maximilian attems wrote: > On Tue, Apr 27, 2010 at 02:47:38PM +1000, Trent W. Buck wrote: > > Package: initramfs-tools > > Version: 0.94.4 > > Severity: wishlist > > > > Following the kernel handbook, I built a mainline kernel using "make > > deb-pkg". > > ok, fine so far. > > >This kernel is configured without modules or initrd support. > > pretty sure that this is not a recommended way by referenced book.
Correct; I didn't mean to imply that. I'm just killing time playing with kernel options, which is why... >> If the problem is in the kernel -- if "make deb-pkg" always invokes >> update-initramfs, but shouldn't -- let me know and I'll take this >> issue upstream. > > if your box does not need initramfs-tools why would it be there. > once you remove it, it can't be run. ...initramfs-tools is installed on this host -- it's needed for the normal (non-play) Debian kernels. > good luck in going down that road without UUID support, nor > md support. Granted. A show-stopper for the general case, but not for the specific toy that I'm playing with right now. > > 2) gave a clearer error message, such as: > > > > update-initramfs: Generating /boot/initrd.img-2.6.34-rc5 > > WARNING: CONFIG_MODULES=n in 2.6.34-rc5, a ramdisk is likely useless. > > FATAL: CONFIG_BLK_DEV_INITRD=n in 2.6.34-rc5, a ramdisk is useless. > > > > In particular, this current warning is unintelligible: > > > > Device driver support needs thus be built-in linux image! > > lol indeed this was a funny minute when I put that in. > the meaning is that you'd better really have everything built in. > for eventual MODULES config settings we have now way to know. I was thinking you could look at /boot/config-<revision>. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
