Quoting r. Sean Hefty <[EMAIL PROTECTED]>: > Subject: Re: [openib-general] [PATCH] ib_addr: local/loopback address handling > > Michael S. Tsirkin wrote: > >>I'm not as sure about this change; specifically the use of LOOPBACK(). > >>If the user specifies 127.0.0.1 for their source address, this will end > >>up changing that address to whatever the destination address is. Is this > >>really what we want? > > > >My testing shows that in sockets 127.0.0.1 is same as 0. > > Maybe I wasn't being clear. If the user tries to connect from 127.0.0.1 to > 192.168.0.101, this patch will change their source address from 127.0.0.1 > to be 192.168.0.101. The current code leaves the source address as > 127.0.0.1.
And then the connection will fail since 127.0.0.1 is not an IB device. So why is this behaviour useful? > What ib_addr is trying to allow is connecting to a remote IP address. > If a source address is not specified, then one is assigned. Right, so it has to have some routing logic. If this code is not in ib_addr then the same logic will have to be placed in all ULPs. That's not too bad, but it just seems a waste. -- Michael S. Tsirkin Staff Engineer, Mellanox Technologies _______________________________________________ openib-general mailing list [email protected] http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
