propagate from branch 'net.randombit.botan.1_8' (head c5ae189464f6ef16e3ce73ea7c563412460d76a3)

            to branch 'net.randombit.botan' (head e2b95b6ad31c7539cf9ac0ebddb1d80bf63b5b21)

26 files changed, 300 insertions, 232 deletions

diff --git a/src/hash/bmw/bmw_512.cpp b/src/hash/bmw/bmw_512.cpp
new file mode 100644
index 000000000..ad3826351
--- /dev/null
+++ b/src/hash/bmw/bmw_512.cpp
@@ -0,0 +1,203 @@
+/*
+* Blue Midnight Wish 512 (Round 2 tweaked)
+* (C) 2009 Jack Lloyd
+*
+* Distributed under the terms of the Botan license
+*/
+
+#include <botan/bmw_512.h>
+#include <botan/loadstor.h>
+#include <botan/rotate.h>
+
+namespace Botan {
+
+namespace {
+
+inline u64bit S0(u64bit X)
+   {
+   return (X >> 1) ^ (X << 3) ^ rotate_left(X, 4) ^ rotate_left(X, 37);
+   }
+
+inline u64bit S1(u64bit X)
+   {
+   return (X >> 1) ^ (X << 2) ^ rotate_left(X, 13) ^ rotate_left(X, 43);
+   }
+
+inline u64bit S2(u64bit X)
+   {
+   return (X >> 2) ^ (X << 1) ^ rotate_left(X, 19) ^ rotate_left(X, 53);
+   }
+
+inline u64bit S3(u64bit X)
+   {
+   return (X >> 2) ^ (X << 2) ^ rotate_left(X, 28) ^ rotate_left(X, 59);
+   }
+
+inline u64bit S4(u64bit X)
+   {
+   return (X >> 1) ^ X;
+   }
+
+/**
+* Blue Midnight Wish 512 compression function
+*/
+void BMW_512_compress(u64bit H[16], const u64bit M[16], u64bit Q[32])
+   {
+   const u32bit EXPAND_1_ROUNDS = 2;
+
+   for(u32bit i = 0; i != 16; ++i)
+      Q[i] = H[i] ^ M[i];
+
+   Q[16] = Q[ 5] - Q[ 7] + Q[10] + Q[13] + Q[14];
+   Q[17] = Q[ 6] - Q[ 8] + Q[11] + Q[14] - Q[15];
+   Q[18] = Q[ 0] + Q[ 7] + Q[ 9] - Q[12] + Q[15];
+   Q[19] = Q[ 0] - Q[ 1] + Q[ 8] - Q[10] + Q[13];
+   Q[20] = Q[ 1] + Q[ 2] + Q[ 9] - Q[11] - Q[14];
+   Q[21] = Q[ 3] - Q[ 2] + Q[10] - Q[12] + Q[15];
+   Q[22] = Q[ 4] - Q[ 0] - Q[ 3] - Q[11] + Q[13];
+   Q[23] = Q[ 1] - Q[ 4] - Q[ 5] - Q[12] - Q[14];
+   Q[24] = Q[ 2] - Q[ 5] - Q[ 6] + Q[13] - Q[15];
+   Q[25] = Q[ 0] - Q[ 3] + Q[ 6] - Q[ 7] + Q[14];
+   Q[26] = Q[ 8] - Q[ 1] - Q[ 4] - Q[ 7] + Q[15];
+   Q[27] = Q[ 8] - Q[ 0] - Q[ 2] - Q[ 5] + Q[ 9];
+   Q[28] = Q[ 1] + Q[ 3] - Q[ 6] - Q[ 9] + Q[10];
+   Q[29] = Q[ 2] + Q[ 4] + Q[ 7] + Q[10] + Q[11];
+   Q[30] = Q[ 3] - Q[ 5] + Q[ 8] - Q[11] - Q[12];
+   Q[31] = Q[12] - Q[ 4] - Q[ 6] - Q[ 9] + Q[13];
+
+   Q[ 0] = S0(Q[16]) + H[ 1];
+   Q[ 1] = S1(Q[17]) + H[ 2];
+   Q[ 2] = S2(Q[18]) + H[ 3];
+   Q[ 3] = S3(Q[19]) + H[ 4];
+   Q[ 4] = S4(Q[20]) + H[ 5];
+   Q[ 5] = S0(Q[21]) + H[ 6];
+   Q[ 6] = S1(Q[22]) + H[ 7];
+   Q[ 7] = S2(Q[23]) + H[ 8];
+   Q[ 8] = S3(Q[24]) + H[ 9];
+   Q[ 9] = S4(Q[25]) + H[10];
+   Q[10] = S0(Q[26]) + H[11];
+   Q[11] = S1(Q[27]) + H[12];
+   Q[12] = S2(Q[28]) + H[13];
+   Q[13] = S3(Q[29]) + H[14];
+   Q[14] = S4(Q[30]) + H[15];
+   Q[15] = S0(Q[31]) + H[ 0];
+
+   for(u32bit i = 16; i != 16 + EXPAND_1_ROUNDS; ++i)
+      {
+      Q[i] = S1(Q[i-16]) + S2(Q[i-15]) + S3(Q[i-14]) + S0(Q[i-13]) +
+             S1(Q[i-12]) + S2(Q[i-11]) + S3(Q[i-10]) + S0(Q[i- 9]) +
+             S1(Q[i- 8]) + S2(Q[i- 7]) + S3(Q[i- 6]) + S0(Q[i- 5]) +
+             S1(Q[i- 4]) + S2(Q[i- 3]) + S3(Q[i- 2]) + S0(Q[i- 1]) +
+             ((rotate_left(M[(i-16) % 16], ((i-16)%16) + 1) +
+               rotate_left(M[(i-13) % 16], ((i-13)%16) + 1) -
+               rotate_left(M[(i- 6) % 16], ((i-6)%16) + 1) +
+               (0x0555555555555555 * i)) ^ H[(i-16+7)%16]);
+      }
+
+   for(u32bit i = 16 + EXPAND_1_ROUNDS; i != 32; ++i)
+      {
+      Q[i] = Q[i-16] + rotate_left(Q[i-15],  5) +
+             Q[i-14] + rotate_left(Q[i-13], 11) +
+             Q[i-12] + rotate_left(Q[i-11], 27) +
+             Q[i-10] + rotate_left(Q[i- 9], 32) +
+             Q[i- 8] + rotate_left(Q[i- 7], 37) +
+             Q[i- 6] + rotate_left(Q[i- 5], 43) +
+             Q[i- 4] + rotate_left(Q[i- 3], 53) +
+             S4(Q[i - 2]) + ((Q[i-1] >> 2) ^ Q[i-1]) +
+             ((rotate_left(M[(i-16) % 16], ((i-16)%16 + 1)) +
+               rotate_left(M[(i-13) % 16], ((i-13)%16 + 1)) -
+               rotate_left(M[(i- 6) % 16], ((i-6)%16 + 1)) +
+               (0x0555555555555555 * i)) ^ H[(i-16+7)%16]);
+      }
+
+   u64bit XL = Q[16] ^ Q[17] ^ Q[18] ^ Q[19] ^
+               Q[20] ^ Q[21] ^ Q[22] ^ Q[23];
+
+   u64bit XH = Q[24] ^ Q[25] ^ Q[26] ^ Q[27] ^
+               Q[28] ^ Q[29] ^ Q[30] ^ Q[31];
+
+   XH ^= XL;
+
+   H[ 0] = ((XH <<  5) ^ (Q[16] >> 5) ^ M[0]) + (XL ^ Q[24] ^ Q[0]);
+   H[ 1] = ((XH >>  7) ^ (Q[17] << 8) ^ M[1]) + (XL ^ Q[25] ^ Q[1]);
+   H[ 2] = ((XH >>  5) ^ (Q[18] << 5) ^ M[2]) + (XL ^ Q[26] ^ Q[2]);
+   H[ 3] = ((XH >>  1) ^ (Q[19] << 5) ^ M[3]) + (XL ^ Q[27] ^ Q[3]);
+   H[ 4] = ((XH >>  3) ^ (Q[20]     ) ^ M[4]) + (XL ^ Q[28] ^ Q[4]);
+   H[ 5] = ((XH <<  6) ^ (Q[21] >> 6) ^ M[5]) + (XL ^ Q[29] ^ Q[5]);
+   H[ 6] = ((XH >>  4) ^ (Q[22] << 6) ^ M[6]) + (XL ^ Q[30] ^ Q[6]);
+   H[ 7] = ((XH >> 11) ^ (Q[23] << 2) ^ M[7]) + (XL ^ Q[31] ^ Q[7]);
+
+   H[ 8] = rotate_left(H[4],  9) + (XH ^ Q[24] ^ M[ 8]) + ((XL << 8) ^ Q[23] ^ Q[ 8]);
+   H[ 9] = rotate_left(H[5], 10) + (XH ^ Q[25] ^ M[ 9]) + ((XL >> 6) ^ Q[16] ^ Q[ 9]);
+   H[10] = rotate_left(H[6], 11) + (XH ^ Q[26] ^ M[10]) + ((XL << 6) ^ Q[17] ^ Q[10]);
+   H[11] = rotate_left(H[7], 12) + (XH ^ Q[27] ^ M[11]) + ((XL << 4) ^ Q[18] ^ Q[11]);
+   H[12] = rotate_left(H[0], 13) + (XH ^ Q[28] ^ M[12]) + ((XL >> 3) ^ Q[19] ^ Q[12]);
+   H[13] = rotate_left(H[1], 14) + (XH ^ Q[29] ^ M[13]) + ((XL >> 4) ^ Q[20] ^ Q[13]);
+   H[14] = rotate_left(H[2], 15) + (XH ^ Q[30] ^ M[14]) + ((XL >> 7) ^ Q[21] ^ Q[14]);
+   H[15] = rotate_left(H[3], 16) + (XH ^ Q[31] ^ M[15]) + ((XL >> 2) ^ Q[22] ^ Q[15]);
+   }
+
+}
+
+void BMW_512::compress_n(const byte input[], u32bit blocks)
+   {
+   for(u32bit i = 0; i != blocks; ++i)
+      {
+      for(u32bit j = 0; j != 16; ++j)
+         M[j] = load_le<u64bit>(input, j);
+
+      BMW_512_compress(H, M, Q);
+
+      input += HASH_BLOCK_SIZE;
+      }
+   }
+
+/*
+* Copy out the digest
+*/
+void BMW_512::copy_out(byte output[])
+   {
+   u64bit final[16] = {
+      0xAAAAAAAAAAAAAAA0,  0xAAAAAAAAAAAAAAA1,
+      0xAAAAAAAAAAAAAAA2,  0xAAAAAAAAAAAAAAA3,
+      0xAAAAAAAAAAAAAAA4,  0xAAAAAAAAAAAAAAA5,
+      0xAAAAAAAAAAAAAAA6,  0xAAAAAAAAAAAAAAA7,
+      0xAAAAAAAAAAAAAAA8,  0xAAAAAAAAAAAAAAA9,
+      0xAAAAAAAAAAAAAAAA,  0xAAAAAAAAAAAAAAAB,
+      0xAAAAAAAAAAAAAAAC,  0xAAAAAAAAAAAAAAAD,
+      0xAAAAAAAAAAAAAAAE,  0xAAAAAAAAAAAAAAAF };
+
+   BMW_512_compress(final, H, Q);
+
+   for(u32bit i = 0; i != OUTPUT_LENGTH; i += 8)
+      store_le(final[8 + i/8], output + i);
+   }
+
+/*
+* Clear memory of sensitive data
+*/
+void BMW_512::clear() throw()
+   {
+   MDx_HashFunction::clear();
+   M.clear();
+   Q.clear();
+
+   H[ 0] = 0x8081828384858687;
+   H[ 1] = 0x88898A8B8C8D8E8F;
+   H[ 2] = 0x9091929394959697;
+   H[ 3] = 0x98999A9B9C9D9E9F;
+   H[ 4] = 0xA0A1A2A3A4A5A6A7;
+   H[ 5] = 0xA8A9AAABACADAEAF;
+   H[ 6] = 0xB0B1B2B3B4B5B6B7;
+   H[ 7] = 0xB8B9BABBBCBDBEBF;
+   H[ 8] = 0xC0C1C2C3C4C5C6C7;
+   H[ 9] = 0xC8C9CACBCCCDCECF;
+   H[10] = 0xD0D1D2D3D4D5D6D7;
+   H[11] = 0xD8D9DADBDCDDDEDF;
+   H[12] = 0xE0E1E2E3E4E5E6E7;
+   H[13] = 0xE8E9EAEBECEDEEEF;
+   H[14] = 0xF0F1F2F3F4F5F6F7;
+   H[15] = 0xF8F9FAFBFCFDFEFF;
+   }
+
+}
diff --git a/src/hash/bmw/bmw_512.h b/src/hash/bmw/bmw_512.h
new file mode 100644
index 000000000..d1f2539e9
--- /dev/null
+++ b/src/hash/bmw/bmw_512.h
@@ -0,0 +1,32 @@
+/*
+* Blue Midnight Wish 512 (Round 2 tweaked)
+* (C) 2009 Jack Lloyd
+*
+* Distributed under the terms of the Botan license
+*/
+
+#ifndef BOTAN_BMW_512_H__
+#define BOTAN_BMW_512_H__
+
+#include <botan/mdx_hash.h>
+
+namespace Botan {
+
+class BMW_512 : public MDx_HashFunction
+   {
+   public:
+      void clear() throw();
+      std::string name() const { return "BMW512"; }
+      HashFunction* clone() const { return new BMW_512; }
+      BMW_512() : MDx_HashFunction(64, 128, false, true) { clear(); }
+   private:
+      void compress_n(const byte input[], u32bit blocks);
+      void copy_out(byte output[]);
+
+      SecureBuffer<u64bit, 16> H, M;
+      SecureBuffer<u64bit, 32> Q;
+   };
+
+}
+
+#endif
diff --git a/src/hash/bmw/info.txt b/src/hash/bmw/info.txt
new file mode 100644
index 000000000..27e069c0a
--- /dev/null
+++ b/src/hash/bmw/info.txt
@@ -0,0 +1,7 @@
+realname "Blue Midnight Wish"
+
+define BMW_512
+
+<requires>
+mdx_hash
+</requires>
diff --git a/src/hash/fork256/info.txt b/src/hash/fork256/info.txt
index ae0c9f1a4..7c3c5bb94 100644
--- a/src/hash/fork256/info.txt
+++ b/src/hash/fork256/info.txt
@@ -2,13 +2,6 @@ realname "FORK-256"
 
 define FORK_256
 
-load_on auto
-
-<add>
-fork256.cpp
-fork256.h
-</add>
-
 <requires>
 mdx_hash
 </requires>
diff --git a/src/hash/gost_3411/gost_3411.cpp b/src/hash/gost_3411/gost_3411.cpp
index 3c7f5f92b..8f3982fca 100644
--- a/src/hash/gost_3411/gost_3411.cpp
+++ b/src/hash/gost_3411/gost_3411.cpp
@@ -1,6 +1,8 @@
 /*
 * GOST 34.11
 * (C) 2009 Jack Lloyd
+*
+* Distributed under the terms of the Botan license
 */
 
 #include <botan/gost_3411.h>
diff --git a/src/hash/gost_3411/gost_3411.h b/src/hash/gost_3411/gost_3411.h
index c69555052..960adaa44 100644
--- a/src/hash/gost_3411/gost_3411.h
+++ b/src/hash/gost_3411/gost_3411.h
@@ -1,6 +1,8 @@
 /**
 * GOST 34.11
 * (C) 2009 Jack Lloyd
+*
+* Distributed under the terms of the Botan license
 */
 
 #ifndef BOTAN_GOST_3411_H__
diff --git a/src/hash/gost_3411/info.txt b/src/hash/gost_3411/info.txt
index 65b9475e1..353ceb3e5 100644
--- a/src/hash/gost_3411/info.txt
+++ b/src/hash/gost_3411/info.txt
@@ -2,13 +2,6 @@ realname "GOST 34.11"
 
 define GOST_34_11
 
-load_on auto
-
-<add>
-gost_3411.cpp
-gost_3411.h
-</add>
-
 <requires>
 gost_28147
 </requires>
diff --git a/src/hash/has160/info.txt b/src/hash/has160/info.txt
index a945f3225..98cbc5155 100644
--- a/src/hash/has160/info.txt
+++ b/src/hash/has160/info.txt
@@ -2,13 +2,6 @@ realname "HAS-160"
 
 define HAS_160
 
-load_on auto
-
-<add>
-has160.cpp
-has160.h
-</add>
-
 <requires>
 mdx_hash
 </requires>
diff --git a/src/hash/md2/info.txt b/src/hash/md2/info.txt
index ff33e1eb3..4428584fd 100644
--- a/src/hash/md2/info.txt
+++ b/src/hash/md2/info.txt
@@ -1,10 +1,3 @@
 realname "MD2"
 
 define MD2
-
-load_on auto
-
-<add>
-md2.cpp
-md2.h
-</add>
diff --git a/src/hash/md4/info.txt b/src/hash/md4/info.txt
index fc9cbe116..6aa4a5d59 100644
--- a/src/hash/md4/info.txt
+++ b/src/hash/md4/info.txt
@@ -2,13 +2,6 @@ realname "MD4"
 
 define MD4
 
-load_on auto
-
-<add>
-md4.cpp
-md4.h
-</add>
-
 <requires>
 mdx_hash
 </requires>
diff --git a/src/hash/md4_ia32/info.txt b/src/hash/md4_ia32/info.txt
index fee7dd1a8..217c3fc21 100644
--- a/src/hash/md4_ia32/info.txt
+++ b/src/hash/md4_ia32/info.txt
@@ -4,12 +4,6 @@ define MD4_IA32
 
 load_on asm_ok
 
-<add>
-md4_ia32_imp.S
-md4_ia32.cpp
-md4_ia32.h
-</add>
-
 <arch>
 ia32
 </arch>
diff --git a/src/hash/md5/info.txt b/src/hash/md5/info.txt
index 525a45ab5..14861d635 100644
--- a/src/hash/md5/info.txt
+++ b/src/hash/md5/info.txt
@@ -2,13 +2,6 @@ realname "MD5"
 
 define MD5
 
-load_on auto
-
-<add>
-md5.cpp
-md5.h
-</add>
-
 <requires>
 mdx_hash
 </requires>
diff --git a/src/hash/md5_ia32/info.txt b/src/hash/md5_ia32/info.txt
index ad9923b26..d298f2005 100644
--- a/src/hash/md5_ia32/info.txt
+++ b/src/hash/md5_ia32/info.txt
@@ -4,12 +4,6 @@ define MD5_IA32
 
 load_on asm_ok
 
-<add>
-md5_ia32_imp.S
-md5_ia32.cpp
-md5_ia32.h
-</add>
-
 <arch>
 ia32
 </arch>
diff --git a/src/hash/mdx_hash/info.txt b/src/hash/mdx_hash/info.txt
index 412c93350..ee4ccd6da 100644
--- a/src/hash/mdx_hash/info.txt
+++ b/src/hash/mdx_hash/info.txt
@@ -3,8 +3,3 @@ realname "MDx Hash Base"
 define MDX_HASH_FUNCTION
 
 load_on dep
-
-<add>
-mdx_hash.cpp
-mdx_hash.h
-</add>
diff --git a/src/hash/par_hash/info.txt b/src/hash/par_hash/info.txt
index 45716aac8..029939cb5 100644
--- a/src/hash/par_hash/info.txt
+++ b/src/hash/par_hash/info.txt
@@ -1,10 +1,3 @@
 realname "Parallel Hash"
 
 define PARALLEL_HASH
-
-load_on auto
-
-<add>
-par_hash.cpp
-par_hash.h
-</add>
diff --git a/src/hash/rmd128/info.txt b/src/hash/rmd128/info.txt
index 402271da5..0a1ef2c74 100644
--- a/src/hash/rmd128/info.txt
+++ b/src/hash/rmd128/info.txt
@@ -2,13 +2,6 @@ realname "RIPEMD-128"
 
 define RIPEMD_128
 
-load_on auto
-
-<add>
-rmd128.cpp
-rmd128.h
-</add>
-
 <requires>
 mdx_hash
 </requires>
diff --git a/src/hash/rmd160/info.txt b/src/hash/rmd160/info.txt
index af4b5c230..225106afc 100644
--- a/src/hash/rmd160/info.txt
+++ b/src/hash/rmd160/info.txt
@@ -2,13 +2,6 @@ realname "RIPEMD-160"
 
 define RIPEMD_160
 
-load_on auto
-
-<add>
-rmd160.cpp
-rmd160.h
-</add>
-
 <requires>
 mdx_hash
 </requires>
diff --git a/src/hash/sha1/info.txt b/src/hash/sha1/info.txt
index a0ae30b4b..8915f9ebd 100644
--- a/src/hash/sha1/info.txt
+++ b/src/hash/sha1/info.txt
@@ -2,13 +2,6 @@ realname "SHA-1"
 
 define SHA1
 
-load_on auto
-
-<add>
-sha160.cpp
-sha160.h
-</add>
-
 <requires>
 mdx_hash
 </requires>
diff --git a/src/hash/sha1_amd64/info.txt b/src/hash/sha1_amd64/info.txt
index ddbc7d380..183658ac7 100644
--- a/src/hash/sha1_amd64/info.txt
+++ b/src/hash/sha1_amd64/info.txt
@@ -4,12 +4,6 @@ define SHA1_AMD64
 
 load_on asm_ok
 
-<add>
-sha1_amd64_imp.S
-sha1_amd64.cpp
-sha1_amd64.h
-</add>
-
 <arch>
 amd64
 </arch>
diff --git a/src/hash/sha1_ia32/info.txt b/src/hash/sha1_ia32/info.txt
index bfb321145..e7709454f 100644
--- a/src/hash/sha1_ia32/info.txt
+++ b/src/hash/sha1_ia32/info.txt
@@ -4,12 +4,6 @@ define SHA1_IA32
 
 load_on asm_ok
 
-<add>
-sha1_ia32_imp.S
-sha1_ia32.cpp
-sha1_ia32.h
-</add>
-
 <arch>
 ia32
 </arch>
diff --git a/src/hash/sha1_sse2/info.txt b/src/hash/sha1_sse2/info.txt
index b8d693b70..4a7a1b5e9 100644
--- a/src/hash/sha1_sse2/info.txt
+++ b/src/hash/sha1_sse2/info.txt
@@ -2,26 +2,6 @@ realname "SHA-1 (SSE2)"
 
 define SHA1_SSE2
 
-load_on auto
-
-<add>
-sha1_sse2_imp.cpp
-sha1_sse2.cpp
-sha1_sse2.h
-</add>
-
-<arch>
-pentium-m
-pentium4
-prescott
-amd64
-</arch>
-
-<cc>
-gcc
-icc
-</cc>
-
 <requires>
 sha1
 sse2_eng
diff --git a/src/hash/sha1_sse2/sha1_sse2_imp.cpp b/src/hash/sha1_sse2/sha1_sse2_imp.cpp
index 90a8dccd5..5ee222cca 100644
--- a/src/hash/sha1_sse2/sha1_sse2_imp.cpp
+++ b/src/hash/sha1_sse2/sha1_sse2_imp.cpp
@@ -1,27 +1,18 @@
-/* this code is public domain.
-
- * dean gaudet <[email protected]>
-
- * this code was inspired by this paper:
-
- *     SHA: A Design for Parallel Architectures?
- *     Antoon Bosselaers, Ren�e Govaerts and Joos Vandewalle
- *     <http://www.esat.kuleuven.ac.be/~cosicart/pdf/AB-9700.pdf>
-
- * more information available on this implementation here:
-
- * 	http://arctic.org/~dean/crypto/sha1.html
-
- * version: 2
- */
+/*
+* SHA-1 using SSE2
+* (C) 2009 Jack Lloyd
+*
+* Distributed under the terms of the Botan license
+*/
 
 /*
- * Lightly modified for Botan, tested under GCC 4.1.1 and ICC 9.1
- * on a Linux/Core2 system.
+* Based on public domain code by Dean Gaudet <[email protected]>
+*   http://arctic.org/~dean/crypto/sha1.html
+*/
 
- */
 #include <botan/sha1_sse2.h>
-#include <xmmintrin.h>
+#include <botan/rotate.h>
+#include <emmintrin.h>
 
 namespace Botan {
 
@@ -30,21 +21,13 @@ namespace {
 typedef union {
    u32bit u32[4];
    __m128i u128;
-   } v4si __attribute__((aligned(16)));
+   } v4si;
 
 static const v4si K00_19 = { { 0x5a827999, 0x5a827999, 0x5a827999, 0x5a827999 } };
 static const v4si K20_39 = { { 0x6ed9eba1, 0x6ed9eba1, 0x6ed9eba1, 0x6ed9eba1 } };
 static const v4si K40_59 = { { 0x8f1bbcdc, 0x8f1bbcdc, 0x8f1bbcdc, 0x8f1bbcdc } };
 static const v4si K60_79 = { { 0xca62c1d6, 0xca62c1d6, 0xca62c1d6, 0xca62c1d6 } };
 
-#define UNALIGNED 1
-#if UNALIGNED
-#define load(p)	_mm_loadu_si128(p)
-#else
-#define load(p) (*p)
-#endif
-
-
 /*
 the first 16 bytes only need byte swapping
 
@@ -57,19 +40,15 @@ and is overwritten with the swapped bytes
 		__m128i r1, r2;						\
 									\
 		r1 = (W);						\
-		if (1) {						\
 		r1 = _mm_shufflehi_epi16(r1, _MM_SHUFFLE(2, 3, 0, 1));	\
 		r1 = _mm_shufflelo_epi16(r1, _MM_SHUFFLE(2, 3, 0, 1));	\
 		r2 = _mm_slli_epi16(r1, 8);				\
 		r1 = _mm_srli_epi16(r1, 8);				\
 		r1 = _mm_or_si128(r1, r2);				\
 		(W) = r1;						\
-		}							\
 		(prep).u128 = _mm_add_epi32(K00_19.u128, r1);		\
 	} while(0)
 
-
-
 /*
 for each multiple of 4, t, we want to calculate this:
 
@@ -115,45 +94,38 @@ W0 = W[t]..W[t+3]
 * efficeon, pentium-m, and opteron but shifts are available in
 * only one unit.
 */
-#define prep(prep, XW0, XW1, XW2, XW3, K) do { 					\
-		__m128i r0, r1, r2, r3;						\
-										\
-		/* load W[t-4] 16-byte aligned, and shift */			\
-		r3 = _mm_srli_si128((XW3), 4);					\
-		r0 = (XW0);							\
-		/* get high 64-bits of XW0 into low 64-bits */			\
-		r1 = _mm_shuffle_epi32((XW0), _MM_SHUFFLE(1,0,3,2));		\
-		/* load high 64-bits of r1 */					\
-		r1 = _mm_unpacklo_epi64(r1, (XW1));				\
-		r2 = (XW2);							\
-										\
-		r0 = _mm_xor_si128(r1, r0);					\
-		r2 = _mm_xor_si128(r3, r2);					\
-		r0 = _mm_xor_si128(r2, r0);					\
-		/* unrotated W[t]..W[t+2] in r0 ... still need W[t+3] */	\
-										\
-		r2 = _mm_slli_si128(r0, 12);					\
-		r1 = _mm_cmplt_epi32(r0, _mm_setzero_si128());			\
-		r0 = _mm_add_epi32(r0, r0);	/* shift left by 1 */		\
-		r0 = _mm_sub_epi32(r0, r1);	/* r0 has W[t]..W[t+2] */	\
-										\
-		r3 = _mm_srli_epi32(r2, 30);					\
-		r2 = _mm_slli_epi32(r2, 2);					\
-										\
-		r0 = _mm_xor_si128(r0, r3);					\
-		r0 = _mm_xor_si128(r0, r2);	/* r0 now has W[t+3] */		\
-										\
-		(XW0) = r0;							\
-		(prep).u128 = _mm_add_epi32(r0, (K).u128);			\
-	} while(0)
-
-
-static inline u32bit rol(u32bit src, u32bit amt)
-   {
-   /* gcc and icc appear to turn this into a rotate */
-   return (src << amt) | (src >> (32 - amt));
-   }
-
+#define prep(prep, XW0, XW1, XW2, XW3, K)                               \
+   do {                                                                 \
+      __m128i r0, r1, r2, r3;						\
+                                                                        \
+      /* load W[t-4] 16-byte aligned, and shift */                      \
+      r3 = _mm_srli_si128((XW3), 4);					\
+      r0 = (XW0);                                                       \
+      /* get high 64-bits of XW0 into low 64-bits */			\
+      r1 = _mm_shuffle_epi32((XW0), _MM_SHUFFLE(1,0,3,2));              \
+      /* load high 64-bits of r1 */					\
+      r1 = _mm_unpacklo_epi64(r1, (XW1));                               \
+      r2 = (XW2);                                                       \
+                                                                        \
+      r0 = _mm_xor_si128(r1, r0);                                       \
+      r2 = _mm_xor_si128(r3, r2);                                       \
+      r0 = _mm_xor_si128(r2, r0);                                       \
+      /* unrotated W[t]..W[t+2] in r0 ... still need W[t+3] */          \
+                                                                        \
+      r2 = _mm_slli_si128(r0, 12);                                      \
+      r1 = _mm_cmplt_epi32(r0, _mm_setzero_si128());			\
+      r0 = _mm_add_epi32(r0, r0);	/* shift left by 1 */           \
+      r0 = _mm_sub_epi32(r0, r1);	/* r0 has W[t]..W[t+2] */       \
+                                                                        \
+      r3 = _mm_srli_epi32(r2, 30);                                      \
+      r2 = _mm_slli_epi32(r2, 2);                                       \
+                                                                        \
+      r0 = _mm_xor_si128(r0, r3);                                       \
+      r0 = _mm_xor_si128(r0, r2);	/* r0 now has W[t+3] */         \
+                                                                        \
+      (XW0) = r0;                                                       \
+      (prep).u128 = _mm_add_epi32(r0, (K).u128);                        \
+   } while(0)
 
 static inline u32bit f00_19(u32bit x, u32bit y, u32bit z)
    {
@@ -184,11 +156,12 @@ static inline u32bit f60_79(u32bit x, u32bit y, u32bit z)
    return f20_39(x, y, z);
    }
 
-#define step(nn_mm, xa, xb, xc, xd, xe, xt, input) do {					\
-		(xt) = (input) + f##nn_mm((xb), (xc), (xd));				\
-		(xb) = rol((xb), 30); 							\
-		(xt) += ((xe) + rol((xa), 5));						\
-	} while(0)
+#define step(nn_mm, xa, xb, xc, xd, xe, xt, input)                      \
+   do {                                                                 \
+      (xt) = (input) + f##nn_mm((xb), (xc), (xd));			\
+      (xb) = rotate_left((xb), 30);                                     \
+      (xt) += ((xe) + rotate_left((xa), 5));                            \
+   } while(0)
 
 }
 
@@ -210,14 +183,14 @@ extern "C" void botan_sha1_sse2_compress(u32bit H[5],
    * steps ahead of the integer code.  12 steps ahead seems
    * to produce the best performance. -dean
    */
-   W0 = load(&input[0]);
+   W0 = _mm_loadu_si128(&input[0]);
    prep00_15(prep0, W0);				/* prepare for 00 through 03 */
-   W1 = load(&input[1]);
+   W1 = _mm_loadu_si128(&input[1]);
    prep00_15(prep1, W1);				/* prepare for 04 through 07 */
-   W2 = load(&input[2]);
+   W2 = _mm_loadu_si128(&input[2]);
    prep00_15(prep2, W2);				/* prepare for 08 through 11 */
 
-   W3 = load(&input[3]);
+   W3 = _mm_loadu_si128(&input[3]);
    step(00_19, a, b, c, d, e, t, prep0.u32[0]);	/* 00 */
    step(00_19, t, a, b, c, d, e, prep0.u32[1]);	/* 01 */
    step(00_19, e, t, a, b, c, d, prep0.u32[2]);	/* 02 */
diff --git a/src/hash/sha2/info.txt b/src/hash/sha2/info.txt
index 9b92ff385..8cc50fa41 100644
--- a/src/hash/sha2/info.txt
+++ b/src/hash/sha2/info.txt
@@ -2,15 +2,6 @@ realname "SHA-2 (224, 256, 384, 512)"
 
 define SHA2
 
-load_on auto
-
-<add>
-sha2_32.cpp
-sha2_32.h
-sha2_64.cpp
-sha2_64.h
-</add>
-
 <requires>
 mdx_hash
 </requires>
diff --git a/src/hash/skein/info.txt b/src/hash/skein/info.txt
index bab8497c5..908033852 100644
--- a/src/hash/skein/info.txt
+++ b/src/hash/skein/info.txt
@@ -2,13 +2,6 @@ realname "Skein"
 
 define SKEIN_512
 
-load_on auto
-
-<add>
-skein_512.cpp
-skein_512.h
-</add>
-
 <requires>
 alloc
 </requires>
diff --git a/src/hash/tiger/info.txt b/src/hash/tiger/info.txt
index 7d4cd7177..92b5519c0 100644
--- a/src/hash/tiger/info.txt
+++ b/src/hash/tiger/info.txt
@@ -2,14 +2,6 @@ realname "Tiger"
 
 define TIGER
 
-load_on auto
-
-<add>
-tig_tab.cpp
-tiger.cpp
-tiger.h
-</add>
-
 <requires>
 mdx_hash
 </requires>
diff --git a/src/hash/whirlpool/info.txt b/src/hash/whirlpool/info.txt
index be55b5c3b..4fe4b2b25 100644
--- a/src/hash/whirlpool/info.txt
+++ b/src/hash/whirlpool/info.txt
@@ -2,14 +2,6 @@ realname "Whirlpool"
 
 define WHIRLPOOL
 
-load_on auto
-
-<add>
-whrl_tab.cpp
-whrlpool.cpp
-whrlpool.h
-</add>
-
 <requires>
 mdx_hash
 </requires>