From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: [PATCH 0/7] sg_ring: a ring of scatterlist arrays Date: Mon, 07 Jan 2008 17:45:47 +0900 Message-ID: <4781E6BB.2010900@gmail.com> References: <200712191731.26512.rusty@rustcorp.com.au> <200801071628.11381.rusty@rustcorp.com.au> <4781C8B5.6050207@gmail.com> <200801071934.01597.rusty@rustcorp.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: Received: from wa-out-1112.google.com ([209.85.146.177]:10119 "EHLO wa-out-1112.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754914AbYAGIpz (ORCPT ); Mon, 7 Jan 2008 03:45:55 -0500 Received: by wa-out-1112.google.com with SMTP id v27so12384040wah.23 for ; Mon, 07 Jan 2008 00:45:54 -0800 (PST) In-Reply-To: <200801071934.01597.rusty@rustcorp.com.au> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Rusty Russell Cc: James Bottomley , Jens Axboe , linux-kernel@vger.kernel.org, linux-scsi@vger.kernel.org, linux-ide@vger.kernel.org Rusty Russell wrote: > On Monday 07 January 2008 17:37:41 Tejun Heo wrote: >> Rusty Russell wrote: >>> Hi Tejun, >>> >>> Nice try! Even ignoring the ugliness of undoing such an operation if >>> the caller doesn't expect you to mangle their chains, consider a >>> one-element sg array. :( >> Heh heh, that can be dealt with by skipping the first chain if the first >> chain is empty after chaining. Please take a look at >> ata_sg_setup_extra() in the following. >> >> http://git.kernel.org/?p=linux/kernel/git/tj/libata-dev.git;a=blob;f=driver >> s/ata/libata-core.c;h=32dde5bbc75ed53e89ac17040da2cd0621a37161;hb=c8847e473a >> 4a2844244784226eb362be10d52ce9 >> >> That said, yeah, it's seriously ugly. Restoring the original sg is ugly >> too. I definitely agree that we need some improvements here. > > Erk, that's beyond ugly, into actual evil. /me agrees. > To make this general you need to find the last N 1-element chains (but SCSI > doesn't do this of course). Oh the horror... /me agrees. > I'd be remiss if I didn't mention that the sg_ring ata patches were > straightforward, and indescribably beautiful if compared to this! /me agrees. As long as this can be made sane using one unified interface, I don't care whether it's sg_ring, table or dish. Thanks. -- tejun