linux-kernel - Re: [RFC PATCH 3/5] mm/vma: add support for peer to peer to device vma

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <39edda88-acd8-c8ec-a6a5-64898006b69f@deltatee.com>
Date:   Wed, 30 Jan 2019 10:32:03 -0700
From:   Logan Gunthorpe <logang@...tatee.com>
To:     Christoph Hellwig <hch@....de>, Jerome Glisse <jglisse@...hat.com>
Cc:     "Koenig, Christian" <Christian.Koenig@....com>,
        Jason Gunthorpe <jgg@...lanox.com>,
        "linux-mm@...ck.org" <linux-mm@...ck.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        "Rafael J . Wysocki" <rafael@...nel.org>,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        "Kuehling, Felix" <Felix.Kuehling@....com>,
        "linux-pci@...r.kernel.org" <linux-pci@...r.kernel.org>,
        "dri-devel@...ts.freedesktop.org" <dri-devel@...ts.freedesktop.org>,
        Marek Szyprowski <m.szyprowski@...sung.com>,
        Robin Murphy <robin.murphy@....com>,
        Joerg Roedel <jroedel@...e.de>,
        "iommu@...ts.linux-foundation.org" <iommu@...ts.linux-foundation.org>
Subject: Re: [RFC PATCH 3/5] mm/vma: add support for peer to peer to device
 vma



On 2019-01-30 10:26 a.m., Christoph Hellwig wrote:
> On Wed, Jan 30, 2019 at 10:55:43AM -0500, Jerome Glisse wrote:
>> Even outside GPU driver, device driver like RDMA just want to share their
>> doorbell to other device and they do not want to see those doorbell page
>> use in direct I/O or anything similar AFAICT.
> 
> At least Mellanox HCA support and inline data feature where you
> can copy data directly into the BAR.  For something like a usrspace
> NVMe target it might be very useful to do direct I/O straight into
> the BAR for that.

Yup, these are things we definitely want to be able to do, and have done
with hacky garbage code: Direct I/O from NVMe to P2P BAR, then we could
Direct I/O to another drive or map it as an MR and send it over an RNIC.

We'd definitely like to move in that direction. And a world where such
userspace mappings are gimpped by the fact that they are only some
special feature of userspace VMAs that can only be used in specialized
userspace interfaces is not useful to us.

Logan