address_review_comments

Author: B Vadlamani

benchmarks/src/hj.rs

@@ -302,6 +302,8 @@ const HASH_QUERIES: &[HashJoinQuery] = &[
         probe_size: "60M",
     },
     // RightSemi Join benchmarks with Int32 keys
+    // Fanout: N/A for semi joins (returns at most one row per probe key)
+    //
     // Q16: RightSemi, Small build (25 rows), 100% Hit rate
     // Build Side: nation (25 rows) | Probe Side: customer (1.5M rows)
     HashJoinQuery {
@@ -342,6 +344,8 @@ const HASH_QUERIES: &[HashJoinQuery] = &[
         probe_size: "60M_RightSemi",
     },
     // RightAnti Join benchmarks with Int32 keys
+    // Fanout: N/A for anti joins (returns at most one row per probe key)
+    //
     // Q19: RightAnti, Small build (25 rows), 100% Hit rate (no output)
     // Build Side: nation (25 rows) | Probe Side: customer (1.5M rows)
     HashJoinQuery {
@@ -401,7 +405,9 @@ impl RunOpt {
             None => 1..=HASH_QUERIES.len(),
         };
 
-        let config = self.common.config()?;
+        let mut config = self.common.config()?;
+        // Disable join reordering to ensure the optimizer doesn't swap join sides
+        config.options_mut().optimizer.join_reordering = false;
         let rt = self.common.build_runtime()?;
         let ctx = SessionContext::new_with_config_rt(config, rt);
 

datafusion/physical-plan/benches/hash_join_semi_anti.rs

@@ -15,10 +15,7 @@
 // specific language governing permissions and limitations
 // under the License.
 
-//! Criterion benchmarks for Hash Join with RightSemi/RightAnti
-//!
-//! These benchmarks measure the hash join kernel for semi/anti joins
-//! with Int32 keys, which can use roaring bitmap optimization.
+//! Criterion benchmarks for Hash Join with RightSemi/RightAnti joins with Int32 keys.
 
 use std::sync::Arc;
 
@@ -33,7 +30,7 @@ use datafusion_physical_plan::joins::{HashJoinExec, PartitionMode, utils::JoinOn
 use datafusion_physical_plan::test::TestMemoryExec;
 use tokio::runtime::Runtime;
 
-/// Build RecordBatches with Int32 keys (for roaring optimization).
+/// Build RecordBatches with Int32 keys.
 ///
 /// Schema: (key: Int32, data: Int32, payload: Utf8)
 ///