diff options
| author | Tom Lane <tgl@sss.pgh.pa.us> | 2022-11-19 11:40:30 -0500 |
|---|---|---|
| committer | Tom Lane <tgl@sss.pgh.pa.us> | 2022-11-19 11:40:30 -0500 |
| commit | 55f30e6c7640c80fbb7be90fad668ee4bffa0e97 (patch) | |
| tree | 5a0fc123adc1d9410e6e861151604d936e7f4a66 /src/backend/executor/nodeHashjoin.c | |
| parent | 038512f71ea807d6c2b8e519f2416a9a5283d9bf (diff) | |
| download | postgresql-55f30e6c7640c80fbb7be90fad668ee4bffa0e97.tar.gz postgresql-55f30e6c7640c80fbb7be90fad668ee4bffa0e97.zip | |
Postpone calls of unsafe server-side functions in pg_dump.
Avoid calling pg_get_partkeydef(), pg_get_expr(relpartbound),
and regtypeout until we have lock on the relevant tables.
The existing coding is at serious risk of failure if there
are any concurrent DROP TABLE commands going on --- including
drops of other sessions' temp tables.
Back-patch of commit e3fcbbd62. That's been in v15/HEAD long enough
to have some confidence about it, so now let's fix the problem in
older branches.
Original patch by me; thanks to Gilles Darold for back-patching
legwork.
Discussion: https://postgr.es/m/2273648.1634764485@sss.pgh.pa.us
Discussion: https://postgr.es/m/7d7eb6128f40401d81b3b7a898b6b4de@W2012-02.nidsa.loc
Discussion: https://postgr.es/m/45c93d57-9973-248e-d2df-e02ca9af48d4@darold.net
Diffstat (limited to 'src/backend/executor/nodeHashjoin.c')
0 files changed, 0 insertions, 0 deletions