dir.c

linux 内核源代码

	desc->ptr = kmap(page);		/* matching kunmap in nfs_do_filldir */
	if (desc->error >= 0) {
		desc->timestamp = timestamp;
		desc->timestamp_valid = 1;
		if ((status = dir_decode(desc)) == 0)
			desc->entry->prev_cookie = *desc->dir_cookie;
	} else
		status = -EIO;
	if (status < 0)
		goto out_release;

	status = nfs_do_filldir(desc, dirent, filldir);

	/* Reset read descriptor so it searches the page cache from
	 * the start upon the next call to readdir_search_pagecache() */
	desc->page_index = 0;
	desc->entry->cookie = desc->entry->prev_cookie = 0;
	desc->entry->eof = 0;
 out:
	dfprintk(DIRCACHE, "NFS: %s: returns %d\n",
			__FUNCTION__, status);
	return status;
 out_release:
	dir_page_release(desc);
	goto out;
}

/* The file offset position represents the dirent entry number.  A
   last cookie cache takes care of the common case of reading the
   whole directory.
 */
static int nfs_readdir(struct file *filp, void *dirent, filldir_t filldir)
{
	struct dentry	*dentry = filp->f_path.dentry;
	struct inode	*inode = dentry->d_inode;
	nfs_readdir_descriptor_t my_desc,
			*desc = &my_desc;
	struct nfs_entry my_entry;
	struct nfs_fh	 fh;
	struct nfs_fattr fattr;
	long		res;

	dfprintk(VFS, "NFS: readdir(%s/%s) starting at cookie %Lu\n",
			dentry->d_parent->d_name.name, dentry->d_name.name,
			(long long)filp->f_pos);
	nfs_inc_stats(inode, NFSIOS_VFSGETDENTS);

	lock_kernel();

	res = nfs_revalidate_mapping_nolock(inode, filp->f_mapping);
	if (res < 0) {
		unlock_kernel();
		return res;
	}

	/*
	 * filp->f_pos points to the dirent entry number.
	 * *desc->dir_cookie has the cookie for the next entry. We have
	 * to either find the entry with the appropriate number or
	 * revalidate the cookie.
	 */
	memset(desc, 0, sizeof(*desc));

	desc->file = filp;
	desc->dir_cookie = &nfs_file_open_context(filp)->dir_cookie;
	desc->decode = NFS_PROTO(inode)->decode_dirent;
	desc->plus = NFS_USE_READDIRPLUS(inode);

	my_entry.cookie = my_entry.prev_cookie = 0;
	my_entry.eof = 0;
	my_entry.fh = &fh;
	my_entry.fattr = &fattr;
	nfs_fattr_init(&fattr);
	desc->entry = &my_entry;

	nfs_block_sillyrename(dentry);
	while(!desc->entry->eof) {
		res = readdir_search_pagecache(desc);

		if (res == -EBADCOOKIE) {
			/* This means either end of directory */
			if (*desc->dir_cookie && desc->entry->cookie != *desc->dir_cookie) {
				/* Or that the server has 'lost' a cookie */
				res = uncached_readdir(desc, dirent, filldir);
				if (res >= 0)
					continue;
			}
			res = 0;
			break;
		}
		if (res == -ETOOSMALL && desc->plus) {
			clear_bit(NFS_INO_ADVISE_RDPLUS, &NFS_FLAGS(inode));
			nfs_zap_caches(inode);
			desc->plus = 0;
			desc->entry->eof = 0;
			continue;
		}
		if (res < 0)
			break;

		res = nfs_do_filldir(desc, dirent, filldir);
		if (res < 0) {
			res = 0;
			break;
		}
	}
	nfs_unblock_sillyrename(dentry);
	unlock_kernel();
	if (res > 0)
		res = 0;
	dfprintk(VFS, "NFS: readdir(%s/%s) returns %ld\n",
			dentry->d_parent->d_name.name, dentry->d_name.name,
			res);
	return res;
}

static loff_t nfs_llseek_dir(struct file *filp, loff_t offset, int origin)
{
	mutex_lock(&filp->f_path.dentry->d_inode->i_mutex);
	switch (origin) {
		case 1:
			offset += filp->f_pos;
		case 0:
			if (offset >= 0)
				break;
		default:
			offset = -EINVAL;
			goto out;
	}
	if (offset != filp->f_pos) {
		filp->f_pos = offset;
		nfs_file_open_context(filp)->dir_cookie = 0;
	}
out:
	mutex_unlock(&filp->f_path.dentry->d_inode->i_mutex);
	return offset;
}

/*
 * All directory operations under NFS are synchronous, so fsync()
 * is a dummy operation.
 */
static int nfs_fsync_dir(struct file *filp, struct dentry *dentry, int datasync)
{
	dfprintk(VFS, "NFS: fsync_dir(%s/%s) datasync %d\n",
			dentry->d_parent->d_name.name, dentry->d_name.name,
			datasync);

	return 0;
}

/*
 * A check for whether or not the parent directory has changed.
 * In the case it has, we assume that the dentries are untrustworthy
 * and may need to be looked up again.
 */
static int nfs_check_verifier(struct inode *dir, struct dentry *dentry)
{
	if (IS_ROOT(dentry))
		return 1;
	if (!nfs_verify_change_attribute(dir, dentry->d_time))
		return 0;
	/* Revalidate nfsi->cache_change_attribute before we declare a match */
	if (nfs_revalidate_inode(NFS_SERVER(dir), dir) < 0)
		return 0;
	if (!nfs_verify_change_attribute(dir, dentry->d_time))
		return 0;
	return 1;
}

/*
 * Return the intent data that applies to this particular path component
 *
 * Note that the current set of intents only apply to the very last
 * component of the path.
 * We check for this using LOOKUP_CONTINUE and LOOKUP_PARENT.
 */
static inline unsigned int nfs_lookup_check_intent(struct nameidata *nd, unsigned int mask)
{
	if (nd->flags & (LOOKUP_CONTINUE|LOOKUP_PARENT))
		return 0;
	return nd->flags & mask;
}

/*
 * Use intent information to check whether or not we're going to do
 * an O_EXCL create using this path component.
 */
static int nfs_is_exclusive_create(struct inode *dir, struct nameidata *nd)
{
	if (NFS_PROTO(dir)->version == 2)
		return 0;
	if (nd == NULL || nfs_lookup_check_intent(nd, LOOKUP_CREATE) == 0)
		return 0;
	return (nd->intent.open.flags & O_EXCL) != 0;
}

/*
 * Inode and filehandle revalidation for lookups.
 *
 * We force revalidation in the cases where the VFS sets LOOKUP_REVAL,
 * or if the intent information indicates that we're about to open this
 * particular file and the "nocto" mount flag is not set.
 *
 */
static inline
int nfs_lookup_verify_inode(struct inode *inode, struct nameidata *nd)
{
	struct nfs_server *server = NFS_SERVER(inode);

	if (nd != NULL) {
		/* VFS wants an on-the-wire revalidation */
		if (nd->flags & LOOKUP_REVAL)
			goto out_force;
		/* This is an open(2) */
		if (nfs_lookup_check_intent(nd, LOOKUP_OPEN) != 0 &&
				!(server->flags & NFS_MOUNT_NOCTO) &&
				(S_ISREG(inode->i_mode) ||
				 S_ISDIR(inode->i_mode)))
			goto out_force;
		return 0;
	}
	return nfs_revalidate_inode(server, inode);
out_force:
	return __nfs_revalidate_inode(server, inode);
}

/*
 * We judge how long we want to trust negative
 * dentries by looking at the parent inode mtime.
 *
 * If parent mtime has changed, we revalidate, else we wait for a
 * period corresponding to the parent's attribute cache timeout value.
 */
static inline
int nfs_neg_need_reval(struct inode *dir, struct dentry *dentry,
		       struct nameidata *nd)
{
	/* Don't revalidate a negative dentry if we're creating a new file */
	if (nd != NULL && nfs_lookup_check_intent(nd, LOOKUP_CREATE) != 0)
		return 0;
	return !nfs_check_verifier(dir, dentry);
}

/*
 * This is called every time the dcache has a lookup hit,
 * and we should check whether we can really trust that
 * lookup.
 *
 * NOTE! The hit can be a negative hit too, don't assume
 * we have an inode!
 *
 * If the parent directory is seen to have changed, we throw out the
 * cached dentry and do a new lookup.
 */
static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd)
{
	struct inode *dir;
	struct inode *inode;
	struct dentry *parent;
	int error;
	struct nfs_fh fhandle;
	struct nfs_fattr fattr;

	parent = dget_parent(dentry);
	lock_kernel();
	dir = parent->d_inode;
	nfs_inc_stats(dir, NFSIOS_DENTRYREVALIDATE);
	inode = dentry->d_inode;

	if (!inode) {
		if (nfs_neg_need_reval(dir, dentry, nd))
			goto out_bad;
		goto out_valid;
	}

	if (is_bad_inode(inode)) {
		dfprintk(LOOKUPCACHE, "%s: %s/%s has dud inode\n",
				__FUNCTION__, dentry->d_parent->d_name.name,
				dentry->d_name.name);
		goto out_bad;
	}

	/* Force a full look up iff the parent directory has changed */
	if (!nfs_is_exclusive_create(dir, nd) && nfs_check_verifier(dir, dentry)) {
		if (nfs_lookup_verify_inode(inode, nd))
			goto out_zap_parent;
		goto out_valid;
	}

	if (NFS_STALE(inode))
		goto out_bad;

	error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, &fhandle, &fattr);
	if (error)
		goto out_bad;
	if (nfs_compare_fh(NFS_FH(inode), &fhandle))
		goto out_bad;
	if ((error = nfs_refresh_inode(inode, &fattr)) != 0)
		goto out_bad;

	nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
 out_valid:
	unlock_kernel();
	dput(parent);
	dfprintk(LOOKUPCACHE, "NFS: %s(%s/%s) is valid\n",
			__FUNCTION__, dentry->d_parent->d_name.name,
			dentry->d_name.name);
	return 1;
out_zap_parent:
	nfs_zap_caches(dir);
 out_bad:
	nfs_mark_for_revalidate(dir);
	if (inode && S_ISDIR(inode->i_mode)) {
		/* Purge readdir caches. */
		nfs_zap_caches(inode);
		/* If we have submounts, don't unhash ! */
		if (have_submounts(dentry))
			goto out_valid;
		shrink_dcache_parent(dentry);
	}
	d_drop(dentry);
	unlock_kernel();
	dput(parent);
	dfprintk(LOOKUPCACHE, "NFS: %s(%s/%s) is invalid\n",
			__FUNCTION__, dentry->d_parent->d_name.name,
			dentry->d_name.name);
	return 0;
}

/*
 * This is called from dput() when d_count is going to 0.
 */
static int nfs_dentry_delete(struct dentry *dentry)
{
	dfprintk(VFS, "NFS: dentry_delete(%s/%s, %x)\n",
		dentry->d_parent->d_name.name, dentry->d_name.name,
		dentry->d_flags);

	if (dentry->d_flags & DCACHE_NFSFS_RENAMED) {
		/* Unhash it, so that ->d_iput() would be called */
		return 1;
	}
	if (!(dentry->d_sb->s_flags & MS_ACTIVE)) {
		/* Unhash it, so that ancestors of killed async unlink
		 * files will be cleaned up during umount */
		return 1;
	}
	return 0;

}

/*
 * Called when the dentry loses inode.
 * We use it to clean up silly-renamed files.
 */
static void nfs_dentry_iput(struct dentry *dentry, struct inode *inode)
{
	nfs_inode_return_delegation(inode);
	if (S_ISDIR(inode->i_mode))
		/* drop any readdir cache as it could easily be old */
		NFS_I(inode)->cache_validity |= NFS_INO_INVALID_DATA;

	if (dentry->d_flags & DCACHE_NFSFS_RENAMED) {
		lock_kernel();
		drop_nlink(inode);
		nfs_complete_unlink(dentry, inode);
		unlock_kernel();
	}
	iput(inode);
}

struct dentry_operations nfs_dentry_operations = {
	.d_revalidate	= nfs_lookup_revalidate,
	.d_delete	= nfs_dentry_delete,
	.d_iput		= nfs_dentry_iput,
};

static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, struct nameidata *nd)
{
	struct dentry *res;
	struct dentry *parent;
	struct inode *inode = NULL;
	int error;
	struct nfs_fh fhandle;
	struct nfs_fattr fattr;

	dfprintk(VFS, "NFS: lookup(%s/%s)\n",
		dentry->d_parent->d_name.name, dentry->d_name.name);
	nfs_inc_stats(dir, NFSIOS_VFSLOOKUP);

	res = ERR_PTR(-ENAMETOOLONG);
	if (dentry->d_name.len > NFS_SERVER(dir)->namelen)
		goto out;

	res = ERR_PTR(-ENOMEM);
	dentry->d_op = NFS_PROTO(dir)->dentry_ops;

	lock_kernel();

	/*
	 * If we're doing an exclusive create, optimize away the lookup
	 * but don't hash the dentry.
	 */
	if (nfs_is_exclusive_create(dir, nd)) {
		d_instantiate(dentry, NULL);
		res = NULL;
		goto out_unlock;
	}

	parent = dentry->d_parent;
	/* Protect against concurrent sillydeletes */
	nfs_block_sillyrename(parent);
	error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, &fhandle, &fattr);
	if (error == -ENOENT)
		goto no_entry;
	if (error < 0) {
		res = ERR_PTR(error);
		goto out_unblock_sillyrename;
	}
	inode = nfs_fhget(dentry->d_sb, &fhandle, &fattr);
	res = (struct dentry *)inode;
	if (IS_ERR(res))
		goto out_unblock_sillyrename;

no_entry:
	res = d_materialise_unique(dentry, inode);
	if (res != NULL) {
		if (IS_ERR(res))
			goto out_unblock_sillyrename;
		dentry = res;
	}
	nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
out_unblock_sillyrename:
	nfs_unblock_sillyrename(parent);
out_unlock:
	unlock_kernel();
out:
	return res;
}

#ifdef CONFIG_NFS_V4
static int nfs_open_revalidate(struct dentry *, struct nameidata *);

struct dentry_operations nfs4_dentry_operations = {
	.d_revalidate	= nfs_open_revalidate,
	.d_delete	= nfs_dentry_delete,
	.d_iput		= nfs_dentry_iput,
};

/*
 * Use intent information to determine whether we need to substitute
 * the NFSv4-style stateful OPEN for the LOOKUP call
 */
static int is_atomic_open(struct inode *dir, struct nameidata *nd)
{
	if (nd == NULL || nfs_lookup_check_intent(nd, LOOKUP_OPEN) == 0)
		return 0;
	/* NFS does not (yet) have a stateful open for directories */
	if (nd->flags & LOOKUP_DIRECTORY)
		return 0;
	/* Are we trying to write to a read only partition? */
	if (IS_RDONLY(dir) && (nd->intent.open.flags & (O_CREAT|O_TRUNC|FMODE_WRITE)))
		return 0;
	return 1;
}

static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
{
	struct dentry *res = NULL;
	int error;

	dfprintk(VFS, "NFS: atomic_lookup(%s/%ld), %s\n",
			dir->i_sb->s_id, dir->i_ino, dentry->d_name.name);

	/* Check that we are indeed trying to open this file */
	if (!is_atomic_open(dir, nd))
		goto no_open;

	if (dentry->d_name.len > NFS_SERVER(dir)->namelen) {
		res = ERR_PTR(-ENAMETOOLONG);
		goto out;
	}
	dentry->d_op = NFS_PROTO(dir)->dentry_ops;

	/* Let vfs_create() deal with O_EXCL. Instantiate, but don't hash
	 * the dentry. */
	if (nd->intent.open.flags & O_EXCL) {
		d_instantiate(dentry, NULL);
		goto out;