Re: path toward faster partition pruning

From: Amit Langote <Langote_Amit_f8(at)lab(dot)ntt(dot)co(dot)jp>
To: amul sul <sulamul(at)gmail(dot)com>
Cc: David Rowley <david(dot)rowley(at)2ndquadrant(dot)com>, Robert Haas <robertmhaas(at)gmail(dot)com>, Dilip Kumar <dilipbalaut(at)gmail(dot)com>, Rajkumar Raghuwanshi <rajkumar(dot)raghuwanshi(at)enterprisedb(dot)com>, Beena Emerson <memissemerson(at)gmail(dot)com>, Pg Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: path toward faster partition pruning
Date: 2017-11-10 00:34:57
Message-ID: 5fcb1a9f-b4ad-119d-14c7-282c30c7f8d1@lab.ntt.co.jp
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi Amul.

On 2017/11/09 20:05, amul sul wrote:
> On Mon, Nov 6, 2017 at 3:31 PM, Amit Langote
> <Langote_Amit_f8(at)lab(dot)ntt(dot)co(dot)jp> wrote:
>> On 2017/11/06 14:32, David Rowley wrote:
>>> On 6 November 2017 at 17:30, Amit Langote wrote:
>>>> On 2017/11/03 13:32, David Rowley wrote:
>>>>> On 31 October 2017 at 21:43, Amit Langote wrote:
> [....]
>>
>> Attached updated set of patches, including the fix to make the new pruning
>> code handle Boolean partitioning.
>>
>
> I am getting following warning on mac os x:

Thanks for the review.

> partition.c:1800:24: warning: comparison of constant -1 with
> expression of type 'NullTestType'
> (aka 'enum NullTestType') is always false
> [-Wtautological-constant-out-of-range-compare]
> if (keynullness[i] == -1)
> ~~~~~~~~~~~~~~ ^ ~~
> partition.c:1932:25: warning: comparison of constant -1 with
> expression of type 'NullTestType'
> (aka 'enum NullTestType') is always false
> [-Wtautological-constant-out-of-range-compare]
> if (keynullness[i] == -1)
> ~~~~~~~~~~~~~~ ^ ~~
> 2 warnings generated.
>
>
> Comment for 0004 patch:
> 270 + /* -1 represents an invalid value of NullTestType. */
> 271 + memset(keynullness, -1, PARTITION_MAX_KEYS * sizeof(NullTestType));
>
> I think we should not use memset to set a value other than 0 or true/false.
> This will work for -1 on the system where values are stored in the 2's
> complement but I am afraid of other architecture.

OK, I will remove all instances of comparing and setting variables of type
NullTestType to a value of -1.

Thanks,
Amit

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Huong Dangminh 2017-11-10 00:41:26 User defined data types in Logical Replication
Previous Message Peter Geoghegan 2017-11-10 00:19:30 Re: [HACKERS] pgsql: Fix freezing of a dead HOT-updated tuple