You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cloudberry.apache.org by ji...@apache.org on 2025/01/16 13:15:55 UTC
(cloudberry) 12/35: Fall back in Orca for queries with RTE of type TableFunc (#14898)
This is an automated email from the ASF dual-hosted git repository.
jiaqizho pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/cloudberry.git
commit 574537722ddaf8d5fc5d7eba891764b41204796b
Author: Chris Hajas <ch...@vmware.com>
AuthorDate: Mon Feb 6 11:55:22 2023 -0800
Fall back in Orca for queries with RTE of type TableFunc (#14898)
Previously, Orca would crash if encountering this. For now, fall back
similarly to what we do for RTE_TABLEFUNCTION.
Example query: explain SELECT * FROM xmltable('/root' passing '' COLUMNS element text);
This actually fell back correctly in retail build, but we were hitting a
`__builtin_unreachable` call, which is undefined behavior, so we were
likely just getting lucky.
Fixes issue reported in https://groups.google.com/a/greenplum.org/g/gpdb-dev/c/4F8GUSs04L8
---
src/backend/gpopt/translate/CTranslatorQueryToDXL.cpp | 1 +
src/test/regress/expected/qp_orca_fallback.out | 8 ++++++++
src/test/regress/expected/qp_orca_fallback_optimizer.out | 10 ++++++++++
src/test/regress/sql/qp_orca_fallback.sql | 3 +++
4 files changed, 22 insertions(+)
diff --git a/src/backend/gpopt/translate/CTranslatorQueryToDXL.cpp b/src/backend/gpopt/translate/CTranslatorQueryToDXL.cpp
index 212be11e35..68b379be26 100644
--- a/src/backend/gpopt/translate/CTranslatorQueryToDXL.cpp
+++ b/src/backend/gpopt/translate/CTranslatorQueryToDXL.cpp
@@ -3342,6 +3342,7 @@ CTranslatorQueryToDXL::UnsupportedRTEKind(RTEKind rtekind)
GPOS_WSZ_LIT("RangeTableEntry of type Void"));
}
case RTE_TABLEFUNCTION:
+ case RTE_TABLEFUNC:
{
GPOS_RAISE(gpdxl::ExmaDXL, gpdxl::ExmiQuery2DXLUnsupportedFeature,
GPOS_WSZ_LIT("RangeTableEntry of type Table Function"));
diff --git a/src/test/regress/expected/qp_orca_fallback.out b/src/test/regress/expected/qp_orca_fallback.out
index d0a7fbb4f2..0d34a26ad3 100644
--- a/src/test/regress/expected/qp_orca_fallback.out
+++ b/src/test/regress/expected/qp_orca_fallback.out
@@ -239,6 +239,14 @@ explain select count(*) from foo group by a;
Optimizer: Postgres query optimizer
(5 rows)
+-- Orca should fallback for RTE_TABLEFUNC RTE type
+explain SELECT * FROM xmltable('/root' passing '' COLUMNS element text);
+ QUERY PLAN
+------------------------------------------------------------------------
+ Table Function Scan on "xmltable" (cost=0.00..1.00 rows=100 width=32)
+ Optimizer: Postgres query optimizer
+(2 rows)
+
create table ext_part(a int) partition by list(a);
NOTICE: Table doesn't have 'DISTRIBUTED BY' clause -- Using column named 'a' as the Cloudberry Database data distribution key for this table.
HINT: The 'DISTRIBUTED BY' clause determines the distribution of data. Make sure column(s) chosen are the optimal data distribution key to minimize skew.
diff --git a/src/test/regress/expected/qp_orca_fallback_optimizer.out b/src/test/regress/expected/qp_orca_fallback_optimizer.out
index 91e64f19cc..385089b6fa 100644
--- a/src/test/regress/expected/qp_orca_fallback_optimizer.out
+++ b/src/test/regress/expected/qp_orca_fallback_optimizer.out
@@ -287,6 +287,16 @@ DETAIL: No plan has been computed for required properties
Optimizer: Postgres query optimizer
(5 rows)
+-- Orca should fallback for RTE_TABLEFUNC RTE type
+explain SELECT * FROM xmltable('/root' passing '' COLUMNS element text);
+INFO: GPORCA failed to produce a plan, falling back to planner
+DETAIL: Feature not supported: RangeTableEntry of type Table Function
+ QUERY PLAN
+------------------------------------------------------------------------
+ Table Function Scan on "xmltable" (cost=0.00..1.00 rows=100 width=32)
+ Optimizer: Postgres query optimizer
+(2 rows)
+
create table ext_part(a int) partition by list(a);
NOTICE: Table doesn't have 'DISTRIBUTED BY' clause -- Using column named 'a' as the Cloudberry Database data distribution key for this table.
HINT: The 'DISTRIBUTED BY' clause determines the distribution of data. Make sure column(s) chosen are the optimal data distribution key to minimize skew.
diff --git a/src/test/regress/sql/qp_orca_fallback.sql b/src/test/regress/sql/qp_orca_fallback.sql
index d25436e090..f31be1997f 100644
--- a/src/test/regress/sql/qp_orca_fallback.sql
+++ b/src/test/regress/sql/qp_orca_fallback.sql
@@ -103,6 +103,9 @@ set optimizer_enable_hashagg = off;
set optimizer_enable_groupagg = off;
explain select count(*) from foo group by a;
+-- Orca should fallback for RTE_TABLEFUNC RTE type
+explain SELECT * FROM xmltable('/root' passing '' COLUMNS element text);
+
create table ext_part(a int) partition by list(a);
create table p1(a int);
create external web table p2_ext (like p1) EXECUTE 'cat something.txt' FORMAT 'TEXT';
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@cloudberry.apache.org
For additional commands, e-mail: commits-help@cloudberry.apache.org