aboutsummaryrefslogtreecommitdiff
path: root/src/backend/access
Commit message (Collapse)AuthorAge
...
* Opclasses live in namespaces. I also took the opportunity to createTom Lane2002-04-17
| | | | | | | an 'opclass owner' column in pg_opclass. Nothing is done with it at present, but since there are plans to invent a CREATE OPERATOR CLASS command soon, we'll probably want DROP OPERATOR CLASS too, which suggests that a notion of ownership would be a good idea.
* Locale support is on by default. The choice of locale is done in initdbPeter Eisentraut2002-04-03
| | | | and/or with GUC variables.
* Create a new GUC variable search_path to control the namespace searchTom Lane2002-04-01
| | | | | | | path. The default behavior if no per-user schemas are created is that all users share a 'public' namespace, thus providing behavior backwards compatible with 7.2 and earlier releases. Probably the semantics and default setting will need to be fine-tuned, but this is a start.
* Reimplement temp tables using schemas. The temp table map is history;Tom Lane2002-03-31
| | | | temp table entries in pg_class have the names the user would expect.
* Mop-up some infelicities in new relation lookup handling.Tom Lane2002-03-29
|
* pg_type has a typnamespace column; system now supports creating typesTom Lane2002-03-29
| | | | | | in different namespaces. Also, cleanup work on relation namespace support: drop, alter, rename commands work for tables in non-default namespaces.
* pg_class has a relnamespace column. You can create and access tablesTom Lane2002-03-26
| | | | | | in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
* Repair two problems with WAL logging of sequence nextvalI() ops, asTom Lane2002-03-15
| | | | | | per recent pghackers discussion: force a new WAL record at first nextval after a checkpoint, and ensure that xlog is flushed to disk if a nextval record is the only thing emitted by a transaction.
* Code review for improved-hashing patch. Fix some portability issuesTom Lane2002-03-09
| | | | | (char != unsigned char, Datum != uint32); make use of new hash code in dynahash hash tables and hash joins.
* I've attached a patch which implements Bob Jenkin's hash function forBruce Momjian2002-03-06
| | | | | | | | | | | | | | | | | | | PostgreSQL. This hash function replaces the one used by hash indexes and the catalog cache. Hash joins use a different, relatively poor-quality hash function, but I'll fix that later. As suggested by Tom Lane, this patch also changes the size of the fixed hash table used by the catalog cache to be a power-of-2 (instead of a prime: I chose 256 instead of 257). This allows the catcache to lookup hash buckets using a simple bitmask. This should improve the performance of the catalog cache slightly, since the previous method (modulo a prime) was slow. In my tests, this improves the performance of hash indexes by between 4% and 8%; the performance when using btree indexes or seqscans is basically unchanged. Neil Conway <neilconway@rogers.com>
* Change made to elog:Bruce Momjian2002-03-06
| | | | | | | | | | | | | | | | | | | o Change all current CVS messages of NOTICE to WARNING. We were going to do this just before 7.3 beta but it has to be done now, as you will see below. o Change current INFO messages that should be controlled by client_min_messages to NOTICE. o Force remaining INFO messages, like from EXPLAIN, VACUUM VERBOSE, etc. to always go to the client. o Remove INFO from the client_min_messages options and add NOTICE. Seems we do need three non-ERROR elog levels to handle the various behaviors we need for these messages. Regression passed.
* I attach a version of my toast-slicing patch, against current CVSBruce Momjian2002-03-05
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (current as of a few hours ago.) This patch: 1. Adds PG_GETARG_xxx_P_SLICE() macros and associated support routines. 2. Adds routines in src/backend/access/tuptoaster.c for fetching only necessary chunks of a toasted value. (Modelled on latest changes to assume chunks are returned in order). 3. Amends text_substr and bytea_substr to use new methods. It now handles multibyte cases -and should still lead to a performance improvement in the multibyte case where the substring is near the beginning of the string. 4. Added new command: ALTER TABLE tabname ALTER COLUMN colname SET STORAGE {PLAIN | EXTERNAL | EXTENDED | MAIN} to parser and documented in alter-table.sgml. (NB I used ColId as the item type for the storage mode string, rather than a new production - I hope this makes sense!). All this does is sets attstorage for the specified column. 4. AlterTableAlterColumnStatistics is now AlterTableAlterColumnFlags and handles both statistics and storage (it uses the subtype code to distinguish). The previous version of my patch also re-arranged other code in backend/commands/command.c but I have dropped that from this patch.(I plan to return to it separately). 5. Documented new macros (and also the PG_GETARG_xxx_P_COPY macros) in xfunc.sgml. ref/alter_table.sgml also contains documentation for ALTER COLUMN SET STORAGE. John Gray
* I was digging through the GiST code, and figured I'd fix up some of theBruce Momjian2002-03-05
| | | | | | | | | | | | "bad smell" in that code. Stuff like function parameters that aren't used, typos in the comments, comparison between signed and unsigned ints, etc. Attached is a pretty trivial patch; it compiles, but beyond that completely untested. Unless anyone sees any problems, please apply for 7.3. Neil Conway
* Catcaches can now store negative entries as well as positive ones, toTom Lane2002-03-03
| | | | | | | | | | speed up repetitive failed searches; per pghackers discussion in late January. inval.c logic substantially simplified, since we can now treat inserts and deletes alike as far as inval events are concerned. Some repair work needed in heap_create_with_catalog, which turns out to have been doing CommandCounterIncrement at a point where the new relation has non-self-consistent catalog entries. With the new inval code, that resulted in assert failures during a relcache entry rebuild.
* Commit to match discussed elog() changes. Only update is that LOG isBruce Momjian2002-03-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | now just below FATAL in server_min_messages. Added more text to highlight ordering difference between it and client_min_messages. --------------------------------------------------------------------------- REALLYFATAL => PANIC STOP => PANIC New INFO level the prints to client by default New LOG level the prints to server log by default Cause VACUUM information to print only to the client NOTICE => INFO where purely information messages are sent DEBUG => LOG for purely server status messages DEBUG removed, kept as backward compatible DEBUG5, DEBUG4, DEBUG3, DEBUG2, DEBUG1 added DebugLvl removed in favor of new DEBUG[1-5] symbols New server_min_messages GUC parameter with values: DEBUG[5-1], INFO, NOTICE, ERROR, LOG, FATAL, PANIC New client_min_messages GUC parameter with values: DEBUG[5-1], LOG, INFO, NOTICE, ERROR, FATAL, PANIC Server startup now logged with LOG instead of DEBUG Remove debug_level GUC parameter elog() numbers now start at 10 Add test to print error message if older elog() values are passed to elog() Bootstrap mode now has a -d that requires an argument, like postmaster
* Clean up BeginCommand and related routines. BeginCommand and EndCommandTom Lane2002-02-27
| | | | | | | | | | | are now both invoked once per received SQL command (raw parsetree) from pg_exec_query_string. BeginCommand is actually just an empty routine at the moment --- all its former operations have been pushed into tuple receiver setup routines in printtup.c. This makes for a clean distinction between BeginCommand/EndCommand (once per command) and the tuple receiver setup/teardown routines (once per ExecutorRun call), whereas the old code was quite ad hoc. Along the way, clean up the calling conventions for ExecutorRun a little bit.
* I've attached a simple patch which should improve the performance ofBruce Momjian2002-02-25
| | | | | | | | | | | | | | hashname() and reduce the penalty incured when NAMEDATALEN is increased. I posted this to -hackers a couple days ago, and there haven't been any major complaints. It passes the regression tests. See -hackers for more discussion, as well as the suggestion from Tom Lane on which this patch is based. Unless anyone sees any problems, please apply for 7.3. Cheers, Neil Conway
* A bunch of changes aimed at reducing backend startup time...Tom Lane2002-02-19
| | | | | | | | | | | | | | | | | | | | | | | | | | Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
* Add better error text:Bruce Momjian2002-02-18
| | | | | elog(LOG, "XLogWrite: new log file created - " "consider increasing 'wal_files' in postgresql.conf.");
* Tweak GiST code to work correctly on machines where 8-byte alignmentTom Lane2002-02-11
| | | | | | of pointers is required. Patch from Teodor Sigaev per pghackers discussion. It's an ugly kluge but avoids forcing initdb; we'll put a better fix into 7.3 or later.
* TOAST needs to do at least minimal time-qual checking in order not toTom Lane2002-01-16
| | | | | | | | | | | mess up after an aborted VACUUM FULL, per today's pghackers discussion. Add a suitable HeapTupleSatisfiesToast routine. Remove useless special- case test in HeapTupleSatisfiesVisibility macro for xmax = BootstrapTransactionId; perhaps that was needed at one time, but it's a waste of cycles now, not to mention actively wrong for SnapshotAny. Along the way, add some much-needed comments to tqual.c, and simplify toast_fetch_datum, which no longer needs to assume it may see chunks out-of-order.
* Add more sanity-checking to PageAddItem and PageIndexTupleDelete,Tom Lane2002-01-15
| | | | | | | to prevent spreading of corruption when page header pointers are bad. Merge PageZero into PageInit, since it was never used separately, and remove separate memset calls used at most other PageInit call points. Remove IndexPageCleanup, which wasn't used at all.
* Reduce severity of 'XLogFlush: request is not satisfied' error condition,Tom Lane2002-01-14
| | | | | | per my proposal of a couple days ago. This will eliminate the unable- to-restart-database class of problem that we have seen reported half a dozen times with 7.1.*.
* Add CHECK_FOR_INTERRUPTS() in various strategic spots, per commentsTom Lane2002-01-06
| | | | from Hiroshi.
* Fix race condition that could allow two concurrent transactionsTom Lane2002-01-01
| | | | | | to insert the same key into a supposedly unique index. The bug is of low probability, and may not explain any of the recent reports of duplicated rows; but a bug is a bug.
* Ensure that all direct uses of spinlock-protected data structures useTom Lane2001-12-28
| | | | | | | | | | | 'volatile' pointers to access those structures, so that optimizing compilers will not decide to move the structure accesses outside of the spinlock-acquire-to-spinlock-release sequence. There are no known bugs in these uses at present, but based on bad experience with lwlock.c, it seems prudent to ensure that we protect these other uses too. Per pghackers discussion around 12-Dec. (Note: it should not be necessary to worry about structures protected by LWLocks, since the LWLock acquire and release operations are not inline macros.)
* Fix mispeling ...Tom Lane2001-12-23
|
* Temporarily dike out GetUndoRecPtr() in checkpoint generation, since weTom Lane2001-12-19
| | | | | | do not use the undo pointer anyway. This is a quick-hack solution for the three-way deadlock condition discussed in pghackers 17-Dec-01. Need to find a better way of doing it.
* Fix boundary condition in btbulkdelete: don't examine high key in caseTom Lane2001-11-23
| | | | | | | where rightmost index page splits while we are waiting to obtain exclusive lock on it. Not clear this would actually hurt (probably the callback would always fail), but better safe than sorry. Also, improve comments describing concurrency considerations in this code.
* Clean up usage-statistics display code (ShowUsage and friends). StatFpTom Lane2001-11-10
| | | | is gone, usage messages now go through elog(DEBUG).
* New pgindent run with fixes suggested by Tom. Patch manually reviewed,Bruce Momjian2001-11-05
| | | | initdb/regression tests pass.
* Merge three existing ways of signaling postmaster from child processes,Tom Lane2001-11-04
| | | | | | | | | | | so that only one signal number is used not three. Flags in shared memory tell the reason(s) for the current signal. This method is extensible to handle more signal reasons without chewing up even more signal numbers, but the immediate reason is to keep pg_pwd reloads separate from SIGHUP processing in the postmaster. Also clean up some problems in the postmaster with delayed response to checkpoint status changes --- basically, it wouldn't schedule a checkpoint if it wasn't getting connection requests on a regular basis.
* Fix problem reported by Alex Korn: if a relation has been dropped andTom Lane2001-11-02
| | | | | | | | | | | | | | | | | | recreated since the start of our transaction, our first reference to it errored out because we'd try to reuse our old relcache entry for it. Do this by accepting SI inval messages just before relcache search in heap_openr, so that dead relcache entries will be flushed before we search. Also, break heap_open/openr into two pairs of routines, relation_open(r) and heap_open(r). The relation_open routines make no tests on relkind and so can be used to open anything that has a pg_class entry. The heap_open routines are wrappers that add a relkind test to preserve their established behavior. Use the relation_open routines in several places that had various kluge solutions for opening rels that might be either heap or index rels. Also, remove the old 'heap stats' code that's been superseded by Jan's stats collector, and clean up some inconsistencies in error reporting between the different types of ALTER TABLE.
* Add missing #include.Tom Lane2001-11-01
|
* Another pgindent run. Fixes enum indenting, and improves #endifBruce Momjian2001-10-28
| | | | spacing. Also adds space for one-line comments.
* Fix a couple of places where lack of parenthesization of a castTom Lane2001-10-25
| | | | | causes pgindent to make weird formatting decisions. Easiest fix seems to be to put in the extra parens...
* pgindent run on all C files. Java run to follow. initdb/regressionBruce Momjian2001-10-25
| | | | tests pass.
* Accept an INTERVAL argument for SET TIME ZONE per SQL99.Thomas G. Lockhart2001-10-18
| | | | | | | | | | | | | | | | | | | | | | | | | Modified the parser and the SET handlers to use full Node structures rather than simply a character string argument. Implement INTERVAL() YEAR TO MONTH (etc) syntax per SQL99. Does not yet accept the goofy string format that goes along with, but this should be fairly straight forward to fix now as a bug or later as a feature. Implement precision for the INTERVAL() type. Use the typmod mechanism for both of INTERVAL features. Fix the INTERVAL syntax in the parser: opt_interval was in the wrong place. INTERVAL is now a reserved word, otherwise we get reduce/reduce errors. Implement an explicit date_part() function for TIMETZ. Should fix coersion problem with INTERVAL reported by Peter E. Fix up some error messages for date/time types. Use all caps for type names within message. Fix recently introduced side-effect bug disabling 'epoch' as a recognized field for date_part() etc. Reported by Peter E. (??) Bump catalog version number. Rename "microseconds" current transaction time field from ...Msec to ...Usec. Duh! date/time regression tests updated for reference platform, but a few changes will be necessary for others.
* Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfoTom Lane2001-10-06
| | | | | | | | | lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
* Further cleanup of dynahash.c API, in pursuit of portability andTom Lane2001-10-05
| | | | | | | | | readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
* Another round of cleanups for dynahash.c (maybe it's finally clean ofTom Lane2001-10-01
| | | | | | | | | portability issues). Caller-visible data structures are now allocated on MAXALIGN boundaries, allowing safe use of datatypes wider than 'long'. Rejigger hash_create API so that caller specifies size of key and total size of entry, not size of key and size of rest of entry. This simplifies life considerably since each number is just a sizeof(), and padding issues etc. are taken care of automatically.
* Tweak btree page split logic so that when splitting a page that isTom Lane2001-09-29
| | | | | | | | | | rightmost on its tree level, we split 2/3 to the left and 1/3 to the new right page, rather than the even split we use elsewhere. The idea is that when faced with a steadily increasing series of inserted keys (such as sequence or timestamp values), we'll end up with a btree that's about 2/3ds full not 1/2 full, which is much closer to the desired steady-state load for a btree. Per suggestion from Ann Harrison of IBPhoenix.
* Implement new 'lightweight lock manager' that's intermediate betweenTom Lane2001-09-29
| | | | | | | | | existing lock manager and spinlocks: it understands exclusive vs shared lock but has few other fancy features. Replace most uses of spinlocks with lightweight locks. All remaining uses of spinlocks have very short lock hold times (a few dozen instructions), so tweak spinlock backoff code to work efficiently given this assumption. All per my proposal on pghackers 26-Sep-01.
* I have made three changes to the rtree code: one bug fix andBruce Momjian2001-09-29
| | | | | | | | | | | | | | | | | two performance improvements. I put an explanation of the changes at http://cs1.cs.nyu.edu/been/postgres-rtree.html The performance improvements are quite significant. All the changes are in the file src/backend/access/rtree/rtree.c I was working with the 7.1.3 code. I'm including the diff output as an attachment. Kenneth Been
* Measure the current transaction time to milliseconds.Thomas G. Lockhart2001-09-28
| | | | | | | | | | | | | | Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
* Remove useless test for time field in pg_control being > 0. We don'tTom Lane2001-09-26
| | | | | | need this, and it will create a Y2038 failure. Per report from David Wheeler, who is evidently running on a platform where time_t is already negative.
* Add some debugging details to some of the elog(STOP) conditions for WAL.Tom Lane2001-09-06
| | | | | | Standardize on %X/%X as the formatting for XLOG position display --- we had a couple of different formats before, and none of 'em were as useful as hex offsets IMHO.
* Transaction IDs wrap around, per my proposal of 13-Aug-01. MoreTom Lane2001-08-26
| | | | documentation to come, but the code is all here. initdb forced.
* Fix portability problem in new CLOG code, per report from Rene Pijlman.Tom Lane2001-08-25
|
* Replace implementation of pg_log as a relation accessed through theTom Lane2001-08-25
| | | | | | | | | | | buffer manager with 'pg_clog', a specialized access method modeled on pg_xlog. This simplifies startup (don't need to play games to open pg_log; among other things, OverrideTransactionSystem goes away), should improve performance a little, and opens the door to recycling commit log space by removing no-longer-needed segments of the commit log. Actual recycling is not there yet, but I felt I should commit this part separately since it'd still be useful if we chose not to do transaction ID wraparound.
generated by cgit v1.2.3 (git 2.25.1) at 2025-08-03 14:26:15 +0000