NetBSD/sbin/fsck_lfs/pass1.c

431 lines
11 KiB
C
Raw Normal View History

Merge the short-lived perseant-lfsv2 branch into the trunk. Kernels and tools understand both v1 and v2 filesystems; newfs_lfs generates v2 by default. Changes for the v2 layout include: - Segments of non-PO2 size and arbitrary block offset, so these can be matched to convenient physical characteristics of the partition (e.g., stripe or track size and offset). - Address by fragment instead of by disk sector, paving the way for non-512-byte-sector devices. In theory fragments can be as large as you like, though in reality they must be smaller than MAXBSIZE in size. - Use serial number and filesystem identifier to ensure that roll-forward doesn't get old data and think it's new. Roll-forward is enabled for v2 filesystems, though not for v1 filesystems by default. - The inode free list is now a tailq, paving the way for undelete (undelete is not yet implemented, but can be without further non-backwards-compatible changes to disk structures). - Inode atime information is kept in the Ifile, instead of on the inode; that is, the inode is never written *just* because atime was changed. Because of this the inodes remain near the file data on the disk, rather than wandering all over as the disk is read repeatedly. This speeds up repeated reads by a small but noticeable amount. Other changes of note include: - The ifile written by newfs_lfs can now be of arbitrary length, it is no longer restricted to a single indirect block. - Fixed an old bug where ctime was changed every time a vnode was created. I need to look more closely to make sure that the times are only updated during write(2) and friends, not after-the-fact during a segment write, and certainly not by the cleaner.
2001-07-14 00:30:18 +04:00
/* $NetBSD: pass1.c,v 1.11 2001/07/13 20:30:19 perseant Exp $ */
/*
* 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.
*/
#include <sys/param.h>
#include <sys/time.h>
#include <ufs/ufs/dinode.h>
#include <ufs/ufs/dir.h>
#include <sys/mount.h>
#include <ufs/lfs/lfs.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <signal.h>
#include "fsck.h"
#include "extern.h"
#include "fsutil.h"
2000-05-23 05:48:52 +04:00
SEGUSE *seg_table;
extern daddr_t *din_table;
2000-05-23 05:48:52 +04:00
static daddr_t badblk;
static daddr_t dupblk;
static void checkinode(ino_t, struct inodesc *);
static int i_d_cmp(const void *, const void *);
struct ino_daddr {
2000-05-23 05:48:52 +04:00
ino_t ino;
daddr_t daddr;
};
static int
i_d_cmp(const void *va, const void *vb)
{
struct ino_daddr *a, *b;
a = *((struct ino_daddr **)va);
b = *((struct ino_daddr **)vb);
if (a->daddr == b->daddr) {
return (a->ino - b->ino);
}
if (a->daddr > b->daddr) {
return 1;
}
return -1;
}
void
pass1()
{
2000-05-23 05:48:52 +04:00
ino_t inumber;
Merge the short-lived perseant-lfsv2 branch into the trunk. Kernels and tools understand both v1 and v2 filesystems; newfs_lfs generates v2 by default. Changes for the v2 layout include: - Segments of non-PO2 size and arbitrary block offset, so these can be matched to convenient physical characteristics of the partition (e.g., stripe or track size and offset). - Address by fragment instead of by disk sector, paving the way for non-512-byte-sector devices. In theory fragments can be as large as you like, though in reality they must be smaller than MAXBSIZE in size. - Use serial number and filesystem identifier to ensure that roll-forward doesn't get old data and think it's new. Roll-forward is enabled for v2 filesystems, though not for v1 filesystems by default. - The inode free list is now a tailq, paving the way for undelete (undelete is not yet implemented, but can be without further non-backwards-compatible changes to disk structures). - Inode atime information is kept in the Ifile, instead of on the inode; that is, the inode is never written *just* because atime was changed. Because of this the inodes remain near the file data on the disk, rather than wandering all over as the disk is read repeatedly. This speeds up repeated reads by a small but noticeable amount. Other changes of note include: - The ifile written by newfs_lfs can now be of arbitrary length, it is no longer restricted to a single indirect block. - Fixed an old bug where ctime was changed every time a vnode was created. I need to look more closely to make sure that the times are only updated during write(2) and friends, not after-the-fact during a segment write, and certainly not by the cleaner.
2001-07-14 00:30:18 +04:00
int i;
2000-05-23 05:48:52 +04:00
struct inodesc idesc;
struct dinode *idinode, *tinode;
struct ifile *ifp;
CLEANERINFO *cp;
struct bufarea *bp;
struct ino_daddr **dins;
2000-05-23 05:48:52 +04:00
idinode = lfs_difind(&sblock, sblock.lfs_ifile, &ifblock);
2000-05-23 05:48:52 +04:00
/*
* We now have the ifile's inode block in core. Read out the
* number of segments.
*/
#if 1
2000-05-23 05:48:52 +04:00
if (pbp != 0)
pbp->b_flags &= ~B_INUSE;
pbp = getddblk(idinode->di_db[0], sblock.lfs_bsize);
2000-05-23 05:48:52 +04:00
cp = (CLEANERINFO *)(pbp->b_un.b_buf);
#endif
/*
* Find all allocated blocks, initialize numdirs.
*/
memset(&idesc, 0, sizeof(struct inodesc));
idesc.id_type = ADDR;
idesc.id_func = pass1check;
idesc.id_lblkno = 0;
inumber = 0;
n_files = n_blks = 0;
if (debug)
printf("creating inode address table...\n");
/* Sort by daddr */
dins = (struct ino_daddr **)malloc(maxino * sizeof(*dins));
for (i = 0; i < maxino; i++) {
dins[i] = malloc(sizeof(**dins));
dins[i]->ino = i;
dins[i]->daddr = lfs_ino_daddr(i);
}
qsort(dins, maxino, sizeof(*dins), i_d_cmp);
/* find a value for numdirs, fill in din_table */
if (debug)
printf("counting dirs...\n");
2000-05-23 05:48:52 +04:00
numdirs = 0;
for (i = 0; i < maxino; i++) {
inumber = dins[i]->ino;
2000-05-23 05:48:52 +04:00
if (inumber == 0 || dins[i]->daddr == 0)
continue;
tinode = lfs_ginode(inumber);
2000-05-23 05:48:52 +04:00
if (tinode && (tinode->di_mode & IFMT) == IFDIR)
numdirs++;
}
/* from setup.c */
2000-05-23 05:48:52 +04:00
inplast = 0;
listmax = numdirs + 10;
inpsort = (struct inoinfo **)calloc((unsigned) listmax,
sizeof(struct inoinfo *));
inphead = (struct inoinfo **)calloc((unsigned) numdirs,
sizeof(struct inoinfo *));
if (inpsort == NULL || inphead == NULL) {
printf("cannot alloc %lu bytes for inphead\n",
(unsigned long)numdirs * sizeof(struct inoinfo *));
exit(1);
2000-05-23 05:48:52 +04:00
}
/* resetinodebuf(); */
if (debug)
printf("counting blocks...\n");
for (i = 0; i < maxino; i++) {
inumber = dins[i]->ino;
if (inumber == 0 || dins[i]->daddr == 0) {
statemap[inumber] = USTATE;
continue;
}
2000-05-23 05:48:52 +04:00
ifp = lfs_ientry(inumber, &bp);
if (ifp && ifp->if_daddr != LFS_UNUSED_DADDR) {
bp->b_flags &= ~B_INUSE;
checkinode(inumber, &idesc);
} else {
bp->b_flags &= ~B_INUSE;
statemap[inumber] = USTATE;
}
free(dins[i]);
2000-05-23 05:48:52 +04:00
}
free(dins);
/* freeinodebuf(); */
}
static void
2000-05-23 05:48:52 +04:00
checkinode(ino_t inumber, struct inodesc * idesc)
{
register struct dinode *dp;
2000-05-23 05:48:52 +04:00
struct zlncnt *zlnp;
int ndb, j;
mode_t mode;
char *symbuf;
/* dp = getnextinode(inumber); */
2000-05-23 05:48:52 +04:00
dp = lfs_ginode(inumber);
2000-05-23 05:48:52 +04:00
if (dp == NULL) {
/* pwarn("Could not find inode %ld\n",(long)inumber); */
statemap[inumber] = USTATE;
return;
}
mode = dp->di_mode & IFMT;
2000-05-23 05:48:52 +04:00
/* XXX - LFS doesn't have this particular problem (?) */
if (mode == 0) {
if (memcmp(dp->di_db, zino.di_db, NDADDR * sizeof(daddr_t)) ||
2000-05-23 05:48:52 +04:00
memcmp(dp->di_ib, zino.di_ib, NIADDR * sizeof(daddr_t)) ||
dp->di_mode || dp->di_size) {
pwarn("mode=o%o, ifmt=o%o\n", dp->di_mode, mode);
pfatal("PARTIALLY ALLOCATED INODE I=%u", inumber);
if (reply("CLEAR") == 1) {
dp = ginode(inumber);
clearinode(dp);
inodirty();
}
}
statemap[inumber] = USTATE;
return;
}
lastino = inumber;
2000-05-23 05:48:52 +04:00
if ( /* dp->di_size < 0 || */
dp->di_size + sblock.lfs_bsize - 1 < dp->di_size) {
if (debug)
printf("bad size %llu:",
(unsigned long long)dp->di_size);
goto unknown;
}
if (!preen && mode == IFMT && reply("HOLD BAD BLOCK") == 1) {
dp = ginode(inumber);
dp->di_size = sblock.lfs_fsize;
2000-05-23 05:48:52 +04:00
dp->di_mode = IFREG | 0600;
inodirty();
}
ndb = howmany(dp->di_size, sblock.lfs_bsize);
if (ndb < 0) {
if (debug)
printf("bad size %llu ndb %d:",
2000-05-23 05:48:52 +04:00
(unsigned long long)dp->di_size, ndb);
goto unknown;
}
if (mode == IFBLK || mode == IFCHR)
ndb++;
if (mode == IFLNK) {
/*
* Note that the old fastlink format always had di_blocks set
* to 0. Other than that we no longer use the `spare' field
2000-05-23 05:48:52 +04:00
* (which is now the extended uid)for sanity checking, the
* new format is the same as the old. We simply ignore the
* conversion altogether. - mycroft, 19MAY1994
*/
if (doinglevel2 &&
dp->di_size > 0 && dp->di_size < MAXSYMLINKLEN &&
dp->di_blocks != 0) {
symbuf = alloca(secsize);
if (bread(fsreadfd, symbuf,
2000-05-23 05:48:52 +04:00
fsbtodb(&sblock, dp->di_db[0]),
(long)secsize) != 0)
errexit("cannot read symlink");
if (debug) {
symbuf[dp->di_size] = 0;
printf("convert symlink %d(%s)of size %lld\n",
2000-05-23 05:48:52 +04:00
inumber, symbuf, (long long)dp->di_size);
}
dp = ginode(inumber);
memcpy(dp->di_shortlink, symbuf, (long)dp->di_size);
dp->di_blocks = 0;
inodirty();
}
/*
* Fake ndb value so direct/indirect block checks below
* will detect any garbage after symlink string.
*/
if (dp->di_size < sblock.lfs_maxsymlinklen ||
(sblock.lfs_maxsymlinklen == 0 && dp->di_blocks == 0)) {
ndb = howmany(dp->di_size, sizeof(daddr_t));
if (ndb > NDADDR) {
j = ndb - NDADDR;
for (ndb = 1; j > 1; j--)
ndb *= NINDIR(&sblock);
ndb += NDADDR;
}
}
}
for (j = ndb; j < NDADDR; j++)
if (dp->di_db[j] != 0) {
if (debug)
printf("bad direct addr: %d\n", dp->di_db[j]);
goto unknown;
}
for (j = 0, ndb -= NDADDR; ndb > 0; j++)
ndb /= NINDIR(&sblock);
for (; j < NIADDR; j++)
if (dp->di_ib[j] != 0) {
if (debug)
printf("bad indirect addr: %d\n",
2000-05-23 05:48:52 +04:00
dp->di_ib[j]);
/* goto unknown; */
}
if (ftypeok(dp) == 0)
goto unknown;
n_files++;
lncntp[inumber] = dp->di_nlink;
if (dp->di_nlink <= 0) {
zlnp = (struct zlncnt *)malloc(sizeof *zlnp);
if (zlnp == NULL) {
pfatal("LINK COUNT TABLE OVERFLOW");
if (reply("CONTINUE") == 0)
errexit("%s", "");
} else {
zlnp->zlncnt = inumber;
zlnp->next = zlnhead;
zlnhead = zlnp;
}
}
if (mode == IFDIR) {
if (dp->di_size == 0)
statemap[inumber] = DCLEAR;
else
statemap[inumber] = DSTATE;
cacheino(dp, inumber);
} else
statemap[inumber] = FSTATE;
typemap[inumber] = IFTODT(mode);
2000-05-23 05:48:52 +04:00
#if 0 /* FFS */
if (doinglevel2 &&
2000-05-23 05:48:52 +04:00
(dp->di_ouid != (u_short) - 1 || dp->di_ogid != (u_short) - 1)) {
dp = ginode(inumber);
dp->di_uid = dp->di_ouid;
dp->di_ouid = -1;
dp->di_gid = dp->di_ogid;
dp->di_ogid = -1;
inodirty();
}
#endif
badblk = dupblk = 0;
idesc->id_number = inumber;
(void)ckinode(dp, idesc);
if (dp->di_blocks != idesc->id_entryno) {
pwarn("INCORRECT BLOCK COUNT I=%u (%d should be %d)",
2000-05-23 05:48:52 +04:00
inumber, dp->di_blocks, idesc->id_entryno);
if (preen)
printf(" (CORRECTED)\n");
else if (reply("CORRECT") == 0)
return;
dp = ginode(inumber);
dp->di_blocks = idesc->id_entryno;
inodirty();
}
return;
unknown:
pfatal("UNKNOWN FILE TYPE I=%u", inumber);
statemap[inumber] = FCLEAR;
if (reply("CLEAR") == 1) {
statemap[inumber] = USTATE;
dp = ginode(inumber);
clearinode(dp);
inodirty();
}
}
int
2000-05-23 05:48:52 +04:00
pass1check(struct inodesc * idesc)
{
2000-05-23 05:48:52 +04:00
int res = KEEPON;
int anyout, ndblks;
2000-05-23 05:48:52 +04:00
daddr_t blkno = idesc->id_blkno;
register struct dups *dlp;
2000-05-23 05:48:52 +04:00
struct dups *new;
Merge the short-lived perseant-lfsv2 branch into the trunk. Kernels and tools understand both v1 and v2 filesystems; newfs_lfs generates v2 by default. Changes for the v2 layout include: - Segments of non-PO2 size and arbitrary block offset, so these can be matched to convenient physical characteristics of the partition (e.g., stripe or track size and offset). - Address by fragment instead of by disk sector, paving the way for non-512-byte-sector devices. In theory fragments can be as large as you like, though in reality they must be smaller than MAXBSIZE in size. - Use serial number and filesystem identifier to ensure that roll-forward doesn't get old data and think it's new. Roll-forward is enabled for v2 filesystems, though not for v1 filesystems by default. - The inode free list is now a tailq, paving the way for undelete (undelete is not yet implemented, but can be without further non-backwards-compatible changes to disk structures). - Inode atime information is kept in the Ifile, instead of on the inode; that is, the inode is never written *just* because atime was changed. Because of this the inodes remain near the file data on the disk, rather than wandering all over as the disk is read repeatedly. This speeds up repeated reads by a small but noticeable amount. Other changes of note include: - The ifile written by newfs_lfs can now be of arbitrary length, it is no longer restricted to a single indirect block. - Fixed an old bug where ctime was changed every time a vnode was created. I need to look more closely to make sure that the times are only updated during write(2) and friends, not after-the-fact during a segment write, and certainly not by the cleaner.
2001-07-14 00:30:18 +04:00
if ((anyout = chkrange(blkno, fragstofsb(&sblock, idesc->id_numfrags))) != 0) {
blkerror(idesc->id_number, "BAD", blkno);
if (badblk++ >= MAXBAD) {
pwarn("EXCESSIVE BAD BLKS I=%u",
2000-05-23 05:48:52 +04:00
idesc->id_number);
if (preen)
printf(" (SKIPPING)\n");
else if (reply("CONTINUE") == 0)
errexit("%s", "");
return (STOP);
}
} else if (!testbmap(blkno)) {
Merge the short-lived perseant-lfsv2 branch into the trunk. Kernels and tools understand both v1 and v2 filesystems; newfs_lfs generates v2 by default. Changes for the v2 layout include: - Segments of non-PO2 size and arbitrary block offset, so these can be matched to convenient physical characteristics of the partition (e.g., stripe or track size and offset). - Address by fragment instead of by disk sector, paving the way for non-512-byte-sector devices. In theory fragments can be as large as you like, though in reality they must be smaller than MAXBSIZE in size. - Use serial number and filesystem identifier to ensure that roll-forward doesn't get old data and think it's new. Roll-forward is enabled for v2 filesystems, though not for v1 filesystems by default. - The inode free list is now a tailq, paving the way for undelete (undelete is not yet implemented, but can be without further non-backwards-compatible changes to disk structures). - Inode atime information is kept in the Ifile, instead of on the inode; that is, the inode is never written *just* because atime was changed. Because of this the inodes remain near the file data on the disk, rather than wandering all over as the disk is read repeatedly. This speeds up repeated reads by a small but noticeable amount. Other changes of note include: - The ifile written by newfs_lfs can now be of arbitrary length, it is no longer restricted to a single indirect block. - Fixed an old bug where ctime was changed every time a vnode was created. I need to look more closely to make sure that the times are only updated during write(2) and friends, not after-the-fact during a segment write, and certainly not by the cleaner.
2001-07-14 00:30:18 +04:00
seg_table[dtosn(&sblock, blkno)].su_nbytes += idesc->id_numfrags * sblock.lfs_fsize;
}
Merge the short-lived perseant-lfsv2 branch into the trunk. Kernels and tools understand both v1 and v2 filesystems; newfs_lfs generates v2 by default. Changes for the v2 layout include: - Segments of non-PO2 size and arbitrary block offset, so these can be matched to convenient physical characteristics of the partition (e.g., stripe or track size and offset). - Address by fragment instead of by disk sector, paving the way for non-512-byte-sector devices. In theory fragments can be as large as you like, though in reality they must be smaller than MAXBSIZE in size. - Use serial number and filesystem identifier to ensure that roll-forward doesn't get old data and think it's new. Roll-forward is enabled for v2 filesystems, though not for v1 filesystems by default. - The inode free list is now a tailq, paving the way for undelete (undelete is not yet implemented, but can be without further non-backwards-compatible changes to disk structures). - Inode atime information is kept in the Ifile, instead of on the inode; that is, the inode is never written *just* because atime was changed. Because of this the inodes remain near the file data on the disk, rather than wandering all over as the disk is read repeatedly. This speeds up repeated reads by a small but noticeable amount. Other changes of note include: - The ifile written by newfs_lfs can now be of arbitrary length, it is no longer restricted to a single indirect block. - Fixed an old bug where ctime was changed every time a vnode was created. I need to look more closely to make sure that the times are only updated during write(2) and friends, not after-the-fact during a segment write, and certainly not by the cleaner.
2001-07-14 00:30:18 +04:00
for (ndblks = fragstofsb(&sblock, idesc->id_numfrags); ndblks > 0; blkno++, ndblks--) {
if (anyout && chkrange(blkno, 1)) {
res = SKIP;
} else if (!testbmap(blkno)) {
n_blks++;
#ifndef VERBOSE_BLOCKMAP
setbmap(blkno);
#else
2000-05-23 05:48:52 +04:00
setbmap(blkno, idesc->id_number);
#endif
} else {
blkerror(idesc->id_number, "DUP", blkno);
#ifdef VERBOSE_BLOCKMAP
2000-05-23 05:48:52 +04:00
pwarn("(lbn %d: Holder is %d)\n", idesc->id_lblkno,
testbmap(blkno));
#endif
if (dupblk++ >= MAXDUP) {
pwarn("EXCESSIVE DUP BLKS I=%u",
2000-05-23 05:48:52 +04:00
idesc->id_number);
if (preen)
printf(" (SKIPPING)\n");
else if (reply("CONTINUE") == 0)
errexit("%s", "");
return (STOP);
}
new = (struct dups *)malloc(sizeof(struct dups));
if (new == NULL) {
pfatal("DUP TABLE OVERFLOW.");
if (reply("CONTINUE") == 0)
errexit("%s", "");
return (STOP);
}
new->dup = blkno;
if (muldup == 0) {
duplist = muldup = new;
new->next = 0;
} else {
new->next = muldup->next;
muldup->next = new;
}
for (dlp = duplist; dlp != muldup; dlp = dlp->next)
if (dlp->dup == blkno)
break;
if (dlp == muldup && dlp->dup != blkno)
muldup = new;
}
/*
* count the number of blocks found in id_entryno
*/
2000-05-23 05:48:52 +04:00
idesc->id_entryno++;
}
return (res);
}