[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: In preparation for 2.6.13 - initrd issues



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Sun, 11 Sep 2005 18:23:25 -0700 (PDT)
Jurij Smakov <jurij@wooyd.org> wrote:

> Hello,
> 
> As you probably know, the 2.6.13 kernel is out, and we are facing some 
> problems with packaging it for Debian. A major change compared to 2.6.12 
> is the discontinued support for devfs, which, I understand, renders 
> current initrd-tools unusable. As I see it, there are two major problems 
> which need to be taken care of:
> 
> 1. Choice of the alternative tool for initrd generation and management.
> The alternatives I've heard about so far are yaird and initramfs (it 
> appears that initramfs is not only the name of the new initrd filesystem 
> but also the tool which is used to generate it). Yaird is heavily 
> advocated by Sven Luther, but it would be nice to hear all pros and cons 
> of both packages. It appears that both of them generate initrds in the new 
> (initramfs) format which is essentially a cpio archive of all files. One 
> difference mentioned is that yaird may be configured to do the 
> install-time detection of required modules, and included only them into 
> the initrd (similarly to initrd-tools). It is also claimed that the 
> yaird code is easier to maintain. On the other hand, initramfs can only 
> build an initrd using a fixed set of modules, so we'll need to include all 
> the modules on initrd if we choose to go that way (is it actually true?). 
> That clearly may have undesired consequences for low-memory installs.

A major drawback of yaird is lack of configuration. this has been added in latest release, and I have now spend all night packaging it, so should be available by tomorrow night in sid. Until then it can be grabbed at http://debian.jones.dk/auryn/pool/official/yaird/ (for powerpc, so recompile if using anything else).


> 2. What changes need to be done to integrate a new initrd-generating tool 
> into the kernel packaging infrastracture. It might be as simple as 
> switching the postinst of kernel packages from running mkinitrd to running 
> yaird. I have no idea what issues may arise in d-i because of that.

yaird has a wrapper that simulates mkinitrd. Currently it is packaged as mkinitrd.yaird and must then manually be installed in /etc/kernel-img.conf as "ramdisk = /usr/sbin/mkinitrd.yaird", but if needed I can repackage with the wrapper as mkinitrd and the package conflicting mkinitrd.


 - Jonas

- -- 
* Jonas Smedegaard - idealist og Internet-arkitekt
* Tlf.: +45 40843136  Website: http://dr.jones.dk/

 - Enden er nær: http://www.shibumi.org/eoti.htm
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQFDJQ3Tn7DbMsAkQLgRAugRAKCA+20zqLFqo9uuIQxyJMB/9iIPRwCcDrdu
agvTRKUZ1H2WaIfZjvm/RF4=
=pWzb
-----END PGP SIGNATURE-----



Reply to: