[BUG] Inconsistency between the time zone in the fallback reason and the actual time zone checked in RapidsMeta.checkTImeZoneId

[gerashegalov]

Describe the bug
RapidsMeta.checkTImeZoneId compares the default time zone of the JVM to UTC in order to fallback on CPU, but quotes the session time zone as the offending actual

Steps/Code to reproduce bug

1. Start a Spark REPL with an unsupported i.e., non UTC default time zone such as GMT-8:

pyspark --driver-java-options -Duser.timezone=GMT-8 \
  --packages com.nvidia:rapids-4-spark_2.12:22.04.0 \
  --conf spark.rapids.sql.enabled=true \
  --conf spark.plugins=com.nvidia.spark.SQLPlugin \
  --conf spark.rapids.sql.explain=ALL

2. Set session's time zone to UTC

>>> spark.conf.set('spark.sql.session.timeZone', 'UTC')

3. Execute a query with a TimestampType
   And observe a confusing message Only UTC zone id is supported. Actual zone id: UTC

>>> spark.createDataFrame([ ['2022-06-01 07:45'] ], 'a string').selectExpr('hour(a)').collect()
22/05/27 08:23:58 WARN GpuOverrides: 
!Exec <ProjectExec> cannot run on GPU because not all expressions can be replaced
  @Expression <Alias> hour(cast(a#4 as timestamp), Some(UTC)) AS hour(a)#6 could run on GPU
    !Expression <Hour> hour(cast(a#4 as timestamp), Some(UTC)) cannot run on GPU because input expression Cast cast(a#4 as timestamp) (TimestampType is not supported when the JVM system timezone is set to GMT-08:00. Set the timezone to UTC to enable TimestampType support); Only UTC zone id is supported. Actual zone id: UTC
      !Expression <Cast> cast(a#4 as timestamp) cannot run on GPU because the GPU only supports a subset of formats when casting strings to timestamps. Refer to the CAST documentation for more details. To enable this operation on the GPU, set spark.rapids.sql.castStringToTimestamp.enabled to true.; Cast from StringType to TimestampType is not supported; Parsing the full rage of supported years is not supported. If your years are limited to 4 positive digits set spark.rapids.sql.hasExtendedYearValues to false.
        @Expression <AttributeReference> a#4 could run on GPU
  ! <RDDScanExec> cannot run on GPU because GPU does not currently support the operator class org.apache.spark.sql.execution.RDDScanExec
    @Expression <AttributeReference> a#4 could run on GPU

[Row(hour(a)=7)]

Expected behavior
A clear actionable message after determining if we can support non-UTC session time zone provided that the JVM time zone is the supported UTC

Environment details (please complete the following information)

• Any

Additional context
rapidsai/cudf#2477
h/t @viadea

[firestarman]

Shall we require both of session time zone id and the JVM time zone id to be "UTC" ?

[gerashegalov]

Shall we require both of session time zone id and the JVM time zone id to be "UTC" ?

+1 It was my vote too on the internal thread.

[firestarman]

+1 It was my vote too on the internal thread.

Thanks for the information. Do we have a deal now ?

[gerashegalov]

Feel free to propose a PR @firestarman

[firestarman]

@gerashegalov Here is the fix #5767, could you help review ?