Re: [PATCH] REPLICA IDENTITY USING INDEX accepts column with invalid NOT NULL
Re: [PATCH] REPLICA IDENTITY USING INDEX accepts column with invalid NOT NULL
Hi all,
I've rebased and updated my patch based on Álvaro Herrera's wording suggestion.
I have left the function refactoring for future work as suggested.
I have left the function refactoring for future work as suggested.
Thanks,
On Mon, Jun 15, 2026 at 8:44 PM Álvaro Herrera <alvherre@kurilemu.de> wrote:
On 2026-May-14, Ante Krešić wrote:
> + if (!conForm->convalidated)
> + ereport(ERROR,
> + errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
> + errmsg("index \"%s\" cannot be used as replica identity because column \"%s\" has an invalid not-null constraint",
> + RelationGetRelationName(indexRel),
> + NameStr(attr->attname)),
> + /*- translator: second %s is a constraint characteristic such as NOT VALID */
> + errdetail("The constraint \"%s\" is marked %s.",
> + NameStr(conForm->conname), "NOT VALID"),
> + errhint("You might need to validate it using %s.",
> + "ALTER TABLE ... VALIDATE CONSTRAINT"));
I'd rather make this error be
+ ereport(ERROR,
+ errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("cannot use index \"%s\" as replica identity",
+ RelationGetRelationName(indexRel),
+ NameStr(attr->attname)),
+ /*- translator: third %s is a constraint characteristic such as NOT VALID */
+ errdetail("The constraint \"%s\" on column \"%s\" is marked %s.",
+ NameStr(conForm->conname), NameStr(attr->attname), "NOT VALID"),
+ errhint("You might need to validate it using %s.",
+ "ALTER TABLE ... VALIDATE CONSTRAINT"));
The other ones which you're mimicking are IMO in poor style.
--
Álvaro Herrera Breisgau, Deutschland — https://www.EnterpriseDB.com/
Officer Krupke, what are we to do?
Gee, officer Krupke, Krup you! (West Side Story, "Gee, Officer Krupke")
Re: [PATCH] REPLICA IDENTITY USING INDEX accepts column with invalid NOT NULL
От:
solai v <solai.cdac@gmail.com>
Дата:
Hi, I tested the patch on my current branch. Before applying the patch i was able to reproduce the reported issue. I created a table containing a NULL value, added a NOT NULL NOT VALID constraint on the column, and created a unique index. PostgreSQL accepted ALTER TABLE REPLICA IDENTITY USING INDEX even though the underlying NOT NULL constraint had not been validated and the column still contained NULL values. The command succeeded and relreplident was set to i, confirming the behavior described in the report. After applying the patch i repeated the same test. PostgreSQL correctly rejected the operation with ERROR: cannot use index "t_idx" as replica identity DETAIL: The constraint "id_nn" on column "id" is marked NOT VALID. HINT: You might need to validate it using ALTER TABLE ... VALIDATE CONSTRAINT. This matches the expected behavior, since a column with an unvalidated NOT NULL constraint cannot guarantee that all rows are non-null and therefore should not be eligible for use as a replica identity. I also ran the regression tests, and all tests passed successfully. Regards, Solai
pgsql: pgcrypto: Add option to revert to prior decryption behavior
От:
Noah Misch <noah@leadboat.com>
Дата:
pgcrypto: Add option to revert to prior decryption behavior The previous commit raises an ERROR during PGP operations if OpenSSL does not support the cipher in use. However, any existing messages created with faulty encryption will no longer be accessible via pgp_[sym|pub]_decrypt(). To help users out of this situation, add a new ignore-cipher-failure option which reverts to the broken behavior during decryption only. A faulty encryption wrapper, created by an OpenSSL configuration that does not support the cipher, can then be stripped back off by that same OpenSSL in order to safely reencrypt it. (Note that when OpenSSL does support the cipher, corrupted messages will not be decrypted regardless of the ignore-cipher-failure setting; this is unchanged.) The new tests add a corrupted Blowfish message for both public- and symmetric-key decryption, resulting in the following test matrix: - Blowfish supported, default behavior: fails to decrypt - Blowfish supported, ignore-cipher-failure: fails to decrypt - Blowfish unsupported, default behavior: fails to load cipher - Blowfish unsupported, ignore-cipher-failure: strips faulty encryption The previous commit's change to the pubkey tests is expanded similarly: correctly encrypted messages cannot be decrypted by an OpenSSL that does not support the cipher, regardless of the option's setting, though the failure mode will change. Suggested-by: Noah Misch Reviewed-by: Daniel Gustafsson Reviewed-by: Noah Misch Security: CVE-2026-14663 Backpatch-through: 14 Branch ------ REL_17_STABLE Details ------- https://git.postgresql.org/pg/commitdiff/7eed42aa88053ce9319f249fc70a46181c026a4b Author: Jacob Champion Modified Files -------------- contrib/pgcrypto/expected/pgp-decrypt.out | 26 ++++++++++++++++++ contrib/pgcrypto/expected/pgp-decrypt_1.out | 30 ++++++++++++++++++++ contrib/pgcrypto/expected/pgp-info.out | 3 +- contrib/pgcrypto/expected/pgp-pubkey-decrypt.out | 30 ++++++++++++++++++++ contrib/pgcrypto/expected/pgp-pubkey-decrypt_1.out | 30 ++++++++++++++++++++ contrib/pgcrypto/pgp-cfb.c | 21 ++++++++++---- contrib/pgcrypto/pgp-decrypt.c | 9 ++++-- contrib/pgcrypto/pgp-encrypt.c | 6 ++-- contrib/pgcrypto/pgp-pgsql.c | 2 ++ contrib/pgcrypto/pgp-pubkey.c | 9 +++++- contrib/pgcrypto/pgp.c | 9 ++++++ contrib/pgcrypto/pgp.h | 7 ++++- contrib/pgcrypto/sql/pgp-decrypt.sql | 26 ++++++++++++++++++ contrib/pgcrypto/sql/pgp-pubkey-decrypt.sql | 27 ++++++++++++++++++ doc/src/sgml/pgcrypto.sgml | 32 ++++++++++++++++++++++ 15 files changed, 254 insertions(+), 13 deletions(-)