native.intel.cfb.cfb128.c Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of bcprov-lts8on Show documentation
Show all versions of bcprov-lts8on Show documentation
The Long Term Stable (LTS) Bouncy Castle Crypto package is a Java implementation of cryptographic algorithms. This jar contains the JCA/JCE provider and low-level API for the BC LTS version 2.73.7 for Java 8 and later.
//
//
//
#include "cfb.h"
static inline void aes128w_cfb128_decrypt(
__m128i *d0, __m128i *d1, __m128i *d2, __m128i *d3,
__m128i *d4, __m128i *d5, __m128i *d6, __m128i *d7,
__m128i *feedback, __m128i *roundKeys, const uint32_t blocks,
const uint32_t num_rounds) {
__m128i tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
if (blocks == 8) {
tmp0 = _mm_xor_si128(*feedback, roundKeys[0]);
tmp1 = _mm_xor_si128(*d0, roundKeys[0]);
tmp2 = _mm_xor_si128(*d1, roundKeys[0]);
tmp3 = _mm_xor_si128(*d2, roundKeys[0]);
tmp4 = _mm_xor_si128(*d3, roundKeys[0]);
tmp5 = _mm_xor_si128(*d4, roundKeys[0]);
tmp6 = _mm_xor_si128(*d5, roundKeys[0]);
tmp7 = _mm_xor_si128(*d6, roundKeys[0]);
*feedback = *d7;
int rounds;
for (rounds = 1; rounds < num_rounds; rounds++) {
tmp0 = _mm_aesenc_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenc_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenc_si128(tmp2, roundKeys[rounds]);
tmp3 = _mm_aesenc_si128(tmp3, roundKeys[rounds]);
tmp4 = _mm_aesenc_si128(tmp4, roundKeys[rounds]);
tmp5 = _mm_aesenc_si128(tmp5, roundKeys[rounds]);
tmp6 = _mm_aesenc_si128(tmp6, roundKeys[rounds]);
tmp7 = _mm_aesenc_si128(tmp7, roundKeys[rounds]);
}
tmp0 = _mm_aesenclast_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenclast_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenclast_si128(tmp2, roundKeys[rounds]);
tmp3 = _mm_aesenclast_si128(tmp3, roundKeys[rounds]);
tmp4 = _mm_aesenclast_si128(tmp4, roundKeys[rounds]);
tmp5 = _mm_aesenclast_si128(tmp5, roundKeys[rounds]);
tmp6 = _mm_aesenclast_si128(tmp6, roundKeys[rounds]);
tmp7 = _mm_aesenclast_si128(tmp7, roundKeys[rounds]);
*d0 = _mm_xor_si128(*d0, tmp0);
*d1 = _mm_xor_si128(*d1, tmp1);
*d2 = _mm_xor_si128(*d2, tmp2);
*d3 = _mm_xor_si128(*d3, tmp3);
*d4 = _mm_xor_si128(*d4, tmp4);
*d5 = _mm_xor_si128(*d5, tmp5);
*d6 = _mm_xor_si128(*d6, tmp6);
*d7 = _mm_xor_si128(*d7, tmp7);
} else if (blocks == 7) {
tmp0 = _mm_xor_si128(*feedback, roundKeys[0]);
tmp1 = _mm_xor_si128(*d0, roundKeys[0]);
tmp2 = _mm_xor_si128(*d1, roundKeys[0]);
tmp3 = _mm_xor_si128(*d2, roundKeys[0]);
tmp4 = _mm_xor_si128(*d3, roundKeys[0]);
tmp5 = _mm_xor_si128(*d4, roundKeys[0]);
tmp6 = _mm_xor_si128(*d5, roundKeys[0]);
*feedback = *d6;
int rounds;
for (rounds = 1; rounds < num_rounds; rounds++) {
tmp0 = _mm_aesenc_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenc_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenc_si128(tmp2, roundKeys[rounds]);
tmp3 = _mm_aesenc_si128(tmp3, roundKeys[rounds]);
tmp4 = _mm_aesenc_si128(tmp4, roundKeys[rounds]);
tmp5 = _mm_aesenc_si128(tmp5, roundKeys[rounds]);
tmp6 = _mm_aesenc_si128(tmp6, roundKeys[rounds]);
}
tmp0 = _mm_aesenclast_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenclast_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenclast_si128(tmp2, roundKeys[rounds]);
tmp3 = _mm_aesenclast_si128(tmp3, roundKeys[rounds]);
tmp4 = _mm_aesenclast_si128(tmp4, roundKeys[rounds]);
tmp5 = _mm_aesenclast_si128(tmp5, roundKeys[rounds]);
tmp6 = _mm_aesenclast_si128(tmp6, roundKeys[rounds]);
*d0 = _mm_xor_si128(*d0, tmp0);
*d1 = _mm_xor_si128(*d1, tmp1);
*d2 = _mm_xor_si128(*d2, tmp2);
*d3 = _mm_xor_si128(*d3, tmp3);
*d4 = _mm_xor_si128(*d4, tmp4);
*d5 = _mm_xor_si128(*d5, tmp5);
*d6 = _mm_xor_si128(*d6, tmp6);
} else if (blocks == 6) {
tmp0 = _mm_xor_si128(*feedback, roundKeys[0]);
tmp1 = _mm_xor_si128(*d0, roundKeys[0]);
tmp2 = _mm_xor_si128(*d1, roundKeys[0]);
tmp3 = _mm_xor_si128(*d2, roundKeys[0]);
tmp4 = _mm_xor_si128(*d3, roundKeys[0]);
tmp5 = _mm_xor_si128(*d4, roundKeys[0]);
*feedback = *d5;
int rounds;
for (rounds = 1; rounds < num_rounds; rounds++) {
tmp0 = _mm_aesenc_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenc_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenc_si128(tmp2, roundKeys[rounds]);
tmp3 = _mm_aesenc_si128(tmp3, roundKeys[rounds]);
tmp4 = _mm_aesenc_si128(tmp4, roundKeys[rounds]);
tmp5 = _mm_aesenc_si128(tmp5, roundKeys[rounds]);
}
tmp0 = _mm_aesenclast_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenclast_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenclast_si128(tmp2, roundKeys[rounds]);
tmp3 = _mm_aesenclast_si128(tmp3, roundKeys[rounds]);
tmp4 = _mm_aesenclast_si128(tmp4, roundKeys[rounds]);
tmp5 = _mm_aesenclast_si128(tmp5, roundKeys[rounds]);
*d0 = _mm_xor_si128(*d0, tmp0);
*d1 = _mm_xor_si128(*d1, tmp1);
*d2 = _mm_xor_si128(*d2, tmp2);
*d3 = _mm_xor_si128(*d3, tmp3);
*d4 = _mm_xor_si128(*d4, tmp4);
*d5 = _mm_xor_si128(*d5, tmp5);
} else if (blocks == 5) {
tmp0 = _mm_xor_si128(*feedback, roundKeys[0]);
tmp1 = _mm_xor_si128(*d0, roundKeys[0]);
tmp2 = _mm_xor_si128(*d1, roundKeys[0]);
tmp3 = _mm_xor_si128(*d2, roundKeys[0]);
tmp4 = _mm_xor_si128(*d3, roundKeys[0]);
*feedback = *d4;
int rounds;
for (rounds = 1; rounds < num_rounds; rounds++) {
tmp0 = _mm_aesenc_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenc_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenc_si128(tmp2, roundKeys[rounds]);
tmp3 = _mm_aesenc_si128(tmp3, roundKeys[rounds]);
tmp4 = _mm_aesenc_si128(tmp4, roundKeys[rounds]);
}
tmp0 = _mm_aesenclast_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenclast_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenclast_si128(tmp2, roundKeys[rounds]);
tmp3 = _mm_aesenclast_si128(tmp3, roundKeys[rounds]);
tmp4 = _mm_aesenclast_si128(tmp4, roundKeys[rounds]);
*d0 = _mm_xor_si128(*d0, tmp0);
*d1 = _mm_xor_si128(*d1, tmp1);
*d2 = _mm_xor_si128(*d2, tmp2);
*d3 = _mm_xor_si128(*d3, tmp3);
*d4 = _mm_xor_si128(*d4, tmp4);
} else if (blocks == 4) {
tmp0 = _mm_xor_si128(*feedback, roundKeys[0]);
tmp1 = _mm_xor_si128(*d0, roundKeys[0]);
tmp2 = _mm_xor_si128(*d1, roundKeys[0]);
tmp3 = _mm_xor_si128(*d2, roundKeys[0]);
*feedback = *d3;
int rounds;
for (rounds = 1; rounds < num_rounds; rounds++) {
tmp0 = _mm_aesenc_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenc_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenc_si128(tmp2, roundKeys[rounds]);
tmp3 = _mm_aesenc_si128(tmp3, roundKeys[rounds]);
}
tmp0 = _mm_aesenclast_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenclast_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenclast_si128(tmp2, roundKeys[rounds]);
tmp3 = _mm_aesenclast_si128(tmp3, roundKeys[rounds]);
*d0 = _mm_xor_si128(*d0, tmp0);
*d1 = _mm_xor_si128(*d1, tmp1);
*d2 = _mm_xor_si128(*d2, tmp2);
*d3 = _mm_xor_si128(*d3, tmp3);
} else if (blocks == 3) {
tmp0 = _mm_xor_si128(*feedback, roundKeys[0]);
tmp1 = _mm_xor_si128(*d0, roundKeys[0]);
tmp2 = _mm_xor_si128(*d1, roundKeys[0]);
*feedback = *d2;
int rounds;
for (rounds = 1; rounds < num_rounds; rounds++) {
tmp0 = _mm_aesenc_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenc_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenc_si128(tmp2, roundKeys[rounds]);
}
tmp0 = _mm_aesenclast_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenclast_si128(tmp1, roundKeys[rounds]);
tmp2 = _mm_aesenclast_si128(tmp2, roundKeys[rounds]);
*d0 = _mm_xor_si128(*d0, tmp0);
*d1 = _mm_xor_si128(*d1, tmp1);
*d2 = _mm_xor_si128(*d2, tmp2);
} else if (blocks == 2) {
tmp0 = _mm_xor_si128(*feedback, roundKeys[0]);
tmp1 = _mm_xor_si128(*d0, roundKeys[0]);
*feedback = *d1;
int rounds;
for (rounds = 1; rounds < num_rounds; rounds++) {
tmp0 = _mm_aesenc_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenc_si128(tmp1, roundKeys[rounds]);
}
tmp0 = _mm_aesenclast_si128(tmp0, roundKeys[rounds]);
tmp1 = _mm_aesenclast_si128(tmp1, roundKeys[rounds]);
*d0 = _mm_xor_si128(*d0, tmp0);
*d1 = _mm_xor_si128(*d1, tmp1);
} else if (blocks == 1) {
tmp0 = _mm_xor_si128(*feedback, roundKeys[0]);
*feedback = *d0;
int rounds;
for (rounds = 1; rounds < num_rounds; rounds++) {
tmp0 = _mm_aesenc_si128(tmp0, roundKeys[rounds]);
}
tmp0 = _mm_aesenclast_si128(tmp0, roundKeys[rounds]);
*d0 = _mm_xor_si128(*d0, tmp0);
}
}
size_t cfb_decrypt(cfb_ctx *cfb, unsigned char *src, size_t len, unsigned char *dest) {
unsigned char *destStart = dest;
//
// Round out buffer.
//
while (cfb->buf_index > 0 && len > 0) {
*dest = cfb_decrypt_byte(cfb, *src);
len--;
dest++;
src++;
}
while (len >= CFB_BLOCK_SIZE * 8) {
__m128i d0 = _mm_loadu_si128((__m128i *) &src[0 * 16]);
__m128i d1 = _mm_loadu_si128((__m128i *) &src[1 * 16]);
__m128i d2 = _mm_loadu_si128((__m128i *) &src[2 * 16]);
__m128i d3 = _mm_loadu_si128((__m128i *) &src[3 * 16]);
__m128i d4 = _mm_loadu_si128((__m128i *) &src[4 * 16]);
__m128i d5 = _mm_loadu_si128((__m128i *) &src[5 * 16]);
__m128i d6 = _mm_loadu_si128((__m128i *) &src[6 * 16]);
__m128i d7 = _mm_loadu_si128((__m128i *) &src[7 * 16]);
aes128w_cfb128_decrypt(&d0, &d1, &d2, &d3, &d4, &d5, &d6, &d7, &cfb->feedback, cfb->roundKeys, 8,
cfb->num_rounds);
_mm_storeu_si128((__m128i *) &dest[0 * 16], d0);
_mm_storeu_si128((__m128i *) &dest[1 * 16], d1);
_mm_storeu_si128((__m128i *) &dest[2 * 16], d2);
_mm_storeu_si128((__m128i *) &dest[3 * 16], d3);
_mm_storeu_si128((__m128i *) &dest[4 * 16], d4);
_mm_storeu_si128((__m128i *) &dest[5 * 16], d5);
_mm_storeu_si128((__m128i *) &dest[6 * 16], d6);
_mm_storeu_si128((__m128i *) &dest[7 * 16], d7);
len -= 16 * 8;
src += 16 * 8;
dest += 16 * 8;
}
//
// Process as many whole blocks as possible.
//
while (len >= CFB_BLOCK_SIZE) {
if (len >= CFB_BLOCK_SIZE * 7) {
__m128i d0 = _mm_loadu_si128((__m128i *) &src[0 * 16]);
__m128i d1 = _mm_loadu_si128((__m128i *) &src[1 * 16]);
__m128i d2 = _mm_loadu_si128((__m128i *) &src[2 * 16]);
__m128i d3 = _mm_loadu_si128((__m128i *) &src[3 * 16]);
__m128i d4 = _mm_loadu_si128((__m128i *) &src[4 * 16]);
__m128i d5 = _mm_loadu_si128((__m128i *) &src[5 * 16]);
__m128i d6 = _mm_loadu_si128((__m128i *) &src[6 * 16]);
aes128w_cfb128_decrypt(&d0, &d1, &d2, &d3, &d4, &d5, &d6, &d6, &cfb->feedback, cfb->roundKeys, 7,
cfb->num_rounds);
_mm_storeu_si128((__m128i *) &dest[0 * 16], d0);
_mm_storeu_si128((__m128i *) &dest[1 * 16], d1);
_mm_storeu_si128((__m128i *) &dest[2 * 16], d2);
_mm_storeu_si128((__m128i *) &dest[3 * 16], d3);
_mm_storeu_si128((__m128i *) &dest[4 * 16], d4);
_mm_storeu_si128((__m128i *) &dest[5 * 16], d5);
_mm_storeu_si128((__m128i *) &dest[6 * 16], d6);
len -= 16 * 7;
src += 16 * 7;
dest += 16 * 7;
} else if (len >= CFB_BLOCK_SIZE * 6) {
__m128i d0 = _mm_loadu_si128((__m128i *) &src[0 * 16]);
__m128i d1 = _mm_loadu_si128((__m128i *) &src[1 * 16]);
__m128i d2 = _mm_loadu_si128((__m128i *) &src[2 * 16]);
__m128i d3 = _mm_loadu_si128((__m128i *) &src[3 * 16]);
__m128i d4 = _mm_loadu_si128((__m128i *) &src[4 * 16]);
__m128i d5 = _mm_loadu_si128((__m128i *) &src[5 * 16]);
aes128w_cfb128_decrypt(&d0, &d1, &d2, &d3, &d4, &d5, &d5, &d5, &cfb->feedback, cfb->roundKeys, 6,
cfb->num_rounds);
_mm_storeu_si128((__m128i *) &dest[0 * 16], d0);
_mm_storeu_si128((__m128i *) &dest[1 * 16], d1);
_mm_storeu_si128((__m128i *) &dest[2 * 16], d2);
_mm_storeu_si128((__m128i *) &dest[3 * 16], d3);
_mm_storeu_si128((__m128i *) &dest[4 * 16], d4);
_mm_storeu_si128((__m128i *) &dest[5 * 16], d5);
len -= 16 * 6;
src += 16 * 6;
dest += 16 * 6;
} else if (len >= CFB_BLOCK_SIZE * 5) {
__m128i d0 = _mm_loadu_si128((__m128i *) &src[0 * 16]);
__m128i d1 = _mm_loadu_si128((__m128i *) &src[1 * 16]);
__m128i d2 = _mm_loadu_si128((__m128i *) &src[2 * 16]);
__m128i d3 = _mm_loadu_si128((__m128i *) &src[3 * 16]);
__m128i d4 = _mm_loadu_si128((__m128i *) &src[4 * 16]);
aes128w_cfb128_decrypt(&d0, &d1, &d2, &d3, &d4, &d4, &d4, &d4, &cfb->feedback, cfb->roundKeys, 5,
cfb->num_rounds);
_mm_storeu_si128((__m128i *) &dest[0 * 16], d0);
_mm_storeu_si128((__m128i *) &dest[1 * 16], d1);
_mm_storeu_si128((__m128i *) &dest[2 * 16], d2);
_mm_storeu_si128((__m128i *) &dest[3 * 16], d3);
_mm_storeu_si128((__m128i *) &dest[4 * 16], d4);
len -= 16 * 5;
src += 16 * 5;
dest += 16 * 5;
} else if (len >= CFB_BLOCK_SIZE * 4) {
__m128i d0 = _mm_loadu_si128((__m128i *) &src[0 * 16]);
__m128i d1 = _mm_loadu_si128((__m128i *) &src[1 * 16]);
__m128i d2 = _mm_loadu_si128((__m128i *) &src[2 * 16]);
__m128i d3 = _mm_loadu_si128((__m128i *) &src[3 * 16]);
aes128w_cfb128_decrypt(&d0, &d1, &d2, &d3, &d3, &d3, &d3, &d3, &cfb->feedback, cfb->roundKeys, 4,
cfb->num_rounds);
_mm_storeu_si128((__m128i *) &dest[0 * 16], d0);
_mm_storeu_si128((__m128i *) &dest[1 * 16], d1);
_mm_storeu_si128((__m128i *) &dest[2 * 16], d2);
_mm_storeu_si128((__m128i *) &dest[3 * 16], d3);
len -= 16 * 4;
src += 16 * 4;
dest += 16 * 4;
} else if (len >= CFB_BLOCK_SIZE * 3) {
__m128i d0 = _mm_loadu_si128((__m128i *) &src[0 * 16]);
__m128i d1 = _mm_loadu_si128((__m128i *) &src[1 * 16]);
__m128i d2 = _mm_loadu_si128((__m128i *) &src[2 * 16]);
aes128w_cfb128_decrypt(&d0, &d1, &d2, &d2, &d2, &d2, &d2, &d2, &cfb->feedback, cfb->roundKeys, 3,
cfb->num_rounds);
_mm_storeu_si128((__m128i *) &dest[0 * 16], d0);
_mm_storeu_si128((__m128i *) &dest[1 * 16], d1);
_mm_storeu_si128((__m128i *) &dest[2 * 16], d2);
len -= 16 * 3;
src += 16 * 3;
dest += 16 * 3;
} else if (len >= CFB_BLOCK_SIZE * 2) {
__m128i d0 = _mm_loadu_si128((__m128i *) &src[0 * 16]);
__m128i d1 = _mm_loadu_si128((__m128i *) &src[1 * 16]);
aes128w_cfb128_decrypt(&d0, &d1, &d1, &d1, &d1, &d1, &d1, &d1, &cfb->feedback, cfb->roundKeys, 2,
cfb->num_rounds);
_mm_storeu_si128((__m128i *) &dest[0 * 16], d0);
_mm_storeu_si128((__m128i *) &dest[1 * 16], d1);
len -= 16 * 2;
src += 16 * 2;
dest += 16 * 2;
} else {
__m128i d0 = _mm_loadu_si128((__m128i *) &src[0 * 16]);
aes128w_cfb128_decrypt(&d0, &d0, &d0, &d0, &d0, &d0, &d0, &d0, &cfb->feedback, cfb->roundKeys, 1,
cfb->num_rounds);
_mm_storeu_si128((__m128i *) &dest[0 * 16], d0);
len -= 16 * 1;
src += 16 * 1;
dest += 16 * 1;
}
}
//
// load any trailing bytes into the buffer, the expectation is that
// whatever is passed in has to be decrypted, ideally callers will
// try and stick to the AES block size for as long as possible.
//
while (len > 0) {
*dest = cfb_decrypt_byte(cfb, *src);
len--;
dest++;
src++;
}
return (size_t) (dest - destStart);
}
unsigned char cfb_decrypt_byte(cfb_ctx *cfb, unsigned char b) {
if (cfb->buf_index == 0) {
// We need to generate a new encrypted feedback block to xor into the data.,
cfb->mask = _mm_xor_si128(cfb->feedback, cfb->roundKeys[0]);
int j;
for (j = 1; j < cfb->num_rounds; j++) {
cfb->mask = _mm_aesenc_si128(cfb->mask, cfb->roundKeys[j]);
}
cfb->mask = _mm_aesenclast_si128(cfb->mask, cfb->roundKeys[j]);
}
//
// incrementally mask becomes the last block of plain text
//
unsigned char pt = ((unsigned char *) &cfb->mask)[cfb->buf_index] ^ b;
((unsigned char *) &cfb->mask)[cfb->buf_index++] = b; // Mask fills with last cipher text directly.
if (cfb->buf_index == CFB_BLOCK_SIZE) {
cfb->buf_index = 0;
cfb->feedback = cfb->mask;
}
return pt;
}
© 2015 - 2024 Weber Informatics LLC | Privacy Policy