yaooqinn commented on code in PR #12077: URL: https://github.com/apache/gluten/pull/12077#discussion_r3231504205
########## backends-velox/src/main/scala/org/apache/gluten/execution/VeloxRDDScanTransformer.scala: ########## @@ -0,0 +1,92 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.gluten.execution + +import org.apache.gluten.backendsapi.velox.VeloxValidatorApi +import org.apache.gluten.config.{GlutenConfig, VeloxConfig} + +import org.apache.spark.rdd.RDD +import org.apache.spark.sql.catalyst.InternalRow +import org.apache.spark.sql.catalyst.expressions.{Attribute, SortOrder} +import org.apache.spark.sql.catalyst.plans.physical.Partitioning +import org.apache.spark.sql.execution.{RDDScanTransformer, SparkPlan} +import org.apache.spark.sql.execution.metric.{SQLMetric, SQLMetrics} +import org.apache.spark.sql.vectorized.ColumnarBatch + +/** + * Velox-backend implementation of RDDScanTransformer. + * + * Converts an RDD[InternalRow] into columnar batches using Velox's native row-to-columnar + * conversion (same JNI path as RowToVeloxColumnarExec). + */ +case class VeloxRDDScanTransformer( + outputAttributes: Seq[Attribute], + rdd: RDD[InternalRow], + name: String, + override val outputPartitioning: Partitioning, + override val outputOrdering: Seq[SortOrder] +) extends RDDScanTransformer(outputAttributes, outputPartitioning, outputOrdering) { + + @transient override lazy val metrics: Map[String, SQLMetric] = Map( + "numInputRows" -> SQLMetrics.createMetric(sparkContext, "number of input rows"), + "numOutputBatches" -> SQLMetrics.createMetric(sparkContext, "number of output batches"), + "convertTime" -> SQLMetrics.createTimingMetric(sparkContext, "time to convert") + ) + + override protected def doValidateInternal(): ValidationResult = { + for (field <- schema.fields) { + val reason = VeloxValidatorApi.validateSchema(field.dataType) + if (reason.isDefined) { + return ValidationResult.failed(reason.get) + } + } + ValidationResult.succeeded + } + + override def doExecuteColumnar(): RDD[ColumnarBatch] = { Review Comment: `RowToVeloxColumnarExec.toColumnarBatchIterator` does `UnsafeProjection.apply(row)`, which throws on a `BatchCarrierRow` since `PlaceholderRow`'s getters all throw `UnsupportedOperationException`. This can show up via `df.checkpoint()` or user code that does `df.queryExecution.toRdd` and re-wraps with `LogicalRDD.fromDataset`, when the upstream Gluten plan ends in `VeloxColumnarToCarrierRowExec`. `CHRDDScanTransformer.scala` L101-104 detects this and unwraps via `findNextTerminalRow.batch()`. Either mirror that, or fail fast with a clear error for carrier rows and add a checkpoint round-trip test to document the current behavior. ########## backends-velox/src/main/scala/org/apache/gluten/execution/VeloxRDDScanTransformer.scala: ########## @@ -0,0 +1,92 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.gluten.execution + +import org.apache.gluten.backendsapi.velox.VeloxValidatorApi +import org.apache.gluten.config.{GlutenConfig, VeloxConfig} + +import org.apache.spark.rdd.RDD +import org.apache.spark.sql.catalyst.InternalRow +import org.apache.spark.sql.catalyst.expressions.{Attribute, SortOrder} +import org.apache.spark.sql.catalyst.plans.physical.Partitioning +import org.apache.spark.sql.execution.{RDDScanTransformer, SparkPlan} +import org.apache.spark.sql.execution.metric.{SQLMetric, SQLMetrics} +import org.apache.spark.sql.vectorized.ColumnarBatch + +/** + * Velox-backend implementation of RDDScanTransformer. + * + * Converts an RDD[InternalRow] into columnar batches using Velox's native row-to-columnar + * conversion (same JNI path as RowToVeloxColumnarExec). + */ +case class VeloxRDDScanTransformer( + outputAttributes: Seq[Attribute], + rdd: RDD[InternalRow], + name: String, + override val outputPartitioning: Partitioning, + override val outputOrdering: Seq[SortOrder] +) extends RDDScanTransformer(outputAttributes, outputPartitioning, outputOrdering) { + + @transient override lazy val metrics: Map[String, SQLMetric] = Map( + "numInputRows" -> SQLMetrics.createMetric(sparkContext, "number of input rows"), + "numOutputBatches" -> SQLMetrics.createMetric(sparkContext, "number of output batches"), + "convertTime" -> SQLMetrics.createTimingMetric(sparkContext, "time to convert") + ) + + override protected def doValidateInternal(): ValidationResult = { + for (field <- schema.fields) { + val reason = VeloxValidatorApi.validateSchema(field.dataType) + if (reason.isDefined) { + return ValidationResult.failed(reason.get) + } + } + ValidationResult.succeeded + } + + override def doExecuteColumnar(): RDD[ColumnarBatch] = { + val numInputRows = longMetric("numInputRows") + val numOutputBatches = longMetric("numOutputBatches") + val convertTime = longMetric("convertTime") + val localSchema = this.schema + val batchSize = GlutenConfig.get.maxBatchSize + val batchBytes = VeloxConfig.get.veloxPreferredBatchBytes + rdd.mapPartitions { + iter => + RowToVeloxColumnarExec.toColumnarBatchIterator( + iter, + localSchema, + numInputRows, + numOutputBatches, + convertTime, + batchSize, + batchBytes) + } + } + + override protected def withNewChildrenInternal(newChildren: IndexedSeq[SparkPlan]): SparkPlan = + copy(outputAttributes, rdd, name, outputPartitioning, outputOrdering) +} + +object VeloxRDDScanTransformer { + def replace(plan: org.apache.spark.sql.execution.RDDScanExec): RDDScanTransformer = Review Comment: CH uses `UnknownPartitioning(0)`; we pass `plan.outputPartitioning` through. If the original `RDDScanExec` declares e.g. `HashPartitioning`, downstream Velox ops might skip a shuffle based on a hint we never verified survives the row→columnar conversion. Worth either justifying with a comment or aligning with CH. ########## backends-velox/src/test/scala/org/apache/gluten/execution/VeloxRDDScanSuite.scala: ########## @@ -0,0 +1,264 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.spark.sql.execution + +import org.apache.gluten.execution._ + +import org.apache.spark.SparkConf +import org.apache.spark.sql.Row +import org.apache.spark.sql.classic.ClassicDataset +import org.apache.spark.sql.execution.adaptive.AdaptiveSparkPlanHelper +import org.apache.spark.sql.types._ + +class VeloxRDDScanSuite extends VeloxWholeStageTransformerSuite with AdaptiveSparkPlanHelper { + + override protected val resourcePath: String = "/tpch-data-parquet" + override protected val fileFormat: String = "parquet" + + override protected def sparkConf: SparkConf = { + super.sparkConf + .set("spark.sql.ansi.enabled", "false") + } + + override def beforeAll(): Unit = { + super.beforeAll() + createTPCHNotNullTables() + } + + test("basic RDDScanExec is replaced by VeloxRDDScanTransformer") { + val data = spark.sql("SELECT l_orderkey, l_partkey FROM lineitem LIMIT 10") + val expectedAnswer = data.collect() + + val node = LogicalRDD.fromDataset( + rdd = data.queryExecution.toRdd, + originDataset = data, + isStreaming = false) + val df = ClassicDataset.ofRows(spark, node).toDF() + + checkAnswer(df, expectedAnswer) + val cnt = collect(df.queryExecution.executedPlan) { case _: VeloxRDDScanTransformer => true } + assert(cnt.nonEmpty, "Expected VeloxRDDScanTransformer in plan") + } + + test("RDDScan with string and numeric types") { + val data = spark.sql("""SELECT l_returnflag, l_linestatus, l_quantity, l_extendedprice + |FROM lineitem LIMIT 20""".stripMargin) + val expectedAnswer = data.collect() + + val node = LogicalRDD.fromDataset( + rdd = data.queryExecution.toRdd, + originDataset = data, + isStreaming = false) + val df = ClassicDataset.ofRows(spark, node).toDF() + + checkAnswer(df, expectedAnswer) + val cnt = collect(df.queryExecution.executedPlan) { case _: VeloxRDDScanTransformer => true } + assert(cnt.nonEmpty, "Expected VeloxRDDScanTransformer in plan") + } + + test("RDDScan with aggregation downstream") { + val query = + """SELECT l_returnflag, sum(l_quantity) AS sum_qty + |FROM lineitem + |WHERE l_shipdate <= date'1998-09-02' + |GROUP BY l_returnflag""".stripMargin + val data = spark.sql(query) + val expectedAnswer = data.collect() + + val node = LogicalRDD.fromDataset( + rdd = data.queryExecution.toRdd, + originDataset = data, + isStreaming = false) + val df = ClassicDataset.ofRows(spark, node).toDF() + + checkAnswer(df, expectedAnswer) Review Comment: This test — and the following `empty RDD` / `multiple re-reads` / `null values` / array / map / struct ones — only does `checkAnswer`. Without a `collectFirst { case _: VeloxRDDScanTransformer => true }` assertion they'd silently pass even if the rewriter stopped offloading (vanilla Spark also gets the right answer). Tests 1 and 2 already assert plan shape; please add the same here. ########## backends-velox/src/main/scala/org/apache/gluten/execution/VeloxRDDScanTransformer.scala: ########## @@ -0,0 +1,92 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.gluten.execution + +import org.apache.gluten.backendsapi.velox.VeloxValidatorApi +import org.apache.gluten.config.{GlutenConfig, VeloxConfig} + +import org.apache.spark.rdd.RDD +import org.apache.spark.sql.catalyst.InternalRow +import org.apache.spark.sql.catalyst.expressions.{Attribute, SortOrder} +import org.apache.spark.sql.catalyst.plans.physical.Partitioning +import org.apache.spark.sql.execution.{RDDScanTransformer, SparkPlan} +import org.apache.spark.sql.execution.metric.{SQLMetric, SQLMetrics} +import org.apache.spark.sql.vectorized.ColumnarBatch + +/** + * Velox-backend implementation of RDDScanTransformer. + * + * Converts an RDD[InternalRow] into columnar batches using Velox's native row-to-columnar + * conversion (same JNI path as RowToVeloxColumnarExec). + */ +case class VeloxRDDScanTransformer( + outputAttributes: Seq[Attribute], + rdd: RDD[InternalRow], + name: String, + override val outputPartitioning: Partitioning, + override val outputOrdering: Seq[SortOrder] +) extends RDDScanTransformer(outputAttributes, outputPartitioning, outputOrdering) { + + @transient override lazy val metrics: Map[String, SQLMetric] = Map( + "numInputRows" -> SQLMetrics.createMetric(sparkContext, "number of input rows"), + "numOutputBatches" -> SQLMetrics.createMetric(sparkContext, "number of output batches"), + "convertTime" -> SQLMetrics.createTimingMetric(sparkContext, "time to convert") + ) + + override protected def doValidateInternal(): ValidationResult = { + for (field <- schema.fields) { + val reason = VeloxValidatorApi.validateSchema(field.dataType) + if (reason.isDefined) { + return ValidationResult.failed(reason.get) + } + } + ValidationResult.succeeded + } + + override def doExecuteColumnar(): RDD[ColumnarBatch] = { + val numInputRows = longMetric("numInputRows") + val numOutputBatches = longMetric("numOutputBatches") + val convertTime = longMetric("convertTime") + val localSchema = this.schema + val batchSize = GlutenConfig.get.maxBatchSize + val batchBytes = VeloxConfig.get.veloxPreferredBatchBytes + rdd.mapPartitions { + iter => + RowToVeloxColumnarExec.toColumnarBatchIterator( + iter, + localSchema, + numInputRows, + numOutputBatches, + convertTime, + batchSize, + batchBytes) + } + } + + override protected def withNewChildrenInternal(newChildren: IndexedSeq[SparkPlan]): SparkPlan = Review Comment: Leaf node — could `assert(newChildren.isEmpty)` here. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
