Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932555AbdDRUbN (ORCPT ); Tue, 18 Apr 2017 16:31:13 -0400 Received: from mail-oi0-f41.google.com ([209.85.218.41]:33437 "EHLO mail-oi0-f41.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754461AbdDRUbL (ORCPT ); Tue, 18 Apr 2017 16:31:11 -0400 MIME-Version: 1.0 In-Reply-To: <1565047873.28680036.1492547352121.JavaMail.zimbra@redhat.com> References: <1492381396.25766.43.camel@kernel.crashing.org> <20170418164557.GA7181@obsidianresearch.com> <20170418190138.GH7181@obsidianresearch.com> <1565047873.28680036.1492547352121.JavaMail.zimbra@redhat.com> From: Dan Williams Date: Tue, 18 Apr 2017 13:31:09 -0700 Message-ID: Subject: Re: [RFC 0/8] Copy Offload with Peer-to-Peer PCI Memory To: Jerome Glisse Cc: Logan Gunthorpe , Jason Gunthorpe , Benjamin Herrenschmidt , Bjorn Helgaas , Christoph Hellwig , Sagi Grimberg , "James E.J. Bottomley" , "Martin K. Petersen" , Jens Axboe , Steve Wise , Stephen Bates , Max Gurtovoy , Keith Busch , linux-pci@vger.kernel.org, linux-scsi , linux-nvme@lists.infradead.org, linux-rdma@vger.kernel.org, linux-nvdimm , "linux-kernel@vger.kernel.org" Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1348 Lines: 28 On Tue, Apr 18, 2017 at 1:29 PM, Jerome Glisse wrote: >> On Tue, Apr 18, 2017 at 12:35 PM, Logan Gunthorpe >> wrote: >> > >> > >> > On 18/04/17 01:01 PM, Jason Gunthorpe wrote: >> >> Ultimately every dma_ops will need special code to support P2P with >> >> the special hardware that ops is controlling, so it makes some sense >> >> to start by pushing the check down there in the first place. This >> >> advice is partially motivated by how dma_map_sg is just a small >> >> wrapper around the function pointer call... >> > >> > Yes, I noticed this problem too and that makes sense. It just means >> > every dma_ops will probably need to be modified to either support p2p >> > pages or fail on them. Though, the only real difficulty there is that it >> > will be a lot of work. >> >> I don't think you need to go touch all dma_ops, I think you can just >> arrange for devices that are going to do dma to get redirected to a >> p2p aware provider of operations that overrides the system default >> dma_ops. I.e. just touch get_dma_ops(). > > This would not work well for everyone, for instance on GPU we usualy > have buffer object with a mix of device memory and regular system > memory but call dma sg map once for the list. > ...and that dma_map goes through get_dma_ops(), so I don't see the conflict?