From d343e2c4a0699bb8a6fd8d47b06fe4f5a157c5e8 Mon Sep 17 00:00:00 2001 From: Jens Steube Date: Sun, 11 Apr 2021 11:53:47 +0200 Subject: [PATCH] Added support for true UTF8 to UTF16 conversion in kernel crypto library --- OpenCL/inc_common.cl | 262 ++++++++++++++++++ OpenCL/inc_common.h | 2 + OpenCL/inc_hash_md4.cl | 244 +--------------- OpenCL/inc_hash_md4.h | 4 - OpenCL/inc_hash_md5.cl | 244 +--------------- OpenCL/inc_hash_md5.h | 4 - OpenCL/inc_hash_ripemd160.cl | 244 +--------------- OpenCL/inc_hash_ripemd160.h | 4 - OpenCL/inc_hash_sha1.cl | 244 +--------------- OpenCL/inc_hash_sha1.h | 4 - OpenCL/inc_hash_sha224.cl | 244 +--------------- OpenCL/inc_hash_sha224.h | 4 - OpenCL/inc_hash_sha256.cl | 244 +--------------- OpenCL/inc_hash_sha256.h | 4 - OpenCL/inc_hash_sha384.cl | 404 +-------------------------- OpenCL/inc_hash_sha384.h | 4 - OpenCL/inc_hash_sha512.cl | 523 ++--------------------------------- OpenCL/inc_hash_sha512.h | 5 +- OpenCL/inc_hash_whirlpool.cl | 244 +--------------- OpenCL/inc_hash_whirlpool.h | 4 - OpenCL/m02100-pure.cl | 5 + docs/changes.txt | 4 +- 22 files changed, 391 insertions(+), 2554 deletions(-) diff --git a/OpenCL/inc_common.cl b/OpenCL/inc_common.cl index ba6edcc56..f8fc15724 100644 --- a/OpenCL/inc_common.cl +++ b/OpenCL/inc_common.cl @@ -1981,6 +1981,268 @@ DECLSPEC int find_hash (const u32 *digest, const u32 digests_cnt, GLOBAL_AS cons } #endif +// Constants and some code snippets from unicode.org's ConvertUTF.c +// Compiler can perfectly translate some of the branches and switch cases this into MOVC +// which is faster than lookup tables + +#define halfShift 10 + +#define halfBase 0x0010000 +#define halfMask 0x3FF + +#define UNI_MAX_BMP 0xFFFF +#define UNI_SUR_HIGH_START 0xD800 +#define UNI_SUR_HIGH_END 0xDBFF +#define UNI_SUR_LOW_START 0xDC00 +#define UNI_SUR_LOW_END 0xDFFF + +/* + * Magic values subtracted from a buffer value during UTF8 conversion. + * This table contains as many values as there might be trailing bytes + * in a UTF-8 sequence. + */ + +#define offsetsFromUTF8_0 0x00000000UL +#define offsetsFromUTF8_1 0x00003080UL +#define offsetsFromUTF8_2 0x000E2080UL +#define offsetsFromUTF8_3 0x03C82080UL +#define offsetsFromUTF8_4 0xFA082080UL +#define offsetsFromUTF8_5 0x82082080UL + +DECLSPEC int utf8_to_utf16le (const u32 *src_buf, int src_len, int src_size, u32 *dst_buf, int dst_size) +{ + const u8 *src_ptr = (const u8 *) src_buf; + u16 *dst_ptr = ( u16 *) dst_buf; + + int src_pos = 0; + int dst_pos = 0; + int dst_len = 0; + + while (src_pos < src_len) + { + const u8 c = src_ptr[src_pos]; + + int extraBytesToRead = 0; + + if (c >= 0xfc) + { + extraBytesToRead = 5; + } + else if (c >= 0xf8) + { + extraBytesToRead = 4; + } + else if (c >= 0xf0) + { + extraBytesToRead = 3; + } + else if (c >= 0xe0) + { + extraBytesToRead = 2; + } + else if (c >= 0xc0) + { + extraBytesToRead = 1; + } + + if ((src_pos + extraBytesToRead) >= src_size) return dst_len; + + u32 ch = 0; + + switch (extraBytesToRead) + { + case 5: + ch += src_ptr[src_pos++]; ch <<= 6; /* remember, illegal UTF-8 */ + ch += src_ptr[src_pos++]; ch <<= 6; /* remember, illegal UTF-8 */ + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_5; + break; + case 4: + ch += src_ptr[src_pos++]; ch <<= 6; /* remember, illegal UTF-8 */ + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_4; + break; + case 3: + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_3; + break; + case 2: + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_2; + break; + case 1: + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_1; + break; + case 0: + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_0; + break; + } + + /* Target is a character <= 0xFFFF */ + if (ch <= UNI_MAX_BMP) + { + if ((dst_len + 2) >= dst_size) return dst_len; + + dst_ptr[dst_pos++] = (u16) ch; + + dst_len += 2; + } + else + { + if ((dst_len + 4) >= dst_size) return dst_len; + + ch -= halfBase; + + dst_ptr[dst_pos++] = (u16) ((ch >> halfShift) + UNI_SUR_HIGH_START); + dst_ptr[dst_pos++] = (u16) ((ch & halfMask) + UNI_SUR_LOW_START); + + dst_len += 4; + } + } + + return dst_len; +} + +DECLSPEC int utf8_to_utf16le_global (GLOBAL_AS const u32 *src_buf, int src_len, int src_size, u32 *dst_buf, int dst_size) +{ + GLOBAL_AS const u8 *src_ptr = (GLOBAL_AS const u8 *) src_buf; + u16 *dst_ptr = ( u16 *) dst_buf; + + int src_pos = 0; + int dst_pos = 0; + int dst_len = 0; + + while (src_pos < src_len) + { + const u8 c = src_ptr[src_pos]; + + int extraBytesToRead = 0; + + if (c >= 0xfc) + { + extraBytesToRead = 5; + } + else if (c >= 0xf8) + { + extraBytesToRead = 4; + } + else if (c >= 0xf0) + { + extraBytesToRead = 3; + } + else if (c >= 0xe0) + { + extraBytesToRead = 2; + } + else if (c >= 0xc0) + { + extraBytesToRead = 1; + } + + if ((src_pos + extraBytesToRead) >= src_size) return dst_len; + + u32 ch = 0; + + switch (extraBytesToRead) + { + case 5: + ch += src_ptr[src_pos++]; ch <<= 6; /* remember, illegal UTF-8 */ + ch += src_ptr[src_pos++]; ch <<= 6; /* remember, illegal UTF-8 */ + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_5; + break; + case 4: + ch += src_ptr[src_pos++]; ch <<= 6; /* remember, illegal UTF-8 */ + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_4; + break; + case 3: + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_3; + break; + case 2: + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_2; + break; + case 1: + ch += src_ptr[src_pos++]; ch <<= 6; + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_1; + break; + case 0: + ch += src_ptr[src_pos++]; + ch -= offsetsFromUTF8_0; + break; + } + + /* Target is a character <= 0xFFFF */ + if (ch <= UNI_MAX_BMP) + { + if ((dst_len + 2) >= dst_size) return dst_len; + + dst_ptr[dst_pos++] = (u16) ch; + + dst_len += 2; + } + else + { + if ((dst_len + 4) >= dst_size) return dst_len; + + ch -= halfBase; + + dst_ptr[dst_pos++] = (u16) ((ch >> halfShift) + UNI_SUR_HIGH_START); + dst_ptr[dst_pos++] = (u16) ((ch & halfMask) + UNI_SUR_LOW_START); + + dst_len += 4; + } + } + + return dst_len; +} + +#undef halfShift + +#undef halfBase +#undef halfMask + +#undef UNI_MAX_BMP +#undef UNI_SUR_HIGH_START +#undef UNI_SUR_HIGH_END +#undef UNI_SUR_LOW_START +#undef UNI_SUR_LOW_END + +#undef offsetsFromUTF8_0 +#undef offsetsFromUTF8_1 +#undef offsetsFromUTF8_2 +#undef offsetsFromUTF8_3 +#undef offsetsFromUTF8_4 +#undef offsetsFromUTF8_5 + DECLSPEC int pkcs_padding_bs8 (const u32 *data_buf, const int data_len) { if (data_len == 0) return -1; // cannot have zero length, is important to avoid out of boundary reads diff --git a/OpenCL/inc_common.h b/OpenCL/inc_common.h index ebdbe4675..9b3437326 100644 --- a/OpenCL/inc_common.h +++ b/OpenCL/inc_common.h @@ -234,6 +234,8 @@ DECLSPEC int hash_comp (const u32 *d1, GLOBAL_AS const u32 *d2); DECLSPEC int find_hash (const u32 *digest, const u32 digests_cnt, GLOBAL_AS const digest_t *digests_buf); #endif +DECLSPEC int utf8_to_utf16le (const u32 *src_buf, int src_len, int src_size, u32 *dst_buf, int dst_size); +DECLSPEC int utf8_to_utf16le_global (GLOBAL_AS const u32 *src_buf, int src_len, int src_size, u32 *dst_buf, int dst_size); DECLSPEC int pkcs_padding_bs8 (const u32 *data_buf, const int data_len); DECLSPEC int pkcs_padding_bs16 (const u32 *data_buf, const int data_len); DECLSPEC int asn1_detect (const u32 *buf, const int len); diff --git a/OpenCL/inc_hash_md4.cl b/OpenCL/inc_hash_md4.cl index eeb28cd17..3fa89680c 100644 --- a/OpenCL/inc_hash_md4.cl +++ b/OpenCL/inc_hash_md4.cl @@ -363,120 +363,20 @@ DECLSPEC void md4_update_swap (md4_ctx_t *ctx, const u32 *w, const int len) DECLSPEC void md4_update_utf16le (md4_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - md4_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - md4_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + md4_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void md4_update_utf16le_swap (md4_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - md4_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - md4_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + md4_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void md4_update_global (md4_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) @@ -619,120 +519,20 @@ DECLSPEC void md4_update_global_swap (md4_ctx_t *ctx, GLOBAL_AS const u32 *w, co DECLSPEC void md4_update_global_utf16le (md4_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; + u32 w_utf16_buf[64] = { 0 }; - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - md4_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - md4_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + md4_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void md4_update_global_utf16le_swap (md4_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + u32 w_utf16_buf[64] = { 0 }; - md4_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - md4_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + md4_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void md4_final (md4_ctx_t *ctx) @@ -1068,16 +868,6 @@ DECLSPEC void md4_hmac_update_swap (md4_hmac_ctx_t *ctx, const u32 *w, const int md4_update_swap (&ctx->ipad, w, len); } -DECLSPEC void md4_hmac_update_utf16le (md4_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - md4_update_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void md4_hmac_update_utf16le_swap (md4_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - md4_update_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void md4_hmac_update_global (md4_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { md4_update_global (&ctx->ipad, w, len); @@ -1088,16 +878,6 @@ DECLSPEC void md4_hmac_update_global_swap (md4_hmac_ctx_t *ctx, GLOBAL_AS const md4_update_global_swap (&ctx->ipad, w, len); } -DECLSPEC void md4_hmac_update_global_utf16le (md4_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - md4_update_global_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void md4_hmac_update_global_utf16le_swap (md4_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - md4_update_global_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void md4_hmac_final (md4_hmac_ctx_t *ctx) { md4_final (&ctx->ipad); diff --git a/OpenCL/inc_hash_md4.h b/OpenCL/inc_hash_md4.h index 7c3b31894..c8b3351a1 100644 --- a/OpenCL/inc_hash_md4.h +++ b/OpenCL/inc_hash_md4.h @@ -102,12 +102,8 @@ DECLSPEC void md4_hmac_init_global_swap (md4_hmac_ctx_t *ctx, GLOBAL_AS const u3 DECLSPEC void md4_hmac_update_64 (md4_hmac_ctx_t *ctx, u32 *w0, u32 *w1, u32 *w2, u32 *w3, const int len); DECLSPEC void md4_hmac_update (md4_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void md4_hmac_update_swap (md4_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void md4_hmac_update_utf16le (md4_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void md4_hmac_update_utf16le_swap (md4_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void md4_hmac_update_global (md4_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void md4_hmac_update_global_swap (md4_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void md4_hmac_update_global_utf16le (md4_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void md4_hmac_update_global_utf16le_swap (md4_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void md4_hmac_final (md4_hmac_ctx_t *ctx); DECLSPEC void md4_transform_vector (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, u32x *digest); DECLSPEC void md4_init_vector (md4_ctx_vector_t *ctx); diff --git a/OpenCL/inc_hash_md5.cl b/OpenCL/inc_hash_md5.cl index 3c52c1f40..73d236467 100644 --- a/OpenCL/inc_hash_md5.cl +++ b/OpenCL/inc_hash_md5.cl @@ -399,120 +399,20 @@ DECLSPEC void md5_update_swap (md5_ctx_t *ctx, const u32 *w, const int len) DECLSPEC void md5_update_utf16le (md5_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - md5_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - md5_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + md5_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void md5_update_utf16le_swap (md5_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - md5_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - md5_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + md5_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void md5_update_global (md5_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) @@ -655,120 +555,20 @@ DECLSPEC void md5_update_global_swap (md5_ctx_t *ctx, GLOBAL_AS const u32 *w, co DECLSPEC void md5_update_global_utf16le (md5_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; + u32 w_utf16_buf[64] = { 0 }; - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - md5_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - md5_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + md5_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void md5_update_global_utf16le_swap (md5_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + u32 w_utf16_buf[64] = { 0 }; - md5_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - md5_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + md5_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void md5_final (md5_ctx_t *ctx) @@ -1104,16 +904,6 @@ DECLSPEC void md5_hmac_update_swap (md5_hmac_ctx_t *ctx, const u32 *w, const int md5_update_swap (&ctx->ipad, w, len); } -DECLSPEC void md5_hmac_update_utf16le (md5_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - md5_update_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void md5_hmac_update_utf16le_swap (md5_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - md5_update_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void md5_hmac_update_global (md5_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { md5_update_global (&ctx->ipad, w, len); @@ -1124,16 +914,6 @@ DECLSPEC void md5_hmac_update_global_swap (md5_hmac_ctx_t *ctx, GLOBAL_AS const md5_update_global_swap (&ctx->ipad, w, len); } -DECLSPEC void md5_hmac_update_global_utf16le (md5_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - md5_update_global_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void md5_hmac_update_global_utf16le_swap (md5_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - md5_update_global_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void md5_hmac_final (md5_hmac_ctx_t *ctx) { md5_final (&ctx->ipad); diff --git a/OpenCL/inc_hash_md5.h b/OpenCL/inc_hash_md5.h index 1e6eaaf93..273a35bb3 100644 --- a/OpenCL/inc_hash_md5.h +++ b/OpenCL/inc_hash_md5.h @@ -109,12 +109,8 @@ DECLSPEC void md5_hmac_init_global_swap (md5_hmac_ctx_t *ctx, GLOBAL_AS const u3 DECLSPEC void md5_hmac_update_64 (md5_hmac_ctx_t *ctx, u32 *w0, u32 *w1, u32 *w2, u32 *w3, const int len); DECLSPEC void md5_hmac_update (md5_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void md5_hmac_update_swap (md5_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void md5_hmac_update_utf16le (md5_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void md5_hmac_update_utf16le_swap (md5_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void md5_hmac_update_global (md5_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void md5_hmac_update_global_swap (md5_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void md5_hmac_update_global_utf16le (md5_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void md5_hmac_update_global_utf16le_swap (md5_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void md5_hmac_final (md5_hmac_ctx_t *ctx); DECLSPEC void md5_transform_vector (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, u32x *digest); DECLSPEC void md5_init_vector (md5_ctx_vector_t *ctx); diff --git a/OpenCL/inc_hash_ripemd160.cl b/OpenCL/inc_hash_ripemd160.cl index bcf1074ac..1b18f01b8 100644 --- a/OpenCL/inc_hash_ripemd160.cl +++ b/OpenCL/inc_hash_ripemd160.cl @@ -497,120 +497,20 @@ DECLSPEC void ripemd160_update_swap (ripemd160_ctx_t *ctx, const u32 *w, const i DECLSPEC void ripemd160_update_utf16le (ripemd160_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - ripemd160_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - ripemd160_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + ripemd160_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void ripemd160_update_utf16le_swap (ripemd160_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - ripemd160_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - ripemd160_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + ripemd160_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void ripemd160_update_global (ripemd160_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) @@ -753,120 +653,20 @@ DECLSPEC void ripemd160_update_global_swap (ripemd160_ctx_t *ctx, GLOBAL_AS cons DECLSPEC void ripemd160_update_global_utf16le (ripemd160_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; + u32 w_utf16_buf[64] = { 0 }; - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - ripemd160_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - ripemd160_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + ripemd160_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void ripemd160_update_global_utf16le_swap (ripemd160_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + u32 w_utf16_buf[64] = { 0 }; - ripemd160_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - ripemd160_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + ripemd160_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void ripemd160_final (ripemd160_ctx_t *ctx) @@ -1202,16 +1002,6 @@ DECLSPEC void ripemd160_hmac_update_swap (ripemd160_hmac_ctx_t *ctx, const u32 * ripemd160_update_swap (&ctx->ipad, w, len); } -DECLSPEC void ripemd160_hmac_update_utf16le (ripemd160_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - ripemd160_update_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void ripemd160_hmac_update_utf16le_swap (ripemd160_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - ripemd160_update_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void ripemd160_hmac_update_global (ripemd160_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { ripemd160_update_global (&ctx->ipad, w, len); @@ -1222,16 +1012,6 @@ DECLSPEC void ripemd160_hmac_update_global_swap (ripemd160_hmac_ctx_t *ctx, GLOB ripemd160_update_global_swap (&ctx->ipad, w, len); } -DECLSPEC void ripemd160_hmac_update_global_utf16le (ripemd160_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - ripemd160_update_global_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void ripemd160_hmac_update_global_utf16le_swap (ripemd160_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - ripemd160_update_global_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void ripemd160_hmac_final (ripemd160_hmac_ctx_t *ctx) { ripemd160_final (&ctx->ipad); diff --git a/OpenCL/inc_hash_ripemd160.h b/OpenCL/inc_hash_ripemd160.h index 25a69ed56..70fa3f60f 100644 --- a/OpenCL/inc_hash_ripemd160.h +++ b/OpenCL/inc_hash_ripemd160.h @@ -122,12 +122,8 @@ DECLSPEC void ripemd160_hmac_init_global_swap (ripemd160_hmac_ctx_t *ctx, GLOBAL DECLSPEC void ripemd160_hmac_update_64 (ripemd160_hmac_ctx_t *ctx, u32 *w0, u32 *w1, u32 *w2, u32 *w3, const int len); DECLSPEC void ripemd160_hmac_update (ripemd160_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void ripemd160_hmac_update_swap (ripemd160_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void ripemd160_hmac_update_utf16le (ripemd160_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void ripemd160_hmac_update_utf16le_swap (ripemd160_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void ripemd160_hmac_update_global (ripemd160_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void ripemd160_hmac_update_global_swap (ripemd160_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void ripemd160_hmac_update_global_utf16le (ripemd160_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void ripemd160_hmac_update_global_utf16le_swap (ripemd160_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void ripemd160_hmac_final (ripemd160_hmac_ctx_t *ctx); DECLSPEC void ripemd160_transform_vector (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, u32x *digest); DECLSPEC void ripemd160_init_vector (ripemd160_ctx_vector_t *ctx); diff --git a/OpenCL/inc_hash_sha1.cl b/OpenCL/inc_hash_sha1.cl index a8f754c1a..f71f04193 100644 --- a/OpenCL/inc_hash_sha1.cl +++ b/OpenCL/inc_hash_sha1.cl @@ -612,120 +612,20 @@ DECLSPEC void sha1_update_swap (sha1_ctx_t *ctx, const u32 *w, const int len) DECLSPEC void sha1_update_utf16le (sha1_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha1_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - sha1_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha1_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha1_update_utf16le_swap (sha1_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - sha1_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - sha1_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha1_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha1_update_utf16be (sha1_ctx_t *ctx, const u32 *w, const int len) @@ -986,120 +886,20 @@ DECLSPEC void sha1_update_global_swap (sha1_ctx_t *ctx, GLOBAL_AS const u32 *w, DECLSPEC void sha1_update_global_utf16le (sha1_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + u32 w_utf16_buf[64] = { 0 }; - sha1_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha1_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha1_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha1_update_global_utf16le_swap (sha1_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - sha1_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } + u32 w_utf16_buf[64] = { 0 }; - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - sha1_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha1_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha1_update_global_utf16be (sha1_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) @@ -1553,16 +1353,6 @@ DECLSPEC void sha1_hmac_update_swap (sha1_hmac_ctx_t *ctx, const u32 *w, const i sha1_update_swap (&ctx->ipad, w, len); } -DECLSPEC void sha1_hmac_update_utf16le (sha1_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - sha1_update_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void sha1_hmac_update_utf16le_swap (sha1_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - sha1_update_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void sha1_hmac_update_global (sha1_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { sha1_update_global (&ctx->ipad, w, len); @@ -1573,16 +1363,6 @@ DECLSPEC void sha1_hmac_update_global_swap (sha1_hmac_ctx_t *ctx, GLOBAL_AS cons sha1_update_global_swap (&ctx->ipad, w, len); } -DECLSPEC void sha1_hmac_update_global_utf16le (sha1_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - sha1_update_global_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void sha1_hmac_update_global_utf16le_swap (sha1_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - sha1_update_global_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void sha1_hmac_final (sha1_hmac_ctx_t *ctx) { sha1_final (&ctx->ipad); diff --git a/OpenCL/inc_hash_sha1.h b/OpenCL/inc_hash_sha1.h index 2ff36fdad..69f6b58d4 100644 --- a/OpenCL/inc_hash_sha1.h +++ b/OpenCL/inc_hash_sha1.h @@ -114,12 +114,8 @@ DECLSPEC void sha1_hmac_init_global_swap (sha1_hmac_ctx_t *ctx, GLOBAL_AS const DECLSPEC void sha1_hmac_update_64 (sha1_hmac_ctx_t *ctx, u32 *w0, u32 *w1, u32 *w2, u32 *w3, const int len); DECLSPEC void sha1_hmac_update (sha1_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha1_hmac_update_swap (sha1_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void sha1_hmac_update_utf16le (sha1_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void sha1_hmac_update_utf16le_swap (sha1_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha1_hmac_update_global (sha1_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha1_hmac_update_global_swap (sha1_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void sha1_hmac_update_global_utf16le (sha1_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void sha1_hmac_update_global_utf16le_swap (sha1_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha1_hmac_final (sha1_hmac_ctx_t *ctx); DECLSPEC void sha1_transform_vector (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, u32x *digest); DECLSPEC void sha1_init_vector (sha1_ctx_vector_t *ctx); diff --git a/OpenCL/inc_hash_sha224.cl b/OpenCL/inc_hash_sha224.cl index 72f3dac99..a5b780376 100644 --- a/OpenCL/inc_hash_sha224.cl +++ b/OpenCL/inc_hash_sha224.cl @@ -414,120 +414,20 @@ DECLSPEC void sha224_update_swap (sha224_ctx_t *ctx, const u32 *w, const int len DECLSPEC void sha224_update_utf16le (sha224_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha224_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - sha224_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha224_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha224_update_utf16le_swap (sha224_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - sha224_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - sha224_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha224_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha224_update_global (sha224_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) @@ -670,120 +570,20 @@ DECLSPEC void sha224_update_global_swap (sha224_ctx_t *ctx, GLOBAL_AS const u32 DECLSPEC void sha224_update_global_utf16le (sha224_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; + u32 w_utf16_buf[64] = { 0 }; - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - sha224_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha224_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha224_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha224_update_global_utf16le_swap (sha224_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + u32 w_utf16_buf[64] = { 0 }; - sha224_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - sha224_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha224_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha224_final (sha224_ctx_t *ctx) @@ -1119,16 +919,6 @@ DECLSPEC void sha224_hmac_update_swap (sha224_hmac_ctx_t *ctx, const u32 *w, con sha224_update_swap (&ctx->ipad, w, len); } -DECLSPEC void sha224_hmac_update_utf16le (sha224_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - sha224_update_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void sha224_hmac_update_utf16le_swap (sha224_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - sha224_update_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void sha224_hmac_update_global (sha224_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { sha224_update_global (&ctx->ipad, w, len); @@ -1139,16 +929,6 @@ DECLSPEC void sha224_hmac_update_global_swap (sha224_hmac_ctx_t *ctx, GLOBAL_AS sha224_update_global_swap (&ctx->ipad, w, len); } -DECLSPEC void sha224_hmac_update_global_utf16le (sha224_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - sha224_update_global_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void sha224_hmac_update_global_utf16le_swap (sha224_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - sha224_update_global_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void sha224_hmac_final (sha224_hmac_ctx_t *ctx) { sha224_final (&ctx->ipad); diff --git a/OpenCL/inc_hash_sha224.h b/OpenCL/inc_hash_sha224.h index d68c79d65..46f03a35d 100644 --- a/OpenCL/inc_hash_sha224.h +++ b/OpenCL/inc_hash_sha224.h @@ -109,12 +109,8 @@ DECLSPEC void sha224_hmac_init_global_swap (sha224_hmac_ctx_t *ctx, GLOBAL_AS co DECLSPEC void sha224_hmac_update_64 (sha224_hmac_ctx_t *ctx, u32 *w0, u32 *w1, u32 *w2, u32 *w3, const int len); DECLSPEC void sha224_hmac_update (sha224_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha224_hmac_update_swap (sha224_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void sha224_hmac_update_utf16le (sha224_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void sha224_hmac_update_utf16le_swap (sha224_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha224_hmac_update_global (sha224_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha224_hmac_update_global_swap (sha224_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void sha224_hmac_update_global_utf16le (sha224_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void sha224_hmac_update_global_utf16le_swap (sha224_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha224_hmac_final (sha224_hmac_ctx_t *ctx); DECLSPEC void sha224_transform_vector (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, u32x *digest); DECLSPEC void sha224_init_vector (sha224_ctx_vector_t *ctx); diff --git a/OpenCL/inc_hash_sha256.cl b/OpenCL/inc_hash_sha256.cl index 430b0e8b9..49bb19f3d 100644 --- a/OpenCL/inc_hash_sha256.cl +++ b/OpenCL/inc_hash_sha256.cl @@ -414,120 +414,20 @@ DECLSPEC void sha256_update_swap (sha256_ctx_t *ctx, const u32 *w, const int len DECLSPEC void sha256_update_utf16le (sha256_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha256_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - sha256_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha256_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha256_update_utf16le_swap (sha256_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - sha256_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - sha256_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha256_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha256_update_global (sha256_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) @@ -670,120 +570,20 @@ DECLSPEC void sha256_update_global_swap (sha256_ctx_t *ctx, GLOBAL_AS const u32 DECLSPEC void sha256_update_global_utf16le (sha256_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; + u32 w_utf16_buf[64] = { 0 }; - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - sha256_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha256_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha256_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha256_update_global_utf16le_swap (sha256_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + u32 w_utf16_buf[64] = { 0 }; - sha256_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - sha256_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + sha256_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha256_final (sha256_ctx_t *ctx) @@ -1119,16 +919,6 @@ DECLSPEC void sha256_hmac_update_swap (sha256_hmac_ctx_t *ctx, const u32 *w, con sha256_update_swap (&ctx->ipad, w, len); } -DECLSPEC void sha256_hmac_update_utf16le (sha256_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - sha256_update_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void sha256_hmac_update_utf16le_swap (sha256_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - sha256_update_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void sha256_hmac_update_global (sha256_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { sha256_update_global (&ctx->ipad, w, len); @@ -1139,16 +929,6 @@ DECLSPEC void sha256_hmac_update_global_swap (sha256_hmac_ctx_t *ctx, GLOBAL_AS sha256_update_global_swap (&ctx->ipad, w, len); } -DECLSPEC void sha256_hmac_update_global_utf16le (sha256_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - sha256_update_global_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void sha256_hmac_update_global_utf16le_swap (sha256_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - sha256_update_global_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void sha256_hmac_final (sha256_hmac_ctx_t *ctx) { sha256_final (&ctx->ipad); diff --git a/OpenCL/inc_hash_sha256.h b/OpenCL/inc_hash_sha256.h index ccf5a79f8..bc655d80b 100644 --- a/OpenCL/inc_hash_sha256.h +++ b/OpenCL/inc_hash_sha256.h @@ -109,12 +109,8 @@ DECLSPEC void sha256_hmac_init_global_swap (sha256_hmac_ctx_t *ctx, GLOBAL_AS co DECLSPEC void sha256_hmac_update_64 (sha256_hmac_ctx_t *ctx, u32 *w0, u32 *w1, u32 *w2, u32 *w3, const int len); DECLSPEC void sha256_hmac_update (sha256_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha256_hmac_update_swap (sha256_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void sha256_hmac_update_utf16le (sha256_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void sha256_hmac_update_utf16le_swap (sha256_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha256_hmac_update_global (sha256_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha256_hmac_update_global_swap (sha256_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void sha256_hmac_update_global_utf16le (sha256_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void sha256_hmac_update_global_utf16le_swap (sha256_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha256_hmac_final (sha256_hmac_ctx_t *ctx); DECLSPEC void sha256_transform_vector (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, u32x *digest); DECLSPEC void sha256_init_vector (sha256_ctx_vector_t *ctx); diff --git a/OpenCL/inc_hash_sha384.cl b/OpenCL/inc_hash_sha384.cl index d63a5ab2b..ef09c26f6 100644 --- a/OpenCL/inc_hash_sha384.cl +++ b/OpenCL/inc_hash_sha384.cl @@ -622,200 +622,20 @@ DECLSPEC void sha384_update_swap (sha384_ctx_t *ctx, const u32 *w, const int len DECLSPEC void sha384_update_utf16le (sha384_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - u32 w4[4]; - u32 w5[4]; - u32 w6[4]; - u32 w7[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 64; pos1 += 64, pos4 += 16) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; - - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha384_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, 64 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - sha384_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, (len - pos1) * 2); + sha384_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha384_update_utf16le_swap (sha384_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - u32 w4[4]; - u32 w5[4]; - u32 w6[4]; - u32 w7[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 64; pos1 += 64, pos4 += 16) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; - - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - w4[0] = hc_swap32_S (w4[0]); - w4[1] = hc_swap32_S (w4[1]); - w4[2] = hc_swap32_S (w4[2]); - w4[3] = hc_swap32_S (w4[3]); - w5[0] = hc_swap32_S (w5[0]); - w5[1] = hc_swap32_S (w5[1]); - w5[2] = hc_swap32_S (w5[2]); - w5[3] = hc_swap32_S (w5[3]); - w6[0] = hc_swap32_S (w6[0]); - w6[1] = hc_swap32_S (w6[1]); - w6[2] = hc_swap32_S (w6[2]); - w6[3] = hc_swap32_S (w6[3]); - w7[0] = hc_swap32_S (w7[0]); - w7[1] = hc_swap32_S (w7[1]); - w7[2] = hc_swap32_S (w7[2]); - w7[3] = hc_swap32_S (w7[3]); - - sha384_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, 64 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - w4[0] = hc_swap32_S (w4[0]); - w4[1] = hc_swap32_S (w4[1]); - w4[2] = hc_swap32_S (w4[2]); - w4[3] = hc_swap32_S (w4[3]); - w5[0] = hc_swap32_S (w5[0]); - w5[1] = hc_swap32_S (w5[1]); - w5[2] = hc_swap32_S (w5[2]); - w5[3] = hc_swap32_S (w5[3]); - w6[0] = hc_swap32_S (w6[0]); - w6[1] = hc_swap32_S (w6[1]); - w6[2] = hc_swap32_S (w6[2]); - w6[3] = hc_swap32_S (w6[3]); - w7[0] = hc_swap32_S (w7[0]); - w7[1] = hc_swap32_S (w7[1]); - w7[2] = hc_swap32_S (w7[2]); - w7[3] = hc_swap32_S (w7[3]); - - sha384_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, (len - pos1) * 2); + sha384_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha384_update_global (sha384_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) @@ -1062,200 +882,20 @@ DECLSPEC void sha384_update_global_swap (sha384_ctx_t *ctx, GLOBAL_AS const u32 DECLSPEC void sha384_update_global_utf16le (sha384_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - u32 w4[4]; - u32 w5[4]; - u32 w6[4]; - u32 w7[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 64; pos1 += 64, pos4 += 16) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; - - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha384_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, 64 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - sha384_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, (len - pos1) * 2); + sha384_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha384_update_global_utf16le_swap (sha384_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - u32 w4[4]; - u32 w5[4]; - u32 w6[4]; - u32 w7[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 64; pos1 += 64, pos4 += 16) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; - - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - w4[0] = hc_swap32_S (w4[0]); - w4[1] = hc_swap32_S (w4[1]); - w4[2] = hc_swap32_S (w4[2]); - w4[3] = hc_swap32_S (w4[3]); - w5[0] = hc_swap32_S (w5[0]); - w5[1] = hc_swap32_S (w5[1]); - w5[2] = hc_swap32_S (w5[2]); - w5[3] = hc_swap32_S (w5[3]); - w6[0] = hc_swap32_S (w6[0]); - w6[1] = hc_swap32_S (w6[1]); - w6[2] = hc_swap32_S (w6[2]); - w6[3] = hc_swap32_S (w6[3]); - w7[0] = hc_swap32_S (w7[0]); - w7[1] = hc_swap32_S (w7[1]); - w7[2] = hc_swap32_S (w7[2]); - w7[3] = hc_swap32_S (w7[3]); + u32 w_utf16_buf[64] = { 0 }; - sha384_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, 64 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; - - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - w4[0] = hc_swap32_S (w4[0]); - w4[1] = hc_swap32_S (w4[1]); - w4[2] = hc_swap32_S (w4[2]); - w4[3] = hc_swap32_S (w4[3]); - w5[0] = hc_swap32_S (w5[0]); - w5[1] = hc_swap32_S (w5[1]); - w5[2] = hc_swap32_S (w5[2]); - w5[3] = hc_swap32_S (w5[3]); - w6[0] = hc_swap32_S (w6[0]); - w6[1] = hc_swap32_S (w6[1]); - w6[2] = hc_swap32_S (w6[2]); - w6[3] = hc_swap32_S (w6[3]); - w7[0] = hc_swap32_S (w7[0]); - w7[1] = hc_swap32_S (w7[1]); - w7[2] = hc_swap32_S (w7[2]); - w7[3] = hc_swap32_S (w7[3]); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - sha384_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, (len - pos1) * 2); + sha384_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha384_final (sha384_ctx_t *ctx) @@ -1787,16 +1427,6 @@ DECLSPEC void sha384_hmac_update_swap (sha384_hmac_ctx_t *ctx, const u32 *w, con sha384_update_swap (&ctx->ipad, w, len); } -DECLSPEC void sha384_hmac_update_utf16le (sha384_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - sha384_update_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void sha384_hmac_update_utf16le_swap (sha384_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - sha384_update_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void sha384_hmac_update_global (sha384_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { sha384_update_global (&ctx->ipad, w, len); @@ -1807,16 +1437,6 @@ DECLSPEC void sha384_hmac_update_global_swap (sha384_hmac_ctx_t *ctx, GLOBAL_AS sha384_update_global_swap (&ctx->ipad, w, len); } -DECLSPEC void sha384_hmac_update_global_utf16le (sha384_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - sha384_update_global_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void sha384_hmac_update_global_utf16le_swap (sha384_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - sha384_update_global_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void sha384_hmac_final (sha384_hmac_ctx_t *ctx) { sha384_final (&ctx->ipad); diff --git a/OpenCL/inc_hash_sha384.h b/OpenCL/inc_hash_sha384.h index 92266b24a..e3705c206 100644 --- a/OpenCL/inc_hash_sha384.h +++ b/OpenCL/inc_hash_sha384.h @@ -123,12 +123,8 @@ DECLSPEC void sha384_hmac_init_global_swap (sha384_hmac_ctx_t *ctx, GLOBAL_AS co DECLSPEC void sha384_hmac_update_128 (sha384_hmac_ctx_t *ctx, u32 *w0, u32 *w1, u32 *w2, u32 *w3, u32 *w4, u32 *w5, u32 *w6, u32 *w7, const int len); DECLSPEC void sha384_hmac_update (sha384_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha384_hmac_update_swap (sha384_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void sha384_hmac_update_utf16le (sha384_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void sha384_hmac_update_utf16le_swap (sha384_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha384_hmac_update_global (sha384_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha384_hmac_update_global_swap (sha384_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void sha384_hmac_update_global_utf16le (sha384_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void sha384_hmac_update_global_utf16le_swap (sha384_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha384_hmac_final (sha384_hmac_ctx_t *ctx); DECLSPEC void sha384_transform_vector (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, const u32x *w4, const u32x *w5, const u32x *w6, const u32x *w7, u64x *digest); DECLSPEC void sha384_init_vector (sha384_ctx_vector_t *ctx); diff --git a/OpenCL/inc_hash_sha512.cl b/OpenCL/inc_hash_sha512.cl index 0f5ca288a..19aee7368 100644 --- a/OpenCL/inc_hash_sha512.cl +++ b/OpenCL/inc_hash_sha512.cl @@ -622,200 +622,20 @@ DECLSPEC void sha512_update_swap (sha512_ctx_t *ctx, const u32 *w, const int len DECLSPEC void sha512_update_utf16le (sha512_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - u32 w4[4]; - u32 w5[4]; - u32 w6[4]; - u32 w7[4]; + u32 w_utf16_buf[64] = { 0 }; - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 64; pos1 += 64, pos4 += 16) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha512_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, 64 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; - - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha512_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, (len - pos1) * 2); + sha512_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha512_update_utf16le_swap (sha512_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - u32 w4[4]; - u32 w5[4]; - u32 w6[4]; - u32 w7[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 64; pos1 += 64, pos4 += 16) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - w4[0] = hc_swap32_S (w4[0]); - w4[1] = hc_swap32_S (w4[1]); - w4[2] = hc_swap32_S (w4[2]); - w4[3] = hc_swap32_S (w4[3]); - w5[0] = hc_swap32_S (w5[0]); - w5[1] = hc_swap32_S (w5[1]); - w5[2] = hc_swap32_S (w5[2]); - w5[3] = hc_swap32_S (w5[3]); - w6[0] = hc_swap32_S (w6[0]); - w6[1] = hc_swap32_S (w6[1]); - w6[2] = hc_swap32_S (w6[2]); - w6[3] = hc_swap32_S (w6[3]); - w7[0] = hc_swap32_S (w7[0]); - w7[1] = hc_swap32_S (w7[1]); - w7[2] = hc_swap32_S (w7[2]); - w7[3] = hc_swap32_S (w7[3]); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - sha512_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, 64 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; - - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - w4[0] = hc_swap32_S (w4[0]); - w4[1] = hc_swap32_S (w4[1]); - w4[2] = hc_swap32_S (w4[2]); - w4[3] = hc_swap32_S (w4[3]); - w5[0] = hc_swap32_S (w5[0]); - w5[1] = hc_swap32_S (w5[1]); - w5[2] = hc_swap32_S (w5[2]); - w5[3] = hc_swap32_S (w5[3]); - w6[0] = hc_swap32_S (w6[0]); - w6[1] = hc_swap32_S (w6[1]); - w6[2] = hc_swap32_S (w6[2]); - w6[3] = hc_swap32_S (w6[3]); - w7[0] = hc_swap32_S (w7[0]); - w7[1] = hc_swap32_S (w7[1]); - w7[2] = hc_swap32_S (w7[2]); - w7[3] = hc_swap32_S (w7[3]); - - sha512_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, (len - pos1) * 2); + sha512_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha512_update_global (sha512_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) @@ -1062,200 +882,20 @@ DECLSPEC void sha512_update_global_swap (sha512_ctx_t *ctx, GLOBAL_AS const u32 DECLSPEC void sha512_update_global_utf16le (sha512_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - u32 w4[4]; - u32 w5[4]; - u32 w6[4]; - u32 w7[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 64; pos1 += 64, pos4 += 16) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; - - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + u32 w_utf16_buf[64] = { 0 }; - sha512_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, 64 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - sha512_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, (len - pos1) * 2); + sha512_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha512_update_global_utf16le_swap (sha512_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - u32 w4[4]; - u32 w5[4]; - u32 w6[4]; - u32 w7[4]; + u32 w_utf16_buf[64] = { 0 }; - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 64; pos1 += 64, pos4 += 16) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; - - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - w4[0] = hc_swap32_S (w4[0]); - w4[1] = hc_swap32_S (w4[1]); - w4[2] = hc_swap32_S (w4[2]); - w4[3] = hc_swap32_S (w4[3]); - w5[0] = hc_swap32_S (w5[0]); - w5[1] = hc_swap32_S (w5[1]); - w5[2] = hc_swap32_S (w5[2]); - w5[3] = hc_swap32_S (w5[3]); - w6[0] = hc_swap32_S (w6[0]); - w6[1] = hc_swap32_S (w6[1]); - w6[2] = hc_swap32_S (w6[2]); - w6[3] = hc_swap32_S (w6[3]); - w7[0] = hc_swap32_S (w7[0]); - w7[1] = hc_swap32_S (w7[1]); - w7[2] = hc_swap32_S (w7[2]); - w7[3] = hc_swap32_S (w7[3]); - - sha512_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, 64 * 2); - } + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - w2[0] = w[pos4 + 8]; - w2[1] = w[pos4 + 9]; - w2[2] = w[pos4 + 10]; - w2[3] = w[pos4 + 11]; - w3[0] = w[pos4 + 12]; - w3[1] = w[pos4 + 13]; - w3[2] = w[pos4 + 14]; - w3[3] = w[pos4 + 15]; - - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - w4[0] = hc_swap32_S (w4[0]); - w4[1] = hc_swap32_S (w4[1]); - w4[2] = hc_swap32_S (w4[2]); - w4[3] = hc_swap32_S (w4[3]); - w5[0] = hc_swap32_S (w5[0]); - w5[1] = hc_swap32_S (w5[1]); - w5[2] = hc_swap32_S (w5[2]); - w5[3] = hc_swap32_S (w5[3]); - w6[0] = hc_swap32_S (w6[0]); - w6[1] = hc_swap32_S (w6[1]); - w6[2] = hc_swap32_S (w6[2]); - w6[3] = hc_swap32_S (w6[3]); - w7[0] = hc_swap32_S (w7[0]); - w7[1] = hc_swap32_S (w7[1]); - w7[2] = hc_swap32_S (w7[2]); - w7[3] = hc_swap32_S (w7[3]); - - sha512_update_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7, (len - pos1) * 2); + sha512_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha512_final (sha512_ctx_t *ctx) @@ -1772,121 +1412,22 @@ DECLSPEC void sha512_hmac_init_global_swap (sha512_hmac_ctx_t *ctx, GLOBAL_AS co sha512_hmac_init_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7); } -DECLSPEC void sha512_hmac_init_global_utf16le_swap (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) +DECLSPEC void sha512_hmac_init_global_ut16le (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - u32 w4[4]; - u32 w5[4]; - u32 w6[4]; - u32 w7[4]; - - const int len_new = len * 2; - - if (len_new > 128) - { - sha512_ctx_t tmp; - - sha512_init (&tmp); + u32 w_utf16_buf[64] = { 0 }; - sha512_update_global_utf16le_swap (&tmp, w, len); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - sha512_final (&tmp); - - w0[0] = h32_from_64_S (tmp.h[0]); - w0[1] = l32_from_64_S (tmp.h[0]); - w0[2] = h32_from_64_S (tmp.h[1]); - w0[3] = l32_from_64_S (tmp.h[1]); - w1[0] = h32_from_64_S (tmp.h[2]); - w1[1] = l32_from_64_S (tmp.h[2]); - w1[2] = h32_from_64_S (tmp.h[3]); - w1[3] = l32_from_64_S (tmp.h[3]); - w2[0] = h32_from_64_S (tmp.h[4]); - w2[1] = l32_from_64_S (tmp.h[4]); - w2[2] = h32_from_64_S (tmp.h[5]); - w2[3] = l32_from_64_S (tmp.h[5]); - w3[0] = h32_from_64_S (tmp.h[6]); - w3[1] = l32_from_64_S (tmp.h[6]); - w3[2] = h32_from_64_S (tmp.h[7]); - w3[3] = l32_from_64_S (tmp.h[7]); - w4[0] = 0; - w4[1] = 0; - w4[2] = 0; - w4[3] = 0; - w5[0] = 0; - w5[1] = 0; - w5[2] = 0; - w5[3] = 0; - w6[0] = 0; - w6[1] = 0; - w6[2] = 0; - w6[3] = 0; - w7[0] = 0; - w7[1] = 0; - w7[2] = 0; - w7[3] = 0; - } - else - { - w0[0] = w[ 0]; - w0[1] = w[ 1]; - w0[2] = w[ 2]; - w0[3] = w[ 3]; - w1[0] = w[ 4]; - w1[1] = w[ 5]; - w1[2] = w[ 6]; - w1[3] = w[ 7]; - w2[0] = w[ 8]; - w2[1] = w[ 9]; - w2[2] = w[10]; - w2[3] = w[11]; - w3[0] = w[12]; - w3[1] = w[13]; - w3[2] = w[14]; - w3[3] = w[15]; + sha512_hmac_init (ctx, w_utf16_buf, w_utf16_len); +} - make_utf16le_S (w3, w6, w7); - make_utf16le_S (w2, w4, w5); - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); +DECLSPEC void sha512_hmac_init_global_utf16le_swap (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) +{ + u32 w_utf16_buf[64] = { 0 }; - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - w4[0] = hc_swap32_S (w4[0]); - w4[1] = hc_swap32_S (w4[1]); - w4[2] = hc_swap32_S (w4[2]); - w4[3] = hc_swap32_S (w4[3]); - w5[0] = hc_swap32_S (w5[0]); - w5[1] = hc_swap32_S (w5[1]); - w5[2] = hc_swap32_S (w5[2]); - w5[3] = hc_swap32_S (w5[3]); - w6[0] = hc_swap32_S (w6[0]); - w6[1] = hc_swap32_S (w6[1]); - w6[2] = hc_swap32_S (w6[2]); - w6[3] = hc_swap32_S (w6[3]); - w7[0] = hc_swap32_S (w7[0]); - w7[1] = hc_swap32_S (w7[1]); - w7[2] = hc_swap32_S (w7[2]); - w7[3] = hc_swap32_S (w7[3]); - } + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - sha512_hmac_init_128 (ctx, w0, w1, w2, w3, w4, w5, w6, w7); + sha512_hmac_init_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void sha512_hmac_update_128 (sha512_hmac_ctx_t *ctx, u32 *w0, u32 *w1, u32 *w2, u32 *w3, u32 *w4, u32 *w5, u32 *w6, u32 *w7, const int len) @@ -1904,16 +1445,6 @@ DECLSPEC void sha512_hmac_update_swap (sha512_hmac_ctx_t *ctx, const u32 *w, con sha512_update_swap (&ctx->ipad, w, len); } -DECLSPEC void sha512_hmac_update_utf16le (sha512_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - sha512_update_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void sha512_hmac_update_utf16le_swap (sha512_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - sha512_update_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void sha512_hmac_update_global (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { sha512_update_global (&ctx->ipad, w, len); @@ -1924,16 +1455,6 @@ DECLSPEC void sha512_hmac_update_global_swap (sha512_hmac_ctx_t *ctx, GLOBAL_AS sha512_update_global_swap (&ctx->ipad, w, len); } -DECLSPEC void sha512_hmac_update_global_utf16le (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - sha512_update_global_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void sha512_hmac_update_global_utf16le_swap (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - sha512_update_global_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void sha512_hmac_final (sha512_hmac_ctx_t *ctx) { sha512_final (&ctx->ipad); diff --git a/OpenCL/inc_hash_sha512.h b/OpenCL/inc_hash_sha512.h index c66aa1fb9..7009b7d4a 100644 --- a/OpenCL/inc_hash_sha512.h +++ b/OpenCL/inc_hash_sha512.h @@ -120,16 +120,13 @@ DECLSPEC void sha512_hmac_init (sha512_hmac_ctx_t *ctx, const u32 *w, const int DECLSPEC void sha512_hmac_init_swap (sha512_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha512_hmac_init_global (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha512_hmac_init_global_swap (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); +DECLSPEC void sha512_hmac_init_global_ut16le (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha512_hmac_init_global_utf16le_swap (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha512_hmac_update_128 (sha512_hmac_ctx_t *ctx, u32 *w0, u32 *w1, u32 *w2, u32 *w3, u32 *w4, u32 *w5, u32 *w6, u32 *w7, const int len); DECLSPEC void sha512_hmac_update (sha512_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha512_hmac_update_swap (sha512_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void sha512_hmac_update_utf16le (sha512_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void sha512_hmac_update_utf16le_swap (sha512_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void sha512_hmac_update_global (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha512_hmac_update_global_swap (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void sha512_hmac_update_global_utf16le (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void sha512_hmac_update_global_utf16le_swap (sha512_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void sha512_hmac_final (sha512_hmac_ctx_t *ctx); DECLSPEC void sha512_transform_vector (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, const u32x *w4, const u32x *w5, const u32x *w6, const u32x *w7, u64x *digest); DECLSPEC void sha512_init_vector (sha512_ctx_vector_t *ctx); diff --git a/OpenCL/inc_hash_whirlpool.cl b/OpenCL/inc_hash_whirlpool.cl index 2835fc72d..5b30615f7 100644 --- a/OpenCL/inc_hash_whirlpool.cl +++ b/OpenCL/inc_hash_whirlpool.cl @@ -1018,120 +1018,20 @@ DECLSPEC void whirlpool_update_swap (whirlpool_ctx_t *ctx, const u32 *w, const i DECLSPEC void whirlpool_update_utf16le (whirlpool_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - whirlpool_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - whirlpool_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + whirlpool_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void whirlpool_update_utf16le_swap (whirlpool_ctx_t *ctx, const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - whirlpool_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; + u32 w_utf16_buf[64] = { 0 }; - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le (w, len, 256, w_utf16_buf, 256); - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); - - whirlpool_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + whirlpool_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void whirlpool_update_global (whirlpool_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) @@ -1274,120 +1174,20 @@ DECLSPEC void whirlpool_update_global_swap (whirlpool_ctx_t *ctx, GLOBAL_AS cons DECLSPEC void whirlpool_update_global_utf16le (whirlpool_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; + u32 w_utf16_buf[64] = { 0 }; - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - whirlpool_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - whirlpool_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + whirlpool_update (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void whirlpool_update_global_utf16le_swap (whirlpool_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { - u32 w0[4]; - u32 w1[4]; - u32 w2[4]; - u32 w3[4]; - - int pos1; - int pos4; - - for (pos1 = 0, pos4 = 0; pos1 < len - 32; pos1 += 32, pos4 += 8) - { - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + u32 w_utf16_buf[64] = { 0 }; - whirlpool_update_64 (ctx, w0, w1, w2, w3, 32 * 2); - } - - w0[0] = w[pos4 + 0]; - w0[1] = w[pos4 + 1]; - w0[2] = w[pos4 + 2]; - w0[3] = w[pos4 + 3]; - w1[0] = w[pos4 + 4]; - w1[1] = w[pos4 + 5]; - w1[2] = w[pos4 + 6]; - w1[3] = w[pos4 + 7]; - - make_utf16le_S (w1, w2, w3); - make_utf16le_S (w0, w0, w1); - - w0[0] = hc_swap32_S (w0[0]); - w0[1] = hc_swap32_S (w0[1]); - w0[2] = hc_swap32_S (w0[2]); - w0[3] = hc_swap32_S (w0[3]); - w1[0] = hc_swap32_S (w1[0]); - w1[1] = hc_swap32_S (w1[1]); - w1[2] = hc_swap32_S (w1[2]); - w1[3] = hc_swap32_S (w1[3]); - w2[0] = hc_swap32_S (w2[0]); - w2[1] = hc_swap32_S (w2[1]); - w2[2] = hc_swap32_S (w2[2]); - w2[3] = hc_swap32_S (w2[3]); - w3[0] = hc_swap32_S (w3[0]); - w3[1] = hc_swap32_S (w3[1]); - w3[2] = hc_swap32_S (w3[2]); - w3[3] = hc_swap32_S (w3[3]); + const int w_utf16_len = utf8_to_utf16le_global (w, len, 256, w_utf16_buf, 256); - whirlpool_update_64 (ctx, w0, w1, w2, w3, (len - pos1) * 2); + whirlpool_update_swap (ctx, w_utf16_buf, w_utf16_len); } DECLSPEC void whirlpool_final (whirlpool_ctx_t *ctx) @@ -1723,16 +1523,6 @@ DECLSPEC void whirlpool_hmac_update_swap (whirlpool_hmac_ctx_t *ctx, const u32 * whirlpool_update_swap (&ctx->ipad, w, len); } -DECLSPEC void whirlpool_hmac_update_utf16le (whirlpool_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - whirlpool_update_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void whirlpool_hmac_update_utf16le_swap (whirlpool_hmac_ctx_t *ctx, const u32 *w, const int len) -{ - whirlpool_update_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void whirlpool_hmac_update_global (whirlpool_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) { whirlpool_update_global (&ctx->ipad, w, len); @@ -1743,16 +1533,6 @@ DECLSPEC void whirlpool_hmac_update_global_swap (whirlpool_hmac_ctx_t *ctx, GLOB whirlpool_update_global_swap (&ctx->ipad, w, len); } -DECLSPEC void whirlpool_hmac_update_global_utf16le (whirlpool_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - whirlpool_update_global_utf16le (&ctx->ipad, w, len); -} - -DECLSPEC void whirlpool_hmac_update_global_utf16le_swap (whirlpool_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) -{ - whirlpool_update_global_utf16le_swap (&ctx->ipad, w, len); -} - DECLSPEC void whirlpool_hmac_final (whirlpool_hmac_ctx_t *ctx) { whirlpool_final (&ctx->ipad); diff --git a/OpenCL/inc_hash_whirlpool.h b/OpenCL/inc_hash_whirlpool.h index e13ec9960..b7600feca 100644 --- a/OpenCL/inc_hash_whirlpool.h +++ b/OpenCL/inc_hash_whirlpool.h @@ -104,12 +104,8 @@ DECLSPEC void whirlpool_hmac_init_global_swap (whirlpool_hmac_ctx_t *ctx, GLOBAL DECLSPEC void whirlpool_hmac_update_64 (whirlpool_hmac_ctx_t *ctx, u32 *w0, u32 *w1, u32 *w2, u32 *w3, const int len); DECLSPEC void whirlpool_hmac_update (whirlpool_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void whirlpool_hmac_update_swap (whirlpool_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void whirlpool_hmac_update_utf16le (whirlpool_hmac_ctx_t *ctx, const u32 *w, const int len); -DECLSPEC void whirlpool_hmac_update_utf16le_swap (whirlpool_hmac_ctx_t *ctx, const u32 *w, const int len); DECLSPEC void whirlpool_hmac_update_global (whirlpool_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void whirlpool_hmac_update_global_swap (whirlpool_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void whirlpool_hmac_update_global_utf16le (whirlpool_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); -DECLSPEC void whirlpool_hmac_update_global_utf16le_swap (whirlpool_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len); DECLSPEC void whirlpool_hmac_final (whirlpool_hmac_ctx_t *ctx); DECLSPEC void whirlpool_transform_vector (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, u32x *digest, SHM_TYPE u64 *s_MT0, SHM_TYPE u64 *s_MT1, SHM_TYPE u64 *s_MT2, SHM_TYPE u64 *s_MT3, SHM_TYPE u64 *s_MT4, SHM_TYPE u64 *s_MT5, SHM_TYPE u64 *s_MT6, SHM_TYPE u64 *s_MT7); DECLSPEC void whirlpool_init_vector (whirlpool_ctx_vector_t *ctx, SHM_TYPE u64 *s_MT0, SHM_TYPE u64 *s_MT1, SHM_TYPE u64 *s_MT2, SHM_TYPE u64 *s_MT3, SHM_TYPE u64 *s_MT4, SHM_TYPE u64 *s_MT5, SHM_TYPE u64 *s_MT6, SHM_TYPE u64 *s_MT7); diff --git a/OpenCL/m02100-pure.cl b/OpenCL/m02100-pure.cl index 9242bf6e6..e9a3bd04e 100644 --- a/OpenCL/m02100-pure.cl +++ b/OpenCL/m02100-pure.cl @@ -28,6 +28,11 @@ typedef struct dcc2_tmp } dcc2_tmp_t; +DECLSPEC void sha1_hmac_update_global_utf16le_swap (sha1_hmac_ctx_t *ctx, GLOBAL_AS const u32 *w, const int len) +{ + sha1_update_global_utf16le_swap (&ctx->ipad, w, len); +} + DECLSPEC void hmac_sha1_run_V (u32x *w0, u32x *w1, u32x *w2, u32x *w3, u32x *ipad, u32x *opad, u32x *digest) { digest[0] = ipad[0]; diff --git a/docs/changes.txt b/docs/changes.txt index c2837ba1d..ccd6f2ee2 100644 --- a/docs/changes.txt +++ b/docs/changes.txt @@ -28,6 +28,7 @@ ## Features ## +- Added support for true UTF8 to UTF16 conversion in kernel crypto library - Added option --hash-info to show generic information for each hash-mode - Removed option --example-hashes, now is an alias of --hash-info @@ -63,7 +64,8 @@ - Hash-Mode 11600 (7-Zip): Improved memory handling (alloc and free) for the hook function - Hash-Mode 13200 (AxCrypt): Changed the name to AxCrypt 1 to avoid confusion - Hash-Mode 13300 (AxCrypt in-memory SHA1): Changed the name to AxCrypt 1 in-memory SHA1 -- OpenCL Runtime: Switched default OpenCL device type on macOS from GPU to CPU. Use -D 2 to enable GPU devices. +- Kernel Crypto Library: Removed unnecessary utf16 conversion functions which would apply on HMAC data portion +- OpenCL Runtime: Switched default OpenCL device type on macOS from GPU to CPU. Use -D 2 to enable GPU devices - Unit tests: Added Python 3 support for all of the Python code in our test framework - Unit tests: Fixed the packaging of test (-p) feature