Re: WIP: further sorting speedup

On Thu, Feb 23, 2006 at 01:15:32PM -0500, Tom Lane wrote:
> "Jim C. Nasby" <jnasby(at)pervasive(dot)com> writes:
> > On Wed, Feb 22, 2006 at 10:18:48PM -0500, Tom Lane wrote:
> >> The Datum is just a pointer into the original tuple in that case.
>
> > That would still result in a speedup since you don't have to figure out
> > where the field begins though, right? I'm curious as to how much this
> > patch would help with sorting text...
>
> Right. It would help just as much as in the integer case as far as
> eliminating the time spent in heap_getattr is concerned. That would be
> a smaller percentage of the whole, because bttextcmp is (a lot) slower
> than btint4cmp, but that's no fault of the patch.

Certainly, thanks for the explanation. Of course this does raise the
question of if it would be worthwhile to cache the first X bytes of a
varlena, but that's a different fish to fry.
--
Jim C. Nasby, Sr. Engineering Consultant jnasby(at)pervasive(dot)com
Pervasive Software http://pervasive.com work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf cell: 512-569-9461