Re: profiling connection overhead

From: Robert Haas <robertmhaas(at)gmail(dot)com>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com>, pgsql-hackers(at)postgresql(dot)org
Subject: Re: profiling connection overhead
Date: 2010-11-24 19:30:47
Message-ID: AANLkTi=8EWfpTpoo7ZHnw1_1-KwrfyWmM1DO6txq3yTp@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox
Thread:
Lists: pgsql-hackers

On Wed, Nov 24, 2010 at 1:20 PM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> I tried configuring oprofile with --callgraph=10 and then running
> oprofile with -c, but it gives kooky looking output I can't interpret.

It looks like the trick is to compile with -fno-omit-frame-pointer.
New profiling run:

27563 10.3470 libc-2.11.2.so memset
15162 5.6917 libc-2.11.2.so memcpy
13471 5.0569 postgres hash_search_with_hash_value
13465 5.0547 postgres AllocSetAlloc
9513 3.5711 libc-2.11.2.so _int_malloc
8729 3.2768 libc-2.11.2.so fread
6336 2.3785 ld-2.11.2.so do_lookup_x
5788 2.1728 ld-2.11.2.so _dl_fixup
4995 1.8751 postgres MemoryContextAllocZero
4978 1.8687 ld-2.11.2.so strcmp

Full results, and call graph, attached. The first obvious fact is
that most of the memset overhead appears to be coming from
InitCatCache.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

Attachment Content-Type Size
reconnect-callgraph.txt.bz2 application/x-bzip2 34.9 KB
reconnect-opreport.txt.bz2 application/x-bzip2 6.8 KB

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Andres Freund 2010-11-24 19:32:20 Re: profiling connection overhead
Previous Message Gerhard Heift 2010-11-24 18:47:13 Re: profiling connection overhead