From: Andrew Morton Subject: Re: [Resubmit][Patch 0/2] Persistent preallocation in ext4 Date: Sun, 25 Feb 2007 02:23:26 -0800 Message-ID: <20070225022326.137b4875.akpm@linux-foundation.org> References: <20070117094658.GA17390@amitarora.in.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: linux-ext4@vger.kernel.org, suparna@in.ibm.com, cmm@us.ibm.com, alex@clusterfs.com, suzuki@in.ibm.com To: "Amit K. Arora" Return-path: Received: from smtp.osdl.org ([65.172.181.24]:58988 "EHLO smtp.osdl.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933690AbXBYKZF (ORCPT ); Sun, 25 Feb 2007 05:25:05 -0500 In-Reply-To: <20070117094658.GA17390@amitarora.in.ibm.com> Sender: linux-ext4-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org > On Wed, 17 Jan 2007 15:16:58 +0530 "Amit K. Arora" wrote: > (1) The final interface is yet to be decided. We have the option of > chosing from one of these: > a> modifying posix_fallocate() in glibc > b> using fcntl > c> using ftruncate, or > d> using the ioctl interface. I'd suggest we add a new syscall, which implements the posix_fallocate() API as per the relevant specs. We can work the final details out with Ulrich and I'm sure that glibc's posix_fallocate() will start using the syscall if it is available. Actually, we should implement asmlinkage long sys_fallocate(int fd, loff_t offset, loff_t len); note: loff_t, not off_t. This probably means that we'll need to implement file_operations.fallocate(). It wouldn't surprise me if XFS was able to implement fallocate() too.