site_archiver@lists.apple.com Delivered-To: macnetworkprog@lists.apple.com On Jan 30, 2008, at 2:57 AM, Hostile Fork wrote: Kind Regards, - Peter Sichel Sustainable Softworks http://www.sustworks.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Macnetworkprog mailing list (Macnetworkprog@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/macnetworkprog/site_archiver%40lists.... I am porting a WinSock DHCP-like server to OS/X, and have run into a wall. I've written a Mac OS X DHCP Server that has been shipping for some time. After using packet sniffers I've found a smoking gun that OS/X may be ignoring broadcast UDP packets that are sent from the serviced network to 255.255.255.255 to the serviced port #8567. (FYI: This machine has two network cards, and the devices that are making DHCP requests are connected on a network through the second card which has a static IP address... the first card is used for plain ol' internet access.) It's a little complicated, but I'll point out that send()s to that static IP subnet are seeming to work and get a reaction, so the connection is at least somewhat "alive". I vaguely recall running into some issues in the BSD stack around which network interface(s) will receive broadcast packets that caused me to use libpcap to receive the packets at the data link layer. Compared to other network stacks (like Mentat/TCP) which just work, broadcast is kind of broken in the BSD stack. It's common practice to use libpcap for receive and bpftransmit for send when implementing DHCP servers or clients. As I remember, in Mentat/TCP broadcast packets were always duplicated and sent upstream or downstream to every broadcast enabled interface. In BSD, the packet is not automatically duplicated. The first broadcast enabled interface receives inbound broadcasts, and sending requires that you use the subnet broadcast address to determine which interface will transmit the packet. The critical flaw in this is that computers with one or more dynamically configured interfaces have become so common that the behavior is unpredictable in the field. This email sent to site_archiver@lists.apple.com