summaryrefslogtreecommitdiffstats
path: root/crypto/bn/bn_asm.c
diff options
context:
space:
mode:
authorMatt Caswell <matt@openssl.org>2015-01-22 03:40:55 +0000
committerMatt Caswell <matt@openssl.org>2015-01-22 09:20:09 +0000
commit0f113f3ee4d629ef9a4a30911b22b224772085e5 (patch)
treee014603da5aed1d0751f587a66d6e270b6bda3de /crypto/bn/bn_asm.c
parent22b52164aaed31d6e93dbd2d397ace041360e6aa (diff)
Run util/openssl-format-source -v -c .
Reviewed-by: Tim Hudson <tjh@openssl.org>
Diffstat (limited to 'crypto/bn/bn_asm.c')
-rw-r--r--crypto/bn/bn_asm.c1830
1 files changed, 945 insertions, 885 deletions
diff --git a/crypto/bn/bn_asm.c b/crypto/bn/bn_asm.c
index 543fc9dc45..cb5a8037b2 100644
--- a/crypto/bn/bn_asm.c
+++ b/crypto/bn/bn_asm.c
@@ -5,21 +5,21 @@
* This package is an SSL implementation written
* by Eric Young (eay@cryptsoft.com).
* The implementation was written so as to conform with Netscapes SSL.
- *
+ *
* This library is free for commercial and non-commercial use as long as
* the following conditions are aheared to. The following conditions
* apply to all code found in this distribution, be it the RC4, RSA,
* lhash, DES, etc., code; not just the SSL code. The SSL documentation
* included with this distribution is covered by the same copyright terms
* except that the holder is Tim Hudson (tjh@cryptsoft.com).
- *
+ *
* Copyright remains Eric Young's, and as such any Copyright notices in
* the code are not to be removed.
* If this package is used in a product, Eric Young should be given attribution
* as the author of the parts of the library used.
* This can be in the form of a textual message at program startup or
* in documentation (online or textual) provided with the package.
- *
+ *
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
@@ -34,10 +34,10 @@
* Eric Young (eay@cryptsoft.com)"
* The word 'cryptographic' can be left out if the rouines from the library
* being used are not cryptographic related :-).
- * 4. If you include any Windows specific code (or a derivative thereof) from
+ * 4. If you include any Windows specific code (or a derivative thereof) from
* the apps directory (application code) you must include an acknowledgement:
* "This product includes software written by Tim Hudson (tjh@cryptsoft.com)"
- *
+ *
* THIS SOFTWARE IS PROVIDED BY ERIC YOUNG ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
@@ -49,7 +49,7 @@
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
- *
+ *
* The licence and distribution terms for any publically available version or
* derivative of this code cannot be changed. i.e. this code cannot simply be
* copied and put under another distribution licence
@@ -57,7 +57,7 @@
*/
#ifndef BN_DEBUG
-# undef NDEBUG /* avoid conflicting definitions */
+# undef NDEBUG /* avoid conflicting definitions */
# define NDEBUG
#endif
@@ -68,772 +68,823 @@
#if defined(BN_LLONG) || defined(BN_UMULT_HIGH)
-BN_ULONG bn_mul_add_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
- {
- BN_ULONG c1=0;
-
- assert(num >= 0);
- if (num <= 0) return(c1);
-
-#ifndef OPENSSL_SMALL_FOOTPRINT
- while (num&~3)
- {
- mul_add(rp[0],ap[0],w,c1);
- mul_add(rp[1],ap[1],w,c1);
- mul_add(rp[2],ap[2],w,c1);
- mul_add(rp[3],ap[3],w,c1);
- ap+=4; rp+=4; num-=4;
- }
-#endif
- while (num)
- {
- mul_add(rp[0],ap[0],w,c1);
- ap++; rp++; num--;
- }
-
- return(c1);
- }
+BN_ULONG bn_mul_add_words(BN_ULONG *rp, const BN_ULONG *ap, int num,
+ BN_ULONG w)
+{
+ BN_ULONG c1 = 0;
+
+ assert(num >= 0);
+ if (num <= 0)
+ return (c1);
+
+# ifndef OPENSSL_SMALL_FOOTPRINT
+ while (num & ~3) {
+ mul_add(rp[0], ap[0], w, c1);
+ mul_add(rp[1], ap[1], w, c1);
+ mul_add(rp[2], ap[2], w, c1);
+ mul_add(rp[3], ap[3], w, c1);
+ ap += 4;
+ rp += 4;
+ num -= 4;
+ }
+# endif
+ while (num) {
+ mul_add(rp[0], ap[0], w, c1);
+ ap++;
+ rp++;
+ num--;
+ }
+
+ return (c1);
+}
BN_ULONG bn_mul_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
- {
- BN_ULONG c1=0;
-
- assert(num >= 0);
- if (num <= 0) return(c1);
-
-#ifndef OPENSSL_SMALL_FOOTPRINT
- while (num&~3)
- {
- mul(rp[0],ap[0],w,c1);
- mul(rp[1],ap[1],w,c1);
- mul(rp[2],ap[2],w,c1);
- mul(rp[3],ap[3],w,c1);
- ap+=4; rp+=4; num-=4;
- }
-#endif
- while (num)
- {
- mul(rp[0],ap[0],w,c1);
- ap++; rp++; num--;
- }
- return(c1);
- }
+{
+ BN_ULONG c1 = 0;
+
+ assert(num >= 0);
+ if (num <= 0)
+ return (c1);
+
+# ifndef OPENSSL_SMALL_FOOTPRINT
+ while (num & ~3) {
+ mul(rp[0], ap[0], w, c1);
+ mul(rp[1], ap[1], w, c1);
+ mul(rp[2], ap[2], w, c1);
+ mul(rp[3], ap[3], w, c1);
+ ap += 4;
+ rp += 4;
+ num -= 4;
+ }
+# endif
+ while (num) {
+ mul(rp[0], ap[0], w, c1);
+ ap++;
+ rp++;
+ num--;
+ }
+ return (c1);
+}
void bn_sqr_words(BN_ULONG *r, const BN_ULONG *a, int n)
- {
- assert(n >= 0);
- if (n <= 0) return;
-
-#ifndef OPENSSL_SMALL_FOOTPRINT
- while (n&~3)
- {
- sqr(r[0],r[1],a[0]);
- sqr(r[2],r[3],a[1]);
- sqr(r[4],r[5],a[2]);
- sqr(r[6],r[7],a[3]);
- a+=4; r+=8; n-=4;
- }
-#endif
- while (n)
- {
- sqr(r[0],r[1],a[0]);
- a++; r+=2; n--;
- }
- }
-
-#else /* !(defined(BN_LLONG) || defined(BN_UMULT_HIGH)) */
-
-BN_ULONG bn_mul_add_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
- {
- BN_ULONG c=0;
- BN_ULONG bl,bh;
-
- assert(num >= 0);
- if (num <= 0) return((BN_ULONG)0);
-
- bl=LBITS(w);
- bh=HBITS(w);
-
-#ifndef OPENSSL_SMALL_FOOTPRINT
- while (num&~3)
- {
- mul_add(rp[0],ap[0],bl,bh,c);
- mul_add(rp[1],ap[1],bl,bh,c);
- mul_add(rp[2],ap[2],bl,bh,c);
- mul_add(rp[3],ap[3],bl,bh,c);
- ap+=4; rp+=4; num-=4;
- }
-#endif
- while (num)
- {
- mul_add(rp[0],ap[0],bl,bh,c);
- ap++; rp++; num--;
- }
- return(c);
- }
+{
+ assert(n >= 0);
+ if (n <= 0)
+ return;
+
+# ifndef OPENSSL_SMALL_FOOTPRINT
+ while (n & ~3) {
+ sqr(r[0], r[1], a[0]);
+ sqr(r[2], r[3], a[1]);
+ sqr(r[4], r[5], a[2]);
+ sqr(r[6], r[7], a[3]);
+ a += 4;
+ r += 8;
+ n -= 4;
+ }
+# endif
+ while (n) {
+ sqr(r[0], r[1], a[0]);
+ a++;
+ r += 2;
+ n--;
+ }
+}
+
+#else /* !(defined(BN_LLONG) ||
+ * defined(BN_UMULT_HIGH)) */
+
+BN_ULONG bn_mul_add_words(BN_ULONG *rp, const BN_ULONG *ap, int num,
+ BN_ULONG w)
+{
+ BN_ULONG c = 0;
+ BN_ULONG bl, bh;
+
+ assert(num >= 0);
+ if (num <= 0)
+ return ((BN_ULONG)0);
+
+ bl = LBITS(w);
+ bh = HBITS(w);
+
+# ifndef OPENSSL_SMALL_FOOTPRINT
+ while (num & ~3) {
+ mul_add(rp[0], ap[0], bl, bh, c);
+ mul_add(rp[1], ap[1], bl, bh, c);
+ mul_add(rp[2], ap[2], bl, bh, c);
+ mul_add(rp[3], ap[3], bl, bh, c);
+ ap += 4;
+ rp += 4;
+ num -= 4;
+ }
+# endif
+ while (num) {
+ mul_add(rp[0], ap[0], bl, bh, c);
+ ap++;
+ rp++;
+ num--;
+ }
+ return (c);
+}
BN_ULONG bn_mul_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
- {
- BN_ULONG carry=0;
- BN_ULONG bl,bh;
-
- assert(num >= 0);
- if (num <= 0) return((BN_ULONG)0);
-
- bl=LBITS(w);
- bh=HBITS(w);
-
-#ifndef OPENSSL_SMALL_FOOTPRINT
- while (num&~3)
- {
- mul(rp[0],ap[0],bl,bh,carry);
- mul(rp[1],ap[1],bl,bh,carry);
- mul(rp[2],ap[2],bl,bh,carry);
- mul(rp[3],ap[3],bl,bh,carry);
- ap+=4; rp+=4; num-=4;
- }
-#endif
- while (num)
- {
- mul(rp[0],ap[0],bl,bh,carry);
- ap++; rp++; num--;
- }
- return(carry);
- }
+{
+ BN_ULONG carry = 0;
+ BN_ULONG bl, bh;
+
+ assert(num >= 0);
+ if (num <= 0)
+ return ((BN_ULONG)0);
+
+ bl = LBITS(w);
+ bh = HBITS(w);
+
+# ifndef OPENSSL_SMALL_FOOTPRINT
+ while (num & ~3) {
+ mul(rp[0], ap[0], bl, bh, carry);
+ mul(rp[1], ap[1], bl, bh, carry);
+ mul(rp[2], ap[2], bl, bh, carry);
+ mul(rp[3], ap[3], bl, bh, carry);
+ ap += 4;
+ rp += 4;
+ num -= 4;
+ }
+# endif
+ while (num) {
+ mul(rp[0], ap[0], bl, bh, carry);
+ ap++;
+ rp++;
+ num--;
+ }
+ return (carry);
+}
void bn_sqr_words(BN_ULONG *r, const BN_ULONG *a, int n)
- {
- assert(n >= 0);
- if (n <= 0) return;
-
-#ifndef OPENSSL_SMALL_FOOTPRINT
- while (n&~3)
- {
- sqr64(r[0],r[1],a[0]);
- sqr64(r[2],r[3],a[1]);
- sqr64(r[4],r[5],a[2]);
- sqr64(r[6],r[7],a[3]);
- a+=4; r+=8; n-=4;
- }
-#endif
- while (n)
- {
- sqr64(r[0],r[1],a[0]);
- a++; r+=2; n--;
- }
- }
-
-#endif /* !(defined(BN_LLONG) || defined(BN_UMULT_HIGH)) */
+{
+ assert(n >= 0);
+ if (n <= 0)
+ return;
+
+# ifndef OPENSSL_SMALL_FOOTPRINT
+ while (n & ~3) {
+ sqr64(r[0], r[1], a[0]);
+ sqr64(r[2], r[3], a[1]);
+ sqr64(r[4], r[5], a[2]);
+ sqr64(r[6], r[7], a[3]);
+ a += 4;
+ r += 8;
+ n -= 4;
+ }
+# endif
+ while (n) {
+ sqr64(r[0], r[1], a[0]);
+ a++;
+ r += 2;
+ n--;
+ }
+}
+
+#endif /* !(defined(BN_LLONG) ||
+ * defined(BN_UMULT_HIGH)) */
#if defined(BN_LLONG) && defined(BN_DIV2W)
BN_ULONG bn_div_words(BN_ULONG h, BN_ULONG l, BN_ULONG d)
- {
- return((BN_ULONG)(((((BN_ULLONG)h)<<BN_BITS2)|l)/(BN_ULLONG)d));
- }
+{
+ return ((BN_ULONG)(((((BN_ULLONG) h) << BN_BITS2) | l) / (BN_ULLONG) d));
+}
#else
/* Divide h,l by d and return the result. */
/* I need to test this some more :-( */
BN_ULONG bn_div_words(BN_ULONG h, BN_ULONG l, BN_ULONG d)
- {
- BN_ULONG dh,dl,q,ret=0,th,tl,t;
- int i,count=2;
-
- if (d == 0) return(BN_MASK2);
-
- i=BN_num_bits_word(d);
- assert((i == BN_BITS2) || (h <= (BN_ULONG)1<<i));
-
- i=BN_BITS2-i;
- if (h >= d) h-=d;
-
- if (i)
- {
- d<<=i;
- h=(h<<i)|(l>>(BN_BITS2-i));
- l<<=i;
- }
- dh=(d&BN_MASK2h)>>BN_BITS4;
- dl=(d&BN_MASK2l);
- for (;;)
- {
- if ((h>>BN_BITS4) == dh)
- q=BN_MASK2l;
- else
- q=h/dh;
-
- th=q*dh;
- tl=dl*q;
- for (;;)
- {
- t=h-th;
- if ((t&BN_MASK2h) ||
- ((tl) <= (
- (t<<BN_BITS4)|
- ((l&BN_MASK2h)>>BN_BITS4))))
- break;
- q--;
- th-=dh;
- tl-=dl;
- }
- t=(tl>>BN_BITS4);
- tl=(tl<<BN_BITS4)&BN_MASK2h;
- th+=t;
-
- if (l < tl) th++;
- l-=tl;
- if (h < th)
- {
- h+=d;
- q--;
- }
- h-=th;
-
- if (--count == 0) break;
-
- ret=q<<BN_BITS4;
- h=((h<<BN_BITS4)|(l>>BN_BITS4))&BN_MASK2;
- l=(l&BN_MASK2l)<<BN_BITS4;
- }
- ret|=q;
- return(ret);
- }
-#endif /* !defined(BN_LLONG) && defined(BN_DIV2W) */
+{
+ BN_ULONG dh, dl, q, ret = 0, th, tl, t;
+ int i, count = 2;
+
+ if (d == 0)
+ return (BN_MASK2);
+
+ i = BN_num_bits_word(d);
+ assert((i == BN_BITS2) || (h <= (BN_ULONG)1 << i));
+
+ i = BN_BITS2 - i;
+ if (h >= d)
+ h -= d;
+
+ if (i) {
+ d <<= i;
+ h = (h << i) | (l >> (BN_BITS2 - i));
+ l <<= i;
+ }
+ dh = (d & BN_MASK2h) >> BN_BITS4;
+ dl = (d & BN_MASK2l);
+ for (;;) {
+ if ((h >> BN_BITS4) == dh)
+ q = BN_MASK2l;
+ else
+ q = h / dh;
+
+ th = q * dh;
+ tl = dl * q;
+ for (;;) {
+ t = h - th;
+ if ((t & BN_MASK2h) ||
+ ((tl) <= ((t << BN_BITS4) | ((l & BN_MASK2h) >> BN_BITS4))))
+ break;
+ q--;
+ th -= dh;
+ tl -= dl;
+ }
+ t = (tl >> BN_BITS4);
+ tl = (tl << BN_BITS4) & BN_MASK2h;
+ th += t;
+
+ if (l < tl)
+ th++;
+ l -= tl;
+ if (h < th) {
+ h += d;
+ q--;
+ }
+ h -= th;
+
+ if (--count == 0)
+ break;
+
+ ret = q << BN_BITS4;
+ h = ((h << BN_BITS4) | (l >> BN_BITS4)) & BN_MASK2;
+ l = (l & BN_MASK2l) << BN_BITS4;
+ }
+ ret |= q;
+ return (ret);
+}
+#endif /* !defined(BN_LLONG) && defined(BN_DIV2W) */
#ifdef BN_LLONG
-BN_ULONG bn_add_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b, int n)
- {
- BN_ULLONG ll=0;
-
- assert(n >= 0);
- if (n <= 0) return((BN_ULONG)0);
-
-#ifndef OPENSSL_SMALL_FOOTPRINT
- while (n&~3)
- {
- ll+=(BN_ULLONG)a[0]+b[0];
- r[0]=(BN_ULONG)ll&BN_MASK2;
- ll>>=BN_BITS2;
- ll+=(BN_ULLONG)a[1]+b[1];
- r[1]=(BN_ULONG)ll&BN_MASK2;
- ll>>=BN_BITS2;
- ll+=(BN_ULLONG)a[2]+b[2];
- r[2]=(BN_ULONG)ll&BN_MASK2;
- ll>>=BN_BITS2;
- ll+=(BN_ULLONG)a[3]+b[3];
- r[3]=(BN_ULONG)ll&BN_MASK2;
- ll>>=BN_BITS2;
- a+=4; b+=4; r+=4; n-=4;
- }
-#endif
- while (n)
- {
- ll+=(BN_ULLONG)a[0]+b[0];
- r[0]=(BN_ULONG)ll&BN_MASK2;
- ll>>=BN_BITS2;
- a++; b++; r++; n--;
- }
- return((BN_ULONG)ll);
- }
-#else /* !BN_LLONG */
-BN_ULONG bn_add_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b, int n)
- {
- BN_ULONG c,l,t;
-
- assert(n >= 0);
- if (n <= 0) return((BN_ULONG)0);
-
- c=0;
-#ifndef OPENSSL_SMALL_FOOTPRINT
- while (n&~3)
- {
- t=a[0];
- t=(t+c)&BN_MASK2;
- c=(t < c);
- l=(t+b[0])&BN_MASK2;
- c+=(l < t);
- r[0]=l;
- t=a[1];
- t=(t+c)&BN_MASK2;
- c=(t < c);
- l=(t+b[1])&BN_MASK2;
- c+=(l < t);
- r[1]=l;
- t=a[2];
- t=(t+c)&BN_MASK2;
- c=(t < c);
- l=(t+b[2])&BN_MASK2;
- c+=(l < t);
- r[2]=l;
- t=a[3];
- t=(t+c)&BN_MASK2;
- c=(t < c);
- l=(t+b[3])&BN_MASK2;
- c+=(l < t);
- r[3]=l;
- a+=4; b+=4; r+=4; n-=4;
- }
-#endif
- while(n)
- {
- t=a[0];
- t=(t+c)&BN_MASK2;
- c=(t < c);
- l=(t+b[0])&BN_MASK2;
- c+=(l < t);
- r[0]=l;
- a++; b++; r++; n--;
- }
- return((BN_ULONG)c);
- }
-#endif /* !BN_LLONG */
-
-BN_ULONG bn_sub_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b, int n)
- {
- BN_ULONG t1,t2;
- int c=0;
-
- assert(n >= 0);
- if (n <= 0) return((BN_ULONG)0);
+BN_ULONG bn_add_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b,
+ int n)
+{
+ BN_ULLONG ll = 0;
+
+ assert(n >= 0);
+ if (n <= 0)
+ return ((BN_ULONG)0);
+
+# ifndef OPENSSL_SMALL_FOOTPRINT
+ while (n & ~3) {
+ ll += (BN_ULLONG) a[0] + b[0];
+ r[0] = (BN_ULONG)ll & BN_MASK2;
+ ll >>= BN_BITS2;
+ ll += (BN_ULLONG) a[1] + b[1];
+ r[1] = (BN_ULONG)ll & BN_MASK2;
+ ll >>= BN_BITS2;
+ ll += (BN_ULLONG) a[2] + b[2];
+ r[2] = (BN_ULONG)ll & BN_MASK2;
+ ll >>= BN_BITS2;
+ ll += (BN_ULLONG) a[3] + b[3];
+ r[3] = (BN_ULONG)ll & BN_MASK2;
+ ll >>= BN_BITS2;
+ a += 4;
+ b += 4;
+ r += 4;
+ n -= 4;
+ }
+# endif
+ while (n) {
+ ll += (BN_ULLONG) a[0] + b[0];
+ r[0] = (BN_ULONG)ll & BN_MASK2;
+ ll >>= BN_BITS2;
+ a++;
+ b++;
+ r++;
+ n--;
+ }
+ return ((BN_ULONG)ll);
+}
+#else /* !BN_LLONG */
+BN_ULONG bn_add_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b,
+ int n)
+{
+ BN_ULONG c, l, t;
+
+ assert(n >= 0);
+ if (n <= 0)
+ return ((BN_ULONG)0);
+
+ c = 0;
+# ifndef OPENSSL_SMALL_FOOTPRINT
+ while (n & ~3) {
+ t = a[0];
+ t = (t + c) & BN_MASK2;
+ c = (t < c);
+ l = (t + b[0]) & BN_MASK2;
+ c += (l < t);
+ r[0] = l;
+ t = a[1];
+ t = (t + c) & BN_MASK2;
+ c = (t < c);
+ l = (t + b[1]) & BN_MASK2;
+ c += (l < t);
+ r[1] = l;
+ t = a[2];
+ t = (t + c) & BN_MASK2;
+ c = (t < c);
+ l = (t + b[2]) & BN_MASK2;
+ c += (l < t);
+ r[2] = l;
+ t = a[3];
+ t = (t + c) & BN_MASK2;
+ c = (t < c);
+ l = (t + b[3]) & BN_MASK2;
+ c += (l < t);
+ r[3] = l;
+ a += 4;
+ b += 4;
+ r += 4;
+ n -= 4;
+ }
+# endif
+ while (n) {
+ t = a[0];
+ t = (t + c) & BN_MASK2;
+ c = (t < c);
+ l = (t + b[0]) & BN_MASK2;
+ c += (l < t);
+ r[0] = l;
+ a++;
+ b++;
+ r++;
+ n--;
+ }
+ return ((BN_ULONG)c);
+}
+#endif /* !BN_LLONG */
+
+BN_ULONG bn_sub_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b,
+ int n)
+{
+ BN_ULONG t1, t2;
+ int c = 0;
+
+ assert(n >= 0);
+ if (n <= 0)
+ return ((BN_ULONG)0);
#ifndef OPENSSL_SMALL_FOOTPRINT
- while (n&~3)
- {
- t1=a[0]; t2=b[0];
- r[0]=(t1-t2-c)&BN_MASK2;
- if (t1 != t2) c=(t1 < t2);
- t1=a[1]; t2=b[1];
- r[1]=(t1-t2-c)&BN_MASK2;
- if (t1 != t2) c=(t1 < t2);
- t1=a[2]; t2=b[2];
- r[2]=(t1-t2-c)&BN_MASK2;
- if (t1 != t2) c=(t1 < t2);
- t1=a[3]; t2=b[3];
- r[3]=(t1-t2-c)&BN_MASK2;
- if (t1 != t2) c=(t1 < t2);
- a+=4; b+=4; r+=4; n-=4;
- }
+ while (n & ~3) {
+ t1 = a[0];
+ t2 = b[0];
+ r[0] = (t1 - t2 - c) & BN_MASK2;
+ if (t1 != t2)
+ c = (t1 < t2);
+ t1 = a[1];
+ t2 = b[1];
+ r[1] = (t1 - t2 - c) & BN_MASK2;
+ if (t1 != t2)
+ c = (t1 < t2);
+ t1 = a[2];
+ t2 = b[2];
+ r[2] = (t1 - t2 - c) & BN_MASK2;
+ if (t1 != t2)
+ c = (t1 < t2);
+ t1 = a[3];
+ t2 = b[3];
+ r[3] = (t1 - t2 - c) & BN_MASK2;
+ if (t1 != t2)
+ c = (t1 < t2);
+ a += 4;
+ b += 4;
+ r += 4;
+ n -= 4;
+ }
#endif
- while (n)
- {
- t1=a[0]; t2=b[0];
- r[0]=(t1-t2-c)&BN_MASK2;
- if (t1 != t2) c=(t1 < t2);
- a++; b++; r++; n--;
- }
- return(c);
- }
+ while (n) {
+ t1 = a[0];
+ t2 = b[0];
+ r[0] = (t1 - t2 - c) & BN_MASK2;
+ if (t1 != t2)
+ c = (t1 < t2);
+ a++;
+ b++;
+ r++;
+ n--;
+ }
+ return (c);
+}
#if defined(BN_MUL_COMBA) && !defined(OPENSSL_SMALL_FOOTPRINT)
-#undef bn_mul_comba8
-#undef bn_mul_comba4
-#undef bn_sqr_comba8
-#undef bn_sqr_comba4
+# undef bn_mul_comba8
+# undef bn_mul_comba4
+# undef bn_sqr_comba8
+# undef bn_sqr_comba4
/* mul_add_c(a,b,c0,c1,c2) -- c+=a*b for three word number c=(c2,c1,c0) */
/* mul_add_c2(a,b,c0,c1,c2) -- c+=2*a*b for three word number c=(c2,c1,c0) */
/* sqr_add_c(a,i,c0,c1,c2) -- c+=a[i]^2 for three word number c=(c2,c1,c0) */
-/* sqr_add_c2(a,i,c0,c1,c2) -- c+=2*a[i]*a[j] for three word number c=(c2,c1,c0) */
+/*
+ * sqr_add_c2(a,i,c0,c1,c2) -- c+=2*a[i]*a[j] for three word number
+ * c=(c2,c1,c0)
+ */
-#ifdef BN_LLONG
+# ifdef BN_LLONG
/*
* Keep in mind that additions to multiplication result can not
* overflow, because its high half cannot be all-ones.
*/
-# define mul_add_c(a,b,c0,c1,c2) do { \
- BN_ULONG hi; \
- BN_ULLONG t = (BN_ULLONG)(a)*(b); \
- t += c0; /* no carry */ \
- c0 = (BN_ULONG)Lw(t); \
- hi = (BN_ULONG)Hw(t); \
- c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
- } while(0)
-
-# define mul_add_c2(a,b,c0,c1,c2) do { \
- BN_ULONG hi; \
- BN_ULLONG t = (BN_ULLONG)(a)*(b); \
- BN_ULLONG tt = t+c0; /* no carry */ \
- c0 = (BN_ULONG)Lw(tt); \
- hi = (BN_ULONG)Hw(tt); \
- c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
- t += c0; /* no carry */ \
- c0 = (BN_ULONG)Lw(t); \
- hi = (BN_ULONG)Hw(t); \
- c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
- } while(0)
-
-# define sqr_add_c(a,i,c0,c1,c2) do { \
- BN_ULONG hi; \
- BN_ULLONG t = (BN_ULLONG)a[i]*a[i]; \
- t += c0; /* no carry */ \
- c0 = (BN_ULONG)Lw(t); \
- hi = (BN_ULONG)Hw(t); \
- c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
- } while(0)
-
-# define sqr_add_c2(a,i,j,c0,c1,c2) \
- mul_add_c2((a)[i],(a)[j],c0,c1,c2)
-
-#elif defined(BN_UMULT_LOHI)
+# define mul_add_c(a,b,c0,c1,c2) do { \
+ BN_ULONG hi; \
+ BN_ULLONG t = (BN_ULLONG)(a)*(b); \
+ t += c0; /* no carry */ \
+ c0 = (BN_ULONG)Lw(t); \
+ hi = (BN_ULONG)Hw(t); \
+ c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
+ } while(0)
+
+# define mul_add_c2(a,b,c0,c1,c2) do { \
+ BN_ULONG hi; \
+ BN_ULLONG t = (BN_ULLONG)(a)*(b); \
+ BN_ULLONG tt = t+c0; /* no carry */ \
+ c0 = (BN_ULONG)Lw(tt); \
+ hi = (BN_ULONG)Hw(tt); \
+ c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
+ t += c0; /* no carry */ \
+ c0 = (BN_ULONG)Lw(t); \
+ hi = (BN_ULONG)Hw(t); \
+ c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
+ } while(0)
+
+# define sqr_add_c(a,i,c0,c1,c2) do { \
+ BN_ULONG hi; \
+ BN_ULLONG t = (BN_ULLONG)a[i]*a[i]; \
+ t += c0; /* no carry */ \
+ c0 = (BN_ULONG)Lw(t); \
+ hi = (BN_ULONG)Hw(t); \
+ c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
+ } while(0)
+
+# define sqr_add_c2(a,i,j,c0,c1,c2) \
+ mul_add_c2((a)[i],(a)[j],c0,c1,c2)
+
+# elif defined(BN_UMULT_LOHI)
/*
* Keep in mind that additions to hi can not overflow, because
* the high word of a multiplication result cannot be all-ones.
*/
-# define mul_add_c(a,b,c0,c1,c2) do { \
- BN_ULONG ta = (a), tb = (b); \
- BN_ULONG lo, hi; \
- BN_UMULT_LOHI(lo,hi,ta,tb); \
- c0 += lo; hi += (c0<lo)?1:0; \
- c1 += hi; c2 += (c1<hi)?1:0; \
- } while(0)
-
-# define mul_add_c2(a,b,c0,c1,c2) do { \
- BN_ULONG ta = (a), tb = (b); \
- BN_ULONG lo, hi, tt; \
- BN_UMULT_LOHI(lo,hi,ta,tb); \
- c0 += lo; tt = hi+((c0<lo)?1:0); \
- c1 += tt; c2 += (c1<tt)?1:0; \
- c0 += lo; hi += (c0<lo)?1:0; \
- c1 += hi; c2 += (c1<hi)?1:0; \
- } while(0)
-
-# define sqr_add_c(a,i,c0,c1,c2) do { \
- BN_ULONG ta = (a)[i]; \
- BN_ULONG lo, hi; \
- BN_UMULT_LOHI(lo,hi,ta,ta); \
- c0 += lo; hi += (c0<lo)?1:0; \
- c1 += hi; c2 += (c1<hi)?1:0; \
- } while(0)
-
-# define sqr_add_c2(a,i,j,c0,c1,c2) \
- mul_add_c2((a)[i],(a)[j],c0,c1,c2)
-
-#elif defined(BN_UMULT_HIGH)
+# define mul_add_c(a,b,c0,c1,c2) do { \
+ BN_ULONG ta = (a), tb = (b); \
+ BN_ULONG lo, hi; \
+ BN_UMULT_LOHI(lo,hi,ta,tb); \
+ c0 += lo; hi += (c0<lo)?1:0; \
+ c1 += hi; c2 += (c1<hi)?1:0; \
+ } while(0)
+
+# define mul_add_c2(a,b,c0,c1,c2) do { \
+ BN_ULONG ta = (a), tb = (b); \
+ BN_ULONG lo, hi, tt; \
+ BN_UMULT_LOHI(lo,hi,ta,tb); \
+ c0 += lo; tt = hi+((c0<lo)?1:0); \
+ c1 += tt; c2 += (c1<tt)?1:0; \
+ c0 += lo; hi += (c0<lo)?1:0; \
+ c1 += hi; c2 += (c1<hi)?1:0; \
+ } while(0)
+
+# define sqr_add_c(a,i,c0,c1,c2) do { \
+ BN_ULONG ta = (a)[i]; \
+ BN_ULONG lo, hi; \
+ BN_UMULT_LOHI(lo,hi,ta,ta); \
+ c0 += lo; hi += (c0<lo)?1:0; \
+ c1 += hi; c2 += (c1<hi)?1:0; \
+ } while(0)
+
+# define sqr_add_c2(a,i,j,c0,c1,c2) \
+ mul_add_c2((a)[i],(a)[j],c0,c1,c2)
+
+# elif defined(BN_UMULT_HIGH)
/*
* Keep in mind that additions to hi can not overflow, because
* the high word of a multiplication result cannot be all-ones.
*/
-# define mul_add_c(a,b,c0,c1,c2) do { \
- BN_ULONG ta = (a), tb = (b); \
- BN_ULONG lo = ta * tb; \
- BN_ULONG hi = BN_UMULT_HIGH(ta,tb); \
- c0 += lo; hi += (c0<lo)?1:0; \
- c1 += hi; c2 += (c1<hi)?1:0; \
- } while(0)
-
-# define mul_add_c2(a,b,c0,c1,c2) do { \
- BN_ULONG ta = (a), tb = (b), tt; \
- BN_ULONG lo = ta * tb; \
- BN_ULONG hi = BN_UMULT_HIGH(ta,tb); \
- c0 += lo; tt = hi + ((c0<lo)?1:0); \
- c1 += tt; c2 += (c1<tt)?1:0; \
- c0 += lo; hi += (c0<lo)?1:0; \
- c1 += hi; c2 += (c1<hi)?1:0; \
- } while(0)
-
-# define sqr_add_c(a,i,c0,c1,c2) do { \
- BN_ULONG ta = (a)[i]; \
- BN_ULONG lo = ta * ta; \
- BN_ULONG hi = BN_UMULT_HIGH(ta,ta); \
- c0 += lo; hi += (c0<lo)?1:0; \
- c1 += hi; c2 += (c1<hi)?1:0; \
- } while(0)
-
-#define sqr_add_c2(a,i,j,c0,c1,c2) \
- mul_add_c2((a)[i],(a)[j],c0,c1,c2)
-
-#else /* !BN_LLONG */
+# define mul_add_c(a,b,c0,c1,c2) do { \
+ BN_ULONG ta = (a), tb = (b); \
+ BN_ULONG lo = ta * tb; \
+ BN_ULONG hi = BN_UMULT_HIGH(ta,tb); \
+ c0 += lo; hi += (c0<lo)?1:0; \
+ c1 += hi; c2 += (c1<hi)?1:0; \
+ } while(0)
+
+# define mul_add_c2(a,b,c0,c1,c2) do { \
+ BN_ULONG ta = (a), tb = (b), tt; \
+ BN_ULONG lo = ta * tb; \
+ BN_ULONG hi = BN_UMULT_HIGH(ta,tb); \
+ c0 += lo; tt = hi + ((c0<lo)?1:0); \
+ c1 += tt; c2 += (c1<tt)?1:0; \
+ c0 += lo; hi += (c0<lo)?1:0; \
+ c1 += hi; c2 += (c1<hi)?1:0; \
+ } while(0)
+
+# define sqr_add_c(a,i,c0,c1,c2) do { \
+ BN_ULONG ta = (a)[i]; \
+ BN_ULONG lo = ta * ta; \
+ BN_ULONG hi = BN_UMULT_HIGH(ta,ta); \
+ c0 += lo; hi += (c0<lo)?1:0; \
+ c1 += hi; c2 += (c1<hi)?1:0; \
+ } while(0)
+
+# define sqr_add_c2(a,i,j,c0,c1,c2) \
+ mul_add_c2((a)[i],(a)[j],c0,c1,c2)
+
+# else /* !BN_LLONG */
/*
* Keep in mind that additions to hi can not overflow, because
* the high word of a multiplication result cannot be all-ones.
*/
-# define mul_add_c(a,b,c0,c1,c2) do { \
- BN_ULONG lo = LBITS(a), hi = HBITS(a); \
- BN_ULONG bl = LBITS(b), bh = HBITS(b); \
- mul64(lo,hi,bl,bh); \
- c0 = (c0+lo)&BN_MASK2; if (c0<lo) hi++; \
- c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
- } while(0)
-
-# define mul_add_c2(a,b,c0,c1,c2) do { \
- BN_ULONG tt; \
- BN_ULONG lo = LBITS(a), hi = HBITS(a); \
- BN_ULONG bl = LBITS(b), bh = HBITS(b); \
- mul64(lo,hi,bl,bh); \
- tt = hi; \
- c0 = (c0+lo)&BN_MASK2; if (c0<lo) tt++; \
- c1 = (c1+tt)&BN_MASK2; if (c1<tt) c2++; \
- c0 = (c0+lo)&BN_MASK2; if (c0<lo) hi++; \
- c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
- } while(0)
-
-# define sqr_add_c(a,i,c0,c1,c2) do { \
- BN_ULONG lo, hi; \
- sqr64(lo,hi,(a)[i]); \
- c0 = (c0+lo)&BN_MASK2; if (c0<lo) hi++; \
- c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
- } while(0)
+# define mul_add_c(a,b,c0,c1,c2) do { \
+ BN_ULONG lo = LBITS(a), hi = HBITS(a); \
+ BN_ULONG bl = LBITS(b), bh = HBITS(b); \
+ mul64(lo,hi,bl,bh); \
+ c0 = (c0+lo)&BN_MASK2; if (c0<lo) hi++; \
+ c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
+ } while(0)
+
+# define mul_add_c2(a,b,c0,c1,c2) do { \
+ BN_ULONG tt; \
+ BN_ULONG lo = LBITS(a), hi = HBITS(a); \
+ BN_ULONG bl = LBITS(b), bh = HBITS(b); \
+ mul64(lo,hi,bl,bh); \
+ tt = hi; \
+ c0 = (c0+lo)&BN_MASK2; if (c0<lo) tt++; \
+ c1 = (c1+tt)&BN_MASK2; if (c1<tt) c2++; \
+ c0 = (c0+lo)&BN_MASK2; if (c0<lo) hi++; \
+ c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
+ } while(0)
+
+# define sqr_add_c(a,i,c0,c1,c2) do { \
+ BN_ULONG lo, hi; \
+ sqr64(lo,hi,(a)[i]); \
+ c0 = (c0+lo)&BN_MASK2; if (c0<lo) hi++; \
+ c1 = (c1+hi)&BN_MASK2; if (c1<hi) c2++; \
+ } while(0)
# define sqr_add_c2(a,i,j,c0,c1,c2) \
- mul_add_c2((a)[i],(a)[j],c0,c1,c2)
-#endif /* !BN_LLONG */
+ mul_add_c2((a)[i],(a)[j],c0,c1,c2)
+# endif /* !BN_LLONG */
void bn_mul_comba8(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
- {
- BN_ULONG c1,c2,c3;
-
- c1=0;
- c2=0;
- c3=0;
- mul_add_c(a[0],b[0],c1,c2,c3);
- r[0]=c1;
- c1=0;
- mul_add_c(a[0],b[1],c2,c3,c1);
- mul_add_c(a[1],b[0],c2,c3,c1);
- r[1]=c2;
- c2=0;
- mul_add_c(a[2],b[0],c3,c1,c2);
- mul_add_c(a[1],b[1],c3,c1,c2);
- mul_add_c(a[0],b[2],c3,c1,c2);
- r[2]=c3;
- c3=0;
- mul_add_c(a[0],b[3],c1,c2,c3);
- mul_add_c(a[1],b[2],c1,c2,c3);
- mul_add_c(a[2],b[1],c1,c2,c3);
- mul_add_c(a[3],b[0],c1,c2,c3);
- r[3]=c1;
- c1=0;
- mul_add_c(a[4],b[0],c2,c3,c1);
- mul_add_c(a[3],b[1],c2,c3,c1);
- mul_add_c(a[2],b[2],c2,c3,c1);
- mul_add_c(a[1],b[3],c2,c3,c1);
- mul_add_c(a[0],b[4],c2,c3,c1);
- r[4]=c2;
- c2=0;
- mul_add_c(a[0],b[5],c3,c1,c2);
- mul_add_c(a[1],b[4],c3,c1,c2);
- mul_add_c(a[2],b[3],c3,c1,c2);
- mul_add_c(a[3],b[2],c3,c1,c2);
- mul_add_c(a[4],b[1],c3,c1,c2);
- mul_add_c(a[5],b[0],c3,c1,c2);
- r[5]=c3;
- c3=0;
- mul_add_c(a[6],b[0],c1,c2,c3);
- mul_add_c(a[5],b[1],c1,c2,c3);
- mul_add_c(a[4],b[2],c1,c2,c3);
- mul_add_c(a[3],b[3],c1,c2,c3);
- mul_add_c(a[2],b[4],c1,c2,c3);
- mul_add_c(a[1],b[5],c1,c2,c3);
- mul_add_c(a[0],b[6],c1,c2,c3);
- r[6]=c1;
- c1=0;
- mul_add_c(a[0],b[7],c2,c3,c1);
- mul_add_c(a[1],b[6],c2,c3,c1);
- mul_add_c(a[2],b[5],c2,c3,c1);
- mul_add_c(a[3],b[4],c2,c3,c1);
- mul_add_c(a[4],b[3],c2,c3,c1);
- mul_add_c(a[5],b[2],c2,c3,c1);
- mul_add_c(a[6],b[1],c2,c3,c1);
- mul_add_c(a[7],b[0],c2,c3,c1);
- r[7]=c2;
- c2=0;
- mul_add_c(a[7],b[1],c3,c1,c2);
- mul_add_c(a[6],b[2],c3,c1,c2);
- mul_add_c(a[5],b[3],c3,c1,c2);
- mul_add_c(a[4],b[4],c3,c1,c2);
- mul_add_c(a[3],b[5],c3,c1,c2);
- mul_add_c(a[2],b[6],c3,c1,c2);
- mul_add_c(a[1],b[7],c3,c1,c2);
- r[8]=c3;
- c3=0;
- mul_add_c(a[2],b[7],c1,c2,c3);
- mul_add_c(a[3],b[6],c1,c2,c3);
- mul_add_c(a[4],b[5],c1,c2,c3);
- mul_add_c(a[5],b[4],c1,c2,c3);
- mul_add_c(a[6],b[3],c1,c2,c3);
- mul_add_c(a[7],b[2],c1,c2,c3);
- r[9]=c1;
- c1=0;
- mul_add_c(a[7],b[3],c2,c3,c1);
- mul_add_c(a[6],b[4],c2,c3,c1);
- mul_add_c(a[5],b[5],c2,c3,c1);
- mul_add_c(a[4],b[6],c2,c3,c1);
- mul_add_c(a[3],b[7],c2,c3,c1);
- r[10]=c2;