U-Boot with enabled secure-boot will not boot images with the
@-character used for hash node-names.
Use the existing separation character configurable for each device.
Signed-off-by: David Bauer <mail@david-bauer.net>
All the tools (e.g fw_setenv, ubiupdatevol) and config (fw_env.config)
needed for sysupgrade are already included in /lib/upgrade/stage2
Signed-off-by: Mathew McBride <matt@traverse.com.au>
The PHY driver needs to read a register containing the values of the
bootstrap pins (which happen to be the PHY LEDs) to determine the LED
polarities. Allow regmap access to first pinctrl bank by adding the
'syscon' compatible, and reference the pinctrl in the MDIO bus where
the PHY driver will look for it.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Safely detect integer overflow in try_addint() and try_subint().
Old code relied on undefined behavior, and recent versions of GCC on x86
optimized away the if-statements.
This caused integer overflow in Lua code instead of falling back to
floating-point numbers.
Signed-off-by: Adam Bailey <aebailey@gmail.com>
This option was removed from upstream kernel back in 2022.
See commits:
2d16803c562ecc644803d42ba98a8e0aef9c014e (>=6.0)
3dd33a09f5dc12ccb0902923c4c784eb0f8c7554 (>=5.15.61 backport)
Signed-off-by: Christian Svensson <blue@cmd.nu>
commit c0611b45a9 ("e2fsprogs: symlink e2fsck to fsck.ext{2, 3, 4},
and tune2fs to findfs") introduced a symlink from tune2fs to findfs.
This only works when the included private libblkid library is used, but
commit 5b1660a538 ("utils/e2fsprogs: Update to 1.43.6") disabled the
usage of this private lib and enabled the shared lib support.
Removing this symlink makes it possible to install tune2fs and findfs
package.
Signed-off-by: Martin Schiller <ms@dev.tdt.de>
The return value of the .remove function pointer has changed from
int to void with Linux 5.18. Use a precompiler macro to allow building
the leds-ws2812b module with both, Linux 5.15 and Linux 6.1.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Stable kernel git log:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/log/?h=v6.1.38
No patches needed to be rebased. Just updated checksum.
Build system: x86_64
Build-tested: x86_64/AMD Ryzen 7
Run-tested: x86_64/AMD Ryzen 7
Signed-off-by: John Audia <therealgraysky@proton.me>
[add link to stable kernel git log]
Signed-off-by: Nick Hainke <vincent@systemli.org>
With Linux 6.1 many of our downstream patches and out-of-tree files
can be removed or at least replaced by backported upstream commits.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
[fix CMDLINE_OVERRIDE for arm64]
Signed-off-by: Bjørn Mork <bjorn@mork.no>
The old RealTek RTL8367S switch driver which is used for some MT7622
devices needs to be modified to no longer free the GPIO after reset
has completed.
This is due to Linux 5.19 removing devm_gpio_free via commit
2b038e786f83 ("gpiolib: devres: Get rid of unused devm_gpio_free()")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
When refreshing the hack patches for Linux 6.1 the part of the uImage.FIT
partition parser patch which takes care of allowing mtdblock and ubiblock
devices to have partitions has been dropped, supposedly by accident.
Re-add a that part to the patch, so devices using a uImage.FIT filesystem
sub-image as rootfs can work with Linux 6.1.
Fixes: 19a246bb65 ("generic: 6.1: manually refresh hack patches")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
This adds support for the i.MX Pixel Pipeline IP block
which is available on some imx6 flavours [1]
This allows to use hardware offloading for operations like:
- Colour conversion
- Scaling
- Rotation
[1] https://en.wikipedia.org/wiki/I.MX#i.MX_6_series
Signed-off-by: Koen Vandeputte <koen.vandeputte@citymesh.com>
This adds support for the Video Processing Unit IP block
which is present in certain i.MX SOC's.
The vpu used in imx6 is the coda960 which supports:
- h264 enc
- h264 dec
- jpeg enc
- jpeg dec
Please note that the required firmware needs to be added
by yourself as it's not available currently in linux-firmware upstream.
The firmware package can be found on the internet
and it will decompress itself exposing all binaries
after accepting the EULA.
The binaries should be placed at exactly these paths:
- /lib/firmware/vpu_fw_imx6d.bin
- /lib/firmware/vpu_fw_imx6q.bin
Following output will be printed at boottime if all is well:
[ 9.769638] coda 2040000.vpu: Firmware code revision: 46076
[ 9.775277] coda 2040000.vpu: Initialized CODA960.
[ 9.780082] coda 2040000.vpu: Firmware version: 3.1.1
[ 9.785312] coda 2040000.vpu: coda-jpeg-encoder registered as video0
[ 9.791859] coda 2040000.vpu: coda-jpeg-decoder registered as video1
[ 9.798375] coda 2040000.vpu: coda-video-encoder registered as video2
[ 9.805013] coda 2040000.vpu: coda-video-decoder registered as video3
gstreamer will automatically detect and use all encoders/decoders.
Please note that a FILES catch-all is required for the videobuf-dma objects
as some modules enabling this could require (and thus generating) only 1 of them.
Signed-off-by: Koen Vandeputte <koen.vandeputte@citymesh.com>
This adds support for videobuf2-dma driver.
This module contains following flavors:
- Contiguous
- Scatter/Gather
Drivers using this can enable 1 of the, or both, depending on their needs.
Due to this, a FILES catch-all is required for the videobuf-dma objects
as depending on requirements, only 1 of them could get generated.
Signed-off-by: Koen Vandeputte <koen.vandeputte@citymesh.com>
This allows addition of devices which use these kernel
modules.
This also adds a package for handling dma within video2buf.
These are only build when selected by a caller
Signed-off-by: Koen Vandeputte <koen.vandeputte@citymesh.com>
Since kernel 5.17+ the mips asm.h includes isa-rev.h, which itself was
added 4.17. Without it, lzma-loader will fail to build:
make[3] -C target/linux compile
make[5]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule.
In file included from head.S:22:
.../staging_dir/toolchain-mips_mips32_gcc-12.3.0_musl/include/asm/asm.h:22:10: fatal error: asm/isa-rev.h: No such file or directory
22 | #include <asm/isa-rev.h>
| ^~~~~~~~~~~~~~~
compilation terminated.
make[6]: *** [Makefile:64: head.o] Error 1
make[5]: *** [Makefile:345: compile] Error 2
make[4]: *** [Makefile:24: compile] Error 2
make[3]: *** [Makefile:11: compile] Error 2
ERROR: target/linux failed to build.
So add the file to the files to install. We can do that unconditionally,
since the oldest supported kernel 5.15 already includes it, even it if
does not need it.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Manually rebased:
generic/hack-6.1/220-arm-gc_sections.patch
armsr/patches-6.1/221-armsr-disable_gc_sections_armv7.patch
All other patches automatically rebased.
Signed-off-by: John Audia <therealgraysky@proton.me>
All patches automatically rebased.
Acknowledgment to @john-tho for the changes to fs.mk to accommodate new paths
introduced in 29429a1f58
Build system: x86_64
Build-tested: bcm2711/RPi4B
Run-tested: bcm2711/RPi4B
Signed-off-by: John Audia <therealgraysky@proton.me>
If CONFIG_USE_MOLD is set, all target packages will use the mold linker.
Except the ones which opted-out via setting PKG_BUILD_FLAGS:=no-mold.
Signed-off-by: Andre Heider <a.heider@gmail.com>
mold is a faster drop-in replacement for existing Unix linkers.
A single binary is able to link various targets, which is why this lives
in tools/.
All toolchain builds then just need to copy the linker over, hence avoiding
multiple builds with the same outcome.
Signed-off-by: Andre Heider <a.heider@gmail.com>
Building it requires gcc >= 10.2 or clang >= 12.
Using sstrip with its -z argument can produce non-working binaries, like
a segfaulting `getrandom`, so don't allow that combination.
Signed-off-by: Andre Heider <a.heider@gmail.com>
sstrip only has one functional arg. Make that a bool option, which can
easily depend on other knobs then.
This is required to be disabled for the mold linker.
Signed-off-by: Andre Heider <a.heider@gmail.com>
This patch will only force mac80211 loss detection upon ath10k by
masking the driver-specific loss-detection bit.
Ref: commit ed816f6ba8 ("mac80211: always use mac80211 loss detection")
Signed-off-by: David Bauer <mail@david-bauer.net>
The upstream board-2.bin file in the linux-firmware.git
repository for the QCA4019 contains a packed board-2.bin
for this device for both 2.4G and 5G wifis. This isn't
something that the ath10k driver supports.
Until this feature either gets implemented - which is
very unlikely -, or the upstream boardfile is mended
(both, the original submitter and ath10k-firmware
custodian have been notified). OpenWrt will go back
and use its own bespoke boardfile. This unfortunately
means that 2.4G and on some revisions the 5G WiFi is
not available in the initramfs image for this device.
Fixes: #12886
Reported-by: Christian Heuff <christian@heuff.at>
Debugged-by: Georgios Kourachanis <geo.kourachanis@gmail.com>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
CONFIG_NVME_HWMON exposes /sys/class/nvme/nvme0/device/hwmon
to allow sensors (and others) to see NVMe drive health
Signed-off-by: John Audia <therealgraysky@proton.me>
