[net-next-2.6.git] / fs / ceph / export.c

#include "ceph_debug.h"

#include <linux/exportfs.h>
#include <linux/slab.h>
#include <asm/unaligned.h>

#include "super.h"

/*
 * NFS export support
 *
 * NFS re-export of a ceph mount is, at present, only semireliable.
 * The basic issue is that the Ceph architectures doesn't lend itself
 * well to generating filehandles that will remain valid forever.
 *
 * So, we do our best.  If you're lucky, your inode will be in the
 * client's cache.  If it's not, and you have a connectable fh, then
 * the MDS server may be able to find it for you.  Otherwise, you get
 * ESTALE.
 *
 * There are ways to this more reliable, but in the non-connectable fh
 * case, we won't every work perfectly, and in the connectable case,
 * some changes are needed on the MDS side to work better.
 */

/*
 * Basic fh
 */
struct ceph_nfs_fh {
	u64 ino;
} __attribute__ ((packed));

/*
 * Larger 'connectable' fh that includes parent ino and name hash.
 * Use this whenever possible, as it works more reliably.
 */
struct ceph_nfs_confh {
	u64 ino, parent_ino;
	u32 parent_name_hash;
} __attribute__ ((packed));

static int ceph_encode_fh(struct dentry *dentry, u32 *rawfh, int *max_len,
			  int connectable)
{
	int type;
	struct ceph_nfs_fh *fh = (void *)rawfh;
	struct ceph_nfs_confh *cfh = (void *)rawfh;
	struct dentry *parent = dentry->d_parent;
	struct inode *inode = dentry->d_inode;
	int connected_handle_length = sizeof(*cfh)/4;
	int handle_length = sizeof(*fh)/4;

	/* don't re-export snaps */
	if (ceph_snap(inode) != CEPH_NOSNAP)
		return -EINVAL;

	if (*max_len >= connected_handle_length) {
		dout("encode_fh %p connectable\n", dentry);
		cfh->ino = ceph_ino(dentry->d_inode);
		cfh->parent_ino = ceph_ino(parent->d_inode);
		cfh->parent_name_hash = parent->d_name.hash;
		*max_len = connected_handle_length;
		type = 2;
	} else if (*max_len >= handle_length) {
		if (connectable)
			return 255;
		dout("encode_fh %p\n", dentry);
		fh->ino = ceph_ino(dentry->d_inode);
		*max_len = handle_length;
		type = 1;
	} else {
		return 255;
	}
	return type;
}

/*
 * convert regular fh to dentry
 *
 * FIXME: we should try harder by querying the mds for the ino.
 */
static struct dentry *__fh_to_dentry(struct super_block *sb,
				     struct ceph_nfs_fh *fh)
{
	struct inode *inode;
	struct dentry *dentry;
	struct ceph_vino vino;
	int err;

	dout("__fh_to_dentry %llx\n", fh->ino);
	vino.ino = fh->ino;
	vino.snap = CEPH_NOSNAP;
	inode = ceph_find_inode(sb, vino);
	if (!inode)
		return ERR_PTR(-ESTALE);

	dentry = d_obtain_alias(inode);
	if (IS_ERR(dentry)) {
		pr_err("fh_to_dentry %llx -- inode %p but ENOMEM\n",
		       fh->ino, inode);
		iput(inode);
		return dentry;
	}
	err = ceph_init_dentry(dentry);

	if (err < 0) {
		iput(inode);
		return ERR_PTR(err);
	}
	dout("__fh_to_dentry %llx %p dentry %p\n", fh->ino, inode, dentry);
	return dentry;
}

/*
 * convert connectable fh to dentry
 */
static struct dentry *__cfh_to_dentry(struct super_block *sb,
				      struct ceph_nfs_confh *cfh)
{
	struct ceph_mds_client *mdsc = &ceph_sb_to_client(sb)->mdsc;
	struct inode *inode;
	struct dentry *dentry;
	struct ceph_vino vino;
	int err;

	dout("__cfh_to_dentry %llx (%llx/%x)\n",
	     cfh->ino, cfh->parent_ino, cfh->parent_name_hash);

	vino.ino = cfh->ino;
	vino.snap = CEPH_NOSNAP;
	inode = ceph_find_inode(sb, vino);
	if (!inode) {
		struct ceph_mds_request *req;

		req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPHASH,
					       USE_ANY_MDS);
		if (IS_ERR(req))
			return ERR_CAST(req);

		req->r_ino1 = vino;
		req->r_ino2.ino = cfh->parent_ino;
		req->r_ino2.snap = CEPH_NOSNAP;
		req->r_path2 = kmalloc(16, GFP_NOFS);
		snprintf(req->r_path2, 16, "%d", cfh->parent_name_hash);
		req->r_num_caps = 1;
		err = ceph_mdsc_do_request(mdsc, NULL, req);
		ceph_mdsc_put_request(req);
		inode = ceph_find_inode(sb, vino);
		if (!inode)
			return ERR_PTR(err ? err : -ESTALE);
	}

	dentry = d_obtain_alias(inode);
	if (IS_ERR(dentry)) {
		pr_err("cfh_to_dentry %llx -- inode %p but ENOMEM\n",
		       cfh->ino, inode);
		iput(inode);
		return dentry;
	}
	err = ceph_init_dentry(dentry);
	if (err < 0) {
		iput(inode);
		return ERR_PTR(err);
	}
	dout("__cfh_to_dentry %llx %p dentry %p\n", cfh->ino, inode, dentry);
	return dentry;
}

static struct dentry *ceph_fh_to_dentry(struct super_block *sb, struct fid *fid,
					int fh_len, int fh_type)
{
	if (fh_type == 1)
		return __fh_to_dentry(sb, (struct ceph_nfs_fh *)fid->raw);
	else
		return __cfh_to_dentry(sb, (struct ceph_nfs_confh *)fid->raw);
}

/*
 * get parent, if possible.
 *
 * FIXME: we could do better by querying the mds to discover the
 * parent.
 */
static struct dentry *ceph_fh_to_parent(struct super_block *sb,
					 struct fid *fid,
					int fh_len, int fh_type)
{
	struct ceph_nfs_confh *cfh = (void *)fid->raw;
	struct ceph_vino vino;
	struct inode *inode;
	struct dentry *dentry;
	int err;

	if (fh_type == 1)
		return ERR_PTR(-ESTALE);

	pr_debug("fh_to_parent %llx/%d\n", cfh->parent_ino,
		 cfh->parent_name_hash);

	vino.ino = cfh->ino;
	vino.snap = CEPH_NOSNAP;
	inode = ceph_find_inode(sb, vino);
	if (!inode)
		return ERR_PTR(-ESTALE);

	dentry = d_obtain_alias(inode);
	if (IS_ERR(dentry)) {
		pr_err("fh_to_parent %llx -- inode %p but ENOMEM\n",
		       cfh->ino, inode);
		iput(inode);
		return dentry;
	}
	err = ceph_init_dentry(dentry);
	if (err < 0) {
		iput(inode);
		return ERR_PTR(err);
	}
	dout("fh_to_parent %llx %p dentry %p\n", cfh->ino, inode, dentry);
	return dentry;
}

const struct export_operations ceph_export_ops = {
	.encode_fh = ceph_encode_fh,
	.fh_to_dentry = ceph_fh_to_dentry,
	.fh_to_parent = ceph_fh_to_parent,
};
Commit	Line	Data
a8e63b7d SW	1	#include "ceph_debug.h"
	2
	3	#include <linux/exportfs.h>
5a0e3ad6	4	#include <linux/slab.h>
a8e63b7d SW	5	#include <asm/unaligned.h>
	6
	7	#include "super.h"
	8
	9	/*
	10	* NFS export support
	11	*
	12	* NFS re-export of a ceph mount is, at present, only semireliable.
	13	* The basic issue is that the Ceph architectures doesn't lend itself
	14	* well to generating filehandles that will remain valid forever.
	15	*
	16	* So, we do our best. If you're lucky, your inode will be in the
	17	* client's cache. If it's not, and you have a connectable fh, then
	18	* the MDS server may be able to find it for you. Otherwise, you get
	19	* ESTALE.
	20	*
	21	* There are ways to this more reliable, but in the non-connectable fh
	22	* case, we won't every work perfectly, and in the connectable case,
	23	* some changes are needed on the MDS side to work better.
	24	*/
	25
	26	/*
	27	* Basic fh
	28	*/
	29	struct ceph_nfs_fh {
	30	u64 ino;
	31	} __attribute__ ((packed));
	32
	33	/*
	34	* Larger 'connectable' fh that includes parent ino and name hash.
	35	* Use this whenever possible, as it works more reliably.
	36	*/
	37	struct ceph_nfs_confh {
	38	u64 ino, parent_ino;
	39	u32 parent_name_hash;
	40	} __attribute__ ((packed));
	41
	42	static int ceph_encode_fh(struct dentry dentry, u32 rawfh, int *max_len,
	43	int connectable)
	44	{
92923dcb	45	int type;
a8e63b7d SW	46	struct ceph_nfs_fh fh = (void )rawfh;
	47	struct ceph_nfs_confh cfh = (void )rawfh;
	48	struct dentry *parent = dentry->d_parent;
	49	struct inode *inode = dentry->d_inode;
92923dcb AK	50	int connected_handle_length = sizeof(*cfh)/4;
92923dcb AK	51	int handle_length = sizeof(*fh)/4;
a8e63b7d SW	52
	53	/* don't re-export snaps */
	54	if (ceph_snap(inode) != CEPH_NOSNAP)
	55	return -EINVAL;
	56
92923dcb	57	if (*max_len >= connected_handle_length) {
a8e63b7d SW	58	dout("encode_fh %p connectable\n", dentry);
	59	cfh->ino = ceph_ino(dentry->d_inode);
	60	cfh->parent_ino = ceph_ino(parent->d_inode);
	61	cfh->parent_name_hash = parent->d_name.hash;
92923dcb	62	*max_len = connected_handle_length;
a8e63b7d	63	type = 2;
92923dcb	64	} else if (*max_len >= handle_length) {
a8e63b7d	65	if (connectable)
92923dcb	66	return 255;
a8e63b7d SW	67	dout("encode_fh %p\n", dentry);
a8e63b7d SW	68	fh->ino = ceph_ino(dentry->d_inode);
92923dcb	69	*max_len = handle_length;
a8e63b7d SW	70	type = 1;
a8e63b7d SW	71	} else {
92923dcb	72	return 255;
a8e63b7d SW	73	}
	74	return type;
	75	}
	76
	77	/*
	78	* convert regular fh to dentry
	79	*
	80	* FIXME: we should try harder by querying the mds for the ino.
	81	*/
	82	static struct dentry __fh_to_dentry(struct super_block sb,
	83	struct ceph_nfs_fh *fh)
	84	{
	85	struct inode *inode;
	86	struct dentry *dentry;
	87	struct ceph_vino vino;
	88	int err;
	89
	90	dout("__fh_to_dentry %llx\n", fh->ino);
	91	vino.ino = fh->ino;
	92	vino.snap = CEPH_NOSNAP;
	93	inode = ceph_find_inode(sb, vino);
	94	if (!inode)
	95	return ERR_PTR(-ESTALE);
	96
	97	dentry = d_obtain_alias(inode);
0d509c94	98	if (IS_ERR(dentry)) {
a8e63b7d SW	99	pr_err("fh_to_dentry %llx -- inode %p but ENOMEM\n",
	100	fh->ino, inode);
	101	iput(inode);
0d509c94	102	return dentry;
a8e63b7d SW	103	}
	104	err = ceph_init_dentry(dentry);
	105
	106	if (err < 0) {
	107	iput(inode);
	108	return ERR_PTR(err);
	109	}
	110	dout("__fh_to_dentry %llx %p dentry %p\n", fh->ino, inode, dentry);
	111	return dentry;
	112	}
	113
	114	/*
	115	* convert connectable fh to dentry
	116	*/
	117	static struct dentry __cfh_to_dentry(struct super_block sb,
	118	struct ceph_nfs_confh *cfh)
	119	{
640ef79d	120	struct ceph_mds_client *mdsc = &ceph_sb_to_client(sb)->mdsc;
a8e63b7d SW	121	struct inode *inode;
	122	struct dentry *dentry;
	123	struct ceph_vino vino;
	124	int err;
	125
	126	dout("__cfh_to_dentry %llx (%llx/%x)\n",
	127	cfh->ino, cfh->parent_ino, cfh->parent_name_hash);
	128
	129	vino.ino = cfh->ino;
	130	vino.snap = CEPH_NOSNAP;
	131	inode = ceph_find_inode(sb, vino);
	132	if (!inode) {
	133	struct ceph_mds_request *req;
	134
	135	req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPHASH,
	136	USE_ANY_MDS);
	137	if (IS_ERR(req))
7e34bc52	138	return ERR_CAST(req);
a8e63b7d SW	139
	140	req->r_ino1 = vino;
	141	req->r_ino2.ino = cfh->parent_ino;
	142	req->r_ino2.snap = CEPH_NOSNAP;
	143	req->r_path2 = kmalloc(16, GFP_NOFS);
	144	snprintf(req->r_path2, 16, "%d", cfh->parent_name_hash);
	145	req->r_num_caps = 1;
	146	err = ceph_mdsc_do_request(mdsc, NULL, req);
	147	ceph_mdsc_put_request(req);
	148	inode = ceph_find_inode(sb, vino);
	149	if (!inode)
	150	return ERR_PTR(err ? err : -ESTALE);
	151	}
	152
	153	dentry = d_obtain_alias(inode);
0d509c94	154	if (IS_ERR(dentry)) {
a8e63b7d SW	155	pr_err("cfh_to_dentry %llx -- inode %p but ENOMEM\n",
	156	cfh->ino, inode);
	157	iput(inode);
0d509c94	158	return dentry;
a8e63b7d SW	159	}
	160	err = ceph_init_dentry(dentry);
	161	if (err < 0) {
	162	iput(inode);
	163	return ERR_PTR(err);
	164	}
	165	dout("__cfh_to_dentry %llx %p dentry %p\n", cfh->ino, inode, dentry);
	166	return dentry;
	167	}
	168
	169	static struct dentry ceph_fh_to_dentry(struct super_block sb, struct fid *fid,
	170	int fh_len, int fh_type)
	171	{
	172	if (fh_type == 1)
	173	return __fh_to_dentry(sb, (struct ceph_nfs_fh *)fid->raw);
	174	else
	175	return __cfh_to_dentry(sb, (struct ceph_nfs_confh *)fid->raw);
	176	}
	177
	178	/*
	179	* get parent, if possible.
	180	*
	181	* FIXME: we could do better by querying the mds to discover the
	182	* parent.
	183	*/
	184	static struct dentry ceph_fh_to_parent(struct super_block sb,
	185	struct fid *fid,
	186	int fh_len, int fh_type)
	187	{
	188	struct ceph_nfs_confh cfh = (void )fid->raw;
	189	struct ceph_vino vino;
	190	struct inode *inode;
	191	struct dentry *dentry;
	192	int err;
	193
	194	if (fh_type == 1)
	195	return ERR_PTR(-ESTALE);
	196
	197	pr_debug("fh_to_parent %llx/%d\n", cfh->parent_ino,
	198	cfh->parent_name_hash);
	199
	200	vino.ino = cfh->ino;
	201	vino.snap = CEPH_NOSNAP;
	202	inode = ceph_find_inode(sb, vino);
	203	if (!inode)
	204	return ERR_PTR(-ESTALE);
	205
	206	dentry = d_obtain_alias(inode);
0d509c94	207	if (IS_ERR(dentry)) {
a8e63b7d SW	208	pr_err("fh_to_parent %llx -- inode %p but ENOMEM\n",
	209	cfh->ino, inode);
	210	iput(inode);
0d509c94	211	return dentry;
a8e63b7d SW	212	}
	213	err = ceph_init_dentry(dentry);
	214	if (err < 0) {
	215	iput(inode);
	216	return ERR_PTR(err);
	217	}
	218	dout("fh_to_parent %llx %p dentry %p\n", cfh->ino, inode, dentry);
	219	return dentry;
	220	}
	221
	222	const struct export_operations ceph_export_ops = {
	223	.encode_fh = ceph_encode_fh,
	224	.fh_to_dentry = ceph_fh_to_dentry,
	225	.fh_to_parent = ceph_fh_to_parent,
	226	};