All of lore.kernel.org
 help / color / mirror / Atom feed
From: Denys Dmytriyenko <denis@denix.org>
To: reatmon@ti.com
Cc: Moteen Shah <m-shah@ti.com>,
	meta-ti@lists.yoctoproject.org, praneeth@ti.com, vigneshr@ti.com,
	anshuld@ti.com, gehariprasath@ti.com, denys@konsulko.com
Subject: Re: [meta-ti] [PATCH] packagegroup-ti-core-initramfs: Remove runtime managers from initramfs image
Date: Thu, 26 Mar 2026 17:02:21 -0400	[thread overview]
Message-ID: <20260326210221.GA11121@denix.org> (raw)
In-Reply-To: <6f06cab4-cafb-49b7-83d1-acb73d2aced8@ti.com>

On Wed, Mar 25, 2026 at 08:33:01AM -0500, Ryan Eatmon via lists.yoctoproject.org wrote:
> 
> 
> On 3/25/2026 1:28 AM, Moteen Shah wrote:
> >Remove runtime managers since ti-core-initramfs does not requires
> >login and other services as it is being used as a read only
> >image during the boot. Modules can be probed using the udev from
> >initrams-module-udev, hence remove the init manager along with
> >others.
> 
> I don't think this is going to pass muster.  We just added those in
> to make the initramfs function correctly with the rest of the system
> post boot.
> 
> What is the more in depth reason you want to remove them?  Are they
> causing you a problem?
> 
> I'll let Denys chime in more since he did the work to figure out
> what all we needed to add in to make it "correct".

It's not really whether it's correct or not. Using busybox instead of systemd 
is also correct, but...

Back when we were initially discussing adding initramfs last year, the topic 
of busybox vs. systemd as init manager came up - once PID 1 process starts by 
the kernel from initramfs, it doesn't change when pivot_root into the full 
rootfs. If PID 1 is busybox, later started systemd won't be PID 1 and that 
changes some of its behavior and causes complications, as I recall. It was 
requested to simplify things and use systemd as init manager in initramfs, 
even if the size would be larger than using only busybox.

That's how ti-core-initramfs was originally implemented. Then systemd got 
upgraded upstream in OE-Core and it had some packaging and dependency issues, 
which blew up initramfs size significantly. Ryan changed the way initramfs 
recipe is done, but that inadvertently removed systemd and other default 
VIRTUAL_RUNTIME components. Eventually systemd packaging got fixed in OE-Core 
and we were able to re-add systemd. Here's the corresponding discussion on the 
list - there were no comments or concerns:

https://lists.yoctoproject.org/g/meta-ti/topic/117180106#msg19431


> >Signed-off-by: Moteen Shah <m-shah@ti.com>
> >---
> >This patch will help in reducing the size of the initramfs to about 7 MB's compressed and 25 MB's uncompressed
> >Boot logs: https://gist.github.com/Jamm02/ebdb124064e253208ee479f46edf127f
> >  .../recipes-ti/initramfs/packagegroup-ti-core-initramfs.bb    | 4 ----
> >  1 file changed, 4 deletions(-)
> >
> >diff --git a/meta-ti-bsp/recipes-ti/initramfs/packagegroup-ti-core-initramfs.bb b/meta-ti-bsp/recipes-ti/initramfs/packagegroup-ti-core-initramfs.bb
> >index a9eff847..f496f21a 100644
> >--- a/meta-ti-bsp/recipes-ti/initramfs/packagegroup-ti-core-initramfs.bb
> >+++ b/meta-ti-bsp/recipes-ti/initramfs/packagegroup-ti-core-initramfs.bb
> >@@ -11,10 +11,6 @@ TI_CORE_INITRAMFS_KERNEL_MODULES ?= ""
> >  RDEPENDS:${PN} += "\
> >      ${TI_CORE_INITRAMFS_KERNEL_MODULES} \
> >      ${VIRTUAL-RUNTIME_base-utils} \
> >-    ${VIRTUAL-RUNTIME_login_manager} \
> >-    ${VIRTUAL-RUNTIME_init_manager} \
> >-    ${VIRTUAL-RUNTIME_dev_manager} \
> >-    ${VIRTUAL-RUNTIME_update-alternatives} \
> >      netbase \
> >      base-files \
> >      base-passwd \


  parent reply	other threads:[~2026-03-26 21:02 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-03-25  6:28 [PATCH] packagegroup-ti-core-initramfs: Remove runtime managers from initramfs image Moteen Shah
2026-03-25  6:42 ` PRC Automation
2026-03-25 13:33 ` Ryan Eatmon
2026-03-26  5:07   ` Moteen Shah
2026-03-26 21:02   ` Denys Dmytriyenko [this message]
2026-03-26 10:54 ` Anshul Dalal

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260326210221.GA11121@denix.org \
    --to=denis@denix.org \
    --cc=anshuld@ti.com \
    --cc=denys@konsulko.com \
    --cc=gehariprasath@ti.com \
    --cc=m-shah@ti.com \
    --cc=meta-ti@lists.yoctoproject.org \
    --cc=praneeth@ti.com \
    --cc=reatmon@ti.com \
    --cc=vigneshr@ti.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.