On Tue, Feb 2, 2010 at 1:13 PM, Thomas Bächler <thomas@archlinux.org> wrote:
Am 24.01.2010 18:05, schrieb Thomas Bächler:
Quoted from http://bugs.archlinux.org/task/17298 (which is also to be used for feedback, besides the threads on these mailing lists):
So, now something can be tested. I recommend using it in conjunction with testing/udev only:
[kill-klibc] Server = http://dev.archlinux.org/~thomas/kill-klibc/i686/
or
[kill-klibc] Server = http://dev.archlinux.org/~thomas/kill-klibc/x86_64/
The standard hooks work, so do keymap, lvm2, encrypt. What does NOT work is raid/mdadm/dmraid and root on NFS. I would appreciate any input and patches w.r.t. those, as well as any testing. The v86d package also needs to be fixed, but that is trivial.
device-mapper, lvm2 and cryptsetup have a weird "1.local" pkgrel so they won't collide with the ones from the repo, sorry if that bothers anyone.
If you upgrade to this repositories, your existing mkinitcpio will be replaced and all klibc packages will be gone too. So, if you try it, please save your existing initramfs to a different filename as a fallback, and do not perform kernel upgrades until you either confirmed that it works, or downgraded to the klibc-based mkinitcpio.
Bump! Virtually nobody has performed any testing (except the people I poked about it). Also, I still don't have fixed raid or dmraid hooks, nor any suggestion for implementing root on NFS (of the last three, I use neither). I could write raid, but I have no idea if it will work. I don't know what exactly needs to be done in the NFS case, because that was all hidden inside "ipconfig" from klibc.
The part that works (standard hooks, keymap, cryptsetup, lvm2) has been 100% stable and has worked perfectly for all of the testers so far (which aren't many, so you never know).
Just tested on my i686 system with lvm2. It booted fine and everything looks OK, in fact I haven't noted any difference. Hooks used in /etc/mkinitcpio.conf: HOOKS="base udev autodetect pata usbinput keymap lvm2 filesystems" I would be willing to test raid support when it will be added. I'm not really familiar with kernels and early userspace stuff so I can't be much help in making a patch for it.
Note that the klibc-based mkinitcpio has problems which will never be fixed (because I can't fix them, or don't want to fix them), some of which are already fixed in the 0.6 branch.
One example is the lack of vfat detection in fstype, which affects people who want to put archiso on vfat volumes - this is no problem because we use blkid now. I am sure there are more examples. I simply refuse to put more time into something that is entirely broken and can't even be rebuilt against our current kernel headers (yes, the last time klibc was successfully rebuilt was against 2.6.29 iirc, the version I built against 2.6.31 works on x86_64, but not on i686).