nfs4proc.c 100.6 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49
/*
 *  fs/nfs/nfs4proc.c
 *
 *  Client-side procedure declarations for NFSv4.
 *
 *  Copyright (c) 2002 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/mm.h>
#include <linux/utsname.h>
#include <linux/delay.h>
#include <linux/errno.h>
#include <linux/string.h>
#include <linux/sunrpc/clnt.h>
#include <linux/nfs.h>
#include <linux/nfs4.h>
#include <linux/nfs_fs.h>
#include <linux/nfs_page.h>
#include <linux/smp_lock.h>
#include <linux/namei.h>
50
#include <linux/mount.h>
L
Linus Torvalds 已提交
51

52
#include "nfs4_fs.h"
L
Linus Torvalds 已提交
53
#include "delegation.h"
54
#include "iostat.h"
L
Linus Torvalds 已提交
55 56 57 58 59 60

#define NFSDBG_FACILITY		NFSDBG_PROC

#define NFS4_POLL_RETRY_MIN	(1*HZ)
#define NFS4_POLL_RETRY_MAX	(15*HZ)

61
struct nfs4_opendata;
62
static int _nfs4_proc_open(struct nfs4_opendata *data);
L
Linus Torvalds 已提交
63
static int nfs4_do_fsinfo(struct nfs_server *, struct nfs_fh *, struct nfs_fsinfo *);
64
static int nfs4_async_handle_error(struct rpc_task *, const struct nfs_server *);
L
Linus Torvalds 已提交
65
static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry);
66
static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception);
67
static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp);
L
Linus Torvalds 已提交
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108

/* Prevent leaks of NFSv4 errors into userland */
int nfs4_map_errors(int err)
{
	if (err < -1000) {
		dprintk("%s could not handle NFSv4 error %d\n",
				__FUNCTION__, -err);
		return -EIO;
	}
	return err;
}

/*
 * This is our standard bitmap for GETATTR requests.
 */
const u32 nfs4_fattr_bitmap[2] = {
	FATTR4_WORD0_TYPE
	| FATTR4_WORD0_CHANGE
	| FATTR4_WORD0_SIZE
	| FATTR4_WORD0_FSID
	| FATTR4_WORD0_FILEID,
	FATTR4_WORD1_MODE
	| FATTR4_WORD1_NUMLINKS
	| FATTR4_WORD1_OWNER
	| FATTR4_WORD1_OWNER_GROUP
	| FATTR4_WORD1_RAWDEV
	| FATTR4_WORD1_SPACE_USED
	| FATTR4_WORD1_TIME_ACCESS
	| FATTR4_WORD1_TIME_METADATA
	| FATTR4_WORD1_TIME_MODIFY
};

const u32 nfs4_statfs_bitmap[2] = {
	FATTR4_WORD0_FILES_AVAIL
	| FATTR4_WORD0_FILES_FREE
	| FATTR4_WORD0_FILES_TOTAL,
	FATTR4_WORD1_SPACE_AVAIL
	| FATTR4_WORD1_SPACE_FREE
	| FATTR4_WORD1_SPACE_TOTAL
};

109
const u32 nfs4_pathconf_bitmap[2] = {
L
Linus Torvalds 已提交
110 111 112 113 114 115 116 117 118 119 120 121
	FATTR4_WORD0_MAXLINK
	| FATTR4_WORD0_MAXNAME,
	0
};

const u32 nfs4_fsinfo_bitmap[2] = { FATTR4_WORD0_MAXFILESIZE
			| FATTR4_WORD0_MAXREAD
			| FATTR4_WORD0_MAXWRITE
			| FATTR4_WORD0_LEASE_TIME,
			0
};

122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140
const u32 nfs4_fs_locations_bitmap[2] = {
	FATTR4_WORD0_TYPE
	| FATTR4_WORD0_CHANGE
	| FATTR4_WORD0_SIZE
	| FATTR4_WORD0_FSID
	| FATTR4_WORD0_FILEID
	| FATTR4_WORD0_FS_LOCATIONS,
	FATTR4_WORD1_MODE
	| FATTR4_WORD1_NUMLINKS
	| FATTR4_WORD1_OWNER
	| FATTR4_WORD1_OWNER_GROUP
	| FATTR4_WORD1_RAWDEV
	| FATTR4_WORD1_SPACE_USED
	| FATTR4_WORD1_TIME_ACCESS
	| FATTR4_WORD1_TIME_METADATA
	| FATTR4_WORD1_TIME_MODIFY
	| FATTR4_WORD1_MOUNTED_ON_FILEID
};

L
Linus Torvalds 已提交
141 142 143 144 145 146 147
static void nfs4_setup_readdir(u64 cookie, u32 *verifier, struct dentry *dentry,
		struct nfs4_readdir_arg *readdir)
{
	u32 *start, *p;

	BUG_ON(readdir->count < 80);
	if (cookie > 2) {
148
		readdir->cookie = cookie;
L
Linus Torvalds 已提交
149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195
		memcpy(&readdir->verifier, verifier, sizeof(readdir->verifier));
		return;
	}

	readdir->cookie = 0;
	memset(&readdir->verifier, 0, sizeof(readdir->verifier));
	if (cookie == 2)
		return;
	
	/*
	 * NFSv4 servers do not return entries for '.' and '..'
	 * Therefore, we fake these entries here.  We let '.'
	 * have cookie 0 and '..' have cookie 1.  Note that
	 * when talking to the server, we always send cookie 0
	 * instead of 1 or 2.
	 */
	start = p = (u32 *)kmap_atomic(*readdir->pages, KM_USER0);
	
	if (cookie == 0) {
		*p++ = xdr_one;                                  /* next */
		*p++ = xdr_zero;                   /* cookie, first word */
		*p++ = xdr_one;                   /* cookie, second word */
		*p++ = xdr_one;                             /* entry len */
		memcpy(p, ".\0\0\0", 4);                        /* entry */
		p++;
		*p++ = xdr_one;                         /* bitmap length */
		*p++ = htonl(FATTR4_WORD0_FILEID);             /* bitmap */
		*p++ = htonl(8);              /* attribute buffer length */
		p = xdr_encode_hyper(p, dentry->d_inode->i_ino);
	}
	
	*p++ = xdr_one;                                  /* next */
	*p++ = xdr_zero;                   /* cookie, first word */
	*p++ = xdr_two;                   /* cookie, second word */
	*p++ = xdr_two;                             /* entry len */
	memcpy(p, "..\0\0", 4);                         /* entry */
	p++;
	*p++ = xdr_one;                         /* bitmap length */
	*p++ = htonl(FATTR4_WORD0_FILEID);             /* bitmap */
	*p++ = htonl(8);              /* attribute buffer length */
	p = xdr_encode_hyper(p, dentry->d_parent->d_inode->i_ino);

	readdir->pgbase = (char *)p - (char *)start;
	readdir->count -= readdir->pgbase;
	kunmap_atomic(start, KM_USER0);
}

196
static void renew_lease(const struct nfs_server *server, unsigned long timestamp)
L
Linus Torvalds 已提交
197
{
198
	struct nfs_client *clp = server->nfs_client;
L
Linus Torvalds 已提交
199 200 201 202 203 204
	spin_lock(&clp->cl_lock);
	if (time_before(clp->cl_last_renewal,timestamp))
		clp->cl_last_renewal = timestamp;
	spin_unlock(&clp->cl_lock);
}

205
static void update_changeattr(struct inode *dir, struct nfs4_change_info *cinfo)
L
Linus Torvalds 已提交
206
{
207
	struct nfs_inode *nfsi = NFS_I(dir);
L
Linus Torvalds 已提交
208

209 210
	spin_lock(&dir->i_lock);
	nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE|NFS_INO_INVALID_DATA;
L
Linus Torvalds 已提交
211 212
	if (cinfo->before == nfsi->change_attr && cinfo->atomic)
		nfsi->change_attr = cinfo->after;
213
	spin_unlock(&dir->i_lock);
L
Linus Torvalds 已提交
214 215
}

216
struct nfs4_opendata {
217
	atomic_t count;
218 219
	struct nfs_openargs o_arg;
	struct nfs_openres o_res;
220 221
	struct nfs_open_confirmargs c_arg;
	struct nfs_open_confirmres c_res;
222 223 224 225 226 227
	struct nfs_fattr f_attr;
	struct nfs_fattr dir_attr;
	struct dentry *dentry;
	struct dentry *dir;
	struct nfs4_state_owner *owner;
	struct iattr attrs;
228
	unsigned long timestamp;
229 230
	int rpc_status;
	int cancelled;
231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247
};

static struct nfs4_opendata *nfs4_opendata_alloc(struct dentry *dentry,
		struct nfs4_state_owner *sp, int flags,
		const struct iattr *attrs)
{
	struct dentry *parent = dget_parent(dentry);
	struct inode *dir = parent->d_inode;
	struct nfs_server *server = NFS_SERVER(dir);
	struct nfs4_opendata *p;

	p = kzalloc(sizeof(*p), GFP_KERNEL);
	if (p == NULL)
		goto err;
	p->o_arg.seqid = nfs_alloc_seqid(&sp->so_seqid);
	if (p->o_arg.seqid == NULL)
		goto err_free;
248
	atomic_set(&p->count, 1);
249 250 251 252 253 254
	p->dentry = dget(dentry);
	p->dir = parent;
	p->owner = sp;
	atomic_inc(&sp->so_count);
	p->o_arg.fh = NFS_FH(dir);
	p->o_arg.open_flags = flags,
255
	p->o_arg.clientid = server->nfs_client->cl_clientid;
256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273
	p->o_arg.id = sp->so_id;
	p->o_arg.name = &dentry->d_name;
	p->o_arg.server = server;
	p->o_arg.bitmask = server->attr_bitmask;
	p->o_arg.claim = NFS4_OPEN_CLAIM_NULL;
	p->o_res.f_attr = &p->f_attr;
	p->o_res.dir_attr = &p->dir_attr;
	p->o_res.server = server;
	nfs_fattr_init(&p->f_attr);
	nfs_fattr_init(&p->dir_attr);
	if (flags & O_EXCL) {
		u32 *s = (u32 *) p->o_arg.u.verifier.data;
		s[0] = jiffies;
		s[1] = current->pid;
	} else if (flags & O_CREAT) {
		p->o_arg.u.attrs = &p->attrs;
		memcpy(&p->attrs, attrs, sizeof(p->attrs));
	}
274 275 276
	p->c_arg.fh = &p->o_res.fh;
	p->c_arg.stateid = &p->o_res.stateid;
	p->c_arg.seqid = p->o_arg.seqid;
277 278 279 280 281 282 283 284 285 286
	return p;
err_free:
	kfree(p);
err:
	dput(parent);
	return NULL;
}

static void nfs4_opendata_free(struct nfs4_opendata *p)
{
287
	if (p != NULL && atomic_dec_and_test(&p->count)) {
288 289 290 291 292 293 294 295
		nfs_free_seqid(p->o_arg.seqid);
		nfs4_put_state_owner(p->owner);
		dput(p->dir);
		dput(p->dentry);
		kfree(p);
	}
}

296
/* Helper for asynchronous RPC calls */
T
Trond Myklebust 已提交
297
static int nfs4_call_async(struct rpc_clnt *clnt,
298
		const struct rpc_call_ops *tk_ops, void *calldata)
299 300 301
{
	struct rpc_task *task;

302
	if (!(task = rpc_new_task(clnt, RPC_TASK_ASYNC, tk_ops, calldata)))
303 304 305 306 307
		return -ENOMEM;
	rpc_execute(task);
	return 0;
}

308 309 310 311 312 313 314 315 316 317 318
static int nfs4_wait_for_completion_rpc_task(struct rpc_task *task)
{
	sigset_t oldset;
	int ret;

	rpc_clnt_sigmask(task->tk_client, &oldset);
	ret = rpc_wait_for_completion_task(task);
	rpc_clnt_sigunmask(task->tk_client, &oldset);
	return ret;
}

319 320 321 322 323 324 325 326 327 328 329 330 331 332
static inline void update_open_stateflags(struct nfs4_state *state, mode_t open_flags)
{
	switch (open_flags) {
		case FMODE_WRITE:
			state->n_wronly++;
			break;
		case FMODE_READ:
			state->n_rdonly++;
			break;
		case FMODE_READ|FMODE_WRITE:
			state->n_rdwr++;
	}
}

L
Linus Torvalds 已提交
333 334 335 336 337
static void update_open_stateid(struct nfs4_state *state, nfs4_stateid *stateid, int open_flags)
{
	struct inode *inode = state->inode;

	open_flags &= (FMODE_READ|FMODE_WRITE);
338
	/* Protect against nfs4_find_state_byowner() */
339
	spin_lock(&state->owner->so_lock);
L
Linus Torvalds 已提交
340
	spin_lock(&inode->i_lock);
341
	memcpy(&state->stateid, stateid, sizeof(state->stateid));
342
	update_open_stateflags(state, open_flags);
343
	nfs4_state_set_mode_locked(state, state->state | open_flags);
L
Linus Torvalds 已提交
344
	spin_unlock(&inode->i_lock);
345
	spin_unlock(&state->owner->so_lock);
L
Linus Torvalds 已提交
346 347
}

348 349 350 351 352 353 354 355
static struct nfs4_state *nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data)
{
	struct inode *inode;
	struct nfs4_state *state = NULL;

	if (!(data->f_attr.valid & NFS_ATTR_FATTR))
		goto out;
	inode = nfs_fhget(data->dir->d_sb, &data->o_res.fh, &data->f_attr);
356
	if (IS_ERR(inode))
357 358 359 360 361 362 363 364 365 366 367
		goto out;
	state = nfs4_get_open_state(inode, data->owner);
	if (state == NULL)
		goto put_inode;
	update_open_stateid(state, &data->o_res.stateid, data->o_arg.open_flags);
put_inode:
	iput(inode);
out:
	return state;
}

368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460
static struct nfs_open_context *nfs4_state_find_open_context(struct nfs4_state *state)
{
	struct nfs_inode *nfsi = NFS_I(state->inode);
	struct nfs_open_context *ctx;

	spin_lock(&state->inode->i_lock);
	list_for_each_entry(ctx, &nfsi->open_files, list) {
		if (ctx->state != state)
			continue;
		get_nfs_open_context(ctx);
		spin_unlock(&state->inode->i_lock);
		return ctx;
	}
	spin_unlock(&state->inode->i_lock);
	return ERR_PTR(-ENOENT);
}

static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, mode_t openflags, nfs4_stateid *stateid)
{
	int ret;

	opendata->o_arg.open_flags = openflags;
	ret = _nfs4_proc_open(opendata);
	if (ret != 0)
		return ret; 
	memcpy(stateid->data, opendata->o_res.stateid.data,
			sizeof(stateid->data));
	return 0;
}

static int nfs4_open_recover(struct nfs4_opendata *opendata, struct nfs4_state *state)
{
	nfs4_stateid stateid;
	struct nfs4_state *newstate;
	int mode = 0;
	int delegation = 0;
	int ret;

	/* memory barrier prior to reading state->n_* */
	smp_rmb();
	if (state->n_rdwr != 0) {
		ret = nfs4_open_recover_helper(opendata, FMODE_READ|FMODE_WRITE, &stateid);
		if (ret != 0)
			return ret;
		mode |= FMODE_READ|FMODE_WRITE;
		if (opendata->o_res.delegation_type != 0)
			delegation = opendata->o_res.delegation_type;
		smp_rmb();
	}
	if (state->n_wronly != 0) {
		ret = nfs4_open_recover_helper(opendata, FMODE_WRITE, &stateid);
		if (ret != 0)
			return ret;
		mode |= FMODE_WRITE;
		if (opendata->o_res.delegation_type != 0)
			delegation = opendata->o_res.delegation_type;
		smp_rmb();
	}
	if (state->n_rdonly != 0) {
		ret = nfs4_open_recover_helper(opendata, FMODE_READ, &stateid);
		if (ret != 0)
			return ret;
		mode |= FMODE_READ;
	}
	clear_bit(NFS_DELEGATED_STATE, &state->flags);
	if (mode == 0)
		return 0;
	if (opendata->o_res.delegation_type == 0)
		opendata->o_res.delegation_type = delegation;
	opendata->o_arg.open_flags |= mode;
	newstate = nfs4_opendata_to_nfs4_state(opendata);
	if (newstate != NULL) {
		if (opendata->o_res.delegation_type != 0) {
			struct nfs_inode *nfsi = NFS_I(newstate->inode);
			int delegation_flags = 0;
			if (nfsi->delegation)
				delegation_flags = nfsi->delegation->flags;
			if (!(delegation_flags & NFS_DELEGATION_NEED_RECLAIM))
				nfs_inode_set_delegation(newstate->inode,
						opendata->owner->so_cred,
						&opendata->o_res);
			else
				nfs_inode_reclaim_delegation(newstate->inode,
						opendata->owner->so_cred,
						&opendata->o_res);
		}
		nfs4_close_state(newstate, opendata->o_arg.open_flags);
	}
	if (newstate != state)
		return -ESTALE;
	return 0;
}

L
Linus Torvalds 已提交
461 462 463 464
/*
 * OPEN_RECLAIM:
 * 	reclaim state on the server after a reboot.
 */
465
static int _nfs4_do_open_reclaim(struct nfs4_state_owner *sp, struct nfs4_state *state, struct dentry *dentry)
L
Linus Torvalds 已提交
466
{
467 468 469
	struct nfs_delegation *delegation = NFS_I(state->inode)->delegation;
	struct nfs4_opendata *opendata;
	int delegation_type = 0;
L
Linus Torvalds 已提交
470 471 472 473 474 475 476 477 478
	int status;

	if (delegation != NULL) {
		if (!(delegation->flags & NFS_DELEGATION_NEED_RECLAIM)) {
			memcpy(&state->stateid, &delegation->stateid,
					sizeof(state->stateid));
			set_bit(NFS_DELEGATED_STATE, &state->flags);
			return 0;
		}
479
		delegation_type = delegation->type;
L
Linus Torvalds 已提交
480
	}
481 482
	opendata = nfs4_opendata_alloc(dentry, sp, 0, NULL);
	if (opendata == NULL)
483
		return -ENOMEM;
484 485 486 487 488 489
	opendata->o_arg.claim = NFS4_OPEN_CLAIM_PREVIOUS;
	opendata->o_arg.fh = NFS_FH(state->inode);
	nfs_copy_fh(&opendata->o_res.fh, opendata->o_arg.fh);
	opendata->o_arg.u.delegation_type = delegation_type;
	status = nfs4_open_recover(opendata, state);
	nfs4_opendata_free(opendata);
L
Linus Torvalds 已提交
490 491 492
	return status;
}

493
static int nfs4_do_open_reclaim(struct nfs4_state_owner *sp, struct nfs4_state *state, struct dentry *dentry)
L
Linus Torvalds 已提交
494 495 496 497 498
{
	struct nfs_server *server = NFS_SERVER(state->inode);
	struct nfs4_exception exception = { };
	int err;
	do {
499
		err = _nfs4_do_open_reclaim(sp, state, dentry);
500 501 502
		if (err != -NFS4ERR_DELAY)
			break;
		nfs4_handle_exception(server, err, &exception);
L
Linus Torvalds 已提交
503 504 505 506
	} while (exception.retry);
	return err;
}

507 508 509 510 511 512 513 514 515 516 517 518 519
static int nfs4_open_reclaim(struct nfs4_state_owner *sp, struct nfs4_state *state)
{
	struct nfs_open_context *ctx;
	int ret;

	ctx = nfs4_state_find_open_context(state);
	if (IS_ERR(ctx))
		return PTR_ERR(ctx);
	ret = nfs4_do_open_reclaim(sp, state, ctx->dentry);
	put_nfs_open_context(ctx);
	return ret;
}

L
Linus Torvalds 已提交
520 521 522
static int _nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state *state)
{
	struct nfs4_state_owner  *sp  = state->owner;
523
	struct nfs4_opendata *opendata;
524
	int ret;
L
Linus Torvalds 已提交
525 526

	if (!test_bit(NFS_DELEGATED_STATE, &state->flags))
527 528
		return 0;
	opendata = nfs4_opendata_alloc(dentry, sp, 0, NULL);
529
	if (opendata == NULL)
530
		return -ENOMEM;
531 532 533
	opendata->o_arg.claim = NFS4_OPEN_CLAIM_DELEGATE_CUR;
	memcpy(opendata->o_arg.u.delegation.data, state->stateid.data,
			sizeof(opendata->o_arg.u.delegation.data));
534
	ret = nfs4_open_recover(opendata, state);
535
	nfs4_opendata_free(opendata);
536
	return ret;
L
Linus Torvalds 已提交
537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552
}

int nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state *state)
{
	struct nfs4_exception exception = { };
	struct nfs_server *server = NFS_SERVER(dentry->d_inode);
	int err;
	do {
		err = _nfs4_open_delegation_recall(dentry, state);
		switch (err) {
			case 0:
				return err;
			case -NFS4ERR_STALE_CLIENTID:
			case -NFS4ERR_STALE_STATEID:
			case -NFS4ERR_EXPIRED:
				/* Don't recall a delegation if it was lost */
553
				nfs4_schedule_state_recovery(server->nfs_client);
L
Linus Torvalds 已提交
554 555 556 557 558 559 560
				return err;
		}
		err = nfs4_handle_exception(server, err, &exception);
	} while (exception.retry);
	return err;
}

561
static void nfs4_open_confirm_prepare(struct rpc_task *task, void *calldata)
L
Linus Torvalds 已提交
562
{
563 564 565 566 567 568
	struct nfs4_opendata *data = calldata;
	struct  rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_CONFIRM],
		.rpc_argp = &data->c_arg,
		.rpc_resp = &data->c_res,
		.rpc_cred = data->owner->so_cred,
L
Linus Torvalds 已提交
569
	};
570
	data->timestamp = jiffies;
571 572 573 574 575 576 577 578 579 580
	rpc_call_setup(task, &msg, 0);
}

static void nfs4_open_confirm_done(struct rpc_task *task, void *calldata)
{
	struct nfs4_opendata *data = calldata;

	data->rpc_status = task->tk_status;
	if (RPC_ASSASSINATED(task))
		return;
581
	if (data->rpc_status == 0) {
582 583
		memcpy(data->o_res.stateid.data, data->c_res.stateid.data,
				sizeof(data->o_res.stateid.data));
584 585
		renew_lease(data->o_res.server, data->timestamp);
	}
586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621
	nfs_increment_open_seqid(data->rpc_status, data->c_arg.seqid);
	nfs_confirm_seqid(&data->owner->so_seqid, data->rpc_status);
}

static void nfs4_open_confirm_release(void *calldata)
{
	struct nfs4_opendata *data = calldata;
	struct nfs4_state *state = NULL;

	/* If this request hasn't been cancelled, do nothing */
	if (data->cancelled == 0)
		goto out_free;
	/* In case of error, no cleanup! */
	if (data->rpc_status != 0)
		goto out_free;
	nfs_confirm_seqid(&data->owner->so_seqid, 0);
	state = nfs4_opendata_to_nfs4_state(data);
	if (state != NULL)
		nfs4_close_state(state, data->o_arg.open_flags);
out_free:
	nfs4_opendata_free(data);
}

static const struct rpc_call_ops nfs4_open_confirm_ops = {
	.rpc_call_prepare = nfs4_open_confirm_prepare,
	.rpc_call_done = nfs4_open_confirm_done,
	.rpc_release = nfs4_open_confirm_release,
};

/*
 * Note: On error, nfs4_proc_open_confirm will free the struct nfs4_opendata
 */
static int _nfs4_proc_open_confirm(struct nfs4_opendata *data)
{
	struct nfs_server *server = NFS_SERVER(data->dir->d_inode);
	struct rpc_task *task;
L
Linus Torvalds 已提交
622 623
	int status;

624
	atomic_inc(&data->count);
625 626 627 628 629
	/*
	 * If rpc_run_task() ends up calling ->rpc_release(), we
	 * want to ensure that it takes the 'error' code path.
	 */
	data->rpc_status = -ENOMEM;
630
	task = rpc_run_task(server->client, RPC_TASK_ASYNC, &nfs4_open_confirm_ops, data);
631
	if (IS_ERR(task))
632 633 634 635 636 637 638 639
		return PTR_ERR(task);
	status = nfs4_wait_for_completion_rpc_task(task);
	if (status != 0) {
		data->cancelled = 1;
		smp_wmb();
	} else
		status = data->rpc_status;
	rpc_release_task(task);
L
Linus Torvalds 已提交
640 641 642
	return status;
}

643
static void nfs4_open_prepare(struct rpc_task *task, void *calldata)
L
Linus Torvalds 已提交
644
{
645 646
	struct nfs4_opendata *data = calldata;
	struct nfs4_state_owner *sp = data->owner;
L
Linus Torvalds 已提交
647 648
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN],
649 650
		.rpc_argp = &data->o_arg,
		.rpc_resp = &data->o_res,
L
Linus Torvalds 已提交
651 652
		.rpc_cred = sp->so_cred,
	};
653 654 655 656 657 658
	
	if (nfs_wait_on_sequence(data->o_arg.seqid, task) != 0)
		return;
	/* Update sequence id. */
	data->o_arg.id = sp->so_id;
	data->o_arg.clientid = sp->so_client->cl_clientid;
659 660
	if (data->o_arg.claim == NFS4_OPEN_CLAIM_PREVIOUS)
		msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_NOATTR];
661
	data->timestamp = jiffies;
662 663
	rpc_call_setup(task, &msg, 0);
}
L
Linus Torvalds 已提交
664

665 666 667
static void nfs4_open_done(struct rpc_task *task, void *calldata)
{
	struct nfs4_opendata *data = calldata;
L
Linus Torvalds 已提交
668

669 670 671 672 673
	data->rpc_status = task->tk_status;
	if (RPC_ASSASSINATED(task))
		return;
	if (task->tk_status == 0) {
		switch (data->o_res.f_attr->mode & S_IFMT) {
674 675 676
			case S_IFREG:
				break;
			case S_IFLNK:
677
				data->rpc_status = -ELOOP;
678 679
				break;
			case S_IFDIR:
680
				data->rpc_status = -EISDIR;
681 682
				break;
			default:
683
				data->rpc_status = -ENOTDIR;
684
		}
685
		renew_lease(data->o_res.server, data->timestamp);
686
	}
687 688
	nfs_increment_open_seqid(data->rpc_status, data->o_arg.seqid);
}
689

690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730
static void nfs4_open_release(void *calldata)
{
	struct nfs4_opendata *data = calldata;
	struct nfs4_state *state = NULL;

	/* If this request hasn't been cancelled, do nothing */
	if (data->cancelled == 0)
		goto out_free;
	/* In case of error, no cleanup! */
	if (data->rpc_status != 0)
		goto out_free;
	/* In case we need an open_confirm, no cleanup! */
	if (data->o_res.rflags & NFS4_OPEN_RESULT_CONFIRM)
		goto out_free;
	nfs_confirm_seqid(&data->owner->so_seqid, 0);
	state = nfs4_opendata_to_nfs4_state(data);
	if (state != NULL)
		nfs4_close_state(state, data->o_arg.open_flags);
out_free:
	nfs4_opendata_free(data);
}

static const struct rpc_call_ops nfs4_open_ops = {
	.rpc_call_prepare = nfs4_open_prepare,
	.rpc_call_done = nfs4_open_done,
	.rpc_release = nfs4_open_release,
};

/*
 * Note: On error, nfs4_proc_open will free the struct nfs4_opendata
 */
static int _nfs4_proc_open(struct nfs4_opendata *data)
{
	struct inode *dir = data->dir->d_inode;
	struct nfs_server *server = NFS_SERVER(dir);
	struct nfs_openargs *o_arg = &data->o_arg;
	struct nfs_openres *o_res = &data->o_res;
	struct rpc_task *task;
	int status;

	atomic_inc(&data->count);
731 732 733 734 735
	/*
	 * If rpc_run_task() ends up calling ->rpc_release(), we
	 * want to ensure that it takes the 'error' code path.
	 */
	data->rpc_status = -ENOMEM;
736
	task = rpc_run_task(server->client, RPC_TASK_ASYNC, &nfs4_open_ops, data);
737
	if (IS_ERR(task))
738 739 740 741 742 743 744 745
		return PTR_ERR(task);
	status = nfs4_wait_for_completion_rpc_task(task);
	if (status != 0) {
		data->cancelled = 1;
		smp_wmb();
	} else
		status = data->rpc_status;
	rpc_release_task(task);
L
Linus Torvalds 已提交
746
	if (status != 0)
747 748
		return status;

749 750 751 752 753
	if (o_arg->open_flags & O_CREAT) {
		update_changeattr(dir, &o_res->cinfo);
		nfs_post_op_update_inode(dir, o_res->dir_attr);
	} else
		nfs_refresh_inode(dir, o_res->dir_attr);
L
Linus Torvalds 已提交
754
	if(o_res->rflags & NFS4_OPEN_RESULT_CONFIRM) {
755
		status = _nfs4_proc_open_confirm(data);
L
Linus Torvalds 已提交
756
		if (status != 0)
757
			return status;
L
Linus Torvalds 已提交
758
	}
759
	nfs_confirm_seqid(&data->owner->so_seqid, 0);
L
Linus Torvalds 已提交
760
	if (!(o_res->f_attr->valid & NFS_ATTR_FATTR))
761 762
		return server->rpc_ops->getattr(server, &o_res->fh, o_res->f_attr);
	return 0;
L
Linus Torvalds 已提交
763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792
}

static int _nfs4_do_access(struct inode *inode, struct rpc_cred *cred, int openflags)
{
	struct nfs_access_entry cache;
	int mask = 0;
	int status;

	if (openflags & FMODE_READ)
		mask |= MAY_READ;
	if (openflags & FMODE_WRITE)
		mask |= MAY_WRITE;
	status = nfs_access_get_cached(inode, cred, &cache);
	if (status == 0)
		goto out;

	/* Be clever: ask server to check for all possible rights */
	cache.mask = MAY_EXEC | MAY_WRITE | MAY_READ;
	cache.cred = cred;
	cache.jiffies = jiffies;
	status = _nfs4_proc_access(inode, &cache);
	if (status != 0)
		return status;
	nfs_access_add_cache(inode, &cache);
out:
	if ((cache.mask & mask) == mask)
		return 0;
	return -EACCES;
}

793 794
int nfs4_recover_expired_lease(struct nfs_server *server)
{
795
	struct nfs_client *clp = server->nfs_client;
796 797 798 799 800 801

	if (test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
		nfs4_schedule_state_recovery(clp);
	return nfs4_wait_clnt_recover(server->client, clp);
}

L
Linus Torvalds 已提交
802 803 804 805 806 807 808 809 810
/*
 * OPEN_EXPIRED:
 * 	reclaim state on the server after a network partition.
 * 	Assumes caller holds the appropriate lock
 */
static int _nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *state, struct dentry *dentry)
{
	struct inode *inode = state->inode;
	struct nfs_delegation *delegation = NFS_I(inode)->delegation;
811
	struct nfs4_opendata *opendata;
812
	int openflags = state->state & (FMODE_READ|FMODE_WRITE);
813
	int ret;
L
Linus Torvalds 已提交
814 815

	if (delegation != NULL && !(delegation->flags & NFS_DELEGATION_NEED_RECLAIM)) {
816 817 818
		ret = _nfs4_do_access(inode, sp->so_cred, openflags);
		if (ret < 0)
			return ret;
L
Linus Torvalds 已提交
819 820
		memcpy(&state->stateid, &delegation->stateid, sizeof(state->stateid));
		set_bit(NFS_DELEGATED_STATE, &state->flags);
821
		return 0;
L
Linus Torvalds 已提交
822
	}
823
	opendata = nfs4_opendata_alloc(dentry, sp, openflags, NULL);
824
	if (opendata == NULL)
825 826 827 828 829 830
		return -ENOMEM;
	ret = nfs4_open_recover(opendata, state);
	if (ret == -ESTALE) {
		/* Invalidate the state owner so we don't ever use it again */
		nfs4_drop_state_owner(sp);
		d_drop(dentry);
L
Linus Torvalds 已提交
831
	}
832
	nfs4_opendata_free(opendata);
833
	return ret;
L
Linus Torvalds 已提交
834 835
}

836 837 838 839 840 841 842 843 844 845 846 847 848 849
static inline int nfs4_do_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *state, struct dentry *dentry)
{
	struct nfs_server *server = NFS_SERVER(dentry->d_inode);
	struct nfs4_exception exception = { };
	int err;

	do {
		err = _nfs4_open_expired(sp, state, dentry);
		if (err == -NFS4ERR_DELAY)
			nfs4_handle_exception(server, err, &exception);
	} while (exception.retry);
	return err;
}

L
Linus Torvalds 已提交
850 851 852
static int nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *state)
{
	struct nfs_open_context *ctx;
853
	int ret;
L
Linus Torvalds 已提交
854

855 856 857 858 859 860
	ctx = nfs4_state_find_open_context(state);
	if (IS_ERR(ctx))
		return PTR_ERR(ctx);
	ret = nfs4_do_open_expired(sp, state, ctx->dentry);
	put_nfs_open_context(ctx);
	return ret;
L
Linus Torvalds 已提交
861 862 863
}

/*
864
 * Returns a referenced nfs4_state if there is an open delegation on the file
L
Linus Torvalds 已提交
865 866 867 868 869
 */
static int _nfs4_open_delegated(struct inode *inode, int flags, struct rpc_cred *cred, struct nfs4_state **res)
{
	struct nfs_delegation *delegation;
	struct nfs_server *server = NFS_SERVER(inode);
870
	struct nfs_client *clp = server->nfs_client;
L
Linus Torvalds 已提交
871 872 873 874 875 876
	struct nfs_inode *nfsi = NFS_I(inode);
	struct nfs4_state_owner *sp = NULL;
	struct nfs4_state *state = NULL;
	int open_flags = flags & (FMODE_READ|FMODE_WRITE);
	int err;

877 878 879 880 881
	err = -ENOMEM;
	if (!(sp = nfs4_get_state_owner(server, cred))) {
		dprintk("%s: nfs4_get_state_owner failed!\n", __FUNCTION__);
		return err;
	}
882 883
	err = nfs4_recover_expired_lease(server);
	if (err != 0)
884
		goto out_put_state_owner;
L
Linus Torvalds 已提交
885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900
	/* Protect against reboot recovery - NOTE ORDER! */
	down_read(&clp->cl_sem);
	/* Protect against delegation recall */
	down_read(&nfsi->rwsem);
	delegation = NFS_I(inode)->delegation;
	err = -ENOENT;
	if (delegation == NULL || (delegation->type & open_flags) != open_flags)
		goto out_err;
	err = -ENOMEM;
	state = nfs4_get_open_state(inode, sp);
	if (state == NULL)
		goto out_err;

	err = -ENOENT;
	if ((state->state & open_flags) == open_flags) {
		spin_lock(&inode->i_lock);
901
		update_open_stateflags(state, open_flags);
L
Linus Torvalds 已提交
902 903 904
		spin_unlock(&inode->i_lock);
		goto out_ok;
	} else if (state->state != 0)
905
		goto out_put_open_state;
L
Linus Torvalds 已提交
906 907 908 909 910

	lock_kernel();
	err = _nfs4_do_access(inode, cred, open_flags);
	unlock_kernel();
	if (err != 0)
911
		goto out_put_open_state;
L
Linus Torvalds 已提交
912 913 914 915 916 917 918
	set_bit(NFS_DELEGATED_STATE, &state->flags);
	update_open_stateid(state, &delegation->stateid, open_flags);
out_ok:
	nfs4_put_state_owner(sp);
	up_read(&nfsi->rwsem);
	up_read(&clp->cl_sem);
	*res = state;
919 920 921
	return 0;
out_put_open_state:
	nfs4_put_open_state(state);
L
Linus Torvalds 已提交
922 923 924
out_err:
	up_read(&nfsi->rwsem);
	up_read(&clp->cl_sem);
925 926
	if (err != -EACCES)
		nfs_inode_return_delegation(inode);
927 928
out_put_state_owner:
	nfs4_put_state_owner(sp);
L
Linus Torvalds 已提交
929 930 931 932 933 934
	return err;
}

static struct nfs4_state *nfs4_open_delegated(struct inode *inode, int flags, struct rpc_cred *cred)
{
	struct nfs4_exception exception = { };
935
	struct nfs4_state *res = ERR_PTR(-EIO);
L
Linus Torvalds 已提交
936 937 938 939 940 941 942 943 944 945 946 947 948
	int err;

	do {
		err = _nfs4_open_delegated(inode, flags, cred, &res);
		if (err == 0)
			break;
		res = ERR_PTR(nfs4_handle_exception(NFS_SERVER(inode),
					err, &exception));
	} while (exception.retry);
	return res;
}

/*
949
 * Returns a referenced nfs4_state
L
Linus Torvalds 已提交
950 951 952 953 954 955
 */
static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, struct iattr *sattr, struct rpc_cred *cred, struct nfs4_state **res)
{
	struct nfs4_state_owner  *sp;
	struct nfs4_state     *state = NULL;
	struct nfs_server       *server = NFS_SERVER(dir);
956
	struct nfs_client *clp = server->nfs_client;
957
	struct nfs4_opendata *opendata;
L
Linus Torvalds 已提交
958 959 960 961 962 963 964 965
	int                     status;

	/* Protect against reboot recovery conflicts */
	status = -ENOMEM;
	if (!(sp = nfs4_get_state_owner(server, cred))) {
		dprintk("nfs4_do_open: nfs4_get_state_owner failed!\n");
		goto out_err;
	}
966 967
	status = nfs4_recover_expired_lease(server);
	if (status != 0)
968
		goto err_put_state_owner;
969 970
	down_read(&clp->cl_sem);
	status = -ENOMEM;
971 972
	opendata = nfs4_opendata_alloc(dentry, sp, flags, sattr);
	if (opendata == NULL)
973
		goto err_release_rwsem;
L
Linus Torvalds 已提交
974

975
	status = _nfs4_proc_open(opendata);
L
Linus Torvalds 已提交
976
	if (status != 0)
977
		goto err_opendata_free;
L
Linus Torvalds 已提交
978 979

	status = -ENOMEM;
980 981
	state = nfs4_opendata_to_nfs4_state(opendata);
	if (state == NULL)
982 983
		goto err_opendata_free;
	if (opendata->o_res.delegation_type != 0)
984
		nfs_inode_set_delegation(state->inode, cred, &opendata->o_res);
985
	nfs4_opendata_free(opendata);
L
Linus Torvalds 已提交
986 987 988 989
	nfs4_put_state_owner(sp);
	up_read(&clp->cl_sem);
	*res = state;
	return 0;
990 991
err_opendata_free:
	nfs4_opendata_free(opendata);
992 993
err_release_rwsem:
	up_read(&clp->cl_sem);
994 995
err_put_state_owner:
	nfs4_put_state_owner(sp);
L
Linus Torvalds 已提交
996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017
out_err:
	*res = NULL;
	return status;
}


static struct nfs4_state *nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, struct iattr *sattr, struct rpc_cred *cred)
{
	struct nfs4_exception exception = { };
	struct nfs4_state *res;
	int status;

	do {
		status = _nfs4_do_open(dir, dentry, flags, sattr, cred, &res);
		if (status == 0)
			break;
		/* NOTE: BAD_SEQID means the server and client disagree about the
		 * book-keeping w.r.t. state-changing operations
		 * (OPEN/CLOSE/LOCK/LOCKU...)
		 * It is actually a sign of a bug on the client or on the server.
		 *
		 * If we receive a BAD_SEQID error in the particular case of
1018
		 * doing an OPEN, we assume that nfs_increment_open_seqid() will
L
Linus Torvalds 已提交
1019 1020 1021 1022 1023 1024 1025 1026 1027
		 * have unhashed the old state_owner for us, and that we can
		 * therefore safely retry using a new one. We should still warn
		 * the user though...
		 */
		if (status == -NFS4ERR_BAD_SEQID) {
			printk(KERN_WARNING "NFS: v4 server returned a bad sequence-id error!\n");
			exception.retry = 1;
			continue;
		}
1028 1029 1030 1031 1032 1033 1034 1035 1036 1037
		/*
		 * BAD_STATEID on OPEN means that the server cancelled our
		 * state before it received the OPEN_CONFIRM.
		 * Recover by retrying the request as per the discussion
		 * on Page 181 of RFC3530.
		 */
		if (status == -NFS4ERR_BAD_STATEID) {
			exception.retry = 1;
			continue;
		}
L
Linus Torvalds 已提交
1038 1039 1040 1041 1042 1043
		res = ERR_PTR(nfs4_handle_exception(NFS_SERVER(dir),
					status, &exception));
	} while (exception.retry);
	return res;
}

1044 1045
static int _nfs4_do_setattr(struct inode *inode, struct nfs_fattr *fattr,
                struct iattr *sattr, struct nfs4_state *state)
L
Linus Torvalds 已提交
1046
{
1047
	struct nfs_server *server = NFS_SERVER(inode);
L
Linus Torvalds 已提交
1048
        struct nfs_setattrargs  arg = {
1049
                .fh             = NFS_FH(inode),
L
Linus Torvalds 已提交
1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062
                .iap            = sattr,
		.server		= server,
		.bitmask = server->attr_bitmask,
        };
        struct nfs_setattrres  res = {
		.fattr		= fattr,
		.server		= server,
        };
        struct rpc_message msg = {
                .rpc_proc       = &nfs4_procedures[NFSPROC4_CLNT_SETATTR],
                .rpc_argp       = &arg,
                .rpc_resp       = &res,
        };
1063
	unsigned long timestamp = jiffies;
1064
	int status;
L
Linus Torvalds 已提交
1065

1066
	nfs_fattr_init(fattr);
L
Linus Torvalds 已提交
1067

1068 1069 1070
	if (nfs4_copy_delegation_stateid(&arg.stateid, inode)) {
		/* Use that stateid */
	} else if (state != NULL) {
L
Linus Torvalds 已提交
1071
		msg.rpc_cred = state->owner->so_cred;
1072 1073
		nfs4_copy_stateid(&arg.stateid, state, current->files);
	} else
L
Linus Torvalds 已提交
1074 1075
		memcpy(&arg.stateid, &zero_stateid, sizeof(arg.stateid));

1076
	status = rpc_call_sync(server->client, &msg, 0);
1077 1078
	if (status == 0 && state != NULL)
		renew_lease(server, timestamp);
1079
	return status;
L
Linus Torvalds 已提交
1080 1081
}

1082 1083
static int nfs4_do_setattr(struct inode *inode, struct nfs_fattr *fattr,
                struct iattr *sattr, struct nfs4_state *state)
L
Linus Torvalds 已提交
1084
{
1085
	struct nfs_server *server = NFS_SERVER(inode);
L
Linus Torvalds 已提交
1086 1087 1088 1089
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(server,
1090
				_nfs4_do_setattr(inode, fattr, sattr, state),
L
Linus Torvalds 已提交
1091 1092 1093 1094 1095 1096 1097 1098 1099 1100
				&exception);
	} while (exception.retry);
	return err;
}

struct nfs4_closedata {
	struct inode *inode;
	struct nfs4_state *state;
	struct nfs_closeargs arg;
	struct nfs_closeres res;
1101
	struct nfs_fattr fattr;
1102
	unsigned long timestamp;
L
Linus Torvalds 已提交
1103 1104
};

1105
static void nfs4_free_closedata(void *data)
1106
{
1107 1108
	struct nfs4_closedata *calldata = data;
	struct nfs4_state_owner *sp = calldata->state->owner;
1109 1110 1111 1112 1113 1114 1115

	nfs4_put_open_state(calldata->state);
	nfs_free_seqid(calldata->arg.seqid);
	nfs4_put_state_owner(sp);
	kfree(calldata);
}

1116
static void nfs4_close_done(struct rpc_task *task, void *data)
L
Linus Torvalds 已提交
1117
{
1118
	struct nfs4_closedata *calldata = data;
L
Linus Torvalds 已提交
1119 1120 1121
	struct nfs4_state *state = calldata->state;
	struct nfs_server *server = NFS_SERVER(calldata->inode);

T
Trond Myklebust 已提交
1122 1123
	if (RPC_ASSASSINATED(task))
		return;
L
Linus Torvalds 已提交
1124 1125 1126
        /* hmm. we are done with the inode, and in the process of freeing
	 * the state_owner. we keep this around to process errors
	 */
1127
	nfs_increment_open_seqid(task->tk_status, calldata->arg.seqid);
L
Linus Torvalds 已提交
1128 1129 1130 1131
	switch (task->tk_status) {
		case 0:
			memcpy(&state->stateid, &calldata->res.stateid,
					sizeof(state->stateid));
1132
			renew_lease(server, calldata->timestamp);
L
Linus Torvalds 已提交
1133 1134 1135
			break;
		case -NFS4ERR_STALE_STATEID:
		case -NFS4ERR_EXPIRED:
1136
			nfs4_schedule_state_recovery(server->nfs_client);
L
Linus Torvalds 已提交
1137 1138 1139 1140 1141 1142 1143
			break;
		default:
			if (nfs4_async_handle_error(task, server) == -EAGAIN) {
				rpc_restart_call(task);
				return;
			}
	}
1144
	nfs_refresh_inode(calldata->inode, calldata->res.fattr);
L
Linus Torvalds 已提交
1145 1146
}

T
Trond Myklebust 已提交
1147
static void nfs4_close_prepare(struct rpc_task *task, void *data)
L
Linus Torvalds 已提交
1148
{
T
Trond Myklebust 已提交
1149
	struct nfs4_closedata *calldata = data;
1150
	struct nfs4_state *state = calldata->state;
L
Linus Torvalds 已提交
1151 1152 1153 1154
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CLOSE],
		.rpc_argp = &calldata->arg,
		.rpc_resp = &calldata->res,
1155
		.rpc_cred = state->owner->so_cred,
L
Linus Torvalds 已提交
1156
	};
1157
	int mode = 0, old_mode;
1158

1159
	if (nfs_wait_on_sequence(calldata->arg.seqid, task) != 0)
1160 1161 1162 1163
		return;
	/* Recalculate the new open mode in case someone reopened the file
	 * while we were waiting in line to be scheduled.
	 */
1164 1165 1166
	spin_lock(&state->owner->so_lock);
	spin_lock(&calldata->inode->i_lock);
	mode = old_mode = state->state;
1167 1168 1169 1170 1171 1172
	if (state->n_rdwr == 0) {
		if (state->n_rdonly == 0)
			mode &= ~FMODE_READ;
		if (state->n_wronly == 0)
			mode &= ~FMODE_WRITE;
	}
1173 1174 1175 1176
	nfs4_state_set_mode_locked(state, mode);
	spin_unlock(&calldata->inode->i_lock);
	spin_unlock(&state->owner->so_lock);
	if (mode == old_mode || test_bit(NFS_DELEGATED_STATE, &state->flags)) {
1177 1178
		/* Note: exit _without_ calling nfs4_close_done */
		task->tk_action = NULL;
1179 1180
		return;
	}
1181
	nfs_fattr_init(calldata->res.fattr);
1182
	if (mode != 0)
L
Linus Torvalds 已提交
1183
		msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_DOWNGRADE];
1184
	calldata->arg.open_flags = mode;
1185
	calldata->timestamp = jiffies;
1186
	rpc_call_setup(task, &msg, 0);
L
Linus Torvalds 已提交
1187 1188
}

1189
static const struct rpc_call_ops nfs4_close_ops = {
T
Trond Myklebust 已提交
1190
	.rpc_call_prepare = nfs4_close_prepare,
1191 1192 1193 1194
	.rpc_call_done = nfs4_close_done,
	.rpc_release = nfs4_free_closedata,
};

L
Linus Torvalds 已提交
1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205
/* 
 * It is possible for data to be read/written from a mem-mapped file 
 * after the sys_close call (which hits the vfs layer as a flush).
 * This means that we can't safely call nfsv4 close on a file until 
 * the inode is cleared. This in turn means that we are not good
 * NFSv4 citizens - we do not indicate to the server to update the file's 
 * share state even when we are done with one of the three share 
 * stateid's in the inode.
 *
 * NOTE: Caller must be holding the sp->so_owner semaphore!
 */
1206
int nfs4_do_close(struct inode *inode, struct nfs4_state *state) 
L
Linus Torvalds 已提交
1207
{
1208
	struct nfs_server *server = NFS_SERVER(inode);
L
Linus Torvalds 已提交
1209
	struct nfs4_closedata *calldata;
1210
	int status = -ENOMEM;
L
Linus Torvalds 已提交
1211

1212
	calldata = kmalloc(sizeof(*calldata), GFP_KERNEL);
L
Linus Torvalds 已提交
1213
	if (calldata == NULL)
1214
		goto out;
L
Linus Torvalds 已提交
1215 1216 1217
	calldata->inode = inode;
	calldata->state = state;
	calldata->arg.fh = NFS_FH(inode);
1218
	calldata->arg.stateid = &state->stateid;
L
Linus Torvalds 已提交
1219
	/* Serialization for the sequence id */
1220
	calldata->arg.seqid = nfs_alloc_seqid(&state->owner->so_seqid);
1221 1222
	if (calldata->arg.seqid == NULL)
		goto out_free_calldata;
1223 1224 1225
	calldata->arg.bitmask = server->attr_bitmask;
	calldata->res.fattr = &calldata->fattr;
	calldata->res.server = server;
1226

T
Trond Myklebust 已提交
1227
	status = nfs4_call_async(server->client, &nfs4_close_ops, calldata);
1228 1229 1230 1231 1232 1233 1234 1235
	if (status == 0)
		goto out;

	nfs_free_seqid(calldata->arg.seqid);
out_free_calldata:
	kfree(calldata);
out:
	return status;
L
Linus Torvalds 已提交
1236 1237
}

1238
static int nfs4_intent_set_file(struct nameidata *nd, struct dentry *dentry, struct nfs4_state *state)
1239 1240 1241 1242 1243 1244 1245 1246
{
	struct file *filp;

	filp = lookup_instantiate_filp(nd, dentry, NULL);
	if (!IS_ERR(filp)) {
		struct nfs_open_context *ctx;
		ctx = (struct nfs_open_context *)filp->private_data;
		ctx->state = state;
1247 1248 1249 1250
		return 0;
	}
	nfs4_close_state(state, nd->intent.open.flags);
	return PTR_ERR(filp);
1251 1252 1253
}

struct dentry *
L
Linus Torvalds 已提交
1254 1255 1256 1257 1258
nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
{
	struct iattr attr;
	struct rpc_cred *cred;
	struct nfs4_state *state;
1259
	struct dentry *res;
L
Linus Torvalds 已提交
1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272

	if (nd->flags & LOOKUP_CREATE) {
		attr.ia_mode = nd->intent.open.create_mode;
		attr.ia_valid = ATTR_MODE;
		if (!IS_POSIXACL(dir))
			attr.ia_mode &= ~current->fs->umask;
	} else {
		attr.ia_valid = 0;
		BUG_ON(nd->intent.open.flags & O_CREAT);
	}

	cred = rpcauth_lookupcred(NFS_SERVER(dir)->client->cl_auth, 0);
	if (IS_ERR(cred))
1273
		return (struct dentry *)cred;
L
Linus Torvalds 已提交
1274 1275
	state = nfs4_do_open(dir, dentry, nd->intent.open.flags, &attr, cred);
	put_rpccred(cred);
1276 1277 1278 1279 1280
	if (IS_ERR(state)) {
		if (PTR_ERR(state) == -ENOENT)
			d_add(dentry, NULL);
		return (struct dentry *)state;
	}
1281
	res = d_add_unique(dentry, igrab(state->inode));
1282 1283 1284 1285
	if (res != NULL)
		dentry = res;
	nfs4_intent_set_file(nd, dentry, state);
	return res;
L
Linus Torvalds 已提交
1286 1287 1288
}

int
1289
nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags, struct nameidata *nd)
L
Linus Torvalds 已提交
1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300
{
	struct rpc_cred *cred;
	struct nfs4_state *state;

	cred = rpcauth_lookupcred(NFS_SERVER(dir)->client->cl_auth, 0);
	if (IS_ERR(cred))
		return PTR_ERR(cred);
	state = nfs4_open_delegated(dentry->d_inode, openflags, cred);
	if (IS_ERR(state))
		state = nfs4_do_open(dir, dentry, openflags, NULL, cred);
	put_rpccred(cred);
1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315
	if (IS_ERR(state)) {
		switch (PTR_ERR(state)) {
			case -EPERM:
			case -EACCES:
			case -EDQUOT:
			case -ENOSPC:
			case -EROFS:
				lookup_instantiate_filp(nd, (struct dentry *)state, NULL);
				return 1;
			case -ENOENT:
				if (dentry->d_inode == NULL)
					return 1;
		}
		goto out_drop;
	}
1316
	if (state->inode == dentry->d_inode) {
1317
		nfs4_intent_set_file(nd, dentry, state);
L
Linus Torvalds 已提交
1318 1319 1320
		return 1;
	}
	nfs4_close_state(state, openflags);
1321 1322
out_drop:
	d_drop(dentry);
L
Linus Torvalds 已提交
1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350
	return 0;
}


static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle)
{
	struct nfs4_server_caps_res res = {};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SERVER_CAPS],
		.rpc_argp = fhandle,
		.rpc_resp = &res,
	};
	int status;

	status = rpc_call_sync(server->client, &msg, 0);
	if (status == 0) {
		memcpy(server->attr_bitmask, res.attr_bitmask, sizeof(server->attr_bitmask));
		if (res.attr_bitmask[0] & FATTR4_WORD0_ACL)
			server->caps |= NFS_CAP_ACLS;
		if (res.has_links != 0)
			server->caps |= NFS_CAP_HARDLINKS;
		if (res.has_symlinks != 0)
			server->caps |= NFS_CAP_SYMLINKS;
		server->acl_bitmask = res.acl_bitmask;
	}
	return status;
}

1351
int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle)
L
Linus Torvalds 已提交
1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(server,
				_nfs4_server_capabilities(server, fhandle),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle,
		struct nfs_fsinfo *info)
{
	struct nfs4_lookup_root_arg args = {
		.bitmask = nfs4_fattr_bitmap,
	};
	struct nfs4_lookup_res res = {
		.server = server,
1371
		.fattr = info->fattr,
L
Linus Torvalds 已提交
1372 1373 1374 1375 1376 1377 1378
		.fh = fhandle,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP_ROOT],
		.rpc_argp = &args,
		.rpc_resp = &res,
	};
1379
	nfs_fattr_init(info->fattr);
L
Linus Torvalds 已提交
1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441
	return rpc_call_sync(server->client, &msg, 0);
}

static int nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle,
		struct nfs_fsinfo *info)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(server,
				_nfs4_lookup_root(server, fhandle, info),
				&exception);
	} while (exception.retry);
	return err;
}

static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
		struct nfs_fsinfo *info)
{
	struct nfs_fattr *	fattr = info->fattr;
	unsigned char *		p;
	struct qstr		q;
	struct nfs4_lookup_arg args = {
		.dir_fh = fhandle,
		.name = &q,
		.bitmask = nfs4_fattr_bitmap,
	};
	struct nfs4_lookup_res res = {
		.server = server,
		.fattr = fattr,
		.fh = fhandle,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP],
		.rpc_argp = &args,
		.rpc_resp = &res,
	};
	int status;

	/*
	 * Now we do a separate LOOKUP for each component of the mount path.
	 * The LOOKUPs are done separately so that we can conveniently
	 * catch an ERR_WRONGSEC if it occurs along the way...
	 */
	status = nfs4_lookup_root(server, fhandle, info);
	if (status)
		goto out;

	p = server->mnt_path;
	for (;;) {
		struct nfs4_exception exception = { };

		while (*p == '/')
			p++;
		if (!*p)
			break;
		q.name = p;
		while (*p && (*p != '/'))
			p++;
		q.len = p - q.name;

		do {
1442
			nfs_fattr_init(fattr);
L
Linus Torvalds 已提交
1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459
			status = nfs4_handle_exception(server,
					rpc_call_sync(server->client, &msg, 0),
					&exception);
		} while (exception.retry);
		if (status == 0)
			continue;
		if (status == -ENOENT) {
			printk(KERN_NOTICE "NFS: mount path %s does not exist!\n", server->mnt_path);
			printk(KERN_NOTICE "NFS: suggestion: try mounting '/' instead.\n");
		}
		break;
	}
	if (status == 0)
		status = nfs4_server_capabilities(server, fhandle);
	if (status == 0)
		status = nfs4_do_fsinfo(server, fhandle, info);
out:
1460
	return nfs4_map_errors(status);
L
Linus Torvalds 已提交
1461 1462
}

M
Manoj Naik 已提交
1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506
/*
 * Get locations and (maybe) other attributes of a referral.
 * Note that we'll actually follow the referral later when
 * we detect fsid mismatch in inode revalidation
 */
static int nfs4_get_referral(struct inode *dir, struct qstr *name, struct nfs_fattr *fattr, struct nfs_fh *fhandle)
{
	int status = -ENOMEM;
	struct page *page = NULL;
	struct nfs4_fs_locations *locations = NULL;
	struct dentry dentry = {};

	page = alloc_page(GFP_KERNEL);
	if (page == NULL)
		goto out;
	locations = kmalloc(sizeof(struct nfs4_fs_locations), GFP_KERNEL);
	if (locations == NULL)
		goto out;

	dentry.d_name.name = name->name;
	dentry.d_name.len = name->len;
	status = nfs4_proc_fs_locations(dir, &dentry, locations, page);
	if (status != 0)
		goto out;
	/* Make sure server returned a different fsid for the referral */
	if (nfs_fsid_equal(&NFS_SERVER(dir)->fsid, &locations->fattr.fsid)) {
		dprintk("%s: server did not return a different fsid for a referral at %s\n", __FUNCTION__, name->name);
		status = -EIO;
		goto out;
	}

	memcpy(fattr, &locations->fattr, sizeof(struct nfs_fattr));
	fattr->valid |= NFS_ATTR_FATTR_V4_REFERRAL;
	if (!fattr->mode)
		fattr->mode = S_IFDIR;
	memset(fhandle, 0, sizeof(struct nfs_fh));
out:
	if (page)
		__free_page(page);
	if (locations)
		kfree(locations);
	return status;
}

L
Linus Torvalds 已提交
1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522
static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr)
{
	struct nfs4_getattr_arg args = {
		.fh = fhandle,
		.bitmask = server->attr_bitmask,
	};
	struct nfs4_getattr_res res = {
		.fattr = fattr,
		.server = server,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_GETATTR],
		.rpc_argp = &args,
		.rpc_resp = &res,
	};
	
1523
	nfs_fattr_init(fattr);
L
Linus Torvalds 已提交
1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559
	return rpc_call_sync(server->client, &msg, 0);
}

static int nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(server,
				_nfs4_proc_getattr(server, fhandle, fattr),
				&exception);
	} while (exception.retry);
	return err;
}

/* 
 * The file is not closed if it is opened due to the a request to change
 * the size of the file. The open call will not be needed once the
 * VFS layer lookup-intents are implemented.
 *
 * Close is called when the inode is destroyed.
 * If we haven't opened the file for O_WRONLY, we
 * need to in the size_change case to obtain a stateid.
 *
 * Got race?
 * Because OPEN is always done by name in nfsv4, it is
 * possible that we opened a different file by the same
 * name.  We can recognize this race condition, but we
 * can't do anything about it besides returning an error.
 *
 * This will be fixed with VFS changes (lookup-intent).
 */
static int
nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr,
		  struct iattr *sattr)
{
1560 1561
	struct rpc_cred *cred;
	struct inode *inode = dentry->d_inode;
1562 1563
	struct nfs_open_context *ctx;
	struct nfs4_state *state = NULL;
L
Linus Torvalds 已提交
1564 1565
	int status;

1566
	nfs_fattr_init(fattr);
L
Linus Torvalds 已提交
1567
	
1568 1569 1570
	cred = rpcauth_lookupcred(NFS_SERVER(inode)->client->cl_auth, 0);
	if (IS_ERR(cred))
		return PTR_ERR(cred);
1571 1572 1573 1574 1575

	/* Search for an existing open(O_WRITE) file */
	ctx = nfs_find_open_context(inode, cred, FMODE_WRITE);
	if (ctx != NULL)
		state = ctx->state;
1576

1577
	status = nfs4_do_setattr(inode, fattr, sattr, state);
1578 1579
	if (status == 0)
		nfs_setattr_update_inode(inode, sattr);
1580 1581
	if (ctx != NULL)
		put_nfs_open_context(ctx);
1582
	put_rpccred(cred);
L
Linus Torvalds 已提交
1583 1584 1585
	return status;
}

D
David Howells 已提交
1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631
static int _nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh,
		struct qstr *name, struct nfs_fh *fhandle,
		struct nfs_fattr *fattr)
{
	int		       status;
	struct nfs4_lookup_arg args = {
		.bitmask = server->attr_bitmask,
		.dir_fh = dirfh,
		.name = name,
	};
	struct nfs4_lookup_res res = {
		.server = server,
		.fattr = fattr,
		.fh = fhandle,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP],
		.rpc_argp = &args,
		.rpc_resp = &res,
	};

	nfs_fattr_init(fattr);

	dprintk("NFS call  lookupfh %s\n", name->name);
	status = rpc_call_sync(server->client, &msg, 0);
	dprintk("NFS reply lookupfh: %d\n", status);
	if (status == -NFS4ERR_MOVED)
		status = -EREMOTE;
	return status;
}

static int nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh,
			      struct qstr *name, struct nfs_fh *fhandle,
			      struct nfs_fattr *fattr)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(server,
				_nfs4_proc_lookupfh(server, dirfh, name,
						    fhandle, fattr),
				&exception);
	} while (exception.retry);
	return err;
}

L
Linus Torvalds 已提交
1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652
static int _nfs4_proc_lookup(struct inode *dir, struct qstr *name,
		struct nfs_fh *fhandle, struct nfs_fattr *fattr)
{
	int		       status;
	struct nfs_server *server = NFS_SERVER(dir);
	struct nfs4_lookup_arg args = {
		.bitmask = server->attr_bitmask,
		.dir_fh = NFS_FH(dir),
		.name = name,
	};
	struct nfs4_lookup_res res = {
		.server = server,
		.fattr = fattr,
		.fh = fhandle,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP],
		.rpc_argp = &args,
		.rpc_resp = &res,
	};
	
1653
	nfs_fattr_init(fattr);
L
Linus Torvalds 已提交
1654 1655 1656
	
	dprintk("NFS call  lookup %s\n", name->name);
	status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
M
Manoj Naik 已提交
1657 1658
	if (status == -NFS4ERR_MOVED)
		status = nfs4_get_referral(dir, name, fattr, fhandle);
L
Linus Torvalds 已提交
1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803
	dprintk("NFS reply lookup: %d\n", status);
	return status;
}

static int nfs4_proc_lookup(struct inode *dir, struct qstr *name, struct nfs_fh *fhandle, struct nfs_fattr *fattr)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(dir),
				_nfs4_proc_lookup(dir, name, fhandle, fattr),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry)
{
	struct nfs4_accessargs args = {
		.fh = NFS_FH(inode),
	};
	struct nfs4_accessres res = { 0 };
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_ACCESS],
		.rpc_argp = &args,
		.rpc_resp = &res,
		.rpc_cred = entry->cred,
	};
	int mode = entry->mask;
	int status;

	/*
	 * Determine which access bits we want to ask for...
	 */
	if (mode & MAY_READ)
		args.access |= NFS4_ACCESS_READ;
	if (S_ISDIR(inode->i_mode)) {
		if (mode & MAY_WRITE)
			args.access |= NFS4_ACCESS_MODIFY | NFS4_ACCESS_EXTEND | NFS4_ACCESS_DELETE;
		if (mode & MAY_EXEC)
			args.access |= NFS4_ACCESS_LOOKUP;
	} else {
		if (mode & MAY_WRITE)
			args.access |= NFS4_ACCESS_MODIFY | NFS4_ACCESS_EXTEND;
		if (mode & MAY_EXEC)
			args.access |= NFS4_ACCESS_EXECUTE;
	}
	status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0);
	if (!status) {
		entry->mask = 0;
		if (res.access & NFS4_ACCESS_READ)
			entry->mask |= MAY_READ;
		if (res.access & (NFS4_ACCESS_MODIFY | NFS4_ACCESS_EXTEND | NFS4_ACCESS_DELETE))
			entry->mask |= MAY_WRITE;
		if (res.access & (NFS4_ACCESS_LOOKUP|NFS4_ACCESS_EXECUTE))
			entry->mask |= MAY_EXEC;
	}
	return status;
}

static int nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(inode),
				_nfs4_proc_access(inode, entry),
				&exception);
	} while (exception.retry);
	return err;
}

/*
 * TODO: For the time being, we don't try to get any attributes
 * along with any of the zero-copy operations READ, READDIR,
 * READLINK, WRITE.
 *
 * In the case of the first three, we want to put the GETATTR
 * after the read-type operation -- this is because it is hard
 * to predict the length of a GETATTR response in v4, and thus
 * align the READ data correctly.  This means that the GETATTR
 * may end up partially falling into the page cache, and we should
 * shift it into the 'tail' of the xdr_buf before processing.
 * To do this efficiently, we need to know the total length
 * of data received, which doesn't seem to be available outside
 * of the RPC layer.
 *
 * In the case of WRITE, we also want to put the GETATTR after
 * the operation -- in this case because we want to make sure
 * we get the post-operation mtime and size.  This means that
 * we can't use xdr_encode_pages() as written: we need a variant
 * of it which would leave room in the 'tail' iovec.
 *
 * Both of these changes to the XDR layer would in fact be quite
 * minor, but I decided to leave them for a subsequent patch.
 */
static int _nfs4_proc_readlink(struct inode *inode, struct page *page,
		unsigned int pgbase, unsigned int pglen)
{
	struct nfs4_readlink args = {
		.fh       = NFS_FH(inode),
		.pgbase	  = pgbase,
		.pglen    = pglen,
		.pages    = &page,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READLINK],
		.rpc_argp = &args,
		.rpc_resp = NULL,
	};

	return rpc_call_sync(NFS_CLIENT(inode), &msg, 0);
}

static int nfs4_proc_readlink(struct inode *inode, struct page *page,
		unsigned int pgbase, unsigned int pglen)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(inode),
				_nfs4_proc_readlink(inode, page, pgbase, pglen),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_proc_read(struct nfs_read_data *rdata)
{
	int flags = rdata->flags;
	struct inode *inode = rdata->inode;
	struct nfs_fattr *fattr = rdata->res.fattr;
	struct nfs_server *server = NFS_SERVER(inode);
	struct rpc_message msg = {
		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_READ],
		.rpc_argp	= &rdata->args,
		.rpc_resp	= &rdata->res,
		.rpc_cred	= rdata->cred,
	};
	unsigned long timestamp = jiffies;
	int status;

	dprintk("NFS call  read %d @ %Ld\n", rdata->args.count,
			(long long) rdata->args.offset);

1804
	nfs_fattr_init(fattr);
L
Linus Torvalds 已提交
1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840
	status = rpc_call_sync(server->client, &msg, flags);
	if (!status)
		renew_lease(server, timestamp);
	dprintk("NFS reply read: %d\n", status);
	return status;
}

static int nfs4_proc_read(struct nfs_read_data *rdata)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(rdata->inode),
				_nfs4_proc_read(rdata),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_proc_write(struct nfs_write_data *wdata)
{
	int rpcflags = wdata->flags;
	struct inode *inode = wdata->inode;
	struct nfs_fattr *fattr = wdata->res.fattr;
	struct nfs_server *server = NFS_SERVER(inode);
	struct rpc_message msg = {
		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_WRITE],
		.rpc_argp	= &wdata->args,
		.rpc_resp	= &wdata->res,
		.rpc_cred	= wdata->cred,
	};
	int status;

	dprintk("NFS call  write %d @ %Ld\n", wdata->args.count,
			(long long) wdata->args.offset);

1841 1842
	wdata->args.bitmask = server->attr_bitmask;
	wdata->res.server = server;
1843
	wdata->timestamp = jiffies;
1844
	nfs_fattr_init(fattr);
L
Linus Torvalds 已提交
1845 1846
	status = rpc_call_sync(server->client, &msg, rpcflags);
	dprintk("NFS reply write: %d\n", status);
1847 1848
	if (status < 0)
		return status;
1849
	renew_lease(server, wdata->timestamp);
1850 1851
	nfs_post_op_update_inode(inode, fattr);
	return wdata->res.count;
L
Linus Torvalds 已提交
1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881
}

static int nfs4_proc_write(struct nfs_write_data *wdata)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(wdata->inode),
				_nfs4_proc_write(wdata),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_proc_commit(struct nfs_write_data *cdata)
{
	struct inode *inode = cdata->inode;
	struct nfs_fattr *fattr = cdata->res.fattr;
	struct nfs_server *server = NFS_SERVER(inode);
	struct rpc_message msg = {
		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_COMMIT],
		.rpc_argp	= &cdata->args,
		.rpc_resp	= &cdata->res,
		.rpc_cred	= cdata->cred,
	};
	int status;

	dprintk("NFS call  commit %d @ %Ld\n", cdata->args.count,
			(long long) cdata->args.offset);

1882 1883
	cdata->args.bitmask = server->attr_bitmask;
	cdata->res.server = server;
1884
	cdata->timestamp = jiffies;
1885
	nfs_fattr_init(fattr);
L
Linus Torvalds 已提交
1886
	status = rpc_call_sync(server->client, &msg, 0);
1887 1888
	if (status >= 0)
		renew_lease(server, cdata->timestamp);
L
Linus Torvalds 已提交
1889
	dprintk("NFS reply commit: %d\n", status);
1890 1891
	if (status >= 0)
		nfs_post_op_update_inode(inode, fattr);
L
Linus Torvalds 已提交
1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923
	return status;
}

static int nfs4_proc_commit(struct nfs_write_data *cdata)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(cdata->inode),
				_nfs4_proc_commit(cdata),
				&exception);
	} while (exception.retry);
	return err;
}

/*
 * Got race?
 * We will need to arrange for the VFS layer to provide an atomic open.
 * Until then, this create/open method is prone to inefficiency and race
 * conditions due to the lookup, create, and open VFS calls from sys_open()
 * placed on the wire.
 *
 * Given the above sorry state of affairs, I'm simply sending an OPEN.
 * The file will be opened again in the subsequent VFS open call
 * (nfs4_proc_file_open).
 *
 * The open for read will just hang around to be used by any process that
 * opens the file O_RDONLY. This will all be resolved with the VFS changes.
 */

static int
nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
1924
                 int flags, struct nameidata *nd)
L
Linus Torvalds 已提交
1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940
{
	struct nfs4_state *state;
	struct rpc_cred *cred;
	int status = 0;

	cred = rpcauth_lookupcred(NFS_SERVER(dir)->client->cl_auth, 0);
	if (IS_ERR(cred)) {
		status = PTR_ERR(cred);
		goto out;
	}
	state = nfs4_do_open(dir, dentry, flags, sattr, cred);
	put_rpccred(cred);
	if (IS_ERR(state)) {
		status = PTR_ERR(state);
		goto out;
	}
1941
	d_instantiate(dentry, igrab(state->inode));
L
Linus Torvalds 已提交
1942 1943
	if (flags & O_EXCL) {
		struct nfs_fattr fattr;
1944
		status = nfs4_do_setattr(state->inode, &fattr, sattr, state);
1945
		if (status == 0)
1946
			nfs_setattr_update_inode(state->inode, sattr);
1947 1948
	}
	if (status == 0 && nd != NULL && (nd->flags & LOOKUP_OPEN))
1949
		status = nfs4_intent_set_file(nd, dentry, state);
1950 1951
	else
		nfs4_close_state(state, flags);
L
Linus Torvalds 已提交
1952 1953 1954 1955 1956 1957
out:
	return status;
}

static int _nfs4_proc_remove(struct inode *dir, struct qstr *name)
{
1958
	struct nfs_server *server = NFS_SERVER(dir);
L
Linus Torvalds 已提交
1959 1960 1961
	struct nfs4_remove_arg args = {
		.fh = NFS_FH(dir),
		.name = name,
1962 1963 1964 1965 1966 1967
		.bitmask = server->attr_bitmask,
	};
	struct nfs_fattr dir_attr;
	struct nfs4_remove_res	res = {
		.server = server,
		.dir_attr = &dir_attr,
L
Linus Torvalds 已提交
1968 1969 1970 1971 1972 1973 1974 1975
	};
	struct rpc_message msg = {
		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_REMOVE],
		.rpc_argp	= &args,
		.rpc_resp	= &res,
	};
	int			status;

1976 1977 1978 1979 1980 1981
	nfs_fattr_init(res.dir_attr);
	status = rpc_call_sync(server->client, &msg, 0);
	if (status == 0) {
		update_changeattr(dir, &res.cinfo);
		nfs_post_op_update_inode(dir, res.dir_attr);
	}
L
Linus Torvalds 已提交
1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
	return status;
}

static int nfs4_proc_remove(struct inode *dir, struct qstr *name)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(dir),
				_nfs4_proc_remove(dir, name),
				&exception);
	} while (exception.retry);
	return err;
}

struct unlink_desc {
	struct nfs4_remove_arg	args;
1999 2000
	struct nfs4_remove_res	res;
	struct nfs_fattr dir_attr;
L
Linus Torvalds 已提交
2001 2002 2003 2004 2005
};

static int nfs4_proc_unlink_setup(struct rpc_message *msg, struct dentry *dir,
		struct qstr *name)
{
2006
	struct nfs_server *server = NFS_SERVER(dir->d_inode);
L
Linus Torvalds 已提交
2007 2008 2009 2010 2011 2012 2013 2014
	struct unlink_desc *up;

	up = (struct unlink_desc *) kmalloc(sizeof(*up), GFP_KERNEL);
	if (!up)
		return -ENOMEM;
	
	up->args.fh = NFS_FH(dir->d_inode);
	up->args.name = name;
2015 2016 2017
	up->args.bitmask = server->attr_bitmask;
	up->res.server = server;
	up->res.dir_attr = &up->dir_attr;
L
Linus Torvalds 已提交
2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031
	
	msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_REMOVE];
	msg->rpc_argp = &up->args;
	msg->rpc_resp = &up->res;
	return 0;
}

static int nfs4_proc_unlink_done(struct dentry *dir, struct rpc_task *task)
{
	struct rpc_message *msg = &task->tk_msg;
	struct unlink_desc *up;
	
	if (msg->rpc_resp != NULL) {
		up = container_of(msg->rpc_resp, struct unlink_desc, res);
2032 2033
		update_changeattr(dir->d_inode, &up->res.cinfo);
		nfs_post_op_update_inode(dir->d_inode, up->res.dir_attr);
L
Linus Torvalds 已提交
2034 2035 2036 2037 2038 2039 2040 2041 2042 2043
		kfree(up);
		msg->rpc_resp = NULL;
		msg->rpc_argp = NULL;
	}
	return 0;
}

static int _nfs4_proc_rename(struct inode *old_dir, struct qstr *old_name,
		struct inode *new_dir, struct qstr *new_name)
{
2044
	struct nfs_server *server = NFS_SERVER(old_dir);
L
Linus Torvalds 已提交
2045 2046 2047 2048 2049
	struct nfs4_rename_arg arg = {
		.old_dir = NFS_FH(old_dir),
		.new_dir = NFS_FH(new_dir),
		.old_name = old_name,
		.new_name = new_name,
2050 2051 2052 2053 2054 2055 2056
		.bitmask = server->attr_bitmask,
	};
	struct nfs_fattr old_fattr, new_fattr;
	struct nfs4_rename_res res = {
		.server = server,
		.old_fattr = &old_fattr,
		.new_fattr = &new_fattr,
L
Linus Torvalds 已提交
2057 2058 2059 2060 2061 2062 2063 2064
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_RENAME],
		.rpc_argp = &arg,
		.rpc_resp = &res,
	};
	int			status;
	
2065 2066 2067
	nfs_fattr_init(res.old_fattr);
	nfs_fattr_init(res.new_fattr);
	status = rpc_call_sync(server->client, &msg, 0);
L
Linus Torvalds 已提交
2068 2069 2070

	if (!status) {
		update_changeattr(old_dir, &res.old_cinfo);
2071
		nfs_post_op_update_inode(old_dir, res.old_fattr);
L
Linus Torvalds 已提交
2072
		update_changeattr(new_dir, &res.new_cinfo);
2073
		nfs_post_op_update_inode(new_dir, res.new_fattr);
L
Linus Torvalds 已提交
2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093
	}
	return status;
}

static int nfs4_proc_rename(struct inode *old_dir, struct qstr *old_name,
		struct inode *new_dir, struct qstr *new_name)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(old_dir),
				_nfs4_proc_rename(old_dir, old_name,
					new_dir, new_name),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_proc_link(struct inode *inode, struct inode *dir, struct qstr *name)
{
2094
	struct nfs_server *server = NFS_SERVER(inode);
L
Linus Torvalds 已提交
2095 2096 2097 2098
	struct nfs4_link_arg arg = {
		.fh     = NFS_FH(inode),
		.dir_fh = NFS_FH(dir),
		.name   = name,
2099 2100 2101 2102 2103 2104 2105
		.bitmask = server->attr_bitmask,
	};
	struct nfs_fattr fattr, dir_attr;
	struct nfs4_link_res res = {
		.server = server,
		.fattr = &fattr,
		.dir_attr = &dir_attr,
L
Linus Torvalds 已提交
2106 2107 2108 2109
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LINK],
		.rpc_argp = &arg,
2110
		.rpc_resp = &res,
L
Linus Torvalds 已提交
2111 2112 2113
	};
	int			status;

2114 2115 2116 2117 2118 2119
	nfs_fattr_init(res.fattr);
	nfs_fattr_init(res.dir_attr);
	status = rpc_call_sync(server->client, &msg, 0);
	if (!status) {
		update_changeattr(dir, &res.cinfo);
		nfs_post_op_update_inode(dir, res.dir_attr);
2120
		nfs_post_op_update_inode(inode, res.fattr);
2121
	}
L
Linus Torvalds 已提交
2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142

	return status;
}

static int nfs4_proc_link(struct inode *inode, struct inode *dir, struct qstr *name)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(inode),
				_nfs4_proc_link(inode, dir, name),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name,
		struct qstr *path, struct iattr *sattr, struct nfs_fh *fhandle,
		struct nfs_fattr *fattr)
{
	struct nfs_server *server = NFS_SERVER(dir);
2143
	struct nfs_fattr dir_fattr;
L
Linus Torvalds 已提交
2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155
	struct nfs4_create_arg arg = {
		.dir_fh = NFS_FH(dir),
		.server = server,
		.name = name,
		.attrs = sattr,
		.ftype = NF4LNK,
		.bitmask = server->attr_bitmask,
	};
	struct nfs4_create_res res = {
		.server = server,
		.fh = fhandle,
		.fattr = fattr,
2156
		.dir_fattr = &dir_fattr,
L
Linus Torvalds 已提交
2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SYMLINK],
		.rpc_argp = &arg,
		.rpc_resp = &res,
	};
	int			status;

	if (path->len > NFS4_MAXPATHLEN)
		return -ENAMETOOLONG;
	arg.u.symlink = path;
2168
	nfs_fattr_init(fattr);
2169
	nfs_fattr_init(&dir_fattr);
L
Linus Torvalds 已提交
2170 2171 2172 2173
	
	status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
	if (!status)
		update_changeattr(dir, &res.dir_cinfo);
2174
	nfs_post_op_update_inode(dir, res.dir_fattr);
L
Linus Torvalds 已提交
2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197
	return status;
}

static int nfs4_proc_symlink(struct inode *dir, struct qstr *name,
		struct qstr *path, struct iattr *sattr, struct nfs_fh *fhandle,
		struct nfs_fattr *fattr)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(dir),
				_nfs4_proc_symlink(dir, name, path, sattr,
					fhandle, fattr),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_proc_mkdir(struct inode *dir, struct dentry *dentry,
		struct iattr *sattr)
{
	struct nfs_server *server = NFS_SERVER(dir);
	struct nfs_fh fhandle;
2198
	struct nfs_fattr fattr, dir_fattr;
L
Linus Torvalds 已提交
2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210
	struct nfs4_create_arg arg = {
		.dir_fh = NFS_FH(dir),
		.server = server,
		.name = &dentry->d_name,
		.attrs = sattr,
		.ftype = NF4DIR,
		.bitmask = server->attr_bitmask,
	};
	struct nfs4_create_res res = {
		.server = server,
		.fh = &fhandle,
		.fattr = &fattr,
2211
		.dir_fattr = &dir_fattr,
L
Linus Torvalds 已提交
2212 2213 2214 2215 2216 2217 2218 2219
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CREATE],
		.rpc_argp = &arg,
		.rpc_resp = &res,
	};
	int			status;

2220
	nfs_fattr_init(&fattr);
2221
	nfs_fattr_init(&dir_fattr);
L
Linus Torvalds 已提交
2222 2223 2224 2225
	
	status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
	if (!status) {
		update_changeattr(dir, &res.dir_cinfo);
2226
		nfs_post_op_update_inode(dir, res.dir_fattr);
L
Linus Torvalds 已提交
2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264
		status = nfs_instantiate(dentry, &fhandle, &fattr);
	}
	return status;
}

static int nfs4_proc_mkdir(struct inode *dir, struct dentry *dentry,
		struct iattr *sattr)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(dir),
				_nfs4_proc_mkdir(dir, dentry, sattr),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_proc_readdir(struct dentry *dentry, struct rpc_cred *cred,
                  u64 cookie, struct page *page, unsigned int count, int plus)
{
	struct inode		*dir = dentry->d_inode;
	struct nfs4_readdir_arg args = {
		.fh = NFS_FH(dir),
		.pages = &page,
		.pgbase = 0,
		.count = count,
		.bitmask = NFS_SERVER(dentry->d_inode)->attr_bitmask,
	};
	struct nfs4_readdir_res res;
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READDIR],
		.rpc_argp = &args,
		.rpc_resp = &res,
		.rpc_cred = cred,
	};
	int			status;

2265 2266 2267 2268
	dprintk("%s: dentry = %s/%s, cookie = %Lu\n", __FUNCTION__,
			dentry->d_parent->d_name.name,
			dentry->d_name.name,
			(unsigned long long)cookie);
L
Linus Torvalds 已提交
2269 2270 2271 2272 2273 2274 2275
	lock_kernel();
	nfs4_setup_readdir(cookie, NFS_COOKIEVERF(dir), dentry, &args);
	res.pgbase = args.pgbase;
	status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
	if (status == 0)
		memcpy(NFS_COOKIEVERF(dir), res.verifier.data, NFS4_VERIFIER_SIZE);
	unlock_kernel();
2276
	dprintk("%s: returns %d\n", __FUNCTION__, status);
L
Linus Torvalds 已提交
2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298
	return status;
}

static int nfs4_proc_readdir(struct dentry *dentry, struct rpc_cred *cred,
                  u64 cookie, struct page *page, unsigned int count, int plus)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(dentry->d_inode),
				_nfs4_proc_readdir(dentry, cred, cookie,
					page, count, plus),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_proc_mknod(struct inode *dir, struct dentry *dentry,
		struct iattr *sattr, dev_t rdev)
{
	struct nfs_server *server = NFS_SERVER(dir);
	struct nfs_fh fh;
2299
	struct nfs_fattr fattr, dir_fattr;
L
Linus Torvalds 已提交
2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310
	struct nfs4_create_arg arg = {
		.dir_fh = NFS_FH(dir),
		.server = server,
		.name = &dentry->d_name,
		.attrs = sattr,
		.bitmask = server->attr_bitmask,
	};
	struct nfs4_create_res res = {
		.server = server,
		.fh = &fh,
		.fattr = &fattr,
2311
		.dir_fattr = &dir_fattr,
L
Linus Torvalds 已提交
2312 2313 2314 2315 2316 2317 2318 2319 2320
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CREATE],
		.rpc_argp = &arg,
		.rpc_resp = &res,
	};
	int			status;
	int                     mode = sattr->ia_mode;

2321
	nfs_fattr_init(&fattr);
2322
	nfs_fattr_init(&dir_fattr);
L
Linus Torvalds 已提交
2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343

	BUG_ON(!(sattr->ia_valid & ATTR_MODE));
	BUG_ON(!S_ISFIFO(mode) && !S_ISBLK(mode) && !S_ISCHR(mode) && !S_ISSOCK(mode));
	if (S_ISFIFO(mode))
		arg.ftype = NF4FIFO;
	else if (S_ISBLK(mode)) {
		arg.ftype = NF4BLK;
		arg.u.device.specdata1 = MAJOR(rdev);
		arg.u.device.specdata2 = MINOR(rdev);
	}
	else if (S_ISCHR(mode)) {
		arg.ftype = NF4CHR;
		arg.u.device.specdata1 = MAJOR(rdev);
		arg.u.device.specdata2 = MINOR(rdev);
	}
	else
		arg.ftype = NF4SOCK;
	
	status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
	if (status == 0) {
		update_changeattr(dir, &res.dir_cinfo);
2344
		nfs_post_op_update_inode(dir, res.dir_fattr);
L
Linus Torvalds 已提交
2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375
		status = nfs_instantiate(dentry, &fh, &fattr);
	}
	return status;
}

static int nfs4_proc_mknod(struct inode *dir, struct dentry *dentry,
		struct iattr *sattr, dev_t rdev)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(dir),
				_nfs4_proc_mknod(dir, dentry, sattr, rdev),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle,
		 struct nfs_fsstat *fsstat)
{
	struct nfs4_statfs_arg args = {
		.fh = fhandle,
		.bitmask = server->attr_bitmask,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_STATFS],
		.rpc_argp = &args,
		.rpc_resp = fsstat,
	};

2376
	nfs_fattr_init(fsstat->fattr);
L
Linus Torvalds 已提交
2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422
	return rpc_call_sync(server->client, &msg, 0);
}

static int nfs4_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsstat *fsstat)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(server,
				_nfs4_proc_statfs(server, fhandle, fsstat),
				&exception);
	} while (exception.retry);
	return err;
}

static int _nfs4_do_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle,
		struct nfs_fsinfo *fsinfo)
{
	struct nfs4_fsinfo_arg args = {
		.fh = fhandle,
		.bitmask = server->attr_bitmask,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_FSINFO],
		.rpc_argp = &args,
		.rpc_resp = fsinfo,
	};

	return rpc_call_sync(server->client, &msg, 0);
}

static int nfs4_do_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsinfo *fsinfo)
{
	struct nfs4_exception exception = { };
	int err;

	do {
		err = nfs4_handle_exception(server,
				_nfs4_do_fsinfo(server, fhandle, fsinfo),
				&exception);
	} while (exception.retry);
	return err;
}

static int nfs4_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsinfo *fsinfo)
{
2423
	nfs_fattr_init(fsinfo->fattr);
L
Linus Torvalds 已提交
2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445
	return nfs4_do_fsinfo(server, fhandle, fsinfo);
}

static int _nfs4_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle,
		struct nfs_pathconf *pathconf)
{
	struct nfs4_pathconf_arg args = {
		.fh = fhandle,
		.bitmask = server->attr_bitmask,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_PATHCONF],
		.rpc_argp = &args,
		.rpc_resp = pathconf,
	};

	/* None of the pathconf attributes are mandatory to implement */
	if ((args.bitmask[0] & nfs4_pathconf_bitmap[0]) == 0) {
		memset(pathconf, 0, sizeof(*pathconf));
		return 0;
	}

2446
	nfs_fattr_init(pathconf->fattr);
L
Linus Torvalds 已提交
2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463
	return rpc_call_sync(server->client, &msg, 0);
}

static int nfs4_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle,
		struct nfs_pathconf *pathconf)
{
	struct nfs4_exception exception = { };
	int err;

	do {
		err = nfs4_handle_exception(server,
				_nfs4_proc_pathconf(server, fhandle, pathconf),
				&exception);
	} while (exception.retry);
	return err;
}

T
Trond Myklebust 已提交
2464
static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data)
L
Linus Torvalds 已提交
2465
{
T
Trond Myklebust 已提交
2466
	struct nfs_server *server = NFS_SERVER(data->inode);
L
Linus Torvalds 已提交
2467

T
Trond Myklebust 已提交
2468
	if (nfs4_async_handle_error(task, server) == -EAGAIN) {
L
Linus Torvalds 已提交
2469
		rpc_restart_call(task);
T
Trond Myklebust 已提交
2470
		return -EAGAIN;
L
Linus Torvalds 已提交
2471 2472
	}
	if (task->tk_status > 0)
T
Trond Myklebust 已提交
2473 2474
		renew_lease(server, data->timestamp);
	return 0;
L
Linus Torvalds 已提交
2475 2476
}

T
Trond Myklebust 已提交
2477
static void nfs4_proc_read_setup(struct nfs_read_data *data)
L
Linus Torvalds 已提交
2478 2479 2480 2481 2482 2483 2484 2485 2486 2487
{
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ],
		.rpc_argp = &data->args,
		.rpc_resp = &data->res,
		.rpc_cred = data->cred,
	};

	data->timestamp   = jiffies;

T
Trond Myklebust 已提交
2488
	rpc_call_setup(&data->task, &msg, 0);
L
Linus Torvalds 已提交
2489 2490
}

2491
static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data)
L
Linus Torvalds 已提交
2492 2493 2494 2495 2496
{
	struct inode *inode = data->inode;
	
	if (nfs4_async_handle_error(task, NFS_SERVER(inode)) == -EAGAIN) {
		rpc_restart_call(task);
2497
		return -EAGAIN;
L
Linus Torvalds 已提交
2498
	}
2499
	if (task->tk_status >= 0) {
L
Linus Torvalds 已提交
2500
		renew_lease(NFS_SERVER(inode), data->timestamp);
2501 2502
		nfs_post_op_update_inode(inode, data->res.fattr);
	}
2503
	return 0;
L
Linus Torvalds 已提交
2504 2505
}

2506
static void nfs4_proc_write_setup(struct nfs_write_data *data, int how)
L
Linus Torvalds 已提交
2507 2508 2509 2510 2511 2512 2513 2514
{
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_WRITE],
		.rpc_argp = &data->args,
		.rpc_resp = &data->res,
		.rpc_cred = data->cred,
	};
	struct inode *inode = data->inode;
2515
	struct nfs_server *server = NFS_SERVER(inode);
L
Linus Torvalds 已提交
2516 2517 2518 2519 2520 2521 2522 2523 2524 2525
	int stable;
	
	if (how & FLUSH_STABLE) {
		if (!NFS_I(inode)->ncommit)
			stable = NFS_FILE_SYNC;
		else
			stable = NFS_DATA_SYNC;
	} else
		stable = NFS_UNSTABLE;
	data->args.stable = stable;
2526 2527
	data->args.bitmask = server->attr_bitmask;
	data->res.server = server;
L
Linus Torvalds 已提交
2528 2529 2530 2531

	data->timestamp   = jiffies;

	/* Finalize the task. */
2532
	rpc_call_setup(&data->task, &msg, 0);
L
Linus Torvalds 已提交
2533 2534
}

2535
static int nfs4_commit_done(struct rpc_task *task, struct nfs_write_data *data)
L
Linus Torvalds 已提交
2536 2537 2538 2539 2540
{
	struct inode *inode = data->inode;
	
	if (nfs4_async_handle_error(task, NFS_SERVER(inode)) == -EAGAIN) {
		rpc_restart_call(task);
2541
		return -EAGAIN;
L
Linus Torvalds 已提交
2542
	}
2543 2544
	if (task->tk_status >= 0)
		nfs_post_op_update_inode(inode, data->res.fattr);
2545
	return 0;
L
Linus Torvalds 已提交
2546 2547
}

2548
static void nfs4_proc_commit_setup(struct nfs_write_data *data, int how)
L
Linus Torvalds 已提交
2549 2550 2551 2552 2553 2554 2555
{
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_COMMIT],
		.rpc_argp = &data->args,
		.rpc_resp = &data->res,
		.rpc_cred = data->cred,
	};	
2556
	struct nfs_server *server = NFS_SERVER(data->inode);
L
Linus Torvalds 已提交
2557
	
2558 2559 2560
	data->args.bitmask = server->attr_bitmask;
	data->res.server = server;

2561
	rpc_call_setup(&data->task, &msg, 0);
L
Linus Torvalds 已提交
2562 2563 2564 2565 2566 2567
}

/*
 * nfs4_proc_async_renew(): This is not one of the nfs_rpc_ops; it is a special
 * standalone procedure for queueing an asynchronous RENEW.
 */
2568
static void nfs4_renew_done(struct rpc_task *task, void *data)
L
Linus Torvalds 已提交
2569
{
2570
	struct nfs_client *clp = (struct nfs_client *)task->tk_msg.rpc_argp;
2571
	unsigned long timestamp = (unsigned long)data;
L
Linus Torvalds 已提交
2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587

	if (task->tk_status < 0) {
		switch (task->tk_status) {
			case -NFS4ERR_STALE_CLIENTID:
			case -NFS4ERR_EXPIRED:
			case -NFS4ERR_CB_PATH_DOWN:
				nfs4_schedule_state_recovery(clp);
		}
		return;
	}
	spin_lock(&clp->cl_lock);
	if (time_before(clp->cl_last_renewal,timestamp))
		clp->cl_last_renewal = timestamp;
	spin_unlock(&clp->cl_lock);
}

2588 2589 2590 2591
static const struct rpc_call_ops nfs4_renew_ops = {
	.rpc_call_done = nfs4_renew_done,
};

2592
int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred)
L
Linus Torvalds 已提交
2593 2594 2595 2596
{
	struct rpc_message msg = {
		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_RENEW],
		.rpc_argp	= clp,
2597
		.rpc_cred	= cred,
L
Linus Torvalds 已提交
2598 2599 2600
	};

	return rpc_call_async(clp->cl_rpcclient, &msg, RPC_TASK_SOFT,
2601
			&nfs4_renew_ops, (void *)jiffies);
L
Linus Torvalds 已提交
2602 2603
}

2604
int nfs4_proc_renew(struct nfs_client *clp, struct rpc_cred *cred)
L
Linus Torvalds 已提交
2605 2606 2607 2608
{
	struct rpc_message msg = {
		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_RENEW],
		.rpc_argp	= clp,
2609
		.rpc_cred	= cred,
L
Linus Torvalds 已提交
2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623
	};
	unsigned long now = jiffies;
	int status;

	status = rpc_call_sync(clp->cl_rpcclient, &msg, 0);
	if (status < 0)
		return status;
	spin_lock(&clp->cl_lock);
	if (time_before(clp->cl_last_renewal,now))
		clp->cl_last_renewal = now;
	spin_unlock(&clp->cl_lock);
	return 0;
}

2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649
static inline int nfs4_server_supports_acls(struct nfs_server *server)
{
	return (server->caps & NFS_CAP_ACLS)
		&& (server->acl_bitmask & ACL4_SUPPORT_ALLOW_ACL)
		&& (server->acl_bitmask & ACL4_SUPPORT_DENY_ACL);
}

/* Assuming that XATTR_SIZE_MAX is a multiple of PAGE_CACHE_SIZE, and that
 * it's OK to put sizeof(void) * (XATTR_SIZE_MAX/PAGE_CACHE_SIZE) bytes on
 * the stack.
 */
#define NFS4ACL_MAXPAGES (XATTR_SIZE_MAX >> PAGE_CACHE_SHIFT)

static void buf_to_pages(const void *buf, size_t buflen,
		struct page **pages, unsigned int *pgbase)
{
	const void *p = buf;

	*pgbase = offset_in_page(buf);
	p -= *pgbase;
	while (p < buf + buflen) {
		*(pages++) = virt_to_page(p);
		p += PAGE_CACHE_SIZE;
	}
}

2650 2651 2652
struct nfs4_cached_acl {
	int cached;
	size_t len;
A
Andrew Morton 已提交
2653
	char data[0];
2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716
};

static void nfs4_set_cached_acl(struct inode *inode, struct nfs4_cached_acl *acl)
{
	struct nfs_inode *nfsi = NFS_I(inode);

	spin_lock(&inode->i_lock);
	kfree(nfsi->nfs4_acl);
	nfsi->nfs4_acl = acl;
	spin_unlock(&inode->i_lock);
}

static void nfs4_zap_acl_attr(struct inode *inode)
{
	nfs4_set_cached_acl(inode, NULL);
}

static inline ssize_t nfs4_read_cached_acl(struct inode *inode, char *buf, size_t buflen)
{
	struct nfs_inode *nfsi = NFS_I(inode);
	struct nfs4_cached_acl *acl;
	int ret = -ENOENT;

	spin_lock(&inode->i_lock);
	acl = nfsi->nfs4_acl;
	if (acl == NULL)
		goto out;
	if (buf == NULL) /* user is just asking for length */
		goto out_len;
	if (acl->cached == 0)
		goto out;
	ret = -ERANGE; /* see getxattr(2) man page */
	if (acl->len > buflen)
		goto out;
	memcpy(buf, acl->data, acl->len);
out_len:
	ret = acl->len;
out:
	spin_unlock(&inode->i_lock);
	return ret;
}

static void nfs4_write_cached_acl(struct inode *inode, const char *buf, size_t acl_len)
{
	struct nfs4_cached_acl *acl;

	if (buf && acl_len <= PAGE_SIZE) {
		acl = kmalloc(sizeof(*acl) + acl_len, GFP_KERNEL);
		if (acl == NULL)
			goto out;
		acl->cached = 1;
		memcpy(acl->data, buf, acl_len);
	} else {
		acl = kmalloc(sizeof(*acl), GFP_KERNEL);
		if (acl == NULL)
			goto out;
		acl->cached = 0;
	}
	acl->len = acl_len;
out:
	nfs4_set_cached_acl(inode, acl);
}

2717
static ssize_t __nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t buflen)
2718 2719 2720 2721 2722 2723 2724 2725
{
	struct page *pages[NFS4ACL_MAXPAGES];
	struct nfs_getaclargs args = {
		.fh = NFS_FH(inode),
		.acl_pages = pages,
		.acl_len = buflen,
	};
	size_t resp_len = buflen;
2726
	void *resp_buf;
2727 2728 2729 2730 2731
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_GETACL],
		.rpc_argp = &args,
		.rpc_resp = &resp_len,
	};
2732
	struct page *localpage = NULL;
2733 2734
	int ret;

2735 2736 2737 2738 2739 2740 2741 2742 2743
	if (buflen < PAGE_SIZE) {
		/* As long as we're doing a round trip to the server anyway,
		 * let's be prepared for a page of acl data. */
		localpage = alloc_page(GFP_KERNEL);
		resp_buf = page_address(localpage);
		if (localpage == NULL)
			return -ENOMEM;
		args.acl_pages[0] = localpage;
		args.acl_pgbase = 0;
J
J. Bruce Fields 已提交
2744
		resp_len = args.acl_len = PAGE_SIZE;
2745 2746 2747 2748
	} else {
		resp_buf = buf;
		buf_to_pages(buf, buflen, args.acl_pages, &args.acl_pgbase);
	}
2749
	ret = rpc_call_sync(NFS_CLIENT(inode), &msg, 0);
2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766
	if (ret)
		goto out_free;
	if (resp_len > args.acl_len)
		nfs4_write_cached_acl(inode, NULL, resp_len);
	else
		nfs4_write_cached_acl(inode, resp_buf, resp_len);
	if (buf) {
		ret = -ERANGE;
		if (resp_len > buflen)
			goto out_free;
		if (localpage)
			memcpy(buf, resp_buf, resp_len);
	}
	ret = resp_len;
out_free:
	if (localpage)
		__free_page(localpage);
2767 2768 2769
	return ret;
}

2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782
static ssize_t nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t buflen)
{
	struct nfs4_exception exception = { };
	ssize_t ret;
	do {
		ret = __nfs4_get_acl_uncached(inode, buf, buflen);
		if (ret >= 0)
			break;
		ret = nfs4_handle_exception(NFS_SERVER(inode), ret, &exception);
	} while (exception.retry);
	return ret;
}

2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798
static ssize_t nfs4_proc_get_acl(struct inode *inode, void *buf, size_t buflen)
{
	struct nfs_server *server = NFS_SERVER(inode);
	int ret;

	if (!nfs4_server_supports_acls(server))
		return -EOPNOTSUPP;
	ret = nfs_revalidate_inode(server, inode);
	if (ret < 0)
		return ret;
	ret = nfs4_read_cached_acl(inode, buf, buflen);
	if (ret != -ENOENT)
		return ret;
	return nfs4_get_acl_uncached(inode, buf, buflen);
}

2799
static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen)
2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816
{
	struct nfs_server *server = NFS_SERVER(inode);
	struct page *pages[NFS4ACL_MAXPAGES];
	struct nfs_setaclargs arg = {
		.fh		= NFS_FH(inode),
		.acl_pages	= pages,
		.acl_len	= buflen,
	};
	struct rpc_message msg = {
		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_SETACL],
		.rpc_argp	= &arg,
		.rpc_resp	= NULL,
	};
	int ret;

	if (!nfs4_server_supports_acls(server))
		return -EOPNOTSUPP;
2817
	nfs_inode_return_delegation(inode);
2818 2819
	buf_to_pages(buf, buflen, arg.acl_pages, &arg.acl_pgbase);
	ret = rpc_call_sync(NFS_SERVER(inode)->client, &msg, 0);
2820 2821
	if (ret == 0)
		nfs4_write_cached_acl(inode, buf, buflen);
2822 2823 2824
	return ret;
}

2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836
static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen)
{
	struct nfs4_exception exception = { };
	int err;
	do {
		err = nfs4_handle_exception(NFS_SERVER(inode),
				__nfs4_proc_set_acl(inode, buf, buflen),
				&exception);
	} while (exception.retry);
	return err;
}

L
Linus Torvalds 已提交
2837
static int
2838
nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server)
L
Linus Torvalds 已提交
2839
{
2840
	struct nfs_client *clp = server->nfs_client;
L
Linus Torvalds 已提交
2841 2842 2843 2844 2845 2846 2847 2848 2849

	if (!clp || task->tk_status >= 0)
		return 0;
	switch(task->tk_status) {
		case -NFS4ERR_STALE_CLIENTID:
		case -NFS4ERR_STALE_STATEID:
		case -NFS4ERR_EXPIRED:
			rpc_sleep_on(&clp->cl_rpcwaitq, task, NULL, NULL);
			nfs4_schedule_state_recovery(clp);
T
Trond Myklebust 已提交
2850
			if (test_bit(NFS4CLNT_STATE_RECOVER, &clp->cl_state) == 0)
L
Linus Torvalds 已提交
2851 2852 2853 2854
				rpc_wake_up_task(task);
			task->tk_status = 0;
			return -EAGAIN;
		case -NFS4ERR_DELAY:
2855 2856 2857
			nfs_inc_server_stats((struct nfs_server *) server,
						NFSIOS_DELAY);
		case -NFS4ERR_GRACE:
L
Linus Torvalds 已提交
2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868
			rpc_delay(task, NFS4_POLL_RETRY_MAX);
			task->tk_status = 0;
			return -EAGAIN;
		case -NFS4ERR_OLD_STATEID:
			task->tk_status = 0;
			return -EAGAIN;
	}
	task->tk_status = nfs4_map_errors(task->tk_status);
	return 0;
}

T
Trond Myklebust 已提交
2869 2870 2871 2872 2873 2874 2875 2876
static int nfs4_wait_bit_interruptible(void *word)
{
	if (signal_pending(current))
		return -ERESTARTSYS;
	schedule();
	return 0;
}

2877
static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp)
L
Linus Torvalds 已提交
2878 2879
{
	sigset_t oldset;
T
Trond Myklebust 已提交
2880
	int res;
L
Linus Torvalds 已提交
2881 2882 2883 2884

	might_sleep();

	rpc_clnt_sigmask(clnt, &oldset);
T
Trond Myklebust 已提交
2885 2886 2887
	res = wait_on_bit(&clp->cl_state, NFS4CLNT_STATE_RECOVER,
			nfs4_wait_bit_interruptible,
			TASK_INTERRUPTIBLE);
L
Linus Torvalds 已提交
2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904
	rpc_clnt_sigunmask(clnt, &oldset);
	return res;
}

static int nfs4_delay(struct rpc_clnt *clnt, long *timeout)
{
	sigset_t oldset;
	int res = 0;

	might_sleep();

	if (*timeout <= 0)
		*timeout = NFS4_POLL_RETRY_MIN;
	if (*timeout > NFS4_POLL_RETRY_MAX)
		*timeout = NFS4_POLL_RETRY_MAX;
	rpc_clnt_sigmask(clnt, &oldset);
	if (clnt->cl_intr) {
2905
		schedule_timeout_interruptible(*timeout);
L
Linus Torvalds 已提交
2906 2907
		if (signalled())
			res = -ERESTARTSYS;
2908 2909
	} else
		schedule_timeout_uninterruptible(*timeout);
L
Linus Torvalds 已提交
2910 2911 2912 2913 2914 2915 2916 2917
	rpc_clnt_sigunmask(clnt, &oldset);
	*timeout <<= 1;
	return res;
}

/* This is the error handling routine for processes that are allowed
 * to sleep.
 */
2918
int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception)
L
Linus Torvalds 已提交
2919
{
2920
	struct nfs_client *clp = server->nfs_client;
L
Linus Torvalds 已提交
2921 2922 2923 2924 2925 2926 2927 2928 2929
	int ret = errorcode;

	exception->retry = 0;
	switch(errorcode) {
		case 0:
			return 0;
		case -NFS4ERR_STALE_CLIENTID:
		case -NFS4ERR_STALE_STATEID:
		case -NFS4ERR_EXPIRED:
T
Trond Myklebust 已提交
2930
			nfs4_schedule_state_recovery(clp);
L
Linus Torvalds 已提交
2931 2932 2933 2934 2935 2936 2937
			ret = nfs4_wait_clnt_recover(server->client, clp);
			if (ret == 0)
				exception->retry = 1;
			break;
		case -NFS4ERR_GRACE:
		case -NFS4ERR_DELAY:
			ret = nfs4_delay(server->client, &exception->timeout);
2938 2939
			if (ret != 0)
				break;
L
Linus Torvalds 已提交
2940
		case -NFS4ERR_OLD_STATEID:
2941
			exception->retry = 1;
L
Linus Torvalds 已提交
2942 2943 2944 2945 2946
	}
	/* We failed to handle the error */
	return nfs4_map_errors(ret);
}

2947
int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, unsigned short port, struct rpc_cred *cred)
L
Linus Torvalds 已提交
2948 2949 2950 2951 2952 2953 2954 2955 2956 2957
{
	nfs4_verifier sc_verifier;
	struct nfs4_setclientid setclientid = {
		.sc_verifier = &sc_verifier,
		.sc_prog = program,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SETCLIENTID],
		.rpc_argp = &setclientid,
		.rpc_resp = clp,
2958
		.rpc_cred = cred,
L
Linus Torvalds 已提交
2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970
	};
	u32 *p;
	int loop = 0;
	int status;

	p = (u32*)sc_verifier.data;
	*p++ = htonl((u32)clp->cl_boot_time.tv_sec);
	*p = htonl((u32)clp->cl_boot_time.tv_nsec);

	for(;;) {
		setclientid.sc_name_len = scnprintf(setclientid.sc_name,
				sizeof(setclientid.sc_name), "%s/%u.%u.%u.%u %s %u",
2971
				clp->cl_ipaddr, NIPQUAD(clp->cl_addr.sin_addr),
2972
				cred->cr_ops->cr_name,
L
Linus Torvalds 已提交
2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993
				clp->cl_id_uniquifier);
		setclientid.sc_netid_len = scnprintf(setclientid.sc_netid,
				sizeof(setclientid.sc_netid), "tcp");
		setclientid.sc_uaddr_len = scnprintf(setclientid.sc_uaddr,
				sizeof(setclientid.sc_uaddr), "%s.%d.%d",
				clp->cl_ipaddr, port >> 8, port & 255);

		status = rpc_call_sync(clp->cl_rpcclient, &msg, 0);
		if (status != -NFS4ERR_CLID_INUSE)
			break;
		if (signalled())
			break;
		if (loop++ & 1)
			ssleep(clp->cl_lease_time + 1);
		else
			if (++clp->cl_id_uniquifier == 0)
				break;
	}
	return status;
}

2994
static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct rpc_cred *cred)
L
Linus Torvalds 已提交
2995 2996 2997 2998 2999 3000
{
	struct nfs_fsinfo fsinfo;
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SETCLIENTID_CONFIRM],
		.rpc_argp = clp,
		.rpc_resp = &fsinfo,
3001
		.rpc_cred = cred,
L
Linus Torvalds 已提交
3002 3003 3004 3005 3006 3007 3008 3009 3010 3011
	};
	unsigned long now;
	int status;

	now = jiffies;
	status = rpc_call_sync(clp->cl_rpcclient, &msg, 0);
	if (status == 0) {
		spin_lock(&clp->cl_lock);
		clp->cl_lease_time = fsinfo.lease_time * HZ;
		clp->cl_last_renewal = now;
3012
		clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
L
Linus Torvalds 已提交
3013 3014 3015 3016 3017
		spin_unlock(&clp->cl_lock);
	}
	return status;
}

3018
int nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct rpc_cred *cred)
3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035
{
	long timeout;
	int err;
	do {
		err = _nfs4_proc_setclientid_confirm(clp, cred);
		switch (err) {
			case 0:
				return err;
			case -NFS4ERR_RESOURCE:
				/* The IBM lawyers misread another document! */
			case -NFS4ERR_DELAY:
				err = nfs4_delay(clp->cl_rpcclient, &timeout);
		}
	} while (err == 0);
	return err;
}

3036 3037
struct nfs4_delegreturndata {
	struct nfs4_delegreturnargs args;
3038
	struct nfs4_delegreturnres res;
3039 3040 3041
	struct nfs_fh fh;
	nfs4_stateid stateid;
	struct rpc_cred *cred;
3042
	unsigned long timestamp;
3043
	struct nfs_fattr fattr;
3044 3045 3046 3047
	int rpc_status;
};

static void nfs4_delegreturn_prepare(struct rpc_task *task, void *calldata)
L
Linus Torvalds 已提交
3048
{
3049
	struct nfs4_delegreturndata *data = calldata;
L
Linus Torvalds 已提交
3050 3051
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_DELEGRETURN],
3052
		.rpc_argp = &data->args,
3053
		.rpc_resp = &data->res,
3054
		.rpc_cred = data->cred,
L
Linus Torvalds 已提交
3055
	};
3056
	nfs_fattr_init(data->res.fattr);
3057 3058
	rpc_call_setup(task, &msg, 0);
}
L
Linus Torvalds 已提交
3059

3060 3061 3062 3063
static void nfs4_delegreturn_done(struct rpc_task *task, void *calldata)
{
	struct nfs4_delegreturndata *data = calldata;
	data->rpc_status = task->tk_status;
3064
	if (data->rpc_status == 0)
3065
		renew_lease(data->res.server, data->timestamp);
3066 3067 3068 3069 3070 3071 3072 3073 3074 3075
}

static void nfs4_delegreturn_release(void *calldata)
{
	struct nfs4_delegreturndata *data = calldata;

	put_rpccred(data->cred);
	kfree(calldata);
}

3076
static const struct rpc_call_ops nfs4_delegreturn_ops = {
3077 3078 3079 3080 3081 3082 3083 3084
	.rpc_call_prepare = nfs4_delegreturn_prepare,
	.rpc_call_done = nfs4_delegreturn_done,
	.rpc_release = nfs4_delegreturn_release,
};

static int _nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4_stateid *stateid)
{
	struct nfs4_delegreturndata *data;
3085
	struct nfs_server *server = NFS_SERVER(inode);
3086 3087 3088 3089 3090 3091 3092 3093
	struct rpc_task *task;
	int status;

	data = kmalloc(sizeof(*data), GFP_KERNEL);
	if (data == NULL)
		return -ENOMEM;
	data->args.fhandle = &data->fh;
	data->args.stateid = &data->stateid;
3094
	data->args.bitmask = server->attr_bitmask;
3095 3096
	nfs_copy_fh(&data->fh, NFS_FH(inode));
	memcpy(&data->stateid, stateid, sizeof(data->stateid));
3097 3098
	data->res.fattr = &data->fattr;
	data->res.server = server;
3099
	data->cred = get_rpccred(cred);
3100
	data->timestamp = jiffies;
3101 3102 3103
	data->rpc_status = 0;

	task = rpc_run_task(NFS_CLIENT(inode), RPC_TASK_ASYNC, &nfs4_delegreturn_ops, data);
3104
	if (IS_ERR(task))
3105 3106
		return PTR_ERR(task);
	status = nfs4_wait_for_completion_rpc_task(task);
3107
	if (status == 0) {
3108
		status = data->rpc_status;
3109 3110 3111
		if (status == 0)
			nfs_post_op_update_inode(inode, &data->fattr);
	}
3112 3113
	rpc_release_task(task);
	return status;
L
Linus Torvalds 已提交
3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125
}

int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4_stateid *stateid)
{
	struct nfs_server *server = NFS_SERVER(inode);
	struct nfs4_exception exception = { };
	int err;
	do {
		err = _nfs4_proc_delegreturn(inode, cred, stateid);
		switch (err) {
			case -NFS4ERR_STALE_STATEID:
			case -NFS4ERR_EXPIRED:
3126
				nfs4_schedule_state_recovery(server->nfs_client);
L
Linus Torvalds 已提交
3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143
			case 0:
				return 0;
		}
		err = nfs4_handle_exception(server, err, &exception);
	} while (exception.retry);
	return err;
}

#define NFS4_LOCK_MINTIMEOUT (1 * HZ)
#define NFS4_LOCK_MAXTIMEOUT (30 * HZ)

/* 
 * sleep, with exponential backoff, and retry the LOCK operation. 
 */
static unsigned long
nfs4_set_lock_task_retry(unsigned long timeout)
{
3144
	schedule_timeout_interruptible(timeout);
L
Linus Torvalds 已提交
3145 3146 3147 3148 3149 3150 3151 3152 3153 3154
	timeout <<= 1;
	if (timeout > NFS4_LOCK_MAXTIMEOUT)
		return NFS4_LOCK_MAXTIMEOUT;
	return timeout;
}

static int _nfs4_proc_getlk(struct nfs4_state *state, int cmd, struct file_lock *request)
{
	struct inode *inode = state->inode;
	struct nfs_server *server = NFS_SERVER(inode);
3155
	struct nfs_client *clp = server->nfs_client;
T
Trond Myklebust 已提交
3156
	struct nfs_lockt_args arg = {
L
Linus Torvalds 已提交
3157
		.fh = NFS_FH(inode),
T
Trond Myklebust 已提交
3158
		.fl = request,
L
Linus Torvalds 已提交
3159
	};
T
Trond Myklebust 已提交
3160 3161
	struct nfs_lockt_res res = {
		.denied = request,
L
Linus Torvalds 已提交
3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172
	};
	struct rpc_message msg = {
		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_LOCKT],
		.rpc_argp       = &arg,
		.rpc_resp       = &res,
		.rpc_cred	= state->owner->so_cred,
	};
	struct nfs4_lock_state *lsp;
	int status;

	down_read(&clp->cl_sem);
T
Trond Myklebust 已提交
3173
	arg.lock_owner.clientid = clp->cl_clientid;
3174 3175 3176 3177
	status = nfs4_set_lock_state(state, request);
	if (status != 0)
		goto out;
	lsp = request->fl_u.nfs4_fl.owner;
T
Trond Myklebust 已提交
3178
	arg.lock_owner.id = lsp->ls_id; 
L
Linus Torvalds 已提交
3179
	status = rpc_call_sync(server->client, &msg, 0);
T
Trond Myklebust 已提交
3180 3181 3182 3183 3184 3185
	switch (status) {
		case 0:
			request->fl_type = F_UNLCK;
			break;
		case -NFS4ERR_DENIED:
			status = 0;
L
Linus Torvalds 已提交
3186
	}
3187
out:
L
Linus Torvalds 已提交
3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220
	up_read(&clp->cl_sem);
	return status;
}

static int nfs4_proc_getlk(struct nfs4_state *state, int cmd, struct file_lock *request)
{
	struct nfs4_exception exception = { };
	int err;

	do {
		err = nfs4_handle_exception(NFS_SERVER(state->inode),
				_nfs4_proc_getlk(state, cmd, request),
				&exception);
	} while (exception.retry);
	return err;
}

static int do_vfs_lock(struct file *file, struct file_lock *fl)
{
	int res = 0;
	switch (fl->fl_flags & (FL_POSIX|FL_FLOCK)) {
		case FL_POSIX:
			res = posix_lock_file_wait(file, fl);
			break;
		case FL_FLOCK:
			res = flock_lock_file_wait(file, fl);
			break;
		default:
			BUG();
	}
	return res;
}

3221
struct nfs4_unlockdata {
T
Trond Myklebust 已提交
3222 3223
	struct nfs_locku_args arg;
	struct nfs_locku_res res;
3224 3225
	struct nfs4_lock_state *lsp;
	struct nfs_open_context *ctx;
T
Trond Myklebust 已提交
3226 3227
	struct file_lock fl;
	const struct nfs_server *server;
3228
	unsigned long timestamp;
3229 3230
};

T
Trond Myklebust 已提交
3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254
static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl,
		struct nfs_open_context *ctx,
		struct nfs4_lock_state *lsp,
		struct nfs_seqid *seqid)
{
	struct nfs4_unlockdata *p;
	struct inode *inode = lsp->ls_state->inode;

	p = kmalloc(sizeof(*p), GFP_KERNEL);
	if (p == NULL)
		return NULL;
	p->arg.fh = NFS_FH(inode);
	p->arg.fl = &p->fl;
	p->arg.seqid = seqid;
	p->arg.stateid = &lsp->ls_stateid;
	p->lsp = lsp;
	atomic_inc(&lsp->ls_count);
	/* Ensure we don't close file until we're done freeing locks! */
	p->ctx = get_nfs_open_context(ctx);
	memcpy(&p->fl, fl, sizeof(p->fl));
	p->server = NFS_SERVER(inode);
	return p;
}

3255
static void nfs4_locku_release_calldata(void *data)
3256
{
3257
	struct nfs4_unlockdata *calldata = data;
T
Trond Myklebust 已提交
3258
	nfs_free_seqid(calldata->arg.seqid);
3259 3260 3261
	nfs4_put_lock_state(calldata->lsp);
	put_nfs_open_context(calldata->ctx);
	kfree(calldata);
3262 3263
}

3264
static void nfs4_locku_done(struct rpc_task *task, void *data)
3265
{
3266
	struct nfs4_unlockdata *calldata = data;
3267

3268 3269
	if (RPC_ASSASSINATED(task))
		return;
T
Trond Myklebust 已提交
3270
	nfs_increment_lock_seqid(task->tk_status, calldata->arg.seqid);
3271 3272 3273
	switch (task->tk_status) {
		case 0:
			memcpy(calldata->lsp->ls_stateid.data,
T
Trond Myklebust 已提交
3274
					calldata->res.stateid.data,
3275
					sizeof(calldata->lsp->ls_stateid.data));
3276
			renew_lease(calldata->server, calldata->timestamp);
3277 3278 3279
			break;
		case -NFS4ERR_STALE_STATEID:
		case -NFS4ERR_EXPIRED:
3280
			nfs4_schedule_state_recovery(calldata->server->nfs_client);
3281 3282
			break;
		default:
T
Trond Myklebust 已提交
3283
			if (nfs4_async_handle_error(task, calldata->server) == -EAGAIN) {
3284 3285 3286 3287 3288
				rpc_restart_call(task);
			}
	}
}

T
Trond Myklebust 已提交
3289
static void nfs4_locku_prepare(struct rpc_task *task, void *data)
3290
{
T
Trond Myklebust 已提交
3291
	struct nfs4_unlockdata *calldata = data;
L
Linus Torvalds 已提交
3292 3293
	struct rpc_message msg = {
		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_LOCKU],
3294 3295 3296
		.rpc_argp       = &calldata->arg,
		.rpc_resp       = &calldata->res,
		.rpc_cred	= calldata->lsp->ls_state->owner->so_cred,
L
Linus Torvalds 已提交
3297
	};
3298

T
Trond Myklebust 已提交
3299
	if (nfs_wait_on_sequence(calldata->arg.seqid, task) != 0)
3300 3301
		return;
	if ((calldata->lsp->ls_flags & NFS_LOCK_INITIALIZED) == 0) {
3302 3303
		/* Note: exit _without_ running nfs4_locku_done */
		task->tk_action = NULL;
3304 3305
		return;
	}
3306
	calldata->timestamp = jiffies;
3307 3308 3309
	rpc_call_setup(task, &msg, 0);
}

3310
static const struct rpc_call_ops nfs4_locku_ops = {
T
Trond Myklebust 已提交
3311
	.rpc_call_prepare = nfs4_locku_prepare,
3312
	.rpc_call_done = nfs4_locku_done,
3313
	.rpc_release = nfs4_locku_release_calldata,
3314 3315
};

3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328
static struct rpc_task *nfs4_do_unlck(struct file_lock *fl,
		struct nfs_open_context *ctx,
		struct nfs4_lock_state *lsp,
		struct nfs_seqid *seqid)
{
	struct nfs4_unlockdata *data;

	data = nfs4_alloc_unlockdata(fl, ctx, lsp, seqid);
	if (data == NULL) {
		nfs_free_seqid(seqid);
		return ERR_PTR(-ENOMEM);
	}

3329
	return rpc_run_task(NFS_CLIENT(lsp->ls_state->inode), RPC_TASK_ASYNC, &nfs4_locku_ops, data);
3330 3331
}

3332 3333
static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock *request)
{
T
Trond Myklebust 已提交
3334
	struct nfs_seqid *seqid;
L
Linus Torvalds 已提交
3335
	struct nfs4_lock_state *lsp;
3336 3337
	struct rpc_task *task;
	int status = 0;
3338

3339
	status = nfs4_set_lock_state(state, request);
3340 3341 3342 3343
	/* Unlock _before_ we do the RPC call */
	request->fl_flags |= FL_EXISTS;
	if (do_vfs_lock(request->fl_file, request) == -ENOENT)
		goto out;
3344
	if (status != 0)
3345 3346 3347 3348
		goto out;
	/* Is this a delegated lock? */
	if (test_bit(NFS_DELEGATED_STATE, &state->flags))
		goto out;
3349
	lsp = request->fl_u.nfs4_fl.owner;
T
Trond Myklebust 已提交
3350
	seqid = nfs_alloc_seqid(&lsp->ls_seqid);
3351
	status = -ENOMEM;
T
Trond Myklebust 已提交
3352
	if (seqid == NULL)
3353
		goto out;
3354 3355 3356
	task = nfs4_do_unlck(request, request->fl_file->private_data, lsp, seqid);
	status = PTR_ERR(task);
	if (IS_ERR(task))
3357
		goto out;
3358 3359
	status = nfs4_wait_for_completion_rpc_task(task);
	rpc_release_task(task);
3360
out:
L
Linus Torvalds 已提交
3361 3362 3363
	return status;
}

3364 3365 3366 3367 3368 3369
struct nfs4_lockdata {
	struct nfs_lock_args arg;
	struct nfs_lock_res res;
	struct nfs4_lock_state *lsp;
	struct nfs_open_context *ctx;
	struct file_lock fl;
3370
	unsigned long timestamp;
3371 3372 3373 3374 3375 3376
	int rpc_status;
	int cancelled;
};

static struct nfs4_lockdata *nfs4_alloc_lockdata(struct file_lock *fl,
		struct nfs_open_context *ctx, struct nfs4_lock_state *lsp)
L
Linus Torvalds 已提交
3377
{
3378 3379
	struct nfs4_lockdata *p;
	struct inode *inode = lsp->ls_state->inode;
L
Linus Torvalds 已提交
3380
	struct nfs_server *server = NFS_SERVER(inode);
3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391

	p = kzalloc(sizeof(*p), GFP_KERNEL);
	if (p == NULL)
		return NULL;

	p->arg.fh = NFS_FH(inode);
	p->arg.fl = &p->fl;
	p->arg.lock_seqid = nfs_alloc_seqid(&lsp->ls_seqid);
	if (p->arg.lock_seqid == NULL)
		goto out_free;
	p->arg.lock_stateid = &lsp->ls_stateid;
3392
	p->arg.lock_owner.clientid = server->nfs_client->cl_clientid;
3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408
	p->arg.lock_owner.id = lsp->ls_id;
	p->lsp = lsp;
	atomic_inc(&lsp->ls_count);
	p->ctx = get_nfs_open_context(ctx);
	memcpy(&p->fl, fl, sizeof(p->fl));
	return p;
out_free:
	kfree(p);
	return NULL;
}

static void nfs4_lock_prepare(struct rpc_task *task, void *calldata)
{
	struct nfs4_lockdata *data = calldata;
	struct nfs4_state *state = data->lsp->ls_state;
	struct nfs4_state_owner *sp = state->owner;
L
Linus Torvalds 已提交
3409
	struct rpc_message msg = {
3410 3411 3412 3413
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOCK],
		.rpc_argp = &data->arg,
		.rpc_resp = &data->res,
		.rpc_cred = sp->so_cred,
L
Linus Torvalds 已提交
3414
	};
3415

3416 3417 3418 3419 3420 3421 3422 3423 3424
	if (nfs_wait_on_sequence(data->arg.lock_seqid, task) != 0)
		return;
	dprintk("%s: begin!\n", __FUNCTION__);
	/* Do we need to do an open_to_lock_owner? */
	if (!(data->arg.lock_seqid->sequence->flags & NFS_SEQID_CONFIRMED)) {
		data->arg.open_seqid = nfs_alloc_seqid(&sp->so_seqid);
		if (data->arg.open_seqid == NULL) {
			data->rpc_status = -ENOMEM;
			task->tk_action = NULL;
3425
			goto out;
3426
		}
3427 3428 3429
		data->arg.open_stateid = &state->stateid;
		data->arg.new_lock_owner = 1;
	}
3430
	data->timestamp = jiffies;
3431
	rpc_call_setup(task, &msg, 0);
3432
out:
3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455
	dprintk("%s: done!, ret = %d\n", __FUNCTION__, data->rpc_status);
}

static void nfs4_lock_done(struct rpc_task *task, void *calldata)
{
	struct nfs4_lockdata *data = calldata;

	dprintk("%s: begin!\n", __FUNCTION__);

	data->rpc_status = task->tk_status;
	if (RPC_ASSASSINATED(task))
		goto out;
	if (data->arg.new_lock_owner != 0) {
		nfs_increment_open_seqid(data->rpc_status, data->arg.open_seqid);
		if (data->rpc_status == 0)
			nfs_confirm_seqid(&data->lsp->ls_seqid, 0);
		else
			goto out;
	}
	if (data->rpc_status == 0) {
		memcpy(data->lsp->ls_stateid.data, data->res.stateid.data,
					sizeof(data->lsp->ls_stateid.data));
		data->lsp->ls_flags |= NFS_LOCK_INITIALIZED;
3456
		renew_lease(NFS_SERVER(data->ctx->dentry->d_inode), data->timestamp);
3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507
	}
	nfs_increment_lock_seqid(data->rpc_status, data->arg.lock_seqid);
out:
	dprintk("%s: done, ret = %d!\n", __FUNCTION__, data->rpc_status);
}

static void nfs4_lock_release(void *calldata)
{
	struct nfs4_lockdata *data = calldata;

	dprintk("%s: begin!\n", __FUNCTION__);
	if (data->arg.open_seqid != NULL)
		nfs_free_seqid(data->arg.open_seqid);
	if (data->cancelled != 0) {
		struct rpc_task *task;
		task = nfs4_do_unlck(&data->fl, data->ctx, data->lsp,
				data->arg.lock_seqid);
		if (!IS_ERR(task))
			rpc_release_task(task);
		dprintk("%s: cancelling lock!\n", __FUNCTION__);
	} else
		nfs_free_seqid(data->arg.lock_seqid);
	nfs4_put_lock_state(data->lsp);
	put_nfs_open_context(data->ctx);
	kfree(data);
	dprintk("%s: done!\n", __FUNCTION__);
}

static const struct rpc_call_ops nfs4_lock_ops = {
	.rpc_call_prepare = nfs4_lock_prepare,
	.rpc_call_done = nfs4_lock_done,
	.rpc_release = nfs4_lock_release,
};

static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *fl, int reclaim)
{
	struct nfs4_lockdata *data;
	struct rpc_task *task;
	int ret;

	dprintk("%s: begin!\n", __FUNCTION__);
	data = nfs4_alloc_lockdata(fl, fl->fl_file->private_data,
			fl->fl_u.nfs4_fl.owner);
	if (data == NULL)
		return -ENOMEM;
	if (IS_SETLKW(cmd))
		data->arg.block = 1;
	if (reclaim != 0)
		data->arg.reclaim = 1;
	task = rpc_run_task(NFS_CLIENT(state->inode), RPC_TASK_ASYNC,
			&nfs4_lock_ops, data);
3508
	if (IS_ERR(task))
3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519
		return PTR_ERR(task);
	ret = nfs4_wait_for_completion_rpc_task(task);
	if (ret == 0) {
		ret = data->rpc_status;
		if (ret == -NFS4ERR_DENIED)
			ret = -EAGAIN;
	} else
		data->cancelled = 1;
	rpc_release_task(task);
	dprintk("%s: done, ret = %d!\n", __FUNCTION__, ret);
	return ret;
L
Linus Torvalds 已提交
3520 3521 3522 3523
}

static int nfs4_lock_reclaim(struct nfs4_state *state, struct file_lock *request)
{
3524 3525 3526 3527 3528
	struct nfs_server *server = NFS_SERVER(state->inode);
	struct nfs4_exception exception = { };
	int err;

	do {
3529 3530 3531
		/* Cache the lock if possible... */
		if (test_bit(NFS_DELEGATED_STATE, &state->flags) != 0)
			return 0;
3532 3533 3534 3535 3536 3537
		err = _nfs4_do_setlk(state, F_SETLK, request, 1);
		if (err != -NFS4ERR_DELAY)
			break;
		nfs4_handle_exception(server, err, &exception);
	} while (exception.retry);
	return err;
L
Linus Torvalds 已提交
3538 3539 3540 3541
}

static int nfs4_lock_expired(struct nfs4_state *state, struct file_lock *request)
{
3542 3543 3544 3545
	struct nfs_server *server = NFS_SERVER(state->inode);
	struct nfs4_exception exception = { };
	int err;

3546 3547 3548
	err = nfs4_set_lock_state(state, request);
	if (err != 0)
		return err;
3549
	do {
3550 3551
		if (test_bit(NFS_DELEGATED_STATE, &state->flags) != 0)
			return 0;
3552 3553 3554 3555 3556 3557
		err = _nfs4_do_setlk(state, F_SETLK, request, 0);
		if (err != -NFS4ERR_DELAY)
			break;
		nfs4_handle_exception(server, err, &exception);
	} while (exception.retry);
	return err;
L
Linus Torvalds 已提交
3558 3559 3560 3561
}

static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *request)
{
3562
	struct nfs_client *clp = state->owner->so_client;
3563
	unsigned char fl_flags = request->fl_flags;
L
Linus Torvalds 已提交
3564 3565
	int status;

3566 3567 3568 3569
	/* Is this a delegated open? */
	status = nfs4_set_lock_state(state, request);
	if (status != 0)
		goto out;
3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587
	request->fl_flags |= FL_ACCESS;
	status = do_vfs_lock(request->fl_file, request);
	if (status < 0)
		goto out;
	down_read(&clp->cl_sem);
	if (test_bit(NFS_DELEGATED_STATE, &state->flags)) {
		struct nfs_inode *nfsi = NFS_I(state->inode);
		/* Yes: cache locks! */
		down_read(&nfsi->rwsem);
		/* ...but avoid races with delegation recall... */
		if (test_bit(NFS_DELEGATED_STATE, &state->flags)) {
			request->fl_flags = fl_flags & ~FL_SLEEP;
			status = do_vfs_lock(request->fl_file, request);
			up_read(&nfsi->rwsem);
			goto out_unlock;
		}
		up_read(&nfsi->rwsem);
	}
3588 3589
	status = _nfs4_do_setlk(state, cmd, request, 0);
	if (status != 0)
3590
		goto out_unlock;
3591
	/* Note: we always want to sleep here! */
3592
	request->fl_flags = fl_flags | FL_SLEEP;
3593 3594
	if (do_vfs_lock(request->fl_file, request) < 0)
		printk(KERN_WARNING "%s: VFS is out of sync with lock manager!\n", __FUNCTION__);
3595
out_unlock:
L
Linus Torvalds 已提交
3596
	up_read(&clp->cl_sem);
3597 3598
out:
	request->fl_flags = fl_flags;
L
Linus Torvalds 已提交
3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650
	return status;
}

static int nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *request)
{
	struct nfs4_exception exception = { };
	int err;

	do {
		err = nfs4_handle_exception(NFS_SERVER(state->inode),
				_nfs4_proc_setlk(state, cmd, request),
				&exception);
	} while (exception.retry);
	return err;
}

static int
nfs4_proc_lock(struct file *filp, int cmd, struct file_lock *request)
{
	struct nfs_open_context *ctx;
	struct nfs4_state *state;
	unsigned long timeout = NFS4_LOCK_MINTIMEOUT;
	int status;

	/* verify open state */
	ctx = (struct nfs_open_context *)filp->private_data;
	state = ctx->state;

	if (request->fl_start < 0 || request->fl_end < 0)
		return -EINVAL;

	if (IS_GETLK(cmd))
		return nfs4_proc_getlk(state, F_GETLK, request);

	if (!(IS_SETLK(cmd) || IS_SETLKW(cmd)))
		return -EINVAL;

	if (request->fl_type == F_UNLCK)
		return nfs4_proc_unlck(state, cmd, request);

	do {
		status = nfs4_proc_setlk(state, cmd, request);
		if ((status != -EAGAIN) || IS_SETLK(cmd))
			break;
		timeout = nfs4_set_lock_task_retry(timeout);
		status = -ERESTARTSYS;
		if (signalled())
			break;
	} while(status < 0);
	return status;
}

3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668
int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl)
{
	struct nfs_server *server = NFS_SERVER(state->inode);
	struct nfs4_exception exception = { };
	int err;

	err = nfs4_set_lock_state(state, fl);
	if (err != 0)
		goto out;
	do {
		err = _nfs4_do_setlk(state, F_SETLK, fl, 0);
		if (err != -NFS4ERR_DELAY)
			break;
		err = nfs4_handle_exception(server, err, &exception);
	} while (exception.retry);
out:
	return err;
}
3669

3670 3671
#define XATTR_NAME_NFSV4_ACL "system.nfs4_acl"

3672 3673 3674
int nfs4_setxattr(struct dentry *dentry, const char *key, const void *buf,
		size_t buflen, int flags)
{
3675 3676 3677 3678 3679 3680 3681 3682 3683 3684
	struct inode *inode = dentry->d_inode;

	if (strcmp(key, XATTR_NAME_NFSV4_ACL) != 0)
		return -EOPNOTSUPP;

	if (!S_ISREG(inode->i_mode) &&
	    (!S_ISDIR(inode->i_mode) || inode->i_mode & S_ISVTX))
		return -EPERM;

	return nfs4_proc_set_acl(inode, buf, buflen);
3685 3686 3687 3688 3689 3690 3691 3692 3693
}

/* The getxattr man page suggests returning -ENODATA for unknown attributes,
 * and that's what we'll do for e.g. user attributes that haven't been set.
 * But we'll follow ext2/ext3's lead by returning -EOPNOTSUPP for unsupported
 * attributes in kernel-managed attribute namespaces. */
ssize_t nfs4_getxattr(struct dentry *dentry, const char *key, void *buf,
		size_t buflen)
{
3694 3695 3696 3697 3698 3699
	struct inode *inode = dentry->d_inode;

	if (strcmp(key, XATTR_NAME_NFSV4_ACL) != 0)
		return -EOPNOTSUPP;

	return nfs4_proc_get_acl(inode, buf, buflen);
3700 3701 3702 3703
}

ssize_t nfs4_listxattr(struct dentry *dentry, char *buf, size_t buflen)
{
3704
	size_t len = strlen(XATTR_NAME_NFSV4_ACL) + 1;
3705

3706 3707
	if (!nfs4_server_supports_acls(NFS_SERVER(dentry->d_inode)))
		return 0;
3708 3709 3710
	if (buf && buflen < len)
		return -ERANGE;
	if (buf)
3711 3712
		memcpy(buf, XATTR_NAME_NFSV4_ACL, len);
	return len;
3713 3714
}

3715
int nfs4_proc_fs_locations(struct inode *dir, struct dentry *dentry,
3716
		struct nfs4_fs_locations *fs_locations, struct page *page)
3717 3718 3719
{
	struct nfs_server *server = NFS_SERVER(dir);
	u32 bitmask[2] = {
3720 3721
		[0] = FATTR4_WORD0_FSID | FATTR4_WORD0_FS_LOCATIONS,
		[1] = FATTR4_WORD1_MOUNTED_ON_FILEID,
3722 3723 3724 3725 3726 3727 3728 3729 3730 3731
	};
	struct nfs4_fs_locations_arg args = {
		.dir_fh = NFS_FH(dir),
		.name = &dentry->d_name,
		.page = page,
		.bitmask = bitmask,
	};
	struct rpc_message msg = {
		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_FS_LOCATIONS],
		.rpc_argp = &args,
3732
		.rpc_resp = fs_locations,
3733 3734 3735 3736 3737 3738
	};
	int status;

	dprintk("%s: start\n", __FUNCTION__);
	fs_locations->fattr.valid = 0;
	fs_locations->server = server;
3739
	fs_locations->nlocations = 0;
3740 3741 3742 3743 3744
	status = rpc_call_sync(server->client, &msg, 0);
	dprintk("%s: returned status = %d\n", __FUNCTION__, status);
	return status;
}

L
Linus Torvalds 已提交
3745 3746 3747 3748 3749 3750 3751 3752 3753 3754
struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops = {
	.recover_open	= nfs4_open_reclaim,
	.recover_lock	= nfs4_lock_reclaim,
};

struct nfs4_state_recovery_ops nfs4_network_partition_recovery_ops = {
	.recover_open	= nfs4_open_expired,
	.recover_lock	= nfs4_lock_expired,
};

3755 3756 3757 3758 3759 3760 3761 3762 3763
static struct inode_operations nfs4_file_inode_operations = {
	.permission	= nfs_permission,
	.getattr	= nfs_getattr,
	.setattr	= nfs_setattr,
	.getxattr	= nfs4_getxattr,
	.setxattr	= nfs4_setxattr,
	.listxattr	= nfs4_listxattr,
};

L
Linus Torvalds 已提交
3764 3765 3766 3767
struct nfs_rpc_ops	nfs_v4_clientops = {
	.version	= 4,			/* protocol version */
	.dentry_ops	= &nfs4_dentry_operations,
	.dir_inode_ops	= &nfs4_dir_inode_operations,
3768
	.file_inode_ops	= &nfs4_file_inode_operations,
L
Linus Torvalds 已提交
3769 3770 3771
	.getroot	= nfs4_proc_get_root,
	.getattr	= nfs4_proc_getattr,
	.setattr	= nfs4_proc_setattr,
D
David Howells 已提交
3772
	.lookupfh	= nfs4_proc_lookupfh,
L
Linus Torvalds 已提交
3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792
	.lookup		= nfs4_proc_lookup,
	.access		= nfs4_proc_access,
	.readlink	= nfs4_proc_readlink,
	.read		= nfs4_proc_read,
	.write		= nfs4_proc_write,
	.commit		= nfs4_proc_commit,
	.create		= nfs4_proc_create,
	.remove		= nfs4_proc_remove,
	.unlink_setup	= nfs4_proc_unlink_setup,
	.unlink_done	= nfs4_proc_unlink_done,
	.rename		= nfs4_proc_rename,
	.link		= nfs4_proc_link,
	.symlink	= nfs4_proc_symlink,
	.mkdir		= nfs4_proc_mkdir,
	.rmdir		= nfs4_proc_remove,
	.readdir	= nfs4_proc_readdir,
	.mknod		= nfs4_proc_mknod,
	.statfs		= nfs4_proc_statfs,
	.fsinfo		= nfs4_proc_fsinfo,
	.pathconf	= nfs4_proc_pathconf,
3793
	.set_capabilities = nfs4_server_capabilities,
L
Linus Torvalds 已提交
3794 3795
	.decode_dirent	= nfs4_decode_dirent,
	.read_setup	= nfs4_proc_read_setup,
T
Trond Myklebust 已提交
3796
	.read_done	= nfs4_read_done,
L
Linus Torvalds 已提交
3797
	.write_setup	= nfs4_proc_write_setup,
3798
	.write_done	= nfs4_write_done,
L
Linus Torvalds 已提交
3799
	.commit_setup	= nfs4_proc_commit_setup,
3800
	.commit_done	= nfs4_commit_done,
3801 3802
	.file_open      = nfs_open,
	.file_release   = nfs_release,
L
Linus Torvalds 已提交
3803
	.lock		= nfs4_proc_lock,
3804
	.clear_acl_cache = nfs4_zap_acl_attr,
L
Linus Torvalds 已提交
3805 3806 3807 3808 3809 3810 3811
};

/*
 * Local variables:
 *  c-basic-offset: 8
 * End:
 */