namei.c

ocfs1.2.7 源码

/* -*- mode: c; c-basic-offset: 8; -*-
 * vim: noexpandtab sw=8 ts=8 sts=0:
 *
 * namei.c
 *
 * Create and rename file, directory, symlinks
 *
 * Copyright (C) 2002, 2004 Oracle.  All rights reserved.
 *
 *  Portions of this code from linux/fs/ext3/dir.c
 *
 *  Copyright (C) 1992, 1993, 1994, 1995
 *  Remy Card ([e-mail])
 *  Laboratoire MASI - Institut Blaise pascal
 *  Universite Pierre et Marie Curie (Paris VI)
 *
 *   from
 *
 *   linux/fs/minix/dir.c
 *
 *   Copyright (C) 1991, 1992 Linux Torvalds
 *
 * 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., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

#include <linux/fs.h>
#include <linux/types.h>
#include <linux/slab.h>
#include <linux/highmem.h>

#define MLOG_MASK_PREFIX ML_NAMEI
#include <cluster/masklog.h>

#include "ocfs2.h"

#include "alloc.h"
#include "dcache.h"
#include "dir.h"
#include "dlmglue.h"
#include "extent_map.h"
#include "file.h"
#include "inode.h"
#include "journal.h"
#include "namei.h"
#include "suballoc.h"
#include "symlink.h"
#include "sysfile.h"
#include "uptodate.h"
#include "vote.h"

#include "buffer_head_io.h"

#define NAMEI_RA_CHUNKS  2
#define NAMEI_RA_BLOCKS  4
#define NAMEI_RA_SIZE        (NAMEI_RA_CHUNKS * NAMEI_RA_BLOCKS)
#define NAMEI_RA_INDEX(c,b)  (((c) * NAMEI_RA_BLOCKS) + (b))

static int inline ocfs2_search_dirblock(struct buffer_head *bh,
					struct inode *dir,
					const char *name, int namelen,
					unsigned long offset,
					struct ocfs2_dir_entry **res_dir);

static int ocfs2_delete_entry(struct ocfs2_journal_handle *handle,
			      struct inode *dir,
			      struct ocfs2_dir_entry *de_del,
			      struct buffer_head *bh);

static int __ocfs2_add_entry(struct ocfs2_journal_handle *handle,
			     struct inode *dir,
			     const char *name, int namelen,
			     struct inode *inode, u64 blkno,
			     struct buffer_head *parent_fe_bh,
			     struct buffer_head *insert_bh);

static int ocfs2_mknod_locked(struct ocfs2_super *osb,
			      struct inode *dir,
			      struct dentry *dentry, int mode,
			      dev_t dev,
			      struct buffer_head **new_fe_bh,
			      struct buffer_head *parent_fe_bh,
			      struct ocfs2_journal_handle *handle,
			      struct inode **ret_inode,
			      struct ocfs2_alloc_context *inode_ac);

static int ocfs2_fill_new_dir(struct ocfs2_super *osb,
			      struct ocfs2_journal_handle *handle,
			      struct inode *parent,
			      struct inode *inode,
			      struct buffer_head *fe_bh,
			      struct ocfs2_alloc_context *data_ac);

static int ocfs2_double_lock(struct ocfs2_super *osb,
			     struct ocfs2_journal_handle *handle,
			     struct buffer_head **bh1,
			     struct inode *inode1,
			     struct buffer_head **bh2,
			     struct inode *inode2);

static int ocfs2_prepare_orphan_dir(struct ocfs2_super *osb,
				    struct ocfs2_journal_handle *handle,
				    struct inode *inode,
				    char *name,
				    struct buffer_head **de_bh);

static int ocfs2_orphan_add(struct ocfs2_super *osb,
			    struct ocfs2_journal_handle *handle,
			    struct inode *inode,
			    struct ocfs2_dinode *fe,
			    char *name,
			    struct buffer_head *de_bh);

static int ocfs2_create_symlink_data(struct ocfs2_super *osb,
				     struct ocfs2_journal_handle *handle,
				     struct inode *inode,
				     const char *symname);

static inline int ocfs2_add_entry(struct ocfs2_journal_handle *handle,
				  struct dentry *dentry,
				  struct inode *inode, u64 blkno,
				  struct buffer_head *parent_fe_bh,
				  struct buffer_head *insert_bh)
{
	return __ocfs2_add_entry(handle, dentry->d_parent->d_inode,
				 dentry->d_name.name, dentry->d_name.len,
				 inode, blkno, parent_fe_bh, insert_bh);
}

/* An orphan dir name is an 8 byte value, printed as a hex string */
#define OCFS2_ORPHAN_NAMELEN ((int)(2 * sizeof(u64)))

static struct dentry *ocfs2_lookup(struct inode *dir, struct dentry *dentry,
				   struct nameidata *nd)
{
	int status;
	u64 blkno;
	struct buffer_head *dirent_bh = NULL;
	struct inode *inode = NULL;
	struct dentry *ret;
	struct ocfs2_dir_entry *dirent;
	struct ocfs2_inode_info *oi;

	mlog_entry("(0x%p, 0x%p, '%.*s')\n", dir, dentry,
		   dentry->d_name.len, dentry->d_name.name);

	if (dentry->d_name.len > OCFS2_MAX_FILENAME_LEN) {
		ret = ERR_PTR(-ENAMETOOLONG);
		goto bail;
	}

	mlog(0, "find name %.*s in directory %"MLFu64"\n", dentry->d_name.len,
	     dentry->d_name.name, OCFS2_I(dir)->ip_blkno);

	status = ocfs2_meta_lock(dir, NULL, NULL, 0);
	if (status < 0) {
		if (status != -ENOENT)
			mlog_errno(status);
		ret = ERR_PTR(status);
		goto bail;
	}

	status = ocfs2_find_files_on_disk(dentry->d_name.name,
					  dentry->d_name.len, &blkno,
					  dir, &dirent_bh, &dirent);
	if (status < 0)
		goto bail_add;

	inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno);
	if (IS_ERR(inode)) {
		mlog(ML_ERROR, "Unable to create inode %"MLFu64"\n", blkno);
		ret = ERR_PTR(-EACCES);
		goto bail_unlock;
	}

	oi = OCFS2_I(inode);
	/* Clear any orphaned state... If we were able to look up the
	 * inode from a directory, it certainly can't be orphaned. We
	 * might have the bad state from a node which intended to
	 * orphan this inode but crashed before it could commit the
	 * unlink. */
	spin_lock(&oi->ip_lock);
	oi->ip_flags &= ~OCFS2_INODE_MAYBE_ORPHANED;
	oi->ip_orphaned_slot = OCFS2_INVALID_SLOT;
	oi->ip_deleting_node = O2NM_INVALID_NODE_NUM;
	spin_unlock(&oi->ip_lock);

bail_add:

	dentry->d_op = &ocfs2_dentry_ops;
	ret = d_splice_alias(inode, dentry);

bail_unlock:
	/* Don't drop the cluster lock until *after* the d_add --
	 * unlink on another node will message us to remove that
	 * dentry under this lock so otherwise we can race this with
	 * the vote thread and have a stale dentry. */
	ocfs2_meta_unlock(dir, 0);

bail:
	if (dirent_bh)
		brelse(dirent_bh);

	mlog_exit_ptr(ret);

	return ret;
}

static int ocfs2_fill_new_dir(struct ocfs2_super *osb,
			      struct ocfs2_journal_handle *handle,
			      struct inode *parent,
			      struct inode *inode,
			      struct buffer_head *fe_bh,
			      struct ocfs2_alloc_context *data_ac)
{
	int status;
	struct buffer_head *new_bh = NULL;
	struct ocfs2_dir_entry *de = NULL;

	mlog_entry_void();

	status = ocfs2_do_extend_dir(osb->sb, handle, inode, fe_bh,
				     data_ac, NULL, &new_bh);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}

	ocfs2_set_new_buffer_uptodate(inode, new_bh);

	status = ocfs2_journal_access(handle, inode, new_bh,
				      OCFS2_JOURNAL_ACCESS_CREATE);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}
	memset(new_bh->b_data, 0, osb->sb->s_blocksize);

	de = (struct ocfs2_dir_entry *) new_bh->b_data;
	de->inode = cpu_to_le64(OCFS2_I(inode)->ip_blkno);
	de->name_len = 1;
	de->rec_len =
		cpu_to_le16(OCFS2_DIR_REC_LEN(de->name_len));
	strcpy(de->name, ".");
	ocfs2_set_de_type(de, S_IFDIR);
	de = (struct ocfs2_dir_entry *) ((char *)de + le16_to_cpu(de->rec_len));
	de->inode = cpu_to_le64(OCFS2_I(parent)->ip_blkno);
	de->rec_len = cpu_to_le16(inode->i_sb->s_blocksize -
				  OCFS2_DIR_REC_LEN(1));
	de->name_len = 2;
	strcpy(de->name, "..");
	ocfs2_set_de_type(de, S_IFDIR);

	status = ocfs2_journal_dirty(handle, new_bh);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}

	i_size_write(inode, inode->i_sb->s_blocksize);
	inode->i_nlink = 2;
	inode->i_blocks = ocfs2_align_bytes_to_sectors(inode->i_sb->s_blocksize);
	status = ocfs2_mark_inode_dirty(handle, inode, fe_bh);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}

	status = 0;
bail:
	if (new_bh)
		brelse(new_bh);

	mlog_exit(status);
	return status;
}

static int ocfs2_mknod(struct inode *dir,
		       struct dentry *dentry,
		       int mode,
		       dev_t dev)
{
	int status = 0;
	struct buffer_head *parent_fe_bh = NULL;
	struct ocfs2_journal_handle *handle = NULL;
	struct ocfs2_super *osb;
	struct ocfs2_dinode *dirfe;
	struct buffer_head *new_fe_bh = NULL;
	struct buffer_head *de_bh = NULL;
	struct inode *inode = NULL;
	struct ocfs2_alloc_context *inode_ac = NULL;
	struct ocfs2_alloc_context *data_ac = NULL;

	mlog_entry("(0x%p, 0x%p, %d, %lu, '%.*s')\n", dir, dentry, mode,
		   (unsigned long)dev, dentry->d_name.len,
		   dentry->d_name.name);

	/* get our super block */
	osb = OCFS2_SB(dir->i_sb);

	if (S_ISDIR(mode) && (dir->i_nlink >= OCFS2_LINK_MAX)) {
		mlog(ML_ERROR, "inode %"MLFu64" has i_nlink of %u\n",
		     OCFS2_I(dir)->ip_blkno, dir->i_nlink);
		status = -EMLINK;
		goto leave;
	}

	handle = ocfs2_alloc_handle(osb);
	if (handle == NULL) {
		status = -ENOMEM;
		mlog_errno(status);
		goto leave;
	}

	status = ocfs2_meta_lock(dir, handle, &parent_fe_bh, 1);
	if (status < 0) {
		if (status != -ENOENT)
			mlog_errno(status);
		goto leave;
	}

	dirfe = (struct ocfs2_dinode *) parent_fe_bh->b_data;
	if (!dirfe->i_links_count) {
		/* can't make a file in a deleted directory. */
		status = -ENOENT;
		goto leave;
	}

	status = ocfs2_check_dir_for_entry(dir, dentry->d_name.name,
					   dentry->d_name.len);
	if (status)
		goto leave;

	/* get a spot inside the dir. */
	status = ocfs2_prepare_dir_for_insert(osb, dir, parent_fe_bh,
					      dentry->d_name.name,
					      dentry->d_name.len, &de_bh);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	/* reserve an inode spot */
	status = ocfs2_reserve_new_inode(osb, handle, &inode_ac);
	if (status < 0) {
		if (status != -ENOSPC)
			mlog_errno(status);
		goto leave;
	}

	/* are we making a directory? If so, reserve a cluster for his
	 * 1st extent. */
	if (S_ISDIR(mode)) {
		status = ocfs2_reserve_clusters(osb, handle, 1, &data_ac);
		if (status < 0) {
			if (status != -ENOSPC)
				mlog_errno(status);
			goto leave;
		}
	}

	handle = ocfs2_start_trans(osb, handle, OCFS2_MKNOD_CREDITS);
	if (IS_ERR(handle)) {
		status = PTR_ERR(handle);
		handle = NULL;
		mlog_errno(status);
		goto leave;
	}

	/* do the real work now. */
	status = ocfs2_mknod_locked(osb, dir, dentry, mode, dev,
				    &new_fe_bh, parent_fe_bh, handle,
				    &inode, inode_ac);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	if (S_ISDIR(mode)) {
		status = ocfs2_fill_new_dir(osb, handle, dir, inode,
					    new_fe_bh, data_ac);
		if (status < 0) {
			mlog_errno(status);
			goto leave;
		}

		status = ocfs2_journal_access(handle, dir, parent_fe_bh,
					      OCFS2_JOURNAL_ACCESS_WRITE);
		if (status < 0) {
			mlog_errno(status);
			goto leave;
		}
		le16_add_cpu(&dirfe->i_links_count, 1);
		status = ocfs2_journal_dirty(handle, parent_fe_bh);
		if (status < 0) {
			mlog_errno(status);
			goto leave;
		}
		dir->i_nlink++;
	}

	status = ocfs2_add_entry(handle, dentry, inode,
				 OCFS2_I(inode)->ip_blkno, parent_fe_bh,
				 de_bh);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	insert_inode_hash(inode);
	dentry->d_op = &ocfs2_dentry_ops;
	d_instantiate(dentry, inode);
	status = 0;
leave:
	if (handle)
		ocfs2_commit_trans(handle);

	if (status == -ENOSPC)
		mlog(0, "Disk is full\n");

	if (new_fe_bh)
		brelse(new_fe_bh);

	if (de_bh)
		brelse(de_bh);

	if (parent_fe_bh)
		brelse(parent_fe_bh);

	if ((status < 0) && inode)
		iput(inode);

	if (inode_ac)
		ocfs2_free_alloc_context(inode_ac);

	if (data_ac)
		ocfs2_free_alloc_context(data_ac);

	mlog_exit(status);

	return status;
}

static int ocfs2_mknod_locked(struct ocfs2_super *osb,
			      struct inode *dir,
			      struct dentry *dentry, int mode,
			      dev_t dev,
			      struct buffer_head **new_fe_bh,
			      struct buffer_head *parent_fe_bh,
			      struct ocfs2_journal_handle *handle,
			      struct inode **ret_inode,
			      struct ocfs2_alloc_context *inode_ac)
{
	int status = 0;
	struct ocfs2_dinode *fe = NULL;
	struct ocfs2_extent_list *fel;
	u64 fe_blkno = 0;
	u16 suballoc_bit;
	struct inode *inode = NULL;

	mlog_entry("(0x%p, 0x%p, %d, %lu, '%.*s')\n", dir, dentry, mode,
		   (unsigned long)dev, dentry->d_name.len,
		   dentry->d_name.name);

	*new_fe_bh = NULL;
	*ret_inode = NULL;

	status = ocfs2_claim_new_inode(osb, handle, inode_ac, &suballoc_bit,
				       &fe_blkno);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	inode = new_inode(dir->i_sb);
	if (IS_ERR(inode)) {
		status = PTR_ERR(inode);
		mlog(ML_ERROR, "new_inode failed!\n");
		goto leave;
	}

	/* populate as many fields early on as possible - many of
	 * these are used by the support functions here and in
	 * callers. */
	inode->i_ino = ino_from_blkno(osb->sb, fe_blkno);
	OCFS2_I(inode)->ip_blkno = fe_blkno;
	if (S_ISDIR(mode))
		inode->i_nlink = 2;
	else
		inode->i_nlink = 1;
	inode->i_mode = mode;
	spin_lock(&osb->osb_lock);
	inode->i_generation = osb->s_next_generation++;
	spin_unlock(&osb->osb_lock);

	*new_fe_bh = sb_getblk(osb->sb, fe_blkno);
	if (!*new_fe_bh) {
		status = -EIO;
		mlog_errno(status);
		goto leave;
	}
	ocfs2_set_new_buffer_uptodate(inode, *new_fe_bh);

	status = ocfs2_journal_access(handle, inode, *new_fe_bh,
				      OCFS2_JOURNAL_ACCESS_CREATE);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	fe = (struct ocfs2_dinode *) (*new_fe_bh)->b_data;
	memset(fe, 0, osb->sb->s_blocksize);

	fe->i_generation = cpu_to_le32(inode->i_generation);
	fe->i_fs_generation = cpu_to_le32(osb->fs_generation);
	fe->i_blkno = cpu_to_le64(fe_blkno);
	fe->i_suballoc_bit = cpu_to_le16(suballoc_bit);
	fe->i_suballoc_slot = cpu_to_le16(osb->slot_num);
	fe->i_uid = cpu_to_le32(current->fsuid);
	if (dir->i_mode & S_ISGID) {
		fe->i_gid = cpu_to_le32(dir->i_gid);
		if (S_ISDIR(mode))
			mode |= S_ISGID;
	} else
		fe->i_gid = cpu_to_le32(current->fsgid);
	fe->i_mode = cpu_to_le16(mode);
	if (S_ISCHR(mode) || S_ISBLK(mode))
		fe->id1.dev1.i_rdev = cpu_to_le64(huge_encode_dev(dev));

	fe->i_links_count = cpu_to_le16(inode->i_nlink);

	fe->i_last_eb_blk = 0;
	strcpy(fe->i_signature, OCFS2_INODE_SIGNATURE);
	le32_add_cpu(&fe->i_flags, OCFS2_VALID_FL);
	fe->i_atime = fe->i_ctime = fe->i_mtime =
		cpu_to_le64(CURRENT_TIME.tv_sec);
	fe->i_mtime_nsec = fe->i_ctime_nsec = fe->i_atime_nsec =
		cpu_to_le32(CURRENT_TIME.tv_nsec);
	fe->i_dtime = 0;

	fel = &fe->id2.i_list;
	fel->l_tree_depth = 0;
	fel->l_next_free_rec = 0;
	fel->l_count = cpu_to_le16(ocfs2_extent_recs_per_inode(osb->sb));

	status = ocfs2_journal_dirty(handle, *new_fe_bh);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	if (ocfs2_populate_inode(inode, fe, 1) < 0) {
		mlog(ML_ERROR, "populate inode failed! bh->b_blocknr=%llu, "
		     "i_blkno=%"MLFu64", i_ino=%lu\n",
		     (unsigned long long) (*new_fe_bh)->b_blocknr,
		     fe->i_blkno, inode->i_ino);