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 NetBSD's virtual memory sub-system
37 * (the well-known UVM) to store file data and metadata in an efficient
38 * way. This means that it does not follow the structure of an on-disk
39 * file 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>
50 #include <sys/kernel.h>
52 #include <sys/systm.h>
53 #include <sys/sysctl.h>
56 #include <vm/vm_object.h>
57 #include <vm/vm_param.h>
59 #include <fs/tmpfs/tmpfs.h>
62 * Default permission for root node
64 #define TMPFS_DEFAULT_ROOT_MODE (S_IRWXU|S_IRGRP|S_IXGRP|S_IROTH|S_IXOTH)
66 MALLOC_DEFINE(M_TMPFSMNT, "tmpfs mount", "tmpfs mount structures");
67 MALLOC_DEFINE(M_TMPFSNAME, "tmpfs name", "tmpfs file names");
69 /* --------------------------------------------------------------------- */
71 static int tmpfs_mount(struct mount *);
72 static int tmpfs_unmount(struct mount *, int);
73 static int tmpfs_root(struct mount *, int flags, struct vnode **);
74 static int tmpfs_fhtovp(struct mount *, struct fid *, struct vnode **);
75 static int tmpfs_statfs(struct mount *, struct statfs *);
77 /* --------------------------------------------------------------------- */
79 static const char *tmpfs_opts[] = {
80 "from", "size", "inodes", "uid", "gid", "mode", "export",
84 /* --------------------------------------------------------------------- */
92 char *sname = "vm.swap_info";
93 int soid[SWI_MAXMIB], oid[2];
94 u_int unswdev, total, dmmax, nswapdev;
100 if (kernel_sysctlbyname(curthread, "vm.dmmax", &dmmax, &len,
101 NULL, 0, NULL, 0) != 0)
104 len = sizeof(nswapdev);
105 if (kernel_sysctlbyname(curthread, "vm.nswapdev",
107 NULL, 0, NULL, 0) != 0)
110 mibi = (SWI_MAXMIB - 1) * sizeof(int);
114 if (kernel_sysctl(curthread, oid, 2,
115 soid, &mibi, (void *)sname, strlen(sname),
119 mibi = (SWI_MAXMIB - 1);
120 for (unswdev = 0; unswdev < nswapdev; ++unswdev) {
121 soid[mibi] = unswdev;
122 len = sizeof(struct xswdev);
123 if (kernel_sysctl(curthread,
124 soid, mibi + 1, &xsd, &len, NULL, 0,
127 if (len == sizeof(struct xswdev))
128 total += (xsd.xsw_nblks - dmmax);
135 /* --------------------------------------------------------------------- */
137 tmpfs_node_ctor(void *mem, int size, void *arg, int flags)
139 struct tmpfs_node *node = (struct tmpfs_node *)mem;
146 node->tn_vnode = NULL;
147 node->tn_vpstate = 0;
153 tmpfs_node_dtor(void *mem, int size, void *arg)
155 struct tmpfs_node *node = (struct tmpfs_node *)mem;
156 node->tn_type = VNON;
160 tmpfs_node_init(void *mem, int size, int flags)
162 struct tmpfs_node *node = (struct tmpfs_node *)mem;
165 mtx_init(&node->tn_interlock, "tmpfs node interlock", NULL, MTX_DEF);
166 node->tn_gen = arc4random();
172 tmpfs_node_fini(void *mem, int size)
174 struct tmpfs_node *node = (struct tmpfs_node *)mem;
176 mtx_destroy(&node->tn_interlock);
180 tmpfs_mount(struct mount *mp)
182 struct tmpfs_mount *tmp;
183 struct tmpfs_node *root;
184 size_t pages, mem_size;
191 /* Root node attributes. */
198 if (vfs_filteropt(mp->mnt_optnew, tmpfs_opts))
201 if (mp->mnt_flag & MNT_UPDATE) {
202 /* XXX: There is no support yet to update file system
203 * settings. Should be added. */
208 printf("WARNING: TMPFS is considered to be a highly experimental "
209 "feature in FreeBSD.\n");
211 vn_lock(mp->mnt_vnodecovered, LK_SHARED | LK_RETRY);
212 error = VOP_GETATTR(mp->mnt_vnodecovered, &va, mp->mnt_cred);
213 VOP_UNLOCK(mp->mnt_vnodecovered, 0);
217 if (mp->mnt_cred->cr_ruid != 0 ||
218 vfs_scanopt(mp->mnt_optnew, "gid", "%d", &root_gid) != 1)
219 root_gid = va.va_gid;
220 if (mp->mnt_cred->cr_ruid != 0 ||
221 vfs_scanopt(mp->mnt_optnew, "uid", "%d", &root_uid) != 1)
222 root_uid = va.va_uid;
223 if (mp->mnt_cred->cr_ruid != 0 ||
224 vfs_scanopt(mp->mnt_optnew, "mode", "%ho", &root_mode) != 1)
225 root_mode = va.va_mode;
226 if (vfs_scanopt(mp->mnt_optnew, "inodes", "%d", &nodes_max) != 1)
228 if (vfs_scanopt(mp->mnt_optnew, "size", "%qu", &size_max) != 1)
231 /* Do not allow mounts if we do not have enough memory to preserve
232 * the minimum reserved pages. */
233 mem_size = cnt.v_free_count + cnt.v_inactive_count + get_swpgtotal();
234 mem_size -= mem_size > cnt.v_wire_count ? cnt.v_wire_count : mem_size;
235 if (mem_size < TMPFS_PAGES_RESERVED)
238 /* Get the maximum number of memory pages this file system is
239 * allowed to use, based on the maximum size the user passed in
240 * the mount structure. A value of zero is treated as if the
241 * maximum available space was requested. */
242 if (size_max < PAGE_SIZE || size_max >= SIZE_MAX)
245 pages = howmany(size_max, PAGE_SIZE);
249 nodes = 3 + pages * PAGE_SIZE / 1024;
254 /* Allocate the tmpfs mount structure and fill it. */
255 tmp = (struct tmpfs_mount *)malloc(sizeof(struct tmpfs_mount),
256 M_TMPFSMNT, M_WAITOK | M_ZERO);
258 mtx_init(&tmp->allnode_lock, "tmpfs allnode lock", NULL, MTX_DEF);
259 tmp->tm_nodes_max = nodes;
260 tmp->tm_nodes_inuse = 0;
261 tmp->tm_maxfilesize = (u_int64_t)(cnt.v_page_count + get_swpgtotal()) * PAGE_SIZE;
262 LIST_INIT(&tmp->tm_nodes_used);
264 tmp->tm_pages_max = pages;
265 tmp->tm_pages_used = 0;
266 tmp->tm_ino_unr = new_unrhdr(2, INT_MAX, &tmp->allnode_lock);
267 tmp->tm_dirent_pool = uma_zcreate("TMPFS dirent",
268 sizeof(struct tmpfs_dirent),
269 NULL, NULL, NULL, NULL,
271 tmp->tm_node_pool = uma_zcreate("TMPFS node",
272 sizeof(struct tmpfs_node),
273 tmpfs_node_ctor, tmpfs_node_dtor,
274 tmpfs_node_init, tmpfs_node_fini,
277 /* Allocate the root node. */
278 error = tmpfs_alloc_node(tmp, VDIR, root_uid,
279 root_gid, root_mode & ALLPERMS, NULL, NULL,
282 if (error != 0 || root == NULL) {
283 uma_zdestroy(tmp->tm_node_pool);
284 uma_zdestroy(tmp->tm_dirent_pool);
285 delete_unrhdr(tmp->tm_ino_unr);
286 free(tmp, M_TMPFSMNT);
289 KASSERT(root->tn_id == 2, ("tmpfs root with invalid ino: %d", root->tn_id));
293 mp->mnt_flag |= MNT_LOCAL;
294 mp->mnt_kern_flag |= MNTK_MPSAFE;
298 mp->mnt_stat.f_namemax = MAXNAMLEN;
300 vfs_mountedfrom(mp, "tmpfs");
305 /* --------------------------------------------------------------------- */
309 tmpfs_unmount(struct mount *mp, int mntflags)
313 struct tmpfs_mount *tmp;
314 struct tmpfs_node *node;
316 /* Handle forced unmounts. */
317 if (mntflags & MNT_FORCE)
320 /* Finalize all pending I/O. */
321 error = vflush(mp, 0, flags, curthread);
325 tmp = VFS_TO_TMPFS(mp);
327 /* Free all associated data. The loop iterates over the linked list
328 * we have containing all used nodes. For each of them that is
329 * a directory, we free all its directory entries. Note that after
330 * freeing a node, it will automatically go to the available list,
331 * so we will later have to iterate over it to release its items. */
332 node = LIST_FIRST(&tmp->tm_nodes_used);
333 while (node != NULL) {
334 struct tmpfs_node *next;
336 if (node->tn_type == VDIR) {
337 struct tmpfs_dirent *de;
339 de = TAILQ_FIRST(&node->tn_dir.tn_dirhead);
341 struct tmpfs_dirent *nde;
343 nde = TAILQ_NEXT(de, td_entries);
344 tmpfs_free_dirent(tmp, de, FALSE);
346 node->tn_size -= sizeof(struct tmpfs_dirent);
350 next = LIST_NEXT(node, tn_entries);
351 tmpfs_free_node(tmp, node);
355 uma_zdestroy(tmp->tm_dirent_pool);
356 uma_zdestroy(tmp->tm_node_pool);
357 delete_unrhdr(tmp->tm_ino_unr);
359 mtx_destroy(&tmp->allnode_lock);
360 MPASS(tmp->tm_pages_used == 0);
361 MPASS(tmp->tm_nodes_inuse == 0);
363 /* Throw away the tmpfs_mount structure. */
364 free(mp->mnt_data, M_TMPFSMNT);
368 mp->mnt_flag &= ~MNT_LOCAL;
373 /* --------------------------------------------------------------------- */
376 tmpfs_root(struct mount *mp, int flags, struct vnode **vpp)
379 error = tmpfs_alloc_vp(mp, VFS_TO_TMPFS(mp)->tm_root, flags, vpp);
382 (*vpp)->v_vflag |= VV_ROOT;
387 /* --------------------------------------------------------------------- */
390 tmpfs_fhtovp(struct mount *mp, struct fid *fhp, struct vnode **vpp)
393 struct tmpfs_fid *tfhp;
394 struct tmpfs_mount *tmp;
395 struct tmpfs_node *node;
397 tmp = VFS_TO_TMPFS(mp);
399 tfhp = (struct tmpfs_fid *)fhp;
400 if (tfhp->tf_len != sizeof(struct tmpfs_fid))
403 if (tfhp->tf_id >= tmp->tm_nodes_max)
409 LIST_FOREACH(node, &tmp->tm_nodes_used, tn_entries) {
410 if (node->tn_id == tfhp->tf_id &&
411 node->tn_gen == tfhp->tf_gen) {
419 return (tmpfs_alloc_vp(mp, node, LK_EXCLUSIVE, vpp));
424 /* --------------------------------------------------------------------- */
428 tmpfs_statfs(struct mount *mp, struct statfs *sbp)
430 fsfilcnt_t freenodes;
431 struct tmpfs_mount *tmp;
433 tmp = VFS_TO_TMPFS(mp);
435 sbp->f_iosize = PAGE_SIZE;
436 sbp->f_bsize = PAGE_SIZE;
438 sbp->f_blocks = TMPFS_PAGES_MAX(tmp);
439 sbp->f_bavail = sbp->f_bfree = TMPFS_PAGES_AVAIL(tmp);
441 freenodes = MIN(tmp->tm_nodes_max - tmp->tm_nodes_inuse,
442 TMPFS_PAGES_AVAIL(tmp) * PAGE_SIZE / sizeof(struct tmpfs_node));
444 sbp->f_files = freenodes + tmp->tm_nodes_inuse;
445 sbp->f_ffree = freenodes;
446 /* sbp->f_owner = tmp->tn_uid; */
451 /* --------------------------------------------------------------------- */
454 * tmpfs vfs operations.
457 struct vfsops tmpfs_vfsops = {
458 .vfs_mount = tmpfs_mount,
459 .vfs_unmount = tmpfs_unmount,
460 .vfs_root = tmpfs_root,
461 .vfs_statfs = tmpfs_statfs,
462 .vfs_fhtovp = tmpfs_fhtovp,
464 VFS_SET(tmpfs_vfsops, tmpfs, 0);