openwrt/staging/pepe2k.git
7 weeks agolibiwinfo: update to Git HEAD (2024-03-23)
Hauke Mehrtens [Sat, 23 Mar 2024 18:34:22 +0000 (19:34 +0100)]
libiwinfo: update to Git HEAD (2024-03-23)

3aa2b6b devices: add device id for MediaTek MT7601U
79a9615 devices: add device id for Realtek RTL8188CU and RTL8188FTV

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
7 weeks agorealtek: support common GPIOs on D-Link DGS-1210-16
Richard Kunze [Wed, 20 Mar 2024 18:26:27 +0000 (19:26 +0100)]
realtek: support common GPIOs on D-Link DGS-1210-16

D-Link DGS-1210-16 hangs when rebooting and has no support for the reset
button.

Fix both by enabling the same GPIOs for reboot and the reset button as
already used for D-Link DGS-1210-20 and D-Link DGS-1210-28.

Signed-off-by: Richard Kunze <kunze@tivano.de>
7 weeks agokernel: fix iptables offload build without nftables
Qingfang Deng [Tue, 12 Mar 2024 02:05:13 +0000 (10:05 +0800)]
kernel: fix iptables offload build without nftables

When rebasing patches on top of 6.1, a change that removed NF_TABLES
from NF_FLOW_TABLE's dependency was dropped accidentally, and iptables
flow offload module can only be built when nftables is selected.
Restore the change to fix it.

Fixes: 19a246bb6503 ("generic: 6.1: manually refresh hack patches")
Signed-off-by: Qingfang Deng <dqfext@gmail.com>
7 weeks agokernel: support RISC-V PMU
Qingfang Deng [Tue, 12 Mar 2024 02:08:30 +0000 (10:08 +0800)]
kernel: support RISC-V PMU

Add RISC-V PMU config symbols and enable them for riscv64.

Signed-off-by: Qingfang Deng <dqfext@gmail.com>
7 weeks agomac80211: add missing config for third 160MHz width for 5GHz radio
Paweł Owoc [Mon, 18 Mar 2024 20:53:51 +0000 (21:53 +0100)]
mac80211: add missing config for third 160MHz width for 5GHz radio

Without this configuration it is not possible to run the radio using HE160 on channels 149-177.

Fixes: #14906
Signed-off-by: Paweł Owoc <frut3k7@gmail.com>
7 weeks agofirmware-utils: update to git HEAD
Hauke Mehrtens [Sat, 23 Mar 2024 14:16:19 +0000 (15:16 +0100)]
firmware-utils: update to git HEAD

9067281 tplink-safeloader: add RE365 v1
e87f238 zycast: new tool for ZyXEL bootloader flashing
c1e69e6 tplink-safeloader: show compat_level with FW info
335d063 tplink-safeloader: bump EAP225-Outdoor v1 compat
c1e06da tplink-safeloader: bump EAP225-V3 compat_level
a5dfb5f tplink-safeloader: add TP-Link RE205 v3 support
17de365 zycast: disable build on non-Linux OS
6b24299 ptgen: fix limitation for active partition in GPT

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
7 weeks agotools: firmware-utils: update to git HEAD
Hauke Mehrtens [Sat, 23 Mar 2024 14:14:01 +0000 (15:14 +0100)]
tools: firmware-utils: update to git HEAD

6b24299 ptgen: fix limitation for active partition in GPT

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
7 weeks agotools: firmware-utils: update to git HEAD
Sander Vanheule [Sat, 23 Mar 2024 14:36:34 +0000 (15:36 +0100)]
tools: firmware-utils: update to git HEAD

Add the new 'zycast' tool for remote flashing of Zyxel devices, support
for factory image generation for two new TP-Link devices, and improved
compatibility with two other devices.

    17de36575f1e zycast: disable build on non-Linux OS
    a5dfb5fb5e6e tplink-safeloader: add TP-Link RE205 v3 support
    c1e06daf4622 tplink-safeloader: bump EAP225-V3 compat_level
    335d063cff23 tplink-safeloader: bump EAP225-Outdoor v1 compat
    c1e69e6f9c0d tplink-safeloader: show compat_level with FW info
    e87f23849790 zycast: new tool for ZyXEL bootloader flashing
    9067281d17da tplink-safeloader: add RE365 v1

Signed-off-by: Sander Vanheule <sander@svanheule.net>
7 weeks agokirkwood: add ix4-200d support to uboot-envtools
Sander van Deijck [Tue, 12 Mar 2024 00:54:38 +0000 (01:54 +0100)]
kirkwood: add ix4-200d support to uboot-envtools

This adds support for the Iomega ix4-200d device in uboot-envtools.

Signed-off-by: Sander van Deijck <sander@vandeijck.com>
7 weeks agomac80211: carl9170: re-fix fortified-memset warning
Mieczyslaw Nalewaj [Fri, 22 Mar 2024 22:15:59 +0000 (23:15 +0100)]
mac80211: carl9170: re-fix fortified-memset warning

The carl9170_tx_release() function sometimes triggers a fortified-memset
warning in my randconfig builds:

In file included from include/linux/string.h:254,
                 from drivers/net/wireless/ath/carl9170/tx.c:40:
In function 'fortify_memset_chk',
    inlined from 'carl9170_tx_release' at drivers/net/wireless/ath/carl9170/tx.c:283:2,
    inlined from 'kref_put' at include/linux/kref.h:65:3,
    inlined from 'carl9170_tx_put_skb' at drivers/net/wireless/ath/carl9170/tx.c:342:9:
include/linux/fortify-string.h:493:25: error: call to '__write_overflow_field' declared with attribute warning: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Werror=attribute-warning]
  493 |                         __write_overflow_field(p_size_field, size);

Kees previously tried to avoid this by using memset_after(), but it seems
this does not fully address the problem. I noticed that the memset_after()
here is done on a different part of the union (status) than the original
cast was from (rate_driver_data), which may confuse the compiler.

Unfortunately, the memset_after() trick does not work on driver_rates[]
because that is part of an anonymous struct, and I could not get
struct_group() to do this either. Using two separate memset() calls
on the two members does address the warning though.

Signed-off-by: Mieczyslaw Nalewaj <namiltd@yahoo.com>
7 weeks agorockchip: add Radxa E25 board support
Marius Durbaca [Wed, 28 Feb 2024 09:39:44 +0000 (09:39 +0000)]
rockchip: add Radxa E25 board support

Radxa E25 is a network application carrier board for the Radxa CM3
Industrial (CM3I) SoM, which is based on the Rockchip RK3568 SoC.

It has the following features:

- MicroSD card socket, on board eMMC flash
- 2x 2.5GbE Realtek RTL8125B Ethernet transceiver
- 1x USB Type-C port (Power and Serial console)
- 1x USB 3.0 OTG port
- mini PCIe socket (USB or PCIe)
- ngff PCIe socket (USB or SATA)
- 1x User LED and 16x RGB LEDs
- 26-pin expansion header

Installation:
Uncompress the OpenWrt sysupgrade and write it to a micro SD card or
internal eMMC using dd.

Signed-off-by: Marius Durbaca <mariusd84@gmail.com>
7 weeks agouboot-rockchip: add Radxa E25 board support
Marius Durbaca [Wed, 28 Feb 2024 08:41:31 +0000 (08:41 +0000)]
uboot-rockchip: add Radxa E25 board support

add Radxa E25 board support in uboot-rockchip

Signed-off-by: Marius Durbaca <mariusd84@gmail.com>
Reviewed-by: Tianling Shen <cnsztl@immortalwrt.org>
7 weeks agokernel: kmod-ata-ahci-platform: enable support for RK35xx
Marius Durbaca [Wed, 28 Feb 2024 08:37:44 +0000 (08:37 +0000)]
kernel: kmod-ata-ahci-platform: enable support for RK35xx

enable support for RK35xx in kmod-ata-ahci-platform kernel module

Suggested-by: Tianling Shen <cnsztl@immortalwrt.org>
Signed-off-by: Marius Durbaca <mariusd84@gmail.com>
7 weeks agotreewide: use APK compatible version schema
Paul Spooren [Sun, 17 Mar 2024 17:43:10 +0000 (18:43 +0100)]
treewide: use APK compatible version schema

Different from OPKG, APK uses a deterministic version schema which chips
the version into chunks and compares them individually. This enforces a
certain schema which was previously entirely flexible.

 - Releases are added at the very and end prefixed with an `r` like
`1.2.3-r3`.
- Hashes are prefixed with a `~` like `1.2.3~abc123`.
- Dates become semantic versions, like `2024.04.01`
- Extra tags are possible like `_git`, `_alpha` and more.

For full details see the APK test list:
https://gitlab.alpinelinux.org/alpine/apk-tools/-/blob/master/test/version.data

Signed-off-by: Paul Spooren <mail@aparcar.org>
7 weeks agoqualcommax: add 6.6 as testing kernel
Robert Marko [Thu, 29 Feb 2024 20:42:06 +0000 (21:42 +0100)]
qualcommax: add 6.6 as testing kernel

Provide kernel 6.6 as the testing kernel for qualcommax.

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agoqualcommax: ipq807x: add WLAN device path migration
Robert Marko [Sat, 9 Mar 2024 10:22:05 +0000 (11:22 +0100)]
qualcommax: ipq807x: add WLAN device path migration

Kernel 6.6 has changed the path of WLAN devices as the soc node was updated
to include an adress as well because according to spec it needed one:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/arch/arm64/boot/dts/qcom/ipq8074.dtsi?h=v6.6.21&id=da6aa1111a17db11367817ddc10c5a6c188cdc44

So, this will break existing configuration as device path was changed for
example:
"platform/soc/c000000.wifi" to "platform/soc@0/c000000.wifi"
"platform/soc/c000000.wifi+1" to "platform/soc@0/c000000.wifi+1"

PCIe attached devices also have their path changed, so lets add a script
that will migrate the paths based on the detected running kernel version
so returning to kernel 6.1 will work as well.

Co-developed-by: Sean Khan <datapronix@protonmail.com>
Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agokernel: qca-nss-dp: enable compiling against 6.6
Robert Marko [Thu, 29 Feb 2024 21:35:52 +0000 (22:35 +0100)]
kernel: qca-nss-dp: enable compiling against 6.6

Since 6.5 netdev_rx_queue was moved out of netdevice.h so include the new
header since that is where it lives now.

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agokernel: qca-ssdk: allow compiling against 6.6
Robert Marko [Thu, 29 Feb 2024 21:18:53 +0000 (22:18 +0100)]
kernel: qca-ssdk: allow compiling against 6.6

Add a patch that makes SSDK recognize kernel 6.6 and thus allows
compiling against it.

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agoqualcommax: 6.6: fix SCM header include path for WCSS
Robert Marko [Thu, 29 Feb 2024 21:04:52 +0000 (22:04 +0100)]
qualcommax: 6.6: fix SCM header include path for WCSS

SCM header was moved to <linux/firmware/qcom/qcom_scm.h> so adjust the
WCSS patch accordingly.

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agoqualcommax: 6.6: refresh config
Robert Marko [Fri, 22 Mar 2024 10:33:33 +0000 (11:33 +0100)]
qualcommax: 6.6: refresh config

Refresh 6.6 qualcommax kernel config via kernel_menuconfig.

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agoqualcommax: 6.6: make patches apply
Robert Marko [Fri, 22 Mar 2024 10:08:54 +0000 (11:08 +0100)]
qualcommax: 6.6: make patches apply

Drop backported patches already included in 6.6 and refresh the rest to
apply.

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agokernel/qualcommax: Restore kernel files for v6.1
Robert Marko [Fri, 22 Mar 2024 09:01:17 +0000 (10:01 +0100)]
kernel/qualcommax: Restore kernel files for v6.1

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.

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

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agokernel/qualcommax: Create kernel files for v6.6 (from v6.1)
Robert Marko [Fri, 22 Mar 2024 09:01:17 +0000 (10:01 +0100)]
kernel/qualcommax: Create kernel files for v6.6 (from v6.1)

This is an automatically generated commit.

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

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agomediatek: mt7629: 6.6: disable LEDS_SMARTRG_LED by default
Robert Marko [Fri, 22 Mar 2024 13:17:50 +0000 (14:17 +0100)]
mediatek: mt7629: 6.6: disable LEDS_SMARTRG_LED by default

When building MT7629 with ALL_KMODS then we get prompted for
LEDS_SMARTRG_LED and this will break CI and in future buildbot compilation.

It depends on I2C so the symbol is hidden until ALL_KMODS is used and I2C
support is available, so disable the LEDS_SMARTRG_LED symbol in 6.6 config
intentionally as is done in the 6.1 mt7629 config.

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agokernel: config: 6.6: add missing ARM64 SM4 CE symbols
Robert Marko [Fri, 22 Mar 2024 10:31:51 +0000 (11:31 +0100)]
kernel: config: 6.6: add missing ARM64 SM4 CE symbols

Disable CONFIG_CRYPTO_SM4_ARM64_CE_CCM and CONFIG_CRYPTO_SM4_ARM64_CE_GCM
in the generic config like we do with similar CE symbols.

This avoids the need to disable them in individual targets.

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agomvebu: puzzle-m90x: optimize thermal zone
Daniel Golle [Thu, 21 Mar 2024 20:59:13 +0000 (20:59 +0000)]
mvebu: puzzle-m90x: optimize thermal zone

 * have more defined PWM steps similar to the vendor's shell script doing
   the same thing
 * replace PWM values with experiencal values provided by forum users
 * increase hyteresis to prevent changing fan speed every second
   (just because it's annoying to listen to)
 * add 'hot' trip point at 75 deg. to warn users that shutdown is imminent

Tested on Puzzle M902 only for now, PWM values for Puzzle M901 might be
slightly different but should quite certainly also be rised.

Suggested-by: Martin Gierschner <martin_gierschner@gmx.de>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
7 weeks agokernel: update deprecated license information
Florian Eckert [Thu, 21 Mar 2024 11:55:01 +0000 (12:55 +0100)]
kernel: update deprecated license information

Update the deprecated license information from GPL-2.0 to GPL-2.0-only
as written in the COPYING file of the linux source tree.

Also add the 'COPYING' file to the PKG_LICENSE_FILES variable.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
7 weeks agomac80211: add missing license information
Florian Eckert [Fri, 15 Mar 2024 10:57:14 +0000 (11:57 +0100)]
mac80211: add missing license information

The lincense information for the packages mac80211 are missing.
This commit adds the missing information.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
7 weeks agoca-certificates: add missing license information
Florian Eckert [Wed, 13 Mar 2024 11:41:38 +0000 (12:41 +0100)]
ca-certificates: add missing license information

The package has no licence information. So let's fix it.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
7 weeks agowireless-regdb: add missing license information
Florian Eckert [Tue, 19 Mar 2024 14:44:44 +0000 (15:44 +0100)]
wireless-regdb: add missing license information

Add the missing license information PKG_LICENSE and PKG_LICENSE_FILES.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
7 weeks agomediatek: mt7622: add a second u-boot for redmi-ax6s
Chuanhong Guo [Sat, 2 Mar 2024 13:12:54 +0000 (21:12 +0800)]
mediatek: mt7622: add a second u-boot for redmi-ax6s

The vendor u-boot knows nothing about UBI, and we used to have a
fixed-size kernel partition for vendor u-boot and UBI for rootfs.
However, that fixed partition becomes too small eventually, and
expanding it requires complicated procedure.

This commit changed the flash layout and added a second u-boot
where the kernel supposed to be.
Now the vendor u-boot chainloads our mainline u-boot, and our
u-boot reads kernel+rootfs from UBI, verifies it, and boot
into OpenWrt.

There are two possible ways to convert from the old fw:
Flash the factory image using mtd (provided by @rany2):

mount -o remount,ro /
mount -o remount,ro /overlay
cd /tmp
dd if=factory.bin bs=1M count=4 | mtd write - kernel
dd if=factory.bin bs=1M skip=4 | mtd -r write - ubi

Or, flash the 2nd u-boot via mtd and upload the firmware
to the 2nd u-boot using tftp:

1. prepare a tftp server at 192.168.1.254 to serve the
   sysupgrade image:
   openwrt-mediatek-mt7622-xiaomi_redmi-router-ax6s-squashfs-sysupgrade.itb
2. upload the ubi-loader.itb to OpenWrt /tmp, and flash it to
   the old kernel partition:
   mtd -r write openwrt-mediatek-mt7622-xiaomi_redmi-router-ax6s-ubi-loader.itb
3. The router should reboot and flash the sysupgrade image via TFTP.

Procedure for flashing from vendor firmware shouldn't change.

Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
7 weeks agouboot-mediatek: add support for GD5F1GQ5UE
Chuanhong Guo [Sat, 9 Mar 2024 04:06:31 +0000 (12:06 +0800)]
uboot-mediatek: add support for GD5F1GQ5UE

This patch adds support for GigaDevice GD5F1GQ5UExxG to the
mtk-snfi driver in u-boot.

Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
7 weeks agouboot-mediatek: add Redmi AX6S as UBI loader
Chuanhong Guo [Sat, 2 Mar 2024 13:07:37 +0000 (21:07 +0800)]
uboot-mediatek: add Redmi AX6S as UBI loader

Add support for Xiaomi Redmi AX6S to be used as a second-stage
UBI loader.
The defconfig/env is minimal: Boot fit from UBI. If that failed,
load and boot initramfs image from TFTP.

Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
7 weeks agoglibc: don't ship /etc/localtime -> /tmp/localtime symlink
Daniel Golle [Wed, 20 Mar 2024 22:56:00 +0000 (22:56 +0000)]
glibc: don't ship /etc/localtime -> /tmp/localtime symlink

The symlink is already part of the base-files package. We don't need
to ship it with glibc again.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
7 weeks agoumdns: fix PKG_MIRROR_HASH
Daniel Golle [Wed, 20 Mar 2024 23:43:05 +0000 (23:43 +0000)]
umdns: fix PKG_MIRROR_HASH

PKG_MIRROR_HASH was accidentally generated with already APK-adapted
version string in the filename. That can't work (yet). Regenerate and
hash the file with the currently used version scheme to fix that.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
7 weeks agoumdns: update to git HEAD
Daniel Golle [Wed, 20 Mar 2024 19:42:46 +0000 (19:42 +0000)]
umdns: update to git HEAD

 e91ed40 ubus: assume that the service iface can be NULL
 4094a3c interface: remove unused peer field
 8a0c9db interface: add missing cache cleanup on interface free
 3b341f4 add the ability to announce additional hostnames

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
7 weeks agoumdns: add /etc/umdns/ to mount namespace jail
Daniel Golle [Wed, 20 Mar 2024 19:22:46 +0000 (19:22 +0000)]
umdns: add /etc/umdns/ to mount namespace jail

Make sure /etc/umdns/ is accessiable for the umdns process if it
exists and umdns is run with umdns.@umdns[0].jail='1'.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
7 weeks agoramips: mt7621: disable Edgerouter X image generation
Robert Marko [Wed, 20 Mar 2024 10:38:58 +0000 (11:38 +0100)]
ramips: mt7621: disable Edgerouter X image generation

With kernel 6.1 image size is too large for Edgerouter X current size
limit and is causing the buildbots to fail building so images for other
devices are not updated as well.

So, disable building Edgerouter X images until a workaround is found.

Signed-off-by: Robert Marko <robimarko@gmail.com>
7 weeks agodnsmasq: Backport 2 upstream patches
Hauke Mehrtens [Mon, 18 Mar 2024 21:39:56 +0000 (22:39 +0100)]
dnsmasq: Backport 2 upstream patches

These two patches are fixing minor problems with DNSSEC found shortly
after the dnsmasq 2.90 release.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
8 weeks agoath11k-firmware: update IPQ6018 to 2.5.0.1-03982
Dirk Buchwalder [Sun, 17 Mar 2024 15:21:45 +0000 (16:21 +0100)]
ath11k-firmware: update IPQ6018 to 2.5.0.1-03982

That new version seems to work more stable including mesh.
On version 2.4.0.1-01746 rproc was immediately crashing if mesh was
active.

Signed-off-by: Dirk Buchwalder <buchwalder@posteo.de>
8 weeks agokernel: bump 6.1 to 6.1.82
John Audia [Sat, 16 Mar 2024 05:40:36 +0000 (01:40 -0400)]
kernel: bump 6.1 to 6.1.82

Changelog: https://cdn.kernel.org/pub/linux/kernel/v6.x/ChangeLog-6.1.82

All patches automatically rebased.

Build system: x86/64
Build-tested: x86/64/AMD Cezanne, ramips/tplink_archer-a6-v3
Run-tested: x86/64/AMD Cezanne, ramips/tplink_archer-a6-v3

Signed-off-by: John Audia <therealgraysky@proton.me>
8 weeks agokernel: config: add symbol introduced with 6.1.82
John Audia [Sat, 16 Mar 2024 05:51:15 +0000 (01:51 -0400)]
kernel: config: add symbol introduced with 6.1.82

Add CONFIG_MITIGATION_RFDS=y to the default config[1]

1. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/diff/arch/x86/Kconfig?id=v6.1.82&id2=v6.1.81

Signed-off-by: John Audia <therealgraysky@proton.me>
8 weeks agolinux-firmware: Update Intel wifi firmware
Hauke Mehrtens [Sun, 17 Mar 2024 15:17:07 +0000 (16:17 +0100)]
linux-firmware: Update Intel wifi firmware

Update Intel wifi firmware to most recent versions supported by the
iwlwifi driver from kernel 6.6.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
8 weeks agogeneric: kernel: fix libata ledtrig support in 6.6
Pawel Dembicki [Sun, 17 Mar 2024 05:55:12 +0000 (06:55 +0100)]
generic: kernel: fix libata ledtrig support in 6.6

Upstream commit e298d8a38b23 [0] changed method how to blink delays are pased
to function. Downstream commit must follow it.

[0] https://lore.kernel.org/r/20230510162234.291439-2-hdegoede@redhat.com

Reported-by: Mieczyslaw Nalewaj <namiltd@yahoo.com>
Signed-off-by: Pawel Dembicki <paweldembicki@gmail.com>
8 weeks agoqualcommax: ipq807x: Fix MAC addresses usage for RAX120v2
Paweł Owoc [Fri, 1 Mar 2024 22:56:36 +0000 (23:56 +0100)]
qualcommax: ipq807x: Fix MAC addresses usage for RAX120v2

Currently, 6 MAC addresses are read from the "boarddata1" partition and set for network interfaces in sequence.
This partition only contains 3 MAC addresses:
1. lan mac
2. wan mac
3. wlan5g mac
As result only lan2, lan3 and lan4 have correct (OUI) MAC addresses.
lan1, lan5 and wan interfaces get MAC addresses with incorrect OUI from random data on "boarddata1" partition.

This commit fix this and use first MAC for lan and second MAC for wan interfaces.

Signed-off-by: Paweł Owoc <frut3k7@gmail.com>
8 weeks agoqualcommax: ipq60xx: set PHY mode to psgmii for port 5
Chukun Pan [Sat, 17 Feb 2024 15:15:39 +0000 (23:15 +0800)]
qualcommax: ipq60xx: set PHY mode to psgmii for port 5

The port 5 of most ipq60xx devices is connected to qca8075,
a few are connected to qca8081. So assume that the default
connection is qca8075 and set the phy mode to psgmii.

Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>
8 weeks agoqualcommax: set phy-mode to sgmii for QCA8081
Chukun Pan [Thu, 15 Feb 2024 15:18:30 +0000 (23:18 +0800)]
qualcommax: set phy-mode to sgmii for QCA8081

The dp5 node uses psgmii by default, corrected
the phy-mode to sgmii for the qca8081 phy.

Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>
8 weeks agoqualcommax: dts: cleanup whitespace
Chukun Pan [Tue, 13 Feb 2024 15:17:10 +0000 (23:17 +0800)]
qualcommax: dts: cleanup whitespace

Replace blanks with tabs, remove extra blank lines.
Also add new lines as appropriate.

Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>
8 weeks agomt76: update to Git HEAD (2024-03-18)
Felix Fietkau [Mon, 18 Mar 2024 12:19:46 +0000 (13:19 +0100)]
mt76: update to Git HEAD (2024-03-18)

a903d3169193 wifi: mt76: mt7921: fix a potential association failure upon resuming
eb0d0ce344f3 wifi: mt76: mt7921: fix suspend issue on MediaTek COB platform
841bf82e9958 wifi: mt76: fix the issue of missing txpwr settings from ch153 to ch177
ce7ccc540168 wifi: mt76: Remove redundant assignment to variable tidno
a238df940d6f wifi: mt76: mt7915: initialize rssi on adding stations
46c7d1849dbd wifi: mt76: replace skb_put with skb_put_zero
b5640b3153c7 wifi: mt76: fix tx packet loss when scanning on DBDC
7b054e5cb3af wifi: mt76: mt7915: fix mcu command format for mt7915 tx stats
3f27a64a8010 wifi: mt76: mt7915: fix bogus Tx/Rx airtime duration values
4f681a8fbc91 wifi: mt76: mt7915: fix HE PHY capabilities IE for station mode
8ede229eb8b5 wifi: mt76: mt7915: only set MT76_MCU_RESET for the main phy
2330781b8c5f wifi: mt76: mt7996: only set MT76_MCU_RESET for the main phy
e5fb6995e7eb wifi: mt76: mt7915: add support for disabling in-band discovery
b4a917417c85 wifi: mt76: mt7915: add mt7986, mt7916 and mt7981 pre-calibration
2135e201e7a9 mt76: mt7915: add fallback in case of missing precal data

Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 weeks agokernel: modules: package driver for MediaTek T7xx PCIe 5G modem
Daniel Golle [Sun, 17 Mar 2024 22:38:58 +0000 (22:38 +0000)]
kernel: modules: package driver for MediaTek T7xx PCIe 5G modem

Build and package driver for MediaTek PCIe 5G WWAN modem T7xx device
available in Linux 6.1 and 6.6 as kmod-mtk-t7xx.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
8 weeks agokernel: backport some useful LED_FUNCTION_* defines for DT
Rafał Miłecki [Sun, 17 Mar 2024 21:20:14 +0000 (22:20 +0100)]
kernel: backport some useful LED_FUNCTION_* defines for DT

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
8 weeks agokernel: bump 6.6 to 6.6.22
John Audia [Sat, 16 Mar 2024 10:49:35 +0000 (06:49 -0400)]
kernel: bump 6.6 to 6.6.22

Changelog: https://cdn.kernel.org/pub/linux/kernel/v6.x/ChangeLog-6.6.22

Affected Patch automatically rebased.

Build system: x86/64
Build-tested: x86/64/AMD Cezanne, flogic/xiaomi_redmi-router-ax6000-ubootmod, flogic/glinet_gl-mt6000, ramips/tplink_archer-a6-v3
Run-tested: x86/64/AMD Cezanne, flogic/xiaomi_redmi-router-ax6000-ubootmod, flogic/glinet_gl-mt6000, ramips/tplink_archer-a6-v3

Signed-off-by: John Audia <therealgraysky@proton.me>
8 weeks agokernel: bump 6.6 to 6.6.21
John Audia [Sat, 16 Mar 2024 10:48:39 +0000 (06:48 -0400)]
kernel: bump 6.6 to 6.6.21

Changelog: https://cdn.kernel.org/pub/linux/kernel/v6.x/ChangeLog-6.6.21

All patches automatically rebased.

Signed-off-by: John Audia <therealgraysky@proton.me>
8 weeks agomvebu: enable thermal zone polling for IEI Puzzle devices
Daniel Golle [Sun, 17 Mar 2024 15:55:09 +0000 (15:55 +0000)]
mvebu: enable thermal zone polling for IEI Puzzle devices

Marvell's thermal sensors do not support interrupts, so we need to
poll them. Reading temperature every second should be enough to
control the fan.
While at it, also make sure fan speed is reduced again if temperature
goes down.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
8 weeks agouci: fix variable declaration
Paul Spooren [Sat, 16 Mar 2024 17:30:37 +0000 (18:30 +0100)]
uci: fix variable declaration

Using := doesn't fly well when including other variables. In fact this
would cause the variable to be empty and break cloning of the git repo.

Fix: "c354c069b3 uci: fix Makefile formatting"

Signed-off-by: Paul Spooren <mail@aparcar.org>
8 weeks agoqualcommax: ipq807x: fix LEDs on Buffalo WXR-5950AX12
INAGAKI Hiroshi [Sat, 16 Mar 2024 10:12:20 +0000 (19:12 +0900)]
qualcommax: ipq807x: fix LEDs on Buffalo WXR-5950AX12

On Buffalo WXR-5590AX12, some LEDs don't have "function" property and
only "color" property is available for the newer binding of LED on
Linux Kernel.
2e659930d3 ("qualcommax: drop redundant label with new LED
color/function format") removes "label" property from LEDs, then,
multiple "<color>:" (ex.: "white:"/"red:") will be appeared as LED names
and renamed to "<color>:_<num>" ("ex.: "white:_1", "green:_2", ...) by
kernel.

logs:

[    5.372863] leds-gpio leds: Led white: renamed to white:_1 due to name collision
[    5.375796] leds-gpio leds: Led red: renamed to red:_1 due to name collision

/sys/class/leds:

root@OpenWrt:/# ls /sys/class/leds/
red:         red:power    white:       white:power
red:_1       red:wlan     white:_1     white:wlan

Fix this issue by adding missing "function" property to LEDs on Buffalo
WXR-5950AX12.
Note: there are no appropriate functions in dt-bindings/leds/common.h
for some LEDs, so use the hardcoded strings for them instead.

Fixes: 2e659930d3 ("qualcommax: drop redundant label with new LED color/function format")
Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
8 weeks agoqualcommax: ipq807x: fix LEDs on QNAP QHora-301W
INAGAKI Hiroshi [Sat, 16 Mar 2024 06:28:31 +0000 (15:28 +0900)]
qualcommax: ipq807x: fix LEDs on QNAP QHora-301W

On QNAP QHora-301W, all LEDs don't have "function" property and only
"color" property is available for the newer binding of LED on Linux
Kernel.
2e659930d3 ("qualcommax: drop redundant label with new LED
color/function format") removes "label" property from LEDs, then,
multiple "<color>:" (ex.: "amber:"/"green:"/"red:") will be appeared as
LED names and renamed to "<color>:_<num>" ("ex.: "green:_1", "green:_2",
...) by kernel.

logs:

[    4.246494] leds-gpio leds: Led green: renamed to green:_1 due to name collision
[    4.248978] leds-gpio leds: Led green: renamed to green:_2 due to name collision
[    4.256345] leds-gpio leds: Led green: renamed to green:_3 due to name collision
[    4.263770] leds-gpio leds: Led green: renamed to green:_4 due to name collision
[    4.271105] leds-gpio leds: Led amber: renamed to amber:_1 due to name collision
[    4.278487] leds-gpio leds: Led green: renamed to green:_5 due to name collision
[    4.285856] leds-gpio leds: Led amber: renamed to amber:_2 due to name collision
[    4.293257] leds-gpio leds: Led green: renamed to green:_6 due to name collision
[    4.300616] leds-gpio leds: Led amber: renamed to amber:_3 due to name collision
[    4.308002] leds-gpio leds: Led green: renamed to green:_7 due to name collision
[    4.315384] leds-gpio leds: Led amber: renamed to amber:_4 due to name collision
[    4.322760] leds-gpio leds: Led green: renamed to green:_8 due to name collision
[    4.330135] leds-gpio leds: Led amber: renamed to amber:_5 due to name collision

/sys/class/leds:

amber:          amber:_4        green:_2        green:_6
amber:_1        green:_5        green:_3        green:_7
amber:_2        green:          green:_4        green:_8
amber:_3        green:_1        green:_5        red:

Fix this issue by adding missing "function" (and "function-enumerator")
property to LEDs on QNAP QHora-301W.
Note: there are no appropriate functions in dt-bindings/leds/common.h
for some LEDs, so use the hardcoded strings for them instead.

Fixes: 2e659930d3 ("qualcommax: drop redundant label with new LED color/function format")
Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
8 weeks agoarchs38: add 6.1 as testing kernel
Robert Marko [Tue, 5 Mar 2024 09:54:11 +0000 (10:54 +0100)]
archs38: add 6.1 as testing kernel

Allow selecting 6.1 as testing kernel on archs38.

Signed-off-by: Robert Marko <robimarko@gmail.com>
8 weeks agoarchs38: 6.1: refresh kernel config
Robert Marko [Tue, 5 Mar 2024 10:21:23 +0000 (11:21 +0100)]
archs38: 6.1: refresh kernel config

Refresh kernel config by running kernel_menuconfig.

Signed-off-by: Robert Marko <robimarko@gmail.com>
8 weeks agokernel/archs38: Restore kernel files for v5.15
Robert Marko [Tue, 5 Mar 2024 09:52:38 +0000 (10:52 +0100)]
kernel/archs38: Restore kernel files for v5.15

This is an automatically generated commit and is needed to outsmart git.

Signed-off-by: Robert Marko <robimarko@gmail.com>
8 weeks agokernel/archs38: Create kernel files for v6.1 (from v5.15)
Robert Marko [Tue, 5 Mar 2024 09:52:38 +0000 (10:52 +0100)]
kernel/archs38: Create kernel files for v6.1 (from v5.15)

This is an automatically generated commit.

During a `git bisect` session, `git bisect --skip` is needed recommended.

Signed-off-by: Robert Marko <robimarko@gmail.com>
8 weeks agokirkwood: add support for Netgear Stora (MS2000/2110) NAS
Zoltan HERPAI [Thu, 5 May 2022 06:47:27 +0000 (08:47 +0200)]
kirkwood: add support for Netgear Stora (MS2000/2110) NAS

Dual-slot NAS based on Marvell Kirkwood.

Specifications:
 - Marvell 88F6281 @1GHz
 - 128Mb RAM
 - 256Mb NAND
 - 1x GbE LAN (Marvell 88E1116)
 - 1x USB 2.0
 - 2x SATA
 - PCF8563 RTC
 - LM75 sensor
 - TC654 PWM fan controller
 - Serial on J2 (115200,8n1)
 - Newer bootROM so kwboot-ing via serial is possible

Installation:

1. Serial console
 - Connect your levelshifter to the serial console
   on J2 (refer to the wiki page for pinout)
2. Update u-boot
 - Download the u-boot.kwb image for the device
 - Powercycle the NAS
 - Run "kwboot -b ./u-boot.kwb /dev/ttyUSB0 -p"
 - Connect to the serial console with minicom
 - tftp 0x0800000 netgear_stora-u-boot.kwb
 - nand erase 0x0 100000
 - nand write 0x0800000 0x0 0x100000
 - reset
3. Install OpenWrt
 - Boot up the initramfs image
 - tftpboot 0x800000 openwrt-kirkwood-netgear_stora-initramfs-uImage; bootm 0x800000
 - Download the sysupgrade image and perform sysupgrade

The fan is controlled in 3 stages by a script running every minute
from cron, measuring the CPU temperature.

Snippets taken from bodhi <mibodhi@gmail.com>

Signed-off-by: Zoltan HERPAI <wigyori@uid0.hu>
8 weeks agohwmon: add TC654 fan-controller support
Zoltan HERPAI [Tue, 5 Mar 2024 12:22:11 +0000 (13:22 +0100)]
hwmon: add TC654 fan-controller support

Add support for the Microchip TC654 PWM fan-controller.

Signed-off-by: Zoltan HERPAI <wigyori@uid0.hu>
8 weeks agomac80211: fix a crash triggered by sta disconnect with per-sta VLANs
Felix Fietkau [Sat, 16 Mar 2024 07:50:05 +0000 (08:50 +0100)]
mac80211: fix a crash triggered by sta disconnect with per-sta VLANs

Reported-by: ranygh@riseup.net
Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 weeks agokernel: xt_FLOWOFFLOAD: fix use of uninitialized dir variable
Qingfang Deng [Wed, 13 Mar 2024 03:48:44 +0000 (11:48 +0800)]
kernel: xt_FLOWOFFLOAD: fix use of uninitialized dir variable

The dir variable has been used uninitialized since the port to 5.10, and
somehow this remains undetected by GCC.

Fixes: b10d6044599d ("kernel: add linux 5.10 support")
Signed-off-by: Qingfang Deng <qingfang.deng@siflower.com.cn>
8 weeks agomediatek: mt7622: simplify 02_network
Chukun Pan [Tue, 16 Jan 2024 15:19:02 +0000 (23:19 +0800)]
mediatek: mt7622: simplify 02_network

Most mt7622 devices use the mt7531 switch, which have been
switched to dsa driver for a long time. So use dsa as the
default configuration and configure these rtl8367s devices
separately. This reduces the amount of code.

Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>
8 weeks agomediatek: fix build failure of mt7623/mt7629 using kernel 6.6
Weijie Gao [Fri, 15 Mar 2024 15:13:55 +0000 (23:13 +0800)]
mediatek: fix build failure of mt7623/mt7629 using kernel 6.6

In kernel 6.6, dts files for mediatek arm target are moved into
arch/arm/boot/dts/mediatek instead of legacy path arch/arm/boot/dts.

To avoid dts compile failure, change DTS_DIR to the mediatek subfolder
for kernel 6.6.

Signed-off-by: Weijie Gao <hackpascal@gmail.com>
8 weeks agods-lite: Add support for IPIP6(RFC2473) tunnel
Arayuki Mago [Tue, 13 Feb 2024 10:08:50 +0000 (19:08 +0900)]
ds-lite: Add support for IPIP6(RFC2473) tunnel

Add Generic Packet Tunneling in IPv6 Specification (RFC 2473) support.

Signed-off-by: Arayuki Mago <ms@missing233.com>
Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
8 weeks agomac80211: fix another regression in the broadcast AQL patch
Felix Fietkau [Fri, 15 Mar 2024 19:34:58 +0000 (20:34 +0100)]
mac80211: fix another regression in the broadcast AQL patch

Add a separate bit in struct ieee80211_tx_info to indicate airtime tracked
as broadcast/multicast. This avoids a race condition where airtime from
stations that were just removed wasn't getting subtracted from the total
PHY airtime.

Fixes: 95e633efbd1b ("mac80211: add AQL support for broadcast/multicast packets")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
8 weeks agogeneric: 6.6: backport support for SDX75-based MHI 5G modems
Daniel Golle [Fri, 15 Mar 2024 18:04:53 +0000 (18:04 +0000)]
generic: 6.6: backport support for SDX75-based MHI 5G modems

Also backport most recent MHI modem additions to Linux 6.6.
Adds support for generic SDX75-based modems.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
8 weeks agogeneric: 6.1: backport support for various MHI 5G modems
Daniel Golle [Fri, 15 Mar 2024 17:18:19 +0000 (17:18 +0000)]
generic: 6.1: backport support for various MHI 5G modems

  * Dell DW5932e
  * Foxconn T99W175 (HP variant)
  * Foxconn T99W510
  * Generic SDX75-based
  * Quectel EM160R-GL (newer variants)
  * Quectel RM520
  * Quectel RM520N-GL (Lenovo variant)
  * SC8280XP Compute Reference Design
  * Telit FE990

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
8 weeks agokernel: modules: video: select drm-exec and drm-suballoc-helper for AMDGPU
Robert Marko [Fri, 15 Mar 2024 11:50:43 +0000 (12:50 +0100)]
kernel: modules: video: select drm-exec and drm-suballoc-helper for AMDGPU

drm-amdgpu and drm-radeon require drm-exec and/or drm-suballoc-helper in
6.6, so since we have them packaged separately include them when required.

Fixes: 5b08b5600720 ("kernel: modules: video: adapt for kernel 6.6")
Signed-off-by: Robert Marko <robimarko@gmail.com>
8 weeks agokernel: modules: video: package DRM suballocation helper
Robert Marko [Fri, 15 Mar 2024 11:50:14 +0000 (12:50 +0100)]
kernel: modules: video: package DRM suballocation helper

Linux 6.4 has split out the previously AMDGPU specific suballocation
helper into a generic one and it has its own symbol now.

So, lets package it as a separate helper as AMDGPU still requires it
for 6.6.

Signed-off-by: Robert Marko <robimarko@gmail.com>
8 weeks agokernel: modules: video: package DRM EXEC
Robert Marko [Fri, 15 Mar 2024 11:14:58 +0000 (12:14 +0100)]
kernel: modules: video: package DRM EXEC

Linux 6.6 has added DRM EXEC as a separate symbol, so package it separately
as AMDGPU requires it.

Signed-off-by: Robert Marko <robimarko@gmail.com>
8 weeks agokernel: modules: video: drm: remove DRM_EXEC and DRM_SUBALLOC_HELPER
Robert Marko [Fri, 15 Mar 2024 10:34:20 +0000 (11:34 +0100)]
kernel: modules: video: drm: remove DRM_EXEC and DRM_SUBALLOC_HELPER

As part of adding kernel 6.6 support, DRM_EXEC and DRM_SUBALLOC_HELPER were
added to the kmod-drm, however these are only used by drm-amdgpu and
drm-radeon which are only supported on x86.

So, lets start fixing building of other targets by removing these from the
main kmod-drm, in follow-up commits they will be packaged separately and
selected when required.

Fixes: 5b08b5600720 ("kernel: modules: video: adapt for kernel 6.6")
Signed-off-by: Robert Marko <robimarko@gmail.com>
8 weeks agoath11k-firmware: update IPQ8074 and QCN9074 to 2.9.0.1-01977
Robert Marko [Thu, 14 Mar 2024 08:20:56 +0000 (09:20 +0100)]
ath11k-firmware: update IPQ8074 and QCN9074 to 2.9.0.1-01977

After a long time QCA has pushed an updated release of 2.9.0.1 firmware
for IPQ8074 and QCN9074, so lets update to 2.9.0.1-01977.

Sadly, still nothing new for IPQ6018.

Signed-off-by: Robert Marko <robimarko@gmail.com>
8 weeks agomediatek: mt7988: add missing clock for PCIe ports
Daniel Golle [Fri, 15 Mar 2024 14:07:58 +0000 (14:07 +0000)]
mediatek: mt7988: add missing clock for PCIe ports

Add missing CLK_TOP_PEXTP_Px_SEL clock for each of the 4 PCIe interfaces
of the MT7988 SoC. Without that clock PCIe doesn't work reliable.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
8 weeks agoqualcommax: ipq60xx: mango-dvk: fix QCA8081 reset timings
Mantas Pucka [Thu, 14 Mar 2024 11:37:02 +0000 (13:37 +0200)]
qualcommax: ipq60xx: mango-dvk: fix QCA8081 reset timings

Add missing reset delay properties for QCA8081 PHY.
Without delays PHY operation is unreliable.

Signed-off-by: Mantas Pucka <mantas@8devices.com>
2 months agokernel: Activate CONFIG_NET_VENDOR_DAVICOM
Hauke Mehrtens [Wed, 13 Mar 2024 20:00:57 +0000 (21:00 +0100)]
kernel: Activate CONFIG_NET_VENDOR_DAVICOM

CONFIG_NET_VENDOR_DAVICOM is needed to activate the CONFIG_DM9000 option
which builds the kmod-dm9000. This fixes the following warning:
logs/package/kernel/linux/compile.txt:WARNING: kmod-dm9000 is not available in the kernel config - generating empty package

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 months agokernel: kmod-hwmon-coretemp: Depend on x86
Hauke Mehrtens [Wed, 13 Mar 2024 20:02:11 +0000 (21:02 +0100)]
kernel: kmod-hwmon-coretemp: Depend on x86

This kernel module depends on x86, it works only on some Intel CPUs.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 months agokernel: kmod-video-cpia2: Depend on kernel 5.15
Hauke Mehrtens [Wed, 13 Mar 2024 20:08:55 +0000 (21:08 +0100)]
kernel: kmod-video-cpia2: Depend on kernel 5.15

This driver was moved to staging in kernel 6.1 and then removed in
kernel 6.3, see these commits:
https://git.kernel.org/linus/be8cebc46d9d38166a1b3fda22a018ae52b0928e
https://git.kernel.org/linus/9ea8a9c72a9b4d24e6045ee25f5e465dc22f9f55

Build it only on older kernel versions.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 months agokernel: kmod-can: Fix build of can drivers
Hauke Mehrtens [Wed, 13 Mar 2024 19:02:01 +0000 (20:02 +0100)]
kernel: kmod-can: Fix build of can drivers

Many can kernel modules are now gated by the newly introduced
CONFIG_CAN_NETLINK configuration option. Activate it to build the can
drivers again.
This was changed in this upstream Linux commit:
https://git.kernel.org/linus/df6ad5dd838e0fa543ca28ca6154901fa65a9443

This should fix these warnings with kernel 6.1 and 6.6:
logs/package/kernel/linux/compile.txt:WARNING: kmod-can-c-can is not available in the kernel config - generating empty package
logs/package/kernel/linux/compile.txt:WARNING: kmod-can-c-can-pci is not available in the kernel config - generating empty package
logs/package/kernel/linux/compile.txt:WARNING: kmod-can-c-can-platform is not available in the kernel config - generating empty package
logs/package/kernel/linux/compile.txt:WARNING: kmod-can-mcp251x is not available in the kernel config - generating empty package
logs/package/kernel/linux/compile.txt:WARNING: kmod-can-slcan is not available in the kernel config - generating empty package
logs/package/kernel/linux/compile.txt:WARNING: kmod-can-usb-8dev is not available in the kernel config - generating empty package
logs/package/kernel/linux/compile.txt:WARNING: kmod-can-usb-ems is not available in the kernel config - generating empty package
logs/package/kernel/linux/compile.txt:WARNING: kmod-can-usb-kvaser is not available in the kernel config - generating empty package
logs/package/kernel/linux/compile.txt:WARNING: kmod-can-usb-peak is not available in the kernel config - generating empty package

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 months agokernel: kmod-can-flexcan: Fix flexcan.ko location
Hauke Mehrtens [Thu, 14 Mar 2024 19:22:59 +0000 (20:22 +0100)]
kernel: kmod-can-flexcan: Fix flexcan.ko location

The flexcan.ko file was moved in upstream Linux:
https://git.kernel.org/linus/bfd00e021cf162049946a9e0047b0997d2b35fec
This fixes the build with kernel >= 5.17. This module was never build
with kernel 6.1 before.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 months agokernel: kmod-can-slcan: Fix can-slcan.ko location
Hauke Mehrtens [Wed, 13 Mar 2024 19:35:19 +0000 (20:35 +0100)]
kernel: kmod-can-slcan: Fix can-slcan.ko location

The slcan.ko file was moved in upstream Linux:
https://git.kernel.org/linus/98b12064591d635db86da4957b547067dc6897cc
This fixes the build with kernel >= 6.0. This module was never build
with kernel 6.1 before.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 months agokernel: kmod-crypto-gf128: Fix build with kernel 6.6
Hauke Mehrtens [Wed, 13 Mar 2024 19:24:23 +0000 (20:24 +0100)]
kernel: kmod-crypto-gf128: Fix build with kernel 6.6

The gf128mul.ko module was moved in kernel 6.2:
https://git.kernel.org/linus/61c581a46a9668747d355436bd4b2505594539bd

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 months agokernel: kmod-crypto-user: Add missing dependency
Hauke Mehrtens [Wed, 13 Mar 2024 17:09:22 +0000 (18:09 +0100)]
kernel: kmod-crypto-user: Add missing dependency

The algif_rng.ko kernel module depends on the rng.ko kernel module with
kernel 6.6 when compiling for MIPS malta.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
2 months agofirmware: intel-microcode: update to 20240312
Christian Lamparter [Thu, 14 Mar 2024 06:13:37 +0000 (07:13 +0100)]
firmware: intel-microcode: update to 20240312

Debian changelog:

intel-microcode (3.20240312.1) unstable; urgency=medium

  * New upstream microcode datafile 20240312 (closes: #1066108)
    - Mitigations for INTEL-SA-INTEL-SA-00972 (CVE-2023-39368):
      Protection mechanism failure of bus lock regulator for some Intel
      Processors may allow an unauthenticated user to potentially enable
      denial of service via network access.
    - Mitigations for INTEL-SA-INTEL-SA-00982 (CVE-2023-38575):
      Non-transparent sharing of return predictor targets between contexts in
      some Intel Processors may allow an authorized user to potentially
      enable information disclosure via local access.  Affects SGX as well.
    - Mitigations for INTEL-SA-INTEL-SA-00898 (CVE-2023-28746), aka RFDS:
      Information exposure through microarchitectural state after transient
      execution from some register files for some Intel Atom Processors and
      E-cores of Intel Core Processors may allow an authenticated user to
      potentially enable information disclosure via local access.  Enhances
      VERW instruction to clear stale register buffers.  Affects SGX as well.
      Requires kernel update to be effective.
    - Mitigations for INTEL-SA-INTEL-SA-00960 (CVE-2023-22655), aka TECRA:
      Protection mechanism failure in some 3rd and 4th Generation Intel Xeon
      Processors when using Intel SGX or Intel TDX may allow a privileged
      user to potentially enable escalation of privilege via local access.
      NOTE: effective only when loaded by firmware.  Allows SMM firmware to
      attack SGX/TDX.
    - Mitigations for INTEL-SA-INTEL-SA-01045 (CVE-2023-43490):
      Incorrect calculation in microcode keying mechanism for some Intel
      Xeon D Processors with Intel SGX may allow a privileged user to
      potentially enable information disclosure via local access.
  * Fixes for other unspecified functional issues on many processors
  * Updated microcodes:
    sig 0x00050653, pf_mask 0x97, 2023-07-28, rev 0x1000191, size 36864
    sig 0x00050656, pf_mask 0xbf, 2023-07-28, rev 0x4003605, size 38912
    sig 0x00050657, pf_mask 0xbf, 2023-07-28, rev 0x5003605, size 37888
    sig 0x0005065b, pf_mask 0xbf, 2023-08-03, rev 0x7002802, size 30720
    sig 0x00050665, pf_mask 0x10, 2023-08-03, rev 0xe000015, size 23552
    sig 0x000506f1, pf_mask 0x01, 2023-10-05, rev 0x003e, size 11264
    sig 0x000606a6, pf_mask 0x87, 2023-09-14, rev 0xd0003d1, size 307200
    sig 0x000606c1, pf_mask 0x10, 2023-12-05, rev 0x1000290, size 299008
    sig 0x000706a1, pf_mask 0x01, 2023-08-25, rev 0x0040, size 76800
    sig 0x000706a8, pf_mask 0x01, 2023-08-25, rev 0x0024, size 76800
    sig 0x000706e5, pf_mask 0x80, 2023-09-14, rev 0x00c4, size 114688
    sig 0x000806c1, pf_mask 0x80, 2023-09-13, rev 0x00b6, size 111616
    sig 0x000806c2, pf_mask 0xc2, 2023-09-13, rev 0x0036, size 98304
    sig 0x000806d1, pf_mask 0xc2, 2023-09-13, rev 0x0050, size 104448
    sig 0x000806ec, pf_mask 0x94, 2023-07-16, rev 0x00fa, size 106496
    sig 0x000806f8, pf_mask 0x87, 2024-01-03, rev 0x2b000590, size 579584
    sig 0x000806f7, pf_mask 0x87, 2024-01-03, rev 0x2b000590
    sig 0x000806f6, pf_mask 0x87, 2024-01-03, rev 0x2b000590
    sig 0x000806f5, pf_mask 0x87, 2024-01-03, rev 0x2b000590
    sig 0x000806f4, pf_mask 0x87, 2024-01-03, rev 0x2b000590
    sig 0x00090661, pf_mask 0x01, 2023-09-26, rev 0x0019, size 20480
    sig 0x00090672, pf_mask 0x07, 2023-09-19, rev 0x0034, size 224256
    sig 0x00090675, pf_mask 0x07, 2023-09-19, rev 0x0034
    sig 0x000b06f2, pf_mask 0x07, 2023-09-19, rev 0x0034
    sig 0x000b06f5, pf_mask 0x07, 2023-09-19, rev 0x0034
    sig 0x000906a3, pf_mask 0x80, 2023-09-19, rev 0x0432, size 222208
    sig 0x000906a4, pf_mask 0x80, 2023-09-19, rev 0x0432
    sig 0x000906c0, pf_mask 0x01, 2023-09-26, rev 0x24000026, size 20480
    sig 0x000906e9, pf_mask 0x2a, 2023-09-28, rev 0x00f8, size 108544
    sig 0x000906ea, pf_mask 0x22, 2023-07-26, rev 0x00f6, size 105472
    sig 0x000906ec, pf_mask 0x22, 2023-07-26, rev 0x00f6, size 106496
    sig 0x000906ed, pf_mask 0x22, 2023-07-27, rev 0x00fc, size 106496
    sig 0x000a0652, pf_mask 0x20, 2023-07-16, rev 0x00fa, size 97280
    sig 0x000a0653, pf_mask 0x22, 2023-07-16, rev 0x00fa, size 97280
    sig 0x000a0655, pf_mask 0x22, 2023-07-16, rev 0x00fa, size 97280
    sig 0x000a0660, pf_mask 0x80, 2023-07-16, rev 0x00fa, size 97280
    sig 0x000a0661, pf_mask 0x80, 2023-07-16, rev 0x00fa, size 96256
    sig 0x000a0671, pf_mask 0x02, 2023-09-14, rev 0x005e, size 108544
    sig 0x000b0671, pf_mask 0x32, 2023-12-14, rev 0x0122, size 215040
    sig 0x000b06a2, pf_mask 0xe0, 2023-12-07, rev 0x4121, size 220160
    sig 0x000b06a3, pf_mask 0xe0, 2023-12-07, rev 0x4121
    sig 0x000b06e0, pf_mask 0x11, 2023-09-25, rev 0x0015, size 138240
  * New microcodes:
    sig 0x000a06a4, pf_mask 0xe6, 2024-01-03, rev 0x001c, size 136192
    sig 0x000b06a8, pf_mask 0xe0, 2023-12-07, rev 0x4121, size 220160
    sig 0x000c06f2, pf_mask 0x87, 2023-11-20, rev 0x21000200, size 549888
    sig 0x000c06f1, pf_mask 0x87, 2023-11-20, rev 0x21000200
  * source: update symlinks to reflect id of the latest release, 20240312
  * changelog, debian/changelog: fix typos

 -- Henrique de Moraes Holschuh <hmh@debian.org>  Tue, 12 Mar 2024 20:28:17 -0300

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
2 months agomxs: switch default kernel to 6.1
Zoltan HERPAI [Thu, 14 Mar 2024 13:21:17 +0000 (14:21 +0100)]
mxs: switch default kernel to 6.1

The testing kernel has been available since Jul/2023, let's switch
to it by default.

Signed-off-by: Zoltan HERPAI <wigyori@uid0.hu>
2 months agoramips: 5.15: delete remaining files
Nick Hainke [Wed, 13 Mar 2024 22:49:32 +0000 (23:49 +0100)]
ramips: 5.15: delete remaining files

Remove the remaining configuration entries that were omitted in the
previous commit.

Fixes: 1576474f55e9 ("ramips: switch to 6.1 kernel")
Signed-off-by: Nick Hainke <vincent@systemli.org>
2 months agomediatek: Add support for GL.iNet X3000 (Spitz AX) and XE3000 (Puli AX)
Jean Thomas [Thu, 9 Nov 2023 10:28:45 +0000 (11:28 +0100)]
mediatek: Add support for GL.iNet X3000 (Spitz AX) and XE3000 (Puli AX)

The GL.iNet X3000 and XE3000 are Wi-Fi 6 5G cellular routers, based on
MediaTek MT7981A SoC. The XE3000 is the same device as the X3000,
except for an additional battery.

Specifications:
 - SoC: Filogic 820 MT7981A (1.3GHz)
 - RAM: DDR4 512M
 - Flash: eMMC 8G, MicroSD card slot
 - WiFi: 2.4GHz and 5GHz with 6 antennas
 - Ethernet:
   - 1x LAN (10/100/1000M)
   - 1x WAN (10/100/1000/2500M)
 - 5G: Quectel RM520N-GL with two nano-SIM card slots
 - USB: 1x USB 2.0 port
 - UART:
   - 3.3V, TX, RX, GND / 115200 8N1

MAC addresses as verified by OEM firmware:
vendor   OpenWrt    address     source
WAN      eth0       label       factory 0x0a (label)
LAN      eth1       label + 1
2g       phy0-ap0   label + 2   factory 0x04
5g       phy1-ap0   label + 3

Installation via U-Boot rescue:
1. Press and hold reset button while booting the device
2. Wait for the Internet led to blink 5 times
3. Release reset button
4. The rescue page is accessible via http://192.168.1.1
5. Select the OpenWrt sysupgrade image and start upgrade
6. Wait for the router to flash new firmware and reboot

Revert to stock firmware:
1. Download the stock firmware from GL.iNet website
2. Use the method explained above to flash the stock firmware

Switch the modem network port between PCIe and USB interfaces:
1. Connect to the AT commands (/dev/ttyUSB2) port using
e.g. minicom: minicom -D /dev/ttyUSB2
2. Check the current modem mode with 'AT+QCFG="data_interface"':
  - 0,0 indicates that the network port uses the USB interface
  - 1,0 indicates that the network port uses the PCIe interface
3. Switch the active interface with:
  - 'AT+QCFG="data_interface",0,0' to use the USB interface
  - 'AT+QCFG="data_interface",1,0' to use the PCIe interface
4. Reboot

Signed-off-by: Jean Thomas <jean.thomas@wifirst.fr>
2 months agomediatek: mt7981: pinctrl: add additional emmc groups
Jean Thomas [Thu, 7 Dec 2023 15:18:02 +0000 (16:18 +0100)]
mediatek: mt7981: pinctrl: add additional emmc groups

Add new emmc groups in the pinctrl driver for the
MediaTek MT7981 SoC:
* emmc reset, with pin 15.
* emmc 4-bit bus-width, with pins 16 to 19, and 24 to 25.
* emmc 8-bit bus-width, with pins 16 to 25.

The existing emmc_45 group is kept for legacy reasons, even
if this is the union of emmc_reset and emmc_8 groups.

Signed-off-by: Jean Thomas <jean.thomas@wifirst.fr>
2 months agomediatek: mt7981: pinctrl: add additional uart group
Jean Thomas [Thu, 7 Dec 2023 15:16:30 +0000 (16:16 +0100)]
mediatek: mt7981: pinctrl: add additional uart group

Add uart1_3 (pins 26, 27) group to the pinctrl driver for the
MediaTek MT7981 SoC.

Signed-off-by: Jean Thomas <jean.thomas@wifirst.fr>
2 months agogeneric: mtk_eth_soc: fix PPE hanging issue
Daniel Golle [Wed, 13 Mar 2024 18:44:52 +0000 (18:44 +0000)]
generic: mtk_eth_soc: fix PPE hanging issue

A patch to resolve an issue was found in MediaTek's GPL-licensed SDK:
In the mtk_ppe_stop() function, the PPE scan mode is not disabled before
disabling the PPE. This can potentially lead to a hang during the process
of disabling the PPE.

Without this patch, the PPE may experience a hang during the reboot test.

Reference: https://git01.mediatek.com/plugins/gitiles/openwrt/feeds/mtk-openwrt-feeds/+/b40da332dfe763932a82f9f62a4709457a15dd6c

Suggested-by: Bc-bocun Chen <bc-bocun.chen@mediatek.com>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
2 months agogeneric: 5.15: mtk_eth_soc: backport fix for hang on link up
Daniel Golle [Wed, 13 Mar 2024 18:43:28 +0000 (18:43 +0000)]
generic: 5.15: mtk_eth_soc: backport fix for hang on link up

Backport commit faa5f17fe2 ("kernel: mtk_eth_soc: release
MAC_MCR_FORCE_LINK only when MAC is up") to Linux 5.15 as well.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
2 months agodnsmasq: reset PKG_RELEASE
Robert Marko [Wed, 13 Mar 2024 11:47:31 +0000 (12:47 +0100)]
dnsmasq: reset PKG_RELEASE

dnsmasq was recently updated to 2.90, but PKG_RELEASE was not reset to 1.

Fixes: 838a27f64f56 ("dnsmasq: version 2.90")
Signed-off-by: Robert Marko <robimarko@gmail.com>
2 months agoath79: add missing kernel config symbols
Shiji Yang [Wed, 13 Mar 2024 06:27:08 +0000 (14:27 +0800)]
ath79: add missing kernel config symbols

Refresh kernel configs to add the missing symbols.

Signed-off-by: Shiji Yang <yangshiji66@qq.com>
2 months agoath79: remove obsolete 5.15 kernel support
Shiji Yang [Wed, 13 Mar 2024 06:27:08 +0000 (14:27 +0800)]
ath79: remove obsolete 5.15 kernel support

The 5.15 kenel config file, patches and version switches will be
removed in this patch. We will introduce kernel 6.6 support soon.

Signed-off-by: Shiji Yang <yangshiji66@qq.com>
2 months agoath79: switch default kernel to 6.1
Shiji Yang [Wed, 13 Mar 2024 06:27:08 +0000 (14:27 +0800)]
ath79: switch default kernel to 6.1

The 6.1 kernel has been tested for over half a year.

Signed-off-by: Shiji Yang <yangshiji66@qq.com>
2 months agolibevent2: make cmake use relative imported path
Liu Dongmiao [Tue, 12 Mar 2024 06:48:08 +0000 (06:48 +0000)]
libevent2: make cmake use relative imported path

libevent2's cmake use absolute path, then cmake cannot find it when cross compiling:

```
-- Found libevent include directory: /builder/staging_dir/target-mips_24kc_musl/usr/include
-- Found libevent component: /builder/staging_dir/target-mips_24kc_musl/usr/lib/libevent_core.so
-- Found libevent component: /builder/staging_dir/target-mips_24kc_musl/usr/lib/libevent_extra.so
-- Found libevent component: /builder/staging_dir/target-mips_24kc_musl/usr/lib/libevent_openssl.so
-- Found libevent 2.1.12 in /builder/staging_dir/target-mips_24kc_musl/usr
CMake Error at /builder/staging_dir/target-mips_24kc_musl/usr/lib/cmake/libevent/LibeventTargets-shared.cmake:102 (message):
  The imported target "libevent::core" references the file

     "/usr/lib/libevent_core-2.1.so.7.0.1"

  but this file does not exist.  Possible reasons include:

  * The file was deleted, renamed, or moved to another location.

  * An install or uninstall procedure did not complete successfully.

  * The installation package was faulty and contained

     "/builder/staging_dir/target-mips_24kc_musl/usr/lib/cmake/libevent/LibeventTargets-shared.cmake"

  but not all the files it references.

Call Stack (most recent call first):
  /builder/staging_dir/target-mips_24kc_musl/usr/lib/cmake/libevent/LibeventConfig.cmake:168 (include)
  CMakeLists.txt:34 (find_package)
```

This patch make cmake use relative imported path, so it can find libevent.

Signed-off-by: Liu Dongmiao <liudongmiao@gmail.com>
2 months agoramips: add support for Z-ROUTER ZR-2660
Mikhail Zhilkin [Sat, 2 Mar 2024 19:58:40 +0000 (19:58 +0000)]
ramips: add support for Z-ROUTER ZR-2660

This commit adds support for Z-ROUTER ZR-2660 (also known as Routerich
AX1800) wireless WiFi 6 router.

Specification
-------------
- SoC       : MediaTek MT7621AT, MIPS, 880 MHz
- RAM       : 256 MiB
- Flash     : NAND 128 MiB (AMD/Spansion S34ML01G2)
- WLAN      :
  - 2.4 GHz : MediaTek MT7905D/MT7975 (14c3:7916), b/g/n/ax, MIMO 2x2
  - 5 GHz   : MediaTek MT7915E (14c3:7915), a/n/ac/ax, MIMO 2x2
- Ethernet  : 10/100/1000 Mbps x4 (1x WAN, 3x LAN)
- USB       : 1x 2.0
- UART      : 3.3V, 115200n8, pins are silkscreened on the pcb
- Buttons   : 1x Reset
- LEDs      : 1x WiFi 2.4 GHz (green)
              1x WiFi 5 GHz (green)
              1x LAN (green)
              1x WAN (green)
              1x WAN no-internet (red)
- Power     : 12 VDC, 1 A

Installation
------------
1. Run tftp server on your PC (IP: 192.168.2.2) and put OpenWrt initramfs
   image (initramfs.bin) to the tftp root dir
2. Open the following link in the browser to enable telnet:
http://192.168.2.1/cgi-bin/telnet_ssh
3. Connect to the router (default IP: 192.168.2.1) using telnet shell
   (credentials - user:admin)
4. Run the following commands in the telnet shell (this will install
   OpenWrt initramfs image on nand flash):
cd /tmp
tftp -g -r initramfs.bin 192.168.2.2
mtd write initramfs.bin firmware
mtd erase firmware_backup
reboot
5. Copy OpenWrt sysupgrade image (sysupgrade.bin) to the /tmp dir of the
   router
6. Connect to the router (IP: 192.168.1.1) using ssh shell and run
   sysupgrade command:
sysupgrade -n /tmp/sysupgrade.bin

Return to stock
---------------
1. Copy stock firmware (stock.bin) to the /tmp dir of the router using scp
2. Run following command in the router shell:
cd /tmp
mtd write stock.bin firmware
reboot

Recovery
--------
Connect uart (pins are silkscreened on the pcb), interrupt boot process by
pressing any key, use u-boot menu to flash stock firmware image or OpenWrt
initramfs image.

MAC addresses
-------------
+---------+-------------------+-----------+
|         | MAC               | Algorithm |
+---------+-------------------+-----------+
| LAN     | 24:0f:5e:xx:xx:4c | label     |
| WAN     | 24:0f:5e:xx:xx:4d | label+1   |
| WLAN 2g | 24:0f:5e:xx:xx:4e | label+2   |
| WLAN 5g | 24:0f:5e:xx:xx:4f | label+3   |
+---------+-------------------+-----------+
The WLAN 2.4 MAC was found in 'factory', 0x4
The LAN MAC was found in 'factory', 0xfff4
The WAN MAC was found in 'factory', 0xfffa

Signed-off-by: Mikhail Zhilkin <csharper2005@gmail.com>