Re: [PATCH] inet: dont set inet_rcv_saddr in connect()

[Eric Dumazet <eric.dumazet@gmail.com>]

Le mardi 07 septembre 2010 à 22:34 -0400, Brian Haley a écrit :

> Is this really the right thing to do?  Linux has been doing this forever,
> right?  Just like BSD has done it forever.  The way I've always "cleared"
> a local address is to set the address family to AF_UNSPEC on the next
> connect() call, as mentioned on the man page.  I just want to make sure
> we're not changing something just to work around a broken application,
> sendto()/sendmsg() work perfect in this case by not setting the local address.
> 
> BTW, it seems as though the reason this might only happen sometimes is
> that if the first connect() is to 127.0.0.1, you won't be able to then
> try and connect to say, 192.168.1.1.  If you first connect() to a remote
> address things will probably just work.


I believe we have the following choice :

1) connect(AF_UNIX) sets the remote address/port
   bind() sets the local port (and optionally address)
   connect(AF_UNSPEC) clears remote addess/port,
                      let local address/port unchanged

2) Correct UDP hashing, when local address changes from 0 to x.y.z.t
   (cost : two locks taken), and a possible packet drop during the
operation.

   Document that connect() also sets local address, and that before
doing a second connect() to change remote address, its mandatory to
first issue a connect(AF_UNSPEC) to clear local address (if not locked
by a prior bind() call)