_mm256_slli_si256 only shifts each 128 bit lane individually

[nerai]

SmithWaterman/src/compare_avx.cpp

Line 139 in 4321d8f

global_max = _mm256_max_epu8 (global_max, _mm256_slli_si256 (global_max, 16));

_mm256_slli_si256 only shifts each 128 bit lane individually