Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Add FullOuter support to GpuShuffledSymmetricHashJoinExec #10272

Merged
merged 2 commits into from
Jan 29, 2024
Merged

Add FullOuter support to GpuShuffledSymmetricHashJoinExec #10272

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1af4060
Add FullOuter support to GpuShuffledSymmetricHashJoinExec
jlowe Oct 30, 2023
40ebdfb
review comments
jlowe Jan 26, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 9 additions & 4 deletions integration_tests/src/main/python/join_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1153,11 +1153,13 @@ def do_join(spark):
"spark.rapids.sql.input." + scan_name: False})

@ignore_order(local=True)
@pytest.mark.parametrize("join_type", ["Inner", "FullOuter"], ids=idfn)
@pytest.mark.parametrize("is_left_host_shuffle", [False, True], ids=idfn)
@pytest.mark.parametrize("is_right_host_shuffle", [False, True], ids=idfn)
@pytest.mark.parametrize("is_left_smaller", [False, True], ids=idfn)
@pytest.mark.parametrize("batch_size", ["1024", "1g"], ids=idfn)
def test_new_inner_join(is_left_host_shuffle, is_right_host_shuffle, is_left_smaller, batch_size):
def test_new_symmetric_join(join_type, is_left_host_shuffle, is_right_host_shuffle,
is_left_smaller, batch_size):
join_conf = {
"spark.rapids.sql.join.useShuffledSymmetricHashJoin": "true",
"spark.sql.autoBroadcastJoinThreshold": "1",
Expand All @@ -1183,14 +1185,17 @@ def do_join(spark):
left_df = left_df.groupBy("key1", "key2").max("ints", "floats")
if not is_right_host_shuffle:
right_df = right_df.groupBy("key1", "key2").max("doubles", "shorts")
return left_df.join(right_df, ["key1", "key2"], "inner")
return left_df.join(right_df, ["key1", "key2"], join_type)
assert_gpu_and_cpu_are_equal_collect(do_join, conf=join_conf)

@ignore_order(local=True)
@pytest.mark.parametrize("join_type", ["Inner", "FullOuter"], ids=idfn)
@pytest.mark.parametrize("is_left_smaller", [False, True], ids=idfn)
@pytest.mark.parametrize("is_ast_supported", [False, True], ids=idfn)
@pytest.mark.parametrize("batch_size", ["1024", "1g"], ids=idfn)
def test_new_inner_join_conditional(is_ast_supported, is_left_smaller, batch_size):
def test_new_symmetric_join_conditional(join_type, is_ast_supported, is_left_smaller, batch_size):
if join_type == "FullOuter" and not is_ast_supported:
pytest.skip("Full outer joins do not support a non-AST condition")
join_conf = {
"spark.rapids.sql.join.useShuffledSymmetricHashJoin": "true",
"spark.sql.autoBroadcastJoinThreshold": "1",
Expand All @@ -1213,5 +1218,5 @@ def do_join(spark):
else:
# AST does not support logarithm yet
cond.append(left_df.ints >= f.log(right_df.ints))
return left_df.join(right_df, cond, "inner")
return left_df.join(right_df, cond, join_type)
assert_gpu_and_cpu_are_equal_collect(do_join, conf=join_conf)
5 changes: 3 additions & 2 deletions sql-plugin/src/main/scala/com/nvidia/spark/rapids/GpuShuffledHashJoinExec.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -28,7 +28,7 @@ import org.apache.spark.internal.Logging
import org.apache.spark.rdd.RDD
import org.apache.spark.sql.catalyst.InternalRow
import org.apache.spark.sql.catalyst.expressions.{Attribute, Expression}
import org.apache.spark.sql.catalyst.plans.{Inner, InnerLike, JoinType, LeftAnti, LeftSemi}
import org.apache.spark.sql.catalyst.plans.{FullOuter, Inner, InnerLike, JoinType, LeftAnti, LeftSemi}
import org.apache.spark.sql.catalyst.plans.physical.Distribution
import org.apache.spark.sql.execution.SparkPlan
import org.apache.spark.sql.execution.joins.ShuffledHashJoinExec
Expand Down Expand Up @@ -70,8 +70,9 @@ class GpuShuffledHashJoinMeta(
}
val Seq(left, right) = childPlans.map(_.convertIfNeeded())
val joinExec = join.joinType match {
case Inner if conf.useShuffledSymmetricHashJoin =>
case Inner | FullOuter if conf.useShuffledSymmetricHashJoin =>
GpuShuffledSymmetricHashJoinExec(
join.joinType,
leftKeys.map(_.convertToGpu()),
rightKeys.map(_.convertToGpu()),
joinCondition,
Expand Down
Loading
Loading