diff --git a/TODO b/TODO
new file mode 100644
index 0000000..deffba1
--- /dev/null
+++ b/TODO
@@ -0,0 +1,16 @@
+things for book in order of importance...
+
+- Fix up pseudo-code [only] for combas that are not consistent with source
+- Start in chapter 3 [basics] and work up...
+   - re-write to prose [less abrupt]
+   - clean up pseudo code [spacing]
+   - more examples where appropriate and figures
+
+Goal:
+   - Get sync done by mid January [roughly 8-12 hours work]
+   - Finish ch3-6 by end of January [roughly 12-16 hours of work]
+   - Finish ch7-end by mid Feb [roughly 20-24 hours of work].
+
+Goal isn't "first edition" but merely cleaner to read.
+
+
diff --git a/bn.pdf b/bn.pdf
index fbd5b2a..9b873e1 100644
Binary files a/bn.pdf and b/bn.pdf differ
diff --git a/bn.tex b/bn.tex
index 74a4f01..962d6ea 100644
--- a/bn.tex
+++ b/bn.tex
@@ -49,7 +49,7 @@
 \begin{document}
 \frontmatter
 \pagestyle{empty}
-\title{LibTomMath User Manual \\ v0.32}
+\title{LibTomMath User Manual \\ v0.33}
 \author{Tom St Denis \\ tomstdenis@iahu.ca}
 \maketitle
 This text, the library and the accompanying textbook are all hereby placed in the public domain.  This book has been 
diff --git a/bn_fast_mp_invmod.c b/bn_fast_mp_invmod.c
index 492a3f1..b5b9f10 100644
--- a/bn_fast_mp_invmod.c
+++ b/bn_fast_mp_invmod.c
@@ -39,20 +39,20 @@ fast_mp_invmod (mp_int * a, mp_int * b, mp_int * c)
 
   /* x == modulus, y == value to invert */
   if ((res = mp_copy (b, &x)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* we need y = |a| */
   if ((res = mp_abs (a, &y)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* 3. u=x, v=y, A=1, B=0, C=0,D=1 */
   if ((res = mp_copy (&x, &u)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
   if ((res = mp_copy (&y, &v)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
   mp_set (&D, 1);
 
@@ -61,17 +61,17 @@ top:
   while (mp_iseven (&u) == 1) {
     /* 4.1 u = u/2 */
     if ((res = mp_div_2 (&u, &u)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     /* 4.2 if B is odd then */
     if (mp_isodd (&B) == 1) {
       if ((res = mp_sub (&B, &x, &B)) != MP_OKAY) {
-        goto __ERR;
+        goto LBL_ERR;
       }
     }
     /* B = B/2 */
     if ((res = mp_div_2 (&B, &B)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -79,18 +79,18 @@ top:
   while (mp_iseven (&v) == 1) {
     /* 5.1 v = v/2 */
     if ((res = mp_div_2 (&v, &v)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     /* 5.2 if D is odd then */
     if (mp_isodd (&D) == 1) {
       /* D = (D-x)/2 */
       if ((res = mp_sub (&D, &x, &D)) != MP_OKAY) {
-        goto __ERR;
+        goto LBL_ERR;
       }
     }
     /* D = D/2 */
     if ((res = mp_div_2 (&D, &D)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -98,20 +98,20 @@ top:
   if (mp_cmp (&u, &v) != MP_LT) {
     /* u = u - v, B = B - D */
     if ((res = mp_sub (&u, &v, &u)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&B, &D, &B)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   } else {
     /* v - v - u, D = D - B */
     if ((res = mp_sub (&v, &u, &v)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&D, &B, &D)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -125,21 +125,21 @@ top:
   /* if v != 1 then there is no inverse */
   if (mp_cmp_d (&v, 1) != MP_EQ) {
     res = MP_VAL;
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* b is now the inverse */
   neg = a->sign;
   while (D.sign == MP_NEG) {
     if ((res = mp_add (&D, b, &D)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
   mp_exch (&D, c);
   c->sign = neg;
   res = MP_OKAY;
 
-__ERR:mp_clear_multi (&x, &y, &u, &v, &B, &D, NULL);
+LBL_ERR:mp_clear_multi (&x, &y, &u, &v, &B, &D, NULL);
   return res;
 }
 #endif
diff --git a/bn_fast_s_mp_mul_digs.c b/bn_fast_s_mp_mul_digs.c
index 92b50bb..e1ff5f3 100644
--- a/bn_fast_s_mp_mul_digs.c
+++ b/bn_fast_s_mp_mul_digs.c
@@ -50,7 +50,7 @@ fast_s_mp_mul_digs (mp_int * a, mp_int * b, mp_int * c, int digs)
 
   /* clear the carry */
   _W = 0;
-  for (ix = 0; ix <= pa; ix++) { 
+  for (ix = 0; ix < pa; ix++) { 
       int      tx, ty;
       int      iy;
       mp_digit *tmpx, *tmpy;
@@ -80,6 +80,9 @@ fast_s_mp_mul_digs (mp_int * a, mp_int * b, mp_int * c, int digs)
       _W = _W >> ((mp_word)DIGIT_BIT);
   }
 
+  /* store final carry */
+  W[ix] = _W;
+
   /* setup dest */
   olduse  = c->used;
   c->used = digs;
diff --git a/bn_fast_s_mp_mul_high_digs.c b/bn_fast_s_mp_mul_high_digs.c
index 9e0cf55..064a9dd 100644
--- a/bn_fast_s_mp_mul_high_digs.c
+++ b/bn_fast_s_mp_mul_high_digs.c
@@ -42,7 +42,7 @@ fast_s_mp_mul_high_digs (mp_int * a, mp_int * b, mp_int * c, int digs)
   /* number of output digits to produce */
   pa = a->used + b->used;
   _W = 0;
-  for (ix = digs; ix <= pa; ix++) { 
+  for (ix = digs; ix < pa; ix++) { 
       int      tx, ty, iy;
       mp_digit *tmpx, *tmpy;
 
@@ -70,6 +70,9 @@ fast_s_mp_mul_high_digs (mp_int * a, mp_int * b, mp_int * c, int digs)
       /* make next carry */
       _W = _W >> ((mp_word)DIGIT_BIT);
   }
+  
+  /* store final carry */
+  W[ix] = _W;
 
   /* setup dest */
   olduse  = c->used;
diff --git a/bn_fast_s_mp_sqr.c b/bn_fast_s_mp_sqr.c
index 9f6962d..d6014ab 100644
--- a/bn_fast_s_mp_sqr.c
+++ b/bn_fast_s_mp_sqr.c
@@ -60,7 +60,7 @@ int fast_s_mp_sqr (mp_int * a, mp_int * b)
 
   /* number of output digits to produce */
   W1 = 0;
-  for (ix = 0; ix <= pa; ix++) { 
+  for (ix = 0; ix < pa; ix++) { 
       int      tx, ty, iy;
       mp_word  _W;
       mp_digit *tmpy;
diff --git a/bn_mp_div.c b/bn_mp_div.c
index 39d921a..6b2b8f0 100644
--- a/bn_mp_div.c
+++ b/bn_mp_div.c
@@ -49,23 +49,23 @@ int mp_div(mp_int * a, mp_int * b, mp_int * c, mp_int * d)
 
   mp_set(&tq, 1);
   n = mp_count_bits(a) - mp_count_bits(b);
-  if (((res = mp_copy(a, &ta)) != MP_OKAY) ||
-      ((res = mp_copy(b, &tb)) != MP_OKAY) || 
+  if (((res = mp_abs(a, &ta)) != MP_OKAY) ||
+      ((res = mp_abs(b, &tb)) != MP_OKAY) || 
       ((res = mp_mul_2d(&tb, n, &tb)) != MP_OKAY) ||
       ((res = mp_mul_2d(&tq, n, &tq)) != MP_OKAY)) {
-      goto __ERR;
+      goto LBL_ERR;
   }
 
   while (n-- >= 0) {
      if (mp_cmp(&tb, &ta) != MP_GT) {
         if (((res = mp_sub(&ta, &tb, &ta)) != MP_OKAY) ||
             ((res = mp_add(&q, &tq, &q)) != MP_OKAY)) {
-           goto __ERR;
+           goto LBL_ERR;
         }
      }
      if (((res = mp_div_2d(&tb, 1, &tb, NULL)) != MP_OKAY) ||
          ((res = mp_div_2d(&tq, 1, &tq, NULL)) != MP_OKAY)) {
-           goto __ERR;
+           goto LBL_ERR;
      }
   }
 
@@ -74,13 +74,13 @@ int mp_div(mp_int * a, mp_int * b, mp_int * c, mp_int * d)
   n2 = (a->sign == b->sign ? MP_ZPOS : MP_NEG);
   if (c != NULL) {
      mp_exch(c, &q);
-     c->sign  = n2;
+     c->sign  = (mp_iszero(c) == MP_YES) ? MP_ZPOS : n2;
   }
   if (d != NULL) {
      mp_exch(d, &ta);
-     d->sign = n;
+     d->sign = (mp_iszero(d) == MP_YES) ? MP_ZPOS : n;
   }
-__ERR:
+LBL_ERR:
    mp_clear_multi(&ta, &tb, &tq, &q, NULL);
    return res;
 }
@@ -129,19 +129,19 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
   q.used = a->used + 2;
 
   if ((res = mp_init (&t1)) != MP_OKAY) {
-    goto __Q;
+    goto LBL_Q;
   }
 
   if ((res = mp_init (&t2)) != MP_OKAY) {
-    goto __T1;
+    goto LBL_T1;
   }
 
   if ((res = mp_init_copy (&x, a)) != MP_OKAY) {
-    goto __T2;
+    goto LBL_T2;
   }
 
   if ((res = mp_init_copy (&y, b)) != MP_OKAY) {
-    goto __X;
+    goto LBL_X;
   }
 
   /* fix the sign */
@@ -153,10 +153,10 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
   if (norm < (int)(DIGIT_BIT-1)) {
      norm = (DIGIT_BIT-1) - norm;
      if ((res = mp_mul_2d (&x, norm, &x)) != MP_OKAY) {
-       goto __Y;
+       goto LBL_Y;
      }
      if ((res = mp_mul_2d (&y, norm, &y)) != MP_OKAY) {
-       goto __Y;
+       goto LBL_Y;
      }
   } else {
      norm = 0;
@@ -168,13 +168,13 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
 
   /* while (x >= y*b**n-t) do { q[n-t] += 1; x -= y*b**{n-t} } */
   if ((res = mp_lshd (&y, n - t)) != MP_OKAY) { /* y = y*b**{n-t} */
-    goto __Y;
+    goto LBL_Y;
   }
 
   while (mp_cmp (&x, &y) != MP_LT) {
     ++(q.dp[n - t]);
     if ((res = mp_sub (&x, &y, &x)) != MP_OKAY) {
-      goto __Y;
+      goto LBL_Y;
     }
   }
 
@@ -216,7 +216,7 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
       t1.dp[1] = y.dp[t];
       t1.used = 2;
       if ((res = mp_mul_d (&t1, q.dp[i - t - 1], &t1)) != MP_OKAY) {
-        goto __Y;
+        goto LBL_Y;
       }
 
       /* find right hand */
@@ -228,27 +228,27 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
 
     /* step 3.3 x = x - q{i-t-1} * y * b**{i-t-1} */
     if ((res = mp_mul_d (&y, q.dp[i - t - 1], &t1)) != MP_OKAY) {
-      goto __Y;
+      goto LBL_Y;
     }
 
     if ((res = mp_lshd (&t1, i - t - 1)) != MP_OKAY) {
-      goto __Y;
+      goto LBL_Y;
     }
 
     if ((res = mp_sub (&x, &t1, &x)) != MP_OKAY) {
-      goto __Y;
+      goto LBL_Y;
     }
 
     /* if x < 0 then { x = x + y*b**{i-t-1}; q{i-t-1} -= 1; } */
     if (x.sign == MP_NEG) {
       if ((res = mp_copy (&y, &t1)) != MP_OKAY) {
-        goto __Y;
+        goto LBL_Y;
       }
       if ((res = mp_lshd (&t1, i - t - 1)) != MP_OKAY) {
-        goto __Y;
+        goto LBL_Y;
       }
       if ((res = mp_add (&x, &t1, &x)) != MP_OKAY) {
-        goto __Y;
+        goto LBL_Y;
       }
 
       q.dp[i - t - 1] = (q.dp[i - t - 1] - 1UL) & MP_MASK;
@@ -275,11 +275,11 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
 
   res = MP_OKAY;
 
-__Y:mp_clear (&y);
-__X:mp_clear (&x);
-__T2:mp_clear (&t2);
-__T1:mp_clear (&t1);
-__Q:mp_clear (&q);
+LBL_Y:mp_clear (&y);
+LBL_X:mp_clear (&x);
+LBL_T2:mp_clear (&t2);
+LBL_T1:mp_clear (&t1);
+LBL_Q:mp_clear (&q);
   return res;
 }
 
diff --git a/bn_mp_dr_reduce.c b/bn_mp_dr_reduce.c
index 308b80a..9bb7ad7 100644
--- a/bn_mp_dr_reduce.c
+++ b/bn_mp_dr_reduce.c
@@ -20,7 +20,7 @@
  * Based on algorithm from the paper
  *
  * "Generating Efficient Primes for Discrete Log Cryptosystems"
- *                 Chae Hoon Lim, Pil Loong Lee,
+ *                 Chae Hoon Lim, Pil Joong Lee,
  *          POSTECH Information Research Laboratories
  *
  * The modulus must be of a special format [see manual]
diff --git a/bn_mp_exptmod.c b/bn_mp_exptmod.c
index da88fec..7309170 100644
--- a/bn_mp_exptmod.c
+++ b/bn_mp_exptmod.c
@@ -61,7 +61,7 @@ int mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
      return err;
 #else 
      /* no invmod */
-     return MP_VAL
+     return MP_VAL;
 #endif
   }
 
diff --git a/bn_mp_exptmod_fast.c b/bn_mp_exptmod_fast.c
index 4351e60..255e9d9 100644
--- a/bn_mp_exptmod_fast.c
+++ b/bn_mp_exptmod_fast.c
@@ -88,11 +88,11 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
 #ifdef BN_MP_MONTGOMERY_SETUP_C     
      /* now setup montgomery  */
      if ((err = mp_montgomery_setup (P, &mp)) != MP_OKAY) {
-        goto __M;
+        goto LBL_M;
      }
 #else
      err = MP_VAL;
-     goto __M;
+     goto LBL_M;
 #endif
 
      /* automatically pick the comba one if available (saves quite a few calls/ifs) */
@@ -108,7 +108,7 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
         redux = mp_montgomery_reduce;
 #else
         err = MP_VAL;
-        goto __M;
+        goto LBL_M;
 #endif
      }
   } else if (redmode == 1) {
@@ -118,24 +118,24 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
      redux = mp_dr_reduce;
 #else
      err = MP_VAL;
-     goto __M;
+     goto LBL_M;
 #endif
   } else {
 #if defined(BN_MP_REDUCE_2K_SETUP_C) && defined(BN_MP_REDUCE_2K_C)
      /* setup DR reduction for moduli of the form 2**k - b */
      if ((err = mp_reduce_2k_setup(P, &mp)) != MP_OKAY) {
-        goto __M;
+        goto LBL_M;
      }
      redux = mp_reduce_2k;
 #else
      err = MP_VAL;
-     goto __M;
+     goto LBL_M;
 #endif
   }
 
   /* setup result */
   if ((err = mp_init (&res)) != MP_OKAY) {
-    goto __M;
+    goto LBL_M;
   }
 
   /* create M table
@@ -149,45 +149,45 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
 #ifdef BN_MP_MONTGOMERY_CALC_NORMALIZATION_C
      /* now we need R mod m */
      if ((err = mp_montgomery_calc_normalization (&res, P)) != MP_OKAY) {
-       goto __RES;
+       goto LBL_RES;
      }
 #else 
      err = MP_VAL;
-     goto __RES;
+     goto LBL_RES;
 #endif
 
      /* now set M[1] to G * R mod m */
      if ((err = mp_mulmod (G, &res, P, &M[1])) != MP_OKAY) {
-       goto __RES;
+       goto LBL_RES;
      }
   } else {
      mp_set(&res, 1);
      if ((err = mp_mod(G, P, &M[1])) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
      }
   }
 
   /* compute the value at M[1<<(winsize-1)] by squaring M[1] (winsize-1) times */
   if ((err = mp_copy (&M[1], &M[1 << (winsize - 1)])) != MP_OKAY) {
-    goto __RES;
+    goto LBL_RES;
   }
 
   for (x = 0; x < (winsize - 1); x++) {
     if ((err = mp_sqr (&M[1 << (winsize - 1)], &M[1 << (winsize - 1)])) != MP_OKAY) {
-      goto __RES;
+      goto LBL_RES;
     }
     if ((err = redux (&M[1 << (winsize - 1)], P, mp)) != MP_OKAY) {
-      goto __RES;
+      goto LBL_RES;
     }
   }
 
   /* create upper table */
   for (x = (1 << (winsize - 1)) + 1; x < (1 << winsize); x++) {
     if ((err = mp_mul (&M[x - 1], &M[1], &M[x])) != MP_OKAY) {
-      goto __RES;
+      goto LBL_RES;
     }
     if ((err = redux (&M[x], P, mp)) != MP_OKAY) {
-      goto __RES;
+      goto LBL_RES;
     }
   }
 
@@ -227,10 +227,10 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
     /* if the bit is zero and mode == 1 then we square */
     if (mode == 1 && y == 0) {
       if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = redux (&res, P, mp)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       continue;
     }
@@ -244,19 +244,19 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
       /* square first */
       for (x = 0; x < winsize; x++) {
         if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
         if ((err = redux (&res, P, mp)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
       }
 
       /* then multiply */
       if ((err = mp_mul (&res, &M[bitbuf], &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = redux (&res, P, mp)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
 
       /* empty window and reset */
@@ -271,10 +271,10 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
     /* square then multiply if the bit is set */
     for (x = 0; x < bitcpy; x++) {
       if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = redux (&res, P, mp)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
 
       /* get next bit of the window */
@@ -282,10 +282,10 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
       if ((bitbuf & (1 << winsize)) != 0) {
         /* then multiply */
         if ((err = mp_mul (&res, &M[1], &res)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
         if ((err = redux (&res, P, mp)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
       }
     }
@@ -299,15 +299,15 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
       * of R.
       */
      if ((err = redux(&res, P, mp)) != MP_OKAY) {
-       goto __RES;
+       goto LBL_RES;
      }
   }
 
   /* swap res with Y */
   mp_exch (&res, Y);
   err = MP_OKAY;
-__RES:mp_clear (&res);
-__M:
+LBL_RES:mp_clear (&res);
+LBL_M:
   mp_clear(&M[1]);
   for (x = 1<<(winsize-1); x < (1 << winsize); x++) {
     mp_clear (&M[x]);
diff --git a/bn_mp_gcd.c b/bn_mp_gcd.c
index 1cd21fc..6265df1 100644
--- a/bn_mp_gcd.c
+++ b/bn_mp_gcd.c
@@ -43,7 +43,7 @@ int mp_gcd (mp_int * a, mp_int * b, mp_int * c)
   }
 
   if ((res = mp_init_copy (&v, b)) != MP_OKAY) {
-    goto __U;
+    goto LBL_U;
   }
 
   /* must be positive for the remainder of the algorithm */
@@ -57,24 +57,24 @@ int mp_gcd (mp_int * a, mp_int * b, mp_int * c)
   if (k > 0) {
      /* divide the power of two out */
      if ((res = mp_div_2d(&u, k, &u, NULL)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      }
 
      if ((res = mp_div_2d(&v, k, &v, NULL)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      }
   }
 
   /* divide any remaining factors of two out */
   if (u_lsb != k) {
      if ((res = mp_div_2d(&u, u_lsb - k, &u, NULL)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      }
   }
 
   if (v_lsb != k) {
      if ((res = mp_div_2d(&v, v_lsb - k, &v, NULL)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      }
   }
 
@@ -87,23 +87,23 @@ int mp_gcd (mp_int * a, mp_int * b, mp_int * c)
      
      /* subtract smallest from largest */
      if ((res = s_mp_sub(&v, &u, &v)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      }
      
      /* Divide out all factors of two */
      if ((res = mp_div_2d(&v, mp_cnt_lsb(&v), &v, NULL)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      } 
   } 
 
   /* multiply by 2**k which we divided out at the beginning */
   if ((res = mp_mul_2d (&u, k, c)) != MP_OKAY) {
-     goto __V;
+     goto LBL_V;
   }
   c->sign = MP_ZPOS;
   res = MP_OKAY;
-__V:mp_clear (&u);
-__U:mp_clear (&v);
+LBL_V:mp_clear (&u);
+LBL_U:mp_clear (&v);
   return res;
 }
 #endif
diff --git a/bn_mp_invmod_slow.c b/bn_mp_invmod_slow.c
index 8ecb009..c1884c0 100644
--- a/bn_mp_invmod_slow.c
+++ b/bn_mp_invmod_slow.c
@@ -34,24 +34,24 @@ int mp_invmod_slow (mp_int * a, mp_int * b, mp_int * c)
 
   /* x = a, y = b */
   if ((res = mp_copy (a, &x)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
   if ((res = mp_copy (b, &y)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* 2. [modified] if x,y are both even then return an error! */
   if (mp_iseven (&x) == 1 && mp_iseven (&y) == 1) {
     res = MP_VAL;
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* 3. u=x, v=y, A=1, B=0, C=0,D=1 */
   if ((res = mp_copy (&x, &u)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
   if ((res = mp_copy (&y, &v)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
   mp_set (&A, 1);
   mp_set (&D, 1);
@@ -61,24 +61,24 @@ top:
   while (mp_iseven (&u) == 1) {
     /* 4.1 u = u/2 */
     if ((res = mp_div_2 (&u, &u)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     /* 4.2 if A or B is odd then */
     if (mp_isodd (&A) == 1 || mp_isodd (&B) == 1) {
       /* A = (A+y)/2, B = (B-x)/2 */
       if ((res = mp_add (&A, &y, &A)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
       if ((res = mp_sub (&B, &x, &B)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
     }
     /* A = A/2, B = B/2 */
     if ((res = mp_div_2 (&A, &A)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     if ((res = mp_div_2 (&B, &B)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -86,24 +86,24 @@ top:
   while (mp_iseven (&v) == 1) {
     /* 5.1 v = v/2 */
     if ((res = mp_div_2 (&v, &v)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     /* 5.2 if C or D is odd then */
     if (mp_isodd (&C) == 1 || mp_isodd (&D) == 1) {
       /* C = (C+y)/2, D = (D-x)/2 */
       if ((res = mp_add (&C, &y, &C)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
       if ((res = mp_sub (&D, &x, &D)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
     }
     /* C = C/2, D = D/2 */
     if ((res = mp_div_2 (&C, &C)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     if ((res = mp_div_2 (&D, &D)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -111,28 +111,28 @@ top:
   if (mp_cmp (&u, &v) != MP_LT) {
     /* u = u - v, A = A - C, B = B - D */
     if ((res = mp_sub (&u, &v, &u)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&A, &C, &A)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&B, &D, &B)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   } else {
     /* v - v - u, C = C - A, D = D - B */
     if ((res = mp_sub (&v, &u, &v)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&C, &A, &C)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&D, &B, &D)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -145,27 +145,27 @@ top:
   /* if v != 1 then there is no inverse */
   if (mp_cmp_d (&v, 1) != MP_EQ) {
     res = MP_VAL;
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* if its too low */
   while (mp_cmp_d(&C, 0) == MP_LT) {
       if ((res = mp_add(&C, b, &C)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
   }
   
   /* too big */
   while (mp_cmp_mag(&C, b) != MP_LT) {
       if ((res = mp_sub(&C, b, &C)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
   }
   
   /* C is now the inverse */
   mp_exch (&C, c);
   res = MP_OKAY;
-__ERR:mp_clear_multi (&x, &y, &u, &v, &A, &B, &C, &D, NULL);
+LBL_ERR:mp_clear_multi (&x, &y, &u, &v, &A, &B, &C, &D, NULL);
   return res;
 }
 #endif
diff --git a/bn_mp_jacobi.c b/bn_mp_jacobi.c
index 1c69cfd..74cbbf3 100644
--- a/bn_mp_jacobi.c
+++ b/bn_mp_jacobi.c
@@ -50,13 +50,13 @@ int mp_jacobi (mp_int * a, mp_int * p, int *c)
   }
 
   if ((res = mp_init (&p1)) != MP_OKAY) {
-    goto __A1;
+    goto LBL_A1;
   }
 
   /* divide out larger power of two */
   k = mp_cnt_lsb(&a1);
   if ((res = mp_div_2d(&a1, k, &a1, NULL)) != MP_OKAY) {
-     goto __P1;
+     goto LBL_P1;
   }
 
   /* step 4.  if e is even set s=1 */
@@ -84,18 +84,18 @@ int mp_jacobi (mp_int * a, mp_int * p, int *c)
   } else {
     /* n1 = n mod a1 */
     if ((res = mp_mod (p, &a1, &p1)) != MP_OKAY) {
-      goto __P1;
+      goto LBL_P1;
     }
     if ((res = mp_jacobi (&p1, &a1, &r)) != MP_OKAY) {
-      goto __P1;
+      goto LBL_P1;
     }
     *c = s * r;
   }
 
   /* done */
   res = MP_OKAY;
-__P1:mp_clear (&p1);
-__A1:mp_clear (&a1);
+LBL_P1:mp_clear (&p1);
+LBL_A1:mp_clear (&a1);
   return res;
 }
 #endif
diff --git a/bn_mp_lcm.c b/bn_mp_lcm.c
index 340d757..8e3a759 100644
--- a/bn_mp_lcm.c
+++ b/bn_mp_lcm.c
@@ -28,20 +28,20 @@ int mp_lcm (mp_int * a, mp_int * b, mp_int * c)
 
   /* t1 = get the GCD of the two inputs */
   if ((res = mp_gcd (a, b, &t1)) != MP_OKAY) {
-    goto __T;
+    goto LBL_T;
   }
 
   /* divide the smallest by the GCD */
   if (mp_cmp_mag(a, b) == MP_LT) {
      /* store quotient in t2 such that t2 * b is the LCM */
      if ((res = mp_div(a, &t1, &t2, NULL)) != MP_OKAY) {
-        goto __T;
+        goto LBL_T;
      }
      res = mp_mul(b, &t2, c);
   } else {
      /* store quotient in t2 such that t2 * a is the LCM */
      if ((res = mp_div(b, &t1, &t2, NULL)) != MP_OKAY) {
-        goto __T;
+        goto LBL_T;
      }
      res = mp_mul(a, &t2, c);
   }
@@ -49,7 +49,7 @@ int mp_lcm (mp_int * a, mp_int * b, mp_int * c)
   /* fix the sign to positive */
   c->sign = MP_ZPOS;
 
-__T:
+LBL_T:
   mp_clear_multi (&t1, &t2, NULL);
   return res;
 }
diff --git a/bn_mp_mod_2d.c b/bn_mp_mod_2d.c
index f81a0d4..589e4ba 100644
--- a/bn_mp_mod_2d.c
+++ b/bn_mp_mod_2d.c
@@ -28,7 +28,7 @@ mp_mod_2d (mp_int * a, int b, mp_int * c)
   }
 
   /* if the modulus is larger than the value than return */
-  if (b > (int) (a->used * DIGIT_BIT)) {
+  if (b >= (int) (a->used * DIGIT_BIT)) {
     res = mp_copy (a, c);
     return res;
   }
diff --git a/bn_mp_n_root.c b/bn_mp_n_root.c
index 9489903..7b11aa2 100644
--- a/bn_mp_n_root.c
+++ b/bn_mp_n_root.c
@@ -40,11 +40,11 @@ int mp_n_root (mp_int * a, mp_digit b, mp_int * c)
   }
 
   if ((res = mp_init (&t2)) != MP_OKAY) {
-    goto __T1;
+    goto LBL_T1;
   }
 
   if ((res = mp_init (&t3)) != MP_OKAY) {
-    goto __T2;
+    goto LBL_T2;
   }
 
   /* if a is negative fudge the sign but keep track */
@@ -57,52 +57,52 @@ int mp_n_root (mp_int * a, mp_digit b, mp_int * c)
   do {
     /* t1 = t2 */
     if ((res = mp_copy (&t2, &t1)) != MP_OKAY) {
-      goto __T3;
+      goto LBL_T3;
     }
 
     /* t2 = t1 - ((t1**b - a) / (b * t1**(b-1))) */
     
     /* t3 = t1**(b-1) */
     if ((res = mp_expt_d (&t1, b - 1, &t3)) != MP_OKAY) {   
-      goto __T3;
+      goto LBL_T3;
     }
 
     /* numerator */
     /* t2 = t1**b */
     if ((res = mp_mul (&t3, &t1, &t2)) != MP_OKAY) {    
-      goto __T3;
+      goto LBL_T3;
     }
 
     /* t2 = t1**b - a */
     if ((res = mp_sub (&t2, a, &t2)) != MP_OKAY) {  
-      goto __T3;
+      goto LBL_T3;
     }
 
     /* denominator */
     /* t3 = t1**(b-1) * b  */
     if ((res = mp_mul_d (&t3, b, &t3)) != MP_OKAY) {    
-      goto __T3;
+      goto LBL_T3;
     }
 
     /* t3 = (t1**b - a)/(b * t1**(b-1)) */
     if ((res = mp_div (&t2, &t3, &t3, NULL)) != MP_OKAY) {  
-      goto __T3;
+      goto LBL_T3;
     }
 
     if ((res = mp_sub (&t1, &t3, &t2)) != MP_OKAY) {
-      goto __T3;
+      goto LBL_T3;
     }
   }  while (mp_cmp (&t1, &t2) != MP_EQ);
 
   /* result can be off by a few so check */
   for (;;) {
     if ((res = mp_expt_d (&t1, b, &t2)) != MP_OKAY) {
-      goto __T3;
+      goto LBL_T3;
     }
 
     if (mp_cmp (&t2, a) == MP_GT) {
       if ((res = mp_sub_d (&t1, 1, &t1)) != MP_OKAY) {
-         goto __T3;
+         goto LBL_T3;
       }
     } else {
       break;
@@ -120,9 +120,9 @@ int mp_n_root (mp_int * a, mp_digit b, mp_int * c)
 
   res = MP_OKAY;
 
-__T3:mp_clear (&t3);
-__T2:mp_clear (&t2);
-__T1:mp_clear (&t1);
+LBL_T3:mp_clear (&t3);
+LBL_T2:mp_clear (&t2);
+LBL_T1:mp_clear (&t1);
   return res;
 }
 #endif
diff --git a/bn_mp_prime_fermat.c b/bn_mp_prime_fermat.c
index fe17aaa..fd74dbe 100644
--- a/bn_mp_prime_fermat.c
+++ b/bn_mp_prime_fermat.c
@@ -43,7 +43,7 @@ int mp_prime_fermat (mp_int * a, mp_int * b, int *result)
 
   /* compute t = b**a mod a */
   if ((err = mp_exptmod (b, a, a, &t)) != MP_OKAY) {
-    goto __T;
+    goto LBL_T;
   }
 
   /* is it equal to b? */
@@ -52,7 +52,7 @@ int mp_prime_fermat (mp_int * a, mp_int * b, int *result)
   }
 
   err = MP_OKAY;
-__T:mp_clear (&t);
+LBL_T:mp_clear (&t);
   return err;
 }
 #endif
diff --git a/bn_mp_prime_is_divisible.c b/bn_mp_prime_is_divisible.c
index 22ec1ae..f85fe7c 100644
--- a/bn_mp_prime_is_divisible.c
+++ b/bn_mp_prime_is_divisible.c
@@ -29,8 +29,8 @@ int mp_prime_is_divisible (mp_int * a, int *result)
   *result = MP_NO;
 
   for (ix = 0; ix < PRIME_SIZE; ix++) {
-    /* what is a mod __prime_tab[ix] */
-    if ((err = mp_mod_d (a, __prime_tab[ix], &res)) != MP_OKAY) {
+    /* what is a mod LBL_prime_tab[ix] */
+    if ((err = mp_mod_d (a, ltm_prime_tab[ix], &res)) != MP_OKAY) {
       return err;
     }
 
diff --git a/bn_mp_prime_is_prime.c b/bn_mp_prime_is_prime.c
index c2354d2..188053a 100644
--- a/bn_mp_prime_is_prime.c
+++ b/bn_mp_prime_is_prime.c
@@ -37,7 +37,7 @@ int mp_prime_is_prime (mp_int * a, int t, int *result)
 
   /* is the input equal to one of the primes in the table? */
   for (ix = 0; ix < PRIME_SIZE; ix++) {
-      if (mp_cmp_d(a, __prime_tab[ix]) == MP_EQ) {
+      if (mp_cmp_d(a, ltm_prime_tab[ix]) == MP_EQ) {
          *result = 1;
          return MP_OKAY;
       }
@@ -60,20 +60,20 @@ int mp_prime_is_prime (mp_int * a, int t, int *result)
 
   for (ix = 0; ix < t; ix++) {
     /* set the prime */
-    mp_set (&b, __prime_tab[ix]);
+    mp_set (&b, ltm_prime_tab[ix]);
 
     if ((err = mp_prime_miller_rabin (a, &b, &res)) != MP_OKAY) {
-      goto __B;
+      goto LBL_B;
     }
 
     if (res == MP_NO) {
-      goto __B;
+      goto LBL_B;
     }
   }
 
   /* passed the test */
   *result = MP_YES;
-__B:mp_clear (&b);
+LBL_B:mp_clear (&b);
   return err;
 }
 #endif
diff --git a/bn_mp_prime_miller_rabin.c b/bn_mp_prime_miller_rabin.c
index 22dec2f..758a2c3 100644
--- a/bn_mp_prime_miller_rabin.c
+++ b/bn_mp_prime_miller_rabin.c
@@ -40,12 +40,12 @@ int mp_prime_miller_rabin (mp_int * a, mp_int * b, int *result)
     return err;
   }
   if ((err = mp_sub_d (&n1, 1, &n1)) != MP_OKAY) {
-    goto __N1;
+    goto LBL_N1;
   }
 
   /* set 2**s * r = n1 */
   if ((err = mp_init_copy (&r, &n1)) != MP_OKAY) {
-    goto __N1;
+    goto LBL_N1;
   }
 
   /* count the number of least significant bits
@@ -55,15 +55,15 @@ int mp_prime_miller_rabin (mp_int * a, mp_int * b, int *result)
 
   /* now divide n - 1 by 2**s */
   if ((err = mp_div_2d (&r, s, &r, NULL)) != MP_OKAY) {
-    goto __R;
+    goto LBL_R;
   }
 
   /* compute y = b**r mod a */
   if ((err = mp_init (&y)) != MP_OKAY) {
-    goto __R;
+    goto LBL_R;
   }
   if ((err = mp_exptmod (b, &r, a, &y)) != MP_OKAY) {
-    goto __Y;
+    goto LBL_Y;
   }
 
   /* if y != 1 and y != n1 do */
@@ -72,12 +72,12 @@ int mp_prime_miller_rabin (mp_int * a, mp_int * b, int *result)
     /* while j <= s-1 and y != n1 */
     while ((j <= (s - 1)) && mp_cmp (&y, &n1) != MP_EQ) {
       if ((err = mp_sqrmod (&y, a, &y)) != MP_OKAY) {
-         goto __Y;
+         goto LBL_Y;
       }
 
       /* if y == 1 then composite */
       if (mp_cmp_d (&y, 1) == MP_EQ) {
-         goto __Y;
+         goto LBL_Y;
       }
 
       ++j;
@@ -85,15 +85,15 @@ int mp_prime_miller_rabin (mp_int * a, mp_int * b, int *result)
 
     /* if y != n1 then composite */
     if (mp_cmp (&y, &n1) != MP_EQ) {
-      goto __Y;
+      goto LBL_Y;
     }
   }
 
   /* probably prime now */
   *result = MP_YES;
-__Y:mp_clear (&y);
-__R:mp_clear (&r);
-__N1:mp_clear (&n1);
+LBL_Y:mp_clear (&y);
+LBL_R:mp_clear (&r);
+LBL_N1:mp_clear (&n1);
   return err;
 }
 #endif
diff --git a/bn_mp_prime_next_prime.c b/bn_mp_prime_next_prime.c
index c478ce5..24f93c4 100644
--- a/bn_mp_prime_next_prime.c
+++ b/bn_mp_prime_next_prime.c
@@ -35,10 +35,10 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
    a->sign = MP_ZPOS;
 
    /* simple algo if a is less than the largest prime in the table */
-   if (mp_cmp_d(a, __prime_tab[PRIME_SIZE-1]) == MP_LT) {
+   if (mp_cmp_d(a, ltm_prime_tab[PRIME_SIZE-1]) == MP_LT) {
       /* find which prime it is bigger than */
       for (x = PRIME_SIZE - 2; x >= 0; x--) {
-          if (mp_cmp_d(a, __prime_tab[x]) != MP_LT) {
+          if (mp_cmp_d(a, ltm_prime_tab[x]) != MP_LT) {
              if (bbs_style == 1) {
                 /* ok we found a prime smaller or
                  * equal [so the next is larger]
@@ -46,17 +46,17 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
                  * however, the prime must be
                  * congruent to 3 mod 4
                  */
-                if ((__prime_tab[x + 1] & 3) != 3) {
+                if ((ltm_prime_tab[x + 1] & 3) != 3) {
                    /* scan upwards for a prime congruent to 3 mod 4 */
                    for (y = x + 1; y < PRIME_SIZE; y++) {
-                       if ((__prime_tab[y] & 3) == 3) {
-                          mp_set(a, __prime_tab[y]);
+                       if ((ltm_prime_tab[y] & 3) == 3) {
+                          mp_set(a, ltm_prime_tab[y]);
                           return MP_OKAY;
                        }
                    }
                 }
              } else {
-                mp_set(a, __prime_tab[x + 1]);
+                mp_set(a, ltm_prime_tab[x + 1]);
                 return MP_OKAY;
              }
           }
@@ -94,7 +94,7 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
 
    /* generate the restable */
    for (x = 1; x < PRIME_SIZE; x++) {
-      if ((err = mp_mod_d(a, __prime_tab[x], res_tab + x)) != MP_OKAY) {
+      if ((err = mp_mod_d(a, ltm_prime_tab[x], res_tab + x)) != MP_OKAY) {
          return err;
       }
    }
@@ -120,8 +120,8 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
              res_tab[x] += kstep;
 
              /* subtract the modulus [instead of using division] */
-             if (res_tab[x] >= __prime_tab[x]) {
-                res_tab[x]  -= __prime_tab[x];
+             if (res_tab[x] >= ltm_prime_tab[x]) {
+                res_tab[x]  -= ltm_prime_tab[x];
              }
 
              /* set flag if zero */
@@ -133,7 +133,7 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
 
       /* add the step */
       if ((err = mp_add_d(a, step, a)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
 
       /* if didn't pass sieve and step == MAX then skip test */
@@ -143,9 +143,9 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
 
       /* is this prime? */
       for (x = 0; x < t; x++) {
-          mp_set(&b, __prime_tab[t]);
+          mp_set(&b, ltm_prime_tab[t]);
           if ((err = mp_prime_miller_rabin(a, &b, &res)) != MP_OKAY) {
-             goto __ERR;
+             goto LBL_ERR;
           }
           if (res == MP_NO) {
              break;
@@ -158,7 +158,7 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
    }
 
    err = MP_OKAY;
-__ERR:
+LBL_ERR:
    mp_clear(&b);
    return err;
 }
diff --git a/bn_mp_prime_random_ex.c b/bn_mp_prime_random_ex.c
index 2c4f4f0..2010ebe 100644
--- a/bn_mp_prime_random_ex.c
+++ b/bn_mp_prime_random_ex.c
@@ -47,7 +47,7 @@ int mp_prime_random_ex(mp_int *a, int t, int size, int flags, ltm_prime_callback
    }
 
    /* calc the byte size */
-   bsize = (size>>3)+(size&7?1:0);
+   bsize = (size>>3) + ((size&7)?1:0);
 
    /* we need a buffer of bsize bytes */
    tmp = OPT_CAST(unsigned char) XMALLOC(bsize);
@@ -56,7 +56,7 @@ int mp_prime_random_ex(mp_int *a, int t, int size, int flags, ltm_prime_callback
    }
 
    /* calc the maskAND value for the MSbyte*/
-   maskAND = 0xFF >> (8 - (size & 7));
+   maskAND = ((size&7) == 0) ? 0xFF : (0xFF >> (8 - (size & 7)));
 
    /* calc the maskOR_msb */
    maskOR_msb        = 0;
@@ -65,7 +65,7 @@ int mp_prime_random_ex(mp_int *a, int t, int size, int flags, ltm_prime_callback
       maskOR_msb     |= 1 << ((size - 2) & 7);
    } else if (flags & LTM_PRIME_2MSB_OFF) {
       maskAND        &= ~(1 << ((size - 2) & 7));
-   }
+   } 
 
    /* get the maskOR_lsb */
    maskOR_lsb         = 0;
diff --git a/bn_prime_tab.c b/bn_prime_tab.c
index 18ecc47..14306c2 100644
--- a/bn_prime_tab.c
+++ b/bn_prime_tab.c
@@ -14,7 +14,7 @@
  *
  * Tom St Denis, tomstdenis@iahu.ca, http://math.libtomcrypt.org
  */
-const mp_digit __prime_tab[] = {
+const mp_digit ltm_prime_tab[] = {
   0x0002, 0x0003, 0x0005, 0x0007, 0x000B, 0x000D, 0x0011, 0x0013,
   0x0017, 0x001D, 0x001F, 0x0025, 0x0029, 0x002B, 0x002F, 0x0035,
   0x003B, 0x003D, 0x0043, 0x0047, 0x0049, 0x004F, 0x0053, 0x0059,
diff --git a/bn_s_mp_exptmod.c b/bn_s_mp_exptmod.c
index 4f1032a..01a766f 100644
--- a/bn_s_mp_exptmod.c
+++ b/bn_s_mp_exptmod.c
@@ -70,10 +70,10 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
 
   /* create mu, used for Barrett reduction */
   if ((err = mp_init (&mu)) != MP_OKAY) {
-    goto __M;
+    goto LBL_M;
   }
   if ((err = mp_reduce_setup (&mu, P)) != MP_OKAY) {
-    goto __MU;
+    goto LBL_MU;
   }
 
   /* create M table
@@ -85,23 +85,23 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
    * computed though accept for M[0] and M[1]
    */
   if ((err = mp_mod (G, P, &M[1])) != MP_OKAY) {
-    goto __MU;
+    goto LBL_MU;
   }
 
   /* compute the value at M[1<<(winsize-1)] by squaring 
    * M[1] (winsize-1) times 
    */
   if ((err = mp_copy (&M[1], &M[1 << (winsize - 1)])) != MP_OKAY) {
-    goto __MU;
+    goto LBL_MU;
   }
 
   for (x = 0; x < (winsize - 1); x++) {
     if ((err = mp_sqr (&M[1 << (winsize - 1)], 
                        &M[1 << (winsize - 1)])) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
     if ((err = mp_reduce (&M[1 << (winsize - 1)], P, &mu)) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
   }
 
@@ -110,16 +110,16 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
    */
   for (x = (1 << (winsize - 1)) + 1; x < (1 << winsize); x++) {
     if ((err = mp_mul (&M[x - 1], &M[1], &M[x])) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
     if ((err = mp_reduce (&M[x], P, &mu)) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
   }
 
   /* setup result */
   if ((err = mp_init (&res)) != MP_OKAY) {
-    goto __MU;
+    goto LBL_MU;
   }
   mp_set (&res, 1);
 
@@ -159,10 +159,10 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
     /* if the bit is zero and mode == 1 then we square */
     if (mode == 1 && y == 0) {
       if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       continue;
     }
@@ -176,19 +176,19 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
       /* square first */
       for (x = 0; x < winsize; x++) {
         if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
         if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
       }
 
       /* then multiply */
       if ((err = mp_mul (&res, &M[bitbuf], &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
 
       /* empty window and reset */
@@ -203,20 +203,20 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
     /* square then multiply if the bit is set */
     for (x = 0; x < bitcpy; x++) {
       if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
 
       bitbuf <<= 1;
       if ((bitbuf & (1 << winsize)) != 0) {
         /* then multiply */
         if ((err = mp_mul (&res, &M[1], &res)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
         if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
       }
     }
@@ -224,9 +224,9 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
 
   mp_exch (&res, Y);
   err = MP_OKAY;
-__RES:mp_clear (&res);
-__MU:mp_clear (&mu);
-__M:
+LBL_RES:mp_clear (&res);
+LBL_MU:mp_clear (&mu);
+LBL_M:
   mp_clear(&M[1]);
   for (x = 1<<(winsize-1); x < (1 << winsize); x++) {
     mp_clear (&M[x]);
diff --git a/callgraph.txt b/callgraph.txt
index 56d4f8b..4dc4cba 100644
--- a/callgraph.txt
+++ b/callgraph.txt
@@ -245,6 +245,7 @@ BN_MP_SQRT_C
 |   |   +--->BN_MP_INIT_MULTI_C
 |   |   |   +--->BN_MP_CLEAR_C
 |   |   +--->BN_MP_COUNT_BITS_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -298,6 +299,7 @@ BN_MP_SQRT_C
 |   |   +--->BN_MP_CLEAR_C
 |   +--->BN_MP_SET_C
 |   +--->BN_MP_COUNT_BITS_C
+|   +--->BN_MP_ABS_C
 |   +--->BN_MP_MUL_2D_C
 |   |   +--->BN_MP_GROW_C
 |   |   +--->BN_MP_LSHD_C
@@ -404,6 +406,7 @@ BN_MP_IS_SQUARE_C
 |   |   |   +--->BN_MP_CLEAR_C
 |   |   +--->BN_MP_SET_C
 |   |   +--->BN_MP_COUNT_BITS_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -700,6 +703,7 @@ BN_MP_IS_SQUARE_C
 |   |   |   +--->BN_MP_INIT_MULTI_C
 |   |   |   |   +--->BN_MP_CLEAR_C
 |   |   |   +--->BN_MP_COUNT_BITS_C
+|   |   |   +--->BN_MP_ABS_C
 |   |   |   +--->BN_MP_MUL_2D_C
 |   |   |   |   +--->BN_MP_GROW_C
 |   |   |   |   +--->BN_MP_LSHD_C
@@ -753,6 +757,7 @@ BN_MP_IS_SQUARE_C
 |   |   |   +--->BN_MP_CLEAR_C
 |   |   +--->BN_MP_SET_C
 |   |   +--->BN_MP_COUNT_BITS_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -2618,6 +2623,7 @@ BN_MP_SUBMOD_C
 |   |   +--->BN_MP_INIT_MULTI_C
 |   |   +--->BN_MP_SET_C
 |   |   +--->BN_MP_COUNT_BITS_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -2838,6 +2844,7 @@ BN_MP_SQRMOD_C
 |   |   +--->BN_MP_INIT_MULTI_C
 |   |   +--->BN_MP_SET_C
 |   |   +--->BN_MP_COUNT_BITS_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -3313,6 +3320,7 @@ BN_MP_N_ROOT_C
 |   +--->BN_MP_INIT_MULTI_C
 |   |   +--->BN_MP_CLEAR_C
 |   +--->BN_MP_COUNT_BITS_C
+|   +--->BN_MP_ABS_C
 |   +--->BN_MP_MUL_2D_C
 |   |   +--->BN_MP_GROW_C
 |   |   +--->BN_MP_LSHD_C
@@ -4322,6 +4330,7 @@ BN_MP_PRIME_RANDOM_EX_C
 |   |   |   |   |   +--->BN_MP_ZERO_C
 |   |   |   |   |   +--->BN_MP_INIT_MULTI_C
 |   |   |   |   |   +--->BN_MP_COUNT_BITS_C
+|   |   |   |   |   +--->BN_MP_ABS_C
 |   |   |   |   |   +--->BN_MP_MUL_2D_C
 |   |   |   |   |   |   +--->BN_MP_GROW_C
 |   |   |   |   |   |   +--->BN_MP_LSHD_C
@@ -4548,6 +4557,7 @@ BN_MP_MOD_C
 |   |   +--->BN_MP_CLEAR_C
 |   +--->BN_MP_SET_C
 |   +--->BN_MP_COUNT_BITS_C
+|   +--->BN_MP_ABS_C
 |   +--->BN_MP_MUL_2D_C
 |   |   +--->BN_MP_GROW_C
 |   |   +--->BN_MP_LSHD_C
@@ -5600,6 +5610,7 @@ BN_MP_PRIME_IS_PRIME_C
 |   |   |   |   +--->BN_MP_ZERO_C
 |   |   |   |   +--->BN_MP_INIT_MULTI_C
 |   |   |   |   +--->BN_MP_COUNT_BITS_C
+|   |   |   |   +--->BN_MP_ABS_C
 |   |   |   |   +--->BN_MP_MUL_2D_C
 |   |   |   |   |   +--->BN_MP_GROW_C
 |   |   |   |   |   +--->BN_MP_LSHD_C
@@ -5809,6 +5820,7 @@ BN_MP_EXPTMOD_FAST_C
 |   |   |   +--->BN_MP_ZERO_C
 |   |   |   +--->BN_MP_INIT_MULTI_C
 |   |   |   +--->BN_MP_SET_C
+|   |   |   +--->BN_MP_ABS_C
 |   |   |   +--->BN_MP_MUL_2D_C
 |   |   |   |   +--->BN_MP_GROW_C
 |   |   |   |   +--->BN_MP_LSHD_C
@@ -5865,6 +5877,7 @@ BN_MP_EXPTMOD_FAST_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   +--->BN_MP_ZERO_C
 |   |   +--->BN_MP_INIT_MULTI_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -6284,6 +6297,7 @@ BN_MP_MULMOD_C
 |   |   +--->BN_MP_INIT_MULTI_C
 |   |   +--->BN_MP_SET_C
 |   |   +--->BN_MP_COUNT_BITS_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -7339,6 +7353,7 @@ BN_MP_PRIME_NEXT_PRIME_C
 |   |   |   |   +--->BN_MP_ZERO_C
 |   |   |   |   +--->BN_MP_INIT_MULTI_C
 |   |   |   |   +--->BN_MP_COUNT_BITS_C
+|   |   |   |   +--->BN_MP_ABS_C
 |   |   |   |   +--->BN_MP_MUL_2D_C
 |   |   |   |   |   +--->BN_MP_GROW_C
 |   |   |   |   |   +--->BN_MP_LSHD_C
@@ -7465,6 +7480,7 @@ BN_MP_LCM_C
 |   +--->BN_MP_ZERO_C
 |   +--->BN_MP_SET_C
 |   +--->BN_MP_COUNT_BITS_C
+|   +--->BN_MP_ABS_C
 |   +--->BN_MP_MUL_2D_C
 |   |   +--->BN_MP_GROW_C
 |   |   +--->BN_MP_LSHD_C
@@ -7928,6 +7944,7 @@ BN_S_MP_EXPTMOD_C
 |   |   +--->BN_MP_ZERO_C
 |   |   +--->BN_MP_INIT_MULTI_C
 |   |   +--->BN_MP_SET_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -7974,6 +7991,7 @@ BN_S_MP_EXPTMOD_C
 |   |   +--->BN_MP_ZERO_C
 |   |   +--->BN_MP_INIT_MULTI_C
 |   |   +--->BN_MP_SET_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -8372,6 +8390,7 @@ BN_MP_DIV_C
 |   +--->BN_MP_CLEAR_C
 +--->BN_MP_SET_C
 +--->BN_MP_COUNT_BITS_C
++--->BN_MP_ABS_C
 +--->BN_MP_MUL_2D_C
 |   +--->BN_MP_GROW_C
 |   +--->BN_MP_LSHD_C
@@ -8465,6 +8484,7 @@ BN_MP_ADDMOD_C
 |   |   +--->BN_MP_INIT_MULTI_C
 |   |   +--->BN_MP_SET_C
 |   |   +--->BN_MP_COUNT_BITS_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -8551,6 +8571,7 @@ BN_MP_REDUCE_C
 |   |   |   +--->BN_MP_CLEAR_C
 |   |   +--->BN_MP_SET_C
 |   |   +--->BN_MP_COUNT_BITS_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -8766,6 +8787,7 @@ BN_MP_JACOBI_C
 |   |   |   +--->BN_MP_CLEAR_C
 |   |   +--->BN_MP_SET_C
 |   |   +--->BN_MP_COUNT_BITS_C
+|   |   +--->BN_MP_ABS_C
 |   |   +--->BN_MP_MUL_2D_C
 |   |   |   +--->BN_MP_GROW_C
 |   |   |   +--->BN_MP_LSHD_C
@@ -8912,6 +8934,7 @@ BN_MP_EXTEUCLID_C
 |   +--->BN_MP_CMP_MAG_C
 |   +--->BN_MP_ZERO_C
 |   +--->BN_MP_COUNT_BITS_C
+|   +--->BN_MP_ABS_C
 |   +--->BN_MP_MUL_2D_C
 |   |   +--->BN_MP_GROW_C
 |   |   +--->BN_MP_LSHD_C
@@ -9078,6 +9101,7 @@ BN_MP_REDUCE_SETUP_C
 |   |   +--->BN_MP_CLEAR_C
 |   +--->BN_MP_SET_C
 |   +--->BN_MP_COUNT_BITS_C
+|   +--->BN_MP_ABS_C
 |   +--->BN_MP_MUL_2D_C
 |   |   +--->BN_MP_GROW_C
 |   |   +--->BN_MP_LSHD_C
@@ -10118,6 +10142,7 @@ BN_MP_PRIME_MILLER_RABIN_C
 |   |   |   +--->BN_MP_INIT_MULTI_C
 |   |   |   +--->BN_MP_SET_C
 |   |   |   +--->BN_MP_COUNT_BITS_C
+|   |   |   +--->BN_MP_ABS_C
 |   |   |   +--->BN_MP_MUL_2D_C
 |   |   |   |   +--->BN_MP_GROW_C
 |   |   |   |   +--->BN_MP_LSHD_C
diff --git a/changes.txt b/changes.txt
index 6a86209..0d1ec2e 100644
--- a/changes.txt
+++ b/changes.txt
@@ -1,3 +1,12 @@
+December 23rd, 2004
+v0.33  -- Fixed "small" variant for mp_div() which would munge with negative dividends...
+       -- Fixed bug in mp_prime_random_ex() which would set the most significant byte to zero when
+          no special flags were set
+       -- Fixed overflow [minor] bug in fast_s_mp_sqr()
+       -- Made the makefiles easier to configure the group/user that ltm will install as
+       -- Fixed "final carry" bug in comba multipliers. (Volkan Ceylan)
+       -- Matt Johnston pointed out a missing semi-colon in mp_exptmod
+
 October 29th, 2004
 v0.32  -- Added "makefile.shared" for shared object support
        -- Added more to the build options/configs in the manual
diff --git a/demo/demo.c b/demo/demo.c
index 53eb3cf..62615cd 100644
--- a/demo/demo.c
+++ b/demo/demo.c
@@ -11,9 +11,9 @@
 
 void ndraw(mp_int *a, char *name)
 {
-   char buf[4096];
+   char buf[16000];
    printf("%s: ", name);
-   mp_toradix(a, buf, 64);
+   mp_toradix(a, buf, 10);
    printf("%s\n", buf);
 }
 
@@ -395,7 +395,7 @@ draw(&a);draw(&b);draw(&c);draw(&d);
 
           mp_div(&a, &b, &e, &f);
           if (mp_cmp(&c, &e) != MP_EQ || mp_cmp(&d, &f) != MP_EQ) {
-             printf("div %lu failure!\n", div_n);
+             printf("div %lu %d, %d, failure!\n", div_n, mp_cmp(&c, &e), mp_cmp(&d, &f));
 draw(&a);draw(&b);draw(&c);draw(&d); draw(&e); draw(&f);
              return 0;
           }
diff --git a/demo/timing.c b/demo/timing.c
index 865c444..7b27d53 100644
--- a/demo/timing.c
+++ b/demo/timing.c
@@ -38,14 +38,13 @@ int lbit(void)
    }
 }
 
-#if defined(__i386__) || defined(_M_IX86) || defined(_M_AMD64)
 /* RDTSC from Scott Duplichan */
 static ulong64 TIMFUNC (void)
    {
    #if defined __GNUC__
-      #ifdef __i386__
-         ulong64 a;
-         __asm__ __volatile__ ("rdtsc ":"=A" (a));
+      #if defined(__i386__) || defined(__x86_64__)
+         unsigned long long a;
+         __asm__ __volatile__ ("rdtsc\nmovl %%eax,%0\nmovl %%edx,4+%0\n"::"m"(a):"%eax","%edx");
          return a;
       #else /* gcc-IA64 version */
          unsigned long result;
@@ -69,9 +68,6 @@ static ulong64 TIMFUNC (void)
      #error need rdtsc function for this build
    #endif
    }
-#else
-#define TIMFUNC clock
-#endif
 
 #define DO(x) x; x;
 //#define DO4(x) DO2(x); DO2(x);
diff --git a/etc/mersenne.c b/etc/mersenne.c
index da6c111..1cd5b50 100644
--- a/etc/mersenne.c
+++ b/etc/mersenne.c
@@ -18,15 +18,15 @@ is_mersenne (long s, int *pp)
   }
 
   if ((res = mp_init (&u)) != MP_OKAY) {
-    goto __N;
+    goto LBL_N;
   }
 
   /* n = 2^s - 1 */
   if ((res = mp_2expt(&n, s)) != MP_OKAY) {
-     goto __MU;
+     goto LBL_MU;
   }
   if ((res = mp_sub_d (&n, 1, &n)) != MP_OKAY) {
-    goto __MU;
+    goto LBL_MU;
   }
 
   /* set u=4 */
@@ -36,22 +36,22 @@ is_mersenne (long s, int *pp)
   for (k = 1; k <= s - 2; k++) {
     /* u = u^2 - 2 mod n */
     if ((res = mp_sqr (&u, &u)) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
     if ((res = mp_sub_d (&u, 2, &u)) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
 
     /* make sure u is positive */
     while (u.sign == MP_NEG) {
       if ((res = mp_add (&u, &n, &u)) != MP_OKAY) {
-         goto __MU;
+         goto LBL_MU;
       }
     }
 
     /* reduce */
     if ((res = mp_reduce_2k (&u, &n, 1)) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
   }
 
@@ -62,8 +62,8 @@ is_mersenne (long s, int *pp)
   }
 
   res = MP_OKAY;
-__MU:mp_clear (&u);
-__N:mp_clear (&n);
+LBL_MU:mp_clear (&u);
+LBL_N:mp_clear (&n);
   return res;
 }
 
diff --git a/etc/pprime.c b/etc/pprime.c
index cccb748..26e0d84 100644
--- a/etc/pprime.c
+++ b/etc/pprime.c
@@ -189,7 +189,7 @@ pprime (int k, int li, mp_int * p, mp_int * q)
   }
 
   if ((res = mp_init (&v)) != MP_OKAY) {
-    goto __C;
+    goto LBL_C;
   }
 
   /* product of first 50 primes */
@@ -197,34 +197,34 @@ pprime (int k, int li, mp_int * p, mp_int * q)
        mp_read_radix (&v,
 		      "19078266889580195013601891820992757757219839668357012055907516904309700014933909014729740190",
 		      10)) != MP_OKAY) {
-    goto __V;
+    goto LBL_V;
   }
 
   if ((res = mp_init (&a)) != MP_OKAY) {
-    goto __V;
+    goto LBL_V;
   }
 
   /* set the prime */
   mp_set (&a, prime_digit ());
 
   if ((res = mp_init (&b)) != MP_OKAY) {
-    goto __A;
+    goto LBL_A;
   }
 
   if ((res = mp_init (&n)) != MP_OKAY) {
-    goto __B;
+    goto LBL_B;
   }
 
   if ((res = mp_init (&x)) != MP_OKAY) {
-    goto __N;
+    goto LBL_N;
   }
 
   if ((res = mp_init (&y)) != MP_OKAY) {
-    goto __X;
+    goto LBL_X;
   }
 
   if ((res = mp_init (&z)) != MP_OKAY) {
-    goto __Y;
+    goto LBL_Y;
   }
 
   /* now loop making the single digit */
@@ -236,25 +236,25 @@ pprime (int k, int li, mp_int * p, mp_int * q)
 
     /* now compute z = a * b * 2 */
     if ((res = mp_mul (&a, &b, &z)) != MP_OKAY) {	/* z = a * b */
-      goto __Z;
+      goto LBL_Z;
     }
 
     if ((res = mp_copy (&z, &c)) != MP_OKAY) {	/* c = a * b */
-      goto __Z;
+      goto LBL_Z;
     }
 
     if ((res = mp_mul_2 (&z, &z)) != MP_OKAY) {	/* z = 2 * a * b */
-      goto __Z;
+      goto LBL_Z;
     }
 
     /* n = z + 1 */
     if ((res = mp_add_d (&z, 1, &n)) != MP_OKAY) {	/* n = z + 1 */
-      goto __Z;
+      goto LBL_Z;
     }
 
     /* check (n, v) == 1 */
     if ((res = mp_gcd (&n, &v, &y)) != MP_OKAY) {	/* y = (n, v) */
-      goto __Z;
+      goto LBL_Z;
     }
 
     if (mp_cmp_d (&y, 1) != MP_EQ)
@@ -266,7 +266,7 @@ pprime (int k, int li, mp_int * p, mp_int * q)
 
       /* compute x^a mod n */
       if ((res = mp_exptmod (&x, &a, &n, &y)) != MP_OKAY) {	/* y = x^a mod n */
-	goto __Z;
+	goto LBL_Z;
       }
 
       /* if y == 1 loop */
@@ -275,7 +275,7 @@ pprime (int k, int li, mp_int * p, mp_int * q)
 
       /* now x^2a mod n */
       if ((res = mp_sqrmod (&y, &n, &y)) != MP_OKAY) {	/* y = x^2a mod n */
-	goto __Z;
+	goto LBL_Z;
       }
 
       if (mp_cmp_d (&y, 1) == MP_EQ)
@@ -283,7 +283,7 @@ pprime (int k, int li, mp_int * p, mp_int * q)
 
       /* compute x^b mod n */
       if ((res = mp_exptmod (&x, &b, &n, &y)) != MP_OKAY) {	/* y = x^b mod n */
-	goto __Z;
+	goto LBL_Z;
       }
 
       /* if y == 1 loop */
@@ -292,7 +292,7 @@ pprime (int k, int li, mp_int * p, mp_int * q)
 
       /* now x^2b mod n */
       if ((res = mp_sqrmod (&y, &n, &y)) != MP_OKAY) {	/* y = x^2b mod n */
-	goto __Z;
+	goto LBL_Z;
       }
 
       if (mp_cmp_d (&y, 1) == MP_EQ)
@@ -300,7 +300,7 @@ pprime (int k, int li, mp_int * p, mp_int * q)
 
       /* compute x^c mod n == x^ab mod n */
       if ((res = mp_exptmod (&x, &c, &n, &y)) != MP_OKAY) {	/* y = x^ab mod n */
-	goto __Z;
+	goto LBL_Z;
       }
 
       /* if y == 1 loop */
@@ -309,7 +309,7 @@ pprime (int k, int li, mp_int * p, mp_int * q)
 
       /* now compute (x^c mod n)^2 */
       if ((res = mp_sqrmod (&y, &n, &y)) != MP_OKAY) {	/* y = x^2ab mod n */
-	goto __Z;
+	goto LBL_Z;
       }
 
       /* y should be 1 */
@@ -346,14 +346,14 @@ pprime (int k, int li, mp_int * p, mp_int * q)
   mp_exch (&n, p);
 
   res = MP_OKAY;
-__Z:mp_clear (&z);
-__Y:mp_clear (&y);
-__X:mp_clear (&x);
-__N:mp_clear (&n);
-__B:mp_clear (&b);
-__A:mp_clear (&a);
-__V:mp_clear (&v);
-__C:mp_clear (&c);
+LBL_Z:mp_clear (&z);
+LBL_Y:mp_clear (&y);
+LBL_X:mp_clear (&x);
+LBL_N:mp_clear (&n);
+LBL_B:mp_clear (&b);
+LBL_A:mp_clear (&a);
+LBL_V:mp_clear (&v);
+LBL_C:mp_clear (&c);
   return res;
 }
 
diff --git a/etc/tune.c b/etc/tune.c
index bc101be..14aace2 100644
--- a/etc/tune.c
+++ b/etc/tune.c
@@ -14,9 +14,9 @@
 #ifndef X86_TIMER
 
 /* generic ISO C timer */
-ulong64 __T;
-void t_start(void) { __T = clock(); }
-ulong64 t_read(void) { return clock() - __T; }
+ulong64 LBL_T;
+void t_start(void) { LBL_T = clock(); }
+ulong64 t_read(void) { return clock() - LBL_T; }
 
 #else
 extern void t_start(void);
diff --git a/logs/add.log b/logs/add.log
index d44c4cd..fa11039 100644
--- a/logs/add.log
+++ b/logs/add.log
@@ -1,16 +1,16 @@
-224       222
-448       330
-672       436
-896       520
-1120       612
-1344       696
-1568       810
-1792       912
-2016      1006
-2240      1116
-2464      1152
-2688      1284
-2912      1348
-3136      1486
-3360      1580
-3584      1636
+480        88
+960       113
+1440       138
+1920       163
+2400       202
+2880       226
+3360       251
+3840       272
+4320       296
+4800       320
+5280       344
+5760       368
+6240       392
+6720       416
+7200       440
+7680       464
diff --git a/logs/expt.log b/logs/expt.log
index e69de29..e65e927 100644
--- a/logs/expt.log
+++ b/logs/expt.log
@@ -0,0 +1,7 @@
+513   1499509
+769   3682671
+1025   8098887
+2049  49332743
+2561  89647783
+3073 149440713
+4097 326135364
diff --git a/logs/expt_2k.log b/logs/expt_2k.log
index e69de29..d106280 100644
--- a/logs/expt_2k.log
+++ b/logs/expt_2k.log
@@ -0,0 +1,6 @@
+521   1423346
+607   1841305
+1279   8375656
+2203  34104708
+3217  83830729
+4253 167916804
diff --git a/logs/expt_dr.log b/logs/expt_dr.log
index e69de29..6cfc874 100644
--- a/logs/expt_dr.log
+++ b/logs/expt_dr.log
@@ -0,0 +1,7 @@
+532   1803110
+784   3607375
+1036   6089790
+1540  14739797
+2072  33251589
+3080  82794331
+4116 165212734
diff --git a/logs/mult.log b/logs/mult.log
index a2c9c18..864de46 100644
--- a/logs/mult.log
+++ b/logs/mult.log
@@ -1,143 +1,143 @@
-140      1272
-195      1428
-252      1996
-307      2586
-364      3464
-420      4420
-476      5260
-532      6430
-588      7692
-644      8704
-699     10226
-755     11670
-812     13190
-865     14834
-924     16738
-979     18362
-1036     20660
-1092     22776
-1148     24848
-1204     27168
-1260     29930
-1316     32258
-1370     35172
-1422     37534
-1482     40390
-1537     43990
-1589     46946
-1652     50438
-1703     52902
-1764     56646
-1820     59892
-1876     63248
-1932     66872
-1988     72596
-2042     74662
-2100     78512
-2156     82944
-2211     87444
-2268     92170
-2324     95534
-2380    100484
-2435    105024
-2491    109460
-2546    114154
-2603    118946
-2660    124110
-2716    129300
-2771    134274
-2828    139594
-2883    145234
-2939    150332
-2996    155750
-3048    161718
-3108    167492
-3162    173882
-3219    179766
-3276    185560
-3330    191826
-3388    197822
-3442    204176
-3500    210682
-3556    217236
-3612    223484
-3666    230714
-3724    237744
-3779    244080
-3835    250970
-3890    257914
-3947    265162
-4001    272128
-4060    279108
-4116    287606
-4171    294716
-4227    302806
-4284    310260
-4340    318564
-4395    326164
-4443    334034
-4508    342108
-4561    351810
-4618    358828
-4675    367332
-4732    376140
-4787    384172
-4841    393308
-4899    402036
-4955    411286
-5010    420290
-5067    429688
-5124    438810
-5180    448130
-5235    457264
-5290    467390
-5348    476586
-5404    486120
-5459    496512
-5516    506624
-5569    516346
-5628    526604
-5684    536544
-5740    546936
-5796    557284
-5852    568106
-5907    578824
-5963    589204
-6019    600176
-6076    610564
-6127    621972
-6188    633564
-6244    644730
-6300    655288
-6354    667402
-6412    678824
-6467    690594
-6522    702718
-6580    714148
-6636    725608
-6690    737834
-6747    750100
-6804    762202
-6860    774184
-6916    787298
-6971    798734
-7028    811162
-7083    824570
-7139    837738
-7196   2579488
-7245   2626714
-7308   2643582
-7364   2698746
-7416   2734106
-7476   2773372
-7530   2816738
-7588   2859204
-7643   2938596
-7698   2919716
-7754   2988542
-7812   3026520
-7867   3058304
-7924   3115790
-7977   3161450
-8035   3203138
-8092   3244056
+271       580
+390       861
+511      1177
+630      1598
+749      2115
+871      2670
+991      3276
+1111      3987
+1231      4722
+1351      5474
+1471      6281
+1589      7126
+1710      8114
+1831      8988
+1946     10038
+2071     10995
+2188     12286
+2310     13152
+2430     14480
+2549     15521
+2671     17171
+2790     18081
+2911     19754
+3031     20809
+3150     22849
+3269     23757
+3391     25772
+3508     26832
+3631     29304
+3750     30149
+3865     32581
+3988     33644
+4111     36565
+4231     37309
+4351     40152
+4471     41188
+4590     44658
+4710     45256
+4827     48538
+4951     49490
+5070     53472
+5190     53902
+5308     57619
+5431     58509
+5550     63044
+5664     63333
+5791     67542
+5911     68279
+6028     73477
+6150     73475
+6271     78189
+6390     78842
+6510     84691
+6631     84444
+6751     89721
+6871     90186
+6991     96665
+7111     96119
+7231    101937
+7350    102212
+7471    109439
+7591    108491
+7709    114965
+7829    115025
+7951    123002
+8071    121630
+8190    128725
+8311    128536
+8430    137298
+8550    135568
+8671    143265
+8791    142793
+8911    152432
+9030    150202
+9151    158616
+9271    157848
+9391    168374
+9511    165651
+9627    174775
+9750    173375
+9871    185067
+9985    181845
+10111    191708
+10229    190239
+10351    202585
+10467    198704
+10591    209193
+10711    207322
+10831    220842
+10950    215882
+11071    227761
+11191    225501
+11311    239669
+11430    234809
+11550    243511
+11671    255947
+11791    255243
+11906    267828
+12029    263437
+12149    276571
+12270    275579
+12390    288963
+12510    284001
+12631    298196
+12751    297018
+12869    310848
+12990    305369
+13111    319086
+13230    318940
+13349    333685
+13471    327495
+13588    343678
+13711    341817
+13831    357181
+13948    350440
+14071    367526
+14189    365330
+14311    381551
+14429    374149
+14549    392203
+14670    389764
+14791    406761
+14910    398652
+15026    417718
+15150    414733
+15269    432759
+15390   1037071
+15511   1053454
+15631   1069198
+15748   1086164
+15871   1112820
+15991   1129676
+16111   1145924
+16230   1163016
+16345   1179911
+16471   1197048
+16586   1214352
+16711   1232095
+16829   1249338
+16947   1266987
+17071   1284181
+17188   1302521
+17311   1320539
diff --git a/logs/sqr.log b/logs/sqr.log
index 3e175ac..0898342 100644
--- a/logs/sqr.log
+++ b/logs/sqr.log
@@ -1,143 +1,143 @@
-139       806
-195      1212
-252      1604
-307      2260
-364      2892
-420      3308
-476      4152
-532      4814
-588      5754
-644      6684
-700      7226
-756      8324
-808      9092
-866     10068
-924     11204
-976     12918
-1036     13656
-1092     15248
-1148     15956
-1204     17270
-1260     19894
-1316     20516
-1370     21864
-1428     25554
-1483     26138
-1540     27086
-1596     29246
-1652     32210
-1707     32704
-1764     35142
-1820     39050
-1876     39256
-1931     41574
-1985     45070
-2044     46352
-2099     48114
-2155     51332
-2212     53268
-2267     55890
-2324     59054
-2380     60206
-2434     63540
-2491     66084
-2547     68590
-2604     74332
-2660     74784
-2715     77974
-2772     79924
-2826     82914
-2884     87210
-2929     89076
-2996     92480
-3052     96814
-3108     99990
-3162    102550
-3219    105396
-3276    109284
-3332    113752
-3387    116628
-3444    120782
-3500    122938
-3556    127940
-3612    303656
-3667    312212
-3724    324376
-3779    329204
-3833    340910
-3892    353850
-3943    362348
-4003    367780
-4056    380448
-4114    393616
-4172    404104
-4227    415148
-4284    409770
-4339    436648
-4394    442970
-4451    463096
-4507    472056
-4564    485780
-4616    496286
-4675    507612
-4732    519524
-4788    536768
-4843    542754
-4899    553090
-4956    571986
-5012    586340
-5068    599606
-5124    613670
-5179    624256
-5235    636266
-5292    655518
-5348    668142
-5403    677266
-5460    696040
-5516    712772
-5570    723942
-5628    739052
-5684    755350
-5739    769962
-5790    775258
-5851    790128
-5908    814536
-5962    827278
-6018    844510
-6076    851606
-6130    865748
-6188    894752
-6244    900474
-6300    928174
-6356    928440
-6410    957758
-6468    981134
-6524    994088
-6580   1011124
-6636   1027178
-6692   1045466
-6747   1056910
-6804   1083784
-6860   1104706
-6915   1116450
-6972   1137894
-7028   1154670
-7084   1158064
-7138   1188734
-7196   1214218
-7249   1226822
-7307   1247528
-7363   1255338
-7420   1291104
-7475   1297940
-7532   1324994
-7587   1340274
-7644   1342596
-7698   1381418
-7756   1382904
-7812   1432588
-7867   1443632
-7922   1465092
-7979   1496804
-8036   1520142
-8092   1539566
+271       552
+389       883
+510      1191
+629      1572
+750      1996
+863      2428
+991      2891
+1108      3539
+1231      4182
+1351      4980
+1471      5771
+1590      6551
+1711      7313
+1830      8240
+1951      9184
+2070     10087
+2191     11140
+2311     12111
+2431     13219
+2550     14247
+2669     15353
+2791     16446
+2911     17692
+3029     18848
+3151     20028
+3268     21282
+3391     22696
+3511     23971
+3631     25303
+3751     26675
+3871     28245
+3990     29736
+4111     31124
+4229     32714
+4347     34397
+4471     35877
+4587     37269
+4710     39011
+4831     40884
+4950     42501
+5070     44005
+5191     46026
+5310     48168
+5431     49801
+5551     51385
+5671     53604
+5787     55942
+5910     57757
+6031     59391
+6151     61754
+6271     64234
+6390     66110
+6511     67845
+6627     70474
+6751     73113
+6871     75064
+6990     76940
+7111     79681
+7230     82548
+7351     84597
+7471     86507
+7591     89497
+7711    225216
+7831    232192
+7951    239583
+8071    247302
+8191    255497
+8308    261587
+8431    271490
+8550    279492
+8671    286927
+8790    294680
+8910    302974
+9030    311300
+9150    318635
+9271    326740
+9390    335304
+9511    344297
+9630    352056
+9748    358652
+9870    369723
+9991    379119
+10111    386982
+10231    396075
+10349    404396
+10470    415375
+10590    424146
+10711    433390
+10829    442662
+10950    453238
+11071    462178
+11186    469811
+11311    482529
+11431    493214
+11550    503210
+11671    513486
+11791    524244
+11911    535277
+12031    544872
+12151    555695
+12271    566893
+12391    578385
+12510    588658
+12628    596914
+12751    611324
+12871    623437
+12991    633907
+13110    645605
+13231    657684
+13351    670037
+13471    680939
+13591    693047
+13710    705363
+13829    718178
+13949    727930
+14069    739641
+14190    754817
+14310    768192
+14431    779875
+14551    792655
+14667    802847
+14791    819806
+14911    831684
+15031    844936
+15151    858813
+15270    873037
+15387    882123
+15510    899117
+15631    913465
+15750    927989
+15870    940790
+15991    954948
+16110    969483
+16231    984544
+16350    997837
+16470   1012445
+16590   1027834
+16710   1043032
+16831   1056394
+16951   1071408
+17069   1097263
+17191   1113364
+17306   1123650
diff --git a/logs/sub.log b/logs/sub.log
index cf2bcd6..a42d91e 100644
--- a/logs/sub.log
+++ b/logs/sub.log
@@ -1,16 +1,16 @@
-224       216
-448       324
-672       428
-896       532
-1120       648
-1344       766
-1568       862
-1792       928
-2016      1070
-2240      1128
-2464      1250
-2688      1344
-2912      1436
-3136      1542
-3360      1628
-3584      1696
+480        87
+960       114
+1440       139
+1920       159
+2400       204
+2880       228
+3360       250
+3840       273
+4320       300
+4800       321
+5280       348
+5760       370
+6240       393
+6720       420
+7200       444
+7680       466
diff --git a/makefile b/makefile
index 4fe2256..164a0ab 100644
--- a/makefile
+++ b/makefile
@@ -1,10 +1,14 @@
 #Makefile for GCC
 #
 #Tom St Denis
+
+#version of library 
+VERSION=0.33
+
 CFLAGS  +=  -I./ -Wall -W -Wshadow -Wsign-compare
 
 #for speed 
-CFLAGS += -O3 -funroll-loops
+CFLAGS += -O3 -funroll-all-loops
 
 #for size 
 #CFLAGS += -Os
@@ -15,13 +19,15 @@ CFLAGS  += -fomit-frame-pointer
 #debug
 #CFLAGS += -g3
 
-VERSION=0.32
+#install as this user
+USER=root
+GROUP=root
 
 default: libtommath.a
 
 #default files to install
 LIBNAME=libtommath.a
-HEADERS=tommath.h
+HEADERS=tommath.h tommath_class.h tommath_superclass.h
 
 #LIBPATH-The directory for libtommath to be installed to.
 #INCPATH-The directory to install the header files for libtommath.
@@ -61,7 +67,6 @@ libtommath.a:  $(OBJECTS)
 	$(AR) $(ARFLAGS) libtommath.a $(OBJECTS)
 	ranlib libtommath.a
 
-
 #make a profiled library (takes a while!!!)
 #
 # This will build the library with profile generation
@@ -86,19 +91,19 @@ profiled_single:
 	ranlib libtommath.a	
 
 install: libtommath.a
-	install -d -g root -o root $(DESTDIR)$(LIBPATH)
-	install -d -g root -o root $(DESTDIR)$(INCPATH)
-	install -g root -o root $(LIBNAME) $(DESTDIR)$(LIBPATH)
-	install -g root -o root $(HEADERS) $(DESTDIR)$(INCPATH)
+	install -d -g $(GROUP) -o $(USER) $(DESTDIR)$(LIBPATH)
+	install -d -g $(GROUP) -o $(USER) $(DESTDIR)$(INCPATH)
+	install -g $(GROUP) -o $(USER) $(LIBNAME) $(DESTDIR)$(LIBPATH)
+	install -g $(GROUP) -o $(USER) $(HEADERS) $(DESTDIR)$(INCPATH)
 
 test: libtommath.a demo/demo.o
-	$(CC) demo/demo.o libtommath.a -o test
+	$(CC) $(CFLAGS) demo/demo.o libtommath.a -o test
 	
 mtest: test	
-	cd mtest ; $(CC) $(CFLAGS) mtest.c -o mtest -s
+	cd mtest ; $(CC) $(CFLAGS) mtest.c -o mtest
         
 timing: libtommath.a
-	$(CC) $(CFLAGS) -DTIMER demo/timing.c libtommath.a -o ltmtest -s
+	$(CC) $(CFLAGS) -DTIMER demo/timing.c libtommath.a -o ltmtest
 
 # makes the LTM book DVI file, requires tetex, perl and makeindex [part of tetex I think]
 docdvi: tommath.src
diff --git a/makefile.icc b/makefile.icc
index 09117b7..3775b20 100644
--- a/makefile.icc
+++ b/makefile.icc
@@ -21,6 +21,10 @@ CFLAGS  +=  -I./
 # Default to just generic max opts
 CFLAGS += -O3 -xN
 
+#install as this user
+USER=root
+GROUP=root
+
 default: libtommath.a
 
 #default files to install
@@ -89,10 +93,10 @@ profiled_single:
 	ranlib libtommath.a	
 
 install: libtommath.a
-	install -d -g root -o root $(DESTDIR)$(LIBPATH)
-	install -d -g root -o root $(DESTDIR)$(INCPATH)
-	install -g root -o root $(LIBNAME) $(DESTDIR)$(LIBPATH)
-	install -g root -o root $(HEADERS) $(DESTDIR)$(INCPATH)
+	install -d -g $(GROUP) -o $(USER) $(DESTDIR)$(LIBPATH)
+	install -d -g $(GROUP) -o $(USER) $(DESTDIR)$(INCPATH)
+	install -g $(GROUP) -o $(USER) $(LIBNAME) $(DESTDIR)$(LIBPATH)
+	install -g $(GROUP) -o $(USER) $(HEADERS) $(DESTDIR)$(INCPATH)
 
 test: libtommath.a demo/demo.o
 	$(CC) demo/demo.o libtommath.a -o test
diff --git a/makefile.shared b/makefile.shared
index 96bbf32..86a3786 100644
--- a/makefile.shared
+++ b/makefile.shared
@@ -1,10 +1,9 @@
 #Makefile for GCC
 #
 #Tom St Denis
-VERSION=0:32
+VERSION=0:33
 
 CC = libtool --mode=compile gcc
-
 CFLAGS  +=  -I./ -Wall -W -Wshadow -Wsign-compare
 
 #for speed 
@@ -16,11 +15,15 @@ CFLAGS += -O3 -funroll-loops
 #x86 optimizations [should be valid for any GCC install though]
 CFLAGS  += -fomit-frame-pointer
 
+#install as this user
+USER=root
+GROUP=root
+
 default: libtommath.la
 
 #default files to install
 LIBNAME=libtommath.la
-HEADERS=tommath.h
+HEADERS=tommath.h tommath_class.h tommath_superclass.h
 
 #LIBPATH-The directory for libtommath to be installed to.
 #INCPATH-The directory to install the header files for libtommath.
@@ -60,8 +63,8 @@ libtommath.la:  $(OBJECTS)
 	libtool --mode=link gcc *.lo -o libtommath.la -rpath $(LIBPATH) -version-info $(VERSION)
 	libtool --mode=link gcc *.o -o libtommath.a 
 	libtool --mode=install install -c libtommath.la $(LIBPATH)/libtommath.la
-	install -d -g root -o root $(DESTDIR)$(INCPATH)
-	install -g root -o root $(HEADERS) $(DESTDIR)$(INCPATH)
+	install -d -g $(GROUP) -o $(USER) $(DESTDIR)$(INCPATH)
+	install -g $(GROUP) -o $(USER) $(HEADERS) $(DESTDIR)$(INCPATH)
 
 test: libtommath.a demo/demo.o
 	gcc $(CFLAGS) -c demo/demo.c -o demo/demo.o
diff --git a/mtest/mtest.c b/mtest/mtest.c
index ef0e093..d46f456 100644
--- a/mtest/mtest.c
+++ b/mtest/mtest.c
@@ -46,7 +46,7 @@ void rand_num(mp_int *a)
    int n, size;
    unsigned char buf[2048];
 
-   size = 1 + ((fgetc(rng)<<8) + fgetc(rng)) % 1031;
+   size = 1 + ((fgetc(rng)<<8) + fgetc(rng)) % 101;
    buf[0] = (fgetc(rng)&1)?1:0;
    fread(buf+1, 1, size, rng);
    while (buf[1] == 0) buf[1] = fgetc(rng);
@@ -58,7 +58,7 @@ void rand_num2(mp_int *a)
    int n, size;
    unsigned char buf[2048];
 
-   size = 10 + ((fgetc(rng)<<8) + fgetc(rng)) % 97;
+   size = 10 + ((fgetc(rng)<<8) + fgetc(rng)) % 101;
    buf[0] = (fgetc(rng)&1)?1:0;
    fread(buf+1, 1, size, rng);
    while (buf[1] == 0) buf[1] = fgetc(rng);
diff --git a/poster.pdf b/poster.pdf
index 60999da..e0b4f84 100644
Binary files a/poster.pdf and b/poster.pdf differ
diff --git a/pre_gen/mpi.c b/pre_gen/mpi.c
index 78a73f0..7d832e7 100644
--- a/pre_gen/mpi.c
+++ b/pre_gen/mpi.c
@@ -87,20 +87,20 @@ fast_mp_invmod (mp_int * a, mp_int * b, mp_int * c)
 
   /* x == modulus, y == value to invert */
   if ((res = mp_copy (b, &x)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* we need y = |a| */
   if ((res = mp_abs (a, &y)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* 3. u=x, v=y, A=1, B=0, C=0,D=1 */
   if ((res = mp_copy (&x, &u)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
   if ((res = mp_copy (&y, &v)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
   mp_set (&D, 1);
 
@@ -109,17 +109,17 @@ top:
   while (mp_iseven (&u) == 1) {
     /* 4.1 u = u/2 */
     if ((res = mp_div_2 (&u, &u)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     /* 4.2 if B is odd then */
     if (mp_isodd (&B) == 1) {
       if ((res = mp_sub (&B, &x, &B)) != MP_OKAY) {
-        goto __ERR;
+        goto LBL_ERR;
       }
     }
     /* B = B/2 */
     if ((res = mp_div_2 (&B, &B)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -127,18 +127,18 @@ top:
   while (mp_iseven (&v) == 1) {
     /* 5.1 v = v/2 */
     if ((res = mp_div_2 (&v, &v)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     /* 5.2 if D is odd then */
     if (mp_isodd (&D) == 1) {
       /* D = (D-x)/2 */
       if ((res = mp_sub (&D, &x, &D)) != MP_OKAY) {
-        goto __ERR;
+        goto LBL_ERR;
       }
     }
     /* D = D/2 */
     if ((res = mp_div_2 (&D, &D)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -146,20 +146,20 @@ top:
   if (mp_cmp (&u, &v) != MP_LT) {
     /* u = u - v, B = B - D */
     if ((res = mp_sub (&u, &v, &u)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&B, &D, &B)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   } else {
     /* v - v - u, D = D - B */
     if ((res = mp_sub (&v, &u, &v)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&D, &B, &D)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -173,21 +173,21 @@ top:
   /* if v != 1 then there is no inverse */
   if (mp_cmp_d (&v, 1) != MP_EQ) {
     res = MP_VAL;
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* b is now the inverse */
   neg = a->sign;
   while (D.sign == MP_NEG) {
     if ((res = mp_add (&D, b, &D)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
   mp_exch (&D, c);
   c->sign = neg;
   res = MP_OKAY;
 
-__ERR:mp_clear_multi (&x, &y, &u, &v, &B, &D, NULL);
+LBL_ERR:mp_clear_multi (&x, &y, &u, &v, &B, &D, NULL);
   return res;
 }
 #endif
@@ -420,7 +420,7 @@ fast_s_mp_mul_digs (mp_int * a, mp_int * b, mp_int * c, int digs)
 
   /* clear the carry */
   _W = 0;
-  for (ix = 0; ix <= pa; ix++) { 
+  for (ix = 0; ix < pa; ix++) { 
       int      tx, ty;
       int      iy;
       mp_digit *tmpx, *tmpy;
@@ -450,6 +450,9 @@ fast_s_mp_mul_digs (mp_int * a, mp_int * b, mp_int * c, int digs)
       _W = _W >> ((mp_word)DIGIT_BIT);
   }
 
+  /* store final carry */
+  W[ix] = _W;
+
   /* setup dest */
   olduse  = c->used;
   c->used = digs;
@@ -519,7 +522,7 @@ fast_s_mp_mul_high_digs (mp_int * a, mp_int * b, mp_int * c, int digs)
   /* number of output digits to produce */
   pa = a->used + b->used;
   _W = 0;
-  for (ix = digs; ix <= pa; ix++) { 
+  for (ix = digs; ix < pa; ix++) { 
       int      tx, ty, iy;
       mp_digit *tmpx, *tmpy;
 
@@ -547,6 +550,9 @@ fast_s_mp_mul_high_digs (mp_int * a, mp_int * b, mp_int * c, int digs)
       /* make next carry */
       _W = _W >> ((mp_word)DIGIT_BIT);
   }
+  
+  /* store final carry */
+  W[ix] = _W;
 
   /* setup dest */
   olduse  = c->used;
@@ -636,7 +642,7 @@ int fast_s_mp_sqr (mp_int * a, mp_int * b)
 
   /* number of output digits to produce */
   W1 = 0;
-  for (ix = 0; ix <= pa; ix++) { 
+  for (ix = 0; ix < pa; ix++) { 
       int      tx, ty, iy;
       mp_word  _W;
       mp_digit *tmpy;
@@ -1539,23 +1545,23 @@ int mp_div(mp_int * a, mp_int * b, mp_int * c, mp_int * d)
 
   mp_set(&tq, 1);
   n = mp_count_bits(a) - mp_count_bits(b);
-  if (((res = mp_copy(a, &ta)) != MP_OKAY) ||
-      ((res = mp_copy(b, &tb)) != MP_OKAY) || 
+  if (((res = mp_abs(a, &ta)) != MP_OKAY) ||
+      ((res = mp_abs(b, &tb)) != MP_OKAY) || 
       ((res = mp_mul_2d(&tb, n, &tb)) != MP_OKAY) ||
       ((res = mp_mul_2d(&tq, n, &tq)) != MP_OKAY)) {
-      goto __ERR;
+      goto LBL_ERR;
   }
 
   while (n-- >= 0) {
      if (mp_cmp(&tb, &ta) != MP_GT) {
         if (((res = mp_sub(&ta, &tb, &ta)) != MP_OKAY) ||
             ((res = mp_add(&q, &tq, &q)) != MP_OKAY)) {
-           goto __ERR;
+           goto LBL_ERR;
         }
      }
      if (((res = mp_div_2d(&tb, 1, &tb, NULL)) != MP_OKAY) ||
          ((res = mp_div_2d(&tq, 1, &tq, NULL)) != MP_OKAY)) {
-           goto __ERR;
+           goto LBL_ERR;
      }
   }
 
@@ -1564,13 +1570,13 @@ int mp_div(mp_int * a, mp_int * b, mp_int * c, mp_int * d)
   n2 = (a->sign == b->sign ? MP_ZPOS : MP_NEG);
   if (c != NULL) {
      mp_exch(c, &q);
-     c->sign  = n2;
+     c->sign  = (mp_iszero(c) == MP_YES) ? MP_ZPOS : n2;
   }
   if (d != NULL) {
      mp_exch(d, &ta);
-     d->sign = n;
+     d->sign = (mp_iszero(d) == MP_YES) ? MP_ZPOS : n;
   }
-__ERR:
+LBL_ERR:
    mp_clear_multi(&ta, &tb, &tq, &q, NULL);
    return res;
 }
@@ -1619,19 +1625,19 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
   q.used = a->used + 2;
 
   if ((res = mp_init (&t1)) != MP_OKAY) {
-    goto __Q;
+    goto LBL_Q;
   }
 
   if ((res = mp_init (&t2)) != MP_OKAY) {
-    goto __T1;
+    goto LBL_T1;
   }
 
   if ((res = mp_init_copy (&x, a)) != MP_OKAY) {
-    goto __T2;
+    goto LBL_T2;
   }
 
   if ((res = mp_init_copy (&y, b)) != MP_OKAY) {
-    goto __X;
+    goto LBL_X;
   }
 
   /* fix the sign */
@@ -1643,10 +1649,10 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
   if (norm < (int)(DIGIT_BIT-1)) {
      norm = (DIGIT_BIT-1) - norm;
      if ((res = mp_mul_2d (&x, norm, &x)) != MP_OKAY) {
-       goto __Y;
+       goto LBL_Y;
      }
      if ((res = mp_mul_2d (&y, norm, &y)) != MP_OKAY) {
-       goto __Y;
+       goto LBL_Y;
      }
   } else {
      norm = 0;
@@ -1658,13 +1664,13 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
 
   /* while (x >= y*b**n-t) do { q[n-t] += 1; x -= y*b**{n-t} } */
   if ((res = mp_lshd (&y, n - t)) != MP_OKAY) { /* y = y*b**{n-t} */
-    goto __Y;
+    goto LBL_Y;
   }
 
   while (mp_cmp (&x, &y) != MP_LT) {
     ++(q.dp[n - t]);
     if ((res = mp_sub (&x, &y, &x)) != MP_OKAY) {
-      goto __Y;
+      goto LBL_Y;
     }
   }
 
@@ -1706,7 +1712,7 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
       t1.dp[1] = y.dp[t];
       t1.used = 2;
       if ((res = mp_mul_d (&t1, q.dp[i - t - 1], &t1)) != MP_OKAY) {
-        goto __Y;
+        goto LBL_Y;
       }
 
       /* find right hand */
@@ -1718,27 +1724,27 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
 
     /* step 3.3 x = x - q{i-t-1} * y * b**{i-t-1} */
     if ((res = mp_mul_d (&y, q.dp[i - t - 1], &t1)) != MP_OKAY) {
-      goto __Y;
+      goto LBL_Y;
     }
 
     if ((res = mp_lshd (&t1, i - t - 1)) != MP_OKAY) {
-      goto __Y;
+      goto LBL_Y;
     }
 
     if ((res = mp_sub (&x, &t1, &x)) != MP_OKAY) {
-      goto __Y;
+      goto LBL_Y;
     }
 
     /* if x < 0 then { x = x + y*b**{i-t-1}; q{i-t-1} -= 1; } */
     if (x.sign == MP_NEG) {
       if ((res = mp_copy (&y, &t1)) != MP_OKAY) {
-        goto __Y;
+        goto LBL_Y;
       }
       if ((res = mp_lshd (&t1, i - t - 1)) != MP_OKAY) {
-        goto __Y;
+        goto LBL_Y;
       }
       if ((res = mp_add (&x, &t1, &x)) != MP_OKAY) {
-        goto __Y;
+        goto LBL_Y;
       }
 
       q.dp[i - t - 1] = (q.dp[i - t - 1] - 1UL) & MP_MASK;
@@ -1765,11 +1771,11 @@ int mp_div (mp_int * a, mp_int * b, mp_int * c, mp_int * d)
 
   res = MP_OKAY;
 
-__Y:mp_clear (&y);
-__X:mp_clear (&x);
-__T2:mp_clear (&t2);
-__T1:mp_clear (&t1);
-__Q:mp_clear (&q);
+LBL_Y:mp_clear (&y);
+LBL_X:mp_clear (&x);
+LBL_T2:mp_clear (&t2);
+LBL_T1:mp_clear (&t1);
+LBL_Q:mp_clear (&q);
   return res;
 }
 
@@ -2199,7 +2205,7 @@ int mp_dr_is_modulus(mp_int *a)
  * Based on algorithm from the paper
  *
  * "Generating Efficient Primes for Discrete Log Cryptosystems"
- *                 Chae Hoon Lim, Pil Loong Lee,
+ *                 Chae Hoon Lim, Pil Joong Lee,
  *          POSTECH Information Research Laboratories
  *
  * The modulus must be of a special format [see manual]
@@ -2457,7 +2463,7 @@ int mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
      return err;
 #else 
      /* no invmod */
-     return MP_VAL
+     return MP_VAL;
 #endif
   }
 
@@ -2588,11 +2594,11 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
 #ifdef BN_MP_MONTGOMERY_SETUP_C     
      /* now setup montgomery  */
      if ((err = mp_montgomery_setup (P, &mp)) != MP_OKAY) {
-        goto __M;
+        goto LBL_M;
      }
 #else
      err = MP_VAL;
-     goto __M;
+     goto LBL_M;
 #endif
 
      /* automatically pick the comba one if available (saves quite a few calls/ifs) */
@@ -2608,7 +2614,7 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
         redux = mp_montgomery_reduce;
 #else
         err = MP_VAL;
-        goto __M;
+        goto LBL_M;
 #endif
      }
   } else if (redmode == 1) {
@@ -2618,24 +2624,24 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
      redux = mp_dr_reduce;
 #else
      err = MP_VAL;
-     goto __M;
+     goto LBL_M;
 #endif
   } else {
 #if defined(BN_MP_REDUCE_2K_SETUP_C) && defined(BN_MP_REDUCE_2K_C)
      /* setup DR reduction for moduli of the form 2**k - b */
      if ((err = mp_reduce_2k_setup(P, &mp)) != MP_OKAY) {
-        goto __M;
+        goto LBL_M;
      }
      redux = mp_reduce_2k;
 #else
      err = MP_VAL;
-     goto __M;
+     goto LBL_M;
 #endif
   }
 
   /* setup result */
   if ((err = mp_init (&res)) != MP_OKAY) {
-    goto __M;
+    goto LBL_M;
   }
 
   /* create M table
@@ -2649,45 +2655,45 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
 #ifdef BN_MP_MONTGOMERY_CALC_NORMALIZATION_C
      /* now we need R mod m */
      if ((err = mp_montgomery_calc_normalization (&res, P)) != MP_OKAY) {
-       goto __RES;
+       goto LBL_RES;
      }
 #else 
      err = MP_VAL;
-     goto __RES;
+     goto LBL_RES;
 #endif
 
      /* now set M[1] to G * R mod m */
      if ((err = mp_mulmod (G, &res, P, &M[1])) != MP_OKAY) {
-       goto __RES;
+       goto LBL_RES;
      }
   } else {
      mp_set(&res, 1);
      if ((err = mp_mod(G, P, &M[1])) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
      }
   }
 
   /* compute the value at M[1<<(winsize-1)] by squaring M[1] (winsize-1) times */
   if ((err = mp_copy (&M[1], &M[1 << (winsize - 1)])) != MP_OKAY) {
-    goto __RES;
+    goto LBL_RES;
   }
 
   for (x = 0; x < (winsize - 1); x++) {
     if ((err = mp_sqr (&M[1 << (winsize - 1)], &M[1 << (winsize - 1)])) != MP_OKAY) {
-      goto __RES;
+      goto LBL_RES;
     }
     if ((err = redux (&M[1 << (winsize - 1)], P, mp)) != MP_OKAY) {
-      goto __RES;
+      goto LBL_RES;
     }
   }
 
   /* create upper table */
   for (x = (1 << (winsize - 1)) + 1; x < (1 << winsize); x++) {
     if ((err = mp_mul (&M[x - 1], &M[1], &M[x])) != MP_OKAY) {
-      goto __RES;
+      goto LBL_RES;
     }
     if ((err = redux (&M[x], P, mp)) != MP_OKAY) {
-      goto __RES;
+      goto LBL_RES;
     }
   }
 
@@ -2727,10 +2733,10 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
     /* if the bit is zero and mode == 1 then we square */
     if (mode == 1 && y == 0) {
       if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = redux (&res, P, mp)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       continue;
     }
@@ -2744,19 +2750,19 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
       /* square first */
       for (x = 0; x < winsize; x++) {
         if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
         if ((err = redux (&res, P, mp)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
       }
 
       /* then multiply */
       if ((err = mp_mul (&res, &M[bitbuf], &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = redux (&res, P, mp)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
 
       /* empty window and reset */
@@ -2771,10 +2777,10 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
     /* square then multiply if the bit is set */
     for (x = 0; x < bitcpy; x++) {
       if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = redux (&res, P, mp)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
 
       /* get next bit of the window */
@@ -2782,10 +2788,10 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
       if ((bitbuf & (1 << winsize)) != 0) {
         /* then multiply */
         if ((err = mp_mul (&res, &M[1], &res)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
         if ((err = redux (&res, P, mp)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
       }
     }
@@ -2799,15 +2805,15 @@ mp_exptmod_fast (mp_int * G, mp_int * X, mp_int * P, mp_int * Y, int redmode)
       * of R.
       */
      if ((err = redux(&res, P, mp)) != MP_OKAY) {
-       goto __RES;
+       goto LBL_RES;
      }
   }
 
   /* swap res with Y */
   mp_exch (&res, Y);
   err = MP_OKAY;
-__RES:mp_clear (&res);
-__M:
+LBL_RES:mp_clear (&res);
+LBL_M:
   mp_clear(&M[1]);
   for (x = 1<<(winsize-1); x < (1 << winsize); x++) {
     mp_clear (&M[x]);
@@ -3059,7 +3065,7 @@ int mp_gcd (mp_int * a, mp_int * b, mp_int * c)
   }
 
   if ((res = mp_init_copy (&v, b)) != MP_OKAY) {
-    goto __U;
+    goto LBL_U;
   }
 
   /* must be positive for the remainder of the algorithm */
@@ -3073,24 +3079,24 @@ int mp_gcd (mp_int * a, mp_int * b, mp_int * c)
   if (k > 0) {
      /* divide the power of two out */
      if ((res = mp_div_2d(&u, k, &u, NULL)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      }
 
      if ((res = mp_div_2d(&v, k, &v, NULL)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      }
   }
 
   /* divide any remaining factors of two out */
   if (u_lsb != k) {
      if ((res = mp_div_2d(&u, u_lsb - k, &u, NULL)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      }
   }
 
   if (v_lsb != k) {
      if ((res = mp_div_2d(&v, v_lsb - k, &v, NULL)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      }
   }
 
@@ -3103,23 +3109,23 @@ int mp_gcd (mp_int * a, mp_int * b, mp_int * c)
      
      /* subtract smallest from largest */
      if ((res = s_mp_sub(&v, &u, &v)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      }
      
      /* Divide out all factors of two */
      if ((res = mp_div_2d(&v, mp_cnt_lsb(&v), &v, NULL)) != MP_OKAY) {
-        goto __V;
+        goto LBL_V;
      } 
   } 
 
   /* multiply by 2**k which we divided out at the beginning */
   if ((res = mp_mul_2d (&u, k, c)) != MP_OKAY) {
-     goto __V;
+     goto LBL_V;
   }
   c->sign = MP_ZPOS;
   res = MP_OKAY;
-__V:mp_clear (&u);
-__U:mp_clear (&v);
+LBL_V:mp_clear (&u);
+LBL_U:mp_clear (&v);
   return res;
 }
 #endif
@@ -3556,24 +3562,24 @@ int mp_invmod_slow (mp_int * a, mp_int * b, mp_int * c)
 
   /* x = a, y = b */
   if ((res = mp_copy (a, &x)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
   if ((res = mp_copy (b, &y)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* 2. [modified] if x,y are both even then return an error! */
   if (mp_iseven (&x) == 1 && mp_iseven (&y) == 1) {
     res = MP_VAL;
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* 3. u=x, v=y, A=1, B=0, C=0,D=1 */
   if ((res = mp_copy (&x, &u)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
   if ((res = mp_copy (&y, &v)) != MP_OKAY) {
-    goto __ERR;
+    goto LBL_ERR;
   }
   mp_set (&A, 1);
   mp_set (&D, 1);
@@ -3583,24 +3589,24 @@ top:
   while (mp_iseven (&u) == 1) {
     /* 4.1 u = u/2 */
     if ((res = mp_div_2 (&u, &u)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     /* 4.2 if A or B is odd then */
     if (mp_isodd (&A) == 1 || mp_isodd (&B) == 1) {
       /* A = (A+y)/2, B = (B-x)/2 */
       if ((res = mp_add (&A, &y, &A)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
       if ((res = mp_sub (&B, &x, &B)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
     }
     /* A = A/2, B = B/2 */
     if ((res = mp_div_2 (&A, &A)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     if ((res = mp_div_2 (&B, &B)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -3608,24 +3614,24 @@ top:
   while (mp_iseven (&v) == 1) {
     /* 5.1 v = v/2 */
     if ((res = mp_div_2 (&v, &v)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     /* 5.2 if C or D is odd then */
     if (mp_isodd (&C) == 1 || mp_isodd (&D) == 1) {
       /* C = (C+y)/2, D = (D-x)/2 */
       if ((res = mp_add (&C, &y, &C)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
       if ((res = mp_sub (&D, &x, &D)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
     }
     /* C = C/2, D = D/2 */
     if ((res = mp_div_2 (&C, &C)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
     if ((res = mp_div_2 (&D, &D)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -3633,28 +3639,28 @@ top:
   if (mp_cmp (&u, &v) != MP_LT) {
     /* u = u - v, A = A - C, B = B - D */
     if ((res = mp_sub (&u, &v, &u)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&A, &C, &A)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&B, &D, &B)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   } else {
     /* v - v - u, C = C - A, D = D - B */
     if ((res = mp_sub (&v, &u, &v)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&C, &A, &C)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
 
     if ((res = mp_sub (&D, &B, &D)) != MP_OKAY) {
-      goto __ERR;
+      goto LBL_ERR;
     }
   }
 
@@ -3667,27 +3673,27 @@ top:
   /* if v != 1 then there is no inverse */
   if (mp_cmp_d (&v, 1) != MP_EQ) {
     res = MP_VAL;
-    goto __ERR;
+    goto LBL_ERR;
   }
 
   /* if its too low */
   while (mp_cmp_d(&C, 0) == MP_LT) {
       if ((res = mp_add(&C, b, &C)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
   }
   
   /* too big */
   while (mp_cmp_mag(&C, b) != MP_LT) {
       if ((res = mp_sub(&C, b, &C)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
   }
   
   /* C is now the inverse */
   mp_exch (&C, c);
   res = MP_OKAY;
-__ERR:mp_clear_multi (&x, &y, &u, &v, &A, &B, &C, &D, NULL);
+LBL_ERR:mp_clear_multi (&x, &y, &u, &v, &A, &B, &C, &D, NULL);
   return res;
 }
 #endif
@@ -3856,13 +3862,13 @@ int mp_jacobi (mp_int * a, mp_int * p, int *c)
   }
 
   if ((res = mp_init (&p1)) != MP_OKAY) {
-    goto __A1;
+    goto LBL_A1;
   }
 
   /* divide out larger power of two */
   k = mp_cnt_lsb(&a1);
   if ((res = mp_div_2d(&a1, k, &a1, NULL)) != MP_OKAY) {
-     goto __P1;
+     goto LBL_P1;
   }
 
   /* step 4.  if e is even set s=1 */
@@ -3890,18 +3896,18 @@ int mp_jacobi (mp_int * a, mp_int * p, int *c)
   } else {
     /* n1 = n mod a1 */
     if ((res = mp_mod (p, &a1, &p1)) != MP_OKAY) {
-      goto __P1;
+      goto LBL_P1;
     }
     if ((res = mp_jacobi (&p1, &a1, &r)) != MP_OKAY) {
-      goto __P1;
+      goto LBL_P1;
     }
     *c = s * r;
   }
 
   /* done */
   res = MP_OKAY;
-__P1:mp_clear (&p1);
-__A1:mp_clear (&a1);
+LBL_P1:mp_clear (&p1);
+LBL_A1:mp_clear (&a1);
   return res;
 }
 #endif
@@ -4227,20 +4233,20 @@ int mp_lcm (mp_int * a, mp_int * b, mp_int * c)
 
   /* t1 = get the GCD of the two inputs */
   if ((res = mp_gcd (a, b, &t1)) != MP_OKAY) {
-    goto __T;
+    goto LBL_T;
   }
 
   /* divide the smallest by the GCD */
   if (mp_cmp_mag(a, b) == MP_LT) {
      /* store quotient in t2 such that t2 * b is the LCM */
      if ((res = mp_div(a, &t1, &t2, NULL)) != MP_OKAY) {
-        goto __T;
+        goto LBL_T;
      }
      res = mp_mul(b, &t2, c);
   } else {
      /* store quotient in t2 such that t2 * a is the LCM */
      if ((res = mp_div(b, &t1, &t2, NULL)) != MP_OKAY) {
-        goto __T;
+        goto LBL_T;
      }
      res = mp_mul(a, &t2, c);
   }
@@ -4248,7 +4254,7 @@ int mp_lcm (mp_int * a, mp_int * b, mp_int * c)
   /* fix the sign to positive */
   c->sign = MP_ZPOS;
 
-__T:
+LBL_T:
   mp_clear_multi (&t1, &t2, NULL);
   return res;
 }
@@ -4402,7 +4408,7 @@ mp_mod_2d (mp_int * a, int b, mp_int * c)
   }
 
   /* if the modulus is larger than the value than return */
-  if (b > (int) (a->used * DIGIT_BIT)) {
+  if (b >= (int) (a->used * DIGIT_BIT)) {
     res = mp_copy (a, c);
     return res;
   }
@@ -5085,11 +5091,11 @@ int mp_n_root (mp_int * a, mp_digit b, mp_int * c)
   }
 
   if ((res = mp_init (&t2)) != MP_OKAY) {
-    goto __T1;
+    goto LBL_T1;
   }
 
   if ((res = mp_init (&t3)) != MP_OKAY) {
-    goto __T2;
+    goto LBL_T2;
   }
 
   /* if a is negative fudge the sign but keep track */
@@ -5102,52 +5108,52 @@ int mp_n_root (mp_int * a, mp_digit b, mp_int * c)
   do {
     /* t1 = t2 */
     if ((res = mp_copy (&t2, &t1)) != MP_OKAY) {
-      goto __T3;
+      goto LBL_T3;
     }
 
     /* t2 = t1 - ((t1**b - a) / (b * t1**(b-1))) */
     
     /* t3 = t1**(b-1) */
     if ((res = mp_expt_d (&t1, b - 1, &t3)) != MP_OKAY) {   
-      goto __T3;
+      goto LBL_T3;
     }
 
     /* numerator */
     /* t2 = t1**b */
     if ((res = mp_mul (&t3, &t1, &t2)) != MP_OKAY) {    
-      goto __T3;
+      goto LBL_T3;
     }
 
     /* t2 = t1**b - a */
     if ((res = mp_sub (&t2, a, &t2)) != MP_OKAY) {  
-      goto __T3;
+      goto LBL_T3;
     }
 
     /* denominator */
     /* t3 = t1**(b-1) * b  */
     if ((res = mp_mul_d (&t3, b, &t3)) != MP_OKAY) {    
-      goto __T3;
+      goto LBL_T3;
     }
 
     /* t3 = (t1**b - a)/(b * t1**(b-1)) */
     if ((res = mp_div (&t2, &t3, &t3, NULL)) != MP_OKAY) {  
-      goto __T3;
+      goto LBL_T3;
     }
 
     if ((res = mp_sub (&t1, &t3, &t2)) != MP_OKAY) {
-      goto __T3;
+      goto LBL_T3;
     }
   }  while (mp_cmp (&t1, &t2) != MP_EQ);
 
   /* result can be off by a few so check */
   for (;;) {
     if ((res = mp_expt_d (&t1, b, &t2)) != MP_OKAY) {
-      goto __T3;
+      goto LBL_T3;
     }
 
     if (mp_cmp (&t2, a) == MP_GT) {
       if ((res = mp_sub_d (&t1, 1, &t1)) != MP_OKAY) {
-         goto __T3;
+         goto LBL_T3;
       }
     } else {
       break;
@@ -5165,9 +5171,9 @@ int mp_n_root (mp_int * a, mp_digit b, mp_int * c)
 
   res = MP_OKAY;
 
-__T3:mp_clear (&t3);
-__T2:mp_clear (&t2);
-__T1:mp_clear (&t1);
+LBL_T3:mp_clear (&t3);
+LBL_T2:mp_clear (&t2);
+LBL_T1:mp_clear (&t1);
   return res;
 }
 #endif
@@ -5304,7 +5310,7 @@ int mp_prime_fermat (mp_int * a, mp_int * b, int *result)
 
   /* compute t = b**a mod a */
   if ((err = mp_exptmod (b, a, a, &t)) != MP_OKAY) {
-    goto __T;
+    goto LBL_T;
   }
 
   /* is it equal to b? */
@@ -5313,7 +5319,7 @@ int mp_prime_fermat (mp_int * a, mp_int * b, int *result)
   }
 
   err = MP_OKAY;
-__T:mp_clear (&t);
+LBL_T:mp_clear (&t);
   return err;
 }
 #endif
@@ -5352,8 +5358,8 @@ int mp_prime_is_divisible (mp_int * a, int *result)
   *result = MP_NO;
 
   for (ix = 0; ix < PRIME_SIZE; ix++) {
-    /* what is a mod __prime_tab[ix] */
-    if ((err = mp_mod_d (a, __prime_tab[ix], &res)) != MP_OKAY) {
+    /* what is a mod LBL_prime_tab[ix] */
+    if ((err = mp_mod_d (a, ltm_prime_tab[ix], &res)) != MP_OKAY) {
       return err;
     }
 
@@ -5410,7 +5416,7 @@ int mp_prime_is_prime (mp_int * a, int t, int *result)
 
   /* is the input equal to one of the primes in the table? */
   for (ix = 0; ix < PRIME_SIZE; ix++) {
-      if (mp_cmp_d(a, __prime_tab[ix]) == MP_EQ) {
+      if (mp_cmp_d(a, ltm_prime_tab[ix]) == MP_EQ) {
          *result = 1;
          return MP_OKAY;
       }
@@ -5433,20 +5439,20 @@ int mp_prime_is_prime (mp_int * a, int t, int *result)
 
   for (ix = 0; ix < t; ix++) {
     /* set the prime */
-    mp_set (&b, __prime_tab[ix]);
+    mp_set (&b, ltm_prime_tab[ix]);
 
     if ((err = mp_prime_miller_rabin (a, &b, &res)) != MP_OKAY) {
-      goto __B;
+      goto LBL_B;
     }
 
     if (res == MP_NO) {
-      goto __B;
+      goto LBL_B;
     }
   }
 
   /* passed the test */
   *result = MP_YES;
-__B:mp_clear (&b);
+LBL_B:mp_clear (&b);
   return err;
 }
 #endif
@@ -5496,12 +5502,12 @@ int mp_prime_miller_rabin (mp_int * a, mp_int * b, int *result)
     return err;
   }
   if ((err = mp_sub_d (&n1, 1, &n1)) != MP_OKAY) {
-    goto __N1;
+    goto LBL_N1;
   }
 
   /* set 2**s * r = n1 */
   if ((err = mp_init_copy (&r, &n1)) != MP_OKAY) {
-    goto __N1;
+    goto LBL_N1;
   }
 
   /* count the number of least significant bits
@@ -5511,15 +5517,15 @@ int mp_prime_miller_rabin (mp_int * a, mp_int * b, int *result)
 
   /* now divide n - 1 by 2**s */
   if ((err = mp_div_2d (&r, s, &r, NULL)) != MP_OKAY) {
-    goto __R;
+    goto LBL_R;
   }
 
   /* compute y = b**r mod a */
   if ((err = mp_init (&y)) != MP_OKAY) {
-    goto __R;
+    goto LBL_R;
   }
   if ((err = mp_exptmod (b, &r, a, &y)) != MP_OKAY) {
-    goto __Y;
+    goto LBL_Y;
   }
 
   /* if y != 1 and y != n1 do */
@@ -5528,12 +5534,12 @@ int mp_prime_miller_rabin (mp_int * a, mp_int * b, int *result)
     /* while j <= s-1 and y != n1 */
     while ((j <= (s - 1)) && mp_cmp (&y, &n1) != MP_EQ) {
       if ((err = mp_sqrmod (&y, a, &y)) != MP_OKAY) {
-         goto __Y;
+         goto LBL_Y;
       }
 
       /* if y == 1 then composite */
       if (mp_cmp_d (&y, 1) == MP_EQ) {
-         goto __Y;
+         goto LBL_Y;
       }
 
       ++j;
@@ -5541,15 +5547,15 @@ int mp_prime_miller_rabin (mp_int * a, mp_int * b, int *result)
 
     /* if y != n1 then composite */
     if (mp_cmp (&y, &n1) != MP_EQ) {
-      goto __Y;
+      goto LBL_Y;
     }
   }
 
   /* probably prime now */
   *result = MP_YES;
-__Y:mp_clear (&y);
-__R:mp_clear (&r);
-__N1:mp_clear (&n1);
+LBL_Y:mp_clear (&y);
+LBL_R:mp_clear (&r);
+LBL_N1:mp_clear (&n1);
   return err;
 }
 #endif
@@ -5594,10 +5600,10 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
    a->sign = MP_ZPOS;
 
    /* simple algo if a is less than the largest prime in the table */
-   if (mp_cmp_d(a, __prime_tab[PRIME_SIZE-1]) == MP_LT) {
+   if (mp_cmp_d(a, ltm_prime_tab[PRIME_SIZE-1]) == MP_LT) {
       /* find which prime it is bigger than */
       for (x = PRIME_SIZE - 2; x >= 0; x--) {
-          if (mp_cmp_d(a, __prime_tab[x]) != MP_LT) {
+          if (mp_cmp_d(a, ltm_prime_tab[x]) != MP_LT) {
              if (bbs_style == 1) {
                 /* ok we found a prime smaller or
                  * equal [so the next is larger]
@@ -5605,17 +5611,17 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
                  * however, the prime must be
                  * congruent to 3 mod 4
                  */
-                if ((__prime_tab[x + 1] & 3) != 3) {
+                if ((ltm_prime_tab[x + 1] & 3) != 3) {
                    /* scan upwards for a prime congruent to 3 mod 4 */
                    for (y = x + 1; y < PRIME_SIZE; y++) {
-                       if ((__prime_tab[y] & 3) == 3) {
-                          mp_set(a, __prime_tab[y]);
+                       if ((ltm_prime_tab[y] & 3) == 3) {
+                          mp_set(a, ltm_prime_tab[y]);
                           return MP_OKAY;
                        }
                    }
                 }
              } else {
-                mp_set(a, __prime_tab[x + 1]);
+                mp_set(a, ltm_prime_tab[x + 1]);
                 return MP_OKAY;
              }
           }
@@ -5653,7 +5659,7 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
 
    /* generate the restable */
    for (x = 1; x < PRIME_SIZE; x++) {
-      if ((err = mp_mod_d(a, __prime_tab[x], res_tab + x)) != MP_OKAY) {
+      if ((err = mp_mod_d(a, ltm_prime_tab[x], res_tab + x)) != MP_OKAY) {
          return err;
       }
    }
@@ -5679,8 +5685,8 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
              res_tab[x] += kstep;
 
              /* subtract the modulus [instead of using division] */
-             if (res_tab[x] >= __prime_tab[x]) {
-                res_tab[x]  -= __prime_tab[x];
+             if (res_tab[x] >= ltm_prime_tab[x]) {
+                res_tab[x]  -= ltm_prime_tab[x];
              }
 
              /* set flag if zero */
@@ -5692,7 +5698,7 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
 
       /* add the step */
       if ((err = mp_add_d(a, step, a)) != MP_OKAY) {
-         goto __ERR;
+         goto LBL_ERR;
       }
 
       /* if didn't pass sieve and step == MAX then skip test */
@@ -5702,9 +5708,9 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
 
       /* is this prime? */
       for (x = 0; x < t; x++) {
-          mp_set(&b, __prime_tab[t]);
+          mp_set(&b, ltm_prime_tab[t]);
           if ((err = mp_prime_miller_rabin(a, &b, &res)) != MP_OKAY) {
-             goto __ERR;
+             goto LBL_ERR;
           }
           if (res == MP_NO) {
              break;
@@ -5717,7 +5723,7 @@ int mp_prime_next_prime(mp_int *a, int t, int bbs_style)
    }
 
    err = MP_OKAY;
-__ERR:
+LBL_ERR:
    mp_clear(&b);
    return err;
 }
@@ -5828,7 +5834,7 @@ int mp_prime_random_ex(mp_int *a, int t, int size, int flags, ltm_prime_callback
    }
 
    /* calc the byte size */
-   bsize = (size>>3)+(size&7?1:0);
+   bsize = (size>>3) + ((size&7)?1:0);
 
    /* we need a buffer of bsize bytes */
    tmp = OPT_CAST(unsigned char) XMALLOC(bsize);
@@ -5837,7 +5843,7 @@ int mp_prime_random_ex(mp_int *a, int t, int size, int flags, ltm_prime_callback
    }
 
    /* calc the maskAND value for the MSbyte*/
-   maskAND = 0xFF >> (8 - (size & 7));
+   maskAND = ((size&7) == 0) ? 0xFF : (0xFF >> (8 - (size & 7)));
 
    /* calc the maskOR_msb */
    maskOR_msb        = 0;
@@ -5846,7 +5852,7 @@ int mp_prime_random_ex(mp_int *a, int t, int size, int flags, ltm_prime_callback
       maskOR_msb     |= 1 << ((size - 2) & 7);
    } else if (flags & LTM_PRIME_2MSB_OFF) {
       maskAND        &= ~(1 << ((size - 2) & 7));
-   }
+   } 
 
    /* get the maskOR_lsb */
    maskOR_lsb         = 0;
@@ -7996,7 +8002,7 @@ mp_zero (mp_int * a)
  *
  * Tom St Denis, tomstdenis@iahu.ca, http://math.libtomcrypt.org
  */
-const mp_digit __prime_tab[] = {
+const mp_digit ltm_prime_tab[] = {
   0x0002, 0x0003, 0x0005, 0x0007, 0x000B, 0x000D, 0x0011, 0x0013,
   0x0017, 0x001D, 0x001F, 0x0025, 0x0029, 0x002B, 0x002F, 0x0035,
   0x003B, 0x003D, 0x0043, 0x0047, 0x0049, 0x004F, 0x0053, 0x0059,
@@ -8261,10 +8267,10 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
 
   /* create mu, used for Barrett reduction */
   if ((err = mp_init (&mu)) != MP_OKAY) {
-    goto __M;
+    goto LBL_M;
   }
   if ((err = mp_reduce_setup (&mu, P)) != MP_OKAY) {
-    goto __MU;
+    goto LBL_MU;
   }
 
   /* create M table
@@ -8276,23 +8282,23 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
    * computed though accept for M[0] and M[1]
    */
   if ((err = mp_mod (G, P, &M[1])) != MP_OKAY) {
-    goto __MU;
+    goto LBL_MU;
   }
 
   /* compute the value at M[1<<(winsize-1)] by squaring 
    * M[1] (winsize-1) times 
    */
   if ((err = mp_copy (&M[1], &M[1 << (winsize - 1)])) != MP_OKAY) {
-    goto __MU;
+    goto LBL_MU;
   }
 
   for (x = 0; x < (winsize - 1); x++) {
     if ((err = mp_sqr (&M[1 << (winsize - 1)], 
                        &M[1 << (winsize - 1)])) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
     if ((err = mp_reduce (&M[1 << (winsize - 1)], P, &mu)) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
   }
 
@@ -8301,16 +8307,16 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
    */
   for (x = (1 << (winsize - 1)) + 1; x < (1 << winsize); x++) {
     if ((err = mp_mul (&M[x - 1], &M[1], &M[x])) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
     if ((err = mp_reduce (&M[x], P, &mu)) != MP_OKAY) {
-      goto __MU;
+      goto LBL_MU;
     }
   }
 
   /* setup result */
   if ((err = mp_init (&res)) != MP_OKAY) {
-    goto __MU;
+    goto LBL_MU;
   }
   mp_set (&res, 1);
 
@@ -8350,10 +8356,10 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
     /* if the bit is zero and mode == 1 then we square */
     if (mode == 1 && y == 0) {
       if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       continue;
     }
@@ -8367,19 +8373,19 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
       /* square first */
       for (x = 0; x < winsize; x++) {
         if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
         if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
       }
 
       /* then multiply */
       if ((err = mp_mul (&res, &M[bitbuf], &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
 
       /* empty window and reset */
@@ -8394,20 +8400,20 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
     /* square then multiply if the bit is set */
     for (x = 0; x < bitcpy; x++) {
       if ((err = mp_sqr (&res, &res)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
       if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) {
-        goto __RES;
+        goto LBL_RES;
       }
 
       bitbuf <<= 1;
       if ((bitbuf & (1 << winsize)) != 0) {
         /* then multiply */
         if ((err = mp_mul (&res, &M[1], &res)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
         if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) {
-          goto __RES;
+          goto LBL_RES;
         }
       }
     }
@@ -8415,9 +8421,9 @@ int s_mp_exptmod (mp_int * G, mp_int * X, mp_int * P, mp_int * Y)
 
   mp_exch (&res, Y);
   err = MP_OKAY;
-__RES:mp_clear (&res);
-__MU:mp_clear (&mu);
-__M:
+LBL_RES:mp_clear (&res);
+LBL_MU:mp_clear (&mu);
+LBL_M:
   mp_clear(&M[1]);
   for (x = 1<<(winsize-1); x < (1 << winsize); x++) {
     mp_clear (&M[x]);
diff --git a/tommath.h b/tommath.h
index 896d389..7cc92c2 100644
--- a/tommath.h
+++ b/tommath.h
@@ -442,7 +442,7 @@ int mp_exptmod(mp_int *a, mp_int *b, mp_int *c, mp_int *d);
 #endif
 
 /* table of first PRIME_SIZE primes */
-extern const mp_digit __prime_tab[];
+extern const mp_digit ltm_prime_tab[];
 
 /* result=1 if a is divisible by one of the first PRIME_SIZE primes */
 int mp_prime_is_divisible(mp_int *a, int *result);
diff --git a/tommath.pdf b/tommath.pdf
index 18cac6f..88e2dc7 100644
Binary files a/tommath.pdf and b/tommath.pdf differ
diff --git a/tommath.tex b/tommath.tex
index d0ac947..9c4dc82 100644
--- a/tommath.tex
+++ b/tommath.tex
@@ -3420,7 +3420,7 @@ is copied to $b$, leading digits are removed and the remaining leading digit is
 027     \}
 028   
 029     /* if the modulus is larger than the value than return */
-030     if (b > (int) (a->used * DIGIT_BIT)) \{
+030     if (b >= (int) (a->used * DIGIT_BIT)) \{
 031       res = mp_copy (a, c);
 032       return res;
 033     \}
@@ -3896,7 +3896,7 @@ and addition operations in the nested loop in parallel.
 049   
 050     /* clear the carry */
 051     _W = 0;
-052     for (ix = 0; ix <= pa; ix++) \{ 
+052     for (ix = 0; ix < pa; ix++) \{ 
 053         int      tx, ty;
 054         int      iy;
 055         mp_digit *tmpx, *tmpy;
@@ -3927,27 +3927,30 @@ and addition operations in the nested loop in parallel.
 079         _W = _W >> ((mp_word)DIGIT_BIT);
 080     \}
 081   
-082     /* setup dest */
-083     olduse  = c->used;
-084     c->used = digs;
-085   
-086     \{
-087       register mp_digit *tmpc;
-088       tmpc = c->dp;
-089       for (ix = 0; ix < digs; ix++) \{
-090         /* now extract the previous digit [below the carry] */
-091         *tmpc++ = W[ix];
-092       \}
-093   
-094       /* clear unused digits [that existed in the old copy of c] */
-095       for (; ix < olduse; ix++) \{
-096         *tmpc++ = 0;
-097       \}
-098     \}
-099     mp_clamp (c);
-100     return MP_OKAY;
-101   \}
-102   #endif
+082     /* store final carry */
+083     W[ix] = _W;
+084   
+085     /* setup dest */
+086     olduse  = c->used;
+087     c->used = digs;
+088   
+089     \{
+090       register mp_digit *tmpc;
+091       tmpc = c->dp;
+092       for (ix = 0; ix < digs; ix++) \{
+093         /* now extract the previous digit [below the carry] */
+094         *tmpc++ = W[ix];
+095       \}
+096   
+097       /* clear unused digits [that existed in the old copy of c] */
+098       for (; ix < olduse; ix++) \{
+099         *tmpc++ = 0;
+100       \}
+101     \}
+102     mp_clamp (c);
+103     return MP_OKAY;
+104   \}
+105   #endif
 \end{alltt}
 \end{small}
 
@@ -3955,7 +3958,7 @@ The memset on line @47,memset@ clears the initial $\hat W$ array to zero in a si
 implementation a series of aliases (\textit{lines 62, 63 and 76}) are used to simplify the inner $O(n^2)$ loop.  
 In this case a new alias $\_\hat W$ has been added which refers to the double precision columns offset by $ix$ in each pass.  
 
-The inner loop on lines 89, 79 and 80 is where the algorithm will spend the majority of the time, which is why it has been 
+The inner loop on lines 92, 79 and 80 is where the algorithm will spend the majority of the time, which is why it has been 
 stripped to the bones of any extra baggage\footnote{Hence the pointer aliases.}.  On x86 processors the multiplication and additions amount to at the 
 very least five instructions (\textit{two loads, two additions, one multiply}) while on the ARMv4 processors they amount to only three 
 (\textit{one load, one store, one multiply-add}).   For both of the x86 and ARMv4 processors the GCC compiler performs a good job at unrolling the loop 
@@ -5100,7 +5103,7 @@ squares in place.
 059   
 060     /* number of output digits to produce */
 061     W1 = 0;
-062     for (ix = 0; ix <= pa; ix++) \{ 
+062     for (ix = 0; ix < pa; ix++) \{ 
 063         int      tx, ty, iy;
 064         mp_word  _W;
 065         mp_digit *tmpy;
@@ -6739,7 +6742,7 @@ at step 3.
 019    * Based on algorithm from the paper
 020    *
 021    * "Generating Efficient Primes for Discrete Log Cryptosystems"
-022    *                 Chae Hoon Lim, Pil Loong Lee,
+022    *                 Chae Hoon Lim, Pil Joong Lee,
 023    *          POSTECH Information Research Laboratories
 024    *
 025    * The modulus must be of a special format [see manual]
@@ -7594,7 +7597,7 @@ algorithm since their arguments are essentially the same (\textit{two mp\_ints a
 060        return err;
 061   #else 
 062        /* no invmod */
-063        return MP_VAL
+063        return MP_VAL;
 064   #endif
 065     \}
 066   
@@ -7866,10 +7869,10 @@ a Left-to-Right algorithm is used to process the remaining few bits.
 069   
 070     /* create mu, used for Barrett reduction */
 071     if ((err = mp_init (&mu)) != MP_OKAY) \{
-072       goto __M;
+072       goto LBL_M;
 073     \}
 074     if ((err = mp_reduce_setup (&mu, P)) != MP_OKAY) \{
-075       goto __MU;
+075       goto LBL_MU;
 076     \}
 077   
 078     /* create M table
@@ -7881,23 +7884,23 @@ a Left-to-Right algorithm is used to process the remaining few bits.
 084      * computed though accept for M[0] and M[1]
 085      */
 086     if ((err = mp_mod (G, P, &M[1])) != MP_OKAY) \{
-087       goto __MU;
+087       goto LBL_MU;
 088     \}
 089   
 090     /* compute the value at M[1<<(winsize-1)] by squaring 
 091      * M[1] (winsize-1) times 
 092      */
 093     if ((err = mp_copy (&M[1], &M[1 << (winsize - 1)])) != MP_OKAY) \{
-094       goto __MU;
+094       goto LBL_MU;
 095     \}
 096   
 097     for (x = 0; x < (winsize - 1); x++) \{
 098       if ((err = mp_sqr (&M[1 << (winsize - 1)], 
 099                          &M[1 << (winsize - 1)])) != MP_OKAY) \{
-100         goto __MU;
+100         goto LBL_MU;
 101       \}
 102       if ((err = mp_reduce (&M[1 << (winsize - 1)], P, &mu)) != MP_OKAY) \{
-103         goto __MU;
+103         goto LBL_MU;
 104       \}
 105     \}
 106   
@@ -7906,16 +7909,16 @@ a Left-to-Right algorithm is used to process the remaining few bits.
 109      */
 110     for (x = (1 << (winsize - 1)) + 1; x < (1 << winsize); x++) \{
 111       if ((err = mp_mul (&M[x - 1], &M[1], &M[x])) != MP_OKAY) \{
-112         goto __MU;
+112         goto LBL_MU;
 113       \}
 114       if ((err = mp_reduce (&M[x], P, &mu)) != MP_OKAY) \{
-115         goto __MU;
+115         goto LBL_MU;
 116       \}
 117     \}
 118   
 119     /* setup result */
 120     if ((err = mp_init (&res)) != MP_OKAY) \{
-121       goto __MU;
+121       goto LBL_MU;
 122     \}
 123     mp_set (&res, 1);
 124   
@@ -7955,10 +7958,10 @@ a Left-to-Right algorithm is used to process the remaining few bits.
 158       /* if the bit is zero and mode == 1 then we square */
 159       if (mode == 1 && y == 0) \{
 160         if ((err = mp_sqr (&res, &res)) != MP_OKAY) \{
-161           goto __RES;
+161           goto LBL_RES;
 162         \}
 163         if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) \{
-164           goto __RES;
+164           goto LBL_RES;
 165         \}
 166         continue;
 167       \}
@@ -7972,19 +7975,19 @@ a Left-to-Right algorithm is used to process the remaining few bits.
 175         /* square first */
 176         for (x = 0; x < winsize; x++) \{
 177           if ((err = mp_sqr (&res, &res)) != MP_OKAY) \{
-178             goto __RES;
+178             goto LBL_RES;
 179           \}
 180           if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) \{
-181             goto __RES;
+181             goto LBL_RES;
 182           \}
 183         \}
 184   
 185         /* then multiply */
 186         if ((err = mp_mul (&res, &M[bitbuf], &res)) != MP_OKAY) \{
-187           goto __RES;
+187           goto LBL_RES;
 188         \}
 189         if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) \{
-190           goto __RES;
+190           goto LBL_RES;
 191         \}
 192   
 193         /* empty window and reset */
@@ -7999,20 +8002,20 @@ a Left-to-Right algorithm is used to process the remaining few bits.
 202       /* square then multiply if the bit is set */
 203       for (x = 0; x < bitcpy; x++) \{
 204         if ((err = mp_sqr (&res, &res)) != MP_OKAY) \{
-205           goto __RES;
+205           goto LBL_RES;
 206         \}
 207         if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) \{
-208           goto __RES;
+208           goto LBL_RES;
 209         \}
 210   
 211         bitbuf <<= 1;
 212         if ((bitbuf & (1 << winsize)) != 0) \{
 213           /* then multiply */
 214           if ((err = mp_mul (&res, &M[1], &res)) != MP_OKAY) \{
-215             goto __RES;
+215             goto LBL_RES;
 216           \}
 217           if ((err = mp_reduce (&res, P, &mu)) != MP_OKAY) \{
-218             goto __RES;
+218             goto LBL_RES;
 219           \}
 220         \}
 221       \}
@@ -8020,9 +8023,9 @@ a Left-to-Right algorithm is used to process the remaining few bits.
 223   
 224     mp_exch (&res, Y);
 225     err = MP_OKAY;
-226   __RES:mp_clear (&res);
-227   __MU:mp_clear (&mu);
-228   __M:
+226   LBL_RES:mp_clear (&res);
+227   LBL_MU:mp_clear (&mu);
+228   LBL_M:
 229     mp_clear(&M[1]);
 230     for (x = 1<<(winsize-1); x < (1 << winsize); x++) \{
 231       mp_clear (&M[x]);
@@ -8386,23 +8389,23 @@ respectively be replaced with a zero.
 048   
 049     mp_set(&tq, 1);
 050     n = mp_count_bits(a) - mp_count_bits(b);
-051     if (((res = mp_copy(a, &ta)) != MP_OKAY) ||
-052         ((res = mp_copy(b, &tb)) != MP_OKAY) || 
+051     if (((res = mp_abs(a, &ta)) != MP_OKAY) ||
+052         ((res = mp_abs(b, &tb)) != MP_OKAY) || 
 053         ((res = mp_mul_2d(&tb, n, &tb)) != MP_OKAY) ||
 054         ((res = mp_mul_2d(&tq, n, &tq)) != MP_OKAY)) \{
-055         goto __ERR;
+055         goto LBL_ERR;
 056     \}
 057   
 058     while (n-- >= 0) \{
 059        if (mp_cmp(&tb, &ta) != MP_GT) \{
 060           if (((res = mp_sub(&ta, &tb, &ta)) != MP_OKAY) ||
 061               ((res = mp_add(&q, &tq, &q)) != MP_OKAY)) \{
-062              goto __ERR;
+062              goto LBL_ERR;
 063           \}
 064        \}
 065        if (((res = mp_div_2d(&tb, 1, &tb, NULL)) != MP_OKAY) ||
 066            ((res = mp_div_2d(&tq, 1, &tq, NULL)) != MP_OKAY)) \{
-067              goto __ERR;
+067              goto LBL_ERR;
 068        \}
 069     \}
 070   
@@ -8411,13 +8414,13 @@ respectively be replaced with a zero.
 073     n2 = (a->sign == b->sign ? MP_ZPOS : MP_NEG);
 074     if (c != NULL) \{
 075        mp_exch(c, &q);
-076        c->sign  = n2;
+076        c->sign  = (mp_iszero(c) == MP_YES) ? MP_ZPOS : n2;
 077     \}
 078     if (d != NULL) \{
 079        mp_exch(d, &ta);
-080        d->sign = n;
+080        d->sign = (mp_iszero(d) == MP_YES) ? MP_ZPOS : n;
 081     \}
-082   __ERR:
+082   LBL_ERR:
 083      mp_clear_multi(&ta, &tb, &tq, &q, NULL);
 084      return res;
 085   \}
@@ -8466,19 +8469,19 @@ respectively be replaced with a zero.
 128     q.used = a->used + 2;
 129   
 130     if ((res = mp_init (&t1)) != MP_OKAY) \{
-131       goto __Q;
+131       goto LBL_Q;
 132     \}
 133   
 134     if ((res = mp_init (&t2)) != MP_OKAY) \{
-135       goto __T1;
+135       goto LBL_T1;
 136     \}
 137   
 138     if ((res = mp_init_copy (&x, a)) != MP_OKAY) \{
-139       goto __T2;
+139       goto LBL_T2;
 140     \}
 141   
 142     if ((res = mp_init_copy (&y, b)) != MP_OKAY) \{
-143       goto __X;
+143       goto LBL_X;
 144     \}
 145   
 146     /* fix the sign */
@@ -8490,10 +8493,10 @@ respectively be replaced with a zero.
 152     if (norm < (int)(DIGIT_BIT-1)) \{
 153        norm = (DIGIT_BIT-1) - norm;
 154        if ((res = mp_mul_2d (&x, norm, &x)) != MP_OKAY) \{
-155          goto __Y;
+155          goto LBL_Y;
 156        \}
 157        if ((res = mp_mul_2d (&y, norm, &y)) != MP_OKAY) \{
-158          goto __Y;
+158          goto LBL_Y;
 159        \}
 160     \} else \{
 161        norm = 0;
@@ -8505,13 +8508,13 @@ respectively be replaced with a zero.
 167   
 168     /* while (x >= y*b**n-t) do \{ q[n-t] += 1; x -= y*b**\{n-t\} \} */
 169     if ((res = mp_lshd (&y, n - t)) != MP_OKAY) \{ /* y = y*b**\{n-t\} */
-170       goto __Y;
+170       goto LBL_Y;
 171     \}
 172   
 173     while (mp_cmp (&x, &y) != MP_LT) \{
 174       ++(q.dp[n - t]);
 175       if ((res = mp_sub (&x, &y, &x)) != MP_OKAY) \{
-176         goto __Y;
+176         goto LBL_Y;
 177       \}
 178     \}
 179   
@@ -8553,7 +8556,7 @@ respectively be replaced with a zero.
 215         t1.dp[1] = y.dp[t];
 216         t1.used = 2;
 217         if ((res = mp_mul_d (&t1, q.dp[i - t - 1], &t1)) != MP_OKAY) \{
-218           goto __Y;
+218           goto LBL_Y;
 219         \}
 220   
 221         /* find right hand */
@@ -8565,27 +8568,27 @@ respectively be replaced with a zero.
 227   
 228       /* step 3.3 x = x - q\{i-t-1\} * y * b**\{i-t-1\} */
 229       if ((res = mp_mul_d (&y, q.dp[i - t - 1], &t1)) != MP_OKAY) \{
-230         goto __Y;
+230         goto LBL_Y;
 231       \}
 232   
 233       if ((res = mp_lshd (&t1, i - t - 1)) != MP_OKAY) \{
-234         goto __Y;
+234         goto LBL_Y;
 235       \}
 236   
 237       if ((res = mp_sub (&x, &t1, &x)) != MP_OKAY) \{
-238         goto __Y;
+238         goto LBL_Y;
 239       \}
 240   
 241       /* if x < 0 then \{ x = x + y*b**\{i-t-1\}; q\{i-t-1\} -= 1; \} */
 242       if (x.sign == MP_NEG) \{
 243         if ((res = mp_copy (&y, &t1)) != MP_OKAY) \{
-244           goto __Y;
+244           goto LBL_Y;
 245         \}
 246         if ((res = mp_lshd (&t1, i - t - 1)) != MP_OKAY) \{
-247           goto __Y;
+247           goto LBL_Y;
 248         \}
 249         if ((res = mp_add (&x, &t1, &x)) != MP_OKAY) \{
-250           goto __Y;
+250           goto LBL_Y;
 251         \}
 252   
 253         q.dp[i - t - 1] = (q.dp[i - t - 1] - 1UL) & MP_MASK;
@@ -8612,11 +8615,11 @@ respectively be replaced with a zero.
 274   
 275     res = MP_OKAY;
 276   
-277   __Y:mp_clear (&y);
-278   __X:mp_clear (&x);
-279   __T2:mp_clear (&t2);
-280   __T1:mp_clear (&t1);
-281   __Q:mp_clear (&q);
+277   LBL_Y:mp_clear (&y);
+278   LBL_X:mp_clear (&x);
+279   LBL_T2:mp_clear (&t2);
+280   LBL_T1:mp_clear (&t1);
+281   LBL_Q:mp_clear (&q);
 282     return res;
 283   \}
 284   
@@ -9130,11 +9133,11 @@ root.  Ideally this algorithm is meant to find the $n$'th root of an input where
 039     \}
 040   
 041     if ((res = mp_init (&t2)) != MP_OKAY) \{
-042       goto __T1;
+042       goto LBL_T1;
 043     \}
 044   
 045     if ((res = mp_init (&t3)) != MP_OKAY) \{
-046       goto __T2;
+046       goto LBL_T2;
 047     \}
 048   
 049     /* if a is negative fudge the sign but keep track */
@@ -9147,52 +9150,52 @@ root.  Ideally this algorithm is meant to find the $n$'th root of an input where
 056     do \{
 057       /* t1 = t2 */
 058       if ((res = mp_copy (&t2, &t1)) != MP_OKAY) \{
-059         goto __T3;
+059         goto LBL_T3;
 060       \}
 061   
 062       /* t2 = t1 - ((t1**b - a) / (b * t1**(b-1))) */
 063       
 064       /* t3 = t1**(b-1) */
 065       if ((res = mp_expt_d (&t1, b - 1, &t3)) != MP_OKAY) \{   
-066         goto __T3;
+066         goto LBL_T3;
 067       \}
 068   
 069       /* numerator */
 070       /* t2 = t1**b */
 071       if ((res = mp_mul (&t3, &t1, &t2)) != MP_OKAY) \{    
-072         goto __T3;
+072         goto LBL_T3;
 073       \}
 074   
 075       /* t2 = t1**b - a */
 076       if ((res = mp_sub (&t2, a, &t2)) != MP_OKAY) \{  
-077         goto __T3;
+077         goto LBL_T3;
 078       \}
 079   
 080       /* denominator */
 081       /* t3 = t1**(b-1) * b  */
 082       if ((res = mp_mul_d (&t3, b, &t3)) != MP_OKAY) \{    
-083         goto __T3;
+083         goto LBL_T3;
 084       \}
 085   
 086       /* t3 = (t1**b - a)/(b * t1**(b-1)) */
 087       if ((res = mp_div (&t2, &t3, &t3, NULL)) != MP_OKAY) \{  
-088         goto __T3;
+088         goto LBL_T3;
 089       \}
 090   
 091       if ((res = mp_sub (&t1, &t3, &t2)) != MP_OKAY) \{
-092         goto __T3;
+092         goto LBL_T3;
 093       \}
 094     \}  while (mp_cmp (&t1, &t2) != MP_EQ);
 095   
 096     /* result can be off by a few so check */
 097     for (;;) \{
 098       if ((res = mp_expt_d (&t1, b, &t2)) != MP_OKAY) \{
-099         goto __T3;
+099         goto LBL_T3;
 100       \}
 101   
 102       if (mp_cmp (&t2, a) == MP_GT) \{
 103         if ((res = mp_sub_d (&t1, 1, &t1)) != MP_OKAY) \{
-104            goto __T3;
+104            goto LBL_T3;
 105         \}
 106       \} else \{
 107         break;
@@ -9210,9 +9213,9 @@ root.  Ideally this algorithm is meant to find the $n$'th root of an input where
 119   
 120     res = MP_OKAY;
 121   
-122   __T3:mp_clear (&t3);
-123   __T2:mp_clear (&t2);
-124   __T1:mp_clear (&t1);
+122   LBL_T3:mp_clear (&t3);
+123   LBL_T2:mp_clear (&t2);
+124   LBL_T1:mp_clear (&t1);
 125     return res;
 126   \}
 127   #endif
@@ -9771,7 +9774,7 @@ must be adjusted by multiplying by the common factors of two ($2^k$) removed ear
 042     \}
 043   
 044     if ((res = mp_init_copy (&v, b)) != MP_OKAY) \{
-045       goto __U;
+045       goto LBL_U;
 046     \}
 047   
 048     /* must be positive for the remainder of the algorithm */
@@ -9785,24 +9788,24 @@ must be adjusted by multiplying by the common factors of two ($2^k$) removed ear
 056     if (k > 0) \{
 057        /* divide the power of two out */
 058        if ((res = mp_div_2d(&u, k, &u, NULL)) != MP_OKAY) \{
-059           goto __V;
+059           goto LBL_V;
 060        \}
 061   
 062        if ((res = mp_div_2d(&v, k, &v, NULL)) != MP_OKAY) \{
-063           goto __V;
+063           goto LBL_V;
 064        \}
 065     \}
 066   
 067     /* divide any remaining factors of two out */
 068     if (u_lsb != k) \{
 069        if ((res = mp_div_2d(&u, u_lsb - k, &u, NULL)) != MP_OKAY) \{
-070           goto __V;
+070           goto LBL_V;
 071        \}
 072     \}
 073   
 074     if (v_lsb != k) \{
 075        if ((res = mp_div_2d(&v, v_lsb - k, &v, NULL)) != MP_OKAY) \{
-076           goto __V;
+076           goto LBL_V;
 077        \}
 078     \}
 079   
@@ -9815,23 +9818,23 @@ must be adjusted by multiplying by the common factors of two ($2^k$) removed ear
 086        
 087        /* subtract smallest from largest */
 088        if ((res = s_mp_sub(&v, &u, &v)) != MP_OKAY) \{
-089           goto __V;
+089           goto LBL_V;
 090        \}
 091        
 092        /* Divide out all factors of two */
 093        if ((res = mp_div_2d(&v, mp_cnt_lsb(&v), &v, NULL)) != MP_OKAY) \{
-094           goto __V;
+094           goto LBL_V;
 095        \} 
 096     \} 
 097   
 098     /* multiply by 2**k which we divided out at the beginning */
 099     if ((res = mp_mul_2d (&u, k, c)) != MP_OKAY) \{
-100        goto __V;
+100        goto LBL_V;
 101     \}
 102     c->sign = MP_ZPOS;
 103     res = MP_OKAY;
-104   __V:mp_clear (&u);
-105   __U:mp_clear (&v);
+104   LBL_V:mp_clear (&u);
+105   LBL_U:mp_clear (&v);
 106     return res;
 107   \}
 108   #endif
@@ -9904,20 +9907,20 @@ dividing the product of the two inputs by their greatest common divisor.
 027   
 028     /* t1 = get the GCD of the two inputs */
 029     if ((res = mp_gcd (a, b, &t1)) != MP_OKAY) \{
-030       goto __T;
+030       goto LBL_T;
 031     \}
 032   
 033     /* divide the smallest by the GCD */
 034     if (mp_cmp_mag(a, b) == MP_LT) \{
 035        /* store quotient in t2 such that t2 * b is the LCM */
 036        if ((res = mp_div(a, &t1, &t2, NULL)) != MP_OKAY) \{
-037           goto __T;
+037           goto LBL_T;
 038        \}
 039        res = mp_mul(b, &t2, c);
 040     \} else \{
 041        /* store quotient in t2 such that t2 * a is the LCM */
 042        if ((res = mp_div(b, &t1, &t2, NULL)) != MP_OKAY) \{
-043           goto __T;
+043           goto LBL_T;
 044        \}
 045        res = mp_mul(a, &t2, c);
 046     \}
@@ -9925,7 +9928,7 @@ dividing the product of the two inputs by their greatest common divisor.
 048     /* fix the sign to positive */
 049     c->sign = MP_ZPOS;
 050   
-051   __T:
+051   LBL_T:
 052     mp_clear_multi (&t1, &t2, NULL);
 053     return res;
 054   \}
@@ -10123,13 +10126,13 @@ $\left ( {p' \over a'} \right )$ which is multiplied against the current Jacobi
 049     \}
 050   
 051     if ((res = mp_init (&p1)) != MP_OKAY) \{
-052       goto __A1;
+052       goto LBL_A1;
 053     \}
 054   
 055     /* divide out larger power of two */
 056     k = mp_cnt_lsb(&a1);
 057     if ((res = mp_div_2d(&a1, k, &a1, NULL)) != MP_OKAY) \{
-058        goto __P1;
+058        goto LBL_P1;
 059     \}
 060   
 061     /* step 4.  if e is even set s=1 */
@@ -10157,18 +10160,18 @@ $\left ( {p' \over a'} \right )$ which is multiplied against the current Jacobi
 083     \} else \{
 084       /* n1 = n mod a1 */
 085       if ((res = mp_mod (p, &a1, &p1)) != MP_OKAY) \{
-086         goto __P1;
+086         goto LBL_P1;
 087       \}
 088       if ((res = mp_jacobi (&p1, &a1, &r)) != MP_OKAY) \{
-089         goto __P1;
+089         goto LBL_P1;
 090       \}
 091       *c = s * r;
 092     \}
 093   
 094     /* done */
 095     res = MP_OKAY;
-096   __P1:mp_clear (&p1);
-097   __A1:mp_clear (&a1);
+096   LBL_P1:mp_clear (&p1);
+097   LBL_A1:mp_clear (&a1);
 098     return res;
 099   \}
 100   #endif
@@ -10406,8 +10409,8 @@ This algorithm attempts to determine if a candidate integer $n$ is composite by
 028     *result = MP_NO;
 029   
 030     for (ix = 0; ix < PRIME_SIZE; ix++) \{
-031       /* what is a mod __prime_tab[ix] */
-032       if ((err = mp_mod_d (a, __prime_tab[ix], &res)) != MP_OKAY) \{
+031       /* what is a mod LBL_prime_tab[ix] */
+032       if ((err = mp_mod_d (a, ltm_prime_tab[ix], &res)) != MP_OKAY) \{
 033         return err;
 034       \}
 035   
@@ -10431,7 +10434,7 @@ mp\_digit.  The table \_\_prime\_tab is defined in the following file.
 \hspace{-5.1mm}{\bf File}: bn\_prime\_tab.c
 \vspace{-3mm}
 \begin{alltt}
-016   const mp_digit __prime_tab[] = \{
+016   const mp_digit ltm_prime_tab[] = \{
 017     0x0002, 0x0003, 0x0005, 0x0007, 0x000B, 0x000D, 0x0011, 0x0013,
 018     0x0017, 0x001D, 0x001F, 0x0025, 0x0029, 0x002B, 0x002F, 0x0035,
 019     0x003B, 0x003D, 0x0043, 0x0047, 0x0049, 0x004F, 0x0053, 0x0059,
@@ -10547,7 +10550,7 @@ determine the result.
 042   
 043     /* compute t = b**a mod a */
 044     if ((err = mp_exptmod (b, a, a, &t)) != MP_OKAY) \{
-045       goto __T;
+045       goto LBL_T;
 046     \}
 047   
 048     /* is it equal to b? */
@@ -10556,7 +10559,7 @@ determine the result.
 051     \}
 052   
 053     err = MP_OKAY;
-054   __T:mp_clear (&t);
+054   LBL_T:mp_clear (&t);
 055     return err;
 056   \}
 057   #endif
@@ -10638,12 +10641,12 @@ composite then it is \textit{probably} prime.
 039       return err;
 040     \}
 041     if ((err = mp_sub_d (&n1, 1, &n1)) != MP_OKAY) \{
-042       goto __N1;
+042       goto LBL_N1;
 043     \}
 044   
 045     /* set 2**s * r = n1 */
 046     if ((err = mp_init_copy (&r, &n1)) != MP_OKAY) \{
-047       goto __N1;
+047       goto LBL_N1;
 048     \}
 049   
 050     /* count the number of least significant bits
@@ -10653,15 +10656,15 @@ composite then it is \textit{probably} prime.
 054   
 055     /* now divide n - 1 by 2**s */
 056     if ((err = mp_div_2d (&r, s, &r, NULL)) != MP_OKAY) \{
-057       goto __R;
+057       goto LBL_R;
 058     \}
 059   
 060     /* compute y = b**r mod a */
 061     if ((err = mp_init (&y)) != MP_OKAY) \{
-062       goto __R;
+062       goto LBL_R;
 063     \}
 064     if ((err = mp_exptmod (b, &r, a, &y)) != MP_OKAY) \{
-065       goto __Y;
+065       goto LBL_Y;
 066     \}
 067   
 068     /* if y != 1 and y != n1 do */
@@ -10670,12 +10673,12 @@ composite then it is \textit{probably} prime.
 071       /* while j <= s-1 and y != n1 */
 072       while ((j <= (s - 1)) && mp_cmp (&y, &n1) != MP_EQ) \{
 073         if ((err = mp_sqrmod (&y, a, &y)) != MP_OKAY) \{
-074            goto __Y;
+074            goto LBL_Y;
 075         \}
 076   
 077         /* if y == 1 then composite */
 078         if (mp_cmp_d (&y, 1) == MP_EQ) \{
-079            goto __Y;
+079            goto LBL_Y;
 080         \}
 081   
 082         ++j;
@@ -10683,15 +10686,15 @@ composite then it is \textit{probably} prime.
 084   
 085       /* if y != n1 then composite */
 086       if (mp_cmp (&y, &n1) != MP_EQ) \{
-087         goto __Y;
+087         goto LBL_Y;
 088       \}
 089     \}
 090   
 091     /* probably prime now */
 092     *result = MP_YES;
-093   __Y:mp_clear (&y);
-094   __R:mp_clear (&r);
-095   __N1:mp_clear (&n1);
+093   LBL_Y:mp_clear (&y);
+094   LBL_R:mp_clear (&r);
+095   LBL_N1:mp_clear (&n1);
 096     return err;
 097   \}
 098   #endif
diff --git a/tommath_class.h b/tommath_class.h
index 2a17d43..53bfa31 100644
--- a/tommath_class.h
+++ b/tommath_class.h
@@ -242,6 +242,7 @@
    #define BN_MP_INIT_MULTI_C
    #define BN_MP_SET_C
    #define BN_MP_COUNT_BITS_C
+   #define BN_MP_ABS_C
    #define BN_MP_MUL_2D_C
    #define BN_MP_CMP_C
    #define BN_MP_SUB_C