[mirror_zfs-debian.git] / module / zfs / zpl_super.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/zfs_ctldir.h>
#include <sys/zpl.h>


static struct inode *
zpl_inode_alloc(struct super_block *sb)
{
	struct inode *ip;

	VERIFY3S(zfs_inode_alloc(sb, &ip), ==, 0);
	ip->i_version = 1;

	return (ip);
}

static void
zpl_inode_destroy(struct inode *ip)
{
        ASSERT(atomic_read(&ip->i_count) == 0);
	zfs_inode_destroy(ip);
}

/*
 * When ->drop_inode() is called its return value indicates if the
 * inode should be evicted from the inode cache.  If the inode is
 * unhashed and has no links the default policy is to evict it
 * immediately.
 *
 * Prior to 2.6.36 this eviction was accomplished by the vfs calling
 * ->delete_inode().  It was ->delete_inode()'s responsibility to
 * truncate the inode pages and call clear_inode().  The call to
 * clear_inode() synchronously invalidates all the buffers and
 * calls ->clear_inode().  It was ->clear_inode()'s responsibility
 * to cleanup and filesystem specific data before freeing the inode.
 *
 * This elaborate mechanism was replaced by ->evict_inode() which
 * does the job of both ->delete_inode() and ->clear_inode().  It
 * will be called exactly once, and when it returns the inode must
 * be in a state where it can simply be freed.i
 *
 * The ->evict_inode() callback must minimally truncate the inode pages,
 * and call clear_inode().  For 2.6.35 and later kernels this will
 * simply update the inode state, with the sync occurring before the
 * truncate in evict().  For earlier kernels clear_inode() maps to
 * end_writeback() which is responsible for completing all outstanding
 * write back.  In either case, once this is done it is safe to cleanup
 * any remaining inode specific data via zfs_inactive().
 * remaining filesystem specific data.
 */
#ifdef HAVE_EVICT_INODE
static void
zpl_evict_inode(struct inode *ip)
{
	truncate_setsize(ip, 0);
	clear_inode(ip);
	zfs_inactive(ip);
}

#else

static void
zpl_clear_inode(struct inode *ip)
{
	zfs_inactive(ip);
}

static void
zpl_inode_delete(struct inode *ip)
{
	truncate_setsize(ip, 0);
	clear_inode(ip);
}

#endif /* HAVE_EVICT_INODE */

static void
zpl_put_super(struct super_block *sb)
{
	int error;

	error = -zfs_umount(sb);
	ASSERT3S(error, <=, 0);
}

static int
zpl_sync_fs(struct super_block *sb, int wait)
{
	cred_t *cr = CRED();
	int error;

	crhold(cr);
	error = -zfs_sync(sb, wait, cr);
	crfree(cr);
	ASSERT3S(error, <=, 0);

	return (error);
}

static int
zpl_statfs(struct dentry *dentry, struct kstatfs *statp)
{
	int error;

	error = -zfs_statvfs(dentry, statp);
	ASSERT3S(error, <=, 0);

	return (error);
}

static int
zpl_remount_fs(struct super_block *sb, int *flags, char *data)
{
	int error;
	error = -zfs_remount(sb, flags, data);
	ASSERT3S(error, <=, 0);

	return (error);
}

static void
zpl_umount_begin(struct super_block *sb)
{
	zfs_sb_t *zsb = sb->s_fs_info;
	int count;

	/*
	 * Best effort to unmount snapshots in .zfs/snapshot/.  Normally this
	 * isn't required because snapshots have the MNT_SHRINKABLE flag set.
	 */
	if (zsb->z_ctldir)
		(void) zfsctl_unmount_snapshots(zsb, MNT_FORCE, &count);
}

/*
 * The Linux VFS automatically handles the following flags:
 * MNT_NOSUID, MNT_NODEV, MNT_NOEXEC, MNT_NOATIME, MNT_READONLY
 */
#ifdef HAVE_SHOW_OPTIONS_WITH_DENTRY
static int
zpl_show_options(struct seq_file *seq, struct dentry *root)
{
	zfs_sb_t *zsb = root->d_sb->s_fs_info;

	seq_printf(seq, ",%s", zsb->z_flags & ZSB_XATTR ? "xattr" : "noxattr");

	return (0);
}
#else
static int
zpl_show_options(struct seq_file *seq, struct vfsmount *vfsp)
{
	zfs_sb_t *zsb = vfsp->mnt_sb->s_fs_info;

	seq_printf(seq, ",%s", zsb->z_flags & ZSB_XATTR ? "xattr" : "noxattr");

	return (0);
}
#endif /* HAVE_SHOW_OPTIONS_WITH_DENTRY */

static int
zpl_fill_super(struct super_block *sb, void *data, int silent)
{
	int error;

	error = -zfs_domount(sb, data, silent);
	ASSERT3S(error, <=, 0);

	return (error);
}

#ifdef HAVE_MOUNT_NODEV
static struct dentry *
zpl_mount(struct file_system_type *fs_type, int flags,
    const char *osname, void *data)
{
	zpl_mount_data_t zmd = { osname, data };

	return mount_nodev(fs_type, flags, &zmd, zpl_fill_super);
}
#else
static int
zpl_get_sb(struct file_system_type *fs_type, int flags,
    const char *osname, void *data, struct vfsmount *mnt)
{
	zpl_mount_data_t zmd = { osname, data };

	return get_sb_nodev(fs_type, flags, &zmd, zpl_fill_super, mnt);
}
#endif /* HAVE_MOUNT_NODEV */

static void
zpl_kill_sb(struct super_block *sb)
{
	zfs_preumount(sb);
	kill_anon_super(sb);
}

#ifdef HAVE_SHRINK
/*
 * Linux 3.1 - 3.x API
 *
 * The Linux 3.1 API introduced per-sb cache shrinkers to replace the
 * global ones.  This allows us a mechanism to cleanly target a specific
 * zfs file system when the dnode and inode caches grow too large.
 *
 * In addition, the 3.0 kernel added the iterate_supers_type() helper
 * function which is used to safely walk all of the zfs file systems.
 */
static void
zpl_prune_sb(struct super_block *sb, void *arg)
{
	int objects = 0;
	int error;

	error = -zfs_sb_prune(sb, *(unsigned long *)arg, &objects);
	ASSERT3S(error, <=, 0);

	return;
}

void
zpl_prune_sbs(int64_t bytes_to_scan, void *private)
{
	unsigned long nr_to_scan = (bytes_to_scan / sizeof(znode_t));

	iterate_supers_type(&zpl_fs_type, zpl_prune_sb, &nr_to_scan);
	kmem_reap();
}
#else
/*
 * Linux 2.6.x - 3.0 API
 *
 * These are best effort interfaces are provided by the SPL to induce
 * the Linux VM subsystem to reclaim a fraction of the both dnode and
 * inode caches.  Ideally, we want to just target the zfs file systems
 * however our only option is to reclaim from them all.
 */
void
zpl_prune_sbs(int64_t bytes_to_scan, void *private)
{
	unsigned long nr_to_scan = (bytes_to_scan / sizeof(znode_t));

        shrink_dcache_memory(nr_to_scan, GFP_KERNEL);
        shrink_icache_memory(nr_to_scan, GFP_KERNEL);
        kmem_reap();
}
#endif /* HAVE_SHRINK */

#ifdef HAVE_NR_CACHED_OBJECTS
static int
zpl_nr_cached_objects(struct super_block *sb)
{
	zfs_sb_t *zsb = sb->s_fs_info;
	int nr;

	mutex_enter(&zsb->z_znodes_lock);
	nr = zsb->z_nr_znodes;
	mutex_exit(&zsb->z_znodes_lock);

	return (nr);
}
#endif /* HAVE_NR_CACHED_OBJECTS */

#ifdef HAVE_FREE_CACHED_OBJECTS
/*
 * Attempt to evict some meta data from the cache.  The ARC operates in
 * terms of bytes while the Linux VFS uses objects.  Now because this is
 * just a best effort eviction and the exact values aren't critical so we
 * extrapolate from an object count to a byte size using the znode_t size.
 */
static void
zpl_free_cached_objects(struct super_block *sb, int nr_to_scan)
{
	arc_adjust_meta(nr_to_scan * sizeof(znode_t), B_FALSE);
}
#endif /* HAVE_FREE_CACHED_OBJECTS */

const struct super_operations zpl_super_operations = {
	.alloc_inode		= zpl_inode_alloc,
	.destroy_inode		= zpl_inode_destroy,
	.dirty_inode		= NULL,
	.write_inode		= NULL,
	.drop_inode		= NULL,
#ifdef HAVE_EVICT_INODE
	.evict_inode		= zpl_evict_inode,
#else
	.clear_inode		= zpl_clear_inode,
	.delete_inode		= zpl_inode_delete,
#endif /* HAVE_EVICT_INODE */
	.put_super		= zpl_put_super,
	.write_super		= NULL,
	.sync_fs		= zpl_sync_fs,
	.statfs			= zpl_statfs,
	.remount_fs		= zpl_remount_fs,
	.umount_begin		= zpl_umount_begin,
	.show_options		= zpl_show_options,
	.show_stats		= NULL,
#ifdef HAVE_NR_CACHED_OBJECTS
	.nr_cached_objects	= zpl_nr_cached_objects,
#endif /* HAVE_NR_CACHED_OBJECTS */
#ifdef HAVE_FREE_CACHED_OBJECTS
	.free_cached_objects	= zpl_free_cached_objects,
#endif /* HAVE_FREE_CACHED_OBJECTS */
};

struct file_system_type zpl_fs_type = {
	.owner			= THIS_MODULE,
	.name			= ZFS_DRIVER,
#ifdef HAVE_MOUNT_NODEV
	.mount			= zpl_mount,
#else
	.get_sb			= zpl_get_sb,
#endif /* HAVE_MOUNT_NODEV */
	.kill_sb		= zpl_kill_sb,
};
Commit	Line	Data
51f0bbe4 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>
	28	#include <sys/zfs_znode.h>
ebe7e575	29	#include <sys/zfs_ctldir.h>
51f0bbe4 BB	30	#include <sys/zpl.h>
	31
	32
	33	static struct inode *
	34	zpl_inode_alloc(struct super_block *sb)
	35	{
	36	struct inode *ip;
	37
	38	VERIFY3S(zfs_inode_alloc(sb, &ip), ==, 0);
	39	ip->i_version = 1;
	40
	41	return (ip);
	42	}
	43
	44	static void
	45	zpl_inode_destroy(struct inode *ip)
	46	{
	47	ASSERT(atomic_read(&ip->i_count) == 0);
	48	zfs_inode_destroy(ip);
	49	}
	50
2c395def BB	51	/*
	52	* When ->drop_inode() is called its return value indicates if the
	53	* inode should be evicted from the inode cache. If the inode is
	54	* unhashed and has no links the default policy is to evict it
	55	* immediately.
	56	*
	57	* Prior to 2.6.36 this eviction was accomplished by the vfs calling
	58	* ->delete_inode(). It was ->delete_inode()'s responsibility to
	59	* truncate the inode pages and call clear_inode(). The call to
	60	* clear_inode() synchronously invalidates all the buffers and
	61	* calls ->clear_inode(). It was ->clear_inode()'s responsibility
	62	* to cleanup and filesystem specific data before freeing the inode.
	63	*
	64	* This elaborate mechanism was replaced by ->evict_inode() which
	65	* does the job of both ->delete_inode() and ->clear_inode(). It
	66	* will be called exactly once, and when it returns the inode must
739a1a82 RY	67	* be in a state where it can simply be freed.i
	68	*
	69	* The ->evict_inode() callback must minimally truncate the inode pages,
	70	* and call clear_inode(). For 2.6.35 and later kernels this will
	71	* simply update the inode state, with the sync occurring before the
	72	* truncate in evict(). For earlier kernels clear_inode() maps to
	73	* end_writeback() which is responsible for completing all outstanding
	74	* write back. In either case, once this is done it is safe to cleanup
	75	* any remaining inode specific data via zfs_inactive().
2c395def BB	76	* remaining filesystem specific data.
	77	*/
	78	#ifdef HAVE_EVICT_INODE
51f0bbe4	79	static void
2c395def	80	zpl_evict_inode(struct inode *ip)
51f0bbe4	81	{
b3129792	82	truncate_setsize(ip, 0);
739a1a82	83	clear_inode(ip);
2c395def	84	zfs_inactive(ip);
51f0bbe4 BB	85	}
51f0bbe4 BB	86
2c395def BB	87	#else
2c395def BB	88
51f0bbe4	89	static void
2c395def	90	zpl_clear_inode(struct inode *ip)
51f0bbe4 BB	91	{
	92	zfs_inactive(ip);
	93	}
	94
2c395def BB	95	static void
	96	zpl_inode_delete(struct inode *ip)
	97	{
b3129792	98	truncate_setsize(ip, 0);
2c395def BB	99	clear_inode(ip);
	100	}
	101
	102	#endif /* HAVE_EVICT_INODE */
	103
51f0bbe4 BB	104	static void
	105	zpl_put_super(struct super_block *sb)
	106	{
	107	int error;
	108
	109	error = -zfs_umount(sb);
	110	ASSERT3S(error, <=, 0);
	111	}
	112
03f9ba9d BB	113	static int
	114	zpl_sync_fs(struct super_block *sb, int wait)
	115	{
0d3ac5e7	116	cred_t *cr = CRED();
03f9ba9d BB	117	int error;
03f9ba9d BB	118
0d3ac5e7	119	crhold(cr);
03f9ba9d	120	error = -zfs_sync(sb, wait, cr);
0d3ac5e7	121	crfree(cr);
03f9ba9d BB	122	ASSERT3S(error, <=, 0);
	123
	124	return (error);
	125	}
	126
51f0bbe4 BB	127	static int
	128	zpl_statfs(struct dentry dentry, struct kstatfs statp)
	129	{
	130	int error;
	131
	132	error = -zfs_statvfs(dentry, statp);
	133	ASSERT3S(error, <=, 0);
	134
	135	return (error);
	136	}
	137
0de19dad BB	138	static int
	139	zpl_remount_fs(struct super_block sb, int flags, char *data)
	140	{
	141	int error;
	142	error = -zfs_remount(sb, flags, data);
	143	ASSERT3S(error, <=, 0);
	144
	145	return (error);
	146	}
	147
ebe7e575 BB	148	static void
	149	zpl_umount_begin(struct super_block *sb)
	150	{
	151	zfs_sb_t *zsb = sb->s_fs_info;
	152	int count;
	153
	154	/*
	155	* Best effort to unmount snapshots in .zfs/snapshot/. Normally this
	156	* isn't required because snapshots have the MNT_SHRINKABLE flag set.
	157	*/
	158	if (zsb->z_ctldir)
	159	(void) zfsctl_unmount_snapshots(zsb, MNT_FORCE, &count);
	160	}
	161
47621f3d BB	162	/*
	163	* The Linux VFS automatically handles the following flags:
	164	* MNT_NOSUID, MNT_NODEV, MNT_NOEXEC, MNT_NOATIME, MNT_READONLY
	165	*/
	166	#ifdef HAVE_SHOW_OPTIONS_WITH_DENTRY
51f0bbe4	167	static int
47621f3d	168	zpl_show_options(struct seq_file seq, struct dentry root)
51f0bbe4	169	{
47621f3d	170	zfs_sb_t *zsb = root->d_sb->s_fs_info;
51f0bbe4	171
47621f3d BB	172	seq_printf(seq, ",%s", zsb->z_flags & ZSB_XATTR ? "xattr" : "noxattr");
	173
	174	return (0);
	175	}
	176	#else
	177	static int
	178	zpl_show_options(struct seq_file seq, struct vfsmount vfsp)
	179	{
	180	zfs_sb_t *zsb = vfsp->mnt_sb->s_fs_info;
51f0bbe4	181
2cf7f52b	182	seq_printf(seq, ",%s", zsb->z_flags & ZSB_XATTR ? "xattr" : "noxattr");
51f0bbe4 BB	183
	184	return (0);
	185	}
47621f3d	186	#endif /* HAVE_SHOW_OPTIONS_WITH_DENTRY */
51f0bbe4 BB	187
	188	static int
	189	zpl_fill_super(struct super_block sb, void data, int silent)
	190	{
	191	int error;
	192
	193	error = -zfs_domount(sb, data, silent);
	194	ASSERT3S(error, <=, 0);
	195
	196	return (error);
	197	}
	198
2cf7f52b BB	199	#ifdef HAVE_MOUNT_NODEV
	200	static struct dentry *
	201	zpl_mount(struct file_system_type *fs_type, int flags,
	202	const char osname, void data)
	203	{
	204	zpl_mount_data_t zmd = { osname, data };
	205
	206	return mount_nodev(fs_type, flags, &zmd, zpl_fill_super);
	207	}
	208	#else
51f0bbe4 BB	209	static int
	210	zpl_get_sb(struct file_system_type *fs_type, int flags,
	211	const char osname, void data, struct vfsmount *mnt)
	212	{
2cf7f52b	213	zpl_mount_data_t zmd = { osname, data };
51f0bbe4 BB	214
	215	return get_sb_nodev(fs_type, flags, &zmd, zpl_fill_super, mnt);
	216	}
2cf7f52b	217	#endif /* HAVE_MOUNT_NODEV */
51f0bbe4 BB	218
	219	static void
	220	zpl_kill_sb(struct super_block *sb)
	221	{
ebe7e575	222	zfs_preumount(sb);
51f0bbe4 BB	223	kill_anon_super(sb);
	224	}
	225
ab26409d BB	226	#ifdef HAVE_SHRINK
	227	/*
	228	* Linux 3.1 - 3.x API
	229	*
	230	* The Linux 3.1 API introduced per-sb cache shrinkers to replace the
	231	* global ones. This allows us a mechanism to cleanly target a specific
	232	* zfs file system when the dnode and inode caches grow too large.
	233	*
	234	* In addition, the 3.0 kernel added the iterate_supers_type() helper
	235	* function which is used to safely walk all of the zfs file systems.
	236	*/
	237	static void
	238	zpl_prune_sb(struct super_block sb, void arg)
	239	{
	240	int objects = 0;
	241	int error;
	242
	243	error = -zfs_sb_prune(sb, (unsigned long )arg, &objects);
	244	ASSERT3S(error, <=, 0);
	245
	246	return;
	247	}
	248
	249	void
	250	zpl_prune_sbs(int64_t bytes_to_scan, void *private)
	251	{
	252	unsigned long nr_to_scan = (bytes_to_scan / sizeof(znode_t));
	253
	254	iterate_supers_type(&zpl_fs_type, zpl_prune_sb, &nr_to_scan);
	255	kmem_reap();
	256	}
	257	#else
	258	/*
	259	* Linux 2.6.x - 3.0 API
	260	*
	261	* These are best effort interfaces are provided by the SPL to induce
	262	* the Linux VM subsystem to reclaim a fraction of the both dnode and
	263	* inode caches. Ideally, we want to just target the zfs file systems
	264	* however our only option is to reclaim from them all.
	265	*/
	266	void
	267	zpl_prune_sbs(int64_t bytes_to_scan, void *private)
	268	{
	269	unsigned long nr_to_scan = (bytes_to_scan / sizeof(znode_t));
	270
	271	shrink_dcache_memory(nr_to_scan, GFP_KERNEL);
	272	shrink_icache_memory(nr_to_scan, GFP_KERNEL);
	273	kmem_reap();
	274	}
	275	#endif /* HAVE_SHRINK */
	276
	277	#ifdef HAVE_NR_CACHED_OBJECTS
	278	static int
	279	zpl_nr_cached_objects(struct super_block *sb)
	280	{
	281	zfs_sb_t *zsb = sb->s_fs_info;
	282	int nr;
	283
	284	mutex_enter(&zsb->z_znodes_lock);
	285	nr = zsb->z_nr_znodes;
	286	mutex_exit(&zsb->z_znodes_lock);
	287
	288	return (nr);
	289	}
290	#endif /* HAVE_NR_CACHED_OBJECTS */
291
292	#ifdef HAVE_FREE_CACHED_OBJECTS
293	/*
294	* Attempt to evict some meta data from the cache. The ARC operates in
295	* terms of bytes while the Linux VFS uses objects. Now because this is
296	* just a best effort eviction and the exact values aren't critical so we
297	* extrapolate from an object count to a byte size using the znode_t size.
298	*/
299	static void
300	zpl_free_cached_objects(struct super_block *sb, int nr_to_scan)
301	{
302	arc_adjust_meta(nr_to_scan * sizeof(znode_t), B_FALSE);
303	}
304	#endif /* HAVE_FREE_CACHED_OBJECTS */
305
51f0bbe4	306	const struct super_operations zpl_super_operations = {
ab26409d BB	307	.alloc_inode = zpl_inode_alloc,
	308	.destroy_inode = zpl_inode_destroy,
	309	.dirty_inode = NULL,
	310	.write_inode = NULL,
	311	.drop_inode = NULL,
2c395def	312	#ifdef HAVE_EVICT_INODE
ab26409d	313	.evict_inode = zpl_evict_inode,
2c395def	314	#else
ab26409d BB	315	.clear_inode = zpl_clear_inode,
ab26409d BB	316	.delete_inode = zpl_inode_delete,
2c395def	317	#endif /* HAVE_EVICT_INODE */
ab26409d BB	318	.put_super = zpl_put_super,
	319	.write_super = NULL,
	320	.sync_fs = zpl_sync_fs,
	321	.statfs = zpl_statfs,
	322	.remount_fs = zpl_remount_fs,
ebe7e575	323	.umount_begin = zpl_umount_begin,
ab26409d BB	324	.show_options = zpl_show_options,
	325	.show_stats = NULL,
	326	#ifdef HAVE_NR_CACHED_OBJECTS
	327	.nr_cached_objects = zpl_nr_cached_objects,
	328	#endif /* HAVE_NR_CACHED_OBJECTS */
	329	#ifdef HAVE_FREE_CACHED_OBJECTS
	330	.free_cached_objects = zpl_free_cached_objects,
	331	#endif /* HAVE_FREE_CACHED_OBJECTS */
51f0bbe4 BB	332	};
51f0bbe4 BB	333
51f0bbe4	334	struct file_system_type zpl_fs_type = {
ab26409d BB	335	.owner = THIS_MODULE,
ab26409d BB	336	.name = ZFS_DRIVER,
2cf7f52b	337	#ifdef HAVE_MOUNT_NODEV
ab26409d	338	.mount = zpl_mount,
2cf7f52b	339	#else
ab26409d	340	.get_sb = zpl_get_sb,
2cf7f52b	341	#endif /* HAVE_MOUNT_NODEV */
ab26409d	342	.kill_sb = zpl_kill_sb,
51f0bbe4	343	};