- Autopatcher was commiting atrocities to the upstream DTs that are so
carefully hand-crafted, by simply rewriting everything, or trying to be
smart and failing at it (`incremental: true`)
- Instead, add `add-only: true` mode, where we simply only ever add to the
end of the Makefile, adding each new .dts and optionally the `subdir-y`
for overlays
The `DEST_LANG` variable affects rootfs cache content (locale is generated via
`locale-gen "${DEST_LANG}"` before cache is packaged), but it was not included in the
cache hash calculation.
This meant that changing `DEST_LANG` would not invalidate the existing cache,
potentially resulting in images with incorrect locale.
Include `DEST_LANG` in the `hash_hooks` calculation in `calculate_rootfs_cache_id()`.
Changes
- `lib/functions/rootfs/create-cache.sh`: Add `LANG=${DEST_LANG}` to the hash input
Result
Changing `DEST_LANG` now properly invalidates rootfs cache, ensuring the correct locale
is generated.
Build process generated numerous locale warnings like:
bash: warning: setlocale: LC_ALL: cannot change locale (en_US.UTF-8)
perl: warning: Setting locale failed
This happened because the Docker container and host environment were
configured to use `en_US.UTF-8`, but the rootfs cache may not contain
this locale (only `C.utf8` and whatever `DEST_LANG` specifies,
e.g. `en_GB.utf8`).
When `chroot_sdcard` runs commands inside the rootfs, environment variables
are inherited from the host/Docker, causing locale lookup failures.
Then use `C.UTF-8` locale for the build environment instead of `en_US.UTF-8`.
This locale is always available in rootfs immediately after mmdebstrap,
requiring no generation.
Changes
`lib/functions/host/docker.sh`: Remove `en_US.UTF-8` locale generation,
set `LANG=C.UTF-8` in container environment
`lib/functions/host/prepare-host.sh`: Change locale exports
from `en_US.UTF-8` to `C.UTF-8`
- turns out everybody was wrong, including me
- some (older?) ATF sources won't work, ever; thus
- introduce ATF_SKIP_LDFLAGS=yes to skip it completely
- introduce ATF_SKIP_LDFLAGS_WL=yes to only skip the `-Wl,` prefix
- this is for ATF's that pass flag directly to linker, not gcc
- artifact-uboot: hash atf-building code into artifact version
This flag was breaking kernel's cc-option detection, causing GCC-specific
warnings (-Wpacked-not-aligned, -Wstringop-truncation, -Wmaybe-uninitialized)
to be incorrectly added to btrfs/drm/coresight builds when using clang.
* rtw88 mainline driver has good enough support for rtl8723cs
so stop patching the kernel with this ancient driver
* fixes rtl8703b (same chip as 8723cs) bluetooth firmware name file
Ubuntu 25.04+ replaced GNU coreutils with uutils coreutils, a Rust-based
reimplementation of Unix core utilities. These are different projects
with the same package name:
- GNU coreutils 9.x (C) - Ubuntu ≤24.04
- uutils coreutils 0.x (Rust) - Ubuntu ≥25.04
The uutils comm doesn't recognize sort output as sorted, causing
"comm: file is not in sorted order" errors.
Replace comm patterns with:
- grep -vxFf for set difference (lines in B but not in A)
- sort | uniq -d for finding duplicates
These alternatives don't depend on comm, ensuring compatibility
with both GNU and uutils coreutils.
* fix(rootfs): pass GIT_FIXED_WORKDIR inline to fetch_from_repo
Instead of using 'export GIT_FIXED_WORKDIR' and then 'unset' after,
pass the variable inline to the fetch_from_repo call. This ensures
the variable only exists in the context of that specific command
and doesn't leak into subsequent calls.
This follows the same pattern used in uboot-git.sh and kernel-git.sh.
---------
Co-authored-by: Viacheslav Bocharov <adeep@lexina.in>
Update all copyright notices in shell scripts from 2025 to 2026.
## Changes
- **Igor Pecovnik**: 2013-2025 → 2013-2026 (129 files)
- **Ricardo Pardini**: 2023-2025 → 2023-2026, 2020-2025 → 2020-2026 (5 files)
## Additional Improvements
Also updated the backtitle in `lib/functions/configuration/interactive.sh`:
- Changed title from "Armbian building script" to "Armbian Linux build framework"
- Removed docs link for cleaner display
- Uses dynamic year calculation with separate declaration (fixes shellcheck SC2155)
Problem: armbian-base-files artifact for jammy fails to build because
apt_find_upstream_package_version_and_download_url() looks for 'jammy-updates'
in https://github.armbian.com/base-files.json, but only 'jammy' key exists.
Root cause:
- For Ubuntu LTS (focal, jammy), code sets package_download_release to '${RELEASE}-updates'
- JSON file from github.armbian.com only has base release keys (jammy, noble, etc)
- jq query returns null for 'jammy-updates'
- Artifact excluded from build matrix after 10 retries
Solution: Add fallback logic
- First try with '-updates' suffix (jammy-updates)
- If not found and release ends with '-updates', retry with base release (jammy)
- This allows using base release data when -updates is not available
Impact:
- Fixes jammy base-files artifact build
- Allows jammy images to build (they depend on this artifact)
- Maintains preference for -updates when available
- No impact on other releases (Debian, non-LTS Ubuntu)
* sunxi-6.18: make the mess even worse
* fixing one of megis patches and add sunxi 32bit to the mess
* rewrite against 6.18
* fix media-ov5640-Don-t-powerup-the-sensor-during-driver-probe.patch
* fix media-sun6i-csi-implement-vidioc_enum_framesizes.patch
* fix misc-modem-power-Power-manager-for-modems.patch
* Fix usb-gadget-Fix-dangling-pointer-in-netdev-private-data.patch, include rewrite
* fix mmc-sunxi-mmc-Remove-runtime-PM.patch, two hunks no longer apply
* re-extract all of megis patches
* remove unneeded branch
* add note to disabled patch
* auto-generated, out of date
* drop megous drm patches in favor of Jernej's work. disable broken patches
* disable patch which breaks compilation for armhf
* disable breaking patch, rewrite everything
* remove patches unrelated to sunxi family
* fix spi dev compatible patch
* fix tsc2007 patch
* drop mainlined patch, adjust x96 mate T95 eth sd card hack
* remove upstreamed patch
* re-enable no longer broken
* another rewrite to align stuff properly
* adjust various comments in series.conf
* recover lost overlays
* uew5622: fix compilation against Linux 6.18
* fix Add-sunxi-addr-driver-Used-to-fix-uwe5622-bluetooth-MAC-address.patch
* adjust patch subject to make sense
* restore fixup creation
restore overlay prefix on opiz2
this needs to be properly sorted at some point
* bump to 6.18.1
sunxi and sunxi64 build just fine
* fix and re-enable drv-mfd-axp20x-add-sysfs-interface.patch
* rewrite patches
- Created patch/misc/rtw88/6.18/ directory with upstream driver patches
- Removed obsolete kernel 6.16 reference from patch conditions
- Updated comment to clarify that RF path detection fix is only needed
for kernel 6.1.x (upstreamed in 6.18+)
This fixes RTL8822CS WiFi initialization failures on kernel 6.18 where
the driver was not being properly patched due to missing version directory.
Fixes SDIO timeout errors during chip initialization.
Use $DOCKER_INFO to check for podman vs dockerd, and conditionally
`exec` and `dev` volume options, which are needed on podman but not
allowed on dockerd.
It hasn't been an issue for Docker because their defaults didn't
conflict, but Podman mounts volumes with nodev (and formerly noexec) by
default, which would break the build. Adding `dev` and `exec` to the
volume mount options ensures the needed setting whatever the defaults.
- stops after artifacts-reduced and adds kernels.ndjson and uboots.njdson via jq
- if one runs `./compile.sh inventory-artifacts` we shall get:
- `output/info/kernels.ndjson`: one line for each kernel build
- `output/info/uboots.ndjson`: one line for each u-boot build
- this is meant as base for other tooling that acts "for each kernel" or "for each u-boot" such as patch/config rewrites, etc
- cli-jsoninfo: better kernels.ndjson; new kernels-duplicate-config.json
- so we can detect mistakes ref LINUXCONFIG
- maybe one day the "Rewrite Kernel Config" GHA can re-use kernels.ndjson
- previous commit introduced targets-all-kernels.yaml, so one can run
- `./compile.sh inventory-artifacts TARGETS_TEMPLATE=targets-all-kernels.yaml`
- Enable bootscript templating using `envsubst`.
- Add generic bootscript.
- Fix: `test` does not support `==`, but deceptively returns `true` in case you use it.
- De-duplicate kernel commandline console options.
- call_extensions_kernel_config() now declares the vars and calls
armbian_kernel_config_apply_opts_from_arrays() at the end
- so each individual hook doesn't need to
- adapt family code using the old way too
- one shall NOT use the contents of .config to determine what's in the .config
- also: introduce `opts_m` for armbian_kernel_config_apply_opts_from_arrays()
* Wireless driver patching: fix inconsistency in matching
Mathing Allwinner family via LINUXFAMILY is giving different results. sunxi doesn't cover sun5* ...
* Re-adding wireless drivers that fall out of this bug + few additional
- two of those were already marked to be removed 6.12+
- for those in need: send fix upstream, wait for merge, update here and re-enable
- in case upstream abandoned: fork and fix, update here and re-enable
In the building when there is not any extensions with given pattern exist, 'enable_extensions_with_hostdeps_builtin_and_user()' will throw errors such as: 'Error 123 occurred in SUBSHELL SUBSHELL at /<some-path>/lib/functions/general/extensions.sh:582'. The changes will eliminate the error message and add prints to show the current searched folder and extensions count found in the folder.
Signed-off-by: diverger <diverger@live.cn>
