Re: inline newNode() - Mailing list pgsql-patches

From Bruce Momjian
Subject Re: inline newNode()
Date
Msg-id 200210120006.g9C06i502964@candle.pha.pa.us
Whole thread Raw
In response to Re: inline newNode()  (Bruce Momjian <pgman@candle.pha.pa.us>)
List pgsql-patches
Bruce Momjian wrote:
> Neil Conway wrote:
> > BTW, regarding the newNode() stuff: so is it agreed that Bruce's patch
> > is a performance win without too high of a code bloat / uglification
> > penalty? If so, is it 7.3 or 7.4 material?
>
> Not sure.  It is a small patch but we normally don't do performance
> fixes during beta unless they are major.

This performance improvement isn't related to a specific user problem
report, so I am going to hold it for 7.4.  At that time we can also
research whether palloc0 should be used in other cases, like palloc
followed by MemSet.

The only issue there is that palloc0 can't optimize away constants used
in the macro so it may be better _not_ to make that change.  Neil, do
you have any performance numbers comparing MemSet with constant vs.
variable parameters, e.g:

    MemSet(ptr, 0, 256)

vs.

    i = 256;
    MemSet(ptr, 0, i)

--
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

pgsql-patches by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: INSTALL updates
Next
From: Bruce Momjian
Date:
Subject: Re: inline newNode()