On 01/08/06, Rodrigo Coacci <[EMAIL PROTECTED]> wrote:
Well, here they are. I guess they should work correctly both with patched and unpatched kernels.
At least it surely worked with my unpatched kernel. Again, it only does boot splash and progress, not console background image (as fbsplash is needed for that).
Indeed you were using the daemon, my mistake ;-)
The trick was to find if kernel had /dev/fbsplash and only if it had does the "splash_util -c on" command, as this sets console background image.
It starts as soon as UDev is done (as per rc.sysinit script)
I guess that to put it to initrd/initcpio we should need also the splash_utils and maybe also the splash script itself... When I said earlier that it should be tricky to put it to work in initrams was because of the FIFO part; I don't really know if we can get that fifo out from ramdisk to disk on-the-fly, or if you do that already some way, because checking out mkinitcpio install scripts, it doesn't copy to ramdisk the splash_util, nor the splash script itself, so we get bootsplash since initram only if we have fbsplash in kernel.
And please, I'm not very good at bash scripts, so if find my solution is not very elegant/clean/portable/etc feel free to change it, I'll be glad if that helps you get with a better solution :-P
And sorry for my bad english....
Fantastic! It's going to need some testing and I'm sure it's not quite right just yet but you have done the hard part, which was figuring out the difference between doing it with and without patches.
I think I now understand what you mean about the initram but we'll have to have phrak/iph/tpowa look at support for that in mkinitcpio.
I'm so impressed that you actually _did_ this and did it so quickly! Great work and thanks!
_______________________________________________ arch mailing list arch@archlinux.org http://www.archlinux.org/mailman/listinfo/arch
