Jouni Malinen [Tue, 5 Feb 2019 18:36:28 +0000 (20:36 +0200)]
nl80211: Clear PMKID add command message buffer
This command has now been extended to include PMK for offload needs, so
the message buffer needs to be cleared explicitly after use to avoid
leaving such material in heap memory unnecessarily.
Fixes: 061a3d3d5300 ("nl80211: Add support for FILS Cache Identifier in add/remove_pmkid()") Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Jouni Malinen [Tue, 5 Feb 2019 18:34:34 +0000 (20:34 +0200)]
nl80211: Clear connect command message buffer
This command can include keys (WEP or PSK for offload), so the message
buffer needs to be cleared explicitly after use to avoid leaving such
material in heap memory unnecessarily.
Jouni Malinen [Tue, 5 Feb 2019 18:30:08 +0000 (20:30 +0200)]
nl80211: Request kernel to trim off payload of netlink requests from acks
We do not need such payload in the acknowledgment, so adding it uses
resources unnecessarily. Furthermore, the original request can include
key material (e.g., NL80211_ATTR_PMK). libnl does not explicitly clear
this received message buffer and it would be inconvenient for
wpa_supplicant/hostapd to try to clear it with the current libnl design
where a duplicated buffer is actually passed to the callback. This means
that keys might be left unnecessarily in heap memory. Avoid this by
requesting the kernel not to copy back the request payload.
Jouni Malinen [Tue, 5 Feb 2019 18:26:50 +0000 (20:26 +0200)]
EAP peer: Clear temporary message buffers before freeing
These buffers in TLS-based EAP methods might contain keys or password
(e.g., when using TTLS-PAP or PEAP-GTC), so clear them explicitly to
avoid leaving such material into heap memory unnecessarily.
Jouni Malinen [Tue, 5 Feb 2019 18:25:21 +0000 (20:25 +0200)]
Clear config item writing buffer before freeing it
This buffer may be used to store items like passwords, so better clean
it explicitly to avoid possibility of leaving such items in heap memory
unnecessarily.
Johannes Berg [Tue, 5 Feb 2019 11:26:58 +0000 (12:26 +0100)]
tests: Optionally start telnet server inside VMs
If telnetd is installed and --telnet <port> is passed on the
vm-run.sh command line, start a telnet server (directly connected
to bash, no login) inside the VM(s) to be able to look into them
when something is wrong. Use a user network in qemu with a single
host forward from the specified port for this, listening only on
'localhost'.
Please note that this provides unauthenticated access to the guest
system from anything that can open a TCP connection on the host system.
The guess system does have access to reading all files on the host that
the user account running kvm has access to (and even write access if the
default ROTAG ,readonly parameter is cleared). In other words, this
option should not be used on any multiuser systems where kvm is run
under user accounts that are not dedicated for testing purposes (i.e.,
do not have access to any files that should not be readable to
everyone).
This needs CONFIG_VIRTIO_NET=y in the guest kernel.
For parallel-vm.py, the --telnet argument specifies the base port
and each VM index (0, 1, ...) is added to it.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Johannes Berg [Tue, 5 Feb 2019 11:26:46 +0000 (12:26 +0100)]
tests: Suppress annoying console reset from VMs
Recently, qemu/seabios grew an annoying console/terminal reset,
which also causes my terminal to be left in a state where long
lines don't work well and less gets confused because of this.
Suppress this by suppressing all output from qemu before a new
magic string printed from inside.sh.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Jouni Malinen [Tue, 5 Feb 2019 00:12:00 +0000 (02:12 +0200)]
tests: Create radio for p2ps_channel_active_go_and_station_different_mcc
Instead of relying on existing configuration (which may conflict
with other tests) and skipping otherwise, create a new radio with
two channels in this test and use it.
Johannes Berg [Sat, 2 Feb 2019 23:04:31 +0000 (00:04 +0100)]
tests: Create radio for p2ps_channel_both_connected_different_mcc
Instead of relying on existing configuration (which may conflict
with other tests) and skipping otherwise, create a new radio with
two channels in this test and use it.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Jouni Malinen [Mon, 4 Feb 2019 17:27:57 +0000 (19:27 +0200)]
tests: Encode Disconnect-Request attributes in sorted order for python3
This is needed to fix issues with dict iteration resulting in different
order of attributes when trying to calculate Message-Authenticator
externally to pyrad.
Jouni Malinen [Mon, 4 Feb 2019 16:26:53 +0000 (18:26 +0200)]
tests: Clean up pyrad test cases for python3 compatibility
All other test cases seem to work, but radius_das_disconnect_time_window
is still failing due to incorrect authenticator or Message-Authenticator
in Disconnect-Request.
Jouni Malinen [Mon, 4 Feb 2019 15:13:54 +0000 (17:13 +0200)]
D-Bus: Fix P2P DeleteService dict iteration
The previous implementation assumed the first entry coming out from the
dict is always service_type. That may not be the case, so properly
iterate over all dict entries in one loop instead of assuming what the
first entry is.
Jouni Malinen [Mon, 4 Feb 2019 14:46:31 +0000 (16:46 +0200)]
tests: Read sigma_dut-ap.conf as binary data for python3 compatibility
Sending UTF-8 encoded data to logger file is currently not working
properly, so create a separate binary file with a copy of
sigma_dut-ap.conf instead to work with both python2 and python3.
Jouni Malinen [Mon, 4 Feb 2019 10:23:45 +0000 (12:23 +0200)]
tests: MAC address ASCII string generation in python3 compatible manner
Use struct.unpack() to get a list of int and then generate a list of
hexstr octets from it for ':'.join() to get consistent behavior for both
python2 and python3.
Jouni Malinen [Sun, 3 Feb 2019 10:24:49 +0000 (12:24 +0200)]
tests: Replace str.translate() with str.replace()
This is needed for python3 since the two argument version of
str.translate() is not available for unicode. Furthermore, these cases
of delete colons from the string are simple enough for replace.
Jouni Malinen [Sat, 2 Feb 2019 10:49:23 +0000 (12:49 +0200)]
wpaspy: Convert to/from str to bytes as needed for python3
The control interface commands use mostly ASCII or UTF-8 strings, so
convert input/output to strings/bytes as needed for the socket
operations with python3.
Jouni Malinen [Fri, 1 Feb 2019 22:01:29 +0000 (00:01 +0200)]
PEAP: Explicitly clear temporary keys from memory when using CMK
The case of PEAPv0 with crypto binding did not clear some of the
temporary keys from stack/heap when those keys were not needed anymore.
Clear those explicitly to avoid unnecessary caching of keying material.
Jouni Malinen [Fri, 1 Feb 2019 21:52:28 +0000 (23:52 +0200)]
EAP-PEAP: Derive EMSK and use 128-octet derivation for MSK
Derive EMSK when using EAP-PEAP to enable ERP. In addition, change the
MSK derivation for EAP-PEAP to always derive 128 octets of key material
instead of the 64 octets to cover just the MSK. This is needed with the
PRF used in TLS 1.3 since the output length is mixed into the PRF
context.
Johannes Berg [Fri, 1 Feb 2019 20:31:59 +0000 (21:31 +0100)]
tests: wpasupplicant: Refactor code duplication in wait_global_event()
This code is identical to the wait_event() code, except for the
mon/global_mon instance. Create a _wait_event() function that
encapsulates this, and use it for both.
While at it, fix the bug in wait_global_event() where in the case
of not having a global_mon it always returns None.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Jouni Malinen [Wed, 30 Jan 2019 10:28:43 +0000 (12:28 +0200)]
DPP: Clear dpp_listen_freq on remain-on-channel failure
If the DPP_LISTEN command failed due to the driver rejecting the
remain-on-channel request, wpa_s->dpp_listen_freq was left set to the
requested listen frequency and this resulted in the next DPP_LISTEN for
the same frequency reporting "DPP: Already listening on .." even when
the driver was not really listening on that frequency. Fix this by
clearing wpa_s->dpp_listen_freq in the error case.
P2P: Allow the avoid channels for P2P discovery/negotiation
The avoid channels are notified through
QCA_NL80211_VENDOR_SUBCMD_AVOID_FREQUENCY allow minimal traffic, so
enhance the P2P behavior accordingly by considering these avoid
frequencies for P2P discovery/negotiation as long as they are not in
disallowed frequencies list.
Additionally, do not return failure when none of social channels are
available as operation channel, rather, mark the op_channel/op_reg_class
to 0 as this would anyway get selected during the group formation in
p2p_prepare_channel.
Jouni Malinen [Mon, 29 Oct 2018 19:10:27 +0000 (21:10 +0200)]
tests: WNM-Sleep Mode Request bounds checking for empty contents
The wnm_sleep_mode_proto test case was already covering number of
invalid WNM-Sleep Mode Request frame cases, but it was missing the
shortest possible case with a missing Dialog Token field. Add that as a
regression test case for bounds checking in
ieee802_11_rx_wnmsleep_req().
Jouni Malinen [Mon, 29 Oct 2018 18:48:07 +0000 (20:48 +0200)]
WNM: Fix WNM-Sleep Mode Request bounds checking
ieee802_11_rx_wnmsleep_req() might have been called for a short frame
that has no more payload after the Public Action field, i.e., with len
== 0. The bounds checking for the payload length was done only for the
information elements while the one octet Dialog Token field was read
unconditionally. In the original implementation, this could have
resulted in reading one octet beyond the end of the received frame data.
This case has not been reachable after the commit e0785ebbbd18 ("Use
more consistent Action frame RX handling in both AP mode paths"), but it
is better to address the specific issue in ieee802_11_rx_wnmsleep_req()
as well for additional protection against accidential removal of the
check and also to have something that can be merged into an older
version (pre-v2.7) if desired. The comments below apply for such older
versions where the case could have been reachable.
Depending on driver interface specific mechanism used for fetching the
frame, this could result in reading one octet beyond the end of a
stack/hash buffer or reading an uninitialized octet from within a
buffer. The actual value that was read as the Dialog Token field is not
used since the function returns immediately after having read this value
when there is no information elements following the field.
This issue was initially added in commit d32d94dbf47a ("WNM: Add
WNM-Sleep Mode implementation for AP") (with CONFIG_IEEE80211V=y build
option) and it remained in place during number of cleanup and fix
changes in this area and renaming of the build parameter to
CONFIG_WNM=y. The impacted function was not included in any default
build without one of the these optional build options being explicitly
enabled. CONFIG_WNM=y is still documented as "experimental and not
complete implementation" in hostapd/defconfig. In addition, commit 114f2830d2c2 ("WNM: Ignore WNM-Sleep Mode Request in wnm_sleep_mode=0
case") made this function exit before the impact read if WNM-Sleep Mode
support was not explicitly enabled in runtime configuration
(wnm_sleep_mode=1 in hostapd.conf). Commit e0785ebbbd18 ("Use more
consistent Action frame RX handling in both AP mode paths") made this
code unreachable in practice.
Add an explicit check that the frame has enough payload before reading
the Dialog Token field in ieee802_11_rx_wnmsleep_req().
Jared Bents [Wed, 16 Jan 2019 15:15:01 +0000 (09:15 -0600)]
crl_reload_interval: Add CRL reloading support
This patch adds a new flag 'crl_reload_interval' to reload CRL
periodically. This can be used to reload ca_cert file and the included
CRL information on every new TLS session if difference between the last
reload and the current time in seconds is greater than
crl_reload_interval.
This reloading is used for cases where check_crl is 1 or 2 and the CRL
is included in the ca_file.
Signed-off-by: Paresh Chaudhary <paresh.chaudhary@rockwellcollins.com> Signed-off-by: Jared Bents <jared.bents@rockwellcollins.com>
Daniel Golle [Wed, 23 Jan 2019 05:18:25 +0000 (06:18 +0100)]
Write multi_ap_backhaul_sta to wpa_supplicant config
The network configration option multi_ap_backhaul_sta was added without
adding it to wpa_config_write_network(). Hence the value of this option
was not included when writing the configuration file. Fix this by
including it in wpa_config_write_network().
Fixes: 5abc7823b ("wpa_supplicant: Add Multi-AP backhaul STA support") Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Lior David [Mon, 21 Jan 2019 16:28:42 +0000 (18:28 +0200)]
Check supported types in wpas_mac_addr_rand_scan_set()
When setting scan with randomized MAC, check the requested scan type
against supported types, to ensure callers will not set an unsupported
type, since this can cause scan/connect failures later. It is better to
do this in wpas_mac_addr_rand_scan_set() instead of control interface
specific code to apply the constraint on all possible interfaces using
this setting.
Lior David [Tue, 22 Jan 2019 15:24:35 +0000 (17:24 +0200)]
Fix test compilation error related to sme_event_unprot_disconnect()
sme_event_unprot_disconnect() is only defined with CONFIG_IEEE80211W, so
the CONFIG_TESTING_OPTIONS command UNPROT_DEAUTH can be defined only
with builds that enable IEEE 802.11w support.
Jouni Malinen [Tue, 22 Jan 2019 23:03:46 +0000 (01:03 +0200)]
HS 2.0 server: Alternative subrem updateNode for certificate credentials
The new subrem field in the users database can now be used to issue an
alternative subscription remediation updateNode for clients using
certificate credentials. The data file for this case is similar to the
policy update files, but it starts with the managementTreeURI value in
the first line.
Jouni Malinen [Tue, 22 Jan 2019 21:31:06 +0000 (23:31 +0200)]
HS 2.0 server: Use noMOUpdate in client certificate subrem
There is no point in trying to update the Credential node with the
existing contents in case of subscription remediation using a client
certificate instead of a username/password credential, so use the
noMOUpdate in that case.
Jouni Malinen [Tue, 22 Jan 2019 11:33:48 +0000 (13:33 +0200)]
FILS: Remove notes about experimental implementation
The standard amendment has been published and there has been sufficient
amount of interoperability testing for FILS to expect the protocol not
to be changed anymore, so remove the notes claiming this to be
experimental and not suitable for production use.