From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: To: Sagi Grimberg , linux-kernel@vger.kernel.org, linux-pci@vger.kernel.org, linux-nvme@lists.infradead.org, linux-rdma@vger.kernel.org, linux-nvdimm@lists.01.org, linux-block@vger.kernel.org Cc: Stephen Bates , Christoph Hellwig , Keith Busch , Bjorn Helgaas , Jason Gunthorpe , Max Gurtovoy , Dan Williams , =?UTF-8?B?SsOpcsO0bWUgR2xpc3Nl?= , Benjamin Herrenschmidt , Alex Williamson , =?UTF-8?Q?Christian_K=c3=b6nig?= , Jens Axboe , Steve Wise References: <20180927165420.5290-1-logang@deltatee.com> <20180927165420.5290-14-logang@deltatee.com> <5ddeed51-0580-9581-cf12-c75e18b4f7cc@grimberg.me> <69cd8aab-b94e-98f5-5397-48bb875e8280@deltatee.com> From: Logan Gunthorpe Message-ID: <587bb6bd-ec07-62ac-6442-cbc7892df37f@deltatee.com> Date: Mon, 1 Oct 2018 17:43:00 -0600 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8 Subject: Re: [PATCH v8 13/13] nvmet: Optionally use PCI P2P memory List-ID: On 01/10/18 04:23 PM, Sagi Grimberg wrote: >> I did not realize the namespace would be available at this time. I guess >> I can give this a try, but it's going to be a fairly big change from >> what's presented here... Though, I agree it'll probably be an >> improvement. > > Thanks, if it turns out to create to much of a churn, we could defer > that to a later stage, but we can at least document it. Yeah, it's going to create a bunch of churn, but it's probably worth doing before merging because I think it will remove a bunch of complexity (ie. the need for the whole p2p client infrastructure because we now only need to worry about only one namespace at a time, instead of needing to find a p2p device that works with all namespaces at once). I'll try to get a v9 with this change published in the next day or two. Logan