Subversion Repositories HelenOS-historic

/*

 * Copyright (C) 2005 Josef Cejka

 * All rights reserved.

 *

 * Redistribution and use in source and binary forms, with or without

 * modification, are permitted provided that the following conditions

 * are met:

 *

 * - Redistributions of source code must retain the above copyright

 *   notice, this list of conditions and the following disclaimer.

 * - Redistributions in binary form must reproduce the above copyright

 *   notice, this list of conditions and the following disclaimer in the

 *   documentation and/or other materials provided with the distribution.

 * - The name of the author may not be used to endorse or promote products

 *   derived from this software without specific prior written permission.

 *

 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR

 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES

 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.

 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,

 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT

 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,

 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY

 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT

 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF

 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

 */

#include<sftypes.h>

#include<add.h>

#include<div.h>

#include<comparison.h>

#include<mul.h>

#include<common.h>

float32 divFloat32(float32 a, float32 b)

{

    float32 result;

    __s32 aexp, bexp, cexp;

    __u64 afrac, bfrac, cfrac;

    result.parts.sign = a.parts.sign ^ b.parts.sign;

    if (isFloat32NaN(a)) {

        if (isFloat32SigNaN(a)) {

            /*FIXME: SigNaN*/

        }

        /*NaN*/

        return a;

    }

    if (isFloat32NaN(b)) {

        if (isFloat32SigNaN(b)) {

            /*FIXME: SigNaN*/

        }

        /*NaN*/

        return b;

    }

    if (isFloat32Infinity(a)) {

        if (isFloat32Infinity(b)) {

            /*FIXME: inf / inf */

            result.binary = FLOAT32_NAN;

            return result;

        }

        /* inf / num */

        result.parts.exp = a.parts.exp;

        result.parts.fraction = a.parts.fraction;

        return result;

    }

    if (isFloat32Infinity(b)) {

        if (isFloat32Zero(a)) {

            /* FIXME 0 / inf */

            result.parts.exp = 0;

            result.parts.fraction = 0;

            return result;

        }

        /* FIXME: num / inf*/

        result.parts.exp = 0;

        result.parts.fraction = 0;

        return result;

    }

    if (isFloat32Zero(b)) {

        if (isFloat32Zero(a)) {

            /*FIXME: 0 / 0*/

            result.binary = FLOAT32_NAN;

            return result;

        }

        /* FIXME: division by zero */

        result.parts.exp = 0;

        result.parts.fraction = 0;

        return result;

    }

    afrac = a.parts.fraction;

    aexp = a.parts.exp;

    bfrac = b.parts.fraction;

    bexp = b.parts.exp;

    /* denormalized numbers */

    if (aexp == 0) {

        if (afrac == 0) {

        result.parts.exp = 0;

        result.parts.fraction = 0;

        return result;

        }

        /* normalize it*/

        afrac <<= 1;

            /* afrac is nonzero => it must stop */ 

        while (! (afrac & FLOAT32_HIDDEN_BIT_MASK) ) {

            afrac <<= 1;

            aexp--;

        }

    }

    if (bexp == 0) {

        bfrac <<= 1;

            /* bfrac is nonzero => it must stop */ 

        while (! (bfrac & FLOAT32_HIDDEN_BIT_MASK) ) {

            bfrac <<= 1;

            bexp--;

        }

    }

    afrac = (afrac | FLOAT32_HIDDEN_BIT_MASK ) << (32 - FLOAT32_FRACTION_SIZE - 1 );

    bfrac = (bfrac | FLOAT32_HIDDEN_BIT_MASK ) << (32 - FLOAT32_FRACTION_SIZE );

    if ( bfrac <= (afrac << 1) ) {

        afrac >>= 1;

        aexp++;

    }

    cexp = aexp - bexp + FLOAT32_BIAS - 2;

    cfrac = (afrac << 32) / bfrac;

    if ((  cfrac & 0x3F ) == 0) {

        cfrac |= ( bfrac * cfrac != afrac << 32 );

    }

    /* pack and round */

    /* find first nonzero digit and shift result and detect possibly underflow */

    while ((cexp > 0) && (cfrac) && (!(cfrac & (FLOAT32_HIDDEN_BIT_MASK << 7 )))) {

        cexp--;

        cfrac <<= 1;

            /* TODO: fix underflow */

    };

    cfrac += (0x1 << 6); /* FIXME: 7 is not sure*/

    if (cfrac & (FLOAT32_HIDDEN_BIT_MASK << 7)) {

        ++cexp;

        cfrac >>= 1;

        }  

    /* check overflow */

    if (cexp >= FLOAT32_MAX_EXPONENT ) {

        /* FIXME: overflow, return infinity */

        result.parts.exp = FLOAT32_MAX_EXPONENT;

        result.parts.fraction = 0;

        return result;

    }

    if (cexp < 0) {

        /* FIXME: underflow */

        result.parts.exp = 0;

        if ((cexp + FLOAT32_FRACTION_SIZE) < 0) {

            result.parts.fraction = 0;

            return result;

        }

        cfrac >>= 1;

        while (cexp < 0) {

            cexp ++;

            cfrac >>= 1;

        }

    } else {

        result.parts.exp = (__u32)cexp;

    }

    result.parts.fraction = ((cfrac >> 6) & (~FLOAT32_HIDDEN_BIT_MASK));

    return result; 

}

float64 divFloat64(float64 a, float64 b)

{

    float64 result;

    __s64 aexp, bexp, cexp;

    __u64 afrac, bfrac, cfrac;

    __u64 remlo, remhi;

    result.parts.sign = a.parts.sign ^ b.parts.sign;

    if (isFloat64NaN(a)) {

        if (isFloat64SigNaN(b)) {

            /*FIXME: SigNaN*/

            return b;

        }

        if (isFloat64SigNaN(a)) {

            /*FIXME: SigNaN*/

        }

        /*NaN*/

        return a;

    }

    if (isFloat64NaN(b)) {

        if (isFloat64SigNaN(b)) {

            /*FIXME: SigNaN*/

        }

        /*NaN*/

        return b;

    }

    if (isFloat64Infinity(a)) {

        if (isFloat64Infinity(b) || isFloat64Zero(b)) {

            /*FIXME: inf / inf */

            result.binary = FLOAT64_NAN;

            return result;

        }

        /* inf / num */

        result.parts.exp = a.parts.exp;

        result.parts.fraction = a.parts.fraction;

        return result;

    }

    if (isFloat64Infinity(b)) {

        if (isFloat64Zero(a)) {

            /* FIXME 0 / inf */

            result.parts.exp = 0;

            result.parts.fraction = 0;

            return result;

        }

        /* FIXME: num / inf*/

        result.parts.exp = 0;

        result.parts.fraction = 0;

        return result;

    }

    if (isFloat64Zero(b)) {

        if (isFloat64Zero(a)) {

            /*FIXME: 0 / 0*/

            result.binary = FLOAT64_NAN;

            return result;

        }

        /* FIXME: division by zero */

        result.parts.exp = 0;

        result.parts.fraction = 0;

        return result;

    }

    afrac = a.parts.fraction;

    aexp = a.parts.exp;

    bfrac = b.parts.fraction;

    bexp = b.parts.exp;

    /* denormalized numbers */

    if (aexp == 0) {

        if (afrac == 0) {

            result.parts.exp = 0;

            result.parts.fraction = 0;

            return result;

        }

        /* normalize it*/

        aexp++;

            /* afrac is nonzero => it must stop */ 

        while (! (afrac & FLOAT64_HIDDEN_BIT_MASK) ) {

            afrac <<= 1;

            aexp--;

        }

    }

    if (bexp == 0) {

        bexp++;

            /* bfrac is nonzero => it must stop */ 

        while (! (bfrac & FLOAT64_HIDDEN_BIT_MASK) ) {

            bfrac <<= 1;

            bexp--;

        }

    }

    afrac = (afrac | FLOAT64_HIDDEN_BIT_MASK ) << (64 - FLOAT64_FRACTION_SIZE - 2 );

    bfrac = (bfrac | FLOAT64_HIDDEN_BIT_MASK ) << (64 - FLOAT64_FRACTION_SIZE - 1);

    if ( bfrac <= (afrac << 1) ) {

        afrac >>= 1;

        aexp++;

    }

    cexp = aexp - bexp + FLOAT64_BIAS - 2;

    cfrac = divFloat64estim(afrac, bfrac);

    if ((  cfrac & 0x1FF ) <= 2) { /*FIXME:?? */

        mul64integers( bfrac, cfrac, &remlo, &remhi);

        /* (__u128)afrac << 64 - ( ((__u128)remhi<<64) + (__u128)remlo )*/ 

        remhi = afrac - remhi - ( remlo > 0);

        remlo = - remlo;

        while ((__s64) remhi < 0) {

            cfrac--;

            remlo += bfrac;

            remhi += ( remlo < bfrac );

        }

        cfrac |= ( remlo != 0 );

    }

    /* round and shift */

    result = finishFloat64(cexp, cfrac, result.parts.sign);

    return result;

}

__u64 divFloat64estim(__u64 a, __u64 b)

{

    __u64 bhi;

    __u64 remhi, remlo;

    __u64 result;

    if ( b <= a ) {

        return 0xFFFFFFFFFFFFFFFFull;

    }

    bhi = b >> 32;

    result = ((bhi << 32) <= a) ?( 0xFFFFFFFFull << 32) : ( a / bhi) << 32;

    mul64integers(b, result, &remlo, &remhi);

    remhi = a - remhi - (remlo > 0);

    remlo = - remlo;

    b <<= 32;

    while ( (__s64) remhi < 0 ) {

            result -= 0x1ll << 32; 

            remlo += b;

            remhi += bhi + ( remlo < b );

        }

    remhi = (remhi << 32) | (remlo >> 32);

    if (( bhi << 32) <= remhi) {

        result |= 0xFFFFFFFF;

    } else {

        result |= remhi / bhi;

    }

    return result;

}