4d4d77e
openssl/crypto/aes/aes_core.c     1.7 -> 1.7.2.1
4d4d77e
4d4d77e
--- openssl/crypto/aes/aes_core.c 2005/01/24 14:22:05 1.7
4d4d77e
+++ openssl/crypto/aes/aes_core.c 2006/06/28 08:58:15 1.7.2.1
4d4d77e
@@ -44,22 +44,14 @@
4d4d77e
 Te1[x] = S [x].[03, 02, 01, 01];
4d4d77e
 Te2[x] = S [x].[01, 03, 02, 01];
4d4d77e
 Te3[x] = S [x].[01, 01, 03, 02];
4d4d77e
-Te4[x] = S [x].[01, 01, 01, 01];
4d4d77e
 
4d4d77e
 Td0[x] = Si[x].[0e, 09, 0d, 0b];
4d4d77e
 Td1[x] = Si[x].[0b, 0e, 09, 0d];
4d4d77e
 Td2[x] = Si[x].[0d, 0b, 0e, 09];
4d4d77e
 Td3[x] = Si[x].[09, 0d, 0b, 0e];
4d4d77e
-Td4[x] = Si[x].[01, 01, 01, 01];
4d4d77e
+Td4[x] = Si[x].[01];
4d4d77e
 */
4d4d77e
 
4d4d77e
-#ifdef AES_ASM
4d4d77e
-extern const u32 AES_Te[5][256];
4d4d77e
-#define Te0 AES_Te[0]
4d4d77e
-#define Te1 AES_Te[1]
4d4d77e
-#define Te2 AES_Te[2]
4d4d77e
-#define Te3 AES_Te[3]
4d4d77e
-#else
4d4d77e
 static const u32 Te0[256] = {
4d4d77e
     0xc66363a5U, 0xf87c7c84U, 0xee777799U, 0xf67b7b8dU,
4d4d77e
     0xfff2f20dU, 0xd66b6bbdU, 0xde6f6fb1U, 0x91c5c554U,
4d4d77e
@@ -324,81 +316,7 @@
4d4d77e
     0x4141c382U, 0x9999b029U, 0x2d2d775aU, 0x0f0f111eU,
4d4d77e
     0xb0b0cb7bU, 0x5454fca8U, 0xbbbbd66dU, 0x16163a2cU,
4d4d77e
 };
4d4d77e
-#endif
4d4d77e
-static const u32 Te4[256] = {
4d4d77e
-    0x63636363U, 0x7c7c7c7cU, 0x77777777U, 0x7b7b7b7bU,
4d4d77e
-    0xf2f2f2f2U, 0x6b6b6b6bU, 0x6f6f6f6fU, 0xc5c5c5c5U,
4d4d77e
-    0x30303030U, 0x01010101U, 0x67676767U, 0x2b2b2b2bU,
4d4d77e
-    0xfefefefeU, 0xd7d7d7d7U, 0xababababU, 0x76767676U,
4d4d77e
-    0xcacacacaU, 0x82828282U, 0xc9c9c9c9U, 0x7d7d7d7dU,
4d4d77e
-    0xfafafafaU, 0x59595959U, 0x47474747U, 0xf0f0f0f0U,
4d4d77e
-    0xadadadadU, 0xd4d4d4d4U, 0xa2a2a2a2U, 0xafafafafU,
4d4d77e
-    0x9c9c9c9cU, 0xa4a4a4a4U, 0x72727272U, 0xc0c0c0c0U,
4d4d77e
-    0xb7b7b7b7U, 0xfdfdfdfdU, 0x93939393U, 0x26262626U,
4d4d77e
-    0x36363636U, 0x3f3f3f3fU, 0xf7f7f7f7U, 0xccccccccU,
4d4d77e
-    0x34343434U, 0xa5a5a5a5U, 0xe5e5e5e5U, 0xf1f1f1f1U,
4d4d77e
-    0x71717171U, 0xd8d8d8d8U, 0x31313131U, 0x15151515U,
4d4d77e
-    0x04040404U, 0xc7c7c7c7U, 0x23232323U, 0xc3c3c3c3U,
4d4d77e
-    0x18181818U, 0x96969696U, 0x05050505U, 0x9a9a9a9aU,
4d4d77e
-    0x07070707U, 0x12121212U, 0x80808080U, 0xe2e2e2e2U,
4d4d77e
-    0xebebebebU, 0x27272727U, 0xb2b2b2b2U, 0x75757575U,
4d4d77e
-    0x09090909U, 0x83838383U, 0x2c2c2c2cU, 0x1a1a1a1aU,
4d4d77e
-    0x1b1b1b1bU, 0x6e6e6e6eU, 0x5a5a5a5aU, 0xa0a0a0a0U,
4d4d77e
-    0x52525252U, 0x3b3b3b3bU, 0xd6d6d6d6U, 0xb3b3b3b3U,
4d4d77e
-    0x29292929U, 0xe3e3e3e3U, 0x2f2f2f2fU, 0x84848484U,
4d4d77e
-    0x53535353U, 0xd1d1d1d1U, 0x00000000U, 0xededededU,
4d4d77e
-    0x20202020U, 0xfcfcfcfcU, 0xb1b1b1b1U, 0x5b5b5b5bU,
4d4d77e
-    0x6a6a6a6aU, 0xcbcbcbcbU, 0xbebebebeU, 0x39393939U,
4d4d77e
-    0x4a4a4a4aU, 0x4c4c4c4cU, 0x58585858U, 0xcfcfcfcfU,
4d4d77e
-    0xd0d0d0d0U, 0xefefefefU, 0xaaaaaaaaU, 0xfbfbfbfbU,
4d4d77e
-    0x43434343U, 0x4d4d4d4dU, 0x33333333U, 0x85858585U,
4d4d77e
-    0x45454545U, 0xf9f9f9f9U, 0x02020202U, 0x7f7f7f7fU,
4d4d77e
-    0x50505050U, 0x3c3c3c3cU, 0x9f9f9f9fU, 0xa8a8a8a8U,
4d4d77e
-    0x51515151U, 0xa3a3a3a3U, 0x40404040U, 0x8f8f8f8fU,
4d4d77e
-    0x92929292U, 0x9d9d9d9dU, 0x38383838U, 0xf5f5f5f5U,
4d4d77e
-    0xbcbcbcbcU, 0xb6b6b6b6U, 0xdadadadaU, 0x21212121U,
4d4d77e
-    0x10101010U, 0xffffffffU, 0xf3f3f3f3U, 0xd2d2d2d2U,
4d4d77e
-    0xcdcdcdcdU, 0x0c0c0c0cU, 0x13131313U, 0xececececU,
4d4d77e
-    0x5f5f5f5fU, 0x97979797U, 0x44444444U, 0x17171717U,
4d4d77e
-    0xc4c4c4c4U, 0xa7a7a7a7U, 0x7e7e7e7eU, 0x3d3d3d3dU,
4d4d77e
-    0x64646464U, 0x5d5d5d5dU, 0x19191919U, 0x73737373U,
4d4d77e
-    0x60606060U, 0x81818181U, 0x4f4f4f4fU, 0xdcdcdcdcU,
4d4d77e
-    0x22222222U, 0x2a2a2a2aU, 0x90909090U, 0x88888888U,
4d4d77e
-    0x46464646U, 0xeeeeeeeeU, 0xb8b8b8b8U, 0x14141414U,
4d4d77e
-    0xdedededeU, 0x5e5e5e5eU, 0x0b0b0b0bU, 0xdbdbdbdbU,
4d4d77e
-    0xe0e0e0e0U, 0x32323232U, 0x3a3a3a3aU, 0x0a0a0a0aU,
4d4d77e
-    0x49494949U, 0x06060606U, 0x24242424U, 0x5c5c5c5cU,
4d4d77e
-    0xc2c2c2c2U, 0xd3d3d3d3U, 0xacacacacU, 0x62626262U,
4d4d77e
-    0x91919191U, 0x95959595U, 0xe4e4e4e4U, 0x79797979U,
4d4d77e
-    0xe7e7e7e7U, 0xc8c8c8c8U, 0x37373737U, 0x6d6d6d6dU,
4d4d77e
-    0x8d8d8d8dU, 0xd5d5d5d5U, 0x4e4e4e4eU, 0xa9a9a9a9U,
4d4d77e
-    0x6c6c6c6cU, 0x56565656U, 0xf4f4f4f4U, 0xeaeaeaeaU,
4d4d77e
-    0x65656565U, 0x7a7a7a7aU, 0xaeaeaeaeU, 0x08080808U,
4d4d77e
-    0xbabababaU, 0x78787878U, 0x25252525U, 0x2e2e2e2eU,
4d4d77e
-    0x1c1c1c1cU, 0xa6a6a6a6U, 0xb4b4b4b4U, 0xc6c6c6c6U,
4d4d77e
-    0xe8e8e8e8U, 0xddddddddU, 0x74747474U, 0x1f1f1f1fU,
4d4d77e
-    0x4b4b4b4bU, 0xbdbdbdbdU, 0x8b8b8b8bU, 0x8a8a8a8aU,
4d4d77e
-    0x70707070U, 0x3e3e3e3eU, 0xb5b5b5b5U, 0x66666666U,
4d4d77e
-    0x48484848U, 0x03030303U, 0xf6f6f6f6U, 0x0e0e0e0eU,
4d4d77e
-    0x61616161U, 0x35353535U, 0x57575757U, 0xb9b9b9b9U,
4d4d77e
-    0x86868686U, 0xc1c1c1c1U, 0x1d1d1d1dU, 0x9e9e9e9eU,
4d4d77e
-    0xe1e1e1e1U, 0xf8f8f8f8U, 0x98989898U, 0x11111111U,
4d4d77e
-    0x69696969U, 0xd9d9d9d9U, 0x8e8e8e8eU, 0x94949494U,
4d4d77e
-    0x9b9b9b9bU, 0x1e1e1e1eU, 0x87878787U, 0xe9e9e9e9U,
4d4d77e
-    0xcecececeU, 0x55555555U, 0x28282828U, 0xdfdfdfdfU,
4d4d77e
-    0x8c8c8c8cU, 0xa1a1a1a1U, 0x89898989U, 0x0d0d0d0dU,
4d4d77e
-    0xbfbfbfbfU, 0xe6e6e6e6U, 0x42424242U, 0x68686868U,
4d4d77e
-    0x41414141U, 0x99999999U, 0x2d2d2d2dU, 0x0f0f0f0fU,
4d4d77e
-    0xb0b0b0b0U, 0x54545454U, 0xbbbbbbbbU, 0x16161616U,
4d4d77e
-};
4d4d77e
 
4d4d77e
-#ifdef AES_ASM
4d4d77e
-extern const u32 AES_Td[5][256];
4d4d77e
-#define Td0 AES_Td[0]
4d4d77e
-#define Td1 AES_Td[1]
4d4d77e
-#define Td2 AES_Td[2]
4d4d77e
-#define Td3 AES_Td[3]
4d4d77e
-#else
4d4d77e
 static const u32 Td0[256] = {
4d4d77e
     0x51f4a750U, 0x7e416553U, 0x1a17a4c3U, 0x3a275e96U,
4d4d77e
     0x3bab6bcbU, 0x1f9d45f1U, 0xacfa58abU, 0x4be30393U,
4d4d77e
@@ -663,72 +581,39 @@
4d4d77e
     0xa8017139U, 0x0cb3de08U, 0xb4e49cd8U, 0x56c19064U,
4d4d77e
     0xcb84617bU, 0x32b670d5U, 0x6c5c7448U, 0xb85742d0U,
4d4d77e
 };
4d4d77e
-#endif
4d4d77e
-static const u32 Td4[256] = {
4d4d77e
-    0x52525252U, 0x09090909U, 0x6a6a6a6aU, 0xd5d5d5d5U,
4d4d77e
-    0x30303030U, 0x36363636U, 0xa5a5a5a5U, 0x38383838U,
4d4d77e
-    0xbfbfbfbfU, 0x40404040U, 0xa3a3a3a3U, 0x9e9e9e9eU,
4d4d77e
-    0x81818181U, 0xf3f3f3f3U, 0xd7d7d7d7U, 0xfbfbfbfbU,
4d4d77e
-    0x7c7c7c7cU, 0xe3e3e3e3U, 0x39393939U, 0x82828282U,
4d4d77e
-    0x9b9b9b9bU, 0x2f2f2f2fU, 0xffffffffU, 0x87878787U,
4d4d77e
-    0x34343434U, 0x8e8e8e8eU, 0x43434343U, 0x44444444U,
4d4d77e
-    0xc4c4c4c4U, 0xdedededeU, 0xe9e9e9e9U, 0xcbcbcbcbU,
4d4d77e
-    0x54545454U, 0x7b7b7b7bU, 0x94949494U, 0x32323232U,
4d4d77e
-    0xa6a6a6a6U, 0xc2c2c2c2U, 0x23232323U, 0x3d3d3d3dU,
4d4d77e
-    0xeeeeeeeeU, 0x4c4c4c4cU, 0x95959595U, 0x0b0b0b0bU,
4d4d77e
-    0x42424242U, 0xfafafafaU, 0xc3c3c3c3U, 0x4e4e4e4eU,
4d4d77e
-    0x08080808U, 0x2e2e2e2eU, 0xa1a1a1a1U, 0x66666666U,
4d4d77e
-    0x28282828U, 0xd9d9d9d9U, 0x24242424U, 0xb2b2b2b2U,
4d4d77e
-    0x76767676U, 0x5b5b5b5bU, 0xa2a2a2a2U, 0x49494949U,
4d4d77e
-    0x6d6d6d6dU, 0x8b8b8b8bU, 0xd1d1d1d1U, 0x25252525U,
4d4d77e
-    0x72727272U, 0xf8f8f8f8U, 0xf6f6f6f6U, 0x64646464U,
4d4d77e
-    0x86868686U, 0x68686868U, 0x98989898U, 0x16161616U,
4d4d77e
-    0xd4d4d4d4U, 0xa4a4a4a4U, 0x5c5c5c5cU, 0xccccccccU,
4d4d77e
-    0x5d5d5d5dU, 0x65656565U, 0xb6b6b6b6U, 0x92929292U,
4d4d77e
-    0x6c6c6c6cU, 0x70707070U, 0x48484848U, 0x50505050U,
4d4d77e
-    0xfdfdfdfdU, 0xededededU, 0xb9b9b9b9U, 0xdadadadaU,
4d4d77e
-    0x5e5e5e5eU, 0x15151515U, 0x46464646U, 0x57575757U,
4d4d77e
-    0xa7a7a7a7U, 0x8d8d8d8dU, 0x9d9d9d9dU, 0x84848484U,
4d4d77e
-    0x90909090U, 0xd8d8d8d8U, 0xababababU, 0x00000000U,
4d4d77e
-    0x8c8c8c8cU, 0xbcbcbcbcU, 0xd3d3d3d3U, 0x0a0a0a0aU,
4d4d77e
-    0xf7f7f7f7U, 0xe4e4e4e4U, 0x58585858U, 0x05050505U,
4d4d77e
-    0xb8b8b8b8U, 0xb3b3b3b3U, 0x45454545U, 0x06060606U,
4d4d77e
-    0xd0d0d0d0U, 0x2c2c2c2cU, 0x1e1e1e1eU, 0x8f8f8f8fU,
4d4d77e
-    0xcacacacaU, 0x3f3f3f3fU, 0x0f0f0f0fU, 0x02020202U,
4d4d77e
-    0xc1c1c1c1U, 0xafafafafU, 0xbdbdbdbdU, 0x03030303U,
4d4d77e
-    0x01010101U, 0x13131313U, 0x8a8a8a8aU, 0x6b6b6b6bU,
4d4d77e
-    0x3a3a3a3aU, 0x91919191U, 0x11111111U, 0x41414141U,
4d4d77e
-    0x4f4f4f4fU, 0x67676767U, 0xdcdcdcdcU, 0xeaeaeaeaU,
4d4d77e
-    0x97979797U, 0xf2f2f2f2U, 0xcfcfcfcfU, 0xcecececeU,
4d4d77e
-    0xf0f0f0f0U, 0xb4b4b4b4U, 0xe6e6e6e6U, 0x73737373U,
4d4d77e
-    0x96969696U, 0xacacacacU, 0x74747474U, 0x22222222U,
4d4d77e
-    0xe7e7e7e7U, 0xadadadadU, 0x35353535U, 0x85858585U,
4d4d77e
-    0xe2e2e2e2U, 0xf9f9f9f9U, 0x37373737U, 0xe8e8e8e8U,
4d4d77e
-    0x1c1c1c1cU, 0x75757575U, 0xdfdfdfdfU, 0x6e6e6e6eU,
4d4d77e
-    0x47474747U, 0xf1f1f1f1U, 0x1a1a1a1aU, 0x71717171U,
4d4d77e
-    0x1d1d1d1dU, 0x29292929U, 0xc5c5c5c5U, 0x89898989U,
4d4d77e
-    0x6f6f6f6fU, 0xb7b7b7b7U, 0x62626262U, 0x0e0e0e0eU,
4d4d77e
-    0xaaaaaaaaU, 0x18181818U, 0xbebebebeU, 0x1b1b1b1bU,
4d4d77e
-    0xfcfcfcfcU, 0x56565656U, 0x3e3e3e3eU, 0x4b4b4b4bU,
4d4d77e
-    0xc6c6c6c6U, 0xd2d2d2d2U, 0x79797979U, 0x20202020U,
4d4d77e
-    0x9a9a9a9aU, 0xdbdbdbdbU, 0xc0c0c0c0U, 0xfefefefeU,
4d4d77e
-    0x78787878U, 0xcdcdcdcdU, 0x5a5a5a5aU, 0xf4f4f4f4U,
4d4d77e
-    0x1f1f1f1fU, 0xddddddddU, 0xa8a8a8a8U, 0x33333333U,
4d4d77e
-    0x88888888U, 0x07070707U, 0xc7c7c7c7U, 0x31313131U,
4d4d77e
-    0xb1b1b1b1U, 0x12121212U, 0x10101010U, 0x59595959U,
4d4d77e
-    0x27272727U, 0x80808080U, 0xececececU, 0x5f5f5f5fU,
4d4d77e
-    0x60606060U, 0x51515151U, 0x7f7f7f7fU, 0xa9a9a9a9U,
4d4d77e
-    0x19191919U, 0xb5b5b5b5U, 0x4a4a4a4aU, 0x0d0d0d0dU,
4d4d77e
-    0x2d2d2d2dU, 0xe5e5e5e5U, 0x7a7a7a7aU, 0x9f9f9f9fU,
4d4d77e
-    0x93939393U, 0xc9c9c9c9U, 0x9c9c9c9cU, 0xefefefefU,
4d4d77e
-    0xa0a0a0a0U, 0xe0e0e0e0U, 0x3b3b3b3bU, 0x4d4d4d4dU,
4d4d77e
-    0xaeaeaeaeU, 0x2a2a2a2aU, 0xf5f5f5f5U, 0xb0b0b0b0U,
4d4d77e
-    0xc8c8c8c8U, 0xebebebebU, 0xbbbbbbbbU, 0x3c3c3c3cU,
4d4d77e
-    0x83838383U, 0x53535353U, 0x99999999U, 0x61616161U,
4d4d77e
-    0x17171717U, 0x2b2b2b2bU, 0x04040404U, 0x7e7e7e7eU,
4d4d77e
-    0xbabababaU, 0x77777777U, 0xd6d6d6d6U, 0x26262626U,
4d4d77e
-    0xe1e1e1e1U, 0x69696969U, 0x14141414U, 0x63636363U,
4d4d77e
-    0x55555555U, 0x21212121U, 0x0c0c0c0cU, 0x7d7d7d7dU,
4d4d77e
+static const u8 Td4[256] = {
4d4d77e
+    0x52U, 0x09U, 0x6aU, 0xd5U, 0x30U, 0x36U, 0xa5U, 0x38U,
4d4d77e
+    0xbfU, 0x40U, 0xa3U, 0x9eU, 0x81U, 0xf3U, 0xd7U, 0xfbU,
4d4d77e
+    0x7cU, 0xe3U, 0x39U, 0x82U, 0x9bU, 0x2fU, 0xffU, 0x87U,
4d4d77e
+    0x34U, 0x8eU, 0x43U, 0x44U, 0xc4U, 0xdeU, 0xe9U, 0xcbU,
4d4d77e
+    0x54U, 0x7bU, 0x94U, 0x32U, 0xa6U, 0xc2U, 0x23U, 0x3dU,
4d4d77e
+    0xeeU, 0x4cU, 0x95U, 0x0bU, 0x42U, 0xfaU, 0xc3U, 0x4eU,
4d4d77e
+    0x08U, 0x2eU, 0xa1U, 0x66U, 0x28U, 0xd9U, 0x24U, 0xb2U,
4d4d77e
+    0x76U, 0x5bU, 0xa2U, 0x49U, 0x6dU, 0x8bU, 0xd1U, 0x25U,
4d4d77e
+    0x72U, 0xf8U, 0xf6U, 0x64U, 0x86U, 0x68U, 0x98U, 0x16U,
4d4d77e
+    0xd4U, 0xa4U, 0x5cU, 0xccU, 0x5dU, 0x65U, 0xb6U, 0x92U,
4d4d77e
+    0x6cU, 0x70U, 0x48U, 0x50U, 0xfdU, 0xedU, 0xb9U, 0xdaU,
4d4d77e
+    0x5eU, 0x15U, 0x46U, 0x57U, 0xa7U, 0x8dU, 0x9dU, 0x84U,
4d4d77e
+    0x90U, 0xd8U, 0xabU, 0x00U, 0x8cU, 0xbcU, 0xd3U, 0x0aU,
4d4d77e
+    0xf7U, 0xe4U, 0x58U, 0x05U, 0xb8U, 0xb3U, 0x45U, 0x06U,
4d4d77e
+    0xd0U, 0x2cU, 0x1eU, 0x8fU, 0xcaU, 0x3fU, 0x0fU, 0x02U,
4d4d77e
+    0xc1U, 0xafU, 0xbdU, 0x03U, 0x01U, 0x13U, 0x8aU, 0x6bU,
4d4d77e
+    0x3aU, 0x91U, 0x11U, 0x41U, 0x4fU, 0x67U, 0xdcU, 0xeaU,
4d4d77e
+    0x97U, 0xf2U, 0xcfU, 0xceU, 0xf0U, 0xb4U, 0xe6U, 0x73U,
4d4d77e
+    0x96U, 0xacU, 0x74U, 0x22U, 0xe7U, 0xadU, 0x35U, 0x85U,
4d4d77e
+    0xe2U, 0xf9U, 0x37U, 0xe8U, 0x1cU, 0x75U, 0xdfU, 0x6eU,
4d4d77e
+    0x47U, 0xf1U, 0x1aU, 0x71U, 0x1dU, 0x29U, 0xc5U, 0x89U,
4d4d77e
+    0x6fU, 0xb7U, 0x62U, 0x0eU, 0xaaU, 0x18U, 0xbeU, 0x1bU,
4d4d77e
+    0xfcU, 0x56U, 0x3eU, 0x4bU, 0xc6U, 0xd2U, 0x79U, 0x20U,
4d4d77e
+    0x9aU, 0xdbU, 0xc0U, 0xfeU, 0x78U, 0xcdU, 0x5aU, 0xf4U,
4d4d77e
+    0x1fU, 0xddU, 0xa8U, 0x33U, 0x88U, 0x07U, 0xc7U, 0x31U,
4d4d77e
+    0xb1U, 0x12U, 0x10U, 0x59U, 0x27U, 0x80U, 0xecU, 0x5fU,
4d4d77e
+    0x60U, 0x51U, 0x7fU, 0xa9U, 0x19U, 0xb5U, 0x4aU, 0x0dU,
4d4d77e
+    0x2dU, 0xe5U, 0x7aU, 0x9fU, 0x93U, 0xc9U, 0x9cU, 0xefU,
4d4d77e
+    0xa0U, 0xe0U, 0x3bU, 0x4dU, 0xaeU, 0x2aU, 0xf5U, 0xb0U,
4d4d77e
+    0xc8U, 0xebU, 0xbbU, 0x3cU, 0x83U, 0x53U, 0x99U, 0x61U,
4d4d77e
+    0x17U, 0x2bU, 0x04U, 0x7eU, 0xbaU, 0x77U, 0xd6U, 0x26U,
4d4d77e
+    0xe1U, 0x69U, 0x14U, 0x63U, 0x55U, 0x21U, 0x0cU, 0x7dU,
4d4d77e
 };
4d4d77e
 static const u32 rcon[] = {
4d4d77e
 	0x01000000, 0x02000000, 0x04000000, 0x08000000,
4d4d77e
@@ -768,10 +653,10 @@
4d4d77e
 		while (1) {
4d4d77e
 			temp  = rk[3];
4d4d77e
 			rk[4] = rk[0] ^
4d4d77e
-				(Te4[(temp >> 16) & 0xff] & 0xff000000) ^
4d4d77e
-				(Te4[(temp >>  8) & 0xff] & 0x00ff0000) ^
4d4d77e
-				(Te4[(temp      ) & 0xff] & 0x0000ff00) ^
4d4d77e
-				(Te4[(temp >> 24)       ] & 0x000000ff) ^
4d4d77e
+				(Te2[(temp >> 16) & 0xff] & 0xff000000) ^
4d4d77e
+				(Te3[(temp >>  8) & 0xff] & 0x00ff0000) ^
4d4d77e
+				(Te0[(temp      ) & 0xff] & 0x0000ff00) ^
4d4d77e
+				(Te1[(temp >> 24)       ] & 0x000000ff) ^
4d4d77e
 				rcon[i];
4d4d77e
 			rk[5] = rk[1] ^ rk[4];
4d4d77e
 			rk[6] = rk[2] ^ rk[5];
4d4d77e
@@ -788,10 +673,10 @@
4d4d77e
 		while (1) {
4d4d77e
 			temp = rk[ 5];
4d4d77e
 			rk[ 6] = rk[ 0] ^
4d4d77e
-				(Te4[(temp >> 16) & 0xff] & 0xff000000) ^
4d4d77e
-				(Te4[(temp >>  8) & 0xff] & 0x00ff0000) ^
4d4d77e
-				(Te4[(temp      ) & 0xff] & 0x0000ff00) ^
4d4d77e
-				(Te4[(temp >> 24)       ] & 0x000000ff) ^
4d4d77e
+				(Te2[(temp >> 16) & 0xff] & 0xff000000) ^
4d4d77e
+				(Te3[(temp >>  8) & 0xff] & 0x00ff0000) ^
4d4d77e
+				(Te0[(temp      ) & 0xff] & 0x0000ff00) ^
4d4d77e
+				(Te1[(temp >> 24)       ] & 0x000000ff) ^
4d4d77e
 				rcon[i];
4d4d77e
 			rk[ 7] = rk[ 1] ^ rk[ 6];
4d4d77e
 			rk[ 8] = rk[ 2] ^ rk[ 7];
4d4d77e
@@ -810,10 +695,10 @@
4d4d77e
 		while (1) {
4d4d77e
 			temp = rk[ 7];
4d4d77e
 			rk[ 8] = rk[ 0] ^
4d4d77e
-				(Te4[(temp >> 16) & 0xff] & 0xff000000) ^
4d4d77e
-				(Te4[(temp >>  8) & 0xff] & 0x00ff0000) ^
4d4d77e
-				(Te4[(temp      ) & 0xff] & 0x0000ff00) ^
4d4d77e
-				(Te4[(temp >> 24)       ] & 0x000000ff) ^
4d4d77e
+				(Te2[(temp >> 16) & 0xff] & 0xff000000) ^
4d4d77e
+				(Te3[(temp >>  8) & 0xff] & 0x00ff0000) ^
4d4d77e
+				(Te0[(temp      ) & 0xff] & 0x0000ff00) ^
4d4d77e
+				(Te1[(temp >> 24)       ] & 0x000000ff) ^
4d4d77e
 				rcon[i];
4d4d77e
 			rk[ 9] = rk[ 1] ^ rk[ 8];
4d4d77e
 			rk[10] = rk[ 2] ^ rk[ 9];
4d4d77e
@@ -823,10 +708,10 @@
4d4d77e
 			}
4d4d77e
 			temp = rk[11];
4d4d77e
 			rk[12] = rk[ 4] ^
4d4d77e
-				(Te4[(temp >> 24)       ] & 0xff000000) ^
4d4d77e
-				(Te4[(temp >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
-				(Te4[(temp >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
-				(Te4[(temp      ) & 0xff] & 0x000000ff);
4d4d77e
+				(Te2[(temp >> 24)       ] & 0xff000000) ^
4d4d77e
+				(Te3[(temp >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
+				(Te0[(temp >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
+				(Te1[(temp      ) & 0xff] & 0x000000ff);
4d4d77e
 			rk[13] = rk[ 5] ^ rk[12];
4d4d77e
 			rk[14] = rk[ 6] ^ rk[13];
4d4d77e
 			rk[15] = rk[ 7] ^ rk[14];
4d4d77e
@@ -865,25 +750,25 @@
4d4d77e
 	for (i = 1; i < (key->rounds); i++) {
4d4d77e
 		rk += 4;
4d4d77e
 		rk[0] =
4d4d77e
-			Td0[Te4[(rk[0] >> 24)       ] & 0xff] ^
4d4d77e
-			Td1[Te4[(rk[0] >> 16) & 0xff] & 0xff] ^
4d4d77e
-			Td2[Te4[(rk[0] >>  8) & 0xff] & 0xff] ^
4d4d77e
-			Td3[Te4[(rk[0]      ) & 0xff] & 0xff];
4d4d77e
+			Td0[Te1[(rk[0] >> 24)       ] & 0xff] ^
4d4d77e
+			Td1[Te1[(rk[0] >> 16) & 0xff] & 0xff] ^
4d4d77e
+			Td2[Te1[(rk[0] >>  8) & 0xff] & 0xff] ^
4d4d77e
+			Td3[Te1[(rk[0]      ) & 0xff] & 0xff];
4d4d77e
 		rk[1] =
4d4d77e
-			Td0[Te4[(rk[1] >> 24)       ] & 0xff] ^
4d4d77e
-			Td1[Te4[(rk[1] >> 16) & 0xff] & 0xff] ^
4d4d77e
-			Td2[Te4[(rk[1] >>  8) & 0xff] & 0xff] ^
4d4d77e
-			Td3[Te4[(rk[1]      ) & 0xff] & 0xff];
4d4d77e
+			Td0[Te1[(rk[1] >> 24)       ] & 0xff] ^
4d4d77e
+			Td1[Te1[(rk[1] >> 16) & 0xff] & 0xff] ^
4d4d77e
+			Td2[Te1[(rk[1] >>  8) & 0xff] & 0xff] ^
4d4d77e
+			Td3[Te1[(rk[1]      ) & 0xff] & 0xff];
4d4d77e
 		rk[2] =
4d4d77e
-			Td0[Te4[(rk[2] >> 24)       ] & 0xff] ^
4d4d77e
-			Td1[Te4[(rk[2] >> 16) & 0xff] & 0xff] ^
4d4d77e
-			Td2[Te4[(rk[2] >>  8) & 0xff] & 0xff] ^
4d4d77e
-			Td3[Te4[(rk[2]      ) & 0xff] & 0xff];
4d4d77e
+			Td0[Te1[(rk[2] >> 24)       ] & 0xff] ^
4d4d77e
+			Td1[Te1[(rk[2] >> 16) & 0xff] & 0xff] ^
4d4d77e
+			Td2[Te1[(rk[2] >>  8) & 0xff] & 0xff] ^
4d4d77e
+			Td3[Te1[(rk[2]      ) & 0xff] & 0xff];
4d4d77e
 		rk[3] =
4d4d77e
-			Td0[Te4[(rk[3] >> 24)       ] & 0xff] ^
4d4d77e
-			Td1[Te4[(rk[3] >> 16) & 0xff] & 0xff] ^
4d4d77e
-			Td2[Te4[(rk[3] >>  8) & 0xff] & 0xff] ^
4d4d77e
-			Td3[Te4[(rk[3]      ) & 0xff] & 0xff];
4d4d77e
+			Td0[Te1[(rk[3] >> 24)       ] & 0xff] ^
4d4d77e
+			Td1[Te1[(rk[3] >> 16) & 0xff] & 0xff] ^
4d4d77e
+			Td2[Te1[(rk[3] >>  8) & 0xff] & 0xff] ^
4d4d77e
+			Td3[Te1[(rk[3]      ) & 0xff] & 0xff];
4d4d77e
 	}
4d4d77e
 	return 0;
4d4d77e
 }
4d4d77e
@@ -1051,31 +936,31 @@
4d4d77e
 	 * map cipher state to byte array block:
4d4d77e
 	 */
4d4d77e
 	s0 =
4d4d77e
-		(Te4[(t0 >> 24)       ] & 0xff000000) ^
4d4d77e
-		(Te4[(t1 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
-		(Te4[(t2 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
-		(Te4[(t3      ) & 0xff] & 0x000000ff) ^
4d4d77e
+		(Te2[(t0 >> 24)       ] & 0xff000000) ^
4d4d77e
+		(Te3[(t1 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
+		(Te0[(t2 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
+		(Te1[(t3      ) & 0xff] & 0x000000ff) ^
4d4d77e
 		rk[0];
4d4d77e
 	PUTU32(out     , s0);
4d4d77e
 	s1 =
4d4d77e
-		(Te4[(t1 >> 24)       ] & 0xff000000) ^
4d4d77e
-		(Te4[(t2 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
-		(Te4[(t3 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
-		(Te4[(t0      ) & 0xff] & 0x000000ff) ^
4d4d77e
+		(Te2[(t1 >> 24)       ] & 0xff000000) ^
4d4d77e
+		(Te3[(t2 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
+		(Te0[(t3 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
+		(Te1[(t0      ) & 0xff] & 0x000000ff) ^
4d4d77e
 		rk[1];
4d4d77e
 	PUTU32(out +  4, s1);
4d4d77e
 	s2 =
4d4d77e
-		(Te4[(t2 >> 24)       ] & 0xff000000) ^
4d4d77e
-		(Te4[(t3 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
-		(Te4[(t0 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
-		(Te4[(t1      ) & 0xff] & 0x000000ff) ^
4d4d77e
+		(Te2[(t2 >> 24)       ] & 0xff000000) ^
4d4d77e
+		(Te3[(t3 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
+		(Te0[(t0 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
+		(Te1[(t1      ) & 0xff] & 0x000000ff) ^
4d4d77e
 		rk[2];
4d4d77e
 	PUTU32(out +  8, s2);
4d4d77e
 	s3 =
4d4d77e
-		(Te4[(t3 >> 24)       ] & 0xff000000) ^
4d4d77e
-		(Te4[(t0 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
-		(Te4[(t1 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
-		(Te4[(t2      ) & 0xff] & 0x000000ff) ^
4d4d77e
+		(Te2[(t3 >> 24)       ] & 0xff000000) ^
4d4d77e
+		(Te3[(t0 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
+		(Te0[(t1 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
+		(Te1[(t2      ) & 0xff] & 0x000000ff) ^
4d4d77e
 		rk[3];
4d4d77e
 	PUTU32(out + 12, s3);
4d4d77e
 }
4d4d77e
@@ -1242,31 +1127,31 @@
4d4d77e
 	 * map cipher state to byte array block:
4d4d77e
 	 */
4d4d77e
    	s0 =
4d4d77e
-   		(Td4[(t0 >> 24)       ] & 0xff000000) ^
4d4d77e
-   		(Td4[(t3 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
-   		(Td4[(t2 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
-   		(Td4[(t1      ) & 0xff] & 0x000000ff) ^
4d4d77e
+   		(Td4[(t0 >> 24)       ] << 24) ^
4d4d77e
+   		(Td4[(t3 >> 16) & 0xff] << 16) ^
4d4d77e
+   		(Td4[(t2 >>  8) & 0xff] <<  8) ^
4d4d77e
+   		(Td4[(t1      ) & 0xff])       ^
4d4d77e
    		rk[0];
4d4d77e
 	PUTU32(out     , s0);
4d4d77e
    	s1 =
4d4d77e
-   		(Td4[(t1 >> 24)       ] & 0xff000000) ^
4d4d77e
-   		(Td4[(t0 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
-   		(Td4[(t3 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
-   		(Td4[(t2      ) & 0xff] & 0x000000ff) ^
4d4d77e
+   		(Td4[(t1 >> 24)       ] << 24) ^
4d4d77e
+   		(Td4[(t0 >> 16) & 0xff] << 16) ^
4d4d77e
+   		(Td4[(t3 >>  8) & 0xff] <<  8) ^
4d4d77e
+   		(Td4[(t2      ) & 0xff])       ^
4d4d77e
    		rk[1];
4d4d77e
 	PUTU32(out +  4, s1);
4d4d77e
    	s2 =
4d4d77e
-   		(Td4[(t2 >> 24)       ] & 0xff000000) ^
4d4d77e
-   		(Td4[(t1 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
-   		(Td4[(t0 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
-   		(Td4[(t3      ) & 0xff] & 0x000000ff) ^
4d4d77e
+   		(Td4[(t2 >> 24)       ] << 24) ^
4d4d77e
+   		(Td4[(t1 >> 16) & 0xff] << 16) ^
4d4d77e
+   		(Td4[(t0 >>  8) & 0xff] <<  8) ^
4d4d77e
+   		(Td4[(t3      ) & 0xff])       ^
4d4d77e
    		rk[2];
4d4d77e
 	PUTU32(out +  8, s2);
4d4d77e
    	s3 =
4d4d77e
-   		(Td4[(t3 >> 24)       ] & 0xff000000) ^
4d4d77e
-   		(Td4[(t2 >> 16) & 0xff] & 0x00ff0000) ^
4d4d77e
-   		(Td4[(t1 >>  8) & 0xff] & 0x0000ff00) ^
4d4d77e
-   		(Td4[(t0      ) & 0xff] & 0x000000ff) ^
4d4d77e
+   		(Td4[(t3 >> 24)       ] << 24) ^
4d4d77e
+   		(Td4[(t2 >> 16) & 0xff] << 16) ^
4d4d77e
+   		(Td4[(t1 >>  8) & 0xff] <<  8) ^
4d4d77e
+   		(Td4[(t0      ) & 0xff])       ^
4d4d77e
    		rk[3];
4d4d77e
 	PUTU32(out + 12, s3);
4d4d77e
 }
4d4d77e
openssl/crypto/aes/asm/aes-586.pl     1.11 -> 1.11.2.1
4d4d77e
4d4d77e
--- openssl/crypto/aes/asm/aes-586.pl 2005/05/09 21:48:00 1.11
4d4d77e
+++ openssl/crypto/aes/asm/aes-586.pl 2006/06/28 09:01:40 1.11.2.1
4d4d77e
@@ -6,7 +6,7 @@
4d4d77e
 # forms are granted according to the OpenSSL license.
4d4d77e
 # ====================================================================
4d4d77e
 #
4d4d77e
-# Version 3.4.
4d4d77e
+# Version 3.6.
4d4d77e
 #
4d4d77e
 # You might fail to appreciate this module performance from the first
4d4d77e
 # try. If compared to "vanilla" linux-ia32-icc target, i.e. considered
4d4d77e
@@ -66,6 +66,13 @@
4d4d77e
 # stack. This unfortunately has rather strong impact on small block CBC
4d4d77e
 # performance, ~2x deterioration on 16-byte block if compared to 3.3.
4d4d77e
 #
4d4d77e
+# Version 3.5 checks if there is L1 cache aliasing between user-supplied
4d4d77e
+# key schedule and S-boxes and abstains from copying the former if
4d4d77e
+# there is no. This allows end-user to consciously retain small block
4d4d77e
+# performance by aligning key schedule in specific manner.
4d4d77e
+#
4d4d77e
+# Version 3.6 compresses Td4 to 256 bytes and prefetches it in ECB.
4d4d77e
+#
4d4d77e
 # Current ECB performance numbers for 128-bit key in CPU cycles per
4d4d77e
 # processed byte [measure commonly used by AES benchmarkers] are:
4d4d77e
 #
4d4d77e
@@ -505,28 +512,27 @@
4d4d77e
 	if($i==3)   {	&mov	($key,&DWP(12,"esp"));		}
4d4d77e
 	else        {	&mov	($out,$s[0]);			}
4d4d77e
 			&and	($out,0xFF);
4d4d77e
-			&mov	($out,&DWP(2048,$td,$out,4));
4d4d77e
-			&and	($out,0x000000ff);
4d4d77e
+			&movz	($out,&DWP(2048,$td,$out,1));
4d4d77e
 
4d4d77e
 	if ($i==3)  {	$tmp=$s[1];				}
4d4d77e
 			&movz	($tmp,&HB($s[1]));
4d4d77e
-			&mov	($tmp,&DWP(2048,$td,$tmp,4));
4d4d77e
-			&and	($tmp,0x0000ff00);
4d4d77e
+			&movz	($tmp,&DWP(2048,$td,$tmp,1));
4d4d77e
+			&shl	($tmp,8);
4d4d77e
 			&xor	($out,$tmp);
4d4d77e
 
4d4d77e
 	if ($i==3)  {	$tmp=$s[2]; &mov ($s[1],$acc);		}
4d4d77e
 	else        {	mov	($tmp,$s[2]);			}
4d4d77e
 			&shr	($tmp,16);
4d4d77e
 			&and	($tmp,0xFF);
4d4d77e
-			&mov	($tmp,&DWP(2048,$td,$tmp,4));
4d4d77e
-			&and	($tmp,0x00ff0000);
4d4d77e
+			&movz	($tmp,&DWP(2048,$td,$tmp,1));
4d4d77e
+			&shl	($tmp,16);
4d4d77e
 			&xor	($out,$tmp);
4d4d77e
 
4d4d77e
 	if ($i==3)  {	$tmp=$s[3]; &mov ($s[2],&DWP(8,"esp"));	}
4d4d77e
 	else        {	&mov	($tmp,$s[3]);			}
4d4d77e
 			&shr	($tmp,24);
4d4d77e
-			&mov	($tmp,&DWP(2048,$td,$tmp,4));
4d4d77e
-			&and	($tmp,0xff000000);
4d4d77e
+			&movz	($tmp,&DWP(2048,$td,$tmp,1));
4d4d77e
+			&shl	($tmp,24);
4d4d77e
 			&xor	($out,$tmp);
4d4d77e
 	if ($i<2)   {	&mov	(&DWP(4+4*$i,"esp"),$out);	}
4d4d77e
 	if ($i==3)  {	&mov	($s[3],&DWP(4,"esp"));		}
4d4d77e
@@ -687,70 +693,38 @@
4d4d77e
 	&_data_word(0x7101a839, 0xdeb30c08, 0x9ce4b4d8, 0x90c15664);
4d4d77e
 	&_data_word(0x6184cb7b, 0x70b632d5, 0x745c6c48, 0x4257b8d0);
4d4d77e
 #Td4:
4d4d77e
-	&data_word(0x52525252, 0x09090909, 0x6a6a6a6a, 0xd5d5d5d5);
4d4d77e
-	&data_word(0x30303030, 0x36363636, 0xa5a5a5a5, 0x38383838);
4d4d77e
-	&data_word(0xbfbfbfbf, 0x40404040, 0xa3a3a3a3, 0x9e9e9e9e);
4d4d77e
-	&data_word(0x81818181, 0xf3f3f3f3, 0xd7d7d7d7, 0xfbfbfbfb);
4d4d77e
-	&data_word(0x7c7c7c7c, 0xe3e3e3e3, 0x39393939, 0x82828282);
4d4d77e
-	&data_word(0x9b9b9b9b, 0x2f2f2f2f, 0xffffffff, 0x87878787);
4d4d77e
-	&data_word(0x34343434, 0x8e8e8e8e, 0x43434343, 0x44444444);
4d4d77e
-	&data_word(0xc4c4c4c4, 0xdededede, 0xe9e9e9e9, 0xcbcbcbcb);
4d4d77e
-	&data_word(0x54545454, 0x7b7b7b7b, 0x94949494, 0x32323232);
4d4d77e
-	&data_word(0xa6a6a6a6, 0xc2c2c2c2, 0x23232323, 0x3d3d3d3d);
4d4d77e
-	&data_word(0xeeeeeeee, 0x4c4c4c4c, 0x95959595, 0x0b0b0b0b);
4d4d77e
-	&data_word(0x42424242, 0xfafafafa, 0xc3c3c3c3, 0x4e4e4e4e);
4d4d77e
-	&data_word(0x08080808, 0x2e2e2e2e, 0xa1a1a1a1, 0x66666666);
4d4d77e
-	&data_word(0x28282828, 0xd9d9d9d9, 0x24242424, 0xb2b2b2b2);
4d4d77e
-	&data_word(0x76767676, 0x5b5b5b5b, 0xa2a2a2a2, 0x49494949);
4d4d77e
-	&data_word(0x6d6d6d6d, 0x8b8b8b8b, 0xd1d1d1d1, 0x25252525);
4d4d77e
-	&data_word(0x72727272, 0xf8f8f8f8, 0xf6f6f6f6, 0x64646464);
4d4d77e
-	&data_word(0x86868686, 0x68686868, 0x98989898, 0x16161616);
4d4d77e
-	&data_word(0xd4d4d4d4, 0xa4a4a4a4, 0x5c5c5c5c, 0xcccccccc);
4d4d77e
-	&data_word(0x5d5d5d5d, 0x65656565, 0xb6b6b6b6, 0x92929292);
4d4d77e
-	&data_word(0x6c6c6c6c, 0x70707070, 0x48484848, 0x50505050);
4d4d77e
-	&data_word(0xfdfdfdfd, 0xedededed, 0xb9b9b9b9, 0xdadadada);
4d4d77e
-	&data_word(0x5e5e5e5e, 0x15151515, 0x46464646, 0x57575757);
4d4d77e
-	&data_word(0xa7a7a7a7, 0x8d8d8d8d, 0x9d9d9d9d, 0x84848484);
4d4d77e
-	&data_word(0x90909090, 0xd8d8d8d8, 0xabababab, 0x00000000);
4d4d77e
-	&data_word(0x8c8c8c8c, 0xbcbcbcbc, 0xd3d3d3d3, 0x0a0a0a0a);
4d4d77e
-	&data_word(0xf7f7f7f7, 0xe4e4e4e4, 0x58585858, 0x05050505);
4d4d77e
-	&data_word(0xb8b8b8b8, 0xb3b3b3b3, 0x45454545, 0x06060606);
4d4d77e
-	&data_word(0xd0d0d0d0, 0x2c2c2c2c, 0x1e1e1e1e, 0x8f8f8f8f);
4d4d77e
-	&data_word(0xcacacaca, 0x3f3f3f3f, 0x0f0f0f0f, 0x02020202);
4d4d77e
-	&data_word(0xc1c1c1c1, 0xafafafaf, 0xbdbdbdbd, 0x03030303);
4d4d77e
-	&data_word(0x01010101, 0x13131313, 0x8a8a8a8a, 0x6b6b6b6b);
4d4d77e
-	&data_word(0x3a3a3a3a, 0x91919191, 0x11111111, 0x41414141);
4d4d77e
-	&data_word(0x4f4f4f4f, 0x67676767, 0xdcdcdcdc, 0xeaeaeaea);
4d4d77e
-	&data_word(0x97979797, 0xf2f2f2f2, 0xcfcfcfcf, 0xcececece);
4d4d77e
-	&data_word(0xf0f0f0f0, 0xb4b4b4b4, 0xe6e6e6e6, 0x73737373);
4d4d77e
-	&data_word(0x96969696, 0xacacacac, 0x74747474, 0x22222222);
4d4d77e
-	&data_word(0xe7e7e7e7, 0xadadadad, 0x35353535, 0x85858585);
4d4d77e
-	&data_word(0xe2e2e2e2, 0xf9f9f9f9, 0x37373737, 0xe8e8e8e8);
4d4d77e
-	&data_word(0x1c1c1c1c, 0x75757575, 0xdfdfdfdf, 0x6e6e6e6e);
4d4d77e
-	&data_word(0x47474747, 0xf1f1f1f1, 0x1a1a1a1a, 0x71717171);
4d4d77e
-	&data_word(0x1d1d1d1d, 0x29292929, 0xc5c5c5c5, 0x89898989);
4d4d77e
-	&data_word(0x6f6f6f6f, 0xb7b7b7b7, 0x62626262, 0x0e0e0e0e);
4d4d77e
-	&data_word(0xaaaaaaaa, 0x18181818, 0xbebebebe, 0x1b1b1b1b);
4d4d77e
-	&data_word(0xfcfcfcfc, 0x56565656, 0x3e3e3e3e, 0x4b4b4b4b);
4d4d77e
-	&data_word(0xc6c6c6c6, 0xd2d2d2d2, 0x79797979, 0x20202020);
4d4d77e
-	&data_word(0x9a9a9a9a, 0xdbdbdbdb, 0xc0c0c0c0, 0xfefefefe);
4d4d77e
-	&data_word(0x78787878, 0xcdcdcdcd, 0x5a5a5a5a, 0xf4f4f4f4);
4d4d77e
-	&data_word(0x1f1f1f1f, 0xdddddddd, 0xa8a8a8a8, 0x33333333);
4d4d77e
-	&data_word(0x88888888, 0x07070707, 0xc7c7c7c7, 0x31313131);
4d4d77e
-	&data_word(0xb1b1b1b1, 0x12121212, 0x10101010, 0x59595959);
4d4d77e
-	&data_word(0x27272727, 0x80808080, 0xecececec, 0x5f5f5f5f);
4d4d77e
-	&data_word(0x60606060, 0x51515151, 0x7f7f7f7f, 0xa9a9a9a9);
4d4d77e
-	&data_word(0x19191919, 0xb5b5b5b5, 0x4a4a4a4a, 0x0d0d0d0d);
4d4d77e
-	&data_word(0x2d2d2d2d, 0xe5e5e5e5, 0x7a7a7a7a, 0x9f9f9f9f);
4d4d77e
-	&data_word(0x93939393, 0xc9c9c9c9, 0x9c9c9c9c, 0xefefefef);
4d4d77e
-	&data_word(0xa0a0a0a0, 0xe0e0e0e0, 0x3b3b3b3b, 0x4d4d4d4d);
4d4d77e
-	&data_word(0xaeaeaeae, 0x2a2a2a2a, 0xf5f5f5f5, 0xb0b0b0b0);
4d4d77e
-	&data_word(0xc8c8c8c8, 0xebebebeb, 0xbbbbbbbb, 0x3c3c3c3c);
4d4d77e
-	&data_word(0x83838383, 0x53535353, 0x99999999, 0x61616161);
4d4d77e
-	&data_word(0x17171717, 0x2b2b2b2b, 0x04040404, 0x7e7e7e7e);
4d4d77e
-	&data_word(0xbabababa, 0x77777777, 0xd6d6d6d6, 0x26262626);
4d4d77e
-	&data_word(0xe1e1e1e1, 0x69696969, 0x14141414, 0x63636363);
4d4d77e
-	&data_word(0x55555555, 0x21212121, 0x0c0c0c0c, 0x7d7d7d7d);
4d4d77e
+	&data_byte(0x52, 0x09, 0x6a, 0xd5, 0x30, 0x36, 0xa5, 0x38);
4d4d77e
+	&data_byte(0xbf, 0x40, 0xa3, 0x9e, 0x81, 0xf3, 0xd7, 0xfb);
4d4d77e
+	&data_byte(0x7c, 0xe3, 0x39, 0x82, 0x9b, 0x2f, 0xff, 0x87);
4d4d77e
+	&data_byte(0x34, 0x8e, 0x43, 0x44, 0xc4, 0xde, 0xe9, 0xcb);
4d4d77e
+	&data_byte(0x54, 0x7b, 0x94, 0x32, 0xa6, 0xc2, 0x23, 0x3d);
4d4d77e
+	&data_byte(0xee, 0x4c, 0x95, 0x0b, 0x42, 0xfa, 0xc3, 0x4e);
4d4d77e
+	&data_byte(0x08, 0x2e, 0xa1, 0x66, 0x28, 0xd9, 0x24, 0xb2);
4d4d77e
+	&data_byte(0x76, 0x5b, 0xa2, 0x49, 0x6d, 0x8b, 0xd1, 0x25);
4d4d77e
+	&data_byte(0x72, 0xf8, 0xf6, 0x64, 0x86, 0x68, 0x98, 0x16);
4d4d77e
+	&data_byte(0xd4, 0xa4, 0x5c, 0xcc, 0x5d, 0x65, 0xb6, 0x92);
4d4d77e
+	&data_byte(0x6c, 0x70, 0x48, 0x50, 0xfd, 0xed, 0xb9, 0xda);
4d4d77e
+	&data_byte(0x5e, 0x15, 0x46, 0x57, 0xa7, 0x8d, 0x9d, 0x84);
4d4d77e
+	&data_byte(0x90, 0xd8, 0xab, 0x00, 0x8c, 0xbc, 0xd3, 0x0a);
4d4d77e
+	&data_byte(0xf7, 0xe4, 0x58, 0x05, 0xb8, 0xb3, 0x45, 0x06);
4d4d77e
+	&data_byte(0xd0, 0x2c, 0x1e, 0x8f, 0xca, 0x3f, 0x0f, 0x02);
4d4d77e
+	&data_byte(0xc1, 0xaf, 0xbd, 0x03, 0x01, 0x13, 0x8a, 0x6b);
4d4d77e
+	&data_byte(0x3a, 0x91, 0x11, 0x41, 0x4f, 0x67, 0xdc, 0xea);
4d4d77e
+	&data_byte(0x97, 0xf2, 0xcf, 0xce, 0xf0, 0xb4, 0xe6, 0x73);
4d4d77e
+	&data_byte(0x96, 0xac, 0x74, 0x22, 0xe7, 0xad, 0x35, 0x85);
4d4d77e
+	&data_byte(0xe2, 0xf9, 0x37, 0xe8, 0x1c, 0x75, 0xdf, 0x6e);
4d4d77e
+	&data_byte(0x47, 0xf1, 0x1a, 0x71, 0x1d, 0x29, 0xc5, 0x89);
4d4d77e
+	&data_byte(0x6f, 0xb7, 0x62, 0x0e, 0xaa, 0x18, 0xbe, 0x1b);
4d4d77e
+	&data_byte(0xfc, 0x56, 0x3e, 0x4b, 0xc6, 0xd2, 0x79, 0x20);
4d4d77e
+	&data_byte(0x9a, 0xdb, 0xc0, 0xfe, 0x78, 0xcd, 0x5a, 0xf4);
4d4d77e
+	&data_byte(0x1f, 0xdd, 0xa8, 0x33, 0x88, 0x07, 0xc7, 0x31);
4d4d77e
+	&data_byte(0xb1, 0x12, 0x10, 0x59, 0x27, 0x80, 0xec, 0x5f);
4d4d77e
+	&data_byte(0x60, 0x51, 0x7f, 0xa9, 0x19, 0xb5, 0x4a, 0x0d);
4d4d77e
+	&data_byte(0x2d, 0xe5, 0x7a, 0x9f, 0x93, 0xc9, 0x9c, 0xef);
4d4d77e
+	&data_byte(0xa0, 0xe0, 0x3b, 0x4d, 0xae, 0x2a, 0xf5, 0xb0);
4d4d77e
+	&data_byte(0xc8, 0xeb, 0xbb, 0x3c, 0x83, 0x53, 0x99, 0x61);
4d4d77e
+	&data_byte(0x17, 0x2b, 0x04, 0x7e, 0xba, 0x77, 0xd6, 0x26);
4d4d77e
+	&data_byte(0xe1, 0x69, 0x14, 0x63, 0x55, 0x21, 0x0c, 0x7d);
4d4d77e
 &function_end_B("_x86_AES_decrypt");
4d4d77e
 
4d4d77e
 # void AES_decrypt (const void *inp,void *out,const AES_KEY *key);
4d4d77e
@@ -770,6 +744,18 @@
4d4d77e
 	&blindpop("ebp");
4d4d77e
 	&lea    ("ebp",&DWP(&label("AES_Td")."-".&label("pic_point"),"ebp"));
4d4d77e
 
4d4d77e
+	# prefetch Td4
4d4d77e
+	&lea	("ebp",&DWP(2048+128,"ebp"));
4d4d77e
+	&mov	($s0,&DWP(0-128,"ebp"));
4d4d77e
+	&mov	($s1,&DWP(32-128,"ebp"));
4d4d77e
+	&mov	($s2,&DWP(64-128,"ebp"));
4d4d77e
+	&mov	($s3,&DWP(96-128,"ebp"));
4d4d77e
+	&mov	($s0,&DWP(128-128,"ebp"));
4d4d77e
+	&mov	($s1,&DWP(160-128,"ebp"));
4d4d77e
+	&mov	($s2,&DWP(192-128,"ebp"));
4d4d77e
+	&mov	($s3,&DWP(224-128,"ebp"));
4d4d77e
+	&lea	("ebp",&DWP(-2048-128,"ebp"));
4d4d77e
+
4d4d77e
 	&mov	($s0,&DWP(0,$acc));		# load input data
4d4d77e
 	&mov	($s1,&DWP(4,$acc));
4d4d77e
 	&mov	($s2,&DWP(8,$acc));
4d4d77e
@@ -805,6 +791,7 @@
4d4d77e
 my $_tmp=&DWP(40,"esp");	#volatile variable
4d4d77e
 my $ivec=&DWP(44,"esp");	#ivec[16]
4d4d77e
 my $aes_key=&DWP(60,"esp");	#copy of aes_key
4d4d77e
+my $mark=&DWP(60+240,"esp");	#copy of aes_key->rounds
4d4d77e
 
4d4d77e
 &public_label("AES_Te");
4d4d77e
 &public_label("AES_Td");
4d4d77e
@@ -865,18 +852,27 @@
4d4d77e
 	&mov	($_key,$s3);		# save copy of key
4d4d77e
 	&mov	($_ivp,$acc);		# save copy of ivp
4d4d77e
 
4d4d77e
+	&mov	($mark,0);		# copy of aes_key->rounds = 0;
4d4d77e
 	if ($compromise) {
4d4d77e
 		&cmp	($s2,$compromise);
4d4d77e
 		&jb	(&label("skip_ecopy"));
4d4d77e
 	}
4d4d77e
-	# copy key schedule to stack
4d4d77e
-	&mov	("ecx",244/4);
4d4d77e
+	# do we copy key schedule to stack?
4d4d77e
+	&mov	($s1 eq "ebx" ? $s1 : "",$s3);
4d4d77e
+	&mov	($s2 eq "ecx" ? $s2 : "",244/4);
4d4d77e
+	&sub	($s1,"ebp");
4d4d77e
 	&mov	("esi",$s3);
4d4d77e
+	&and	($s1,0xfff);
4d4d77e
 	&lea	("edi",$aes_key);
4d4d77e
-	&mov	($_key,"edi");
4d4d77e
+	&cmp	($s1,2048);
4d4d77e
+	&jb	(&label("do_ecopy"));
4d4d77e
+	&cmp	($s1,4096-244);
4d4d77e
+	&jb	(&label("skip_ecopy"));
4d4d77e
 	&align	(4);
4d4d77e
-	&data_word(0xF689A5F3);	# rep movsd
4d4d77e
-	&set_label("skip_ecopy") if ($compromise);
4d4d77e
+	&set_label("do_ecopy");
4d4d77e
+		&mov	($_key,"edi");
4d4d77e
+		&data_word(0xA5F3F689);	# rep movsd
4d4d77e
+	&set_label("skip_ecopy");
4d4d77e
 
4d4d77e
 	&mov	($acc,$s0);
4d4d77e
 	&mov	($key,16);
4d4d77e
@@ -942,18 +938,16 @@
4d4d77e
 	&mov	(&DWP(8,$acc),$s2);
4d4d77e
 	&mov	(&DWP(12,$acc),$s3);
4d4d77e
 
4d4d77e
+	&cmp	($mark,0);		# was the key schedule copied?
4d4d77e
 	&mov	("edi",$_key);
4d4d77e
 	&mov	("esp",$_esp);
4d4d77e
-	if ($compromise) {
4d4d77e
-		&cmp	(&wparam(2),$compromise);
4d4d77e
-		&jb	(&label("skip_ezero"));
4d4d77e
-	}
4d4d77e
+	&je	(&label("skip_ezero"));
4d4d77e
 	# zero copy of key schedule
4d4d77e
 	&mov	("ecx",240/4);
4d4d77e
 	&xor	("eax","eax");
4d4d77e
 	&align	(4);
4d4d77e
-	&data_word(0xF689ABF3);	# rep stosd
4d4d77e
-	&set_label("skip_ezero") if ($compromise);
4d4d77e
+	&data_word(0xABF3F689);	# rep stosd
4d4d77e
+	&set_label("skip_ezero")
4d4d77e
 	&popf	();
4d4d77e
     &set_label("enc_out");
4d4d77e
 	&function_end_A();
4d4d77e
@@ -968,7 +962,7 @@
4d4d77e
 	&cmp	($key,$acc);			# compare with inp
4d4d77e
 	&je	(&label("enc_in_place"));
4d4d77e
 	&align	(4);
4d4d77e
-	&data_word(0xF689A4F3);	# rep movsb	# copy input
4d4d77e
+	&data_word(0xA4F3F689);	# rep movsb	# copy input
4d4d77e
 	&jmp	(&label("enc_skip_in_place"));
4d4d77e
     &set_label("enc_in_place");
4d4d77e
 	&lea	($key,&DWP(0,$key,$s2));
4d4d77e
@@ -976,7 +970,7 @@
4d4d77e
 	&mov	($s2,$s1);
4d4d77e
 	&xor	($s0,$s0);
4d4d77e
 	&align	(4);
4d4d77e
-	&data_word(0xF689AAF3);	# rep stosb	# zero tail
4d4d77e
+	&data_word(0xAAF3F689);	# rep stosb	# zero tail
4d4d77e
 	&pop	($key);				# pop ivp
4d4d77e
 
4d4d77e
 	&mov	($acc,$_out);			# output as input
4d4d77e
@@ -996,10 +990,10 @@
4d4d77e
 
4d4d77e
 	# ... and make sure it doesn't alias with AES_Td modulo 4096
4d4d77e
 	&mov	($s0,"ebp");
4d4d77e
-	&lea	($s1,&DWP(3072,"ebp"));
4d4d77e
+	&lea	($s1,&DWP(2048+256,"ebp"));
4d4d77e
 	&mov	($s3,$key);
4d4d77e
 	&and	($s0,0xfff);		# s = %ebp&0xfff
4d4d77e
-	&and	($s1,0xfff);		# e = (%ebp+3072)&0xfff
4d4d77e
+	&and	($s1,0xfff);		# e = (%ebp+2048+256)&0xfff
4d4d77e
 	&and	($s3,0xfff);		# p = %esp&0xfff
4d4d77e
 
4d4d77e
 	&cmp	($s3,$s1);		# if (p>=e) %esp =- (p-e);
4d4d77e
@@ -1030,21 +1024,30 @@
4d4d77e
 	&mov	($_key,$s3);		# save copy of key
4d4d77e
 	&mov	($_ivp,$acc);		# save copy of ivp
4d4d77e
 
4d4d77e
+	&mov	($mark,0);		# copy of aes_key->rounds = 0;
4d4d77e
 	if ($compromise) {
4d4d77e
 		&cmp	($s2,$compromise);
4d4d77e
 		&jb	(&label("skip_dcopy"));
4d4d77e
 	}
4d4d77e
-	# copy key schedule to stack
4d4d77e
-	&mov	("ecx",244/4);
4d4d77e
+	# do we copy key schedule to stack?
4d4d77e
+	&mov	($s1 eq "ebx" ? $s1 : "",$s3);
4d4d77e
+	&mov	($s2 eq "ecx" ? $s2 : "",244/4);
4d4d77e
+	&sub	($s1,"ebp");
4d4d77e
 	&mov	("esi",$s3);
4d4d77e
+	&and	($s1,0xfff);
4d4d77e
 	&lea	("edi",$aes_key);
4d4d77e
-	&mov	($_key,"edi");
4d4d77e
+	&cmp	($s1,2048+256);
4d4d77e
+	&jb	(&label("do_dcopy"));
4d4d77e
+	&cmp	($s1,4096-244);
4d4d77e
+	&jb	(&label("skip_dcopy"));
4d4d77e
 	&align	(4);
4d4d77e
-	&data_word(0xF689A5F3);	# rep movsd
4d4d77e
-	&set_label("skip_dcopy") if ($compromise);
4d4d77e
+	&set_label("do_dcopy");
4d4d77e
+		&mov	($_key,"edi");
4d4d77e
+		&data_word(0xA5F3F689);	# rep movsd
4d4d77e
+	&set_label("skip_dcopy");
4d4d77e
 
4d4d77e
 	&mov	($acc,$s0);
4d4d77e
-	&mov	($key,24);
4d4d77e
+	&mov	($key,18);
4d4d77e
 	&align	(4);
4d4d77e
 	&set_label("prefetch_td");
4d4d77e
 		&mov	($s0,&DWP(0,"ebp"));
4d4d77e
@@ -1054,7 +1057,7 @@
4d4d77e
 		&lea	("ebp",&DWP(128,"ebp"));
4d4d77e
 		&dec	($key);
4d4d77e
 	&jnz	(&label("prefetch_td"));
4d4d77e
-	&sub	("ebp",3072);
4d4d77e
+	&sub	("ebp",2048+256);
4d4d77e
 
4d4d77e
 	&cmp	($acc,$_out);
4d4d77e
 	&je	(&label("dec_in_place"));	# in-place processing...
4d4d77e
@@ -1121,7 +1124,7 @@
4d4d77e
 	&lea	($s2 eq "ecx" ? $s2 : "",&DWP(16,$acc));
4d4d77e
 	&mov	($acc eq "esi" ? $acc : "",$key);
4d4d77e
 	&mov	($key eq "edi" ? $key : "",$_out);	# load out
4d4d77e
-	&data_word(0xF689A4F3);	# rep movsb		# copy output
4d4d77e
+	&data_word(0xA4F3F689);	# rep movsb		# copy output
4d4d77e
 	&mov	($key,$_inp);				# use inp as temp ivp
4d4d77e
 	&jmp	(&label("dec_end"));
4d4d77e
 
4d4d77e
@@ -1188,22 +1191,20 @@
4d4d77e
 	&lea	($key,&DWP(0,$key,$s2));
4d4d77e
 	&lea	($acc,&DWP(16,$acc,$s2));
4d4d77e
 	&neg	($s2 eq "ecx" ? $s2 : "");
4d4d77e
-	&data_word(0xF689A4F3);	# rep movsb	# restore tail
4d4d77e
+	&data_word(0xA4F3F689);	# rep movsb	# restore tail
4d4d77e
 
4d4d77e
     &align	(4);
4d4d77e
     &set_label("dec_out");
4d4d77e
+    &cmp	($mark,0);		# was the key schedule copied?
4d4d77e
     &mov	("edi",$_key);
4d4d77e
     &mov	("esp",$_esp);
4d4d77e
-    if ($compromise) {
4d4d77e
-	&cmp	(&wparam(2),$compromise);
4d4d77e
-	&jb	(&label("skip_dzero"));
4d4d77e
-    }
4d4d77e
+    &je		(&label("skip_dzero"));
4d4d77e
     # zero copy of key schedule
4d4d77e
     &mov	("ecx",240/4);
4d4d77e
     &xor	("eax","eax");
4d4d77e
     &align	(4);
4d4d77e
-    &data_word(0xF689ABF3);	# rep stosd
4d4d77e
-    &set_label("skip_dzero") if ($compromise);
4d4d77e
+    &data_word(0xABF3F689);	# rep stosd
4d4d77e
+    &set_label("skip_dzero")
4d4d77e
     &popf	();
4d4d77e
 &function_end("AES_cbc_encrypt");
4d4d77e
 }