Module Name: src
Committed By: ad
Date: Fri Mar 27 00:14:25 UTC 2020
Modified Files:
src/sys/kern: vfs_cache.c
Log Message:
- Stats: add a cast to make sure 32-bit wraparound works properly.
- cache_purge_children(): just use RB_TREE_MIN().
- Change the name of the pool to "namecache".
- Update comments.
To generate a diff of this commit:
cvs rdiff -u -r1.134 -r1.135 src/sys/kern/vfs_cache.c
Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.
Modified files:
Index: src/sys/kern/vfs_cache.c
diff -u src/sys/kern/vfs_cache.c:1.134 src/sys/kern/vfs_cache.c:1.135
--- src/sys/kern/vfs_cache.c:1.134 Mon Mar 23 20:02:13 2020
+++ src/sys/kern/vfs_cache.c Fri Mar 27 00:14:25 2020
@@ -1,4 +1,4 @@
-/* $NetBSD: vfs_cache.c,v 1.134 2020/03/23 20:02:13 ad Exp $ */
+/* $NetBSD: vfs_cache.c,v 1.135 2020/03/27 00:14:25 ad Exp $ */
/*-
* Copyright (c) 2008, 2019, 2020 The NetBSD Foundation, Inc.
@@ -172,7 +172,7 @@
*/
#include <sys/cdefs.h>
-__KERNEL_RCSID(0, "$NetBSD: vfs_cache.c,v 1.134 2020/03/23 20:02:13 ad Exp $");
+__KERNEL_RCSID(0, "$NetBSD: vfs_cache.c,v 1.135 2020/03/27 00:14:25 ad Exp $");
#define __NAMECACHE_PRIVATE
#ifdef _KERNEL_OPT
@@ -250,7 +250,7 @@ static kmutex_t cache_stat_lock __cachel
#define UPDATE(nchcpu, f) do { \
uint32_t cur = atomic_load_relaxed(&nchcpu->cur.f); \
- nchstats.f += cur - nchcpu->last.f; \
+ nchstats.f += (uint32_t)(cur - nchcpu->last.f); \
nchcpu->last.f = cur; \
} while (/* CONSTCOND */ 0)
@@ -409,9 +409,9 @@ cache_lookup_entry(struct vnode *dvp, co
* tailored for exactly what's needed here (64-bit key and so on)
* that is quite a bit faster than using rb_tree_find_node().
*
- * In the fast path memcmp() needs to be called at least once to
- * confirm that the correct name has been found. If there has been
- * a hash value collision (very rare) the search will continue on.
+ * For a matching key memcmp() needs to be called once to confirm
+ * that the correct name has been found. Very rarely there will be
+ * a key value collision and the search will continue.
*/
for (;;) {
if (__predict_false(RB_SENTINEL_P(node))) {
@@ -791,7 +791,10 @@ cache_revlookup(struct vnode *vp, struct
}
}
- /* Record a hit on the entry. This is an unlocked read. */
+ /*
+ * Record a hit on the entry. This is an unlocked read but
+ * even if wrong it doesn't matter too much.
+ */
lrulist = atomic_load_relaxed(&ncp->nc_lrulist);
if (lrulist != LRU_ACTIVE) {
cache_activate(ncp);
@@ -905,8 +908,8 @@ cache_enter(struct vnode *dvp, struct vn
/*
* With the directory lock still held, insert to the tail of the
- * ACTIVE LRU list (new) and with the LRU lock held take the to
- * opportunity to incrementally balance the lists.
+ * ACTIVE LRU list (new) and take the opportunity to incrementally
+ * balance the lists.
*/
mutex_enter(&cache_lru_lock);
ncp->nc_lrulist = LRU_ACTIVE;
@@ -916,8 +919,10 @@ cache_enter(struct vnode *dvp, struct vn
mutex_exit(&cache_lru_lock);
/*
- * Finally, insert to the vnode, and unlock. Partially sort the
- * per-vnode list: dots go to back.
+ * Finally, insert to the vnode and unlock. With everything set up
+ * it's safe to let cache_revlookup() see the entry. Partially sort
+ * the per-vnode list: dots go to back so cache_revlookup() doesn't
+ * have to consider them.
*/
if (vp != NULL) {
vnode_impl_t *vi = VNODE_TO_VIMPL(vp);
@@ -987,7 +992,7 @@ nchinit(void)
{
cache_pool = pool_cache_init(sizeof(struct namecache),
- coherency_unit, 0, 0, "nchentry", NULL, IPL_NONE, NULL,
+ coherency_unit, 0, 0, "namecache", NULL, IPL_NONE, NULL,
NULL, NULL);
KASSERT(cache_pool != NULL);
@@ -1126,11 +1131,7 @@ cache_purge_children(struct vnode *dvp)
SDT_PROBE(vfs, namecache, purge, children, dvp, 0, 0, 0, 0);
rw_enter(&dvi->vi_nc_lock, RW_WRITER);
- for (;;) {
- ncp = rb_tree_iterate(&dvi->vi_nc_tree, NULL, RB_DIR_RIGHT);
- if (ncp == NULL) {
- break;
- }
+ while ((ncp = RB_TREE_MIN(&dvi->vi_nc_tree)) != NULL) {
cache_remove(ncp, true);
}
rw_exit(&dvi->vi_nc_lock);
@@ -1210,14 +1211,14 @@ cache_purgevfs(struct mount *mp)
}
/*
- * Re-queue an entry onto the correct LRU list, after it has scored a hit.
+ * Re-queue an entry onto the tail of the active LRU list, after it has
+ * scored a hit.
*/
static void
cache_activate(struct namecache *ncp)
{
mutex_enter(&cache_lru_lock);
- /* Put on tail of ACTIVE list, since it just scored a hit. */
TAILQ_REMOVE(&cache_lru.list[ncp->nc_lrulist], ncp, nc_lru);
TAILQ_INSERT_TAIL(&cache_lru.list[LRU_ACTIVE], ncp, nc_lru);
cache_lru.count[ncp->nc_lrulist]--;
