From 332b43063351758055f68c6227444e6003dba769 Mon Sep 17 00:00:00 2001 From: Amit Kapila Date: Mon, 11 Dec 2023 08:23:33 +0530 Subject: [PATCH] Fix an undetected deadlock due to apply worker. The apply worker needs to update the state of the subscription tables to 'READY' during the synchronization phase which requires locking the corresponding subscription. The apply worker also waits for the subscription tables to reach the 'SYNCDONE' state after holding the locks on the subscription and the wait is done using WaitLatch. The 'SYNCDONE' state is changed by tablesync workers again by locking the corresponding subscription. Both the state updates use AccessShareLock mode to lock the subscription, so they can't block each other. However, a backend can simultaneously try to acquire a lock on the same subscription using AccessExclusiveLock mode to alter the subscription. Now, the backend's wait on a lock can sneak in between the apply worker and table sync worker causing deadlock. In other words, apply_worker waits for tablesync worker which waits for backend, and backend waits for apply worker. This is not detected by the deadlock detector because apply worker uses WaitLatch. The fix is to release existing locks in apply worker before it starts to wait for tablesync worker to change the state. Reported-by: Tomas Vondra Author: Shlok Kyal Reviewed-by: Amit Kapila, Peter Smith Backpatch-through: 12 Discussion: https://postgr.es/m/d291bb50-12c4-e8af-2af2-7bb9bb4d8e3e@enterprisedb.com --- src/backend/replication/logical/tablesync.c | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) diff --git a/src/backend/replication/logical/tablesync.c b/src/backend/replication/logical/tablesync.c index 9d6699f4d31..c3dd902f8c8 100644 --- a/src/backend/replication/logical/tablesync.c +++ b/src/backend/replication/logical/tablesync.c @@ -518,15 +518,25 @@ process_syncing_tables_for_apply(XLogRecPtr current_lsn) /* Now safe to release the LWLock */ LWLockRelease(LogicalRepWorkerLock); + if (started_tx) + { + /* + * We must commit the existing transaction to release + * the existing locks before entering a busy loop. + * This is required to avoid any undetected deadlocks + * due to any existing lock as deadlock detector won't + * be able to detect the waits on the latch. + */ + CommitTransactionCommand(); + pgstat_report_stat(false); + } + /* * Enter busy loop and wait for synchronization worker to * reach expected state (or die trying). */ - if (!started_tx) - { - StartTransactionCommand(); - started_tx = true; - } + StartTransactionCommand(); + started_tx = true; wait_for_relation_state_change(rstate->relid, SUBREL_STATE_SYNCDONE);