Return-Path: Received: from mail-qt0-f196.google.com ([209.85.216.196]:33584 "EHLO mail-qt0-f196.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751041AbdFTThp (ORCPT ); Tue, 20 Jun 2017 15:37:45 -0400 MIME-Version: 1.0 In-Reply-To: <3836534F-8360-4EF4-926B-C5322A44D95E@oracle.com> References: <5bca6687-ac03-72ef-f38e-6759a0fbb1d6@gmail.com> <20170614185335.58193-1-kolga@netapp.com> <20170615035923.GA4521@birch.djwong.org> <20170615205357.GA13715@birch.djwong.org> <20170619193903.GA12803@infradead.org> <3836534F-8360-4EF4-926B-C5322A44D95E@oracle.com> From: Olga Kornievskaia Date: Tue, 20 Jun 2017 15:37:43 -0400 Message-ID: Subject: Re: [PATCH 1/1] [RFC] 64bit copy_file_range system call To: Chuck Lever Cc: Christoph Hellwig , "Darrick J. Wong" , Olga Kornievskaia , "linux-fsdevel@vger.kernel.org" , Linux NFS Mailing List Content-Type: text/plain; charset="UTF-8" Sender: linux-nfs-owner@vger.kernel.org List-ID: On Tue, Jun 20, 2017 at 3:33 PM, Chuck Lever wrote: > >> On Jun 19, 2017, at 3:39 PM, Christoph Hellwig wrote: >> >> On Mon, Jun 19, 2017 at 02:34:58PM -0400, Olga Kornievskaia wrote: >>> Yes NFS community does need one for doing a server-to-server copy >>> (performance) feature. >> >> s/NFS community/NetApp/ > > I don't have an opinion on the API details, but Oracle is interested > in server-to-server copy, especially if there is an implementation > of it in the Linux NFS server. The proposed patch set is the linux client AND linux server. We need removal of the VFS cross-device check on both side (client and destination server) to make the server-to-server copy work. On the server side, we also call vfs_copy_file_range() that (when check is removed) ends up reading from the NFS source file and writing to the local file system.