This patch backports verbatim the commit from Linux 5.2-rc7 that fixes
the warnings about invalid lpm related parameters on hardware which
don't that.
This is the case for e.g. lantiq xrx200 targets.
Supported only in Linux 4.17 an later.
Signed-off-by: Martin Schiller <ms@dev.tdt.de>
[refresh patches, fix commit title]
Signed-off-by: Mathias Kresin <dev@kresin.me>
The following warnings are introduced by our own patches:
lantiq_etop.c:173:1: warning: no semicolon at end of struct or union
};
^
lantiq_etop.c: In function 'ltq_etop_change_mtu':
lantiq_etop.c:725:2: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
int max = ETH_HLEN + VLAN_HLEN + new_mtu + ETH_FCS_LEN;
^~~
Signed-off-by: Mathias Kresin <dev@kresin.me>
In DTS Checklist[1] we're now demanding proper generic node names, as
the name of a node should reflect the function of the device and use
generic name for that[2]. Everybody seems to be copy&pasting from DTS
files available in the repository today, so let's unify that naming
there as well and provide proper examples.
1. https://openwrt.org/submitting-patches#dts_checklist
2. https://github.com/devicetree-org/devicetree-specification/blob/master/source/devicetree-basics.rst#generic-names-recommendation
Signed-off-by: Petr Štetiar <ynezz@true.cz>
[split out treewide patch, integrated P2812HNUF1 rename patch]
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
[merged key and led changes, use generic node name for flash, wifi and
gpio]
Signed-off-by: Mathias Kresin <dev@kresin.me>
It is pretty ovbious a misuse of the simple-bus binding as no child has
a address.
Luckly the clock node isn't used at all so it is safe to get rid of it.
Signed-off-by: Mathias Kresin <dev@kresin.me>
Set a unit address matching the reg property for the EASY98000 ethernet
node.
Fixes a unit mismatch devicetree compiler warning.
Signed-off-by: Mathias Kresin <dev@kresin.me>
The localbus childs are highly board specific and need to be part of the
board dts. Otherwise we run into duplicate unit address issues if a
board has something else (NAND) connected to the bus id.
Signed-off-by: Mathias Kresin <dev@kresin.me>
The register address is the same for danube und vr9. The register size
is 0x400 byte for danube and 0x300 byte for vr9.
The change doesn't have an impact as the vmmc driver doesn't use the
devicetree and has hardcoded register offsets.
Signed-off-by: Mathias Kresin <dev@kresin.me>
Fixes the following devicetree compiler warning:
unnecessary #address-cells/#size-cells without "ranges" or child "reg" property
Signed-off-by: Mathias Kresin <dev@kresin.me>
Add size and address cells where missing to fix the following devicetree
compiler warning:
Warning (avoid_default_addr_size): Relying on default #address-cells value
Warning (avoid_default_addr_size): Relying on default #size-cells value
Signed-off-by: Mathias Kresin <dev@kresin.me>
The cpu temperature driver has the register offset hardcode and doesn't
need a reg.
The mdio bus node is only used as a well known name and doesn't need a
reg or unit address.
syscon-reboot doesn't have a unit address or a reg either. The unit name
collides with reset-controller@10 anyway.
Signed-off-by: Mathias Kresin <dev@kresin.me>
Add the reg property if the node has an unit address. Fixes the
following device tree compiler warning:
Warning (unit_address_vs_reg): node has a unit name, but no reg property
Signed-off-by: Mathias Kresin <dev@kresin.me>
This allows the definition of DEVICE_VENDOR, DEVICE_MODEL and
DEVICE_VARIANT. All three are merged together to look the same as the
current DEVICE_TITLE. Also, if DEVICE_TITLE is set it's directly used as
a *fallback* for devices which weren't upgraded yet.
This is based on the work of @sudhanshu16.
The motivation is to create JSON files based on provided metadata,
needing clear differentiation between vendor, model and revision.
Signed-off-by: Paul Spooren <mail@aparcar.org>
The current state of the kernel 4.14 support is in the openwrt-19.07
branch. No need to keep a not default used kernel in this branch.
Signed-off-by: Mathias Kresin <dev@kresin.me>
With kernel 4.19 the DMA API relies on a struct device to work properly.
Backport upstream patches which passed a struct device instead of NULL
to the DMA API.
Fixes kernel panics during boot.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Mathias Kresin <dev@kresin.me>
In case a PCI device is physical present, the Lantiq PCI driver need to
be loaded prior to the Lantiq PCIe driver. Otherwise none of them will
work because updating the BAR 0 of a device attached to the PCIe Bus
will fail.
Till kernel commit 9667bb039bf6 ("MIPS: lantiq: remove unnecessary
of_platform_default_populate call"), the Lantiq PCI driver was probed
early in the boot process via plat_of_setup().
Since the mentioned commit, both drivers are probed via
of_platform_default_populate_init() and they are probed according to
their order in the devicetree source file.
If the Lantiq PCI driver is enabled in the devicetree, defer the probing
og the Lantiq PCIe driver till the Lantiq PCI driver is probed. Use the
presence of a PCU bus as indicater for a loaded Lantiq PCI driver.
This way we don't need to shuffle the PCI/PCIe nodes in the devicetree
source file, as they are already (correctly) ordered based on the unit
address.
Signed-off-by: Mathias Kresin <dev@kresin.me>
With kernel 4.19 dwc2 would not want to initialize due to reset
timeouts, while it worked fine with 4.14.
Increase the reset timeouts to 1 second, as it was used by the old
lantiq ifxhcd usb driver.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Mathias Kresin <dev@kresin.me>
The following patches are dropped because they are now upstreamed:
- 0002-gpio-stp-xway-Implement-get-callback.patch
upstreamed with commit 5b9b2b5284f819 ("gpio: stp-xway: Implement get
callback")
- 0027-01-net-phy-intel-xway-add-VR9-version-number.patch
upstreamed with commit 5b73d9955fb4b0 ("net: phy: intel-xway: add VR9
version number")
- 0027-02-net-phy-intel-xway-add-VR9-v1.1-phy-ids.patch
upstreamed with commit f452518c982e57 ("net: phy: intel-xway: add VR9
v1.1 phy ids")
The following patches were updated:
- 0018-MTD-nand-lots-of-xrx200-fixes.patch
the mainline driver now resides in drivers/mtd/nand/raw/xway_nand.c
(instead of drivers/mtd/nand/xway_nand.c)
- 0025-NET-MIPS-lantiq-adds-xrx200-net.patch
the DMA API now requires a valid device to be passed to all operations
- 0028-NET-lantiq-various-etop-fixes.patch
the DMA API now requires a valid device to be passed to all operations
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Mathias Kresin <dev@kresin.me>
This just copies the files from the kernel 4.14 specific folders into
the kernel 4.19 specific folder, no changes are done to the files in
this commit.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Linux 4.15 removes the init_timer() API. It's replaced by two functions:
- timer_setup() is used instead of init_timer() and also replaces the
timer "function" (callback) setup.
- from_timer() has to be used to obtain the use-case specific data from
a struct timer_list, which is now passed to the timer callback.
Update the timer API to be compatible with Linux 4.15+ so it compiles
with the upcoming Linux 4.19 kernel update.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Since we switched to 4.19, the kernel build checks for libelf to decide if
it should build tools for stack validation.
On macOS, this check fails during target/compile, but succeeds during package
build (because of the pkg-config path picking up target libraries).
Signed-off-by: Felix Fietkau <nbd@nbd.name>
While moving common used parts to dtsi files, the was disabled by
default but not enabled for all boards using the STP.
Fixes: f519fea4c6 ("lantiq: kernel 4.14: cleanup dts files")
Reported-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Mathias Kresin <dev@kresin.me>
Reviewed-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
This addresses an issue where the list option specified in
/etc/config/openvpn i.e. 'tls_cipher' would instead show up in the
generated openvpn-<name>.conf as 'ncp-ciphers'. For context,
'ncp_ciphers' appears after 'tls_cipher' in OPENVPN_LIST from
openvpn.options.
Also, the ordering of the options in the UCI config file is now
preserved when generating the OpenVPN config. The two currently
supported list options deal with cipher preferences.
Signed-off-by: Matt Merhar <mattmerhar@protonmail.com>
All protos for wwan (ncm,qmi,mbim) do have a delay option.
To standardize that add also the missing delay option to the 3g proto.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Restart is in default implemented so it calls stop and start. This is
pretty unsafe to call on umount service. This service should not do
anything on restart the same way as on start. Only use of this service
is on stop.
Signed-off-by: Karel Kočí <cynerd@email.cz>
There are mobile carrier who have different MTU size in their network.
With this change it is now possible to configure this with the qmi
proto handler.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Apply the same approach as in commit 3b53d6fdbc ("ar71xx: fix pci irq
init on kernel 4.14") to fix IRQ initialization for ath79-based chipsets
on rb4xx.
Ref: PR#2182
Acked-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Signed-off-by: W. Michael Petullo <mike@flyn.org>
[commit ref fix]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
In ath79, identifiers tplink_tl-wdr3600 and tplink_tl-wdr4300 have
been used while most other TP-Link devices include the revision.
Although there actually is only one major revision of these
devices, they bear the revision on their bottom (v1.x). TP-Link
also refers to the devices as V1 on its web page.
This patch thus adds -v1 to both so it is more consistent
with other devices and with what you would expect from reading
the on-device sticker and the support pages.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Having the build time written on stderr make it appear with V=w
although it is not an error or warning. Just write the time on stdout
to have it part of the build log like all the rest, but not clutter
the output when only warnings and errors should be shown.
Signed-off-by: Alban Bedel <albeu@free.fr>
Fix wrong paths in InstallDev which cause Lua 5.3 headers to be staged
in /usr/include/, overwriting Lua 5.1 headers and leading to widespread
build failures in all Lua related packages.
Fixes: FS#2348
Fixes: 6b161bb8d5 ("lua5.3: package Lua 5.3 version")
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
* curve25519: not all linkers support bmi2 and adx
This should allow WireGuard to build on older toolchains.
* global: switch to coarse ktime
Our prior use of fast ktime before meant that sometimes, depending on how
broken the motherboard was, we'd wind up calling into the HPET slow path. Here
we move to coarse ktime which is always super speedy. In the process we had to
fix the resolution of the clock, as well as introduce a new interface for it,
landing in 5.3. Older kernels fall back to a fast-enough mechanism based on
jiffies.
https://lore.kernel.org/lkml/tip-e3ff9c3678b4d80e22d2557b68726174578eaf52@git.kernel.org/https://lore.kernel.org/lkml/20190621203249.3909-3-Jason@zx2c4.com/
* netlink: cast struct over cb->args for type safety
This follow recent upstream changes such as:
https://lore.kernel.org/lkml/20190628144022.31376-1-Jason@zx2c4.com/
* peer: use LIST_HEAD macro
Style nit.
* receive: queue dead packets to napi queue instead of empty rx_queue
This mitigates a WARN_ON being triggered by the workqueue code. It was quite
hard to trigger, except sporadically, or reliably with a PC Engines ALIX, an
extremely slow board with an AMD LX800 that Ryan Whelan of Axatrax was kind
enough to mail me.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
The linux kernel is not reproducible because the build user
and domain is included into the kernel. Set the build user
to `builder` and build domain to buildhost.
It's also possible to build reproducible builds by setting
KERNEL_BUILD_USER KERNEL_BUILD_DOMAIN to static values.
Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
TP-Link TL-WR841n v14 is a router based on MediaTek MT7628N.
- MediaTek MT7628NN
- 32 MB of RAM
- 4 MB of FLASH
- 2T2R 2.4 GHz
- 5x 10/100 Mbps Ethernet
Installation:
- copy the
'openwrt-ramips-mt76x8-tl-wr841n-v14-squashfs-tftp-recovery.bin'
file to your tftp server root and rename it to 'tp_recovery.bin'.
- configure your PC running the TFTP server with the static IP address
192.168.0.66/24
- push the reset button and plug in the power connector. Wait until
the orange led starts blinking (~6sec)
Signed-off-by: Alexander Müller <donothingloop@gmail.com>
Signed-off-by: Alexander Couzens <lynxis@fe80.eu> [small modifications gpio-hog]
Similiar to the lantiq target use a dts alias to define the wlan led
instead of static mapping in /etc/board.d/01_leds. Reduce code
duplication.
A device tree must define the alias "led-wlan" similiar to "led-boot".
/ {
aliases {
led-wlan = &led_wlan;
};
[..]
led_wlan: wlan {
label = "tl-wr841n-v14:green:wlan";
gpios = <&gpio1 9 GPIO_ACTIVE_LOW>;
};
};
Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
The src-include method allows recursive inclusion of feeds.conf snippets.
This can for example be used for adding static local feeds to
feeds.conf.default without ever having to update the local feeds.conf:
src-include defaults feeds.conf.default
src-link custom /usr/local/src/lede/custom
Signed-off-by: Bjørn Mork <bjorn@mork.no>
