From: Jamal Hadi Salim <jhs@mojatatu•com>
To: John Fastabend <john.fastabend@gmail•com>,
Tom Herbert <tom@herbertland•com>
Cc: Hannes Frederic Sowa <hannes@stressinduktion•org>,
"John W. Linville" <linville@tuxdriver•com>,
Jesse Gross <jesse@kernel•org>,
David Miller <davem@davemloft•net>,
Anjali Singhai Jain <anjali.singhai@intel•com>,
Linux Kernel Network Developers <netdev@vger•kernel.org>,
Kiran Patil <kiran.patil@intel•com>
Subject: Re: [PATCH v1 1/6] net: Generalize udp based tunnel offload
Date: Tue, 8 Dec 2015 09:23:39 -0500 [thread overview]
Message-ID: <5666E7EB.40709@mojatatu.com> (raw)
In-Reply-To: <566687DC.7040805@gmail.com>
On 15-12-08 02:33 AM, John Fastabend wrote:
> On 15-12-02 04:15 PM, Tom Herbert wrote:
>>>
>>> Just keying off the last statement there...
>>>
>>> I think BPF programs are going to be hard to translate into hardware
>>> for most devices. The problem is the BPF programs in general lack
>>> structure. A parse graph would be much more friendly for hardware or
>>> at minimum the BPF program would need to be a some sort of
>>> well-structured program so a driver could turn that into a parse graph.
>>>
>> This might be relevant:
>> http://richard.systems/research/pdf/IEEE_HPSR_BPF_OPENFLOW.pdf
>>
>
> Thanks Tom interesting read but they seem to argue for a BPF engine in
> hardware which I'm still not convinced is necessary and the numbers
> provided are for a 1Gbps link where 10Gpbs/100Gbps+ would be more
> valuable.
>
> I am still leaning towards a fully programmable parse graph and a set
> of basic actions push/pop/set/fwd/etc. This would be useful for other
> features not just checksum offloads. I guess it doesn't necessarily
> exclude also having 1s complement logic though.
;-> I feel a little vindicated with this discussion.
Of course you can implement hardware using BPF! I think there is an
opportunity for someone to build such hardware, if one is not in
progress of being built yet.
A BPF hardware implementation is just a very different approach;
instead of it being a series of TCAM table hardware implementation
(and/or other types of implementation which use DRAM etc), it becomes
CPU instructions. Surely one can cast the EBPF bytecode into an ASIC.
My disagreement with Tom is laying a stake that this is how hardware
features are to be exposed.
My disagreement with you is laying a stake in the ground that hardware
oughta be implemented using a series of Tubes^WTCAMS.
When i build a graphics card the API is not how the internal
implementation works. Everbody conforms to the same driver APIs.
Likewise, we have Linux APIs - and switchdev
is the right direction. Write your driver to switchdev interfaces.
Let a thousand flowers bloom.
BTW: It is bordering on the abstraction-ridiculous when I see the
P4 claims to use ebpf and then somehow translate to use
classifier/actions.
cheers,
jamal
next prev parent reply other threads:[~2015-12-08 14:23 UTC|newest]
Thread overview: 94+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-11-23 21:02 [PATCH 0/6] Generalize udp based tunnels and add geneve offload Anjali Singhai Jain
2015-11-23 21:02 ` [PATCH v1 1/6] net: Generalize udp based tunnel offload Anjali Singhai Jain
2015-11-23 20:57 ` kbuild test robot
2015-11-23 20:58 ` kbuild test robot
2015-11-23 21:53 ` Tom Herbert
2015-11-23 22:49 ` Jesse Gross
2015-11-24 0:32 ` Singhai, Anjali
2015-11-24 0:38 ` Tom Herbert
2015-11-24 1:11 ` Jesse Brandeburg
2015-11-24 17:32 ` Tom Herbert
2015-11-24 17:43 ` Hannes Frederic Sowa
2015-11-24 17:52 ` Tom Herbert
2015-11-24 18:16 ` Hannes Frederic Sowa
2015-11-24 18:37 ` David Miller
2015-11-24 18:42 ` Hannes Frederic Sowa
2015-11-24 18:43 ` Tom Herbert
2015-11-30 3:22 ` David Miller
2015-11-30 21:42 ` Singhai, Anjali
2015-11-30 21:48 ` Tom Herbert
2015-12-01 3:51 ` David Miller
2015-12-01 3:48 ` David Miller
2015-12-01 6:33 ` Alexander Duyck
2015-11-30 3:21 ` David Miller
2015-11-30 21:33 ` Singhai, Anjali
2015-12-01 0:25 ` Jesse Gross
2015-12-01 1:02 ` Tom Herbert
2015-12-01 1:28 ` Jesse Gross
2015-12-01 5:26 ` Tom Herbert
2015-12-01 15:44 ` John W. Linville
2015-12-01 15:49 ` Hannes Frederic Sowa
2015-12-01 16:08 ` John W. Linville
2015-12-02 0:40 ` Singhai, Anjali
2015-12-02 3:50 ` Tom Herbert
2015-12-02 16:35 ` Hannes Frederic Sowa
2015-12-02 19:15 ` Tom Herbert
2015-12-02 23:35 ` John Fastabend
2015-12-03 0:15 ` Tom Herbert
2015-12-08 7:33 ` John Fastabend
2015-12-08 14:23 ` Jamal Hadi Salim [this message]
2015-12-08 15:10 ` Jamal Hadi Salim
2015-12-09 1:40 ` Thomas Graf
2015-12-09 5:45 ` Alexei Starovoitov
2015-12-09 12:58 ` Thomas Graf
2015-12-09 17:38 ` Alexei Starovoitov
2015-12-09 20:03 ` David Miller
2015-12-09 22:03 ` Thomas Graf
2015-12-09 22:21 ` David Miller
2015-12-09 22:25 ` Thomas Graf
2015-12-03 2:08 ` Alexei Starovoitov
2015-12-03 15:59 ` Hannes Frederic Sowa
2015-12-03 16:35 ` Andreas Schultz
2015-12-03 16:43 ` Hannes Frederic Sowa
2015-12-04 18:28 ` Tom Herbert
2015-12-04 19:54 ` John Fastabend
2015-12-04 19:59 ` Hannes Frederic Sowa
2015-12-04 20:02 ` Hannes Frederic Sowa
2015-12-04 20:06 ` David Miller
2015-12-04 20:13 ` Tom Herbert
2015-12-04 21:37 ` David Miller
2015-12-04 20:26 ` Hannes Frederic Sowa
2015-12-04 20:43 ` Tom Herbert
2015-12-04 21:11 ` Hannes Frederic Sowa
2015-12-04 20:44 ` Jesse Gross
2015-12-04 22:44 ` Alexander Duyck
2015-12-05 0:53 ` Tom Herbert
2015-12-05 5:45 ` Alexander Duyck
2015-12-05 6:49 ` David Miller
2015-12-05 8:24 ` Alexander Duyck
2015-12-05 17:53 ` Tom Herbert
2015-12-05 19:34 ` Alexander Duyck
2015-12-05 18:03 ` David Miller
2015-12-05 19:34 ` Alexander Duyck
2015-12-05 22:27 ` David Miller
2015-12-06 2:13 ` Alexander Duyck
2015-12-06 16:31 ` Tom Herbert
2015-12-06 18:44 ` Alexander Duyck
2015-12-06 21:30 ` Tom Herbert
2015-12-07 1:20 ` Alexander Duyck
2015-12-07 3:02 ` David Ahern
2015-12-07 16:20 ` Jesse Gross
2015-12-05 4:50 ` David Miller
2015-12-05 6:50 ` Alexander Duyck
2015-11-24 5:41 ` Alexander Duyck
2015-11-30 16:35 ` Tom Herbert
2015-11-30 21:53 ` Singhai, Anjali
2015-12-01 3:52 ` David Miller
2015-11-23 21:02 ` [PATCH v1 2/6] net: Add a generic udp_offload_get_port function Anjali Singhai Jain
2015-11-24 6:08 ` Alexander Duyck
2015-11-24 6:37 ` Alexander Duyck
2015-11-24 19:35 ` Singhai, Anjali
2015-11-23 21:02 ` [PATCH v1 3/6] i40e: Generalize the flow for udp based tunnels Anjali Singhai Jain
2015-11-23 21:02 ` [PATCH v1 4/6] i40e: Remove CONFIG_I40E_VXLAN Anjali Singhai Jain
2015-11-23 21:02 ` [PATCH v1 5/6] net: Refactor udp_offload and add Geneve port offload support Anjali Singhai Jain
2015-11-23 21:02 ` [PATCH v1 6/6] i40e:Add geneve tunnel " Anjali Singhai Jain
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=5666E7EB.40709@mojatatu.com \
--to=jhs@mojatatu$(echo .)com \
--cc=anjali.singhai@intel$(echo .)com \
--cc=davem@davemloft$(echo .)net \
--cc=hannes@stressinduktion$(echo .)org \
--cc=jesse@kernel$(echo .)org \
--cc=john.fastabend@gmail$(echo .)com \
--cc=kiran.patil@intel$(echo .)com \
--cc=linville@tuxdriver$(echo .)com \
--cc=netdev@vger$(echo .)kernel.org \
--cc=tom@herbertland$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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