initramfs-framework: remove hard dependency with busybox
A message that is not in the archive (sent off-list or not yet received): [email protected]
PATCH v2 1/1 initramfs-framework: remove hard dependency with busybox
In 'finish'script the Switch_root command with option -c is busybox specific. When package-core-base-utils is selected the boot is breaking with message "Switch_root: invalid option — ‘c’.It was reproduced using qemu. Remove -c option to make it compatible with both providers.
It is tested using qemu with busybox and package-core-base-utils. In both cases qemu boots successfully and filesystem is accessible.
Fixes [YOCTO #15716]
---v1->v2
* the commit message is extended with more info
--- meta/recipes-core/initrdscripts/initramfs-framework/finish | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
meta/recipes-core/initrdscripts/initramfs-framework/finish +1 -1
@@ -41,7 +41,7 @@ finish_run() { mount --move /sys $ROOTFS_DIR/sys cd $ROOTFS_DIR - exec switch_root -c /dev/console $ROOTFS_DIR ${bootparam_init:-/sbin/init} + exec switch_root $ROOTFS_DIR ${bootparam_init:-/sbin/init} else debug "No rootfs has been set" fi
Signature · 1 line
2.34.1
Quoted 9 lines · Christos Gavros
In 'finish'script the
Switch_rootcommand with option -c is busybox specific. When package-core-base-utils is selected the boot is breaking with message "Switch_root: invalid option — ‘c’.It was reproduced using qemu. Remove -c option to make it compatible with both providers.
It is tested using qemu with busybox and package-core-base-utils. In both cases qemu boots successfully and filesystem is accessible.
Fixes [YOCTO #15716]
Looks good
Quoted 29 lines
CC Yoann Congal <[email protected]>CC Randy MacLeod <[email protected]>CC jbk <[email protected]>CC Raj Khem <[email protected]>CC Adrian Freihofer <[email protected]>Signed-off-by Christos Gavros <[email protected]>---
v1->v2
* the commit message is extended with more info--- meta/recipes-core/initrdscripts/initramfs-framework/finish | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)meta/recipes-core/initrdscripts/initramfs-framework/finish +1
-2@@ -41,7 +41,7 @@ finish_run() { mount --move /sys $ROOTFS_DIR/sys cd $ROOTFS_DIR - exec switch_root -c /dev/console $ROOTFS_DIR ${bootparam_init:-/sbin/init} + exec switch_root $ROOTFS_DIR ${bootparam_init:-/sbin/init} else debug "No rootfs has been set" fi --2.34.1
Quoted 10 lines · Khem Raj
Quoted 8 lines · Christos Gavros
In 'finish'script the
Switch_rootcommand with option -c is busybox specific. When package-core-base-utils is selected the boot is breaking with message "Switch_root: invalid option — ‘c’.It was reproduced using qemu. Remove -c option to make it compatible with both providers.
It is tested using qemu with busybox and package-core-base-utils. In both cases qemu boots successfully and filesystem is accessible.
Fixes [YOCTO #15716]Looks good
Probably, and it's a small change, but :
1. Usually the busybox version has fewer features but here util-linux doesn't have the option, which is a bit odd.
I also checked the git repo/master branch and it hasn't been added.
Note that toybox, a BSD-Zero licensed, re-implementation of busybox *does* have the -c option.
I wonder if anyone has asked util-linux if they'd like that feature for embedded boards...
No sign of that as PR or issue:
https://github.com/util-linux/util-linux/pulls?q=is%3Apr++switch_root
https://github.com/util-linux/util-linux/issues?q=is%3Aissue%20switch_root
and
2. More importantly, embedded folks are often emotionally attached to there serial lines/ console for doing BSP development and low-level debug work.
CCing Kexin who will hopefully comment on whether he and the WR team would care if we drop the console specification.
If not, then I have no objection since one can easily enable the -c option when debugging and then ship without it.
../Randy
Quoted 30 lines
Quoted 30 lines
CC Yoann Congal<[email protected]>CC Randy MacLeod<[email protected]>CC jbk<[email protected]>CC Raj Khem<[email protected]>CC Adrian Freihofer<[email protected]>Signed-off-by Christos Gavros<[email protected]>---
v1->v2
* the commit message is extended with more info--- meta/recipes-core/initrdscripts/initramfs-framework/finish | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)meta/recipes-core/initrdscripts/initramfs-framework/finish +1
-2@@ -41,7 +41,7 @@ finish_run() { mount --move /sys $ROOTFS_DIR/sys cd $ROOTFS_DIR - exec switch_root -c /dev/console $ROOTFS_DIR ${bootparam_init:-/sbin/init} + exec switch_root $ROOTFS_DIR ${bootparam_init:-/sbin/init} else debug "No rootfs has been set" fi --2.34.1
Signature · 2 lines
# Randy MacLeod # Wind River Linux
Quoted 32 lines · Randy MacLeod
On 2025-04-07 1:04 p.m., Khem Raj wrote:
On Mon, Apr 7, 2025 at 9:45 AM Christos Gavros <[email protected]> wrote:
In 'finish'script the
Switch_rootcommand with option -c is busybox specific. When package-core-base-utils is selected the boot is breaking with message "Switch_root: invalid option — ‘c’.It was reproduced using qemu. Remove -c option to make it compatible with both providers.
It is tested using qemu with busybox and package-core-base-utils. In both cases qemu boots successfully and filesystem is accessible.
Fixes [YOCTO #15716]Looks good
Probably, and it's a small change, but :
1. Usually the busybox version has fewer features but here util-linux doesn't have the option, which is a bit odd.
I also checked the git repo/master branch and it hasn't been added. Note that toybox, a BSD-Zero licensed, re-implementation of busybox *does* have the -c option.
I wonder if anyone has asked util-linux if they'd like that feature for embedded boards...
No sign of that as PR or issue:
https://github.com/util-linux/util-linux/pulls?q=is%3Apr++switch_root
https://github.com/util-linux/util-linux/issues?q=is%3Aissue%20switch_root
Quoted 8 lines
and
2. More importantly, embedded folks are often emotionally attached to there serial lines/ console
for doing BSP development and low-level debug work.CCing Kexin who will hopefully comment on whether he and the WR team would care if we drop the console specification.
Appreciate the notification. This change appears safe for our use case, as our primary focus is kernel console functionality. No objections from my side.
Thanks,
Kevin
Quoted 46 lines
If not, then I have no objection since one can easily enable the -c option when debugging and then ship without it.
../Randy
CC Yoann Congal <[email protected]>CC Randy MacLeod <[email protected]>CC jbk <[email protected]>CC Raj Khem <[email protected]>CC Adrian Freihofer <[email protected]>Signed-off-by Christos Gavros <[email protected]>---
v1->v2
* the commit message is extended with more info
---
meta/recipes-core/initrdscripts/initramfs-framework/finish | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)diff --git a/meta/recipes-core/initrdscripts/initramfs-framework/finish b/meta/recipes-core/initrdscripts/initramfs-framework/finish index
ac0de9f996..716c2f44af100755
--- a/meta/recipes-core/initrdscripts/initramfs-framework/finish
+++ b/meta/recipes-core/initrdscripts/initramfs-framework/finish
@@ -41,7 +41,7 @@finish_run(){
mount --move /sys $ROOTFS_DIR/syscd $ROOTFS_DIR
- execswitch_root-c/dev/console$ROOTFS_DIR ${bootparam_init:-/sbin/init}
+ execswitch_root$ROOTFS_DIR ${bootparam_init:-/sbin/init}
else
debug "No rootfs has been set"
fi
--
2.34.1--
# Randy MacLeod
# Wind River Linux