[959] in SIPB_Linux_Development
Re: probs still with Linux/Kerberos
daemon@ATHENA.MIT.EDU (Bill Sommerfeld)
Thu Mar 2 09:28:05 1995
To: "Theodore Ts'o" <tytso@MIT.EDU>
Cc: "Charles M. Hannum" <mycroft@ai.mit.edu>, ghudson@MIT.EDU,
linux-dev@MIT.EDU, netbsd-dev@MIT.EDU
In-Reply-To: Your message of "Wed, 1 Mar 1995 10:59:05 +0500 ."
<9503011559.AA29113@dcl.MIT.EDU>
Date: Thu, 02 Mar 1995 09:14:49 -0500
From: Bill Sommerfeld <sommerfeld@orchard.medford.ma.us>
> I suspect Linux may be doing this even when the UDP socket isn't
> connected.
>
> That sounds right. Off hand, can anyone think of any potentially bad
> effects of this behavior? If so, I'll submit it as a bug to the Linux
> networking developers.
There are two possible problems with doing this:
1) There may be problems related to cases where multiple sockets are
bound to a UDP port... (which it *appears* you can do if all the
sockets set the SO_REUSEADDR option before binding; at least there's
code in udp_usrreq.c to support that case); in that case, it's a
matter of "how do you know what the right one to get the error is".
2) There's no way for the kernel to tell user mode *where* the ICMP
error came from, so in the unconnected case, it could just be a ICMP
error from Mars unconnected to anything else going on on the host.
If you *really* want that info, open up an ICMP socket and do the
sorting/sanity checking yourself.
- Bill
