Commit b36935be1a ('nl80211: Fix EAPOL
frames not being delivered') and commit
147848ec4d ('nl80211: Do not add all
virtual interfaces to drv->if_indices') were not fully in sync and it
was possible for some non-hostapd use cases to end up adding undesired
ifindexes into the list of interfaces from which events and EAPOL frames
are processed on the parent interface. This could result, e.g., in P2P
Device management interface on getting unexpected events, including
RTM_NEWLINK event that could end up getting interpreted as an
indication of the interface being down and unavailable.
Make both add_ifidx() calls use the same criteria for adding interfaces
to the local list. This is not really a complete solution, but it is
good enough for now to fix the most visible side effects of this issue.
Signed-off-by: Jithu Jance <jithu@broadcom.com>
The driver might not be able to add the TDLS STA. Fail if this happens.
Also fix the error path to always reset the TDLS peer data.
Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
If a link is unreachable, the specification mandates we should send a
teardown packet via the AP with a specific teardown reason. Force this
by first disabling the link and only then sending the teardown packet
for the LOW_ACK event.
Rename the TDLS LOW_ACK event handler to better reflect its purpose.
Signed-off-by: Arik Nemtsov <arikx.nemtsov@intel.com>
Earlier commit related to MAC address based access control list
offloaded to the driver was not sending ACL configuration to the driver
if the MAC address list was empty. Remove this check as empty access
control list is a valid use case and sending ACL parameters should not
be dependent on whether the list is empty.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Performing a P2P Device flow such as p2p_listen or
p2p_find, can degrade the performance of an active interface
connection, if the listen frequency is different than the
frequency used by that interface.
To reduce the effect of P2P Device flows on other interfaces,
try changing the listen channel of the P2P Device to match the
operating channel of one of the other active interfaces. This change
will be possible only in case that the listen channel is not forced
externally, and will be delayed to a point where the P2P Device
state machine is idle.
The optimization can be configured in the configuration file and
is disabled by default.
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
In p2p_get_pref_freq, if the channels argument is NULL, select a
preferred channel that is also one of the P2P Device configured
channels.
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Use a single location in wpa_sm_rx_eapol() for preparing the pointer to
the Key Data field and to its validated length instead of fetching that
information in number of processing functions separately.
Signed-off-by: Jouni Malinen <j@w1.fi>
Re-order wpa_sm_rx_eapol() to first go through all EAPOL (802.1X) header
validation steps using the original message buffer and re-allocate and
copy the frame only if this is a valid EAPOL frame that contains an
EAPOL-Key. This makes the implementation easier to understand and saves
unnecessary memory allocations and copying should other types of EAPOL
frames get here.
Signed-off-by: Jouni Malinen <j@w1.fi>
These error messages had an incorrect frame name (likely copy-pasted
from the commit message handler) and couple of typos.
Signed-off-by: Jouni Malinen <j@w1.fi>
While this call cannot really fail, check the return value to be more
consistent with all the other wps_build_wfa_ext() calls.
Signed-off-by: Jouni Malinen <j@w1.fi>
While the hash functions would be very unlikely to fail in practice,
they do have option of returning an error. Check that return value more
consistently.
Signed-off-by: Jouni Malinen <j@w1.fi>
The additional eight octet field was removed from keydatalen without
proper validation of the Key Data Length field. It would have been
possible for an invalid EAPOL-Key frame to be processed in a way that
ends up reading beyond the buffer. In theory, this could have also
resulted in writing beyond the EAPOL-Key frame buffer, but that is
unlikely to be feasible due to the AES key wrap validation step on
arbitrary memory contents.
Signed-off-by: Jouni Malinen <j@w1.fi>
Since there is a limit on the EAP exchange due to maximum number of
roundtrips, there is no point in allowing excessively large buffers to
be allocated based on what the peer device claims the total message to
be. Instead, reject the message if it would not be possible to receive
it in full anyway.
Signed-off-by: Jouni Malinen <j@w1.fi>
The Action code field is in a fixed location, so the IEEE80211_HDRLEN
can be used here to clean up bounds checking to avoid false reports from
static analyzer.
Signed-off-by: Jouni Malinen <j@w1.fi>
There seemed to be an off-by-one error in the validation of GAS request
frames. If a Public Action frame without the Action code field would
have reached this function, the length could have been passed as
(size_t) -1 which would likely have resulted in a crash due to reading
beyond the buffer. However, it looks like such frame would not be
delivered to hostapd at least with mac80211-based drivers. Anyway, this
function better be more careful with length validation should some other
driver end up reporting invalid Action frames.
In addition, the Action code field is in a fixed location, so the
IEEE80211_HDRLEN can be used here to clean up bounds checking to avoid
false reports from static analyzer.
Signed-off-by: Jouni Malinen <j@w1.fi>
The Action code field is in a fixed location, so the IEEE80211_HDRLEN
can be used here to clean up bounds checking to avoid false reports from
static analyzer.
Signed-off-by: Jouni Malinen <j@w1.fi>
The implementation here was doing what it was supposed to, but the code
was indented in a way that made it quite confusing in the context of a
single line if statement body.
Signed-off-by: Jouni Malinen <j@w1.fi>
In tncc_read_config(), the memory allocted for the config
did not get freed if an error occured.
Signed-off-by: Philippe De Swert <philippe.deswert@jollamobile.com>
The driver may reject offchannel TX operation if the requested wait time
is longer than what the driver indicates as the maximum
remain-on-channel time. Two of the P2P action frame cases used long
enough wait times (1000 ms for DevDisc and 5000 ms for SD) that could go
beyond the limit with some drivers. Fix these to limit the maximum wait
to what the driver indicates as supported.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Extend the set of supported cipher suites to include CCMP-256, GCMP,
GCMP-256, CMAC-256, GMAC, and GMAC-256 when ATH_GCM_SUPPORT=y is set in
the build configuration.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
It does not look like there is going to be any additional use for this
old build option that could be used to build the EAP-IKEv2 peer
implementation in a way that interoperates with the eap-ikev2.ccns.pl
project. Remove the workarounds that matches incorrect implementation in
that project to clean up implementation.
Signed-off-by: Jouni Malinen <j@w1.fi>
While it does not look like that eapol_sm_get_key() would ever be called
with sm == NULL, the current implementation is inconsistent on whether
that is allowed or not. Check sm != NULL consistently to avoid warnings
from static analyzers.
Signed-off-by: Jouni Malinen <j@w1.fi>
The following kde.rsn_ie_len != peer->rsnie_i_len was already taking
care of enforcing the length to be within the target buffer length.
Anyway, this explicit check makes this clearer and matches the design in
TPK M1 processing.
Signed-off-by: Jouni Malinen <j@w1.fi>
In theory, gmtime() could return NULL if the year value would not fit
into an integer. However, that cannot really happen with the current
time() value in practice. Anyway, clean up static analyzer reports by
checking for this corner case.
Signed-off-by: Jouni Malinen <j@w1.fi>
Must check that UDN was present before trying to parse it. Avoid a NULL
pointer dereference by checking the result before using os_strstr() when
parsing device description from an AP.
Signed-off-by: Jouni Malinen <j@w1.fi>
prev cannot be NULL here in the hostapd_eid_country_add() call since
prev is set whenever start becomes non-NULL. That seems to be a bit too
difficult for some static analyzers, so check the prev pointer
explicitly to avoid false warnings.
Signed-off-by: Jouni Malinen <j@w1.fi>
This fixes an issue where a driver using the deprecated set_ieee8021x()
callback did not include rsn_pairwise bits in the driver configuration
even if mixed WPA+WPA2 configuration was used. This could result, e.g.,
in CCMP not being enabled properly when wpa_pairwise=TKIP and
rsn_pairwise=CCMP was used in the configuration. Fix this by using
bitwise OR of the wpa_pairwise and rsn_pairwise values to allow the
driver to enable all pairwise ciphers.
In addition, make the newer set_ap() driver_ops use the same bitwise OR
design instead of picking between rsn_pairwise and wpa_pairwise. This
makes the code paths consistent and can also fix issues with mixed mode
configuration with set_ap().
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
If the peer provides a username with large part of it being non-ASCII
characters, the previously used buffers may not have been long enough to
include the full string in debug logs and database search due to forced
truncation of the string by printf_encode(). Avoid this by increasing
the buffer sizes to fit in the maximum result.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
The off-by-one error in printf_encode() bounds checking could have
allowed buffer overflow with 0x00 being written to the memory position
following the last octet of the target buffer. Since this output is used
as \0-terminated string, the following operation would likely read past
the buffer as well. Either of these operations can result in the process
dying either due to buffer overflow protection or by a read from
unallowed address.
This has been seen to cause wpa_supplicant crash on OpenBSD when control
interface client attaches (debug print shows the client socket address).
Similarly, it may be possible to trigger the issue in RADIUS/EAP server
implementation within hostapd with a suitable constructed user name.
Signed-off-by: Stuart Henderson <sthen@openbsd.org>
If PMF was enabled, the validation step for EAPOL-Key descriptor version
ended up rejecting the message if GCMP had been negotiated as the
pairwise cipher. Fix this by making the GCMP check skipped similarly to
the CCMP case if a SHA256-based AKM is used.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Move to PTKINITDONE state and mark PTK valid after successful completion
of FT protocol. This allows the AP/Authenticator to start GTK rekeying
when FT protocol is used. Previously, the station using FT protocol did
not get the new GTK which would break delivery of group addressed
frames.
Signed-off-by: Jouni Malinen <j@w1.fi>
I got the below prints on a particular Android platform:
I/wpa_supplicant( 2637): nl80211: send_and_recv->nl_recvmsgs failed: 20
I/wpa_supplicant( 2637): nl80211: send_and_recv->nl_recvmsgs failed: 20
In JellyBean libnl_2 code, I see that the nl_recvmsgs returns postive values
too. In some cases, nl_recvmgs return the output of nl_recv function. nl_recv
function can return Number of bytes read, 0 or a negative error code.
Looks like this positive return value for nl_recvmsgs may be specific to
Android. While this is not how the API is supposed to work, this does no
harm with upstream libnl which returns only 0 or -1 from the function.
Signed-off-by: Jithu Jance <jithu@broadcom.com>
The dev_found() callback from NFC connection handover message processing
ended up using the p2p_dev_addr pointer that points to the parsed
message. However, that parsed data was freed just before the call. Fix
this by reordering the calls.
Signed-off-by: Jouni Malinen <j@w1.fi>
If the randomly generated bignum does not meet the validation steps, the
iteration loop in sae_get_rand() did not free the data properly. Fix the
memory leak by freeing the temporary bignum before starting the next
attempt at generating the value.
Signed-off-by: Jouni Malinen <j@w1.fi>
Couple of code paths in hostapd.c could have left hapd->drv_priv
pointing to memory that was freed in driver_nl80211.c when a secondary
BSS interface is removed. This could result in use of freed memory and
segfault when the next driver operation (likely during interface
deinit/removal). Fix this by clearing hapd->drv_priv when there is
reason to believe that the old value is not valid within the driver
wrapper anymore.
Signed-off-by: Jouni Malinen <j@w1.fi>
These three calls were used already in three different paths. Use a
helper function to avoid adding even more copies of this.
Signed-off-by: Jouni Malinen <j@w1.fi>
If for some reason interface setup fails mid-way when setting up
multi-BSS AP it was possible to get segmentation fault because driver
was not properly cleaned up.
One possible trigger, when using nl80211 driver, was udev renaming an
interface created by hostapd causing, e.g., linux_set_iface_flags() to
fail.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
This variable is updated when calling hostapd_if_add(), so it makes
sense to do the same thing when calling hostapd_if_remove().
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Performing, e.g. `wpa_cli -p /var/run/hostapd raw DISABLE` twice led to
hostapd segmentation fault if multiple BSSes were configured. Fix this
by checking if there is anything to disable at all before trying.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Make sure to not remove interfaces that were not created by
hostapd/wpa_supplicant. This was already done on number of the error
paths, but not all.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
It is possible for the connect() call to fail (e.g., due to unreachable
network based on local routing table), so the current auth/acct_sock may
be left to -1. Use that as an addition trigger to allow server failover
operation to be performed more quickly if it is known that the
retransmission attempt will not succeed anyway.
Signed-off-by: Jouni Malinen <j@w1.fi>
The re-open socket to the current RADIUS server code path did not work
in the expected way here. The pending authentication messages do not
need to be flushed in that case and neither should the retransmission
parameters be cleared. Fix this by performing these operations only if
the server did actually change as a part of a failover operation.
Signed-off-by: Jouni Malinen <j@w1.fi>
