From: Jason Gunthorpe <jgg@nvidia.com>
To: Jakub Kicinski <kuba@kernel.org>
Cc: Saeed Mahameed <saeedm@nvidia.com>, Eli Cohen <elic@nvidia.com>,
Leon Romanovsky <leonro@mellanox.com>, <netdev@vger.kernel.org>,
<linux-rdma@vger.kernel.org>, Eli Cohen <eli@mellanox.com>,
Mark Bloch <mbloch@nvidia.com>, Maor Gottlieb <maorg@nvidia.com>
Subject: Re: [PATCH mlx5-next 11/16] net/mlx5: Add VDPA priority to NIC RX namespace
Date: Tue, 24 Nov 2020 14:02:10 -0400 [thread overview]
Message-ID: <20201124180210.GJ5487@ziepe.ca> (raw)
In-Reply-To: <20201124091219.5900e7bb@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
On Tue, Nov 24, 2020 at 09:12:19AM -0800, Jakub Kicinski wrote:
> On Sun, 22 Nov 2020 08:41:58 +0200 Eli Cohen wrote:
> > On Sat, Nov 21, 2020 at 04:01:55PM -0800, Jakub Kicinski wrote:
> > > On Fri, 20 Nov 2020 15:03:34 -0800 Saeed Mahameed wrote:
> > > > From: Eli Cohen <eli@mellanox.com>
> > > >
> > > > Add a new namespace type to the NIC RX root namespace to allow for
> > > > inserting VDPA rules before regular NIC but after bypass, thus allowing
> > > > DPDK to have precedence in packet processing.
> > >
> > > How does DPDK and VDPA relate in this context?
> >
> > mlx5 steering is hierarchical and defines precedence amongst namespaces.
> > Up till now, the VDPA implementation would insert a rule into the
> > MLX5_FLOW_NAMESPACE_BYPASS hierarchy which is used by DPDK thus taking
> > all the incoming traffic.
> >
> > The MLX5_FLOW_NAMESPACE_VDPA hirerachy comes after
> > MLX5_FLOW_NAMESPACE_BYPASS.
>
> Our policy was no DPDK driver bifurcation. There's no asterisk saying
> "unless you pretend you need flow filters for RDMA, get them upstream
> and then drop the act".
Huh?
mlx5 DPDK is an *RDMA* userspace application. It links to
libibverbs. It runs on the RDMA stack. It uses RDMA flow filtering and
RDMA raw ethernet QPs. It has been like this for years, it is not some
"act".
It is long standing uABI that accelerators like RDMA/etc get to take
the traffic before netdev. This cannot be reverted. I don't really
understand what you are expecting here?
Jason
next prev parent reply other threads:[~2020-11-24 18:02 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-11-20 23:03 [PATCH mlx5-next 00/16] mlx5 next updates 2020-11-20 Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 01/16] net/mlx5: Add sample offload hardware bits and structures Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 02/16] net/mlx5: Add sampler destination type Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 03/16] net/mlx5: Check dr mask size against mlx5_match_param size Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 04/16] net/mlx5: Add misc4 to mlx5_ifc_fte_match_param_bits Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 05/16] net/mlx5: Add ts_cqe_to_dest_cqn related bits Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 06/16] net/mlx5: Avoid exposing driver internal command helpers Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 07/16] net/mlx5: Update the list of the PCI supported devices Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 08/16] net/mlx5: Update the hardware interface definition for vhca state Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 09/16] net/mlx5: Expose IP-in-IP TX and RX capability bits Saeed Mahameed
2020-11-21 23:58 ` Jakub Kicinski
2020-11-22 15:17 ` Aya Levin
2020-11-23 21:15 ` Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 10/16] net/mlx5: Expose other function ifc bits Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 11/16] net/mlx5: Add VDPA priority to NIC RX namespace Saeed Mahameed
2020-11-22 0:01 ` Jakub Kicinski
2020-11-22 6:41 ` Eli Cohen
2020-11-24 17:12 ` Jakub Kicinski
2020-11-24 18:02 ` Jason Gunthorpe [this message]
2020-11-24 18:41 ` Jakub Kicinski
2020-11-24 19:44 ` Jason Gunthorpe
2020-11-25 6:19 ` Eli Cohen
2020-11-25 19:04 ` Saeed Mahameed
2020-11-25 18:54 ` Jakub Kicinski
2020-11-25 19:28 ` Saeed Mahameed
2020-11-25 21:22 ` Jason Gunthorpe
2020-11-20 23:03 ` [PATCH mlx5-next 12/16] net/mlx5: Export steering related functions Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 13/16] net/mlx5: Make API mlx5_core_is_ecpf accept const pointer Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 14/16] net/mlx5: Rename peer_pf to host_pf Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 15/16] net/mlx5: Enable host PF HCA after eswitch is initialized Saeed Mahameed
2020-11-20 23:03 ` [PATCH mlx5-next 16/16] net/mlx5: Treat host PF vport as other (non eswitch manager) vport Saeed Mahameed
2020-11-30 18:42 ` [PATCH mlx5-next 00/16] mlx5 next updates 2020-11-20 Saeed Mahameed
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=20201124180210.GJ5487@ziepe.ca \
--to=jgg@nvidia.com \
--cc=eli@mellanox.com \
--cc=elic@nvidia.com \
--cc=kuba@kernel.org \
--cc=leonro@mellanox.com \
--cc=linux-rdma@vger.kernel.org \
--cc=maorg@nvidia.com \
--cc=mbloch@nvidia.com \
--cc=netdev@vger.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 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).