Skip to content
Snippets Groups Projects
Commit 274a9630 authored by Michael Kuron's avatar Michael Kuron :mortar_board:
Browse files

Philox AVX512: fix shuffle mask

parent f8b730e4
No related branches found
No related tags found
No related merge requests found
...@@ -436,10 +436,10 @@ QUALIFIERS void _philox4x32round(__m512i* ctr, __m512i* key) ...@@ -436,10 +436,10 @@ QUALIFIERS void _philox4x32round(__m512i* ctr, __m512i* key)
__m512i lohi1a = _mm512_mul_epu32(ctr[2], _mm512_set1_epi32(PHILOX_M4x32_1)); __m512i lohi1a = _mm512_mul_epu32(ctr[2], _mm512_set1_epi32(PHILOX_M4x32_1));
__m512i lohi1b = _mm512_mul_epu32(_mm512_srli_epi64(ctr[2], 32), _mm512_set1_epi32(PHILOX_M4x32_1)); __m512i lohi1b = _mm512_mul_epu32(_mm512_srli_epi64(ctr[2], 32), _mm512_set1_epi32(PHILOX_M4x32_1));
lohi0a = _mm512_shuffle_epi32(lohi0a, 0xD8); lohi0a = _mm512_shuffle_epi32(lohi0a, _MM_PERM_DBCA);
lohi0b = _mm512_shuffle_epi32(lohi0b, 0xD8); lohi0b = _mm512_shuffle_epi32(lohi0b, _MM_PERM_DBCA);
lohi1a = _mm512_shuffle_epi32(lohi1a, 0xD8); lohi1a = _mm512_shuffle_epi32(lohi1a, _MM_PERM_DBCA);
lohi1b = _mm512_shuffle_epi32(lohi1b, 0xD8); lohi1b = _mm512_shuffle_epi32(lohi1b, _MM_PERM_DBCA);
__m512i lo0 = _mm512_unpacklo_epi32(lohi0a, lohi0b); __m512i lo0 = _mm512_unpacklo_epi32(lohi0a, lohi0b);
__m512i hi0 = _mm512_unpackhi_epi32(lohi0a, lohi0b); __m512i hi0 = _mm512_unpackhi_epi32(lohi0a, lohi0b);
......
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment