[SPARK-55757][CORE] Improve spark.task.cpus validation

[dongjoon-hyun]

What changes were proposed in this pull request?

This PR aim to improve spark.task.cpus validation.

Why are the changes needed?

Currently, Apache Spark throws java.lang.ArithmeticException for the invalid spark.task.cpus. We had better provide actionable direction.

BEFORE

$ bin/spark-shell -c spark.task.cpus=0
26/02/28 09:00:56 ERROR SparkContext: Error initializing SparkContext.
java.lang.ArithmeticException: / by zero
	at org.apache.spark.resource.ResourceUtils$.warnOnWastedResources(ResourceUtils.scala:444)

AFTER

$ bin/spark-shell -c spark.task.cpus=0
...
26/02/28 09:02:15 ERROR SparkContext: Error initializing SparkContext.
org.apache.spark.SparkIllegalArgumentException: [INVALID_CONF_VALUE.REQUIREMENT] The value '0' in the config "spark.task.cpus" is invalid. The number of task CPUs must be positive. SQLSTATE: 22022

Does this PR introduce any user-facing change?

This only changes the error message before starting Spark job.

How was this patch tested?

Pass the CIs with newly added test case.

Was this patch authored or co-authored using generative AI tooling?

Generated-by: Gemini 3.1 Pro (High) on Antigravity.

[dongjoon-hyun]

Thank you, @pan3793 .

[dongjoon-hyun]

Also, FYI, cc @peter-toth

[dongjoon-hyun]

Thank you, @peter-toth ! Merged to master.