Fix SIGSEGV in rfc1035.c answer_request() line 1228 where memset()
is called with header & limit pointing at the same address and thus
tries to clear memory from before the buffer begins.
answer_request() is called with an invalid edns packet size provided by
the client. Ensure the udp_size provided by the client is bounded by
512 and configured maximum as per RFC 6891 6.2.3 "Values lower than 512
MUST be treated as equal to 512"
The client that exposed the problem provided a payload udp size of 0.
Signed-off-by: Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk> Acked-by: Hans Dedecker <dedeckeh@gmail.com>
Rafał Miłecki [Wed, 8 Mar 2017 11:38:43 +0000 (12:38 +0100)]
kernel: fix of_node handling in LEDs core code
This backports fixes for setting of_node and making it possible to read
extra info from DT. This was partially fixed by:
[PATCH] leds: leds-gpio: Set of_node for created LED devices
but it didn't work during initialization.
Refresh patches.
Minor update 704-phy-no-genphy-soft-reset.patch which was partially
accepted upstream.
Compile-tested on ar71xx.
Runtime-tested on ar71xx.
Fixes the following vulnerabilities:
- CVE-2017-7533 (4.4.80)
- CVE-2017-1000111 (4.4.82)
- CVE-2017-1000112 (4.4.82)
Signed-off-by: Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>
Since mbedtls 2.5.1, SHA1 has been disallowed in TLS certificates.
This breaks openvpn clients that try to connect to servers that
present a TLS certificate signed with SHA1, which is fairly common.
Run-tested with openvpn-mbedtls 2.4.3, LEDE 17.01.2, on ar71xx.
Rafał Miłecki [Mon, 7 Aug 2017 09:09:33 +0000 (11:09 +0200)]
base-files: don't setup network in preinit if failsafe is disabled
With failsafe disabled there is no point in early network setup. We
don't send announcement over UDP and there is no way to ssh to the
device.
A side effect of this is avoiding a possibly incorrect network config
(only with failsafe disabled). This problem is related to possible
changes made by user in /etc/config/network.
Thibaut VARENE [Fri, 4 Aug 2017 15:22:03 +0000 (17:22 +0200)]
ramips: ArcherC50v1: fix wlan2g MAC address
By default the wlan eprom contains the generic ralink MAC which is not
the vendor (TP-Link) one. Based on OFW bootlog, it appears that addresses
are decremented from the ethernet MAC.
This patch fixes the MAC address for wlan2g in line with OFW.
John Crispin [Tue, 1 Aug 2017 04:53:38 +0000 (06:53 +0200)]
ralink: fix rcu_sched stalls on mt7621
there were 2 bugs
*) core1 came up with a bad bogo mips, looks like the clock needed time to stabilize
*) HPT frequency was not set making r4k timers not come up properly
Backport of 9551d91b1d6 "ralink: fix rcu_sched stalls on mt7621".
01_leds had a workaround for the power led to compensate for the
inverted GPIO state. This patch was missing from my previous commit.
Signed-off-by: Thibaut VARENE <hacks@slashdirt.org>
[add the power led default-state which was omitted in the last commit
by me] Signed-off-by: Mathias Kresin <dev@kresin.me>
With d2b6bf141662 ("ramips: fix image validation errors") the board
name was changed to fix an image validation error. But this change
wasn't applied to all other files using the board name, which broke
sysupgrade.
Revert this change and use the former board name in the metadata
instead.
Make the behaviour of clk_get_rate consistent with common clk's
clk_get_rate by accepting NULL clocks as parameter. Some device
drivers rely on this, and will cause an OOPS otherwise.
Make the behaviour of clk_get_rate consistent with common clk's
clk_get_rate by accepting NULL clocks as parameter. Some device
drivers rely on this, and will cause an OOPS otherwise.
Make the behaviour of clk_get_rate consistent with common clk's
clk_get_rate by accepting NULL clocks as parameter. Some device
drivers rely on this, and will cause an OOPS otherwise.
This fixes the following security problems:
* CVE-2017-7407: https://curl.haxx.se/docs/adv_20170403.html
* CVE-2017-7468: https://curl.haxx.se/docs/adv_20170419.html
Paul Wassi [Sat, 22 Jul 2017 09:15:55 +0000 (11:15 +0200)]
ramips: DTS: VoCore2 improvements/fixes
The VoCore2 features 128MB of RAM, therefore set
memory in DTS to 128*1024*1024 = 0x8000000
The board's LED is connected to GND, set it to
ACTIVE_HIGH here.
Make serial console working again on kernel 4.9 by
change of pinmux configuration.
Piotr Dymacz [Fri, 31 Mar 2017 11:37:31 +0000 (13:37 +0200)]
ar71xx: add support for ALFA Network AP121F
ALFA Network AP121F is a pocket-size router dedicated for VPN/TOR users.
Device is based on Atheros AR9331 WiSoC and is running a custom version
(updated from OpenWrt CC to LEDE 17.01 release) of NetAidKit firmware.
Specification:
- 400/400/200 MHz (CPU/DDR/AHB)
- 64 MB of RAM (DDR1)
- 16 MB of FLASH (SPI NOR)
- 1x 10/100 Mbps Ethernet
- 1T1R 2.4 GHz
- 1x microSD (optional, on separate PCB)
- 3x LED, 1x button, 1x switch
- UART header on PCB
Flash instruction (under U-Boot web recovery mode):
1. Configure PC with static IP 192.168.1.2/24.
2. Connect PC with RJ45 port, press the reset button, power up device,
wait for first blink of all LEDs (indicates network setup), then keep
button for 3 following blinks and release it.
3. Open 192.168.1.1 address in your browser and upload sysupgrade image.
Even the commit message of the patch adding support for the MiWiFi Nano
says that a 16 MB flash chip is used. Extend the firmware partition to
make use of all available flash space.
Sergey Sergeev [Wed, 31 May 2017 08:00:01 +0000 (11:00 +0300)]
ar71xx: Fix UBIFS work on Mikrotik RB95x devices
If nand chip has no NAND_NO_SUBPAGE_WRITE flag on its options
ubifs can't use it mtd devices and the kernel crashes with error:
__nand_correct_data: uncorrectable ECC error
Mathias Kresin [Wed, 28 Jun 2017 21:36:37 +0000 (23:36 +0200)]
lantiq: use img file extension for DGN3500 factory images
The Netgear UI in basic mode refuses the upgrade file if the the
fileextension is not img. The expert/advanced mode accepts any
fileextension. Use img to make it work in any case.
Hans Dedecker [Mon, 26 Jun 2017 08:23:08 +0000 (10:23 +0200)]
dnsmasq: backport tweak ICMP ping logic for DHCPv4
Don't start ping-check of address in DHCP discover if there already
exists a lease for the address. It has been reported under some
circumstances android and netbooted windows devices can reply to
ICMP pings if they have a lease and thus block the allocation of
the IP address the device already has during boot.
Rafał Miłecki [Fri, 16 Jun 2017 11:23:22 +0000 (13:23 +0200)]
base-files: fix PKG_CONFIG_DEPENDS to include version.mk entries
Including version.mk sets PKG_CONFIG_DEPENDS to config entries used for
VERSION_SED command. We should keep these configs to make sure package
gets refreshed when needed.
Magnus Kroken [Thu, 22 Jun 2017 21:01:01 +0000 (23:01 +0200)]
openvpn: update to 2.4.3
Fixes for security and other issues. See security announcement for more details:
https://community.openvpn.net/openvpn/wiki/VulnerabilitiesFixedInOpenVPN243
* Remotely-triggerable ASSERT() on malformed IPv6 packet (CVE-2017-7508)
* Pre-authentication remote crash/information disclosure for clients (CVE-2017-7520)
* Potential double-free in --x509-alt-username (CVE-2017-7521)
* Remote-triggerable memory leaks (CVE-2017-7512)
* Post-authentication remote DoS when using the --x509-track option (CVE-2017-7522)
* Null-pointer dereference in establish_http_proxy_passthru()
* Restrict --x509-alt-username extension types
* Fix potential 1-byte overread in TCP option parsing
* Fix mbedtls fingerprint calculation
* openssl: fix overflow check for long --tls-cipher option
* Ensure option array p[] is always NULL-terminated
* Pass correct buffer size to GetModuleFileNameW() (Quarkslabs finding 5.6)
Magnus Kroken [Wed, 21 Jun 2017 19:05:09 +0000 (21:05 +0200)]
mbedtls: update to 2.5.1
Fixes some security issues (no remote exploits), and introduces
some changes. See release notes for details:
https://tls.mbed.org/tech-updates/releases/mbedtls-2.5.1-2.1.8-and-1.3.20-released
* Fixes an unlimited overread of heap-based buffers in mbedtls_ssl_read()
* Adds exponent blinding to RSA private operations
* Wipes stack buffers in RSA private key operations (rsa_rsaes_pkcs1_v15_decrypt(), rsa_rsaes_oaep_decrypt())
* Removes SHA-1 and RIPEMD-160 from the default hash algorithms for certificate verification.
* Fixes offset in FALLBACK_SCSV parsing that caused TLS server to fail to detect it sometimes.
* Tighten parsing of RSA PKCS#1 v1.5 signatures, to avoid a potential Bleichenbacher/BERserk-style attack.
This breaks local dns resolving in case noresolv=1 as resolv.conf is not
populated anymore with 127.0.0.1 as resolvfile does not equal
/tmp/resolv.conf.auto anymore.
When git-https request a service (e.g. github) which ask for credentials
git will pass this request to the user resulting download.pl to wait for
user input. Set GIT_ASKPASS to stop asking.
Jo-Philipp Wich [Thu, 8 Jun 2017 17:27:46 +0000 (19:27 +0200)]
base-files: network.sh: fix a number of IPv6 logic flaws
* Change network_get_subnet6() to sensibly guess a suitable prefix
Attempt to return the first non-linklocal, non-ula range, then attempt
to return the first non-linklocal range and finally fall back to the
previous behaviour of simply returning the first found item.
* Fix network_get_ipaddrs_all()
Instead of replicating the flawed logic appending a fixed ":1" suffix
to IPv6 addresses, rely on network_get_ipaddrs() and network_get_ipaddrs6()
to build a single list of all interface addresses.
* Fix network_get_subnets6()
Instead of replicating the flawed logic appending a fixed ":1" suffix
to IPv6 addresses, rely on the ipv6-prefix-assignment.local-address
field to figure out the proper network address.
Daniel Golle [Wed, 7 Jun 2017 17:39:33 +0000 (19:39 +0200)]
automake: import upstream fix for perl 5.26
Build broke as distributions now include Perl 5.26 and automake
triggered an "Unescaped left brace in regex" error.
Import upstream commit 13f00eb449 to fix that.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Jo-Philipp Wich [Thu, 8 Jun 2017 10:02:36 +0000 (12:02 +0200)]
base-files: network.sh: properly report local IPv6 addresses
Rework the network_get_ipaddr6() and network_get_ipaddrs6() functions to
fetch the effective local IPv6 address of delegated prefix from the
"local-address" field instead of naively hardcoding ":1" as static suffix.
Jo-Philipp Wich [Wed, 7 Jun 2017 19:24:41 +0000 (21:24 +0200)]
kernel: update kernel 4.4 to 4.4.71
Fixes the following security vulnerabilities:
CVE-2017-8890
The inet_csk_clone_lock function in net/ipv4/inet_connection_sock.c in the
Linux kernel through 4.10.15 allows attackers to cause a denial of service
(double free) or possibly have unspecified other impact by leveraging use
of the accept system call.
CVE-2017-9074
The IPv6 fragmentation implementation in the Linux kernel through 4.11.1
does not consider that the nexthdr field may be associated with an invalid
option, which allows local users to cause a denial of service (out-of-bounds
read and BUG) or possibly have unspecified other impact via crafted socket
and send system calls.
CVE-2017-9075
The sctp_v6_create_accept_sk function in net/sctp/ipv6.c in the Linux kernel
through 4.11.1 mishandles inheritance, which allows local users to cause a
denial of service or possibly have unspecified other impact via crafted
system calls, a related issue to CVE-2017-8890.
CVE-2017-9076
The dccp_v6_request_recv_sock function in net/dccp/ipv6.c in the Linux
kernel through 4.11.1 mishandles inheritance, which allows local users to
cause a denial of service or possibly have unspecified other impact via
crafted system calls, a related issue to CVE-2017-8890.
CVE-2017-9077
The tcp_v6_syn_recv_sock function in net/ipv6/tcp_ipv6.c in the Linux kernel
through 4.11.1 mishandles inheritance, which allows local users to cause a
denial of service or possibly have unspecified other impact via crafted
system calls, a related issue to CVE-2017-8890.
CVE-2017-9242
The __ip6_append_data function in net/ipv6/ip6_output.c in the Linux kernel
through 4.11.3 is too late in checking whether an overwrite of an skb data
structure may occur, which allows local users to cause a denial of service
(system crash) via crafted system calls.
x86 board.d only contains a case for the APU2, not the APU1. This
causes, for example, network configuration not to be created correctly.
Even though the APU1 seems to reaching EOL, there a still a lot of them
out there.
The APU1 and APU2 is configured in the same way and this patch should
also be considered for stable, as the error also exists there.
Signed-off-by: Kristian Evensen <kristian.evensen@gmail.com>
Mathias Kresin [Wed, 15 Feb 2017 07:39:05 +0000 (08:39 +0100)]
base-files: always set proto passed to _ucidef_set_interface()
Overwrite an already set proto if a new one is passed to
_ucidef_set_interface() similar to what is done for the interface.
It is required when using ""ucidef_set_interface_wan 'ptm0' 'pppoe'"
after some initial wan interface configuration is already done by
ucidef_add_switch.
The "json_is_a protocol string" guard is meant to not reset an earlier
set interface proto in case something like
"ucidef_set_interface_lan 'eth0'" is used afterwards.
Mathias Kresin [Fri, 24 Feb 2017 08:19:49 +0000 (09:19 +0100)]
lantiq: fix broadcasts and vlans in two iface mode
The two phy operation mode where one phy is assigned to an interface
without lantiq,* device tree property and the other phy is assigned to
an interface with the lantiq,wan device property was broken with the
multicast package leaks between vlans fixes.
Move the multicast packages relevant portmap settings to the condition
which handles multicast packages for better readability.
Replace the priv->port_map based port_map only for the interface which
has the lantiq,switch device tree property set, to allow tagged
multicast packages in two phy mode where the lantiq,switch device tree
property isn't used.
Using the lantiq,wan device tree property for one interface node and
the lantiq,switch device tree property for another interface node at
the same time was never intended/isn't supported at the moment.
The property is meant to be used in two phy operation mode where one
phy is assigned to an interface without lantiq,* device tree property
and the other phy is assigned to an interface with the lantiq,wan
device property to have two netdevs.
If both properties are used at the same time, the lantiq,wan interface
is shown as independent netdev but not able to operate independent. The
port needs to be managed via swconfig. These dependency is not obvious
and fooled already a lot of users.
Add a default WAN vlan for xrx200 devices having an ethernet WAN port
and remove the lantiq,wan device tree property. Leave it up to the user
to set the ethernet WAN port as default WAN interface or to use this
port as additional LAN port.
Jo-Philipp Wich [Thu, 1 Jun 2017 23:26:20 +0000 (01:26 +0200)]
umdns: remove superfluous include in init script
The umdns init script includes function/network.sh globally, outside of any
service procedure. This causes init script activation to fail in buildroot
and IB context if umdns is set to builtin.
Additionally, the network.sh helper is not actually used.
Drop the entire include in order to repair init script activation in build
host context. Fixes FS#658.
Alberto Bursi [Tue, 2 May 2017 17:31:17 +0000 (19:31 +0200)]
dnsmasq: make tftp root if not existing
If there's a TFTP root directory configured, create it with mkdir -p
(which does not throw an error if the folder exists already)
before starting dnsmasq. This is useful for TFTP roots in /tmp, for example.
Originally submitted by nfw user aka Nathaniel Wesley Filardo
Signed-off-by: Alberto Bursi <alberto.bursi@outlook.it>
Karl Vogel [Wed, 29 Mar 2017 09:39:35 +0000 (11:39 +0200)]
dnsmasq: use logical interface name for dhcp relay config
The relay section should use the logical interface name and
not the linux network device name directly. This to be
consistent with other sections of the dnsmasq config where
'interface' means the logical interface.
Piotr Dymacz [Mon, 29 May 2017 19:32:11 +0000 (21:32 +0200)]
ar71xx: fix Wallys DR344 GPIO-connected LEDs and button
This fixes wrong GPIO numbers for LEDs and button in Wallys DR344 board
and sets color of all LEDs to green as the mass production boards have
only green one.
Actually, DR344 has 6 GPIO-connected LEDs and one button:
Piotr Dymacz [Mon, 29 May 2017 17:18:52 +0000 (19:18 +0200)]
ar71xx: fix GE interface support in Wallys DR344
GMAC0 interface of AR9344 SOC in Wallys DR344 board is connected with
AR8035, not with AR8327. Without this fix, GE interface doesn't work at
all or shows high packet loss ratio.
Jo-Philipp Wich [Sun, 19 Feb 2017 18:04:09 +0000 (19:04 +0100)]
firewall: resync with master
Update to latest Git HEAD in order to import a number of fixes and other
improvements:
a4d98ae options: remove stray continue statement 3d2c18a options: improve handling of negations when parsing space separated values 0e5dd73 iptables: support -i, -o, -s and -d in option extra 4cb06c7 ubus: increase ubus network interface dump timeout e5dfc82 iptables: add exception handling f625954 firewall3: add check_snat() function 7d3d9dc firewall3: display the section type for UBUS rules 53ef9f1 firewall3: add UBUS support for include scripts 5cd4af4 firewall3: add UBUS support for ipset sections 02d6832 firewall3: add UBUS support for forwarding sections 0a7d36d firewall3: add UBUS support for redirect sections d44f418 firewall3: add fw3_attr_parse_name_type() function e264c8e firewall3: replace warn_rule() by warn_section() 6039c7f firewall3: check the return value of fw3_parse_options() c328d1f build: use -Wno-format-truncation instead of -Wno-error=format-truncation e06e537 utils: replace sprintf use with snprintf to avoid overflows 533f834 build: disable the format-truncation warning error to fix gcc 7 build errors e751cde zones: drop outgoing invalid traffic in masqueraded zones d596f72 rules: fix UCI context in error reporting 1d0564c ubus: fix interface name and proto lookup 82ccd9e firewall3: fix handling of UTC times 1949e0c iptables: support xtables API > 11
mac80211, hostapd: always explicitly set beacon interval
One of the latest mac80211 updates added sanity checks, requiring the
beacon intervals of all VIFs of the same radio to match. This often broke
AP+11s setups, as these modes use different default intervals, at least in
some configurations (observed on ath9k).
Instead of relying on driver or hostapd defaults, change the scripts to
always explicitly set the beacon interval, defaulting to 100. This also
applies the beacon interval to 11s interfaces, which had been forgotten
before. VIF-specific beacon_int setting is removed from hostapd.sh.
projects / thirdparty / openwrt.git / log
