hostapd/doc/eap_server.doxygen
Jouni Malinen 5eb513c3ba doc: Disable Doxygen autolink support
The way autolink support is implementing in Doxygen is a bit
inconvenient with wpa_supplicant being recognized as something that
would always be linked to struct wpa_supplicant. In addition, number of
links were not really noticed automatically. To get this working more
robustly and without having to use the %wpa_supplicant workaround (which
had its own issues, e.g., with titles), disable autolinking and use
explicit \ref commands instead.

This is also updating some of the obsolete notes to point to correct
file names, etc. changes in the source code tree.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-01-03 15:44:35 +02:00

56 lines
2.8 KiB
Text

/**
\page eap_server_module EAP server implementation
Extensible Authentication Protocol (EAP) is an authentication framework
defined in RFC 3748. hostapd uses a separate code module for EAP server
implementation. This module was designed to use only a minimal set of
direct function calls (mainly, to debug/event functions) in order for
it to be usable in other programs. The design of the EAP
implementation is based loosely on RFC 4137. The state machine is
defined in this RFC and so is the interface between the server state
machine and methods. As such, this RFC provides useful information for
understanding the EAP server implementation in hostapd.
Some of the terminology used in EAP state machine is referring to
EAPOL (IEEE 802.1X), but there is no strict requirement on the lower
layer being IEEE 802.1X if EAP module is built for other programs than
wpa_supplicant. These terms should be understood to refer to the
lower layer as defined in RFC 4137.
\section adding_eap_methods Adding EAP methods
Each EAP method is implemented as a separate module, usually as one C
file named eap_server_<name of the method>.c, e.g., \ref eap_server_md5.c. All EAP
methods use the same interface between the server state machine and
method specific functions. This allows new EAP methods to be added
without modifying the core EAP state machine implementation.
New EAP methods need to be registered by adding them into the build
(Makefile) and the EAP method registration list in the
\ref eap_server_register_methods() function of \ref eap_server_methods.c. Each EAP
method should use a build-time configuration option, e.g., EAP_TLS, in
order to make it possible to select which of the methods are included
in the build.
EAP methods must implement the interface defined in \ref eap_i.h. struct
\ref eap_method defines the needed function pointers that each EAP method
must provide. In addition, the EAP type and name are registered using
this structure. This interface is based on section 4.4 of RFC 4137.
It is recommended that the EAP methods would use generic helper
functions, \ref eap_msg_alloc() and \ref eap_hdr_validate() when processing
messages. This allows code sharing and can avoid missing some of the
needed validation steps for received packets. In addition, these
functions make it easier to change between expanded and legacy EAP
header, if needed.
When adding an EAP method that uses a vendor specific EAP type
(Expanded Type as defined in RFC 3748, Chapter 5.7), the new method
must be registered by passing vendor id instead of EAP_VENDOR_IETF to
\ref eap_server_method_alloc(). These methods must not try to emulate
expanded types by registering a legacy EAP method for type 254. See
\ref eap_server_vendor_test.c for an example of an EAP method implementation that
is implemented as an expanded type.
*/