From: Simon Horman <horms@kernel.org>
To: Jakub Kicinski <kuba@kernel.org>
Cc: Amit Cohen <amcohen@nvidia.com>,
Alexei Starovoitov <alexei.starovoitov@gmail.com>,
Petr Machata <petrm@nvidia.com>,
"David S. Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@google.com>,
Paolo Abeni <pabeni@redhat.com>,
Andrew Lunn <andrew+netdev@lunn.ch>,
Network Development <netdev@vger.kernel.org>,
Ido Schimmel <idosch@nvidia.com>,
Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
Jesper Dangaard Brouer <hawk@kernel.org>,
John Fastabend <john.fastabend@gmail.com>,
bpf <bpf@vger.kernel.org>, mlxsw <mlxsw@nvidia.com>
Subject: Re: [PATCH net-next 00/12] mlxsw: Preparations for XDP support
Date: Sat, 15 Feb 2025 14:02:52 +0000 [thread overview]
Message-ID: <20250215140252.GP1615191@kernel.org> (raw)
In-Reply-To: <20250205090958.278ffaff@kernel.org>
On Wed, Feb 05, 2025 at 09:09:58AM -0800, Jakub Kicinski wrote:
> On Tue, 4 Feb 2025 17:26:43 +0000 Amit Cohen wrote:
> > > > You're right, most of packets should be handled by HW, XDP is
> > > > mainly useful for telemetry.
> > >
> > > Why skb path is not enough?
> >
> > We get better packet rates using XDP, this can be useful to redirect
> > packets to a server for analysis for example.
>
> TBH I also feel a little ambivalent about adding advanced software
> features to mlxsw. You have a dummy device off which you hang the NAPIs,
> the page pools, and now the RXQ objects. That already works poorly with
> our APIs. How are you going to handle the XDP side? Program per port,
> I hope? But the basic fact remains that only fallback traffic goes thru
> the XDP program which is not the normal Linux model, routing is after
> XDP.
>
> On one hand it'd be great if upstream switch drivers could benefit from
> the advanced features. On the other the HW is clearly not capable of
> delivering in line with how NICs work, so we're signing up for a stream
> of corner cases, bugs and incompatibility. Dunno.
FWIIW, I do think that as this driver is actively maintained by the vendor,
and this is a grey zone, it is reasonable to allow the vendor to decide if
they want the burden of this complexity to gain some performance.
next prev parent reply other threads:[~2025-02-15 14:02 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-04 11:04 [PATCH net-next 00/12] mlxsw: Preparations for XDP support Petr Machata
2025-02-04 11:04 ` [PATCH net-next 01/12] mlxsw: core: Remove debug prints Petr Machata
2025-02-04 11:04 ` [PATCH net-next 02/12] mlxsw: Check Rx local port in PCI code Petr Machata
2025-02-04 11:04 ` [PATCH net-next 03/12] mlxsw: Add struct mlxsw_pci_rx_pkt_info Petr Machata
2025-02-04 11:04 ` [PATCH net-next 04/12] mlxsw: pci: Use mlxsw_pci_rx_pkt_info Petr Machata
2025-02-04 11:05 ` [PATCH net-next 05/12] mlxsw: pci: Add a separate function for syncing buffers for CPU Petr Machata
2025-02-04 11:05 ` [PATCH net-next 06/12] mlxsw: pci: Store maximum number of ports Petr Machata
2025-02-04 11:05 ` [PATCH net-next 07/12] mlxsw: pci: Add PCI ports array Petr Machata
2025-02-04 11:05 ` [PATCH net-next 08/12] mlxsw: Add APIs to init/fini PCI port Petr Machata
2025-02-04 11:05 ` [PATCH net-next 09/12] mlxsw: pci: Initialize XDP Rx queue info per RDQ Petr Machata
2025-02-04 11:05 ` [PATCH net-next 10/12] mlxsw: spectrum: Initialize PCI port with the relevant netdevice Petr Machata
2025-02-04 11:05 ` [PATCH net-next 11/12] mlxsw: Set some SKB fields in bus driver Petr Machata
2025-02-04 11:05 ` [PATCH net-next 12/12] mlxsw: Validate local port from CQE in PCI code Petr Machata
2025-02-04 15:56 ` [PATCH net-next 00/12] mlxsw: Preparations for XDP support Alexei Starovoitov
2025-02-04 15:59 ` Amit Cohen
2025-02-04 16:02 ` Alexei Starovoitov
2025-02-04 17:26 ` Amit Cohen
2025-02-05 17:09 ` Jakub Kicinski
2025-02-15 14:02 ` Simon Horman [this message]
2025-02-15 16:10 ` Jakub Kicinski
2025-02-16 9:26 ` Simon Horman
2025-02-17 9:35 ` Ido Schimmel
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=20250215140252.GP1615191@kernel.org \
--to=horms@kernel.org \
--cc=alexei.starovoitov@gmail.com \
--cc=amcohen@nvidia.com \
--cc=andrew+netdev@lunn.ch \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=hawk@kernel.org \
--cc=idosch@nvidia.com \
--cc=john.fastabend@gmail.com \
--cc=kuba@kernel.org \
--cc=mlxsw@nvidia.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=petrm@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).