Fix a panic on SMP systems, caused by sleeping while holding a

simple-lock.

The reviewer raises the following caveat: "I believe these changes
open a non-critical race condition when adding memory to the pool
for the zone. I think what will happen is that you could have two
threads that are simultaneously adding additional memory when the
pool runs out. This appears to not be a problem, however, since
the re-aquisition of the lock will protect the list pointers."
The submitter agrees that the race is non-critical, and points out
that it already existed for the non-SMP case.  He suggests that
perhaps a sleep lock (using the lock manager) should be used to
close that race.  This might be worth revisiting after 3.0 is
released.

Reviewed by:	dg (David Greenman)
Submitted by:	tegge (Tor Egge)

1 files changed, 13 insertions, 1 deletions

diff --git a/sys/vm/vm_zone.c b/sys/vm/vm_zone.c
index 478ed48..11a7ae0 100644
--- a/sys/vm/vm_zone.c
+++ b/sys/vm/vm_zone.c
@@ -11,7 +11,7 @@
  * 2. Absolutely no warranty of function or purpose is made by the author
  *	John S. Dyson.
  *
- * $Id: vm_zone.c,v 1.20 1998/04/15 17:47:40 bde Exp $
+ * $Id: vm_zone.c,v 1.21 1998/04/25 04:50:01 dyson Exp $
  */
 
 #include <sys/param.h>
@@ -326,11 +326,23 @@ _zget(vm_zone_t z)
 		if (lockstatus(&kernel_map->lock)) {
 			int s;
 			s = splvm();
+#ifdef SMP
+			simple_unlock(&z->zlock);
+#endif
 			item = (void *) kmem_malloc(kmem_map, nbytes, M_WAITOK);
+#ifdef SMP
+			simple_lock(&z->zlock);
+#endif
 			zone_kmem_pages += z->zalloc;
 			splx(s);
 		} else {
+#ifdef SMP
+			simple_unlock(&z->zlock);
+#endif
 			item = (void *) kmem_alloc(kernel_map, nbytes);
+#ifdef SMP
+			simple_lock(&z->zlock);
+#endif
 			zone_kern_pages += z->zalloc;
 		}
 		bzero(item, nbytes);