airoha: add pending patch for USB support on AN7581
Add pending patch for USB support on AN7581 SoC. This is also required
to make operational the 3rd PCIe line that use the USB2 Serdes for PCIe
operations.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Felix Baumann [Sun, 26 Oct 2025 01:54:14 +0000 (02:54 +0100)]
realtek: rtl838x: fix regression in enable_phy_polling
Fix regression from back when support for RTL930x was added.
While at it replace 0x8000 by BIT(15).
Fixes: 27029277f98ddd0006175bdc5beec8b0b150f187 Tested-by: Jonas Jelonek <jelonek.jonas@gmail.com> Signed-off-by: Felix Baumann <felix.bau@gmx.de> Link: https://github.com/openwrt/openwrt/pull/20549 Signed-off-by: Robert Marko <robimarko@gmail.com>
Jonas Jelonek [Thu, 16 Oct 2025 16:01:11 +0000 (16:01 +0000)]
realtek: dsa,pcs: rtl930x: let PCS driver setup SerDes
Remove SerDes initialization/configuration calls from the DSA driver in
'rtl93xx_phylink_mac_config' and let our PCS driver setup the SerDes now
that the driver is able to do that.
Adjust some details in rtl93xx_phylink_mac_config to ensure the MAC is
properly disabled MAC before configuring the SerDes. This was done
within the SerDes code before.
Jonas Jelonek [Mon, 20 Oct 2025 22:06:33 +0000 (22:06 +0000)]
realtek: pcs: rtl930x: import SerDes setup code from PHY driver
Import SerDes configuration code from PHY driver into the PCS driver.
Only do mandatory adjustments, rename the function to adhere to the
naming scheme, adjust all SerDes access calls.
Comparing the partition contents of a bricked and a working unit showed
that on the bricked one the fake uImage header was missing. The UBI
partition also showed significant changes. Both are fixed when the
base DniImage receipt is used.
Christoph Krapp [Sun, 26 Oct 2025 20:16:50 +0000 (21:16 +0100)]
ipq40xx: improve Netgear LBR20 LED config
This aligns the LED behaviour with other Orbi devices. Orbi devices have
multiple RBG LEDs at the top of the device and two status LEDs at the
back next to the barrel jack.
The current behaviour of other Orbi devices is to use the multi-color
LEDs at the top for status indication and the green/red LEDs at the back
for running/panic-indication. This matches the vendor behaviour except
the color choice.
Other devices use green as running, blue on upgrade, red on failsafe and
white on bootup, so this aligns the LBR20 behaviour to the rest.
fc27940fe993 dhcpv6: support a configurable DUID ae16476bcad2 dhcpv6-ia: split reconf_msg struct into partial structs a3bf57a39958 dhcpv6: split clientid to a separate struct 14da3727ac0c dhcpv6-ia: simplify/fix IID calculations f49574e42e78 odhcpd: change "-c" cmd line arg to take a dir 0900202c2754 dhcpv4: dhcpv4_lease() - clarify which variables are requests 2c1e6e7c3ebf dhcpv4: dhcpv4_lease() - simplification 48487aedead8 dhcpv4: dhcpv4_lease() - convert to switch statement 3ad54baa4adc dhcpv4: dhcpv4_assign() - explain address assignment fab8f67818e6 dhcpv6: send timezone only when client requests
realtek: add 10G_QXGMII serdes mode support for RTL930x
In Realtek implementation USXGMII is divided in submodes:
- USXGMII_SX: 10G single link, equivalent of PHY_INTERFACE_MODE_USXGMII
- USXGMII_DX: 10G two links (2*5G ?),
- USXGMII_QX: 10G four links, presumably 4*2.5G, used with the RTL8224,
equivalent of PHY_INTERFACE_MODE_10G_QXGMII.
This CL adds the 10_GQXGMII modes to the RTL930x implementation. In
particular the "mode set" function is extended to support both simple
mode set, and force mode set depending on the mode according to
dal_longan_sds_mode_set [1].
Damien Dejean [Wed, 27 Aug 2025 14:05:39 +0000 (16:05 +0200)]
realtek: add serdes patch for 10G_QXGMII
Adds the serdes patch sequence [1] and configuration [2] for the
PHY_INTERFACE_MODE_10G_QXGMII mode (aka USXGMII_QX in Realtek sources).
It is required by devices with light bootloaders (ie not u-boot) that
does not initialize the hardware before booting the kernel.
Damien Dejean [Sat, 18 Oct 2025 15:01:24 +0000 (17:01 +0200)]
realtek: add serdes mapping for rtl930x
On the RTL930x series the serdes #3 is backed by serdes #10 when pages
0, 1, 2 or 3 are accessed [1]. This changeset modifies the sds mapping
function from a single implementation for the 3 families to one
implementation per chip family. In particular it implements the mapping
required for the rtl930x one.
The root file system is getting a bit bigger.
Without this commit: 5250380 bin/targets/armsr/armv8/openwrt-armsr-armv8-generic-rootfs.tar.gz 4919997 bin/targets/malta/be/openwrt-malta-be-default-rootfs.tar.gz
With this commit: 5263407 bin/targets/armsr/armv8/openwrt-armsr-armv8-generic-rootfs.tar.gz 4930011 bin/targets/malta/be/openwrt-malta-be-default-rootfs.tar.gz
Hauke Mehrtens [Sat, 25 Oct 2025 12:27:00 +0000 (14:27 +0200)]
uboot-sifiveu: fix build with swig 4.3.0
Fixes the following error by backporting upstream update:
```
scripts/dtc/pylibfdt/libfdt_wrap.c: In function ‘_wrap_fdt_next_node’:
scripts/dtc/pylibfdt/libfdt_wrap.c:5581:17: error: too few arguments to function ‘SWIG_Python_AppendOutput’
5581 | resultobj = SWIG_Python_AppendOutput(resultobj, val);
| ^~~~~~~~~~~~~~~~~~~~~~~~
```
This is the same fix as applied in 4e68103c4eb9 ("uboot-rockchip: fix build with swig 4.3.0")
Hauke Mehrtens [Fri, 24 Oct 2025 22:34:04 +0000 (00:34 +0200)]
valgrind: Update to version 3.26.0
The following patches were removed:
* 010-486180-Valgrind-MIPS-VexGuestArchState-has-no-member.patch [1]
* 020-mips-skip-using-shared-syscall-numbers-for-mips32.patch [2]
* 021-mips-skip-using-shared-syscall-numbers-for-mips64.patch [3]
Jonas Jelonek [Fri, 24 Oct 2025 21:40:44 +0000 (21:40 +0000)]
realtek: fix SFP GPIOs for XikeStor SKS8310-8X
Fix the GPIO assignment of RX-LOS and TX-DISABLE for all SFP ports. Both
were actually swapped when adding support for the device. Apparently,
this didn't cause any issues.
These paths have changed in recent kernels, so we need to adapte.
The CONFIG_GOOGLE_VPD kernel module (provided by kmod-google-firmware)
is present on OnHub, and provides alternative means to locate this
information, in /sys/firmware/vpd/ro/wifi_base64_calibration{0,1,2}. Use
that instead.
airoha: an7581: add Aeonsemi AS21xxx PHY to RFB eMMC board
Airoha AN7581 eMMC RFB board mount 2 Aeonsemi AS21xxx PHY. One is
usually connected to GDM4 and the other is optionally connected to GDM2.
Add the relevant nodes to support the one connected to GDM4.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Jonas Jelonek [Wed, 22 Oct 2025 07:21:09 +0000 (07:21 +0000)]
realtek: phy: rtl931x: remove SerDes code from PHY driver
Since ddf94f7489 and 4a5de35dba, a SerDes is configured by the PCS
driver. All code from PHY and DSA related to this has been imported and
adjusted into the PCS driver. Thus, remove the unused code from the PHY
driver now.
As mkits.sh only generates the relevant DTS for mkimage and is
specific for the building image, we can move it outside the locked
section as it doesn't do any operation that can be used concurrently by
others.
This won't have any real impact but clean the code making it clear what
needs to be protected and what can be executed concurrently.
image: fix race between mkits.sh and mkimage in .itb squashfs generation
With further investigation it was found a race in generating .itb images
that include a RootFS caused by the mkits.sh and the mkimage.
Due to the fact that mkits.sh generates a .pagesync image of the passed
rootfs, it can happen that, concurrently, mkimage can be called at the
same time mkits.sh is creating another .pagesync for the same rootfs.
This cause mkimage to use an half made rootfs.pagesync creating a
corrupted image.
To address this, also protect the mkimage with the same lock used for
mkits.sh preventing any kind of concurrent usage/generation of the
rootfs.pagesync blob
e2fsprogs: add pending patch that improve reproducibility
Add pending patch that disable libmagic support as this cause
reproducibility issue due to timing of presence of libmagic library at
times e2fsprogs gets compiled.
The new patch adds control on libmagic inclusion and we add the
--without-libmagic to the configure args by default.
Drop redundant patch for host build as this has been fixed in 1.44.1
but for some reason it wasn't notice with the patch directly refreshed
(with the fix right up)
airoha: backport additional upstream patch for NPU support for AN7583
Backport additional patch required for NPU support of Airoha AN7583.
These are specific for the NPU module with some minor fixes and to adds
upport for loading the specific Airoha AN7583 NPU firmware.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
airoha: an7583: drop bootloader artifacts entry and add default pkg
Fix the current image Makefile for Airoha AN7583 by dropping the
artifacts entry for bootloader and add default PKG for the PHY mounted
on the RFB boards.
The artifacts for the bootloader will be readded later once ATF-A
support will be pushed mainline.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
airoha: an7583: add additional comments on EVB board
On the Airoha AN7583 EVB there are 2 Aeonsemi AS21xxx PHY but only the
one connected to Serdes Ethernet is usable as the other connected to
Serdes PON is not actually HW connected (as the serdes is connected to
the BOSA port)
Add comments and disable the relevant PHY.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
0d0fac30075f dhcpv4: bump problem scenario up to warn bf6137092346 config: properly set log level from uci 7956f4271b4e dhcpv6: RFC4833 timezones 7000557cd8f6 dhcpv6-ia: respect prefix assigned to interface (>= /64) e1e60601ffeb odhcpd: improve odhcpd_urandom() c2eb4b59f107 config: fix erroneous clamp message if clamp value == max 54b9e729b00d dhcpv4: bail earlier on release/decline 417f4b11d352 dhcpv4: don't hardcode options array length d63fa3c3612c dhcpv4: shrink struct dhcpv4_message 9653b43617e3 dhcpv4: use iovec for forcereconf messages, fix hash bf41f4edfbe3 dhcpv4: fix padding of iovec message in dhcpv4_handle_msg() be68f423c528 dhcpv4: some minor cleanups post-iovec e24a371ef714 dhcpv4: use iovec for forcerenew opts bd353e891ae6 dhcpv4: use iovec for router and DNS server b81cfaa7859e dhcpv4: use iovec for DNS search and MTU 578a9289440b dhcpv4: use iovec for netmask/hostname/broadcast 5bafc17b79d8 dhcpv4: use iovec for leasetime/renew/rebind b63448ffe447 dhcpv4: introduce a reply_opts array 0533eaea0a94 dhcpv4: use iovec for DNR 6329e37d595d dhcpv4: use iovec for NTP 87fee619205d dhcpv4: use iovec for message and serverid 2f97bf0b56de dhcpv4: reorder some more variables in dhcpv4_handle_msg() 18c1b02bdc20 dhcpv4: remove one more variable from dhcpv4_handle_msg() 6fd691ff29cd dhcpv4: move dest handling from dhcpv4_handle_msg() 1f803caf9a1f dhcpv4: don't copy reqopts around b1be3984ebf8 dhcpv4: more refactoring of dhcpv4_handle_msg() 85717bedf8ce dhcpv4: clarify variable names in dhcpv4_handle_msg() be864ccf9919 dhcpv4: some more cleanups to dhcpv4_handle_msg() f87464520564 dhcpv4: preparations for iovec usage f48e1c205af3 odhcdp: use a more suitable clock 7e78caac4eae dhcpv6: change dhcpv6 message type check in relay 288abd9c4046 dhcpv6: move dhcpv6 message type check for early exit d504458ef515 odhcpd: add a simple build script 4ee309a54011 github: improve CI ff3a241ccc98 odhcpd: shrink binary size by creating a logging function e2ecf7ba6d72 odhcpd: support stderr logging 5de3b0d5b509 odhcpd: add log helpers 398d03a1a236 config: cap dhcpv6_pd_min_len to max instead of only logging error 4f54738d3ae7 config: clamp dhcpv6_hostid_len instead of only logging an error 465f19c9c2e3 config: clamp ra_mtu into 1280-65535 range 434b06133997 config: cap ra_retranstime and warn instead of only logging an error e5f58a90a147 config: cap ra_hoplimit to maximum and warn instead of logging an error 208eb10307c1 config: cap ra_reachabletime to RFC maximum instead of logging error 93449f1513b4 config: drop double size lease times; they are all UINT32_MAX; 439c0ceab131 router: redefine ra_mininterval and ra_maxinterval as uint32_t 84b4dfe81363 config: clamp ra_mininterval, ra_maxinterval, ra_lifetime at load time aa4f26232e05 router: refactor calc_ra_lifetime; redefine ra_lifetime as uint32_t 6ece28ffd475 config: do MaxRtrAdvInterval init at (ra_maxinterval) init time dc03e02d973e router: Apply updated values from RFC9096 (updates RFC4861) to RA/ND cc7766c12abe router: Apply updated values from RFC8319 (updates RFC4861) to RA/ND 964da13e758c config: refactor parse_leasetime() - branch amount remains same 9646c749467b github: fix CMAKE_SYSTEM_PROCESSOR copy&paste 288206c9a2ed github: add CI build 30780debd691 odhcpd: fix a compilation error e0b2c3cf9476 odhcpd: allow assignments to be reassigned 01e5e311b0db odhcpd: support multiple per-client DUIDs aebc647a6b7b odhcpd: support assignments on the basis of IAID cc3ec9c20c61 odhcpd: support IAIDs for static DHCPv6 leases e42c62725942 odhcpd: break up complex matching logic e1123906a4bc odhcpd: document the ubus interface c69200195263 dhcpv4: generate dbus events on lease expiry dd7a2d474d0d dhcpv4: fix ubus events 22481d848e0d odhcpd: remove mac_len argument to ubus_bcast_dhcp_event() d31d64efd56c odhcpd: fix ubus support flag in help msg 9bc1b4e26e10 odhcpd: reduce use of WITH_UBUS defines in code d402cdae4316 ndp: fix macOS IPv6 compatibility by using link-local source addresses
77e1ae21e67f odhcp6c: set server address from Information-request reply e24ac1cf8c27 dhcpv6: use a per-interface IAID for IA_NA requests 053e6c4b3863 github: improve CI b65ff293a969 dhcpv6: Check status code for IA_ADDR operations 3c7e425169e1 dhcpv6: Add error checking to handle renew failure for IA_PD operations 8de25373a860 dhcpv6: dhcpv6_parse_ia refactor to switch case ca3cd525f447 odhcp6c: fix deamon raw buffer inc d7afeea2b965 dhcpv6c: fix illegal DHCPV6_OPT_FQDN 70f7467db301 script: don't ignore RA with zero router lifetime 6ca1552cb3fe odhcp6c: sync and accumulate RA & DHCPv6 events as fast as possible 29b58cfb4711 odhcp6c: update cmake file ee6d4a82bbeb github: add CI build
Gain telnet access:
1. Login into web interface, and download the configuration.
2. Uncompress the configuration:
* Enter fakeroot if you are not login as root.
tar -zxf <filename>.tar.gz
3. Edit 'etc/passwd', remove root password: 'root::1:0:99999:7:::'.
4. Edit 'etc/rc.local', insert telnetd command before 'exit 0':
( sleep 3s; telnetd; ) &
5. Repack the configuration:
tar -zcf backup.tar.gz etc/
6. Upload new configuration via web interface, now you can connect to
KOMI A31 via telnet.
Flash instructions:
1. Connect to KOMI A31, backup everything, especially 'Factory' part.
2. Write new BL2:
mtd write openwrt-mediatek-filogic-konka_komi-a31-preloader.bin BL2
3. Write new FIP:
mtd write openwrt-mediatek-filogic-konka_komi-a31-bl31-uboot.fip FIP
4. Set static IP on your PC:
IP 192.168.1.254/24, GW 192.168.1.1
5. Serve OpenWrt initramfs image using TFTP server.
6. Cut off the power and re-engage, wait for TFTP recovery to complete.
7. After OpenWrt has booted, perform sysupgrade.
generic: backport phylink patches for PCS/PHY caps OPs
Backport phylink patches for PCS/PHY caps OPs. This makes it easier to
align future generic PCS patch and permit supporting special PHY that
needs specific tune if "in-band" mode is enabled (for example Aeonsemi
PHYs)
This is also mainly using the upstream version of the Mediatek patch
739-net-add-negotiation-of-in-band-capabilities.
Christoph Krapp [Thu, 25 Sep 2025 10:26:45 +0000 (12:26 +0200)]
ath79: add calibration variant for TP-Link Archer C60 v2
Now that we have a board file, add calibration variant for TP-Link
Archer C60 v2 and add ipq-wifi package for it.
Tested-by: Semih Baskan <strstgs@gmail.com> Signed-off-by: Christoph Krapp <achterin@gmail.com> Link: https://github.com/openwrt/openwrt/pull/20496 Signed-off-by: Robert Marko <robimarko@gmail.com>
L3 Offloading caused DHCP packets to be dropped at hardware level
And potentially buggy route implementation can cause a crash
Signed-off-by: Harshal Gohel <hg@simonwunderlich.de> Co-developed-by: Sven Eckelmann <se@simonwunderlich.de> Signed-off-by: Sven Eckelmann <se@simonwunderlich.de> Link: https://github.com/openwrt/openwrt/pull/20208 Signed-off-by: Robert Marko <robimarko@gmail.com>
realtek: rtl931x: Disable callbacks for l3 hw routing
The RTL931x is not supporting L3 offloading at the moment. To avoid crashes
when using this switch, simply disable L3 offloading completely.
Signed-off-by: Harshal Gohel <hg@simonwunderlich.de> Co-developed-by: Sven Eckelmann <se@simonwunderlich.de> Signed-off-by: Sven Eckelmann <se@simonwunderlich.de> Link: https://github.com/openwrt/openwrt/pull/20208 Signed-off-by: Robert Marko <robimarko@gmail.com>
Add more explict guards around the unsupported generation of the initramfs
images and suppress requirement for certain others.
Fixes: https://github.com/openwrt/openwrt/issues/20151 Fixes: https://github.com/openwrt/asu/issues/1499 Signed-off-by: Eric Fahlgren <ericfahlgren@gmail.com> Link: https://github.com/openwrt/openwrt/pull/20460 Signed-off-by: Robert Marko <robimarko@gmail.com>
Felix Fietkau [Tue, 21 Oct 2025 05:24:46 +0000 (07:24 +0200)]
netifd: update to Git HEAD (2025-10-20)
df2f5c9a30f8 veth: delete device before creation to handle stale state b6d371f30770 system-linux: handle RTM_DELLINK events for device state tracking 777f5942fa7d system-linux: only parse device status for AF_UNSPEC link messages
Rework the package SOURCE entry handling to account for the --root feeds
script feature.
Move the SOURCE entry string manipulation logic outside package-defaults.mk
in package.mk and limit only to non DUMP scenario to not pollute the .mk
too much.
Restructure the previous logic and add a new additional condition.
If we detect the package comes from a feed, replace any feed path that
have the _root prefix to the feed name with the non-root variant (the
feeds script create a symbolic link to it) and point the package SOURCE
entry to what the symbolic link points to.
Mark Mentovai [Mon, 20 Oct 2025 03:30:11 +0000 (23:30 -0400)]
ipq806x: restore recent changes made to 6.6 dts files
a33d59f7af8f (2025-10-14) restored dts files for kernel 6.6 to files-6.6
after the preceding 15fa59c41f0d moved files-6.6 to files-6.12 to be
used by the new testing kernel, 6.12. This restoration omitted 1a3f05eb2b8e (2025-10-08), which fixed important aspects of the wifi
device definitions in qcom-ipq8064-eax500.dtsi and
qcom-ipq8064-unifi-ac-hd.dts.
The 1a3f05eb2b8e fix persisted into the 6.12 dts files. This change now
restores it for 6.6. Note that ipq806x is currently using kernel 6.6 for
most purposes, as 6.12 is only in testing status for this target.
This change was prepared with `git diff a33d59f7af8f:target/linux/ipq806x/files-6.6 15fa59c41f0d^:target/linux/ipq806x/files-6.6`. This identified c186d17fa518 (2025-10-08) as another change that was not present in the
restored files-6.6, but it has since been superseded by 2a709d108e80
(2025-10-20), so no action is necessary to restore it. There were no
diffs present between these revisions in config-6.6 or patches-6.6,
which were also moved and restored in the 6.12 bring-up.
Signed-off-by: Mark Mentovai <mark@mentovai.com> Cc: Shiji Yang <yangshiji66@outlook.com> Cc: Rosen Penev <rosenp@gmail.com> Link: https://github.com/openwrt/openwrt/pull/20464 Signed-off-by: Robert Marko <robimarko@gmail.com>
3. Download the sysupgrade firmware at openwrt.bin.
4. The firmware size should be 7995392 bytes. Fix the size of your firmware putting zeros to the end, with:
wifi-scripts: ucode: export HE and EHT operation in scan results
Export WiFi 6E (HE) and WiFi 7 (EHT) operation data in scan results.
These additional data can be useful to check wifi channel utilization
by nearby stations.
Example:
Cell 32 - Address: xx:xx:xx:xx:xx:xx
Mode: Master Frequency: 6.115 GHz Band: 6 GHz Channel: 33
Signal: -14 dBm Quality: 70/70
Encryption: SAE (CCMP)
HE Operation:
Center Frequency 1: 39
Center Frequency 2: 47
Channel Width: 160 MHz
EHT Operation:
Center Frequency 1: 47
Center Frequency 2: 63
Channel Width: 320 MHz
The SPI chip select GPIO polarity is active low by default. We must
use "spi-cs-high" dts property to toggle the polarity. The polarity
on "cs-gpios" won't take effect at all[1]. Fix these incorrect GPIO
polarities to silence the kernel warnings.
[1] Refer to Linux/Documentation/devicetree/bindings/spi/spi-controller.yaml
```
device node | cs-gpio | CS pin state active | Note
================+===============+=====================+=====
spi-cs-high | - | H |
- | - | L |
spi-cs-high | ACTIVE_HIGH | H |
- | ACTIVE_HIGH | L | 1
spi-cs-high | ACTIVE_LOW | H | 2
- | ACTIVE_LOW | L |
Notes:
1) Should print a warning about polarity inversion.
Here it would be wise to avoid and define the gpio as
ACTIVE_LOW.
2) Should print a warning about polarity inversion
because ACTIVE_LOW is overridden by spi-cs-high.
Should be generally avoided and be replaced by
spi-cs-high + ACTIVE_HIGH.
```
These devices only have one SPI peripheral. And the chip select pin is
directly controlled by the SPI host hardware. Hence we don't need to
assign empty GPIO phandle for them. This patch also adjust the reg
address of the SPI peripheral node to follow the cs-gpios changes.
Shiji Yang [Sun, 24 Aug 2025 12:57:18 +0000 (20:57 +0800)]
ipq806x: dts: fix SPI chip select GPIO polarity
The SPI chip select GPIO polarity is active low by default. We must
use "spi-cs-high" dts property to toggle the polarity. The polarity
on "cs-gpios" won't take effect at all[1]. Fix these incorrect GPIO
polarities to silence the kernel warnings.
[1] Refer to Linux/Documentation/devicetree/bindings/spi/spi-controller.yaml
```
device node | cs-gpio | CS pin state active | Note
================+===============+=====================+=====
spi-cs-high | - | H |
- | - | L |
spi-cs-high | ACTIVE_HIGH | H |
- | ACTIVE_HIGH | L | 1
spi-cs-high | ACTIVE_LOW | H | 2
- | ACTIVE_LOW | L |
Notes:
1) Should print a warning about polarity inversion.
Here it would be wise to avoid and define the gpio as
ACTIVE_LOW.
2) Should print a warning about polarity inversion
because ACTIVE_LOW is overridden by spi-cs-high.
Should be generally avoided and be replaced by
spi-cs-high + ACTIVE_HIGH.
```
projects / thirdparty / openwrt.git / log
