From: Vlad Yasevich <vyasevic@redhat.com>
To: Jerry Chu <hkchu@google.com>, Herbert Xu <herbert@gondor.hengli.com.au>
Cc: Eric Dumazet <eric.dumazet@gmail.com>,
David Miller <davem@davemloft.net>,
Christoph Paasch <christoph.paasch@uclouvain.be>,
"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
Michael Dalton <mwdalton@google.com>
Subject: Re: [PATCH v2 net-next] net: introduce gro_frag_list_enable sysctl
Date: Wed, 30 Oct 2013 14:09:54 -0400 [thread overview]
Message-ID: <52714B72.2060200@redhat.com> (raw)
In-Reply-To: <CAPshTCg8j-y6fGfh0tRMir34-Lzy77qyzibvwjMSMpR0Rr1rsA@mail.gmail.com>
On 10/30/2013 01:39 PM, Jerry Chu wrote:
> On Tue, Oct 29, 2013 at 9:42 PM, Herbert Xu <herbert@gondor.apana.org.au> wrote:
>>
>> On Tue, Oct 29, 2013 at 09:34:41PM -0700, Eric Dumazet wrote:
>>>
>>> What matters ?
>>>
>>> GRO ?
>>
>> What matters is that you should not treat the forwarding case
>> separately from the host case.
>>
>> For virtualisation the host case looks exactly like the forwarding
>> case.
>
>
> Not sure I agree - there are two different "forwarding" cases - forwarding
> to another physical NIC (to go out to the wire hence need to do GSO),
> and (for virtualization) forwarding to a virtual NIC and consumed internally
> (e.g., VM).
I don't think you can really differentiate these 2 case. VM are
very commonly used as routers/forwarders. In some cases, to get
better throughput, VFs are assigned to the VMs as the externally
facing ports. So, you still end up forwarding to another physical
NIC.
-vlad
> For the latter we should strive to push GSO pkts all the way
> to the VM stack w/o breaking them up. So for virtualization GRO is all
> goodness but not sure about the regular forwarding path. (From the
> perf perspective it boils down to if the cost of GSO/GRO will offset
> the benefit of GRO. Sure if one manages to get the cost close to zero
> than there is not reason to leave GRO always on. But it's still a big if for
> now.)
>
> Best,
>
> Jerry
>
>>
>>
>> IOW, if having a 64KB packet matters for the host, then it matters
>> for forwarding as well.
>>
>>> Before my patch, GRO packets were 17 MSS, and nobody complained packets
>>> were too small, so what are you saying exactly ?
>>
>> I'm not criticsing your mega-GRO patch at all. That one is great
>> and means that we'll get aggregated packets up to 64K. What we need
>> to do is just to patch up the GSO code so that it can handle these
>> mega-packets properly.
>>
>> Cheers,
>> --
>> Email: Herbert Xu <herbert@gondor.apana.org.au>
>> Home Page: http://gondor.apana.org.au/~herbert/
>> PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
next prev parent reply other threads:[~2013-10-30 18:10 UTC|newest]
Thread overview: 163+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-10-28 11:55 Bug in skb_segment: fskb->len != len Christoph Paasch
2013-10-28 13:21 ` Eric Dumazet
2013-10-28 13:28 ` Christoph Paasch
2013-10-29 1:15 ` Eric Dumazet
2013-10-29 9:08 ` Christoph Paasch
2013-10-29 12:57 ` Eric Dumazet
2013-10-29 13:06 ` [PATCH net-next] net: introduce gro_frag_list_enable sysctl Eric Dumazet
2013-10-29 13:48 ` Christoph Paasch
2013-10-29 15:12 ` [PATCH v2 " Eric Dumazet
2013-10-29 23:44 ` David Miller
2013-10-30 0:06 ` Ben Hutchings
2013-11-02 14:01 ` [PATCH v3 net-next] net: introduce dev_set_forwarding() Eric Dumazet
2013-11-02 15:46 ` Ben Hutchings
2013-11-02 18:20 ` Eric Dumazet
2013-11-02 19:58 ` [PATCH v4 " Eric Dumazet
2013-11-03 17:18 ` Christoph Paasch
2013-11-04 16:55 ` Ben Hutchings
2013-11-07 21:17 ` David Miller
2013-11-07 21:31 ` Herbert Xu
2013-11-07 21:54 ` Eric Dumazet
2013-11-08 3:59 ` Herbert Xu
2013-11-08 4:25 ` Eric Dumazet
2013-11-10 14:05 ` Herbert Xu
2013-11-11 14:36 ` Herbert Xu
2013-11-07 22:06 ` David Miller
2013-11-08 2:17 ` Herbert Xu
2013-11-08 2:42 ` Eric Dumazet
2013-11-08 2:51 ` Eric Dumazet
2013-11-08 3:23 ` Herbert Xu
2013-11-08 4:21 ` Eric Dumazet
2013-11-08 4:24 ` Herbert Xu
2013-11-08 4:40 ` Eric Dumazet
2013-11-08 4:43 ` Herbert Xu
2013-11-08 5:08 ` Eric Dumazet
2013-11-08 5:21 ` Herbert Xu
2013-11-08 5:40 ` Eric Dumazet
2013-11-11 18:58 ` Herbert Xu
2013-11-08 3:22 ` Herbert Xu
2013-11-08 4:06 ` Eric Dumazet
2013-11-08 4:10 ` Herbert Xu
2013-11-08 4:24 ` Eric Dumazet
2013-11-08 4:28 ` Herbert Xu
2013-11-21 18:29 ` David Miller
2013-11-21 18:38 ` Eric Dumazet
2013-11-03 12:28 ` [PATCH v3 " Herbert Xu
2013-11-03 16:28 ` Eric Dumazet
2013-11-03 16:31 ` Herbert Xu
2013-11-03 17:26 ` Eric Dumazet
2013-11-04 4:11 ` Herbert Xu
2013-11-04 4:23 ` Eric Dumazet
2013-11-04 4:29 ` Herbert Xu
2013-11-04 5:00 ` Eric Dumazet
2013-11-04 5:23 ` Herbert Xu
2013-11-04 6:05 ` Eric Dumazet
2013-11-04 6:22 ` Herbert Xu
2013-11-04 6:26 ` Herbert Xu
2013-11-04 7:10 ` Eric Dumazet
2013-11-04 7:21 ` Herbert Xu
2013-11-04 13:58 ` Eric Dumazet
2013-11-04 6:46 ` Eric Dumazet
2013-11-04 7:03 ` Herbert Xu
2013-11-06 1:30 ` gso: Attempt to handle mega-GRO packets Herbert Xu
2013-11-06 1:45 ` Eric Dumazet
2013-11-06 4:07 ` Herbert Xu
2013-11-06 4:23 ` Eric Dumazet
2013-11-06 4:28 ` Herbert Xu
2013-11-06 5:20 ` Eric Dumazet
2013-11-06 8:04 ` Herbert Xu
2013-11-06 8:16 ` Herbert Xu
2013-11-06 13:12 ` Herbert Xu
2013-11-06 15:01 ` Eric Dumazet
2013-11-07 0:36 ` Herbert Xu
2013-11-07 1:03 ` Eric Dumazet
2013-11-07 1:47 ` Herbert Xu
2013-11-07 2:02 ` Eric Dumazet
2013-11-07 2:08 ` Eric Dumazet
2013-11-07 2:15 ` Herbert Xu
2013-11-07 2:37 ` Eric Dumazet
2013-11-07 2:41 ` Herbert Xu
2013-11-07 5:56 ` Michael S. Tsirkin
2013-11-07 7:07 ` Eric Dumazet
2013-11-07 2:52 ` Jason Wang
2013-11-06 15:05 ` Eric Dumazet
2013-11-07 0:39 ` Herbert Xu
2013-11-06 12:39 ` Herbert Xu
2013-11-06 13:30 ` Herbert Xu
2013-11-06 14:39 ` Herbert Xu
2013-11-06 15:06 ` Eric Dumazet
2013-11-06 17:25 ` Joe Perches
2013-11-06 19:47 ` Eric Dumazet
2013-11-07 0:15 ` Eric Dumazet
2013-11-07 0:47 ` Herbert Xu
2013-11-07 0:56 ` Eric Dumazet
2013-11-07 1:00 ` Herbert Xu
2013-11-07 1:08 ` Eric Dumazet
2013-11-07 1:13 ` Hannes Frederic Sowa
2013-11-07 1:21 ` Eric Dumazet
2013-11-07 1:34 ` Eric Dumazet
2013-11-07 2:03 ` Hannes Frederic Sowa
2013-11-07 3:05 ` Eric Dumazet
2013-11-07 6:59 ` Eric Dumazet
2013-11-07 0:43 ` Herbert Xu
2013-11-07 6:22 ` Herbert Xu
2013-11-07 7:03 ` [1/3] gso: Add to segs at end of loop in skb_segment Herbert Xu
2013-11-07 7:06 ` [2/3] gso: Handle new frag_list of frags GRO packets Herbert Xu
2013-11-07 7:08 ` [3/3] gso: Handle malicious GRO packets without crashing Herbert Xu
2013-11-07 18:18 ` Ben Hutchings
2013-11-07 19:13 ` Sergei Shtylyov
2013-11-11 18:55 ` Herbert Xu
2013-11-07 18:16 ` [2/3] gso: Handle new frag_list of frags GRO packets Ben Hutchings
2013-11-11 18:54 ` Herbert Xu
2013-11-11 18:52 ` Herbert Xu
2013-11-12 10:12 ` David Laight
2013-11-13 1:13 ` gso: " Eric Dumazet
2013-11-13 1:29 ` Herbert Xu
2013-11-13 2:14 ` Eric Dumazet
2013-11-13 2:17 ` Eric Dumazet
2013-11-13 2:22 ` Herbert Xu
2013-11-13 2:25 ` Herbert Xu
2013-11-13 2:45 ` Eric Dumazet
2013-11-13 14:26 ` Herbert Xu
2013-11-13 15:06 ` Eric Dumazet
2013-11-14 8:11 ` Herbert Xu
2013-11-15 4:37 ` Eric Dumazet
2013-11-13 2:31 ` Eric Dumazet
2013-11-07 7:11 ` gso: Attempt to handle mega-GRO packets Eric Dumazet
2013-11-07 7:15 ` Herbert Xu
2013-11-07 7:17 ` Herbert Xu
2013-11-07 7:31 ` Eric Dumazet
2013-11-07 7:33 ` Herbert Xu
2013-11-03 23:23 ` [PATCH v3 net-next] net: introduce dev_set_forwarding() David Miller
2013-10-30 0:53 ` [PATCH v2 net-next] net: introduce gro_frag_list_enable sysctl Eric Dumazet
2013-10-30 2:02 ` David Miller
2013-10-30 2:05 ` Herbert Xu
2013-10-30 2:13 ` Jerry Chu
2013-10-30 2:19 ` Herbert Xu
2013-10-30 2:34 ` David Miller
2013-10-30 2:33 ` David Miller
[not found] ` <44571383414236@web13j.yandex.ru>
2013-11-02 18:28 ` Eric Dumazet
2013-11-03 23:19 ` David Miller
2013-10-30 19:39 ` Ben Hutchings
2013-10-30 19:53 ` Eric Dumazet
2013-10-30 20:05 ` Ben Hutchings
2013-10-30 20:12 ` Eric Dumazet
2013-10-30 4:06 ` Eric Dumazet
2013-10-30 4:08 ` Herbert Xu
2013-10-30 4:09 ` Herbert Xu
2013-10-30 4:15 ` Jerry Chu
2013-10-30 4:16 ` Eric Dumazet
2013-10-30 4:19 ` Herbert Xu
2013-10-30 4:34 ` Eric Dumazet
2013-10-30 4:42 ` Herbert Xu
2013-10-30 17:39 ` Jerry Chu
2013-10-30 18:09 ` Vlad Yasevich [this message]
2013-10-30 19:12 ` David Miller
2013-10-30 0:03 ` Jerry Chu
2013-10-29 14:41 ` Bug in skb_segment: fskb->len != len Herbert Xu
2013-10-29 15:08 ` Eric Dumazet
2013-10-30 1:50 ` Herbert Xu
2013-10-30 4:03 ` Eric Dumazet
2013-10-30 4:06 ` Herbert Xu
2013-10-30 4:37 ` Eric Dumazet
2013-10-30 4:47 ` Herbert Xu
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=52714B72.2060200@redhat.com \
--to=vyasevic@redhat.com \
--cc=christoph.paasch@uclouvain.be \
--cc=davem@davemloft.net \
--cc=eric.dumazet@gmail.com \
--cc=herbert@gondor.hengli.com.au \
--cc=hkchu@google.com \
--cc=mwdalton@google.com \
--cc=netdev@vger.kernel.org \
/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).