Subject: Re: RFC: migration to a fully dynamically linked system
To: Bill Studenmund <email@example.com>
From: Todd Vierling <firstname.lastname@example.org>
Date: 12/29/2001 17:54:17
On Sat, 29 Dec 2001, Bill Studenmund wrote:
: > - If you reboot to single user mode, you can specify
: > /recovery/sh as the shell. If /sbin/init was dynamically
: > linked (see above), the kernel could just try
: > /recovery/init as an alternative (see above).
: The kernel trying /recovery/init isn't 100% as I understand things. That
: list of inits the kernel looks for are ones the kernel tries to start. But
: once one of them starts, it is init.
: And if it exits, the kernel panics.
: We could have the kernel try /recovery/init at that point, but how does
: the kernel know if we're in a, "libc is messed up, try recovery" mode, or
: a, "we're doomed, panic until someone comes to help" mode?
As a point of information, do realize that ld.elf_so is loaded and set up by
the kernel (before the exec call succeeds). Of course, libc.so is loaded by
ld.elf_so, so that can be a failure point.
I'm not entirely sure there is a sane way to detect that init has
*successfully* started, short of proc0 restarting init if it exits with the
exitcode != 0.
Perhaps the question should be rephrased as, "do we *need* the ability for
init to have external auth methods, and so forth?" If not, then init can
stay static without pain.
-- Todd Vierling <email@example.com> * Wasabi & NetBSD: Run with it.
-- CDs, Integration, Embedding, Support -- http://www.wasabisystems.com/