Commit graph

4885 commits

Author SHA1 Message Date
Veerendranath Jakkam
5fdacce465 Allow wildcard SSID to be enforced for a specific BSSID scan
Specific BSSID scan was replacing wildcard SSID with the known SSID if
any BSS with the specified BSSID is available in the known BSSes list.

Add control interface support to force use of a wildcard SSID in a
specific BSSID scan by user with the new "wildcard_ssid=1" argument to
the SCAN command.

Signed-off-by: Veerendranath Jakkam <vjakkam@codeaurora.org>
2020-03-02 20:33:52 +02:00
Pradeep Kumar Chitrapu
43282f7329 mesh: Fix HE enablement on 5 GHz with VHT
Incorrect he_enabled parameter was being passed to
hostapd_set_freq_params() in mesh which caused HE to be not fully
enabled on the 5 GHz band. Fix this by setting freq->he_enabled instead
of vht_freq.he_enabled so that the hostapd_set_freq_params() uses the
correct he_enabled value (and then ends up copying this to
vht_freq.he_enabled in the success case).

Fixes: 6e711e7ab3 ("mesh: Do not enable HE on 5 GHz without VHT")
Signed-off-by: Pradeep Kumar Chitrapu <pradeepc@codeaurora.org>
2020-03-02 19:54:35 +02:00
Sachin Shelke
21f835e640 SAE: Allow SAE-only network profile with sae_password to be written
The commit a34ca59e (SAE: Allow SAE password to be configured separately
(STA)) added sae_password configuration option. We should also consider
sae_password in the wpa_config_write() function which stores the valid
network block details to an external database.

Fixes: a34ca59e4d ("SAE: Allow SAE password to be configured separately (STA)")
Signed-off-by: Sachin Shelke <sachin.shelke@nxp.com>
Signed-off-by: Cathy Luo <xiaohua.luo@nxp.com>
Signed-off-by: Ganapathi Bhat <ganapathi.bhat@nxp.com>
2020-03-02 11:49:19 +02:00
Alexander Wetzel
c1a6b1e47e privsep: Add key_flag to set_key()
Pass through the new key_flag to wpa_priv.

Signed-off-by: Alexander Wetzel <alexander@wetzel-home.de>
2020-03-01 21:11:07 +02:00
Jouni Malinen
852d370f65 Silence a compiler warning in no-WEP and no-EAP builds
wep_keys_set was not used in wpas_start_assoc_cb() without
IEEE8021X_EAPOL, so need to make this local variable conditional on
build options.

Signed-off-by: Jouni Malinen <j@w1.fi>
2020-03-01 20:57:14 +02:00
Ilan Peer
c4988e73c0 driver: Extend send_mlme() with wait option
PASN authentication can be performed while a station interface is
connected to an AP. To allow sending PASN frames while connected, extend
the send_mlme() driver callback to also allow a wait option. Update the
relevant drivers and wpa_supplicant accordingly.

hostapd calls for send_mlme() are left unchanged, since the wait option
is not required there.

Signed-off-by: Ilan Peer <ilan.peer@intel.com>
2020-02-29 23:03:20 +02:00
Brian Norris
7a9c367225 DBus: Add "sae" to interface key_mgmt capabilities
This will be present when the driver supports SAE and it's included in
the wpa_supplicant build.

Signed-off-by: Brian Norris <briannorris@chromium.org>
2020-02-29 18:01:23 +02:00
Jouni Malinen
200c7693c9 Make WEP functionality an optional build parameter
WEP should not be used for anything anymore. As a step towards removing
it completely, move all WEP related functionality to be within
CONFIG_WEP blocks. This will be included in builds only if CONFIG_WEP=y
is explicitly set in build configuration.

Signed-off-by: Jouni Malinen <j@w1.fi>
2020-02-29 17:45:25 +02:00
Jouni Malinen
3fadb1dcc0 WPS: Ignore other APs if PBC is used with a specific BSSID
While the WSC specification requires the Enrollee to stop PBC
provisioning if the scan sees multiple APs in active PBC mode, this is
problematic due to some deployed devices continuing to advertise PBC
mode for extended duration (or even permanently). Such an environment
will still need to prevent wildcard AP selection with PBC since an
incorrect device could be selected. However, if the Enrollee device has
been explicitly requested to connect to a specific AP based on its
BSSID, the other APs in scan results can be ignored without affecting
which AP would be selected (only the one matching the specified BSSID is
acceptable).

Start filtering scan results for PBC session overlap check based on the
locally specified constraint on the BSSID, if one is set. This allows
PBC to be used with "WPS_PBC <BSSID>" command in environment where
another AP device is claiming to be in active PBC mode while "WPS_PBC"
command will still continue to reject provisioning since the correct AP
cannot be selected.

This will also cover the P2P cases where P2P_CONNECT is used to start or
authorize GO Negotiation and joining-a-GO with a specific P2P GO
Interface Address (BSSID).

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-02-24 19:53:49 +02:00
Jouni Malinen
0f84a93f65 Fix a type in wpa_supplicant defconfig
ap_mode=1 explanation in CONFIG_NO_SCAN_PROCESSING=y was really supposed
to be talking about ap_scan=1.

Signed-off-by: Jouni Malinen <j@w1.fi>
2020-02-24 00:05:20 +02:00
Alexander Wetzel
0e05e8781a Simplify wpa_deny_ptk0_rekey documentation
Signed-off-by: Alexander Wetzel <alexander@wetzel-home.de>
2020-02-23 23:59:58 +02:00
Janusz Dziedzic
12fb9698ab Use IFNAME= prefix for global UDP control interface events
There does not seem to be a good reason for using the different IFACE=
prefix on the UDP control interface. This got added when the UDP
interface in wpa_supplicant was extended in commit f0e5d3b5c6
("wpa_supplicant: Share attach/detach/send UDP ctrl_iface functions")
and that was then extended to hostapd in commit e920805685 ("hostapd:
Extend global control interface notifications").

Replace the IFACE= prefix in UDP case with IFNAME= to be consistent with
the UNIX domain socket based control interface.

This fixes a problem when at least one test case fail (hapd_ctrl_sta)
when remote/udp used. This also fixes test_connectivity().

Signed-off-by: Janusz Dziedzic <janusz.dziedzic@gmail.com>
2020-02-23 15:40:58 +02:00
Alexander Wetzel
293631f177 IBSS RSN: Coding style cleanup
Use consistent style for pointers.

Signed-off-by: Alexander Wetzel <alexander@wetzel-home.de>
2020-02-23 13:12:28 +02:00
Alexander Wetzel
1f90a49d02 STA: Allow PTK rekeying without Ext KeyID to be disabled as a workaround
Rekeying a pairwise key using only keyid 0 (PTK0 rekey) has many broken
implementations and should be avoided when using or interacting with
one. The effects can be triggered by either end of the connection and
range from hardly noticeable disconnects over long connection freezes up
to leaking clear text MPDUs.

To allow affected users to mitigate the issues, add a new configuration
option "wpa_deny_ptk0_rekey" to replace all PTK0 rekeys with fast
reconnects.

Signed-off-by: Alexander Wetzel <alexander@wetzel-home.de>
2020-02-23 13:05:19 +02:00
Jouni Malinen
e9ac44fcbc Make INTERWORKING_CONNECT more reliable in testing environment
Pick the most recently added BSS entry based on BSSID matching to avoid
issues in testing environment where the SSID of the AP may have changed
and both the old and new BSS is still present in the scan results.

Signed-off-by: Jouni Malinen <j@w1.fi>
2020-02-22 19:20:44 +02:00
Jouni Malinen
f3bcd69603 Remove CONFIG_IEEE80211N build option
Hardcoded CONFIG_IEEE80211N to be included to clean up implementation.
More or less all new devices support IEEE 802.11n (HT) and there is not
much need for being able to remove that functionality from the build.
Included this unconditionally to get rid of one more build options and
to keep things simpler.

Signed-off-by: Jouni Malinen <j@w1.fi>
2020-02-22 19:20:44 +02:00
Jouni Malinen
2d4c78aef7 Configure received BIGTK on station/supplicant side
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-02-18 00:18:47 +02:00
Jouni Malinen
ecbf59e693 wpa_supplicant configuration for Beacon protection
Add a new wpa_supplicant network profile configuration parameter
beacon_prot=<0/1> to allow Beacon protection to be enabled.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-02-18 00:18:47 +02:00
Didier Raboud
8155b36fae Fix VERSION_STR printf() calls in case the postfix strings include %
Do not use VERSION_STR directly as the format string to printf() since
it is possible for that string to contain '%'.

Signed-off-by: Didier Raboud <odyx@debian.org>
2020-02-17 19:17:10 +02:00
Jouni Malinen
466e48dcd7 HT: Remove SMPS in AP mode
SM Power Save was described in somewhat unclear manner in IEEE Std
802.11n-2009 as far the use of it locally in an AP to save power. That
was clarified in IEEE Std 802.11-2016 to allow only a non-AP STA to use
SMPS while the AP is required to support an associated STA doing so. The
AP itself cannot use SMPS locally and the HT Capability advertisement
for this is not appropriate.

Remove the parts of SMPS support that involve the AP using it locally.
In practice, this reverts the following commits:
04ee647d58 ("HT: Let the driver advertise its supported SMPS modes for AP mode")
8f461b50cf ("HT: Pass the smps_mode in AP parameters")
da1080d721 ("nl80211: Advertise and configure SMPS modes")

Signed-off-by: Jouni Malinen <j@w1.fi>
2020-02-16 13:58:54 +02:00
Jouni Malinen
b056275111 Fix exception checking in a wpa_supplicant P2P example script
Signed-off-by: Jouni Malinen <j@w1.fi>
2020-02-15 18:36:29 +02:00
Vamsi Krishna
75d0ec4702 P2P: Fix a possible buffer overflow in struct p2p_reg_class
Avoid adding more than P2P_MAX_REG_CLASSES operating classes or
P2P_MAX_REG_CLASS_CHANNELS channels while populating P2P channels. The
current limits on the operating classes or channels per operating class
could be hit in some case (mainly, with 6 GHz, but in theory, with a
2.4/5/60 GHz capable device as well).

If the local driver advertised a larger number of supported operarting
classes or channels per operating class, the construction of the struct
p2p_reg_class instances could have resulted in writing beyond the end of
the buffer and ending up corrupting memory around the struct p2p_config.
This could result in unexpected behavior in some other operations that
used corrupted memory, e.g., generation of a P2P Channel List failing
(with validation code stopping the process to avoid writing beyond the
end of the message buffer) due to not having sufficient buffer space for
the corrupted data.

This issue is triggered only based on information from the local driver
(mainly based on addition of support for 6 GHz band operating classes),
so the issue cannot be triggered based on received frames or any other
remote information.

The issue was introduced by commit d7c2c5c98c ("AP: Add initial
support for 6 GHz band") which added the operating class 131 which has
sufficiently large number of channels to go beyond the
P2P_MAX_REG_CLASS_CHANNELS limit.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-02-12 23:17:24 +02:00
Jouni Malinen
c304bddcf9 DPP: Stop Action frame sequence on DPP_STOP_LISTEN and PKEX failure
Previously it was possible for the PKEX/DPP exchange to terminate with
an error and the ongoing Action frame TX/RX offchannel operation not
getting terminated. This could leave the driver waiting on offchannel
until timeout and failing following operations before that timeout
happens. Fix this by explicitly stopping the Action frame sequence in
the driver in the previously missed cases.

This fixes a case that was showing up with the following test sequence
every now and then:
dpp_qr_code_chan_list_unicast dpp_pkex_test_fail dpp_enrollee_reject_config

dpp_pkex_test_fail was adding a large number of pending offchannel
operations and dpp_enrollee_reject_config could fail if those pending
operations were blocking new remain-on-channel or offchannel TX
operation for a sufficiently long time.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-02-11 07:09:47 +02:00
Jouni Malinen
c7cc80fbc8 DPP: Reset DPP_AUTH_INIT netrole back to STA by default
Previously DPP_AUTH_INIT command update wpa_s->dpp_netrole only if the
netrole parameter was included. This could leave AP or configurator
network in place for the next DPP_AUTH_INIT command. This would be
unexpected behavior, so reset wpa_s->dpp_netrole back to the
DPP_NETROLE_STA default if no explicit netrole parameter is included.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-02-11 06:33:24 +02:00
Veerendranath Jakkam
adf3de44ca Add check to consider band in enabling connection_vht flag
connection_vht flag was set to true when both Association Request and
Response frame IEs have VHT capability. Thus all devices that have
support for the vendor specific partial VHT support in the 2.4 GHz band
were also being reported as VHT capable. However, IEEE Std 802.11ac-2013
defines VHT STA to operate in frequency bands below 6 GHz excluding the
2.4 GHz band.

Do not set connection_vht when the operating band is 2.4 GHz. This
avoids reporting wifi_generation 5 on the 2.4 GHz band and reserves the
generation value 5 for full VHT as defined in the IEEE 802.11 standard.

Signed-off-by: Veerendranath Jakkam <vjakkam@codeaurora.org>
2020-02-11 05:19:16 +02:00
Matthew Wang
23dc196fde Check for FT support when selecting FT suites
A driver supports FT if it either supports SME or the
NL80211_CMD_UPDATE_FT_IES command. When selecting AKM suites,
wpa_supplicant currently doesn't take into account whether or not either
of those conditions are met. This can cause association failures, e.g.,
when an AP supports both WPA-EAP and FT-EAP but the driver doesn't
support FT (wpa_supplicant will decide to do FT-EAP since it is unaware
the driver doesn't support it). This change allows an FT suite to be
selected only when the driver also supports FT.

Signed-off-by: Matthew Wang <matthewmwang@chromium.org>
Reviewed-by: Brian Norris <briannorris@chromium.org>
2020-02-10 06:43:38 +02:00
Jouni Malinen
641d79f165 SAE: Special test mode sae_pwe=3 for looping with password identifier
The new sae_pwe=3 mode can be used to test non-compliant behavior with
SAE Password Identifiers. This can be used to force use of
hunting-and-pecking loop for PWE derivation when Password Identifier is
used. This is not allowed by the standard and as such, this
functionality is aimed at compliance testing.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-02-10 05:13:13 +02:00
Jouni Malinen
ca1cecc54d SAE: Verify that appropriate Status Code is used in SAE commit (SME)
Previous version accepted both 0 and 126 values in SAE commit message
from the AP. Explicitly check that the value the AP uses matches what
the STA started with to avoid unexpected cases.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-02-08 07:19:53 +02:00
Jouni Malinen
4dc3b70deb DPP: Try to negotiate different parameters if NFC handover fails
This is mainly for NFC testing purposes now since the own DPP parameters
are not yet updated.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-02-08 07:19:53 +02:00
Jouni Malinen
61c049da3d DPP: Allow local channel list to be set
The new dpp-nfc.py command line argument --chan can be used to replace
the local channel list default (81/1).

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-02-08 07:19:53 +02:00
Matthew Wang
fee28410de scan_est_throughput: Use ie_len instead of res->ie_len
Local variable should be used. This fixes an issue where IEs are
available only from a Beacon frame.

Fixes: ad06ac0b0 ("Move throughput estimation into a helper function")
Signed-off-by: Matthew Wang <matthewmwang@chromium.org>
2020-02-02 21:55:42 +02:00
Jouni Malinen
7d9e320054 DPP: Received Configurator backup processing
Add local Configurator instance for each received Configurator backup.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-31 23:16:05 +02:00
Jouni Malinen
312eac1d1e DPP: Add ASN.1 support into build
This will be needed in following patches to process DPPEnvelopedData.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-31 23:16:05 +02:00
Jouni Malinen
703c2b6457 DPP: Example script for NFC bootstrapping method
This Python script is an example on how nfcpy can be used to drive an
NFC Device to perform DPP bootstrapping operations over DPP (tag with
NFC URI and negotiated connection handover).

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-27 21:58:45 +02:00
Jouni Malinen
5e287724ee DPP: NFC negotiated connection handover
Add new control interface commands "DPP_NFC_HANDOVER_REQ own=<id>
uri=<URI>" and "DPP_NFC_HANDOVER_SEL own=<id> uri=<URI>" to support NFC
negotiated connection handover. These commands are used to report a DPP
URI received from a peer NFC Device in Handover Request and Handover
Select messages. The commands return peer bootstrapping information ID
or FAIL on failure. The returned ID is used similarly to any other
bootstrapping information to initiate DPP authentication.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-27 20:36:09 +02:00
Jouni Malinen
8b138d2826 OWE: PTK derivation workaround in STA mode
Initial OWE implementation used SHA256 when deriving the PTK for all OWE
groups. This was supposed to change to SHA384 for group 20 and SHA512
for group 21. The new owe_ptk_workaround=1 network parameter can be used
to enable older behavior mainly for testing purposes. There is no impact
to group 19 behavior, but if enabled, this will make group 20 and 21
cases use SHA256-based PTK derivation which will not work with the
updated OWE implementation on the AP side.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-24 00:47:41 +02:00
Vamsi Krishna
d0e116f61f Enhance get_mode() to return correct hw_mode with 6 GHz support
The 5 GHz channels are stored in one hw_features set with mode
HOSTAPD_MODE_IEEE80211A while the 6 GHz channels will need to be stored
in a separate hw_features set (but with same mode
HOSTAPD_MODE_IEEE80211A) due to possibility of different HT/VHT/HE
capabilities being available between the 5 GHz and 6 GHz bands.

Iterate through all hw_features sets and check and match the band of
channel supported by the hw_features set while getting the hw_features
set in get_mode(). This allows both the 5 GHz and 6 GHz channels to be
found and correct capabilities to be used in cases where the driver
reports different capability values between 5 and 6 GHz channels.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-23 15:48:51 +02:00
Chaitanya Tata
4658eb77d6 Remove deprecated text for ap_scan=0
Users might be tempted to try ap_scan=0 for offloading scan,
ap_selection and, WPA to driver. Update the text to reflect that this is
deprecated.

Jouni confirmed deprecation in
https://www.spinics.net/lists/hostap/msg06482.html

Signed-off-by: Chaitanya Tata <chaitanya.tata@bluwireless.com>
2020-01-21 18:17:07 +02:00
Jouni Malinen
5e32fb0170 SAE: Use Anti-Clogging Token Container element with H2E
IEEE P802.11-REVmd was modified to use a container IE for anti-clogging
token whenver H2E is used so that parsing of the SAE Authentication
frames can be simplified.

See this document for more details of the approved changes:
https://mentor.ieee.org/802.11/dcn/19/11-19-2154-02-000m-sae-anti-clogging-token.docx

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-21 13:13:56 +02:00
Jouni Malinen
e36a5894d0 SAE: Use H2E whenever Password Identifier is used
IEEE P802.11-REVmd was modified to require H2E to be used whenever
Password Identifier is used with SAE.

See this document for more details of the approved changes:
https://mentor.ieee.org/802.11/dcn/19/11-19-2154-02-000m-sae-anti-clogging-token.docx

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-21 13:13:56 +02:00
Jouni Malinen
29dd0b3164 SAE H2E: Check H2E-only BSS membership selector only if SAE is enabled
This BSS membership selector has impact only for SAE functionality, so
ignore it when configured not to use SAE. This allows WPA-PSK connection
to and AP that advertises WPA-PSK and SAE while requiring H2E for SAE.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-20 21:17:28 +02:00
Jouni Malinen
dd530b8739 Silence compiler warning with CONFIG_NO_ROAMING=y
Comment out unused static functions if CONFIG_NO_ROAMING is defined.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-09 12:40:17 +02:00
Alexander Wetzel
a919a26035 Introduce and add key_flag
Add the new set_key() parameter "key_flag" to provide more specific
description of what type of a key is being configured. This is needed to
be able to add support for "Extended Key ID for Individually Addressed
Frames" from IEEE Std 802.11-2016. In addition, this may be used to
replace the set_tx boolean eventually once all the driver wrappers have
moved to using the new key_flag.

The following flag are defined:

  KEY_FLAG_MODIFY
    Set when an already installed key must be updated.
    So far the only use-case is changing RX/TX status of installed
    keys. Must not be set when deleting a key.

  KEY_FLAG_DEFAULT
    Set when the key is also a default key. Must not be set when
    deleting a key. (This is the replacement for set_tx.)

  KEY_FLAG_RX
    The key is valid for RX. Must not be set when deleting a key.

  KEY_FLAG_TX
    The key is valid for TX. Must not be set when deleting a key.

  KEY_FLAG_GROUP
    The key is a broadcast or group key.

  KEY_FLAG_PAIRWISE
    The key is a pairwise key.

  KEY_FLAG_PMK
    The key is a Pairwise Master Key (PMK).

Predefined and needed flag combinations so far are:

  KEY_FLAG_GROUP_RX_TX
    WEP key not used as default key (yet).

  KEY_FLAG_GROUP_RX_TX_DEFAULT
    Default WEP or WPA-NONE key.

  KEY_FLAG_GROUP_RX
    GTK key valid for RX only.

  KEY_FLAG_GROUP_TX_DEFAULT
    GTK key valid for TX only, immediately taking over TX.

  KEY_FLAG_PAIRWISE_RX_TX
    Pairwise key immediately becoming the active pairwise key.

  KEY_FLAG_PAIRWISE_RX
    Pairwise key not yet valid for TX. (Only usable with Extended Key ID
    support.)

  KEY_FLAG_PAIRWISE_RX_TX_MODIFY
    Enable TX for a pairwise key installed with KEY_FLAG_PAIRWISE_RX.

  KEY_FLAG_RX_TX
    Not a valid standalone key type and can only used in combination
    with other flags to mark a key for RX/TX.

This commit is not changing any functionality. It just adds the new
key_flag to all hostapd/wpa_supplicant set_key() functions without using
it, yet.

Signed-off-by: Alexander Wetzel <alexander@wetzel-home.de>
2020-01-09 12:38:36 +02:00
Jouni Malinen
f822546451 driver: Move set_key() parameters into a struct
This makes it more convenient to add, remove, and modify the parameters
without always having to update every single driver_*.c implementation
of this callback function.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-09 00:48:57 +02:00
Vamsi Krishna
33c8a10498 Do not select APs found on disabled channels for connection
If a channel list changed event is received after a scan and before
selecting a BSS for connection, a BSS found on a now disabled channel
may get selected for connection. The connect request issued with the BSS
found on a disabled channel is rejected by cfg80211. Filter out the BSSs
found on disabled channels and select from the other BSSs found on
enabled channels to avoid unnecessary connection attempts that are bound
to fail.

The channel list information will be updated by the driver in cases like
country code update, disabling/enabling specific bands, etc. which can
occur between the scan and connection attempt.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2020-01-08 16:19:47 +02:00
Markus Theil
8296ee1805 RSN IBSS: Fix EAPOL TX using control port
This was previously done only in supplicant role, but a similar change
is needed for the authenticator role.

Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
2020-01-06 22:33:45 +02:00
Markus Theil
a79ed06871 Add no_encrypt flag for control port TX
In order to correctly encrypt rekeying frames, wpa_supplicant now checks
if a PTK is currently installed and sets the corresponding encrypt
option for tx_control_port().

Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
2020-01-05 20:34:50 +02:00
Brendan Jackman
144314eaa7 wpa_supplicant: Send EAPOL frames over nl80211 where available
Linux kernel v4.17 added the ability to request sending control port
frames via nl80211 instead of a normal network socket. Doing this
provides the device driver with ordering information between the
control port frames and the installation of keys. This empowers it to
avoid race conditions between, for example, PTK replacement and the
sending of frame 4 of the 4-way rekeying handshake in an RSNA. The
key difference between a TX_CONTROL_PORT and normal socket send is
that the device driver will certainly get any EAPOL frames comprising
a 4-way handshake before it gets the key installation call
for the derived key. By flushing its TX buffers it can then ensure
that no pending EAPOL frames are inadvertently encrypted with a key
that the peer will not yet have installed.

Update the RSN supplicant system to use this new operation for sending
EAPOL-Key frames when the driver reports that this capability is
available; otherwise, fall back to a normal Ethernet TX.

I have tested this on DMG (11ad/ay) devices with an out-of-tree Linux
driver that does not use mac80211. Without this patch I consistently see
PTK rekeying fail if message 4/4 shares a stream with other in-flight
traffic. With this patch, and the driver updated to flush the relevant TX
queue before overwriting a PTK (knowing, now, that if there was a message
4/4 related to the key installation, it has already entered the driver
queue), rekeying is reliable.

There is still data loss surrounding key installation - this problem is
alluded to in IEEE Std 802.11-2016, 12.6.21, where extended Key ID
support is described as the eventual solution. This patch aims to at
least prevent rekeying from totally breaking the association, in a way
that works on kernels as far back as 4.17 (as per Alexander Wetzel
extended Key ID support should be possible on 5.2).

See http://lists.infradead.org/pipermail/hostap/2019-May/040089.html for
a little more context.

Signed-off-by: Brendan Jackman <brendan.jackman@bluwireless.co.uk>
2020-01-05 20:34:15 +02:00
Brendan Jackman
8759e9116a nl80211: Control port over nl80211 helpers
Linux kernel v4.17 added the ability to request sending controlled port
frames (e.g., IEEE 802.1X controlled port EAPOL frames) via nl80211
instead of a normal network socket. Doing this provides the device
driver with ordering information between the control port frames and the
installation of keys. This empowers it to avoid race conditions between,
for example, PTK replacement and the sending of frame 4 of the 4-way
rekeying handshake in an RSNA. The key difference between the specific
control port and normal socket send is that the device driver will
certainly get any EAPOL frames comprising a 4-way handshake before it
gets the key installation call for the derived key. By flushing its TX
buffers it can then ensure that no pending EAPOL frames are
inadvertently encrypted with a key that the peer will not yet have
installed.

Add a CONTROL_PORT flag to the hostap driver API to report driver
capability for using a separate control port for EAPOL frames. This
operation is exactly like an Ethernet send except for the extra ordering
information it provides for device drivers. The nl80211 driver is
updated to support this operation when the device reports support for
NL80211_EXT_FEATURE_CONTROL_PORT_OVER_NL80211. Also add a driver op
tx_control_port() for request a frame to be sent over the controlled
port.

Signed-off-by: Brendan Jackman <brendan.jackman@bluwireless.co.uk>
2020-01-05 19:43:52 +02:00
Jouni Malinen
b7bb2c0204 P2P: Move p2p_long_listen into struct wpa_global
This variable is not specific to any P2P group interface and since it
was already used through global->p2p_init_wpa_s, it is cleaner to simply
move this to the global structure so that there is a single variable
instead of per-interface variables and need to pick the correct
interface.

Signed-off-by: Jouni Malinen <j@w1.fi>
2020-01-04 23:11:28 +02:00