This solution that is needed for some routers to provide proper
LED activity when controlled directly by the MV88E6xxx switch,
has just been merged in upstream Linux.
Make this patch 901 as other backports from earlier kernels
are in patch 896 and this is the first free number after
that.
Patch offsets in pending patches are augmented as part of
the refresh.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
The SPI driver framework can handle it automatically in
spi.c:spi_setup():spi_set_cs():gpiod_set_value_cansleep().
Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16636
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The ase subtarget supports two devices. ALLNET ALL0333CJ has one
Ethernet port and uses only the built-in MAC and PHY. The second
device is the Netgear DGN1000, which uses an AR8216 switch.
The driver for the adm6996 switch was probably accidentally
enabled instead of the ar8216 in commit that restores ase support
9b321bc60d ("lantiq: add Amazon-SE subtarget").
Before removing the ase subtarget, only ar8216 was enabled
c821836395 ("lantiq: remove unmaintained code").
Reduce uncompressed kernel size by 15.6 kB.
Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Link: https://github.com/openwrt/openwrt/pull/15332
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
This was done by executing these command:
$ make kernel_oldconfig CONFIG_TARGET=subtarget
Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Link: https://github.com/openwrt/openwrt/pull/15332
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Backport BLOCK OF support patch merged upstream and refresh pending
BLOCK patches.
This is a new way to declare partition table for BLOCK device (eMMC
currently supported) with the use of DTS.
Current pending patch are adapted to not cause regression with current
downstream implementation of a similar functionality.
Also enable the new OF_PARTITION config by default.
Link: https://github.com/openwrt/openwrt/pull/16663
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
The bootloader can support zImage linux kernel which can decrease
the firmware image size.
Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16662
Signed-off-by: Robert Marko <robimarko@gmail.com>
Instead of forcing 2.5G PHYs into rate-adapter mode which results higher
energy consumption, lack of support for half-duplex modes and typically
worse performance when linked at speeds less than 2.5G, use SGMII mode
which allows the MAC to follow the PHY speed.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Follow the advise of Russell King allows to greatly improve the driver
for RealTek's 1G and 2.5G Ethernet PHYs. The results are full/half
duplex as well as Gbit master/slave property being read from PHY
Specific Status Register (PHYSR), and fixes regarding link-partner
advertisement.
Fixes: #14504
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
The previous iteration of MediaTek's PHY patches caused various weird bugs.
Drop culprit patch 733-10-net-phy-mediatek-Extend-1G-TX-RX-link-pulse-time.patch
and use the most recent iteration of the patchset which has been posted to the
netdev mailing list.
Link: https://patchwork.kernel.org/project/netdevbpf/list/?series=895513&state=*
Fixes: #16448
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
mutex_destroy is not called in any error paths or in _remove. Just use
devm to do so.
Removed a pointless platform_set_devdata call. Not needed with all of
the devm conversions.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16630
Signed-off-by: Robert Marko <robimarko@gmail.com>
Also use devm for gpiochip_add_data.
Allows completely removing the _remove function.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16622
Signed-off-by: Robert Marko <robimarko@gmail.com>
Specification
-------------
- SoC : Qualcomm IPQ4019
- RAM : 256 MiB DDR3 (NT5CC128M16JR-EK)
- Flash : 64 MiB SPI NOR (Winbond W25Q512JVFQ)
- WLAN : IPQ4019 built-in
- 2.4 GHz : 2x2 MIMO WiFi4
- 5 GHz : 2x2 MIMO WiFi5
- Ethernet : QCA8075 10/100/1000 Mbps 1x WAN (ETH1, PoE); 1x LAN (ETH2)
- USB : 1x 2.0 Type-A
- UART : 3.3V, 115200n8
- Buttons : 1x Reset
- LEDs : 1x RUN (lime & red)
1x WiFi 2.4 GHz (lime)
1x WiFi 5 GHz (lime)
2x ETH (lime), controlled by the QCA8075 phy
- Power : DC 12V & 802.3at PoE
- FCC ID : 2AHKT-WIA3300-20
- TFTP IP :
- client : 192.168.18.254
- router : 192.168.18.1
Installation
------------
1. Open uart console and start TFTP server. Copy initramfs image to
the TFTP root directory and rename it to 'ipqinitramfs.bin'.
2. Power on and press 'Enter' to exit to the u-boot console according
to the TTL log prompt.
3. Execute commands to load the initramfs image:
tftpboot && bootm
4. Enter into OpenWrt to backup the partitions if you want to restore
the stock firmware one day.
5. Override default 'bootcmd' environment variable in u-boot console:
env set bootcmd 'sf probe && sf read $loadaddr 0x980000 0x800000 && bootm $loadaddr'
env save
6. Repeat step 3 and flash 'sysupgrade' image in OpenWrt.
Recovery and return to stock
----------------------------
1. Restore the backup firmware partitions in the installation step 4.
2. Restore `bootcmd` environment variable via commands:
env set bootcmd bootipq && env save
MAC addresses
-------------
+---------+-------------------+
| | MAC example |
+---------+-------------------+
| LABEL | xx:xx:xx:xx:xx:25 |
| LAN | xx:xx:xx:xx:xx:26 |
| WAN | xx:xx:xx:xx:xx:25 |
| WLAN 2g | xx:xx:xx:xx:xx:28 |
| WLAN 5g | xx:xx:xx:xx:xx:29 |
+---------+-------------------+
Notice
-----------
1. Some CH340 USB-TTL module doesn't work on this device.
2. The 'firmware' partition consists of four parts in the vendor
layout:
* Name Start Size
* rootfs 0x980000 0x1680000
* 0:HLOS1 0x2000000 0x800000
* rootfs_1 0x2800000 0x1400000
* rootfs_data 0x3c00000 0x350000
3. User can control the USB power supply via commands:
echo enabled > /sys/devices/platform/output-usb-power/state
echo disabled > /sys/devices/platform/output-usb-power/state
Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16476
Signed-off-by: Robert Marko <robimarko@gmail.com>
It can be used to control the power on/off status for the fixed
voltage regulator via sysfs.
Signed-off-by: Shiji Yang <yangshiji66@qq.com>
Link: https://github.com/openwrt/openwrt/pull/16476
Signed-off-by: Robert Marko <robimarko@gmail.com>
Even though optional is used, there's are still pointless NULL
assignments.
Use dev_err_probe to avoid manually handling EPROBE_DEFER.
Use devm_platform_iomap_resource. No struct resource needed.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Tested-by: Shiji Yang <yangshiji66@qq.com> on NETGEAR R6100.
Link: https://github.com/openwrt/openwrt/pull/16519
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
...conversion.
Commit 20736013e9 ("kernel: backport nvmem v6.6 fixes and v6.7 changes")
has causedt he device to no longer correctly read MAC address from its
onboard 24c64 EEPROM, because "at24" driver doesn't support legacy
nvmem-cell bindings [1] - and there was an explicit config option added
to mandate that behaviour in the following patch:
820-v6.7-0002-nvmem-add-explicit-config-option-to-read-old-syntax-.patch
But some of the devices, MR33 and MR74 included, weren't converted with
that as well.
Convert the definition to use proper fixed-layout binding to fix it.
The offending change was introduced between v23.05.0 and v23.05.1, and
found by bisection:
git bisect start
# status: waiting for both good and bad commits
# good: [bd4f415efa] OpenWrt v23.05.0: adjust config defaults
git bisect good bd4f415efa
# status: waiting for bad commit, 1 good commit known
# bad: [a58a86693f] OpenWrt v23.05.1: adjust config defaults
git bisect bad a58a86693f
# good: [3d0a78add2] qualcommax: only build initramfs if CONFIG_TARGET_ROOTFS_INITRAMFS is set
git bisect good 3d0a78add2
# bad: [21e5db97c4] build: add CycloneDX SBOM JSON support
git bisect bad 21e5db97c4
# good: [89184b15cf] mediatek: add build for MT7981 RFB
git bisect good 89184b15cf
# bad: [41f27bbb6d] bcm53xx: add the latest fix version of brcm_nvram
git bisect bad 41f27bbb6d
# good: [b649b0bf71] kernel: nvmem: fix "fixed-layout" & support "mac-base"
git bisect good b649b0bf71
# bad: [20736013e9] kernel: backport nvmem v6.6 fixes and v6.7 changes
git bisect bad 20736013e9
# good: [066971615f] kernel: backport v6.6 nvmem changes
git bisect good 066971615f
# first bad commit: [20736013e9] kernel: backport nvmem v6.6 fixes and v6.7 changes
Link: [1] https://github.com/openwrt/openwrt/issues/15393#issuecomment-2212300849
Fixes: 20736013e9 ("kernel: backport nvmem v6.6 fixes and v6.7 changes")
Fixes: https://github.com/openwrt/openwrt/issues/15393
Signed-off-by: Lech Perczak <lech.perczak@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16623
Signed-off-by: Robert Marko <robimarko@gmail.com>
Avoids needing to call mutex_destroy in error paths, which is missing
anyway.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16620
Signed-off-by: Robert Marko <robimarko@gmail.com>
The u-boot,env driver does not have layout in compatible.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16616
Signed-off-by: Robert Marko <robimarko@gmail.com>
This fixes the build of the gemini and the apm821xx target.
The e2fsck application returns an error code now and that makes the
build fail. The tune2fs command adds an extra option and the e2fsck
should later fix the file system. It is intentionally broken in this
place.
e2fsprogs was patched before to ignore this error.
Fixes: 95e4664b5e ("tools: e2fsprogs: drop e2fsck patch")
Link: https://github.com/openwrt/openwrt/pull/16607
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Use the absolute path to access the e2fsprogs applications. It is also
working with relative paths, but this makes sure that we use our
versions.
Link: https://github.com/openwrt/openwrt/pull/16607
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Set compatible string and specify redundant type U-boot environment.
Fixes: fea2264d9f ("ramips: mt7621: Add DNA Valokuitu Plus EX400")
Signed-off-by: Mauri Sandberg <maukka@ext.kapsi.fi>
Link: https://github.com/openwrt/openwrt/pull/16573
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Enable above drivers for mt7621 targets. They are needed for retrieving a MAC
address stored in a U-boot environment residing on an UBI volume.
Signed-off-by: Mauri Sandberg <maukka@ext.kapsi.fi>
Link: https://github.com/openwrt/openwrt/pull/16573
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
There's no specific fwnode handling here. Simpler to just use pdev.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16506
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Use devm_gpiochip_add_data to get rid of the remove function. No need
for it.
Also use dev_err_probe to simplify the error path and avoid having to
handle -EPROBE_DEFER manually.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16506
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Add CONFIG_ARCH_CONFIGURES_CPU_MITIGATIONS=y which was introduced
back in April of this year[1]
1. ce0abef6a1
Signed-off-by: John Audia <therealgraysky@proton.me>
Link: https://github.com/openwrt/openwrt/pull/16535
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Use more devm and remove _remove function.
Remove owner as the kernel sets it automatically.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16546
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Currently eth1 (which is the first "lan" interface) doesn't work on this device.
During boot the following can be seen in logs:
```
[ 2.252804] mt7530-mdio mdio-bus:1f: MT7530 adapts as multi-chip module
[ 2.266060] mtk_soc_eth 1e100000.ethernet eth0: mediatek frame engine at 0xbe100000, irq 19
[ 2.277889] mtk_soc_eth 1e100000.ethernet eth1: mediatek frame engine at 0xbe100000, irq 19
...
[ 2.355157] mt7530-mdio mdio-bus:1f: MT7530 adapts as multi-chip module
[ 2.390312] mt7530-mdio mdio-bus:1f: configuring for fixed/rgmii link mode
[ 2.398597] mt7530-mdio mdio-bus:1f: Link is Up - 1Gbps/Full - flow control rx/tx
[ 2.403872] mt7530-mdio mdio-bus:1f eth1 (uninitialized): PHY [mt7530-0:01] driver [MediaTek MT7530 PHY] (irq=21)
[ 2.416988] mtk_soc_eth 1e100000.ethernet eth0: error -17 registering interface eth1
[ 2.426973] mt7530-mdio mdio-bus:1f eth2 (uninitialized): PHY [mt7530-0:02] driver [MediaTek MT7530 PHY] (irq=22)
[ 2.440996] mt7530-mdio mdio-bus:1f eth3 (uninitialized): PHY [mt7530-0:03] driver [MediaTek MT7530 PHY] (irq=23)
[ 2.454405] mt7530-mdio mdio-bus:1f eth4 (uninitialized): PHY [mt7530-0:04] driver [MediaTek MT7530 PHY] (irq=24)
[ 2.467198] mtk_soc_eth 1e100000.ethernet eth0: entered promiscuous mode
[ 2.474117] DSA: tree 0 setup
...
[ 6.820998] mtk_soc_eth 1e100000.ethernet dsa: renamed from eth0
[ 6.919904] mtk_soc_eth 1e100000.ethernet wan: renamed from eth1
```
So the problem seems to be the fact that built-in gmacs get default
names (eth0/eth1) and are renamed after switch ports are initialized. This means
that when switch port with name `eth1` is brought up this name is still used by
gmac1 causing switch port's init to fail.
This patch just renames the ports to avoid name collision.
Note: this will break existing configs for this device because it renames all
the ports. This should not be major problem because this device doesn't have a
proper OEM image and is only flashable with serial access, meaning there should
not be many users.
Signed-off-by: Nikolay Martynov <mar.kolya@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/15865
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Hardware specification
----------------------
* RTL8380M SoC, 1 MIPS 4KEc core @ 500MHz
* 256MB DRAM
* 32MB NOR Flash
* 8 x 10/100/1000BASE-T ports
* 2 x SFP ports
* Power LED, Fault LED
* Reset button on front panel
* UART (115200 8N1) via populated standard pin header marked JP1
TODO: The SFP ports use a shared SCL GPIO that the driver cannot handle.
The left SFP port (lan9) is defined and fully functional while the laser
on the right SFP port (lan10) is off by default.
UART pinout
-----------
[o]ooo|JP1
| ||`------ GND
| |`------- RX
| `-------- TX
`---------- Vcc (3V3)
Installation using OEM webinterface
-----------------------------------
1. Make sure you are running OEM firmware in secondary slot
2. Install squashfs-factory.imag to primary slot by upload via http
Installation using serial interface
-----------------------------------
1. Press "a" "c" "p" during message "Enter correct key to stop autoboot"
2. Load image with "upgrade runtime <TFTP IP>:squashfs-sysupgrade.bin" command
3. Switch to primary slot with "setsys bootpartition 0"
4. Store config with "savesys"
5. Boot the image with `boota` command
Dual-boot with stock firmware using writable u-boot-env
-------------------------------------------------------
From stock to OpenWrt / primary image 1 (CLI as admin):
- > boot system image1
- > reboot
From OpenWrt to stock / boot image 2: (shell as root)
- # fw_setsys bootpartition 1
- # reboot
Debrick using serial interface
------------------------------
1. Press "a" "c" "p" during message "Enter correct key to stop autoboot"
2. Load vendor image with "upgrade runtime <TFTP IP>:LGS310xxxxx.imag"
3. switch to primary partition "setsys bootpartition 0"
4. safe config "savesys"
Further documentation
---------------------
See https://openwrt.org/toh/linksys/lgs352c
It has been developed and tested on device with v1 revision.
Signed-off-by: Markus Stockhausen <markus.stockhausen@gmx.de>
Link: https://github.com/openwrt/openwrt/pull/16068
[Add missing 'w' in name of firmware partition]
Signed-off-by: Sander Vanheule <sander@svanheule.net>
Backport patch that fixes memory disclosure in packet padding.
The downstream driver supports statistics, so when a packet
cannot be padded the statistics of dropped packets are incremented.
The other patches do not introduce any functional changes.
Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Link: https://github.com/openwrt/openwrt/pull/16563
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The cbr-reg DTS property uses a wrong name causing the RAC kernel
panic again on BCM6358 BCM6368 boards.
Use the correct cbr-reg name property.
Fixes: 7c9644a7b5 ("bmips: backport upstreamed RAC patches")
Signed-off-by: Daniel González Cabanelas <dgcbueu@gmail.com>
Link: https://github.com/openwrt/openwrt/pull/16561
Signed-off-by: Robert Marko <robimarko@gmail.com>
Select MT7996 2+3+3 firmware for the Adtran SmartRG SDG-8733A device
and also add it to the image for the BPi-R4 which supports both,
BE19000 (4+5+5) as well as BE14000 (2+3+3) Wi-Fi 7 modules.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
This driver does not support more than 1 queue.
Fixes: ba24b94e0b ("ramips: use more devm in mtk_eth_soc")
Signed-off-by: Rosen Penev <rosenp@gmail.com>
There are cases where an unavailable port is not an error, making
this error message a false-positive. The kernel log is flooded with
the messages like:
OF: graph: no port node found in /soc@0/bus@42000000/i2c@42530000/usb-typec@50
Silence this message by making it a debug message.
Signed-off-by: Tianling Shen <cnsztl@immortalwrt.org>
Link: https://github.com/openwrt/openwrt/pull/16524
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
This fixes some corner cases triggered by enabling fraglist GRO, where some
protocols may accidentally or intentionally linearize fraglist skbs.
Previously, these skbs became unusable and segmenting them led to crashes.
With this patch, they are properly handled by passing them to skb_segment
instead of skb_segment_list.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
