[PATCH 2/4] cast5: read Kr four blocks at time and shift for current round

[Jussi Kivilinna]

* cipher/cast5.c (do_encrypt_block, do_decrypt_block): Read Kr as
32-bit words instead of bytes and shift value for each round.
--

Signed-off-by: Jussi Kivilinna <jussi.kivilinna at iki.fi>
---
 cipher/cast5.c |   72 ++++++++++++++++++++++++++++----------------------------
 1 file changed, 36 insertions(+), 36 deletions(-)

diff --git a/cipher/cast5.c b/cipher/cast5.c
index cc5bd9d66..65485ba23 100644
--- a/cipher/cast5.c
+++ b/cipher/cast5.c
@@ -483,10 +483,10 @@ do_encrypt_block( CAST5_context *c, byte *outbuf, const byte *inbuf )
     u32 l, r, t;
     u32 I;   /* used by the Fx macros */
     u32 *Km;
-    byte *Kr;
+    u32 Kr;
 
     Km = c->Km;
-    Kr = c->Kr;
+    Kr = buf_get_le32(c->Kr + 0);
 
     /* (L0,R0) <-- (m1...m64).	(Split the plaintext into left and
      * right 32-bit halves L0 = m1...m32 and R0 = m33...m64.)
@@ -502,22 +502,22 @@ do_encrypt_block( CAST5_context *c, byte *outbuf, const byte *inbuf )
      * Rounds 3, 6, 9, 12, and 15 use f function Type 3.
      */
 
-    t = l; l = r; r = t ^ F1(r, Km[ 0], Kr[ 0]);
-    t = l; l = r; r = t ^ F2(r, Km[ 1], Kr[ 1]);
-    t = l; l = r; r = t ^ F3(r, Km[ 2], Kr[ 2]);
-    t = l; l = r; r = t ^ F1(r, Km[ 3], Kr[ 3]);
-    t = l; l = r; r = t ^ F2(r, Km[ 4], Kr[ 4]);
-    t = l; l = r; r = t ^ F3(r, Km[ 5], Kr[ 5]);
-    t = l; l = r; r = t ^ F1(r, Km[ 6], Kr[ 6]);
-    t = l; l = r; r = t ^ F2(r, Km[ 7], Kr[ 7]);
-    t = l; l = r; r = t ^ F3(r, Km[ 8], Kr[ 8]);
-    t = l; l = r; r = t ^ F1(r, Km[ 9], Kr[ 9]);
-    t = l; l = r; r = t ^ F2(r, Km[10], Kr[10]);
-    t = l; l = r; r = t ^ F3(r, Km[11], Kr[11]);
-    t = l; l = r; r = t ^ F1(r, Km[12], Kr[12]);
-    t = l; l = r; r = t ^ F2(r, Km[13], Kr[13]);
-    t = l; l = r; r = t ^ F3(r, Km[14], Kr[14]);
-    t = l; l = r; r = t ^ F1(r, Km[15], Kr[15]);
+    t = l; l = r; r = t ^ F1(r, Km[ 0], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F2(r, Km[ 1], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F3(r, Km[ 2], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F1(r, Km[ 3], Kr & 31); Kr = buf_get_le32(c->Kr + 4);
+    t = l; l = r; r = t ^ F2(r, Km[ 4], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F3(r, Km[ 5], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F1(r, Km[ 6], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F2(r, Km[ 7], Kr & 31); Kr = buf_get_le32(c->Kr + 8);
+    t = l; l = r; r = t ^ F3(r, Km[ 8], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F1(r, Km[ 9], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F2(r, Km[10], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F3(r, Km[11], Kr & 31); Kr = buf_get_le32(c->Kr + 12);
+    t = l; l = r; r = t ^ F1(r, Km[12], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F2(r, Km[13], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F3(r, Km[14], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F1(r, Km[15], Kr & 31);
 
     /* c1...c64 <-- (R16,L16).	(Exchange final blocks L16, R16 and
      *	concatenate to form the ciphertext.) */
@@ -540,30 +540,30 @@ do_decrypt_block (CAST5_context *c, byte *outbuf, const byte *inbuf )
     u32 l, r, t;
     u32 I;
     u32 *Km;
-    byte *Kr;
+    u32 Kr;
 
     Km = c->Km;
-    Kr = c->Kr;
+    Kr = buf_get_be32(c->Kr + 12);
 
     l = buf_get_be32(inbuf + 0);
     r = buf_get_be32(inbuf + 4);
 
-    t = l; l = r; r = t ^ F1(r, Km[15], Kr[15]);
-    t = l; l = r; r = t ^ F3(r, Km[14], Kr[14]);
-    t = l; l = r; r = t ^ F2(r, Km[13], Kr[13]);
-    t = l; l = r; r = t ^ F1(r, Km[12], Kr[12]);
-    t = l; l = r; r = t ^ F3(r, Km[11], Kr[11]);
-    t = l; l = r; r = t ^ F2(r, Km[10], Kr[10]);
-    t = l; l = r; r = t ^ F1(r, Km[ 9], Kr[ 9]);
-    t = l; l = r; r = t ^ F3(r, Km[ 8], Kr[ 8]);
-    t = l; l = r; r = t ^ F2(r, Km[ 7], Kr[ 7]);
-    t = l; l = r; r = t ^ F1(r, Km[ 6], Kr[ 6]);
-    t = l; l = r; r = t ^ F3(r, Km[ 5], Kr[ 5]);
-    t = l; l = r; r = t ^ F2(r, Km[ 4], Kr[ 4]);
-    t = l; l = r; r = t ^ F1(r, Km[ 3], Kr[ 3]);
-    t = l; l = r; r = t ^ F3(r, Km[ 2], Kr[ 2]);
-    t = l; l = r; r = t ^ F2(r, Km[ 1], Kr[ 1]);
-    t = l; l = r; r = t ^ F1(r, Km[ 0], Kr[ 0]);
+    t = l; l = r; r = t ^ F1(r, Km[15], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F3(r, Km[14], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F2(r, Km[13], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F1(r, Km[12], Kr & 31); Kr = buf_get_be32(c->Kr + 8);
+    t = l; l = r; r = t ^ F3(r, Km[11], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F2(r, Km[10], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F1(r, Km[ 9], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F3(r, Km[ 8], Kr & 31); Kr = buf_get_be32(c->Kr + 4);
+    t = l; l = r; r = t ^ F2(r, Km[ 7], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F1(r, Km[ 6], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F3(r, Km[ 5], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F2(r, Km[ 4], Kr & 31); Kr = buf_get_be32(c->Kr + 0);
+    t = l; l = r; r = t ^ F1(r, Km[ 3], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F3(r, Km[ 2], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F2(r, Km[ 1], Kr & 31); Kr >>= 8;
+    t = l; l = r; r = t ^ F1(r, Km[ 0], Kr & 31);
 
     buf_put_be32(outbuf + 0, r);
     buf_put_be32(outbuf + 4, l);