Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Thu, 11 Feb 2021 08:50:21 +0000
From:   Christoph Hellwig <hch@infradead.org>
To:     Max Gurtovoy <mgurtovoy@nvidia.com>
Cc:     Alexey Kardashevskiy <aik@ozlabs.ru>,
        Cornelia Huck <cohuck@redhat.com>,
        Alex Williamson <alex.williamson@redhat.com>,
        Matthew Rosato <mjrosato@linux.ibm.com>, jgg@nvidia.com,
        kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
        liranl@nvidia.com, oren@nvidia.com, tzahio@nvidia.com,
        leonro@nvidia.com, yarong@nvidia.com, aviadye@nvidia.com,
        shahafs@nvidia.com, artemp@nvidia.com, kwankhede@nvidia.com,
        ACurrid@nvidia.com, gmataev@nvidia.com, cjia@nvidia.com,
        yishaih@nvidia.com
Subject: Re: [PATCH 8/9] vfio/pci: use x86 naming instead of igd
Message-ID: <20210211085021.GD2378134@infradead.org>
References: <20210201162828.5938-1-mgurtovoy@nvidia.com>
 <20210201162828.5938-9-mgurtovoy@nvidia.com>
 <20210201181454.22112b57.cohuck@redhat.com>
 <599c6452-8ba6-a00a-65e7-0167f21eac35@linux.ibm.com>
 <20210201114230.37c18abd@omen.home.shazbot.org>
 <20210202170659.1c62a9e8.cohuck@redhat.com>
 <a413334c-3319-c6a3-3d8a-0bb68a10b9c1@nvidia.com>
 <806c138e-685c-0955-7c15-93cb1d4fe0d9@ozlabs.ru>
 <6c96f41a-0daa-12d4-528e-6db44df1a5a6@nvidia.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <6c96f41a-0daa-12d4-528e-6db44df1a5a6@nvidia.com>
Precedence: bulk

On Thu, Feb 04, 2021 at 11:12:49AM +0200, Max Gurtovoy wrote:
> But the PCI function (the bounded BDF) is GPU function or NVLINK function ?
> 
> If it's NVLINK function then we should fail probing in the host vfio-pci
> driver.
> 
> if its a GPU function so it shouldn't been called nvlink2 vfio-pci driver.
> Its just an extension in the GPU vfio-pci driver.

I suspect the trivial and correct answer is that we should just drop the
driver entirely.  It is for obsolete hardware that never had upstream
support for even using it in the guest.  It also is the reason for
keeping cruft in the always built-in powernv platform code alive that
is otherwise dead wood.