[DocDB] alma8 fastdebug gcc11 compilation: ‘prev_op’ may be used uninitialized in this function #21811
Closed
1 task done
Labels
area/docdb
YugabyteDB core features
kind/bug
This issue is a bug
priority/medium
Medium priority issue
status/awaiting-triage
Issue awaiting triage
Comments
jasonyb
added
area/docdb
yugabyte-ci
added
kind/bug
jasonyb
pushed a commit
that referenced
this issue
Apr 25, 2024
Summary: Commit 061dc69 introduces code causing compilation failure ../../src/yb/integration-tests/cdcsdk_ysql_test_base.cc: In member function ‘void yb::cdc::CDCSDKYsqlTest::CheckRecordsConsistencyFromVWAL(const std::vector<yb::cdc::CDCSDKProtoRecordPB>&)’: ../../src/yb/integration-tests/cdcsdk_ysql_test_base.cc:3879:9: error: ‘prev_op’ may be used uninitialized in this function [-Werror=maybe-uninitialized] 3879 | if (prev_op == RowMessage::DDL) { | ^~ cc1plus: all warnings being treated as errors Almalinux 8 + fastdebug + gcc11 is affected, but other build combos such as those in jenkins_jobs.yml are fine. Fix the issue by initializing prev_op to UNKNOWN. This should be fine according to the logic as far as I can tell. Jira: DB-10712 Test Plan: On Almalinux 8: ./yb_build.sh fastdebug --gcc11 Close: #21811 Original commit: 193589f / D33830 Reviewers: siddharth.shah Reviewed By: siddharth.shah Tags: #jenkins-ready Differential Revision: https://phorge.dev.yugabyte.com/D34494
jasonyb
pushed a commit
that referenced
this issue
May 14, 2024
…ter-merge Merge YB master commit 193589f titled [#21811] docdb: fix alma8 fastdebug gcc11 compilation and committed 2024-04-04T10:20:10-07:00 to YB pg15. - TestPgReplicationSlot.java: - consumptionOnSubsetOfColocatedTables: YB pg15 merge 5fe2012 ignores this test. YB master fddfec6 moves this test lower. YB master 71ad62a adds new test dynamicTableAdditionForTablesCreatedBeforeStreamCreation in the old location. This causes merge conflict. Apply the ignore to the new location. - *: add @ignore for a bunch of new tests introduced by various YB master commits because they fail. This is along the lines of what has been already happening to this file in previous merges. - nodeYbBitmapTablescan.c: - ExecReScanYbBitmapTableScan: YB master 7ac6dc2 adds logic to pfree ss_currentScanDesc. In YB pg15, it is already being freed by ybc_heap_endscan(tsdesc). In YB pg15, the YbScanDesc contains TableScanDesc whereas in YB master, YbScanDesc is contained in HeapScanDesc. All in all, the incoming logic from YB master is not needed because things are already being freed properly. - logical.c: - function declarations: YB master 459e6b6 adds yb_schema_change_cb_wrapper declaration in the middle of PG-owned declarations. Upstream PG 45fdc9738b36d1068d3ad8fdb06436d6fd14436b and 0aa8a01d04c8fe200b7a106878eebc3d0af9105c add extra functions in the same area. Adjacent lines conflict. - StartupDecodingContext: again, upstream PG 45fdc9738b36d1068d3ad8fdb06436d6fd14436b and 0aa8a01d04c8fe200b7a106878eebc3d0af9105c add code where YB master 459e6b6 adds code. Adjacent lines conflict. - yb_schema_change_cb_wrapper: (same) - proto.c: - function declarations: upstream PG 52e4f0cd472d39d07732b99559989ea3b615be78 and 923def9a533a7d986acfb524139d8b9e5466d0a5 change the signature of logicalrep_write_tuple. YB master fddfec6 also changes signature. Apply both. - logicalrep_write_insert: (same) - logicalrep_write_update signature: upstream PG 464824323e57dc4b397e8b05854d779908b55304, 52e4f0cd472d39d07732b99559989ea3b615be78, and 923def9a533a7d986acfb524139d8b9e5466d0a5 change the signature. YB master fddfec6 also changes the signature. Apply both. - logicalrep_write_update body: (same as logicalrep_write_insert) - logicalrep_write_delete: (same) - logicalrep_write_tuple: (same) - yb_virtual_wal_client.c: - YBCReadRecord: YB master 71ad62a adds publication_names parameter while YB pg15 merge 417e9b3 removes RecPtr parameter. Apply both. - pgoutput.c: - _PG_output_plugin_init: upstream PG 464824323e57dc4b397e8b05854d779908b55304 and ac4645c0157fc5fcef0af8ff571512aa284a2cec add code where YB master 459e6b6 adds code. Adjacent lines conflict. - pgoutput_change case REORDER_BUFFER_CHANGE_UPDATE: upstream PG 83fd4532a72179c370e318075a10e0e2aa832024 and 52e4f0cd472d39d07732b99559989ea3b615be78 significantly change the code. YB master fddfec6 adds code to determine yb_old_is_omitted and yb_new_is_omitted, and it passes them to the logicalrep_write_update call. Major conflict. - slot.c: - ReplicationSlotAcquire: major conflict: many commits appear to be involved on both sides. Follow the resolution in YB pg15 merge ee08bd6 to change "slot" to "s" and take YB master's version. - walsender.c: - XLogSendLogical: (same as yb_virtual_wal_client.c YBCReadRecord) - pg_yb_utils.c: - YbGetDdlMode T_VacuumStmt: YB pg15 0a5cbae and YB master 9a27aff add logic in the same area. Keep both. - logicalproto.h: - logicalrep_write_update: (same as proto.c logicalrep_write_update signature) - output_plugin.h: - YBLogicalDecodeSchemaChangeCB: (same as logical.c function declarations) - OutputPluginCallbacks: (same as logical.c function declarations) - reorderbuffer.h: - YBReorderBufferSchemaChangeCB: (same as logical.c function declarations) - yb_virtual_wal_client.h: - YBCReadRecord: (same as yb_virtual_wal_client.c YBCReadRecord) - aggregates.sql: - YB master b2f8822 is an upstream PG point import of 44e95b5728a4569c494fa4ea4317f8a2f50a206b, which is already part of upstream PG REL_15_2. Drop incoming YB master changes because YB pg15 is already up to date. - aggregates.out: - (same as aggregates.sql) - yb_decode.c: - YBDecodeUpdate: like YB pg15 merge d68d96c, pass false for the new argument of the new ReorderBufferQueueChange call. - build_postgres.py: - make_postgres: add new extensions yb_ycql_utils (YB master ce7d880) and yb_xcluster_ddl_replication (YB master 0185d16) to the list of directories to build. YB pg15 initial merge disabled building all contrib extensions automatically, so each extension has to be added to this list for now. - yb_xcluster_ddl_replication.c: - HandleSourceDDLEnd: this file is introduced by YB master 0185d16. In various places, adjust code according to upstream PG 2f9661311b83dc481fc19f6e3bda015392010a40, which makes trig_data->tag of CommandTag type rather than char*. Use GetCommandTagName where str is needed and switch to comparing CommandTag to CommandTag instead of str to str where possible. - IsInIgnoreList: (same)
jasonyb
pushed a commit
that referenced
this issue
May 14, 2024
…9b0af3c' into pg15 Summary: Merge YB master commit 193589f titled [#21811] docdb: fix alma8 fastdebug gcc11 compilation and committed 2024-04-04T10:20:10-07:00 to YB pg15. - TestPgReplicationSlot.java: - consumptionOnSubsetOfColocatedTables: YB pg15 merge 5fe2012 ignores this test. YB master fddfec6 moves this test lower. YB master 71ad62a adds new test dynamicTableAdditionForTablesCreatedBeforeStreamCreation in the old location. This causes merge conflict. Apply the ignore to the new location. - *: add @ignore for a bunch of new tests introduced by various YB master commits because they fail. This is along the lines of what has been already happening to this file in previous merges. - nodeYbBitmapTablescan.c: - ExecReScanYbBitmapTableScan: YB master 7ac6dc2 adds logic to pfree ss_currentScanDesc. In YB pg15, it is already being freed by ybc_heap_endscan(tsdesc). In YB pg15, the YbScanDesc contains TableScanDesc whereas in YB master, YbScanDesc is contained in HeapScanDesc. All in all, the incoming logic from YB master is not needed because things are already being freed properly. - logical.c: - function declarations: YB master 459e6b6 adds yb_schema_change_cb_wrapper declaration in the middle of PG-owned declarations. Upstream PG 45fdc9738b36d1068d3ad8fdb06436d6fd14436b and 0aa8a01d04c8fe200b7a106878eebc3d0af9105c add extra functions in the same area. Adjacent lines conflict. - StartupDecodingContext: again, upstream PG 45fdc9738b36d1068d3ad8fdb06436d6fd14436b and 0aa8a01d04c8fe200b7a106878eebc3d0af9105c add code where YB master 459e6b6 adds code. Adjacent lines conflict. - yb_schema_change_cb_wrapper: (same) - proto.c: - function declarations: upstream PG 52e4f0cd472d39d07732b99559989ea3b615be78 and 923def9a533a7d986acfb524139d8b9e5466d0a5 change the signature of logicalrep_write_tuple. YB master fddfec6 also changes signature. Apply both. - logicalrep_write_insert: (same) - logicalrep_write_update signature: upstream PG 464824323e57dc4b397e8b05854d779908b55304, 52e4f0cd472d39d07732b99559989ea3b615be78, and 923def9a533a7d986acfb524139d8b9e5466d0a5 change the signature. YB master fddfec6 also changes the signature. Apply both. - logicalrep_write_update body: (same as logicalrep_write_insert) - logicalrep_write_delete: (same) - logicalrep_write_tuple: (same) - yb_virtual_wal_client.c: - YBCReadRecord: YB master 71ad62a adds publication_names parameter while YB pg15 merge 417e9b3 removes RecPtr parameter. Apply both. - pgoutput.c: - _PG_output_plugin_init: upstream PG 464824323e57dc4b397e8b05854d779908b55304 and ac4645c0157fc5fcef0af8ff571512aa284a2cec add code where YB master 459e6b6 adds code. Adjacent lines conflict. - pgoutput_change case REORDER_BUFFER_CHANGE_UPDATE: upstream PG 83fd4532a72179c370e318075a10e0e2aa832024 and 52e4f0cd472d39d07732b99559989ea3b615be78 significantly change the code. YB master fddfec6 adds code to determine yb_old_is_omitted and yb_new_is_omitted, and it passes them to the logicalrep_write_update call. Major conflict. - slot.c: - ReplicationSlotAcquire: major conflict: many commits appear to be involved on both sides. Follow the resolution in YB pg15 merge ee08bd6 to change "slot" to "s" and take YB master's version. - walsender.c: - XLogSendLogical: (same as yb_virtual_wal_client.c YBCReadRecord) - pg_yb_utils.c: - YbGetDdlMode T_VacuumStmt: YB pg15 0a5cbae and YB master 9a27aff add logic in the same area. Keep both. - logicalproto.h: - logicalrep_write_update: (same as proto.c logicalrep_write_update signature) - output_plugin.h: - YBLogicalDecodeSchemaChangeCB: (same as logical.c function declarations) - OutputPluginCallbacks: (same as logical.c function declarations) - reorderbuffer.h: - YBReorderBufferSchemaChangeCB: (same as logical.c function declarations) - yb_virtual_wal_client.h: - YBCReadRecord: (same as yb_virtual_wal_client.c YBCReadRecord) - aggregates.sql: - YB master b2f8822 is an upstream PG point import of 44e95b5728a4569c494fa4ea4317f8a2f50a206b, which is already part of upstream PG REL_15_2. Drop incoming YB master changes because YB pg15 is already up to date. - aggregates.out: - (same as aggregates.sql) - yb_decode.c: - YBDecodeUpdate: like YB pg15 merge d68d96c, pass false for the new argument of the new ReorderBufferQueueChange call. - build_postgres.py: - make_postgres: add new extensions yb_ycql_utils (YB master ce7d880) and yb_xcluster_ddl_replication (YB master 0185d16) to the list of directories to build. YB pg15 initial merge disabled building all contrib extensions automatically, so each extension has to be added to this list for now. - yb_xcluster_ddl_replication.c: - HandleSourceDDLEnd: this file is introduced by YB master 0185d16. In various places, adjust code according to upstream PG 2f9661311b83dc481fc19f6e3bda015392010a40, which makes trig_data->tag of CommandTag type rather than char*. Use GetCommandTagName where str is needed and switch to comparing CommandTag to CommandTag instead of str to str where possible. - IsInIgnoreList: (same) Test Plan: On Almalinux 8: #!/usr/bin/env bash set -eu ./yb_build.sh fastdebug --gcc11 pg15_tests/run_tests.sh Get the following results using pg15_tests/run_tests.sh: 0 2024-05-13T14:05:03-07:00 JAVA org.yb.pgsql.TestPgViewYCQLStats 0 2024-05-13T14:05:33-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestVWALConsumptionWitDDLStatements 0 2024-05-13T14:06:05-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestVWALConsumptionWithMultipleAlter 0 2024-05-13T14:06:35-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestVWALConsumptionWitDDLStatementsAndRestart 0 2024-05-13T14:26:41-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.* 0 2024-05-13T14:27:32-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestTabletSplitDuringConsumptionFromVWAL 0 2024-05-13T14:28:38-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestRecordCountsAfterMultipleTabletSplits 0 2024-05-13T14:29:25-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestTabletSplitDuringConsumptionFromVWALWithRestart 0 2024-05-13T14:30:07-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestTabletSplitDuringConsumptionFromVWALWithRestartOnPartialAck 0 2024-05-13T14:30:33-07:00 integration-tests_master_path_handlers-itest MasterPathHandlersItest.TestHiddenSplitParentTablet 1 2024-05-13T14:39:24-07:00 JAVA org.yb.pgsql.TestPgWaitQueuesRegress 0 2024-05-13T14:39:46-07:00 integration-tests_xcluster_ysql_index-test XClusterDbScopedYsqlIndexTest.IndexCheckpointLocation 0 2024-05-13T14:40:43-07:00 integration-tests_minicluster-snapshot-test Colocation/MasterExportSnapshotTest.ExportSnapshotAsOfTimeWithHiddenTables/0 0 2024-05-13T14:41:32-07:00 integration-tests_minicluster-snapshot-test Colocation/MasterExportSnapshotTest.ExportSnapshotAsOfTimeWithHiddenTables/1 0 2024-05-13T14:49:45-07:00 integration-tests_wait_states-itest * 1 2024-05-13T15:14:59-07:00 JAVA org.yb.pgsql.TestPgRegressIndex 0 2024-05-13T15:15:38-07:00 JAVA org.yb.pgsql.TestPgRegressThirdPartyExtensionsYBXClusterDDLReplication 0 2024-05-13T15:16:17-07:00 integration-tests_xcluster_ddl_replication-test XClusterDDLReplicationTest.CreateTable 0 2024-05-13T15:16:35-07:00 pgwrapper_pg_mini-test PgMiniTest.SmallParallelScan 1 2024-05-13T15:17:54-07:00 JAVA org.yb.pgsql.TestPgRegressMisc#testPgRegressMiscIndependent 0 2024-05-13T15:18:25-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestDynamicTablesAddition 0 2024-05-13T15:18:55-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestDynamicTablesRemoval 0 2024-05-13T15:19:15-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestCommitTimeTieWithoutSeparateResponseForPublicationRefreshNotification 0 2024-05-13T15:19:35-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestCommitTimeTieWithSeparateResponseForPublicationRefreshNotification 0 2024-05-13T15:19:58-07:00 integration-tests_cdcsdk_consumption_consistent_changes-test CDCSDKConsumptionConsistentChangesTest.TestLSNDeterminismWithSpecialRecordOnRestartWithPartialAck 0 2024-05-13T15:22:29-07:00 integration-tests_xcluster_db_scoped-test XClusterDBScopedTest.TestCreateWithCheckpoint 0 2024-05-13T15:23:16-07:00 integration-tests_xcluster_db_scoped-test XClusterDBScopedTest.ColocatedDB Also run some tests that enable YSQL Connection Manager: #./yb_build.sh fastdebug --gcc11 --enable-ysql-conn-mgr-test --java-test 'org.yb.pgsql.TestPgYbStat#testYbTerminatedQueriesMultipleCauses' ./yb_build.sh fastdebug --gcc11 --enable-ysql-conn-mgr-test --java-test 'org.yb.pgsql.TestPgAlterTable' ./yb_build.sh fastdebug --gcc11 --enable-ysql-conn-mgr-test --java-test 'org.yb.pgsql.TestPgRegressReplicaIdentity#testPgRegressTable' ./yb_build.sh fastdebug --gcc11 --enable-ysql-conn-mgr-test --java-test 'org.yb.pgsql.TestPgConnection' ./yb_build.sh fastdebug --gcc11 --enable-ysql-conn-mgr-test --java-test 'org.yb.pgsql.TestYbPgStatActivity#testMemUsageFuncsWithMultipleBackends' ./yb_build.sh fastdebug --gcc11 --enable-ysql-conn-mgr-test --java-test 'org.yb.pgsql.TestPgBackendMemoryContext#testLogBackendMemoryTexts' ./yb_build.sh fastdebug --gcc11 --enable-ysql-conn-mgr-test --java-test 'org.yb.ysqlconnmgr.TestMisc#testNegSetYsqlConnMgr' ./yb_build.sh fastdebug --gcc11 --enable-ysql-conn-mgr-test --java-test 'org.yb.pgsql.TestPgRegressPublication#testPgRegressPublication' ./yb_build.sh fastdebug --gcc11 --enable-ysql-conn-mgr-test --java-test 'org.yb.pgsql.TestPgUniqueConstraint#createIndexViolatingUniqueness' C++ test ClientTest.TestCreateTableWithRangePartition may be fixed by merging over YB master 84c4ae5. Jenkins: rebase: pg15 Reviewers: tfoucher Reviewed By: tfoucher Subscribers: fizaa, aagrawal, yql Differential Revision: https://phorge.dev.yugabyte.com/D34968
ZhenYongFan
pushed a commit
to ZhenYongFan/yugabyte-db
that referenced
this issue
Jun 15, 2024
…mpilation Summary: Commit 061dc69 introduces code causing compilation failure ../../src/yb/integration-tests/cdcsdk_ysql_test_base.cc: In member function ‘void yb::cdc::CDCSDKYsqlTest::CheckRecordsConsistencyFromVWAL(const std::vector<yb::cdc::CDCSDKProtoRecordPB>&)’: ../../src/yb/integration-tests/cdcsdk_ysql_test_base.cc:3879:9: error: ‘prev_op’ may be used uninitialized in this function [-Werror=maybe-uninitialized] 3879 | if (prev_op == RowMessage::DDL) { | ^~ cc1plus: all warnings being treated as errors Almalinux 8 + fastdebug + gcc11 is affected, but other build combos such as those in jenkins_jobs.yml are fine. Fix the issue by initializing prev_op to UNKNOWN. This should be fine according to the logic as far as I can tell. Jira: DB-10712 Test Plan: On Almalinux 8: ./yb_build.sh fastdebug --gcc11 Close: yugabyte#21811 Original commit: 193589f / D33830 Reviewers: siddharth.shah Reviewed By: siddharth.shah Tags: #jenkins-ready Differential Revision: https://phorge.dev.yugabyte.com/D34494
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Jira Link: DB-10712
Description
Almalinux 8, fastdebug, gcc11 compilation fails on master 8239595.
Likely due to commit 061dc69.
Issue Type
kind/bug
Warning: Please confirm that this issue does not contain any sensitive information
The text was updated successfully, but these errors were encountered: