Date: Sat, 24 Feb 1996 16:43:01 -0500
From: "Charles M. Hannum" <mycroft@NetBSD.ORG>
Message-Id: <199602242143.QAA08828@pain.lcs.mit.edu>
To: ragge@NetBSD.ORG
Subject: Re: NetBSD master CVS tree commits
Cc: netbsd-developers@NetBSD.ORG
Sender: owner-netbsd-developers@NetBSD.ORG
Precedence: first-class


   Remove __FORK_BRAINDAMAGE, and the reason to why it was there.

You did *not* fix it.  __FORK_BRAINDAMAGE refers to copying the kernel
stack on fork.  Most of the ports don't do this any more; they create
a new kernel stack for the forked process, and arrange for it to exit
directly to user mode, without rewinding through all the fork()
machinery.  This is also why the return type is void; there's no longer
a need to distinguish in sys_fork() whether or not we're returning to
the parent or the child.

[CCed to -developers for the benefit of Jonathon and Paul...]