try alternative 2 for mp_sqr

libtom · Oct 30, 2019 · 44caa78 · 44caa78
1 parent ead719c
commit 44caa78
Show file tree

Hide file tree

Showing 5 changed files with 74 additions and 50 deletions.
diff --git a/demo/test.c b/demo/test.c
@@ -1,6 +1,31 @@
 #include <inttypes.h>
 #include "shared.h"
 
+static mp_err mp_sqr_simple(const mp_int *a, mp_int *b)
+{
+   mp_err err;
+   if (b->dp == a->dp) {
+      mp_int t;
+      if ((err = mp_init_size(&t, (2 * a->used) + 1)) != MP_OKAY) {
+         return err;
+      }
+      if ((err = s_mp_sqr(a, &t)) != MP_OKAY) {
+         mp_clear(&t);
+         return err;
+      }
+      mp_exch(b, &t);
+      mp_clear(&t);
+   } else {
+      if ((err = mp_grow(b, (2 * a->used) + 1)) != MP_OKAY) {
+         return err;
+      }
+      mp_zero(b);
+      err = s_mp_sqr(a, b);
+   }
+   b->sign = MP_ZPOS;
+   return err;
+}
+
 static long rand_long(void)
 {
    long x;
@@ -1930,7 +1955,7 @@ static int test_s_mp_sqr_karatsuba(void)
    for (size = MP_SQR_KARATSUBA_CUTOFF; size < MP_SQR_KARATSUBA_CUTOFF + 20; size++) {
       DO(mp_rand(&a, size));
       DO(s_mp_sqr_karatsuba(&a, &b));
-      DO(s_mp_sqr(&a, &c));
+      DO(mp_sqr_simple(&a, &c));
       if (mp_cmp(&b, &c) != MP_EQ) {
          fprintf(stderr, "Karatsuba squaring failed at size %d\n", size);
          goto LBL_ERR;
@@ -2003,7 +2028,7 @@ static int test_s_mp_sqr_toom(void)
    for (size = MP_SQR_TOOM_CUTOFF; size < MP_SQR_TOOM_CUTOFF + 20; size++) {
       DO(mp_rand(&a, size));
       DO(s_mp_sqr_toom(&a, &b));
-      DO(s_mp_sqr(&a, &c));
+      DO(mp_sqr_simple(&a, &c));
       if (mp_cmp(&b, &c) != MP_EQ) {
          fprintf(stderr, "Toom-Cook 3-way squaring failed at size %d\n", size);
          goto LBL_ERR;

diff --git a/mp_sqr.c b/mp_sqr.c
@@ -3,6 +3,17 @@
 /* LibTomMath, multiple-precision integer library -- Tom St Denis */
 /* SPDX-License-Identifier: Unlicense */
 
+static mp_err s_mp_sqr_noalias(const mp_int *a, mp_int *b)
+{
+   if (MP_HAS(S_MP_SQR_COMBA) && (a->used < (MP_MAX_COMBA / 2))) {
+      return s_mp_sqr_comba(a, b);
+   } else if (MP_HAS(S_MP_SQR)) {
+      return s_mp_sqr(a, b);
+   } else {
+      return MP_VAL;
+   }
+}
+
 /* computes b = a*a */
 mp_err mp_sqr(const mp_int *a, mp_int *b)
 {
@@ -13,14 +24,23 @@ mp_err mp_sqr(const mp_int *a, mp_int *b)
    } else if (MP_HAS(S_MP_SQR_KARATSUBA) &&  /* Karatsuba? */
               (a->used >= MP_SQR_KARATSUBA_CUTOFF)) {
       err = s_mp_sqr_karatsuba(a, b);
-   } else if (MP_HAS(S_MP_SQR_COMBA) && /* can we use the fast comba multiplier? */
-              (a->dp != b->dp) &&
-              (a->used < (MP_MAX_COMBA / 2))) {
-      err = s_mp_sqr_comba(a, b);
-   } else if (MP_HAS(S_MP_SQR)) {
-      err = s_mp_sqr(a, b);
+   } else if (b->dp == a->dp) {
+      mp_int t;
+      if ((err = mp_init_size(&t, (2 * a->used) + 1)) != MP_OKAY) {
+         return err;
+      }
+      if ((err = s_mp_sqr_noalias(a, &t)) != MP_OKAY) {
+         mp_clear(&t);
+         return err;
+      }
+      mp_exch(b, &t);
+      mp_clear(&t);
    } else {
-      err = MP_VAL;
+      if ((err = mp_grow(b, (2 * a->used) + 1)) != MP_OKAY) {
+         return err;
+      }
+      mp_zero(b);
+      err = s_mp_sqr_noalias(a, b);
    }
    b->sign = MP_ZPOS;
    return err;

diff --git a/s_mp_sqr.c b/s_mp_sqr.c
@@ -6,29 +6,19 @@
 /* low level squaring, b = a*a, HAC pp.596-597, Algorithm 14.16 */
 mp_err s_mp_sqr(const mp_int *a, mp_int *b)
 {
-   mp_int   t;
-   int      ix, pa;
-   mp_err   err;
-
-   pa = a->used;
-   if ((err = mp_init_size(&t, (2 * pa) + 1)) != MP_OKAY) {
-      return err;
-   }
-
-   /* default used is maximum possible size */
-   t.used = (2 * pa) + 1;
+   int ix, pa = a->used;
 
    for (ix = 0; ix < pa; ix++) {
       mp_digit u;
       int iy;
 
       /* first calculate the digit at 2*ix */
       /* calculate double precision result */
-      mp_word r = (mp_word)t.dp[2*ix] +
+      mp_word r = (mp_word)b->dp[2*ix] +
                   ((mp_word)a->dp[ix] * (mp_word)a->dp[ix]);
 
       /* store lower part in result */
-      t.dp[ix+ix] = (mp_digit)(r & (mp_word)MP_MASK);
+      b->dp[ix+ix] = (mp_digit)(r & (mp_word)MP_MASK);
 
       /* get the carry */
       u           = (mp_digit)(r >> (mp_word)MP_DIGIT_BIT);
@@ -40,26 +30,25 @@ mp_err s_mp_sqr(const mp_int *a, mp_int *b)
          /* now calculate the double precision result, note we use
           * addition instead of *2 since it's easier to optimize
           */
-         r       = (mp_word)t.dp[ix + iy] + r + r + (mp_word)u;
+         r       = (mp_word)b->dp[ix + iy] + r + r + (mp_word)u;
 
          /* store lower part */
-         t.dp[ix + iy] = (mp_digit)(r & (mp_word)MP_MASK);
+         b->dp[ix + iy] = (mp_digit)(r & (mp_word)MP_MASK);
 
          /* get carry */
          u       = (mp_digit)(r >> (mp_word)MP_DIGIT_BIT);
       }
       /* propagate upwards */
       while (u != 0uL) {
-         r       = (mp_word)t.dp[ix + iy] + (mp_word)u;
-         t.dp[ix + iy] = (mp_digit)(r & (mp_word)MP_MASK);
+         r       = (mp_word)b->dp[ix + iy] + (mp_word)u;
+         b->dp[ix + iy] = (mp_digit)(r & (mp_word)MP_MASK);
          u       = (mp_digit)(r >> (mp_word)MP_DIGIT_BIT);
          ++iy;
       }
    }
 
-   mp_clamp(&t);
-   mp_exch(&t, b);
-   mp_clear(&t);
+   b->used = (2 * pa) + 1;
+   mp_clamp(b);
    return MP_OKAY;
 }
 #endif
diff --git a/s_mp_sqr_comba.c b/s_mp_sqr_comba.c
@@ -15,33 +15,26 @@ After that loop you do the squares and add them in.
 
 mp_err s_mp_sqr_comba(const mp_int *a, mp_int *b)
 {
-   int       oldused, pa, ix;
+   int       ix, pa = a->used, pb = 2 * pa;
    mp_word   W1;
-   mp_err err;
-
-   /* grow the destination as required */
-   pa = a->used + a->used;
-   if ((err = mp_grow(b, pa)) != MP_OKAY) {
-      return err;
-   }
 
    /* number of output digits to produce */
    W1 = 0;
-   for (ix = 0; ix < pa; ix++) {
+   for (ix = 0; ix < pb; ix++) {
       int      tx, ty, iy, iz;
       mp_word  W;
 
       /* clear counter */
       W = 0;
 
       /* get offsets into the two bignums */
-      ty = MP_MIN(a->used-1, ix);
+      ty = MP_MIN(pa-1, ix);
       tx = ix - ty;
 
       /* this is the number of times the loop will iterrate, essentially
-         while (tx++ < a->used && ty-- >= 0) { ... }
+         while (tx++ < pa && ty-- >= 0) { ... }
        */
-      iy = MP_MIN(a->used-tx, ty+1);
+      iy = MP_MIN(pa-tx, ty+1);
 
       /* now for squaring tx can never equal ty
        * we halve the distance since they approach at a rate of 2x
@@ -69,11 +62,7 @@ mp_err s_mp_sqr_comba(const mp_int *a, mp_int *b)
       W1 = W >> (mp_word)MP_DIGIT_BIT;
    }
 
-   /* clear unused digits [that existed in the old copy of c] */
-   oldused = b->used;
-   b->used = a->used + a->used;
-   s_mp_zero_digs(b->dp + b->used, oldused - b->used);
-
+   b->used = pb;
    mp_clamp(b);
    return MP_OKAY;
 }

diff --git a/tommath_class.h b/tommath_class.h
@@ -889,9 +889,14 @@
 #endif
 
 #if defined(MP_SQR_C)
+#   define MP_CLEAR_C
+#   define MP_EXCH_C
+#   define MP_GROW_C
+#   define MP_INIT_SIZE_C
 #   define S_MP_SQR_C
 #   define S_MP_SQR_COMBA_C
 #   define S_MP_SQR_KARATSUBA_C
+#   define S_MP_SQR_NOALIAS_C
 #   define S_MP_SQR_TOOM_C
 #endif
 
@@ -1222,14 +1227,10 @@
 
 #if defined(S_MP_SQR_C)
 #   define MP_CLAMP_C
-#   define MP_CLEAR_C
-#   define MP_EXCH_C
-#   define MP_INIT_SIZE_C
 #endif
 
 #if defined(S_MP_SQR_COMBA_C)
 #   define MP_CLAMP_C
-#   define MP_GROW_C
 #   define S_MP_ZERO_DIGS_C
 #endif