From: | Mark Kirkwood <markir(at)paradise(dot)net(dot)nz> |
---|---|
To: | Luke Lonergan <llonergan(at)greenplum(dot)com> |
Cc: | Qingqing Zhou <zhouqq(at)cs(dot)toronto(dot)edu>, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: Which qsort is used |
Date: | 2005-12-17 02:15:01 |
Message-ID: | 43A374A5.2060205@paradise.net.nz |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Luke Lonergan wrote:
> Qingqing,
>
>
> On 12/15/05 6:33 PM, "Qingqing Zhou" <zhouqq(at)cs(dot)toronto(dot)edu> wrote:
>
>>Thanks for Greg "let" me take a second look at qsortB.c - there is
>>paste-and-copy error there, so when it perform recursive sort, it calls
>>glibc's qsort ... Really sorry, and feel a little bit gun-shy now ...
>>
>>After I re-tested it, now BSD qsort is the obvious winner in most
>>situations.
>
>
> :-D
>
> Can you post the new results like the last post?
>
> - Luke
>
Here is a result from a dual 0.8G x86 running Freebsd 6.0-RELEASE:
http://homepages.paradise.net.nz/markir/download/sort-fbsd.out
(after patching the bug with qsortB calling qsort). Clearly in this
case, there is no glibc version, hence I've relabeled the 1st case as
"native qsort".
Cheers
Mark
From | Date | Subject | |
---|---|---|---|
Next Message | Bruce Momjian | 2005-12-17 02:44:19 | Re: How much expensive are row level statistics? |
Previous Message | Bruce Momjian | 2005-12-17 01:47:53 | Re: number of loaded/unloaded COPY rows |