[2/5] AArch64: Improve A64FX memset for large sizes

Improve performance of large memsets. Simplify alignment code. For zero memset
use DC ZVA, which almost doubles performance. For non-zero memsets use the
unroll8 loop which is about 10% faster.

Reviewed-by: Naohiro Tamura <naohirot@fujitsu.com>