linux/meta-virtualization.git/recipes-kernel, branch master-next

kernel: skip kcfg search path injection for kernels with kernel-cache

2026-03-12T19:06:40+00:00

The kcfg sysroot search path (type=kmeta) was being injected unconditionally, which caused fragment conflicts for linux-yocto kernels that already have their own kernel-cache meta branch. The injected search path would roll back fragments to an older set. Make the injection conditional and fix the root cause: - kernel_cache_feature() now checks SRC_URI for type=kmeta entries. Kernels with kernel-cache get short paths (e.g. cfg/container.scc) that resolve directly via the kernel-cache's search path. Kernels without kernel-cache get the sysroot-relative ../../ paths as before. - distro_cond_feature() now passes d.getVar('SRC_URI') instead of an empty string, so kernel_cache_feature() can actually see whether the kernel has a kernel-cache. - SRC_URI append and inject_kcfg_search_path prefunc skip when type=kmeta is already present. - Remove unused kernel_cache_cond_feature() which was the original conditional implementation but was never wired up. This restores the original design intent from commit 5c212911 ("allow conditional use of yocto-cfg-fragments"): kernel-cache is the first choice for fragment resolution, sysroot is the fallback. Signed-off-by: Bruce Ashfield

yocto-cfg-fragments: bump to -tip

2026-03-12T17:43:27+00:00

bc293057 bcm-2xxx-rpi: Fix configuration audit warnings (LEDS, BACKLIGHT) 8ef043d5 kver: bumping to v6.18.16 104a5c9d bcm-2xxx-rpi: Add raspberry pi5 config and add input feature by default 9b173d3a genericarm64: drop CONFIG_PINCTRL_SM8350_LPASS_LPI 0b1ea0e5 genericarm64.cfg: enable CONFIG_INTERCONNECT_QCOM_QCS8300 2fa1b419 genericarm64-pinctrl.cfg: enable CONFIG_PINCTRL_QCS8300 53f6a3a8 kver: bumping to v6.18.13 98a36302 cfgs/nfc: create an intel-nfc-vendor specific feature 6350b56a netfilter.cfg: enable NF_CONNTRACK_SNMP a425fdd0 pmem.cfg: enable DEV_DAX and DEV_DAX_HMEM 70eeaeff genericarm64.scc: enable TRANSPARENT_HUGEPAGE support d62a1267 security-arm64.cfg: rename CFI_CLANG to CFI b925ef94 security-arm64.cfg: correctly enable KASAN 1fdb98a4 security-arm64.cfg: add comment for ARM64_BTI_KERNEL 6d8bf6f0 security.cfg: move RANDOMIZE_MEMORY to x86_64 18fcb6dc arm.scc: move ARM_CPUIDLE to arm only 7e1d9b2b numa_x86_64.scc: move x86_64 config from numa.cfg to numa_x86_64.cfg e93e999e sound.cfg: split SND_SOC_WM8731 to I2C and SPI variants 5a567776 sound_x86.scc: take over x86 configs from sound.scc 1ae7ce94 genericarm64.cfg: enable MTD NAND ECC support 15d48f71 genericarm64.cfg: enable TI MTD NAND support 40a67254 genericarm64-serial.cfg: set SERIAL_8250_CONSOLE to y 5ee6148e qemu-kvm.cfg: remove extra space 9f780417 firmware.cfg: whitespace fixes bc221187 genericarm64.cfg: fix CONFIG_PM_DEVFREQ_EVENT aef69bf3 Revert "genericarm64.scc: enable OP-TEE support" 6f597e1a genericarm64.cfg: enable SCSI support for HiSilicon etc f1f313cf genericarm64.scc: enable RPMB support a4565911 genericarm64.scc: enable GNSS support 391566e2 gnss.scc: add feature d40a455d genericarm64.cfg: enable more bus drivers 20e41104 genericarm64.cfg: enable PCI_PASID support 0b2967f5 genericarm64.scc: enable NFC support 2643b37e nfc-vendor.scc: move Intel configs to nfc-vendor-intel.cfg b1d00b01 genericarm64.scc: enable RFKILL LED, INPUT and GPIO 92610953 rfkill-extra.scc: add fragment for RFKILL LEDs, input and GPIO support fc7d5f45 bluetooth.cfg: enable BT_LEDS support f18ec272 genericarm64.scc: enable HSR 79fa2c17 hsr.scc: add feature for High-availability Seamless Redundancy (HSR & PRP) 1313923a genericarm64.scc: enable Time Sensitive Networking e190eeff intel-x86.scc: enable hibernation with feature c941f4ee genericarm64.scc: enable hibernation support ce5c6d9d hibernation.scc: add feature 83620a7d bluetooth-usb.cfg: enable MediaTek and RealTek support cf6e1261 bluetooth-hw.cfg: enable BT_MTK ffdd0769 genericarm64.cfg: enable Microchip PHY support 9e9536d1 features/iommu: remove CONFIG_INTEL_IOMMU_FLOPPY_WA=y 61265288 genericarm64.cfg: enable CONFIG_QCA808X_PHY 8e6a09c7 kver: bumping to v6.18.11 Signed-off-by: Bruce Ashfield

linux-yocto: add vxn distro feature and Xen PV kernel config

2026-02-26T01:05:01+00:00

Add 'vxn' to DISTRO_FEATURES in vruntime.conf to enable Xen PV guest kernel configuration. When set, linux-yocto_virtualization.inc includes vxn.cfg which enables Xen PVH frontend drivers and the 9p Xen transport (NET_9P_XEN). The base 9p stack comes from the existing vcontainer feature via cfg/container.scc. Signed-off-by: Bruce Ashfield

vxn: add Xen DomU container runtime with OCI image support

2026-02-26T01:05:01+00:00

vxn runs OCI containers as Xen DomU guests — the VM IS the container. No Docker/containerd runs inside the guest; the init script directly mounts the container rootfs and execs the entrypoint via chroot. Host-side (Dom0): - vxn.sh: Docker-like CLI wrapper (sets HYPERVISOR=xen) - vrunner-backend-xen.sh: Xen xl backend for vrunner - hv_prepare_container(): pulls OCI images via skopeo, resolves entrypoint from OCI config using jq on host - xl create for VM lifecycle (PVH on aarch64, PV on x86_64) - Bridge networking with iptables DNAT for port forwards - Console capture via xl console for ephemeral mode Guest-side (DomU): - vxn-init.sh: mounts container rootfs from input disk, extracts OCI layers, execs entrypoint via chroot - Supports containers with or without /bin/sh - grep/sed fallback for OCI config parsing (no jq needed) - Daemon mode with command loop on hvc1 - vcontainer-init-common.sh: hypervisor detection, head -n fix - vcontainer-preinit.sh: init selection via vcontainer.init= Build system: - vxn-initramfs-create.inc: assembles boot blobs from vruntime multiconfig, injects vxn-init.sh into rootfs squashfs - vxn_1.0.bb: Dom0 package with scripts + blobs - nostamp on install/package chain (blobs from DEPLOY_DIR are untracked by sstate) - vxn.cfg: Xen PV kernel config fragment Tested: vxn -it --no-daemon run --rm hello-world Signed-off-by: Bruce Ashfield

kernel: drop 6.12 .inc and append

2026-02-25T18:30:39+00:00

6.12 has been removed from OEcore, so we drop our appends as well. Signed-off-by: Bruce Ashfield

kernel: inject local search path to fragment processing

2026-02-10T21:04:20+00:00

Inject the kcfg sysroot root into the scc search path so compound .scc files (like container.scc) can resolve their include directives. kernel-yocto.bbclass adds type=kmeta directories to -I includes via find_kernel_feature_dirs/feat_dirs. The prefunc replaces the placeholder directory with a symlink to the kcfg sysroot so that spp can resolve include directives within scc files that reference other fragments by relative path (e.g. "include cfg/9p.scc" inside container.scc). Signed-off-by: Bruce Ashfield

kernel: consolidate kernel configuration options

2026-02-10T21:04:20+00:00

Move from the meta-virt hosted fragments to common fragments that are in the kernel-cache. Signed-off-by: Bruce Ashfield

yocto-cfg-fragments: update to 6.18-tip

2026-02-10T21:04:20+00:00

In particular we pick up the consolidated container configuration options. Signed-off-by: Bruce Ashfield

kernel: drop 6.10, 6.16 and 6.17 support

2026-02-10T21:04:20+00:00

These kernels are EOL upstream and in OE core. Dropping their enablement in meta-virt. Signed-off-by: Bruce Ashfield

linux-yocto: add iptables legacy kernel config for Docker

2026-02-09T03:34:12+00:00

Kernel 6.18+ split iptables into legacy/nftables backends. Docker requires the legacy iptables support, so add the kernel configuration for the full dependency chain: - CONFIG_NETFILTER_XTABLES_LEGACY=y - CONFIG_IP_NF_IPTABLES_LEGACY=m - CONFIG_IP_NF_FILTER=m - CONFIG_IP_NF_NAT=m - CONFIG_IP_NF_TARGET_MASQUERADE=m Without these, Docker's iptables rules fail to load on 6.18+ kernels. Signed-off-by: Bruce Ashfield