Re: [PATCH] net: improve ipv4 performances

2018-04-04 Thread Paolo Abeni
On Sun, 2018-04-01 at 20:31 +0200, Anton Gary Ceph wrote: > After a few profiling and analysis, turned out that the ethertype field > of the packets has the following distribution: [...] > 0.6% don't know/no opinion Am I the only one finding the submission date and the above info suspicious ?

Re: [PATCH] net: improve ipv4 performances

2018-04-03 Thread Douglas Caetano dos Santos
Hi Anton, everyone, On 04/01/18 15:31, Anton Gary Ceph wrote: > As the Linux networking stack is growing, more and more protocols are > added, increasing the complexity of stack itself. > Modern processors, contrary to common belief, are very bad in branch > prediction, so it's our task to give hi

Re: [PATCH] net: improve ipv4 performances

2018-04-02 Thread kbuild test robot
Hi Anton, Thank you for the patch! Perhaps something to improve: [auto build test WARNING on net/master] [also build test WARNING on v4.16 next-20180329] [if your patch is applied to the wrong git tree, please drop us a note to help improve the system] url: https://github.com/0day-ci/linux/

Re: [PATCH] net: improve ipv4 performances

2018-04-01 Thread Eric Dumazet
On 04/01/2018 11:31 AM, Anton Gary Ceph wrote: > As the Linux networking stack is growing, more and more protocols are > added, increasing the complexity of stack itself. > Modern processors, contrary to common belief, are very bad in branch > prediction, so it's our task to give hints to the com

Re: [PATCH] net: improve ipv4 performances

2018-04-01 Thread Md. Islam
Yes, I'm also seeing good performance improvement after adding likely() and prefetch(). On Sun, Apr 1, 2018 at 2:50 PM, Stephen Hemminger wrote: > On Sun, 1 Apr 2018 20:31:21 +0200 > Anton Gary Ceph wrote: > >> As the Linux networking stack is growing, more and more protocols are >> added, incr

Re: [PATCH] net: improve ipv4 performances

2018-04-01 Thread Stephen Hemminger
On Sun, 1 Apr 2018 20:31:21 +0200 Anton Gary Ceph wrote: > As the Linux networking stack is growing, more and more protocols are > added, increasing the complexity of stack itself. > Modern processors, contrary to common belief, are very bad in branch > prediction, so it's our task to give hints

[PATCH] net: improve ipv4 performances

2018-04-01 Thread Anton Gary Ceph
As the Linux networking stack is growing, more and more protocols are added, increasing the complexity of stack itself. Modern processors, contrary to common belief, are very bad in branch prediction, so it's our task to give hints to the compiler when possible. After a few profiling and analysis,