linux/meta-freescale.git/recipes-kernel, branch master

linux-qoriq: Order FILES before RDEPENDS per oelint.var.order

2026-08-11T13:14:40+00:00

Variable-definition order among lazily assigned variables does not affect BitBake expansion, so this is behaviour-preserving (verified with bitbake -e before/after showing identical variable expansion). Verified behaviour-preserving: expanded metadata (bitbake -e) unchanged vs master. Signed-off-by: Luciano Dittgen

recipes: Document the remaining custom tasks per oelint

2026-08-03T19:13:30+00:00

Set doc strings on the custom functions oelint-adv reports via 'oelint.task.docstrings': the three i.MX boot firmware deploy helpers, the libubootenv uboot-config fixup handler, the qemu-qoriq wrapper generator, the testfloat SoftFloat unpack step, the linux-imx device tree compatibility handler and the imx-mp4-parser QA-skip hook. imx-mp4-parser reuses the wording already used for the identical hook in imx-parser, so the two prebuilt-binary recipes read the same. opencv and libcamera are deliberately left alone. Both recipes carry their upstream recipe as a marked verbatim block ("meta-openembedded copy"), and the two functions oelint flags -- do_unpack_extra and do_package_recalculate_ipa_signatures -- sit inside those blocks, where meta-openembedded sets no doc string. Annotating them would break the diff against upstream for no functional gain, so those two findings stay visible. bitbake never reads the 'doc' flag when computing task checksums -- bb/data.py build_dependencies() only consults vardeps, vardepvalue, vardepsexclude, exports, postfuncs, prefuncs, lineno and filename -- so these additions cannot change a signature or invalidate sstate. Verified by re-linting each file. Signed-off-by: Luciano Dittgen

linux: Set SECTION on the i.MX kernel and header recipes

2026-08-03T19:13:30+00:00

oelint-adv reports 'oelint.var.suggestedvar: Variable SECTION should be set' for linux-fslc-imx, linux-fslc-lts (6.1 and 6.6), linux-imx and linux-imx-headers. The finding is genuine: kernel.bbclass sets no SECTION, so these recipes fall back to the bitbake.conf default SECTION = "base". The layer already answers this for its other kernels -- linux-fslc_6.12 sets SECTION = "kernel" directly after HOMEPAGE, and linux-qoriq.inc does the same -- so follow that idiom rather than inventing one. The four kernel recipes get "kernel"; linux-imx-headers only installs UAPI headers into ${includedir}/imx, so it gets "devel", matching oe-core's linux-libc-headers.inc. Set per recipe rather than in linux-imx.inc, because linux-fslc_6.12 already requires that include and sets SECTION itself; putting it in the include would report the variable as overridden. Verified by re-linting all five recipes: the SECTION findings are gone, no ordering finding is introduced, and linux-imx-headers is now free of findings entirely. This changes the Section/Group field of the generated packages from "base" to "kernel"/"devel". That field is not tracked by buildhistory, whose package "latest" files record only PV, PR, RPROVIDES, RDEPENDS, RRECOMMENDS, PKGSIZE, FILES and FILELIST; it is visible in tmp/pkgdata//runtime/ and in the package control metadata. Validation: - Image build on an i.MX 8M target; signature A/B on an i.MX 6 and a QorIQ target - Task graph and packaged output unchanged; pkgdata shows 656 packages moving base -> kernel (647 of them kernel-module-*) and 2 -> devel Signed-off-by: Luciano Dittgen

kernel-module-ar: Install the auto-response tools instead of copying them

2026-08-03T19:13:21+00:00

do_install copied everything matching ${S}/bin/ar_* with a bare "cp -f", so the packaged modes came from the build tree filtered through the building user's umask rather than from the recipe. oelint-adv reports this as 'oelint.task.nocopy'. A single mode would be wrong here, because ${S}/bin mixes two kinds of file. The directory is not in git at all -- the top-level Makefile creates it -- and the sub-Makefiles fill it with both compiled executables and configuration data: usr/ar-wakeup-lib/Makefile cp ar_wakeup_tool $(PWD)/bin/. cp ar_wakeup_rule_cfg $(PWD)/bin/. usr/ar-snmp-lib/lib-tool/ cp ar_snmp_tool $(PWD)/bin/. cp ../ar_snmp_cfg $(PWD)/bin/. So ar_snmp_tool and ar_wakeup_tool are linker output and want 0755, while ar_snmp_cfg and ar_wakeup_rule_cfg are git-tracked 0644 data files copied verbatim. Split the one "cp" into two "install" calls that say so. The two globs together match exactly the four files the old "ar_*" matched -- ar.ko has no underscore and was never picked up by it -- so the set of packaged files does not change, only the stated modes. Signed-off-by: Luciano Dittgen

Merge pull request #2610 from lucianogdittgen/oelint-cleanup-16

2026-07-27T21:23:35+00:00

oelint: Recipe SECTION/DEPENDS metadata and patch Upstream-Status cleanups

Merge pull request #2606 from lucianogdittgen/oelint-cleanup-15

2026-07-27T17:37:29+00:00

oelint: Order recipe variables and document custom tasks

kernel-module-isp-vvcam: Add oelint reason to patch Upstream-Status

2026-07-27T17:34:05+00:00

oelint.file.inappropriatemsg requires a reason token after an "Inappropriate" status. This local patch is a compatibility shim that keeps file->private_data management in the driver for pre-6.18 kernels; mainline already handles this differently, so it has no upstream destination and none of the specific reason keywords apply. Tag it "[other]"; the commit message above documents the rationale. Signed-off-by: Luciano Dittgen

kernel-module-nxp-wlan: Order SRC_URI before runtime provides

2026-07-27T12:26:33+00:00

oelint flagged SRC_URI placed after RCONFLICTS. Move the SRC_URI/SRCREV fetch block ahead of the RPROVIDES/RCONFLICTS runtime-packaging group per the canonical variable order. Pure reordering, no functional change. Signed-off-by: Luciano Dittgen

kernel-module-imx-gpu-viv: Order SRC_URI before SRCREV

2026-07-27T12:20:02+00:00

oelint flagged SRC_URI placed after SRCREV. Swap to the canonical SRC_URI/SRCREV order. Pure reordering, no functional change. Signed-off-by: Luciano Dittgen

linux-imx-headers: Order DEPENDS before the fetch group

2026-07-27T12:20:02+00:00

oelint flagged DEPENDS placed after SRCREV. Move the INHIBIT_DEFAULT_DEPS/DEPENDS pair ahead of SRC_URI per the canonical variable order. Pure reordering, no functional change. Signed-off-by: Luciano Dittgen