Updating the driver patches for ipq40xx (correctly) removed the
ethernet0 alias from qcom-ipq4019.dtsi; however, on some devices this
alias is needed for the bootloader to set MAC addresses in the FDT.
As it is unknown which devices actually need the alias, simply add it to
all devices trees for now that enable the &gmac now to avoid regressions
from previous OpenWrt releases. The additional alias should not cause any
issues even when it is not needed.
A TODO comment is added to the same Device Trees to document that the
alias may not be needed (hopefully preventing it from being copied
unnecessarily to newly added devices in the future). The following
devices are known to need the alias for correct MAC address assignment,
so no TODO comment is added:
- FRITZ!Box 4040
- FRITZ!Box 7530
Fixes: cd9c721124 ("ipq40xx: 6.1: use latest DSA and ethernet patches")
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
Link: https://github.com/openwrt/openwrt/pull/17442
Signed-off-by: Robert Marko <robimarko@gmail.com>
(cherry picked from commit d09dd75fbd)
This patch is needed on bmips since it fixes issues with GPIOs not being
properly configured due to gpio_request_enable not being called on bcm63xx
devices. Therefore we can now drop the bcm63268 gpio function patch.
Backported from f5b1d340be with the exception of
the realtek patch removal.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit f5b1d340be)
Basic rates were not set for mesh-interfaces, resulting in the undesired
behavior where 11s frames might be sent with a rate which was not
configured.
Depending on the driver, the basic rate might also be used to determine
the beacon rate configured to the chip. One such example are MediaTek
MT7915 platforms.
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit 3ec7f3a512)
According to datasheet, on AR9344 the switch and switch analog need to
be reset first before initiating a full reset.
Resetting these systems fixes spurious reset hangs on Atheros AR9344
SoCs.
Link: https://github.com/freifunk-gluon/gluon/issues/2904
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit 916af73fc3)
Hardware:
SoC: MT7981b
RAM: 512 MB
Flash: 256 MB SPI NAND
Ethernet:
1x2.5Gbps (rtl8221b)
WiFi: 2x2 MT7981
Button: Reset
LED: 1x multicolor
Installation
------------
At the moment, firmware installation is only possible via a transition firmware.
It's can be requested from the manufacturer by email to support@cudy.com
Signed-off-by: Maxim Anisimov <maxim.anisimov.ua@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/17225
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 6992d6e51a)
I think I implemented the U-Boot handling incorrectly on M30 (saw the issue while porting M60 to OpenWrt). Maybe someone with more U-Boot experience can have a look at it.
What I understood until now:
Before flashing, `sw_tryactive` must be set to 0 because OpenWrt runs on partition 0
During reset after flashing, U-Boot executes the following line:
`boot_rd_auto_sw_img=if itest.s ${sw_tryactive} == 2; then run boot_by_part; else run boot_by_tryactive; fi`
As `sw_tryactive` was set to 0 before flashing, `boot_by_tryactive` will be executed:
`boot_by_tryactive=if itest.s ${sw_tryactive} == 0; then setenv sw_tryactive 2; setenv sw_active 1; saveenv; run ub0; else setenv sw_tryactive 2; setenv sw_active 2; saveenv; run ub1; fi`
As `sw_tryactive` was set to 0 before flashing, `sw_active` will be set to 1 and `ub0` will be executed:
`ub0=setenv bootpart 0; mtkboardboot; run ub0to1; uip main; reset`
If the OpenWrt boot is successful, `ub0to1` and `uip` main will never be executed. Only in case OpenWrt cannot be loaded, `mtkboardboot` will return and the fallback `ub0to1` is executed.
Conclusion: It's sufficient to set `sw_tryacitve` to 0 before flashing, the added code in `target/linux/mediatek/filogic/base-files/etc/init.d/bootcount` is useless.
In the worst case (/proc/cmdline doesn't contain `bootpart=ubi0` as expected), the bootpart variable would be set to 1 and causes starting the firmware from the second partition instead of the one on the first partition.
Signed-off-by: Roland Reinl <reinlroland+github@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/17298
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 70610a5240)
Images for certain devices are staring to become too large, as some
device only have 6MB available in their vendor partition layout for the
initial install. This is especially pressing for bootloaders only
supporting gzip compression.
Drop some packages from DEFAULT_PACKAGES that aren't strictly required
for a factory install. The user can always install more packages later
using opkg/apk, or via a sysupgrade to a custom build.
firewall4 is kept to ensure the most recent firewall package is selected
in builds including LuCI.
ethtool is kept as a frequently used diagnostics tool.
Link: https://github.com/openwrt/openwrt/pull/17450
Signed-off-by: Sander Vanheule <sander@svanheule.net>
(cherry picked from commit c9ae39b2d1)
Just one of the devices uses the Marvell MV88E6060 DSA
switch so break this out from the generic kernel config
and into a package selected only by that single device
and probed at boot instead.
The big win is from being able to drop the dsa_core
(~600KB) kernel module out of the common kernel on
devices with no DSA switch.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
(cherry picked from commit 2456a2fd7f)
On the AVM 5490/5491, lan1, lan2 and wan ports are connected directly
to the internal GSWIP switch. The lan3 and lan4 ports are connected via
an external QCA8334 switch. This commit adds the missing entries in dts
and adds the driver module.
Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Link: https://github.com/openwrt/openwrt/pull/17473
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 14be320291)
These two packages are SDXC drivers for Mediatek mt762x series SoCs.
One is upstream implementation, and the other is downstream driver.
Installing them together will result in conflicts.
Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/17446
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 0764e30082)
There are only 5 devices in mt76x8 sub-target selected the MTK SDXC
driver package. And they are all single ethernet port routers or dev
boards:
* LinkIt Smart 7688
* Onion Omega2+
* RAVPower RP-WD009
* VoCore VoCore2
* VoCore VoCore2-Lite
For these devices, they are using the ephy p1 - p4 as the SDXC IO
pins. Therefore, these GPIO pads must be configured in "digital"
IO mode.
Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/17446
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit f5996ae947)
In the MTK vendor driver, mt762x SDXC registers MSDC_PATCH_BIT and
MSDC_PATCH_BIT1 have different init values than upstream driver.
These magical values should have some help for the stability.
Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/17446
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit f70cdfd682)
Correct a few mistakes around dependencies and naming and unset
CONFIG_FIRMWARE_RP1 in RPi5B's config and instead of a builtin, build it
as a module.
Without this change, there are two entries for rp1.ko in
/lib/modules/$(uname -r)/modules.builtin due to how we strip the leading
directories when we generate it. See: package/kernel/linux/Makefile
around line 63.
% grep rp1.ko /lib/modules/$(uname -r)/modules.builtin
pwm-rp1.ko
clk-rp1.ko
rp1.ko
rp1.ko
The kernel log gets spammed with tons of superfluous warnings as a
results of the double entry:
daemon.warn modprobe: found duplicate builtin module rp1
Signed-off-by: John Audia <therealgraysky@proton.me>
Link: https://github.com/openwrt/openwrt/pull/17461
Signed-off-by: Robert Marko <robimarko@gmail.com>
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit f105d1a9a9)
On some drivers, setting the tx power on the interface is not enough.
Set it for the phy as well.
Fixes: 04fb05914e ("wifi-scripts: add multi-radio config support")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry picked from commit b795e5cbcf)
35bcf68a6297 wifi: mt76: scan: fix setting tx_info fields
4d8d6e2e7710 wifi: mt76: scan: set vif offchannel link for scanning/roc
e354436db440 wifi: mt76: mt7996: use the correct vif link for scanning/roc
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry picked from commit f11bc4f201)
7a3bcd39ae1f r8169: use helper r8169_mod_reg8_cond to simplify rtl_jumbo_config
e3e9e9039fa6 r8169: align WAKE_PHY handling with r8125/r8126 vendor drivers
330dc2297c82 r8169: improve rtl_set_d3_pll_down
c507e96b5763 r8169: improve __rtl8169_set_wol
83cb4b470c66 r8169: remove leftover locks after reverted change
2cd02f2fdd8a r8169: improve initialization of RSS registers on RTL8125/RTL8126
a3d8520e6a19 r8169: align RTL8126 EEE config with vendor driver
4af2f60bf737 r8169: align RTL8125/RTL8126 PHY config with vendor driver
eb90f876b796 r8169: align RTL8125 EEE config with vendor driver
b8bd8c44a266 r8169: fix inconsistent indenting in rtl8169_get_eth_mac_stats
f75d1fbe7809 r8169: add support for RTL8125D
c4e64095c00c r8169: enable EEE at 2.5G per default on RTL8125B
d64113c6bb5e r8169: remove rtl_dash_loop_wait_high/low
1c105bacb160 r8169: avoid duplicated messages if loading firmware fails and switch to warn level
ac48430368c1 r8169: don't take RTNL lock in rtl_task()
e3fc5139bd8f r8169: implement additional ethtool stats ops
b8bf38440ba9 r8169: enable SG/TSO on selected chip versions per default
854d71c555df r8169: remove original workaround for RTL8125 broken rx issue
1ffcc8d41306 r8169: add support for the temperature sensor being available from RTL8125B
The following patches require backporting additional linux patches:
e2015942e90a r8169: replace custom flag with disable_work() et al
e340bff27e63 r8169: copy vendor driver 2.5G/5G EEE advertisement constraints
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
(cherry picked from commit 0eeba04a16)
The ARIA block cipher is pretty uncommon in TLS, deactivate it for now.
This saves some space and reduces the possible variations and attack
vectors of mbedtls.
ARIA support was deactivated in OpenWrt 23.05 by default.
Link: https://github.com/openwrt/openwrt/pull/17342
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit 3c0ef48bc8)
The commit 7160820d742a ("phy: rockchip: naneng-combphy: fix phy reset")
was backported to kernel 6.6 branch by upstream, however the correspond
dtsi fixes was not, resulting the following error:
```
[ 0.225521] rockchip-naneng-combphy fe830000.phy: error -ENOENT: failed to get phy reset
[ 0.227467] rockchip-naneng-combphy fe840000.phy: error -ENOENT: failed to get phy reset
```
So backport the dtsi fixes here manually.
Fixes: 89b2356b8c ("kernel: bump 6.6 to 6.6.69")
Signed-off-by: Tianling Shen <cnsztl@immortalwrt.org>
Link: https://github.com/openwrt/openwrt/pull/17468
Signed-off-by: Robert Marko <robimarko@gmail.com>
(cherry picked from commit 8a477bafb4)
defer_list skbs held by NAPI can block releasing page pools.
Work around this by scheduling rx softirq on all CPUs while trying to release
a page pool.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry picked from commit 2b70b32aef)
d996988ae55b libubus: close file descriptor after sending it from a request
afa57cce0aff libubus: add support for using channels
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry picked from commit f0df6e3a4a)
It seems that we have some kind of a symbol name conflict which causes
CONFIG_SECCOMP to always be read as y.
Unfortunatelly, I could not figure out what is causing this, but simply
renaming SECCOMP to USE_SECCOMP seems to properly work and leaves the
symbol unset unless arch dependencies are satisfied.
This fixes qoriq and others that dont support seccomp from failing due
to procd-seccomp package being selected to get included but it cannot be
built for them:
ERROR: unable to select packages:
procd-seccomp (no such package):
required by: base-files-1637~52b6c92479[procd-seccomp]
Fixes: 4c65359af4 ("build: fix including busybox, procd and apk/opkg in imagebuilder")
Link: https://github.com/openwrt/openwrt/pull/17048
Signed-off-by: Robert Marko <robimarko@gmail.com>
(cherry picked from commit a48ec449cc)
Robert reported, that in firmware images generated by ASU, there is
`apk` package manager missing after the commit 44598c233d ("build:
remove broken dependency of metadata on toplevel .config variables").
That is happening, because apk got removed from `default_packages` list in
`profiles.json`, which is being generated by `json_overview_image_info` Make
target, which uses `scripts/json_overview_image_info.py` helper script,
which gets the information from `DEFAULT_PACKAGES` Make variable.
So lets fix it by providing `DEFAULT_PACKAGES` variable when its needed.
The reason why we didn't added those packages as a dependency to
base-files like any other packages, was to allow disabling them (in
order to save space).
Fixes: #16969
Fixes: openwrt/asu/issues/1084
Fixes: 44598c233d ("build: remove broken dependency of metadata on toplevel .config variables")
Reported-by: Robert Marko <robimarko@gmail.com>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Link: https://github.com/openwrt/openwrt/pull/16986
Signed-off-by: Robert Marko <robimarko@gmail.com>
(cherry picked from commit 90f0be8521)
Since the image builder pulls package lists from metadata directly,
add procd and busybox as depdendencies to base-files.
As for the package manager itself, since it can be disabled it needs
to be added directly in the image builder makefile
Fixes: 44598c233d ("build: remove broken dependency of metadata on toplevel .config variables")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry picked from commit 4c65359af4)
Instead of relying on .config symbols for metadata, alter the DEFAULT
variable of affected packages. Fixes enabling opkg vs apk among others.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry picked from commit 44598c233d)
