Dan Smith wrote: > OL> I figured that if we called a sock_create() for a type that isn't > OL> supported by checkpoint/restart, we will catch it down the code by > OL> testing sock->ops->restore. That way, I don't need to "teach" > OL> sock_file_restore() about socket types and protocols. > > Okay, yeah I suppose so. In that case, why even have the check? > certainly sock_create() checks that somewhere down the line, right? Hmm... true. I was distracted by the test in socket() syscalls. Will remove the test. Oren. > > OL> Do you see any reason why it should work ? The only difference > OL> (for afunix) between dgram and seqpacket is the requirement that > OL> the socket be connected - see unix_seqpacket_sendmsg(). > > Not for AF_UNIX, no. I just said I hadn't tested it :) > _______________________________________________ Containers mailing list Containers@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linux-foundation.org/mailman/listinfo/containers
