cache: improve scalability by introducing bucket locks

An array of bucket locks is added.

All modifications still require the global cache_lock to be held for
writing. However, most readers only need the relevant bucket lock and in
effect can run concurrently to the writer as long as they use a
different lock. See the added comment for more details.

This is an intermediate step towards removal of the global lock.

Reviewed by:	kib
Tested by:	pho

1 files changed, 8 insertions, 0 deletions

diff --git a/sys/kern/subr_witness.c b/sys/kern/subr_witness.c
index d4359efbd227..ed1203578c2c 100644
--- a/sys/kern/subr_witness.c
+++ b/sys/kern/subr_witness.c
@@ -623,6 +623,14 @@ static struct witness_order_list_entry order_lists[] = {
 	{ "vnode interlock", &lock_class_mtx_sleep },
 	{ NULL, NULL },
 	/*
+	 * VFS namecache
+	 */
+	{ "ncglobal", &lock_class_rw },
+	{ "ncbuc", &lock_class_rw },
+	{ "vnode interlock", &lock_class_mtx_sleep },
+	{ "ncneg", &lock_class_mtx_sleep },
+	{ NULL, NULL },
+	/*
 	 * ZFS locking
 	 */
 	{ "dn->dn_mtx", &lock_class_sx },