Following the initial support of the Cudy WR3000H with PHY C22 for the 2.5G WAN
interface, several improvements fixing issues with RealTek RTL8221B PHYs have
been merged.
Therefore we can now bring the DT in line with other equipment and declare the
2.5G WAN PHY as C45.
Fixes: 9d66b8b ("mediatek: filogic: Add support for cudy wr3000h")
Signed-off-by: Juan Pedro Paredes Caballero <juanpedro.paredes@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/17739
[reword commit description]
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit dcc5587374)
Link: https://github.com/openwrt/openwrt/pull/19124
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The manufacturer Cudy usually releases signed openwrt firmware, to
facilitate the migration from the proprietary version to the official
versions of openwrt. In contact with the manufacturer tells me that only
releases the firmware of the WR3000H if and only if
there is an official version. With this proposal I pretend to have an
initial operative version so that they do their part, and facilitate to
the users the possibility of using openwrt. In the present state, it is
only possible to use this firmware by uploading and installing it with
UART connection.
AX3000 2.5G Dual Band Wi-Fi 6 Mesh Router (WR3000H)
Hardware
--------
MediaTek MT7981 WiSoC
256MB DDR3 RAM
128MB SPI-NAND (XMC XM25QH128C)
MediaTek MT7981 2x2 DBDC 802.11ax 2T2R (2.4 / 5)
4 LAN MediaTek MT7531 PHY
1 WAN RTL8221B-VB-CG 2.5Gbps PHY (C22)
2 Radios MT7976CN
UART: 115200 8N1 3.3V
MAC:
LAN MAC: label mac
WAN MAC: label mac + 1
2.4G MAC: label mac
5G MAC: label mac + 1 with LA bit set
Installation
------------
1. Connect to the serial port as described in the "Hardware" section.
2. Power on the device + press reset pin. Keep pressing reset pin to
enter the U-Boot shell (The recovery.bin image load process must fail).
3. Download the OpenWrt initramfs image. Place it on an TFTP server
connected to the Cudy LAN ports. Make sure the server is reachable at
192.168.1.88. Rename the image to "cudy3000h.bin"
4. Download and boot the OpenWrt initramfs image.
$ tftpboot 0x46000000 cudy3000h.bin; bootm 0x46000000
5. IMPORTANT: Make backup from original firmware. System -> Backup
/Flash Firmware -> Save mtdblock contents. All mtdblock one by one,
keep unaltered (BL2, u-boot-env, Factory, bdinfo, FIP, and ubi).
6. Transfer the OpenWrt sysupgrade image to the device using scp.
Install with sysupgrade.
Warning for BL2 and U-BOOT developers
-------------------------------------
The nand partition layout from vendor is slightly diferent from "standard".
The FIP partition starts at 0x3c0000 be carefull with BL2 to BL31.
The UBI partition start at 0x5c0000 be carefull.
DO NOT OVERWRITE bdinfo partition it contains hardware MAC definition
Layout is start-end (not start size)
- 0x000000000000-0x000007800000 : "nmbm0"
- 0x000000000000-0x000000100000 : "bl2"
- 0x000000100000-0x000000180000 : "u-boot-env"
- 0x000000180000-0x000000380000 : "factory"
- 0x000000380000-0x0000003c0000 : "bdinfo"
- 0x0000003c0000-0x0000005c0000 : "fip"
- 0x0000005c0000-0x0000045c0000 : "ubi"
ALLWAYS for U-BOOT operations check this
setenv mtdids nmbm0=nmbm0
setenv mtdparts nmbm0:1024k(bl2),512k(u-boot-env),2048k(factory),256k(bdinfo),2048k(fip),65536k(ubi)
Signed-off-by: Juan Pedro Paredes Caballero <juanpedro.paredes@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/17458
(cherry picked from commit 9d66b8b312)
Link: https://github.com/openwrt/openwrt/pull/19124
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Multiple users have reported a regression [1] in OpenWRT 24.10 with the
ramips/mt7621 target, which has the MT7530 PHYs: the Ethernet link is
periodically going down for a brief period of time:
mt7530-mdio mdio-bus:1f lan1: Link is Down
br-lan: port 1(lan1) entered disabled state
mt7530-mdio mdio-bus:1f lan1: Link is Up - 1Gbps/Full - flow control rx/tx
The symptoms stop after disabling EEE and it was reported by Mediatek in
2021 that EEE is unstable for the MT7530 PHYs [2]:
> EEE of the 10-year-old MT7530 internal gephy has many IOT problems, so
> it is recommended to disable its EEE.
EEE is enabled by default for these devices in OpenWRT 24.10 whereas in the
previous version (OpenWRT 23.05, Linux 5.15) it was not. It was determined
that in Linux 6.6, the PHY driver tries to disable EEE in
mtk_gephy_config_init() in drivers/net/phy/mediatek-ge.c, but this is later
overridden by a subsequent execution of the genphy_c45_write_eee_adv()
function, which enables every EEE mode supported.
The best way forward for now seems to be to mark EEE as broken directly in
the devicetree, which affects the genphy_c45_write_eee_adv() function.
There are some devices, like GnuBee GB-PC2, that define additional PHYs,
for example ethernet-phy@5 or ethernet-phy@7. As reported by Chester A.
Unal, these are not MT7530 PHYs and they are not affected.
This would need to be cherrypicked for the OpenWRT 24.10 branch.
[1] https://github.com/openwrt/openwrt/issues/17351
[2] https://lore.kernel.org/all/0adde34f936a2dafca40b06b408d82afe0852327.camel@mediatek.com/
Tested-by: Darren Tucker
Signed-off-by: Mateusz Jończyk <mat.jonczyk@o2.pl>
Closes: https://github.com/openwrt/openwrt/issues/17351
Link: https://github.com/openwrt/openwrt/pull/18585
(cherry picked from commit b7fa9d92ae)
Signed-off-by: Mateusz Jończyk <mat.jonczyk@o2.pl>
Tested-by: Mateusz Jończyk <mat.jonczyk@o2.pl>
Link: https://github.com/openwrt/openwrt/pull/19150
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The I2C and SPI packages required for each RPi generation is different.
Therefore, in order to avoid confusion let's select them by default.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit 9117ff7b39)
Add kernel package for Broadcom Settop/DSL I2C controller.
This controller is used on RPi devices.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit 59f8312400)
Add kernel package for DesignWare I2C platform controller.
This controller is used on the RP1 SoC found on RPi 5 devices.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit 84922a1e34)
Add kernel packages for DesignWare SPI core and MMIO controllers.
This is needed for the RP1 SoC found on RPi 5 devices.
Tested with a Microchip ENC28J60 Ethernet controller on a RPi 5.
Signed-off-by: Albrecht Lohofener <albrechtloh@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/19049
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit 82d20a0fbd)
Since the early beginning of the Realtek DSA driver there is an uncovered
locking issue between the standard (parent) mdio bus and the DSA (child)
mdio bus. This comes from the fact that the DSA bus simply links to the
parent read and write functions and calls them directly. This leads to
the following lock issue.
- Child bus calls phy_read/write functions and uses its internal lock
- Parent bus calls phy_read/write functions and uses its internal lock
It becomes clear that critical section can be accessed twice without
knowing that a operation from the other bus is currently active. This
can lead to critical malfunctions because the mdio driver needs a lot of
internal magic to get page selection done right. Effects are:
- The original page is lost after a phy_write/read_paged() call
- dmesg like "Realtek RTL8218B (external) rtl838x slave mii-0:00:
Expected external RTL8218B, found PHY-ID 6b23"
Other DSA drivers simply use the read/write functions from the parent bus
and thus avoid locking issues. Do it the same way.
Fixes: 2b88563ee5 ("realtek: update the tree to the latest refactored version")
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/18824
Signed-off-by: Robert Marko <robimarko@gmail.com>
(cherry picked from commit 461fc06f9d)
Link: https://github.com/openwrt/openwrt/pull/18755
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Cleanup the code of the RTL83xx packet receive interrupt handler. Not
only for better readability but to avoid inconsistencies and stalls on
the RTL839x targets.
The current implementation seems to come from the GPL source code.
Calling the existing cleanup() function inside the interrupt context
without any locks conflicts with SMP & NAPI polling and makes things
worse instead of giving any benefit. Simply ignore RX buffer overruns
and let the device handle packet dropping itself.
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/18855
Signed-off-by: Robert Marko <robimarko@gmail.com>
(cherry picked from commit 8dde1e4638)
Link: https://github.com/openwrt/openwrt/pull/18755
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The RTL8214FC currently uses generic PHY functions. That makes it look like a copper
device. Switching to/from fibre works fortunately but the autonegotiation handling
still works on MII_LPA (PHY register 5) as if a copper link is used. Fix that by
- advertising a superset of TP/FIBRE features
- using clause 37 functions when on fibre
Additionally enhance the code of the driver to assist further development.
- log the speed of the inserted module to detect wrongly inserted 10gbase-r modules
- order phy driver functions alphabetically (keep match/name on top)
- remove genphy_loopback as the kernel uses it if not provided
Remark! The driver internally uses PORT_MII for the TP port. Align with that and
report MII to ethtool instead of TP. Other drivers do the same and it can be
changed in the future if needed.
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/18724
Signed-off-by: Robert Marko <robimarko@gmail.com>
(cherry picked from commit 4cfd1c4501)
Link: https://github.com/openwrt/openwrt/pull/18755
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The Arcadyan WE410443 is a WiFi AC access point distributed by various ISPs
under various names, including KPN SuperWifi and BT Whole Home Wi-Fi. It
features one ethernet port, dual MT7615N radios and four internal antennas.
Hardware:
- SoC: Mediatek MT7621AT
- Flash: 32 MB
- RAM: 128 MB
- Ethernet: 1x 10/100/1000 Mbps, built into the SoC
- WLAN: 2x MediaTek MT7615N
- Buttons: 1 Reset button, 1 WPS button
- LEDs: 1x Green, 1x Blue, 1x Red, all unmarked
- Power: 12 VDC, 1.5A barrel plug
Installation:
The bootloader is locked with a password, so the image needs to be written
directly to the SPI flash chip. To do this, you need to open up the case,
remove the heatsink and connect the flash chip to a Raspberry Pi. Use the
following connections:
Flash chip --> Raspberry Pi
VCC --> 3v3
RESET --> 3v3
/CS --> GPIO 8
DO --> GPIO 9
CLK --> GPIO 11
DI --> GPIO 10
GND --> Ground
You can solder wires to the flash chip, or use a SOIC16 clip. More details on
the Raspberry Pi and SPI chip pinouts are available on the wiki [1]
When you have the Raspberry Pi connected to the flash chip, boot your Pi and
follow the instructions:
1) Make sure your Pi has SPI enabled with sudo raspi-config
2) Install necessary tools: sudo apt install xxd libubootenv-tool mtd-utils
3) Upload overlay and execute:
sudo dtc -@ -I dts -O dtb -o
/boot/overlays/we410443.dtbo we410443-overlay.dts
4) Enable in /boot/firmware/config.txt by adding a new line containing
dtoverlay=we410443
5) Reboot your Pi and verify the mtd partitions with
cat /proc/mtd, you should see:
dev: size erasesize name
mtd0: 02000000 00001000 "all"
mtd1: 00030000 00001000 "u-boot"
mtd2: 00010000 00001000 "u-boot-env"
mtd3: 00010000 00001000 "factory"
mtd4: 01f60000 00001000 "firmware"
mtd5: 00010000 00001000 "glbcfg"
mtd6: 00010000 00001000 "config"
mtd7: 00010000 00001000 "glbcfg2"
mtd8: 00010000 00001000 "config2"
6) Optionally (but recommended), make a backup:
sudo dd if=/dev/mtd0 of=backup.bin
It can be restored with: sudo flashcp backup.bin /dev/mtd0
7) Set the variables for the bootloader:
echo '/dev/mtd2 0x0 0x1000 0x1000' > fw_env.config
sudo fw_setenv -c fw_env.config bootpartition 0
8) Finally, flash the image:
sudo flashcp openwrt-ramips-mt7621-arcadyan_we410443-
squashfs-sysupgrade.bin /dev/mtd4
MAC addresses
The label address is stored in ASCII in the config partition
Use --> Address
Device --> label
Ethernet --> label
WLAN 2g --> + 1
WLAN 5g --> + 2
References:
[1] https://openwrt.org/toh/arcadyan/astoria/we410443
Signed-off-by: Sander van Deijck <sander@vandeijck.com>
Link: https://github.com/openwrt/openwrt/pull/17981
(cherry picked from commit b6a07dd091)
Link: https://github.com/openwrt/openwrt/pull/19068
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Commit 7cace002ba added a generic kernel patch that exposes a new
symbol REALTEK_PHY_HWMON when REALTEK_PHY and HWMON are enabled. The new
symbol was added to kmod-phy-realtek, but the kmod is not used in the
rockchip target.
Signed-off-by: Tianling Shen <cnsztl@immortalwrt.org>
Link: https://github.com/openwrt/openwrt/pull/18921
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Now that we have a board file, add calibration variant for TP-Link
Archer C6 v2 and add ipq-wifi package for it.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit 427c4aa266)
Signed-off-by: Leopoldo Pla <leopoldoplasempere@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/19035
Signed-off-by: Nick Hainke <vincent@systemli.org>
Enable activity on the WiFi LEDs of the Teltonika RUTX50
like other boards in the ipq40xx target.
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit 54463f1e2e)
The Teltonika RUTX50 mac-addresses on its wired interfaces are currently
random on every boot.
Setting the mac-addresses from device-tree using nvmem does not work, as
the vendor bootloader mangles the mtd partitions, removing the
nvmem-cells property.
To remedy the random mac-addresse, set the correct ones in preinit.
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit 08c93512fe)
Turn on the 5G modem of the RUTX50 on by default.
This allows to make the modem detectable on a fresh
installation OOTB without further intervention.
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit ac1ad1a7ad)
Backport a small fix for brcm legacy dsa tags that has been accepted for
linux v6.16.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit 84eefb493b)
Fixes SPI and HSSPI coexistance on bmips bcm6362 and bcm63268.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit ba1017bf74)
The RPi upstream repo switched to the upstream SDHOST driver some time ago, so
it's time to do the same in our configs.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit cd79b6ede9)
Import pending PXA I2C recovery fixes so that if I2C recovery is enabled in
the DTS it does not completely break I2C as it currently does since kernel
6.6.
Signed-off-by: Robert Marko <robert.marko@sartura.hr>
(cherry picked from commit dfacae3049)
Backport upstream pinctrl fixes, these are required for the follow-up I2C
PXA recovery fixes.
Signed-off-by: Robert Marko <robert.marko@sartura.hr>
(cherry picked from commit 831994e258)
Commit d7e82c78d7 added a generic kernel patch that exposes a new
symbol REALTEK_PHY_HWMON when REALTEK_PHY and HWMON are enabled. The new
symbol was added to kmod-phy-realtek, but the kmod is not used in the
realtek target.
Fixes: d7e82c78d7 ("generic: backport Realtek PHY patches from upstream")
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
(cherry picked from commit ab87087672)
This fixes the handling of some FS copper SFP modules using the RollBall
protocol and needing some extra treatment.
Signed-off-by: Martin Schiller <ms@dev.tdt.de>
(cherry picked from commit bbe58f9830)
Sync jitterentropy source code with linux-6.12 to solve the
issue of jitterentropy initialization failed:
[ 9.523489] jitterentropy: Initialization failed with host not compliant with requirements: 9
[ 9.661916] kmodloader: 1 module could not be probed
[ 9.662377] kmodloader: - jitterentropy_rng - 0
In linux upstream commit cf27d9475f37 ("crypto: jitter - use
permanent health test storage"), when FIPS crypto is disabled,
the health test results are always explicitly skipped. That means
it will never return error code 9 (health test failed) again.
Fixes: https://github.com/openwrt/openwrt/issues/16684
Signed-off-by: Shiji Yang <yangshiji66@outlook.com>
Link: https://github.com/openwrt/openwrt/pull/18399
(cherry picked from commit eec11fbbb6)
Link: https://github.com/openwrt/openwrt/pull/18822
Signed-off-by: Robert Marko <robimarko@gmail.com>
Fixes the following warning:
[ 180.314652] NOHZ tick-stop error: local softirq work is pending, handler #08!!!
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry picked from commit 610dd871aa)
The web-recovery of the Genexis EX400 validates uploaded images to fit
in the rootf_0 partition.
With OpenWrt, only the kernel is stored in this partition, leaving the
partition very small. Currently, the first factory release image won't
be accepted by the recovery interface after the OpenWrt installation.
Pad the image of the ubifs to 10MB. This allows the 24.10 release image
to be uploaded, enabling device recovery.
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit eea4689654)
bcm6328, bcm6362 and bcm6368 targets are missing a key config symbol which
prevents the NAND from working.
Fixes: f6c02b014d ("bmips: 6.1: refresh config and add missing symbols")
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit 61da745817)
Add missing kmod-7915e package to get both wireless interfaces working.
Link: https://github.com/openwrt/openwrt/issues/18690
Reviewed-by: Steffen Förster <nemesis@chemnitz.freifunk.net>
Signed-off-by: Christoph Krapp <achterin@gmail.com>
(cherry picked from commit 0d6ba33181)
Configure the i2c pins as GPIO outputs and use the i2c-gpio driver to
control the Semtech SX9512 touch controller.
This fixes spurious errors in i2c transactions even at 1kHz with the
native i2c driver.
leds green:wan: Setting an LED's brightness failed (-6)
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit f828be4d10)
Add the necessary package dependencies as well as device-tree properties
to support the touch-inputs as well as missing LEDs on the Genexis Pulse
EX400 range extender.
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit 49a1781d74)
This driver is required for the touch-inputs as well as some LEDs on the
Genexis Pulse EX400.
The driver was also sent upstream. As the EX400 is currently the only
consumer, the driver is added target-specific for ramips.
Once the driver has been accepted upstream and is provided in a kernel
release used by OpenWrt, the package should be moved to the global input
drivers.
Link: https://www.spinics.net/lists/kernel/msg5669349.html
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit 0842e79338)
The alternative model name does not need to be represented in the DTS.
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit ff55daf1eb)
The Xiaomi MiWiFi 3A wireless router has a similar system architecture as the Xiaomi Mi 4A router, which is already officially supported by OpenWrt.
Product website: https://www.mi.com/miwifi3a
Device specification
--------------------
SoC: MT7628AN MIPS_24KEc @ 580 MHz 2.4G-bgn 2x2
WiFi: MT7612EN 5G-an, ac 80 MHz 2T2R
Flash: 16 MB
DRAM: 64 MB
Switch: MT7628AN (integrated in SoC)
Ethernet: 1 x 10 /100 Mbps
USB: None
Antennas: 2 x 2,4 GHz and 2 x 5 GHz (all are external and non-detachable)
LEDs: blue/red/amber
Buttons: Reset
Serial: 115200,8n1
MAC addresses as verified by OEM firmware:
------------------------------------------
use address source
LAN *:DD factory 0x28
WAN *:DD factory 0x28
2g *:DE factory 0x4
5g *:DF factory 0x8004
OEM firmware uses VLAN's to create the network interface for WAN and LAN.
Bootloader info:
----------------
The stock bootloader uses a "Dual ROM Partition System".
OS1 is a deep copy of OS2.
The bootloader starts OS2 by default.
To force start OS1 it is needed to set "flag_try_sys2_failed=1".
How to install:
---------------
1- Use OpenWRTInvasion to gain Telnet, SSH and FTP access: https://github.com/acecilia/OpenWRTInvasion
[IP: 192.168.31.1 | Username: root | Password: root | FTP-Port: 21]
2- Connect to router using telnet or ssh.
3- Backup all partitions. Use command "dd if=/dev/mtd0 of=/tmp/mtd0". Copy /tmp/mtd0 to computer using ftp.
4- Copy openwrt-ramips-mt76x8-xiaomi_miwifi-3a-squashfs-sysupgrade.bin to /tmp in router using ftp.
5- Enable UART access and change start image to OS1.
nvram set uart_en=1
nvram set flag_last_success=1
nvram set boot_wait=on
nvram set flag_try_sys2_failed=1
nvram commit
6- Erase OS1 & OS2 and install OpenWrt
mtd erase OS1
mtd erase OS2
mtd -r write /tmp/openwrt-ramips-mt76x8-xiaomi_miwifi-3a-squashfs-sysupgrade.bin OS1
Credits:
--------
This PR is based on the work of Zehao Zhang (Github: @ZZH-Finalize) that he had published in the PR: #15698
Signed-off-by: Olgun Demir <olgun.demir@mail.com.tr>
Link: https://github.com/openwrt/openwrt/pull/18427
(cherry picked from commit c3b8108a2b)
Link: https://github.com/openwrt/openwrt/pull/18550
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The WL-WN573HX3 is an AX3000 outdoor Access Point by WAVLINK,
also sold in Europe as 7Links WLR-1300 (ZX-5612).
Specifications:
- MT7981B + MT7976 AX3000 2x2 DBDC (160 MHz)
- 16 MiB SPI NOR, 256 MiB RAM
- Gigabit ethernet port, 802.3af PoE
- IP67 outdoor case for wall or pole mounting with
four single band RP-SMA fiberglass antennas (8 dBi)
Installation:
- OEM Web UI is at 192.168.30.1 which will forward to
http://netlogin.link (using a captive portal)
- login with default password `admin`
- skip setup wizard by navigating directly to
http://netlogin.link/html/meshUpgrade.html
- upload WN573HX3-sysupgrade.bin
- reset to factory defaults to discard OEM UCI settings
MAC address assignment:
LAN 80:xx:xx:76:xx:25 hw 0x44e
WLAN 2.4G 80:xx:xx:76:xx:27 factory 0x04 (label MAC)
WLAN 5G 82:xx:xx:46:xx:27
pair key 8a:xx:xx:76:xx:27 also on label, not used by OpenWrt
Signed-off-by: Sebastian Schaper <openwrt@sebastianschaper.net>
Link: https://github.com/openwrt/openwrt/pull/18206
(cherry-picked from commit 907cb88e10)
Link: https://github.com/openwrt/openwrt/pull/18557
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
