hfsutil_jnl.c   [plain text]

/*
 * Copyright (c) 1999-2001 Apple Computer, Inc. All rights reserved.
 *
 * @APPLE_LICENSE_HEADER_START@
 * 
 * The contents of this file constitute Original Code as defined in and
 * are subject to the Apple Public Source License Version 1.2 (the
 * "License").  You may not use this file except in compliance with the
 * License.  Please obtain a copy of the License at
 * http://www.apple.com/publicsource and read it before using this file.
 * 
 * This Original Code and all software distributed under the License are
 * distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER
 * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
 * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT.  Please see the
 * License for the specific language governing rights and limitations
 * under the License.
 * 
 * @APPLE_LICENSE_HEADER_END@
 */
/*
 Copyright (c) 2002 Apple Computer, Inc.
 All Rights Reserved.

 This file contains the routine to make an HFS+ volume journaled
 and a corresponding routine to turn it off.
 
 */

#include <sys/types.h>
#include <sys/attr.h>
#include <sys/stat.h>
#include <sys/time.h>
#include <sys/sysctl.h>
#include <sys/resource.h>
#include <sys/vmmeter.h>
#include <sys/mount.h>
#include <sys/wait.h>
#include <sys/ioctl.h>

#include <sys/disk.h>
#include <sys/loadable_fs.h>
#include <hfs/hfs_format.h>
#include <hfs/hfs_mount.h>    /* for hfs sysctl values */

#include <errno.h>
#include <fcntl.h>
#include <libgen.h>
#include <pwd.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>

#include <architecture/byte_order.h>

// just in case these aren't in <hfs/hfs_mount.h> yet
#ifndef HFS_ENABLE_JOURNALING
#define HFS_ENABLE_JOURNALING   0x082969
#endif
#ifndef HFS_DISABLE_JOURNALING
#define HFS_DISABLE_JOURNALING 0x031272
#endif
#ifndef HFS_GET_JOURNAL_INFO
#define HFS_GET_JOURNAL_INFO    0x6a6e6c69
#endif

/* getattrlist buffers start with an extra length field */
struct ExtentsAttrBuf {
	unsigned long	infoLength;
	HFSPlusExtentRecord	extents;
};
typedef struct ExtentsAttrBuf ExtentsAttrBuf;



#define kIsInvisible 0x4000

/*
 * Generic Finder file/dir data
 */
struct FinderInfo {
	u_int32_t 	opaque_1[2];
	u_int16_t 	fdFlags;	/* Finder flags */
	int16_t 	opaque_2[11];
};
typedef struct FinderInfo FinderInfo;

/* getattrlist buffers start with an extra length field */
struct FinderAttrBuf {
	unsigned long	infoLength;
	FinderInfo	finderInfo;
};
typedef struct FinderAttrBuf FinderAttrBuf;


int hide_file(const char * file)
{
    struct attrlist alist = {0};
    FinderAttrBuf finderInfoBuf = {0};
    int result;
    
    alist.bitmapcount = ATTR_BIT_MAP_COUNT;
    alist.commonattr = ATTR_CMN_FNDRINFO;

    result = getattrlist(file, &alist, &finderInfoBuf, sizeof(finderInfoBuf), 0);
    if (result) {
	return (errno);
    }
	
    if (finderInfoBuf.finderInfo.fdFlags & kIsInvisible) {
	printf("hide: %s is alreadly invisible\n", file);
	return (0);
    }

    finderInfoBuf.finderInfo.fdFlags |= kIsInvisible;

    result = setattrlist(file, &alist, &finderInfoBuf.finderInfo, sizeof(FinderInfo), 0);
    
    return (result == -1 ? errno : result);
}

off_t
get_start_block(const char *file, uint32_t fs_block_size)
{
    off_t cur_pos, phys_start, len;
    int fd, err;
    struct log2phys l2p;
    struct stat st;

    fd = open(file, O_RDONLY);
    if (fd < 0) {
	return -1;
    }

    if (fstat(fd, &st) < 0) {
	fprintf(stderr, "can't stat %s (%s)\n", file, strerror(errno));
	close(fd);
	return -1;
    }

    fs_block_size = st.st_blksize; // XXXdbg quick hack for now

    phys_start = len = 0;
    for(cur_pos=0; cur_pos < st.st_size; cur_pos += fs_block_size) {
	memset(&l2p, 0, sizeof(l2p));
	lseek(fd, cur_pos, SEEK_SET);
	err = fcntl(fd, F_LOG2PHYS, &l2p);

	if (phys_start == 0) {
	    phys_start = l2p.l2p_devoffset;
	    len = fs_block_size;
	} else if (l2p.l2p_devoffset != (phys_start + len)) {
	    // printf("    %lld : %lld - %lld\n", cur_pos, phys_start / fs_block_size, len / fs_block_size);
	    fprintf(stderr, "%s : is not contiguous!\n", file);
	    close(fd);
	    return -1;
	    // phys_start = l2p.l2p_devoffset;
	    // len = fs_block_size;
	} else {
	    len += fs_block_size;
	}
    }

    close(fd);

    //printf("%s start offset %lld; byte len %lld (blksize %d)\n",
    // file, phys_start, len, fs_block_size);

    if ((phys_start / (unsigned)fs_block_size) & 0xffffffff00000000LL) {
	fprintf(stderr, "%s : starting block is > 32bits!\n", file);
	return -1;
    }
	
    return phys_start;
}


//
// Get the embedded offset (if any) for an hfs+ volume.
// This is pretty skanky that we have to do this but
// that's life...
//
#include <sys/disk.h>
#include <hfs/hfs_format.h>

#include <machine/endian.h>

#define HFS_PRI_SECTOR(blksize)          (1024 / (blksize))
#define HFS_PRI_OFFSET(blksize)          ((blksize) > 1024 ? 1024 : 0)

#define SWAP_BE16(x) ntohs(x)
#define SWAP_BE32(x) ntohl(x)


off_t
get_embedded_offset(char *devname)
{
    int fd = -1;
    off_t ret = 0;
    char *buff = NULL, rawdev[256];
    u_int64_t blkcnt;
    u_int32_t blksize;
    HFSMasterDirectoryBlock *mdbp;
    off_t embeddedOffset;
    struct statfs sfs;
    struct stat   st;
	
  restart:
    if (stat(devname, &st) != 0) {
	fprintf(stderr, "Could not access %s (%s)\n", devname, strerror(errno));
	ret = -1;
	goto out;
    }

    if (S_ISCHR(st.st_mode) == 0) {
	// hmmm, it's not the character special raw device so we
	// should try to figure out the real device.
	if (statfs(devname, &sfs) != 0) {
	    fprintf(stderr, "Can't find out any info about the fs for path %s (%s)\n",
		devname, strerror(errno));
	    ret = -1;
	    goto out;
	}

	// copy the "/dev/"
	strncpy(rawdev, sfs.f_mntfromname, 5);
	rawdev[5] = 'r';
	strcpy(&rawdev[6], &sfs.f_mntfromname[5]);
	devname = &rawdev[0];
	goto restart;
    }

    fd = open(devname, O_RDONLY);
    if (fd < 0) {
	fprintf(stderr, "can't open: %s (%s)\n", devname, strerror(errno));
	ret = -1;
	goto out;
    }

    /* Get the real physical block size. */
    if (ioctl(fd, DKIOCGETBLOCKSIZE, (caddr_t)&blksize) != 0) {
	fprintf(stderr, "can't get the device block size (%s). assuming 512\n", strerror(errno));
	blksize = 512;
	ret = -1;
	goto out;
    }

    /* Get the number of physical blocks. */
    if (ioctl(fd, DKIOCGETBLOCKCOUNT, (caddr_t)&blkcnt)) {
	struct stat st;
	fprintf(stderr, "failed to get block count. trying stat().\n");
	if (fstat(fd, &st) != 0) {
	    ret = -1;
	    goto out;
	}

	blkcnt = st.st_size / blksize;
    }

    /*
     * At this point:
     *   blksize has our prefered physical block size
     *   blkcnt has the total number of physical blocks
     */

    buff = (char *)malloc(blksize);
	
    if (pread(fd, buff, blksize, HFS_PRI_SECTOR(blksize)*blksize) != blksize) {
	fprintf(stderr, "failed to read volume header @ offset %d (%s)\n",
	    HFS_PRI_SECTOR(blksize), strerror(errno));
	ret = -1;
	goto out;
    }

    mdbp = (HFSMasterDirectoryBlock *)buff;
    if (   (SWAP_BE16(mdbp->drSigWord) != kHFSSigWord) 
        && (SWAP_BE16(mdbp->drSigWord) != kHFSPlusSigWord)
        && (SWAP_BE16(mdbp->drSigWord) != kHFSXSigWord)) {
	ret = -1;
	goto out;
    }

    if ((SWAP_BE16(mdbp->drSigWord) == kHFSSigWord) && (SWAP_BE16(mdbp->drEmbedSigWord) != kHFSPlusSigWord)) {
	ret = -1;
	goto out;
    } else if (SWAP_BE16(mdbp->drEmbedSigWord) == kHFSPlusSigWord) {
	/* Get the embedded Volume Header */
	embeddedOffset = SWAP_BE16(mdbp->drAlBlSt) * 512;
	embeddedOffset += (u_int64_t)SWAP_BE16(mdbp->drEmbedExtent.startBlock) *
                          (u_int64_t)SWAP_BE32(mdbp->drAlBlkSiz);

	/*
	 * If the embedded volume doesn't start on a block
	 * boundary, then switch the device to a 512-byte
	 * block size so everything will line up on a block
	 * boundary.
	 */
	if ((embeddedOffset % blksize) != 0) {
	    fprintf(stderr, "HFS Mount: embedded volume offset not"
		" a multiple of physical block size (%d);"
		" switching to 512\n", blksize);
		
	    blkcnt  *= (blksize / 512);
	    blksize  = 512;
	}

    } else { /* pure HFS+ */ 
	embeddedOffset = 0;
    }

    ret = embeddedOffset;

  out:
    if (buff) {
	free(buff);
    }
    if (fd >= 0) 
	close(fd);

    return ret;
}



static const char *journal_fname = ".journal";
static const char *jib_fname = ".journal_info_block";

int
DoMakeJournaled(char *volname, int jsize)
{
    int              fd, i, block_size, journal_size = 8*1024*1024;
    char            *buf;
    int              ret;
    fstore_t         fst;
    int32_t          jstart_block, jinfo_block;
    int              sysctl_info[8];
    JournalInfoBlock jib;
    struct statfs    sfs;
    static char      tmpname[MAXPATHLEN];
    off_t            start_block, embedded_offset;

    if (statfs(volname, &sfs) != 0) {
	fprintf(stderr, "Can't stat volume %s (%s).\n", volname, strerror(errno));
	return 10;
    }

    // Make sure that we're HFS+.  First we check the fstypename.
    // If that's ok then we try to create a symlink (which won't
    // work on plain hfs volumes but will work on hfs+ volumes).
    //
    sprintf(tmpname, "%s/is_vol_hfs_plus", volname);
    if (strcmp(sfs.f_fstypename, "hfs") != 0 ||
	((ret = symlink(tmpname, tmpname)) != 0 && errno == ENOTSUP)) {
	fprintf(stderr, "%s is not an HFS+ volume.  Journaling only works on HFS+ volumes.\n",
		volname);
	return 10;
    }
    unlink(tmpname);

    if (sfs.f_flags & MNT_JOURNALED) {
	fprintf(stderr, "Volume %s is already journaled.\n", volname);
	return 1;
    }

    if (jsize != 0) {
	journal_size = jsize;
    } else {
	int scale;

	//
	// we want at least 8 megs of journal for each 100 gigs of
	// disk space.  We cap the size at 512 megs though.
	//
	scale = ((long long)sfs.f_bsize * (long long)((unsigned)sfs.f_blocks)) / (100*1024*1024*1024ULL);
	journal_size *= (scale + 1);
	if (journal_size > 512 * 1024 * 1024) {
	    journal_size = 512 * 1024 * 1024;
	}
    }

    if (chdir(volname) != 0) {
	fprintf(stderr, "Can't locate volume %s to make it journaled (%s).\n",
		volname, strerror(errno));
	return 10;
    }


    embedded_offset = get_embedded_offset(volname);
    if (embedded_offset < 0) {
	fprintf(stderr, "Can't calculate the embedded offset (if any) for %s.\n", volname);
	fprintf(stderr, "Journal creation failure.\n");
	return 15;
    }
    // printf("Embedded offset == 0x%llx\n", embedded_offset);

    fd = open(journal_fname, O_CREAT|O_TRUNC|O_RDWR, 000);
    if (fd < 0) {
	fprintf(stderr, "Can't create journal file on volume %s (%s)\n",
		volname, strerror(errno));
	return 5;
    }

    // make sure that it has no r/w/x privs (only could happen if
    // the file already existed since open() doesn't reset the mode
    // bits).
    //
    fchmod(fd, 0);

    block_size = sfs.f_bsize;
    if ((journal_size % block_size) != 0) {
	fprintf(stderr, "Journal size %dk is not a multiple of volume %s block size (%d).\n",
		journal_size/1024, volname, block_size);
	close(fd);
	unlink(journal_fname);
	return 5;
    }

  retry:
    memset(&fst, 0, sizeof(fst));
    fst.fst_flags   = F_ALLOCATECONTIG|F_ALLOCATEALL;
    fst.fst_length  = journal_size;
    fst.fst_posmode = F_PEOFPOSMODE;
    
    ret = fcntl(fd, F_PREALLOCATE, &fst);
    if (ret < 0) {
	if (journal_size >= 2*1024*1024) {
	    fprintf(stderr, "Not enough contiguous space for a %d k journal.  Retrying.\n",
		    journal_size/1024);
	    journal_size /= 2;
	    ftruncate(fd, 0);     // make sure the file is zero bytes long.
	    goto retry;
	} else {
	    fprintf(stderr, "Disk too fragmented to enable journaling.\n");
	    fprintf(stderr, "Please run a defragmenter on %s.\n", volname);
	    close(fd);
	    unlink(journal_fname);
	    return 10;
	}
    }

    printf("Allocated %lldK for journal file.\n", fst.fst_bytesalloc/1024LL);
    buf = (char *)calloc(block_size, 1);
    if (buf) {
	for(i=0; i < journal_size/block_size; i++) {
	    ret = write(fd, buf, block_size);
	    if (ret != block_size) {
		break;
	    }
	}
		
	if (i*block_size != journal_size) {
	    fprintf(stderr, "Failed to write %dk to journal on volume %s (%s)\n",
		    journal_size/1024, volname, strerror(errno));
	}
    } else {
	printf("Could not allocate memory to write to the journal on volume %s (%s)\n",
	       volname, strerror(errno));
    }

    fsync(fd);
    close(fd);
    hide_file(journal_fname);

    start_block = get_start_block(journal_fname, block_size);
    if (start_block == (off_t)-1) {
	fprintf(stderr, "Failed to get start block for %s (%s)\n",
		journal_fname, strerror(errno));
	unlink(journal_fname);
	return 20;
    }
    jstart_block = (start_block / block_size) - (embedded_offset / block_size);

    memset(&jib, 'Z', sizeof(jib));
    jib.flags  = kJIJournalInFSMask;
    jib.offset = (off_t)((unsigned)jstart_block) * (off_t)((unsigned)block_size);
    jib.size   = (off_t)((unsigned)journal_size);

    fd = open(jib_fname, O_CREAT|O_TRUNC|O_RDWR, 000);
    if (fd < 0) {
	fprintf(stderr, "Could not create journal info block file on volume %s (%s)\n",
		volname, strerror(errno));
	unlink(journal_fname);
	return 5;
    }
    
    // swap the data before we copy it
    jib.flags  = NXSwapBigLongToHost(jib.flags);
    jib.offset = NXSwapBigLongLongToHost(jib.offset);
    jib.size   = NXSwapBigLongLongToHost(jib.size);
    
    memcpy(buf, &jib, sizeof(jib));

    // now put it back the way it was
    jib.size   = NXSwapBigLongLongToHost(jib.size);
    jib.offset = NXSwapBigLongLongToHost(jib.offset);
    jib.flags  = NXSwapBigLongToHost(jib.flags);

    if (write(fd, buf, block_size) != block_size) {
	fprintf(stderr, "Failed to write journal info block on volume %s (%s)!\n",
		volname, strerror(errno));
	unlink(journal_fname);
	return 10;
    }

    fsync(fd);
    close(fd);
    hide_file(jib_fname);

    start_block = get_start_block(jib_fname, block_size);
    if (start_block == (off_t)-1) {
	fprintf(stderr, "Failed to get start block for %s (%s)\n",
		jib_fname, strerror(errno));
	unlink(journal_fname);
	unlink(jib_fname);
	return 20;
    }
    jinfo_block = (start_block / block_size) - (embedded_offset / block_size);


    //
    // Now make the volume journaled!
    //
    memset(sysctl_info, 0, sizeof(sysctl_info));
    sysctl_info[0] = CTL_VFS;
    sysctl_info[1] = sfs.f_fsid.val[1];
    sysctl_info[2] = HFS_ENABLE_JOURNALING;
    sysctl_info[3] = jinfo_block;
    sysctl_info[4] = jstart_block;
    sysctl_info[5] = journal_size;
     
    //printf("fs type: 0x%x\n", sysctl_info[1]);
    //printf("jinfo block : 0x%x\n", jinfo_block);
    //printf("jstart block: 0x%x\n", jstart_block);
    //printf("journal size: 0x%x\n", journal_size);

    ret = sysctl((void *)sysctl_info, 6, NULL, NULL, NULL, 0);
    if (ret != 0) {
	fprintf(stderr, "Failed to make volume %s journaled (%s)\n",
		volname, strerror(errno));
	unlink(journal_fname);
	unlink(jib_fname);
	return 20;
    }
    
    return 0;
}


int
DoUnJournal(char *volname)
{
    int           result;
    int           sysctl_info[8];
    struct statfs sfs;
    char          jbuf[MAXPATHLEN];
	
    if (statfs(volname, &sfs) != 0) {
	fprintf(stderr, "Can't stat volume %s (%s).\n", volname, strerror(errno));
	return 10;
    }

    if ((sfs.f_flags & MNT_JOURNALED) == 0) {
	fprintf(stderr, "Volume %s is not journaled.\n", volname);
	return 1;
    }

    if (chdir(volname) != 0) {
	fprintf(stderr, "Can't locate volume %s to turn off journaling (%s).\n",
		volname, strerror(errno));
	return 10;
    }
	
    memset(sysctl_info, 0, sizeof(sysctl_info));
    sysctl_info[0] = CTL_VFS;
    sysctl_info[1] = sfs.f_fsid.val[1];
    sysctl_info[2] = HFS_DISABLE_JOURNALING;
	
    result = sysctl((void *)sysctl_info, 3, NULL, NULL, NULL, 0);
    if (result != 0) {
	fprintf(stderr, "Failed to make volume %s UN-journaled (%s)\n",
		volname, strerror(errno));
	return 20;
    }

    sprintf(jbuf, "%s/%s", volname, journal_fname);
    if (unlink(jbuf) != 0) {
	fprintf(stderr, "Failed to remove the journal %s (%s)\n",
		jbuf, strerror(errno));
    }

    sprintf(jbuf, "%s/%s", volname, jib_fname);
    if (unlink(jbuf) != 0) {
	fprintf(stderr, "Failed to remove the journal info block %s (%s)\n",
		jbuf, strerror(errno));
    }

    printf("Journaling disabled on %s\n", volname);
	    
    return 0;
}


int
DoGetJournalInfo(char *volname)
{
    int           result;
    int           sysctl_info[8];
    struct statfs sfs;
    off_t         jstart, jsize;
	
    if (statfs(volname, &sfs) != 0) {
	fprintf(stderr, "Can't stat volume %s (%s).\n", volname, strerror(errno));
	return 10;
    }

    if ((sfs.f_flags & MNT_JOURNALED) == 0) {
	fprintf(stderr, "Volume %s is not journaled.\n", volname);
	return 1;
    }

    if (chdir(volname) != 0) {
	fprintf(stderr, "Can't cd to volume %s to get journal info (%s).\n",
		volname, strerror(errno));
	return 10;
    }
	
    memset(sysctl_info, 0, sizeof(sysctl_info));
    sysctl_info[0] = CTL_VFS;
    sysctl_info[1] = sfs.f_fsid.val[1];
    sysctl_info[2] = HFS_GET_JOURNAL_INFO;
    sysctl_info[3] = (int)&jstart;
    sysctl_info[4] = (int)&jsize;
	
    result = sysctl((void *)sysctl_info, 5, NULL, NULL, NULL, 0);
    if (result != 0) {
	fprintf(stderr, "Failed to get journal info for volume %s (%s)\n",
		volname, strerror(errno));
	return 20;
    }

    if (jsize == 0) {
	printf("%s : not journaled.\n", volname);
    } else {
	printf("%s : journal size %lld k at offset 0x%llx\n", volname, jsize/1024, jstart);
    }

    return 0;
}