openwrt/staging/blocktrron.git
2 years agokernel: bump 5.10 to 5.10.149
John Audia [Mon, 17 Oct 2022 18:34:40 +0000 (14:34 -0400)]
kernel: bump 5.10 to 5.10.149

No patches required modification.

Signed-off-by: John Audia <therealgraysky@proton.me>
(cherry picked from commit 25fba4c375f081d175bf94079d5d713b6bc4bbc0)

2 years agokernel: bump 5.10 to 5.10.148
John Audia [Sat, 15 Oct 2022 10:31:21 +0000 (06:31 -0400)]
kernel: bump 5.10 to 5.10.148

No patches required modification.

Signed-off-by: John Audia <therealgraysky@proton.me>
(cherry picked from commit 59b5d59edf4479197099e3bf6b034f085067b8e5)

2 years agokernel: bump 5.10 to 5.10.147
John Audia [Wed, 5 Oct 2022 11:58:32 +0000 (07:58 -0400)]
kernel: bump 5.10 to 5.10.147

Removed upstreamed:
  bcm53xx/patches-5.10/083-v6.0-clk-iproc-Do-not-rely-on-node-name-for-correct-PLL-s.patch[1]

All other patches automatically rebased.

1. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.10.147&id=a8e6cde5062fb2aff81f86cc0770591714bee545

Signed-off-by: John Audia <therealgraysky@proton.me>
(cherry picked from commit e2da6a0a59a81a4fc0fdffde31abf22ee121e9f5)

2 years agomt7621: hiwifi_hc5962: fix reboot loop by using LZMA loader
Petr Štetiar [Tue, 18 Oct 2022 06:55:46 +0000 (08:55 +0200)]
mt7621: hiwifi_hc5962: fix reboot loop by using LZMA loader

This fixes a well known "LZMA ERROR 1" error, reported previously on
numerous of similar devices.

References: https://github.com/openwrt/openwrt/issues/10645#issuecomment-1282607274
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit b63d6d4730fd0dc30ce6707338c398e8b9d61d86)

2 years agoucode: update to latest Git HEAD
Jo-Philipp Wich [Tue, 18 Oct 2022 09:06:31 +0000 (11:06 +0200)]
ucode: update to latest Git HEAD

00af065 fs: expose `getdelim()` functionality through `fd.read()`
21ace5e lexer: fixes for regex literal parsing

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit 1b90c7441b81aee7b1212e8918e3ec7144375d96)

2 years agofirewall4: update to latest Git HEAD
Jo-Philipp Wich [Tue, 18 Oct 2022 07:44:42 +0000 (09:44 +0200)]
firewall4: update to latest Git HEAD

7ae5e14 fw4: gracefully handle `null` return values from `fd.read("line")`

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit 5e2e048c0e7c16d7967ec7a0cd8a9c01aa0f12b1)

2 years agomt7621: netgear_ex6150: fix reboot loop by using LZMA loader
Petr Štetiar [Tue, 18 Oct 2022 06:55:46 +0000 (08:55 +0200)]
mt7621: netgear_ex6150: fix reboot loop by using LZMA loader

This fixes a well known "LZMA ERROR 1" error, reported previously on
numerous of similar devices.

Fixes: #10645
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit 7dd1cab1c16f374716b1ee7a5bf99f849b74c1dc)

2 years agoOpenWrt v22.03.2: revert to branch defaults
Hauke Mehrtens [Sat, 15 Oct 2022 12:46:15 +0000 (14:46 +0200)]
OpenWrt v22.03.2: revert to branch defaults

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 years agoOpenWrt v22.03.2: adjust config defaults
Hauke Mehrtens [Sat, 15 Oct 2022 12:45:57 +0000 (14:45 +0200)]
OpenWrt v22.03.2: adjust config defaults

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 years agofirewall4: update to latest Git HEAD
Jo-Philipp Wich [Fri, 14 Oct 2022 15:32:28 +0000 (17:32 +0200)]
firewall4: update to latest Git HEAD

4fbf6d7 ruleset.uc: log forwarded traffic not matched by zone policies
c7201a3 main.uc: reintroduce set reload restriction
756f1e2 ruleset: fix emitting set_mark/set_xmark rules with masks
3db4741 ruleset: properly handle zone names starting with a digit
43d8ef5 fw4: fix formatting of default log prefix
592ba45 main.uc: remove uneeded/wrong set reload restrictions
b0a6bff tests: fix testcases
145e159 fw4: recognize `option log` and `option counter` in `config nat` sections
ce050a8 fw4: fall back to device if l3_device is not available in ifstatus

Fixes: #10639, #10965
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(cherry picked from commit fdfa9d8f7469626d2dc8e4b46a6ad56a3b27c16b)

2 years agoucode: update to latest Git HEAD
Jo-Philipp Wich [Mon, 5 Sep 2022 08:50:39 +0000 (10:50 +0200)]
ucode: update to latest Git HEAD

4ae7072 fs: use `getline()` for line wise read operations
21ace5e lexer: fixes for regex literal parsing
00965fa lib: implement slice() function
76d396d main: implement print mode
7bbba78 compiler: optimize function return opcode generation
a45f2a3 lexer: improve regex literal handling
d64d5d6 vm: maintain export symbol tables per program
f4b4ded uloop: task: gracefully handle absent output callback
a58fe47 ubus: hold reference to underlying connection until deferred is concluded
e23b58a lib: uc_system(): retry waitpid() on EINTR
cc4eb79 ubus: support obtaining numeric error code
01c412c ubus: add toplevel constants for ubus status codes
8e240fa ubus: allow object method call handlers to return a numeric status code
5cdddd3 lib: add limit support to split() and replace()
0ba9c3e fs: add optional third permission argument to fs.open()
c1f7b3b lib: remove fixed capture group limit in match() and regex replace()

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(backported from commits 639754e36d849553e288f8e34f51f793761c07db
 and 5110dcb1fa44fc1aac737c63b31474daa471de89)

2 years agorpcd: update to latest Git HEAD
Jo-Philipp Wich [Mon, 12 Sep 2022 11:48:40 +0000 (13:48 +0200)]
rpcd: update to latest Git HEAD

8c852b6 ucode: write ucode runtime exceptions to stderr
e80d0b2 ucode: pass-through `ubus_rpc_session` argument
0d02243 ucode: initialize module search path early

Signed-off-by: Jo-Philipp Wich <jo@mein.io>
(backported from commits 94129cbefb6027cdfe2b7801a6e27a36d4ec58b8
 and db17c7527107c1dae190608a1313a3977fe4f23f)

2 years agoimage: always rebuild kernel loaders
Matthias Schiffer [Sun, 27 Mar 2022 21:15:48 +0000 (23:15 +0200)]
image: always rebuild kernel loaders

Kernel loaders like the lzma-loader currently don't track changes to
their sources. This can lead to an old version of a loader to be used
when a build tree is not clean between builds.

As the loaders are tiny and the build times are insignificant, simply
force rebuilding them on every build to avoid this problem.

Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
(cherry picked from commit a01d23e755ba46f41e667d558d82d4871d7f5450)

2 years agompc85xx: p1010: make TP-Link WDR4900 v1 build again
Matthias Schiffer [Sun, 27 Mar 2022 19:00:30 +0000 (21:00 +0200)]
mpc85xx: p1010: make TP-Link WDR4900 v1 build again

Add the spi-loader as a pre-kernel stage, so we can lift the kernel size
limit.

Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
(cherry picked from commit 2fa53c9214b7b93fd82ad1ff885145b9e1c1f71b)

2 years agompc85xx: add SPI kernel loader for TP-Link TL-WDR4900 v1
Matthias Schiffer [Sun, 27 Mar 2022 18:57:01 +0000 (20:57 +0200)]
mpc85xx: add SPI kernel loader for TP-Link TL-WDR4900 v1

Similar to the lzma-loader on our MIPS targets, the spi-loader acts as
a second-stage loader that will then load and start the actual kernel.
As the TL-WDR4900 uses SPI-NOR and the P1010 family does not have support
for memory mapping of this type of flash, this loader needs to contain a
basic driver for the FSL ESPI controller.

Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
(cherry picked from commit a296055b82fbb20457273492069ce9d62009e2a1)

2 years agoramips: fix ZyXEL NWA55AXE model name
David Bauer [Thu, 18 Aug 2022 16:02:48 +0000 (18:02 +0200)]
ramips: fix ZyXEL NWA55AXE model name

The model name was missing a letter.

Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit 9c8605dee238cdf52e88b6a1aa64d5b7bf5dd846)

2 years agoramips: add support for ZyXEL NWA50AX / NWA55AXE
David Bauer [Fri, 27 May 2022 15:48:06 +0000 (17:48 +0200)]
ramips: add support for ZyXEL NWA50AX / NWA55AXE

Hardware
--------
CPU:    Mediatek MT7621
RAM:    256M DDR3
FLASH:  128M NAND
ETH:    1x Gigabit Ethernet
WiFi:   Mediatek MT7915 (2.4/5GHz 802.11ax 2x2 DBDC)
BTN:    1x Reset (NWA50AX only)
LED:    1x Multi-Color (NWA50AX only)

UART Console
------------
NWA50AX:
Available below the rubber cover next to the ethernet port.
NWA55AXE:
Available on the board when disassembling the device.

Settings: 115200 8N1

Layout:

<12V> <LAN> GND-RX-TX-VCC

Logic-Level is 3V3. Don't connect VCC to your UART adapter!

Installation Web-UI
-------------------
Upload the Factory image using the devices Web-Interface.

As the device uses a dual-image partition layout, OpenWrt can only
installed on Slot A. This requires the current active image prior
flashing the device to be on Slot B.

If the currently installed image is started from Slot A, the device will
flash OpenWrt to Slot B. OpenWrt will panic upon first boot in this case
and the device will return to the ZyXEL firmware upon next boot.

If this happens, first install a ZyXEL firmware upgrade of any version
and install OpenWrt after that.

Installation TFTP
-----------------
This installation routine is especially useful in case
 * unknown device password (NWA55AXE lacks reset button)
 * bricked device

Attach to the UART console header of the device. Interrupt the boot
procedure by pressing Enter.

The bootloader has a reduced command-set available from CLI, but more
commands can be executed by abusing the atns command.

Boot a OpenWrt initramfs image available on a TFTP server at
192.168.1.66. Rename the image to owrt.bin

 $ atnf owrt.bin
 $ atna 192.168.1.88
 $ atns "192.168.1.66; tftpboot; bootm"

Upon booting, set the booted image to the correct slot:

 $ zyxel-bootconfig /dev/mtd10 get-status
 $ zyxel-bootconfig /dev/mtd10 set-image-status 0 valid
 $ zyxel-bootconfig /dev/mtd10 set-active-image 0

Copy the OpenWrt ramboot-factory image to the device using scp.
Write the factory image to NAND and reboot the device.

 $ mtd write ramboot-factory.bin firmware
 $ reboot

Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit a0b7fef0ffe4cd9cca39a652a37e4f3ce8f0a681)

2 years agobusybox: nslookup: ensure unique transaction IDs for the DNS queries
Uwe Kleine-König [Fri, 14 Oct 2022 10:23:07 +0000 (12:23 +0200)]
busybox: nslookup: ensure unique transaction IDs for the DNS queries

On machines with a coarse monotonic clock (here: TP-Link RE200 powered
by a MediaTek MT7620A) it can happen that the two DNS requests (for A
and AAAA) share the same transaction ID. If this happens the second
reply is wrongly dropped and nslookup reports "No answer".

Fix this by ensuring that the transaction IDs are unique.

Signed-off-by: Uwe Kleine-König <uwe@kleine-koenig.org>
(cherry picked from commit 63e5ba8e69f03a584b707520db0a0821eda3024f)
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
2 years agomac80211: backport security fixes
Felix Fietkau [Thu, 13 Oct 2022 12:29:53 +0000 (14:29 +0200)]
mac80211: backport security fixes

This mainly affects scanning and beacon parsing, especially with MBSSID enabled

Fixes: CVE-2022-41674
Fixes: CVE-2022-42719
Fixes: CVE-2022-42720
Fixes: CVE-2022-42721
Fixes: CVE-2022-42722
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 26f400210d6b3780fcc0deb89b9741837df9c8b8)

2 years agomac80211: merge upstream fixes
Koen Vandeputte [Mon, 22 Aug 2022 09:18:35 +0000 (11:18 +0200)]
mac80211: merge upstream fixes

fetched from upstream kernel v5.15.67

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
(cherry-picked from commit aa9be386d40f3a5e559c0f2183c772175a45cf0d)

2 years agoramips: skip bbt scan on mt7621
Felix Fietkau [Thu, 10 Mar 2022 14:20:29 +0000 (15:20 +0100)]
ramips: skip bbt scan on mt7621

reduces unnecessary flash reads and speeds up boot time

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 55e8d52157f191bf578cf716983764e64c6f94e4)

2 years agoramips: enable support for mtk_bmt in the nand flash driver
Felix Fietkau [Thu, 10 Mar 2022 10:45:00 +0000 (11:45 +0100)]
ramips: enable support for mtk_bmt in the nand flash driver

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 4947623d6c801365a60f383217c187e3d9dae953)

2 years agoramips: mt7621_nand: initialize ECC_FDMADDR
Felix Fietkau [Wed, 9 Mar 2022 19:46:21 +0000 (20:46 +0100)]
ramips: mt7621_nand: initialize ECC_FDMADDR

This is needed for the ECC controller to access FDM data

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 73b2a4ca033bbd84d3e0373d4fd21c559ddc090b)

2 years agoramips: mt7621_nand: reduce log verbosity
Stijn Tintel [Wed, 19 Jan 2022 15:59:51 +0000 (17:59 +0200)]
ramips: mt7621_nand: reduce log verbosity

Avoid flooding the log with the message below by increasing the log
level to debug:

  mt7621-nand 1e003000.nand: Using programmed access timing: 31c07388

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
(cherry-picked from commit 89c195925109d2b59f284bfdd23a3d853c67e27b)

2 years agoramips: move mt7621_nand driver to files
Stijn Tintel [Wed, 19 Jan 2022 13:44:58 +0000 (15:44 +0200)]
ramips: move mt7621_nand driver to files

The patch was rejected by upstream. The mtk_nand driver should be
modified to support the mt7621 flash controller instead. As there is no
newer version to backport, or no upstream version to fix bugs, let's
move the driver to the files dir under the ramips target. This makes it
easier to make changes to the driver while waiting for mt7621 support to
land in mtk_nand.

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
(cherry-picked from commit 2f2e81a4ea110328c5434054d1412b4d1d8fde81)

2 years agokernel: mtdsplit: support UBI after FIT images
Felix Fietkau [Thu, 10 Mar 2022 16:32:20 +0000 (17:32 +0100)]
kernel: mtdsplit: support UBI after FIT images

Change the partition name accordingly. Same behavior as mtdsplit_uimage

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 62fd9f97090d05637a283d594f2d02958fd36a80)

2 years agokernel: mtk_bmt: skip bitflip check if threshold isn't set
Chuanhong Guo [Thu, 7 Apr 2022 01:56:02 +0000 (09:56 +0800)]
kernel: mtk_bmt: skip bitflip check if threshold isn't set

kernel spi-nand driver leaves this field empty and let mtd set it later.

Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
(cherry-picked from commit 6fa50e26e7c1221085d8e71cde7e9148c540c752)

2 years agokernel: mtk_bmt: add debugfs file to attempt repair of remapped sectors
Felix Fietkau [Thu, 10 Mar 2022 15:23:25 +0000 (16:23 +0100)]
kernel: mtk_bmt: add debugfs file to attempt repair of remapped sectors

This can be used for sectors that are not physically damaged

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 2a8a333ee96fb0c8d9875abf9fcd5c412f4b19ae)

2 years agokernel: add support for mediatek NMBM flash mapping support
Felix Fietkau [Thu, 10 Mar 2022 14:31:25 +0000 (15:31 +0100)]
kernel: add support for mediatek NMBM flash mapping support

This NAND flash remapping method is used on newer MediaTek devices with NAND
flash.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 06382d1af7b2c5afcde605db436522326fbf5467)

2 years agokernel: mtk_bmt: on error, do not attempt to remap out-of-range blocks
Felix Fietkau [Wed, 9 Mar 2022 15:14:12 +0000 (16:14 +0100)]
kernel: mtk_bmt: on error, do not attempt to remap out-of-range blocks

Pass errors to caller instead

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit be1f2b4d9d01594c2fd20f37b8e63e7cb7a81572)

2 years agokernel: mtk_bmt: fix block copying on remap with bmt v2
Felix Fietkau [Wed, 9 Mar 2022 15:04:39 +0000 (16:04 +0100)]
kernel: mtk_bmt: fix block copying on remap with bmt v2

Copy from the previously mapped block (in case it was remapped already)

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 7d1e2be1605eda85e1a607a3d3ddefed1b961d3d)

2 years agokernel: mtk_bmt: allow get_mapping_block to return an error
Felix Fietkau [Wed, 9 Mar 2022 14:54:52 +0000 (15:54 +0100)]
kernel: mtk_bmt: allow get_mapping_block to return an error

Used by the mapping implementation to indicate that no backing block is
available

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit b4c7f8c5f7a2dab76fe90849e112c6bd8f80ab19)

2 years agokernel: split up mtk_bmt driver code
Felix Fietkau [Wed, 9 Mar 2022 10:30:58 +0000 (11:30 +0100)]
kernel: split up mtk_bmt driver code

Keep a separate source file per variant

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 601c7b4adb0f91005a8755bf9575c68ac623cc33)

2 years agoOpenWrt v22.03.1: revert to branch defaults
Hauke Mehrtens [Sun, 9 Oct 2022 17:32:23 +0000 (19:32 +0200)]
OpenWrt v22.03.1: revert to branch defaults

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 years agoOpenWrt v22.03.1: adjust config defaults
Hauke Mehrtens [Sun, 9 Oct 2022 17:32:19 +0000 (19:32 +0200)]
OpenWrt v22.03.1: adjust config defaults

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 years agoath79: fix model name of Extreme Networks WS-AP3805i
Tom Herbers [Fri, 7 Oct 2022 22:08:52 +0000 (00:08 +0200)]
ath79: fix model name of Extreme Networks WS-AP3805i

Everywhere else the device is referred to as WS-AP3805i,
only the model name wrongly only said AP3805i.

Signed-off-by: Tom Herbers <mail@tomherbers.de>
(cherry picked from commit 7d6032f310058d7e9b96d7e1dc4d49c8232beff7)

2 years agoath79: add low_mem to tiny image
Nick Hainke [Fri, 19 Aug 2022 16:59:36 +0000 (18:59 +0200)]
ath79: add low_mem to tiny image

Devices with SMALL_FLASH enabled have "SQUASHFS_BLOCK_SIZE=1024" in
their config. This significantly increases the cache memory required by
squashfs [0]. This commit enables low_mem leading to a much better
performance because the SQUASHFS_BLOCK_SIZE is reduced to 256.

Example Nanostation M5 (XM):
The image size increases by 128 KiB. However, the memory statisitcs look
much better:

Default tiny build:
------
MemTotal:          26020 kB
MemFree:            5648 kB
MemAvailable:       6112 kB
Buffers:               0 kB
Cached:             3044 kB

low_mem enabled:
-----
MemTotal:          26976 kB
MemFree:            6748 kB
MemAvailable:      11504 kB
Buffers:               0 kB
Cached:             7204 kB

[0] - https://github.com/freifunk-gluon/gluon/commit/7e8af99cf504ca1dc389f282a0c94f4a911571be

Signed-off-by: Nick Hainke <vincent@systemli.org>
(cherry picked from commit f54ac98f8cec676761e5144ae06640b8007b4b04)

2 years agoath79: move ubnt-xm to tiny
Nick Hainke [Sat, 25 Dec 2021 08:19:52 +0000 (09:19 +0100)]
ath79: move ubnt-xm to tiny

ath79 has was bumped to 5.10. With this, as with every kernel change,
the kernel has become larger. However, although the kernel gets bigger,
there are still enough flash resources. But the RAM reaches its capacity
limits. The tiny image comes with fewer kernel flags enabled and
fewer daemons.

Improves: 15aa53d7ee65 ("ath79: switch to Kernel 5.10")

Tested-by: Robert Foss <me@robertfoss.se>
Signed-off-by: Nick Hainke <vincent@systemli.org>
(cherry picked from commit f4415f7635164ec07ddc22f56df93555804b5767)

2 years agokernel: fix possible mtd NULL pointer dereference
Rafał Miłecki [Tue, 4 Oct 2022 10:04:37 +0000 (12:04 +0200)]
kernel: fix possible mtd NULL pointer dereference

Fixes: cae4d089bc1d3 ("kernel: backport mtd dynamic partition patch")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
(cherry picked from commit a5265497a4f6da158e95d6a450cb2cb6dc085cab)

2 years agotreewide: fix security issues by bumping all packages using libwolfssl
Petr Štetiar [Thu, 29 Sep 2022 16:45:40 +0000 (18:45 +0200)]
treewide: fix security issues by bumping all packages using libwolfssl

As wolfSSL is having hard time maintaining ABI compatibility between
releases, we need to manually force rebuild of packages depending on
libwolfssl and thus force their upgrade. Otherwise due to the ABI
handling we would endup with possibly two libwolfssl libraries in the
system, including the patched libwolfssl-5.5.1, but still have
vulnerable services running using the vulnerable libwolfssl-5.4.0.

So in order to propagate update of libwolfssl to latest stable release
done in commit ec8fb542ec3e4 ("wolfssl: fix TLSv1.3 RCE in uhttpd by
using 5.5.1-stable (CVE-2022-39173)") which fixes several remotely
exploitable vulnerabilities, we need to bump PKG_RELEASE of all
packages using wolfSSL library.

Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit f1b7e1434f66a3cb09cb9e70b40add354a22e458)

2 years agowolfssl: fix TLSv1.3 RCE in uhttpd by using 5.5.1-stable (CVE-2022-39173)
Petr Štetiar [Wed, 28 Sep 2022 09:28:06 +0000 (11:28 +0200)]
wolfssl: fix TLSv1.3 RCE in uhttpd by using 5.5.1-stable (CVE-2022-39173)

Fixes denial of service attack and buffer overflow against TLS 1.3
servers using session ticket resumption. When built with
--enable-session-ticket and making use of TLS 1.3 server code in
wolfSSL, there is the possibility of a malicious client to craft a
malformed second ClientHello packet that causes the server to crash.

This issue is limited to when using both --enable-session-ticket and TLS
1.3 on the server side. Users with TLS 1.3 servers, and having
--enable-session-ticket, should update to the latest version of wolfSSL.

Thanks to Max at Trail of Bits for the report and "LORIA, INRIA, France"
for research on tlspuffin.

Complete release notes https://github.com/wolfSSL/wolfssl/releases/tag/v5.5.1-stable

Fixes: CVE-2022-39173
Fixes: https://github.com/openwrt/luci/issues/5962
References: https://github.com/wolfSSL/wolfssl/issues/5629
Tested-by: Kien Truong <duckientruong@gmail.com>
Reported-by: Kien Truong <duckientruong@gmail.com>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit ec8fb542ec3e4f584444a97de5ac05dbc2a9cde5)

2 years agowolfssl: refresh patches
Petr Štetiar [Wed, 28 Sep 2022 09:58:22 +0000 (11:58 +0200)]
wolfssl: refresh patches

So they're tidy and apply cleanly.

Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit 8ad9a72cbed07643c7a8e4febbea71c7122b29a4)

2 years agowolfssl: bump to 5.5.0
Ivan Pavlov [Wed, 31 Aug 2022 05:04:42 +0000 (08:04 +0300)]
wolfssl: bump to 5.5.0

Remove upstreamed: 101-update-sp_rand_prime-s-preprocessor-gating-to-match.patch

Some low severity vulnerabilities fixed
OpenVPN compatibility fixed (broken in 5.4.0)
Other fixes && improvements

Signed-off-by: Ivan Pavlov <AuthorReflex@gmail.com>
(cherry picked from commit 3d88f26d74f7771b808082cef541ed8286c40491)

2 years agokernel: bump 5.10 to 5.10.146
John Audia [Wed, 28 Sep 2022 20:05:00 +0000 (16:05 -0400)]
kernel: bump 5.10 to 5.10.146

All patches automatically rebased.

Signed-off-by: John Audia <therealgraysky@proton.me>
(cherry picked from commit eed0a31b90e4feeb65f3b54853bd4db1f5bcd524)

2 years agokernel: bump 5.10 to 5.10.145
John Audia [Fri, 23 Sep 2022 20:58:05 +0000 (16:58 -0400)]
kernel: bump 5.10 to 5.10.145

Manually rebased:
  hack-5.10/780-usb-net-MeigLink_modem_support.patch

Removed upstreamed:
  patches-5.10/110-gpio-mpc8xxx-Fix-support-for-IRQ_TYPE_LEVEL_LOW-flow.patch[1]

All other patches automatically rebased.

1. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.10.145&id=24196210b198e8e39296e277bb93b362aa207775

Signed-off-by: John Audia <therealgraysky@proton.me>
(cherry picked from commit 8fe67fae1d619467b2b3300178d4309f17cfdb54)

2 years agokernel: bump 5.10 to 5.10.144
John Audia [Wed, 21 Sep 2022 11:52:34 +0000 (07:52 -0400)]
kernel: bump 5.10 to 5.10.144

All patches automatically rebased.

Signed-off-by: John Audia <therealgraysky@proton.me>
(cherry picked from commit eff4f8b2f0a2900c945c21c8183d1faa0ac35ec1)

2 years agoramips: fix switch setup for ASUS RT-AX53U
Matthias Schiffer [Mon, 3 Oct 2022 11:23:41 +0000 (13:23 +0200)]
ramips: fix switch setup for ASUS RT-AX53U

The device has only 1 WAN + 3 LAN ports. Remove "lan4" interface
corresponding to the non-existing port.

Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
(cherry picked from commit 149fc3a269b435483b31df03d6fd9679286cf9e7)

2 years agouboot-mvebu: backport LibreSSL patches for older version of LibreSSL
Josef Schlehofer [Wed, 14 Sep 2022 13:23:45 +0000 (15:23 +0200)]
uboot-mvebu: backport LibreSSL patches for older version of LibreSSL

If you would like to compile the newest version of U-boot together with the stable
OpenWrt version, which does not have LibreSSL >= 3.5, which was updated
in the master branch by commit 5451b03b7ceb2315445c683fe174e28bbdd49c2f
("tools/libressl: bump to v3.5.3"), then you need these two patches to
fix it. They are backported from U-boot repository.

This should be backported to stable OpenWrt versions.

Reported-by: Michal Vasilek <michal.vasilek@nic.cz>
Signed-off-by: Josef Schlehofer <pepe.schlehofer@gmail.com>
(cherry picked from commit 185541f50ff59c0a5e0663ad612f0f5eb31926cf)

2 years agouboot-mvebu: backport patch to fix compilation on non glibc system
Josef Schlehofer [Wed, 14 Sep 2022 13:19:56 +0000 (15:19 +0200)]
uboot-mvebu: backport patch to fix compilation on non glibc system

This issue was reported by @paper42, who is using Void Linux with musl
to compile OpenWrt and its packages and found out it is not possible to
compile U-boot for Turris Omnia (neither any other).

It fixes following output:
```
  HOSTCC  tools/kwboot
tools/kwboot.c: In function 'kwboot_tty_change_baudrate':
tools/kwboot.c:662:6: error: 'struct termios' has no member named 'c_ospeed'
  662 |   tio.c_ospeed = tio.c_ispeed = baudrate;
      |      ^
tools/kwboot.c:662:21: error: 'struct termios' has no member named 'c_ispeed'
  662 |   tio.c_ospeed = tio.c_ispeed = baudrate;
      |                     ^
tools/kwboot.c:690:31: error: 'struct termios' has no member named 'c_ospeed'
  690 |  if (!_is_within_tolerance(tio.c_ospeed, baudrate, 3))
      |                               ^
tools/kwboot.c:693:31: error: 'struct termios' has no member named 'c_ispeed'
  693 |  if (!_is_within_tolerance(tio.c_ispeed, baudrate, 3))
      |
```

Tested-by: Michal Vasilek <michal.vasilek@nic.cz>
Signed-off-by: Josef Schlehofer <pepe.schlehofer@gmail.com>
(cherry picked from commit 9c7472950b01c5b3a461f4e29b3b62bac9e35b46)

2 years agoramips: enable LZMA loader to fix Linksys RE6500 boot
Mark King [Sun, 25 Sep 2022 20:53:52 +0000 (21:53 +0100)]
ramips: enable LZMA loader to fix Linksys RE6500 boot

At some point after 21.02.3 and before 22.03.0, the size limits of the
Linksys RE6500 were reached and prevent booting from the 22.03.0 release
or builds of current SNAPSHOT. This patch allows builds of master to boot
again and has been tested on my device.

Fixes: #8577
Signed-off-by: Mark King <mark@vemek.co>
(cherry picked from commit bf5b1a53d4ff3f1d742f9ece1ab16555280417ec)

2 years agotools/meson: backport WSL2 fix
Rosen Penev [Thu, 22 Sep 2022 22:41:28 +0000 (15:41 -0700)]
tools/meson: backport WSL2 fix

For some reason, Microsoft's Plan9 driver returns IOError on missing
file.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
(cherry picked from commit 875e17774bafb132a93d66f1d7b2c6a2deec2030)

2 years agofirmware: intel-microcode: update to 20220809
Christian Lamparter [Mon, 22 Aug 2022 08:35:28 +0000 (10:35 +0200)]
firmware: intel-microcode: update to 20220809

Debian's changelog by Henrique de Moraes Holschuh <hmh@debian.org>:

  * New upstream microcode datafile 20220809
    * Fixes INTEL-SA-00657, CVE-2022-21233
      Stale data from APIC leaks SGX memory (AEPIC leak)
    * Fixes unspecified errata (functional issues) on Xeon Scalable
    * Updated Microcodes:
      sig 0x00050653, pf_mask 0x97, 2022-03-14, rev 0x100015e, size 34816
      sig 0x00050654, pf_mask 0xb7, 2022-03-08, rev 0x2006e05, size 44032
      sig 0x000606a6, pf_mask 0x87, 2022-04-07, rev 0xd000375, size 293888
      sig 0x000706a1, pf_mask 0x01, 2022-03-23, rev 0x003c, size 75776
      sig 0x000706a8, pf_mask 0x01, 2022-03-23, rev 0x0020, size 75776
      sig 0x000706e5, pf_mask 0x80, 2022-03-17, rev 0x00b2, size 112640
      sig 0x000806c2, pf_mask 0xc2, 2022-03-19, rev 0x0028, size 97280
      sig 0x000806d1, pf_mask 0xc2, 2022-03-28, rev 0x0040, size 102400
      sig 0x00090672, pf_mask 0x03, 2022-06-07, rev 0x0022, size 216064
      sig 0x00090675, pf_mask 0x03, 2022-06-07, rev 0x0022, size 216064
      sig 0x000906a3, pf_mask 0x80, 2022-06-15, rev 0x0421, size 216064
      sig 0x000906a4, pf_mask 0x80, 2022-06-15, rev 0x0421, size 216064
      sig 0x000a0671, pf_mask 0x02, 2022-03-17, rev 0x0054, size 103424
      sig 0x000b06f2, pf_mask 0x03, 2022-06-07, rev 0x0022, size 216064
      sig 0x000b06f5, pf_mask 0x03, 2022-06-07, rev 0x0022, size 216064

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
(cherry picked from commit bb73828b89def128f26ae1cdff0d08569d261f1b)

2 years agotoolchain: Include ./include/fortify for external musl toolchain
Hauke Mehrtens [Sun, 11 Sep 2022 22:31:43 +0000 (00:31 +0200)]
toolchain: Include ./include/fortify for external musl toolchain

When building with an external toolcahin with musl also include
./include/fortify by default. This is also done when we build with the
internal toolchain using musl libc.

Without this extra include the fortify source feature is not working
when using an external musl toolchain. All binaries were compiled
without fortify source when an external musl toolchain was used. All
binaries release done by the OpenWrt project use the internal toolcahin
where fortify source is working.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit b21ddbfa18cd1e437e31dc0fa29408721a186c2a)

2 years agotoolchain: Select USE_SSTRIP with external musl toolchain
Hauke Mehrtens [Sun, 11 Sep 2022 22:29:00 +0000 (00:29 +0200)]
toolchain: Select USE_SSTRIP with external musl toolchain

When we use the internal toolchain USE_SSTRIP will be selected by
default for musl libc and USE_STRIP when glibc is used. Do the same when
an external toolchain is used. USE_GLIBC will also be set for external
toolchain builds based on the EXTERNAL_TOOLCHAIN_LIBC_USE_GLIBC setting.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 9403810c020cca136149973a3929bf77a1f501aa)

2 years agoscripts: ext-toolchain: add support for musl
Christian Marangi [Sun, 17 Jul 2022 15:56:59 +0000 (17:56 +0200)]
scripts: ext-toolchain: add support for musl

Openwrt now supports only glibc and musl. Add support for musl and
rework the libc check to handle the new config flags and correctly
compile package basend on that.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit 7be01fe13b4517e5edb8a4818f437d60144cdcb4)

2 years agoscripts: ext-toolchain: add support for info.mk in probe_cc
Christian Marangi [Sun, 17 Jul 2022 15:56:36 +0000 (17:56 +0200)]
scripts: ext-toolchain: add support for info.mk in probe_cc

Openwrt generate info.mk that contains the libc type. For probe_cc check
if the file exist and parse directly it for LIBC type.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit 75311977f5ff64b491cb57ac713d75e0e410d786)

2 years agoscripts: ext-toolchain: actually probe libc type on config generation
Christian Marangi [Sun, 17 Jul 2022 15:53:58 +0000 (17:53 +0200)]
scripts: ext-toolchain: actually probe libc type on config generation

Currently we never call probe_cc before config generation, this cause
the script to never actually detect the correct libc type.
Call probe_cc before config generation to correctl set the .config file.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit ddeabc75ebe3151ff7da302cb1aae702b3ad7eba)

2 years agoscripts: ext-toolchain: add option to overwrite config
Christian Marangi [Mon, 4 Jul 2022 16:22:18 +0000 (18:22 +0200)]
scripts: ext-toolchain: add option to overwrite config

It can be useful to overwrite an already generated config.
Option are simply added at the end of the config and make defconfig
will overwrite the relevant option with the new one.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit f4dd18ca39c42a324e34633c8ee553717531bc3b)

2 years agoscripts: ext-toolchain: fix wrong prefix in print_config generation
Christian Marangi [Sun, 3 Jul 2022 00:20:11 +0000 (02:20 +0200)]
scripts: ext-toolchain: fix wrong prefix in print_config generation

The parsed prefix in print_config is wrong and this produce broken
generated .config that won't work with any external toolchain.

Currently the prefix from a CC of

'arm-openwrt-linux-muslgnueabi-gcc-12.1.0'

produce a prefix

'arm-openwrt-linux-muslgnueabi-gcc-'

This is wrong as the real prefix should be

'arm-openwrt-linux-muslgnueabi-'

This is probably caused by a change in how the toolchain is now handled
that now append also the gcc version. Probably in ancient days the
version wasn't part of the name and the prefix generation stripped the
'-gcc' instead of the gcc version.

Fix this and correctly strip the gcc version and the gcc suffix to
correctly call toolchain bins.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit 53c293262fce844c8291ab82e6726a8489d3c57b)

2 years agorules_mk: don't include wrapped bin with external toolchains
Christian Marangi [Sun, 3 Jul 2022 00:06:21 +0000 (02:06 +0200)]
rules_mk: don't include wrapped bin with external toolchains

Don't add wrapped bin to the TARGET_PATH as it does cause compilation
error.

cmake.mk will use the "command -v" and will use the wrapped bin instead
of the external toolchain bin as they have the same name and command
will select the first result.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit a90eabf60255773231ed0259e5da5eb6a36fe9ce)

2 years agorules_mk: use gcc versions for external toolchain
Vincent Wiemann [Mon, 28 Dec 2020 15:00:13 +0000 (16:00 +0100)]
rules_mk: use gcc versions for external toolchain

When using the OpenWrt toolchain as an external toolchain the build
failed due to missing LTO support. By choosing the GCC wrappers of
the tools this commit makes sure that the LTO-enabled executables
are being used.

Signed-off-by: Vincent Wiemann <vincent.wiemann@ironai.com>
[ wrap the commit description to 72 char ]
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit 2555ffb4536e2727465e9a0d426ad3c4f1ef003a)

2 years agobcm53xx: backport clk driver fix for DT nodes names
Rafał Miłecki [Thu, 29 Sep 2022 03:51:25 +0000 (05:51 +0200)]
bcm53xx: backport clk driver fix for DT nodes names

It allows dropping downstream patch renaming DT nodes.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
(cherry picked from commit 77d9cce604d32005ddb90e91c6cc9b9cf35068d7)

2 years agoath79: fix LibreRouter-v1 watchdog and poe_pass
Santiago Piccinini [Fri, 30 Sep 2022 13:44:04 +0000 (10:44 -0300)]
ath79: fix LibreRouter-v1 watchdog and poe_pass

Watchdog and poe_passthrough gpios require the jtag disabled.

Signed-off-by: Santiago Piccinini <spiccinini@altermundi.net>
(cherry picked from commit 2ad949b11dbaa4c634868d55a4452d5a558776bd)

2 years agobuild: fix warnings from grep
Chris Osgood [Thu, 15 Sep 2022 12:09:38 +0000 (08:09 -0400)]
build: fix warnings from grep

Fixes build warnings when using newer versions of grep.

Signed-off-by: Chris Osgood <chris_github@functionalfuture.com>
Tested-by: Georgi Valkov <gvalkov@gmail.com>
(cherry picked from commit c5e167e0d6075c46ca92c30b425c4dcb05fec5ed)
[ fix conflict error ]
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
2 years agoMakefile: fix stray \ warnings with grep-3.8
Nick Hainke [Sun, 11 Sep 2022 06:15:52 +0000 (08:15 +0200)]
Makefile: fix stray \ warnings with grep-3.8

We simply grep for "/usr". So no need for "-E" or "\/". Furthermore, in
the new grep versions this creates warnings.

As written in the grep-3.8 announcement:
  Regular expressions with stray backslashes now cause warnings, as
  their unspecified behavior can lead to unexpected results.
  For example, '\a' and 'a' are not always equivalent
  <https://bugs.gnu.org/39678>.

Fixes warnings in the form of:
  grep: warning: stray \ before /

Signed-off-by: Nick Hainke <vincent@systemli.org>
(cherry picked from commit a29d3bc48c40c6a2a93ae1806bea2ac26455cdbb)
[ fix conflict error ]
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
2 years agobuild: fix issues with targets installed via feeds
Felix Fietkau [Tue, 27 Sep 2022 11:39:12 +0000 (13:39 +0200)]
build: fix issues with targets installed via feeds

- fix including modules.mk when a target is being replaced
- fix calling make targets from target/linux

Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 3a8825ad6acbf18b2b472ace56be58868af78be7)

2 years agobuild: fix including modules.mk for targets pulled in from feeds
Felix Fietkau [Sun, 18 Sep 2022 04:27:09 +0000 (06:27 +0200)]
build: fix including modules.mk for targets pulled in from feeds

Fixes: ebc36ebb2349 ("scripts/feeds: install targets to target/linux/feeds and support overriding")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 00094efec33f07c9dc16cce23be492430c40b3cc)

2 years agompc85xx: add patch to fix gpio mpc8xxx
Josef Schlehofer [Tue, 6 Sep 2022 11:17:02 +0000 (13:17 +0200)]
mpc85xx: add patch to fix gpio mpc8xxx

Backports Linux kernel patch [1] for kernel 5.10 where it applies cleanly.
This was tested on CZ.NIC Turris 1.1 router running OpenWrt 22.03
release.

Before:

- In /var/log/messages:
```
[   16.392988] lm90 0-004c: cannot request IRQ 48
[   16.398280] lm90: probe of 0-004c failed with error -22
```

- Sensors does not work:
```
root@turris:~# sensors
No sensors found!
Make sure you loaded all the kernel drivers you need.
Try sensors-detect to find out which these are.
```

After:

```
root@turris:/# sensors
sa56004-i2c-0-4c
Adapter: MPC adapter (i2c@3000)
temp1:        +44.0°C  (low  =  +0.0°C, high = +70.0°C)
                       (crit = +85.0°C, hyst = +75.0°C)
temp2:        +73.8°C  (low  =  +0.0°C, high = +70.0°C)  ALARM (HIGH)
                       (crit = +85.0°C, hyst = +75.0°C)
```

[1] https://lore.kernel.org/linux-gpio/20220906105431.30911-1-pali@kernel.org/

Signed-off-by: Josef Schlehofer <pepe.schlehofer@gmail.com>
(cherry picked from commit 2f496c34b6dfbedbfbff8d2ade5d629f096e40bd)
(c&p commit ID from openwrt's master)
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
2 years agoramips: fix fw_setsys
Wenli Looi [Mon, 1 Aug 2022 03:23:22 +0000 (03:23 +0000)]
ramips: fix fw_setsys

This change was included in the original pull request but later omitted
for some reason:

https://github.com/openwrt/openwrt/pull/4936

Signed-off-by: Wenli Looi <wlooi@ucalgary.ca>
(cherry picked from commit 4cccea02a60aee0dd77c4db35672c92e2fe384a1)

2 years agokernel: add missing symbol
Stijn Tintel [Sat, 2 Apr 2022 20:41:27 +0000 (23:41 +0300)]
kernel: add missing symbol

Enabling KERNEL_KPROBES exposes KERNEL_BPF_KPROBE_OVERRIDE. Add a build
option for it to fix build failures with KERNEL_KPROBES enabled.

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
(cherry picked from commit 500c37c56ff60b46c30bb0ea7c92676bea23331a)

2 years agobcm4908: fix -EPROBE_DEFER support in bcm4908_enet
Rafał Miłecki [Wed, 21 Sep 2022 07:24:47 +0000 (09:24 +0200)]
bcm4908: fix -EPROBE_DEFER support in bcm4908_enet

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
(cherry picked from commit 1722e23ffcf9038114142a2129c29eb3cdec8ff9)

2 years agokernel: update U-Boot NVMEM driver
Rafał Miłecki [Wed, 21 Sep 2022 07:24:45 +0000 (09:24 +0200)]
kernel: update U-Boot NVMEM driver

1. Fix casting
2. Support DT-defined variables

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
(cherry picked from commit 5652f378c6c607f99a15e6472cdca46c9c4b6162)

2 years agobcm4908: backport mtd parser for Broadcom's U-Boot partition
Rafał Miłecki [Wed, 21 Sep 2022 07:24:41 +0000 (09:24 +0200)]
bcm4908: backport mtd parser for Broadcom's U-Boot partition

Broadcom's U-Boot contains environment data blocks. They need to be
found (offsets aren't predefined) to access env variables.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
(cherry picked from commit 137149847d0f374515f38952ce0986b03a97f2e4)

2 years agomediatek: fix Unifi 6LR network config
Stijn Tintel [Tue, 20 Sep 2022 16:27:34 +0000 (19:27 +0300)]
mediatek: fix Unifi 6LR network config

When the v1 and v2 variants of the U6LR were introduced, the board
network config was not adapted to the new device names. Due to this, the
wrong network config is applied during initial boot. The resulting
config has lan, wan and a switch, while this device only has a single
ethernet interface without a switch.

Fix this by using a wildcard that matches all the variants.

Fixes: 15a02471bb85 ("mediatek: new target mt7622-ubnt-unifi-6-lr-v1")
Fixes: 5c8d3893a78f ("mediatek: new target ubnt_unifi-6-lr-v1-ubootmod")
Fixes: 31d86a1a1192 ("mediatek: add Ubiquiti UniFi 6 LR v2 targets")
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
Acked-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 117f41ee95ec445dce5ab9332bc27a6311747c0d)

2 years agompc85xx: fix unset kernel symbols
David Bauer [Tue, 20 Sep 2022 19:08:23 +0000 (21:08 +0200)]
mpc85xx: fix unset kernel symbols

Signed-off-by: David Bauer <mail@david-bauer.net>
2 years agomac80211: rt2x00: experimental improvements for MT7620 wifi
Daniel Golle [Thu, 28 Jul 2022 11:40:12 +0000 (13:40 +0200)]
mac80211: rt2x00: experimental improvements for MT7620 wifi

Serge Vasilugin reports:

To improve mt7620 built-in wifi performance some changes:
1. Correct BW20/BW40 switching (see comments with mark (1))
2. Correct TX_SW_CFG1 MAC reg from v3 of vendor driver see
https://gitlab.com/dm38/padavan-ng/-/blob/master/trunk/proprietary/rt_wifi/rtpci/3.0.X.X/mt76x2/chips/rt6352.c#L531
3. Set bbp66 for all chains.
4. US_CYC_CNT init based on Programming guide, default value was 33 (pci),
   set chipset bus clock with fallback to cpu clock/3.
5. Don't overwrite default values for mt7620.
6. Correct some typos.
7. Add support for external LNA:
    a) RF and BBP regs never be corrected for this mode
    b) eLNA is driven the same way as ePA with mt7620's pin PA
but vendor driver explicitly pin PA to gpio mode (for forrect calibration?)
so I'm not sure that request for pa_pin in dts-file will be enough

First 5 changes (really 2) improve performance for boards w/o eLNA/ePA.
Changes 7 add support for eLNA

Configuration w/o eLAN/ePA and with eLNA show results
tx/rx (from router point of view) for each stream:
 35-40/30-35 Mbps for HT20
 65-70/60-65 Mbps for HT40

Yes. Max results for 2T2R client is 140-145/135-140
with peaks 160/150, It correspond to mediatek driver results.
Boards with ePA untested.

Reported-by: Serge Vasilugin <vasilugin@yandex.ru>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
[directly include v3 of the patchset submitted upstream]
(cherry picked from commit 31a6605de04218e1c04bd5c2436c24d7d1c07506)
(cherry picked from commit e785ca05e9f0502894772f5df92192b816ba5d7c)
(cherry picked from commit 412fcf3d4400f84551f3ead0514834c62d94a251)

2 years agomac80211: rt2x00: fix typo
Sungbo Eo [Mon, 31 Jan 2022 17:53:59 +0000 (02:53 +0900)]
mac80211: rt2x00: fix typo

Add missing semicolon and refresh patches.

Signed-off-by: Sungbo Eo <mans0n@gorani.run>
(cherry picked from commit d826c91704d2baa5e389c225791740e4c61d62c4)

2 years agomac80211: add patch descriptions to rt2x00 patches
Daniel Golle [Mon, 12 Sep 2022 20:21:05 +0000 (21:21 +0100)]
mac80211: add patch descriptions to rt2x00 patches

Prepare patches for sending upstream by adding patch descriptions
generated from the original OpenWrt commits adding each patch.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit d4feb66048f6a8f387eedfb162a1184cdae9d756)

2 years agomediatek: fix ledbar of UniFi 6 LR when running custom U-Boot
Daniel Golle [Sun, 11 Sep 2022 19:24:05 +0000 (20:24 +0100)]
mediatek: fix ledbar of UniFi 6 LR when running custom U-Boot

The RGB LED of the UniFi 6 LR v1 doesn't work when using the Openwrt-
built U-Boot. This is because the vendor loader resets the ledbar
controller while our U-Boot doesn't care.
Add reset-gpio so the ledbar driver in Linux will always reset the
ledbar controller.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 0b5cf952cf5b70d3af2424d05b15ce2730a3dc11)

2 years agoramips: add support for Ubiquiti UniFi FlexHD
Sven Wegener [Sat, 1 May 2021 11:51:51 +0000 (13:51 +0200)]
ramips: add support for Ubiquiti UniFi FlexHD

Hardware
--------

- SoC: MediaTek MT7621AT with 128 MiB RAM and 32 MiB Flash
- Wi-Fi: MediaTek MT7603 (b/g/n, 2x2) and MediaTek MT7615 (ac, 4x4)
- Bluetooth: CSR8811 (internal USB, install kmod-bluetooth)

Installation
------------

1. Connect to the booted device at 192.168.1.20 using username/password
   "ubnt".

2. Update the bootloader environment.

   $ fw_setenv devmode TRUE
   $ fw_setenv boot_openwrt "fdt addr \$(fdtcontroladdr);
     fdt rm /signature; bootubnt"
   $ fw_setenv bootcmd "run boot_openwrt"

3. Transfer the OpenWrt sysupgrade image to the device using SCP.

4. Check the mtd partition number for bs / kernel0 / kernel1

   $ cat /proc/mtd

5. Set the bootselect flag to boot from kernel0

   $ dd if=/dev/zero bs=1 count=1 of=/dev/mtdblock4

6. Write the OpenWrt sysupgrade image to both kernel0 as well as kernel1

   $ dd if=openwrt.bin of=/dev/mtdblock6
   $ dd if=openwrt.bin of=/dev/mtdblock7

7. Reboot the device. It should boot into OpenWrt.

Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 820f0c07c544652c1c53a45262a481ac5dbe139b)

2 years agoath79: Make patches apply again
Daniel Golle [Sun, 18 Sep 2022 14:38:47 +0000 (15:38 +0100)]
ath79: Make patches apply again

The patch adding support for LEDs connected to a reset controller did
not apply any more, refresh it on top of current master.

Fixes: 53fc987b2552 ("generic: move ledbar driver from mediatek target")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 76fc277917fc847892dbbf525aea9f90a035d296)

2 years agogeneric: move ledbar driver from mediatek target
Sven Wegener [Mon, 2 May 2022 20:13:14 +0000 (22:13 +0200)]
generic: move ledbar driver from mediatek target

This moves the ledbar driver to generic, to be also used by the ramips target.

Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 53fc987b25524c28525313347f0e2e5653dce43c)

2 years agomediatek: add led count
Sven Wegener [Sat, 12 Jun 2021 19:56:22 +0000 (21:56 +0200)]
mediatek: add led count

The LEDs connected to the MCU are so-called smart LEDs and their signal is
daisy-chained. Because of this, the MCU needs to be told how many LEDs are
connected. It also means the LEDs could be individually controlled, if the MCU
has a command for this.

Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 76198e8f09a607c6aca2621f59d7a4cd7f6653e3)

2 years agomediatek: add initialization after reset
Sven Wegener [Sat, 12 Jun 2021 19:56:22 +0000 (21:56 +0200)]
mediatek: add initialization after reset

During GPIO initialization the pin state flips and triggers a reset of
the ledbar MCU. It needs to be moved through an initialization sequence
before working correctly.

Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 013a956f086dc5d88f6e9cbd852ed4d900d436e0)

2 years agomediatek: add support for reset gpio
Sven Wegener [Sat, 12 Jun 2021 19:56:22 +0000 (21:56 +0200)]
mediatek: add support for reset gpio

Some versions of the ledbar MCU have a reset pin. It needs to be
correctly initialized or we might keep the MCU in reset state.

Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 84e4bbf5f049cdb862eb25c446bd6973ed4238aa)

2 years agomediatek: support reading more than one byte of response
Sven Wegener [Sat, 12 Jun 2021 19:56:21 +0000 (21:56 +0200)]
mediatek: support reading more than one byte of response

There are commands that return more than one byte of response.

Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 7c852e7df50e370b26b060c96c3cc4491cee88eb)

2 years agomediatek: cast literal value to char
Sven Wegener [Sat, 12 Jun 2021 19:56:21 +0000 (21:56 +0200)]
mediatek: cast literal value to char

Or the comparison against a signed char is always true, because the
literal 0xaa is treated as an unsigned int, to which the signed char is
casted during comparison. 0xaa is above the positive values of a signed
char and negative signed char values result in values larger than 0xaa
when casted to unsigned int.

Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit e9a22ce2535c4a0faf83d5595304fe0e2c7bd1b2)

2 years agomediatek: correctly log i2c response
Sven Wegener [Sat, 12 Jun 2021 19:56:21 +0000 (21:56 +0200)]
mediatek: correctly log i2c response

The read response is in the i2c_response variable. Also use %hhx format,
because we're dealing with a single char.

Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit c4f9f9b44c82ffeacd9d3c00ba07171c35831337)

2 years agomediatek: remove gpiod_direction_output()
Sven Wegener [Sat, 12 Jun 2021 19:56:21 +0000 (21:56 +0200)]
mediatek: remove gpiod_direction_output()

It's already set to output with GPIOD_OUT_LOW.

Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit a188356484f434c72ea536d91d7f93f59c562eb5)

2 years agomediatek: do not use gpiod_set_raw_value()
Sven Wegener [Sat, 12 Jun 2021 19:56:21 +0000 (21:56 +0200)]
mediatek: do not use gpiod_set_raw_value()

The polarity of the signal is set in the device dts.

Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 56c2d15587dfe5285f1cf3bfb236c827bc446465)

2 years agokernel: modules: package kmod-crypto-essiv
Daniel Golle [Sun, 11 Sep 2022 13:26:25 +0000 (14:26 +0100)]
kernel: modules: package kmod-crypto-essiv

Package kernel module providing ESSIV support for block encryption.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 4133102898502c9bb453e8603b6c891aa103bce4)

2 years agomediatek: fix typo in bpi-r64 image recipe
Daniel Golle [Thu, 8 Sep 2022 18:29:44 +0000 (19:29 +0100)]
mediatek: fix typo in bpi-r64 image recipe

Janusz Dziedzic reported a typo introduced by a recent commit. Fix it.

Fixes: 50c892d67b ("mediatek: bpi-r64: make initramfs/recovery optional")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 007c8809c1e888748af70c1b878f4d0105fe7531)

2 years agomediatek: bpi-r64: make initramfs/recovery optional
Daniel Golle [Thu, 8 Sep 2022 01:44:30 +0000 (02:44 +0100)]
mediatek: bpi-r64: make initramfs/recovery optional

Only include recovery image in SD card image generated for the
BananaPi BPi-R64 if building with CONFIG_TARGET_ROOTFS_INITRAMFS
This allows to build images larger than 32 MB (the limit for
initramfs/recovery image) by deselecting initramfs.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 50c892d67bb4af90861be9fbb6831c2eeb62de11)

2 years agoramips: fix GB-PC1 and GB-PC2 LEDs
Arınç ÜNAL [Wed, 14 Sep 2022 18:30:59 +0000 (21:30 +0300)]
ramips: fix GB-PC1 and GB-PC2 LEDs

Add the missing LEDs for GB-PC2. Some of these LEDs don't exist on the
device schematics. Tests on a GB-PC2 by me and Petr proved otherwise.

Remove ethblack-green and ethblue-green LEDs for GB-PC1. They are not wired
to GPIO 3 or 4 and the wiring is currently unknown.

Set ethyellow-orange to display link state and activity of the ethyellow
interface for GB-PC2.

Link: https://github.com/ngiger/GnuBee_Docs/blob/master/GB-PCx/Documents/GB-PC2_V1.1_schematic.pdf
Tested-by: Petr Louda <petr.louda@outlook.cz>
Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
(cherry picked from commit 2a6ef7f53d7b96e4ee0200196c28ed6b0a7c8465)

2 years agowireless-regdb: update to 2022-08-12
Nick Hainke [Sat, 17 Sep 2022 06:05:47 +0000 (08:05 +0200)]
wireless-regdb: update to 2022-08-12

Changes:
9dc9c89 wireless-regdb: update regulatory database based on preceding changes
442bc25 wireless-regdb: update 5 GHz rules for PK and add 60 GHz rule
daee7f3 wireless-regdb: add 5 GHz rules for GY

Signed-off-by: Nick Hainke <vincent@systemli.org>
(cherry picked from commit 1d2d69c810261308652a577fc136e7327e0e1358)

2 years agokernel: bump 5.10 to 5.10.143
John Audia [Thu, 15 Sep 2022 13:56:36 +0000 (09:56 -0400)]
kernel: bump 5.10 to 5.10.143

All patches automatically rebased.

Signed-off-by: John Audia <therealgraysky@proton.me>
(cherry picked from commit b77217d916ba48accbdb46589e6e5e3a9972c44e)

2 years agokernel: bump 5.10 to 5.10.142
John Audia [Thu, 8 Sep 2022 10:19:20 +0000 (06:19 -0400)]
kernel: bump 5.10 to 5.10.142

All patches automatically rebased.

Signed-off-by: John Audia <therealgraysky@proton.me>
(cherry picked from commit 837fd23c80b22c566b40181a28630ac29af8f96e)

2 years agokernel: bump 5.10 to 5.10.141
John Audia [Mon, 5 Sep 2022 10:47:34 +0000 (06:47 -0400)]
kernel: bump 5.10 to 5.10.141

All patches automatically rebased.

Signed-off-by: John Audia <therealgraysky@proton.me>
(cherry picked from commit e8a62a1e6040de71bdcb1be1a1c30c064acbc059)

2 years agokernel: bump 5.10 to 5.10.139
John Audia [Mon, 29 Aug 2022 15:30:27 +0000 (11:30 -0400)]
kernel: bump 5.10 to 5.10.139

All patches automatically rebased.

Signed-off-by: John Audia <therealgraysky@proton.me>
(cherry picked from commit e0753c5d5cef5b03c60601364188afb262ccd02e)

2 years agokernel: build crypto md5/sha1/sha256 modules for powerpc
Josef Schlehofer [Tue, 30 Aug 2022 07:02:32 +0000 (09:02 +0200)]
kernel: build crypto md5/sha1/sha256 modules for powerpc

This builds and enables kernel optimized modules for mpc85xx target:
- CONFIG_CRYPTO_MD5_PPC [1]
- CONFIG_CRYPTO_SHA1_PPC_SPE [2]
- CONFIG_CRYPTO_SHA256_PPC_SPE [3]

Where it was possible, then use Signal Processing Engine, because
CONFIG_SPE is already enabled in mpc85xx config.

[1] https://cateee.net/lkddb/web-lkddb/CRYPTO_MD5_PPC.html
[2] https://cateee.net/lkddb/web-lkddb/CRYPTO_SHA1_PPC.html
[3] https://cateee.net/lkddb/web-lkddb/CRYPTO_SHA256_PPC_SPE.html

Signed-off-by: Josef Schlehofer <pepe.schlehofer@gmail.com>
(cherry picked from commit 3a702f8733ff371f30e9e3ba1e1aed5f4686b6b4)