aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
* doc: Re-order Table 28.2 "Collected Statistics Views"Peter Eisentraut2023-01-02
| | | | | | | | | Make the order more sensible and repeatable. Comments have been added to guide future additions. Author: Peter Smith <peter.b.smith@fujitsu.com> Reviewed-by: David G. Johnston <david.g.johnston@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/CAHut+Pv8Oa7v06hJb3+HzCtM2u-3oHWMdvXVHhvi7ofB83pNbg@mail.gmail.com
* Accept "+infinity" in date and timestamp[tz] input.Tom Lane2023-01-01
| | | | | | | | | The float and numeric types accept this variant spelling of "infinity", so it seems like the datetime types should too. Vik Fearing, some cosmetic mods by me Discussion: https://postgr.es/m/d0bef637-2dbd-0a5d-e539-48243b6f6c5e@postgresfriends.org
* In plpgsql, don't preassign portal names to bound cursor variables.Tom Lane2023-01-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | A refcursor variable that is bound to a specific query (by declaring it with "CURSOR FOR") now chooses a portal name in the same way as an unbound, plain refcursor variable. Its string value starts out as NULL, and unless that's overridden by manual assignment, it will be replaced by a unique-within-session portal name during OPEN. The previous behavior was to initialize such variables to contain their own name, resulting in that also being the portal name unless the user overwrote it before OPEN. The trouble with this is that it causes failures due to conflicting portal names if the same cursor variable name is used in different functions. It is pretty non-orthogonal to have bound and unbound refcursor variables behave differently on this point, too, so let's change it. This change can cause compatibility problems for applications that open a bound cursor in a plpgsql function and then use it in the calling code without explicitly passing back the refcursor value (portal name). If the calling code simply assumes that the portal name matches the called function's variable name, it will now fail. That can be fixed by explicitly assigning a string value to the refcursor variable before OPEN, e.g. DECLARE myc CURSOR FOR SELECT ...; BEGIN myc := 'myc'; -- add this OPEN myc; We have no documentation examples showing the troublesome usage pattern, so we can hope it's rare in practice. Patch by me; thanks to Pavel Stehule and Jan Wieck for review. Discussion: https://postgr.es/m/1465101.1667345983@sss.pgh.pa.us
* ci: Change macOS builds from Intel to ARM.Thomas Munro2023-01-01
| | | | | | | | | | | | Cirrus is about to shut down its macOS-on-Intel support, so it's time to move our CI testing over to ARM instances. The Homebrew package manager changed its default installation prefix for the new architecture, so a couple of tests need tweaks to find binaries. Back-patch to 15, where in-tree CI began. Author: Justin Pryzby <pryzby@telsasoft.com> Discussion: https://postgr.es/m/20221122225744.GF11463%40telsasoft.com
* Sample postgres_fdw tables remotely during ANALYZETomas Vondra2022-12-30
| | | | | | | | | | | | | | | | | When collecting ANALYZE sample on foreign tables, postgres_fdw fetched all rows and performed the sampling locally. For large tables this means transferring and immediately discarding large amounts of data. This commit allows the sampling to be performed on the remote server, transferring only the much smaller sample. The sampling is performed using the built-in TABLESAMPLE methods (system, bernoulli) or random() function, depending on the remote server version. Remote sampling can be enabled by analyze_sampling on the foreign server and/or foreign table, with supported values 'off', 'auto', 'system', 'bernoulli' and 'random'. The default value is 'auto' which uses either 'bernoulli' (TABLESAMPLE method) or 'random' (for remote servers without TABLESAMPLE support).
* Fix assert in BRIN build_distancesTomas Vondra2022-12-30
| | | | | | | | | | | | | | | | | | | | | When brin_minmax_multi_union merges summaries, we may end up with just a single range after merge_overlapping_ranges. The summaries may contain just one range each, and they may overlap (or be exactly the same). With a single range there's no distance to calculate, but we happen to call build_distances anyway - which is fine, we don't calculate the distance in this case, except that with asserts this failed due to a check there are at least two ranges. The assert is unnecessarily strict, so relax it a bit and bail out if there's just a single range. The relaxed assert would be enough, but this way we don't allocate unnecessary memory for distance. Backpatch to 14, where minmax-multi opclasses were introduced. Reported-by: Jaime Casanova Backpatch-through: 14 Discussion: https://postgr.es/m/YzVA55qS0hgz8P3r@ahch-to
* Fix precision handling for some COERCE_SQL_SYNTAX functionsMichael Paquier2022-12-30
| | | | | | | | | | | | | | | | | | | | | | | f193883 has been incorrectly setting up the precision used in the timestamp compilations returned by the following functions: - LOCALTIME - LOCALTIMESTAMP - CURRENT_TIME - CURRENT_TIMESTAMP Specifying an out-of-range precision for CURRENT_TIMESTAMP and LOCALTIMESTAMP was raising a WARNING without adjusting the precision, leading to a subsequent error. LOCALTIME and CURRENT_TIME raised a WARNING without an error, still the precision given to the internal routines was not correct, so let's be clean. Ian has reported the problems in timestamp.c, while I have noticed the ones in date.c. Regression tests are added for all of them with precisions high enough to provide coverage for the warnings, something that went missing up to this commit. Author: Ian Lawrence Barwick, Michael Paquier Discussion: https://postgr.es/m/CAB8KJ=jQEnn9sYG+N752spt68wMrhmT-ocHCh4oeNmHF82QMWA@mail.gmail.com
* Change argument of appendBinaryStringInfo from char * to void *Peter Eisentraut2022-12-30
| | | | | | | | | There is some code that uses this function to assemble some kind of packed binary layout, which requires a bunch of casts because of this. Functions taking binary data plus length should take void * instead, like memcpy() for example. Discussion: https://www.postgresql.org/message-id/flat/a0086cfc-ff0f-2827-20fe-52b591d2666c%40enterprisedb.com
* Use appendStringInfoString instead of appendBinaryStringInfo where possiblePeter Eisentraut2022-12-30
| | | | | | | | | For the jsonpath output, we don't need to squeeze out every bit of performance, so instead use a more robust coding style. There are similar calls in jsonb.c, which we leave alone here since there is indeed a performance impact for bulk exports. Discussion: https://www.postgresql.org/message-id/flat/a0086cfc-ff0f-2827-20fe-52b591d2666c%40enterprisedb.com
* Add const to BufFileWritePeter Eisentraut2022-12-30
| | | | | | | | Make data buffer argument to BufFileWrite a const pointer and bubble this up to various callers and related APIs. This makes the APIs clearer and more consistent. Discussion: https://www.postgresql.org/message-id/flat/11dda853-bb5b-59ba-a746-e168b1ce4bdb%40enterprisedb.com
* Remove unnecessary castsPeter Eisentraut2022-12-30
| | | | | | | | Some code carefully cast all data buffer arguments for data write and read function calls to void *, even though the respective arguments are already void *. Remove this unnecessary clutter. Discussion: https://www.postgresql.org/message-id/flat/11dda853-bb5b-59ba-a746-e168b1ce4bdb%40enterprisedb.com
* perl: Hide warnings inside perl.h when using gcc compatible compilerAndres Freund2022-12-29
| | | | | | | | | | | | | | | | | | | | | | | | | New versions of perl trigger warnings within perl.h with our compiler flags. At least -Wdeclaration-after-statement, -Wshadow=compatible-local are known to be problematic. To avoid these warnings, conditionally use #pragma GCC system_header before including plperl.h. Alternatively, we could add the include paths for problematic headers with -isystem, but that is a larger hammer and is harder to search for. A more granular alternative would be to use #pragma GCC diagnostic push/ignored/pop, but gcc warns about unknown warnings being ignored, so every to-be-ignored-temporarily compiler warning would require its own pg_config.h symbol and #ifdef. As the warnings are voluminous, it makes sense to backpatch this change. But don't do so yet, we first want gather buildfarm coverage - it's e.g. possible that some compiler claiming to be gcc compatible has issues with the pragma. Author: Andres Freund <andres@anarazel.de> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: Discussion: https://postgr.es/m/20221228182455.hfdwd22zztvkojy2@awork3.anarazel.de
* Fix oversight in 7a05425d96Andrew Dunstan2022-12-29
| | | | | This patch was changed as a result of review but one line didn't get the message. Mea Culpa.
* Add page-level freezing to VACUUM.Peter Geoghegan2022-12-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Teach VACUUM to decide on whether or not to trigger freezing at the level of whole heap pages. Individual XIDs and MXIDs fields from tuple headers now trigger freezing of whole pages, rather than independently triggering freezing of each individual tuple header field. Managing the cost of freezing over time now significantly influences when and how VACUUM freezes. The overall amount of WAL written is the single most important freezing related cost, in general. Freezing each page's tuples together in batch allows VACUUM to take full advantage of the freeze plan WAL deduplication optimization added by commit 9e540599. Also teach VACUUM to trigger page-level freezing whenever it detects that heap pruning generated an FPI. We'll have already written a large amount of WAL just to do that much, so it's very likely a good idea to get freezing out of the way for the page early. This only happens in cases where it will directly lead to marking the page all-frozen in the visibility map. In most cases "freezing a page" removes all XIDs < OldestXmin, and all MXIDs < OldestMxact. It doesn't quite work that way in certain rare cases involving MultiXacts, though. It is convenient to define "freeze the page" in a way that gives FreezeMultiXactId the leeway to put off the work of processing an individual tuple's xmax whenever it happens to be a MultiXactId that would require an expensive second pass to process aggressively (allocating a new multi is especially worth avoiding here). FreezeMultiXactId is eager when processing is cheap (as it usually is), and lazy in the event of an individual multi that happens to require expensive second pass processing. This avoids regressions related to processing of multis that page-level freezing might otherwise cause. Author: Peter Geoghegan <pg@bowt.ie> Reviewed-By: Jeff Davis <pgsql@j-davis.com> Reviewed-By: Andres Freund <andres@anarazel.de> Discussion: https://postgr.es/m/CAH2-WzkFok_6EAHuK39GaW4FjEFQsY=3J0AAd6FXk93u-Xq3Fg@mail.gmail.com
* Convert contrib/ltree's input functions to report errors softlyAndrew Dunstan2022-12-28
| | | | | | Reviewed by Tom Lane and Amul Sul Discussion: https://postgr.es/m/49e598c2-cfe8-0928-b6fb-d0cc51aab626@dunslane.net
* Convert contrib/intarray's bqarr_in() to report errors softlyAndrew Dunstan2022-12-28
| | | | | | Reviewed by Tom Lane and Amul Sul Discussion: https://postgr.es/m/49e598c2-cfe8-0928-b6fb-d0cc51aab626@dunslane.net
* Reorder some object files in makefilesPeter Eisentraut2022-12-28
| | | | | This restores some once-intended alphabetical orders and makes the lists consistent between the different build systems.
* Suppress uninitialized-variable warning from a61b1f748.Tom Lane2022-12-27
| | | | | | | | | | | | Some compilers complain about sub_rteperminfos not being initialized, evidently because they don't detect that it is only used and set if isGeneralSelect is true. Make it follow the long-established pattern for its sibling variable sub_rtable. Per reports from Pavel Stehule and the buildfarm. Discussion: https://postgr.es/m/CAFj8pRDOvGOi-n616kM0Cc7qSbg_nGoS=-haB+D785sUXADqSg@mail.gmail.com
* Convert hstore_in to report errors softly.Tom Lane2022-12-27
| | | | | | | | The error reporting here was not only old and crufty, but untested. I took the opportunity to bring the messages into some sort of compliance with our message style guidelines. Discussion: https://postgr.es/m/6B6A5C77-60AD-4A71-9F3A-B2C026A281A6@dunslane.net
* Remove new locale dependency in regproc regression test.Tom Lane2022-12-27
| | | | | | | | | | | The modified error message for regcollationin failure includes the database encoding, which it should've occurred to me is a portability hazard for the regression tests. Adjust the test so the expected output doesn't include that. In passing, fix a comment typo introduced in b8c0ffbd2. Per buildfarm.
* Simplify the implementations of the to_reg* functions.Tom Lane2022-12-27
| | | | | | | | | | | | | | | Given the soft-input-error feature, we can reduce these functions to be just thin wrappers around a soft-error call of the corresponding datatype input function. This means less code and more certainty that the to_reg* functions match the normal input behavior. Notably, it also means that they will accept numeric OID input, which they didn't before. It's not clear to me if that omission had more than laziness behind it, but it doesn't seem like something we need to work hard to preserve. Discussion: https://postgr.es/m/3910031.1672095600@sss.pgh.pa.us
* Convert the reg* input functions to report (most) errors softly.Tom Lane2022-12-27
| | | | | | | | | | | | | | | | | | | | | | | | This is not really complete, but it catches most cases of practical interest. The main omissions are: * regtype, regprocedure, and regoperator parse type names by calling the main grammar, so any grammar-detected syntax error will still be a hard error. Also, if one includes a type modifier in such a type specification, errors detected by the typmodin function will be hard errors. * Lookup errors are handled just by passing missing_ok = true to the relevant catalog lookup function. Because we've used quite a restrictive definition of "missing_ok", this means that edge cases such as "the named schema exists, but you lack USAGE permission on it" are still hard errors. It would make sense to me to replace most/all missing_ok parameters with an escontext parameter and then allow these additional lookup failure cases to be trapped too. But that's a job for some other day. Discussion: https://postgr.es/m/3342239.1671988406@sss.pgh.pa.us
* Convert tsqueryin and tsvectorin to report errors softly.Tom Lane2022-12-27
| | | | | | | | | | | | | | | This is slightly tedious because the adjustments cascade through a couple of levels of subroutines, but it's not very hard. I chose to avoid changing function signatures more than absolutely necessary, by passing the escontext pointer in existing structs where possible. tsquery's nuisance NOTICEs about empty queries are suppressed in soft-error mode, since they're not errors and we surely don't want them to be shown to the user anyway. Maybe that whole behavior should be reconsidered. Discussion: https://postgr.es/m/3824377.1672076822@sss.pgh.pa.us
* Detect bad input for types xid, xid8, and cid.Tom Lane2022-12-27
| | | | | | | | | | | | | | | | | | | | | | Historically these input functions just called strtoul or strtoull and returned the result, with no error detection whatever. Upgrade them to reject garbage input and out-of-range values, similarly to our other numeric input routines. To share the code for this with type oid, adjust the existing "oidin_subr" to be agnostic about the SQL name of the type it is handling, and move it to numutils.c; then clone it for 64-bit types. Because the xid types previously accepted hex and octal input by reason of calling strtoul[l] with third argument zero, I made the common subroutine do that too, with the consequence that type oid now also accepts hex and octal input. In view of 6fcda9aba, that seems like a good thing. While at it, simplify the existing over-complicated handling of syntax errors from strtoul: we only need one ereturn not three. Discussion: https://postgr.es/m/3526121.1672000729@sss.pgh.pa.us
* Remove overzealous MultiXact freeze assertion.Peter Geoghegan2022-12-26
| | | | | | | | | | | | | When VACUUM determines that an existing MultiXact should use a freeze plan that sets xmax to InvalidTransactionId, the original Multi may or may not be before OldestMxact. Remove an incorrect assertion that expected it to always be from before OldestMxact. Oversight in commit 4ce3af. Author: Peter Geoghegan <pg@bowt.ie> Reported-By: Hayato Kuroda <kuroda.hayato@fujitsu.com> Discussion: https://postgr.es/m/TYAPR01MB5866B24104FD80B5D7E65C3EF5ED9@TYAPR01MB5866.jpnprd01.prod.outlook.com
* Add custom filtering rules to the TAP tests of pg_upgradeMichael Paquier2022-12-27
| | | | | | | | | | | | | | | | | | | | 002_pg_upgrade.pl gains support for a new environment variable called "filter_rules", that can be used to point to a file that includes a set of custom regular expressions that would be applied to the dumps of the origin and target clusters when doing a cross-version test (aka when defining olddump and oldinstall), to give the possibility to reshape dynamically the dumps in the same way as the internals of the buildfarm code so as the tests are able to pass in scenarios where one expects them to even if pg_dump generates slightly-different outputs depending on the versions involved. This option is not used when pg_upgrade runs with the same version for the origin and target clusters, and it is the last piece I see as required to be able to plug-in more efficiently the TAP tests of pg_upgrade with the buildfarm or just a CI. Author: Anton A. Melnikov Discussion: https://postgr.es/m/49f389ba-95ce-8a9b-09ae-f60650c0e7c7@inbox.ru
* Fix incorrect copy-pasto in error message of pg_waldump.cMichael Paquier2022-12-27
| | | | | The error message used on fclose() failure was incorrect, so fix it. Oversight in d497093, that I have somehow managed to miss.
* pg_waldump: Add --save-fullpage=PATH to save full page images from WAL recordsMichael Paquier2022-12-27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This option extracts (potentially decompressing) full-page images included in WAL records into a given target directory. These images are subject to the same filtering rules as the normal display of the WAL records, hence with --relation one can for example extract only the FPIs issued on the relation defined. By default, the records are printed or their stats computed (--stats), using --quiet would only save the images without any output generated. This is a tool aimed mostly for very experienced users, useful for fixing page-level corruption or just analyzing the past state of a page, and there were no easy way to do that with the in-core tools up to now when looking at WAL. Each block is saved in a separate file, to ease their manipulation, with the file respecting <lsn>.<ts>.<db>.<rel>.<blk>_<fork> with as format. For instance, 00000000-010000C0.1663.1.6117.123_main refers to: - WAL record LSN in hexa format (00000000-010000C0). - Tablespace OID (1663). - Database OID (1). - Relfilenode (6117). - Block number (123). - Fork name of the file this block came from (_main). Author: David Christensen Reviewed-by: Sho Kato, Justin Pryzby, Bharath Rupireddy, Matthias van de Meent Discussion: https://postgr.es/m/CAOxo6XKjQb2bMSBRpePf3ZpzfNTwjQUc4Tafh21=jzjX6bX8CA@mail.gmail.com
* Add 'logical_decoding_mode' GUC.Amit Kapila2022-12-26
| | | | | | | | | | | | | | | | This enables streaming or serializing changes immediately in logical decoding. This parameter is intended to be used to test logical decoding and replication of large transactions for which otherwise we need to generate the changes till logical_decoding_work_mem is reached. This helps in reducing the timing of existing tests related to logical replication of in-progress transactions and will help in writing tests for for the upcoming feature for parallelly applying large in-progress transactions. Author: Shi yu Reviewed-by: Sawada Masahiko, Shveta Mallik, Amit Kapila, Dilip Kumar, Kuroda Hayato, Kyotaro Horiguchi Discussion: https://postgr.es/m/OSZPR01MB63104E7449DBE41932DB19F1FD1B9@OSZPR01MB6310.jpnprd01.prod.outlook.com
* Switch query fixing aclitems in ~15 from O(N^2) to O(N) in upgrade_adapt.sqlMichael Paquier2022-12-26
| | | | | | | | | | | | f4f2f2b was doing a sequential scan of pg_class before checking if a relation had attributes dependent on aclitem as data typewhen building the set of ALTER TABLE queries, but it would be costly on a regression database. While on it, make the query style more consistent with the rest. Reported-by: Justin Pryzby Discussion: https://postgr.es/m/20221223032724.GQ1153@telsasoft.com
* Convert enum_in() to report errors softly.Tom Lane2022-12-25
| | | | | | | | I missed this in my initial survey, probably because I examined the contents of pg_type in the postgres database, which lacks any enumerated types. Discussion: https://postgr.es/m/CAAJ_b97KeDWUdpTKGOaFYPv0OicjOu6EW+QYWj-Ywrgj_aEy1g@mail.gmail.com
* Fix thinko in 720e0327bcAndrew Dunstan2022-12-25
|
* Convert contrib/isn's input functions to report errors softlyAndrew Dunstan2022-12-24
|
* Convert jsonpath's input function to report errors softlyAndrew Dunstan2022-12-24
| | | | | | Reviewed by Tom Lane Discussion: https://postgr.es/m/a8dc5700-c341-3ba8-0507-cc09881e6200@dunslane.net
* Make the numeric-OID cases of regprocin and friends be non-throwing.Tom Lane2022-12-24
| | | | | | | | | While at it, use a common subroutine already. This doesn't move the needle very far in terms of making these functions non-throwing; the only case we're now able to trap is numeric-OID-is-out-of-range. Still, it seems like a pretty non-controversial step in that direction.
* Fix recent accidental omission in pg_proc.datDavid Rowley2022-12-24
| | | | | | | | | | | | | ed1a88dda added support functions for the ntile(), percent_rank() and cume_dist() window functions but neglected to actually add these support functions to the pg_proc entry for the corresponding window function. Also, take this opportunity to add these window functions to one of the regression tests added in ed1a88dda to give the support functions a little bit of exercise. If I'd done that in the first place then the omission would have been more obvious. Bump the catversion, again.
* Rework <warning> box about column list combining in logical replicationAlvaro Herrera2022-12-23
| | | | | | | | | | | | | | After some copy-edit I made in commit 3a06a79cd137, we have a <sect2> that only contains a warning box. This doesn't look good. Rework by moving the sect2 title to be the warning's title, and put the 'id' to it as well, so that the external reference continues to work. Backpatch to 15. In branch master, I also take the opportunity to add titles to a couple of other warning boxes elsewhere in the documentation. Discussion: https://postgr.es/m/20221219164713.ccnlvtkyj6lmshqq@alvherre.pgsql
* Fix end LSN determination in recently added testAlvaro Herrera2022-12-23
| | | | | | | | | | | | | | | | | | The test added in commit e44dae07f931 has a thinko: it wants to read info about a few WAL records, but it obtains the LSN of the final record to read by asking for the WAL insert position; however, pg_get_wal_records_info only accepts to read up to the flush position (cf. IsFutureLSN()). In normal conditions there is no difference, since the last record written by the preceding loop is known flushed and it's the one the test wants; but it's possible to have some other process insert another WAL record that isn't flushed, and that causes the whole test to explode. Fix by having pg_get_wal_records_info() read only up to the flushed position. Backpatch to 15, which is where pg_walinspect appeared. Author: Karina Litskevich <litskevichkarina@gmail.com> Discussion: https://postgr.es/m/a5559c95-52c3-5eea-cd63-9b4f1c70ff96@gmail.com
* Convert contrib/seg's input function to report errors softlyAndrew Dunstan2022-12-23
| | | | | | Reviewed by Tom Lane Discussion: https://postgr.es/m/a8dc5700-c341-3ba8-0507-cc09881e6200@dunslane.net
* Fix event trigger exampleAlvaro Herrera2022-12-23
| | | | | | | | | | | | | Commit 2f9661311b changed command tags from strings to numbers, but forgot to adjust the code in the event trigger example, which consequently failed to compile. While fixing that, improve the indentation to adhere to pgindent style. Backpatch to v13, where the change was introduced. Author: Laurenz Albe Discussion: https://postgr.es/m/81e36ac17dc80489e74dc5b6914afa6ccdb1a99d.camel@cybertec.at
* Fix bug in translate_col_privs_multilevelDavid Rowley2022-12-24
| | | | | | | | | | | | | Fix incorrect code which was trying to convert a Bitmapset of columns at the attnums according to a parent table and transform them into the equivalent Bitmapset with same attnums according to the given child table. This code is new as of a61b1f748 and was failing to do the correct translation when there was an intermediate parent table between 'rel' and 'top_parent_rel'. Reported-by: Ranier Vilela Author: Richard Guo, Amit Langote Discussion: https://postgr.es/m/CAEudQArohfB_Gy%2BhcH2-bANUkxgjJiP%3DABq01_LgTNTbcNijag%40mail.gmail.com
* Allow parent's WaitEventSets to be freed after fork().Thomas Munro2022-12-23
| | | | | | | | | | | | | An epoll fd belonging to the parent should be closed in the child. A kqueue fd is automatically closed by fork(), but we should still adjust our counter. For poll and Windows systems, nothing special is required. On all systems we free the memory. No caller yet, but we'll need this if we start using WaitEventSet in the postmaster as planned. Reviewed-by: Andres Freund <andres@anarazel.de> Discussion: https://postgr.es/m/CA%2BhUKG%2BZ-HpOj1JsO9eWUP%2Bar7npSVinsC_npxSy%2BjdOMsx%3DGg%40mail.gmail.com
* Don't leak a signalfd when using latches in the postmaster.Thomas Munro2022-12-23
| | | | | | | | | | | At the time of commit 6a2a70a02 we didn't use latch infrastructure in the postmaster. We're planning to start doing that, so we'd better make sure that the signalfd inherited from a postmaster is not duplicated and then leaked in the child. Reviewed-by: Andres Freund <andres@anarazel.de> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Discussion: https://postgr.es/m/CA%2BhUKG%2BZ-HpOj1JsO9eWUP%2Bar7npSVinsC_npxSy%2BjdOMsx%3DGg%40mail.gmail.com
* Add WL_SOCKET_ACCEPT event to WaitEventSet API.Thomas Munro2022-12-23
| | | | | | | | | | | | | | | To be able to handle incoming connections on a server socket with the WaitEventSet API, we'll need a new kind of event to indicate that the the socket is ready to accept a connection. On Unix, it's just the same as WL_SOCKET_READABLE, but on Windows there is a different underlying kernel event that we need to map our abstraction to. No user yet, but a proposed patch would use this. Reviewed-by: Andres Freund <andres@anarazel.de> Discussion: https://postgr.es/m/CA%2BhUKG%2BZ-HpOj1JsO9eWUP%2Bar7npSVinsC_npxSy%2BjdOMsx%3DGg%40mail.gmail.com
* Update upgrade_adapt.sql to handle tables using aclitem as data typeMichael Paquier2022-12-23
| | | | | | | | | | | | | | | | | | | | | The regression test suite includes a table called "tab_core_types" that has one attribute based on the type "aclitem". Keeping this attribute as-is causes hard failures when running pg_upgrade with an origin on ~15. This commit updates upgrade_adapt.sql to automatically detect the tables with such attributes and switch them to text so as pg_upgrade is able to go through its run. This does not provide the same detection coverage as pg_upgrade, where we are able to find out aclitems used in arrays, domains or even composite types, but this is (I guess) enough for most things like an instance that had installcheck run on before the upgrade with a dump generated from it. Note that the buildfarm code has taken the simplest approach of just dropping "tab_core_types", so what we have here is more modular. Author: Anton A. Melnikov Discussion: https://postgr.es/m/49f389ba-95ce-8a9b-09ae-f60650c0e7c7@inbox.ru
* Fix some incorrectness in upgrade_adapt.sql on query for WITH OIDSMichael Paquier2022-12-23
| | | | | | | | | | | | | The query used to disable WITH OIDS in all the relations making use of it was checking for materialized views, but this is not a supported operation. On the contrary, this needs to be done on foreign tables. While on it, use quote_ident() in the ALTER TABLE strings built on the relation name. Author: Anton A. Melnikov, Michael Paquier Discussion: https://postgr.es/m/49f389ba-95ce-8a9b-09ae-f60650c0e7c7@inbox.ru Backpatch-through: 12
* Fix come incorrect elog() messages in aclchk.cMichael Paquier2022-12-23
| | | | | | | | | | | | | | Three error strings used with cache lookup failures were referring to incorrect object types for ACL checks: - Schemas - Types - Foreign Servers There errors should never be triggered, but if they do incorrect information would be reported. Author: Justin Pryzby Discussion: https://postgr.es/m/20221222153041.GN1153@telsasoft.com Backpatch-through: 11
* Rename pg_dissect_walfile_name() to pg_split_walfile_name()Michael Paquier2022-12-23
| | | | | | | | | | | | | | | The former name was discussed as being confusing, so use "split", as per a suggestion from Magnus Hagander. While on it, one of the output arguments is renamed from "segno" to "segment_number", as per a suggestion from Kyotaro Horiguchi. The documentation is updated to reflect all these changes. Bump catalog version. Author: Bharath Rupireddy, Michael Paquier Discussion: https://postgr.es/m/CABUevEytQVaOOhGdoh0D7hGwe3fuKcRF6NthsSW7ww04EmtFgQ@mail.gmail.com
* Allow window functions to adjust their frameOptionsDavid Rowley2022-12-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | WindowFuncs such as row_number() don't care if it's called with ROWS UNBOUNDED PRECEDING AND CURRENT ROW or with RANGE UNBOUNDED PRECEDING AND CURRENT ROW. The latter is less efficient as the RANGE option requires that the executor check for peer rows, so using the ROW option instead would cause less overhead. Because RANGE is part of the default frame options for WindowClauses, it means WindowAgg is, by default, working much harder than it needs to for window functions where the ROWS / RANGE option has no effect on the window function's result. On a test query from the discussion thread, a performance improvement of 344% was seen by using ROWS instead of RANGE. Here we add a new support function node type to allow support functions to be called for window functions so that the most optimal version of the frame options can be set. The planner has been adjusted so that the frame options are changed only if all window functions sharing the same window clause agree on what the optimized frame options are. Here we give the ability for row_number(), rank(), dense_rank(), percent_rank(), cume_dist() and ntile() to alter their WindowClause's frameOptions. Reviewed-by: Vik Fearing, Erwin Brandstetter, Zhihong Yu Discussion: https://postgr.es/m/CAGHENJ7LBBszxS+SkWWFVnBmOT2oVsBhDMB1DFrgerCeYa_DyA@mail.gmail.com Discussion: https://postgr.es/m/CAApHDvohAKEtTXxq7Pc-ic2dKT8oZfbRKeEJP64M0B6+S88z+A@mail.gmail.com
* Improve notation of cacheinfo table in syscache.c.Thomas Munro2022-12-23
| | | | | | | | | | Use C99 designated initializer syntax for the array elements, instead of writing the enumerator name and position in a comment. Replace nkeys and key with a local variadic macro, for a shorter notation. Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Discussion: https://postgr.es/m/CA%2BhUKGKdpDjKL2jgC-GpoL4DGZU1YPqnOFHbDqFkfRQcPaR5DQ%40mail.gmail.com
generated by cgit v1.2.3 (git 2.25.1) at 2025-08-03 06:53:24 +0000