xattr.c

linux 内核源代码

/*
 * linux/fs/reiserfs/xattr.c
 *
 * Copyright (c) 2002 by Jeff Mahoney, <[e-mail]>
 *
 */

/*
 * In order to implement EA/ACLs in a clean, backwards compatible manner,
 * they are implemented as files in a "private" directory.
 * Each EA is in it's own file, with the directory layout like so (/ is assumed
 * to be relative to fs root). Inside the /.reiserfs_priv/xattrs directory,
 * directories named using the capital-hex form of the objectid and
 * generation number are used. Inside each directory are individual files
 * named with the name of the extended attribute.
 *
 * So, for objectid 12648430, we could have:
 * /.reiserfs_priv/xattrs/C0FFEE.0/system.posix_acl_access
 * /.reiserfs_priv/xattrs/C0FFEE.0/system.posix_acl_default
 * /.reiserfs_priv/xattrs/C0FFEE.0/user.Content-Type
 * .. or similar.
 *
 * The file contents are the text of the EA. The size is known based on the
 * stat data describing the file.
 *
 * In the case of system.posix_acl_access and system.posix_acl_default, since
 * these are special cases for filesystem ACLs, they are interpreted by the
 * kernel, in addition, they are negatively and positively cached and attached
 * to the inode so that unnecessary lookups are avoided.
 */

#include <linux/reiserfs_fs.h>
#include <linux/capability.h>
#include <linux/dcache.h>
#include <linux/namei.h>
#include <linux/errno.h>
#include <linux/fs.h>
#include <linux/file.h>
#include <linux/pagemap.h>
#include <linux/xattr.h>
#include <linux/reiserfs_xattr.h>
#include <linux/reiserfs_acl.h>
#include <asm/uaccess.h>
#include <net/checksum.h>
#include <linux/smp_lock.h>
#include <linux/stat.h>
#include <asm/semaphore.h>

#define FL_READONLY 128
#define FL_DIR_SEM_HELD 256
#define PRIVROOT_NAME ".reiserfs_priv"
#define XAROOT_NAME   "xattrs"

static struct reiserfs_xattr_handler *find_xattr_handler_prefix(const char
								*prefix);

/* Returns the dentry referring to the root of the extended attribute
 * directory tree. If it has already been retrieved, it is used. If it
 * hasn't been created and the flags indicate creation is allowed, we
 * attempt to create it. On error, we return a pointer-encoded error.
 */
static struct dentry *get_xa_root(struct super_block *sb, int flags)
{
	struct dentry *privroot = dget(REISERFS_SB(sb)->priv_root);
	struct dentry *xaroot;

	/* This needs to be created at mount-time */
	if (!privroot)
		return ERR_PTR(-ENODATA);

	mutex_lock_nested(&privroot->d_inode->i_mutex, I_MUTEX_XATTR);
	if (REISERFS_SB(sb)->xattr_root) {
		xaroot = dget(REISERFS_SB(sb)->xattr_root);
		goto out;
	}

	xaroot = lookup_one_len(XAROOT_NAME, privroot, strlen(XAROOT_NAME));
	if (IS_ERR(xaroot)) {
		goto out;
	} else if (!xaroot->d_inode) {
		int err = -ENODATA;
		if (flags == 0 || flags & XATTR_CREATE)
			err = privroot->d_inode->i_op->mkdir(privroot->d_inode,
			                                     xaroot, 0700);
		if (err) {
			dput(xaroot);
			xaroot = ERR_PTR(err);
			goto out;
		}
	}
	REISERFS_SB(sb)->xattr_root = dget(xaroot);

      out:
	mutex_unlock(&privroot->d_inode->i_mutex);
	dput(privroot);
	return xaroot;
}

/* Opens the directory corresponding to the inode's extended attribute store.
 * If flags allow, the tree to the directory may be created. If creation is
 * prohibited, -ENODATA is returned. */
static struct dentry *open_xa_dir(const struct inode *inode, int flags)
{
	struct dentry *xaroot, *xadir;
	char namebuf[17];

	xaroot = get_xa_root(inode->i_sb, flags);
	if (IS_ERR(xaroot))
		return xaroot;

	/* ok, we have xaroot open */
	snprintf(namebuf, sizeof(namebuf), "%X.%X",
		 le32_to_cpu(INODE_PKEY(inode)->k_objectid),
		 inode->i_generation);
	xadir = lookup_one_len(namebuf, xaroot, strlen(namebuf));
	if (IS_ERR(xadir)) {
		dput(xaroot);
		return xadir;
	}

	if (!xadir->d_inode) {
		int err;
		if (flags == 0 || flags & XATTR_CREATE) {
			/* Although there is nothing else trying to create this directory,
			 * another directory with the same hash may be created, so we need
			 * to protect against that */
			err =
			    xaroot->d_inode->i_op->mkdir(xaroot->d_inode, xadir,
							 0700);
			if (err) {
				dput(xaroot);
				dput(xadir);
				return ERR_PTR(err);
			}
		}
		if (!xadir->d_inode) {
			dput(xaroot);
			dput(xadir);
			return ERR_PTR(-ENODATA);
		}
	}

	dput(xaroot);
	return xadir;
}

/* Returns a dentry corresponding to a specific extended attribute file
 * for the inode. If flags allow, the file is created. Otherwise, a
 * valid or negative dentry, or an error is returned. */
static struct dentry *get_xa_file_dentry(const struct inode *inode,
					 const char *name, int flags)
{
	struct dentry *xadir, *xafile;
	int err = 0;

	xadir = open_xa_dir(inode, flags);
	if (IS_ERR(xadir)) {
		return ERR_PTR(PTR_ERR(xadir));
	} else if (xadir && !xadir->d_inode) {
		dput(xadir);
		return ERR_PTR(-ENODATA);
	}

	xafile = lookup_one_len(name, xadir, strlen(name));
	if (IS_ERR(xafile)) {
		dput(xadir);
		return ERR_PTR(PTR_ERR(xafile));
	}

	if (xafile->d_inode) {	/* file exists */
		if (flags & XATTR_CREATE) {
			err = -EEXIST;
			dput(xafile);
			goto out;
		}
	} else if (flags & XATTR_REPLACE || flags & FL_READONLY) {
		goto out;
	} else {
		/* inode->i_mutex is down, so nothing else can try to create
		 * the same xattr */
		err = xadir->d_inode->i_op->create(xadir->d_inode, xafile,
						   0700 | S_IFREG, NULL);

		if (err) {
			dput(xafile);
			goto out;
		}
	}

      out:
	dput(xadir);
	if (err)
		xafile = ERR_PTR(err);
	return xafile;
}

/* Opens a file pointer to the attribute associated with inode */
static struct file *open_xa_file(const struct inode *inode, const char *name,
				 int flags)
{
	struct dentry *xafile;
	struct file *fp;

	xafile = get_xa_file_dentry(inode, name, flags);
	if (IS_ERR(xafile))
		return ERR_PTR(PTR_ERR(xafile));
	else if (!xafile->d_inode) {
		dput(xafile);
		return ERR_PTR(-ENODATA);
	}

	fp = dentry_open(xafile, NULL, O_RDWR);
	/* dentry_open dputs the dentry if it fails */

	return fp;
}

/*
 * this is very similar to fs/reiserfs/dir.c:reiserfs_readdir, but
 * we need to drop the path before calling the filldir struct.  That
 * would be a big performance hit to the non-xattr case, so I've copied
 * the whole thing for now. --clm
 *
 * the big difference is that I go backwards through the directory,
 * and don't mess with f->f_pos, but the idea is the same.  Do some
 * action on each and every entry in the directory.
 *
 * we're called with i_mutex held, so there are no worries about the directory
 * changing underneath us.
 */
static int __xattr_readdir(struct file *filp, void *dirent, filldir_t filldir)
{
	struct inode *inode = filp->f_path.dentry->d_inode;
	struct cpu_key pos_key;	/* key of current position in the directory (key of directory entry) */
	INITIALIZE_PATH(path_to_entry);
	struct buffer_head *bh;
	int entry_num;
	struct item_head *ih, tmp_ih;
	int search_res;
	char *local_buf;
	loff_t next_pos;
	char small_buf[32];	/* avoid kmalloc if we can */
	struct reiserfs_de_head *deh;
	int d_reclen;
	char *d_name;
	off_t d_off;
	ino_t d_ino;
	struct reiserfs_dir_entry de;

	/* form key for search the next directory entry using f_pos field of
	   file structure */
	next_pos = max_reiserfs_offset(inode);

	while (1) {
	      research:
		if (next_pos <= DOT_DOT_OFFSET)
			break;
		make_cpu_key(&pos_key, inode, next_pos, TYPE_DIRENTRY, 3);

		search_res =
		    search_by_entry_key(inode->i_sb, &pos_key, &path_to_entry,
					&de);
		if (search_res == IO_ERROR) {
			// FIXME: we could just skip part of directory which could
			// not be read
			pathrelse(&path_to_entry);
			return -EIO;
		}

		if (search_res == NAME_NOT_FOUND)
			de.de_entry_num--;

		set_de_name_and_namelen(&de);
		entry_num = de.de_entry_num;
		deh = &(de.de_deh[entry_num]);

		bh = de.de_bh;
		ih = de.de_ih;

		if (!is_direntry_le_ih(ih)) {
			reiserfs_warning(inode->i_sb, "not direntry %h", ih);
			break;
		}
		copy_item_head(&tmp_ih, ih);

		/* we must have found item, that is item of this directory, */
		RFALSE(COMP_SHORT_KEYS(&(ih->ih_key), &pos_key),
		       "vs-9000: found item %h does not match to dir we readdir %K",
		       ih, &pos_key);

		if (deh_offset(deh) <= DOT_DOT_OFFSET) {
			break;
		}

		/* look for the previous entry in the directory */
		next_pos = deh_offset(deh) - 1;

		if (!de_visible(deh))
			/* it is hidden entry */
			continue;

		d_reclen = entry_length(bh, ih, entry_num);
		d_name = B_I_DEH_ENTRY_FILE_NAME(bh, ih, deh);
		d_off = deh_offset(deh);
		d_ino = deh_objectid(deh);

		if (!d_name[d_reclen - 1])
			d_reclen = strlen(d_name);

		if (d_reclen > REISERFS_MAX_NAME(inode->i_sb->s_blocksize)) {
			/* too big to send back to VFS */
			continue;
		}

		/* Ignore the .reiserfs_priv entry */
		if (reiserfs_xattrs(inode->i_sb) &&
		    !old_format_only(inode->i_sb) &&
		    deh_objectid(deh) ==
		    le32_to_cpu(INODE_PKEY
				(REISERFS_SB(inode->i_sb)->priv_root->d_inode)->
				k_objectid))
			continue;

		if (d_reclen <= 32) {
			local_buf = small_buf;
		} else {
			local_buf = kmalloc(d_reclen, GFP_NOFS);
			if (!local_buf) {
				pathrelse(&path_to_entry);
				return -ENOMEM;
			}
			if (item_moved(&tmp_ih, &path_to_entry)) {
				kfree(local_buf);

				/* sigh, must retry.  Do this same offset again */
				next_pos = d_off;
				goto research;
			}
		}

		// Note, that we copy name to user space via temporary
		// buffer (local_buf) because filldir will block if
		// user space buffer is swapped out. At that time
		// entry can move to somewhere else
		memcpy(local_buf, d_name, d_reclen);

		/* the filldir function might need to start transactions,
		 * or do who knows what.  Release the path now that we've
		 * copied all the important stuff out of the deh
		 */
		pathrelse(&path_to_entry);

		if (filldir(dirent, local_buf, d_reclen, d_off, d_ino,
			    DT_UNKNOWN) < 0) {
			if (local_buf != small_buf) {
				kfree(local_buf);
			}
			goto end;
		}
		if (local_buf != small_buf) {
			kfree(local_buf);
		}
	}			/* while */

      end:
	pathrelse(&path_to_entry);
	return 0;
}

/*
 * this could be done with dedicated readdir ops for the xattr files,
 * but I want to get something working asap
 * this is stolen from vfs_readdir
 *
 */
static
int xattr_readdir(struct file *file, filldir_t filler, void *buf)
{
	struct inode *inode = file->f_path.dentry->d_inode;
	int res = -ENOTDIR;
	if (!file->f_op || !file->f_op->readdir)
		goto out;
	mutex_lock_nested(&inode->i_mutex, I_MUTEX_XATTR);
//        down(&inode->i_zombie);
	res = -ENOENT;
	if (!IS_DEADDIR(inode)) {
		lock_kernel();
		res = __xattr_readdir(file, buf, filler);
		unlock_kernel();
	}
//        up(&inode->i_zombie);
	mutex_unlock(&inode->i_mutex);
      out:
	return res;
}

/* Internal operations on file data */
static inline void reiserfs_put_page(struct page *page)
{
	kunmap(page);
	page_cache_release(page);
}

static struct page *reiserfs_get_page(struct inode *dir, unsigned long n)
{
	struct address_space *mapping = dir->i_mapping;
	struct page *page;
	/* We can deadlock if we try to free dentries,
	   and an unlink/rmdir has just occured - GFP_NOFS avoids this */
	mapping_set_gfp_mask(mapping, GFP_NOFS);
	page = read_mapping_page(mapping, n, NULL);
	if (!IS_ERR(page)) {
		kmap(page);
		if (PageError(page))
			goto fail;
	}
	return page;

      fail:
	reiserfs_put_page(page);
	return ERR_PTR(-EIO);
}

static inline __u32 xattr_hash(const char *msg, int len)
{
	return csum_partial(msg, len, 0);
}

int reiserfs_commit_write(struct file *f, struct page *page,
			  unsigned from, unsigned to);
int reiserfs_prepare_write(struct file *f, struct page *page,
			   unsigned from, unsigned to);


/* Generic extended attribute operations that can be used by xa plugins */

/*
 * inode->i_mutex: down
 */
int
reiserfs_xattr_set(struct inode *inode, const char *name, const void *buffer,