summaryrefslogtreecommitdiff
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* Remove stray semicolons in old ecpg preproc grammar ... modern bisonREL7_1_STABLETom Lane2004-10-131-2/+2
| | | | | versions won't compile it at all with those there. Probably of only academic interest now, but ...
* Fix breakage in hashjoin from recent backpatch of left-join bug fix.Tom Lane2004-10-131-10/+10
| | | | (That's what I get for not testing the back branches *before* committing.)
* Hashed LEFT JOIN would miss outer tuples with no inner match if the joinTom Lane2004-09-171-5/+7
| | | | | | was large enough to be batched and the tuples fell into a batch where there were no inner tuples at all. Thanks to Xiaoyu Wang for finding a test case that exposed this long-standing bug.
* Repair multiple memory leaks in getTables(), including one that couldTom Lane2004-03-201-1/+4
| | | | | | easily exhaust memory on databases with more than a few hundred triggers. I don't expect any more releases of these old versions, but let's put the fix in CVS just so it's archived.
* Keep the contents of TIDs not the pointers.Hiroshi Inoue2001-10-012-18/+13
| | | | Tid scan has been broken for 7.1.
* Back-patch deadlock recovery fix into 7.1 tree, in case someone needs it.Tom Lane2001-09-121-9/+27
|
* This patch fixes the well-known but unfixed bug that fetchone() alwaysBruce Momjian2001-08-161-2/+2
| | | | | | | returns the first result in the DB-API compliant wrapper. It turned out that the bug was way down in the C code. Gerhard Häring
* Make hashjoin give the right answer with toasted input data.Tom Lane2001-08-131-18/+32
|
* fix my old fault.Hiroshi Inoue2001-08-094-17/+24
|
* Update branding for 7.1.3, no HISTORY yet.Bruce Momjian2001-08-092-6/+6
|
* Fix Cygwin build, per Jason Tishler.Tom Lane2001-08-091-1/+2
|
* Back-patch fix to disallow COPY TO/FROM a view (or anything else that'sTom Lane2001-08-081-3/+19
| | | | not a plain relation).
* Back-patch fix to prevent infinite loop when $PGDATA is not writable.Tom Lane2001-08-081-4/+7
|
* Commit BEOS patch to 7.1.X.Bruce Momjian2001-08-073-20/+53
|
* Back-patch fixes for dumping user-defined types and dumping commentsTom Lane2001-08-032-30/+83
| | | | on views.
* Fix optimizer to not try to push WHERE clauses down into a sub-SELECT thatTom Lane2001-07-314-38/+69
| | | | | | | has a DISTINCT ON clause, per bug report from Anthony Wood. While at it, improve the DISTINCT-ON-clause recognizer routine to not be fooled by out- of-order DISTINCT lists. Also, back-patch earlier fix to not push down into sub-SELECT with LIMIT.
* Arrange for GRANT/REVOKE on a view to be dumped at the right time,Tom Lane2001-07-291-2/+8
| | | | | namely after the view definition rather than before it. Bug introduced in 7.1 by changes to dump stuff in OID ordering.
* In any case, it seems the REL7_1 branch needs the update too...Tom Lane2001-07-022-4/+2
|
* Fix longstanding error in VACUUM: sometimes would examine a buffer pageTom Lane2001-06-291-12/+9
| | | | | after writing/unpinning it. An actual failure is unlikely, unless the system is tremendously short of buffers ... but a bug is a bug.
* Back-patch fix for attempt to pfree a value that's not palloc'dTom Lane2001-06-131-7/+1
| | | | | (it's a field of a tuple). I see Jan has already fixed this in current sources, but 7.1.* is pretty badly broken here.
* Repair problem with multi-action rules in combination with any nontrivialTom Lane2001-06-121-20/+19
| | | | | | manipulation of rtable/jointree by planner. Rewriter was generating actions that shared rtable/jointree substructure, which caused havoc when planner got to the later actions that it'd already mucked up.
* Back-patch change to not keep WAL segments just for UNDO information.Tom Lane2001-06-061-5/+38
|
* RI triggers would fail for datatypes using old-style equal function,Tom Lane2001-05-311-7/+21
| | | | | | because cached fmgr info contained reference to a shorter-lived data structure. Also guard against possibility that fmgr_info could fail, leaving an incomplete entry present in the hash table.
* Fix a message error in utf_to_localTatsuo Ishii2001-05-281-2/+2
|
* Patch from Barry Lind to correctly decode time zones in timestamp results.REL7_1_2Thomas G. Lockhart2001-05-221-4/+4
| | | | | | Without patch, the time zone field is ignored and the returned time is not correct. Already applied to the development tree...
* Correct recently-broken avg(interval) definition. We can't force anTom Lane2001-05-181-2/+2
| | | | | initdb to fix this in 7.1 installations, but it seems better to be shipping a correct entry than a wrong one.
* Back-patch fix for race condition in heap_update (make sure we holdTom Lane2001-05-171-11/+57
| | | | the buffer lock while checking page free space).
* Update crypto with patch for computation.Bruce Momjian2001-05-152-11/+18
|
* EvalPlanQual was thoroughly broken for concurrent update/delete on inheritanceTom Lane2001-05-153-161/+243
| | | | | | | | trees (mostly my fault). Repair. Also fix long-standing bug in ExecReplace: after recomputing a concurrently updated tuple, we must recheck constraints. Make EvalPlanQual leak memory with somewhat less enthusiasm than before, although plugging leaks fully will require more changes than I care to risk in a dot-release.
* Current implementation of FOR UPDATE has no hope of working correctlyTom Lane2001-05-141-2/+18
| | | | | for relations on the nullable side of an OUTER JOIN. For now I think we'd better refuse such queries.
* Cope with configure arguments that contain spaces.Peter Eisentraut2001-05-131-2/+2
|
* proisstrict must be assumed FALSE when dumping from a 7.0 database,Tom Lane2001-05-121-11/+12
| | | | | | not TRUE. Otherwise we break pl call handler functions. fmgr_oldstyle will take care of making sure the semantics are the same for C functions. Clean up some slightly grotty coding in 7.0 pg_class reading, also.
* Backpatch jdbc1 compile failure fix.Bruce Momjian2001-05-121-1/+1
|
* - Don't dump COMMENTs in data-only dumpsPhilip Warner2001-05-123-66/+122
| | | | | | | | | - Fix view dumping SQL for V7.0 - Fix bug when getting view oid with long view names - Treat SEQUENCE SET TOC entries as data entries rather than schema entries. - Make allowance for data entries that did not have a data dumper routine (eg. SEQUENCE SET)
* Stamp for 7.1.2. Ready when you are...Bruce Momjian2001-05-112-6/+6
|
* Fix remaining RI permission problems (cascaded update/delete, restrict,Peter Eisentraut2001-05-091-3/+62
| | | | set null/default).
* Append and SubqueryScan nodes were not passing changed-parameter signals downTom Lane2001-05-082-8/+30
| | | | | | | | | to their children, leading to misbehavior if they had any children that paid attention to chgParam (most plan node types don't). Append's bug has been there a long time, but nobody had noticed because it used to be difficult to create a query where an Append would be used below the top level of a plan; so there were never any parameters getting passed down. SubqueryScan is new in 7.1 ... and I'd modeled its behavior on Append :-(
* Un-break exec_move_row() for case that a NULL tuple and tupdesc areTom Lane2001-05-081-14/+11
| | | | | passed, which occurs when no rows are retrieved by a SELECT. Mea maxima culpa ... I should have caught this.
* Fix collateral damage from previous (rev 1.49) patch.Peter Eisentraut2001-05-061-1/+4
|
* Makefile should have automatic dependency for parser.o too, if it'sTom Lane2001-05-041-2/+4
| | | | going to have any at all.
* Seems like we should not hold off cancel/die interrupts while we areTom Lane2001-05-041-4/+4
| | | | | running deferred triggers. They are really part of the regular transaction, and they could take awhile.
* Consolidate several near-identical uses of mktime() into a singleTom Lane2001-05-034-122/+69
| | | | | | | | | routine DetermineLocalTimeZone(). In that routine, be more wary of broken mktime() implementations than the original code was: don't allow mktime to change the already-set y/m/d/h/m/s information, and don't use tm_gmtoff if mktime failed. Possibly this will resolve some of the complaints we've been hearing from users of Middle Eastern timezones on RedHat.
* BTW it does not add encodign it just patches existing one (KOI8) toBruce Momjian2001-05-034-19/+20
| | | | | | | support two - KOI8-R and KOI8-U (latter is superset of the former if not to take to the account pseudographics) Andy Rysin
* Permission checking wasn't quite right for insert/update/delete rules,Tom Lane2001-05-031-13/+12
| | | | either :-(.
* apply a little patch:Bruce Momjian2001-05-031-1/+1
| | | | Oleg Bartunov
* Ensure that btree sort ordering functions and boolean comparison operatorsTom Lane2001-05-0313-651/+440
| | | | | | | | | | give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
* Permissions were not checked correctly when one view invokes another.Tom Lane2001-05-031-1/+3
| | | | Per bug report from Lieven Van Acker, 5/2/01.
* Improve list.Bruce Momjian2001-05-031-1/+1
|
* Updates for 7.1.1. Not done yet.Bruce Momjian2001-05-033-5/+5
|
* Correct pg_description entry for type macaddr.Tom Lane2001-05-031-2/+2
|
View Lorry Controller Status