Security fixes for:
* CVE-2020-10932
* a potentially remotely exploitable buffer overread in a DTLS client
* bug in DTLS handling of new associations with the same parameters
Full release announement:
https://tls.mbed.org/tech-updates/releases/mbedtls-2.16.6-and-2.7.15-released
Signed-off-by: Magnus Kroken <mkroken@gmail.com>
Some FullMAC cfg80211 wireless devices do not support virtual
interfaces, hence there is script logic to keep the existing network
device. Improve this to support renaming the interface if needed and
make sure the existing interface actually belongs to the right phy.
Change calls to 'iw' to avoid outputing warnings and errors to not
confuse users of such devices.
Also bump PKG_RELEASE which has been forgotten in the previous two
mac80211 changes.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Refreshed all patches, run tested on apalis.
Cc: Vladimir Vid <vladimir.vid@sartura.hr>
Cc: Tim Harvey <tharvey@gateworks.com>
Cc: Koen Vandeputte <koen.vandeputte@ncentric.com>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
If we know that we have an encrypted link (based on having had
a key configured for TX in the past) then drop all data frames
in the key selection handler if there's no key anymore.
This fixes an issue with mac80211 internal TXQs - there we can
buffer frames for an encrypted link, but then if the key is no
longer there when they're dequeued, the frames are sent without
encryption. This happens if a station is disconnected while the
frames are still on the TXQ.
Detecting that a link should be encrypted based on a first key
having been configured for TX is fine as there are no use cases
for a connection going from with encryption to no encryption.
With extended key IDs, however, there is a case of having a key
configured for only decryption, so we can't just trigger this
behaviour on a key being configured.
Cc: stable@vger.kernel.org
Reported-by: Jouni Malinen <j@w1.fi>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: David Bauer <mail@david-bauer.net>
Instead of using the actual interface name, a hard-coded 'wlan0' has
slipped into the script. Replace it.
Fixes: ccf2aa9d4b ("mac80211: detect existing interface before adding")
Reported-by: John Crispin <john@phrozen.org>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
32c717e jail: only mess with rootfs if CLONE_NEWNS was set
b275a62 instance: harmonize instance API
511fd97 jail: make /proc more secure
4953b7c jail: mount /sys read-only
a4d6442 jail: replace /etc/resolv.conf with symlink in extroot+overlay
a4cc165 jail: always mount /dev as additional tmpfs
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
This reverts commit 1b973b54ea.
It turns out act_police is included in the kmod-sched package so this
package turns out to be superfluous and causes file provision conflicts.
Ooooops! Best revert it then.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
The previous commit introduced a regression for netns jails without
jail_ifname set. Fix that.
Fixes: 4e4f7c6d2d ("netifd: network namespace jail improvements")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
aaaca2e interface: allocate and free memory for jail name
d93126d interface: allow renaming interface when moving to jail netns
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Removed sys/cdefs usage. The header is deprecated.
Removed canonicalize_file_name define. It's already fixed upstream.
Added --disable-debuginfod. Seems to be needed.
Modified patch 005 to build more stuff. It was failing before. It still
only builds libraries.
Modified patch 100 to use strerror under non-glibc. It is used under
glibc as strerror is not thread safe. It is under musl and uClibc-ng.
strerror_l is not available under uClibc-ng.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
CONFIG_WRITE functionality is not used and could be removed.
Looks helpful for devices with small flash because wpad is also affected.
Little testing shows that about 6 KB could be saved.
Signed-off-by: Kirill Lukonin <klukonin@gmail.com>
Updates the 88W8964 firmware used in the Linksys WRT3200ACM and WRT32X
[v9.3.2.6 -> v9.3.2.12]
Removes 0c43219 ("mwlwifi: Fix loading with backports v5.3")
as it has been merged upstream.
Unfortunately, there is a bug wherein Kaloz's repo, the version
detection mechanism for fixing vendor commands doesn't work.
It pulls in the Linux kernel version, which as of this time is
"4.14.y" or "4.19.y"
However, the proper behaviour is that it should pull in the mac80211
backports version which as of now is "5.4.27"
The included patch works around this using a backports define found
only on versions >5.3, "VENDOR_CMD_RAW_DATA".
Signed-off-by: Jose Olivera <oliverajeo@gmail.com>
Including the local build key in /etc/opkg/keys isn't feasible when
building on the buildbot: The included key collides with its copy
already in openwrt-keyring which breaks the ImageBuilder.
Not including a locally generated key also makes the base-files package
more reproducible.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
current preinit code in base-files doesn't config switch when there are
no port roles defined. But this kind of configuration exists on single
port devices where switch vlan is simply disabled.
configure reset and enable_vlan property when a switch node exist.
Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
The file /lib/functions/system.sh depends on find_mtd_index() and
find_mtd_part() located in /lib/function.sh, so let's source that
file.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
The default_postinst() function in /lib/functions.sh sources
/lib/functions/system.sh before cycling through uci-defaults files.
This creates a pseudo-cyclic dependency as system.sh also uses
functions that are located in functions.sh. Despite that, there
is actually only one uci-defaults file in the entire repo that needs
system.sh, and this one contains an explicit source for system.sh
anyway.
Consequently, this patch removes the sourcing of system.sh in
functions.sh. There are no relevant uses in packages, routing and
luci repositories.
This may require adjustments for downstream, though.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
This adds two fixes for compilation with kernel 5.4:
1. dev_open from include/linux/netdevice.h needs a second parameter
since kernel 5.0:
00f54e68924e ("net: core: dev: Add extack argument to dev_open()")
2. get_ds() macro definition has been dropped since kernel 5.1:
736706bee329 ("get rid of legacy 'get_ds()' function")
Since get_ds() has been just a macro before, replace it in
the driver instead of creating a version switch.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Since kernel 4.15, init_timer is not available anymore, and has been
replaced by timer_setup. The fixes compilation of wl_linuc.c, which
returned the following errors beforehand (line-wrapped manually):
.../broadcom-wl-5.10.56.27.3/driver/wl_linux.c: In function 'wl_init_timer':
.../broadcom-wl-5.10.56.27.3/driver/wl_linux.c:2576:2: error: implicit
declaration of function 'init_timer'; did you mean 'init_timers'?
[-Werror=implicit-function-declaration]
init_timer(&t->timer);
^~~~~~~~~~
init_timers
.../broadcom-wl-5.10.56.27.3/driver/wl_linux.c:2577:10: error:
'struct timer_list' has no member named 'data'
t->timer.data = (ulong) t;
^
.../broadcom-wl-5.10.56.27.3/driver/wl_linux.c:2578:20: error: assignment
to 'void (*)(struct timer_list *)' from incompatible pointer type
'void (*)(ulong)' {aka 'void (*)(long unsigned int)'}
[-Werror=incompatible-pointer-types]
t->timer.function = wl_timer;
This should fix build of several devices on bcm63xx with testing
kernel (4.19).
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Cell C RTL30VW is a LTE router with tho gigabit ethernets and integrated
QMI mPCIE modem.
This is stripped version of ASKEY RTL0030VW.
Hardware:
Specification:
-CPU: IPQ4019
-RAM: 256MB
-Flash: NAND 128MB + NOR 16MB
-WiFi: Integrated bgn/ac
-LTE: mPCIe card (Modem chipset MDM9230)
-LAN: 2 Gigabit Ports
-USB: 2x USB2.0
-Serial console: RJ-45 115200 8n1
-Unsupported VoIP
Known issues:
None so far.
Instruction install:
There are two methods: Factory web-gui and serial + tftp.
Web-gui:
1. Apply factory image via stock web-gui.
Serial + initramfs:
1. Rename OpenWrt initramfs image to "image"
2. Connect serial console (115200,8n1)
3. Set IP to different than 192.168.1.11, but 24 bit mask, eg. 192.168.1.4.
4. U-Boot commands:
sf probe && sf read 0x80000000 0x180000 0x10000
setenv serverip 192.168.1.4
set fdt_high 0x85000000
tftpboot 0x84000000 image
bootm 0x84000000
5. Install sysupgrade image via "sysupgrade -n"
Back to stock:
All is needed is swap 0x4c byte in mtd8 from 0 to 1 or 1 to 0,
do firstboot and factory reset with OFW:
1. read mtd8:
dd if=/dev/mtd8 of=/tmp/mtd8
2. go to tmp:
cd /tmp/
3. write first part of partition:
dd if=mtd8 of=mtd8.new bs=1 count=76
4. check which layout uses bootloader:
cat /proc/mtd
5a. If first are kernel_1 and rootfs_1 write 0:
echo -n -e '\x00' >> mtd8.new
5b. If first are kernel and rootfs write 1:
echo -n -e '\x01' >> mtd8.new
6. fill with rest of data:
dd if=mtd8 bs=1 skip=77 >> mtd8.new
7. CHECK IF mtd8.new HAVE CHANGED ONLY ONE BYTE! e.g with:
hexdump mtd8.new
8. write new mtd8 to flash:
mtd write mtd8.new /dev/mtd8
9. do firstboot
10.reboot
11. Do back to factory defaults in OFW GUI.
Based on work: Cezary Jackiewicz <cezary@eko.one.pl>
Signed-off-by: Pawel Dembicki <paweldembicki@gmail.com>
MobiPromo CM520-79F is an AC1300 dual band router based on IPQ4019
Specification:
SoC/Wireless: QCA IPQ4019
RAM: 512MiB
Flash: 128MiB SLC NAND
Ethernet PHY: QCA8075
Ethernet ports: 1x WAN, 2x LAN
LEDs: 7 LEDs
2 (USB, CAN) are GPIO
other 5 (2.4G, 5G, LAN1, LAN2, WAN) are connected to a shift register
Button: Reset
Flash instruction:
Disassemble the router, connect UART pins like this:
GND TX RX
[x x . . x .]
[. . . . . .]
(QCA8075 and IPQ4019 below)
Baud-rate: 115200
Set up TFTP server: IP 192.168.1.188/24
Power on the router and interrupt the booting with UART console
env backup (in case you want to go back to stock and need it there):
printenv
(Copy the output to somewhere save)
Set bootenv:
setenv set_ubi 'set mtdids nand0=nand0; set mtdparts mtdparts=nand0:0x7480000@0xb80000(fs); ubi part fs'
setenv bootkernel 'ubi read 0x84000000 kernel; bootm 0x84000000#config@1'
setenv cm520_boot 'run set_ubi; run bootkernel'
setenv bootcmd 'run cm520_boot'
setenv bootargs
saveenv
Boot initramfs from TFTP:
tftpboot openwrt-ipq40xx-generic-mobipromo_cm520-79f-initramfs-fit-zImage.itb
bootm
After initramfs image is booted, backup rootfs partition in case of reverting to stock image
cat /dev/mtd12 > /tmp/mtd12.bin
Then fetch it via SCP
Upload nand-factory.ubi to /tmp via SCP, then run
mtd erase rootfs
mtd write /tmp/*nand-factory.ubi rootfs
reboot
To revert to stock image, restore default bootenv in uboot UART console
setenv bootcmd 'bootipq'
printenv
use the saved dump you did back when you installed OpenWrt to verify that
there are no other differences from back in the day.
saveenv
upload the backed up mtd12.bin and run
tftpboot mtd12.bin
nand erase 0xb80000 0x7480000
nand write 0x84000000 0xb80000 0x7480000
The BOOTCONFIG may have been configured to boot from alternate partition (rootfs_1) instead
In case of this, set it back to rootfs:
cd /tmp
cat /dev/mtd7 > mtd7.bin
echo -ne '\x0b' | dd of=mtd7.bin conv=notrunc bs=1 count=1 seek=4
for i in 28 48 68 108; do
dd if=/dev/zero of=mtd7.bin conv=notrunc bs=1 count=1 seek=$i
done
mtd write mtd7.bin BOOTCONFIG
mtd write mtd7.bin BOOTCONFIG1
Signed-off-by: DENG Qingfang <dengqf6@mail2.sysu.edu.cn>
[renamed volume to ubi to support autoboot,
as per David Lam's test in PR#2432]
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
"Whoop whoop, sound of da police"
Add an ingress capable traffic policer module configurable with tc.
From the man page:
The police action allows to limit bandwidth of traffic matched by the
filter it is attached to. Basically there are two different algorithms
available to measure the packet rate: The first one uses an internal
dual token bucket and is configured using the rate, burst, mtu,
peakrate, overhead and linklayer parameters. The second one uses an
in-kernel sampling mechanism. It can be fine-tuned using the estimator
filter parameter.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
2188d81 jail: add support for launching extroot containers
6f3dbd2 jail: add support for userns and cgroupsns
28a06e5 jail: add support for (ram-)overlayfs
Add handling for extroot, overlaydir and tmpoverlaysize as well as
jail flags for userns and cgroupsns to OpenWrt's shell script to
allow their use in init scripts.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Newer versions of the kconfig program requires quoting the arguments of
the 'source' directive. These are the last ones not using them.
Signed-off-by: Eneas U de Queiroz <cotequeiroz@gmail.com>
IPv6 modules should all depend on @IPV6, to avoid circular dependencies
problems, especially if they select a module that depends on IPV6 as
well. In theory, if a package A depends on IPV6, any package doing
'select A' (DEPENDS+= A) should also depend on IPV6; otherwise selecting
A will fail. Sometimes the build system is forgiving this, but
eventually, and unexpectedly, it may blow up on some other commit.
Alternatively one can conditionally add IPv6 dependencies only if
CONFIG_IPV6 is selected: (DEPENDS+= +IPV6:package6).
Signed-off-by: Eneas U de Queiroz <cotequeiroz@gmail.com>
Since linux kernel commit 0768e17073dc527ccd18ed5f96ce85f9985e9115
(2019-04-19) the asm-generic/sockios.h header no longer defines
SIOCGSTAMP. Instead it provides only SIOCGSTAMP_OLD.
The linux/sockios.h header now defines SIOCGSTAMP using either
SIOCGSTAMP_OLD or SIOCGSTAMP_NEW as appropriate. This linux only
header file is not included so we get a build failure.
Signed-off-by: Norbert van Bolhuis <nvbolhuis@aimvalley.nl>
This commit add patch with 14c3:7610 pci id addition.
It was sent upstream.
Signed-off-by: Pawel Dembicki <paweldembicki@gmail.com>
[bumped PKG_RELEASE]
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Allows proper usage of the ss tool. Otherwise, several errors and bad
data gets thrown:
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Cannot open netlink socket: Protocol not supported
Originally reported here: https://github.com/openwrt/packages/issues/8232
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Use upstream latest git HEAD as it allows to remove the patches
700-radius-Prevent-buffer-overflow-in-rc_mksid,
701-pppd-Fix-bounds-check-in-EAP-code and
702-pppd-Ignore-received-EAP-messages-when-not-doing-EAP and
take in other fixes.
41a7323 pppd: Fixed spelling 'unkown' => 'unknown' (#141)
6b014be pppd: Print version information to stdout instead of stderr (#133)
cba2736 pppd: Add RFC1990 (Multilink) to the See Also section of the man page
f2f9554 pppd: Add mppe.h to the list of headers to install if MPPE is defined
ae54fcf pppd: Obfuscate password argument string
8d45443 pppd: Ignore received EAP messages when not doing EAP
8d7970b pppd: Fix bounds check in EAP code
858976b radius: Prevent buffer overflow in rc_mksid()
Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
The mlk5 kmod lacks all necessary build symbols
for kernel 4.14 (again).
Add missing symbols to avoid build failure on these targets.
Signed-off-by: Tan Zien <nabsdh9@gmail.com>
[rewrite commit message - reorder symbols]
Signed-off-by: David Bauer <mail@david-bauer.net>
Bump to 2.81rc5 and re-work ipset-remove-old-kernel-support.
More runtime kernel version checking is done in 2.81rc5 in various parts
of the code, so expand the ipset patch' scope to inlude those new areas
and rename to something a bit more generic.:wq
Upstream changes from rc4
532246f Tweak to DNSSEC logging.
8caf3d7 Fix rare problem allocating frec for DNSSEC.
d162bee Allow overriding of ubus service name.
b43585c Fix nameserver list in auth mode.
3f60ecd Fixed resource leak on ubus_init failure.
0506a5e Handle old kernels that don't do NETLINK_NO_ENOBUFS.
e7ee1aa Extend stop-dns-rebind to reject IPv6 LL and ULA addresses. We also reject the loopback address if rebind-localhost-ok is NOT set.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
201e359 Handle early exit when addrstring isn't set
fa4c464 Improve address logging on early exit messages (#83)
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
It seemed very confusing when trying to select the different variants of
hostapd which are somewhat scattered about under the menu 'Network'.
Moving all hostapd variants under a common submenu helps avoid
confusion.
Inspired-by: Kevin Mahoney <kevin.mahoney@zenotec.net>
[Fixup badly formatted patch, change menu name]
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
ab7a39a umdns: fix unused error
45c4953 dns: explicitly endian-convert all fields in header and question
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
Missing config symbols could lead to build failures on kernel
4.14/4.19.
Signed-off-by: Tan Zien <nabsdh9@gmail.com>
[rephrase commit message - reorder symbols]
Signed-off-by: David Bauer <mail@david-bauer.net>
gcc 8 & 9 appear to be more picky with regards access alignment to
packed structures, leading to this warning in dns.c:
dns.c:261:2: error: converting a packed ‘struct dns_question’ pointer
(alignment 1) to a ‘uint16_t’ {aka ‘short unsigned int’} pointer
(alignment 2) may result in an unaligned pointer value
[-Werror=address-of-packed-member]
261 | uint16_t *swap = (uint16_t *) q;
Work around what I think is a false positive by turning the warning off.
Not ideal, but not quite as not ideal as build failure.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
