Re: Netperf TCP_RR(loopback) 10% regression in 2.6.24-rc6, comparing with 2.6.22

[Eric Dumazet <dada1@cosmosbay.com>]

Zhang, Yanmin a écrit :
> On Mon, 2008-01-21 at 22:22 -0800, David Miller wrote:
>> From: "Zhang, Yanmin" <yanmin_zhang@linux.intel.com>
>> Date: Tue, 22 Jan 2008 14:07:19 +0800
>>
>>> I am wondering if UDP stack in kernel has a bug.
>> If one server binds to INADDR_ANY with port N, then any other socket
>> can be bound to a specific IP address with port N.  When packets
>> come in destined for port N, the delivery will be prioritized
>> to whichever socket has the more specific and matching binding.
> What does 'more specific' mean here? I assume 127.0.0.1 should be
> prioritized before 0.0.0.0 which means packets should be queued to
> 127.0.0.1 firstly.

vi +278 net/ipv4/udp.c

                         int score = (sk->sk_family == PF_INET ? 1 : 0);
                         if (inet->rcv_saddr) {
                                 if (inet->rcv_saddr != daddr)
                                         continue;
                                 score+=2;
                         }
                         if (inet->daddr) {
                                 if (inet->daddr != saddr)
                                         continue;
                                 score+=2;
                         }
                         if (inet->dport) {
                                 if (inet->dport != sport)
                                         continue;
                                 score+=2;
                         }
                         if (sk->sk_bound_dev_if) {
                                 if (sk->sk_bound_dev_if != dif)
                                         continue;
                                 score+=2;
                         }

So in your case, socket bound to 127.0.0.1 should have a better score (+2) 
than other one, unless the other one got an >= score because of another match 
(rcv_saddr set or bounded to an interface)