Re: raw socket, IP_HDRINCL byteorder

Bill Fenner <[email protected]> Thu, 26 Sep 2002 08:32:25 -0700

Newsgroups	gmane.os.bsd.api.general
Message-ID	<[email protected]>

>sensible, though if you wish to prevent a program from sending ip
>fragments, would it not make sense to prevent it from receiving them
>as well, and reassemble packets before passing them up?

In fact, it'd be awfully hard to get ip_input() to supply a fragment.
If we're not proposing massive changes to the stack, then raw input
only returns reassembled packets, presumably with ip_off and MF = 0.

  Bill

---------------------------------------------------------------------
The BSD APIs Discussion Mailing List
To unsubscribe:
send "unsubscribe bsd-api-discuss" to majordomo-DZEk9q7Sfju/3pe1ocb+swC/[email protected]