osd_client.c 80.0 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
#define OSD_OPREPLY_FRONT_LEN	512
23

24 25
static struct kmem_cache	*ceph_osd_request_cache;

26
static const struct ceph_connection_operations osd_con_ops;
S
Sage Weil 已提交
27

A
Alex Elder 已提交
28
static void __send_queued(struct ceph_osd_client *osdc);
29
static int __reset_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd);
30 31
static void __register_request(struct ceph_osd_client *osdc,
			       struct ceph_osd_request *req);
32 33
static void __unregister_request(struct ceph_osd_client *osdc,
				 struct ceph_osd_request *req);
34 35
static void __unregister_linger_request(struct ceph_osd_client *osdc,
					struct ceph_osd_request *req);
36
static void __enqueue_request(struct ceph_osd_request *req);
37 38
static void __send_request(struct ceph_osd_client *osdc,
			   struct ceph_osd_request *req);
S
Sage Weil 已提交
39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

/*
 * 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.
 */
65
static int calc_layout(struct ceph_file_layout *layout, u64 off, u64 *plen,
66
			u64 *objnum, u64 *objoff, u64 *objlen)
S
Sage Weil 已提交
67
{
68
	u64 orig_len = *plen;
S
Sage Weil 已提交
69
	int r;
S
Sage Weil 已提交
70

71
	/* object extent? */
72 73
	r = ceph_calc_file_object_mapping(layout, off, orig_len, objnum,
					  objoff, objlen);
S
Sage Weil 已提交
74 75
	if (r < 0)
		return r;
76 77
	if (*objlen < orig_len) {
		*plen = *objlen;
78 79 80 81
		dout(" skipping last %llu, final file extent %llu~%llu\n",
		     orig_len - *plen, off, *plen);
	}

82
	dout("calc_layout objnum=%llx %llu~%llu\n", *objnum, *objoff, *objlen);
S
Sage Weil 已提交
83

84
	return 0;
S
Sage Weil 已提交
85 86
}

A
Alex Elder 已提交
87 88 89 90 91 92
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;
}

93
static void ceph_osd_data_pages_init(struct ceph_osd_data *osd_data,
94 95 96 97 98 99 100 101 102 103 104
			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;
}

105
static void ceph_osd_data_pagelist_init(struct ceph_osd_data *osd_data,
106 107 108 109 110 111 112
			struct ceph_pagelist *pagelist)
{
	osd_data->type = CEPH_OSD_DATA_TYPE_PAGELIST;
	osd_data->pagelist = pagelist;
}

#ifdef CONFIG_BLOCK
113
static void ceph_osd_data_bio_init(struct ceph_osd_data *osd_data,
114 115 116 117 118 119 120 121
			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 */

122 123 124 125 126 127 128
#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;					\
})
129

A
Alex Elder 已提交
130 131 132 133 134 135 136 137
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;
}

138 139
struct ceph_osd_data *
osd_req_op_extent_osd_data(struct ceph_osd_request *osd_req,
140
			unsigned int which)
141
{
142
	return osd_req_op_data(osd_req, which, extent, osd_data);
143 144 145 146 147 148 149
}
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)
{
150
	return osd_req_op_data(osd_req, which, cls, response_data);
151 152 153
}
EXPORT_SYMBOL(osd_req_op_cls_response_data);	/* ??? */

A
Alex Elder 已提交
154 155 156 157 158 159 160 161 162 163 164 165 166
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);

167
void osd_req_op_extent_osd_data_pages(struct ceph_osd_request *osd_req,
168 169
			unsigned int which, struct page **pages,
			u64 length, u32 alignment,
170 171 172 173
			bool pages_from_pool, bool own_pages)
{
	struct ceph_osd_data *osd_data;

174
	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
175 176 177 178 179 180
	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,
181
			unsigned int which, struct ceph_pagelist *pagelist)
182 183 184
{
	struct ceph_osd_data *osd_data;

185
	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
186 187 188 189 190 191
	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,
192
			unsigned int which, struct bio *bio, size_t bio_length)
193 194
{
	struct ceph_osd_data *osd_data;
195 196

	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
197 198 199 200 201 202 203 204 205 206 207
	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;

208
	osd_data = osd_req_op_data(osd_req, which, cls, request_info);
209 210 211
	ceph_osd_data_pagelist_init(osd_data, pagelist);
}

212 213 214 215 216 217
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;

218
	osd_data = osd_req_op_data(osd_req, which, cls, request_data);
219 220 221 222
	ceph_osd_data_pagelist_init(osd_data, pagelist);
}
EXPORT_SYMBOL(osd_req_op_cls_request_data_pagelist);

223 224 225 226 227 228 229 230 231 232 233 234
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);

235 236 237 238 239 240
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;

241
	osd_data = osd_req_op_data(osd_req, which, cls, response_data);
242 243 244 245 246
	ceph_osd_data_pages_init(osd_data, pages, length, alignment,
				pages_from_pool, own_pages);
}
EXPORT_SYMBOL(osd_req_op_cls_response_data_pages);

247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
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
266 267
static void ceph_osd_data_release(struct ceph_osd_data *osd_data)
{
268
	if (osd_data->type == CEPH_OSD_DATA_TYPE_PAGES && osd_data->own_pages) {
A
Alex Elder 已提交
269 270 271 272 273 274
		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);
	}
275 276 277 278 279 280 281 282 283 284 285 286 287 288
	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:
289
	case CEPH_OSD_OP_WRITEFULL:
290 291 292 293
		ceph_osd_data_release(&op->extent.osd_data);
		break;
	case CEPH_OSD_OP_CALL:
		ceph_osd_data_release(&op->cls.request_info);
294
		ceph_osd_data_release(&op->cls.request_data);
295 296
		ceph_osd_data_release(&op->cls.response_data);
		break;
297 298 299 300
	case CEPH_OSD_OP_SETXATTR:
	case CEPH_OSD_OP_CMPXATTR:
		ceph_osd_data_release(&op->xattr.osd_data);
		break;
301 302 303
	case CEPH_OSD_OP_STAT:
		ceph_osd_data_release(&op->raw_data_in);
		break;
304 305 306
	default:
		break;
	}
A
Alex Elder 已提交
307 308
}

S
Sage Weil 已提交
309 310 311
/*
 * requests
 */
312
static void ceph_osdc_release_request(struct kref *kref)
S
Sage Weil 已提交
313
{
314 315
	struct ceph_osd_request *req = container_of(kref,
					    struct ceph_osd_request, r_kref);
316
	unsigned int which;
S
Sage Weil 已提交
317

318 319
	dout("%s %p (r_request %p r_reply %p)\n", __func__, req,
	     req->r_request, req->r_reply);
320 321 322 323 324 325
	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);
326

S
Sage Weil 已提交
327 328
	if (req->r_request)
		ceph_msg_put(req->r_request);
329
	if (req->r_reply) {
330
		ceph_msg_revoke_incoming(req->r_reply);
331
		ceph_msg_put(req->r_reply);
332
	}
333

334 335
	for (which = 0; which < req->r_num_ops; which++)
		osd_req_op_data_release(req, which);
336

S
Sage Weil 已提交
337 338 339
	ceph_put_snap_context(req->r_snapc);
	if (req->r_mempool)
		mempool_free(req, req->r_osdc->req_mempool);
340
	else if (req->r_num_ops <= CEPH_OSD_SLAB_OPS)
341
		kmem_cache_free(ceph_osd_request_cache, req);
342 343
	else
		kfree(req);
S
Sage Weil 已提交
344
}
345 346 347 348 349 350 351 352 353 354 355

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)
{
356 357 358 359 360
	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);
	}
361 362
}
EXPORT_SYMBOL(ceph_osdc_put_request);
363

364
struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc,
S
Sage Weil 已提交
365
					       struct ceph_snap_context *snapc,
366
					       unsigned int num_ops,
367
					       bool use_mempool,
368
					       gfp_t gfp_flags)
S
Sage Weil 已提交
369 370
{
	struct ceph_osd_request *req;
371

S
Sage Weil 已提交
372
	if (use_mempool) {
373
		BUG_ON(num_ops > CEPH_OSD_SLAB_OPS);
374
		req = mempool_alloc(osdc->req_mempool, gfp_flags);
375 376
	} else if (num_ops <= CEPH_OSD_SLAB_OPS) {
		req = kmem_cache_alloc(ceph_osd_request_cache, gfp_flags);
S
Sage Weil 已提交
377
	} else {
378 379 380
		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
381
	}
382
	if (unlikely(!req))
383
		return NULL;
S
Sage Weil 已提交
384

385 386 387
	/* req only, each op is zeroed in _osd_req_op_init() */
	memset(req, 0, sizeof(*req));

S
Sage Weil 已提交
388 389
	req->r_osdc = osdc;
	req->r_mempool = use_mempool;
390
	req->r_num_ops = num_ops;
391 392
	req->r_snapid = CEPH_NOSNAP;
	req->r_snapc = ceph_get_snap_context(snapc);
393

S
Sage Weil 已提交
394
	kref_init(&req->r_kref);
S
Sage Weil 已提交
395 396
	init_completion(&req->r_completion);
	init_completion(&req->r_safe_completion);
397
	RB_CLEAR_NODE(&req->r_node);
S
Sage Weil 已提交
398
	INIT_LIST_HEAD(&req->r_unsafe_item);
399
	INIT_LIST_HEAD(&req->r_linger_item);
400
	INIT_LIST_HEAD(&req->r_linger_osd_item);
401
	INIT_LIST_HEAD(&req->r_req_lru_item);
402 403
	INIT_LIST_HEAD(&req->r_osd_item);

404
	req->r_base_oloc.pool = -1;
405
	req->r_target_oloc.pool = -1;
406

407 408 409 410
	dout("%s req %p\n", __func__, req);
	return req;
}
EXPORT_SYMBOL(ceph_osdc_alloc_request);
411

412 413 414 415 416
int ceph_osdc_alloc_messages(struct ceph_osd_request *req, gfp_t gfp)
{
	struct ceph_osd_client *osdc = req->r_osdc;
	struct ceph_msg *msg;
	int msg_size;
S
Sage Weil 已提交
417

418
	/* create request message */
419 420 421 422
	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 */
423 424
	msg_size += 4 + req->r_base_oid.name_len; /* oid */
	msg_size += 2 + req->r_num_ops * sizeof(struct ceph_osd_op);
425 426
	msg_size += 8; /* snapid */
	msg_size += 8; /* snap_seq */
427
	msg_size += 4 + 8 * (req->r_snapc ? req->r_snapc->num_snaps : 0);
428 429
	msg_size += 4; /* retry_attempt */

430
	if (req->r_mempool)
431
		msg = ceph_msgpool_get(&osdc->msgpool_op, 0);
S
Sage Weil 已提交
432
	else
433 434 435
		msg = ceph_msg_new(CEPH_MSG_OSD_OP, msg_size, gfp, true);
	if (!msg)
		return -ENOMEM;
436

S
Sage Weil 已提交
437
	memset(msg->front.iov_base, 0, msg->front.iov_len);
438 439
	req->r_request = msg;

440 441
	/* create reply message */
	msg_size = OSD_OPREPLY_FRONT_LEN;
442 443
	msg_size += req->r_base_oid.name_len;
	msg_size += req->r_num_ops * sizeof(struct ceph_osd_op);
444 445 446 447 448 449 450 451 452 453 454

	if (req->r_mempool)
		msg = ceph_msgpool_get(&osdc->msgpool_op_reply, 0);
	else
		msg = ceph_msg_new(CEPH_MSG_OSD_OPREPLY, msg_size, gfp, true);
	if (!msg)
		return -ENOMEM;

	req->r_reply = msg;

	return 0;
455
}
456
EXPORT_SYMBOL(ceph_osdc_alloc_messages);
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);
831 832 833 834
	if (!req) {
		r = -ENOMEM;
		goto fail;
	}
835

836
	req->r_flags = flags;
837 838

	/* calculate max write size */
839
	r = calc_layout(layout, off, plen, &objnum, &objoff, &objlen);
840 841
	if (r)
		goto fail;
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

867 868 869 870
	r = ceph_osdc_alloc_messages(req, GFP_NOFS);
	if (r)
		goto fail;

S
Sage Weil 已提交
871
	return req;
872 873 874 875

fail:
	ceph_osdc_put_request(req);
	return ERR_PTR(r);
S
Sage Weil 已提交
876
}
877
EXPORT_SYMBOL(ceph_osdc_new_request);
S
Sage Weil 已提交
878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943

/*
 * 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;
}

944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974
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);
}

975 976 977 978 979 980
/*
 * Resubmit requests pending on the given osd.
 */
static void __kick_osd_requests(struct ceph_osd_client *osdc,
				struct ceph_osd *osd)
{
981
	struct ceph_osd_request *req, *nreq;
982
	LIST_HEAD(resend);
983
	LIST_HEAD(resend_linger);
984 985
	int err;

986
	dout("%s osd%d\n", __func__, osd->o_osd);
987
	err = __reset_osd(osdc, osd);
988
	if (err)
989
		return;
990

991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007
	/*
	 * 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.
	 */
1008
	list_for_each_entry(req, &osd->o_requests, r_osd_item) {
1009 1010
		if (!req->r_sent)
			break;
1011 1012 1013 1014 1015

		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);
1016
			req->r_flags |= CEPH_OSD_FLAG_RETRY;
1017 1018 1019
		} else {
			list_move_tail(&req->r_req_lru_item, &resend_linger);
		}
1020
	}
1021
	list_splice(&resend, &osdc->req_unsent);
1022

1023
	/*
1024 1025 1026 1027
	 * 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.
1028
	 */
1029
	list_for_each_entry_safe(req, nreq, &osd->o_linger_requests,
1030
				 r_linger_osd_item) {
1031 1032
		WARN_ON(!list_empty(&req->r_req_lru_item));
		__kick_linger_request(req);
1033
	}
1034 1035 1036

	list_for_each_entry_safe(req, nreq, &resend_linger, r_req_lru_item)
		__kick_linger_request(req);
1037 1038
}

S
Sage Weil 已提交
1039
/*
1040
 * If the osd connection drops, we need to resubmit all requests.
S
Sage Weil 已提交
1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051
 */
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
1052 1053
	mutex_lock(&osdc->request_mutex);
	__kick_osd_requests(osdc, osd);
A
Alex Elder 已提交
1054
	__send_queued(osdc);
S
Sage Weil 已提交
1055
	mutex_unlock(&osdc->request_mutex);
S
Sage Weil 已提交
1056 1057 1058 1059 1060 1061
	up_read(&osdc->map_sem);
}

/*
 * Track open sessions with osds.
 */
1062
static struct ceph_osd *create_osd(struct ceph_osd_client *osdc, int onum)
S
Sage Weil 已提交
1063 1064 1065 1066 1067 1068 1069 1070 1071
{
	struct ceph_osd *osd;

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

	atomic_set(&osd->o_ref, 1);
	osd->o_osdc = osdc;
1072
	osd->o_osd = onum;
1073
	RB_CLEAR_NODE(&osd->o_node);
S
Sage Weil 已提交
1074
	INIT_LIST_HEAD(&osd->o_requests);
1075
	INIT_LIST_HEAD(&osd->o_linger_requests);
1076
	INIT_LIST_HEAD(&osd->o_osd_lru);
S
Sage Weil 已提交
1077 1078
	osd->o_incarnation = 1;

1079
	ceph_con_init(&osd->o_con, osd, &osd_con_ops, &osdc->client->msgr);
1080

1081
	INIT_LIST_HEAD(&osd->o_keepalive_item);
S
Sage Weil 已提交
1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100
	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);
1101 1102
	if (atomic_dec_and_test(&osd->o_ref)) {
		if (osd->o_auth.authorizer)
1103
			ceph_auth_destroy_authorizer(osd->o_auth.authorizer);
S
Sage Weil 已提交
1104
		kfree(osd);
S
Sage Weil 已提交
1105
	}
S
Sage Weil 已提交
1106 1107 1108 1109 1110
}

/*
 * remove an osd from our map
 */
1111
static void __remove_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
S
Sage Weil 已提交
1112
{
1113
	dout("%s %p osd%d\n", __func__, osd, osd->o_osd);
1114 1115
	WARN_ON(!list_empty(&osd->o_requests));
	WARN_ON(!list_empty(&osd->o_linger_requests));
1116

1117
	list_del_init(&osd->o_osd_lru);
1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130
	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
1131 1132
}

1133 1134
static void remove_all_osds(struct ceph_osd_client *osdc)
{
1135
	dout("%s %p\n", __func__, osdc);
1136 1137 1138 1139
	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);
1140
		remove_osd(osdc, osd);
1141 1142 1143 1144
	}
	mutex_unlock(&osdc->request_mutex);
}

1145 1146 1147
static void __move_osd_to_lru(struct ceph_osd_client *osdc,
			      struct ceph_osd *osd)
{
1148
	dout("%s %p\n", __func__, osd);
1149
	BUG_ON(!list_empty(&osd->o_osd_lru));
1150

1151
	list_add_tail(&osd->o_osd_lru, &osdc->osd_lru);
1152
	osd->lru_ttl = jiffies + osdc->client->options->osd_idle_ttl;
1153 1154
}

1155 1156 1157 1158 1159 1160 1161 1162 1163 1164
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);
}

1165 1166 1167 1168 1169 1170 1171
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);
}

1172
static void remove_old_osds(struct ceph_osd_client *osdc)
1173 1174 1175 1176 1177 1178
{
	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) {
1179
		if (time_before(jiffies, osd->lru_ttl))
1180
			break;
1181
		remove_osd(osdc, osd);
1182 1183 1184 1185
	}
	mutex_unlock(&osdc->request_mutex);
}

S
Sage Weil 已提交
1186 1187 1188
/*
 * reset osd connect
 */
1189
static int __reset_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
S
Sage Weil 已提交
1190
{
A
Alex Elder 已提交
1191
	struct ceph_entity_addr *peer_addr;
S
Sage Weil 已提交
1192

1193
	dout("__reset_osd %p osd%d\n", osd, osd->o_osd);
1194 1195
	if (list_empty(&osd->o_requests) &&
	    list_empty(&osd->o_linger_requests)) {
1196
		remove_osd(osdc, osd);
A
Alex Elder 已提交
1197 1198 1199 1200 1201 1202 1203 1204
		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;

1205 1206
		dout("osd addr hasn't changed and connection never opened, "
		     "letting msgr retry\n");
1207 1208 1209
		/* 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
1210 1211

		return -EAGAIN;
S
Sage Weil 已提交
1212
	}
A
Alex Elder 已提交
1213 1214 1215 1216 1217 1218

	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
1219 1220 1221 1222 1223 1224 1225 1226
}

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;

1227
	dout("__insert_osd %p osd%d\n", new, new->o_osd);
S
Sage Weil 已提交
1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259
	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;
}

1260 1261 1262
static void __schedule_osd_timeout(struct ceph_osd_client *osdc)
{
	schedule_delayed_work(&osdc->timeout_work,
1263
			      osdc->client->options->osd_keepalive_timeout);
1264 1265 1266 1267 1268 1269
}

static void __cancel_osd_timeout(struct ceph_osd_client *osdc)
{
	cancel_delayed_work(&osdc->timeout_work);
}
S
Sage Weil 已提交
1270 1271 1272 1273 1274

/*
 * Register request, assign tid.  If this is the first request, set up
 * the timeout event.
 */
1275 1276
static void __register_request(struct ceph_osd_client *osdc,
			       struct ceph_osd_request *req)
S
Sage Weil 已提交
1277 1278
{
	req->r_tid = ++osdc->last_tid;
1279
	req->r_request->hdr.tid = cpu_to_le64(req->r_tid);
1280
	dout("__register_request %p tid %lld\n", req, req->r_tid);
S
Sage Weil 已提交
1281 1282 1283 1284
	__insert_request(osdc, req);
	ceph_osdc_get_request(req);
	osdc->num_requests++;
	if (osdc->num_requests == 1) {
1285 1286
		dout(" first request, scheduling timeout\n");
		__schedule_osd_timeout(osdc);
S
Sage Weil 已提交
1287
	}
1288 1289
}

S
Sage Weil 已提交
1290 1291 1292 1293 1294 1295
/*
 * called under osdc->request_mutex
 */
static void __unregister_request(struct ceph_osd_client *osdc,
				 struct ceph_osd_request *req)
{
1296 1297 1298 1299 1300 1301
	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
1302 1303
	dout("__unregister_request %p tid %lld\n", req, req->r_tid);
	rb_erase(&req->r_node, &osdc->requests);
1304
	RB_CLEAR_NODE(&req->r_node);
S
Sage Weil 已提交
1305 1306
	osdc->num_requests--;

1307 1308
	if (req->r_osd) {
		/* make sure the original request isn't in flight. */
1309
		ceph_msg_revoke(req->r_request);
1310 1311

		list_del_init(&req->r_osd_item);
1312
		maybe_move_osd_to_lru(osdc, req->r_osd);
1313
		if (list_empty(&req->r_linger_osd_item))
1314
			req->r_osd = NULL;
1315
	}
S
Sage Weil 已提交
1316

A
Alex Elder 已提交
1317
	list_del_init(&req->r_req_lru_item);
S
Sage Weil 已提交
1318 1319
	ceph_osdc_put_request(req);

1320 1321 1322
	if (osdc->num_requests == 0) {
		dout(" no requests, canceling timeout\n");
		__cancel_osd_timeout(osdc);
S
Sage Weil 已提交
1323 1324 1325 1326 1327 1328 1329 1330
	}
}

/*
 * Cancel a previously queued request message
 */
static void __cancel_request(struct ceph_osd_request *req)
{
1331
	if (req->r_sent && req->r_osd) {
1332
		ceph_msg_revoke(req->r_request);
S
Sage Weil 已提交
1333 1334 1335 1336
		req->r_sent = 0;
	}
}

1337 1338 1339
static void __register_linger_request(struct ceph_osd_client *osdc,
				    struct ceph_osd_request *req)
{
1340 1341 1342
	dout("%s %p tid %llu\n", __func__, req, req->r_tid);
	WARN_ON(!req->r_linger);

1343
	ceph_osdc_get_request(req);
1344
	list_add_tail(&req->r_linger_item, &osdc->req_linger);
1345
	if (req->r_osd)
1346
		list_add_tail(&req->r_linger_osd_item,
1347
			      &req->r_osd->o_linger_requests);
1348 1349 1350 1351 1352
}

static void __unregister_linger_request(struct ceph_osd_client *osdc,
					struct ceph_osd_request *req)
{
1353 1354 1355 1356 1357 1358 1359 1360 1361
	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
1362
	list_del_init(&req->r_linger_item);
1363

1364
	if (req->r_osd) {
1365
		list_del_init(&req->r_linger_osd_item);
1366
		maybe_move_osd_to_lru(osdc, req->r_osd);
1367 1368
		if (list_empty(&req->r_osd_item))
			req->r_osd = NULL;
1369
	}
1370
	ceph_osdc_put_request(req);
1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382
}

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);

1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398
/*
 * 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);
}

1399 1400 1401 1402 1403 1404 1405
/*
 * 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)
{
1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419
	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) {
1420 1421
		struct ceph_pg_pool_info *pi;

1422
		pi = ceph_pg_pool_by_id(osdmap, req->r_target_oloc.pool);
1423 1424 1425
		if (pi) {
			if ((req->r_flags & CEPH_OSD_FLAG_READ) &&
			    pi->read_tier >= 0)
1426
				req->r_target_oloc.pool = pi->read_tier;
1427 1428
			if ((req->r_flags & CEPH_OSD_FLAG_WRITE) &&
			    pi->write_tier >= 0)
1429
				req->r_target_oloc.pool = pi->write_tier;
1430 1431 1432 1433
		}
		/* !pi is caught in ceph_oloc_oid_to_pg() */
	}

1434 1435
	return ceph_oloc_oid_to_pg(osdmap, &req->r_target_oloc,
				   &req->r_target_oid, pg_out);
1436 1437
}

1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453
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
1454 1455 1456
/*
 * 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
1457
 * no up osd, set r_osd to NULL.  Move the request to the appropriate list
1458
 * (unsent, homeless) or leave on in-flight lru.
S
Sage Weil 已提交
1459 1460 1461 1462 1463
 *
 * Return 0 if unchanged, 1 if changed, or negative on error.
 *
 * Caller should hold map_sem for read and request_mutex.
 */
1464
static int __map_request(struct ceph_osd_client *osdc,
1465
			 struct ceph_osd_request *req, int force_resend)
S
Sage Weil 已提交
1466
{
1467
	struct ceph_pg pgid;
1468
	int acting[CEPH_PG_MAX_SIZE];
1469
	int num, o;
S
Sage Weil 已提交
1470
	int err;
1471
	bool was_paused;
S
Sage Weil 已提交
1472

1473
	dout("map_request %p tid %lld\n", req, req->r_tid);
1474 1475

	err = __calc_request_pg(osdc->osdmap, req, &pgid);
1476 1477
	if (err) {
		list_move(&req->r_req_lru_item, &osdc->req_notarget);
S
Sage Weil 已提交
1478
		return err;
1479
	}
1480 1481
	req->r_pgid = pgid;

1482 1483 1484
	num = ceph_calc_pg_acting(osdc->osdmap, pgid, acting, &o);
	if (num < 0)
		num = 0;
S
Sage Weil 已提交
1485

1486 1487 1488 1489 1490
	was_paused = req->r_paused;
	req->r_paused = __req_should_be_paused(osdc, req);
	if (was_paused && !req->r_paused)
		force_resend = 1;

1491 1492
	if ((!force_resend &&
	     req->r_osd && req->r_osd->o_osd == o &&
1493 1494 1495
	     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) ||
1496 1497
	    (req->r_osd == NULL && o == -1) ||
	    req->r_paused)
S
Sage Weil 已提交
1498 1499
		return 0;  /* no change */

1500 1501
	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
1502 1503
	     req->r_osd ? req->r_osd->o_osd : -1);

1504 1505 1506 1507
	/* 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
1508 1509 1510
	if (req->r_osd) {
		__cancel_request(req);
		list_del_init(&req->r_osd_item);
1511
		list_del_init(&req->r_linger_osd_item);
S
Sage Weil 已提交
1512 1513 1514 1515 1516
		req->r_osd = NULL;
	}

	req->r_osd = __lookup_osd(osdc, o);
	if (!req->r_osd && o >= 0) {
1517
		err = -ENOMEM;
1518
		req->r_osd = create_osd(osdc, o);
1519 1520
		if (!req->r_osd) {
			list_move(&req->r_req_lru_item, &osdc->req_notarget);
1521
			goto out;
1522
		}
S
Sage Weil 已提交
1523

1524
		dout("map_request osd %p is osd%d\n", req->r_osd, o);
S
Sage Weil 已提交
1525 1526
		__insert_osd(osdc, req->r_osd);

1527 1528 1529
		ceph_con_open(&req->r_osd->o_con,
			      CEPH_ENTITY_TYPE_OSD, o,
			      &osdc->osdmap->osd_addr[o]);
S
Sage Weil 已提交
1530 1531
	}

1532
	__enqueue_request(req);
1533
	err = 1;   /* osd or pg changed */
S
Sage Weil 已提交
1534 1535 1536 1537 1538 1539 1540 1541

out:
	return err;
}

/*
 * caller should hold map_sem (for read) and request_mutex
 */
1542 1543
static void __send_request(struct ceph_osd_client *osdc,
			   struct ceph_osd_request *req)
S
Sage Weil 已提交
1544
{
1545
	void *p;
S
Sage Weil 已提交
1546

1547 1548 1549 1550 1551 1552 1553
	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);
1554
	put_unaligned_le64(req->r_target_oloc.pool, req->r_request_pool);
1555 1556 1557 1558 1559 1560
	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));
1561

S
Sage Weil 已提交
1562
	req->r_stamp = jiffies;
1563
	list_move_tail(&req->r_req_lru_item, &osdc->req_lru);
S
Sage Weil 已提交
1564 1565

	ceph_msg_get(req->r_request); /* send consumes a ref */
1566

S
Sage Weil 已提交
1567
	req->r_sent = req->r_osd->o_incarnation;
1568 1569

	ceph_con_send(&req->r_osd->o_con, req->r_request);
S
Sage Weil 已提交
1570 1571
}

1572 1573 1574
/*
 * Send any requests in the queue (req_unsent).
 */
A
Alex Elder 已提交
1575
static void __send_queued(struct ceph_osd_client *osdc)
1576 1577 1578
{
	struct ceph_osd_request *req, *tmp;

A
Alex Elder 已提交
1579 1580
	dout("__send_queued\n");
	list_for_each_entry_safe(req, tmp, &osdc->req_unsent, r_req_lru_item)
1581 1582 1583
		__send_request(osdc, req);
}

1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617
/*
 * 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
1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630
/*
 * 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);
1631
	struct ceph_options *opts = osdc->client->options;
S
Sage Weil 已提交
1632
	struct ceph_osd_request *req;
S
Sage Weil 已提交
1633
	struct ceph_osd *osd;
1634
	struct list_head slow_osds;
S
Sage Weil 已提交
1635 1636 1637 1638 1639 1640 1641
	dout("timeout\n");
	down_read(&osdc->map_sem);

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

	mutex_lock(&osdc->request_mutex);

1642 1643 1644 1645 1646 1647 1648
	/*
	 * 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) {
1649 1650
		if (time_before(jiffies,
				req->r_stamp + opts->osd_keepalive_timeout))
1651 1652 1653 1654 1655
			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
1656
		     req->r_tid, osd->o_osd);
1657 1658 1659 1660 1661 1662
		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
1663 1664 1665
		ceph_con_keepalive(&osd->o_con);
	}

1666
	__schedule_osd_timeout(osdc);
A
Alex Elder 已提交
1667
	__send_queued(osdc);
S
Sage Weil 已提交
1668 1669 1670 1671
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
}

1672 1673 1674 1675 1676
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);
1677
	unsigned long delay = osdc->client->options->osd_idle_ttl / 4;
1678 1679 1680

	dout("osds timeout\n");
	down_read(&osdc->map_sem);
1681
	remove_old_osds(osdc);
1682 1683 1684 1685 1686 1687
	up_read(&osdc->map_sem);

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

1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790
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
1791 1792 1793 1794 1795
static void complete_request(struct ceph_osd_request *req)
{
	complete_all(&req->r_safe_completion);  /* fsync waiter */
}

S
Sage Weil 已提交
1796 1797 1798 1799
/*
 * handle osd op reply.  either call the callback if it is specified,
 * or do the completion to wake up the waiting thread.
 */
1800
static void handle_reply(struct ceph_osd_client *osdc, struct ceph_msg *msg)
S
Sage Weil 已提交
1801
{
1802
	void *p, *end;
S
Sage Weil 已提交
1803
	struct ceph_osd_request *req;
1804
	struct ceph_request_redirect redir;
S
Sage Weil 已提交
1805
	u64 tid;
1806
	int object_len;
1807 1808
	unsigned int numops;
	int payload_len, flags;
1809
	s32 result;
1810 1811 1812 1813 1814 1815
	s32 retry_attempt;
	struct ceph_pg pg;
	int err;
	u32 reassert_epoch;
	u64 reassert_version;
	u32 osdmap_epoch;
1816
	int already_completed;
1817
	u32 bytes;
I
Ilya Dryomov 已提交
1818
	u8 decode_redir;
1819
	unsigned int i;
S
Sage Weil 已提交
1820

1821
	tid = le64_to_cpu(msg->hdr.tid);
1822 1823 1824 1825 1826 1827 1828 1829 1830 1831
	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;

1832
	err = ceph_decode_pgid(&p, end, &pg);
1833
	if (err)
S
Sage Weil 已提交
1834
		goto bad;
1835 1836 1837 1838 1839 1840 1841 1842

	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
1843
	/* lookup */
1844
	down_read(&osdc->map_sem);
S
Sage Weil 已提交
1845 1846 1847 1848
	mutex_lock(&osdc->request_mutex);
	req = __lookup_request(osdc, tid);
	if (req == NULL) {
		dout("handle_reply tid %llu dne\n", tid);
1849
		goto bad_mutex;
S
Sage Weil 已提交
1850 1851
	}
	ceph_osdc_get_request(req);
1852 1853 1854 1855

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

1856
	ceph_decode_need(&p, end, 4, bad_put);
1857
	numops = ceph_decode_32(&p);
1858
	if (numops > CEPH_OSD_MAX_OPS)
1859 1860 1861 1862
		goto bad_put;
	if (numops != req->r_num_ops)
		goto bad_put;
	payload_len = 0;
1863
	ceph_decode_need(&p, end, numops * sizeof(struct ceph_osd_op), bad_put);
1864 1865 1866 1867 1868
	for (i = 0; i < numops; i++) {
		struct ceph_osd_op *op = p;
		int len;

		len = le32_to_cpu(op->payload_len);
1869
		req->r_ops[i].outdata_len = len;
1870 1871 1872 1873
		dout(" op %d has %d bytes\n", i, len);
		payload_len += len;
		p += sizeof(*op);
	}
1874 1875
	bytes = le32_to_cpu(msg->hdr.data_len);
	if (payload_len != bytes) {
1876 1877
		pr_warn("sum of op payload lens %d != data_len %d\n",
			payload_len, bytes);
1878 1879 1880
		goto bad_put;
	}

1881
	ceph_decode_need(&p, end, 4 + numops * 4, bad_put);
1882 1883
	retry_attempt = ceph_decode_32(&p);
	for (i = 0; i < numops; i++)
1884
		req->r_ops[i].rval = ceph_decode_32(&p);
S
Sage Weil 已提交
1885

1886 1887 1888
	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
1889

I
Ilya Dryomov 已提交
1890 1891 1892 1893 1894 1895 1896 1897 1898
		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) {
1899 1900 1901 1902 1903 1904
		err = ceph_redirect_decode(&p, end, &redir);
		if (err)
			goto bad_put;
	} else {
		redir.oloc.pool = -1;
	}
S
Sage Weil 已提交
1905

1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920
	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.
		 */
1921
		err = __ceph_osdc_start_request(osdc, req, true);
1922 1923
		BUG_ON(err);

1924
		goto out_unlock;
1925 1926 1927 1928
	}

	already_completed = req->r_got_reply;
	if (!req->r_got_reply) {
1929
		req->r_result = result;
S
Sage Weil 已提交
1930 1931 1932 1933 1934 1935
		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, */
1936 1937
		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
1938 1939 1940 1941

		req->r_got_reply = 1;
	} else if ((flags & CEPH_OSD_FLAG_ONDISK) == 0) {
		dout("handle_reply tid %llu dup ack\n", tid);
1942
		goto out_unlock;
S
Sage Weil 已提交
1943 1944 1945 1946
	}

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

1947 1948 1949
	if (req->r_linger && (flags & CEPH_OSD_FLAG_ONDISK))
		__register_linger_request(osdc, req);

S
Sage Weil 已提交
1950
	/* either this is a read, or we got the safe response */
1951 1952
	if (result < 0 ||
	    (flags & CEPH_OSD_FLAG_ONDISK) ||
S
Sage Weil 已提交
1953 1954 1955 1956
	    ((flags & CEPH_OSD_FLAG_WRITE) == 0))
		__unregister_request(osdc, req);

	mutex_unlock(&osdc->request_mutex);
1957
	up_read(&osdc->map_sem);
S
Sage Weil 已提交
1958

Y
Yan, Zheng 已提交
1959
	if (!already_completed) {
1960 1961 1962
		if (req->r_unsafe_callback &&
		    result >= 0 && !(flags & CEPH_OSD_FLAG_ONDISK))
			req->r_unsafe_callback(req, true);
Y
Yan, Zheng 已提交
1963 1964 1965 1966 1967
		if (req->r_callback)
			req->r_callback(req, msg);
		else
			complete_all(&req->r_completion);
	}
S
Sage Weil 已提交
1968

1969 1970 1971
	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
1972
		complete_request(req);
1973
	}
S
Sage Weil 已提交
1974

1975
out:
1976
	dout("req=%p req->r_linger=%d\n", req, req->r_linger);
S
Sage Weil 已提交
1977 1978
	ceph_osdc_put_request(req);
	return;
1979 1980 1981 1982
out_unlock:
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
	goto out;
S
Sage Weil 已提交
1983

1984
bad_put:
1985 1986 1987 1988 1989 1990 1991
	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);
1992
	ceph_osdc_put_request(req);
1993 1994
bad_mutex:
	mutex_unlock(&osdc->request_mutex);
1995
	up_read(&osdc->map_sem);
S
Sage Weil 已提交
1996
bad:
1997 1998
	pr_err("corrupt osd_op_reply got %d %d\n",
	       (int)msg->front.iov_len, le32_to_cpu(msg->hdr.front_len));
1999
	ceph_msg_dump(msg);
S
Sage Weil 已提交
2000 2001
}

2002
static void reset_changed_osds(struct ceph_osd_client *osdc)
S
Sage Weil 已提交
2003 2004 2005
{
	struct rb_node *p, *n;

2006
	dout("%s %p\n", __func__, osdc);
2007 2008
	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
2009

2010 2011 2012 2013 2014 2015 2016
		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
2017
	}
2018 2019 2020
}

/*
2021 2022
 * Requeue requests whose mapping to an OSD has changed.  If requests map to
 * no osd, request a new map.
2023
 *
2024
 * Caller should hold map_sem for read.
2025
 */
2026 2027
static void kick_requests(struct ceph_osd_client *osdc, bool force_resend,
			  bool force_resend_writes)
2028
{
2029
	struct ceph_osd_request *req, *nreq;
2030 2031 2032
	struct rb_node *p;
	int needmap = 0;
	int err;
2033
	bool force_resend_req;
2034

2035 2036
	dout("kick_requests %s %s\n", force_resend ? " (force resend)" : "",
		force_resend_writes ? " (force resend writes)" : "");
2037
	mutex_lock(&osdc->request_mutex);
2038
	for (p = rb_first(&osdc->requests); p; ) {
2039
		req = rb_entry(p, struct ceph_osd_request, r_node);
2040
		p = rb_next(p);
2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053

		/*
		 * 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);
2054
			ceph_osdc_get_request(req);
2055 2056
			__unregister_request(osdc, req);
			__register_linger_request(osdc, req);
2057
			ceph_osdc_put_request(req);
2058 2059 2060
			continue;
		}

2061 2062 2063 2064
		force_resend_req = force_resend ||
			(force_resend_writes &&
				req->r_flags & CEPH_OSD_FLAG_WRITE);
		err = __map_request(osdc, req, force_resend_req);
2065 2066 2067 2068 2069 2070
		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) {
2071 2072 2073 2074
			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);
2075
				req->r_flags |= CEPH_OSD_FLAG_RETRY;
2076 2077
			}
		}
2078 2079 2080 2081 2082 2083
	}

	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);

2084 2085
		err = __map_request(osdc, req,
				    force_resend || force_resend_writes);
2086
		dout("__map_request returned %d\n", err);
2087 2088
		if (err < 0)
			continue;  /* hrm! */
2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103
		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++;
			}
2104

2105 2106 2107 2108 2109
			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);
		}
2110
	}
2111
	reset_changed_osds(osdc);
S
Sage Weil 已提交
2112 2113 2114 2115 2116 2117
	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);
	}
2118
}
2119 2120


S
Sage Weil 已提交
2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135
/*
 * 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;
2136
	bool was_full;
S
Sage Weil 已提交
2137 2138 2139 2140 2141 2142 2143 2144

	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));
2145 2146
	if (ceph_check_fsid(osdc->client, &fsid) < 0)
		return;
S
Sage Weil 已提交
2147 2148 2149

	down_write(&osdc->map_sem);

2150 2151
	was_full = ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_FULL);

S
Sage Weil 已提交
2152 2153 2154 2155 2156
	/* 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);
2157 2158
		epoch = ceph_decode_32(&p);
		maplen = ceph_decode_32(&p);
S
Sage Weil 已提交
2159 2160 2161 2162 2163 2164 2165
		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,
2166
							  &osdc->client->msgr);
S
Sage Weil 已提交
2167 2168 2169 2170
			if (IS_ERR(newmap)) {
				err = PTR_ERR(newmap);
				goto bad;
			}
2171
			BUG_ON(!newmap);
S
Sage Weil 已提交
2172 2173 2174 2175
			if (newmap != osdc->osdmap) {
				ceph_osdmap_destroy(osdc->osdmap);
				osdc->osdmap = newmap;
			}
2176 2177 2178 2179
			was_full = was_full ||
				ceph_osdmap_flag(osdc->osdmap,
						 CEPH_OSDMAP_FULL);
			kick_requests(osdc, 0, was_full);
S
Sage Weil 已提交
2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194
		} 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);
2195 2196
		epoch = ceph_decode_32(&p);
		maplen = ceph_decode_32(&p);
S
Sage Weil 已提交
2197 2198 2199 2200 2201 2202 2203 2204 2205
		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 {
2206 2207
			int skipped_map = 0;

S
Sage Weil 已提交
2208
			dout("taking full map %u len %d\n", epoch, maplen);
2209
			newmap = ceph_osdmap_decode(&p, p+maplen);
S
Sage Weil 已提交
2210 2211 2212 2213
			if (IS_ERR(newmap)) {
				err = PTR_ERR(newmap);
				goto bad;
			}
2214
			BUG_ON(!newmap);
S
Sage Weil 已提交
2215 2216
			oldmap = osdc->osdmap;
			osdc->osdmap = newmap;
2217 2218 2219
			if (oldmap) {
				if (oldmap->epoch + 1 < newmap->epoch)
					skipped_map = 1;
S
Sage Weil 已提交
2220
				ceph_osdmap_destroy(oldmap);
2221
			}
2222 2223 2224 2225
			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
2226 2227 2228 2229 2230
		}
		p += maplen;
		nr_maps--;
	}

2231 2232
	if (!osdc->osdmap)
		goto bad;
S
Sage Weil 已提交
2233 2234
done:
	downgrade_write(&osdc->map_sem);
2235 2236
	ceph_monc_got_map(&osdc->client->monc, CEPH_SUB_OSDMAP,
			  osdc->osdmap->epoch);
2237 2238 2239 2240 2241 2242

	/*
	 * subscribe to subsequent osdmap updates if full to ensure
	 * we find out when we are no longer full and stop returning
	 * ENOSPC.
	 */
2243 2244 2245
	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))
2246 2247
		ceph_monc_request_next_osdmap(&osdc->client->monc);

A
Alex Elder 已提交
2248 2249 2250
	mutex_lock(&osdc->request_mutex);
	__send_queued(osdc);
	mutex_unlock(&osdc->request_mutex);
S
Sage Weil 已提交
2251
	up_read(&osdc->map_sem);
2252
	wake_up_all(&osdc->client->auth_wq);
S
Sage Weil 已提交
2253 2254 2255 2256
	return;

bad:
	pr_err("osdc handle_map corrupt msg\n");
2257
	ceph_msg_dump(msg);
S
Sage Weil 已提交
2258 2259 2260
	up_write(&osdc->map_sem);
}

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 2335 2336 2337 2338 2339 2340 2341 2342
/*
 * 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 *),
2343
			   void *data, struct ceph_osd_event **pevent)
2344 2345 2346 2347 2348 2349 2350 2351 2352
{
	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;
2353
	event->one_shot = 0;
2354 2355 2356
	event->data = data;
	event->osdc = osdc;
	INIT_LIST_HEAD(&event->osd_node);
2357
	RB_CLEAR_NODE(&event->node);
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 2396 2397 2398 2399 2400 2401 2402 2403
	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
 */
2404 2405
static void handle_watch_notify(struct ceph_osd_client *osdc,
				struct ceph_msg *msg)
2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425
{
	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) {
2426
		BUG_ON(event->one_shot);
2427 2428 2429 2430 2431 2432 2433 2434
		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) {
2435 2436 2437
			pr_err("couldn't allocate event_work\n");
			ceph_osdc_put_event(event);
			return;
2438
		}
2439
		INIT_WORK(&event_work->work, do_event_work);
2440 2441 2442 2443 2444
		event_work->event = event;
		event_work->ver = ver;
		event_work->notify_id = notify_id;
		event_work->opcode = opcode;

2445 2446
		queue_work(osdc->notify_wq, &event_work->work);
	}
2447 2448 2449 2450 2451 2452 2453

	return;

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

2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469
/*
 * 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;
2470
	WARN_ON(snapc != req->r_snapc);
2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501

	/* 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 */
2502 2503 2504 2505 2506
	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;
2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520

	/* 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) {
2521
		for (i = 0; i < req->r_snapc->num_snaps; i++) {
2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552
			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);

2553 2554 2555 2556 2557 2558 2559
/*
 * Register request, send initial attempt.
 */
int ceph_osdc_start_request(struct ceph_osd_client *osdc,
			    struct ceph_osd_request *req,
			    bool nofail)
{
2560
	int rc;
2561

S
Sage Weil 已提交
2562 2563
	down_read(&osdc->map_sem);
	mutex_lock(&osdc->request_mutex);
2564 2565 2566

	rc = __ceph_osdc_start_request(osdc, req, nofail);

S
Sage Weil 已提交
2567 2568
	mutex_unlock(&osdc->request_mutex);
	up_read(&osdc->map_sem);
2569

S
Sage Weil 已提交
2570 2571
	return rc;
}
2572
EXPORT_SYMBOL(ceph_osdc_start_request);
S
Sage Weil 已提交
2573

2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592
/*
 * 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
2593 2594 2595 2596 2597 2598 2599 2600
/*
 * wait for a request to complete
 */
int ceph_osdc_wait_request(struct ceph_osd_client *osdc,
			   struct ceph_osd_request *req)
{
	int rc;

2601 2602
	dout("%s %p tid %llu\n", __func__, req, req->r_tid);

S
Sage Weil 已提交
2603 2604
	rc = wait_for_completion_interruptible(&req->r_completion);
	if (rc < 0) {
2605 2606
		dout("%s %p tid %llu interrupted\n", __func__, req, req->r_tid);
		ceph_osdc_cancel_request(req);
S
Sage Weil 已提交
2607
		complete_request(req);
S
Sage Weil 已提交
2608 2609 2610
		return rc;
	}

2611 2612
	dout("%s %p tid %llu result %d\n", __func__, req, req->r_tid,
	     req->r_result);
S
Sage Weil 已提交
2613 2614
	return req->r_result;
}
2615
EXPORT_SYMBOL(ceph_osdc_wait_request);
S
Sage Weil 已提交
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 2641 2642 2643 2644 2645 2646 2647 2648

/*
 * 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);
}
2649
EXPORT_SYMBOL(ceph_osdc_sync);
S
Sage Weil 已提交
2650

2651 2652 2653 2654
/*
 * 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
2655
void ceph_osdc_flush_notifies(struct ceph_osd_client *osdc)
2656 2657 2658 2659 2660 2661
{
	flush_workqueue(osdc->notify_wq);
}
EXPORT_SYMBOL(ceph_osdc_flush_notifies);


S
Sage Weil 已提交
2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677
/*
 * 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;
2678
	INIT_LIST_HEAD(&osdc->osd_lru);
S
Sage Weil 已提交
2679
	osdc->requests = RB_ROOT;
2680
	INIT_LIST_HEAD(&osdc->req_lru);
2681 2682
	INIT_LIST_HEAD(&osdc->req_unsent);
	INIT_LIST_HEAD(&osdc->req_notarget);
2683
	INIT_LIST_HEAD(&osdc->req_linger);
S
Sage Weil 已提交
2684 2685
	osdc->num_requests = 0;
	INIT_DELAYED_WORK(&osdc->timeout_work, handle_timeout);
2686
	INIT_DELAYED_WORK(&osdc->osds_timeout_work, handle_osds_timeout);
2687 2688 2689
	spin_lock_init(&osdc->event_lock);
	osdc->event_tree = RB_ROOT;
	osdc->event_count = 0;
2690 2691

	schedule_delayed_work(&osdc->osds_timeout_work,
2692
	    round_jiffies_relative(osdc->client->options->osd_idle_ttl));
S
Sage Weil 已提交
2693

2694
	err = -ENOMEM;
2695 2696
	osdc->req_mempool = mempool_create_slab_pool(10,
						     ceph_osd_request_cache);
S
Sage Weil 已提交
2697
	if (!osdc->req_mempool)
2698
		goto out;
S
Sage Weil 已提交
2699

2700
	err = ceph_msgpool_init(&osdc->msgpool_op, CEPH_MSG_OSD_OP,
2701
				PAGE_SIZE, 10, true, "osd_op");
S
Sage Weil 已提交
2702
	if (err < 0)
2703
		goto out_mempool;
2704
	err = ceph_msgpool_init(&osdc->msgpool_op_reply, CEPH_MSG_OSD_OPREPLY,
2705
				PAGE_SIZE, 10, true, "osd_op_reply");
S
Sage Weil 已提交
2706 2707
	if (err < 0)
		goto out_msgpool;
2708

2709
	err = -ENOMEM;
2710
	osdc->notify_wq = create_singlethread_workqueue("ceph-watch-notify");
2711
	if (!osdc->notify_wq)
2712 2713
		goto out_msgpool_reply;

S
Sage Weil 已提交
2714
	return 0;
2715

2716 2717
out_msgpool_reply:
	ceph_msgpool_destroy(&osdc->msgpool_op_reply);
S
Sage Weil 已提交
2718 2719
out_msgpool:
	ceph_msgpool_destroy(&osdc->msgpool_op);
2720 2721 2722 2723
out_mempool:
	mempool_destroy(osdc->req_mempool);
out:
	return err;
S
Sage Weil 已提交
2724 2725 2726 2727
}

void ceph_osdc_stop(struct ceph_osd_client *osdc)
{
2728 2729
	flush_workqueue(osdc->notify_wq);
	destroy_workqueue(osdc->notify_wq);
S
Sage Weil 已提交
2730
	cancel_delayed_work_sync(&osdc->timeout_work);
2731
	cancel_delayed_work_sync(&osdc->osds_timeout_work);
S
Sage Weil 已提交
2732 2733 2734 2735
	if (osdc->osdmap) {
		ceph_osdmap_destroy(osdc->osdmap);
		osdc->osdmap = NULL;
	}
2736
	remove_all_osds(osdc);
S
Sage Weil 已提交
2737 2738
	mempool_destroy(osdc->req_mempool);
	ceph_msgpool_destroy(&osdc->msgpool_op);
S
Sage Weil 已提交
2739
	ceph_msgpool_destroy(&osdc->msgpool_op_reply);
S
Sage Weil 已提交
2740 2741 2742 2743 2744 2745 2746 2747 2748 2749
}

/*
 * 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,
2750
			struct page **pages, int num_pages, int page_align)
S
Sage Weil 已提交
2751 2752 2753 2754 2755 2756
{
	struct ceph_osd_request *req;
	int rc = 0;

	dout("readpages on ino %llx.%llx on %llu~%llu\n", vino.ino,
	     vino.snap, off, *plen);
2757
	req = ceph_osdc_new_request(osdc, layout, vino, off, plen, 0, 1,
S
Sage Weil 已提交
2758
				    CEPH_OSD_OP_READ, CEPH_OSD_FLAG_READ,
2759
				    NULL, truncate_seq, truncate_size,
2760
				    false);
2761 2762
	if (IS_ERR(req))
		return PTR_ERR(req);
S
Sage Weil 已提交
2763 2764

	/* it may be a short read due to an object boundary */
2765

2766
	osd_req_op_extent_osd_data_pages(req, 0,
2767
				pages, *plen, page_align, false, false);
S
Sage Weil 已提交
2768

2769
	dout("readpages  final extent is %llu~%llu (%llu bytes align %d)\n",
2770
	     off, *plen, *plen, page_align);
S
Sage Weil 已提交
2771

2772
	ceph_osdc_build_request(req, off, NULL, vino.snap, NULL);
2773

S
Sage Weil 已提交
2774 2775 2776 2777 2778 2779 2780 2781
	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;
}
2782
EXPORT_SYMBOL(ceph_osdc_readpages);
S
Sage Weil 已提交
2783 2784 2785 2786 2787 2788 2789 2790 2791 2792

/*
 * 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,
2793
			 struct page **pages, int num_pages)
S
Sage Weil 已提交
2794 2795 2796
{
	struct ceph_osd_request *req;
	int rc = 0;
2797
	int page_align = off & ~PAGE_MASK;
S
Sage Weil 已提交
2798

2799
	BUG_ON(vino.snap != CEPH_NOSNAP);	/* snapshots aren't writeable */
2800
	req = ceph_osdc_new_request(osdc, layout, vino, off, &len, 0, 1,
S
Sage Weil 已提交
2801
				    CEPH_OSD_OP_WRITE,
2802
				    CEPH_OSD_FLAG_ONDISK | CEPH_OSD_FLAG_WRITE,
2803
				    snapc, truncate_seq, truncate_size,
2804
				    true);
2805 2806
	if (IS_ERR(req))
		return PTR_ERR(req);
S
Sage Weil 已提交
2807 2808

	/* it may be a short write due to an object boundary */
2809
	osd_req_op_extent_osd_data_pages(req, 0, pages, len, page_align,
2810 2811
				false, false);
	dout("writepages %llu~%llu (%llu bytes)\n", off, len, len);
S
Sage Weil 已提交
2812

2813
	ceph_osdc_build_request(req, off, snapc, CEPH_NOSNAP, mtime);
2814

2815
	rc = ceph_osdc_start_request(osdc, req, true);
S
Sage Weil 已提交
2816 2817 2818 2819 2820 2821 2822 2823 2824
	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;
}
2825
EXPORT_SYMBOL(ceph_osdc_writepages);
S
Sage Weil 已提交
2826

2827 2828
int ceph_osdc_setup(void)
{
2829 2830 2831
	size_t size = sizeof(struct ceph_osd_request) +
	    CEPH_OSD_SLAB_OPS * sizeof(struct ceph_osd_req_op);

2832
	BUG_ON(ceph_osd_request_cache);
2833 2834
	ceph_osd_request_cache = kmem_cache_create("ceph_osd_request", size,
						   0, 0, NULL);
2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847

	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
2848 2849 2850 2851 2852 2853
/*
 * handle incoming message
 */
static void dispatch(struct ceph_connection *con, struct ceph_msg *msg)
{
	struct ceph_osd *osd = con->private;
2854
	struct ceph_osd_client *osdc;
S
Sage Weil 已提交
2855 2856 2857
	int type = le16_to_cpu(msg->hdr.type);

	if (!osd)
S
Sage Weil 已提交
2858
		goto out;
2859
	osdc = osd->o_osdc;
S
Sage Weil 已提交
2860 2861 2862 2863 2864 2865

	switch (type) {
	case CEPH_MSG_OSD_MAP:
		ceph_osdc_handle_map(osdc, msg);
		break;
	case CEPH_MSG_OSD_OPREPLY:
2866
		handle_reply(osdc, msg);
S
Sage Weil 已提交
2867
		break;
2868 2869 2870
	case CEPH_MSG_WATCH_NOTIFY:
		handle_watch_notify(osdc, msg);
		break;
S
Sage Weil 已提交
2871 2872 2873 2874 2875

	default:
		pr_err("received unknown message type %d %s\n", type,
		       ceph_msg_type_name(type));
	}
S
Sage Weil 已提交
2876
out:
S
Sage Weil 已提交
2877 2878 2879
	ceph_msg_put(msg);
}

2880
/*
2881 2882 2883
 * 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.
2884 2885
 */
static struct ceph_msg *get_reply(struct ceph_connection *con,
2886 2887
				  struct ceph_msg_header *hdr,
				  int *skip)
S
Sage Weil 已提交
2888 2889 2890
{
	struct ceph_osd *osd = con->private;
	struct ceph_osd_client *osdc = osd->o_osdc;
2891
	struct ceph_msg *m;
2892
	struct ceph_osd_request *req;
2893
	int front_len = le32_to_cpu(hdr->front_len);
2894
	int data_len = le32_to_cpu(hdr->data_len);
2895
	u64 tid;
S
Sage Weil 已提交
2896

2897 2898 2899 2900
	tid = le64_to_cpu(hdr->tid);
	mutex_lock(&osdc->request_mutex);
	req = __lookup_request(osdc, tid);
	if (!req) {
2901 2902
		dout("%s osd%d tid %llu unknown, skipping\n", __func__,
		     osd->o_osd, tid);
2903
		m = NULL;
2904
		*skip = 1;
2905 2906
		goto out;
	}
S
Sage Weil 已提交
2907

2908
	ceph_msg_revoke_incoming(req->r_reply);
2909

2910
	if (front_len > req->r_reply->front_alloc_len) {
2911 2912 2913
		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);
2914 2915
		m = ceph_msg_new(CEPH_MSG_OSD_OPREPLY, front_len, GFP_NOFS,
				 false);
2916
		if (!m)
S
Sage Weil 已提交
2917 2918 2919 2920
			goto out;
		ceph_msg_put(req->r_reply);
		req->r_reply = m;
	}
2921

2922 2923 2924 2925 2926 2927 2928
	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;
2929
	}
2930 2931

	m = ceph_msg_get(req->r_reply);
S
Sage Weil 已提交
2932
	dout("get_reply tid %lld %p\n", tid, m);
2933 2934 2935

out:
	mutex_unlock(&osdc->request_mutex);
2936
	return m;
2937 2938 2939 2940 2941 2942 2943 2944 2945 2946
}

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
2947
	*skip = 0;
2948 2949
	switch (type) {
	case CEPH_MSG_OSD_MAP:
2950
	case CEPH_MSG_WATCH_NOTIFY:
2951
		return ceph_msg_new(type, front, GFP_NOFS, false);
2952 2953 2954 2955 2956 2957 2958 2959
	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
2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978
}

/*
 * 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);
}

2979 2980 2981
/*
 * authentication
 */
2982 2983 2984 2985 2986
/*
 * 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,
2987
					int *proto, int force_new)
2988 2989 2990 2991
{
	struct ceph_osd *o = con->private;
	struct ceph_osd_client *osdc = o->o_osdc;
	struct ceph_auth_client *ac = osdc->client->monc.auth;
2992
	struct ceph_auth_handshake *auth = &o->o_auth;
2993

2994
	if (force_new && auth->authorizer) {
2995
		ceph_auth_destroy_authorizer(auth->authorizer);
2996 2997
		auth->authorizer = NULL;
	}
2998 2999 3000
	if (!auth->authorizer) {
		int ret = ceph_auth_create_authorizer(ac, CEPH_ENTITY_TYPE_OSD,
						      auth);
3001
		if (ret)
3002
			return ERR_PTR(ret);
3003 3004
	} else {
		int ret = ceph_auth_update_authorizer(ac, CEPH_ENTITY_TYPE_OSD,
3005 3006 3007
						     auth);
		if (ret)
			return ERR_PTR(ret);
3008 3009
	}
	*proto = ac->protocol;
3010

3011
	return auth;
3012 3013 3014 3015 3016 3017 3018 3019 3020
}


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;

3021
	return ceph_auth_verify_authorizer_reply(ac, o->o_auth.authorizer, len);
3022 3023
}

3024 3025 3026 3027 3028 3029
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;

3030
	ceph_auth_invalidate_authorizer(ac, CEPH_ENTITY_TYPE_OSD);
3031 3032
	return ceph_monc_validate_auth(&osdc->client->monc);
}
3033

3034
static int osd_sign_message(struct ceph_msg *msg)
Y
Yan, Zheng 已提交
3035
{
3036
	struct ceph_osd *o = msg->con->private;
Y
Yan, Zheng 已提交
3037
	struct ceph_auth_handshake *auth = &o->o_auth;
3038

Y
Yan, Zheng 已提交
3039 3040 3041
	return ceph_auth_sign_message(auth, msg);
}

3042
static int osd_check_message_signature(struct ceph_msg *msg)
Y
Yan, Zheng 已提交
3043
{
3044
	struct ceph_osd *o = msg->con->private;
Y
Yan, Zheng 已提交
3045
	struct ceph_auth_handshake *auth = &o->o_auth;
3046

Y
Yan, Zheng 已提交
3047 3048 3049
	return ceph_auth_check_message_signature(auth, msg);
}

3050
static const struct ceph_connection_operations osd_con_ops = {
S
Sage Weil 已提交
3051 3052 3053
	.get = get_osd_con,
	.put = put_osd_con,
	.dispatch = dispatch,
3054 3055
	.get_authorizer = get_authorizer,
	.verify_authorizer_reply = verify_authorizer_reply,
3056
	.invalidate_authorizer = invalidate_authorizer,
S
Sage Weil 已提交
3057
	.alloc_msg = alloc_msg,
3058 3059
	.sign_message = osd_sign_message,
	.check_message_signature = osd_check_message_signature,
3060
	.fault = osd_reset,
S
Sage Weil 已提交
3061
};