[mirror_zfs.git] / module / zfs / zpl_inode.c

/*
 * CDDL HEADER START
 *
 * The contents of this file are subject to the terms of the
 * Common Development and Distribution License (the "License").
 * You may not use this file except in compliance with the License.
 *
 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
 * or http://www.opensolaris.org/os/licensing.
 * See the License for the specific language governing permissions
 * and limitations under the License.
 *
 * When distributing Covered Code, include this CDDL HEADER in each
 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
 * If applicable, add the following below this CDDL HEADER, with the
 * fields enclosed by brackets "[]" replaced with your own identifying
 * information: Portions Copyright [yyyy] [name of copyright owner]
 *
 * CDDL HEADER END
 */
/*
 * Copyright (c) 2011, Lawrence Livermore National Security, LLC.
 */


#include <sys/zfs_vfsops.h>
#include <sys/zfs_vnops.h>
#include <sys/zfs_znode.h>
#include <sys/vfs.h>
#include <sys/zpl.h>


static struct dentry *
#ifdef HAVE_LOOKUP_NAMEIDATA
zpl_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
#else
zpl_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags)
#endif
{
	cred_t *cr = CRED();
	struct inode *ip;
	int error;

	crhold(cr);
	error = -zfs_lookup(dir, dname(dentry), &ip, 0, cr, NULL, NULL);
	ASSERT3S(error, <=, 0);
	crfree(cr);

	spin_lock(&dentry->d_lock);
	dentry->d_time = jiffies;
#ifndef HAVE_S_D_OP
	d_set_d_op(dentry, &zpl_dentry_operations);
#endif /* HAVE_S_D_OP */
	spin_unlock(&dentry->d_lock);

	if (error) {
		if (error == -ENOENT)
			return d_splice_alias(NULL, dentry);
		else
			return ERR_PTR(error);
	}

	return d_splice_alias(ip, dentry);
}

void
zpl_vap_init(vattr_t *vap, struct inode *dir, zpl_umode_t mode, cred_t *cr)
{
	vap->va_mask = ATTR_MODE;
	vap->va_mode = mode;
	vap->va_uid = crgetfsuid(cr);

	if (dir && dir->i_mode & S_ISGID) {
		vap->va_gid = dir->i_gid;
		if (S_ISDIR(mode))
			vap->va_mode |= S_ISGID;
	} else {
		vap->va_gid = crgetfsgid(cr);
	}
}

static int
#ifdef HAVE_CREATE_NAMEIDATA
zpl_create(struct inode *dir, struct dentry *dentry, zpl_umode_t mode,
    struct nameidata *nd)
#else
zpl_create(struct inode *dir, struct dentry *dentry, zpl_umode_t mode,
    bool flag)
#endif
{
	cred_t *cr = CRED();
	struct inode *ip;
	vattr_t *vap;
	int error;

	crhold(cr);
	vap = kmem_zalloc(sizeof(vattr_t), KM_SLEEP);
	zpl_vap_init(vap, dir, mode, cr);

	error = -zfs_create(dir, dname(dentry), vap, 0, mode, &ip, cr, 0, NULL);
	if (error == 0) {
		error = zpl_xattr_security_init(ip, dir, &dentry->d_name);
		VERIFY3S(error, ==, 0);
		d_instantiate(dentry, ip);
	}

	kmem_free(vap, sizeof(vattr_t));
	crfree(cr);
	ASSERT3S(error, <=, 0);

	return (error);
}

static int
zpl_mknod(struct inode *dir, struct dentry *dentry, zpl_umode_t mode,
    dev_t rdev)
{
	cred_t *cr = CRED();
	struct inode *ip;
	vattr_t *vap;
	int error;

	/*
	 * We currently expect Linux to supply rdev=0 for all sockets
	 * and fifos, but we want to know if this behavior ever changes.
	 */
	if (S_ISSOCK(mode) || S_ISFIFO(mode))
		ASSERT(rdev == 0);

	crhold(cr);
	vap = kmem_zalloc(sizeof(vattr_t), KM_SLEEP);
	zpl_vap_init(vap, dir, mode, cr);
	vap->va_rdev = rdev;

	error = -zfs_create(dir, dname(dentry), vap, 0, mode, &ip, cr, 0, NULL);
	if (error == 0)
		d_instantiate(dentry, ip);

	kmem_free(vap, sizeof(vattr_t));
	crfree(cr);
	ASSERT3S(error, <=, 0);

	return (-error);
}

static int
zpl_unlink(struct inode *dir, struct dentry *dentry)
{
	cred_t *cr = CRED();
	int error;

	crhold(cr);
	error = -zfs_remove(dir, dname(dentry), cr);
	crfree(cr);
	ASSERT3S(error, <=, 0);

	return (error);
}

static int
zpl_mkdir(struct inode *dir, struct dentry *dentry, zpl_umode_t mode)
{
	cred_t *cr = CRED();
	vattr_t *vap;
	struct inode *ip;
	int error;

	crhold(cr);
	vap = kmem_zalloc(sizeof(vattr_t), KM_SLEEP);
	zpl_vap_init(vap, dir, mode | S_IFDIR, cr);

	error = -zfs_mkdir(dir, dname(dentry), vap, &ip, cr, 0, NULL);
	if (error == 0)
		d_instantiate(dentry, ip);

	kmem_free(vap, sizeof(vattr_t));
	crfree(cr);
	ASSERT3S(error, <=, 0);

	return (error);
}

static int
zpl_rmdir(struct inode * dir, struct dentry *dentry)
{
	cred_t *cr = CRED();
	int error;

	crhold(cr);
	error = -zfs_rmdir(dir, dname(dentry), NULL, cr, 0);
	crfree(cr);
	ASSERT3S(error, <=, 0);

	return (error);
}

static int
zpl_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
{
	boolean_t issnap = ITOZSB(dentry->d_inode)->z_issnap;
	int error;

	/*
	 * Ensure MNT_SHRINKABLE is set on snapshots to ensure they are
	 * unmounted automatically with the parent file system.  This
	 * is done on the first getattr because it's not easy to get the
	 * vfsmount structure at mount time.  This call path is explicitly
	 * marked unlikely to avoid any performance impact.  FWIW, ext4
	 * resorts to a similar trick for sysadmin convenience.
	 */
	if (unlikely(issnap && !(mnt->mnt_flags & MNT_SHRINKABLE)))
		mnt->mnt_flags |= MNT_SHRINKABLE;

	error = -zfs_getattr_fast(dentry->d_inode, stat);
	ASSERT3S(error, <=, 0);

	return (error);
}

static int
zpl_setattr(struct dentry *dentry, struct iattr *ia)
{
	cred_t *cr = CRED();
	vattr_t *vap;
	int error;

	error = inode_change_ok(dentry->d_inode, ia);
	if (error)
		return (error);

	crhold(cr);
	vap = kmem_zalloc(sizeof(vattr_t), KM_SLEEP);
	vap->va_mask = ia->ia_valid & ATTR_IATTR_MASK;
	vap->va_mode = ia->ia_mode;
	vap->va_uid = ia->ia_uid;
	vap->va_gid = ia->ia_gid;
	vap->va_size = ia->ia_size;
	vap->va_atime = ia->ia_atime;
	vap->va_mtime = ia->ia_mtime;
	vap->va_ctime = ia->ia_ctime;

	error = -zfs_setattr(dentry->d_inode, vap, 0, cr);

	kmem_free(vap, sizeof(vattr_t));
	crfree(cr);
	ASSERT3S(error, <=, 0);

	return (error);
}

static int
zpl_rename(struct inode *sdip, struct dentry *sdentry,
    struct inode *tdip, struct dentry *tdentry)
{
	cred_t *cr = CRED();
	int error;

	crhold(cr);
	error = -zfs_rename(sdip, dname(sdentry), tdip, dname(tdentry), cr, 0);
	crfree(cr);
	ASSERT3S(error, <=, 0);

	return (error);
}

static int
zpl_symlink(struct inode *dir, struct dentry *dentry, const char *name)
{
	cred_t *cr = CRED();
	vattr_t *vap;
	struct inode *ip;
	int error;

	crhold(cr);
	vap = kmem_zalloc(sizeof(vattr_t), KM_SLEEP);
	zpl_vap_init(vap, dir, S_IFLNK | S_IRWXUGO, cr);

	error = -zfs_symlink(dir, dname(dentry), vap, (char *)name, &ip, cr, 0);
	if (error == 0)
		d_instantiate(dentry, ip);

	kmem_free(vap, sizeof(vattr_t));
	crfree(cr);
	ASSERT3S(error, <=, 0);

	return (error);
}

static void *
zpl_follow_link(struct dentry *dentry, struct nameidata *nd)
{
	cred_t *cr = CRED();
	struct inode *ip = dentry->d_inode;
	struct iovec iov;
	uio_t uio;
	char *link;
	int error;

	crhold(cr);

	iov.iov_len = MAXPATHLEN;
	iov.iov_base = link = kmem_zalloc(MAXPATHLEN, KM_SLEEP);

	uio.uio_iov = &iov;
	uio.uio_iovcnt = 1;
	uio.uio_resid = (MAXPATHLEN - 1);
	uio.uio_segflg = UIO_SYSSPACE;

	error = -zfs_readlink(ip, &uio, cr);
	if (error) {
		kmem_free(link, MAXPATHLEN);
		nd_set_link(nd, ERR_PTR(error));
	} else {
		nd_set_link(nd, link);
	}

	crfree(cr);
	return (NULL);
}

static void
zpl_put_link(struct dentry *dentry, struct nameidata *nd, void *ptr)
{
	const char *link = nd_get_link(nd);

	if (!IS_ERR(link))
		kmem_free(link, MAXPATHLEN);
}

static int
zpl_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry)
{
	cred_t *cr = CRED();
	struct inode *ip = old_dentry->d_inode;
	int error;

	if (ip->i_nlink >= ZFS_LINK_MAX)
		return -EMLINK;

	crhold(cr);
	ip->i_ctime = CURRENT_TIME_SEC;
	igrab(ip); /* Use ihold() if available */

	error = -zfs_link(dir, ip, dname(dentry), cr);
	if (error) {
		iput(ip);
		goto out;
	}

	d_instantiate(dentry, ip);
out:
	crfree(cr);
	ASSERT3S(error, <=, 0);

	return (error);
}

#ifdef HAVE_INODE_TRUNCATE_RANGE
static void
zpl_truncate_range(struct inode* ip, loff_t start, loff_t end)
{
	cred_t *cr = CRED();
	flock64_t bf;

	ASSERT3S(start, <=, end);

	/*
	 * zfs_freesp() will interpret (len == 0) as meaning "truncate until
	 * the end of the file". We don't want that.
	 */
	if (start == end)
		return;

	crhold(cr);

	bf.l_type = F_WRLCK;
	bf.l_whence = 0;
	bf.l_start = start;
	bf.l_len = end - start;
	bf.l_pid = 0;
	zfs_space(ip, F_FREESP, &bf, FWRITE, start, cr);

	crfree(cr);
}
#endif /* HAVE_INODE_TRUNCATE_RANGE */

#ifdef HAVE_INODE_FALLOCATE
static long
zpl_fallocate(struct inode *ip, int mode, loff_t offset, loff_t len)
{
	return zpl_fallocate_common(ip, mode, offset, len);
}
#endif /* HAVE_INODE_FALLOCATE */

static int
#ifdef HAVE_D_REVALIDATE_NAMEIDATA
zpl_revalidate(struct dentry *dentry, struct nameidata *nd)
{
	unsigned int flags = (nd ? nd->flags : 0);
#else
zpl_revalidate(struct dentry *dentry, unsigned int flags)
{
#endif /* HAVE_D_REVALIDATE_NAMEIDATA */
	zfs_sb_t *zsb = dentry->d_sb->s_fs_info;
	int error;

	if (flags & LOOKUP_RCU)
		return (-ECHILD);

	/*
	 * After a rollback negative dentries created before the rollback
	 * time must be invalidated.  Otherwise they can obscure files which
	 * are only present in the rolled back dataset.
	 */
	if (dentry->d_inode == NULL) {
		spin_lock(&dentry->d_lock);
		error = time_before(dentry->d_time, zsb->z_rollback_time);
		spin_unlock(&dentry->d_lock);

		if (error)
			return (0);
	}

	/*
	 * The dentry may reference a stale inode if a mounted file system
	 * was rolled back to a point in time where the object didn't exist.
	 */
	if (dentry->d_inode && ITOZ(dentry->d_inode)->z_is_stale)
		return (0);

	return (1);
}

const struct inode_operations zpl_inode_operations = {
	.create		= zpl_create,
	.link		= zpl_link,
	.unlink		= zpl_unlink,
	.symlink	= zpl_symlink,
	.mkdir		= zpl_mkdir,
	.rmdir		= zpl_rmdir,
	.mknod		= zpl_mknod,
	.rename		= zpl_rename,
	.setattr	= zpl_setattr,
	.getattr	= zpl_getattr,
	.setxattr	= generic_setxattr,
	.getxattr	= generic_getxattr,
	.removexattr	= generic_removexattr,
	.listxattr	= zpl_xattr_list,
#ifdef HAVE_INODE_TRUNCATE_RANGE
	.truncate_range = zpl_truncate_range,
#endif /* HAVE_INODE_TRUNCATE_RANGE */
#ifdef HAVE_INODE_FALLOCATE
	.fallocate	= zpl_fallocate,
#endif /* HAVE_INODE_FALLOCATE */
};

const struct inode_operations zpl_dir_inode_operations = {
	.create		= zpl_create,
	.lookup		= zpl_lookup,
	.link		= zpl_link,
	.unlink		= zpl_unlink,
	.symlink	= zpl_symlink,
	.mkdir		= zpl_mkdir,
	.rmdir		= zpl_rmdir,
	.mknod		= zpl_mknod,
	.rename		= zpl_rename,
	.setattr	= zpl_setattr,
	.getattr	= zpl_getattr,
	.setxattr	= generic_setxattr,
	.getxattr	= generic_getxattr,
	.removexattr	= generic_removexattr,
	.listxattr	= zpl_xattr_list,
};

const struct inode_operations zpl_symlink_inode_operations = {
	.readlink	= generic_readlink,
	.follow_link	= zpl_follow_link,
	.put_link	= zpl_put_link,
	.setattr	= zpl_setattr,
	.getattr	= zpl_getattr,
	.setxattr	= generic_setxattr,
	.getxattr	= generic_getxattr,
	.removexattr	= generic_removexattr,
	.listxattr	= zpl_xattr_list,
};

const struct inode_operations zpl_special_inode_operations = {
	.setattr	= zpl_setattr,
	.getattr	= zpl_getattr,
	.setxattr	= generic_setxattr,
	.getxattr	= generic_getxattr,
	.removexattr	= generic_removexattr,
	.listxattr	= zpl_xattr_list,
};

dentry_operations_t zpl_dentry_operations = {
	.d_revalidate	= zpl_revalidate,
};
Commit	Line	Data
ee154f01 BB	1	/*
	2	* CDDL HEADER START
	3	*
	4	* The contents of this file are subject to the terms of the
	5	* Common Development and Distribution License (the "License").
	6	* You may not use this file except in compliance with the License.
	7	*
	8	* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
	9	* or http://www.opensolaris.org/os/licensing.
	10	* See the License for the specific language governing permissions
	11	* and limitations under the License.
	12	*
	13	* When distributing Covered Code, include this CDDL HEADER in each
	14	* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
	15	* If applicable, add the following below this CDDL HEADER, with the
	16	* fields enclosed by brackets "[]" replaced with your own identifying
	17	* information: Portions Copyright [yyyy] [name of copyright owner]
	18	*
	19	* CDDL HEADER END
	20	*/
	21	/*
	22	* Copyright (c) 2011, Lawrence Livermore National Security, LLC.
	23	*/
	24
	25
	26	#include <sys/zfs_vfsops.h>
	27	#include <sys/zfs_vnops.h>
ebe7e575	28	#include <sys/zfs_znode.h>
ee154f01 BB	29	#include <sys/vfs.h>
	30	#include <sys/zpl.h>
	31
	32
	33	static struct dentry *
8f195a90	34	#ifdef HAVE_LOOKUP_NAMEIDATA
ee154f01	35	zpl_lookup(struct inode dir, struct dentry dentry, struct nameidata *nd)
8f195a90 YS	36	#else
	37	zpl_lookup(struct inode dir, struct dentry dentry, unsigned int flags)
	38	#endif
ee154f01	39	{
81e97e21	40	cred_t *cr = CRED();
ee154f01	41	struct inode *ip;
ee154f01 BB	42	int error;
ee154f01 BB	43
81e97e21	44	crhold(cr);
ee154f01 BB	45	error = -zfs_lookup(dir, dname(dentry), &ip, 0, cr, NULL, NULL);
ee154f01 BB	46	ASSERT3S(error, <=, 0);
81e97e21	47	crfree(cr);
ee154f01	48
7b3e34ba BB	49	spin_lock(&dentry->d_lock);
7b3e34ba BB	50	dentry->d_time = jiffies;
ee930353 BB	51	#ifndef HAVE_S_D_OP
	52	d_set_d_op(dentry, &zpl_dentry_operations);
	53	#endif /* HAVE_S_D_OP */
7b3e34ba BB	54	spin_unlock(&dentry->d_lock);
7b3e34ba BB	55
ee154f01 BB	56	if (error) {
	57	if (error == -ENOENT)
	58	return d_splice_alias(NULL, dentry);
	59	else
	60	return ERR_PTR(error);
	61	}
	62
	63	return d_splice_alias(ip, dentry);
	64	}
	65
ebe7e575	66	void
7b3e34ba	67	zpl_vap_init(vattr_t vap, struct inode dir, zpl_umode_t mode, cred_t *cr)
9fd91dae BB	68	{
	69	vap->va_mask = ATTR_MODE;
	70	vap->va_mode = mode;
9fd91dae BB	71	vap->va_uid = crgetfsuid(cr);
	72
	73	if (dir && dir->i_mode & S_ISGID) {
	74	vap->va_gid = dir->i_gid;
	75	if (S_ISDIR(mode))
	76	vap->va_mode \|= S_ISGID;
	77	} else {
	78	vap->va_gid = crgetfsgid(cr);
	79	}
	80	}
	81
ee154f01	82	static int
558ef6d0	83	#ifdef HAVE_CREATE_NAMEIDATA
b39d3b9f	84	zpl_create(struct inode dir, struct dentry dentry, zpl_umode_t mode,
ee154f01	85	struct nameidata *nd)
558ef6d0 YS	86	#else
	87	zpl_create(struct inode dir, struct dentry dentry, zpl_umode_t mode,
	88	bool flag)
	89	#endif
ee154f01	90	{
81e97e21	91	cred_t *cr = CRED();
ee154f01 BB	92	struct inode *ip;
	93	vattr_t *vap;
	94	int error;
	95
81e97e21	96	crhold(cr);
ee154f01	97	vap = kmem_zalloc(sizeof(vattr_t), KM_SLEEP);
7b3e34ba	98	zpl_vap_init(vap, dir, mode, cr);
ee154f01	99
e89260a1 BB	100	error = -zfs_create(dir, dname(dentry), vap, 0, mode, &ip, cr, 0, NULL);
	101	if (error == 0) {
	102	error = zpl_xattr_security_init(ip, dir, &dentry->d_name);
	103	VERIFY3S(error, ==, 0);
7b3e34ba	104	d_instantiate(dentry, ip);
e89260a1 BB	105	}
e89260a1 BB	106
ee154f01	107	kmem_free(vap, sizeof(vattr_t));
81e97e21	108	crfree(cr);
ee154f01 BB	109	ASSERT3S(error, <=, 0);
	110
	111	return (error);
	112	}
	113
	114	static int
b39d3b9f BB	115	zpl_mknod(struct inode dir, struct dentry dentry, zpl_umode_t mode,
b39d3b9f BB	116	dev_t rdev)
ee154f01	117	{
81e97e21	118	cred_t *cr = CRED();
ee154f01 BB	119	struct inode *ip;
	120	vattr_t *vap;
	121	int error;
	122
aa6d8c10 NB	123	/*
	124	* We currently expect Linux to supply rdev=0 for all sockets
	125	* and fifos, but we want to know if this behavior ever changes.
	126	*/
	127	if (S_ISSOCK(mode) \|\| S_ISFIFO(mode))
	128	ASSERT(rdev == 0);
	129
81e97e21	130	crhold(cr);
ee154f01	131	vap = kmem_zalloc(sizeof(vattr_t), KM_SLEEP);
7b3e34ba	132	zpl_vap_init(vap, dir, mode, cr);
ee154f01	133	vap->va_rdev = rdev;
ee154f01	134
7b3e34ba	135	error = -zfs_create(dir, dname(dentry), vap, 0, mode, &ip, cr, 0, NULL);
ee930353	136	if (error == 0)
7b3e34ba	137	d_instantiate(dentry, ip);
7b3e34ba	138
ee154f01	139	kmem_free(vap, sizeof(vattr_t));
81e97e21	140	crfree(cr);
ee154f01 BB	141	ASSERT3S(error, <=, 0);
	142
	143	return (-error);
	144	}
	145
	146	static int
	147	zpl_unlink(struct inode dir, struct dentry dentry)
	148	{
81e97e21	149	cred_t *cr = CRED();
ee154f01 BB	150	int error;
ee154f01 BB	151
81e97e21	152	crhold(cr);
ee154f01	153	error = -zfs_remove(dir, dname(dentry), cr);
81e97e21	154	crfree(cr);
ee154f01 BB	155	ASSERT3S(error, <=, 0);
	156
	157	return (error);
	158	}
	159
	160	static int
b39d3b9f	161	zpl_mkdir(struct inode dir, struct dentry dentry, zpl_umode_t mode)
ee154f01	162	{
81e97e21	163	cred_t *cr = CRED();
ee154f01 BB	164	vattr_t *vap;
	165	struct inode *ip;
	166	int error;
	167
81e97e21	168	crhold(cr);
ee154f01	169	vap = kmem_zalloc(sizeof(vattr_t), KM_SLEEP);
7b3e34ba	170	zpl_vap_init(vap, dir, mode \| S_IFDIR, cr);
ee154f01 BB	171
ee154f01 BB	172	error = -zfs_mkdir(dir, dname(dentry), vap, &ip, cr, 0, NULL);
ee930353	173	if (error == 0)
7b3e34ba	174	d_instantiate(dentry, ip);
7b3e34ba	175
ee154f01	176	kmem_free(vap, sizeof(vattr_t));
81e97e21	177	crfree(cr);
ee154f01 BB	178	ASSERT3S(error, <=, 0);
	179
	180	return (error);
	181	}
	182
	183	static int
	184	zpl_rmdir(struct inode * dir, struct dentry *dentry)
	185	{
81e97e21	186	cred_t *cr = CRED();
ee154f01 BB	187	int error;
ee154f01 BB	188
81e97e21	189	crhold(cr);
ee154f01	190	error = -zfs_rmdir(dir, dname(dentry), NULL, cr, 0);
81e97e21	191	crfree(cr);
ee154f01 BB	192	ASSERT3S(error, <=, 0);
	193
	194	return (error);
	195	}
	196
	197	static int
	198	zpl_getattr(struct vfsmount mnt, struct dentry dentry, struct kstat *stat)
	199	{
ebe7e575	200	boolean_t issnap = ITOZSB(dentry->d_inode)->z_issnap;
ee154f01 BB	201	int error;
ee154f01 BB	202
ebe7e575 BB	203	/*
	204	* Ensure MNT_SHRINKABLE is set on snapshots to ensure they are
	205	* unmounted automatically with the parent file system. This
	206	* is done on the first getattr because it's not easy to get the
	207	* vfsmount structure at mount time. This call path is explicitly
	208	* marked unlikely to avoid any performance impact. FWIW, ext4
	209	* resorts to a similar trick for sysadmin convenience.
	210	*/
	211	if (unlikely(issnap && !(mnt->mnt_flags & MNT_SHRINKABLE)))
	212	mnt->mnt_flags \|= MNT_SHRINKABLE;
	213
057e8eee	214	error = -zfs_getattr_fast(dentry->d_inode, stat);
ee154f01 BB	215	ASSERT3S(error, <=, 0);
	216
	217	return (error);
	218	}
	219
	220	static int
5484965a	221	zpl_setattr(struct dentry dentry, struct iattr ia)
ee154f01	222	{
81e97e21	223	cred_t *cr = CRED();
5484965a	224	vattr_t *vap;
ee154f01 BB	225	int error;
ee154f01 BB	226
5484965a	227	error = inode_change_ok(dentry->d_inode, ia);
ee154f01 BB	228	if (error)
	229	return (error);
	230
81e97e21	231	crhold(cr);
5484965a BB	232	vap = kmem_zalloc(sizeof(vattr_t), KM_SLEEP);
	233	vap->va_mask = ia->ia_valid & ATTR_IATTR_MASK;
	234	vap->va_mode = ia->ia_mode;
	235	vap->va_uid = ia->ia_uid;
	236	vap->va_gid = ia->ia_gid;
	237	vap->va_size = ia->ia_size;
	238	vap->va_atime = ia->ia_atime;
	239	vap->va_mtime = ia->ia_mtime;
	240	vap->va_ctime = ia->ia_ctime;
	241
	242	error = -zfs_setattr(dentry->d_inode, vap, 0, cr);
	243
	244	kmem_free(vap, sizeof(vattr_t));
81e97e21	245	crfree(cr);
ee154f01 BB	246	ASSERT3S(error, <=, 0);
ee154f01 BB	247
5484965a	248	return (error);
ee154f01 BB	249	}
	250
	251	static int
	252	zpl_rename(struct inode sdip, struct dentry sdentry,
	253	struct inode tdip, struct dentry tdentry)
	254	{
81e97e21	255	cred_t *cr = CRED();
ee154f01 BB	256	int error;
ee154f01 BB	257
81e97e21	258	crhold(cr);
ee154f01	259	error = -zfs_rename(sdip, dname(sdentry), tdip, dname(tdentry), cr, 0);
81e97e21	260	crfree(cr);
ee154f01 BB	261	ASSERT3S(error, <=, 0);
	262
	263	return (error);
	264	}
	265
	266	static int
	267	zpl_symlink(struct inode dir, struct dentry dentry, const char *name)
	268	{
81e97e21	269	cred_t *cr = CRED();
ee154f01 BB	270	vattr_t *vap;
	271	struct inode *ip;
	272	int error;
	273
81e97e21	274	crhold(cr);
ee154f01	275	vap = kmem_zalloc(sizeof(vattr_t), KM_SLEEP);
7b3e34ba	276	zpl_vap_init(vap, dir, S_IFLNK \| S_IRWXUGO, cr);
ee154f01 BB	277
ee154f01 BB	278	error = -zfs_symlink(dir, dname(dentry), vap, (char *)name, &ip, cr, 0);
ee930353	279	if (error == 0)
7b3e34ba	280	d_instantiate(dentry, ip);
7b3e34ba	281
ee154f01	282	kmem_free(vap, sizeof(vattr_t));
81e97e21	283	crfree(cr);
ee154f01 BB	284	ASSERT3S(error, <=, 0);
	285
	286	return (error);
	287	}
	288
	289	static void *
	290	zpl_follow_link(struct dentry dentry, struct nameidata nd)
	291	{
81e97e21	292	cred_t *cr = CRED();
8b4f9a2d BB	293	struct inode *ip = dentry->d_inode;
	294	struct iovec iov;
	295	uio_t uio;
	296	char *link;
8b4f9a2d BB	297	int error;
8b4f9a2d BB	298
81e97e21	299	crhold(cr);
8b4f9a2d BB	300
	301	iov.iov_len = MAXPATHLEN;
	302	iov.iov_base = link = kmem_zalloc(MAXPATHLEN, KM_SLEEP);
	303
	304	uio.uio_iov = &iov;
	305	uio.uio_iovcnt = 1;
	306	uio.uio_resid = (MAXPATHLEN - 1);
	307	uio.uio_segflg = UIO_SYSSPACE;
	308
50950001	309	error = -zfs_readlink(ip, &uio, cr);
8b4f9a2d BB	310	if (error) {
	311	kmem_free(link, MAXPATHLEN);
	312	nd_set_link(nd, ERR_PTR(error));
	313	} else {
	314	nd_set_link(nd, link);
	315	}
	316
81e97e21	317	crfree(cr);
8b4f9a2d	318	return (NULL);
ee154f01 BB	319	}
	320
	321	static void
	322	zpl_put_link(struct dentry dentry, struct nameidata nd, void *ptr)
	323	{
0a6b03d3	324	const char *link = nd_get_link(nd);
ee154f01	325
ee154f01	326	if (!IS_ERR(link))
8b4f9a2d	327	kmem_free(link, MAXPATHLEN);
ee154f01 BB	328	}
	329
	330	static int
	331	zpl_link(struct dentry old_dentry, struct inode dir, struct dentry *dentry)
	332	{
81e97e21	333	cred_t *cr = CRED();
ee154f01	334	struct inode *ip = old_dentry->d_inode;
ee154f01 BB	335	int error;
	336
	337	if (ip->i_nlink >= ZFS_LINK_MAX)
	338	return -EMLINK;
	339
81e97e21	340	crhold(cr);
ee154f01 BB	341	ip->i_ctime = CURRENT_TIME_SEC;
	342	igrab(ip); /* Use ihold() if available */
	343
	344	error = -zfs_link(dir, ip, dname(dentry), cr);
	345	if (error) {
	346	iput(ip);
	347	goto out;
	348	}
	349
	350	d_instantiate(dentry, ip);
	351	out:
81e97e21	352	crfree(cr);
ee154f01 BB	353	ASSERT3S(error, <=, 0);
	354
	355	return (error);
	356	}
	357
ea1fdf46	358	#ifdef HAVE_INODE_TRUNCATE_RANGE
5cb63a57 ED	359	static void
	360	zpl_truncate_range(struct inode* ip, loff_t start, loff_t end)
	361	{
	362	cred_t *cr = CRED();
	363	flock64_t bf;
	364
	365	ASSERT3S(start, <=, end);
	366
	367	/*
	368	* zfs_freesp() will interpret (len == 0) as meaning "truncate until
	369	* the end of the file". We don't want that.
	370	*/
	371	if (start == end)
	372	return;
	373
	374	crhold(cr);
	375
	376	bf.l_type = F_WRLCK;
	377	bf.l_whence = 0;
	378	bf.l_start = start;
	379	bf.l_len = end - start;
	380	bf.l_pid = 0;
	381	zfs_space(ip, F_FREESP, &bf, FWRITE, start, cr);
	382
	383	crfree(cr);
	384	}
ea1fdf46	385	#endif /* HAVE_INODE_TRUNCATE_RANGE */
5cb63a57	386
cb2d1901 ED	387	#ifdef HAVE_INODE_FALLOCATE
	388	static long
	389	zpl_fallocate(struct inode *ip, int mode, loff_t offset, loff_t len)
	390	{
	391	return zpl_fallocate_common(ip, mode, offset, len);
	392	}
	393	#endif /* HAVE_INODE_FALLOCATE */
	394
7b3e34ba BB	395	static int
	396	#ifdef HAVE_D_REVALIDATE_NAMEIDATA
	397	zpl_revalidate(struct dentry dentry, struct nameidata nd)
	398	{
09a661e9	399	unsigned int flags = (nd ? nd->flags : 0);
7b3e34ba BB	400	#else
	401	zpl_revalidate(struct dentry *dentry, unsigned int flags)
	402	{
	403	#endif /* HAVE_D_REVALIDATE_NAMEIDATA */
	404	zfs_sb_t *zsb = dentry->d_sb->s_fs_info;
	405	int error;
	406
	407	if (flags & LOOKUP_RCU)
	408	return (-ECHILD);
	409
	410	/*
	411	* After a rollback negative dentries created before the rollback
	412	* time must be invalidated. Otherwise they can obscure files which
	413	* are only present in the rolled back dataset.
	414	*/
	415	if (dentry->d_inode == NULL) {
	416	spin_lock(&dentry->d_lock);
	417	error = time_before(dentry->d_time, zsb->z_rollback_time);
	418	spin_unlock(&dentry->d_lock);
	419
	420	if (error)
	421	return (0);
	422	}
	423
	424	/*
	425	* The dentry may reference a stale inode if a mounted file system
	426	* was rolled back to a point in time where the object didn't exist.
	427	*/
	428	if (dentry->d_inode && ITOZ(dentry->d_inode)->z_is_stale)
	429	return (0);
	430
	431	return (1);
	432	}
cb2d1901	433
ee154f01	434	const struct inode_operations zpl_inode_operations = {
ee154f01 BB	435	.create = zpl_create,
	436	.link = zpl_link,
	437	.unlink = zpl_unlink,
	438	.symlink = zpl_symlink,
	439	.mkdir = zpl_mkdir,
	440	.rmdir = zpl_rmdir,
	441	.mknod = zpl_mknod,
	442	.rename = zpl_rename,
	443	.setattr = zpl_setattr,
	444	.getattr = zpl_getattr,
	445	.setxattr = generic_setxattr,
	446	.getxattr = generic_getxattr,
	447	.removexattr = generic_removexattr,
	448	.listxattr = zpl_xattr_list,
ea1fdf46	449	#ifdef HAVE_INODE_TRUNCATE_RANGE
5cb63a57	450	.truncate_range = zpl_truncate_range,
ea1fdf46	451	#endif /* HAVE_INODE_TRUNCATE_RANGE */
cb2d1901 ED	452	#ifdef HAVE_INODE_FALLOCATE
	453	.fallocate = zpl_fallocate,
	454	#endif /* HAVE_INODE_FALLOCATE */
ee154f01 BB	455	};
	456
	457	const struct inode_operations zpl_dir_inode_operations = {
ee154f01 BB	458	.create = zpl_create,
	459	.lookup = zpl_lookup,
	460	.link = zpl_link,
	461	.unlink = zpl_unlink,
	462	.symlink = zpl_symlink,
	463	.mkdir = zpl_mkdir,
	464	.rmdir = zpl_rmdir,
	465	.mknod = zpl_mknod,
	466	.rename = zpl_rename,
	467	.setattr = zpl_setattr,
a6695d83 BB	468	.getattr = zpl_getattr,
	469	.setxattr = generic_setxattr,
	470	.getxattr = generic_getxattr,
	471	.removexattr = generic_removexattr,
	472	.listxattr = zpl_xattr_list,
ee154f01 BB	473	};
	474
	475	const struct inode_operations zpl_symlink_inode_operations = {
ee154f01 BB	476	.readlink = generic_readlink,
	477	.follow_link = zpl_follow_link,
	478	.put_link = zpl_put_link,
6f2255ba BB	479	.setattr = zpl_setattr,
6f2255ba BB	480	.getattr = zpl_getattr,
f31b3ebe BB	481	.setxattr = generic_setxattr,
	482	.getxattr = generic_getxattr,
	483	.removexattr = generic_removexattr,
	484	.listxattr = zpl_xattr_list,
ee154f01 BB	485	};
	486
	487	const struct inode_operations zpl_special_inode_operations = {
a6695d83 BB	488	.setattr = zpl_setattr,
	489	.getattr = zpl_getattr,
	490	.setxattr = generic_setxattr,
	491	.getxattr = generic_getxattr,
	492	.removexattr = generic_removexattr,
	493	.listxattr = zpl_xattr_list,
ee154f01	494	};
7b3e34ba BB	495
	496	dentry_operations_t zpl_dentry_operations = {
	497	.d_revalidate = zpl_revalidate,
	498	};