Package "libpq-dev"

  postgresql-10 (10.19-0ubuntu0.18.04.1) bionic-security; urgency=medium

  * New upstream version (LP: #1950268).

    + Make the server reject extraneous data after an SSL or GSS
      encryption handshake
      CVE-2021-23214

    + Make libpq reject extraneous data after an SSL or GSS
      encryption handshake
      CVE-2021-23222

    + A dump/restore is not required for those running 10.X.

    + However, note that installations using physical replication should
      update standby servers before the primary server, details in the
      release notes linked below.

    + Also, several bugs have been found that may have resulted in corrupted
      indexes, explained in detail in the release notes linked below. If any
      of those cases apply to you, it's recommended to reindex
      possibly-affected indexes after updating.

    + Also, if you are upgrading from a version earlier than 10.16,
      see those release notes as well please.

    + Details about these and many further changes can be found at:
      https://www.postgresql.org/docs/10/release-10-19.html

 -- Christian Ehrhardt <email address hidden> Tue, 09 Nov 2021 09:39:50 +0100

  postgresql-10 (10.18-0ubuntu0.18.04.1) bionic-security; urgency=medium

  * New upstream version (LP: #1939396).

    + Disallow SSL renegotiation more completely (Michael Paquier)

      SSL renegotiation has been disabled for some time, but the server
      would still cooperate with a client-initiated renegotiation request.
      A maliciously crafted renegotiation request could result in a server
      crash (see OpenSSL issue CVE-2021-3449). Disable the feature
      altogether on OpenSSL versions that permit doing so, which are
      1.1.0h and newer.
      (CVE-2021-3449)

    + Details about these and many further changes can be found at:
      https://www.postgresql.org/docs/10/release-10-18.html

    + d/p/reproducible-bki: refreshed for 10.18

 -- Christian Ehrhardt <email address hidden> Tue, 10 Aug 2021 14:18:35 +0200

  postgresql-10 (10.17-0ubuntu0.18.04.1) bionic-security; urgency=medium

  * New upstream version (LP: #1928773).

    + Prevent integer overflows in array subscripting calculations (Tom Lane)

      The array code previously did not complain about cases where an array's
      lower bound plus length overflows an integer. This resulted in later
      entries in the array becoming inaccessible (since their subscripts could
      not be written as integers), but more importantly it confused subsequent
      assignment operations. This could lead to memory overwrites, with
      ensuing crashes or unwanted data modifications. (CVE-2021-32027)

    + Fix mishandling of junk columns in INSERT ... ON CONFLICT ... UPDATE
      target lists (Tom Lane)

      If the UPDATE list contains any multi-column sub-selects (which give
      rise to junk columns in addition to the results proper), the UPDATE path
      would end up storing tuples that include the values of the extra junk
      columns. That's fairly harmless in the short run, but if new columns are
      added to the table then the values would become accessible, possibly
      leading to malfunctions if they don't match the datatypes of the added
      columns.

      In addition, in versions supporting cross-partition updates, a
      cross-partition update triggered by such a case had the reverse problem:
      the junk columns were removed from the target list, typically causing an
      immediate crash due to malfunction of the multi-column sub-select
      mechanism. (CVE-2021-32028)

    + Details about these and many further changes can be found at:
      https://www.postgresql.org/docs/10/release-10-17.html

 -- Christian Ehrhardt <email address hidden> Tue, 18 May 2021 12:17:37 +0200

  postgresql-10 (10.16-0ubuntu0.18.04.1) bionic; urgency=medium

  * New upstream version (LP: #1915254)
    + Fix CREATE INDEX CONCURRENTLY to wait for
      concurrent prepared transactions (Andrey Borodin)

      At the point where CREATE INDEX CONCURRENTLY
      waits for all concurrent transactions to complete so that it can see
      rows they inserted, it must also wait for all prepared transactions
      to complete, for the same reason. Its failure to do so meant that
      rows inserted by prepared transactions might be omitted from the new
      index, causing queries relying on the index to miss such rows.
      In installations that have enabled prepared transactions
      (max_prepared_transactions > 0),
      it's recommended to reindex any concurrently-built indexes in
      case this problem occurred when they were built.

    + Details about this and many further changes can be found at:
      https://www.postgresql.org/docs/10/static/release-10-16.html

 -- Christian Ehrhardt <email address hidden> Wed, 10 Feb 2021 11:47:35 +0100

  postgresql-10 (10.15-0ubuntu0.18.04.1) bionic-security; urgency=medium

  * New upstream version.
    + Block DECLARE CURSOR ... WITH HOLD and firing of deferred triggers
      within index expressions and materialized view queries (Noah Misch)

      This is essentially a leak in the security restricted operation sandbox
      mechanism. An attacker having permission to create non-temporary SQL
      objects could parlay this leak to execute arbitrary SQL code as a
      superuser.

      The PostgreSQL Project thanks Etienne Stalmans for reporting this
      problem. (CVE-2020-25695)

    + Fix usage of complex connection-string parameters in pg_dump,
      pg_restore, clusterdb, reindexdb, and vacuumdb (Tom Lane)

      The -d parameter of pg_dump and pg_restore, or the --maintenance-db
      parameter of the other programs mentioned, can be a connection string
      containing multiple connection parameters rather than just a database
      name. In cases where these programs need to initiate additional
      connections, such as parallel processing or processing of multiple
      databases, the connection string was forgotten and just the basic
      connection parameters (database name, host, port, and username) were
      used for the additional connections. This could lead to connection
      failures if the connection string included any other essential
      information, such as non-default SSL or GSS parameters. Worse, the
      connection might succeed but not be encrypted as intended, or be
      vulnerable to man-in-the-middle attacks that the intended connection
      parameters would have prevented. (CVE-2020-25694)

    + When psql's \connect command re-uses connection parameters, ensure that
      all non-overridden parameters from a previous connection string are
      re-used (Tom Lane)

      This avoids cases where reconnection might fail due to omission of
      relevant parameters, such as non-default SSL or GSS options. Worse, the
      reconnection might succeed but not be encrypted as intended, or be
      vulnerable to man-in-the-middle attacks that the intended connection
      parameters would have prevented. This is largely the same problem as
      just cited for pg_dump et al, although psql's behavior is more complex
      since the user may intentionally override some connection parameters.
      (CVE-2020-25694)

    + Prevent psql's \gset command from modifying specially-treated variables
      (Noah Misch)

      \gset without a prefix would overwrite whatever variables the server
      told it to. Thus, a compromised server could set specially-treated
      variables such as PROMPT1, giving the ability to execute arbitrary shell
      code in the user's session.

      The PostgreSQL Project thanks Nick Cleaton for reporting this problem.
      (CVE-2020-25696)

    + Details about these and many further changes can be found at:
      https://www.postgresql.org/docs/10/static/release-10-15.html

 -- Christian Ehrhardt <email address hidden> Thu, 12 Nov 2020 12:13:22 +0100