1 /*- 2 * Copyright (c) 2005, 2006 The NetBSD Foundation, Inc. 3 * All rights reserved. 4 * 5 * This code is derived from software contributed to The NetBSD Foundation 6 * by Julio M. Merino Vidal, developed as part of Google's Summer of Code 7 * 2005 program. 8 * 9 * Redistribution and use in source and binary forms, with or without 10 * modification, are permitted provided that the following conditions 11 * are met: 12 * 1. Redistributions of source code must retain the above copyright 13 * notice, this list of conditions and the following disclaimer. 14 * 2. Redistributions in binary form must reproduce the above copyright 15 * notice, this list of conditions and the following disclaimer in the 16 * documentation and/or other materials provided with the distribution. 17 * 18 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 19 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 20 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 21 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 22 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 23 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 24 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 25 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 26 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 27 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 28 * POSSIBILITY OF SUCH DAMAGE. 29 * 30 * $NetBSD: tmpfs_vnops.c,v 1.39 2007/07/23 15:41:01 jmmv Exp $ 31 */ 32 33 /* 34 * tmpfs vnode interface. 35 */ 36 37 #include <sys/kernel.h> 38 #include <sys/kern_syscall.h> 39 #include <sys/param.h> 40 #include <sys/fcntl.h> 41 #include <sys/lockf.h> 42 #include <sys/priv.h> 43 #include <sys/proc.h> 44 #include <sys/resourcevar.h> 45 #include <sys/sched.h> 46 #include <sys/stat.h> 47 #include <sys/systm.h> 48 #include <sys/unistd.h> 49 #include <sys/vfsops.h> 50 #include <sys/vnode.h> 51 #include <sys/mountctl.h> 52 53 #include <vm/vm.h> 54 #include <vm/vm_object.h> 55 #include <vm/vm_page.h> 56 #include <vm/vm_pager.h> 57 #include <vm/swap_pager.h> 58 59 #include <sys/buf2.h> 60 61 #include <vfs/fifofs/fifo.h> 62 #include <vfs/tmpfs/tmpfs_vnops.h> 63 #include <vfs/tmpfs/tmpfs.h> 64 65 MALLOC_DECLARE(M_TMPFS); 66 67 static void tmpfs_strategy_done(struct bio *bio); 68 69 static __inline 70 void 71 tmpfs_knote(struct vnode *vp, int flags) 72 { 73 if (flags) 74 KNOTE(&vp->v_pollinfo.vpi_kqinfo.ki_note, flags); 75 } 76 77 78 /* --------------------------------------------------------------------- */ 79 80 static int 81 tmpfs_nresolve(struct vop_nresolve_args *v) 82 { 83 struct vnode *dvp = v->a_dvp; 84 struct vnode *vp = NULL; 85 struct namecache *ncp = v->a_nch->ncp; 86 struct tmpfs_node *tnode; 87 88 int error; 89 struct tmpfs_dirent *de; 90 struct tmpfs_node *dnode; 91 92 dnode = VP_TO_TMPFS_DIR(dvp); 93 94 de = tmpfs_dir_lookup(dnode, NULL, ncp); 95 if (de == NULL) { 96 error = ENOENT; 97 } else { 98 /* 99 * Allocate a vnode for the node we found. 100 */ 101 tnode = de->td_node; 102 error = tmpfs_alloc_vp(dvp->v_mount, tnode, 103 LK_EXCLUSIVE | LK_RETRY, &vp); 104 if (error) 105 goto out; 106 KKASSERT(vp); 107 } 108 109 out: 110 /* 111 * Store the result of this lookup in the cache. Avoid this if the 112 * request was for creation, as it does not improve timings on 113 * emprical tests. 114 */ 115 if (vp) { 116 vn_unlock(vp); 117 cache_setvp(v->a_nch, vp); 118 vrele(vp); 119 } else if (error == ENOENT) { 120 cache_setvp(v->a_nch, NULL); 121 } 122 return error; 123 } 124 125 static int 126 tmpfs_nlookupdotdot(struct vop_nlookupdotdot_args *v) 127 { 128 struct vnode *dvp = v->a_dvp; 129 struct vnode **vpp = v->a_vpp; 130 struct tmpfs_node *dnode = VP_TO_TMPFS_NODE(dvp); 131 struct ucred *cred = v->a_cred; 132 int error; 133 134 *vpp = NULL; 135 /* Check accessibility of requested node as a first step. */ 136 error = VOP_ACCESS(dvp, VEXEC, cred); 137 if (error != 0) 138 return error; 139 140 if (dnode->tn_dir.tn_parent != NULL) { 141 /* Allocate a new vnode on the matching entry. */ 142 error = tmpfs_alloc_vp(dvp->v_mount, dnode->tn_dir.tn_parent, 143 LK_EXCLUSIVE | LK_RETRY, vpp); 144 145 if (*vpp) 146 vn_unlock(*vpp); 147 } 148 149 return (*vpp == NULL) ? ENOENT : 0; 150 } 151 152 /* --------------------------------------------------------------------- */ 153 154 static int 155 tmpfs_ncreate(struct vop_ncreate_args *v) 156 { 157 struct vnode *dvp = v->a_dvp; 158 struct vnode **vpp = v->a_vpp; 159 struct namecache *ncp = v->a_nch->ncp; 160 struct vattr *vap = v->a_vap; 161 struct ucred *cred = v->a_cred; 162 int error; 163 164 KKASSERT(vap->va_type == VREG || vap->va_type == VSOCK); 165 166 error = tmpfs_alloc_file(dvp, vpp, vap, ncp, cred, NULL); 167 if (error == 0) { 168 cache_setunresolved(v->a_nch); 169 cache_setvp(v->a_nch, *vpp); 170 tmpfs_knote(dvp, NOTE_WRITE); 171 } 172 173 return error; 174 } 175 /* --------------------------------------------------------------------- */ 176 177 static int 178 tmpfs_nmknod(struct vop_nmknod_args *v) 179 { 180 struct vnode *dvp = v->a_dvp; 181 struct vnode **vpp = v->a_vpp; 182 struct namecache *ncp = v->a_nch->ncp; 183 struct vattr *vap = v->a_vap; 184 struct ucred *cred = v->a_cred; 185 int error; 186 187 if (vap->va_type != VBLK && vap->va_type != VCHR && 188 vap->va_type != VFIFO) 189 return EINVAL; 190 191 error = tmpfs_alloc_file(dvp, vpp, vap, ncp, cred, NULL); 192 if (error == 0) { 193 cache_setunresolved(v->a_nch); 194 cache_setvp(v->a_nch, *vpp); 195 tmpfs_knote(dvp, NOTE_WRITE); 196 } 197 198 return error; 199 } 200 201 /* --------------------------------------------------------------------- */ 202 203 static int 204 tmpfs_open(struct vop_open_args *v) 205 { 206 struct vnode *vp = v->a_vp; 207 int mode = v->a_mode; 208 209 int error; 210 struct tmpfs_node *node; 211 212 node = VP_TO_TMPFS_NODE(vp); 213 214 /* The file is still active but all its names have been removed 215 * (e.g. by a "rmdir $(pwd)"). It cannot be opened any more as 216 * it is about to die. */ 217 if (node->tn_links < 1) 218 return (ENOENT); 219 220 /* If the file is marked append-only, deny write requests. */ 221 if ((node->tn_flags & APPEND) && 222 (mode & (FWRITE | O_APPEND)) == FWRITE) { 223 error = EPERM; 224 } else { 225 return (vop_stdopen(v)); 226 } 227 return error; 228 } 229 230 /* --------------------------------------------------------------------- */ 231 232 static int 233 tmpfs_close(struct vop_close_args *v) 234 { 235 struct vnode *vp = v->a_vp; 236 struct tmpfs_node *node; 237 238 node = VP_TO_TMPFS_NODE(vp); 239 240 if (node->tn_links > 0) { 241 /* Update node times. No need to do it if the node has 242 * been deleted, because it will vanish after we return. */ 243 tmpfs_update(vp); 244 } 245 246 return vop_stdclose(v); 247 } 248 249 /* --------------------------------------------------------------------- */ 250 251 int 252 tmpfs_access(struct vop_access_args *v) 253 { 254 struct vnode *vp = v->a_vp; 255 int error; 256 struct tmpfs_node *node; 257 258 node = VP_TO_TMPFS_NODE(vp); 259 260 switch (vp->v_type) { 261 case VDIR: 262 /* FALLTHROUGH */ 263 case VLNK: 264 /* FALLTHROUGH */ 265 case VREG: 266 if ((v->a_mode & VWRITE) && (vp->v_mount->mnt_flag & MNT_RDONLY)) { 267 error = EROFS; 268 goto out; 269 } 270 break; 271 272 case VBLK: 273 /* FALLTHROUGH */ 274 case VCHR: 275 /* FALLTHROUGH */ 276 case VSOCK: 277 /* FALLTHROUGH */ 278 case VFIFO: 279 break; 280 281 default: 282 error = EINVAL; 283 goto out; 284 } 285 286 if ((v->a_mode & VWRITE) && (node->tn_flags & IMMUTABLE)) { 287 error = EPERM; 288 goto out; 289 } 290 291 error = vop_helper_access(v, node->tn_uid, node->tn_gid, node->tn_mode, 0); 292 293 out: 294 295 return error; 296 } 297 298 /* --------------------------------------------------------------------- */ 299 300 int 301 tmpfs_getattr(struct vop_getattr_args *v) 302 { 303 struct vnode *vp = v->a_vp; 304 struct vattr *vap = v->a_vap; 305 struct tmpfs_node *node; 306 307 node = VP_TO_TMPFS_NODE(vp); 308 309 lwkt_gettoken(&vp->v_mount->mnt_token); 310 tmpfs_update(vp); 311 312 vap->va_type = vp->v_type; 313 vap->va_mode = node->tn_mode; 314 vap->va_nlink = node->tn_links; 315 vap->va_uid = node->tn_uid; 316 vap->va_gid = node->tn_gid; 317 vap->va_fsid = vp->v_mount->mnt_stat.f_fsid.val[0]; 318 vap->va_fileid = node->tn_id; 319 vap->va_size = node->tn_size; 320 vap->va_blocksize = PAGE_SIZE; 321 vap->va_atime.tv_sec = node->tn_atime; 322 vap->va_atime.tv_nsec = node->tn_atimensec; 323 vap->va_mtime.tv_sec = node->tn_mtime; 324 vap->va_mtime.tv_nsec = node->tn_mtimensec; 325 vap->va_ctime.tv_sec = node->tn_ctime; 326 vap->va_ctime.tv_nsec = node->tn_ctimensec; 327 vap->va_gen = node->tn_gen; 328 vap->va_flags = node->tn_flags; 329 if (vp->v_type == VBLK || vp->v_type == VCHR) 330 { 331 vap->va_rmajor = umajor(node->tn_rdev); 332 vap->va_rminor = uminor(node->tn_rdev); 333 } 334 vap->va_bytes = round_page(node->tn_size); 335 vap->va_filerev = 0; 336 337 lwkt_reltoken(&vp->v_mount->mnt_token); 338 339 return 0; 340 } 341 342 /* --------------------------------------------------------------------- */ 343 344 int 345 tmpfs_setattr(struct vop_setattr_args *v) 346 { 347 struct vnode *vp = v->a_vp; 348 struct vattr *vap = v->a_vap; 349 struct ucred *cred = v->a_cred; 350 struct tmpfs_node *node = VP_TO_TMPFS_NODE(vp); 351 int error = 0; 352 int kflags = 0; 353 354 if (error == 0 && (vap->va_flags != VNOVAL)) { 355 error = tmpfs_chflags(vp, vap->va_flags, cred); 356 kflags |= NOTE_ATTRIB; 357 } 358 359 if (error == 0 && (vap->va_size != VNOVAL)) { 360 if (vap->va_size > node->tn_size) 361 kflags |= NOTE_WRITE | NOTE_EXTEND; 362 else 363 kflags |= NOTE_WRITE; 364 error = tmpfs_chsize(vp, vap->va_size, cred); 365 } 366 367 if (error == 0 && (vap->va_uid != (uid_t)VNOVAL || 368 vap->va_gid != (gid_t)VNOVAL)) { 369 error = tmpfs_chown(vp, vap->va_uid, vap->va_gid, cred); 370 kflags |= NOTE_ATTRIB; 371 } 372 373 if (error == 0 && (vap->va_mode != (mode_t)VNOVAL)) { 374 error = tmpfs_chmod(vp, vap->va_mode, cred); 375 kflags |= NOTE_ATTRIB; 376 } 377 378 if (error == 0 && ((vap->va_atime.tv_sec != VNOVAL && 379 vap->va_atime.tv_nsec != VNOVAL) || 380 (vap->va_mtime.tv_sec != VNOVAL && 381 vap->va_mtime.tv_nsec != VNOVAL) )) { 382 error = tmpfs_chtimes(vp, &vap->va_atime, &vap->va_mtime, 383 vap->va_vaflags, cred); 384 kflags |= NOTE_ATTRIB; 385 } 386 387 /* Update the node times. We give preference to the error codes 388 * generated by this function rather than the ones that may arise 389 * from tmpfs_update. */ 390 tmpfs_update(vp); 391 tmpfs_knote(vp, kflags); 392 393 return error; 394 } 395 396 /* --------------------------------------------------------------------- */ 397 398 /* 399 * fsync is usually a NOP, but we must take action when unmounting or 400 * when recycling. 401 */ 402 static int 403 tmpfs_fsync(struct vop_fsync_args *v) 404 { 405 struct tmpfs_node *node; 406 struct vnode *vp = v->a_vp; 407 408 node = VP_TO_TMPFS_NODE(vp); 409 410 tmpfs_update(vp); 411 if (vp->v_type == VREG) { 412 if (vp->v_flag & VRECLAIMED) { 413 if (node->tn_links == 0) 414 tmpfs_truncate(vp, 0); 415 else 416 vfsync(v->a_vp, v->a_waitfor, 1, NULL, NULL); 417 } 418 } 419 return 0; 420 } 421 422 /* --------------------------------------------------------------------- */ 423 424 static int 425 tmpfs_read (struct vop_read_args *ap) 426 { 427 struct buf *bp; 428 struct vnode *vp = ap->a_vp; 429 struct uio *uio = ap->a_uio; 430 struct tmpfs_node *node; 431 off_t base_offset; 432 size_t offset; 433 size_t len; 434 int error; 435 436 error = 0; 437 if (uio->uio_resid == 0) { 438 return error; 439 } 440 441 node = VP_TO_TMPFS_NODE(vp); 442 443 if (uio->uio_offset < 0) 444 return (EINVAL); 445 if (vp->v_type != VREG) 446 return (EINVAL); 447 448 while (uio->uio_resid > 0 && uio->uio_offset < node->tn_size) { 449 /* 450 * Use buffer cache I/O (via tmpfs_strategy) 451 */ 452 offset = (size_t)uio->uio_offset & BMASK; 453 base_offset = (off_t)uio->uio_offset - offset; 454 bp = getcacheblk(vp, base_offset, BSIZE, 0); 455 if (bp == NULL) { 456 lwkt_gettoken(&vp->v_mount->mnt_token); 457 error = bread(vp, base_offset, BSIZE, &bp); 458 if (error) { 459 brelse(bp); 460 lwkt_reltoken(&vp->v_mount->mnt_token); 461 kprintf("tmpfs_read bread error %d\n", error); 462 break; 463 } 464 lwkt_reltoken(&vp->v_mount->mnt_token); 465 } 466 467 /* 468 * Figure out how many bytes we can actually copy this loop. 469 */ 470 len = BSIZE - offset; 471 if (len > uio->uio_resid) 472 len = uio->uio_resid; 473 if (len > node->tn_size - uio->uio_offset) 474 len = (size_t)(node->tn_size - uio->uio_offset); 475 476 error = uiomove((char *)bp->b_data + offset, len, uio); 477 bqrelse(bp); 478 if (error) { 479 kprintf("tmpfs_read uiomove error %d\n", error); 480 break; 481 } 482 } 483 484 TMPFS_NODE_LOCK(node); 485 node->tn_status |= TMPFS_NODE_ACCESSED; 486 TMPFS_NODE_UNLOCK(node); 487 488 return(error); 489 } 490 491 static int 492 tmpfs_write (struct vop_write_args *ap) 493 { 494 struct buf *bp; 495 struct vnode *vp = ap->a_vp; 496 struct uio *uio = ap->a_uio; 497 struct thread *td = uio->uio_td; 498 struct tmpfs_node *node; 499 boolean_t extended; 500 off_t oldsize; 501 int error; 502 off_t base_offset; 503 size_t offset; 504 size_t len; 505 struct rlimit limit; 506 int trivial = 0; 507 int kflags = 0; 508 509 error = 0; 510 if (uio->uio_resid == 0) { 511 return error; 512 } 513 514 node = VP_TO_TMPFS_NODE(vp); 515 516 if (vp->v_type != VREG) 517 return (EINVAL); 518 519 lwkt_gettoken(&vp->v_mount->mnt_token); 520 521 oldsize = node->tn_size; 522 if (ap->a_ioflag & IO_APPEND) 523 uio->uio_offset = node->tn_size; 524 525 /* 526 * Check for illegal write offsets. 527 */ 528 if (uio->uio_offset + uio->uio_resid > 529 VFS_TO_TMPFS(vp->v_mount)->tm_maxfilesize) { 530 lwkt_reltoken(&vp->v_mount->mnt_token); 531 return (EFBIG); 532 } 533 534 if (vp->v_type == VREG && td != NULL) { 535 error = kern_getrlimit(RLIMIT_FSIZE, &limit); 536 if (error != 0) { 537 lwkt_reltoken(&vp->v_mount->mnt_token); 538 return error; 539 } 540 if (uio->uio_offset + uio->uio_resid > limit.rlim_cur) { 541 ksignal(td->td_proc, SIGXFSZ); 542 lwkt_reltoken(&vp->v_mount->mnt_token); 543 return (EFBIG); 544 } 545 } 546 547 548 /* 549 * Extend the file's size if necessary 550 */ 551 extended = ((uio->uio_offset + uio->uio_resid) > node->tn_size); 552 553 while (uio->uio_resid > 0) { 554 /* 555 * Use buffer cache I/O (via tmpfs_strategy) 556 */ 557 offset = (size_t)uio->uio_offset & BMASK; 558 base_offset = (off_t)uio->uio_offset - offset; 559 len = BSIZE - offset; 560 if (len > uio->uio_resid) 561 len = uio->uio_resid; 562 563 if ((uio->uio_offset + len) > node->tn_size) { 564 trivial = (uio->uio_offset <= node->tn_size); 565 error = tmpfs_reg_resize(vp, uio->uio_offset + len, trivial); 566 if (error) 567 break; 568 } 569 570 /* 571 * Read to fill in any gaps. Theoretically we could 572 * optimize this if the write covers the entire buffer 573 * and is not a UIO_NOCOPY write, however this can lead 574 * to a security violation exposing random kernel memory 575 * (whatever junk was in the backing VM pages before). 576 * 577 * So just use bread() to do the right thing. 578 */ 579 error = bread(vp, base_offset, BSIZE, &bp); 580 error = uiomove((char *)bp->b_data + offset, len, uio); 581 if (error) { 582 kprintf("tmpfs_write uiomove error %d\n", error); 583 brelse(bp); 584 break; 585 } 586 587 if (uio->uio_offset > node->tn_size) { 588 node->tn_size = uio->uio_offset; 589 kflags |= NOTE_EXTEND; 590 } 591 kflags |= NOTE_WRITE; 592 593 /* 594 * Always try to flush the page if the request is coming 595 * from the pageout daemon (IO_ASYNC), else buwrite() the 596 * buffer. 597 * 598 * buwrite() dirties the underlying VM pages instead of 599 * dirtying the buffer, releasing the buffer as a clean 600 * buffer. This allows tmpfs to use essentially all 601 * available memory to cache file data. If we used bdwrite() 602 * the buffer cache would wind up flushing the data to 603 * swap too quickly. 604 */ 605 bp->b_flags |= B_AGE; 606 if (ap->a_ioflag & IO_ASYNC) { 607 bawrite(bp); 608 } else { 609 buwrite(bp); 610 } 611 612 if (bp->b_error) { 613 kprintf("tmpfs_write bwrite error %d\n", bp->b_error); 614 break; 615 } 616 } 617 618 if (error) { 619 if (extended) { 620 (void)tmpfs_reg_resize(vp, oldsize, trivial); 621 kflags &= ~NOTE_EXTEND; 622 } 623 goto done; 624 } 625 626 /* 627 * Currently we don't set the mtime on files modified via mmap() 628 * because we can't tell the difference between those modifications 629 * and an attempt by the pageout daemon to flush tmpfs pages to 630 * swap. 631 * 632 * This is because in order to defer flushes as long as possible 633 * buwrite() works by marking the underlying VM pages dirty in 634 * order to be able to dispose of the buffer cache buffer without 635 * flushing it. 636 */ 637 TMPFS_NODE_LOCK(node); 638 if (uio->uio_segflg != UIO_NOCOPY) 639 node->tn_status |= TMPFS_NODE_ACCESSED | TMPFS_NODE_MODIFIED; 640 if (extended) 641 node->tn_status |= TMPFS_NODE_CHANGED; 642 643 if (node->tn_mode & (S_ISUID | S_ISGID)) { 644 if (priv_check_cred(ap->a_cred, PRIV_VFS_RETAINSUGID, 0)) 645 node->tn_mode &= ~(S_ISUID | S_ISGID); 646 } 647 TMPFS_NODE_UNLOCK(node); 648 done: 649 650 tmpfs_knote(vp, kflags); 651 652 653 lwkt_reltoken(&vp->v_mount->mnt_token); 654 return(error); 655 } 656 657 static int 658 tmpfs_advlock (struct vop_advlock_args *ap) 659 { 660 struct tmpfs_node *node; 661 struct vnode *vp = ap->a_vp; 662 663 node = VP_TO_TMPFS_NODE(vp); 664 665 return (lf_advlock(ap, &node->tn_advlock, node->tn_size)); 666 } 667 668 /* 669 * The strategy function is typically only called when memory pressure 670 * forces the system to attempt to pageout pages. It can also be called 671 * by [n]vtruncbuf() when a truncation cuts a page in half. Normal write 672 * operations 673 */ 674 static int 675 tmpfs_strategy(struct vop_strategy_args *ap) 676 { 677 struct bio *bio = ap->a_bio; 678 struct bio *nbio; 679 struct buf *bp = bio->bio_buf; 680 struct vnode *vp = ap->a_vp; 681 struct tmpfs_node *node; 682 vm_object_t uobj; 683 vm_page_t m; 684 int i; 685 686 if (vp->v_type != VREG) { 687 bp->b_resid = bp->b_bcount; 688 bp->b_flags |= B_ERROR | B_INVAL; 689 bp->b_error = EINVAL; 690 biodone(bio); 691 return(0); 692 } 693 694 lwkt_gettoken(&vp->v_mount->mnt_token); 695 node = VP_TO_TMPFS_NODE(vp); 696 697 uobj = node->tn_reg.tn_aobj; 698 699 /* 700 * Don't bother flushing to swap if there is no swap, just 701 * ensure that the pages are marked as needing a commit (still). 702 */ 703 if (bp->b_cmd == BUF_CMD_WRITE && vm_swap_size == 0) { 704 for (i = 0; i < bp->b_xio.xio_npages; ++i) { 705 m = bp->b_xio.xio_pages[i]; 706 vm_page_need_commit(m); 707 } 708 bp->b_resid = 0; 709 bp->b_error = 0; 710 biodone(bio); 711 } else { 712 nbio = push_bio(bio); 713 nbio->bio_done = tmpfs_strategy_done; 714 nbio->bio_offset = bio->bio_offset; 715 swap_pager_strategy(uobj, nbio); 716 } 717 718 lwkt_reltoken(&vp->v_mount->mnt_token); 719 return 0; 720 } 721 722 /* 723 * If we were unable to commit the pages to swap make sure they are marked 724 * as needing a commit (again). If we were, clear the flag to allow the 725 * pages to be freed. 726 */ 727 static void 728 tmpfs_strategy_done(struct bio *bio) 729 { 730 struct buf *bp; 731 vm_page_t m; 732 int i; 733 734 bp = bio->bio_buf; 735 736 if (bp->b_flags & B_ERROR) { 737 bp->b_flags &= ~B_ERROR; 738 bp->b_error = 0; 739 bp->b_resid = 0; 740 for (i = 0; i < bp->b_xio.xio_npages; ++i) { 741 m = bp->b_xio.xio_pages[i]; 742 vm_page_need_commit(m); 743 } 744 } else { 745 for (i = 0; i < bp->b_xio.xio_npages; ++i) { 746 m = bp->b_xio.xio_pages[i]; 747 vm_page_clear_commit(m); 748 } 749 } 750 bio = pop_bio(bio); 751 biodone(bio); 752 } 753 754 static int 755 tmpfs_bmap(struct vop_bmap_args *ap) 756 { 757 if (ap->a_doffsetp != NULL) 758 *ap->a_doffsetp = ap->a_loffset; 759 if (ap->a_runp != NULL) 760 *ap->a_runp = 0; 761 if (ap->a_runb != NULL) 762 *ap->a_runb = 0; 763 764 return 0; 765 } 766 767 /* --------------------------------------------------------------------- */ 768 769 static int 770 tmpfs_nremove(struct vop_nremove_args *v) 771 { 772 struct vnode *dvp = v->a_dvp; 773 struct namecache *ncp = v->a_nch->ncp; 774 struct vnode *vp; 775 int error; 776 struct tmpfs_dirent *de; 777 struct tmpfs_mount *tmp; 778 struct tmpfs_node *dnode; 779 struct tmpfs_node *node; 780 781 /* 782 * We have to acquire the vp from v->a_nch because we will likely 783 * unresolve the namecache entry, and a vrele/vput is needed to 784 * trigger the tmpfs_inactive/tmpfs_reclaim sequence. 785 * 786 * We have to use vget to clear any inactive state on the vnode, 787 * otherwise the vnode may remain inactive and thus tmpfs_inactive 788 * will not get called when we release it. 789 */ 790 error = cache_vget(v->a_nch, v->a_cred, LK_SHARED, &vp); 791 KKASSERT(error == 0); 792 vn_unlock(vp); 793 794 if (vp->v_type == VDIR) { 795 error = EISDIR; 796 goto out; 797 } 798 799 dnode = VP_TO_TMPFS_DIR(dvp); 800 node = VP_TO_TMPFS_NODE(vp); 801 tmp = VFS_TO_TMPFS(vp->v_mount); 802 de = tmpfs_dir_lookup(dnode, node, ncp); 803 if (de == NULL) { 804 error = ENOENT; 805 goto out; 806 } 807 808 /* Files marked as immutable or append-only cannot be deleted. */ 809 if ((node->tn_flags & (IMMUTABLE | APPEND | NOUNLINK)) || 810 (dnode->tn_flags & APPEND)) { 811 error = EPERM; 812 goto out; 813 } 814 815 /* Remove the entry from the directory; as it is a file, we do not 816 * have to change the number of hard links of the directory. */ 817 tmpfs_dir_detach(dnode, de); 818 819 /* Free the directory entry we just deleted. Note that the node 820 * referred by it will not be removed until the vnode is really 821 * reclaimed. */ 822 tmpfs_free_dirent(tmp, de); 823 824 if (node->tn_links > 0) { 825 TMPFS_NODE_LOCK(node); 826 node->tn_status |= TMPFS_NODE_ACCESSED | TMPFS_NODE_CHANGED | \ 827 TMPFS_NODE_MODIFIED; 828 TMPFS_NODE_UNLOCK(node); 829 } 830 831 cache_setunresolved(v->a_nch); 832 cache_setvp(v->a_nch, NULL); 833 tmpfs_knote(vp, NOTE_DELETE); 834 /*cache_inval_vp(vp, CINV_DESTROY);*/ 835 tmpfs_knote(dvp, NOTE_WRITE); 836 error = 0; 837 838 out: 839 vrele(vp); 840 841 return error; 842 } 843 844 /* --------------------------------------------------------------------- */ 845 846 static int 847 tmpfs_nlink(struct vop_nlink_args *v) 848 { 849 struct vnode *dvp = v->a_dvp; 850 struct vnode *vp = v->a_vp; 851 struct namecache *ncp = v->a_nch->ncp; 852 struct tmpfs_dirent *de; 853 struct tmpfs_node *node; 854 struct tmpfs_node *dnode; 855 int error; 856 857 KKASSERT(dvp != vp); /* XXX When can this be false? */ 858 859 node = VP_TO_TMPFS_NODE(vp); 860 dnode = VP_TO_TMPFS_NODE(dvp); 861 862 /* XXX: Why aren't the following two tests done by the caller? */ 863 864 /* Hard links of directories are forbidden. */ 865 if (vp->v_type == VDIR) { 866 error = EPERM; 867 goto out; 868 } 869 870 /* Cannot create cross-device links. */ 871 if (dvp->v_mount != vp->v_mount) { 872 error = EXDEV; 873 goto out; 874 } 875 876 /* Ensure that we do not overflow the maximum number of links imposed 877 * by the system. */ 878 KKASSERT(node->tn_links <= LINK_MAX); 879 if (node->tn_links == LINK_MAX) { 880 error = EMLINK; 881 goto out; 882 } 883 884 /* We cannot create links of files marked immutable or append-only. */ 885 if (node->tn_flags & (IMMUTABLE | APPEND)) { 886 error = EPERM; 887 goto out; 888 } 889 890 /* Allocate a new directory entry to represent the node. */ 891 error = tmpfs_alloc_dirent(VFS_TO_TMPFS(vp->v_mount), node, 892 ncp->nc_name, ncp->nc_nlen, &de); 893 if (error != 0) 894 goto out; 895 896 /* Insert the new directory entry into the appropriate directory. */ 897 tmpfs_dir_attach(dnode, de); 898 899 /* vp link count has changed, so update node times. */ 900 901 TMPFS_NODE_LOCK(node); 902 node->tn_status |= TMPFS_NODE_CHANGED; 903 TMPFS_NODE_UNLOCK(node); 904 tmpfs_update(vp); 905 906 tmpfs_knote(vp, NOTE_LINK); 907 cache_setunresolved(v->a_nch); 908 cache_setvp(v->a_nch, vp); 909 tmpfs_knote(dvp, NOTE_WRITE); 910 error = 0; 911 912 out: 913 return error; 914 } 915 916 /* --------------------------------------------------------------------- */ 917 918 static int 919 tmpfs_nrename(struct vop_nrename_args *v) 920 { 921 struct vnode *fdvp = v->a_fdvp; 922 struct namecache *fncp = v->a_fnch->ncp; 923 struct vnode *fvp = fncp->nc_vp; 924 struct vnode *tdvp = v->a_tdvp; 925 struct namecache *tncp = v->a_tnch->ncp; 926 struct vnode *tvp; 927 struct tmpfs_dirent *de; 928 struct tmpfs_mount *tmp; 929 struct tmpfs_node *fdnode; 930 struct tmpfs_node *fnode; 931 struct tmpfs_node *tnode; 932 struct tmpfs_node *tdnode; 933 char *newname; 934 char *oldname; 935 int error; 936 937 /* 938 * Because tvp can get overwritten we have to vget it instead of 939 * just vref or use it, otherwise it's VINACTIVE flag may not get 940 * cleared and the node won't get destroyed. 941 */ 942 error = cache_vget(v->a_tnch, v->a_cred, LK_SHARED, &tvp); 943 if (error == 0) { 944 tnode = VP_TO_TMPFS_NODE(tvp); 945 vn_unlock(tvp); 946 } else { 947 tnode = NULL; 948 } 949 950 /* Disallow cross-device renames. 951 * XXX Why isn't this done by the caller? */ 952 if (fvp->v_mount != tdvp->v_mount || 953 (tvp != NULL && fvp->v_mount != tvp->v_mount)) { 954 error = EXDEV; 955 goto out; 956 } 957 958 tmp = VFS_TO_TMPFS(tdvp->v_mount); 959 tdnode = VP_TO_TMPFS_DIR(tdvp); 960 961 /* If source and target are the same file, there is nothing to do. */ 962 if (fvp == tvp) { 963 error = 0; 964 goto out; 965 } 966 967 fdnode = VP_TO_TMPFS_DIR(fdvp); 968 fnode = VP_TO_TMPFS_NODE(fvp); 969 de = tmpfs_dir_lookup(fdnode, fnode, fncp); 970 971 /* Avoid manipulating '.' and '..' entries. */ 972 if (de == NULL) { 973 error = ENOENT; 974 goto out_locked; 975 } 976 KKASSERT(de->td_node == fnode); 977 978 /* 979 * If replacing an entry in the target directory and that entry 980 * is a directory, it must be empty. 981 * 982 * Kern_rename gurantees the destination to be a directory 983 * if the source is one (it does?). 984 */ 985 if (tvp != NULL) { 986 KKASSERT(tnode != NULL); 987 988 if ((tnode->tn_flags & (NOUNLINK | IMMUTABLE | APPEND)) || 989 (tdnode->tn_flags & (APPEND | IMMUTABLE))) { 990 error = EPERM; 991 goto out_locked; 992 } 993 994 if (fnode->tn_type == VDIR && tnode->tn_type == VDIR) { 995 if (tnode->tn_size > 0) { 996 error = ENOTEMPTY; 997 goto out_locked; 998 } 999 } else if (fnode->tn_type == VDIR && tnode->tn_type != VDIR) { 1000 error = ENOTDIR; 1001 goto out_locked; 1002 } else if (fnode->tn_type != VDIR && tnode->tn_type == VDIR) { 1003 error = EISDIR; 1004 goto out_locked; 1005 } else { 1006 KKASSERT(fnode->tn_type != VDIR && 1007 tnode->tn_type != VDIR); 1008 } 1009 } 1010 1011 if ((fnode->tn_flags & (NOUNLINK | IMMUTABLE | APPEND)) || 1012 (fdnode->tn_flags & (APPEND | IMMUTABLE))) { 1013 error = EPERM; 1014 goto out_locked; 1015 } 1016 1017 /* 1018 * Ensure that we have enough memory to hold the new name, if it 1019 * has to be changed. 1020 */ 1021 if (fncp->nc_nlen != tncp->nc_nlen || 1022 bcmp(fncp->nc_name, tncp->nc_name, fncp->nc_nlen) != 0) { 1023 newname = kmalloc(tncp->nc_nlen + 1, tmp->tm_name_zone, 1024 M_WAITOK | M_NULLOK); 1025 if (newname == NULL) { 1026 error = ENOSPC; 1027 goto out_locked; 1028 } 1029 bcopy(tncp->nc_name, newname, tncp->nc_nlen); 1030 newname[tncp->nc_nlen] = '\0'; 1031 } else { 1032 newname = NULL; 1033 } 1034 1035 /* 1036 * Unlink entry from source directory. Note that the kernel has 1037 * already checked for illegal recursion cases (renaming a directory 1038 * into a subdirectory of itself). 1039 */ 1040 if (fdnode != tdnode) 1041 tmpfs_dir_detach(fdnode, de); 1042 1043 /* 1044 * Handle any name change. Swap with newname, we will 1045 * deallocate it at the end. 1046 */ 1047 if (newname != NULL) { 1048 #if 0 1049 TMPFS_NODE_LOCK(fnode); 1050 fnode->tn_status |= TMPFS_NODE_CHANGED; 1051 TMPFS_NODE_UNLOCK(fnode); 1052 #endif 1053 oldname = de->td_name; 1054 de->td_name = newname; 1055 de->td_namelen = (uint16_t)tncp->nc_nlen; 1056 newname = oldname; 1057 } 1058 1059 /* 1060 * Link entry to target directory. If the entry 1061 * represents a directory move the parent linkage 1062 * as well. 1063 */ 1064 if (fdnode != tdnode) { 1065 if (de->td_node->tn_type == VDIR) { 1066 TMPFS_VALIDATE_DIR(fnode); 1067 1068 TMPFS_NODE_LOCK(tdnode); 1069 tdnode->tn_links++; 1070 tdnode->tn_status |= TMPFS_NODE_MODIFIED; 1071 TMPFS_NODE_UNLOCK(tdnode); 1072 1073 TMPFS_NODE_LOCK(fnode); 1074 fnode->tn_dir.tn_parent = tdnode; 1075 fnode->tn_status |= TMPFS_NODE_CHANGED; 1076 TMPFS_NODE_UNLOCK(fnode); 1077 1078 TMPFS_NODE_LOCK(fdnode); 1079 fdnode->tn_links--; 1080 fdnode->tn_status |= TMPFS_NODE_MODIFIED; 1081 TMPFS_NODE_UNLOCK(fdnode); 1082 } 1083 tmpfs_dir_attach(tdnode, de); 1084 } else { 1085 TMPFS_NODE_LOCK(tdnode); 1086 tdnode->tn_status |= TMPFS_NODE_MODIFIED; 1087 TMPFS_NODE_UNLOCK(tdnode); 1088 } 1089 1090 /* 1091 * If we are overwriting an entry, we have to remove the old one 1092 * from the target directory. 1093 */ 1094 if (tvp != NULL) { 1095 /* Remove the old entry from the target directory. */ 1096 de = tmpfs_dir_lookup(tdnode, tnode, tncp); 1097 tmpfs_dir_detach(tdnode, de); 1098 tmpfs_knote(tdnode->tn_vnode, NOTE_DELETE); 1099 1100 /* 1101 * Free the directory entry we just deleted. Note that the 1102 * node referred by it will not be removed until the vnode is 1103 * really reclaimed. 1104 */ 1105 tmpfs_free_dirent(VFS_TO_TMPFS(tvp->v_mount), de); 1106 /*cache_inval_vp(tvp, CINV_DESTROY);*/ 1107 } 1108 1109 /* 1110 * Finish up 1111 */ 1112 if (newname) { 1113 kfree(newname, tmp->tm_name_zone); 1114 newname = NULL; 1115 } 1116 cache_rename(v->a_fnch, v->a_tnch); 1117 tmpfs_knote(v->a_fdvp, NOTE_WRITE); 1118 tmpfs_knote(v->a_tdvp, NOTE_WRITE); 1119 if (fnode->tn_vnode) 1120 tmpfs_knote(fnode->tn_vnode, NOTE_RENAME); 1121 error = 0; 1122 1123 out_locked: 1124 ; 1125 1126 out: 1127 if (tvp) 1128 vrele(tvp); 1129 1130 return error; 1131 } 1132 1133 /* --------------------------------------------------------------------- */ 1134 1135 static int 1136 tmpfs_nmkdir(struct vop_nmkdir_args *v) 1137 { 1138 struct vnode *dvp = v->a_dvp; 1139 struct vnode **vpp = v->a_vpp; 1140 struct namecache *ncp = v->a_nch->ncp; 1141 struct vattr *vap = v->a_vap; 1142 struct ucred *cred = v->a_cred; 1143 int error; 1144 1145 KKASSERT(vap->va_type == VDIR); 1146 1147 error = tmpfs_alloc_file(dvp, vpp, vap, ncp, cred, NULL); 1148 if (error == 0) { 1149 cache_setunresolved(v->a_nch); 1150 cache_setvp(v->a_nch, *vpp); 1151 tmpfs_knote(dvp, NOTE_WRITE | NOTE_LINK); 1152 } 1153 1154 return error; 1155 } 1156 1157 /* --------------------------------------------------------------------- */ 1158 1159 static int 1160 tmpfs_nrmdir(struct vop_nrmdir_args *v) 1161 { 1162 struct vnode *dvp = v->a_dvp; 1163 struct namecache *ncp = v->a_nch->ncp; 1164 struct vnode *vp; 1165 struct tmpfs_dirent *de; 1166 struct tmpfs_mount *tmp; 1167 struct tmpfs_node *dnode; 1168 struct tmpfs_node *node; 1169 int error; 1170 1171 /* 1172 * We have to acquire the vp from v->a_nch because we will likely 1173 * unresolve the namecache entry, and a vrele/vput is needed to 1174 * trigger the tmpfs_inactive/tmpfs_reclaim sequence. 1175 * 1176 * We have to use vget to clear any inactive state on the vnode, 1177 * otherwise the vnode may remain inactive and thus tmpfs_inactive 1178 * will not get called when we release it. 1179 */ 1180 error = cache_vget(v->a_nch, v->a_cred, LK_SHARED, &vp); 1181 KKASSERT(error == 0); 1182 vn_unlock(vp); 1183 1184 /* 1185 * Prevalidate so we don't hit an assertion later 1186 */ 1187 if (vp->v_type != VDIR) { 1188 error = ENOTDIR; 1189 goto out; 1190 } 1191 1192 tmp = VFS_TO_TMPFS(dvp->v_mount); 1193 dnode = VP_TO_TMPFS_DIR(dvp); 1194 node = VP_TO_TMPFS_DIR(vp); 1195 1196 /* Directories with more than two entries ('.' and '..') cannot be 1197 * removed. */ 1198 if (node->tn_size > 0) { 1199 error = ENOTEMPTY; 1200 goto out; 1201 } 1202 1203 if ((dnode->tn_flags & APPEND) 1204 || (node->tn_flags & (NOUNLINK | IMMUTABLE | APPEND))) { 1205 error = EPERM; 1206 goto out; 1207 } 1208 1209 /* This invariant holds only if we are not trying to remove "..". 1210 * We checked for that above so this is safe now. */ 1211 KKASSERT(node->tn_dir.tn_parent == dnode); 1212 1213 /* Get the directory entry associated with node (vp). This was 1214 * filled by tmpfs_lookup while looking up the entry. */ 1215 de = tmpfs_dir_lookup(dnode, node, ncp); 1216 KKASSERT(TMPFS_DIRENT_MATCHES(de, 1217 ncp->nc_name, 1218 ncp->nc_nlen)); 1219 1220 /* Check flags to see if we are allowed to remove the directory. */ 1221 if ((dnode->tn_flags & APPEND) || 1222 node->tn_flags & (NOUNLINK | IMMUTABLE | APPEND)) { 1223 error = EPERM; 1224 goto out; 1225 } 1226 1227 1228 /* Detach the directory entry from the directory (dnode). */ 1229 tmpfs_dir_detach(dnode, de); 1230 1231 /* No vnode should be allocated for this entry from this point */ 1232 TMPFS_NODE_LOCK(node); 1233 TMPFS_ASSERT_ELOCKED(node); 1234 TMPFS_NODE_LOCK(dnode); 1235 TMPFS_ASSERT_ELOCKED(dnode); 1236 1237 #if 0 1238 /* handled by tmpfs_free_node */ 1239 KKASSERT(node->tn_links > 0); 1240 node->tn_links--; 1241 node->tn_dir.tn_parent = NULL; 1242 #endif 1243 node->tn_status |= TMPFS_NODE_ACCESSED | TMPFS_NODE_CHANGED | \ 1244 TMPFS_NODE_MODIFIED; 1245 1246 #if 0 1247 /* handled by tmpfs_free_node */ 1248 KKASSERT(dnode->tn_links > 0); 1249 dnode->tn_links--; 1250 #endif 1251 dnode->tn_status |= TMPFS_NODE_ACCESSED | \ 1252 TMPFS_NODE_CHANGED | TMPFS_NODE_MODIFIED; 1253 1254 TMPFS_NODE_UNLOCK(dnode); 1255 TMPFS_NODE_UNLOCK(node); 1256 1257 /* Free the directory entry we just deleted. Note that the node 1258 * referred by it will not be removed until the vnode is really 1259 * reclaimed. */ 1260 tmpfs_free_dirent(tmp, de); 1261 1262 /* Release the deleted vnode (will destroy the node, notify 1263 * interested parties and clean it from the cache). */ 1264 1265 TMPFS_NODE_LOCK(dnode); 1266 dnode->tn_status |= TMPFS_NODE_CHANGED; 1267 TMPFS_NODE_UNLOCK(dnode); 1268 tmpfs_update(dvp); 1269 1270 cache_setunresolved(v->a_nch); 1271 cache_setvp(v->a_nch, NULL); 1272 /*cache_inval_vp(vp, CINV_DESTROY);*/ 1273 tmpfs_knote(dvp, NOTE_WRITE | NOTE_LINK); 1274 error = 0; 1275 1276 out: 1277 vrele(vp); 1278 1279 return error; 1280 } 1281 1282 /* --------------------------------------------------------------------- */ 1283 1284 static int 1285 tmpfs_nsymlink(struct vop_nsymlink_args *v) 1286 { 1287 struct vnode *dvp = v->a_dvp; 1288 struct vnode **vpp = v->a_vpp; 1289 struct namecache *ncp = v->a_nch->ncp; 1290 struct vattr *vap = v->a_vap; 1291 struct ucred *cred = v->a_cred; 1292 char *target = v->a_target; 1293 int error; 1294 1295 vap->va_type = VLNK; 1296 error = tmpfs_alloc_file(dvp, vpp, vap, ncp, cred, target); 1297 if (error == 0) { 1298 tmpfs_knote(*vpp, NOTE_WRITE); 1299 cache_setunresolved(v->a_nch); 1300 cache_setvp(v->a_nch, *vpp); 1301 } 1302 1303 return error; 1304 } 1305 1306 /* --------------------------------------------------------------------- */ 1307 1308 static int 1309 tmpfs_readdir(struct vop_readdir_args *v) 1310 { 1311 struct vnode *vp = v->a_vp; 1312 struct uio *uio = v->a_uio; 1313 int *eofflag = v->a_eofflag; 1314 off_t **cookies = v->a_cookies; 1315 int *ncookies = v->a_ncookies; 1316 struct tmpfs_mount *tmp; 1317 int error; 1318 off_t startoff; 1319 off_t cnt = 0; 1320 struct tmpfs_node *node; 1321 1322 /* This operation only makes sense on directory nodes. */ 1323 if (vp->v_type != VDIR) 1324 return ENOTDIR; 1325 1326 tmp = VFS_TO_TMPFS(vp->v_mount); 1327 node = VP_TO_TMPFS_DIR(vp); 1328 startoff = uio->uio_offset; 1329 1330 if (uio->uio_offset == TMPFS_DIRCOOKIE_DOT) { 1331 error = tmpfs_dir_getdotdent(node, uio); 1332 if (error != 0) 1333 goto outok; 1334 cnt++; 1335 } 1336 1337 if (uio->uio_offset == TMPFS_DIRCOOKIE_DOTDOT) { 1338 error = tmpfs_dir_getdotdotdent(tmp, node, uio); 1339 if (error != 0) 1340 goto outok; 1341 cnt++; 1342 } 1343 1344 error = tmpfs_dir_getdents(node, uio, &cnt); 1345 1346 outok: 1347 KKASSERT(error >= -1); 1348 1349 if (error == -1) 1350 error = 0; 1351 1352 if (eofflag != NULL) 1353 *eofflag = 1354 (error == 0 && uio->uio_offset == TMPFS_DIRCOOKIE_EOF); 1355 1356 /* Update NFS-related variables. */ 1357 if (error == 0 && cookies != NULL && ncookies != NULL) { 1358 off_t i; 1359 off_t off = startoff; 1360 struct tmpfs_dirent *de = NULL; 1361 1362 *ncookies = cnt; 1363 *cookies = kmalloc(cnt * sizeof(off_t), M_TEMP, M_WAITOK); 1364 1365 for (i = 0; i < cnt; i++) { 1366 KKASSERT(off != TMPFS_DIRCOOKIE_EOF); 1367 if (off == TMPFS_DIRCOOKIE_DOT) { 1368 off = TMPFS_DIRCOOKIE_DOTDOT; 1369 } else { 1370 if (off == TMPFS_DIRCOOKIE_DOTDOT) { 1371 de = TAILQ_FIRST(&node->tn_dir.tn_dirhead); 1372 } else if (de != NULL) { 1373 de = TAILQ_NEXT(de, td_entries); 1374 } else { 1375 de = tmpfs_dir_lookupbycookie(node, 1376 off); 1377 KKASSERT(de != NULL); 1378 de = TAILQ_NEXT(de, td_entries); 1379 } 1380 if (de == NULL) 1381 off = TMPFS_DIRCOOKIE_EOF; 1382 else 1383 off = tmpfs_dircookie(de); 1384 } 1385 1386 (*cookies)[i] = off; 1387 } 1388 KKASSERT(uio->uio_offset == off); 1389 } 1390 1391 return error; 1392 } 1393 1394 /* --------------------------------------------------------------------- */ 1395 1396 static int 1397 tmpfs_readlink(struct vop_readlink_args *v) 1398 { 1399 struct vnode *vp = v->a_vp; 1400 struct uio *uio = v->a_uio; 1401 1402 int error; 1403 struct tmpfs_node *node; 1404 1405 KKASSERT(uio->uio_offset == 0); 1406 KKASSERT(vp->v_type == VLNK); 1407 1408 node = VP_TO_TMPFS_NODE(vp); 1409 1410 error = uiomove(node->tn_link, MIN(node->tn_size, uio->uio_resid), 1411 uio); 1412 TMPFS_NODE_LOCK(node); 1413 node->tn_status |= TMPFS_NODE_ACCESSED; 1414 TMPFS_NODE_UNLOCK(node); 1415 1416 return error; 1417 } 1418 1419 /* --------------------------------------------------------------------- */ 1420 1421 static int 1422 tmpfs_inactive(struct vop_inactive_args *v) 1423 { 1424 struct vnode *vp = v->a_vp; 1425 struct tmpfs_node *node; 1426 1427 node = VP_TO_TMPFS_NODE(vp); 1428 1429 /* 1430 * Degenerate case 1431 */ 1432 if (node == NULL) { 1433 vrecycle(vp); 1434 return(0); 1435 } 1436 1437 /* 1438 * Get rid of unreferenced deleted vnodes sooner rather than 1439 * later so the data memory can be recovered immediately. 1440 * 1441 * We must truncate the vnode to prevent the normal reclamation 1442 * path from flushing the data for the removed file to disk. 1443 */ 1444 TMPFS_NODE_LOCK(node); 1445 if ((node->tn_vpstate & TMPFS_VNODE_ALLOCATING) == 0 && 1446 (node->tn_links == 0 || 1447 (node->tn_links == 1 && node->tn_type == VDIR && 1448 node->tn_dir.tn_parent))) 1449 { 1450 node->tn_vpstate = TMPFS_VNODE_DOOMED; 1451 TMPFS_NODE_UNLOCK(node); 1452 if (node->tn_type == VREG) 1453 tmpfs_truncate(vp, 0); 1454 vrecycle(vp); 1455 } else { 1456 TMPFS_NODE_UNLOCK(node); 1457 } 1458 1459 return 0; 1460 } 1461 1462 /* --------------------------------------------------------------------- */ 1463 1464 int 1465 tmpfs_reclaim(struct vop_reclaim_args *v) 1466 { 1467 struct vnode *vp = v->a_vp; 1468 struct tmpfs_mount *tmp; 1469 struct tmpfs_node *node; 1470 1471 node = VP_TO_TMPFS_NODE(vp); 1472 tmp = VFS_TO_TMPFS(vp->v_mount); 1473 1474 tmpfs_free_vp(vp); 1475 1476 /* 1477 * If the node referenced by this vnode was deleted by the 1478 * user, we must free its associated data structures now that 1479 * the vnode is being reclaimed. 1480 * 1481 * Directories have an extra link ref. 1482 */ 1483 TMPFS_NODE_LOCK(node); 1484 if ((node->tn_vpstate & TMPFS_VNODE_ALLOCATING) == 0 && 1485 (node->tn_links == 0 || 1486 (node->tn_links == 1 && node->tn_type == VDIR && 1487 node->tn_dir.tn_parent))) 1488 { 1489 node->tn_vpstate = TMPFS_VNODE_DOOMED; 1490 tmpfs_free_node(tmp, node); 1491 /* eats the lock */ 1492 } else { 1493 TMPFS_NODE_UNLOCK(node); 1494 } 1495 1496 KKASSERT(vp->v_data == NULL); 1497 return 0; 1498 } 1499 1500 /* --------------------------------------------------------------------- */ 1501 1502 static int 1503 tmpfs_mountctl(struct vop_mountctl_args *ap) 1504 { 1505 struct tmpfs_mount *tmp; 1506 struct mount *mp; 1507 int rc; 1508 1509 switch (ap->a_op) { 1510 case (MOUNTCTL_SET_EXPORT): 1511 mp = ap->a_head.a_ops->head.vv_mount; 1512 tmp = (struct tmpfs_mount *) mp->mnt_data; 1513 1514 if (ap->a_ctllen != sizeof(struct export_args)) 1515 rc = (EINVAL); 1516 else 1517 rc = vfs_export(mp, &tmp->tm_export, 1518 (const struct export_args *) ap->a_ctl); 1519 break; 1520 default: 1521 rc = vop_stdmountctl(ap); 1522 break; 1523 } 1524 return (rc); 1525 } 1526 1527 /* --------------------------------------------------------------------- */ 1528 1529 static int 1530 tmpfs_print(struct vop_print_args *v) 1531 { 1532 struct vnode *vp = v->a_vp; 1533 1534 struct tmpfs_node *node; 1535 1536 node = VP_TO_TMPFS_NODE(vp); 1537 1538 kprintf("tag VT_TMPFS, tmpfs_node %p, flags 0x%x, links %d\n", 1539 node, node->tn_flags, node->tn_links); 1540 kprintf("\tmode 0%o, owner %d, group %d, size %ju, status 0x%x\n", 1541 node->tn_mode, node->tn_uid, node->tn_gid, 1542 (uintmax_t)node->tn_size, node->tn_status); 1543 1544 if (vp->v_type == VFIFO) 1545 fifo_printinfo(vp); 1546 1547 kprintf("\n"); 1548 1549 return 0; 1550 } 1551 1552 /* --------------------------------------------------------------------- */ 1553 1554 static int 1555 tmpfs_pathconf(struct vop_pathconf_args *v) 1556 { 1557 int name = v->a_name; 1558 register_t *retval = v->a_retval; 1559 1560 int error; 1561 1562 error = 0; 1563 1564 switch (name) { 1565 case _PC_LINK_MAX: 1566 *retval = LINK_MAX; 1567 break; 1568 1569 case _PC_NAME_MAX: 1570 *retval = NAME_MAX; 1571 break; 1572 1573 case _PC_PATH_MAX: 1574 *retval = PATH_MAX; 1575 break; 1576 1577 case _PC_PIPE_BUF: 1578 *retval = PIPE_BUF; 1579 break; 1580 1581 case _PC_CHOWN_RESTRICTED: 1582 *retval = 1; 1583 break; 1584 1585 case _PC_NO_TRUNC: 1586 *retval = 1; 1587 break; 1588 1589 case _PC_SYNC_IO: 1590 *retval = 1; 1591 break; 1592 1593 case _PC_FILESIZEBITS: 1594 *retval = 0; /* XXX Don't know which value should I return. */ 1595 break; 1596 1597 default: 1598 error = EINVAL; 1599 } 1600 1601 return error; 1602 } 1603 1604 /************************************************************************ 1605 * KQFILTER OPS * 1606 ************************************************************************/ 1607 1608 static void filt_tmpfsdetach(struct knote *kn); 1609 static int filt_tmpfsread(struct knote *kn, long hint); 1610 static int filt_tmpfswrite(struct knote *kn, long hint); 1611 static int filt_tmpfsvnode(struct knote *kn, long hint); 1612 1613 static struct filterops tmpfsread_filtops = 1614 { FILTEROP_ISFD, NULL, filt_tmpfsdetach, filt_tmpfsread }; 1615 static struct filterops tmpfswrite_filtops = 1616 { FILTEROP_ISFD, NULL, filt_tmpfsdetach, filt_tmpfswrite }; 1617 static struct filterops tmpfsvnode_filtops = 1618 { FILTEROP_ISFD, NULL, filt_tmpfsdetach, filt_tmpfsvnode }; 1619 1620 static int 1621 tmpfs_kqfilter (struct vop_kqfilter_args *ap) 1622 { 1623 struct vnode *vp = ap->a_vp; 1624 struct knote *kn = ap->a_kn; 1625 1626 switch (kn->kn_filter) { 1627 case EVFILT_READ: 1628 kn->kn_fop = &tmpfsread_filtops; 1629 break; 1630 case EVFILT_WRITE: 1631 kn->kn_fop = &tmpfswrite_filtops; 1632 break; 1633 case EVFILT_VNODE: 1634 kn->kn_fop = &tmpfsvnode_filtops; 1635 break; 1636 default: 1637 return (EOPNOTSUPP); 1638 } 1639 1640 kn->kn_hook = (caddr_t)vp; 1641 1642 knote_insert(&vp->v_pollinfo.vpi_kqinfo.ki_note, kn); 1643 1644 return(0); 1645 } 1646 1647 static void 1648 filt_tmpfsdetach(struct knote *kn) 1649 { 1650 struct vnode *vp = (void *)kn->kn_hook; 1651 1652 knote_remove(&vp->v_pollinfo.vpi_kqinfo.ki_note, kn); 1653 } 1654 1655 static int 1656 filt_tmpfsread(struct knote *kn, long hint) 1657 { 1658 struct vnode *vp = (void *)kn->kn_hook; 1659 struct tmpfs_node *node = VP_TO_TMPFS_NODE(vp); 1660 off_t off; 1661 1662 if (hint == NOTE_REVOKE) { 1663 kn->kn_flags |= (EV_EOF | EV_NODATA | EV_ONESHOT); 1664 return(1); 1665 } 1666 1667 /* 1668 * Interlock against MP races when performing this function. 1669 */ 1670 lwkt_gettoken(&vp->v_mount->mnt_token); 1671 off = node->tn_size - kn->kn_fp->f_offset; 1672 kn->kn_data = (off < INTPTR_MAX) ? off : INTPTR_MAX; 1673 if (kn->kn_sfflags & NOTE_OLDAPI) { 1674 lwkt_reltoken(&vp->v_mount->mnt_token); 1675 return(1); 1676 } 1677 1678 if (kn->kn_data == 0) { 1679 kn->kn_data = (off < INTPTR_MAX) ? off : INTPTR_MAX; 1680 } 1681 lwkt_reltoken(&vp->v_mount->mnt_token); 1682 return (kn->kn_data != 0); 1683 } 1684 1685 static int 1686 filt_tmpfswrite(struct knote *kn, long hint) 1687 { 1688 if (hint == NOTE_REVOKE) 1689 kn->kn_flags |= (EV_EOF | EV_NODATA | EV_ONESHOT); 1690 kn->kn_data = 0; 1691 return (1); 1692 } 1693 1694 static int 1695 filt_tmpfsvnode(struct knote *kn, long hint) 1696 { 1697 if (kn->kn_sfflags & hint) 1698 kn->kn_fflags |= hint; 1699 if (hint == NOTE_REVOKE) { 1700 kn->kn_flags |= (EV_EOF | EV_NODATA); 1701 return (1); 1702 } 1703 return (kn->kn_fflags != 0); 1704 } 1705 1706 1707 /* --------------------------------------------------------------------- */ 1708 1709 /* 1710 * vnode operations vector used for files stored in a tmpfs file system. 1711 */ 1712 struct vop_ops tmpfs_vnode_vops = { 1713 .vop_default = vop_defaultop, 1714 .vop_getpages = vop_stdgetpages, 1715 .vop_putpages = vop_stdputpages, 1716 .vop_ncreate = tmpfs_ncreate, 1717 .vop_nresolve = tmpfs_nresolve, 1718 .vop_nlookupdotdot = tmpfs_nlookupdotdot, 1719 .vop_nmknod = tmpfs_nmknod, 1720 .vop_open = tmpfs_open, 1721 .vop_close = tmpfs_close, 1722 .vop_access = tmpfs_access, 1723 .vop_getattr = tmpfs_getattr, 1724 .vop_setattr = tmpfs_setattr, 1725 .vop_read = tmpfs_read, 1726 .vop_write = tmpfs_write, 1727 .vop_fsync = tmpfs_fsync, 1728 .vop_mountctl = tmpfs_mountctl, 1729 .vop_nremove = tmpfs_nremove, 1730 .vop_nlink = tmpfs_nlink, 1731 .vop_nrename = tmpfs_nrename, 1732 .vop_nmkdir = tmpfs_nmkdir, 1733 .vop_nrmdir = tmpfs_nrmdir, 1734 .vop_nsymlink = tmpfs_nsymlink, 1735 .vop_readdir = tmpfs_readdir, 1736 .vop_readlink = tmpfs_readlink, 1737 .vop_inactive = tmpfs_inactive, 1738 .vop_reclaim = tmpfs_reclaim, 1739 .vop_print = tmpfs_print, 1740 .vop_pathconf = tmpfs_pathconf, 1741 .vop_bmap = tmpfs_bmap, 1742 .vop_strategy = tmpfs_strategy, 1743 .vop_advlock = tmpfs_advlock, 1744 .vop_kqfilter = tmpfs_kqfilter 1745 }; 1746