[C++] AVX512 part for SIMD operations of DecodeSpaced/EncodeSpaced

[asfimport]

As part of https://issues.apache.org/jira/browse/PARQUET-1841, AVX512 path identified with the helper of mask_compress_/mask_expand_  API.

This Jira created for spaced benchmark, unittest and AVX512 path and other basic support of further potential SIMD chance of SSE/AVX2. 

Reporter: Frank Du / @jianxind
Assignee: Frank Du / @jianxind

PRs and other links:

• GitHub Pull Request #7029

_{Note: This issue was originally created as ARROW-8579. Please see the migration documentation for further details.}

[asfimport]

Frank Du / @jianxind:
No need now. The typical 0.01% is optimized  by #7531 already.