netdev - PACKET_TX_RING: packet size is too long

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:	Fri, 9 Oct 2009 11:07:11 +0200
From:	Gabor Gombas <gombasg@...aki.hu>
To:	netdev@...r.kernel.org
Cc:	johann.baudy@...-log.net
Subject: PACKET_TX_RING: packet size is too long

Hi,

I have added PACKET_TX_RING support to ggaoed
(http://code.google.com/p/ggaoed) and it have worked nice when I have
tested it with a virtual ethernet pair (veth). However when testing it
on real hardware, I get:

Oct  8 17:30:52 storage2 kernel: [ 1083.714204] packet size is too long (8740 > 8472)

Kernel is 2.6.31 from Debian experimental. The socket is SOCK_RAW, the
MTU is 9000, I'm using TPACKET_V2 format, and the ring have been created
with the following parameters:

block_nr = 256, block_size = 65536, frame_nr = 1792, frame_size = 9056

Documentation/networking/packet_mmap.txt says that a frame for
PACKET_TX_RING consists of a struct tpacket[2]_hdr followed by the data
to send, so IMHO a frame size of 9056 should be enough to send 8740
bytes of data. However, net/packet/af_packet.c has in tpacket_snd():

	size_max = po->tx_ring.frame_size
		- sizeof(struct skb_shared_info)
		- po->tp_hdrlen
		- LL_ALLOCATED_SPACE(dev)
		- sizeof(struct sockaddr_ll);

which is much more than just the struct tpacket[2]_hdr. So which is
right? The code or the documentation? What is the official formula
to compute the required frame length for a SOCK_RAW socket, given the
MTU?

Gabor

-- 
     ---------------------------------------------------------
     MTA SZTAKI Computer and Automation Research Institute
                Hungarian Academy of Sciences
     ---------------------------------------------------------
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html