+ * lib/netdev-vport.c provides support for "virtual ports"
+ implemented by the Open vSwitch datapath module for the Linux
+ kernel. This may serve as a model for minimal netdev
+ implementations.
+
+ * lib/netdev-dummy.c is a fake netdev implementation useful only
+ for testing.
+
+
+Porting Strategies
+------------------
+
+After a netdev provider has been implemented for a system's network
+devices, you may choose among three basic porting strategies.
+
+The lowest-effort strategy is to use the "userspace switch"
+implementation built into Open vSwitch. This ought to work, without
+writing any more code, as long as the netdev provider that you
+implemented supports receiving packets. It yields poor performance,
+however, because every packet passes through the ovs-vswitchd process.
+See INSTALL.userspace for instructions on how to configure a userspace
+switch.
+
+If the userspace switch is not the right choice for your port, then
+you will have to write more code. You may implement either an
+"ofproto provider" or a "dpif provider". Which you should choose
+depends on a few different factors:
+
+ * Only an ofproto provider can take full advantage of hardware
+ with built-in support for wildcards (e.g. an ACL table or a
+ TCAM).
+
+ * A dpif provider can take advantage of the Open vSwitch built-in
+ implementations of bonding, LACP, 802.1ag, 802.1Q VLANs, and
+ other features. An ofproto provider has to provide its own
+ implementations, if the hardware can support them at all.
+
+ * A dpif provider is usually easier to implement, but most
+ appropriate for software switching. It "explodes" wildcard
+ rules into exact-match entries. This allows fast hash lookups
+ in software, but makes inefficient use of TCAMs in hardware
+ that support wildcarding.
+
+The following sections describe how to implement each kind of port.
+
+
+ofproto Providers
+-----------------
+
+An "ofproto provider" is what ofproto uses to directly monitor and
+control an OpenFlow-capable switch. struct ofproto_class, in
+ofproto/ofproto-provider.h, defines the interfaces to implement an
+ofproto provider for new hardware or software. That structure contains
+many function pointers, each of which has a comment that is meant to
+describe its behavior in detail. If the requirements are unclear,
+please report this as a bug.
+
+The ofproto provider interface is preliminary. Please let us know if
+it seems unsuitable for your purpose. We will try to improve it.
+
+
+Writing a dpif Provider
+-----------------------
+
+Open vSwitch has a built-in ofproto provider named "ofproto-dpif",
+which is built on top of a library for manipulating datapaths, called
+"dpif". A "datapath" is a simple flow table, one that supports only
+exact-match flows, that is, flows without wildcards. When a packet
+arrives on a network device, the datapath looks for it in this
+exact-match table. If there is a match, then it performs the
+associated actions. If there is no match, the datapath passes the
+packet up to ofproto-dpif, which maintains an OpenFlow flow table
+(that supports wildcards). If the packet matches in this flow table,
+then ofproto-dpif executes its actions and inserts a new exact-match
+entry into the dpif flow table. (Otherwise, ofproto-dpif passes the
+packet up to ofproto to send the packet to the OpenFlow controller, if
+one is configured.)
+
+The "dpif" library in turn delegates much of its functionality to a
+"dpif provider". The following diagram shows how dpif providers fit
+into the Open vSwitch architecture:
+
+ _
+ | +-------------------+
+ | | ovs-vswitchd |<-->ovsdb-server
+ | +-------------------+
+ | | ofproto |<-->OpenFlow controllers
+ | +--------+-+--------+ _
+ | | netdev | |ofproto-| |
+ userspace | +--------+ | dpif | |
+ | | netdev | +--------+ |
+ | |provider| | dpif | |
+ | +---||---+ +--------+ |
+ | || | dpif | | implementation of
+ | || |provider| | ofproto provider
+ |_ || +---||---+ |
+ || || |
+ _ +---||-----+---||---+ |
+ | | |datapath| |
+ kernel | | +--------+ _|
+ | | |
+ |_ +--------||---------+
+ ||
+ physical
+ NIC
+
+struct dpif_class, in lib/dpif-provider.h, defines the interfaces
+required to implement a dpif provider for new hardware or software.
+That structure contains many function pointers, each of which has a
+comment that is meant to describe its behavior in detail. If the
+requirements are unclear, please report this as a bug.
+
+There are two existing dpif implementations that may serve as
+useful examples during a port:
+
+ * lib/dpif-linux.c is a Linux-specific dpif implementation that
+ talks to an Open vSwitch-specific kernel module (whose sources
+ are in the "datapath" directory). The kernel module performs
+ all of the switching work, passing packets that do not match any
+ flow table entry up to userspace. This dpif implementation is
+ essentially a wrapper around calls into the kernel module.
+
+ * lib/dpif-netdev.c is a generic dpif implementation that performs
+ all switching internally. This is how the Open vSwitch
+ userspace switch is implemented.
+
+
+Miscellaneous Notes
+-------------------
+
+Open vSwitch source code uses uint16_t, uint32_t, and uint64_t as
+fixed-width types in host byte order, and ovs_be16, ovs_be32, and
+ovs_be64 as fixed-width types in network byte order. Each of the
+latter is equivalent to the one of the former, but the difference in
+name makes the intended use obvious.
+
+The default "fail-mode" for Open vSwitch bridges is "standalone",
+meaning that, when the OpenFlow controllers cannot be contacted, Open
+vSwitch acts as a regular MAC-learning switch. This works well in
+virtualization environments where there is normally just one uplink
+(either a single physical interface or a bond). In a more general
+environment, it can create loops. So, if you are porting to a
+general-purpose switch platform, you should consider changing the
+default "fail-mode" to "secure", which does not behave this way. See
+documentation for the "fail-mode" column in the Bridge table in
+ovs-vswitchd.conf.db(5) for more information.
+
+lib/entropy.c assumes that it can obtain high-quality random number
+seeds at startup by reading from /dev/urandom. You will need to
+modify it if this is not true on your platform.
+
+vswitchd/system-stats.c only knows how to obtain some statistics on
+Linux. Optionally you may implement them for your platform as well.
+
+
+Why OVS Does Not Support Hybrid Providers
+-----------------------------------------
+
+The "Porting Strategies" section above describes the "ofproto
+provider" and "dpif provider" porting strategies. Only an ofproto
+provider can take advantage of hardware TCAM support, and only a dpif
+provider can take advantage of the OVS built-in implementations of
+various features. It is therefore tempting to suggest a hybrid
+approach that shares the advantages of both strategies.
+
+However, Open vSwitch does not support a hybrid approach. Doing so
+may be possible, with a significant amount of extra development work,
+but it does not yet seem worthwhile, for the reasons explained below.
+
+First, user surprise is likely when a switch supports a feature only
+with a high performance penalty. For example, one user questioned why
+adding a particular OpenFlow action to a flow caused a 1,058x slowdown
+on a hardware OpenFlow implementation [1]. The action required the
+flow to be implemented in software.
+
+Given that implementing a flow in software on the slow management CPU
+of a hardware switch causes a major slowdown, software-implemented
+flows would only make sense for very low-volume traffic. But many of
+the features built into the OVS software switch implementation would
+need to apply to every flow to be useful. There is no value, for
+example, in applying bonding or 802.1Q VLAN support only to low-volume
+traffic.
+
+Besides supporting features of OpenFlow actions, a hybrid approach
+could also support forms of matching not supported by particular
+switching hardware, by sending all packets that might match a rule to
+software. But again this can cause an unacceptable slowdown by
+forcing bulk traffic through software in the hardware switch's slow
+management CPU. Consider, for example, a hardware switch that can
+match on the IPv6 Ethernet type but not on fields in IPv6 headers. An
+OpenFlow table that matched on the IPv6 Ethernet type would perform
+well, but adding a rule that matched only UDPv6 would force every IPv6
+packet to software, slowing down not just UDPv6 but all IPv6
+processing.
+
+[1] Aaron Rosen, "Modify packet fields extremely slow",
+ openflow-discuss mailing list, June 26, 2011, archived at
+ https://mailman.stanford.edu/pipermail/openflow-discuss/2011-June/002386.html.
+