Revision 1d567e19cc9810a9bd67b1ccab54b68d86dadb76 authored by Linus Torvalds on 16 November 2012, 22:10:15 UTC, committed by Linus Torvalds on 16 November 2012, 22:10:15 UTC
Pull networking updates from David Miller: 1) tx_filtered/ps_tx_buf queues need to be accessed with the SKB queue lock, from Arik Nemtsov. 2) Don't call 802.11 driver's filter configure method until it's actually open, from Felix Fietkau. 3) Use ieee80211_free_txskb otherwise we leak control information. From Johannes Berg. 4) Fix memory leak in bluetooth UUID removal,f rom Johan Hedberg. 5) The shift mask trick doesn't work properly when 'optname' is out of range in do_ip_setsockopt(). Use a straightforward switch statement instead, the compiler emits essentially the same code but without the missing range check. From Xi Wang. 6) Fix when we call tcp_replace_ts_recent() otherwise we can erroneously accept a too-high tsval. From Eric Dumazet. 7) VXLAN bug fixes, mostly to do with VLAN header length handling, from Alexander Duyck. 8) Missing return value initialization for IPV6_MINHOPCOUNT socket option handling. From Hannes Frederic. 9) Fix regression in tasklet handling in jme/ksz884x/xilinx drivers, from Xiaotian Feng. 10) At smsc911x driver init time, we don't know if the chip is in word swap mode or not. However we do need to wait for the control register's ready bit to be set before we program any other part of the chip. Adjust the wait loop to account for this. From Kamlakant Patel. 11) Revert erroneous MDIO bus unregister change to mdio-bitbang.c 12) Fix memory leak in /proc/net/sctp/, from Tommi Rantala. 13) tilegx driver registers IRQ with NULL name, oops, from Simon Marchi. 14) TCP metrics hash table kzalloc() based allocation can fail, back down to using vmalloc() if it does. From Eric Dumazet. 15) Fix packet steering out-of-order delivery regression, from Tom Herbert. * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (40 commits) net-rps: Fix brokeness causing OOO packets tcp: handle tcp_net_metrics_init() order-5 memory allocation failures batman-adv: process broadcast packets in BLA earlier batman-adv: don't add TEMP clients belonging to other backbone nodes batman-adv: correctly pass the client flag on tt_response batman-adv: fix tt_global_entries flags update tilegx: request_irq with a non-null device name net: correct check in dev_addr_del() tcp: fix retransmission in repair mode sctp: fix /proc/net/sctp/ memory leak Revert "drivers/net/phy/mdio-bitbang.c: Call mdiobus_unregister before mdiobus_free" net/smsc911x: Fix ready check in cases where WORD_SWAP is needed drivers/net: fix tasklet misuse issue ipv4/ip_vti.c: VTI fix post-decryption forwarding brcmfmac: fix typo in CONFIG_BRCMISCAN vxlan: Update hard_header_len based on lowerdev when instantiating VXLAN vxlan: fix a typo. ipv6: setsockopt(IPIPPROTO_IPV6, IPV6_MINHOPCOUNT) forgot to set return value doc/net: Fix typo in netdev-features.txt vxlan: Fix error that was resulting in VXLAN MTU size being 10 bytes too large ...
Kconfig.net
menu "UML Network Devices"
depends on NET
# UML virtual driver
config UML_NET
bool "Virtual network device"
help
While the User-Mode port cannot directly talk to any physical
hardware devices, this choice and the following transport options
provide one or more virtual network devices through which the UML
kernels can talk to each other, the host, and with the host's help,
machines on the outside world.
For more information, including explanations of the networking and
sample configurations, see
<http://user-mode-linux.sourceforge.net/old/networking.html>.
If you'd like to be able to enable networking in the User-Mode
linux environment, say Y; otherwise say N. Note that you must
enable at least one of the following transport options to actually
make use of UML networking.
config UML_NET_ETHERTAP
bool "Ethertap transport"
depends on UML_NET
help
The Ethertap User-Mode Linux network transport allows a single
running UML to exchange packets with its host over one of the
host's Ethertap devices, such as /dev/tap0. Additional running
UMLs can use additional Ethertap devices, one per running UML.
While the UML believes it's on a (multi-device, broadcast) virtual
Ethernet network, it's in fact communicating over a point-to-point
link with the host.
To use this, your host kernel must have support for Ethertap
devices. Also, if your host kernel is 2.4.x, it must have
CONFIG_NETLINK_DEV configured as Y or M.
For more information, see
<http://user-mode-linux.sourceforge.net/old/networking.html> That site
has examples of the UML command line to use to enable Ethertap
networking.
If you'd like to set up an IP network with the host and/or the
outside world, say Y to this, the Daemon Transport and/or the
Slip Transport. You'll need at least one of them, but may choose
more than one without conflict. If you don't need UML networking,
say N.
config UML_NET_TUNTAP
bool "TUN/TAP transport"
depends on UML_NET
help
The UML TUN/TAP network transport allows a UML instance to exchange
packets with the host over a TUN/TAP device. This option will only
work with a 2.4 host, unless you've applied the TUN/TAP patch to
your 2.2 host kernel.
To use this transport, your host kernel must have support for TUN/TAP
devices, either built-in or as a module.
config UML_NET_SLIP
bool "SLIP transport"
depends on UML_NET
help
The slip User-Mode Linux network transport allows a running UML to
network with its host over a point-to-point link. Unlike Ethertap,
which can carry any Ethernet frame (and hence even non-IP packets),
the slip transport can only carry IP packets.
To use this, your host must support slip devices.
For more information, see
<http://user-mode-linux.sourceforge.net/old/networking.html>.
has examples of the UML command line to use to enable slip
networking, and details of a few quirks with it.
The Ethertap Transport is preferred over slip because of its
limitations. If you prefer slip, however, say Y here. Otherwise
choose the Multicast transport (to network multiple UMLs on
multiple hosts), Ethertap (to network with the host and the
outside world), and/or the Daemon transport (to network multiple
UMLs on a single host). You may choose more than one without
conflict. If you don't need UML networking, say N.
config UML_NET_DAEMON
bool "Daemon transport"
depends on UML_NET
help
This User-Mode Linux network transport allows one or more running
UMLs on a single host to communicate with each other, but not to
the host.
To use this form of networking, you'll need to run the UML
networking daemon on the host.
For more information, see
<http://user-mode-linux.sourceforge.net/old/networking.html> That site
has examples of the UML command line to use to enable Daemon
networking.
If you'd like to set up a network with other UMLs on a single host,
say Y. If you need a network between UMLs on multiple physical
hosts, choose the Multicast Transport. To set up a network with
the host and/or other IP machines, say Y to the Ethertap or Slip
transports. You'll need at least one of them, but may choose
more than one without conflict. If you don't need UML networking,
say N.
config UML_NET_VDE
bool "VDE transport"
depends on UML_NET
help
This User-Mode Linux network transport allows one or more running
UMLs on a single host to communicate with each other and also
with the rest of the world using Virtual Distributed Ethernet,
an improved fork of uml_switch.
You must have libvdeplug installed in order to build the vde
transport into UML.
To use this form of networking, you will need to run vde_switch
on the host.
For more information, see <http://wiki.virtualsquare.org/>
That site has a good overview of what VDE is and also examples
of the UML command line to use to enable VDE networking.
If you need UML networking with VDE,
say Y.
config UML_NET_MCAST
bool "Multicast transport"
depends on UML_NET
help
This Multicast User-Mode Linux network transport allows multiple
UMLs (even ones running on different host machines!) to talk to
each other over a virtual ethernet network. However, it requires
at least one UML with one of the other transports to act as a
bridge if any of them need to be able to talk to their hosts or any
other IP machines.
To use this, your host kernel(s) must support IP Multicasting.
For more information, see
<http://user-mode-linux.sourceforge.net/old/networking.html> That site
has examples of the UML command line to use to enable Multicast
networking, and notes about the security of this approach.
If you need UMLs on multiple physical hosts to communicate as if
they shared an Ethernet network, say Y. If you need to communicate
with other IP machines, make sure you select one of the other
transports (possibly in addition to Multicast; they're not
exclusive). If you don't need to network UMLs say N to each of
the transports.
config UML_NET_PCAP
bool "pcap transport"
depends on UML_NET && EXPERIMENTAL
help
The pcap transport makes a pcap packet stream on the host look
like an ethernet device inside UML. This is useful for making
UML act as a network monitor for the host. You must have libcap
installed in order to build the pcap transport into UML.
For more information, see
<http://user-mode-linux.sourceforge.net/old/networking.html> That site
has examples of the UML command line to use to enable this option.
If you intend to use UML as a network monitor for the host, say
Y here. Otherwise, say N.
config UML_NET_SLIRP
bool "SLiRP transport"
depends on UML_NET
help
The SLiRP User-Mode Linux network transport allows a running UML
to network by invoking a program that can handle SLIP encapsulated
packets. This is commonly (but not limited to) the application
known as SLiRP, a program that can re-socket IP packets back onto
the host on which it is run. Only IP packets are supported,
unlike other network transports that can handle all Ethernet
frames. In general, slirp allows the UML the same IP connectivity
to the outside world that the host user is permitted, and unlike
other transports, SLiRP works without the need of root level
privleges, setuid binaries, or SLIP devices on the host. This
also means not every type of connection is possible, but most
situations can be accommodated with carefully crafted slirp
commands that can be passed along as part of the network device's
setup string. The effect of this transport on the UML is similar
that of a host behind a firewall that masquerades all network
connections passing through it (but is less secure).
To use this you should first have slirp compiled somewhere
accessible on the host, and have read its documentation. If you
don't need UML networking, say N.
Startup example: "eth0=slirp,FE:FD:01:02:03:04,/usr/local/bin/slirp"
endmenu
Computing file changes ...