| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <ddd8b0f3569a302d4242dafd155c7330dd4c7461.1412056800.git.eric.rost@mybabylon.net>
Date: Tue, 30 Sep 2014 01:13:51 -0500
From: Eric Rost <eric.rost@...abylon.net>
To: gregkh@...uxfoundation.org, jason@...edaemon.net, jake@....net,
antonysaraev@...il.com
Cc: devel@...verdev.osuosl.org, linux-kernel@...r.kernel.org
Subject: [PATCH v2 4/4] staging: skein: File Reorg
Reorganizes file to remove #defines from middle of functions. Also
removes #if'd loop declarations and adds ternary if driven loops.
Signed-off-by: Eric Rost <eric.rost@...abylon.net>
---
drivers/staging/skein/skein_block.c | 568 ++++++++++++++++++-----------------
1 file changed, 291 insertions(+), 277 deletions(-)
diff --git a/drivers/staging/skein/skein_block.c b/drivers/staging/skein/skein_block.c
index b8e4bd8..88bc718 100644
--- a/drivers/staging/skein/skein_block.c
+++ b/drivers/staging/skein/skein_block.c
@@ -122,8 +122,248 @@ do { \
SKEIN_256_ROUNDS_TOTAL / 8 > (NN)) || \
(SKEIN_UNROLL_256 > (NN)))
+#if (SKEIN_UNROLL_256 > 14)
+#error "need more unrolling in skein_256_process_block"
+#endif
+#endif
+
+#if !(SKEIN_USE_ASM & 512)
+#undef RCNT
+#define RCNT (SKEIN_512_ROUNDS_TOTAL/8)
+
+#ifdef SKEIN_LOOP /* configure how much to unroll the loop */
+#define SKEIN_UNROLL_512 (((SKEIN_LOOP)/10)%10)
+#else
+#define SKEIN_UNROLL_512 (0)
+#endif
+
+#if SKEIN_UNROLL_512
+#if (RCNT % SKEIN_UNROLL_512)
+#error "Invalid SKEIN_UNROLL_512" /* sanity check on unroll count */
+#endif
+#endif
+#define ROUND512(p0, p1, p2, p3, p4, p5, p6, p7, ROT, r_num) \
+do { \
+ X##p0 += X##p1; \
+ X##p1 = rotl_64(X##p1, ROT##_0); \
+ X##p1 ^= X##p0; \
+ X##p2 += X##p3; \
+ X##p3 = rotl_64(X##p3, ROT##_1); \
+ X##p3 ^= X##p2; \
+ X##p4 += X##p5; \
+ X##p5 = rotl_64(X##p5, ROT##_2); \
+ X##p5 ^= X##p4; \
+ X##p6 += X##p7; X##p7 = rotl_64(X##p7, ROT##_3); \
+ X##p7 ^= X##p6; \
+} while (0)
+
+#if SKEIN_UNROLL_512 == 0
+#define R512(p0, p1, p2, p3, p4, p5, p6, p7, ROT, r_num) /* unrolled */ \
+do { \
+ ROUND512(p0, p1, p2, p3, p4, p5, p6, p7, ROT, r_num) \
+ skein_show_r_ptr(BLK_BITS, &ctx->h, r_num, X_ptr); \
+} while (0)
+
+#define I512(R) \
+do { \
+ /* inject the key schedule value */ \
+ X0 += ks[((R) + 1) % 9]; \
+ X1 += ks[((R) + 2) % 9]; \
+ X2 += ks[((R) + 3) % 9]; \
+ X3 += ks[((R) + 4) % 9]; \
+ X4 += ks[((R) + 5) % 9]; \
+ X5 += ks[((R) + 6) % 9] + ts[((R) + 1) % 3]; \
+ X6 += ks[((R) + 7) % 9] + ts[((R) + 2) % 3]; \
+ X7 += ks[((R) + 8) % 9] + (R) + 1; \
+ skein_show_r_ptr(BLK_BITS, &ctx->h, SKEIN_RND_KEY_INJECT, X_ptr); \
+} while (0)
+
+#else /* looping version */
+#define R512(p0, p1, p2, p3, p4, p5, p6, p7, ROT, r_num) \
+do { \
+ ROUND512(p0, p1, p2, p3, p4, p5, p6, p7, ROT, r_num); \
+ skein_show_r_ptr(BLK_BITS, &ctx->h, 4 * (r - 1) + r_num, X_ptr); \
+} while (0)
+
+#define I512(R) \
+do { \
+ /* inject the key schedule value */ \
+ X0 += ks[r + (R) + 0]; \
+ X1 += ks[r + (R) + 1]; \
+ X2 += ks[r + (R) + 2]; \
+ X3 += ks[r + (R) + 3]; \
+ X4 += ks[r + (R) + 4]; \
+ X5 += ks[r + (R) + 5] + ts[r + (R) + 0]; \
+ X6 += ks[r + (R) + 6] + ts[r + (R) + 1]; \
+ X7 += ks[r + (R) + 7] + r + (R); \
+ /* rotate key schedule */ \
+ ks[r + (R) + 8] = ks[r + (R) - 1]; \
+ ts[r + (R) + 2] = ts[r + (R) - 1]; \
+ skein_show_r_ptr(BLK_BITS, &ctx->h, SKEIN_RND_KEY_INJECT, X_ptr); \
+} while (0)
+#endif /* end of looped code definitions */
+#define R512_8_ROUNDS(R) /* do 8 full rounds */ \
+do { \
+ R512(0, 1, 2, 3, 4, 5, 6, 7, R_512_0, 8 * (R) + 1); \
+ R512(2, 1, 4, 7, 6, 5, 0, 3, R_512_1, 8 * (R) + 2); \
+ R512(4, 1, 6, 3, 0, 5, 2, 7, R_512_2, 8 * (R) + 3); \
+ R512(6, 1, 0, 7, 2, 5, 4, 3, R_512_3, 8 * (R) + 4); \
+ I512(2 * (R)); \
+ R512(0, 1, 2, 3, 4, 5, 6, 7, R_512_4, 8 * (R) + 5); \
+ R512(2, 1, 4, 7, 6, 5, 0, 3, R_512_5, 8 * (R) + 6); \
+ R512(4, 1, 6, 3, 0, 5, 2, 7, R_512_6, 8 * (R) + 7); \
+ R512(6, 1, 0, 7, 2, 5, 4, 3, R_512_7, 8 * (R) + 8); \
+ I512(2 * (R) + 1); /* and key injection */ \
+} while (0)
+#define R512_UNROLL_R(NN) \
+ ((SKEIN_UNROLL_512 == 0 && \
+ SKEIN_512_ROUNDS_TOTAL/8 > (NN)) || \
+ (SKEIN_UNROLL_512 > (NN)))
+
+#if (SKEIN_UNROLL_512 > 14)
+#error "need more unrolling in skein_512_process_block"
+#endif
+#endif
+
+#if !(SKEIN_USE_ASM & 1024)
+#undef RCNT
+#define RCNT (SKEIN_1024_ROUNDS_TOTAL/8)
+#ifdef SKEIN_LOOP /* configure how much to unroll the loop */
+#define SKEIN_UNROLL_1024 ((SKEIN_LOOP)%10)
+#else
+#define SKEIN_UNROLL_1024 (0)
+#endif
+
+#if (SKEIN_UNROLL_1024 != 0)
+#if (RCNT % SKEIN_UNROLL_1024)
+#error "Invalid SKEIN_UNROLL_1024" /* sanity check on unroll count */
+#endif
+#endif
+#define ROUND1024(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pA, pB, pC, pD, pE, \
+ pF, ROT, r_num) \
+do { \
+ X##p0 += X##p1; \
+ X##p1 = rotl_64(X##p1, ROT##_0); \
+ X##p1 ^= X##p0; \
+ X##p2 += X##p3; \
+ X##p3 = rotl_64(X##p3, ROT##_1); \
+ X##p3 ^= X##p2; \
+ X##p4 += X##p5; \
+ X##p5 = rotl_64(X##p5, ROT##_2); \
+ X##p5 ^= X##p4; \
+ X##p6 += X##p7; \
+ X##p7 = rotl_64(X##p7, ROT##_3); \
+ X##p7 ^= X##p6; \
+ X##p8 += X##p9; \
+ X##p9 = rotl_64(X##p9, ROT##_4); \
+ X##p9 ^= X##p8; \
+ X##pA += X##pB; \
+ X##pB = rotl_64(X##pB, ROT##_5); \
+ X##pB ^= X##pA; \
+ X##pC += X##pD; \
+ X##pD = rotl_64(X##pD, ROT##_6); \
+ X##pD ^= X##pC; \
+ X##pE += X##pF; \
+ X##pF = rotl_64(X##pF, ROT##_7); \
+ X##pF ^= X##pE; \
+} while (0)
+
+#if SKEIN_UNROLL_1024 == 0
+#define R1024(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pA, pB, pC, pD, pE, pF, \
+ ROT, rn) \
+do { \
+ ROUND1024(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pA, pB, pC, pD, pE, \
+ pF, ROT, rn); \
+ skein_show_r_ptr(BLK_BITS, &ctx->h, rn, X_ptr); \
+} while (0)
+
+#define I1024(R) \
+do { \
+ /* inject the key schedule value */ \
+ X00 += ks[((R) + 1) % 17]; \
+ X01 += ks[((R) + 2) % 17]; \
+ X02 += ks[((R) + 3) % 17]; \
+ X03 += ks[((R) + 4) % 17]; \
+ X04 += ks[((R) + 5) % 17]; \
+ X05 += ks[((R) + 6) % 17]; \
+ X06 += ks[((R) + 7) % 17]; \
+ X07 += ks[((R) + 8) % 17]; \
+ X08 += ks[((R) + 9) % 17]; \
+ X09 += ks[((R) + 10) % 17]; \
+ X10 += ks[((R) + 11) % 17]; \
+ X11 += ks[((R) + 12) % 17]; \
+ X12 += ks[((R) + 13) % 17]; \
+ X13 += ks[((R) + 14) % 17] + ts[((R) + 1) % 3]; \
+ X14 += ks[((R) + 15) % 17] + ts[((R) + 2) % 3]; \
+ X15 += ks[((R) + 16) % 17] + (R) + 1; \
+ skein_show_r_ptr(BLK_BITS, &ctx->h, SKEIN_RND_KEY_INJECT, X_ptr); \
+} while (0)
+#else /* looping version */
+#define R1024(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pA, pB, pC, pD, pE, pF, \
+ ROT, rn) \
+do { \
+ ROUND1024(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pA, pB, pC, pD, pE, \
+ pF, ROT, rn); \
+ skein_show_r_ptr(BLK_BITS, &ctx->h, 4 * (r - 1) + rn, X_ptr); \
+} while (0)
+
+#define I1024(R) \
+do { \
+ /* inject the key schedule value */ \
+ X00 += ks[r + (R) + 0]; \
+ X01 += ks[r + (R) + 1]; \
+ X02 += ks[r + (R) + 2]; \
+ X03 += ks[r + (R) + 3]; \
+ X04 += ks[r + (R) + 4]; \
+ X05 += ks[r + (R) + 5]; \
+ X06 += ks[r + (R) + 6]; \
+ X07 += ks[r + (R) + 7]; \
+ X08 += ks[r + (R) + 8]; \
+ X09 += ks[r + (R) + 9]; \
+ X10 += ks[r + (R) + 10]; \
+ X11 += ks[r + (R) + 11]; \
+ X12 += ks[r + (R) + 12]; \
+ X13 += ks[r + (R) + 13] + ts[r + (R) + 0]; \
+ X14 += ks[r + (R) + 14] + ts[r + (R) + 1]; \
+ X15 += ks[r + (R) + 15] + r + (R); \
+ /* rotate key schedule */ \
+ ks[r + (R) + 16] = ks[r + (R) - 1]; \
+ ts[r + (R) + 2] = ts[r + (R) - 1]; \
+ skein_show_r_ptr(BLK_BITSi, &ctx->h, SKEIN_RND_KEY_INJECT, X_ptr); \
+} while (0)
+
#endif
+#define R1024_8_ROUNDS(R) \
+do { \
+ R1024(00, 01, 02, 03, 04, 05, 06, 07, 08, 09, 10, 11, 12, 13, 14, 15, \
+ R1024_0, 8*(R) + 1); \
+ R1024(00, 09, 02, 13, 06, 11, 04, 15, 10, 07, 12, 03, 14, 05, 08, 01, \
+ R1024_1, 8*(R) + 2); \
+ R1024(00, 07, 02, 05, 04, 03, 06, 01, 12, 15, 14, 13, 08, 11, 10, 09, \
+ R1024_2, 8*(R) + 3); \
+ R1024(00, 15, 02, 11, 06, 13, 04, 09, 14, 01, 08, 05, 10, 03, 12, 07, \
+ R1024_3, 8*(R) + 4); \
+ I1024(2*(R)); \
+ R1024(00, 01, 02, 03, 04, 05, 06, 07, 08, 09, 10, 11, 12, 13, 14, 15, \
+ R1024_4, 8*(R) + 5); \
+ R1024(00, 09, 02, 13, 06, 11, 04, 15, 10, 07, 12, 03, 14, 05, 08, 01, \
+ R1024_5, 8*(R) + 6); \
+ R1024(00, 07, 02, 05, 04, 03, 06, 01, 12, 15, 14, 13, 08, 11, 10, 09, \
+ R1024_6, 8*(R) + 7); \
+ R1024(00, 15, 02, 11, 06, 13, 04, 09, 14, 01, 08, 05, 10, 03, 12, 07, \
+ R1024_7, 8*(R) + 8); \
+ I1024(2*(R)+1); \
+} while (0)
+#define R1024_UNROLL_R(NN) \
+ ((SKEIN_UNROLL_1024 == 0 && \
+ SKEIN_1024_ROUNDS_TOTAL/8 > (NN)) || \
+ (SKEIN_UNROLL_1024 > (NN)))
+
+#if (SKEIN_UNROLL_1024 > 14)
+#error "need more unrolling in Skein_1024_Process_Block"
+#endif
+#endif
/***************************** SKEIN_256 ******************************/
#if !(SKEIN_USE_ASM & 256)
@@ -133,19 +373,23 @@ void skein_256_process_block(struct skein_256_ctx *ctx, const u8 *blk_ptr,
enum {
WCNT = SKEIN_256_STATE_WORDS
};
-
+ size_t r;
#if SKEIN_UNROLL_256
- size_t r;
- u64 kw[WCNT+4+RCNT*2]; /* key schedule: chaining vars + tweak + "rot"*/
+ /* key schedule: chaining vars + tweak + "rot"*/
+ u64 kw[WCNT+4+RCNT*2];
#else
- u64 kw[WCNT+4]; /* key schedule words : chaining vars + tweak */
+ /* key schedule words : chaining vars + tweak */
+ u64 kw[WCNT+4];
#endif
u64 X0, X1, X2, X3; /* local copy of context vars, for speed */
u64 w[WCNT]; /* local copy of input block */
#ifdef SKEIN_DEBUG
const u64 *X_ptr[4]; /* use for debugging (help cc put Xn in regs) */
- X_ptr[0] = &X0; X_ptr[1] = &X1; X_ptr[2] = &X2; X_ptr[3] = &X3;
+ X_ptr[0] = &X0;
+ X_ptr[1] = &X1;
+ X_ptr[2] = &X2;
+ X_ptr[3] = &X3;
#endif
skein_assert(blk_cnt != 0); /* never call with blk_cnt == 0! */
ts[0] = ctx->h.tweak[0];
@@ -184,58 +428,51 @@ void skein_256_process_block(struct skein_256_ctx *ctx, const u8 *blk_ptr,
blk_ptr += SKEIN_256_BLOCK_BYTES;
/* run the rounds */
-#if SKEIN_UNROLL_256
-
- for (r = 1; r < 2 * RCNT; r += 2 * SKEIN_UNROLL_256)
-#endif
- {
-
- R256_8_ROUNDS(0);
-
+ for (r = 1;
+ r < (SKEIN_UNROLL_256 ? 2 * RCNT : 2);
+ r += (SKEIN_UNROLL_256 ? 2 * SKEIN_UNROLL_256 : 1)) {
+ R256_8_ROUNDS(0);
#if R256_UNROLL_R(1)
- R256_8_ROUNDS(1);
+ R256_8_ROUNDS(1);
#endif
#if R256_UNROLL_R(2)
- R256_8_ROUNDS(2);
+ R256_8_ROUNDS(2);
#endif
#if R256_UNROLL_R(3)
- R256_8_ROUNDS(3);
+ R256_8_ROUNDS(3);
#endif
#if R256_UNROLL_R(4)
- R256_8_ROUNDS(4);
+ R256_8_ROUNDS(4);
#endif
#if R256_UNROLL_R(5)
- R256_8_ROUNDS(5);
+ R256_8_ROUNDS(5);
#endif
#if R256_UNROLL_R(6)
- R256_8_ROUNDS(6);
+ R256_8_ROUNDS(6);
#endif
#if R256_UNROLL_R(7)
- R256_8_ROUNDS(7);
+ R256_8_ROUNDS(7);
#endif
#if R256_UNROLL_R(8)
- R256_8_ROUNDS(8);
+ R256_8_ROUNDS(8);
#endif
#if R256_UNROLL_R(9)
- R256_8_ROUNDS(9);
+ R256_8_ROUNDS(9);
#endif
#if R256_UNROLL_R(10)
- R256_8_ROUNDS(10);
+ R256_8_ROUNDS(10);
#endif
#if R256_UNROLL_R(11)
- R256_8_ROUNDS(11);
+ R256_8_ROUNDS(11);
#endif
#if R256_UNROLL_R(12)
- R256_8_ROUNDS(12);
+ R256_8_ROUNDS(12);
#endif
#if R256_UNROLL_R(13)
- R256_8_ROUNDS(13);
+ R256_8_ROUNDS(13);
#endif
#if R256_UNROLL_R(14)
- R256_8_ROUNDS(14);
-#endif
-#if (SKEIN_UNROLL_256 > 14)
-#error "need more unrolling in skein_256_process_block"
+ R256_8_ROUNDS(14);
#endif
}
/* do the final "feedforward" xor, update context chaining */
@@ -273,20 +510,8 @@ void skein_512_process_block(struct skein_512_ctx *ctx, const u8 *blk_ptr,
enum {
WCNT = SKEIN_512_STATE_WORDS
};
-#undef RCNT
-#define RCNT (SKEIN_512_ROUNDS_TOTAL/8)
-
-#ifdef SKEIN_LOOP /* configure how much to unroll the loop */
-#define SKEIN_UNROLL_512 (((SKEIN_LOOP)/10)%10)
-#else
-#define SKEIN_UNROLL_512 (0)
-#endif
-
-#if SKEIN_UNROLL_512
-#if (RCNT % SKEIN_UNROLL_512)
-#error "Invalid SKEIN_UNROLL_512" /* sanity check on unroll count */
-#endif
size_t r;
+#if SKEIN_UNROLL_512
u64 kw[WCNT+4+RCNT*2]; /* key sched: chaining vars + tweak + "rot"*/
#else
u64 kw[WCNT+4]; /* key schedule words : chaining vars + tweak */
@@ -296,8 +521,14 @@ void skein_512_process_block(struct skein_512_ctx *ctx, const u8 *blk_ptr,
#ifdef SKEIN_DEBUG
const u64 *X_ptr[8]; /* use for debugging (help cc put Xn in regs) */
- X_ptr[0] = &X0; X_ptr[1] = &X1; X_ptr[2] = &X2; X_ptr[3] = &X3;
- X_ptr[4] = &X4; X_ptr[5] = &X5; X_ptr[6] = &X6; X_ptr[7] = &X7;
+ X_ptr[0] = &X0;
+ X_ptr[1] = &X1;
+ X_ptr[2] = &X2;
+ X_ptr[3] = &X3;
+ X_ptr[4] = &X4;
+ X_ptr[5] = &X5;
+ X_ptr[6] = &X6;
+ X_ptr[7] = &X7;
#endif
skein_assert(blk_cnt != 0); /* never call with blk_cnt == 0! */
@@ -330,104 +561,26 @@ void skein_512_process_block(struct skein_512_ctx *ctx, const u8 *blk_ptr,
skein_show_block(BLK_BITS, &ctx->h, ctx->x, blk_ptr, w, ks, ts);
/* do the first full key injection */
- X0 = w[0] + ks[0];
- X1 = w[1] + ks[1];
- X2 = w[2] + ks[2];
- X3 = w[3] + ks[3];
- X4 = w[4] + ks[4];
- X5 = w[5] + ks[5] + ts[0];
- X6 = w[6] + ks[6] + ts[1];
- X7 = w[7] + ks[7];
+ X0 = w[0] + ks[0];
+ X1 = w[1] + ks[1];
+ X2 = w[2] + ks[2];
+ X3 = w[3] + ks[3];
+ X4 = w[4] + ks[4];
+ X5 = w[5] + ks[5] + ts[0];
+ X6 = w[6] + ks[6] + ts[1];
+ X7 = w[7] + ks[7];
blk_ptr += SKEIN_512_BLOCK_BYTES;
skein_show_r_ptr(BLK_BITS, &ctx->h, SKEIN_RND_KEY_INITIAL,
X_ptr);
/* run the rounds */
-#define ROUND512(p0, p1, p2, p3, p4, p5, p6, p7, ROT, r_num) \
-do { \
- X##p0 += X##p1; \
- X##p1 = rotl_64(X##p1, ROT##_0); \
- X##p1 ^= X##p0; \
- X##p2 += X##p3; \
- X##p3 = rotl_64(X##p3, ROT##_1); \
- X##p3 ^= X##p2; \
- X##p4 += X##p5; \
- X##p5 = rotl_64(X##p5, ROT##_2); \
- X##p5 ^= X##p4; \
- X##p6 += X##p7; X##p7 = rotl_64(X##p7, ROT##_3); \
- X##p7 ^= X##p6; \
-} while (0)
-
-#if SKEIN_UNROLL_512 == 0
-#define R512(p0, p1, p2, p3, p4, p5, p6, p7, ROT, r_num) /* unrolled */ \
-do { \
- ROUND512(p0, p1, p2, p3, p4, p5, p6, p7, ROT, r_num) \
- skein_show_r_ptr(BLK_BITS, &ctx->h, r_num, X_ptr); \
-} while (0)
-
-#define I512(R) \
-do { \
- /* inject the key schedule value */ \
- X0 += ks[((R) + 1) % 9]; \
- X1 += ks[((R) + 2) % 9]; \
- X2 += ks[((R) + 3) % 9]; \
- X3 += ks[((R) + 4) % 9]; \
- X4 += ks[((R) + 5) % 9]; \
- X5 += ks[((R) + 6) % 9] + ts[((R) + 1) % 3]; \
- X6 += ks[((R) + 7) % 9] + ts[((R) + 2) % 3]; \
- X7 += ks[((R) + 8) % 9] + (R) + 1; \
- skein_show_r_ptr(BLK_BITS, &ctx->h, SKEIN_RND_KEY_INJECT, X_ptr); \
-} while (0)
-
-#else /* looping version */
-#define R512(p0, p1, p2, p3, p4, p5, p6, p7, ROT, r_num) \
-do { \
- ROUND512(p0, p1, p2, p3, p4, p5, p6, p7, ROT, r_num); \
- skein_show_r_ptr(BLK_BITS, &ctx->h, 4 * (r - 1) + r_num, X_ptr); \
-} while (0)
-
-#define I512(R) \
-do { \
- /* inject the key schedule value */ \
- X0 += ks[r + (R) + 0]; \
- X1 += ks[r + (R) + 1]; \
- X2 += ks[r + (R) + 2]; \
- X3 += ks[r + (R) + 3]; \
- X4 += ks[r + (R) + 4]; \
- X5 += ks[r + (R) + 5] + ts[r + (R) + 0]; \
- X6 += ks[r + (R) + 6] + ts[r + (R) + 1]; \
- X7 += ks[r + (R) + 7] + r + (R); \
- /* rotate key schedule */ \
- ks[r + (R) + 8] = ks[r + (R) - 1]; \
- ts[r + (R) + 2] = ts[r + (R) - 1]; \
- skein_show_r_ptr(BLK_BITS, &ctx->h, SKEIN_RND_KEY_INJECT, X_ptr); \
-} while (0)
-
- for (r = 1; r < 2 * RCNT; r += 2 * SKEIN_UNROLL_512)
-#endif /* end of looped code definitions */
- {
-#define R512_8_ROUNDS(R) /* do 8 full rounds */ \
-do { \
- R512(0, 1, 2, 3, 4, 5, 6, 7, R_512_0, 8 * (R) + 1); \
- R512(2, 1, 4, 7, 6, 5, 0, 3, R_512_1, 8 * (R) + 2); \
- R512(4, 1, 6, 3, 0, 5, 2, 7, R_512_2, 8 * (R) + 3); \
- R512(6, 1, 0, 7, 2, 5, 4, 3, R_512_3, 8 * (R) + 4); \
- I512(2 * (R)); \
- R512(0, 1, 2, 3, 4, 5, 6, 7, R_512_4, 8 * (R) + 5); \
- R512(2, 1, 4, 7, 6, 5, 0, 3, R_512_5, 8 * (R) + 6); \
- R512(4, 1, 6, 3, 0, 5, 2, 7, R_512_6, 8 * (R) + 7); \
- R512(6, 1, 0, 7, 2, 5, 4, 3, R_512_7, 8 * (R) + 8); \
- I512(2 * (R) + 1); /* and key injection */ \
-} while (0)
+ for (r = 1;
+ r < (SKEIN_UNROLL_512 ? 2 * RCNT : 2);
+ r += (SKEIN_UNROLL_512 ? 2 * SKEIN_UNROLL_512 : 1)) {
R512_8_ROUNDS(0);
-#define R512_UNROLL_R(NN) \
- ((SKEIN_UNROLL_512 == 0 && \
- SKEIN_512_ROUNDS_TOTAL/8 > (NN)) || \
- (SKEIN_UNROLL_512 > (NN)))
-
#if R512_UNROLL_R(1)
R512_8_ROUNDS(1);
#endif
@@ -470,9 +623,6 @@ do { \
#if R512_UNROLL_R(14)
R512_8_ROUNDS(14);
#endif
-#if (SKEIN_UNROLL_512 > 14)
-#error "need more unrolling in skein_512_process_block"
-#endif
}
/* do the final "feedforward" xor, update context chaining */
@@ -513,20 +663,8 @@ void skein_1024_process_block(struct skein_1024_ctx *ctx, const u8 *blk_ptr,
enum {
WCNT = SKEIN_1024_STATE_WORDS
};
-#undef RCNT
-#define RCNT (SKEIN_1024_ROUNDS_TOTAL/8)
-
-#ifdef SKEIN_LOOP /* configure how much to unroll the loop */
-#define SKEIN_UNROLL_1024 ((SKEIN_LOOP)%10)
-#else
-#define SKEIN_UNROLL_1024 (0)
-#endif
-
-#if (SKEIN_UNROLL_1024 != 0)
-#if (RCNT % SKEIN_UNROLL_1024)
-#error "Invalid SKEIN_UNROLL_1024" /* sanity check on unroll count */
-#endif
size_t r;
+#if (SKEIN_UNROLL_1024 != 0)
u64 kw[WCNT+4+RCNT*2]; /* key sched: chaining vars + tweak + "rot" */
#else
u64 kw[WCNT+4]; /* key schedule words : chaining vars + tweak */
@@ -617,131 +755,10 @@ void skein_1024_process_block(struct skein_1024_ctx *ctx, const u8 *blk_ptr,
skein_show_r_ptr(BLK_BITS, &ctx->h, SKEIN_RND_KEY_INITIAL,
X_ptr);
-#define ROUND1024(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pA, pB, pC, pD, pE, \
- pF, ROT, r_num) \
-do { \
- X##p0 += X##p1; \
- X##p1 = rotl_64(X##p1, ROT##_0); \
- X##p1 ^= X##p0; \
- X##p2 += X##p3; \
- X##p3 = rotl_64(X##p3, ROT##_1); \
- X##p3 ^= X##p2; \
- X##p4 += X##p5; \
- X##p5 = rotl_64(X##p5, ROT##_2); \
- X##p5 ^= X##p4; \
- X##p6 += X##p7; \
- X##p7 = rotl_64(X##p7, ROT##_3); \
- X##p7 ^= X##p6; \
- X##p8 += X##p9; \
- X##p9 = rotl_64(X##p9, ROT##_4); \
- X##p9 ^= X##p8; \
- X##pA += X##pB; \
- X##pB = rotl_64(X##pB, ROT##_5); \
- X##pB ^= X##pA; \
- X##pC += X##pD; \
- X##pD = rotl_64(X##pD, ROT##_6); \
- X##pD ^= X##pC; \
- X##pE += X##pF; \
- X##pF = rotl_64(X##pF, ROT##_7); \
- X##pF ^= X##pE; \
-} while (0)
-
-#if SKEIN_UNROLL_1024 == 0
-#define R1024(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pA, pB, pC, pD, pE, pF, \
- ROT, rn) \
-do { \
- ROUND1024(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pA, pB, pC, pD, pE, \
- pF, ROT, rn); \
- skein_show_r_ptr(BLK_BITS, &ctx->h, rn, X_ptr); \
-} while (0)
-
-#define I1024(R) \
-do { \
- /* inject the key schedule value */ \
- X00 += ks[((R) + 1) % 17]; \
- X01 += ks[((R) + 2) % 17]; \
- X02 += ks[((R) + 3) % 17]; \
- X03 += ks[((R) + 4) % 17]; \
- X04 += ks[((R) + 5) % 17]; \
- X05 += ks[((R) + 6) % 17]; \
- X06 += ks[((R) + 7) % 17]; \
- X07 += ks[((R) + 8) % 17]; \
- X08 += ks[((R) + 9) % 17]; \
- X09 += ks[((R) + 10) % 17]; \
- X10 += ks[((R) + 11) % 17]; \
- X11 += ks[((R) + 12) % 17]; \
- X12 += ks[((R) + 13) % 17]; \
- X13 += ks[((R) + 14) % 17] + ts[((R) + 1) % 3]; \
- X14 += ks[((R) + 15) % 17] + ts[((R) + 2) % 3]; \
- X15 += ks[((R) + 16) % 17] + (R) + 1; \
- skein_show_r_ptr(BLK_BITS, &ctx->h, SKEIN_RND_KEY_INJECT, X_ptr); \
-} while (0)
-#else /* looping version */
-#define R1024(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pA, pB, pC, pD, pE, pF, \
- ROT, rn) \
-do { \
- ROUND1024(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pA, pB, pC, pD, pE, \
- pF, ROT, rn); \
- skein_show_r_ptr(BLK_BITS, &ctx->h, 4 * (r - 1) + rn, X_ptr); \
-} while (0)
-
-#define I1024(R) \
-do { \
- /* inject the key schedule value */ \
- X00 += ks[r + (R) + 0]; \
- X01 += ks[r + (R) + 1]; \
- X02 += ks[r + (R) + 2]; \
- X03 += ks[r + (R) + 3]; \
- X04 += ks[r + (R) + 4]; \
- X05 += ks[r + (R) + 5]; \
- X06 += ks[r + (R) + 6]; \
- X07 += ks[r + (R) + 7]; \
- X08 += ks[r + (R) + 8]; \
- X09 += ks[r + (R) + 9]; \
- X10 += ks[r + (R) + 10]; \
- X11 += ks[r + (R) + 11]; \
- X12 += ks[r + (R) + 12]; \
- X13 += ks[r + (R) + 13] + ts[r + (R) + 0]; \
- X14 += ks[r + (R) + 14] + ts[r + (R) + 1]; \
- X15 += ks[r + (R) + 15] + r + (R); \
- /* rotate key schedule */ \
- ks[r + (R) + 16] = ks[r + (R) - 1]; \
- ts[r + (R) + 2] = ts[r + (R) - 1]; \
- skein_show_r_ptr(BLK_BITSi, &ctx->h, SKEIN_RND_KEY_INJECT, X_ptr); \
-} while (0)
-
- for (r = 1; r <= 2 * RCNT; r += 2 * SKEIN_UNROLL_1024)
-#endif
- {
-#define R1024_8_ROUNDS(R) \
-do { \
- R1024(00, 01, 02, 03, 04, 05, 06, 07, 08, 09, 10, 11, 12, 13, 14, 15, \
- R1024_0, 8*(R) + 1); \
- R1024(00, 09, 02, 13, 06, 11, 04, 15, 10, 07, 12, 03, 14, 05, 08, 01, \
- R1024_1, 8*(R) + 2); \
- R1024(00, 07, 02, 05, 04, 03, 06, 01, 12, 15, 14, 13, 08, 11, 10, 09, \
- R1024_2, 8*(R) + 3); \
- R1024(00, 15, 02, 11, 06, 13, 04, 09, 14, 01, 08, 05, 10, 03, 12, 07, \
- R1024_3, 8*(R) + 4); \
- I1024(2*(R)); \
- R1024(00, 01, 02, 03, 04, 05, 06, 07, 08, 09, 10, 11, 12, 13, 14, 15, \
- R1024_4, 8*(R) + 5); \
- R1024(00, 09, 02, 13, 06, 11, 04, 15, 10, 07, 12, 03, 14, 05, 08, 01, \
- R1024_5, 8*(R) + 6); \
- R1024(00, 07, 02, 05, 04, 03, 06, 01, 12, 15, 14, 13, 08, 11, 10, 09, \
- R1024_6, 8*(R) + 7); \
- R1024(00, 15, 02, 11, 06, 13, 04, 09, 14, 01, 08, 05, 10, 03, 12, 07, \
- R1024_7, 8*(R) + 8); \
- I1024(2*(R)+1); \
-} while (0)
-
+ for (r = 1;
+ r < (SKEIN_UNROLL_1024 ? 2 * RCNT : 2);
+ r += (SKEIN_UNROLL_1024 ? 2 * SKEIN_UNROLL_1024 : 1)) {
R1024_8_ROUNDS(0);
-
-#define R1024_UNROLL_R(NN) \
- ((SKEIN_UNROLL_1024 == 0 && \
- SKEIN_1024_ROUNDS_TOTAL/8 > (NN)) || \
- (SKEIN_UNROLL_1024 > (NN)))
-
#if R1024_UNROLL_R(1)
R1024_8_ROUNDS(1);
#endif
@@ -784,9 +801,6 @@ do { \
#if R1024_UNROLL_R(14)
R1024_8_ROUNDS(14);
#endif
-#if (SKEIN_UNROLL_1024 > 14)
-#error "need more unrolling in Skein_1024_Process_Block"
- #endif
}
/* do the final "feedforward" xor, update context chaining */
--
1.7.10.4
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists