Thread (9 messages) flat view 9 messages, 4 authors, 2015-05-13

Re: [PATCH net-next v2] net: core: set and refine qdisc pkt len before tc_classify

From: Daniel Borkmann <daniel@iogearbox.net>
Date: 2015-05-13 21:36:58

On 05/13/2015 11:20 PM, Eric Dumazet wrote:
...
Is qdisc_pkt_len_init() still inlined, now it has 2 callers ?

People using policers on ingress are very often disabling GRO for
various historical reasons.
Yes, tested policer and for getting it working on my side, GRO needs
to be off.
Since handle_ing() is very uncommon, I would prefer not slowing down tx
just so that GRO ~5% error in ingress is fixed.
I see a callq dc0 <qdisc_pkt_len_init> now. Ok, in that case Florian's
original version is fine:

   http://patchwork.ozlabs.org/patch/471346/
I noticed you did not CC me, don't be afraid to include me next time one
of my commit is referred in a changelog ;)
Hmm, I did Cc you. ;)

To:	davem@davemloft.net
Cc:	fw@strlen.de, ast@plumgrid.com, jhs@mojatatu.com,
	eric.dumazet@gmail.com, netdev@vger.kernel.org,
	daniel@iogearbox.net

Thanks,
Daniel
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help