The etop driver is used by all targets except xrx200.
Remove the UBIFS compression support for zlib and zstd from the xway
subtarget. The hardware is EOL for a long time and it's unlikely to ever
see a board shipped with an UBIFS using these compression methods.
Remove the JFFS2 support. Support for jjfs2 images was dropped years
ago.
It shrinks the compressed kernel up to 130 KByte.
Signed-off-by: Mathias Kresin <dev@kresin.me>
Switch to the mainline Lantiq PCIe PHY driver and update the vr9.dtsi
accordingly.
The Lantiq IRQ SMP support added upstream required changes to the SoC
dtsi as well.
Following changes are made to the Lantiq kernel patches:
0005-lantiq_etop-pass-struct-device-to-DMA-API-functions.patch
0006-MIPS-lantiq-pass-struct-device-to-DMA-API-functions.patch
applied upstream
0008-MIPS-lantiq-backport-old-timer-code.patch
access_ok API update because it lost it's type (which was the first)
parameter in upstream commit 96d4f267e40f95 ("Remove 'type' argument
from access_ok() function")
0024-MIPS-lantiq-autoselect-soc-rev-matching-fw.patch
merged into 0026-MIPS-lantiq-Add-GPHY-Firmware-loader.patch
0024-MIPS-lantiq-revert-DSA-switch-driver-PMU-clock-chang.patch
revert upstream changes required for upstream xrx200 ethernet and
xrx200 (DSA) switch driver but breaking our driver
0026-MIPS-lantiq-Add-GPHY-Firmware-loader.patch
required for our driver but dropped upstream, add former upstream
version
0028-NET-lantiq-various-etop-fixes.patch
now has to use the phy_set_max_speed API instead of modifying
phydev->supported. Also call ltq_dma_enable_irq() in
ltq_etop_open() based on upstream commit cc973aecf0b054 ("MIPS:
lantiq: Do not enable IRQs in dma open")
Signed-off-by: Mathias Kresin <dev@kresin.me>
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
This is referenced in two .dts but never actually used from userspace.
Drop support for this property because it's not used and because it
makes updating to Linux 5.4 harder (as the patch doesn't apply anymore).
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Duplicate kernel 4.19 config and patches for kernel 5.4.
Duplicate the devicetree source files as well, they need kernel 5.4
specific adjustments.
Signed-off-by: Mathias Kresin <dev@kresin.me>
Compiling the ltq-tapi driver against Linux 5.4 results in a compile
error complaining that the size of struct sched_param is not known.
Switch the existing "sched/types.h" include to reference
include/uapi/linux/sched/types.h to fix compilation against Linux 5.4.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Upstream commit 84ede58dfcd1d ("crypto: hash - remove
CRYPTO_ALG_TYPE_DIGEST") drops the CRYPTO_ALG_TYPE_DIGEST define because
it has the same value as CRYPTO_ALG_TYPE_HASH. This was the case for
earlier kernels as well. Switch to CRYPTO_ALG_TYPE_HASH to fix building
against Linux 5.4.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Upstream commit e4b92b108c6cd6 ("timekeeping: remove obsolete time
accessors") removed do_gettimeofday(). In Linux 4.19 this was only a
wrapper around ktime_get_real_ts64(). Use ktime_get_real_ts64() now that
the wrapper is gone to fix compilation against Linux 5.4.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Upstream commit e4b92b108c6cd6 ("timekeeping: remove obsolete time
accessors") removed do_gettimeofday(). In Linux 4.19 this was only a
wrapper around ktime_get_real_ts64(). Use ktime_get_real_ts64() now that
the wrapper is gone to fix compilation against Linux 5.4.
Move the ifxmips_mei_interface header to the include directory, it can't
be found otherwise during compilation. The reason for the changed
behaviour is not yet clear, however having header files in an include
directory is more straight forward.
To use the of_device_id struct, the mod_devicetable header need to be
included. Instead of including this header, include the of_platform
header, which includes the mod_devicetable on its own.
Signed-off-by: Mathias Kresin <dev@kresin.me>
Upstream commit 96d4f267e40f95 ("Remove 'type' argument from access_ok()
function") removes the first argument to access_ok.
Adjust the code so it builds with Linux 5.4.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Mathias Kresin <dev@kresin.me>
Beginning with linux 5.3, kmod-serial-8250 uses functions provided by
serial_mctrl_gpio.ko if GPIO support is enabled.
Signed-off-by: Mathias Kresin <dev@kresin.me>
It adjusts the code to upstream changes from the commit 3c1bcc8614db
("net: ethernet: Convert phydev advertize and supported from u32 to link
mode")
Signed-off-by: Mathias Kresin <dev@kresin.me>
For some reason pin 7 (gphy0_led1_pins) need to be set to output prior
to loading the PCI driver. Otherwise the wireless doesn't appear on the
PCI bus. Of course, it doesn't make much sense, since pin 7 is used to
drive the LAN1 led.
It can either be done by setting the pins function to GPHY or GPIO +
direction output. However, the pinctrl driver doesn't provide a way to
switch a pin to GPIO. It is done indirectly by the pinctrl driver at the
time a GPIO is requested (requesting a GPIO always resets the function
to GPIO).
Do it via pinmux driver, as it is always loaded first. Use the GPHY
function as it's the pins intended purpose for this board.
Fixes: FS#2895
Signed-off-by: Mathias Kresin <dev@kresin.me>
There is no such role as target maintainer anymore, one should always
send corresponding changes for the review and anyone from the commiters
is allowed to merge them or eventually use the hand break and NACK them.
Lets make it clear, that it is solely a community doing the maintenance
tasks.
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Acked-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Acked-by: Piotr Dymacz <pepe2k@gmail.com>
In 5.4 kernel old u32 array way of setting network features was dropped and linkmode is now the only way.
So lets migrate the EDMA driver to support linkmode.
Also, old get/set settings API for ethtool is also dropped so lets convert to new ksettings API while at it as it demands linkmode.
Now, gigabit works properly as well as ethtool.
Previously you would get this in ethtool:
root@OpenWrt:/# ethtool eth1
Settings for eth1:
Supports Wake-on: d
Wake-on: d
Current message level: 0x00000000 (0)
Link detected: yes
Now, features are properly advertised:
root@OpenWrt:/# ethtool eth1
Settings for eth1:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supported pause frame use: Symmetric Receive-only
Supports auto-negotiation: Yes
Supported FEC modes: Not reported
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised pause frame use: Symmetric Receive-only
Advertised auto-negotiation: Yes
Advertised FEC modes: Not reported
Link partner advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
Link partner advertised pause frame use: No
Link partner advertised auto-negotiation: No
Link partner advertised FEC modes: Not reported
Speed: 1000Mb/s
Duplex: Full
Port: Twisted Pair
PHYAD: 4
Transceiver: internal
Auto-negotiation: on
MDI-X: Unknown
Supports Wake-on: d
Wake-on: d
Current message level: 0x00000000 (0)
Link detected: yes
Signed-off-by: Robert Marko <robert.marko@sartura.hr>
In 5.4 kernel old u32 array way of setting network features was dropped and linkmode is now the only way.
So lets migrate the PHY driver to support linkmode.
Also, now in order for gigabit to work, PHY driver needs to advertise PHY_GBIT_FEATURES instead of PHY_BASIC_FEATURES
Signed-off-by: Robert Marko <robert.marko@sartura.hr>
of_get_mac_address returns valid pointer or ERR_PTR since 5.2 via commit
d01f449c008a ("of_net: add NVMEM support to of_get_mac_address") so the
patch fixes following OOPs on nbg6617:
Unable to handle kernel paging request at virtual address ffffffed
CPU: 1 PID: 1 Comm: swapper/0 Not tainted 5.4.24 #0
PC is at edma_axi_probe+0x444/0x1114
LR is at bus_find_device+0x88/0x9c
Where the PC can be resolved to:
>>> l *edma_axi_probe+0x444
0xc067be5c is in edma_axi_probe (./include/linux/string.h:378).
>>> l *edma_axi_probe+0x43f
0xc067be57 is in edma_axi_probe (drivers/net/ethernet/qualcomm/essedma/edma_axi.c:936)
Which leads to the following code fragment:
935 mac_addr = of_get_mac_address(pnp);
936 if (mac_addr)
937 memcpy(edma_netdev[idx_mac]->dev_addr, mac_addr, ETH_ALEN);
Where using mac_addr=0xffffffed (-ENODEV) as source address in memcpy()
is causing the OOPs.
Acked-by: John Crispin <john@phrozen.org>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
These new options are common for all subtargets and can be moved into target kconfig.
Signed-off-by: Aleksander Jan Bajkowski <A.Bajkowski@stud.elka.pw.edu.pl>
This was done by executing these command for all subtargets:
$ time make -j9 kernel_oldconfig CONFIG_TARGET=subtarget
Signed-off-by: Aleksander Jan Bajkowski <A.Bajkowski@stud.elka.pw.edu.pl>
These patches are already in mainline 5.4 kernel:
* 010-v5.3-drivers-ata-ahci_sunxi-Increased-SATA-AHCI-DMA-TX-RX.patch
* 101-arm64-dts-allwinner-a64-Enable-A64-timer-workaround.patch
Signed-off-by: Aleksander Jan Bajkowski <A.Bajkowski@stud.elka.pw.edu.pl>
Seems to be working fine on apu2 and QEMU, it's 99.9% upstream stuff
anyway. Lets start wider userbase testing.
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Commit 432ec292cc ("rpcd: add respawn param") has introduced infinite
restarting of the service which could be reached over network. This is
not recommended security practice as it might give potential adversary
infinite number of tries in case there might be some issue in the rpcd
or its surrounding stack.
So lets remove the currently bogus `respawn_retry` variable (it wasn't
possible to override it anyway), reverting to the previous default max.
of 5 service restarts which could be now overriden via system's UCI
settings if desired.
Cc: Jo-Philip Wich <jow@mein.io>
Cc: Florian Eckert <fe@dev.tdt.de>
Cc: Hauke Mehrtens <hauke@hauke-m.de>
Fixes: 432ec292cc ("rpcd: add respawn param")
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Been running it for a few days on RPi0W PMS sensor node and on RPi4.
Lets start wider userbase testing.
Cc: Stijn Tintel <stijn@linux-ipv6.be>
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Signed-off-by: Petr Štetiar <ynezz@true.cz>
WG2600HP has its WLAN MAC addresses at PRODUCTDATA 0xc and 0x12, so
use them both directly.
Signed-off-by: Yanase Yuki <dev@zpc.sakura.ne.jp>
[commit title/message clarification]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
The default case will catch anything left, there is no need to
explicitly add any device to it.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Solve missing references to nftnl_set_list_lookup_byname when building
iptables with Nftables support enabled (CONFIG_IPTABLES_NFTABLES)
Bump the ABI version to force everything to match.
/Users/kevin/wrt/staging_dir/toolchain-x86_64_gcc-9.2.0_musl/lib/gcc/x86_64-openwrt-linux-musl/9.2.0/../../../../x86_64-openwrt-linux-musl/bin/ld: xtables_nft_multi-nft-bridge.o: in function `nft_bridge_parse_lookup':
nft-bridge.c:(.text.nft_bridge_parse_lookup+0xcd): undefined reference to `nftnl_set_list_lookup_byname'
/Users/kevin/wrt/staging_dir/toolchain-x86_64_gcc-9.2.0_musl/lib/gcc/x86_64-openwrt-linux-musl/9.2.0/../../../../x86_64-openwrt-linux-musl/bin/ld: xtables_nft_multi-nft-cache.o: in function `nftnl_set_list_cb':
nft-cache.c:(.text.nftnl_set_list_cb+0x80): undefined reference to `nftnl_set_list_lookup_byname'
/Users/kevin/wrt/staging_dir/toolchain-x86_64_gcc-9.2.0_musl/lib/gcc/x86_64-openwrt-linux-musl/9.2.0/../../../../x86_64-openwrt-linux-musl/bin/ld: xtables_nft_multi-nft-cache.o: in function `fetch_set_cache':
nft-cache.c:(.text.fetch_set_cache+0x10a): undefined reference to `nftnl_set_list_lookup_byname'
collect2: error: ld returned 1 exit status
make[6]: *** [xtables-nft-multi] Error 1
make[5]: *** [all] Error 2
make[4]: *** [all-recursive] Error 1
make[3]: *** [all] Error 2
make[2]: *** [/Users/kevin/wrt/build_dir/target-x86_64_musl/linux-x86_64/iptables-1.8.4/.built] Error 2
make[2]: Leaving directory `/Users/kevin/wrt/package/network/utils/iptables'
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
Currently kmod-i2c-mux-* will not get into images unless kmod-i2c-mux is added
to DEVICE_PACKAGES as well. By changing the dependencies from "depends on" to
"select", we do not have the issue anymore.
Furthermore, we can remove most occurrences of the package from DEVICE_PACKAGES
and similar variables, as it is now pulled by dependent modules such as:
- kmod-i2c-mux-pca954x
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
As it seeems these patches are for OpenWrt and they are not going to be merged
to mainline, let's merge them to improve maintainability.
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
Follow the guideline suggested in e417ff88f1 ("ramips: harmonize line breaks
in image Makefiles").
While at it, fix typos in the device title of Akitio MyCloud.
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
The mux need to be defined in a subnode to be considered by the pinctrl
framework. These muxes aren't set as expected and might cause not
working subsystems.
Fixes: 8e7b573b7a ("lantiq: dts: assign the PCI pins to the PCI controller node")
Fixes: dcb5e52209 ("lantiq: dts: assign the STP pins to the STP GPIO controller node")
Fixes: 660200e53d ("lantiq: dts: assign the GPHY LED pins to the Ethernet controller node")
Signed-off-by: Mathias Kresin <dev@kresin.me>
This reverts commit a6da3f9ef7.
The libcap isn't as optional as the commit messages suggests. A hard
dependency to the libcap package is added, which is only available in
the external packages feed. Therefore it is impossible to package
ip-full without having the external packages feed up and running, which
is a regression to the former behaviour.
Signed-off-by: Mathias Kresin <dev@kresin.me>
Acked-by: Hans Dedecker <dedeckeh@gmail.com>
The pkgconf fork filters -I and -L flag values from .pc files which match
pkgconf's builtin system directory value.
During configure, pkgconf derives the default system include and library
search path values from exec_prefix, which is set to staging_dir/host in
the host tool build phase.
Due to that, pkgconf will drop all -I and -L flags pointing to
staging_dir/host/include or staging_dir/host/lib, unless invoked with
--keep-system-cflags and --keep-system-libs respectively, breaking our
kernel libelf discovery / stack validation workarounds.
In order to inhibit the filtering, add --keep-system-cflags and
--keep-system-libs to our pkg-config shell wrapper.
Fixes: GH#2832
Fixes: 867298cf47 ("tools/pkg-config: Replace with pkgconf")
Ref: https://lists.infradead.org/pipermail/openwrt-devel/2020-March/022182.html
Ref: https://git.openwrt.org/fe43969336201f2cc7d103b68fd6e65989bee184
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Acked-by: Petr Štetiar <ynezz@true.cz>
6594c6b ubus: use dhcpv6 ia assignment flag
a90cc2e dhcpv6-ia: avoid setting lifetime to infinite for static assignments
bb07fa4 dhcpv4: avoid setting lifetime to infinite for static assignments
Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
