From: Jason Gunthorpe <jgg@nvidia.com>
To: Jakub Kicinski <kuba@kernel.org>
Cc: Edward Cree <ecree.xilinx@gmail.com>,
David Ahern <dsahern@kernel.org>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Christoph Hellwig <hch@infradead.org>,
Saeed Mahameed <saeed@kernel.org>, Arnd Bergmann <arnd@arndb.de>,
Leon Romanovsky <leonro@nvidia.com>, Jiri Pirko <jiri@nvidia.com>,
Leonid Bloch <lbloch@nvidia.com>,
Itay Avraham <itayavr@nvidia.com>,
Saeed Mahameed <saeedm@nvidia.com>,
Aron Silverton <aron.silverton@oracle.com>,
linux-kernel@vger.kernel.org,
"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
Andy Gospodarek <andrew.gospodarek@broadcom.com>
Subject: Re: [PATCH V4 0/5] mlx5 ConnectX control misc driver
Date: Fri, 5 Apr 2024 08:13:06 -0300 [thread overview]
Message-ID: <20240405111306.GB5383@nvidia.com> (raw)
In-Reply-To: <20240404143407.64b44a88@kernel.org>
On Thu, Apr 04, 2024 at 02:34:07PM -0700, Jakub Kicinski wrote:
> On Thu, 4 Apr 2024 17:44:54 -0300 Jason Gunthorpe wrote:
> > > To my knowledge the "customizations" are mostly around fitting into OCP
> > > servers.
> >
> > Nope. I understand it is significant. If Meta had to work with a COTS
> > environment like a HP/Dell customer then Meta would have a list of
> > flash configurables to set. I think you greatly underestimate the
> > privilege of being at a hyperscaler and having vendors create custom
> > products just for you..
> >
> > > Those unfamiliar with how hyperscalers operate can mentally
> > > replace $hyperscaler with HP or Dell in your message. Minus all the
> > > proprietary OOB management stuff those guys also provide.
> >
> > A significant Dell customer will get a server pre-populated with a NIC
> > with some generic Dell configuration. In most cases the customer will
> > have to then program the flash to match their needs. Set a specific FW
> > version, set site specific configurables, etc.
> >
> > Similar to how a Dell customer will have to change the BIOS settings
> > in the Dell to match their needs.
>
> I can only guess that you are again thinking about RDMA/HPC.
> Flashing tunables is not a workable solution for extremely varied
> and ephemeral TCP/IP workloads :|
As I answered to Anderew, a lot is functional behavior not so much
"tunables". The same way many BIOS settings are not all tunables but
have functional impacts to the machine. Like enable SRIOV, for
instance.
Even for dataplane tunables - you know there are micro-architectural
performance tunables set in the special Meta NICs that are wired just
for Meta's special use case? Apparently that is actually perfectly
workable.
It is really strange to hear you act like "Meta doesn't need
provisioning or tuning" when the NIC Meta uses is *highly* customized
specifically for Meta to the point it is an entirely different
product. Of course you don't need provisioning, alot of other people
did alot of hard work to make it that way.
So please don't use that as a justification to pull up the ladder so
nobody else can enjoy even a semi-customized device.
Jason
next prev parent reply other threads:[~2024-04-05 11:13 UTC|newest]
Thread overview: 102+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-02-07 7:24 [PATCH V4 0/5] mlx5 ConnectX control misc driver Saeed Mahameed
2024-02-07 7:24 ` [PATCH V4 1/5] mlx5: Add aux dev for ctl interface Saeed Mahameed
2024-02-07 7:24 ` [PATCH V4 2/5] misc: mlx5ctl: Add mlx5ctl misc driver Saeed Mahameed
2024-02-29 11:44 ` Vegard Nossum
2024-03-02 8:04 ` Saeed Mahameed
2024-02-07 7:24 ` [PATCH V4 3/5] misc: mlx5ctl: Add info ioctl Saeed Mahameed
2024-02-29 11:47 ` Vegard Nossum
2024-03-02 8:03 ` Saeed Mahameed
2024-02-07 7:24 ` [PATCH V4 4/5] misc: mlx5ctl: Add command rpc ioctl Saeed Mahameed
2024-02-29 11:49 ` Vegard Nossum
2024-03-02 7:48 ` Saeed Mahameed
2024-02-07 7:24 ` [PATCH V4 5/5] misc: mlx5ctl: Add umem reg/unreg ioctl Saeed Mahameed
2024-02-29 11:51 ` Vegard Nossum
2024-02-07 15:03 ` [PATCH V4 0/5] mlx5 ConnectX control misc driver Jakub Kicinski
2024-02-08 5:03 ` Saeed Mahameed
2024-02-09 2:15 ` Jakub Kicinski
2024-02-09 6:55 ` Jiri Pirko
2024-02-09 22:42 ` David Ahern
2024-02-09 22:58 ` Jakub Kicinski
2024-02-10 5:01 ` David Ahern
2024-02-11 11:03 ` Greg Kroah-Hartman
2024-02-11 17:01 ` David Ahern
2024-02-14 20:31 ` David Ahern
2024-02-15 0:46 ` Jason Gunthorpe
2024-02-10 1:01 ` Jason Gunthorpe
2024-02-11 16:59 ` David Ahern
2024-02-14 8:29 ` Christoph Hellwig
2024-02-14 15:48 ` Jakub Kicinski
2024-02-15 7:00 ` Christoph Hellwig
2024-02-15 12:08 ` Jiri Pirko
2024-02-16 1:00 ` Jakub Kicinski
2024-02-16 15:05 ` Jason Gunthorpe
2024-02-15 13:21 ` Jason Gunthorpe
2024-02-16 1:10 ` Jakub Kicinski
2024-02-16 4:20 ` David Ahern
2024-02-16 19:04 ` Jason Gunthorpe
2024-02-14 16:17 ` Andy Gospodarek
2024-02-14 17:57 ` Jason Gunthorpe
2024-02-14 18:11 ` Jakub Kicinski
2024-02-14 18:37 ` Jason Gunthorpe
2024-02-16 1:40 ` Jakub Kicinski
2024-02-16 14:27 ` Jason Gunthorpe
2024-03-04 16:02 ` Jason Gunthorpe
2024-03-22 3:23 ` David Ahern
2024-03-22 7:32 ` Greg Kroah-Hartman
2024-03-22 15:24 ` David Ahern
2024-03-22 15:46 ` Andy Gospodarek
2024-03-22 20:58 ` Jakub Kicinski
2024-03-22 21:18 ` David Ahern
2024-03-22 22:40 ` Jakub Kicinski
2024-03-26 14:57 ` David Ahern
2024-04-01 12:30 ` Leon Romanovsky
2024-04-01 14:50 ` Jakub Kicinski
2024-04-01 18:10 ` Leon Romanovsky
2024-04-01 19:04 ` Jakub Kicinski
2024-04-02 19:20 ` Leon Romanovsky
2024-04-02 18:45 ` Jason Gunthorpe
2024-04-02 21:36 ` Jakub Kicinski
2024-04-02 22:46 ` Jason Gunthorpe
2024-04-02 23:21 ` Jakub Kicinski
2024-04-03 0:15 ` Jakub Kicinski
2024-04-03 6:57 ` Leon Romanovsky
2024-04-02 16:32 ` Edward Cree
2024-04-02 18:40 ` Jason Gunthorpe
2024-04-03 19:28 ` David Ahern
2024-04-04 17:35 ` Edward Cree
2024-04-04 18:33 ` Jason Gunthorpe
2024-04-04 19:31 ` Edward Cree
2024-04-05 11:21 ` Jason Gunthorpe
2024-04-04 19:53 ` Jakub Kicinski
2024-04-04 20:44 ` Jason Gunthorpe
2024-04-04 21:34 ` Jakub Kicinski
2024-04-05 11:13 ` Jason Gunthorpe [this message]
2024-04-05 15:38 ` Jakub Kicinski
2024-04-05 17:48 ` Jakub Kicinski
2024-04-08 16:45 ` Jason Gunthorpe
2024-04-08 16:41 ` Jason Gunthorpe
2024-04-04 18:44 ` Andrew Lunn
2024-04-04 20:25 ` Jason Gunthorpe
2024-04-04 20:53 ` Edward Cree
2024-04-05 11:00 ` Jason Gunthorpe
2024-04-02 18:48 ` Leon Romanovsky
2024-04-03 12:26 ` Edward Cree
2024-04-03 19:00 ` Leon Romanovsky
2024-04-03 19:31 ` David Ahern
2024-04-04 0:01 ` Jakub Kicinski
2024-04-04 3:57 ` David Ahern
2024-04-04 12:23 ` Jason Gunthorpe
2024-04-04 14:48 ` Jakub Kicinski
2024-04-04 17:47 ` Jason Gunthorpe
2024-04-04 18:06 ` Edward Cree
2024-04-04 18:35 ` Leon Romanovsky
2024-04-04 19:46 ` Edward Cree
2024-04-05 10:41 ` Leon Romanovsky
2024-04-08 8:02 ` Przemek Kitszel
2024-03-22 21:44 ` Jason Gunthorpe
2024-03-22 22:29 ` Jakub Kicinski
2024-03-23 1:27 ` Saeed Mahameed
2024-03-23 1:33 ` Jason Gunthorpe
2024-03-22 14:53 ` Aron Silverton
2024-04-30 1:36 ` David Ahern
2024-04-30 7:09 ` Greg Kroah-Hartman
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=20240405111306.GB5383@nvidia.com \
--to=jgg@nvidia.com \
--cc=andrew.gospodarek@broadcom.com \
--cc=arnd@arndb.de \
--cc=aron.silverton@oracle.com \
--cc=dsahern@kernel.org \
--cc=ecree.xilinx@gmail.com \
--cc=gregkh@linuxfoundation.org \
--cc=hch@infradead.org \
--cc=itayavr@nvidia.com \
--cc=jiri@nvidia.com \
--cc=kuba@kernel.org \
--cc=lbloch@nvidia.com \
--cc=leonro@nvidia.com \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=saeed@kernel.org \
--cc=saeedm@nvidia.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.