Commits on Aug 21, 2026
-
Fix replication slot leak on error caught in a subtransaction.
The SQL-callable replication slot functions, such as pg_replication_slot_advance(), acquire a slot and release it before returning. If one throws an error in between, the slot is released by the top-level error handler (the sigsetjmp block in PostgresMain()). But PL/pgSQL, PL/Perl, PL/Python and PL/Tcl run an error-handling block in an internal subtransaction: when an error is raised there, they abort the subtransaction and, if a handler matches, catch the error without re-throwing it. The top-level handler is then never reached, and nothing in the (sub)transaction abort path releases the slot either. The slot is therefore left acquired after the caught error. The next slot operation in the session then fails an assertion, or in a non-assertion build silently overwrites the reference and leaks the slot, which keeps holding back WAL removal and the catalog xmin (blocking vacuum) and can no longer be acquired. An error that does reach the top-level handler is unaffected. That is the case for a plain ROLLBACK TO SAVEPOINT, or for a PL/pgSQL exception handler whose conditions do not match the error, which then re-throws it. Fix by recording the subtransaction that acquires the slot and releasing the slot when that subtransaction aborts, the same way other subtransaction-scoped resources are released, for example in AtEOSubXact_LargeObject() and AtEOSubXact_Files(). The slot cannot simply be released on every subtransaction abort, because logical decoding starts and aborts an internal transaction or subtransaction for each decoded transaction while holding the slot. Those internal aborts are below the acquiring subtransaction and so are left alone. Releasing a temporary slot does not drop it, so it would keep holding resources until the session ends. To avoid that, when releasing the held slot on abort also drop the session's temporary slots, the same way the top-level error handler already does. Reported-by: Satya Narlapuram <satyanarlapuram@gmail.com> Author: Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com> Co-authored-by: Satya Narlapuram <satyanarlapuram@gmail.com> Co-authored-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Fujii Masao <masao.fujii@gmail.com> Reviewed-by: shveta malik <shveta.malik@gmail.com> Reviewed-by: Hou Zhijie <houzj.fnst@fujitsu.com> Reviewed-by: Kyotaro Horiguchi <horikyota.ntt@gmail.com> Reviewed-by: Ashutosh Sharma <ashu.coek88@gmail.com> Discussion: https://postgr.es/m/CAHg+QDeuf9tCq3ce=kgFMJP0m=PZC+wi6B=yS+7V0vNXjLS31w@mail.gmail.com Backpatch-through: 14
3 people authored and committed
Aug 21, 2026
