use std::simd::u32x4;
#[inline]
pub fn simd_process_blocks_sha256(data: &[u8]) -> Vec<u8> {
let mut result = Vec::with_capacity(32);
let padded_len = if data.is_empty() {
64
} else {
(data.len() + 9).div_ceil(64) * 64
};
let mut processed = Vec::with_capacity(padded_len);
if !data.is_empty() {
processed.extend_from_slice(data);
}
while processed.len() < padded_len - 8 {
processed.push(0);
}
let msg_len_bits = (data.len() as u64) * 8;
processed.extend_from_slice(&msg_len_bits.to_be_bytes());
let mut offset = 0;
while offset + 64 <= processed.len() {
let chunk = &processed[offset..offset + 64];
let hash = simd_sha256_block(chunk);
result.extend_from_slice(&hash);
offset += 64;
}
result
}
#[inline]
fn simd_sha256_block(block: &[u8]) -> [u8; 32] {
let mut w = [0u32; 64];
for i in 0..16 {
let bytes = &block[i * 4..i * 4 + 4];
w[i] = u32::from_be_bytes([bytes[0], bytes[1], bytes[2], bytes[3]]);
}
for i in 16..64 {
let s0 = w[i - 15].rotate_right(7) ^ w[i - 15].rotate_right(18) ^ w[i - 15] >> 3;
let s1 = w[i - 2].rotate_right(17) ^ w[i - 2].rotate_right(19) ^ w[i - 2] >> 10;
w[i] = w[i - 16]
.wrapping_add(s0)
.wrapping_add(w[i - 7])
.wrapping_add(s1);
}
let k: [u32; 64] = [
0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5, 0x3956c25b, 0x59f111f1, 0x923f82a4,
0xab1c5ed5, 0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3, 0x72be5d74, 0x80deb1fe,
0x9bdc06a7, 0xc19bf174, 0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc, 0x2de92c6f,
0x4a7484aa, 0x5cb0a9dc, 0x76f988da, 0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7,
0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967, 0x27b70a85, 0x2e1b2138, 0x4d2c6dfc,
0x53380d13, 0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85, 0xa2bfe8a1, 0xa81a664b,
0xc24b8b70, 0xc76c51a3, 0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070, 0x19a4c116,
0x1e376c08, 0x2748774c, 0x34b0bcb5, 0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3,
0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208, 0x90befffa, 0xa4506ceb, 0xbef9a3f7,
0xc67178f2,
];
let h_init: [u32; 8] = [
0x6a09e667, 0xbb67ae85, 0x3c6ef372, 0xa54ff53a, 0x510e527f, 0x9b05688c, 0x1f83d9ab,
0x5be0cd19,
];
let mut a = h_init[0];
let mut b = h_init[1];
let mut c = h_init[2];
let mut d = h_init[3];
let mut e = h_init[4];
let mut f = h_init[5];
let mut g = h_init[6];
let mut h = h_init[7];
for i in 0..64 {
let s0 = a.rotate_right(2) ^ a.rotate_right(13) ^ a.rotate_right(22);
let maj = (a & b) ^ (a & c) ^ (b & c);
let t2 = s0.wrapping_add(maj);
let s1 = e.rotate_right(6) ^ e.rotate_right(11) ^ e.rotate_right(25);
let ch = (e & f) ^ (!e & g);
let t1 = h
.wrapping_add(s1)
.wrapping_add(ch)
.wrapping_add(k[i])
.wrapping_add(w[i]);
h = g;
g = f;
f = e;
e = d.wrapping_add(t1);
d = c;
c = b;
b = a;
a = t1.wrapping_add(t2);
}
let mut result = [0u8; 32];
let h_final: [u32; 8] = [
h_init[0].wrapping_add(a),
h_init[1].wrapping_add(b),
h_init[2].wrapping_add(c),
h_init[3].wrapping_add(d),
h_init[4].wrapping_add(e),
h_init[5].wrapping_add(f),
h_init[6].wrapping_add(g),
h_init[7].wrapping_add(h),
];
for (i, &val) in h_final.iter().enumerate() {
result[i * 4..i * 4 + 4].copy_from_slice(&val.to_be_bytes());
}
result
}
#[inline]
pub fn simd_sha256_finalize(hash: &[u8], data: &[u8]) -> Vec<u8> {
let mut combined = Vec::with_capacity(hash.len() + data.len());
combined.extend_from_slice(hash);
combined.extend_from_slice(data);
simd_process_blocks_sha256(&combined)
}
#[inline]
pub fn simd_combine_hashes(hashes: &[&[u8]]) -> Vec<u8> {
let mut result = Vec::with_capacity(hashes.len() * 32);
for &hash in hashes {
result.extend_from_slice(hash);
}
simd_process_blocks_sha256(&result)
}
#[inline]
pub fn simd_sha256_block_vectorized(block: &[u8]) -> [u8; 32] {
let k_vals: [u32x4; 16] = [
u32x4::from_array([0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5]),
u32x4::from_array([0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5]),
u32x4::from_array([0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3]),
u32x4::from_array([0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174]),
u32x4::from_array([0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc]),
u32x4::from_array([0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da]),
u32x4::from_array([0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7]),
u32x4::from_array([0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967]),
u32x4::from_array([0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13]),
u32x4::from_array([0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85]),
u32x4::from_array([0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3]),
u32x4::from_array([0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070]),
u32x4::from_array([0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5]),
u32x4::from_array([0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3]),
u32x4::from_array([0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208]),
u32x4::from_array([0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2]),
];
let mut w = [0u32; 64];
for i in 0..16 {
let bytes = &block[i * 4..i * 4 + 4];
w[i] = u32::from_be_bytes([bytes[0], bytes[1], bytes[2], bytes[3]]);
}
for i in 16..64 {
let s0 = w[i - 15].rotate_right(7) ^ w[i - 15].rotate_right(18) ^ w[i - 15] >> 3;
let s1 = w[i - 2].rotate_right(17) ^ w[i - 2].rotate_right(19) ^ w[i - 2] >> 10;
w[i] = w[i - 16]
.wrapping_add(s0)
.wrapping_add(w[i - 7])
.wrapping_add(s1);
}
let mut a = 0x6a09e667u32;
let mut b = 0xbb67ae85u32;
let mut c = 0x3c6ef372u32;
let mut d = 0xa54ff53au32;
let mut e = 0x510e527fu32;
let mut f = 0x9b05688cu32;
let mut g = 0x1f83d9abu32;
let mut h = 0x5be0cd19u32;
for (i, &wi) in w.iter().enumerate() {
let idx = i / 4;
let k = k_vals[idx];
let offset = i % 4;
let s0 = a.rotate_right(2) ^ a.rotate_right(13) ^ a.rotate_right(22);
let maj = (a & b) ^ (a & c) ^ (b & c);
let t2 = s0.wrapping_add(maj);
let s1 = e.rotate_right(6) ^ e.rotate_right(11) ^ e.rotate_right(25);
let ch = (e & f) ^ (!e & g);
let t1 = h
.wrapping_add(s1)
.wrapping_add(ch)
.wrapping_add(k[offset])
.wrapping_add(wi);
h = g;
g = f;
f = e;
e = d.wrapping_add(t1);
d = c;
c = b;
b = a;
a = t1.wrapping_add(t2);
}
let h_init: [u32; 8] = [
0x6a09e667, 0xbb67ae85, 0x3c6ef372, 0xa54ff53a, 0x510e527f, 0x9b05688c, 0x1f83d9ab,
0x5be0cd19,
];
let mut result = [0u8; 32];
let h_final: [u32; 8] = [
h_init[0].wrapping_add(a),
h_init[1].wrapping_add(b),
h_init[2].wrapping_add(c),
h_init[3].wrapping_add(d),
h_init[4].wrapping_add(e),
h_init[5].wrapping_add(f),
h_init[6].wrapping_add(g),
h_init[7].wrapping_add(h),
];
for (i, &val) in h_final.iter().enumerate() {
result[i * 4..i * 4 + 4].copy_from_slice(&val.to_be_bytes());
}
result
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_sha256_basic() {
let data = b"hello world";
let result = simd_process_blocks_sha256(data);
assert_eq!(result.len(), 32);
}
#[test]
fn test_sha256_empty() {
let result = simd_process_blocks_sha256(b"");
assert_eq!(result.len(), 32);
}
#[test]
fn test_sha256_finalize() {
let hash = simd_process_blocks_sha256(b"initial");
let result = simd_sha256_finalize(&hash, b" more data");
assert_eq!(result.len(), 32);
}
#[test]
fn test_combine_hashes() {
let hash1 = simd_process_blocks_sha256(b"data1");
let hash2 = simd_process_blocks_sha256(b"data2");
let result = simd_combine_hashes(&[&hash1, &hash2]);
assert_eq!(result.len(), 64);
}
#[test]
fn test_sha256_vectorized() {
let data = b"hello world";
let padded_len = (data.len() + 9).div_ceil(64) * 64;
let mut processed = Vec::with_capacity(padded_len);
processed.extend_from_slice(data);
while processed.len() < padded_len - 8 {
processed.push(0);
}
let msg_len_bits = (data.len() as u64) * 8;
processed.extend_from_slice(&msg_len_bits.to_be_bytes());
let result = simd_sha256_block_vectorized(&processed);
assert_eq!(result.len(), 32);
let expected = simd_process_blocks_sha256(data);
assert_eq!(result.to_vec(), expected);
}
}