bsd/sys/mount_internal.h

   1 /*
   2  * Copyright (c) 2000-2007 Apple Inc. All rights reserved.
   3  *
   4  * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
   5  *
   6  * This file contains Original Code and/or Modifications of Original Code
   7  * as defined in and that are subject to the Apple Public Source License
   8  * Version 2.0 (the 'License'). You may not use this file except in
   9  * compliance with the License. The rights granted to you under the License
  10  * may not be used to create, or enable the creation or redistribution of,
  11  * unlawful or unlicensed copies of an Apple operating system, or to
  12  * circumvent, violate, or enable the circumvention or violation of, any
  13  * terms of an Apple operating system software license agreement.
  14  *
  15  * Please obtain a copy of the License at
  16  * http://www.opensource.apple.com/apsl/ and read it before using this file.
  17  *
  18  * The Original Code and all software distributed under the License are
  19  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  20  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  21  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  22  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  23  * Please see the License for the specific language governing rights and
  24  * limitations under the License.
  25  *
  26  * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
  27  */
  28 /* Copyright (c) 1995 NeXT Computer, Inc. All Rights Reserved */
  29 /*
  30  * Copyright (c) 1989, 1991, 1993
  31  *      The Regents of the University of California.  All rights reserved.
  32  *
  33  * Redistribution and use in source and binary forms, with or without
  34  * modification, are permitted provided that the following conditions
  35  * are met:
  36  * 1. Redistributions of source code must retain the above copyright
  37  *    notice, this list of conditions and the following disclaimer.
  38  * 2. Redistributions in binary form must reproduce the above copyright
  39  *    notice, this list of conditions and the following disclaimer in the
  40  *    documentation and/or other materials provided with the distribution.
  41  * 3. All advertising materials mentioning features or use of this software
  42  *    must display the following acknowledgement:
  43  *      This product includes software developed by the University of
  44  *      California, Berkeley and its contributors.
  45  * 4. Neither the name of the University nor the names of its contributors
  46  *    may be used to endorse or promote products derived from this software
  47  *    without specific prior written permission.
  48  *
  49  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  50  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  51  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  52  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  53  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  54  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  55  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  56  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  57  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  58  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  59  * SUCH DAMAGE.
  60  *
  61  *      @(#)mount.h     8.21 (Berkeley) 5/20/95
  62  */
  63 /*
  64  * NOTICE: This file was modified by McAfee Research in 2004 to introduce
  65  * support for mandatory and extensible security protections.  This notice
  66  * is included in support of clause 2.2 (b) of the Apple Public License,
  67  * Version 2.0.
  68  */
  69
  70 #ifndef _SYS_MOUNT_INTERNAL_H_
  71 #define _SYS_MOUNT_INTERNAL_H_
  72
  73 #include <sys/appleapiopts.h>
  74 #ifndef KERNEL
  75 #include <sys/ucred.h>
  76 #else
  77 #include <sys/kernel_types.h>
  78 #include <sys/namei.h>
  79 #endif
  80 #include <sys/queue.h>
  81 #include <sys/lock.h>
  82 #include <net/radix.h>
  83 #include <sys/socket.h>         /* XXX for AF_MAX */
  84 #include <sys/vfs_context.h>            /* XXX for AF_MAX */
  85 #include <sys/mount.h>
  86 #include <sys/cdefs.h>
  87
  88 struct label;
  89
  90 #if defined(__i386__) || defined(__x86_64__)
  91 typedef uint64_t  pending_io_t;
  92 #define INCR_PENDING_IO(a, b) OSAddAtomic64((int64_t)(a), (int64_t *)&(b));
  93 #else
  94 typedef uint32_t  pending_io_t;
  95 #define INCR_PENDING_IO(a, b) OSAddAtomic((int32_t)(a), (int32_t *)&(b));
  96 #endif
  97
  98
  99 /*
 100  * Structure per mounted file system.  Each mounted file system has an
 101  * array of operations and an instance record.  The file systems are
 102  * put on a doubly linked list.
 103  */
 104 TAILQ_HEAD(vnodelst, vnode);
 105
 106 struct mount {
 107         TAILQ_ENTRY(mount) mnt_list;            /* mount list */
 108         int32_t         mnt_count;              /* reference on the mount */
 109         lck_mtx_t       mnt_mlock;              /* mutex that protects mount point */
 110         struct vfsops   *mnt_op;                /* operations on fs */
 111         struct vfstable *mnt_vtable;            /* configuration info */
 112         struct vnode    *mnt_vnodecovered;      /* vnode we mounted on */
 113         struct vnodelst mnt_vnodelist;          /* list of vnodes this mount */
 114         struct vnodelst mnt_workerqueue;                /* list of vnodes this mount */
 115         struct vnodelst mnt_newvnodes;          /* list of vnodes this mount */
 116         uint32_t                mnt_flag;               /* flags */
 117         uint32_t                mnt_kern_flag;          /* kernel only flags */
 118         uint32_t                mnt_lflag;                      /* mount life cycle flags */
 119         uint32_t                mnt_maxsymlinklen;      /* max size of short symlink */
 120         struct vfsstatfs        mnt_vfsstat;            /* cache of filesystem stats */
 121         qaddr_t         mnt_data;               /* private data */
 122         /* Cached values of the IO constraints for the device */
 123         uint32_t        mnt_maxreadcnt;         /* Max. byte count for read */
 124         uint32_t        mnt_maxwritecnt;        /* Max. byte count for write */
 125         uint32_t        mnt_segreadcnt;         /* Max. segment count for read */
 126         uint32_t        mnt_segwritecnt;        /* Max. segment count for write */
 127         uint32_t        mnt_maxsegreadsize;     /* Max. segment read size  */
 128         uint32_t        mnt_maxsegwritesize;    /* Max. segment write size */
 129         uint32_t        mnt_alignmentmask;      /* Mask of bits that aren't addressable via DMA */
 130         uint32_t        mnt_devblocksize;       /* the underlying device block size */
 131         uint32_t        mnt_ioqueue_depth;      /* the maxiumum number of commands a device can accept */
 132         uint32_t        mnt_ioscale;            /* scale the various throttles/limits imposed on the amount of I/O in flight */
 133         uint32_t        mnt_ioflags;            /* flags for  underlying device */
 134         pending_io_t    mnt_pending_write_size; /* byte count of pending writes */
 135         pending_io_t    mnt_pending_read_size;  /* byte count of pending reads */
 136
 137         lck_rw_t        mnt_rwlock;             /* mutex readwrite lock */
 138         lck_mtx_t       mnt_renamelock;         /* mutex that serializes renames that change shape of tree */
 139         vnode_t         mnt_devvp;              /* the device mounted on for local file systems */
 140         uint32_t        mnt_devbsdunit;         /* the BSD unit number of the device */
 141         void            *mnt_throttle_info;     /* used by the throttle code */
 142         int32_t         mnt_crossref;           /* refernces to cover lookups  crossing into mp */
 143         int32_t         mnt_iterref;            /* refernces to cover iterations; drained makes it -ve  */
 144
 145         /* XXX 3762912 hack to support HFS filesystem 'owner' */
 146         uid_t           mnt_fsowner;
 147         gid_t           mnt_fsgroup;
 148
 149         struct label    *mnt_mntlabel;          /* MAC mount label */
 150         struct label    *mnt_fslabel;           /* MAC default fs label */
 151
 152         /*
 153          * cache the rootvp of the last mount point
 154          * in the chain in the mount struct pointed
 155          * to by the vnode sitting in '/'
 156          * this cache is used to shortcircuit the
 157          * mount chain traversal and allows us
 158          * to traverse to the true underlying rootvp
 159          * in 1 easy step inside of 'cache_lookup_path'
 160          *
 161          * make sure to validate against the cached vid
 162          * in case the rootvp gets stolen away since
 163          * we don't take an explicit long term reference
 164          * on it when we mount it
 165          */
 166         vnode_t         mnt_realrootvp;
 167         uint32_t        mnt_realrootvp_vid;
 168         /*
 169          * bumped each time a mount or unmount
 170          * occurs... its used to invalidate
 171          * 'mnt_realrootvp' from the cache
 172          */
 173         uint32_t             mnt_generation;
 174         /*
 175          * if 'MNTK_AUTH_CACHE_TIMEOUT' is
 176          * set, then 'mnt_authcache_ttl' is
 177          * the time-to-live for the per-vnode authentication cache
 178          * on this mount... if zero, no cache is maintained...
 179          * if 'MNTK_AUTH_CACHE_TIMEOUT' isn't set, its the
 180          * time-to-live for the cached lookup right for
 181          * volumes marked 'MNTK_AUTH_OPAQUE'.
 182          */
 183         int             mnt_authcache_ttl;
 184         /*
 185          * The proc structure pointer and process ID form a
 186          * sufficiently unique duple identifying the process
 187          * hosting this mount point. Set by vfs_markdependency()
 188          * and utilized in new_vnode() to avoid reclaiming vnodes
 189          * with this dependency (radar 5192010).
 190          */
 191         pid_t           mnt_dependent_pid;
 192         void            *mnt_dependent_process;
 193 };
 194
 195 /*
 196  * default number of seconds to keep cached lookup
 197  * rights valid on mounts marked MNTK_AUTH_OPAQUE
 198  */
 199 #define CACHED_LOOKUP_RIGHT_TTL         2
 200
 201 /*
 202  * ioflags
 203  */
 204 #define MNT_IOFLAGS_FUA_SUPPORTED       0x00000001
 205
 206 /*
 207  * ioqueue depth for devices that don't report one
 208  */
 209 #define MNT_DEFAULT_IOQUEUE_DEPTH       32
 210
 211
 212 /* XXX 3762912 hack to support HFS filesystem 'owner' */
 213 #define vfs_setowner(_mp, _uid, _gid)   do {(_mp)->mnt_fsowner = (_uid); (_mp)->mnt_fsgroup = (_gid); } while (0)
 214
 215
 216 /* mount point to which dead vps point to */
 217 extern struct mount * dead_mountp;
 218
 219 /*
 220  * Internal filesystem control flags stored in mnt_kern_flag.
 221  *
 222  * MNTK_UNMOUNT locks the mount entry so that name lookup cannot proceed
 223  * past the mount point.  This keeps the subtree stable during mounts
 224  * and unmounts.
 225  *
 226  * Note:        We are counting down on new bit assignments.  This is
 227  *              because the bits here were broken out from the high bits
 228  *              of the mount flags.
 229  */
 230 #ifdef CONFIG_IMGSRC_ACCESS
 231 #define MNTK_HAS_MOVED          0x00002000
 232 #define MNTK_BACKS_ROOT         0x00004000
 233 #endif /* CONFIG_IMGSRC_ACCESS */
 234 #define MNTK_AUTH_CACHE_TTL     0x00008000      /* rights cache has TTL - TTL of 0 disables cache */
 235 #define MNTK_PATH_FROM_ID       0x00010000      /* mounted file system supports id-to-path lookups */
 236 #define MNTK_UNMOUNT_PREFLIGHT  0x00020000      /* mounted file system wants preflight check during unmount */
 237 #define MNTK_NAMED_STREAMS      0x00040000      /* mounted file system supports Named Streams VNOPs */
 238 #define MNTK_EXTENDED_ATTRS     0x00080000      /* mounted file system supports Extended Attributes VNOPs */
 239 #define MNTK_LOCK_LOCAL         0x00100000      /* advisory locking is done above the VFS itself */
 240 #define MNTK_VIRTUALDEV         0x00200000      /* mounted on a virtual device i.e. a disk image */
 241 #define MNTK_ROOTDEV            0x00400000      /* this filesystem resides on the same device as the root */
 242 #define MNTK_SSD                0x00800000      /* underlying device is of the solid state variety */
 243 #define MNTK_UNMOUNT            0x01000000      /* unmount in progress */
 244 #define MNTK_MWAIT              0x02000000      /* waiting for unmount to finish */
 245 #define MNTK_WANTRDWR           0x04000000      /* upgrade to read/write requested */
 246 #if REV_ENDIAN_FS
 247 #define MNT_REVEND              0x08000000      /* Reverse endian FS */
 248 #endif /* REV_ENDIAN_FS */
 249 #define MNTK_FRCUNMOUNT         0x10000000      /* Forced unmount wanted. */
 250 #define MNTK_AUTH_OPAQUE        0x20000000      /* authorisation decisions are not made locally */
 251 #define MNTK_AUTH_OPAQUE_ACCESS 0x40000000      /* VNOP_ACCESS is reliable for remote auth */
 252 #define MNTK_EXTENDED_SECURITY  0x80000000      /* extended security supported */
 253
 254 #define MNT_LBUSY               0x00000001      /* mount is busy */
 255 #define MNT_LUNMOUNT            0x00000002      /* mount in unmount */
 256 #define MNT_LFORCE              0x00000004      /* mount in forced unmount */
 257 #define MNT_LDRAIN              0x00000008      /* mount in drain */
 258 #define MNT_LITER               0x00000010      /* mount in iteration */
 259 #define MNT_LNEWVN              0x00000020      /* mount has new vnodes created */
 260 #define MNT_LWAIT               0x00000040      /* wait for unmount op */
 261 #define MNT_LITERWAIT           0x00000080      /* mount in iteration */
 262 #define MNT_LDEAD               0x00000100      /* mount already unmounted*/
 263
 264
 265 /*
 266  * Generic file handle
 267  */
 268 #define NFS_MAX_FH_SIZE         NFSV4_MAX_FH_SIZE
 269 #define NFSV4_MAX_FH_SIZE       128
 270 #define NFSV3_MAX_FH_SIZE       64
 271 #define NFSV2_MAX_FH_SIZE       32
 272 struct fhandle {
 273         int             fh_len;                         /* length of file handle */
 274         unsigned char   fh_data[NFS_MAX_FH_SIZE];       /* file handle value */
 275 };
 276 typedef struct fhandle  fhandle_t;
 277
 278
 279
 280 /*
 281  * Filesystem configuration information. One of these exists for each
 282  * type of filesystem supported by the kernel. These are searched at
 283  * mount time to identify the requested filesystem.
 284  */
 285 struct vfstable {
 286         struct  vfsops *vfc_vfsops;     /* filesystem operations vector */
 287         char    vfc_name[MFSNAMELEN];   /* filesystem type name */
 288         int     vfc_typenum;            /* historic filesystem type number */
 289         int     vfc_refcount;           /* number mounted of this type */
 290         int     vfc_flags;              /* permanent flags */
 291         int     (*vfc_mountroot)(mount_t, vnode_t, vfs_context_t);      /* if != NULL, routine to mount root */
 292         struct  vfstable *vfc_next;     /* next in list */
 293         int32_t vfc_reserved1;
 294         int32_t vfc_reserved2;
 295         int             vfc_vfsflags;   /* for optional types */
 296         void *          vfc_descptr;    /* desc table allocated address */
 297         int                     vfc_descsize;   /* size allocated for desc table */
 298 };
 299
 300 /* vfc_vfsflags: */
 301 #define VFC_VFSLOCALARGS        0x002
 302 #define VFC_VFSGENERICARGS      0x004
 303 #define VFC_VFSNATIVEXATTR      0x010
 304 #define VFC_VFSDIRLINKS         0x020
 305 #define VFC_VFSPREFLIGHT        0x040
 306 #define VFC_VFSREADDIR_EXTENDED 0x080
 307 #define VFC_VFS64BITREADY       0x100
 308 #ifndef __LP64__
 309 #define VFC_VFSTHREADSAFE       0x200
 310 #endif /* __LP64__ */
 311 #define VFC_VFSNOMACLABEL       0x1000
 312 #define VFC_VFSVNOP_PAGEINV2    0x2000
 313 #define VFC_VFSVNOP_PAGEOUTV2   0x4000
 314
 315
 316 extern int maxvfsconf;          /* highest defined filesystem type */
 317 extern struct vfstable  *vfsconf;       /* head of list of filesystem types */
 318 extern int maxvfsslots;         /* Maximum slots available to be used */
 319 extern int numused_vfsslots;    /* number of slots already used */
 320
 321 /* the following two are xnu private */
 322 struct vfstable *       vfstable_add(struct     vfstable *);
 323 int     vfstable_del(struct vfstable *);
 324
 325
 326 struct vfsmount_args {
 327         union {
 328                 struct {
 329                         char * mnt_fspec;
 330                         void * mnt_fsdata;
 331                 } mnt_localfs_args;
 332                 struct {
 333                         void *  mnt_fsdata;             /* FS specific */
 334                 } mnt_remotefs_args;
 335         } mountfs_args;
 336 };
 337
 338
 339 /*
 340  * LP64 *user* version of statfs structure.
 341  * NOTE - must be kept in sync with struct statfs in mount.h
 342  */
 343 struct user64_statfs {
 344         short           f_otype;                /* TEMPORARY SHADOW COPY OF f_type */
 345         short           f_oflags;               /* TEMPORARY SHADOW COPY OF f_flags */
 346         user64_long_t   f_bsize;                /* fundamental file system block size */
 347         user64_long_t   f_iosize;               /* optimal transfer block size */
 348         user64_long_t   f_blocks;               /* total data blocks in file system */
 349         user64_long_t   f_bfree;                /* free blocks in fs */
 350         user64_long_t   f_bavail;               /* free blocks avail to non-superuser */
 351         user64_long_t   f_files;                /* total file nodes in file system */
 352         user64_long_t   f_ffree;                /* free file nodes in fs */
 353         fsid_t          f_fsid;                 /* file system id */
 354         uid_t           f_owner;                /* user that mounted the filesystem */
 355         short           f_reserved1;    /* spare for later */
 356         short           f_type;                 /* type of filesystem */
 357     user64_long_t       f_flags;                /* copy of mount exported flags */
 358         user64_long_t f_reserved2[2];   /* reserved for future use */
 359         char            f_fstypename[MFSNAMELEN]; /* fs type name */
 360         char            f_mntonname[MNAMELEN];  /* directory on which mounted */
 361         char            f_mntfromname[MNAMELEN];/* mounted filesystem */
 362         char            f_reserved3;    /* For alignment */
 363         user64_long_t   f_reserved4[4]; /* For future use */
 364 };
 365
 366 /*
 367  * ILP32 *user* version of statfs structure.
 368  * NOTE - must be kept in sync with struct statfs in mount.h
 369  */
 370 struct user32_statfs {
 371         short           f_otype;                /* TEMPORARY SHADOW COPY OF f_type */
 372         short           f_oflags;               /* TEMPORARY SHADOW COPY OF f_flags */
 373         user32_long_t   f_bsize;                /* fundamental file system block size */
 374         user32_long_t   f_iosize;               /* optimal transfer block size */
 375         user32_long_t   f_blocks;               /* total data blocks in file system */
 376         user32_long_t   f_bfree;                /* free blocks in fs */
 377         user32_long_t   f_bavail;               /* free blocks avail to non-superuser */
 378         user32_long_t   f_files;                /* total file nodes in file system */
 379         user32_long_t   f_ffree;                /* free file nodes in fs */
 380         fsid_t          f_fsid;                 /* file system id */
 381         uid_t           f_owner;                /* user that mounted the filesystem */
 382         short           f_reserved1;    /* spare for later */
 383         short           f_type;                 /* type of filesystem */
 384     user32_long_t       f_flags;                /* copy of mount exported flags */
 385         user32_long_t f_reserved2[2];   /* reserved for future use */
 386         char            f_fstypename[MFSNAMELEN]; /* fs type name */
 387         char            f_mntonname[MNAMELEN];  /* directory on which mounted */
 388         char            f_mntfromname[MNAMELEN];/* mounted filesystem */
 389         char            f_reserved3;    /* For alignment */
 390         user32_long_t   f_reserved4[4]; /* For future use */
 391 };
 392
 393 /*
 394  * throttle I/Os are affected only by normal I/Os happening on the same bsd device node.  For example, disk1s3 and
 395  * disk1s5 are the same device node, while disk1s3 and disk2 are not (although disk2 might be a mounted disk image file
 396  * and the disk image file resides on a partition in disk1).  The following constant defines the maximum number of
 397  * different bsd device nodes the algorithm can consider, and larger numbers are rounded by this maximum.  Since
 398  * throttled I/O is usually useful in non-server environment only, a small number 16 is enough in most cases
 399  */
 400 #define LOWPRI_MAX_NUM_DEV 16
 401
 402 __BEGIN_DECLS
 403
 404 extern uint32_t mount_generation;
 405 extern TAILQ_HEAD(mntlist, mount) mountlist;
 406 void mount_list_lock(void);
 407 void mount_list_unlock(void);
 408 void mount_lock_init(mount_t);
 409 void mount_lock_destroy(mount_t);
 410 void mount_lock(mount_t);
 411 void mount_lock_spin(mount_t);
 412 void mount_unlock(mount_t);
 413 void mount_lock_renames(mount_t);
 414 void mount_unlock_renames(mount_t);
 415 void mount_ref(mount_t, int);
 416 void mount_drop(mount_t, int);
 417 int  mount_refdrain(mount_t);
 418
 419 /* vfs_rootmountalloc should be kept as a private api */
 420 errno_t vfs_rootmountalloc(const char *, const char *, mount_t *mpp);
 421
 422 int     vfs_mountroot(void);
 423 void    vfs_unmountall(void);
 424 int     safedounmount(struct mount *, int, vfs_context_t);
 425 int     dounmount(struct mount *, int, int, vfs_context_t);
 426
 427 /* xnuy internal api */
 428 void  mount_dropcrossref(mount_t, vnode_t, int);
 429 mount_t mount_lookupby_volfsid(int, int);
 430 mount_t mount_list_lookupby_fsid(fsid_t *, int, int);
 431 int  mount_list_add(mount_t);
 432 void mount_list_remove(mount_t);
 433 int  mount_iterref(mount_t, int);
 434 int  mount_isdrained(mount_t, int);
 435 void mount_iterdrop(mount_t);
 436 void mount_iterdrain(mount_t);
 437 void mount_iterreset(mount_t);
 438
 439 /* throttled I/O api */
 440 int throttle_get_io_policy(struct uthread **ut);
 441 extern void throttle_lowpri_io(boolean_t ok_to_sleep);
 442 int throttle_io_will_be_throttled(int lowpri_window_msecs, mount_t mp);
 443
 444 __END_DECLS
 445
 446 #endif /* !_SYS_MOUNT_INTERNAL_H_ */