1 /* $NetBSD: tmpfs_vfsops.c,v 1.10 2005/12/11 12:24:29 christos Exp $ */
4 * SPDX-License-Identifier: BSD-2-Clause-NetBSD
6 * Copyright (c) 2005 The NetBSD Foundation, Inc.
9 * This code is derived from software contributed to The NetBSD Foundation
10 * by Julio M. Merino Vidal, developed as part of Google's Summer of Code
13 * Redistribution and use in source and binary forms, with or without
14 * modification, are permitted provided that the following conditions
16 * 1. Redistributions of source code must retain the above copyright
17 * notice, this list of conditions and the following disclaimer.
18 * 2. Redistributions in binary form must reproduce the above copyright
19 * notice, this list of conditions and the following disclaimer in the
20 * documentation and/or other materials provided with the distribution.
22 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
23 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
24 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
25 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
26 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
27 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
28 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
29 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
30 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
31 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
32 * POSSIBILITY OF SUCH DAMAGE.
36 * Efficient memory file system.
38 * tmpfs is a file system that uses FreeBSD's virtual memory
39 * sub-system to store file data and metadata in an efficient way.
40 * This means that it does not follow the structure of an on-disk file
41 * system because it simply does not need to. Instead, it uses
42 * memory-specific data structures and algorithms to automatically
43 * allocate and release resources.
45 #include <sys/cdefs.h>
46 __FBSDID("$FreeBSD$");
48 #include <sys/param.h>
49 #include <sys/dirent.h>
50 #include <sys/limits.h>
52 #include <sys/mount.h>
53 #include <sys/mutex.h>
56 #include <sys/kernel.h>
57 #include <sys/rwlock.h>
59 #include <sys/systm.h>
60 #include <sys/sysctl.h>
61 #include <sys/vnode.h>
64 #include <vm/vm_object.h>
65 #include <vm/vm_param.h>
67 #include <fs/tmpfs/tmpfs.h>
70 * Default permission for root node
72 #define TMPFS_DEFAULT_ROOT_MODE (S_IRWXU|S_IRGRP|S_IXGRP|S_IROTH|S_IXOTH)
74 MALLOC_DEFINE(M_TMPFSMNT, "tmpfs mount", "tmpfs mount structures");
75 MALLOC_DEFINE(M_TMPFSNAME, "tmpfs name", "tmpfs file names");
77 static int tmpfs_mount(struct mount *);
78 static int tmpfs_unmount(struct mount *, int);
79 static int tmpfs_root(struct mount *, int flags, struct vnode **);
80 static int tmpfs_fhtovp(struct mount *, struct fid *, int,
82 static int tmpfs_statfs(struct mount *, struct statfs *);
83 static void tmpfs_susp_clean(struct mount *);
85 static const char *tmpfs_opts[] = {
86 "from", "size", "maxfilesize", "inodes", "uid", "gid", "mode", "export",
90 static const char *tmpfs_updateopts[] = {
91 "from", "export", NULL
95 tmpfs_node_ctor(void *mem, int size, void *arg, int flags)
97 struct tmpfs_node *node = (struct tmpfs_node *)mem;
104 node->tn_vnode = NULL;
105 node->tn_vpstate = 0;
111 tmpfs_node_dtor(void *mem, int size, void *arg)
113 struct tmpfs_node *node = (struct tmpfs_node *)mem;
114 node->tn_type = VNON;
118 tmpfs_node_init(void *mem, int size, int flags)
120 struct tmpfs_node *node = (struct tmpfs_node *)mem;
123 mtx_init(&node->tn_interlock, "tmpfs node interlock", NULL, MTX_DEF);
124 node->tn_gen = arc4random();
130 tmpfs_node_fini(void *mem, int size)
132 struct tmpfs_node *node = (struct tmpfs_node *)mem;
134 mtx_destroy(&node->tn_interlock);
138 tmpfs_mount(struct mount *mp)
140 const size_t nodes_per_page = howmany(PAGE_SIZE,
141 sizeof(struct tmpfs_dirent) + sizeof(struct tmpfs_node));
142 struct tmpfs_mount *tmp;
143 struct tmpfs_node *root;
148 off_t nodes_max, size_max, maxfilesize;
150 /* Root node attributes. */
157 if (vfs_filteropt(mp->mnt_optnew, tmpfs_opts))
160 if (mp->mnt_flag & MNT_UPDATE) {
161 /* Only support update mounts for certain options. */
162 if (vfs_filteropt(mp->mnt_optnew, tmpfs_updateopts) != 0)
164 if (vfs_flagopt(mp->mnt_optnew, "ro", NULL, 0) !=
165 ((struct tmpfs_mount *)mp->mnt_data)->tm_ronly)
170 vn_lock(mp->mnt_vnodecovered, LK_SHARED | LK_RETRY);
171 error = VOP_GETATTR(mp->mnt_vnodecovered, &va, mp->mnt_cred);
172 VOP_UNLOCK(mp->mnt_vnodecovered, 0);
176 if (mp->mnt_cred->cr_ruid != 0 ||
177 vfs_scanopt(mp->mnt_optnew, "gid", "%d", &root_gid) != 1)
178 root_gid = va.va_gid;
179 if (mp->mnt_cred->cr_ruid != 0 ||
180 vfs_scanopt(mp->mnt_optnew, "uid", "%d", &root_uid) != 1)
181 root_uid = va.va_uid;
182 if (mp->mnt_cred->cr_ruid != 0 ||
183 vfs_scanopt(mp->mnt_optnew, "mode", "%ho", &root_mode) != 1)
184 root_mode = va.va_mode;
185 if (vfs_getopt_size(mp->mnt_optnew, "inodes", &nodes_max) != 0)
187 if (vfs_getopt_size(mp->mnt_optnew, "size", &size_max) != 0)
189 if (vfs_getopt_size(mp->mnt_optnew, "maxfilesize", &maxfilesize) != 0)
191 nonc = vfs_getopt(mp->mnt_optnew, "nonc", NULL, NULL) == 0;
193 /* Do not allow mounts if we do not have enough memory to preserve
194 * the minimum reserved pages. */
195 if (tmpfs_mem_avail() < TMPFS_PAGES_MINRESERVED)
198 /* Get the maximum number of memory pages this file system is
199 * allowed to use, based on the maximum size the user passed in
200 * the mount structure. A value of zero is treated as if the
201 * maximum available space was requested. */
202 if (size_max == 0 || size_max > OFF_MAX - PAGE_SIZE ||
203 (SIZE_MAX < OFF_MAX && size_max / PAGE_SIZE >= SIZE_MAX))
206 size_max = roundup(size_max, PAGE_SIZE);
207 pages = howmany(size_max, PAGE_SIZE);
211 if (nodes_max <= 3) {
212 if (pages < INT_MAX / nodes_per_page)
213 nodes_max = pages * nodes_per_page;
217 if (nodes_max > INT_MAX)
219 MPASS(nodes_max >= 3);
221 /* Allocate the tmpfs mount structure and fill it. */
222 tmp = (struct tmpfs_mount *)malloc(sizeof(struct tmpfs_mount),
223 M_TMPFSMNT, M_WAITOK | M_ZERO);
225 mtx_init(&tmp->tm_allnode_lock, "tmpfs allnode lock", NULL, MTX_DEF);
226 tmp->tm_nodes_max = nodes_max;
227 tmp->tm_nodes_inuse = 0;
228 tmp->tm_refcount = 1;
229 tmp->tm_maxfilesize = maxfilesize > 0 ? maxfilesize : OFF_MAX;
230 LIST_INIT(&tmp->tm_nodes_used);
232 tmp->tm_pages_max = pages;
233 tmp->tm_pages_used = 0;
234 tmp->tm_ino_unr = new_unrhdr(2, INT_MAX, &tmp->tm_allnode_lock);
235 tmp->tm_dirent_pool = uma_zcreate("TMPFS dirent",
236 sizeof(struct tmpfs_dirent), NULL, NULL, NULL, NULL,
238 tmp->tm_node_pool = uma_zcreate("TMPFS node",
239 sizeof(struct tmpfs_node), tmpfs_node_ctor, tmpfs_node_dtor,
240 tmpfs_node_init, tmpfs_node_fini, UMA_ALIGN_PTR, 0);
241 tmp->tm_ronly = (mp->mnt_flag & MNT_RDONLY) != 0;
244 /* Allocate the root node. */
245 error = tmpfs_alloc_node(mp, tmp, VDIR, root_uid, root_gid,
246 root_mode & ALLPERMS, NULL, NULL, VNOVAL, &root);
248 if (error != 0 || root == NULL) {
249 uma_zdestroy(tmp->tm_node_pool);
250 uma_zdestroy(tmp->tm_dirent_pool);
251 delete_unrhdr(tmp->tm_ino_unr);
252 free(tmp, M_TMPFSMNT);
255 KASSERT(root->tn_id == 2,
256 ("tmpfs root with invalid ino: %ju", (uintmax_t)root->tn_id));
260 mp->mnt_flag |= MNT_LOCAL;
261 mp->mnt_kern_flag |= MNTK_LOOKUP_SHARED | MNTK_EXTENDED_SHARED;
265 mp->mnt_stat.f_namemax = MAXNAMLEN;
267 vfs_mountedfrom(mp, "tmpfs");
274 tmpfs_unmount(struct mount *mp, int mntflags)
276 struct tmpfs_mount *tmp;
277 struct tmpfs_node *node;
280 flags = (mntflags & MNT_FORCE) != 0 ? FORCECLOSE : 0;
281 tmp = VFS_TO_TMPFS(mp);
284 error = vfs_write_suspend_umnt(mp);
288 * At this point, nodes cannot be destroyed by any other
289 * thread because write suspension is started.
293 error = vflush(mp, 0, flags, curthread);
295 vfs_write_resume(mp, VR_START_WRITE);
299 if (mp->mnt_nvnodelistsize == 0) {
304 if ((mntflags & MNT_FORCE) == 0) {
305 vfs_write_resume(mp, VR_START_WRITE);
311 while ((node = LIST_FIRST(&tmp->tm_nodes_used)) != NULL) {
312 TMPFS_NODE_LOCK(node);
313 if (node->tn_type == VDIR)
314 tmpfs_dir_destroy(tmp, node);
315 if (tmpfs_free_node_locked(tmp, node, true))
318 TMPFS_NODE_UNLOCK(node);
323 vfs_write_resume(mp, VR_START_WRITE);
326 mp->mnt_flag &= ~MNT_LOCAL;
333 tmpfs_free_tmp(struct tmpfs_mount *tmp)
336 MPASS(tmp->tm_refcount > 0);
338 if (tmp->tm_refcount > 0) {
344 uma_zdestroy(tmp->tm_dirent_pool);
345 uma_zdestroy(tmp->tm_node_pool);
346 clear_unrhdr(tmp->tm_ino_unr);
347 delete_unrhdr(tmp->tm_ino_unr);
349 mtx_destroy(&tmp->tm_allnode_lock);
350 MPASS(tmp->tm_pages_used == 0);
351 MPASS(tmp->tm_nodes_inuse == 0);
353 free(tmp, M_TMPFSMNT);
357 tmpfs_root(struct mount *mp, int flags, struct vnode **vpp)
361 error = tmpfs_alloc_vp(mp, VFS_TO_TMPFS(mp)->tm_root, flags, vpp);
363 (*vpp)->v_vflag |= VV_ROOT;
368 tmpfs_fhtovp(struct mount *mp, struct fid *fhp, int flags,
371 struct tmpfs_fid *tfhp;
372 struct tmpfs_mount *tmp;
373 struct tmpfs_node *node;
376 tmp = VFS_TO_TMPFS(mp);
378 tfhp = (struct tmpfs_fid *)fhp;
379 if (tfhp->tf_len != sizeof(struct tmpfs_fid))
382 if (tfhp->tf_id >= tmp->tm_nodes_max)
386 LIST_FOREACH(node, &tmp->tm_nodes_used, tn_entries) {
387 if (node->tn_id == tfhp->tf_id &&
388 node->tn_gen == tfhp->tf_gen) {
389 tmpfs_ref_node(node);
396 error = tmpfs_alloc_vp(mp, node, LK_EXCLUSIVE, vpp);
397 tmpfs_free_node(tmp, node);
405 tmpfs_statfs(struct mount *mp, struct statfs *sbp)
407 struct tmpfs_mount *tmp;
410 tmp = VFS_TO_TMPFS(mp);
412 sbp->f_iosize = PAGE_SIZE;
413 sbp->f_bsize = PAGE_SIZE;
415 used = tmpfs_pages_used(tmp);
416 if (tmp->tm_pages_max != ULONG_MAX)
417 sbp->f_blocks = tmp->tm_pages_max;
419 sbp->f_blocks = used + tmpfs_mem_avail();
420 if (sbp->f_blocks <= used)
423 sbp->f_bavail = sbp->f_blocks - used;
424 sbp->f_bfree = sbp->f_bavail;
425 used = tmp->tm_nodes_inuse;
426 sbp->f_files = tmp->tm_nodes_max;
427 if (sbp->f_files <= used)
430 sbp->f_ffree = sbp->f_files - used;
431 /* sbp->f_owner = tmp->tn_uid; */
437 tmpfs_sync(struct mount *mp, int waitfor)
439 struct vnode *vp, *mvp;
440 struct vm_object *obj;
442 if (waitfor == MNT_SUSPEND) {
444 mp->mnt_kern_flag |= MNTK_SUSPEND2 | MNTK_SUSPENDED;
446 } else if (waitfor == MNT_LAZY) {
448 * Handle lazy updates of mtime from writes to mmaped
449 * regions. Use MNT_VNODE_FOREACH_ALL instead of
450 * MNT_VNODE_FOREACH_ACTIVE, since unmap of the
451 * tmpfs-backed vnode does not call vinactive(), due
452 * to vm object type is OBJT_SWAP.
454 MNT_VNODE_FOREACH_ALL(vp, mp, mvp) {
455 if (vp->v_type != VREG) {
460 KASSERT((obj->flags & (OBJ_TMPFS_NODE | OBJ_TMPFS)) ==
461 (OBJ_TMPFS_NODE | OBJ_TMPFS), ("non-tmpfs obj"));
464 * Unlocked read, avoid taking vnode lock if
465 * not needed. Lost update will be handled on
468 if ((obj->flags & OBJ_TMPFS_DIRTY) == 0) {
472 if (vget(vp, LK_EXCLUSIVE | LK_RETRY | LK_INTERLOCK,
475 tmpfs_check_mtime(vp);
483 * The presence of a susp_clean method tells the VFS to track writes.
486 tmpfs_susp_clean(struct mount *mp __unused)
491 * tmpfs vfs operations.
494 struct vfsops tmpfs_vfsops = {
495 .vfs_mount = tmpfs_mount,
496 .vfs_unmount = tmpfs_unmount,
497 .vfs_root = tmpfs_root,
498 .vfs_statfs = tmpfs_statfs,
499 .vfs_fhtovp = tmpfs_fhtovp,
500 .vfs_sync = tmpfs_sync,
501 .vfs_susp_clean = tmpfs_susp_clean,
503 VFS_SET(tmpfs_vfsops, tmpfs, VFCF_JAIL);