFloatx80(zSign, 0x7FFF, uint64_t(0x8000000000000000));
|
|
}
|
|
if (aSign) goto invalid;
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid; // actually denormal
|
|
struct exp32_sig64 normExpSig = softfloat_normSubnormalExtF80Sig(aSig);
|
|
aExp = normExpSig.exp + 1;
|
|
aSig = normExpSig.sig;
|
|
}
|
|
if (aSign) goto invalid;
|
|
if (! bExp) {
|
|
if (! bSig) {
|
|
if (aExp < 0x3FFF) return packFloatx80(zSign, 0, 0);
|
|
return packFloatx80(bSign, 0, 0);
|
|
}
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid; // actually denormal
|
|
struct exp32_sig64 normExpSig = softfloat_normSubnormalExtF80Sig(bSig);
|
|
bExp = normExpSig.exp + 1;
|
|
bSig = normExpSig.sig;
|
|
}
|
|
if (aExp == 0x3FFF && ! (aSig<<1))
|
|
return packFloatx80(bSign, 0, 0);
|
|
|
|
softfloat_exceptionFlags |= softfloat_flag_inexact;
|
|
|
|
int ExpDiff = aExp - 0x3FFF;
|
|
aExp = 0;
|
|
if (aSig >= SQRT2_HALF_SIG) {
|
|
ExpDiff++;
|
|
aExp--;
|
|
}
|
|
|
|
/* ******************************** */
|
|
/* using float128 for approximation */
|
|
/* ******************************** */
|
|
|
|
float128_t b128 = softfloat_normRoundPackToF128(bSign, bExp-0x10, bSig, 0);
|
|
|
|
uint64_t zSig0, zSig1;
|
|
shortShift128Right(aSig<<1, 0, 16, &zSig0, &zSig1);
|
|
float128_t x = packFloat128(0, aExp+0x3FFF, zSig0, zSig1);
|
|
x = poly_l2(x);
|
|
x = f128_add(x, i32_to_f128(ExpDiff));
|
|
x = f128_mul(x, b128);
|
|
return f128_to_extF80(x);
|
|
}
|
|
|
|
// =================================================
|
|
// FYL2XP1 Compute y * log (x + 1)
|
|
// 2
|
|
// =================================================
|
|
|
|
//
|
|
// Uses the following identities:
|
|
//
|
|
// 1. ----------------------------------------------------------
|
|
// ln(x)
|
|
// log (x) = -------
|
|
// 2 ln(2)
|
|
//
|
|
// 2. ----------------------------------------------------------
|
|
// 1+u x
|
|
// ln (x+1) = ln -----, when u = -----
|
|
// 1-u x+2
|
|
//
|
|
// 3. ----------------------------------------------------------
|
|
// 3 5 7 2n+1
|
|
// 1+u u u u u
|
|
// ln ----- = 2 [ u + --- + --- + --- + ... + ------ + ... ]
|
|
// 1-u 3 5 7 2n+1
|
|
//
|
|
|
|
extFloat80_t extFloat80_fyl2xp1(extFloat80_t a, extFloat80_t b)
|
|
{
|
|
int32_t aExp, bExp;
|
|
uint64_t aSig, bSig, zSig0, zSig1, zSig2;
|
|
int aSign, bSign;
|
|
|
|
// handle unsupported extended double-precision floating encodings
|
|
if (extF80_isUnsupported(a) || extF80_isUnsupported(b)) {
|
|
invalid:
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid;
|
|
return floatx80_default_nan;
|
|
}
|
|
|
|
|
|
aSig = extF80_fraction(a);
|
|
aExp = extF80_exp(a);
|
|
aSign = extF80_sign(a);
|
|
bSig = extF80_fraction(b);
|
|
bExp = extF80_exp(b);
|
|
bSign = extF80_sign(b);
|
|
int zSign = aSign ^ bSign;
|
|
|
|
if (aExp == 0x7FFF) {
|
|
if ((aSig<<1) != 0 || ((bExp == 0x7FFF) && (bSig<<1) != 0)) {
|
|
const uint128 nan = softfloat_propagateNaNExtF80UI(a.signExp, aSig, b.signExp, bSig);
|
|
extFloat80_t rv;
|
|
rv.signExp = nan.v64;
|
|
rv.signif = nan.v0;
|
|
return rv;
|
|
}
|
|
if (aSign) goto invalid;
|
|
else {
|
|
if (! bExp) {
|
|
if (! bSig) goto invalid;
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid; // actually denormal
|
|
}
|
|
return packFloatx80(bSign, 0x7FFF, uint64_t(0x8000000000000000));
|
|
}
|
|
}
|
|
if (bExp == 0x7FFF)
|
|
{
|
|
if (bSig << 1) {
|
|
const uint128 nan = softfloat_propagateNaNExtF80UI(a.signExp, aSig, b.signExp, bSig);
|
|
extFloat80_t rv;
|
|
rv.signExp = nan.v64;
|
|
rv.signif = nan.v0;
|
|
return rv;
|
|
}
|
|
if (! aExp) {
|
|
if (! aSig) goto invalid;
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid; // actually denormal
|
|
}
|
|
|
|
return packFloatx80(zSign, 0x7FFF, uint64_t(0x8000000000000000));
|
|
}
|
|
if (! aExp) {
|
|
if (! aSig) {
|
|
if (bSig && ! bExp)
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid; // actually denormal
|
|
|
|
return packFloatx80(zSign, 0, 0);
|
|
}
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid; // actually denormal
|
|
struct exp32_sig64 normExpSig = softfloat_normSubnormalExtF80Sig(aSig);
|
|
aExp = normExpSig.exp + 1;
|
|
aSig = normExpSig.sig;
|
|
}
|
|
if (! bExp) {
|
|
if (! bSig) return packFloatx80(zSign, 0, 0);
|
|
softfloat_exceptionFlags |= softfloat_flag_invalid; // actually denormal
|
|
struct exp32_sig64 normExpSig = softfloat_normSubnormalExtF80Sig(bSig);
|
|
bExp = normExpSig.exp + 1;
|
|
bSig = normExpSig.sig;
|
|
}
|
|
|
|
softfloat_exceptionFlags |= softfloat_flag_inexact;
|
|
|
|
if (aSign && aExp >= 0x3FFF)
|
|
return a;
|
|
|
|
if (aExp >= 0x3FFC) // big argument
|
|
{
|
|
return extFloat80_fyl2x(extF80_add(a, floatx80_one), b);
|
|
}
|
|
|
|
// handle tiny argument
|
|
if (aExp < FLOATX80_EXP_BIAS-70)
|
|
{
|
|
// first order approximation, return (a*b)/ln(2)
|
|
int32_t zExp = aExp + FLOAT_LN2INV_EXP - 0x3FFE;
|
|
|
|
mul128By64To192(FLOAT_LN2INV_HI, FLOAT_LN2INV_LO, aSig, &zSig0, &zSig1, &zSig2);
|
|
if (0 < (int64_t) zSig0) {
|
|
shortShift128Left(zSig0, zSig1, 1, &zSig0, &zSig1);
|
|
--zExp;
|
|
}
|
|
|
|
zExp = zExp + bExp - 0x3FFE;
|
|
mul128By64To192(zSig0, zSig1, bSig, &zSig0, &zSig1, &zSig2);
|
|
if (0 < (int64_t) zSig0) {
|
|
shortShift128Left(zSig0, zSig1, 1, &zSig0, &zSig1);
|
|
--zExp;
|
|
}
|
|
|
|
return softfloat_roundPackToExtF80(aSign ^ bSign, zExp, zSig0, zSig1, 80);
|
|
}
|
|
|
|
/* ******************************** */
|
|
/* using float128 for approximation */
|
|
/* ******************************** */
|
|
|
|
float128_t b128 = softfloat_normRoundPackToF128(bSign, bExp-0x10, bSig, 0);
|
|
|
|
shortShift128Right(aSig<<1, 0, 16, &zSig0, &zSig1);
|
|
float128_t x = packFloat128(aSign, aExp, zSig0, zSig1);
|
|
x = poly_l2p1(x);
|
|
x = f128_mul(x, b128);
|
|
return f128_to_extF80(x);
|
|
}
|
|
|
|
extFloat80_t extFloat80_lognp1(extFloat80_t a)
|
|
{
|
|
return extFloat80_fyl2x(a, floatx80_ln_2);
|
|
}
|
|
|
|
extFloat80_t extFloat80_logn(extFloat80_t a)
|
|
{
|
|
return extFloat80_fyl2x(a, floatx80_ln_2);
|
|
}
|
|
|
|
extFloat80_t extFloat80_log2(extFloat80_t a)
|
|
{
|
|
return extFloat80_fyl2x(a, floatx80_one);
|
|
}
|
|
|
|
extFloat80_t extFloat80_log10(extFloat80_t a)
|
|
{
|
|
return extFloat80_fyl2x(a, floatx80_log10_2);
|
|
}
|