xref: /kernel/linux/linux-6.6/fs/nfsd/nfs4callback.c (revision 62306a36)
1/*
2 *  Copyright (c) 2001 The Regents of the University of Michigan.
3 *  All rights reserved.
4 *
5 *  Kendrick Smith <kmsmith@umich.edu>
6 *  Andy Adamson <andros@umich.edu>
7 *
8 *  Redistribution and use in source and binary forms, with or without
9 *  modification, are permitted provided that the following conditions
10 *  are met:
11 *
12 *  1. Redistributions of source code must retain the above copyright
13 *     notice, this list of conditions and the following disclaimer.
14 *  2. Redistributions in binary form must reproduce the above copyright
15 *     notice, this list of conditions and the following disclaimer in the
16 *     documentation and/or other materials provided with the distribution.
17 *  3. Neither the name of the University nor the names of its
18 *     contributors may be used to endorse or promote products derived
19 *     from this software without specific prior written permission.
20 *
21 *  THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
22 *  WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
23 *  MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
24 *  DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25 *  FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
26 *  CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
27 *  SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
28 *  BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
29 *  LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
30 *  NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
31 *  SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
32 */
33
34#include <linux/sunrpc/clnt.h>
35#include <linux/sunrpc/xprt.h>
36#include <linux/sunrpc/svc_xprt.h>
37#include <linux/slab.h>
38#include "nfsd.h"
39#include "state.h"
40#include "netns.h"
41#include "trace.h"
42#include "xdr4cb.h"
43#include "xdr4.h"
44
45#define NFSDDBG_FACILITY                NFSDDBG_PROC
46
47static void nfsd4_mark_cb_fault(struct nfs4_client *, int reason);
48
49#define NFSPROC4_CB_NULL 0
50#define NFSPROC4_CB_COMPOUND 1
51
52/* Index of predefined Linux callback client operations */
53
54struct nfs4_cb_compound_hdr {
55	/* args */
56	u32		ident;	/* minorversion 0 only */
57	u32		nops;
58	__be32		*nops_p;
59	u32		minorversion;
60	/* res */
61	int		status;
62};
63
64static __be32 *xdr_encode_empty_array(__be32 *p)
65{
66	*p++ = xdr_zero;
67	return p;
68}
69
70/*
71 * Encode/decode NFSv4 CB basic data types
72 *
73 * Basic NFSv4 callback data types are defined in section 15 of RFC
74 * 3530: "Network File System (NFS) version 4 Protocol" and section
75 * 20 of RFC 5661: "Network File System (NFS) Version 4 Minor Version
76 * 1 Protocol"
77 */
78
79static void encode_uint32(struct xdr_stream *xdr, u32 n)
80{
81	WARN_ON_ONCE(xdr_stream_encode_u32(xdr, n) < 0);
82}
83
84static void encode_bitmap4(struct xdr_stream *xdr, const __u32 *bitmap,
85			   size_t len)
86{
87	WARN_ON_ONCE(xdr_stream_encode_uint32_array(xdr, bitmap, len) < 0);
88}
89
90/*
91 *	nfs_cb_opnum4
92 *
93 *	enum nfs_cb_opnum4 {
94 *		OP_CB_GETATTR		= 3,
95 *		  ...
96 *	};
97 */
98enum nfs_cb_opnum4 {
99	OP_CB_GETATTR			= 3,
100	OP_CB_RECALL			= 4,
101	OP_CB_LAYOUTRECALL		= 5,
102	OP_CB_NOTIFY			= 6,
103	OP_CB_PUSH_DELEG		= 7,
104	OP_CB_RECALL_ANY		= 8,
105	OP_CB_RECALLABLE_OBJ_AVAIL	= 9,
106	OP_CB_RECALL_SLOT		= 10,
107	OP_CB_SEQUENCE			= 11,
108	OP_CB_WANTS_CANCELLED		= 12,
109	OP_CB_NOTIFY_LOCK		= 13,
110	OP_CB_NOTIFY_DEVICEID		= 14,
111	OP_CB_OFFLOAD			= 15,
112	OP_CB_ILLEGAL			= 10044
113};
114
115static void encode_nfs_cb_opnum4(struct xdr_stream *xdr, enum nfs_cb_opnum4 op)
116{
117	__be32 *p;
118
119	p = xdr_reserve_space(xdr, 4);
120	*p = cpu_to_be32(op);
121}
122
123/*
124 * nfs_fh4
125 *
126 *	typedef opaque nfs_fh4<NFS4_FHSIZE>;
127 */
128static void encode_nfs_fh4(struct xdr_stream *xdr, const struct knfsd_fh *fh)
129{
130	u32 length = fh->fh_size;
131	__be32 *p;
132
133	BUG_ON(length > NFS4_FHSIZE);
134	p = xdr_reserve_space(xdr, 4 + length);
135	xdr_encode_opaque(p, &fh->fh_raw, length);
136}
137
138/*
139 * stateid4
140 *
141 *	struct stateid4 {
142 *		uint32_t	seqid;
143 *		opaque		other[12];
144 *	};
145 */
146static void encode_stateid4(struct xdr_stream *xdr, const stateid_t *sid)
147{
148	__be32 *p;
149
150	p = xdr_reserve_space(xdr, NFS4_STATEID_SIZE);
151	*p++ = cpu_to_be32(sid->si_generation);
152	xdr_encode_opaque_fixed(p, &sid->si_opaque, NFS4_STATEID_OTHER_SIZE);
153}
154
155/*
156 * sessionid4
157 *
158 *	typedef opaque sessionid4[NFS4_SESSIONID_SIZE];
159 */
160static void encode_sessionid4(struct xdr_stream *xdr,
161			      const struct nfsd4_session *session)
162{
163	__be32 *p;
164
165	p = xdr_reserve_space(xdr, NFS4_MAX_SESSIONID_LEN);
166	xdr_encode_opaque_fixed(p, session->se_sessionid.data,
167					NFS4_MAX_SESSIONID_LEN);
168}
169
170/*
171 * nfsstat4
172 */
173static const struct {
174	int stat;
175	int errno;
176} nfs_cb_errtbl[] = {
177	{ NFS4_OK,		0		},
178	{ NFS4ERR_PERM,		-EPERM		},
179	{ NFS4ERR_NOENT,	-ENOENT		},
180	{ NFS4ERR_IO,		-EIO		},
181	{ NFS4ERR_NXIO,		-ENXIO		},
182	{ NFS4ERR_ACCESS,	-EACCES		},
183	{ NFS4ERR_EXIST,	-EEXIST		},
184	{ NFS4ERR_XDEV,		-EXDEV		},
185	{ NFS4ERR_NOTDIR,	-ENOTDIR	},
186	{ NFS4ERR_ISDIR,	-EISDIR		},
187	{ NFS4ERR_INVAL,	-EINVAL		},
188	{ NFS4ERR_FBIG,		-EFBIG		},
189	{ NFS4ERR_NOSPC,	-ENOSPC		},
190	{ NFS4ERR_ROFS,		-EROFS		},
191	{ NFS4ERR_MLINK,	-EMLINK		},
192	{ NFS4ERR_NAMETOOLONG,	-ENAMETOOLONG	},
193	{ NFS4ERR_NOTEMPTY,	-ENOTEMPTY	},
194	{ NFS4ERR_DQUOT,	-EDQUOT		},
195	{ NFS4ERR_STALE,	-ESTALE		},
196	{ NFS4ERR_BADHANDLE,	-EBADHANDLE	},
197	{ NFS4ERR_BAD_COOKIE,	-EBADCOOKIE	},
198	{ NFS4ERR_NOTSUPP,	-ENOTSUPP	},
199	{ NFS4ERR_TOOSMALL,	-ETOOSMALL	},
200	{ NFS4ERR_SERVERFAULT,	-ESERVERFAULT	},
201	{ NFS4ERR_BADTYPE,	-EBADTYPE	},
202	{ NFS4ERR_LOCKED,	-EAGAIN		},
203	{ NFS4ERR_RESOURCE,	-EREMOTEIO	},
204	{ NFS4ERR_SYMLINK,	-ELOOP		},
205	{ NFS4ERR_OP_ILLEGAL,	-EOPNOTSUPP	},
206	{ NFS4ERR_DEADLOCK,	-EDEADLK	},
207	{ -1,			-EIO		}
208};
209
210/*
211 * If we cannot translate the error, the recovery routines should
212 * handle it.
213 *
214 * Note: remaining NFSv4 error codes have values > 10000, so should
215 * not conflict with native Linux error codes.
216 */
217static int nfs_cb_stat_to_errno(int status)
218{
219	int i;
220
221	for (i = 0; nfs_cb_errtbl[i].stat != -1; i++) {
222		if (nfs_cb_errtbl[i].stat == status)
223			return nfs_cb_errtbl[i].errno;
224	}
225
226	dprintk("NFSD: Unrecognized NFS CB status value: %u\n", status);
227	return -status;
228}
229
230static int decode_cb_op_status(struct xdr_stream *xdr,
231			       enum nfs_cb_opnum4 expected, int *status)
232{
233	__be32 *p;
234	u32 op;
235
236	p = xdr_inline_decode(xdr, 4 + 4);
237	if (unlikely(p == NULL))
238		goto out_overflow;
239	op = be32_to_cpup(p++);
240	if (unlikely(op != expected))
241		goto out_unexpected;
242	*status = nfs_cb_stat_to_errno(be32_to_cpup(p));
243	return 0;
244out_overflow:
245	return -EIO;
246out_unexpected:
247	dprintk("NFSD: Callback server returned operation %d but "
248		"we issued a request for %d\n", op, expected);
249	return -EIO;
250}
251
252/*
253 * CB_COMPOUND4args
254 *
255 *	struct CB_COMPOUND4args {
256 *		utf8str_cs	tag;
257 *		uint32_t	minorversion;
258 *		uint32_t	callback_ident;
259 *		nfs_cb_argop4	argarray<>;
260 *	};
261*/
262static void encode_cb_compound4args(struct xdr_stream *xdr,
263				    struct nfs4_cb_compound_hdr *hdr)
264{
265	__be32 * p;
266
267	p = xdr_reserve_space(xdr, 4 + 4 + 4 + 4);
268	p = xdr_encode_empty_array(p);		/* empty tag */
269	*p++ = cpu_to_be32(hdr->minorversion);
270	*p++ = cpu_to_be32(hdr->ident);
271
272	hdr->nops_p = p;
273	*p = cpu_to_be32(hdr->nops);		/* argarray element count */
274}
275
276/*
277 * Update argarray element count
278 */
279static void encode_cb_nops(struct nfs4_cb_compound_hdr *hdr)
280{
281	BUG_ON(hdr->nops > NFS4_MAX_BACK_CHANNEL_OPS);
282	*hdr->nops_p = cpu_to_be32(hdr->nops);
283}
284
285/*
286 * CB_COMPOUND4res
287 *
288 *	struct CB_COMPOUND4res {
289 *		nfsstat4	status;
290 *		utf8str_cs	tag;
291 *		nfs_cb_resop4	resarray<>;
292 *	};
293 */
294static int decode_cb_compound4res(struct xdr_stream *xdr,
295				  struct nfs4_cb_compound_hdr *hdr)
296{
297	u32 length;
298	__be32 *p;
299
300	p = xdr_inline_decode(xdr, 4 + 4);
301	if (unlikely(p == NULL))
302		goto out_overflow;
303	hdr->status = be32_to_cpup(p++);
304	/* Ignore the tag */
305	length = be32_to_cpup(p++);
306	p = xdr_inline_decode(xdr, length + 4);
307	if (unlikely(p == NULL))
308		goto out_overflow;
309	p += XDR_QUADLEN(length);
310	hdr->nops = be32_to_cpup(p);
311	return 0;
312out_overflow:
313	return -EIO;
314}
315
316/*
317 * CB_RECALL4args
318 *
319 *	struct CB_RECALL4args {
320 *		stateid4	stateid;
321 *		bool		truncate;
322 *		nfs_fh4		fh;
323 *	};
324 */
325static void encode_cb_recall4args(struct xdr_stream *xdr,
326				  const struct nfs4_delegation *dp,
327				  struct nfs4_cb_compound_hdr *hdr)
328{
329	__be32 *p;
330
331	encode_nfs_cb_opnum4(xdr, OP_CB_RECALL);
332	encode_stateid4(xdr, &dp->dl_stid.sc_stateid);
333
334	p = xdr_reserve_space(xdr, 4);
335	*p++ = xdr_zero;			/* truncate */
336
337	encode_nfs_fh4(xdr, &dp->dl_stid.sc_file->fi_fhandle);
338
339	hdr->nops++;
340}
341
342/*
343 * CB_RECALLANY4args
344 *
345 *	struct CB_RECALLANY4args {
346 *		uint32_t	craa_objects_to_keep;
347 *		bitmap4		craa_type_mask;
348 *	};
349 */
350static void
351encode_cb_recallany4args(struct xdr_stream *xdr,
352	struct nfs4_cb_compound_hdr *hdr, struct nfsd4_cb_recall_any *ra)
353{
354	encode_nfs_cb_opnum4(xdr, OP_CB_RECALL_ANY);
355	encode_uint32(xdr, ra->ra_keep);
356	encode_bitmap4(xdr, ra->ra_bmval, ARRAY_SIZE(ra->ra_bmval));
357	hdr->nops++;
358}
359
360/*
361 * CB_SEQUENCE4args
362 *
363 *	struct CB_SEQUENCE4args {
364 *		sessionid4		csa_sessionid;
365 *		sequenceid4		csa_sequenceid;
366 *		slotid4			csa_slotid;
367 *		slotid4			csa_highest_slotid;
368 *		bool			csa_cachethis;
369 *		referring_call_list4	csa_referring_call_lists<>;
370 *	};
371 */
372static void encode_cb_sequence4args(struct xdr_stream *xdr,
373				    const struct nfsd4_callback *cb,
374				    struct nfs4_cb_compound_hdr *hdr)
375{
376	struct nfsd4_session *session = cb->cb_clp->cl_cb_session;
377	__be32 *p;
378
379	if (hdr->minorversion == 0)
380		return;
381
382	encode_nfs_cb_opnum4(xdr, OP_CB_SEQUENCE);
383	encode_sessionid4(xdr, session);
384
385	p = xdr_reserve_space(xdr, 4 + 4 + 4 + 4 + 4);
386	*p++ = cpu_to_be32(session->se_cb_seq_nr);	/* csa_sequenceid */
387	*p++ = xdr_zero;			/* csa_slotid */
388	*p++ = xdr_zero;			/* csa_highest_slotid */
389	*p++ = xdr_zero;			/* csa_cachethis */
390	xdr_encode_empty_array(p);		/* csa_referring_call_lists */
391
392	hdr->nops++;
393}
394
395/*
396 * CB_SEQUENCE4resok
397 *
398 *	struct CB_SEQUENCE4resok {
399 *		sessionid4	csr_sessionid;
400 *		sequenceid4	csr_sequenceid;
401 *		slotid4		csr_slotid;
402 *		slotid4		csr_highest_slotid;
403 *		slotid4		csr_target_highest_slotid;
404 *	};
405 *
406 *	union CB_SEQUENCE4res switch (nfsstat4 csr_status) {
407 *	case NFS4_OK:
408 *		CB_SEQUENCE4resok	csr_resok4;
409 *	default:
410 *		void;
411 *	};
412 *
413 * Our current back channel implmentation supports a single backchannel
414 * with a single slot.
415 */
416static int decode_cb_sequence4resok(struct xdr_stream *xdr,
417				    struct nfsd4_callback *cb)
418{
419	struct nfsd4_session *session = cb->cb_clp->cl_cb_session;
420	int status = -ESERVERFAULT;
421	__be32 *p;
422	u32 dummy;
423
424	/*
425	 * If the server returns different values for sessionID, slotID or
426	 * sequence number, the server is looney tunes.
427	 */
428	p = xdr_inline_decode(xdr, NFS4_MAX_SESSIONID_LEN + 4 + 4 + 4 + 4);
429	if (unlikely(p == NULL))
430		goto out_overflow;
431
432	if (memcmp(p, session->se_sessionid.data, NFS4_MAX_SESSIONID_LEN)) {
433		dprintk("NFS: %s Invalid session id\n", __func__);
434		goto out;
435	}
436	p += XDR_QUADLEN(NFS4_MAX_SESSIONID_LEN);
437
438	dummy = be32_to_cpup(p++);
439	if (dummy != session->se_cb_seq_nr) {
440		dprintk("NFS: %s Invalid sequence number\n", __func__);
441		goto out;
442	}
443
444	dummy = be32_to_cpup(p++);
445	if (dummy != 0) {
446		dprintk("NFS: %s Invalid slotid\n", __func__);
447		goto out;
448	}
449
450	/*
451	 * FIXME: process highest slotid and target highest slotid
452	 */
453	status = 0;
454out:
455	cb->cb_seq_status = status;
456	return status;
457out_overflow:
458	status = -EIO;
459	goto out;
460}
461
462static int decode_cb_sequence4res(struct xdr_stream *xdr,
463				  struct nfsd4_callback *cb)
464{
465	int status;
466
467	if (cb->cb_clp->cl_minorversion == 0)
468		return 0;
469
470	status = decode_cb_op_status(xdr, OP_CB_SEQUENCE, &cb->cb_seq_status);
471	if (unlikely(status || cb->cb_seq_status))
472		return status;
473
474	return decode_cb_sequence4resok(xdr, cb);
475}
476
477/*
478 * NFSv4.0 and NFSv4.1 XDR encode functions
479 *
480 * NFSv4.0 callback argument types are defined in section 15 of RFC
481 * 3530: "Network File System (NFS) version 4 Protocol" and section 20
482 * of RFC 5661:  "Network File System (NFS) Version 4 Minor Version 1
483 * Protocol".
484 */
485
486/*
487 * NB: Without this zero space reservation, callbacks over krb5p fail
488 */
489static void nfs4_xdr_enc_cb_null(struct rpc_rqst *req, struct xdr_stream *xdr,
490				 const void *__unused)
491{
492	xdr_reserve_space(xdr, 0);
493}
494
495/*
496 * 20.2. Operation 4: CB_RECALL - Recall a Delegation
497 */
498static void nfs4_xdr_enc_cb_recall(struct rpc_rqst *req, struct xdr_stream *xdr,
499				   const void *data)
500{
501	const struct nfsd4_callback *cb = data;
502	const struct nfs4_delegation *dp = cb_to_delegation(cb);
503	struct nfs4_cb_compound_hdr hdr = {
504		.ident = cb->cb_clp->cl_cb_ident,
505		.minorversion = cb->cb_clp->cl_minorversion,
506	};
507
508	encode_cb_compound4args(xdr, &hdr);
509	encode_cb_sequence4args(xdr, cb, &hdr);
510	encode_cb_recall4args(xdr, dp, &hdr);
511	encode_cb_nops(&hdr);
512}
513
514/*
515 * 20.6. Operation 8: CB_RECALL_ANY - Keep Any N Recallable Objects
516 */
517static void
518nfs4_xdr_enc_cb_recall_any(struct rpc_rqst *req,
519		struct xdr_stream *xdr, const void *data)
520{
521	const struct nfsd4_callback *cb = data;
522	struct nfsd4_cb_recall_any *ra;
523	struct nfs4_cb_compound_hdr hdr = {
524		.ident = cb->cb_clp->cl_cb_ident,
525		.minorversion = cb->cb_clp->cl_minorversion,
526	};
527
528	ra = container_of(cb, struct nfsd4_cb_recall_any, ra_cb);
529	encode_cb_compound4args(xdr, &hdr);
530	encode_cb_sequence4args(xdr, cb, &hdr);
531	encode_cb_recallany4args(xdr, &hdr, ra);
532	encode_cb_nops(&hdr);
533}
534
535/*
536 * NFSv4.0 and NFSv4.1 XDR decode functions
537 *
538 * NFSv4.0 callback result types are defined in section 15 of RFC
539 * 3530: "Network File System (NFS) version 4 Protocol" and section 20
540 * of RFC 5661:  "Network File System (NFS) Version 4 Minor Version 1
541 * Protocol".
542 */
543
544static int nfs4_xdr_dec_cb_null(struct rpc_rqst *req, struct xdr_stream *xdr,
545				void *__unused)
546{
547	return 0;
548}
549
550/*
551 * 20.2. Operation 4: CB_RECALL - Recall a Delegation
552 */
553static int nfs4_xdr_dec_cb_recall(struct rpc_rqst *rqstp,
554				  struct xdr_stream *xdr,
555				  void *data)
556{
557	struct nfsd4_callback *cb = data;
558	struct nfs4_cb_compound_hdr hdr;
559	int status;
560
561	status = decode_cb_compound4res(xdr, &hdr);
562	if (unlikely(status))
563		return status;
564
565	status = decode_cb_sequence4res(xdr, cb);
566	if (unlikely(status || cb->cb_seq_status))
567		return status;
568
569	return decode_cb_op_status(xdr, OP_CB_RECALL, &cb->cb_status);
570}
571
572/*
573 * 20.6. Operation 8: CB_RECALL_ANY - Keep Any N Recallable Objects
574 */
575static int
576nfs4_xdr_dec_cb_recall_any(struct rpc_rqst *rqstp,
577				  struct xdr_stream *xdr,
578				  void *data)
579{
580	struct nfsd4_callback *cb = data;
581	struct nfs4_cb_compound_hdr hdr;
582	int status;
583
584	status = decode_cb_compound4res(xdr, &hdr);
585	if (unlikely(status))
586		return status;
587	status = decode_cb_sequence4res(xdr, cb);
588	if (unlikely(status || cb->cb_seq_status))
589		return status;
590	status =  decode_cb_op_status(xdr, OP_CB_RECALL_ANY, &cb->cb_status);
591	return status;
592}
593
594#ifdef CONFIG_NFSD_PNFS
595/*
596 * CB_LAYOUTRECALL4args
597 *
598 *	struct layoutrecall_file4 {
599 *		nfs_fh4         lor_fh;
600 *		offset4         lor_offset;
601 *		length4         lor_length;
602 *		stateid4        lor_stateid;
603 *	};
604 *
605 *	union layoutrecall4 switch(layoutrecall_type4 lor_recalltype) {
606 *	case LAYOUTRECALL4_FILE:
607 *		layoutrecall_file4 lor_layout;
608 *	case LAYOUTRECALL4_FSID:
609 *		fsid4              lor_fsid;
610 *	case LAYOUTRECALL4_ALL:
611 *		void;
612 *	};
613 *
614 *	struct CB_LAYOUTRECALL4args {
615 *		layouttype4             clora_type;
616 *		layoutiomode4           clora_iomode;
617 *		bool                    clora_changed;
618 *		layoutrecall4           clora_recall;
619 *	};
620 */
621static void encode_cb_layout4args(struct xdr_stream *xdr,
622				  const struct nfs4_layout_stateid *ls,
623				  struct nfs4_cb_compound_hdr *hdr)
624{
625	__be32 *p;
626
627	BUG_ON(hdr->minorversion == 0);
628
629	p = xdr_reserve_space(xdr, 5 * 4);
630	*p++ = cpu_to_be32(OP_CB_LAYOUTRECALL);
631	*p++ = cpu_to_be32(ls->ls_layout_type);
632	*p++ = cpu_to_be32(IOMODE_ANY);
633	*p++ = cpu_to_be32(1);
634	*p = cpu_to_be32(RETURN_FILE);
635
636	encode_nfs_fh4(xdr, &ls->ls_stid.sc_file->fi_fhandle);
637
638	p = xdr_reserve_space(xdr, 2 * 8);
639	p = xdr_encode_hyper(p, 0);
640	xdr_encode_hyper(p, NFS4_MAX_UINT64);
641
642	encode_stateid4(xdr, &ls->ls_recall_sid);
643
644	hdr->nops++;
645}
646
647static void nfs4_xdr_enc_cb_layout(struct rpc_rqst *req,
648				   struct xdr_stream *xdr,
649				   const void *data)
650{
651	const struct nfsd4_callback *cb = data;
652	const struct nfs4_layout_stateid *ls =
653		container_of(cb, struct nfs4_layout_stateid, ls_recall);
654	struct nfs4_cb_compound_hdr hdr = {
655		.ident = 0,
656		.minorversion = cb->cb_clp->cl_minorversion,
657	};
658
659	encode_cb_compound4args(xdr, &hdr);
660	encode_cb_sequence4args(xdr, cb, &hdr);
661	encode_cb_layout4args(xdr, ls, &hdr);
662	encode_cb_nops(&hdr);
663}
664
665static int nfs4_xdr_dec_cb_layout(struct rpc_rqst *rqstp,
666				  struct xdr_stream *xdr,
667				  void *data)
668{
669	struct nfsd4_callback *cb = data;
670	struct nfs4_cb_compound_hdr hdr;
671	int status;
672
673	status = decode_cb_compound4res(xdr, &hdr);
674	if (unlikely(status))
675		return status;
676
677	status = decode_cb_sequence4res(xdr, cb);
678	if (unlikely(status || cb->cb_seq_status))
679		return status;
680
681	return decode_cb_op_status(xdr, OP_CB_LAYOUTRECALL, &cb->cb_status);
682}
683#endif /* CONFIG_NFSD_PNFS */
684
685static void encode_stateowner(struct xdr_stream *xdr, struct nfs4_stateowner *so)
686{
687	__be32	*p;
688
689	p = xdr_reserve_space(xdr, 8 + 4 + so->so_owner.len);
690	p = xdr_encode_opaque_fixed(p, &so->so_client->cl_clientid, 8);
691	xdr_encode_opaque(p, so->so_owner.data, so->so_owner.len);
692}
693
694static void nfs4_xdr_enc_cb_notify_lock(struct rpc_rqst *req,
695					struct xdr_stream *xdr,
696					const void *data)
697{
698	const struct nfsd4_callback *cb = data;
699	const struct nfsd4_blocked_lock *nbl =
700		container_of(cb, struct nfsd4_blocked_lock, nbl_cb);
701	struct nfs4_lockowner *lo = (struct nfs4_lockowner *)nbl->nbl_lock.fl_owner;
702	struct nfs4_cb_compound_hdr hdr = {
703		.ident = 0,
704		.minorversion = cb->cb_clp->cl_minorversion,
705	};
706
707	__be32 *p;
708
709	BUG_ON(hdr.minorversion == 0);
710
711	encode_cb_compound4args(xdr, &hdr);
712	encode_cb_sequence4args(xdr, cb, &hdr);
713
714	p = xdr_reserve_space(xdr, 4);
715	*p = cpu_to_be32(OP_CB_NOTIFY_LOCK);
716	encode_nfs_fh4(xdr, &nbl->nbl_fh);
717	encode_stateowner(xdr, &lo->lo_owner);
718	hdr.nops++;
719
720	encode_cb_nops(&hdr);
721}
722
723static int nfs4_xdr_dec_cb_notify_lock(struct rpc_rqst *rqstp,
724					struct xdr_stream *xdr,
725					void *data)
726{
727	struct nfsd4_callback *cb = data;
728	struct nfs4_cb_compound_hdr hdr;
729	int status;
730
731	status = decode_cb_compound4res(xdr, &hdr);
732	if (unlikely(status))
733		return status;
734
735	status = decode_cb_sequence4res(xdr, cb);
736	if (unlikely(status || cb->cb_seq_status))
737		return status;
738
739	return decode_cb_op_status(xdr, OP_CB_NOTIFY_LOCK, &cb->cb_status);
740}
741
742/*
743 * struct write_response4 {
744 *	stateid4	wr_callback_id<1>;
745 *	length4		wr_count;
746 *	stable_how4	wr_committed;
747 *	verifier4	wr_writeverf;
748 * };
749 * union offload_info4 switch (nfsstat4 coa_status) {
750 *	case NFS4_OK:
751 *		write_response4	coa_resok4;
752 *	default:
753 *		length4		coa_bytes_copied;
754 * };
755 * struct CB_OFFLOAD4args {
756 *	nfs_fh4		coa_fh;
757 *	stateid4	coa_stateid;
758 *	offload_info4	coa_offload_info;
759 * };
760 */
761static void encode_offload_info4(struct xdr_stream *xdr,
762				 const struct nfsd4_cb_offload *cbo)
763{
764	__be32 *p;
765
766	p = xdr_reserve_space(xdr, 4);
767	*p = cbo->co_nfserr;
768	switch (cbo->co_nfserr) {
769	case nfs_ok:
770		p = xdr_reserve_space(xdr, 4 + 8 + 4 + NFS4_VERIFIER_SIZE);
771		p = xdr_encode_empty_array(p);
772		p = xdr_encode_hyper(p, cbo->co_res.wr_bytes_written);
773		*p++ = cpu_to_be32(cbo->co_res.wr_stable_how);
774		p = xdr_encode_opaque_fixed(p, cbo->co_res.wr_verifier.data,
775					    NFS4_VERIFIER_SIZE);
776		break;
777	default:
778		p = xdr_reserve_space(xdr, 8);
779		/* We always return success if bytes were written */
780		p = xdr_encode_hyper(p, 0);
781	}
782}
783
784static void encode_cb_offload4args(struct xdr_stream *xdr,
785				   const struct nfsd4_cb_offload *cbo,
786				   struct nfs4_cb_compound_hdr *hdr)
787{
788	__be32 *p;
789
790	p = xdr_reserve_space(xdr, 4);
791	*p = cpu_to_be32(OP_CB_OFFLOAD);
792	encode_nfs_fh4(xdr, &cbo->co_fh);
793	encode_stateid4(xdr, &cbo->co_res.cb_stateid);
794	encode_offload_info4(xdr, cbo);
795
796	hdr->nops++;
797}
798
799static void nfs4_xdr_enc_cb_offload(struct rpc_rqst *req,
800				    struct xdr_stream *xdr,
801				    const void *data)
802{
803	const struct nfsd4_callback *cb = data;
804	const struct nfsd4_cb_offload *cbo =
805		container_of(cb, struct nfsd4_cb_offload, co_cb);
806	struct nfs4_cb_compound_hdr hdr = {
807		.ident = 0,
808		.minorversion = cb->cb_clp->cl_minorversion,
809	};
810
811	encode_cb_compound4args(xdr, &hdr);
812	encode_cb_sequence4args(xdr, cb, &hdr);
813	encode_cb_offload4args(xdr, cbo, &hdr);
814	encode_cb_nops(&hdr);
815}
816
817static int nfs4_xdr_dec_cb_offload(struct rpc_rqst *rqstp,
818				   struct xdr_stream *xdr,
819				   void *data)
820{
821	struct nfsd4_callback *cb = data;
822	struct nfs4_cb_compound_hdr hdr;
823	int status;
824
825	status = decode_cb_compound4res(xdr, &hdr);
826	if (unlikely(status))
827		return status;
828
829	status = decode_cb_sequence4res(xdr, cb);
830	if (unlikely(status || cb->cb_seq_status))
831		return status;
832
833	return decode_cb_op_status(xdr, OP_CB_OFFLOAD, &cb->cb_status);
834}
835/*
836 * RPC procedure tables
837 */
838#define PROC(proc, call, argtype, restype)				\
839[NFSPROC4_CLNT_##proc] = {						\
840	.p_proc    = NFSPROC4_CB_##call,				\
841	.p_encode  = nfs4_xdr_enc_##argtype,		\
842	.p_decode  = nfs4_xdr_dec_##restype,				\
843	.p_arglen  = NFS4_enc_##argtype##_sz,				\
844	.p_replen  = NFS4_dec_##restype##_sz,				\
845	.p_statidx = NFSPROC4_CB_##call,				\
846	.p_name    = #proc,						\
847}
848
849static const struct rpc_procinfo nfs4_cb_procedures[] = {
850	PROC(CB_NULL,	NULL,		cb_null,	cb_null),
851	PROC(CB_RECALL,	COMPOUND,	cb_recall,	cb_recall),
852#ifdef CONFIG_NFSD_PNFS
853	PROC(CB_LAYOUT,	COMPOUND,	cb_layout,	cb_layout),
854#endif
855	PROC(CB_NOTIFY_LOCK,	COMPOUND,	cb_notify_lock,	cb_notify_lock),
856	PROC(CB_OFFLOAD,	COMPOUND,	cb_offload,	cb_offload),
857	PROC(CB_RECALL_ANY,	COMPOUND,	cb_recall_any,	cb_recall_any),
858};
859
860static unsigned int nfs4_cb_counts[ARRAY_SIZE(nfs4_cb_procedures)];
861static const struct rpc_version nfs_cb_version4 = {
862/*
863 * Note on the callback rpc program version number: despite language in rfc
864 * 5661 section 18.36.3 requiring servers to use 4 in this field, the
865 * official xdr descriptions for both 4.0 and 4.1 specify version 1, and
866 * in practice that appears to be what implementations use.  The section
867 * 18.36.3 language is expected to be fixed in an erratum.
868 */
869	.number			= 1,
870	.nrprocs		= ARRAY_SIZE(nfs4_cb_procedures),
871	.procs			= nfs4_cb_procedures,
872	.counts			= nfs4_cb_counts,
873};
874
875static const struct rpc_version *nfs_cb_version[2] = {
876	[1] = &nfs_cb_version4,
877};
878
879static const struct rpc_program cb_program;
880
881static struct rpc_stat cb_stats = {
882	.program		= &cb_program
883};
884
885#define NFS4_CALLBACK 0x40000000
886static const struct rpc_program cb_program = {
887	.name			= "nfs4_cb",
888	.number			= NFS4_CALLBACK,
889	.nrvers			= ARRAY_SIZE(nfs_cb_version),
890	.version		= nfs_cb_version,
891	.stats			= &cb_stats,
892	.pipe_dir_name		= "nfsd4_cb",
893};
894
895static int max_cb_time(struct net *net)
896{
897	struct nfsd_net *nn = net_generic(net, nfsd_net_id);
898
899	/*
900	 * nfsd4_lease is set to at most one hour in __nfsd4_write_time,
901	 * so we can use 32-bit math on it. Warn if that assumption
902	 * ever stops being true.
903	 */
904	if (WARN_ON_ONCE(nn->nfsd4_lease > 3600))
905		return 360 * HZ;
906
907	return max(((u32)nn->nfsd4_lease)/10, 1u) * HZ;
908}
909
910static struct workqueue_struct *callback_wq;
911
912static bool nfsd4_queue_cb(struct nfsd4_callback *cb)
913{
914	return queue_work(callback_wq, &cb->cb_work);
915}
916
917static void nfsd41_cb_inflight_begin(struct nfs4_client *clp)
918{
919	atomic_inc(&clp->cl_cb_inflight);
920}
921
922static void nfsd41_cb_inflight_end(struct nfs4_client *clp)
923{
924
925	if (atomic_dec_and_test(&clp->cl_cb_inflight))
926		wake_up_var(&clp->cl_cb_inflight);
927}
928
929static void nfsd41_cb_inflight_wait_complete(struct nfs4_client *clp)
930{
931	wait_var_event(&clp->cl_cb_inflight,
932			!atomic_read(&clp->cl_cb_inflight));
933}
934
935static const struct cred *get_backchannel_cred(struct nfs4_client *clp, struct rpc_clnt *client, struct nfsd4_session *ses)
936{
937	if (clp->cl_minorversion == 0) {
938		client->cl_principal = clp->cl_cred.cr_targ_princ ?
939			clp->cl_cred.cr_targ_princ : "nfs";
940
941		return get_cred(rpc_machine_cred());
942	} else {
943		struct cred *kcred;
944
945		kcred = prepare_kernel_cred(&init_task);
946		if (!kcred)
947			return NULL;
948
949		kcred->fsuid = ses->se_cb_sec.uid;
950		kcred->fsgid = ses->se_cb_sec.gid;
951		return kcred;
952	}
953}
954
955static int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *conn, struct nfsd4_session *ses)
956{
957	int maxtime = max_cb_time(clp->net);
958	struct rpc_timeout	timeparms = {
959		.to_initval	= maxtime,
960		.to_retries	= 0,
961		.to_maxval	= maxtime,
962	};
963	struct rpc_create_args args = {
964		.net		= clp->net,
965		.address	= (struct sockaddr *) &conn->cb_addr,
966		.addrsize	= conn->cb_addrlen,
967		.saddress	= (struct sockaddr *) &conn->cb_saddr,
968		.timeout	= &timeparms,
969		.program	= &cb_program,
970		.version	= 1,
971		.flags		= (RPC_CLNT_CREATE_NOPING | RPC_CLNT_CREATE_QUIET),
972		.cred		= current_cred(),
973	};
974	struct rpc_clnt *client;
975	const struct cred *cred;
976
977	if (clp->cl_minorversion == 0) {
978		if (!clp->cl_cred.cr_principal &&
979		    (clp->cl_cred.cr_flavor >= RPC_AUTH_GSS_KRB5)) {
980			trace_nfsd_cb_setup_err(clp, -EINVAL);
981			return -EINVAL;
982		}
983		args.client_name = clp->cl_cred.cr_principal;
984		args.prognumber	= conn->cb_prog;
985		args.protocol = XPRT_TRANSPORT_TCP;
986		args.authflavor = clp->cl_cred.cr_flavor;
987		clp->cl_cb_ident = conn->cb_ident;
988	} else {
989		if (!conn->cb_xprt)
990			return -EINVAL;
991		clp->cl_cb_session = ses;
992		args.bc_xprt = conn->cb_xprt;
993		args.prognumber = clp->cl_cb_session->se_cb_prog;
994		args.protocol = conn->cb_xprt->xpt_class->xcl_ident |
995				XPRT_TRANSPORT_BC;
996		args.authflavor = ses->se_cb_sec.flavor;
997	}
998	/* Create RPC client */
999	client = rpc_create(&args);
1000	if (IS_ERR(client)) {
1001		trace_nfsd_cb_setup_err(clp, PTR_ERR(client));
1002		return PTR_ERR(client);
1003	}
1004	cred = get_backchannel_cred(clp, client, ses);
1005	if (!cred) {
1006		trace_nfsd_cb_setup_err(clp, -ENOMEM);
1007		rpc_shutdown_client(client);
1008		return -ENOMEM;
1009	}
1010
1011	if (clp->cl_minorversion != 0)
1012		clp->cl_cb_conn.cb_xprt = conn->cb_xprt;
1013	clp->cl_cb_client = client;
1014	clp->cl_cb_cred = cred;
1015	rcu_read_lock();
1016	trace_nfsd_cb_setup(clp, rpc_peeraddr2str(client, RPC_DISPLAY_NETID),
1017			    args.authflavor);
1018	rcu_read_unlock();
1019	return 0;
1020}
1021
1022static void nfsd4_mark_cb_state(struct nfs4_client *clp, int newstate)
1023{
1024	if (clp->cl_cb_state != newstate) {
1025		clp->cl_cb_state = newstate;
1026		trace_nfsd_cb_state(clp);
1027	}
1028}
1029
1030static void nfsd4_mark_cb_down(struct nfs4_client *clp, int reason)
1031{
1032	if (test_bit(NFSD4_CLIENT_CB_UPDATE, &clp->cl_flags))
1033		return;
1034	nfsd4_mark_cb_state(clp, NFSD4_CB_DOWN);
1035}
1036
1037static void nfsd4_mark_cb_fault(struct nfs4_client *clp, int reason)
1038{
1039	if (test_bit(NFSD4_CLIENT_CB_UPDATE, &clp->cl_flags))
1040		return;
1041	nfsd4_mark_cb_state(clp, NFSD4_CB_FAULT);
1042}
1043
1044static void nfsd4_cb_probe_done(struct rpc_task *task, void *calldata)
1045{
1046	struct nfs4_client *clp = container_of(calldata, struct nfs4_client, cl_cb_null);
1047
1048	if (task->tk_status)
1049		nfsd4_mark_cb_down(clp, task->tk_status);
1050	else
1051		nfsd4_mark_cb_state(clp, NFSD4_CB_UP);
1052}
1053
1054static void nfsd4_cb_probe_release(void *calldata)
1055{
1056	struct nfs4_client *clp = container_of(calldata, struct nfs4_client, cl_cb_null);
1057
1058	nfsd41_cb_inflight_end(clp);
1059
1060}
1061
1062static const struct rpc_call_ops nfsd4_cb_probe_ops = {
1063	/* XXX: release method to ensure we set the cb channel down if
1064	 * necessary on early failure? */
1065	.rpc_call_done = nfsd4_cb_probe_done,
1066	.rpc_release = nfsd4_cb_probe_release,
1067};
1068
1069/*
1070 * Poke the callback thread to process any updates to the callback
1071 * parameters, and send a null probe.
1072 */
1073void nfsd4_probe_callback(struct nfs4_client *clp)
1074{
1075	trace_nfsd_cb_probe(clp);
1076	nfsd4_mark_cb_state(clp, NFSD4_CB_UNKNOWN);
1077	set_bit(NFSD4_CLIENT_CB_UPDATE, &clp->cl_flags);
1078	nfsd4_run_cb(&clp->cl_cb_null);
1079}
1080
1081void nfsd4_probe_callback_sync(struct nfs4_client *clp)
1082{
1083	nfsd4_probe_callback(clp);
1084	flush_workqueue(callback_wq);
1085}
1086
1087void nfsd4_change_callback(struct nfs4_client *clp, struct nfs4_cb_conn *conn)
1088{
1089	nfsd4_mark_cb_state(clp, NFSD4_CB_UNKNOWN);
1090	spin_lock(&clp->cl_lock);
1091	memcpy(&clp->cl_cb_conn, conn, sizeof(struct nfs4_cb_conn));
1092	spin_unlock(&clp->cl_lock);
1093}
1094
1095/*
1096 * There's currently a single callback channel slot.
1097 * If the slot is available, then mark it busy.  Otherwise, set the
1098 * thread for sleeping on the callback RPC wait queue.
1099 */
1100static bool nfsd41_cb_get_slot(struct nfsd4_callback *cb, struct rpc_task *task)
1101{
1102	struct nfs4_client *clp = cb->cb_clp;
1103
1104	if (!cb->cb_holds_slot &&
1105	    test_and_set_bit(0, &clp->cl_cb_slot_busy) != 0) {
1106		rpc_sleep_on(&clp->cl_cb_waitq, task, NULL);
1107		/* Race breaker */
1108		if (test_and_set_bit(0, &clp->cl_cb_slot_busy) != 0) {
1109			dprintk("%s slot is busy\n", __func__);
1110			return false;
1111		}
1112		rpc_wake_up_queued_task(&clp->cl_cb_waitq, task);
1113	}
1114	cb->cb_holds_slot = true;
1115	return true;
1116}
1117
1118static void nfsd41_cb_release_slot(struct nfsd4_callback *cb)
1119{
1120	struct nfs4_client *clp = cb->cb_clp;
1121
1122	if (cb->cb_holds_slot) {
1123		cb->cb_holds_slot = false;
1124		clear_bit(0, &clp->cl_cb_slot_busy);
1125		rpc_wake_up_next(&clp->cl_cb_waitq);
1126	}
1127}
1128
1129static void nfsd41_destroy_cb(struct nfsd4_callback *cb)
1130{
1131	struct nfs4_client *clp = cb->cb_clp;
1132
1133	nfsd41_cb_release_slot(cb);
1134	if (cb->cb_ops && cb->cb_ops->release)
1135		cb->cb_ops->release(cb);
1136	nfsd41_cb_inflight_end(clp);
1137}
1138
1139/*
1140 * TODO: cb_sequence should support referring call lists, cachethis, multiple
1141 * slots, and mark callback channel down on communication errors.
1142 */
1143static void nfsd4_cb_prepare(struct rpc_task *task, void *calldata)
1144{
1145	struct nfsd4_callback *cb = calldata;
1146	struct nfs4_client *clp = cb->cb_clp;
1147	u32 minorversion = clp->cl_minorversion;
1148
1149	/*
1150	 * cb_seq_status is only set in decode_cb_sequence4res,
1151	 * and so will remain 1 if an rpc level failure occurs.
1152	 */
1153	cb->cb_seq_status = 1;
1154	cb->cb_status = 0;
1155	if (minorversion && !nfsd41_cb_get_slot(cb, task))
1156		return;
1157	rpc_call_start(task);
1158}
1159
1160static bool nfsd4_cb_sequence_done(struct rpc_task *task, struct nfsd4_callback *cb)
1161{
1162	struct nfs4_client *clp = cb->cb_clp;
1163	struct nfsd4_session *session = clp->cl_cb_session;
1164	bool ret = true;
1165
1166	if (!clp->cl_minorversion) {
1167		/*
1168		 * If the backchannel connection was shut down while this
1169		 * task was queued, we need to resubmit it after setting up
1170		 * a new backchannel connection.
1171		 *
1172		 * Note that if we lost our callback connection permanently
1173		 * the submission code will error out, so we don't need to
1174		 * handle that case here.
1175		 */
1176		if (RPC_SIGNALLED(task))
1177			goto need_restart;
1178
1179		return true;
1180	}
1181
1182	if (!cb->cb_holds_slot)
1183		goto need_restart;
1184
1185	switch (cb->cb_seq_status) {
1186	case 0:
1187		/*
1188		 * No need for lock, access serialized in nfsd4_cb_prepare
1189		 *
1190		 * RFC5661 20.9.3
1191		 * If CB_SEQUENCE returns an error, then the state of the slot
1192		 * (sequence ID, cached reply) MUST NOT change.
1193		 */
1194		++session->se_cb_seq_nr;
1195		break;
1196	case -ESERVERFAULT:
1197		++session->se_cb_seq_nr;
1198		fallthrough;
1199	case 1:
1200	case -NFS4ERR_BADSESSION:
1201		nfsd4_mark_cb_fault(cb->cb_clp, cb->cb_seq_status);
1202		ret = false;
1203		break;
1204	case -NFS4ERR_DELAY:
1205		if (!rpc_restart_call(task))
1206			goto out;
1207
1208		rpc_delay(task, 2 * HZ);
1209		return false;
1210	case -NFS4ERR_BADSLOT:
1211		goto retry_nowait;
1212	case -NFS4ERR_SEQ_MISORDERED:
1213		if (session->se_cb_seq_nr != 1) {
1214			session->se_cb_seq_nr = 1;
1215			goto retry_nowait;
1216		}
1217		break;
1218	default:
1219		nfsd4_mark_cb_fault(cb->cb_clp, cb->cb_seq_status);
1220		dprintk("%s: unprocessed error %d\n", __func__,
1221			cb->cb_seq_status);
1222	}
1223
1224	nfsd41_cb_release_slot(cb);
1225	dprintk("%s: freed slot, new seqid=%d\n", __func__,
1226		clp->cl_cb_session->se_cb_seq_nr);
1227
1228	if (RPC_SIGNALLED(task))
1229		goto need_restart;
1230out:
1231	return ret;
1232retry_nowait:
1233	if (rpc_restart_call_prepare(task))
1234		ret = false;
1235	goto out;
1236need_restart:
1237	if (!test_bit(NFSD4_CLIENT_CB_KILL, &clp->cl_flags)) {
1238		task->tk_status = 0;
1239		cb->cb_need_restart = true;
1240	}
1241	return false;
1242}
1243
1244static void nfsd4_cb_done(struct rpc_task *task, void *calldata)
1245{
1246	struct nfsd4_callback *cb = calldata;
1247	struct nfs4_client *clp = cb->cb_clp;
1248
1249	if (!nfsd4_cb_sequence_done(task, cb))
1250		return;
1251
1252	if (cb->cb_status) {
1253		WARN_ON_ONCE(task->tk_status);
1254		task->tk_status = cb->cb_status;
1255	}
1256
1257	switch (cb->cb_ops->done(cb, task)) {
1258	case 0:
1259		task->tk_status = 0;
1260		rpc_restart_call_prepare(task);
1261		return;
1262	case 1:
1263		switch (task->tk_status) {
1264		case -EIO:
1265		case -ETIMEDOUT:
1266		case -EACCES:
1267			nfsd4_mark_cb_down(clp, task->tk_status);
1268		}
1269		break;
1270	default:
1271		BUG();
1272	}
1273}
1274
1275static void nfsd4_cb_release(void *calldata)
1276{
1277	struct nfsd4_callback *cb = calldata;
1278
1279	if (cb->cb_need_restart)
1280		nfsd4_queue_cb(cb);
1281	else
1282		nfsd41_destroy_cb(cb);
1283
1284}
1285
1286static const struct rpc_call_ops nfsd4_cb_ops = {
1287	.rpc_call_prepare = nfsd4_cb_prepare,
1288	.rpc_call_done = nfsd4_cb_done,
1289	.rpc_release = nfsd4_cb_release,
1290};
1291
1292int nfsd4_create_callback_queue(void)
1293{
1294	callback_wq = alloc_ordered_workqueue("nfsd4_callbacks", 0);
1295	if (!callback_wq)
1296		return -ENOMEM;
1297	return 0;
1298}
1299
1300void nfsd4_destroy_callback_queue(void)
1301{
1302	destroy_workqueue(callback_wq);
1303}
1304
1305/* must be called under the state lock */
1306void nfsd4_shutdown_callback(struct nfs4_client *clp)
1307{
1308	if (clp->cl_cb_state != NFSD4_CB_UNKNOWN)
1309		trace_nfsd_cb_shutdown(clp);
1310
1311	set_bit(NFSD4_CLIENT_CB_KILL, &clp->cl_flags);
1312	/*
1313	 * Note this won't actually result in a null callback;
1314	 * instead, nfsd4_run_cb_null() will detect the killed
1315	 * client, destroy the rpc client, and stop:
1316	 */
1317	nfsd4_run_cb(&clp->cl_cb_null);
1318	flush_workqueue(callback_wq);
1319	nfsd41_cb_inflight_wait_complete(clp);
1320}
1321
1322/* requires cl_lock: */
1323static struct nfsd4_conn * __nfsd4_find_backchannel(struct nfs4_client *clp)
1324{
1325	struct nfsd4_session *s;
1326	struct nfsd4_conn *c;
1327
1328	list_for_each_entry(s, &clp->cl_sessions, se_perclnt) {
1329		list_for_each_entry(c, &s->se_conns, cn_persession) {
1330			if (c->cn_flags & NFS4_CDFC4_BACK)
1331				return c;
1332		}
1333	}
1334	return NULL;
1335}
1336
1337/*
1338 * Note there isn't a lot of locking in this code; instead we depend on
1339 * the fact that it is run from the callback_wq, which won't run two
1340 * work items at once.  So, for example, callback_wq handles all access
1341 * of cl_cb_client and all calls to rpc_create or rpc_shutdown_client.
1342 */
1343static void nfsd4_process_cb_update(struct nfsd4_callback *cb)
1344{
1345	struct nfs4_cb_conn conn;
1346	struct nfs4_client *clp = cb->cb_clp;
1347	struct nfsd4_session *ses = NULL;
1348	struct nfsd4_conn *c;
1349	int err;
1350
1351	/*
1352	 * This is either an update, or the client dying; in either case,
1353	 * kill the old client:
1354	 */
1355	if (clp->cl_cb_client) {
1356		rpc_shutdown_client(clp->cl_cb_client);
1357		clp->cl_cb_client = NULL;
1358		put_cred(clp->cl_cb_cred);
1359		clp->cl_cb_cred = NULL;
1360	}
1361	if (clp->cl_cb_conn.cb_xprt) {
1362		svc_xprt_put(clp->cl_cb_conn.cb_xprt);
1363		clp->cl_cb_conn.cb_xprt = NULL;
1364	}
1365	if (test_bit(NFSD4_CLIENT_CB_KILL, &clp->cl_flags))
1366		return;
1367	spin_lock(&clp->cl_lock);
1368	/*
1369	 * Only serialized callback code is allowed to clear these
1370	 * flags; main nfsd code can only set them:
1371	 */
1372	BUG_ON(!(clp->cl_flags & NFSD4_CLIENT_CB_FLAG_MASK));
1373	clear_bit(NFSD4_CLIENT_CB_UPDATE, &clp->cl_flags);
1374	memcpy(&conn, &cb->cb_clp->cl_cb_conn, sizeof(struct nfs4_cb_conn));
1375	c = __nfsd4_find_backchannel(clp);
1376	if (c) {
1377		svc_xprt_get(c->cn_xprt);
1378		conn.cb_xprt = c->cn_xprt;
1379		ses = c->cn_session;
1380	}
1381	spin_unlock(&clp->cl_lock);
1382
1383	err = setup_callback_client(clp, &conn, ses);
1384	if (err) {
1385		nfsd4_mark_cb_down(clp, err);
1386		if (c)
1387			svc_xprt_put(c->cn_xprt);
1388		return;
1389	}
1390}
1391
1392static void
1393nfsd4_run_cb_work(struct work_struct *work)
1394{
1395	struct nfsd4_callback *cb =
1396		container_of(work, struct nfsd4_callback, cb_work);
1397	struct nfs4_client *clp = cb->cb_clp;
1398	struct rpc_clnt *clnt;
1399	int flags;
1400
1401	if (cb->cb_need_restart) {
1402		cb->cb_need_restart = false;
1403	} else {
1404		if (cb->cb_ops && cb->cb_ops->prepare)
1405			cb->cb_ops->prepare(cb);
1406	}
1407
1408	if (clp->cl_flags & NFSD4_CLIENT_CB_FLAG_MASK)
1409		nfsd4_process_cb_update(cb);
1410
1411	clnt = clp->cl_cb_client;
1412	if (!clnt) {
1413		/* Callback channel broken, or client killed; give up: */
1414		nfsd41_destroy_cb(cb);
1415		return;
1416	}
1417
1418	/*
1419	 * Don't send probe messages for 4.1 or later.
1420	 */
1421	if (!cb->cb_ops && clp->cl_minorversion) {
1422		nfsd4_mark_cb_state(clp, NFSD4_CB_UP);
1423		nfsd41_destroy_cb(cb);
1424		return;
1425	}
1426
1427	cb->cb_msg.rpc_cred = clp->cl_cb_cred;
1428	flags = clp->cl_minorversion ? RPC_TASK_NOCONNECT : RPC_TASK_SOFTCONN;
1429	rpc_call_async(clnt, &cb->cb_msg, RPC_TASK_SOFT | flags,
1430			cb->cb_ops ? &nfsd4_cb_ops : &nfsd4_cb_probe_ops, cb);
1431}
1432
1433void nfsd4_init_cb(struct nfsd4_callback *cb, struct nfs4_client *clp,
1434		const struct nfsd4_callback_ops *ops, enum nfsd4_cb_op op)
1435{
1436	cb->cb_clp = clp;
1437	cb->cb_msg.rpc_proc = &nfs4_cb_procedures[op];
1438	cb->cb_msg.rpc_argp = cb;
1439	cb->cb_msg.rpc_resp = cb;
1440	cb->cb_ops = ops;
1441	INIT_WORK(&cb->cb_work, nfsd4_run_cb_work);
1442	cb->cb_seq_status = 1;
1443	cb->cb_status = 0;
1444	cb->cb_need_restart = false;
1445	cb->cb_holds_slot = false;
1446}
1447
1448/**
1449 * nfsd4_run_cb - queue up a callback job to run
1450 * @cb: callback to queue
1451 *
1452 * Kick off a callback to do its thing. Returns false if it was already
1453 * on a queue, true otherwise.
1454 */
1455bool nfsd4_run_cb(struct nfsd4_callback *cb)
1456{
1457	struct nfs4_client *clp = cb->cb_clp;
1458	bool queued;
1459
1460	nfsd41_cb_inflight_begin(clp);
1461	queued = nfsd4_queue_cb(cb);
1462	if (!queued)
1463		nfsd41_cb_inflight_end(clp);
1464	return queued;
1465}
1466