osd_client.c 78.2 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
		ceph_osd_data_release(&op->cls.response_data);
		break;
295 296 297 298
	case CEPH_OSD_OP_SETXATTR:
	case CEPH_OSD_OP_CMPXATTR:
		ceph_osd_data_release(&op->xattr.osd_data);
		break;
299 300 301
	default:
		break;
	}
A
Alex Elder 已提交
302 303
}

S
Sage Weil 已提交
304 305 306
/*
 * requests
 */
307
static void ceph_osdc_release_request(struct kref *kref)
S
Sage Weil 已提交
308
{
309 310
	struct ceph_osd_request *req = container_of(kref,
					    struct ceph_osd_request, r_kref);
311
	unsigned int which;
S
Sage Weil 已提交
312

313 314
	dout("%s %p (r_request %p r_reply %p)\n", __func__, req,
	     req->r_request, req->r_reply);
315 316 317 318 319 320
	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);
321

S
Sage Weil 已提交
322 323
	if (req->r_request)
		ceph_msg_put(req->r_request);
324
	if (req->r_reply) {
325
		ceph_msg_revoke_incoming(req->r_reply);
326
		ceph_msg_put(req->r_reply);
327
	}
328

329 330
	for (which = 0; which < req->r_num_ops; which++)
		osd_req_op_data_release(req, which);
331

S
Sage Weil 已提交
332 333 334 335
	ceph_put_snap_context(req->r_snapc);
	if (req->r_mempool)
		mempool_free(req, req->r_osdc->req_mempool);
	else
336 337
		kmem_cache_free(ceph_osd_request_cache, req);

S
Sage Weil 已提交
338
}
339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354

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);
355

356
struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc,
S
Sage Weil 已提交
357
					       struct ceph_snap_context *snapc,
358
					       unsigned int num_ops,
359
					       bool use_mempool,
360
					       gfp_t gfp_flags)
S
Sage Weil 已提交
361 362 363
{
	struct ceph_osd_request *req;
	struct ceph_msg *msg;
364 365
	size_t msg_size;

366 367 368
	BUILD_BUG_ON(CEPH_OSD_MAX_OP > U16_MAX);
	BUG_ON(num_ops > CEPH_OSD_MAX_OP);

369 370 371
	msg_size = 4 + 4 + 8 + 8 + 4+8;
	msg_size += 2 + 4 + 8 + 4 + 4; /* oloc */
	msg_size += 1 + 8 + 4 + 4;     /* pg_t */
372
	msg_size += 4 + CEPH_MAX_OID_NAME_LEN; /* oid */
373 374 375 376 377
	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
378 379

	if (use_mempool) {
380
		req = mempool_alloc(osdc->req_mempool, gfp_flags);
S
Sage Weil 已提交
381 382
		memset(req, 0, sizeof(*req));
	} else {
383
		req = kmem_cache_zalloc(ceph_osd_request_cache, gfp_flags);
S
Sage Weil 已提交
384 385
	}
	if (req == NULL)
386
		return NULL;
S
Sage Weil 已提交
387 388 389

	req->r_osdc = osdc;
	req->r_mempool = use_mempool;
390
	req->r_num_ops = num_ops;
391

S
Sage Weil 已提交
392
	kref_init(&req->r_kref);
S
Sage Weil 已提交
393 394
	init_completion(&req->r_completion);
	init_completion(&req->r_safe_completion);
395
	RB_CLEAR_NODE(&req->r_node);
S
Sage Weil 已提交
396
	INIT_LIST_HEAD(&req->r_unsafe_item);
397
	INIT_LIST_HEAD(&req->r_linger_item);
398
	INIT_LIST_HEAD(&req->r_linger_osd_item);
399
	INIT_LIST_HEAD(&req->r_req_lru_item);
400 401
	INIT_LIST_HEAD(&req->r_osd_item);

402
	req->r_base_oloc.pool = -1;
403
	req->r_target_oloc.pool = -1;
404

S
Sage Weil 已提交
405 406 407 408 409
	/* create reply message */
	if (use_mempool)
		msg = ceph_msgpool_get(&osdc->msgpool_op_reply, 0);
	else
		msg = ceph_msg_new(CEPH_MSG_OSD_OPREPLY,
410
				   OSD_OPREPLY_FRONT_LEN, gfp_flags, true);
411
	if (!msg) {
S
Sage Weil 已提交
412
		ceph_osdc_put_request(req);
413
		return NULL;
S
Sage Weil 已提交
414 415 416 417
	}
	req->r_reply = msg;

	/* create request message; allow space for oid */
S
Sage Weil 已提交
418
	if (use_mempool)
419
		msg = ceph_msgpool_get(&osdc->msgpool_op, 0);
S
Sage Weil 已提交
420
	else
421
		msg = ceph_msg_new(CEPH_MSG_OSD_OP, msg_size, gfp_flags, true);
422
	if (!msg) {
S
Sage Weil 已提交
423
		ceph_osdc_put_request(req);
424
		return NULL;
S
Sage Weil 已提交
425
	}
426

S
Sage Weil 已提交
427
	memset(msg->front.iov_base, 0, msg->front.iov_len);
428 429 430 431 432

	req->r_request = msg;

	return req;
}
433
EXPORT_SYMBOL(ceph_osdc_alloc_request);
434

435
static bool osd_req_opcode_valid(u16 opcode)
436
{
437
	switch (opcode) {
438 439 440
#define GENERATE_CASE(op, opcode, str)	case CEPH_OSD_OP_##op: return true;
__CEPH_FORALL_OSD_OPS(GENERATE_CASE)
#undef GENERATE_CASE
441 442 443 444 445
	default:
		return false;
	}
}

446 447 448 449 450
/*
 * 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.
 */
451
static struct ceph_osd_req_op *
A
Alex Elder 已提交
452
_osd_req_op_init(struct ceph_osd_request *osd_req, unsigned int which,
453
				u16 opcode)
454
{
455 456 457
	struct ceph_osd_req_op *op;

	BUG_ON(which >= osd_req->r_num_ops);
458 459
	BUG_ON(!osd_req_opcode_valid(opcode));

460
	op = &osd_req->r_ops[which];
461 462
	memset(op, 0, sizeof (*op));
	op->op = opcode;
463 464

	return op;
465 466
}

A
Alex Elder 已提交
467 468 469 470 471 472 473
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);

474 475
void osd_req_op_extent_init(struct ceph_osd_request *osd_req,
				unsigned int which, u16 opcode,
476 477 478
				u64 offset, u64 length,
				u64 truncate_size, u32 truncate_seq)
{
A
Alex Elder 已提交
479
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
480 481
	size_t payload_len = 0;

L
Li Wang 已提交
482
	BUG_ON(opcode != CEPH_OSD_OP_READ && opcode != CEPH_OSD_OP_WRITE &&
483
	       opcode != CEPH_OSD_OP_ZERO && opcode != CEPH_OSD_OP_TRUNCATE);
484 485 486 487 488 489 490 491 492 493 494 495

	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);

496 497
void osd_req_op_extent_update(struct ceph_osd_request *osd_req,
				unsigned int which, u64 length)
498
{
499 500 501 502 503 504
	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;
505 506 507 508 509 510 511 512 513 514

	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);

515
void osd_req_op_cls_init(struct ceph_osd_request *osd_req, unsigned int which,
516
			u16 opcode, const char *class, const char *method)
517
{
A
Alex Elder 已提交
518
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
519
	struct ceph_pagelist *pagelist;
520 521 522 523 524
	size_t payload_len = 0;
	size_t size;

	BUG_ON(opcode != CEPH_OSD_OP_CALL);

525 526 527 528
	pagelist = kmalloc(sizeof (*pagelist), GFP_NOFS);
	BUG_ON(!pagelist);
	ceph_pagelist_init(pagelist);

529 530 531 532
	op->cls.class_name = class;
	size = strlen(class);
	BUG_ON(size > (size_t) U8_MAX);
	op->cls.class_len = size;
533
	ceph_pagelist_append(pagelist, class, size);
534 535 536 537 538 539
	payload_len += size;

	op->cls.method_name = method;
	size = strlen(method);
	BUG_ON(size > (size_t) U8_MAX);
	op->cls.method_len = size;
540
	ceph_pagelist_append(pagelist, method, size);
541 542
	payload_len += size;

543
	osd_req_op_cls_request_info_pagelist(osd_req, which, pagelist);
544

545 546 547 548 549
	op->cls.argc = 0;	/* currently unused */

	op->payload_len = payload_len;
}
EXPORT_SYMBOL(osd_req_op_cls_init);
550

551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583
int osd_req_op_xattr_init(struct ceph_osd_request *osd_req, unsigned int which,
			  u16 opcode, const char *name, const void *value,
			  size_t size, u8 cmp_op, u8 cmp_mode)
{
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
	struct ceph_pagelist *pagelist;
	size_t payload_len;

	BUG_ON(opcode != CEPH_OSD_OP_SETXATTR && opcode != CEPH_OSD_OP_CMPXATTR);

	pagelist = kmalloc(sizeof(*pagelist), GFP_NOFS);
	if (!pagelist)
		return -ENOMEM;

	ceph_pagelist_init(pagelist);

	payload_len = strlen(name);
	op->xattr.name_len = payload_len;
	ceph_pagelist_append(pagelist, name, payload_len);

	op->xattr.value_len = size;
	ceph_pagelist_append(pagelist, value, size);
	payload_len += size;

	op->xattr.cmp_op = cmp_op;
	op->xattr.cmp_mode = cmp_mode;

	ceph_osd_data_pagelist_init(&op->xattr.osd_data, pagelist);
	op->payload_len = payload_len;
	return 0;
}
EXPORT_SYMBOL(osd_req_op_xattr_init);

584 585
void osd_req_op_watch_init(struct ceph_osd_request *osd_req,
				unsigned int which, u16 opcode,
586 587
				u64 cookie, u64 version, int flag)
{
A
Alex Elder 已提交
588
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
589

590
	BUG_ON(opcode != CEPH_OSD_OP_NOTIFY_ACK && opcode != CEPH_OSD_OP_WATCH);
591 592

	op->watch.cookie = cookie;
A
Alex Elder 已提交
593
	op->watch.ver = version;
594
	if (opcode == CEPH_OSD_OP_WATCH && flag)
595
		op->watch.flag = (u8)1;
596 597 598
}
EXPORT_SYMBOL(osd_req_op_watch_init);

599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618
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);

619
static void ceph_osdc_msg_data_add(struct ceph_msg *msg,
620 621 622 623 624 625 626
				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)
627
			ceph_msg_data_add_pages(msg, osd_data->pages,
628 629 630
					length, osd_data->alignment);
	} else if (osd_data->type == CEPH_OSD_DATA_TYPE_PAGELIST) {
		BUG_ON(!length);
631
		ceph_msg_data_add_pagelist(msg, osd_data->pagelist);
632 633
#ifdef CONFIG_BLOCK
	} else if (osd_data->type == CEPH_OSD_DATA_TYPE_BIO) {
634
		ceph_msg_data_add_bio(msg, osd_data->bio, length);
635 636 637 638 639 640
#endif
	} else {
		BUG_ON(osd_data->type != CEPH_OSD_DATA_TYPE_NONE);
	}
}

641
static u64 osd_req_encode_op(struct ceph_osd_request *req,
642
			      struct ceph_osd_op *dst, unsigned int which)
643
{
644
	struct ceph_osd_req_op *src;
645
	struct ceph_osd_data *osd_data;
646
	u64 request_data_len = 0;
647
	u64 data_length;
648

649 650
	BUG_ON(which >= req->r_num_ops);
	src = &req->r_ops[which];
651 652 653 654 655 656 657 658
	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
659 660
		osd_data = &src->raw_data_in;
		ceph_osdc_msg_data_add(req->r_reply, osd_data);
661 662 663
		break;
	case CEPH_OSD_OP_READ:
	case CEPH_OSD_OP_WRITE:
L
Li Wang 已提交
664 665
	case CEPH_OSD_OP_ZERO:
	case CEPH_OSD_OP_TRUNCATE:
666
		if (src->op == CEPH_OSD_OP_WRITE)
667
			request_data_len = src->extent.length;
668 669 670 671 672 673
		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);
674
		osd_data = &src->extent.osd_data;
675
		if (src->op == CEPH_OSD_OP_WRITE)
676
			ceph_osdc_msg_data_add(req->r_request, osd_data);
677
		else
678
			ceph_osdc_msg_data_add(req->r_reply, osd_data);
679 680 681 682
		break;
	case CEPH_OSD_OP_CALL:
		dst->cls.class_len = src->cls.class_len;
		dst->cls.method_len = src->cls.method_len;
683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698
		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);
699 700 701 702 703 704 705 706 707
		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;
708 709 710 711 712 713
	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;
714 715 716 717 718 719 720 721 722 723
	case CEPH_OSD_OP_SETXATTR:
	case CEPH_OSD_OP_CMPXATTR:
		dst->xattr.name_len = cpu_to_le32(src->xattr.name_len);
		dst->xattr.value_len = cpu_to_le32(src->xattr.value_len);
		dst->xattr.cmp_op = src->xattr.cmp_op;
		dst->xattr.cmp_mode = src->xattr.cmp_mode;
		osd_data = &src->xattr.osd_data;
		ceph_osdc_msg_data_add(req->r_request, osd_data);
		request_data_len = osd_data->pagelist->length;
		break;
724 725 726
	case CEPH_OSD_OP_CREATE:
	case CEPH_OSD_OP_DELETE:
		break;
727
	default:
728
		pr_err("unsupported osd opcode %s\n",
729
			ceph_osd_op_name(src->op));
730
		WARN_ON(1);
731 732

		return 0;
733
	}
734

735
	dst->op = cpu_to_le16(src->op);
736
	dst->flags = cpu_to_le32(src->flags);
737
	dst->payload_len = cpu_to_le32(src->payload_len);
738

739
	return request_data_len;
740 741
}

742 743 744 745 746 747 748 749 750 751 752 753 754 755
/*
 * 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,
756
					       u64 off, u64 *plen, int num_ops,
757 758 759 760
					       int opcode, int flags,
					       struct ceph_snap_context *snapc,
					       u32 truncate_seq,
					       u64 truncate_size,
761
					       bool use_mempool)
762
{
763
	struct ceph_osd_request *req;
764 765 766
	u64 objnum = 0;
	u64 objoff = 0;
	u64 objlen = 0;
767
	int r;
768

L
Li Wang 已提交
769
	BUG_ON(opcode != CEPH_OSD_OP_READ && opcode != CEPH_OSD_OP_WRITE &&
770 771
	       opcode != CEPH_OSD_OP_ZERO && opcode != CEPH_OSD_OP_TRUNCATE &&
	       opcode != CEPH_OSD_OP_CREATE && opcode != CEPH_OSD_OP_DELETE);
772

773
	req = ceph_osdc_alloc_request(osdc, snapc, num_ops, use_mempool,
A
Alex Elder 已提交
774
					GFP_NOFS);
775
	if (!req)
776
		return ERR_PTR(-ENOMEM);
777

778
	req->r_flags = flags;
779 780

	/* calculate max write size */
781
	r = calc_layout(layout, off, plen, &objnum, &objoff, &objlen);
782 783
	if (r < 0) {
		ceph_osdc_put_request(req);
784
		return ERR_PTR(r);
785
	}
786

787 788 789 790 791 792 793 794 795 796 797 798 799
	if (opcode == CEPH_OSD_OP_CREATE || opcode == CEPH_OSD_OP_DELETE) {
		osd_req_op_init(req, 0, opcode);
	} else {
		u32 object_size = le32_to_cpu(layout->fl_object_size);
		u32 object_base = off - objoff;
		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;
			}
800
		}
801

802 803 804
		osd_req_op_extent_init(req, 0, opcode, objoff, objlen,
				       truncate_size, truncate_seq);
	}
805 806 807 808 809 810
	/*
	 * 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)
811
		osd_req_op_init(req, 1, CEPH_OSD_OP_STARTSYNC);
812

813
	req->r_base_oloc.pool = ceph_file_layout_pg_pool(*layout);
814

815
	snprintf(req->r_base_oid.name, sizeof(req->r_base_oid.name),
816
		 "%llx.%08llx", vino.ino, objnum);
817
	req->r_base_oid.name_len = strlen(req->r_base_oid.name);
818

S
Sage Weil 已提交
819 820
	return req;
}
821
EXPORT_SYMBOL(ceph_osdc_new_request);
S
Sage Weil 已提交
822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887

/*
 * 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;
}

888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918
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);
}

919 920 921 922 923 924
/*
 * Resubmit requests pending on the given osd.
 */
static void __kick_osd_requests(struct ceph_osd_client *osdc,
				struct ceph_osd *osd)
{
925
	struct ceph_osd_request *req, *nreq;
926
	LIST_HEAD(resend);
927
	LIST_HEAD(resend_linger);
928 929
	int err;

930
	dout("%s osd%d\n", __func__, osd->o_osd);
931
	err = __reset_osd(osdc, osd);
932
	if (err)
933
		return;
934

935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951
	/*
	 * 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.
	 */
952
	list_for_each_entry(req, &osd->o_requests, r_osd_item) {
953 954
		if (!req->r_sent)
			break;
955 956 957 958 959

		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);
960
			req->r_flags |= CEPH_OSD_FLAG_RETRY;
961 962 963
		} else {
			list_move_tail(&req->r_req_lru_item, &resend_linger);
		}
964
	}
965
	list_splice(&resend, &osdc->req_unsent);
966

967
	/*
968 969 970 971
	 * 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.
972
	 */
973
	list_for_each_entry_safe(req, nreq, &osd->o_linger_requests,
974
				 r_linger_osd_item) {
975 976
		WARN_ON(!list_empty(&req->r_req_lru_item));
		__kick_linger_request(req);
977
	}
978 979 980

	list_for_each_entry_safe(req, nreq, &resend_linger, r_req_lru_item)
		__kick_linger_request(req);
981 982
}

S
Sage Weil 已提交
983
/*
984
 * If the osd connection drops, we need to resubmit all requests.
S
Sage Weil 已提交
985 986 987 988 989 990 991 992 993 994 995
 */
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
996 997
	mutex_lock(&osdc->request_mutex);
	__kick_osd_requests(osdc, osd);
A
Alex Elder 已提交
998
	__send_queued(osdc);
S
Sage Weil 已提交
999
	mutex_unlock(&osdc->request_mutex);
S
Sage Weil 已提交
1000 1001 1002 1003 1004 1005
	up_read(&osdc->map_sem);
}

/*
 * Track open sessions with osds.
 */
1006
static struct ceph_osd *create_osd(struct ceph_osd_client *osdc, int onum)
S
Sage Weil 已提交
1007 1008 1009 1010 1011 1012 1013 1014 1015
{
	struct ceph_osd *osd;

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

	atomic_set(&osd->o_ref, 1);
	osd->o_osdc = osdc;
1016
	osd->o_osd = onum;
1017
	RB_CLEAR_NODE(&osd->o_node);
S
Sage Weil 已提交
1018
	INIT_LIST_HEAD(&osd->o_requests);
1019
	INIT_LIST_HEAD(&osd->o_linger_requests);
1020
	INIT_LIST_HEAD(&osd->o_osd_lru);
S
Sage Weil 已提交
1021 1022
	osd->o_incarnation = 1;

1023
	ceph_con_init(&osd->o_con, osd, &osd_con_ops, &osdc->client->msgr);
1024

1025
	INIT_LIST_HEAD(&osd->o_keepalive_item);
S
Sage Weil 已提交
1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044
	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);
1045
	if (atomic_dec_and_test(&osd->o_ref) && osd->o_auth.authorizer) {
S
Sage Weil 已提交
1046 1047
		struct ceph_auth_client *ac = osd->o_osdc->client->monc.auth;

1048
		ceph_auth_destroy_authorizer(ac, osd->o_auth.authorizer);
S
Sage Weil 已提交
1049
		kfree(osd);
S
Sage Weil 已提交
1050
	}
S
Sage Weil 已提交
1051 1052 1053 1054 1055
}

/*
 * remove an osd from our map
 */
1056
static void __remove_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
S
Sage Weil 已提交
1057
{
1058
	dout("__remove_osd %p\n", osd);
1059 1060
	WARN_ON(!list_empty(&osd->o_requests));
	WARN_ON(!list_empty(&osd->o_linger_requests));
1061

S
Sage Weil 已提交
1062
	rb_erase(&osd->o_node, &osdc->osds);
1063
	list_del_init(&osd->o_osd_lru);
S
Sage Weil 已提交
1064 1065 1066 1067
	ceph_con_close(&osd->o_con);
	put_osd(osd);
}

1068 1069
static void remove_all_osds(struct ceph_osd_client *osdc)
{
1070
	dout("%s %p\n", __func__, osdc);
1071 1072 1073 1074 1075 1076 1077 1078 1079
	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);
}

1080 1081 1082
static void __move_osd_to_lru(struct ceph_osd_client *osdc,
			      struct ceph_osd *osd)
{
1083
	dout("%s %p\n", __func__, osd);
1084
	BUG_ON(!list_empty(&osd->o_osd_lru));
1085

1086
	list_add_tail(&osd->o_osd_lru, &osdc->osd_lru);
1087
	osd->lru_ttl = jiffies + osdc->client->options->osd_idle_ttl * HZ;
1088 1089
}

1090 1091 1092 1093 1094 1095 1096 1097 1098 1099
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);
}

1100 1101 1102 1103 1104 1105 1106
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);
}

1107
static void remove_old_osds(struct ceph_osd_client *osdc)
1108 1109 1110 1111 1112 1113
{
	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) {
1114
		if (time_before(jiffies, osd->lru_ttl))
1115 1116 1117 1118 1119 1120
			break;
		__remove_osd(osdc, osd);
	}
	mutex_unlock(&osdc->request_mutex);
}

S
Sage Weil 已提交
1121 1122 1123
/*
 * reset osd connect
 */
1124
static int __reset_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
S
Sage Weil 已提交
1125
{
A
Alex Elder 已提交
1126
	struct ceph_entity_addr *peer_addr;
S
Sage Weil 已提交
1127

1128
	dout("__reset_osd %p osd%d\n", osd, osd->o_osd);
1129 1130
	if (list_empty(&osd->o_requests) &&
	    list_empty(&osd->o_linger_requests)) {
1131
		__remove_osd(osdc, osd);
A
Alex Elder 已提交
1132 1133 1134 1135 1136 1137 1138 1139 1140

		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;

1141 1142
		dout("osd addr hasn't changed and connection never opened, "
		     "letting msgr retry\n");
1143 1144 1145
		/* 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
1146 1147

		return -EAGAIN;
S
Sage Weil 已提交
1148
	}
A
Alex Elder 已提交
1149 1150 1151 1152 1153 1154

	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
1155 1156 1157 1158 1159 1160 1161 1162
}

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;

1163
	dout("__insert_osd %p osd%d\n", new, new->o_osd);
S
Sage Weil 已提交
1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195
	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;
}

1196 1197 1198
static void __schedule_osd_timeout(struct ceph_osd_client *osdc)
{
	schedule_delayed_work(&osdc->timeout_work,
1199
			osdc->client->options->osd_keepalive_timeout * HZ);
1200 1201 1202 1203 1204 1205
}

static void __cancel_osd_timeout(struct ceph_osd_client *osdc)
{
	cancel_delayed_work(&osdc->timeout_work);
}
S
Sage Weil 已提交
1206 1207 1208 1209 1210

/*
 * Register request, assign tid.  If this is the first request, set up
 * the timeout event.
 */
1211 1212
static void __register_request(struct ceph_osd_client *osdc,
			       struct ceph_osd_request *req)
S
Sage Weil 已提交
1213 1214
{
	req->r_tid = ++osdc->last_tid;
1215
	req->r_request->hdr.tid = cpu_to_le64(req->r_tid);
1216
	dout("__register_request %p tid %lld\n", req, req->r_tid);
S
Sage Weil 已提交
1217 1218 1219 1220
	__insert_request(osdc, req);
	ceph_osdc_get_request(req);
	osdc->num_requests++;
	if (osdc->num_requests == 1) {
1221 1222
		dout(" first request, scheduling timeout\n");
		__schedule_osd_timeout(osdc);
S
Sage Weil 已提交
1223
	}
1224 1225
}

S
Sage Weil 已提交
1226 1227 1228 1229 1230 1231
/*
 * called under osdc->request_mutex
 */
static void __unregister_request(struct ceph_osd_client *osdc,
				 struct ceph_osd_request *req)
{
1232 1233 1234 1235 1236 1237
	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
1238 1239
	dout("__unregister_request %p tid %lld\n", req, req->r_tid);
	rb_erase(&req->r_node, &osdc->requests);
1240
	RB_CLEAR_NODE(&req->r_node);
S
Sage Weil 已提交
1241 1242
	osdc->num_requests--;

1243 1244
	if (req->r_osd) {
		/* make sure the original request isn't in flight. */
1245
		ceph_msg_revoke(req->r_request);
1246 1247

		list_del_init(&req->r_osd_item);
1248
		maybe_move_osd_to_lru(osdc, req->r_osd);
1249
		if (list_empty(&req->r_linger_osd_item))
1250
			req->r_osd = NULL;
1251
	}
S
Sage Weil 已提交
1252

A
Alex Elder 已提交
1253
	list_del_init(&req->r_req_lru_item);
S
Sage Weil 已提交
1254 1255
	ceph_osdc_put_request(req);

1256 1257 1258
	if (osdc->num_requests == 0) {
		dout(" no requests, canceling timeout\n");
		__cancel_osd_timeout(osdc);
S
Sage Weil 已提交
1259 1260 1261 1262 1263 1264 1265 1266
	}
}

/*
 * Cancel a previously queued request message
 */
static void __cancel_request(struct ceph_osd_request *req)
{
1267
	if (req->r_sent && req->r_osd) {
1268
		ceph_msg_revoke(req->r_request);
S
Sage Weil 已提交
1269 1270 1271 1272
		req->r_sent = 0;
	}
}

1273 1274 1275
static void __register_linger_request(struct ceph_osd_client *osdc,
				    struct ceph_osd_request *req)
{
1276 1277 1278
	dout("%s %p tid %llu\n", __func__, req, req->r_tid);
	WARN_ON(!req->r_linger);

1279
	ceph_osdc_get_request(req);
1280
	list_add_tail(&req->r_linger_item, &osdc->req_linger);
1281
	if (req->r_osd)
1282
		list_add_tail(&req->r_linger_osd_item,
1283
			      &req->r_osd->o_linger_requests);
1284 1285 1286 1287 1288
}

static void __unregister_linger_request(struct ceph_osd_client *osdc,
					struct ceph_osd_request *req)
{
1289 1290 1291 1292 1293 1294 1295 1296 1297
	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
1298
	list_del_init(&req->r_linger_item);
1299

1300
	if (req->r_osd) {
1301
		list_del_init(&req->r_linger_osd_item);
1302
		maybe_move_osd_to_lru(osdc, req->r_osd);
1303 1304
		if (list_empty(&req->r_osd_item))
			req->r_osd = NULL;
1305
	}
1306 1307

	list_del_init(&req->r_req_lru_item); /* can be on notarget */
1308
	ceph_osdc_put_request(req);
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320
}

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);

1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336
/*
 * 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);
}

1337 1338 1339 1340 1341 1342 1343
/*
 * 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)
{
1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357
	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) {
1358 1359
		struct ceph_pg_pool_info *pi;

1360
		pi = ceph_pg_pool_by_id(osdmap, req->r_target_oloc.pool);
1361 1362 1363
		if (pi) {
			if ((req->r_flags & CEPH_OSD_FLAG_READ) &&
			    pi->read_tier >= 0)
1364
				req->r_target_oloc.pool = pi->read_tier;
1365 1366
			if ((req->r_flags & CEPH_OSD_FLAG_WRITE) &&
			    pi->write_tier >= 0)
1367
				req->r_target_oloc.pool = pi->write_tier;
1368 1369 1370 1371
		}
		/* !pi is caught in ceph_oloc_oid_to_pg() */
	}

1372 1373
	return ceph_oloc_oid_to_pg(osdmap, &req->r_target_oloc,
				   &req->r_target_oid, pg_out);
1374 1375
}

1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391
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
1392 1393 1394
/*
 * 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
1395
 * no up osd, set r_osd to NULL.  Move the request to the appropriate list
1396
 * (unsent, homeless) or leave on in-flight lru.
S
Sage Weil 已提交
1397 1398 1399 1400 1401
 *
 * Return 0 if unchanged, 1 if changed, or negative on error.
 *
 * Caller should hold map_sem for read and request_mutex.
 */
1402
static int __map_request(struct ceph_osd_client *osdc,
1403
			 struct ceph_osd_request *req, int force_resend)
S
Sage Weil 已提交
1404
{
1405
	struct ceph_pg pgid;
1406
	int acting[CEPH_PG_MAX_SIZE];
1407
	int num, o;
S
Sage Weil 已提交
1408
	int err;
1409
	bool was_paused;
S
Sage Weil 已提交
1410

1411
	dout("map_request %p tid %lld\n", req, req->r_tid);
1412 1413

	err = __calc_request_pg(osdc->osdmap, req, &pgid);
1414 1415
	if (err) {
		list_move(&req->r_req_lru_item, &osdc->req_notarget);
S
Sage Weil 已提交
1416
		return err;
1417
	}
1418 1419
	req->r_pgid = pgid;

1420 1421 1422
	num = ceph_calc_pg_acting(osdc->osdmap, pgid, acting, &o);
	if (num < 0)
		num = 0;
S
Sage Weil 已提交
1423

1424 1425 1426 1427 1428
	was_paused = req->r_paused;
	req->r_paused = __req_should_be_paused(osdc, req);
	if (was_paused && !req->r_paused)
		force_resend = 1;

1429 1430
	if ((!force_resend &&
	     req->r_osd && req->r_osd->o_osd == o &&
1431 1432 1433
	     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) ||
1434 1435
	    (req->r_osd == NULL && o == -1) ||
	    req->r_paused)
S
Sage Weil 已提交
1436 1437
		return 0;  /* no change */

1438 1439
	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
1440 1441
	     req->r_osd ? req->r_osd->o_osd : -1);

1442 1443 1444 1445
	/* 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
1446 1447 1448
	if (req->r_osd) {
		__cancel_request(req);
		list_del_init(&req->r_osd_item);
1449
		list_del_init(&req->r_linger_osd_item);
S
Sage Weil 已提交
1450 1451 1452 1453 1454
		req->r_osd = NULL;
	}

	req->r_osd = __lookup_osd(osdc, o);
	if (!req->r_osd && o >= 0) {
1455
		err = -ENOMEM;
1456
		req->r_osd = create_osd(osdc, o);
1457 1458
		if (!req->r_osd) {
			list_move(&req->r_req_lru_item, &osdc->req_notarget);
1459
			goto out;
1460
		}
S
Sage Weil 已提交
1461

1462
		dout("map_request osd %p is osd%d\n", req->r_osd, o);
S
Sage Weil 已提交
1463 1464
		__insert_osd(osdc, req->r_osd);

1465 1466 1467
		ceph_con_open(&req->r_osd->o_con,
			      CEPH_ENTITY_TYPE_OSD, o,
			      &osdc->osdmap->osd_addr[o]);
S
Sage Weil 已提交
1468 1469
	}

1470
	__enqueue_request(req);
1471
	err = 1;   /* osd or pg changed */
S
Sage Weil 已提交
1472 1473 1474 1475 1476 1477 1478 1479

out:
	return err;
}

/*
 * caller should hold map_sem (for read) and request_mutex
 */
1480 1481
static void __send_request(struct ceph_osd_client *osdc,
			   struct ceph_osd_request *req)
S
Sage Weil 已提交
1482
{
1483
	void *p;
S
Sage Weil 已提交
1484

1485 1486 1487 1488 1489 1490 1491
	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);
1492
	put_unaligned_le64(req->r_target_oloc.pool, req->r_request_pool);
1493 1494 1495 1496 1497 1498
	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));
1499

S
Sage Weil 已提交
1500
	req->r_stamp = jiffies;
1501
	list_move_tail(&req->r_req_lru_item, &osdc->req_lru);
S
Sage Weil 已提交
1502 1503

	ceph_msg_get(req->r_request); /* send consumes a ref */
1504

S
Sage Weil 已提交
1505
	req->r_sent = req->r_osd->o_incarnation;
1506 1507

	ceph_con_send(&req->r_osd->o_con, req->r_request);
S
Sage Weil 已提交
1508 1509
}

1510 1511 1512
/*
 * Send any requests in the queue (req_unsent).
 */
A
Alex Elder 已提交
1513
static void __send_queued(struct ceph_osd_client *osdc)
1514 1515 1516
{
	struct ceph_osd_request *req, *tmp;

A
Alex Elder 已提交
1517 1518
	dout("__send_queued\n");
	list_for_each_entry_safe(req, tmp, &osdc->req_unsent, r_req_lru_item)
1519 1520 1521
		__send_request(osdc, req);
}

1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555
/*
 * 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
1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568
/*
 * 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
1569
	struct ceph_osd_request *req;
S
Sage Weil 已提交
1570
	struct ceph_osd *osd;
1571
	unsigned long keepalive =
1572
		osdc->client->options->osd_keepalive_timeout * HZ;
1573
	struct list_head slow_osds;
S
Sage Weil 已提交
1574 1575 1576 1577 1578 1579 1580
	dout("timeout\n");
	down_read(&osdc->map_sem);

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

	mutex_lock(&osdc->request_mutex);

1581 1582 1583 1584 1585 1586 1587
	/*
	 * 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
1588
		if (time_before(jiffies, req->r_stamp + keepalive))
1589 1590 1591 1592 1593
			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
1594
		     req->r_tid, osd->o_osd);
1595 1596 1597 1598 1599 1600
		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
1601 1602 1603
		ceph_con_keepalive(&osd->o_con);
	}

1604
	__schedule_osd_timeout(osdc);
A
Alex Elder 已提交
1605
	__send_queued(osdc);
S
Sage Weil 已提交
1606 1607 1608 1609
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
}

1610 1611 1612 1613 1614 1615
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 =
1616
		osdc->client->options->osd_idle_ttl * HZ >> 2;
1617 1618 1619

	dout("osds timeout\n");
	down_read(&osdc->map_sem);
1620
	remove_old_osds(osdc);
1621 1622 1623 1624 1625 1626
	up_read(&osdc->map_sem);

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

1627 1628 1629 1630 1631 1632 1633 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
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
1730 1731 1732 1733 1734
static void complete_request(struct ceph_osd_request *req)
{
	complete_all(&req->r_safe_completion);  /* fsync waiter */
}

S
Sage Weil 已提交
1735 1736 1737 1738
/*
 * handle osd op reply.  either call the callback if it is specified,
 * or do the completion to wake up the waiting thread.
 */
1739 1740
static void handle_reply(struct ceph_osd_client *osdc, struct ceph_msg *msg,
			 struct ceph_connection *con)
S
Sage Weil 已提交
1741
{
1742
	void *p, *end;
S
Sage Weil 已提交
1743
	struct ceph_osd_request *req;
1744
	struct ceph_request_redirect redir;
S
Sage Weil 已提交
1745
	u64 tid;
1746
	int object_len;
1747 1748
	unsigned int numops;
	int payload_len, flags;
1749
	s32 result;
1750 1751 1752 1753 1754 1755
	s32 retry_attempt;
	struct ceph_pg pg;
	int err;
	u32 reassert_epoch;
	u64 reassert_version;
	u32 osdmap_epoch;
1756
	int already_completed;
1757
	u32 bytes;
1758
	unsigned int i;
S
Sage Weil 已提交
1759

1760
	tid = le64_to_cpu(msg->hdr.tid);
1761 1762 1763 1764 1765 1766 1767 1768 1769 1770
	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;

1771
	err = ceph_decode_pgid(&p, end, &pg);
1772
	if (err)
S
Sage Weil 已提交
1773
		goto bad;
1774 1775 1776 1777 1778 1779 1780 1781

	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
1782
	/* lookup */
1783
	down_read(&osdc->map_sem);
S
Sage Weil 已提交
1784 1785 1786 1787
	mutex_lock(&osdc->request_mutex);
	req = __lookup_request(osdc, tid);
	if (req == NULL) {
		dout("handle_reply tid %llu dne\n", tid);
1788
		goto bad_mutex;
S
Sage Weil 已提交
1789 1790
	}
	ceph_osdc_get_request(req);
1791 1792 1793 1794

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

1795
	ceph_decode_need(&p, end, 4, bad_put);
1796 1797 1798 1799 1800 1801
	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;
1802
	ceph_decode_need(&p, end, numops * sizeof(struct ceph_osd_op), bad_put);
1803 1804 1805 1806 1807 1808 1809 1810 1811 1812
	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);
	}
1813 1814
	bytes = le32_to_cpu(msg->hdr.data_len);
	if (payload_len != bytes) {
1815 1816
		pr_warn("sum of op payload lens %d != data_len %d\n",
			payload_len, bytes);
1817 1818 1819
		goto bad_put;
	}

1820
	ceph_decode_need(&p, end, 4 + numops * 4, bad_put);
1821 1822 1823
	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
1824

1825 1826 1827
	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
1828

1829 1830 1831 1832 1833 1834
		err = ceph_redirect_decode(&p, end, &redir);
		if (err)
			goto bad_put;
	} else {
		redir.oloc.pool = -1;
	}
S
Sage Weil 已提交
1835

1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850
	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.
		 */
1851
		err = __ceph_osdc_start_request(osdc, req, true);
1852 1853
		BUG_ON(err);

1854
		goto out_unlock;
1855 1856 1857 1858
	}

	already_completed = req->r_got_reply;
	if (!req->r_got_reply) {
1859
		req->r_result = result;
S
Sage Weil 已提交
1860 1861 1862 1863 1864 1865
		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, */
1866 1867
		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
1868 1869 1870 1871

		req->r_got_reply = 1;
	} else if ((flags & CEPH_OSD_FLAG_ONDISK) == 0) {
		dout("handle_reply tid %llu dup ack\n", tid);
1872
		goto out_unlock;
S
Sage Weil 已提交
1873 1874 1875 1876
	}

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

1877 1878 1879
	if (req->r_linger && (flags & CEPH_OSD_FLAG_ONDISK))
		__register_linger_request(osdc, req);

S
Sage Weil 已提交
1880
	/* either this is a read, or we got the safe response */
1881 1882
	if (result < 0 ||
	    (flags & CEPH_OSD_FLAG_ONDISK) ||
S
Sage Weil 已提交
1883 1884 1885 1886
	    ((flags & CEPH_OSD_FLAG_WRITE) == 0))
		__unregister_request(osdc, req);

	mutex_unlock(&osdc->request_mutex);
1887
	up_read(&osdc->map_sem);
S
Sage Weil 已提交
1888

Y
Yan, Zheng 已提交
1889
	if (!already_completed) {
1890 1891 1892
		if (req->r_unsafe_callback &&
		    result >= 0 && !(flags & CEPH_OSD_FLAG_ONDISK))
			req->r_unsafe_callback(req, true);
Y
Yan, Zheng 已提交
1893 1894 1895 1896 1897
		if (req->r_callback)
			req->r_callback(req, msg);
		else
			complete_all(&req->r_completion);
	}
S
Sage Weil 已提交
1898

1899 1900 1901
	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
1902
		complete_request(req);
1903
	}
S
Sage Weil 已提交
1904

1905
out:
1906
	dout("req=%p req->r_linger=%d\n", req, req->r_linger);
S
Sage Weil 已提交
1907 1908
	ceph_osdc_put_request(req);
	return;
1909 1910 1911 1912
out_unlock:
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
	goto out;
S
Sage Weil 已提交
1913

1914
bad_put:
1915 1916 1917 1918 1919 1920 1921
	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);
1922
	ceph_osdc_put_request(req);
1923 1924
bad_mutex:
	mutex_unlock(&osdc->request_mutex);
1925
	up_read(&osdc->map_sem);
S
Sage Weil 已提交
1926
bad:
1927 1928
	pr_err("corrupt osd_op_reply got %d %d\n",
	       (int)msg->front.iov_len, le32_to_cpu(msg->hdr.front_len));
1929
	ceph_msg_dump(msg);
S
Sage Weil 已提交
1930 1931
}

1932
static void reset_changed_osds(struct ceph_osd_client *osdc)
S
Sage Weil 已提交
1933 1934 1935
{
	struct rb_node *p, *n;

1936 1937
	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
1938

1939 1940 1941 1942 1943 1944 1945
		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
1946
	}
1947 1948 1949
}

/*
1950 1951
 * Requeue requests whose mapping to an OSD has changed.  If requests map to
 * no osd, request a new map.
1952
 *
1953
 * Caller should hold map_sem for read.
1954
 */
1955 1956
static void kick_requests(struct ceph_osd_client *osdc, bool force_resend,
			  bool force_resend_writes)
1957
{
1958
	struct ceph_osd_request *req, *nreq;
1959 1960 1961
	struct rb_node *p;
	int needmap = 0;
	int err;
1962
	bool force_resend_req;
1963

1964 1965
	dout("kick_requests %s %s\n", force_resend ? " (force resend)" : "",
		force_resend_writes ? " (force resend writes)" : "");
1966
	mutex_lock(&osdc->request_mutex);
1967
	for (p = rb_first(&osdc->requests); p; ) {
1968
		req = rb_entry(p, struct ceph_osd_request, r_node);
1969
		p = rb_next(p);
1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982

		/*
		 * 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);
1983
			ceph_osdc_get_request(req);
1984 1985
			__unregister_request(osdc, req);
			__register_linger_request(osdc, req);
1986
			ceph_osdc_put_request(req);
1987 1988 1989
			continue;
		}

1990 1991 1992 1993
		force_resend_req = force_resend ||
			(force_resend_writes &&
				req->r_flags & CEPH_OSD_FLAG_WRITE);
		err = __map_request(osdc, req, force_resend_req);
1994 1995 1996 1997 1998 1999
		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) {
2000 2001 2002 2003
			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);
2004
				req->r_flags |= CEPH_OSD_FLAG_RETRY;
2005 2006
			}
		}
2007 2008 2009 2010 2011 2012
	}

	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);

2013 2014
		err = __map_request(osdc, req,
				    force_resend || force_resend_writes);
2015
		dout("__map_request returned %d\n", err);
2016 2017 2018 2019 2020 2021 2022 2023
		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;
2024
		}
2025 2026 2027 2028

		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);
2029
		__unregister_linger_request(osdc, req);
2030
	}
2031
	reset_changed_osds(osdc);
S
Sage Weil 已提交
2032 2033 2034 2035 2036 2037
	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);
	}
2038
}
2039 2040


S
Sage Weil 已提交
2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055
/*
 * 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;
2056
	bool was_full;
S
Sage Weil 已提交
2057 2058 2059 2060 2061 2062 2063 2064

	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));
2065 2066
	if (ceph_check_fsid(osdc->client, &fsid) < 0)
		return;
S
Sage Weil 已提交
2067 2068 2069

	down_write(&osdc->map_sem);

2070 2071
	was_full = ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_FULL);

S
Sage Weil 已提交
2072 2073 2074 2075 2076
	/* 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);
2077 2078
		epoch = ceph_decode_32(&p);
		maplen = ceph_decode_32(&p);
S
Sage Weil 已提交
2079 2080 2081 2082 2083 2084 2085
		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,
2086
							  &osdc->client->msgr);
S
Sage Weil 已提交
2087 2088 2089 2090
			if (IS_ERR(newmap)) {
				err = PTR_ERR(newmap);
				goto bad;
			}
2091
			BUG_ON(!newmap);
S
Sage Weil 已提交
2092 2093 2094 2095
			if (newmap != osdc->osdmap) {
				ceph_osdmap_destroy(osdc->osdmap);
				osdc->osdmap = newmap;
			}
2096 2097 2098 2099
			was_full = was_full ||
				ceph_osdmap_flag(osdc->osdmap,
						 CEPH_OSDMAP_FULL);
			kick_requests(osdc, 0, was_full);
S
Sage Weil 已提交
2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114
		} 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);
2115 2116
		epoch = ceph_decode_32(&p);
		maplen = ceph_decode_32(&p);
S
Sage Weil 已提交
2117 2118 2119 2120 2121 2122 2123 2124 2125
		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 {
2126 2127
			int skipped_map = 0;

S
Sage Weil 已提交
2128
			dout("taking full map %u len %d\n", epoch, maplen);
2129
			newmap = ceph_osdmap_decode(&p, p+maplen);
S
Sage Weil 已提交
2130 2131 2132 2133
			if (IS_ERR(newmap)) {
				err = PTR_ERR(newmap);
				goto bad;
			}
2134
			BUG_ON(!newmap);
S
Sage Weil 已提交
2135 2136
			oldmap = osdc->osdmap;
			osdc->osdmap = newmap;
2137 2138 2139
			if (oldmap) {
				if (oldmap->epoch + 1 < newmap->epoch)
					skipped_map = 1;
S
Sage Weil 已提交
2140
				ceph_osdmap_destroy(oldmap);
2141
			}
2142 2143 2144 2145
			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
2146 2147 2148 2149 2150
		}
		p += maplen;
		nr_maps--;
	}

2151 2152
	if (!osdc->osdmap)
		goto bad;
S
Sage Weil 已提交
2153 2154 2155
done:
	downgrade_write(&osdc->map_sem);
	ceph_monc_got_osdmap(&osdc->client->monc, osdc->osdmap->epoch);
2156 2157 2158 2159 2160 2161

	/*
	 * subscribe to subsequent osdmap updates if full to ensure
	 * we find out when we are no longer full and stop returning
	 * ENOSPC.
	 */
2162 2163 2164
	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))
2165 2166
		ceph_monc_request_next_osdmap(&osdc->client->monc);

A
Alex Elder 已提交
2167 2168 2169
	mutex_lock(&osdc->request_mutex);
	__send_queued(osdc);
	mutex_unlock(&osdc->request_mutex);
S
Sage Weil 已提交
2170
	up_read(&osdc->map_sem);
2171
	wake_up_all(&osdc->client->auth_wq);
S
Sage Weil 已提交
2172 2173 2174 2175
	return;

bad:
	pr_err("osdc handle_map corrupt msg\n");
2176
	ceph_msg_dump(msg);
S
Sage Weil 已提交
2177 2178 2179
	up_write(&osdc->map_sem);
}

2180 2181 2182 2183 2184 2185 2186 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
/*
 * 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 *),
2262
			   void *data, struct ceph_osd_event **pevent)
2263 2264 2265 2266 2267 2268 2269 2270 2271
{
	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;
2272
	event->one_shot = 0;
2273 2274 2275
	event->data = data;
	event->osdc = osdc;
	INIT_LIST_HEAD(&event->osd_node);
2276
	RB_CLEAR_NODE(&event->node);
2277 2278 2279 2280 2281 2282 2283 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
	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
 */
2323 2324
static void handle_watch_notify(struct ceph_osd_client *osdc,
				struct ceph_msg *msg)
2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344
{
	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) {
2345
		BUG_ON(event->one_shot);
2346 2347 2348 2349 2350 2351 2352 2353
		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) {
2354 2355 2356
			pr_err("couldn't allocate event_work\n");
			ceph_osdc_put_event(event);
			return;
2357
		}
2358
		INIT_WORK(&event_work->work, do_event_work);
2359 2360 2361 2362 2363
		event_work->event = event;
		event_work->ver = ver;
		event_work->notify_id = notify_id;
		event_work->opcode = opcode;

2364 2365
		queue_work(osdc->notify_wq, &event_work->work);
	}
2366 2367 2368 2369 2370 2371 2372

	return;

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

2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420
/*
 * 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 */
2421 2422 2423 2424 2425
	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;
2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 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

	/* 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);

2472 2473 2474 2475 2476 2477 2478
/*
 * Register request, send initial attempt.
 */
int ceph_osdc_start_request(struct ceph_osd_client *osdc,
			    struct ceph_osd_request *req,
			    bool nofail)
{
2479
	int rc;
2480

S
Sage Weil 已提交
2481 2482
	down_read(&osdc->map_sem);
	mutex_lock(&osdc->request_mutex);
2483 2484 2485

	rc = __ceph_osdc_start_request(osdc, req, nofail);

S
Sage Weil 已提交
2486 2487
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
2488

S
Sage Weil 已提交
2489 2490
	return rc;
}
2491
EXPORT_SYMBOL(ceph_osdc_start_request);
S
Sage Weil 已提交
2492

2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511
/*
 * 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
2512 2513 2514 2515 2516 2517 2518 2519
/*
 * wait for a request to complete
 */
int ceph_osdc_wait_request(struct ceph_osd_client *osdc,
			   struct ceph_osd_request *req)
{
	int rc;

2520 2521
	dout("%s %p tid %llu\n", __func__, req, req->r_tid);

S
Sage Weil 已提交
2522 2523
	rc = wait_for_completion_interruptible(&req->r_completion);
	if (rc < 0) {
2524 2525
		dout("%s %p tid %llu interrupted\n", __func__, req, req->r_tid);
		ceph_osdc_cancel_request(req);
S
Sage Weil 已提交
2526
		complete_request(req);
S
Sage Weil 已提交
2527 2528 2529
		return rc;
	}

2530 2531
	dout("%s %p tid %llu result %d\n", __func__, req, req->r_tid,
	     req->r_result);
S
Sage Weil 已提交
2532 2533
	return req->r_result;
}
2534
EXPORT_SYMBOL(ceph_osdc_wait_request);
S
Sage Weil 已提交
2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567

/*
 * 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);
}
2568
EXPORT_SYMBOL(ceph_osdc_sync);
S
Sage Weil 已提交
2569

2570 2571 2572 2573
/*
 * 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
2574
void ceph_osdc_flush_notifies(struct ceph_osd_client *osdc)
2575 2576 2577 2578 2579 2580
{
	flush_workqueue(osdc->notify_wq);
}
EXPORT_SYMBOL(ceph_osdc_flush_notifies);


S
Sage Weil 已提交
2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596
/*
 * 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;
2597
	INIT_LIST_HEAD(&osdc->osd_lru);
S
Sage Weil 已提交
2598
	osdc->requests = RB_ROOT;
2599
	INIT_LIST_HEAD(&osdc->req_lru);
2600 2601
	INIT_LIST_HEAD(&osdc->req_unsent);
	INIT_LIST_HEAD(&osdc->req_notarget);
2602
	INIT_LIST_HEAD(&osdc->req_linger);
S
Sage Weil 已提交
2603 2604
	osdc->num_requests = 0;
	INIT_DELAYED_WORK(&osdc->timeout_work, handle_timeout);
2605
	INIT_DELAYED_WORK(&osdc->osds_timeout_work, handle_osds_timeout);
2606 2607 2608
	spin_lock_init(&osdc->event_lock);
	osdc->event_tree = RB_ROOT;
	osdc->event_count = 0;
2609 2610

	schedule_delayed_work(&osdc->osds_timeout_work,
2611
	   round_jiffies_relative(osdc->client->options->osd_idle_ttl * HZ));
S
Sage Weil 已提交
2612

2613
	err = -ENOMEM;
S
Sage Weil 已提交
2614 2615 2616
	osdc->req_mempool = mempool_create_kmalloc_pool(10,
					sizeof(struct ceph_osd_request));
	if (!osdc->req_mempool)
2617
		goto out;
S
Sage Weil 已提交
2618

2619 2620
	err = ceph_msgpool_init(&osdc->msgpool_op, CEPH_MSG_OSD_OP,
				OSD_OP_FRONT_LEN, 10, true,
2621
				"osd_op");
S
Sage Weil 已提交
2622
	if (err < 0)
2623
		goto out_mempool;
2624
	err = ceph_msgpool_init(&osdc->msgpool_op_reply, CEPH_MSG_OSD_OPREPLY,
2625 2626
				OSD_OPREPLY_FRONT_LEN, 10, true,
				"osd_op_reply");
S
Sage Weil 已提交
2627 2628
	if (err < 0)
		goto out_msgpool;
2629

2630
	err = -ENOMEM;
2631
	osdc->notify_wq = create_singlethread_workqueue("ceph-watch-notify");
2632
	if (!osdc->notify_wq)
2633 2634
		goto out_msgpool_reply;

S
Sage Weil 已提交
2635
	return 0;
2636

2637 2638
out_msgpool_reply:
	ceph_msgpool_destroy(&osdc->msgpool_op_reply);
S
Sage Weil 已提交
2639 2640
out_msgpool:
	ceph_msgpool_destroy(&osdc->msgpool_op);
2641 2642 2643 2644
out_mempool:
	mempool_destroy(osdc->req_mempool);
out:
	return err;
S
Sage Weil 已提交
2645 2646 2647 2648
}

void ceph_osdc_stop(struct ceph_osd_client *osdc)
{
2649 2650
	flush_workqueue(osdc->notify_wq);
	destroy_workqueue(osdc->notify_wq);
S
Sage Weil 已提交
2651
	cancel_delayed_work_sync(&osdc->timeout_work);
2652
	cancel_delayed_work_sync(&osdc->osds_timeout_work);
S
Sage Weil 已提交
2653 2654 2655 2656
	if (osdc->osdmap) {
		ceph_osdmap_destroy(osdc->osdmap);
		osdc->osdmap = NULL;
	}
2657
	remove_all_osds(osdc);
S
Sage Weil 已提交
2658 2659
	mempool_destroy(osdc->req_mempool);
	ceph_msgpool_destroy(&osdc->msgpool_op);
S
Sage Weil 已提交
2660
	ceph_msgpool_destroy(&osdc->msgpool_op_reply);
S
Sage Weil 已提交
2661 2662 2663 2664 2665 2666 2667 2668 2669 2670
}

/*
 * 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,
2671
			struct page **pages, int num_pages, int page_align)
S
Sage Weil 已提交
2672 2673 2674 2675 2676 2677
{
	struct ceph_osd_request *req;
	int rc = 0;

	dout("readpages on ino %llx.%llx on %llu~%llu\n", vino.ino,
	     vino.snap, off, *plen);
2678
	req = ceph_osdc_new_request(osdc, layout, vino, off, plen, 1,
S
Sage Weil 已提交
2679
				    CEPH_OSD_OP_READ, CEPH_OSD_FLAG_READ,
2680
				    NULL, truncate_seq, truncate_size,
2681
				    false);
2682 2683
	if (IS_ERR(req))
		return PTR_ERR(req);
S
Sage Weil 已提交
2684 2685

	/* it may be a short read due to an object boundary */
2686

2687
	osd_req_op_extent_osd_data_pages(req, 0,
2688
				pages, *plen, page_align, false, false);
S
Sage Weil 已提交
2689

2690
	dout("readpages  final extent is %llu~%llu (%llu bytes align %d)\n",
2691
	     off, *plen, *plen, page_align);
S
Sage Weil 已提交
2692

2693
	ceph_osdc_build_request(req, off, NULL, vino.snap, NULL);
2694

S
Sage Weil 已提交
2695 2696 2697 2698 2699 2700 2701 2702
	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;
}
2703
EXPORT_SYMBOL(ceph_osdc_readpages);
S
Sage Weil 已提交
2704 2705 2706 2707 2708 2709 2710 2711 2712 2713

/*
 * 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,
2714
			 struct page **pages, int num_pages)
S
Sage Weil 已提交
2715 2716 2717
{
	struct ceph_osd_request *req;
	int rc = 0;
2718
	int page_align = off & ~PAGE_MASK;
S
Sage Weil 已提交
2719

2720
	BUG_ON(vino.snap != CEPH_NOSNAP);	/* snapshots aren't writeable */
2721
	req = ceph_osdc_new_request(osdc, layout, vino, off, &len, 1,
S
Sage Weil 已提交
2722
				    CEPH_OSD_OP_WRITE,
2723
				    CEPH_OSD_FLAG_ONDISK | CEPH_OSD_FLAG_WRITE,
2724
				    snapc, truncate_seq, truncate_size,
2725
				    true);
2726 2727
	if (IS_ERR(req))
		return PTR_ERR(req);
S
Sage Weil 已提交
2728 2729

	/* it may be a short write due to an object boundary */
2730
	osd_req_op_extent_osd_data_pages(req, 0, pages, len, page_align,
2731 2732
				false, false);
	dout("writepages %llu~%llu (%llu bytes)\n", off, len, len);
S
Sage Weil 已提交
2733

2734
	ceph_osdc_build_request(req, off, snapc, CEPH_NOSNAP, mtime);
2735

2736
	rc = ceph_osdc_start_request(osdc, req, true);
S
Sage Weil 已提交
2737 2738 2739 2740 2741 2742 2743 2744 2745
	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;
}
2746
EXPORT_SYMBOL(ceph_osdc_writepages);
S
Sage Weil 已提交
2747

2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767
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
2768 2769 2770 2771 2772 2773
/*
 * handle incoming message
 */
static void dispatch(struct ceph_connection *con, struct ceph_msg *msg)
{
	struct ceph_osd *osd = con->private;
2774
	struct ceph_osd_client *osdc;
S
Sage Weil 已提交
2775 2776 2777
	int type = le16_to_cpu(msg->hdr.type);

	if (!osd)
S
Sage Weil 已提交
2778
		goto out;
2779
	osdc = osd->o_osdc;
S
Sage Weil 已提交
2780 2781 2782 2783 2784 2785

	switch (type) {
	case CEPH_MSG_OSD_MAP:
		ceph_osdc_handle_map(osdc, msg);
		break;
	case CEPH_MSG_OSD_OPREPLY:
2786
		handle_reply(osdc, msg, con);
S
Sage Weil 已提交
2787
		break;
2788 2789 2790
	case CEPH_MSG_WATCH_NOTIFY:
		handle_watch_notify(osdc, msg);
		break;
S
Sage Weil 已提交
2791 2792 2793 2794 2795

	default:
		pr_err("received unknown message type %d %s\n", type,
		       ceph_msg_type_name(type));
	}
S
Sage Weil 已提交
2796
out:
S
Sage Weil 已提交
2797 2798 2799
	ceph_msg_put(msg);
}

2800
/*
2801 2802
 * lookup and return message for incoming reply.  set up reply message
 * pages.
2803 2804
 */
static struct ceph_msg *get_reply(struct ceph_connection *con,
2805 2806
				  struct ceph_msg_header *hdr,
				  int *skip)
S
Sage Weil 已提交
2807 2808 2809
{
	struct ceph_osd *osd = con->private;
	struct ceph_osd_client *osdc = osd->o_osdc;
2810
	struct ceph_msg *m;
2811
	struct ceph_osd_request *req;
2812
	int front_len = le32_to_cpu(hdr->front_len);
2813
	int data_len = le32_to_cpu(hdr->data_len);
2814
	u64 tid;
S
Sage Weil 已提交
2815

2816 2817 2818 2819 2820 2821
	tid = le64_to_cpu(hdr->tid);
	mutex_lock(&osdc->request_mutex);
	req = __lookup_request(osdc, tid);
	if (!req) {
		*skip = 1;
		m = NULL;
2822 2823
		dout("get_reply unknown tid %llu from osd%d\n", tid,
		     osd->o_osd);
2824 2825
		goto out;
	}
S
Sage Weil 已提交
2826

2827
	if (req->r_reply->con)
2828
		dout("%s revoking msg %p from old con %p\n", __func__,
2829 2830
		     req->r_reply, req->r_reply->con);
	ceph_msg_revoke_incoming(req->r_reply);
2831

2832
	if (front_len > req->r_reply->front_alloc_len) {
2833 2834 2835 2836
		pr_warn("get_reply front %d > preallocated %d (%u#%llu)\n",
			front_len, req->r_reply->front_alloc_len,
			(unsigned int)con->peer_name.type,
			le64_to_cpu(con->peer_name.num));
2837 2838
		m = ceph_msg_new(CEPH_MSG_OSD_OPREPLY, front_len, GFP_NOFS,
				 false);
2839
		if (!m)
S
Sage Weil 已提交
2840 2841 2842 2843 2844 2845
			goto out;
		ceph_msg_put(req->r_reply);
		req->r_reply = m;
	}
	m = ceph_msg_get(req->r_reply);

2846
	if (data_len > 0) {
2847
		struct ceph_osd_data *osd_data;
2848

2849 2850 2851 2852 2853
		/*
		 * 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.
		 */
2854
		osd_data = osd_req_op_extent_osd_data(req, 0);
2855 2856
		if (osd_data->type == CEPH_OSD_DATA_TYPE_PAGES) {
			if (osd_data->pages &&
2857
				unlikely(osd_data->length < data_len)) {
2858

2859
				pr_warn("tid %lld reply has %d bytes we had only %llu bytes ready\n",
2860
					tid, data_len, osd_data->length);
2861 2862 2863 2864 2865 2866
				*skip = 1;
				ceph_msg_put(m);
				m = NULL;
				goto out;
			}
		}
2867
	}
2868
	*skip = 0;
S
Sage Weil 已提交
2869
	dout("get_reply tid %lld %p\n", tid, m);
2870 2871 2872

out:
	mutex_unlock(&osdc->request_mutex);
2873
	return m;
2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884

}

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
2885
	*skip = 0;
2886 2887
	switch (type) {
	case CEPH_MSG_OSD_MAP:
2888
	case CEPH_MSG_WATCH_NOTIFY:
2889
		return ceph_msg_new(type, front, GFP_NOFS, false);
2890 2891 2892 2893 2894 2895 2896 2897
	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
2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916
}

/*
 * 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);
}

2917 2918 2919
/*
 * authentication
 */
2920 2921 2922 2923 2924
/*
 * 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,
2925
					int *proto, int force_new)
2926 2927 2928 2929
{
	struct ceph_osd *o = con->private;
	struct ceph_osd_client *osdc = o->o_osdc;
	struct ceph_auth_client *ac = osdc->client->monc.auth;
2930
	struct ceph_auth_handshake *auth = &o->o_auth;
2931

2932
	if (force_new && auth->authorizer) {
2933
		ceph_auth_destroy_authorizer(ac, auth->authorizer);
2934 2935
		auth->authorizer = NULL;
	}
2936 2937 2938
	if (!auth->authorizer) {
		int ret = ceph_auth_create_authorizer(ac, CEPH_ENTITY_TYPE_OSD,
						      auth);
2939
		if (ret)
2940
			return ERR_PTR(ret);
2941 2942
	} else {
		int ret = ceph_auth_update_authorizer(ac, CEPH_ENTITY_TYPE_OSD,
2943 2944 2945
						     auth);
		if (ret)
			return ERR_PTR(ret);
2946 2947
	}
	*proto = ac->protocol;
2948

2949
	return auth;
2950 2951 2952 2953 2954 2955 2956 2957 2958
}


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;

2959
	return ceph_auth_verify_authorizer_reply(ac, o->o_auth.authorizer, len);
2960 2961
}

2962 2963 2964 2965 2966 2967
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;

2968
	ceph_auth_invalidate_authorizer(ac, CEPH_ENTITY_TYPE_OSD);
2969 2970
	return ceph_monc_validate_auth(&osdc->client->monc);
}
2971

Y
Yan, Zheng 已提交
2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985
static int sign_message(struct ceph_connection *con, struct ceph_msg *msg)
{
	struct ceph_osd *o = con->private;
	struct ceph_auth_handshake *auth = &o->o_auth;
	return ceph_auth_sign_message(auth, msg);
}

static int check_message_signature(struct ceph_connection *con, struct ceph_msg *msg)
{
	struct ceph_osd *o = con->private;
	struct ceph_auth_handshake *auth = &o->o_auth;
	return ceph_auth_check_message_signature(auth, msg);
}

2986
static const struct ceph_connection_operations osd_con_ops = {
S
Sage Weil 已提交
2987 2988 2989
	.get = get_osd_con,
	.put = put_osd_con,
	.dispatch = dispatch,
2990 2991
	.get_authorizer = get_authorizer,
	.verify_authorizer_reply = verify_authorizer_reply,
2992
	.invalidate_authorizer = invalidate_authorizer,
S
Sage Weil 已提交
2993
	.alloc_msg = alloc_msg,
Y
Yan, Zheng 已提交
2994 2995
	.sign_message = sign_message,
	.check_message_signature = check_message_signature,
2996
	.fault = osd_reset,
S
Sage Weil 已提交
2997
};