From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755383Ab1KUIIf (ORCPT ); Mon, 21 Nov 2011 03:08:35 -0500 Received: from nat28.tlf.novell.com ([130.57.49.28]:44295 "EHLO nat28.tlf.novell.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754414Ab1KUIId convert rfc822-to-8bit (ORCPT ); Mon, 21 Nov 2011 03:08:33 -0500 Message-Id: <4ECA150C0200007800062072@nat28.tlf.novell.com> X-Mailer: Novell GroupWise Internet Agent 12.0.0 Date: Mon, 21 Nov 2011 08:08:28 +0000 From: "Jan Beulich" To: "Kevin Cernekee" , "Rusty Russell" Cc: , Subject: Re: [PATCH V2 2/2] module: Fix performance regression on modules with large symbol tables References: <4EC0E4040200007800060B1C@nat28.tlf.novell.com> <63fcc782681c8b3bd330164871ad4785@localhost> <87ipme9i9o.fsf@rustcorp.com.au> In-Reply-To: <87ipme9i9o.fsf@rustcorp.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8BIT Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org >>> On 21.11.11 at 06:06, Rusty Russell wrote: > On Thu, 17 Nov 2011 19:15:02 -0800, Kevin Cernekee wrote: >> Potential memory savings from reusing EXACT strtab matches: 36,475 bytes >> total; 35,432 from nvidia.ko. For comparison, discarding non-core >> strtab entries saved about 91KB on my system. > ... >> Original scheme: 1.230s >> With patch V2: 0.280s >> With naive copying: 0.058s > > I'm deeply tempted. It's very simple, 46 lines shorter, preserves exact > matches, and doesn't have any strange slowdowns on corner cases. > > But like Kevin, I could be convinced either way. Jan? I'd be fine with the simpler solution, as long it is clearly understood (and marked as intended to be that was for the time being) that it has room for improvement, should it ever turn out to yield more significant savings in other contexts. Jan