1/*-
2 * SPDX-License-Identifier: BSD-2-Clause
3 *
4 * Copyright (c) 2005 Poul-Henning Kamp
5 * All rights reserved.
6 *
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
9 * are met:
10 * 1. Redistributions of source code must retain the above copyright
11 *    notice, this list of conditions and the following disclaimer.
12 * 2. Redistributions in binary form must reproduce the above copyright
13 *    notice, this list of conditions and the following disclaimer in the
14 *    documentation and/or other materials provided with the distribution.
15 *
16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
17 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
20 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26 * SUCH DAMAGE.
27 *
28 */
29
30#include <sys/param.h>
31#include <sys/systm.h>
32#include <sys/kernel.h>
33#include <sys/malloc.h>
34#include <sys/mount.h>
35#include <sys/rwlock.h>
36#include <sys/vnode.h>
37
38static MALLOC_DEFINE(M_VFS_HASH, "vfs_hash", "VFS hash table");
39
40static LIST_HEAD(vfs_hash_head, vnode)	*vfs_hash_tbl;
41static LIST_HEAD(,vnode)		vfs_hash_side;
42static u_long				vfs_hash_mask;
43static struct rwlock __exclusive_cache_line vfs_hash_lock;
44
45static void
46vfs_hashinit(void *dummy __unused)
47{
48
49	vfs_hash_tbl = hashinit(desiredvnodes, M_VFS_HASH, &vfs_hash_mask);
50	rw_init(&vfs_hash_lock, "vfs hash");
51	LIST_INIT(&vfs_hash_side);
52}
53
54/* Must be SI_ORDER_SECOND so desiredvnodes is available */
55SYSINIT(vfs_hash, SI_SUB_VFS, SI_ORDER_SECOND, vfs_hashinit, NULL);
56
57u_int
58vfs_hash_index(struct vnode *vp)
59{
60
61	return (vp->v_hash + vp->v_mount->mnt_hashseed);
62}
63
64static struct vfs_hash_head *
65vfs_hash_bucket(const struct mount *mp, u_int hash)
66{
67
68	return (&vfs_hash_tbl[(hash + mp->mnt_hashseed) & vfs_hash_mask]);
69}
70
71int
72vfs_hash_get(const struct mount *mp, u_int hash, int flags, struct thread *td,
73    struct vnode **vpp, vfs_hash_cmp_t *fn, void *arg)
74{
75	struct vnode *vp;
76	enum vgetstate vs;
77	int error;
78
79	while (1) {
80		rw_rlock(&vfs_hash_lock);
81		LIST_FOREACH(vp, vfs_hash_bucket(mp, hash), v_hashlist) {
82			if (vp->v_hash != hash)
83				continue;
84			if (vp->v_mount != mp)
85				continue;
86			if (fn != NULL && fn(vp, arg))
87				continue;
88			vs = vget_prep(vp);
89			rw_runlock(&vfs_hash_lock);
90			error = vget_finish(vp, flags, vs);
91			if (error == ENOENT && (flags & LK_NOWAIT) == 0)
92				break;
93			if (error != 0)
94				return (error);
95			if (vp->v_hash != hash ||
96			    (fn != NULL && fn(vp, arg))) {
97				vput(vp);
98				/* Restart the bucket walk. */
99				break;
100			}
101			*vpp = vp;
102			return (0);
103		}
104		if (vp == NULL) {
105			rw_runlock(&vfs_hash_lock);
106			*vpp = NULL;
107			return (0);
108		}
109	}
110}
111
112void
113vfs_hash_ref(const struct mount *mp, u_int hash, struct thread *td,
114    struct vnode **vpp, vfs_hash_cmp_t *fn, void *arg)
115{
116	struct vnode *vp;
117
118	while (1) {
119		rw_rlock(&vfs_hash_lock);
120		LIST_FOREACH(vp, vfs_hash_bucket(mp, hash), v_hashlist) {
121			if (vp->v_hash != hash)
122				continue;
123			if (vp->v_mount != mp)
124				continue;
125			if (fn != NULL && fn(vp, arg))
126				continue;
127			vhold(vp);
128			rw_runlock(&vfs_hash_lock);
129			vref(vp);
130			vdrop(vp);
131			*vpp = vp;
132			return;
133		}
134		if (vp == NULL) {
135			rw_runlock(&vfs_hash_lock);
136			*vpp = NULL;
137			return;
138		}
139	}
140}
141
142void
143vfs_hash_remove(struct vnode *vp)
144{
145
146	rw_wlock(&vfs_hash_lock);
147	LIST_REMOVE(vp, v_hashlist);
148	rw_wunlock(&vfs_hash_lock);
149}
150
151int
152vfs_hash_insert(struct vnode *vp, u_int hash, int flags, struct thread *td,
153    struct vnode **vpp, vfs_hash_cmp_t *fn, void *arg)
154{
155	struct vnode *vp2;
156	enum vgetstate vs;
157	int error;
158
159	*vpp = NULL;
160	while (1) {
161		rw_wlock(&vfs_hash_lock);
162		LIST_FOREACH(vp2,
163		    vfs_hash_bucket(vp->v_mount, hash), v_hashlist) {
164			if (vp2->v_hash != hash)
165				continue;
166			if (vp2->v_mount != vp->v_mount)
167				continue;
168			if (fn != NULL && fn(vp2, arg))
169				continue;
170			vs = vget_prep(vp2);
171			rw_wunlock(&vfs_hash_lock);
172			error = vget_finish(vp2, flags, vs);
173			if (error == ENOENT && (flags & LK_NOWAIT) == 0)
174				break;
175			rw_wlock(&vfs_hash_lock);
176			LIST_INSERT_HEAD(&vfs_hash_side, vp, v_hashlist);
177			rw_wunlock(&vfs_hash_lock);
178			vgone(vp);
179			vput(vp);
180			if (!error)
181				*vpp = vp2;
182			return (error);
183		}
184		if (vp2 == NULL)
185			break;
186	}
187	vp->v_hash = hash;
188	LIST_INSERT_HEAD(vfs_hash_bucket(vp->v_mount, hash), vp, v_hashlist);
189	rw_wunlock(&vfs_hash_lock);
190	return (0);
191}
192
193void
194vfs_hash_rehash(struct vnode *vp, u_int hash)
195{
196	ASSERT_VOP_ELOCKED(vp, "rehash requires excl lock");
197
198	rw_wlock(&vfs_hash_lock);
199	LIST_REMOVE(vp, v_hashlist);
200	LIST_INSERT_HEAD(vfs_hash_bucket(vp->v_mount, hash), vp, v_hashlist);
201	vp->v_hash = hash;
202	rw_wunlock(&vfs_hash_lock);
203}
204
205void
206vfs_hash_changesize(u_long newmaxvnodes)
207{
208	struct vfs_hash_head *vfs_hash_newtbl, *vfs_hash_oldtbl;
209	u_long vfs_hash_newmask, vfs_hash_oldmask;
210	struct vnode *vp;
211	int i;
212
213	vfs_hash_newtbl = hashinit(newmaxvnodes, M_VFS_HASH,
214		&vfs_hash_newmask);
215	/* If same hash table size, nothing to do */
216	if (vfs_hash_mask == vfs_hash_newmask) {
217		free(vfs_hash_newtbl, M_VFS_HASH);
218		return;
219	}
220	/*
221	 * Move everything from the old hash table to the new table.
222	 * None of the vnodes in the table can be recycled because to
223	 * do so, they have to be removed from the hash table.
224	 */
225	rw_wlock(&vfs_hash_lock);
226	vfs_hash_oldtbl = vfs_hash_tbl;
227	vfs_hash_oldmask = vfs_hash_mask;
228	vfs_hash_tbl = vfs_hash_newtbl;
229	vfs_hash_mask = vfs_hash_newmask;
230	for (i = 0; i <= vfs_hash_oldmask; i++) {
231		while ((vp = LIST_FIRST(&vfs_hash_oldtbl[i])) != NULL) {
232			LIST_REMOVE(vp, v_hashlist);
233			LIST_INSERT_HEAD(
234			    vfs_hash_bucket(vp->v_mount, vp->v_hash),
235			    vp, v_hashlist);
236		}
237	}
238	rw_wunlock(&vfs_hash_lock);
239	free(vfs_hash_oldtbl, M_VFS_HASH);
240}
241