Add HiveHash support for nested types (#11660)

* Add HiveHash support for nested types Signed-off-by: ustcfy <[email protected]> * Primitive types do not enter the stack Signed-off-by: Yan Feng <[email protected]> * Generate docs Signed-off-by: Yan Feng <[email protected]> * Generate docs Signed-off-by: Yan Feng <[email protected]> * Add fallback test case for exceeding nesting depth limit Signed-off-by: Yan Feng <[email protected]> * Retrieve supportedDepth from JNI Signed-off-by: Yan Feng <[email protected]> --------- Signed-off-by: ustcfy <[email protected]> Signed-off-by: Yan Feng <[email protected]>
NVIDIA · Dec 14, 2024 · ab3111a · ab3111a
1 parent 0764c18
commit ab3111a
Show file tree

Hide file tree

Showing 28 changed files with 98 additions and 34 deletions.
diff --git a/docs/supported_ops.md b/docs/supported_ops.md
@@ -8480,9 +8480,9 @@ are limited.
 <td>S</td>
 <td><b>NS</b></td>
 <td><b>NS</b></td>
+<td><em>PS<br/>The nesting depth has a certain limit;<br/>UTC is only supported TZ for child TIMESTAMP;<br/>unsupported child types DECIMAL, BINARY, CALENDAR, MAP, UDT, DAYTIME, YEARMONTH</em></td>
 <td><b>NS</b></td>
-<td><b>NS</b></td>
-<td><b>NS</b></td>
+<td><em>PS<br/>The nesting depth has a certain limit;<br/>UTC is only supported TZ for child TIMESTAMP;<br/>unsupported child types DECIMAL, BINARY, CALENDAR, MAP, UDT, DAYTIME, YEARMONTH</em></td>
 <td><b>NS</b></td>
 <td><b>NS</b></td>
 <td><b>NS</b></td>

diff --git a/integration_tests/src/main/python/datasourcev2_write_test.py b/integration_tests/src/main/python/datasourcev2_write_test.py
@@ -15,10 +15,10 @@
 import pytest
 
 from asserts import assert_gpu_fallback_collect, assert_equal_with_local_sort
-from data_gen import gen_df, decimal_gens, non_utc_allow
+from data_gen import gen_df, decimal_gens, non_utc_allow, StructGen, ArrayGen, string_gen
 from marks import *
 from spark_session import is_hive_available, is_spark_330_or_later, with_cpu_session, with_gpu_session
-from hive_parquet_write_test import _hive_bucket_gens_sans_bools, _hive_array_gens, _hive_struct_gens
+from hive_parquet_write_test import _hive_bucket_gens_sans_bools
 from hive_parquet_write_test import read_single_bucket
 
 _hive_write_conf = {
@@ -77,7 +77,7 @@ def write_hive_table(spark, out_table):
 @pytest.mark.skipif(not (is_hive_available() and is_spark_330_or_later()),
                     reason="Must have Hive on Spark 3.3+")
 @pytest.mark.parametrize('file_format', ['parquet', 'orc'])
-@pytest.mark.parametrize('gen', decimal_gens + _hive_array_gens + _hive_struct_gens)
+@pytest.mark.parametrize('gen', decimal_gens)
 def test_write_hive_bucketed_unsupported_types_fallback(spark_tmp_table_factory, file_format, gen):
     out_table = spark_tmp_table_factory.get()
 
@@ -94,3 +94,31 @@ def create_hive_table(spark):
             "insert into {0} select * from {1}".format(out_table, input_table)),
         'DataWritingCommandExec',
         _hive_write_conf)
+
+# The calculation of nesting depth here does not take basic types into account.
+nested_gen_depth_9 = [StructGen([('depth_8', StructGen([('depth_7', StructGen([('depth_6',
+    StructGen([('depth_5', ArrayGen(StructGen([('depth_3', StructGen([('depth_2',
+    ArrayGen(StructGen([('depth_0', string_gen)]), max_length=5))]))]), max_length=5))]))]))]))])]
+
+@ignore_order
+@allow_non_gpu('DataWritingCommandExec,ExecutedCommandExec,SortExec,WriteFilesExec')
+@pytest.mark.skipif(not (is_hive_available() and is_spark_330_or_later()),
+                    reason="Must have Hive on Spark 3.3+")
+@pytest.mark.parametrize('file_format', ['parquet', 'orc'])
+@pytest.mark.parametrize('gen', nested_gen_depth_9)
+def test_write_hive_bucketed_nesting_depth_exceed_limit_fallback(spark_tmp_table_factory, file_format, gen):
+    out_table = spark_tmp_table_factory.get()
+
+    def create_hive_table(spark):
+        spark.sql("create table {0} (a {1}) stored as {2} clustered by (a) into 3 buckets".format(
+            out_table, gen.data_type.simpleString(), file_format))
+        data_table = spark_tmp_table_factory.get()
+        gen_df(spark, [('a', gen)], length=1).createOrReplaceTempView(data_table)
+        return data_table
+
+    input_table = with_cpu_session(create_hive_table, _hive_write_conf)
+    assert_gpu_fallback_collect(
+        lambda spark: spark.sql(
+            "insert into {0} select * from {1}".format(out_table, input_table)),
+        'DataWritingCommandExec',
+        _hive_write_conf)
diff --git a/integration_tests/src/main/python/hive_parquet_write_test.py b/integration_tests/src/main/python/hive_parquet_write_test.py
@@ -25,12 +25,27 @@
 # "GpuInsertIntoHiveTable" for Parquet write.
 _write_to_hive_conf = {"spark.sql.hive.convertMetastoreParquet": False}
 
-_hive_bucket_gens_sans_bools = [
+_hive_bucket_basic_gens_sans_bools = [
     byte_gen, short_gen, int_gen, long_gen, string_gen, float_gen, double_gen,
     DateGen(start=date(1590, 1, 1)), _restricted_timestamp()]
-_hive_bucket_gens = [boolean_gen] + _hive_bucket_gens_sans_bools
+_hive_bucket_basic_gens = [boolean_gen] + _hive_bucket_basic_gens_sans_bools
 
-_hive_basic_gens = _hive_bucket_gens + [
+_hive_bucket_basic_struct_gen = StructGen(
+    [['c'+str(ind), c_gen] for ind, c_gen in enumerate(_hive_bucket_basic_gens_sans_bools)])
+
+_hive_bucket_struct_gens = [
+    _hive_bucket_basic_struct_gen,
+    StructGen([['child0', byte_gen], ['child1', _hive_bucket_basic_struct_gen]]),
+    StructGen([['child0', ArrayGen(short_gen)], ['child1', double_gen]])]
+
+_hive_bucket_array_gens = [ArrayGen(sub_gen) for sub_gen in _hive_bucket_basic_gens_sans_bools] + [
+    ArrayGen(ArrayGen(short_gen, max_length=10), max_length=10),
+    ArrayGen(ArrayGen(string_gen, max_length=10), max_length=10),
+    ArrayGen(StructGen([['child0', byte_gen], ['child1', string_gen], ['child2', float_gen]]))]
+
+_hive_bucket_gens_sans_bools = _hive_bucket_basic_gens_sans_bools + _hive_bucket_struct_gens + _hive_bucket_array_gens
+
+_hive_basic_gens = _hive_bucket_basic_gens + [
     DecimalGen(precision=19, scale=1, nullable=True),
     DecimalGen(precision=23, scale=5, nullable=True),
     DecimalGen(precision=36, scale=3, nullable=True)]
@@ -197,7 +212,7 @@ def test_insert_hive_bucketed_table(spark_tmp_table_factory):
     num_rows = 2048
 
     def gen_table(spark):
-        gen_list = [('_c' + str(i), gen) for i, gen in enumerate(_hive_bucket_gens)]
+        gen_list = [('_c' + str(i), gen) for i, gen in enumerate(_hive_bucket_gens_sans_bools)]
         types_sql_str = ','.join('{} {}'.format(
             name, gen.data_type.simpleString()) for name, gen in gen_list)
         col_names_str = ','.join(name for name, gen in gen_list)

diff --git a/sql-plugin/src/main/scala/com/nvidia/spark/rapids/GpuOverrides.scala b/sql-plugin/src/main/scala/com/nvidia/spark/rapids/GpuOverrides.scala
@@ -25,6 +25,7 @@ import scala.util.control.NonFatal
 import ai.rapids.cudf.DType
 import com.nvidia.spark.rapids.RapidsConf.{SUPPRESS_PLANNING_FAILURE, TEST_CONF}
 import com.nvidia.spark.rapids.jni.GpuTimeZoneDB
+import com.nvidia.spark.rapids.jni.Hash
 import com.nvidia.spark.rapids.lore.GpuLore
 import com.nvidia.spark.rapids.shims._
 import com.nvidia.spark.rapids.window.{GpuDenseRank, GpuLag, GpuLead, GpuPercentRank, GpuRank, GpuRowNumber, GpuSpecialFrameBoundary, GpuWindowExecMeta, GpuWindowSpecDefinitionMeta}
@@ -3320,8 +3321,28 @@ object GpuOverrides extends Logging {
       "hive hash operator",
       ExprChecks.projectOnly(TypeSig.INT, TypeSig.INT,
         repeatingParamCheck = Some(RepeatingParamCheck("input",
-          TypeSig.commonCudfTypes + TypeSig.NULL, TypeSig.all))),
+          (TypeSig.commonCudfTypes + TypeSig.NULL + TypeSig.STRUCT + TypeSig.ARRAY).nested() +
+              TypeSig.psNote(TypeEnum.ARRAY, "The nesting depth has a certain limit") +
+              TypeSig.psNote(TypeEnum.STRUCT, "The nesting depth has a certain limit"),
+          TypeSig.all))),
       (a, conf, p, r) => new ExprMeta[HiveHash](a, conf, p, r) {
+        override def tagExprForGpu(): Unit = {
+          def getMaxStackDepth(inputType: DataType): Int = {
+            inputType match {
+              case at: ArrayType => 1 + getMaxStackDepth(at.elementType)
+              case st: StructType =>
+                1 + st.map(f => getMaxStackDepth(f.dataType)).max
+              case _ => 0 // primitive types
+            }
+          }
+          val maxDepth = a.children.map(c => getMaxStackDepth(c.dataType)).max
+          val supportedDepth = Hash.MAX_STACK_DEPTH
+          if (maxDepth > supportedDepth) {
+            willNotWorkOnGpu(s"the data type requires a stack size of $maxDepth, " +
+              s"which exceeds the GPU limit of $supportedDepth")
+          }
+        }
+
         def convertToGpu(): GpuExpression =
           GpuHiveHash(childExprs.map(_.convertToGpu()))
       }),

diff --git a/tools/generated_files/320/supportedExprs.csv b/tools/generated_files/320/supportedExprs.csv
@@ -258,7 +258,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/321/supportedExprs.csv b/tools/generated_files/321/supportedExprs.csv
@@ -258,7 +258,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/321cdh/supportedExprs.csv b/tools/generated_files/321cdh/supportedExprs.csv
@@ -258,7 +258,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/322/supportedExprs.csv b/tools/generated_files/322/supportedExprs.csv
@@ -258,7 +258,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/323/supportedExprs.csv b/tools/generated_files/323/supportedExprs.csv
@@ -258,7 +258,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/324/supportedExprs.csv b/tools/generated_files/324/supportedExprs.csv
@@ -258,7 +258,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/330/supportedExprs.csv b/tools/generated_files/330/supportedExprs.csv
@@ -267,7 +267,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/330cdh/supportedExprs.csv b/tools/generated_files/330cdh/supportedExprs.csv
@@ -267,7 +267,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/331/supportedExprs.csv b/tools/generated_files/331/supportedExprs.csv
@@ -269,7 +269,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/332/supportedExprs.csv b/tools/generated_files/332/supportedExprs.csv
@@ -269,7 +269,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/332cdh/supportedExprs.csv b/tools/generated_files/332cdh/supportedExprs.csv
@@ -269,7 +269,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/333/supportedExprs.csv b/tools/generated_files/333/supportedExprs.csv
@@ -269,7 +269,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/334/supportedExprs.csv b/tools/generated_files/334/supportedExprs.csv
@@ -269,7 +269,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/340/supportedExprs.csv b/tools/generated_files/340/supportedExprs.csv
@@ -269,7 +269,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA

diff --git a/tools/generated_files/341/supportedExprs.csv b/tools/generated_files/341/supportedExprs.csv
@@ -269,7 +269,7 @@ GreaterThanOrEqual,S,`>=`,None,AST,rhs,S,S,S,S,S,NS,NS,S,PS,S,NS,NS,NS,NS,NS,NA,
 GreaterThanOrEqual,S,`>=`,None,AST,result,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Greatest,S,`greatest`,None,project,param,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
 Greatest,S,`greatest`,None,project,result,S,S,S,S,S,S,S,S,PS,S,S,S,NS,NS,NS,NA,NS,NS,NA,NA
-HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,NS,NS,NS,NS,NS,NS
+HiveHash,S, ,None,project,input,S,S,S,S,S,S,S,S,PS,S,NS,S,NS,NS,PS,NS,PS,NS,NS,NS
 HiveHash,S, ,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,input,NA,NA,NA,NA,NA,NA,NA,NA,PS,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA
 Hour,S,`hour`,None,project,result,NA,NA,NA,S,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA,NA