lfs.h revision 1.37
1 /* $NetBSD: lfs.h,v 1.37 2001/07/13 20:30:22 perseant Exp $ */ 2 3 /*- 4 * Copyright (c) 1999, 2000 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Konrad E. Schroder <perseant (at) hhhh.org>. 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. All advertising materials mentioning features or use of this software 19 * must display the following acknowledgement: 20 * This product includes software developed by the NetBSD 21 * Foundation, Inc. and its contributors. 22 * 4. Neither the name of The NetBSD Foundation nor the names of its 23 * contributors may be used to endorse or promote products derived 24 * from this software without specific prior written permission. 25 * 26 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 27 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 28 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 29 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 30 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 31 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 32 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 33 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 34 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 35 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 36 * POSSIBILITY OF SUCH DAMAGE. 37 */ 38 /*- 39 * Copyright (c) 1991, 1993 40 * The Regents of the University of California. All rights reserved. 41 * 42 * Redistribution and use in source and binary forms, with or without 43 * modification, are permitted provided that the following conditions 44 * are met: 45 * 1. Redistributions of source code must retain the above copyright 46 * notice, this list of conditions and the following disclaimer. 47 * 2. Redistributions in binary form must reproduce the above copyright 48 * notice, this list of conditions and the following disclaimer in the 49 * documentation and/or other materials provided with the distribution. 50 * 3. All advertising materials mentioning features or use of this software 51 * must display the following acknowledgement: 52 * This product includes software developed by the University of 53 * California, Berkeley and its contributors. 54 * 4. Neither the name of the University nor the names of its contributors 55 * may be used to endorse or promote products derived from this software 56 * without specific prior written permission. 57 * 58 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 59 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 60 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 61 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 62 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 63 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 64 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 65 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 66 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 67 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 68 * SUCH DAMAGE. 69 * 70 * @(#)lfs.h 8.9 (Berkeley) 5/8/95 71 */ 72 73 /* 74 * Compile-time options for LFS. 75 */ 76 #define LFS_EAGAIN_FAIL /* markv fail with EAGAIN if ino is locked */ 77 #define LFS_TRACK_IOS /* attempt to avoid cleaning segments not yet fully written to disk */ 78 #define LFS_DEBUG_RFW /* print roll-forward debugging info */ 79 80 /* #define DEBUG_LFS */ /* Intensive debugging of LFS subsystem */ 81 82 /* 83 * Parameters and generic definitions 84 */ 85 #define BW_CLEAN 1 86 #define MIN_FREE_SEGS 2 87 #define LFS_MAX_ACTIVE 10 88 #define LFS_MAXDIROP (desiredvnodes>>2) 89 #ifndef LFS_ATIME_IFILE 90 # define LFS_ATIME_IFILE 0 91 #endif 92 93 /* 94 * #define WRITE_THRESHHOLD ((nbuf >> 1) - 10) 95 * #define WAIT_THRESHHOLD (nbuf - (nbuf >> 2) - 10) 96 */ 97 #define LFS_MAX_BUFS ((nbuf >> 2) - 10) 98 #define LFS_WAIT_BUFS ((nbuf >> 1) - (nbuf >> 3) - 10) 99 /* These are new ... is LFS taking up too much memory in its buffers? */ 100 #define LFS_MAX_BYTES (((bufpages >> 2) - 10) * NBPG) 101 #define LFS_WAIT_BYTES (((bufpages >> 1) - (bufpages >> 3) - 10) * NBPG) 102 #define LFS_BUFWAIT 2 103 104 #define LFS_LOCK_BUF(bp) do { \ 105 if (((bp)->b_flags & (B_LOCKED | B_CALL)) == 0) { \ 106 ++locked_queue_count; \ 107 locked_queue_bytes += bp->b_bufsize; \ 108 } \ 109 (bp)->b_flags |= B_LOCKED; \ 110 } while(0) 111 112 #define LFS_UNLOCK_BUF(bp) do { \ 113 if (((bp)->b_flags & (B_LOCKED | B_CALL)) == B_LOCKED) { \ 114 --locked_queue_count; \ 115 locked_queue_bytes -= bp->b_bufsize; \ 116 if (locked_queue_count < LFS_WAIT_BUFS && \ 117 locked_queue_bytes < LFS_WAIT_BYTES) \ 118 wakeup(&locked_queue_count); \ 119 } \ 120 (bp)->b_flags &= ~B_LOCKED; \ 121 } while(0) 122 123 /* For convenience */ 124 #define IN_ALLMOD (IN_MODIFIED|IN_ACCESS|IN_CHANGE|IN_UPDATE|IN_ACCESSED|IN_CLEANING) 125 126 #define LFS_SET_UINO(ip, flags) do { \ 127 if (((flags) & IN_ACCESSED) && !((ip)->i_flag & IN_ACCESSED)) \ 128 ++(ip)->i_lfs->lfs_uinodes; \ 129 if (((flags) & IN_CLEANING) && !((ip)->i_flag & IN_CLEANING)) \ 130 ++(ip)->i_lfs->lfs_uinodes; \ 131 if (((flags) & IN_MODIFIED) && !((ip)->i_flag & IN_MODIFIED)) \ 132 ++(ip)->i_lfs->lfs_uinodes; \ 133 (ip)->i_flag |= (flags); \ 134 } while(0) 135 136 #define LFS_CLR_UINO(ip, flags) do { \ 137 if (((flags) & IN_ACCESSED) && ((ip)->i_flag & IN_ACCESSED)) \ 138 --(ip)->i_lfs->lfs_uinodes; \ 139 if (((flags) & IN_CLEANING) && ((ip)->i_flag & IN_CLEANING)) \ 140 --(ip)->i_lfs->lfs_uinodes; \ 141 if (((flags) & IN_MODIFIED) && ((ip)->i_flag & IN_MODIFIED)) \ 142 --(ip)->i_lfs->lfs_uinodes; \ 143 (ip)->i_flag &= ~(flags); \ 144 if ((ip)->i_lfs->lfs_uinodes < 0) { \ 145 panic("lfs_uinodes < 0"); \ 146 } \ 147 } while(0) 148 149 #define LFS_ITIMES(ip, acc, mod, cre) do { \ 150 if ((ip)->i_flag & IN_ACCESS) { \ 151 (ip)->i_ffs_atime = (acc)->tv_sec; \ 152 (ip)->i_ffs_atimensec = (acc)->tv_nsec; \ 153 if ((ip)->i_lfs->lfs_version > 1) { \ 154 struct buf *ibp; \ 155 IFILE *ifp; \ 156 \ 157 LFS_IENTRY(ifp, ip->i_lfs, ip->i_number, ibp); \ 158 ifp->if_atime_sec = (acc)->tv_sec; \ 159 ifp->if_atime_nsec = (acc)->tv_nsec; \ 160 VOP_BWRITE(ibp); \ 161 } else { \ 162 LFS_SET_UINO(ip, IN_ACCESSED); \ 163 } \ 164 } \ 165 if ((ip)->i_flag & (IN_CHANGE | IN_UPDATE)) { \ 166 if ((ip)->i_flag & IN_UPDATE) { \ 167 (ip)->i_ffs_mtime = (mod)->tv_sec; \ 168 (ip)->i_ffs_mtimensec = (mod)->tv_nsec; \ 169 (ip)->i_modrev++; \ 170 } \ 171 if ((ip)->i_flag & IN_CHANGE) { \ 172 (ip)->i_ffs_ctime = (cre)->tv_sec; \ 173 (ip)->i_ffs_ctimensec = (cre)->tv_nsec; \ 174 } \ 175 LFS_SET_UINO(ip, IN_MODIFIED); \ 176 } \ 177 (ip)->i_flag &= ~(IN_ACCESS | IN_CHANGE | IN_UPDATE); \ 178 } while(0) 179 180 #define WRITEINPROG(vp) (vp->v_dirtyblkhd.lh_first && !(VTOI(vp)->i_flag & \ 181 (IN_MODIFIED | IN_ACCESSED | IN_CLEANING))) 182 183 /* Here begins the berkeley code */ 184 185 #define LFS_LABELPAD 8192 /* LFS label size */ 186 #define LFS_SBPAD 8192 /* LFS superblock size */ 187 188 /* On-disk and in-memory checkpoint segment usage structure. */ 189 typedef struct segusage SEGUSE; 190 struct segusage { 191 u_int32_t su_nbytes; /* 0: number of live bytes */ 192 u_int32_t su_olastmod; /* 4: SEGUSE last modified timestamp */ 193 u_int16_t su_nsums; /* 8: number of summaries in segment */ 194 u_int16_t su_ninos; /* 10: number of inode blocks in seg */ 195 196 #define SEGUSE_ACTIVE 0x01 /* segment currently being written */ 197 #define SEGUSE_DIRTY 0x02 /* segment has data in it */ 198 #define SEGUSE_SUPERBLOCK 0x04 /* segment contains a superblock */ 199 #define SEGUSE_ERROR 0x08 /* cleaner: do not clean segment */ 200 u_int32_t su_flags; /* 12: segment flags */ 201 u_int64_t su_lastmod; /* 16: last modified timestamp */ 202 }; 203 204 typedef struct segusage_v1 SEGUSE_V1; 205 struct segusage_v1 { 206 u_int32_t su_nbytes; /* 0: number of live bytes */ 207 u_int32_t su_lastmod; /* 4: SEGUSE last modified timestamp */ 208 u_int16_t su_nsums; /* 8: number of summaries in segment */ 209 u_int16_t su_ninos; /* 10: number of inode blocks in seg */ 210 u_int32_t su_flags; /* 12: segment flags */ 211 }; 212 213 #define SEGUPB(fs) (fs->lfs_sepb) 214 #define SEGTABSIZE_SU(fs) \ 215 (((fs)->lfs_nseg + SEGUPB(fs) - 1) / (fs)->lfs_sepb) 216 217 /* On-disk file information. One per file with data blocks in the segment. */ 218 typedef struct finfo FINFO; 219 struct finfo { 220 u_int32_t fi_nblocks; /* number of blocks */ 221 u_int32_t fi_version; /* version number */ 222 u_int32_t fi_ino; /* inode number */ 223 u_int32_t fi_lastlength; /* length of last block in array */ 224 ufs_daddr_t fi_blocks[1]; /* array of logical block numbers */ 225 }; 226 227 228 /* On-disk super block. */ 229 struct dlfs { 230 #define LFS_MAGIC 0x070162 231 u_int32_t dlfs_magic; /* 0: magic number */ 232 #define LFS_VERSION 2 233 u_int32_t dlfs_version; /* 4: version number */ 234 235 u_int32_t dlfs_size; /* 8: number of blocks in fs (v1) */ 236 /* number of frags in fs (v2) */ 237 u_int32_t dlfs_ssize; /* 12: number of blocks per segment (v1) */ 238 /* number of bytes per segment (v2) */ 239 u_int32_t dlfs_dsize; /* 16: number of disk blocks in fs */ 240 u_int32_t dlfs_bsize; /* 20: file system block size */ 241 u_int32_t dlfs_fsize; /* 24: size of frag blocks in fs */ 242 u_int32_t dlfs_frag; /* 28: number of frags in a block in fs */ 243 244 /* Checkpoint region. */ 245 u_int32_t dlfs_free; /* 32: start of the free list */ 246 u_int32_t dlfs_bfree; /* 36: number of free disk blocks */ 247 u_int32_t dlfs_nfiles; /* 40: number of allocated inodes */ 248 int32_t dlfs_avail; /* 44: blocks available for writing */ 249 int32_t dlfs_uinodes; /* 48: inodes in cache not yet on disk */ 250 ufs_daddr_t dlfs_idaddr; /* 52: inode file disk address */ 251 u_int32_t dlfs_ifile; /* 56: inode file inode number */ 252 ufs_daddr_t dlfs_lastseg; /* 60: address of last segment written */ 253 ufs_daddr_t dlfs_nextseg; /* 64: address of next segment to write */ 254 ufs_daddr_t dlfs_curseg; /* 68: current segment being written */ 255 ufs_daddr_t dlfs_offset; /* 72: offset in curseg for next partial */ 256 ufs_daddr_t dlfs_lastpseg; /* 76: address of last partial written */ 257 u_int32_t dlfs_inopf; /* 80: v1: time stamp; v2: inodes per frag */ 258 #define dlfs_otstamp dlfs_inopf 259 260 /* These are configuration parameters. */ 261 u_int32_t dlfs_minfree; /* 84: minimum percentage of free blocks */ 262 263 /* These fields can be computed from the others. */ 264 u_int64_t dlfs_maxfilesize; /* 88: maximum representable file size */ 265 u_int32_t dlfs_fsbpseg; /* 96: fsb per segment */ 266 u_int32_t dlfs_inopb; /* 100: inodes per block */ 267 u_int32_t dlfs_ifpb; /* 104: IFILE entries per block */ 268 u_int32_t dlfs_sepb; /* 108: SEGUSE entries per block */ 269 u_int32_t dlfs_nindir; /* 112: indirect pointers per block */ 270 u_int32_t dlfs_nseg; /* 116: number of segments */ 271 u_int32_t dlfs_nspf; /* 120: number of sectors per fragment */ 272 u_int32_t dlfs_cleansz; /* 124: cleaner info size in blocks */ 273 u_int32_t dlfs_segtabsz; /* 128: segment table size in blocks */ 274 u_int32_t dlfs_segmask; /* 132: calculate offset within a segment */ 275 u_int32_t dlfs_segshift; /* 136: fast mult/div for segments */ 276 u_int32_t dlfs_bshift; /* 140: calc block number from file offset */ 277 u_int32_t dlfs_ffshift; /* 144: fast mult/div for frag from file */ 278 u_int32_t dlfs_fbshift; /* 148: fast mult/div for frag from block */ 279 u_int64_t dlfs_bmask; /* 152: calc block offset from file offset */ 280 u_int64_t dlfs_ffmask; /* 160: calc frag offset from file offset */ 281 u_int64_t dlfs_fbmask; /* 168: calc frag offset from block offset */ 282 u_int32_t dlfs_blktodb; /* 176: blktodb and dbtoblk shift constant */ 283 u_int32_t dlfs_sushift; /* 180: fast mult/div for segusage table */ 284 285 int32_t dlfs_maxsymlinklen; /* 184: max length of an internal symlink */ 286 #define LFS_MIN_SBINTERVAL 5 /* minimum superblock segment spacing */ 287 #define LFS_MAXNUMSB 10 /* 188: superblock disk offsets */ 288 ufs_daddr_t dlfs_sboffs[LFS_MAXNUMSB]; 289 290 u_int32_t dlfs_nclean; /* 228: Number of clean segments */ 291 u_char dlfs_fsmnt[MNAMELEN]; /* 232: name mounted on */ 292 #define LFS_PF_CLEAN 0x1 293 u_int16_t dlfs_pflags; /* 322: file system persistent flags */ 294 int32_t dlfs_dmeta; /* 324: total number of dirty summaries */ 295 u_int32_t dlfs_minfreeseg; /* 328: segs reserved for cleaner */ 296 u_int32_t dlfs_sumsize; /* 332: size of summary blocks */ 297 u_int64_t dlfs_serial; /* 336: serial number */ 298 u_int32_t dlfs_ibsize; /* 344: size of inode blocks */ 299 ufs_daddr_t dlfs_start; /* 348: start of segment 0 */ 300 u_int64_t dlfs_tstamp; /* 352: time stamp */ 301 #define LFS_44INODEFMT 0 302 #define LFS_MAXINODEFMT 0 303 u_int32_t dlfs_inodefmt; /* 360: inode format version */ 304 u_int32_t dlfs_interleave; /* 364: segment interleave */ 305 u_int32_t dlfs_ident; /* 368: per-fs identifier */ 306 u_int32_t dlfs_fsbtodb; /* 372: fsbtodb abd dbtodsb shift constant */ 307 int8_t dlfs_pad[132]; /* 376: round to 512 bytes */ 308 /* Checksum -- last valid disk field. */ 309 u_int32_t dlfs_cksum; /* 508: checksum for superblock checking */ 310 }; 311 312 /* Maximum number of io's we can have pending at once */ 313 #define LFS_THROTTLE 16 /* XXX should be better paramtrized - ? */ 314 315 /* In-memory super block. */ 316 struct lfs { 317 struct dlfs lfs_dlfs; /* on-disk parameters */ 318 #define lfs_magic lfs_dlfs.dlfs_magic 319 #define lfs_version lfs_dlfs.dlfs_version 320 #define lfs_size lfs_dlfs.dlfs_size 321 #define lfs_ssize lfs_dlfs.dlfs_ssize 322 #define lfs_dsize lfs_dlfs.dlfs_dsize 323 #define lfs_bsize lfs_dlfs.dlfs_bsize 324 #define lfs_fsize lfs_dlfs.dlfs_fsize 325 #define lfs_frag lfs_dlfs.dlfs_frag 326 #define lfs_free lfs_dlfs.dlfs_free 327 #define lfs_bfree lfs_dlfs.dlfs_bfree 328 #define lfs_nfiles lfs_dlfs.dlfs_nfiles 329 #define lfs_avail lfs_dlfs.dlfs_avail 330 #define lfs_uinodes lfs_dlfs.dlfs_uinodes 331 #define lfs_idaddr lfs_dlfs.dlfs_idaddr 332 #define lfs_ifile lfs_dlfs.dlfs_ifile 333 #define lfs_lastseg lfs_dlfs.dlfs_lastseg 334 #define lfs_nextseg lfs_dlfs.dlfs_nextseg 335 #define lfs_curseg lfs_dlfs.dlfs_curseg 336 #define lfs_offset lfs_dlfs.dlfs_offset 337 #define lfs_lastpseg lfs_dlfs.dlfs_lastpseg 338 #define lfs_otstamp lfs_dlfs.dlfs_inopf 339 #define lfs_inopf lfs_dlfs.dlfs_inopf 340 #define lfs_minfree lfs_dlfs.dlfs_minfree 341 #define lfs_maxfilesize lfs_dlfs.dlfs_maxfilesize 342 #define lfs_fsbpseg lfs_dlfs.dlfs_fsbpseg 343 #define lfs_inopb lfs_dlfs.dlfs_inopb 344 #define lfs_ifpb lfs_dlfs.dlfs_ifpb 345 #define lfs_sepb lfs_dlfs.dlfs_sepb 346 #define lfs_nindir lfs_dlfs.dlfs_nindir 347 #define lfs_nseg lfs_dlfs.dlfs_nseg 348 #define lfs_nspf lfs_dlfs.dlfs_nspf 349 #define lfs_cleansz lfs_dlfs.dlfs_cleansz 350 #define lfs_segtabsz lfs_dlfs.dlfs_segtabsz 351 #define lfs_segmask lfs_dlfs.dlfs_segmask 352 #define lfs_segshift lfs_dlfs.dlfs_segshift 353 #define lfs_bmask lfs_dlfs.dlfs_bmask 354 #define lfs_bshift lfs_dlfs.dlfs_bshift 355 #define lfs_ffmask lfs_dlfs.dlfs_ffmask 356 #define lfs_ffshift lfs_dlfs.dlfs_ffshift 357 #define lfs_fbmask lfs_dlfs.dlfs_fbmask 358 #define lfs_fbshift lfs_dlfs.dlfs_fbshift 359 #define lfs_blktodb lfs_dlfs.dlfs_blktodb 360 #define lfs_fsbtodb lfs_dlfs.dlfs_fsbtodb 361 #define lfs_sushift lfs_dlfs.dlfs_sushift 362 #define lfs_maxsymlinklen lfs_dlfs.dlfs_maxsymlinklen 363 #define lfs_sboffs lfs_dlfs.dlfs_sboffs 364 #define lfs_cksum lfs_dlfs.dlfs_cksum 365 #define lfs_pflags lfs_dlfs.dlfs_pflags 366 #define lfs_fsmnt lfs_dlfs.dlfs_fsmnt 367 #define lfs_nclean lfs_dlfs.dlfs_nclean 368 #define lfs_dmeta lfs_dlfs.dlfs_dmeta 369 #define lfs_minfreeseg lfs_dlfs.dlfs_minfreeseg 370 #define lfs_sumsize lfs_dlfs.dlfs_sumsize 371 #define lfs_serial lfs_dlfs.dlfs_serial 372 #define lfs_ibsize lfs_dlfs.dlfs_ibsize 373 #define lfs_start lfs_dlfs.dlfs_start 374 #define lfs_tstamp lfs_dlfs.dlfs_tstamp 375 #define lfs_inodefmt lfs_dlfs.dlfs_inodefmt 376 #define lfs_interleave lfs_dlfs.dlfs_interleave 377 #define lfs_ident lfs_dlfs.dlfs_ident 378 379 /* These fields are set at mount time and are meaningless on disk. */ 380 struct segment *lfs_sp; /* current segment being written */ 381 struct vnode *lfs_ivnode; /* vnode for the ifile */ 382 u_int32_t lfs_seglock; /* single-thread the segment writer */ 383 pid_t lfs_lockpid; /* pid of lock holder */ 384 u_int32_t lfs_iocount; /* number of ios pending */ 385 u_int32_t lfs_writer; /* don't allow any dirops to start */ 386 u_int32_t lfs_dirops; /* count of active directory ops */ 387 u_int32_t lfs_doifile; /* Write ifile blocks on next write */ 388 u_int32_t lfs_nactive; /* Number of segments since last ckp */ 389 int8_t lfs_fmod; /* super block modified flag */ 390 int8_t lfs_ronly; /* mounted read-only flag */ 391 #define LFS_NOTYET 0x01 392 int8_t lfs_flags; /* currently unused flag */ 393 u_int16_t lfs_activesb; /* toggle between superblocks */ 394 #ifdef LFS_TRACK_IOS 395 daddr_t lfs_pending[LFS_THROTTLE]; /* daddrs of pending writes */ 396 #endif /* LFS_TRACK_IOS */ 397 daddr_t lfs_sbactive; /* disk address of in-progress sb write */ 398 struct vnode *lfs_flushvp; /* vnode being flushed */ 399 struct vnode *lfs_unlockvp; /* being inactivated in lfs_segunlock */ 400 u_int32_t lfs_diropwait; /* # procs waiting on dirop flush */ 401 size_t lfs_devbsize; /* Device block size */ 402 size_t lfs_devbshift; /* Device block shift */ 403 struct lock lfs_freelock; 404 pid_t lfs_rfpid; /* Process ID of roll-forward agent */ 405 int lfs_nadirop; /* number of active dirop nodes */ 406 long lfs_ravail; /* blocks pre-reserved for writing */ 407 }; 408 409 /* 410 * Inode 0: out-of-band inode number 411 * Inode 1: IFILE inode number 412 * Inode 2: root inode 413 * Inode 3: lost+found inode number 414 */ 415 #define LFS_UNUSED_INUM 0 /* out of band inode number */ 416 #define LFS_IFILE_INUM 1 /* IFILE inode number */ 417 #define LOSTFOUNDINO 3 /* lost+found inode number */ 418 #define LFS_FIRST_INUM 4 /* first free inode number */ 419 420 /* Address calculations for metadata located in the inode */ 421 #define S_INDIR(fs) -NDADDR 422 #define D_INDIR(fs) (S_INDIR(fs) - NINDIR(fs) - 1) 423 #define T_INDIR(fs) (D_INDIR(fs) - NINDIR(fs) * NINDIR(fs) - 1) 424 425 /* Unassigned disk addresses. */ 426 #define UNASSIGNED -1 427 #define UNWRITTEN -2 428 429 /* Unused logical block number */ 430 #define LFS_UNUSED_LBN -1 431 432 typedef struct ifile IFILE; 433 struct ifile { 434 u_int32_t if_version; /* inode version number */ 435 #define LFS_UNUSED_DADDR 0 /* out-of-band daddr */ 436 ufs_daddr_t if_daddr; /* inode disk address */ 437 ino_t if_nextfree; /* next-unallocated inode */ 438 /* XXX - when inode format changes, this changes too */ 439 u_int32_t if_atime_sec; /* Last access time, seconds */ 440 u_int32_t if_atime_nsec; /* and nanoseconds */ 441 }; 442 443 typedef struct ifile_v1 IFILE_V1; 444 struct ifile_v1 { 445 u_int32_t if_version; /* inode version number */ 446 ufs_daddr_t if_daddr; /* inode disk address */ 447 ino_t if_nextfree; /* next-unallocated inode */ 448 #if LFS_ATIME_IFILE 449 struct timespec if_atime; /* Last access time */ 450 #endif 451 }; 452 453 /* 454 * Cleaner information structure. This resides in the ifile and is used 455 * to pass information between the cleaner and the kernel. 456 */ 457 typedef struct _cleanerinfo { 458 u_int32_t clean; /* number of clean segments */ 459 u_int32_t dirty; /* number of dirty segments */ 460 u_int32_t bfree; /* disk blocks free */ 461 int32_t avail; /* disk blocks available */ 462 u_int32_t free_head; /* head of the inode free list */ 463 u_int32_t free_tail; /* tail of the inode free list */ 464 } CLEANERINFO; 465 466 #define CLEANSIZE_SU(fs) \ 467 ((sizeof(CLEANERINFO) + (fs)->lfs_bsize - 1) >> (fs)->lfs_bshift) 468 469 /* 470 * All summary blocks are the same size, so we can always read a summary 471 * block easily from a segment. 472 */ 473 #define LFS_V1_SUMMARY_SIZE 512 474 #define LFS_DFL_SUMMARY_SIZE 512 475 476 /* On-disk segment summary information */ 477 typedef struct segsum_v1 SEGSUM_V1; 478 struct segsum_v1 { 479 u_int32_t ss_sumsum; /* 0: check sum of summary block */ 480 u_int32_t ss_datasum; /* 4: check sum of data */ 481 u_int32_t ss_magic; /* 8: segment summary magic number */ 482 #define SS_MAGIC 0x061561 483 ufs_daddr_t ss_next; /* 12: next segment */ 484 u_int32_t ss_create; /* 16: creation time stamp */ 485 u_int16_t ss_nfinfo; /* 20: number of file info structures */ 486 u_int16_t ss_ninos; /* 22: number of inodes in summary */ 487 488 #define SS_DIROP 0x01 /* segment begins a dirop */ 489 #define SS_CONT 0x02 /* more partials to finish this write*/ 490 u_int16_t ss_flags; /* 24: used for directory operations */ 491 u_int16_t ss_pad; /* 26: extra space */ 492 /* FINFO's and inode daddr's... */ 493 }; 494 495 typedef struct segsum SEGSUM; 496 struct segsum { 497 u_int32_t ss_sumsum; /* 0: check sum of summary block */ 498 u_int32_t ss_datasum; /* 4: check sum of data */ 499 u_int32_t ss_magic; /* 8: segment summary magic number */ 500 ufs_daddr_t ss_next; /* 12: next segment */ 501 u_int32_t ss_ident; /* 16: roll-forward fsid */ 502 #define ss_ocreate ss_ident /* ident is where create was in v1 */ 503 u_int16_t ss_nfinfo; /* 20: number of file info structures */ 504 u_int16_t ss_ninos; /* 22: number of inodes in summary */ 505 u_int16_t ss_flags; /* 24: used for directory operations */ 506 u_int8_t ss_pad[6]; /* 26: extra space */ 507 u_int64_t ss_serial; /* 32: serial number */ 508 u_int64_t ss_create; /* 40: time stamp */ 509 /* FINFO's and inode daddr's... */ 510 }; 511 512 #define SEGSUM_SIZE(fs) ((fs)->lfs_version == 1 ? sizeof(SEGSUM_V1) : sizeof(SEGSUM)) 513 514 /* NINDIR is the number of indirects in a file system block. */ 515 #define NINDIR(fs) ((fs)->lfs_nindir) 516 517 /* INOPB is the number of inodes in a secondary storage block. */ 518 #define INOPB(fs) ((fs)->lfs_inopb) 519 /* INOPF is the number of inodes in a fragment. */ 520 #define INOPF(fs) ((fs)->lfs_inopf) 521 522 #define blksize(fs, ip, lbn) \ 523 (((lbn) >= NDADDR || (ip)->i_ffs_size >= ((lbn) + 1) << (fs)->lfs_bshift) \ 524 ? (fs)->lfs_bsize \ 525 : (fragroundup(fs, blkoff(fs, (ip)->i_ffs_size)))) 526 #define blkoff(fs, loc) ((int)(loc) & (fs)->lfs_bmask) 527 #define fragoff(fs, loc) /* calculates (loc % fs->lfs_fsize) */ \ 528 ((int)((loc) & (fs)->lfs_ffmask)) 529 #define fsbtodb(fs, b) ((b) << (fs)->lfs_fsbtodb) 530 #define dbtofsb(fs, b) ((b) >> (fs)->lfs_fsbtodb) 531 #define fragstodb(fs, b) ((b) << ((fs)->lfs_blktodb - (fs)->lfs_fbshift)) 532 #define dbtofrags(fs, b) ((b) >> ((fs)->lfs_blktodb - (fs)->lfs_fbshift)) 533 #define lblkno(fs, loc) ((loc) >> (fs)->lfs_bshift) 534 #define lblktosize(fs, blk) ((blk) << (fs)->lfs_bshift) 535 /* Same as above, but named like dbtob(), btodb() */ 536 #define fsbtob(fs, b) ((b) << ((fs)->lfs_bshift - \ 537 (fs)->lfs_blktodb + (fs)->lfs_fsbtodb)) 538 #define btofsb(fs, b) ((b) >> ((fs)->lfs_bshift - \ 539 (fs)->lfs_blktodb + (fs)->lfs_fsbtodb)) 540 #define fsbtofrags(fs, b) ((b) >> ((fs)->lfs_blktodb - (fs)->lfs_fbshift - \ 541 (fs)->lfs_fsbtodb)) 542 #define fragstofsb(fs, b) ((b) << ((fs)->lfs_blktodb - (fs)->lfs_fbshift - \ 543 (fs)->lfs_fsbtodb)) 544 #define btofrags(fs, b) ((b) >> (fs)->lfs_ffshift) 545 #define numfrags(fs, loc) /* calculates (loc / fs->lfs_fsize) */ \ 546 ((loc) >> (fs)->lfs_ffshift) 547 #define blkroundup(fs, size) /* calculates roundup(size, fs->lfs_bsize) */ \ 548 ((int)(((size) + (fs)->lfs_bmask) & (~(fs)->lfs_bmask))) 549 #define fragroundup(fs, size) /* calculates roundup(size, fs->lfs_fsize) */ \ 550 ((int)(((size) + (fs)->lfs_ffmask) & (~(fs)->lfs_ffmask))) 551 #define fragstoblks(fs, frags) /* calculates (frags / fs->lfs_frag) */ \ 552 ((frags) >> (fs)->lfs_fbshift) 553 #define blkstofrags(fs, blks) /* calculates (blks * fs->lfs_frag) */ \ 554 ((blks) << (fs)->lfs_fbshift) 555 #define fragnum(fs, fsb) /* calculates (fsb % fs->lfs_frag) */ \ 556 ((fsb) & ((fs)->lfs_frag - 1)) 557 #define blknum(fs, fsb) /* calculates rounddown(fsb, fs->lfs_frag) */ \ 558 ((fsb) &~ ((fs)->lfs_frag - 1)) 559 #define dblksize(fs, dip, lbn) \ 560 (((lbn) >= NDADDR || (dip)->di_size >= ((lbn) + 1) << (fs)->lfs_bshift)\ 561 ? (fs)->lfs_bsize \ 562 : (fragroundup(fs, blkoff(fs, (dip)->di_size)))) 563 564 #define segtod(fs, seg) (((fs)->lfs_version == 1 ? \ 565 (fs)->lfs_ssize << (fs)->lfs_blktodb : \ 566 btofsb((fs), (fs)->lfs_ssize)) * (seg)) 567 #define dtosn(fs, daddr) /* block address to segment number */ \ 568 (((daddr) - (fs)->lfs_start) / segtod((fs), 1)) 569 #define sntod(fs, sn) /* segment number to disk address */ \ 570 ((ufs_daddr_t)(segtod((fs), (sn)) + (fs)->lfs_start)) 571 572 /* Read in the block with the cleaner info from the ifile. */ 573 #define LFS_CLEANERINFO(CP, F, BP) { \ 574 VTOI((F)->lfs_ivnode)->i_flag |= IN_ACCESS; \ 575 if (bread((F)->lfs_ivnode, \ 576 (ufs_daddr_t)0, (F)->lfs_bsize, NOCRED, &(BP))) \ 577 panic("lfs: ifile read"); \ 578 (CP) = (CLEANERINFO *)(BP)->b_data; \ 579 } 580 581 /* Synchronize the Ifile cleaner info with current avail and bfree */ 582 #define LFS_SYNC_CLEANERINFO(cip, fs, bp, w) do { \ 583 if ((w) || (cip)->bfree != (fs)->lfs_bfree || \ 584 (cip)->avail != (fs)->lfs_avail - (fs)->lfs_ravail) { \ 585 (cip)->bfree = (fs)->lfs_bfree; \ 586 (cip)->avail = (fs)->lfs_avail - (fs)->lfs_ravail; \ 587 (void) VOP_BWRITE(bp); /* Ifile */ \ 588 } else \ 589 brelse(bp); \ 590 } while(0) 591 592 #define LFS_GET_HEADFREE(FS, CIP, BP, FREEP) do { \ 593 if ((FS)->lfs_version > 1) { \ 594 LFS_CLEANERINFO((CIP), (FS), (BP)); \ 595 (FS)->lfs_free = (CIP)->free_head; \ 596 brelse(BP); \ 597 } \ 598 *(FREEP) = (FS)->lfs_free; \ 599 } while (0) 600 601 #define LFS_PUT_HEADFREE(FS, CIP, BP, VAL) do { \ 602 (FS)->lfs_free = (VAL); \ 603 if ((FS)->lfs_version > 1) { \ 604 LFS_CLEANERINFO((CIP), (FS), (BP)); \ 605 (CIP)->free_head = (VAL); \ 606 VOP_BWRITE(BP); \ 607 } \ 608 } while (0) 609 610 #define LFS_GET_TAILFREE(FS, CIP, BP, FREEP) do { \ 611 LFS_CLEANERINFO((CIP), (FS), (BP)); \ 612 *(FREEP) = (CIP)->free_tail; \ 613 brelse(BP); \ 614 } while (0) 615 616 #define LFS_PUT_TAILFREE(FS, CIP, BP, VAL) do { \ 617 LFS_CLEANERINFO((CIP), (FS), (BP)); \ 618 (CIP)->free_tail = (VAL); \ 619 VOP_BWRITE(BP); \ 620 } while (0) 621 622 /* 623 * XXX - v1 compatibility code is not allowed to touch if_atime, since it 624 * may not be mapped! 625 */ 626 /* Read in the block with a specific inode from the ifile. */ 627 #define LFS_IENTRY(IP, F, IN, BP) { \ 628 int _e; \ 629 VTOI((F)->lfs_ivnode)->i_flag |= IN_ACCESS; \ 630 if ((_e = bread((F)->lfs_ivnode, \ 631 (IN) / (F)->lfs_ifpb + (F)->lfs_cleansz + (F)->lfs_segtabsz, \ 632 (F)->lfs_bsize, NOCRED, &(BP))) != 0) \ 633 panic("lfs: ifile read %d", _e); \ 634 if((F)->lfs_version == 1) \ 635 (IP) = (IFILE *)((IFILE_V1 *)(BP)->b_data + (IN) % (F)->lfs_ifpb); \ 636 else \ 637 (IP) = (IFILE *)(BP)->b_data + (IN) % (F)->lfs_ifpb; \ 638 } 639 640 /* Read in the block with a specific segment usage entry from the ifile. */ 641 #define LFS_SEGENTRY(SP, F, IN, BP) { \ 642 int _e; \ 643 VTOI((F)->lfs_ivnode)->i_flag |= IN_ACCESS; \ 644 if ((_e = bread((F)->lfs_ivnode, \ 645 ((IN) / (F)->lfs_sepb) + (F)->lfs_cleansz, \ 646 (F)->lfs_bsize, NOCRED, &(BP))) != 0) \ 647 panic("lfs: ifile read: %d", _e); \ 648 if ((F)->lfs_version == 1) \ 649 (SP) = (SEGUSE *)((SEGUSE_V1 *)(BP)->b_data + \ 650 ((IN) & ((F)->lfs_sepb - 1))); \ 651 else \ 652 (SP) = (SEGUSE *)(BP)->b_data + ((IN) % (F)->lfs_sepb); \ 653 } 654 655 /* Determine if a buffer belongs to the ifile */ 656 #define IS_IFILE(bp) (VTOI(bp->b_vp)->i_number == LFS_IFILE_INUM) 657 658 /* 659 * Structures used by lfs_bmapv and lfs_markv to communicate information 660 * about inodes and data blocks. 661 */ 662 typedef struct block_info { 663 ino_t bi_inode; /* inode # */ 664 ufs_daddr_t bi_lbn; /* logical block w/in file */ 665 ufs_daddr_t bi_daddr; /* disk address of block */ 666 u_int64_t bi_segcreate; /* origin segment create time */ 667 int bi_version; /* file version number */ 668 void *bi_bp; /* data buffer */ 669 int bi_size; /* size of the block (if fragment) */ 670 } BLOCK_INFO; 671 672 /* Compatibility for 1.5 binaries */ 673 typedef struct block_info_15 { 674 ino_t bi_inode; /* inode # */ 675 ufs_daddr_t bi_lbn; /* logical block w/in file */ 676 ufs_daddr_t bi_daddr; /* disk address of block */ 677 u_int32_t bi_segcreate; /* origin segment create time */ 678 int bi_version; /* file version number */ 679 void *bi_bp; /* data buffer */ 680 int bi_size; /* size of the block (if fragment) */ 681 } BLOCK_INFO_15; 682 683 /* In-memory description of a segment about to be written. */ 684 struct segment { 685 struct lfs *fs; /* file system pointer */ 686 struct buf **bpp; /* pointer to buffer array */ 687 struct buf **cbpp; /* pointer to next available bp */ 688 struct buf **start_bpp; /* pointer to first bp in this set */ 689 struct buf *ibp; /* buffer pointer to inode page */ 690 struct dinode *idp; /* pointer to ifile dinode */ 691 struct finfo *fip; /* current fileinfo pointer */ 692 struct vnode *vp; /* vnode being gathered */ 693 void *segsum; /* segment summary info */ 694 u_int32_t ninodes; /* number of inodes in this segment */ 695 u_int32_t seg_bytes_left; /* bytes left in segment */ 696 u_int32_t sum_bytes_left; /* bytes left in summary block */ 697 u_int32_t seg_number; /* number of this segment */ 698 ufs_daddr_t *start_lbp; /* beginning lbn for this set */ 699 700 #define SEGM_CKP 0x01 /* doing a checkpoint */ 701 #define SEGM_CLEAN 0x02 /* cleaner call; don't sort */ 702 #define SEGM_SYNC 0x04 /* wait for segment */ 703 #define SEGM_PROT 0x08 /* don't inactivate at segunlock */ 704 u_int16_t seg_flags; /* run-time flags for this segment */ 705 }; 706 707 /* 708 * Macros for determining free space on the disk, with the variable metadata 709 * of segment summaries and inode blocks taken into account. 710 */ 711 /* Estimate number of clean blocks not available for writing */ 712 #define LFS_EST_CMETA(F) (int32_t)((((F)->lfs_dmeta * \ 713 (int64_t)(F)->lfs_nclean) / \ 714 ((F)->lfs_nseg - (F)->lfs_nclean))) 715 716 /* Estimate total size of the disk not including metadata */ 717 #define LFS_EST_NONMETA(F) ((F)->lfs_dsize - (F)->lfs_dmeta - LFS_EST_CMETA(F)) 718 719 /* Estimate number of blocks actually available for writing */ 720 #define LFS_EST_BFREE(F) ((F)->lfs_bfree - LFS_EST_CMETA(F) - (F)->lfs_dmeta) 721 722 /* Amount of non-meta space not available to mortal man */ 723 #define LFS_EST_RSVD(F) (int32_t)((LFS_EST_NONMETA(F) * \ 724 (u_int64_t)(F)->lfs_minfree) / \ 725 100) 726 727 /* Can credential C write BB blocks */ 728 #define ISSPACE(F, BB, C) \ 729 ((((C) == NOCRED || (C)->cr_uid == 0) && \ 730 LFS_EST_BFREE(F) >= (BB)) || \ 731 ((C)->cr_uid != 0 && IS_FREESPACE(F, BB))) 732 733 /* Can an ordinary user write BB blocks */ 734 #define IS_FREESPACE(F, BB) \ 735 (LFS_EST_BFREE(F) >= (BB) + LFS_EST_RSVD(F)) 736 737 /* Statistics Counters */ 738 struct lfs_stats { 739 u_int segsused; 740 u_int psegwrites; 741 u_int psyncwrites; 742 u_int pcleanwrites; 743 u_int blocktot; 744 u_int cleanblocks; 745 u_int ncheckpoints; 746 u_int nwrites; 747 u_int nsync_writes; 748 u_int wait_exceeded; 749 u_int write_exceeded; 750 u_int flush_invoked; 751 u_int vflush_invoked; 752 }; 753 extern struct lfs_stats lfs_stats; 754
Indexes created Mon Sep 29 21:09:56 GMT 2025