Bug#646284: dropping applets-fallback breaks initramfs images
On 23.10.2011 01:49, Stefan Lippers-Hollmann wrote:
> Package: busybox
> Version: 1:1.19.2-1
> Severity: grave
> Justification: Breaks system booting using initramfs-tools in non-trivial ways.
> Tags: patch
> X-Debbugs-CC: Debian kernel team <email@example.com>
> Initramfs images generated by initramfs-tools 0.99 after busybox got
> upgraded to 1:1.19.2-1 fail to boot with the following error messages:
> Loading, please wait...
> /init: line 11: mount: not found
Yes, applets-fallback was a dirty hack, and I tried to get rid of it,
and I asked kernel team several times about using regular ways to invoke
busybox commands in initramfs, but only got "Later" so far. I tempted
to reassign this bug to initramfs-tools, but wont do that yet.
And no, it wasn't really expected - I hoped that CONFIG_FEATURE_PREFER_APPLETS
will take care of this. Apparently not, and I wonder why it is so.....
.. ah, I see now: it is because initramfs does not put busybox as /bin/busybox
as it expects to be found, but copies it to /bin/sh. So, with previously
applied applets-fallback, it tried to execute /proc/self/exe AND /bin/busybox,
and without that patch, it tries just /bin/busybox. Sigh.
> Re-instating applets-fallback.patch in busybox 1:1.19.2-1 however fixes
> this reproducable boot failure, quick'n'dirty rediff (tested on
> amd64 && i386, using systems with mdadm, lvm2 or nothing special at
> all) attached.
Thank you for the quick fix - something I tried to do yesterday but failed.
It is my intention to get rid of this ugly hack before wheezy, but unfortunately
kernel team does not want to cooperate. So for now, I'll include this hack
again, and file a bug against initramfs-tools properly.