diff options
| author | Heikki Linnakangas <heikki.linnakangas@iki.fi> | 2024-02-05 11:01:30 +0200 |
|---|---|---|
| committer | Heikki Linnakangas <heikki.linnakangas@iki.fi> | 2024-02-05 11:03:43 +0200 |
| commit | a45c950ae35e53de4007c9dd0991d5005a4bee7b (patch) | |
| tree | b578e715712036aeb5ba4eecd4e9d6c60cfc4ac1 /src/backend/commands/matview.c | |
| parent | f4f2883521fc3f81765ab82eab2ffa31574f0a07 (diff) | |
Fix assertion if index is dropped during REFRESH CONCURRENTLY
When assertions are disabled, the built SQL statement is invalid and
you get a "syntax error". So this isn't a serious problem, but let's
avoid the assertion failure.
Backpatch to all supported versions.
Reviewed-by: Noah Misch
Diffstat (limited to 'src/backend/commands/matview.c')
| -rw-r--r-- | src/backend/commands/matview.c | 7 |
1 files changed, 5 insertions, 2 deletions
diff --git a/src/backend/commands/matview.c b/src/backend/commands/matview.c index 0aa0a749dab..0ab8ba1a58d 100644 --- a/src/backend/commands/matview.c +++ b/src/backend/commands/matview.c @@ -803,9 +803,12 @@ refresh_by_match_merge(Oid matviewOid, Oid tempOid, Oid relowner, * * ExecRefreshMatView() checks that after taking the exclusive lock on the * matview. So at least one unique index is guaranteed to exist here - * because the lock is still being held; so an Assert seems sufficient. + * because the lock is still being held. (One known exception is if a + * function called as part of refreshing the matview drops the index. + * That's a pretty silly thing to do.) */ - Assert(foundUniqueIndex); + if (!foundUniqueIndex) + elog(ERROR, "could not find suitable unique index on materialized view"); appendStringInfoString(&querybuf, " AND newdata.* OPERATOR(pg_catalog.*=) mv.*) " |