This reverts commit 79af0593a3.
A hack adjusting fw_devlink value was added to workaround issue with
probing device drivers caused by of_platform_populate(). With upstream
mtd commit (the one adding OF_POPULATED) backported there is no need for
that hack anymore.
Ref: 3eebb91317 ("kernel: backport proper fix for mtd preventing devices probing")
Ref: #10232
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
DK01 and DK04 board support has been in a form of 2 patches that we have
been carrying for a long time.
Both of the patches contain weird changes, dont follow any DT syntax and I
honestly doubt they are even valid.
DK01 and DK04 also have not been converted to DSA even after a long time
and I doubt that anybody in the community even has these boards as they are
QCA reference boards that are not even obtainable anymore.
Since patches for these 2 boards have been just causing us pain when trying
to update the kernel to a new major release or even point releases lets
remove the support for these boards, and if there are users they can easily
be reinstated.
Signed-off-by: Robert Marko <robimarko@gmail.com>
Teltonika RUTX currently is the only device pulling in DK01 DTSI and thus
preventing removal of DK01 and DK04 support.
So, lets add the missing nodes from DK01 DTSI and use the SoC DTSI instead.
Signed-off-by: Robert Marko <robimarko@gmail.com>
Lets add a proper commit title and description to the SCM cold boot
patch so it applies with a git apply or git-am.
Signed-off-by: Robert Marko <robimarko@gmail.com>
SCM SDI disable support is pending upstream, so lets use that instead.
Since the board check needs to be split out, export it with a header so
it applies with git-am.
Signed-off-by: Robert Marko <robimarko@gmail.com>
Update to the latest stable version.
This update changes the default lockfile directory from /var/lock to
/run [1]. In OpenWRT we still use the "legacy" /var/lock and /run might
not even exist, so we add a patch to revert this particular change.
[1] aeb40f1166
Signed-off-by: Stefan Kalscheuer <stefan@stklcode.de>
This change makes it possible to use the GPIO_OPEN_DRAIN /
GPIO_OPEN_SOURCE Flags when exporting GPIOs via dts.
We need to emulate the open-source or open-drain functionalities for the
initial value, because the used functions (gpiod_direction_output_raw)
do not take this into account.
Signed-off-by: Martin Schiller <ms@dev.tdt.de>
The kernel FSL_ENETC_QOS option is only a compile time
option, it does not result in a separate module being built.
Set it to 'y' to resolve a warning from the kernel compile:
.config:2654:warning: symbol value 'm' invalid for FSL_ENETC_QOS
Signed-off-by: Mathew McBride <matt@traverse.com.au>
Fixes: c3151b6f04 ("armvirt: 64: add support for other SystemReady-compatible vendors")
When comparing the generated OpenWrt .config to the Linux arm64
defconfig, I noticed these SATA controllers were not included.
As they may be used as a boot drive, they should be built into
the kernel.
CONFIG_SATA_MVEBU is for Marvell platforms.
CONFIG_SATA_QORIQ is for NXP Layerscape.
CONFIG_SATA_SIL24 is for Arm's Juno development board, see Linux
kernel commit d7c38ff1cd86 ("arm64: defconfig: Add Juno SATA
controller").
Signed-off-by: Mathew McBride <matt@traverse.com.au>
This MDIO driver was already being built, but not installed due
to being selected by the ThunderX Ethernet driver.
Signed-off-by: Mathew McBride <matt@traverse.com.au>
The initial armv8 module incorrectly labelled the Thunder(v1) as
supporting the ThunderX2, when they have different drivers.
Add kmod-octeon-tx2 to support the newer devices.
Signed-off-by: Mathew McBride <matt@traverse.com.au>
This turns on various PCI related options which are enabled
in the Linux kernel arch/arm64/configs/defconfig but not
yet in the OpenWrt config.
Signed-off-by: Mathew McBride <matt@traverse.com.au>
These are used by common Broadcom SoC's like
the BCM2711 (RPi4) and iProc network processor.
Tested on the RPi4B using the Raspberry Pi
UEFI+ACPI firmware[1].
Signed-off-by: Mathew McBride <matt@traverse.com.au>
[1] - https://github.com/pftf/RPi4
This is part of an effort to reduce differences between
the OpenWrt armsr/armv8 config and Linux arm64 defconfig.
This enables CONFIG_ARCH_BCM and downstream
CONFIG_ARCH_BCM2835 (= BCM2711 like Raspberry Pi 4)
and CONFIG_ARCH_BCM_IPROC (Broadcom iProc packet processors).
The broadband specific SoC's (ARCH_BCMBCA) are left out
as it is assumed these will not be doing EFI boot.
Signed-off-by: Mathew McBride <matt@traverse.com.au>
Renesas markets several embedded Arm64 SoCs in the
RZ series (RZ/G, RZ/V), so should be enabled in
a general purpose target.
Automotive (R-Car) SoC's are not enabled by this change.
Signed-off-by: Mathew McBride <matt@traverse.com.au>
Due to an error on my part, Anton Antonov's
i.MX changes[1] did not fully make it into my
armvirt kernel 6.1 EFI pull request. I have updated
them using the options he supplied[1] as well
as comparing to the Linux arm64 defconfig.
The notable exception is:
CONFIG_USB_DWC3_OF_SIMPLE currently disabled
due to an issue with i.MX8P and i.MX8Q.
Fixes: 3efb3b8 ("armvirt: 64: Add NXP i.MX 8M Mini/Nano/Quad/Plus EVK support")
Signed-off-by: Mathew McBride <matt@traverse.com.au>
[1] - ccf826c344
A review of the generated OpenWrt kernel .config
vs the Linux arm64 defconfig showed that this
option was not being enabled, as it is disabled
in OpenWrt's generic config.
ACPI_BUTTON is needed to report and respond to
power button events, so it should be enabled.
Signed-off-by: Mathew McBride <matt@traverse.com.au>
To bring the armsr/armv8 kernel configuration closer to the Linux
arm64 defconfig, synchronize options related to CPU features
(especially more recent Armv8.X variants), scheduler, EFI vars,
CMA and scheduler options.
Signed-off-by: Mathew McBride <matt@traverse.com.au>
CONFIG_ARM_PMU (Arm Performance Monitor Unit) is a requirement
to use KVM (virtualization) from Linux 5.11+, as the virtualised
guest has virtualized PMU access.
Signed-off-by: Mathew McBride <matt@traverse.com.au>
x86/64 enables support for KVM so I can't see a reason why
not on armsr/armv8 as well.
Arm CPU errata workaround items related to virtualization
are also enabled by this change.
Signed-off-by: Mathew McBride <matt@traverse.com.au>
To reduce differences with the Linux arm64 defconfig,
sync the enabled erratum items with defconfig.
There are still some options not selected due to
CONFIG_KVM or other options not enabled in OpenWrt
by default.
Signed-off-by: Mathew McBride <matt@traverse.com.au>
The ide-drive option was renamed to ide-hd in qemu 6.0.
With this change qemu is starting again on Debian 12.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Changes from 2.0.0 to 2.0.1:
----------------------------
* The behavior of --modversion was largely reverted back to the traditional
pkg-config behavior, but still operates on a solved dependency graph.
The order of --modversion output is based on the dependency resolution
queue which is passed to the solver, which itself generally maps to the
order of the constrants provided on the command line.
* A new flag, --verbose, has been added. When used with `--modversion`, it
is possible to disambiguate which version belongs to which module:
% pkgconf --modversion --verbose foo bar
foo: 1.2.3
bar: 1.3
Changes from 1.9.5 to 2.0.0:
----------------------------
* When flattening the dependency graph, retain the latest seen edges
rather than the earliest.
* Fix a long-standing bug where the dependency resolution queue was
evaluated in reverse. This bug masked the aforementioned dependency
flattening bug in many cases.
* Fix handling of --with-path, which was appending paths to the search
list rather than prepending them as intended.
* Error when --modversion is requested with more than one package, as
the output is ambiguous.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
The current minimum OS requirement for OpenWrt is Ubuntu 18.04, which
includes 7 and 8. 8 is necessary for ccache.
gcc and g+++ are now symlinked to staging_dir, similar to Python.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
v2.0.0:
- transition from AGPL to MIT
- Previously, mold could not produce an object file with more than 65520
sections using the --relocatable option. Now the bug has been fixed.
- mold now interprets -undefined as a synonym for --undefined instead of
-u ndefined. This seems inconsistent, as -ufoo is generally treated as
-u foo (which is an alias for --undefined foo), but this is the behavior
of the GNU linkers and LLVM lld, so we prioritize compatibility over
consistency.
- -nopie is now handled as a synonym for --no-pie.
- [RISC-V] R_RISCV_SET_ULEB128 and R_RISCV_SUB_ULEB128 relocation types are
now supported (4bffe26, 1ac5fe7)
- [PPC64] R_PPC64_REL32 relocation type is now supported. (ebd780e)
v2.1.0:
- Loongson's LoongArch CPU has been supported. (03b1a1c)
- -z nosectionheader has been added to eliminate section headers from the
output file. (084ca55)
- Previously, linking with the -z pack-relative-relocs option produces an
executable that glibc 2.38 refuses to run with DT_RELR without
GLIBC_ABI_DT_RELR dependency error. Now, mold produces binaries compatible
with glibc 2.38. (f467ad1)
- [ARM64] R_AARCH64_ADR_PREL_PG_HI21_NC relocation type has been supported.
(17a5c3e)
- [ARM64] R_AARCH64_MOVW_UABS_G3 relocation type has now been handled as a
PLT-generating relocation to fix an issue when main is not defined in the
main executable but rather in a .so file. (e764557)
- [RISC-V] We now merge input .riscv.attributes contents. Previously, we
just concatenated them. (aa64491)
Signed-off-by: Andre Heider <a.heider@gmail.com>
This compiles the CONFIG_PTP_1588_CLOCK support into the kernel binary
and activates the drivers for KVM and VMware which allow syncing the
host time with the VM when OpenWrt is running in a VM. With this change
the CONFIG_HYPERV_UTILS driver is now build into the kernel, because it
depends on the PTP framework being compiled in. CONFIG_HYPERV_UTILS was
build as a module, but not packages before.
Fixes: #13277
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
This activates PCI Express ASPM control in Linux. Without this option it
is completely controlled by the BIOS, now Linux will take over and apply
some workarounds if needed.
Fixes: #13248
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
It seems that the Meraki bootloader does not respect the kernel ARM booting
specification[1] that requires that address where DTB is located needs to
be 64-bit aligned and often places the DTB on a non 64-bit aligned address
and then kernel fails to find the DTB magic and fails to boot.
Even worse, there is no prints until early printk is enabled and then its
visible that kernel is trying to find the ATAG-s as DTB was not found or
is invalid.
Unifi 6 devices had the same issue and it can be solved by passing the
load adress as part of the FIT image.
It seems that the vendor was aware of the issue and is always relocating
the DTB to 0x89000000, so lets just do the same.
Now that booting is reliable, reenable default images for the Meraki MR33
and MR74 devices.
Reviewed-by: Lech Perczak lech.perczak@gmail.com
Signed-off-by: Robert Marko <robimarko@gmail.com>
ipq40xx was converted to DSA and swconfig is not being included at all in
the default packages so there is no need to drop it from device packages.
Signed-off-by: Robert Marko <robimarko@gmail.com>
MR33 and MR74 share pretty much everything in the image recipe, so lets
extract a common recipe to avoid duplication.
Signed-off-by: Robert Marko <robimarko@gmail.com>
This updates the Intel iwlwifi firmware for AX200 and AX210 from version
66 to version 72. Version 72 is the latest version supported by iwlwifi
from kernel 6.1.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Reviewed-by: Robert Marko <robimarko@gmail.com>
* liblzma and xzdec can now build against WASI SDK when threading
support is disabled. xz and tests don't build yet.
* CMake:
- Fixed a bug preventing other projects from including liblzma
multiple times using find_package().
- Don't create broken symlinks in Cygwin and MSYS2 unless
supported by the environment. This prevented building for the
default MSYS2 environment. The problem was introduced in
xz 5.4.0.
* Documentation:
- Small improvements to man pages.
- Small improvements and typo fixes for liblzma API
documentation.
* Tests:
- Added a new section to INSTALL to describe basic test usage
and address recent questions about building the tests when
cross compiling.
- Small fixes and improvements to the tests.
* Translations:
- Fixed a mistake that caused one of the error messages to not
be translated. This only affected versions 5.4.2 and 5.4.3.
- Updated the Chinese (simplified), Croatian, Esperanto, German,
Korean, Polish, Romanian, Spanish, Swedish, Ukrainian, and
Vietnamese translations.
- Updated the German, Korean, Romanian, and Ukrainian man page
translations.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Rename two patches which were only accepted in Linux 6.2, but were
marked as if they were accepted in Linux 6.1.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
the Marvell 10G PHY driver is no way specific to ARM SystemReady
systems, it frequently occurs on SFP+ copper modules and is useful on
many targets.
Hence it been added to package/kernel/linux/modules/netdevices and we
can remove the now redundant target-specific module.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Instead of reading only a single 4kiB page, read the first 128kiB to
determine the size of an uImage.FIT using 'imsz' or 'imszb'.
This will be needed once we add more Device Tree Overlays, which may
happen for the BPi-R3 mini.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
A dependency of the MT7988 MMC host controller on the SoC's RTC clock
being running has been discovered. Mark RTC clock as critical to fix
MMC host on MT7988.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
The Bananapro board has an Ampak 6181 onboard (BCM43362/1), enable
the firmware files in the device profile, and add wpad-basic-mbedtls.
Signed-off-by: Zoltan HERPAI <wigyori@uid0.hu>
