Add unique mac for each AP-UP interface asociated with batman #1200
Open
javierbrk wants to merge 3519 commits into
Open
Add unique mac for each AP-UP interface asociated with batman #1200javierbrk wants to merge 3519 commits into
javierbrk wants to merge 3519 commits into
Conversation
…alization Easing libremesh virtualization
Follow-up to libremesh#938: Fix README link and improve clusterssh command
…ages packages: drop abandoned luci-mod-status and luci-lib-jquery-flot
docs: update changelog
links to new generic kernel and rootfs for QEMU testing.
tools/qemu_dev_stop [NODE_ID] will stop the node-id vm
The -N option is not supported by the netcat version shipped by default with debian. Use nc -w 1 instead, this will cause the connection to timeout after 1 second. system_powerdown command can take some time to finish, this prevents the terminal idling from the user perspective.
Previous link led to a 404 page, new one to the official documentation of busted
Update testing.md
…hcat Replacing deferrable reboot with watchcat
APuP support for LibreMesh
Ethernet configuration using lime-app
javierbrk/openwrt-packages net/vwifi is missing PKG_MIRROR_HASH, which OpenWrt 24.10 SDK requires (Package HASH check failed). Our fork already has the hash; Javier needs to add it before his package can be used as a direct replacement.
…ly belkin Signed-off-by: Franco <franco.riba@mi.unc.edu.ar>
Signed-off-by: Franco <franco.riba@mi.unc.edu.ar>
…collision Reverts f87e112. Pairing two identical-model belkins exercises a LibreMesh identity collision (primary_mac() reads eth0, and same-model devices in initramfs share the DTS default eth0 MAC) that cannot be fixed in test glue. Restore the original walking-chain layout used by the May 5 schedule run (25364365026) where bananapi_bpi-r4 (wired) is paired with belkin_rt3200_3 (wifi mesh via VLAN 200). The same-model MAC collision should be addressed in lime-packages itself (e.g. a uci-defaults seed before lime-config) when a stable identity source is available on the device.
…e is skipped test-mesh and test-mesh-pairs keep test-firmware in `needs` only to serialise lab access on the labgrid lock; they do not consume any artifact or result from it. Default GitHub Actions behaviour, however, skips a dependent job whenever any `needs` job is in `skipped` state, which made `workflow_dispatch` with `physical_mesh_pairs=true` (but `physical_single=false`) skip the mesh-pairs run that the user explicitly asked for. Wrap the existing `if:` with `!cancelled()` and add explicit `needs.<job>.result` checks so the mesh jobs run when test-firmware is either `success` or `skipped`, and skip only when test-firmware truly failed.
Signed-off-by: Franco <franco.riba@mi.unc.edu.ar>
Signed-off-by: Franco <franco.riba@mi.unc.edu.ar>
Member
|
@javierbrk @G10h4ck what's the status of this? It would be nice to have working APuP for the next release! |
Member
Author
|
There are some pending requests for changes. I'll try to do it soon. If
anyone wants to pick it up in the meantime, that's ok with me.
Saludos
Ing. Javier Alejandro Jorge
…On Mon, 15 Jun 2026 at 10:02, Ilario Gelmetti ***@***.***> wrote:
*ilario* left a comment (libremesh/lime-packages#1200)
<#1200 (comment)>
@javierbrk <https://github.com/javierbrk> @G10h4ck
<https://github.com/G10h4ck> what's the status of this? It would be nice
to have working APuP for the next release!
—
Reply to this email directly, view it on GitHub
<#1200?email_source=notifications&email_token=ABE2I2GAO43MG3YFTUMZNHD477XXHA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTINZQHAYTQOBSG4ZKM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#issuecomment-4708188272>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ABE2I2E55F4537NJYHGWDT3477XXHAVCNFSNUABEKJSXA33TNF2G64TZHMYTAMBQGM4DEMB3JFZXG5LFHMZTCOJTG4YTCMJXG2QXMAQ>
.
Triage notifications, keep track of coding agent tasks and review pull
requests on the go with GitHub Mobile for iOS
<https://github.com/notifications/mobile/ios/ABE2I2DAFHXIZPLSJSFQHBD477XXHA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTINZQHAYTQOBSG4ZKM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJKTGN5XXIZLSL5UW64Y>
and Android
<https://github.com/notifications/mobile/android/ABE2I2E6SDRVMCJVXQ33FSD477XXHA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTINZQHAYTQOBSG4ZKM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLTGN5XXIZLSL5QW4ZDSN5UWI>.
Download it today!
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
Add IMAGE_FORMAT=dual-tftp to build_image.sh that emits kernel.bin and rootfs.cpio as separate artifacts. U-Boot TFTP-loads each to a distinct RAM address and passes rd_start/rd_size via bootargs so the MIPS kernel mounts the external CPIO as rootfs. - build_image.sh: new dual-tftp format (kernel + CPIO, no FIT repacking) - targets.yml: librerouter_v1 entry (mips_24kc, ath79-generic) - lab_stage_firmware.sh: detect and stage dual-file artifacts - build-firmware.yml: forward LG_IMAGE_INITRD to test step
The Atheros U-Boot 1.1.x on the LibreRouter does not pass
rd_start/rd_size bootargs to the kernel (${var} expansion
not supported, and the kernel may have CONFIG_CMDLINE_OVERRIDE).
Wrap the rootfs CPIO as a uImage ramdisk (mkimage -T ramdisk)
so `bootm <kernel> <ramdisk>` makes U-Boot pass initrd info
natively through the MIPS boot params mechanism.
base-files creates these but verify and create if missing to prevent mount_root / preinit failures when /proc is not available.
…boot The OpenWrt preinit hook 80_mount_root gates do_mount_root behind [ "$INITRAMFS" = "1" ]. Without this variable set, mount_root runs on initramfs-booted devices, finds rootfs_data on flash, attempts a jffs2 overlay pivot_root which fails (rootfs doesn't support it), and the ramoverlay fallback loses /proc - cascading into lime-config never running and the device booting as root@(none). The upstream OpenWrt build system places a proper /init script at target/linux/generic/other-files/init that exports INITRAMFS=1 before exec'ing /sbin/init. Our build_image.sh was creating a bare symlink /init -> /sbin/init instead, so procd started without INITRAMFS in the environment. Replace the symlink with the equivalent 3-line script that sets INITRAMFS=1 and execs /sbin/init. This ensures mount_root is skipped on all initramfs targets (dual-tftp, FIT, multi-uimage).
feat(ci): add LibreRouter v1 support via dual-TFTP boot
Signed-off-by: Franco <franco.riba@mi.unc.edu.ar>
|
Picked this up per @javierbrk's note above. I've opened javierbrk#16 against the
Full unit suite passes (305/305). Happy to fold/squash however suits the maintainers — main goal is to get APuP + batman-adv ready for the next release. |
Updated versioning to use a specific commit hash and date.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
15 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
APuP creates an interface for each radio named "wlanX-apup" then for each peer Y in that radio creates a new interface called wlanX-peerY. This interface is then associated to a new device called wlanX-peerY_WZ were WZ are numbers associated with protocols. In my particular case 29 is associated with batman and derived from the network name.
Each time a new peer appears the device linked with the peer WiFi interface has the same mac address and batman requires that all the interfaces have different macs. So this pull request adds the new mac feature to the each interface.