From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from aws-us-west-2-korg-lkml-1.web.codeaurora.org (localhost.localdomain [127.0.0.1]) by smtp.lore.kernel.org (Postfix) with ESMTP id DB980CCD190 for ; Wed, 18 Sep 2024 08:39:46 +0000 (UTC) Subject: Re: [yocto] read-only-rootfs and opkg To: "Alex Stewart" , yocto@lists.yoctoproject.org From: "Martin Siegumfeldt" X-Originating-Location: Copenhagen, Capital Region, DK (79.171.149.174) X-Originating-Platform: Linux Chrome 128 User-Agent: GROUPS.IO Web Poster MIME-Version: 1.0 Date: Wed, 18 Sep 2024 01:39:42 -0700 References: In-Reply-To: Message-ID: <28131.1726648782868302764@lists.yoctoproject.org> Content-Type: multipart/alternative; boundary="K5ZDEqH7RUkxoMbMZDqg" List-Id: X-Webhook-Received: from li982-79.members.linode.com [45.33.32.79] by aws-us-west-2-korg-lkml-1.web.codeaurora.org with HTTPS for ; Wed, 18 Sep 2024 08:39:46 -0000 X-Groupsio-URL: https://lists.yoctoproject.org/g/yocto/message/63858 --K5ZDEqH7RUkxoMbMZDqg Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Hi Alex, Thanks for your input - highly appreciated. It appears you have a point regarding the variable assignment - 'bitbake-ge= tvar OPKGLIBDIR' indeed indicates the intended assignment does not occur: $ bitbake-getvar --value OPKGLIBDIR NOTE: Starting bitbake server... /var/lib Doing the strong assignment (OPKGLIBDIR =3D "/data-os/opkg") does not make = a difference, which puzzles me as all assigments (I am able to grep) are we= ak: $ rg ^OPKGLIBDIR meta/classes-recipe/rootfs_ipk.bbclass 30:OPKGLIBDIR ??=3D "${localstatedir}/lib" meta/recipes-devtools/opkg/opkg_0.6.1.bb 32:OPKGLIBDIR ??=3D "${target_localstatedir}/lib" meta/classes-global/package_ipk.bbclass 24:OPKGLIBDIR ??=3D "${localstatedir}/lib" However, based on your reply, I am starting to wonder whether what I am try= ing to achieve is at all possible. The mount '/data-os' is a dedicated part= ition mounted R/W and used by overlayfs-etc, my assumption was to configure= opkg to use this for the meta-data. But if the meta-data is rendered build= -time during rootfs creation, I do not see how this is achievable? Thanks, Martin --K5ZDEqH7RUkxoMbMZDqg Content-Type: text/html; charset="utf-8" Content-Transfer-Encoding: quoted-printable
Hi Alex,
 
Thanks for your input - highly appreciated.
 
It appears you have a point regarding the variable assignment - 'bitba= ke-getvar OPKGLIBDIR' indeed indicates the intended assignment does not occ= ur:
$ bitbake-getvar --value OPKGLIBDIR
NOTE: Starting bitbake server= ...
/var/lib
 
Doing the strong assignment (OPKGLIBDIR =3D "/data-os/opkg") does not = make a difference, which puzzles me as all assigments (I am able to grep) a= re weak:

$ rg ^OPKGLIBDIR 
meta/classes-recipe/rootfs_ipk.bbcla= ss
30:OPKGLIBDIR ??=3D "${localstatedir}/lib"
meta/recipes-devtools/opkg/opkg_0.6.1.bb
32:OPKGLIBDIR ??=3D "${t= arget_localstatedir}/lib"
meta/classes-global/package_ipk.bbclass
24:OPKGLIBDIR ??=3D "${lo= calstatedir}/lib"
 
However, based on your reply, I am starting to wonder whether what I a= m trying to achieve is at all possible. The mount '/data-os' is a dedicated= partition mounted R/W and used by overlayfs-etc, my assumption was to conf= igure opkg to use this for the meta-data. But if the meta-data is rendered = build-time during rootfs creation, I do not see how this is achievable?
 
Thanks,
Martin
 
--K5ZDEqH7RUkxoMbMZDqg--