Chore: Fix Scala code warnings - common module

[andy-hf-kwok]

Which issue does this PR close?

Partially closes #.#2255

Rationale for this change

• To perform code clean / refactor on the codebase, in order to comply with the new maven profile chore: Fix Scala code warnings  #2255, one module at a time.

What changes are included in this PR?

• Add explicit cast (int -> long) to avoid the warning
• Add a suppress warning on shim and put up a comment to justify it.

How are these changes tested?

• mvn test -pl :comet-common-spark3.5_2.12 -Pstrict-warnings to make sure the common module is now comply with the strict-warnings profile setting
• CI is green

[codecov-commenter]

Codecov Report

❌ Patch coverage is 0% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 59.00%. Comparing base (f09f8af) to head (e2a67ae).
⚠️ Report is 579 commits behind head on main.

Files with missing lines	Patch %	Lines
.../scala/org/apache/spark/sql/comet/util/Utils.scala	0.00%	2 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##               main    #2527      +/-   ##
============================================
+ Coverage     56.12%   59.00%   +2.88%     
- Complexity      976     1457     +481     
============================================
  Files           119      146      +27     
  Lines         11743    13566    +1823     
  Branches       2251     2358     +107     
============================================
+ Hits           6591     8005    +1414     
- Misses         4012     4340     +328     
- Partials       1140     1221      +81     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[comphead]

what was the warning for this change?

[andy-hf-kwok]

/scala/org/apache/spark/sql/comet/util/Utils.scala:226: implicit numeric widening

The type mismatch of batch.numRows() returning int, compare to the method return type Iterator[(Long, ChunkedByteBuffer)], which is long.
By working on this EPIC, I realized there are some more widening issue on the spark module also, sure we can probably suppress the warning, since this is a up cast, but if adding .toLong or equivalent would dismiss the warning all together I would reckon it would be better.
Being explicit on the intension is always better.