summaryrefslogtreecommitdiff
path: root/bpf_insns.h
AgeCommit message (Collapse)AuthorFilesLines
2013-06-15bpf: use Linux' define of BPF_MAXINSNSDaniel Borkmann1-0/+4
Do not necessarily define our own, only in case the Linux one is not available. Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
2013-06-04bpf: split up instructions and extensions from headerDaniel Borkmann1-0/+58
Move them into a separate header file, so that it is more clean. Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
o-mapping.txt?id=4e28e05aac78dca759c4b2b374ec99306f7ae8ae'>patch) treed13b7d82d6263784654faeaf60882e68460d5797 /Documentation/io-mapping.txt parent93d8be41c619ea6d4b29a1cefd15b60d2d82e764 (diff)
net: packet: Introduce PACKET_XMIT_RECEIVE socket optionpacket-loop-back
The PACKET_XMIT_RECEIVE socket options allows to immediately pump packets sent through PF_PACKET back into the network stack via dev_forward_skb() instead of submitting them to the device driver using dev_queue_xmit(). This feature might be useful to load the network stack with a lot of packets, similar to pktgen. Enable the setting using: int one = 1; setsockopt(fd, SOL_PACKET, PACKET_XMIT_RECEIVE, &one, sizeof(one)); By default, PACKET_XMIT_RECEIVE is disabled and needs to be explicitly enabled on PF_PACKET sockets. Joint work with Daniel Borkmann. Suggested-by: Daniel Borkmann <daniel@iogearbox.net> Cc: Willem de Bruijn <willemdebruijn.kernel@gmail.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Diffstat (limited to 'Documentation/io-mapping.txt')