1 /* $NetBSD: vfs_init.c,v 1.43 2009/01/17 07:02:35 yamt Exp $ */ 2 3 /*- 4 * Copyright (c) 1998, 2000, 2008 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Jason R. Thorpe of the Numerical Aerospace Simulation Facility, 9 * NASA Ames Research Center. 10 * 11 * Redistribution and use in source and binary forms, with or without 12 * modification, are permitted provided that the following conditions 13 * are met: 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. 19 * 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. 31 */ 32 33 /* 34 * Copyright (c) 1989, 1993 35 * The Regents of the University of California. All rights reserved. 36 * 37 * This code is derived from software contributed 38 * to Berkeley by John Heidemann of the UCLA Ficus project. 39 * 40 * Source: * @(#)i405_init.c 2.10 92/04/27 UCLA Ficus project 41 * 42 * Redistribution and use in source and binary forms, with or without 43 * modification, are permitted provided that the following conditions 44 * are met: 45 * 1. Redistributions of source code must retain the above copyright 46 * notice, this list of conditions and the following disclaimer. 47 * 2. Redistributions in binary form must reproduce the above copyright 48 * notice, this list of conditions and the following disclaimer in the 49 * documentation and/or other materials provided with the distribution. 50 * 3. Neither the name of the University nor the names of its contributors 51 * may be used to endorse or promote products derived from this software 52 * without specific prior written permission. 53 * 54 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 55 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 56 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 57 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 58 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 59 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 60 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 61 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 62 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 63 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 64 * SUCH DAMAGE. 65 * 66 * @(#)vfs_init.c 8.5 (Berkeley) 5/11/95 67 */ 68 69 #include <sys/cdefs.h> 70 __KERNEL_RCSID(0, "$NetBSD: vfs_init.c,v 1.43 2009/01/17 07:02:35 yamt Exp $"); 71 72 #include <sys/param.h> 73 #include <sys/mount.h> 74 #include <sys/time.h> 75 #include <sys/vnode.h> 76 #include <sys/stat.h> 77 #include <sys/namei.h> 78 #include <sys/ucred.h> 79 #include <sys/buf.h> 80 #include <sys/errno.h> 81 #include <sys/kmem.h> 82 #include <sys/systm.h> 83 #include <sys/module.h> 84 #include <sys/dirhash.h> 85 #include <sys/sysctl.h> 86 87 /* 88 * Sigh, such primitive tools are these... 89 */ 90 #if 0 91 #define DODEBUG(A) A 92 #else 93 #define DODEBUG(A) 94 #endif 95 96 /* 97 * The global list of vnode operations. 98 */ 99 extern const struct vnodeop_desc * const vfs_op_descs[]; 100 101 /* 102 * These vnodeopv_descs are listed here because they are not 103 * associated with any particular file system, and thus cannot 104 * be initialized by vfs_attach(). 105 */ 106 extern const struct vnodeopv_desc dead_vnodeop_opv_desc; 107 extern const struct vnodeopv_desc fifo_vnodeop_opv_desc; 108 extern const struct vnodeopv_desc spec_vnodeop_opv_desc; 109 extern const struct vnodeopv_desc sync_vnodeop_opv_desc; 110 111 const struct vnodeopv_desc * const vfs_special_vnodeopv_descs[] = { 112 &dead_vnodeop_opv_desc, 113 &fifo_vnodeop_opv_desc, 114 &spec_vnodeop_opv_desc, 115 &sync_vnodeop_opv_desc, 116 NULL, 117 }; 118 119 struct vfs_list_head vfs_list = /* vfs list */ 120 LIST_HEAD_INITIALIZER(vfs_list); 121 122 /* 123 * This code doesn't work if the defn is **vnodop_defns with cc. 124 * The problem is because of the compiler sometimes putting in an 125 * extra level of indirection for arrays. It's an interesting 126 * "feature" of C. 127 */ 128 typedef int (*PFI)(void *); 129 130 /* 131 * A miscellaneous routine. 132 * A generic "default" routine that just returns an error. 133 */ 134 /*ARGSUSED*/ 135 int 136 vn_default_error(void *v) 137 { 138 139 return (EOPNOTSUPP); 140 } 141 142 static struct sysctllog *vfs_sysctllog; 143 144 /* 145 * Top level filesystem related information gathering. 146 */ 147 static void 148 sysctl_vfs_setup(void) 149 { 150 extern int dovfsusermount; 151 extern int vfs_magiclinks; 152 153 sysctl_createv(&vfs_sysctllog, 0, NULL, NULL, 154 CTLFLAG_PERMANENT, 155 CTLTYPE_NODE, "vfs", NULL, 156 NULL, 0, NULL, 0, 157 CTL_VFS, CTL_EOL); 158 sysctl_createv(&vfs_sysctllog, 0, NULL, NULL, 159 CTLFLAG_PERMANENT, 160 CTLTYPE_NODE, "generic", 161 SYSCTL_DESCR("Non-specific vfs related information"), 162 NULL, 0, NULL, 0, 163 CTL_VFS, VFS_GENERIC, CTL_EOL); 164 sysctl_createv(&vfs_sysctllog, 0, NULL, NULL, 165 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 166 CTLTYPE_INT, "usermount", 167 SYSCTL_DESCR("Whether unprivileged users may mount " 168 "filesystems"), 169 NULL, 0, &dovfsusermount, 0, 170 CTL_VFS, VFS_GENERIC, VFS_USERMOUNT, CTL_EOL); 171 sysctl_createv(&vfs_sysctllog, 0, NULL, NULL, 172 CTLFLAG_PERMANENT, 173 CTLTYPE_STRING, "fstypes", 174 SYSCTL_DESCR("List of file systems present"), 175 sysctl_vfs_generic_fstypes, 0, NULL, 0, 176 CTL_VFS, VFS_GENERIC, CTL_CREATE, CTL_EOL); 177 sysctl_createv(&vfs_sysctllog, 0, NULL, NULL, 178 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 179 CTLTYPE_INT, "magiclinks", 180 SYSCTL_DESCR("Whether \"magic\" symlinks are expanded"), 181 NULL, 0, &vfs_magiclinks, 0, 182 CTL_VFS, VFS_GENERIC, VFS_MAGICLINKS, CTL_EOL); 183 } 184 185 186 /* 187 * vfs_init.c 188 * 189 * Allocate and fill in operations vectors. 190 * 191 * An undocumented feature of this approach to defining operations is that 192 * there can be multiple entries in vfs_opv_descs for the same operations 193 * vector. This allows third parties to extend the set of operations 194 * supported by another layer in a binary compatibile way. For example, 195 * assume that NFS needed to be modified to support Ficus. NFS has an entry 196 * (probably nfs_vnopdeop_decls) declaring all the operations NFS supports by 197 * default. Ficus could add another entry (ficus_nfs_vnodeop_decl_entensions) 198 * listing those new operations Ficus adds to NFS, all without modifying the 199 * NFS code. (Of couse, the OTW NFS protocol still needs to be munged, but 200 * that is a(whole)nother story.) This is a feature. 201 */ 202 203 /* 204 * Init the vector, if it needs it. 205 * Also handle backwards compatibility. 206 */ 207 static void 208 vfs_opv_init_explicit(const struct vnodeopv_desc *vfs_opv_desc) 209 { 210 int (**opv_desc_vector)(void *); 211 const struct vnodeopv_entry_desc *opve_descp; 212 213 opv_desc_vector = *(vfs_opv_desc->opv_desc_vector_p); 214 215 for (opve_descp = vfs_opv_desc->opv_desc_ops; 216 opve_descp->opve_op; 217 opve_descp++) { 218 /* 219 * Sanity check: is this operation listed 220 * in the list of operations? We check this 221 * by seeing if its offset is zero. Since 222 * the default routine should always be listed 223 * first, it should be the only one with a zero 224 * offset. Any other operation with a zero 225 * offset is probably not listed in 226 * vfs_op_descs, and so is probably an error. 227 * 228 * A panic here means the layer programmer 229 * has committed the all-too common bug 230 * of adding a new operation to the layer's 231 * list of vnode operations but 232 * not adding the operation to the system-wide 233 * list of supported operations. 234 */ 235 if (opve_descp->opve_op->vdesc_offset == 0 && 236 opve_descp->opve_op->vdesc_offset != VOFFSET(vop_default)) { 237 printf("operation %s not listed in %s.\n", 238 opve_descp->opve_op->vdesc_name, "vfs_op_descs"); 239 panic ("vfs_opv_init: bad operation"); 240 } 241 242 /* 243 * Fill in this entry. 244 */ 245 opv_desc_vector[opve_descp->opve_op->vdesc_offset] = 246 opve_descp->opve_impl; 247 } 248 } 249 250 static void 251 vfs_opv_init_default(const struct vnodeopv_desc *vfs_opv_desc) 252 { 253 int j; 254 int (**opv_desc_vector)(void *); 255 256 opv_desc_vector = *(vfs_opv_desc->opv_desc_vector_p); 257 258 /* 259 * Force every operations vector to have a default routine. 260 */ 261 if (opv_desc_vector[VOFFSET(vop_default)] == NULL) 262 panic("vfs_opv_init: operation vector without default routine."); 263 264 for (j = 0; j < VNODE_OPS_COUNT; j++) 265 if (opv_desc_vector[j] == NULL) 266 opv_desc_vector[j] = 267 opv_desc_vector[VOFFSET(vop_default)]; 268 } 269 270 void 271 vfs_opv_init(const struct vnodeopv_desc * const *vopvdpp) 272 { 273 int (**opv_desc_vector)(void *); 274 int i; 275 276 /* 277 * Allocate the vectors. 278 */ 279 for (i = 0; vopvdpp[i] != NULL; i++) { 280 opv_desc_vector = 281 kmem_alloc(VNODE_OPS_COUNT * sizeof(PFI), KM_SLEEP); 282 memset(opv_desc_vector, 0, VNODE_OPS_COUNT * sizeof(PFI)); 283 *(vopvdpp[i]->opv_desc_vector_p) = opv_desc_vector; 284 DODEBUG(printf("vector at %p allocated\n", 285 opv_desc_vector_p)); 286 } 287 288 /* 289 * ...and fill them in. 290 */ 291 for (i = 0; vopvdpp[i] != NULL; i++) 292 vfs_opv_init_explicit(vopvdpp[i]); 293 294 /* 295 * Finally, go back and replace unfilled routines 296 * with their default. 297 */ 298 for (i = 0; vopvdpp[i] != NULL; i++) 299 vfs_opv_init_default(vopvdpp[i]); 300 } 301 302 void 303 vfs_opv_free(const struct vnodeopv_desc * const *vopvdpp) 304 { 305 int i; 306 307 /* 308 * Free the vectors allocated in vfs_opv_init(). 309 */ 310 for (i = 0; vopvdpp[i] != NULL; i++) { 311 kmem_free(*(vopvdpp[i]->opv_desc_vector_p), 312 VNODE_OPS_COUNT * sizeof(PFI)); 313 *(vopvdpp[i]->opv_desc_vector_p) = NULL; 314 } 315 } 316 317 #ifdef DEBUG 318 static void 319 vfs_op_check(void) 320 { 321 int i; 322 323 DODEBUG(printf("Vnode_interface_init.\n")); 324 325 /* 326 * Check offset of each op. 327 */ 328 for (i = 0; vfs_op_descs[i]; i++) { 329 if (vfs_op_descs[i]->vdesc_offset != i) 330 panic("vfs_op_check: vfs_op_desc[] offset mismatch"); 331 } 332 333 if (i != VNODE_OPS_COUNT) { 334 panic("vfs_op_check: vnode ops count mismatch (%d != %d)", 335 i, VNODE_OPS_COUNT); 336 } 337 338 DODEBUG(printf ("vfs_opv_numops=%d\n", VNODE_OPS_COUNT)); 339 } 340 #endif /* DEBUG */ 341 342 /* 343 * Initialize the vnode structures and initialize each file system type. 344 */ 345 void 346 vfsinit(void) 347 { 348 349 /* 350 * Attach sysctl nodes 351 */ 352 sysctl_vfs_setup(); 353 354 /* 355 * Initialize the namei pathname buffer pool and cache. 356 */ 357 pnbuf_cache = pool_cache_init(MAXPATHLEN, 0, 0, 0, "pnbufpl", 358 NULL, IPL_NONE, NULL, NULL, NULL); 359 KASSERT(pnbuf_cache != NULL); 360 361 /* 362 * Initialize the vnode table 363 */ 364 vntblinit(); 365 366 /* 367 * Initialize the vnode name cache 368 */ 369 nchinit(); 370 371 #ifdef DEBUG 372 /* 373 * Check the list of vnode operations. 374 */ 375 vfs_op_check(); 376 #endif 377 378 /* 379 * Initialize the special vnode operations. 380 */ 381 vfs_opv_init(vfs_special_vnodeopv_descs); 382 383 /* 384 * Initialise generic dirhash. 385 */ 386 dirhash_init(); 387 388 /* 389 * Initialise VFS hooks. 390 */ 391 vfs_hooks_init(); 392 393 /* 394 * Establish each file system which was statically 395 * included in the kernel. 396 */ 397 module_init_class(MODULE_CLASS_VFS); 398 } 399 400 /* 401 * Drop a reference to a file system type. 402 */ 403 void 404 vfs_delref(struct vfsops *vfs) 405 { 406 407 mutex_enter(&vfs_list_lock); 408 vfs->vfs_refcount--; 409 mutex_exit(&vfs_list_lock); 410 } 411 412 /* 413 * Establish a file system and initialize it. 414 */ 415 int 416 vfs_attach(struct vfsops *vfs) 417 { 418 struct vfsops *v; 419 int error = 0; 420 421 mutex_enter(&vfs_list_lock); 422 423 /* 424 * Make sure this file system doesn't already exist. 425 */ 426 LIST_FOREACH(v, &vfs_list, vfs_list) { 427 if (strcmp(vfs->vfs_name, v->vfs_name) == 0) { 428 error = EEXIST; 429 goto out; 430 } 431 } 432 433 /* 434 * Initialize the vnode operations for this file system. 435 */ 436 vfs_opv_init(vfs->vfs_opv_descs); 437 438 /* 439 * Now initialize the file system itself. 440 */ 441 (*vfs->vfs_init)(); 442 443 /* 444 * ...and link it into the kernel's list. 445 */ 446 LIST_INSERT_HEAD(&vfs_list, vfs, vfs_list); 447 448 /* 449 * Sanity: make sure the reference count is 0. 450 */ 451 vfs->vfs_refcount = 0; 452 out: 453 mutex_exit(&vfs_list_lock); 454 return (error); 455 } 456 457 /* 458 * Remove a file system from the kernel. 459 */ 460 int 461 vfs_detach(struct vfsops *vfs) 462 { 463 struct vfsops *v; 464 int error = 0; 465 466 mutex_enter(&vfs_list_lock); 467 468 /* 469 * Make sure no one is using the filesystem. 470 */ 471 if (vfs->vfs_refcount != 0) { 472 error = EBUSY; 473 goto out; 474 } 475 476 /* 477 * ...and remove it from the kernel's list. 478 */ 479 LIST_FOREACH(v, &vfs_list, vfs_list) { 480 if (v == vfs) { 481 LIST_REMOVE(v, vfs_list); 482 break; 483 } 484 } 485 486 if (v == NULL) { 487 error = ESRCH; 488 goto out; 489 } 490 491 /* 492 * Now run the file system-specific cleanups. 493 */ 494 (*vfs->vfs_done)(); 495 496 /* 497 * Free the vnode operations vector. 498 */ 499 vfs_opv_free(vfs->vfs_opv_descs); 500 out: 501 mutex_exit(&vfs_list_lock); 502 return (error); 503 } 504 505 void 506 vfs_reinit(void) 507 { 508 struct vfsops *vfs; 509 510 mutex_enter(&vfs_list_lock); 511 LIST_FOREACH(vfs, &vfs_list, vfs_list) { 512 if (vfs->vfs_reinit) { 513 vfs->vfs_refcount++; 514 mutex_exit(&vfs_list_lock); 515 (*vfs->vfs_reinit)(); 516 mutex_enter(&vfs_list_lock); 517 vfs->vfs_refcount--; 518 } 519 } 520 mutex_exit(&vfs_list_lock); 521 } 522