変更履歴

toolchain: binutils: add version 2.45

Patches automatically rebased.

Signed-off-by: Rui Salvaterra <rsalvaterra@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/19547
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: 40203c5)

toolchain: binutils: default to 2.44 and drop 2.40

2.42 has been the default for over a year. 2.44 is well tested by now. While at
it, drop support for 2.40, as it's over two years old.

Signed-off-by: Rui Salvaterra <rsalvaterra@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/19547
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: 854d88b)

realtek: switch Netgear NGE devices to rt-loader

The bootloader on these devices uses 0x81000000 as load address for the
compressed image. Since the kernel uses a load address 0x80100000, this
only leaves a space of 15 MiB for the uncompressed image. For larger
images, the compressed data starts to get overwritten, and at some point
the boot will fail:

    ## Booting image from partition ... 0
    ## Booting kernel from Legacy Image at 81000000 ...
       Version:      9.9.9.9
       Created:      2025-08-07  14:56:09 UTC
       Data Size:    6756645 Bytes = 6.4 MB
       Checksum ... OK
       Uncompressing ... LZMA: uncompress or overwrite error 1 - must RESET board to recover

Currently, initramfs images with default config are already over the
limit. And while they still happen to work regardless, adding additional
packages easily pushes the size so much that the boot fails.

Fix this by switching to rt-loader (which relocates the data to the
upper end of the RAM before decompression). The switch includes regular
kernel images to avoid this becoming an issue again in the future.

Signed-off-by: Jan Hoffmann <jan@3e8.eu>
Link: https://github.com/openwrt/openwrt/pull/19734
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: 3e09991)

mediatek: remove loglevel in bootargs

`loglevel=8` causes the kernel to output all logs, including debug logs, at boot time

It is enabled by default on the upstream eval board because it is aimed at developer debugging.

Most devices reference the eval board directly without modification, and the debug log should be hidden at release

Signed-off-by: Coia Prant <coiaprant@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/19714
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: e596356)

tools: bash: fix PKG_HASH

Invalid hash was added during the 5.3 update and CI will fail since none
of the available sources has the tarball with incorrect hash.

Fixes: 423ed9fc5aea ("tools: bash: update to 5.3")
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: 2876106)

linux-firmware: update to 20250808

% git log --no-merges --pretty=oneline --abbrev-commit 20250708...20250808
53f89ed819bc qcom: Add QDSP firmware file for Qualcomm QDU100 device.
35abebaa15c1 ath12k: WCN7850 hw2.0: update to WLAN.HMT.1.1.c5-00302-QCAHMTSWPL_V1.0_V2.0_SILICONZ-1.115823.3
7aa2aba3a90e ath12k: QCN9274 hw2.0: update to WLAN.WBE.1.5-01651-QCAHKSWPL_SILICONZ-1
db7327128eff ath11k: WCN6855 hw2.0: update board-2.bin
14794133088c ath11k: QCA6698AQ hw2.1: update to WLAN.HSP.1.1-04650-QCAHSPSWPL_V1_V2_SILICONZ_IOE-2
effe7d041975 ath11k: QCA2066 hw2.1: update to WLAN.HSP.1.1-03926.13-QCAHSPSWPL_V2_SILICONZ_CE-2.52297.9
d4fde9ca4ecb ath11k: QCA2066 hw2.1: update board-2.bin
9188b156478e qcom: Update xbl_config firmware file.
ba2549fc9730 amdgpu: Update GCN 4.0.5 microcode
61d6a5f9f1e8 amdgpu: Update SDMA 6.1.0 microcode
67ee1b80467c amdgpu: Update GC 11.5.0 microcode
b2acd913eb00 qcom: Add QDU100 firmware image files required for booting.
224186ce2e8e linux-firmware: Add firmware for airoha-npu driver
ebf45cd0f8d0 linux-firmware: update firmware for MT7925 WiFi device
2ccae12c4f73 mediatek MT7925: update bluetooth firmware to 20250721233113
1d2771d839ba qcom: Update DSP firmware for qcm6490 platform
8ecf764788f8 qcom: update Venus firmware file for v6.0
c3855905442a i915: Xe3LPD DMC v2.29
3768c184de68 linux-firmware: Update AMD cpu microcode
02ad2a17abc4 qcom: Add QCS6490 symlink for QUPv3 firmware
530f28320603 qcom: Add firmware binary for SM8750.
dce0f1f8a3d2 amdgpu: update dmcub fw for dcn314
53d585cd763b cirrus: cs35l41: Add Firmware for various ASUS commercial Laptops using CS35L41 HDA
28964e023a39 cirrus: cs35l41: Update Firmware for Dell Oasis
5be0792a63e9 cirrus: cs35l56: Add firmware for Cirrus CS35L56 for various Dell laptops
b4cd8b7e766a qcom: Add Audio topology for QCS6490 RB3Gen2
dcd36f8607e5 intel_vpu: Update NPU firmware
43c969c12234 amdgpu: update dmcub fw for various DCN version
2aaf3bf0702c WHENCE: extract more license statements
f62714f9a0ac WHENCE: clarify io_ti origin
e2c1b151087b amdgpu: Update GC 11.5.1 microcode
eb03c8ce595d rtw89: 8852b: update fw to v0.29.128.0
351eb5c7643f rtw89: 8852bt: update fw to v0.29.127.0
ee12f2433585 rtw89: 8922a: add regd fw element with version R72-R6
75e7b2b760ae rtw89: 8852c: add regd fw element with version R72-R57
d300c8670364 rtw89: 8922a: update BB parameter V49
5bcef62c4107 qcom: Update gpu firmwares of QCS615 chipset
40c52dd245b4 linux-firmware: Update firmware file for Intel Solar core
a7be448d356e linux-firmware: Update firmware file for Intel BlazarU core
9ec63315f4c1 linux-firmware: Update firmware file for Intel BlazarI core

Build system: x86/64
Build-tested: x86/64-glibc
Run-tested: x86/64-glibc (Intel N150-based system)

Signed-off-by: John Audia <therealgraysky@proton.me>
Link: https://github.com/openwrt/openwrt/pull/19721
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
 (commit: 6152381)

realtek: RTL838x harden ethernet driver against fuzzying

While checking setup routines for stability and completeness,
random RTL838x SoC I/O areas were intentionally overwritten.
As soon as L2_CTRL_1->FAST_AGE_OUT is set to 1, the system
stalls during bootup. Analysis shows that it loops endlessly
in rtl838x_hw_stop()

/* Flush L2 address cache */
if (priv->family_id == RTL8380_FAMILY_ID) {
	for (int i = 0; i <= priv->cpu_port; i++) {
		sw_w32(1 << 26 | 1 << 23 | i << 5, priv->r->l2_tbl_flush_ctrl);
		do { } while (sw_r32(priv->r->l2_tbl_flush_ctrl) & (1 << 26));
	}

This is exactly the same logic as in the vendor GPL. There
are no hints about possible timeouts or issues. The reason is
still unclear. Nevertheless harden the function for further fuzzy
tests. Do this by resetting the configuration value to its SoC
default.

Additionally convert some shifts to BIT() for better readability.

Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/19679
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: 0724412)

qualcommax: ipq6018 rework nss_port5 clock to multiple conf

This fixes intermittent dmesg errors
"nss_port5_rx_clk_src: rcg didn't update its configuration."

Signed-off-by: Marko Mäkelä <marko.makela@iki.fi>
Link: https://github.com/openwrt/openwrt/pull/19620
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: 57bd20b)

tools: gnulib: always use std-gnu23 module

The new "std-gnu23" module has the stated goal of:
"...to update the c99 module to depend on std-gnu23
instead of on std-gnu11, and to make std-gnu11 obsolete."
in upstream commit 8990abb50 ("std-gnu23: new module").

However, for now, it's design is optional, so that
definitions of the latest standard module overrides the former.
At some point, upstream gnulib will replace the dependency
instead of add it alongside the older one.

Because all macros are copied to the aclocal directory,
for complex projects, not including the module
may cause the macros to apply only to some subdirectories
rather than all of them and top-level together.

For projects that import source from local gnulib,
always include the std-gnu23.m4 macros for consistency.

Signed-off-by: Michael Pratt <mcpratt@pm.me>
Link: https://github.com/openwrt/openwrt/pull/19748
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: a808086)

tools: libtool: always install all auxillary files

Remove the needless restriction
for which auxillary files are installed.
This aligns the behavior of libtoolize
with the existing behavior of automake.

This can prevent the seemingly random error
if autoreconf happens to be ran multiple times:

  configure: error: cannot find required auxiliary files: compile

Signed-off-by: Michael Pratt <mcpratt@pm.me>
Link: https://github.com/openwrt/openwrt/pull/19748
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: 4c0d64c)

gettext-full: force compiling target in C23 standard

Currently, the build system overrides the value of the CC variable
for actual compilation after configuring for target builds.
However, the configure script now modifies the CC variable
to include "-std=gnu23" when the test for C23 features is successful.
The configure script also tests for the ability to use alignof
without including the stdalign.h header, and only includes it if necessary.
The test in the configure script is being done with the C23 standard option
but the compilation is being done without the C23 standard option,
leading to an unusual build error where alignof() is not defined.

Resolving the conflict between the autoconf macros and the build system
causes several other packages to fail, so instead in the meantime,
force the use of C23 standard to compile as part of the new standard
includes alignof as a keyword to deprecate the stdalign.h macro.

Forcing use of the new standard is safe for target builds
as the toolchain is known to support the option
and is always within our scope of version control.

Signed-off-by: Michael Pratt <mcpratt@pm.me>
Link: https://github.com/openwrt/openwrt/pull/19748
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: b6f67ca)

gettext-full: remove usage of autoreconf

Running the bootstrap script autogen.sh
handles the execution of autotools already,
so calling autoreconf before configure
makes this happen twice, which is unnecessary
and can lead to an occasional build problem.

Signed-off-by: Michael Pratt <mcpratt@pm.me>
Link: https://github.com/openwrt/openwrt/pull/19748
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: d119f1b)

sunxi: change target to use FIT kernel images

Modernize the target slightly to use kernel+dtb FIT images in all
subtargets. LZMA compression will be used for the cortexa53 devices,
and we'll stay conservative and use gzip for the cortexa7/a8 devices
due to performance differences.

Tested-on:
 - Linksprite pcDuino v2 (cortexa8 / A10)
 - Olinuxino Micro (cortexa7 / A20)
 - Banana Pi M2 Berry (cortexa7 / V40)
 - Banana Pi P2 Zero (cortexa7 / H2+)
 - Xunlong Orange Pi 2 (cortexa7/ H3)
 - OrangePi PC Zero 2 (cortexa53 / H616)

Signed-off-by: Zoltan HERPAI <wigyori@uid0.hu>
 (commit: 8abb6db)

sunxi: enable ramdisk feature

Build initramfs images for this target.

Signed-off-by: Zoltan HERPAI <wigyori@uid0.hu>
 (commit: 3ded6d2)

ipq-wifi: update to Git HEAD (2025-08-12)

967cfe977926 qca9988: add BDFs for Xiaomi AIoT Router AC2350
6db497116709 qcn9274: update board file for 8devices Kiwi-DVK
7a161011d89d Add IPQ5018 and QCN9074 BDF for Yuncore AX850.
30b99fc9cc13 ipq6018: add jdcloud_re-ss-01 BDF

Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: f9a8a36)

procd: update to Git HEAD (2025-08-12)

dafdf98b03bf service: fix object nesting in the ubus get_data call

Signed-off-by: Felix Fietkau <nbd@nbd.name>
 (commit: 8130021)

netifd: update to Git HEAD (2025-08-12)

843a736a4e26 ubus: fix type check in procd ubus data callback
94c3869941b0 interface: avoid memleaks on invalid interfaces
6ead304877ad global: use blobmsg_parse_attr
813b9882dbab config: add support for defining devices via procd data
575b04326e32 config: add support for defining interfaces via procd data
a0be934399cf examples: sync wireless.uc

Signed-off-by: Felix Fietkau <nbd@nbd.name>
 (commit: be1c301)

mpc85xx: tl-wdr4900: add back 5ghz LED

In the conversion to dts, qca,led-pin was used for both interfaces.
Unfortunately, it's mutually exclusive with gpio-controller which made
it not do anything.

Fixes: 949e1a0 ("mpc85xx: tl-wdr4900: move platform code to dts")

Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/19758
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: cc0ff28)

wifi-scripts: ucode: get the phy capabilities for the correct band

On multi-band phys, capabilities can differ between bands.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
 (commit: d223d29)

mac80211: update to version 6.16

Signed-off-by: Felix Fietkau <nbd@nbd.name>
 (commit: e005cde)

realtek: backport upstream RTL9300 I2C driver

Backport/add all patches for the upstream RTL9300 I2C driver.

The upstream driver was added in 6.13 and was heavily based on our
downstream driver except for how the multiplexing behaviour is handled.
This driver was working fine with basic SFP operation on RTL930x devices
but there was no support for RTL931x though.

Major advantage over our downstream driver is: The multiplexing
behaviour is handled completely by the driver. Thus, there's no need for
a separate rtl9300-mux driver as we had it downstream. Moreover, this
simplifies the DTS of affected devices a lot since we can now move the
controller definition - which is in the DTS of each device so far - to
the base DTSI.

Currently pending patches are also included because the progress on
getting this upstream seems really slow right now, albeit upstream
maintainers may require several changes to the current state.

These include:
- patches fixing several issues in the driver
- patches doing a refactoring of the driver and adding support for RTL931x

See the commit messages included in each patch to have details on the
changes.

Signed-off-by: Jonas Jelonek <jelonek.jonas@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/19736
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: 44655c9)

realtek: adapt devices to backported i2c driver

Adapt the device tree definitions of rtl93xx devices and the base dtsi
for rtl930x and rtl931x to match with what's expected by the recently
backported RTL9300 I2C driver.

Signed-off-by: Jonas Jelonek <jelonek.jonas@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/19736
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: dc54af8)

realtek: drop downstream rtl93xx i2c driver

Drop our downstream driver in favor of an upstream existing driver which
is available starting from v6.13.

Signed-off-by: Jonas Jelonek <jelonek.jonas@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/19736
Signed-off-by: Robert Marko <robimarko@gmail.com>
 (commit: d0c64a1)

mt76: update to Git HEAD (2025-08-13)

00275548ba8f wifi: mt76: mt7996: Overwrite unspecified link_id in mt7996_tx()
2601bb91e078 wifi: mt76: mt7996: Fix mt7996_mcu_sta_ba wcid configuration
110ecc331123 wifi: mt76: mt7996: Fix mt7996_mcu_bss_mld_tlv routine
449a1aca9f72 wifi: mt76: mt7996: Initialize hdr before passing to skb_put_data()
1de707e73fc7 wifi: mt76: mt7921: don't disconnect when CSA to DFS chan
fce683dbcfc9 wifi: mt76: mt7925: fix locking in mt7925_change_vif_links()
11f721bce235 wifi: mt76: mt7925: fix the wrong bss cleanup for SAP
b3613d9a317d wifi: mt76: mt7996: Set def_wcid pointer in mt7996_mac_sta_init_link()
d1a2900a390d wifi: mt76: mt7996: Set proper link destination address in mt7996_tx()
56990ce9289a wifi: mt76: mt7925u: Add VID/PID for Netgear A9000
e796542a0585 wifi: mt76: mt7925: add MBSSID support
e35e4f8fcf8a wifi: mt76: fix potential memory leak in mt76_wmac_probe()
aa2f13c3be79 wifi: mt76: mt7996: disable beacons when going offchannel
630e166c958d wifi: mt76: prevent non-offchannel mgmt tx during scan/roc
d3e7c2051aae wifi: mt76: mt7996: use the correct vif link for scanning/roc
903a7e0ace79 wifi: mt76: mt7921: add MBSSID support

Signed-off-by: Felix Fietkau <nbd@nbd.name>
 (commit: 4010984)