On Wed, Dec 3, 2025 at 4:47 PM tianbing <[email protected]> wrote: > > Hi， Peter， > I have reviewed the v21 patch and noticed that there seems to be a memory > leak. > > +static bool > +check_publication_exists(PGconn *conn, const char *pubname, const char > *dbname) > +{ > + PGresult *res; > + bool exists; > + char *query; > + > + query = psprintf("SELECT 1 FROM pg_publication WHERE pubname = %s", > + PQescapeLiteral(conn, pubname, strlen(pubname))); > + res = PQexec(conn, query); > + > + if (PQresultStatus(res) != PGRES_TUPLES_OK) > + pg_fatal("could not check for publication \"%s\" in database \"%s\": %s", > + pubname, dbname, PQerrorMessage(conn)); > + > + exists = (PQntuples(res) == 1); > + > + PQclear(res); > + pg_free(query); > + return exists; > +} > > The PQescapeLiteral() function through malloc to allocate memmory，and should > be free by PQfreemem。 > > I suggest making the following modifications： > > + char *pub = PQescapeLiteral(conn, pubname, strlen(pubname); > + query = psprintf("SELECT 1 FROM pg_publication WHERE pubname = %s", pub); > ...... > + PQfreemem(pub); >
Fixed in v22. ====== Kind Regards, Peter Smith. Fujitsu Australia
v22-0001-Support-existing-publications-in-pg_createsubscr.patch
Description: Binary data
