popov/cpuminer-opt-gpu
1
0
Fork 0
mirror of https://github.com/JayDDee/cpuminer-opt.git synced 2025-09-17 23:44:27 +00:00
Code Issues Packages Projects Releases Wiki Activity

v24.2

Browse Source
This commit is contained in:
Jay D Dee
2024-05-20 23:08:50 -04:00
parent 4f930574cc
commit 042d13d1e1
129 changed files with 835 additions and 538 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
algo/scrypt/scrypt-core-4way.c
View File

@@ -745,7 +745,7 @@ do{ \
SALSA_2ROUNDS; SALSA_2ROUNDS; SALSA_2ROUNDS; SALSA_2ROUNDS;
#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
#if defined(SIMD512)
// Tested OK but very slow
// 16 way parallel, requires 16x32 interleaving
@@ -2487,7 +2487,7 @@ static void salsa8_simd128_2buf( uint32_t * const ba, uint32_t * const bb,
XA3 = BA[3] = v128_xor( BA[3], CA[3] );
XB3 = BB[3] = v128_xor( BB[3], CB[3] );
#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
#if defined(SIMD512)
SALSA_8ROUNDS_SIMD128_2BUF;
@@ -2886,7 +2886,7 @@ static void salsa8_simd128_3buf( uint32_t *ba, uint32_t *bb, uint32_t *bc,
XB3 = BB[3] = v128_xor( BB[3], CB[3] );
XC3 = BC[3] = v128_xor( BC[3], CC[3] );
#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
#if defined(SIMD512)
SALSA_8ROUNDS_SIMD128_3BUF;