James Coleman writes:
> On Mon, Apr 19, 2021 at 7:10 PM Tom Lane wrote:
>> After some more testing, that seems like a good thing to do,
>> so here's a v4.
> This all looks good to me.
Pushed, thanks for reviewing!
regards, tom lane
ilm...@ilmari.org (Dagfinn Ilmari =?utf-8?Q?Manns=C3=A5ker?=) writes:
> ilm...@ilmari.org (Dagfinn Ilmari Mannsåker) writes:
>> There are 10 instances of this exact loop scattered around the codebase.
>> Is it worth it turning it into a static inline function?
> Something like the attached, maybe?
On Tue, Apr 20, 2021 at 7:11 AM Dagfinn Ilmari Mannsåker
wrote:
>
> ilm...@ilmari.org (Dagfinn Ilmari Mannsåker) writes:
>
> > Tom Lane writes:
> >
> >> +/* We ignore binary-compatible relabeling on both ends */
> >> +while (expr && IsA(expr, RelabelType))
> >> +expr = ((Relab
ilm...@ilmari.org (Dagfinn Ilmari Mannsåker) writes:
> Tom Lane writes:
>
>> +/* We ignore binary-compatible relabeling on both ends */
>> +while (expr && IsA(expr, RelabelType))
>> +expr = ((RelabelType *) expr)->arg;
>
> There are 10 instances of this exact loop scattered ar
Tom Lane writes:
> + /* We ignore binary-compatible relabeling on both ends */
> + while (expr && IsA(expr, RelabelType))
> + expr = ((RelabelType *) expr)->arg;
There are 10 instances of this exact loop scattered around the codebase.
Is it worth it turning it into a static i
On Mon, Apr 19, 2021 at 7:10 PM Tom Lane wrote:
>
> I wrote:
> > Anyway I'm now inclined to remove that behavior from
> > find_computable_ec_member, and adjust comments accordingly.
>
> After some more testing, that seems like a good thing to do,
> so here's a v4.
This all looks good to me.
Jame
I wrote:
> Anyway I'm now inclined to remove that behavior from
> find_computable_ec_member, and adjust comments accordingly.
After some more testing, that seems like a good thing to do,
so here's a v4.
regards, tom lane
diff --git a/src/backend/optimizer/path/equivclass.
James Coleman writes:
> Two things I wonder:
> 1. Should we add tests for the relabel code path?
As far as that goes, the Relabel-stripping loops in
find_ec_member_matching_expr are already exercised in the core
regression tests (I didn't bother to discover exactly where, but
a quick coverage tes
I wrote:
> I'm not wedded to that name, certainly, but it seems like neither
> of these is quite getting at the issue. An EC can be sorted on,
> by definition, but there are some things we don't want to sort
> on till the final output step. I was trying to think of something
> using the terminolo
James Coleman writes:
> I forgot to comment on this in my previous email, but it seems to me
> that relation_has_safe_ec_member, while less wordy, isn't quite
> descriptive enough. Perhaps something like
> relation_has_sort_safe_ec_member?
I'm not wedded to that name, certainly, but it seems like
On Sat, Apr 17, 2021 at 3:39 PM Tom Lane wrote:
> ...
> Also, I don't much care for either the name or API of
> find_em_expr_usable_for_sorting_rel. The sole current caller only
> really needs a boolean result, and if it did need more than that
> it'd likely need the whole EquivalenceMember not j
On Sun, Apr 18, 2021 at 1:21 PM Tom Lane wrote:
>
> I wrote:
> > I think it's time for some refactoring of this code so that we can
> > actually share the logic. Accordingly, I propose the attached.
>
> After sleeping on it, here's an improved version that gets rid of
> an unnecessary assumption
I wrote:
> I think it's time for some refactoring of this code so that we can
> actually share the logic. Accordingly, I propose the attached.
After sleeping on it, here's an improved version that gets rid of
an unnecessary assumption about ECs usually not containing both
parallel-safe and parall
[ sorry for not getting to this thread till now ]
Tomas Vondra writes:
> 3) Shouldn't find_em_expr_usable_for_sorting_rel now mostly mimic what
> prepare_sort_from_pathkeys does? That is, try to match the entries
> directly first, before the new pull_vars() business?
Yeah. I concur that the pro
On Thu, Apr 15, 2021 at 6:27 PM Tomas Vondra
wrote:
>
>
> On 4/15/21 7:35 PM, James Coleman wrote:
> > On Thu, Apr 15, 2021 at 5:33 AM Luc Vlaming wrote:
> >>
> >> On 15-04-2021 04:01, James Coleman wrote:
> >>> On Wed, Apr 14, 2021 at 5:42 PM James Coleman
> wrote:
>
> On Mon, Apr 12
On 4/15/21 7:35 PM, James Coleman wrote:
> On Thu, Apr 15, 2021 at 5:33 AM Luc Vlaming wrote:
>>
>> On 15-04-2021 04:01, James Coleman wrote:
>>> On Wed, Apr 14, 2021 at 5:42 PM James Coleman wrote:
On Mon, Apr 12, 2021 at 8:37 AM Tomas Vondra
wrote:
>
> On 4/12/21 2:24
On 4/15/21 2:21 AM, Robert Haas wrote:
> On Wed, Apr 14, 2021 at 8:20 PM James Coleman wrote:
>>> Hmm, could be. Although, the stack trace at issue doesn't seem to show
>>> a call to create_incrementalsort_plan().
>>
>> The changes to gather merge path generation made it possible to use
>> thos
On Thu, Apr 15, 2021 at 5:33 AM Luc Vlaming wrote:
>
> On 15-04-2021 04:01, James Coleman wrote:
> > On Wed, Apr 14, 2021 at 5:42 PM James Coleman wrote:
> >>
> >> On Mon, Apr 12, 2021 at 8:37 AM Tomas Vondra
> >> wrote:
> >>>
> >>> On 4/12/21 2:24 PM, Luc Vlaming wrote:
> Hi,
>
>
On 15-04-2021 04:01, James Coleman wrote:
On Wed, Apr 14, 2021 at 5:42 PM James Coleman wrote:
On Mon, Apr 12, 2021 at 8:37 AM Tomas Vondra
wrote:
On 4/12/21 2:24 PM, Luc Vlaming wrote:
Hi,
When trying to run on master (but afaik also PG-13) TPC-DS queries 94,
95 and 96 on a SF10 I get th
On Wed, Apr 14, 2021 at 5:42 PM James Coleman wrote:
>
> On Mon, Apr 12, 2021 at 8:37 AM Tomas Vondra
> wrote:
> >
> > On 4/12/21 2:24 PM, Luc Vlaming wrote:
> > > Hi,
> > >
> > > When trying to run on master (but afaik also PG-13) TPC-DS queries 94,
> > > 95 and 96 on a SF10 I get the error "cou
On Wed, Apr 14, 2021 at 8:21 PM Robert Haas wrote:
>
> On Wed, Apr 14, 2021 at 5:43 PM James Coleman wrote:
> > The query in question is:
> > select count(*)
> > from store_sales
> > ,household_demographics
> > ,time_dim, store
> > where ss_sold_time_sk =
On Wed, Apr 14, 2021 at 8:21 PM Robert Haas wrote:
>
> On Wed, Apr 14, 2021 at 5:43 PM James Coleman wrote:
> > The query in question is:
> > select count(*)
> > from store_sales
> > ,household_demographics
> > ,time_dim, store
> > where ss_sold_time_sk =
On Wed, Apr 14, 2021 at 8:20 PM James Coleman wrote:
> > Hmm, could be. Although, the stack trace at issue doesn't seem to show
> > a call to create_incrementalsort_plan().
>
> The changes to gather merge path generation made it possible to use
> those paths in more cases for both incremental sort
On Wed, Apr 14, 2021 at 5:43 PM James Coleman wrote:
> The query in question is:
> select count(*)
> from store_sales
> ,household_demographics
> ,time_dim, store
> where ss_sold_time_sk = time_dim.t_time_sk
> and ss_hdemo_sk = household_demogra
On Wed, Apr 14, 2021 at 8:16 PM Robert Haas wrote:
>
> On Mon, Apr 12, 2021 at 8:37 AM Tomas Vondra
> wrote:
> > Could be related to incremental sort, which allowed some gather merge
> > paths that were impossible before. We had a couple issues related to
> > that fixed in November, IIRC.
>
> Hmm
On Mon, Apr 12, 2021 at 8:37 AM Tomas Vondra
wrote:
> Could be related to incremental sort, which allowed some gather merge
> paths that were impossible before. We had a couple issues related to
> that fixed in November, IIRC.
Hmm, could be. Although, the stack trace at issue doesn't seem to show
On Mon, Apr 12, 2021 at 8:37 AM Tomas Vondra
wrote:
>
> On 4/12/21 2:24 PM, Luc Vlaming wrote:
> > Hi,
> >
> > When trying to run on master (but afaik also PG-13) TPC-DS queries 94,
> > 95 and 96 on a SF10 I get the error "could not find pathkey item to sort".
> > When I disable enable_gathermerge
On 4/12/21 2:24 PM, Luc Vlaming wrote:
> Hi,
>
> When trying to run on master (but afaik also PG-13) TPC-DS queries 94,
> 95 and 96 on a SF10 I get the error "could not find pathkey item to sort".
> When I disable enable_gathermerge the problem goes away and then the
> plan for query 94 looks like
Hi,
When trying to run on master (but afaik also PG-13) TPC-DS queries 94,
95 and 96 on a SF10 I get the error "could not find pathkey item to sort".
When I disable enable_gathermerge the problem goes away and then the
plan for query 94 looks like below. I tried figuring out what the
problem i
29 matches
Mail list logo
