mirror of
https://github.com/Atmosphere-NX/Atmosphere.git
synced 2024-11-24 21:02:03 +00:00
225 lines
7.3 KiB
C++
225 lines
7.3 KiB
C++
/*
|
|
* Copyright (c) Atmosphère-NX
|
|
*
|
|
* This program is free software; you can redistribute it and/or modify it
|
|
* under the terms and conditions of the GNU General Public License,
|
|
* version 2, as published by the Free Software Foundation.
|
|
*
|
|
* This program is distributed in the hope it will be useful, but WITHOUT
|
|
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
|
|
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
|
|
* more details.
|
|
*
|
|
* You should have received a copy of the GNU General Public License
|
|
* along with this program. If not, see <http://www.gnu.org/licenses/>.
|
|
*/
|
|
#include <vapours.hpp>
|
|
|
|
namespace ams::crypto::impl {
|
|
|
|
namespace {
|
|
|
|
constexpr const u32 RoundConstants[4] = {
|
|
0x5A827999, 0x6ED9EBA1, 0x8F1BBCDC, 0xCA62C1D6
|
|
};
|
|
|
|
constexpr ALWAYS_INLINE u32 Choose(u32 x, u32 y, u32 z) {
|
|
return (x & y) ^ ((~x) & z);
|
|
}
|
|
|
|
constexpr ALWAYS_INLINE u32 Majority(u32 x, u32 y, u32 z) {
|
|
return (x & y) ^ (x & z) ^ (y & z);
|
|
}
|
|
|
|
constexpr ALWAYS_INLINE u32 Parity(u32 x, u32 y, u32 z) {
|
|
return x ^ y ^ z;
|
|
}
|
|
|
|
}
|
|
|
|
void Sha1Impl::Initialize() {
|
|
/* Reset buffered bytes/bits. */
|
|
m_buffered_bytes = 0;
|
|
m_bits_consumed = 0;
|
|
|
|
/* Set intermediate hash. */
|
|
m_intermediate_hash[0] = 0x67452301;
|
|
m_intermediate_hash[1] = 0xEFCDAB89;
|
|
m_intermediate_hash[2] = 0x98BADCFE;
|
|
m_intermediate_hash[3] = 0x10325476;
|
|
m_intermediate_hash[4] = 0xC3D2E1F0;
|
|
|
|
/* Set state. */
|
|
m_state = State_Initialized;
|
|
}
|
|
|
|
void Sha1Impl::Update(const void *data, size_t size) {
|
|
/* Verify we're in a state to update. */
|
|
AMS_ASSERT(m_state == State_Initialized);
|
|
|
|
/* Advance our input bit count. */
|
|
m_bits_consumed += BITSIZEOF(u8) * (((m_buffered_bytes + size) / BlockSize) * BlockSize);
|
|
|
|
/* Process anything we have buffered. */
|
|
const u8 *data8 = static_cast<const u8 *>(data);
|
|
size_t remaining = size;
|
|
|
|
if (m_buffered_bytes > 0) {
|
|
const size_t copy_size = std::min(BlockSize - m_buffered_bytes, remaining);
|
|
std::memcpy(m_buffer + m_buffered_bytes, data8, copy_size);
|
|
|
|
data8 += copy_size;
|
|
remaining -= copy_size;
|
|
m_buffered_bytes += copy_size;
|
|
|
|
/* Process a block, if we filled one. */
|
|
if (m_buffered_bytes == BlockSize) {
|
|
this->ProcessBlock(m_buffer);
|
|
m_buffered_bytes = 0;
|
|
}
|
|
}
|
|
|
|
/* Process blocks, while we have any. */
|
|
while (remaining >= BlockSize) {
|
|
this->ProcessBlock(data8);
|
|
data8 += BlockSize;
|
|
remaining -= BlockSize;
|
|
}
|
|
|
|
/* Copy any leftover data to our buffer. */
|
|
if (remaining > 0) {
|
|
m_buffered_bytes = remaining;
|
|
std::memcpy(m_buffer, data8, remaining);
|
|
}
|
|
}
|
|
|
|
void Sha1Impl::GetHash(void *dst, size_t size) {
|
|
/* Verify we're in a state to get hash. */
|
|
AMS_ASSERT(m_state == State_Initialized || m_state == State_Done);
|
|
AMS_ASSERT(size >= HashSize);
|
|
AMS_UNUSED(size);
|
|
|
|
/* If we need to, process the last block. */
|
|
if (m_state == State_Initialized) {
|
|
this->ProcessLastBlock();
|
|
m_state = State_Done;
|
|
}
|
|
|
|
/* Copy the output hash. */
|
|
if constexpr (util::IsLittleEndian()) {
|
|
static_assert(HashSize % sizeof(u32) == 0);
|
|
|
|
u32 *dst_32 = static_cast<u32 *>(dst);
|
|
for (size_t i = 0; i < HashSize / sizeof(u32); ++i) {
|
|
dst_32[i] = util::LoadBigEndian<u32>(m_intermediate_hash + i);
|
|
}
|
|
} else {
|
|
std::memcpy(dst, m_intermediate_hash, HashSize);
|
|
}
|
|
}
|
|
|
|
void Sha1Impl::ProcessBlock(const void *data) {
|
|
/* Load work variables. */
|
|
u32 a = m_intermediate_hash[0];
|
|
u32 b = m_intermediate_hash[1];
|
|
u32 c = m_intermediate_hash[2];
|
|
u32 d = m_intermediate_hash[3];
|
|
u32 e = m_intermediate_hash[4];
|
|
u32 tmp;
|
|
size_t i;
|
|
|
|
/* Copy the input. */
|
|
u32 w[80];
|
|
if constexpr (util::IsLittleEndian()) {
|
|
static_assert(BlockSize % sizeof(u32) == 0);
|
|
|
|
const u32 *src_32 = static_cast<const u32 *>(data);
|
|
for (size_t i = 0; i < BlockSize / sizeof(u32); ++i) {
|
|
w[i] = util::LoadBigEndian<u32>(src_32 + i);
|
|
}
|
|
} else {
|
|
std::memcpy(w, data, BlockSize);
|
|
}
|
|
|
|
/* Initialize the rest of w. */
|
|
for (i = BlockSize / sizeof(u32); i < util::size(w); ++i) {
|
|
const u32 *prev = w + (i - BlockSize / sizeof(u32));
|
|
w[i] = util::RotateLeft<u32>(prev[0] ^ prev[2] ^ prev[8] ^ prev[13], 1);
|
|
}
|
|
|
|
/* Perform rounds. */
|
|
for (i = 0; i < 20; ++i) {
|
|
tmp = util::RotateLeft<u32>(a, 5) + Choose(b, c, d) + e + w[i] + RoundConstants[0];
|
|
e = d;
|
|
d = c;
|
|
c = util::RotateLeft<u32>(b, 30);
|
|
b = a;
|
|
a = tmp;
|
|
}
|
|
|
|
for (/* ... */; i < 40; ++i) {
|
|
tmp = util::RotateLeft<u32>(a, 5) + Parity(b, c, d) + e + w[i] + RoundConstants[1];
|
|
e = d;
|
|
d = c;
|
|
c = util::RotateLeft<u32>(b, 30);
|
|
b = a;
|
|
a = tmp;
|
|
}
|
|
|
|
for (/* ... */; i < 60; ++i) {
|
|
tmp = util::RotateLeft<u32>(a, 5) + Majority(b, c, d) + e + w[i] + RoundConstants[2];
|
|
e = d;
|
|
d = c;
|
|
c = util::RotateLeft<u32>(b, 30);
|
|
b = a;
|
|
a = tmp;
|
|
}
|
|
|
|
for (/* ... */; i < 80; ++i) {
|
|
tmp = util::RotateLeft<u32>(a, 5) + Parity(b, c, d) + e + w[i] + RoundConstants[3];
|
|
e = d;
|
|
d = c;
|
|
c = util::RotateLeft<u32>(b, 30);
|
|
b = a;
|
|
a = tmp;
|
|
}
|
|
|
|
/* Update intermediate hash. */
|
|
m_intermediate_hash[0] += a;
|
|
m_intermediate_hash[1] += b;
|
|
m_intermediate_hash[2] += c;
|
|
m_intermediate_hash[3] += d;
|
|
m_intermediate_hash[4] += e;
|
|
}
|
|
|
|
void Sha1Impl::ProcessLastBlock() {
|
|
/* Setup the final block. */
|
|
constexpr const auto BlockSizeWithoutSizeField = BlockSize - sizeof(u64);
|
|
|
|
/* Increment our bits consumed. */
|
|
m_bits_consumed += BITSIZEOF(u8) * m_buffered_bytes;
|
|
|
|
/* Add 0x80 terminator. */
|
|
m_buffer[m_buffered_bytes++] = 0x80;
|
|
|
|
/* If we can process the size field directly, do so, otherwise set up to process it. */
|
|
if (m_buffered_bytes <= BlockSizeWithoutSizeField) {
|
|
/* Clear up to size field. */
|
|
std::memset(m_buffer + m_buffered_bytes, 0, BlockSizeWithoutSizeField - m_buffered_bytes);
|
|
} else {
|
|
/* Consume full block */
|
|
std::memset(m_buffer + m_buffered_bytes, 0, BlockSize - m_buffered_bytes);
|
|
this->ProcessBlock(m_buffer);
|
|
|
|
/* Clear up to size field. */
|
|
std::memset(m_buffer, 0, BlockSizeWithoutSizeField);
|
|
}
|
|
|
|
/* Store the size field. */
|
|
util::StoreBigEndian<u64>(reinterpret_cast<u64 *>(m_buffer + BlockSizeWithoutSizeField), m_bits_consumed);
|
|
|
|
/* Process the final block. */
|
|
this->ProcessBlock(m_buffer);
|
|
}
|
|
|
|
}
|