Added in fallback tests (NVIDIA#174)

integration_tests/src/main/python/asserts.py

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from conftest import is_incompat, should_sort_on_spark, should_sort_locally, get_float_check, get_limit
+from conftest import is_incompat, should_sort_on_spark, should_sort_locally, get_float_check, get_limit, spark_jvm
 from datetime import date, datetime
 import math
 from pyspark.sql import Row
@@ -235,6 +235,32 @@ def assert_gpu_and_cpu_writes_are_equal_iterator(write_func, read_func, base_pat
     """
     _assert_gpu_and_cpu_writes_are_equal(write_func, read_func, base_path, False, conf=conf)
 
+def assert_gpu_fallback_collect(func,
+        cpu_fallback_class_name,
+        conf={}):
+    (bring_back, collect_type) = _prep_func_for_compare(func, True)
+    conf = _prep_incompat_conf(conf)
+
+    print('### CPU RUN ###')
+    cpu_start = time.time()
+    from_cpu = with_cpu_session(bring_back, conf=conf)
+    cpu_end = time.time()
+    print('### GPU RUN ###')
+    jvm = spark_jvm()
+    jvm.ai.rapids.spark.ExecutionPlanCaptureCallback.startCapture()
+    gpu_start = time.time()
+    from_gpu = with_gpu_session(bring_back,
+            conf=conf)
+    gpu_end = time.time()
+    jvm.ai.rapids.spark.ExecutionPlanCaptureCallback.assertCapturedAndGpuFellBack(cpu_fallback_class_name, 2000)
+    print('### {}: GPU TOOK {} CPU TOOK {} ###'.format(collect_type, 
+        gpu_end - gpu_start, cpu_end - cpu_start))
+    if should_sort_locally():
+        from_cpu.sort(key=_RowCmp)
+        from_gpu.sort(key=_RowCmp)
+
+    assert_equal(from_cpu, from_gpu)
+
 def _assert_gpu_and_cpu_are_equal(func,
         should_collect,
         conf={}):

integration_tests/src/main/python/conftest.py

@@ -176,6 +176,9 @@ def _get_jvm_session(spark):
 def _get_jvm(spark):
     return spark.sparkContext._jvm
 
+def spark_jvm():
+    return _get_jvm(spark)
+
 class TpchRunner:
   def __init__(self, tpch_format, tpch_path):
     self.tpch_format = tpch_format

integration_tests/src/main/python/csv_test.py

@@ -14,7 +14,7 @@
 
 import pytest
 
-from asserts import assert_gpu_and_cpu_are_equal_collect
+from asserts import assert_gpu_and_cpu_are_equal_collect, assert_gpu_fallback_collect
 from datetime import datetime, timezone
 from data_gen import *
 from marks import *
@@ -166,8 +166,9 @@ def test_csv_fallback(spark_tmp_path, read_func, disable_conf):
     reader = read_func(data_path, schema, False, ',')
     with_cpu_session(
             lambda spark : gen_df(spark, gen).write.csv(data_path))
-    assert_gpu_and_cpu_are_equal_collect(
+    assert_gpu_fallback_collect(
             lambda spark : reader(spark).select(f.col('*'), f.col('_c2') + f.col('_c3')),
+            'FileSourceScanExec',
             conf={disable_conf: 'false'})
 
 csv_supported_date_formats = ['yyyy-MM-dd', 'yyyy/MM/dd', 'yyyy-MM', 'yyyy/MM',

integration_tests/src/main/python/orc_test.py

@@ -14,7 +14,7 @@
 
 import pytest
 
-from asserts import assert_gpu_and_cpu_are_equal_collect, assert_gpu_and_cpu_writes_are_equal_collect
+from asserts import assert_gpu_and_cpu_are_equal_collect, assert_gpu_and_cpu_writes_are_equal_collect, assert_gpu_fallback_collect
 from datetime import date, datetime, timezone
 from data_gen import *
 from marks import *
@@ -52,8 +52,9 @@ def test_orc_fallback(spark_tmp_path, read_func, disable_conf):
     reader = read_func(data_path)
     with_cpu_session(
             lambda spark : gen_df(spark, gen).write.orc(data_path))
-    assert_gpu_and_cpu_are_equal_collect(
+    assert_gpu_fallback_collect(
             lambda spark : reader(spark).select(f.col('*'), f.col('_c2') + f.col('_c3')),
+            'FileSourceScanExec',
             conf={disable_conf: 'false'})
 
 @pytest.mark.parametrize('orc_gens', orc_gens_list, ids=idfn)

integration_tests/src/main/python/parquet_test.py

@@ -14,7 +14,7 @@
 
 import pytest
 
-from asserts import assert_gpu_and_cpu_are_equal_collect, assert_gpu_and_cpu_writes_are_equal_collect
+from asserts import assert_gpu_and_cpu_are_equal_collect, assert_gpu_and_cpu_writes_are_equal_collect, assert_gpu_fallback_collect
 from datetime import date, datetime, timezone
 from data_gen import *
 from marks import *
@@ -56,8 +56,9 @@ def test_parquet_fallback(spark_tmp_path, read_func, disable_conf):
     reader = read_func(data_path)
     with_cpu_session(
             lambda spark : gen_df(spark, gen).write.parquet(data_path))
-    assert_gpu_and_cpu_are_equal_collect(
+    assert_gpu_fallback_collect(
             lambda spark : reader(spark).select(f.col('*'), f.col('_c2') + f.col('_c3')),
+            'FileSourceScanExec',
             conf={disable_conf: 'false'})
 
 parquet_compress_options = ['none', 'uncompressed', 'snappy', 'gzip']

integration_tests/src/main/python/spark_init_internal.py

@@ -21,6 +21,7 @@ def _spark__init():
     # can be reset in the middle of a test if specific operations are done (some types of cast etc)
     _s = SparkSession.builder \
             .config('spark.plugins', 'ai.rapids.spark.SQLPlugin') \
+            .config('spark.sql.queryExecutionListeners', 'ai.rapids.spark.ExecutionPlanCaptureCallback')\
             .appName('rapids spark plugin integration tests (python)').getOrCreate()
     #TODO catch the ClassNotFound error that happens if the classpath is not set up properly and
     # make it a better error message

integration_tests/src/test/scala/ai/rapids/spark/JoinsSuite.scala

@@ -60,10 +60,9 @@ class JoinsSuite extends SparkQueryCompareTestSuite {
     conf = new SparkConf()
       .set("spark.sql.autoBroadcastJoinThreshold", "-1")
       .set("spark.sql.join.preferSortMergeJoin", "true")
-      .set("spark.sql.shuffle.partitions", "2"), // hack to try and work around bug in cudf
+      .set("spark.sql.shuffle.partitions", "2"),
     incompat = true,
-    sort = true,
-    execsAllowedNonGpu = Seq("SortExec", "SortOrder")) {
+    sort = true) {
     (A, B) => A.join(B, A("longs") === B("longs"))
   }
 

shuffle-plugin/pom.xml

@@ -52,11 +52,6 @@
             <type>test-jar</type>
             <scope>test</scope>
         </dependency>
-        <dependency>
-            <groupId>org.scalatest</groupId>
-            <artifactId>scalatest_${scala.binary.version}</artifactId>
-            <scope>test</scope>
-        </dependency>
         <dependency>
             <groupId>ai.rapids</groupId>
             <artifactId>rapids-4-spark-sql_${scala.binary.version}</artifactId>
@@ -113,7 +108,7 @@
                 </execution>
               </executions>
             </plugin>
-            <!-- disable surefire as we are using scalatest only -->
+            <!-- disable surefire as tests are some place else -->
             <plugin>
                 <groupId>org.apache.maven.plugins</groupId>
                 <artifactId>maven-surefire-plugin</artifactId>
@@ -125,10 +120,6 @@
                 <groupId>net.alchim31.maven</groupId>
                 <artifactId>scala-maven-plugin</artifactId>
             </plugin>
-            <plugin>
-                <groupId>org.scalatest</groupId>
-                <artifactId>scalatest-maven-plugin</artifactId>
-            </plugin>
             <plugin>
                 <groupId>org.apache.rat</groupId>
                 <artifactId>apache-rat-plugin</artifactId>

sql-plugin/src/main/scala/ai/rapids/spark/Plugin.scala

@@ -17,27 +17,27 @@
 package ai.rapids.spark
 
 import java.util
-import java.util.concurrent.atomic.{AtomicLong, AtomicReference}
+import java.util.concurrent.atomic.{AtomicBoolean, AtomicReference}
 
 import scala.collection.JavaConverters._
 import scala.collection.mutable.ArrayBuffer
 
 import ai.rapids.cudf._
 import ai.rapids.spark.RapidsPluginImplicits._
-import org.apache.commons.lang3.mutable.MutableLong
 
 import org.apache.spark.{SparkConf, SparkContext, TaskContext}
 import org.apache.spark.api.plugin.{DriverPlugin, ExecutorPlugin, PluginContext, SparkPlugin}
 import org.apache.spark.internal.Logging
 import org.apache.spark.serializer.{JavaSerializer, KryoSerializer}
 import org.apache.spark.sql.SparkSessionExtensions
+import org.apache.spark.sql.catalyst.expressions.Expression
 import org.apache.spark.sql.catalyst.plans.physical.Partitioning
 import org.apache.spark.sql.catalyst.rules.Rule
 import org.apache.spark.sql.execution._
 import org.apache.spark.sql.internal.StaticSQLConf
 import org.apache.spark.sql.rapids.GpuShuffleEnv
+import org.apache.spark.sql.util.QueryExecutionListener
 import org.apache.spark.sql.vectorized.ColumnarBatch
-import org.apache.spark.util.TaskCompletionListener
 
 trait GpuPartitioning extends Partitioning {
 
@@ -255,6 +255,84 @@ class RapidsExecutorPlugin extends ExecutorPlugin with Logging {
   }
 }
 
+object ExecutionPlanCaptureCallback {
+  private[this] val shouldCapture: AtomicBoolean = new AtomicBoolean(false)
+  private[this] val execPlan: AtomicReference[SparkPlan] = new AtomicReference[SparkPlan]()
+
+  private def captureIfNeeded(qe: QueryExecution): Unit = {
+    if (shouldCapture.get()) {
+      execPlan.set(qe.executedPlan)
+    }
+  }
+
+  def startCapture(): Unit = {
+    execPlan.set(null)
+    shouldCapture.set(true)
+  }
+
+  def getResultWithTimeout(timeoutMs: Long = 2000): Option[SparkPlan] = {
+    try {
+      val endTime = System.currentTimeMillis() + timeoutMs
+      var plan = execPlan.getAndSet(null)
+      while (plan == null) {
+        if (System.currentTimeMillis() > endTime) {
+          return None
+        }
+        Thread.sleep(10)
+        plan = execPlan.getAndSet(null)
+      }
+      Some(plan)
+    } finally {
+      shouldCapture.set(false)
+      execPlan.set(null)
+    }
+  }
+
+  def assertCapturedAndGpuFellBack(fallbackCpuClass: String, timeoutMs: Long = 2000): Unit = {
+    val gpuPlan = getResultWithTimeout(timeoutMs=timeoutMs)
+    assert(gpuPlan.isDefined, "Did not capture a GPU plan")
+    assertDidFallBack(gpuPlan.get, fallbackCpuClass)
+  }
+
+  def assertDidFallBack(gpuPlan: SparkPlan, fallbackCpuClass: String): Unit = {
+    assert(gpuPlan.find(didFallBack(_, fallbackCpuClass)).isDefined,
+      s"Could not find $fallbackCpuClass in the GPU plan\n$gpuPlan")
+  }
+
+  private def getBaseNameFromClass(planClassStr: String): String = {
+    val firstDotIndex = planClassStr.lastIndexOf(".")
+    if (firstDotIndex != -1) planClassStr.substring(firstDotIndex + 1) else planClassStr
+  }
+
+  private def didFallBack(exp: Expression, fallbackCpuClass: String): Boolean = {
+    if (!exp.isInstanceOf[GpuExpression] &&
+      getBaseNameFromClass(exp.getClass.getName) == fallbackCpuClass) {
+      true
+    } else {
+      exp.children.exists(didFallBack(_, fallbackCpuClass))
+    }
+  }
+
+  private def didFallBack(plan: SparkPlan, fallbackCpuClass: String): Boolean = {
+    if (!plan.isInstanceOf[GpuExec] &&
+      getBaseNameFromClass(plan.getClass.getName) == fallbackCpuClass) {
+      true
+    } else {
+      plan.expressions.exists(didFallBack(_, fallbackCpuClass))
+    }
+  }
+}
+
+class ExecutionPlanCaptureCallback extends QueryExecutionListener {
+  import ExecutionPlanCaptureCallback._
+
+  override def onSuccess(funcName: String, qe: QueryExecution, durationNs: Long): Unit =
+    captureIfNeeded(qe)
+
+  override def onFailure(funcName: String, qe: QueryExecution, exception: Exception): Unit =
+    captureIfNeeded(qe)
+}
+
 /**
  * The RAPIDS plugin for Spark.
  * To enable this plugin, set the config "spark.plugins" to ai.rapids.spark.SQLPlugin

tests/src/test/scala/ai/rapids/spark/AnsiCastOpSuite.scala

@@ -605,24 +605,6 @@ class AnsiCastOpSuite extends GpuExpressionTestSuite {
     spark.sql(s"SELECT c0 FROM $t3")
   }
 
-  /** Test that a transformation is not supported on GPU */
-  private def testNotSupportedOnGpu(testName: String, frame: SparkSession => DataFrame,
-    sparkConf: SparkConf)(transformation: DataFrame => DataFrame): Unit = {
-
-    test(testName) {
-      try {
-        withGpuSparkSession(spark => {
-          val input = frame(spark).repartition(1)
-          transformation(input).collect()
-        }, sparkConf)
-        fail("should not run on GPU")
-      } catch {
-        case e: IllegalArgumentException =>
-          assert(e.getMessage.startsWith("Part of the plan is not columnar"))
-      }
-    }
-  }
-
   /**
    * Perform a transformation that is expected to fail due to values not being valid for
    * an ansi_cast

tests/src/test/scala/ai/rapids/spark/HashAggregatesSuite.scala

@@ -896,9 +896,7 @@ class HashAggregatesSuite extends SparkQueryCompareTestSuite {
       count("*"))
   }
 
-  testSparkResultsAreEqual("Agg expression with filter fall back", longsFromCSVDf,
-    execsAllowedNonGpu = Seq("HashAggregateExec", "AggregateExpression", "AttributeReference",
-        "Alias", "Literal", "Count", "GreaterThan")) {
+  testSparkResultsAreEqual("Agg expression with filter", longsFromCSVDf) {
     frame => frame.selectExpr("count(1) filter (where longs > 20)")
   }