In other words, it can be said that the difference in efficiency between 'copy_user_enhanced_fast_string', which accounts for most of the processing by normal system calls, and '__memmove_avx ...