Compromise for critical*()/cpu_critical*() recommit. Cleanup the interrupt

disablement assumptions in kern_fork.c by adding another API call,
cpu_critical_fork_exit().  Cleanup the td_savecrit field by moving it
from MI to MD.  Temporarily move cpu_critical*() from <arch>/include/cpufunc.h
to <arch>/<arch>/critical.c (stage-2 will clean this up).

Implement interrupt deferral for i386 that allows interrupts to remain
enabled inside critical sections.  This also fixes an IPI interlock bug,
and requires uses of icu_lock to be enclosed in a true interrupt disablement.

This is the stage-1 commit.  Stage-2 will occur after stage-1 has stabilized,
and will move cpu_critical*() into its own header file(s) + other things.
This commit may break non-i386 architectures in trivial ways.  This should
be temporary.

Reviewed by:	core
Approved by:	core

1 files changed, 7 insertions, 3 deletions

diff --git a/sys/kern/kern_fork.c b/sys/kern/kern_fork.c
index d508cbf..723fd62 100644
--- a/sys/kern/kern_fork.c
+++ b/sys/kern/kern_fork.c
@@ -774,12 +774,16 @@ fork_exit(callout, arg, frame)
 
 	td->td_kse->ke_oncpu = PCPU_GET(cpuid);
 	/*
-	 * Setup the sched_lock state so that we can release it.
+	 * Finish setting up thread glue.  We need to initialize
+	 * the thread into a td_critnest=1 state.  Some platforms
+	 * may have already partially or fully initialized td_critnest
+	 * and/or td_md.md_savecrit (when applciable).
+	 *
+	 * see <arch>/<arch>/critical.c
 	 */
 	sched_lock.mtx_lock = (uintptr_t)td;
 	sched_lock.mtx_recurse = 0;
-	td->td_critnest = 1;
-	td->td_savecrit = CRITICAL_FORK;
+	cpu_critical_fork_exit();
 	CTR3(KTR_PROC, "fork_exit: new proc %p (pid %d, %s)", p, p->p_pid,
 	    p->p_comm);
 	if (PCPU_GET(switchtime.sec) == 0)