stage2/fsys_reiserfs.c

/* fsys_reiserfs.c - an implementation for the ReiserFS filesystem */
/*
 *  GRUB  --  GRand Unified Bootloader
 *  Copyright (C) 2000, 2001  Free Software Foundation, Inc.
 *
 *  This program is free software; you can redistribute it and/or modify
 *  it under the terms of the GNU General Public License as published by
 *  the Free Software Foundation; either version 2 of the License, or
 *  (at your option) any later version.
 *
 *  This program is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *  GNU General Public License for more details.
 *
 *  You should have received a copy of the GNU General Public License
 *  along with this program; if not, write to the Free Software
 *  Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */

#ifdef FSYS_REISERFS
#include "shared.h"
#include "filesys.h"

#undef REISERDEBUG

/* Some parts of this code (mainly the structures and defines) are
 * from the original reiser fs code, as found in the linux kernel.
 */

/* include/asm-i386/types.h */
typedef __signed__ char __s8;
typedef unsigned char __u8;
typedef __signed__ short __s16;
typedef unsigned short __u16;
typedef __signed__ int __s32;
typedef unsigned int __u32;
typedef unsigned long long __u64;

/* linux/posix_type.h */
typedef long linux_off_t;

/* linux/little_endian.h */
#define __cpu_to_le64(x) ((__u64) (x))
#define __le64_to_cpu(x) ((__u64) (x))
#define __cpu_to_le32(x) ((__u32) (x))
#define __le32_to_cpu(x) ((__u32) (x))
#define __cpu_to_le16(x) ((__u16) (x))
#define __le16_to_cpu(x) ((__u16) (x))

/* include/linux/reiser_fs.h */
/* This is the new super block of a journaling reiserfs system */
struct reiserfs_super_block
{
  __u32 s_block_count;			/* blocks count         */
  __u32 s_free_blocks;                  /* free blocks count    */
  __u32 s_root_block;           	/* root block number    */
  __u32 s_journal_block;           	/* journal block number    */
  __u32 s_journal_dev;           	/* journal device number  */
  __u32 s_orig_journal_size; 		/* size of the journal on FS creation.  used to make sure they don't overflow it */
  __u32 s_journal_trans_max ;           /* max number of blocks in a transaction.  */
  __u32 s_journal_block_count ;         /* total size of the journal. can change over time  */
  __u32 s_journal_max_batch ;           /* max number of blocks to batch into a trans */
  __u32 s_journal_max_commit_age ;      /* in seconds, how old can an async commit be */
  __u32 s_journal_max_trans_age ;       /* in seconds, how old can a transaction be */
  __u16 s_blocksize;                   	/* block size           */
  __u16 s_oid_maxsize;			/* max size of object id array  */
  __u16 s_oid_cursize;			/* current size of object id array */
  __u16 s_state;                       	/* valid or error       */
  char s_magic[16];                     /* reiserfs magic string indicates that file system is reiserfs */
  __u16 s_tree_height;                  /* height of disk tree */
  __u16 s_bmap_nr;                      /* amount of bitmap blocks needed to address each block of file system */
  __u16 s_version;
  char s_unused[128];			/* zero filled by mkreiserfs */
};

#define REISERFS_MAX_SUPPORTED_VERSION 2
#define REISERFS_SUPER_MAGIC_STRING "ReIsErFs"
#define REISER2FS_SUPER_MAGIC_STRING "ReIsEr2Fs"

#define MAX_HEIGHT 7

/* must be correct to keep the desc and commit structs at 4k */
#define JOURNAL_TRANS_HALF 1018

/* first block written in a commit.  */
struct reiserfs_journal_desc {
  __u32 j_trans_id;			/* id of commit */
  __u32 j_len;				/* length of commit. len +1 is the commit block */
  __u32 j_mount_id;			/* mount id of this trans*/
  __u32 j_realblock[JOURNAL_TRANS_HALF]; /* real locations for the first blocks */
  char j_magic[12];
};

/* last block written in a commit */
struct reiserfs_journal_commit {
  __u32 j_trans_id;			/* must match j_trans_id from the desc block */
  __u32 j_len;			/* ditto */
  __u32 j_realblock[JOURNAL_TRANS_HALF]; /* real locations for the last blocks */
  char j_digest[16];			/* md5 sum of all the blocks involved, including desc and commit. not used, kill it */
};

/* this header block gets written whenever a transaction is considered
   fully flushed, and is more recent than the last fully flushed
   transaction.  
   fully flushed means all the log blocks and all the real blocks are
   on disk, and this transaction does not need to be replayed.  
*/
struct reiserfs_journal_header {
  /* id of last fully flushed transaction */
  __u32 j_last_flush_trans_id;
  /* offset in the log of where to start replay after a crash */
  __u32 j_first_unflushed_offset;
  /* mount id to detect very old transactions */
  __u32 long j_mount_id;
};

/* magic string to find desc blocks in the journal */
#define JOURNAL_DESC_MAGIC "ReIsErLB" 


/*
 * directories use this key as well as old files
 */
struct offset_v1
{
  /*
   * for regular files this is the offset to the first byte of the
   * body, contained in the object-item, as measured from the start of
   * the entire body of the object.
   *
   * for directory entries, k_offset consists of hash derived from
   * hashing the name and using few bits (23 or more) of the resulting
   * hash, and generation number that allows distinguishing names with
   * hash collisions. If number of collisions overflows generation
   * number, we return EEXIST.  High order bit is 0 always 
   */
  __u32 k_offset;
  __u32 k_uniqueness;
};

struct offset_v2
{
  /*
   * for regular files this is the offset to the first byte of the
   * body, contained in the object-item, as measured from the start of
   * the entire body of the object.
   *
   * for directory entries, k_offset consists of hash derived from
   * hashing the name and using few bits (23 or more) of the resulting
   * hash, and generation number that allows distinguishing names with
   * hash collisions. If number of collisions overflows generation
   * number, we return EEXIST.  High order bit is 0 always 
   */
  __u64 k_offset:60;
  __u64 k_type: 4;
};


struct key
{
  /* packing locality: by default parent directory object id */
  __u32 k_dir_id;
  /* object identifier */
  __u32 k_objectid;
  /* the offset and node type (old and new form) */
  union
  {
    struct offset_v1 v1;
    struct offset_v2 v2;
  }
  u;
};

#define KEY_SIZE (sizeof (struct key))

/* Header of a disk block.  More precisely, header of a formatted leaf
   or internal node, and not the header of an unformatted node. */
struct block_head
{       
  __u16 blk_level;        /* Level of a block in the tree. */
  __u16 blk_nr_item;      /* Number of keys/items in a block. */
  __u16 blk_free_space;   /* Block free space in bytes. */
  struct key  blk_right_delim_key; /* Right delimiting key for this block (supported for leaf level nodes
				      only) */
};
#define BLKH_SIZE (sizeof (struct block_head))
#define DISK_LEAF_NODE_LEVEL  1 /* Leaf node level.                       */

struct item_head
{
  struct key ih_key; 	/* Everything in the tree is found by searching for it based on its key.*/
  
  union
  {
    __u16 ih_free_space; /* The free space in the last unformatted node of an indirect item if this
			    is an indirect item.  This equals 0xFFFF iff this is a direct item or
			    stat data item. Note that the key, not this field, is used to determine
			    the item type, and thus which field this union contains. */
    __u16 ih_entry_count; /* Iff this is a directory item, this field equals the number of directory
			     entries in the directory item. */
  }
  u;
  __u16 ih_item_len;           /* total size of the item body                  */
  __u16 ih_item_location;      /* an offset to the item body within the block  */
  __u16 ih_version;	       /* ITEM_VERSION_1 for all old items, 
				  ITEM_VERSION_2 for new ones. 
				  Highest bit is set by fsck
                                  temporary, cleaned after all done */
};
/* size of item header     */
#define IH_SIZE (sizeof (struct item_head))

#define ITEM_VERSION_1 0
#define ITEM_VERSION_2 1
#define IH_KEY_OFFSET(ih) (INFO->version < 2 \
			   || (ih)->ih_version == ITEM_VERSION_1 \
			   ? (ih)->ih_key.u.v1.k_offset \
			   : (ih)->ih_key.u.v2.k_offset)

#define IH_KEY_ISTYPE(ih, type) (INFO->version < 2 \
				 || (ih)->ih_version == ITEM_VERSION_1 \
				 ? (ih)->ih_key.u.v1.k_uniqueness == V1_##type \
				 : (ih)->ih_key.u.v2.k_type == V2_##type)

struct disk_child
{
  unsigned long       dc_block_number;              /* Disk child's block number. */
  unsigned short      dc_size;		            /* Disk child's used space.   */
};

#define DC_SIZE (sizeof (struct disk_child))

/* Stat Data on disk.
 *
 * Note that reiserfs has two different forms of stat data.  Luckily
 * the fields needed by grub are at the same position.
 */
struct stat_data
{
  __u16 sd_mode;	/* file type, permissions */
  __u16 sd_notused1[3]; /* fields not needed by reiserfs */
  __u32 sd_size;	/* file size */
  __u32 sd_size_hi;	/* file size high 32 bits (since version 2) */
};

struct reiserfs_de_head
{
  __u32 deh_offset;  /* third component of the directory entry key */
  __u32 deh_dir_id;  /* objectid of the parent directory of the
			object, that is referenced by directory entry */
  __u32 deh_objectid;/* objectid of the object, that is referenced by
                        directory entry */
  __u16 deh_location;/* offset of name in the whole item */
  __u16 deh_state;   /* whether 1) entry contains stat data (for
			future), and 2) whether entry is hidden
			(unlinked) */
};

#define DEH_SIZE (sizeof (struct reiserfs_de_head))

#define DEH_Statdata (1 << 0)			/* not used now */
#define DEH_Visible  (1 << 2)

#define SD_OFFSET  0
#define SD_UNIQUENESS 0
#define DOT_OFFSET 1
#define DOT_DOT_OFFSET 2
#define DIRENTRY_UNIQUENESS 500

#define V1_TYPE_STAT_DATA 0x0
#define V1_TYPE_DIRECT 0xffffffff
#define V1_TYPE_INDIRECT 0xfffffffe
#define V1_TYPE_DIRECTORY_MAX 0xfffffffd
#define V2_TYPE_STAT_DATA 0
#define V2_TYPE_INDIRECT 1
#define V2_TYPE_DIRECT 2
#define V2_TYPE_DIRENTRY 3 

#define REISERFS_ROOT_OBJECTID 2
#define REISERFS_ROOT_PARENT_OBJECTID 1
#define REISERFS_DISK_OFFSET_IN_BYTES (64 * 1024)
/* the spot for the super in versions 3.5 - 3.5.11 (inclusive) */
#define REISERFS_OLD_DISK_OFFSET_IN_BYTES (8 * 1024)
#define REISERFS_OLD_BLOCKSIZE 4096

#define S_ISREG(mode) (((mode) & 0170000) == 0100000)
#define S_ISDIR(mode) (((mode) & 0170000) == 0040000)
#define S_ISLNK(mode) (((mode) & 0170000) == 0120000)

#define PATH_MAX       1024	/* include/linux/limits.h */
#define MAX_LINK_COUNT    5	/* number of symbolic links to follow */

/* The size of the node cache */
#define FSYSREISER_CACHE_SIZE 24*1024
#define FSYSREISER_MIN_BLOCKSIZE SECTOR_SIZE
#define FSYSREISER_MAX_BLOCKSIZE FSYSREISER_CACHE_SIZE / 3

struct fsys_reiser_fileinfo
{
  __u32 k_dir_id;
  __u32 k_objectid;
};

struct fsys_reiser_info
{
  /* The last read item head */
  struct item_head *current_ih;
  /* The last read item */
  char *current_item;
  /* The information for the currently open file */
  struct fsys_reiser_fileinfo fileinfo;
  /* The start of the journal */
  __u32 journal_block;
  /* The size of the journal */
  __u32 journal_block_count;
  /* The first valid descriptor block in journal
     (relative to journal_block) */
  __u32 journal_first_desc;

  /* The ReiserFS version. */
  __u16 version;
  /* The current depth of the reiser tree. */
  __u16 tree_depth;
  /* SECTOR_SIZE << blocksize_shift == blocksize. */
  __u8  blocksize_shift;
  /* 1 << full_blocksize_shift == blocksize. */
  __u8  fullblocksize_shift;
  /* The reiserfs block size  (must be a power of 2) */
  __u16 blocksize;
  /* The number of cached tree nodes */
  __u16 cached_slots;
  /* The number of valid transactions in journal */
  __u16 journal_transactions;
  
  unsigned int blocks[MAX_HEIGHT];
  unsigned int next_key_nr[MAX_HEIGHT];
};

/* The cached s+tree blocks in FSYS_BUF,  see below
 * for a more detailed description.
 */
#define ROOT     ((char *) ((int) FSYS_BUF))
#define CACHE(i) (ROOT + ((i) << INFO->fullblocksize_shift))
#define LEAF     CACHE (DISK_LEAF_NODE_LEVEL)

#define BLOCKHEAD(cache) ((struct block_head *) cache)
#define ITEMHEAD         ((struct item_head  *) ((int) LEAF + BLKH_SIZE))
#define KEY(cache)       ((struct key        *) ((int) cache + BLKH_SIZE))
#define DC(cache)        ((struct disk_child *) \
			  ((int) cache + BLKH_SIZE + KEY_SIZE * nr_item))
/* The fsys_reiser_info block.
 */
#define INFO \
    ((struct fsys_reiser_info *) ((int) FSYS_BUF + FSYSREISER_CACHE_SIZE))
/* 
 * The journal cache.  For each transaction it contains the number of
 * blocks followed by the real block numbers of this transaction.  
 *
 * If the block numbers of some transaction won't fit in this space,
 * this list is stopped with a 0xffffffff marker and the remaining
 * uncommitted transactions aren't cached.  
 */
#define JOURNAL_START    ((__u32 *) (INFO + 1))
#define JOURNAL_END      ((__u32 *) (FSYS_BUF + FSYS_BUFLEN))


static __inline__ unsigned long
log2 (unsigned long word)
{
  __asm__ ("bsfl %1,%0"
	   : "=r" (word)
	   : "r" (word));
  return word;
}

static __inline__ int
is_power_of_two (unsigned long word)
{
  return (word & -word) == word;
}

static int 
journal_read (int block, int len, char *buffer) 
{
  return devread ((INFO->journal_block + block) << INFO->blocksize_shift, 
		  0, len, buffer);
}

/* Read a block from ReiserFS file system, taking the journal into
 * account.  If the block nr is in the journal, the block from the
 * journal taken.  
 */
static int
block_read (int blockNr, int start, int len, char *buffer)
{
  int transactions = INFO->journal_transactions;
  int desc_block = INFO->journal_first_desc;
  int journal_mask = INFO->journal_block_count - 1;
  int translatedNr = blockNr;
  __u32 *journal_table = JOURNAL_START;
  while (transactions-- > 0) 
    {
      int i = 0;
      int j_len;
      if (*journal_table != 0xffffffff)
	{
	  /* Search for the blockNr in cached journal */
	  j_len = *journal_table++;
	  while (i++ < j_len)
	    {
	      if (*journal_table++ == blockNr)
		{
		  journal_table += j_len - i;
		  goto found;
		}
	    }
	}
      else
	{
	  /* This is the end of cached journal marker.  The remaining
	   * transactions are still on disk.
	   */
	  struct reiserfs_journal_desc   desc;
	  struct reiserfs_journal_commit commit;

	  if (! journal_read (desc_block, sizeof (desc), (char *) &desc))
	    return 0;

	  j_len = desc.j_len;
	  while (i < j_len && i < JOURNAL_TRANS_HALF)
	    if (desc.j_realblock[i++] == blockNr)
	      goto found;
	  
	  if (j_len >= JOURNAL_TRANS_HALF)
	    {
	      int commit_block = (desc_block + 1 + j_len) & journal_mask;
	      if (! journal_read (commit_block, 
				  sizeof (commit), (char *) &commit))
		return 0;
	      while (i < j_len)
		if (commit.j_realblock[i++ - JOURNAL_TRANS_HALF] == blockNr)
		  goto found;
	    }
	}
      goto not_found;
      
    found:
      translatedNr = INFO->journal_block + ((desc_block + i) & journal_mask);
#ifdef REISERDEBUG
      printf ("block_read: block %d is mapped to journal block %d.\n", 
	      blockNr, translatedNr - INFO->journal_block);
#endif
      /* We must continue the search, as this block may be overwritten
       * in later transactions.
       */
    not_found:
      desc_block = (desc_block + 2 + j_len) & journal_mask;
    }
  return devread (translatedNr << INFO->blocksize_shift, start, len, buffer);
}

/* Init the journal data structure.  We try to cache as much as
 * possible in the JOURNAL_START-JOURNAL_END space, but if it is full
 * we can still read the rest from the disk on demand.
 *
 * The first number of valid transactions and the descriptor block of the
 * first valid transaction are held in INFO.  The transactions are all 
 * adjacent, but we must take care of the journal wrap around. 
 */
static int
journal_init (void)
{
  unsigned int block_count = INFO->journal_block_count;
  unsigned int desc_block;
  unsigned int commit_block;
  unsigned int next_trans_id;
  struct reiserfs_journal_header header;
  struct reiserfs_journal_desc   desc;
  struct reiserfs_journal_commit commit;
  __u32 *journal_table = JOURNAL_START;

  journal_read (block_count, sizeof (header), (char *) &header);
  desc_block = header.j_first_unflushed_offset;
  if (desc_block >= block_count)
    return 0;

  INFO->journal_transactions = 0;
  INFO->journal_first_desc = desc_block;
  next_trans_id = header.j_last_flush_trans_id + 1;

#ifdef REISERDEBUG
  printf ("journal_init: last flushed %d\n", 
	  header.j_last_flush_trans_id);
#endif

  while (1) 
    {
      journal_read (desc_block, sizeof (desc), (char *) &desc);
      if (substring (JOURNAL_DESC_MAGIC, desc.j_magic) > 0
	  || desc.j_trans_id != next_trans_id
	  || desc.j_mount_id != header.j_mount_id)
	/* no more valid transactions */
	break;
      
      commit_block = (desc_block + desc.j_len + 1) & (block_count - 1);
      journal_read (commit_block, sizeof (commit), (char *) &commit);
      if (desc.j_trans_id != commit.j_trans_id
	  || desc.j_len != commit.j_len)
	/* no more valid transactions */
	break;
      
#ifdef REISERDEBUG
      printf ("Found valid transaction %d/%d at %d.\n", 
	      desc.j_trans_id, desc.j_mount_id, desc_block);
#endif

      next_trans_id++;
      if (journal_table < JOURNAL_END)
	{
	  if ((journal_table + 1 + desc.j_len) >= JOURNAL_END)
	    {
	      /* The table is almost full; mark the end of the cached
	       * journal.*/
	      *journal_table = 0xffffffff;
	      journal_table = JOURNAL_END;
	    }
	  else
	    {
	      int i;
	      /* Cache the length and the realblock numbers in the table.
	       * The block number of descriptor can easily be computed.
	       * and need not to be stored here.
	       */
	      *journal_table++ = desc.j_len;
	      for (i = 0; i < desc.j_len && i < JOURNAL_TRANS_HALF; i++)
		{
		  *journal_table++ = desc.j_realblock[i];
#ifdef REISERDEBUG
		  printf ("block %d is in journal %d.\n", 
			  desc.j_realblock[i], desc_block);
#endif
		}
	      for (     ; i < desc.j_len; i++)
		{
		  *journal_table++ = commit.j_realblock[i-JOURNAL_TRANS_HALF];
#ifdef REISERDEBUG
		  printf ("block %d is in journal %d.\n", 
			  commit.j_realblock[i-JOURNAL_TRANS_HALF], 
			  desc_block);
#endif
		}
	    }
	}
      desc_block = (commit_block + 1) & (block_count - 1);
    }
#ifdef REISERDEBUG
  printf ("Transaction %d/%d at %d isn't valid.\n", 
	  desc.j_trans_id, desc.j_mount_id, desc_block);
#endif

  INFO->journal_transactions
    = next_trans_id - header.j_last_flush_trans_id - 1;
  return errnum == 0;
}

/* check filesystem types and read superblock into memory buffer */
int
reiserfs_mount (void)
{
  struct reiserfs_super_block super;
  int superblock = REISERFS_DISK_OFFSET_IN_BYTES >> SECTOR_BITS;

  if (part_length < superblock + (sizeof (super) >> SECTOR_BITS)
      || ! devread (superblock, 0, sizeof (struct reiserfs_super_block), 
		(char *) &super)
      || (substring (REISER2FS_SUPER_MAGIC_STRING, super.s_magic) > 0
	  && substring (REISERFS_SUPER_MAGIC_STRING, super.s_magic) > 0)
      || (/* check that this is not a copy inside the journal log */
	  super.s_journal_block * super.s_blocksize
	  <= REISERFS_DISK_OFFSET_IN_BYTES))
    {
      /* Try old super block position */
      superblock = REISERFS_OLD_DISK_OFFSET_IN_BYTES >> SECTOR_BITS;
      if (part_length < superblock + (sizeof (super) >> SECTOR_BITS)
	  || ! devread (superblock, 0, sizeof (struct reiserfs_super_block), 
			(char *) &super))
	return 0;
      
      if (substring (REISER2FS_SUPER_MAGIC_STRING, super.s_magic) > 0
	  && substring (REISERFS_SUPER_MAGIC_STRING, super.s_magic) > 0)
	{
	  /* pre journaling super block ? */
	  if (substring (REISERFS_SUPER_MAGIC_STRING, 
			 (char*) ((int) &super + 20)) > 0)
	    return 0;
	  
	  super.s_blocksize = REISERFS_OLD_BLOCKSIZE;
	  super.s_journal_block = 0;
	  super.s_version = 0;
	}
    }
  
  /* check the version number.  */
  if (super.s_version > REISERFS_MAX_SUPPORTED_VERSION)
    return 0;
  
  INFO->version = super.s_version;
  INFO->blocksize = super.s_blocksize;
  INFO->fullblocksize_shift = log2 (super.s_blocksize);
  INFO->blocksize_shift = INFO->fullblocksize_shift - SECTOR_BITS;
  INFO->cached_slots = 
    (FSYSREISER_CACHE_SIZE >> INFO->fullblocksize_shift) - 1;
  
  if (super.s_blocksize < FSYSREISER_MIN_BLOCKSIZE
      || super.s_blocksize > FSYSREISER_MAX_BLOCKSIZE
      || (SECTOR_SIZE << INFO->blocksize_shift) != super.s_blocksize)
    return 0;

  if (super.s_journal_block != 0)
    {
      INFO->journal_block = super.s_journal_block;
      INFO->journal_block_count = super.s_journal_block_count;
      if (INFO->journal_block_count == 0)
	INFO->journal_block_count = super.s_orig_journal_size;
      if (! is_power_of_two (INFO->journal_block_count))
	return 0;

      journal_init ();
      /* Read in super block again, maybe it is in the journal */
      block_read (superblock >> INFO->blocksize_shift, 
		  0, sizeof (struct reiserfs_super_block), (char *) &super);
    }

  if (! block_read (super.s_root_block, 0, INFO->blocksize, (char*) ROOT))
    return 0;
  
  INFO->tree_depth = BLOCKHEAD (ROOT)->blk_level;
  
#ifdef REISERDEBUG
  printf ("root read_in: block=%d, depth=%d\n", 
	  super.s_root_block, INFO->tree_depth);
#endif /* REISERDEBUG */
  
  if (INFO->tree_depth >= MAX_HEIGHT)
    return 0;
  if (INFO->tree_depth == DISK_LEAF_NODE_LEVEL)
    {
      /* There is only one node in the whole filesystem, 
       * which is simultanously leaf and root */
      memcpy (LEAF, ROOT, INFO->blocksize);
    }
  return 1;
}

/***************** TREE ACCESSING METHODS *****************************/

/* I assume you are familiar with the ReiserFS tree, if not go to
 * http://devlinux.com/projects/reiserfs/
 *
 * My tree node cache is organized as following
 *   0   ROOT node
 *   1   LEAF node  (if the ROOT is also a LEAF it is copied here
 *   2-n other nodes on current path from bottom to top.  
 *       if there is not enough space in the cache, the top most are
 *       omitted.
 *
 * I have only two methods to find a key in the tree:
 *   search_stat(dir_id, objectid) searches for the stat entry (always
 *       the first entry) of an object.
 *   next_key() gets the next key in tree order.
 *
 * This means, that I can only sequential reads of files are
 * efficient, but this really doesn't hurt for grub.  
 */

/* Read in the node at the current path and depth into the node cache.
 * You must set INFO->blocks[depth] before.
 */
static char *
read_tree_node (unsigned int blockNr, int depth)
{
  char* cache = CACHE(depth);
  int num_cached = INFO->cached_slots;
  if (depth < num_cached)
    {
      /* This is the cached part of the path.  Check if same block is
       * needed.  
       */
      if (blockNr == INFO->blocks[depth])
	return cache;
    }
  else
    cache = CACHE(num_cached);

#ifdef REISERDEBUG
  printf ("  next read_in: block=%d (depth=%d)\n",
	  blockNr, depth);
#endif /* REISERDEBUG */
  if (! block_read (blockNr, 0, INFO->blocksize, cache))
    return 0;
  /* Make sure it has the right node level */
  if (BLOCKHEAD (cache)->blk_level != depth)
    {
      errnum = ERR_FSYS_CORRUPT;
      return 0;
    }

  INFO->blocks[depth] = blockNr;
  return cache;
}

/* Get the next key, i.e. the key following the last retrieved key in
 * tree order.  INFO->current_ih and 
 * INFO->current_info are adapted accordingly.  */
static int
next_key (void)
{
  int depth;
  struct item_head *ih = INFO->current_ih + 1;
  char *cache;
  
#ifdef REISERDEBUG
  printf ("next_key:\n  old ih: key %d:%d:%d:%d version:%d\n", 
	  INFO->current_ih->ih_key.k_dir_id, 
	  INFO->current_ih->ih_key.k_objectid, 
	  INFO->current_ih->ih_key.u.v1.k_offset,
	  INFO->current_ih->ih_key.u.v1.k_uniqueness,
	  INFO->current_ih->ih_version);
#endif /* REISERDEBUG */
  
  if (ih == &ITEMHEAD[BLOCKHEAD (LEAF)->blk_nr_item])
    {
      depth = DISK_LEAF_NODE_LEVEL;
      /* The last item, was the last in the leaf node.  
       * Read in the next block */
      do
	{
	  if (depth == INFO->tree_depth)
	    {
	      /* There are no more keys at all.
	       * Return a dummy item with MAX_KEY */
	      ih = (struct item_head *) &BLOCKHEAD (LEAF)->blk_right_delim_key;
	      goto found;
	    }
	  depth++;
#ifdef REISERDEBUG
	  printf ("  depth=%d, i=%d\n", depth, INFO->next_key_nr[depth]);
#endif /* REISERDEBUG */
	}
      while (INFO->next_key_nr[depth] == 0);
      
      if (depth == INFO->tree_depth)
	cache = ROOT;
      else if (depth <= INFO->cached_slots)
	cache = CACHE (depth);
      else 
	{
	  cache = read_tree_node (INFO->blocks[depth], --depth);
	  if (! cache)
	    return 0;
	}
      
      do
	{
	  int nr_item = BLOCKHEAD (cache)->blk_nr_item;
	  int key_nr = INFO->next_key_nr[depth]++;
#ifdef REISERDEBUG
	  printf ("  depth=%d, i=%d/%d\n", depth, key_nr, nr_item);
#endif /* REISERDEBUG */
	  if (key_nr == nr_item)
	    /* This is the last item in this block, set the next_key_nr to 0 */
	    INFO->next_key_nr[depth] = 0;

	  cache = read_tree_node (DC (cache)[key_nr].dc_block_number, --depth);
	  if (! cache)
	    return 0;
	}
      while (depth > DISK_LEAF_NODE_LEVEL);
      
      ih = ITEMHEAD;
    }
 found:
  INFO->current_ih   = ih;
  INFO->current_item = &LEAF[ih->ih_item_location];
#ifdef REISERDEBUG
  printf ("  new ih: key %d:%d:%d:%d version:%d\n", 
	  INFO->current_ih->ih_key.k_dir_id, 
	  INFO->current_ih->ih_key.k_objectid, 
	  INFO->current_ih->ih_key.u.v1.k_offset,
	  INFO->current_ih->ih_key.u.v1.k_uniqueness,
	  INFO->current_ih->ih_version);
#endif /* REISERDEBUG */
  return 1;
}

/* preconditions: reiserfs_mount already executed, therefore 
 *   INFO block is valid
 * returns: 0 if error (errnum is set), 
 *   nonzero iff we were able to find the key successfully.
 * postconditions: on a nonzero return, the current_ih and 
 *   current_item fields describe the key that equals the
 *   searched key.  INFO->next_key contains the next key after
 *   the searched key.
 * side effects: messes around with the cache.
 */
static int
search_stat (__u32 dir_id, __u32 objectid) 
{
  char *cache;
  int depth;
  int nr_item;
  int i;
  struct item_head *ih;
#ifdef REISERDEBUG
  printf ("search_stat:\n  key %d:%d:0:0\n", dir_id, objectid);
#endif /* REISERDEBUG */
  
  depth = INFO->tree_depth;
  cache = ROOT;
  
  while (depth > DISK_LEAF_NODE_LEVEL)
    {
      struct key *key;
      nr_item = BLOCKHEAD (cache)->blk_nr_item;
      
      key = KEY (cache);
      
      for (i = 0; i < nr_item; i++) 
	{
	  if (key->k_dir_id > dir_id
	      || (key->k_dir_id == dir_id 
		  && (key->k_objectid > objectid
		      || (key->k_objectid == objectid
			  && (key->u.v1.k_offset
			      | key->u.v1.k_uniqueness) > 0))))
	    break;
	  key++;
	}
      
#ifdef REISERDEBUG
      printf ("  depth=%d, i=%d/%d\n", depth, i, nr_item);
#endif /* REISERDEBUG */
      INFO->next_key_nr[depth] = (i == nr_item) ? 0 : i+1;
      cache = read_tree_node (DC (cache)[i].dc_block_number, --depth);
      if (! cache)
	return 0;
    }
  
  /* cache == LEAF */
  nr_item = BLOCKHEAD (LEAF)->blk_nr_item;
  ih = ITEMHEAD;
  for (i = 0; i < nr_item; i++) 
    {
      if (ih->ih_key.k_dir_id == dir_id 
	  && ih->ih_key.k_objectid == objectid
	  && ih->ih_key.u.v1.k_offset == 0
	  && ih->ih_key.u.v1.k_uniqueness == 0)
	{
#ifdef REISERDEBUG
	  printf ("  depth=%d, i=%d/%d\n", depth, i, nr_item);
#endif /* REISERDEBUG */
	  INFO->current_ih   = ih;
	  INFO->current_item = &LEAF[ih->ih_item_location];
	  return 1;
	}
      ih++;
    }
  errnum = ERR_FSYS_CORRUPT;
  return 0;
}

int
reiserfs_read (char *buf, int len)
{
  unsigned int blocksize;
  unsigned int offset;
  unsigned int to_read;
  char *prev_buf = buf;
  
#ifdef REISERDEBUG
  printf ("reiserfs_read: filepos=%d len=%d, offset=%x:%x\n",
	  filepos, len, (__u64) IH_KEY_OFFSET (INFO->current_ih) - 1);
#endif /* REISERDEBUG */
  
  if (INFO->current_ih->ih_key.k_objectid != INFO->fileinfo.k_objectid
      || IH_KEY_OFFSET (INFO->current_ih) > filepos + 1)
    {
      search_stat (INFO->fileinfo.k_dir_id, INFO->fileinfo.k_objectid);
      goto get_next_key;
    }
  
  while (! errnum)
    {
      if (INFO->current_ih->ih_key.k_objectid != INFO->fileinfo.k_objectid)
	break;
      
      offset = filepos - IH_KEY_OFFSET (INFO->current_ih) + 1;
      blocksize = INFO->current_ih->ih_item_len;
      
#ifdef REISERDEBUG
      printf ("  loop: filepos=%d len=%d, offset=%d blocksize=%d\n",
	      filepos, len, offset, blocksize);
#endif /* REISERDEBUG */
      
      if (IH_KEY_ISTYPE(INFO->current_ih, TYPE_DIRECT)
	  && offset < blocksize)
	{
	  to_read = blocksize - offset;
	  if (to_read > len)
	    to_read = len;
	  
	  if (disk_read_hook != NULL)
	    {
	      disk_read_func = disk_read_hook;
	      
	      block_read (INFO->blocks[DISK_LEAF_NODE_LEVEL],
			  (INFO->current_item - LEAF + offset), to_read, buf);
	      
	      disk_read_func = NULL;
	    }
	  else
	    memcpy (buf, INFO->current_item + offset, to_read);
	  goto update_buf_len;
	}
      else if (IH_KEY_ISTYPE(INFO->current_ih, TYPE_INDIRECT))
	{
	  blocksize = (blocksize >> 2) << INFO->fullblocksize_shift;
	  
	  while (offset < blocksize)
	    {
	      __u32 blocknr = ((__u32 *) INFO->current_item)
		[offset >> INFO->fullblocksize_shift];
	      int blk_offset = offset & (INFO->blocksize-1);
	      
	      to_read = INFO->blocksize - blk_offset;
	      if (to_read > len)
		to_read = len;
	      
	      disk_read_func = disk_read_hook;
	      
	      /* Journal is only for meta data.  Data blocks can be read
	       * directly without using block_read
	       */
	      devread (blocknr << INFO->blocksize_shift,
		       blk_offset, to_read, buf);
	      
	      disk_read_func = NULL;
	    update_buf_len:
	      len -= to_read;
	      buf += to_read;
	      offset += to_read;
	      filepos += to_read;
	      if (len == 0)
		goto done;
	    }
	}
    get_next_key:
      next_key ();
    }
 done:
  return errnum ? 0 : buf - prev_buf;
}


/* preconditions: reiserfs_mount already executed, therefore 
 *   INFO block is valid
 * returns: 0 if error, nonzero iff we were able to find the file successfully
 * postconditions: on a nonzero return, INFO->fileinfo contains the info
 *   of the file we were trying to look up, filepos is 0 and filemax is 
 *   the size of the file.
 */
int
reiserfs_dir (char *dirname)
{
  struct reiserfs_de_head *de_head;
  char *rest, ch;
  __u32 dir_id, objectid, parent_dir_id = 0, parent_objectid = 0;
#ifndef STAGE1_5
  int do_possibilities = 0;
#endif /* ! STAGE1_5 */
  char linkbuf[PATH_MAX];	/* buffer for following symbolic links */
  int link_count = 0;
  int mode;

  dir_id = REISERFS_ROOT_PARENT_OBJECTID;
  objectid = REISERFS_ROOT_OBJECTID;
  
  while (1)
    {
#ifdef REISERDEBUG
      printf ("dirname=%s\n", dirname);
#endif /* REISERDEBUG */
      
      /* Search for the stat info first. */
      if (! search_stat (dir_id, objectid))
	return 0;
      
#ifdef REISERDEBUG
      printf ("sd_mode=%x sd_size=%d\n", 
	      ((struct stat_data *) INFO->current_item)->sd_mode,
	      ((struct stat_data *) INFO->current_item)->sd_size);
#endif /* REISERDEBUG */
      
      mode = ((struct stat_data *) INFO->current_item)->sd_mode;

      /* If we've got a symbolic link, then chase it. */
      if (S_ISLNK (mode))
	{
	  int len;
	  if (++link_count > MAX_LINK_COUNT)
	    {
	      errnum = ERR_SYMLINK_LOOP;
	      return 0;
	    }

	  /* Get the symlink size. */
	  filemax = ((struct stat_data *) INFO->current_item)->sd_size;

	  /* Find out how long our remaining name is. */
	  len = 0;
	  while (dirname[len] && !isspace (dirname[len]))
	    len++;

	  if (filemax + len > sizeof (linkbuf) - 1)
	    {
	      errnum = ERR_FILELENGTH;
	      return 0;
	    }
 	  
	  /* Copy the remaining name to the end of the symlink data.
	     Note that DIRNAME and LINKBUF may overlap! */
	  grub_memmove (linkbuf + filemax, dirname, len+1);

	  INFO->fileinfo.k_dir_id = dir_id;
	  INFO->fileinfo.k_objectid = objectid;
  	  filepos = 0;
	  if (! next_key ()
	      || reiserfs_read (linkbuf, filemax) != filemax)
	    return 0;

#ifdef REISERDEBUG
	  printf ("symlink=%s\n", linkbuf);
#endif /* REISERDEBUG */

	  dirname = linkbuf;
	  if (*dirname == '/')
	    {
	      /* It's an absolute link, so look it up in root. */
	      dir_id = REISERFS_ROOT_PARENT_OBJECTID;
	      objectid = REISERFS_ROOT_OBJECTID;
	    }
	  else
	    {
	      /* Relative, so look it up in our parent directory. */
	      dir_id   = parent_dir_id;
	      objectid = parent_objectid;
	    }

	  /* Now lookup the new name. */
	  continue;
	}

      /* if we have a real file (and we're not just printing possibilities),
	 then this is where we want to exit */
      
      if (! *dirname || isspace (*dirname))
	{
	  if (! S_ISREG (mode))
	    {
	      errnum = ERR_BAD_FILETYPE;
	      return 0;
	    }
	  
	  filepos = 0;
	  filemax = ((struct stat_data *) INFO->current_item)->sd_size;
	  
	  /* If this is a new stat data and size is > 4GB set filemax to 
	   * maximum
	   */
	  if (INFO->version >= 2
	      && INFO->current_ih->ih_version == ITEM_VERSION_2
	      && ((struct stat_data *) INFO->current_item)->sd_size_hi > 0)
	    filemax = 0xffffffff;
	  
	  INFO->fileinfo.k_dir_id = dir_id;
	  INFO->fileinfo.k_objectid = objectid;
	  return next_key ();
	}
      
      /* continue with the file/directory name interpretation */
      while (*dirname == '/')
	dirname++;
      if (! S_ISDIR (mode))
	{
	  errnum = ERR_BAD_FILETYPE;
	  return 0;
	}
      for (rest = dirname; (ch = *rest) && ! isspace (ch) && ch != '/'; rest++);
      *rest = 0;
      
# ifndef STAGE1_5
      if (print_possibilities && ch != '/')
	do_possibilities = 1;
# endif /* ! STAGE1_5 */
      
      while (1)
	{
	  char *name_end;
	  int num_entries;
	  
	  if (! next_key ())
	    return 0;
#ifdef REISERDEBUG
	  printf ("ih: key %d:%d:%d:%d version:%d\n", 
		  INFO->current_ih->ih_key.k_dir_id, 
		  INFO->current_ih->ih_key.k_objectid, 
		  INFO->current_ih->ih_key.u.v1.k_offset,
		  INFO->current_ih->ih_key.u.v1.k_uniqueness,
		  INFO->current_ih->ih_version);
#endif /* REISERDEBUG */
	  
	  if (INFO->current_ih->ih_key.k_objectid != objectid)
	    break;
	  
	  name_end = INFO->current_item + INFO->current_ih->ih_item_len;
	  de_head = (struct reiserfs_de_head *) INFO->current_item;
	  num_entries = INFO->current_ih->u.ih_entry_count;
	  while (num_entries > 0)
	    {
	      char *filename = INFO->current_item + de_head->deh_location;
	      char  tmp = *name_end;
	      if ((de_head->deh_state & DEH_Visible))
		{
		  int cmp;
		  /* Directory names in ReiserFS are not null
		   * terminated.  We write a temporary 0 behind it.
		   * NOTE: that this may overwrite the first block in
		   * the tree cache.  That doesn't hurt as long as we
		   * don't call next_key () in between.  
		   */
		  *name_end = 0;
		  cmp = substring (dirname, filename);
		  *name_end = tmp;
# ifndef STAGE1_5
		  if (do_possibilities)
		    {
		      if (cmp <= 0)
			{
			  if (print_possibilities > 0)
			    print_possibilities = -print_possibilities;
			  *name_end = 0;
			  print_a_completion (filename);
			  *name_end = tmp;
			}
		    }
		  else
# endif /* ! STAGE1_5 */
		    if (cmp == 0)
		      goto found;
		}
	      /* The beginning of this name marks the end of the next name.
	       */
	      name_end = filename;
	      de_head++;
	      num_entries--;
	    }
	}
      
# ifndef STAGE1_5
      if (print_possibilities < 0)
	return 1;
# endif /* ! STAGE1_5 */
      
      errnum = ERR_FILE_NOT_FOUND;
      *rest = ch;
      return 0;
      
    found:
      
      *rest = ch;
      dirname = rest;

      parent_dir_id = dir_id;
      parent_objectid = objectid;
      dir_id = de_head->deh_dir_id;
      objectid = de_head->deh_objectid;
    }
}

int
reiserfs_embed (int *start_sector, int needed_sectors)
{
  struct reiserfs_super_block super;
  int num_sectors;
  
  if (! devread (REISERFS_DISK_OFFSET_IN_BYTES >> SECTOR_BITS, 0, 
		 sizeof (struct reiserfs_super_block), (char *) &super))
    return 0;
  
  *start_sector = 1; /* reserve first sector for stage1 */
  if ((substring (REISERFS_SUPER_MAGIC_STRING, super.s_magic) <= 0
       || substring (REISER2FS_SUPER_MAGIC_STRING, super.s_magic) <= 0)
      && (/* check that this is not a super block copy inside
	   * the journal log */
	  super.s_journal_block * super.s_blocksize 
	  > REISERFS_DISK_OFFSET_IN_BYTES))
    num_sectors = (REISERFS_DISK_OFFSET_IN_BYTES >> SECTOR_BITS) - 1;
  else
    num_sectors = (REISERFS_OLD_DISK_OFFSET_IN_BYTES >> SECTOR_BITS) - 1;
  
  return (needed_sectors <= num_sectors);
}
#endif /* FSYS_REISERFS */