Re: [HACKERS] [POC] Faster processing at Gather node

On Sun, Nov 5, 2017 at 6:54 AM, Andres Freund wrote > On 2017-11-05 01:05:59 +0100, Robert Haas wrote: >> skip-gather-project-v1.patch does what it says on the tin. I still >> don't have a test case for this, and I didn't find that it helped very >> much, I am also wondering

Re: [HACKERS] Flexible configuration for full-text search

On Sat, Oct 21, 2017 at 1:39 AM, Aleksandr Parfenov wrote: > In attachment updated patch with fixes of empty XML tags in > documentation. Hi Aleksandr, I'm not sure if this is expected at this stage, but just in case you aren't aware, with this version of the patch

Re: [HACKERS] [PATCH] Add ALWAYS DEFERRED option for constraints

On Fri, Oct 20, 2017 at 9:05 AM, Nico Williams wrote: > Rebased (there were conflicts in the SGML files). Hi Nico FYI that version has some stray absolute paths in constraints.source: -COPY COPY_TBL FROM '@abs_srcdir@/data/constro.data'; +COPY COPY_TBL FROM

Re: [Sender Address Forgery]Re: [HACKERS] path toward faster partition pruning

On 2017/11/06 12:53, David Rowley wrote: > On 3 November 2017 at 17:32, David Rowley > wrote: >> 2. This code is way more complex than it needs to be. >> >> if (num_parts > 0) >> { >> int j; >> >> all_indexes = (int *) palloc(num_parts * sizeof(int)); >> j = 0; >>

Re: [HACKERS] [PATCH] Overestimated filter cost and its mitigation

On Mon, Sep 11, 2017 at 7:43 PM, Yuto Hayamizu wrote: > Suppose that there are three qual clauses in a scan node, current > postgres estimates per-tuple cost of the filter to be: >cost(A) + cost(B) + cost(C) > > And basic idea of the attached patch is: >cost(A) +

[HACKERS] PATCH: psql tab completion for SELECT

Hi pgsql-hackers, Here's a little draft patch to add *some* tab completion ability for SELECT in psql. I have often missed the ability, especially with invocations of utility functions. It would be nice to be able to suggest column names from the relevant tables in the query, but, as the SQL

Re: [HACKERS] Setting pd_lower in GIN metapage

On Sat, Nov 4, 2017 at 2:03 AM, Tom Lane wrote: > Michael Paquier writes: >> On Fri, Nov 3, 2017 at 1:10 AM, Amit Kapila wrote: >>> On Fri, Nov 3, 2017 at 2:54 AM, Tom Lane wrote: I've marked the

Re: [HACKERS] path toward faster partition pruning

On 31 October 2017 at 21:43, Amit Langote wrote: > Attached updated version of the patches match_clauses_to_partkey() needs to allow for the way quals on Bool columns are represented. create table pt (a bool not null) partition by list (a); create table pt_true

Re: [HACKERS] How to implement a SP-GiST index as a extension module?

Never mind, it turns out the issue boiled down to me declaring the wrong prefixType in my config function. TL;DR - PEBKAC On Sun, Nov 5, 2017 at 1:09 AM, Connor Wolf wrote: > Ok, I've got everything compiling and it installs properly, but I'm > running into

Re: [HACKERS] path toward faster partition pruning

On 3 November 2017 at 17:32, David Rowley wrote: > 2. This code is way more complex than it needs to be. > > if (num_parts > 0) > { > int j; > > all_indexes = (int *) palloc(num_parts * sizeof(int)); > j = 0; > if (min_part_idx >= 0 && max_part_idx >= 0) > { > for (i

Re: [HACKERS] Early locking option to parallel backup

Andres Freund writes: > Well, the current approach afaics requires #relations * 2 locks, whereas > acquiring them in every worker would scale that with the number of > workers. Yeah, that's gonna be a problem with this proposal. > IIUC the problem here is that even though a

Re: [HACKERS] why not parallel seq scan for slow functions

Robert Haas writes: > This looks like it's on the right track to me. I hope Tom will look > into it, but if he doesn't I may try to get it committed myself. I do plan to take a look at it during this CF. > +/* Set or update cheapest_total_path and related fields

Re: [HACKERS] Restricting maximum keep segments by repslots

On 31 October 2017 at 17:43, Kyotaro HORIGUCHI wrote: > Hello, this is a rebased version. > > It gets a change of the meaning of monitoring value along with > rebasing. > > In previous version, the "live" column mysteriously predicts the > necessary segments will

Re: [HACKERS] Display number of heap accesses for index scans

Peter Geoghegan writes: > Andres Freund wrote: >> The number of index lookups that failed to return anything can be a >> critical performance factor in OLTP workloads. Therefore it seems like >> it'd be a good idea to extend the explain analyze output to

Re: [HACKERS] Custom compression methods

Robert Haas writes: > A basic problem here is that, as proposed, DROP COMPRESSION METHOD may > break your database irretrievably. If there's no data compressed > using the compression method you dropped, everything is cool - > otherwise everything is broken and there's no

Re: [HACKERS] dropping partitioned tables without CASCADE

On 2017/11/03 21:39, Alvaro Herrera wrote: > Ashutosh Bapat wrote: >> On Fri, Nov 3, 2017 at 1:42 PM, Alvaro Herrera >> wrote: > >>> I think adding "is partitioned" at end of line isn't good; looks like a >>> phrase but isn't translatable. Maybe add keyword PARTITIONED

Re: [HACKERS] Secondary index access optimizations

On Fri, Sep 8, 2017 at 3:58 AM, Konstantin Knizhnik wrote: > Updated version of the patch is attached to this mail. > Also I added support of date type to operator_predicate_proof to be able to > imply (logdate <= '2017-03-31') from (logdate < '2017-04-01') . Hi

Re: [HACKERS] Setting pd_lower in GIN metapage

On 2017/11/03 6:24, Tom Lane wrote: > Amit Langote writes: >> On 2017/09/26 16:30, Michael Paquier wrote: >>> Cool, let's switch it back to a ready for committer status then. > >> Sure, thanks. > > Pushed with some cosmetic adjustments --- mostly, making the

Re: [HACKERS] Restricting maximum keep segments by repslots

On Tue, Oct 31, 2017 at 10:43 PM, Kyotaro HORIGUCHI wrote: > Hello, this is a rebased version. Hello Horiguchi-san, I think the "ddl" test under contrib/test_decoding also needs to be updated because it looks at pg_replication_slots and doesn't expect your new

Re: [HACKERS] Custom compression methods

> If there's no data compressed > using the compression method you dropped, everything is cool - > otherwise everything is broken and there's no way to recover. > The only obvious alternative is to disallow DROP altogether (or make it > not really DROP). Wouldn't whatever was using the

Re: [HACKERS] Statement-level rollback

From: pgsql-hackers-ow...@postgresql.org > [mailto:pgsql-hackers-ow...@postgresql.org] On Behalf Of Simon Riggs > A backend-based solution is required for PL procedures and functions. > > We could put this as an option into PL/pgSQL, but it seems like it is > a function of the transaction manager

Re: [HACKERS] Early locking option to parallel backup

On 2017-11-05 17:38:39 -0500, Robert Haas wrote: > On Sun, Nov 5, 2017 at 5:17 AM, Lucas wrote: > > The patch creates a "--lock-early" option which will make pg_dump to issue > > shared locks on all tables on the backup TOC on each parallel worker start. > > That way, the

Re: [HACKERS] Early locking option to parallel backup

On Sun, Nov 5, 2017 at 5:17 AM, Lucas wrote: > The patch creates a "--lock-early" option which will make pg_dump to issue > shared locks on all tables on the backup TOC on each parallel worker start. > That way, the backup has a very small chance of failing. When it does, >

Re: [HACKERS] Custom compression methods

On Sun, Nov 5, 2017 at 2:22 PM, Oleg Bartunov wrote: >> IIRC there were some concerns about what happened with pg_upgrade, >> with consuming precious toast bits, and a few other things. > > yes, pg_upgrade may be a problem. A basic problem here is that, as proposed, DROP

Re: [HACKERS] why not parallel seq scan for slow functions

On Sun, Nov 5, 2017 at 12:57 AM, Amit Kapila wrote: > Thanks for the confirmation. Find rebased patch attached. This looks like it's on the right track to me. I hope Tom will look into it, but if he doesn't I may try to get it committed myself. -if

Re: [HACKERS] [POC] Faster processing at Gather node

Hi, On November 5, 2017 1:33:24 PM PST, Jim Van Fleet wrote: >Ran this change with hammerdb on a power 8 firestone > >with 2 socket, 20 core >9.6 base-- 451991 NOPM >0926_master -- 464385 NOPM >11_04master -- 449177 NOPM >11_04_patch -- 431423 NOPM >-- two socket

Re: [HACKERS] [POC] Faster processing at Gather node

Ran this change with hammerdb on a power 8 firestone with 2 socket, 20 core 9.6 base-- 451991 NOPM 0926_master -- 464385 NOPM 11_04master -- 449177 NOPM 11_04_patch -- 431423 NOPM -- two socket patch is a little down from previous base runs With one socket 9.6 base -- 393727

Re: [HACKERS] Display number of heap accesses for index scans

Andres Freund wrote: The number of index lookups that failed to return anything can be a critical performance factor in OLTP workloads. Therefore it seems like it'd be a good idea to extend the explain analyze output to include that information. I certainly agree. I've

Re: [HACKERS] WIP: long transactions on hot standby feedback replica / proof of concept

On Fri, Nov 3, 2017 at 5:57 PM, Alexander Korotkov wrote: > On Thu, Nov 2, 2017 at 5:56 AM, Robert Haas wrote: >> > I can propose following alternative approach: teach read-only queries on >> > hot >> > standby to tolerate concurrent relation

Re: [HACKERS] Custom compression methods

On Thu, Nov 2, 2017 at 6:02 PM, Craig Ringer wrote: > On 2 November 2017 at 17:41, Ildus Kurbangaliev > wrote: > >> In this patch compression methods is suitable for MAIN and EXTENDED >> storages like in current implementation in postgres.

Re: [HACKERS] Small improvement to compactify_tuples

On Sat, Nov 4, 2017 at 8:07 PM, Юрий Соколов wrote: > 2017-11-03 5:46 GMT+03:00 Tom Lane : >> >> Sokolov Yura writes: >> > [ 0001-Improve-compactify_tuples.patch, v5 or thereabouts ] >> >> I went to check the shellsort

[HACKERS] Re: [COMMITTERS] pgsql: Account for catalog snapshot in PGXACT->xmin updates.

On Sat, Nov 04, 2017 at 12:23:36PM -0400, Tom Lane wrote: > Noah Misch writes: > > I plan to use the attached patch after the minor release tags land. If > > there's significant support, I could instead push before the wrap. > > This looks fine to me --- I think you should

Re: [HACKERS] possible encoding issues with libxml2 functions

Hi 2017-11-05 4:07 GMT+01:00 Noah Misch : > On Tue, Oct 17, 2017 at 06:06:40AM +0200, Pavel Stehule wrote: > > Please, if you can, try it write. I am little bit lost :) > > I'm attaching the patch I desired. Please review. This will probably miss > this week's minor

Re: [HACKERS] Exclude pg_internal.init from base backup

On Sat, Nov 4, 2017 at 4:04 AM, Michael Paquier wrote: > On Fri, Nov 3, 2017 at 4:04 PM, Petr Jelinek > wrote: > > Not specific problem to this patch, but I wonder if it should be made > > more clear that those files (there are couple

Re: [HACKERS] Minor comment issue in receivelog.c

On Thu, Nov 2, 2017 at 5:18 PM, Bernd Helmle wrote: > Please find a minor comment fix for receivelog.c, HandleCopyStream(). > > The comments talks about a START_STREAMING command, but i think > START_REPLICATION is what's meant here. > Yeah, it is. Confusingly enough,

Re: [HACKERS] Parallel Plans and Cost of non-filter functions

On Sat, Nov 4, 2017 at 10:02 PM, Amit Kapila wrote: > On Sat, Nov 4, 2017 at 4:43 AM, Tom Lane wrote: > > Paul Ramsey writes: > >>> Whether I get a parallel aggregate seems entirely determined by the > number > >>> of

Re: [HACKERS] [POC] Faster processing at Gather node

On Sun, Nov 5, 2017 at 2:24 AM, Andres Freund wrote: >> shm-mq-reduce-receiver-latch-set-v1.patch causes the receiver to only >> consume input from the shared queue when the amount of unconsumed >> input exceeds 1/4 of the queue size. This caused a large performance >>

Re: [HACKERS] Early locking option to parallel backup

On Sun, Nov 5, 2017 at 7:17 PM, Lucas wrote: > The patch creates a "--lock-early" option which will make pg_dump to issue > shared locks on all tables on the backup TOC on each parallel worker start. > That way, the backup has a very small chance of failing. When it does, >

Re: [HACKERS] How to implement a SP-GiST index as a extension module?

Ok, I've got everything compiling and it installs properly, but I'm running into problems that I think are either a side-effect of implementing picksplit incorrectly (likely), or a bug in SP-GiST(?). Program received signal SIGSEGV, Segmentation fault. __memcpy_sse2_unaligned () at

Re: [HACKERS] Parallel Plans and Cost of non-filter functions

On Sat, Nov 4, 2017 at 4:43 AM, Tom Lane wrote: > Paul Ramsey writes: >>> Whether I get a parallel aggregate seems entirely determined by the number >>> of rows, not the cost of preparing those rows. > >> This is true, as far as I can tell and

Re: [HACKERS] why not parallel seq scan for slow functions

On Thu, Sep 21, 2017 at 2:35 AM, Jeff Janes wrote: > On Tue, Sep 19, 2017 at 9:15 PM, Amit Kapila > wrote: >> >> On Wed, Sep 20, 2017 at 3:05 AM, Jeff Janes wrote: >> > On Tue, Sep 19, 2017 at 1:17 PM, Thomas Munro >> >

Re: [HACKERS] possible encoding issues with libxml2 functions

On Tue, Oct 17, 2017 at 06:06:40AM +0200, Pavel Stehule wrote: > Please, if you can, try it write. I am little bit lost :) I'm attaching the patch I desired. Please review. This will probably miss this week's minor releases. If there's significant support, I could instead push before the wrap.

Re: [HACKERS] [POC] Faster processing at Gather node

On 2017-11-05 01:05:59 +0100, Robert Haas wrote: > skip-gather-project-v1.patch does what it says on the tin. I still > don't have a test case for this, and I didn't find that it helped very > much, but it would probably help more in a test case with more > columns, and you said this looked like

Re: [HACKERS] [POC] Faster processing at Gather node

On Sat, Nov 4, 2017 at 5:55 PM, Andres Freund wrote: >> master: 21436.745, 20978.355, 19918.617 >> patch: 15896.573, 15880.652, 15967.176 >> >> Median-to-median, that's about a 24% improvement. > > Neat! With the attached stack of 4 patches, I get: 10811.768 ms, 10743.424 ms,

Re: [HACKERS] Small improvement to compactify_tuples

Юрий Соколов wrote: tps is also reflects changes: ~17ktps with qsort ~19ktps with bucket sort Also vacuum of benchmark's table is also improved: ~3s with qsort, ~2.4s with bucket sort One thing that you have to be careful with when it comes to our qsort with partially

Re: [HACKERS] Small improvement to compactify_tuples

2017-11-03 5:46 GMT+03:00 Tom Lane : > > Sokolov Yura writes: > > [ 0001-Improve-compactify_tuples.patch, v5 or thereabouts ] > > I went to check the shellsort algorithm against Wikipedia's entry, > and found that this appears to be an incorrect

[HACKERS] Release notes for next week's minor releases are up for review

... at https://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=42de8a0255c2509bf179205e94b9d65f9d6f3cf9 regards, tom lane -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription:

Re: [HACKERS] Support to COMMENT ON DATABASE CURRENT_DATABASE

On 10/5/17, 11:53 PM, "Jing Wang" wrote: > The patch has been updated according to Nathan's comments.  > Thanks Nathan's review. Thanks for the new versions of the patches. I apologize for the long delay for this new review. It looks like the no-pgdump patch needs a

Re: [HACKERS] Add some const decorations to prototypes

Just leave it as char*. If you change the endptr argument you're going to force every call site to change their return variable, and some of them would end up having to cast away the const on their end. OK, here is an updated patch with the controversial bits removed. I'm in general favor

[HACKERS] Display number of heap accesses for index scans

Hi, right now it's hard to figure out whether a plain indexscan returns matches that then get eliminated due to visibility checks in the heap. For both index only scans (via "Heap Fetches") and bitmapscans (via row mismatches between bitmap heap and index scans) one can gather that to some degree

Re: [HACKERS] [PATCH] A hook for session start

On Sat, Nov 4, 2017 at 1:23 AM, Michael Paquier wrote: > > On Fri, Nov 3, 2017 at 1:55 PM, Fabrízio de Royes Mello > wrote: > >> Passing the database name and user name does not look much useful to > >> me. You can have access to this data

Re: [HACKERS] [COMMITTERS] pgsql: Fix freezing of a dead HOT-updated tuple

On 2017-11-04 06:15:00 -0700, Andres Freund wrote: > The current testcase, and I think the descriptions in the relevant > threads, all actually fail to point out the precise way the bug is > triggered. As e.g. evidenced that the freeze-the-dead case appears to > not cause any failures in

Re: [HACKERS] [POC] Faster processing at Gather node

Hi, On 2017-11-04 16:38:31 +0530, Robert Haas wrote: > On hydra (PPC), these changes didn't help much. Timings: > > master: 29605.582, 29753.417, 30160.485 > patch: 28218.396, 27986.951, 26465.584 > > That's about a 5-6% improvement. On my MacBook, though, the > improvement was quite a bit

Re: [HACKERS] [COMMITTERS] pgsql: Account for catalog snapshot in PGXACT->xmin updates.

Noah Misch writes: > I plan to use the attached patch after the minor release tags land. If > there's significant support, I could instead push before the wrap. This looks fine to me --- I think you should push now. (which reminds me, I'd better get on with making release

Re: [HACKERS] taking stdbool.h into use

On 10/29/17 08:50, Michael Paquier wrote: > On Thu, Oct 26, 2017 at 9:25 AM, Peter Eisentraut > wrote: >> Here is an updated patch set. This is just a rebase of the previous >> set, no substantial changes. Based on the discussion so far, I'm >> proposing that

Re: [HACKERS] [COMMITTERS] pgsql: Fix freezing of a dead HOT-updated tuple

On 2017-11-03 12:36:59 -0700, Peter Geoghegan wrote: > Andres Freund wrote: > > Here's that patch. I've stared at this some, and Robert did too. Robert > > mentioned that the commit message might need some polish and I'm not > > 100% sure about the error message texts yet. >

Re: [HACKERS] Dynamic result sets from procedures

Peter Eisentraut wrote: > > CREATE PROCEDURE test() > > LANGUAGE plpgsql > > AS $$ > > RETURN QUERY EXECUTE 'SELECT 1 AS col1, 2 AS col2'; > > END; > > $$; > > > > Or is that not possible or not desirable? > > RETURN means the execution ends there, so how would you return

Re: [HACKERS] Parallel Plans and Cost of non-filter functions

Moin, On Fri, November 3, 2017 7:13 pm, Tom Lane wrote: > Paul Ramsey writes: >>> Whether I get a parallel aggregate seems entirely determined by the >>> number >>> of rows, not the cost of preparing those rows. > >> This is true, as far as I can tell and unfortunate.

Re: [HACKERS] Add some const decorations to prototypes

On 11/3/17 13:54, Tom Lane wrote: >> Would you prefer leaving the input argument as char *, or change the >> endptr argument to const as well? > > Just leave it as char*. If you change the endptr argument you're going to > force every call site to change their return variable, and some of them >

Re: [HACKERS] pow support for pgbench

Hello Raúl, Sorry about the patch. Attaching it now so it can be considered as submitted. There is a typo in the XML doc: 1024.0/ Please check that the documentation compiles. I'm at odds with having the integer version rely on a double pow(), even if it works. I think that there

Re: [HACKERS] [POC] Faster processing at Gather node

On Wed, Oct 18, 2017 at 3:09 AM, Andres Freund wrote: > 2) The spinlocks both on the the sending and receiving side a quite hot: > >rafia query leader: > + 36.16% postgres postgres[.] shm_mq_receive > + 19.49% postgres postgres[.] s_lock > +

Re: [HACKERS] pgbench - allow to store select results into variables

Hello Pavel, Here is a v13, which is just a rebase after the documentation xml-ization. Here is a v14, after yet another rebase, and some comments added to answer your new comments. I am looking to this patch. Not sure if "cset" is best name - maybe "eset" .. like embeded set? I used

[HACKERS] Re: [COMMITTERS] pgsql: Account for catalog snapshot in PGXACT->xmin updates.

On Mon, Aug 21, 2017 at 07:41:52AM +0100, Simon Riggs wrote: > On 19 August 2017 at 20:54, Noah Misch wrote: > > On Tue, Dec 06, 2016 at 01:59:18PM -0500, Robert Haas wrote: > >> On Tue, Dec 6, 2016 at 1:36 PM, Tom Lane wrote: > >> > Robert Haas

Re: [HACKERS] [PATCH] A hook for session start

On Fri, Nov 3, 2017 at 1:55 PM, Fabrízio de Royes Mello wrote: >> Passing the database name and user name does not look much useful to >> me. You can have access to this data already with CurrentUserId and >> MyDatabaseId. > > This way we don't need to convert oid to

Re: [HACKERS] Exclude pg_internal.init from base backup

On Fri, Nov 3, 2017 at 4:04 PM, Petr Jelinek wrote: > Not specific problem to this patch, but I wonder if it should be made > more clear that those files (there are couple above of what you added) > are skipped no matter which directory they reside in. Agreed, it is

Re: [HACKERS] Small improvement to compactify_tuples

On Fri, Nov 3, 2017 at 4:30 PM, Tom Lane wrote: > Claudio Freire writes: >> On Thu, Nov 2, 2017 at 11:46 PM, Tom Lane wrote: >>> BTW, the originally given test case shows no measurable improvement >>> on my box. > >> I did manage

Re: [HACKERS] Parallel Plans and Cost of non-filter functions

Paul Ramsey writes: >> Whether I get a parallel aggregate seems entirely determined by the number >> of rows, not the cost of preparing those rows. > This is true, as far as I can tell and unfortunate. Feeding tables with > 100ks of rows, I get parallel plans, feeding

Re: [HACKERS] Parallel Plans and Cost of non-filter functions

Just clarifying myself a little, since I made a dumb error partway through. On Thu, Nov 2, 2017 at 10:09 AM, Paul Ramsey wrote: > I'm working on a custom aggregate, that generates a serialized data > format. The preparation of the geometry before being formatted is

Re: [HACKERS] How to implement a SP-GiST index as a extension module?

On Fri, Nov 3, 2017 at 12:37 PM, Connor Wolf < conn...@imaginaryindustries.com> wrote: > EDIT: That's actually exactly how the example I'm working off of works. > DERP. The SQL is > > CREATE TYPE vptree_area AS > ( > center _int4, > distance float8 > ); > > CREATE OR REPLACE FUNCTION

Re: [HACKERS] WIP: long transactions on hot standby feedback replica / proof of concept

On Wed, Nov 1, 2017 at 5:55 AM, Masahiko Sawada wrote: > On Tue, Oct 31, 2017 at 6:17 PM, Alexander Korotkov > wrote: > > On Tue, Oct 31, 2017 at 5:16 AM, Masahiko Sawada > > wrote: > >> > >> On Mon, Oct 30, 2017 at 10:16

Re: [HACKERS] WIP: long transactions on hot standby feedback replica / proof of concept

On Thu, Nov 2, 2017 at 5:56 AM, Robert Haas wrote: > On Tue, Oct 31, 2017 at 2:47 PM, Alexander Korotkov > wrote: > > However, from user prospective of view, current behavior of > > hot_standby_feedback is just broken, because it both increases

Re: [HACKERS] Re: PANIC: invalid index offnum: 186 when processing BRIN indexes in VACUUM

On Fri, Nov 3, 2017 at 1:34 PM, Tom Lane wrote: > Jeff Janes writes: > > With this v3 patch (assuming this is the one you just committed > > as ec42a1dcb30de235b252f6d4) am now getting make check failures. > > There's a followup commit already :-( > >

Re: [HACKERS] Small improvement to compactify_tuples

I wrote: > Have not looked at the 0002 patch yet. I looked at that one, and it seems to be a potential win with no downside, so pushed. (I tweaked it slightly to avoid an unnecessary conflict with the test patch I posted earlier.) regards, tom lane -- Sent via

Re: [HACKERS] Re: PANIC: invalid index offnum: 186 when processing BRIN indexes in VACUUM

Jeff Janes writes: > With this v3 patch (assuming this is the one you just committed > as ec42a1dcb30de235b252f6d4) am now getting make check failures. There's a followup commit already :-( regards, tom lane -- Sent via pgsql-hackers mailing list

Re: [HACKERS] [COMMITTERS] pgsql: Fix freezing of a dead HOT-updated tuple

On November 4, 2017 1:22:04 AM GMT+05:30, Alvaro Herrera wrote: >Peter Geoghegan wrote: >> Andres Freund wrote: > >> > Staring at the vacuumlazy hunk I think I might have found a related >bug: >> > heap_update_tuple() just copies the old xmax to

Re: [HACKERS] Setting pd_lower in GIN metapage

Michael Paquier writes: > On Fri, Nov 3, 2017 at 1:10 AM, Amit Kapila wrote: >> On Fri, Nov 3, 2017 at 2:54 AM, Tom Lane wrote: >>> I've marked the CF entry closed. However, I'm not sure if we're quite >>> done with this

Re: [HACKERS] [COMMITTERS] pgsql: Fix freezing of a dead HOT-updated tuple

Alvaro Herrera wrote: He means that the tuple that heap_update moves to page 1 (which will no longer be processed by vacuum) will contain a multixact that's older than relminmxid -- because it is copied unchanged by heap_update instead of properly checking against age

Re: [HACKERS] Re: PANIC: invalid index offnum: 186 when processing BRIN indexes in VACUUM

Hi Alvaro, With this v3 patch (assuming this is the one you just committed as ec42a1dcb30de235b252f6d4) am now getting make check failures. brin_summarize_range is returning unexpected values. CentOS6, PostgreSQL 11devel on x86_64-pc-linux-gnu, compiled by gcc (GCC) 4.4.7 20120313 (Red Hat

Re: [HACKERS] LDAPS

On Sat, Nov 4, 2017 at 2:05 AM, Thomas Munro wrote: > That > said, I've only tested the attached lightly on FreeBSD + OpenLDAP and > don't know if it'll work elsewhere. Oops, that version's TAP test was a little too dependent on my system's ldap.conf file. Here's

Re: [HACKERS] Re: PANIC: invalid index offnum: 186 when processing BRIN indexes in VACUUM

Alvaro Herrera wrote: > I think your argument is sensible for some uses (where people run manual > VACUUM after loading data) but not others (where people just use manual > VACUUM in place of autovacuuming -- because they don't trust autovac, or > the schedule isn't convenient, or whatever other

Re: [HACKERS] [COMMITTERS] pgsql: Fix freezing of a dead HOT-updated tuple

Peter Geoghegan wrote: > Andres Freund wrote: > > Staring at the vacuumlazy hunk I think I might have found a related bug: > > heap_update_tuple() just copies the old xmax to the new tuple's xmax if > > a multixact and still running. It does so without verifying liveliness >

Re: [HACKERS] ucs_wcwidth vintage

Alvaro Herrera writes: > Thomas Munro wrote: >> src/backend/utils/mb/wchar.c contains a ~16 year old wcwidth >> implementation that originally arrived in commit df4cba68, but the >> upstream code[1] apparently continued evolving and there have been >> more Unicode

Re: [HACKERS] [COMMITTERS] pgsql: Fix freezing of a dead HOT-updated tuple

Andres Freund wrote: Here's that patch. I've stared at this some, and Robert did too. Robert mentioned that the commit message might need some polish and I'm not 100% sure about the error message texts yet. The commit message should probably say that the bug involves the

Re: [HACKERS] Small improvement to compactify_tuples

Claudio Freire writes: > On Thu, Nov 2, 2017 at 11:46 PM, Tom Lane wrote: >> BTW, the originally given test case shows no measurable improvement >> on my box. > I did manage to reproduce the original test and got a consistent improvement. It occurred

Re: [HACKERS] Proposal: Local indexes for partitioned table

Alvaro Herrera writes: > Tom Lane wrote: >> It might work to build the new key in a context that's initially a >> child of CurrentMemoryContext, then reparent it to be a child of >> CacheMemoryContext when done. > That's another way (than the PG_TRY block), but I think

Re: [HACKERS] ucs_wcwidth vintage

Thomas Munro wrote: > Hi hackers, > > src/backend/utils/mb/wchar.c contains a ~16 year old wcwidth > implementation that originally arrived in commit df4cba68, but the > upstream code[1] apparently continued evolving and there have been > more Unicode revisions since. It probably doesn't matter

Re: [HACKERS] ucs_wcwidth vintage

Thomas Munro wrote: > Hi hackers, > > src/backend/utils/mb/wchar.c contains a ~16 year old wcwidth > implementation that originally arrived in commit df4cba68, but the > upstream code[1] apparently continued evolving and there have been > more Unicode revisions since. I think we should update it

Re: [HACKERS] Proposal: Local indexes for partitioned table

> Robert Haas writes: > > We could do that, but the motivation for the current system was to > > avoid leaking memory in a long-lived context. Yeah, my approach here is to use a CATCH block that deletes the memory context just created, thus avoiding a long-lived leak.

Re: [HACKERS] Where is it documented what role executes constraint triggers?

Chapman Flack writes: > From a little experimenting in 9.5, it seems that a referential > integrity trigger is executed with the identity of the referencED > table's owner, but I have not been able to find this covered in > the docs. Is this a documentation oversight, or is

Re: [HACKERS] Where is it documented what role executes constraint triggers?

On Fri, Nov 03, 2017 at 02:09:00PM -0400, Chapman Flack wrote: > From a little experimenting in 9.5, it seems that a referential > integrity trigger is executed with the identity of the referencED > table's owner, but I have not been able to find this covered in > the docs. Is this a documentation

[HACKERS] Where is it documented what role executes constraint triggers?

>From a little experimenting in 9.5, it seems that a referential integrity trigger is executed with the identity of the referencED table's owner, but I have not been able to find this covered in the docs. Is this a documentation oversight, or is it explained somewhere I didn't look (or may have

Re: [HACKERS] [PATCH] Add ALWAYS DEFERRED option for constraints

On Fri, Nov 03, 2017 at 01:41:45PM -0400, Peter Eisentraut wrote: > On 11/2/17 16:54, Nico Williams wrote: > > Replacing condeferred and condeferrable with a char columns also > > occurred to me, and though I assume backwards-incompatible changes to > > pg_catalog tables are fair game, I assumed

Re: [HACKERS] Dynamic result sets from procedures

On 11/2/17 16:40, Daniel Verite wrote: > But instead of having procedures not return anything, > couldn't they return whatever resultset(s) they want to > ("no resultset" being just a particular case of "anything"), > so that we could leave out cursors and simply write: We could in general design

Re: [HACKERS] Add some const decorations to prototypes

Peter Eisentraut writes: > On 10/31/17 10:56, Tom Lane wrote: >>> Some functions have a strtol()-like behavior >>> where they take in a const char * and return a pointer into that as >>> another argument. In those cases, I added a cast or two. >> ... but I'm

Re: [HACKERS] Dynamic result sets from procedures

On 11/1/17 22:40, Robert Haas wrote: > That seems like it is at least arguably a wire protocol break. Today, > if you send a string containing only one command, you will only get > one answer. The wire protocol already supports this. And the wire protocol doesn't really know about statements,

Re: [HACKERS] Dynamic result sets from procedures

On 11/1/17 06:23, Pavel Stehule wrote: > We need to think about how the \timing option should work in such > scenarios.  Right now it does > Has the total time sense  in this case? > > should not be total time related to any fetched result? The \timing option in psql measures from the

Re: [HACKERS] Linking libpq statically to libssl

On 11/2/17 19:52, Stephen Frost wrote: > Uh, libpq doesn't actually have symbol versioning, at least not on the > installed Ubuntu packages for PG10, as shown by objdump -T : You're right, I was dreaming. But in any case, he would need symbol versioning of libssl. -- Peter Eisentraut

Re: [HACKERS] [PATCH] Add ALWAYS DEFERRED option for constraints

On 11/2/17 16:54, Nico Williams wrote: > Replacing condeferred and condeferrable with a char columns also > occurred to me, and though I assume backwards-incompatible changes to > pg_catalog tables are fair game, I assumed everyone would prefer > avoiding such changes where possible. I don't

Re: [HACKERS] Add some const decorations to prototypes

On 10/31/17 10:56, Tom Lane wrote: >> Some functions have a strtol()-like behavior >> where they take in a const char * and return a pointer into that as >> another argument. In those cases, I added a cast or two. > ... but I'm not sure that it's an improvement in cases where you have to > cast

Re: [HACKERS] MERGE SQL Statement for PG11

Simon Riggs wrote: The *only* behavioural difference I have proposed would be the *lack* of an ERROR in (some) concurrent cases. I think that's a big difference. Error vs. non-error is a big deal by itself; Are you saying avoiding an ERROR is a bad or good thing?

<    1   2   3   4   5   6   7   8   9   10   >