Remove one parameter from word_madd2 and word_madd3

4 files changed, 47 insertions, 40 deletions

diff --git a/include/mp_asm.h b/include/mp_asm.h
index 5f0747d88..1166d9c8d 100644
--- a/include/mp_asm.h
+++ b/include/mp_asm.h
@@ -27,20 +27,20 @@ extern "C" {
 /*************************************************
 * Word Multiply/Add                              *
 *************************************************/
-inline word word_madd2(word a, word b, word c, word* carry)
+inline word word_madd2(word a, word b, word* c)
    {
-   dword z = (dword)a * b + c;
-   *carry = (word)(z >> BOTAN_MP_WORD_BITS);
+   dword z = (dword)a * b + *c;
+   *c = (word)(z >> BOTAN_MP_WORD_BITS);
    return (word)z;
    }
 
 /*************************************************
 * Word Multiply/Add                              *
 *************************************************/
-inline word word_madd3(word a, word b, word c, word d, word* carry)
+inline word word_madd3(word a, word b, word c, word *d)
    {
-   dword z = (dword)a * b + c + d;
-   *carry = (word)(z >> BOTAN_MP_WORD_BITS);
+   dword z = (dword)a * b + c + *d;
+   *d = (word)(z >> BOTAN_MP_WORD_BITS);
    return (word)z;
    }
 
@@ -49,7 +49,11 @@ inline word word_madd3(word a, word b, word c, word d, word* carry)
 *************************************************/
 inline void word3_muladd(word* w2, word* w1, word* w0, word a, word b)
    {
-   *w0 = word_madd2(a, b, *w0, &b);
+   dword z = (dword)a * b + *w0;
+
+   b = (word)(z >> BOTAN_MP_WORD_BITS);
+   *w0 = (word)z;
+
    *w1 += b;
    *w2 += (*w1 < b) ? 1 : 0;
    }
@@ -59,7 +63,10 @@ inline void word3_muladd(word* w2, word* w1, word* w0, word a, word b)
 *************************************************/
 inline void word3_muladd_2(word* w2, word* w1, word* w0, word a, word b)
    {
-   a = word_madd2(a, b, 0, &b);
+   dword z = (dword)a * b;
+
+   b = (word)(z >> BOTAN_MP_WORD_BITS);
+   a = (word)z;
 
    *w0 += a;
    *w1 += b + ((*w0 < a) ? 1 : 0);
diff --git a/include/mp_asmi.h b/include/mp_asmi.h
index b05199d0a..70e25d07b 100644
--- a/include/mp_asmi.h
+++ b/include/mp_asmi.h
@@ -107,14 +107,14 @@ inline word word8_sub3(word z[8], const word x[8],
 *************************************************/
 inline word word8_linmul2(word x[4], word y, word carry)
    {
-   x[0] = word_madd2(x[0], y, carry, &carry);
-   x[1] = word_madd2(x[1], y, carry, &carry);
-   x[2] = word_madd2(x[2], y, carry, &carry);
-   x[3] = word_madd2(x[3], y, carry, &carry);
-   x[4] = word_madd2(x[4], y, carry, &carry);
-   x[5] = word_madd2(x[5], y, carry, &carry);
-   x[6] = word_madd2(x[6], y, carry, &carry);
-   x[7] = word_madd2(x[7], y, carry, &carry);
+   x[0] = word_madd2(x[0], y, &carry);
+   x[1] = word_madd2(x[1], y, &carry);
+   x[2] = word_madd2(x[2], y, &carry);
+   x[3] = word_madd2(x[3], y, &carry);
+   x[4] = word_madd2(x[4], y, &carry);
+   x[5] = word_madd2(x[5], y, &carry);
+   x[6] = word_madd2(x[6], y, &carry);
+   x[7] = word_madd2(x[7], y, &carry);
    return carry;
    }
 
@@ -123,14 +123,14 @@ inline word word8_linmul2(word x[4], word y, word carry)
 *************************************************/
 inline word word8_linmul3(word z[8], const word x[8], word y, word carry)
    {
-   z[0] = word_madd2(x[0], y, carry, &carry);
-   z[1] = word_madd2(x[1], y, carry, &carry);
-   z[2] = word_madd2(x[2], y, carry, &carry);
-   z[3] = word_madd2(x[3], y, carry, &carry);
-   z[4] = word_madd2(x[4], y, carry, &carry);
-   z[5] = word_madd2(x[5], y, carry, &carry);
-   z[6] = word_madd2(x[6], y, carry, &carry);
-   z[7] = word_madd2(x[7], y, carry, &carry);
+   z[0] = word_madd2(x[0], y, &carry);
+   z[1] = word_madd2(x[1], y, &carry);
+   z[2] = word_madd2(x[2], y, &carry);
+   z[3] = word_madd2(x[3], y, &carry);
+   z[4] = word_madd2(x[4], y, &carry);
+   z[5] = word_madd2(x[5], y, &carry);
+   z[6] = word_madd2(x[6], y, &carry);
+   z[7] = word_madd2(x[7], y, &carry);
    return carry;
    }
 
@@ -139,14 +139,14 @@ inline word word8_linmul3(word z[8], const word x[8], word y, word carry)
 *************************************************/
 inline word word8_madd3(word z[8], const word x[8], word y, word carry)
    {
-   z[0] = word_madd3(x[0], y, z[0], carry, &carry);
-   z[1] = word_madd3(x[1], y, z[1], carry, &carry);
-   z[2] = word_madd3(x[2], y, z[2], carry, &carry);
-   z[3] = word_madd3(x[3], y, z[3], carry, &carry);
-   z[4] = word_madd3(x[4], y, z[4], carry, &carry);
-   z[5] = word_madd3(x[5], y, z[5], carry, &carry);
-   z[6] = word_madd3(x[6], y, z[6], carry, &carry);
-   z[7] = word_madd3(x[7], y, z[7], carry, &carry);
+   z[0] = word_madd3(x[0], y, z[0], &carry);
+   z[1] = word_madd3(x[1], y, z[1], &carry);
+   z[2] = word_madd3(x[2], y, z[2], &carry);
+   z[3] = word_madd3(x[3], y, z[3], &carry);
+   z[4] = word_madd3(x[4], y, z[4], &carry);
+   z[5] = word_madd3(x[5], y, z[5], &carry);
+   z[6] = word_madd3(x[6], y, z[6], &carry);
+   z[7] = word_madd3(x[7], y, z[7], &carry);
    return carry;
    }
 
diff --git a/src/mp_asm.cpp b/src/mp_asm.cpp
index ec2e31011..ef8292083 100644
--- a/src/mp_asm.cpp
+++ b/src/mp_asm.cpp
@@ -148,7 +148,7 @@ void bigint_linmul2(word x[], u32bit x_size, word y)
       carry = word8_linmul2(x + j, y, carry);
 
    for(u32bit j = blocks; j != x_size; ++j)
-      x[j] = word_madd2(x[j], y, carry, &carry);
+      x[j] = word_madd2(x[j], y, &carry);
 
    x[x_size] = carry;
    }
@@ -166,7 +166,7 @@ void bigint_linmul3(word z[], const word x[], u32bit x_size, word y)
       carry = word8_linmul3(z + j, x + j, y, carry);
 
    for(u32bit j = blocks; j != x_size; ++j)
-      z[j] = word_madd2(x[j], y, carry, &carry);
+      z[j] = word_madd2(x[j], y, &carry);
 
    z[x_size] = carry;
    }
@@ -190,7 +190,7 @@ void bigint_simple_mul(word z[], const word x[], u32bit x_size,
          carry = word8_madd3(z + j + k, y + k, x_j, carry);
 
       for(u32bit k = blocks; k != y_size; ++k)
-         z[j+k] = word_madd3(x_j, y[k], z[j+k], carry, &carry);
+         z[j+k] = word_madd3(x_j, y[k], z[j+k], &carry);
 
       z[j+y_size] = carry;
       }
@@ -215,7 +215,7 @@ void montgomery_reduce(word z[], u32bit z_size,
          carry = word8_madd3(z_j + k, x + k, y, carry);
 
       for(u32bit k = blocks; k != x_size; ++k)
-         z_j[k] = word_madd3(x[k], y, z_j[k], carry, &carry);
+         z_j[k] = word_madd3(x[k], y, z_j[k], &carry);
 
       word carry2 = 0;
       z_j[x_size] = word_add(z_j[x_size], carry, &carry2);
diff --git a/src/mp_misc.cpp b/src/mp_misc.cpp
index f551948b7..782ad9d15 100644
--- a/src/mp_misc.cpp
+++ b/src/mp_misc.cpp
@@ -17,8 +17,8 @@ u32bit bigint_divcore(word q, word y1, word y2,
                       word x1, word x2, word x3)
    {
    word y0 = 0;
-   y2 = word_madd2(q, y2, y0, &y0);
-   y1 = word_madd2(q, y1, y0, &y0);
+   y2 = word_madd2(q, y2, &y0);
+   y1 = word_madd2(q, y1, &y0);
 
    if(y0 > x1) return 1;
    if(y0 < x1) return 0;
@@ -80,8 +80,8 @@ word bigint_divop(word n1, word n0, word d)
 word bigint_modop(word n1, word n0, word d)
    {
    word z = bigint_divop(n1, n0, d);
-   word dummy = 0;
-   z = word_madd2(z, d, dummy, &dummy);
+   word carry = 0;
+   z = word_madd2(z, d, &carry);
    return (n0-z);
    }