caps.c 84.3 KB
Newer Older
1
#include <linux/ceph/ceph_debug.h>
S
Sage Weil 已提交
2 3 4 5

#include <linux/fs.h>
#include <linux/kernel.h>
#include <linux/sched.h>
6
#include <linux/slab.h>
S
Sage Weil 已提交
7 8
#include <linux/vmalloc.h>
#include <linux/wait.h>
9
#include <linux/writeback.h>
S
Sage Weil 已提交
10 11

#include "super.h"
12 13 14
#include "mds_client.h"
#include <linux/ceph/decode.h>
#include <linux/ceph/messenger.h>
S
Sage Weil 已提交
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 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116

/*
 * Capability management
 *
 * The Ceph metadata servers control client access to inode metadata
 * and file data by issuing capabilities, granting clients permission
 * to read and/or write both inode field and file data to OSDs
 * (storage nodes).  Each capability consists of a set of bits
 * indicating which operations are allowed.
 *
 * If the client holds a *_SHARED cap, the client has a coherent value
 * that can be safely read from the cached inode.
 *
 * In the case of a *_EXCL (exclusive) or FILE_WR capabilities, the
 * client is allowed to change inode attributes (e.g., file size,
 * mtime), note its dirty state in the ceph_cap, and asynchronously
 * flush that metadata change to the MDS.
 *
 * In the event of a conflicting operation (perhaps by another
 * client), the MDS will revoke the conflicting client capabilities.
 *
 * In order for a client to cache an inode, it must hold a capability
 * with at least one MDS server.  When inodes are released, release
 * notifications are batched and periodically sent en masse to the MDS
 * cluster to release server state.
 */


/*
 * Generate readable cap strings for debugging output.
 */
#define MAX_CAP_STR 20
static char cap_str[MAX_CAP_STR][40];
static DEFINE_SPINLOCK(cap_str_lock);
static int last_cap_str;

static char *gcap_string(char *s, int c)
{
	if (c & CEPH_CAP_GSHARED)
		*s++ = 's';
	if (c & CEPH_CAP_GEXCL)
		*s++ = 'x';
	if (c & CEPH_CAP_GCACHE)
		*s++ = 'c';
	if (c & CEPH_CAP_GRD)
		*s++ = 'r';
	if (c & CEPH_CAP_GWR)
		*s++ = 'w';
	if (c & CEPH_CAP_GBUFFER)
		*s++ = 'b';
	if (c & CEPH_CAP_GLAZYIO)
		*s++ = 'l';
	return s;
}

const char *ceph_cap_string(int caps)
{
	int i;
	char *s;
	int c;

	spin_lock(&cap_str_lock);
	i = last_cap_str++;
	if (last_cap_str == MAX_CAP_STR)
		last_cap_str = 0;
	spin_unlock(&cap_str_lock);

	s = cap_str[i];

	if (caps & CEPH_CAP_PIN)
		*s++ = 'p';

	c = (caps >> CEPH_CAP_SAUTH) & 3;
	if (c) {
		*s++ = 'A';
		s = gcap_string(s, c);
	}

	c = (caps >> CEPH_CAP_SLINK) & 3;
	if (c) {
		*s++ = 'L';
		s = gcap_string(s, c);
	}

	c = (caps >> CEPH_CAP_SXATTR) & 3;
	if (c) {
		*s++ = 'X';
		s = gcap_string(s, c);
	}

	c = caps >> CEPH_CAP_SFILE;
	if (c) {
		*s++ = 'F';
		s = gcap_string(s, c);
	}

	if (s == cap_str[i])
		*s++ = '-';
	*s = 0;
	return cap_str[i];
}

117
void ceph_caps_init(struct ceph_mds_client *mdsc)
S
Sage Weil 已提交
118
{
119 120
	INIT_LIST_HEAD(&mdsc->caps_list);
	spin_lock_init(&mdsc->caps_list_lock);
S
Sage Weil 已提交
121 122
}

123
void ceph_caps_finalize(struct ceph_mds_client *mdsc)
S
Sage Weil 已提交
124 125 126
{
	struct ceph_cap *cap;

127 128 129 130
	spin_lock(&mdsc->caps_list_lock);
	while (!list_empty(&mdsc->caps_list)) {
		cap = list_first_entry(&mdsc->caps_list,
				       struct ceph_cap, caps_item);
S
Sage Weil 已提交
131 132 133
		list_del(&cap->caps_item);
		kmem_cache_free(ceph_cap_cachep, cap);
	}
134 135 136 137 138 139
	mdsc->caps_total_count = 0;
	mdsc->caps_avail_count = 0;
	mdsc->caps_use_count = 0;
	mdsc->caps_reserve_count = 0;
	mdsc->caps_min_count = 0;
	spin_unlock(&mdsc->caps_list_lock);
140 141
}

142
void ceph_adjust_min_caps(struct ceph_mds_client *mdsc, int delta)
143
{
144 145 146 147
	spin_lock(&mdsc->caps_list_lock);
	mdsc->caps_min_count += delta;
	BUG_ON(mdsc->caps_min_count < 0);
	spin_unlock(&mdsc->caps_list_lock);
S
Sage Weil 已提交
148 149
}

150 151
int ceph_reserve_caps(struct ceph_mds_client *mdsc,
		      struct ceph_cap_reservation *ctx, int need)
S
Sage Weil 已提交
152 153 154 155 156 157 158 159 160 161 162
{
	int i;
	struct ceph_cap *cap;
	int have;
	int alloc = 0;
	LIST_HEAD(newcaps);
	int ret = 0;

	dout("reserve caps ctx=%p need=%d\n", ctx, need);

	/* first reserve any caps that are already allocated */
163 164
	spin_lock(&mdsc->caps_list_lock);
	if (mdsc->caps_avail_count >= need)
S
Sage Weil 已提交
165 166
		have = need;
	else
167 168 169 170 171 172 173
		have = mdsc->caps_avail_count;
	mdsc->caps_avail_count -= have;
	mdsc->caps_reserve_count += have;
	BUG_ON(mdsc->caps_total_count != mdsc->caps_use_count +
					 mdsc->caps_reserve_count +
					 mdsc->caps_avail_count);
	spin_unlock(&mdsc->caps_list_lock);
S
Sage Weil 已提交
174 175 176 177 178 179 180 181 182 183 184 185

	for (i = have; i < need; i++) {
		cap = kmem_cache_alloc(ceph_cap_cachep, GFP_NOFS);
		if (!cap) {
			ret = -ENOMEM;
			goto out_alloc_count;
		}
		list_add(&cap->caps_item, &newcaps);
		alloc++;
	}
	BUG_ON(have + alloc != need);

186 187 188 189
	spin_lock(&mdsc->caps_list_lock);
	mdsc->caps_total_count += alloc;
	mdsc->caps_reserve_count += alloc;
	list_splice(&newcaps, &mdsc->caps_list);
S
Sage Weil 已提交
190

191 192 193 194
	BUG_ON(mdsc->caps_total_count != mdsc->caps_use_count +
					 mdsc->caps_reserve_count +
					 mdsc->caps_avail_count);
	spin_unlock(&mdsc->caps_list_lock);
S
Sage Weil 已提交
195 196 197

	ctx->count = need;
	dout("reserve caps ctx=%p %d = %d used + %d resv + %d avail\n",
198 199
	     ctx, mdsc->caps_total_count, mdsc->caps_use_count,
	     mdsc->caps_reserve_count, mdsc->caps_avail_count);
S
Sage Weil 已提交
200 201 202 203 204 205 206 207 208
	return 0;

out_alloc_count:
	/* we didn't manage to reserve as much as we needed */
	pr_warning("reserve caps ctx=%p ENOMEM need=%d got=%d\n",
		   ctx, need, have);
	return ret;
}

209 210
int ceph_unreserve_caps(struct ceph_mds_client *mdsc,
			struct ceph_cap_reservation *ctx)
S
Sage Weil 已提交
211 212 213
{
	dout("unreserve caps ctx=%p count=%d\n", ctx, ctx->count);
	if (ctx->count) {
214 215 216 217
		spin_lock(&mdsc->caps_list_lock);
		BUG_ON(mdsc->caps_reserve_count < ctx->count);
		mdsc->caps_reserve_count -= ctx->count;
		mdsc->caps_avail_count += ctx->count;
S
Sage Weil 已提交
218 219
		ctx->count = 0;
		dout("unreserve caps %d = %d used + %d resv + %d avail\n",
220 221 222 223 224 225
		     mdsc->caps_total_count, mdsc->caps_use_count,
		     mdsc->caps_reserve_count, mdsc->caps_avail_count);
		BUG_ON(mdsc->caps_total_count != mdsc->caps_use_count +
						 mdsc->caps_reserve_count +
						 mdsc->caps_avail_count);
		spin_unlock(&mdsc->caps_list_lock);
S
Sage Weil 已提交
226 227 228 229
	}
	return 0;
}

230 231
static struct ceph_cap *get_cap(struct ceph_mds_client *mdsc,
				struct ceph_cap_reservation *ctx)
S
Sage Weil 已提交
232 233 234 235
{
	struct ceph_cap *cap = NULL;

	/* temporary, until we do something about cap import/export */
236 237 238
	if (!ctx) {
		cap = kmem_cache_alloc(ceph_cap_cachep, GFP_NOFS);
		if (cap) {
239 240
			mdsc->caps_use_count++;
			mdsc->caps_total_count++;
241 242 243
		}
		return cap;
	}
S
Sage Weil 已提交
244

245
	spin_lock(&mdsc->caps_list_lock);
S
Sage Weil 已提交
246
	dout("get_cap ctx=%p (%d) %d = %d used + %d resv + %d avail\n",
247 248
	     ctx, ctx->count, mdsc->caps_total_count, mdsc->caps_use_count,
	     mdsc->caps_reserve_count, mdsc->caps_avail_count);
S
Sage Weil 已提交
249
	BUG_ON(!ctx->count);
250 251
	BUG_ON(ctx->count > mdsc->caps_reserve_count);
	BUG_ON(list_empty(&mdsc->caps_list));
S
Sage Weil 已提交
252 253

	ctx->count--;
254 255
	mdsc->caps_reserve_count--;
	mdsc->caps_use_count++;
S
Sage Weil 已提交
256

257
	cap = list_first_entry(&mdsc->caps_list, struct ceph_cap, caps_item);
S
Sage Weil 已提交
258 259
	list_del(&cap->caps_item);

260 261 262
	BUG_ON(mdsc->caps_total_count != mdsc->caps_use_count +
	       mdsc->caps_reserve_count + mdsc->caps_avail_count);
	spin_unlock(&mdsc->caps_list_lock);
S
Sage Weil 已提交
263 264 265
	return cap;
}

266
void ceph_put_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap)
S
Sage Weil 已提交
267
{
268
	spin_lock(&mdsc->caps_list_lock);
S
Sage Weil 已提交
269
	dout("put_cap %p %d = %d used + %d resv + %d avail\n",
270 271 272
	     cap, mdsc->caps_total_count, mdsc->caps_use_count,
	     mdsc->caps_reserve_count, mdsc->caps_avail_count);
	mdsc->caps_use_count--;
S
Sage Weil 已提交
273
	/*
274 275
	 * Keep some preallocated caps around (ceph_min_count), to
	 * avoid lots of free/alloc churn.
S
Sage Weil 已提交
276
	 */
277 278 279
	if (mdsc->caps_avail_count >= mdsc->caps_reserve_count +
				      mdsc->caps_min_count) {
		mdsc->caps_total_count--;
S
Sage Weil 已提交
280 281
		kmem_cache_free(ceph_cap_cachep, cap);
	} else {
282 283
		mdsc->caps_avail_count++;
		list_add(&cap->caps_item, &mdsc->caps_list);
S
Sage Weil 已提交
284 285
	}

286 287 288
	BUG_ON(mdsc->caps_total_count != mdsc->caps_use_count +
	       mdsc->caps_reserve_count + mdsc->caps_avail_count);
	spin_unlock(&mdsc->caps_list_lock);
S
Sage Weil 已提交
289 290
}

291
void ceph_reservation_status(struct ceph_fs_client *fsc,
292 293
			     int *total, int *avail, int *used, int *reserved,
			     int *min)
S
Sage Weil 已提交
294
{
295
	struct ceph_mds_client *mdsc = fsc->mdsc;
296

S
Sage Weil 已提交
297
	if (total)
298
		*total = mdsc->caps_total_count;
S
Sage Weil 已提交
299
	if (avail)
300
		*avail = mdsc->caps_avail_count;
S
Sage Weil 已提交
301
	if (used)
302
		*used = mdsc->caps_use_count;
S
Sage Weil 已提交
303
	if (reserved)
304
		*reserved = mdsc->caps_reserve_count;
305
	if (min)
306
		*min = mdsc->caps_min_count;
S
Sage Weil 已提交
307 308 309 310 311
}

/*
 * Find ceph_cap for given mds, if any.
 *
312
 * Called with i_ceph_lock held.
S
Sage Weil 已提交
313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330
 */
static struct ceph_cap *__get_cap_for_mds(struct ceph_inode_info *ci, int mds)
{
	struct ceph_cap *cap;
	struct rb_node *n = ci->i_caps.rb_node;

	while (n) {
		cap = rb_entry(n, struct ceph_cap, ci_node);
		if (mds < cap->mds)
			n = n->rb_left;
		else if (mds > cap->mds)
			n = n->rb_right;
		else
			return cap;
	}
	return NULL;
}

331 332 333 334
struct ceph_cap *ceph_get_cap_for_mds(struct ceph_inode_info *ci, int mds)
{
	struct ceph_cap *cap;

335
	spin_lock(&ci->i_ceph_lock);
336
	cap = __get_cap_for_mds(ci, mds);
337
	spin_unlock(&ci->i_ceph_lock);
338 339 340
	return cap;
}

S
Sage Weil 已提交
341
/*
342
 * Return id of any MDS with a cap, preferably FILE_WR|BUFFER|EXCL, else -1.
S
Sage Weil 已提交
343
 */
344
static int __ceph_get_cap_mds(struct ceph_inode_info *ci)
S
Sage Weil 已提交
345 346 347 348 349
{
	struct ceph_cap *cap;
	int mds = -1;
	struct rb_node *p;

350
	/* prefer mds with WR|BUFFER|EXCL caps */
S
Sage Weil 已提交
351 352 353 354 355 356 357 358 359 360 361 362 363
	for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
		cap = rb_entry(p, struct ceph_cap, ci_node);
		mds = cap->mds;
		if (cap->issued & (CEPH_CAP_FILE_WR |
				   CEPH_CAP_FILE_BUFFER |
				   CEPH_CAP_FILE_EXCL))
			break;
	}
	return mds;
}

int ceph_get_cap_mds(struct inode *inode)
{
364
	struct ceph_inode_info *ci = ceph_inode(inode);
S
Sage Weil 已提交
365
	int mds;
366
	spin_lock(&ci->i_ceph_lock);
367
	mds = __ceph_get_cap_mds(ceph_inode(inode));
368
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
369 370 371 372
	return mds;
}

/*
373
 * Called under i_ceph_lock.
S
Sage Weil 已提交
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
 */
static void __insert_cap_node(struct ceph_inode_info *ci,
			      struct ceph_cap *new)
{
	struct rb_node **p = &ci->i_caps.rb_node;
	struct rb_node *parent = NULL;
	struct ceph_cap *cap = NULL;

	while (*p) {
		parent = *p;
		cap = rb_entry(parent, struct ceph_cap, ci_node);
		if (new->mds < cap->mds)
			p = &(*p)->rb_left;
		else if (new->mds > cap->mds)
			p = &(*p)->rb_right;
		else
			BUG();
	}

	rb_link_node(&new->ci_node, parent, p);
	rb_insert_color(&new->ci_node, &ci->i_caps);
}

/*
 * (re)set cap hold timeouts, which control the delayed release
 * of unused caps back to the MDS.  Should be called on cap use.
 */
static void __cap_set_timeouts(struct ceph_mds_client *mdsc,
			       struct ceph_inode_info *ci)
{
404
	struct ceph_mount_options *ma = mdsc->fsc->mount_options;
S
Sage Weil 已提交
405 406 407 408 409 410 411 412 413 414 415 416 417 418

	ci->i_hold_caps_min = round_jiffies(jiffies +
					    ma->caps_wanted_delay_min * HZ);
	ci->i_hold_caps_max = round_jiffies(jiffies +
					    ma->caps_wanted_delay_max * HZ);
	dout("__cap_set_timeouts %p min %lu max %lu\n", &ci->vfs_inode,
	     ci->i_hold_caps_min - jiffies, ci->i_hold_caps_max - jiffies);
}

/*
 * (Re)queue cap at the end of the delayed cap release list.
 *
 * If I_FLUSH is set, leave the inode at the front of the list.
 *
419
 * Caller holds i_ceph_lock
S
Sage Weil 已提交
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
 *    -> we take mdsc->cap_delay_lock
 */
static void __cap_delay_requeue(struct ceph_mds_client *mdsc,
				struct ceph_inode_info *ci)
{
	__cap_set_timeouts(mdsc, ci);
	dout("__cap_delay_requeue %p flags %d at %lu\n", &ci->vfs_inode,
	     ci->i_ceph_flags, ci->i_hold_caps_max);
	if (!mdsc->stopping) {
		spin_lock(&mdsc->cap_delay_lock);
		if (!list_empty(&ci->i_cap_delay_list)) {
			if (ci->i_ceph_flags & CEPH_I_FLUSH)
				goto no_change;
			list_del_init(&ci->i_cap_delay_list);
		}
		list_add_tail(&ci->i_cap_delay_list, &mdsc->cap_delay_list);
no_change:
		spin_unlock(&mdsc->cap_delay_lock);
	}
}

/*
 * Queue an inode for immediate writeback.  Mark inode with I_FLUSH,
 * indicating we should send a cap message to flush dirty metadata
 * asap, and move to the front of the delayed cap list.
 */
static void __cap_delay_requeue_front(struct ceph_mds_client *mdsc,
				      struct ceph_inode_info *ci)
{
	dout("__cap_delay_requeue_front %p\n", &ci->vfs_inode);
	spin_lock(&mdsc->cap_delay_lock);
	ci->i_ceph_flags |= CEPH_I_FLUSH;
	if (!list_empty(&ci->i_cap_delay_list))
		list_del_init(&ci->i_cap_delay_list);
	list_add(&ci->i_cap_delay_list, &mdsc->cap_delay_list);
	spin_unlock(&mdsc->cap_delay_lock);
}

/*
 * Cancel delayed work on cap.
 *
461
 * Caller must hold i_ceph_lock.
S
Sage Weil 已提交
462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485
 */
static void __cap_delay_cancel(struct ceph_mds_client *mdsc,
			       struct ceph_inode_info *ci)
{
	dout("__cap_delay_cancel %p\n", &ci->vfs_inode);
	if (list_empty(&ci->i_cap_delay_list))
		return;
	spin_lock(&mdsc->cap_delay_lock);
	list_del_init(&ci->i_cap_delay_list);
	spin_unlock(&mdsc->cap_delay_lock);
}

/*
 * Common issue checks for add_cap, handle_cap_grant.
 */
static void __check_cap_issue(struct ceph_inode_info *ci, struct ceph_cap *cap,
			      unsigned issued)
{
	unsigned had = __ceph_caps_issued(ci, NULL);

	/*
	 * Each time we receive FILE_CACHE anew, we increment
	 * i_rdcache_gen.
	 */
486 487
	if ((issued & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) &&
	    (had & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) == 0)
S
Sage Weil 已提交
488 489 490
		ci->i_rdcache_gen++;

	/*
S
Sage Weil 已提交
491
	 * if we are newly issued FILE_SHARED, clear D_COMPLETE; we
S
Sage Weil 已提交
492 493 494 495 496 497
	 * don't know what happened to this directory while we didn't
	 * have the cap.
	 */
	if ((issued & CEPH_CAP_FILE_SHARED) &&
	    (had & CEPH_CAP_FILE_SHARED) == 0) {
		ci->i_shared_gen++;
S
Sage Weil 已提交
498 499
		if (S_ISDIR(ci->vfs_inode.i_mode))
			ceph_dir_clear_complete(&ci->vfs_inode);
S
Sage Weil 已提交
500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517
	}
}

/*
 * Add a capability under the given MDS session.
 *
 * Caller should hold session snap_rwsem (read) and s_mutex.
 *
 * @fmode is the open file mode, if we are opening a file, otherwise
 * it is < 0.  (This is so we can atomically add the cap and add an
 * open file reference to it.)
 */
int ceph_add_cap(struct inode *inode,
		 struct ceph_mds_session *session, u64 cap_id,
		 int fmode, unsigned issued, unsigned wanted,
		 unsigned seq, unsigned mseq, u64 realmino, int flags,
		 struct ceph_cap_reservation *caps_reservation)
{
518
	struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc;
S
Sage Weil 已提交
519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535
	struct ceph_inode_info *ci = ceph_inode(inode);
	struct ceph_cap *new_cap = NULL;
	struct ceph_cap *cap;
	int mds = session->s_mds;
	int actual_wanted;

	dout("add_cap %p mds%d cap %llx %s seq %d\n", inode,
	     session->s_mds, cap_id, ceph_cap_string(issued), seq);

	/*
	 * If we are opening the file, include file mode wanted bits
	 * in wanted.
	 */
	if (fmode >= 0)
		wanted |= ceph_caps_for_mode(fmode);

retry:
536
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
537 538 539 540 541 542
	cap = __get_cap_for_mds(ci, mds);
	if (!cap) {
		if (new_cap) {
			cap = new_cap;
			new_cap = NULL;
		} else {
543
			spin_unlock(&ci->i_ceph_lock);
544
			new_cap = get_cap(mdsc, caps_reservation);
S
Sage Weil 已提交
545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570
			if (new_cap == NULL)
				return -ENOMEM;
			goto retry;
		}

		cap->issued = 0;
		cap->implemented = 0;
		cap->mds = mds;
		cap->mds_wanted = 0;

		cap->ci = ci;
		__insert_cap_node(ci, cap);

		/* clear out old exporting info?  (i.e. on cap import) */
		if (ci->i_cap_exporting_mds == mds) {
			ci->i_cap_exporting_issued = 0;
			ci->i_cap_exporting_mseq = 0;
			ci->i_cap_exporting_mds = -1;
		}

		/* add to session cap list */
		cap->session = session;
		spin_lock(&session->s_cap_lock);
		list_add_tail(&cap->session_caps, &session->s_caps);
		session->s_nr_caps++;
		spin_unlock(&session->s_cap_lock);
S
Sage Weil 已提交
571 572
	} else if (new_cap)
		ceph_put_cap(mdsc, new_cap);
S
Sage Weil 已提交
573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589

	if (!ci->i_snap_realm) {
		/*
		 * add this inode to the appropriate snap realm
		 */
		struct ceph_snap_realm *realm = ceph_lookup_snap_realm(mdsc,
							       realmino);
		if (realm) {
			ceph_get_snap_realm(mdsc, realm);
			spin_lock(&realm->inodes_with_caps_lock);
			ci->i_snap_realm = realm;
			list_add(&ci->i_snap_realm_item,
				 &realm->inodes_with_caps);
			spin_unlock(&realm->inodes_with_caps_lock);
		} else {
			pr_err("ceph_add_cap: couldn't find snap realm %llx\n",
			       realmino);
S
Sage Weil 已提交
590
			WARN_ON(!realm);
S
Sage Weil 已提交
591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624
		}
	}

	__check_cap_issue(ci, cap, issued);

	/*
	 * If we are issued caps we don't want, or the mds' wanted
	 * value appears to be off, queue a check so we'll release
	 * later and/or update the mds wanted value.
	 */
	actual_wanted = __ceph_caps_wanted(ci);
	if ((wanted & ~actual_wanted) ||
	    (issued & ~actual_wanted & CEPH_CAP_ANY_WR)) {
		dout(" issued %s, mds wanted %s, actual %s, queueing\n",
		     ceph_cap_string(issued), ceph_cap_string(wanted),
		     ceph_cap_string(actual_wanted));
		__cap_delay_requeue(mdsc, ci);
	}

	if (flags & CEPH_CAP_FLAG_AUTH)
		ci->i_auth_cap = cap;
	else if (ci->i_auth_cap == cap)
		ci->i_auth_cap = NULL;

	dout("add_cap inode %p (%llx.%llx) cap %p %s now %s seq %d mds%d\n",
	     inode, ceph_vinop(inode), cap, ceph_cap_string(issued),
	     ceph_cap_string(issued|cap->issued), seq, mds);
	cap->cap_id = cap_id;
	cap->issued = issued;
	cap->implemented |= issued;
	cap->mds_wanted |= wanted;
	cap->seq = seq;
	cap->issue_seq = seq;
	cap->mseq = mseq;
625
	cap->cap_gen = session->s_cap_gen;
S
Sage Weil 已提交
626 627 628

	if (fmode >= 0)
		__ceph_get_fmode(ci, fmode);
629
	spin_unlock(&ci->i_ceph_lock);
630
	wake_up_all(&ci->i_cap_wq);
S
Sage Weil 已提交
631 632 633 634 635 636 637 638 639 640 641
	return 0;
}

/*
 * Return true if cap has not timed out and belongs to the current
 * generation of the MDS session (i.e. has not gone 'stale' due to
 * us losing touch with the mds).
 */
static int __cap_is_valid(struct ceph_cap *cap)
{
	unsigned long ttl;
S
Sage Weil 已提交
642
	u32 gen;
S
Sage Weil 已提交
643

644
	spin_lock(&cap->session->s_gen_ttl_lock);
S
Sage Weil 已提交
645 646
	gen = cap->session->s_cap_gen;
	ttl = cap->session->s_cap_ttl;
647
	spin_unlock(&cap->session->s_gen_ttl_lock);
S
Sage Weil 已提交
648

649
	if (cap->cap_gen < gen || time_after_eq(jiffies, ttl)) {
S
Sage Weil 已提交
650 651
		dout("__cap_is_valid %p cap %p issued %s "
		     "but STALE (gen %u vs %u)\n", &cap->ci->vfs_inode,
652
		     cap, ceph_cap_string(cap->issued), cap->cap_gen, gen);
S
Sage Weil 已提交
653 654 655 656 657 658 659 660 661 662 663 664 665
		return 0;
	}

	return 1;
}

/*
 * Return set of valid cap bits issued to us.  Note that caps time
 * out, and may be invalidated in bulk if the client session times out
 * and session->s_cap_gen is bumped.
 */
int __ceph_caps_issued(struct ceph_inode_info *ci, int *implemented)
{
666
	int have = ci->i_snap_caps | ci->i_cap_exporting_issued;
S
Sage Weil 已提交
667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713
	struct ceph_cap *cap;
	struct rb_node *p;

	if (implemented)
		*implemented = 0;
	for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
		cap = rb_entry(p, struct ceph_cap, ci_node);
		if (!__cap_is_valid(cap))
			continue;
		dout("__ceph_caps_issued %p cap %p issued %s\n",
		     &ci->vfs_inode, cap, ceph_cap_string(cap->issued));
		have |= cap->issued;
		if (implemented)
			*implemented |= cap->implemented;
	}
	return have;
}

/*
 * Get cap bits issued by caps other than @ocap
 */
int __ceph_caps_issued_other(struct ceph_inode_info *ci, struct ceph_cap *ocap)
{
	int have = ci->i_snap_caps;
	struct ceph_cap *cap;
	struct rb_node *p;

	for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
		cap = rb_entry(p, struct ceph_cap, ci_node);
		if (cap == ocap)
			continue;
		if (!__cap_is_valid(cap))
			continue;
		have |= cap->issued;
	}
	return have;
}

/*
 * Move a cap to the end of the LRU (oldest caps at list head, newest
 * at list tail).
 */
static void __touch_cap(struct ceph_cap *cap)
{
	struct ceph_mds_session *s = cap->session;

	spin_lock(&s->s_cap_lock);
S
Sage Weil 已提交
714
	if (s->s_cap_iterator == NULL) {
715 716 717 718 719 720 721
		dout("__touch_cap %p cap %p mds%d\n", &cap->ci->vfs_inode, cap,
		     s->s_mds);
		list_move_tail(&cap->session_caps, &s->s_caps);
	} else {
		dout("__touch_cap %p cap %p mds%d NOP, iterating over caps\n",
		     &cap->ci->vfs_inode, cap, s->s_mds);
	}
S
Sage Weil 已提交
722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767
	spin_unlock(&s->s_cap_lock);
}

/*
 * Check if we hold the given mask.  If so, move the cap(s) to the
 * front of their respective LRUs.  (This is the preferred way for
 * callers to check for caps they want.)
 */
int __ceph_caps_issued_mask(struct ceph_inode_info *ci, int mask, int touch)
{
	struct ceph_cap *cap;
	struct rb_node *p;
	int have = ci->i_snap_caps;

	if ((have & mask) == mask) {
		dout("__ceph_caps_issued_mask %p snap issued %s"
		     " (mask %s)\n", &ci->vfs_inode,
		     ceph_cap_string(have),
		     ceph_cap_string(mask));
		return 1;
	}

	for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
		cap = rb_entry(p, struct ceph_cap, ci_node);
		if (!__cap_is_valid(cap))
			continue;
		if ((cap->issued & mask) == mask) {
			dout("__ceph_caps_issued_mask %p cap %p issued %s"
			     " (mask %s)\n", &ci->vfs_inode, cap,
			     ceph_cap_string(cap->issued),
			     ceph_cap_string(mask));
			if (touch)
				__touch_cap(cap);
			return 1;
		}

		/* does a combination of caps satisfy mask? */
		have |= cap->issued;
		if ((have & mask) == mask) {
			dout("__ceph_caps_issued_mask %p combo issued %s"
			     " (mask %s)\n", &ci->vfs_inode,
			     ceph_cap_string(cap->issued),
			     ceph_cap_string(mask));
			if (touch) {
				struct rb_node *q;

L
Lucas De Marchi 已提交
768
				/* touch this + preceding caps */
S
Sage Weil 已提交
769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795
				__touch_cap(cap);
				for (q = rb_first(&ci->i_caps); q != p;
				     q = rb_next(q)) {
					cap = rb_entry(q, struct ceph_cap,
						       ci_node);
					if (!__cap_is_valid(cap))
						continue;
					__touch_cap(cap);
				}
			}
			return 1;
		}
	}

	return 0;
}

/*
 * Return true if mask caps are currently being revoked by an MDS.
 */
int ceph_caps_revoking(struct ceph_inode_info *ci, int mask)
{
	struct inode *inode = &ci->vfs_inode;
	struct ceph_cap *cap;
	struct rb_node *p;
	int ret = 0;

796
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
797 798 799 800 801 802 803 804
	for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
		cap = rb_entry(p, struct ceph_cap, ci_node);
		if (__cap_is_valid(cap) &&
		    (cap->implemented & ~cap->issued & mask)) {
			ret = 1;
			break;
		}
	}
805
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
806 807 808 809 810 811 812 813 814 815 816 817
	dout("ceph_caps_revoking %p %s = %d\n", inode,
	     ceph_cap_string(mask), ret);
	return ret;
}

int __ceph_caps_used(struct ceph_inode_info *ci)
{
	int used = 0;
	if (ci->i_pin_ref)
		used |= CEPH_CAP_PIN;
	if (ci->i_rd_ref)
		used |= CEPH_CAP_FILE_RD;
818
	if (ci->i_rdcache_ref || ci->vfs_inode.i_data.nrpages)
S
Sage Weil 已提交
819 820 821
		used |= CEPH_CAP_FILE_CACHE;
	if (ci->i_wr_ref)
		used |= CEPH_CAP_FILE_WR;
822
	if (ci->i_wb_ref || ci->i_wrbuffer_ref)
S
Sage Weil 已提交
823 824 825 826 827 828 829 830 831 832 833
		used |= CEPH_CAP_FILE_BUFFER;
	return used;
}

/*
 * wanted, by virtue of open file modes
 */
int __ceph_caps_file_wanted(struct ceph_inode_info *ci)
{
	int want = 0;
	int mode;
834
	for (mode = 0; mode < CEPH_FILE_MODE_NUM; mode++)
S
Sage Weil 已提交
835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858
		if (ci->i_nr_by_mode[mode])
			want |= ceph_caps_for_mode(mode);
	return want;
}

/*
 * Return caps we have registered with the MDS(s) as 'wanted'.
 */
int __ceph_caps_mds_wanted(struct ceph_inode_info *ci)
{
	struct ceph_cap *cap;
	struct rb_node *p;
	int mds_wanted = 0;

	for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
		cap = rb_entry(p, struct ceph_cap, ci_node);
		if (!__cap_is_valid(cap))
			continue;
		mds_wanted |= cap->mds_wanted;
	}
	return mds_wanted;
}

/*
859
 * called under i_ceph_lock
S
Sage Weil 已提交
860 861 862 863 864 865 866
 */
static int __ceph_is_any_caps(struct ceph_inode_info *ci)
{
	return !RB_EMPTY_ROOT(&ci->i_caps) || ci->i_cap_exporting_mds >= 0;
}

/*
S
Sage Weil 已提交
867 868
 * Remove a cap.  Take steps to deal with a racing iterate_session_caps.
 *
869
 * caller should hold i_ceph_lock.
870
 * caller will not hold session s_mutex if called from destroy_inode.
S
Sage Weil 已提交
871
 */
S
Sage Weil 已提交
872
void __ceph_remove_cap(struct ceph_cap *cap)
S
Sage Weil 已提交
873 874 875
{
	struct ceph_mds_session *session = cap->session;
	struct ceph_inode_info *ci = cap->ci;
876
	struct ceph_mds_client *mdsc =
877
		ceph_sb_to_client(ci->vfs_inode.i_sb)->mdsc;
S
Sage Weil 已提交
878
	int removed = 0;
S
Sage Weil 已提交
879 880 881

	dout("__ceph_remove_cap %p from %p\n", cap, &ci->vfs_inode);

S
Sage Weil 已提交
882 883 884 885 886 887 888 889 890 891
	/* remove from session list */
	spin_lock(&session->s_cap_lock);
	if (session->s_cap_iterator == cap) {
		/* not yet, we are iterating over this very cap */
		dout("__ceph_remove_cap  delaying %p removal from session %p\n",
		     cap, cap->session);
	} else {
		list_del_init(&cap->session_caps);
		session->s_nr_caps--;
		cap->session = NULL;
S
Sage Weil 已提交
892
		removed = 1;
S
Sage Weil 已提交
893
	}
S
Sage Weil 已提交
894 895
	/* protect backpointer with s_cap_lock: see iterate_session_caps */
	cap->ci = NULL;
S
Sage Weil 已提交
896 897
	spin_unlock(&session->s_cap_lock);

S
Sage Weil 已提交
898 899 900 901 902 903
	/* remove from inode list */
	rb_erase(&cap->ci_node, &ci->i_caps);
	if (ci->i_auth_cap == cap)
		ci->i_auth_cap = NULL;

	if (removed)
904
		ceph_put_cap(mdsc, cap);
S
Sage Weil 已提交
905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930

	if (!__ceph_is_any_caps(ci) && ci->i_snap_realm) {
		struct ceph_snap_realm *realm = ci->i_snap_realm;
		spin_lock(&realm->inodes_with_caps_lock);
		list_del_init(&ci->i_snap_realm_item);
		ci->i_snap_realm_counter++;
		ci->i_snap_realm = NULL;
		spin_unlock(&realm->inodes_with_caps_lock);
		ceph_put_snap_realm(mdsc, realm);
	}
	if (!__ceph_is_any_real_caps(ci))
		__cap_delay_cancel(mdsc, ci);
}

/*
 * Build and send a cap message to the given MDS.
 *
 * Caller should be holding s_mutex.
 */
static int send_cap_msg(struct ceph_mds_session *session,
			u64 ino, u64 cid, int op,
			int caps, int wanted, int dirty,
			u32 seq, u64 flush_tid, u32 issue_seq, u32 mseq,
			u64 size, u64 max_size,
			struct timespec *mtime, struct timespec *atime,
			u64 time_warp_seq,
A
Al Viro 已提交
931
			uid_t uid, gid_t gid, umode_t mode,
S
Sage Weil 已提交
932 933 934 935 936 937 938 939 940 941 942 943 944 945 946
			u64 xattr_version,
			struct ceph_buffer *xattrs_buf,
			u64 follows)
{
	struct ceph_mds_caps *fc;
	struct ceph_msg *msg;

	dout("send_cap_msg %s %llx %llx caps %s wanted %s dirty %s"
	     " seq %u/%u mseq %u follows %lld size %llu/%llu"
	     " xattr_ver %llu xattr_len %d\n", ceph_cap_op_name(op),
	     cid, ino, ceph_cap_string(caps), ceph_cap_string(wanted),
	     ceph_cap_string(dirty),
	     seq, issue_seq, mseq, follows, size, max_size,
	     xattr_version, xattrs_buf ? (int)xattrs_buf->vec.iov_len : 0);

947
	msg = ceph_msg_new(CEPH_MSG_CLIENT_CAPS, sizeof(*fc), GFP_NOFS, false);
948 949
	if (!msg)
		return -ENOMEM;
S
Sage Weil 已提交
950

951
	msg->hdr.tid = cpu_to_le64(flush_tid);
S
Sage Weil 已提交
952

953
	fc = msg->front.iov_base;
S
Sage Weil 已提交
954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989
	memset(fc, 0, sizeof(*fc));

	fc->cap_id = cpu_to_le64(cid);
	fc->op = cpu_to_le32(op);
	fc->seq = cpu_to_le32(seq);
	fc->issue_seq = cpu_to_le32(issue_seq);
	fc->migrate_seq = cpu_to_le32(mseq);
	fc->caps = cpu_to_le32(caps);
	fc->wanted = cpu_to_le32(wanted);
	fc->dirty = cpu_to_le32(dirty);
	fc->ino = cpu_to_le64(ino);
	fc->snap_follows = cpu_to_le64(follows);

	fc->size = cpu_to_le64(size);
	fc->max_size = cpu_to_le64(max_size);
	if (mtime)
		ceph_encode_timespec(&fc->mtime, mtime);
	if (atime)
		ceph_encode_timespec(&fc->atime, atime);
	fc->time_warp_seq = cpu_to_le32(time_warp_seq);

	fc->uid = cpu_to_le32(uid);
	fc->gid = cpu_to_le32(gid);
	fc->mode = cpu_to_le32(mode);

	fc->xattr_version = cpu_to_le64(xattr_version);
	if (xattrs_buf) {
		msg->middle = ceph_buffer_get(xattrs_buf);
		fc->xattr_len = cpu_to_le32(xattrs_buf->vec.iov_len);
		msg->hdr.middle_len = cpu_to_le32(xattrs_buf->vec.iov_len);
	}

	ceph_con_send(&session->s_con, msg);
	return 0;
}

990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007
static void __queue_cap_release(struct ceph_mds_session *session,
				u64 ino, u64 cap_id, u32 migrate_seq,
				u32 issue_seq)
{
	struct ceph_msg *msg;
	struct ceph_mds_cap_release *head;
	struct ceph_mds_cap_item *item;

	spin_lock(&session->s_cap_lock);
	BUG_ON(!session->s_num_cap_releases);
	msg = list_first_entry(&session->s_cap_releases,
			       struct ceph_msg, list_head);

	dout(" adding %llx release to mds%d msg %p (%d left)\n",
	     ino, session->s_mds, msg, session->s_num_cap_releases);

	BUG_ON(msg->front.iov_len + sizeof(*item) > PAGE_CACHE_SIZE);
	head = msg->front.iov_base;
W
Wei Yongjun 已提交
1008
	le32_add_cpu(&head->num, 1);
1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029
	item = msg->front.iov_base + msg->front.iov_len;
	item->ino = cpu_to_le64(ino);
	item->cap_id = cpu_to_le64(cap_id);
	item->migrate_seq = cpu_to_le32(migrate_seq);
	item->seq = cpu_to_le32(issue_seq);

	session->s_num_cap_releases--;

	msg->front.iov_len += sizeof(*item);
	if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) {
		dout(" release msg %p full\n", msg);
		list_move_tail(&msg->list_head, &session->s_cap_releases_done);
	} else {
		dout(" release msg %p at %d/%d (%d)\n", msg,
		     (int)le32_to_cpu(head->num),
		     (int)CEPH_CAPS_PER_RELEASE,
		     (int)msg->front.iov_len);
	}
	spin_unlock(&session->s_cap_lock);
}

S
Sage Weil 已提交
1030
/*
1031
 * Queue cap releases when an inode is dropped from our cache.  Since
1032
 * inode is about to be destroyed, there is no need for i_ceph_lock.
S
Sage Weil 已提交
1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043
 */
void ceph_queue_caps_release(struct inode *inode)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	struct rb_node *p;

	p = rb_first(&ci->i_caps);
	while (p) {
		struct ceph_cap *cap = rb_entry(p, struct ceph_cap, ci_node);
		struct ceph_mds_session *session = cap->session;

1044 1045
		__queue_cap_release(session, ceph_ino(inode), cap->cap_id,
				    cap->mseq, cap->issue_seq);
S
Sage Weil 已提交
1046
		p = rb_next(p);
S
Sage Weil 已提交
1047
		__ceph_remove_cap(cap);
S
Sage Weil 已提交
1048 1049 1050 1051 1052
	}
}

/*
 * Send a cap msg on the given inode.  Update our caps state, then
1053
 * drop i_ceph_lock and send the message.
S
Sage Weil 已提交
1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064
 *
 * Make note of max_size reported/requested from mds, revoked caps
 * that have now been implemented.
 *
 * Make half-hearted attempt ot to invalidate page cache if we are
 * dropping RDCACHE.  Note that this will leave behind locked pages
 * that we'll then need to deal with elsewhere.
 *
 * Return non-zero if delayed release, or we experienced an error
 * such that the caller should requeue + retry later.
 *
1065
 * called with i_ceph_lock, then drops it.
S
Sage Weil 已提交
1066 1067 1068 1069 1070
 * caller should hold snap_rwsem (read), s_mutex.
 */
static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap,
		      int op, int used, int want, int retain, int flushing,
		      unsigned *pflush_tid)
1071
	__releases(cap->ci->i_ceph_lock)
S
Sage Weil 已提交
1072 1073 1074 1075
{
	struct ceph_inode_info *ci = cap->ci;
	struct inode *inode = &ci->vfs_inode;
	u64 cap_id = cap->cap_id;
1076
	int held, revoking, dropping, keep;
S
Sage Weil 已提交
1077 1078 1079 1080
	u64 seq, issue_seq, mseq, time_warp_seq, follows;
	u64 size, max_size;
	struct timespec mtime, atime;
	int wake = 0;
A
Al Viro 已提交
1081
	umode_t mode;
S
Sage Weil 已提交
1082 1083 1084 1085
	uid_t uid;
	gid_t gid;
	struct ceph_mds_session *session;
	u64 xattr_version = 0;
S
Sage Weil 已提交
1086
	struct ceph_buffer *xattr_blob = NULL;
S
Sage Weil 已提交
1087 1088 1089 1090 1091
	int delayed = 0;
	u64 flush_tid = 0;
	int i;
	int ret;

1092 1093 1094 1095 1096
	held = cap->issued | cap->implemented;
	revoking = cap->implemented & ~cap->issued;
	retain &= ~revoking;
	dropping = cap->issued & ~retain;

S
Sage Weil 已提交
1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146
	dout("__send_cap %p cap %p session %p %s -> %s (revoking %s)\n",
	     inode, cap, cap->session,
	     ceph_cap_string(held), ceph_cap_string(held & retain),
	     ceph_cap_string(revoking));
	BUG_ON((retain & CEPH_CAP_PIN) == 0);

	session = cap->session;

	/* don't release wanted unless we've waited a bit. */
	if ((ci->i_ceph_flags & CEPH_I_NODELAY) == 0 &&
	    time_before(jiffies, ci->i_hold_caps_min)) {
		dout(" delaying issued %s -> %s, wanted %s -> %s on send\n",
		     ceph_cap_string(cap->issued),
		     ceph_cap_string(cap->issued & retain),
		     ceph_cap_string(cap->mds_wanted),
		     ceph_cap_string(want));
		want |= cap->mds_wanted;
		retain |= cap->issued;
		delayed = 1;
	}
	ci->i_ceph_flags &= ~(CEPH_I_NODELAY | CEPH_I_FLUSH);

	cap->issued &= retain;  /* drop bits we don't want */
	if (cap->implemented & ~cap->issued) {
		/*
		 * Wake up any waiters on wanted -> needed transition.
		 * This is due to the weird transition from buffered
		 * to sync IO... we need to flush dirty pages _before_
		 * allowing sync writes to avoid reordering.
		 */
		wake = 1;
	}
	cap->implemented &= cap->issued | used;
	cap->mds_wanted = want;

	if (flushing) {
		/*
		 * assign a tid for flush operations so we can avoid
		 * flush1 -> dirty1 -> flush2 -> flushack1 -> mark
		 * clean type races.  track latest tid for every bit
		 * so we can handle flush AxFw, flush Fw, and have the
		 * first ack clean Ax.
		 */
		flush_tid = ++ci->i_cap_flush_last_tid;
		if (pflush_tid)
			*pflush_tid = flush_tid;
		dout(" cap_flush_tid %d\n", (int)flush_tid);
		for (i = 0; i < CEPH_CAP_BITS; i++)
			if (flushing & (1 << i))
				ci->i_cap_flush_tid[i] = flush_tid;
1147 1148 1149 1150

		follows = ci->i_head_snapc->seq;
	} else {
		follows = 0;
S
Sage Weil 已提交
1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167
	}

	keep = cap->implemented;
	seq = cap->seq;
	issue_seq = cap->issue_seq;
	mseq = cap->mseq;
	size = inode->i_size;
	ci->i_reported_size = size;
	max_size = ci->i_wanted_max_size;
	ci->i_requested_max_size = max_size;
	mtime = inode->i_mtime;
	atime = inode->i_atime;
	time_warp_seq = ci->i_time_warp_seq;
	uid = inode->i_uid;
	gid = inode->i_gid;
	mode = inode->i_mode;

S
Sage Weil 已提交
1168
	if (flushing & CEPH_CAP_XATTR_EXCL) {
S
Sage Weil 已提交
1169
		__ceph_build_xattrs_blob(ci);
S
Sage Weil 已提交
1170 1171
		xattr_blob = ci->i_xattrs.blob;
		xattr_version = ci->i_xattrs.version;
S
Sage Weil 已提交
1172 1173
	}

1174
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1175 1176 1177 1178

	ret = send_cap_msg(session, ceph_vino(inode).ino, cap_id,
		op, keep, want, flushing, seq, flush_tid, issue_seq, mseq,
		size, max_size, &mtime, &atime, time_warp_seq,
S
Sage Weil 已提交
1179
		uid, gid, mode, xattr_version, xattr_blob,
S
Sage Weil 已提交
1180 1181 1182 1183 1184 1185 1186
		follows);
	if (ret < 0) {
		dout("error sending cap msg, must requeue %p\n", inode);
		delayed = 1;
	}

	if (wake)
1187
		wake_up_all(&ci->i_cap_wq);
S
Sage Weil 已提交
1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198

	return delayed;
}

/*
 * When a snapshot is taken, clients accumulate dirty metadata on
 * inodes with capabilities in ceph_cap_snaps to describe the file
 * state at the time the snapshot was taken.  This must be flushed
 * asynchronously back to the MDS once sync writes complete and dirty
 * data is written out.
 *
1199 1200 1201
 * Unless @again is true, skip cap_snaps that were already sent to
 * the MDS (i.e., during this session).
 *
1202
 * Called under i_ceph_lock.  Takes s_mutex as needed.
S
Sage Weil 已提交
1203 1204
 */
void __ceph_flush_snaps(struct ceph_inode_info *ci,
1205 1206
			struct ceph_mds_session **psession,
			int again)
1207 1208
		__releases(ci->i_ceph_lock)
		__acquires(ci->i_ceph_lock)
S
Sage Weil 已提交
1209 1210 1211 1212 1213
{
	struct inode *inode = &ci->vfs_inode;
	int mds;
	struct ceph_cap_snap *capsnap;
	u32 mseq;
1214
	struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc;
S
Sage Weil 已提交
1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234
	struct ceph_mds_session *session = NULL; /* if session != NULL, we hold
						    session->s_mutex */
	u64 next_follows = 0;  /* keep track of how far we've gotten through the
			     i_cap_snaps list, and skip these entries next time
			     around to avoid an infinite loop */

	if (psession)
		session = *psession;

	dout("__flush_snaps %p\n", inode);
retry:
	list_for_each_entry(capsnap, &ci->i_cap_snaps, ci_item) {
		/* avoid an infiniute loop after retry */
		if (capsnap->follows < next_follows)
			continue;
		/*
		 * we need to wait for sync writes to complete and for dirty
		 * pages to be written out.
		 */
		if (capsnap->dirty_pages || capsnap->writing)
1235
			break;
S
Sage Weil 已提交
1236

1237 1238 1239 1240 1241 1242
		/*
		 * if cap writeback already occurred, we should have dropped
		 * the capsnap in ceph_put_wrbuffer_cap_refs.
		 */
		BUG_ON(capsnap->dirty == 0);

S
Sage Weil 已提交
1243
		/* pick mds, take s_mutex */
1244 1245 1246 1247
		if (ci->i_auth_cap == NULL) {
			dout("no auth cap (migrating?), doing nothing\n");
			goto out;
		}
1248 1249 1250 1251 1252 1253 1254

		/* only flush each capsnap once */
		if (!again && !list_empty(&capsnap->flushing_item)) {
			dout("already flushed %p, skipping\n", capsnap);
			continue;
		}

1255 1256 1257
		mds = ci->i_auth_cap->session->s_mds;
		mseq = ci->i_auth_cap->mseq;

S
Sage Weil 已提交
1258 1259 1260 1261 1262 1263 1264
		if (session && session->s_mds != mds) {
			dout("oops, wrong session %p mutex\n", session);
			mutex_unlock(&session->s_mutex);
			ceph_put_mds_session(session);
			session = NULL;
		}
		if (!session) {
1265
			spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1266 1267 1268 1269 1270 1271 1272 1273 1274 1275
			mutex_lock(&mdsc->mutex);
			session = __ceph_lookup_mds_session(mdsc, mds);
			mutex_unlock(&mdsc->mutex);
			if (session) {
				dout("inverting session/ino locks on %p\n",
				     session);
				mutex_lock(&session->s_mutex);
			}
			/*
			 * if session == NULL, we raced against a cap
1276 1277
			 * deletion or migration.  retry, and we'll
			 * get a better @mds value next time.
S
Sage Weil 已提交
1278
			 */
1279
			spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1280 1281 1282 1283 1284 1285 1286 1287 1288
			goto retry;
		}

		capsnap->flush_tid = ++ci->i_cap_flush_last_tid;
		atomic_inc(&capsnap->nref);
		if (!list_empty(&capsnap->flushing_item))
			list_del_init(&capsnap->flushing_item);
		list_add_tail(&capsnap->flushing_item,
			      &session->s_cap_snaps_flushing);
1289
		spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1290

1291 1292
		dout("flush_snaps %p cap_snap %p follows %lld tid %llu\n",
		     inode, capsnap, capsnap->follows, capsnap->flush_tid);
S
Sage Weil 已提交
1293 1294 1295 1296 1297 1298 1299
		send_cap_msg(session, ceph_vino(inode).ino, 0,
			     CEPH_CAP_OP_FLUSHSNAP, capsnap->issued, 0,
			     capsnap->dirty, 0, capsnap->flush_tid, 0, mseq,
			     capsnap->size, 0,
			     &capsnap->mtime, &capsnap->atime,
			     capsnap->time_warp_seq,
			     capsnap->uid, capsnap->gid, capsnap->mode,
1300
			     capsnap->xattr_version, capsnap->xattr_blob,
S
Sage Weil 已提交
1301 1302 1303 1304 1305
			     capsnap->follows);

		next_follows = capsnap->follows + 1;
		ceph_put_cap_snap(capsnap);

1306
		spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1307 1308 1309 1310 1311 1312 1313 1314
		goto retry;
	}

	/* we flushed them all; remove this inode from the queue */
	spin_lock(&mdsc->snap_flush_lock);
	list_del_init(&ci->i_snap_flush_item);
	spin_unlock(&mdsc->snap_flush_lock);

1315
out:
S
Sage Weil 已提交
1316 1317 1318 1319 1320 1321 1322 1323 1324 1325
	if (psession)
		*psession = session;
	else if (session) {
		mutex_unlock(&session->s_mutex);
		ceph_put_mds_session(session);
	}
}

static void ceph_flush_snaps(struct ceph_inode_info *ci)
{
1326
	spin_lock(&ci->i_ceph_lock);
1327
	__ceph_flush_snaps(ci, NULL, 0);
1328
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1329 1330
}

S
Sage Weil 已提交
1331
/*
1332 1333 1334
 * Mark caps dirty.  If inode is newly dirty, return the dirty flags.
 * Caller is then responsible for calling __mark_inode_dirty with the
 * returned flags value.
S
Sage Weil 已提交
1335
 */
1336
int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask)
S
Sage Weil 已提交
1337
{
1338
	struct ceph_mds_client *mdsc =
1339
		ceph_sb_to_client(ci->vfs_inode.i_sb)->mdsc;
S
Sage Weil 已提交
1340 1341 1342 1343 1344 1345 1346 1347 1348
	struct inode *inode = &ci->vfs_inode;
	int was = ci->i_dirty_caps;
	int dirty = 0;

	dout("__mark_dirty_caps %p %s dirty %s -> %s\n", &ci->vfs_inode,
	     ceph_cap_string(mask), ceph_cap_string(was),
	     ceph_cap_string(was | mask));
	ci->i_dirty_caps |= mask;
	if (was == 0) {
1349 1350 1351 1352 1353
		if (!ci->i_head_snapc)
			ci->i_head_snapc = ceph_get_snap_context(
				ci->i_snap_realm->cached_context);
		dout(" inode %p now dirty snapc %p\n", &ci->vfs_inode,
			ci->i_head_snapc);
S
Sage Weil 已提交
1354 1355 1356 1357 1358
		BUG_ON(!list_empty(&ci->i_dirty_item));
		spin_lock(&mdsc->cap_dirty_lock);
		list_add(&ci->i_dirty_item, &mdsc->cap_dirty);
		spin_unlock(&mdsc->cap_dirty_lock);
		if (ci->i_flushing_caps == 0) {
1359
			ihold(inode);
S
Sage Weil 已提交
1360 1361 1362 1363 1364 1365 1366 1367
			dirty |= I_DIRTY_SYNC;
		}
	}
	BUG_ON(list_empty(&ci->i_dirty_item));
	if (((was | ci->i_flushing_caps) & CEPH_CAP_FILE_BUFFER) &&
	    (mask & CEPH_CAP_FILE_BUFFER))
		dirty |= I_DIRTY_DATASYNC;
	__cap_delay_requeue(mdsc, ci);
1368
	return dirty;
S
Sage Weil 已提交
1369 1370
}

S
Sage Weil 已提交
1371 1372 1373
/*
 * Add dirty inode to the flushing list.  Assigned a seq number so we
 * can wait for caps to flush without starving.
1374
 *
1375
 * Called under i_ceph_lock.
S
Sage Weil 已提交
1376
 */
1377
static int __mark_caps_flushing(struct inode *inode,
S
Sage Weil 已提交
1378 1379
				 struct ceph_mds_session *session)
{
1380
	struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc;
S
Sage Weil 已提交
1381
	struct ceph_inode_info *ci = ceph_inode(inode);
1382
	int flushing;
S
Sage Weil 已提交
1383

1384
	BUG_ON(ci->i_dirty_caps == 0);
S
Sage Weil 已提交
1385
	BUG_ON(list_empty(&ci->i_dirty_item));
1386 1387 1388 1389 1390 1391 1392 1393

	flushing = ci->i_dirty_caps;
	dout("__mark_caps_flushing flushing %s, flushing_caps %s -> %s\n",
	     ceph_cap_string(flushing),
	     ceph_cap_string(ci->i_flushing_caps),
	     ceph_cap_string(ci->i_flushing_caps | flushing));
	ci->i_flushing_caps |= flushing;
	ci->i_dirty_caps = 0;
1394
	dout(" inode %p now !dirty\n", inode);
1395

S
Sage Weil 已提交
1396
	spin_lock(&mdsc->cap_dirty_lock);
1397 1398 1399
	list_del_init(&ci->i_dirty_item);

	ci->i_cap_flush_seq = ++mdsc->cap_flush_seq;
S
Sage Weil 已提交
1400 1401 1402
	if (list_empty(&ci->i_flushing_item)) {
		list_add_tail(&ci->i_flushing_item, &session->s_cap_flushing);
		mdsc->num_cap_flushing++;
1403 1404 1405 1406 1407
		dout(" inode %p now flushing seq %lld\n", inode,
		     ci->i_cap_flush_seq);
	} else {
		list_move_tail(&ci->i_flushing_item, &session->s_cap_flushing);
		dout(" inode %p now flushing (more) seq %lld\n", inode,
S
Sage Weil 已提交
1408 1409 1410
		     ci->i_cap_flush_seq);
	}
	spin_unlock(&mdsc->cap_dirty_lock);
1411 1412

	return flushing;
S
Sage Weil 已提交
1413 1414
}

1415 1416 1417 1418 1419 1420 1421 1422
/*
 * try to invalidate mapping pages without blocking.
 */
static int try_nonblocking_invalidate(struct inode *inode)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	u32 invalidating_gen = ci->i_rdcache_gen;

1423
	spin_unlock(&ci->i_ceph_lock);
1424
	invalidate_mapping_pages(&inode->i_data, 0, -1);
1425
	spin_lock(&ci->i_ceph_lock);
1426

1427
	if (inode->i_data.nrpages == 0 &&
1428 1429 1430
	    invalidating_gen == ci->i_rdcache_gen) {
		/* success. */
		dout("try_nonblocking_invalidate %p success\n", inode);
1431 1432
		/* save any racing async invalidate some trouble */
		ci->i_rdcache_revoking = ci->i_rdcache_gen - 1;
1433 1434 1435 1436 1437 1438
		return 0;
	}
	dout("try_nonblocking_invalidate %p failed\n", inode);
	return -1;
}

S
Sage Weil 已提交
1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452
/*
 * Swiss army knife function to examine currently used and wanted
 * versus held caps.  Release, flush, ack revoked caps to mds as
 * appropriate.
 *
 *  CHECK_CAPS_NODELAY - caller is delayed work and we should not delay
 *    cap release further.
 *  CHECK_CAPS_AUTHONLY - we should only check the auth cap
 *  CHECK_CAPS_FLUSH - we should flush any dirty caps immediately, without
 *    further delay.
 */
void ceph_check_caps(struct ceph_inode_info *ci, int flags,
		     struct ceph_mds_session *session)
{
1453 1454
	struct ceph_fs_client *fsc = ceph_inode_to_client(&ci->vfs_inode);
	struct ceph_mds_client *mdsc = fsc->mdsc;
S
Sage Weil 已提交
1455 1456 1457 1458
	struct inode *inode = &ci->vfs_inode;
	struct ceph_cap *cap;
	int file_wanted, used;
	int took_snap_rwsem = 0;             /* true if mdsc->snap_rwsem held */
S
Sage Weil 已提交
1459
	int issued, implemented, want, retain, revoking, flushing = 0;
S
Sage Weil 已提交
1460 1461 1462 1463 1464
	int mds = -1;   /* keep track of how far we've gone through i_caps list
			   to avoid an infinite loop on retry */
	struct rb_node *p;
	int tried_invalidate = 0;
	int delayed = 0, sent = 0, force_requeue = 0, num;
S
Sage Weil 已提交
1465
	int queue_invalidate = 0;
S
Sage Weil 已提交
1466 1467 1468 1469 1470 1471
	int is_delayed = flags & CHECK_CAPS_NODELAY;

	/* if we are unmounting, flush any unused caps immediately. */
	if (mdsc->stopping)
		is_delayed = 1;

1472
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1473 1474 1475 1476 1477 1478

	if (ci->i_ceph_flags & CEPH_I_FLUSH)
		flags |= CHECK_CAPS_FLUSH;

	/* flush snaps first time around only */
	if (!list_empty(&ci->i_cap_snaps))
1479
		__ceph_flush_snaps(ci, &session, 0);
S
Sage Weil 已提交
1480 1481
	goto retry_locked;
retry:
1482
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1483 1484 1485 1486
retry_locked:
	file_wanted = __ceph_caps_file_wanted(ci);
	used = __ceph_caps_used(ci);
	want = file_wanted | used;
S
Sage Weil 已提交
1487 1488
	issued = __ceph_caps_issued(ci, &implemented);
	revoking = implemented & ~issued;
S
Sage Weil 已提交
1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506

	retain = want | CEPH_CAP_PIN;
	if (!mdsc->stopping && inode->i_nlink > 0) {
		if (want) {
			retain |= CEPH_CAP_ANY;       /* be greedy */
		} else {
			retain |= CEPH_CAP_ANY_SHARED;
			/*
			 * keep RD only if we didn't have the file open RW,
			 * because then the mds would revoke it anyway to
			 * journal max_size=0.
			 */
			if (ci->i_max_size == 0)
				retain |= CEPH_CAP_ANY_RD;
		}
	}

	dout("check_caps %p file_want %s used %s dirty %s flushing %s"
S
Sage Weil 已提交
1507
	     " issued %s revoking %s retain %s %s%s%s\n", inode,
S
Sage Weil 已提交
1508 1509 1510
	     ceph_cap_string(file_wanted),
	     ceph_cap_string(used), ceph_cap_string(ci->i_dirty_caps),
	     ceph_cap_string(ci->i_flushing_caps),
S
Sage Weil 已提交
1511
	     ceph_cap_string(issued), ceph_cap_string(revoking),
S
Sage Weil 已提交
1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523
	     ceph_cap_string(retain),
	     (flags & CHECK_CAPS_AUTHONLY) ? " AUTHONLY" : "",
	     (flags & CHECK_CAPS_NODELAY) ? " NODELAY" : "",
	     (flags & CHECK_CAPS_FLUSH) ? " FLUSH" : "");

	/*
	 * If we no longer need to hold onto old our caps, and we may
	 * have cached pages, but don't want them, then try to invalidate.
	 * If we fail, it's because pages are locked.... try again later.
	 */
	if ((!is_delayed || mdsc->stopping) &&
	    ci->i_wrbuffer_ref == 0 &&               /* no dirty pages... */
1524
	    inode->i_data.nrpages &&                 /* have cached pages */
S
Sage Weil 已提交
1525
	    (file_wanted == 0 ||                     /* no open files */
1526 1527
	     (revoking & (CEPH_CAP_FILE_CACHE|
			  CEPH_CAP_FILE_LAZYIO))) && /*  or revoking cache */
S
Sage Weil 已提交
1528 1529
	    !tried_invalidate) {
		dout("check_caps trying to invalidate on %p\n", inode);
1530
		if (try_nonblocking_invalidate(inode) < 0) {
1531 1532
			if (revoking & (CEPH_CAP_FILE_CACHE|
					CEPH_CAP_FILE_LAZYIO)) {
1533 1534 1535 1536 1537 1538 1539 1540 1541 1542
				dout("check_caps queuing invalidate\n");
				queue_invalidate = 1;
				ci->i_rdcache_revoking = ci->i_rdcache_gen;
			} else {
				dout("check_caps failed to invalidate pages\n");
				/* we failed to invalidate pages.  check these
				   caps again later. */
				force_requeue = 1;
				__cap_set_timeouts(mdsc, ci);
			}
S
Sage Weil 已提交
1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560
		}
		tried_invalidate = 1;
		goto retry_locked;
	}

	num = 0;
	for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
		cap = rb_entry(p, struct ceph_cap, ci_node);
		num++;

		/* avoid looping forever */
		if (mds >= cap->mds ||
		    ((flags & CHECK_CAPS_AUTHONLY) && cap != ci->i_auth_cap))
			continue;

		/* NOTE: no side-effects allowed, until we take s_mutex */

		revoking = cap->implemented & ~cap->issued;
1561 1562 1563 1564
		dout(" mds%d cap %p issued %s implemented %s revoking %s\n",
		     cap->mds, cap, ceph_cap_string(cap->issued),
		     ceph_cap_string(cap->implemented),
		     ceph_cap_string(revoking));
S
Sage Weil 已提交
1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 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

		if (cap == ci->i_auth_cap &&
		    (cap->issued & CEPH_CAP_FILE_WR)) {
			/* request larger max_size from MDS? */
			if (ci->i_wanted_max_size > ci->i_max_size &&
			    ci->i_wanted_max_size > ci->i_requested_max_size) {
				dout("requesting new max_size\n");
				goto ack;
			}

			/* approaching file_max? */
			if ((inode->i_size << 1) >= ci->i_max_size &&
			    (ci->i_reported_size << 1) < ci->i_max_size) {
				dout("i_size approaching max_size\n");
				goto ack;
			}
		}
		/* flush anything dirty? */
		if (cap == ci->i_auth_cap && (flags & CHECK_CAPS_FLUSH) &&
		    ci->i_dirty_caps) {
			dout("flushing dirty caps\n");
			goto ack;
		}

		/* completed revocation? going down and there are no caps? */
		if (revoking && (revoking & used) == 0) {
			dout("completed revocation of %s\n",
			     ceph_cap_string(cap->implemented & ~cap->issued));
			goto ack;
		}

		/* want more caps from mds? */
		if (want & ~(cap->mds_wanted | cap->issued))
			goto ack;

		/* things we might delay */
		if ((cap->issued & ~retain) == 0 &&
		    cap->mds_wanted == want)
			continue;     /* nope, all good */

		if (is_delayed)
			goto ack;

		/* delay? */
		if ((ci->i_ceph_flags & CEPH_I_NODELAY) == 0 &&
		    time_before(jiffies, ci->i_hold_caps_max)) {
			dout(" delaying issued %s -> %s, wanted %s -> %s\n",
			     ceph_cap_string(cap->issued),
			     ceph_cap_string(cap->issued & retain),
			     ceph_cap_string(cap->mds_wanted),
			     ceph_cap_string(want));
			delayed++;
			continue;
		}

ack:
1621 1622 1623 1624 1625
		if (ci->i_ceph_flags & CEPH_I_NOFLUSH) {
			dout(" skipping %p I_NOFLUSH set\n", inode);
			continue;
		}

S
Sage Weil 已提交
1626 1627 1628 1629 1630 1631 1632 1633 1634 1635
		if (session && session != cap->session) {
			dout("oops, wrong session %p mutex\n", session);
			mutex_unlock(&session->s_mutex);
			session = NULL;
		}
		if (!session) {
			session = cap->session;
			if (mutex_trylock(&session->s_mutex) == 0) {
				dout("inverting session/ino locks on %p\n",
				     session);
1636
				spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649
				if (took_snap_rwsem) {
					up_read(&mdsc->snap_rwsem);
					took_snap_rwsem = 0;
				}
				mutex_lock(&session->s_mutex);
				goto retry;
			}
		}
		/* take snap_rwsem after session mutex */
		if (!took_snap_rwsem) {
			if (down_read_trylock(&mdsc->snap_rwsem) == 0) {
				dout("inverting snap/in locks on %p\n",
				     inode);
1650
				spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1651 1652 1653 1654 1655 1656 1657
				down_read(&mdsc->snap_rwsem);
				took_snap_rwsem = 1;
				goto retry;
			}
			took_snap_rwsem = 1;
		}

1658 1659
		if (cap == ci->i_auth_cap && ci->i_dirty_caps)
			flushing = __mark_caps_flushing(inode, session);
1660 1661
		else
			flushing = 0;
S
Sage Weil 已提交
1662 1663 1664 1665

		mds = cap->mds;  /* remember mds, so we don't repeat */
		sent++;

1666
		/* __send_cap drops i_ceph_lock */
S
Sage Weil 已提交
1667 1668
		delayed += __send_cap(mdsc, cap, CEPH_CAP_OP_UPDATE, used, want,
				      retain, flushing, NULL);
1669
		goto retry; /* retake i_ceph_lock and restart our cap scan. */
S
Sage Weil 已提交
1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682
	}

	/*
	 * Reschedule delayed caps release if we delayed anything,
	 * otherwise cancel.
	 */
	if (delayed && is_delayed)
		force_requeue = 1;   /* __send_cap delayed release; requeue */
	if (!delayed && !is_delayed)
		__cap_delay_cancel(mdsc, ci);
	else if (!is_delayed || force_requeue)
		__cap_delay_requeue(mdsc, ci);

1683
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1684

S
Sage Weil 已提交
1685
	if (queue_invalidate)
1686
		ceph_queue_invalidate(inode);
S
Sage Weil 已提交
1687

1688
	if (session)
S
Sage Weil 已提交
1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699
		mutex_unlock(&session->s_mutex);
	if (took_snap_rwsem)
		up_read(&mdsc->snap_rwsem);
}

/*
 * Try to flush dirty caps back to the auth mds.
 */
static int try_flush_caps(struct inode *inode, struct ceph_mds_session *session,
			  unsigned *flush_tid)
{
1700
	struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc;
S
Sage Weil 已提交
1701 1702 1703 1704 1705
	struct ceph_inode_info *ci = ceph_inode(inode);
	int unlock_session = session ? 0 : 1;
	int flushing = 0;

retry:
1706
	spin_lock(&ci->i_ceph_lock);
1707 1708 1709 1710
	if (ci->i_ceph_flags & CEPH_I_NOFLUSH) {
		dout("try_flush_caps skipping %p I_NOFLUSH set\n", inode);
		goto out;
	}
S
Sage Weil 已提交
1711 1712 1713 1714 1715 1716 1717
	if (ci->i_dirty_caps && ci->i_auth_cap) {
		struct ceph_cap *cap = ci->i_auth_cap;
		int used = __ceph_caps_used(ci);
		int want = __ceph_caps_wanted(ci);
		int delayed;

		if (!session) {
1718
			spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1719 1720 1721 1722 1723 1724 1725 1726
			session = cap->session;
			mutex_lock(&session->s_mutex);
			goto retry;
		}
		BUG_ON(session != cap->session);
		if (cap->session->s_state < CEPH_MDS_SESSION_OPEN)
			goto out;

1727
		flushing = __mark_caps_flushing(inode, session);
S
Sage Weil 已提交
1728

1729
		/* __send_cap drops i_ceph_lock */
S
Sage Weil 已提交
1730 1731 1732 1733 1734 1735
		delayed = __send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH, used, want,
				     cap->issued | cap->implemented, flushing,
				     flush_tid);
		if (!delayed)
			goto out_unlocked;

1736
		spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1737 1738 1739
		__cap_delay_requeue(mdsc, ci);
	}
out:
1740
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752
out_unlocked:
	if (session && unlock_session)
		mutex_unlock(&session->s_mutex);
	return flushing;
}

/*
 * Return true if we've flushed caps through the given flush_tid.
 */
static int caps_are_flushed(struct inode *inode, unsigned tid)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
1753
	int i, ret = 1;
S
Sage Weil 已提交
1754

1755
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1756 1757 1758 1759 1760 1761 1762
	for (i = 0; i < CEPH_CAP_BITS; i++)
		if ((ci->i_flushing_caps & (1 << i)) &&
		    ci->i_cap_flush_tid[i] <= tid) {
			/* still flushing this bit */
			ret = 0;
			break;
		}
1763
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
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 1804 1805 1806 1807 1808 1809 1810
	return ret;
}

/*
 * Wait on any unsafe replies for the given inode.  First wait on the
 * newest request, and make that the upper bound.  Then, if there are
 * more requests, keep waiting on the oldest as long as it is still older
 * than the original request.
 */
static void sync_write_wait(struct inode *inode)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	struct list_head *head = &ci->i_unsafe_writes;
	struct ceph_osd_request *req;
	u64 last_tid;

	spin_lock(&ci->i_unsafe_lock);
	if (list_empty(head))
		goto out;

	/* set upper bound as _last_ entry in chain */
	req = list_entry(head->prev, struct ceph_osd_request,
			 r_unsafe_item);
	last_tid = req->r_tid;

	do {
		ceph_osdc_get_request(req);
		spin_unlock(&ci->i_unsafe_lock);
		dout("sync_write_wait on tid %llu (until %llu)\n",
		     req->r_tid, last_tid);
		wait_for_completion(&req->r_safe_completion);
		spin_lock(&ci->i_unsafe_lock);
		ceph_osdc_put_request(req);

		/*
		 * from here on look at first entry in chain, since we
		 * only want to wait for anything older than last_tid
		 */
		if (list_empty(head))
			break;
		req = list_entry(head->next, struct ceph_osd_request,
				 r_unsafe_item);
	} while (req->r_tid < last_tid);
out:
	spin_unlock(&ci->i_unsafe_lock);
}

1811
int ceph_fsync(struct file *file, loff_t start, loff_t end, int datasync)
S
Sage Weil 已提交
1812
{
1813
	struct inode *inode = file->f_mapping->host;
S
Sage Weil 已提交
1814 1815 1816 1817 1818 1819 1820 1821
	struct ceph_inode_info *ci = ceph_inode(inode);
	unsigned flush_tid;
	int ret;
	int dirty;

	dout("fsync %p%s\n", inode, datasync ? " datasync" : "");
	sync_write_wait(inode);

1822
	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
S
Sage Weil 已提交
1823 1824
	if (ret < 0)
		return ret;
1825
	mutex_lock(&inode->i_mutex);
S
Sage Weil 已提交
1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841

	dirty = try_flush_caps(inode, NULL, &flush_tid);
	dout("fsync dirty caps are %s\n", ceph_cap_string(dirty));

	/*
	 * only wait on non-file metadata writeback (the mds
	 * can recover size and mtime, so we don't need to
	 * wait for that)
	 */
	if (!datasync && (dirty & ~CEPH_CAP_ANY_FILE_WR)) {
		dout("fsync waiting for flush_tid %u\n", flush_tid);
		ret = wait_event_interruptible(ci->i_cap_wq,
				       caps_are_flushed(inode, flush_tid));
	}

	dout("fsync %p%s done\n", inode, datasync ? " datasync" : "");
1842
	mutex_unlock(&inode->i_mutex);
S
Sage Weil 已提交
1843 1844 1845 1846 1847 1848 1849 1850 1851
	return ret;
}

/*
 * Flush any dirty caps back to the mds.  If we aren't asked to wait,
 * queue inode for flush but don't do so immediately, because we can
 * get by with fewer MDS messages if we wait for data writeback to
 * complete first.
 */
1852
int ceph_write_inode(struct inode *inode, struct writeback_control *wbc)
S
Sage Weil 已提交
1853 1854 1855 1856 1857
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	unsigned flush_tid;
	int err = 0;
	int dirty;
1858
	int wait = wbc->sync_mode == WB_SYNC_ALL;
S
Sage Weil 已提交
1859 1860 1861 1862 1863 1864 1865 1866

	dout("write_inode %p wait=%d\n", inode, wait);
	if (wait) {
		dirty = try_flush_caps(inode, NULL, &flush_tid);
		if (dirty)
			err = wait_event_interruptible(ci->i_cap_wq,
				       caps_are_flushed(inode, flush_tid));
	} else {
1867
		struct ceph_mds_client *mdsc =
1868
			ceph_sb_to_client(inode->i_sb)->mdsc;
S
Sage Weil 已提交
1869

1870
		spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1871 1872
		if (__ceph_caps_dirty(ci))
			__cap_delay_requeue_front(mdsc, ci);
1873
		spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895
	}
	return err;
}

/*
 * After a recovering MDS goes active, we need to resend any caps
 * we were flushing.
 *
 * Caller holds session->s_mutex.
 */
static void kick_flushing_capsnaps(struct ceph_mds_client *mdsc,
				   struct ceph_mds_session *session)
{
	struct ceph_cap_snap *capsnap;

	dout("kick_flushing_capsnaps mds%d\n", session->s_mds);
	list_for_each_entry(capsnap, &session->s_cap_snaps_flushing,
			    flushing_item) {
		struct ceph_inode_info *ci = capsnap->ci;
		struct inode *inode = &ci->vfs_inode;
		struct ceph_cap *cap;

1896
		spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1897 1898 1899 1900
		cap = ci->i_auth_cap;
		if (cap && cap->session == session) {
			dout("kick_flushing_caps %p cap %p capsnap %p\n", inode,
			     cap, capsnap);
1901
			__ceph_flush_snaps(ci, &session, 1);
S
Sage Weil 已提交
1902 1903 1904 1905
		} else {
			pr_err("%p auth cap %p not mds%d ???\n", inode,
			       cap, session->s_mds);
		}
1906
		spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922
	}
}

void ceph_kick_flushing_caps(struct ceph_mds_client *mdsc,
			     struct ceph_mds_session *session)
{
	struct ceph_inode_info *ci;

	kick_flushing_capsnaps(mdsc, session);

	dout("kick_flushing_caps mds%d\n", session->s_mds);
	list_for_each_entry(ci, &session->s_cap_flushing, i_flushing_item) {
		struct inode *inode = &ci->vfs_inode;
		struct ceph_cap *cap;
		int delayed = 0;

1923
		spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1924 1925 1926 1927 1928 1929 1930 1931 1932 1933
		cap = ci->i_auth_cap;
		if (cap && cap->session == session) {
			dout("kick_flushing_caps %p cap %p %s\n", inode,
			     cap, ceph_cap_string(ci->i_flushing_caps));
			delayed = __send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH,
					     __ceph_caps_used(ci),
					     __ceph_caps_wanted(ci),
					     cap->issued | cap->implemented,
					     ci->i_flushing_caps, NULL);
			if (delayed) {
1934
				spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1935
				__cap_delay_requeue(mdsc, ci);
1936
				spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1937 1938 1939 1940
			}
		} else {
			pr_err("%p auth cap %p not mds%d ???\n", inode,
			       cap, session->s_mds);
1941
			spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
1942 1943 1944 1945
		}
	}
}

1946 1947 1948 1949 1950 1951 1952 1953
static void kick_flushing_inode_caps(struct ceph_mds_client *mdsc,
				     struct ceph_mds_session *session,
				     struct inode *inode)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	struct ceph_cap *cap;
	int delayed = 0;

1954
	spin_lock(&ci->i_ceph_lock);
1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
	cap = ci->i_auth_cap;
	dout("kick_flushing_inode_caps %p flushing %s flush_seq %lld\n", inode,
	     ceph_cap_string(ci->i_flushing_caps), ci->i_cap_flush_seq);
	__ceph_flush_snaps(ci, &session, 1);
	if (ci->i_flushing_caps) {
		delayed = __send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH,
				     __ceph_caps_used(ci),
				     __ceph_caps_wanted(ci),
				     cap->issued | cap->implemented,
				     ci->i_flushing_caps, NULL);
		if (delayed) {
1966
			spin_lock(&ci->i_ceph_lock);
1967
			__cap_delay_requeue(mdsc, ci);
1968
			spin_unlock(&ci->i_ceph_lock);
1969 1970
		}
	} else {
1971
		spin_unlock(&ci->i_ceph_lock);
1972 1973 1974
	}
}

S
Sage Weil 已提交
1975 1976 1977 1978 1979

/*
 * Take references to capabilities we hold, so that we don't release
 * them to the MDS prematurely.
 *
1980
 * Protected by i_ceph_lock.
S
Sage Weil 已提交
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992
 */
static void __take_cap_refs(struct ceph_inode_info *ci, int got)
{
	if (got & CEPH_CAP_PIN)
		ci->i_pin_ref++;
	if (got & CEPH_CAP_FILE_RD)
		ci->i_rd_ref++;
	if (got & CEPH_CAP_FILE_CACHE)
		ci->i_rdcache_ref++;
	if (got & CEPH_CAP_FILE_WR)
		ci->i_wr_ref++;
	if (got & CEPH_CAP_FILE_BUFFER) {
1993
		if (ci->i_wb_ref == 0)
1994
			ihold(&ci->vfs_inode);
1995 1996 1997
		ci->i_wb_ref++;
		dout("__take_cap_refs %p wb %d -> %d (?)\n",
		     &ci->vfs_inode, ci->i_wb_ref-1, ci->i_wb_ref);
S
Sage Weil 已提交
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
	}
}

/*
 * Try to grab cap references.  Specify those refs we @want, and the
 * minimal set we @need.  Also include the larger offset we are writing
 * to (when applicable), and check against max_size here as well.
 * Note that caller is responsible for ensuring max_size increases are
 * requested from the MDS.
 */
static int try_get_cap_refs(struct ceph_inode_info *ci, int need, int want,
			    int *got, loff_t endoff, int *check_max, int *err)
{
	struct inode *inode = &ci->vfs_inode;
	int ret = 0;
	int have, implemented;
2014
	int file_wanted;
S
Sage Weil 已提交
2015 2016 2017

	dout("get_cap_refs %p need %s want %s\n", inode,
	     ceph_cap_string(need), ceph_cap_string(want));
2018
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2019

2020 2021 2022 2023 2024
	/* make sure file is actually open */
	file_wanted = __ceph_caps_file_wanted(ci);
	if ((file_wanted & need) == 0) {
		dout("try_get_cap_refs need %s file_wanted %s, EBADF\n",
		     ceph_cap_string(need), ceph_cap_string(file_wanted));
S
Sage Weil 已提交
2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078
		*err = -EBADF;
		ret = 1;
		goto out;
	}

	if (need & CEPH_CAP_FILE_WR) {
		if (endoff >= 0 && endoff > (loff_t)ci->i_max_size) {
			dout("get_cap_refs %p endoff %llu > maxsize %llu\n",
			     inode, endoff, ci->i_max_size);
			if (endoff > ci->i_wanted_max_size) {
				*check_max = 1;
				ret = 1;
			}
			goto out;
		}
		/*
		 * If a sync write is in progress, we must wait, so that we
		 * can get a final snapshot value for size+mtime.
		 */
		if (__ceph_have_pending_cap_snap(ci)) {
			dout("get_cap_refs %p cap_snap_pending\n", inode);
			goto out;
		}
	}
	have = __ceph_caps_issued(ci, &implemented);

	/*
	 * disallow writes while a truncate is pending
	 */
	if (ci->i_truncate_pending)
		have &= ~CEPH_CAP_FILE_WR;

	if ((have & need) == need) {
		/*
		 * Look at (implemented & ~have & not) so that we keep waiting
		 * on transition from wanted -> needed caps.  This is needed
		 * for WRBUFFER|WR -> WR to avoid a new WR sync write from
		 * going before a prior buffered writeback happens.
		 */
		int not = want & ~(have & need);
		int revoking = implemented & ~have;
		dout("get_cap_refs %p have %s but not %s (revoking %s)\n",
		     inode, ceph_cap_string(have), ceph_cap_string(not),
		     ceph_cap_string(revoking));
		if ((revoking & not) == 0) {
			*got = need | (have & want);
			__take_cap_refs(ci, *got);
			ret = 1;
		}
	} else {
		dout("get_cap_refs %p have %s needed %s\n", inode,
		     ceph_cap_string(have), ceph_cap_string(need));
	}
out:
2079
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095
	dout("get_cap_refs %p ret %d got %s\n", inode,
	     ret, ceph_cap_string(*got));
	return ret;
}

/*
 * Check the offset we are writing up to against our current
 * max_size.  If necessary, tell the MDS we want to write to
 * a larger offset.
 */
static void check_max_size(struct inode *inode, loff_t endoff)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	int check = 0;

	/* do we need to explicitly request a larger max_size? */
2096
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2097 2098 2099 2100 2101 2102 2103 2104
	if ((endoff >= ci->i_max_size ||
	     endoff > (inode->i_size << 1)) &&
	    endoff > ci->i_wanted_max_size) {
		dout("write %p at large endoff %llu, req max_size\n",
		     inode, endoff);
		ci->i_wanted_max_size = endoff;
		check = 1;
	}
2105
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141
	if (check)
		ceph_check_caps(ci, CHECK_CAPS_AUTHONLY, NULL);
}

/*
 * Wait for caps, and take cap references.  If we can't get a WR cap
 * due to a small max_size, make sure we check_max_size (and possibly
 * ask the mds) so we don't get hung up indefinitely.
 */
int ceph_get_caps(struct ceph_inode_info *ci, int need, int want, int *got,
		  loff_t endoff)
{
	int check_max, ret, err;

retry:
	if (endoff > 0)
		check_max_size(&ci->vfs_inode, endoff);
	check_max = 0;
	err = 0;
	ret = wait_event_interruptible(ci->i_cap_wq,
				       try_get_cap_refs(ci, need, want,
							got, endoff,
							&check_max, &err));
	if (err)
		ret = err;
	if (check_max)
		goto retry;
	return ret;
}

/*
 * Take cap refs.  Caller must already know we hold at least one ref
 * on the caps in question or we don't know this is safe.
 */
void ceph_get_cap_refs(struct ceph_inode_info *ci, int caps)
{
2142
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2143
	__take_cap_refs(ci, caps);
2144
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161
}

/*
 * Release cap refs.
 *
 * If we released the last ref on any given cap, call ceph_check_caps
 * to release (or schedule a release).
 *
 * If we are releasing a WR cap (from a sync write), finalize any affected
 * cap_snap, and wake up any waiters.
 */
void ceph_put_cap_refs(struct ceph_inode_info *ci, int had)
{
	struct inode *inode = &ci->vfs_inode;
	int last = 0, put = 0, flushsnaps = 0, wake = 0;
	struct ceph_cap_snap *capsnap;

2162
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2163 2164 2165 2166 2167 2168 2169 2170 2171
	if (had & CEPH_CAP_PIN)
		--ci->i_pin_ref;
	if (had & CEPH_CAP_FILE_RD)
		if (--ci->i_rd_ref == 0)
			last++;
	if (had & CEPH_CAP_FILE_CACHE)
		if (--ci->i_rdcache_ref == 0)
			last++;
	if (had & CEPH_CAP_FILE_BUFFER) {
2172
		if (--ci->i_wb_ref == 0) {
S
Sage Weil 已提交
2173 2174 2175
			last++;
			put++;
		}
2176 2177
		dout("put_cap_refs %p wb %d -> %d (?)\n",
		     inode, ci->i_wb_ref+1, ci->i_wb_ref);
S
Sage Weil 已提交
2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194
	}
	if (had & CEPH_CAP_FILE_WR)
		if (--ci->i_wr_ref == 0) {
			last++;
			if (!list_empty(&ci->i_cap_snaps)) {
				capsnap = list_first_entry(&ci->i_cap_snaps,
						     struct ceph_cap_snap,
						     ci_item);
				if (capsnap->writing) {
					capsnap->writing = 0;
					flushsnaps =
						__ceph_finish_cap_snap(ci,
								       capsnap);
					wake = 1;
				}
			}
		}
2195
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2196

2197 2198
	dout("put_cap_refs %p had %s%s%s\n", inode, ceph_cap_string(had),
	     last ? " last" : "", put ? " put" : "");
S
Sage Weil 已提交
2199 2200 2201 2202 2203 2204

	if (last && !flushsnaps)
		ceph_check_caps(ci, 0, NULL);
	else if (flushsnaps)
		ceph_flush_snaps(ci);
	if (wake)
2205
		wake_up_all(&ci->i_cap_wq);
S
Sage Weil 已提交
2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221
	if (put)
		iput(inode);
}

/*
 * Release @nr WRBUFFER refs on dirty pages for the given @snapc snap
 * context.  Adjust per-snap dirty page accounting as appropriate.
 * Once all dirty data for a cap_snap is flushed, flush snapped file
 * metadata back to the MDS.  If we dropped the last ref, call
 * ceph_check_caps.
 */
void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr,
				struct ceph_snap_context *snapc)
{
	struct inode *inode = &ci->vfs_inode;
	int last = 0;
2222 2223
	int complete_capsnap = 0;
	int drop_capsnap = 0;
S
Sage Weil 已提交
2224 2225 2226
	int found = 0;
	struct ceph_cap_snap *capsnap = NULL;

2227
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2228 2229 2230 2231 2232
	ci->i_wrbuffer_ref -= nr;
	last = !ci->i_wrbuffer_ref;

	if (ci->i_head_snapc == snapc) {
		ci->i_wrbuffer_ref_head -= nr;
2233 2234 2235
		if (ci->i_wrbuffer_ref_head == 0 &&
		    ci->i_dirty_caps == 0 && ci->i_flushing_caps == 0) {
			BUG_ON(!ci->i_head_snapc);
S
Sage Weil 已提交
2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251
			ceph_put_snap_context(ci->i_head_snapc);
			ci->i_head_snapc = NULL;
		}
		dout("put_wrbuffer_cap_refs on %p head %d/%d -> %d/%d %s\n",
		     inode,
		     ci->i_wrbuffer_ref+nr, ci->i_wrbuffer_ref_head+nr,
		     ci->i_wrbuffer_ref, ci->i_wrbuffer_ref_head,
		     last ? " LAST" : "");
	} else {
		list_for_each_entry(capsnap, &ci->i_cap_snaps, ci_item) {
			if (capsnap->context == snapc) {
				found = 1;
				break;
			}
		}
		BUG_ON(!found);
2252 2253 2254 2255 2256 2257 2258 2259
		capsnap->dirty_pages -= nr;
		if (capsnap->dirty_pages == 0) {
			complete_capsnap = 1;
			if (capsnap->dirty == 0)
				/* cap writeback completed before we created
				 * the cap_snap; no FLUSHSNAP is needed */
				drop_capsnap = 1;
		}
S
Sage Weil 已提交
2260
		dout("put_wrbuffer_cap_refs on %p cap_snap %p "
2261
		     " snap %lld %d/%d -> %d/%d %s%s%s\n",
S
Sage Weil 已提交
2262 2263 2264 2265
		     inode, capsnap, capsnap->context->seq,
		     ci->i_wrbuffer_ref+nr, capsnap->dirty_pages + nr,
		     ci->i_wrbuffer_ref, capsnap->dirty_pages,
		     last ? " (wrbuffer last)" : "",
2266 2267 2268 2269 2270 2271 2272 2273
		     complete_capsnap ? " (complete capsnap)" : "",
		     drop_capsnap ? " (drop capsnap)" : "");
		if (drop_capsnap) {
			ceph_put_snap_context(capsnap->context);
			list_del(&capsnap->ci_item);
			list_del(&capsnap->flushing_item);
			ceph_put_cap_snap(capsnap);
		}
S
Sage Weil 已提交
2274 2275
	}

2276
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2277 2278 2279 2280

	if (last) {
		ceph_check_caps(ci, CHECK_CAPS_AUTHONLY, NULL);
		iput(inode);
2281
	} else if (complete_capsnap) {
S
Sage Weil 已提交
2282
		ceph_flush_snaps(ci);
2283
		wake_up_all(&ci->i_cap_wq);
S
Sage Weil 已提交
2284
	}
2285 2286
	if (drop_capsnap)
		iput(inode);
S
Sage Weil 已提交
2287 2288 2289 2290 2291 2292
}

/*
 * Handle a cap GRANT message from the MDS.  (Note that a GRANT may
 * actually be a revocation if it specifies a smaller cap set.)
 *
2293
 * caller holds s_mutex and i_ceph_lock, we drop both.
2294
 *
S
Sage Weil 已提交
2295 2296 2297 2298 2299
 * return value:
 *  0 - ok
 *  1 - check_caps on auth cap only (writeback)
 *  2 - check_caps (ack revoke)
 */
2300 2301 2302 2303
static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
			     struct ceph_mds_session *session,
			     struct ceph_cap *cap,
			     struct ceph_buffer *xattr_buf)
2304
		__releases(ci->i_ceph_lock)
S
Sage Weil 已提交
2305 2306 2307
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	int mds = session->s_mds;
2308
	int seq = le32_to_cpu(grant->seq);
S
Sage Weil 已提交
2309 2310 2311 2312 2313
	int newcaps = le32_to_cpu(grant->caps);
	int issued, implemented, used, wanted, dirty;
	u64 size = le64_to_cpu(grant->size);
	u64 max_size = le64_to_cpu(grant->max_size);
	struct timespec mtime, atime, ctime;
2314
	int check_caps = 0;
S
Sage Weil 已提交
2315 2316 2317
	int wake = 0;
	int writeback = 0;
	int revoked_rdcache = 0;
2318
	int queue_invalidate = 0;
S
Sage Weil 已提交
2319

2320 2321
	dout("handle_cap_grant inode %p cap %p mds%d seq %d %s\n",
	     inode, cap, mds, seq, ceph_cap_string(newcaps));
S
Sage Weil 已提交
2322 2323 2324 2325 2326 2327 2328 2329
	dout(" size %llu max_size %llu, i_size %llu\n", size, max_size,
		inode->i_size);

	/*
	 * If CACHE is being revoked, and we have no dirty buffers,
	 * try to invalidate (once).  (If there are dirty buffers, we
	 * will invalidate _after_ writeback.)
	 */
2330 2331
	if (((cap->issued & ~newcaps) & CEPH_CAP_FILE_CACHE) &&
	    (newcaps & CEPH_CAP_FILE_LAZYIO) == 0 &&
2332
	    !ci->i_wrbuffer_ref) {
2333 2334 2335
		if (try_nonblocking_invalidate(inode) == 0) {
			revoked_rdcache = 1;
		} else {
S
Sage Weil 已提交
2336 2337 2338
			/* there were locked pages.. invalidate later
			   in a separate thread. */
			if (ci->i_rdcache_revoking != ci->i_rdcache_gen) {
2339
				queue_invalidate = 1;
S
Sage Weil 已提交
2340 2341 2342 2343 2344 2345 2346 2347 2348 2349
				ci->i_rdcache_revoking = ci->i_rdcache_gen;
			}
		}
	}

	/* side effects now are allowed */

	issued = __ceph_caps_issued(ci, &implemented);
	issued |= implemented | __ceph_caps_dirty(ci);

2350
	cap->cap_gen = session->s_cap_gen;
S
Sage Weil 已提交
2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362

	__check_cap_issue(ci, cap, newcaps);

	if ((issued & CEPH_CAP_AUTH_EXCL) == 0) {
		inode->i_mode = le32_to_cpu(grant->mode);
		inode->i_uid = le32_to_cpu(grant->uid);
		inode->i_gid = le32_to_cpu(grant->gid);
		dout("%p mode 0%o uid.gid %d.%d\n", inode, inode->i_mode,
		     inode->i_uid, inode->i_gid);
	}

	if ((issued & CEPH_CAP_LINK_EXCL) == 0)
M
Miklos Szeredi 已提交
2363
		set_nlink(inode, le32_to_cpu(grant->nlink));
S
Sage Weil 已提交
2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 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

	if ((issued & CEPH_CAP_XATTR_EXCL) == 0 && grant->xattr_len) {
		int len = le32_to_cpu(grant->xattr_len);
		u64 version = le64_to_cpu(grant->xattr_version);

		if (version > ci->i_xattrs.version) {
			dout(" got new xattrs v%llu on %p len %d\n",
			     version, inode, len);
			if (ci->i_xattrs.blob)
				ceph_buffer_put(ci->i_xattrs.blob);
			ci->i_xattrs.blob = ceph_buffer_get(xattr_buf);
			ci->i_xattrs.version = version;
		}
	}

	/* size/ctime/mtime/atime? */
	ceph_fill_file_size(inode, issued,
			    le32_to_cpu(grant->truncate_seq),
			    le64_to_cpu(grant->truncate_size), size);
	ceph_decode_timespec(&mtime, &grant->mtime);
	ceph_decode_timespec(&atime, &grant->atime);
	ceph_decode_timespec(&ctime, &grant->ctime);
	ceph_fill_file_time(inode, issued,
			    le32_to_cpu(grant->time_warp_seq), &ctime, &mtime,
			    &atime);

	/* max size increase? */
	if (max_size != ci->i_max_size) {
		dout("max_size %lld -> %llu\n", ci->i_max_size, max_size);
		ci->i_max_size = max_size;
		if (max_size >= ci->i_wanted_max_size) {
			ci->i_wanted_max_size = 0;  /* reset */
			ci->i_requested_max_size = 0;
		}
		wake = 1;
	}

	/* check cap bits */
	wanted = __ceph_caps_wanted(ci);
	used = __ceph_caps_used(ci);
	dirty = __ceph_caps_dirty(ci);
	dout(" my wanted = %s, used = %s, dirty %s\n",
	     ceph_cap_string(wanted),
	     ceph_cap_string(used),
	     ceph_cap_string(dirty));
	if (wanted != le32_to_cpu(grant->wanted)) {
		dout("mds wanted %s -> %s\n",
		     ceph_cap_string(le32_to_cpu(grant->wanted)),
		     ceph_cap_string(wanted));
		grant->wanted = cpu_to_le32(wanted);
	}

	cap->seq = seq;

	/* file layout may have changed */
	ci->i_layout = grant->layout;

	/* revocation, grant, or no-op? */
	if (cap->issued & ~newcaps) {
2423 2424 2425 2426 2427 2428
		int revoking = cap->issued & ~newcaps;

		dout("revocation: %s -> %s (revoking %s)\n",
		     ceph_cap_string(cap->issued),
		     ceph_cap_string(newcaps),
		     ceph_cap_string(revoking));
2429
		if (revoking & used & CEPH_CAP_FILE_BUFFER)
2430 2431 2432 2433 2434 2435 2436 2437 2438
			writeback = 1;  /* initiate writeback; will delay ack */
		else if (revoking == CEPH_CAP_FILE_CACHE &&
			 (newcaps & CEPH_CAP_FILE_LAZYIO) == 0 &&
			 queue_invalidate)
			; /* do nothing yet, invalidation will be queued */
		else if (cap == ci->i_auth_cap)
			check_caps = 1; /* check auth cap only */
		else
			check_caps = 2; /* check all caps */
S
Sage Weil 已提交
2439
		cap->issued = newcaps;
2440
		cap->implemented |= newcaps;
S
Sage Weil 已提交
2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452
	} else if (cap->issued == newcaps) {
		dout("caps unchanged: %s -> %s\n",
		     ceph_cap_string(cap->issued), ceph_cap_string(newcaps));
	} else {
		dout("grant: %s -> %s\n", ceph_cap_string(cap->issued),
		     ceph_cap_string(newcaps));
		cap->issued = newcaps;
		cap->implemented |= newcaps; /* add bits only, to
					      * avoid stepping on a
					      * pending revocation */
		wake = 1;
	}
2453
	BUG_ON(cap->issued & ~cap->implemented);
S
Sage Weil 已提交
2454

2455
	spin_unlock(&ci->i_ceph_lock);
2456
	if (writeback)
S
Sage Weil 已提交
2457 2458 2459 2460 2461
		/*
		 * queue inode for writeback: we can't actually call
		 * filemap_write_and_wait, etc. from message handler
		 * context.
		 */
2462 2463 2464
		ceph_queue_writeback(inode);
	if (queue_invalidate)
		ceph_queue_invalidate(inode);
S
Sage Weil 已提交
2465
	if (wake)
2466
		wake_up_all(&ci->i_cap_wq);
2467 2468 2469 2470 2471 2472 2473 2474

	if (check_caps == 1)
		ceph_check_caps(ci, CHECK_CAPS_NODELAY|CHECK_CAPS_AUTHONLY,
				session);
	else if (check_caps == 2)
		ceph_check_caps(ci, CHECK_CAPS_NODELAY, session);
	else
		mutex_unlock(&session->s_mutex);
S
Sage Weil 已提交
2475 2476 2477 2478 2479 2480
}

/*
 * Handle FLUSH_ACK from MDS, indicating that metadata we sent to the
 * MDS has been safely committed.
 */
2481
static void handle_cap_flush_ack(struct inode *inode, u64 flush_tid,
S
Sage Weil 已提交
2482 2483 2484
				 struct ceph_mds_caps *m,
				 struct ceph_mds_session *session,
				 struct ceph_cap *cap)
2485
	__releases(ci->i_ceph_lock)
S
Sage Weil 已提交
2486 2487
{
	struct ceph_inode_info *ci = ceph_inode(inode);
2488
	struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc;
S
Sage Weil 已提交
2489 2490 2491
	unsigned seq = le32_to_cpu(m->seq);
	int dirty = le32_to_cpu(m->dirty);
	int cleaned = 0;
2492
	int drop = 0;
S
Sage Weil 已提交
2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520
	int i;

	for (i = 0; i < CEPH_CAP_BITS; i++)
		if ((dirty & (1 << i)) &&
		    flush_tid == ci->i_cap_flush_tid[i])
			cleaned |= 1 << i;

	dout("handle_cap_flush_ack inode %p mds%d seq %d on %s cleaned %s,"
	     " flushing %s -> %s\n",
	     inode, session->s_mds, seq, ceph_cap_string(dirty),
	     ceph_cap_string(cleaned), ceph_cap_string(ci->i_flushing_caps),
	     ceph_cap_string(ci->i_flushing_caps & ~cleaned));

	if (ci->i_flushing_caps == (ci->i_flushing_caps & ~cleaned))
		goto out;

	ci->i_flushing_caps &= ~cleaned;

	spin_lock(&mdsc->cap_dirty_lock);
	if (ci->i_flushing_caps == 0) {
		list_del_init(&ci->i_flushing_item);
		if (!list_empty(&session->s_cap_flushing))
			dout(" mds%d still flushing cap on %p\n",
			     session->s_mds,
			     &list_entry(session->s_cap_flushing.next,
					 struct ceph_inode_info,
					 i_flushing_item)->vfs_inode);
		mdsc->num_cap_flushing--;
2521
		wake_up_all(&mdsc->cap_flushing_wq);
S
Sage Weil 已提交
2522
		dout(" inode %p now !flushing\n", inode);
2523 2524 2525 2526 2527

		if (ci->i_dirty_caps == 0) {
			dout(" inode %p now clean\n", inode);
			BUG_ON(!list_empty(&ci->i_dirty_item));
			drop = 1;
2528 2529 2530 2531 2532
			if (ci->i_wrbuffer_ref_head == 0) {
				BUG_ON(!ci->i_head_snapc);
				ceph_put_snap_context(ci->i_head_snapc);
				ci->i_head_snapc = NULL;
			}
S
Sage Weil 已提交
2533 2534
		} else {
			BUG_ON(list_empty(&ci->i_dirty_item));
2535
		}
S
Sage Weil 已提交
2536 2537
	}
	spin_unlock(&mdsc->cap_dirty_lock);
2538
	wake_up_all(&ci->i_cap_wq);
S
Sage Weil 已提交
2539 2540

out:
2541
	spin_unlock(&ci->i_ceph_lock);
2542
	if (drop)
S
Sage Weil 已提交
2543 2544 2545 2546 2547 2548 2549 2550 2551
		iput(inode);
}

/*
 * Handle FLUSHSNAP_ACK.  MDS has flushed snap data to disk and we can
 * throw away our cap_snap.
 *
 * Caller hold s_mutex.
 */
2552
static void handle_cap_flushsnap_ack(struct inode *inode, u64 flush_tid,
S
Sage Weil 已提交
2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563
				     struct ceph_mds_caps *m,
				     struct ceph_mds_session *session)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	u64 follows = le64_to_cpu(m->snap_follows);
	struct ceph_cap_snap *capsnap;
	int drop = 0;

	dout("handle_cap_flushsnap_ack inode %p ci %p mds%d follows %lld\n",
	     inode, ci, session->s_mds, follows);

2564
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2565 2566 2567 2568 2569 2570 2571 2572 2573
	list_for_each_entry(capsnap, &ci->i_cap_snaps, ci_item) {
		if (capsnap->follows == follows) {
			if (capsnap->flush_tid != flush_tid) {
				dout(" cap_snap %p follows %lld tid %lld !="
				     " %lld\n", capsnap, follows,
				     flush_tid, capsnap->flush_tid);
				break;
			}
			WARN_ON(capsnap->dirty_pages || capsnap->writing);
2574 2575
			dout(" removing %p cap_snap %p follows %lld\n",
			     inode, capsnap, follows);
S
Sage Weil 已提交
2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586
			ceph_put_snap_context(capsnap->context);
			list_del(&capsnap->ci_item);
			list_del(&capsnap->flushing_item);
			ceph_put_cap_snap(capsnap);
			drop = 1;
			break;
		} else {
			dout(" skipping cap_snap %p follows %lld\n",
			     capsnap, capsnap->follows);
		}
	}
2587
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599
	if (drop)
		iput(inode);
}

/*
 * Handle TRUNC from MDS, indicating file truncation.
 *
 * caller hold s_mutex.
 */
static void handle_cap_trunc(struct inode *inode,
			     struct ceph_mds_caps *trunc,
			     struct ceph_mds_session *session)
2600
	__releases(ci->i_ceph_lock)
S
Sage Weil 已提交
2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	int mds = session->s_mds;
	int seq = le32_to_cpu(trunc->seq);
	u32 truncate_seq = le32_to_cpu(trunc->truncate_seq);
	u64 truncate_size = le64_to_cpu(trunc->truncate_size);
	u64 size = le64_to_cpu(trunc->size);
	int implemented = 0;
	int dirty = __ceph_caps_dirty(ci);
	int issued = __ceph_caps_issued(ceph_inode(inode), &implemented);
	int queue_trunc = 0;

	issued |= implemented | dirty;

	dout("handle_cap_trunc inode %p mds%d seq %d to %lld seq %d\n",
	     inode, mds, seq, truncate_size, truncate_seq);
	queue_trunc = ceph_fill_file_size(inode, issued,
					  truncate_seq, truncate_size, size);
2619
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2620 2621

	if (queue_trunc)
2622
		ceph_queue_vmtruncate(inode);
S
Sage Weil 已提交
2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633
}

/*
 * Handle EXPORT from MDS.  Cap is being migrated _from_ this mds to a
 * different one.  If we are the most recent migration we've seen (as
 * indicated by mseq), make note of the migrating cap bits for the
 * duration (until we see the corresponding IMPORT).
 *
 * caller holds s_mutex
 */
static void handle_cap_export(struct inode *inode, struct ceph_mds_caps *ex,
2634 2635
			      struct ceph_mds_session *session,
			      int *open_target_sessions)
S
Sage Weil 已提交
2636
{
S
Sage Weil 已提交
2637
	struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc;
S
Sage Weil 已提交
2638 2639 2640 2641 2642 2643 2644 2645 2646 2647
	struct ceph_inode_info *ci = ceph_inode(inode);
	int mds = session->s_mds;
	unsigned mseq = le32_to_cpu(ex->migrate_seq);
	struct ceph_cap *cap = NULL, *t;
	struct rb_node *p;
	int remember = 1;

	dout("handle_cap_export inode %p ci %p mds%d mseq %d\n",
	     inode, ci, mds, mseq);

2648
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667

	/* make sure we haven't seen a higher mseq */
	for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
		t = rb_entry(p, struct ceph_cap, ci_node);
		if (ceph_seq_cmp(t->mseq, mseq) > 0) {
			dout(" higher mseq on cap from mds%d\n",
			     t->session->s_mds);
			remember = 0;
		}
		if (t->session->s_mds == mds)
			cap = t;
	}

	if (cap) {
		if (remember) {
			/* make note */
			ci->i_cap_exporting_mds = mds;
			ci->i_cap_exporting_mseq = mseq;
			ci->i_cap_exporting_issued = cap->issued;
2668 2669 2670 2671 2672 2673

			/*
			 * make sure we have open sessions with all possible
			 * export targets, so that we get the matching IMPORT
			 */
			*open_target_sessions = 1;
S
Sage Weil 已提交
2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686

			/*
			 * we can't flush dirty caps that we've seen the
			 * EXPORT but no IMPORT for
			 */
			spin_lock(&mdsc->cap_dirty_lock);
			if (!list_empty(&ci->i_dirty_item)) {
				dout(" moving %p to cap_dirty_migrating\n",
				     inode);
				list_move(&ci->i_dirty_item,
					  &mdsc->cap_dirty_migrating);
			}
			spin_unlock(&mdsc->cap_dirty_lock);
S
Sage Weil 已提交
2687
		}
S
Sage Weil 已提交
2688
		__ceph_remove_cap(cap);
S
Sage Weil 已提交
2689
	}
2690
	/* else, we already released it */
S
Sage Weil 已提交
2691

2692
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723
}

/*
 * Handle cap IMPORT.  If there are temp bits from an older EXPORT,
 * clean them up.
 *
 * caller holds s_mutex.
 */
static void handle_cap_import(struct ceph_mds_client *mdsc,
			      struct inode *inode, struct ceph_mds_caps *im,
			      struct ceph_mds_session *session,
			      void *snaptrace, int snaptrace_len)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	int mds = session->s_mds;
	unsigned issued = le32_to_cpu(im->caps);
	unsigned wanted = le32_to_cpu(im->wanted);
	unsigned seq = le32_to_cpu(im->seq);
	unsigned mseq = le32_to_cpu(im->migrate_seq);
	u64 realmino = le64_to_cpu(im->realm);
	u64 cap_id = le64_to_cpu(im->cap_id);

	if (ci->i_cap_exporting_mds >= 0 &&
	    ceph_seq_cmp(ci->i_cap_exporting_mseq, mseq) < 0) {
		dout("handle_cap_import inode %p ci %p mds%d mseq %d"
		     " - cleared exporting from mds%d\n",
		     inode, ci, mds, mseq,
		     ci->i_cap_exporting_mds);
		ci->i_cap_exporting_issued = 0;
		ci->i_cap_exporting_mseq = 0;
		ci->i_cap_exporting_mds = -1;
S
Sage Weil 已提交
2724 2725 2726 2727 2728 2729 2730

		spin_lock(&mdsc->cap_dirty_lock);
		if (!list_empty(&ci->i_dirty_item)) {
			dout(" moving %p back to cap_dirty\n", inode);
			list_move(&ci->i_dirty_item, &mdsc->cap_dirty);
		}
		spin_unlock(&mdsc->cap_dirty_lock);
S
Sage Weil 已提交
2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742
	} else {
		dout("handle_cap_import inode %p ci %p mds%d mseq %d\n",
		     inode, ci, mds, mseq);
	}

	down_write(&mdsc->snap_rwsem);
	ceph_update_snap_trace(mdsc, snaptrace, snaptrace+snaptrace_len,
			       false);
	downgrade_write(&mdsc->snap_rwsem);
	ceph_add_cap(inode, session, cap_id, -1,
		     issued, wanted, seq, mseq, realmino, CEPH_CAP_FLAG_AUTH,
		     NULL /* no caps context */);
2743
	kick_flushing_inode_caps(mdsc, session, inode);
S
Sage Weil 已提交
2744
	up_read(&mdsc->snap_rwsem);
2745 2746

	/* make sure we re-request max_size, if necessary */
2747
	spin_lock(&ci->i_ceph_lock);
2748
	ci->i_requested_max_size = 0;
2749
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761
}

/*
 * Handle a caps message from the MDS.
 *
 * Identify the appropriate session, inode, and call the right handler
 * based on the cap op.
 */
void ceph_handle_caps(struct ceph_mds_session *session,
		      struct ceph_msg *msg)
{
	struct ceph_mds_client *mdsc = session->s_mdsc;
2762
	struct super_block *sb = mdsc->fsc->sb;
S
Sage Weil 已提交
2763
	struct inode *inode;
2764
	struct ceph_inode_info *ci;
S
Sage Weil 已提交
2765 2766
	struct ceph_cap *cap;
	struct ceph_mds_caps *h;
2767
	int mds = session->s_mds;
S
Sage Weil 已提交
2768
	int op;
2769
	u32 seq, mseq;
S
Sage Weil 已提交
2770 2771 2772
	struct ceph_vino vino;
	u64 cap_id;
	u64 size, max_size;
2773
	u64 tid;
2774
	void *snaptrace;
2775 2776 2777
	size_t snaptrace_len;
	void *flock;
	u32 flock_len;
2778
	int open_target_sessions = 0;
S
Sage Weil 已提交
2779 2780 2781 2782

	dout("handle_caps from mds%d\n", mds);

	/* decode */
2783
	tid = le64_to_cpu(msg->hdr.tid);
S
Sage Weil 已提交
2784 2785 2786 2787 2788 2789 2790 2791
	if (msg->front.iov_len < sizeof(*h))
		goto bad;
	h = msg->front.iov_base;
	op = le32_to_cpu(h->op);
	vino.ino = le64_to_cpu(h->ino);
	vino.snap = CEPH_NOSNAP;
	cap_id = le64_to_cpu(h->cap_id);
	seq = le32_to_cpu(h->seq);
2792
	mseq = le32_to_cpu(h->migrate_seq);
S
Sage Weil 已提交
2793 2794 2795
	size = le64_to_cpu(h->size);
	max_size = le64_to_cpu(h->max_size);

2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810
	snaptrace = h + 1;
	snaptrace_len = le32_to_cpu(h->snap_trace_len);

	if (le16_to_cpu(msg->hdr.version) >= 2) {
		void *p, *end;

		p = snaptrace + snaptrace_len;
		end = msg->front.iov_base + msg->front.iov_len;
		ceph_decode_32_safe(&p, end, flock_len, bad);
		flock = p;
	} else {
		flock = NULL;
		flock_len = 0;
	}

S
Sage Weil 已提交
2811 2812 2813 2814 2815 2816 2817
	mutex_lock(&session->s_mutex);
	session->s_seq++;
	dout(" mds%d seq %lld cap seq %u\n", session->s_mds, session->s_seq,
	     (unsigned)seq);

	/* lookup ino */
	inode = ceph_find_inode(sb, vino);
2818
	ci = ceph_inode(inode);
S
Sage Weil 已提交
2819 2820 2821 2822
	dout(" op %s ino %llx.%llx inode %p\n", ceph_cap_op_name(op), vino.ino,
	     vino.snap, inode);
	if (!inode) {
		dout(" i don't have ino %llx\n", vino.ino);
2823 2824 2825 2826

		if (op == CEPH_CAP_OP_IMPORT)
			__queue_cap_release(session, vino.ino, cap_id,
					    mseq, seq);
2827
		goto flush_cap_releases;
S
Sage Weil 已提交
2828 2829 2830 2831 2832
	}

	/* these will work even if we don't have a cap yet */
	switch (op) {
	case CEPH_CAP_OP_FLUSHSNAP_ACK:
2833
		handle_cap_flushsnap_ack(inode, tid, h, session);
S
Sage Weil 已提交
2834 2835 2836
		goto done;

	case CEPH_CAP_OP_EXPORT:
2837
		handle_cap_export(inode, h, session, &open_target_sessions);
S
Sage Weil 已提交
2838 2839 2840 2841
		goto done;

	case CEPH_CAP_OP_IMPORT:
		handle_cap_import(mdsc, inode, h, session,
2842
				  snaptrace, snaptrace_len);
2843
		ceph_check_caps(ceph_inode(inode), 0, session);
2844
		goto done_unlocked;
S
Sage Weil 已提交
2845 2846 2847
	}

	/* the rest require a cap */
2848
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2849 2850
	cap = __get_cap_for_mds(ceph_inode(inode), mds);
	if (!cap) {
2851
		dout(" no cap on %p ino %llx.%llx from mds%d\n",
S
Sage Weil 已提交
2852
		     inode, ceph_ino(inode), ceph_snap(inode), mds);
2853
		spin_unlock(&ci->i_ceph_lock);
2854
		goto flush_cap_releases;
S
Sage Weil 已提交
2855 2856
	}

2857
	/* note that each of these drops i_ceph_lock for us */
S
Sage Weil 已提交
2858 2859 2860
	switch (op) {
	case CEPH_CAP_OP_REVOKE:
	case CEPH_CAP_OP_GRANT:
2861 2862
		handle_cap_grant(inode, h, session, cap, msg->middle);
		goto done_unlocked;
S
Sage Weil 已提交
2863 2864

	case CEPH_CAP_OP_FLUSH_ACK:
2865
		handle_cap_flush_ack(inode, tid, h, session, cap);
S
Sage Weil 已提交
2866 2867 2868 2869 2870 2871 2872
		break;

	case CEPH_CAP_OP_TRUNC:
		handle_cap_trunc(inode, h, session);
		break;

	default:
2873
		spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2874 2875 2876 2877
		pr_err("ceph_handle_caps: unknown cap op %d %s\n", op,
		       ceph_cap_op_name(op));
	}

2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888
	goto done;

flush_cap_releases:
	/*
	 * send any full release message to try to move things
	 * along for the mds (who clearly thinks we still have this
	 * cap).
	 */
	ceph_add_cap_releases(mdsc, session);
	ceph_send_cap_releases(mdsc, session);

S
Sage Weil 已提交
2889
done:
2890 2891
	mutex_unlock(&session->s_mutex);
done_unlocked:
S
Sage Weil 已提交
2892 2893
	if (inode)
		iput(inode);
2894 2895
	if (open_target_sessions)
		ceph_mdsc_open_export_target_sessions(mdsc, session);
S
Sage Weil 已提交
2896 2897 2898 2899
	return;

bad:
	pr_err("ceph_handle_caps: corrupt message\n");
2900
	ceph_msg_dump(msg);
S
Sage Weil 已提交
2901 2902 2903 2904 2905 2906
	return;
}

/*
 * Delayed work handler to process end of delayed cap release LRU list.
 */
2907
void ceph_check_delayed_caps(struct ceph_mds_client *mdsc)
S
Sage Weil 已提交
2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930
{
	struct ceph_inode_info *ci;
	int flags = CHECK_CAPS_NODELAY;

	dout("check_delayed_caps\n");
	while (1) {
		spin_lock(&mdsc->cap_delay_lock);
		if (list_empty(&mdsc->cap_delay_list))
			break;
		ci = list_first_entry(&mdsc->cap_delay_list,
				      struct ceph_inode_info,
				      i_cap_delay_list);
		if ((ci->i_ceph_flags & CEPH_I_FLUSH) == 0 &&
		    time_before(jiffies, ci->i_hold_caps_max))
			break;
		list_del_init(&ci->i_cap_delay_list);
		spin_unlock(&mdsc->cap_delay_lock);
		dout("check_delayed_caps on %p\n", &ci->vfs_inode);
		ceph_check_caps(ci, flags, NULL);
	}
	spin_unlock(&mdsc->cap_delay_lock);
}

2931 2932 2933 2934 2935
/*
 * Flush all dirty caps to the mds
 */
void ceph_flush_dirty_caps(struct ceph_mds_client *mdsc)
{
S
Sage Weil 已提交
2936 2937
	struct ceph_inode_info *ci;
	struct inode *inode;
2938 2939 2940

	dout("flush_dirty_caps\n");
	spin_lock(&mdsc->cap_dirty_lock);
S
Sage Weil 已提交
2941 2942 2943
	while (!list_empty(&mdsc->cap_dirty)) {
		ci = list_first_entry(&mdsc->cap_dirty, struct ceph_inode_info,
				      i_dirty_item);
2944 2945
		inode = &ci->vfs_inode;
		ihold(inode);
S
Sage Weil 已提交
2946
		dout("flush_dirty_caps %p\n", inode);
2947
		spin_unlock(&mdsc->cap_dirty_lock);
2948 2949
		ceph_check_caps(ci, CHECK_CAPS_NODELAY|CHECK_CAPS_FLUSH, NULL);
		iput(inode);
2950 2951 2952
		spin_lock(&mdsc->cap_dirty_lock);
	}
	spin_unlock(&mdsc->cap_dirty_lock);
S
Sage Weil 已提交
2953
	dout("flush_dirty_caps done\n");
2954 2955
}

S
Sage Weil 已提交
2956 2957 2958 2959 2960 2961 2962 2963 2964 2965
/*
 * Drop open file reference.  If we were the last open file,
 * we may need to release capabilities to the MDS (or schedule
 * their delayed release).
 */
void ceph_put_fmode(struct ceph_inode_info *ci, int fmode)
{
	struct inode *inode = &ci->vfs_inode;
	int last = 0;

2966
	spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2967 2968 2969 2970 2971
	dout("put_fmode %p fmode %d %d -> %d\n", inode, fmode,
	     ci->i_nr_by_mode[fmode], ci->i_nr_by_mode[fmode]-1);
	BUG_ON(ci->i_nr_by_mode[fmode] == 0);
	if (--ci->i_nr_by_mode[fmode] == 0)
		last++;
2972
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991

	if (last && ci->i_vino.snap == CEPH_NOSNAP)
		ceph_check_caps(ci, 0, NULL);
}

/*
 * Helpers for embedding cap and dentry lease releases into mds
 * requests.
 *
 * @force is used by dentry_release (below) to force inclusion of a
 * record for the directory inode, even when there aren't any caps to
 * drop.
 */
int ceph_encode_inode_release(void **p, struct inode *inode,
			      int mds, int drop, int unless, int force)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	struct ceph_cap *cap;
	struct ceph_mds_request_release *rel = *p;
2992
	int used, dirty;
S
Sage Weil 已提交
2993 2994
	int ret = 0;

2995
	spin_lock(&ci->i_ceph_lock);
2996
	used = __ceph_caps_used(ci);
2997
	dirty = __ceph_caps_dirty(ci);
2998

2999 3000
	dout("encode_inode_release %p mds%d used|dirty %s drop %s unless %s\n",
	     inode, mds, ceph_cap_string(used|dirty), ceph_cap_string(drop),
3001 3002
	     ceph_cap_string(unless));

3003 3004
	/* only drop unused, clean caps */
	drop &= ~(used | dirty);
3005

S
Sage Weil 已提交
3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049
	cap = __get_cap_for_mds(ci, mds);
	if (cap && __cap_is_valid(cap)) {
		if (force ||
		    ((cap->issued & drop) &&
		     (cap->issued & unless) == 0)) {
			if ((cap->issued & drop) &&
			    (cap->issued & unless) == 0) {
				dout("encode_inode_release %p cap %p %s -> "
				     "%s\n", inode, cap,
				     ceph_cap_string(cap->issued),
				     ceph_cap_string(cap->issued & ~drop));
				cap->issued &= ~drop;
				cap->implemented &= ~drop;
				if (ci->i_ceph_flags & CEPH_I_NODELAY) {
					int wanted = __ceph_caps_wanted(ci);
					dout("  wanted %s -> %s (act %s)\n",
					     ceph_cap_string(cap->mds_wanted),
					     ceph_cap_string(cap->mds_wanted &
							     ~wanted),
					     ceph_cap_string(wanted));
					cap->mds_wanted &= wanted;
				}
			} else {
				dout("encode_inode_release %p cap %p %s"
				     " (force)\n", inode, cap,
				     ceph_cap_string(cap->issued));
			}

			rel->ino = cpu_to_le64(ceph_ino(inode));
			rel->cap_id = cpu_to_le64(cap->cap_id);
			rel->seq = cpu_to_le32(cap->seq);
			rel->issue_seq = cpu_to_le32(cap->issue_seq),
			rel->mseq = cpu_to_le32(cap->mseq);
			rel->caps = cpu_to_le32(cap->issued);
			rel->wanted = cpu_to_le32(cap->mds_wanted);
			rel->dname_len = 0;
			rel->dname_seq = 0;
			*p += sizeof(*rel);
			ret = 1;
		} else {
			dout("encode_inode_release %p cap %p %s\n",
			     inode, cap, ceph_cap_string(cap->issued));
		}
	}
3050
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064
	return ret;
}

int ceph_encode_dentry_release(void **p, struct dentry *dentry,
			       int mds, int drop, int unless)
{
	struct inode *dir = dentry->d_parent->d_inode;
	struct ceph_mds_request_release *rel = *p;
	struct ceph_dentry_info *di = ceph_dentry(dentry);
	int force = 0;
	int ret;

	/*
	 * force an record for the directory caps if we have a dentry lease.
3065
	 * this is racy (can't take i_ceph_lock and d_lock together), but it
S
Sage Weil 已提交
3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083
	 * doesn't have to be perfect; the mds will revoke anything we don't
	 * release.
	 */
	spin_lock(&dentry->d_lock);
	if (di->lease_session && di->lease_session->s_mds == mds)
		force = 1;
	spin_unlock(&dentry->d_lock);

	ret = ceph_encode_inode_release(p, dir, mds, drop, unless, force);

	spin_lock(&dentry->d_lock);
	if (ret && di->lease_session && di->lease_session->s_mds == mds) {
		dout("encode_dentry_release %p mds%d seq %d\n",
		     dentry, mds, (int)di->lease_seq);
		rel->dname_len = cpu_to_le32(dentry->d_name.len);
		memcpy(*p, dentry->d_name.name, dentry->d_name.len);
		*p += dentry->d_name.len;
		rel->dname_seq = cpu_to_le32(di->lease_seq);
S
Sage Weil 已提交
3084
		__ceph_mdsc_drop_dentry_lease(dentry);
S
Sage Weil 已提交
3085 3086 3087 3088
	}
	spin_unlock(&dentry->d_lock);
	return ret;
}