From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Sat, 4 Dec 2010 19:55:36 +0300 From: Alexey Tourbin To: ALT Devel discussion list Message-ID: <20101204165535.GD15470@altlinux.org> References: <1290722666-24606-1-git-send-email-kirill@shutemov.name> <20101126083537.GA26344@shutemov.name> <20101130003541.GA20093@altlinux.org> <20101204130217.GA3406@altlinux.org> <20101204163048.GB12275@altlinux.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20101204163048.GB12275@altlinux.org> Subject: Re: [devel] [PATCH 0/3] optimize rpmsetcmp() X-BeenThere: devel@lists.altlinux.org X-Mailman-Version: 2.1.12 Precedence: list Reply-To: ALT Linux Team development discussions List-Id: ALT Linux Team development discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 04 Dec 2010 16:55:36 -0000 Archived-At: List-Archive: List-Post: On Sat, Dec 04, 2010 at 07:30:49PM +0300, Dmitry V. Levin wrote: > > I think I've made more amazing a change. > > http://git.altlinux.org/gears/r/rpm.git?a=commitdiff;h=fc5b5a5f > > Have you tried to combine both optimizations? What's the criteria of doing good enough? I always thought that 'char bitv[]' is good enough. Let's call it a reference implementation. Kirill argued that changing bitv to bitmap could actually win both some space and some time, as there's a possibility to handle 6 bits at a time, and even Mshift bits at a time. The problem is that, well, *bitv++ is replaced with something else. The code will be no longer a reference implementation.