Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S269719AbUICRrl (ORCPT ); Fri, 3 Sep 2004 13:47:41 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S269724AbUICRqm (ORCPT ); Fri, 3 Sep 2004 13:46:42 -0400 Received: from fed1rmmtao05.cox.net ([68.230.241.34]:32978 "EHLO fed1rmmtao05.cox.net") by vger.kernel.org with ESMTP id S269719AbUICRkj (ORCPT ); Fri, 3 Sep 2004 13:40:39 -0400 Date: Fri, 3 Sep 2004 10:40:10 -0700 From: Tom Rini To: Paul Mackerras Cc: Kernel Mailing List Subject: Re: [PATCH,RFC 2.6.9-rc1 2/2] zlib_inflate: Add __BOOTER__ around zlib_inflate_trees_fixed(...) Message-ID: <20040903174010.GB6290@smtp.west.cox.net> References: <20040901231659.GA20624@smtp.west.cox.net> <20040902173626.GB26144@smtp.west.cox.net> <20040902174707.GC26144@smtp.west.cox.net> <16695.50554.389435.137893@cargo.ozlabs.ibm.com> <20040903014520.GE26144@smtp.west.cox.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20040903014520.GE26144@smtp.west.cox.net> User-Agent: Mutt/1.5.6+20040818i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1329 Lines: 32 On Thu, Sep 02, 2004 at 06:45:20PM -0700, Tom Rini wrote: > On Fri, Sep 03, 2004 at 11:14:34AM +1000, Paul Mackerras wrote: > > > Tom Rini writes: > > > > > This is the second part of what I found. zlib_inflate_trees_fixed(...) > > > isn't called in decompressing a kernel. Dropping this, and the call to > > > > I think it is just luck that gzip hasn't used the fixed table in > > compressing the kernel. I don't think we have any guarantee that gzip > > won't use the fixed table. > > That is a good point. Looking at arch/ppc/boot/lib/zlib.c it generates > the table, instead of having a static one. So perhaps we should move to > that instead for lib/zlib_inflate ? I've moved lib/zlib_inflate over to generating the table. But, in trying to test it, I haven't been able to find anything that will actually use this (it appears that using this static table was something that pkzip introduced long ago as an alternative method of good compression / speed) so I can't actually test it. But the image size is back in line. -- Tom Rini http://gate.crashing.org/~trini/ - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/