This is being discussed on lkml but Jens says it is an early user space problem not a kernel issue. Thread is: current linus bk, error mounting root I was loading the SATA driver, ata_piix as a module on initrd. This has been working fine for about a year. Something that went in after 2.6.11 has changed the timing. The nash script loads ata_piix and then does a mount. The device doesn't exist for the mount to work. Adding a 1 sec pause in the nash script before mkrootdev avoids the problem. There are comments that modprobe can now return before the device has actually been created so this may impact all modular disk drivers.
After more discussion on lkml it looks like this is a problem with mkinitrd and udev. The script does udevstart followed by raidautorun. In my case the drive nodes are not all there and some of the volumes don't get added to the raid group. mkinitrd needs to be modified to provide some synchronization with udev. The sleep works because it lets udev run. The sleep is not a reliable solution since it may take more than 1 sec for udev to run. The change after 2.6.11 has altered things so that the raidautorun task now runs before udev.
An update has been released for Fedora Core 3 (kernel-2.6.12-1.1372_FC3) which may contain a fix for your problem. Please update to this new kernel, and report whether or not it fixes your problem. If you have updated to Fedora Core 4 since this bug was opened, and the problem still occurs with the latest updates for that release, please change the version field of this bug to 'fc4'. Thank you.
This is a problem with nash, initrd and udev. A new kernel isn't going to fix it.
An mkinitrd update has been pushed. Update that, then remove and reinstall the kernel, and it should work. *** This bug has been marked as a duplicate of 163407 ***