uboot-ath79: add initial support for QCA955x SoCs

Add initial support for Qualcomm Atheros QCA955x series SoCs.
This support was based on the QCA956x support, QSDK, GPL tar of TP-Link
Archer C5 v1.20.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16297
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

mac80211: realtek: rtlwifi: add RTL8192DU support

Add Realtek RTL8192DU support to the rtlwifi package.
The RTL8192DU chipset is a 802.11a/b/g/n chip which supports 2.4 and 5 GHz at
up to 40 MHz channel bandwidth, three hardware variations exist:

* single MAC/ single PHY
* single MAC/ double PHY
* double MAC/ double PHY

This driver has been successully tested on a single MAC/ single PHY variant
0bda:8194 (DeLock 88540, https://www.delock.com/produkt/88540/merkmale.html)

rtl_usb: rx_max_size 15360, rx_urb_num 8, in_ep 1
rtl8192du: Driver for Realtek RTL8192DU WLAN interface
rtl8192du: Loading firmware file rtlwifi/rtl8192dufw.bin
ieee80211 phy0: Selected rate control algorithm 'rtl_rc'
usbcore: registered new interface driver rtl8192du

Supported interface modes:
* IBSS
* managed
* AP
* AP/VLAN
* monitor
* mesh point
* P2P-client
* P2P-GO

interface combinations are not supported

The new rtlwifi based driver rtl8192du has been merged in kernel v6.11.

Signed-off-by: Stefan Lippers-Hollmann <s.l-h@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/16721
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

linux-firmware: add firmware for Realtek RTL8192DU

This commit packages the newly merged firmware (v39.0) for Realtek RTL8192DU
802.11a/b/g/n USB wireless cards.

Signed-off-by: Stefan Lippers-Hollmann <s.l-h@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/16721
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

linux-firmware: Update to version 20241017

This updates the following firmware files:

amdgpu/gc_11_5_2_imu.bin
amdgpu/gc_11_5_2_me.bin
amdgpu/gc_11_5_2_mec.bin
amdgpu/gc_11_5_2_mes1.bin
amdgpu/gc_11_5_2_mes_2.bin
amdgpu/gc_11_5_2_pfp.bin
amdgpu/gc_11_5_2_rlc.bin
amdgpu/psp_14_0_4_ta.bin
amdgpu/psp_14_0_4_toc.bin
amdgpu/sdma_6_1_2.bin
amdgpu/vpe_6_1_3.bin
ath12k/QCN9274/hw2.0/board-2.bin
ath12k/QCN9274/hw2.0/firmware-2.bin
ath12k/QCN9274/hw2.0/Notice.txt
brcm/brcmfmac4354-sdio.nvidia,p2371-2180.txt
cirrus/cs35l54-b0-dsp1-misc-103c8d85-amp1.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d85-amp2.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d85-amp3.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d85-amp4.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d86-amp1.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d86-amp2.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d86-amp3.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d86-amp4.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d8c-amp1.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d8c-amp2.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d8d-amp1.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d8d-amp2.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d8e-amp1.bin
cirrus/cs35l54-b0-dsp1-misc-103c8d8e-amp2.bin
cirrus/cs35l54-b0-dsp1-misc-103c8e14-amp1.bin
cirrus/cs35l54-b0-dsp1-misc-103c8e14-amp2.bin
cirrus/cs35l54-b0-dsp1-misc-103c8e15-amp1.bin
cirrus/cs35l54-b0-dsp1-misc-103c8e15-amp2.bin
cirrus/cs35l54-b0-dsp1-misc-103c8e1b-amp1.bin
cirrus/cs35l54-b0-dsp1-misc-103c8e1b-amp2.bin
cirrus/cs35l56-b0-dsp1-misc-103c8d90-amp1.bin
cirrus/cs35l56-b0-dsp1-misc-103c8d90-amp2.bin
cirrus/cs35l56-b0-dsp1-misc-10431e13-amp1.bin
cirrus/cs35l56-b0-dsp1-misc-10431e13-amp2.bin
cirrus/cs35l56-b0-dsp1-misc-10431e13-amp3.bin
cirrus/cs35l56-b0-dsp1-misc-10431e13-amp4.bin
cirrus/cs35l56-b0-dsp1-misc-10431f43-amp1.bin
cirrus/cs35l56-b0-dsp1-misc-10431f43-amp2.bin
cirrus/cs35l56-b0-dsp1-misc-10431f43-amp3.bin
cirrus/cs35l56-b0-dsp1-misc-10431f43-amp4.bin
cirrus/cs35l56-b0-dsp1-misc-17aa380e-spkid0-ampl.bin
cirrus/cs35l56-b0-dsp1-misc-17aa380e-spkid0-ampr.bin
cirrus/cs35l56-b0-dsp1-misc-17aa380e-spkid1-ampl.bin
cirrus/cs35l56-b0-dsp1-misc-17aa380e-spkid1-ampr.bin
cirrus/cs35l56/CS35L56_Rev3.11.11.1.wmfw
cirrus/cs35l56/CS35L56_Rev3.11.20.wmfw
iwlwifi-bz-b0-fm-c0-94.ucode
iwlwifi-gl-c0-fm-c0-94.ucode
mediatek/BT_RAM_CODE_MT7961_1a_2_hdr.bin
mediatek/mt7988/mt7988_wo_0.bin
mediatek/mt7988/mt7988_wo_1.bin
mediatek/mt7996/mt7992_dsp_23.bin
mediatek/mt7996/mt7992_dsp.bin
mediatek/mt7996/mt7992_eeprom_23_2i5i.bin
mediatek/mt7996/mt7992_eeprom_23.bin
mediatek/mt7996/mt7992_eeprom_2i5e.bin
mediatek/mt7996/mt7992_eeprom_2i5i.bin
mediatek/mt7996/mt7992_eeprom.bin
mediatek/mt7996/mt7992_rom_patch_23.bin
mediatek/mt7996/mt7992_rom_patch.bin
mediatek/mt7996/mt7992_wa_23.bin
mediatek/mt7996/mt7992_wa.bin
mediatek/mt7996/mt7992_wm_23.bin
mediatek/mt7996/mt7992_wm.bin
mediatek/mt7996/mt7996_eeprom_233_2i5i6i.bin
mediatek/mt7996/mt7996_eeprom_233.bin
mediatek/mt7996/mt7996_eeprom_2i5i6i.bin
mediatek/mt7996/mt7996_rom_patch_233.bin
mediatek/mt7996/mt7996_wa_233.bin
mediatek/mt7996/mt7996_wm_233.bin
mediatek/WIFI_MT7961_patch_mcu_1a_2_hdr.bin
mediatek/WIFI_RAM_CODE_MT7961_1a.bin
qca/hmtbtfw20.tlv
qca/hmtnv20.b10f
qca/hmtnv20.b112
qca/hmtnv20.bin
qcom/a663_gmu.bin
qcom/qcm6490/a660_zap.mbn
qcom/qcm6490/adsp.mbn
qcom/qcm6490/adspr.jsn
qcom/qcm6490/adsps.jsn
qcom/qcm6490/adspua.jsn
qcom/qcm6490/battmgr.jsn
qcom/qcm6490/cdsp.mbn
qcom/qcm6490/cdspr.jsn
qcom/qcm6490/ipa_fws.mbn
qcom/sa8775p/a663_zap.mbn
realtek/rt1320/rt1320-patch-code-vab.bin
realtek/rt1320/rt1320-patch-code-vc.bin
rtl_nic/rtl8125d-1.fw
rtlwifi/rtl8192dufw.bin
rtw89/rtw8922a_fw-2.bin
ti/tas2781/TAS2XXX1EB30.bin
ti/tas2781/TAS2XXX1EB31.bin

Signed-off-by: Stefan Lippers-Hollmann <s.l-h@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/16721
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

ath79: apply qca,tx-gain-buffalo

It seems this was only used by one device in ar71xx. Might as well
reapply it.

Make use of the separate TX gain table for WZR-HP-G450H:
https://github.com/torvalds/linux/commit/0f978bfaf2600040f9826f394eb12e023bef3068

Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/15949
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

ramips: mt7621: add support for Gemtek WVRTM-127ACN and WVRTM-130ACN

The Gemtek WVRTM-127ACN is an indoor dual band wifi router
with internal antennas and 3 Gigabit Ethernet ports.

The Gemtek WVRTM-130ACN is an indoor dual band wifi router
with external antennas and 5 Gigabit Ethernet ports.

Hardware of WVRTM-127ACN:
- SoC: Mediatek MT7621AT (880 MHz, dual core)
- RAM: 128 MB
- Storage: 128 MB NAND SLC flash
- Ethernet: 3x 10/100/1000 Mbps LAN1,LAN2 & WAN
- Wireless: 2.4GHz: Mediatek MT7603EN (802.11b/g/n)
- Wireless: 5GHz: Mediatek MT7612EN (802.11n/ac)
- LEDs: 11x
- Buttons: 2x WPS, reset
- USB: 1x 3.0
- Power: 56 VDC, 0.54 A, PoE+ IN (WAN)
- PoE: 1x PoE+ 802.3af/at (WAN)
- Uart: GND RX TX VCC - J2 (GND near WAN)
- Board silkscreen: "WVRTM-127ACN_V02" "19K-513-8500R" "RoHS" "1717"

Hardware of WVRTM-130ACN:
- SoC: Mediatek MT7621AT (880 MHz, dual core)
- RAM: 128 MB (Kioxia TC58BVG0S3HTA00)
- Storage: 128 MB NAND SLC (Winbond W971GG6SB-25)
- Ethernet: 5x 10/100/1000 Mbps LAN1,LAN2,LAN3,LAN4 & WAN
- Wireless: 2.4GHz and 5GHz Mediatek MT7615DN (802.11ac/b/g/n) (DBDC)
- LEDs: 10x
- Buttons: 3x Power, WPS, reset
- USB: 1x 3.0
- Power: 56 VDC, 0.54 A, PoE+ (WAN)
- PoE: 1x PoE+ 802.3af/at (WAN)
- Uart: GND RX TX VCC - J2 (GND near WAN)
- Board silkscreen: "WVRTM-130ACN_V01" "19K-515-4500R" "RoHS" "2112"

Enable access to uboot menu (needed in wvrtm-130acn):
- The access to uboot menu is blocked by `bootdelay = 0` set in ubootenv.
With stock firmware version 01.01.02.163 and previous, you can use CVE 2020-24365
command injection https://nvd.nist.gov/vuln/detail/CVE-2020-24365
    python3 exploit.py -t 192.168.1.1 -c "fw_setenv bootdelay 3; fw_saveenv"

Backup the stock firmware:
- Connect via uart
- Connect via ethernet and assign your pc the address 192.168.15.x/24
- Power on the device; and start typing '4' to enter uboot menu
- Set factory mode and boot
    MT7621 # setenv factory 2; saveenv
    MT7621 # nand read 2800000 2000000 81000000; bootm
- Telnet and copy all mtd blocks
    telnet 192.168.15.1
- Copy all mtd blocks and start webserver
    for N in $(seq 0 6); do dd if=/dev/mtd$N of=/tmp/eeprom_mtd$N.bin; done
    mount -o bind /tmp /www
    lighttpd -f /etc/lighttpd.conf
- Backup stock rootfs_data (optional)
    dd if=/dev/mtd7 of=/tmp/eeprom_mtd7.bin
    dd if=/dev/mtd8 of=/tmp/eeprom_mtd8.bin
- Download to your pc from http://192.168.15.1/eeprom_mtd$N.bin

Installation:
- Connect via uart
- Connect via ethernet and assign your pc the address 10.10.10.3/24
- Start a tftp server and serve the image initramfs-kernel.bin
    mkdir /tmp/ftpd;
    cp initramfs-kernel.bin /tmp/ftpd/kernel.bin
    dnsmasq --enable-tftp --tftp-root=/tmp/ftpd
- Power on the device; and start typing '4' to halt the bootloader
- Change the active mtd partition from mtd6 to mtd5 (needed by uboot)
    MT7621 # setenv mtddevnum 5; saveenv
- Write the openwrt initramfs in ram via tftp and boot it
    MT7621 # tftpboot 81000000 kernel.bin; bootm
- From the initramfs create the ubi device and install openwrt via sysupgrade
    ubiformat /dev/mtd11 -y
    sysupgrade -n -v /tmp/sysupgrade.bin

Recovery:
Restore the stock firmware from the backup of the mtd blocks
    mtd write eeprom_mtd5.bin firmware
    mtd write eeprom_mtd6.bin Kernel2
    mtd write eeprom_mtd7.bin Storage1
    mtd write eeprom_mtd8.bin Storage2
    ubiformat /dev/mtd8 -y
    reboot

Links to previous works on wvrtm-127acn:
https://github.com/digiampietro/hacking-gemtek
https://forum.openwrt.org/t/add-support-for-gemtek-wvrtm-127acn-linkem-provider/168757

Signed-off-by: Samuele Longhi <agave@dracaena.it>
Link: https://github.com/openwrt/openwrt/pull/16685
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

kernel: bump 6.6 to 6.6.57

1. Update target/linux/generic/config-6.6 for new ksym
2. Refresh patches

Changelog: https://cdn.kernel.org/pub/linux/kernel/v6.x/ChangeLog-6.6.57

Added:
generic/backport-6.6/777-netfilter-xtables-fix-typo-causing-some-targets-to-not-load-on-IPv6.patch[1]

Manually rebased:
generic/hack-6.6/645-netfilter-connmark-introduce-set-dscpmark.patch

Removed upstreamed:
gemini/patches-6.6/0001-net-ethernet-cortina-Drop-TSO-support.patch[2]
gemini/patches-6.6/0004-net-ethernet-cortina-Restore-TSO-support.patch[3]

All other patches automatically rebased.

1. https://lore.kernel.org/all/20241019-xtables-typos-v2-1-6b8b1735dc8e@0upti.me/
2 https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v6.6.57&id=452c0740d72c6a77a41f6ddc318a48f18c3d2346
3. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v6.6.57&id=611f74b0e7fb93ee2366d9d7edca546806b220e9

Build system: x86/64
Build-tested: x86/64/AMD Cezanne, flogic/xiaomi_redmi-router-ax6000-ubootmod, ramips/tplink_archer-a6-v3
Run-tested: x86/64/AMD Cezanne, flogic/xiaomi_redmi-router-ax6000-ubootmod, ramips/tplink_archer-a6-v3

Signed-off-by: John Audia <therealgraysky@proton.me>
Link: https://github.com/openwrt/openwrt/pull/16726
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

base-files: increase default system log size to 128 kB

Increase the default system log buffer size option
in /etc/config/system from 64 kB to 128 kB.

64 kB is barely enough for the boot items of a modern router
with a few add-on packages, but any subsequent logging will
quickly cause the early boot items to get overwritten in the
round-robin log buffer. Double the buffer size to 128 kB.

(Note: built-in default in ubox logd itself is still 16 kB)

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
Link: https://github.com/openwrt/openwrt/pull/16723
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

ath79: Upstream b53 DSA driver for Ubiquiti EdgeSwitch 8XP

The swconfig-based b53 driver for the BCM53128 switch stopped working
after commits b2cfed48f6 (Revert "swconfig: fix Broadcom b53 support")
and e4e410733f (kernel: export switch_generic_set_link() symbol). This
rendered the 8 LAN ports of the EdgeSwitch 8XP non-functional, so the
image compilation for the device was disabled (5a1d7d8c1b).

This commit adds the kmod-dsa-b53-mdio and kmod-dsa-b53 packages
with the upstream B53 DSA driver, replacing the swconfig-based
kmod and kmod-switch-bcm53xx-mdio downstream ones that are not used by
any other device.

The 8 LAN ports of the EdgeSwitch 8XP are usable again. The 02_network
init script has been updated with the new DSA interfaces lan1 .. lan8.

Image building has been reenabled for the device, adding the usual DSA
incompatibility notice.

Tested on a Ubiquiti EdgeSwitch 8XP.

Signed-off-by: Roger Pueyo Centelles <roger.pueyo@guifi.net>
Link: https://github.com/openwrt/openwrt/pull/11680
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

mediatek: filogic: convert Acer Predator W6 to use NVMEM framework

Read WiFi calibration data via NVMEM framework. The MAC addresses are
stored inside a file on a filesystem and hence still have to be
extracted in userspace.

WiFI EEPROM extraction has already accidentally been partially removed
by commit 3e6de5d77a ("mediatek: use NVMEM framework on all Adtran
devices").

Fixes: 3e6de5d77a ("mediatek: use NVMEM framework on all Adtran devices")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>

firewall: update to Git HEAD (2024-10-18)

ffba75c9cd8f iptables: free xtables_match if found in need_protomatch
bf1d5fdf6234 iptables: fix regression with unintended free in need_protomatch
1aef9791a21e defaults.c: fix ipv6 flow offloading

Signed-off-by: Felix Fietkau <nbd@nbd.name>

lantiq: refresh patches

CI says that they need to be refreshed, so do so.

Link: https://github.com/openwrt/openwrt/pull/16722
Signed-off-by: Robert Marko <robimarko@gmail.com>

realtek: refresh patches

CI says that they need to be refreshed, so do so.

Link: https://github.com/openwrt/openwrt/pull/16722
Signed-off-by: Robert Marko <robimarko@gmail.com>

package: busybox: update download link

There is no file named busybox-1.36.1.tar.bz2 in the root directory of the website
The actual download link is “https://sources.buildroot.net/busybox/busybox-1.36.1.tar.bz2”

Signed-off-by: Jack Sun <sunjiazheng321521@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16719
Signed-off-by: Robert Marko <robimarko@gmail.com>

base-files: Increase failsafe trigger wait time from 2 to 4 sec

Increase the failsafe waiting timeout period from 2 seconds
to 4 seconds.

Since commit 29207748b in 2015 we have had a rapid LED blinking
indication for the failsafe triggering period. But the really short
timeout of 2 seconds requires snappy reaction time from the user to
notice the LED blinking and to push button inside the short window.

Relax the timeout to 4 seconds, which more easily allows to notice
the change in LED blinking and push a button.

Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
Link: https://github.com/openwrt/openwrt/pull/11852
Signed-off-by: Robert Marko <robimarko@gmail.com>

ath11k-firmware: update IPQ8074 and QCN9074 to 2.9.0.1-02146

After a long time QCA has pushed an updated release of 2.9.0.1 firmware
for IPQ8074 and QCN9074, so lets update to 2.9.0.1-02146.

Sadly, still nothing new for IPQ6018.

QCA has also moved the repository where they will be posting firmware to
their CodeLinaro instance, so we move to using that and it allows us to
remove the manual download of QCN9074 board-2.bin.

Link: https://github.com/openwrt/openwrt/pull/16720
Signed-off-by: Robert Marko <robimarko@gmail.com>

generic: import and move accepted RealTek PHY patches

Import commits from upstream Linux and move/rename existing patches
which have been acceped.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

mpc85xx: convert Watchguard T10 mac-assignment to NVMEM

MAC-addresses are stored with colons, thus they can now be referenced
using NVMEM.

Signed-off-by: David Bauer <mail@david-bauer.net>
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16310
Signed-off-by: Robert Marko <robimarko@gmail.com>

mpc85xx: p1010: fix NAND devices

Upstream commit 9ba0cae3cac07c21c583f9ff194f74043f90d29c made FSL_IFC
visible and selectable, which means that in order for MTD_NAND_FSL_IFC
to work, it needs these two extra CONFIG options.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16717
Signed-off-by: Robert Marko <robimarko@gmail.com>

uboot-ath79: update to U-Boot v2024.10

Update uboot-ath79 package to U-Boot v2024.10.

Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16703
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

mac80211: realtek: rtw89: add RTL8922AE support

Add the Realtek RTL8922AE support to rtw89 package.
The 8922A is a 802.11be chip that can support 2/5/6GHz 160MHz.

Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>
Link: https://github.com/openwrt/openwrt/pull/16700
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

linux-firmware: update Realtek RTW89 firmware

Add RealTek RTL8922AE firmware package. The rtw89 driver uses
rtw8852b_fw-1.bin starting from kernel 6.4[1], remove the old
one. Also note that in the kernel 6.12, rtw8852c and rtw8922a
should use fw-1.bin [2][3].

[1] https://github.com/torvalds/linux/commit/ffde7f3476
[2] https://github.com/torvalds/linux/commit/45ae0e8cf8
[3] https://github.com/torvalds/linux/commit/6ca6b918f2

Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>
Suggested-by: Hauke Mehrtens <hauke@hauke-m.de>
Link: https://github.com/openwrt/openwrt/pull/16700
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

prereq-build: limit python distutils check to <v3.12

v3.12 won't have it anymore.

Signed-off-by: Andre Heider <a.heider@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16699
Signed-off-by: Robert Marko <robimarko@gmail.com>

prereq-build: add Python 3.12 support

This reverts commit 09b7e1955c92e0b05197174b7bd4aef2c4073a35.

Signed-off-by: Paul Spooren <mail@aparcar.org>
Link: https://github.com/openwrt/openwrt/pull/16699
Signed-off-by: Robert Marko <robimarko@gmail.com>

opkg: update to Git HEAD (2024-10-16)

38eccbb1fd69 libopkg: allow install package from https and ftps

Link: https://github.com/openwrt/openwrt/pull/16715
Signed-off-by: Robert Marko <robimarko@gmail.com>

ramips: disable CONFIG_LIST_HARDENED on MT7620

In #16396, crashes were reported on MT7620, which were introduced by enabling
CONFIG_LIST_HARDENED. The cause seems to be random memory corruption somewhere
in the kernel. Unfortunately the crash traces do not point to the real cause
of the crash in any way.
Since MT7620 is really ancient hardware that likely only has few users left,
I don't expect that anybody will invest a significant amount of time to track
down the real cause. Because of that, let's disable CONFIG_LIST_HARDENED on
this target only, and leave it enabled on all other platforms.

Fixes: #16396
Signed-off-by: Felix Fietkau <nbd@nbd.name>

uboot-mediatek: add noncached_set_region prototype to fix build

Building uboot-mediatek fails with GCC-14, uboot v2024.10 and
CONFIG_SYS_NONCACHED_MEMORY defined with error:
cmd/cache.c: In function 'do_dcache':
cmd/cache.c:57:25: error: implicit declaration of function
'noncached_set_region' [-Wimplicit-function-declaration]

This is caused by upstream commit 7d6cee2cd0e2e2507aca1e3a6fe0e2cb241a116e
("cmd: cache: Remove weak functions") as this removes weak functions in
favor of arch-specific definitions.

This patch adds the function prototype for `noncached_set_region` to
arch-specific header for ARM. It also adds an include in cmd/cache.c to
make the function available there.

Fixes: #16697
Fixes: f8c22c9bff ("uboot-mediatek: update to U-Boot 2024.10")
Signed-off-by: Jonas Jelonek <jelonek.jonas@gmail.com>
[@dangowrt refreshed patch]
Signed-off-by: Daniel Golle <daniel@makrotopia.org>

mediatek: mt7622: add netdev rename preinit script

Add preinit script to rename network interfaces according to the
label provided in Device Tree.

Fixes: b6d8297510 ("mediatek: mt7622: add Adtran SmartRG SDG-841t6 device")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>

mediatek: filogic: reorder alphabetically

Reorder scripts to keep alphabetical order.

Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>

mediatek: filogic: fix network config for mt7986b-rfb

Fix the network configuration according to the device tree.

Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>

generic: fix kernel version and ordering of Aquantia backports

Two patches declared as accepted in v6.13 were already accepted for
v6.12. Fix filenames and order of patches applied.

Reported-by: Mieczyslaw Nalewaj <namiltd@yahoo.com>
Fixes: 8cc049cec2 ("generic: phy: aquantia: move accepted patches to backport-6.6")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>

mediatek: use NVMEM framework on all Adtran devices

Use NVMEM framework to assign MAC addresses and WiFi EEPROM data on
all Adtran SmartRG devices.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

mediatek: mt7622: add Adtran SmartRG SDG-841t6 device

Specification
 SoC:   MediaTek MT7622A (2x Cortex-A53)
 RAM:   NANYA NT5CC256M16ER-EK (512 MiB)
 MMC:   Kioxia THGBMNG5D1LBAIT (4 GiB)
 ETH0:  Intel/MaxLinear GPY211 (2500M/1000M/100M/10M)
 ETH1:  Lantiq/Intel/MaxLinear PEF7071/GPY111 (1000M/100M/10M)
 WLAN0: MediaTek MT7915E + MT7975D 2T2R 802.11bgn/ax (2.4G)
                                   2T2R 802.11an/ac/ax (5G L)
 WLAN1: MediaTek MT7915E + MT7975A 4T4R 802.11an/ac/ax (5G H)
 BT:    BT5 provided by MT7915E+MT7975D (missing firmware)

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

mediatek: add driver for Adtran MFGINFO NVMEM layout

This driver allows retrieving MAC addresses and other factory information
via the NVMEM framework on Adtran devices.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

mediatek: mt7622: enable HWMON in kernel config

Building HWMON as module prevents some other drivers from getting built
into the kernel, e.g. mxl-gpy. Always enable HWMON and also set
THERMAL_HWMON.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

generic: net: phy: intel-way: add support for PHY LEDs

Add driver support for PHY LEDs for MaxLinear/Intel/Lantiq XWAY
Ethernet PHYs.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

generic: net: phy: mxl-gpy: add support for PHY LEDs

Add driver support for PHY LEDs for MaxLinear/Intel GPY Ethernet PHYs.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

generic: phy: aquantia: correctly describe LED polarity

Backport newly introduced support for 'active-high' property and use
it to correctly implement polarity assignment for Aquantia PHY LEDs.
Previously the 'active-low' property was used to switch a LED PIN to
active-high ("drive VDD" in Aquantia-speak) mode.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

generic: phy: aquantia: move accepted patches to backport-6.6

Move patches accepted upstream from pending-6.6 to backport-6.6.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

generic: net: phy: be more conservative about MAC interface selection

Only add SerDes interface modes at lower speeds to host_interfaces.

Fixes: 5281033831 ("generic: net: phy: use all SerDes MAC interface modes")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>

qoriq: add squashfs to FEATURES

The snapshot builds for qoriq only contain ext4 images. Add squashfs to
FEATURES to also have the buildbots build squashfs images.

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>

kernel: fix mtk_eth_soc memory corruption issue on fq dma init

Signed-off-by: Felix Fietkau <nbd@nbd.name>

qoriq: add missing kernel config symbol

Dropping the source-only flag from qoriq and adding it to the buildbots
exposed a missing kernel config symbol: CONFIG_DRM_OFDRM. Addd it as
disabled to the qoriq kernel config, as the only device currently
supported in the qoriq target does not have any display hardware.

Reported-by: Mieczyslaw Nalewaj <namiltd@yahoo.com>
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>

tools/ninja: update to 1.12.1

Update to the most recent stable release,
including a new job scheduler and bug fixes.

Backport a patch to assist in reworking jobserver support.

The original patch for jobserver support requires reworking
due to the function CanRunMore() and its usage
being significantly changed, including the return
becoming an integer instead of a boolean.

AcquireToken() must now be used in CanRunMore()
in order to quantify how much the job searching loop
is able to run at any time, and in order to do so
CanRunMore() cannot be a constant function anymore.

Added:
 - 010-backport-gtest.patch

Manually Adjust:
 - 100-make_jobserver_support.patch

Signed-off-by: Michael Pratt <mcpratt@pm.me>
Link: https://github.com/openwrt/openwrt/pull/16693
Signed-off-by: Robert Marko <robimarko@gmail.com>

tools/ninja: use existing ninja to build ninja

Add methods to skip the building and execution of Ninja by python
in order to allow Make to execute Ninja after the configure script.

This allows the user to build Ninja only once
if they already have an older version of Ninja built.

It also allows the user to test the jobserver functionality
by having Ninja built twice (clean then compile).

Signed-off-by: Michael Pratt <mcpratt@pm.me>
Link: https://github.com/openwrt/openwrt/pull/16693
Signed-off-by: Robert Marko <robimarko@gmail.com>

tools/ninja: cleanup Makefile

Use default CXXFLAGS and LDFLAGS variables.

Directly add --verbose flag without a variable.

Split up lines for script arguments.

Delete unnecessary lines.

Signed-off-by: Michael Pratt <mcpratt@pm.me>
Link: https://github.com/openwrt/openwrt/pull/16693
Signed-off-by: Robert Marko <robimarko@gmail.com>

mvebu: cortex-a9: add upgrade methode to nas1dual

The blamed commit adds a upgrade recipe for nas1dual to specify the
firmware partition name, but does not actually include the recipe that
will be called.

Since it previously relied on the default one, add that one.

Fixes: d21720fa90c8 ("mvebu: fix default partition name")
Link: https://github.com/openwrt/openwrt/pull/16704
Signed-off-by: Robert Marko <robimarko@gmail.com>

xdp-tools: bump to 1.4.3

Release notes:
https://github.com/xdp-project/xdp-tools/releases/tag/v1.4.3

Remove upstreamed patch
023-libxdp-fix-compilation-on-multiarch-systems.patch.

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>

libbpf: bump to 1.4.6

Release notes:
https://github.com/libbpf/libbpf/releases/tag/v1.4.6

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>

qoriq: drop source-only flag

While the qoriq target currently supports only one device, the
WatchGuard Firebox M300, the feedback from people using it is that it's
a great bang for the buck. Additionally, people suggest having no
snapshots available hurts adoption.

Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>

qoriq: use nvmem for MAC addresses

Userspace handling is deprecated.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16308
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>

rtl83xx: fix typo

Removes an unwanted special character in a debug-message.

Signed-off-by: Peter Körner <peter@mazdermind.de>

mvebu: fix sysupgrade failure

Linksys devices do no pass $cmd at all.

Fixes: #16148
Fixes: 715634e6d144 ("base-files: upgrade: nand: use "cmd" argument for extracting command")
Signed-off-by: Boris Krasnovskiy <borkra@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16690
Signed-off-by: Robert Marko <robimarko@gmail.com>

mvebu: fix default partition name

The firmware partition name is specifc to ipTIME NAS1dual and should not be
set globally.

Fixes: #16148
Fixes: 6ff970bb51f5 ("mvebu: add support for ipTIME NAS1dual")
Signed-off-by: Boris Krasnovskiy <borkra@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16690
Signed-off-by: Robert Marko <robimarko@gmail.com>

mac80211: fix letting drivers iterate over keys

Skip non-uploaded keys to avoid corner cases

Signed-off-by: Felix Fietkau <nbd@nbd.name>

tools/cmake: update to 3.30.5

Release notes: https://cmake.org/cmake/help/latest/release/3.30.html

Signed-off-by: Jack Sun <sunjiazheng321521@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16657
Signed-off-by: Robert Marko <robimarko@gmail.com>

uboot-mediatek: update to U-Boot 2024.10

Update to latest U-Boot release.
Patches refreshed and fixed when needed.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

ramips: use nvmem to get mikrotik mac

73a9f9f857947bc63fef1aefb7e2e4b906fb9d2b added support. Let's use it.
Calibration data is not yet supported.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16256
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

mediatek: ASUS TUF-AX6000: Add Green & Red LEDs

Beside the already existing control of the
Blue cover LED, this will also add the
Green and the Red cover LEDS available in the
Asus TUF-AX6000.

Signed-off-by: Magnus Sandin <magnus.sandin@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16632
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

toolchain/binutils: remove old versions

Reduce supported versions to 3.
Since 2.40 is supported by previous release, this commit
keeps it.

Signed-off-by: Yanase Yuki <dev@zpc.st>
Link: https://github.com/openwrt/openwrt/pull/16597
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

linux: generic: dsa: mv88e6xx: Drop obsolete patch

As described by Tobias Waldekranz this patch is obsolete
and the issue is already fixed in the upstream kernels we
are using.

Link: http://lists.openwrt.org/pipermail/openwrt-devel/2024-October/043238.html
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

linux: generic: net: dsa: mv88e6xxx LED support

This solution that is needed for some routers to provide proper
LED activity when controlled directly by the MV88E6xxx switch,
has just been merged in upstream Linux.

Make this patch 901 as other backports from earlier kernels
are in patch 896 and this is the first free number after
that.

Patch offsets in pending patches are augmented as part of
the refresh.

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

ramips: remove useless cs gpio code for rt2800 SPI driver

The SPI driver framework can handle it automatically in
spi.c:spi_setup():spi_set_cs():gpiod_set_value_cansleep().

Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16636
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

scripts/download.pl: use perl builtins instead of system()

Perl natively supports renaming files and create directories.
Do it without calling system().

Signed-off-by: Matteo Croce <teknoraver@meta.com>
Link: https://github.com/openwrt/openwrt/pull/16542
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

wireless-regdb: Update to version 2024.10.07

b66b9a1 wireless-regdb: update regulatory database based on preceding changes
5097b4a wireless-regdb: Update regulatory info for Tanzania (TZ) for 2024
29633a6 wireless-regdb: Update regulatory info for Pakistan (PK) for 2024
b44edb2 wireless-regdb: Update regulatory info for Serbia (RS) for 2024
dbfae47 Revert "wireless-regdb: Update regulatory info for Serbia (SR) for 2024"
8e3d27c wireless-regdb: Correct regulatory rules of 6GHz frequency for Türkiye (TR)
8760bc3 wireless-regdb: Update regulatory info for Honduras (HN) for 2023
3ba2c53 wireless-regdb: Update regulatory info for Israel (IL) for 2021
83c175c wireless-regdb: Update regulatory info for Kuwait (KW) for 2022
388c80c wireless-regdb: Update regulatory info for Serbia (SR) for 2024
bf55ed4 wireless-regdb: Add .b4-config
3afe172 wireless-regdb: Update .gitignore
3b34761 wireless-regdb: Correct regulatory rules for China (CN)
003c282 wireless-regdb: Update regulatory info for Philippines (PH) on 6GHz
21fcb86 wireless-regdb: Update regulatory info for Guatemala (GT) for 2020
158f105 wireless-regdb: Update regulatory info for Bahrain (BH) for 2024
218d146 wireless-regdb: Add regulatory info for Namibia (NA) for 2023
aad0c26 wireless-regdb: Update regulatory info for Togo (TG) for 2022
983f551 wireless-regdb: Update regulatory info for El Salvador (SV) on 6GHz
58575b4 wireless-regdb: Update regulatory info for Peru (PE) on 6GHz
bad3985 wireless-regdb: Update regulatory info for New Zealand (NZ) for 2022
c7d1083 wireless-regdb: Update regulatory info for Qatar (QA) on 6GHz

Signed-off-by: Itay Shoshani <itai.sho@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16678
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

gdb: Remove MIPS NSIG patch

This bug was fixed in the libc:
https://sourceware.org/bugzilla/show_bug.cgi?id=14523
https://git.busybox.net/uClibc/commit/?id=2da958760f798224065508431787e3a83b7fe2ae
https://git.musl-libc.org/cgit/musl/commit/?id=7c440977db9444d7e6b1c3dcb1fdf4ee49ca4158

Link: https://github.com/openwrt/openwrt/pull/16665
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

gdb: Remove upstreamed SIGINT patch

The problem was fixed upstream, see:
https://sourceware.org/bugzilla/show_bug.cgi?id=18945
https://sourceware.org/git/?p=binutils-gdb.git;a=commitdiff;h=4c35c4c6a779c79e456b7a5311f74aafc9026bd5

Link: https://github.com/openwrt/openwrt/pull/16665
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

gdb: Update to version 15.2

See news file for changes:
https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=blob_plain;f=gdb/NEWS;hb=gdb-15.1-release

Link: https://github.com/openwrt/openwrt/pull/16665
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

toolchain: gdb: Update to version 15.2

See news file for changes:
https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=blob_plain;f=gdb/NEWS;hb=gdb-15.1-release

This requires a C++17 compiler compatible compiler.

Link: https://github.com/openwrt/openwrt/pull/16665
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

lantiq: ase: disable adm6996 and swconfig

The ase subtarget supports two devices. ALLNET ALL0333CJ has one
Ethernet port and uses only the built-in MAC and PHY. The second
device is the Netgear DGN1000, which uses an AR8216 switch.

The driver for the adm6996 switch was probably accidentally
enabled instead of the ar8216 in commit that restores ase support
9b321bc60de3 ("lantiq: add Amazon-SE subtarget").

Before removing the ase subtarget, only ar8216 was enabled
c82183639512 ("lantiq: remove unmaintained code").

Reduce uncompressed kernel size by 15.6 kB.

Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Link: https://github.com/openwrt/openwrt/pull/15332
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

lantiq: ase: refresh config

This was done by executing these command:
$ make kernel_oldconfig CONFIG_TARGET=subtarget

Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Link: https://github.com/openwrt/openwrt/pull/15332
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

ppp: clean up makefile

The latest ppp version seems to no longer require these ancient
build fixes.

Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16605
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

ppp: remove discovery phase timeout hack patch

In the original code, the entire time delay of the discovery phase
is only 5+5x2+5x2x2 = 35s. Increasing timeout may be necessary if
discovery phase fails on first attempt. There is a chance to fix
the "Timeout waiting for PADO packets" issue by removing this patch.

Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16605
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

ppp: remove uClibc wtmp hack patch

The uClibc library support was removed since commit:
57fe7d5401e3 ("toolchain: remove uClibc install stuff")

Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16605
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

ppp: update to 2.5.1

ChangeLog:
https://github.com/ppp-project/ppp/blob/ppp-2.5.1/ChangeLog

Suppressed patches:
010-use_target_for_configure.patch [1]
510-pptp_compile_fix.patch [2]
520-u_int_bsd_fix.patch [3]

Upstreamed patches:
330-retain_foreign_default_routes.patch [4]
521-remove_unused_openssl_dep.patch [5]

[1] https://github.com/ppp-project/ppp/commit/e48a9b5de46b6d3a7fda98ac64de56092471ab98
[2] Merged into "500-add-pptp-plugin.patch"
[3] https://github.com/ppp-project/ppp/commit/797cdae57c27ed74a5ec3b4675f31f2c6cfc98ae
[4] https://github.com/ppp-project/ppp/commit/9856f47063c01096fe6649a21be9a021e42a45fe
[5] https://github.com/ppp-project/ppp/commit/59342ab622a96393d25d3a2f0592c8675e771295

Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16605
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

ppp: update to 2.5.0

ChangeLog:
https://github.com/ppp-project/ppp/blob/ppp-2.5.0/ChangeLog

Upstreamed patches:
120-debian_ipv6_updown_option.patch [1]
133-fix_sha1_include.patch [2]
140-pppd-Fix-compilation-with-older-glibc-or-kernel-head.patch [3]
141-Expand-byte-count-statistics-to-64-bits-298.patch [4]
142-pppd-Add-support-for-registering-ppp-interface-via-L.patch [5]
143-pppd-Workaround-for-generating-ppp-unit-id-on-Linux-.patch [6]
144-pppd-Retry-registering-interface-when-on-rtnetlink-E.patch [7]

Suppressed patches:
200-makefile.patch [8]
201-mppe_mppc_1.1.patch [9]
203-opt_flags.patch [10]
300-filter-pcap-includes-lib.patch [11]
511-pptp_cflags.patch [12]
600-Revert-pppd-Use-openssl-for-the-DES-instead-of-the-l.patch [13]
610-pppd_compile_fix.patch [14]

[1] https://github.com/ppp-project/ppp/commit/7f8c1a1f8e486b232340fd9a0a19c5d34f1c5ae0
[2] https://github.com/ppp-project/ppp/commit/ba7f7e053daae846a54a1d08d3d133a5f1266ace
[3] https://github.com/ppp-project/ppp/commit/98ec18f098e5ef68e3a8cc6954fcaf5a7fb8b7be
[4] https://github.com/ppp-project/ppp/commit/81ad945630120cc1c27c8bb00503be42b76ff202
[5] https://github.com/ppp-project/ppp/commit/4a54e34cf5629f9fed61f0b7d69ee3ba4d874bc6
[6] https://github.com/ppp-project/ppp/commit/44609bfc974bdafc0316d069aabf5e2903efa805
[7] https://github.com/ppp-project/ppp/commit/089687fbcc6524809ae9f4b2f8145fe3c2a91147
[8] enable_eaptls=no, with_pcap=no, HAVE_CRYPT_H=1 in configure
[9] enable_microsoft_extensions=yes, MPPC support is removed.
[10] fPIC ignored so far
[11] done by autotools
[12] in main patch for pptp plugin
[13] with_openssl=no, already in upstream ppp-des.c
[14] with_static_pcap=yes from patch 310

Signed-off-by: Sergey Ivanov <icegood1980@gmail.com>
* Fix package hash.
* Fix multilink variant build.
* Fix some compile errors.
* Some code format fixes.
* Refactor commit message.
* Rebase git and fix conflicts.

Co-authored-by: Shiji Yang <yangshiji66@qq.com>
Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16605
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

openwrt-keyring: add apk public key for signing of package indexes

apk/openwrt-snapshots.pem contains Elliptic Curve based public key which
is going to be used for signing of apk's package.adb package indexes
after the builds using `apk adbsign --sign-key <key> packages.adb`
command on the buildbot.

References: https://github.com/openwrt/buildbot/pull/46
Link: https://github.com/openwrt/openwrt/pull/16539
Signed-off-by: Petr Štetiar <ynezz@true.cz>

realtek: scripts: fix error in belkin-header

For some reason the new belkin-header.py script works without issues
in a local Fedora build environment. In the OpenWrt build pipeline it
produces the following errors:

Traceback (most recent call last):
  File "/builder/shared-workdir/build/scripts/belkin-header.py", line 92, in <module>
    head = create_header(buf, args.belkin_header, args.belkin_model)
  File "/builder/shared-workdir/build/scripts/belkin-header.py", line 68, in create_header
    head[28:29] = VERSION1.to_bytes(1)
TypeError: to_bytes() missing required argument 'byteorder' (pos 2)

This may be related due to different python version. Fix this by
handing over the needed parameters

Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/16667
Signed-off-by: Robert Marko <robimarko@gmail.com>

mt76: update to Git HEAD (2024-10-11.1)

ecca0e77b4bc wifi: mt76: mt7915: fix omac index assignment after hardware reset

Signed-off-by: Felix Fietkau <nbd@nbd.name>

kernel: bump 6.6 to 6.6.56

Changelog: https://cdn.kernel.org/pub/linux/kernel/v6.x/ChangeLog-6.6.56

No patches needed a rebase

Build system: x86/64
Build-tested: x86/64/AMD Cezanne, flogic/xiaomi_redmi-router-ax6000-ubootmod, ramips/tplink_archer-a6-v3
Run-tested: x86/64/AMD Cezanne, flogic/xiaomi_redmi-router-ax6000-ubootmod, ramips/tplink_archer-a6-v3

Signed-off-by: John Audia <therealgraysky@proton.me>
Link: https://github.com/openwrt/openwrt/pull/16655
Signed-off-by: Nick Hainke <vincent@systemli.org>

kernel: bump 6.6 to 6.6.55

Changelog: https://cdn.kernel.org/pub/linux/kernel/v6.x/ChangeLog-6.6.55

Added the following default ksym to target/linux/generic/config-6.6:
CONFIG_PROC_MEM_ALWAYS_FORCE=y
# CONFIG_PROC_MEM_FORCE_PTRACE is not set
# CONFIG_PROC_MEM_NO_FORCE is not set

Removed upstreamed:
generic/backport-6.6/780-23-v6.12-r8169-Fix-spelling-mistake-tx_underun-tx_underrun.patch[1]
generic/backport-6.6/780-25-v6.12-r8169-add-tally-counter-fields-added-with-RTL8125.patch[2]
generic/pending-6.6/684-gso-fix-gso-fraglist-segmentation-after-pull-from-fr.patch[3]
lantiq/patches-6.6/0025-v6.12-net-ethernet-lantiq_etop-fix-memory-disclosure.patch[4]

Manually rebased:
bcm27xx/patches-6.6/950-0086-Main-bcm2708-bcm2709-linux-port.patch
bcm27xx/patches-6.6/950-0998-i2c-designware-Add-support-for-bus-clear-feature.patch

All other patches automatically rebased.

1. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v6.6.56&id=f02fcb7283b1c25f7e3ae07d7a2c830e06eb1a62
2. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v6.6.56&id=1c723d785adb711496bc64c24240f952f4faaabf
3. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v6.6.56&id=af3122f5fdc0d00581d6e598a668df6bf54c9daa
4. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v6.6.56&id=e66e38d07b31e177ca430758ed97fbc79f27d966

Build system: x86/64
Build-tested: x86/64/AMD Cezanne, flogic/xiaomi_redmi-router-ax6000-ubootmod, ramips/tplink_archer-a6-v3
Run-tested: x86/64/AMD Cezanne, flogic/xiaomi_redmi-router-ax6000-ubootmod, ramips/tplink_archer-a6-v3

Signed-off-by: John Audia <therealgraysky@proton.me>
Link: https://github.com/openwrt/openwrt/pull/16655
Signed-off-by: Nick Hainke <vincent@systemli.org>

mt76: update to Git HEAD (2024-10-11)

513bd7683746 Revert "wifi: mt76: mt7915: disable the second PCIe link for MT7915"
703c6b78c133 wifi: mt76: mt7915: firmware restart on devices with a second pcie link

Signed-off-by: Felix Fietkau <nbd@nbd.name>

generic: backport BLOCK OF support patch

Backport BLOCK OF support patch merged upstream and refresh pending
BLOCK patches.

This is a new way to declare partition table for BLOCK device (eMMC
currently supported) with the use of DTS.

Current pending patch are adapted to not cause regression with current
downstream implementation of a similar functionality.

Also enable the new OF_PARTITION config by default.

Link: https://github.com/openwrt/openwrt/pull/16663
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>

ipq40xx: switch to zImage kernel for SKSpruce WIA3300-20

The bootloader can support zImage linux kernel which can decrease
the firmware image size.

Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16662
Signed-off-by: Robert Marko <robimarko@gmail.com>

netifd: update to Git HEAD (2024-10-06)

3c6265fdbd02 wireless: reload wireless device if any vif network bridge changes ifindex

Signed-off-by: Felix Fietkau <nbd@nbd.name>

uboot-rockchip: Update to 2024.10

Removed upstreamed patches.

Signed-off-by: Tianling Shen <cnsztl@immortalwrt.org>
Link: https://github.com/openwrt/openwrt/pull/16647
Signed-off-by: Nick Hainke <vincent@systemli.org>

scripts/feeds: force kernel package scan after a target installation

When a target is installed from a feed, the linux kernel package is scanned
before the installation of this target.
If some kernel module packages are defined in this feeds at the target
level, there were not parsed during the scan of linux kernel package, as
the target didn't exist yet. So these kernel module packages don't exist.

Once the target is installed, clean the linux kernel packageinfo to force
the scan of the linux kernel package next time this script (or the make
command) is called.

Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://patchwork.ozlabs.org/project/openwrt/patch/mailman.117548.1727195440.1280.openwrt-devel@lists.openwrt.org/
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

realtek: refresh patches

CI is saying that patches need to be refreshed, so refresh them.

Link: https://github.com/openwrt/openwrt/pull/16653
Signed-off-by: Robert Marko <robimarko@gmail.com>

mt76: update to Git HEAD (2024-10-10)

86b41a32a5ac wifi: mac80211: don't use rate mask for offchannel TX either
82a452919f52 wifi: mac80211: handle ieee80211_radar_detected() for MLO
02c07fe539b4 wifi: mt76: mt7996: fix rx filter setting for bfee functionality
a902bfa6db2f wifi: mt76: mt7915: disable the second PCIe link for MT7915
a7d48cd3ed34 wifi: mt76: mt7915: reduce the number of command retries
0d8474602a9b wifi: mt76: mt7915: decrease timeout for commonly issued MCU commands
5f6051569639 wifi: mt76: mt7915: add module param to select 5 GHz or 6 GHz on MT7916
f5bccb122305 wifi: mt76: mt7996: fix invalid interface combinations
6a1199463fc8 wifi: mt76: mt7915: Fix an error handling path in mt7915_add_interface()

Signed-off-by: Felix Fietkau <nbd@nbd.name>

mpc85xx: handle MSM460 wifi mac in dts

Userspace handling is deprecated.

Signed-off-by: Rosen Penev <rosenp@gmail.com>

kernel: netdevices: add RmNet MAP support

RmNet driver provides a transport agnostic MAP (multiplexing and
aggregation protocol) support in embedded module. Module provides
virtual network devices which can be attached to any IP-mode
physical device.

This is commonly used on Qualcomm based modems for data aggregation.

Signed-off-by: Robert Marko <robert.marko@sartura.hr>

kernel: backport upstream commit that allows leaving out .ndo_get_stats64

Fixes mac80211 backport device statistics.

Fixes: #16642
Signed-off-by: Felix Fietkau <nbd@nbd.name>

generic: net: phy: use all SerDes MAC interface modes

Instead of forcing 2.5G PHYs into rate-adapter mode which results higher
energy consumption, lack of support for half-duplex modes and typically
worse performance when linked at speeds less than 2.5G, use SGMII mode
which allows the MAC to follow the PHY speed.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>

generic: net: phy: realtek: various improvements

Follow the advise of Russell King allows to greatly improve the driver
for RealTek's 1G and 2.5G Ethernet PHYs. The results are full/half
duplex as well as Gbit master/slave property being read from PHY
Specific Status Register (PHYSR), and fixes regarding link-partner
advertisement.

Fixes: #14504
Signed-off-by: Daniel Golle <daniel@makrotopia.org>

mediatek: update MediaTek PHY patchset

The previous iteration of MediaTek's PHY patches caused various weird bugs.
Drop culprit patch 733-10-net-phy-mediatek-Extend-1G-TX-RX-link-pulse-time.patch
and use the most recent iteration of the patchset which has been posted to the
netdev mailing list.

Link: https://patchwork.kernel.org/project/netdevbpf/list/?series=895513&state=*
Fixes: #16448
Signed-off-by: Daniel Golle <daniel@makrotopia.org>

toplevel.mk: add target feeds in 'make distclean'

When 'make distclean' is called, remove the target feeds.
Otherwise the directory target/linux/feeds/ (with broken symlinks) still
exists after a distclean.

Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://patchwork.ozlabs.org/project/openwrt/patch/mailman.115093.1726741692.1280.openwrt-devel@lists.openwrt.org/
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

gpio-button-hotplug: skip disabled buttons

Ignore buttons which are disabled in the devicetree.

Reviewed-by: Rosen Penev <rosenp@gmail.com>
Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://patchwork.ozlabs.org/project/openwrt/patch/mailman.117522.1727194512.1280.openwrt-devel@lists.openwrt.org/
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

ubnt-ledbar: fix compilation error

Wrong variable name from copy/paste.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16640
Signed-off-by: Robert Marko <robimarko@gmail.com>

valgrind: Remove unsupported configure options.

These options are not recognized by the valgrind configure script,
remove them.

Link: https://github.com/openwrt/openwrt/pull/16633
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>

valgrind: Fix build on ARMv7

The valgrind configure script checks if host_cpu is set to armv7 or arm.
By default --host is set to arm-openwrt-linux and the host_cpu variable
is set to arm. Then the valgrind build tries to compile valgrind for
armv6 and fails. Set it explicitly to armv7 to compile valgrind with
armv7 support.

Fixes: 1a55d90320c1 ("valgrind: Update to version 3.23")
Link: https://github.com/openwrt/openwrt/pull/16633
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>