From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756867AbbAZSCJ (ORCPT ); Mon, 26 Jan 2015 13:02:09 -0500 Received: from mga03.intel.com ([134.134.136.65]:61019 "EHLO mga03.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755991AbbAZSCG (ORCPT ); Mon, 26 Jan 2015 13:02:06 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.09,470,1418112000"; d="scan'208";a="642764951" Date: Mon, 26 Jan 2015 18:02:04 +0000 (UTC) From: Keith Busch X-X-Sender: vmware@localhost.lm.intel.com To: Christoph Hellwig cc: Yan Liu , Matthew Wilcox , linux-kernel@vger.kernel.org, linux-nvme@lists.infradead.org Subject: Re: [PATCH 1/1] NVMe: Do not take nsid while a passthrough IO command is being issued via a block device file descriptor In-Reply-To: <20150125145900.GA15231@infradead.org> Message-ID: References: <1422057426-8014-1-git-send-email-yan@purestorage.com> <20150125145900.GA15231@infradead.org> User-Agent: Alpine 2.00 (LNX 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sun, 25 Jan 2015, Christoph Hellwig wrote: > On Fri, Jan 23, 2015 at 03:57:06PM -0800, Yan Liu wrote: >> When a passthrough IO command is issued with a specific block device file descriptor. It should be applied at >> the namespace which is associated with that block device file descriptor. This patch makes such passthrough >> command ignore nsid in nvme_passthru_cmd structure. Instead it takes the namespace ID asscoiated with the >> block device descriptor. > > This looks good to me. If Keith can point to a use case for magic or > hidden nsids we'll have to find a bypass for them, but this certainly > is the right short term fix: 'Flush' is the only command from the spec where specifying all namespaces might be desirable. Otherwise, I can't disclose vendor specific behavior to gain concensus on operating within spec, silly as this desire may seem. Why empower the driver to make a feature unreachable? From mboxrd@z Thu Jan 1 00:00:00 1970 From: keith.busch@intel.com (Keith Busch) Date: Mon, 26 Jan 2015 18:02:04 +0000 (UTC) Subject: [PATCH 1/1] NVMe: Do not take nsid while a passthrough IO command is being issued via a block device file descriptor In-Reply-To: <20150125145900.GA15231@infradead.org> References: <1422057426-8014-1-git-send-email-yan@purestorage.com> <20150125145900.GA15231@infradead.org> Message-ID: On Sun, 25 Jan 2015, Christoph Hellwig wrote: > On Fri, Jan 23, 2015@03:57:06PM -0800, Yan Liu wrote: >> When a passthrough IO command is issued with a specific block device file descriptor. It should be applied at >> the namespace which is associated with that block device file descriptor. This patch makes such passthrough >> command ignore nsid in nvme_passthru_cmd structure. Instead it takes the namespace ID asscoiated with the >> block device descriptor. > > This looks good to me. If Keith can point to a use case for magic or > hidden nsids we'll have to find a bypass for them, but this certainly > is the right short term fix: 'Flush' is the only command from the spec where specifying all namespaces might be desirable. Otherwise, I can't disclose vendor specific behavior to gain concensus on operating within spec, silly as this desire may seem. Why empower the driver to make a feature unreachable?