netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jason Gunthorpe <jgg@nvidia.com>
To: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>,
	Jonathan Cameron <Jonathan.Cameron@huawei.com>,
	ksummit@lists.linux.dev, linux-cxl@vger.kernel.org,
	linux-rdma@vger.kernel.org, netdev@vger.kernel.org,
	shiju.jose@huawei.com, Borislav Petkov <bp@alien8.de>,
	Mauro Carvalho Chehab <mchehab@kernel.org>
Subject: Re: [MAINTAINERS SUMMIT] Device Passthrough Considered Harmful?
Date: Tue, 6 Aug 2024 10:04:10 -0300	[thread overview]
Message-ID: <20240806130410.GK478300@nvidia.com> (raw)
In-Reply-To: <ZrHNTBJV5aybQrum@phenom.ffwll.local>

On Tue, Aug 06, 2024 at 09:14:20AM +0200, Daniel Vetter wrote:
> On Thu, Aug 01, 2024 at 11:22:23AM -0300, Jason Gunthorpe wrote:
> > On Tue, Jul 30, 2024 at 09:13:00AM +0200, Daniel Vetter wrote:
> > > I think a solid consensus on the topics above would be really useful for
> > > gpu/accel too. We're still busy with more pressing community/ecosystem
> > > building needs, but gpu fw has become rather complex and it's not
> > > stopping. And there's random other devices attached too nowadays, so fwctl
> > > makes a ton of sense.
> > 
> > Yeah, I'm pretty sure GPU is going to need fwctl too, the GPU's are
> > going to have the same issues as NIC does. I see people are already
> > struggling with topics like how to get debug traces out of the GPU FW.
> > 
> > > But for me the more important stuff would be some clear guidelines like
> > > what should be in other more across-devices subsystems like edac (or other
> > > ras features), what should be in functional subsystems like netdev, rdma,
> > > gpu/accel, ... whatever else, and what should be exposed through some
> > > special purpose subsystems like hwmon.
> > 
> > In my mind the most important part is that fwctl is not exclusive, the
> > FW interface and things being manipulated must be sharable or blocked
> > from fwctl. We should never get in a situation where a fwctl
> > implementation becomes a reason we cannot have a functional subsystem
> > interface.
> 
> Hm still not clear to me how you want to achive that, but I guess best
> I'll jump over to the fwctl thread and ask about those details
> there.

I'm looking at it from the perspective of mlx5 which has deep
multi-user support in the FW. There is almost nothing in the interface
that is "global" and would become a problem. Everything else can, and
often already is, reasonably be shared.

I think that would have to be the baseline for what you could expose
here.

Like with the memory scrubbing example. It would be fine if fwctl can
read any related counters concurrently with the EDAC driver reading
the same counters. But fwctl shouldn't clear counters or program a
single global scrubber unit.

This limitation has to be baked into the FW/driver on the fwctl side
to undertsand and block these things.

Jason

      reply	other threads:[~2024-08-06 13:04 UTC|newest]

Thread overview: 126+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-07-08 22:26 [MAINTAINERS SUMMIT] Device Passthrough Considered Harmful? Dan Williams
2024-07-09  6:09 ` Christoph Hellwig
2024-07-09 19:43   ` Dan Williams
2024-07-10 13:05     ` Jason Gunthorpe
2024-07-21 18:51       ` Laurent Pinchart
2024-07-22 13:27         ` Jason Gunthorpe
2024-07-09 10:01 ` Greg KH
2024-07-09 12:25   ` Leon Romanovsky
2024-07-09 12:33     ` Greg KH
2024-07-09 12:47       ` Leon Romanovsky
2024-07-11 14:07         ` Jason Gunthorpe
2024-07-09 20:09   ` Dan Williams
2024-07-09 16:02 ` James Bottomley
2024-07-09 22:15   ` Dan Williams
2024-07-10 13:22     ` Jonathan Cameron
2024-07-11 11:00       ` Jonathan Cameron
2024-07-11 15:05       ` Jason Gunthorpe
2024-07-11 17:01         ` Jonathan Cameron
2024-07-11 17:45           ` Jason Gunthorpe
2024-07-11 16:36       ` Dan Williams
2024-07-12 10:37         ` Jonathan Cameron
2024-07-21 19:25     ` Laurent Pinchart
2024-07-22  7:31       ` Leon Romanovsky
2024-07-22  8:53         ` Laurent Pinchart
2024-07-22 10:44           ` Leon Romanovsky
2024-07-22 11:10             ` Laurent Pinchart
2024-07-22 13:28               ` Leon Romanovsky
2024-07-22 14:13                 ` Laurent Pinchart
2024-07-22 14:43                   ` Jason Gunthorpe
2024-07-22 10:42       ` Ricardo Ribalda Delgado
2024-07-22 11:18         ` Laurent Pinchart
2024-07-22 11:56           ` Ricardo Ribalda Delgado
2024-07-25 20:01             ` Laurent Pinchart
2024-07-26  8:04               ` Ricardo Ribalda Delgado
2024-07-26 10:59                 ` Laurent Pinchart
2024-07-26 15:40                   ` Ricardo Ribalda Delgado
2024-07-28 17:18                     ` Laurent Pinchart
2024-07-29  9:58                       ` Ricardo Ribalda Delgado
2024-07-29 10:31                         ` Laurent Pinchart
2024-07-31 11:54                         ` Sakari Ailus
2024-07-31 13:15                           ` Daniel Vetter
2024-08-02 15:07                             ` Laurent Pinchart
2024-08-13 10:17                             ` Tomasz Figa
2024-08-13 10:26                               ` Laurent Pinchart
2024-08-13 10:33                                 ` Tomasz Figa
2024-08-13 10:58                                   ` Laurent Pinchart
2024-07-11 13:50   ` Jason Gunthorpe
2024-07-11 15:16     ` James Bottomley
2024-07-11 16:29       ` Jason Gunthorpe
2024-07-23 11:20 ` Jiri Kosina
2024-07-23 11:36   ` James Bottomley
2024-07-23 23:22     ` Jiri Kosina
2024-07-24 20:12       ` James Bottomley
2024-07-24 20:00     ` Laurent Pinchart
2024-07-24 20:37       ` James Bottomley
2024-07-24 21:10         ` Steven Rostedt
2024-07-25 19:31         ` Laurent Pinchart
2024-07-25 19:43           ` Jason Gunthorpe
2024-07-25 20:07             ` Laurent Pinchart
2024-07-25 23:39               ` Jason Gunthorpe
2024-07-26  8:04               ` Ricardo Ribalda Delgado
2024-07-26 12:49                 ` Laurent Pinchart
2024-07-26 13:11                   ` Jason Gunthorpe
2024-07-26 14:22                     ` Laurent Pinchart
2024-07-26 15:43                       ` Ricardo Ribalda Delgado
2024-07-28 15:25                         ` Laurent Pinchart
2024-07-29  9:57                           ` Ricardo Ribalda Delgado
2024-07-29 14:20                       ` Jason Gunthorpe
2024-07-26  8:03           ` Ricardo Ribalda Delgado
2024-07-26 13:22             ` Laurent Pinchart
2024-07-26 15:44               ` Ricardo Ribalda Delgado
2024-07-28 17:02                 ` Laurent Pinchart
2024-07-26 16:49           ` James Bottomley
2024-07-28 16:44             ` Laurent Pinchart
2024-07-26 17:33         ` Daniel Vetter
2024-07-29 14:10           ` Jason Gunthorpe
2024-07-25  9:26       ` Ricardo Ribalda Delgado
2024-07-25 10:51         ` Wolfram Sang
2024-07-25 12:23         ` Leon Romanovsky
2024-07-25 13:02           ` Ricardo Ribalda Delgado
2024-07-25 13:20             ` Leon Romanovsky
2024-07-25 13:29               ` Mark Brown
2024-07-25 14:18                 ` Leon Romanovsky
2024-07-25 14:22                   ` James Bottomley
2024-07-25 17:37                     ` Leon Romanovsky
2024-07-26 13:58                       ` James Bottomley
2024-07-25 19:42               ` Laurent Pinchart
2024-07-26  8:02                 ` Ricardo Ribalda Delgado
2024-07-26 13:11                   ` Laurent Pinchart
2024-07-26 15:40                     ` Ricardo Ribalda Delgado
2024-07-28 11:23                       ` Laurent Pinchart
2024-07-29  9:56                         ` Ricardo Ribalda Delgado
2024-07-29 10:38                           ` Laurent Pinchart
2024-07-26 16:01                     ` James Bottomley
2024-07-26 17:56                       ` Laurent Pinchart
2024-07-25 13:44             ` Steven Rostedt
2024-07-26 14:27 ` Laurent Pinchart
2024-07-26 15:34   ` Steven Rostedt
2024-07-28 16:03     ` Laurent Pinchart
2024-07-27  0:16   ` Dan Williams
2024-07-28 11:18     ` Laurent Pinchart
2024-07-28 15:16       ` Greg KH
2024-07-28 15:34         ` Laurent Pinchart
2024-07-28 15:49         ` James Bottomley
2024-07-29  6:10           ` Greg KH
2024-07-31 12:33             ` James Bottomley
2024-07-31 12:45               ` Laurent Pinchart
2024-08-01 14:41               ` Jason Gunthorpe
2024-08-07  0:06                 ` Dan Williams
2024-08-07  0:13                   ` James Bottomley
2024-08-16 11:12                   ` Hannes Reinecke
2024-07-29 14:56           ` Jakub Kicinski
2024-07-29 15:16             ` Greg KH
2024-07-29 15:29             ` Jason Gunthorpe
2024-07-29 12:45 ` Jonathan Cameron
2024-07-29 13:38   ` Borislav Petkov
2024-07-29 14:29     ` Jonathan Cameron
2024-07-29 14:58       ` Jason Gunthorpe
2024-07-30 13:19         ` Borislav Petkov
2024-08-01 14:23           ` Jason Gunthorpe
2024-07-29 15:42   ` Jason Gunthorpe
2024-07-29 22:37     ` Dan Williams
2024-07-30  7:13       ` Daniel Vetter
2024-08-01 14:22         ` Jason Gunthorpe
2024-08-06  7:14           ` Daniel Vetter
2024-08-06 13:04             ` Jason Gunthorpe [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20240806130410.GK478300@nvidia.com \
    --to=jgg@nvidia.com \
    --cc=Jonathan.Cameron@huawei.com \
    --cc=bp@alien8.de \
    --cc=dan.j.williams@intel.com \
    --cc=daniel.vetter@ffwll.ch \
    --cc=ksummit@lists.linux.dev \
    --cc=linux-cxl@vger.kernel.org \
    --cc=linux-rdma@vger.kernel.org \
    --cc=mchehab@kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=shiju.jose@huawei.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).