tools/quilt: update to 0.69

Refresh patches.

Signed-off-by: Rosen Penev <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18858
Signed-off-by: Robert Marko <[email protected]>

boot: uboot-armsr: update to U-Boot 2025.04

This U-Boot version is used to generate a 'reference'
U-Boot binary for QEMU's 'virt' machine on both
armv7 and armv8.

It has not been updated since EFI support was merged
into the then-armvirt target, so we should bring it up
to the latest version.

The 'mkeficapsule' tool is disabled due to a host-side
dependency on GnuTLS.

Signed-off-by: Mathew McBride <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18862
Signed-off-by: Robert Marko <[email protected]>

ramips: fix wrg-header image recipe

Before generating the factory image, check if the input file
exists. Fix the build error when sysupgrade image is too big:

[mkwrgimg] *** error: stat failed on /builder/shared-workdir/build/build_dir/target-mipsel_24kc_musl/linux-ramips_rt288x/tmp/openwrt-ramips-rt288x-airlink101_ar670w-squashfs-factory.bin, No such file or directory

Signed-off-by: Shiji Yang <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18836
Signed-off-by: Robert Marko <[email protected]>

d1: add 6.12 kernel as testing

Allow selecting 6.12 as testing kernel on d1.

Runtime-tested:
 - Nezha D1

Signed-off-by: Zoltan HERPAI <[email protected]>

d1: 6.12: refresh kernel config

Refresh kernel config.

Signed-off-by: Zoltan HERPAI <[email protected]>

d1: 6.12: remove upstreamed patches and refresh remaining ones

Remove patches that have been upstreamed or dropped, and refresh
the remaining ones as required.

Signed-off-by: Zoltan HERPAI <[email protected]>

kernel/d1: Restore kernel files for v6.6

This is an automatically generated commit which aids following Kernel patch
history, as git will see the move and copy as a rename thus defeating the
purpose.

For the original discussion see:
https://lists.openwrt.org/pipermail/openwrt-devel/2023-October/041673.html

Signed-off-by: Zoltan HERPAI <[email protected]>

kernel/d1: Create kernel files for v6.12 (from v6.6)

This is an automatically generated commit.

When doing `git bisect`, consider `git bisect --skip`.

Signed-off-by: Zoltan HERPAI <[email protected]>

generic: backport upstream v6.16 Realtek PHY patches

83d962316128 net: phy: realtek: add RTL8127-internal PHY
708686132ba0 net: phy: realtek: Add support for PHY LEDs on RTL8211E
be1cc96ddf82 net: phy: realtek: use __set_bit() in rtl8211f_led_hw_control_get()
8c4d0172657c net: phy: realtek: Group RTL82* macro definitions
12d40df259e3 net: phy: realtek: add RTL8211F register defines
7c6fa3ffd265 net: phy: realtek: Clean up RTL821x ExtPage access
f3b265358b91 net: phy: realtek: remove unsed RTL821x_PHYSR* macros
7840e4d6f48a net: phy: realtek: Add support for WOL magic packet on RTL8211F

Signed-off-by: Álvaro Fernández Rojas <[email protected]>

generic: backport upstream v6.16 r8169 patches

f24f7b2f3af9 r8169: add support for RTL8127A
4f51e7d370a0 r8169: merge chip versions 52 and 53 (RTL8117)
f372ef6ed5a6 r8169: merge chip versions 64 and 65 (RTL8125D)
4dec0702b862 r8169: merge chip versions 70 and 71 (RTL8126A)
b7ed5d5a78fc r8169: use pci_prepare_to_sleep in rtl_shutdown
fe733618b27a r8169: add RTL_GIGA_MAC_VER_LAST to facilitate adding support for new chip versions
2b065c098c37 r8169: refactor chip version detection
0c49baf099ba r8169: add helper rtl8125_phy_param
8c40d99e5f43 r8169: add helper rtl_csi_mod for accessing extended config space

Signed-off-by: Álvaro Fernández Rojas <[email protected]>

ath79: remove platform_device_id from drivers

This was needed while ar71xx was in tree. This is no longer the case.
Not only that, these have already been converted to OF.

Signed-off-by: Rosen Penev <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18860
Signed-off-by: Robert Marko <[email protected]>

mxs: add 6.12 kernel as testing

Allow selecting 6.12 as testing kernel on mxs.

Runtime-tested:
 - Olinuxino Micro

Signed-off-by: Zoltan HERPAI <[email protected]>

mxs: refresh config for 6.12

Refresh kernel config for 6.12.

Signed-off-by: Zoltan HERPAI <[email protected]>

kernel/mxs: Restore kernel files for v6.6

This is an automatically generated commit which aids following Kernel patch
history, as git will see the move and copy as a rename thus defeating the
purpose.

For the original discussion see:
https://lists.openwrt.org/pipermail/openwrt-devel/2023-October/041673.html

Signed-off-by: Zoltan HERPAI <[email protected]>

kernel/mxs: Create kernel files for v6.12 (from v6.6)

This is an automatically generated commit.

When doing `git bisect`, consider `git bisect --skip`.

Signed-off-by: Zoltan HERPAI <[email protected]>

generic: 6.12: add missing symbols

These were found while working on mxs and sunxi 6.12.

Signed-off-by: Zoltan HERPAI <[email protected]>

realtek: harmonize power functions for RTL8214FC

There are currently two power functions for the RTL8214FC in the driver

- rtl8380_rtl8214fc_on_off(): powers on/off both ports
- rtl8214fc_power_set(): powers on/off fibre or copper individually

While it seems obvious to merge these, one of them uses extended register
29 and the other register 30 to achieve the update. From looking at the
GPL source drops both methods exist with exactly the same operations. So
where to go from here? The descision was made based on:

- register 30 is only used during patching, 29 during normal operation
- Avoid raw page if possible to use function for RTL839x

So the attached patch was created. The result in the power down bits
was compared for the two old and the new implementation. Result is
always the same.

Signed-off-by: Markus Stockhausen <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18707
Signed-off-by: Robert Marko <[email protected]>

realtek: Identify RTL9311 properly

RTL9311 switches (like the Linksys LGS352C) will boot but have not set
the right family id in the soc_info structure. Fix that so drivers do
their work right when checking for that.

Signed-off-by: Markus Stockhausen <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18871
Signed-off-by: Robert Marko <[email protected]>

bcm27xx: refresh patches for v6.12.29

Commit 0912d3f59547 ("kernel: bump 6.12 to 6.12.29") was merged right after
bcm27xx 6.12 support commits, causing conflicts.

Refresh patches and removed upstreamed patch in v6.12.29:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=linux-6.12.y&id=a5f162727b91e480656da1876247a91f651f76de

Fixes: 0912d3f59547 ("kernel: bump 6.12 to 6.12.29")
Signed-off-by: Álvaro Fernández Rojas <[email protected]>

mold: update to 2.39.1

Update to latest mold linker version 2.39.1.

Link: https://github.com/openwrt/openwrt/pull/18865
Signed-off-by: Robert Marko <[email protected]>

bcm27xx: drop 6.6 support

Drop configs, files and patches for Linux 6.6.

Signed-off-by: Álvaro Fernández Rojas <[email protected]>

bcm27xx: use kernel 6.12 by default

Switch to Linux kernel version 6.12.

Signed-off-by: Álvaro Fernández Rojas <[email protected]>

bcm27xx: switch to upstream SDHOST driver

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 <[email protected]>

realtek: refactor net rx interrupt handler rtl83xx_net_irq()

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 <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18855
Signed-off-by: Robert Marko <[email protected]>

kernel: bump 6.12 to 6.12.29

Changelog: https://cdn.kernel.org/pub/linux/kernel/v6.x/ChangeLog-6.12.29

Removed upstreamed:
        generic/pending-6.12/730-net-ethernet-mtk_eth_soc-reset-all-TX-queues-on-DMA-.patch[1]

Added settings:
        x86: CONFIG_MITIGATION_ITS=y[2]

1. https://web.git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=linux-6.12.y&id=aac9d5fa537b0c19f6d3af3c8a124028510bb647
2. https://web.git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=linux-6.12.y&id=68d59e9ba38424389168eddf8791265818f67292

All other patches automatically rebased.

Build system: x86/64
Build-tested: x86/64
Run-tested: x86/64

Signed-off-by: Mieczyslaw Nalewaj <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18837
Signed-off-by: Robert Marko <[email protected]>

toolchain: glibc: Update glibc 2.38 to recent HEAD

3be3728d elf: Ignore LD_LIBRARY_PATH and debug env var for setuid for static
454f24e9 support: Use const char * argument in support_capture_subprogram_self_sgid
4335cd9b support: Add support_record_failure_barrier
31fa0f73 elf: Test case for bug 32976 (CVE-2025-4802)
ca41fe44 x86_64: Fix typo in ifunc-impl-list.c.

Signed-off-by: Konstantin Demin <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18863
Signed-off-by: Robert Marko <[email protected]>

bcm27xx: enable 6.12 kernel testing

bcm27xx target now supports 6.12 kernel as testing.

Tested-by: Zoltan HERPAI <[email protected]>
Signed-off-by: Álvaro Fernández Rojas <[email protected]>

bcm27xx: update config for kernel 6.12

Update bcm27xx target config for kernel 6.12.

Signed-off-by: Álvaro Fernández Rojas <[email protected]>

bcm27xx: add 6.12 patches from RPi repo

These patches were generated from:
https://github.com/raspberrypi/linux/commits/rpi-6.12.y
With the following command:
git format-patch -N v6.12.27..HEAD
(HEAD -> 8d3206ee456a5ecdf9ddbfd8e5e231e4f0cd716e)

Exceptions:
- (def)configs patches
- github workflows patches
- applied & reverted patches
- readme patches
- wireless patches

Signed-off-by: Álvaro Fernández Rojas <[email protected]>

bcm27xx: restore kernel config for v6.6

This is an automatically generated commit which aids following Kernel patch
history, as git will see the move and copy as a rename thus defeating the
purpose.

For the original discussion see:
https://lists.openwrt.org/pipermail/openwrt-devel/2023-October/041673.html

Signed-off-by: Álvaro Fernández Rojas <[email protected]>

bcm27xx: create kernel config for v6.12 (from v6.6)

This is an automatically generated commit.

When doing `git bisect`, consider `git bisect --skip`.

Signed-off-by: Álvaro Fernández Rojas <[email protected]>

mvebu: armada37xx: add pending PXA I2C recovery fix

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 <[email protected]>

mvebu: armada37xx: backport pinctrl fixes

Backport upstream pinctrl fixes, these are required for the follow-up I2C
PXA recovery fixes.

Signed-off-by: Robert Marko <[email protected]>

unetmsg: fix host name in response messages

Avoid using loop iterator variable when generating the closure

Signed-off-by: Felix Fietkau <[email protected]>

unetmsg: add support for sending requests/messages to specific hosts

Makes it possible to implement unetmsg APIs for host control

Signed-off-by: Felix Fietkau <[email protected]>

ramips: mt7621: fix broken hexdump expression in iodata.sh

Use the "1/1" as a iteration count of hexdump expression instead of the
broken "1", to fix the issue that the invalid bootnum will be obtained.

Currently, the hexdump command always outputs "0" when the decimal format
and the iteration count "1" are specified[0]. This is unexpected
behaviour, but the cause is unknown and use this fix as a workaround.

[0]: https://github.com/openwrt/openwrt/issues/18808

Signed-off-by: INAGAKI Hiroshi <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18827
Signed-off-by: Robert Marko <[email protected]>

kernel/i2c: x86: remove kernel 6.12 conditions checking

After removing kernel 6.6 support for x86 targets,
kernel 6.12 conditions no longer need to be checked.

Signed-off-by: Mieczyslaw Nalewaj <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18848
Signed-off-by: Robert Marko <[email protected]>

x86: drop 6.6 support

Drop configs, files and patches for Linux 6.6.

Signed-off-by: Mieczyslaw Nalewaj <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18848
Signed-off-by: Robert Marko <[email protected]>

x86: use kernel 6.12 by default

Switch to Linux kernel version 6.12.

Signed-off-by: Mieczyslaw Nalewaj <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18848
Signed-off-by: Robert Marko <[email protected]>

toolchain: glibc: set min kernel version to 6.6

Minimum kernel version is 6.6 so it's safe to increase minimum kernel version to 6.6 for glibc too.
This change effectively removes last compatibility quirk in [1] which leads to better/faster detection of SVE support in string ifuncs for aarch64.
Furthermore, this change will allow to use "fchmodat2" syscall within fchmodat() wrapper in future glibc versions (starting from 2.39) as noted in [2] and [3].

[1] https://sourceware.org/git/?p=glibc.git;a=blob;f=sysdeps/unix/sysv/linux/aarch64/cpu-features.c;h=4a205a6b35050ffcfac771b4b4ff2d9fb7581f93;hb=refs/heads/release/2.38/master
[2] https://sourceware.org/git/?p=glibc.git;a=blob;f=sysdeps/unix/sysv/linux/kernel-features.h;h=86b2d3ce512588c8823c11076c91116092836402;hb=refs/heads/release/2.41/master
[3] https://sourceware.org/git/?p=glibc.git;a=blob;f=sysdeps/unix/sysv/linux/fchmodat.c;h=dd1fa5db86bde99fe3eb4804b06c5adf11914b94;hb=refs/heads/release/2.41/master

Signed-off-by: Konstantin Demin <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18850
Signed-off-by: Robert Marko <[email protected]>

qualcommax: default to 6.12

Default qualcommax target to 6.12 kernel.

It makes no sense for us to double the effort required for backporting
upstreamed patches or for downstream modifications.

Link: https://github.com/openwrt/openwrt/pull/18840
Signed-off-by: Robert Marko <[email protected]>

treewide: strip trailing whitespace

Strip trailing whitespace in all code:
find . -type f | grep "\.c$" | xargs sed -i 's/[ \t]\+$//'
find . -type f | grep "\.h$" | xargs sed -i 's/[ \t]\+$//'
find . -type f | grep "\.dts$" | xargs sed -i 's/[ \t]\+$//'
find . -type f | grep "\.dtsi$" | xargs sed -i 's/[ \t]\+$//'

Signed-off-by: Mieczyslaw Nalewaj <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18626
Signed-off-by: Robert Marko <[email protected]>

mediatek: mt7622: add the missing phy interrupt-parent for WAX206

Based on other mt7622 dts, it is not difficult to infer that the
missing interrupt controller is "pio". Fix dtc warning:

mt7622.dtsi:282.3-26: Warning (interrupts_property): /ethernet@1b100000/mdio-bus/ethernet-phy@7:#interrupt-cells: size is (8), expected multiple of 12

Signed-off-by: Shiji Yang <[email protected]>

mediatek: mt7622: various dtc warning fixes for D-Link EAGLE PRO AI

- Add missing #address-cells and #size-cells to the partition nodes.
- Add missing #address-cells and #size-cells to the switch ports nodes.

This patch fixes the following dtc warnings:

../dts/mt7622-dlink-eagle-pro-ai-r32-a1.dts:74.3-20: Warning (reg_format): /spi@1100d000/flash@0/partitions/partition@180000/nvmem-layout/macaddr@81:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-ax3200-a1.dtsi:289.6-36: Warning (reg_format): /spi@1100d000/flash@0/partitions/partition@2C0000/partition@0:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-ax3200-a1.dtsi:294.6-36: Warning (reg_format): /spi@1100d000/flash@0/partitions/partition@2C0000/partition@800000:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-ax3200-a1.dtsi:306.6-36: Warning (reg_format): /spi@1100d000/flash@0/partitions/partition@2FC0000/partition@0:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-ax3200-a1.dtsi:311.6-36: Warning (reg_format): /spi@1100d000/flash@0/partitions/partition@2FC0000/partition@800000:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-ax3200-a1.dtsi:90.6-16: Warning (reg_format): /ethernet@1b100000/mdio-bus/switch@1f/ports/port@4:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-ax3200-a1.dtsi:97.6-16: Warning (reg_format): /ethernet@1b100000/mdio-bus/switch@1f/ports/port@6:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-r32-a1.dts:50.4-14: Warning (reg_format): /ethernet@1b100000/mdio-bus/switch@1f/ports/port@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-r32-a1.dts:55.4-14: Warning (reg_format): /ethernet@1b100000/mdio-bus/switch@1f/ports/port@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-m32-a1.dts:45.4-14: Warning (reg_format): /ethernet@1b100000/mdio-bus/switch@1f/ports/port@2:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-m32-a1.dts:50.4-14: Warning (reg_format): /ethernet@1b100000/mdio-bus/switch@1f/ports/port@3:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-r32-a1.dts:60.4-14: Warning (reg_format): /ethernet@1b100000/mdio-bus/switch@1f/ports/port@2:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7622-dlink-eagle-pro-ai-r32-a1.dts:65.4-14: Warning (reg_format): /ethernet@1b100000/mdio-bus/switch@1f/ports/port@3:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: mt7622: fix mmc dtc warnings

Add missing #address-cells and #size-cells to fix the
following dtc warnings:

mt7622-bananapi-bpi-r64.dts:243.3-13: Warning (reg_format): /mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
mt7622-bananapi-bpi-r64.dts:278.3-13: Warning (reg_format): /mmc@11240000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: mt7623: fix mmc dtc warnings

Add missing #address-cells and #size-cells to fix the
following dtc warnings:

mt7623n-bananapi-bpi-r2.dts:346.3-13: Warning (reg_format): /mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
mt7623n-bananapi-bpi-r2.dts:373.3-13: Warning (reg_format): /mmc@11240000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7623a-unielec-u7623-02.dts:23.3-13: Warning (reg_format): /mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix mxl phy node dtc warnings

Add missing #address-cells and #size-cells to fix the
following dtc warnings:

mt7981-rfb-mxl-2p5g-phy-eth1.dtso:26.5-15: Warning (reg_format): /fragment@1/__overlay__/ethernet-phy@5:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
mt7981-rfb-mxl-2p5g-phy-swp5.dtso:27.5-15: Warning (reg_format): /fragment@1/__overlay__/ethernet-phy@5:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix SPI node reg size dtc warnings

Add missing #address-cells and #size-cells to fix the
following dtc warnings:

../dts/mt7981b-cmcc-rax3000m-nand.dtso:67.5-15: Warning (reg_format): /fragment@4/__overlay__/spi_nand@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-rfb-snfi-nand.dtso:20.5-15: Warning (reg_format): /fragment@0/__overlay__/flash@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-rfb-spim-nand.dtso:22.5-15: Warning (reg_format): /fragment@0/__overlay__/flash@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-rfb-spim-nor.dtso:31.5-15: Warning (reg_format): /fragment@0/__overlay__/flash@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix bpi-r4 eeprom dtc warnings

Add missing #address-cells and #size-cells to fix the
following dtc warnings:

mt7988a-bananapi-bpi-r4-wifi-mt7996a.dtso:31.5-18: Warning (reg_format): /fragment@1/__overlay__/wifi_eeprom@51:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-bananapi-bpi-r4-wifi-mt7996a.dtso:50.5-18: Warning (reg_format): /fragment@1/__overlay__/wifi_eeprom@52:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix PCIe node dtc warnings

Add missing #address-cells and #size-cells to fix the
following dtc warnings:

../dts/mt7986a-smartrg-bonanza-peak.dtsi:359.4-27: Warning (reg_format): /soc/pcie@11280000/pcie@0,0/mt7915@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988a-smartrg-mt-stuart.dtsi:551.4-27: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988a-arcadyan-mozart.dts:331.4-27: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@1,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988d-asus-zenwifi-bt8.dtsi:291.4-27: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-bananapi-bpi-r4-wifi-mt7996a.dtso:72.5-28: Warning (reg_format): /fragment@2/__overlay__/pcie@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-bananapi-bpi-r4-wifi-mt7996a.dtso:76.6-29: Warning (reg_format): /fragment@2/__overlay__/pcie@0,0/wifi@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-bananapi-bpi-r4-wifi-mt7996a.dtso:88.5-28: Warning (reg_format): /fragment@3/__overlay__/pcie@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-bananapi-bpi-r4-wifi-mt7996a.dtso:92.6-29: Warning (reg_format): /fragment@3/__overlay__/pcie@0,0/wifi@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-rfb-spim-nand-factory.dtso:47.5-28: Warning (reg_format): /fragment@1/__overlay__/pcie@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-rfb-spim-nand-factory.dtso:51.6-29: Warning (reg_format): /fragment@1/__overlay__/pcie@0,0/wifi@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-tplink-tl-xtr8488.dts:305.4-27: Warning (reg_format): /soc/pcie@11280000/pcie@0,0/wifi@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-zyxel-ex5700-telenor.dts:268.4-27: Warning (reg_format): /soc/pcie@11280000/pcie@0,0/wifi@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-bananapi-bpi-r4-wifi-mt7996a.dtso:77.5-28: Warning (reg_format): /fragment@2/__overlay__/pcie@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-bananapi-bpi-r4-wifi-mt7996a.dtso:95.5-28: Warning (reg_format): /fragment@3/__overlay__/pcie@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)
mt7988a-rfb-spim-nand-factory.dtso:49.5-28: Warning (reg_format): /fragment@1/__overlay__/pcie@0,0:reg: property has invalid length (20 bytes) (#address-cells == 2, #size-cells == 1)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix wm8960 dtc warnings

Add missing #address-cells, #size-cells and #sound-dai-cells to
fix the following dtc warnings:

mt7986a-bananapi-bpi-r3-respeaker-2mics.dtso:46.5-18: Warning (reg_format): /fragment@1/__overlay__/wm8960@1a:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
mt7986a-bananapi-bpi-r3-respeaker-2mics.dtso:29.11-31.7: Warning (sound_dai_property): /fragment@0/__overlay__/sound_wm8960/codec: Missing property '#sound-dai-cells' in node /fragment@1/__overlay__/wm8960@1a or bad phandle (referred from sound-dai[0])
mt7986a-rfb-spim-nand.dts:17.12-19.5: Warning (sound_dai_property): /sound_wm8960/platform: Missing property '#sound-dai-cells' in node /soc/audio-controller@11210000 or bad phandle (referred from sound-dai[0])
mt7986a-rfb-spim-nand.dts:21.9-23.5: Warning (sound_dai_property): /sound_wm8960/codec: Missing property '#sound-dai-cells' in node /soc/i2c@11008000/wm8960@1a or bad phandle (referred from sound-dai[0])

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix WiFi MAC address invalid length dtc warnings

Add missing #address-cells and #size-cells to fix the
following dtc warnings:

../dts/mt7988a-smartrg-mt-stuart.dtsi:558.5-15: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0/band@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988a-smartrg-mt-stuart.dtsi:565.5-15: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988a-smartrg-mt-stuart.dtsi:572.5-15: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0/band@2:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988d-asus-zenwifi-bt8.dts:78.3-13: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0/band@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988d-asus-zenwifi-bt8.dts:85.3-13: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988d-asus-zenwifi-bt8.dts:92.3-13: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0/band@2:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988d-asus-zenwifi-bt8-ubootmod.dts:82.3-13: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0/band@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988d-asus-zenwifi-bt8-ubootmod.dts:89.3-13: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988d-asus-zenwifi-bt8-ubootmod.dts:96.3-13: Warning (reg_format): /soc/pcie@11300000/pcie@0,0/mt7996@0,0/band@2:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-cmcc-a10.dtsi:250.3-13: Warning (reg_format): /soc/wifi@18000000/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-dlink-aquila-pro-ai-m60-a1.dts:324.3-13: Warning (reg_format): /soc/wifi@18000000/band@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-dlink-aquila-pro-ai-m60-a1.dts:330.3-13: Warning (reg_format): /soc/wifi@18000000/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-gatonetworks-gdsp.dts:247.3-13: Warning (reg_format): /soc/wifi@18000000/band@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-gatonetworks-gdsp.dts:253.3-13: Warning (reg_format): /soc/wifi@18000000/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-huasifei-wh3000.dts:182.3-13: Warning (reg_format): /soc/wifi@18000000/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986b-mercusys-mr90x-v1-ubi.dts:114.3-13: Warning (reg_format): /soc/wifi@18000000/band@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986b-mercusys-mr90x-v1-ubi.dts:120.3-13: Warning (reg_format): /soc/wifi@18000000/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-openembed-som7981.dts:287.3-13: Warning (reg_format): /soc/wifi@18000000/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-openwrt-one.dts:446.3-13: Warning (reg_format): /soc/wifi@18000000/band@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-openwrt-one.dts:452.3-13: Warning (reg_format): /soc/wifi@18000000/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-routerich-ax3000-v1.dts:323.3-13: Warning (reg_format): /soc/wifi@18000000/band@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-routerich-ax3000-v1.dts:329.3-13: Warning (reg_format): /soc/wifi@18000000/band@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix "unknown,gca230718" dtc warnings

Add missing #address-cells and #size-cells to fix the
following dtc warnings:

../dts/mt7981b-dlink-aquila-pro-ai-m30-a1.dts:289.4-14: Warning (reg_format): /soc/i2c@11007000/gca230718@40/led@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-dlink-aquila-pro-ai-m30-a1.dts:295.4-14: Warning (reg_format): /soc/i2c@11007000/gca230718@40/led@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-dlink-aquila-pro-ai-m30-a1.dts:301.4-14: Warning (reg_format): /soc/i2c@11007000/gca230718@40/led@2:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-dlink-aquila-pro-ai-m30-a1.dts:307.4-14: Warning (reg_format): /soc/i2c@11007000/gca230718@40/led@3:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-dlink-aquila-pro-ai-m60-a1.dts:356.4-14: Warning (reg_format): /soc/i2c@11008000/gca230718@40/led@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-dlink-aquila-pro-ai-m60-a1.dts:362.4-14: Warning (reg_format): /soc/i2c@11008000/gca230718@40/led@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-dlink-aquila-pro-ai-m60-a1.dts:368.4-14: Warning (reg_format): /soc/i2c@11008000/gca230718@40/led@2:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-dlink-aquila-pro-ai-m60-a1.dts:374.4-14: Warning (reg_format): /soc/i2c@11008000/gca230718@40/led@3:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix memory node invalid length dtc warnings

Correct the dts memory reg property to fix the following dtc
warnings:

../dts/mt7986a-zyxel-ex5601-t0-stock.dts:19.3-33: Warning (reg_format): /memory@40000000:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 2)
../dts/mt7986a-zyxel-ex5601-t0-ubootmod.dts:19.3-33: Warning (reg_format): /memory@40000000:reg: property has invalid length (8 bytes) (#address-cells == 2, #size-cells == 2)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix unit name leading 0s dtc warnings

Trim unnecessary 0s to fix the following dtc warnings:

../dts/mt7981b-cudy-re3000-v1.dts:135.20-139.6: Warning (unit_address_format): /soc/spi@11009000/flash@0/partitions/partition@00000: unit name should not have leading 0s
../dts/mt7981b-cudy-wr3000-v1.dts:153.20-157.6: Warning (unit_address_format): /soc/spi@11009000/flash@0/partitions/partition@00000: unit name should not have leading 0s
../dts/mt7981b-gatonetworks-gdsp.dts:276.20-280.6: Warning (unit_address_format): /soc/spi@11009000/flash@0/partitions/partition@00000: unit name should not have leading 0s
mt7988a-rfb-spim-nor.dtso:36.21-39.7: Warning (unit_address_format): /fragment@0/__overlay__/flash@0/partition@00000: unit name should not have leading 0s
../dts/mt7981a-ubnt-unifi-6-plus.dts:113.28-131.6: Warning (unit_address_format): /soc/spi@11009000/flash@0/partitions/partition@00000: unit name should not have leading 0s
../dts/mt7981b-wavlink-wl-wn586x3.dts:147.20-151.6: Warning (unit_address_format): /soc/spi@11009000/flash@0/partitions/partition@00000: unit name should not have leading 0s
../dts/mt7981b-wavlink-wl-wn573hx3.dts:102.20-106.6: Warning (unit_address_format): /soc/spi@11009000/flash@0/partitions/partition@00000: unit name should not have leading 0s
../dts/mt7981b-yuncore-ax835.dts:161.20-165.6: Warning (unit_address_format): /soc/spi@11009000/flash@0/partitions/partition@00000: unit name should not have leading 0s
../dts/mt7986a-zyxel-ex5601-t0-stock.dts:57.39-61.6: Warning (unit_address_format): /soc/spi@1100a000/spi_nand@0/partitions/partition@180000/nvmem-layout/macaddr@0004: unit name should not have leading 0s
../dts/mt7986a-zyxel-ex5601-t0-stock.dts:63.39-67.6: Warning (unit_address_format): /soc/spi@1100a000/spi_nand@0/partitions/partition@180000/nvmem-layout/macaddr@0024: unit name should not have leading 0s
../dts/mt7986a-zyxel-ex5601-t0-stock.dts:69.39-73.6: Warning (unit_address_format): /soc/spi@1100a000/spi_nand@0/partitions/partition@180000/nvmem-layout/macaddr@002a: unit name should not have leading 0s
../dts/mt7986a-zyxel-ex5601-t0-ubootmod.dts:56.39-60.6: Warning (unit_address_format): /soc/spi@1100a000/spi_nand@0/partitions/partition@180000/nvmem-layout/macaddr@0004: unit name should not have leading 0s
../dts/mt7986a-zyxel-ex5601-t0-ubootmod.dts:62.39-66.6: Warning (unit_address_format): /soc/spi@1100a000/spi_nand@0/partitions/partition@180000/nvmem-layout/macaddr@0024: unit name should not have leading 0s
../dts/mt7986a-zyxel-ex5601-t0-ubootmod.dts:68.39-72.6: Warning (unit_address_format): /soc/spi@1100a000/spi_nand@0/partitions/partition@180000/nvmem-layout/macaddr@002a: unit name should not have leading 0s

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix flash node name dtc warnings

Correct dts node name based on reg property to fix the
following dtc warnings:

mt7981-rfb-spim-nand.dtso:20.25-67.6: Warning (spi_bus_reg): /fragment@1/__overlay__/spi_nand@0: SPI bus unit address format error, expected "1"
mt7986a-rfb-spim-nand.dts:41.23-76.4: Warning (spi_bus_reg): /soc/spi@1100a000/spi_nand@0: SPI bus unit address format error, expected "1"
../dts/mt7981a-comfast-cf-e393ax.dts:135.23-206.4: Warning (spi_bus_reg): /soc/spi@1100a000/spi_nand@0: SPI bus unit address format error, expected "1"
../dts/mt7981b-openwrt-one.dts:305.10-342.4: Warning (spi_bus_reg): /soc/spi@1100a000/flash@0: SPI bus unit address format error, expected "1"
../dts/mt7986a-zyxel-ex5601-t0-common.dtsi:166.30-180.4: Warning (spi_bus_reg): /soc/spi@1100a000/spi_nand@0: SPI bus unit address format error, expected "1"

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix mmc dtc warnings

Add missing #address-cells and #size-cells to fix the
following dtc warnings:

../dts/mt7986a-smartrg-bonanza-peak.dtsi:310.3-13: Warning (reg_format): /soc/mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
mt7986a-bananapi-bpi-r3-emmc.dtso:29.5-15: Warning (reg_format): /fragment@0/__overlay__/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
mt7986a-bananapi-bpi-r3-sd.dtso:23.5-15: Warning (reg_format): /fragment@0/__overlay__/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-bananapi-bpi-r3-mini.dts:274.3-13: Warning (reg_format): /soc/mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-cmcc-rax3000m-emmc.dtso:48.5-15: Warning (reg_format): /fragment@3/__overlay__/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-glinet-gl-mt2500.dts:158.3-13: Warning (reg_format): /soc/mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-glinet-gl-mt6000.dts:319.3-13: Warning (reg_format): /soc/mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981a-glinet-gl-x3000-xe3000-common.dtsi:142.3-13: Warning (reg_format): /soc/mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-huasifei-wh3000.dts:118.3-13: Warning (reg_format): /soc/mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-jdcloud-re-cp-03.dts:171.3-13: Warning (reg_format): /soc/mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-nradio-c8-668gl.dts:110.3-13: Warning (reg_format): /soc/mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7981b-unielec-u7981-01-emmc.dts:26.3-13: Warning (reg_format): /soc/mmc@11230000/card@0:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)

Signed-off-by: Shiji Yang <[email protected]>

mediatek: filogic: fix "srg,sysled" dtc warnings

Add missing #address-cells and #size-cells to fix the
following dtc warnings:

../dts/mt7986a-smartrg-bonanza-peak.dtsi:562.4-14: Warning (reg_format): /soc/i2c@11008000/system-leds/led@1:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-smartrg-bonanza-peak.dtsi:568.4-14: Warning (reg_format): /soc/i2c@11008000/system-leds/led@2:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-smartrg-bonanza-peak.dtsi:574.4-14: Warning (reg_format): /soc/i2c@11008000/system-leds/led@3:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-smartrg-bonanza-peak.dtsi:580.4-14: Warning (reg_format): /soc/i2c@11008000/system-leds/led@4:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988a-smartrg-mt-stuart.dtsi:394.4-14: Warning (reg_format): /soc/i2c@11004000/system-leds/system_red:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988a-smartrg-mt-stuart.dtsi:399.4-14: Warning (reg_format): /soc/i2c@11004000/system-leds/system_green:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988a-smartrg-mt-stuart.dtsi:404.4-14: Warning (reg_format): /soc/i2c@11004000/system-leds/system_blue:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7988a-smartrg-mt-stuart.dtsi:409.4-14: Warning (reg_format): /soc/i2c@11004000/system-leds/system_white:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
../dts/mt7986a-smartrg-bonanza-peak.dtsi:559.14-588.4: Warning (i2c_bus_reg): /soc/i2c@11008000/system-leds: I2C bus unit address format error, expected "30"
../dts/mt7988a-smartrg-mt-stuart.dtsi:388.14-413.4: Warning (i2c_bus_reg): /soc/i2c@11004000/system-leds: I2C bus unit address format error, expected "30"

Signed-off-by: Shiji Yang <[email protected]>

kernel: modules: define fb-io-fops module

Define define fb-io-fops module - Fbdev helpers for framebuffers in I/O memory
By creating a hidden module it is possible to load fb_io_fops.ko
according to the kernel version.

Fixes: 5048c9d ("kernel: modules: adjust the object files of kmod-fb ")
Package kmod-drm-radeon is missing dependencies for the following libraries:
fb_io_fops.ko
make[2]: *** [modules/video.mk:620: /__w/openwrt/openwrt/openwrt/bin/targets/x86/geode/packages/kmod-drm-radeon-6.12.28-r1.apk] Error 1
time: package/kernel/linux/compile#45.35#40.01#75.34

Signed-off-by: Mieczyslaw Nalewaj <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18771
Signed-off-by: Robert Marko <[email protected]>

treewide: dts: trim "#size-cells" from "gpio-export" node

The "gpio-export" driver doesn't require a "reg" property in the
device tree, hence we don't need to use the "#size-cells" property
to describe the size of "reg".

Signed-off-by: Shiji Yang <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18290
Signed-off-by: Robert Marko <[email protected]>

qualcommax: TP-Link EAP620HD fix factory data

The re-factoring of the preinit factory partition mount script used the
incorrect mtd name for the EAP620HD and (I assume incorrect) for EAP660HD.
This corrects it to "factory_data".

Signed-off-by: Alex Thompson <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18834
Signed-off-by: Robert Marko <[email protected]>

openssl: Update to 3.5.0

Refresh patch:
- 100-Configure-afalg-support.patch
- 120-strip-cflags-from-binary.patch
- 140-allow-prefer-chacha20.patch
- 500-e_devcrypto-default-to-not-use-digests-in-engine.patch
- 510-e_devcrypto-ignore-error-when-closing-session.patch

Signed-off-by: Ryan Keane <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/15804
Signed-off-by: Robert Marko <[email protected]>

generic: refresh patches

Made: 'make target/linux/refresh'

Fixes: fbd31da8402e ("generic: make xhci-pci-renesas a proper modular driver")
Signed-off-by: Mieczyslaw Nalewaj <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18832
Signed-off-by: Robert Marko <[email protected]>

rtl8812au-ct: fix netdevice notifier call warning

rtw_ndev_notifier_call should ignore network devices created by other drivers, and it does so,
but when CONFIG_WIRELESS_EXT=n, it does not behave as expected.

Signed-off-by: Liangbin Lian <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18801
Signed-off-by: Hauke Mehrtens <[email protected]>

mvebu: replace kmod-usb3 with kmod-usb-xhci-pci-renesas

The dependency of kmod-usb3 and kmod-usb-xhci-pci-renesas have
been reversed. Now we need to select kmod-usb-xhci-pci-renesas
instead of kmod-usb3.

Signed-off-by: Shiji Yang <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18768
Signed-off-by: Hauke Mehrtens <[email protected]>

lantiq: replace kmod-usb3 with kmod-usb-xhci-pci-renesas

The dependency of kmod-usb3 and kmod-usb-xhci-pci-renesas have
been reversed. Now we need to select kmod-usb-xhci-pci-renesas
instead of kmod-usb3.

Signed-off-by: Shiji Yang <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18768
Signed-off-by: Hauke Mehrtens <[email protected]>

apm821xx: replace kmod-usb3 with kmod-usb-xhci-pci-renesas

The dependency of kmod-usb3 and kmod-usb-xhci-pci-renesas have
been reversed. Now we need to select kmod-usb-xhci-pci-renesas
instead of kmod-usb3.

Signed-off-by: Shiji Yang <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18768
Signed-off-by: Hauke Mehrtens <[email protected]>

kernel: modules: update dependency for kmod-usb-xhci-pci-renesas

Now kmod-usb-xhci-pci-renesas is depend on kmod-usb3 on both
6.6 and 6.12 kernels. Also remove the hidden flag so that users
can select it from menuconfig. Fix the following error:

Package kmod-usb-xhci-pci-renesas is missing dependencies for the following libraries:
xhci-pci.ko

Link: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=linux-6.12.y&id=25f51b76f90f10f9bf2fbc05fc51cf685da7ccad
Signed-off-by: Shiji Yang <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18768
Signed-off-by: Hauke Mehrtens <[email protected]>

generic: make xhci-pci-renesas a proper modular driver

In the new 6.12 kernel, the dependencies of kmod-usb3 and
kmod-usb-xhci-pci-renesas have been reversed. To address the
issue of package recursive dependencies, we need to backport
this patchset to make the dependencies consistent between the
6.6 and 6.12 kernels.

Links:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=linux-6.10.y&id=d6b2b694dd536d16566424c395d5642783ed6f34
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=linux-6.11.y&id=9fdce69f674e49be669e97919d24b1ab5d2c9684
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=linux-6.12.y&id=25f51b76f90f10f9bf2fbc05fc51cf685da7ccad
Signed-off-by: Shiji Yang <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18768
Signed-off-by: Hauke Mehrtens <[email protected]>

generic: fix disable common USB quirks patch

Remove modifications for codes guarded by CONFIG_USB_PCI_AMD.
This kernel symbol is only visible on desktop level platform,
e.g. x86. They are not sensitive to firmware image size. This
patch also fixes function redefinition build errors:

drivers/usb/host/pci-quirks.c:621:6: error: redefinition of 'usb_asmedia_modifyflowcontrol'
  621 | void usb_asmedia_modifyflowcontrol(struct pci_dev *pdev)
      |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from drivers/usb/host/pci-quirks.c:22:
drivers/usb/host/pci-quirks.h:49:20: note: previous definition of 'usb_asmedia_modifyflowcontrol' with type 'void(struct pci_dev *)'
   49 | static inline void usb_asmedia_modifyflowcontrol(struct pci_dev *pdev) {}
      |                    ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Fixes: be6753dda00c ("generic: 6.12: manually rebuild pending patches")
Signed-off-by: Shiji Yang <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18637
Signed-off-by: Hauke Mehrtens <[email protected]>

tools: m4: update to 1.4.20

New m4 release was made after a long time, for us most importantly
it ships newer gnulib so it compiles against GCC15 without forcing GNU17.

* Noteworthy changes in release 1.4.20 (2025-05-10) [stable]

** Fix a bug in the `eval' builtin where it does not suppress warnings
   about division by zero that occurs within a more complex expression on
   the right hand side of || or && (present since short-circuiting was
   introduced in 1.4.8b).

** The `syscmd' and `esyscmd' builtins no longer mishandle a command line
   starting with `-' or `+' (present since "the beginning").

** Fix regression introduced in 1.4.19 where trace output (such as with
   `debugmode(t)') could read invalid memory when tracing a series of
   pushed macros that are popped during argument collection.

** Fix regression introduced in 1.4.19 where the `format' builtin
   inadvertently took on locale-dependent parsing and output of floating
   point numbers as a side-effect of introducing message translations.
   While it would be nice for m4 to be fully locale-aware, such a behavior
   change belongs in a major version release such as 1.6, and not a minor
   release.

** Fix regression introduced in 1.4.11 where the experimental `changeword'
   builtin could cause a crash if given a regex that does not match all
   one-byte prefixes of valid longer matches.  As a reminder, `changeword'
   is not recommended for production use, and will likely not be present
   in the next major version release.

** On non-Unix platforms where binary files differ from text, loading a
   frozen file (which should be cross-platform compatible) now correctly
   uses binary mode.

** Several documentation improvements to the manual.

** Update to comply with newer C standards, and inherit portability
   improvements from gnulib.

Link: https://github.com/openwrt/openwrt/pull/18821
Signed-off-by: Robert Marko <[email protected]>

x86: generic: add missing CONFIG_DRM_ACCEL_QAIC

CONFIG_DRM_ACCEL_QAIC was not added to x86 generic config, but it needs to
be added as CONFIG_DRM_ACCEL_QAIC does not depend on x86_64 like others.

Fixes: 51b78aa ("kernel: package intel_vpu driver")
Signed-off-by: Robert Marko <[email protected]>

rockchip: backport PCIe MSI fixes for RK356x SoC

Backport GIC ITS support for RK356x SoC, which fixes long-standing
MSI/MSI-X bug. (Previously MSI-X could only work on one PCIe node)

e.g. [    7.250882] r8125 0002:01:00.0: no MSI/MSI-X. Back to INTx.

Tested on Radxa E25 with kmod-r8125-rss driver.

Signed-off-by: Chukun Pan <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18800
Signed-off-by: Robert Marko <[email protected]>

rockchip: replace RK356x RNG patch with upstream

Replace RK356x RNG patch with upstream version to
add the tag flag them as upstreamed.

Signed-off-by: Chukun Pan <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18800
Signed-off-by: Robert Marko <[email protected]>

rockchip: armv8: Refresh kernel config

Run 'make kernel_oldconfig' to get an up-to-date config.

Signed-off-by: Chukun Pan <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18800
Signed-off-by: Robert Marko <[email protected]>

rules.mk: respect the empty CONFIG_HOST_FLAGS_OPT

Currently, the empty CONFIG_HOST_FLAGS_OPT will be overridden by
"-O2". Fix this issue by checking if its parent menuconfig symbol
CONFIG_OPTIMIZE_HOST_TOOLS was defined.

Fixes: e3bec5692c50 ("rules.mk: set default host tools GCC optimization level to -O2")
Reported-by: Christian Marangi <[email protected]>
Signed-off-by: Shiji Yang <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18815
Signed-off-by: Robert Marko <[email protected]>

image: respect DEFAULT and BROKEN when Default profile is selected

Currently, when you select the Default profile it does not honor DEFAULT:=n
nor BROKEN:=y in device profiles but rather just tries to build all of them.

This may work when building directly, but when using Image Builder it will
always fail since no kernel or anything else is present for devices that
have DEFAULT:=n or BROKEN:=Y set since those are skipped during build.

So, lets look for DEFAULT being set to "n" or BROKEN being set to "y" and
then remove clear _PROFILE_SET so they dont end up being marked for
installation.

Fixes: #18410
Link: https://github.com/openwrt/openwrt/pull/18814
Signed-off-by: Robert Marko <[email protected]>

qualcommax: ipq50xx: spi-qpic-snand: default to 4-bit ECC

There are NAND IC-s that define 1-bit ECC as the minimal strength,
however that is unsupported by QPIC-SNAND as it only supports 4 or 8 bit
ECC.

Since most of these chips also support 4-bit ECC just fine, instead of
erroring out if 1-bit ECC is requested lets instead default to 4-bit ECC.

Fixes: 01b72ce61e8f ("qualcommax: ipq50xx: remove ECC user config from board files")
Signed-off-by: George Moussalem <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qualcommax: add 6.12 as testing kernel

Allow testing 6.12 as the testing kernel.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qca-ssdk: update for 6.12 compatibility

Update to current SSDK head for 6.12 compatibility.

We must disable warnings for missing declarations and prototypes since
whole SSDK is litered with them and actually fixing them is like doing
whack a mole.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qca-nss-dp: update for 6.12 compatibility

Update to current QCA-NSS-DP head for 6.12 compatibility.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qualcommax: 6.12: MPD: add missing <linux/of_platform.h>

It seems that kernel cleaned up its kernels so we need to include
<linux/of_platform.h> for of_platform_device_create().

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qualcommax: 6.12: pwm: use devm_clk_get_enabled()

Using devm_clk_get_enabled() allows simplification of the driver and
dropping of the .remove OP as well.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qualcommax: 6.12: pwm: fixup for 6.12

6.12 PWM core introduced a bunch of incompatible changes, namely removal
of manual module owner assignment, complete PWM struct allocation and usage
refactor, etc.

So, update the driver to follow other drivers in 6.12 so it compiles.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qualcommax: 6.12: cpr: update for 6.12

Remove op is now of a void type and strlcpy was removed.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qualcommax: 6.12: apm: change remove to void

Remove op is now of a void type, update APM.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qualcommax: 6.12: refresh config

Refresh kernel config by running kernel_menuconfig.

Manually enable CONFIG_HW_RANDOM and CONFIG_CRYPTO_DEV_QCOM_RNG as these
got somehow disabled by the refresh.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qualcommax: 6.12: refresh patches

Refresh 6.12 patches, those that failed automatic refresh were refreshed
manually.

DT bindings patches that failed were dropped as we dont use them in practice.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

qualcommax: 6.12: drop upstreamed patches

Drop patches that are already present in kernel 6.12.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

kernel/qualcommax: Restore kernel files for v6.6

This is an automatically generated commit which aids following Kernel patch
history, as git will see the move and copy as a rename thus defeating the
purpose.

For the original discussion see:
https://lists.openwrt.org/pipermail/openwrt-devel/2023-October/041673.html

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

kernel/qualcommax: Create kernel files for v6.12 (from v6.6)

This is an automatically generated commit.

When doing `git bisect`, consider `git bisect --skip`.

Link: https://github.com/openwrt/openwrt/pull/18795
Signed-off-by: Robert Marko <[email protected]>

rules.mk: set default host tools GCC optimization level to -O2

This patch can help save over 60% of package build time. -O2
optimization level is the default value before the commit
e8b470139c78 ("tools: add options to optimize host binaries").

Signed-off-by: Shiji Yang <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18812
Signed-off-by: Robert Marko <[email protected]>

unetmsg: list remotely published ids via ubus call

Before this change, only locally published ids were listed

Signed-off-by: Felix Fietkau <[email protected]>

ubus: update to Git HEAD (2025-05-16)

88e6325092bf libubus: flush all pending requests on connection loss

Signed-off-by: Felix Fietkau <[email protected]>

unetmsg: return the request data when issuing async requests

Makes it possible to track and cancel requests

Signed-off-by: Felix Fietkau <[email protected]>

ethtool: update to 6.14

Version 6.14 - April 7, 2025
* Feature: list PHYs (--show-phys)
* Feature: target a specific PHY with some commands (--phy)
* Feature: more attributes for C33 PSE (--show-pse, --set-pse)
* Feature: source information for cable tests (--cable-test[-tdr])
* Feature: JSON output for module info (-m)
* Feature: misc RSS hash info improvements (-x)
* Feature: tsinfo hwtstamp provider (--{get,set}-hwtimestamp-cfg)
* Fix: fix wrong auto-negotiation state (no option)
* Fix: more explicit RSS context action (-n)
* Fix: print PHY address as decimal (no option)
* Fix: fix return value on flow hashing error (-N)
* Fix: fix JSON output for IRQ coalescing
* Fix: fix MDI-X info output (no option)
* Misc: code cleanup in module parsers
* Misc: provide module_info JSON schema
* Misc: add '-j' alias for --json
* Misc: provide AppStream metainfo XML
* Misc: update message descriptions for debugging output

Signed-off-by: Piotr Kubik <[email protected]>
Signed-off-by: Chad Monroe <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18803
Signed-off-by: Robert Marko <[email protected]>

generic: 6.12: add CONFIG_IDPF

CONFIG_IDPF is Intel(R) Infrastructure Data Path Function Support and it
should be disabled by default, if needed it can be packaged as a kmod.

Disable to prevent spreading out through target configs.

Signed-off-by: Robert Marko <[email protected]>

generic: 6.12: add CONFIG_HISILICON_ERRATUM_162100801

Noticed that CONFIG_HISILICON_ERRATUM_162100801 was being added to
qualcommax 6.12 config.

Since we dont support any Hisilicon chips, disable it in generic config
before it starts spreading in target configs.

Signed-off-by: Robert Marko <[email protected]>

x86: add missing DRM_ACCEL symbols

DRM_ACCEL symbols were not added to geode and legacy thus making the kernel
prompt for them on compilation.

Fixes: 51b78aaaae17 ("kernel: package intel_vpu driver")
Signed-off-by: Robert Marko <[email protected]>

realtek: fix flooding of unsnoopable multicast addresses

RFC4541, section 2.1.2 says:

  Packets with a destination IP (DIP) address in the 224.0.0.X range
  which are not IGMP must be forwarded on all ports.

And section 3 says:

  In IPv6, the data forwarding rules are more straight forward because
  MLD is mandated for addresses with scope 2 (link-scope) or greater.
  The only exception is the address FF02::1 which is the all hosts
  link-scope address for which MLD messages are never sent.  Packets
  with the all hosts link-scope address should be forwarded on all
  ports.

However, currently when a listener on FF12::1 or FF12::1234:0:1 for
example joins then not only packets to these addresses but also for
FF02::1 won't be flooded to all ports anymore, too. Which violates
RFC4541.

This happens because A): They all map to the same ethernet multicast
address, that is 33:33:00:00:00:01. And B) the VLAN profile L2
unknown MC flood setting will only apply flooding of 33:33:00:00:00:01
if there is no specific listener registered for it.

So to fix this, avoid registering an MDB entry in the switch for
33:33:00:00:00:01 at all.

The downside of this is that FF12::1, FF12::1234:0:1 etc.
will always be flooded, too. However fixing the handling of 224.0.0.X
and FF02::1 and adhering to RFC4541 must have priority to avoid
undesired packetloss, to avoid breaking IPv4/IPv6.

Tested-on: ZyXEL GS1900-24HP v1

Fixes: cde31976e375 ("realtek: Add support for Layer 2 Multicast")
Signed-off-by: Linus Lüssing <[email protected]>
Link: https://github.com/openwrt/openwrt/pull/18769
Signed-off-by: Robert Marko <[email protected]>