Edit: The fixed package has been built and is available in the repos. It is now safe to update.

Heads up for edge users, a recent losetup update broke the initramfs.

"This was fixed in postmarketos-initramfs 3.10.2. If you upgraded your system recently and are not running that version, **do not reboot** until you upgrade to postmarketos-initramfs 3.10.2"

It is important to note that the upgraded package is not built yet. If you would like, you can follow https://build.postmarketos.org/#queued to see when it is ready.

https://postmarketos.org/edge/2026/05/15/losetup-option-removal-leading-to-boot-failure/

@postmarketOS
Issues like this should not happen.

@oausi @postmarketOS

And another friendly reminder that most people *should not* run edge, and *should* run the latest stable release.

@craftyguy @oausi yeppers, unfortunate though it is.

with any luck this will be a lot better once we improve our infra, faster builders, integrated testing of packages on actual devices before they get published

@cas @craftyguy @oausi how though? This is an Alpine package upgrade breaking our use downstream, none of the proposed changes would prevent this from happening

@craftyguy @oausi @postmarketOS yep you should run edge only if you know that you may get a broken system at any point in time. edge is by definition unstable.

@aelin @craftyguy @oausi @cas yeah not sure faster builders and integrated testing of our packages before getting published would have done much to prevent this from happening.

It was an Alpine package upgrade, not an initramfs upgrade that broke booting.

@aelin @cas @craftyguy @oausi maybe regular booting of pmOS every few hours would've

@postmarketOS I was interested to test it too.
So I just flashed edge phosh on my test fp5 and got this error. Hmm, I'll put on pause updating my main device for a while.

@postmarketOS Interesting,
In the morning I've upgrade my main device and I do have postmarketos-initramfs 3.10.2. on it.
But now fresh flash os is broken.

@Oleksii @postmarketOS Reminder, use stable and not edge if you're not a developer ready to fix problems.

https://postmarketos.org/edge/2026/05/15/losetup-option-removal-leading-to-boot-failure/

@fun @aelin @craftyguy @oausi @cas

This class of bugs can be completely eliminated with immutable images and boot-testing them before publishing them, something that we are working towards with Duranium (https://postmarketos.org/blog/2026/03/17/introducing-duranium/) and hardware testing (https://postmarketos.org/blog/2026/01/21/hw-ci-mvp/).

In any case, I think @fun did a great job with fixing the bug and writing this edge blog post that not only describers the problem in detail, but also how users could work around it if needed. Thank you very much!

@craftyguy @oausi @postmarketOS

I think this comment is also a very good opportunity to post https://postmarketos.org/state/ again.

A community of volunteers identified an issue on Friday night and got it fixed in under 24h. Honestly, kudos to everybody involved. Anybody thinking such community of volunteers funded by donations that cover an absolutely tiny fraction of the work done (yet are critical for the sustainability of the project) should not be allowed to make mistakes is in my opinion, breaching the social contract under which the work is done.

@craftyguy define the group not( most people ). It's hard, I bet. I'd count me in, but I'm sure there'll be the moment I'll swear... 🙃

@chris edge is for people who can deal with things breaking sometimes, and help report, debug, and/or fix them.

@pabloyoyoista

excellent work, PostmarketOS team!!!!

@craftyguy @oausi @postmarketOS

@oausi @postmarketOS egde is tempting 'cause you get latest features, etc. But it's to test those features before going into stable. Occasional breakage can happen in any test suit. They are called: edge, testing, unstable, etc for that reason.

@oausi @postmarketOS They should - if they don't it means you've stagnated - and they will, so you design processes to handle them well. Which seem to have worked fine in this case, as the issue was contained to a version that's explicitly unstable, right?