X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (av.hub.org [200.46.204.144]) by postgresql.org (Postfix) with ESMTP id 7322F9DCAC9 for ; Fri, 20 Jan 2006 16:21:48 -0400 (AST) Received: from postgresql.org ([200.46.204.71]) by localhost (av.hub.org [200.46.204.144]) (amavisd-new, port 10024) with ESMTP id 56401-01 for ; Fri, 20 Jan 2006 16:21:48 -0400 (AST) X-Greylist: from auto-whitelisted by SQLgrey- Received: from sss.pgh.pa.us (sss.pgh.pa.us [66.207.139.130]) by postgresql.org (Postfix) with ESMTP id 1466D9DCA39 for ; Fri, 20 Jan 2006 16:21:45 -0400 (AST) Received: from sss2.sss.pgh.pa.us (tgl@localhost [127.0.0.1]) by sss.pgh.pa.us (8.13.1/8.13.1) with ESMTP id k0KKLjRt026584; Fri, 20 Jan 2006 15:21:45 -0500 (EST) To: Martijn van Oosterhout cc: Stephan Vollmer , pgsql-performance@postgreSQL.org Subject: Re: [GENERAL] Creation of tsearch2 index is very slow In-reply-to: <20060120201036.GE31908@svana.org> References: <43D0ED57.1020608@gmx.de> <17454.1137771321@sss.pgh.pa.us> <20060120170452.GC31908@svana.org> <26153.1137784469@sss.pgh.pa.us> <20060120201036.GE31908@svana.org> Comments: In-reply-to Martijn van Oosterhout message dated "Fri, 20 Jan 2006 21:10:36 +0100" Date: Fri, 20 Jan 2006 15:21:45 -0500 Message-ID: <26583.1137788505@sss.pgh.pa.us> From: Tom Lane X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, score=0.094 required=5 tests=[AWL=0.094] X-Spam-Score: 0.094 X-Spam-Level: X-Archive-Number: 200601/360 X-Sequence-Number: 16838 Martijn van Oosterhout writes: > Something I'm missing is the calls to tsearch functions. I'm not 100% > familiar with gprof, but is it possible those costs have been added > somewhere else because it's in a shared library? Perhaps the costs went > into FunctionCall1/3? I think that the tsearch functions must be the stuff charged as "data_start" (which is not actually any symbol in our source code). That is showing as being called by FunctionCallN which is what you'd expect. If the totals given by gprof are correct, then it's down in the noise. I don't think I trust that too much ... but I don't see anything in the gprof manual about how to include a dynamically loaded library in the profile. (I did compile tsearch2 with -pg, but that's evidently not enough.) I'll see if I can link tsearch2 statically to resolve this question. regards, tom lane