Description of problem: With an initrd created by recent versions of mkinitrd, boot from a "mkbootdisk --iso"-created boot CD fails on my system. Relevant bit of boot messages (verbose version using nash-test for more info): Switching to new root exec of init (/sbin/init) failed!!!: 14 Kernel panic: Attempted to kill init! Now unless I'm looking in the wrong header file, 14 is EFAULT, which from the execve man page means "filename points outside your accessible address space". I wish that made sense to me... Version-Release number of selected component (if applicable): I've had this problem from current mkinitrd (4.0.5) back at least as far as 4.0.3. 4.0.0 worked fine (at least in this regard). Not sure about 4.0.1 or 4.0.2; if you can point me to a place I can get those old versions, I'd be happy to retest with them. How reproducible: Using an affected version of mkinitrd, on my system, always. Unfortunately I have no idea which aspects of my system's configuration are causing the behavior. I always boot off CD, having no boot partition configured, so I don't know if the CD booting is actually a factor. Additional info: Here's what things look like once I'm successfully booted: $ cat /etc/fstab LABEL=/ / ext3 defaults 1 1 /dev/sda2 /media/temp ext3 defaults 1 2 none /dev/pts devpts gid=5,mode=620 0 0 none /proc proc defaults 0 0 none /dev/shm tmpfs defaults 0 0 /dev/sda1 swap swap defaults 0 0 /dev/fd0 /mnt/floppy auto noauto,owner,kudzu 0 0 /dev/hdc /mnt/cdrom auto user 0 0 none /var/lib/rpc_pipes rpc_pipefs defaults 0 0 $ mount /dev/md0 on / type ext3 (rw) none on /proc type proc (rw) none on /dev/pts type devpts (rw,gid=5,mode=620) usbdevfs on /proc/bus/usb type usbdevfs (rw) /dev/sda2 on /media/temp type ext3 (rw) none on /dev/shm type tmpfs (rw) /dev/hdc on /mnt/cdrom type iso9660 (ro,noexec,nosuid,nodev) sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw) none on /var/lib/rpc_pipes type rpc_pipefs (rw)
I've added some debugging printfs to nash's exec of init. New nash binary is at http://people.redhat.com/~katzj/nash-test-2. If you could grab that, copy it over /sbin/nash, and remake your initrd and let me know what output you get, that would be helpful.
Ok, with nash-test-2: initargs[1]: initrd=initrd.img initargs[2]: ro initargs[3]: psmouse.proto=bare initargs[4]: selinux=0 initargs[5]: vga=ask initargs[6]: root=/dev/md0 initargs[7]: BOOT_IMAGE=vmlinuz initargs[8]: single Then the same error 14. Parameter 8 came from me typing "linux single" at the isolinux boot prompt, and the rest appear to be consistent with the entry in my grub.conf: title Fedora Core (2.6.7-1.517) root (hd0,0) kernel /boot/vmlinuz-2.6.7-1.517 ro root=LABEL=/ psmouse.proto=bare selinux=0 vga=ask initrd /boot/initrd-2.6.7-1.517.img The only idea I get out of that is that I have an awful lot of boot parameters; are we overrunning a buffer used to pass them to init or something?
I also have the "exec of init failed!!!: 14" problem. I am trying to boot using mkinitrd-4.0.5-1, patched to boot an encrypted filesystem (see bug #124789). Things worked fine with the 3.5 series of mkinitrd (no initramfs). I can confirm that the encryped filesystem is properly mounted by initrd. I can also confirm that /sysroot/sbin/init exits at the time of the switchroot call. I modified nash.c to print debug messages similar to those above: initargs[0]: /sbin/init initargs[1]: root=/dev/hda5 initargs[2]: ro initargs[3]: enforcing=1 I am running Rawhide on an iBook.
This all seems to be fixed by the patch submitted for bug #129673. https://bugzilla.redhat.com/bugzilla/attachment.cgi?id=102734&action=view
I just tested the above patch, and it has fixed my problems too. Review it and roll it in! (Looking at what the patch fixes, all I can say is... Dude?!)
Should be happier with mkinitrd-4.0.6
Yes, mkinitrd-4.0.6 works fine; thanks for your work. Closing this bug now...