Re: [yocto] overlayfs-etc on top of dm-verity?

Michael Opdenacker <[email protected]> Thu, 9 Apr 2026 21:18:21 +0200
Newsgroups org.yoctoproject.lists.yocto
Message-ID <[email protected]>
Hi Francesco,

On 3/26/26 11:19 PM, Francesco Valla wrote:
>> The code looks right though, I need to understand why this happens.
>>
>> Another weirdness that remains is these volatile mounts for /var/cache/,
>> /var/lib, /var/spool and /srv, which I didn't have with a regular read-only
>> root filesystem.
> AFAIK, this should be the regular behavior on a read-only root
> filesystem. The overlayfs mounts are created by services generated by:
>
>    meta/recipes-core/volatile-binds/volatile-binds.bb
>
> depending on the content of the VOLATILE_BINDS variable. For each couple
> of upperdir-lowerdir specified there, a service is generated that
> starts only if upperdir's parent is writable and lowerdir is not.
>
> E.g.:
>    lowerdir=/srv
>    upperdir=/var/volatile/srv
>
> In a vanilla openembedded-core system, a tmpfs is mounted on /var/volatile
> by the fstab (that is, by the fstab systemd generator), so the
> upperdir's parent directory (which is the same /var/volatile) is writable.
>
> You can force a copy+bind behavior setting AVOID_OVERLAYFS=1.

I didn't thank you for these great clarifications and tips. It's good to 
understand the "magic".
I eventually dropped overlayfs-etc and started implementing specific 
mount points using VOLATILE_BINDS indeed (with AVOID_OVERLAYFS=1).

It's a very easy to use mechanism thanks to the volatile-binds.bb recipe.
Thanks again, you made my day!
Cheers
Michael.

-- 
Root Commit
Embedded Linux Training and Consulting
https://rootcommit.com