Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

update f2k #28

Merged
merged 1 commit into from
Mar 21, 2023
Merged

update f2k #28

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
62 changes: 4 additions & 58 deletions emp-zk/extensions/ram-zk/gfmul.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,66 +24,12 @@ inline void mul128_0(__m128i *a, __m128i *b, __m128i *res1, __m128i *res2) {
}
}

inline void reduction0(__m128i tmp3, __m128i tmp6, __m128i* res) {
/*__m128i tmp7, tmp8, tmp9, tmp10, tmp11, tmp12;
__m128i XMMMASK = _mm_setr_epi32(0xffffffff, 0x0, 0x0, 0x0);
tmp7 = _mm_srli_epi32(tmp6, 31);
tmp8 = _mm_srli_epi32(tmp6, 30);
tmp9 = _mm_srli_epi32(tmp6, 25);
tmp7 = _mm_xor_si128(tmp7, tmp8);
tmp7 = _mm_xor_si128(tmp7, tmp9);
tmp8 = _mm_shuffle_epi32(tmp7, 147);

tmp7 = _mm_and_si128(XMMMASK, tmp8);
tmp8 = _mm_andnot_si128(XMMMASK, tmp8);
tmp3 = _mm_xor_si128(tmp3, tmp8);
tmp6 = _mm_xor_si128(tmp6, tmp7);
tmp10 = _mm_slli_epi32(tmp6, 1);
tmp3 = _mm_xor_si128(tmp3, tmp10);
tmp11 = _mm_slli_epi32(tmp6, 2);
tmp3 = _mm_xor_si128(tmp3, tmp11);
tmp12 = _mm_slli_epi32(tmp6, 7);
tmp3 = _mm_xor_si128(tmp3, tmp12);

*res = _mm_xor_si128(tmp3, tmp6);*/
__m128i tmp2, tmp4, tmp5, tmp7, tmp8, tmp9;
tmp7 = _mm_srli_epi32(tmp3, 31);
tmp8 = _mm_srli_epi32(tmp6, 31);
tmp3 = _mm_slli_epi32(tmp3, 1);
tmp6 = _mm_slli_epi32(tmp6, 1);

tmp9 = _mm_srli_si128(tmp7, 12);
tmp8 = _mm_slli_si128(tmp8, 4);
tmp7 = _mm_slli_si128(tmp7, 4);
tmp3 = _mm_or_si128(tmp3, tmp7);
tmp6 = _mm_or_si128(tmp6, tmp8);
tmp6 = _mm_or_si128(tmp6, tmp9);

tmp7 = _mm_slli_epi32(tmp3, 31);
tmp8 = _mm_slli_epi32(tmp3, 30);
tmp9 = _mm_slli_epi32(tmp3, 25);
tmp7 = _mm_xor_si128(tmp7, tmp8);
tmp7 = _mm_xor_si128(tmp7, tmp9);
tmp8 = _mm_srli_si128(tmp7, 4);
tmp7 = _mm_slli_si128(tmp7, 12);
tmp3 = _mm_xor_si128(tmp3, tmp7);

tmp2 = _mm_srli_epi32(tmp3, 1);
tmp4 = _mm_srli_epi32(tmp3, 2);
tmp5 = _mm_srli_epi32(tmp3, 7);
tmp2 = _mm_xor_si128(tmp2, tmp4);
tmp2 = _mm_xor_si128(tmp2, tmp5);
tmp2 = _mm_xor_si128(tmp2, tmp8);
tmp3 = _mm_xor_si128(tmp3, tmp2);
*res = _mm_xor_si128(tmp6, tmp3);
}

template<int N>
inline void gfmul_1(__m128i *a, __m128i *b, __m128i *res) {
__m128i tmp3[N], tmp6[N];
mul128_0<N>(a, b, tmp3, tmp6);
for(int i = 0; i < N; ++i)
reduction0(tmp3[i], tmp6[i], res+i);
res[i] = reduce(tmp3[i], tmp6[i]);
}

/* batch 3 */
Expand Down Expand Up @@ -120,7 +66,7 @@ inline void gfmul3(__m128i &a1, __m128i &a2, __m128i &a3,
__m128i tmp3[3], tmp6[3];
mul128_3(a1, a2, a3, b1, b2, b3, tmp3, tmp6);
for(int i = 0; i < 3; ++i)
reduction0(tmp3[i], tmp6[i], res+i);
res[i] = reduce(tmp3[i], tmp6[i]);
}


Expand Down Expand Up @@ -162,7 +108,7 @@ inline void gfmul4(__m128i &a1, __m128i &a2, __m128i &a3, __m128i &a4,
__m128i tmp3[4], tmp6[4];
mul128_4(a1, a2, a3, a4, b1, b2, b3, b4, tmp3, tmp6);
for(int i = 0; i < 4; ++i)
reduction0(tmp3[i], tmp6[i], res+i);
res[i] = reduce(tmp3[i], tmp6[i]);
}


Expand Down Expand Up @@ -216,7 +162,7 @@ inline void gfmul6(__m128i &a1, __m128i &a2, __m128i &a3,
__m128i tmp3[6], tmp6[6];
mul128_6(a1, a2, a3, a4, a5, a6, b1, b2, b3, b4, b5, b6, tmp3, tmp6);
for(int i = 0; i < 6; ++i)
reduction0(tmp3[i], tmp6[i], res+i);
res[i] = reduce(tmp3[i], tmp6[i]);
}
#endif

Expand Down