[pkg-cryptsetup-devel] Bug#659688: Bug#546610: cryptsetup: root= cmdline call is not honoured by initrd in crypto+lvm (Was initramfs-tools: ...)

[Agustin Martin]

On Thu, Mar 01, 2012 at 01:25:20AM +0100, Jonas Meurer wrote:
> 
> first rough tests seem to work properly, but I didn't find time to
> test with luks-encrypted rootfs on top of several physical lvm volumes
> // raid devices yet.
> 
> unfortunately I'm rather busy at the moment. will see whether I'm able
> to do the remaining tests within the next week.

Hi, Jonas and Roland,

Thanks for the info.

Take all the time you need, I am absolutely not in a hurry about this. 
I have prepared personal packages with the relevant changes and put them in
a personal repo, so I am using them in the meantime.

> but before actually uploading, I'd like to find a fix for the
> regression described in bug#659688. and I didn't find time to work on
> that one yet at all.

That is fine too.

Roland Mas wrote:
> I did a casual diff between 2:1.3.0-3.1 and 2:1.4.1-2 and found
> nothing obviously relevant to LVM partial stuff.  I suppose the change
> occurred in the way the device dependencies are calculated, around lines
> 150-160 of cryptroot-hook.

I had a quick look at #659688 and I have a similar feeling. I am no familiar
with cryptsetup besides what I looked for #546610, but as a blind guess to
confirm the above, I'd suggest start looking at "/conf/conf.d/cryptroot"
inside the relevant initrds.

Roland, if you keep somewhere a 2:1.4.1-2 non-working initrd for your old
setup, what are the contents of /conf/conf.d/cryptroot in that initrd? 
Same if you have a 2:1.3.0-3.1 initrd that worked for your old setup.

Just a guess, hope it helps,

Regards,

-- 
Agustin