[FEA] support ‘yyyyMMdd’ format in from_unixtime function

[nvliyuan]

It would be nice if we support 'yyyyMMdd' format for from_unixtime function. I know we could avoid the fallback by setting spark.rapids.sql.incompatibleDateFormats.enabled=true, but it would simplify the configs for customers if we support more LEGACY format.
For example:

create a hive table and insert some rows:
CREATE TABLE `xxx_timestamp`(
  `id` int,
  `name` string,
  `created_at` timestamp)
ROW FORMAT SERDE
  'org.apache.hadoop.hive.ql.io.parquet.serde.ParquetHiveSerDe'
STORED AS INPUTFORMAT
  'org.apache.hadoop.hive.ql.io.parquet.MapredParquetInputFormat'
OUTPUTFORMAT
  'org.apache.hadoop.hive.ql.io.parquet.MapredParquetOutputFormat'

run below SQL:
spark.sql("select from_unixtime(unix_timestamp(created_at, 'yyyy-MM-dd HH:mm:ss'), 'yyyyMMdd') from xxx_timestamp ").show

fallback info:

!Expression <FromUnixTime> from_unixtime(unix_timestamp(created_at#27, yyyy-MM-dd HH:mm:ss, Some(UTC), false), yyyyMMdd, Some(UTC)) cannot run on GPU because CORRECTED format 'yyyyMMdd' on the GPU is not guaranteed to produce the same results as Spark on CPU. Set spark.rapids.sql.incompatibleDateFormats.enabled=true to force onto GPU.

[nvliyuan]

Low priority for customers.

[nvliyuan]

A similar fallback:
!Expression <FromUnixTime> from_unixtime(report_create_time#174L, yyyyMMdd, Some(Etc/UTC)) cannot run on GPU because LEGACY format 'yyyyMMdd' is not supported on the GPU.

[nvliyuan]

CC @firestarman @winningsix