summaryrefslogtreecommitdiff
path: root/src/backend/optimizer
AgeCommit message (Expand)Author
2003-09-07Remove geqo_random_seed parameter. Having geqo reset the global random()Tom Lane
2003-08-27Department of second thoughts: can't reorder merge-join clauses.Tom Lane
2003-08-26Push subplan clauses to the back in qual lists for join plans, notTom Lane
2003-08-17Fix ARRAY[] construct so that in multidimensional case, elements canTom Lane
2003-08-17Create a 'type cache' that keeps track of the data needed for any particularTom Lane
2003-08-12Change some frequently-reached elog(DEBUG...) calls to ereport(DEBUG...)Tom Lane
2003-08-11Rewriter and planner should use only resno, not resname, to identifyTom Lane
2003-08-11Code cleanup inspired by recent resname bug report (doesn't fix the bugTom Lane
2003-08-08Another pgindent run with updated typedefs.Bruce Momjian
2003-08-07create_unique_plan() should not discard existing output columns of theTom Lane
2003-08-04Update copyrights to 2003.Bruce Momjian
2003-08-04pgindent run.Bruce Momjian
2003-08-03Tighten inline_function's test for overly complex parameters. ThisTom Lane
2003-07-28Add error stack traceback support for SQL-language functions.Tom Lane
2003-07-28A visit from the message-style police ...Tom Lane
2003-07-25Error message editing in backend/optimizer, backend/rewrite.Tom Lane
2003-07-20Another round of error message editing, covering backend/commands/.Tom Lane
2003-07-16Add defenses against trying to attach qual conditions to a setOperationTom Lane
2003-07-14Make cost estimates for SubqueryScan more realistic: charge cpu_tuple_costTom Lane
2003-07-03Do honest transformation and preprocessing of LIMIT/OFFSET clauses,Tom Lane
2003-07-03Code review for UPDATE tab SET col = DEFAULT patch ... whack it aroundTom Lane
2003-07-01Dept. of second thoughts: supporting inlining of polymorphic SQL functionsTom Lane
2003-07-01SQL functions can have arguments and results declared ANYARRAY orTom Lane
2003-06-29Restructure building of join relation targetlists so that a join planTom Lane
2003-06-29Support expressions of the form 'scalar op ANY (array)' andTom Lane
2003-06-25Back out array mega-patch.Bruce Momjian
2003-06-24Array mega-patch.Bruce Momjian
2003-06-16Allow GROUP BY, ORDER BY, DISTINCT targets to be unknown literals,Tom Lane
2003-06-15Adjust nestloop-with-inner-indexscan plan generation so that we catchTom Lane
2003-06-06Implement outer-level aggregates to conform to the SQL spec, withTom Lane
2003-05-28Knock down a couple more lappend() hotspots for large WHERE clauses.Tom Lane
2003-05-28Fix some planner performance problems with large WHERE clauses, byTom Lane
2003-05-28Replace functional-index facility with expressional indexes. Any columnTom Lane
2003-05-27Make debug_ GUC varables output DEBUG1 rather than LOG, and mention inBruce Momjian
2003-05-26Cause CHAR(n) to TEXT or VARCHAR conversion to automatically strip trailingTom Lane
2003-05-15Add missing #include.Tom Lane
2003-05-15Indexing support for pattern matching operations via separate operatorPeter Eisentraut
2003-05-13Change pg_amop's index on (amopclaid,amopopr) to index (amopopr,amopclaid).Tom Lane
2003-05-12Apply fixes for problems with dropped columns whose types have also beenTom Lane
2003-05-11Revert previous change --- turns out the underlying problem affectsTom Lane
2003-05-11Disable the recently-added use_physical_tlist optimization in casesTom Lane
2003-05-06Implement feature of new FE/BE protocol whereby RowDescription identifiesTom Lane
2003-05-02Portal and memory management infrastructure for extended query protocol.Tom Lane
2003-05-02Repair sloppiness about where cached mergeclause pathkeys are allocated.Tom Lane
2003-04-29Infrastructure for deducing Param types from context, in the same wayTom Lane
2003-04-27Put back encoding-conversion step in processing of incoming queries;Tom Lane
2003-04-24Adjust subquery qual pushdown rules so that we can push down a qualTom Lane
2003-04-08First phase of work on array improvements. ARRAY[x,y,z] constructorTom Lane
2003-03-23Tweak selectivity and related routines to cope with domains. Per reportTom Lane
2003-03-22Department of second thoughts: probably shouldn't use nth() to get theTom Lane