[VL] No reason for native validation failed with round function

[dcoliversun]

Backend

VL (Velox)

Bug description

Two problems:

• Velox support round but Gluten couldn't push into native
• No reason for fallback

Test case:

test("round") {
    withTempPath {
      path =>
        Seq[Integer](
          25,
          36
        )
          .toDF("value")
          .write
          .parquet(path.getCanonicalPath)

        spark.read.parquet(path.getCanonicalPath).createOrReplaceTempView("view")

        runQueryAndCompare("SELECT round(value, -1) from view") {
          checkGlutenOperatorMatch[ProjectExecTransformer]
        }
    }
  }

The log is as follow:

16:05:46.980 WARN org.apache.hadoop.util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable

W20250507 16:05:56.172621 1131222 MemoryArbitrator.cpp:84] Query memory capacity[460.50MB] is set for NOOP arbitrator which has no capacity enforcement
E20250507 16:06:37.676007 1131695 Exceptions.h:66] Line: /opt/gluten/ep/build-velox/build/velox_ep/velox/exec/Task.cpp:2068, Function:terminate, Expression:  Cancelled, Source: RUNTIME, ErrorCode: INVALID_STATE

E20250507 16:06:46.796408 1131695 Exceptions.h:66] Line: /opt/gluten/ep/build-velox/build/velox_ep/velox/exec/Task.cpp:2068, Function:terminate, Expression:  Cancelled, Source: RUNTIME, ErrorCode: INVALID_STATE
16:06:53.176 WARN org.apache.spark.sql.execution.GlutenFallbackReporter: Validation failed for plan: Project, due to: 
 - Native validation failed: 
   |- 

16:06:54.658 WARN org.apache.spark.sql.execution.GlutenFallbackReporter: Validation failed for plan: Project[QueryId=13], due to: 
 - Native validation failed: 
   |- 



executedPlan.exists(((plan: org.apache.spark.sql.execution.SparkPlan) => tag.runtimeClass.isInstance(plan))) was false Expect ProjectExecTransformer exists in executedPlan:
 Project [round(value#144, -1) AS round(value, -1)#151]
+- VeloxColumnarToRow
   +- ^(3) BatchScanTransformer parquet file:/tmp/spark-c676d58b-b13a-43f2-82e6-cbdb3abcf21b[value#144] ParquetScan DataFilters: [], Format: parquet, Location: InMemoryFileIndex(1 paths)[file:/tmp/spark-c676d58b-b13a-43f2-82e6-cbdb3abcf21b], PartitionFilters: [], PushedAggregation: [], PushedFilters: [], PushedGroupBy: [], ReadSchema: struct<value:int> RuntimeFilters: [] NativeFilters: []

ScalaTestFailureLocation: org.apache.spark.sql.GlutenQueryTest at (GlutenQueryTest.scala:421)
org.scalatest.exceptions.TestFailedException: executedPlan.exists(((plan: org.apache.spark.sql.execution.SparkPlan) => tag.runtimeClass.isInstance(plan))) was false Expect ProjectExecTransformer exists in executedPlan:
 Project [round(value#144, -1) AS round(value, -1)#151]
+- VeloxColumnarToRow
   +- ^(3) BatchScanTransformer parquet file:/tmp/spark-c676d58b-b13a-43f2-82e6-cbdb3abcf21b[value#144] ParquetScan DataFilters: [], Format: parquet, Location: InMemoryFileIndex(1 paths)[file:/tmp/spark-c676d58b-b13a-43f2-82e6-cbdb3abcf21b], PartitionFilters: [], PushedAggregation: [], PushedFilters: [], PushedGroupBy: [], ReadSchema: struct<value:int> RuntimeFilters: [] NativeFilters: []

Gluten version

Gluten-1.4

Spark version

Spark-3.5.x

Spark configurations

No response

System information

No response

Relevant logs

[jinchengchenghh]

The scale in velox is uint8_t, so it does not support the negative scale, but I don't know why the reason cannot be returned