On Fri, Jul 2, 2021 at 12:03 PM Dilip Kumar wrote:
>
> Yeah, this looks like a bug. I will look at the patch.
>
While looking into this, I think the main cause of the problem is that
schema rename does not invalidate the relation cache right? I also
tried other cases e.g. if there is an open
On Sat, Jul 3, 2021 at 2:19 AM Tom Lane wrote:
>
> I wrote:
> > Didn't look at 0002 yet.
>
> ... and now that I have, I don't like it much. It adds a lot of
> complexity, plus some lookup cycles that might be wasted. I experimented
> with looking into the range table as I suggested previously,
On Sat, Jul 3, 2021 at 1:37 AM Tom Lane wrote:
>
> Bharath Rupireddy writes:
> >> << Attaching v5-0001 here again for completion >>
> >> I'm attaching v5-0001 patch that avoids printing the column type names
> >> in the context message thus no cache lookups have to be done in the
> >> error
On Fri, Jul 02, 2021 at 11:20:17AM -0400, Tom Lane wrote:
> Alvaro Herrera writes:
> > Now it's hoverfly:
> > ! nm -A -u libpq.so.5 2>/dev/null | grep -v __cxa_atexit | grep exit
> > libpq.so.5: atexit U -
> > libpq.so.5: pthread_exit U -
>
> Ugh. What
"alvhe...@alvh.no-ip.org" writes:
> gcc here is 8.3.0.
Hmmm ... mine is 8.4.1.
I'm about to go out to dinner, but will check into this with some
newer gcc versions later.
regards, tom lane
On Fri, 2 Jul 2021 at 12:56, David Rowley wrote:
>
> On Fri, 2 Jul 2021 at 22:55, Dean Rasheed wrote:
> > Here's an update with the
> > last set of changes discussed.
>
> Looks good to me.
Thanks for the review and testing!
> Just the question of if we have any problems changing the serialized
On 2021-Jul-02, Jacob Champion wrote:
> Only src/common:
>
> controldata_utils_shlib.o:
> U close
> U __errno_location
> U exit
Actually, I do see these in the .o files as well, but they don't make it
to the .a file.
gcc here is 8.3.0.
--
On Fri, 2021-07-02 at 18:45 -0400, Tom Lane wrote:
> Jacob Champion writes:
> > On Fri, 2021-07-02 at 18:20 -0400, Tom Lane wrote:
> > > What configure options are you using?
> > Just `./configure --enable-coverage`, nothing else. I distclean'd right
> > before for good measure.
>
> Hmph.
Jacob Champion writes:
> On Fri, 2021-07-02 at 18:20 -0400, Tom Lane wrote:
>> What configure options are you using?
> Just `./configure --enable-coverage`, nothing else. I distclean'd right
> before for good measure.
Hmph. There's *something* different about your setup from what
either Alvaro
On Fri, 2021-07-02 at 18:20 -0400, Tom Lane wrote:
> What configure options are you using?
Just `./configure --enable-coverage`, nothing else. I distclean'd right
before for good measure.
> Does "nm -u" report "exit" being referenced from any *.o in libpq,
> or from any *_shlib.o in src/port/ or
Jacob Champion writes:
> With latest HEAD, building with --enable-coverage still fails on my
> Ubuntu 20.04:
> ! nm -A -u libpq.so.5.15 2>/dev/null | grep -v __cxa_atexit | grep exit
> libpq.so.5.15: U exit@@GLIBC_2.2.5
Hm, weird. I don't see that here on RHEL8, and
On Wed, 2021-06-30 at 10:42 -0400, Tom Lane wrote:
> Alvaro Herrera writes:
> > On 2021-Jun-30, Tom Lane wrote:
> > > You mentioned __gcov_exit, but I'm not sure if we need an
> > > exception for that. I see it referenced by the individual .o
> > > files, but the completed .so has no such
Hello Dean,
It may be true that the bias is of the same magnitude as FP multiply,
but it is not of the same nature. With FP multiply, the
more-likely-to-be-chosen values are more-or-less evenly distributed
across the range, whereas modulo concentrates them all at one end,
making it more
"-- # QUERY\n%s\n\n"
Attached an attempt along those lines. I found another duplicate of the
ascii-art printing in another function.
Completion queries seems to be out of the echo/echo hidden feature.
Incredible, there is a (small) impact on regression tests for the \gexec
case. All
Alvaro Herrera writes:
> I think the most interesting case for decoration is the "step by step"
> mode, where you want the "title" that precedes each query be easily
> visible.
I'm okay with leaving the step-by-step prompt as-is, personally.
It's the inconsistency of the other ones that bugs me.
Em qui., 1 de jul. de 2021 às 17:20, Mahendra Singh Thalor <
mahi6...@gmail.com> escreveu:
> On Fri, 2 Jul 2021 at 01:13, Ranier Vilela wrote:
> >
> > Hi,
> >
> > The function FreePageManagerPutInternal can access an uninitialized
> variable,
> > if the following conditions occur:
>
> Patch
On 2021-Jul-02, Tom Lane wrote:
> Fabien COELHO writes:
> > Yes. Maybe decorations should be SQL comments, and the purpose/origin of
> > the query could be made clear as you suggest, eg something like markdown
> > in a comment:
> >"-- # QUERY\n%s\n\n"
>
> If we keep the decoration, I'd
Fabien COELHO writes:
> Yes. Maybe decorations should be SQL comments, and the purpose/origin of
> the query could be made clear as you suggest, eg something like markdown
> in a comment:
>"-- # QUERY\n%s\n\n"
If we keep the decoration, I'd agree with dropping all the asterisks.
I'd vote
Gavin Flower writes:
> On 2/07/21 8:39 pm, David Rowley wrote:
>> That's a good question. There was an argument in [1] that mentions
>> that there might be a group of people who rely on aggregation being
>> done in a certain order where they're not specifying an ORDER BY
>> clause in the
I wrote:
> Didn't look at 0002 yet.
... and now that I have, I don't like it much. It adds a lot of
complexity, plus some lookup cycles that might be wasted. I experimented
with looking into the range table as I suggested previously, and
that seems to work; see attached. (This includes a
On 2/07/21 8:39 pm, David Rowley wrote:
On Fri, 2 Jul 2021 at 19:54, Ronan Dunklau wrote:
I don't know if it's acceptable, but in the case where you add both an
aggregate with an ORDER BY clause, and another aggregate without the clause,
the output for the unordered one will change and use the
On Fri, 2021-07-02 at 11:39 +0500, Andrey Borodin wrote:
> If the failover happens due to unresponsive node we cannot just turn
> off sync rep. We need to have some spare connections for that (number
> of stuck backends will skyrocket during network partitioning). We
> need available descriptors
Bharath Rupireddy writes:
>> << Attaching v5-0001 here again for completion >>
>> I'm attaching v5-0001 patch that avoids printing the column type names
>> in the context message thus no cache lookups have to be done in the
>> error context callback. I think the column name is enough to know on
Hello Tom,
I went to commit this, figuring that it was a trivial bit of code
consolidation, but as I looked around in common.c I got rather
unhappy with the inconsistent behavior of things. Examining
the various places that implement "echo"-related logic, we have
the three places this patch
On Tue, Apr 06, 2021 at 11:01:31AM -0500, Justin Pryzby wrote:
> On Wed, Dec 23, 2020 at 01:17:10PM -0600, Justin Pryzby wrote:
> > On Mon, Nov 23, 2020 at 04:14:18PM -0500, Tom Lane wrote:
> > > * I noticed that you did s/stat/lstat/. That's fine on Unix systems,
> > > but it won't have any
On Fri, 2 Jul 2021 at 22:55, Dean Rasheed
wrote:
> Here's an update with the
> last set of changes discussed.
If you allow me a small suggestion.
Move the initializations of the variable tmp_var to after check if the
function can run.
Saves some cycles, when not running.
/* Ensure we disallow
Here's a more fleshed-out version of this patch. I ran around and
fixed all the places where INNER_VAR etc. were being assigned directly to
a variable or parameter of type Index, and also grepped for 'Index.*varno'
to find suspicious declarations. (I didn't change every last instance
of the
Em sex., 2 de jul. de 2021 às 13:29, Alvaro Herrera
escreveu:
> On 2021-Jul-02, Justin Pryzby wrote:
>
> > On Fri, Jul 02, 2021 at 12:09:23PM +0200, Peter Eisentraut wrote:
>
> > > The use of InvalidBlockNumber with vac_update_relstats() looks a bit
> fishy
> > > to me. We are using in the same
In my git workflow, I normally use scripts to simplify and check things.
Previously, most of my workfload was on master, with patches migrated to
appropriate back branches.
FYI, now that we have the release notes only in the major version
branches, I have had to adjust my scripts to allow for
On 2021-Jul-02, Justin Pryzby wrote:
> On Fri, Jul 02, 2021 at 12:09:23PM +0200, Peter Eisentraut wrote:
> > The use of InvalidBlockNumber with vac_update_relstats() looks a bit fishy
> > to me. We are using in the same call 0 as the default for
> > num_all_visible_pages, and we generally
On 2021-Jun-24, Peter Eisentraut wrote:
> There might be room for some wordsmithing in a few places, but generally I
> think this is complete.
This looks good to me. I am +0.1 on your proposal of "cannot have
triggers" vs Michael's "cannot create triggers", but really I could go
with either.
Alvaro Herrera writes:
> Now it's hoverfly:
> ! nm -A -u libpq.so.5 2>/dev/null | grep -v __cxa_atexit | grep exit
> libpq.so.5: atexit U -
> libpq.so.5: pthread_exit U -
Ugh. What in the world is producing those references?
(As I mentioned upthread,
Fabien COELHO writes:
> [ psql-echo-2.patch ]
I went to commit this, figuring that it was a trivial bit of code
consolidation, but as I looked around in common.c I got rather
unhappy with the inconsistent behavior of things. Examining
the various places that implement "echo"-related logic, we
Now it's hoverfly:
! nm -A -u libpq.so.5 2>/dev/null | grep -v __cxa_atexit | grep exit
libpq.so.5: atexit U -
libpq.so.5: pthread_exit U -
https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=hoverfly=2021-07-02%2010%3A10%3A29
--
Álvaro Herrera
On Fri, Jul 2, 2021 at 5:34 AM Craig Ringer
wrote:
>
> On Fri, 2 Jul 2021 at 00:19, Simon Riggs wrote:
>
>>
>> > So yeah. I think it'd be better to log the info you want at start-of-txn
>> > unless there's a compelling reason not so, and I don't see one yet.
>>
>> AFAIK, XLOG_XACT_ASSIGNMENT
On Fri, Jul 02, 2021 at 12:09:23PM +0200, Peter Eisentraut wrote:
> On 16.06.21 10:48, Peter Eisentraut wrote:
> > On 15.06.21 10:17, Kyotaro Horiguchi wrote:
> > > The definitions are not ((type) -1) but ((type) 0x) so
> > > actually they might be different if we forget to widen the
On Wed, Jun 30, 2021 at 11:46 PM Greg Nancarrow wrote:
> I personally think "(b) provide an option to the user to specify
> whether inserts can be parallelized on a relation" is the preferable
> option.
> There seems to be too many issues with the alternative of trying to
> determine the
On Tue, Jun 22, 2021 at 8:00 AM torikoshia wrote:
> Updated the patch.
Thanks for the patch. Here are some comments on the v4 patch:
1) Can we do + ExplainState *es = NewExplainState(); and es
assignments after if (!ActivePortal || !ActivePortal->queryDesc), just
to avoid unnecessary call in
On Fri, 2 Jul 2021 at 7:06 PM, vignesh C wrote:
> On Fri, Jul 2, 2021 at 6:05 PM Ibrar Ahmed wrote:
> >
> >
> >
> > On Fri, 2 Jul 2021 at 1:47 PM, David Rowley
> wrote:
> >>
> >> On Fri, 2 Jul 2021 at 15:04, vignesh C wrote:
> >> > I'm interested in assisting Ibrar Ahmed.
> >>
> >> It might
On Fri, Jul 2, 2021 at 6:05 PM Ibrar Ahmed wrote:
>
>
>
> On Fri, 2 Jul 2021 at 1:47 PM, David Rowley wrote:
>>
>> On Fri, 2 Jul 2021 at 15:04, vignesh C wrote:
>> > I'm interested in assisting Ibrar Ahmed.
>>
>> It might be worth talking to Ibrar to see where you can lend a hand. I
>> think in
On 7/1/21 9:53 PM, Michael Paquier wrote:
> On Thu, Jul 01, 2021 at 10:03:10AM -0400, Greg Sabino Mullane wrote:
>> Searching about the TAP tests often leads to this page, but there is no
>> easy link or mention of the fact that the sample invocations will not work
>> without the special config
Dear Hackers,
I revised my patch.
> Please also ensure that you're generating the patch against git HEAD.
> The cfbot shows it as failing to apply, likely because you're looking
> at something predating ad8305a43d1890768a613d3fb586b44f17360f29.
Maybe there was something wrong with my local
On Fri, Jul 2, 2021 at 1:55 AM Jeff Davis wrote:
> On Wed, 2021-06-30 at 12:25 -0400, Robert Haas wrote:
> > I am not sure whether this works or not. Holding off cancel
> > interrupts
> > across possible network I/O seems like a non-starter. We have to be
> > able to kill off connections that
On Fri, 2021-07-02 at 23:31 +1200, David Rowley wrote:
> I had a look at the patch in [1] and I find it a bit weird that we'd
> write the following about autovacuum_work_mem in our docs:
>
> +
> +Note that VACUUM has a hard-coded limit of 1GB
> +for the amount of memory
On Fri, 2 Jul 2021 at 1:47 PM, David Rowley wrote:
> On Fri, 2 Jul 2021 at 15:04, vignesh C wrote:
> > I'm interested in assisting Ibrar Ahmed.
>
> It might be worth talking to Ibrar to see where you can lend a hand. I
> think in terms of the number of patches, this might be our biggest
>
On Fri, 2 Jul 2021 at 22:55, Dean Rasheed wrote:
> Here's an update with the
> last set of changes discussed.
Looks good to me.
Just the question of if we have any problems changing the serialized
format in the back branches. I'm not sure if that's something we've
done before. I only had a
On Fri, 21 May 2021 at 03:52, Laurenz Albe wrote:
> Just sending a reply to -hackers so I can add it to the commitfest.
I had a look at the patch in [1] and I find it a bit weird that we'd
write the following about autovacuum_work_mem in our docs:
+
+Note that VACUUM has a
so 12. 6. 2021 v 8:00 odesílatel Pavel Stehule
napsal:
> Hi
>
> rebase
>
>
rebase only
Regards
Pavel
Regards
>
> Pavel
>
schema-variables-20210702.patch.gz
Description: application/gzip
Em sex., 2 de jul. de 2021 às 07:09, Peter Eisentraut <
peter.eisentr...@enterprisedb.com> escreveu:
> On 16.06.21 10:48, Peter Eisentraut wrote:
> > On 15.06.21 10:17, Kyotaro Horiguchi wrote:
> >> The definitions are not ((type) -1) but ((type) 0x) so
> >> actually they might be
On Fri, 2 Jul 2021 at 10:24, David Rowley wrote:
>
> I ran this again with a few different worker counts after tuning a few
> memory settings so there was no spilling to disk and so everything was
> in RAM. Mostly so I could get consistent results.
>
> Here's the results. Average over 3 runs on
On 02.07.21 08:25, Michael Paquier wrote:
+ ereport(ERROR,
+ (errcode(ERRCODE_WRONG_OBJECT_TYPE),
+errmsg("ALTER action %s cannot be performed on relation
\"%s\"",
+ action_str, RelationGetRelationName(rel)),
+
On Thu, 1 Jul 2021 at 11:09, Tom Lane wrote:
> Just reading it over quickly, I noticed a comment mentioning
> "aggcombinedfn" which I suppose should be "aggcombinefn".
Thanks. I've fixed that locally.
> No particular opinion on whether this is a net reduction
> of logical complexity.
I had
On Thursday, July 1, 2021 11:48 AM Ajin Cherian
>
> Adding a new patch (0004) to this patch-set that handles skipping of
> empty streamed transactions. patch-0003 did not
> handle empty streamed transactions. To support this, added a new flag
> "sent_stream_start" to PGOutputTxnData.
> Also
On 16.06.21 10:48, Peter Eisentraut wrote:
On 15.06.21 10:17, Kyotaro Horiguchi wrote:
The definitions are not ((type) -1) but ((type) 0x) so
actually they might be different if we forget to widen the constant
when widening the types. Regarding to the compiler behavior, I think
we are
On Fri, Jul 2, 2021 at 10:18 AM tanghy.f...@fujitsu.com
wrote:
>
> On Wednesday, June 30, 2021 7:43 PM vignesh C wrote:
> >
> > Thanks for reporting this issue, the attached v9 patch fixes this issue.
> > This also fixes the other issue you reported at [1].
>
> A comment on v9:
>
>
On Fri, 2 Jul 2021 at 00:28, Dean Rasheed wrote:
>
> On Thu, 1 Jul 2021 at 06:43, David Rowley wrote:
> >
> > Master @ 3788c6678
> >
> > Execution Time: 8306.319 ms
> > Execution Time: 8407.785 ms
> > Execution Time: 8491.056 ms
> >
> > Master + numeric-agg-sumX2-overflow-v2.patch
> > Execution
On Fri, 2 Jul 2021 at 15:04, vignesh C wrote:
> I'm interested in assisting Ibrar Ahmed.
It might be worth talking to Ibrar to see where you can lend a hand. I
think in terms of the number of patches, this might be our biggest
commitfest yet.
2020-07 246
2020-09 235
2020-11 244
2021-01 260
On Fri, 2 Jul 2021 at 19:54, Ronan Dunklau wrote:
> I don't know if it's acceptable, but in the case where you add both an
> aggregate with an ORDER BY clause, and another aggregate without the clause,
> the output for the unordered one will change and use the same ordering, maybe
> suprising the
On Thu, 1 Jul 2021 at 22:18, Fabien COELHO wrote:
>
> > However, this patch seems to be replacing that with a simple
> > modulo operation, which is perhaps the worst possible way to do it.
>
> The modulo operation is biased for large ranges close to the limit, sure.
> Also, the bias is somehow of
>
> This allows us to give presorted input to both aggregates in the following
> case:
>
> SELECT agg(a ORDER BY a),agg2(a ORDER BY a,b) ...
>
> but just the first agg in this one:
>
> SELECT agg(a ORDER BY a),agg2(a ORDER BY c) ...
I don't know if it's acceptable, but in the case where you
‐‐‐ Original Message ‐‐‐
On Friday, July 2nd, 2021 at 03:10, Michael Paquier wrote:
> On Thu, Jul 01, 2021 at 02:10:17PM +, gkokola...@pm.me wrote:
>
> > Micheal suggested on the same thread to move my entry in the help output so
> > that
> >
> > the output remains ordered. I
Hi.
I have been looking at the latest patch set (v16). Below are my review
comments and some patches.
The patches are:
v16-0001. This is identical to your previously posted 0001 patch. (I
only attached it again hoping it can allow the cfbot to keep working
OK).
v16-0002,0003. These are for
On Fri, Jul 2, 2021 at 8:35 AM Alvaro Herrera wrote:
>
> > The latest patch sent by Bharath looks good to me. Would you like to
> > commit it or shall I take care of it?
>
> Please, go ahead.
>
Okay, I'll push it early next week (by Tuesday) unless there are more
comments or suggestions. Thanks!
Hi,
I agree that —drop-cascade does not make sense for pg_dumpall, so I removed
them.
> are we expecting more things to appear after the semi-colon?
No, I was just trying to “reuse” original statement as much as possible. Append
“\n” manually should also do the job, and I’ve updated the
> 2 июля 2021 г., в 10:59, Jeff Davis написал(а):
>
> On Wed, 2021-06-30 at 17:28 +0500, Andrey Borodin wrote:
>>> My patch also covers the backend termination case. Is there a
>>> reason
>>> you left that case out?
>>
>> Yes, backend termination is used by HA tool before rewinding the
>>
On Fri, Jul 2, 2021 at 11:11 AM vignesh C wrote:
>
> Hi,
>
> I found a strange behavior when there is an insert after renaming the
> schema. The test steps for the same are given below, Here after the
> schema is renamed, the renamed schema table data should not be sent,
> but the data was being
On Fri, Jul 2, 2021 at 9:53 AM Michael Paquier wrote:
>
> On Mon, Jun 28, 2021 at 08:21:06PM +0530, Bharath Rupireddy wrote:
> > The recoveryWakeupLatch and procLatch/MyLatch are being used for WAL
> > replay and recovery conflict, respectively. Actually, I was earlier
> > using
On Thu, Jun 24, 2021 at 10:12:49AM +0200, Peter Eisentraut wrote:
> There might be room for some wordsmithing in a few places, but generally I
> think this is complete.
I have been looking at that, and it seems to me that you nailed it.
That's a nice improvement compared to the existing error
On Wed, 2021-06-30 at 17:28 +0500, Andrey Borodin wrote:
> > My patch also covers the backend termination case. Is there a
> > reason
> > you left that case out?
>
> Yes, backend termination is used by HA tool before rewinding the
> node.
Can't you just disable sync rep first (using ALTER SYSTEM
69 matches
Mail list logo
