freebsd-dev/sbin/fsck_ffs/inode.c
Kirk McKusick 1c85e6a35d This commit adds basic support for the UFS2 filesystem. The UFS2
filesystem expands the inode to 256 bytes to make space for 64-bit
block pointers. It also adds a file-creation time field, an ability
to use jumbo blocks per inode to allow extent like pointer density,
and space for extended attributes (up to twice the filesystem block
size worth of attributes, e.g., on a 16K filesystem, there is space
for 32K of attributes). UFS2 fully supports and runs existing UFS1
filesystems. New filesystems built using newfs can be built in either
UFS1 or UFS2 format using the -O option. In this commit UFS1 is
the default format, so if you want to build UFS2 format filesystems,
you must specify -O 2. This default will be changed to UFS2 when
UFS2 proves itself to be stable. In this commit the boot code for
reading UFS2 filesystems is not compiled (see /sys/boot/common/ufsread.c)
as there is insufficient space in the boot block. Once the size of the
boot block is increased, this code can be defined.

Things to note: the definition of SBSIZE has changed to SBLOCKSIZE.
The header file <ufs/ufs/dinode.h> must be included before
<ufs/ffs/fs.h> so as to get the definitions of ufs2_daddr_t and
ufs_lbn_t.

Still TODO:
Verify that the first level bootstraps work for all the architectures.
Convert the utility ffsinfo to understand UFS2 and test growfs.
Add support for the extended attribute storage. Update soft updates
to ensure integrity of extended attribute storage. Switch the
current extended attribute interfaces to use the extended attribute
storage. Add the extent like functionality (framework is there,
but is currently never used).

Sponsored by: DARPA & NAI Labs.
Reviewed by:	Poul-Henning Kamp <phk@freebsd.org>
2002-06-21 06:18:05 +00:00

677 lines
17 KiB
C

/*
* Copyright (c) 1980, 1986, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* California, Berkeley and its contributors.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#ifndef lint
#if 0
static const char sccsid[] = "@(#)inode.c 8.8 (Berkeley) 4/28/95";
#endif
static const char rcsid[] =
"$FreeBSD$";
#endif /* not lint */
#include <sys/param.h>
#include <sys/stdint.h>
#include <sys/time.h>
#include <sys/sysctl.h>
#include <ufs/ufs/dinode.h>
#include <ufs/ufs/dir.h>
#include <ufs/ffs/fs.h>
#include <err.h>
#include <pwd.h>
#include <string.h>
#include "fsck.h"
static ino_t startinum;
static int iblock(struct inodesc *, long ilevel, off_t isize);
int
ckinode(union dinode *dp, struct inodesc *idesc)
{
off_t remsize, sizepb;
int i, offset, ret;
union dinode dino;
ufs2_daddr_t ndb;
mode_t mode;
char pathbuf[MAXPATHLEN + 1];
if (idesc->id_fix != IGNORE)
idesc->id_fix = DONTKNOW;
idesc->id_lbn = -1;
idesc->id_entryno = 0;
idesc->id_filesize = DIP(dp, di_size);
mode = DIP(dp, di_mode) & IFMT;
if (mode == IFBLK || mode == IFCHR || (mode == IFLNK &&
DIP(dp, di_size) < (unsigned)sblock.fs_maxsymlinklen))
return (KEEPON);
if (sblock.fs_magic == FS_UFS1_MAGIC)
dino.dp1 = dp->dp1;
else
dino.dp2 = dp->dp2;
ndb = howmany(DIP(&dino, di_size), sblock.fs_bsize);
for (i = 0; i < NDADDR; i++) {
idesc->id_lbn++;
if (--ndb == 0 &&
(offset = blkoff(&sblock, DIP(&dino, di_size))) != 0)
idesc->id_numfrags =
numfrags(&sblock, fragroundup(&sblock, offset));
else
idesc->id_numfrags = sblock.fs_frag;
if (DIP(&dino, di_db[i]) == 0) {
if (idesc->id_type == DATA && ndb >= 0) {
/* An empty block in a directory XXX */
getpathname(pathbuf, idesc->id_number,
idesc->id_number);
pfatal("DIRECTORY %s: CONTAINS EMPTY BLOCKS",
pathbuf);
if (reply("ADJUST LENGTH") == 1) {
dp = ginode(idesc->id_number);
DIP(dp, di_size) = i * sblock.fs_bsize;
printf(
"YOU MUST RERUN FSCK AFTERWARDS\n");
rerun = 1;
inodirty();
}
}
continue;
}
idesc->id_blkno = DIP(&dino, di_db[i]);
if (idesc->id_type != DATA)
ret = (*idesc->id_func)(idesc);
else
ret = dirscan(idesc);
if (ret & STOP)
return (ret);
}
idesc->id_numfrags = sblock.fs_frag;
remsize = DIP(&dino, di_size) - sblock.fs_bsize * NDADDR;
sizepb = sblock.fs_bsize;
for (i = 0; i < NIADDR; i++) {
sizepb *= NINDIR(&sblock);
if (DIP(&dino, di_ib[i])) {
idesc->id_blkno = DIP(&dino, di_ib[i]);
ret = iblock(idesc, i + 1, remsize);
if (ret & STOP)
return (ret);
} else {
idesc->id_lbn += sizepb / sblock.fs_bsize;
if (idesc->id_type == DATA && remsize > 0) {
/* An empty block in a directory XXX */
getpathname(pathbuf, idesc->id_number,
idesc->id_number);
pfatal("DIRECTORY %s: CONTAINS EMPTY BLOCKS",
pathbuf);
if (reply("ADJUST LENGTH") == 1) {
dp = ginode(idesc->id_number);
DIP(dp, di_size) -= remsize;
remsize = 0;
printf(
"YOU MUST RERUN FSCK AFTERWARDS\n");
rerun = 1;
inodirty();
break;
}
}
}
remsize -= sizepb;
}
return (KEEPON);
}
static int
iblock(struct inodesc *idesc, long ilevel, off_t isize)
{
struct bufarea *bp;
int i, n, (*func)(), nif;
off_t sizepb;
char buf[BUFSIZ];
char pathbuf[MAXPATHLEN + 1];
union dinode *dp;
if (idesc->id_type != DATA) {
func = idesc->id_func;
if (((n = (*func)(idesc)) & KEEPON) == 0)
return (n);
} else
func = dirscan;
if (chkrange(idesc->id_blkno, idesc->id_numfrags))
return (SKIP);
bp = getdatablk(idesc->id_blkno, sblock.fs_bsize);
ilevel--;
for (sizepb = sblock.fs_bsize, i = 0; i < ilevel; i++)
sizepb *= NINDIR(&sblock);
if (howmany(isize, sizepb) > NINDIR(&sblock))
nif = NINDIR(&sblock);
else
nif = howmany(isize, sizepb);
if (idesc->id_func == pass1check && nif < NINDIR(&sblock)) {
for (i = nif; i < NINDIR(&sblock); i++) {
if (IBLK(bp, i) == 0)
continue;
(void)sprintf(buf, "PARTIALLY TRUNCATED INODE I=%lu",
(u_long)idesc->id_number);
if (preen) {
pfatal("%s", buf);
} else if (dofix(idesc, buf)) {
IBLK(bp, i) = 0;
dirty(bp);
}
}
flush(fswritefd, bp);
}
for (i = 0; i < nif; i++) {
if (ilevel == 0)
idesc->id_lbn++;
if (IBLK(bp, i)) {
idesc->id_blkno = IBLK(bp, i);
if (ilevel == 0)
n = (*func)(idesc);
else
n = iblock(idesc, ilevel, isize);
if (n & STOP) {
bp->b_flags &= ~B_INUSE;
return (n);
}
} else {
if (idesc->id_type == DATA && isize > 0) {
/* An empty block in a directory XXX */
getpathname(pathbuf, idesc->id_number,
idesc->id_number);
pfatal("DIRECTORY %s: CONTAINS EMPTY BLOCKS",
pathbuf);
if (reply("ADJUST LENGTH") == 1) {
dp = ginode(idesc->id_number);
DIP(dp, di_size) -= isize;
isize = 0;
printf(
"YOU MUST RERUN FSCK AFTERWARDS\n");
rerun = 1;
inodirty();
bp->b_flags &= ~B_INUSE;
return(STOP);
}
}
}
isize -= sizepb;
}
bp->b_flags &= ~B_INUSE;
return (KEEPON);
}
/*
* Check that a block in a legal block number.
* Return 0 if in range, 1 if out of range.
*/
int
chkrange(ufs2_daddr_t blk, int cnt)
{
int c;
if (cnt <= 0 || blk <= 0 || blk > maxfsblock ||
cnt - 1 > maxfsblock - blk)
return (1);
if (cnt > sblock.fs_frag ||
fragnum(&sblock, blk) + cnt > sblock.fs_frag) {
if (debug)
printf("bad size: blk %ld, offset %i, size %d\n",
(long)blk, (int)fragnum(&sblock, blk), cnt);
return (1);
}
c = dtog(&sblock, blk);
if (blk < cgdmin(&sblock, c)) {
if ((blk + cnt) > cgsblock(&sblock, c)) {
if (debug) {
printf("blk %ld < cgdmin %ld;",
(long)blk, (long)cgdmin(&sblock, c));
printf(" blk + cnt %ld > cgsbase %ld\n",
(long)(blk + cnt),
(long)cgsblock(&sblock, c));
}
return (1);
}
} else {
if ((blk + cnt) > cgbase(&sblock, c+1)) {
if (debug) {
printf("blk %ld >= cgdmin %ld;",
(long)blk, (long)cgdmin(&sblock, c));
printf(" blk + cnt %ld > sblock.fs_fpg %ld\n",
(long)(blk + cnt), (long)sblock.fs_fpg);
}
return (1);
}
}
return (0);
}
/*
* General purpose interface for reading inodes.
*/
union dinode *
ginode(ino_t inumber)
{
ufs2_daddr_t iblk;
if (inumber < ROOTINO || inumber > maxino)
errx(EEXIT, "bad inode number %d to ginode", inumber);
if (startinum == 0 ||
inumber < startinum || inumber >= startinum + INOPB(&sblock)) {
iblk = ino_to_fsba(&sblock, inumber);
if (pbp != 0)
pbp->b_flags &= ~B_INUSE;
pbp = getdatablk(iblk, sblock.fs_bsize);
startinum = (inumber / INOPB(&sblock)) * INOPB(&sblock);
}
if (sblock.fs_magic == FS_UFS1_MAGIC)
return ((union dinode *)
&pbp->b_un.b_dinode1[inumber % INOPB(&sblock)]);
return ((union dinode *)&pbp->b_un.b_dinode2[inumber % INOPB(&sblock)]);
}
/*
* Special purpose version of ginode used to optimize first pass
* over all the inodes in numerical order.
*/
static ino_t nextino, lastinum, lastvalidinum;
static long readcnt, readpercg, fullcnt, inobufsize, partialcnt, partialsize;
static caddr_t inodebuf;
union dinode *
getnextinode(ino_t inumber)
{
long size;
ufs2_daddr_t dblk;
union dinode *dp;
static caddr_t nextinop;
if (inumber != nextino++ || inumber > lastvalidinum)
errx(EEXIT, "bad inode number %d to nextinode", inumber);
if (inumber >= lastinum) {
readcnt++;
dblk = fsbtodb(&sblock, ino_to_fsba(&sblock, lastinum));
if (readcnt % readpercg == 0) {
size = partialsize;
lastinum += partialcnt;
} else {
size = inobufsize;
lastinum += fullcnt;
}
/*
* If bread returns an error, it will already have zeroed
* out the buffer, so we do not need to do so here.
*/
(void)bread(fsreadfd, inodebuf, dblk, size);
nextinop = inodebuf;
}
dp = (union dinode *)nextinop;
if (sblock.fs_magic == FS_UFS1_MAGIC)
nextinop += sizeof(struct ufs1_dinode);
else
nextinop += sizeof(struct ufs2_dinode);
return (dp);
}
void
setinodebuf(ino_t inum)
{
if (inum % sblock.fs_ipg != 0)
errx(EEXIT, "bad inode number %d to setinodebuf", inum);
lastvalidinum = inum + sblock.fs_ipg - 1;
startinum = 0;
nextino = inum;
lastinum = inum;
readcnt = 0;
if (inodebuf != NULL)
return;
inobufsize = blkroundup(&sblock, INOBUFSIZE);
fullcnt = inobufsize / ((sblock.fs_magic == FS_UFS1_MAGIC) ?
sizeof(struct ufs1_dinode) : sizeof(struct ufs2_dinode));
readpercg = sblock.fs_ipg / fullcnt;
partialcnt = sblock.fs_ipg % fullcnt;
partialsize = partialcnt * ((sblock.fs_magic == FS_UFS1_MAGIC) ?
sizeof(struct ufs1_dinode) : sizeof(struct ufs2_dinode));
if (partialcnt != 0) {
readpercg++;
} else {
partialcnt = fullcnt;
partialsize = inobufsize;
}
if ((inodebuf = malloc((unsigned)inobufsize)) == NULL)
errx(EEXIT, "cannot allocate space for inode buffer");
}
void
freeinodebuf(void)
{
if (inodebuf != NULL)
free((char *)inodebuf);
inodebuf = NULL;
}
/*
* Routines to maintain information about directory inodes.
* This is built during the first pass and used during the
* second and third passes.
*
* Enter inodes into the cache.
*/
void
cacheino(union dinode *dp, ino_t inumber)
{
struct inoinfo *inp, **inpp;
int i, blks;
if (howmany(DIP(dp, di_size), sblock.fs_bsize) > NDADDR)
blks = NDADDR + NIADDR;
else
blks = howmany(DIP(dp, di_size), sblock.fs_bsize);
inp = (struct inoinfo *)
malloc(sizeof(*inp) + (blks - 1) * sizeof(ufs2_daddr_t));
if (inp == NULL)
errx(EEXIT, "cannot increase directory list");
inpp = &inphead[inumber % dirhash];
inp->i_nexthash = *inpp;
*inpp = inp;
inp->i_parent = inumber == ROOTINO ? ROOTINO : (ino_t)0;
inp->i_dotdot = (ino_t)0;
inp->i_number = inumber;
inp->i_isize = DIP(dp, di_size);
inp->i_numblks = blks;
for (i = 0; i < (blks < NDADDR ? blks : NDADDR); i++)
inp->i_blks[i] = DIP(dp, di_db[i]);
if (blks > NDADDR)
for (i = 0; i < NIADDR; i++)
inp->i_blks[NDADDR + i] = DIP(dp, di_ib[i]);
if (inplast == listmax) {
listmax += 100;
inpsort = (struct inoinfo **)realloc((char *)inpsort,
(unsigned)listmax * sizeof(struct inoinfo *));
if (inpsort == NULL)
errx(EEXIT, "cannot increase directory list");
}
inpsort[inplast++] = inp;
}
/*
* Look up an inode cache structure.
*/
struct inoinfo *
getinoinfo(ino_t inumber)
{
struct inoinfo *inp;
for (inp = inphead[inumber % dirhash]; inp; inp = inp->i_nexthash) {
if (inp->i_number != inumber)
continue;
return (inp);
}
errx(EEXIT, "cannot find inode %d", inumber);
return ((struct inoinfo *)0);
}
/*
* Clean up all the inode cache structure.
*/
void
inocleanup(void)
{
struct inoinfo **inpp;
if (inphead == NULL)
return;
for (inpp = &inpsort[inplast - 1]; inpp >= inpsort; inpp--)
free((char *)(*inpp));
free((char *)inphead);
free((char *)inpsort);
inphead = inpsort = NULL;
}
void
inodirty(void)
{
dirty(pbp);
}
void
clri(struct inodesc *idesc, char *type, int flag)
{
union dinode *dp;
dp = ginode(idesc->id_number);
if (flag == 1) {
pwarn("%s %s", type,
(DIP(dp, di_mode) & IFMT) == IFDIR ? "DIR" : "FILE");
pinode(idesc->id_number);
}
if (preen || reply("CLEAR") == 1) {
if (preen)
printf(" (CLEARED)\n");
n_files--;
if (bkgrdflag == 0) {
(void)ckinode(dp, idesc);
inoinfo(idesc->id_number)->ino_state = USTATE;
clearinode(dp);
inodirty();
} else {
cmd.value = idesc->id_number;
cmd.size = -DIP(dp, di_nlink);
if (debug)
printf("adjrefcnt ino %ld amt %ld\n",
(long)cmd.value, cmd.size);
if (sysctl(adjrefcnt, MIBSIZE, 0, 0,
&cmd, sizeof cmd) == -1)
rwerror("ADJUST INODE", cmd.value);
}
}
}
int
findname(struct inodesc *idesc)
{
struct direct *dirp = idesc->id_dirp;
if (dirp->d_ino != idesc->id_parent || idesc->id_entryno < 2) {
idesc->id_entryno++;
return (KEEPON);
}
memmove(idesc->id_name, dirp->d_name, (size_t)dirp->d_namlen + 1);
return (STOP|FOUND);
}
int
findino(struct inodesc *idesc)
{
struct direct *dirp = idesc->id_dirp;
if (dirp->d_ino == 0)
return (KEEPON);
if (strcmp(dirp->d_name, idesc->id_name) == 0 &&
dirp->d_ino >= ROOTINO && dirp->d_ino <= maxino) {
idesc->id_parent = dirp->d_ino;
return (STOP|FOUND);
}
return (KEEPON);
}
int
clearentry(struct inodesc *idesc)
{
struct direct *dirp = idesc->id_dirp;
if (dirp->d_ino != idesc->id_parent || idesc->id_entryno < 2) {
idesc->id_entryno++;
return (KEEPON);
}
dirp->d_ino = 0;
return (STOP|FOUND|ALTERED);
}
void
pinode(ino_t ino)
{
union dinode *dp;
char *p;
struct passwd *pw;
time_t t;
printf(" I=%lu ", (u_long)ino);
if (ino < ROOTINO || ino > maxino)
return;
dp = ginode(ino);
printf(" OWNER=");
if ((pw = getpwuid((int)DIP(dp, di_uid))) != 0)
printf("%s ", pw->pw_name);
else
printf("%u ", (unsigned)DIP(dp, di_uid));
printf("MODE=%o\n", DIP(dp, di_mode));
if (preen)
printf("%s: ", cdevname);
printf("SIZE=%qu ", DIP(dp, di_size));
t = DIP(dp, di_mtime);
p = ctime(&t);
printf("MTIME=%12.12s %4.4s ", &p[4], &p[20]);
}
void
blkerror(ino_t ino, char *type, ufs2_daddr_t blk)
{
pfatal("%lld %s I=%lu", (intmax_t)blk, type, (u_long)ino);
printf("\n");
switch (inoinfo(ino)->ino_state) {
case FSTATE:
inoinfo(ino)->ino_state = FCLEAR;
return;
case DSTATE:
inoinfo(ino)->ino_state = DCLEAR;
return;
case FCLEAR:
case DCLEAR:
return;
default:
errx(EEXIT, "BAD STATE %d TO BLKERR", inoinfo(ino)->ino_state);
/* NOTREACHED */
}
}
/*
* allocate an unused inode
*/
ino_t
allocino(ino_t request, int type)
{
ino_t ino;
union dinode *dp;
struct cg *cgp = &cgrp;
int cg;
if (request == 0)
request = ROOTINO;
else if (inoinfo(request)->ino_state != USTATE)
return (0);
for (ino = request; ino < maxino; ino++)
if (inoinfo(ino)->ino_state == USTATE)
break;
if (ino == maxino)
return (0);
cg = ino_to_cg(&sblock, ino);
getblk(&cgblk, cgtod(&sblock, cg), sblock.fs_cgsize);
if (!cg_chkmagic(cgp))
pfatal("CG %d: BAD MAGIC NUMBER\n", cg);
setbit(cg_inosused(cgp), ino % sblock.fs_ipg);
cgp->cg_cs.cs_nifree--;
switch (type & IFMT) {
case IFDIR:
inoinfo(ino)->ino_state = DSTATE;
cgp->cg_cs.cs_ndir++;
break;
case IFREG:
case IFLNK:
inoinfo(ino)->ino_state = FSTATE;
break;
default:
return (0);
}
cgdirty();
dp = ginode(ino);
DIP(dp, di_db[0]) = allocblk((long)1);
if (DIP(dp, di_db[0]) == 0) {
inoinfo(ino)->ino_state = USTATE;
return (0);
}
DIP(dp, di_mode) = type;
DIP(dp, di_flags) = 0;
DIP(dp, di_atime) = time(NULL);
DIP(dp, di_mtime) = DIP(dp, di_ctime) = DIP(dp, di_atime);
DIP(dp, di_mtimensec) = 0;
DIP(dp, di_ctimensec) = 0;
DIP(dp, di_atimensec) = 0;
DIP(dp, di_size) = sblock.fs_fsize;
DIP(dp, di_blocks) = btodb(sblock.fs_fsize);
n_files++;
inodirty();
inoinfo(ino)->ino_type = IFTODT(type);
return (ino);
}
/*
* deallocate an inode
*/
void
freeino(ino_t ino)
{
struct inodesc idesc;
union dinode *dp;
memset(&idesc, 0, sizeof(struct inodesc));
idesc.id_type = ADDR;
idesc.id_func = pass4check;
idesc.id_number = ino;
dp = ginode(ino);
(void)ckinode(dp, &idesc);
clearinode(dp);
inodirty();
inoinfo(ino)->ino_state = USTATE;
n_files--;
}