The AQL limit for buffered broadcast packets is higher than the maximum
total pending airtime limit. This can get unicast data stuck whenever there
is too much pending broadcast data. Fix this by excluding broadcast AQL from
the total limit.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Fixes libssh, which requires it. Bump ABI_VERSION, since enabling this
option affects data structures in mbedtls include files.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Cleanup uart1-related node on ELECOM WAB-I1750-PS and enable it for
"SERIAL" port on the case.
"SERIAL" port can be used for OpenWrt console by adding the following
line to /etc/inittab and rebooting:
ttyATH1::askfirst:/usr/libexec/login.sh
Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
Add aliases with "serialN = &uartN;" of uart0/1 on QCA955x SoCs to
qca955x.dtsi, to enable uart1 on Linux Kernel.
without this:
[ 0.342915] ar933x-uart 18500000.uart: unable to get alias id, err=-19
Additionally, remove "serial0 = &uart;" alias from QCA955x device
dts/dtsi files.
Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
Add HighSpeed UART support to QCA955x series SoCs as a secondary UART
(uart1). This UART is compatible with qca,ar9330-uart.
Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
Rename the DT label of the primary UART on Qualcomm Atheros QCA955x
series SoCs to "uart0" from "uart" for the preparation to add HighSpeed
UART (uart1) support.
Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
with 6.1, the kernel no longer fitted into the 16 MiB and
kicking down the can and increasing KERNEL_SIZE to 20 MiB
didn't help as the device failed to boot.
Using 'kernel-bin | gzip | uimage gzip' didn't work since the
uboot does not have enough heap to decompress these big kernels.
And finally playing around with uboot was more a hassle than
converting this device to take the simpleImage-boot-route in
the future.
Note: The device now takes even longer on the first boot-up after
the flash due to JFFS2 initializing all the remaining flash.
Be prepared to wait up to 10 minutes before the green status LED
stops blinking and will shine a solid green!
(On the plus site: the device now has ~10 MiB of additional
space for rootfs+rootfs_data).
Note2: This patch includes a kernel patch refresh.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Kernel and initramfs size grows. Now uncompressed initramfs image and
regular kernel image overlaps configured area:
Initramfs:
WRAP arch/powerpc/boot/simpleImage.br200-wp
INFO: Uncompressed kernel (size 0x1428688) overlaps the address of the
wrapper(0x1000000)
INFO: Fixing the link_address of wrapper to (0x1500000)
WRAP arch/powerpc/boot/simpleImage.tl-wdr4900-v1
INFO: Uncompressed kernel (size 0x1428688) overlaps the address of the
wrapper(0x1000000)
INFO: Fixing the link_address of wrapper to (0x1500000)
WRAP arch/powerpc/boot/simpleImage.ws-ap3715i
INFO: Uncompressed kernel (size 0x1428688) overlaps the address of the
wrapper(0x1000000)
INFO: Fixing the link_address of wrapper to (0x1500000)
Regular image:
WRAP arch/powerpc/boot/simpleImage.br200-wp
INFO: Uncompressed kernel (size 0x10e0688) overlaps the address of the
wrapper(0x1000000)
INFO: Fixing the link_address of wrapper to (0x1100000)
WRAP arch/powerpc/boot/simpleImage.tl-wdr4900-v1
INFO: Uncompressed kernel (size 0x10e0688) overlaps the address of the
wrapper(0x1000000)
INFO: Fixing the link_address of wrapper to (0x1100000)
WRAP arch/powerpc/boot/simpleImage.ws-ap3715i
INFO: Uncompressed kernel (size 0x10e0688) overlaps the address of the
wrapper(0x1000000)
INFO: Fixing the link_address of wrapper to (0x1100000)
Let's change wrapper address to safe value.
Tested on: TL-WDR4900, BR200-WP
Signed-off-by: Pawel Dembicki <paweldembicki@gmail.com>
This patch replaces 'of_flat_dt_is_compatible' with 'of_machine_is_compatible'.
The TL-WDR4900 platform file won't compile in the 6.1 kernel. The platform
files for the rest of the routers have been reworked or based on newer
solutions.
Let's make the TL-WDR4900 consistent with them.
Signed-off-by: Pawel Dembicki <paweldembicki@gmail.com>
At start I only set qualcommax to use the Crypto Extensions optimized
version of SHA256 as I knew it supports the optional Crypto Extensions.
However, after looking into the tree there are more targets/subtargets
that I could find at least a specification sheet that says the support
Cryptographic Extensions, so lets add them.
Signed-off-by: Robert Marko <robimarko@gmail.com>
At start I only set qualcommax to use the Crypto Extensions optimized
version of SHA1 as I knew it supports the optional Crypto Extensions.
However, after looking into the tree there are more targets/subtargets
that I could find at least a specification sheet that says the support
Cryptographic Extensions, so lets add them.
Signed-off-by: Robert Marko <robimarko@gmail.com>
This fixes the following compile problem:
````
CC arch/mips/kernel/vpe-mt.o
arch/mips/kernel/vpe-mt.c: In function 'vpe_run':
arch/mips/kernel/vpe-mt.c:32:23: error: unused variable 'physical_memsize' [-Werror=unused-variable]
32 | unsigned long physical_memsize = 0L;
| ^~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
make[6]: *** [scripts/Makefile.build:289: arch/mips/kernel/vpe-mt.o] Error 1
````
physical_memsize was removed from upstream kernel, see:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=linux-5.15.y&id=5b4f6c5ff65c8551018ccea40c569afd759734c0
Fixes: 387fde0da0 ("kernel: bump 5.15 to 5.15.150")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
KASAN has supported more architectures, such as ARM, PPC32 and RISC-V 64.
Enable KASAN option for those architectures.
Signed-off-by: Qingfang Deng <dqfext@gmail.com>
Commit 2d63d42f5e ("mediatek: convert to new LED color/function
format where possible") leaves Xiaomi Redmi AX6000 un-converted,
the two LEDs become dead.
Now, LEDs are alive again.
Fixes: 2d63d42f5e ("mediatek: convert to new LED color/function
format where possible")
Signed-off-by: Furong Xu <xfr@outlook.com>
ath25 has been on life support for the last couple of releases, eventually
leading to marking it as source-only in 2023.
It has been basically only touched to do a kernel bump so that we can make
the new OpenWrt release which was a challenge due to small RAM amount.
However, with the attempt of kernel 6.1 update it turns out that kernel
cannot even finish booting due to RAM constraints, so its time to let this
target go.
Signed-off-by: Robert Marko <robimarko@gmail.com>
It was announced [1] that the original staging repositories are no longer
used for staging of new firmware binaries. And that the old repository will
be removed [2] in June 2024.
The ath11k-firmware package must therefore point to the new repository
before the old one is no longer accessible.
[1] https://lore.kernel.org/r/bac97f31-4a70-4c4c-8179-4ede0b32f869@quicinc.com
[2] 8d2cc160f3
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Kernel has an ASM optimized version of SHA512 that was ported from
OpenSSL, so lets package it as it provides significant perfomance
improvement compared to the generic implementation.
There is a Cryptographic Extension based version as well, but that relies
on ARMv8.2 ISA which I am not aware any of the OpenWrt supported SoC-s use.
Signed-off-by: Robert Marko <robimarko@gmail.com>
Kernel has optimized version of SHA2(224 and 256) using the ARMv8 Crypto
Extensions, so lets package it.
Use it by default for qualcommax as it uses Cortex-A53 core and has ARMv8
CE extensions present.
Signed-off-by: Robert Marko <robimarko@gmail.com>
Kernel has an ASM optimized version of SHA256 that was ported from
OpenSSL, so lets package it as it provides significant perfomance
improvement compared to the generic implementation.
Signed-off-by: Robert Marko <robimarko@gmail.com>
Kernel has optimized version of SHA1 using the ARMv8 Crypto Extensions,
so lets package it.
Use it by default for qualcommax as it uses Cortex-A53 core and has ARMv8
CE extensions present.
Signed-off-by: Robert Marko <robimarko@gmail.com>
SHA3 is now required by jitterentropy_rng in kernel 6.6, so lets start
preparing by packaging SHA3 support as its supported in 5.15 and 6.1
kernels as well.
AFAIK, only ARMv8.2 has a crypto extension for SHA3, however I am not aware
of any SoC we support that uses ARMv8.2 ISA so its not enabled currently.
Signed-off-by: Robert Marko <robimarko@gmail.com>
1. Return error if any step of generating tar file fails
2. Use pipefail to avoid calling "gzip" if tar failed
Fixes: e36cc53092 ("base-files: sysupgrade: use tar helper to include installed_packages.txt")
Reported-by: Luiz Angelo Daros de Luca <luizluca@gmail.com>
Cc: Luiz Angelo Daros de Luca <luizluca@gmail.com>
Cc: Jo-Philipp Wich <jo@mein.io>
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Netgear WAX214 is a 802.11 ax dual-band AP
with PoE. (similar to Engenius EWS357APV3)
Specifications:
• CPU: Qualcomm IPQ6010 Quad core Cortex-A53
• RAM: 512MB of DDR3
• Storage: 128MB NAND (Macronix MX30UF1G18AC)
• Ethernet: 1x 1G RJ45 port (QCA8072) PoE
• WIFI:
2.4GHz: Qualcomm QCN5022 2x2 802.11b/g/n/ax 574 Mbps PHY rate
5GHz: Qualcomm QCN5052 2x2 802.11a/b/g/n/ac/ax 1201 PHY rate
• LEDs:
4 x GPIO-controlled LEDs
- 1 Power LED (orange)
- 1 LAN LED (blue)
- 1 WIFI 5g LED (blue)
- 1 WIFI 2g LED (blue)
black_small_square Buttons: 1x soft reset
black_small_square Power: 12V DC jack or PoE (802.3af )
An populated serial header is onboard, format is
1.25mm 4p (DF13A-4P-1.25H)
RX/TX is working, bootwait is active, secure boot is not
enabled.
The root password of the stock firmware is unknown,
but failsafe mode can be entered to reset the password.
Installation Instructions:
- obtain serial access
- stop auto boot (press "4", Entr boot command line
interface)
- setenv active_fw 0 (to boot from the primary rootfs,
or set to 1 to boot from the secondary rootfs
partition)
- saveenv
- tftpboot the initramfs image
- bootm
- copy
openwrt-qualcommax-ipq60xx-netgear_wax214-squashfs-factory.ubi
to the device
- write the image to the NAND:
- cat /proc/mtd and look for rootfs partition (should
be mtd11,
or mtd12 if you choose active_fw 1)
- ubiformat /dev/mtd11 -f -y
openwrt-qualcommax-ipq60xx-netgear_wax214-squashfs-factory.ubi
- reboot
Note: the firmware is senao-based. But I was unable to build
a valid senao-header into the image.
Maybe they changed the header format and senaoFW isn't
working any more.
Signed-off-by: Dirk Buchwalder <buchwalder@posteo.de>
The dependency can't be satisfied when building using the SDK, breaking
package builds. As the staging and bin dirs are distributed with the SDK
archive, ignoring the dependency is fine when SDK is set.
Fixes: fbb924abff ("build: add $(STAGING_DIR) and $(BIN_DIR) ...")
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
Some platforms like Raspberry Pi require patching some backup files like
cmdline.txt in order to set the correct root PARTUUID.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
BCM63xx maintainance is a PITA since there's no full Device Tree support and
therefore every board needs it's own definition in board_bcm963xx.c
There's no DSA support for this board and there have been very few
contributions in the last years.
Moreover, BCM63xx SoCs >= BCM6358 are already supported in bmips target, so
any efforts to keep this target alive aren't worth the time.
Let's focus the community efforts on improving bmips instead.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Due to previous refactoring in sysupgrade, writing backup archives to
stdout became impossible since the hardcoded gzip output redirection
did not account for the `-` special case filename.
Fix this issue by substituting `-` with `/proc/self/fd/1` in the tar
archive output path variable.
Also remove a redundant `rm -f` of the target file path that occurs
before the file could've possibly been written.
Fixes: #14773
Fixes: https://github.com/openwrt/luci/issues/6961
Fixes: e36cc53092 ("base-files: sysupgrade: use tar helper to include installed_packages.txt")
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Felix Fietkau