openwrt/svn-archive/openwrt.git
10 years agotarget/imagebuilder: remove obsolete !TARGET_ROOTFS_INITRAMFS dependency
Felix Fietkau [Fri, 31 Oct 2014 12:59:42 +0000 (12:59 +0000)]
target/imagebuilder: remove obsolete !TARGET_ROOTFS_INITRAMFS dependency

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r43128

SVN-Revision: 43129

10 years agolantiq: vgv7519: fix profile, this board have a rt2800-pci board
John Crispin [Thu, 30 Oct 2014 08:13:05 +0000 (08:13 +0000)]
lantiq: vgv7519: fix profile, this board have a rt2800-pci board

Signed-off-by: Eddi De Pieri <eddi@depieri.net>
SVN-Revision: 43116

10 years agolantiq: vgv7519: fix board_config on dts, read mac address and set it
John Crispin [Wed, 29 Oct 2014 20:51:43 +0000 (20:51 +0000)]
lantiq: vgv7519: fix board_config on dts, read mac address and set it

Signed-off-by: Eddi De Pieri <eddi@depieri.net>
SVN-Revision: 43110

10 years agoBB: busybox: make high ASCII chars printable (#7993)
Jo-Philipp Wich [Mon, 27 Oct 2014 11:22:41 +0000 (11:22 +0000)]
BB: busybox: make high ASCII chars printable (#7993)

Currently busybox utils like "ls" fail to display filenames containing UTF-8
characters, replacing any special characters with "?".

Change libbb's printable_string() function to allow high ASCII characters so
that unicode filenames are displayed correctls.

Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
Backport of r43084

SVN-Revision: 43085

10 years agoBB: px5g-standalone: use /dev/urandom to initialize serial (#18232)
Jo-Philipp Wich [Mon, 27 Oct 2014 11:09:37 +0000 (11:09 +0000)]
BB: px5g-standalone: use /dev/urandom to initialize serial (#18232)

Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
Backport of r43080

SVN-Revision: 43083

10 years agoBB: uhttpd: fix HTTP incompatibilities in file handler
Jo-Philipp Wich [Mon, 27 Oct 2014 11:09:09 +0000 (11:09 +0000)]
BB: uhttpd: fix HTTP incompatibilities in file handler

 * Fixes sending an extraneous message body for 204 and 304 resoponses which
   breaks Chrome in keep-alive mode.

 * Adds mimetypes for JSON and JSONP.

Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
Backport of r43078

SVN-Revision: 43082

10 years agomac80211: fix key flush handling in AP+STA
Felix Fietkau [Mon, 27 Oct 2014 11:07:45 +0000 (11:07 +0000)]
mac80211: fix key flush handling in AP+STA

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r43079

SVN-Revision: 43081

10 years agompc85xx: fix WAN/LAN-MAC for TP-LINK TL-WDR4900 v1
John Crispin [Sun, 26 Oct 2014 20:33:34 +0000 (20:33 +0000)]
mpc85xx: fix WAN/LAN-MAC for TP-LINK TL-WDR4900 v1

This works around a bootloader issue where every device
has the same lan/wan-mac 00:04:9f:ef:01:01 - with this patch
we read the macs from config-partition during initial network
setup. We have 9 valid macs stored in the partition, the
1st two are used for the radios, 3 and 4 are now used for WAN/LAN.

on an already setup / running device we can get the real macs with
. /lib/functions.sh
. /lib/functions/system.sh
echo "LAN = $(mtd_get_mac_binary config 338)"
echo "WAN = $(mtd_get_mac_binary config 344)"

see:
https://dev.openwrt.org/ticket/14714

from the ticket / user klondike:
U-Boot passed this commit https://gitorious.org/0xlab-kernel/u-boot/commit/ecd1a09b81f2ed6e6ba7bd1d0bfb0cc3d0ea2ad0
http://u-boot.10912.n7.nabble.com/U-Boot-PATCH-mpc83xx-remove-hardcoded-network-addresses-from-config-files-td44372.html
I suppose to prevent this particular issue, but the WDR4900 may be using an old bootloader still affected.
                                                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

I have been checking the contents of the dtb on the flash, this particular bit is quite revealing:

ethernet@b0000 {
                        #address-cells = <0x1>;
                        #size-cells = <0x1>;
                        device_type = "network";
                        model = "eTSEC";
                        compatible = "fsl,etsec2";
                        fsl,num_rx_queues = <0x8>;
                        fsl,num_tx_queues = <0x8>;
                        local-mac-address = [00 00 00 00 00 00];
                                             ^^^^^^^^^^^^^^^^^

                        interrupt-parent = <0x2>;
                        phy-handle = <0x3>;
                        phy-connection-type = "rgmii-id";
                        ptimer-handle = <0x4>;

                        queue-group@0 {
                                #address-cells = <0x1>;
                                #size-cells = <0x1>;
                                reg = <0xb0000 0x1000>;
                                rx-bit-map = <0xff>;
                                tx-bit-map = <0xff>;
                                interrupts = <0x1d 0x2 0x1e 0x2 0x22 0x2>;
                        };
                };

I also have been checking the live device map to find this:

root@GHS-AP3:~# hexdump -C /proc/device-tree/soc@ffe00000/ethernet@b0000/local-mac-address
00000000  00 04 9f ef 01 01                                 |......|
00000006
root@GHS-AP3:~# hexdump -C /proc/device-tree/soc@ffe00000/ethernet@b1000/local-mac-address
*
root@GHS-AP3:~# hexdump -C /proc/device-tree/soc@ffe00000/ethernet@b2000/local-mac-address
*

My conclussion is that U-Boot most likely finds the device and (as no valid MAC-address is provided)
falls back to the default MAC provided by the old code, the kernel then receives thee modified
device map from U-Boot and assumes this is the correct MAC for the device despite it obviously isn't.

This can be seen at
target/linux/mpc85xx/patches-3.10/140-powerpc-85xx-tl-wdr4900-v1-support.patch

The enetaddr is filled up by using the device tree data by the process_boot_dtb
function and used by the platform_fixups function to set the eth0 address
(by calling dt_fixup_mac_address_by_alias("ethernet0", enetaddr); ).
But instead we should be used the device address which to my understanding is
provided in the mtd.

Signed-off-by: Bastian Bittorf <bittorf@bluebottle.com>
Backport of r43074

SVN-Revision: 43077

10 years agoBB: vgv7519: support more recent board release with vrx288 v1.2
John Crispin [Sun, 26 Oct 2014 17:59:55 +0000 (17:59 +0000)]
BB: vgv7519: support more recent board release with vrx288 v1.2

The existance of this new board has been confirmed and tested by the guy on this commit

https://github.com/openwrt-vgv7519/uboot-bin/commit/76fdd8e638f81e324ffe77856481bd147398f6e8

SVN-Revision: 43075

10 years agokernel: makefiles: make use of new var KERNEL_PATCHVER
Rafał Miłecki [Sun, 26 Oct 2014 13:27:27 +0000 (13:27 +0000)]
kernel: makefiles: make use of new var KERNEL_PATCHVER

Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Inspired by r43047

SVN-Revision: 43067

10 years agokernel: allow targets to specify KERNEL_PATCHVER instead of LINUX_VERSION
Rafał Miłecki [Sun, 26 Oct 2014 13:26:59 +0000 (13:26 +0000)]
kernel: allow targets to specify KERNEL_PATCHVER instead of LINUX_VERSION

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Backport of r43043

SVN-Revision: 43066

10 years agofindutils: add dependency on bison
Felix Fietkau [Sun, 26 Oct 2014 11:27:15 +0000 (11:27 +0000)]
findutils: add dependency on bison

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r43064

SVN-Revision: 43065

10 years agobase-files: add /etc/shadow to list of essential files to keep in sysupgrade (bug...
Felix Fietkau [Sat, 25 Oct 2014 17:23:14 +0000 (17:23 +0000)]
base-files: add /etc/shadow to list of essential files to keep in sysupgrade (bug 18206)

Based on bug #18206 sysupgrade can lead to loss of password information in
certain situations. Most likely all users who will upgrade from versions
r43017-43040, will lose their current passwords. :-(
https://dev.openwrt.org/ticket/18206

Currently /etc/shadow is defined as a conffile in base-files:
https://dev.openwrt.org/browser/trunk/package/base-files/Makefile#L37

But it is not defined in the default list of essential files to keep in
sysupgrade:
https://dev.openwrt.org/browser/trunk/package/base-files/files/lib/upgrade/keep.d/base-files-essential

If exporting conffiles info fails, /etc/shadow can get lost.

Shadow passwords are now the default, so saying that preserving /etc/passwd
is essential while /etc/shadow is not, makes no sense.

The attached patch adds /etc/shadow to the list of essential files.

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
Backport of r43061

SVN-Revision: 43062

10 years agoodhcp6c: avoid some unnecessary big mallocs
Steven Barth [Sat, 25 Oct 2014 10:39:31 +0000 (10:39 +0000)]
odhcp6c: avoid some unnecessary big mallocs

SVN-Revision: 43057

10 years agoath9k: fix regulatory notifier calls with CONFIG_ATH_USER_REGD=y (#18211)
Felix Fietkau [Sat, 25 Oct 2014 09:31:04 +0000 (09:31 +0000)]
ath9k: fix regulatory notifier calls with CONFIG_ATH_USER_REGD=y (#18211)

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 43055

10 years agomac80211: refresh patches
Felix Fietkau [Sat, 25 Oct 2014 09:31:01 +0000 (09:31 +0000)]
mac80211: refresh patches

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 43054

10 years agoath9k: backport tx power reporting fix
Felix Fietkau [Sat, 25 Oct 2014 09:30:59 +0000 (09:30 +0000)]
ath9k: backport tx power reporting fix

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 43053

10 years agokernel/adc: backport essential fixes for ad799x
John Crispin [Mon, 20 Oct 2014 06:32:40 +0000 (06:32 +0000)]
kernel/adc: backport essential fixes for ad799x

This patch ports back some fixes of the ad799x driver, which were
committed between 3.10 and 3.15.

Signed-off-by: Hartmut Knaack <knaack.h@gmx.de>
SVN-Revision: 42993

10 years agoopenssl: add ABI_VERSION to fix package rebuild issues (fixes #18169)
Felix Fietkau [Sun, 19 Oct 2014 16:54:33 +0000 (16:54 +0000)]
openssl: add ABI_VERSION to fix package rebuild issues (fixes #18169)

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42963

SVN-Revision: 42964

10 years agopolarssl: do not build the unused programs
Felix Fietkau [Sun, 19 Oct 2014 09:43:49 +0000 (09:43 +0000)]
polarssl: do not build the unused programs

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42960

SVN-Revision: 42961

10 years agoar71xx: disable irq on reboot to fix hang issues (fixes #17839)
Felix Fietkau [Sat, 18 Oct 2014 19:26:22 +0000 (19:26 +0000)]
ar71xx: disable irq on reboot to fix hang issues (fixes #17839)

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42955

SVN-Revision: 42956

10 years agopolarssl: add missing version bump
Felix Fietkau [Sat, 18 Oct 2014 09:20:57 +0000 (09:20 +0000)]
polarssl: add missing version bump

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42949

SVN-Revision: 42950

10 years agopolarssl: disable SSLv3 support, fixes CVE-2014-3566 (POODLE)
Felix Fietkau [Sat, 18 Oct 2014 09:19:51 +0000 (09:19 +0000)]
polarssl: disable SSLv3 support, fixes CVE-2014-3566 (POODLE)

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42947

SVN-Revision: 42948

10 years agohostapd: CVE-2014-3686 fixes
Steven Barth [Fri, 17 Oct 2014 06:16:07 +0000 (06:16 +0000)]
hostapd: CVE-2014-3686 fixes

Signed-off-by: Steven Barth <steven@midlink.org>
SVN-Revision: 42943

10 years agoBB: include: correctly predefine other version variables as well
Jo-Philipp Wich [Thu, 16 Oct 2014 16:13:16 +0000 (16:13 +0000)]
BB: include: correctly predefine other version variables as well

Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
SVN-Revision: 42933

10 years agoBB: include: predefine correct repo url template for 14.07
Jo-Philipp Wich [Thu, 16 Oct 2014 14:42:28 +0000 (14:42 +0000)]
BB: include: predefine correct repo url template for 14.07

Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
SVN-Revision: 42932

10 years agoBB: openssl: update to v1.0.1j (CVE-2014-3513, CVE-2014-3567, CVE-2014-3568)
Jo-Philipp Wich [Thu, 16 Oct 2014 08:34:38 +0000 (08:34 +0000)]
BB: openssl: update to v1.0.1j (CVE-2014-3513, CVE-2014-3567, CVE-2014-3568)

Also refresh patches and bump copyright year in Makefile.

Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
Backport of r42929

SVN-Revision: 42930

10 years agolantiq: lantiq: start using device-tree on ifxmips_pcie
John Crispin [Tue, 14 Oct 2014 11:03:53 +0000 (11:03 +0000)]
lantiq: lantiq: start using device-tree on ifxmips_pcie

Signed-off-by: Eddi De Pieri <eddi@depieri.net>
SVN-Revision: 42906

10 years agolantiq: vgv7519: fix irq and disable pcie
John Crispin [Tue, 14 Oct 2014 11:03:47 +0000 (11:03 +0000)]
lantiq: vgv7519: fix irq and disable pcie

Signed-off-by: Eddi De Pieri <eddi@depieri.net>
SVN-Revision: 42905

10 years agolantiq: lantiq: enable pci bus
John Crispin [Tue, 14 Oct 2014 11:03:36 +0000 (11:03 +0000)]
lantiq: lantiq: enable pci bus

Signed-off-by: Eddi De Pieri <eddi@depieri.net>
SVN-Revision: 42904

10 years agolantiq: mac80211: add support for rf5362 (needed by vgv7519 wifi)
John Crispin [Tue, 14 Oct 2014 11:03:30 +0000 (11:03 +0000)]
lantiq: mac80211: add support for rf5362 (needed by vgv7519 wifi)

Signed-off-by: Eddi De Pieri <eddi@depieri.net>
SVN-Revision: 42903

10 years agolantiq: vgv7519: extract wifi firmware from mtd
John Crispin [Tue, 14 Oct 2014 11:03:23 +0000 (11:03 +0000)]
lantiq: vgv7519: extract wifi firmware from mtd

Signed-off-by: Eddi De Pieri <eddi@depieri.net>
SVN-Revision: 42902

10 years agolantiq: lantiq: remove strip-fcs patch that seems to introduce regression on multiple...
John Crispin [Tue, 14 Oct 2014 11:03:16 +0000 (11:03 +0000)]
lantiq: lantiq: remove strip-fcs patch that seems to introduce regression on multiple devices

Signed-off-by: Eddi De Pieri <eddi@depieri.net>
SVN-Revision: 42901

10 years agolantiq: lantiq: fix pinctrl-xway
John Crispin [Tue, 14 Oct 2014 11:03:11 +0000 (11:03 +0000)]
lantiq: lantiq: fix pinctrl-xway

Signed-off-by: Eddi De Pieri <eddi@depieri.net>
SVN-Revision: 42900

10 years agolantiq: vgv7519: backport vgv7519 dts
John Crispin [Tue, 14 Oct 2014 11:03:07 +0000 (11:03 +0000)]
lantiq: vgv7519: backport vgv7519 dts

Signed-off-by: Eddi De Pieri <eddi@depieri.net>
SVN-Revision: 42899

10 years agompc85xx: use wpad-mini in WDR4900 profile instead of wpad
John Crispin [Tue, 14 Oct 2014 10:36:08 +0000 (10:36 +0000)]
mpc85xx: use wpad-mini in WDR4900 profile instead of wpad

This change makes the profile of WDR4900 more consistent with other router
specific profiles.

Signed-off-by: Philipp Borgers <borgers@mi.fu-berlin.de>
SVN-Revision: 42898

10 years agoprocd: fix initial nand sysupgrade
Felix Fietkau [Sun, 12 Oct 2014 15:51:35 +0000 (15:51 +0000)]
procd: fix initial nand sysupgrade

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42884

SVN-Revision: 42887

10 years agoprocd: fix parsing of the has_env parameter
Felix Fietkau [Sun, 12 Oct 2014 15:51:33 +0000 (15:51 +0000)]
procd: fix parsing of the has_env parameter

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42883

SVN-Revision: 42886

10 years agomac80211: remove NO-IR from channel 36-48
Felix Fietkau [Sun, 12 Oct 2014 15:51:31 +0000 (15:51 +0000)]
mac80211: remove NO-IR from channel 36-48

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42882

SVN-Revision: 42885

10 years agoodhcp6c: fix parse errors with some dhcpv6 servers
Steven Barth [Sun, 12 Oct 2014 12:18:58 +0000 (12:18 +0000)]
odhcp6c: fix parse errors with some dhcpv6 servers

SVN-Revision: 42870

10 years agoodhcp6c: fix typo in last commit
Steven Barth [Wed, 8 Oct 2014 15:27:38 +0000 (15:27 +0000)]
odhcp6c: fix typo in last commit

SVN-Revision: 42853

10 years agoodhcp6c: don't assign addresses or prefixes prematurely
Steven Barth [Wed, 8 Oct 2014 15:16:41 +0000 (15:16 +0000)]
odhcp6c: don't assign addresses or prefixes prematurely

SVN-Revision: 42851

10 years agocns3xxx: Adopt irq_domain support for cns3xxx gpio driver
Felix Fietkau [Wed, 8 Oct 2014 10:29:32 +0000 (10:29 +0000)]
cns3xxx: Adopt irq_domain support for cns3xxx gpio driver

Have gpio driver adopt irqdomain support so that there are
non-overlapping allocations of irq numbers mapped to gpio's.

Signed-off-by: Pushpal Sidhu <psidhu@gateworks.com>
Backport of r42844

SVN-Revision: 42847

10 years agocns3xxx: fix shared PCI interrupt mapping
Felix Fietkau [Wed, 8 Oct 2014 10:29:30 +0000 (10:29 +0000)]
cns3xxx: fix shared PCI interrupt mapping

This patch originally failed to combine INTA/B/C/D onto a single ARM CPU
interrupt. Instead, it mapped INTA/B/C and excluded D. This patch
corrects the issue by mapping all four interrupts to the single ARM CPU
interrupt. The original intent of the patch still holds as the newer PCB
take advantage of isolated interrupts. This fix only applies to older
PCB's that do not route INTA/B/C/D to unique external ARM CPU
interrupts.

Signed-off-by: Pushpal Sidhu <psidhu@gateworks.com>
Backport of r42830

SVN-Revision: 42846

10 years agocns3xxx: fix register for enabling MMC/SD pins
Felix Fietkau [Wed, 8 Oct 2014 10:29:27 +0000 (10:29 +0000)]
cns3xxx: fix register for enabling MMC/SD pins

also fixes a GPIO related regression from the upgrade to 3.10

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42829

SVN-Revision: 42845

10 years agoodhcp6c: ensure signal-safety of signal handlers
Steven Barth [Wed, 8 Oct 2014 10:00:40 +0000 (10:00 +0000)]
odhcp6c: ensure signal-safety of signal handlers

SVN-Revision: 42843

10 years agoodhcp6c: add route-workaround for broken IPv6-ISPs
Steven Barth [Wed, 8 Oct 2014 10:00:36 +0000 (10:00 +0000)]
odhcp6c: add route-workaround for broken IPv6-ISPs

Some ISP seem to only do stateful DHCPv6 and not sending RAs.
This is technically broken because plain DHCPv6 doesn't carry routes.

We work around here by faking a default route to the DHCPv6 server
if we do not receive a useful RA from the ISP.

This workaround can be turned off with: option fakeroutes 0

Signed-off-by: Steven Barth <steven@midlink.org>
SVN-Revision: 42842

10 years agoralink: add dwc2 reset patch
John Crispin [Mon, 6 Oct 2014 06:11:37 +0000 (06:11 +0000)]
ralink: add dwc2 reset patch

this allows people building custom images to use dwc2 instead of dwc_otg

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 42801

10 years agohostapd: merge an upstream patch for pmksa cache
Felix Fietkau [Sun, 5 Oct 2014 10:39:52 +0000 (10:39 +0000)]
hostapd: merge an upstream patch for pmksa cache

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42762

SVN-Revision: 42763

10 years agomap: fix portsets starting with 0 and use regular NAT for 1:1 MAP
Steven Barth [Thu, 2 Oct 2014 19:16:17 +0000 (19:16 +0000)]
map: fix portsets starting with 0 and use regular NAT for 1:1 MAP

Signed-off-by: Steven Barth <steven@midlink.org>
SVN-Revision: 42742

10 years agobrcm2708: update 3.10 patches with raspberrypi/rpi-3.10.y of 27 Apr. 2014
Florian Fainelli [Sat, 27 Sep 2014 19:13:32 +0000 (19:13 +0000)]
brcm2708: update 3.10 patches with raspberrypi/rpi-3.10.y of 27 Apr. 2014

Backport r42678

Signed-off-by: Florian Fainelli <florian@openwrt.org>
SVN-Revision: 42682

10 years agobrcm2708-gpu-fw: update to Sep. 27 2014 snapshot
Florian Fainelli [Sat, 27 Sep 2014 17:43:54 +0000 (17:43 +0000)]
brcm2708-gpu-fw: update to Sep. 27 2014 snapshot

Backport of r42676

Signed-off-by: Florian Fainelli <florian@openwrt.org>
SVN-Revision: 42677

10 years agoodhcp6c: work around weird ISP RS behaviour
Steven Barth [Sat, 27 Sep 2014 16:58:15 +0000 (16:58 +0000)]
odhcp6c: work around weird ISP RS behaviour

SVN-Revision: 42675

10 years agoodhcpd: fix segfault when parsing domain options in UCI
Steven Barth [Thu, 25 Sep 2014 11:54:08 +0000 (11:54 +0000)]
odhcpd: fix segfault when parsing domain options in UCI

SVN-Revision: 42664

10 years agotoolchain: remove another bogus copy of libiberty.a
Felix Fietkau [Sat, 20 Sep 2014 11:12:19 +0000 (11:12 +0000)]
toolchain: remove another bogus copy of libiberty.a

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42624

SVN-Revision: 42625

10 years agoBB: firewall: allow '*' as synonym for any / all in family and proto options
Jo-Philipp Wich [Fri, 19 Sep 2014 18:23:20 +0000 (18:23 +0000)]
BB: firewall: allow '*' as synonym for any / all in family and proto options

Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
Backport of r42620

SVN-Revision: 42623

10 years agoBB: firewall: fix heap corruption in fw3_bitlen2netmask() with IPv6 addresses
Jo-Philipp Wich [Fri, 19 Sep 2014 18:23:17 +0000 (18:23 +0000)]
BB: firewall: fix heap corruption in fw3_bitlen2netmask() with IPv6 addresses

Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
Backport of r42610

SVN-Revision: 42622

10 years agoBB: firewall: fix invalid memory access when processing /128 IPv6 addresses from...
Jo-Philipp Wich [Fri, 19 Sep 2014 18:23:15 +0000 (18:23 +0000)]
BB: firewall: fix invalid memory access when processing /128 IPv6 addresses from ubus, properly emit REDIRECT rules for local port forwards

Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
Backport of r42604

SVN-Revision: 42621

10 years agocyassl: update to version 3.2.0
John Crispin [Thu, 18 Sep 2014 15:24:45 +0000 (15:24 +0000)]
cyassl: update to version 3.2.0

This fixes a security problem:
Security fix for RSA Padding check vulnerability

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Backport of r42526

SVN-Revision: 42614

10 years agoppp: update to version 2.4.7
John Crispin [Thu, 18 Sep 2014 15:24:22 +0000 (15:24 +0000)]
ppp: update to version 2.4.7

This fixes: CVE-2014-3158 and some other bugs.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Backport of r42525

SVN-Revision: 42613

10 years agocurl: 7.36.0 -> 7.38.0
John Crispin [Thu, 18 Sep 2014 15:24:07 +0000 (15:24 +0000)]
curl: 7.36.0 -> 7.38.0

Main changes:
- URL parser: IPv6 zone identifiers are now supported
- cyassl: Use error-ssl.h when available (drop local patch)
- polarssl: support CURLOPT_CAPATH / --capath
- mkhelp: generate code for --disable-manual as well (drop local patch)

Full release notes: http://curl.haxx.se/changes.html

MIPS 34kc binary size:
- 7.36.0 before: 82,539 bytes
- 7.38.0 after: 83,321 bytes

Signed-off-by: Catalin Patulea <cat@vv.carleton.ca>
Backport of r42517

SVN-Revision: 42612

10 years agoubus: update to latest version, adds various fixes and improvements
Felix Fietkau [Wed, 17 Sep 2014 11:34:40 +0000 (11:34 +0000)]
ubus: update to latest version, adds various fixes and improvements

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42594

SVN-Revision: 42595

10 years agomdns: update to latest git HEAD
John Crispin [Wed, 17 Sep 2014 11:01:26 +0000 (11:01 +0000)]
mdns: update to latest git HEAD

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 42593

10 years agosdk: predefine kernel version magic
John Crispin [Wed, 17 Sep 2014 11:01:24 +0000 (11:01 +0000)]
sdk: predefine kernel version magic

Kernel modules which are built using the SDK do not have access
to the .vermagic file in the Linux build dir, therefore the
generated .ipk metadata depends on kernel versions like
'3.10.49-1-unknown' which are not satisfiable by the kernel
meta package in the image.

Fix this problem by substituting 'unknown' with the current
version magic in include/kernel.mk when packing the SDK.

Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
Backport of r42590

SVN-Revision: 42592

10 years agohostapd: do not remove foreign wpa_supplicant sockets
John Crispin [Wed, 17 Sep 2014 11:01:23 +0000 (11:01 +0000)]
hostapd: do not remove foreign wpa_supplicant sockets

https://dev.openwrt.org/ticket/17886

Signed-off-by: John Crispin <blogic@openwrt.org>
Backport of r42586

SVN-Revision: 42591

10 years agoralink: disable wifi on DIR-810L and WHR-300HP2
John Crispin [Wed, 17 Sep 2014 07:06:12 +0000 (07:06 +0000)]
ralink: disable wifi on DIR-810L and WHR-300HP2

both units have a lna/pna combo that the driver so far does not support

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 42585

10 years agotoolchain/binutils: do not install host libiberty.a into target lib dir
John Crispin [Wed, 17 Sep 2014 06:29:42 +0000 (06:29 +0000)]
toolchain/binutils: do not install host libiberty.a into target lib dir

Binutils target build will pick it up and break build on e.g. arm:

libtool: link: arm-openwrt-linux-uclibcgnueabi-gcc -shared  .libs/archive.o .libs/archures.o .libs/bfd.o .libs/bfdio.o .libs/bfdwin.o .libs/cache.o .libs/coffgen.o .libs/corefile.o .libs/format.o .libs/init.o .libs/libbfd.o .libs/opncls.o .libs/reloc.o .libs/section.o .libs/syms.o .libs/targets.o .libs/hash.o .libs/linker.o .libs/srec.o .libs/binary.o .libs/tekhex.o .libs/ihex.o .libs/stabs.o .libs/stab-syms.o .libs/merge.o .libs/dwarf2.o .libs/simple.o .libs/compress.o .libs/verilog.o .libs/elf32-arm.o .libs/elf32.o .libs/elf-nacl.o .libs/elf-vxworks.o .libs/elf.o .libs/elflink.o .libs/elf-attrs.o .libs/elf-strtab.o .libs/elf-eh-frame.o .libs/dwarf1.o .libs/elf32-gen.o .libs/cpu-arm.o   -L/mnt/dl/slave/at91/build/staging_dir/target-arm_arm926ej-s_uClibc-0.9.33.2_eabi/usr/lib -L/mnt/dl/slave/at91/build/staging_dir/target-arm_arm926ej-s_uClibc-0.9.33.2_eabi/lib -L/mnt/dl/slave/at91/build/staging_dir/toolchain-arm_arm926ej-s_gcc-4.8-linaro_uClibc-0.9.33.2_eabi/usr/lib -L/mnt/dl/slave/at91/build/staging_dir/toolchain-arm_arm926ej-s_gcc-4.8-linaro_uClibc-0.9.33.2_eabi/lib -L/mnt/dl/slave/at91/build/build_dir/target-arm_arm926ej-s_uClibc-0.9.33.2_eabi/binutils-2.24/bfd/../libiberty/pic -liberty -lz  -march=armv5te -mtune=arm926ej-s -mfloat-abi=soft -Wl,-lc -Wl,--as-needed -Wl,-lm -Wl,--no-as-needed   -Wl,-soname -Wl,libbfd-2.24.so -o .libs/libbfd-2.24.so
/mnt/dl/slave/at91/build/staging_dir/toolchain-arm_arm926ej-s_gcc-4.8-linaro_uClibc-0.9.33.2_eabi/lib/libiberty.a: could not read symbols: File format not recognized

Signed-off-by: Jonas Gorski <jogo@openwrt.org>
Backport of r42430

SVN-Revision: 42584

10 years agoubox: update to latest git HEAD
John Crispin [Tue, 16 Sep 2014 18:35:15 +0000 (18:35 +0000)]
ubox: update to latest git HEAD

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 42582

10 years agoprocd: update to latest git HEAD
John Crispin [Tue, 16 Sep 2014 18:35:09 +0000 (18:35 +0000)]
procd: update to latest git HEAD

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 42581

10 years agoLantiq: fix build warning from dsl_fs script
John Crispin [Tue, 16 Sep 2014 18:35:05 +0000 (18:35 +0000)]
Lantiq: fix build warning from dsl_fs script

fix inclusion of functions.sh in dsl_fs init script

without this, the following error is seen during build:
./etc/init.d/dsl_fs: line 4: /lib/functions.sh: No such file or directory

Signed-off-by: Thomas Langer <thomas.langer@lantiq.com>
Backport of r42575

SVN-Revision: 42580

10 years agofs-tools: dont use mips16
John Crispin [Tue, 16 Sep 2014 18:35:03 +0000 (18:35 +0000)]
fs-tools: dont use mips16

this causes problems on 47xx/74k

Signed-off-by: John Crispin <blogic@openwrt.org>
Backport of r42570

SVN-Revision: 42579

10 years agolantiq: fix switch init on vr9 SoC (3.10)
John Crispin [Tue, 16 Sep 2014 18:35:00 +0000 (18:35 +0000)]
lantiq: fix switch init on vr9 SoC (3.10)

Signed-off-by: John Crispin <blogic@openwrt.org>
Backport of r42544
Backport of r42551
Backport of r42574

SVN-Revision: 42578

10 years agocomgt: add directip proto
John Crispin [Tue, 16 Sep 2014 18:34:44 +0000 (18:34 +0000)]
comgt: add directip proto

add a proto handler for sierra wireless directip modems

Signed-off-by: John Crispin <blogic@openwrt.org>
Backport of r42465

SVN-Revision: 42577

10 years agohostapd: remove bogus default setting for wps_pin (#17873)
Felix Fietkau [Mon, 15 Sep 2014 16:09:36 +0000 (16:09 +0000)]
hostapd: remove bogus default setting for wps_pin (#17873)

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42553

SVN-Revision: 42554

10 years agoar71xx: ack completed tx descriptors only after the full frame has been completed
Felix Fietkau [Thu, 11 Sep 2014 15:14:05 +0000 (15:14 +0000)]
ar71xx: ack completed tx descriptors only after the full frame has been completed

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42457

SVN-Revision: 42475

10 years agoatheros: enable the low_mem flag to better deal with 16M RAM devices
Felix Fietkau [Wed, 10 Sep 2014 13:50:53 +0000 (13:50 +0000)]
atheros: enable the low_mem flag to better deal with 16M RAM devices

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42460

SVN-Revision: 42461

10 years agoprocd: allow adding empty strings to arrays
Felix Fietkau [Wed, 10 Sep 2014 12:57:07 +0000 (12:57 +0000)]
procd: allow adding empty strings to arrays

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42456

SVN-Revision: 42458

10 years agoath9k: fix pll clock initialization on newer soc devices (fixes #14916)
Felix Fietkau [Tue, 9 Sep 2014 08:03:34 +0000 (08:03 +0000)]
ath9k: fix pll clock initialization on newer soc devices (fixes #14916)

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42453

SVN-Revision: 42454

10 years agoath9k: add some ani related stability / performance fixes
Felix Fietkau [Mon, 8 Sep 2014 16:45:58 +0000 (16:45 +0000)]
ath9k: add some ani related stability / performance fixes

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42443

SVN-Revision: 42444

10 years agomac80211: add mt7620a/n to the rt2x00 depends list
John Crispin [Mon, 8 Sep 2014 15:14:03 +0000 (15:14 +0000)]
mac80211: add mt7620a/n to the rt2x00 depends list

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 42441

10 years agonetifd: work-around kernel IPv6 on-link route issue
Steven Barth [Mon, 8 Sep 2014 14:47:26 +0000 (14:47 +0000)]
netifd: work-around kernel IPv6 on-link route issue

SVN-Revision: 42440

10 years agolantiq: add ath9k led pin support
John Crispin [Mon, 8 Sep 2014 14:30:00 +0000 (14:30 +0000)]
lantiq: add ath9k led pin support

Signed-off-by: John Crispin <blogic@openwrt.org>
Backport of r42437

SVN-Revision: 42438

10 years agoath79: dev-eth: Don't advertise 1gbit in link code word on ar9331
Felix Fietkau [Sun, 7 Sep 2014 09:46:41 +0000 (09:46 +0000)]
ath79: dev-eth: Don't advertise 1gbit in link code word on ar9331

While the AR9331 has a gigabit MAC towards the internal switch, the
integrated PHYs however are only 100-base-tx capable.  The existing code
however advertieses gigabit capability in the link status word.  If you
attach such a PHY to a gigabit capable switch on the remote end, with
some probability it attempts to negotiate gigabit and fails, falling
baco to the AR9331 assuming a 10mbit half-duplex link.  This has been
observed quite frequently with the Carambola2 and gigabit capable
switches.

In ath79_register_eth(), "pdata->has_gbit = 1;" is set unconditionally
for both AR9331 ethernet ports. This is most likely wrong. Despite the
two MAC IP cores being gigabit MACs, the MAC for eth1 is connected to a
100base-T PHY via MII. The has_gbit attribute is used in the ethernet
driver to determine the supported link modes.

So either pdata->has_gbit is not set to 1 anymore, or the ethernet
driver needs to be modified to determine the advertised link code word
on another criteria than pdata->has_gbit.  This patch implements the
former solution.

Signed-off-by: Harald Welte <laforge@gnumonks.org>
Backport of r42432

SVN-Revision: 42434

10 years agoar71xx: merge ethernet tx descriptor splitting from trunk (as of r42429) to fix ar716...
Felix Fietkau [Sun, 7 Sep 2014 09:46:39 +0000 (09:46 +0000)]
ar71xx: merge ethernet tx descriptor splitting from trunk (as of r42429) to fix ar716x performance issues

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
SVN-Revision: 42433

10 years agouclient: update to the latest version, fixes some crash issues
Felix Fietkau [Thu, 4 Sep 2014 17:46:39 +0000 (17:46 +0000)]
uclient: update to the latest version, fixes some crash issues

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42419

SVN-Revision: 42420

10 years agomtd: disable trx_check() for ar71xx, fixes a regression introduced in r42403
Felix Fietkau [Tue, 2 Sep 2014 21:06:59 +0000 (21:06 +0000)]
mtd: disable trx_check() for ar71xx, fixes a regression introduced in r42403

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42405

SVN-Revision: 42406

10 years agomtd: Fix trx check after partition rename ("linux" to "firmware")
Felix Fietkau [Tue, 2 Sep 2014 17:14:24 +0000 (17:14 +0000)]
mtd: Fix trx check after partition rename ("linux" to "firmware")

On brcm47xx (Huawei E970), I noticed that sysupgrade now tries to flash trx images which don't fit into the flash, resulting in non-booting system.

I found that mtd does trx size checking only when flashing to a partition called "linux". This patch changes this to be "firmware".

This matches the following patch which changed partition name in bcm47xxpart.c driver to "firmware":

commit 86b4d5ef68ca77ff6724ebb9bddd6b93239c87fc
Author: hauke <hauke@3c298f89-4303-0410-b956-a3cf2f4a3e73>
 Date:   Sun Dec 1 16:32:32 2013 +0000

    brcm47xx: use "firmware" partition name
    ...
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
    Backport of r38973

Signed-off-by: Mathias Adam <m.adam--openwrt@adamis.de>
Backport of r42403

SVN-Revision: 42404

10 years agocns3xxx: support isolated PCI interrupts on newer Laguna PCBs
Felix Fietkau [Tue, 2 Sep 2014 17:12:34 +0000 (17:12 +0000)]
cns3xxx: support isolated PCI interrupts on newer Laguna PCBs

The cns3xxx uses irq61 for pcie0_intr which in the case of a PCIe-to-PCI
bridge ends up combining INTA/B/C/D on a single ARM CPU interrupt. This is
not optimal when you have multiple cores. To overcome this limitation an
enhancement was made on newer Laguna PCB's that support miniPCI cards
to route the INTA/B/C/D signals to unique external ARM CPU interrupts which
can help balance CPU core utilization and in some cases increase overall
system performance or responsiveness.

For more details see:
 http://trac.gateworks.com/wiki/multicoreprocessing#PCIInterruptsteering

Signed-off-by: Tim Harvey <tharvey@gateworks.com>
Backport of r42400

SVN-Revision: 42402

10 years agoBB: point old.packages at the real tree
John Crispin [Tue, 2 Sep 2014 15:51:05 +0000 (15:51 +0000)]
BB: point old.packages at the real tree

Signed-off-by: John Crispin <blogic@openwrt.org>
SVN-Revision: 42399

10 years agosdk: keep $(STAGING_DIR)/usr/bin, at least curl needs it
Felix Fietkau [Mon, 1 Sep 2014 22:34:49 +0000 (22:34 +0000)]
sdk: keep $(STAGING_DIR)/usr/bin, at least curl needs it

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Backport of r42397

SVN-Revision: 42398

10 years agomac80211: rt2x00 correctly set ht20/ht40 filter
John Crispin [Mon, 1 Sep 2014 13:22:16 +0000 (13:22 +0000)]
mac80211: rt2x00 correctly set ht20/ht40 filter

Simple patch to correct ht20/ht40 switching. Tested with rt3290/rt3352/rt5350

Signed-off-by: Serge Vasilugin <vasilugin@yandex.ru>
Backport of r42348

SVN-Revision: 42396

10 years agoramips: fix switch config for Edimax 3G-6200n
John Crispin [Mon, 1 Sep 2014 13:22:14 +0000 (13:22 +0000)]
ramips: fix switch config for Edimax 3G-6200n

Switch in Edimax 3G-6200n also require full switch reset, not only vlan definitions. Tested on Edimax 3G-6200n.

Signed-off-by: Cezary Jackiewicz <cezary.jackiewicz@gmail.com>
Backport of r42194

SVN-Revision: 42395

10 years agoramips: Hame: change LED name mpr-x -> hame
John Crispin [Mon, 1 Sep 2014 13:22:13 +0000 (13:22 +0000)]
ramips: Hame: change LED name mpr-x -> hame

Signed-off-by: Cezary Jackiewicz <cezary.jackiewicz@gmail.com>
Backport of r42190

SVN-Revision: 42394

10 years agoadd support for the NC-LINK SMART-300
John Crispin [Mon, 1 Sep 2014 13:22:12 +0000 (13:22 +0000)]
add support for the NC-LINK SMART-300

Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
Backport of r42130

SVN-Revision: 42393

10 years agoAdd 'subversion' build dependency to README
John Crispin [Mon, 1 Sep 2014 13:22:10 +0000 (13:22 +0000)]
Add 'subversion' build dependency to README

Fixes #17725.

Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Backport of r42343

SVN-Revision: 42392

10 years agoralink: DIR-645 picks up incorrect MAC address in DTS file
John Crispin [Mon, 1 Sep 2014 13:22:09 +0000 (13:22 +0000)]
ralink: DIR-645 picks up incorrect MAC address in DTS file

Signed-off-by: John Crispin <blogic@openwrt.org>
Backport of r42321

SVN-Revision: 42391

10 years agodnsmasq: Make the --dhcp-host logic easier to understand
John Crispin [Mon, 1 Sep 2014 13:22:08 +0000 (13:22 +0000)]
dnsmasq: Make the --dhcp-host logic easier to understand

Use an if/else statement to cover the two different syntaxes.  Add
comments explaining what the end results should look like.

This patch should not change the script's output.

Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Backport of r42320

SVN-Revision: 42390

10 years agodnsmasq: Fix hosts file format when MAC address is not specified
John Crispin [Mon, 1 Sep 2014 13:22:06 +0000 (13:22 +0000)]
dnsmasq: Fix hosts file format when MAC address is not specified

An entry like this in /etc/config/dhcp:

    config 'host'
        option 'name' 'pc2'
        option 'ip' '192.168.100.56'
        option 'dns' '1'

results in a /tmp/hosts/dhcp entry that looks like this:

    192.168.100.56 .lan

Obviously it should say "pc2.lan".

This happens because $name is set to "" in order to support the MAC-less
syntax: "--dhcp-host=lap,192.168.0.199".  Fix this by reordering the
operations.  Also, refuse to add a DNS entry if the hostname or IP is
missing.

Fixes #17683

Reported-by: Kostas Papadopoulos <kpapad75@travelguide.gr>
Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Backport of r42319

SVN-Revision: 42389

10 years agodnsmasq: Create rDNS records for LuCI "Hostnames"
John Crispin [Mon, 1 Sep 2014 13:22:05 +0000 (13:22 +0000)]
dnsmasq: Create rDNS records for LuCI "Hostnames"

LuCI creates "domain" UCI config sections, which the dnsmasq init file
then, currently, translates into "address" config lines. This is not
the correct usage of "address" (see r36943), and also causes rDNS
records to not be created. This patches dnsmasq.init to utilize the
additional hosts file introduced in r40799 for such domain names,
resolving both issues.

Signed-off-by: Tyler Fenby <tylerf@securecominc.com>
Backport of r42318

SVN-Revision: 42388

10 years agoar71xx: add Support for RB2011UiAS-2HnD
John Crispin [Mon, 1 Sep 2014 13:22:04 +0000 (13:22 +0000)]
ar71xx: add Support for RB2011UiAS-2HnD

Backport of r42317

SVN-Revision: 42387