hcq/postgresql
1
0
Fork 0
mirror of https://git.postgresql.org/git/postgresql.git synced 2026-04-06 07:28:26 +08:00
Code Issues Packages Projects Releases Wiki Activity

Commit Graph

  • 3c8c3ccf99 Add item, not fixed: Bruce Momjian 2006-06-14 21:40:00 +00:00
  • ca3d2c31f9 Remove, already in back branches: Bruce Momjian 2006-06-14 21:38:21 +00:00
  • bcc97255f0 Done: Bruce Momjian 2006-06-14 21:26:49 +00:00
  • c135db1d4c Add information schema item: Bruce Momjian 2006-06-14 21:10:50 +00:00
  • 63b35a8cb1 Add COPY views patch URL: Bruce Momjian 2006-06-14 20:40:20 +00:00
  • 744697317e Add: Bruce Momjian 2006-06-14 19:32:48 +00:00
  • 82009572ac Keep TODO.detail until item is done: Bruce Momjian 2006-06-14 19:02:06 +00:00
  • 22fd9a0ff8 Add URL, remove TODO.detail: Bruce Momjian 2006-06-14 19:01:42 +00:00
  • 5b50ede670 Add locale/encoding references: Bruce Momjian 2006-06-14 18:47:53 +00:00
  • bc29957202 Add locale TODO.detail emails. Bruce Momjian 2006-06-14 18:42:51 +00:00
  • 9ae8690ad2 Update link to TODO.detail: Bruce Momjian 2006-06-14 18:39:18 +00:00
  • f944f88021 Add: Bruce Momjian 2006-06-14 18:36:57 +00:00
  • 1e52c5eb96 Added: Bruce Momjian 2006-06-14 17:59:33 +00:00
  • b6f5676c60 Add entry: Bruce Momjian 2006-06-14 17:51:55 +00:00
  • 2193beb761 Further thoughts about lo_export/lo_import error handling: if one of the lower-level large object functions fails, it will have already set a suitable error message --- probably something from the backend --- and it is not useful to overwrite that with a generic 'error while reading large object' message. So remove redundant messages. Tom Lane 2006-06-14 17:49:25 +00:00
  • f3164c0200 Clean up psql's control-C handling to avoid longjmp'ing out of random places --- that risks corrupting data structures, losing sync with the backend, etc. We now longjmp only from calls to readline, fgets, and fread, which we assume are coded to protect themselves against interrupts at undesirable times. This requires adding explicit tests for cancel_pressed in long-running loops, but on the whole it's far cleaner. Martijn van Oosterhout and Tom Lane. Tom Lane 2006-06-14 16:49:03 +00:00
  • ace93353ea Improve error detection and recovery in lo_import()/lo_export(). Problems noted while experimenting with new control-C code for psql. Tom Lane 2006-06-14 01:28:55 +00:00
  • 067e40cc6d Do a CHECK_FOR_INTERRUPTS immediately before terminating a fastpath function call. Previously, there may have been no CHECK_FOR_INTERRUPTS at all in the fastpath code path, making it impossible to cancel an operation such as \lo_import externally. This addition doesn't ensure you can cancel, since your SIGINT may arrive while the backend is idle waiting for the client, but it gives the largest window we can easily provide. Noted while experimenting with new control-C code for psql. Tom Lane 2006-06-14 01:24:14 +00:00
  • aa30e0e555 Remove: Bruce Momjian 2006-06-13 21:08:22 +00:00
  • 07c25723da Add BY clause to PL/PgSQL FOR loop, to control the iteration increment. Bruce Momjian 2006-06-12 16:45:30 +00:00
  • 4cfe1fadad Avoid use of C commment inside C comment from recent Win32 int overflow patch. Bruce Momjian 2006-06-12 16:29:08 +00:00
  • e6a7b01930 Avoid use of C commment inside C comment from recent Win32 int overflow patch. Bruce Momjian 2006-06-12 16:28:52 +00:00
  • 27bff7502f Prevent Win32 from displaying a popup box on backend crash. Instead let the postmaster deal with it. Bruce Momjian 2006-06-12 16:17:20 +00:00
  • 11a0c1de9f Add URL for bitmap AM: Bruce Momjian 2006-06-12 16:15:07 +00:00
  • f7a0b645f5 Win32 can't catch the exception thrown by INT_MIN / -1 or INT_MIN * -1, so on that platform we test for those before the computation and throw an "out of range" error. Bruce Momjian 2006-06-12 16:09:39 +00:00
  • 9fc6f4e1ae Win32 can't catch the exception thrown by INT_MIN / -1 or INT_MIN * -1, so on that platform we test for those before the computation and throw an "out of range" error. Bruce Momjian 2006-06-12 16:09:11 +00:00
  • e71e02d926 Minor code cleanup: make the WIN32 case less gratuitously different from the other platform-specific cases in ps_status. Tom Lane 2006-06-12 02:39:49 +00:00
  • e2e2a9db4c Code review for psql multiline history patch(es). Fix memory leak, failure to enter commands in history if canceled by control-C, other infelicities. Tom Lane 2006-06-11 23:06:00 +00:00
  • e2d201ecdb Fix Assert failure when a fastpath function call is attempted inside an already-aborted transaction block. GetSnapshotData throws an Assert if not in a valid transaction; hence we mustn't attempt to set a snapshot for the function until after checking for aborted transaction. This is harmless AFAICT if Asserts aren't enabled (GetSnapshotData will compute a bogus snapshot, but it doesn't matter since HandleFunctionRequest will throw an error shortly anywy). Hence, not a major bug. Tom Lane 2006-06-11 15:49:46 +00:00
  • b45f5bb726 Fix Assert failure when a fastpath function call is attempted inside an already-aborted transaction block. GetSnapshotData throws an Assert if not in a valid transaction; hence we mustn't attempt to set a snapshot for the function until after checking for aborted transaction. This is harmless AFAICT if Asserts aren't enabled (GetSnapshotData will compute a bogus snapshot, but it doesn't matter since HandleFunctionRequest will throw an error shortly anywy). Hence, not a major bug. Tom Lane 2006-06-11 15:49:36 +00:00
  • bf64a37909 Fix Assert failure when a fastpath function call is attempted inside an already-aborted transaction block. GetSnapshotData throws an Assert if not in a valid transaction; hence we mustn't attempt to set a snapshot for the function until after checking for aborted transaction. This is harmless AFAICT if Asserts aren't enabled (GetSnapshotData will compute a bogus snapshot, but it doesn't matter since HandleFunctionRequest will throw an error shortly anywy). Hence, not a major bug. Tom Lane 2006-06-11 15:49:28 +00:00
  • 807bfa581c Remove pg_dump wording: Bruce Momjian 2006-06-10 23:33:27 +00:00
  • 0d001c86f8 Add: Bruce Momjian 2006-06-10 23:17:21 +00:00
  • 98365a1e34 Repair remarkably-inefficient query for dumping large object comments: it was invoking obj_description() for each large object chunk, instead of once per large object. This code is new as of 8.1, which may explain why the problem hadn't been noticed already. Tom Lane 2006-06-09 19:46:17 +00:00
  • b773987fc2 Repair remarkably-inefficient query for dumping large object comments: it was invoking obj_description() for each large object chunk, instead of once per large object. This code is new as of 8.1, which may explain why the problem hadn't been noticed already. Tom Lane 2006-06-09 19:46:09 +00:00
  • 5de0cbdf0c Revert sampling patch for EXPLAIN ANALYZE; it turns out to be too unreliable because node timing is much less predictable than the patch expects. I kept the API change for InstrStopNode, however. Tom Lane 2006-06-09 19:30:56 +00:00
  • 9143144935 Cleanups: Bruce Momjian 2006-06-09 18:41:57 +00:00
  • c0ba4829a9 Cleanup: Bruce Momjian 2006-06-09 18:12:29 +00:00
  • 04e9704b9e Now ispell dictionary can eat dictionaries in MySpell format, used by OpenOffice. Dictionaries are placed at http://lingucomponent.openoffice.org/spell_dic.html Dictionary automatically recognizes format of files. Teodor Sigaev 2006-06-09 13:25:59 +00:00
  • 1a1326d64d Add URL. Bruce Momjian 2006-06-09 02:11:10 +00:00
  • fa23204c5d Update: Bruce Momjian 2006-06-09 01:02:22 +00:00
  • 2f12903eb9 Remove: Bruce Momjian 2006-06-09 01:00:36 +00:00
  • 03c1be0663 Fix bootstrap.c so that database startup process and bgwriter properly release LWLocks during a panic exit. This avoids the possible self-deadlock pointed out by Qingqing Zhou. Also, I noted that an error during LoadFreeSpaceMap() or BuildFlatFiles() would result in exit(0) which would leave the postmaster thinking all is well. Added a critical section to ensure such errors don't allow startup to proceed. Tom Lane 2006-06-08 23:55:54 +00:00
  • 1358466a8c Fix bootstrap.c so that database startup process and bgwriter properly release LWLocks during a panic exit. This avoids the possible self-deadlock pointed out by Qingqing Zhou. Also, I noted that an error during LoadFreeSpaceMap() or BuildFlatFiles() would result in exit(0) which would leave the postmaster thinking all is well. Added a critical section to ensure such errors don't allow startup to proceed. Tom Lane 2006-06-08 23:55:48 +00:00
  • e4725e35be Add URL. Bruce Momjian 2006-06-08 18:59:01 +00:00
  • 51aa05a437 Use simple URL's rather than text and a URL: Bruce Momjian 2006-06-08 16:07:23 +00:00
  • 2b5b47605c Add URL. Bruce Momjian 2006-06-08 15:41:22 +00:00
  • e36d0a0cbf Remove Win32 file, moved to /port. Bruce Momjian 2006-06-08 15:15:13 +00:00
  • 8ff80c1bd3 Remove obsolete comment about VACUUM FULL: it takes buffer content locks now, and must do so to ensure bgwriter doesn't write a page that is in process of being compacted. Tom Lane 2006-06-08 14:58:33 +00:00
  • 1758b3ec96 /contrib/adminpack: More clearly identify renaming of existing backend functions. I also found that pg_file_length was incorrectly documented in the README as pg_file_size. Bruce Momjian 2006-06-08 14:32:11 +00:00
  • ea0390b682 Spelling fix. Bruce Momjian 2006-06-08 13:45:36 +00:00
  • 4369b0b521 Done: Bruce Momjian 2006-06-08 13:44:09 +00:00
  • 12632d3c31 Add missing 3rd argument to open(). Bruce Momjian 2006-06-08 03:29:30 +00:00
  • 149008dfe7 Add missing 3rd parameter to open(). Bruce Momjian 2006-06-08 03:28:01 +00:00
  • 44f62decad Add URL for ALTER CONSTRAINT RENAME. Bruce Momjian 2006-06-08 03:17:11 +00:00
  • ae76c3ace9 Add URL. Bruce Momjian 2006-06-08 02:59:02 +00:00
  • 366231f42f Add URL: Bruce Momjian 2006-06-08 02:42:44 +00:00
  • 8a75c07626 Update wording: Bruce Momjian 2006-06-08 01:08:45 +00:00
  • 0f1a7c55b4 Add entry: Bruce Momjian 2006-06-08 01:02:53 +00:00
  • 18aee7968a Allow timezone names in SQL strings, Bruce Momjian 2006-06-07 22:32:31 +00:00
  • 51dfe35121 Done: Bruce Momjian 2006-06-07 22:31:38 +00:00
  • 399a36a75d Prepare code to be built by MSVC: Bruce Momjian 2006-06-07 22:24:46 +00:00
  • 877e296306 Spelling fix. Bruce Momjian 2006-06-07 21:23:01 +00:00
  • 9ca4c153bf Mention GPL /contrib modules: Bruce Momjian 2006-06-07 20:39:10 +00:00
  • d8364f3f8f Per previous analysis, the most correct notion of SampleOverhead is that it is just the total time to do INSTR_TIME_SET_CURRENT(), and not any of the other code involved in InstrStartNode/InstrStopNode. Even though I fear we may end up reverting this patch altogether, we may as well have the most correct version in our CVS archive. Tom Lane 2006-06-07 18:49:03 +00:00
  • e9f6679ce2 Have REINDEX use NOTICE, so it can be suppressed. Bruce Momjian 2006-06-07 17:20:17 +00:00
  • 735529ce24 Remove "fuzzy comparison" logic in qsort comparison function for choose_bitmap_and(). It was way too fuzzy --- per comment, it was meant to be 1% relative difference, but was actually coded as 0.01 absolute difference, thus causing selectivities of say 0.001 and 0.000000000001 to be treated as equal. I believe this thinko explains Maxim Boguk's recent complaint. While we could change it to a relative test coded like compare_fuzzy_path_costs(), there's a bigger problem here, which is that any fuzziness at all renders the comparison function non-transitive, which could confuse qsort() to the point of delivering completely wrong results. So forget the whole thing and just do an exact comparison. Tom Lane 2006-06-07 17:08:15 +00:00
  • ae0c8d09fb Remove "fuzzy comparison" logic in qsort comparison function for choose_bitmap_and(). It was way too fuzzy --- per comment, it was meant to be 1% relative difference, but was actually coded as 0.01 absolute difference, thus causing selectivities of say 0.001 and 0.000000000001 to be treated as equal. I believe this thinko explains Maxim Boguk's recent complaint. While we could change it to a relative test coded like compare_fuzzy_path_costs(), there's a bigger problem here, which is that any fuzziness at all renders the comparison function non-transitive, which could confuse qsort() to the point of delivering completely wrong results. So forget the whole thing and just do an exact comparison. Tom Lane 2006-06-07 17:08:07 +00:00
  • ca9d50304f For psql multi-line history, save backslash commands right away, rather than only if there is already history. Bruce Momjian 2006-06-07 13:18:37 +00:00
  • 9c351695dc Use INFO-level for reindex messages, rather than NOTICE. Bruce Momjian 2006-06-07 13:13:16 +00:00
  • c944851a79 Backpatch comment about --sysconfdir. Bruce Momjian 2006-06-07 12:19:37 +00:00
  • 6ca917adff Update comment. Bruce Momjian 2006-06-06 22:41:09 +00:00
  • 87632c1b8e On Win32, return original patch if GetShortPathName() fails (no short name, path does not exist), rather than returning nothing. Bruce Momjian 2006-06-06 22:32:56 +00:00
  • 62e7ad6603 On Win32, return original patch if GetShortPathName() fails (no short name, path does not exist), rather than returning nothing. Bruce Momjian 2006-06-06 22:32:19 +00:00
  • 8a30cc2127 Make the planner estimate costs for nestloop inner indexscans on the basis that the Mackert-Lohmann formula applies across all the repetitions of the nestloop, not just each scan independently. We use the M-L formula to estimate the number of pages fetched from the index as well as from the table; that isn't what it was designed for, but it seems reasonably applicable anyway. This makes large numbers of repetitions look much cheaper than before, which accords with many reports we've received of overestimation of the cost of a nestloop. Also, change the index access cost model to charge random_page_cost per index leaf page touched, while explicitly not counting anything for access to metapage or upper tree pages. This may all need tweaking after we get some field experience, but in simple tests it seems to be giving saner results than before. The main thing is to get the infrastructure in place to let cost_index() and amcostestimate functions take repeated scans into account at all. Per my recent proposal. Tom Lane 2006-06-06 17:59:58 +00:00
  • 05631354f3 Add URL to RESET CONNECTION: Bruce Momjian 2006-06-06 16:27:23 +00:00
  • 92bcb5abe0 Allow do not lexize words in substitution. Docs will be submitted some later, now it's at http://www.sai.msu.su/~megera/oddmuse/index.cgi/Thesaurus_dictionary Teodor Sigaev 2006-06-06 16:25:55 +00:00
  • 63e464a5e6 Remove ancient AIX structure workaround. Bruce Momjian 2006-06-06 16:20:11 +00:00
  • 06a1807e5e Have BCC echo output "" using "", rather than \"\" as earlier reported. Backpatch to 8.1.X. Bruce Momjian 2006-06-06 16:17:31 +00:00
  • ed3b2e696b Have BCC echo output "" using "", rather than \"\" as earlier reported. Backpatch to 8.1.X. Bruce Momjian 2006-06-06 16:17:20 +00:00
  • 491a8c27c9 Done: Bruce Momjian 2006-06-06 13:52:21 +00:00
  • aac12de9b6 Fixed two memory leaks in ecpglib. Michael Meskes 2006-06-06 11:36:22 +00:00
  • 3066e8ab03 Fixed two memory leaks in ecpglib. Michael Meskes 2006-06-06 11:36:12 +00:00
  • 8fc6b3dbf3 Fixed two memory leaks in ecpglib. Michael Meskes 2006-06-06 11:36:00 +00:00
  • 0251602fc9 Fixed two more memory leaks in ecpglib. Synced parser. Michael Meskes 2006-06-06 11:31:55 +00:00
  • ff0ac57db9 Remove extremely old, incomplete, broken example code. Per my proposal a few days ago. Tom Lane 2006-06-05 21:16:32 +00:00
  • 7868590c61 While making the seq_page_cost changes, I was struck by the fact that cost_nonsequential_access() is really totally inappropriate for its only remaining use, namely estimating I/O costs in cost_sort(). The routine was designed on the assumption that disk caching might eliminate the need for some re-reads on a random basis, but there's nothing very random in that sense about sort's access pattern --- it'll always be picking up the oldest outputs. If we had a good fix on the effective cache size we might consider charging zero for I/O unless the sort temp file size exceeds it, but that's probably putting much too much faith in the parameter. Instead just drop the logic in favor of a fixed compromise between seq_page_cost and random_page_cost per page of sort I/O. Tom Lane 2006-06-05 20:56:33 +00:00
  • b7af62e4a9 Avoid longjump/vfork warning about line_saved_in_history variable by making it volatile. Bruce Momjian 2006-06-05 14:47:38 +00:00
  • a073a70415 Have psql history save \e editor contents to history before query is executed, like it did for 8.1.X. Bruce Momjian 2006-06-05 03:55:00 +00:00
  • e4de635a2b Increase the default value of cpu_index_tuple_cost from 0.001 to 0.005. This shouldn't affect simple indexscans much, while for bitmap scans that are touching a lot of index rows, this seems to bring the estimates more in line with reality. Per recent discussion. Tom Lane 2006-06-05 03:03:42 +00:00
  • eed6c9ed7e Add a GUC parameter seq_page_cost, and use that everywhere we formerly assumed that a sequential page fetch has cost 1.0. This patch doesn't in itself change the system's behavior at all, but it opens the door to people adopting other units of measurement for EXPLAIN costs. Also, if we ever decide it's worth inventing per-tablespace access cost settings, this change provides a workable intellectual framework for that. Tom Lane 2006-06-05 02:49:58 +00:00
  • a837851dc0 Fix psql so \r is separate in the history, rather than at the end of the query. Bruce Momjian 2006-06-04 04:35:55 +00:00
  • f6dc20e109 Update: Bruce Momjian 2006-06-04 02:11:04 +00:00
  • d7739850e4 Update: Bruce Momjian 2006-06-04 01:33:39 +00:00
  • e5b198066d Update GUI FAQ. Bruce Momjian 2006-06-04 00:17:43 +00:00
  • 1309c6f547 Update GUI tools URL. Bruce Momjian 2006-06-04 00:13:43 +00:00
  • f2de12e50e Don't choke during startup if the environment offers an invalid value for LC_MESSAGES; instead, just press forward, leaving the effective setting at 'C'. There is not any very good reason to complain when we are going to replace the value soon with whatever postgresql.conf says. This change should solve the occasionally-reported problem of initdb failing with 'failed to initialize lc_messages'; the current theory is that that is a reflection of either wrong LANG/LC_MESSAGES or completely broken locale support. Tom Lane 2006-06-03 17:36:10 +00:00
  • c8fc48e257 Fix copy-and-pasteo in Russian translation: message complaining about HAVE_INT64_TIMESTAMP was mentioning PG_CONTROL_VERSION instead. Victor Snezhko Tom Lane 2006-06-03 16:49:46 +00:00
  • b734fc67a9 Fix copy-and-pasteo in Russian translation: message complaining about HAVE_INT64_TIMESTAMP was mentioning PG_CONTROL_VERSION instead. Victor Snezhko Tom Lane 2006-06-03 16:49:32 +00:00
  • b58ed707ce Fix copy-and-pasteo in Russian translation: message complaining about HAVE_INT64_TIMESTAMP was mentioning PG_CONTROL_VERSION instead. Victor Snezhko Tom Lane 2006-06-03 16:49:15 +00:00