Bug#338405: another patch
On Sunday 31 December 2006 20:33, maximilian attems <maks@sternwelten.at>
wrote:
> well the patch description is different to the patch itself.
> kinit is not used in initramfs-tools and is not removed in
> the patch below, what you remove are dash and minips from klibc-utils.
Sorry for the mis-description, I presume that the intent is clear though.
> no this patch has the same fundamental problems than the one
> that was merged for 0.85d.
It is different in that it's a second option, so it is possible to have
busybox without replacing all the non-busybox parts.
> * ugly: once busybox has a busybox --list, there would be no need
> to invoke tail, tr, sed and grep.
Good point, I filed a but report against busybox about that.
> * wrong: the busybox counterparts to the klibc-utils are not functinality
> equivalent. as written in the changelog documenting the revert for
> example sleep doesn't take subsecond args. quite sure that there is
> other breakage too as experienced by the sheer number of boot failures of
> 0.85d.
Such things of course will be less likely to be found if there is no easy
option of creating an initramfs that uses it. The last patch that I sent
will give no different result unless you set ONLY_BUSYBOX=y, so I can't
imagine it causing any problems for anyone.
The sleep bug is reported as below:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=341403
Having this code in mkinitramfs makes it more likely that the busybox bugs
will get fixed.
Reply to: