- build-iso.yml: on tag push (v*.*.*), split ISO into 1.9G parts, GPG-sign
the sha256 with GPG_PRIVATE_KEY secret, and auto-create release with
softprops/action-gh-release@v2 attaching part files + sig + reassembly
instructions. Falls back to legacy release.published path.
- build-iso.yml: optional EFI Secure Boot signing step. If MOK_PRIVATE_KEY
+ MOK_CERT secrets are present, sbsign each .efi inside the ISO and
repack with xorriso; otherwise warn and ship unsigned. Refresh sha256.
- release-checksums.yml: new PR-time gate. Validates source + generated
CI kickstart, shellchecks scripts, parses every workflow YAML, and
asserts the split size stays under GitHub'''s 2 GiB asset cap.
- scripts/gen-mok-key.sh: idempotent MOK keypair generator (RSA-4096,
10y), outputs to gitignored build/keys/. Header documents mokutil
enrollment and gh secret upload. exec bit set in index.
- .gitignore: add build/keys/, *.priv, *.der.
User must add GitHub secrets before the next tagged release:
GPG_PRIVATE_KEY — armored private key for sha256 signing
MOK_PRIVATE_KEY — sbsign EFI signing key (PEM)
MOK_CERT — public cert (DER) for sbsign + mokutil enrollment
Live ISO stalled at dracut for 5+min on KVM with init_on_alloc=1
init_on_free=1 — kernel zeroes every page on alloc/free, brutal in
virtualized memory. Keep slab_nomerge + lockdown=integrity +
randomize_kstack + vsyscall=none for live (cheap). Re-add memory
init flags on installed system via veilor-firstboot post-install
GRUB edit (planned v0.3).
Upstream bug in /usr/lib/python3.14/site-packages/imgcreate/live.py:
if self._isDracut:
args["rootlabel"] = "live:LABEL=%(fslabel)s" # WRONG
else:
args["rootlabel"] = "CDLABEL=%(fslabel)s"
For dracut path on EFI grub it writes `root=live:LABEL=...` but
dracut needs `live:CDLABEL=...` to look up ISO9660 by CD volume id.
Result: parse-livenet hook stalls indefinitely.
CI now sed-patches the file in-place before build. Reported upstream
livecd-tools as separate task.
- kde-theme-apply.sh: search /etc/os-release.d/veilor (where overlay
put it) before falling back to $REPO/overlay path. Rewire symlinks
cleanly: /etc/os-release → ../usr/lib/os-release.
- Kickstart: useradd admin in chroot %post since livecd-creator skips
the `user` directive (no installer phase). Blank pw + expired = forced
reset at first login same as before.
Found via debugfs: overlay copy succeeds (veilor-power, tuned profiles,
sshd-hardening, sudoers, systemd units all present in v0.2.1 rootfs) but
`mkdir + cp assets/scripts` aborted with set -eu — leaves /usr/share/
veilor-os missing → all chroot %post scripts fail. Switch to set +e on cp
plus persist trace log to /var/log/veilor-nochroot.log for next debug.
POSTTRANS ldconfig hit ENOSPC/ROFS — KDE install + dnf cache + scriptlet
working set exceeds 16G tmpfs. Move livecd-creator tmpdir to /var/lmc on
runner's host ext4 disk.
Agent A: missing livesys-scripts + anaconda-live = lorax can't build EFI/BOOT.
Agent B: livecd-creator ignores url=, only reads repo.repoList — added
explicit repo --name=fedora to feed it the base.
Both Fedora's own pipeline + livecd-creator now have what they need.
Live image plumbing in %post: enable livesys.service livesys-late.service
tmp.mount, reset machine-id.
Past grub2-mkimage. Failed at xorrisofs final ISO assembly because EFI/BOOT
dir not built — needs grub2-efi-x64-modules to compile standalone grubx64.efi.
CI made it through full install, configure, %post, squashfs build,
initrd rebuild — failed at final boot.iso wrap because grub2-mkimage
needed /usr/lib/grub/i386-pc/moddep.lst (BIOS legacy boot modules).
Hybrid BIOS+UEFI ISO requires both grub variants.
Local builds need fix-repo because host has stale libselinux vs newer pcre2.
CI fresh container has matched libs, fix-repo unnecessary and refs invalid
(file:///tmp/veilor-fix-repo not present in CI). sed strips that ks line.
CI builds in fresh Fedora 43 container — matched pcre2/libselinux/selinux-policy
versions, no fix-repo hack needed. Container starts every run from clean
state, no zombie collisions. Fastest path to first green ISO.
