Skip to content

Commit

Permalink
Fix self test NEON compile with MSVC compiler (GH #776)
Browse files Browse the repository at this point in the history
  • Loading branch information
@noloader
noloader committed Jan 4, 2019
1 parent a20c56e commit 9255740
Show file tree
Hide file tree
Showing 3 changed files with 32 additions and 34 deletions.
60 changes: 29 additions & 31 deletions adv_simd.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -513,13 +513,11 @@ inline size_t AdvancedProcessBlocks128_6x1_NEON(F1 func1, F6 func6,
/// \tparam F1 function to process 1 128-bit block
/// \tparam F4 function to process 4 128-bit blocks
/// \tparam W word type of the subkey table
/// \tparam V vector type of the NEON datatype
/// \details AdvancedProcessBlocks128_4x1_NEON processes 4 and 1 NEON SIMD words
/// at a time.
/// \details The subkey type is usually word32 or word64. V is the vector type and it is
/// usually uint32x4_t or uint64x2_t. F1, F4, W and V must use the same word and
/// vector type. The V parameter is used to avoid template argument
/// deduction/substitution failures.
/// usually uint32x4_t or uint32x4_t. F1, F4, and W must use the same word and
/// vector type.
template <typename F1, typename F4, typename W>
inline size_t AdvancedProcessBlocks128_4x1_NEON(F1 func1, F4 func4,
const W *subKeys, size_t rounds, const byte *inBlocks,
Expand Down Expand Up @@ -564,62 +562,62 @@ inline size_t AdvancedProcessBlocks128_4x1_NEON(F1 func1, F4 func4,
{
while (length >= 4*blockSize)
{
uint64x2_t block0, block1, block2, block3;
uint32x4_t block0, block1, block2, block3;
if (flags & BT_InBlockIsCounter)
{
const uint64x2_t one = vreinterpretq_u64_u32(s_one);
block0 = vreinterpretq_u64_u8(vld1q_u8(inBlocks));
block1 = vaddq_u64(block0, one);
block2 = vaddq_u64(block1, one);
block3 = vaddq_u64(block2, one);
const uint32x4_t one = s_one;
block0 = vreinterpretq_u32_u8(vld1q_u8(inBlocks));
block1 = vaddq_u32(block0, one);
block2 = vaddq_u32(block1, one);
block3 = vaddq_u32(block2, one);
vst1q_u8(const_cast<byte*>(inBlocks),
vreinterpretq_u8_u64(vaddq_u64(block3, one)));
vreinterpretq_u8_u32(vaddq_u32(block3, one)));
}
else
{
block0 = vreinterpretq_u64_u8(vld1q_u8(inBlocks));
block0 = vreinterpretq_u32_u8(vld1q_u8(inBlocks));
inBlocks = PtrAdd(inBlocks, inIncrement);
block1 = vreinterpretq_u64_u8(vld1q_u8(inBlocks));
block1 = vreinterpretq_u32_u8(vld1q_u8(inBlocks));
inBlocks = PtrAdd(inBlocks, inIncrement);
block2 = vreinterpretq_u64_u8(vld1q_u8(inBlocks));
block2 = vreinterpretq_u32_u8(vld1q_u8(inBlocks));
inBlocks = PtrAdd(inBlocks, inIncrement);
block3 = vreinterpretq_u64_u8(vld1q_u8(inBlocks));
block3 = vreinterpretq_u32_u8(vld1q_u8(inBlocks));
inBlocks = PtrAdd(inBlocks, inIncrement);
}

if (xorInput)
{
block0 = veorq_u64(block0, vreinterpretq_u64_u8(vld1q_u8(xorBlocks)));
block0 = veorq_u32(block0, vreinterpretq_u32_u8(vld1q_u8(xorBlocks)));
xorBlocks = PtrAdd(xorBlocks, xorIncrement);
block1 = veorq_u64(block1, vreinterpretq_u64_u8(vld1q_u8(xorBlocks)));
block1 = veorq_u32(block1, vreinterpretq_u32_u8(vld1q_u8(xorBlocks)));
xorBlocks = PtrAdd(xorBlocks, xorIncrement);
block2 = veorq_u64(block2, vreinterpretq_u64_u8(vld1q_u8(xorBlocks)));
block2 = veorq_u32(block2, vreinterpretq_u32_u8(vld1q_u8(xorBlocks)));
xorBlocks = PtrAdd(xorBlocks, xorIncrement);
block3 = veorq_u64(block3, vreinterpretq_u64_u8(vld1q_u8(xorBlocks)));
block3 = veorq_u32(block3, vreinterpretq_u32_u8(vld1q_u8(xorBlocks)));
xorBlocks = PtrAdd(xorBlocks, xorIncrement);
}

func4(block0, block1, block2, block3, subKeys, static_cast<unsigned int>(rounds));

if (xorOutput)
{
block0 = veorq_u64(block0, vreinterpretq_u64_u8(vld1q_u8(xorBlocks)));
block0 = veorq_u32(block0, vreinterpretq_u32_u8(vld1q_u8(xorBlocks)));
xorBlocks = PtrAdd(xorBlocks, xorIncrement);
block1 = veorq_u64(block1, vreinterpretq_u64_u8(vld1q_u8(xorBlocks)));
block1 = veorq_u32(block1, vreinterpretq_u32_u8(vld1q_u8(xorBlocks)));
xorBlocks = PtrAdd(xorBlocks, xorIncrement);
block2 = veorq_u64(block2, vreinterpretq_u64_u8(vld1q_u8(xorBlocks)));
block2 = veorq_u32(block2, vreinterpretq_u32_u8(vld1q_u8(xorBlocks)));
xorBlocks = PtrAdd(xorBlocks, xorIncrement);
block3 = veorq_u64(block3, vreinterpretq_u64_u8(vld1q_u8(xorBlocks)));
block3 = veorq_u32(block3, vreinterpretq_u32_u8(vld1q_u8(xorBlocks)));
xorBlocks = PtrAdd(xorBlocks, xorIncrement);
}

vst1q_u8(outBlocks, vreinterpretq_u8_u64(block0));
vst1q_u8(outBlocks, vreinterpretq_u8_u32(block0));
outBlocks = PtrAdd(outBlocks, outIncrement);
vst1q_u8(outBlocks, vreinterpretq_u8_u64(block1));
vst1q_u8(outBlocks, vreinterpretq_u8_u32(block1));
outBlocks = PtrAdd(outBlocks, outIncrement);
vst1q_u8(outBlocks, vreinterpretq_u8_u64(block2));
vst1q_u8(outBlocks, vreinterpretq_u8_u32(block2));
outBlocks = PtrAdd(outBlocks, outIncrement);
vst1q_u8(outBlocks, vreinterpretq_u8_u64(block3));
vst1q_u8(outBlocks, vreinterpretq_u8_u32(block3));
outBlocks = PtrAdd(outBlocks, outIncrement);

length -= 4*blockSize;
Expand All @@ -628,20 +626,20 @@ inline size_t AdvancedProcessBlocks128_4x1_NEON(F1 func1, F4 func4,

while (length >= blockSize)
{
uint64x2_t block = vreinterpretq_u64_u8(vld1q_u8(inBlocks));
uint32x4_t block = vreinterpretq_u32_u8(vld1q_u8(inBlocks));

if (xorInput)
block = veorq_u64(block, vreinterpretq_u64_u8(vld1q_u8(xorBlocks)));
block = veorq_u32(block, vreinterpretq_u32_u8(vld1q_u8(xorBlocks)));

if (flags & BT_InBlockIsCounter)
const_cast<byte *>(inBlocks)[15]++;

func1(block, subKeys, static_cast<unsigned int>(rounds));

if (xorOutput)
block = veorq_u64(block, vreinterpretq_u64_u8(vld1q_u8(xorBlocks)));
block = veorq_u32(block, vreinterpretq_u32_u8(vld1q_u8(xorBlocks)));

vst1q_u8(outBlocks, vreinterpretq_u8_u64(block));
vst1q_u8(outBlocks, vreinterpretq_u8_u32(block));

inBlocks = PtrAdd(inBlocks, inIncrement);
outBlocks = PtrAdd(outBlocks, outIncrement);
Expand Down
2 changes: 1 addition & 1 deletion naclite.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -53,7 +53,7 @@
#include "config.h"
#include "stdcpp.h"

#if defined(NO_OS_DEPENDENCE)
#if defined(NO_OS_DEPENDENCE) || !defined(OS_RNG_AVAILABLE)
# define CRYPTOPP_DISABLE_NACL 1
#endif

Expand Down
4 changes: 2 additions & 2 deletions validat7.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -404,7 +404,7 @@ bool TestX25519()
int ret2 = Donna::curve25519_mult(share2, priv2, pub1);
int ret3 = std::memcmp(share1, share2, 32);

#if defined(NO_OS_DEPENDENCE)
#if defined(CRYPTOPP_DISABLE_NACL)
int ret4=0, ret5=0, ret6=0;
#else
// Bernstein's NaCl requires DefaultAutoSeededRNG.
Expand Down Expand Up @@ -436,7 +436,7 @@ bool TestEd25519()
std::cout << "\nTesting ed25519 Signatures...\n\n";
bool pass = true;

#ifndef NO_OS_DEPENDENCE
#ifndef CRYPTOPP_DISABLE_NACL
const unsigned int SIGN_COUNT = 64, MSG_SIZE=128;
const unsigned int NACL_EXTRA=NaCl::crypto_sign_BYTES;

Expand Down

0 comments on commit 9255740

Please sign in to comment.