struct ieee80211_local needs to be passed in separately instead of
dereferencing the (potentially NULL) sdata
Signed-off-by: Felix Fietkau <nbd@nbd.name>
new features:
- priority-inheritance mutexes
- membarrier syscall, pre-registration to use it, fallback emulation
- header-level support for new linux features in 4.19, 4.20, 5.0
major internal changes:
- complete, async-safe view of all existent threads as global list
- robust __synccall based on new thread list
- new dynamic TLS is installed synchronously at dlopen
- TLSDESC resolver functions no longer make bad ABI assumptions to call C
- resolved shared library dependencies are now recorded
compatibility & conformance:
- dependency-order shared library constructor execution
- sigaltstack no longer rejects SS_AUTODISARM, future flags
- FILE is now a complete (dummy) type in pre-C11 feature profiles
- setvbuf reports failure on invalid arguments
- TSVTX is exposed unconditionally in tar.h
- multithreaded set*id() no longer depends on /proc
- key slot reuse after pthread_key_delete no longer depends on /proc
bugs fixed:
- failures in multithreaded set*id() with concurrent thread creation/exit
- interposed free was called from invalid/inconsistent contexts
- freeaddrinfo performed invalid free of some partial results lists
- dlsym dependency order search had false negatives and false positives
- dn_skipname gave wrong results for labels with 8-bit content
- dcngettext clobbered errno, often breaking printing of error messages
- sscanf read past end of buffer under certain conditions (1.1.21 regression)
- pthread_key_create spuriously failed under race condition (1.1.21 regression)
- fdopendir wrongly succeeded with O_PATH file descriptors
- gets behaved incorrectly in presence of null bytes
- namespace violations in c11 tsd and mutex function dependencies
- incorrect prototype for makecontext (unimplemented)
arch-specfic bugs fixed:
- s390x had wrong values for POSIX_FADV_DONTNEED/_NOREUSE
Extensively tested on dozens of devices, covering most popular architectures.
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Signed-off-by: Dainis Jonitis <dainis.jonitis@ubnt.com>
Signed-off-by: Roman Yeryomin <roman@advem.lv>
9cd701a4f028 ath10k-ct: Add PN get/set API for wave-2 firmware.
5c8a4668323b ath10k-ct: Support over-riding the power ctl table in eeprom
75e2705f31bb ath10k-ct: CCA, eeprom, other changes.
a696e602a0fc ath10k-ct: Attempt to fix-out-of-tree compile for 4.16
a2aec62262df ath10k: Improve beacon tx status for 4.20 kernel.
be5c21a82b15 ath10k-ct: Fix out-of-tree compile for 4.20, pull in stable changes for 4.19
Fixes compile errors when using the 4.20 flavour.
Also the amount of beacon errors seems to have dropped.
Tested on a Mikrotik RB912UAGS-5HPacD
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
This reverts commit c6aa9ff388.
Further testing has revealed that we will need to allow concurrent
requests after all, especially for situations where CGI processes
initiate further HTTP requests to the local host.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Refreshed all patches.
New symbols:
- CONFIG_LDISC_AUTOLOAD
- CONFIG_PPC_BARRIER_NOSPEC
Compile-tested on: ar7
Runtime-tested on: none
Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
This applies an upstream patch that fixes a OPENSSL_config() bug that
causes SSL initialization to fail when the openssl.cnf file is not
found. The config file is not installed by default.
Signed-off-by: Eneas U de Queiroz <cote2004-github@yahoo.com>
4a9d2005 Update manual pages
acf6a922 Bump up version number to 1.38.0, LT revision to 31:3:17
4ff45821 Update AUTHORS
42dce01e Merge branch 'nghttpx-fix-backend-selection-on-retry'
a35059e3 nghttpx: Fix bug that altered authority and path affect backend selection
5a30fafd Merge branch 'nghttpx-fix-chunked-request-stall'
dce91ad3 Merge branch 'nghttpx-dont-log-authorization'
2cff8b43 nghttpx: Fix bug that chunked request stalls
be96654d nghttpx: Don't log authorization request header field value with -LINFO
ce962c3f Merge branch 'update-http-parser'
f931504e Update http-parser to v2.9.1
d978f351 Fix bug that on_header callback is still called after stream is closed
ec519f22 Merge pull request #1270 from baitisj/master
e8b213e3 Bump up version number to 1.38.0-DEV
Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
This patch doesn't seem to have any more users.
The only one used to be target in
`target/linux/etrax/image/e100boot/Makefile`.
That target has since been removed via commit
1080f68b2b .
There doesn't seem to be any reason left for this workaround.
Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
This patch fixes disfunctional WLAN LED on TP-Link W8970. The LED was
reported working in the CC release[1], but doesn't work anymore in 18.06.2.
1. 420cb24d41
Tested-by: Damian Janarek <dzanar18@o2.pl>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
The sender domain has a DMARC Reject/Quarantine policy which disallows
sending mailing list messages using the original "From" header.
To mitigate this problem, the original message has been wrapped
automatically by the mailing list software.
Enable engine support by default. Right now, some packages require
this, so it is always enabled by the bots. Many packages will compile
differently when engine support is detected, needing engine symbols from
the libraries.
However, being off by default, a user compiling its own image will fail
to run some popular packages from the official repo.
Note that disabling engines did not work in 1.0.2, so this problem never
showed up before.
NPN support has been removed in major browsers & servers, and has become
a small bloat, so it does not make sense to leave it on by default.
Remove deprecated CONFIG_ENGINE_CRYPTO symbol that is no longer needed.
Signed-off-by: Eneas U de Queiroz <cote2004-github@yahoo.com>
Currently the Geode builds fails on following kernel module missing
dependencies:
Package kmod-drm-amdgpu is missing dependencies for the following libraries:
backlight.ko
drm_kms_helper.ko
fb.ko
ttm.ko
So this patch tries to fix the kmod-drm-amdgpu module dependecies.
Fixes: 2f239c0 ("x86: video: add amdgpu DRM kernel package")
Fixes: 2f6918e ("x86: video: add radeon DRM module support")
Tested-by: Tomasz Maciej Nowak <tomek_n@o2.pl>
Signed-off-by: Lucian Cristian <lucian.cristian@gmail.com>
Commit 6e060bd62c introduced a dependency to the dialout group.
Adding this group to the "group" file in the base-files package is not
enough to handle this dependency, because after a sysupgrade this entry
will be missing in the "group" file.
To address this problem the dependencies to the required groups needs to
be set in the Makefile of the procd package.
Then, the uci-default script "13_fix_group_user" will add the groups
on first boot-up after a sysupgrade.
Fixes: 6e060bd62c ("base-files/hotplug: fix dedicated group for tty devices")
Tested-by: Michael Heimpold <mhei@heimpold.de>
Signed-off-by: Martin Schiller <ms@dev.tdt.de>
Both targets have their own idea of how to use ehci-fsl.
This patch reverts part of commit
68b8d3b079 ("kernel: usb: add FSL EHCI package") and moves
ehci-fsl back into kmod-usb2, while also making it hopefully
useable for the mpc85xx target.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
The current mpc85xx build is failing because the
TL-WDR4900v1 kernel image no longer fits into the
partition. Extending the kernel is not possible
without updating u-boot's kernel loader commands.
This patch disables the WDR4900v1 until the kernel
image size issue is fixed so the buildbot can still
compile the Sophos RED 15w Rev.1 . Installing the
WDR4900v1 images would cause the routers to get
bricked.
For the discussion, please go to:
<https://github.com/openwrt/openwrt/pull/1773>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Currently it's not possible to use perf on ath79 due to genirq flags
mismatch happening on static virtual IRQ 13 which is used for
performance counters hardware IRQ 5.
On TP-Link Archer C7v5:
CPU0
2: 0 MIPS 2 ath9k
4: 318 MIPS 4 19000000.eth
7: 55034 MIPS 7 timer
8: 1236 MISC 3 ttyS0
12: 0 INTC 1 ehci_hcd:usb1
13: 0 gpio-ath79 2 keys
14: 0 gpio-ath79 5 keys
15: 31 AR724X PCI 1 ath10k_pci
$ perf top
genirq: Flags mismatch irq 13. 00014c83 (mips_perf_pmu) vs. 00002003 (keys)
On TP-Link Archer C7v4:
CPU0
4: 0 MIPS 4 19000000.eth
5: 7135 MIPS 5 1a000000.eth
7: 98379 MIPS 7 timer
8: 30 MISC 3 ttyS0
12: 90028 INTC 0 ath9k
13: 5520 INTC 1 ehci_hcd:usb1
14: 4623 INTC 2 ehci_hcd:usb2
15: 32844 AR724X PCI 1 ath10k_pci
16: 0 gpio-ath79 16 keys
23: 0 gpio-ath79 23 keys
$ perf top
genirq: Flags mismatch irq 13. 00014c80 (mips_perf_pmu) vs. 00000080 (ehci_hcd:usb1)
This problem is happening, because currently statically assigned virtual
IRQ 13 for performance counters is not claimed during the initialization
of MIPS PMU during the bootup, so the IRQ subsystem doesn't know, that
this interrupt isn't available for further use.
So this patch fixes the issue by simply booking hardware IRQ 5 for MIPS PMU.
Signed-off-by: Petr Štetiar <ynezz@true.cz>
24f9dc7 Iron out all extra compiler warnings
9d8dbc9 Enable extra compiler checks
ff8d356 mbim-proxy support
ccca03f umbim: add registration set support
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Reverting this commit as I've missed the fact, that the button is
already present in the included DTSI file.
Signed-off-by: Petr Štetiar <ynezz@true.cz>
It's OEM module with 2*26 pin header, similar to LinkIt Smart 7688 or
Vocore2.
Specification:
CPU: MT7628 580 MHz. MIPS 24K
RAM: 64 MB
Flash: 8 MB
WIFI: 802.11n/g/b 20/40 MHz
USB: 1x Port USB 2.0
Ethernet: 5 Port ethernet switch
UART: 2x
Installation: Use the installed uboot Bootloader. Connect a serial cable
to serialport 0. Turn power on. Choose the option: "Load system code
then write to Flash via TFTP". Choose the local device IP and the TFTP
server IP and the file name of the system image. After if the
Bootloader will copy the image to the local flash.
Notes: The I2C Kernel module work not correctly. You can send and
receive data. But the command i2cdetect doesn’t work. FS#845
Signed-off-by: Eike Feldmann <eike.feldmann@outlook.com>
[commit subject and message touches, DTS whitespace fixes, wifi LED
rename, pinctrl fixes, network settings fixes, lan/wmac mac addresses,
removed i2c kernel modules]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Hardware
--------
SoC: MediaTek MT7628NN
RAM: 64M DDR2 (Etron EM68B16CWQD-25H)
FLASH: 8M (Winbond W25Q64JVSIQ)
LED: Power - WLAN
BTN: Reset
UART: 115200 8N1
TX and RX are labled on the board as pads next to the SoC
Installation via web-interface
------------------------------
1. Visit the web-interface at 192.168.8.1
Note: The ethernet port is by default WAN. So you need to connect to
the router via WiFi
2. Navigate to the Update tab on the left side.
3. Select "Local Update"
4. Upload the OpenWrt sysupgrade image.
Note: Make sure you select not to preserve the configuration.
Installation via U-Boot
-----------------------
1. Hold down the reset button while powering on the device.
Wait for the LED to flash 5 times.
2. Assign yourself a static IPv4 in 192.168.1.0/24
3. Upload the OpenWrt sysupgrade image at 192.168.1.1.
Signed-off-by: David Bauer <mail@david-bauer.net>
TP-Link TL-WDR 4900 have u-boot with read-only env.
Boot command read only 0x29F000 data from flash.
Bigger images causes crc error. It can't be changed.
This patch add kernel size checking.
Signed-off-by: Pawel Dembicki <paweldembicki@gmail.com>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> [utilize KERNEL_SIZE]
Remove wireless and USB packages from the device-specific package
selection as they are already selected by the target itself.
Signed-off-by: David Bauer <mail@david-bauer.net>
This removes USB support from the compiled kernel. Because of this, the
kernel is just small enough for the TP-Link WDR4900 to boot the
resulting kernel.
This is necessary to support the WDR4900 in the upcoming 19.xx release.
In the long run, this should be fixed with a second stage bootloader, as
the vendor bootloader only loads the first 2684k bytes.
Signed-off-by: David Bauer <mail@david-bauer.net>
I-O DATA ETG3-R is a wired router. So wireless-related packages are
unnecessary and remove those packages from default configuration to
reduce flash usage.
Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
This adds support for the TP-Link WR842N v3 which is already supported on ar71xx
target (0b45bec22c).
Specification:
* SoC: QCA9533 ver 2 rev 0
* 16 MB Flash (gd25q128)
* 64 MB RAM
* 1 WAN 10/100 MBit/s (blue connector)
* 4 LAN 10/100 MBit/s (AR8229; 4 ports; yellow connectors)
* Atheros AR9531 (2,4GHz, two fixed antennas)
* USB
* Reset / WPS button
* WiFi button (rf kill)
* 8 green leds; 1 red/green led
* serial console (115200 8N1, according to the OpenWrt-wiki some soldering is needed)
Installation:
* flash via vendor WebUI (the filename must not exceed certain length)
* sysupgrade from installed OpenWrt (also ar71xx)
Thanks to Holger Drefs for providing the hardware
Tested-by: @kofec (github)
Signed-off-by: Sven Roederer <devel-sven@geroedel.de>
Replace the patch introduced by commit d0b969eee8 ("mac80211: rt2x00:
do not increment sequence number while re-transmitting") was merged
into wireless-drivers.git. Replace our version with the merged version.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Instead of assigning I2C pins as GPIOs by default, leave it up to the
user whether to install kmod-i2c-mt7621 and use them for hardware I2C
instead.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
ar71xx is in the process of being deprecated as a target accepting new
devices. The replacement target for the same hardware is DTS
based ath79.
Switch the default build target selection from ar71xx to ath79.
This is intended to encourage DTS takeup & support for ath79 and longer
term will also aid kernel upstream support.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
EAP-pwd missing commit validation
Published: April 10, 2019
Identifiers:
- CVE-2019-9497 (EAP-pwd server not checking for reflection attack)
- CVE-2019-9498 (EAP-pwd server missing commit validation for
scalar/element)
- CVE-2019-9499 (EAP-pwd peer missing commit validation for
scalar/element)
Latest version available from: https://w1.fi/security/2019-4/
Vulnerability
EAP-pwd implementation in hostapd (EAP server) and wpa_supplicant (EAP
peer) was discovered not to validate the received scalar and element
values in EAP-pwd-Commit messages properly. This could result in attacks
that would be able to complete EAP-pwd authentication exchange without
the attacker having to know the used password.
A reflection attack is possible against the EAP-pwd server since the
hostapd EAP server did not verify that the EAP-pwd-Commit contains
scalar/element values that differ from the ones the server sent out
itself. This allows the attacker to complete EAP-pwd authentication
without knowing the password, but this does not result in the attacker
being able to derive the session key (MSK), i.e., the attacker would not
be able to complete the following key exchange (e.g., 4-way handshake in
RSN/WPA).
An attack using invalid scalar/element values is possible against both
the EAP-pwd server and peer since hostapd and wpa_supplicant did not
validate these values in the received EAP-pwd-Commit messages. If the
used crypto library does not implement additional checks for the element
(EC point), this could result in attacks where the attacker could use a
specially crafted commit message values to manipulate the exchange to
result in deriving a session key value from a very small set of possible
values. This could further be used to attack the EAP-pwd server in a
practical manner. An attack against the EAP-pwd peer is slightly more
complex, but still consider practical. These invalid scalar/element
attacks could result in the attacker being able to complete
authentication and learn the session key and MSK to allow the key
exchange to be completed as well, i.e., the attacker gaining access to
the network in case of the attack against the EAP server or the attacker
being able to operate a rogue AP in case of the attack against the EAP
peer.
While similar attacks might be applicable against SAE, it should be
noted that the SAE implementation in hostapd and wpa_supplicant does
have the validation steps that were missing from the EAP-pwd
implementation and as such, these attacks do not apply to the current
SAE implementation. Old versions of wpa_supplicant/hostapd did not
include the reflection attack check in the SAE implementation, though,
since that was added in June 2015 for v2.5 (commit 6a58444d27fd 'SAE:
Verify that own/peer commit-scalar and COMMIT-ELEMENT are different').
Vulnerable versions/configurations
All hostapd versions with EAP-pwd support (CONFIG_EAP_PWD=y in the build
configuration and EAP-pwd being enabled in the runtime configuration)
are vulnerable against the reflection attack.
All wpa_supplicant and hostapd versions with EAP-pwd support
(CONFIG_EAP_PWD=y in the build configuration and EAP-pwd being enabled
in the runtime configuration) are vulnerable against the invalid
scalar/element attack when built against a crypto library that does not
have an explicit validation step on imported EC points. The following
list indicates which cases are vulnerable/not vulnerable:
- OpenSSL v1.0.2 or older: vulnerable
- OpenSSL v1.1.0 or newer: not vulnerable
- BoringSSL with commit 38feb990a183 ('Require that EC points are on the
curve.') from September 2015: not vulnerable
- BoringSSL without commit 38feb990a183: vulnerable
- LibreSSL: vulnerable
- wolfssl: vulnerable
Acknowledgments
Thanks to Mathy Vanhoef (New York University Abu Dhabi) for discovering
and reporting the issues and for proposing changes to address them in
the implementation.
Possible mitigation steps
- Merge the following commits to wpa_supplicant/hostapd and rebuild:
CVE-2019-9497:
EAP-pwd server: Detect reflection attacks
CVE-2019-9498:
EAP-pwd server: Verify received scalar and element
EAP-pwd: Check element x,y coordinates explicitly
CVE-2019-9499:
EAP-pwd client: Verify received scalar and element
EAP-pwd: Check element x,y coordinates explicitly
These patches are available from https://w1.fi/security/2019-4/
- Update to wpa_supplicant/hostapd v2.8 or newer, once available
Signed-off-by: Stefan Lippers-Hollmann <s.l-h@gmx.de>
[bump PKG_RELEASE]
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
hostapd: fix SAE confirm missing state validation
Published: April 10, 2019
Identifiers:
- CVE-2019-9496 (SAE confirm missing state validation in hostapd/AP)
Latest version available from: https://w1.fi/security/2019-3/
Vulnerability
When hostapd is used to operate an access point with SAE (Simultaneous
Authentication of Equals; also known as WPA3-Personal), an invalid
authentication sequence could result in the hostapd process terminating
due to a NULL pointer dereference when processing SAE confirm
message. This was caused by missing state validation steps when
processing the SAE confirm message in hostapd/AP mode.
Similar cases against the wpa_supplicant SAE station implementation had
already been tested by the hwsim test cases, but those sequences did not
trigger this specific code path in AP mode which is why the issue was
not discovered earlier.
An attacker in radio range of an access point using hostapd in SAE
configuration could use this issue to perform a denial of service attack
by forcing the hostapd process to terminate.
Vulnerable versions/configurations
All hostapd versions with SAE support (CONFIG_SAE=y in the build
configuration and SAE being enabled in the runtime configuration).
Possible mitigation steps
- Merge the following commit to hostapd and rebuild:
SAE: Fix confirm message validation in error cases
These patches are available from https://w1.fi/security/2019-3/
- Update to hostapd v2.8 or newer, once available
Signed-off-by: Stefan Lippers-Hollmann <s.l-h@gmx.de>
[bump PKG_RELEASE]
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
EAP-pwd side-channel attack
Published: April 10, 2019
Identifiers:
- CVE-2019-9495 (cache attack against EAP-pwd)
Latest version available from: https://w1.fi/security/2019-2/
Vulnerability
Number of potential side channel attacks were recently discovered in the
SAE implementations used by both hostapd and wpa_supplicant (see
security advisory 2019-1 and VU#871675). EAP-pwd uses a similar design
for deriving PWE from the password and while a specific attack against
EAP-pwd is not yet known to be tested, there is no reason to believe
that the EAP-pwd implementation would be immune against the type of
cache attack that was identified for the SAE implementation. Since the
EAP-pwd implementation in hostapd (EAP server) and wpa_supplicant (EAP
peer) does not support MODP groups, the timing attack described against
SAE is not applicable for the EAP-pwd implementation.
A novel cache-based attack against SAE handshake would likely be
applicable against the EAP-pwd implementation. Even though the
wpa_supplicant/hostapd PWE derivation iteration for EAP-pwd has
protections against timing attacks, this new cache-based attack might
enable an attacker to determine which code branch is taken in the
iteration if the attacker is able to run unprivileged code on the victim
machine (e.g., an app installed on a smart phone or potentially a
JavaScript code on a web site loaded by a web browser). This depends on
the used CPU not providing sufficient protection to prevent unprivileged
applications from observing memory access patterns through the shared
cache (which is the most likely case with today's designs).
The attacker could use information about the selected branch to learn
information about the password and combine this information from number
of handshake instances with an offline dictionary attack. With
sufficient number of handshakes and sufficiently weak password, this
might result in full recovery of the used password if that password is
not strong enough to protect against dictionary attacks.
This attack requires the attacker to be able to run a program on the
target device. This is not commonly the case on an authentication server
(EAP server), so the most likely target for this would be a client
device using EAP-pwd.
The commits listed in the end of this advisory change the EAP-pwd
implementation shared by hostapd and wpa_supplicant to perform the PWE
derivation loop using operations that use constant time and memory
access pattern to minimize the externally observable differences from
operations that depend on the password even for the case where the
attacker might be able to run unprivileged code on the same device.
Vulnerable versions/configurations
All wpa_supplicant and hostapd versions with EAP-pwd support
(CONFIG_EAP_PWD=y in the build configuration and EAP-pwd being enabled
in the runtime configuration).
It should also be noted that older versions of wpa_supplicant/hostapd
prior to v2.7 did not include additional protection against certain
timing differences. The definition of the EAP-pwd (RFC 5931) does not
describe such protection, but the same issue that was addressed in SAE
earlier can be applicable against EAP-pwd as well and as such, that
implementation specific extra protection (commit 22ac3dfebf7b, "EAP-pwd:
Mask timing of PWE derivation") is needed to avoid showing externally
visible timing differences that could leak information about the
password. Any uses of older wpa_supplicant/hostapd versions with EAP-pwd
are recommended to update to v2.7 or newer in addition to the mitigation
steps listed below for the more recently discovered issue.
Possible mitigation steps
- Merge the following commits to wpa_supplicant/hostapd and rebuild:
OpenSSL: Use constant time operations for private bignums
Add helper functions for constant time operations
OpenSSL: Use constant time selection for crypto_bignum_legendre()
EAP-pwd: Use constant time and memory access for finding the PWE
These patches are available from https://w1.fi/security/2019-2/
- Update to wpa_supplicant/hostapd v2.8 or newer, once available
- Use strong passwords to prevent dictionary attacks
Signed-off-by: Stefan Lippers-Hollmann <s.l-h@gmx.de>
[bump PKG_RELEASE]
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
SAE side-channel attacks
Published: April 10, 2019
Identifiers:
- VU#871675
- CVE-2019-9494 (cache attack against SAE)
Latest version available from: https://w1.fi/security/2019-1/
Vulnerability
Number of potential side channel attacks were discovered in the SAE
implementations used by both hostapd (AP) and wpa_supplicant
(infrastructure BSS station/mesh station). SAE (Simultaneous
Authentication of Equals) is also known as WPA3-Personal. The discovered
side channel attacks may be able to leak information about the used
password based on observable timing differences and cache access
patterns. This might result in full password recovery when combined with
an offline dictionary attack and if the password is not strong enough to
protect against dictionary attacks.
Cache attack
A novel cache-based attack against SAE handshake was discovered. This
attack targets SAE with ECC groups. ECC group 19 being the mandatory
group to support and the most likely used group for SAE today, so this
attack applies to the most common SAE use case. Even though the PWE
derivation iteration in SAE has protections against timing attacks, this
new cache-based attack enables an attacker to determine which code
branch is taken in the iteration if the attacker is able to run
unprivileged code on the victim machine (e.g., an app installed on a
smart phone or potentially a JavaScript code on a web site loaded by a
web browser). This depends on the used CPU not providing sufficient
protection to prevent unprivileged applications from observing memory
access patterns through the shared cache (which is the most likely case
with today's designs).
The attacker can use information about the selected branch to learn
information about the password and combine this information from number
of handshake instances with an offline dictionary attack. With
sufficient number of handshakes and sufficiently weak password, this
might result in full discovery of the used password.
This attack requires the attacker to be able to run a program on the
target device. This is not commonly the case on access points, so the
most likely target for this would be a client device using SAE in an
infrastructure BSS or mesh BSS.
The commits listed in the end of this advisory change the SAE
implementation shared by hostapd and wpa_supplicant to perform the PWE
derivation loop using operations that use constant time and memory
access pattern to minimize the externally observable differences from
operations that depend on the password even for the case where the
attacker might be able to run unprivileged code on the same device.
Timing attack
The timing attack applies to the MODP groups 22, 23, and 24 where the
PWE generation algorithm defined for SAE can have sufficient timing
differences for an attacker to be able to determine how many rounds were
needed to find the PWE based on the used password and MAC
addresses. When the attack is repeated with multiple times, the attacker
may be able to gather enough information about the password to be able
to recover it fully using an offline dictionary attack if the password
is not strong enough to protect against dictionary attacks. This attack
could be performed by an attacker in radio range of an access point or a
station enabling the specific MODP groups.
This timing attack requires the applicable MODP groups to be enabled
explicitly in hostapd/wpa_supplicant configuration (sae_groups
parameter). All versions of hostapd/wpa_supplicant have disabled these
groups by default.
While this security advisory lists couple of commits introducing
additional protection for MODP groups in SAE, it should be noted that
the groups 22, 23, and 24 are not considered strong enough to meet the
current expectation for a secure system. As such, their use is
discouraged even if the additional protection mechanisms in the
implementation are included.
Vulnerable versions/configurations
All wpa_supplicant and hostapd versions with SAE support (CONFIG_SAE=y
in the build configuration and SAE being enabled in the runtime
configuration).
Acknowledgments
Thanks to Mathy Vanhoef (New York University Abu Dhabi) and Eyal Ronen
(Tel Aviv University) for discovering the issues and for discussions on
how to address them.
Possible mitigation steps
- Merge the following commits to wpa_supplicant/hostapd and rebuild:
OpenSSL: Use constant time operations for private bignums
Add helper functions for constant time operations
OpenSSL: Use constant time selection for crypto_bignum_legendre()
SAE: Minimize timing differences in PWE derivation
SAE: Avoid branches in is_quadratic_residue_blind()
SAE: Mask timing of MODP groups 22, 23, 24
SAE: Use const_time selection for PWE in FFC
SAE: Use constant time operations in sae_test_pwd_seed_ffc()
These patches are available from https://w1.fi/security/2019-1/
- Update to wpa_supplicant/hostapd v2.8 or newer, once available
- In addition to either of the above alternatives, disable MODP groups
1, 2, 5, 22, 23, and 24 by removing them from hostapd/wpa_supplicant
sae_groups runtime configuration parameter, if they were explicitly
enabled since those groups are not considered strong enough to meet
current security expectations. The groups 22, 23, and 24 are related
to the discovered side channel (timing) attack. The other groups in
the list are consider too weak to provide sufficient security. Note
that all these groups have been disabled by default in all
hostapd/wpa_supplicant versions and these would be used only if
explicitly enabled in the configuration.
- Use strong passwords to prevent dictionary attacks
Signed-off-by: Stefan Lippers-Hollmann <s.l-h@gmx.de>
[bump PKG_RELEASE]
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Felix Fietkau