From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Bottomley Subject: Re: Request for review of Linux iSCSI driver version 4.0.0.1 Date: 02 Dec 2003 17:55:14 -0600 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <1070409314.1955.11.camel@mulgrave> References: <03120118001300.08627@naveenb-lnx.cisco.com> <03120217260300.01630@naveenb-lnx.cisco.com> <1070383028.2345.8.camel@mulgrave> <20031202174202.GA1960@beaverton.ibm.com> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from nat9.steeleye.com ([65.114.3.137]:7172 "EHLO hancock.sc.steeleye.com") by vger.kernel.org with ESMTP id S264471AbTLBX4d (ORCPT ); Tue, 2 Dec 2003 18:56:33 -0500 In-Reply-To: <20031202174202.GA1960@beaverton.ibm.com> List-Id: linux-scsi@vger.kernel.org To: Mike Anderson Cc: naveenb@cisco.com, Roman Zippel , hch@infradead.org, SCSI Mailing List On Tue, 2003-12-02 at 11:42, Mike Anderson wrote: > The user would need to generate another write request and the > "guarantee" is up to the user (correct?)? Well, not necessarily. There are APIs to flush mmapped data which the user can call. However, the most likely thing (especially if the user is careless---how many times do you code your apps to do fflush() for instance?) is simply that automatic reaping will flush the page for you, or failing that it will be flushed on process exit. For mmapped file descriptors, there are other events which trigger flushing, but I forget exactly what they are. James