Nick Hainke [Tue, 25 Mar 2025 16:31:24 +0000 (17:31 +0100)]
libtracefs: update to 1.8.1
ChangeLog 417c2e3 libtracefs: version 1.8.1 41efd9e libtracefs: Add meson build targets to Makefile 310b796 libtracefs utest: Add better logic to cause missed events b589e32 libtracefs: Add cpu-map sample to trace mapped buffer 4ede86e libtracefs: Enable mmapped ring buffer e6737d4 libtracefs: Initialize val in build_filter() 590e452 libtracefs: Close dir in the error path in tracefs_event_systems() 0309a87 libtracefs: Close dir in the error path in tracefs_system_events() f34fb1f libtracefs: Prevent memory leak in tracefs_dynevent_get_all() 48e906b libtracefs: my_yyinput() should return 0 when no data can be read 5e5b2a7 libtracefs: Prevent memory leak in tracefs_instance_create() 8f2593f libtracefs: Prevent a memory leak in open_cpu_files() 7d77b83 libtracefs: Prevent a memory leak in tracefs_system_events() 7fcd8d2 libtracefs: Prevent a memory leak in add_func_str() a01d0ba libtracefs: Don't leak socket file descriptor in open_vsock() efdf7f7 libtracefs: Prevent memory leak in tracefs_event_systems() 2342293 libtracefs: Prevent a memory leak in tracefs_synth_add_end_field() 1c95394 libtracefs: Prevent a memory leak in update_fields() 6b35665 libtracefs: Prevent memory leak in append_filer() aecc0b7 libtracefs: Call va_end() before exiting tracefs_hist_set_sort_key() a5e37f7 libtracefs: Add trace-mmap.c to meson build 8f62e96 libtracefs meson: Use SemVer in the build description e04fa01 meson: Add utest option fb213a4 libtracefs: Update trace_buffer_meta 04505a0 libtracefs utest: Include libgen.h for basename() 7b4a9c9 libtracefs utest: Define _LARGEFILE64_SOURCE for lseek64() with musl ba75081 libtracefs utest: Add PATH_MAX if it is not already defined 5f27b7f libtracefs: Update the kbuf for previous read in trace_mmap_load_subbuf() 73ac9c1 libtracefs: Fix tracefs_instance_reset() of triggers 7d15d77 libtracefs meson: build tracefs-mmap by default
Nick Hainke [Tue, 25 Mar 2025 16:27:01 +0000 (17:27 +0100)]
libtraceevent: update to 1.8.4
ChangeLog: bd47bd5 libtraceevent: 1.8.4 fe0bc49 libtraceevent: Print function pointer address when TEP_EVENT_FL_PRINTRAW is specified f2224d5 libtraceevent: Have sizeof() parsing handle u8/s8 through u64/s64 5f570de libtraceevent: Print arrays like Linux does 645a883 libtraceevent: 1.8.3 d4c1fb4 libtraceevent: Add meson build targets to Makefile c3dc220 libtraceevent: Fix a double free in process_op() 021da90 libtraceevent: Do not return a local stack pointer in get_field_str() 340e2e6 libtraceevent: Have unit test fail when any tests fail c84155f libtraceevent: prevent a memory leak in tep_plugin_add_option() 03551eb libtraceevent: Prevent a memory leak in process_fields() 34ece90 libtraceevent: Close shared object in the error path of load_plugin() 8802f0f libtraceevent: Avoid a simple asprintf case 76a0eb8 libtraceevent: Fix event-parse memory leak in process_cond 5bc98bd libtraceevent: Have single quotes represent characters ec8e0cc libtraceevent: Fix tests running on big endian arch 60ed6c3 libtraceevent: build: Various fixes for the Meson build of libtraceevent 0351241 libtraceevent utest: Include libgen.h for basename() with musl
Robert Marko [Sun, 23 Mar 2025 09:29:28 +0000 (10:29 +0100)]
uboot-tools: dont build tools unconditionally
Currently, both envtools and the rest of U-Boot tools are being built
regardless if the dumpimage package has been selected.
This will fail if only envtools are selected since the rest of tools
require OpenSSL while envtools do not require them.
So, only build tools if dumpimage is selected.
Fixes: 46e376c93514 ("uboot-tools: migrate uboot-envtools to uboot-tools") Fixes: #18327 Tested-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com> Link: https://github.com/openwrt/openwrt/pull/18329 Signed-off-by: Robert Marko <robimarko@gmail.com>
Sat Mar 16 19:11:39 2024 daemon.info lldpd[10916]: custom TLV op replace oui 33:44:55 subtype fe
Sat Mar 16 19:11:39 2024 daemon.info lldpd[10916]: custom TLV op add oui 33:44:44 subtype e8
Sat Mar 16 19:11:39 2024 daemon.info lldpd[10916]: custom TLV op add oui 33:44:33 subtype 42
Sat Mar 16 19:11:39 2024 daemon.info lldpd[10916]: custom TLV op add oui 33:44:33 subtype 42
Sat Mar 16 19:11:39 2024 daemon.info lldpd[10916]: custom TLV op add oui 33:44:31 subtype 2c
Sat Mar 16 19:11:39 2024 daemon.info lldpd[10916]: custom TLV op add oui 33:44:31 subtype 2c
Sat Mar 16 19:11:39 2024 daemon.warn lldpcli[10915]: invalid OUI value '3322'
Sat Mar 16 19:11:39 2024 daemon.info lldpcli[10915]: an error occurred while executing last command
Sat Mar 16 19:11:39 2024 daemon.warn lldpcli[10915]: invalid OUI value '3312'
Sat Mar 16 19:11:39 2024 daemon.info lldpcli[10915]: an error occurred while executing last command
Sat Mar 16 19:11:39 2024 daemon.info lldpcli[10915]: lldpd should resume operations
( The last two TLV are invalid: their oui must be three hex bytes, comma
separated. Only the first hex byte of oui-info 5555555555 is used )
As described in #13873, from 23.05.0 onwards logging to a file on an
external filesystem fails under some conditions.
This occurs because the log initscript had code added to prevent start
logging to an external filesystem on boot, and added a trigger to start
said logging when the external filesystem gets mounted.
The issue is that for filesystems mount with fstab uci, the fstab
scripts runs at START=11, while log runs at START=12, which means the
external filesystem may already be mounted by the time the log initscript
runs. Since the external filesystem is already mounted it does not
trigger a hotplug event to trigger the trigger to start logging. This
combination means the logging never automatically starts when the log
file is on an external filesystem.
We therefore add a check for the presence of a mounted filesystem when
the log file is being sent to an fstab mounted filesystem. If the
filesystem is mounted, we don't skip starting logging during boot.
If the filesystem is not mounted then file logging is not started and
the trigger will start the logging when the filesystem is mounted.
Signed-off-by: Daniel F. Dickinson <dfdpublic@wildtechgarden.ca>
[improved commit message] Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Daniel Golle [Fri, 21 Mar 2025 11:39:07 +0000 (11:39 +0000)]
mediatek: filogic: fix case statement in 01_leds
Remove a stray '|' character from 01_leds which has accidentally
been added.
Reported-by: Chukun Pan <amadeus@jmu.edu.cn> Fixes: 63d56af6c6 ("mediatek: filogic: migrate Netgate N60 to upstream PHY LED control") Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Daniel Golle [Fri, 21 Mar 2025 01:04:39 +0000 (01:04 +0000)]
mediatek: filogic: PHY LEDs do have an address, gpio-leds don't
Other than GPIO LEDs, PHY LEDs do have an address.
Fix node names such that all gpio-leds do *not* contain an '@' sign and
PHY leds which do have an address also do contain the '@' sign.
This is done to prevent more copy&paste'ry of non-complaint DT
fragments.
Fixes: 7cbe34170e ("mediatek: add support for the GL.iNet GL-MT3000") Fixes: fe10f97439 ("filogic: add support for GL.iNet GL-MT6000") Fixes: e8f7597317 ("mediatek: filogic: add support for Cudy RE3000 v1") Fixes: c9cb6411c1 ("mediatek: add support for Cudy WR3000 v1") Fixes: 7560af7647 ("mediatek: filogic: migrate ASUS TUF AX6000 to upstream PHY LED control") Fixes: 25ea7ff393 ("mediatek: filogic: migrate Acer W6/W6d to upstream PHY LED control") Fixes: d50d51d74e ("mediatek: filogic: migrate Zyxel NWA50AX Pro to upstream PHY LED control") Fixes: b88de5d507 ("mediatek: filogic: migrate Zyxel EX5700 to upstream PHY LED control") Fixes: 63d56af6c6 ("mediatek: filogic: migrate Netgate N60 to upstream PHY LED control") Fixes: fd76a38190 ("mediatek: filogic: migrate SmartRG Bonanza to upstream PHY LED control") Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Lech Perczak [Tue, 28 Feb 2023 01:26:40 +0000 (02:26 +0100)]
ath79: support switch LEDs on TL-WR1043ND v2/v3
Add switch LED definitions for TP-Link TL-WR1043ND family, based on data
extracted from ar71xx board file. Update the LED labels to match current
pattern, i.e. drop the "tp-link:" prefix.
Lech Perczak [Tue, 4 Feb 2025 20:15:21 +0000 (21:15 +0100)]
ath79: TP-link TL-WR1043ND v2/v3: use fixed-link for AR8327 switch connection
Attaching PHY driver to the switch, while adding LEDs binding causes the
PHY driver to create additional LED instances, handled incorrectly by
the PHY driver, which are non-functional. Use fixed-link to attach the
switch driver, instead of PHY driver, to prevent that.
This has a side effect of not logging switch port up/down events in the kernel
log.
Add switch LED definitions for TP-Link Archer C7 v1/2/3 family, based on data
extracted from ar71xx board file. Update the LED labels to match current
pattern, i.e. drop the "tp-link:" prefix.
Lech Perczak [Tue, 4 Feb 2025 20:15:16 +0000 (21:15 +0100)]
ath79: TP-link Archer C7v2: use fixed-link for AR8327 switch connection
Attaching PHY driver to the switch, while adding LEDs binding causes the
PHY driver to create additional LED instances, handled incorrectly by
the PHY driver, which are non-functional. Use fixed-link to attach the
switch driver, instead of PHY driver, to prevent that.
This has a side effect of not logging switch port up/down events in the kernel
log.
Add switch LED definitions for TP-Link TL-WDR4300 family, based on data
extracted from ar71xx board file. Update the LED labels to match current
pattern, i.e. drop the "tp-link:" prefix.
Lech Perczak [Tue, 4 Feb 2025 20:14:52 +0000 (21:14 +0100)]
ath79: TP-link TL-WDR4300: use fixed-link for AR8327 switch connection
Attaching PHY driver to the switch, while adding LEDs binding causes the
PHY driver to create additional LED instances, handled incorrectly by
the PHY driver, which are non-functional. Use fixed-link to attach the
switch driver, instead of PHY driver, to prevent that.
This has a side effect of not logging switch port up/down events in the kernel
log.
The ar8216 switch driver supports exposing configuration of AR8327 and
AR8337 switch LEDs to the userspace, however it is only configurable
through platform data, causing the devices ported from ar71xx target to
lack the support.
Since there is still a long way to go until we can migrate the target to
qca8k, an interim solution is needed.
Extend ar8327_hw_config_of function to parse a "leds"
subnode, which will populate the missing platform data based on device
tree contents, and restore the existing support for the LEDs.
Standard bindings apply, mapping "reg" property to LED index, with
addition of "qca,led-mode" property, which selects HW (0) or SW (1)
mode, defaulting to HW mode.
Lech Perczak [Tue, 4 Feb 2025 20:14:45 +0000 (21:14 +0100)]
kernel: ar8327: fix active-low LED initialization
Switch LEDs configured as active-low remain low instead of high upon
initialization, because in ar8327_leds_init, no distinction is made with
regards to LED pattern based on active_low property - only whether HW
mode is active. Select the proper initial pattern based also on
active_low to fix that.
While at that, simplify the equation ruling pattern selection for
setting brightness, avoiding unnecessary binary XOR operation, not
really valid for 'bool' type.
ramips: add missing LEDs and modem control for ASUS 4G-AX56
Add missing LEDs and modem control for ASUS 4G-AX56
- wifi2.4G white
- wifi5G white
- wan two-coloured, white and red
- modem four-coloured white, blue, yellow and red
change
label = "xxxx:modem";
to
color = <LED_COLOR_ID_xxxx>;
function = LED_FUNCTION_MOBILE;
- rssi-1 white
- rssi-2 white
- rssi-3 white
Felix Fietkau [Fri, 7 Mar 2025 17:20:23 +0000 (18:20 +0100)]
unetmsg: add unet pub/sub message broker based on ubus
This service automatically establishes connections to any hosts that are members
of the same unet network, and allows publish/subscribe exchanges via ubus channels.
Felix Fietkau [Fri, 14 Mar 2025 13:05:06 +0000 (14:05 +0100)]
provision: add script for managing device provisioning data
This is useful for keeping specific data on a device across factory reset.
It uses a separate partition (only UBI supported at the moment) to store
the data. The primary use case is storing sensitive data like cryptographic
keys for maintaining a device as part of a network.
Felix Fietkau [Sat, 15 Mar 2025 20:55:10 +0000 (21:55 +0100)]
unetd: cli: add support for changing network password
This does not actually create a new private key. Instead, the salt is replaced,
and a xor key is generated which when merged with the key derived from the new
password transforms into the original private key.
With upstream changes hitting kernel 6.4 the dtb for u7623 ends up with
both mac (gmac) disabled, since this is now the default status in
mt7623.dtsi. Fix this by including mt7623a.dtsi (which already has all
necessary bits) and enabling all revlevant ports. This will also do
a side hustle of assigning proper clocks for power controller and
specifying proper power domain for few devices.
Apparently U-Boot will discard whole node if requested pin function is
unknown to the driver. This resulted in inability to interact with
U-Boot on the said board, as U-Boot always assumed the recovery key
pressed and issued recovery procedure. Log snippet:
Recovery procedure also booted recovery image, which didn't affect much
the 23.05.x release, since the root fs argument was valid, so changes
persisted. But as 24.10.x hit with fitblk, the board will boot only
recovery image (initramfs) because of default bootargs will reset on each
boot and U-Boot provided bootargs took precedence.
Like #15865, it seems that gmac0 does not rename eth0 to dsa until after the
switch ports are initialized, leading to a name collision (error -17 = EEXIST).
This patch follows #17062 by using openwrt,netdev-name to fix the collision.
Til Kaiser [Sun, 16 Feb 2025 17:03:01 +0000 (18:03 +0100)]
x86: add Supermicro SuperServer E302-9D
This adds a default network configuration for the
Supermicro SuperServer SYS-E302-9D by adding all
onboard network ports to the default `lan` interface.
The network ports `eth0` till `eth3` use the `igb`
driver, whereas `eth4` till `eth7` use `i40e`.
Also, increase those priorities because I think they
are too low since there is currently no "room" for
built-in network device drivers.
That can cause interface order, i.e., name inconsistencies,
when Mellanox ConnectX cards are inserted or removed.
Lech Perczak [Mon, 3 Feb 2025 23:24:21 +0000 (00:24 +0100)]
ath79: ZTE MF281: use specific board definition file for qca9888
Using board definition file extracted from stock firmware yields 50%
throughput improvement in RX direction under iperf3 test.
Make the device use temporary files from firmware_qca-wireless.git
temporarily, as well as select the specific variant in the device tree
files. The device uses same board file as the MF286C.
Lech Perczak [Sun, 24 Nov 2024 23:36:24 +0000 (00:36 +0100)]
ath79: support ZTE MF286C
ZTE MF286 is an indoor LTE category 12 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.
Software-wise it's compatible with previous MF286A, save for different
5GHz Wi-Fi board definition file, requiring a separate image.
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: MDM9250-based category 12 internal LTE modem
in extended mini-PCIE form factor, with 5 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.
- Label MAC device: eth0
Internal modem of MF286C is supported via uqmi.
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
ath79: support ZTE MF286C
STEP 1: 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:
(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.
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:
- Connect your USB-UART adapter
- Dump stock firmware located on stock kernel and ubi partitions:
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:
Differences might indicate that this is NOT a MF286C 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 MF286C 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: Actual installation:
- Set your computer IP to 192.168.1.22/24
- scp the sysupgrade image to the device:
STEP 4: 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:
- Boot to initramfs as in step 3:
- Completely detach ubi0 partition using ubidetach /dev/ubi0_0
- 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 /tmp/mtd4_kernel.bin
rm /tmp/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:
- If the write was correct, force a device reboot with
reboot -f
Quirks and known issues
- 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.
- 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.
Lech Perczak [Sun, 24 Nov 2024 23:22:27 +0000 (00:22 +0100)]
ath79: ZTE MF286A: use specific board definition file for qca9888
Using board definition file extracted from stock firmware yields 50%
throughput improvement in RX direction under iperf3 test.
Make the device use temporary files from firmware_qca-wireless.git
temporarily, as well as select the specific variant in the device tree
files.
SoC: MediaTek MT7621
Flash: 16 MB SPI Flash
RAM: 128 MB RAM
Ethernet:
2x 1G RJ45 ports
WLAN:
2.4GHz: MediaTek MT7603E
5GHz: MediaTek MT7613BE
LEDs: Red and blue status lights
Power: 12V DC
UART: 3.3V, 115200 baud, 8N1, like printed on silkscreen (GND,TX,RX,3.3V)
MAC addresses
-------------
+---------+-------------------+
| | MAC example |
+---------+-------------------+
| LAN | 80:3F:5D:xx:xx:72 |
| WAN | 80:3F:5D:xx:xx:73 |
| WLAN 2g | 80:3F:5D:xx:xx:74 |
| WLAN 5g | 80:3F:5D:xx:xx:75 |
+---------+-------------------+
Installation:
The firmware can be flashed via the U-Boot recovery web interface.
To access it, hold the reset button while powering on the device.
U-Boot recovery web interface is then avaiable at 192.168.10.1.
Alternatively, the image can be loaded using the U-Boot serial interface and TFTP.
Adds latest 6.6 patches from the Raspberry Pi repository.
These patches were generated from:
https://github.com/raspberrypi/linux/commits/rpi-6.6.y/
With the following command:
git format-patch -N v6.6.83..HEAD
(HEAD -> 08d4e8f52256bd422d8a1f876411603f627d0a82)
Hauke Mehrtens [Sun, 19 Jan 2025 18:47:23 +0000 (19:47 +0100)]
kernel: Activate CONFIG_NET_SWITCHDEV in generic config
The CONFIG_NET_SWITCHDEV option is needed by CONFIG_DSA and some other
options. It is boolean, we have to compile it into the kernel it self.
Activate it for all targets in the generic configuration, it is already
activated for most of them. This allows to install DSA drivers as a
module.
On the ramips/mt7620 target the kernel would grown by 4.5kB.
For some small targets which do not support a DSA switch by default the
option is deactivated.
This has several advantages:
* reduction in the size of the kernel and the complete image. Individual
devices only need two of the four binaries. In combination with the second
commit it reduces kernel size by 64.2 kB and image size by 22.8 kB,
* the option to extend this package with firmware for future SoCs,
* combining the kernel and binary blobs with another licence may not be
fully compatible with the licence used by Linux. The current PHY firmware
is built into the kernel. This comit converts it to a package.
The next commit converts the firmware used by this driver
to a package. Due to the fact that the driver is shipped
as a package the firmware is already available.
Thomas Richard [Mon, 24 Feb 2025 10:15:51 +0000 (11:15 +0100)]
optee-os.mk: override default PATH to not use hostpkg python
In some cases hostpkg python from packages feed is used (hostpkg has higher
priority in PATH) which causes build failure (cryptography module is
missing). So override PATH to not use hostpkg python.
Eric Schäfer [Mon, 10 Mar 2025 21:36:18 +0000 (22:36 +0100)]
ramips: add support for Zyxel LTE7490-M904
The Zyxel LTE7490-M904 is an 802.3at PoE powered LTE outdoor (IP68) CPE
with integrated directional antennas.
Specifications:
- SoC: MediaTek MT7621AT
- RAM: 256 MB
- Flash: 128 MB MB NAND (MX30LF1G18AC)
- WiFi: MediaTek MT7603E 802.11b/g/n
- Switch: 1 LAN port (1 Gbps)
- LTE/3G/2G: Quectel EG18-EA LTE-A Cat. 18 connected by USB3 to SoC
- SIM: 1 micro-SIM slots under transparent cover
- Buttons: Reset, WLAN under same cover
- LEDs: Multicolour green/red/amber under same cover (visible)
- Power: 802.3at PoE via LAN port
The device is built as an outdoor ethernet to LTE bridge or router.
The wifi interface is intended for installation and/or temporary
management purposes only.
Remove the SIM/button/LED cover and 12 screws holding the back plate
and antenna cover together. Be careful with the cables.
Installation from OEM web GUI:
- Log in as "admin" on OEM web GUI
- Upload OpenWrt initramfs-recovery.bin image on the
Maintenance -> Firmware page
- Wait for OpenWrt to boot and ssh to root@192.168.1.1
- Sysupgrade to the OpenWrt sysupgrade image and reboot
For more details about flashing see: 2449a63208 (ramips: mt7621: Add support for ZyXEL NR7101, 2021-04-19)
Main porting work done by Ernesto Castellotti <ernesto@castellotti.net>: bf1c12f68b (ramips: add support for ZyXEL LTE7490-M904, 2023-12-20)
Roy H [Tue, 11 Mar 2025 04:35:34 +0000 (11:35 +0700)]
ath79: add support for Longdata APS256
Forum discussion : https://forum.openwrt.org/t/aps-256va-help-for-identification/143653/52
Specification:
Power: 12-36V input via 5,5/2,1 DC barrel jack, or 5V Micro USB-B
CPU: Atheros AR9344 rev 2
RAM: 128MB
Flash: 16MB
WI-Fi: 2.4GHz
Fast Ethernet: 1 WAN and 2 LAN
USB: 2 x USB-A, 1 x micro-USB-B (for power input)
WWAN: 3G modem via extended mini-PCIE form factor (can be replaced with Wifi 5GHz card)
The device come with custom openwrt BB an CC.
Because of limited LAN port, I disable GMAC0, so the WAN port can be connected to GMAC1 and function as LAN port as well.
Enable ssh access and Backup:
1. open router admin page via LAN cable
2. browse 192.168.111.1:8000
3. login with password 123456
4. click wifi icon on top menu
5. change the path at the end of the url (after random hash) with /admin/system/flashops
it will looks like this:
http://192.168.111.1:8000/cgi-bin/luci/;stok=29698152cf64c980177a04f86c99ea0d/admin/system/flashops
(the hash after "stok=" will be different)
6. restore the config with this modified backup (can be created manually by changing dropbear config to allow ssh)
https://drive.google.com/file/d/1Vs-k7DHBSRZFfkxv1cMOmgAPZfB-RUen/view?usp=sharing
7. now you can login to ssh with root user and 123456 password, and backup all partition and upgrade firmware
!!! BACKUP EVERY PARTITION !!!
Flashing instructions:
- Flash directly from factory web interface accessed from "Enable ssh access" step 5
Shiji Yang [Thu, 20 Feb 2025 12:03:26 +0000 (20:03 +0800)]
uboot-mediatek: move custom uart config symbol to board defconfigs
This helps to solve the issue of waiting for "SERIAL_RX_BUFFER_SIZE"
input when enabling verbose log output option (V=s).
Fixes: https://github.com/openwrt/openwrt/issues/18036 Signed-off-by: Shiji Yang <yangshiji66@qq.com> Link: https://github.com/openwrt/openwrt/pull/18043 Signed-off-by: Robert Marko <robimarko@gmail.com>
MAC Addresses:
- There is one on the label, e.g. xx:xx:xx:xx:xx:A4
- LAN (bottom connector) is the same as the label, e.g. xx:xx:xx:xx:xx:A4
- WAN (top connector) is label + 1, e.g. xx:xx:xx:xx:xx:A5
- WLAN (2.4G) is the same as the label, e.g. xx:xx:xx:xx:xx:A4
- WLAN (5G) is label + 2, e.g. xx:xx:xx:xx:xx:A6
UART:
- is available via the pin holes on the board
- The pinout is printed to the board: P: VCC, G: GND, R: RX, T:TX
- RX and TX require solder bridges to be installed
- Do NOT connect VCC
- Settings: 3.3V, 115200, 8N1
GPIO:
- There are two LEDs: Red (GPIO 4) and White (GPIO 0)
- There are two buttons: Reset (GPIO 11) and WPS (GPIO 5)
Migration to OpenWrt:
- Download the migration image from the Cudy website (it should be available as soon as OpenWrt officially supports the device)
- Connect computer to LAN (bottom connector) and flash the migration image via OEM web interface
- OpenWrt is now accessible via 192.168.1.1
Revert back to OEM firmware:
- Set up a TFTP server on IP 192.168.1.88 and connect to the WAN port (upper port)
- Provide the Cudy firmware as recovery.bin in the TFTP server
- Press the reset button while powering on the device
- Recovery process is started now
- When recovery process is done, OEM firmware is accessible via 192.168.10.1 again
General information:
- No possibility to load a initramfs image via U-Boot because there is no option to interrupt U-Boot
This allows the network interface naming to be stable, free from any
possible interaction from external USB network devices that might
claim usb* interface names.
(This was a real problem I encountered with a nanopi R6S device and
an external rtl8152 usb3 network controller - the USB controller would
claim the eth1 name, causing much confusion).
rockchip: set network IRQ affinity to fast CPU cores
The nanopi R6S, R6C and nanopc T6 platforms are based on rk3588(s) SoC,
which has fast and slow CPU cores. Set up network interrupt affinity to be
on the fast CPU cores by default. This is similar to the way this was
already configured on nanopi R4S.
rockchip: show boot stages on nanopi R6 system LED
Set up openwrt to show boot progress on the nanopi R6S or R6C system LED.
The LED blinking states indicate the boot stage. The LED is defined as
a power LED, but can still be set to heartbeat in /etc/config/system
after the system is done booting.
Rename 852-stable-bus-mhi-host-pci_generic-constify-modem_telit_fn980_.patch
to 852-v6.9-stable-bus-mhi-host-pci_generic-constify-modem_telit_fn980.patch
because it is used since kernel 6.9-rc1 (https://lore.kernel.org/lkml/Zfwv2y7P7BneKqMZ@kroah.com/).
Mauri Sandberg [Thu, 30 Jan 2025 16:08:25 +0000 (18:08 +0200)]
ramips: Cleanup Genexis EX400 upgrade script
The code can be made more efficient by not extracting the sysupgrade.tar but
rather just querying for the filesize within the archive. Resorting to
manual update of UBI volume is extra work too, setting CI_KERNPART=rootfs_0
is enough.
Suggested-by: Andreas Gnau <andreas.gnau@iopsys.eu> Signed-off-by: Mauri Sandberg <maukka@ext.kapsi.fi> Link: https://github.com/openwrt/openwrt/pull/17806 Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
local access.
- Mitigations for INTEL-SA-01079 (CVE-2024-23918)
Potential security vulnerabilities in some Intel Xeon processors
using Intel SGX may allow escalation of privilege. Intel disclosed
that some processor models were already fixed by a previous
microcode update.
- Updated mitigations for INTEL-SA-01097 (CVE-2024-24968)
Improper finite state machines (FSMs) in hardware logic in some
Intel Processors may allow an privileged user to potentially enable a
denial of service via local access.
- Mitigations for INTEL-SA-01103 (CVE-2024-23984)
A potential security vulnerability in the Running Average Power Limit
(RAPL) interface for some Intel Processors may allow information
disclosure. Added mitigations for more processor models.
* Updated Microcodes:
sig 0x000806f8, pf_mask 0x87, 2024-06-20, rev 0x2b000603, size 588800
sig 0x000806f7, pf_mask 0x87, 2024-06-20, rev 0x2b000603
sig 0x000806f6, pf_mask 0x87, 2024-06-20, rev 0x2b000603
sig 0x000806f5, pf_mask 0x87, 2024-06-20, rev 0x2b000603
sig 0x000806f4, pf_mask 0x87, 2024-06-20, rev 0x2b000603
sig 0x00090672, pf_mask 0x07, 2024-05-29, rev 0x0037, size 224256
sig 0x00090675, pf_mask 0x07, 2024-05-29, rev 0x0037
sig 0x000b06f2, pf_mask 0x07, 2024-05-29, rev 0x0037
sig 0x000b06f5, pf_mask 0x07, 2024-05-29, rev 0x0037
sig 0x000906a3, pf_mask 0x80, 2024-06-03, rev 0x0435, size 223232
sig 0x000906a4, pf_mask 0x80, 2024-06-03, rev 0x0435
sig 0x000a06a4, pf_mask 0xe6, 2024-08-02, rev 0x0020, size 138240
sig 0x000b06a2, pf_mask 0xe0, 2024-05-29, rev 0x4123, size 220160
sig 0x000b06a3, pf_mask 0xe0, 2024-05-29, rev 0x4123
sig 0x000b06a8, pf_mask 0xe0, 2024-05-29, rev 0x4123
sig 0x000c06f2, pf_mask 0x87, 2024-06-20, rev 0x21000283, size 560128
sig 0x000c06f1, pf_mask 0x87, 2024-06-20, rev 0x21000283
* source: update symlinks to reflect id of the latest release, 20241112
* Update changelog for 3.20240910.1 and 3.20240813.1 with new information:
INTEL-SA-1103 was addressed by 3.20240813.1 for some processor models,
and not by 3.20240910. INTEL-SA-1079 was addressed by 3.20240910.1 for
some processor models.
-- Henrique de Moraes Holschuh <hmh@debian.org> Thu, 14 Nov 2024 15:37:40 -0300
* New upstream microcode datafile 20241029
- Not relevant for operating system microcode updates
- Only when loaded from firmware, this update fixes the critical,
potentially hardware-damaging errata RPL061: Incorrect Internal
Voltage Request on Raptor Lake (Core 13th/14th gen) Intel
processors.
* Updated Microcodes:
sig 0x000b0671, pf_mask 0x32, 2024-08-29, rev 0x012b, size 211968
-- Henrique de Moraes Holschuh <hmh@debian.org> Thu, 14 Nov 2024 14:49:03 -0300
* New upstream microcode datafile 20240910 (closes: #1081363)
- Mitigations for INTEL-SA-01097 (CVE-2024-24968)
Improper finite state machines (FSMs) in hardware logic in some
Intel Processors may allow an privileged user to potentially enable a
denial of service via local access.
- Fixes for unspecified functional issues on several processor models
- The processor voltage limit issue on Core 13rd/14th gen REQUIRES A
FIRMWARE UPDATE. It is present in this release for sig 0xb0671, but
THE VOLTAGE ISSUE FIX ONLY WORKS WHEN THE MICROCODE UPDATE IS LOADED
THROUGH THE FIT TABLE IN FIRMWARE. Contact your system vendor for a
firmware update that includes the appropriate microcode update for
your processor.
* Updated Microcodes:
sig 0x00090672, pf_mask 0x07, 2024-02-22, rev 0x0036, size 224256
sig 0x00090675, pf_mask 0x07, 2024-02-22, rev 0x0036
sig 0x000b06f2, pf_mask 0x07, 2024-02-22, rev 0x0036
sig 0x000b06f5, pf_mask 0x07, 2024-02-22, rev 0x0036
sig 0x000906a3, pf_mask 0x80, 2024-02-22, rev 0x0434, size 222208
sig 0x000906a4, pf_mask 0x80, 2024-02-22, rev 0x0434
sig 0x000a06a4, pf_mask 0xe6, 2024-06-17, rev 0x001f, size 137216
sig 0x000b0671, pf_mask 0x32, 2024-07-18, rev 0x0129, size 215040
sig 0x000b06a2, pf_mask 0xe0, 2024-02-22, rev 0x4122, size 220160
sig 0x000b06a3, pf_mask 0xe0, 2024-02-22, rev 0x4122
sig 0x000b06a8, pf_mask 0xe0, 2024-02-22, rev 0x4122
sig 0x000b06e0, pf_mask 0x19, 2024-03-25, rev 0x001a, size 138240
* Update changelog for 3.20240813.1 with new information
* Update changelog for 3.20240514.1 with new information
* source: update symlinks to reflect id of the latest release, 20240910
* New upstream microcode datafile 20240813 (closes: #1078742)
- Mitigations for INTEL-SA-01083 (CVE-2024-24853)
Incorrect behavior order in transition between executive monitor and SMI
transfer monitor (STM) in some Intel Processors may allow a privileged
user to potentially enable escalation of privilege via local access.
- Mitigations for INTEL-SA-01118 (CVE-2024-25939)
Mirrored regions with different values in 3rd Generation Intel Xeon
Scalable Processors may allow a privileged user to potentially enable
denial of service via local access.
- Mitigations for INTEL-SA-01100 (CVE-2024-24980)
Protection mechanism failure in some 3rd, 4th, and 5th Generation Intel
Xeon Processors may allow a privileged user to potentially enable
escalation of privilege via local access.
- Mitigations for INTEL-SA-01038 (CVE-2023-42667)
Improper isolation in the Intel Core Ultra Processor stream cache
mechanism may allow an authenticated user to potentially enable
escalation of privilege via local access. Intel disclosed that some
processor models were already fixed by the previous microcode update.
- Mitigations for INTEL-SA-01046 (CVE-2023-49141)
Improper isolation in some Intel Processors stream cache mechanism may
allow an authenticated user to potentially enable escalation of
privilege via local access. Intel disclosed that some processor models
were already fixed by the previous microcode update.
- Mitigations for INTEL-SA-01079 (CVE-2024-23918)
Potential security vulnerabilities in some Intel Xeon processors
using Intel SGX may allow escalation of privilege. Intel released this
information during the full disclosure for the 20241112 update.
Processor signatures 0x606a6 and 0x606c1.
- Mitigations for INTEL-SA-01103 (CVE-2024-23984)
A potential security vulnerability in the Running Average Power Limit
(RAPL) interface for some Intel Processors may allow information
disclosure. Intel released this information during the full disclosure
for the 20240910 update. Processor signatures 0x5065b, 0x606a6,
0x606c1.
- Fix for unspecified functional issues on several processor models
- Fix for errata TGL068/ADL075/ICL088/... "Processor may hang during a
microcode update". It is not clear which processors were fixed by this
release, or by one of the microcode updates from 2024-05.
- Mitigations for INTEL-SA-01213 (CVE-2024-36293)
Improper access control in the EDECCSSA user leaf function for some
Intel Processors with Intel SGX may allow an authenticated user to
potentially enable denial of service via local access. Intel released
this information during the full disclosure for the 20250211 update.
Processor signature 0x906ec (9th Generation Intel Core processor).
* Updated microcodes:
sig 0x00050657, pf_mask 0xbf, 2024-03-01, rev 0x5003707, size 39936
sig 0x0005065b, pf_mask 0xbf, 2024-04-01, rev 0x7002904, size 30720
sig 0x000606a6, pf_mask 0x87, 2024-04-01, rev 0xd0003e7, size 308224
sig 0x000606c1, pf_mask 0x10, 2024-04-03, rev 0x10002b0, size 300032
sig 0x000706e5, pf_mask 0x80, 2024-02-15, rev 0x00c6, size 114688
sig 0x000806c1, pf_mask 0x80, 2024-02-15, rev 0x00b8, size 112640
sig 0x000806c2, pf_mask 0xc2, 2024-02-15, rev 0x0038, size 99328
sig 0x000806d1, pf_mask 0xc2, 2024-02-15, rev 0x0052, size 104448
sig 0x000806e9, pf_mask 0xc0, 2024-02-01, rev 0x00f6, size 106496
sig 0x000806e9, pf_mask 0x10, 2024-02-01, rev 0x00f6, size 106496
sig 0x000806ea, pf_mask 0xc0, 2024-02-01, rev 0x00f6, size 105472
sig 0x000806eb, pf_mask 0xd0, 2024-02-01, rev 0x00f6, size 106496
sig 0x000806ec, pf_mask 0x94, 2024-02-05, rev 0x00fc, size 106496
sig 0x00090661, pf_mask 0x01, 2024-04-05, rev 0x001a, size 20480
sig 0x000906ea, pf_mask 0x22, 2024-02-01, rev 0x00f8, size 105472
sig 0x000906eb, pf_mask 0x02, 2024-02-01, rev 0x00f6, size 106496
sig 0x000906ec, pf_mask 0x22, 2024-02-01, rev 0x00f8, size 106496
sig 0x000906ed, pf_mask 0x22, 2024-02-05, rev 0x0100, size 106496
sig 0x000a0652, pf_mask 0x20, 2024-02-01, rev 0x00fc, size 97280
sig 0x000a0653, pf_mask 0x22, 2024-02-01, rev 0x00fc, size 98304
sig 0x000a0655, pf_mask 0x22, 2024-02-01, rev 0x00fc, size 97280
sig 0x000a0660, pf_mask 0x80, 2024-02-01, rev 0x00fe, size 97280
sig 0x000a0661, pf_mask 0x80, 2024-02-01, rev 0x00fc, size 97280
sig 0x000a0671, pf_mask 0x02, 2024-03-07, rev 0x0062, size 108544
sig 0x000a06a4, pf_mask 0xe6, 2024-04-15, rev 0x001e, size 137216
* source: update symlinks to reflect id of the latest release, 20240813
* postinst, postrm: switch to dpkg-trigger to run update-initramfs
-- Henrique de Moraes Holschuh <hmh@debian.org> Thu, 15 Aug 2024 14:41:50 -0300
projects / thirdparty / openwrt.git / log
