Re: RFC: live-initramfs 2.x features

[Philippe Lelédy <debian@leledy.fr>]

Daniel Baumann wrote:
> Philippe Lelédy wrote:
>   
> > 1) It would be nice to specify that internal disks are not to be used at
> > all, even read/only:
> >     
>
> currently, if you don't enable swap nor persistency, internal disks
> arenot used, except for searching the live-media.
>   
And persistent partitions files or snapshots !
That's exactly what I would like to avoid.
> > 2.1) break=all would be useful to debug initramfs. Currently you can
> > only have *one* breakpoint.
> >     
>
> yes, major drawpack. i expect to get that solved by itself when we do
> the whole live script from scratch. but added to the list to keep paying
> attention to it.
>
>   
Even better having break=extented regular expression  ?
I use that simple patch :

maybe_break()
{
-    if [ "${break}" = "$1" ]; then
+    if [ "${break}" = "$1" ] || echo "$1" | egrep -q  "${break}" ; then

> > 2.3)  lzma compression
> >     
>
> see mailinglist archives, 
I was not speaking  about squashfs using lzma (known and big issue !) 
but simply -- a little out of topic -- about a mkinitramfs issue: 
produce a lzma-compressed initrd. (depends on linux kernel version)

Ph.L.