Always explicitly panic in propogate_priority() if we try to propogate

a lock's priority to a sleeping thread.  When we panic, dump a stack
trace of the thread that is asleep if DDB is compiled into the kernel
just before calling panic().  This is much more informative and useful
for debugging than the current behavior of getting a page fault and not
having an easy way of determining which thread caused the original problem.

MFC after:	1 week

1 files changed, 14 insertions, 8 deletions

diff --git a/sys/kern/subr_turnstile.c b/sys/kern/subr_turnstile.c
index 2b0fc2e..0185d4a 100644
--- a/sys/kern/subr_turnstile.c
+++ b/sys/kern/subr_turnstile.c
@@ -186,15 +186,21 @@ propagate_priority(struct thread *td)
 		MPASS(td->td_proc->p_magic == P_MAGIC);
 
 		/*
-		 * XXX: The owner of a turnstile can be stale if it is the
-		 * first thread to grab a rlock of a rw lock.  In that case
-		 * it is possible for us to be at SSLEEP or some other
-		 * weird state.  We should probably just return if the state
-		 * isn't SRUN or SLOCK.
+		 * If the thread is asleep, then we are probably about
+		 * to deadlock.  To make debugging this easier, just
+		 * panic and tell the user which thread misbehaved so
+		 * they can hopefully get a stack trace from the truly
+		 * misbehaving thread.
 		 */
-		KASSERT(!TD_IS_SLEEPING(td),
-		    ("sleeping thread (tid %d) owns a non-sleepable lock",
-		    td->td_tid));
+		if (TD_IS_SLEEPING(td)) {
+			printf(
+		"Sleeping thread (tid %d, pid %d) owns a non-sleepable lock\n",
+			    td->td_tid, td->td_proc->p_pid);
+#ifdef DDB
+			db_trace_thread(td, -1);
+#endif
+			panic("sleeping thread");
+		}
 
 		/*
 		 * If this thread already has higher priority than the