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 3CF04C4345F for ; Wed, 24 Apr 2024 20:08:54 +0000 (UTC) Received: from mailout4.zoneedit.com (mailout4.zoneedit.com [64.68.198.64]) by mx.groups.io with SMTP id smtpd.web10.6698.1713989327955380656 for ; Wed, 24 Apr 2024 13:08:48 -0700 Authentication-Results: mx.groups.io; dkim=none (message not signed); spf=pass (domain: denix.org, ip: 64.68.198.64, mailfrom: denis@denix.org) Received: from localhost (localhost [127.0.0.1]) by mailout4.zoneedit.com (Postfix) with ESMTP id 045F240C8A; Wed, 24 Apr 2024 20:08:47 +0000 (UTC) Received: from mailout4.zoneedit.com ([127.0.0.1]) by localhost (zmo14-pco.easydns.vpn [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id O3h-m-Gk3_ee; Wed, 24 Apr 2024 20:08:46 +0000 (UTC) Received: from mail.denix.org (pool-100-15-87-159.washdc.fios.verizon.net [100.15.87.159]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by mailout4.zoneedit.com (Postfix) with ESMTPSA id 3EAAD40BE6; Wed, 24 Apr 2024 20:08:39 +0000 (UTC) Received: by mail.denix.org (Postfix, from userid 1000) id E46A9163F33; Wed, 24 Apr 2024 16:08:38 -0400 (EDT) Date: Wed, 24 Apr 2024 16:08:38 -0400 From: Denys Dmytriyenko To: Jon Cormier Cc: Chirag Shilwant , Ryan Eatmon , meta-arago@lists.yoctoproject.org, Gyan Gupta , dannenberg@ti.com Subject: Re: [meta-arago][master/scarthgap][PATCH] meta-arago-distro: recipes-core: Add no-fstab-update option to WIC_CREATE_EXTRA_ARG Message-ID: <20240424200838.GD28791@denix.org> References: <20240420125318.397900-1-c-shilwant@ti.com> <20240421030736.GZ28791@denix.org> <1e3b95f7-e43f-44c9-92c5-15fdbc5ce8ab@ti.com> <20240422223206.GB28791@denix.org> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.20 (2009-06-14) Content-Transfer-Encoding: quoted-printable 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, 24 Apr 2024 20:08:54 -0000 X-Groupsio-URL: https://lists.yoctoproject.org/g/meta-arago/message/15308 On Mon, Apr 22, 2024 at 02:05:46PM -1000, Jon Cormier wrote: > Looks good to me. >=20 > Though i'm curious what the /boot -> boot part does. When we did it we > deleted the /boot entirely. wic --help says: * 'part' or 'partition' This command creates a partition on the system and uses the following syntax: part [] The is where the partition will be mounted and must take of one of the following forms: /: For example: /, /usr, or /home swap: The partition will be used as swap space. If a is not specified the partition will be created but will not be mounted. Partitions with a specified will be automatically mou= nted. So, the part is optional, you can definitely fully omit it. = But,=20 I've seen many layers use it as a name, mostly in cases of --source=3Draw= copy=20 or --source=3Dempty as those don't have a --label parameter. But it behav= es the=20 same with --source=3Dbootimg-partition and won't treat it as actual mount= point=20 when it doesn't start with a / character. It's rather undocumented, but i= t=20 works. I guess we can also remove it to just follow the docs, since we ha= ve=20 label as well to avoid confusion... > Also while we are touching the file. Could add the M to "128" in the no= n > efi file so it matches and is more clear. I guess I can send a v2 for that... --=20 Denys > On Mon, Apr 22, 2024, 12:32=E2=80=AFPM Denys Dmytriyenko via lists.yoct= oproject.org > wrote: >=20 > > On Sun, Apr 21, 2024 at 02:42:23PM +0530, Chirag Shilwant wrote: > > > + Andreas > > > On 21/04/24 08:37, Denys Dmytriyenko wrote: > > > >On Sat, Apr 20, 2024 at 06:23:18PM +0530, Chirag Shilwant wrote: > > > >>- Ideally running `ls /boot/` in filesystem should list contents = from > > /rootfs/boot partition > > > >>but currently it lists the contents of /boot partition of the WIC > > Image. > > > >> > > > >>Eg - Incase of AM64x, > > > >>``` > > > >> root@am64xx-evm:~# ls /boot/ > > > >> EFI Image tiboot3-am64x-gp-evm.bin > > tiboot3-am64x_sr2-hs-evm.bin tiboot3-am64x_sr2-hs-fs-evm.bin tiboot= 3.bin > > tispl.bin u-boot.img uEnv.txt > > > >Will this change affect EFI? As there's EFI directory in that part= ition. > > > > > > The EFI partition under /rootfs/boot will still be present. > > > Following is the output of `ls /boot` with this patch, > > > > > > ``` > > > root@am64xx-evm:~# ls /boot > > > EFI Image Image-6.1.80-ti-g2e423244f8c0 dtb fitImage > > > fitImage-6.1.80-ti-g2e423244f8c0 > > > > > > ``` > > > > Thanks, I'll do some additional testing with EFI a bit later. > > > > > > > >>- To fix this issue, add --no-fstab-update flag to > > WIC_CREATE_EXTRA_ARG in arago-image.inc > > > >>to prevent WIC image from adding an additional entry in /etc/fsta= b. > > > >This flag is global and will prevent all other partitions from > > populating > > > >/etc/fstab properly. Sure, meta-ti only has /boot and / defined, b= ut > > other > > > >layers could have own wic files with e.g. /home or /data partition= s. > > Mounting > > > >those will require a custom /etc/fstab instead of wic creating > > corresponding > > > >entries automatically. > > > > > > > > > > > > > Apparently, this change was being carried in processor sdk layer fr= om > > > quite a long time (from dunfell) to resolve a customer issue > > > reported by Andreas. > > > Diving into our internal JIRA system showed that this fix was > > > suggested by Andreas > > > & was resolved by the SDK team back then by keeping this change in > > > processor sdk layer [1]. > > > Not sure why this was in processor sdk layer though. Maybe Andreas > > > can provide additional details. > > > > I tested and submitted a different fix last night to adjust WIC templ= ates > > to > > not mount boot partition under /boot directory[1]. I still believe do= ing > > it > > just for that one partition is better than disabling fstab updates > > globally. > > > > [1] > > https://patchwork.yoctoproject.org/project/ti/patch/20240421191458.22= 21303-1-denis@denix.org/ > > > > Can you please take a look and maybe test this option for your use ca= se? > > Thanks.