--- /dev/null
+/* -*-c-*-
+ *
+ * $Id: square.c,v 1.1 2000/07/15 20:51:58 mdw Exp $
+ *
+ * The Square block cipher
+ *
+ * (c) 2000 Straylight/Edgeware
+ */
+
+/*----- Licensing notice --------------------------------------------------*
+ *
+ * This file is part of Catacomb.
+ *
+ * Catacomb is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU Library General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * Catacomb is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU Library General Public License for more details.
+ *
+ * You should have received a copy of the GNU Library General Public
+ * License along with Catacomb; if not, write to the Free
+ * Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+ * MA 02111-1307, USA.
+ */
+
+/*----- Revision history --------------------------------------------------*
+ *
+ * $Log: square.c,v $
+ * Revision 1.1 2000/07/15 20:51:58 mdw
+ * New block cipher.
+ *
+ */
+
+/*----- Header files ------------------------------------------------------*/
+
+#include <assert.h>
+#include <stdio.h>
+
+#include <mLib/bits.h>
+
+#include "blkc.h"
+#include "gcipher.h"
+#include "paranoia.h"
+#include "square.h"
+#include "square-tab.h"
+
+/*----- Global variables --------------------------------------------------*/
+
+const octet square_keysz[] = { KSZ_RANGE, SQUARE_KEYSZ, 4, 16, 4 };
+
+/*----- Constant tables ---------------------------------------------------*/
+
+static const octet S[256] = SQUARE_S, SI[256] = SQUARE_SI;
+static const uint32 T[4][256] = SQUARE_T, TI[4][256] = SQUARE_TI;
+static const uint32 U[4][256] = SQUARE_U;
+static const octet rcon[] = SQUARE_RCON;
+
+/*----- Main code ---------------------------------------------------------*/
+
+#define BYTESUB(x, s) \
+ (s[U8((x) >> 24)] << 24 | s[U8((x) >> 16)] << 16 | \
+ s[U8((x) >> 8)] << 8 | s[U8((x) >> 0)] << 0)
+
+/* --- @square_init@ --- *
+ *
+ * Arguments: @square_ctx *k@ = pointer to context to initialize
+ * @const void *buf@ = pointer to buffer of key material
+ * @size_t sz@ = size of the key material
+ *
+ * Returns: ---
+ *
+ * Use: Initializes a Square context with a particular key. Square
+ * keys must be a multiple of 32 bits long, and may be at most
+ * 128 bits.
+ */
+
+void square_init(square_ctx *k, const void *buf, size_t sz)
+{
+ unsigned nk, nr, nw;
+ unsigned i, j, jj;
+ const octet *p;
+ uint32 ww;
+ uint32 kk[SQUARE_KWORDS];
+
+ /* --- Sort out the key size --- */
+
+ KSZ_ASSERT(square, sz);
+ nk = sz / 4;
+
+ /* --- Fetch the first key words out --- */
+
+ p = buf;
+ for (i = 0; i < nk; i++) {
+ kk[i] = LOAD32_L(p);
+ p += 4;
+ }
+ nr = 8;
+
+ /* --- Expand this material to fill the rest of the table --- */
+
+ nw = (nr + 1) * 4;
+ ww = kk[i - 1];
+ p = rcon;
+ for (; i < nw; i++) {
+ uint32 w = kk[i - nk];
+ if (i % nk == 0) {
+ ww = ROR32(ww, 8);
+ w ^= ww ^ *p++;
+ } else
+ w ^= ww;
+ kk[i] = ww = w;
+ }
+
+ /* --- Make the encryption and decryption keys --- */
+
+ for (i = 0; i < nr * 4; i++) {
+ uint32 w = kk[i];
+ k->w[i] = (U[0][U8(w >> 0)] ^ U[1][U8(w >> 8)] ^
+ U[2][U8(w >> 16)] ^ U[3][U8(w >> 24)]);
+ }
+ for (; i < nw; i++)
+ k->w[i] = kk[i];
+
+ jj = nw;
+ for (i = 0; i < nr * 4; i += 4) {
+ jj -= 4;
+ for (j = 0; j < 4; j++)
+ k->wi[i + j] = kk[jj + j];
+ }
+ for (j = 0; j < 4; j++)
+ k->wi[i + j] = k->w[j];
+
+ BURN(kk);
+}
+
+/* --- @square_eblk@, @square_dblk@ --- *
+ *
+ * Arguments: @const square_ctx *k@ = pointer to Square context
+ * @const uint32 s[4]@ = pointer to source block
+ * @uint32 d[4]@ = pointer to destination block
+ *
+ * Returns: ---
+ *
+ * Use: Low-level block encryption and decryption.
+ */
+
+#define EROUND(aa, bb, cc, dd, a, b, c, d, w) do { \
+ aa = (T[0][U8(a >> 0)] ^ T[1][U8(b >> 0)] ^ \
+ T[2][U8(c >> 0)] ^ T[3][U8(d >> 0)]) ^ *w++; \
+ bb = (T[0][U8(a >> 8)] ^ T[1][U8(b >> 8)] ^ \
+ T[2][U8(c >> 8)] ^ T[3][U8(d >> 8)]) ^ *w++; \
+ cc = (T[0][U8(a >> 16)] ^ T[1][U8(b >> 16)] ^ \
+ T[2][U8(c >> 16)] ^ T[3][U8(d >> 16)]) ^ *w++; \
+ dd = (T[0][U8(a >> 24)] ^ T[1][U8(b >> 24)] ^ \
+ T[2][U8(c >> 24)] ^ T[3][U8(d >> 24)]) ^ *w++; \
+} while (0)
+
+#define DROUND(aa, bb, cc, dd, a, b, c, d, w) do { \
+ aa = (TI[0][U8(a >> 0)] ^ TI[1][U8(b >> 0)] ^ \
+ TI[2][U8(c >> 0)] ^ TI[3][U8(d >> 0)]) ^ *w++; \
+ bb = (TI[0][U8(a >> 8)] ^ TI[1][U8(b >> 8)] ^ \
+ TI[2][U8(c >> 8)] ^ TI[3][U8(d >> 8)]) ^ *w++; \
+ cc = (TI[0][U8(a >> 16)] ^ TI[1][U8(b >> 16)] ^ \
+ TI[2][U8(c >> 16)] ^ TI[3][U8(d >> 16)]) ^ *w++; \
+ dd = (TI[0][U8(a >> 24)] ^ TI[1][U8(b >> 24)] ^ \
+ TI[2][U8(c >> 24)] ^ TI[3][U8(d >> 24)]) ^ *w++; \
+} while (0)
+
+void square_eblk(const square_ctx *k, const uint32 *s, uint32 *dst)
+{
+ uint32 a = s[0], b = s[1], c = s[2], d = s[3];
+ uint32 aa, bb, cc, dd;
+ uint32 *w = k->w;
+
+ a ^= *w++; b ^= *w++; c ^= *w++; d ^= *w++;
+
+ EROUND(aa, bb, cc, dd, a, b, c, d, w);
+ EROUND(a, b, c, d, aa, bb, cc, dd, w);
+ EROUND(aa, bb, cc, dd, a, b, c, d, w);
+ EROUND(a, b, c, d, aa, bb, cc, dd, w);
+ EROUND(aa, bb, cc, dd, a, b, c, d, w);
+ EROUND(a, b, c, d, aa, bb, cc, dd, w);
+ EROUND(aa, bb, cc, dd, a, b, c, d, w);
+
+ a = ((S[U8(aa >> 0)] << 0) ^ (S[U8(bb >> 0)] << 8) ^
+ (S[U8(cc >> 0)] << 16) ^ (S[U8(dd >> 0)] << 24)) ^ *w++;
+ b = ((S[U8(aa >> 8)] << 0) ^ (S[U8(bb >> 8)] << 8) ^
+ (S[U8(cc >> 8)] << 16) ^ (S[U8(dd >> 8)] << 24)) ^ *w++;
+ c = ((S[U8(aa >> 16)] << 0) ^ (S[U8(bb >> 16)] << 8) ^
+ (S[U8(cc >> 16)] << 16) ^ (S[U8(dd >> 16)] << 24)) ^ *w++;
+ d = ((S[U8(aa >> 24)] << 0) ^ (S[U8(bb >> 24)] << 8) ^
+ (S[U8(cc >> 24)] << 16) ^ (S[U8(dd >> 24)] << 24)) ^ *w++;
+
+ dst[0] = a; dst[1] = b; dst[2] = c; dst[3] = d;
+}
+
+void square_dblk(const square_ctx *k, const uint32 *s, uint32 *dst)
+{
+ uint32 a = s[0], b = s[1], c = s[2], d = s[3];
+ uint32 aa, bb, cc, dd;
+ uint32 *w = k->wi;
+
+ a ^= *w++; b ^= *w++; c ^= *w++; d ^= *w++;
+
+ DROUND(aa, bb, cc, dd, a, b, c, d, w);
+ DROUND(a, b, c, d, aa, bb, cc, dd, w);
+ DROUND(aa, bb, cc, dd, a, b, c, d, w);
+ DROUND(a, b, c, d, aa, bb, cc, dd, w);
+ DROUND(aa, bb, cc, dd, a, b, c, d, w);
+ DROUND(a, b, c, d, aa, bb, cc, dd, w);
+ DROUND(aa, bb, cc, dd, a, b, c, d, w);
+
+ a = ((SI[U8(aa >> 0)] << 0) ^ (SI[U8(bb >> 0)] << 8) ^
+ (SI[U8(cc >> 0)] << 16) ^ (SI[U8(dd >> 0)] << 24)) ^ *w++;
+ b = ((SI[U8(aa >> 8)] << 0) ^ (SI[U8(bb >> 8)] << 8) ^
+ (SI[U8(cc >> 8)] << 16) ^ (SI[U8(dd >> 8)] << 24)) ^ *w++;
+ c = ((SI[U8(aa >> 16)] << 0) ^ (SI[U8(bb >> 16)] << 8) ^
+ (SI[U8(cc >> 16)] << 16) ^ (SI[U8(dd >> 16)] << 24)) ^ *w++;
+ d = ((SI[U8(aa >> 24)] << 0) ^ (SI[U8(bb >> 24)] << 8) ^
+ (SI[U8(cc >> 24)] << 16) ^ (SI[U8(dd >> 24)] << 24)) ^ *w++;
+
+ dst[0] = a; dst[1] = b; dst[2] = c; dst[3] = d;
+}
+
+BLKC_TEST(SQUARE, square)
+
+/*----- That's all, folks -------------------------------------------------*/