This functionality fits better with src/tls (i.e., internal TLS
implementation), so move it there to make crypto_internal.c more
of a wrapper like other crypto_*.c files.
Private keys can now be used in either unencrypted or encrypted
PKCS #8 encoding. Only the pbeWithMD5AndDES-CBC algorithm (PKCS #5)
is currently supported.
Recognize the PEM header "BEGIN PRIVATE KEY" as base64-decode the data
to be able to use PEM encoded, unencrypted PKCS#8 private keys with the
internal TLS implementation. Previously, only DER encoding of the
PKCS#8 private key was supported.
OpenSSL 0.9.7 does not include get_rfc3526_prime_1536() function, so
provide that functionality internally if needed. In addition, make
sha256_vector() building depend on whether SHA256 support is included
in the OpenSSL library. This with CONFIG_INTERNAL_SHA256=y in .config
allows OpenSSL without SHA256 support to be used.
Both hostapd/wpa_supplicant compilation fails on DragonFly BSD.
This patch solves this issue.
I have tested only compilation. Not functionality.
Because I don't have any device which can work on DragonFly BSD.
mac80211 does not currently allow (re)authentication when we are already
authenticated. In order to work around this, force deauthentication if
nl80211 authentication command fails with EALREADY. Unfortunately, the
workaround code in driver_nl80211.c alone is not enough since the
following disconnection event would clear wpa_supplicant authentication
state. To handle this, add some code to restore authentication state
when using userspace SME.
This workaround will hopefully become unnecessary in some point should
mac80211 start accepting new authentication requests even when in
authenticated state.
Reorganize the TLS/crypto library segments into a single set of blocks
for each library instead of multiple locations handling library-specific
operations. Group crypto functionality together and get wpa_supplicant
and hostapd Makefile closer to eachother in order to make it easier to
eventually move this into a shared makefile.
Crypto library wrappers can now override the internal DH (group 5)
implementation. As a starting point, this is done with OpenSSL. The
new mechanism is currently available only for WPS (i.e., IKEv2 still
depends on the internal DH implementation).
Replace "nl80211: Ignored unknown event (cmd=33)" with
"nl80211: Scan trigger" to make debug output clearer. We do not
currently do anything with this event apart from showing it in
the debug log.
This allows NSS to be used to derive EAP-TLS/PEAP/TTLS keying material.
NSS requires a patch from
https://bugzilla.mozilla.org/show_bug.cgi?id=507359
to provide the new API. In addition, that patch needs to be modified to
add the 16-bit context length value in SSL_ExportKeyingMaterial() only if
contextlen != 0 in order to match with the EAP-TLS/PEAP/TTLS use cases.
This issue seems to be coming from the unfortunate incompatibility in
draft-ietf-tls-extractor-07.txt (draft-ietf-tls-extractor-00.txt would
have used compatible PRF construction).
At this point, it is unclear how this will be resolved eventually, but
anyway, this shows a mechanism that can be used to implement EAP key
derivation with NSS with a small patch to NSS.
This brings in the first step in adding support for using NSS
(Mozilla Network Security Services) as the crypto and TLS library
with wpa_supplicant. This version is able to run through EAP-PEAP
and EAP-TTLS authentication, but does not yet implement any
certificate/private key configuration. In addition, this does not
implement proper key fetching functions either, so the end result
is not really of much use in real world yet.
On DragonFly BSD, wired IEEE 802.1X fails with this message:
ioctl[SIOC{ADD/DEL}MULTI]: Invalid argument
This patch solves this issue.
I have tested with these:
OS : DragonFly BSD 2.4.0
EAP : EAP-TLS
Switch : Cisco Catalyst 2950
This reverts commit 5aa9cb5cca.
The nested key attribute is using different attribute values
(NL80211_KEY_* vs. NL80211_ATTR_KEY_*), so cannot share the same routine
for these purposes..
Current wpa_supplicant has a bug with WEP keys, it adds a zero-length
sequence counter field to netlink which the kernel doesn't accept.
Additionally, the kernel API slightly changed to accept keys only when
connected, so we need to send it the keys after that. For that to work
with shared key authentication, we also include the default WEP TX key
in the authentication command.
To upload the keys properly _after_ associating, add a new flag
WPA_DRIVER_FLAGS_SET_KEYS_AFTER_ASSOC_DONE indicating that the driver
needs the keys at that point and not earlier.
Commit 0b55b934ee broke this by not
initializing drv->ap = 1 in hostapd case since the mode updating
code ended up unlinking the socket file. Setting drv->ap = 1
removes the mode change and as such, unlinking of the socket file.
It is simpler to just build in all the test driver code regardless
of whether this is for hostapd or wpa_supplicant (which will eventually
get AP mode support with driver_test, too).
radius_server_encapsulate_eap() resets sess->eap->if->eap{Success,Fail}
to FALSE, such that the completion condition is never true.
The net effect is that completed sessions would linger for
RADIUS_SESSION_TIMEOUT seconds.
Signed-off-by: Alex Badea <vamposdecampos@gmail.com>
Previously, the default settings allowed 100 sessions in 60 seconds.
With this fix, the default limit is now 100 sessions per 10 seconds.
[Bug 329]
Store a copy of device attributes during WPS protocol run and make it
available for external programs via the control interface STA MIB
command for associated stations. This gives access to device name and
type which can be useful when showing user information about associated
stations.
wpa_supplicant can now reconfigure the AP by acting as an External
Registrar with the wps_reg command. Previously, this was only used
to fetch the current AP settings, but now the wps_reg command has
optional arguments which can be used to provide the new AP
configuration. When the new parameters are set, the WPS protocol run
is allowed to continue through M8 to reconfigure the AP instead of
stopping at M7.
This is an initial step in adding support for the new connect command.
For now, we just add the capability query. The actual use of the new
command will be added separately.
This is a patch for OpenBSD wired IEEE 802.1X. This is only for wired,
not wireless, because OpenBSD uses wpa_supplicant only on wired now.
http://www.openbsd.org/cgi-bin/cvsweb/ports/security/wpa_supplicant/
I have tested with these.
OS : OpenBSD 4.5
EAP : EAP-TLS
Switch : CentreCOM 8724SL
Previously, both the command replies and unsolicited events were
received from the same socket. This could cause problems if an event
message is received between a command and the response to that command.
Using two sockets avoids this issue.
When Linux has Path MTU discovery enabled, it sets by default the DF bit
on all outgoing datagrams, also UDP ones. If a RADIUS message is bigger
than the smallest MTU size to the target, it will be discarded.
This effectively limits RADIUS messages to ~ 1500 Bytes, while they can
be up to 4k according to RFC2865. In practice, this can mean trouble
when doing EAP-TLS with many RADIUS attributes besides the EAP-Message.
[Bug 326]
When Linux has Path MTU discovery enabled, it sets by default the DF bit
on all outgoing datagrams, also UDP ones. If a RADIUS message is bigger
than the smallest MTU size to the target, it will be discarded.
This effectively limits RADIUS messages to ~ 1500 Bytes, while they can
be up to 4k according to RFC2865. In practice, this can mean trouble
when doing EAP-TLS with many RADIUS attributes besides the EAP-Message.
[Bug 326]
These could, at least in theory, be used to generate unexpected common
name or subject alternative name matches should a CA sign strings with
NUL (C string termination) in them. For now, just reject the certificate
if an embedded NUL is detected. In theory, all the comparison routines
could be made to compare these strings as binary blobs (with additional
X.509 rules to handle some exceptions) and display NUL characters
somehow. Anyway, just rejecting the certificate will get rid of
potential problems with the C string getting terminated and it should
not really be used in certificates, so this should not break valid use
cases.
The BLOCK_SIZE define can be made more specific by using AES_ prefix and
by moving it to aes.h. After this, most aes-*.c do not really need to
include anything from the internal aes_i.h header file. In other words,
aes_i.h can now be used only for the code that uses the internal AES
block operation implementation and none of the code that can use AES
implementation from an external library do not need to include this
header file.
Better not specify EVP_CIPHER again for the second init call since that
will override key length with the default value. The previous version
was likely to work since most use cases would be likely to use the
default key length. Anyway, better make this handle variable length
ciphers (mainly, RC4), too, just in case it is needed in the future.
This is not really of that much use since rc4_skip() can be used as
easily. In addition, rc4 has caused some symbol conflicts in the past,
so it is easier to live without that as an exported symbol.
wpa_supplicant can now be built with FIPS capable OpenSSL for FIPS mode
operation. Currently, this is only enabling the FIPS mode in OpenSSL
without providing any higher level enforcement in wpa_supplicant.
Consequently, invalid configuration will fail during the authentication
run. Proper configuration (e.g., WPA2-Enterprise with EAP-TLS) allows
the connection to be completed.
Instead of using low level, digest-specific functions, use the generic
EVP interface for digest functions. In addition, report OpenSSL errors
in more detail.
Some crypto libraries can return in these functions (e.g., if a specific
hash function is disabled), so we better provide the caller a chance to
check whether the call failed. The return values are not yet used
anywhere, but they will be needed for future changes.
On NetBSD 5.0, when I use wired 802.1X, "Invalid argument" occurs
on SIOCADDMULTI ioctl and 802.1X fails.
I tried FreeBSD code, but "Address family not supported by protocol family"
occurs on SIOCADDMULTI ioctl and 802.1X fails, too.
This patch solves this issue.
I have tested with these:
OS : NetBSD 5.0
EAP : EAP-MD5
Switch : CentreCOM 8724SL
This is a (hopefully) temporary workaround to allow the same source code
tree to be used for building hostapd and wpa_supplicant without having
to manually force recompilation of some files. Currently, some of the
driver wrapper files need to be built separately for hostapd and
wpa_supplicant (#ifdef's in the files based on AP functionality).
This is somewhat racy as far as parallel make execution is concerned,
i.e., it may be necessary to run "make -j#" twice (plain "make" works
fine. Since this is supposed to be a temporary workaround, there is not
much point in trying to fix this with any more complex make processing.
Instead of having all driver stuff collected across wpa_supplicant
and hostapd, create a common snippet that they both include and
that handles the build configuration.
If base64_encode() were to be used with a huge data array, the
previous version could have resulted in overwriting the allocated
buffer due to an integer overflow as pointed out in
http://www.freebsd.org/cgi/query-pr.cgi?pr=137484. However, there
are no know use cases in hostapd or wpa_supplicant that would do that.
Anyway, the recommended change looks reasonable and provides additional
protection should the base64_encode() function be used for something
else in the future.
Current wpa_supplicant has a bug with WEP keys, it adds a zero-length
sequence counter field to netlink which the kernel doesn't accept.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
hostapd currently tries to encrypt all auth frames,
except for "OPEN" and "SHARED with transaction 3".
This means that it will send an encrypted "unknown
auth algorithm" reply for all other algorithsm. Fix
this by changing the logic to only encrypt shared
key auth frames with transaction 3.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Change existing CONFIG_LIBNL20 compatibility code in
driver_nl80211.c to be used by both wpa_supplicant
and hostapd, but take care of nl_handle too now.
Propagate CONFIG_LIBNL20 out of .config file and onto
CFLAGS in the Makefile.
Use libnl-gen now too.
Signed-off-by: Jon Loeliger <jdl@bigfootnetworks.com>
---
This removes need for local configuration to ignore *.o and *~
and allows the src/*/.gitignore files to be removed (subdirectories
will inherit the rules from the root .gitignore).
This attribute is not supposed to be used in the response frame (i.e.,
it is only in the EAP-Request/SIM-Notification frame) per RFC 4186
chapters 10.1 and 9.9. This is a minor bug since the server is required
to ignore the contents of the EAP-Response/SIM-Notification during
protected result indication per chapter 6.2.
EAP-AKA peer was already following the similar specification in RFC 4187,
but this was somehow missed in the EAP-SIM peer implementation.
The driver wrappers should not need to include wps_hostapd.h, so let's
make this easier by introducing a driver callback for reporting Probe
Request frames.
Since we do not currently support changing the AP settings received
from M7, there is no point in actually sending out the M8 that would
likely trigger the AP to reconfigure itself and potentially reboot.
For now, we just receive the AP settings in M7 and add a local network
configuration block based on those, but NACK the message. This makes
wps_reg work like wps_pin, but by using the AP PIN instead of a client
PIN.
Old way does not work with all drivers on NetBSD and FreeBSD are
also using this so should be a safe change. [Bug 312]
Signed-off-by: Roy Marples <roy@marples.name>
Add generic functions to get/set 80211 vars, set 80211 params and
get/sid ssid.
Change NetBSD defines to match the ioctl used for portability.
Check size we're copying into instead of assuming IFNAMSIZ.
Signed-off-by: Roy Marples <roy@marples.name>
If hostapd segfaults, or is killed with -9, or the interface already exists,
when the interface is created, it will fail.
Configuration file: /tmp/hostapd/hostapd.conf
Failed to create interface mon.wlan0_0.
Using interface wlan0_0 with hwaddr 00:13:01:01:08:0a and ssid 'IG_0405_LAN'
Failed to set beacon head/tail or DTIM period
Failed to create interface wlan0_1.
Try to remove the interface and re-create it before aborting.
To ensure the supplicant starts and ends with a clean slate (keys are
already cleaned up at init and deinit time), force a null BSSID and
bogus SSID to ensure the driver isn't connected to anything.
Signed-off-by: Dan Williams <dcbw@redhat.com>
4853d5ac84 had a small bug in the order
of these function calls in _wext_deauthenticate() (_disassociate()
did have the correct order). The deauthentication frame is supposed
to go out (if driver supports that) before we disconnect more
forcefully.
Otherwise the driver might interpret the request as a request to
create/join a new adhoc network with the bogus SSID.
Signed-off-by: Dan Williams <dcbw@redhat.com>
hostapd_cli wps_pin command can now have an optional timeout
parameter that sets the PIN lifetime in seconds. This can be used
to reduce the likelihood of someone else using the PIN should an
active PIN be left in the Registrar.
The attached patch will replace get_scan_results with get_scan_results2.
This is a preparation for WPS on BSD.
And I erased "wpa_scan_result_compar". Because scan result
sorting will be done with "wpa_scan_result_compar" on
"scan_helpers.c".
I have done below tests on NetBSD with an atheros card.
- WPA2-PSK(CCMP)
- WPA-PSK(TKIP)
- PEAP(MSCHAPv2)
- EAP-TLS
- EAP-TTLS(MSCHAPv2)
This adds a cleaner mechanism for protecting against unauthorized
data frames than the previously used drop_unencrypted mechanism
(which is not even available with nl80211 and had to use a WEXT
ioctl.
The old drop_unencrypted code is left in for now as the final
surviving WEXT use in driver_nl80211.c. However, it can be removed
from the build by defining NO_WEXT. It may also be removed
eventually when most users are expected to be using recent enough
kernel version.
For now, the old code using NL80211_ATTR_STA_FLAGS is left in for
backwards compatibility with older kernel versions. It may be removed
eventually when most users are expected to be running with new
enough kernel version.
Some deployed implementations seem to advertise incorrect information
in this attribute. For example, Linksys WRT350N seems to have a
byteorder bug that breaks this negotiation. In order to interoperate
with existing implementations, assume that the Enrollee supports
everything we do.
Attached is a patch for the RoboSwitch driver in trunk. It is a
general revision of the source code.
Changes:
- Improved IEEE 802.1X conformance ([1])
- Better conformity to Broadcom specifications
- Fixed compatibility with different chipset revisions
It is worth noting that performance may drop a little using the new
driver. This can be overcome by using "multicast_only=1" as a
parameter. In that case only packets to the PAE group address are
regarded, as the previous revision of the driver did. A more detailed
description of the parameter and it's consequences is available at [2]
(summary: use "multicast_only=1" whenever possible).
[1] http://lists.shmoo.com/pipermail/hostap/2009-February/019398.html
[2] http://forum.openwrt.org/viewtopic.php?id=19873
None of the driver wrappers user this. hostapd-controlled broadcast SSID
hiding can only be used with drivers that use hostapd for handling
Beacon and Probe Request/Response frames.
None of the driver wrappers use this. Only the drivers that use hostapd
for Beacon and Probe Request/Response handling can now use IEEE 802.11d
properly.
Note: This changes values for existing NL80211_ATTR_MAX_SCAN_IE_LEN
and NL80211_ATTR_KEY_TYPE (for some reason, they ended up getting
swapped in wireless-testing.git) and as such, could break Michael
MIC error reporting (well, at least partially).
When in AP mode, wpa_supplicant is now enabling WPS (only Internal
Registrar). wpa_cli wps_pbc and wps_pin commands can be used to initiate
WPS negotiation similarly to how this is done in station mode (though,
the BSSID parameter is ignored).
There is no point in trying to continue if a 4-way handshake frame is
discarded or if PTK/GTK/IGTK configuration fails. Force the client to
disconnect in such a case to avoid confusing user by claiming the
connection was successfully completed.
These driver wrappers use UNIX domain sockets and as such, won't be
built with Microsoft compiler. Consequently, use of C99-style designated
initializers can be used to make these files easier to maintain. Only
driver_ndis.c needs to support the old C89-style due to Microsoft
compiler not supporting newer C standard.
This was not really supported by any of the included driver wrappers. If
this functionality is desired in the future, this (or something similar)
can be added with the changes needed into a driver wrapper to use the
mechanism.
This simplifies driver wrapper operations and removes last direct
struct hostapd_data dereferences from driver_nl80211.c. In addition,
some of the TX callbacks are now fixed for secondary BSSes.
This fixes deauth/disassoc frames in secondary BSSes when using
multi-BSSID. In addition, it reduces need to dereference
struct hostapd_data inside driver wrappers.
NO_WEXT can be defined to disable all remaining WEXT uses from
driver_nl80211.c. This breaks some functionality, so the code is still
in use by default. Anyway, the new #ifdef blocks make it easy to search
for areas that need to be converted to nl80211.
Instead of adding a new driver_ops for fetching neighbor BSS data (that
nl80211 driver interface had to scan during initialization), share the
same scan operations that wpa_supplicant is using. This gets rid of
duplicated scan code in driver_nl80211.c (and better yet, removes large
part of old WEXT code).
hostapd interface initialization is now completed in a callback, if
needed, i.e., he_features channel/hw_mode selection can use as much time
as needed. This can also help with radar detection in the future.
hostapd ended up with unregistered send_mlme() in commit
9f324b61ba. Fix this by registering the
handler both for wpa_supplicant-AP and hostapd use. [Bug 310]
Use a parameter structure to pass in information that can be more easily
extended in the future. Include some of the parameters that were
previously read directly from hapd->conf in order to reduce need for
including hostapd/config.h into driver wrappers.
This was not documented properly and was not really used nor would it be
suitable to be used in generic way as it was implemented. It is better
to just remove the parameter since there does not seem to be any
reasonable use for it.
nl80211 interface has a trivial bug that prevents it to work correctly
with channel 14. Channel frequency is erroneously 2848 instead of 2484.
To correct it just apply this patch. [Bug 308]
This merges the driver wrapper implementations to use the same
implementation both for hostapd and wpa_supplicant operations to avoid
code duplication.
This commit merges the driver_ops structures and implementations from
hostapd/driver*.[ch] into src/drivers. This is only an initial step and
there is room for number of cleanups to share code between the hostapd
and wpa_supplicant parts of the wrappers to avoid unnecessary source
code duplication.
Need to set WEP keys before requesting authentication in order to get
Shared Key authentication working. Previously, the WEP keys were not set
at all when using SME in wpa_supplicant.
IEEE 802.11r KDF uses key length in the derivation and as such, the PTK
length must be specified correctly. The previous version was deriving
using 512-bit PTK regardless of the negotiated cipher suite; this works
for TKIP, but not for CCMP. Update the code to use proper PTK length
based on the pairwise cipher.
This fixed PTK derivation for both IEEE 802.11r and IEEE 802.11w (when
using AKMP that specifies SHA-256-based key derivation). The fixed
version does not interoperate with the previous versions. [Bug 307]
mac80211 does not use this type of Michael MIC event, so this is not
really used and proper Michael MIC failure processing will be added
using an nl80211 event once that gets added into wireless-testing.git.
It looks like both new nl80211 operations (scan and auth/assoc) end up
being added in Linux 2.6.30, so we do not need to maintain the backwards
compatibility code in wpa_supplicant.
Update credential to only include a single authentication and
encryption type in case the AP configuration includes more than one
option. Without this, the credential would be rejected if the AP was
configured to allow more than one authentication type.
A new network block parameter, scan_freq, can be used to specify subset
of frequencies to scan. This can speed up scanning process considerably
if it is known that only a small subset of channels is actually used in
the network. A union of configured frequencies for all enabled network
blocks is used in scan requests.
Currently, only driver_nl80211.c has support for this functionality.
For example, following parameter marks 2.4 GHz channels 1, 6, 11 to be
scanned: scan_freq=2412 2437 2462
Move the shared IEEE 802.11w enum definition into src/common/defs.h to
avoid redefinition when both configuration structures are included into
the same file.
The memcpy was using incorrect size (of pointer; should have been of
data structure) for the client address. This ended up working on 64-bit
targets, but left part of the address uninitialized on 32-bit targets.
This version is adding the configuration option (mode=2) for this and
driver capability reporting to figure out whether AP mode can be used.
However, this does not actually implement any real functionality yet.
Need to set drv->ifindex before calling set_mode(). In addition, set the
mode before setting the interface up to avoid having to set it down
again.
Add more useful error message on mode changes.
All these driver handlers can be implemented in associate() handler
which gets all the needed information in the parameters structure. The
old functions that provided only a single parameter will be removed
eventually to clean up the driver_ops structure, so driver wrappers
should start using the newer mechanism.
This can be used, e.g., with mac80211-based Linux drivers with
nl80211. This allows over-the-air FT protocol to be used (IEEE
802.11r).
Since the nl80211 interface needed for this is very recent (added
today into wireless-testing.git), driver_nl80211.c has backwards
compatibility code that uses WEXT for association if the kernel does
not support the new commands. This compatibility code can be
disabled by defining NO_WEXT_COMPAT. That code will also be removed
at some point to clean up driver_nl80211.c.
On "eap_tnc_process" function error case, data->in_buf keeps reference
to a local scope variable. For example this will cause segmentation
fault in "eap_tnc_deinit" function "wpabuf_free(data->in_buf)"
statement.
This was broken by 510c02d4a3 which added
validation of eap_ttls_phase2_eap_init() return value. The main problem
in the code trying to initialize a new phase 2 EAP method
unconditionally; this should only happen if there is a new method in the
inner method sequence.
This is based on a patch and report by Masashi Honma
<honma@ictec.co.jp>. The issue is more generic than just TNC, though,
since failure to initialize any phase 2 EAP method can result in NULL
dereference.
On PEAP(TNC), hostapd integrated RADIUS server doesn't return
EAP-Failure when "Recommendation = none". So, EAP data retransmittion
occurs.
My co-worker "Ryuji Ohba" made below patch.
hostapd will now go through the RIC Request and process each RDIE. Only
WMM TSPEC requests are currently supported; all other request
descriptors will be declined.
RIC Response is written by hostapd and verified by wpa_supplicant (MIC
validation). wpa_supplicant does not yet have code to notify the driver
about the resource request results.
This adds first part of FT resource request as part of Reassocition
Request frame (i.e., FT Protocol, not FT Resource Request Protocol).
wpa_supplicant can generate a test resource request when driver_test.c
is used with internal MLME code and hostapd can verify the FTIE MIC
properly with the included RIC Request.
The actual RIC Request IEs are not processed yet and hostapd does not
yet reply with RIC Response (nor would wpa_supplicant be able to
validate the FTIE MIC for a frame with RIC Response).
Windows XP and Vista clients can get confused about EAP-Identity/Request
when they probe the network with EAPOL-Start. In such a case, they may
assume the network is using IEEE 802.1X and prompt user for a
certificate while the correct (non-WPS) behavior would be to ask for the
static WEP key. As a workaround, use Microsoft Provisioning IE to
advertise that legacy 802.1X is not supported.
This seems to make Windows ask for a static WEP key when adding a new
network, but at least Windows XP SP3 was still marking IEEE 802.1X
enabled for the network. Anyway, this is better than just leaving the
network configured with IEEE 802.1X and automatic WEP key distribution.
When using the internal TLS implementation, EAP-FAST unauthenticated
provisioning ends up proposing multiple cipher suites. It looks like
Cisco AP (at least 350 and 1200 series) local authentication server does
not know how to search cipher suites from the list and seem to require
that the last entry in the list is the one that it wants to use.
However, TLS specification requires the list to be in the client
preference order. As a workaround, ass anon-DH AES-128-SHA1 again at the
end of the list to allow the Cisco code to find it.
This fixed EAP-FAST provisioning with the following IOS version:
Cisco IOS Software, C350 Software (C350-K9W7-M), Version 12.3(8)JEA3,
RELEASE SOFTWARE (fc2)
Compiled Wed 21-Nov-07 14:08 by ccai
This is just there very first step on being able to do something with
wireless LAN on Vista. There is some example code for requesting a scan,
but it does not work in its current form. Anyway, this adds a wpa_printf
noting that Native 802.11 drivers are not yet supported.
This is done with wired interfaces to fix IEEE 802.1X authentication
when the authenticator uses the group address (which should be happening
with wired Ethernet authentication).
This allows wpa_supplicant to complete wired authentication successfully
on Vista with a NDIS 6 driver, but the change is likely needed for
Windows XP, too.
Do not use just the driver name for this since driver_ndis.c supports
both wired and wireless NDIS drivers and needs to indicate the driver
type after initialization.
Calculate the estimated medium time using integer variables since there
is no real need to use floating point arithmetics here. In addition,
make sure there is no division by zero regardless of how invalid the
request from the station is. Reject invalid parameters and refuse
requests that would take most of the bandwidth by themselves.
Add test code into wpa_supplicant mlme.c to allow WMM-AC negotiation to
be tested with driver_test.
The new file wps_nfc.c and ndef.c implements NFC device independent
operation, wps_nfc_pn531.c implements NFC device dependent operation.
This patch is only for the following use case:
- Enrollee = wpa_supplicant
- Registrar = hostapd internal Registrar
Following NFC methods can be used:
- Enrollee PIN with NFC
- Registrar PIN with NFC
- unencrypted credential with NFC
Encrypted credentials are not supported.
Enrollee side operation:
Registrar side operation:
Example configuration.
CONFIG_WPS=y
CONFIG_WPS_NFC=y
CONFIG_WPS_NFC_PN531=y
I used NFC device "NXP PN531". The NFC device access method is
confidential, so I used outer library. Please download below files from
https://www.saice-wpsnfc.bz/index.php
[WPS NFC Library]
WpsNfcLibrary/WpsNfc.h
WpsNfcLibrary/WpsNfcType.h
WpsNfcLibrary/WpsNfcVersion.h
WpsNfcLibrary/linux/libnfc_mapping_pn53x.dll
WpsNfcLibrary/linux/wpsnfc.dll
[NFC Reader/Writer Kernel Driver]
NFCKernelDriver-1.0.3/linux/kobj/sonyrw.ko
<WiFi test>
The hostapd/wpa_supplicant with this patch passed below tests on
"Wi-Fi WPS Test Plan Version 1.6".
4.2.5 Add device using NFC Method with password token
(I used SONY STA instead of NXP STA.)
4.2.6 Add device using NFC Method with configuration token
5.1.9 Add to AP using NFC Method with password token
through internal registrar
(I used SONY AP instead of NXP AP.)
5.1.10 Add to AP using NFC Method with configuration token
through internal registrar
This updates the terminogy to match with the final WMM specification. In
addition, some of the WMM TSPEC structure fields were in incorrect order
and used without byte order swapping. Those are also taken care of this
cleanup patch.
The EAP server state machine will need to have special code in
getDecision() to avoid starting passthrough operations before having
completed Identity round in the beginning of reauthentication. This was
broken when moving into using the full authenticator state machine from
RFC 4137 in 0.6.x.
The old behavior of generating new DH keys can be maintained for non-OOB
cases and only OOB (in this case, with UFD) will use the pre-configured
DH keys to allow the public key hash to be checked.
Not all embedded devices have USB interface and it is useful to be able
to remove unneeded functionality from the binary. In addition, the
current implementation has some UNIX specific calls in it which may make
it not compile cleanly on all target systems.
Reduce startWhen from 3 to 1 second if WPS is included in the build.
While this is done regardless of runtime WPS configuration, it is fine
to use a smaller value here in general. This cuts two seconds out from
WPS negotiation if the driver does not support addition of WPS IE into
the (Re)Association Request frame.
Previously, both CONFIG_IEEE80211W=y and CONFIG_IEEE80211R=y were needed
to enable SHA256-based key handshake (WPA-PSK-SHA256 and
WPA-EAP-SHA256). This can now be done with just CONFIG_IEEE80211W=y.
These flags are used to mark which values (level, noise, qual) are
invalid (not available from the driver) and whether level is using dBm.
D-Bus interface will now only report the values that were available.
If session resumption fails for any reason, do not try it again because
that is just likely to fail. Instead, drop back to using full
authentication which may work. This is a workaround for servers that do
not like session resumption, but do not know how to fall back to full
authentication properly.
This fixes an issue where two AKA'/Challenge messages are received when
resynchronizing SEQ#. Previously, this used to trigger an authentication
failure since the second Challenge message did not duplicate AT_KDF.
This does not actually send out separate scan requests for each SSID,
but the debug output can be used to test scan2() functionality with
multiple SSIDs.
This can be used to provide support for scanning multiple SSIDs at a
time to optimize scan_ssid=1 operations. In addition, Probe Request IEs
will be available to scan2() (e.g., for WPS PBC scanning).
For example, -Dnl80211,wext could be used to automatically select
between nl80211 and wext. The first driver wrapper that is able to
initialize the interface will be used.
This workaround was needed with some drivers that used WEXT, but there
is no known nl80211-enabled driver that would need this, so lets get rid
of the extra delay.
The driver wrappers can now inform wpa_supplicant how many SSIDs can
be used in a single scan request (i.e., send multiple Probe Requests
per channel). This value is not yet used, but it can eventually be used
to allow a new scan command to specify multiple SSIDs to speed up
scan_ssid=1 operations. In addition, a warning could be printed if
scan_ssid=1 is used with a driver that does not support it
(max_scan_ssids=0).
Some drivers (for example ipw2100) do not report signal level but only
signal quality. wpa_supplicant already uses the signal quality if no
level is reported and all other comparision parameters are equal to sort
the scan results. However, if two APs have different max rates and the
signal level does not differ much wpa_supplicant chooses the AP with the
higher max rate.
In case of ipw2100 no signal level is reported and thus wpa_supplicant
always takes the AP with higher max rate even if its signal quality is
really low. For example if AP1 (max rate 11Mb/s, 80% signal quality) and
AP2 (max rate 54 Mb/s, 20% signal quality) are found by a scan
wpa_supplicant would choose AP2.
Hence, if no signal level is reported depend on the signal quality if
max rate should be compared. A quality difference of 10% is considered
acceptable in favor of the higher max rate.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Really disassociate when tearing stuff down; drivers may sometimes
(legally) keep trying to reassociate unless the BSSID is unlocked. If
the SSID is unlocked too, under WEXT drivers are able to pick an SSID to
associate, so kill that behavior by setting a bogus SSID. Unfortunately
WEXT doesn't provide an easy method to say "stop whatever doing and just
idle".
Signed-off-by: Dan Williams <dcbw@redhat.com>
It looks like GnuTLS (at least newer versions) is using random padding
on the application data and the previously used 100 byte extra buffer
for tls_connection_encrypt() calls was not enough to handle all cases.
This resulted in semi-random authentication failures with EAP-PEAP and
EAP-TTLS during Phase 2.
Increase the extra space for encryption from 100 to 300 bytes and add an
error message into tls_gnutls.c to make it easier to notice this issue
should it ever show up again even with the larger buffer.
I tried PBC with the hostapd registrar.
I pushed the button with "hostap_cli WPS_PBC".
But hostapd registrar always sends Selected Registrar Config Methods
attribute=0x0000 in beacon/probe response.
Allow more than one pending PutWLANMessage data to be stored (M2/M2D
from multiple external Registrars) and drop pending M2/M2D messages when
the Enrollee replies with M3.
Version attribute processing details are not described in the WPS spec,
but it is safer to allow minor version to change and only refuse to
process the message if major version is different from ours. This
matches with the behavior used in the Intel reference implementation.
It looks like we don't set correspondent Device Password ID attribute in
M2 message during PBC registration. Without it TG185n STA was not able
to connect to our AP in PBC mode. Attached patch fixes this.
The madwifi driver has interface to set probe request ie.
Attached patch will enable the functionality.
I could see probe request includes WSC IE with this patch.
We can now handle up to 65535 byte result buffer which is the maximum
due to WEXT using 16-bit length field. Previously, this was limited to
32768 bytes in practice even through we tried with 65536 and 131072
buffers which we just truncated into 0 in the 16-bit variable.
This more or less doubles the number of BSSes we can received from scan
results.
I am terribly sorry, but because of a lack of testing equipment the
patch was submitted not properly tested.
Because the chipset documentation is not publicly available all
behaviour has to be found out by experimentation. The other day, I
made some incorrect assumptions based on my findings.
I do believe the attached patch does support the whole RoboSwitch line
(5325, 5350, 5352, 5365 and others). It is a drop-in substitution for
my previous submission.
The RoboSwitch driver of wpa_supplicant had one shortcoming: not
supporting the 5365 series. I believe the patch attached fixes this
problem.
Furthermore it contains a small readability rewrite. It basically is an
explicit loop-rollout so that the wpa_driver_roboswitch_leave style
matches that of wpa_driver_roboswitch_join.
If the bind() on /tmp/wpa_ctrl_<pid>_<in-proc-counter> fails with
EADDRINUSE, there is an existing socket file with the name we are trying
to create. Since getpid() is unique, there cannot be another process
using that socket and we can just unlink the file and try again. This
can speed up client connection if wpa_cli is killed without allowing it
to clean up the socket file. [Bug 288]
This adds mostly feature complete external Registrar support with the
main missing part being proper support for multiple external Registrars
working at the same time and processing of concurrent registrations when
using an external Registrar.
This code is based on Sony/Saice implementation
(https://www.saice-wpsnfc.bz/) and the changes made by Ted Merrill
(Atheros) to make it more suitable for hostapd design and embedded
systems. Some of the UPnP code is based on Intel's libupnp. Copyrights
and licensing are explained in src/wps/wps_upnp.c in more detail.
We must not replace M1 with M2D as the last_msg since we need M1 to
validate a possible M2 after M2D. Since M2D and ACK/NACK replies do not
include Authenticator attribute, we can just ignore M2D as far as
updating last_msg is concerned.
External Registrar in Vista may include NULL termination in the Network
Key when encoding an ASCII passphrase for WPA/WPA2-PSK. As a workaround,
remove this extra octet if present.
If a Registrar tries to configure the AP, but fails to validate the
device password (AP PIN), lock the AP setup after four failures. This
protects the AP PIN against brute force guessing attacks.
This optional configuration parameter can be used to override AP
Settings attributes in M7 similarly to extra_cred option for Credential
attribute(s) in M8.
Buffalo WHR-G125 Ver.1.47 seems to send EAP-WPS packets with too short
EAPOL header length field (14 octets regardless of EAP frame length).
This is fixed in firmware Ver.1.49, but the broken version is included
in many deployed APs. As a workaround, fix the EAPOL header based on the
correct length in the EAP packet. This workaround can be disabled with
eap_workaround=0 option in the network configuration.
Some WPS APs do not set Selected Registrar attribute to 1 properly when
using an external Registrar. Allow such an AP to be selected for PIN
registration after couple of scan runs that do not find APs marked with
Selected Registrar = 1. This allows wpa_supplicant to iterate through
all APs that advertise WPS support without delaying connection with
implementations that set Selected Registrar = 1 properly.
WPS spec is not very specific on the presentation used for the DH
values. The Public Key attribute is described to be 192 octets long, so
that could be interpreted to imply that other places use fixed length
presentation for the DH keys. Change the DH derivation to use fixed
length bufferd by zero padding them from beginning if needed. This can
resolve infrequent (about 1/256 chance for both Public Key and Shared
Key being shorter) interop issues.
This is needed to allow external Registrar (at least the implementation
in Windows Vista) to configure the Enrollee.
With this patch and my previous patch (for wps.c) , I could pass "Wi-Fi
WPS Test Plan Version 1.0 [5.1.4. Add to AP using PIN Config method and
PASS PHRASE through wired external registrar]".
I can't pass the "Wi-Fi WPS Test Plan Version 1.0 [5.1.4. Add to AP
using PIN Config method and PASS PHRASE through wired external
registrar]". The wpa_supplicant-0.6.7 can't recoginize the testbed
AP(BCM94704AGRRev-E.2.4) as WPS PIN AP. Because after PIN entered, the
AP sends Selected Registrar attribute=0 and not send Device Password ID
attribute.
The proposed change as-is removed validation of Selected Registrar
attribute completely. However, that part is not included in this commit
since it can cause problems for environments with multiple WPS-enabled
APs. Another workaround for this will be considered in wpa_supplicant
scanning process (e.g., start trying to use WPS with APs that do not set
Selected Registrar to TRUE after couple of scan runs that do not find
any APs with Selected Registrar TRUE).
This operation can now be moved into an external program by configuring
hostapd with wps_cred_processing=1 and skip_cred_build=1. A new
ctrl_iface message (WPS-REG-SUCCESS <Enrollee MAC addr> <UUID-E>) will
be used to notify external programs of each successful registration and
that can be used as a tricker to move from unconfigured to configured
state.
The separate Association Comeback Time IE was removed from IEEE 802.11w
and the Timeout Interval IE (from IEEE 802.11r) is used instead. The
editing on this is still somewhat incomplete in IEEE 802.11w/D7.0, but
still, the use of Timeout Interval IE is the expected mechanism.
This makes it easier to pass the credential data to external programs
(e.g., Network Manager) for processing. The actual use of this data is
not yet included in hostapd/wpa_supplicant.
This commit adds a new build option, CONFIG_IBSS_RSN=y, that can be used
to enable RSN support for IBSS. This links in RSN Authenticator code
from hostapd and adds code for managing per-peer information for IBSS. A
new wpa_cli command or driver event can be used to request RSN
authentication with an IBSS peer. New RSN Authenticator and Supplicant
will be allocated for each peer.
The basic state machine setup code is included in this commit, but the
state machines are not properly started yet. In addition, some of the
callback functions are not yet complete.
This is needed for airo driver to work correctly and no other driver
seems to care, so the change is safe to make. This has been in number of
distro releases for a long time and no issues have been reported.
The configuration parsing functions seemed to have worked fine before,
but these were real bugs even if they did not show up in practice.
hostapd_ip_diff() was broken for IPv6 addresses (overwrote address and
always returned 1.
Generate a SHA1 hash -based UUID from the local MAC address if the UUID
was not configured. This makes it easier to prepare for WPS since there
is no need to generate an UUID.
IEEE 802.11w/D7.0 incorrectly changed the Action Category from 8 to 7
when renaming Ping to SA Query. Category 7 is reserved for HT (IEEE
802.11n) and IEEE 802.11w will need to continue to use the category 8
that was allocated for it.
Previous version assumed that the Flags field is always present and
ended up reading one octet past the end of the buffer should the Flags
field be missing. The message length would also be set incorrectly
(size_t)-1 or (size_t)-5, but it looks like reassembly code ended up
failing in malloc before actually using this huge length to read data.
RFC 2716 uses a somewhat unclear description on what exactly is included
in the TLS Ack message ("no data" can refer to either Data field in 4.1
or TLS Data field in 4.2), so in theory, it would be possible for some
implementations to not include Flags field. However,
EAP-{PEAP,TTLS,FAST} need the Flags field in Ack messages, too, for
indicating the used version.
The EAP peer code will now accept the no-Flags case as an Ack message if
EAP workarounds are enabled (which is the default behavior). If
workarounds are disabled, the message without Flags field will be
rejected.
[Bug 292]
It looks like this never survived the move from IEEE 802.1X-2001 to
IEEE 802.1X-2004 and EAP state machine (RFC 4137). The retransmission
scheduling and control is now in EAP authenticator and the
calculateTimeout() producedure is used to determine timeout for
retransmission (either dynamic backoff or value from EAP method hint).
The recommended calculations based on SRTT and RTTVAR (RFC 2988) are not
yet implemented since there is no round-trip time measurement available
yet.
This should make EAP authentication much more robust in environments
where initial packets are lost for any reason. If the EAP method does
not provide a hint on timeout, default schedule of 3, 6, 12, 20, 20, 20,
... seconds will be used.
Previously, only the delivery option 1 from RFC 4284
(EAP-Request/Identity from the AP) was supported. Now option 3
(subsequent EAP-Request/Identity from RADIUS server) can also be used
when hostapd is used as a RADIUS server. The eap_user file will need to
have a Phase 1 user entry pointing to Identity method in order for this
to happen (e.g., "* Identity" in the end of the file). The identity hint
is configured in the same was as for AP/Authenticator case (eap_message
in hostapd.conf).
This commit changes just the name and Action category per D7.0. The
retransmit/timeout processing in the AP is not yet updated with the
changes in D7.0.
Some deployed WPS implementations fail to parse zero-length attributes.
As a workaround, send a null character if the device attribute string is
empty. This allows default values (empty strings) to be used without
interop issues.
Before this change, it looked like an AP that was using wsccmd did not
get activated since wsccmd left the Selected Registrar Config Methods
attribute to be zero. Since Device Password ID can be used to
distinguish PBC from any other method, use only it to figure out whether
PBC or PIN method is active.
The new INTERFACE_LIST global control interface command can be used to
request a list of all available network interfaces that could be used
with the enabled driver wrappers. This could be used to enable
interfaces automatically by external programs (e.g., wpa_gui).
Driver wrappers can now register global_init() and global_deinit()
driver_ops handlers to get a global data structure that can be shared
for all interfaces. This allows driver wrappers to initialize some
functionality (e.g., interface monitoring) before any interfaces have
been initialized.
This callback is now used to stop wpa_supplicant from trying to continue
using parameters (most likely, device password) that do not work in a
loop. In addition, wpa_gui can now notify user of failed registration.
The event callback will be used for various event messages and the M2D
notification is the first such message. It is used to notify wpa_gui
about Registrar not yet knowing the device password (PIN).
We need to be a bit more careful when removing the WPS configuration
block since wpa_s->current_ssid may still be pointing at it. In
addition, registrar pointer in wps_context will need to be cleared
since the context data is now maintained over multiple EAP-WSC runs.
Without this, certain WPS operations could have used freed memory.
Instead of sending out EAP-Failure on errors (on AP) or stopping (on
Supplicant), send a NACK message based on the allowed EAP state machine
transitions for EAP-WSC.
Whenever new scan results include WPS AP(s) and the client is not
associated, send a notification message to control interface monitors.
This makes it easier for GUIs to notify the user about possible WPS
availability without having to go through the scan results.
Windows Server 2008 NPS gets very confused if the TLS Message Length is
not included in the Phase 1 messages even if fragmentation is not used.
If the TLS Message Length field is not included in ClientHello message,
NPS seems to decide to use the ClientHello data (excluding first six
octets, i.e., EAP header, type, Flags) as the OuterTLVs data in
Cryptobinding Compound_MAC calculation (per PEAPv2; not MS-PEAP)..
Lets add the TLS Message Length to PEAPv0 Phase 1 messages to get rid of
this issue. This seems to fix Cryptobinding issues with NPS and PEAPv0
is now using optional Cryptobinding by default (again) since there are
no known interop issues with it anymore.
Changed peer to derive the full key (both MS-MPPE-Recv-Key and
MS-MPPE-Send-Key for total of 32 octets) to match with server
implementation.
Swapped the order of MPPE keys in MSK derivation since server
MS-MPPE-Recv-Key | MS-MPPE-Send-Key matches with the order specified for
EAP-TLS MSK derivation. This means that PEAPv0 cryptobinding is now
using EAP-MSCHAPv2 MSK as-is for ISK while EAP-FAST will need to swap
the order of the MPPE keys to get ISK in a way that interoperates with
Cisco EAP-FAST implementation.
driver_test can now be used either over UNIX domain socket or UDP
socket. This makes it possible to run the test over network and makes it
easier to port driver_test to Windows.
hostapd configuration: test_socket=UDP:<listen port>
wpa_supplicant configuration: driver_param=test_udp=<dst IP addr>:<port>
This adds support for setting of a regulatory domain to wpa_supplicant
drivers. It also adds regulatory domain setting for the nl80211 driver.
We expect an ISO / IEC 3166 alpha2 in the wpa configuration file as a
global.
Since only one KDF is currently supported, the negotiation is not
allowed and peer must be rejected if it tries to send KDF selection in a
Challenge message. The negotiation code is left in the file and just
commented out since it was tested to work and can be used in the future
if another KDF is added.
The attribute uses 'Actual Identity Length' field to indicate the exact
(pre-padding) length of the Identity. This actual length should be used
as the length, not the remaining attribute length.
This was previously worked around by stripping null termination away
from the end of the identity string at EAP-SIM and EAP-AKA server code.
However, it is likely that that workaround is not really needed and the
real problem was in AT_IDENTITY parsing. Anyway, the workaround is left
in just in case it was really needed with some implementations.
This IE is not (at least yet) actually used for anything, but parsing it
cleans up verbose debug log a bit since thie previously unknown, but
commonly used, vendor IE was being reported as unknown.
This allows the same source code file to be shared for both methods. For
now, this is only in eap_aka_prime.c, but eventually, changes in
eap_aka_prime.c are likely to be merged into eap_aka.c at which point
the separate eap_aka_prime.c can be removed.
This is just making an as-is copy of EAP-AKA server and peer
implementation into a new file and by using the different EAP method
type that is allocated for EAP-AKA' (50). None of the other differences
between EAP-AKA and EAP-AKA' are not yet included.
It is likely that once EAP-AKA' implementation is done and is found to
work correctly, large part of the EAP-AKA and EAP-AKA' code will be
shared. However, it is not reasonable to destabilize EAP-AKA
implementation at this point before it is clearer what the final
differences will be.
Since the Registrar may not yet know the UUID-E when a new PIN is
entered, use of a wildcard PIN that works with any UUID-E can be useful.
Such a PIN will be bound to the first Enrollee trying to use it and it
will be invalidated after the first use.
Fragment WPS IE if needed to fit into the IE length limits in hostapd
and Reassemble WPS IE data from multiple IEs in wpa_supplicant.
In addition, moved WPS code from events.c into wps_supplicant.c to clean
up module interfaces.
These functions fit in better with the category of functions included in
wps.c. wps_common.c is now used for generic helper functions (currently,
only crypto code).
WPS IE is now passed from hostapd association processing into EAP-WSC
and WPS processing. Request Type attribute is parsed from this
information and if the request is for a WLAN Manager Registrar,
additional management keys are derived (to be used with UPnP).
The wps_context data is now managed at wpa_supplicant, not EAP-WSC. This
makes wpa_supplicant design for WPS match with hostapd one and also
makes it easier configure whatever parameters and callbacks are needed
for WPS.
Previously, wpa_supplicant as Enrollee case was handled using a
different callback function pointer. However, now that the wps_context
structure is allocated for all cases, the same variable can be used in
all cases.
Previously, hardcoded values were used in wps_enrollee.c. These are now
moved into shared data in struct wps_context. In case of
AP/Authenticator, these are initialized in wps_hostapd.c. In case of
client/supplicant, these are now initialized in EAP-WSC peer method,
but will probably end up being moved into higher layer for better
configuration.
EAP-WSC peer method for
This allows the network to be used after the Registrar configuration
step. The local WPS network is replaced with a new network block
similarly to the case of acting as an Enrollee.
This makes it easier to store old AP settings into wps->cred (and allow
them to modified and taken into use in the future). Separation between
Credential and AP Settings building is also cleaner in this design.
The old (i.e., currently used) AP Settings are processed. For now, they
are copied as-is into M8 as new AP Settings to avoid changing
configuration. This should be changed to allow external programs (e.g.,
GUI) to fetch the old AP settings over ctrl_iface and then allow
settings to be changed before sending M8 with the new settings.
The core processing of attributes into struct wps_credential is now in
wps_common.c (was in wps_enrollee.c), so that the same code can be
shared with Registrar.
This adds WPS support for both hostapd and wpa_supplicant. Both programs
can be configured to act as WPS Enrollee and Registrar. Both PBC and PIN
methods are supported.
Currently, hostapd has more complete configuration option for WPS
parameters and wpa_supplicant configuration style will likely change in
the future. External Registrars are not yet supported in hostapd or
wpa_supplicant. While wpa_supplicant has initial support for acting as
an Registrar to configure an AP, this is still using number of hardcoded
parameters which will need to be made configurable for proper operation.
It looks like some Windows NDIS drivers (e.g., Intel) do not clear the
PMKID list even when wpa_supplicant explicitly sets the list to be
empty. In such a case, the driver ends up trying to use PMKSA caching
with the AP and wpa_supplicant may not have the PMK that would be needed
to complete 4-way handshake.
RSN processing already had some code for aborting PMKSA caching by
sending EAPOL-Start. However, this was not triggered in this particular
case where the driver generates the RSN IE. With this change, this case
is included, too, and the failed PMKSA caching attempt is cleanly
canceled and wpa_supplicant can fall back to full EAP authentication.
It the message was large enough to require fragmentation (e.g., if a large
Session Ticket data is included), More Fragment flag was set, but no
more fragments were actually sent (i.e., Access-Accept was sent out).
It looks like [MS-PEAP] 3.2.5.6 points towards this being the expected
behavior (however, that chapter is very confusing).
In addition, remove Cryptobinding TLV from response if the received
Cryptobinding TLV is not valid. Add some more debug messages to the case
where the received Cryptobinding TLV is found invalid.
I fixed the engine issue in phase2 of EAP-TTLS. The problem was that you
only defined one engine variable, which was read already in phase1. I
defined some new variables:
engine2
engine2_id
pin2
and added support to read those in phase2 wheres all the engine
variables without number are only read in phase1. That solved it and I
am now able to use an engine also in EAP-TTLS phase2.
Find attached the patch that creates a new driver: roboswitch. This
driver adds support for wired authentication with a Broadcom
RoboSwitch chipset. For example it is now possible to do wired
authentication with a Linksys WRT54G router running OpenWRT.
LIMITATIONS
- At the moment the driver does not support the BCM5365 series (though
adding it requires just some register tweaks).
- The driver is also limited to Linux (this is a far more technical
restriction).
- In order to compile against a 2.4 series you need to edit
include/linux/mii.h and change all references to "u16" in "__u16". I
have submitted a patch upstream that will fix this in a future version
of the 2.4 kernel. [These modifications (and more) are now included in
the kernel source and can be found in versions 2.4.37-rc2 and up.]
USAGE
- Usage is similar to the wired driver. Choose the interfacename of
the vlan that contains your desired authentication port on the router.
This name must be formatted as <interface>.<vlan>, which is the
default on all systems I know.
Remove the old code from driver_wext.c since the private ioctl interface is
never going to be used with mac80211. driver_nl80211.c has an
implementation than can be used with mac80211 (with two external patches to
enable userspace MLME configuration are still required, though).
Updated OpenSSL code for EAP-FAST to use an updated version of the
session ticket overriding API that was included into the upstream
OpenSSL 0.9.9 tree on 2008-11-15 (no additional OpenSSL patch is
needed with that version anymore).
It looks like ACS did not like PAC Acknowledgment TLV before Result TLV, so
reorder the TLVs to match the order shown in a
draft-cam-winget-eap-fast-provisioning-09.txt example. This allows
authenticated provisioning to be terminated with Access-Accept (if ACS has
that option enabled). Previously, provisioning was otherwise successful,
but the server rejected connection due to not understanding the PAC Ack
("Invalid TEAP Data recieved").
Previously, hardcoded identity in the network configuration skipped both
IMSI reading and PIN verification. This broke cases where PIN is needed for
GSM/UMTS authentication. Now, only IMSI reading is skipped if identity is
hardcoded.
This change breaks interoperability with older wpa_supplicant versions
(everything up to and including wpa_supplicant 0.5.10 and 0.6.5) which
incorrectly used this field as number of bytes, not bits, in RES.
Instead of falling back to full TLS handshake on expired PAC, allow the
PAC to be used to allow a PAC update with some level of server
authentication (i.e., do not fall back to full TLS handshake since we
cannot be sure that the peer would be able to validate server certificate
now). However, reject the authentication since the PAC was not valid
anymore. Peer can connect again with the newly provisioned PAC after this.
Added a new configuration option, wpa_ptk_rekey, that can be used to
enforce frequent PTK rekeying, e.g., to mitigate some attacks against TKIP
deficiencies. This can be set either by the Authenticator (to initiate
periodic 4-way handshake to rekey PTK) or by the Supplicant (to request
Authenticator to rekey PTK).
With both wpa_ptk_rekey and wpa_group_rekey (in hostapd) set to 600, TKIP
keys will not be used for more than 10 minutes which may make some attacks
against TKIP more difficult to implement.
A driver was found to remove SSID IE from NDIS_WLAN_BSSID_EX IEs, but the
correct SSID is included in NDIS_802_11_SSID structure inside the BSSID
data. If this is seen in scan results, create a matching SSID IE and add it
to the end of IEs to fix scan result parsing.
Need to make sure that portValid is TRUE in order to avoid PAE state
machine going into DISCONNECTED state on eapol_sm_step(). This could be
triggered at least with OKC.
Changed EAP-FAST configuration to use separate fields for A-ID and
A-ID-Info (eap_fast_a_id_info) to allow A-ID to be set to a fixed
16-octet len binary value for better interoperability with some peer
implementations; eap_fast_a_id is now configured as a hex string.
eap_fast_prov config parameter can now be used to enable/disable different
EAP-FAST provisioning modes:
0 = provisioning disabled
1 = only anonymous provisioning allowed
2 = only authenticated provisioning allowed
3 = both provisioning modes allowed
draft-cam-winget-eap-fast-provisioning-06.txt or RFC 4851 do not seem to
mandate any particular order for TLVs, but some interop issues were noticed
with an EAP-FAST peer implementation when Result TLV followed PAC TLV. The
example in draft-cam-winget-eap-fast-provisioning-06.txt shows the TLVs in
the other order, so change the order here, too, to make it less likely to
hit this type of interop issues.
This adds all the attributes that are marked as mandatory for SoH in
IF-TNCCS-SOH v1.0. MS-Machine-Inventory does not contain correct data
(i.e., all version fields are just marked as inapplicable) and
MS-MachineName is hardcoded to wpa_supplicant@w1.fi for now.
It is possible that the initialization of the Phase 2 EAP method fails and
if that happens, we need to stop EAP-TTLS server from trying to continue
using the uninitialized EAP method. Otherwise, the server could trigger
a segmentation fault when dereferencing a NULL pointer.
A bug just got reported as a result of this for mac80211 drivers.
https://bugzilla.redhat.com/show_bug.cgi?id=459399
The basic problem is that since taking the device down clears the keys
from the driver on many mac80211-based cards, and since the mode gets
set _after_ the keys have been set in the driver, the keys get cleared
on a mode switch and the resulting association is wrong. The report is
about ad-hoc mode specifically, but this could happen when switching
from adhoc back to managed mode.
IEEE 802.11w/D6.0 defines new AKMPs to indicate SHA256-based algorithms for
key derivation (and AES-CMAC for EAPOL-Key MIC). Add support for using new
AKMPs and clean up AKMP processing with helper functions in defs.h.
This updates management frame protection to use the assocition ping process
from the latest draft (D6.0) to protect against unauthenticated
authenticate or (re)associate frames dropping association.
This adds most of the new frame format and identifier definitions from IEEE
802.11w/D6.0. In addition, the RSN IE capability field values for MFP is
replaced with the new two-bit version with MFPC (capable) and MFPR
(required) processing.
If IWEVGENIE or custom event wpa_ie/rsn_ie is received in scan with empty
buffer, the previous version ended up calling realloc(NULL, 0) which seems
to return a non-NULL value in some cases. When this return value is passed
again into realloc with realloc(ptr, 0), the returned value could be NULL.
If the ptr is then freed (os_free(data.ie) in SIOCGIWAP handling), glibc
may crash due to invalid pointer being freed (or double-freed?). The
non-NULL realloc(NULL, 0) return value from glibc looks a bit odd behavior,
but anyway, better avoid this case completely and just skip the IE events
that have an empty buffer.
This issue should not show up with drivers that produce proper scan results
since the IEs will always include the two-octet header. However, it seems
to be possible to see this when using 64-bit kernel and 32-bit userspace
with incorrect compat-ioctl processing.
When the TLS handshake had been completed earlier by the server in case of
abbreviated handshake, the output buffer length was left uninitialized. It
must be initialized to zero in this case. This code is used by EAP-FAST
server and the uninitialized length could have caused it to try to send a
very large frame (though, this would be terminated by the 50 roundtrip EAP
limit). This broke EAP-FAST server code in some cases when PAC was used to
establish the tunnel.
This commit brings in cleaned up version of IEEE 802.11n implementation
from Intel (1). The Intel tarball includes number of other changes, too,
and only the changes specific to IEEE 802.11n are brought in here. In
addition, this does not include all the changes (e.g., some of the
configuration parameters are still missing and driver wrapper changes for
mac80211 were not included).
(1)
http://www.kernel.org/pub/linux/kernel/people/chuyee/wireless/iwl4965_ap/hostap_0_6_0_intel_0.0.13.1.tgz
These functions are based on the hostapd implementation and complete
the userspace MLME code in wpa_supplicant (though, mac80211 will still need
couple of pending patches to be integrated in order to get userspace client
MLME working again).
This adds some parts needed to use usermode MLME with the current mac80211
(plus a patch to add a new cfg80211 command; not yet submitted to
wireless-testing). This version creates a monitor interface for management
frames and is able to send Probe Request frames during scan. However, it
looks like management frame reception is not yet working properly. In
addition, mlme_{add,remove}_sta() handlers are still missing.
Network device ifindex will change when the interface is re-inserted.
driver_nl80211.c will need to accept netlink events from "unknown" (based
on ifindex) interfaces when a previously used card was removed earlier. If
the previously removed interface is added back, the driver_wext data need
to be updated to match with the new ifindex value. In addition, the initial
setup tasks for the card (set interface up, update ifindex, set mode, etc.)
from wpa_driver_nl80211_init() need to be run again.
This is the changes from commit 3fbda8f943
(driver_wext.c) ported for driver_nl80211.c.
wpa_sm_set_config() can be called even if the network block does not
change. However, the previous version ended up calling
pmksa_cache_notify_reconfig() every time and this cleared the network
context from PMKSA cache entries. This prevented OKC from ever being used.
Do not call pmksa_cache_notify_reconfig() if the network context remains
unchanged to allow OKC to be used.
Network device ifindex will change when the interface is re-inserted.
driver_wext.c will need to accept netlink events from "unknown" (based on
ifindex) interfaces when a previously used card was removed earlier. If the
previously removed interface is added back, the driver_wext data need to be
updated to match with the new ifindex value. In addition, the initial setup
tasks for the card (set interface up, update ifindex, set mode, etc.) from
wpa_driver_wext_init() need to be run again.
The change to support fragmentation added extra function to generate the
EAP header, but forgot to remove the original code and ended up getting two
EAP headers and TNC flags field in the generated message. These header
fields need to be added only in the function that builds the final message
(and if necessary, fragments the data).
When scan results got moved from wpa_scan_result -> wpa_scan_res, the
'maxrate' member was dropped from wpa_scan_res. The D-Bus interface
used 'maxrate', which was replaced with wpa_scan_get_max_rate().
Unfortunately, wpa_scan_get_max_rate() returns 802.11 rate values
directly from the IE, where 'maxrate' was the rate in bits/second. The
supplicant internally fakes an IE for wpa_scan_res from the value of
wpa_scan_result->maxrate, but interprets ->maxrate as an 802.11 rate
index.
As a side-effect, this fixes a soft-break of the D-Bus control API since
the wpa_scan_res change was introduced.
Function 'wpa_sm_set_config' used the argument 'config' as the network
context which is a pointer to a local variable of the function
'wpa_supplicant_rsn_supp_set_config'.
This is one reason why no proactive key was generated. This network
context never matched with the network context saved in the pmksa cache
entries.
The structure 'rsn_supp_config' has already a member 'network_ctx' which
is now filled in by this patch with 'ssid'.
Signed-off-by: Michael Bernhard <michael.bernhard@bfh.ch>
Just in case, do not use the not-yet-approved WEXT changes even if someone
where to build wpa_supplicant with IEEE 802.11w support unless this new
macro has been defined explicitly.
Added configuration of MFP related parameters with WEXT. The changes to
linux/wireless.h have not yet been applied to the Linux kernel tree, so the
code using them is still open to changes and is ifdef'ed out if
CONFIG_IEEE80211W is not set.
Add the support for the Linux wireless drivers which want to do
4-way handshake and need to know the PSK before the handshake.
Signed-off-by: Masakazu Mokuno <mokuno@sm.sony.co.jp>
Add the new flags which are supposed to be included in Linux 2.6.27
for the drivers which want to do 4-way handshake and to know PMK.
Signed-off-by: Masakazu Mokuno <mokuno@sm.sony.co.jp>
This avoids getting stuck in state where wpa_supplicant has canceled scans,
but the driver is actually in disassociated state. The previously used code
that controlled scan timeout from WPA module is not really needed anymore
(and has not been needed for past four years since authentication timeout
was separated from scan request timeout), so this can simply be removed to
resolved the race condition. As an extra bonus, this simplifies the
interface to WPA module.
[Bug 261]
driver_nl80211.c is based on driver_wext.c and it is still using Linux
wireless extensions for many functions. Over time, the new driver interface
code should be modified to use nl80211/cfg80211 for everything.
Don't cast pointers to int in definitions of PRISM2_HOSTAPD_RID_HDR_LEN
and PRISM2_HOSTAPD_GENERIC_ELEMENT_HDR_LEN. Use size_t instead. That's
actually what the code needs.
CONFIG_INTERNAL_LIBTOMMATH_FAST=y in .config can now be used to enable all
optimized routines at a cost of about 4 kB. This is small enough increase
in size to justify simplified configuration.
At the cost of about 1 kB of additional binary size, the internal
LibTomMath can be configured to include faster div routine to speed up DH
and RSA. This can be enabled with CONFIG_INTERNAL_LIBTOMMATH_FAST_DIV=y in
.config.
This gets rid of potential warnings about buffer bounds errors. The earlier
code works fine, but it is not the cleanest way of using the struct wpa_ptk
definition for TK1/TK2.
At the cost of about 0.5 kB of additional binary size, the internal
LibTomMath can be configured to include faster sqr routine to speed up DH
and RSA. This can be enabled with CONFIG_INTERNAL_LIBTOMMATH_FAST_SQR=y in
.config.
struct wpa_ie_hdr had separate fields for 24-bit OUI and 8-bit oui_type
for WPA/RSN selectors. {WPA,RSN}_SELECTOR_{GET,PUT} access these four
octets through oui and the "out-of-bounds" access for the fourth octet is
actually reading/writing oui_type. This works fine, but some tools complain
about the array bounds "failure". Since oui_type is never accessed
separately, the simplest fix is to just combine these into a single 4-octet
field.
Since mac80211 requires that the device be !IFF_UP to change the mode
(and I think the old prism54 fullmac driver does too), do that. This
shouldn't harm fullmac devices since they can handle mode switches on
the fly and usually don't care about up/down that much.
Add a cost of about 2.5 kB of additional cost, the internal LibTomMath can
be configured to include fast exptmod routine to speed up DH and RSA.
This can be enabled with CONFIG_INTERNAL_LIBTOMMATH_FAST_EXPTMOD=y in
.config.
In situations where the driver does background scanning and sends a
steady stream of scan results, wpa_supplicant would continually
reschedule the scan. This resulted in specific SSID scans never
happening for a hidden AP, and the supplicant never connecting to the AP
because it never got found. Instead, if there's an already scheduled
scan, and a request comes in to reschedule it, and there are enabled
scan_ssid=1 network blocks, let the scan happen anyway so the hidden
SSID has a chance to be found.
mac80211 sends _both_ channel and frequency in it's scan results, with
frequency first and channel second (it's since been fixed to send
channel first and frequency second to work around this issue). This
results in wpa_supplicant getting the right value when the frequency
comes, but overwriting the value with '0' when the channel comes because
wpa_supplicant can't handle 5GHz channel numbers. So if a valid
previous SIOCGIWFREQ event came in, don't try to overwrite it.
The internal TLS implementation can now use both PKCS #1 RSA private key
and PKCS #8 encapsulated RSA private key. PKCS #8 encrypted private key is
not yet supported.
The server handshake processing was still using SSL_read() to get OpenSSL
to perform the handshake. While this works for most cases, it caused some
issues for re-authentication. This is now changed to use SSL_accept() which
is more approriate here since we know that the handshake is still going on
and there will not be any tunneled data available. This resolves some of
the re-authentication issues and makes it possible for the server to notice
if TLS processing fails (SSL_read() did not return an error in many of
these cases while SSL_accept() does).
Set session id context to a unique value in order to avoid fatal errors
when client tries session resumption (SSL_set_session_id_context() must be
called for that to work), but disable session resumption with the unique
value for the time being since not all server side code is ready for it yet
(e.g., EAP-TTLS needs special Phase 2 processing when using abbreviated
handshake).
Changed EAP-TLS server not to call TLS library when processing the final
ACK (empty data) from the client in order to avoid starting a new TLS
handshake with SSL_accept().
Move the basic processing of received frames into eap_tls_common.c and use
callback functions to handle EAP type specific processing of the version
field and payload.
Fragmentation is now done as a separate step to clean up the design and to
allow the same code to be used in both Phase 1 and Phase 2. This adds
support for fragmenting EAP-PEAP/TTLS/FAST Phase 2 (tunneled) data.
Need to clear the state back to MSG after having processed all incoming
fragments. Without this, the server got stuck in sending the fragment ACK
even after having received the full message.
Even though we try to disable TLS compression, it is possible that this
cannot be done with all TLS libraries. For example, OpenSSL 0.9.8 does not
seem to have a configuration item for disabling all compression (0.9.9 has
such an option). If compression is used, Phase 2 decryption may end up
producing more data than the input buffer due to compressed data. This
shows up especially with EAP-TNC that uses very compressible data format.
As a workaround, increase the decryption buffer length to (orig_len+500)*3.
This is a hack, but at least it handles most cases. TLS compression should
really be disabled for EAP use of TLS, but since this can show up with
common setups, it is better to handle this case.
It looks like this SSL_set_options() value was added in 0.9.9 and it does
not exist in stable releases of OpenSSL. Fix build by using #ifdef on this
variable before use.
The middle byte of the secret (key for PRF) is shared with key halfs in
case the key length is odd. This does not happen in any of the current
tls_prf() uses, but it's better to fix this function to avoid future issues
should someone end up defining a use that uses an odd length for the key.
Process SoH Request in SoH EAP Extension method and reply with SoH TLV. The
contents of SSoH is not yet complete (i.e., some of the required attributes
are still missing). Use of SoH is disabled by default; it can be enabled
with tnc=soh in phase1 parameter for PEAP.
crypto_binding=# in phase1 network parameter can now be used to change the
PEAPv0 cryptobinding behavior (0 = do not use, 1 = optional (default),
2 = required).
Allow the user to set the IP address of the eapol_test client. This if
very useful when you have a machine with many interfaces and want to use a
particular one for testing RADIUS connectivity. For instance when I run the
national eduroam proxy I can only connect to other server from a particular
address, an our machine happens to have several IPs. So if I want to run
connectivity tests, I must make sure that my test uses a particular
interface. The -A option allows one to set this).
(jm: cleaned up to use radius configuration structure instead of global
variable for the address and added IPv6 support)
Implements the Chargable-User-Identity (CUI), as defined in RFC 4372.
Option "-i" causes the eapol_test to send a NUL CUI - which is a request to
send a CUI back. Capital "-I" allows to specify the value of the CUI.
This has been defined for cases where the client wants to reauthenticate.
If TNC is enabled, PEAPv0 server is now sending out SoH request to initiate
IF-TNCCS-SOH (Microsoft NAP / Statement of Health) handshake. The results
are currently only shown in debug log and they do not affect authentication
or authorization result.
The previous version was deriving unnecessarily large amount of output data
for the implicit challenge. With the TLS PRF, this was not causing any
problems because the output length is not explicitly bound into the
derivation. Anyway, it is better to use the correct output length should
the PRF be ever changed to a one that uses the output length to change the
beginning of the output data.