Commit d5a05e69ac6e4 ("net: stmmac: Use hrtimer for TX coalescing") causes
high CPU usage due to hrtimer raw spin locks.
Fixes: #11676
Signed-off-by: Oskari Lemmela <oskari@lemmela.net>
[ renumber and rename revert patch ]
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
There's a typo in here: board_name is a function, not a variable. This
issue was pointed out on the OpenWrt forum.
Closes: #13409
Reviewed-by: Robert Marko <robimarko@gmail.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
(cherry picked from commit 7b78a19e6a)
Backport initial LEDs hw control support. Currently this is limited to
only rx/tx and link events for the netdev trigger but the API got
accepted and the additional modes are working on and will be backported
later.
Refresh every patch and add the additional config flag for QCA8K new
LEDs support.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
(cherry picked from commit 0a4b309f41)
1. Add new symbols to generic config
2. Bump kernel
Changelog: https://lore.kernel.org/stable/2023080818-groin-gradient-a031@gregkh/
All patches automatically rebased.
Signed-off-by: John Audia <therealgraysky@proton.me>
[Refreshed on top of OpenWrt 23.05]
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
(cherry picked from commit daed3322d3)
Move default cpufreq governor from ONDEMAND to PERFORMANCE. The temp
increase is just 2°C and Watt usage the change is minimal in the order
of additional millwatt. The SoC and krait in general looks to suffer for
some problem with cache scaling. To have better system stability, force
cpu freq and cache freq to the max value supported by the system. This
follows mvebu platform where cpufreq is broken and cause minimal
temp/watt increase.
User can still tweak the governor to ondemand using sysfs entry if
needed.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit 6f5ea752d7)
This reverts commit 60fc93b359.
Reenable devfreq and revert for both 5.15 and 6.1.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
(cherry picked from commit 37e4593213)
commit 0c45ad41e1 changes ipq806x usb kmod name
from usb-phy-qcom-dwc3 to phy-qcom-ipq806x-usb, so
use new name.
Signed-off-by: Yanase Yuki <dev@zpc.sakura.ne.jp>
(cherry picked from commit 9314744350)
This allows loading modules with large memory requirements, recently needed
while testing on armvirt/32. Past forum discussions [1] and bug reports [2]
also raised this and the ipq806x target already set it in response [3].
Given this increases kernel image size by only ~1KB, is generally useful on
multi-platform kernels, and enabled by default on upstream arm32 Linux, add
it to the generic config.
The setting has similar utility on arm64, is a requirement for KASLR, and
already enabled on most OpenWrt aarch64 targets, so pull this into the
top-level generic config.
[1]: https://forum.openwrt.org/t/vmap-allocation-for-size-442368-failed-use-vmalloc-size-to-increase-size/34545/7
[2]: https://github.com/openwrt/openwrt/issues/8282
[3]: f81e148eb6 ("ipq806x: update 4.19 kernel config").
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
(cherry picked from commit c2d194a34e)
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
should have been part of the ipq-wifi update.
Fixes: 8217f02a1c ("ipq-wifi: drop upstreamed board-2.bin")
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
The BDFs for the:
Aruba AP-365
Devolo Magic 2 WiFi next
Edgecore ECW5410
Edgecore OAP100
Extreme Networks WS-AP3915i
GL.iNet GL-A1300
GL.iNet GL-AP1300
GL.iNet GL-S1300
Linksys EA8300
Linksys WHW03v2
Nokia Wi4A AC400i
P&W R619AC
Pakedge WR-1
Qxwlan E2600AC C1
Sony NCP-HG100/Cellular
Teltonika RUTX10
ZTE MF18A
were upstreamed to the ath10k-firmware repository
and landed in linux-firmware.git.
Furthermore the BDFs for the:
8devices Habanero
8devices Jalapeno
Qxwlan E2600AC C2
have been updated.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
All targets are bumped to 5.15. Remove the old 5.10 patches, configs
and files using:
find target/linux -iname '*-5.10' -exec rm -r {} \;
Further, remove the 5.10 include.
Signed-off-by: Nick Hainke <vincent@systemli.org>
Drop arm override compile patch. The use is dubious and at times (2016)
it was never explained why this was required.
Krait CPU and Cortex-a15 may be similar but they have some intrinsic
difference. While very similar they differ a lot in additional feature
and also cache configuration.
Also these conflict with raid6 libraty compilation that use specific
workaround not compatible with cortex-a15 and produce compilation
warning.
Fix compilation warning:
cc1: error: switch '-mcpu=cortex-a15' conflicts with switch '-march=armv7-a+simd' [-Werror]
cc1: all warnings being treated as errors
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Fix compilation warning from mangle bootargs patch. Now that we flag
warning as error these cause compilation error.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Fix following error when building 32bit arm targets with kmod-crypto-sha512
ERROR: module '/home/user/openwrt/build_dir/target-arm_xscale_musl_eabi/linux-kirkwood_generic/linux-5.15.109/arch/arm/crypto/sha512-arm.ko' is missing.
Signed-off-by: Lu jicong <jiconglu58@gmail.com>
Since CONFIG_DYNAMIC_DEBUG is already managed via the KERNEL_DYNAMIC_DEBUG
setting in Config-kernel.in (default N), remove or disable it in target
configs which unconditionally enable it, along with the related setting
CONFIG_DYNAMIC_DEBUG_CORE. This saves several KB in the kernels for
ipq40xx, ipq806x, filogic, mt7622, qoriq, and sunxi.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
It appears that the refactor of the upgrade process for NAND devices resulted in the nand_do_upgrade_success step not being called for
devices using the linksys.sh script. As a result, configuration was
not preserved over sysupgrade steps.
This corrects a typo in the call of nand_do_upgrade_failed for ipq40xx
and ipq806x devices using the linksys.sh script.
Fixes: 8634c1080d ("ipq40xx: Fix Linksys upgrade, restore config step")
Fixes: 2715aff5df ("ipq806x: Fix Linksys upgrade, restore config step")
Signed-off-by: Michael Trinidad <trinidude4@hotmail.com>
It appears that the refactor of the upgrade process for NAND devices
resulted in the nand_do_upgrade_success step not being called for
devices using the linksys.sh script. As a result, configuration
was not preserved over sysupgrade steps.
This restores the preservation of configs for ipq806x devices using the
linksys.sh script. Other devices and targets have not been examined.
This commit uses the same functionality and terminology used in commit
8634c10 ("ipq40xx: Fix Linksys upgrade, restore config step")
Fixes: e25e6d8 ("base-files: fix and clean up nand sysupgrade code")
Tested-on: EA8500
Signed-off-by: Jacob Aharon <ah.jacob@gmail.com>
Manually rebased:
ramips/patches-5.10/810-uvc-add-iPassion-iP2970-support.patch
All other patches automatically rebased.
Signed-off-by: John Audia <therealgraysky@proton.me>
Hardware
--------
SoC: Qualcomm IPQ8065
RAM: 512 MB DDR3
Flash: 256 MB NAND (Macronix MX30UF2G18AC) (split into 2x128MB)
4 MB SPI-NOR (Macronix MX25U3235F)
WLAN: Qualcomm Atheros QCA9984 - 2.4Ghz
Qualcomm Atheros QCA9984 - 5Ghz
ETH: eth0 - POE (100Mbps in U-Boot, 1000Mbps in OpenWrt)
eth1 - (1000Mbps in both)
Auto-negotiation broken on both.
USB: USB 2.0
LED: 5G, 2.4G, ETH1, ETH2, CTRL, PWR (All support green and red)
BTN: Reset
Other: SD card slot (non-functional)
Serial: 115200bps, near the Ethernet transformers, labeled 9X.
Connections from the arrow to the 9X text:
[NC] - [TXD] - [GND] - [RXD] - [NC]
Installation
------------
0. Connect to the device
Plug your computer into LAN2 (1000Mbps connection required).
If you use the LAN1/POE port, set your computer to force a 100Mbps link.
Connect to the device via TTL (Serial) 115200n8.
Locate the header (or solder pads) labeled 9X,
near the Ethernet jacks/transformers.
There should be an arrow on the other side of the header marking.
The connections should go like this:
(from the arrow to the 9X text): NC - TXD - GND - RXD - NC
1. Prepare for installation
While the AP is powering up, interrupt the startup process.
MAKE SURE TO CHECK YOUR CURRENT PARTITION!
If you see: "Current Partition is : partB" or
"Need to switch partition from partA to partB",
you have to force the device into partA mode, before continuing.
This can be done by changing the PKRstCnt to 5 and resetting the device.
setenv PKRstCnt 5
saveenv
reset
After you interrupt the startup process again,
you should see: Need to switch partition from partB to partA
You can now continue to the next step.
If you see: "Current Partition is : partA",
you can continue to the next step.
2. Prevent partition switching.
To prevent the device from switching partitions,
we are going to modify the startup command.
set bootcmd "setenv PKRstCnt 0; saveenv; bootipq"
setenv
3. First boot
Now, we have to boot the OpenWrt intifs.
The easiest way to do this is by using Tiny PXE.
You can also use the normal U-Boot tftp method.
Run "bootp" this will get an IP from the DHCP server
and possibly the firmware image.
If it doesn't download the firmware image, run "tftpboot".
Now run "bootm" to run the image.
You might see:
"ERROR: new format image overwritten - must RESET the board to recover"
this means that the image you are trying to load is too big.
Use a smaller image for the initial boot.
4. Install OpenWrt from initfs
Once you are booted into OpenWrt,
transfer the OpenWrt upgrade image and
use sysupgrade to install OpenWrt to the device.
Signed-off-by: Kristjan Krušič <kristjan.krusic@krusic22.com>
These are the factory reset button (external) and "developer mode"
button (hidden inside the case (ASUS) or under a screw in the base
(TP-Link)) found on the TP-Link and ASUS OnHub devices.
Signed-off-by: Alan Luck <luckyhome2008@gmail.com>
[Brian: add description; factor out for both ASUS and TP-Link; use
existing pinmux definitions; add keycode for dev button]
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Chromium devices (like OnHub) have ramoops memory reserved by the
bootloader. Let's enable the ramoops kernel module by default, so we get
better crash logging.
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
With 5.15 kernel version Linksys EAX500 family devices suffered from a
big regression where the Ethernet switch became silent and started to
malfunction.
It was discovered later that the cause was not really the kernel upgrade
itself but a hackish implementation of the hw implementation of these
special routers.
In the original Linksys source code, GPIO 63 was handled in a special way
and was reset on reboot.
Normally GPIO 63 is used for pcie2 reset but in every device we support,
pcie2 is actually never used as nothing is attached to it.
Linksys rerouted GPIO 63 to the switch reset pin and deviates from
common hw implementation.
Till now it was used an hack to handle this case... It was set pcie3 as
working (while actually nothing was connected), set it to output low
(for assert-deassert from the pcie init code) and be done with it.
The result was that the GPIO was reset for enough time in early boot and
everything worked correctly.
This hack implementation was born to fail from the very start and in
kernel 5.15 finally problem arised.
In 5.15 pcie code changed and now the GPIO reset pin is not asserted as
probe won't fail if nothing is connected to the line (the old behaviour)
This result in the switch hold the reset pin and the Ethernet switch
dead.
On top of that with 5.15 code got optimized and simply attaching the
GPIO reset to the mdio wasn't enough as the switch require at least 10ms
to be correctly reset.
So implement finally a correct solution where:
- pcie2 is correctly disabled (nothing attached, unused)
- drop the wrong output-low for pcie2 reset pin
- define GPIO 63 as switch reset
- Add the reset-gpios to the mdio0 node
- Set the reset-post-delay-us to 12ms to correctly give time the switch
to reset
Fixes: #10983
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Starting from Linux Kernel version 6.3 UBI devices will no longer be
considered virtual, but rather have an MTD device parent. Hence they
will no longer be listed under /sys/devices/virtual/ubi which is
used in multiple places in OpenWrt. Prepare for future kernels by
using /sys/class/ubi instead of /sys/devuces/virtual/ubi.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Most of the time when booting kernel prints a warning from
mm/page_alloc.c when pstore/ramoops is being initialized and ramoops is
not functional.
Fix this by moving ramopps node into reserved-memory block as described
in kernel documentation.
Fixes: 2964e5024c ("ipq806x: kernel ramoops storage for C2600/AD7200")
Signed-off-by: Filip Matijević <filip.matijevic.pz@gmail.com>
When fstools is unable to parse our root=<...> arg correctly, it can
fall back to scanning all block devices for a 'rootfs_data' partition.
This fallback was deemed wrong (or at least, a breaking/incompatible
change) for some targets, so we're forced to opt back into it with
fstools_partname_fallback_scan=1.
Without this, OnHub devices will use a rootfs-appended loop device for
rootfs_data instead of the intended 3rd partition.
While I'm at it, just move all the boot args into the 'cros-vboot'
build rule, instead of using the custom bootargs-append. All cros-vboot
subtargets here are using the same rootwait (to support both eMMC and
USB boot) and root/partition args.
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
[ drop unrelated comments in commit description ]
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
TP-Link and ASUS OnHub devices are very similar, sharing many of the
same characteristics and much of their Device Tree. They both run a
version of ChromeOS for their factory firmware, and so installation
instructions look very similar to Google Wifi [1].
Things I've tested, and are working:
* Ethernet
* WiFi (2.4 and 5 GHz)
* LEDs
* USB
* eMMC
* Serial console (if you wire it up yourself)
* 2x CPU
* Speaker
== Installation instructions summary ==
1. Flash *-factory.bin to a USB drive (e.g., with `dd`)
2. Insert USB drive, to boot OpenWrt from USB
3. Copy the same *-factory.bin over to device, and flash it to eMMC to
make OpenWrt permanent
== Developer mode, booting from USB (Step 2) ==
To enter Developer Mode and boot OpenWrt from a USB stick:
1. Unplug power
2. Gain access to the "developer switch" through the bottom of the
device
3. Hold down the "reset switch" (near the USB port / power plug)
4. Plug power back in
5. The LED on the device should turn white, then blink orange, then
red. Release the reset switch.
6. Insert USB drive with OpenWrt factory.bin
7. Press the hidden developer switch under the device to boot to USB;
you should see some activity lights (if you have any) on your USB
drive
8. Depending on your configuration, the router's LED(s) should come on.
You're now running OpenWrt off a USB stick.
These instructions are derived from:
https://www.exploitee.rs/index.php/Rooting_The_Google_OnHub#Enabling_%22Developer_Mode%22_on_the_OnHubhttps://www.exploitee.rs/index.php/Asus_OnHub#Enabling_%22Developer_Mode%22_on_the_OnHub
~~Finding the developer switch:~~ for TP-Link, the developer switch is
on the bottom of the device, underneath some of the rubber padding and a
screw. For ASUS, remove the entire base, via 4 screws under the rubber
feet. See the Exploitee instructions for more info and photos.
== Making OpenWrt permanent (on eMMC) (Step 3) ==
Once you're running OpenWrt via USB:
1. Connect Ethernet to the LAN port; router's LAN address should be at
192.168.1.1
2. Connect another system to the router's LAN, and copy the factory.bin
image over, via SCP and SSH:
scp -O openwrt-ipq806x-chromium-tplink_onhub-squashfs-factory.bin root@192.168.1.1:
ssh root@192.168.1.1 -C "dd if=/dev/zero bs=512 seek=7552991 of=/dev/mmcblk0 count=33 && \
dd if=/root/openwrt-ipq806x-chromium-tplink_onhub-squashfs-factory.bin of=/dev/mmcblk0"
3. Reboot and remove the USB drive.
== Developer mode beep ==
Note that every time you boot the OnHub in developer mode, the device
will play a loud "beep" after a few seconds. This is described in the
Chromium docs [2], and is intended to make it clear that the device is
not running Google software. It is nontrivial to completely disable this
beep, although it's possible to "acknowledge" developer mode (and skip
the beep) by using a USB keyboard to press CTRL+D every time you boot.
[1] https://openwrt.org/toh/google/wifi
[2] https://chromium.googlesource.com/chromiumos/docs/+/HEAD/developer_mode.md
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
For IPQ8064 systems based off the "Google Storm" reference platform,
such as the TP-Link OnHub.
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
This fixes device tree registration for 'qcom,lpass-cpu' as used by
qcom-ipq8064 SoCs, and allows speaker audio to function.
This patch has been submitted (and merged, for -next; likely v6.3)
upstream.
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Refresh target config with `make kernel_menuconfig`, then save the
result. This drops missing symbols or otherwise accounts for defaults.
It should not change any functionality.
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Similar to commit 4d8b42d8a7 ("ipq40xx: point to externally compiled
dtbs in recipes").
Currently, we patch our DTS files into the kernel source tree, so the
kernel build process will produce DTBs for us. The kernel-to-DTS
dependency can cause buildroot to perform excessive rebuilds of the
kernel though, which slows down device development iteration.
Buildroot also compiles DTBs on its own, to
$(KDIR)/image-$(DEVICE_DTS).dtb. With small adjustments, we can leverage
this, and stop patching DTS files into the kernel Makefile at the same
time.
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
