On Sat, Aug 12, 2023 at 5:01 AM Peter Smith wrote:
>
> On Fri, Aug 11, 2023 at 9:13 PM Amit Kapila wrote:
> >
> > On Fri, Aug 11, 2023 at 3:41 PM Peter Smith wrote:
> > >
> > > On Fri, Aug 11, 2023 at 7:33 PM Amit Kapila
> > > wrote:
> > > >
> > > > On Thu, Aug 10, 2023 at 7:50 AM Peter Smith
Thanks. I realised that it's easy enough to test that theory about
cleanup locks by hacking ConditionalLockBufferForCleanup() to return
false randomly. Then the test occasionally fails as described. Seems
like we'll need to fix that test, but it's not evidence of a server
bug, and my signal
The attached patch implements a new SEARCH clause for CREATE FUNCTION.
The SEARCH clause controls the search_path used when executing
functions that were created without a SET clause.
Background:
Controlling search_path is critical for the correctness and security of
functions. Right now, the
On Fri, Aug 11, 2023 at 05:55:26PM +0200, Dmitry Dolgov wrote:
>
> Another confusing example was this one at the end of set_session_variable:
>
> + /*
> + * XXX While unlikely, an error here is possible. It wouldn't leak
> memory
> + * as the allocated chunk has already been
Hi,
On 2023-08-11 20:11:34 -0400, Tom Lane wrote:
> Andres Freund writes:
> > On 2023-08-11 18:30:02 -0400, Tom Lane wrote:
> >> +1 for including this in CI tests
>
> > I didn't even mean CI - I meant 'make check-world' / 'meson test'. Which of
> > course would include CI automatically.
>
> Hmm.
Hi
Commit 31966b15 invented a way for functions dealing with relation
extension to accept a Relation in online code and an SMgrRelation in
recovery code (instead of using the earlier FakeRelcacheEntry
concept). It seems highly likely that future new bufmgr.c interfaces
will face the same
On Fri, Aug 11, 2023 at 08:11:34PM -0400, Tom Lane wrote:
> Hmm. I'm allergic to anything that significantly increases the cost
> of check-world, and this seems like it'd do that.
>
> Maybe we could automate it, but not as part of check-world per se?
It does not have to be part of check-world
Andres Freund writes:
> On 2023-08-11 18:30:02 -0400, Tom Lane wrote:
>> +1 for including this in CI tests
> I didn't even mean CI - I meant 'make check-world' / 'meson test'. Which of
> course would include CI automatically.
Hmm. I'm allergic to anything that significantly increases the cost
The previous patch was accidentally not resetting the boolean limit
flags to false for retries.
Fixed in v2.
--
Kind Regards,
Peter Smith.
Fujitsu Australia
v2-0001-logicalrep_worker_launch-limit-checks.patch
Description: Binary data
On Fri, Aug 11, 2023 at 9:13 PM Amit Kapila wrote:
>
> On Fri, Aug 11, 2023 at 3:41 PM Peter Smith wrote:
> >
> > On Fri, Aug 11, 2023 at 7:33 PM Amit Kapila wrote:
> > >
> > > On Thu, Aug 10, 2023 at 7:50 AM Peter Smith wrote:
> > > >
> > > > >
> > > > > * If you do the above then there won't
Hi,
On 2023-08-11 18:30:02 -0400, Tom Lane wrote:
> Peter Geoghegan writes:
> > On Fri, Aug 11, 2023 at 2:25 PM Andres Freund wrote:
> >> We could a test that fails when there's some mis-indented code. That seems
> >> like it might catch things earlier?
>
> +1 for including this in CI tests
I
On Fri, 11 Aug 2023 at 23:00, Peter Geoghegan wrote:
> I'm starting to have doubts about this policy. There have now been
> quite a few follow-up "fixes" to indentation issues that koel
> complained about.
I think one thing that would help a lot in reducing the is for
committers to set up the
Peter Geoghegan writes:
> I'm starting to have doubts about this policy. There have now been
> quite a few follow-up "fixes" to indentation issues that koel
> complained about. None of these fixups have been included in
> .git-blame-ignore-revs. If things continue like this then "git blame"
> is
Peter Geoghegan writes:
> My workflow up until now has avoiding making updates to typedefs.list
> in patches. I only update typedefs locally, for long enough to indent
> my code. The final patch doesn't retain any typedefs.list changes.
Yeah, I've done the same and will have to stop.
> I guess
On Fri, Aug 11, 2023 at 3:30 PM Tom Lane wrote:
> No. I presume koel is using src/tools/pgindent/typedefs.list,
> which has always been the "canonical" list but up to now we've
> been lazy about maintaining it. Part of the new regime is that
> typedefs.list should now be updated on-the-fly by
Peter Geoghegan writes:
> On Fri, Aug 11, 2023 at 2:25 PM Andres Freund wrote:
>> We could a test that fails when there's some mis-indented code. That seems
>> like it might catch things earlier?
+1 for including this in CI tests
> It definitely would. That would go a long way towards
On Fri, Aug 11, 2023 at 2:25 PM Andres Freund wrote:
> > I don't think that it makes sense to invent yet another rule for
> > .git-blame-ignore-revs, though. Will we need another buildfarm member
> > to enforce that rule, too?
>
> We could a test that fails when there's some mis-indented code.
On Fri, Aug 11, 2023 at 2:04 PM Andres Freund wrote:
> Where supported, a crash (distinguishing from assertion failures) will now
> report something like:
>
> process with pid: 2900527 received signal: SIGSEGV, si_code: 1, si_addr:
> 0xdeadbeef
> [0x5628ec45212f]
Hi,
On 2023-08-11 13:59:40 -0700, Peter Geoghegan wrote:
> On Sat, Jun 17, 2023 at 7:08 AM Andrew Dunstan wrote:
> > I have set up a new buildfarm animal called koel which will run the module.
>
> I'm starting to have doubts about this policy. There have now been
> quite a few follow-up "fixes"
Hi,
On 2023-08-11 13:19:34 -0700, Peter Geoghegan wrote:
> On Fri, Aug 11, 2023 at 12:26 PM Andres Freund wrote:
> > > For example, dealing with core dumps left behind by the regression
> > > tests can be annoying.
> >
> > Hm. I don't have a significant problem with that. But I can see it being
In my case, my product has a very controlled environment.
We build all our infrastructure from source and we avoid dynamic linking by
design, except where technically not viable (e.g.: pgsql extensions).
LLVM is one of the libraries we're specifically required to statically link.
Unfortunately I
Andres, Tom, I found your names in the git history for JIT and LLVM.
Any chance one of you could take a look at the patch?
-mj
On Thu, Aug 10, 2023 at 2:45 PM Marcelo Juchem wrote:
> By default, PostgreSQL doesn't explicitly choose whether to link
> statically or dynamically against LLVM when
On Sat, Jun 17, 2023 at 7:08 AM Andrew Dunstan wrote:
> I have set up a new buildfarm animal called koel which will run the module.
I'm starting to have doubts about this policy. There have now been
quite a few follow-up "fixes" to indentation issues that koel
complained about. None of these
Hi,
On 2023-08-12 07:51:09 +1200, Thomas Munro wrote:
> Oh, I see what's happening. Maybe commit b91dd9de wasn't the best
> idea, but bc971f4025c broke an assumption, since it doesn't use
> ConditionVariableSleep(). That is confusing the signal forwarding
> logic which expects to find our entry
On Fri, Aug 11, 2023 at 12:26 PM Andres Freund wrote:
> > For example, dealing with core dumps left behind by the regression
> > tests can be annoying.
>
> Hm. I don't have a significant problem with that. But I can see it being
> problematic. Unfortunately, short of preventing core dumps from
On Sat, Aug 12, 2023 at 5:51 AM Andres Freund wrote:
> On 2023-08-11 15:31:43 +0200, Tomas Vondra wrote:
> > It seems to me the issue is in WalSndWait, which was reworked to use
> > ConditionVariableCancelSleep() in bc971f4025c. The walsenders end up
> > waking each other in a busy loop, until
Applied v3 patch to master and verified it with below commands,
#Alter view
postgres=# alter view v
ALTER COLUMN OWNER TO RENAME RESET ( SET
postgres=# alter view v set
( SCHEMA
postgres=# alter view v set (
CHECK_OPTION SECURITY_BARRIER SECURITY_INVOKER
Hi,
On 2023-08-11 11:56:27 -0700, Peter Geoghegan wrote:
> On Fri, Aug 11, 2023 at 11:23 AM Andres Freund wrote:
> > > Couldn't you say the same thing about defensive "can't happen" ERRORs?
> > > They are essentially a form of assertion that isn't limited to
> > > assert-enabled builds.
> >
> >
On Fri, Aug 11, 2023 at 11:23 AM Andres Freund wrote:
> > Couldn't you say the same thing about defensive "can't happen" ERRORs?
> > They are essentially a form of assertion that isn't limited to
> > assert-enabled builds.
>
> Yes. A lot of them I hate them with the passion of a thousand suns ;).
Hi,
On 2023-08-11 11:14:34 -0700, Peter Geoghegan wrote:
> On Fri, Aug 11, 2023 at 10:57 AM Andres Freund wrote:
> > I am quite strongly against this. This will lead to assertions being hit in
> > tests without that being noticed, e.g. because they happen in a background
> > process that just
On Fri, Aug 11, 2023 at 10:57 AM Andres Freund wrote:
> I am quite strongly against this. This will lead to assertions being hit in
> tests without that being noticed, e.g. because they happen in a background
> process that just restarts.
Couldn't you say the same thing about defensive "can't
Hi,
On 2023-08-11 12:59:27 -0500, Marcelo Juchem wrote:
> In my case, my product has a very controlled environment.
> We build all our infrastructure from source and we avoid dynamic linking by
> design, except where technically not viable (e.g.: pgsql extensions).
>
> LLVM is one of the
Hi,
On 2023-05-21 15:01:41 +1200, Thomas Munro wrote:
> *For anyone working with this type of IR generation code and
> questioning their sanity, I can pass on some excellent advice I got
> from Andres: build LLVM yourself with assertions enabled, as they
> catch some classes of silly mistake that
On Fri, Aug 11, 2023 at 10:46:31AM +0530, Amit Kapila wrote:
> On Thu, Aug 10, 2023 at 7:07 PM Masahiko Sawada wrote:
> > What I imagined is that we do this check before
> > check_and_dump_old_cluster() while the server is 'off'. Reading the
> > slot state file would be simple and I guess we
On Fri, Aug 11, 2023 at 11:18:09AM +0530, Amit Kapila wrote:
> On Fri, Aug 11, 2023 at 10:43 AM Julien Rouhaud wrote:
> > I disagree. As I mentioned before any module registered in
> > shared_preload_libraries can spawn background workers which can perform any
> > activity. There were previous
Hi,
On 2023-08-10 16:56:54 +0200, Ronan Dunklau wrote:
> I tried my hand at backporting it to previous versions, and not knowing
> anything about it made me indeed question my sanity. It's quite easy for PG
> 15, 14, 13. PG 12 is nothing insurmontable either, but PG 11 is a bit hairier
> most
Hi,
On 2023-08-11 17:59:37 +0530, Ashutosh Bapat wrote:
> Most of the Asserts are recoverable by rolling back the transaction
> without crashing the backend. So an elog(ERROR, ) is enough. But just
> by themselves elogs are compiled into non-debug binary and the
> condition check can waste CPU
Hi,
On 2023-08-11 15:31:43 +0200, Tomas Vondra wrote:
> That's an awful lot of CPU for a cluster doing essentially "nothing"
> (there's no WAL to decode/replicate, etc.). It usually disappears after
> a couple seconds, but sometimes it's a rather persistent state.
Ugh, that's not great.
> The
Hi,
On 2023-08-10 14:45:47 -0500, Marcelo Juchem wrote:
> By default, PostgreSQL doesn't explicitly choose whether to link
> statically or dynamically against LLVM when LLVM JIT is enabled (e.g.:
> `./configure --with-llvm`).
>
> `llvm-config` will choose to dynamically link by default.
>
> In
Re: Thomas Munro
> On Thu, Aug 10, 2023 at 9:15 PM Christoph Berg wrote:
> > No XXX lines this time either, but I've seen then im logfiles that
> > went through successfully.
>
> Do you still have the data directories around from that run, so we can
> see if the expected Heap2/PRUNE was actually
> On Thu, Aug 03, 2023 at 08:15:13AM +0200, Pavel Stehule wrote:
> Hi
>
> fresh rebase
Thanks for continuing efforts. The new patch structure looks better to
me (although the boundary between patches 0001 and 0002 is somewhat
fuzzy, e.g. the function NameListToString is used already in the first
On Fri, Aug 11, 2023 at 14:31 Amit Langote wrote:
> On Wed, Aug 9, 2023 at 1:05 AM Robert Haas wrote:
> > On Tue, Aug 8, 2023 at 10:32 AM Amit Langote
> wrote:
> > > But should ExecInitNode() subroutines return the partially initialized
> > > PlanState node or NULL on detecting invalidation?
Hi Peter,
Peter Smith , 11 Ağu 2023 Cum, 01:26 tarihinde şunu
yazdı:
> No, I meant what I wrote there. When I ran the tests the HEAD included
> the v25-0001 refactoring patch, but v26 did not yet exist.
>
> For now, we are only performance testing the first
> "Reuse-Tablesyc-Workers" patch, but
Hi,
while working on some logical replication stuff, I noticed that on PG16
I often end up with a completely idle publisher (no user activity), that
however looks like this in top:
... %CPU COMMAND
... 17.9 postgres: walsender user test ::1(43064) START_REPLICATION
... 16.6
On Fri, Aug 11, 2023 at 10:41 AM Andrey Lepikhov
wrote:
> >>
> >
> > This won't just affect planner but every subsystem of PostgreSQL, not
> > just planner, unless we device a new context type for planner.
> >
> > My point is what's relevant here is how much net memory planner asked
> > for.
On Fri, Aug 11, 2023 at 11:48:18AM +0200, Juan José Santamaría Flecha wrote:
> On Fri, Aug 11, 2023 at 7:29 AM Noah Misch wrote:
>
> >
> > The LCMapStringEx() solution is elegant. I do see
> >
> > https://learn.microsoft.com/en-us/windows/win32/intl/handling-sorting-in-your-applications
> >
I spent some time on 4th point below but also looked at other points.
Here's what I have found so far
On Thu, Jul 27, 2023 at 7:35 PM Ashutosh Bapat
wrote:
>
> 1. The patch uses RestrictInfo::required_relids as the key for
> searching child RelOptInfos. I am not sure which of the two viz.
>
Hi All,
PostgreSQL code uses Assert() as a way to
1. document the assumption or conditions which must be true at a given
place in code
2. make sure that some bug elsewhere does not break these assumptions or rules
3. conditions which can not be (easily) induced by user actions
E.g. following
On Fri, 11 Aug 2023 at 16:26, vignesh C wrote:
>
> On Wed, 9 Aug 2023 at 09:51, vignesh C wrote:
> >
> > Hi Melih,
> >
> > Here is a patch to help in getting the execution at various phases
> > like: a) replication slot creation time, b) Wal reading c) Number of
> > WAL records read d)
On Thu, Jul 27, 2023 at 8:28 AM David Geier wrote:
> Hi,
>
> On 6/7/23 23:37, Andres Freund wrote:
> > I think we're starting to hit quite a few limits related to the process
> model,
> > particularly on bigger machines. The overhead of cross-process context
> > switches is inherently higher
> On 10 Aug 2023, at 19:47, Jonah H. Harris wrote:
>
> Pinging to see if anyone has continued to work on this behind-the-scenes or
> whether this is the latest patch set there is.
It's still on my TODO list, but I haven't done much review cycles yet. And the
patch series already needs
On Fri, Aug 11, 2023 at 3:41 PM Peter Smith wrote:
>
> On Fri, Aug 11, 2023 at 7:33 PM Amit Kapila wrote:
> >
> > On Thu, Aug 10, 2023 at 7:50 AM Peter Smith wrote:
> > >
> > > >
> > > > * If you do the above then there won't be a need to change the
> > > > variable name
On Wed, 9 Aug 2023 at 09:51, vignesh C wrote:
>
> Hi Melih,
>
> Here is a patch to help in getting the execution at various phases
> like: a) replication slot creation time, b) Wal reading c) Number of
> WAL records read d) subscription relation state change etc
> Couple of observation while we
On Fri, Aug 11, 2023 at 7:33 PM Amit Kapila wrote:
>
> On Thu, Aug 10, 2023 at 7:50 AM Peter Smith wrote:
> >
> > >
> > > * If you do the above then there won't be a need to change the
> > > variable name is_parallel_apply_worker in logicalrep_worker_launch.
> > >
> >
> > Done.
> >
>
> I don't
On Fri, Aug 11, 2023 at 7:29 AM Noah Misch wrote:
>
> The LCMapStringEx() solution is elegant. I do see
>
> https://learn.microsoft.com/en-us/windows/win32/intl/handling-sorting-in-your-applications
> says, "If an application calls the function to create a sort key for a
> string
> containing
On Thu, Aug 10, 2023 at 7:50 AM Peter Smith wrote:
>
> >
> > * If you do the above then there won't be a need to change the
> > variable name is_parallel_apply_worker in logicalrep_worker_launch.
> >
>
> Done.
>
I don't think the addition of two new macros isTablesyncWorker() and
While reviewing other threads I have been looking more closely at the
the logicalrep_worker_launch() function. IMO the logic of that
function seems not quite right.
Here are a few things I felt are strange:
1. The function knows exactly what type of worker it is launching, but
still, it is
čt 10. 8. 2023 v 16:31 odesílatel Jelte Fennema napsal:
> On Thu, 10 Aug 2023 at 14:44, Pavel Stehule
> wrote:
> > čt 10. 8. 2023 v 14:05 odesílatel Jelte Fennema
> napsal:
> >> That it is not rolled-back
> >> in a case like this?
> >>
> >> BEGIN;
> >> \set PROMPT '%N'
> >> ROLLBACK;
> >
> >
>
On Tue, Aug 1, 2023 at 8:46 PM Tomas Vondra
wrote:
>
> Anyway, I think this is "just" a matter of efficiency, not correctness.
> IMHO there are bigger questions regarding the "going back" behavior
> after apply restart.
sequence_decode() has the following code
/* Skip the change if already
59 matches
Mail list logo
