Renumber cursor option flags
Move the planner-control flags up so that there is more room for parse
options. Some pending patches need some room there, so do this
renumbering separately so that there is less potential for conflicts.
Branch
--
master
Details
---
https://git.postgresql.or
Michael Paquier writes:
> Refactor all TAP test suites doing connection checks
Some of the buildfarm seems not to like this. gaur failed here:
==~_~===-=-===~_~==
pgsql.build/src/test/authentication/tmp_check/log/regress_log_001_password
==~_~===-=-===~_~==
...
ok 8 - authentication success f
Align some terms in arch-dev.sgml to glossary
This mostly adds links to the glossary to the existing text, instead of
using . Heikki left this out of 29ad6595ef7f out of
stylistic concerns; these have since been addressed.
Author: Jürgen Purtz
Discussion: https://postgr.es/m/67d7240f-8596-83fc-
Shut down transaction tracking at startup process exit.
Maxim Orlov reported that the shutdown of standby server could result in
the following assertion failure. The cause of this issue was that,
when the shutdown caused the startup process to exit, recovery-time
transaction tracking was not shut
Shut down transaction tracking at startup process exit.
Maxim Orlov reported that the shutdown of standby server could result in
the following assertion failure. The cause of this issue was that,
when the shutdown caused the startup process to exit, recovery-time
transaction tracking was not shut
Shut down transaction tracking at startup process exit.
Maxim Orlov reported that the shutdown of standby server could result in
the following assertion failure. The cause of this issue was that,
when the shutdown caused the startup process to exit, recovery-time
transaction tracking was not shut
Shut down transaction tracking at startup process exit.
Maxim Orlov reported that the shutdown of standby server could result in
the following assertion failure. The cause of this issue was that,
when the shutdown caused the startup process to exit, recovery-time
transaction tracking was not shut
Shut down transaction tracking at startup process exit.
Maxim Orlov reported that the shutdown of standby server could result in
the following assertion failure. The cause of this issue was that,
when the shutdown caused the startup process to exit, recovery-time
transaction tracking was not shut
Shut down transaction tracking at startup process exit.
Maxim Orlov reported that the shutdown of standby server could result in
the following assertion failure. The cause of this issue was that,
when the shutdown caused the startup process to exit, recovery-time
transaction tracking was not shut
Add pg_read_all_data and pg_write_all_data roles
A commonly requested use-case is to have a role who can run an
unfettered pg_dump without having to explicitly GRANT that user access
to all tables, schemas, et al, without that role being a superuser.
This address that by adding a "pg_read_all_data
Simplify state managed by VACUUM.
Reorganize the state struct used by VACUUM -- group related items
together to make it easier to understand. Also stop relying on stack
variables inside lazy_scan_heap() -- move those into the state struct
instead. Doing things this way simplifies large groups of
On Mon, Apr 5, 2021 at 1:27 PM Peter Geoghegan wrote:
> Simplify state managed by VACUUM.
I'm investigating the buildfarm failures on elver, curculio, and morepork.
--
Peter Geoghegan
Propagate parallel VACUUM's buffer access strategy.
Parallel VACUUM relied on global variable state from the leader process
being propagated to workers on fork(). Commit b4af70cb removed most
uses of global variables inside vacuumlazy.c, but did not account for
the buffer access strategy state.
Support INCLUDE'd columns in SP-GiST.
Not much to say here: does what it says on the tin.
We steal a previously-always-zero bit from the nextOffset
field of leaf index tuples in order to track whether there
is a nulls bitmap. Otherwise it works about like included
columns in other index types.
P
On Mon, Apr 05, 2021 at 10:57:22AM -0400, Tom Lane wrote:
> Some of the buildfarm seems not to like this. gaur failed here:
Thanks. I did not notice these.
> I didn't look at the code, but I bet this is similar to the race
> conditions we've seen before where a server process may exit before
>
Michael Paquier writes:
> On Mon, Apr 05, 2021 at 10:57:22AM -0400, Tom Lane wrote:
>> I didn't look at the code, but I bet this is similar to the race
>> conditions we've seen before where a server process may exit before
>> the perl test script finishes stuffing a command down the pipe.
> Hmm.
On Mon, Apr 05, 2021 at 06:57:47PM -0400, Tom Lane wrote:
> No, I bet it is with "now defines a SQL query". That means there's
> something to be stuffed down the pipe.
>
> I find I can reproduce this pretty reliably on slower machines
> here, if you want me to dig into it more carefully.
Now tha
I wrote:
> Michael Paquier writes:
>> So, the problem is with on_error_stop.
> No, I bet it is with "now defines a SQL query". That means there's
> something to be stuffed down the pipe.
I can confirm that removing the on_error_stop parameter as you
propose does *not* fix it.
What does seem to fix it reliably is the attached.
regards, tom lane
diff --git a/src/test/perl/PostgresNode.pm b/src/test/perl/PostgresNode.pm
index bbde34c..07a9167 100644
--- a/src/test/perl/PostgresNode.pm
+++ b/src/test/perl/PostgresNode.pm
@@ -1938,7 +1938,7 @@ sub c
On Mon, Apr 05, 2021 at 07:23:43PM -0400, Tom Lane wrote:
> What does seem to fix it reliably is the attached.
Thanks for confirming. I am still surprised that this has not popped
up as an issue until now. Could that be because those animals have
never run the LDAP, SSL or kerberos test suites o
Michael Paquier writes:
> On Mon, Apr 05, 2021 at 07:23:43PM -0400, Tom Lane wrote:
>> What does seem to fix it reliably is the attached.
> Thanks for confirming. I am still surprised that this has not popped
> up as an issue until now. Could that be because those animals have
> never run the L
Allocate access strategy in parallel VACUUM workers.
Commit 49f49def took entirely the wrong approach to fixing this issue.
Just allocate a local buffer access strategy in each individual worker
instead of trying to propagate state. This state was never propagated
by parallel VACUUM in the first
On Mon, Apr 05, 2021 at 08:15:31PM -0400, Tom Lane wrote:
> Dunno. Certainly most animals that enable TAP at all should be
> running the "authentication" test, but those others all need to
> be manually enabled, I believe. My own animals run the SSL test
> but not LDAP or kerberos.
The SSL tests
On Tue, Apr 6, 2021 at 12:15 PM Tom Lane wrote:
> Michael Paquier writes:
> > On Mon, Apr 05, 2021 at 07:23:43PM -0400, Tom Lane wrote:
> >> What does seem to fix it reliably is the attached.
>
> > Thanks for confirming. I am still surprised that this has not popped
> > up as an issue until now.
Michael Paquier writes:
> On Mon, Apr 05, 2021 at 08:15:31PM -0400, Tom Lane wrote:
>> Dunno. Certainly most animals that enable TAP at all should be
>> running the "authentication" test, but those others all need to
>> be manually enabled, I believe. My own animals run the SSL test
>> but not L
Change PostgresNode::connect_fails() to never send down queries
This type of failure is similar to what has been fixed in c757a3da,
where an authentication failure combined with psql pushing a command
down its communication pipe causes a test failure. This routine is
designed to fail, so sending
On Tue, Apr 06, 2021 at 12:30:48PM +1200, Thomas Munro wrote:
> Commit f44b9b62 springs to mind.
Right. The kerberos suite does not do that though, so wouldn't it be
better to use the same trick with -c? For example, the attached for
~13 does so.
--
Michael
diff --git a/src/test/kerberos/t/001_a
I wrote:
> Well, since it's a race condition it's obviously going to be sensitive to
> machine speed. I think that Perl version, or more likely version of
> IPC::Run or a related module, may enter into it too. prairiedog has
> gotten through the tests several times since yesterday without failing
Refactor function parse_output_parameters.
Instead of using multiple parameters in parse_ouput_parameters function
signature, use the struct PGOutputData that encapsulates all pgoutput
options. It will be useful for future work where we need to add other
options in pgoutput.
Author: Euler Taveira
Allow pgoutput to send logical decoding messages.
The output plugin accepts a new parameter (messages) that controls if
logical decoding messages are written into the replication stream. It is
useful for those clients that use pgoutput as an output plugin and needs
to process messages that were wr
Fix some issues with SSL and Kerberos tests
The recent refactoring done in c50624c accidentally broke a portion of
the kerberos tests checking after a query, so add its functionality
back. Some inactive SSL tests had their arguments in an incorrect
order, which would cause them to fail if they we
On Wed, 31 Mar 2021 at 12:17, Alvaro Herrera wrote:
>
> Improve PQtrace() output format
I see the Visual Studio buildfarm members have been getting a compiler
warning since this commit [1]
(ClCompile target) ->
src/interfaces/libpq/fe-trace.c(87): warning C4133: 'function':
incompatible types
Add function to log the memory contexts of specified backend process.
Commit 3e98c0bafb added pg_backend_memory_contexts view to display
the memory contexts of the backend process. However its target process
is limited to the backend that is accessing to the view. So this is
not so convenient when
David Rowley writes:
> I see the Visual Studio buildfarm members have been getting a compiler
> warning since this commit [1]
> (ClCompile target) ->
> src/interfaces/libpq/fe-trace.c(87): warning C4133: 'function':
> incompatible types - from 'long *' to 'const time_t *const '
> [C:\\pgbuildfar
Fix typo in pgstat.c.
Introduced by 9868167500.
Author: Vignesh C
Discussion:
https://postgr.es/m/caldanm1dqgalbajrtgznkk1sr1mh-augmp7lfgvxwwtuhah...@mail.gmail.com
Branch
--
master
Details
---
https://git.postgresql.org/pg/commitdiff/f5d94e405e17a49487672316610630be2f9d0bb7
Modified
I wrote:
> As best I can tell from these warnings, VS has the tv_sec field
> of struct timeval defined as something other than time_t. Because
> VS would never stoop to actually complying with the plain text of
> the POSIX standard.
Ah, seems like it's really Winsock's fault:
https://docs.micros
Change return type of EXTRACT to numeric
The previous implementation of EXTRACT mapped internally to
date_part(), which returned type double precision (since it was
implemented long before the numeric type existed). This can lead to
imprecise output in some cases, so returning numeric would be
pr
Fix compiler warning in fe-trace.c for MSVC
It seems that in MSVC timeval's tv_sec field is of type long.
localtime() takes a time_t pointer. Since long is 32-bit even on 64-bit
builds in MSVC, passing a long pointer instead of the correct time_t
pointer generated a compiler warning. Fix that.
On Tue, 6 Apr 2021 at 17:21, Tom Lane wrote:
>
> I wrote:
> > As best I can tell from these warnings, VS has the tv_sec field
> > of struct timeval defined as something other than time_t. Because
> > VS would never stoop to actually complying with the plain text of
> > the POSIX standard.
>
> Ah,
39 matches
Mail list logo
