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

Switch to VectorLoad overload with one arg

pull/739/head
Jeffrey Walton 2018-11-12 23:38:46 -05:00
parent 7515b4d661
commit 57b2bf5708
No known key found for this signature in database
GPG Key ID: B36AB348921B1838
2 changed files with 16 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
simon64_simd.cpp
View File

@ -587,8 +587,8 @@ inline void SIMON64_Enc_Block(uint32x4_p &block0, uint32x4_p &block1,
const uint32x4_p rk2 = vec_splats(subkeys[i+1]); const uint32x4_p rk2 = vec_splats(subkeys[i+1]);
#else #else
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk1 = VectorLoad(0, subkeys+i); uint32x4_p rk1 = VectorLoad(subkeys+i);
uint32x4_p rk2 = VectorLoad(0, subkeys+i+1); uint32x4_p rk2 = VectorLoad(subkeys+i+1);
rk1 = vec_perm(rk1, rk1, m); rk1 = vec_perm(rk1, rk1, m);
rk2 = vec_perm(rk2, rk2, m); rk2 = vec_perm(rk2, rk2, m);
#endif #endif
@ -602,7 +602,7 @@ inline void SIMON64_Enc_Block(uint32x4_p &block0, uint32x4_p &block1,
const uint32x4_p rk = vec_splats(subkeys[rounds-1]); const uint32x4_p rk = vec_splats(subkeys[rounds-1]);
#else #else
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk = VectorLoad(0, subkeys+rounds-1); uint32x4_p rk = VectorLoad(subkeys+rounds-1);
rk = vec_perm(rk, rk, m); rk = vec_perm(rk, rk, m);
#endif #endif
y1 = VectorXor(VectorXor(y1, SIMON64_f(x1)), rk); y1 = VectorXor(VectorXor(y1, SIMON64_f(x1)), rk);
@ -644,7 +644,7 @@ inline void SIMON64_Dec_Block(uint32x4_p &block0, uint32x4_p &block1,
const uint32x4_p rk = vec_splats(subkeys[rounds-1]); const uint32x4_p rk = vec_splats(subkeys[rounds-1]);
#else #else
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk = VectorLoad(0, subkeys+rounds-1); uint32x4_p rk = VectorLoad(subkeys+rounds-1);
rk = vec_perm(rk, rk, m); rk = vec_perm(rk, rk, m);
#endif #endif
y1 = VectorXor(VectorXor(y1, rk), SIMON64_f(x1)); y1 = VectorXor(VectorXor(y1, rk), SIMON64_f(x1));
@ -658,8 +658,8 @@ inline void SIMON64_Dec_Block(uint32x4_p &block0, uint32x4_p &block1,
const uint32x4_p rk2 = vec_splats(subkeys[i]); const uint32x4_p rk2 = vec_splats(subkeys[i]);
#else #else
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk1 = VectorLoad(0, subkeys+i+1); uint32x4_p rk1 = VectorLoad(subkeys+i+1);
uint32x4_p rk2 = VectorLoad(0, subkeys+i); uint32x4_p rk2 = VectorLoad(subkeys+i);
rk1 = vec_perm(rk1, rk1, m); rk1 = vec_perm(rk1, rk1, m);
rk2 = vec_perm(rk2, rk2, m); rk2 = vec_perm(rk2, rk2, m);
#endif #endif
@ -707,8 +707,8 @@ inline void SIMON64_Enc_6_Blocks(uint32x4_p &block0, uint32x4_p &block1,
const uint32x4_p rk2 = vec_splats(subkeys[i+1]); const uint32x4_p rk2 = vec_splats(subkeys[i+1]);
#else #else
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk1 = VectorLoad(0, subkeys+i); uint32x4_p rk1 = VectorLoad(subkeys+i);
uint32x4_p rk2 = VectorLoad(0, subkeys+i+1); uint32x4_p rk2 = VectorLoad(subkeys+i+1);
rk1 = vec_perm(rk1, rk1, m); rk1 = vec_perm(rk1, rk1, m);
rk2 = vec_perm(rk2, rk2, m); rk2 = vec_perm(rk2, rk2, m);
#endif #endif
@ -727,7 +727,7 @@ inline void SIMON64_Enc_6_Blocks(uint32x4_p &block0, uint32x4_p &block1,
const uint32x4_p rk = vec_splats(subkeys[rounds-1]); const uint32x4_p rk = vec_splats(subkeys[rounds-1]);
#else #else
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk = VectorLoad(0, subkeys+rounds-1); uint32x4_p rk = VectorLoad(subkeys+rounds-1);
rk = vec_perm(rk, rk, m); rk = vec_perm(rk, rk, m);
#endif #endif
y1 = VectorXor(VectorXor(y1, SIMON64_f(x1)), rk); y1 = VectorXor(VectorXor(y1, SIMON64_f(x1)), rk);
@ -781,7 +781,7 @@ inline void SIMON64_Dec_6_Blocks(uint32x4_p &block0, uint32x4_p &block1,
const uint32x4_p rk = vec_splats(subkeys[rounds-1]); const uint32x4_p rk = vec_splats(subkeys[rounds-1]);
#else #else
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk = VectorLoad(0, subkeys+rounds-1); uint32x4_p rk = VectorLoad(subkeys+rounds-1);
rk = vec_perm(rk, rk, m); rk = vec_perm(rk, rk, m);
#endif #endif
y1 = VectorXor(VectorXor(y1, rk), SIMON64_f(x1)); y1 = VectorXor(VectorXor(y1, rk), SIMON64_f(x1));
@ -797,8 +797,8 @@ inline void SIMON64_Dec_6_Blocks(uint32x4_p &block0, uint32x4_p &block1,
const uint32x4_p rk2 = vec_splats(subkeys[i]); const uint32x4_p rk2 = vec_splats(subkeys[i]);
#else #else
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk1 = VectorLoad(0, subkeys+i+1); uint32x4_p rk1 = VectorLoad(subkeys+i+1);
uint32x4_p rk2 = VectorLoad(0, subkeys+i); uint32x4_p rk2 = VectorLoad(subkeys+i);
rk1 = vec_perm(rk1, rk1, m); rk1 = vec_perm(rk1, rk1, m);
rk2 = vec_perm(rk2, rk2, m); rk2 = vec_perm(rk2, rk2, m);
#endif #endif

8
speck64_simd.cpp
View File

@ -526,7 +526,7 @@ void SPECK64_Enc_Block(uint32x4_p &block0, uint32x4_p &block1,
#else #else
// subkeys has extra elements so memory backs the last subkey // subkeys has extra elements so memory backs the last subkey
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk = VectorLoad(0, subkeys+i); uint32x4_p rk = VectorLoad(subkeys+i);
rk = vec_perm(rk, rk, m); rk = vec_perm(rk, rk, m);
#endif #endif
@ -573,7 +573,7 @@ void SPECK64_Dec_Block(uint32x4_p &block0, uint32x4_p &block1,
#else #else
// subkeys has extra elements so memory backs the last subkey // subkeys has extra elements so memory backs the last subkey
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk = VectorLoad(0, subkeys+i); uint32x4_p rk = VectorLoad(subkeys+i);
rk = vec_perm(rk, rk, m); rk = vec_perm(rk, rk, m);
#endif #endif
@ -625,7 +625,7 @@ void SPECK64_Enc_6_Blocks(uint32x4_p &block0, uint32x4_p &block1,
#else #else
// subkeys has extra elements so memory backs the last subkey // subkeys has extra elements so memory backs the last subkey
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk = VectorLoad(0, subkeys+i); uint32x4_p rk = VectorLoad(subkeys+i);
rk = vec_perm(rk, rk, m); rk = vec_perm(rk, rk, m);
#endif #endif
@ -694,7 +694,7 @@ void SPECK64_Dec_6_Blocks(uint32x4_p &block0, uint32x4_p &block1,
#else #else
// subkeys has extra elements so memory backs the last subkey // subkeys has extra elements so memory backs the last subkey
const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3}; const uint8x16_p m = {0,1,2,3, 0,1,2,3, 0,1,2,3, 0,1,2,3};
uint32x4_p rk = VectorLoad(0, subkeys+i); uint32x4_p rk = VectorLoad(subkeys+i);
rk = vec_perm(rk, rk, m); rk = vec_perm(rk, rk, m);
#endif #endif