RE: BPF sk_lookup v5 - TCP SYN and UDP 0-len flood benchmarks

From: Jakub Sitnicki

Sent: 20 August 2020 11:30
Subject: Re: BPF sk_lookup v5 - TCP SYN and UDP 0-len flood benchmarks

On Tue, Aug 18, 2020 at 08:19 PM CEST, Alexei Starovoitov wrote:

quoted

On Tue, Aug 18, 2020 at 8:49 AM Jakub Sitnicki [off-list ref] wrote:

quoted

         :                      rcu_read_lock();
         :                      run_array = rcu_dereference(net-

bpf.run_array[NETNS_BPF_SK_LOOKUP]);

quoted

    0.01 :   ffffffff817f8624:       mov    0xd68(%r12),%rsi
         :                      if (run_array) {
    0.00 :   ffffffff817f862c:       test   %rsi,%rsi
    0.00 :   ffffffff817f862f:       je     ffffffff817f87a9 <__udp4_lib_lookup+0x2c9>
         :                      struct bpf_sk_lookup_kern ctx = {
    1.05 :   ffffffff817f8635:       xor    %eax,%eax
    0.00 :   ffffffff817f8637:       mov    $0x6,%ecx
    0.01 :   ffffffff817f863c:       movl   $0x110002,0x40(%rsp)
    0.00 :   ffffffff817f8644:       lea    0x48(%rsp),%rdi
   18.76 :   ffffffff817f8649:       rep stos %rax,%es:(%rdi)
    1.12 :   ffffffff817f864c:       mov    0xc(%rsp),%eax
    0.00 :   ffffffff817f8650:       mov    %ebp,0x48(%rsp)
    0.00 :   ffffffff817f8654:       mov    %eax,0x44(%rsp)
    0.00 :   ffffffff817f8658:       movzwl 0x10(%rsp),%eax
    1.21 :   ffffffff817f865d:       mov    %ax,0x60(%rsp)
    0.00 :   ffffffff817f8662:       movzwl 0x20(%rsp),%eax
    0.00 :   ffffffff817f8667:       mov    %ax,0x62(%rsp)
         :                      .sport          = sport,
         :                      .dport          = dport,
         :                      };

Such heavy hit to zero init 56-byte structure is surprising.
There are two 4-byte holes in this struct. You can try to pack it and
make sure that 'rep stoq' is used instead of 'rep stos' (8 byte at a time vs 4).

Thanks for the tip. I'll give it a try.

You probably don't want to use 'rep stos' in any of its forms.
The instruction 'setup' time is horrid on most cpu variants.
For a 48 byte structure six writes of a zero register will be faster.

If gcc is generating the 'rep stos' then the compiler source code for that
pessimisation needs deleting...

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)