Serge Vasilugin reports:
To improve mt7620 built-in wifi performance some changes:
1. Correct BW20/BW40 switching (see comments with mark (1))
2. Correct TX_SW_CFG1 MAC reg from v3 of vendor driver see
https://gitlab.com/dm38/padavan-ng/-/blob/master/trunk/proprietary/rt_wifi/rtpci/3.0.X.X/mt76x2/chips/rt6352.c#L531
3. Set bbp66 for all chains.
4. US_CYC_CNT init based on Programming guide, default value was 33 (pci),
set chipset bus clock with fallback to cpu clock/3.
5. Don't overwrite default values for mt7620.
6. Correct some typos.
7. Add support for external LNA:
a) RF and BBP regs never be corrected for this mode
b) eLNA is driven the same way as ePA with mt7620's pin PA
but vendor driver explicitly pin PA to gpio mode (for forrect calibration?)
so I'm not sure that request for pa_pin in dts-file will be enough
First 5 changes (really 2) improve performance for boards w/o eLNA/ePA.
Changes 7 add support for eLNA
Configuration w/o eLAN/ePA and with eLNA show results
tx/rx (from router point of view) for each stream:
35-40/30-35 Mbps for HT20
65-70/60-65 Mbps for HT40
Yes. Max results for 2T2R client is 140-145/135-140
with peaks 160/150, It correspond to mediatek driver results.
Boards with ePA untested.
Reported-by: Serge Vasilugin <vasilugin@yandex.ru>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
[directly include v3 of the patchset submitted upstream]
(cherry picked from commit 31a6605de0)
(cherry picked from commit e785ca05e9)
(cherry picked from commit 412fcf3d44)
Prepare patches for sending upstream by adding patch descriptions
generated from the original OpenWrt commits adding each patch.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit d4feb66048)
The RGB LED of the UniFi 6 LR v1 doesn't work when using the Openwrt-
built U-Boot. This is because the vendor loader resets the ledbar
controller while our U-Boot doesn't care.
Add reset-gpio so the ledbar driver in Linux will always reset the
ledbar controller.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 0b5cf952cf)
Hardware
--------
- SoC: MediaTek MT7621AT with 128 MiB RAM and 32 MiB Flash
- Wi-Fi: MediaTek MT7603 (b/g/n, 2x2) and MediaTek MT7615 (ac, 4x4)
- Bluetooth: CSR8811 (internal USB, install kmod-bluetooth)
Installation
------------
1. Connect to the booted device at 192.168.1.20 using username/password
"ubnt".
2. Update the bootloader environment.
$ fw_setenv devmode TRUE
$ fw_setenv boot_openwrt "fdt addr \$(fdtcontroladdr);
fdt rm /signature; bootubnt"
$ fw_setenv bootcmd "run boot_openwrt"
3. Transfer the OpenWrt sysupgrade image to the device using SCP.
4. Check the mtd partition number for bs / kernel0 / kernel1
$ cat /proc/mtd
5. Set the bootselect flag to boot from kernel0
$ dd if=/dev/zero bs=1 count=1 of=/dev/mtdblock4
6. Write the OpenWrt sysupgrade image to both kernel0 as well as kernel1
$ dd if=openwrt.bin of=/dev/mtdblock6
$ dd if=openwrt.bin of=/dev/mtdblock7
7. Reboot the device. It should boot into OpenWrt.
Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 820f0c07c5)
The patch adding support for LEDs connected to a reset controller did
not apply any more, refresh it on top of current master.
Fixes: 53fc987b25 ("generic: move ledbar driver from mediatek target")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 76fc277917)
This moves the ledbar driver to generic, to be also used by the ramips target.
Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 53fc987b25)
The LEDs connected to the MCU are so-called smart LEDs and their signal is
daisy-chained. Because of this, the MCU needs to be told how many LEDs are
connected. It also means the LEDs could be individually controlled, if the MCU
has a command for this.
Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 76198e8f09)
During GPIO initialization the pin state flips and triggers a reset of
the ledbar MCU. It needs to be moved through an initialization sequence
before working correctly.
Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 013a956f08)
Some versions of the ledbar MCU have a reset pin. It needs to be
correctly initialized or we might keep the MCU in reset state.
Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 84e4bbf5f0)
There are commands that return more than one byte of response.
Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit 7c852e7df5)
Or the comparison against a signed char is always true, because the
literal 0xaa is treated as an unsigned int, to which the signed char is
casted during comparison. 0xaa is above the positive values of a signed
char and negative signed char values result in values larger than 0xaa
when casted to unsigned int.
Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit e9a22ce253)
The read response is in the i2c_response variable. Also use %hhx format,
because we're dealing with a single char.
Signed-off-by: Sven Wegener <sven.wegener@stealer.net>
(cherry picked from commit c4f9f9b44c)
Package kernel module providing ESSIV support for block encryption.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 4133102898)
Janusz Dziedzic reported a typo introduced by a recent commit. Fix it.
Fixes: 50c892d67b ("mediatek: bpi-r64: make initramfs/recovery optional")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 007c8809c1)
Only include recovery image in SD card image generated for the
BananaPi BPi-R64 if building with CONFIG_TARGET_ROOTFS_INITRAMFS
This allows to build images larger than 32 MB (the limit for
initramfs/recovery image) by deselecting initramfs.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 50c892d67b)
Add the missing LEDs for GB-PC2. Some of these LEDs don't exist on the
device schematics. Tests on a GB-PC2 by me and Petr proved otherwise.
Remove ethblack-green and ethblue-green LEDs for GB-PC1. They are not wired
to GPIO 3 or 4 and the wiring is currently unknown.
Set ethyellow-orange to display link state and activity of the ethyellow
interface for GB-PC2.
Link: https://github.com/ngiger/GnuBee_Docs/blob/master/GB-PCx/Documents/GB-PC2_V1.1_schematic.pdf
Tested-by: Petr Louda <petr.louda@outlook.cz>
Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
(cherry picked from commit 2a6ef7f53d)
Several users of wget for downloads (curl is not available in the
system) have reported broken download functionality:
wget --tries=5 --timeout=20 --output-document=- https://cdn.kernel.org/pub/linux/kernel/v5.x/linux-5.10.142.tar.xz
http://: Invalid host name.
Thats all happening due to '' was passed as an argument, which got later
expanded to http://.
In the context of a list constructor '' is not nothing, it is an empty
string element. So fix it by using () as it will yield "nothing" and
thus not introduce an empty string element.
Fixes: #10692
Fixes: 90c6e3aedf ("scripts: always check certificates")
Signed-off-by: Jo-Philipp Wich <jo@mein.io> [shellwords() -> ()]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit 50a48faa1b)
When running build in verbose mode `make V=s` we can see a lot of
following warnings when curl is not available in the system:
Can't exec "curl": No such file or directory at scripts/download.pl line 77.
So lets fix it by redirecting of the stderr to null hole.
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit c836ca84e8)
Remove flags from wget and curl instructing them to ignore bad server
certificates. Although other mechanisms can protect against malicious
modifications of downloads, other vectors of attack may be available
to an adversary.
TLS certificate verification can be disabled by turning oof the
"Enable TLS certificate verification during package download" option
enabled by default in the "Global build settings" in "make menuconfig"
Signed-off-by: Josh Roys <roysjosh@gmail.com>
[ add additional info on how to disable this option ]
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit 90c6e3aedf)
It shouldn't be needed anymore as we've now `scripts/xxdi.pl`, which
should be self contained and fully compatible `xxd -i` replacement.
Fixes: #10555
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit 88c9056a70)
Dependency on xxd was added in commit c4dd2441e7 ("tools: add xxd
(from vim)") as U-Boot requires xxd to create the default environment
from an external file.
Later in commit 2b94aac7a1 ("tools: xxd: use more convenient source
tarball"), xxd from another source was used instead, but that source is
currently unavailable, so let's fix it by using simple xxdi.pl Perl
script instead.
Fixes: #10555
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit eae2fb8027)
So it can serve as a standalone drop in replacement for xxd utility used
currently mostly in U-Boot packages with `xxd -i` mode which outputs C
include file style, with aim for byte to byte identical output, so the
eventual difference in the generated output is easily spottable.
Fixes: #10555
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Signed-off-by: Jo-Philipp Wich <jo@mein.io> [perl-fu]
(cherry picked from commit 06e01e817e)
In order to make it more portable.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit 8b278a76d9)
xxdi.pl is a Perl script that implements vim's 'xxd -i' mode so that
packages do not have to use all of vim just to get this functionality.
References: #10555
Source: 97a6bd5cee/xxdi.pl
Signed-off-by: Petr Štetiar <ynezz@true.cz>
(cherry picked from commit 2117d04a3a)
Re-introduce the queue wake fix that was reverted due to a regression,
but this time with the follow-up fixes that take care of the regression.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry picked from commit 9a93b62f31)
(cherry-picked from commit 8b804cae5e)
(cherry-picked from commit 8b06e06832)
Backport upstream solution that permits to declare nvmem cells with
dynamic partition defined by special parser.
This provide an OF node for NVMEM and connect it to the defined dynamic
partition.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
(cherry picked from commit 1a9ee36734)
Add the aliases sections required to detect LEDs specific to OpenWrt
boot / update indication for the NanoPi R4S.
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit 1588069612)
Ensure the MAC address for all NanoPi R4S boards is assigned unique for
each board.
FriendlyElec ship two versions of the R4S: The standard as well as the
enterprise edition with only the enterprise edition including the EEPROM
chip that stores the unique MAC address.
In order to assign both board types unique MAC addresses, fall back on
the same method used for the NanoPi R2S in case the EEPROM chip is not
present by generating the board MAC from the SD card CID.
[0] https://wiki.friendlyelec.com/wiki/index.php/NanoPi_R4S#Differences_Between_R4S_Standard_Version_.26_R4S_Enterprise_Version
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit b5675f500d)
The previous fixup was incomplete, and the offsets for the
queue and crc_error cpu_tag bitfields were still wrong on
RTL839x.
Fixes: 545c6113c9 ("realtek: fix RTL838x receive tag decoding")
Suggested-by: Jan Hoffmann <jan@3e8.eu>
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Add LTE packages required for operating the LTE modem optionally shipped
with the GL-AP1300.
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit c7c3509226)
Add support for the ZTE MF281 battery-powered WiFi router.
Hardware
--------
SoC: Qualcomm Atheros QCA9563
RAM: 128M DDR2
FLASH: 2M SPI-NOR (GigaDevice GD25Q16)
128M SPI-NAND (GigaDevice)
WLAN: QCA9563 2T2R 802.11 abgn
QCA9886 2T2R 802.11 nac
WWAN: ASRMicro ASR1826
ETH: Qualcomm Atheros QCA8337
UART: 115200 8n1
Unpopulated connector next to SIM slot
(SIM) GND - RX - TX - 3V3
Don't connect 3V3
BUTTON: Reset - WPS
LED: 1x debug-LED (internal)
LEDs on front of the device are controlled
using the modem CPU and can not be controlled
by OpenWrt
Installation
------------
1. Connect to the serial console. Power up the device and interrupt
autoboot when prompted
2. Connect a TFTP server reachable at 192.168.1.66 to the ethernet port.
Serve the OpenWrt initramfs image as "speedbox-2.bin"
3. Boot the initramfs image using U-Boot
$ setenv serverip 192.168.1.66
$ setenv ipaddr 192.168.1.154
$ tftpboot 0x84000000 speedbox-2.bin
$ bootm
4. Copy the OpenWrt factory image to the device using scp and write to
the NAND flash
$ mtd write /path/to/openwrt/factory.bin firmware
WWAN
----
The WWAN card can be used with OpenWrt. Example configuration for
connection with a unauthenticated dual-stack APN:
network.lte=interface
network.lte.proto='ncm'
network.lte.device='/dev/ttyACM0'
network.lte.pdptype='IPV4V6'
network.lte.apn='internet.telekom'
network.lte.ipv6='auto'
network.lte.delay='10'
The WWAN card is running a modified version of OpenWrt and handles
power-management as well as the LED controller (AW9523). A root shell
can be acquired by installing adb using opkg and executing "adb shell".
Signed-off-by: David Bauer <mail@david-bauer.net>
(cherry picked from commit 1e1695f959)
Commit dc9cc0d3e2 ("realtek: add QoS and rate control") replaced a
16 bit reserved field in the RTL83xx packet header with the initial
cpu_tag word, shifting the real cpu_tag fields by one. Adjusting for
this new shift was partially forgotten in the new RX tag decoders.
This caused the switch to block IGMP, effectively blocking IPv4
multicast.
The bug was partially fixed by commit 9d847244d9 ("realtek: fix
RTL839X receive tag decoding")
Fix on RTL838x too, including correct NIC_RX_REASON_SPECIAL_TRAP value.
Suggested-by: Jan Hoffmann <jan@3e8.eu>
Fixes: dc9cc0d3e2 ("realtek: add QoS and rate control")
Signed-off-by: Bjørn Mork <bjorn@mork.no>
(cherry picked from commit 545c6113c9)
There are two feature currently altered by the multicast_to_unicast option.
1. bridge level multicast_to_unicast via IGMP snooping
2. hostapd/mac80211 config multicast_to_unicast setting
The hostapd/mac80211 setting has the side effect of converting *all* multicast
or broadcast traffic into per-station duplicated unicast traffic, which can
in some cases break expectations of various protocols.
It also has been observed to cause ARP lookup failure between stations
connected to the same interface.
The bridge level feature is much more useful, since it only covers actual
multicast traffic managed by IGMP, and it implicitly defaults to 1 already.
Renaming the hostapd/mac80211 option to multicast_to_unicast_all should avoid
unintentionally enabling this feature
Signed-off-by: Felix Fietkau <nbd@nbd.name>
(cherry-picked from commit 09ea1db93b)
