[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CADnq5_PH3xR_nnN2wQ2NjydUoT+X0z_9OvHtaabwcPbwP-o+9A@mail.gmail.com>
Date: Tue, 25 Oct 2022 10:21:34 -0400
From: Alex Deucher <alexdeucher@...il.com>
To: Jason Gunthorpe <jgg@...dia.com>
Cc: Dave Airlie <airlied@...il.com>,
Tvrtko Ursulin <tvrtko.ursulin@...ux.intel.com>,
Jiho Chu <jiho.chu@...sung.com>,
Jeffrey Hugo <quic_jhugo@...cinc.com>,
Thomas Zimmermann <tzimmermann@...e.de>,
Arnd Bergmann <arnd@...db.de>,
John Hubbard <jhubbard@...dia.com>,
Oded Gabbay <ogabbay@...nel.org>, linux-kernel@...r.kernel.org,
dri-devel@...ts.freedesktop.org,
Christoph Hellwig <hch@...radead.org>,
Jacek Lawrynowicz <jacek.lawrynowicz@...ux.intel.com>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Alex Deucher <alexander.deucher@....com>,
Yuji Ishikawa <yuji2.ishikawa@...hiba.co.jp>,
Kevin Hilman <khilman@...libre.com>,
Maciej Kwapulinski <maciej.kwapulinski@...ux.intel.com>,
Jagan Teki <jagan@...rulasolutions.com>
Subject: Re: [RFC PATCH 0/3] new subsystem for compute accelerator devices
On Tue, Oct 25, 2022 at 7:15 AM Jason Gunthorpe <jgg@...dia.com> wrote:
>
> On Tue, Oct 25, 2022 at 12:27:11PM +1000, Dave Airlie wrote:
>
> > The userspace for those is normally bespoke like ROCm, which uses
> > amdkfd, and amdkfd doesn't operate like most device files from what I
> > know, so I'm not sure we'd want it to operate as an accel device.
>
> I intensely dislike this direction that drivers will create their own
> char devs buried inside their device driver with no support or
> supervision.
>
> We've been here before with RDMA and it is just a complete mess.
>
> Whatever special non-drm stuff amdkfd need to do should be supported
> through the new subsystem, in a proper maintainable way.
We plan to eventually move ROCm over the drm interfaces once we get
user mode queues working on non-compute queues which is already in
progress. ROCm already uses the existing drm nodes and libdrm for a
number of things today (buffer sharing, media and compute command
submission in certain cases, etc.). I don't see much value in the
accel nodes for AMD products at this time. Even when we transition,
there are still a bunch of things that we'd need to think about, so
the current kfd node may stick around until we figure out a plan for
those areas. E.g., the kfd node provides platform level compute
topology information; e.g., the NUMA details for connected GPUs and
CPUs, non-GPU compute node information, cache level topologies, etc.
Alex
>
> Jason
Powered by blists - more mailing lists