The sysupgrade formware of the Puzzle series is a slightly strange
dual-boot approach while remaining compatible with Marvell's SDK
firmware upgrade binary format -- which happens to be a full-disk
image with GPT partition table. Hence that /lib/upgrade/emmc-puzzle.sh
script is like an exotic disease which results from those decisions,
and as we also want to somehow stay compatible with the IEI-World
stock firmware we got to use it in that same way (we are not
compatible with the QNAP-branded identical hardware device anyway).
Currently, on sysupgrade the result is that one ends up with the old
content of rootfs_data (a GPT partition on those devices) as nothing
ever wipes or in any way re-creates the filesystem there. As a simple
work-around, let's kill the filesystem on rootfs_data so fstools
re-formats it on the next boot.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Trying to link certain kernel modules like dahdi-linux when building with
the OpenWrt SDK will fail with:
openwrt-sdk-apm821xx-sata_gcc-13.2.0_musl.Linux-x86_64/staging_dir/toolchain-powerpc_464fp_gcc-13.2.0_musl/bin/powerpc-openwrt-linux-musl-ld: cannot find arch/powerpc/lib/crtsavres.o: No such file or directory
Previously this worked with the PowerPC SDK since we carried a hack that
was passing --save-restore-funcs to module LDFLAGS so the linker provided
the required functions automatically as without --save-restore-funcs it
doesnt do so automatically on relocatable links and as a sideffect did not
require the kernel provided crtsaves.o to link against.
Now that hack has been removed as upstream kernel now compiles crtsaves.o
by default so it can be linked against but its not included in the SDK.
So, lets include lib/crtsavres.o when SDK is generated for PowerPC.
Fixes: 99c9d8abd6 ("kernel: bump 5.15 to 5.15.148")
Signed-off-by: Robert Marko <robimarko@gmail.com>
Currently the compile phase of the kernel builds `Image dtbs modules`.
However, none of the dtbs that get built are used for the final image.
This ends up unnecessarily taking CPU cycles and produces a lot of
`WARNINGS` that can lead users to believe there's cause for concern. I
believe the same principle can be applied to other targets.
```
DTC arch/arm64/boot/dts/qcom/msm8996-mtp.dtb
arch/arm64/boot/dts/qcom/msm8996.dtsi:2954.36-2962.5: Warning (clocks_property):
/soc/clock-controller@6400000: Missing property '#clock-cells' in node
/soc/mailbox@9820000 or bad phandle (referred from clocks[2])
DTC arch/arm64/boot/dts/qcom/msm8996-sony-xperia-tone-dora.dtb
arch/arm64/boot/dts/qcom/msm8996.dtsi:2954.36-2962.5: Warning (clocks_property):
/soc/clock-controller@6400000: Missing property '#clock-cells' in node
/soc/mailbox@9820000 or bad phandle (referred from clocks[2])
DTC arch/arm64/boot/dts/qcom/msm8996-sony-xperia-tone-kagura.dtb
arch/arm64/boot/dts/qcom/msm8996.dtsi:2954.36-2962.5: Warning (clocks_property):
/soc/clock-controller@6400000: Missing property '#clock-cells' in node
/soc/mailbox@9820000 or bad phandle (referred from clocks[2])
DTC arch/arm64/boot/dts/qcom/msm8996-sony-xperia-tone-keyaki.dtb
arch/arm64/boot/dts/qcom/msm8996.dtsi:2954.36-2962.5: Warning (clocks_property):
/soc/clock-controller@6400000: Missing property '#clock-cells' in node
/soc/mailbox@9820000 or bad phandle (referred from clocks[2])
DTC arch/arm64/boot/dts/qcom/msm8996-xiaomi-gemini.dtb
arch/arm64/boot/dts/qcom/msm8996.dtsi:2954.36-2962.5: Warning (clocks_property):
/soc/clock-controller@6400000: Missing property '#clock-cells' in node
/soc/mailbox@9820000 or bad phandle (referred from clocks[2])
DTC arch/arm64/boot/dts/qcom/msm8996-xiaomi-natrium.dtb
arch/arm64/boot/dts/qcom/msm8996.dtsi:2954.36-2962.5: Warning (clocks_property):
/soc/clock-controller@6400000: Missing property '#clock-cells' in node
/soc/mailbox@9820000 or bad phandle (referred from clocks[2])
DTC arch/arm64/boot/dts/qcom/msm8996-xiaomi-scorpio.dtb
arch/arm64/boot/dts/qcom/msm8996.dtsi:2954.36-2962.5: Warning (clocks_property):
/soc/clock-controller@6400000: Missing property '#clock-cells' in node
/soc/mailbox@9820000 or bad phandle (referred from clocks[2])
```
Signed-off-by: Sean Khan <datapronix@protonmail.com>
Historically it's possible to leave the `SUBTARGETS` undefined and
automatically fallback to a "generic" subtarget. This however breaks
various downstream scripts which may have expectations around filenames:
While some targets with an explicit generic subtarget contain `generic`
in the filenames of artifacts, implicit "subtargets" don't.
Right now this breaks the CI[1], possibly also scripts using the ImageBuilders.
This commit removes all code that support implicit handling of
subtargets and instead requires every target to define "SUBTARGETS".
[1]: https://github.com/openwrt/openwrt/actions/runs/8592821105/job/23548273630
Signed-off-by: Paul Spooren <mail@aparcar.org>
Historically it's possible to leave the `SUBTARGETS` undefined and
automatically fallback to a "generic" subtarget. This however breaks
various downstream scripts which may have expectations around filenames:
While some targets with an explicit generic subtarget contain `generic`
in the filenames of artifacts, implicit "subtargets" don't.
Right now this breaks the CI[1], possibly also scripts using the ImageBuilders.
Do to the D1 target what's done to other target, explicitly define the
"generic" subtarget.
[1]: https://github.com/openwrt/openwrt/actions/runs/8592821105/job/23548273630
Signed-off-by: Paul Spooren <mail@aparcar.org>
This device is very similar to the GS1900-24E switch (added in b515ad1),
except that the first 12 of 24 ethernet ports are capable of PoE and the
physical jacks are in the right order - unlike for the GS1900-24E, where
even and uneven ports are flipped (up <-> down on panel).
Zyxel version code for this device (-24EP) is: ABTO
Signed-off-by: Mirko Vogt <mirko-openwrt@nanl.de>
At the moment we have to manually follow the default GCC version
also in config/Config-kernel.in. This tends to be forgotten at GCC
version bumps (just happened when switching from version 12 to 13).
Instead, introduce a hidden Kconfig symbol which implies KERNEL_WERROR
in toolchain/gcc/Config.in where it is visible for developers changing
the default version.
Also remove the explicit default on BUILDBOT to avoid a circular
dependency and also because buildbots anyway implicitly always select
the default GCC version.
Reference: https://github.com/openwrt/openwrt/pull/15064
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
e209a4ced1d8 add strdupa macro for compatibility
af1962b9a609 uclient: add helper function for getting ustream-ssl context/ops
488f1d52cfd2 http: add helper function for checking redirect status
b6e5548a3ecc uclient: defer read notifications to uloop timer
352fb3eeb408 http: call ustream_poll if not enough read data is available
e611e6d0ff0b add ucode binding
ddb18d265757 uclient: add function for getting the amount of pending read/write data
980220ad1762 ucode: fix a few ucode binding issues
6c16331e4bf5 ucode: add support for using a prototype for cb, pass it to callbacks
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Sync 6.1 patches with the RPi foundation.
Since rpi-6.6.y is now the main branch of the RPi foundation, there won't be
any new patches for linux 6.1.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
This is mostly a cosmetic cleanup. The absence of
the return statement was not causing any problems.
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
So, when updating the hash for at91bootstrap it was done via CHECK_ALL=1
so that updated the PKG_MIRROR_HASH for the main v4 version hash, but
at91bootstrap checkout version depends on the subtarget as well.
Choosing to build for sam9x will change the at91bootstrap version to v3
and this hash was not refreshed thus causing the CI to fail.
Fixes: 6918c637b7 ("treewide: package: update missed hashes after switch to ZSTD")
Signed-off-by: Robert Marko <robimarko@gmail.com>
With the switch to ZSTD for git clone packaging, hashes have changed so
fixup remaining package hashes that were missed in the inital update.
Fixes: b3c1c57 ("treewide: update PKG_MIRROR_HASH to zst")
Signed-off-by: Robert Marko <robimarko@gmail.com>
Kernel 6.6 added dynamic SWIOTLB allocation, but with it also started
allocating 64MB of the SWIOTLB bounce buffer by default which is quite a
lot of memory on most OpenWrt devices.
Luckily in kernel 6.7 arm64 received an optimization that reduces that
default size to 1MB per 1GB of RAM if certain criteria was met.
So in order to reclaim back 63MB of RAM which brought some ipq807x devices
close to OOM under load lets backport the upstream commit.
Signed-off-by: Robert Marko <robimarko@gmail.com>
When using zst instead of xz, the hash changes.
This was missed in the initial treewide updated.
Fixes: b3c1c57a35 ("treewide: update PKG_MIRROR_HASH to zst")
Signed-off-by: Robert Marko <robimarko@gmail.com>
Add options removed by `make kernel_oldconfig CONFIG_TARGET=subtarget', missing which causes compilation to stop and cause an error.
Fixes: 2a86425de1 ("x86: 6.6: refresh kernel config")
Signed-off-by: Mieczyslaw Nalewaj <namiltd@yahoo.com>
When an IBBS interface is configured for IBSS legacy mode, wdev.htmode
is empty. This is empty string results in an empty positional argument
to the "ibbs join" command, for example:
iw dev phy0-ibss0 ibss join crymesh 2412 '' fixed-freq beacon-interval 100
This empty argument is interpreted as an invalid HT mode by 'iw',
causing the entire command to fail and print a "usage" message:
daemon.notice netifd: radio0 (4527): Usage: iw [options] \
dev <devname> ibss join <SSID> <freq in MHz> ...
Although nobody will ever need more than 640K of IBSS, explicitly use
"NOHT" if an HT mode is not given. This fixes the problem.
Fixes: e56c5f7b27 ("hostapd: add ucode support, use ucode for the main ubus object")
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name> [extend to cover more cases]
The current .pc generated by make build system for zstd is wrong and
always assume a single-threaded library is used.
This wasn't the case for the meson build system that used his own logic
to generate the pkg-config file.
Add a patch to fix this by introducing install-mt-pc make target to
generate the correct pkg-config gile.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
This reverts commit 8d934c1196.
The update seems to be causing issues that need to be further
explored [0]. Let's revert it.
[0] - https://github.com/openwrt/openwrt/pull/15078
Signed-off-by: Nick Hainke <vincent@systemli.org>
We forgot to make sure install-includes is called for the libzstd in
order for it to install the required headers.
Fixes: 4b920e799f ("tools: zstd: convert to make and drop meson dependency")
Signed-off-by: Robert Marko <robimarko@gmail.com>
When using zst instead of xz, the hash changes. This commit fixes the
hash for packages and tools in core.
Signed-off-by: Paul Spooren <mail@aparcar.org>
In the light of recent XZ events, and fundamental XZ issues lets work on
moving away from using XZ.
So, use gz compressed tarballs as sources whenever possible.
dwarves only offers bz2 compressed tarballs, so use those as size
difference is minor compared to XZ.
Signed-off-by: Robert Marko <robimarko@gmail.com>
dwarves
libdeflate is currently intentionally being fetched via GIT.
However, with the move to using ZSTD to compress the cloned GIT repo
tarballs it means that we would first need to compile ZSTD.
But that means that we need to be able to unpack gzipped tarballs first
which we currently do by using libdeflate-gzip.
So, in order to do so lets fetch libdeflate as a tarball, use gzip to
extract it and then use libdeflate as regular for all other tools.
Signed-off-by: Robert Marko <robimarko@gmail.com>
ZSTD and libdeflate do not depend on SED nor flock, so instead of the whole
for loop that will filter 2 out of 4 core packages just specify that patch
and tar depend on sed explicitly.
ZSTD now depends on libdeflate since libdeflate-gzip will then be used to
unpack ZSTD as well as most tool archives.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Signed-off-by: Robert Marko <robimarko@gmail.com>
We dont really have a reason to deviate from the upstream default for
ZSTD_LEGACY_SUPPORT value of 5, as it will save a bit of space but
prevent decompressing data compressed with legacy ZSTD versions.
Signed-off-by: Robert Marko <robimarko@gmail.com>
ZTSD limits the safe compression level to a max of 19 as 20 to 22 cause
increased RAM usage. Higher levels require --ultra arg passed.
There isn't currently a way to set --ultra using ENV options similar to
ZSTD_CLEVEL and ZSTD_CLEVEL is limited to 19.
To fix this, we can increase the max safe compression level by providing
a custom ZSTDCLI_CLEVEL_MAX value with CFLAGS.
The max safe level is increased to 20.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Convert to make and drop meson dependency since it's not a core tools
and can't depend on advanced build system like cmake or meson.
On top of this make is the official build support and cmake/meson are
supported by 3rd parties.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
The compression is faster to (un)pack files, make use of it.
Also add a new build prerequirement, the `zstd` to (un)pack files.
Signed-off-by: Paul Spooren <mail@aparcar.org>
[ improve commit title ]
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
This fix a long lasting bug/inconsistency with rawgit method and
dl_github_archive script.
The dl_github_archive script works by using the github generated tar.gz
instead of cloning and checkout and the tar.gz is generated by using git
archive command that parse and apply .gitattributes rules.
rawgit command never handled .gitattributes and instead made a simple git
clone and checkout causing the inconsistency.
To fix the inconsistency, add extra steps to call git archive command
and generate an intermediate tar to apply .gitattributes rules.
Also for git log format, Github shorthash length is 8 instead of the
default 7, also apply this locally for each cloned repo to produce
consistent tar.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Add:
110-pwm-img-fix-clock-lookup.patch
- patch to fix a clock lookup issue from upstream
Update:
401-mtd-nor-support-mtd-name-from-device-tree.patch
- mtd-name lookup hack to reflect the updated spi_nor_scan function
Signed-off-by: Zoltan HERPAI <wigyori@uid0.hu>
On Fedora 40 system, some compile error happens when
building iconv-ostream.c. Linking to libiconv-full
fixes this.
Signed-off-by: Yanase Yuki <dev@zpc.st>
hostapd packages were accidentally left out. Clean up this mess by
changing the dependencies to hostapd-common
Signed-off-by: Felix Fietkau <nbd@nbd.name>
refreshed kernel config + patches
otherwise same as 6.1/5.15.
Tested on: WNDAP620, WNDAP660, MyBook Live Single, MR24, MX60, WNDR4700
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
With the default BUILD_BOT configuration on a linux 6.6 kernel,
the WNDR4700's kernel no longer fits into the alloted ~3.5MiB,
even with LZMA compression.
Bigger kernels are possible, but there's a problem with Netgear's
"bootcmd":
> if loadn_dniimg 0 0x180000 0x4e0000 && chk_dniimg 0x4e0000; then nand read 0x800000 0x180000 0x20000;bootm 0x500000 - 0x800040;else fw_recovery; fi"
This loads the dni-image starting offset 0x180000 from the NAND
flash (which is the DTB partition) to 0x4e0000 in the RAM. It then
checks whenever the provided image is "valid". If it is then it
reads the DTB again to 0x800000 in the RAM and starts the extraction
and boot process. (If the image wasn't valid then it starts the
automated firmware recovery).
The issues here are that first: the kernel image gets "squeezed"
between 0x500040 and 0x7fffff... And second, the decompressor
only has area 0x0 - 0x500000 for decompression.
Hence the image now requires to update the bootcmd by providing
new values (which have been successfully tested with the original
Netgear WNDR4700 v1.0.0.56 firmware) for the RAM locations and
make full use of the fact that loadn_dniimg loads the DTB as well.
This needs to be done only once. Just connect a serial adapter to
interface with uboot and overwrite (and save) the new bootcmd.
WARNING: The serial port needs a TTL/RS-232 3.3v level converter!
Steps:
0. Power-off the WNDR4700
1. Connect the serial interface (you need to open the WNDR4700)
2. Power-up the WNDR4700
3. Monitor the boot-sequence and hit "Enter"-key when it says:
"Hit any key to stop autoboot" (Be quick, you have a ~2 second window)
4. in the Prompt enter the following commands (copy & paste)
setenv bootcmd "if loadn_dniimg 0 0x180000 0xce0000 && chk_dniimg 0xce0000; then bootm 0xd00000 - 0xce0040;else fw_recovery; fi"
saveenv
run bootcmd
Note: This new bootcmd will also unbrick devices that were bricked
by the bigger 4.19-6.1 kernels.
Note2: This method was tested with a WNDR4700. A big kernel with most
debug features enabled on v6.6.22 measured 4.30 MiB when compressed
with lzma. The uncompressed kernel is 12.34 MiB. This is over the 3 MiB,
the device reserves for the kernel... But it booted! For bigger kernels,
the device needs repartitioning of the the ubi partition due to the
kernel+dtb not fitting into the partition.
Note3: For initramfs development. I would advice to load the initramfs
images to 0x800000 (or higher). i.e.: tftp 800000 wndr4700.bin
Note4: the fw_recovery uboot command to transfer the factory image to
the flash still works.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Daniel Golle