I think your first thought was good.
How high ? I think it's a matter of convention, certainly more than default
100.
czw., 21 lis 2019 o 02:05 Andy Fan napisał(a):
>
>
> On Thu, Oct 31, 2019 at 11:07 PM Tom Lane wrote:
>
>>
>>
>> Possibly this could be finessed by only trying to find duplica
On Thu, Oct 31, 2019 at 11:07 PM Tom Lane wrote:
>
>
> Possibly this could be finessed by only trying to find duplicates of
> functions that have high cost estimates. Not sure how high is high
> enough.
can we just add a flag on pg_proc to show if the cost is high or not, if
user are not happ
Hi
I need advice.
ResetExprContext(econtext) is defined as
MemoryContextReset((econtext)->ecxt_per_tuple_memory).
I can register callback in MemoryContext but it is always cleaned on every
call to MemoryContextReset().
How to reset some fields of ExprContext ( living in per_query_memory ) when
Res
Hi,
On 2019-11-03 21:56:31 +0100, Andrzej Barszcz wrote:
> Main goal of this patch is to avoid repeated calls of immutable/stable
> functions.
> This patch is against version 10.10.
> I guess same logic could be implemented up till version 12.
If you want actual development feedback, you're more
Hi,
On October 31, 2019 8:51:11 AM PDT, Andrzej Barszcz wrote:
>x <> 0 is evaluated first, 1/x only when x <> 0, not ?
>
>czw., 31 paź 2019 o 16:45 Tom Lane napisał(a):
>
>> Andres Freund writes:
>> > Potentially related note: for nodes like seqscan, combining the
>qual and
>> projection proce
x <> 0 is evaluated first, 1/x only when x <> 0, not ?
czw., 31 paź 2019 o 16:45 Tom Lane napisał(a):
> Andres Freund writes:
> > Potentially related note: for nodes like seqscan, combining the qual and
> projection processing into one expression seems to be a noticable win (at
> least when tak
Hi,
On October 31, 2019 8:45:26 AM PDT, Tom Lane wrote:
>Andres Freund writes:
>> Potentially related note: for nodes like seqscan, combining the qual
>and projection processing into one expression seems to be a noticable
>win (at least when taking care do emit two different sets of deform
>exp
Andres Freund writes:
> Potentially related note: for nodes like seqscan, combining the qual and
> projection processing into one expression seems to be a noticable win (at
> least when taking care do emit two different sets of deform expression steps).
There's just one problem: that violates S
Hi,
On October 31, 2019 8:06:50 AM PDT, Tom Lane wrote:
>Andres Freund writes:
>> On October 31, 2019 7:45:26 AM PDT, Tom Lane
>wrote:
>>> We've typically supposed that the cost of searching for duplicate
>>> subexpressions would outweigh the benefits of sometimes finding
>them.
>
>> Based on
Andres Freund writes:
> On October 31, 2019 7:45:26 AM PDT, Tom Lane wrote:
>> We've typically supposed that the cost of searching for duplicate
>> subexpressions would outweigh the benefits of sometimes finding them.
> Based on profiles I've seen I'm not sure that's the right choice. Both for
On 10/31/19 3:45 PM, Tom Lane wrote:
Andres Freund writes:
On October 31, 2019 7:06:13 AM PDT, Andrzej Barszcz wrote:
Any pros and cons ?
Depends on the actual way of implementing this proposal. Think we need more
details than what you idea here.
We've typically supposed that the cost
Hi
On October 31, 2019 7:53:20 AM PDT, Andres Freund wrote:
>On October 31, 2019 7:45:26 AM PDT, Tom Lane wrote:
>>We've typically supposed that the cost of searching for duplicate
>>subexpressions would outweigh the benefits of sometimes finding them.
>
>Based on profiles I've seen I'm not sure
Hi,
On October 31, 2019 7:45:26 AM PDT, Tom Lane wrote:
>Andres Freund writes:
>> On October 31, 2019 7:06:13 AM PDT, Andrzej Barszcz
> wrote:
>>> I almost finished patch optimizing non volatile function calls.
>>>
>>> select f(t.n) from t where f(t.n) > 10 and f(t.n) < 100; needs 3
>calls
>>
Andres Freund writes:
> On October 31, 2019 7:06:13 AM PDT, Andrzej Barszcz wrote:
>> I almost finished patch optimizing non volatile function calls.
>>
>> select f(t.n) from t where f(t.n) > 10 and f(t.n) < 100; needs 3 calls
>> of
>> f() for each tuple,
>> after applying patch only 1.
>>
>>
Hi,
On October 31, 2019 7:06:13 AM PDT, Andrzej Barszcz wrote:
>Hi
>
>I almost finished patch optimizing non volatile function calls.
>
>select f(t.n) from t where f(t.n) > 10 and f(t.n) < 100; needs 3 calls
>of
>f() for each tuple,
>after applying patch only 1.
>
>Any pros and cons ?
Depends
15 matches
Mail list logo
