arne
/
cryptopp
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Cleared -Wcast-align (Issue 122). No asserts added because X86/X32/X64 uses unaligned accesses with the compressed tables

pull/136/head
Jeffrey Walton 2016-01-25 19:30:35 -05:00
parent 1f1fecce88
commit c76114705c
1 changed files with 14 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
rijndael.cpp
View File

@ -130,8 +130,8 @@ static volatile bool s_TeFilled = false, s_TdFilled = false;
#define QUARTER_ROUND_FE(t, a, b, c, d) QUARTER_ROUND(TL_F, Te, t, d, c, b, a) #define QUARTER_ROUND_FE(t, a, b, c, d) QUARTER_ROUND(TL_F, Te, t, d, c, b, a)
#define QUARTER_ROUND_FD(t, a, b, c, d) QUARTER_ROUND(TL_F, Td, t, d, c, b, a) #define QUARTER_ROUND_FD(t, a, b, c, d) QUARTER_ROUND(TL_F, Td, t, d, c, b, a)
#if defined(CRYPTOPP_ALLOW_UNALIGNED_DATA_ACCESS) || defined(CRYPTOPP_ALLOW_RIJNDAEL_UNALIGNED_DATA_ACCESS) #if defined(CRYPTOPP_ALLOW_UNALIGNED_DATA_ACCESS) || defined(CRYPTOPP_ALLOW_RIJNDAEL_UNALIGNED_DATA_ACCESS)
#define TL_F(T, i, x) (*(word32 *)((byte *)T + x*8 + (6-i)%4+1)) #define TL_F(T, i, x) (*(word32 *)(void *)((byte *)T + x*8 + (6-i)%4+1))
#define TL_M(T, i, x) (*(word32 *)((byte *)T + x*8 + (i+3)%4+1)) #define TL_M(T, i, x) (*(word32 *)(void *)((byte *)T + x*8 + (i+3)%4+1))
#else #else
#define TL_F(T, i, x) rotrFixed(T[x], (3-i)*8) #define TL_F(T, i, x) rotrFixed(T[x], (3-i)*8)
#define TL_M(T, i, x) T[i*256 + x] #define TL_M(T, i, x) T[i*256 + x]
@ -140,7 +140,7 @@ static volatile bool s_TeFilled = false, s_TdFilled = false;
#define QUARTER_ROUND_FE(t, a, b, c, d) QUARTER_ROUND(TL_F, Te, t, a, b, c, d) #define QUARTER_ROUND_FE(t, a, b, c, d) QUARTER_ROUND(TL_F, Te, t, a, b, c, d)
#define QUARTER_ROUND_FD(t, a, b, c, d) QUARTER_ROUND(TL_F, Td, t, a, b, c, d) #define QUARTER_ROUND_FD(t, a, b, c, d) QUARTER_ROUND(TL_F, Td, t, a, b, c, d)
#if defined(CRYPTOPP_ALLOW_UNALIGNED_DATA_ACCESS) || defined(CRYPTOPP_ALLOW_RIJNDAEL_UNALIGNED_DATA_ACCESS) #if defined(CRYPTOPP_ALLOW_UNALIGNED_DATA_ACCESS) || defined(CRYPTOPP_ALLOW_RIJNDAEL_UNALIGNED_DATA_ACCESS)
#define TL_F(T, i, x) (*(word32 *)((byte *)T + x*8 + (4-i)%4)) #define TL_F(T, i, x) (*(word32 *)(void *)((byte *)T + x*8 + (4-i)%4))
#define TL_M TL_F #define TL_M TL_F
#else #else
#define TL_F(T, i, x) rotrFixed(T[x], i*8) #define TL_F(T, i, x) rotrFixed(T[x], i*8)
@ -261,16 +261,16 @@ void Rijndael::Base::UncheckedSetKey(const byte *userKey, unsigned int keylen, c
rk = m_key; rk = m_key;
unsigned int i, j; unsigned int i, j;
std::swap(*(__m128i *)(rk), *(__m128i *)(rk+4*m_rounds)); std::swap(*(__m128i *)(void *)(rk), *(__m128i *)(void *)(rk+4*m_rounds));
for (i = 4, j = 4*m_rounds-4; i < j; i += 4, j -= 4) for (i = 4, j = 4*m_rounds-4; i < j; i += 4, j -= 4)
{ {
temp = _mm_aesimc_si128(*(__m128i *)(rk+i)); temp = _mm_aesimc_si128(*(__m128i *)(void *)(rk+i));
*(__m128i *)(rk+i) = _mm_aesimc_si128(*(__m128i *)(rk+j)); *(__m128i *)(void *)(rk+i) = _mm_aesimc_si128(*(__m128i *)(void *)(rk+j));
*(__m128i *)(rk+j) = temp; *(__m128i *)(void *)(rk+j) = temp;
} }
*(__m128i *)(rk+i) = _mm_aesimc_si128(*(__m128i *)(rk+i)); *(__m128i *)(void *)(rk+i) = _mm_aesimc_si128(*(__m128i *)(void *)(rk+i));
} }
return; return;
@ -391,7 +391,7 @@ void Rijndael::Enc::ProcessAndXorBlock(const byte *inBlock, const byte *xorBlock
#else #else
for (i=0; i<1024; i+=cacheLineSize) for (i=0; i<1024; i+=cacheLineSize)
#endif #endif
u &= *(const word32 *)(((const byte *)Te)+i); u &= *(const word32 *)(void *)(((const byte *)Te)+i);
u &= Te[255]; u &= Te[255];
s0 |= u; s1 |= u; s2 |= u; s3 |= u; s0 |= u; s1 |= u; s2 |= u; s3 |= u;
@ -467,7 +467,7 @@ void Rijndael::Dec::ProcessAndXorBlock(const byte *inBlock, const byte *xorBlock
#else #else
for (i=0; i<1024; i+=cacheLineSize) for (i=0; i<1024; i+=cacheLineSize)
#endif #endif
u &= *(const word32 *)(((const byte *)Td)+i); u &= *(const word32 *)(void *)(((const byte *)Td)+i);
u &= Td[255]; u &= Td[255];
s0 |= u; s1 |= u; s2 |= u; s3 |= u; s0 |= u; s1 |= u; s2 |= u; s3 |= u;
@ -503,8 +503,8 @@ void Rijndael::Dec::ProcessAndXorBlock(const byte *inBlock, const byte *xorBlock
// QUARTER_ROUND_LD will use Td, which is already preloaded. // QUARTER_ROUND_LD will use Td, which is already preloaded.
u = 0; u = 0;
for (i=0; i<256; i+=cacheLineSize) for (i=0; i<256; i+=cacheLineSize)
u &= *(const word32 *)(Sd+i); u &= *(const word32 *)(void *)(Sd+i);
u &= *(const word32 *)(Sd+252); u &= *(const word32 *)(void *)(Sd+252);
t0 |= u; t1 |= u; t2 |= u; t3 |= u; t0 |= u; t1 |= u; t2 |= u; t3 |= u;
#endif #endif
@ -1121,7 +1121,7 @@ inline size_t AESNI_AdvancedProcessBlocks(F1 func1, F4 func4, const __m128i *sub
__m128i block0 = _mm_loadu_si128((const __m128i *)inBlocks), block1, block2, block3; __m128i block0 = _mm_loadu_si128((const __m128i *)inBlocks), block1, block2, block3;
if (flags & BlockTransformation::BT_InBlockIsCounter) if (flags & BlockTransformation::BT_InBlockIsCounter)
{ {
const __m128i be1 = *(const __m128i *)s_one; const __m128i be1 = *(const __m128i *)(void *)s_one;
block1 = _mm_add_epi32(block0, be1); block1 = _mm_add_epi32(block0, be1);
block2 = _mm_add_epi32(block1, be1); block2 = _mm_add_epi32(block1, be1);
block3 = _mm_add_epi32(block2, be1); block3 = _mm_add_epi32(block2, be1);
@ -1251,7 +1251,7 @@ size_t Rijndael::Enc::AdvancedProcessBlocks(const byte *inBlocks, const byte *xo
increment = 0-increment; increment = 0-increment;
} }
Locals &locals = *(Locals *)space; Locals &locals = *(Locals *)(void *)space;
locals.inBlocks = inBlocks; locals.inBlocks = inBlocks;
locals.inXorBlocks = (flags & BT_XorInput) && xorBlocks ? xorBlocks : zeros; locals.inXorBlocks = (flags & BT_XorInput) && xorBlocks ? xorBlocks : zeros;