Date: Thu, 22 Aug 2002 07:59:03 -0500 (CDT) From: mark tinguely <tinguely@web.cs.ndsu.nodak.edu> To: freebsd-atm@FreeBSD.ORG, vjardin@wanadoo.fr Subject: Re: BPF into the HARP stack Message-ID: <200208221259.g7MCx3I34803@web.cs.ndsu.nodak.edu> In-Reply-To: <20020821225810.D8DD615039F@venus.vincentjardin.net>
next in thread | previous in thread | raw e-mail | index | archive | help
on Thu, 22 Aug 2002 00:58:10 +0200, Vincent Jardin <vjardin@wanadoo.fr> said: > I am wondering where the best place could be to add a bpf_[m]tap support into > the HARP stack. > > For example, I think that the incoming mbufs could be tapped into atm_subr.c: > atm_intr() > > About's the outgoing mbufs, I think that they could be tapped into atm_if.c: > atm_ifoutput(). > > However, which DLT_XXX should I use ? I do not think that DLT_ATM_RFC1483, > nor DLT_ATM_CLIP are the right one because it is not useful for the ILMI, > ARP, ... cells. > > Moreover, which header shoud be appended in order to be compatible with > tcpdump and ethereal ? In my Nicstar driver, I put the hooks into the device driver, right after transmission, and as it recieved a new packet. In this way we are not HARP specific. I added a pseudo-header consisting of the following 5 octects: DIRECTION VPI VCI GFC Payload Type Cell Loss Priority 8bits 8bits 16 bits 4 bits 3 bits 1 bit 0x80 = transmit 0x00 = recieve New pcap rules have to be added to use the ATM packets with tcpdump and family. when I supplied the DMA address for the first buffer of the PDU, I left 8 octects, so I knew I had room to add the pseudo-header without needing a m_pullup(). --mark tinguely. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-atm" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200208221259.g7MCx3I34803>