From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757641AbaHGNv0 (ORCPT ); Thu, 7 Aug 2014 09:51:26 -0400 Received: from mx1.redhat.com ([209.132.183.28]:51793 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757334AbaHGNvZ (ORCPT ); Thu, 7 Aug 2014 09:51:25 -0400 Date: Thu, 7 Aug 2014 15:51:10 +0200 From: Karel Zak To: Boaz Harrosh Cc: Ross Zwisler , Boaz Harrosh , Jens Axboe , Matthew Wilcox , linux-kernel , linux-fsdevel Subject: Re: [PATCH 4/4] brd: Request from fdisk 4k alignment Message-ID: <20140807135110.GC9401@x2.net.home> References: <53E21125.9080308@plexistor.com> <53E21317.7080403@plexistor.com> <1407362634.2869.2.camel@rzwisler-mobl1.amr.corp.intel.com> <53E36E53.3090008@gmail.com> <20140807130042.GB9401@x2.net.home> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20140807130042.GB9401@x2.net.home> User-Agent: Mutt/1.5.22.1 (2013-10-16) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Aug 07, 2014 at 03:00:42PM +0200, Karel Zak wrote: > On Thu, Aug 07, 2014 at 03:17:23PM +0300, Boaz Harrosh wrote: > > > > Is there an error case that this patch fixes? I've had page alignment checks > > > in my PRD direct_access code forever, and I don't know if they've ever > > > tripped. > > > > > > > Yes! as I said above fix fdisk. You never tripped on it because partitions never > > worked and you never tried them. With current code fdisk is very trigger happy > > What do you mean with fdisk? which version? Oh, I read all your email more carefully now, and if I good understand the problem is what I/O limits the device provides to userspace rather than with fdisk. Anyway, I think you're right that alignment_offset is not the right thing. It was introduced for backward compatibility with DOS-like partitioning tools (~magical sector 63). I have doubts it's usable for something else. For normal use-case should be enough to set proper phy-sector size or min/optimal I/O limits (e.g. zram has all the limits set to 4K(PAGE_SIZE)). $ lsblk --topology /dev/zram0 NAME ALIGNMENT MIN-IO OPT-IO PHY-SEC LOG-SEC ROTA SCHED RQ-SIZE RA WSAME zram0 0 4096 4096 4096 4096 0 128 128 0B Karel -- Karel Zak http://karelzak.blogspot.com