1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
|
/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */
/*
* Copyright (c) 2014-2017 Oracle. All rights reserved.
* Copyright (c) 2003-2007 Network Appliance, Inc. All rights reserved.
*
* This software is available to you under a choice of one of two
* licenses. You may choose to be licensed under the terms of the GNU
* General Public License (GPL) Version 2, available from the file
* COPYING in the main directory of this source tree, or the BSD-type
* license below:
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
*
* Redistributions in binary form must reproduce the above
* copyright notice, this list of conditions and the following
* disclaimer in the documentation and/or other materials provided
* with the distribution.
*
* Neither the name of the Network Appliance, Inc. nor the names of
* its contributors may be used to endorse or promote products
* derived from this software without specific prior written
* permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#ifndef _LINUX_SUNRPC_XPRT_RDMA_H
#define _LINUX_SUNRPC_XPRT_RDMA_H
#include <linux/wait.h> /* wait_queue_head_t, etc */
#include <linux/spinlock.h> /* spinlock_t, etc */
#include <linux/atomic.h> /* atomic_t, etc */
#include <linux/workqueue.h> /* struct work_struct */
#include <rdma/rdma_cm.h> /* RDMA connection api */
#include <rdma/ib_verbs.h> /* RDMA verbs api */
#include <linux/sunrpc/clnt.h> /* rpc_xprt */
#include <linux/sunrpc/rpc_rdma.h> /* RPC/RDMA protocol */
#include <linux/sunrpc/xprtrdma.h> /* xprt parameters */
#define RDMA_RESOLVE_TIMEOUT (5000) /* 5 seconds */
#define RDMA_CONNECT_RETRY_MAX (2) /* retries if no listener backlog */
#define RPCRDMA_BIND_TO (60U * HZ)
#define RPCRDMA_INIT_REEST_TO (5U * HZ)
#define RPCRDMA_MAX_REEST_TO (30U * HZ)
#define RPCRDMA_IDLE_DISC_TO (5U * 60 * HZ)
/*
* Interface Adapter -- one per transport instance
*/
struct rpcrdma_ia {
struct ib_device *ri_device;
struct rdma_cm_id *ri_id;
struct ib_pd *ri_pd;
struct completion ri_done;
struct completion ri_remove_done;
int ri_async_rc;
unsigned int ri_max_segs;
unsigned int ri_max_frwr_depth;
unsigned int ri_max_inline_write;
unsigned int ri_max_inline_read;
unsigned int ri_max_send_sges;
bool ri_implicit_roundup;
enum ib_mr_type ri_mrtype;
unsigned long ri_flags;
};
enum {
RPCRDMA_IAF_REMOVING = 0,
};
/*
* RDMA Endpoint -- one per transport instance
*/
struct rpcrdma_ep {
unsigned int rep_send_count;
unsigned int rep_send_batch;
int rep_connected;
struct ib_qp_init_attr rep_attr;
wait_queue_head_t rep_connect_wait;
struct rpcrdma_connect_private rep_cm_private;
struct rdma_conn_param rep_remote_cma;
int rep_receive_count;
};
/* Pre-allocate extra Work Requests for handling backward receives
* and sends. This is a fixed value because the Work Queues are
* allocated when the forward channel is set up.
*/
#if defined(CONFIG_SUNRPC_BACKCHANNEL)
#define RPCRDMA_BACKWARD_WRS (8)
#else
#define RPCRDMA_BACKWARD_WRS (0)
#endif
/* Registered buffer -- registered kmalloc'd memory for RDMA SEND/RECV
*
* The below structure appears at the front of a large region of kmalloc'd
* memory, which always starts on a good alignment boundary.
*/
struct rpcrdma_regbuf {
struct ib_sge rg_iov;
struct ib_device *rg_device;
enum dma_data_direction rg_direction;
void *rg_data;
};
static inline u64 rdmab_addr(struct rpcrdma_regbuf *rb)
{
return rb->rg_iov.addr;
}
static inline u32 rdmab_length(struct rpcrdma_regbuf *rb)
{
return rb->rg_iov.length;
}
static inline u32 rdmab_lkey(struct rpcrdma_regbuf *rb)
{
return rb->rg_iov.lkey;
}
static inline struct ib_device *rdmab_device(struct rpcrdma_regbuf *rb)
{
return rb->rg_device;
}
static inline void *rdmab_data(const struct rpcrdma_regbuf *rb)
{
return rb->rg_data;
}
#define RPCRDMA_DEF_GFP (GFP_NOIO | __GFP_NOWARN)
/* To ensure a transport can always make forward progress,
* the number of RDMA segments allowed in header chunk lists
* is capped at 8. This prevents less-capable devices and
* memory registrations from overrunning the Send buffer
* while building chunk lists.
*
* Elements of the Read list take up more room than the
* Write list or Reply chunk. 8 read segments means the Read
* list (or Write list or Reply chunk) cannot consume more
* than
*
* ((8 + 2) * read segment size) + 1 XDR words, or 244 bytes.
*
* And the fixed part of the header is another 24 bytes.
*
* The smallest inline threshold is 1024 bytes, ensuring that
* at least 750 bytes are available for RPC messages.
*/
enum {
RPCRDMA_MAX_HDR_SEGS = 8,
RPCRDMA_HDRBUF_SIZE = 256,
};
/*
* struct rpcrdma_rep -- this structure encapsulates state required
* to receive and complete an RPC Reply, asychronously. It needs
* several pieces of state:
*
* o receive buffer and ib_sge (donated to provider)
* o status of receive (success or not, length, inv rkey)
* o bookkeeping state to get run by reply handler (XDR stream)
*
* These structures are allocated during transport initialization.
* N of these are associated with a transport instance, managed by
* struct rpcrdma_buffer. N is the max number of outstanding RPCs.
*/
struct rpcrdma_rep {
struct ib_cqe rr_cqe;
__be32 rr_xid;
__be32 rr_vers;
__be32 rr_proc;
int rr_wc_flags;
u32 rr_inv_rkey;
bool rr_temp;
struct rpcrdma_regbuf *rr_rdmabuf;
struct rpcrdma_xprt *rr_rxprt;
struct work_struct rr_work;
struct xdr_buf rr_hdrbuf;
struct xdr_stream rr_stream;
struct rpc_rqst *rr_rqst;
struct list_head rr_list;
struct ib_recv_wr rr_recv_wr;
};
/* To reduce the rate at which a transport invokes ib_post_recv
* (and thus the hardware doorbell rate), xprtrdma posts Receive
* WRs in batches.
*
* Setting this to zero disables Receive post batching.
*/
enum {
RPCRDMA_MAX_RECV_BATCH = 7,
};
/* struct rpcrdma_sendctx - DMA mapped SGEs to unmap after Send completes
*/
struct rpcrdma_req;
struct rpcrdma_xprt;
struct rpcrdma_sendctx {
struct ib_send_wr sc_wr;
struct ib_cqe sc_cqe;
struct rpcrdma_xprt *sc_xprt;
struct rpcrdma_req *sc_req;
unsigned int sc_unmap_count;
struct ib_sge sc_sges[];
};
/* Limit the number of SGEs that can be unmapped during one
* Send completion. This caps the amount of work a single
* completion can do before returning to the provider.
*
* Setting this to zero disables Send completion batching.
*/
enum {
RPCRDMA_MAX_SEND_BATCH = 7,
};
/*
* struct rpcrdma_mr - external memory region metadata
*
* An external memory region is any buffer or page that is registered
* on the fly (ie, not pre-registered).
*
* Each rpcrdma_buffer has a list of free MWs anchored in rb_mrs. During
* call_allocate, rpcrdma_buffer_get() assigns one to each segment in
* an rpcrdma_req. Then rpcrdma_register_external() grabs these to keep
* track of registration metadata while each RPC is pending.
* rpcrdma_deregister_external() uses this metadata to unmap and
* release these resources when an RPC is complete.
*/
enum rpcrdma_frwr_state {
FRWR_IS_INVALID, /* ready to be used */
FRWR_IS_VALID, /* in use */
FRWR_FLUSHED_FR, /* flushed FASTREG WR */
FRWR_FLUSHED_LI, /* flushed LOCALINV WR */
};
struct rpcrdma_frwr {
struct ib_mr *fr_mr;
struct ib_cqe fr_cqe;
enum rpcrdma_frwr_state fr_state;
struct completion fr_linv_done;
union {
struct ib_reg_wr fr_regwr;
struct ib_send_wr fr_invwr;
};
};
struct rpcrdma_mr {
struct list_head mr_list;
struct scatterlist *mr_sg;
int mr_nents;
enum dma_data_direction mr_dir;
struct rpcrdma_frwr frwr;
struct rpcrdma_xprt *mr_xprt;
u32 mr_handle;
u32 mr_length;
u64 mr_offset;
struct work_struct mr_recycle;
struct list_head mr_all;
};
/*
* struct rpcrdma_req -- structure central to the request/reply sequence.
*
* N of these are associated with a transport instance, and stored in
* struct rpcrdma_buffer. N is the max number of outstanding requests.
*
* It includes pre-registered buffer memory for send AND recv.
* The recv buffer, however, is not owned by this structure, and
* is "donated" to the hardware when a recv is posted. When a
* reply is handled, the recv buffer used is given back to the
* struct rpcrdma_req associated with the request.
*
* In addition to the basic memory, this structure includes an array
* of iovs for send operations. The reason is that the iovs passed to
* ib_post_{send,recv} must not be modified until the work request
* completes.
*/
/* Maximum number of page-sized "segments" per chunk list to be
* registered or invalidated. Must handle a Reply chunk:
*/
enum {
RPCRDMA_MAX_IOV_SEGS = 3,
RPCRDMA_MAX_DATA_SEGS = ((1 * 1024 * 1024) / PAGE_SIZE) + 1,
RPCRDMA_MAX_SEGS = RPCRDMA_MAX_DATA_SEGS +
RPCRDMA_MAX_IOV_SEGS,
};
struct rpcrdma_mr_seg { /* chunk descriptors */
u32 mr_len; /* length of chunk or segment */
struct page *mr_page; /* owning page, if any */
char *mr_offset; /* kva if no page, else offset */
};
/* The Send SGE array is provisioned to send a maximum size
* inline request:
* - RPC-over-RDMA header
* - xdr_buf head iovec
* - RPCRDMA_MAX_INLINE bytes, in pages
* - xdr_buf tail iovec
*
* The actual number of array elements consumed by each RPC
* depends on the device's max_sge limit.
*/
enum {
RPCRDMA_MIN_SEND_SGES = 3,
RPCRDMA_MAX_PAGE_SGES = RPCRDMA_MAX_INLINE >> PAGE_SHIFT,
RPCRDMA_MAX_SEND_SGES = 1 + 1 + RPCRDMA_MAX_PAGE_SGES + 1,
};
struct rpcrdma_buffer;
struct
|