Re: klibc

[Joshua Hudson]

I've come to the conclusion that there is no good way to return to the
initramfs at all
after init moves to the real root device. What I have found is that the only way
is for another process to keep a cwd or open file handle on the initramfs which
plays very badly with killall.

Anybody got a way to make a user process other than init involunerable
to kill -9? <g>

It would be dirt-simple if I could mount --rbind / /root/initrd where
/ is the initramfs and /root is a mounted filesystem, but that creates
cycles and so breaks other things.

Oh, and mount / followed by ls / returns the contents of the initramfs. Weird.
umount -l / has the exteremely bizarre effect of leaving the process stranded in
/ unless it currently has pwd or open directory handle elsewhere.

Anybody want a patch that dumps the executor of umount / in the
initramfs, then does
a lazy unmount? That, however, carries risks of its own so might not
be the best move.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/