nfs_subs.c revision 1.201.4.1
1 /* $NetBSD: nfs_subs.c,v 1.201.4.1 2008/04/27 12:52:50 yamt Exp $ */ 2 3 /* 4 * Copyright (c) 1989, 1993 5 * The Regents of the University of California. All rights reserved. 6 * 7 * This code is derived from software contributed to Berkeley by 8 * Rick Macklem at The University of Guelph. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 3. Neither the name of the University nor the names of its contributors 19 * may be used to endorse or promote products derived from this software 20 * without specific prior written permission. 21 * 22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 25 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 32 * SUCH DAMAGE. 33 * 34 * @(#)nfs_subs.c 8.8 (Berkeley) 5/22/95 35 */ 36 37 /* 38 * Copyright 2000 Wasabi Systems, Inc. 39 * All rights reserved. 40 * 41 * Written by Frank van der Linden for Wasabi Systems, Inc. 42 * 43 * Redistribution and use in source and binary forms, with or without 44 * modification, are permitted provided that the following conditions 45 * are met: 46 * 1. Redistributions of source code must retain the above copyright 47 * notice, this list of conditions and the following disclaimer. 48 * 2. Redistributions in binary form must reproduce the above copyright 49 * notice, this list of conditions and the following disclaimer in the 50 * documentation and/or other materials provided with the distribution. 51 * 3. All advertising materials mentioning features or use of this software 52 * must display the following acknowledgement: 53 * This product includes software developed for the NetBSD Project by 54 * Wasabi Systems, Inc. 55 * 4. The name of Wasabi Systems, Inc. may not be used to endorse 56 * or promote products derived from this software without specific prior 57 * written permission. 58 * 59 * THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``AS IS'' AND 60 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 61 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 62 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL WASABI SYSTEMS, INC 63 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 64 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 65 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 66 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 67 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 68 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 69 * POSSIBILITY OF SUCH DAMAGE. 70 */ 71 72 #include <sys/cdefs.h> 73 __KERNEL_RCSID(0, "$NetBSD: nfs_subs.c,v 1.201.4.1 2008/04/27 12:52:50 yamt Exp $"); 74 75 #include "fs_nfs.h" 76 #include "opt_nfs.h" 77 #include "opt_nfsserver.h" 78 #include "opt_iso.h" 79 #include "opt_inet.h" 80 81 /* 82 * These functions support the macros and help fiddle mbuf chains for 83 * the nfs op functions. They do things like create the rpc header and 84 * copy data between mbuf chains and uio lists. 85 */ 86 #include <sys/param.h> 87 #include <sys/proc.h> 88 #include <sys/systm.h> 89 #include <sys/kernel.h> 90 #include <sys/kmem.h> 91 #include <sys/mount.h> 92 #include <sys/vnode.h> 93 #include <sys/namei.h> 94 #include <sys/mbuf.h> 95 #include <sys/socket.h> 96 #include <sys/stat.h> 97 #include <sys/malloc.h> 98 #include <sys/filedesc.h> 99 #include <sys/time.h> 100 #include <sys/dirent.h> 101 #include <sys/once.h> 102 #include <sys/kauth.h> 103 104 #include <uvm/uvm_extern.h> 105 106 #include <nfs/rpcv2.h> 107 #include <nfs/nfsproto.h> 108 #include <nfs/nfsnode.h> 109 #include <nfs/nfs.h> 110 #include <nfs/xdr_subs.h> 111 #include <nfs/nfsm_subs.h> 112 #include <nfs/nfsmount.h> 113 #include <nfs/nfsrtt.h> 114 #include <nfs/nfs_var.h> 115 116 #include <miscfs/specfs/specdev.h> 117 118 #include <netinet/in.h> 119 #ifdef ISO 120 #include <netiso/iso.h> 121 #endif 122 123 /* 124 * Data items converted to xdr at startup, since they are constant 125 * This is kinda hokey, but may save a little time doing byte swaps 126 */ 127 u_int32_t nfs_xdrneg1; 128 u_int32_t rpc_call, rpc_vers, rpc_reply, rpc_msgdenied, rpc_autherr, 129 rpc_mismatch, rpc_auth_unix, rpc_msgaccepted, 130 rpc_auth_kerb; 131 u_int32_t nfs_prog, nfs_true, nfs_false; 132 133 /* And other global data */ 134 const nfstype nfsv2_type[9] = 135 { NFNON, NFREG, NFDIR, NFBLK, NFCHR, NFLNK, NFNON, NFCHR, NFNON }; 136 const nfstype nfsv3_type[9] = 137 { NFNON, NFREG, NFDIR, NFBLK, NFCHR, NFLNK, NFSOCK, NFFIFO, NFNON }; 138 const enum vtype nv2tov_type[8] = 139 { VNON, VREG, VDIR, VBLK, VCHR, VLNK, VNON, VNON }; 140 const enum vtype nv3tov_type[8] = 141 { VNON, VREG, VDIR, VBLK, VCHR, VLNK, VSOCK, VFIFO }; 142 int nfs_ticks; 143 int nfs_commitsize; 144 145 MALLOC_DEFINE(M_NFSDIROFF, "NFS diroff", "NFS directory cookies"); 146 147 /* NFS client/server stats. */ 148 struct nfsstats nfsstats; 149 150 /* 151 * Mapping of old NFS Version 2 RPC numbers to generic numbers. 152 */ 153 const int nfsv3_procid[NFS_NPROCS] = { 154 NFSPROC_NULL, 155 NFSPROC_GETATTR, 156 NFSPROC_SETATTR, 157 NFSPROC_NOOP, 158 NFSPROC_LOOKUP, 159 NFSPROC_READLINK, 160 NFSPROC_READ, 161 NFSPROC_NOOP, 162 NFSPROC_WRITE, 163 NFSPROC_CREATE, 164 NFSPROC_REMOVE, 165 NFSPROC_RENAME, 166 NFSPROC_LINK, 167 NFSPROC_SYMLINK, 168 NFSPROC_MKDIR, 169 NFSPROC_RMDIR, 170 NFSPROC_READDIR, 171 NFSPROC_FSSTAT, 172 NFSPROC_NOOP, 173 NFSPROC_NOOP, 174 NFSPROC_NOOP, 175 NFSPROC_NOOP, 176 NFSPROC_NOOP 177 }; 178 179 /* 180 * and the reverse mapping from generic to Version 2 procedure numbers 181 */ 182 const int nfsv2_procid[NFS_NPROCS] = { 183 NFSV2PROC_NULL, 184 NFSV2PROC_GETATTR, 185 NFSV2PROC_SETATTR, 186 NFSV2PROC_LOOKUP, 187 NFSV2PROC_NOOP, 188 NFSV2PROC_READLINK, 189 NFSV2PROC_READ, 190 NFSV2PROC_WRITE, 191 NFSV2PROC_CREATE, 192 NFSV2PROC_MKDIR, 193 NFSV2PROC_SYMLINK, 194 NFSV2PROC_CREATE, 195 NFSV2PROC_REMOVE, 196 NFSV2PROC_RMDIR, 197 NFSV2PROC_RENAME, 198 NFSV2PROC_LINK, 199 NFSV2PROC_READDIR, 200 NFSV2PROC_NOOP, 201 NFSV2PROC_STATFS, 202 NFSV2PROC_NOOP, 203 NFSV2PROC_NOOP, 204 NFSV2PROC_NOOP, 205 NFSV2PROC_NOOP, 206 }; 207 208 /* 209 * Maps errno values to nfs error numbers. 210 * Use NFSERR_IO as the catch all for ones not specifically defined in 211 * RFC 1094. 212 */ 213 static const u_char nfsrv_v2errmap[ELAST] = { 214 NFSERR_PERM, NFSERR_NOENT, NFSERR_IO, NFSERR_IO, NFSERR_IO, 215 NFSERR_NXIO, NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, 216 NFSERR_IO, NFSERR_IO, NFSERR_ACCES, NFSERR_IO, NFSERR_IO, 217 NFSERR_IO, NFSERR_EXIST, NFSERR_IO, NFSERR_NODEV, NFSERR_NOTDIR, 218 NFSERR_ISDIR, NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, 219 NFSERR_IO, NFSERR_FBIG, NFSERR_NOSPC, NFSERR_IO, NFSERR_ROFS, 220 NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, 221 NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, 222 NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, 223 NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, 224 NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, 225 NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, 226 NFSERR_IO, NFSERR_IO, NFSERR_NAMETOL, NFSERR_IO, NFSERR_IO, 227 NFSERR_NOTEMPTY, NFSERR_IO, NFSERR_IO, NFSERR_DQUOT, NFSERR_STALE, 228 NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, 229 NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, NFSERR_IO, 230 NFSERR_IO, NFSERR_IO, 231 }; 232 233 /* 234 * Maps errno values to nfs error numbers. 235 * Although it is not obvious whether or not NFS clients really care if 236 * a returned error value is in the specified list for the procedure, the 237 * safest thing to do is filter them appropriately. For Version 2, the 238 * X/Open XNFS document is the only specification that defines error values 239 * for each RPC (The RFC simply lists all possible error values for all RPCs), 240 * so I have decided to not do this for Version 2. 241 * The first entry is the default error return and the rest are the valid 242 * errors for that RPC in increasing numeric order. 243 */ 244 static const short nfsv3err_null[] = { 245 0, 246 0, 247 }; 248 249 static const short nfsv3err_getattr[] = { 250 NFSERR_IO, 251 NFSERR_IO, 252 NFSERR_STALE, 253 NFSERR_BADHANDLE, 254 NFSERR_SERVERFAULT, 255 0, 256 }; 257 258 static const short nfsv3err_setattr[] = { 259 NFSERR_IO, 260 NFSERR_PERM, 261 NFSERR_IO, 262 NFSERR_ACCES, 263 NFSERR_INVAL, 264 NFSERR_NOSPC, 265 NFSERR_ROFS, 266 NFSERR_DQUOT, 267 NFSERR_STALE, 268 NFSERR_BADHANDLE, 269 NFSERR_NOT_SYNC, 270 NFSERR_SERVERFAULT, 271 0, 272 }; 273 274 static const short nfsv3err_lookup[] = { 275 NFSERR_IO, 276 NFSERR_NOENT, 277 NFSERR_IO, 278 NFSERR_ACCES, 279 NFSERR_NOTDIR, 280 NFSERR_NAMETOL, 281 NFSERR_STALE, 282 NFSERR_BADHANDLE, 283 NFSERR_SERVERFAULT, 284 0, 285 }; 286 287 static const short nfsv3err_access[] = { 288 NFSERR_IO, 289 NFSERR_IO, 290 NFSERR_STALE, 291 NFSERR_BADHANDLE, 292 NFSERR_SERVERFAULT, 293 0, 294 }; 295 296 static const short nfsv3err_readlink[] = { 297 NFSERR_IO, 298 NFSERR_IO, 299 NFSERR_ACCES, 300 NFSERR_INVAL, 301 NFSERR_STALE, 302 NFSERR_BADHANDLE, 303 NFSERR_NOTSUPP, 304 NFSERR_SERVERFAULT, 305 0, 306 }; 307 308 static const short nfsv3err_read[] = { 309 NFSERR_IO, 310 NFSERR_IO, 311 NFSERR_NXIO, 312 NFSERR_ACCES, 313 NFSERR_INVAL, 314 NFSERR_STALE, 315 NFSERR_BADHANDLE, 316 NFSERR_SERVERFAULT, 317 NFSERR_JUKEBOX, 318 0, 319 }; 320 321 static const short nfsv3err_write[] = { 322 NFSERR_IO, 323 NFSERR_IO, 324 NFSERR_ACCES, 325 NFSERR_INVAL, 326 NFSERR_FBIG, 327 NFSERR_NOSPC, 328 NFSERR_ROFS, 329 NFSERR_DQUOT, 330 NFSERR_STALE, 331 NFSERR_BADHANDLE, 332 NFSERR_SERVERFAULT, 333 NFSERR_JUKEBOX, 334 0, 335 }; 336 337 static const short nfsv3err_create[] = { 338 NFSERR_IO, 339 NFSERR_IO, 340 NFSERR_ACCES, 341 NFSERR_EXIST, 342 NFSERR_NOTDIR, 343 NFSERR_NOSPC, 344 NFSERR_ROFS, 345 NFSERR_NAMETOL, 346 NFSERR_DQUOT, 347 NFSERR_STALE, 348 NFSERR_BADHANDLE, 349 NFSERR_NOTSUPP, 350 NFSERR_SERVERFAULT, 351 0, 352 }; 353 354 static const short nfsv3err_mkdir[] = { 355 NFSERR_IO, 356 NFSERR_IO, 357 NFSERR_ACCES, 358 NFSERR_EXIST, 359 NFSERR_NOTDIR, 360 NFSERR_NOSPC, 361 NFSERR_ROFS, 362 NFSERR_NAMETOL, 363 NFSERR_DQUOT, 364 NFSERR_STALE, 365 NFSERR_BADHANDLE, 366 NFSERR_NOTSUPP, 367 NFSERR_SERVERFAULT, 368 0, 369 }; 370 371 static const short nfsv3err_symlink[] = { 372 NFSERR_IO, 373 NFSERR_IO, 374 NFSERR_ACCES, 375 NFSERR_EXIST, 376 NFSERR_NOTDIR, 377 NFSERR_NOSPC, 378 NFSERR_ROFS, 379 NFSERR_NAMETOL, 380 NFSERR_DQUOT, 381 NFSERR_STALE, 382 NFSERR_BADHANDLE, 383 NFSERR_NOTSUPP, 384 NFSERR_SERVERFAULT, 385 0, 386 }; 387 388 static const short nfsv3err_mknod[] = { 389 NFSERR_IO, 390 NFSERR_IO, 391 NFSERR_ACCES, 392 NFSERR_EXIST, 393 NFSERR_NOTDIR, 394 NFSERR_NOSPC, 395 NFSERR_ROFS, 396 NFSERR_NAMETOL, 397 NFSERR_DQUOT, 398 NFSERR_STALE, 399 NFSERR_BADHANDLE, 400 NFSERR_NOTSUPP, 401 NFSERR_SERVERFAULT, 402 NFSERR_BADTYPE, 403 0, 404 }; 405 406 static const short nfsv3err_remove[] = { 407 NFSERR_IO, 408 NFSERR_NOENT, 409 NFSERR_IO, 410 NFSERR_ACCES, 411 NFSERR_NOTDIR, 412 NFSERR_ROFS, 413 NFSERR_NAMETOL, 414 NFSERR_STALE, 415 NFSERR_BADHANDLE, 416 NFSERR_SERVERFAULT, 417 0, 418 }; 419 420 static const short nfsv3err_rmdir[] = { 421 NFSERR_IO, 422 NFSERR_NOENT, 423 NFSERR_IO, 424 NFSERR_ACCES, 425 NFSERR_EXIST, 426 NFSERR_NOTDIR, 427 NFSERR_INVAL, 428 NFSERR_ROFS, 429 NFSERR_NAMETOL, 430 NFSERR_NOTEMPTY, 431 NFSERR_STALE, 432 NFSERR_BADHANDLE, 433 NFSERR_NOTSUPP, 434 NFSERR_SERVERFAULT, 435 0, 436 }; 437 438 static const short nfsv3err_rename[] = { 439 NFSERR_IO, 440 NFSERR_NOENT, 441 NFSERR_IO, 442 NFSERR_ACCES, 443 NFSERR_EXIST, 444 NFSERR_XDEV, 445 NFSERR_NOTDIR, 446 NFSERR_ISDIR, 447 NFSERR_INVAL, 448 NFSERR_NOSPC, 449 NFSERR_ROFS, 450 NFSERR_MLINK, 451 NFSERR_NAMETOL, 452 NFSERR_NOTEMPTY, 453 NFSERR_DQUOT, 454 NFSERR_STALE, 455 NFSERR_BADHANDLE, 456 NFSERR_NOTSUPP, 457 NFSERR_SERVERFAULT, 458 0, 459 }; 460 461 static const short nfsv3err_link[] = { 462 NFSERR_IO, 463 NFSERR_IO, 464 NFSERR_ACCES, 465 NFSERR_EXIST, 466 NFSERR_XDEV, 467 NFSERR_NOTDIR, 468 NFSERR_INVAL, 469 NFSERR_NOSPC, 470 NFSERR_ROFS, 471 NFSERR_MLINK, 472 NFSERR_NAMETOL, 473 NFSERR_DQUOT, 474 NFSERR_STALE, 475 NFSERR_BADHANDLE, 476 NFSERR_NOTSUPP, 477 NFSERR_SERVERFAULT, 478 0, 479 }; 480 481 static const short nfsv3err_readdir[] = { 482 NFSERR_IO, 483 NFSERR_IO, 484 NFSERR_ACCES, 485 NFSERR_NOTDIR, 486 NFSERR_STALE, 487 NFSERR_BADHANDLE, 488 NFSERR_BAD_COOKIE, 489 NFSERR_TOOSMALL, 490 NFSERR_SERVERFAULT, 491 0, 492 }; 493 494 static const short nfsv3err_readdirplus[] = { 495 NFSERR_IO, 496 NFSERR_IO, 497 NFSERR_ACCES, 498 NFSERR_NOTDIR, 499 NFSERR_STALE, 500 NFSERR_BADHANDLE, 501 NFSERR_BAD_COOKIE, 502 NFSERR_NOTSUPP, 503 NFSERR_TOOSMALL, 504 NFSERR_SERVERFAULT, 505 0, 506 }; 507 508 static const short nfsv3err_fsstat[] = { 509 NFSERR_IO, 510 NFSERR_IO, 511 NFSERR_STALE, 512 NFSERR_BADHANDLE, 513 NFSERR_SERVERFAULT, 514 0, 515 }; 516 517 static const short nfsv3err_fsinfo[] = { 518 NFSERR_STALE, 519 NFSERR_STALE, 520 NFSERR_BADHANDLE, 521 NFSERR_SERVERFAULT, 522 0, 523 }; 524 525 static const short nfsv3err_pathconf[] = { 526 NFSERR_STALE, 527 NFSERR_STALE, 528 NFSERR_BADHANDLE, 529 NFSERR_SERVERFAULT, 530 0, 531 }; 532 533 static const short nfsv3err_commit[] = { 534 NFSERR_IO, 535 NFSERR_IO, 536 NFSERR_STALE, 537 NFSERR_BADHANDLE, 538 NFSERR_SERVERFAULT, 539 0, 540 }; 541 542 static const short * const nfsrv_v3errmap[] = { 543 nfsv3err_null, 544 nfsv3err_getattr, 545 nfsv3err_setattr, 546 nfsv3err_lookup, 547 nfsv3err_access, 548 nfsv3err_readlink, 549 nfsv3err_read, 550 nfsv3err_write, 551 nfsv3err_create, 552 nfsv3err_mkdir, 553 nfsv3err_symlink, 554 nfsv3err_mknod, 555 nfsv3err_remove, 556 nfsv3err_rmdir, 557 nfsv3err_rename, 558 nfsv3err_link, 559 nfsv3err_readdir, 560 nfsv3err_readdirplus, 561 nfsv3err_fsstat, 562 nfsv3err_fsinfo, 563 nfsv3err_pathconf, 564 nfsv3err_commit, 565 }; 566 567 extern struct nfsrtt nfsrtt; 568 extern struct nfsnodehashhead *nfsnodehashtbl; 569 extern u_long nfsnodehash; 570 571 u_long nfsdirhashmask; 572 573 static kmutex_t nfs_xidlock; 574 575 int nfs_webnamei __P((struct nameidata *, struct vnode *, struct proc *)); 576 577 /* 578 * Create the header for an rpc request packet 579 * The hsiz is the size of the rest of the nfs request header. 580 * (just used to decide if a cluster is a good idea) 581 */ 582 struct mbuf * 583 nfsm_reqh(struct nfsnode *np, u_long procid, int hsiz, char **bposp) 584 { 585 struct mbuf *mb; 586 char *bpos; 587 588 mb = m_get(M_WAIT, MT_DATA); 589 MCLAIM(mb, &nfs_mowner); 590 if (hsiz >= MINCLSIZE) 591 m_clget(mb, M_WAIT); 592 mb->m_len = 0; 593 bpos = mtod(mb, void *); 594 595 /* Finally, return values */ 596 *bposp = bpos; 597 return (mb); 598 } 599 600 /* 601 * Build the RPC header and fill in the authorization info. 602 * The authorization string argument is only used when the credentials 603 * come from outside of the kernel. 604 * Returns the head of the mbuf list. 605 */ 606 struct mbuf * 607 nfsm_rpchead(cr, nmflag, procid, auth_type, auth_len, auth_str, verf_len, 608 verf_str, mrest, mrest_len, mbp, xidp) 609 kauth_cred_t cr; 610 int nmflag; 611 int procid; 612 int auth_type; 613 int auth_len; 614 char *auth_str; 615 int verf_len; 616 char *verf_str; 617 struct mbuf *mrest; 618 int mrest_len; 619 struct mbuf **mbp; 620 u_int32_t *xidp; 621 { 622 struct mbuf *mb; 623 u_int32_t *tl; 624 char *bpos; 625 int i; 626 struct mbuf *mreq; 627 int siz, grpsiz, authsiz; 628 629 authsiz = nfsm_rndup(auth_len); 630 mb = m_gethdr(M_WAIT, MT_DATA); 631 MCLAIM(mb, &nfs_mowner); 632 if ((authsiz + 10 * NFSX_UNSIGNED) >= MINCLSIZE) { 633 m_clget(mb, M_WAIT); 634 } else if ((authsiz + 10 * NFSX_UNSIGNED) < MHLEN) { 635 MH_ALIGN(mb, authsiz + 10 * NFSX_UNSIGNED); 636 } else { 637 MH_ALIGN(mb, 8 * NFSX_UNSIGNED); 638 } 639 mb->m_len = 0; 640 mreq = mb; 641 bpos = mtod(mb, void *); 642 643 /* 644 * First the RPC header. 645 */ 646 nfsm_build(tl, u_int32_t *, 8 * NFSX_UNSIGNED); 647 648 *tl++ = *xidp = nfs_getxid(); 649 *tl++ = rpc_call; 650 *tl++ = rpc_vers; 651 *tl++ = txdr_unsigned(NFS_PROG); 652 if (nmflag & NFSMNT_NFSV3) 653 *tl++ = txdr_unsigned(NFS_VER3); 654 else 655 *tl++ = txdr_unsigned(NFS_VER2); 656 if (nmflag & NFSMNT_NFSV3) 657 *tl++ = txdr_unsigned(procid); 658 else 659 *tl++ = txdr_unsigned(nfsv2_procid[procid]); 660 661 /* 662 * And then the authorization cred. 663 */ 664 *tl++ = txdr_unsigned(auth_type); 665 *tl = txdr_unsigned(authsiz); 666 switch (auth_type) { 667 case RPCAUTH_UNIX: 668 nfsm_build(tl, u_int32_t *, auth_len); 669 *tl++ = 0; /* stamp ?? */ 670 *tl++ = 0; /* NULL hostname */ 671 *tl++ = txdr_unsigned(kauth_cred_geteuid(cr)); 672 *tl++ = txdr_unsigned(kauth_cred_getegid(cr)); 673 grpsiz = (auth_len >> 2) - 5; 674 *tl++ = txdr_unsigned(grpsiz); 675 for (i = 0; i < grpsiz; i++) 676 *tl++ = txdr_unsigned(kauth_cred_group(cr, i)); /* XXX elad review */ 677 break; 678 case RPCAUTH_KERB4: 679 siz = auth_len; 680 while (siz > 0) { 681 if (M_TRAILINGSPACE(mb) == 0) { 682 struct mbuf *mb2; 683 mb2 = m_get(M_WAIT, MT_DATA); 684 MCLAIM(mb2, &nfs_mowner); 685 if (siz >= MINCLSIZE) 686 m_clget(mb2, M_WAIT); 687 mb->m_next = mb2; 688 mb = mb2; 689 mb->m_len = 0; 690 bpos = mtod(mb, void *); 691 } 692 i = min(siz, M_TRAILINGSPACE(mb)); 693 memcpy(bpos, auth_str, i); 694 mb->m_len += i; 695 auth_str += i; 696 bpos += i; 697 siz -= i; 698 } 699 if ((siz = (nfsm_rndup(auth_len) - auth_len)) > 0) { 700 for (i = 0; i < siz; i++) 701 *bpos++ = '\0'; 702 mb->m_len += siz; 703 } 704 break; 705 }; 706 707 /* 708 * And the verifier... 709 */ 710 nfsm_build(tl, u_int32_t *, 2 * NFSX_UNSIGNED); 711 if (verf_str) { 712 *tl++ = txdr_unsigned(RPCAUTH_KERB4); 713 *tl = txdr_unsigned(verf_len); 714 siz = verf_len; 715 while (siz > 0) { 716 if (M_TRAILINGSPACE(mb) == 0) { 717 struct mbuf *mb2; 718 mb2 = m_get(M_WAIT, MT_DATA); 719 MCLAIM(mb2, &nfs_mowner); 720 if (siz >= MINCLSIZE) 721 m_clget(mb2, M_WAIT); 722 mb->m_next = mb2; 723 mb = mb2; 724 mb->m_len = 0; 725 bpos = mtod(mb, void *); 726 } 727 i = min(siz, M_TRAILINGSPACE(mb)); 728 memcpy(bpos, verf_str, i); 729 mb->m_len += i; 730 verf_str += i; 731 bpos += i; 732 siz -= i; 733 } 734 if ((siz = (nfsm_rndup(verf_len) - verf_len)) > 0) { 735 for (i = 0; i < siz; i++) 736 *bpos++ = '\0'; 737 mb->m_len += siz; 738 } 739 } else { 740 *tl++ = txdr_unsigned(RPCAUTH_NULL); 741 *tl = 0; 742 } 743 mb->m_next = mrest; 744 mreq->m_pkthdr.len = authsiz + 10 * NFSX_UNSIGNED + mrest_len; 745 mreq->m_pkthdr.rcvif = (struct ifnet *)0; 746 *mbp = mb; 747 return (mreq); 748 } 749 750 /* 751 * copies mbuf chain to the uio scatter/gather list 752 */ 753 int 754 nfsm_mbuftouio(mrep, uiop, siz, dpos) 755 struct mbuf **mrep; 756 struct uio *uiop; 757 int siz; 758 char **dpos; 759 { 760 char *mbufcp, *uiocp; 761 int xfer, left, len; 762 struct mbuf *mp; 763 long uiosiz, rem; 764 int error = 0; 765 766 mp = *mrep; 767 mbufcp = *dpos; 768 len = mtod(mp, char *) + mp->m_len - mbufcp; 769 rem = nfsm_rndup(siz)-siz; 770 while (siz > 0) { 771 if (uiop->uio_iovcnt <= 0 || uiop->uio_iov == NULL) 772 return (EFBIG); 773 left = uiop->uio_iov->iov_len; 774 uiocp = uiop->uio_iov->iov_base; 775 if (left > siz) 776 left = siz; 777 uiosiz = left; 778 while (left > 0) { 779 while (len == 0) { 780 mp = mp->m_next; 781 if (mp == NULL) 782 return (EBADRPC); 783 mbufcp = mtod(mp, void *); 784 len = mp->m_len; 785 } 786 xfer = (left > len) ? len : left; 787 error = copyout_vmspace(uiop->uio_vmspace, mbufcp, 788 uiocp, xfer); 789 if (error) { 790 return error; 791 } 792 left -= xfer; 793 len -= xfer; 794 mbufcp += xfer; 795 uiocp += xfer; 796 uiop->uio_offset += xfer; 797 uiop->uio_resid -= xfer; 798 } 799 if (uiop->uio_iov->iov_len <= siz) { 800 uiop->uio_iovcnt--; 801 uiop->uio_iov++; 802 } else { 803 uiop->uio_iov->iov_base = 804 (char *)uiop->uio_iov->iov_base + uiosiz; 805 uiop->uio_iov->iov_len -= uiosiz; 806 } 807 siz -= uiosiz; 808 } 809 *dpos = mbufcp; 810 *mrep = mp; 811 if (rem > 0) { 812 if (len < rem) 813 error = nfs_adv(mrep, dpos, rem, len); 814 else 815 *dpos += rem; 816 } 817 return (error); 818 } 819 820 /* 821 * copies a uio scatter/gather list to an mbuf chain. 822 * NOTE: can ony handle iovcnt == 1 823 */ 824 int 825 nfsm_uiotombuf(uiop, mq, siz, bpos) 826 struct uio *uiop; 827 struct mbuf **mq; 828 int siz; 829 char **bpos; 830 { 831 char *uiocp; 832 struct mbuf *mp, *mp2; 833 int xfer, left, mlen; 834 int uiosiz, clflg, rem; 835 char *cp; 836 int error; 837 838 #ifdef DIAGNOSTIC 839 if (uiop->uio_iovcnt != 1) 840 panic("nfsm_uiotombuf: iovcnt != 1"); 841 #endif 842 843 if (siz > MLEN) /* or should it >= MCLBYTES ?? */ 844 clflg = 1; 845 else 846 clflg = 0; 847 rem = nfsm_rndup(siz)-siz; 848 mp = mp2 = *mq; 849 while (siz > 0) { 850 left = uiop->uio_iov->iov_len; 851 uiocp = uiop->uio_iov->iov_base; 852 if (left > siz) 853 left = siz; 854 uiosiz = left; 855 while (left > 0) { 856 mlen = M_TRAILINGSPACE(mp); 857 if (mlen == 0) { 858 mp = m_get(M_WAIT, MT_DATA); 859 MCLAIM(mp, &nfs_mowner); 860 if (clflg) 861 m_clget(mp, M_WAIT); 862 mp->m_len = 0; 863 mp2->m_next = mp; 864 mp2 = mp; 865 mlen = M_TRAILINGSPACE(mp); 866 } 867 xfer = (left > mlen) ? mlen : left; 868 cp = mtod(mp, char *) + mp->m_len; 869 error = copyin_vmspace(uiop->uio_vmspace, uiocp, cp, 870 xfer); 871 if (error) { 872 /* XXX */ 873 } 874 mp->m_len += xfer; 875 left -= xfer; 876 uiocp += xfer; 877 uiop->uio_offset += xfer; 878 uiop->uio_resid -= xfer; 879 } 880 uiop->uio_iov->iov_base = (char *)uiop->uio_iov->iov_base + 881 uiosiz; 882 uiop->uio_iov->iov_len -= uiosiz; 883 siz -= uiosiz; 884 } 885 if (rem > 0) { 886 if (rem > M_TRAILINGSPACE(mp)) { 887 mp = m_get(M_WAIT, MT_DATA); 888 MCLAIM(mp, &nfs_mowner); 889 mp->m_len = 0; 890 mp2->m_next = mp; 891 } 892 cp = mtod(mp, char *) + mp->m_len; 893 for (left = 0; left < rem; left++) 894 *cp++ = '\0'; 895 mp->m_len += rem; 896 *bpos = cp; 897 } else 898 *bpos = mtod(mp, char *) + mp->m_len; 899 *mq = mp; 900 return (0); 901 } 902 903 /* 904 * Get at least "siz" bytes of correctly aligned data. 905 * When called the mbuf pointers are not necessarily correct, 906 * dsosp points to what ought to be in m_data and left contains 907 * what ought to be in m_len. 908 * This is used by the macros nfsm_dissect and nfsm_dissecton for tough 909 * cases. (The macros use the vars. dpos and dpos2) 910 */ 911 int 912 nfsm_disct(mdp, dposp, siz, left, cp2) 913 struct mbuf **mdp; 914 char **dposp; 915 int siz; 916 int left; 917 char **cp2; 918 { 919 struct mbuf *m1, *m2; 920 struct mbuf *havebuf = NULL; 921 char *src = *dposp; 922 char *dst; 923 int len; 924 925 #ifdef DEBUG 926 if (left < 0) 927 panic("nfsm_disct: left < 0"); 928 #endif 929 m1 = *mdp; 930 /* 931 * Skip through the mbuf chain looking for an mbuf with 932 * some data. If the first mbuf found has enough data 933 * and it is correctly aligned return it. 934 */ 935 while (left == 0) { 936 havebuf = m1; 937 *mdp = m1 = m1->m_next; 938 if (m1 == NULL) 939 return (EBADRPC); 940 src = mtod(m1, void *); 941 left = m1->m_len; 942 /* 943 * If we start a new mbuf and it is big enough 944 * and correctly aligned just return it, don't 945 * do any pull up. 946 */ 947 if (left >= siz && nfsm_aligned(src)) { 948 *cp2 = src; 949 *dposp = src + siz; 950 return (0); 951 } 952 } 953 if ((m1->m_flags & M_EXT) != 0) { 954 if (havebuf && M_TRAILINGSPACE(havebuf) >= siz && 955 nfsm_aligned(mtod(havebuf, char *) + havebuf->m_len)) { 956 /* 957 * If the first mbuf with data has external data 958 * and there is a previous mbuf with some trailing 959 * space, use it to move the data into. 960 */ 961 m2 = m1; 962 *mdp = m1 = havebuf; 963 *cp2 = mtod(m1, char *) + m1->m_len; 964 } else if (havebuf) { 965 /* 966 * If the first mbuf has a external data 967 * and there is no previous empty mbuf 968 * allocate a new mbuf and move the external 969 * data to the new mbuf. Also make the first 970 * mbuf look empty. 971 */ 972 m2 = m1; 973 *mdp = m1 = m_get(M_WAIT, MT_DATA); 974 MCLAIM(m1, m2->m_owner); 975 if ((m2->m_flags & M_PKTHDR) != 0) { 976 /* XXX MOVE */ 977 M_COPY_PKTHDR(m1, m2); 978 m_tag_delete_chain(m2, NULL); 979 m2->m_flags &= ~M_PKTHDR; 980 } 981 if (havebuf) { 982 havebuf->m_next = m1; 983 } 984 m1->m_next = m2; 985 MRESETDATA(m1); 986 m1->m_len = 0; 987 m2->m_data = src; 988 m2->m_len = left; 989 *cp2 = mtod(m1, char *); 990 } else { 991 struct mbuf **nextp = &m1->m_next; 992 993 m1->m_len -= left; 994 do { 995 m2 = m_get(M_WAIT, MT_DATA); 996 MCLAIM(m2, m1->m_owner); 997 if (left >= MINCLSIZE) { 998 MCLGET(m2, M_WAIT); 999 } 1000 m2->m_next = *nextp; 1001 *nextp = m2; 1002 nextp = &m2->m_next; 1003 len = (m2->m_flags & M_EXT) != 0 ? 1004 MCLBYTES : MLEN; 1005 if (len > left) { 1006 len = left; 1007 } 1008 memcpy(mtod(m2, char *), src, len); 1009 m2->m_len = len; 1010 src += len; 1011 left -= len; 1012 } while (left > 0); 1013 *mdp = m1 = m1->m_next; 1014 m2 = m1->m_next; 1015 *cp2 = mtod(m1, char *); 1016 } 1017 } else { 1018 /* 1019 * If the first mbuf has no external data 1020 * move the data to the front of the mbuf. 1021 */ 1022 MRESETDATA(m1); 1023 dst = mtod(m1, char *); 1024 if (dst != src) { 1025 memmove(dst, src, left); 1026 } 1027 m1->m_len = left; 1028 m2 = m1->m_next; 1029 *cp2 = m1->m_data; 1030 } 1031 *dposp = *cp2 + siz; 1032 /* 1033 * Loop through mbufs pulling data up into first mbuf until 1034 * the first mbuf is full or there is no more data to 1035 * pullup. 1036 */ 1037 dst = mtod(m1, char *) + m1->m_len; 1038 while ((len = M_TRAILINGSPACE(m1)) != 0 && m2) { 1039 if ((len = min(len, m2->m_len)) != 0) { 1040 memcpy(dst, mtod(m2, char *), len); 1041 } 1042 m1->m_len += len; 1043 dst += len; 1044 m2->m_data += len; 1045 m2->m_len -= len; 1046 m2 = m2->m_next; 1047 } 1048 if (m1->m_len < siz) 1049 return (EBADRPC); 1050 return (0); 1051 } 1052 1053 /* 1054 * Advance the position in the mbuf chain. 1055 */ 1056 int 1057 nfs_adv(mdp, dposp, offs, left) 1058 struct mbuf **mdp; 1059 char **dposp; 1060 int offs; 1061 int left; 1062 { 1063 struct mbuf *m; 1064 int s; 1065 1066 m = *mdp; 1067 s = left; 1068 while (s < offs) { 1069 offs -= s; 1070 m = m->m_next; 1071 if (m == NULL) 1072 return (EBADRPC); 1073 s = m->m_len; 1074 } 1075 *mdp = m; 1076 *dposp = mtod(m, char *) + offs; 1077 return (0); 1078 } 1079 1080 /* 1081 * Copy a string into mbufs for the hard cases... 1082 */ 1083 int 1084 nfsm_strtmbuf(mb, bpos, cp, siz) 1085 struct mbuf **mb; 1086 char **bpos; 1087 const char *cp; 1088 long siz; 1089 { 1090 struct mbuf *m1 = NULL, *m2; 1091 long left, xfer, len, tlen; 1092 u_int32_t *tl; 1093 int putsize; 1094 1095 putsize = 1; 1096 m2 = *mb; 1097 left = M_TRAILINGSPACE(m2); 1098 if (left > 0) { 1099 tl = ((u_int32_t *)(*bpos)); 1100 *tl++ = txdr_unsigned(siz); 1101 putsize = 0; 1102 left -= NFSX_UNSIGNED; 1103 m2->m_len += NFSX_UNSIGNED; 1104 if (left > 0) { 1105 memcpy((void *) tl, cp, left); 1106 siz -= left; 1107 cp += left; 1108 m2->m_len += left; 1109 left = 0; 1110 } 1111 } 1112 /* Loop around adding mbufs */ 1113 while (siz > 0) { 1114 m1 = m_get(M_WAIT, MT_DATA); 1115 MCLAIM(m1, &nfs_mowner); 1116 if (siz > MLEN) 1117 m_clget(m1, M_WAIT); 1118 m1->m_len = NFSMSIZ(m1); 1119 m2->m_next = m1; 1120 m2 = m1; 1121 tl = mtod(m1, u_int32_t *); 1122 tlen = 0; 1123 if (putsize) { 1124 *tl++ = txdr_unsigned(siz); 1125 m1->m_len -= NFSX_UNSIGNED; 1126 tlen = NFSX_UNSIGNED; 1127 putsize = 0; 1128 } 1129 if (siz < m1->m_len) { 1130 len = nfsm_rndup(siz); 1131 xfer = siz; 1132 if (xfer < len) 1133 *(tl+(xfer>>2)) = 0; 1134 } else { 1135 xfer = len = m1->m_len; 1136 } 1137 memcpy((void *) tl, cp, xfer); 1138 m1->m_len = len+tlen; 1139 siz -= xfer; 1140 cp += xfer; 1141 } 1142 *mb = m1; 1143 *bpos = mtod(m1, char *) + m1->m_len; 1144 return (0); 1145 } 1146 1147 /* 1148 * Directory caching routines. They work as follows: 1149 * - a cache is maintained per VDIR nfsnode. 1150 * - for each offset cookie that is exported to userspace, and can 1151 * thus be thrown back at us as an offset to VOP_READDIR, store 1152 * information in the cache. 1153 * - cached are: 1154 * - cookie itself 1155 * - blocknumber (essentially just a search key in the buffer cache) 1156 * - entry number in block. 1157 * - offset cookie of block in which this entry is stored 1158 * - 32 bit cookie if NFSMNT_XLATECOOKIE is used. 1159 * - entries are looked up in a hash table 1160 * - also maintained is an LRU list of entries, used to determine 1161 * which ones to delete if the cache grows too large. 1162 * - if 32 <-> 64 translation mode is requested for a filesystem, 1163 * the cache also functions as a translation table 1164 * - in the translation case, invalidating the cache does not mean 1165 * flushing it, but just marking entries as invalid, except for 1166 * the <64bit cookie, 32bitcookie> pair which is still valid, to 1167 * still be able to use the cache as a translation table. 1168 * - 32 bit cookies are uniquely created by combining the hash table 1169 * entry value, and one generation count per hash table entry, 1170 * incremented each time an entry is appended to the chain. 1171 * - the cache is invalidated each time a direcory is modified 1172 * - sanity checks are also done; if an entry in a block turns 1173 * out not to have a matching cookie, the cache is invalidated 1174 * and a new block starting from the wanted offset is fetched from 1175 * the server. 1176 * - directory entries as read from the server are extended to contain 1177 * the 64bit and, optionally, the 32bit cookies, for sanity checking 1178 * the cache and exporting them to userspace through the cookie 1179 * argument to VOP_READDIR. 1180 */ 1181 1182 u_long 1183 nfs_dirhash(off) 1184 off_t off; 1185 { 1186 int i; 1187 char *cp = (char *)&off; 1188 u_long sum = 0L; 1189 1190 for (i = 0 ; i < sizeof (off); i++) 1191 sum += *cp++; 1192 1193 return sum; 1194 } 1195 1196 #define _NFSDC_MTX(np) (&NFSTOV(np)->v_interlock) 1197 #define NFSDC_LOCK(np) mutex_enter(_NFSDC_MTX(np)) 1198 #define NFSDC_UNLOCK(np) mutex_exit(_NFSDC_MTX(np)) 1199 #define NFSDC_ASSERT_LOCKED(np) KASSERT(mutex_owned(_NFSDC_MTX(np))) 1200 1201 void 1202 nfs_initdircache(vp) 1203 struct vnode *vp; 1204 { 1205 struct nfsnode *np = VTONFS(vp); 1206 struct nfsdirhashhead *dircache; 1207 1208 dircache = hashinit(NFS_DIRHASHSIZ, HASH_LIST, M_NFSDIROFF, 1209 M_WAITOK, &nfsdirhashmask); 1210 1211 NFSDC_LOCK(np); 1212 if (np->n_dircache == NULL) { 1213 np->n_dircachesize = 0; 1214 np->n_dircache = dircache; 1215 dircache = NULL; 1216 TAILQ_INIT(&np->n_dirchain); 1217 } 1218 NFSDC_UNLOCK(np); 1219 if (dircache) 1220 hashdone(dircache, M_NFSDIROFF); 1221 } 1222 1223 void 1224 nfs_initdirxlatecookie(vp) 1225 struct vnode *vp; 1226 { 1227 struct nfsnode *np = VTONFS(vp); 1228 unsigned *dirgens; 1229 1230 KASSERT(VFSTONFS(vp->v_mount)->nm_flag & NFSMNT_XLATECOOKIE); 1231 1232 dirgens = kmem_zalloc(NFS_DIRHASHSIZ * sizeof(unsigned), KM_SLEEP); 1233 NFSDC_LOCK(np); 1234 if (np->n_dirgens == NULL) { 1235 np->n_dirgens = dirgens; 1236 dirgens = NULL; 1237 } 1238 NFSDC_UNLOCK(np); 1239 if (dirgens) 1240 kmem_free(dirgens, NFS_DIRHASHSIZ * sizeof(unsigned)); 1241 } 1242 1243 static const struct nfsdircache dzero; 1244 1245 static void nfs_unlinkdircache __P((struct nfsnode *np, struct nfsdircache *)); 1246 static void nfs_putdircache_unlocked __P((struct nfsnode *, 1247 struct nfsdircache *)); 1248 1249 static void 1250 nfs_unlinkdircache(np, ndp) 1251 struct nfsnode *np; 1252 struct nfsdircache *ndp; 1253 { 1254 1255 NFSDC_ASSERT_LOCKED(np); 1256 KASSERT(ndp != &dzero); 1257 1258 if (LIST_NEXT(ndp, dc_hash) == (void *)-1) 1259 return; 1260 1261 TAILQ_REMOVE(&np->n_dirchain, ndp, dc_chain); 1262 LIST_REMOVE(ndp, dc_hash); 1263 LIST_NEXT(ndp, dc_hash) = (void *)-1; /* mark as unlinked */ 1264 1265 nfs_putdircache_unlocked(np, ndp); 1266 } 1267 1268 void 1269 nfs_putdircache(np, ndp) 1270 struct nfsnode *np; 1271 struct nfsdircache *ndp; 1272 { 1273 int ref; 1274 1275 if (ndp == &dzero) 1276 return; 1277 1278 KASSERT(ndp->dc_refcnt > 0); 1279 NFSDC_LOCK(np); 1280 ref = --ndp->dc_refcnt; 1281 NFSDC_UNLOCK(np); 1282 1283 if (ref == 0) 1284 kmem_free(ndp, sizeof(*ndp)); 1285 } 1286 1287 static void 1288 nfs_putdircache_unlocked(struct nfsnode *np, struct nfsdircache *ndp) 1289 { 1290 int ref; 1291 1292 NFSDC_ASSERT_LOCKED(np); 1293 1294 if (ndp == &dzero) 1295 return; 1296 1297 KASSERT(ndp->dc_refcnt > 0); 1298 ref = --ndp->dc_refcnt; 1299 if (ref == 0) 1300 kmem_free(ndp, sizeof(*ndp)); 1301 } 1302 1303 struct nfsdircache * 1304 nfs_searchdircache(vp, off, do32, hashent) 1305 struct vnode *vp; 1306 off_t off; 1307 int do32; 1308 int *hashent; 1309 { 1310 struct nfsdirhashhead *ndhp; 1311 struct nfsdircache *ndp = NULL; 1312 struct nfsnode *np = VTONFS(vp); 1313 unsigned ent; 1314 1315 /* 1316 * Zero is always a valid cookie. 1317 */ 1318 if (off == 0) 1319 /* XXXUNCONST */ 1320 return (struct nfsdircache *)__UNCONST(&dzero); 1321 1322 if (!np->n_dircache) 1323 return NULL; 1324 1325 /* 1326 * We use a 32bit cookie as search key, directly reconstruct 1327 * the hashentry. Else use the hashfunction. 1328 */ 1329 if (do32) { 1330 ent = (u_int32_t)off >> 24; 1331 if (ent >= NFS_DIRHASHSIZ) 1332 return NULL; 1333 ndhp = &np->n_dircache[ent]; 1334 } else { 1335 ndhp = NFSDIRHASH(np, off); 1336 } 1337 1338 if (hashent) 1339 *hashent = (int)(ndhp - np->n_dircache); 1340 1341 NFSDC_LOCK(np); 1342 if (do32) { 1343 LIST_FOREACH(ndp, ndhp, dc_hash) { 1344 if (ndp->dc_cookie32 == (u_int32_t)off) { 1345 /* 1346 * An invalidated entry will become the 1347 * start of a new block fetched from 1348 * the server. 1349 */ 1350 if (ndp->dc_flags & NFSDC_INVALID) { 1351 ndp->dc_blkcookie = ndp->dc_cookie; 1352 ndp->dc_entry = 0; 1353 ndp->dc_flags &= ~NFSDC_INVALID; 1354 } 1355 break; 1356 } 1357 } 1358 } else { 1359 LIST_FOREACH(ndp, ndhp, dc_hash) { 1360 if (ndp->dc_cookie == off) 1361 break; 1362 } 1363 } 1364 if (ndp != NULL) 1365 ndp->dc_refcnt++; 1366 NFSDC_UNLOCK(np); 1367 return ndp; 1368 } 1369 1370 1371 struct nfsdircache * 1372 nfs_enterdircache(struct vnode *vp, off_t off, off_t blkoff, int en, 1373 daddr_t blkno) 1374 { 1375 struct nfsnode *np = VTONFS(vp); 1376 struct nfsdirhashhead *ndhp; 1377 struct nfsdircache *ndp = NULL; 1378 struct nfsdircache *newndp = NULL; 1379 struct nfsmount *nmp = VFSTONFS(vp->v_mount); 1380 int hashent = 0, gen, overwrite; /* XXX: GCC */ 1381 1382 /* 1383 * XXX refuse entries for offset 0. amd(8) erroneously sets 1384 * cookie 0 for the '.' entry, making this necessary. This 1385 * isn't so bad, as 0 is a special case anyway. 1386 */ 1387 if (off == 0) 1388 /* XXXUNCONST */ 1389 return (struct nfsdircache *)__UNCONST(&dzero); 1390 1391 if (!np->n_dircache) 1392 /* 1393 * XXX would like to do this in nfs_nget but vtype 1394 * isn't known at that time. 1395 */ 1396 nfs_initdircache(vp); 1397 1398 if ((nmp->nm_flag & NFSMNT_XLATECOOKIE) && !np->n_dirgens) 1399 nfs_initdirxlatecookie(vp); 1400 1401 retry: 1402 ndp = nfs_searchdircache(vp, off, 0, &hashent); 1403 1404 NFSDC_LOCK(np); 1405 if (ndp && (ndp->dc_flags & NFSDC_INVALID) == 0) { 1406 /* 1407 * Overwriting an old entry. Check if it's the same. 1408 * If so, just return. If not, remove the old entry. 1409 */ 1410 if (ndp->dc_blkcookie == blkoff && ndp->dc_entry == en) 1411 goto done; 1412 nfs_unlinkdircache(np, ndp); 1413 nfs_putdircache_unlocked(np, ndp); 1414 ndp = NULL; 1415 } 1416 1417 ndhp = &np->n_dircache[hashent]; 1418 1419 if (!ndp) { 1420 if (newndp == NULL) { 1421 NFSDC_UNLOCK(np); 1422 newndp = kmem_alloc(sizeof(*newndp), KM_SLEEP); 1423 newndp->dc_refcnt = 1; 1424 LIST_NEXT(newndp, dc_hash) = (void *)-1; 1425 goto retry; 1426 } 1427 ndp = newndp; 1428 newndp = NULL; 1429 overwrite = 0; 1430 if (nmp->nm_flag & NFSMNT_XLATECOOKIE) { 1431 /* 1432 * We're allocating a new entry, so bump the 1433 * generation number. 1434 */ 1435 KASSERT(np->n_dirgens); 1436 gen = ++np->n_dirgens[hashent]; 1437 if (gen == 0) { 1438 np->n_dirgens[hashent]++; 1439 gen++; 1440 } 1441 ndp->dc_cookie32 = (hashent << 24) | (gen & 0xffffff); 1442 } 1443 } else 1444 overwrite = 1; 1445 1446 ndp->dc_cookie = off; 1447 ndp->dc_blkcookie = blkoff; 1448 ndp->dc_entry = en; 1449 ndp->dc_flags = 0; 1450 1451 if (overwrite) 1452 goto done; 1453 1454 /* 1455 * If the maximum directory cookie cache size has been reached 1456 * for this node, take one off the front. The idea is that 1457 * directories are typically read front-to-back once, so that 1458 * the oldest entries can be thrown away without much performance 1459 * loss. 1460 */ 1461 if (np->n_dircachesize == NFS_MAXDIRCACHE) { 1462 nfs_unlinkdircache(np, TAILQ_FIRST(&np->n_dirchain)); 1463 } else 1464 np->n_dircachesize++; 1465 1466 KASSERT(ndp->dc_refcnt == 1); 1467 LIST_INSERT_HEAD(ndhp, ndp, dc_hash); 1468 TAILQ_INSERT_TAIL(&np->n_dirchain, ndp, dc_chain); 1469 ndp->dc_refcnt++; 1470 done: 1471 KASSERT(ndp->dc_refcnt > 0); 1472 NFSDC_UNLOCK(np); 1473 if (newndp) 1474 nfs_putdircache(np, newndp); 1475 return ndp; 1476 } 1477 1478 void 1479 nfs_invaldircache(vp, flags) 1480 struct vnode *vp; 1481 int flags; 1482 { 1483 struct nfsnode *np = VTONFS(vp); 1484 struct nfsdircache *ndp = NULL; 1485 struct nfsmount *nmp = VFSTONFS(vp->v_mount); 1486 const bool forcefree = flags & NFS_INVALDIRCACHE_FORCE; 1487 1488 #ifdef DIAGNOSTIC 1489 if (vp->v_type != VDIR) 1490 panic("nfs: invaldircache: not dir"); 1491 #endif 1492 1493 if ((flags & NFS_INVALDIRCACHE_KEEPEOF) == 0) 1494 np->n_flag &= ~NEOFVALID; 1495 1496 if (!np->n_dircache) 1497 return; 1498 1499 NFSDC_LOCK(np); 1500 if (!(nmp->nm_flag & NFSMNT_XLATECOOKIE) || forcefree) { 1501 while ((ndp = TAILQ_FIRST(&np->n_dirchain)) != NULL) { 1502 KASSERT(!forcefree || ndp->dc_refcnt == 1); 1503 nfs_unlinkdircache(np, ndp); 1504 } 1505 np->n_dircachesize = 0; 1506 if (forcefree && np->n_dirgens) { 1507 kmem_free(np->n_dirgens, 1508 NFS_DIRHASHSIZ * sizeof(unsigned)); 1509 np->n_dirgens = NULL; 1510 } 1511 } else { 1512 TAILQ_FOREACH(ndp, &np->n_dirchain, dc_chain) 1513 ndp->dc_flags |= NFSDC_INVALID; 1514 } 1515 1516 NFSDC_UNLOCK(np); 1517 } 1518 1519 /* 1520 * Called once before VFS init to initialize shared and 1521 * server-specific data structures. 1522 */ 1523 static int 1524 nfs_init0(void) 1525 { 1526 1527 nfsrtt.pos = 0; 1528 rpc_vers = txdr_unsigned(RPC_VER2); 1529 rpc_call = txdr_unsigned(RPC_CALL); 1530 rpc_reply = txdr_unsigned(RPC_REPLY); 1531 rpc_msgdenied = txdr_unsigned(RPC_MSGDENIED); 1532 rpc_msgaccepted = txdr_unsigned(RPC_MSGACCEPTED); 1533 rpc_mismatch = txdr_unsigned(RPC_MISMATCH); 1534 rpc_autherr = txdr_unsigned(RPC_AUTHERR); 1535 rpc_auth_unix = txdr_unsigned(RPCAUTH_UNIX); 1536 rpc_auth_kerb = txdr_unsigned(RPCAUTH_KERB4); 1537 nfs_prog = txdr_unsigned(NFS_PROG); 1538 nfs_true = txdr_unsigned(true); 1539 nfs_false = txdr_unsigned(false); 1540 nfs_xdrneg1 = txdr_unsigned(-1); 1541 nfs_ticks = (hz * NFS_TICKINTVL + 500) / 1000; 1542 if (nfs_ticks < 1) 1543 nfs_ticks = 1; 1544 #ifdef NFSSERVER 1545 nfsrv_init(0); /* Init server data structures */ 1546 nfsrv_initcache(); /* Init the server request cache */ 1547 { 1548 extern krwlock_t netexport_lock; /* XXX */ 1549 rw_init(&netexport_lock); 1550 } 1551 #endif /* NFSSERVER */ 1552 1553 #if defined(NFSSERVER) || (defined(NFS) && !defined(NFS_V2_ONLY)) 1554 nfsdreq_init(); 1555 #endif /* defined(NFSSERVER) || (defined(NFS) && !defined(NFS_V2_ONLY)) */ 1556 1557 /* 1558 * Initialize reply list and start timer 1559 */ 1560 mutex_init(&nfs_reqq_lock, MUTEX_DEFAULT, IPL_SOFTCLOCK); 1561 mutex_init(&nfs_xidlock, MUTEX_DEFAULT, IPL_NONE); 1562 TAILQ_INIT(&nfs_reqq); 1563 nfs_timer_init(); 1564 MOWNER_ATTACH(&nfs_mowner); 1565 1566 #ifdef NFS 1567 /* Initialize the kqueue structures */ 1568 nfs_kqinit(); 1569 /* Initialize the iod structures */ 1570 nfs_iodinit(); 1571 #endif 1572 return 0; 1573 } 1574 1575 void 1576 nfs_init(void) 1577 { 1578 static ONCE_DECL(nfs_init_once); 1579 1580 RUN_ONCE(&nfs_init_once, nfs_init0); 1581 } 1582 1583 #ifdef NFS 1584 /* 1585 * Called once at VFS init to initialize client-specific data structures. 1586 */ 1587 void 1588 nfs_vfs_init() 1589 { 1590 /* Initialize NFS server / client shared data. */ 1591 nfs_init(); 1592 1593 nfs_nhinit(); /* Init the nfsnode table */ 1594 nfs_commitsize = uvmexp.npages << (PAGE_SHIFT - 4); 1595 } 1596 1597 void 1598 nfs_vfs_reinit() 1599 { 1600 nfs_nhreinit(); 1601 } 1602 1603 void 1604 nfs_vfs_done() 1605 { 1606 nfs_nhdone(); 1607 } 1608 1609 /* 1610 * Attribute cache routines. 1611 * nfs_loadattrcache() - loads or updates the cache contents from attributes 1612 * that are on the mbuf list 1613 * nfs_getattrcache() - returns valid attributes if found in cache, returns 1614 * error otherwise 1615 */ 1616 1617 /* 1618 * Load the attribute cache (that lives in the nfsnode entry) with 1619 * the values on the mbuf list and 1620 * Iff vap not NULL 1621 * copy the attributes to *vaper 1622 */ 1623 int 1624 nfsm_loadattrcache(vpp, mdp, dposp, vaper, flags) 1625 struct vnode **vpp; 1626 struct mbuf **mdp; 1627 char **dposp; 1628 struct vattr *vaper; 1629 int flags; 1630 { 1631 int32_t t1; 1632 char *cp2; 1633 int error = 0; 1634 struct mbuf *md; 1635 int v3 = NFS_ISV3(*vpp); 1636 1637 md = *mdp; 1638 t1 = (mtod(md, char *) + md->m_len) - *dposp; 1639 error = nfsm_disct(mdp, dposp, NFSX_FATTR(v3), t1, &cp2); 1640 if (error) 1641 return (error); 1642 return nfs_loadattrcache(vpp, (struct nfs_fattr *)cp2, vaper, flags); 1643 } 1644 1645 int 1646 nfs_loadattrcache(vpp, fp, vaper, flags) 1647 struct vnode **vpp; 1648 struct nfs_fattr *fp; 1649 struct vattr *vaper; 1650 int flags; 1651 { 1652 struct vnode *vp = *vpp; 1653 struct vattr *vap; 1654 int v3 = NFS_ISV3(vp); 1655 enum vtype vtyp; 1656 u_short vmode; 1657 struct timespec mtime; 1658 struct timespec ctime; 1659 int32_t rdev; 1660 struct nfsnode *np; 1661 extern int (**spec_nfsv2nodeop_p) __P((void *)); 1662 uid_t uid; 1663 gid_t gid; 1664 1665 if (v3) { 1666 vtyp = nfsv3tov_type(fp->fa_type); 1667 vmode = fxdr_unsigned(u_short, fp->fa_mode); 1668 rdev = makedev(fxdr_unsigned(u_int32_t, fp->fa3_rdev.specdata1), 1669 fxdr_unsigned(u_int32_t, fp->fa3_rdev.specdata2)); 1670 fxdr_nfsv3time(&fp->fa3_mtime, &mtime); 1671 fxdr_nfsv3time(&fp->fa3_ctime, &ctime); 1672 } else { 1673 vtyp = nfsv2tov_type(fp->fa_type); 1674 vmode = fxdr_unsigned(u_short, fp->fa_mode); 1675 if (vtyp == VNON || vtyp == VREG) 1676 vtyp = IFTOVT(vmode); 1677 rdev = fxdr_unsigned(int32_t, fp->fa2_rdev); 1678 fxdr_nfsv2time(&fp->fa2_mtime, &mtime); 1679 ctime.tv_sec = fxdr_unsigned(u_int32_t, 1680 fp->fa2_ctime.nfsv2_sec); 1681 ctime.tv_nsec = 0; 1682 1683 /* 1684 * Really ugly NFSv2 kludge. 1685 */ 1686 if (vtyp == VCHR && rdev == 0xffffffff) 1687 vtyp = VFIFO; 1688 } 1689 1690 vmode &= ALLPERMS; 1691 1692 /* 1693 * If v_type == VNON it is a new node, so fill in the v_type, 1694 * n_mtime fields. Check to see if it represents a special 1695 * device, and if so, check for a possible alias. Once the 1696 * correct vnode has been obtained, fill in the rest of the 1697 * information. 1698 */ 1699 np = VTONFS(vp); 1700 if (vp->v_type == VNON) { 1701 vp->v_type = vtyp; 1702 if (vp->v_type == VFIFO) { 1703 extern int (**fifo_nfsv2nodeop_p) __P((void *)); 1704 vp->v_op = fifo_nfsv2nodeop_p; 1705 } else if (vp->v_type == VREG) { 1706 mutex_init(&np->n_commitlock, MUTEX_DEFAULT, IPL_NONE); 1707 } else if (vp->v_type == VCHR || vp->v_type == VBLK) { 1708 vp->v_op = spec_nfsv2nodeop_p; 1709 spec_node_init(vp, (dev_t)rdev); 1710 } 1711 np->n_mtime = mtime; 1712 } 1713 uid = fxdr_unsigned(uid_t, fp->fa_uid); 1714 gid = fxdr_unsigned(gid_t, fp->fa_gid); 1715 vap = np->n_vattr; 1716 1717 mutex_enter(&np->n_attrlock); 1718 /* 1719 * Invalidate access cache if uid, gid, mode or ctime changed. 1720 */ 1721 if (np->n_accstamp != -1 && 1722 (gid != vap->va_gid || uid != vap->va_uid || vmode != vap->va_mode 1723 || timespeccmp(&ctime, &vap->va_ctime, !=))) 1724 np->n_accstamp = -1; 1725 1726 vap->va_type = vtyp; 1727 vap->va_mode = vmode; 1728 vap->va_rdev = (dev_t)rdev; 1729 vap->va_mtime = mtime; 1730 vap->va_ctime = ctime; 1731 vap->va_fsid = vp->v_mount->mnt_stat.f_fsidx.__fsid_val[0]; 1732 switch (vtyp) { 1733 case VDIR: 1734 vap->va_blocksize = NFS_DIRFRAGSIZ; 1735 break; 1736 case VBLK: 1737 vap->va_blocksize = BLKDEV_IOSIZE; 1738 break; 1739 case VCHR: 1740 vap->va_blocksize = MAXBSIZE; 1741 break; 1742 default: 1743 vap->va_blocksize = v3 ? vp->v_mount->mnt_stat.f_iosize : 1744 fxdr_unsigned(int32_t, fp->fa2_blocksize); 1745 break; 1746 } 1747 if (v3) { 1748 vap->va_nlink = fxdr_unsigned(u_short, fp->fa_nlink); 1749 vap->va_uid = uid; 1750 vap->va_gid = gid; 1751 vap->va_size = fxdr_hyper(&fp->fa3_size); 1752 vap->va_bytes = fxdr_hyper(&fp->fa3_used); 1753 vap->va_fileid = fxdr_hyper(&fp->fa3_fileid); 1754 fxdr_nfsv3time(&fp->fa3_atime, &vap->va_atime); 1755 vap->va_flags = 0; 1756 vap->va_filerev = 0; 1757 } else { 1758 vap->va_nlink = fxdr_unsigned(u_short, fp->fa_nlink); 1759 vap->va_uid = uid; 1760 vap->va_gid = gid; 1761 vap->va_size = fxdr_unsigned(u_int32_t, fp->fa2_size); 1762 vap->va_bytes = fxdr_unsigned(int32_t, fp->fa2_blocks) 1763 * NFS_FABLKSIZE; 1764 vap->va_fileid = fxdr_unsigned(int32_t, fp->fa2_fileid); 1765 fxdr_nfsv2time(&fp->fa2_atime, &vap->va_atime); 1766 vap->va_flags = 0; 1767 vap->va_gen = fxdr_unsigned(u_int32_t,fp->fa2_ctime.nfsv2_usec); 1768 vap->va_filerev = 0; 1769 } 1770 if (vap->va_size > VFSTONFS(vp->v_mount)->nm_maxfilesize) { 1771 mutex_exit(&np->n_attrlock); 1772 return EFBIG; 1773 } 1774 if (vap->va_size != np->n_size) { 1775 if ((np->n_flag & NMODIFIED) && vap->va_size < np->n_size) { 1776 vap->va_size = np->n_size; 1777 } else { 1778 np->n_size = vap->va_size; 1779 if (vap->va_type == VREG) { 1780 /* 1781 * we can't free pages if NAC_NOTRUNC because 1782 * the pages can be owned by ourselves. 1783 */ 1784 if (flags & NAC_NOTRUNC) { 1785 np->n_flag |= NTRUNCDELAYED; 1786 } else { 1787 mutex_exit(&np->n_attrlock); /* XXX */ 1788 genfs_node_wrlock(vp); 1789 mutex_enter(&vp->v_interlock); 1790 (void)VOP_PUTPAGES(vp, 0, 1791 0, PGO_SYNCIO | PGO_CLEANIT | 1792 PGO_FREE | PGO_ALLPAGES); 1793 uvm_vnp_setsize(vp, np->n_size); 1794 genfs_node_unlock(vp); 1795 mutex_enter(&np->n_attrlock); 1796 } 1797 } 1798 } 1799 } 1800 np->n_attrstamp = time_second; 1801 if (vaper != NULL) { 1802 memcpy((void *)vaper, (void *)vap, sizeof(*vap)); 1803 if (np->n_flag & NCHG) { 1804 if (np->n_flag & NACC) 1805 vaper->va_atime = np->n_atim; 1806 if (np->n_flag & NUPD) 1807 vaper->va_mtime = np->n_mtim; 1808 } 1809 } 1810 mutex_exit(&np->n_attrlock); 1811 return (0); 1812 } 1813 1814 /* 1815 * Check the time stamp 1816 * If the cache is valid, copy contents to *vap and return 0 1817 * otherwise return an error 1818 */ 1819 int 1820 nfs_getattrcache(vp, vaper) 1821 struct vnode *vp; 1822 struct vattr *vaper; 1823 { 1824 struct nfsnode *np = VTONFS(vp); 1825 struct nfsmount *nmp = VFSTONFS(vp->v_mount); 1826 struct vattr *vap; 1827 1828 mutex_enter(&np->n_attrlock); 1829 if (np->n_attrstamp == 0 || 1830 (time_second - np->n_attrstamp) >= nfs_attrtimeo(nmp, np)) { 1831 mutex_exit(&np->n_attrlock); 1832 nfsstats.attrcache_misses++; 1833 return (ENOENT); 1834 } 1835 nfsstats.attrcache_hits++; 1836 1837 vap = np->n_vattr; 1838 if (vap->va_size != np->n_size) { 1839 if (vap->va_type == VREG) { 1840 voff_t size; 1841 1842 if ((np->n_flag & NMODIFIED) != 0 && 1843 vap->va_size < np->n_size) { 1844 vap->va_size = np->n_size; 1845 } else { 1846 np->n_size = vap->va_size; 1847 } 1848 size = np->n_size; 1849 mutex_exit(&np->n_attrlock); /* XXX */ 1850 genfs_node_wrlock(vp); 1851 uvm_vnp_setsize(vp, size); 1852 genfs_node_unlock(vp); 1853 mutex_enter(&np->n_attrlock); 1854 } else 1855 np->n_size = vap->va_size; 1856 } 1857 memcpy(vaper, vap, sizeof(struct vattr)); 1858 if (np->n_flag & NCHG) { 1859 if (np->n_flag & NACC) 1860 vaper->va_atime = np->n_atim; 1861 if (np->n_flag & NUPD) 1862 vaper->va_mtime = np->n_mtim; 1863 } 1864 mutex_exit(&np->n_attrlock); 1865 return (0); 1866 } 1867 1868 void 1869 nfs_delayedtruncate(vp) 1870 struct vnode *vp; 1871 { 1872 struct nfsnode *np = VTONFS(vp); 1873 1874 if (np->n_flag & NTRUNCDELAYED) { 1875 np->n_flag &= ~NTRUNCDELAYED; 1876 genfs_node_wrlock(vp); 1877 mutex_enter(&vp->v_interlock); 1878 (void)VOP_PUTPAGES(vp, 0, 1879 0, PGO_SYNCIO | PGO_CLEANIT | PGO_FREE | PGO_ALLPAGES); 1880 uvm_vnp_setsize(vp, np->n_size); 1881 genfs_node_unlock(vp); 1882 } 1883 } 1884 1885 #define NFS_WCCKLUDGE_TIMEOUT (24 * 60 * 60) /* 1 day */ 1886 #define NFS_WCCKLUDGE(nmp, now) \ 1887 (((nmp)->nm_iflag & NFSMNT_WCCKLUDGE) && \ 1888 ((now) - (nmp)->nm_wcckludgetime - NFS_WCCKLUDGE_TIMEOUT) < 0) 1889 1890 /* 1891 * nfs_check_wccdata: check inaccurate wcc_data 1892 * 1893 * => return non-zero if we shouldn't trust the wcc_data. 1894 * => NFS_WCCKLUDGE_TIMEOUT is for the case that the server is "fixed". 1895 */ 1896 1897 int 1898 nfs_check_wccdata(struct nfsnode *np, const struct timespec *ctime, 1899 struct timespec *mtime, bool docheck) 1900 { 1901 int error = 0; 1902 1903 #if !defined(NFS_V2_ONLY) 1904 1905 if (docheck) { 1906 struct vnode *vp = NFSTOV(np); 1907 struct nfsmount *nmp; 1908 long now = time_second; 1909 const struct timespec *omtime = &np->n_vattr->va_mtime; 1910 const struct timespec *octime = &np->n_vattr->va_ctime; 1911 const char *reason = NULL; /* XXX: gcc */ 1912 1913 if (timespeccmp(omtime, mtime, <=)) { 1914 reason = "mtime"; 1915 error = EINVAL; 1916 } 1917 1918 if (vp->v_type == VDIR && timespeccmp(octime, ctime, <=)) { 1919 reason = "ctime"; 1920 error = EINVAL; 1921 } 1922 1923 nmp = VFSTONFS(vp->v_mount); 1924 if (error) { 1925 1926 /* 1927 * despite of the fact that we've updated the file, 1928 * timestamps of the file were not updated as we 1929 * expected. 1930 * it means that the server has incompatible 1931 * semantics of timestamps or (more likely) 1932 * the server time is not precise enough to 1933 * track each modifications. 1934 * in that case, we disable wcc processing. 1935 * 1936 * yes, strictly speaking, we should disable all 1937 * caching. it's a compromise. 1938 */ 1939 1940 mutex_enter(&nmp->nm_lock); 1941 if (!NFS_WCCKLUDGE(nmp, now)) { 1942 printf("%s: inaccurate wcc data (%s) detected," 1943 " disabling wcc" 1944 " (ctime %u.%09u %u.%09u," 1945 " mtime %u.%09u %u.%09u)\n", 1946 vp->v_mount->mnt_stat.f_mntfromname, 1947 reason, 1948 (unsigned int)octime->tv_sec, 1949 (unsigned int)octime->tv_nsec, 1950 (unsigned int)ctime->tv_sec, 1951 (unsigned int)ctime->tv_nsec, 1952 (unsigned int)omtime->tv_sec, 1953 (unsigned int)omtime->tv_nsec, 1954 (unsigned int)mtime->tv_sec, 1955 (unsigned int)mtime->tv_nsec); 1956 } 1957 nmp->nm_iflag |= NFSMNT_WCCKLUDGE; 1958 nmp->nm_wcckludgetime = now; 1959 mutex_exit(&nmp->nm_lock); 1960 } else if (NFS_WCCKLUDGE(nmp, now)) { 1961 error = EPERM; /* XXX */ 1962 } else if (nmp->nm_iflag & NFSMNT_WCCKLUDGE) { 1963 mutex_enter(&nmp->nm_lock); 1964 if (nmp->nm_iflag & NFSMNT_WCCKLUDGE) { 1965 printf("%s: re-enabling wcc\n", 1966 vp->v_mount->mnt_stat.f_mntfromname); 1967 nmp->nm_iflag &= ~NFSMNT_WCCKLUDGE; 1968 } 1969 mutex_exit(&nmp->nm_lock); 1970 } 1971 } 1972 1973 #endif /* !defined(NFS_V2_ONLY) */ 1974 1975 return error; 1976 } 1977 1978 /* 1979 * Heuristic to see if the server XDR encodes directory cookies or not. 1980 * it is not supposed to, but a lot of servers may do this. Also, since 1981 * most/all servers will implement V2 as well, it is expected that they 1982 * may return just 32 bits worth of cookie information, so we need to 1983 * find out in which 32 bits this information is available. We do this 1984 * to avoid trouble with emulated binaries that can't handle 64 bit 1985 * directory offsets. 1986 */ 1987 1988 void 1989 nfs_cookieheuristic(vp, flagp, l, cred) 1990 struct vnode *vp; 1991 int *flagp; 1992 struct lwp *l; 1993 kauth_cred_t cred; 1994 { 1995 struct uio auio; 1996 struct iovec aiov; 1997 char *tbuf, *cp; 1998 struct dirent *dp; 1999 off_t *cookies = NULL, *cop; 2000 int error, eof, nc, len; 2001 2002 MALLOC(tbuf, void *, NFS_DIRFRAGSIZ, M_TEMP, M_WAITOK); 2003 2004 aiov.iov_base = tbuf; 2005 aiov.iov_len = NFS_DIRFRAGSIZ; 2006 auio.uio_iov = &aiov; 2007 auio.uio_iovcnt = 1; 2008 auio.uio_rw = UIO_READ; 2009 auio.uio_resid = NFS_DIRFRAGSIZ; 2010 auio.uio_offset = 0; 2011 UIO_SETUP_SYSSPACE(&auio); 2012 2013 error = VOP_READDIR(vp, &auio, cred, &eof, &cookies, &nc); 2014 2015 len = NFS_DIRFRAGSIZ - auio.uio_resid; 2016 if (error || len == 0) { 2017 FREE(tbuf, M_TEMP); 2018 if (cookies) 2019 free(cookies, M_TEMP); 2020 return; 2021 } 2022 2023 /* 2024 * Find the first valid entry and look at its offset cookie. 2025 */ 2026 2027 cp = tbuf; 2028 for (cop = cookies; len > 0; len -= dp->d_reclen) { 2029 dp = (struct dirent *)cp; 2030 if (dp->d_fileno != 0 && len >= dp->d_reclen) { 2031 if ((*cop >> 32) != 0 && (*cop & 0xffffffffLL) == 0) { 2032 *flagp |= NFSMNT_SWAPCOOKIE; 2033 nfs_invaldircache(vp, 0); 2034 nfs_vinvalbuf(vp, 0, cred, l, 1); 2035 } 2036 break; 2037 } 2038 cop++; 2039 cp += dp->d_reclen; 2040 } 2041 2042 FREE(tbuf, M_TEMP); 2043 free(cookies, M_TEMP); 2044 } 2045 #endif /* NFS */ 2046 2047 #ifdef NFSSERVER 2048 /* 2049 * Set up nameidata for a lookup() call and do it. 2050 * 2051 * If pubflag is set, this call is done for a lookup operation on the 2052 * public filehandle. In that case we allow crossing mountpoints and 2053 * absolute pathnames. However, the caller is expected to check that 2054 * the lookup result is within the public fs, and deny access if 2055 * it is not. 2056 */ 2057 int 2058 nfs_namei(ndp, nsfh, len, slp, nam, mdp, dposp, retdirp, l, kerbflag, pubflag) 2059 struct nameidata *ndp; 2060 nfsrvfh_t *nsfh; 2061 uint32_t len; 2062 struct nfssvc_sock *slp; 2063 struct mbuf *nam; 2064 struct mbuf **mdp; 2065 char **dposp; 2066 struct vnode **retdirp; 2067 struct lwp *l; 2068 int kerbflag, pubflag; 2069 { 2070 int i, rem; 2071 struct mbuf *md; 2072 char *fromcp, *tocp, *cp; 2073 struct iovec aiov; 2074 struct uio auio; 2075 struct vnode *dp; 2076 int error, rdonly, linklen; 2077 struct componentname *cnp = &ndp->ni_cnd; 2078 2079 *retdirp = NULL; 2080 2081 if ((len + 1) > MAXPATHLEN) 2082 return (ENAMETOOLONG); 2083 if (len == 0) 2084 return (EACCES); 2085 cnp->cn_pnbuf = PNBUF_GET(); 2086 2087 /* 2088 * Copy the name from the mbuf list to ndp->ni_pnbuf 2089 * and set the various ndp fields appropriately. 2090 */ 2091 fromcp = *dposp; 2092 tocp = cnp->cn_pnbuf; 2093 md = *mdp; 2094 rem = mtod(md, char *) + md->m_len - fromcp; 2095 for (i = 0; i < len; i++) { 2096 while (rem == 0) { 2097 md = md->m_next; 2098 if (md == NULL) { 2099 error = EBADRPC; 2100 goto out; 2101 } 2102 fromcp = mtod(md, void *); 2103 rem = md->m_len; 2104 } 2105 if (*fromcp == '\0' || (!pubflag && *fromcp == '/')) { 2106 error = EACCES; 2107 goto out; 2108 } 2109 *tocp++ = *fromcp++; 2110 rem--; 2111 } 2112 *tocp = '\0'; 2113 *mdp = md; 2114 *dposp = fromcp; 2115 len = nfsm_rndup(len)-len; 2116 if (len > 0) { 2117 if (rem >= len) 2118 *dposp += len; 2119 else if ((error = nfs_adv(mdp, dposp, len, rem)) != 0) 2120 goto out; 2121 } 2122 2123 /* 2124 * Extract and set starting directory. 2125 */ 2126 error = nfsrv_fhtovp(nsfh, false, &dp, ndp->ni_cnd.cn_cred, slp, 2127 nam, &rdonly, kerbflag, pubflag); 2128 if (error) 2129 goto out; 2130 if (dp->v_type != VDIR) { 2131 vrele(dp); 2132 error = ENOTDIR; 2133 goto out; 2134 } 2135 2136 if (rdonly) 2137 cnp->cn_flags |= RDONLY; 2138 2139 *retdirp = dp; 2140 2141 if (pubflag) { 2142 /* 2143 * Oh joy. For WebNFS, handle those pesky '%' escapes, 2144 * and the 'native path' indicator. 2145 */ 2146 cp = PNBUF_GET(); 2147 fromcp = cnp->cn_pnbuf; 2148 tocp = cp; 2149 if ((unsigned char)*fromcp >= WEBNFS_SPECCHAR_START) { 2150 switch ((unsigned char)*fromcp) { 2151 case WEBNFS_NATIVE_CHAR: 2152 /* 2153 * 'Native' path for us is the same 2154 * as a path according to the NFS spec, 2155 * just skip the escape char. 2156 */ 2157 fromcp++; 2158 break; 2159 /* 2160 * More may be added in the future, range 0x80-0xff 2161 */ 2162 default: 2163 error = EIO; 2164 vrele(dp); 2165 PNBUF_PUT(cp); 2166 goto out; 2167 } 2168 } 2169 /* 2170 * Translate the '%' escapes, URL-style. 2171 */ 2172 while (*fromcp != '\0') { 2173 if (*fromcp == WEBNFS_ESC_CHAR) { 2174 if (fromcp[1] != '\0' && fromcp[2] != '\0') { 2175 fromcp++; 2176 *tocp++ = HEXSTRTOI(fromcp); 2177 fromcp += 2; 2178 continue; 2179 } else { 2180 error = ENOENT; 2181 vrele(dp); 2182 PNBUF_PUT(cp); 2183 goto out; 2184 } 2185 } else 2186 *tocp++ = *fromcp++; 2187 } 2188 *tocp = '\0'; 2189 PNBUF_PUT(cnp->cn_pnbuf); 2190 cnp->cn_pnbuf = cp; 2191 } 2192 2193 ndp->ni_pathlen = (tocp - cnp->cn_pnbuf) + 1; 2194 ndp->ni_segflg = UIO_SYSSPACE; 2195 ndp->ni_rootdir = rootvnode; 2196 ndp->ni_erootdir = NULL; 2197 2198 if (pubflag) { 2199 ndp->ni_loopcnt = 0; 2200 if (cnp->cn_pnbuf[0] == '/') 2201 dp = rootvnode; 2202 } else { 2203 cnp->cn_flags |= NOCROSSMOUNT; 2204 } 2205 2206 VREF(dp); 2207 vn_lock(dp, LK_EXCLUSIVE | LK_RETRY); 2208 2209 for (;;) { 2210 cnp->cn_nameptr = cnp->cn_pnbuf; 2211 ndp->ni_startdir = dp; 2212 2213 /* 2214 * And call lookup() to do the real work 2215 */ 2216 error = lookup(ndp); 2217 if (error) { 2218 if (ndp->ni_dvp) { 2219 vput(ndp->ni_dvp); 2220 } 2221 PNBUF_PUT(cnp->cn_pnbuf); 2222 return (error); 2223 } 2224 2225 /* 2226 * Check for encountering a symbolic link 2227 */ 2228 if ((cnp->cn_flags & ISSYMLINK) == 0) { 2229 if ((cnp->cn_flags & LOCKPARENT) == 0 && ndp->ni_dvp) { 2230 if (ndp->ni_dvp == ndp->ni_vp) { 2231 vrele(ndp->ni_dvp); 2232 } else { 2233 vput(ndp->ni_dvp); 2234 } 2235 } 2236 if (cnp->cn_flags & (SAVENAME | SAVESTART)) 2237 cnp->cn_flags |= HASBUF; 2238 else 2239 PNBUF_PUT(cnp->cn_pnbuf); 2240 return (0); 2241 } else { 2242 if (!pubflag) { 2243 error = EINVAL; 2244 break; 2245 } 2246 if (ndp->ni_loopcnt++ >= MAXSYMLINKS) { 2247 error = ELOOP; 2248 break; 2249 } 2250 if (ndp->ni_vp->v_mount->mnt_flag & MNT_SYMPERM) { 2251 error = VOP_ACCESS(ndp->ni_vp, VEXEC, cnp->cn_cred); 2252 if (error != 0) 2253 break; 2254 } 2255 if (ndp->ni_pathlen > 1) 2256 cp = PNBUF_GET(); 2257 else 2258 cp = cnp->cn_pnbuf; 2259 aiov.iov_base = cp; 2260 aiov.iov_len = MAXPATHLEN; 2261 auio.uio_iov = &aiov; 2262 auio.uio_iovcnt = 1; 2263 auio.uio_offset = 0; 2264 auio.uio_rw = UIO_READ; 2265 auio.uio_resid = MAXPATHLEN; 2266 UIO_SETUP_SYSSPACE(&auio); 2267 error = VOP_READLINK(ndp->ni_vp, &auio, cnp->cn_cred); 2268 if (error) { 2269 badlink: 2270 if (ndp->ni_pathlen > 1) 2271 PNBUF_PUT(cp); 2272 break; 2273 } 2274 linklen = MAXPATHLEN - auio.uio_resid; 2275 if (linklen == 0) { 2276 error = ENOENT; 2277 goto badlink; 2278 } 2279 if (linklen + ndp->ni_pathlen >= MAXPATHLEN) { 2280 error = ENAMETOOLONG; 2281 goto badlink; 2282 } 2283 if (ndp->ni_pathlen > 1) { 2284 memcpy(cp + linklen, ndp->ni_next, ndp->ni_pathlen); 2285 PNBUF_PUT(cnp->cn_pnbuf); 2286 cnp->cn_pnbuf = cp; 2287 } else 2288 cnp->cn_pnbuf[linklen] = '\0'; 2289 ndp->ni_pathlen += linklen; 2290 vput(ndp->ni_vp); 2291 dp = ndp->ni_dvp; 2292 2293 /* 2294 * Check if root directory should replace current directory. 2295 */ 2296 if (cnp->cn_pnbuf[0] == '/') { 2297 vput(dp); 2298 dp = ndp->ni_rootdir; 2299 VREF(dp); 2300 vn_lock(dp, LK_EXCLUSIVE | LK_RETRY); 2301 } 2302 } 2303 } 2304 vput(ndp->ni_dvp); 2305 vput(ndp->ni_vp); 2306 ndp->ni_vp = NULL; 2307 out: 2308 PNBUF_PUT(cnp->cn_pnbuf); 2309 return (error); 2310 } 2311 #endif /* NFSSERVER */ 2312 2313 /* 2314 * A fiddled version of m_adj() that ensures null fill to a 32-bit 2315 * boundary and only trims off the back end 2316 * 2317 * 1. trim off 'len' bytes as m_adj(mp, -len). 2318 * 2. add zero-padding 'nul' bytes at the end of the mbuf chain. 2319 */ 2320 void 2321 nfs_zeropad(mp, len, nul) 2322 struct mbuf *mp; 2323 int len; 2324 int nul; 2325 { 2326 struct mbuf *m; 2327 int count; 2328 2329 /* 2330 * Trim from tail. Scan the mbuf chain, 2331 * calculating its length and finding the last mbuf. 2332 * If the adjustment only affects this mbuf, then just 2333 * adjust and return. Otherwise, rescan and truncate 2334 * after the remaining size. 2335 */ 2336 count = 0; 2337 m = mp; 2338 for (;;) { 2339 count += m->m_len; 2340 if (m->m_next == NULL) 2341 break; 2342 m = m->m_next; 2343 } 2344 2345 KDASSERT(count >= len); 2346 2347 if (m->m_len >= len) { 2348 m->m_len -= len; 2349 } else { 2350 count -= len; 2351 /* 2352 * Correct length for chain is "count". 2353 * Find the mbuf with last data, adjust its length, 2354 * and toss data from remaining mbufs on chain. 2355 */ 2356 for (m = mp; m; m = m->m_next) { 2357 if (m->m_len >= count) { 2358 m->m_len = count; 2359 break; 2360 } 2361 count -= m->m_len; 2362 } 2363 KASSERT(m && m->m_next); 2364 m_freem(m->m_next); 2365 m->m_next = NULL; 2366 } 2367 2368 KDASSERT(m->m_next == NULL); 2369 2370 /* 2371 * zero-padding. 2372 */ 2373 if (nul > 0) { 2374 char *cp; 2375 int i; 2376 2377 if (M_ROMAP(m) || M_TRAILINGSPACE(m) < nul) { 2378 struct mbuf *n; 2379 2380 KDASSERT(MLEN >= nul); 2381 n = m_get(M_WAIT, MT_DATA); 2382 MCLAIM(n, &nfs_mowner); 2383 n->m_len = nul; 2384 n->m_next = NULL; 2385 m->m_next = n; 2386 cp = mtod(n, void *); 2387 } else { 2388 cp = mtod(m, char *) + m->m_len; 2389 m->m_len += nul; 2390 } 2391 for (i = 0; i < nul; i++) 2392 *cp++ = '\0'; 2393 } 2394 return; 2395 } 2396 2397 /* 2398 * Make these functions instead of macros, so that the kernel text size 2399 * doesn't get too big... 2400 */ 2401 void 2402 nfsm_srvwcc(nfsd, before_ret, before_vap, after_ret, after_vap, mbp, bposp) 2403 struct nfsrv_descript *nfsd; 2404 int before_ret; 2405 struct vattr *before_vap; 2406 int after_ret; 2407 struct vattr *after_vap; 2408 struct mbuf **mbp; 2409 char **bposp; 2410 { 2411 struct mbuf *mb = *mbp; 2412 char *bpos = *bposp; 2413 u_int32_t *tl; 2414 2415 if (before_ret) { 2416 nfsm_build(tl, u_int32_t *, NFSX_UNSIGNED); 2417 *tl = nfs_false; 2418 } else { 2419 nfsm_build(tl, u_int32_t *, 7 * NFSX_UNSIGNED); 2420 *tl++ = nfs_true; 2421 txdr_hyper(before_vap->va_size, tl); 2422 tl += 2; 2423 txdr_nfsv3time(&(before_vap->va_mtime), tl); 2424 tl += 2; 2425 txdr_nfsv3time(&(before_vap->va_ctime), tl); 2426 } 2427 *bposp = bpos; 2428 *mbp = mb; 2429 nfsm_srvpostopattr(nfsd, after_ret, after_vap, mbp, bposp); 2430 } 2431 2432 void 2433 nfsm_srvpostopattr(nfsd, after_ret, after_vap, mbp, bposp) 2434 struct nfsrv_descript *nfsd; 2435 int after_ret; 2436 struct vattr *after_vap; 2437 struct mbuf **mbp; 2438 char **bposp; 2439 { 2440 struct mbuf *mb = *mbp; 2441 char *bpos = *bposp; 2442 u_int32_t *tl; 2443 struct nfs_fattr *fp; 2444 2445 if (after_ret) { 2446 nfsm_build(tl, u_int32_t *, NFSX_UNSIGNED); 2447 *tl = nfs_false; 2448 } else { 2449 nfsm_build(tl, u_int32_t *, NFSX_UNSIGNED + NFSX_V3FATTR); 2450 *tl++ = nfs_true; 2451 fp = (struct nfs_fattr *)tl; 2452 nfsm_srvfattr(nfsd, after_vap, fp); 2453 } 2454 *mbp = mb; 2455 *bposp = bpos; 2456 } 2457 2458 void 2459 nfsm_srvfattr(nfsd, vap, fp) 2460 struct nfsrv_descript *nfsd; 2461 struct vattr *vap; 2462 struct nfs_fattr *fp; 2463 { 2464 2465 fp->fa_nlink = txdr_unsigned(vap->va_nlink); 2466 fp->fa_uid = txdr_unsigned(vap->va_uid); 2467 fp->fa_gid = txdr_unsigned(vap->va_gid); 2468 if (nfsd->nd_flag & ND_NFSV3) { 2469 fp->fa_type = vtonfsv3_type(vap->va_type); 2470 fp->fa_mode = vtonfsv3_mode(vap->va_mode); 2471 txdr_hyper(vap->va_size, &fp->fa3_size); 2472 txdr_hyper(vap->va_bytes, &fp->fa3_used); 2473 fp->fa3_rdev.specdata1 = txdr_unsigned(major(vap->va_rdev)); 2474 fp->fa3_rdev.specdata2 = txdr_unsigned(minor(vap->va_rdev)); 2475 fp->fa3_fsid.nfsuquad[0] = 0; 2476 fp->fa3_fsid.nfsuquad[1] = txdr_unsigned(vap->va_fsid); 2477 txdr_hyper(vap->va_fileid, &fp->fa3_fileid); 2478 txdr_nfsv3time(&vap->va_atime, &fp->fa3_atime); 2479 txdr_nfsv3time(&vap->va_mtime, &fp->fa3_mtime); 2480 txdr_nfsv3time(&vap->va_ctime, &fp->fa3_ctime); 2481 } else { 2482 fp->fa_type = vtonfsv2_type(vap->va_type); 2483 fp->fa_mode = vtonfsv2_mode(vap->va_type, vap->va_mode); 2484 fp->fa2_size = txdr_unsigned(vap->va_size); 2485 fp->fa2_blocksize = txdr_unsigned(vap->va_blocksize); 2486 if (vap->va_type == VFIFO) 2487 fp->fa2_rdev = 0xffffffff; 2488 else 2489 fp->fa2_rdev = txdr_unsigned(vap->va_rdev); 2490 fp->fa2_blocks = txdr_unsigned(vap->va_bytes / NFS_FABLKSIZE); 2491 fp->fa2_fsid = txdr_unsigned(vap->va_fsid); 2492 fp->fa2_fileid = txdr_unsigned(vap->va_fileid); 2493 txdr_nfsv2time(&vap->va_atime, &fp->fa2_atime); 2494 txdr_nfsv2time(&vap->va_mtime, &fp->fa2_mtime); 2495 txdr_nfsv2time(&vap->va_ctime, &fp->fa2_ctime); 2496 } 2497 } 2498 2499 #ifdef NFSSERVER 2500 /* 2501 * nfsrv_fhtovp() - convert a fh to a vnode ptr (optionally locked) 2502 * - look up fsid in mount list (if not found ret error) 2503 * - get vp and export rights by calling VFS_FHTOVP() 2504 * - if cred->cr_uid == 0 or MNT_EXPORTANON set it to credanon 2505 * - if not lockflag unlock it with VOP_UNLOCK() 2506 */ 2507 int 2508 nfsrv_fhtovp(nfsrvfh_t *nsfh, int lockflag, struct vnode **vpp, 2509 kauth_cred_t cred, struct nfssvc_sock *slp, struct mbuf *nam, int *rdonlyp, 2510 int kerbflag, int pubflag) 2511 { 2512 struct mount *mp; 2513 kauth_cred_t credanon; 2514 int error, exflags; 2515 struct sockaddr_in *saddr; 2516 fhandle_t *fhp; 2517 2518 fhp = NFSRVFH_FHANDLE(nsfh); 2519 *vpp = (struct vnode *)0; 2520 2521 if (nfs_ispublicfh(nsfh)) { 2522 if (!pubflag || !nfs_pub.np_valid) 2523 return (ESTALE); 2524 fhp = nfs_pub.np_handle; 2525 } 2526 2527 error = netexport_check(&fhp->fh_fsid, nam, &mp, &exflags, &credanon); 2528 if (error) { 2529 return error; 2530 } 2531 2532 error = VFS_FHTOVP(mp, &fhp->fh_fid, vpp); 2533 if (error) 2534 return (error); 2535 2536 if (!(exflags & (MNT_EXNORESPORT|MNT_EXPUBLIC))) { 2537 saddr = mtod(nam, struct sockaddr_in *); 2538 if ((saddr->sin_family == AF_INET) && 2539 ntohs(saddr->sin_port) >= IPPORT_RESERVED) { 2540 vput(*vpp); 2541 return (NFSERR_AUTHERR | AUTH_TOOWEAK); 2542 } 2543 #ifdef INET6 2544 if ((saddr->sin_family == AF_INET6) && 2545 ntohs(saddr->sin_port) >= IPV6PORT_RESERVED) { 2546 vput(*vpp); 2547 return (NFSERR_AUTHERR | AUTH_TOOWEAK); 2548 } 2549 #endif 2550 } 2551 /* 2552 * Check/setup credentials. 2553 */ 2554 if (exflags & MNT_EXKERB) { 2555 if (!kerbflag) { 2556 vput(*vpp); 2557 return (NFSERR_AUTHERR | AUTH_TOOWEAK); 2558 } 2559 } else if (kerbflag) { 2560 vput(*vpp); 2561 return (NFSERR_AUTHERR | AUTH_TOOWEAK); 2562 } else if (kauth_cred_geteuid(cred) == 0 || /* NFS maproot, see below */ 2563 (exflags & MNT_EXPORTANON)) { 2564 /* 2565 * This is used by the NFS maproot option. While we can change 2566 * the secmodel on our own host, we can't change it on the 2567 * clients. As means of least surprise, we're doing the 2568 * traditional thing here. 2569 * Should look into adding a "mapprivileged" or similar where 2570 * the users can be explicitly specified... 2571 * [elad, yamt 2008-03-05] 2572 */ 2573 kauth_cred_clone(credanon, cred); 2574 } 2575 if (exflags & MNT_EXRDONLY) 2576 *rdonlyp = 1; 2577 else 2578 *rdonlyp = 0; 2579 if (!lockflag) 2580 VOP_UNLOCK(*vpp, 0); 2581 return (0); 2582 } 2583 2584 /* 2585 * WebNFS: check if a filehandle is a public filehandle. For v3, this 2586 * means a length of 0, for v2 it means all zeroes. 2587 */ 2588 int 2589 nfs_ispublicfh(const nfsrvfh_t *nsfh) 2590 { 2591 const char *cp = (const void *)(NFSRVFH_DATA(nsfh)); 2592 int i; 2593 2594 if (NFSRVFH_SIZE(nsfh) == 0) { 2595 return true; 2596 } 2597 if (NFSRVFH_SIZE(nsfh) != NFSX_V2FH) { 2598 return false; 2599 } 2600 for (i = 0; i < NFSX_V2FH; i++) 2601 if (*cp++ != 0) 2602 return false; 2603 return true; 2604 } 2605 #endif /* NFSSERVER */ 2606 2607 /* 2608 * This function compares two net addresses by family and returns true 2609 * if they are the same host. 2610 * If there is any doubt, return false. 2611 * The AF_INET family is handled as a special case so that address mbufs 2612 * don't need to be saved to store "struct in_addr", which is only 4 bytes. 2613 */ 2614 int 2615 netaddr_match(family, haddr, nam) 2616 int family; 2617 union nethostaddr *haddr; 2618 struct mbuf *nam; 2619 { 2620 struct sockaddr_in *inetaddr; 2621 2622 switch (family) { 2623 case AF_INET: 2624 inetaddr = mtod(nam, struct sockaddr_in *); 2625 if (inetaddr->sin_family == AF_INET && 2626 inetaddr->sin_addr.s_addr == haddr->had_inetaddr) 2627 return (1); 2628 break; 2629 #ifdef INET6 2630 case AF_INET6: 2631 { 2632 struct sockaddr_in6 *sin6_1, *sin6_2; 2633 2634 sin6_1 = mtod(nam, struct sockaddr_in6 *); 2635 sin6_2 = mtod(haddr->had_nam, struct sockaddr_in6 *); 2636 if (sin6_1->sin6_family == AF_INET6 && 2637 IN6_ARE_ADDR_EQUAL(&sin6_1->sin6_addr, &sin6_2->sin6_addr)) 2638 return 1; 2639 } 2640 #endif 2641 #ifdef ISO 2642 case AF_ISO: 2643 { 2644 struct sockaddr_iso *isoaddr1, *isoaddr2; 2645 2646 isoaddr1 = mtod(nam, struct sockaddr_iso *); 2647 isoaddr2 = mtod(haddr->had_nam, struct sockaddr_iso *); 2648 if (isoaddr1->siso_family == AF_ISO && 2649 isoaddr1->siso_nlen > 0 && 2650 isoaddr1->siso_nlen == isoaddr2->siso_nlen && 2651 SAME_ISOADDR(isoaddr1, isoaddr2)) 2652 return (1); 2653 break; 2654 } 2655 #endif /* ISO */ 2656 default: 2657 break; 2658 }; 2659 return (0); 2660 } 2661 2662 /* 2663 * The write verifier has changed (probably due to a server reboot), so all 2664 * PG_NEEDCOMMIT pages will have to be written again. Since they are marked 2665 * as dirty or are being written out just now, all this takes is clearing 2666 * the PG_NEEDCOMMIT flag. Once done the new write verifier can be set for 2667 * the mount point. 2668 */ 2669 void 2670 nfs_clearcommit(mp) 2671 struct mount *mp; 2672 { 2673 struct vnode *vp; 2674 struct nfsnode *np; 2675 struct vm_page *pg; 2676 struct nfsmount *nmp = VFSTONFS(mp); 2677 2678 rw_enter(&nmp->nm_writeverflock, RW_WRITER); 2679 mutex_enter(&mntvnode_lock); 2680 TAILQ_FOREACH(vp, &mp->mnt_vnodelist, v_mntvnodes) { 2681 KASSERT(vp->v_mount == mp); 2682 if (vp->v_type != VREG) 2683 continue; 2684 np = VTONFS(vp); 2685 np->n_pushlo = np->n_pushhi = np->n_pushedlo = 2686 np->n_pushedhi = 0; 2687 np->n_commitflags &= 2688 ~(NFS_COMMIT_PUSH_VALID | NFS_COMMIT_PUSHED_VALID); 2689 mutex_enter(&vp->v_uobj.vmobjlock); 2690 TAILQ_FOREACH(pg, &vp->v_uobj.memq, listq) { 2691 pg->flags &= ~PG_NEEDCOMMIT; 2692 } 2693 mutex_exit(&vp->v_uobj.vmobjlock); 2694 } 2695 mutex_exit(&mntvnode_lock); 2696 mutex_enter(&nmp->nm_lock); 2697 nmp->nm_iflag &= ~NFSMNT_STALEWRITEVERF; 2698 mutex_exit(&nmp->nm_lock); 2699 rw_exit(&nmp->nm_writeverflock); 2700 } 2701 2702 void 2703 nfs_merge_commit_ranges(vp) 2704 struct vnode *vp; 2705 { 2706 struct nfsnode *np = VTONFS(vp); 2707 2708 KASSERT(np->n_commitflags & NFS_COMMIT_PUSH_VALID); 2709 2710 if (!(np->n_commitflags & NFS_COMMIT_PUSHED_VALID)) { 2711 np->n_pushedlo = np->n_pushlo; 2712 np->n_pushedhi = np->n_pushhi; 2713 np->n_commitflags |= NFS_COMMIT_PUSHED_VALID; 2714 } else { 2715 if (np->n_pushlo < np->n_pushedlo) 2716 np->n_pushedlo = np->n_pushlo; 2717 if (np->n_pushhi > np->n_pushedhi) 2718 np->n_pushedhi = np->n_pushhi; 2719 } 2720 2721 np->n_pushlo = np->n_pushhi = 0; 2722 np->n_commitflags &= ~NFS_COMMIT_PUSH_VALID; 2723 2724 #ifdef NFS_DEBUG_COMMIT 2725 printf("merge: committed: %u - %u\n", (unsigned)np->n_pushedlo, 2726 (unsigned)np->n_pushedhi); 2727 #endif 2728 } 2729 2730 int 2731 nfs_in_committed_range(vp, off, len) 2732 struct vnode *vp; 2733 off_t off, len; 2734 { 2735 struct nfsnode *np = VTONFS(vp); 2736 off_t lo, hi; 2737 2738 if (!(np->n_commitflags & NFS_COMMIT_PUSHED_VALID)) 2739 return 0; 2740 lo = off; 2741 hi = lo + len; 2742 2743 return (lo >= np->n_pushedlo && hi <= np->n_pushedhi); 2744 } 2745 2746 int 2747 nfs_in_tobecommitted_range(vp, off, len) 2748 struct vnode *vp; 2749 off_t off, len; 2750 { 2751 struct nfsnode *np = VTONFS(vp); 2752 off_t lo, hi; 2753 2754 if (!(np->n_commitflags & NFS_COMMIT_PUSH_VALID)) 2755 return 0; 2756 lo = off; 2757 hi = lo + len; 2758 2759 return (lo >= np->n_pushlo && hi <= np->n_pushhi); 2760 } 2761 2762 void 2763 nfs_add_committed_range(vp, off, len) 2764 struct vnode *vp; 2765 off_t off, len; 2766 { 2767 struct nfsnode *np = VTONFS(vp); 2768 off_t lo, hi; 2769 2770 lo = off; 2771 hi = lo + len; 2772 2773 if (!(np->n_commitflags & NFS_COMMIT_PUSHED_VALID)) { 2774 np->n_pushedlo = lo; 2775 np->n_pushedhi = hi; 2776 np->n_commitflags |= NFS_COMMIT_PUSHED_VALID; 2777 } else { 2778 if (hi > np->n_pushedhi) 2779 np->n_pushedhi = hi; 2780 if (lo < np->n_pushedlo) 2781 np->n_pushedlo = lo; 2782 } 2783 #ifdef NFS_DEBUG_COMMIT 2784 printf("add: committed: %u - %u\n", (unsigned)np->n_pushedlo, 2785 (unsigned)np->n_pushedhi); 2786 #endif 2787 } 2788 2789 void 2790 nfs_del_committed_range(vp, off, len) 2791 struct vnode *vp; 2792 off_t off, len; 2793 { 2794 struct nfsnode *np = VTONFS(vp); 2795 off_t lo, hi; 2796 2797 if (!(np->n_commitflags & NFS_COMMIT_PUSHED_VALID)) 2798 return; 2799 2800 lo = off; 2801 hi = lo + len; 2802 2803 if (lo > np->n_pushedhi || hi < np->n_pushedlo) 2804 return; 2805 if (lo <= np->n_pushedlo) 2806 np->n_pushedlo = hi; 2807 else if (hi >= np->n_pushedhi) 2808 np->n_pushedhi = lo; 2809 else { 2810 /* 2811 * XXX There's only one range. If the deleted range 2812 * is in the middle, pick the largest of the 2813 * contiguous ranges that it leaves. 2814 */ 2815 if ((np->n_pushedlo - lo) > (hi - np->n_pushedhi)) 2816 np->n_pushedhi = lo; 2817 else 2818 np->n_pushedlo = hi; 2819 } 2820 #ifdef NFS_DEBUG_COMMIT 2821 printf("del: committed: %u - %u\n", (unsigned)np->n_pushedlo, 2822 (unsigned)np->n_pushedhi); 2823 #endif 2824 } 2825 2826 void 2827 nfs_add_tobecommitted_range(vp, off, len) 2828 struct vnode *vp; 2829 off_t off, len; 2830 { 2831 struct nfsnode *np = VTONFS(vp); 2832 off_t lo, hi; 2833 2834 lo = off; 2835 hi = lo + len; 2836 2837 if (!(np->n_commitflags & NFS_COMMIT_PUSH_VALID)) { 2838 np->n_pushlo = lo; 2839 np->n_pushhi = hi; 2840 np->n_commitflags |= NFS_COMMIT_PUSH_VALID; 2841 } else { 2842 if (lo < np->n_pushlo) 2843 np->n_pushlo = lo; 2844 if (hi > np->n_pushhi) 2845 np->n_pushhi = hi; 2846 } 2847 #ifdef NFS_DEBUG_COMMIT 2848 printf("add: tobecommitted: %u - %u\n", (unsigned)np->n_pushlo, 2849 (unsigned)np->n_pushhi); 2850 #endif 2851 } 2852 2853 void 2854 nfs_del_tobecommitted_range(vp, off, len) 2855 struct vnode *vp; 2856 off_t off, len; 2857 { 2858 struct nfsnode *np = VTONFS(vp); 2859 off_t lo, hi; 2860 2861 if (!(np->n_commitflags & NFS_COMMIT_PUSH_VALID)) 2862 return; 2863 2864 lo = off; 2865 hi = lo + len; 2866 2867 if (lo > np->n_pushhi || hi < np->n_pushlo) 2868 return; 2869 2870 if (lo <= np->n_pushlo) 2871 np->n_pushlo = hi; 2872 else if (hi >= np->n_pushhi) 2873 np->n_pushhi = lo; 2874 else { 2875 /* 2876 * XXX There's only one range. If the deleted range 2877 * is in the middle, pick the largest of the 2878 * contiguous ranges that it leaves. 2879 */ 2880 if ((np->n_pushlo - lo) > (hi - np->n_pushhi)) 2881 np->n_pushhi = lo; 2882 else 2883 np->n_pushlo = hi; 2884 } 2885 #ifdef NFS_DEBUG_COMMIT 2886 printf("del: tobecommitted: %u - %u\n", (unsigned)np->n_pushlo, 2887 (unsigned)np->n_pushhi); 2888 #endif 2889 } 2890 2891 /* 2892 * Map errnos to NFS error numbers. For Version 3 also filter out error 2893 * numbers not specified for the associated procedure. 2894 */ 2895 int 2896 nfsrv_errmap(nd, err) 2897 struct nfsrv_descript *nd; 2898 int err; 2899 { 2900 const short *defaulterrp, *errp; 2901 2902 if (nd->nd_flag & ND_NFSV3) { 2903 if (nd->nd_procnum <= NFSPROC_COMMIT) { 2904 errp = defaulterrp = nfsrv_v3errmap[nd->nd_procnum]; 2905 while (*++errp) { 2906 if (*errp == err) 2907 return (err); 2908 else if (*errp > err) 2909 break; 2910 } 2911 return ((int)*defaulterrp); 2912 } else 2913 return (err & 0xffff); 2914 } 2915 if (err <= ELAST) 2916 return ((int)nfsrv_v2errmap[err - 1]); 2917 return (NFSERR_IO); 2918 } 2919 2920 u_int32_t 2921 nfs_getxid() 2922 { 2923 static u_int32_t base; 2924 static u_int32_t nfs_xid = 0; 2925 u_int32_t newxid; 2926 2927 mutex_enter(&nfs_xidlock); 2928 /* 2929 * derive initial xid from system time 2930 * XXX time is invalid if root not yet mounted 2931 */ 2932 if (__predict_false(!base && (rootvp))) { 2933 struct timeval tv; 2934 2935 microtime(&tv); 2936 base = tv.tv_sec << 12; 2937 nfs_xid = base; 2938 } 2939 2940 /* 2941 * Skip zero xid if it should ever happen. 2942 */ 2943 if (__predict_false(++nfs_xid == 0)) 2944 nfs_xid++; 2945 newxid = nfs_xid; 2946 mutex_exit(&nfs_xidlock); 2947 2948 return txdr_unsigned(newxid); 2949 } 2950 2951 /* 2952 * assign a new xid for existing request. 2953 * used for NFSERR_JUKEBOX handling. 2954 */ 2955 void 2956 nfs_renewxid(struct nfsreq *req) 2957 { 2958 u_int32_t xid; 2959 int off; 2960 2961 xid = nfs_getxid(); 2962 if (req->r_nmp->nm_sotype == SOCK_STREAM) 2963 off = sizeof(u_int32_t); /* RPC record mark */ 2964 else 2965 off = 0; 2966 2967 m_copyback(req->r_mreq, off, sizeof(xid), (void *)&xid); 2968 req->r_xid = xid; 2969 } 2970 2971 #if defined(NFSSERVER) 2972 int 2973 nfsrv_composefh(struct vnode *vp, nfsrvfh_t *nsfh, bool v3) 2974 { 2975 int error; 2976 size_t fhsize; 2977 2978 fhsize = NFSD_MAXFHSIZE; 2979 error = vfs_composefh(vp, (void *)NFSRVFH_DATA(nsfh), &fhsize); 2980 if (NFSX_FHTOOBIG_P(fhsize, v3)) { 2981 error = EOPNOTSUPP; 2982 } 2983 if (error != 0) { 2984 return error; 2985 } 2986 if (!v3 && fhsize < NFSX_V2FH) { 2987 memset((char *)NFSRVFH_DATA(nsfh) + fhsize, 0, 2988 NFSX_V2FH - fhsize); 2989 fhsize = NFSX_V2FH; 2990 } 2991 if ((fhsize % NFSX_UNSIGNED) != 0) { 2992 return EOPNOTSUPP; 2993 } 2994 nsfh->nsfh_size = fhsize; 2995 return 0; 2996 } 2997 2998 int 2999 nfsrv_comparefh(const nfsrvfh_t *fh1, const nfsrvfh_t *fh2) 3000 { 3001 3002 if (NFSRVFH_SIZE(fh1) != NFSRVFH_SIZE(fh2)) { 3003 return NFSRVFH_SIZE(fh2) - NFSRVFH_SIZE(fh1); 3004 } 3005 return memcmp(NFSRVFH_DATA(fh1), NFSRVFH_DATA(fh2), NFSRVFH_SIZE(fh1)); 3006 } 3007 3008 void 3009 nfsrv_copyfh(nfsrvfh_t *fh1, const nfsrvfh_t *fh2) 3010 { 3011 size_t size; 3012 3013 fh1->nsfh_size = size = NFSRVFH_SIZE(fh2); 3014 memcpy(NFSRVFH_DATA(fh1), NFSRVFH_DATA(fh2), size); 3015 } 3016 #endif /* defined(NFSSERVER) */ 3017 3018 #if defined(NFS) 3019 /* 3020 * Set the attribute timeout based on how recently the file has been modified. 3021 */ 3022 3023 time_t 3024 nfs_attrtimeo(struct nfsmount *nmp, struct nfsnode *np) 3025 { 3026 time_t timeo; 3027 3028 if ((nmp->nm_flag & NFSMNT_NOAC) != 0) 3029 return 0; 3030 3031 if (((np)->n_flag & NMODIFIED) != 0) 3032 return NFS_MINATTRTIMO; 3033 3034 timeo = (time_second - np->n_mtime.tv_sec) / 10; 3035 timeo = max(timeo, NFS_MINATTRTIMO); 3036 timeo = min(timeo, NFS_MAXATTRTIMO); 3037 return timeo; 3038 } 3039 #endif /* defined(NFS) */ 3040
Indexes created Thu Oct 02 14:10:14 GMT 2025