Update the SAE-PK implementation to match the changes in the protocol
design:
- allow only Sec values 3 and 5 and encode this as a single bit field
with multiple copies
- add a checksum character
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This was clarified in the draft specification to not be a mandatory
requirement for the AP and STA to enforce, i.e., matching security level
is a recommendation for AP configuration rather than a protocol
requirement.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This maintains knowledge of whether H2E or PK was used as part of the
SAE authentication beyond the removal of temporary state needed during
that authentication. This makes it easier to use information about which
kind of SAE authentication was used at higher layer functionality.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This adds core SAE functionality for a new mode of using SAE with a
specially constructed password that contains a fingerprint for an AP
public key and that public key being used to validate an additional
signature in SAE confirm from the AP.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Check whether an error is reported from any of the functions that could
in theory fail and if so, do not proceed with the partially filled SAE
commit buffer.
Signed-off-by: Jouni Malinen <j@w1.fi>
Only one peer-commit-scalar value was stored for a specific STA (i.e.,
one per MAC address) and that value got replaced when the next SAE
Authentication exchange was started. This ended up breaking the check
against re-use of peer-commit-scalar from an Accepted instance when
anti-clogging token was requested. The first SAE commit message (the one
without anti-clogging token) ended up overwriting the cached
peer-commit-scalar value while leaving that instance in Accepted state.
The second SAE commit message (with anti-clogging token) added ended up
getting rejected if it used the same value again (and re-use is expected
in this particular case where the value was not used in Accepted
instance).
Fix this by using a separate pointer for storing the peer-commit-scalar
value that was used in an Accepted instance. There is no need to
allocate memory for two values, i.e., it is sufficient to maintain
separate pointers to the value and move the stored value to the special
Accepted state pointer when moving to the Accepted state.
This fixes issues where a peer STA ends up running back-to-back SAE
authentication within couple of seconds, i.e., without hostapd timing
out the STA entry for a case where anti-clogging token is required.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
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>
Remove support for performing full sqrt(), i.e., only support curves
that use prime with p = 3 mod 4. In practice, this drops only group 26
with SAE H2E. This seems acceptable since there does not seem to be any
strong use case for that group taken into account the limits being
placed on acceptable prime lengths.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
The KCK, PMK, and PMKID derivation fix broke SAE key derivation for all
FFC groups. Fix that by setting sae->tmp->order_len for FFC groups (it
was only set for ECC groups).
Fixes: ac734a342e ("SAE: Fix KCK, PMK, and PMKID derivation for groups 22, 23, 24")
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
IEEE Std 802.11-2016 is not exactly clear on the encoding of the bit
string that is needed for KCK, PMK, and PMKID derivation, but it seems
to make most sense to encode the (commit-scalar + peer-commit-scalar)
mod r part as a bit string by zero padding it from left to the length of
the order (in full octets).
The previous implementation used the length of the prime (in full
octets). This would work for KCK/PMK, but this results in deriving all
zero PMKIDs for the groups where the size of the order is smaller than
the size of the prime. This is the case for groups 22, 23, and 24.
However, those groups have been marked as being unsuitable for use with
SAE, so this fix should not really have a practical impact anymore.
Anyway, better fix it and document this clearly in the implementation
taken into account the unclarity of the standard in this area.
Signed-off-by: Jouni Malinen <j@w1.fi>
Use a shared function to determine the k parameter, i.e., the minimum
number of iterations of the PWE derivation loop, for SAE and EAP-pwd.
This makes it easier to fine-tune the parameter based on the negotiated
group, if desired.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This reduces differences in timing and memory access within the
hunting-and-pecking loop for ECC groups that have a prime that is not
close to a power of two (e.g., Brainpool curves).
Signed-off-by: Jouni Malinen <j@w1.fi>
Start sharing common SAE and EAP-pwd functionality by adding a new
source code file that can be included into both. This first step is
bringing in a shared function to check whether a group is suitable.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Move the identical function used by both SAE and EAP-pwd to
src/utils/common.c to avoid duplicated implementation.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This reduces timing and memory access pattern differences for an
operation that could depend on the used password.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Explicitly verify that own and peer commit scalar/element are available
when trying to check SAE confirm message. It could have been possible to
hit a NULL pointer dereference if the peer element could not have been
parsed. (CVE-2019-9496)
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Try to avoid showing externally visible timing or memory access
differences regardless of whether the derived pwd-value is smaller than
the group prime.
This is related to CVE-2019-9494.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This is an initial step towards making the FFC case use strictly
constant time operations similarly to the ECC case.
sae_test_pwd_seed_ffc() does not yet have constant time behavior,
though.
This is related to CVE-2019-9494.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
These groups have significant probability of coming up with pwd-value
that is equal or greater than the prime and as such, need for going
through the PWE derivation loop multiple times. This can result in
sufficient timing different to allow an external observer to determine
how many rounds are needed and that can leak information about the used
password.
Force at least 40 loop rounds for these MODP groups similarly to the ECC
group design to mask timing. This behavior is not described in IEEE Std
802.11-2016 for SAE, but it does not result in different values (i.e.,
only different timing), so such implementation specific countermeasures
can be done without breaking interoperability with other implementation.
Note: These MODP groups 22, 23, and 24 are not considered sufficiently
strong to be used with SAE (or more or less anything else). As such,
they should never be enabled in runtime configuration for any production
use cases. These changes to introduce additional protection to mask
timing is only for completeness of implementation and not an indication
that these groups should be used.
This is related to CVE-2019-9494.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Make the non-failure path in the function proceed without branches based
on r_odd and in constant time to minimize risk of observable differences
in timing or cache use. (CVE-2019-9494)
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
The QR test result can provide information about the password to an
attacker, so try to minimize differences in how the
sae_test_pwd_seed_ecc() result is used. (CVE-2019-9494)
Use heap memory for the dummy password to allow the same password length
to be used even with long passwords.
Use constant time selection functions to track the real vs. dummy
variables so that the exact same operations can be performed for both QR
test results.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
The rules defining which DH groups are suitable for SAE use were
accepted into IEEE 802.11 REVmd based on this document:
https://mentor.ieee.org/802.11/dcn/19/11-19-0387-02-000m-addressing-some-sae-comments.docx
Enforce those rules in production builds of wpa_supplicant and hostapd.
CONFIG_TESTING_OPTIONS=y builds can still be used to select any o the
implemented groups to maintain testing coverage.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This extends the SAE implementation in both infrastructure and mesh BSS
cases to allow an optional Password Identifier to be used. This uses the
mechanism added in P802.11REVmd/D1.0. The Password Identifier is
configured in a wpa_supplicant network profile as a new string parameter
sae_password_id. In hostapd configuration, the existing sae_password
parameter has been extended to allow the password identifier (and also a
peer MAC address) to be set. In addition, multiple sae_password entries
can now be provided to hostapd to allow multiple per-peer and
per-identifier passwords to be set.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This makes it easier to understand why "SAE: Failed to select group"
debug entry shows up in cases the selected crypto library does not
support a specific group.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Set Sc to 2^16-1 when moving to Accepted state per IEEE Std 802.11-2016,
12.4.8.6.5 (Protocol instance behavior - Confirmed state). This allows
the peer in Accepted state to silently ignore unnecessary
retransmissions of the Confirm message.
Signed-off-by: Jouni Malinen <j@w1.fi>
IEEE Std 802.11-2012 11.3.5.4 specifies the PMKID for SAE-derived keys
as:
L((commit-scalar + peer-commit-scalar) mod r, 0, 128)
This is already calculated in the SAE code when the PMK is derived, but
not saved anywhere. Later, when generating the PMKID for plink action
frames, the definition for PMKID from 11.6.1.3 is incorrectly used.
Correct this by saving the PMKID when the key is generated and use it
subsequently.
Signed-off-by: Bob Copeland <me@bobcopeland.com>
Reorder terms in a way that no invalid pointers are generated with
pos+len operations. end-pos is always defined (with a valid pos pointer)
while pos+len could end up pointing beyond the end pointer which would
be undefined behavior.
Signed-off-by: Jouni Malinen <j@w1.fi>
The current P802.11 description of SAE uses "1 < element < p" as the
required range. However, this is not correct and does not match the
Dragonfly description of "1 < element < p-1". SAE definition will need
to change here. Update the implementation to reject p-1 based on the
correct rule here.
Signed-off-by: Jouni Malinen <j@w1.fi>
IEEE Std 802.11-2012 description of SAE does not require this, i.e., it
describes the requirement as 0 < scalar < r for processing the Commit
message. However, this is not correct and will be changes to 1 < scalar
< r to match the Dragonfly description so that a trivial secret case
will be avoided explicitly.
This is not much of an issue for the locally generated commit-scalar
since it would be very unlikely to get the value of 1. For Commit
message processing, a peer with knowledge of the password could
potentially force the exchange to expose key material without this
check.
Signed-off-by: Jouni Malinen <j@w1.fi>
If PWE is discovered before the minimum number of loops (k) is reached,
the extra iterations use a random "password" to further obfuscate the
cost of discovering PWE.
Signed-off-by: Jouni Malinen <j@w1.fi>
This replaces the earlier IEEE Std 802.11-2012 algorithm with the design
from P802.11-REVmc/D4.0. Things brings in a blinding technique for
determining whether the pwd-seed results in a suitable PWE value.
Signed-off-by: Jouni Malinen <j@w1.fi>
draft-irtf-cfrg-dragonfly recommends implementation to set the security
parameter, k, to a value of at least 40. This will make PWE generation
take significantly more resources, but makes it more likely to hide
timing differences due to different number of loops needed to find a
suitable PWE.
Signed-off-by: Jouni Malinen <j@w1.fi>
