osd_client.c 79.9 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 */

123 124 125 126 127 128 129
#define osd_req_op_data(oreq, whch, typ, fld)				\
({									\
	struct ceph_osd_request *__oreq = (oreq);			\
	unsigned int __whch = (whch);					\
	BUG_ON(__whch >= __oreq->r_num_ops);				\
	&__oreq->r_ops[__whch].typ.fld;					\
})
130

A
Alex Elder 已提交
131 132 133 134 135 136 137 138
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;
}

139 140
struct ceph_osd_data *
osd_req_op_extent_osd_data(struct ceph_osd_request *osd_req,
141
			unsigned int which)
142
{
143
	return osd_req_op_data(osd_req, which, extent, osd_data);
144 145 146 147 148 149 150
}
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)
{
151
	return osd_req_op_data(osd_req, which, cls, response_data);
152 153 154
}
EXPORT_SYMBOL(osd_req_op_cls_response_data);	/* ??? */

A
Alex Elder 已提交
155 156 157 158 159 160 161 162 163 164 165 166 167
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);

168
void osd_req_op_extent_osd_data_pages(struct ceph_osd_request *osd_req,
169 170
			unsigned int which, struct page **pages,
			u64 length, u32 alignment,
171 172 173 174
			bool pages_from_pool, bool own_pages)
{
	struct ceph_osd_data *osd_data;

175
	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
176 177 178 179 180 181
	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,
182
			unsigned int which, struct ceph_pagelist *pagelist)
183 184 185
{
	struct ceph_osd_data *osd_data;

186
	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
187 188 189 190 191 192
	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,
193
			unsigned int which, struct bio *bio, size_t bio_length)
194 195
{
	struct ceph_osd_data *osd_data;
196 197

	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
198 199 200 201 202 203 204 205 206 207 208
	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;

209
	osd_data = osd_req_op_data(osd_req, which, cls, request_info);
210 211 212
	ceph_osd_data_pagelist_init(osd_data, pagelist);
}

213 214 215 216 217 218
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;

219
	osd_data = osd_req_op_data(osd_req, which, cls, request_data);
220 221 222 223
	ceph_osd_data_pagelist_init(osd_data, pagelist);
}
EXPORT_SYMBOL(osd_req_op_cls_request_data_pagelist);

224 225 226 227 228 229 230 231 232 233 234 235
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);

236 237 238 239 240 241
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;

242
	osd_data = osd_req_op_data(osd_req, which, cls, response_data);
243 244 245 246 247
	ceph_osd_data_pages_init(osd_data, pages, length, alignment,
				pages_from_pool, own_pages);
}
EXPORT_SYMBOL(osd_req_op_cls_response_data_pages);

248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266
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
267 268
static void ceph_osd_data_release(struct ceph_osd_data *osd_data)
{
269
	if (osd_data->type == CEPH_OSD_DATA_TYPE_PAGES && osd_data->own_pages) {
A
Alex Elder 已提交
270 271 272 273 274 275
		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);
	}
276 277 278 279 280 281 282 283 284 285 286 287 288 289
	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:
290
	case CEPH_OSD_OP_WRITEFULL:
291 292 293 294
		ceph_osd_data_release(&op->extent.osd_data);
		break;
	case CEPH_OSD_OP_CALL:
		ceph_osd_data_release(&op->cls.request_info);
295
		ceph_osd_data_release(&op->cls.request_data);
296 297
		ceph_osd_data_release(&op->cls.response_data);
		break;
298 299 300 301
	case CEPH_OSD_OP_SETXATTR:
	case CEPH_OSD_OP_CMPXATTR:
		ceph_osd_data_release(&op->xattr.osd_data);
		break;
302 303 304
	case CEPH_OSD_OP_STAT:
		ceph_osd_data_release(&op->raw_data_in);
		break;
305 306 307
	default:
		break;
	}
A
Alex Elder 已提交
308 309
}

S
Sage Weil 已提交
310 311 312
/*
 * requests
 */
313
static void ceph_osdc_release_request(struct kref *kref)
S
Sage Weil 已提交
314
{
315 316
	struct ceph_osd_request *req = container_of(kref,
					    struct ceph_osd_request, r_kref);
317
	unsigned int which;
S
Sage Weil 已提交
318

319 320
	dout("%s %p (r_request %p r_reply %p)\n", __func__, req,
	     req->r_request, req->r_reply);
321 322 323 324 325 326
	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);
327

S
Sage Weil 已提交
328 329
	if (req->r_request)
		ceph_msg_put(req->r_request);
330
	if (req->r_reply) {
331
		ceph_msg_revoke_incoming(req->r_reply);
332
		ceph_msg_put(req->r_reply);
333
	}
334

335 336
	for (which = 0; which < req->r_num_ops; which++)
		osd_req_op_data_release(req, which);
337

S
Sage Weil 已提交
338 339 340
	ceph_put_snap_context(req->r_snapc);
	if (req->r_mempool)
		mempool_free(req, req->r_osdc->req_mempool);
341
	else if (req->r_num_ops <= CEPH_OSD_SLAB_OPS)
342
		kmem_cache_free(ceph_osd_request_cache, req);
343 344
	else
		kfree(req);
S
Sage Weil 已提交
345
}
346 347 348 349 350 351 352 353 354 355 356

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)
{
357 358 359 360 361
	if (req) {
		dout("%s %p (was %d)\n", __func__, req,
		     atomic_read(&req->r_kref.refcount));
		kref_put(&req->r_kref, ceph_osdc_release_request);
	}
362 363
}
EXPORT_SYMBOL(ceph_osdc_put_request);
364

365
struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc,
S
Sage Weil 已提交
366
					       struct ceph_snap_context *snapc,
367
					       unsigned int num_ops,
368
					       bool use_mempool,
369
					       gfp_t gfp_flags)
S
Sage Weil 已提交
370 371 372
{
	struct ceph_osd_request *req;
	struct ceph_msg *msg;
373 374
	size_t msg_size;

S
Sage Weil 已提交
375
	if (use_mempool) {
376
		BUG_ON(num_ops > CEPH_OSD_SLAB_OPS);
377
		req = mempool_alloc(osdc->req_mempool, gfp_flags);
378 379
	} else if (num_ops <= CEPH_OSD_SLAB_OPS) {
		req = kmem_cache_alloc(ceph_osd_request_cache, gfp_flags);
S
Sage Weil 已提交
380
	} else {
381 382 383
		BUG_ON(num_ops > CEPH_OSD_MAX_OPS);
		req = kmalloc(sizeof(*req) + num_ops * sizeof(req->r_ops[0]),
			      gfp_flags);
S
Sage Weil 已提交
384
	}
385
	if (unlikely(!req))
386
		return NULL;
S
Sage Weil 已提交
387

388 389 390
	/* req only, each op is zeroed in _osd_req_op_init() */
	memset(req, 0, sizeof(*req));

S
Sage Weil 已提交
391 392
	req->r_osdc = osdc;
	req->r_mempool = use_mempool;
393
	req->r_num_ops = num_ops;
394 395
	req->r_snapid = CEPH_NOSNAP;
	req->r_snapc = ceph_get_snap_context(snapc);
396

S
Sage Weil 已提交
397
	kref_init(&req->r_kref);
S
Sage Weil 已提交
398 399
	init_completion(&req->r_completion);
	init_completion(&req->r_safe_completion);
400
	RB_CLEAR_NODE(&req->r_node);
S
Sage Weil 已提交
401
	INIT_LIST_HEAD(&req->r_unsafe_item);
402
	INIT_LIST_HEAD(&req->r_linger_item);
403
	INIT_LIST_HEAD(&req->r_linger_osd_item);
404
	INIT_LIST_HEAD(&req->r_req_lru_item);
405 406
	INIT_LIST_HEAD(&req->r_osd_item);

407
	req->r_base_oloc.pool = -1;
408
	req->r_target_oloc.pool = -1;
409

410 411 412 413 414 415 416
	msg_size = OSD_OPREPLY_FRONT_LEN;
	if (num_ops > CEPH_OSD_SLAB_OPS) {
		/* ceph_osd_op and rval */
		msg_size += (num_ops - CEPH_OSD_SLAB_OPS) *
			    (sizeof(struct ceph_osd_op) + 4);
	}

S
Sage Weil 已提交
417 418 419 420
	/* create reply message */
	if (use_mempool)
		msg = ceph_msgpool_get(&osdc->msgpool_op_reply, 0);
	else
421 422
		msg = ceph_msg_new(CEPH_MSG_OSD_OPREPLY, msg_size,
				   gfp_flags, true);
423
	if (!msg) {
S
Sage Weil 已提交
424
		ceph_osdc_put_request(req);
425
		return NULL;
S
Sage Weil 已提交
426 427 428
	}
	req->r_reply = msg;

429 430 431 432 433 434 435 436 437 438 439
	msg_size = 4 + 4 + 4; /* client_inc, osdmap_epoch, flags */
	msg_size += 4 + 4 + 4 + 8; /* mtime, reassert_version */
	msg_size += 2 + 4 + 8 + 4 + 4; /* oloc */
	msg_size += 1 + 8 + 4 + 4; /* pgid */
	msg_size += 4 + CEPH_MAX_OID_NAME_LEN; /* oid */
	msg_size += 2 + num_ops * sizeof(struct ceph_osd_op);
	msg_size += 8; /* snapid */
	msg_size += 8; /* snap_seq */
	msg_size += 4 + 8 * (snapc ? snapc->num_snaps : 0); /* snaps */
	msg_size += 4; /* retry_attempt */

S
Sage Weil 已提交
440
	/* create request message; allow space for oid */
S
Sage Weil 已提交
441
	if (use_mempool)
442
		msg = ceph_msgpool_get(&osdc->msgpool_op, 0);
S
Sage Weil 已提交
443
	else
444
		msg = ceph_msg_new(CEPH_MSG_OSD_OP, msg_size, gfp_flags, true);
445
	if (!msg) {
S
Sage Weil 已提交
446
		ceph_osdc_put_request(req);
447
		return NULL;
S
Sage Weil 已提交
448
	}
449

S
Sage Weil 已提交
450
	memset(msg->front.iov_base, 0, msg->front.iov_len);
451 452 453 454 455

	req->r_request = msg;

	return req;
}
456
EXPORT_SYMBOL(ceph_osdc_alloc_request);
457

458
static bool osd_req_opcode_valid(u16 opcode)
459
{
460
	switch (opcode) {
461 462 463
#define GENERATE_CASE(op, opcode, str)	case CEPH_OSD_OP_##op: return true;
__CEPH_FORALL_OSD_OPS(GENERATE_CASE)
#undef GENERATE_CASE
464 465 466 467 468
	default:
		return false;
	}
}

469 470 471 472 473
/*
 * 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.
 */
474
static struct ceph_osd_req_op *
A
Alex Elder 已提交
475
_osd_req_op_init(struct ceph_osd_request *osd_req, unsigned int which,
476
		 u16 opcode, u32 flags)
477
{
478 479 480
	struct ceph_osd_req_op *op;

	BUG_ON(which >= osd_req->r_num_ops);
481 482
	BUG_ON(!osd_req_opcode_valid(opcode));

483
	op = &osd_req->r_ops[which];
484 485
	memset(op, 0, sizeof (*op));
	op->op = opcode;
486
	op->flags = flags;
487 488

	return op;
489 490
}

A
Alex Elder 已提交
491
void osd_req_op_init(struct ceph_osd_request *osd_req,
492
		     unsigned int which, u16 opcode, u32 flags)
A
Alex Elder 已提交
493
{
494
	(void)_osd_req_op_init(osd_req, which, opcode, flags);
A
Alex Elder 已提交
495 496 497
}
EXPORT_SYMBOL(osd_req_op_init);

498 499
void osd_req_op_extent_init(struct ceph_osd_request *osd_req,
				unsigned int which, u16 opcode,
500 501 502
				u64 offset, u64 length,
				u64 truncate_size, u32 truncate_seq)
{
503 504
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
						      opcode, 0);
505 506
	size_t payload_len = 0;

L
Li Wang 已提交
507
	BUG_ON(opcode != CEPH_OSD_OP_READ && opcode != CEPH_OSD_OP_WRITE &&
508 509
	       opcode != CEPH_OSD_OP_WRITEFULL && opcode != CEPH_OSD_OP_ZERO &&
	       opcode != CEPH_OSD_OP_TRUNCATE);
510 511 512 513 514

	op->extent.offset = offset;
	op->extent.length = length;
	op->extent.truncate_size = truncate_size;
	op->extent.truncate_seq = truncate_seq;
515
	if (opcode == CEPH_OSD_OP_WRITE || opcode == CEPH_OSD_OP_WRITEFULL)
516 517
		payload_len += length;

518
	op->indata_len = payload_len;
519 520 521
}
EXPORT_SYMBOL(osd_req_op_extent_init);

522 523
void osd_req_op_extent_update(struct ceph_osd_request *osd_req,
				unsigned int which, u64 length)
524
{
525 526 527 528 529 530
	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;
531 532 533 534 535 536

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

	op->extent.length = length;
537
	op->indata_len -= previous - length;
538 539 540
}
EXPORT_SYMBOL(osd_req_op_extent_update);

541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562
void osd_req_op_extent_dup_last(struct ceph_osd_request *osd_req,
				unsigned int which, u64 offset_inc)
{
	struct ceph_osd_req_op *op, *prev_op;

	BUG_ON(which + 1 >= osd_req->r_num_ops);

	prev_op = &osd_req->r_ops[which];
	op = _osd_req_op_init(osd_req, which + 1, prev_op->op, prev_op->flags);
	/* dup previous one */
	op->indata_len = prev_op->indata_len;
	op->outdata_len = prev_op->outdata_len;
	op->extent = prev_op->extent;
	/* adjust offset */
	op->extent.offset += offset_inc;
	op->extent.length -= offset_inc;

	if (op->op == CEPH_OSD_OP_WRITE || op->op == CEPH_OSD_OP_WRITEFULL)
		op->indata_len -= offset_inc;
}
EXPORT_SYMBOL(osd_req_op_extent_dup_last);

563
void osd_req_op_cls_init(struct ceph_osd_request *osd_req, unsigned int which,
564
			u16 opcode, const char *class, const char *method)
565
{
566 567
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
						      opcode, 0);
568
	struct ceph_pagelist *pagelist;
569 570 571 572 573
	size_t payload_len = 0;
	size_t size;

	BUG_ON(opcode != CEPH_OSD_OP_CALL);

574 575 576 577
	pagelist = kmalloc(sizeof (*pagelist), GFP_NOFS);
	BUG_ON(!pagelist);
	ceph_pagelist_init(pagelist);

578 579 580 581
	op->cls.class_name = class;
	size = strlen(class);
	BUG_ON(size > (size_t) U8_MAX);
	op->cls.class_len = size;
582
	ceph_pagelist_append(pagelist, class, size);
583 584 585 586 587 588
	payload_len += size;

	op->cls.method_name = method;
	size = strlen(method);
	BUG_ON(size > (size_t) U8_MAX);
	op->cls.method_len = size;
589
	ceph_pagelist_append(pagelist, method, size);
590 591
	payload_len += size;

592
	osd_req_op_cls_request_info_pagelist(osd_req, which, pagelist);
593

594 595
	op->cls.argc = 0;	/* currently unused */

596
	op->indata_len = payload_len;
597 598
}
EXPORT_SYMBOL(osd_req_op_cls_init);
599

600 601 602 603
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)
{
604 605
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
						      opcode, 0);
606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628
	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);
629
	op->indata_len = payload_len;
630 631 632 633
	return 0;
}
EXPORT_SYMBOL(osd_req_op_xattr_init);

634 635
void osd_req_op_watch_init(struct ceph_osd_request *osd_req,
				unsigned int which, u16 opcode,
636 637
				u64 cookie, u64 version, int flag)
{
638 639
	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
						      opcode, 0);
640

641
	BUG_ON(opcode != CEPH_OSD_OP_NOTIFY_ACK && opcode != CEPH_OSD_OP_WATCH);
642 643

	op->watch.cookie = cookie;
A
Alex Elder 已提交
644
	op->watch.ver = version;
645
	if (opcode == CEPH_OSD_OP_WATCH && flag)
646
		op->watch.flag = (u8)1;
647 648 649
}
EXPORT_SYMBOL(osd_req_op_watch_init);

650 651 652 653 654 655
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,
656 657
						      CEPH_OSD_OP_SETALLOCHINT,
						      0);
658 659 660 661 662 663 664 665 666 667 668 669 670

	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);

671
static void ceph_osdc_msg_data_add(struct ceph_msg *msg,
672 673 674 675 676 677 678
				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)
679
			ceph_msg_data_add_pages(msg, osd_data->pages,
680 681 682
					length, osd_data->alignment);
	} else if (osd_data->type == CEPH_OSD_DATA_TYPE_PAGELIST) {
		BUG_ON(!length);
683
		ceph_msg_data_add_pagelist(msg, osd_data->pagelist);
684 685
#ifdef CONFIG_BLOCK
	} else if (osd_data->type == CEPH_OSD_DATA_TYPE_BIO) {
686
		ceph_msg_data_add_bio(msg, osd_data->bio, length);
687 688 689 690 691 692
#endif
	} else {
		BUG_ON(osd_data->type != CEPH_OSD_DATA_TYPE_NONE);
	}
}

693
static u64 osd_req_encode_op(struct ceph_osd_request *req,
694
			      struct ceph_osd_op *dst, unsigned int which)
695
{
696
	struct ceph_osd_req_op *src;
697
	struct ceph_osd_data *osd_data;
698
	u64 request_data_len = 0;
699
	u64 data_length;
700

701 702
	BUG_ON(which >= req->r_num_ops);
	src = &req->r_ops[which];
703 704 705 706 707 708 709 710
	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
711 712
		osd_data = &src->raw_data_in;
		ceph_osdc_msg_data_add(req->r_reply, osd_data);
713 714 715
		break;
	case CEPH_OSD_OP_READ:
	case CEPH_OSD_OP_WRITE:
716
	case CEPH_OSD_OP_WRITEFULL:
L
Li Wang 已提交
717 718
	case CEPH_OSD_OP_ZERO:
	case CEPH_OSD_OP_TRUNCATE:
719 720
		if (src->op == CEPH_OSD_OP_WRITE ||
		    src->op == CEPH_OSD_OP_WRITEFULL)
721
			request_data_len = src->extent.length;
722 723 724 725 726 727
		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);
728
		osd_data = &src->extent.osd_data;
729 730
		if (src->op == CEPH_OSD_OP_WRITE ||
		    src->op == CEPH_OSD_OP_WRITEFULL)
731
			ceph_osdc_msg_data_add(req->r_request, osd_data);
732
		else
733
			ceph_osdc_msg_data_add(req->r_reply, osd_data);
734 735 736 737
		break;
	case CEPH_OSD_OP_CALL:
		dst->cls.class_len = src->cls.class_len;
		dst->cls.method_len = src->cls.method_len;
738 739 740 741 742 743 744 745 746 747 748
		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);
749
			src->indata_len += data_length;
750 751 752 753
			request_data_len += data_length;
		}
		osd_data = &src->cls.response_data;
		ceph_osdc_msg_data_add(req->r_reply, osd_data);
754 755 756 757 758 759 760 761 762
		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;
763 764 765 766 767 768
	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;
769 770 771 772 773 774 775 776 777 778
	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;
779 780 781
	case CEPH_OSD_OP_CREATE:
	case CEPH_OSD_OP_DELETE:
		break;
782
	default:
783
		pr_err("unsupported osd opcode %s\n",
784
			ceph_osd_op_name(src->op));
785
		WARN_ON(1);
786 787

		return 0;
788
	}
789

790
	dst->op = cpu_to_le16(src->op);
791
	dst->flags = cpu_to_le32(src->flags);
792
	dst->payload_len = cpu_to_le32(src->indata_len);
793

794
	return request_data_len;
795 796
}

797 798 799 800 801 802 803 804 805 806 807 808 809 810
/*
 * 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,
811 812
					       u64 off, u64 *plen,
					       unsigned int which, int num_ops,
813 814 815 816
					       int opcode, int flags,
					       struct ceph_snap_context *snapc,
					       u32 truncate_seq,
					       u64 truncate_size,
817
					       bool use_mempool)
818
{
819
	struct ceph_osd_request *req;
820 821 822
	u64 objnum = 0;
	u64 objoff = 0;
	u64 objlen = 0;
823
	int r;
824

L
Li Wang 已提交
825
	BUG_ON(opcode != CEPH_OSD_OP_READ && opcode != CEPH_OSD_OP_WRITE &&
826 827
	       opcode != CEPH_OSD_OP_ZERO && opcode != CEPH_OSD_OP_TRUNCATE &&
	       opcode != CEPH_OSD_OP_CREATE && opcode != CEPH_OSD_OP_DELETE);
828

829
	req = ceph_osdc_alloc_request(osdc, snapc, num_ops, use_mempool,
A
Alex Elder 已提交
830
					GFP_NOFS);
831
	if (!req)
832
		return ERR_PTR(-ENOMEM);
833

834
	req->r_flags = flags;
835 836

	/* calculate max write size */
837
	r = calc_layout(layout, off, plen, &objnum, &objoff, &objlen);
838 839
	if (r < 0) {
		ceph_osdc_put_request(req);
840
		return ERR_PTR(r);
841
	}
842

843
	if (opcode == CEPH_OSD_OP_CREATE || opcode == CEPH_OSD_OP_DELETE) {
844
		osd_req_op_init(req, which, opcode, 0);
845 846 847 848 849 850 851 852 853 854 855
	} 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;
			}
856
		}
857
		osd_req_op_extent_init(req, which, opcode, objoff, objlen,
858 859
				       truncate_size, truncate_seq);
	}
860

861
	req->r_base_oloc.pool = ceph_file_layout_pg_pool(*layout);
862

863
	snprintf(req->r_base_oid.name, sizeof(req->r_base_oid.name),
864
		 "%llx.%08llx", vino.ino, objnum);
865
	req->r_base_oid.name_len = strlen(req->r_base_oid.name);
866

S
Sage Weil 已提交
867 868
	return req;
}
869
EXPORT_SYMBOL(ceph_osdc_new_request);
S
Sage Weil 已提交
870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935

/*
 * 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;
}

936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966
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);
}

967 968 969 970 971 972
/*
 * Resubmit requests pending on the given osd.
 */
static void __kick_osd_requests(struct ceph_osd_client *osdc,
				struct ceph_osd *osd)
{
973
	struct ceph_osd_request *req, *nreq;
974
	LIST_HEAD(resend);
975
	LIST_HEAD(resend_linger);
976 977
	int err;

978
	dout("%s osd%d\n", __func__, osd->o_osd);
979
	err = __reset_osd(osdc, osd);
980
	if (err)
981
		return;
982

983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999
	/*
	 * 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.
	 */
1000
	list_for_each_entry(req, &osd->o_requests, r_osd_item) {
1001 1002
		if (!req->r_sent)
			break;
1003 1004 1005 1006 1007

		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);
1008
			req->r_flags |= CEPH_OSD_FLAG_RETRY;
1009 1010 1011
		} else {
			list_move_tail(&req->r_req_lru_item, &resend_linger);
		}
1012
	}
1013
	list_splice(&resend, &osdc->req_unsent);
1014

1015
	/*
1016 1017 1018 1019
	 * 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.
1020
	 */
1021
	list_for_each_entry_safe(req, nreq, &osd->o_linger_requests,
1022
				 r_linger_osd_item) {
1023 1024
		WARN_ON(!list_empty(&req->r_req_lru_item));
		__kick_linger_request(req);
1025
	}
1026 1027 1028

	list_for_each_entry_safe(req, nreq, &resend_linger, r_req_lru_item)
		__kick_linger_request(req);
1029 1030
}

S
Sage Weil 已提交
1031
/*
1032
 * If the osd connection drops, we need to resubmit all requests.
S
Sage Weil 已提交
1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043
 */
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
1044 1045
	mutex_lock(&osdc->request_mutex);
	__kick_osd_requests(osdc, osd);
A
Alex Elder 已提交
1046
	__send_queued(osdc);
S
Sage Weil 已提交
1047
	mutex_unlock(&osdc->request_mutex);
S
Sage Weil 已提交
1048 1049 1050 1051 1052 1053
	up_read(&osdc->map_sem);
}

/*
 * Track open sessions with osds.
 */
1054
static struct ceph_osd *create_osd(struct ceph_osd_client *osdc, int onum)
S
Sage Weil 已提交
1055 1056 1057 1058 1059 1060 1061 1062 1063
{
	struct ceph_osd *osd;

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

	atomic_set(&osd->o_ref, 1);
	osd->o_osdc = osdc;
1064
	osd->o_osd = onum;
1065
	RB_CLEAR_NODE(&osd->o_node);
S
Sage Weil 已提交
1066
	INIT_LIST_HEAD(&osd->o_requests);
1067
	INIT_LIST_HEAD(&osd->o_linger_requests);
1068
	INIT_LIST_HEAD(&osd->o_osd_lru);
S
Sage Weil 已提交
1069 1070
	osd->o_incarnation = 1;

1071
	ceph_con_init(&osd->o_con, osd, &osd_con_ops, &osdc->client->msgr);
1072

1073
	INIT_LIST_HEAD(&osd->o_keepalive_item);
S
Sage Weil 已提交
1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092
	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);
1093 1094
	if (atomic_dec_and_test(&osd->o_ref)) {
		if (osd->o_auth.authorizer)
1095
			ceph_auth_destroy_authorizer(osd->o_auth.authorizer);
S
Sage Weil 已提交
1096
		kfree(osd);
S
Sage Weil 已提交
1097
	}
S
Sage Weil 已提交
1098 1099 1100 1101 1102
}

/*
 * remove an osd from our map
 */
1103
static void __remove_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
S
Sage Weil 已提交
1104
{
1105
	dout("%s %p osd%d\n", __func__, osd, osd->o_osd);
1106 1107
	WARN_ON(!list_empty(&osd->o_requests));
	WARN_ON(!list_empty(&osd->o_linger_requests));
1108

1109
	list_del_init(&osd->o_osd_lru);
1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122
	rb_erase(&osd->o_node, &osdc->osds);
	RB_CLEAR_NODE(&osd->o_node);
}

static void remove_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
{
	dout("%s %p osd%d\n", __func__, osd, osd->o_osd);

	if (!RB_EMPTY_NODE(&osd->o_node)) {
		ceph_con_close(&osd->o_con);
		__remove_osd(osdc, osd);
		put_osd(osd);
	}
S
Sage Weil 已提交
1123 1124
}

1125 1126
static void remove_all_osds(struct ceph_osd_client *osdc)
{
1127
	dout("%s %p\n", __func__, osdc);
1128 1129 1130 1131
	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);
1132
		remove_osd(osdc, osd);
1133 1134 1135 1136
	}
	mutex_unlock(&osdc->request_mutex);
}

1137 1138 1139
static void __move_osd_to_lru(struct ceph_osd_client *osdc,
			      struct ceph_osd *osd)
{
1140
	dout("%s %p\n", __func__, osd);
1141
	BUG_ON(!list_empty(&osd->o_osd_lru));
1142

1143
	list_add_tail(&osd->o_osd_lru, &osdc->osd_lru);
1144
	osd->lru_ttl = jiffies + osdc->client->options->osd_idle_ttl;
1145 1146
}

1147 1148 1149 1150 1151 1152 1153 1154 1155 1156
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);
}

1157 1158 1159 1160 1161 1162 1163
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);
}

1164
static void remove_old_osds(struct ceph_osd_client *osdc)
1165 1166 1167 1168 1169 1170
{
	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) {
1171
		if (time_before(jiffies, osd->lru_ttl))
1172
			break;
1173
		remove_osd(osdc, osd);
1174 1175 1176 1177
	}
	mutex_unlock(&osdc->request_mutex);
}

S
Sage Weil 已提交
1178 1179 1180
/*
 * reset osd connect
 */
1181
static int __reset_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
S
Sage Weil 已提交
1182
{
A
Alex Elder 已提交
1183
	struct ceph_entity_addr *peer_addr;
S
Sage Weil 已提交
1184

1185
	dout("__reset_osd %p osd%d\n", osd, osd->o_osd);
1186 1187
	if (list_empty(&osd->o_requests) &&
	    list_empty(&osd->o_linger_requests)) {
1188
		remove_osd(osdc, osd);
A
Alex Elder 已提交
1189 1190 1191 1192 1193 1194 1195 1196
		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;

1197 1198
		dout("osd addr hasn't changed and connection never opened, "
		     "letting msgr retry\n");
1199 1200 1201
		/* 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
1202 1203

		return -EAGAIN;
S
Sage Weil 已提交
1204
	}
A
Alex Elder 已提交
1205 1206 1207 1208 1209 1210

	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
1211 1212 1213 1214 1215 1216 1217 1218
}

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;

1219
	dout("__insert_osd %p osd%d\n", new, new->o_osd);
S
Sage Weil 已提交
1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251
	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;
}

1252 1253 1254
static void __schedule_osd_timeout(struct ceph_osd_client *osdc)
{
	schedule_delayed_work(&osdc->timeout_work,
1255
			      osdc->client->options->osd_keepalive_timeout);
1256 1257 1258 1259 1260 1261
}

static void __cancel_osd_timeout(struct ceph_osd_client *osdc)
{
	cancel_delayed_work(&osdc->timeout_work);
}
S
Sage Weil 已提交
1262 1263 1264 1265 1266

/*
 * Register request, assign tid.  If this is the first request, set up
 * the timeout event.
 */
1267 1268
static void __register_request(struct ceph_osd_client *osdc,
			       struct ceph_osd_request *req)
S
Sage Weil 已提交
1269 1270
{
	req->r_tid = ++osdc->last_tid;
1271
	req->r_request->hdr.tid = cpu_to_le64(req->r_tid);
1272
	dout("__register_request %p tid %lld\n", req, req->r_tid);
S
Sage Weil 已提交
1273 1274 1275 1276
	__insert_request(osdc, req);
	ceph_osdc_get_request(req);
	osdc->num_requests++;
	if (osdc->num_requests == 1) {
1277 1278
		dout(" first request, scheduling timeout\n");
		__schedule_osd_timeout(osdc);
S
Sage Weil 已提交
1279
	}
1280 1281
}

S
Sage Weil 已提交
1282 1283 1284 1285 1286 1287
/*
 * called under osdc->request_mutex
 */
static void __unregister_request(struct ceph_osd_client *osdc,
				 struct ceph_osd_request *req)
{
1288 1289 1290 1291 1292 1293
	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
1294 1295
	dout("__unregister_request %p tid %lld\n", req, req->r_tid);
	rb_erase(&req->r_node, &osdc->requests);
1296
	RB_CLEAR_NODE(&req->r_node);
S
Sage Weil 已提交
1297 1298
	osdc->num_requests--;

1299 1300
	if (req->r_osd) {
		/* make sure the original request isn't in flight. */
1301
		ceph_msg_revoke(req->r_request);
1302 1303

		list_del_init(&req->r_osd_item);
1304
		maybe_move_osd_to_lru(osdc, req->r_osd);
1305
		if (list_empty(&req->r_linger_osd_item))
1306
			req->r_osd = NULL;
1307
	}
S
Sage Weil 已提交
1308

A
Alex Elder 已提交
1309
	list_del_init(&req->r_req_lru_item);
S
Sage Weil 已提交
1310 1311
	ceph_osdc_put_request(req);

1312 1313 1314
	if (osdc->num_requests == 0) {
		dout(" no requests, canceling timeout\n");
		__cancel_osd_timeout(osdc);
S
Sage Weil 已提交
1315 1316 1317 1318 1319 1320 1321 1322
	}
}

/*
 * Cancel a previously queued request message
 */
static void __cancel_request(struct ceph_osd_request *req)
{
1323
	if (req->r_sent && req->r_osd) {
1324
		ceph_msg_revoke(req->r_request);
S
Sage Weil 已提交
1325 1326 1327 1328
		req->r_sent = 0;
	}
}

1329 1330 1331
static void __register_linger_request(struct ceph_osd_client *osdc,
				    struct ceph_osd_request *req)
{
1332 1333 1334
	dout("%s %p tid %llu\n", __func__, req, req->r_tid);
	WARN_ON(!req->r_linger);

1335
	ceph_osdc_get_request(req);
1336
	list_add_tail(&req->r_linger_item, &osdc->req_linger);
1337
	if (req->r_osd)
1338
		list_add_tail(&req->r_linger_osd_item,
1339
			      &req->r_osd->o_linger_requests);
1340 1341 1342 1343 1344
}

static void __unregister_linger_request(struct ceph_osd_client *osdc,
					struct ceph_osd_request *req)
{
1345 1346 1347 1348 1349 1350 1351 1352 1353
	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
1354
	list_del_init(&req->r_linger_item);
1355

1356
	if (req->r_osd) {
1357
		list_del_init(&req->r_linger_osd_item);
1358
		maybe_move_osd_to_lru(osdc, req->r_osd);
1359 1360
		if (list_empty(&req->r_osd_item))
			req->r_osd = NULL;
1361
	}
1362
	ceph_osdc_put_request(req);
1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374
}

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);

1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390
/*
 * 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);
}

1391 1392 1393 1394 1395 1396 1397
/*
 * 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)
{
1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411
	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) {
1412 1413
		struct ceph_pg_pool_info *pi;

1414
		pi = ceph_pg_pool_by_id(osdmap, req->r_target_oloc.pool);
1415 1416 1417
		if (pi) {
			if ((req->r_flags & CEPH_OSD_FLAG_READ) &&
			    pi->read_tier >= 0)
1418
				req->r_target_oloc.pool = pi->read_tier;
1419 1420
			if ((req->r_flags & CEPH_OSD_FLAG_WRITE) &&
			    pi->write_tier >= 0)
1421
				req->r_target_oloc.pool = pi->write_tier;
1422 1423 1424 1425
		}
		/* !pi is caught in ceph_oloc_oid_to_pg() */
	}

1426 1427
	return ceph_oloc_oid_to_pg(osdmap, &req->r_target_oloc,
				   &req->r_target_oid, pg_out);
1428 1429
}

1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445
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
1446 1447 1448
/*
 * 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
1449
 * no up osd, set r_osd to NULL.  Move the request to the appropriate list
1450
 * (unsent, homeless) or leave on in-flight lru.
S
Sage Weil 已提交
1451 1452 1453 1454 1455
 *
 * Return 0 if unchanged, 1 if changed, or negative on error.
 *
 * Caller should hold map_sem for read and request_mutex.
 */
1456
static int __map_request(struct ceph_osd_client *osdc,
1457
			 struct ceph_osd_request *req, int force_resend)
S
Sage Weil 已提交
1458
{
1459
	struct ceph_pg pgid;
1460
	int acting[CEPH_PG_MAX_SIZE];
1461
	int num, o;
S
Sage Weil 已提交
1462
	int err;
1463
	bool was_paused;
S
Sage Weil 已提交
1464

1465
	dout("map_request %p tid %lld\n", req, req->r_tid);
1466 1467

	err = __calc_request_pg(osdc->osdmap, req, &pgid);
1468 1469
	if (err) {
		list_move(&req->r_req_lru_item, &osdc->req_notarget);
S
Sage Weil 已提交
1470
		return err;
1471
	}
1472 1473
	req->r_pgid = pgid;

1474 1475 1476
	num = ceph_calc_pg_acting(osdc->osdmap, pgid, acting, &o);
	if (num < 0)
		num = 0;
S
Sage Weil 已提交
1477

1478 1479 1480 1481 1482
	was_paused = req->r_paused;
	req->r_paused = __req_should_be_paused(osdc, req);
	if (was_paused && !req->r_paused)
		force_resend = 1;

1483 1484
	if ((!force_resend &&
	     req->r_osd && req->r_osd->o_osd == o &&
1485 1486 1487
	     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) ||
1488 1489
	    (req->r_osd == NULL && o == -1) ||
	    req->r_paused)
S
Sage Weil 已提交
1490 1491
		return 0;  /* no change */

1492 1493
	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
1494 1495
	     req->r_osd ? req->r_osd->o_osd : -1);

1496 1497 1498 1499
	/* 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
1500 1501 1502
	if (req->r_osd) {
		__cancel_request(req);
		list_del_init(&req->r_osd_item);
1503
		list_del_init(&req->r_linger_osd_item);
S
Sage Weil 已提交
1504 1505 1506 1507 1508
		req->r_osd = NULL;
	}

	req->r_osd = __lookup_osd(osdc, o);
	if (!req->r_osd && o >= 0) {
1509
		err = -ENOMEM;
1510
		req->r_osd = create_osd(osdc, o);
1511 1512
		if (!req->r_osd) {
			list_move(&req->r_req_lru_item, &osdc->req_notarget);
1513
			goto out;
1514
		}
S
Sage Weil 已提交
1515

1516
		dout("map_request osd %p is osd%d\n", req->r_osd, o);
S
Sage Weil 已提交
1517 1518
		__insert_osd(osdc, req->r_osd);

1519 1520 1521
		ceph_con_open(&req->r_osd->o_con,
			      CEPH_ENTITY_TYPE_OSD, o,
			      &osdc->osdmap->osd_addr[o]);
S
Sage Weil 已提交
1522 1523
	}

1524
	__enqueue_request(req);
1525
	err = 1;   /* osd or pg changed */
S
Sage Weil 已提交
1526 1527 1528 1529 1530 1531 1532 1533

out:
	return err;
}

/*
 * caller should hold map_sem (for read) and request_mutex
 */
1534 1535
static void __send_request(struct ceph_osd_client *osdc,
			   struct ceph_osd_request *req)
S
Sage Weil 已提交
1536
{
1537
	void *p;
S
Sage Weil 已提交
1538

1539 1540 1541 1542 1543 1544 1545
	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);
1546
	put_unaligned_le64(req->r_target_oloc.pool, req->r_request_pool);
1547 1548 1549 1550 1551 1552
	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));
1553

S
Sage Weil 已提交
1554
	req->r_stamp = jiffies;
1555
	list_move_tail(&req->r_req_lru_item, &osdc->req_lru);
S
Sage Weil 已提交
1556 1557

	ceph_msg_get(req->r_request); /* send consumes a ref */
1558

S
Sage Weil 已提交
1559
	req->r_sent = req->r_osd->o_incarnation;
1560 1561

	ceph_con_send(&req->r_osd->o_con, req->r_request);
S
Sage Weil 已提交
1562 1563
}

1564 1565 1566
/*
 * Send any requests in the queue (req_unsent).
 */
A
Alex Elder 已提交
1567
static void __send_queued(struct ceph_osd_client *osdc)
1568 1569 1570
{
	struct ceph_osd_request *req, *tmp;

A
Alex Elder 已提交
1571 1572
	dout("__send_queued\n");
	list_for_each_entry_safe(req, tmp, &osdc->req_unsent, r_req_lru_item)
1573 1574 1575
		__send_request(osdc, req);
}

1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609
/*
 * 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
1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622
/*
 * 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);
1623
	struct ceph_options *opts = osdc->client->options;
S
Sage Weil 已提交
1624
	struct ceph_osd_request *req;
S
Sage Weil 已提交
1625
	struct ceph_osd *osd;
1626
	struct list_head slow_osds;
S
Sage Weil 已提交
1627 1628 1629 1630 1631 1632 1633
	dout("timeout\n");
	down_read(&osdc->map_sem);

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

	mutex_lock(&osdc->request_mutex);

1634 1635 1636 1637 1638 1639 1640
	/*
	 * 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) {
1641 1642
		if (time_before(jiffies,
				req->r_stamp + opts->osd_keepalive_timeout))
1643 1644 1645 1646 1647
			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
1648
		     req->r_tid, osd->o_osd);
1649 1650 1651 1652 1653 1654
		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
1655 1656 1657
		ceph_con_keepalive(&osd->o_con);
	}

1658
	__schedule_osd_timeout(osdc);
A
Alex Elder 已提交
1659
	__send_queued(osdc);
S
Sage Weil 已提交
1660 1661 1662 1663
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
}

1664 1665 1666 1667 1668
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);
1669
	unsigned long delay = osdc->client->options->osd_idle_ttl / 4;
1670 1671 1672

	dout("osds timeout\n");
	down_read(&osdc->map_sem);
1673
	remove_old_osds(osdc);
1674 1675 1676 1677 1678 1679
	up_read(&osdc->map_sem);

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

1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782
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
1783 1784 1785 1786 1787
static void complete_request(struct ceph_osd_request *req)
{
	complete_all(&req->r_safe_completion);  /* fsync waiter */
}

S
Sage Weil 已提交
1788 1789 1790 1791
/*
 * handle osd op reply.  either call the callback if it is specified,
 * or do the completion to wake up the waiting thread.
 */
1792
static void handle_reply(struct ceph_osd_client *osdc, struct ceph_msg *msg)
S
Sage Weil 已提交
1793
{
1794
	void *p, *end;
S
Sage Weil 已提交
1795
	struct ceph_osd_request *req;
1796
	struct ceph_request_redirect redir;
S
Sage Weil 已提交
1797
	u64 tid;
1798
	int object_len;
1799 1800
	unsigned int numops;
	int payload_len, flags;
1801
	s32 result;
1802 1803 1804 1805 1806 1807
	s32 retry_attempt;
	struct ceph_pg pg;
	int err;
	u32 reassert_epoch;
	u64 reassert_version;
	u32 osdmap_epoch;
1808
	int already_completed;
1809
	u32 bytes;
I
Ilya Dryomov 已提交
1810
	u8 decode_redir;
1811
	unsigned int i;
S
Sage Weil 已提交
1812

1813
	tid = le64_to_cpu(msg->hdr.tid);
1814 1815 1816 1817 1818 1819 1820 1821 1822 1823
	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;

1824
	err = ceph_decode_pgid(&p, end, &pg);
1825
	if (err)
S
Sage Weil 已提交
1826
		goto bad;
1827 1828 1829 1830 1831 1832 1833 1834

	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
1835
	/* lookup */
1836
	down_read(&osdc->map_sem);
S
Sage Weil 已提交
1837 1838 1839 1840
	mutex_lock(&osdc->request_mutex);
	req = __lookup_request(osdc, tid);
	if (req == NULL) {
		dout("handle_reply tid %llu dne\n", tid);
1841
		goto bad_mutex;
S
Sage Weil 已提交
1842 1843
	}
	ceph_osdc_get_request(req);
1844 1845 1846 1847

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

1848
	ceph_decode_need(&p, end, 4, bad_put);
1849
	numops = ceph_decode_32(&p);
1850
	if (numops > CEPH_OSD_MAX_OPS)
1851 1852 1853 1854
		goto bad_put;
	if (numops != req->r_num_ops)
		goto bad_put;
	payload_len = 0;
1855
	ceph_decode_need(&p, end, numops * sizeof(struct ceph_osd_op), bad_put);
1856 1857 1858 1859 1860
	for (i = 0; i < numops; i++) {
		struct ceph_osd_op *op = p;
		int len;

		len = le32_to_cpu(op->payload_len);
1861
		req->r_ops[i].outdata_len = len;
1862 1863 1864 1865
		dout(" op %d has %d bytes\n", i, len);
		payload_len += len;
		p += sizeof(*op);
	}
1866 1867
	bytes = le32_to_cpu(msg->hdr.data_len);
	if (payload_len != bytes) {
1868 1869
		pr_warn("sum of op payload lens %d != data_len %d\n",
			payload_len, bytes);
1870 1871 1872
		goto bad_put;
	}

1873
	ceph_decode_need(&p, end, 4 + numops * 4, bad_put);
1874 1875
	retry_attempt = ceph_decode_32(&p);
	for (i = 0; i < numops; i++)
1876
		req->r_ops[i].rval = ceph_decode_32(&p);
S
Sage Weil 已提交
1877

1878 1879 1880
	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
1881

I
Ilya Dryomov 已提交
1882 1883 1884 1885 1886 1887 1888 1889 1890
		if (le16_to_cpu(msg->hdr.version) >= 7)
			ceph_decode_8_safe(&p, end, decode_redir, bad_put);
		else
			decode_redir = 1;
	} else {
		decode_redir = 0;
	}

	if (decode_redir) {
1891 1892 1893 1894 1895 1896
		err = ceph_redirect_decode(&p, end, &redir);
		if (err)
			goto bad_put;
	} else {
		redir.oloc.pool = -1;
	}
S
Sage Weil 已提交
1897

1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912
	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.
		 */
1913
		err = __ceph_osdc_start_request(osdc, req, true);
1914 1915
		BUG_ON(err);

1916
		goto out_unlock;
1917 1918 1919 1920
	}

	already_completed = req->r_got_reply;
	if (!req->r_got_reply) {
1921
		req->r_result = result;
S
Sage Weil 已提交
1922 1923 1924 1925 1926 1927
		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, */
1928 1929
		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
1930 1931 1932 1933

		req->r_got_reply = 1;
	} else if ((flags & CEPH_OSD_FLAG_ONDISK) == 0) {
		dout("handle_reply tid %llu dup ack\n", tid);
1934
		goto out_unlock;
S
Sage Weil 已提交
1935 1936 1937 1938
	}

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

1939 1940 1941
	if (req->r_linger && (flags & CEPH_OSD_FLAG_ONDISK))
		__register_linger_request(osdc, req);

S
Sage Weil 已提交
1942
	/* either this is a read, or we got the safe response */
1943 1944
	if (result < 0 ||
	    (flags & CEPH_OSD_FLAG_ONDISK) ||
S
Sage Weil 已提交
1945 1946 1947 1948
	    ((flags & CEPH_OSD_FLAG_WRITE) == 0))
		__unregister_request(osdc, req);

	mutex_unlock(&osdc->request_mutex);
1949
	up_read(&osdc->map_sem);
S
Sage Weil 已提交
1950

Y
Yan, Zheng 已提交
1951
	if (!already_completed) {
1952 1953 1954
		if (req->r_unsafe_callback &&
		    result >= 0 && !(flags & CEPH_OSD_FLAG_ONDISK))
			req->r_unsafe_callback(req, true);
Y
Yan, Zheng 已提交
1955 1956 1957 1958 1959
		if (req->r_callback)
			req->r_callback(req, msg);
		else
			complete_all(&req->r_completion);
	}
S
Sage Weil 已提交
1960

1961 1962 1963
	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
1964
		complete_request(req);
1965
	}
S
Sage Weil 已提交
1966

1967
out:
1968
	dout("req=%p req->r_linger=%d\n", req, req->r_linger);
S
Sage Weil 已提交
1969 1970
	ceph_osdc_put_request(req);
	return;
1971 1972 1973 1974
out_unlock:
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
	goto out;
S
Sage Weil 已提交
1975

1976
bad_put:
1977 1978 1979 1980 1981 1982 1983
	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);
1984
	ceph_osdc_put_request(req);
1985 1986
bad_mutex:
	mutex_unlock(&osdc->request_mutex);
1987
	up_read(&osdc->map_sem);
S
Sage Weil 已提交
1988
bad:
1989 1990
	pr_err("corrupt osd_op_reply got %d %d\n",
	       (int)msg->front.iov_len, le32_to_cpu(msg->hdr.front_len));
1991
	ceph_msg_dump(msg);
S
Sage Weil 已提交
1992 1993
}

1994
static void reset_changed_osds(struct ceph_osd_client *osdc)
S
Sage Weil 已提交
1995 1996 1997
{
	struct rb_node *p, *n;

1998
	dout("%s %p\n", __func__, osdc);
1999 2000
	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
2001

2002 2003 2004 2005 2006 2007 2008
		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
2009
	}
2010 2011 2012
}

/*
2013 2014
 * Requeue requests whose mapping to an OSD has changed.  If requests map to
 * no osd, request a new map.
2015
 *
2016
 * Caller should hold map_sem for read.
2017
 */
2018 2019
static void kick_requests(struct ceph_osd_client *osdc, bool force_resend,
			  bool force_resend_writes)
2020
{
2021
	struct ceph_osd_request *req, *nreq;
2022 2023 2024
	struct rb_node *p;
	int needmap = 0;
	int err;
2025
	bool force_resend_req;
2026

2027 2028
	dout("kick_requests %s %s\n", force_resend ? " (force resend)" : "",
		force_resend_writes ? " (force resend writes)" : "");
2029
	mutex_lock(&osdc->request_mutex);
2030
	for (p = rb_first(&osdc->requests); p; ) {
2031
		req = rb_entry(p, struct ceph_osd_request, r_node);
2032
		p = rb_next(p);
2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045

		/*
		 * 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);
2046
			ceph_osdc_get_request(req);
2047 2048
			__unregister_request(osdc, req);
			__register_linger_request(osdc, req);
2049
			ceph_osdc_put_request(req);
2050 2051 2052
			continue;
		}

2053 2054 2055 2056
		force_resend_req = force_resend ||
			(force_resend_writes &&
				req->r_flags & CEPH_OSD_FLAG_WRITE);
		err = __map_request(osdc, req, force_resend_req);
2057 2058 2059 2060 2061 2062
		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) {
2063 2064 2065 2066
			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);
2067
				req->r_flags |= CEPH_OSD_FLAG_RETRY;
2068 2069
			}
		}
2070 2071 2072 2073 2074 2075
	}

	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);

2076 2077
		err = __map_request(osdc, req,
				    force_resend || force_resend_writes);
2078
		dout("__map_request returned %d\n", err);
2079 2080
		if (err < 0)
			continue;  /* hrm! */
2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095
		if (req->r_osd == NULL || err > 0) {
			if (req->r_osd == NULL) {
				dout("lingering %p tid %llu maps to no osd\n",
				     req, req->r_tid);
				/*
				 * A homeless lingering request makes
				 * no sense, as it's job is to keep
				 * a particular OSD connection open.
				 * Request a newer map and kick the
				 * request, knowing that it won't be
				 * resent until we actually get a map
				 * that can tell us where to send it.
				 */
				needmap++;
			}
2096

2097 2098 2099 2100 2101
			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);
			__unregister_linger_request(osdc, req);
		}
2102
	}
2103
	reset_changed_osds(osdc);
S
Sage Weil 已提交
2104 2105 2106 2107 2108 2109
	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);
	}
2110
}
2111 2112


S
Sage Weil 已提交
2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127
/*
 * 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;
2128
	bool was_full;
S
Sage Weil 已提交
2129 2130 2131 2132 2133 2134 2135 2136

	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));
2137 2138
	if (ceph_check_fsid(osdc->client, &fsid) < 0)
		return;
S
Sage Weil 已提交
2139 2140 2141

	down_write(&osdc->map_sem);

2142 2143
	was_full = ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_FULL);

S
Sage Weil 已提交
2144 2145 2146 2147 2148
	/* 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);
2149 2150
		epoch = ceph_decode_32(&p);
		maplen = ceph_decode_32(&p);
S
Sage Weil 已提交
2151 2152 2153 2154 2155 2156 2157
		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,
2158
							  &osdc->client->msgr);
S
Sage Weil 已提交
2159 2160 2161 2162
			if (IS_ERR(newmap)) {
				err = PTR_ERR(newmap);
				goto bad;
			}
2163
			BUG_ON(!newmap);
S
Sage Weil 已提交
2164 2165 2166 2167
			if (newmap != osdc->osdmap) {
				ceph_osdmap_destroy(osdc->osdmap);
				osdc->osdmap = newmap;
			}
2168 2169 2170 2171
			was_full = was_full ||
				ceph_osdmap_flag(osdc->osdmap,
						 CEPH_OSDMAP_FULL);
			kick_requests(osdc, 0, was_full);
S
Sage Weil 已提交
2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186
		} 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);
2187 2188
		epoch = ceph_decode_32(&p);
		maplen = ceph_decode_32(&p);
S
Sage Weil 已提交
2189 2190 2191 2192 2193 2194 2195 2196 2197
		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 {
2198 2199
			int skipped_map = 0;

S
Sage Weil 已提交
2200
			dout("taking full map %u len %d\n", epoch, maplen);
2201
			newmap = ceph_osdmap_decode(&p, p+maplen);
S
Sage Weil 已提交
2202 2203 2204 2205
			if (IS_ERR(newmap)) {
				err = PTR_ERR(newmap);
				goto bad;
			}
2206
			BUG_ON(!newmap);
S
Sage Weil 已提交
2207 2208
			oldmap = osdc->osdmap;
			osdc->osdmap = newmap;
2209 2210 2211
			if (oldmap) {
				if (oldmap->epoch + 1 < newmap->epoch)
					skipped_map = 1;
S
Sage Weil 已提交
2212
				ceph_osdmap_destroy(oldmap);
2213
			}
2214 2215 2216 2217
			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
2218 2219 2220 2221 2222
		}
		p += maplen;
		nr_maps--;
	}

2223 2224
	if (!osdc->osdmap)
		goto bad;
S
Sage Weil 已提交
2225 2226
done:
	downgrade_write(&osdc->map_sem);
2227 2228
	ceph_monc_got_map(&osdc->client->monc, CEPH_SUB_OSDMAP,
			  osdc->osdmap->epoch);
2229 2230 2231 2232 2233 2234

	/*
	 * subscribe to subsequent osdmap updates if full to ensure
	 * we find out when we are no longer full and stop returning
	 * ENOSPC.
	 */
2235 2236 2237
	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))
2238 2239
		ceph_monc_request_next_osdmap(&osdc->client->monc);

A
Alex Elder 已提交
2240 2241 2242
	mutex_lock(&osdc->request_mutex);
	__send_queued(osdc);
	mutex_unlock(&osdc->request_mutex);
S
Sage Weil 已提交
2243
	up_read(&osdc->map_sem);
2244
	wake_up_all(&osdc->client->auth_wq);
S
Sage Weil 已提交
2245 2246 2247 2248
	return;

bad:
	pr_err("osdc handle_map corrupt msg\n");
2249
	ceph_msg_dump(msg);
S
Sage Weil 已提交
2250 2251 2252
	up_write(&osdc->map_sem);
}

2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 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 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334
/*
 * 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 *),
2335
			   void *data, struct ceph_osd_event **pevent)
2336 2337 2338 2339 2340 2341 2342 2343 2344
{
	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;
2345
	event->one_shot = 0;
2346 2347 2348
	event->data = data;
	event->osdc = osdc;
	INIT_LIST_HEAD(&event->osd_node);
2349
	RB_CLEAR_NODE(&event->node);
2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395
	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
 */
2396 2397
static void handle_watch_notify(struct ceph_osd_client *osdc,
				struct ceph_msg *msg)
2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417
{
	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) {
2418
		BUG_ON(event->one_shot);
2419 2420 2421 2422 2423 2424 2425 2426
		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) {
2427 2428 2429
			pr_err("couldn't allocate event_work\n");
			ceph_osdc_put_event(event);
			return;
2430
		}
2431
		INIT_WORK(&event_work->work, do_event_work);
2432 2433 2434 2435 2436
		event_work->event = event;
		event_work->ver = ver;
		event_work->notify_id = notify_id;
		event_work->opcode = opcode;

2437 2438
		queue_work(osdc->notify_wq, &event_work->work);
	}
2439 2440 2441 2442 2443 2444 2445

	return;

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

2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461
/*
 * 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;
2462
	WARN_ON(snapc != req->r_snapc);
2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493

	/* 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 */
2494 2495 2496 2497 2498
	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;
2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512

	/* 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) {
2513
		for (i = 0; i < req->r_snapc->num_snaps; i++) {
2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544
			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);

2545 2546 2547 2548 2549 2550 2551
/*
 * Register request, send initial attempt.
 */
int ceph_osdc_start_request(struct ceph_osd_client *osdc,
			    struct ceph_osd_request *req,
			    bool nofail)
{
2552
	int rc;
2553

S
Sage Weil 已提交
2554 2555
	down_read(&osdc->map_sem);
	mutex_lock(&osdc->request_mutex);
2556 2557 2558

	rc = __ceph_osdc_start_request(osdc, req, nofail);

S
Sage Weil 已提交
2559 2560
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
2561

S
Sage Weil 已提交
2562 2563
	return rc;
}
2564
EXPORT_SYMBOL(ceph_osdc_start_request);
S
Sage Weil 已提交
2565

2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584
/*
 * 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
2585 2586 2587 2588 2589 2590 2591 2592
/*
 * wait for a request to complete
 */
int ceph_osdc_wait_request(struct ceph_osd_client *osdc,
			   struct ceph_osd_request *req)
{
	int rc;

2593 2594
	dout("%s %p tid %llu\n", __func__, req, req->r_tid);

S
Sage Weil 已提交
2595 2596
	rc = wait_for_completion_interruptible(&req->r_completion);
	if (rc < 0) {
2597 2598
		dout("%s %p tid %llu interrupted\n", __func__, req, req->r_tid);
		ceph_osdc_cancel_request(req);
S
Sage Weil 已提交
2599
		complete_request(req);
S
Sage Weil 已提交
2600 2601 2602
		return rc;
	}

2603 2604
	dout("%s %p tid %llu result %d\n", __func__, req, req->r_tid,
	     req->r_result);
S
Sage Weil 已提交
2605 2606
	return req->r_result;
}
2607
EXPORT_SYMBOL(ceph_osdc_wait_request);
S
Sage Weil 已提交
2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640

/*
 * 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);
}
2641
EXPORT_SYMBOL(ceph_osdc_sync);
S
Sage Weil 已提交
2642

2643 2644 2645 2646
/*
 * 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
2647
void ceph_osdc_flush_notifies(struct ceph_osd_client *osdc)
2648 2649 2650 2651 2652 2653
{
	flush_workqueue(osdc->notify_wq);
}
EXPORT_SYMBOL(ceph_osdc_flush_notifies);


S
Sage Weil 已提交
2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669
/*
 * 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;
2670
	INIT_LIST_HEAD(&osdc->osd_lru);
S
Sage Weil 已提交
2671
	osdc->requests = RB_ROOT;
2672
	INIT_LIST_HEAD(&osdc->req_lru);
2673 2674
	INIT_LIST_HEAD(&osdc->req_unsent);
	INIT_LIST_HEAD(&osdc->req_notarget);
2675
	INIT_LIST_HEAD(&osdc->req_linger);
S
Sage Weil 已提交
2676 2677
	osdc->num_requests = 0;
	INIT_DELAYED_WORK(&osdc->timeout_work, handle_timeout);
2678
	INIT_DELAYED_WORK(&osdc->osds_timeout_work, handle_osds_timeout);
2679 2680 2681
	spin_lock_init(&osdc->event_lock);
	osdc->event_tree = RB_ROOT;
	osdc->event_count = 0;
2682 2683

	schedule_delayed_work(&osdc->osds_timeout_work,
2684
	    round_jiffies_relative(osdc->client->options->osd_idle_ttl));
S
Sage Weil 已提交
2685

2686
	err = -ENOMEM;
2687 2688
	osdc->req_mempool = mempool_create_slab_pool(10,
						     ceph_osd_request_cache);
S
Sage Weil 已提交
2689
	if (!osdc->req_mempool)
2690
		goto out;
S
Sage Weil 已提交
2691

2692 2693
	err = ceph_msgpool_init(&osdc->msgpool_op, CEPH_MSG_OSD_OP,
				OSD_OP_FRONT_LEN, 10, true,
2694
				"osd_op");
S
Sage Weil 已提交
2695
	if (err < 0)
2696
		goto out_mempool;
2697
	err = ceph_msgpool_init(&osdc->msgpool_op_reply, CEPH_MSG_OSD_OPREPLY,
2698 2699
				OSD_OPREPLY_FRONT_LEN, 10, true,
				"osd_op_reply");
S
Sage Weil 已提交
2700 2701
	if (err < 0)
		goto out_msgpool;
2702

2703
	err = -ENOMEM;
2704
	osdc->notify_wq = create_singlethread_workqueue("ceph-watch-notify");
2705
	if (!osdc->notify_wq)
2706 2707
		goto out_msgpool_reply;

S
Sage Weil 已提交
2708
	return 0;
2709

2710 2711
out_msgpool_reply:
	ceph_msgpool_destroy(&osdc->msgpool_op_reply);
S
Sage Weil 已提交
2712 2713
out_msgpool:
	ceph_msgpool_destroy(&osdc->msgpool_op);
2714 2715 2716 2717
out_mempool:
	mempool_destroy(osdc->req_mempool);
out:
	return err;
S
Sage Weil 已提交
2718 2719 2720 2721
}

void ceph_osdc_stop(struct ceph_osd_client *osdc)
{
2722 2723
	flush_workqueue(osdc->notify_wq);
	destroy_workqueue(osdc->notify_wq);
S
Sage Weil 已提交
2724
	cancel_delayed_work_sync(&osdc->timeout_work);
2725
	cancel_delayed_work_sync(&osdc->osds_timeout_work);
S
Sage Weil 已提交
2726 2727 2728 2729
	if (osdc->osdmap) {
		ceph_osdmap_destroy(osdc->osdmap);
		osdc->osdmap = NULL;
	}
2730
	remove_all_osds(osdc);
S
Sage Weil 已提交
2731 2732
	mempool_destroy(osdc->req_mempool);
	ceph_msgpool_destroy(&osdc->msgpool_op);
S
Sage Weil 已提交
2733
	ceph_msgpool_destroy(&osdc->msgpool_op_reply);
S
Sage Weil 已提交
2734 2735 2736 2737 2738 2739 2740 2741 2742 2743
}

/*
 * 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,
2744
			struct page **pages, int num_pages, int page_align)
S
Sage Weil 已提交
2745 2746 2747 2748 2749 2750
{
	struct ceph_osd_request *req;
	int rc = 0;

	dout("readpages on ino %llx.%llx on %llu~%llu\n", vino.ino,
	     vino.snap, off, *plen);
2751
	req = ceph_osdc_new_request(osdc, layout, vino, off, plen, 0, 1,
S
Sage Weil 已提交
2752
				    CEPH_OSD_OP_READ, CEPH_OSD_FLAG_READ,
2753
				    NULL, truncate_seq, truncate_size,
2754
				    false);
2755 2756
	if (IS_ERR(req))
		return PTR_ERR(req);
S
Sage Weil 已提交
2757 2758

	/* it may be a short read due to an object boundary */
2759

2760
	osd_req_op_extent_osd_data_pages(req, 0,
2761
				pages, *plen, page_align, false, false);
S
Sage Weil 已提交
2762

2763
	dout("readpages  final extent is %llu~%llu (%llu bytes align %d)\n",
2764
	     off, *plen, *plen, page_align);
S
Sage Weil 已提交
2765

2766
	ceph_osdc_build_request(req, off, NULL, vino.snap, NULL);
2767

S
Sage Weil 已提交
2768 2769 2770 2771 2772 2773 2774 2775
	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;
}
2776
EXPORT_SYMBOL(ceph_osdc_readpages);
S
Sage Weil 已提交
2777 2778 2779 2780 2781 2782 2783 2784 2785 2786

/*
 * 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,
2787
			 struct page **pages, int num_pages)
S
Sage Weil 已提交
2788 2789 2790
{
	struct ceph_osd_request *req;
	int rc = 0;
2791
	int page_align = off & ~PAGE_MASK;
S
Sage Weil 已提交
2792

2793
	BUG_ON(vino.snap != CEPH_NOSNAP);	/* snapshots aren't writeable */
2794
	req = ceph_osdc_new_request(osdc, layout, vino, off, &len, 0, 1,
S
Sage Weil 已提交
2795
				    CEPH_OSD_OP_WRITE,
2796
				    CEPH_OSD_FLAG_ONDISK | CEPH_OSD_FLAG_WRITE,
2797
				    snapc, truncate_seq, truncate_size,
2798
				    true);
2799 2800
	if (IS_ERR(req))
		return PTR_ERR(req);
S
Sage Weil 已提交
2801 2802

	/* it may be a short write due to an object boundary */
2803
	osd_req_op_extent_osd_data_pages(req, 0, pages, len, page_align,
2804 2805
				false, false);
	dout("writepages %llu~%llu (%llu bytes)\n", off, len, len);
S
Sage Weil 已提交
2806

2807
	ceph_osdc_build_request(req, off, snapc, CEPH_NOSNAP, mtime);
2808

2809
	rc = ceph_osdc_start_request(osdc, req, true);
S
Sage Weil 已提交
2810 2811 2812 2813 2814 2815 2816 2817 2818
	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;
}
2819
EXPORT_SYMBOL(ceph_osdc_writepages);
S
Sage Weil 已提交
2820

2821 2822
int ceph_osdc_setup(void)
{
2823 2824 2825
	size_t size = sizeof(struct ceph_osd_request) +
	    CEPH_OSD_SLAB_OPS * sizeof(struct ceph_osd_req_op);

2826
	BUG_ON(ceph_osd_request_cache);
2827 2828
	ceph_osd_request_cache = kmem_cache_create("ceph_osd_request", size,
						   0, 0, NULL);
2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841

	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
2842 2843 2844 2845 2846 2847
/*
 * handle incoming message
 */
static void dispatch(struct ceph_connection *con, struct ceph_msg *msg)
{
	struct ceph_osd *osd = con->private;
2848
	struct ceph_osd_client *osdc;
S
Sage Weil 已提交
2849 2850 2851
	int type = le16_to_cpu(msg->hdr.type);

	if (!osd)
S
Sage Weil 已提交
2852
		goto out;
2853
	osdc = osd->o_osdc;
S
Sage Weil 已提交
2854 2855 2856 2857 2858 2859

	switch (type) {
	case CEPH_MSG_OSD_MAP:
		ceph_osdc_handle_map(osdc, msg);
		break;
	case CEPH_MSG_OSD_OPREPLY:
2860
		handle_reply(osdc, msg);
S
Sage Weil 已提交
2861
		break;
2862 2863 2864
	case CEPH_MSG_WATCH_NOTIFY:
		handle_watch_notify(osdc, msg);
		break;
S
Sage Weil 已提交
2865 2866 2867 2868 2869

	default:
		pr_err("received unknown message type %d %s\n", type,
		       ceph_msg_type_name(type));
	}
S
Sage Weil 已提交
2870
out:
S
Sage Weil 已提交
2871 2872 2873
	ceph_msg_put(msg);
}

2874
/*
2875 2876 2877
 * Lookup and return message for incoming reply.  Don't try to do
 * anything about a larger than preallocated data portion of the
 * message at the moment - for now, just skip the message.
2878 2879
 */
static struct ceph_msg *get_reply(struct ceph_connection *con,
2880 2881
				  struct ceph_msg_header *hdr,
				  int *skip)
S
Sage Weil 已提交
2882 2883 2884
{
	struct ceph_osd *osd = con->private;
	struct ceph_osd_client *osdc = osd->o_osdc;
2885
	struct ceph_msg *m;
2886
	struct ceph_osd_request *req;
2887
	int front_len = le32_to_cpu(hdr->front_len);
2888
	int data_len = le32_to_cpu(hdr->data_len);
2889
	u64 tid;
S
Sage Weil 已提交
2890

2891 2892 2893 2894
	tid = le64_to_cpu(hdr->tid);
	mutex_lock(&osdc->request_mutex);
	req = __lookup_request(osdc, tid);
	if (!req) {
2895 2896
		dout("%s osd%d tid %llu unknown, skipping\n", __func__,
		     osd->o_osd, tid);
2897
		m = NULL;
2898
		*skip = 1;
2899 2900
		goto out;
	}
S
Sage Weil 已提交
2901

2902
	ceph_msg_revoke_incoming(req->r_reply);
2903

2904
	if (front_len > req->r_reply->front_alloc_len) {
2905 2906 2907
		pr_warn("%s osd%d tid %llu front %d > preallocated %d\n",
			__func__, osd->o_osd, req->r_tid, front_len,
			req->r_reply->front_alloc_len);
2908 2909
		m = ceph_msg_new(CEPH_MSG_OSD_OPREPLY, front_len, GFP_NOFS,
				 false);
2910
		if (!m)
S
Sage Weil 已提交
2911 2912 2913 2914
			goto out;
		ceph_msg_put(req->r_reply);
		req->r_reply = m;
	}
2915

2916 2917 2918 2919 2920 2921 2922
	if (data_len > req->r_reply->data_length) {
		pr_warn("%s osd%d tid %llu data %d > preallocated %zu, skipping\n",
			__func__, osd->o_osd, req->r_tid, data_len,
			req->r_reply->data_length);
		m = NULL;
		*skip = 1;
		goto out;
2923
	}
2924 2925

	m = ceph_msg_get(req->r_reply);
S
Sage Weil 已提交
2926
	dout("get_reply tid %lld %p\n", tid, m);
2927 2928 2929

out:
	mutex_unlock(&osdc->request_mutex);
2930
	return m;
2931 2932 2933 2934 2935 2936 2937 2938 2939 2940
}

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
2941
	*skip = 0;
2942 2943
	switch (type) {
	case CEPH_MSG_OSD_MAP:
2944
	case CEPH_MSG_WATCH_NOTIFY:
2945
		return ceph_msg_new(type, front, GFP_NOFS, false);
2946 2947 2948 2949 2950 2951 2952 2953
	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
2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972
}

/*
 * 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);
}

2973 2974 2975
/*
 * authentication
 */
2976 2977 2978 2979 2980
/*
 * 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,
2981
					int *proto, int force_new)
2982 2983 2984 2985
{
	struct ceph_osd *o = con->private;
	struct ceph_osd_client *osdc = o->o_osdc;
	struct ceph_auth_client *ac = osdc->client->monc.auth;
2986
	struct ceph_auth_handshake *auth = &o->o_auth;
2987

2988
	if (force_new && auth->authorizer) {
2989
		ceph_auth_destroy_authorizer(auth->authorizer);
2990 2991
		auth->authorizer = NULL;
	}
2992 2993 2994
	if (!auth->authorizer) {
		int ret = ceph_auth_create_authorizer(ac, CEPH_ENTITY_TYPE_OSD,
						      auth);
2995
		if (ret)
2996
			return ERR_PTR(ret);
2997 2998
	} else {
		int ret = ceph_auth_update_authorizer(ac, CEPH_ENTITY_TYPE_OSD,
2999 3000 3001
						     auth);
		if (ret)
			return ERR_PTR(ret);
3002 3003
	}
	*proto = ac->protocol;
3004

3005
	return auth;
3006 3007 3008 3009 3010 3011 3012 3013 3014
}


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;

3015
	return ceph_auth_verify_authorizer_reply(ac, o->o_auth.authorizer, len);
3016 3017
}

3018 3019 3020 3021 3022 3023
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;

3024
	ceph_auth_invalidate_authorizer(ac, CEPH_ENTITY_TYPE_OSD);
3025 3026
	return ceph_monc_validate_auth(&osdc->client->monc);
}
3027

3028
static int osd_sign_message(struct ceph_msg *msg)
Y
Yan, Zheng 已提交
3029
{
3030
	struct ceph_osd *o = msg->con->private;
Y
Yan, Zheng 已提交
3031
	struct ceph_auth_handshake *auth = &o->o_auth;
3032

Y
Yan, Zheng 已提交
3033 3034 3035
	return ceph_auth_sign_message(auth, msg);
}

3036
static int osd_check_message_signature(struct ceph_msg *msg)
Y
Yan, Zheng 已提交
3037
{
3038
	struct ceph_osd *o = msg->con->private;
Y
Yan, Zheng 已提交
3039
	struct ceph_auth_handshake *auth = &o->o_auth;
3040

Y
Yan, Zheng 已提交
3041 3042 3043
	return ceph_auth_check_message_signature(auth, msg);
}

3044
static const struct ceph_connection_operations osd_con_ops = {
S
Sage Weil 已提交
3045 3046 3047
	.get = get_osd_con,
	.put = put_osd_con,
	.dispatch = dispatch,
3048 3049
	.get_authorizer = get_authorizer,
	.verify_authorizer_reply = verify_authorizer_reply,
3050
	.invalidate_authorizer = invalidate_authorizer,
S
Sage Weil 已提交
3051
	.alloc_msg = alloc_msg,
3052 3053
	.sign_message = osd_sign_message,
	.check_message_signature = osd_check_message_signature,
3054
	.fault = osd_reset,
S
Sage Weil 已提交
3055
};