summaryrefslogtreecommitdiff
path: root/src/backend
AgeCommit message (Expand)Author
2004-01-09Add WIN32 pipe implementation that uses sockets.Bruce Momjian
2004-01-09Remove duplicate include of value.h. I think it got in because Tom and I bot...Bruce Momjian
2004-01-09Allow libpq to do thread-safe SIGPIPE handling. This allows it toBruce Momjian
2004-01-08[Forced commit to add log message that I fat-fingered last time]Tom Lane
2004-01-07makeTom Lane
2004-01-07Make some improvements in the intelligence of the partial-indexTom Lane
2004-01-07Add missing inclusion of nodes/value.h.Tom Lane
2004-01-07More janitorial work: remove the explicit casting of NULL literals to aNeil Conway
2004-01-07Cleanup: move the 'Value' node into a separate file, rather than puttingNeil Conway
2004-01-06Apply the core parts of Dennis Bjorklund's patch to allow functionTom Lane
2004-01-06Final rearrangement of main postgresql child process (ie.Bruce Momjian
2004-01-06Suppress compiler warning (xlog_outrec is unused if not WAL_DEBUG).Tom Lane
2004-01-06Code cleanup, mostly in the smgr:Neil Conway
2004-01-06Remove unused 'noversion' command-line option processing from theNeil Conway
2004-01-06Make the 'wal_debug' GUC variable a boolean (rather than an integer), andNeil Conway
2004-01-06Instead of rechecking lossy index operators by putting them into theTom Lane
2004-01-05Adjust indexscan planning logic to keep RestrictInfo nodes associatedTom Lane
2004-01-05Trivial refactoring: move analysis of ViewStmt into its own function forNeil Conway
2004-01-05Fix three trivial typos in comments.Neil Conway
2004-01-05Improve UniquePath logic to detect the case where the input is alreadyTom Lane
2004-01-05Dept. of second thoughts: clause_selectivity shouldn't try to cache itsTom Lane
2004-01-05Add the ability to extract OR indexscan conditions from OR-of-ANDTom Lane
2004-01-04There's no longer any good reason for genbki.sh and Gen_fmgrtab.sh toTom Lane
2004-01-04Fix discrepancy in prototypes for HPUX pg_dlerror.Tom Lane
2004-01-04Merge restrictlist_selectivity into clauselist_selectivity byTom Lane
2004-01-04Rewrite OR indexscan processing to be more flexible. We can now for theTom Lane
2003-12-30Adjust the definition of RestrictInfo's left_relids and right_relidsTom Lane
2003-12-30Instead of trying to force WHERE clauses into CNF or DNF normal form,Tom Lane
2003-12-30Avoid running out of memory during hash_create, by not passing aTom Lane
2003-12-30ShmemInitHash forgot to specify HASH_ALLOC flag bit in its hash_createTom Lane
2003-12-29Use hash table name, not one-size-fits-all 'DynaHashTable', to identifyTom Lane
2003-12-29Using canonicalize_qual() to get rid of duplicate index predicateTom Lane
2003-12-29Improve comment.Tom Lane
2003-12-29Tweak OpernameGetCandidates() to reduce palloc overhead --- profilingTom Lane
2003-12-28Clean up the usage of canonicalize_qual(): in particular, be consistentTom Lane
2003-12-27Improve spinlock code for recent x86 processors: insert a PAUSETom Lane
2003-12-25Continued rearrangement to permit pgstat + BootstrapMain processes to beBruce Momjian
2003-12-25Patch that makes quoting "sameuser", "samegroup", and "all" removeBruce Momjian
2003-12-25> > I have no idea if this in Oracle or not. But it's something IBruce Momjian
2003-12-23Use inlined TAS() on PA-RISC, if we are compiling with gcc.Tom Lane
2003-12-23Fix a number of places where reconfiguring with a different installationTom Lane
2003-12-23Mop-up for HAS_TEST_AND_SET refactoring. Un-break two or three platformsTom Lane
2003-12-23Move slock_t typdefs into s_lock.h from include/port files forBruce Momjian
2003-12-23Remove __alpha__ additions in main.c but document that they are missing.Bruce Momjian
2003-12-23Remove NEED_I386_TAS_ASM and just test for compiler defines.Bruce Momjian
2003-12-22This applied patch remove NEED_SPARC_TAS_ASM and instead uses __sparc ||Bruce Momjian
2003-12-22Test for __alpha and __alpha__.Bruce Momjian
2003-12-21Further cleanup in _bt_first: eliminate duplicate code paths.Tom Lane
2003-12-21Back out:Bruce Momjian
2003-12-21[ This description should have been on the earlier fork/execBruce Momjian