lp:ubuntu/oneiric-proposed/postgresql-9.1

Branch merges

Propose for merging

No branches dependent on this one.

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

15. By Martin Pitt on 2012-12-10

* New upstream bug fix release: (LP: #1088393)
  - Fix multiple bugs associated with "CREATE INDEX CONCURRENTLY".
    Fix "CREATE INDEX CONCURRENTLY" to use in-place updates when
    changing the state of an index's pg_index row. This prevents race
    conditions that could cause concurrent sessions to miss updating
    the target index, thus resulting in corrupt concurrently-created
    indexes.
    Also, fix various other operations to ensure that they ignore
    invalid indexes resulting from a failed "CREATE INDEX CONCURRENTLY"
    command. The most important of these is "VACUUM", because an
    auto-vacuum could easily be launched on the table before corrective
    action can be taken to fix or remove the invalid index.
  - Fix buffer locking during WAL replay.
    The WAL replay code was insufficiently careful about locking
    buffers when replaying WAL records that affect more than one page.
    This could result in hot standby queries transiently seeing
    inconsistent states, resulting in wrong answers or unexpected
    failures.
  - See HISTORY/changelog.gz for the other bug fixes.

14. By Martin Pitt on 2012-09-25

* New upstream bug fix release: (LP: #1055944)
  - Fix persistence marking of shared buffers during WAL replay.
    This mistake can result in buffers not being written out during
    checkpoints, resulting in data corruption if the server later
    crashes without ever having written those buffers. Corruption can
    occur on any server following crash recovery, but it is
    significantly more likely to occur on standby slave servers since
    those perform much more WAL replay. There is a low probability of
    corruption of btree and GIN indexes. There is a much higher
    probability of corruption of table "visibility maps". Fortunately,
    visibility maps are non-critical data in 9.1, so the worst
    consequence of such corruption in 9.1 installations is transient
    inefficiency of vacuuming. Table data proper cannot be corrupted by
    this bug.
    While no index corruption due to this bug is known to have occurred
    in the field, as a precautionary measure it is recommended that
    production installations "REINDEX" all btree and GIN indexes at a
    convenient time after upgrading to 9.1.6.
    Also, if you intend to do an in-place upgrade to 9.2.X, before
    doing so it is recommended to perform a "VACUUM" of all tables
    while having vacuum_freeze_table_age set to zero. This will ensure
    that any lingering wrong data in the visibility maps is corrected
    before 9.2.X can depend on it. vacuum_cost_delay can be adjusted to
    reduce the performance impact of vacuuming, while causing it to
    take longer to finish.
  - See HISTORY/changelog.gz for the other bug fixes.

13. By Martin Pitt on 2012-01-05

Add 00git_inet_cidr_unpack.patch: Revert the behavior of inet/cidr
functions to not unpack the arguments. This fixes the memory leak when
sorting inet values. Patch taken from upstream git HEAD. Spotted during
testing in LP #904631.

12. By Martin Pitt on 2011-12-16

01-armel-tas.patch: Turn slock_t datatype into an int, and define
S_UNLOCK() to call __sync_lock_release() instead of using the default
implementation. This complies to the gcc built-in atomic operations
specifiction more strictly and now also works on the Panda boards.
(LP: #904828)

11. By Martin Pitt on 2011-12-15

* New upstream release (LP: #904631):
  - Fix bugs in information_schema.referential_constraints view.
    This view was being insufficiently careful about matching the
    foreign-key constraint to the depended-on primary or unique key
    constraint. That could result in failure to show a foreign key
    constraint at all, or showing it multiple times, or claiming that
    it depends on a different constraint than the one it really does.
    Since the view definition is installed by initdb, merely upgrading
    will not fix the problem. If you need to fix this in an existing
    installation, you can (as a superuser) drop the information_schema
    schema then re-create it by sourcing
    "SHAREDIR/information_schema.sql". (Run pg_config --sharedir if
    you're uncertain where "SHAREDIR" is.) This must be repeated in
    each database to be fixed.
  - Make "contrib/citext"'s upgrade script fix collations of citext
    columns and indexes.
    Existing citext columns and indexes aren't correctly marked as
    being of a collatable data type during pg_upgrade from a pre-9.1
    server. That leads to operations on them failing with errors such
    as "could not determine which collation to use for string
    comparison". This change allows them to be fixed by the same script
    that upgrades the citext module into a proper 9.1 extension during
    CREATE EXTENSION citext FROM unpackaged.
    If you have a previously-upgraded database that is suffering from
    this problem, and you already ran the "CREATE EXTENSION" command,
    you can manually run (as superuser) the "UPDATE" commands found at
    the end of "SHAREDIR/extension/citext--unpackaged--1.0.sql". (Run
    pg_config --sharedir if you're uncertain where "SHAREDIR" is.)
  - Fix possible crash during "UPDATE" or "DELETE" that joins to the
    output of a scalar-returning function.
  - Fix incorrect replay of WAL records for GIN index updates.
  - Fix TOAST-related data corruption during CREATE TABLE dest AS
    SELECT - FROM src or INSERT INTO dest SELECT * FROM src.
  - Fix possible failures during hot standby startup.
  - Start hot standby faster when initial snapshot is incomplete.
  - Fix race condition during toast table access from stale syscache
    entries. The typical symptom was transient errors like "missing chunk
    number 0 for toast value NNNNN in pg_toast_2619", where the cited toast
    table would always belong to a system catalog.
  - Track dependencies of functions on items used in parameter default
    expressions. Previously, a referenced object could be dropped without
    having dropped or modified the function, leading to misbehavior when
    the function was used. Note that merely installing this update will not
    fix the missing dependency entries; to do that, you'd need to
    "CREATE OR REPLACE" each such function afterwards. If you have
    functions whose defaults depend on non-built-in objects, doing so
    is recommended.
  - Fix incorrect management of placeholder variables in nestloop joins.
    This bug is known to lead to "variable not found in subplan target
    list" planner errors, and could possibly result in wrong query
    output when outer joins are involved.
  - Fix window functions that sort by expressions involving aggregates.
  - Fix "MergeAppend child's targetlist doesn't match MergeAppend"
    planner errors.
  - Fix index matching for operators with both collatable and
    noncollatable inputs. In 9.1.0, an indexable operator that has a
    non-collatable left-hand input type and a collatable right-hand input
    type would not be recognized as matching the left-hand column's index.
    An example is the hstore ? text operator.
  - Allow inlining of set-returning SQL functions with multiple OUT
    parameters.
  - Don't trust deferred-unique indexes for join removal.
  - Make DatumGetInetP() unpack inet datums that have a 1-byte header,
    and add a new macro, DatumGetInetPP(), that does not.
  - Improve locale support in money type's input and output.
    Aside from not supporting all standard lc_monetary formatting
    options, the input and output functions were inconsistent, meaning
    there were locales in which dumped money values could not be
    re-read.
  - Don't let transform_null_equals affect CASE foo WHEN NULL ...
    constructs. transform_null_equals is only supposed to affect foo =
    NULL expressions written directly by the user, not equality checks
    generated internally by this form of CASE.
  - Change foreign-key trigger creation order to better support
    self-referential foreign keys.
  - Fix IF EXISTS to work correctly in "DROP OPERATOR FAMILY".
  - Disallow dropping of an extension from within its own script.
  - Don't mark auto-generated types as extension members.
  - Cope with invalid pre-existing search_path settings during "CREATE
    EXTENSION".
  - Avoid floating-point underflow while tracking buffer allocation
    rate.
  - Prevent autovacuum transactions from running in serializable mode.
    Autovacuum formerly used the cluster-wide default transaction
    isolation level, but there is no need for it to use anything higher
    than READ COMMITTED, and using SERIALIZABLE could result in
    unnecessary delays for other processes.
  - Ensure walsender processes respond promptly to SIGTERM.
  - Exclude "postmaster.opts" from base backups.
  - Fix incorrect field alignment in ecpg's SQLDA area.
  - Preserve blank lines within commands in psql's command history.
    The former behavior could cause problems if an empty line was
    removed from within a string literal, for example.
  - Avoid platform-specific infinite loop in pg_dump.
  - Fix compression of plain-text output format in pg_dump.
    pg_dump has historically understood -Z with no -F switch to mean
    that it should emit a gzip-compressed version of its plain text
    output. Restore that behavior.
  - Fix pg_dump to dump user-defined casts between auto-generated
    types, such as table rowtypes.
  - Fix missed quoting of foreign server names in pg_dump.
  - Assorted fixes for pg_upgrade. Handle exclusion constraints correctly,
    avoid failures on Windows, don't complain about mismatched toast table
    names in 8.4 databases.
  - In PL/pgSQL, allow foreign tables to define row types.
  - Fix up conversions of PL/Perl functions' results.
    Restore the pre-9.1 behavior that PL/Perl functions returning void
    ignore the result value of their last Perl statement; 9.1.0 would
    throw an error if that statement returned a reference. Also, make
    sure it works to return a string value for a composite type, so
    long as the string meets the type's input format. In addition,
    throw errors for attempts to return Perl arrays or hashes when the
    function's declared result type is not an array or composite type,
    respectively. (Pre-9.1 versions rather uselessly returned strings
    like ARRAY(0x221a9a0) or HASH(0x221aa90) in such cases.)
  - Ensure PL/Perl strings are always correctly UTF8-encoded.
  - Use the preferred version of xsubpp to build PL/Perl, not
    necessarily the operating system's main copy.
  - Correctly propagate SQLSTATE in PL/Python exceptions.
  - Do not install PL/Python extension files for Python major versions
    other than the one built against.
  - Change all the "contrib" extension script files to report a useful
    error message if they are fed to psql. This should help teach people
    about the new method of using "CREATE EXTENSION" to load these files.
    In most cases, sourcing the scripts directly would fail anyway, but
    with harder-to-interpret messages.
  - Fix incorrect coding in "contrib/dict_int" and "contrib/dict_xsyn".
  - Remove "contrib/sepgsql" tests from the regular regression test
    mechanism. Since these tests require root privileges for setup, they're
    impractical to run automatically. Switch over to a manual approach
    instead, and provide a testing script to help with that.
  - Fix assorted errors in "contrib/unaccent"'s configuration file
    parsing.
  - Honor query cancel interrupts promptly in pgstatindex().
  - Revert unintentional enabling of WAL_DEBUG. Fortunately, as debugging
    tools go, this one is pretty cheap; but it's not intended to be enabled
    by default, so revert.
  - Ensure VPATH builds properly install all server header files.
  - Shorten file names reported in verbose error messages.
    Regular builds have always reported just the name of the C file
    containing the error message call, but VPATH builds formerly
    reported an absolute path name.

10. By Martin Pitt on 2011-09-26

* New upstream bug fix release:
  - Make pg_options_to_table return NULL for an option with no value.
    Previously such cases would result in a server crash.
  - Fix memory leak at end of a GiST index scan.
    Commands that perform many separate GiST index scans, such as
    verification of a new GiST-based exclusion constraint on a table
    already containing many rows, could transiently require large
    amounts of memory due to this leak.
  - Fix explicit reference to pg_temp schema in "CREATE TEMPORARY
    TABLE". This used to be allowed, but failed in 9.1.0.

9. By Martin Pitt on 2011-09-12

* Final 9.1 release.
* 02-relax-sslkey-permscheck.patch, 50-per-version-dirs.patch: Refresh to
  apply cleanly.
* debian/control: Tighten the dependencies of the -pl* extensions/-contrib
  to postgresql-9.1 to the same binary version. (Closes: #640335)

8. By Martin Pitt on 2011-08-27

* debian/watch: Fix for mangling ~rc, thanks Peter Eisentraut.
  (Closes: #639357)
* debian/control: Add versionless Provides: to the PL* extensions, as per
  request from Christoph Berg.
* debian/control: Add "Replaces: postgresql-9.0-dbg" to fix file conflict.
  (Closes: #639258)
* debian/control: Drop the versionless metapackages, they are built from
  postgresql-common now. This behaves better with backports. Thanks to
  Christoph Berg for the suggestion.

7. By Martin Pitt on 2011-08-25

debian/control: Build the versionless metapackages again, and point them
to 9.1.

6. By Martin Pitt on 2011-07-14

* New upstream beta release.
  - Works around gcc 4.6.0 bug. (Closes: #633086)

  Note that this does not change the data format since Beta 2, so no need
  to dump/reload clusters.