Commit graph

17118 commits

Author SHA1 Message Date
Sreeramya Soratkal
9b50746f50 P2P: Introduce 6 GHz band capability bit in P2P Device Capability
Introduce P2P 6 GHz band capable information in Device Capability
Bitmap of P2P Capability sub-attribute.

Signed-off-by: Sreeramya Soratkal <ssramya@codeaurora.org>
2021-06-10 23:08:22 +03:00
Gurumoorthi Gnanasambandhan
9f901e65b4 WNM: Ignore SSID check for hidden SSID in transition candidates
Do not skip scan results with zero length SSID (i.e., a hidden SSID)
when searching for potential BSS transition candidates since such
entries might be for the same ESS (i.e., for the current SSID). Use only
the BSSID check for such cases.

Signed-off-by: Gurumoorthi Gnanasambandhan <gguru@codeaurora.org>
2021-06-09 20:55:39 +03:00
Jouni Malinen
525ec045f3 P2P: Use correct return type for has_channel()
This helper function returns enum chan_allowed values, so use it as the
return type instead of unnecessarily generic int.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-06-08 12:46:45 +03:00
Hu Wang
09fb9b0cb0 DFS offload: Use hostapd_is_dfs_required() to check if DFS required
hostapd_handle_dfs_offload() is the DFS handler for the offloaded case,
in which ieee80211_is_dfs() is used to check if the configured frequency
requires DFS or not.

When the configured channel width is not 20 (e.g., 160),
ieee80211_is_dfs() will not checked adjacent freqs, so it possibly makes
wrong conclusion for whether DFS is required.

hostapd_is_dfs_required() does similar thing with ieee80211_is_dfs()
except it supports checking whether the configured frequency and its
adjacent frequencies require DFS. So hostapd_is_dfs_required() is a more
robust and better option than ieee80211_is_dfs() to check DFS.

The issue is hostapd_is_dfs_required() is for non-offload case due to
the check of the configuration parameter ieee80211h. Add a check for
WPA_DRIVER_FLAGS_DFS_OFFLOAD to make it support the DFS offload case
(i.e., ieee80211h=0) as well.

For example, configuring the AP to start at freq=5240 with channel width
160:
- Existing hostapd checks freq=5240 is non-DFS, hence skip DFS CAC and
  transition to AP-Enabled which volatiles DFS-RADAR detection.

  LOG: "hostapd : hostapd_handle_dfs_offload: freq 5240 MHz does not
        require DFS. Continue channel/AP setup"

- This commit checks freq=5240 and its adjacent freqs are DFS required,
  hence remains in DFS state until DFS CAC completed.

  LOG: "hostapd : hostapd_handle_dfs_offload: freq 5240 MHz requires
        DFS for 4 chans"

Signed-off-by: Hu Wang <huw@codeaurora.org>
2021-06-07 17:42:56 +03:00
Jouni Malinen
e8662e9d44 Use a helper function to remove struct wpa_bss_tmp_disallowed entries
It is safer to remove and free these entries with a shared helper
function to avoid issues with potentially forgetting to unregister or
free something if this structure is extended in the future.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-06-03 00:11:18 +03:00
Hu Wang
ecaacb47b7 OCE: Remove AP from driver disallow list with sufficient AP RSSI
When a STA makes an association request that is rejected by an OCE AP
due to the RSSI being insufficient, the AP is added to the driver
disallow list by wpa_set_driver_tmp_disallow_list().

Once the AP increases TX power which makes the AP RSSI higher than
Association Rejection RSSI threshold, the AP is supposed to be removed
from the driver disallow list but that was not the case.

wpa_is_bss_tmp_disallowed() is called in the scan result handler, so it
is the best place to put the logic of removing the AP from the driver
disallow list with sufficient AP RSSI.

This is needed with drivers that use the temporarily disallowed BSS list
(which is currently supported only with a QCA vendor command). The
wpa_supplicant internal functionality was already taking care of this
with the wpa_is_bss_tmp_disallowed() return value even for cases where
the entry remaining in the list.

Signed-off-by: Hu Wang <huw@codeaurora.org>
2021-06-03 00:06:00 +03:00
Hu Wang
c25b50306e hostapd: Reject 40 MHz channel config if regulatory rules do not allow it
When regulatory rules are configured not to support 40 MHz channels on
the 2.4 GHz band, hostapd_is_usable_chans() still allowed 40 MHz
channels (i.e., 1-9) to be used with ht_capab=[HT40+][HT40-].

Looking into hostapd_is_usable_chans():
1) Validate primary channel using hostapd_is_usable_chan()
2) Try to pick a default secondary channel if hostapd_is_usable_chan()
3) Try to pick a valid secondary channel if both HT40+/HT40- set, and
   further validate primary channel's allowed bandwidth mask.
4) Return channel not usable.

For example, for the 2.4 GHz channel 9 in Japan, its default secondary
channel is 13, which is valid per hostapd_is_usable_chan(), so step (2)
returns channel usable.

Add a more strict check to step (2) to clearly reject 40 MHz channel
configuration if regulatory rules do not allow the 40 MHz bandwidth,
which is similarly done in commit ce6d9ce15b ("hostapd: Add supported
channel bandwidth checking infrastructure").

Signed-off-by: Hu Wang <huw@codeaurora.org>
2021-06-02 00:06:20 +03:00
Mohammad Asaad Akram
20a522b9eb AP: Add user configuration for TWT responder role
Add user configuration he_twt_responder for enabling/disabling TWT
responder role, in addition to checking the driver's capability. The
default configuration is to enable TWT responder role when the driver
supports this.

Signed-off-by: Mohammad Asaad Akram <asadkrm@codeaurora.org>
2021-06-01 00:17:03 +03:00
Mohammad Asaad Akram
9efed66845 Android: Pass the vendor events to $(BOARD_HOSTAPD_PRIVATE_LIB)
Introduce a new board configuration via
$(BOARD_HOSTAPD_PRIVATE_LIB_EVENT) rather than reusing
$(BOARD_HOSTAPD_PRIVATE_LIB) to pass vendor events handling in the
hostapd private library. This is to avoid compilation issues for
wpa_driver_nl80211_driver_event() with the already existing private
library implementations defined with $(BOARD_HOSTAPD_PRIVATE_LIB).

This is similar to the existing BOARD_WPA_SUPPLICANT_PRIVATE_LIB_EVENT
parameter for the wpa_supplicant build.

Signed-off-by: Mohammad Asaad Akram <asadkrm@codeaurora.org>
2021-05-27 20:58:09 +03:00
Vamsi Krishna
39a1d55b76 Add QCA vendor interface to transport CFR data using netlink events
Add QCA vendor interface to configure the driver which transport mode to
use for sending CFR data to userspace. Currently, relayfs and netlink
event modes are supported.

Signed-off-by: Vamsi Krishna <vamsin@codeaurora.org>
2021-05-27 20:43:45 +03:00
Kiran Kumar Lokere
da3335c925 QCA vendor attribute to configure keep alive data type
Add a QCA vendor attribute to configure the driver to use Data or
Management frames for keep alive data. This attribute is used for
testing purpose.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-27 20:31:28 +03:00
Kiran Kumar Lokere
98f1259cd1 QCA vendor attribute to configure ER SU PPDU type
Add QCA vendor attribute to configure the driver to transmit the
Data frames with ER SU PPDU type format. This attribute is used
for testing purpose.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-27 20:29:43 +03:00
Jouni Malinen
b102f19bcc tests: Opportunistic Wireless Encryption - SA Query
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-22 00:18:57 +03:00
Jouni Malinen
d98f4c4660 tests: FILS SK and opportunistic key caching
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-21 21:07:28 +03:00
Veerendranath Jakkam
e2e2655ce8 FILS: Fix PMKID derivation for OKC
FILS authentication derives PMK differently from the EAP cases. The PMK
value does not bind in the MAC addresses of the STAs. As such, the same
PMKID is used with different BSSIDs. Fix both the hostapd and
wpa_supplicant to use the previous PMKID as is for OKC instead of
deriving a new PMKID using an incorrect derivation method when using an
FILS AKM.

Signed-off-by: Veerendranath Jakkam <vjakkam@codeaurora.org>
2021-05-21 21:05:41 +03:00
Vamsi Krishna
6abfb1418c Use estimated throughputs irrespective of RSSI delta for 6 GHz APs
APs in 6 GHz operating with LPI/VLP rules will have significantly lower
SNR values compared to 2.4/5 GHz band APs. Earlier, the estimated
throughputs were used for comparison only when the delta of SNRs between
both the APs was not greater than 7 and as a result for comparing 6 GHz
APs with 2.4/5 GHz APs, estimated throughputs were not getting used.

The estimated throughput calculations takes SNR value also into
consideration, hence remove RSSI delta check if any of the APs are from
the 6 GHz band. This change is limited to the 6 GHz band only in order
to avoid possible regressions with 2.4/5 GHz APs.

Signed-off-by: Vamsi Krishna <vamsin@codeaurora.org>
2021-05-21 18:41:59 +03:00
Vinita S. Maloo
5e1e4cceb6 Add QCA vendor attribute to enable/disable FT over DS
Add QCA vendor attribute QCA_WLAN_VENDOR_ATTR_CONFIG_FT_OVER_DS
to configure FT over DS to the driver/firmware.

Signed-off-by: Vinita S. Maloo<vmaloo@codeaurora.org>
2021-05-21 18:21:07 +03:00
Vamsi Krishna
586afb8fa8 Add QCA interface to configure band specific RSSI thresholds for roaming
Add QCA interface to specify the RSSI thresholds separately for candidate
APs from different bands.

Signed-off-by: Vamsi Krishna <vamsin@codeaurora.org>
2021-05-21 18:15:57 +03:00
Mohammad Asaad Akram
800e345502 Fix documentation for QCA_WLAN_VENDOR_ATTR_TWT_SETUP_MAC_ADDR
Mention that for AP mode this attribute is required in
response for TWT SET, TWT GET, TWT SUSPEND, and TWT
TERMINATE. And is required in request for TWT GET, and
TWT TERMINATE.

For STA mode, the usage of this attribute remains unchanged.

Signed-off-by: Mohammad Asaad Akram <asadkrm@codeaurora.org>
2021-05-21 17:59:48 +03:00
Jouni Malinen
7d2302f878 Add EAPOL-4WAY-HS-COMPLETED indication to AP
This makes it easier for test scripts to track completion of 4-way
handshake from hostapd, e.g., when going through PTK rekeying.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-11 21:13:56 +03:00
Jouni Malinen
1c5aa2579d Add EAPOL_TX command to extend ext_eapol_frame_io possibilities
This makes it convenient for an external test script to use
ext_eapol_frame_io=1 to delay and/or modify transmission of EAPOL-Key
msg 1/4 without having to use separate frame injection mechanisms.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-11 21:13:56 +03:00
Jouni Malinen
7f0a2e4225 Report EAPOL-RX events for testing purposes
This makes it more convenient to track EAPOL frame reception from an
external test script.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-11 21:13:56 +03:00
Jouni Malinen
04283cf36b Add REKEY_PTK to allow upper layer request to force PTK rekeying
"REKEY_PTK <STA MAC address>" can now be used to force rekeying of the
PTK for the specified associated STA.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-11 21:13:56 +03:00
Jouni Malinen
ced15c8ba8 wlantest: TKIP frame reassembly for Michael MIC check in fragmented case
Reassemble the full MSDU when processing TKIP protected fragmented
frames so that the Michael MIC can be validated once the last fragment
has been received.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-11 21:13:56 +03:00
Jouni Malinen
3332657d69 wlantest: Report decrypted TKIP frames even if cannot check Michael MIC
This can be useful for debugging, so return successfully decrypted TKIP
frame even if the Michael MIC cannot be verified (fragment reassembly
not yet supported) or if the Michael MIC value is incorrect. Add a note
in the frame to point out that the Michael MIC was not verified or is
incorrect.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-11 21:13:56 +03:00
Jouni Malinen
81169ebc48 tests: WPA2-PSK AP and GTK rekey request from multiple stations
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-11 01:05:22 +03:00
Jouni Malinen
82d8d631ec Skip GTK rekeying request if rekeying already in process
Do not start yet another rekeying of GTK when receiving an EAPOL-Key
request frame at the point when the GTK is already being rekeyed. This
fixes issues where the AP might end up configuring a different GTK than
the one it sends to the associated stations.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-11 01:03:57 +03:00
Hu Wang
de4d62dbcd Add QCA vendor definitions for DFS radar history
Introduce definitions for QCA vendor specific subcommands and attributes
to report DFS radar history.

Signed-off-by: Hu Wang <huw@codeaurora.org>
2021-05-07 13:10:52 +03:00
Jouni Malinen
2b6915a139 tests: Extend estimated throughput testing
Cover VHT160 and HE cases.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-07 13:05:58 +03:00
Vamsi Krishna
46f8976196 Prefer 6 GHz APs for connection in BSS selection
Prefer 6 GHz APs when estimated throughputs are equal with APs from the
2.4/5 GHz bands while selecting APs for connection. Also add a 6 GHz
specific noise floor default value for the 6 GHz band (with the same
value as was used for 5 GHz previously) to make this step clearer.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-07 13:05:58 +03:00
Vamsi Krishna
84008457ed Add support to calculate estimated throughputs for HE rates
Add support to consider HE rates while estimating throughputs for the
scan results from HE enabled APs. HE 0.8 usec GI rates are used in all
tables. The minimum SNR values for HE rates (1024-QAM) are derived by
adding the existing minimum SNR values of 256-QAM rates from VHT tables
and the difference between the values of minimum sensitivity levels of
256-QAM rates and 1024-QAM rates defined in Table 27-51 (Receiver
minimum input level sensitivity) in IEEE P802.11ax/D8.0.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-07 13:05:58 +03:00
Vamsi Krishna
658b6a0b08 Add support to estimate throughput for VHT 160/80+80 MHz supporting APs
Add support to calculate estimated throughputs for APs which support the
160 MHz (including 80+80 MHz) mode in VHT. The minimum SNR values for
VHT 160 MHz mode are derived from minimum SNR values used for VHT 80 MHz
mode + 3 dBm. The min-SNR values are derived relatively based on the
information that the minimum sensitivity levels defined in Table 21-25
(Receiver minimum input level sensitivity) in IEEE Std 802.11-2020 for
the 160 MHz mode are higher by 3 dBm compared to the values of the 80
MHz mode for each rate.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-07 12:28:46 +03:00
Vamsi Krishna
1d2118b509 Check local supported features for estimating BSS throughputs accurately
Add checks for features supported by the specific hardware mode of the
local device that has the channel for which the throughput is being
estimated instead of assuming the local device supports all optional
features. This is more accurate for cases where the local capabilities
might differ based on the band. In addition, this is in preparation for
extending rate estimates to cover optional VHT and HE features.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-07 12:27:21 +03:00
Jouni Malinen
3cefdf0d4a tests: Fix eap_proto_sake_server with Session ID = 255
Need to wrap back to 0 when changing value 255 to avoid generating a too
large value to fit an octet field. This was resulting in errors due to a
python exception (likely for about every 256th run).

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-06 23:49:12 +03:00
Jouni Malinen
73f65cc6c4 wlantest: Support HT Control field in QoS Data frames
Extend Data frame processing (and decryption) to handle +HTC frames by
skipping the HT Control field at the end of the frame header. While this
is not an exact match of the rules in IEEE Std 802.11-2020 for when the
HT Control field is present in frames (e.g., no check of the TXVECTOR
value), this is good enough to cover the most likely used cases.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-06 23:32:54 +03:00
Jouni Malinen
2950851ace Rename the Frame Control field subfield Order define to +HTC
This moves the implementation closer to the current IEEE 802.11 standard
since B15 of Frame Control field was renamed to +HTC to match it newer
uses.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-06 12:12:51 +03:00
Vamsi Krishna
11821ab3d1 Add QCA vendor interface to query usable channels
Add QCA vendor interface for userspace to get information of usable
channels for different interface types from the driver/firmware.

Signed-off-by: Vamsi Krishna <vamsin@codeaurora.org>
2021-05-04 22:39:44 +03:00
Jouni Malinen
fd4580e1e8 tests: Limit maximum number of frames in fuzzing tests
This limits the EAP-SIM and EAP-AKA tests to 100 test frames to avoid
undesired timeouts in automated fuzz testing. The real world uses are
limited to 50 rounds, so there is not really any point in trying with
thousands of frames.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-04 18:19:23 +03:00
Pradeep Kumar Chitrapu
6ae0d78b8e Determine 6 GHz bandwidth in AP mode ACS using op_class parameter
Determine bandwidth from op_class parameter when set in config. When not
configured, use he_oper_chwidth for determining 80 MHz or 160 MHz. When
both are not set, fall back to 20 MHz by default. This helps in removing
the dependency on op_class parameter in 6 GHz ACS.

Signed-off-by: Pradeep Kumar Chitrapu <pradeepc@codeaurora.org>
2021-05-03 23:24:13 +03:00
Pradeep Kumar Chitrapu
0822de037c Add AP mode ACS support for the 6 GHz band
Add support for the 6 GHz frequencies using 40, 80, and 160 MHz
bandwidths in the AP mode ACS.

Signed-off-by: Pradeep Kumar Chitrapu <pradeepc@codeaurora.org>
2021-05-03 23:22:25 +03:00
Pradeep Kumar Chitrapu
bef5eee4f7 Convert channel to frequency based selection for AP mode ACS
Convert channel based selection to frequency based selection for AP mode
ACS to accommodate for the 6 GHz band needs.

Signed-off-by: Pradeep Kumar Chitrapu <pradeepc@codeaurora.org>
2021-05-03 23:19:03 +03:00
Pradeep Kumar Chitrapu
15742566fd 6 GHz: Fix operating class in Supported Operating Classes element
Previously, the secondary channel was set only in presence of HT
capabilities based on HT40+ or HT40-. As HT capabilities and
secondary_channel are not present for the 6 GHz bamd, this causes
incorrect operating class indication in the Supported Operating Classes
element.

Fix this by assigning the secondary channel for bandwidths greater than
20 MHz in the 6 GHz band.

Signed-off-by: Pradeep Kumar Chitrapu <pradeepc@codeaurora.org>
2021-05-03 23:04:55 +03:00
Jouni Malinen
6162890798 tests: Update server and user certificates (2020)
The previous versions expired, so need to re-sign these to fix number of
the EAP test cases. This contains updates from running
tests/hwsim/auth_server/update.sh.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-05-03 20:03:28 +03:00
P Praneesh
79e8f0c164 hostapd: Update 160 MHz center freq calculation in 6 GHz
In the 6 GHz Operation Information field, the Channel Center Frequency
Segment 0 field indicates the channel center frequency index for
the 20 MHz, 40 MHz, 80 MHz, or 80+80 MHz channel on which the
BSS operates in the 6 GHz band. If the BSS channel width is 160 MHz
then the Channel Center Frequency Segment 0 field indicates the
channel center frequency index of the primary 80 MHz.

The Channel Center Frequency Segment 1 field indicates the channel
center frequency index of the 160 MHz channel on which the BSS operates
in the 6 GHz band or the channel center frequency of the secondary 80
MHz for the 80+80 MHz channel.

Since Channel Center Frequency Segment 1 was 0 for 160 MHz, 6 GHz STA
associated using 80 MHz. Update seg0 and seg1 fields per standard (IEEE
P802.11ax/D8.0: 9.4.2.249 HE Operation element).

Signed-off-by: P Praneesh <ppranees@codeaurora.org>
2021-05-03 17:35:59 +03:00
Lavanya Suresh
9c6b0a9416 hostapd: Disable VHT/HE when WMM is not enabled
When WMM is disabled, HT/VHT/HE capabilities should not be used for any
STA. If any STA advertises these capabilities, hostapd AP disables HT
capabilities in STA flags during STA assoc, but VHT/HE was not handled
similarly. This could allow a STA to associate in VHT/HE mode even in
WMM disable case.

To avoid this, disable VHT/HE capabilities similarly to HT during STA
association, if WMM is not enabled by the STA.

Signed-off-by: Lavanya Suresh <lavaks@codeaurora.org>
2021-05-03 17:27:22 +03:00
Jouni Malinen
15b1831a2c nl80211: Map internal TDLS_PEER_* to NL80211_TDLS_PEER_*
Even though these enum definitions are currently identical, it is better
to explicitly map these bits to the kernel interface instead of using
the internal definition for this. This makes it much clearer that new
enum tdls_peer_capability value needs to be assigned in nl80211 before
they can be added into wpa_supplicant.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-04-23 12:40:07 +03:00
Sreeramya Soratkal
742018f44d Add support to indicate TDLS peer's HE capability to driver
Indicate TDLS peer's capability to driver after processing TDLS setup
response frame. This information can be used by the driver to decide
whether to include HE operation IE in TLDS setup confirmation frame.

Signed-off-by: Sreeramya Soratkal <ssramya@codeaurora.org>
2021-04-23 12:32:37 +03:00
Jouni Malinen
2be5777a99 Sync with mac80211-next.git include/uapi/linux/nl80211.h
This brings in nl80211 definitions as of 2021-04-19.

Signed-off-by: Jouni Malinen <j@w1.fi>
2021-04-23 12:31:50 +03:00
Jouni Malinen
d5ce180a60 tests: UPDATE_BEACON on disabled interface
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2021-04-21 23:24:58 +03:00
Kani M
1f2fbf41d0 Fix UPDATE_BEACON processing when disabled
The hostapd process crashed when the UPDATE_BEACON control interface
command was issue after the interface was disabled. Check for this case
and return an error if the interface is disabled.

Signed-off-by: Kani M <kanisumi@codeaurora.org>
2021-04-21 23:23:54 +03:00