[net-next-2.6.git] / fs / nfsd / nfs4callback.c

/*
 *  linux/fs/nfsd/nfs4callback.c
 *
 *  Copyright (c) 2001 The Regents of the University of Michigan.
 *  All rights reserved.
 *
 *  Kendrick Smith <kmsmith@umich.edu>
 *  Andy Adamson <andros@umich.edu>
 *
 *  Redistribution and use in source and binary forms, with or without
 *  modification, are permitted provided that the following conditions
 *  are met:
 *
 *  1. Redistributions of source code must retain the above copyright
 *     notice, this list of conditions and the following disclaimer.
 *  2. Redistributions in binary form must reproduce the above copyright
 *     notice, this list of conditions and the following disclaimer in the
 *     documentation and/or other materials provided with the distribution.
 *  3. Neither the name of the University nor the names of its
 *     contributors may be used to endorse or promote products derived
 *     from this software without specific prior written permission.
 *
 *  THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
 *  WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
 *  MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 *  DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 *  FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 *  CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 *  SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
 *  BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
 *  LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
 *  NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 *  SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#include <linux/module.h>
#include <linux/list.h>
#include <linux/inet.h>
#include <linux/errno.h>
#include <linux/delay.h>
#include <linux/sched.h>
#include <linux/kthread.h>
#include <linux/sunrpc/xdr.h>
#include <linux/sunrpc/svc.h>
#include <linux/sunrpc/clnt.h>
#include <linux/nfsd/nfsd.h>
#include <linux/nfsd/state.h>
#include <linux/sunrpc/sched.h>
#include <linux/nfs4.h>

#define NFSDDBG_FACILITY                NFSDDBG_PROC

#define NFSPROC4_CB_NULL 0
#define NFSPROC4_CB_COMPOUND 1

/* Index of predefined Linux callback client operations */

enum {
        NFSPROC4_CLNT_CB_NULL = 0,
	NFSPROC4_CLNT_CB_RECALL,
};

enum nfs_cb_opnum4 {
	OP_CB_RECALL            = 4,
};

#define NFS4_MAXTAGLEN		20

#define NFS4_enc_cb_null_sz		0
#define NFS4_dec_cb_null_sz		0
#define cb_compound_enc_hdr_sz		4
#define cb_compound_dec_hdr_sz		(3 + (NFS4_MAXTAGLEN >> 2))
#define op_enc_sz			1
#define op_dec_sz			2
#define enc_nfs4_fh_sz			(1 + (NFS4_FHSIZE >> 2))
#define enc_stateid_sz			(NFS4_STATEID_SIZE >> 2)
#define NFS4_enc_cb_recall_sz		(cb_compound_enc_hdr_sz +       \
					1 + enc_stateid_sz +            \
					enc_nfs4_fh_sz)

#define NFS4_dec_cb_recall_sz		(cb_compound_dec_hdr_sz  +      \
					op_dec_sz)

/*
* Generic encode routines from fs/nfs/nfs4xdr.c
*/
static inline __be32 *
xdr_writemem(__be32 *p, const void *ptr, int nbytes)
{
	int tmp = XDR_QUADLEN(nbytes);
	if (!tmp)
		return p;
	p[tmp-1] = 0;
	memcpy(p, ptr, nbytes);
	return p + tmp;
}

#define WRITE32(n)               *p++ = htonl(n)
#define WRITEMEM(ptr,nbytes)     do {                           \
	p = xdr_writemem(p, ptr, nbytes);                       \
} while (0)
#define RESERVE_SPACE(nbytes)   do {                            \
	p = xdr_reserve_space(xdr, nbytes);                     \
	if (!p) dprintk("NFSD: RESERVE_SPACE(%d) failed in function %s\n", (int) (nbytes), __func__); \
	BUG_ON(!p);                                             \
} while (0)

/*
 * Generic decode routines from fs/nfs/nfs4xdr.c
 */
#define DECODE_TAIL                             \
	status = 0;                             \
out:                                            \
	return status;                          \
xdr_error:                                      \
	dprintk("NFSD: xdr error! (%s:%d)\n", __FILE__, __LINE__); \
	status = -EIO;                          \
	goto out

#define READ32(x)         (x) = ntohl(*p++)
#define READ64(x)         do {                  \
	(x) = (u64)ntohl(*p++) << 32;           \
	(x) |= ntohl(*p++);                     \
} while (0)
#define READTIME(x)       do {                  \
	p++;                                    \
	(x.tv_sec) = ntohl(*p++);               \
	(x.tv_nsec) = ntohl(*p++);              \
} while (0)
#define READ_BUF(nbytes)  do { \
	p = xdr_inline_decode(xdr, nbytes); \
	if (!p) { \
		dprintk("NFSD: %s: reply buffer overflowed in line %d.\n", \
			__func__, __LINE__); \
		return -EIO; \
	} \
} while (0)

struct nfs4_cb_compound_hdr {
	int		status;
	u32		ident;
	u32		nops;
	__be32		*nops_p;
	u32		minorversion;
	u32		taglen;
	char		*tag;
};

static struct {
int stat;
int errno;
} nfs_cb_errtbl[] = {
	{ NFS4_OK,		0               },
	{ NFS4ERR_PERM,		EPERM           },
	{ NFS4ERR_NOENT,	ENOENT          },
	{ NFS4ERR_IO,		EIO             },
	{ NFS4ERR_NXIO,		ENXIO           },
	{ NFS4ERR_ACCESS,	EACCES          },
	{ NFS4ERR_EXIST,	EEXIST          },
	{ NFS4ERR_XDEV,		EXDEV           },
	{ NFS4ERR_NOTDIR,	ENOTDIR         },
	{ NFS4ERR_ISDIR,	EISDIR          },
	{ NFS4ERR_INVAL,	EINVAL          },
	{ NFS4ERR_FBIG,		EFBIG           },
	{ NFS4ERR_NOSPC,	ENOSPC          },
	{ NFS4ERR_ROFS,		EROFS           },
	{ NFS4ERR_MLINK,	EMLINK          },
	{ NFS4ERR_NAMETOOLONG,	ENAMETOOLONG    },
	{ NFS4ERR_NOTEMPTY,	ENOTEMPTY       },
	{ NFS4ERR_DQUOT,	EDQUOT          },
	{ NFS4ERR_STALE,	ESTALE          },
	{ NFS4ERR_BADHANDLE,	EBADHANDLE      },
	{ NFS4ERR_BAD_COOKIE,	EBADCOOKIE      },
	{ NFS4ERR_NOTSUPP,	ENOTSUPP        },
	{ NFS4ERR_TOOSMALL,	ETOOSMALL       },
	{ NFS4ERR_SERVERFAULT,	ESERVERFAULT    },
	{ NFS4ERR_BADTYPE,	EBADTYPE        },
	{ NFS4ERR_LOCKED,	EAGAIN          },
	{ NFS4ERR_RESOURCE,	EREMOTEIO       },
	{ NFS4ERR_SYMLINK,	ELOOP           },
	{ NFS4ERR_OP_ILLEGAL,	EOPNOTSUPP      },
	{ NFS4ERR_DEADLOCK,	EDEADLK         },
	{ -1,                   EIO             }
};

static int
nfs_cb_stat_to_errno(int stat)
{
	int i;
	for (i = 0; nfs_cb_errtbl[i].stat != -1; i++) {
		if (nfs_cb_errtbl[i].stat == stat)
			return nfs_cb_errtbl[i].errno;
	}
	/* If we cannot translate the error, the recovery routines should
	* handle it.
	* Note: remaining NFSv4 error codes have values > 10000, so should
	* not conflict with native Linux error codes.
	*/
	return stat;
}

/*
 * XDR encode
 */

static void
encode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr)
{
	__be32 * p;

	RESERVE_SPACE(16);
	WRITE32(0);            /* tag length is always 0 */
	WRITE32(hdr->minorversion);
	WRITE32(hdr->ident);
	hdr->nops_p = p;
	WRITE32(hdr->nops);
}

static void encode_cb_nops(struct nfs4_cb_compound_hdr *hdr)
{
	*hdr->nops_p = htonl(hdr->nops);
}

static void
encode_cb_recall(struct xdr_stream *xdr, struct nfs4_delegation *dp,
		struct nfs4_cb_compound_hdr *hdr)
{
	__be32 *p;
	int len = dp->dl_fh.fh_size;

	RESERVE_SPACE(12+sizeof(dp->dl_stateid) + len);
	WRITE32(OP_CB_RECALL);
	WRITE32(dp->dl_stateid.si_generation);
	WRITEMEM(&dp->dl_stateid.si_opaque, sizeof(stateid_opaque_t));
	WRITE32(0); /* truncate optimization not implemented */
	WRITE32(len);
	WRITEMEM(&dp->dl_fh.fh_base, len);
	hdr->nops++;
}

static int
nfs4_xdr_enc_cb_null(struct rpc_rqst *req, __be32 *p)
{
	struct xdr_stream xdrs, *xdr = &xdrs;

	xdr_init_encode(&xdrs, &req->rq_snd_buf, p);
        RESERVE_SPACE(0);
	return 0;
}

static int
nfs4_xdr_enc_cb_recall(struct rpc_rqst *req, __be32 *p, struct nfs4_delegation *args)
{
	struct xdr_stream xdr;
	struct nfs4_cb_compound_hdr hdr = {
		.ident = args->dl_ident,
	};

	xdr_init_encode(&xdr, &req->rq_snd_buf, p);
	encode_cb_compound_hdr(&xdr, &hdr);
	encode_cb_recall(&xdr, args, &hdr);
	encode_cb_nops(&hdr);
	return 0;
}


static int
decode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr){
        __be32 *p;

        READ_BUF(8);
        READ32(hdr->status);
        READ32(hdr->taglen);
        READ_BUF(hdr->taglen + 4);
        hdr->tag = (char *)p;
        p += XDR_QUADLEN(hdr->taglen);
        READ32(hdr->nops);
        return 0;
}

static int
decode_cb_op_hdr(struct xdr_stream *xdr, enum nfs_opnum4 expected)
{
	__be32 *p;
	u32 op;
	int32_t nfserr;

	READ_BUF(8);
	READ32(op);
	if (op != expected) {
		dprintk("NFSD: decode_cb_op_hdr: Callback server returned "
		         " operation %d but we issued a request for %d\n",
		         op, expected);
		return -EIO;
	}
	READ32(nfserr);
	if (nfserr != NFS_OK)
		return -nfs_cb_stat_to_errno(nfserr);
	return 0;
}

static int
nfs4_xdr_dec_cb_null(struct rpc_rqst *req, __be32 *p)
{
	return 0;
}

static int
nfs4_xdr_dec_cb_recall(struct rpc_rqst *rqstp, __be32 *p)
{
	struct xdr_stream xdr;
	struct nfs4_cb_compound_hdr hdr;
	int status;

	xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p);
	status = decode_cb_compound_hdr(&xdr, &hdr);
	if (status)
		goto out;
	status = decode_cb_op_hdr(&xdr, OP_CB_RECALL);
out:
	return status;
}

/*
 * RPC procedure tables
 */
#define PROC(proc, call, argtype, restype)                              \
[NFSPROC4_CLNT_##proc] = {                                      	\
        .p_proc   = NFSPROC4_CB_##call,					\
        .p_encode = (kxdrproc_t) nfs4_xdr_##argtype,                    \
        .p_decode = (kxdrproc_t) nfs4_xdr_##restype,                    \
        .p_arglen = NFS4_##argtype##_sz,                                \
        .p_replen = NFS4_##restype##_sz,                                \
        .p_statidx = NFSPROC4_CB_##call,				\
	.p_name   = #proc,                                              \
}

static struct rpc_procinfo     nfs4_cb_procedures[] = {
    PROC(CB_NULL,      NULL,     enc_cb_null,     dec_cb_null),
    PROC(CB_RECALL,    COMPOUND,   enc_cb_recall,      dec_cb_recall),
};

static struct rpc_version       nfs_cb_version4 = {
        .number                 = 1,
        .nrprocs                = ARRAY_SIZE(nfs4_cb_procedures),
        .procs                  = nfs4_cb_procedures
};

static struct rpc_version *	nfs_cb_version[] = {
	NULL,
	&nfs_cb_version4,
};

static struct rpc_program cb_program;

static struct rpc_stat cb_stats = {
		.program	= &cb_program
};

#define NFS4_CALLBACK 0x40000000
static struct rpc_program cb_program = {
		.name 		= "nfs4_cb",
		.number		= NFS4_CALLBACK,
		.nrvers		= ARRAY_SIZE(nfs_cb_version),
		.version	= nfs_cb_version,
		.stats		= &cb_stats,
		.pipe_dir_name  = "/nfsd4_cb",
};

static int max_cb_time(void)
{
	return max(NFSD_LEASE_TIME/10, (time_t)1) * HZ;
}

/* Reference counting, callback cleanup, etc., all look racy as heck.
 * And why is cb_set an atomic? */

int setup_callback_client(struct nfs4_client *clp)
{
	struct nfs4_cb_conn *cb = &clp->cl_cb_conn;
	struct rpc_timeout	timeparms = {
		.to_initval	= max_cb_time(),
		.to_retries	= 0,
	};
	struct rpc_create_args args = {
		.protocol	= IPPROTO_TCP,
		.address	= (struct sockaddr *) &cb->cb_addr,
		.addrsize	= cb->cb_addrlen,
		.timeout	= &timeparms,
		.program	= &cb_program,
		.prognumber	= cb->cb_prog,
		.version	= nfs_cb_version[1]->number,
		.authflavor	= clp->cl_flavor,
		.flags		= (RPC_CLNT_CREATE_NOPING | RPC_CLNT_CREATE_QUIET),
		.client_name    = clp->cl_principal,
	};
	struct rpc_clnt *client;

	if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5))
		return -EINVAL;

	/* Create RPC client */
	client = rpc_create(&args);
	if (IS_ERR(client)) {
		dprintk("NFSD: couldn't create callback client: %ld\n",
			PTR_ERR(client));
		return PTR_ERR(client);
	}
	cb->cb_client = client;
	return 0;

}

static void warn_no_callback_path(struct nfs4_client *clp, int reason)
{
	dprintk("NFSD: warning: no callback path to client %.*s: error %d\n",
		(int)clp->cl_name.len, clp->cl_name.data, reason);
}

static void nfsd4_cb_probe_done(struct rpc_task *task, void *calldata)
{
	struct nfs4_client *clp = calldata;

	if (task->tk_status)
		warn_no_callback_path(clp, task->tk_status);
	else
		atomic_set(&clp->cl_cb_conn.cb_set, 1);
	put_nfs4_client(clp);
}

static const struct rpc_call_ops nfsd4_cb_probe_ops = {
	.rpc_call_done = nfsd4_cb_probe_done,
};

static struct rpc_cred *lookup_cb_cred(struct nfs4_cb_conn *cb)
{
	struct auth_cred acred = {
		.machine_cred = 1
	};

	/*
	 * Note in the gss case this doesn't actually have to wait for a
	 * gss upcall (or any calls to the client); this just creates a
	 * non-uptodate cred which the rpc state machine will fill in with
	 * a refresh_upcall later.
	 */
	return rpcauth_lookup_credcache(cb->cb_client->cl_auth, &acred,
							RPCAUTH_LOOKUP_NEW);
}

void do_probe_callback(struct nfs4_client *clp)
{
	struct nfs4_cb_conn *cb = &clp->cl_cb_conn;
	struct rpc_message msg = {
		.rpc_proc       = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL],
		.rpc_argp       = clp,
	};
	struct rpc_cred *cred;
	int status;

	cred = lookup_cb_cred(cb);
	if (IS_ERR(cred)) {
		status = PTR_ERR(cred);
		goto out;
	}
	cb->cb_cred = cred;
	msg.rpc_cred = cb->cb_cred;
	status = rpc_call_async(cb->cb_client, &msg, RPC_TASK_SOFT,
				&nfsd4_cb_probe_ops, (void *)clp);
out:
	if (status) {
		warn_no_callback_path(clp, status);
		put_nfs4_client(clp);
	}
}

/*
 * Set up the callback client and put a NFSPROC4_CB_NULL on the wire...
 */
void
nfsd4_probe_callback(struct nfs4_client *clp)
{
	int status;

	BUG_ON(atomic_read(&clp->cl_cb_conn.cb_set));

	status = setup_callback_client(clp);
	if (status) {
		warn_no_callback_path(clp, status);
		return;
	}

	/* the task holds a reference to the nfs4_client struct */
	atomic_inc(&clp->cl_count);

	do_probe_callback(clp);
}

static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata)
{
	struct nfs4_delegation *dp = calldata;
	struct nfs4_client *clp = dp->dl_client;

	switch (task->tk_status) {
	case -EIO:
		/* Network partition? */
		atomic_set(&clp->cl_cb_conn.cb_set, 0);
		warn_no_callback_path(clp, task->tk_status);
	case -EBADHANDLE:
	case -NFS4ERR_BAD_STATEID:
		/* Race: client probably got cb_recall
		 * before open reply granting delegation */
		break;
	default:
		/* success, or error we can't handle */
		return;
	}
	if (dp->dl_retries--) {
		rpc_delay(task, 2*HZ);
		task->tk_status = 0;
		rpc_restart_call(task);
	} else {
		atomic_set(&clp->cl_cb_conn.cb_set, 0);
		warn_no_callback_path(clp, task->tk_status);
	}
}

static void nfsd4_cb_recall_release(void *calldata)
{
	struct nfs4_delegation *dp = calldata;
	struct nfs4_client *clp = dp->dl_client;

	nfs4_put_delegation(dp);
	put_nfs4_client(clp);
}

static const struct rpc_call_ops nfsd4_cb_recall_ops = {
	.rpc_call_done = nfsd4_cb_recall_done,
	.rpc_release = nfsd4_cb_recall_release,
};

/*
 * called with dp->dl_count inc'ed.
 */
void
nfsd4_cb_recall(struct nfs4_delegation *dp)
{
	struct nfs4_client *clp = dp->dl_client;
	struct rpc_clnt *clnt = clp->cl_cb_conn.cb_client;
	struct rpc_message msg = {
		.rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_RECALL],
		.rpc_argp = dp,
		.rpc_cred = clp->cl_cb_conn.cb_cred
	};
	int status;

	dp->dl_retries = 1;
	status = rpc_call_async(clnt, &msg, RPC_TASK_SOFT,
				&nfsd4_cb_recall_ops, dp);
	if (status) {
		put_nfs4_client(clp);
		nfs4_put_delegation(dp);
	}
}
Commit	Line	Data
1da177e4 LT	1	/*
	2	* linux/fs/nfsd/nfs4callback.c
	3	*
	4	* Copyright (c) 2001 The Regents of the University of Michigan.
	5	* All rights reserved.
	6	*
	7	* Kendrick Smith <kmsmith@umich.edu>
	8	* Andy Adamson <andros@umich.edu>
	9	*
	10	* Redistribution and use in source and binary forms, with or without
	11	* modification, are permitted provided that the following conditions
	12	* are met:
	13	*
	14	* 1. Redistributions of source code must retain the above copyright
	15	* notice, this list of conditions and the following disclaimer.
	16	* 2. Redistributions in binary form must reproduce the above copyright
	17	* notice, this list of conditions and the following disclaimer in the
	18	* documentation and/or other materials provided with the distribution.
	19	* 3. Neither the name of the University nor the names of its
	20	* contributors may be used to endorse or promote products derived
	21	* from this software without specific prior written permission.
	22	*
	23	* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
	24	* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
	25	* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
	26	* DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
	27	* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
	28	* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
	29	* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
	30	* BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
	31	* LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
	32	* NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
	33	* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
	34	*/
	35
1da177e4 LT	36	#include <linux/module.h>
	37	#include <linux/list.h>
	38	#include <linux/inet.h>
	39	#include <linux/errno.h>
	40	#include <linux/delay.h>
e8edc6e0	41	#include <linux/sched.h>
2b47eece	42	#include <linux/kthread.h>
1da177e4 LT	43	#include <linux/sunrpc/xdr.h>
	44	#include <linux/sunrpc/svc.h>
	45	#include <linux/sunrpc/clnt.h>
	46	#include <linux/nfsd/nfsd.h>
	47	#include <linux/nfsd/state.h>
	48	#include <linux/sunrpc/sched.h>
	49	#include <linux/nfs4.h>
	50
	51	#define NFSDDBG_FACILITY NFSDDBG_PROC
	52
	53	#define NFSPROC4_CB_NULL 0
	54	#define NFSPROC4_CB_COMPOUND 1
	55
1da177e4 LT	56	/* Index of predefined Linux callback client operations */
	57
	58	enum {
	59	NFSPROC4_CLNT_CB_NULL = 0,
	60	NFSPROC4_CLNT_CB_RECALL,
	61	};
	62
	63	enum nfs_cb_opnum4 {
	64	OP_CB_RECALL = 4,
	65	};
	66
	67	#define NFS4_MAXTAGLEN 20
	68
	69	#define NFS4_enc_cb_null_sz 0
	70	#define NFS4_dec_cb_null_sz 0
	71	#define cb_compound_enc_hdr_sz 4
	72	#define cb_compound_dec_hdr_sz (3 + (NFS4_MAXTAGLEN >> 2))
	73	#define op_enc_sz 1
	74	#define op_dec_sz 2
	75	#define enc_nfs4_fh_sz (1 + (NFS4_FHSIZE >> 2))
0ac68d17	76	#define enc_stateid_sz (NFS4_STATEID_SIZE >> 2)
1da177e4 LT	77	#define NFS4_enc_cb_recall_sz (cb_compound_enc_hdr_sz + \
	78	1 + enc_stateid_sz + \
	79	enc_nfs4_fh_sz)
	80
	81	#define NFS4_dec_cb_recall_sz (cb_compound_dec_hdr_sz + \
	82	op_dec_sz)
	83
	84	/*
	85	* Generic encode routines from fs/nfs/nfs4xdr.c
	86	*/
f00f328f AV	87	static inline __be32 *
f00f328f AV	88	xdr_writemem(__be32 p, const void ptr, int nbytes)
1da177e4 LT	89	{
	90	int tmp = XDR_QUADLEN(nbytes);
	91	if (!tmp)
	92	return p;
	93	p[tmp-1] = 0;
	94	memcpy(p, ptr, nbytes);
	95	return p + tmp;
	96	}
	97
	98	#define WRITE32(n) *p++ = htonl(n)
	99	#define WRITEMEM(ptr,nbytes) do { \
	100	p = xdr_writemem(p, ptr, nbytes); \
	101	} while (0)
	102	#define RESERVE_SPACE(nbytes) do { \
	103	p = xdr_reserve_space(xdr, nbytes); \
8e24eea7	104	if (!p) dprintk("NFSD: RESERVE_SPACE(%d) failed in function %s\n", (int) (nbytes), __func__); \
1da177e4 LT	105	BUG_ON(!p); \
	106	} while (0)
	107
	108	/*
	109	* Generic decode routines from fs/nfs/nfs4xdr.c
	110	*/
	111	#define DECODE_TAIL \
	112	status = 0; \
	113	out: \
	114	return status; \
	115	xdr_error: \
	116	dprintk("NFSD: xdr error! (%s:%d)\n", __FILE__, __LINE__); \
	117	status = -EIO; \
	118	goto out
	119
	120	#define READ32(x) (x) = ntohl(*p++)
	121	#define READ64(x) do { \
	122	(x) = (u64)ntohl(*p++) << 32; \
	123	(x) \|= ntohl(*p++); \
	124	} while (0)
	125	#define READTIME(x) do { \
	126	p++; \
	127	(x.tv_sec) = ntohl(*p++); \
	128	(x.tv_nsec) = ntohl(*p++); \
	129	} while (0)
	130	#define READ_BUF(nbytes) do { \
	131	p = xdr_inline_decode(xdr, nbytes); \
	132	if (!p) { \
3e3b4800	133	dprintk("NFSD: %s: reply buffer overflowed in line %d.\n", \
8e24eea7	134	__func__, __LINE__); \
1da177e4 LT	135	return -EIO; \
	136	} \
	137	} while (0)
	138
	139	struct nfs4_cb_compound_hdr {
	140	int status;
	141	u32 ident;
	142	u32 nops;
ef52bff8	143	__be32 *nops_p;
ab52ae6d	144	u32 minorversion;
1da177e4	145	u32 taglen;
ef52bff8	146	char *tag;
1da177e4 LT	147	};
	148
	149	static struct {
	150	int stat;
	151	int errno;
	152	} nfs_cb_errtbl[] = {
	153	{ NFS4_OK, 0 },
	154	{ NFS4ERR_PERM, EPERM },
	155	{ NFS4ERR_NOENT, ENOENT },
	156	{ NFS4ERR_IO, EIO },
	157	{ NFS4ERR_NXIO, ENXIO },
	158	{ NFS4ERR_ACCESS, EACCES },
	159	{ NFS4ERR_EXIST, EEXIST },
	160	{ NFS4ERR_XDEV, EXDEV },
	161	{ NFS4ERR_NOTDIR, ENOTDIR },
	162	{ NFS4ERR_ISDIR, EISDIR },
	163	{ NFS4ERR_INVAL, EINVAL },
	164	{ NFS4ERR_FBIG, EFBIG },
	165	{ NFS4ERR_NOSPC, ENOSPC },
	166	{ NFS4ERR_ROFS, EROFS },
	167	{ NFS4ERR_MLINK, EMLINK },
	168	{ NFS4ERR_NAMETOOLONG, ENAMETOOLONG },
	169	{ NFS4ERR_NOTEMPTY, ENOTEMPTY },
	170	{ NFS4ERR_DQUOT, EDQUOT },
	171	{ NFS4ERR_STALE, ESTALE },
	172	{ NFS4ERR_BADHANDLE, EBADHANDLE },
	173	{ NFS4ERR_BAD_COOKIE, EBADCOOKIE },
	174	{ NFS4ERR_NOTSUPP, ENOTSUPP },
	175	{ NFS4ERR_TOOSMALL, ETOOSMALL },
	176	{ NFS4ERR_SERVERFAULT, ESERVERFAULT },
	177	{ NFS4ERR_BADTYPE, EBADTYPE },
	178	{ NFS4ERR_LOCKED, EAGAIN },
	179	{ NFS4ERR_RESOURCE, EREMOTEIO },
	180	{ NFS4ERR_SYMLINK, ELOOP },
	181	{ NFS4ERR_OP_ILLEGAL, EOPNOTSUPP },
	182	{ NFS4ERR_DEADLOCK, EDEADLK },
	183	{ -1, EIO }
	184	};
	185
	186	static int
	187	nfs_cb_stat_to_errno(int stat)
	188	{
	189	int i;
	190	for (i = 0; nfs_cb_errtbl[i].stat != -1; i++) {
	191	if (nfs_cb_errtbl[i].stat == stat)
	192	return nfs_cb_errtbl[i].errno;
	193	}
	194	/* If we cannot translate the error, the recovery routines should
	195	* handle it.
	196	* Note: remaining NFSv4 error codes have values > 10000, so should
	197	* not conflict with native Linux error codes.
	198	*/
	199	return stat;
	200	}
	201
	202	/*
	203	* XDR encode
	204	*/
	205
ef52bff8	206	static void
1da177e4 LT	207	encode_cb_compound_hdr(struct xdr_stream xdr, struct nfs4_cb_compound_hdr hdr)
1da177e4 LT	208	{
f00f328f	209	__be32 * p;
1da177e4 LT	210
	211	RESERVE_SPACE(16);
	212	WRITE32(0); /* tag length is always 0 */
ab52ae6d	213	WRITE32(hdr->minorversion);
1da177e4	214	WRITE32(hdr->ident);
ef52bff8	215	hdr->nops_p = p;
1da177e4	216	WRITE32(hdr->nops);
1da177e4 LT	217	}
1da177e4 LT	218
ef52bff8 AA	219	static void encode_cb_nops(struct nfs4_cb_compound_hdr *hdr)
	220	{
	221	*hdr->nops_p = htonl(hdr->nops);
	222	}
	223
	224	static void
	225	encode_cb_recall(struct xdr_stream xdr, struct nfs4_delegation dp,
	226	struct nfs4_cb_compound_hdr *hdr)
1da177e4	227	{
f00f328f	228	__be32 *p;
b53d40c5	229	int len = dp->dl_fh.fh_size;
1da177e4	230
b53d40c5	231	RESERVE_SPACE(12+sizeof(dp->dl_stateid) + len);
1da177e4	232	WRITE32(OP_CB_RECALL);
b53d40c5 BF	233	WRITE32(dp->dl_stateid.si_generation);
b53d40c5 BF	234	WRITEMEM(&dp->dl_stateid.si_opaque, sizeof(stateid_opaque_t));
6707bd3d	235	WRITE32(0); /* truncate optimization not implemented */
1da177e4	236	WRITE32(len);
b53d40c5	237	WRITEMEM(&dp->dl_fh.fh_base, len);
ef52bff8	238	hdr->nops++;
1da177e4 LT	239	}
	240
	241	static int
f00f328f	242	nfs4_xdr_enc_cb_null(struct rpc_rqst req, __be32 p)
1da177e4 LT	243	{
	244	struct xdr_stream xdrs, *xdr = &xdrs;
	245
	246	xdr_init_encode(&xdrs, &req->rq_snd_buf, p);
	247	RESERVE_SPACE(0);
	248	return 0;
	249	}
	250
	251	static int
b53d40c5	252	nfs4_xdr_enc_cb_recall(struct rpc_rqst req, __be32 p, struct nfs4_delegation *args)
1da177e4 LT	253	{
	254	struct xdr_stream xdr;
	255	struct nfs4_cb_compound_hdr hdr = {
b53d40c5	256	.ident = args->dl_ident,
1da177e4 LT	257	};
	258
	259	xdr_init_encode(&xdr, &req->rq_snd_buf, p);
	260	encode_cb_compound_hdr(&xdr, &hdr);
ef52bff8 AA	261	encode_cb_recall(&xdr, args, &hdr);
	262	encode_cb_nops(&hdr);
	263	return 0;
1da177e4 LT	264	}
	265
	266
	267	static int
	268	decode_cb_compound_hdr(struct xdr_stream xdr, struct nfs4_cb_compound_hdr hdr){
f00f328f	269	__be32 *p;
1da177e4 LT	270
	271	READ_BUF(8);
	272	READ32(hdr->status);
	273	READ32(hdr->taglen);
	274	READ_BUF(hdr->taglen + 4);
	275	hdr->tag = (char *)p;
	276	p += XDR_QUADLEN(hdr->taglen);
	277	READ32(hdr->nops);
	278	return 0;
	279	}
	280
	281	static int
	282	decode_cb_op_hdr(struct xdr_stream *xdr, enum nfs_opnum4 expected)
	283	{
f00f328f	284	__be32 *p;
1da177e4 LT	285	u32 op;
	286	int32_t nfserr;
	287
	288	READ_BUF(8);
	289	READ32(op);
	290	if (op != expected) {
	291	dprintk("NFSD: decode_cb_op_hdr: Callback server returned "
	292	" operation %d but we issued a request for %d\n",
	293	op, expected);
	294	return -EIO;
	295	}
	296	READ32(nfserr);
	297	if (nfserr != NFS_OK)
	298	return -nfs_cb_stat_to_errno(nfserr);
	299	return 0;
	300	}
	301
	302	static int
f00f328f	303	nfs4_xdr_dec_cb_null(struct rpc_rqst req, __be32 p)
1da177e4 LT	304	{
	305	return 0;
	306	}
	307
	308	static int
f00f328f	309	nfs4_xdr_dec_cb_recall(struct rpc_rqst rqstp, __be32 p)
1da177e4 LT	310	{
	311	struct xdr_stream xdr;
	312	struct nfs4_cb_compound_hdr hdr;
	313	int status;
	314
	315	xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p);
	316	status = decode_cb_compound_hdr(&xdr, &hdr);
	317	if (status)
	318	goto out;
	319	status = decode_cb_op_hdr(&xdr, OP_CB_RECALL);
	320	out:
	321	return status;
	322	}
	323
	324	/*
	325	* RPC procedure tables
	326	*/
1da177e4 LT	327	#define PROC(proc, call, argtype, restype) \
	328	[NFSPROC4_CLNT_##proc] = { \
	329	.p_proc = NFSPROC4_CB_##call, \
	330	.p_encode = (kxdrproc_t) nfs4_xdr_##argtype, \
	331	.p_decode = (kxdrproc_t) nfs4_xdr_##restype, \
2bea90d4 CL	332	.p_arglen = NFS4_##argtype##_sz, \
2bea90d4 CL	333	.p_replen = NFS4_##restype##_sz, \
cc0175c1 CL	334	.p_statidx = NFSPROC4_CB_##call, \
cc0175c1 CL	335	.p_name = #proc, \
1da177e4 LT	336	}
1da177e4 LT	337
fd39ca9a	338	static struct rpc_procinfo nfs4_cb_procedures[] = {
1da177e4 LT	339	PROC(CB_NULL, NULL, enc_cb_null, dec_cb_null),
	340	PROC(CB_RECALL, COMPOUND, enc_cb_recall, dec_cb_recall),
	341	};
	342
fd39ca9a	343	static struct rpc_version nfs_cb_version4 = {
1da177e4	344	.number = 1,
e8c96f8c	345	.nrprocs = ARRAY_SIZE(nfs4_cb_procedures),
1da177e4 LT	346	.procs = nfs4_cb_procedures
	347	};
	348
	349	static struct rpc_version * nfs_cb_version[] = {
	350	NULL,
	351	&nfs_cb_version4,
	352	};
	353
ff7d9756 OK	354	static struct rpc_program cb_program;
	355
	356	static struct rpc_stat cb_stats = {
	357	.program = &cb_program
	358	};
	359
	360	#define NFS4_CALLBACK 0x40000000
	361	static struct rpc_program cb_program = {
	362	.name = "nfs4_cb",
	363	.number = NFS4_CALLBACK,
	364	.nrvers = ARRAY_SIZE(nfs_cb_version),
	365	.version = nfs_cb_version,
	366	.stats = &cb_stats,
61054b14	367	.pipe_dir_name = "/nfsd4_cb",
ff7d9756 OK	368	};
ff7d9756 OK	369
595947ac BF	370	static int max_cb_time(void)
	371	{
	372	return max(NFSD_LEASE_TIME/10, (time_t)1) * HZ;
	373	}
	374
2b47eece BF	375	/* Reference counting, callback cleanup, etc., all look racy as heck.
	376	* And why is cb_set an atomic? */
	377
e1cab5a5	378	int setup_callback_client(struct nfs4_client *clp)
2b47eece	379	{
c237dc03	380	struct nfs4_cb_conn *cb = &clp->cl_cb_conn;
ae5c7947	381	struct rpc_timeout timeparms = {
595947ac BF	382	.to_initval = max_cb_time(),
595947ac BF	383	.to_retries = 0,
ae5c7947	384	};
ae5c7947 CL	385	struct rpc_create_args args = {
ae5c7947 CL	386	.protocol = IPPROTO_TCP,
aa9a4ec7 JL	387	.address = (struct sockaddr *) &cb->cb_addr,
aa9a4ec7 JL	388	.addrsize = cb->cb_addrlen,
ae5c7947	389	.timeout = &timeparms,
ff7d9756	390	.program = &cb_program,
d5b337b4	391	.prognumber = cb->cb_prog,
ae5c7947	392	.version = nfs_cb_version[1]->number,
61054b14	393	.authflavor = clp->cl_flavor,
b6b6152c	394	.flags = (RPC_CLNT_CREATE_NOPING \| RPC_CLNT_CREATE_QUIET),
608207e8	395	.client_name = clp->cl_principal,
ae5c7947	396	};
63c86716	397	struct rpc_clnt *client;
1da177e4	398
418cd20a	399	if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5))
e1cab5a5	400	return -EINVAL;
608207e8	401
ae5c7947	402	/* Create RPC client */
63c86716	403	client = rpc_create(&args);
e1cab5a5	404	if (IS_ERR(client)) {
a601caed BF	405	dprintk("NFSD: couldn't create callback client: %ld\n",
a601caed BF	406	PTR_ERR(client));
e1cab5a5 BF	407	return PTR_ERR(client);
	408	}
	409	cb->cb_client = client;
	410	return 0;
a601caed BF	411
	412	}
	413
ecdd03b7 BF	414	static void warn_no_callback_path(struct nfs4_client *clp, int reason)
	415	{
	416	dprintk("NFSD: warning: no callback path to client %.*s: error %d\n",
	417	(int)clp->cl_name.len, clp->cl_name.data, reason);
	418	}
	419
e300a63c BF	420	static void nfsd4_cb_probe_done(struct rpc_task task, void calldata)
	421	{
	422	struct nfs4_client *clp = calldata;
	423
	424	if (task->tk_status)
	425	warn_no_callback_path(clp, task->tk_status);
	426	else
c237dc03	427	atomic_set(&clp->cl_cb_conn.cb_set, 1);
e300a63c BF	428	put_nfs4_client(clp);
	429	}
	430
	431	static const struct rpc_call_ops nfsd4_cb_probe_ops = {
	432	.rpc_call_done = nfsd4_cb_probe_done,
	433	};
	434
c237dc03	435	static struct rpc_cred lookup_cb_cred(struct nfs4_cb_conn cb)
3cef9ab2 BF	436	{
	437	struct auth_cred acred = {
	438	.machine_cred = 1
	439	};
	440
	441	/*
	442	* Note in the gss case this doesn't actually have to wait for a
	443	* gss upcall (or any calls to the client); this just creates a
	444	* non-uptodate cred which the rpc state machine will fill in with
	445	* a refresh_upcall later.
	446	*/
	447	return rpcauth_lookup_credcache(cb->cb_client->cl_auth, &acred,
	448	RPCAUTH_LOOKUP_NEW);
	449	}
	450
e300a63c	451	void do_probe_callback(struct nfs4_client *clp)
a601caed	452	{
c237dc03	453	struct nfs4_cb_conn *cb = &clp->cl_cb_conn;
a601caed BF	454	struct rpc_message msg = {
	455	.rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL],
	456	.rpc_argp = clp,
	457	};
3cef9ab2	458	struct rpc_cred *cred;
a601caed BF	459	int status;
a601caed BF	460
3cef9ab2 BF	461	cred = lookup_cb_cred(cb);
	462	if (IS_ERR(cred)) {
	463	status = PTR_ERR(cred);
	464	goto out;
	465	}
	466	cb->cb_cred = cred;
	467	msg.rpc_cred = cb->cb_cred;
e300a63c BF	468	status = rpc_call_async(cb->cb_client, &msg, RPC_TASK_SOFT,
e300a63c BF	469	&nfsd4_cb_probe_ops, (void *)clp);
3cef9ab2	470	out:
e300a63c	471	if (status) {
ecdd03b7	472	warn_no_callback_path(clp, status);
e300a63c BF	473	put_nfs4_client(clp);
e300a63c BF	474	}
63c86716 BF	475	}
	476
	477	/*
	478	* Set up the callback client and put a NFSPROC4_CB_NULL on the wire...
	479	*/
	480	void
	481	nfsd4_probe_callback(struct nfs4_client *clp)
	482	{
ecdd03b7	483	int status;
63c86716	484
c237dc03	485	BUG_ON(atomic_read(&clp->cl_cb_conn.cb_set));
63c86716	486
ecdd03b7 BF	487	status = setup_callback_client(clp);
	488	if (status) {
	489	warn_no_callback_path(clp, status);
	490	return;
	491	}
	492
541e0e09	493	/* the task holds a reference to the nfs4_client struct */
1da177e4 LT	494	atomic_inc(&clp->cl_count);
1da177e4 LT	495
e300a63c	496	do_probe_callback(clp);
1da177e4 LT	497	}
1da177e4 LT	498
63e4863f BF	499	static void nfsd4_cb_recall_done(struct rpc_task task, void calldata)
	500	{
	501	struct nfs4_delegation *dp = calldata;
	502	struct nfs4_client *clp = dp->dl_client;
	503
	504	switch (task->tk_status) {
	505	case -EIO:
	506	/* Network partition? */
	507	atomic_set(&clp->cl_cb_conn.cb_set, 0);
	508	warn_no_callback_path(clp, task->tk_status);
	509	case -EBADHANDLE:
	510	case -NFS4ERR_BAD_STATEID:
	511	/* Race: client probably got cb_recall
	512	* before open reply granting delegation */
	513	break;
	514	default:
	515	/* success, or error we can't handle */
	516	return;
	517	}
	518	if (dp->dl_retries--) {
	519	rpc_delay(task, 2*HZ);
	520	task->tk_status = 0;
	521	rpc_restart_call(task);
	522	} else {
	523	atomic_set(&clp->cl_cb_conn.cb_set, 0);
	524	warn_no_callback_path(clp, task->tk_status);
	525	}
	526	}
	527
	528	static void nfsd4_cb_recall_release(void *calldata)
	529	{
	530	struct nfs4_delegation *dp = calldata;
	531	struct nfs4_client *clp = dp->dl_client;
	532
	533	nfs4_put_delegation(dp);
	534	put_nfs4_client(clp);
	535	}
	536
	537	static const struct rpc_call_ops nfsd4_cb_recall_ops = {
	538	.rpc_call_done = nfsd4_cb_recall_done,
	539	.rpc_release = nfsd4_cb_recall_release,
	540	};
	541
1da177e4 LT	542	/*
1da177e4 LT	543	* called with dp->dl_count inc'ed.
1da177e4 LT	544	*/
	545	void
	546	nfsd4_cb_recall(struct nfs4_delegation *dp)
	547	{
	548	struct nfs4_client *clp = dp->dl_client;
c237dc03	549	struct rpc_clnt *clnt = clp->cl_cb_conn.cb_client;
1da177e4 LT	550	struct rpc_message msg = {
1da177e4 LT	551	.rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_RECALL],
b53d40c5	552	.rpc_argp = dp,
c237dc03	553	.rpc_cred = clp->cl_cb_conn.cb_cred
1da177e4	554	};
63e4863f	555	int status;
1da177e4	556
3aea09dc	557	dp->dl_retries = 1;
63e4863f BF	558	status = rpc_call_async(clnt, &msg, RPC_TASK_SOFT,
	559	&nfsd4_cb_recall_ops, dp);
	560	if (status) {
	561	put_nfs4_client(clp);
	562	nfs4_put_delegation(dp);
1da177e4	563	}
1da177e4	564	}