osd_client.c 77.7 KB
Newer Older
1

2
#include <linux/ceph/ceph_debug.h>
S
Sage Weil 已提交
3

4
#include <linux/module.h>
S
Sage Weil 已提交
5 6 7 8 9 10
#include <linux/err.h>
#include <linux/highmem.h>
#include <linux/mm.h>
#include <linux/pagemap.h>
#include <linux/slab.h>
#include <linux/uaccess.h>
11 12 13
#ifdef CONFIG_BLOCK
#include <linux/bio.h>
#endif
S
Sage Weil 已提交
14

15 16 17 18 19 20
#include <linux/ceph/libceph.h>
#include <linux/ceph/osd_client.h>
#include <linux/ceph/messenger.h>
#include <linux/ceph/decode.h>
#include <linux/ceph/auth.h>
#include <linux/ceph/pagelist.h>
S
Sage Weil 已提交
21

S
Sage Weil 已提交
22 23
#define OSD_OP_FRONT_LEN	4096
#define OSD_OPREPLY_FRONT_LEN	512
24

25 26
static struct kmem_cache	*ceph_osd_request_cache;

27
static const struct ceph_connection_operations osd_con_ops;
S
Sage Weil 已提交
28

A
Alex Elder 已提交
29
static void __send_queued(struct ceph_osd_client *osdc);
30
static int __reset_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd);
31 32
static void __register_request(struct ceph_osd_client *osdc,
			       struct ceph_osd_request *req);
33 34
static void __unregister_request(struct ceph_osd_client *osdc,
				 struct ceph_osd_request *req);
35 36
static void __unregister_linger_request(struct ceph_osd_client *osdc,
					struct ceph_osd_request *req);
37
static void __enqueue_request(struct ceph_osd_request *req);
38 39
static void __send_request(struct ceph_osd_client *osdc,
			   struct ceph_osd_request *req);
S
Sage Weil 已提交
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

/*
 * Implement client access to distributed object storage cluster.
 *
 * All data objects are stored within a cluster/cloud of OSDs, or
 * "object storage devices."  (Note that Ceph OSDs have _nothing_ to
 * do with the T10 OSD extensions to SCSI.)  Ceph OSDs are simply
 * remote daemons serving up and coordinating consistent and safe
 * access to storage.
 *
 * Cluster membership and the mapping of data objects onto storage devices
 * are described by the osd map.
 *
 * We keep track of pending OSD requests (read, write), resubmit
 * requests to different OSDs when the cluster topology/data layout
 * change, or retry the affected requests when the communications
 * channel with an OSD is reset.
 */

/*
 * calculate the mapping of a file extent onto an object, and fill out the
 * request accordingly.  shorten extent as necessary if it crosses an
 * object boundary.
 *
 * fill osd op in request message.
 */
66
static int calc_layout(struct ceph_file_layout *layout, u64 off, u64 *plen,
67
			u64 *objnum, u64 *objoff, u64 *objlen)
S
Sage Weil 已提交
68
{
69
	u64 orig_len = *plen;
S
Sage Weil 已提交
70
	int r;
S
Sage Weil 已提交
71

72
	/* object extent? */
73 74
	r = ceph_calc_file_object_mapping(layout, off, orig_len, objnum,
					  objoff, objlen);
S
Sage Weil 已提交
75 76
	if (r < 0)
		return r;
77 78
	if (*objlen < orig_len) {
		*plen = *objlen;
79 80 81 82
		dout(" skipping last %llu, final file extent %llu~%llu\n",
		     orig_len - *plen, off, *plen);
	}

83
	dout("calc_layout objnum=%llx %llu~%llu\n", *objnum, *objoff, *objlen);
S
Sage Weil 已提交
84

85
	return 0;
S
Sage Weil 已提交
86 87
}

A
Alex Elder 已提交
88 89 90 91 92 93
static void ceph_osd_data_init(struct ceph_osd_data *osd_data)
{
	memset(osd_data, 0, sizeof (*osd_data));
	osd_data->type = CEPH_OSD_DATA_TYPE_NONE;
}

94
static void ceph_osd_data_pages_init(struct ceph_osd_data *osd_data,
95 96 97 98 99 100 101 102 103 104 105
			struct page **pages, u64 length, u32 alignment,
			bool pages_from_pool, bool own_pages)
{
	osd_data->type = CEPH_OSD_DATA_TYPE_PAGES;
	osd_data->pages = pages;
	osd_data->length = length;
	osd_data->alignment = alignment;
	osd_data->pages_from_pool = pages_from_pool;
	osd_data->own_pages = own_pages;
}

106
static void ceph_osd_data_pagelist_init(struct ceph_osd_data *osd_data,
107 108 109 110 111 112 113
			struct ceph_pagelist *pagelist)
{
	osd_data->type = CEPH_OSD_DATA_TYPE_PAGELIST;
	osd_data->pagelist = pagelist;
}

#ifdef CONFIG_BLOCK
114
static void ceph_osd_data_bio_init(struct ceph_osd_data *osd_data,
115 116 117 118 119 120 121 122
			struct bio *bio, size_t bio_length)
{
	osd_data->type = CEPH_OSD_DATA_TYPE_BIO;
	osd_data->bio = bio;
	osd_data->bio_length = bio_length;
}
#endif /* CONFIG_BLOCK */

123 124 125 126 127 128
#define osd_req_op_data(oreq, whch, typ, fld)	\
	({						\
		BUG_ON(whch >= (oreq)->r_num_ops);	\
		&(oreq)->r_ops[whch].typ.fld;		\
	})

A
Alex Elder 已提交
129 130 131 132 133 134 135 136
static struct ceph_osd_data *
osd_req_op_raw_data_in(struct ceph_osd_request *osd_req, unsigned int which)
{
	BUG_ON(which >= osd_req->r_num_ops);

	return &osd_req->r_ops[which].raw_data_in;
}

137 138
struct ceph_osd_data *
osd_req_op_extent_osd_data(struct ceph_osd_request *osd_req,
139
			unsigned int which)
140
{
141
	return osd_req_op_data(osd_req, which, extent, osd_data);
142 143 144 145 146 147 148
}
EXPORT_SYMBOL(osd_req_op_extent_osd_data);

struct ceph_osd_data *
osd_req_op_cls_response_data(struct ceph_osd_request *osd_req,
			unsigned int which)
{
149
	return osd_req_op_data(osd_req, which, cls, response_data);
150 151 152
}
EXPORT_SYMBOL(osd_req_op_cls_response_data);	/* ??? */

A
Alex Elder 已提交
153 154 155 156 157 158 159 160 161 162 163 164 165
void osd_req_op_raw_data_in_pages(struct ceph_osd_request *osd_req,
			unsigned int which, struct page **pages,
			u64 length, u32 alignment,
			bool pages_from_pool, bool own_pages)
{
	struct ceph_osd_data *osd_data;

	osd_data = osd_req_op_raw_data_in(osd_req, which);
	ceph_osd_data_pages_init(osd_data, pages, length, alignment,
				pages_from_pool, own_pages);
}
EXPORT_SYMBOL(osd_req_op_raw_data_in_pages);

166
void osd_req_op_extent_osd_data_pages(struct ceph_osd_request *osd_req,
167 168
			unsigned int which, struct page **pages,
			u64 length, u32 alignment,
169 170 171 172
			bool pages_from_pool, bool own_pages)
{
	struct ceph_osd_data *osd_data;

173
	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
174 175 176 177 178 179
	ceph_osd_data_pages_init(osd_data, pages, length, alignment,
				pages_from_pool, own_pages);
}
EXPORT_SYMBOL(osd_req_op_extent_osd_data_pages);

void osd_req_op_extent_osd_data_pagelist(struct ceph_osd_request *osd_req,
180
			unsigned int which, struct ceph_pagelist *pagelist)
181 182 183
{
	struct ceph_osd_data *osd_data;

184
	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
185 186 187 188 189 190
	ceph_osd_data_pagelist_init(osd_data, pagelist);
}
EXPORT_SYMBOL(osd_req_op_extent_osd_data_pagelist);

#ifdef CONFIG_BLOCK
void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *osd_req,
191
			unsigned int which, struct bio *bio, size_t bio_length)
192 193
{
	struct ceph_osd_data *osd_data;
194 195

	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
196 197 198 199 200 201 202 203 204 205 206
	ceph_osd_data_bio_init(osd_data, bio, bio_length);
}
EXPORT_SYMBOL(osd_req_op_extent_osd_data_bio);
#endif /* CONFIG_BLOCK */

static void osd_req_op_cls_request_info_pagelist(
			struct ceph_osd_request *osd_req,
			unsigned int which, struct ceph_pagelist *pagelist)
{
	struct ceph_osd_data *osd_data;

207
	osd_data = osd_req_op_data(osd_req, which, cls, request_info);
208 209 210
	ceph_osd_data_pagelist_init(osd_data, pagelist);
}

211 212 213 214 215 216
void osd_req_op_cls_request_data_pagelist(
			struct ceph_osd_request *osd_req,
			unsigned int which, struct ceph_pagelist *pagelist)
{
	struct ceph_osd_data *osd_data;

217
	osd_data = osd_req_op_data(osd_req, which, cls, request_data);
218 219 220 221
	ceph_osd_data_pagelist_init(osd_data, pagelist);
}
EXPORT_SYMBOL(osd_req_op_cls_request_data_pagelist);

222 223 224 225 226 227 228 229 230 231 232 233
void osd_req_op_cls_request_data_pages(struct ceph_osd_request *osd_req,
			unsigned int which, struct page **pages, u64 length,
			u32 alignment, bool pages_from_pool, bool own_pages)
{
	struct ceph_osd_data *osd_data;

	osd_data = osd_req_op_data(osd_req, which, cls, request_data);
	ceph_osd_data_pages_init(osd_data, pages, length, alignment,
				pages_from_pool, own_pages);
}
EXPORT_SYMBOL(osd_req_op_cls_request_data_pages);

234 235 236 237 238 239
void osd_req_op_cls_response_data_pages(struct ceph_osd_request *osd_req,
			unsigned int which, struct page **pages, u64 length,
			u32 alignment, bool pages_from_pool, bool own_pages)
{
	struct ceph_osd_data *osd_data;

240
	osd_data = osd_req_op_data(osd_req, which, cls, response_data);
241 242 243 244 245
	ceph_osd_data_pages_init(osd_data, pages, length, alignment,
				pages_from_pool, own_pages);
}
EXPORT_SYMBOL(osd_req_op_cls_response_data_pages);

246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264
static u64 ceph_osd_data_length(struct ceph_osd_data *osd_data)
{
	switch (osd_data->type) {
	case CEPH_OSD_DATA_TYPE_NONE:
		return 0;
	case CEPH_OSD_DATA_TYPE_PAGES:
		return osd_data->length;
	case CEPH_OSD_DATA_TYPE_PAGELIST:
		return (u64)osd_data->pagelist->length;
#ifdef CONFIG_BLOCK
	case CEPH_OSD_DATA_TYPE_BIO:
		return (u64)osd_data->bio_length;
#endif /* CONFIG_BLOCK */
	default:
		WARN(true, "unrecognized data type %d\n", (int)osd_data->type);
		return 0;
	}
}

A
Alex Elder 已提交
265 266
static void ceph_osd_data_release(struct ceph_osd_data *osd_data)
{
267
	if (osd_data->type == CEPH_OSD_DATA_TYPE_PAGES && osd_data->own_pages) {
A
Alex Elder 已提交
268 269 270 271 272 273
		int num_pages;

		num_pages = calc_pages_for((u64)osd_data->alignment,
						(u64)osd_data->length);
		ceph_release_page_vector(osd_data->pages, num_pages);
	}
274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291
	ceph_osd_data_init(osd_data);
}

static void osd_req_op_data_release(struct ceph_osd_request *osd_req,
			unsigned int which)
{
	struct ceph_osd_req_op *op;

	BUG_ON(which >= osd_req->r_num_ops);
	op = &osd_req->r_ops[which];

	switch (op->op) {
	case CEPH_OSD_OP_READ:
	case CEPH_OSD_OP_WRITE:
		ceph_osd_data_release(&op->extent.osd_data);
		break;
	case CEPH_OSD_OP_CALL:
		ceph_osd_data_release(&op->cls.request_info);
292
		ceph_osd_data_release(&op->cls.request_data);
293 294 295 296 297
		ceph_osd_data_release(&op->cls.response_data);
		break;
	default:
		break;
	}
A
Alex Elder 已提交
298 299
}

S
Sage Weil 已提交
300 301 302
/*
 * requests
 */
303
static void ceph_osdc_release_request(struct kref *kref)
S
Sage Weil 已提交
304
{
305 306
	struct ceph_osd_request *req = container_of(kref,
					    struct ceph_osd_request, r_kref);
307
	unsigned int which;
S
Sage Weil 已提交
308

309 310
	dout("%s %p (r_request %p r_reply %p)\n", __func__, req,
	     req->r_request, req->r_reply);
311 312 313 314 315 316
	WARN_ON(!RB_EMPTY_NODE(&req->r_node));
	WARN_ON(!list_empty(&req->r_req_lru_item));
	WARN_ON(!list_empty(&req->r_osd_item));
	WARN_ON(!list_empty(&req->r_linger_item));
	WARN_ON(!list_empty(&req->r_linger_osd_item));
	WARN_ON(req->r_osd);
317

S
Sage Weil 已提交
318 319
	if (req->r_request)
		ceph_msg_put(req->r_request);
320
	if (req->r_reply) {
321
		ceph_msg_revoke_incoming(req->r_reply);
322
		ceph_msg_put(req->r_reply);
323
	}
324

325 326
	for (which = 0; which < req->r_num_ops; which++)
		osd_req_op_data_release(req, which);
327

S
Sage Weil 已提交
328 329 330 331
	ceph_put_snap_context(req->r_snapc);
	if (req->r_mempool)
		mempool_free(req, req->r_osdc->req_mempool);
	else
332 333
		kmem_cache_free(ceph_osd_request_cache, req);

S
Sage Weil 已提交
334
}
335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350

void ceph_osdc_get_request(struct ceph_osd_request *req)
{
	dout("%s %p (was %d)\n", __func__, req,
	     atomic_read(&req->r_kref.refcount));
	kref_get(&req->r_kref);
}
EXPORT_SYMBOL(ceph_osdc_get_request);

void ceph_osdc_put_request(struct ceph_osd_request *req)
{
	dout("%s %p (was %d)\n", __func__, req,
	     atomic_read(&req->r_kref.refcount));
	kref_put(&req->r_kref, ceph_osdc_release_request);
}
EXPORT_SYMBOL(ceph_osdc_put_request);
351

352
struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc,
S
Sage Weil 已提交
353
					       struct ceph_snap_context *snapc,
354
					       unsigned int num_ops,
355
					       bool use_mempool,
356
					       gfp_t gfp_flags)
S
Sage Weil 已提交
357 358 359
{
	struct ceph_osd_request *req;
	struct ceph_msg *msg;
360 361
	size_t msg_size;

362 363 364
	BUILD_BUG_ON(CEPH_OSD_MAX_OP > U16_MAX);
	BUG_ON(num_ops > CEPH_OSD_MAX_OP);

365 366 367
	msg_size = 4 + 4 + 8 + 8 + 4+8;
	msg_size += 2 + 4 + 8 + 4 + 4; /* oloc */
	msg_size += 1 + 8 + 4 + 4;     /* pg_t */
368
	msg_size += 4 + CEPH_MAX_OID_NAME_LEN; /* oid */
369 370 371 372 373
	msg_size += 2 + num_ops*sizeof(struct ceph_osd_op);
	msg_size += 8;  /* snapid */
	msg_size += 8;  /* snap_seq */
	msg_size += 8 * (snapc ? snapc->num_snaps : 0);  /* snaps */
	msg_size += 4;
S
Sage Weil 已提交
374 375

	if (use_mempool) {
376
		req = mempool_alloc(osdc->req_mempool, gfp_flags);
S
Sage Weil 已提交
377 378
		memset(req, 0, sizeof(*req));
	} else {
379
		req = kmem_cache_zalloc(ceph_osd_request_cache, gfp_flags);
S
Sage Weil 已提交
380 381
	}
	if (req == NULL)
382
		return NULL;
S
Sage Weil 已提交
383 384 385

	req->r_osdc = osdc;
	req->r_mempool = use_mempool;
386
	req->r_num_ops = num_ops;
387

S
Sage Weil 已提交
388
	kref_init(&req->r_kref);
S
Sage Weil 已提交
389 390
	init_completion(&req->r_completion);
	init_completion(&req->r_safe_completion);
391
	RB_CLEAR_NODE(&req->r_node);
S
Sage Weil 已提交
392
	INIT_LIST_HEAD(&req->r_unsafe_item);
393
	INIT_LIST_HEAD(&req->r_linger_item);
394
	INIT_LIST_HEAD(&req->r_linger_osd_item);
395
	INIT_LIST_HEAD(&req->r_req_lru_item);
396 397
	INIT_LIST_HEAD(&req->r_osd_item);

398
	req->r_base_oloc.pool = -1;
399
	req->r_target_oloc.pool = -1;
400

S
Sage Weil 已提交
401 402 403 404 405
	/* create reply message */
	if (use_mempool)
		msg = ceph_msgpool_get(&osdc->msgpool_op_reply, 0);
	else
		msg = ceph_msg_new(CEPH_MSG_OSD_OPREPLY,
406
				   OSD_OPREPLY_FRONT_LEN, gfp_flags, true);
407
	if (!msg) {
S
Sage Weil 已提交
408
		ceph_osdc_put_request(req);
409
		return NULL;
S
Sage Weil 已提交
410 411 412 413
	}
	req->r_reply = msg;

	/* create request message; allow space for oid */
S
Sage Weil 已提交
414
	if (use_mempool)
415
		msg = ceph_msgpool_get(&osdc->msgpool_op, 0);
S
Sage Weil 已提交
416
	else
417
		msg = ceph_msg_new(CEPH_MSG_OSD_OP, msg_size, gfp_flags, true);
418
	if (!msg) {
S
Sage Weil 已提交
419
		ceph_osdc_put_request(req);
420
		return NULL;
S
Sage Weil 已提交
421
	}
422

S
Sage Weil 已提交
423
	memset(msg->front.iov_base, 0, msg->front.iov_len);
424 425 426 427 428

	req->r_request = msg;

	return req;
}
429
EXPORT_SYMBOL(ceph_osdc_alloc_request);
430

431
static bool osd_req_opcode_valid(u16 opcode)
432
{
433
	switch (opcode) {
434
	case CEPH_OSD_OP_READ:
435
	case CEPH_OSD_OP_STAT:
436 437 438
	case CEPH_OSD_OP_MAPEXT:
	case CEPH_OSD_OP_MASKTRUNC:
	case CEPH_OSD_OP_SPARSE_READ:
439
	case CEPH_OSD_OP_NOTIFY:
440
	case CEPH_OSD_OP_NOTIFY_ACK:
441
	case CEPH_OSD_OP_ASSERT_VER:
442
	case CEPH_OSD_OP_WRITE:
443 444 445 446 447
	case CEPH_OSD_OP_WRITEFULL:
	case CEPH_OSD_OP_TRUNCATE:
	case CEPH_OSD_OP_ZERO:
	case CEPH_OSD_OP_DELETE:
	case CEPH_OSD_OP_APPEND:
448
	case CEPH_OSD_OP_STARTSYNC:
449 450 451 452 453 454
	case CEPH_OSD_OP_SETTRUNC:
	case CEPH_OSD_OP_TRIMTRUNC:
	case CEPH_OSD_OP_TMAPUP:
	case CEPH_OSD_OP_TMAPPUT:
	case CEPH_OSD_OP_TMAPGET:
	case CEPH_OSD_OP_CREATE:
455
	case CEPH_OSD_OP_ROLLBACK:
456
	case CEPH_OSD_OP_WATCH:
457 458 459 460 461 462 463 464 465
	case CEPH_OSD_OP_OMAPGETKEYS:
	case CEPH_OSD_OP_OMAPGETVALS:
	case CEPH_OSD_OP_OMAPGETHEADER:
	case CEPH_OSD_OP_OMAPGETVALSBYKEYS:
	case CEPH_OSD_OP_OMAPSETVALS:
	case CEPH_OSD_OP_OMAPSETHEADER:
	case CEPH_OSD_OP_OMAPCLEAR:
	case CEPH_OSD_OP_OMAPRMKEYS:
	case CEPH_OSD_OP_OMAP_CMP:
466
	case CEPH_OSD_OP_SETALLOCHINT:
467 468 469
	case CEPH_OSD_OP_CLONERANGE:
	case CEPH_OSD_OP_ASSERT_SRC_VERSION:
	case CEPH_OSD_OP_SRC_CMPXATTR:
470
	case CEPH_OSD_OP_GETXATTR:
471
	case CEPH_OSD_OP_GETXATTRS:
472 473
	case CEPH_OSD_OP_CMPXATTR:
	case CEPH_OSD_OP_SETXATTR:
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491
	case CEPH_OSD_OP_SETXATTRS:
	case CEPH_OSD_OP_RESETXATTRS:
	case CEPH_OSD_OP_RMXATTR:
	case CEPH_OSD_OP_PULL:
	case CEPH_OSD_OP_PUSH:
	case CEPH_OSD_OP_BALANCEREADS:
	case CEPH_OSD_OP_UNBALANCEREADS:
	case CEPH_OSD_OP_SCRUB:
	case CEPH_OSD_OP_SCRUB_RESERVE:
	case CEPH_OSD_OP_SCRUB_UNRESERVE:
	case CEPH_OSD_OP_SCRUB_STOP:
	case CEPH_OSD_OP_SCRUB_MAP:
	case CEPH_OSD_OP_WRLOCK:
	case CEPH_OSD_OP_WRUNLOCK:
	case CEPH_OSD_OP_RDLOCK:
	case CEPH_OSD_OP_RDUNLOCK:
	case CEPH_OSD_OP_UPLOCK:
	case CEPH_OSD_OP_DNLOCK:
492
	case CEPH_OSD_OP_CALL:
493 494
	case CEPH_OSD_OP_PGLS:
	case CEPH_OSD_OP_PGLS_FILTER:
495 496 497 498 499 500
		return true;
	default:
		return false;
	}
}

501 502 503 504 505
/*
 * This is an osd op init function for opcodes that have no data or
 * other information associated with them.  It also serves as a
 * common init routine for all the other init functions, below.
 */
506
static struct ceph_osd_req_op *
A
Alex Elder 已提交
507
_osd_req_op_init(struct ceph_osd_request *osd_req, unsigned int which,
508
				u16 opcode)
509
{
510 511 512
	struct ceph_osd_req_op *op;

	BUG_ON(which >= osd_req->r_num_ops);
513 514
	BUG_ON(!osd_req_opcode_valid(opcode));

515
	op = &osd_req->r_ops[which];
516 517
	memset(op, 0, sizeof (*op));
	op->op = opcode;
518 519

	return op;
520 521
}

A
Alex Elder 已提交
522 523 524 525 526 527 528
void osd_req_op_init(struct ceph_osd_request *osd_req,
				unsigned int which, u16 opcode)
{
	(void)_osd_req_op_init(osd_req, which, opcode);
}
EXPORT_SYMBOL(osd_req_op_init);

529 530
void osd_req_op_extent_init(struct ceph_osd_request *osd_req,
				unsigned int which, u16 opcode,
531 532 533
				u64 offset, u64 length,
				u64 truncate_size, u32 truncate_seq)
{
A
Alex Elder 已提交
534
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
535 536
	size_t payload_len = 0;

L
Li Wang 已提交
537 538 539
	BUG_ON(opcode != CEPH_OSD_OP_READ && opcode != CEPH_OSD_OP_WRITE &&
	       opcode != CEPH_OSD_OP_DELETE && opcode != CEPH_OSD_OP_ZERO &&
	       opcode != CEPH_OSD_OP_TRUNCATE);
540 541 542 543 544 545 546 547 548 549 550 551

	op->extent.offset = offset;
	op->extent.length = length;
	op->extent.truncate_size = truncate_size;
	op->extent.truncate_seq = truncate_seq;
	if (opcode == CEPH_OSD_OP_WRITE)
		payload_len += length;

	op->payload_len = payload_len;
}
EXPORT_SYMBOL(osd_req_op_extent_init);

552 553
void osd_req_op_extent_update(struct ceph_osd_request *osd_req,
				unsigned int which, u64 length)
554
{
555 556 557 558 559 560
	struct ceph_osd_req_op *op;
	u64 previous;

	BUG_ON(which >= osd_req->r_num_ops);
	op = &osd_req->r_ops[which];
	previous = op->extent.length;
561 562 563 564 565 566 567 568 569 570

	if (length == previous)
		return;		/* Nothing to do */
	BUG_ON(length > previous);

	op->extent.length = length;
	op->payload_len -= previous - length;
}
EXPORT_SYMBOL(osd_req_op_extent_update);

571
void osd_req_op_cls_init(struct ceph_osd_request *osd_req, unsigned int which,
572
			u16 opcode, const char *class, const char *method)
573
{
A
Alex Elder 已提交
574
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
575
	struct ceph_pagelist *pagelist;
576 577 578 579 580
	size_t payload_len = 0;
	size_t size;

	BUG_ON(opcode != CEPH_OSD_OP_CALL);

581 582 583 584
	pagelist = kmalloc(sizeof (*pagelist), GFP_NOFS);
	BUG_ON(!pagelist);
	ceph_pagelist_init(pagelist);

585 586 587 588
	op->cls.class_name = class;
	size = strlen(class);
	BUG_ON(size > (size_t) U8_MAX);
	op->cls.class_len = size;
589
	ceph_pagelist_append(pagelist, class, size);
590 591 592 593 594 595
	payload_len += size;

	op->cls.method_name = method;
	size = strlen(method);
	BUG_ON(size > (size_t) U8_MAX);
	op->cls.method_len = size;
596
	ceph_pagelist_append(pagelist, method, size);
597 598
	payload_len += size;

599
	osd_req_op_cls_request_info_pagelist(osd_req, which, pagelist);
600

601 602 603 604 605
	op->cls.argc = 0;	/* currently unused */

	op->payload_len = payload_len;
}
EXPORT_SYMBOL(osd_req_op_cls_init);
606

607 608
void osd_req_op_watch_init(struct ceph_osd_request *osd_req,
				unsigned int which, u16 opcode,
609 610
				u64 cookie, u64 version, int flag)
{
A
Alex Elder 已提交
611
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
612

613
	BUG_ON(opcode != CEPH_OSD_OP_NOTIFY_ACK && opcode != CEPH_OSD_OP_WATCH);
614 615

	op->watch.cookie = cookie;
A
Alex Elder 已提交
616
	op->watch.ver = version;
617
	if (opcode == CEPH_OSD_OP_WATCH && flag)
618
		op->watch.flag = (u8)1;
619 620 621
}
EXPORT_SYMBOL(osd_req_op_watch_init);

622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641
void osd_req_op_alloc_hint_init(struct ceph_osd_request *osd_req,
				unsigned int which,
				u64 expected_object_size,
				u64 expected_write_size)
{
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
						      CEPH_OSD_OP_SETALLOCHINT);

	op->alloc_hint.expected_object_size = expected_object_size;
	op->alloc_hint.expected_write_size = expected_write_size;

	/*
	 * CEPH_OSD_OP_SETALLOCHINT op is advisory and therefore deemed
	 * not worth a feature bit.  Set FAILOK per-op flag to make
	 * sure older osds don't trip over an unsupported opcode.
	 */
	op->flags |= CEPH_OSD_OP_FLAG_FAILOK;
}
EXPORT_SYMBOL(osd_req_op_alloc_hint_init);

642
static void ceph_osdc_msg_data_add(struct ceph_msg *msg,
643 644 645 646 647 648 649
				struct ceph_osd_data *osd_data)
{
	u64 length = ceph_osd_data_length(osd_data);

	if (osd_data->type == CEPH_OSD_DATA_TYPE_PAGES) {
		BUG_ON(length > (u64) SIZE_MAX);
		if (length)
650
			ceph_msg_data_add_pages(msg, osd_data->pages,
651 652 653
					length, osd_data->alignment);
	} else if (osd_data->type == CEPH_OSD_DATA_TYPE_PAGELIST) {
		BUG_ON(!length);
654
		ceph_msg_data_add_pagelist(msg, osd_data->pagelist);
655 656
#ifdef CONFIG_BLOCK
	} else if (osd_data->type == CEPH_OSD_DATA_TYPE_BIO) {
657
		ceph_msg_data_add_bio(msg, osd_data->bio, length);
658 659 660 661 662 663
#endif
	} else {
		BUG_ON(osd_data->type != CEPH_OSD_DATA_TYPE_NONE);
	}
}

664
static u64 osd_req_encode_op(struct ceph_osd_request *req,
665
			      struct ceph_osd_op *dst, unsigned int which)
666
{
667
	struct ceph_osd_req_op *src;
668
	struct ceph_osd_data *osd_data;
669
	u64 request_data_len = 0;
670
	u64 data_length;
671

672 673
	BUG_ON(which >= req->r_num_ops);
	src = &req->r_ops[which];
674 675 676 677 678 679 680 681
	if (WARN_ON(!osd_req_opcode_valid(src->op))) {
		pr_err("unrecognized osd opcode %d\n", src->op);

		return 0;
	}

	switch (src->op) {
	case CEPH_OSD_OP_STAT:
A
Alex Elder 已提交
682 683
		osd_data = &src->raw_data_in;
		ceph_osdc_msg_data_add(req->r_reply, osd_data);
684 685 686
		break;
	case CEPH_OSD_OP_READ:
	case CEPH_OSD_OP_WRITE:
L
Li Wang 已提交
687 688 689
	case CEPH_OSD_OP_ZERO:
	case CEPH_OSD_OP_DELETE:
	case CEPH_OSD_OP_TRUNCATE:
690
		if (src->op == CEPH_OSD_OP_WRITE)
691
			request_data_len = src->extent.length;
692 693 694 695 696 697
		dst->extent.offset = cpu_to_le64(src->extent.offset);
		dst->extent.length = cpu_to_le64(src->extent.length);
		dst->extent.truncate_size =
			cpu_to_le64(src->extent.truncate_size);
		dst->extent.truncate_seq =
			cpu_to_le32(src->extent.truncate_seq);
698
		osd_data = &src->extent.osd_data;
699
		if (src->op == CEPH_OSD_OP_WRITE)
700
			ceph_osdc_msg_data_add(req->r_request, osd_data);
701
		else
702
			ceph_osdc_msg_data_add(req->r_reply, osd_data);
703 704 705 706
		break;
	case CEPH_OSD_OP_CALL:
		dst->cls.class_len = src->cls.class_len;
		dst->cls.method_len = src->cls.method_len;
707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722
		osd_data = &src->cls.request_info;
		ceph_osdc_msg_data_add(req->r_request, osd_data);
		BUG_ON(osd_data->type != CEPH_OSD_DATA_TYPE_PAGELIST);
		request_data_len = osd_data->pagelist->length;

		osd_data = &src->cls.request_data;
		data_length = ceph_osd_data_length(osd_data);
		if (data_length) {
			BUG_ON(osd_data->type == CEPH_OSD_DATA_TYPE_NONE);
			dst->cls.indata_len = cpu_to_le32(data_length);
			ceph_osdc_msg_data_add(req->r_request, osd_data);
			src->payload_len += data_length;
			request_data_len += data_length;
		}
		osd_data = &src->cls.response_data;
		ceph_osdc_msg_data_add(req->r_reply, osd_data);
723 724 725 726 727 728 729 730 731
		break;
	case CEPH_OSD_OP_STARTSYNC:
		break;
	case CEPH_OSD_OP_NOTIFY_ACK:
	case CEPH_OSD_OP_WATCH:
		dst->watch.cookie = cpu_to_le64(src->watch.cookie);
		dst->watch.ver = cpu_to_le64(src->watch.ver);
		dst->watch.flag = src->watch.flag;
		break;
732 733 734 735 736 737
	case CEPH_OSD_OP_SETALLOCHINT:
		dst->alloc_hint.expected_object_size =
		    cpu_to_le64(src->alloc_hint.expected_object_size);
		dst->alloc_hint.expected_write_size =
		    cpu_to_le64(src->alloc_hint.expected_write_size);
		break;
738
	default:
739
		pr_err("unsupported osd opcode %s\n",
740
			ceph_osd_op_name(src->op));
741
		WARN_ON(1);
742 743

		return 0;
744
	}
745

746
	dst->op = cpu_to_le16(src->op);
747
	dst->flags = cpu_to_le32(src->flags);
748
	dst->payload_len = cpu_to_le32(src->payload_len);
749

750
	return request_data_len;
751 752
}

753 754 755 756 757 758 759 760 761 762 763 764 765 766
/*
 * build new request AND message, calculate layout, and adjust file
 * extent as needed.
 *
 * if the file was recently truncated, we include information about its
 * old and new size so that the object can be updated appropriately.  (we
 * avoid synchronously deleting truncated objects because it's slow.)
 *
 * if @do_sync, include a 'startsync' command so that the osd will flush
 * data quickly.
 */
struct ceph_osd_request *ceph_osdc_new_request(struct ceph_osd_client *osdc,
					       struct ceph_file_layout *layout,
					       struct ceph_vino vino,
767
					       u64 off, u64 *plen, int num_ops,
768 769 770 771
					       int opcode, int flags,
					       struct ceph_snap_context *snapc,
					       u32 truncate_seq,
					       u64 truncate_size,
772
					       bool use_mempool)
773
{
774
	struct ceph_osd_request *req;
775 776 777
	u64 objnum = 0;
	u64 objoff = 0;
	u64 objlen = 0;
778 779
	u32 object_size;
	u64 object_base;
780
	int r;
781

L
Li Wang 已提交
782 783 784
	BUG_ON(opcode != CEPH_OSD_OP_READ && opcode != CEPH_OSD_OP_WRITE &&
	       opcode != CEPH_OSD_OP_DELETE && opcode != CEPH_OSD_OP_ZERO &&
	       opcode != CEPH_OSD_OP_TRUNCATE);
785

786
	req = ceph_osdc_alloc_request(osdc, snapc, num_ops, use_mempool,
A
Alex Elder 已提交
787
					GFP_NOFS);
788
	if (!req)
789
		return ERR_PTR(-ENOMEM);
790

791
	req->r_flags = flags;
792 793

	/* calculate max write size */
794
	r = calc_layout(layout, off, plen, &objnum, &objoff, &objlen);
795 796
	if (r < 0) {
		ceph_osdc_put_request(req);
797
		return ERR_PTR(r);
798
	}
799

800 801
	object_size = le32_to_cpu(layout->fl_object_size);
	object_base = off - objoff;
802 803 804 805 806 807 808 809
	if (!(truncate_seq == 1 && truncate_size == -1ULL)) {
		if (truncate_size <= object_base) {
			truncate_size = 0;
		} else {
			truncate_size -= object_base;
			if (truncate_size > object_size)
				truncate_size = object_size;
		}
810
	}
811

812
	osd_req_op_extent_init(req, 0, opcode, objoff, objlen,
813
				truncate_size, truncate_seq);
814

815 816 817 818 819 820
	/*
	 * A second op in the ops array means the caller wants to
	 * also issue a include a 'startsync' command so that the
	 * osd will flush data quickly.
	 */
	if (num_ops > 1)
821
		osd_req_op_init(req, 1, CEPH_OSD_OP_STARTSYNC);
822

823
	req->r_base_oloc.pool = ceph_file_layout_pg_pool(*layout);
824

825
	snprintf(req->r_base_oid.name, sizeof(req->r_base_oid.name),
826
		 "%llx.%08llx", vino.ino, objnum);
827
	req->r_base_oid.name_len = strlen(req->r_base_oid.name);
828

S
Sage Weil 已提交
829 830
	return req;
}
831
EXPORT_SYMBOL(ceph_osdc_new_request);
S
Sage Weil 已提交
832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897

/*
 * We keep osd requests in an rbtree, sorted by ->r_tid.
 */
static void __insert_request(struct ceph_osd_client *osdc,
			     struct ceph_osd_request *new)
{
	struct rb_node **p = &osdc->requests.rb_node;
	struct rb_node *parent = NULL;
	struct ceph_osd_request *req = NULL;

	while (*p) {
		parent = *p;
		req = rb_entry(parent, struct ceph_osd_request, r_node);
		if (new->r_tid < req->r_tid)
			p = &(*p)->rb_left;
		else if (new->r_tid > req->r_tid)
			p = &(*p)->rb_right;
		else
			BUG();
	}

	rb_link_node(&new->r_node, parent, p);
	rb_insert_color(&new->r_node, &osdc->requests);
}

static struct ceph_osd_request *__lookup_request(struct ceph_osd_client *osdc,
						 u64 tid)
{
	struct ceph_osd_request *req;
	struct rb_node *n = osdc->requests.rb_node;

	while (n) {
		req = rb_entry(n, struct ceph_osd_request, r_node);
		if (tid < req->r_tid)
			n = n->rb_left;
		else if (tid > req->r_tid)
			n = n->rb_right;
		else
			return req;
	}
	return NULL;
}

static struct ceph_osd_request *
__lookup_request_ge(struct ceph_osd_client *osdc,
		    u64 tid)
{
	struct ceph_osd_request *req;
	struct rb_node *n = osdc->requests.rb_node;

	while (n) {
		req = rb_entry(n, struct ceph_osd_request, r_node);
		if (tid < req->r_tid) {
			if (!n->rb_left)
				return req;
			n = n->rb_left;
		} else if (tid > req->r_tid) {
			n = n->rb_right;
		} else {
			return req;
		}
	}
	return NULL;
}

898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928
static void __kick_linger_request(struct ceph_osd_request *req)
{
	struct ceph_osd_client *osdc = req->r_osdc;
	struct ceph_osd *osd = req->r_osd;

	/*
	 * Linger requests need to be resent with a new tid to avoid
	 * the dup op detection logic on the OSDs.  Achieve this with
	 * a re-register dance instead of open-coding.
	 */
	ceph_osdc_get_request(req);
	if (!list_empty(&req->r_linger_item))
		__unregister_linger_request(osdc, req);
	else
		__unregister_request(osdc, req);
	__register_request(osdc, req);
	ceph_osdc_put_request(req);

	/*
	 * Unless request has been registered as both normal and
	 * lingering, __unregister{,_linger}_request clears r_osd.
	 * However, here we need to preserve r_osd to make sure we
	 * requeue on the same OSD.
	 */
	WARN_ON(req->r_osd || !osd);
	req->r_osd = osd;

	dout("%s requeueing %p tid %llu\n", __func__, req, req->r_tid);
	__enqueue_request(req);
}

929 930 931 932 933 934
/*
 * Resubmit requests pending on the given osd.
 */
static void __kick_osd_requests(struct ceph_osd_client *osdc,
				struct ceph_osd *osd)
{
935
	struct ceph_osd_request *req, *nreq;
936
	LIST_HEAD(resend);
937
	LIST_HEAD(resend_linger);
938 939
	int err;

940
	dout("%s osd%d\n", __func__, osd->o_osd);
941
	err = __reset_osd(osdc, osd);
942
	if (err)
943
		return;
944

945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961
	/*
	 * Build up a list of requests to resend by traversing the
	 * osd's list of requests.  Requests for a given object are
	 * sent in tid order, and that is also the order they're
	 * kept on this list.  Therefore all requests that are in
	 * flight will be found first, followed by all requests that
	 * have not yet been sent.  And to resend requests while
	 * preserving this order we will want to put any sent
	 * requests back on the front of the osd client's unsent
	 * list.
	 *
	 * So we build a separate ordered list of already-sent
	 * requests for the affected osd and splice it onto the
	 * front of the osd client's unsent list.  Once we've seen a
	 * request that has not yet been sent we're done.  Those
	 * requests are already sitting right where they belong.
	 */
962
	list_for_each_entry(req, &osd->o_requests, r_osd_item) {
963 964
		if (!req->r_sent)
			break;
965 966 967 968 969

		if (!req->r_linger) {
			dout("%s requeueing %p tid %llu\n", __func__, req,
			     req->r_tid);
			list_move_tail(&req->r_req_lru_item, &resend);
970
			req->r_flags |= CEPH_OSD_FLAG_RETRY;
971 972 973
		} else {
			list_move_tail(&req->r_req_lru_item, &resend_linger);
		}
974
	}
975
	list_splice(&resend, &osdc->req_unsent);
976

977
	/*
978 979 980 981
	 * Both registered and not yet registered linger requests are
	 * enqueued with a new tid on the same OSD.  We add/move them
	 * to req_unsent/o_requests at the end to keep things in tid
	 * order.
982
	 */
983
	list_for_each_entry_safe(req, nreq, &osd->o_linger_requests,
984
				 r_linger_osd_item) {
985 986
		WARN_ON(!list_empty(&req->r_req_lru_item));
		__kick_linger_request(req);
987
	}
988 989 990

	list_for_each_entry_safe(req, nreq, &resend_linger, r_req_lru_item)
		__kick_linger_request(req);
991 992
}

S
Sage Weil 已提交
993
/*
994
 * If the osd connection drops, we need to resubmit all requests.
S
Sage Weil 已提交
995 996 997 998 999 1000 1001 1002 1003 1004 1005
 */
static void osd_reset(struct ceph_connection *con)
{
	struct ceph_osd *osd = con->private;
	struct ceph_osd_client *osdc;

	if (!osd)
		return;
	dout("osd_reset osd%d\n", osd->o_osd);
	osdc = osd->o_osdc;
	down_read(&osdc->map_sem);
S
Sage Weil 已提交
1006 1007
	mutex_lock(&osdc->request_mutex);
	__kick_osd_requests(osdc, osd);
A
Alex Elder 已提交
1008
	__send_queued(osdc);
S
Sage Weil 已提交
1009
	mutex_unlock(&osdc->request_mutex);
S
Sage Weil 已提交
1010 1011 1012 1013 1014 1015
	up_read(&osdc->map_sem);
}

/*
 * Track open sessions with osds.
 */
1016
static struct ceph_osd *create_osd(struct ceph_osd_client *osdc, int onum)
S
Sage Weil 已提交
1017 1018 1019 1020 1021 1022 1023 1024 1025
{
	struct ceph_osd *osd;

	osd = kzalloc(sizeof(*osd), GFP_NOFS);
	if (!osd)
		return NULL;

	atomic_set(&osd->o_ref, 1);
	osd->o_osdc = osdc;
1026
	osd->o_osd = onum;
1027
	RB_CLEAR_NODE(&osd->o_node);
S
Sage Weil 已提交
1028
	INIT_LIST_HEAD(&osd->o_requests);
1029
	INIT_LIST_HEAD(&osd->o_linger_requests);
1030
	INIT_LIST_HEAD(&osd->o_osd_lru);
S
Sage Weil 已提交
1031 1032
	osd->o_incarnation = 1;

1033
	ceph_con_init(&osd->o_con, osd, &osd_con_ops, &osdc->client->msgr);
1034

1035
	INIT_LIST_HEAD(&osd->o_keepalive_item);
S
Sage Weil 已提交
1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
	return osd;
}

static struct ceph_osd *get_osd(struct ceph_osd *osd)
{
	if (atomic_inc_not_zero(&osd->o_ref)) {
		dout("get_osd %p %d -> %d\n", osd, atomic_read(&osd->o_ref)-1,
		     atomic_read(&osd->o_ref));
		return osd;
	} else {
		dout("get_osd %p FAIL\n", osd);
		return NULL;
	}
}

static void put_osd(struct ceph_osd *osd)
{
	dout("put_osd %p %d -> %d\n", osd, atomic_read(&osd->o_ref),
	     atomic_read(&osd->o_ref) - 1);
1055
	if (atomic_dec_and_test(&osd->o_ref) && osd->o_auth.authorizer) {
S
Sage Weil 已提交
1056 1057
		struct ceph_auth_client *ac = osd->o_osdc->client->monc.auth;

1058
		ceph_auth_destroy_authorizer(ac, osd->o_auth.authorizer);
S
Sage Weil 已提交
1059
		kfree(osd);
S
Sage Weil 已提交
1060
	}
S
Sage Weil 已提交
1061 1062 1063 1064 1065
}

/*
 * remove an osd from our map
 */
1066
static void __remove_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
S
Sage Weil 已提交
1067
{
1068
	dout("__remove_osd %p\n", osd);
S
Sage Weil 已提交
1069
	BUG_ON(!list_empty(&osd->o_requests));
1070 1071
	BUG_ON(!list_empty(&osd->o_linger_requests));

S
Sage Weil 已提交
1072
	rb_erase(&osd->o_node, &osdc->osds);
1073
	list_del_init(&osd->o_osd_lru);
S
Sage Weil 已提交
1074 1075 1076 1077
	ceph_con_close(&osd->o_con);
	put_osd(osd);
}

1078 1079
static void remove_all_osds(struct ceph_osd_client *osdc)
{
1080
	dout("%s %p\n", __func__, osdc);
1081 1082 1083 1084 1085 1086 1087 1088 1089
	mutex_lock(&osdc->request_mutex);
	while (!RB_EMPTY_ROOT(&osdc->osds)) {
		struct ceph_osd *osd = rb_entry(rb_first(&osdc->osds),
						struct ceph_osd, o_node);
		__remove_osd(osdc, osd);
	}
	mutex_unlock(&osdc->request_mutex);
}

1090 1091 1092
static void __move_osd_to_lru(struct ceph_osd_client *osdc,
			      struct ceph_osd *osd)
{
1093
	dout("%s %p\n", __func__, osd);
1094
	BUG_ON(!list_empty(&osd->o_osd_lru));
1095

1096
	list_add_tail(&osd->o_osd_lru, &osdc->osd_lru);
1097
	osd->lru_ttl = jiffies + osdc->client->options->osd_idle_ttl * HZ;
1098 1099
}

1100 1101 1102 1103 1104 1105 1106 1107 1108 1109
static void maybe_move_osd_to_lru(struct ceph_osd_client *osdc,
				  struct ceph_osd *osd)
{
	dout("%s %p\n", __func__, osd);

	if (list_empty(&osd->o_requests) &&
	    list_empty(&osd->o_linger_requests))
		__move_osd_to_lru(osdc, osd);
}

1110 1111 1112 1113 1114 1115 1116
static void __remove_osd_from_lru(struct ceph_osd *osd)
{
	dout("__remove_osd_from_lru %p\n", osd);
	if (!list_empty(&osd->o_osd_lru))
		list_del_init(&osd->o_osd_lru);
}

1117
static void remove_old_osds(struct ceph_osd_client *osdc)
1118 1119 1120 1121 1122 1123
{
	struct ceph_osd *osd, *nosd;

	dout("__remove_old_osds %p\n", osdc);
	mutex_lock(&osdc->request_mutex);
	list_for_each_entry_safe(osd, nosd, &osdc->osd_lru, o_osd_lru) {
1124
		if (time_before(jiffies, osd->lru_ttl))
1125 1126 1127 1128 1129 1130
			break;
		__remove_osd(osdc, osd);
	}
	mutex_unlock(&osdc->request_mutex);
}

S
Sage Weil 已提交
1131 1132 1133
/*
 * reset osd connect
 */
1134
static int __reset_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
S
Sage Weil 已提交
1135
{
A
Alex Elder 已提交
1136
	struct ceph_entity_addr *peer_addr;
S
Sage Weil 已提交
1137

1138
	dout("__reset_osd %p osd%d\n", osd, osd->o_osd);
1139 1140
	if (list_empty(&osd->o_requests) &&
	    list_empty(&osd->o_linger_requests)) {
1141
		__remove_osd(osdc, osd);
A
Alex Elder 已提交
1142 1143 1144 1145 1146 1147 1148 1149 1150

		return -ENODEV;
	}

	peer_addr = &osdc->osdmap->osd_addr[osd->o_osd];
	if (!memcmp(peer_addr, &osd->o_con.peer_addr, sizeof (*peer_addr)) &&
			!ceph_con_opened(&osd->o_con)) {
		struct ceph_osd_request *req;

1151 1152
		dout("osd addr hasn't changed and connection never opened, "
		     "letting msgr retry\n");
1153 1154 1155
		/* touch each r_stamp for handle_timeout()'s benfit */
		list_for_each_entry(req, &osd->o_requests, r_osd_item)
			req->r_stamp = jiffies;
A
Alex Elder 已提交
1156 1157

		return -EAGAIN;
S
Sage Weil 已提交
1158
	}
A
Alex Elder 已提交
1159 1160 1161 1162 1163 1164

	ceph_con_close(&osd->o_con);
	ceph_con_open(&osd->o_con, CEPH_ENTITY_TYPE_OSD, osd->o_osd, peer_addr);
	osd->o_incarnation++;

	return 0;
S
Sage Weil 已提交
1165 1166 1167 1168 1169 1170 1171 1172
}

static void __insert_osd(struct ceph_osd_client *osdc, struct ceph_osd *new)
{
	struct rb_node **p = &osdc->osds.rb_node;
	struct rb_node *parent = NULL;
	struct ceph_osd *osd = NULL;

1173
	dout("__insert_osd %p osd%d\n", new, new->o_osd);
S
Sage Weil 已提交
1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205
	while (*p) {
		parent = *p;
		osd = rb_entry(parent, struct ceph_osd, o_node);
		if (new->o_osd < osd->o_osd)
			p = &(*p)->rb_left;
		else if (new->o_osd > osd->o_osd)
			p = &(*p)->rb_right;
		else
			BUG();
	}

	rb_link_node(&new->o_node, parent, p);
	rb_insert_color(&new->o_node, &osdc->osds);
}

static struct ceph_osd *__lookup_osd(struct ceph_osd_client *osdc, int o)
{
	struct ceph_osd *osd;
	struct rb_node *n = osdc->osds.rb_node;

	while (n) {
		osd = rb_entry(n, struct ceph_osd, o_node);
		if (o < osd->o_osd)
			n = n->rb_left;
		else if (o > osd->o_osd)
			n = n->rb_right;
		else
			return osd;
	}
	return NULL;
}

1206 1207 1208
static void __schedule_osd_timeout(struct ceph_osd_client *osdc)
{
	schedule_delayed_work(&osdc->timeout_work,
1209
			osdc->client->options->osd_keepalive_timeout * HZ);
1210 1211 1212 1213 1214 1215
}

static void __cancel_osd_timeout(struct ceph_osd_client *osdc)
{
	cancel_delayed_work(&osdc->timeout_work);
}
S
Sage Weil 已提交
1216 1217 1218 1219 1220

/*
 * Register request, assign tid.  If this is the first request, set up
 * the timeout event.
 */
1221 1222
static void __register_request(struct ceph_osd_client *osdc,
			       struct ceph_osd_request *req)
S
Sage Weil 已提交
1223 1224
{
	req->r_tid = ++osdc->last_tid;
1225
	req->r_request->hdr.tid = cpu_to_le64(req->r_tid);
1226
	dout("__register_request %p tid %lld\n", req, req->r_tid);
S
Sage Weil 已提交
1227 1228 1229 1230
	__insert_request(osdc, req);
	ceph_osdc_get_request(req);
	osdc->num_requests++;
	if (osdc->num_requests == 1) {
1231 1232
		dout(" first request, scheduling timeout\n");
		__schedule_osd_timeout(osdc);
S
Sage Weil 已提交
1233
	}
1234 1235
}

S
Sage Weil 已提交
1236 1237 1238 1239 1240 1241
/*
 * called under osdc->request_mutex
 */
static void __unregister_request(struct ceph_osd_client *osdc,
				 struct ceph_osd_request *req)
{
1242 1243 1244 1245 1246 1247
	if (RB_EMPTY_NODE(&req->r_node)) {
		dout("__unregister_request %p tid %lld not registered\n",
			req, req->r_tid);
		return;
	}

S
Sage Weil 已提交
1248 1249
	dout("__unregister_request %p tid %lld\n", req, req->r_tid);
	rb_erase(&req->r_node, &osdc->requests);
1250
	RB_CLEAR_NODE(&req->r_node);
S
Sage Weil 已提交
1251 1252
	osdc->num_requests--;

1253 1254
	if (req->r_osd) {
		/* make sure the original request isn't in flight. */
1255
		ceph_msg_revoke(req->r_request);
1256 1257

		list_del_init(&req->r_osd_item);
1258
		maybe_move_osd_to_lru(osdc, req->r_osd);
1259
		if (list_empty(&req->r_linger_osd_item))
1260
			req->r_osd = NULL;
1261
	}
S
Sage Weil 已提交
1262

A
Alex Elder 已提交
1263
	list_del_init(&req->r_req_lru_item);
S
Sage Weil 已提交
1264 1265
	ceph_osdc_put_request(req);

1266 1267 1268
	if (osdc->num_requests == 0) {
		dout(" no requests, canceling timeout\n");
		__cancel_osd_timeout(osdc);
S
Sage Weil 已提交
1269 1270 1271 1272 1273 1274 1275 1276
	}
}

/*
 * Cancel a previously queued request message
 */
static void __cancel_request(struct ceph_osd_request *req)
{
1277
	if (req->r_sent && req->r_osd) {
1278
		ceph_msg_revoke(req->r_request);
S
Sage Weil 已提交
1279 1280 1281 1282
		req->r_sent = 0;
	}
}

1283 1284 1285
static void __register_linger_request(struct ceph_osd_client *osdc,
				    struct ceph_osd_request *req)
{
1286 1287 1288
	dout("%s %p tid %llu\n", __func__, req, req->r_tid);
	WARN_ON(!req->r_linger);

1289
	ceph_osdc_get_request(req);
1290
	list_add_tail(&req->r_linger_item, &osdc->req_linger);
1291
	if (req->r_osd)
1292
		list_add_tail(&req->r_linger_osd_item,
1293
			      &req->r_osd->o_linger_requests);
1294 1295 1296 1297 1298
}

static void __unregister_linger_request(struct ceph_osd_client *osdc,
					struct ceph_osd_request *req)
{
1299 1300 1301 1302 1303 1304 1305 1306 1307
	WARN_ON(!req->r_linger);

	if (list_empty(&req->r_linger_item)) {
		dout("%s %p tid %llu not registered\n", __func__, req,
		     req->r_tid);
		return;
	}

	dout("%s %p tid %llu\n", __func__, req, req->r_tid);
A
Alex Elder 已提交
1308
	list_del_init(&req->r_linger_item);
1309

1310
	if (req->r_osd) {
1311
		list_del_init(&req->r_linger_osd_item);
1312
		maybe_move_osd_to_lru(osdc, req->r_osd);
1313 1314
		if (list_empty(&req->r_osd_item))
			req->r_osd = NULL;
1315
	}
1316
	ceph_osdc_put_request(req);
1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328
}

void ceph_osdc_set_request_linger(struct ceph_osd_client *osdc,
				  struct ceph_osd_request *req)
{
	if (!req->r_linger) {
		dout("set_request_linger %p\n", req);
		req->r_linger = 1;
	}
}
EXPORT_SYMBOL(ceph_osdc_set_request_linger);

1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344
/*
 * Returns whether a request should be blocked from being sent
 * based on the current osdmap and osd_client settings.
 *
 * Caller should hold map_sem for read.
 */
static bool __req_should_be_paused(struct ceph_osd_client *osdc,
				   struct ceph_osd_request *req)
{
	bool pauserd = ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_PAUSERD);
	bool pausewr = ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_PAUSEWR) ||
		ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_FULL);
	return (req->r_flags & CEPH_OSD_FLAG_READ && pauserd) ||
		(req->r_flags & CEPH_OSD_FLAG_WRITE && pausewr);
}

1345 1346 1347 1348 1349 1350 1351
/*
 * Calculate mapping of a request to a PG.  Takes tiering into account.
 */
static int __calc_request_pg(struct ceph_osdmap *osdmap,
			     struct ceph_osd_request *req,
			     struct ceph_pg *pg_out)
{
1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365
	bool need_check_tiering;

	need_check_tiering = false;
	if (req->r_target_oloc.pool == -1) {
		req->r_target_oloc = req->r_base_oloc; /* struct */
		need_check_tiering = true;
	}
	if (req->r_target_oid.name_len == 0) {
		ceph_oid_copy(&req->r_target_oid, &req->r_base_oid);
		need_check_tiering = true;
	}

	if (need_check_tiering &&
	    (req->r_flags & CEPH_OSD_FLAG_IGNORE_OVERLAY) == 0) {
1366 1367
		struct ceph_pg_pool_info *pi;

1368
		pi = ceph_pg_pool_by_id(osdmap, req->r_target_oloc.pool);
1369 1370 1371
		if (pi) {
			if ((req->r_flags & CEPH_OSD_FLAG_READ) &&
			    pi->read_tier >= 0)
1372
				req->r_target_oloc.pool = pi->read_tier;
1373 1374
			if ((req->r_flags & CEPH_OSD_FLAG_WRITE) &&
			    pi->write_tier >= 0)
1375
				req->r_target_oloc.pool = pi->write_tier;
1376 1377 1378 1379
		}
		/* !pi is caught in ceph_oloc_oid_to_pg() */
	}

1380 1381
	return ceph_oloc_oid_to_pg(osdmap, &req->r_target_oloc,
				   &req->r_target_oid, pg_out);
1382 1383
}

1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399
static void __enqueue_request(struct ceph_osd_request *req)
{
	struct ceph_osd_client *osdc = req->r_osdc;

	dout("%s %p tid %llu to osd%d\n", __func__, req, req->r_tid,
	     req->r_osd ? req->r_osd->o_osd : -1);

	if (req->r_osd) {
		__remove_osd_from_lru(req->r_osd);
		list_add_tail(&req->r_osd_item, &req->r_osd->o_requests);
		list_move_tail(&req->r_req_lru_item, &osdc->req_unsent);
	} else {
		list_move_tail(&req->r_req_lru_item, &osdc->req_notarget);
	}
}

S
Sage Weil 已提交
1400 1401 1402
/*
 * Pick an osd (the first 'up' osd in the pg), allocate the osd struct
 * (as needed), and set the request r_osd appropriately.  If there is
L
Lucas De Marchi 已提交
1403
 * no up osd, set r_osd to NULL.  Move the request to the appropriate list
1404
 * (unsent, homeless) or leave on in-flight lru.
S
Sage Weil 已提交
1405 1406 1407 1408 1409
 *
 * Return 0 if unchanged, 1 if changed, or negative on error.
 *
 * Caller should hold map_sem for read and request_mutex.
 */
1410
static int __map_request(struct ceph_osd_client *osdc,
1411
			 struct ceph_osd_request *req, int force_resend)
S
Sage Weil 已提交
1412
{
1413
	struct ceph_pg pgid;
1414
	int acting[CEPH_PG_MAX_SIZE];
1415
	int num, o;
S
Sage Weil 已提交
1416
	int err;
1417
	bool was_paused;
S
Sage Weil 已提交
1418

1419
	dout("map_request %p tid %lld\n", req, req->r_tid);
1420 1421

	err = __calc_request_pg(osdc->osdmap, req, &pgid);
1422 1423
	if (err) {
		list_move(&req->r_req_lru_item, &osdc->req_notarget);
S
Sage Weil 已提交
1424
		return err;
1425
	}
1426 1427
	req->r_pgid = pgid;

1428 1429 1430
	num = ceph_calc_pg_acting(osdc->osdmap, pgid, acting, &o);
	if (num < 0)
		num = 0;
S
Sage Weil 已提交
1431

1432 1433 1434 1435 1436
	was_paused = req->r_paused;
	req->r_paused = __req_should_be_paused(osdc, req);
	if (was_paused && !req->r_paused)
		force_resend = 1;

1437 1438
	if ((!force_resend &&
	     req->r_osd && req->r_osd->o_osd == o &&
1439 1440 1441
	     req->r_sent >= req->r_osd->o_incarnation &&
	     req->r_num_pg_osds == num &&
	     memcmp(req->r_pg_osds, acting, sizeof(acting[0])*num) == 0) ||
1442 1443
	    (req->r_osd == NULL && o == -1) ||
	    req->r_paused)
S
Sage Weil 已提交
1444 1445
		return 0;  /* no change */

1446 1447
	dout("map_request tid %llu pgid %lld.%x osd%d (was osd%d)\n",
	     req->r_tid, pgid.pool, pgid.seed, o,
S
Sage Weil 已提交
1448 1449
	     req->r_osd ? req->r_osd->o_osd : -1);

1450 1451 1452 1453
	/* record full pg acting set */
	memcpy(req->r_pg_osds, acting, sizeof(acting[0]) * num);
	req->r_num_pg_osds = num;

S
Sage Weil 已提交
1454 1455 1456 1457 1458 1459 1460 1461
	if (req->r_osd) {
		__cancel_request(req);
		list_del_init(&req->r_osd_item);
		req->r_osd = NULL;
	}

	req->r_osd = __lookup_osd(osdc, o);
	if (!req->r_osd && o >= 0) {
1462
		err = -ENOMEM;
1463
		req->r_osd = create_osd(osdc, o);
1464 1465
		if (!req->r_osd) {
			list_move(&req->r_req_lru_item, &osdc->req_notarget);
1466
			goto out;
1467
		}
S
Sage Weil 已提交
1468

1469
		dout("map_request osd %p is osd%d\n", req->r_osd, o);
S
Sage Weil 已提交
1470 1471
		__insert_osd(osdc, req->r_osd);

1472 1473 1474
		ceph_con_open(&req->r_osd->o_con,
			      CEPH_ENTITY_TYPE_OSD, o,
			      &osdc->osdmap->osd_addr[o]);
S
Sage Weil 已提交
1475 1476
	}

1477
	__enqueue_request(req);
1478
	err = 1;   /* osd or pg changed */
S
Sage Weil 已提交
1479 1480 1481 1482 1483 1484 1485 1486

out:
	return err;
}

/*
 * caller should hold map_sem (for read) and request_mutex
 */
1487 1488
static void __send_request(struct ceph_osd_client *osdc,
			   struct ceph_osd_request *req)
S
Sage Weil 已提交
1489
{
1490
	void *p;
S
Sage Weil 已提交
1491

1492 1493 1494 1495 1496 1497 1498
	dout("send_request %p tid %llu to osd%d flags %d pg %lld.%x\n",
	     req, req->r_tid, req->r_osd->o_osd, req->r_flags,
	     (unsigned long long)req->r_pgid.pool, req->r_pgid.seed);

	/* fill in message content that changes each time we send it */
	put_unaligned_le32(osdc->osdmap->epoch, req->r_request_osdmap_epoch);
	put_unaligned_le32(req->r_flags, req->r_request_flags);
1499
	put_unaligned_le64(req->r_target_oloc.pool, req->r_request_pool);
1500 1501 1502 1503 1504 1505
	p = req->r_request_pgid;
	ceph_encode_64(&p, req->r_pgid.pool);
	ceph_encode_32(&p, req->r_pgid.seed);
	put_unaligned_le64(1, req->r_request_attempts);  /* FIXME */
	memcpy(req->r_request_reassert_version, &req->r_reassert_version,
	       sizeof(req->r_reassert_version));
1506

S
Sage Weil 已提交
1507
	req->r_stamp = jiffies;
1508
	list_move_tail(&req->r_req_lru_item, &osdc->req_lru);
S
Sage Weil 已提交
1509 1510

	ceph_msg_get(req->r_request); /* send consumes a ref */
1511

S
Sage Weil 已提交
1512
	req->r_sent = req->r_osd->o_incarnation;
1513 1514

	ceph_con_send(&req->r_osd->o_con, req->r_request);
S
Sage Weil 已提交
1515 1516
}

1517 1518 1519
/*
 * Send any requests in the queue (req_unsent).
 */
A
Alex Elder 已提交
1520
static void __send_queued(struct ceph_osd_client *osdc)
1521 1522 1523
{
	struct ceph_osd_request *req, *tmp;

A
Alex Elder 已提交
1524 1525
	dout("__send_queued\n");
	list_for_each_entry_safe(req, tmp, &osdc->req_unsent, r_req_lru_item)
1526 1527 1528
		__send_request(osdc, req);
}

1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562
/*
 * Caller should hold map_sem for read and request_mutex.
 */
static int __ceph_osdc_start_request(struct ceph_osd_client *osdc,
				     struct ceph_osd_request *req,
				     bool nofail)
{
	int rc;

	__register_request(osdc, req);
	req->r_sent = 0;
	req->r_got_reply = 0;
	rc = __map_request(osdc, req, 0);
	if (rc < 0) {
		if (nofail) {
			dout("osdc_start_request failed map, "
				" will retry %lld\n", req->r_tid);
			rc = 0;
		} else {
			__unregister_request(osdc, req);
		}
		return rc;
	}

	if (req->r_osd == NULL) {
		dout("send_request %p no up osds in pg\n", req);
		ceph_monc_request_next_osdmap(&osdc->client->monc);
	} else {
		__send_queued(osdc);
	}

	return 0;
}

S
Sage Weil 已提交
1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575
/*
 * Timeout callback, called every N seconds when 1 or more osd
 * requests has been active for more than N seconds.  When this
 * happens, we ping all OSDs with requests who have timed out to
 * ensure any communications channel reset is detected.  Reset the
 * request timeouts another N seconds in the future as we go.
 * Reschedule the timeout event another N seconds in future (unless
 * there are no open requests).
 */
static void handle_timeout(struct work_struct *work)
{
	struct ceph_osd_client *osdc =
		container_of(work, struct ceph_osd_client, timeout_work.work);
S
Sage Weil 已提交
1576
	struct ceph_osd_request *req;
S
Sage Weil 已提交
1577
	struct ceph_osd *osd;
1578
	unsigned long keepalive =
1579
		osdc->client->options->osd_keepalive_timeout * HZ;
1580
	struct list_head slow_osds;
S
Sage Weil 已提交
1581 1582 1583 1584 1585 1586 1587
	dout("timeout\n");
	down_read(&osdc->map_sem);

	ceph_monc_request_next_osdmap(&osdc->client->monc);

	mutex_lock(&osdc->request_mutex);

1588 1589 1590 1591 1592 1593 1594
	/*
	 * ping osds that are a bit slow.  this ensures that if there
	 * is a break in the TCP connection we will notice, and reopen
	 * a connection with that osd (from the fault callback).
	 */
	INIT_LIST_HEAD(&slow_osds);
	list_for_each_entry(req, &osdc->req_lru, r_req_lru_item) {
S
Sage Weil 已提交
1595
		if (time_before(jiffies, req->r_stamp + keepalive))
1596 1597 1598 1599 1600
			break;

		osd = req->r_osd;
		BUG_ON(!osd);
		dout(" tid %llu is slow, will send keepalive on osd%d\n",
S
Sage Weil 已提交
1601
		     req->r_tid, osd->o_osd);
1602 1603 1604 1605 1606 1607
		list_move_tail(&osd->o_keepalive_item, &slow_osds);
	}
	while (!list_empty(&slow_osds)) {
		osd = list_entry(slow_osds.next, struct ceph_osd,
				 o_keepalive_item);
		list_del_init(&osd->o_keepalive_item);
S
Sage Weil 已提交
1608 1609 1610
		ceph_con_keepalive(&osd->o_con);
	}

1611
	__schedule_osd_timeout(osdc);
A
Alex Elder 已提交
1612
	__send_queued(osdc);
S
Sage Weil 已提交
1613 1614 1615 1616
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
}

1617 1618 1619 1620 1621 1622
static void handle_osds_timeout(struct work_struct *work)
{
	struct ceph_osd_client *osdc =
		container_of(work, struct ceph_osd_client,
			     osds_timeout_work.work);
	unsigned long delay =
1623
		osdc->client->options->osd_idle_ttl * HZ >> 2;
1624 1625 1626

	dout("osds timeout\n");
	down_read(&osdc->map_sem);
1627
	remove_old_osds(osdc);
1628 1629 1630 1631 1632 1633
	up_read(&osdc->map_sem);

	schedule_delayed_work(&osdc->osds_timeout_work,
			      round_jiffies_relative(delay));
}

1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736
static int ceph_oloc_decode(void **p, void *end,
			    struct ceph_object_locator *oloc)
{
	u8 struct_v, struct_cv;
	u32 len;
	void *struct_end;
	int ret = 0;

	ceph_decode_need(p, end, 1 + 1 + 4, e_inval);
	struct_v = ceph_decode_8(p);
	struct_cv = ceph_decode_8(p);
	if (struct_v < 3) {
		pr_warn("got v %d < 3 cv %d of ceph_object_locator\n",
			struct_v, struct_cv);
		goto e_inval;
	}
	if (struct_cv > 6) {
		pr_warn("got v %d cv %d > 6 of ceph_object_locator\n",
			struct_v, struct_cv);
		goto e_inval;
	}
	len = ceph_decode_32(p);
	ceph_decode_need(p, end, len, e_inval);
	struct_end = *p + len;

	oloc->pool = ceph_decode_64(p);
	*p += 4; /* skip preferred */

	len = ceph_decode_32(p);
	if (len > 0) {
		pr_warn("ceph_object_locator::key is set\n");
		goto e_inval;
	}

	if (struct_v >= 5) {
		len = ceph_decode_32(p);
		if (len > 0) {
			pr_warn("ceph_object_locator::nspace is set\n");
			goto e_inval;
		}
	}

	if (struct_v >= 6) {
		s64 hash = ceph_decode_64(p);
		if (hash != -1) {
			pr_warn("ceph_object_locator::hash is set\n");
			goto e_inval;
		}
	}

	/* skip the rest */
	*p = struct_end;
out:
	return ret;

e_inval:
	ret = -EINVAL;
	goto out;
}

static int ceph_redirect_decode(void **p, void *end,
				struct ceph_request_redirect *redir)
{
	u8 struct_v, struct_cv;
	u32 len;
	void *struct_end;
	int ret;

	ceph_decode_need(p, end, 1 + 1 + 4, e_inval);
	struct_v = ceph_decode_8(p);
	struct_cv = ceph_decode_8(p);
	if (struct_cv > 1) {
		pr_warn("got v %d cv %d > 1 of ceph_request_redirect\n",
			struct_v, struct_cv);
		goto e_inval;
	}
	len = ceph_decode_32(p);
	ceph_decode_need(p, end, len, e_inval);
	struct_end = *p + len;

	ret = ceph_oloc_decode(p, end, &redir->oloc);
	if (ret)
		goto out;

	len = ceph_decode_32(p);
	if (len > 0) {
		pr_warn("ceph_request_redirect::object_name is set\n");
		goto e_inval;
	}

	len = ceph_decode_32(p);
	*p += len; /* skip osd_instructions */

	/* skip the rest */
	*p = struct_end;
out:
	return ret;

e_inval:
	ret = -EINVAL;
	goto out;
}

S
Sage Weil 已提交
1737 1738 1739 1740 1741
static void complete_request(struct ceph_osd_request *req)
{
	complete_all(&req->r_safe_completion);  /* fsync waiter */
}

S
Sage Weil 已提交
1742 1743 1744 1745
/*
 * handle osd op reply.  either call the callback if it is specified,
 * or do the completion to wake up the waiting thread.
 */
1746 1747
static void handle_reply(struct ceph_osd_client *osdc, struct ceph_msg *msg,
			 struct ceph_connection *con)
S
Sage Weil 已提交
1748
{
1749
	void *p, *end;
S
Sage Weil 已提交
1750
	struct ceph_osd_request *req;
1751
	struct ceph_request_redirect redir;
S
Sage Weil 已提交
1752
	u64 tid;
1753
	int object_len;
1754 1755
	unsigned int numops;
	int payload_len, flags;
1756
	s32 result;
1757 1758 1759 1760 1761 1762
	s32 retry_attempt;
	struct ceph_pg pg;
	int err;
	u32 reassert_epoch;
	u64 reassert_version;
	u32 osdmap_epoch;
1763
	int already_completed;
1764
	u32 bytes;
1765
	unsigned int i;
S
Sage Weil 已提交
1766

1767
	tid = le64_to_cpu(msg->hdr.tid);
1768 1769 1770 1771 1772 1773 1774 1775 1776 1777
	dout("handle_reply %p tid %llu\n", msg, tid);

	p = msg->front.iov_base;
	end = p + msg->front.iov_len;

	ceph_decode_need(&p, end, 4, bad);
	object_len = ceph_decode_32(&p);
	ceph_decode_need(&p, end, object_len, bad);
	p += object_len;

1778
	err = ceph_decode_pgid(&p, end, &pg);
1779
	if (err)
S
Sage Weil 已提交
1780
		goto bad;
1781 1782 1783 1784 1785 1786 1787 1788

	ceph_decode_need(&p, end, 8 + 4 + 4 + 8 + 4, bad);
	flags = ceph_decode_64(&p);
	result = ceph_decode_32(&p);
	reassert_epoch = ceph_decode_32(&p);
	reassert_version = ceph_decode_64(&p);
	osdmap_epoch = ceph_decode_32(&p);

S
Sage Weil 已提交
1789
	/* lookup */
1790
	down_read(&osdc->map_sem);
S
Sage Weil 已提交
1791 1792 1793 1794
	mutex_lock(&osdc->request_mutex);
	req = __lookup_request(osdc, tid);
	if (req == NULL) {
		dout("handle_reply tid %llu dne\n", tid);
1795
		goto bad_mutex;
S
Sage Weil 已提交
1796 1797
	}
	ceph_osdc_get_request(req);
1798 1799 1800 1801

	dout("handle_reply %p tid %llu req %p result %d\n", msg, tid,
	     req, result);

1802
	ceph_decode_need(&p, end, 4, bad_put);
1803 1804 1805 1806 1807 1808
	numops = ceph_decode_32(&p);
	if (numops > CEPH_OSD_MAX_OP)
		goto bad_put;
	if (numops != req->r_num_ops)
		goto bad_put;
	payload_len = 0;
1809
	ceph_decode_need(&p, end, numops * sizeof(struct ceph_osd_op), bad_put);
1810 1811 1812 1813 1814 1815 1816 1817 1818 1819
	for (i = 0; i < numops; i++) {
		struct ceph_osd_op *op = p;
		int len;

		len = le32_to_cpu(op->payload_len);
		req->r_reply_op_len[i] = len;
		dout(" op %d has %d bytes\n", i, len);
		payload_len += len;
		p += sizeof(*op);
	}
1820 1821
	bytes = le32_to_cpu(msg->hdr.data_len);
	if (payload_len != bytes) {
1822
		pr_warning("sum of op payload lens %d != data_len %d",
1823
			   payload_len, bytes);
1824 1825 1826
		goto bad_put;
	}

1827
	ceph_decode_need(&p, end, 4 + numops * 4, bad_put);
1828 1829 1830
	retry_attempt = ceph_decode_32(&p);
	for (i = 0; i < numops; i++)
		req->r_reply_op_result[i] = ceph_decode_32(&p);
S
Sage Weil 已提交
1831

1832 1833 1834
	if (le16_to_cpu(msg->hdr.version) >= 6) {
		p += 8 + 4; /* skip replay_version */
		p += 8; /* skip user_version */
Y
Yan, Zheng 已提交
1835

1836 1837 1838 1839 1840 1841
		err = ceph_redirect_decode(&p, end, &redir);
		if (err)
			goto bad_put;
	} else {
		redir.oloc.pool = -1;
	}
S
Sage Weil 已提交
1842

1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857
	if (redir.oloc.pool != -1) {
		dout("redirect pool %lld\n", redir.oloc.pool);

		__unregister_request(osdc, req);

		req->r_target_oloc = redir.oloc; /* struct */

		/*
		 * Start redirect requests with nofail=true.  If
		 * mapping fails, request will end up on the notarget
		 * list, waiting for the new osdmap (which can take
		 * a while), even though the original request mapped
		 * successfully.  In the future we might want to follow
		 * original request's nofail setting here.
		 */
1858
		err = __ceph_osdc_start_request(osdc, req, true);
1859 1860
		BUG_ON(err);

1861
		goto out_unlock;
1862 1863 1864 1865
	}

	already_completed = req->r_got_reply;
	if (!req->r_got_reply) {
1866
		req->r_result = result;
S
Sage Weil 已提交
1867 1868 1869 1870 1871 1872
		dout("handle_reply result %d bytes %d\n", req->r_result,
		     bytes);
		if (req->r_result == 0)
			req->r_result = bytes;

		/* in case this is a write and we need to replay, */
1873 1874
		req->r_reassert_version.epoch = cpu_to_le32(reassert_epoch);
		req->r_reassert_version.version = cpu_to_le64(reassert_version);
S
Sage Weil 已提交
1875 1876 1877 1878

		req->r_got_reply = 1;
	} else if ((flags & CEPH_OSD_FLAG_ONDISK) == 0) {
		dout("handle_reply tid %llu dup ack\n", tid);
1879
		goto out_unlock;
S
Sage Weil 已提交
1880 1881 1882 1883
	}

	dout("handle_reply tid %llu flags %d\n", tid, flags);

1884 1885 1886
	if (req->r_linger && (flags & CEPH_OSD_FLAG_ONDISK))
		__register_linger_request(osdc, req);

S
Sage Weil 已提交
1887
	/* either this is a read, or we got the safe response */
1888 1889
	if (result < 0 ||
	    (flags & CEPH_OSD_FLAG_ONDISK) ||
S
Sage Weil 已提交
1890 1891 1892 1893
	    ((flags & CEPH_OSD_FLAG_WRITE) == 0))
		__unregister_request(osdc, req);

	mutex_unlock(&osdc->request_mutex);
1894
	up_read(&osdc->map_sem);
S
Sage Weil 已提交
1895

Y
Yan, Zheng 已提交
1896
	if (!already_completed) {
1897 1898 1899
		if (req->r_unsafe_callback &&
		    result >= 0 && !(flags & CEPH_OSD_FLAG_ONDISK))
			req->r_unsafe_callback(req, true);
Y
Yan, Zheng 已提交
1900 1901 1902 1903 1904
		if (req->r_callback)
			req->r_callback(req, msg);
		else
			complete_all(&req->r_completion);
	}
S
Sage Weil 已提交
1905

1906 1907 1908
	if (flags & CEPH_OSD_FLAG_ONDISK) {
		if (req->r_unsafe_callback && already_completed)
			req->r_unsafe_callback(req, false);
S
Sage Weil 已提交
1909
		complete_request(req);
1910
	}
S
Sage Weil 已提交
1911

1912
out:
1913
	dout("req=%p req->r_linger=%d\n", req, req->r_linger);
S
Sage Weil 已提交
1914 1915
	ceph_osdc_put_request(req);
	return;
1916 1917 1918 1919
out_unlock:
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
	goto out;
S
Sage Weil 已提交
1920

1921
bad_put:
1922 1923 1924 1925 1926 1927 1928
	req->r_result = -EIO;
	__unregister_request(osdc, req);
	if (req->r_callback)
		req->r_callback(req, msg);
	else
		complete_all(&req->r_completion);
	complete_request(req);
1929
	ceph_osdc_put_request(req);
1930 1931
bad_mutex:
	mutex_unlock(&osdc->request_mutex);
1932
	up_read(&osdc->map_sem);
S
Sage Weil 已提交
1933
bad:
1934 1935
	pr_err("corrupt osd_op_reply got %d %d\n",
	       (int)msg->front.iov_len, le32_to_cpu(msg->hdr.front_len));
1936
	ceph_msg_dump(msg);
S
Sage Weil 已提交
1937 1938
}

1939
static void reset_changed_osds(struct ceph_osd_client *osdc)
S
Sage Weil 已提交
1940 1941 1942
{
	struct rb_node *p, *n;

1943 1944
	for (p = rb_first(&osdc->osds); p; p = n) {
		struct ceph_osd *osd = rb_entry(p, struct ceph_osd, o_node);
S
Sage Weil 已提交
1945

1946 1947 1948 1949 1950 1951 1952
		n = rb_next(p);
		if (!ceph_osd_is_up(osdc->osdmap, osd->o_osd) ||
		    memcmp(&osd->o_con.peer_addr,
			   ceph_osd_addr(osdc->osdmap,
					 osd->o_osd),
			   sizeof(struct ceph_entity_addr)) != 0)
			__reset_osd(osdc, osd);
S
Sage Weil 已提交
1953
	}
1954 1955 1956
}

/*
1957 1958
 * Requeue requests whose mapping to an OSD has changed.  If requests map to
 * no osd, request a new map.
1959
 *
1960
 * Caller should hold map_sem for read.
1961
 */
1962 1963
static void kick_requests(struct ceph_osd_client *osdc, bool force_resend,
			  bool force_resend_writes)
1964
{
1965
	struct ceph_osd_request *req, *nreq;
1966 1967 1968
	struct rb_node *p;
	int needmap = 0;
	int err;
1969
	bool force_resend_req;
1970

1971 1972
	dout("kick_requests %s %s\n", force_resend ? " (force resend)" : "",
		force_resend_writes ? " (force resend writes)" : "");
1973
	mutex_lock(&osdc->request_mutex);
1974
	for (p = rb_first(&osdc->requests); p; ) {
1975
		req = rb_entry(p, struct ceph_osd_request, r_node);
1976
		p = rb_next(p);
1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989

		/*
		 * For linger requests that have not yet been
		 * registered, move them to the linger list; they'll
		 * be sent to the osd in the loop below.  Unregister
		 * the request before re-registering it as a linger
		 * request to ensure the __map_request() below
		 * will decide it needs to be sent.
		 */
		if (req->r_linger && list_empty(&req->r_linger_item)) {
			dout("%p tid %llu restart on osd%d\n",
			     req, req->r_tid,
			     req->r_osd ? req->r_osd->o_osd : -1);
1990
			ceph_osdc_get_request(req);
1991 1992
			__unregister_request(osdc, req);
			__register_linger_request(osdc, req);
1993
			ceph_osdc_put_request(req);
1994 1995 1996
			continue;
		}

1997 1998 1999 2000
		force_resend_req = force_resend ||
			(force_resend_writes &&
				req->r_flags & CEPH_OSD_FLAG_WRITE);
		err = __map_request(osdc, req, force_resend_req);
2001 2002 2003 2004 2005 2006
		if (err < 0)
			continue;  /* error */
		if (req->r_osd == NULL) {
			dout("%p tid %llu maps to no osd\n", req, req->r_tid);
			needmap++;  /* request a newer map */
		} else if (err > 0) {
2007 2008 2009 2010
			if (!req->r_linger) {
				dout("%p tid %llu requeued on osd%d\n", req,
				     req->r_tid,
				     req->r_osd ? req->r_osd->o_osd : -1);
2011
				req->r_flags |= CEPH_OSD_FLAG_RETRY;
2012 2013
			}
		}
2014 2015 2016 2017 2018 2019
	}

	list_for_each_entry_safe(req, nreq, &osdc->req_linger,
				 r_linger_item) {
		dout("linger req=%p req->r_osd=%p\n", req, req->r_osd);

2020 2021
		err = __map_request(osdc, req,
				    force_resend || force_resend_writes);
2022
		dout("__map_request returned %d\n", err);
2023 2024 2025 2026 2027 2028 2029 2030
		if (err == 0)
			continue;  /* no change and no osd was specified */
		if (err < 0)
			continue;  /* hrm! */
		if (req->r_osd == NULL) {
			dout("tid %llu maps to no valid osd\n", req->r_tid);
			needmap++;  /* request a newer map */
			continue;
2031
		}
2032 2033 2034 2035

		dout("kicking lingering %p tid %llu osd%d\n", req, req->r_tid,
		     req->r_osd ? req->r_osd->o_osd : -1);
		__register_request(osdc, req);
2036
		__unregister_linger_request(osdc, req);
2037
	}
2038
	reset_changed_osds(osdc);
S
Sage Weil 已提交
2039 2040 2041 2042 2043 2044
	mutex_unlock(&osdc->request_mutex);

	if (needmap) {
		dout("%d requests for down osds, need new map\n", needmap);
		ceph_monc_request_next_osdmap(&osdc->client->monc);
	}
2045
}
2046 2047


S
Sage Weil 已提交
2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062
/*
 * Process updated osd map.
 *
 * The message contains any number of incremental and full maps, normally
 * indicating some sort of topology change in the cluster.  Kick requests
 * off to different OSDs as needed.
 */
void ceph_osdc_handle_map(struct ceph_osd_client *osdc, struct ceph_msg *msg)
{
	void *p, *end, *next;
	u32 nr_maps, maplen;
	u32 epoch;
	struct ceph_osdmap *newmap = NULL, *oldmap;
	int err;
	struct ceph_fsid fsid;
2063
	bool was_full;
S
Sage Weil 已提交
2064 2065 2066 2067 2068 2069 2070 2071

	dout("handle_map have %u\n", osdc->osdmap ? osdc->osdmap->epoch : 0);
	p = msg->front.iov_base;
	end = p + msg->front.iov_len;

	/* verify fsid */
	ceph_decode_need(&p, end, sizeof(fsid), bad);
	ceph_decode_copy(&p, &fsid, sizeof(fsid));
2072 2073
	if (ceph_check_fsid(osdc->client, &fsid) < 0)
		return;
S
Sage Weil 已提交
2074 2075 2076

	down_write(&osdc->map_sem);

2077 2078
	was_full = ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_FULL);

S
Sage Weil 已提交
2079 2080 2081 2082 2083
	/* incremental maps */
	ceph_decode_32_safe(&p, end, nr_maps, bad);
	dout(" %d inc maps\n", nr_maps);
	while (nr_maps > 0) {
		ceph_decode_need(&p, end, 2*sizeof(u32), bad);
2084 2085
		epoch = ceph_decode_32(&p);
		maplen = ceph_decode_32(&p);
S
Sage Weil 已提交
2086 2087 2088 2089 2090 2091 2092
		ceph_decode_need(&p, end, maplen, bad);
		next = p + maplen;
		if (osdc->osdmap && osdc->osdmap->epoch+1 == epoch) {
			dout("applying incremental map %u len %d\n",
			     epoch, maplen);
			newmap = osdmap_apply_incremental(&p, next,
							  osdc->osdmap,
2093
							  &osdc->client->msgr);
S
Sage Weil 已提交
2094 2095 2096 2097
			if (IS_ERR(newmap)) {
				err = PTR_ERR(newmap);
				goto bad;
			}
2098
			BUG_ON(!newmap);
S
Sage Weil 已提交
2099 2100 2101 2102
			if (newmap != osdc->osdmap) {
				ceph_osdmap_destroy(osdc->osdmap);
				osdc->osdmap = newmap;
			}
2103 2104 2105 2106
			was_full = was_full ||
				ceph_osdmap_flag(osdc->osdmap,
						 CEPH_OSDMAP_FULL);
			kick_requests(osdc, 0, was_full);
S
Sage Weil 已提交
2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121
		} else {
			dout("ignoring incremental map %u len %d\n",
			     epoch, maplen);
		}
		p = next;
		nr_maps--;
	}
	if (newmap)
		goto done;

	/* full maps */
	ceph_decode_32_safe(&p, end, nr_maps, bad);
	dout(" %d full maps\n", nr_maps);
	while (nr_maps) {
		ceph_decode_need(&p, end, 2*sizeof(u32), bad);
2122 2123
		epoch = ceph_decode_32(&p);
		maplen = ceph_decode_32(&p);
S
Sage Weil 已提交
2124 2125 2126 2127 2128 2129 2130 2131 2132
		ceph_decode_need(&p, end, maplen, bad);
		if (nr_maps > 1) {
			dout("skipping non-latest full map %u len %d\n",
			     epoch, maplen);
		} else if (osdc->osdmap && osdc->osdmap->epoch >= epoch) {
			dout("skipping full map %u len %d, "
			     "older than our %u\n", epoch, maplen,
			     osdc->osdmap->epoch);
		} else {
2133 2134
			int skipped_map = 0;

S
Sage Weil 已提交
2135
			dout("taking full map %u len %d\n", epoch, maplen);
2136
			newmap = ceph_osdmap_decode(&p, p+maplen);
S
Sage Weil 已提交
2137 2138 2139 2140
			if (IS_ERR(newmap)) {
				err = PTR_ERR(newmap);
				goto bad;
			}
2141
			BUG_ON(!newmap);
S
Sage Weil 已提交
2142 2143
			oldmap = osdc->osdmap;
			osdc->osdmap = newmap;
2144 2145 2146
			if (oldmap) {
				if (oldmap->epoch + 1 < newmap->epoch)
					skipped_map = 1;
S
Sage Weil 已提交
2147
				ceph_osdmap_destroy(oldmap);
2148
			}
2149 2150 2151 2152
			was_full = was_full ||
				ceph_osdmap_flag(osdc->osdmap,
						 CEPH_OSDMAP_FULL);
			kick_requests(osdc, skipped_map, was_full);
S
Sage Weil 已提交
2153 2154 2155 2156 2157
		}
		p += maplen;
		nr_maps--;
	}

2158 2159
	if (!osdc->osdmap)
		goto bad;
S
Sage Weil 已提交
2160 2161 2162
done:
	downgrade_write(&osdc->map_sem);
	ceph_monc_got_osdmap(&osdc->client->monc, osdc->osdmap->epoch);
2163 2164 2165 2166 2167 2168

	/*
	 * subscribe to subsequent osdmap updates if full to ensure
	 * we find out when we are no longer full and stop returning
	 * ENOSPC.
	 */
2169 2170 2171
	if (ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_FULL) ||
		ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_PAUSERD) ||
		ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_PAUSEWR))
2172 2173
		ceph_monc_request_next_osdmap(&osdc->client->monc);

A
Alex Elder 已提交
2174 2175 2176
	mutex_lock(&osdc->request_mutex);
	__send_queued(osdc);
	mutex_unlock(&osdc->request_mutex);
S
Sage Weil 已提交
2177
	up_read(&osdc->map_sem);
2178
	wake_up_all(&osdc->client->auth_wq);
S
Sage Weil 已提交
2179 2180 2181 2182
	return;

bad:
	pr_err("osdc handle_map corrupt msg\n");
2183
	ceph_msg_dump(msg);
S
Sage Weil 已提交
2184 2185 2186
	up_write(&osdc->map_sem);
}

2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268
/*
 * watch/notify callback event infrastructure
 *
 * These callbacks are used both for watch and notify operations.
 */
static void __release_event(struct kref *kref)
{
	struct ceph_osd_event *event =
		container_of(kref, struct ceph_osd_event, kref);

	dout("__release_event %p\n", event);
	kfree(event);
}

static void get_event(struct ceph_osd_event *event)
{
	kref_get(&event->kref);
}

void ceph_osdc_put_event(struct ceph_osd_event *event)
{
	kref_put(&event->kref, __release_event);
}
EXPORT_SYMBOL(ceph_osdc_put_event);

static void __insert_event(struct ceph_osd_client *osdc,
			     struct ceph_osd_event *new)
{
	struct rb_node **p = &osdc->event_tree.rb_node;
	struct rb_node *parent = NULL;
	struct ceph_osd_event *event = NULL;

	while (*p) {
		parent = *p;
		event = rb_entry(parent, struct ceph_osd_event, node);
		if (new->cookie < event->cookie)
			p = &(*p)->rb_left;
		else if (new->cookie > event->cookie)
			p = &(*p)->rb_right;
		else
			BUG();
	}

	rb_link_node(&new->node, parent, p);
	rb_insert_color(&new->node, &osdc->event_tree);
}

static struct ceph_osd_event *__find_event(struct ceph_osd_client *osdc,
					        u64 cookie)
{
	struct rb_node **p = &osdc->event_tree.rb_node;
	struct rb_node *parent = NULL;
	struct ceph_osd_event *event = NULL;

	while (*p) {
		parent = *p;
		event = rb_entry(parent, struct ceph_osd_event, node);
		if (cookie < event->cookie)
			p = &(*p)->rb_left;
		else if (cookie > event->cookie)
			p = &(*p)->rb_right;
		else
			return event;
	}
	return NULL;
}

static void __remove_event(struct ceph_osd_event *event)
{
	struct ceph_osd_client *osdc = event->osdc;

	if (!RB_EMPTY_NODE(&event->node)) {
		dout("__remove_event removed %p\n", event);
		rb_erase(&event->node, &osdc->event_tree);
		ceph_osdc_put_event(event);
	} else {
		dout("__remove_event didn't remove %p\n", event);
	}
}

int ceph_osdc_create_event(struct ceph_osd_client *osdc,
			   void (*event_cb)(u64, u64, u8, void *),
2269
			   void *data, struct ceph_osd_event **pevent)
2270 2271 2272 2273 2274 2275 2276 2277 2278
{
	struct ceph_osd_event *event;

	event = kmalloc(sizeof(*event), GFP_NOIO);
	if (!event)
		return -ENOMEM;

	dout("create_event %p\n", event);
	event->cb = event_cb;
2279
	event->one_shot = 0;
2280 2281 2282
	event->data = data;
	event->osdc = osdc;
	INIT_LIST_HEAD(&event->osd_node);
2283
	RB_CLEAR_NODE(&event->node);
2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329
	kref_init(&event->kref);   /* one ref for us */
	kref_get(&event->kref);    /* one ref for the caller */

	spin_lock(&osdc->event_lock);
	event->cookie = ++osdc->event_count;
	__insert_event(osdc, event);
	spin_unlock(&osdc->event_lock);

	*pevent = event;
	return 0;
}
EXPORT_SYMBOL(ceph_osdc_create_event);

void ceph_osdc_cancel_event(struct ceph_osd_event *event)
{
	struct ceph_osd_client *osdc = event->osdc;

	dout("cancel_event %p\n", event);
	spin_lock(&osdc->event_lock);
	__remove_event(event);
	spin_unlock(&osdc->event_lock);
	ceph_osdc_put_event(event); /* caller's */
}
EXPORT_SYMBOL(ceph_osdc_cancel_event);


static void do_event_work(struct work_struct *work)
{
	struct ceph_osd_event_work *event_work =
		container_of(work, struct ceph_osd_event_work, work);
	struct ceph_osd_event *event = event_work->event;
	u64 ver = event_work->ver;
	u64 notify_id = event_work->notify_id;
	u8 opcode = event_work->opcode;

	dout("do_event_work completing %p\n", event);
	event->cb(ver, notify_id, opcode, event->data);
	dout("do_event_work completed %p\n", event);
	ceph_osdc_put_event(event);
	kfree(event_work);
}


/*
 * Process osd watch notifications
 */
2330 2331
static void handle_watch_notify(struct ceph_osd_client *osdc,
				struct ceph_msg *msg)
2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351
{
	void *p, *end;
	u8 proto_ver;
	u64 cookie, ver, notify_id;
	u8 opcode;
	struct ceph_osd_event *event;
	struct ceph_osd_event_work *event_work;

	p = msg->front.iov_base;
	end = p + msg->front.iov_len;

	ceph_decode_8_safe(&p, end, proto_ver, bad);
	ceph_decode_8_safe(&p, end, opcode, bad);
	ceph_decode_64_safe(&p, end, cookie, bad);
	ceph_decode_64_safe(&p, end, ver, bad);
	ceph_decode_64_safe(&p, end, notify_id, bad);

	spin_lock(&osdc->event_lock);
	event = __find_event(osdc, cookie);
	if (event) {
2352
		BUG_ON(event->one_shot);
2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363
		get_event(event);
	}
	spin_unlock(&osdc->event_lock);
	dout("handle_watch_notify cookie %lld ver %lld event %p\n",
	     cookie, ver, event);
	if (event) {
		event_work = kmalloc(sizeof(*event_work), GFP_NOIO);
		if (!event_work) {
			dout("ERROR: could not allocate event_work\n");
			goto done_err;
		}
2364
		INIT_WORK(&event_work->work, do_event_work);
2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384
		event_work->event = event;
		event_work->ver = ver;
		event_work->notify_id = notify_id;
		event_work->opcode = opcode;
		if (!queue_work(osdc->notify_wq, &event_work->work)) {
			dout("WARNING: failed to queue notify event work\n");
			goto done_err;
		}
	}

	return;

done_err:
	ceph_osdc_put_event(event);
	return;

bad:
	pr_err("osdc handle_watch_notify corrupt msg\n");
}

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 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432
/*
 * build new request AND message
 *
 */
void ceph_osdc_build_request(struct ceph_osd_request *req, u64 off,
				struct ceph_snap_context *snapc, u64 snap_id,
				struct timespec *mtime)
{
	struct ceph_msg *msg = req->r_request;
	void *p;
	size_t msg_size;
	int flags = req->r_flags;
	u64 data_len;
	unsigned int i;

	req->r_snapid = snap_id;
	req->r_snapc = ceph_get_snap_context(snapc);

	/* encode request */
	msg->hdr.version = cpu_to_le16(4);

	p = msg->front.iov_base;
	ceph_encode_32(&p, 1);   /* client_inc  is always 1 */
	req->r_request_osdmap_epoch = p;
	p += 4;
	req->r_request_flags = p;
	p += 4;
	if (req->r_flags & CEPH_OSD_FLAG_WRITE)
		ceph_encode_timespec(p, mtime);
	p += sizeof(struct ceph_timespec);
	req->r_request_reassert_version = p;
	p += sizeof(struct ceph_eversion); /* will get filled in */

	/* oloc */
	ceph_encode_8(&p, 4);
	ceph_encode_8(&p, 4);
	ceph_encode_32(&p, 8 + 4 + 4);
	req->r_request_pool = p;
	p += 8;
	ceph_encode_32(&p, -1);  /* preferred */
	ceph_encode_32(&p, 0);   /* key len */

	ceph_encode_8(&p, 1);
	req->r_request_pgid = p;
	p += 8 + 4;
	ceph_encode_32(&p, -1);  /* preferred */

	/* oid */
2433 2434 2435 2436 2437
	ceph_encode_32(&p, req->r_base_oid.name_len);
	memcpy(p, req->r_base_oid.name, req->r_base_oid.name_len);
	dout("oid '%.*s' len %d\n", req->r_base_oid.name_len,
	     req->r_base_oid.name, req->r_base_oid.name_len);
	p += req->r_base_oid.name_len;
2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483

	/* ops--can imply data */
	ceph_encode_16(&p, (u16)req->r_num_ops);
	data_len = 0;
	for (i = 0; i < req->r_num_ops; i++) {
		data_len += osd_req_encode_op(req, p, i);
		p += sizeof(struct ceph_osd_op);
	}

	/* snaps */
	ceph_encode_64(&p, req->r_snapid);
	ceph_encode_64(&p, req->r_snapc ? req->r_snapc->seq : 0);
	ceph_encode_32(&p, req->r_snapc ? req->r_snapc->num_snaps : 0);
	if (req->r_snapc) {
		for (i = 0; i < snapc->num_snaps; i++) {
			ceph_encode_64(&p, req->r_snapc->snaps[i]);
		}
	}

	req->r_request_attempts = p;
	p += 4;

	/* data */
	if (flags & CEPH_OSD_FLAG_WRITE) {
		u16 data_off;

		/*
		 * The header "data_off" is a hint to the receiver
		 * allowing it to align received data into its
		 * buffers such that there's no need to re-copy
		 * it before writing it to disk (direct I/O).
		 */
		data_off = (u16) (off & 0xffff);
		req->r_request->hdr.data_off = cpu_to_le16(data_off);
	}
	req->r_request->hdr.data_len = cpu_to_le32(data_len);

	BUG_ON(p > msg->front.iov_base + msg->front.iov_len);
	msg_size = p - msg->front.iov_base;
	msg->front.iov_len = msg_size;
	msg->hdr.front_len = cpu_to_le32(msg_size);

	dout("build_request msg_size was %d\n", (int)msg_size);
}
EXPORT_SYMBOL(ceph_osdc_build_request);

2484 2485 2486 2487 2488 2489 2490
/*
 * Register request, send initial attempt.
 */
int ceph_osdc_start_request(struct ceph_osd_client *osdc,
			    struct ceph_osd_request *req,
			    bool nofail)
{
2491
	int rc;
2492

S
Sage Weil 已提交
2493 2494
	down_read(&osdc->map_sem);
	mutex_lock(&osdc->request_mutex);
2495 2496 2497

	rc = __ceph_osdc_start_request(osdc, req, nofail);

S
Sage Weil 已提交
2498 2499
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
2500

S
Sage Weil 已提交
2501 2502
	return rc;
}
2503
EXPORT_SYMBOL(ceph_osdc_start_request);
S
Sage Weil 已提交
2504

2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523
/*
 * Unregister a registered request.  The request is not completed (i.e.
 * no callbacks or wakeups) - higher layers are supposed to know what
 * they are canceling.
 */
void ceph_osdc_cancel_request(struct ceph_osd_request *req)
{
	struct ceph_osd_client *osdc = req->r_osdc;

	mutex_lock(&osdc->request_mutex);
	if (req->r_linger)
		__unregister_linger_request(osdc, req);
	__unregister_request(osdc, req);
	mutex_unlock(&osdc->request_mutex);

	dout("%s %p tid %llu canceled\n", __func__, req, req->r_tid);
}
EXPORT_SYMBOL(ceph_osdc_cancel_request);

S
Sage Weil 已提交
2524 2525 2526 2527 2528 2529 2530 2531
/*
 * wait for a request to complete
 */
int ceph_osdc_wait_request(struct ceph_osd_client *osdc,
			   struct ceph_osd_request *req)
{
	int rc;

2532 2533
	dout("%s %p tid %llu\n", __func__, req, req->r_tid);

S
Sage Weil 已提交
2534 2535
	rc = wait_for_completion_interruptible(&req->r_completion);
	if (rc < 0) {
2536 2537
		dout("%s %p tid %llu interrupted\n", __func__, req, req->r_tid);
		ceph_osdc_cancel_request(req);
S
Sage Weil 已提交
2538
		complete_request(req);
S
Sage Weil 已提交
2539 2540 2541
		return rc;
	}

2542 2543
	dout("%s %p tid %llu result %d\n", __func__, req, req->r_tid,
	     req->r_result);
S
Sage Weil 已提交
2544 2545
	return req->r_result;
}
2546
EXPORT_SYMBOL(ceph_osdc_wait_request);
S
Sage Weil 已提交
2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579

/*
 * sync - wait for all in-flight requests to flush.  avoid starvation.
 */
void ceph_osdc_sync(struct ceph_osd_client *osdc)
{
	struct ceph_osd_request *req;
	u64 last_tid, next_tid = 0;

	mutex_lock(&osdc->request_mutex);
	last_tid = osdc->last_tid;
	while (1) {
		req = __lookup_request_ge(osdc, next_tid);
		if (!req)
			break;
		if (req->r_tid > last_tid)
			break;

		next_tid = req->r_tid + 1;
		if ((req->r_flags & CEPH_OSD_FLAG_WRITE) == 0)
			continue;

		ceph_osdc_get_request(req);
		mutex_unlock(&osdc->request_mutex);
		dout("sync waiting on tid %llu (last is %llu)\n",
		     req->r_tid, last_tid);
		wait_for_completion(&req->r_safe_completion);
		mutex_lock(&osdc->request_mutex);
		ceph_osdc_put_request(req);
	}
	mutex_unlock(&osdc->request_mutex);
	dout("sync done (thru tid %llu)\n", last_tid);
}
2580
EXPORT_SYMBOL(ceph_osdc_sync);
S
Sage Weil 已提交
2581

2582 2583 2584 2585
/*
 * Call all pending notify callbacks - for use after a watch is
 * unregistered, to make sure no more callbacks for it will be invoked
 */
S
stephen hemminger 已提交
2586
void ceph_osdc_flush_notifies(struct ceph_osd_client *osdc)
2587 2588 2589 2590 2591 2592
{
	flush_workqueue(osdc->notify_wq);
}
EXPORT_SYMBOL(ceph_osdc_flush_notifies);


S
Sage Weil 已提交
2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608
/*
 * init, shutdown
 */
int ceph_osdc_init(struct ceph_osd_client *osdc, struct ceph_client *client)
{
	int err;

	dout("init\n");
	osdc->client = client;
	osdc->osdmap = NULL;
	init_rwsem(&osdc->map_sem);
	init_completion(&osdc->map_waiters);
	osdc->last_requested_map = 0;
	mutex_init(&osdc->request_mutex);
	osdc->last_tid = 0;
	osdc->osds = RB_ROOT;
2609
	INIT_LIST_HEAD(&osdc->osd_lru);
S
Sage Weil 已提交
2610
	osdc->requests = RB_ROOT;
2611
	INIT_LIST_HEAD(&osdc->req_lru);
2612 2613
	INIT_LIST_HEAD(&osdc->req_unsent);
	INIT_LIST_HEAD(&osdc->req_notarget);
2614
	INIT_LIST_HEAD(&osdc->req_linger);
S
Sage Weil 已提交
2615 2616
	osdc->num_requests = 0;
	INIT_DELAYED_WORK(&osdc->timeout_work, handle_timeout);
2617
	INIT_DELAYED_WORK(&osdc->osds_timeout_work, handle_osds_timeout);
2618 2619 2620
	spin_lock_init(&osdc->event_lock);
	osdc->event_tree = RB_ROOT;
	osdc->event_count = 0;
2621 2622

	schedule_delayed_work(&osdc->osds_timeout_work,
2623
	   round_jiffies_relative(osdc->client->options->osd_idle_ttl * HZ));
S
Sage Weil 已提交
2624

2625
	err = -ENOMEM;
S
Sage Weil 已提交
2626 2627 2628
	osdc->req_mempool = mempool_create_kmalloc_pool(10,
					sizeof(struct ceph_osd_request));
	if (!osdc->req_mempool)
2629
		goto out;
S
Sage Weil 已提交
2630

2631 2632
	err = ceph_msgpool_init(&osdc->msgpool_op, CEPH_MSG_OSD_OP,
				OSD_OP_FRONT_LEN, 10, true,
2633
				"osd_op");
S
Sage Weil 已提交
2634
	if (err < 0)
2635
		goto out_mempool;
2636
	err = ceph_msgpool_init(&osdc->msgpool_op_reply, CEPH_MSG_OSD_OPREPLY,
2637 2638
				OSD_OPREPLY_FRONT_LEN, 10, true,
				"osd_op_reply");
S
Sage Weil 已提交
2639 2640
	if (err < 0)
		goto out_msgpool;
2641

2642
	err = -ENOMEM;
2643
	osdc->notify_wq = create_singlethread_workqueue("ceph-watch-notify");
2644
	if (!osdc->notify_wq)
2645 2646
		goto out_msgpool_reply;

S
Sage Weil 已提交
2647
	return 0;
2648

2649 2650
out_msgpool_reply:
	ceph_msgpool_destroy(&osdc->msgpool_op_reply);
S
Sage Weil 已提交
2651 2652
out_msgpool:
	ceph_msgpool_destroy(&osdc->msgpool_op);
2653 2654 2655 2656
out_mempool:
	mempool_destroy(osdc->req_mempool);
out:
	return err;
S
Sage Weil 已提交
2657 2658 2659 2660
}

void ceph_osdc_stop(struct ceph_osd_client *osdc)
{
2661 2662
	flush_workqueue(osdc->notify_wq);
	destroy_workqueue(osdc->notify_wq);
S
Sage Weil 已提交
2663
	cancel_delayed_work_sync(&osdc->timeout_work);
2664
	cancel_delayed_work_sync(&osdc->osds_timeout_work);
S
Sage Weil 已提交
2665 2666 2667 2668
	if (osdc->osdmap) {
		ceph_osdmap_destroy(osdc->osdmap);
		osdc->osdmap = NULL;
	}
2669
	remove_all_osds(osdc);
S
Sage Weil 已提交
2670 2671
	mempool_destroy(osdc->req_mempool);
	ceph_msgpool_destroy(&osdc->msgpool_op);
S
Sage Weil 已提交
2672
	ceph_msgpool_destroy(&osdc->msgpool_op_reply);
S
Sage Weil 已提交
2673 2674 2675 2676 2677 2678 2679 2680 2681 2682
}

/*
 * Read some contiguous pages.  If we cross a stripe boundary, shorten
 * *plen.  Return number of bytes read, or error.
 */
int ceph_osdc_readpages(struct ceph_osd_client *osdc,
			struct ceph_vino vino, struct ceph_file_layout *layout,
			u64 off, u64 *plen,
			u32 truncate_seq, u64 truncate_size,
2683
			struct page **pages, int num_pages, int page_align)
S
Sage Weil 已提交
2684 2685 2686 2687 2688 2689
{
	struct ceph_osd_request *req;
	int rc = 0;

	dout("readpages on ino %llx.%llx on %llu~%llu\n", vino.ino,
	     vino.snap, off, *plen);
2690
	req = ceph_osdc_new_request(osdc, layout, vino, off, plen, 1,
S
Sage Weil 已提交
2691
				    CEPH_OSD_OP_READ, CEPH_OSD_FLAG_READ,
2692
				    NULL, truncate_seq, truncate_size,
2693
				    false);
2694 2695
	if (IS_ERR(req))
		return PTR_ERR(req);
S
Sage Weil 已提交
2696 2697

	/* it may be a short read due to an object boundary */
2698

2699
	osd_req_op_extent_osd_data_pages(req, 0,
2700
				pages, *plen, page_align, false, false);
S
Sage Weil 已提交
2701

2702
	dout("readpages  final extent is %llu~%llu (%llu bytes align %d)\n",
2703
	     off, *plen, *plen, page_align);
S
Sage Weil 已提交
2704

2705
	ceph_osdc_build_request(req, off, NULL, vino.snap, NULL);
2706

S
Sage Weil 已提交
2707 2708 2709 2710 2711 2712 2713 2714
	rc = ceph_osdc_start_request(osdc, req, false);
	if (!rc)
		rc = ceph_osdc_wait_request(osdc, req);

	ceph_osdc_put_request(req);
	dout("readpages result %d\n", rc);
	return rc;
}
2715
EXPORT_SYMBOL(ceph_osdc_readpages);
S
Sage Weil 已提交
2716 2717 2718 2719 2720 2721 2722 2723 2724 2725

/*
 * do a synchronous write on N pages
 */
int ceph_osdc_writepages(struct ceph_osd_client *osdc, struct ceph_vino vino,
			 struct ceph_file_layout *layout,
			 struct ceph_snap_context *snapc,
			 u64 off, u64 len,
			 u32 truncate_seq, u64 truncate_size,
			 struct timespec *mtime,
2726
			 struct page **pages, int num_pages)
S
Sage Weil 已提交
2727 2728 2729
{
	struct ceph_osd_request *req;
	int rc = 0;
2730
	int page_align = off & ~PAGE_MASK;
S
Sage Weil 已提交
2731

2732
	BUG_ON(vino.snap != CEPH_NOSNAP);	/* snapshots aren't writeable */
2733
	req = ceph_osdc_new_request(osdc, layout, vino, off, &len, 1,
S
Sage Weil 已提交
2734
				    CEPH_OSD_OP_WRITE,
2735
				    CEPH_OSD_FLAG_ONDISK | CEPH_OSD_FLAG_WRITE,
2736
				    snapc, truncate_seq, truncate_size,
2737
				    true);
2738 2739
	if (IS_ERR(req))
		return PTR_ERR(req);
S
Sage Weil 已提交
2740 2741

	/* it may be a short write due to an object boundary */
2742
	osd_req_op_extent_osd_data_pages(req, 0, pages, len, page_align,
2743 2744
				false, false);
	dout("writepages %llu~%llu (%llu bytes)\n", off, len, len);
S
Sage Weil 已提交
2745

2746
	ceph_osdc_build_request(req, off, snapc, CEPH_NOSNAP, mtime);
2747

2748
	rc = ceph_osdc_start_request(osdc, req, true);
S
Sage Weil 已提交
2749 2750 2751 2752 2753 2754 2755 2756 2757
	if (!rc)
		rc = ceph_osdc_wait_request(osdc, req);

	ceph_osdc_put_request(req);
	if (rc == 0)
		rc = len;
	dout("writepages result %d\n", rc);
	return rc;
}
2758
EXPORT_SYMBOL(ceph_osdc_writepages);
S
Sage Weil 已提交
2759

2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779
int ceph_osdc_setup(void)
{
	BUG_ON(ceph_osd_request_cache);
	ceph_osd_request_cache = kmem_cache_create("ceph_osd_request",
					sizeof (struct ceph_osd_request),
					__alignof__(struct ceph_osd_request),
					0, NULL);

	return ceph_osd_request_cache ? 0 : -ENOMEM;
}
EXPORT_SYMBOL(ceph_osdc_setup);

void ceph_osdc_cleanup(void)
{
	BUG_ON(!ceph_osd_request_cache);
	kmem_cache_destroy(ceph_osd_request_cache);
	ceph_osd_request_cache = NULL;
}
EXPORT_SYMBOL(ceph_osdc_cleanup);

S
Sage Weil 已提交
2780 2781 2782 2783 2784 2785
/*
 * handle incoming message
 */
static void dispatch(struct ceph_connection *con, struct ceph_msg *msg)
{
	struct ceph_osd *osd = con->private;
2786
	struct ceph_osd_client *osdc;
S
Sage Weil 已提交
2787 2788 2789
	int type = le16_to_cpu(msg->hdr.type);

	if (!osd)
S
Sage Weil 已提交
2790
		goto out;
2791
	osdc = osd->o_osdc;
S
Sage Weil 已提交
2792 2793 2794 2795 2796 2797

	switch (type) {
	case CEPH_MSG_OSD_MAP:
		ceph_osdc_handle_map(osdc, msg);
		break;
	case CEPH_MSG_OSD_OPREPLY:
2798
		handle_reply(osdc, msg, con);
S
Sage Weil 已提交
2799
		break;
2800 2801 2802
	case CEPH_MSG_WATCH_NOTIFY:
		handle_watch_notify(osdc, msg);
		break;
S
Sage Weil 已提交
2803 2804 2805 2806 2807

	default:
		pr_err("received unknown message type %d %s\n", type,
		       ceph_msg_type_name(type));
	}
S
Sage Weil 已提交
2808
out:
S
Sage Weil 已提交
2809 2810 2811
	ceph_msg_put(msg);
}

2812
/*
2813 2814
 * lookup and return message for incoming reply.  set up reply message
 * pages.
2815 2816
 */
static struct ceph_msg *get_reply(struct ceph_connection *con,
2817 2818
				  struct ceph_msg_header *hdr,
				  int *skip)
S
Sage Weil 已提交
2819 2820 2821
{
	struct ceph_osd *osd = con->private;
	struct ceph_osd_client *osdc = osd->o_osdc;
2822
	struct ceph_msg *m;
2823
	struct ceph_osd_request *req;
2824
	int front_len = le32_to_cpu(hdr->front_len);
2825
	int data_len = le32_to_cpu(hdr->data_len);
2826
	u64 tid;
S
Sage Weil 已提交
2827

2828 2829 2830 2831 2832 2833
	tid = le64_to_cpu(hdr->tid);
	mutex_lock(&osdc->request_mutex);
	req = __lookup_request(osdc, tid);
	if (!req) {
		*skip = 1;
		m = NULL;
2834 2835
		dout("get_reply unknown tid %llu from osd%d\n", tid,
		     osd->o_osd);
2836 2837
		goto out;
	}
S
Sage Weil 已提交
2838

2839
	if (req->r_reply->con)
2840
		dout("%s revoking msg %p from old con %p\n", __func__,
2841 2842
		     req->r_reply, req->r_reply->con);
	ceph_msg_revoke_incoming(req->r_reply);
2843

2844
	if (front_len > req->r_reply->front_alloc_len) {
2845
		pr_warning("get_reply front %d > preallocated %d (%u#%llu)\n",
2846
			   front_len, req->r_reply->front_alloc_len,
2847 2848
			   (unsigned int)con->peer_name.type,
			   le64_to_cpu(con->peer_name.num));
2849 2850
		m = ceph_msg_new(CEPH_MSG_OSD_OPREPLY, front_len, GFP_NOFS,
				 false);
2851
		if (!m)
S
Sage Weil 已提交
2852 2853 2854 2855 2856 2857
			goto out;
		ceph_msg_put(req->r_reply);
		req->r_reply = m;
	}
	m = ceph_msg_get(req->r_reply);

2858
	if (data_len > 0) {
2859
		struct ceph_osd_data *osd_data;
2860

2861 2862 2863 2864 2865
		/*
		 * XXX This is assuming there is only one op containing
		 * XXX page data.  Probably OK for reads, but this
		 * XXX ought to be done more generally.
		 */
2866
		osd_data = osd_req_op_extent_osd_data(req, 0);
2867 2868
		if (osd_data->type == CEPH_OSD_DATA_TYPE_PAGES) {
			if (osd_data->pages &&
2869
				unlikely(osd_data->length < data_len)) {
2870

2871 2872 2873
				pr_warning("tid %lld reply has %d bytes "
					"we had only %llu bytes ready\n",
					tid, data_len, osd_data->length);
2874 2875 2876 2877 2878 2879
				*skip = 1;
				ceph_msg_put(m);
				m = NULL;
				goto out;
			}
		}
2880
	}
2881
	*skip = 0;
S
Sage Weil 已提交
2882
	dout("get_reply tid %lld %p\n", tid, m);
2883 2884 2885

out:
	mutex_unlock(&osdc->request_mutex);
2886
	return m;
2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897

}

static struct ceph_msg *alloc_msg(struct ceph_connection *con,
				  struct ceph_msg_header *hdr,
				  int *skip)
{
	struct ceph_osd *osd = con->private;
	int type = le16_to_cpu(hdr->type);
	int front = le32_to_cpu(hdr->front_len);

A
Alex Elder 已提交
2898
	*skip = 0;
2899 2900
	switch (type) {
	case CEPH_MSG_OSD_MAP:
2901
	case CEPH_MSG_WATCH_NOTIFY:
2902
		return ceph_msg_new(type, front, GFP_NOFS, false);
2903 2904 2905 2906 2907 2908 2909 2910
	case CEPH_MSG_OSD_OPREPLY:
		return get_reply(con, hdr, skip);
	default:
		pr_info("alloc_msg unexpected msg type %d from osd%d\n", type,
			osd->o_osd);
		*skip = 1;
		return NULL;
	}
S
Sage Weil 已提交
2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929
}

/*
 * Wrappers to refcount containing ceph_osd struct
 */
static struct ceph_connection *get_osd_con(struct ceph_connection *con)
{
	struct ceph_osd *osd = con->private;
	if (get_osd(osd))
		return con;
	return NULL;
}

static void put_osd_con(struct ceph_connection *con)
{
	struct ceph_osd *osd = con->private;
	put_osd(osd);
}

2930 2931 2932
/*
 * authentication
 */
2933 2934 2935 2936 2937
/*
 * Note: returned pointer is the address of a structure that's
 * managed separately.  Caller must *not* attempt to free it.
 */
static struct ceph_auth_handshake *get_authorizer(struct ceph_connection *con,
2938
					int *proto, int force_new)
2939 2940 2941 2942
{
	struct ceph_osd *o = con->private;
	struct ceph_osd_client *osdc = o->o_osdc;
	struct ceph_auth_client *ac = osdc->client->monc.auth;
2943
	struct ceph_auth_handshake *auth = &o->o_auth;
2944

2945
	if (force_new && auth->authorizer) {
2946
		ceph_auth_destroy_authorizer(ac, auth->authorizer);
2947 2948
		auth->authorizer = NULL;
	}
2949 2950 2951
	if (!auth->authorizer) {
		int ret = ceph_auth_create_authorizer(ac, CEPH_ENTITY_TYPE_OSD,
						      auth);
2952
		if (ret)
2953
			return ERR_PTR(ret);
2954 2955
	} else {
		int ret = ceph_auth_update_authorizer(ac, CEPH_ENTITY_TYPE_OSD,
2956 2957 2958
						     auth);
		if (ret)
			return ERR_PTR(ret);
2959 2960
	}
	*proto = ac->protocol;
2961

2962
	return auth;
2963 2964 2965 2966 2967 2968 2969 2970 2971
}


static int verify_authorizer_reply(struct ceph_connection *con, int len)
{
	struct ceph_osd *o = con->private;
	struct ceph_osd_client *osdc = o->o_osdc;
	struct ceph_auth_client *ac = osdc->client->monc.auth;

2972
	return ceph_auth_verify_authorizer_reply(ac, o->o_auth.authorizer, len);
2973 2974
}

2975 2976 2977 2978 2979 2980
static int invalidate_authorizer(struct ceph_connection *con)
{
	struct ceph_osd *o = con->private;
	struct ceph_osd_client *osdc = o->o_osdc;
	struct ceph_auth_client *ac = osdc->client->monc.auth;

2981
	ceph_auth_invalidate_authorizer(ac, CEPH_ENTITY_TYPE_OSD);
2982 2983
	return ceph_monc_validate_auth(&osdc->client->monc);
}
2984

2985
static const struct ceph_connection_operations osd_con_ops = {
S
Sage Weil 已提交
2986 2987 2988
	.get = get_osd_con,
	.put = put_osd_con,
	.dispatch = dispatch,
2989 2990
	.get_authorizer = get_authorizer,
	.verify_authorizer_reply = verify_authorizer_reply,
2991
	.invalidate_authorizer = invalidate_authorizer,
S
Sage Weil 已提交
2992
	.alloc_msg = alloc_msg,
2993
	.fault = osd_reset,
S
Sage Weil 已提交
2994
};