Drop custom 'mtd-cal-data' and switch to 'nvmem-cells' based solution
for fetching radio calibration data and its MAC address.
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
All the QCA9531 based boards from ALFA Network are based on the same
design and share a common DTSI: 'qca9531_alfa-network_r36a.dtsi'.
Instead of defining 'nvmem-cells' for the MAC address in every device's
DTS, move definition to the common DTSI file.
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
Bump the last missing target to Kernel 5.10. While this requires a work
around to boot it will allow more people to test the new Kernel before
the upcomming release.
Signed-off-by: Paul Spooren <mail@aparcar.org>
This is a workaround to make the target overall bootable. With this more
people should be able to test the Kernel 5.10 and report further issues.
Suggested-by: Daniel González Cabanelas <dgcbueu@gmail.com>
Signed-off-by: Paul Spooren <mail@aparcar.org>
Add support for the TP-Link EAP615-Wall, an AX1800 Wall Plate WiFi 6 AP.
The device is very similar to the TP-Link EAP235-Wall.
Hardware:
* SoC: MediaTek MT7621AT
* RAM: 128MiB
* Flash: 16MiB SPI-NOR
* Ethernet: 4x GbE
* Back: ETH0 (PoE-PD)
* Bottom: ETH1, ETH2, ETH3 (PoE passthrough)
* WiFi: MT7905DAN/MT7975DN 2.4/5 GHz 2T2R
* LEDS: 1x white
* Buttons: 1x LED, 1x reset
Stock firmware uses a random MAC address for ethernet. OpenWrt uses the
MAC address that is on the device label for ethernet and the wireless
interfaces. MAC address must not be incremented, as this will cause MAC
address conflicts in case you have two devices with consecutive MAC
addresses. Instead, different locally administered addresses will be
generated automatically, based on the MAC on the label.
Installation via stock firmware:
* Enable SSH in the TP-Link web interface
* SSH to the device
* Run `cliclientd stopcs`
* Upload the OpenWrt factory image via the TP-Link web interface
Installation via bootloader:
* Solder TTL header. Pinout: 1: TX, 2: RX, 3: GND, 4: VCC, with pin 1
closest to ETH1. Baud rate 115200
* Interrupt boot process by holding a key during boot
* Boot the OpenWrt initramfs:
# tftpboot 0x84000000 openwrt-ramips-mt7621-tplink_eap615-wall-v1-initramfs-kernel.bin
# bootm
* Copy openwrt-ramips-mt7621-tplink_eap615-wall-v1-squashfs-sysupgrade.bin
to /tmp and use sysupgrade to install it
Thanks to Sander Vanheule for his work on the EAP235-Wall, which made
adding support for the EAP615-Wall very easy.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
Reviewed-by: Sander Vanheule <sander@svanheule.net>
Acked-by: Arınç ÜNAL <arinc.unal@arinc9.com>
Remove PM debug features from sama5 kernel config. It is not
necessary to have it on production code. This also fixes the
build for sama5 target after commit 97158fe10e ("kernel:
package ramoops pstore-ram crash log storage)
Fixes: 97158fe10e ("kernel: package ramoops pstore-ram crash log storage")
Signed-off-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Increase the kernel size from 3 MB to 4 MB for EA8500 and EA7500v1.
* modify the common .dtsi
* modify the kernel size in the image recipes
Define compat-version 2.0 to force factory image usage for sysupgrade.
Add explanation message. Reenable both devices.
As for 4MiB (and not more): Hannu Nyman noted that:
"We have lots of ipq806x devices with 4 MB kernel, so will
need action at that point in future in any case.
(Assuming that the bootloader did not have a 4 MB limit that
has been tested...)"
Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
(squashed, added 4MiB notice of support in ipq806x)
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
ZTE MF286A and MF286R are indoor LTE category 6/7 CPE router with simultaneous
dual-band 802.11ac plus 802.11n Wi-Fi radios and quad-port gigabit
Ethernet switch, FXS and external USB 2.0 port.
Hardware highlights:
- CPU: QCA9563 SoC at 775MHz,
- RAM: 128MB DDR2,
- NOR Flash: MX25L1606E 2MB SPI Flash, for U-boot only,
- NAND Flash: W25N01GV 128MB SPI NAND-Flash, for all other data,
- Wi-Fi 5GHz: QCA9886 2x2 MIMO 802.11ac Wave2 radio,
- WI-Fi 2.4GHz: QCA9563 3x3 MIMO 802.11n radio,
- Switch: QCA8337v2 4-port gigabit Ethernet, with single SGMII CPU port,
- WWAN:
[MF286A] MDM9230-based category 6 internal LTE modem
[MF286R] PXA1826-based category 7 internal LTE modem
in extended mini-PCIE form factor, with 3 internal antennas and
2 external antenna connections, single mini-SIM slot.
- FXS: one external ATA port (handled entirely by modem part) with two
physical connections in parallel,
- USB: Single external USB 2.0 port,
- Switches: power switch, WPS, Wi-Fi and reset buttons,
- LEDs: Wi-Fi, Test (internal). Rest of LEDs (Phone, WWAN, Battery,
Signal state) handled entirely by modem. 4 link status LEDs handled by
the switch on the backside.
- Battery: 3Ah 1-cell Li-Ion replaceable battery, with charging and
monitoring handled by modem.
- Label MAC device: eth0
The device shares many components with previous model, MF286, differing
mostly by a Wave2 5GHz radio, flash layout and internal LED color.
In case of MF286A, the modem is the same as in MF286. MF286R uses a
different modem based on Marvell PXA1826 chip.
Internal modem of MF286A is supported via uqmi, MF286R modem isn't fully
supported, but it is expected to use comgt-ncm for connection, as it
uses standard 3GPP AT commands for connection establishment.
Console connection: connector X2 is the console port, with the following
pinout, starting from pin 1, which is the topmost pin when the board is
upright:
- VCC (3.3V). Do not use unless you need to source power for the
converer from it.
- TX
- RX
- GND
Default port configuration in U-boot as well as in stock firmware is
115200-8-N-1.
Installation:
Due to different flash layout from stock firmware, sysupgrade from
within stock firmware is impossible, despite it's based on QSDK which
itself is based on OpenWrt.
STEP 0: Stock firmware update:
As installing OpenWrt cuts you off from official firmware updates for
the modem part, it is recommended to update the stock firmware to latest
version before installation, to have built-in modem at the latest firmware
version.
STEP 1: gaining root shell:
Method 1:
This works if busybox has telnetd compiled in the binary.
If this does not work, try method 2.
Using well-known exploit to start telnetd on your router - works
only if Busybox on stock firmware has telnetd included:
- Open stock firmware web interface
- Navigate to "URL filtering" section by going to "Advanced settings",
then "Firewall" and finally "URL filter".
- Add an entry ending with "&&telnetd&&", for example
"http://hostname/&&telnetd&&".
- telnetd will immediately listen on port 4719.
- After connecting to telnetd use "admin/admin" as credentials.
Method 2:
This works if busybox does not have telnetd compiled in. Notably, this
is the case in DNA.fi firmware.
If this does not work, try method 3.
- Set IP of your computer to 192.168.0.22. (or appropriate subnet if
changed)
- Have a TFTP server running at that address
- Download MIPS build of busybox including telnetd, for example from:
https://busybox.net/downloads/binaries/1.21.1/busybox-mips
and put it in it's root directory. Rename it as "telnetd".
- As previously, login to router's web UI and navigate to "URL
filtering"
- Using "Inspect" feature, extend "maxlength" property of the input
field named "addURLFilter", so it looks like this:
<input type="text" name="addURLFilter" id="addURLFilter" maxlength="332"
class="required form-control">
- Stay on the page - do not navigate anywhere
- Enter "http://aa&zte_debug.sh 192.168.0.22 telnetd" as a filter.
- Save the settings. This will download the telnetd binary over tftp and
execute it. You should be able to log in at port 23, using
"admin/admin" as credentials.
Method 3:
If the above doesn't work, use the serial console - it exposes root shell
directly without need for login. Some stock firmwares, notably one from
finnish DNA operator lack telnetd in their builds.
STEP 2: Backing up original software:
As the stock firmware may be customized by the carrier and is not
officially available in the Internet, IT IS IMPERATIVE to back up the
stock firmware, if you ever plan to returning to stock firmware.
It is highly recommended to perform backup using both methods, to avoid
hassle of reassembling firmware images in future, if a restore is
needed.
Method 1: after booting OpenWrt initramfs image via TFTP:
PLEASE NOTE: YOU CANNOT DO THIS IF USING INTERMEDIATE FIRMWARE FOR INSTALLATION.
- Dump stock firmware located on stock kernel and ubi partitions:
ssh root@192.168.1.1: cat /dev/mtd4 > mtd4_kernel.bin
ssh root@192.168.1.1: cat /dev/mtd9 > mtd9_ubi.bin
And keep them in a safe place, should a restore be needed in future.
Method 2: using stock firmware:
- Connect an external USB drive formatted with FAT or ext4 to the USB
port.
- The drive will be auto-mounted to /var/usb_disk
- Check the flash layout of the device:
cat /proc/mtd
It should show the following:
mtd0: 000a0000 00010000 "u-boot"
mtd1: 00020000 00010000 "u-boot-env"
mtd2: 00140000 00010000 "reserved1"
mtd3: 000a0000 00020000 "fota-flag"
mtd4: 00080000 00020000 "art"
mtd5: 00080000 00020000 "mac"
mtd6: 000c0000 00020000 "reserved2"
mtd7: 00400000 00020000 "cfg-param"
mtd8: 00400000 00020000 "log"
mtd9: 000a0000 00020000 "oops"
mtd10: 00500000 00020000 "reserved3"
mtd11: 00800000 00020000 "web"
mtd12: 00300000 00020000 "kernel"
mtd13: 01a00000 00020000 "rootfs"
mtd14: 01900000 00020000 "data"
mtd15: 03200000 00020000 "fota"
mtd16: 01d00000 00020000 "firmware"
Differences might indicate that this is NOT a MF286A device but
one of other variants.
- Copy over all MTD partitions, for example by executing the following:
for i in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15; do cat /dev/mtd$i > \
/var/usb_disk/mtd$i; done
"Firmware" partition can be skipped, it is a concatenation
of "kernel" and "rootfs".
- If the count of MTD partitions is different, this might indicate that
this is not a MF286A device, but one of its other variants.
- (optionally) rename the files according to MTD partition names from
/proc/mtd
- Unmount the filesystem:
umount /var/usb_disk; sync
and then remove the drive.
- Store the files in safe place if you ever plan to return to stock
firmware. This is especially important, because stock firmware for
this device is not available officially, and is usually customized by
the mobile providers.
STEP 3: Booting initramfs image:
Method 1: using serial console (RECOMMENDED):
- Have TFTP server running, exposing the OpenWrt initramfs image, and
set your computer's IP address as 192.168.0.22. This is the default
expected by U-boot. You may wish to change that, and alter later
commands accordingly.
- Connect the serial console if you haven't done so already,
- Interrupt boot sequence by pressing any key in U-boot when prompted
- Use the following commands to boot OpenWrt initramfs through TFTP:
setenv serverip 192.168.0.22
setenv ipaddr 192.168.0.1
tftpboot 0x81000000 openwrt-ath79-nand-zte_mf286a-initramfs-kernel.bin
bootm 0x81000000
(Replace server IP and router IP as needed). There is no emergency
TFTP boot sequence triggered by buttons, contrary to MF283+.
- When OpenWrt initramfs finishes booting, proceed to actual
installation.
Method 2: using initramfs image as temporary boot kernel
This exploits the fact, that kernel and rootfs MTD devices are
consecutive on NAND flash, so from within stock image, an initramfs can
be written to this area and booted by U-boot on next reboot, because it
uses "nboot" command which isn't limited by kernel partition size.
- Download the initramfs-kernel.bin image
- After backing up the previous MTD contents, write the images to the
"firmware" MTD device, which conveniently concatenates "kernel" and
"rootfs" partitions that can fit the initramfs image:
nandwrite -p /dev/<firmware-mtd> \
/var/usb_disk/openwrt-ath79-zte_mf286a-initramfs-kernel.bin
- If write is OK, reboot the device, it will reboot to OpenWrt
initramfs:
reboot -f
- After rebooting, SSH into the device and use sysupgrade to perform
proper installation.
Method 3: using built-in TFTP recovery (LAST RESORT):
- With that method, ensure you have complete backup of system's NAND
flash first. It involves deliberately erasing the kernel.
- Download "-initramfs-kernel.bin" image for the device.
- Prepare the recovery image by prepending 8MB of zeroes to the image,
and name it root_uImage:
dd if=/dev/zero of=padding.bin bs=8M count=1
cat padding.bin openwrt-ath79-nand-zte_mf286a-initramfs-kernel.bin >
root_uImage
- Set up a TFTP server at 192.0.0.1/8. Router will use random address
from that range.
- Put the previously generated "root_uImage" into TFTP server root
directory.
- Deliberately erase "kernel" partition" using stock firmware after
taking backup. THIS IS POINT OF NO RETURN.
- Restart the device. U-boot will attempt flashing the recovery
initramfs image, which will let you perform actual installation using
sysupgrade. This might take a considerable time, sometimes the router
doesn't establish Ethernet link properly right after booting. Be
patient.
- After U-boot finishes flashing, the LEDs of switch ports will all
light up. At this moment, perform power-on reset, and wait for OpenWrt
initramfs to finish booting. Then proceed to actual installation.
STEP 4: Actual installation:
- Set your computer IP to 192.168.1.22/24
- scp the sysupgrade image to the device:
scp openwrt-ath79-nand-zte_mf286a-squashfs-sysupgrade.bin \
root@192.168.1.1:/tmp/
- ssh into the device and execute sysupgrade:
sysupgrade -n /tmp/openwrt-ath79-nand-zte_mf286a-squashfs-sysupgrade.bin
- Wait for router to reboot to full OpenWrt.
STEP 5: WAN connection establishment
Since the router is equipped with LTE modem as its main WAN interface, it
might be useful to connect to the Internet right away after
installation. To do so, please put the following entries in
/etc/config/network, replacing the specific configuration entries with
one needed for your ISP:
config interface 'wan'
option proto 'qmi'
option device '/dev/cdc-wdm0'
option auth '<auth>' # As required, usually 'none'
option pincode '<pin>' # If required by SIM
option apn '<apn>' # As required by ISP
option pdptype '<pdp>' # Typically 'ipv4', or 'ipv4v6' or 'ipv6'
For example, the following works for most polish ISPs
config interface 'wan'
option proto 'qmi'
option device '/dev/cdc-wdm0'
option auth 'none'
option apn 'internet'
option pdptype 'ipv4'
The required minimum is:
config interface 'wan'
option proto 'qmi'
option device '/dev/cdc-wdm0'
In this case, the modem will use last configured APN from stock
firmware - this should work out of the box, unless your SIM requires
PIN which can't be switched off.
If you have build with LuCI, installing luci-proto-qmi helps with this
task.
Restoring the stock firmware:
Preparation:
If you took your backup using stock firmware, you will need to
reassemble the partitions into images to be restored onto the flash. The
layout might differ from ISP to ISP, this example is based on generic stock
firmware
The only partitions you really care about are "web", "kernel", and
"rootfs". These are required to restore the stock firmware through
factory TFTP recovery.
Because kernel partition was enlarged, compared to stock
firmware, the kernel and rootfs MTDs don't align anymore, and you need
to carve out required data if you only have backup from stock FW:
- Prepare kernel image
cat mtd12_kernel.bin mtd13_rootfs.bin > owrt_kernel.bin
truncate -s 4M owrt_kernel_restore.bin
- Cut off first 1MB from rootfs
dd if=mtd13_rootfs.bin of=owrt_rootfs.bin bs=1M skip=1
- Prepare image to write to "ubi" meta-partition:
cat mtd6_reserved2.bi mtd7_cfg-param.bin mtd8_log.bin mtd9_oops.bin \
mtd10_reserved3.bin mtd11_web.bin owrt_rootfs.bin > \
owrt_ubi_ubi_restore.bin
You can skip the "fota" partition altogether,
it is used only for stock firmware update purposes and can be overwritten
safely anyway. The same is true for "data" partition which on my device
was found to be unused at all. Restoring mtd5_cfg-param.bin will restore
the stock firmware configuration you had before.
Method 1: Using initramfs:
This method is recmmended if you took your backup from within OpenWrt
initramfs, as the reassembly is not needed.
- Boot to initramfs as in step 3:
- Completely detach ubi0 partition using ubidetach /dev/ubi0_0
- Look up the kernel and ubi partitions in /proc/mtd
- Copy over the stock kernel image using scp to /tmp
- Erase kernel and restore stock kernel:
(scp mtd4_kernel.bin root@192.168.1.1:/tmp/)
mtd write <kernel_mtd> mtd4_kernel.bin
rm mtd4_kernel.bin
- Copy over the stock partition backups one-by-one using scp to /tmp, and
restore them individually. Otherwise you might run out of space in
tmpfs:
(scp mtd3_ubiconcat0.bin root@192.168.1.1:/tmp/)
mtd write <ubiconcat0_mtd> mtd3_ubiconcat0.bin
rm mtd3_ubiconcat0.bin
(scp mtd5_ubiconcat1.bin root@192.168.1.1:/tmp/)
mtd write <ubiconcat1_mtd> mtd5_ubiconcat1.bin
rm mtd5_ubiconcat1.bin
- If the write was correct, force a device reboot with
reboot -f
Method 2: Using live OpenWrt system (NOT RECOMMENDED):
- Prepare a USB flash drive contatining MTD backup files
- Ensure you have kmod-usb-storage and filesystem driver installed for
your drive
- Mount your flash drive
mkdir /tmp/usb
mount /dev/sda1 /tmp/usb
- Remount your UBI volume at /overlay to R/O
mount -o remount,ro /overlay
- Write back the kernel and ubi partitions from USB drive
cd /tmp/usb
mtd write mtd4_kernel.bin /dev/<kernel_mtd>
mtd write mtd9_ubi.bin /dev/<kernel_ubi>
- If everything went well, force a device reboot with
reboot -f
Last image may be truncated a bit due to lack of space in RAM, but this will happen over "fota"
MTD partition which may be safely erased after reboot anyway.
Method 3: using built-in TFTP recovery:
This method is recommended if you took backups using stock firmware.
- Assemble a recovery rootfs image from backup of stock partitions by
concatenating "web", "kernel", "rootfs" images dumped from the device,
as "root_uImage"
- Use it in place of "root_uImage" recovery initramfs image as in the
TFTP pre-installation method.
Quirks and known issuesa
- It was observed, that CH340-based USB-UART converters output garbage
during U-boot phase of system boot. At least CP2102 is known to work
properly.
- Kernel partition size is increased to 4MB compared to stock 3MB, to
accomodate future kernel updates - at this moment OpenWrt 5.10 kernel
image is at 2.5MB which is dangerously close to the limit. This has no
effect on booting the system - but keep that in mind when reassembling
an image to restore stock firmware.
- uqmi seems to be unable to change APN manually, so please use the one
you used before in stock firmware first. If you need to change it,
please use protocok '3g' to establish connection once, or use the
following command to change APN (and optionally IP type) manually:
echo -ne 'AT+CGDCONT=1,"IP","<apn>' > /dev/ttyUSB0
- The only usable LED as a "system LED" is the blue debug LED hidden
inside the case. All other LEDs are controlled by modem, on which the
router part has some influence only on Wi-Fi LED.
- Wi-Fi LED currently doesn't work while under OpenWrt, despite having
correct GPIO mapping. All other LEDs are controlled by modem,
including this one in stock firmware. GPIO19, mapped there only acts
as a gate, while the actual signal source seems to be 5GHz Wi-Fi
radio, however it seems it is not the LED exposed by ath10k as
ath10k-phy0.
- GPIO5 used for modem reset is a suicide switch, causing a hardware
reset of whole board, not only the modem. It is attached to
gpio-restart driver, to restart the modem on reboot as well, to ensure
QMI connectivity after reboot, which tends to fail otherwise.
- Modem, as in MF283+, exposes root shell over ADB - while not needed
for OpenWrt operation at all - have fun lurking around.
The same modem module is used as in older MF286.
Signed-off-by: Lech Perczak <lech.perczak@gmail.com>
Add the missing pinctrl properties on the ethernet node.
GMAC1 will start working with this change.
Link: https://lore.kernel.org/netdev/83a35aa3-6cb8-2bc4-2ff4-64278bbcd8c8@arinc9.com/
Overwrite pinctrl-0 property without rgmii2_pins on devicetrees which use
the rgmii2 pins as GPIO (22 - 33).
Give gpio function to rgmii2 pin group on mt7621_tplink_archer-x6-v3.dtsi
which uses GPIO 28.
Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
Flow control needs to be enabled on both sides to work.
It is already enabled on gmac0, enable it on port@6 too.
Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
Tested-by: Sungbo Eo <mans0n@gorani.run>
Remove reg property from ports node to fix this warning:
Warning (unit_address_vs_reg): /ethernet@1e100000/mdio-bus/switch@1f/ports: node has a reg or ranges property, but no unit name
Another warning surfaces afterwards. Remove #address-cells and #size-cells
from switch@1f node to fix this warning:
Warning (avoid_unnecessary_addr_size): /ethernet@1e100000/mdio-bus/switch@1f: unnecessary #address-cells/#size-cells without "ranges" or child "reg" property
Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
The Zyxel EMG2926-Q10A is 99% the Zyxel NBG6716, but the bootloader
expects a different product name when flashing over TFTP. Also, the
EMG2926-Q10A always has 128 MiB of NAND flash whereas the NBG6716
reportedly can have either 128 MiB or 256 MiB.
Signed-off-by: Alex Henrie <alexhenrie24@gmail.com>
The Sagem/Plusnet F@ST2704N has a red label in ethernet port 4. Its purpose is
to be used as Fibre/WAN with the stock firmware.
Configure the Eth4 as WAN.
Fixes: fbbb977772 (brcm63xx: Tune the network configuration for several
routers)
Signed-off-by: Daniel González Cabanelas <dgcbueu@gmail.com>
the Aerohive HiveAP-330 and HiveAP-350 come equipped
with an TI TMP125 temperature chip. This patch wires
up the necessary support for this sensor and exposes
it through hwmon / thermal sensor framework. Upstream
support is coming, but it has to go through hwmon-next
first.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
The last remaining 5.4 target currently chokes because the
symbols haven't been disabled like for 5.10.
Fixes: 97158fe10e ("kernel: package ramoops pstore-ram crash log storage")
Reported-by: Hannu Nyman <hannu.nyman@iki.fi>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Add the following kconfig symbols (disabled):
CONFIG_DEFAULT_FQ
CONFIG_DEFAULT_CODEL
CONFIG_DEFAULT_SFQ
Also resort the config with the kconfig.pl script.
Fixes: f39872d966 ("kernel: generic: select the fq_codel qdisc by default")
Tested-by: Christian Lamparter <chunkeey@gmail.com>
Signed-off-by: Rui Salvaterra <rsalvaterra@gmail.com>
Define the kernel crash log storage ramoops/pstore feature
for R7800 and its sister XR500.
Reference to the ramoops admin guide in upstream Linux:
https://www.kernel.org/doc/html/v5.10/admin-guide/ramoops.html
Tested with R7800.
Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
Package the ability to log kernel crashes to 'ramoops' pstore
files into RAM in /sys/fs/pstore
Reference to the ramoops admin guide in upstream Linux:
https://www.kernel.org/doc/html/v5.10/admin-guide/ramoops.html
The files in RAM survive a warm reboot, but not a cold reboot.
Note: kmod-ramoops selects kmod-pstore and kmod-reed-solomon.
The feature can be used by selecting the kmod-ramoops and
adding a ramoops reserved-memory definition to the device DTS.
Example from R7800:
reserved-memory {
rsvd@5fe00000 {
reg = <0x5fe00000 0x200000>;
reusable;
};
ramoops@42100000 {
compatible = "ramoops";
reg = <0x42100000 0x40000>;
record-size = <0x4000>;
console-size = <0x4000>;
ftrace-size = <0x4000>;
pmsg-size = <0x4000>;
};
};
If no definition has been made in DTS, no crash log is stored
for the device.
Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
(added CONFIG_EFI_VARS_PSTORE disable)
Previously, grub2 was hardcoded to always look on "hd0" for the
kernel.
This works well when the system only had a single disk.
But if there was a second disk/stick present, it may have look
on the wrong drive because of enumeration races.
This patch utilizes grub2 search function to look for a filesystem
with the label "kernel". This works thanks to existing setup in
scripts/gen_image_generic.sh. Which sets the "kernel" label on
both the fat and ext4 filesystem variants.
Signed-off-by: Jax Jiang <jax.jiang.007@gmail.com>
Suggested-by: Alberto Bursi <bobafetthotmail@gmail.com> (MX100 WA)
(word wrapped, slightly rewritten commit message, removed MX100 WA)
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
This reverts all four commits
dbb45421ba "bcm27xx: bcm2708: update defconfig"
332f69583a "bcm27xx: bcm2709: update defconfig"
a478202d74 "bcm27xx: bcm2710: update defconfig"
82da1dfd69 "bcm27xx: bcm2711: update defconfig"
this also highlighted an unrelated kconfig failure
that warrants investigation. But for now it is important
for the bcm27xx target to come back again.
|*
|* Restart config...
|*
|*
|* Allow override default queue discipline
|*
|Allow override default queue discipline (NET_SCH_DEFAULT) [Y/n/?] y
| Default queuing discipline
| 1. Fair Queue (DEFAULT_FQ) (NEW)
| 2. Controlled Delay (DEFAULT_CODEL) (NEW)
| > 3. Fair Queue Controlled Delay (DEFAULT_FQ_CODEL)
| 4. Stochastic Fair Queue (DEFAULT_SFQ) (NEW)
| 5. Priority FIFO Fast (DEFAULT_PFIFO_FAST)
| choice[1-5?]:
|Error in reading or end of file.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Add kernel support for SAMA7G5 by back-porting mainline kernel patches.
Among SAMA7G5 features could be remembered:
- ARM Cortex-A7
- double data rate multi-port dynamic RAM controller supporting DDR2,
DDR3, DDR3L, LPDDR2, LPDDR3 up to 533MHz
- peripherals for audio, video processing
- 1 gigabit + 1 megabit Ethernet controllers
- 6 CAN controllers
- trust zone support
- DVFS for CPU
- criptography IPs
Signed-off-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Adding the feature flag automatically creates a a rootfs.tar.gz files
which can be used for Docker rootfs containers.
Signed-off-by: Paul Spooren <mail@aparcar.org>
fwtool is now always part of the sysupgrade stage2 ramdisk, so drop
the no longer needed RAMFS_COPY_BIN variable.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Now that both, fw_printenv/fw_setenv and fwtool are always present
during stage2 sysupgrade, we no longer need to list them in
RAMFS_COPY_BIN and RAMFS_COPY_DATA in platform.sh.
Drop both variables as they are now unneeded.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
The 'BOARDNAME' variable is part of target configuration and shouldn't
be part of a device's image recipe.
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
The I/O base address for the timers was hardcoded into the driver,
or derived from the HW IRQ number as an even more horrible hack. All
supported SoC families have these timers, but with hardcoded addresses
the code cannot be reused right now.
Request the timer's base address from the DT specification, and store it
in a private struct for future reference.
Matching the second interrupt specifier, the address range for the
second timer is added to the DT specification.
Signed-off-by: Sander Vanheule <sander@svanheule.net>
The Realtek timer node for RTL930x doesn't have any child nodes, making
the use of '#address-cells' quite pointless. It is also not an interrupt
controller, meaning it makes no sense to define '#interrupt-cells'.
The I/O address for this node is also wrong, but this is hidden by the
fact that the driver associated with this node bypasses the usual DT
machinery and does it's own thing. Correct the address to have a sane
value, even though it isn't actually used.
Fixes: a75b9e3ecb ("realtek: Adding RTL930X sub-target")
Signed-off-by: Sander Vanheule <sander@svanheule.net>
When driven by a GPIO pin, the system LED needs to be configured as
active high. Otherwise the LED switches off after booting and
initialisation.
Fixes: 47f5a0a3ee ("realtek: Add support for ZyXEL GS1900-48 Switch")
Signed-off-by: Sander Vanheule <sander@svanheule.net>
The default value for a DT node's status property is already "okay", so
there's no need to specify it again. Drop the status property to clean
up the DTS.
Signed-off-by: Sander Vanheule <sander@svanheule.net>
The assigned output index for the event timers was quite low, lower even
than the ethernet interrupt. This means that high network load could
preempt timer interrupts, possibly leading to all sorts of strange
behaviour.
Increase the interrupt output index of the event timers to 5, which is
the highest priority output and corresponds to the (otherwise unused)
MIPS CPU timer interrupt.
Fixes: a75b9e3ecb ("realtek: Adding RTL930X sub-target")
Signed-off-by: Sander Vanheule <sander@svanheule.net>
The RTL8231 is an external chip, and not part of the SoC. That means
it is more appropriate to define it in the board specific (base) files,
instead of the DT include for the SoC itself.
Moving the RTL8231 definition also ensures that boards with no GPIO
expander, or an alternative one, don't have a useless gpio1 node label
defined.
Tested on a Netgear GS110TPPv1.
Signed-off-by: Sander Vanheule <sander@svanheule.net>
The address in some node names doesn't match the actual offset specified
in the DT node. Update the names to fix this.
While fixing the node names, also drop the unused node labels.
Fixes: 0a7565e536 ("realtek: Update rtl839x.dtsi for realtek,rtl-intc, new gpio controller remove RTL8231 node")
Signed-off-by: Sander Vanheule <sander@svanheule.net>
Bootargs for devices in the realtek target were previously consolidated
in commit af2cfbda2b ("realtek: Consolidate bootargs"), since all
devices currently use the same arguments.
Commit a75b9e3ecb ("realtek: Adding RTL930X sub-target") reverted this
without any argumentation, so let's undo that.
Commit 0b8dfe0851 ("realtek: Add RTL931X sub-target") introduced the
old bootargs also for RTL931x, without providing any actual device
support. Until that is done, let's assume vendors will have done what
they did before, and use a baud rate of 115200.
Fixes: a75b9e3ecb ("realtek: Adding RTL930X sub-target")
Signed-off-by: Sander Vanheule <sander@svanheule.net>
Enable the AP806's cpufreq driver. This driver is compatible with the
Armada 7K and 8K platforms.
Tested-by: Stijn Segers <foss@volatilesystems.org> (RB5009UG+S+IN)
Signed-off-by: Matthew Hagan <mnhagan88@gmail.com>
Enabled CONFIG_ALL_KMODS and ran make kernel_menuconfig against
bcm2711 to update defconfig. Some of the removed symbols are
present in target/linux/generic/config-5.10 while others were
removed by the make target.
Signed-off-by: John Audia <graysky@archlinux.us>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> (wrapped)
Enabled CONFIG_ALL_KMODS and ran make kernel_menuconfig against
bcm2710 to update defconfig. Some of the removed symbols are
present in target/linux/generic/config-5.10 while others were
removed by the make target.
Signed-off-by: John Audia <graysky@archlinux.us>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> (wrapped)
Enabled CONFIG_ALL_KMODS and ran make kernel_menuconfig against
bcm2709 to update defconfig. Some of the removed symbols are
present in target/linux/generic/config-5.10 while others were
removed by the make target.
Signed-off-by: John Audia <graysky@archlinux.us>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> (wrapped)
Enabled CONFIG_ALL_KMODS and ran make kernel_menuconfig against
bcm2708 to update defconfig. Some of the removed symbols are
present in target/linux/generic/config-5.10 while others were
removed by the make target.
Signed-off-by: John Audia <graysky@archlinux.us>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> (wrapped)
Fix the missing ;; after the cAP ac case in /e/b/01_leds.
Fixes: 93d9119 ("ipq40xx: add MikroTik cAP ac support")
Signed-off-by: Robert Marko <robimarko@gmail.com>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> (minor touch-up)
The kernel of both images will no longer fit into
the 3072KiB / 3MiB kernel partition:
|Image Name: ARM OpenWrt Linux-5.10.100
|Created: Sat Feb 19 00:11:55 2022
|Image Type: ARM Linux Kernel Image (uncompressed)
|Data Size: 3147140 Bytes = 3073.38 KiB = 3.00 MiB
Disable both targets for now, until a solution is available.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Use correct indent in target/linux/ramips/image/mt7621.mk
to be consistent with the rest of the file.
Signed-off-by: Nick McKinney <nick@ndmckinney.net>
[rephrase commit message as Adrian suggested, fix a6004ns-m indent]
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
The LED and LAN port numbering on the case of wndr4500v3 devices are
reversed relative to the wndr4300v2. I created this patch to so that the
ordering in OpenWRT will be consistent with that.
Signed-off-by: Graham Cole <diakka@gmail.com>
Add support for ipTIME A3002MESH.
Hardware:
- SoC: MediaTek MT7621AT (880MHz, Duel-Core)
- RAM: DDR3 128MB
- Flash: XMC XM25QH128AHIG (SPI-NOR 16MB)
- WiFi: MediaTek MT7615D (2.4GHz, 5GHz, DBDC)
- Ethernet: MediaTek MT7530 (WAN x1, LAN x2, SoC built-in)
- UART: [GND, RX, TX, 3.3V] (57600 8N1, J4)
MAC addresses:
| interface | MAC | source | comment
|-----------|-------------------|----------------|----------
| LAN | 70:XX:XX:5X:XX:X3 | |
| WAN | 70:XX:XX:5X:XX:X1 | u-boot 0x1fc40 |
| WLAN 2G | 72:XX:XX:4X:XX:X0 | |
| WLAN 5G | 70:XX:XX:5X:XX:X0 | factory 0x4 |
| | 70:XX:XX:5X:XX:X0 | u-boot 0x1fc20 | unknown
| | 70:XX:XX:5X:XX:X2 | factory 0x8004 | unknown
- WLAN 2G MAC address is not the same as stock firmware since OpenWrt
uses LAN MAC address with local bit sets.
Installation:
1. Flash initramfs image. This can be done using stock web ui or TFTP
2. Connect to OpenWrt with an SSH connection to 192.168.1.1
3. Perform sysupgrade with sysupgrade image
Revert to stock firmware:
- Flash stock firmware via OEM TFTP Recovery mode
- Perform sysupgrade with stock image
TFTP Recovery method:
1. Unplug the router
2. Hold the reset button and plug in
3. Release when the power LED stops flashing and go off
4. Set your computer IP address manually to 192.168.0.x / 255.255.255.0
5. Flash image with TFTP client to 192.168.0.1
Signed-off-by: Yoonji Park <koreapyj@dcmys.kr>
[wrap/rephrase commit message]
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
This reverts commit 13a185bf8a.
There was a report that one A1004ns device fails to detect its flash
chip correctly:
[ 1.470297] spi-nor spi0.0: unrecognized JEDEC id bytes: e0 10 0c 40 10 08
[ 1.484110] spi-nor: probe of spi0.0 failed with error -2
It also uses a different flash chip model:
* in my hand: Winbond W25Q128FVSIG (SOIC-8)
* reported: Macronix MX25L12845EMI-10G (SOP-16)
Reducing spi-max-frequency solved the detection failure. Hence revert.
Reported-by: Koasing <koasing@gmail.com>
Tested-by: Koasing <koasing@gmail.com>
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
mtd-mac-address should no longer be used after commit 5ae2e78639
("kernel: drop support for mtd-mac-address"). Convert it to nvmem-cells.
While at it, also convert OpenWrt's custom mtd-cal-data property and
userspace pre-calibration data extraction to the nvmem implementation.
Note: nvmem-cells in QCN5502 wmac has not been tested.
Fixes: c32008a37b ("ath79: add partial support for Netgear EX7300v2")
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
Upstream hwmon-maintainer had various comments about
the changes to the tc654 driver. These have been
addressed and the cooling device support is destined
for inclusion.
One of the comments was the change of the cooling states
scaling. No longer the driver uses the same values as the
hwmon interface, instead the states are now the 17 states
the tc654 supports.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
include the device-tree binding headers that provide definitions
for keys codes and gpios in the device-tree files.
Random bonus: merge tl-wdr4900-v1's uboot with the nvmem-node.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
silences the following message:
> eeprom 0-0051: eeprom driver is deprecated, please use at24 instead
The chip was likely a Dallas Semiconductor and later MAXIM part
before Analog Devices, Inc. bought MAXIM.
From the datasheet:
"The DS28CN01 combines 1024 bits of EEPROM with challenge-and-response
authentication security implemented with the FIPS 180-1/180-2 and
ISO/IEC 10118-3 Secure Hash Algorithm (SHA-1)."
...
"Write Access Requires Knowledge of the Secret
and the Capability of Computing and Transmitting
a 160-Bit MAC as Authorization"
OpenWrt doesn't use it. There's no in-kernel driver
from what I know. Let's document that the chip is
at the location.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
The tricolor LED which is controlled by a lp5521 needed
some maintenance as the driver failed to load in the
current v5.10 image:
| lp5521: probe of 0-0032 failed with error -22
This is because the device-tree needed to be updated
to match the latest led coloring and function trends.
- removed the device name from the label
- added color/function properties
- added required reg and cells properties
For reference a disabled multicolor/RGB is added since this
reflects the real hardware. Unfortunately, the multicolor
sysfs interface isn't supported by yet.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
When Kernel 5.10 was enabled for mpc85xx, the kernel once again became too
large upon decompression (>7MB or so) to decompress itself on boot (see
FS#4110[1]).
There have been many attempts to fix booting from a compressed kernel on
the HiveAP-330:
- b683f1c36d ("mpc85xx: Use gzip compressed kernel on HiveAP-330")
- 98089bb8ba ("mpc85xx: Use uncompressed kernel on the HiveAP-330")
- 26cb167a5c ("mpc85xx: Fix Aerohive HiveAP-330 initramfs image")
We can no longer compress the kernel due to size, and the stock bootloader
does not support any other types of compression. Since an uncompressed
kernel no longer fits in the 8MiB kernel partition at 0x2840000, we need to
patch u-boot to autoboot by running variable which isn't set by the
bootloader on each autoboot.
This commit repartitions the HiveAP, requiring a new COMPAT_VERSION,
and uses the DEVICE_COMPAT_MESSAGE to guide the user to patch u-boot,
which changes the variable run on boot to be `owrt_boot`; the user can
then set the value of that variable appropriately.
The following has been documented in the device's OpenWrt wiki page:
<https://openwrt.org/toh/aerohive/hiveap-330>. Please look there
first/too for more information.
The from-stock and upgrade from a previous installation now becomes:
0) setup a network with a dhcp server and a tftp server at serverip
(192.168.1.101) with the initramfs image in the servers root directory.
1) Hook into UART (9600 baud) and enter U-Boot. You may need to enter
a password of administrator or AhNf?d@ta06 if prompted. If the password
doesn't work. Try reseting the device by pressing and holding the reset
button with the stock OS.
2) Once in U-Boot, set the new owrt_boot and tftp+boot the initramfs image:
Use copy and paste!
# fw_setenv owrt_boot 'setenv bootargs \"console=ttyS0,$baudrate\";bootm 0xEC040000 - 0xEC000000'
# save
# dhcp
# setenv bootargs console=ttyS0,$baudrate
# tftpboot 0x1000000 192.168.1.101:openwrt-mpc85xx-p1020-aerohive_hiveap-330-initramfs-kernel.bin
# bootm
3) Once openwrt booted:
carefully copy and paste this into the root shell. One step at a time
# 3.0 install kmod-mtd-rw from the internet and load it
opkg update; opkg install kmod-mtd-rw
insmod mtd-rw i_want_a_brick=y
# 3.1 create scripts that modifies uboot
cat <<- "EOF" > /tmp/uboot-update.sh
. /lib/functions/system.sh
cp "/dev/mtd$(find_mtd_index 'u-boot')" /tmp/uboot
cp /tmp/uboot /tmp/uboot_patched
ofs=$(strings -n80 -td < /tmp/uboot | grep '^ [0-9]* setenv bootargs.*cp\.l' | cut -f2 -d' ')
for off in $ofs; do
printf "run owrt_boot; " | dd of=/tmp/uboot_patched bs=1 seek=${off} conv=notrunc
done
md5sum /tmp/uboot*
EOF
# 3.2 run the script to do the modification
sh /tmp/uboot-update.sh
# verify that /tmp/uboot and /tmp/uboot_patched are good
#
# my uboot was: (is printed during boot)
# U-Boot 2009.11 (Jan 12 2017 - 00:27:25), Build: jenkins-HiveOS-Honolulu_AP350_Rel-245
#
# d84b45a2e8aca60d630fbd422efc6b39 /tmp/uboot
# 6dc420f24c2028b9cf7f0c62c0c7f692 /tmp/uboot_patched
# 98ebc7e7480ce9148cd2799357a844b0 /tmp/uboot-update.sh <-- just for reference
# 3.3 this produces the /tmp/u-boot_patched file.
mtd write /tmp/uboot_patched u-boot
3) scp over the sysupgrade file to /tmp/ and run sysupgrade to flash OpenWrt:
sysupgrade -n /tmp/openwrt-mpc85xx-p1020-aerohive_hiveap-330-squashfs-sysupgrade.bin
4) after the reboot, you are good to go.
Other notes:
- Note that after this sysupgrade, the AP will be unavailable for 7 minutes
to reformat flash. The tri-color LED does not blink in any way to
indicate this, though there is no risk in interrupting this process,
other than the jffs2 reformat being reset.
- Add a uci-default to fix the compat version. This will prevent updates
from previous versions without going through the installation process.
- Enable CONFIG_MTD_SPLIT_UIMAGE_FW and adjust partitioning to combine
the kernel and rootfs into a single dts partition to maximize storage
space, though in practice the kernel can grow no larger than 16MiB due
to constraints of the older mpc85xx u-boot platform.
- Because of that limit, KERNEL_SIZE has been raised to 16m.
- A .tar.gz of the u-boot source for the AP330 (a.k.a. Goldengate) can
be found here[2].
- The stock-jffs2 partition is also removed to make more space -- this
is possible only now that it is no longer split away from the rootfs.
- the console-override is gone. The device will now get the console
through the bootargs. This has the advantage that you can set a different
baudrate in uboot and the linux kernel will stick with it!
- due to the repartitioning, the partition layout and names got a makeover.
- the initramfs+fdt method is now combined into a MultiImage initramfs.
The separate fdt download is no longer needed.
- added uboot-envtools to the mpc85xx target. All targets have uboot and
this way its available in the initramfs.
[1]: https://bugs.openwrt.org/index.php?do=details&task_id=4110
[2]: magnet:?xt=urn:btih:e53b27006979afb632af5935fa0f2affaa822a59
Tested-by: Martin Kennedy <hurricos@gmail.com>
Signed-off-by: Martin Kennedy <hurricos@gmail.com>
(rewrote parts of the commit message, Initramfs-MultiImage,
dropped bootargs-override, added wiki entry + link, uboot-envtools)
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Correct typo that caused network interfaces for Sophos
SG/XG wireless devices to not be configured properly.
Tested on Sophos SG 135wr2, Sophos XG 125wr2 and
Sophos SG 105wr1
Signed-off-by: Raylynn Knight <rayknight@me.com>
On this device, two of the three defined MTD partitions are
automatically set to read-only, since they do not end at an
erase/write block boundary.
In particular, the only partition remaining writable is the
one holding the u-boot bootloader.
Mark all of the partitions read-only, at least until a better
understanding of why the layout has been laid out this way is
gained.
Signed-off-by: Enrico Mioso <mrkiko.rs@gmail.com>
Enabled `CONFIG_ALL_KMODS` and ran `make kernel_menuconfig` against
ipq806x to update defconfig.
The removed symbols are in fact present in
target/linux/generic/config-5.10. CONFIG_MDIO_DEVRES
was likely added due to this:
<https://elixir.bootlin.com/linux/v5.10.100/source/drivers/net/phy/Kconfig#L16>
Signed-off-by: John Audia <graysky@archlinux.us>
This device still had the legacy flash partitioning.
This is a problem, because neither the nvmem-cells
for mac-address and calibration. Nor the denx,uimage
mtd-splitter compatible would be picked up.
The patch also changes the node-names of the flash
and partition nodes to hopefully meet all the
current FDT trends.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Enrico provided a bootlog that shows the chip is a WAVE-2 QCA9888v2:
> pci 0000:01:00.0: [168c:0056] type 00 class 0x028000
> [...]
> ath10k 5.15 driver, optimized for CT firmware, probing pci device: 0x56.
> ath10k_pci 0000:01:00.0: qca9888 hw2.0 target 0x01000000 [...] chip_id 0x00000000 sub 0000:0000
> ath10k_pci 0000:01:00.0: firmware ver 10.4b-ct-9888-fW-13-5ae337bb1 api 5 features mfp,[...]
> ath10k_pci 0000:01:00.0: board_file api 2 bmi_id N/A crc32 6535d835
> ath10k_pci 0000:01:00.0: htt-ver 2.2 wmi-op 6 htt-op 4 cal file max-sta 32 raw 0 hwcrypto 1
this patch switches the device over to pre-calibration.
(this is more or less cosmetic)
Reported-by: Enrico Mioso <mrkiko.rs@gmail.com>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
The PCIe and built-in 5GHZ radios are meant to operate on different
frequency bands. The hardware enforces this via RF filters.
Add this information to allow software enforcing it as well.
Credits to Piotr Dymacz for the invaluable help.
Signed-off-by: Enrico Mioso <mrkiko.rs@gmail.com>
with the introduction of the DEVICE_ALTX_VENDOR, DEVICE_ALTX_MODEL
multiple/sibiling devices can seemingly supported by one device entry.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
The Ubiquiti EdgePoint R6 is identical to the EdgeRouter X SFP.
However, it fits well into outdoor environments due to its water-proven
case.
More specifications: 9715beb04c ("ramips: add support for Ubiquiti
EdgeRouter X-SFP")
Signed-off-by: Nick Hainke <vincent@systemli.org>
Add the missing IPv6 flow offloading support for routing only.
Hardware flow offloading is done by the packet processing engine (PPE)
of the Ethernet MAC and as it doesn't support mangling of IPv6 packets,
IPv6 NAT cannot be supported.
Signed-off-by: David Bentham <db260179@gmail.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Module kmod-crypto-hw-geode provides accelerated cbc(aes) and ecb(aes)
but the software implementation is also needed when AES key size isn't
128 so that the operation can fall back. Add the kmod so that it would
all work as expected out of the box.
Tested-by: timur_davletshin
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
Switch to a generic GPIO cascade driver.
Signed-off-by: Mauri Sandberg <maukka@ext.kapsi.fi>
Signed-off-by: Petr Štetiar <ynezz@true.cz> [missing commit description]
A Locking bug in the packet receive path was introduced with PR
#4973. The following patch prevents the driver from locking
after a few minutes with an endless flow of
[ 1434.185085] rtl838x-eth 1b00a300.ethernet eth0: Ring contention: r: 0, last a28000f4, cur a28000f8
[ 1434.208971] rtl838x-eth 1b00a300.ethernet eth0: Ring contention: r: 0, last a28000f4, cur a28000fc
[ 1434.794800] rtl838x-eth 1b00a300.ethernet eth0: Ring contention: r: 0, last a28000f4, cur a28000fc
[ 1435.049187] rtl838x-eth 1b00a300.ethernet eth0: Ring contention: r: 0, last a28000f4, cur a28000fc
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: Birger Koblitz <mail@birger-koblitz.de>
When initialising the driver, check if the RTL8231 chip is actually
present at the specified address. If the READY_CODE value does not match
the expected value, return -ENXIO to fail probing.
This should help users to figure out which address an RTL8231 is
configured to use, if measuring pull-up/-down resistors is not an
option.
On an unsuccesful probe, the driver will log:
[ 0.795364] Probing RTL8231 GPIOs
[ 0.798978] rtl8231_init called, MDIO bus ID: 30
[ 0.804194] rtl8231-gpio rtl8231-gpio: no device found at bus address 30
When a device is found, only the first two lines will be logged:
[ 0.453698] Probing RTL8231 GPIOs
[ 0.457312] rtl8231_init called, MDIO bus ID: 31
Signed-off-by: Sander Vanheule <sander@svanheule.net>
Tested-by: Stijn Tintel <stijn@linux-ipv6.be>
The SMI bus ID for RTL8231 currently defaults to 0, and can be
overridden from the devicetree. However, there is no value check on the
DT-provided value, aside from masking which would only cause value
wrap-around.
Change the driver to always require the "indirect-access-bus-id"
property, as there is no real reason to use 0 as default, and perform a
sanity check on the value when probing. This allows the other parts of
the driver to be simplified a bit.
Signed-off-by: Sander Vanheule <sander@svanheule.net>
Tested-by: Stijn Tintel <stijn@linux-ipv6.be>
Set the gpio_chip.base to -1 to use automatic GPIO line indexing.
Setting base to 0 or a positive number is deprecated and should not be
used.
Signed-off-by: Sander Vanheule <sander@svanheule.net>
Tested-by: Stijn Tintel <stijn@linux-ipv6.be>
The RTL8231's gpio_chip.ngpio was set to 36, which is the largest valid
GPIO index. Fix the allowed number of GPIOs by setting ngpio to 37, the
actual line count.
Reported-by: INAGAKI Hiroshi <musashino.open@gmail.com>
Signed-off-by: Sander Vanheule <sander@svanheule.net>
Tested-by: Stijn Tintel <stijn@linux-ipv6.be>
Replace magic values with more self-descriptive code now that I start
to understand more about the design of the PHY (and MDIO controller).
Remove one line before reading RTL8214FC internal PHY id which turned
out to be a no-op and can hence safely be removed (confirmed by
INAGAKI Hiroshi[1])
[1]: df8e6be59a (r66890713)
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Instead of directly calling SoC-specific functions in order to access
(paged) MII registers or MMD registers, create infrastructure to allow
using the generic phy_*, phy_*_paged and phy_*_mmd functions.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
* Add missing Clause-45 write support for rtl931x
* Switch to use helper functions in all Clause-45 access functions to
make the code more readable.
* More meaningful/unified debugging output (dynamic kprintf)
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Import commit ("c6af53f038aa3 net: mdio: add helpers to extract clause
45 regad and devad fields") from Linux 5.17 to allow making the MDIO
code in the ethernet driver more readable.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Using the led-set attribute of a port in the dts we allow configuration
of the port leds. Each led-set is being defined in the led-set configuration
of the .dts, giving a specific configuration to steer the port LEDs via a serial
connection.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
The RTL8221B PHY is a newer version of the RTL8226, also supporting
2.5GBit Ethernet. It is found with RTL931X devices such as the
EdgeCore ECS4125-10P
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Both the Aquantia AQR113c and the RTL8226 PHYs in the Zyxel XGS1250 and the
Zyxel XGS1210 require special polling configuration settings in the
RTL930X_SMI_10GPHY_POLLING_REGxx_CFG configuration registers. Set them.
Additionally, for RTL 1GBit phys set the RTL930X_SMI_PRVTE_POLLING_CTRL bits
in the poll mask.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
For SFP slots on the RTL9302, the link status is not correctly detected.
Use the link media status instead.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
We add the RTL931X sub-target with kernel configuration for
a dual core MIPS InterAptive CPU.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
We add HW support routines for the RTL931X SoC family for handling
the Packet Inspection Engine, L2 table handling and STP aging.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
We need to store and restore MC memberships in HW when a port joins or
leaves a bridge as well as when it is enabled or disabled, as these
properties should not change in these situations.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
In order to receive STP information at the kernel level, we make sure
that all Bridge Protocol Data Units are copied to the CPU-Port.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Instead of a generic L2 aging configuration function with complex
logic, we implement an individual function for all SoC types.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Add functionality to enable or disable L2 learning offload and port flooding
for RTL83XX.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Adds the DSA API for bridge configuration (flooding, L2 learning,
and aging) offload as found in Linux 5.12 so that we can implement
it in our drivver.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
This adds LAG support for all 4 SoC families, including support
ofr the use of different distribution algorithm for the load-
balancing between individual links.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Add the LAG configuration API for DSA as found in Linux 5.12 so that we
can implement it in the dsa driver.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Use setting functions instead of register numbers in order to clean up the code.
Also use enums to define inner/outer VLAN types and the filter type.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
The ZyXEL XGS1250-12 Switch is a 11 + 1 port multi-GBit switch with
8 x 1000BaseT, 3 x 1000/2500/5000/10000BaseT Ethernet ports and
1 SFP+ module slot.
Hardware:
- RTL9302B SoC
- Macronix MX25L12833F (16MB flash)
- Nanja NT5CC64M16GP-1 (128MB DDR3 SDRAM)
- RTL8231 GPIO extender to control the port LEDs
- RTL8218D 8x Gigabit PHY
- Aquantia AQR113c 1/2.5/5/10 Gigabit PHYs
- SFP+ 10GBit slot
Power is supplied via a 12V 2A standard barrel connector. At the
right side behind the grid is UART serial connector. A Serial
header can be connected to from the outside of the switch trough
the airvents with a standard 2.54mm header.
Pins are from top to bottom Vcc(3.3V), TX, RX and GND. Serial
connection is via 115200 baud, 8N1.
A reset button is accessble through a hole in the front panel
At the time of this commit, all ethernet ports work under OpenWRT,
including the various NBaseT modes, however the 10GBit SFP+ slot is not
supported.
Installation
--------------
* Connect serial as per the layout above. Connection parameters: 115200 8N1.
* Navigate to 'Management' in the OEM web interface and click on 'Firmware upgrade'
to the left.
* Upload the OpenWrt initramfs image, and wait till the switch reboots.
* Connect to the device through serial and change the U-boot boot command.
> fw_setenv bootcmd 'rtk network on; boota'
* Reboot, scp the sysupgrade image to /tmp, verify the checksum and flash it:
> sysupgrade /tmp/openwrt-realtek-rtl930x-zyxel_xgs1250-12-squashfs-sysupgrade.bin
* Upon reboot, you have a functional OpenWrt installation. Leave the bootcmd
value as is - without 'rtk network on' the switch will fail to initialise
the network.
Web recovery
------------
The XGS1250-12 has a handy web recovery that will load when U-boot does
not find a bootable kernel. In case you would like to trigger the web
recovery manually, partially overwrite the firmware partition with some
zeroes:
# dd if=/dev/zero of=/dev/mtd5 bs=1M count=2
If you have serial connected you'll see U-boot will start the web recovery
and print it's listening on 192.168.1.1, but by default it seems to be on
the OEM default IP for the switch - 192.168.1.3. The web recovery only
listens on HTTP (80) and *not* on 443 (HTTPS) unlike the web UI.
Return to stock
---------------
You can flash the ZyXEL firmware images to return to stock:
# sysupgrade -F -n XGS1250-12_Firmware_V1.00(ABWE.1)C0.bix
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Adds configuration routines for the internal SerDes of the
RTL930X and RTL931X.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Adds a rtl931x_phylink_mac_config for the RTL931X and improve
the handling of the RTL930X phylink configuration. Add separate
handling of the RTL839x since some configurations are different
from the RTL838X.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
We were using the PHY-ids (the reg entries in the PHY
sections of the .dts) as the port numbers. Now scan the
ports section in the .dts, and use the actual port numbers,
following the phy-handle to the PHY properties.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
When a port is brought up, read the SDS-id via the phy_device
for a given port and use this to configure the SDS when it
is brought up.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
The RTL839X does not have an internal phy and thus does not need to have any
firmware as part of the kernel, especially not firmware for the RTL838X.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Selects the new CEVT timer for Realtek instead of the previous
timer driver. While we are at it, we explicitily state we do
not use the I2C driver of the RTL9300.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
The RTL9300 has a broken R4K MIPS timer interrupt, however, the
R4K clocksource works. We replace the RTL9300 timer with a
Clock Event Timer (CEVT), which is VSMP aware and can be instantiated
as part of brining a VSMTP cpu up instead of the R4K CEVT source.
For this we place the RTL9300 CEVT timer in arch/mips/kernel
together with other MIPS CEVT timers, initialize the SoC IRQs
from a modified smp-mt.c and instantiate each timer as part
of the MIPS time setup in arch/mips/include/asm/time.h instead
of the R4K CEVT, similarly as is done by other MIPS CEVT timers.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Various fixes to enable Ethernet on the RTL931X:
- Network start and stop sequence for RTL931X HW
- MDIO access on RTL931X SoC
- Chip initialization
- SerDes setup
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Do not lock the register structure in IRQ context. It is not
necessary and leads to lockups under SMP load.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Rename the SoC-specific rtl838x_reg structure in the Ethernet
driver to avoid confusion with the structure of the same name
in the DSA driver. New name is: rtl838x_eth_reg
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Setting bits 20 and 23 in a u16 is obviously wrong.
According to https://www.svanheule.net/realtek/cypress/cputag
cpu_tag[2] starts at bit 48 in the cpu-tag structure, so
bit 43 is bit 5 in cpu_tag[2] and bit 40 is bit 8 in
cpu_tag[2].
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Set CONFIG_FORCE_MAX_ZONEORDER setting to 13 to allow larger
contiguous memory allocation for the DMA of the Ethernet
driver. Increase the number of entries in the RX ring
to 300 making use of the larger DMA region now possible for
receiveing packets.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
The GS1900-48 is a 48 + 2 port Gigabit L2 switch with 48 gigabit ports.
Hardware:
RTL8393M SoC
Macronix MX25l12805D (16MB flash)
128MB RAM
6 * RTL8218B external PHY
2 * RTL8231 GPIO extenders to control the port LEDs, system LED and
Reset button
2 Uplink ports are SFP cages which support 1000 Base-X mini GBIC modules.
Power is supplied via a 230 volt mains connector.
The board has a hard reset switch SW1, which is is not reachable from the outside.
J4 provides a 12V RS232 serial connector which is connected through U8 to
the 3.3V UART of the RTL8393. Conversion is done by U8, a SIPEX 3232EC.
To connect to the UART, wires can be soldered to R603 (TX) and R602 (RX).
Installation:
Install the squashfs image via Realtek's original Web-Interface.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Update the IRQ configuration to work with the new rtl-intc controller.
Also change all KSEG1 addresses in reg = <> of the devics to physical
addresses.
Use the new gpio-otto controller instead of the legacy driver.
Also remove the memory node as this is better put into a device .dts.
Also remove the RTL8231 GPIO controller node from this base file
since the chip might not be found in all Realtek RTL839x devices.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Replace the interrupt controller node with the new realtek,rtl-intc
node and change all device interrupts to use the 2 field notation:
interrupts = <[SoC IRQ] [Index to MIPS IRQ]>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
In order to support VSMP, enable support for both VPEs
of the RTL839X and RTL930X SoCs in the irq-realtek-rtl
driver. Add support for IRQ affinity setting.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
In order for the Platform includes to be available on
all sub-targets, make them dependent on CONFIG_RTL83XX.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
The RTL838X SoCs do not use Aquantia PHYs, remove this.
Also the RTL838X uses a high resolution R4K timer.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Creates RTL83XX as a basic kernel config parameter for the
RTL838X, RTL839x, RTL930X and RTL931X platforms with respective
configurations for the SoCs, which are introduced in addition.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Create the RTL838x specific Makefiles. Move CPU-type into
rtl838x.mk as this is specifc to that platform. Add
rtl838x subtarget into main Makefile.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
mv generic/target.mk to rtl838x/target.mk in order to create
an initial makefile for the rtl838x sub-architecture
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
The EEPROMs on SFP modules are compatible both to I2C as well
as SMBus. However, the kernel so far only supports I2C
access. We add SMBus access routines, because the I2C driver
for the RTL9300 HW only supports that protocol. At the same
time we disable I2C access to PHYs on SFP modules as otherwise
detection of any SFP module would fail. This is not in any
way problematic at this point in time since the RTL93XX
platform so far does not support PHYs on SFP modules.
The patches are copied and rebased version of:
https://bootlin.com/blog/sfp-modules-on-a-board-running-linux/
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
The RTL9300/RTL9310 I2C controllers have support for 2 independent I2C
masters, each with a fixed SCL pin, that cannot be changed. Each of these
masters can use 8 (RTL9300) or 16 (RTL9310) different pins for SDA.
This multiplexer directly controls the two masters and their shared
IO configuration registers to allow multiplexing between any of these
busses. The two masters cannot be used in parallel as the multiplex
is protected by a standard multiplex lock.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
This adds support for the RTL9300 and RTL9310 I2C controller.
The controller implements the SMBus protocol for SMBus transfers
over an I2C bus. The driver supports selecting one of the 2 possible
SCL pins and any of the 8 possible SDA pins. Bus speeds of
100kHz (standard speed) and 400kHz (high speed I2C) are supported.
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
This patch removes support for the legacy GPIO driver, since now
the gpio-otto driver can be used on all platforms
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
We add support for the RTL930X and RTL931X architectures
in the gpio-realtek-otto.c driver.
Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Signed-off-by: Birger Koblitz <git@birger-koblitz.de>
Drop patches and files for Linux 5.4 now that we've been using 5.10
for a while and support for Linux 5.4 has gone out-of-sync.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
The bit position mask was accidentally made too wide, overlapping with the LSB
from the byte position mask. This caused ECC calculation to fail for odd bytes
Signed-off-by: Chad Monroe <chad.monroe@smartrg.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
flowtable->net was initialized too late, and this could be triggered even
without hardware offload support on the device
Signed-off-by: Felix Fietkau <nbd@nbd.name>
It's reported that current memory detection code occasionally detects
larger memory under some bootloaders.
Current memory detection code tests whether address space wraps around
on KSEG0, which is unreliable because it's cached.
Rewrite memory size detection to perform the same test on KSEG1 instead.
While at it, this patch also does the following two things:
1. use a fixed pattern instead of a random function pointer as the magic
value.
2. add an additional memory write and a second comparison as part of the
test to prevent possible smaller memory detection result due to
leftover values in memory.
Fixes: 6d91ddf517 ("ramips: mt7621: add support for memory detection")
Reported-by: Rui Salvaterra <rsalvaterra@gmail.com>
Tested-by: Rui Salvaterra <rsalvaterra@gmail.com>
Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
The issue of EAP frames sent to group address (or the wrong address) has been
addressed in mac80211, so this hack is no longer needed
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This caches flows between MAC addresses on separate ports, including their VLAN
in order to bypass the normal bridge forwarding code.
In my test on MT7622, this reduces LAN->WLAN bridging CPU usage by 6-10%,
potentially even more on weaker platforms
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Had to update generic defconfig (make kernel_menuconfig CONFIG_TARGET=generic)
for this bump, but since that only modifies the target defined in .config,
and since that target also needed to be updated for unrelated reasons, manually
propagated the newly added symbol to the generic config.
Removed upstreamed:
pending-5.10/860-Revert-ASoC-mediatek-Check-for-error-clk-pointer.patch[1]
All other patches automatically rebased.
1. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.10.99&id=080f371d984e8039c66db87f3c54804b0d172329
Build system: x86_64
Build-tested: bcm2711/RPi4B, mt7622/RT3200
Run-tested: bcm2711/RPi4B, mt7622/RT3200
Signed-off-by: John Audia <graysky@archlinux.us>
