1 /* $NetBSD: tmpfs_vfsops.c,v 1.10 2005/12/11 12:24:29 christos Exp $ */
4 * Copyright (c) 2005 The NetBSD Foundation, Inc.
7 * This code is derived from software contributed to The NetBSD Foundation
8 * by Julio M. Merino Vidal, developed as part of Google's Summer of Code
11 * Redistribution and use in source and binary forms, with or without
12 * modification, are permitted provided that the following conditions
14 * 1. Redistributions of source code must retain the above copyright
15 * notice, this list of conditions and the following disclaimer.
16 * 2. Redistributions in binary form must reproduce the above copyright
17 * notice, this list of conditions and the following disclaimer in the
18 * documentation and/or other materials provided with the distribution.
20 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
21 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
22 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
23 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
24 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
25 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
26 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
27 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
28 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
29 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
30 * POSSIBILITY OF SUCH DAMAGE.
34 * Efficient memory file system.
36 * tmpfs is a file system that uses FreeBSD's virtual memory
37 * sub-system to store file data and metadata in an efficient way.
38 * This means that it does not follow the structure of an on-disk file
39 * system because it simply does not need to. Instead, it uses
40 * memory-specific data structures and algorithms to automatically
41 * allocate and release resources.
43 #include <sys/cdefs.h>
44 __FBSDID("$FreeBSD$");
46 #include <sys/param.h>
47 #include <sys/limits.h>
49 #include <sys/mutex.h>
52 #include <sys/kernel.h>
53 #include <sys/rwlock.h>
55 #include <sys/systm.h>
56 #include <sys/sysctl.h>
59 #include <vm/vm_object.h>
60 #include <vm/vm_param.h>
62 #include <fs/tmpfs/tmpfs.h>
65 * Default permission for root node
67 #define TMPFS_DEFAULT_ROOT_MODE (S_IRWXU|S_IRGRP|S_IXGRP|S_IROTH|S_IXOTH)
69 MALLOC_DEFINE(M_TMPFSMNT, "tmpfs mount", "tmpfs mount structures");
70 MALLOC_DEFINE(M_TMPFSNAME, "tmpfs name", "tmpfs file names");
72 static int tmpfs_mount(struct mount *);
73 static int tmpfs_unmount(struct mount *, int);
74 static int tmpfs_root(struct mount *, int flags, struct vnode **);
75 static int tmpfs_fhtovp(struct mount *, struct fid *, int,
77 static int tmpfs_statfs(struct mount *, struct statfs *);
79 static const char *tmpfs_opts[] = {
80 "from", "size", "maxfilesize", "inodes", "uid", "gid", "mode", "export",
84 static const char *tmpfs_updateopts[] = {
85 "from", "export", NULL
89 tmpfs_node_ctor(void *mem, int size, void *arg, int flags)
91 struct tmpfs_node *node = (struct tmpfs_node *)mem;
98 node->tn_vnode = NULL;
105 tmpfs_node_dtor(void *mem, int size, void *arg)
107 struct tmpfs_node *node = (struct tmpfs_node *)mem;
108 node->tn_type = VNON;
112 tmpfs_node_init(void *mem, int size, int flags)
114 struct tmpfs_node *node = (struct tmpfs_node *)mem;
117 mtx_init(&node->tn_interlock, "tmpfs node interlock", NULL, MTX_DEF);
118 node->tn_gen = arc4random();
124 tmpfs_node_fini(void *mem, int size)
126 struct tmpfs_node *node = (struct tmpfs_node *)mem;
128 mtx_destroy(&node->tn_interlock);
132 tmpfs_mount(struct mount *mp)
134 const size_t nodes_per_page = howmany(PAGE_SIZE,
135 sizeof(struct tmpfs_dirent) + sizeof(struct tmpfs_node));
136 struct tmpfs_mount *tmp;
137 struct tmpfs_node *root;
138 struct thread *td = curthread;
143 off_t nodes_max, size_max, maxfilesize;
145 /* Root node attributes. */
152 if (!prison_allow(td->td_ucred, PR_ALLOW_MOUNT_TMPFS))
155 if (vfs_filteropt(mp->mnt_optnew, tmpfs_opts))
158 if (mp->mnt_flag & MNT_UPDATE) {
159 /* Only support update mounts for certain options. */
160 if (vfs_filteropt(mp->mnt_optnew, tmpfs_updateopts) != 0)
162 if (vfs_flagopt(mp->mnt_optnew, "ro", NULL, 0) !=
163 ((struct tmpfs_mount *)mp->mnt_data)->tm_ronly)
168 vn_lock(mp->mnt_vnodecovered, LK_SHARED | LK_RETRY);
169 error = VOP_GETATTR(mp->mnt_vnodecovered, &va, mp->mnt_cred);
170 VOP_UNLOCK(mp->mnt_vnodecovered, 0);
174 if (mp->mnt_cred->cr_ruid != 0 ||
175 vfs_scanopt(mp->mnt_optnew, "gid", "%d", &root_gid) != 1)
176 root_gid = va.va_gid;
177 if (mp->mnt_cred->cr_ruid != 0 ||
178 vfs_scanopt(mp->mnt_optnew, "uid", "%d", &root_uid) != 1)
179 root_uid = va.va_uid;
180 if (mp->mnt_cred->cr_ruid != 0 ||
181 vfs_scanopt(mp->mnt_optnew, "mode", "%ho", &root_mode) != 1)
182 root_mode = va.va_mode;
183 if (vfs_getopt_size(mp->mnt_optnew, "inodes", &nodes_max) != 0)
185 if (vfs_getopt_size(mp->mnt_optnew, "size", &size_max) != 0)
187 if (vfs_getopt_size(mp->mnt_optnew, "maxfilesize", &maxfilesize) != 0)
189 nonc = vfs_getopt(mp->mnt_optnew, "nonc", NULL, NULL) == 0;
191 /* Do not allow mounts if we do not have enough memory to preserve
192 * the minimum reserved pages. */
193 if (tmpfs_mem_avail() < TMPFS_PAGES_MINRESERVED)
196 /* Get the maximum number of memory pages this file system is
197 * allowed to use, based on the maximum size the user passed in
198 * the mount structure. A value of zero is treated as if the
199 * maximum available space was requested. */
200 if (size_max == 0 || size_max > OFF_MAX - PAGE_SIZE ||
201 (SIZE_MAX < OFF_MAX && size_max / PAGE_SIZE >= SIZE_MAX))
204 size_max = roundup(size_max, PAGE_SIZE);
205 pages = howmany(size_max, PAGE_SIZE);
209 if (nodes_max <= 3) {
210 if (pages < INT_MAX / nodes_per_page)
211 nodes_max = pages * nodes_per_page;
215 if (nodes_max > INT_MAX)
217 MPASS(nodes_max >= 3);
219 /* Allocate the tmpfs mount structure and fill it. */
220 tmp = (struct tmpfs_mount *)malloc(sizeof(struct tmpfs_mount),
221 M_TMPFSMNT, M_WAITOK | M_ZERO);
223 mtx_init(&tmp->tm_allnode_lock, "tmpfs allnode lock", NULL, MTX_DEF);
224 tmp->tm_nodes_max = nodes_max;
225 tmp->tm_nodes_inuse = 0;
226 tmp->tm_refcount = 1;
227 tmp->tm_maxfilesize = maxfilesize > 0 ? maxfilesize : OFF_MAX;
228 LIST_INIT(&tmp->tm_nodes_used);
230 tmp->tm_pages_max = pages;
231 tmp->tm_pages_used = 0;
232 tmp->tm_ino_unr = new_unrhdr(2, INT_MAX, &tmp->tm_allnode_lock);
233 tmp->tm_dirent_pool = uma_zcreate("TMPFS dirent",
234 sizeof(struct tmpfs_dirent), NULL, NULL, NULL, NULL,
236 tmp->tm_node_pool = uma_zcreate("TMPFS node",
237 sizeof(struct tmpfs_node), tmpfs_node_ctor, tmpfs_node_dtor,
238 tmpfs_node_init, tmpfs_node_fini, UMA_ALIGN_PTR, 0);
239 tmp->tm_ronly = (mp->mnt_flag & MNT_RDONLY) != 0;
242 /* Allocate the root node. */
243 error = tmpfs_alloc_node(mp, tmp, VDIR, root_uid, root_gid,
244 root_mode & ALLPERMS, NULL, NULL, VNOVAL, &root);
246 if (error != 0 || root == NULL) {
247 uma_zdestroy(tmp->tm_node_pool);
248 uma_zdestroy(tmp->tm_dirent_pool);
249 delete_unrhdr(tmp->tm_ino_unr);
250 free(tmp, M_TMPFSMNT);
253 KASSERT(root->tn_id == 2,
254 ("tmpfs root with invalid ino: %ju", (uintmax_t)root->tn_id));
258 mp->mnt_flag |= MNT_LOCAL;
262 mp->mnt_stat.f_namemax = MAXNAMLEN;
264 vfs_mountedfrom(mp, "tmpfs");
271 tmpfs_unmount(struct mount *mp, int mntflags)
273 struct tmpfs_mount *tmp;
274 struct tmpfs_node *node;
277 flags = (mntflags & MNT_FORCE) != 0 ? FORCECLOSE : 0;
278 tmp = VFS_TO_TMPFS(mp);
281 error = vfs_write_suspend_umnt(mp);
285 * At this point, nodes cannot be destroyed by any other
286 * thread because write suspension is started.
290 error = vflush(mp, 0, flags, curthread);
292 vfs_write_resume(mp, VR_START_WRITE);
296 if (mp->mnt_nvnodelistsize == 0) {
301 if ((mntflags & MNT_FORCE) == 0) {
302 vfs_write_resume(mp, VR_START_WRITE);
308 while ((node = LIST_FIRST(&tmp->tm_nodes_used)) != NULL) {
309 TMPFS_NODE_LOCK(node);
310 if (node->tn_type == VDIR)
311 tmpfs_dir_destroy(tmp, node);
312 if (tmpfs_free_node_locked(tmp, node, true))
315 TMPFS_NODE_UNLOCK(node);
320 vfs_write_resume(mp, VR_START_WRITE);
323 mp->mnt_flag &= ~MNT_LOCAL;
330 tmpfs_free_tmp(struct tmpfs_mount *tmp)
333 MPASS(tmp->tm_refcount > 0);
335 if (tmp->tm_refcount > 0) {
341 uma_zdestroy(tmp->tm_dirent_pool);
342 uma_zdestroy(tmp->tm_node_pool);
343 delete_unrhdr(tmp->tm_ino_unr);
345 mtx_destroy(&tmp->tm_allnode_lock);
346 MPASS(tmp->tm_pages_used == 0);
347 MPASS(tmp->tm_nodes_inuse == 0);
349 free(tmp, M_TMPFSMNT);
353 tmpfs_root(struct mount *mp, int flags, struct vnode **vpp)
357 error = tmpfs_alloc_vp(mp, VFS_TO_TMPFS(mp)->tm_root, flags, vpp);
359 (*vpp)->v_vflag |= VV_ROOT;
364 tmpfs_fhtovp(struct mount *mp, struct fid *fhp, int flags,
367 struct tmpfs_fid *tfhp;
368 struct tmpfs_mount *tmp;
369 struct tmpfs_node *node;
372 tmp = VFS_TO_TMPFS(mp);
374 tfhp = (struct tmpfs_fid *)fhp;
375 if (tfhp->tf_len != sizeof(struct tmpfs_fid))
378 if (tfhp->tf_id >= tmp->tm_nodes_max)
382 LIST_FOREACH(node, &tmp->tm_nodes_used, tn_entries) {
383 if (node->tn_id == tfhp->tf_id &&
384 node->tn_gen == tfhp->tf_gen) {
385 tmpfs_ref_node(node);
392 error = tmpfs_alloc_vp(mp, node, LK_EXCLUSIVE, vpp);
393 tmpfs_free_node(tmp, node);
401 tmpfs_statfs(struct mount *mp, struct statfs *sbp)
403 struct tmpfs_mount *tmp;
406 tmp = VFS_TO_TMPFS(mp);
408 sbp->f_iosize = PAGE_SIZE;
409 sbp->f_bsize = PAGE_SIZE;
411 used = tmpfs_pages_used(tmp);
412 if (tmp->tm_pages_max != ULONG_MAX)
413 sbp->f_blocks = tmp->tm_pages_max;
415 sbp->f_blocks = used + tmpfs_mem_avail();
416 if (sbp->f_blocks <= used)
419 sbp->f_bavail = sbp->f_blocks - used;
420 sbp->f_bfree = sbp->f_bavail;
421 used = tmp->tm_nodes_inuse;
422 sbp->f_files = tmp->tm_nodes_max;
423 if (sbp->f_files <= used)
426 sbp->f_ffree = sbp->f_files - used;
427 /* sbp->f_owner = tmp->tn_uid; */
433 tmpfs_sync(struct mount *mp, int waitfor)
435 struct vnode *vp, *mvp;
436 struct vm_object *obj;
438 if (waitfor == MNT_SUSPEND) {
440 mp->mnt_kern_flag |= MNTK_SUSPEND2 | MNTK_SUSPENDED;
442 } else if (waitfor == MNT_LAZY) {
444 * Handle lazy updates of mtime from writes to mmaped
445 * regions. Use MNT_VNODE_FOREACH_ALL instead of
446 * MNT_VNODE_FOREACH_ACTIVE, since unmap of the
447 * tmpfs-backed vnode does not call vinactive(), due
448 * to vm object type is OBJT_SWAP.
450 MNT_VNODE_FOREACH_ALL(vp, mp, mvp) {
451 if (vp->v_type != VREG) {
456 KASSERT((obj->flags & (OBJ_TMPFS_NODE | OBJ_TMPFS)) ==
457 (OBJ_TMPFS_NODE | OBJ_TMPFS), ("non-tmpfs obj"));
460 * Unlocked read, avoid taking vnode lock if
461 * not needed. Lost update will be handled on
464 if ((obj->flags & OBJ_TMPFS_DIRTY) == 0) {
468 if (vget(vp, LK_EXCLUSIVE | LK_RETRY | LK_INTERLOCK,
471 tmpfs_check_mtime(vp);
479 * tmpfs vfs operations.
482 struct vfsops tmpfs_vfsops = {
483 .vfs_mount = tmpfs_mount,
484 .vfs_unmount = tmpfs_unmount,
485 .vfs_root = tmpfs_root,
486 .vfs_statfs = tmpfs_statfs,
487 .vfs_fhtovp = tmpfs_fhtovp,
488 .vfs_sync = tmpfs_sync,
490 VFS_SET(tmpfs_vfsops, tmpfs, VFCF_JAIL);