summaryrefslogtreecommitdiff
path: root/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
blob: 9e884383134f6536f6ba5a81e916c5999386ca98 (plain)
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
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
/*
 * Copyright (c) 2005-2006 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.
 *
 * Author: Tom Tucker <tom@opengridcomputing.com>
 */

#include <linux/sunrpc/debug.h>
#include <linux/sunrpc/rpc_rdma.h>
#include <linux/spinlock.h>
#include <asm/unaligned.h>
#include <rdma/ib_verbs.h>
#include <rdma/rdma_cm.h>
#include <linux/sunrpc/svc_rdma.h>

#define RPCDBG_FACILITY	RPCDBG_SVCXPRT

/*
 * Replace the pages in the rq_argpages array with the pages from the SGE in
 * the RDMA_RECV completion. The SGL should contain full pages up until the
 * last one.
 */
static void rdma_build_arg_xdr(struct svc_rqst *rqstp,
			       struct svc_rdma_op_ctxt *ctxt,
			       u32 byte_count)
{
	struct page *page;
	u32 bc;
	int sge_no;

	/* Swap the page in the SGE with the page in argpages */
	page = ctxt->pages[0];
	put_page(rqstp->rq_pages[0]);
	rqstp->rq_pages[0] = page;

	/* Set up the XDR head */
	rqstp->rq_arg.head[0].iov_base = page_address(page);
	rqstp->rq_arg.head[0].iov_len = min(byte_count, ctxt->sge[0].length);
	rqstp->rq_arg.len = byte_count;
	rqstp->rq_arg.buflen = byte_count;

	/* Compute bytes past head in the SGL */
	bc = byte_count - rqstp->rq_arg.head[0].iov_len;

	/* If data remains, store it in the pagelist */
	rqstp->rq_arg.page_len = bc;
	rqstp->rq_arg.page_base = 0;
	rqstp->rq_arg.pages = &rqstp->rq_pages[1];
	sge_no = 1;
	while (bc && sge_no < ctxt->count) {
		page = ctxt->pages[sge_no];
		put_page(rqstp->rq_pages[sge_no]);
		rqstp->rq_pages[sge_no] = page;
		bc -= min(bc, ctxt->sge[sge_no].length);
		rqstp->rq_arg.buflen += ctxt->sge[sge_no].length;
		sge_no++;
	}
	rqstp->rq_respages = &rqstp->rq_pages[sge_no];

	/* We should never run out of SGE because the limit is defined to
	 * support the max allowed RPC data length
	 */
	BUG_ON(bc && (sge_no == ctxt->count));
	BUG_ON((rqstp->rq_arg.head[0].iov_len + rqstp->rq_arg.page_len)
	       != byte_count);
	BUG_ON(rqstp->rq_arg.len != byte_count);

	/* If not all pages were used from the SGL, free the remaining ones */
	bc = sge_no;
	while (sge_no < ctxt->count) {
		page = ctxt->pages[sge_no++];
		put_page(page);
	}
	ctxt->count = bc;

	/* Set up tail */
	rqstp->rq_arg.tail[0].iov_base = NULL;
	rqstp->rq_arg.tail[0].iov_len = 0;
}

/* Encode a read-chunk-list as an array of IB SGE
 *
 * Assumptions:
 * - chunk[0]->position points to pages[0] at an offset of 0
 * - pages[] is not physically or virtually contiguous and consists of
 *   PAGE_SIZE elements.
 *
 * Output:
 * - sge array pointing into pages[] array.
 * - chunk_sge array specifying sge index and count for each
 *   chunk in the read list
 *
 */
static int map_read_chunks(struct svcxprt_rdma *xprt,
			   struct svc_rqst *rqstp,
			   struct svc_rdma_op_ctxt *head,
			   struct rpcrdma_msg *rmsgp,
			   struct svc_rdma_req_map *rpl_map,
			   struct svc_rdma_req_map *chl_map,
			   int ch_count,
			   int byte_count)
{
	int sge_no;
	int sge_bytes;
	int page_off;
	int page_no;
	int ch_bytes;
	int ch_no;
	struct rpcrdma_read_chunk *ch;

	sge_no = 0;
	page_no = 0;
	page_off = 0;
	ch = (struct rpcrdma_read_chunk *)&rmsgp->rm_body.rm_chunks[0];
	ch_no = 0;
	ch_bytes = ch->rc_target.rs_length;
	head->arg.head[0] = rqstp->rq_arg.head[0];
	head->arg.tail[0] = rqstp->rq_arg.tail[0];
	head->arg.pages = &head->pages[head->count];
	head->hdr_count = head->count; /* save count of hdr pages */
	head->arg.page_base = 0;
	head->arg.page_len = ch_bytes;
	head->arg.len = rqstp->rq_arg.len + ch_bytes;
	head->arg.buflen = rqstp->rq_arg.buflen + ch_bytes;
	head->count++;
	chl_map->ch[0].start = 0;
	while (byte_count) {
		rpl_map->sge[sge_no].iov_base =
			page_address(rqstp->rq_arg.pages[page_no]) + page_off;
		sge_bytes = min_t(int, PAGE_SIZE-page_off, ch_bytes);
		rpl_map->sge[sge_no].iov_len = sge_bytes;
		/*
		 * Don't bump head->count here because the same page
		 * may be used by multiple SGE.
		 */
		head->arg.pages[page_no] = rqstp->rq_arg.pages[page_no];
		rqstp->rq_respages = &rqstp->rq_arg.pages[page_no+1];

		byte_count -= sge_bytes;
		ch_bytes -= sge_bytes;
		sge_no++;
		/*
		 * If all bytes for this chunk have been mapped to an
		 * SGE, move to the next SGE
		 */
		if (ch_bytes == 0) {
			chl_map->ch[ch_no].count =
				sge_no - chl_map->ch[ch_no].start;
			ch_no++;
			ch++;
			chl_map->ch[ch_no].start = sge_no;
			ch_bytes = ch->rc_target.rs_length;
			/* If bytes remaining account for next chunk */
			if (byte_count) {
				head->arg.page_len += ch_bytes;
				head->arg.len += ch_bytes;
				head->arg.buflen += ch_bytes;
			}
		}
		/*
		 * If this SGE consumed all of the page, move to the
		 * next page
		 */
		if ((sge_bytes + page_off) == PAGE_SIZE) {
			page_no++;
			page_off = 0;
			/*
			 * If there are still bytes left to map, bump
			 * the page count
			 */
			if (byte_count)
				head->count++;
		} else
			page_off += sge_bytes;
	}
	BUG_ON(byte_count != 0);
	return sge_no;
}

/* Map a read-chunk-list to an XDR and fast register the page-list.
 *
 * Assumptions:
 * - chunk[0]	position points to pages[0] at an offset of 0
 * - pages[]	will be made physically contiguous by creating a one-off memory
 *		region using the fastreg verb.
 * - byte_count is # of bytes in read-chunk-list
 * - ch_count	is # of chunks in read-chunk-list
 *
 * Output:
 * - sge array pointing into pages[] array.
 * - chunk_sge array specifying sge index and count for each
 *   chunk in the read list
 */
static int fast_reg_read_chunks(struct svcxprt_rdma *xprt,
				struct svc_rqst *rqstp,
				struct svc_rdma_op_ctxt *head,
				struct rpcrdma_msg *rmsgp,
				struct svc_rdma_req_map *rpl_map,
				struct svc_rdma_req_map *chl_map,
				int ch_count,
				int byte_count)
{
	int page_no;
	int ch_no;
	u32 offset;
	struct rpcrdma_read_chunk *ch;
	struct svc_rdma_fastreg_mr *frmr;
	int ret = 0;

	frmr = svc_rdma_get_frmr(xprt);
	if (IS_ERR(frmr))
		return -ENOMEM;

	head->frmr = frmr;
	head->arg.head[0] = rqstp->rq_arg.head[0];
	head->arg.tail[0] = rqstp->rq_arg.tail[0];
	head->arg.pages = &head->pages[head->count];
	head->hdr_count = head->count; /* save count of hdr pages */
	head->arg.page_base = 0;
	head->arg.page_len = byte_count;
	head->arg.len = rqstp->rq_arg.len + byte_count;
	head->arg.buflen = rqstp->rq_arg.buflen + byte_count;

	/* Fast register the page list */
	frmr->kva = page_address(rqstp->rq_arg.pages[0]);
	frmr->direction = DMA_FROM_DEVICE;
	frmr->access_flags = (IB_ACCESS_LOCAL_WRITE|IB_ACCESS_REMOTE_WRITE);
	frmr->map_len = byte_count;
	frmr->page_list_len = PAGE_ALIGN(byte_count) >> PAGE_SHIFT;
	for (page_no = 0; page_no < frmr->page_list_len; page_no++) {
		frmr->page_list->page_list[page_no] =
			ib_dma_map_single(xprt->sc_cm_id->device,
					  page_address(rqstp->rq_arg.pages[page_no]),
					  PAGE_SIZE, DMA_FROM_DEVICE);
		if (ib_dma_mapping_error(xprt->sc_cm_id->device,
					 frmr->page_list->page_list[page_no]))
			goto fatal_err;
		atomic_inc(&xprt->sc_dma_used);
		head->arg.pages[page_no] = rqstp->rq_arg.pages[page_no];
	}
	head->count += page_no;

	/* rq_respages points one past arg pages */
	rqstp->rq_respages = &rqstp->rq_arg.pages[page_no];

	/* Create the reply and chunk maps */
	offset = 0;
	ch = (struct rpcrdma_read_chunk *)&rmsgp->rm_body.rm_chunks[0];
	for (ch_no = 0; ch_no < ch_count; ch_no++) {
		rpl_map->sge[ch_no].iov_base = frmr->kva + offset;
		rpl_map->sge[ch_no].iov_len = ch->rc_target.rs_length;
		chl_map->ch[ch_no].count = 1;
		chl_map->ch[ch_no].start = ch_no;
		offset += ch->rc_target.rs_length;
		ch++;
	}

	ret = svc_rdma_fastreg(xprt, frmr);
	if (ret)
		goto fatal_err;

	return ch_no;

 fatal_err:
	printk("svcrdma: error fast registering xdr for xprt %p", xprt);
	svc_rdma_put_frmr(xprt, frmr);
	return -EIO;
}

static int rdma_set_ctxt_sge(struct svcxprt_rdma *xprt,
			     struct svc_rdma_op_ctxt *ctxt,
			     struct svc_rdma_fastreg_mr *frmr,
			     struct kvec *vec,
			     u64 *sgl_offset,
			     int count)
{
	int i;

	ctxt->count = count;
	ctxt->direction = DMA_FROM_DEVICE;
	for (i = 0; i < count; i++) {
		ctxt->sge[i].length = 0; /* in case map fails */
		if (!frmr) {
			ctxt->sge[i].addr =
				ib_dma_map_single(xprt->sc_cm_id->device,
						  vec[i].iov_base,
						  vec[i].iov_len,
						  DMA_FROM_DEVICE);
			if (ib_dma_mapping_error(xprt->sc_cm_id->device,
						 ctxt->sge[i].addr))
				return -EINVAL;
			ctxt->sge[i].lkey = xprt->sc_dma_lkey;
			atomic_inc(&xprt->sc_dma_used);
		} else {
			ctxt->sge[i].addr = (unsigned long)vec[i].iov_base;
			ctxt->sge[i].lkey = frmr->mr->lkey;
		}
		ctxt->sge[i].length = vec[i].iov_len;
		*sgl_offset = *sgl_offset + vec[i].iov_len;
	}
	return 0;
}

static int rdma_read_max_sge(struct svcxprt_rdma *xprt, int sge_count)
{
	if ((RDMA_TRANSPORT_IWARP ==
	     rdma_node_get_transport(xprt->sc_cm_id->
				     device->node_type))
	    && sge_count > 1)
		return 1;
	else
		return min_t(int, sge_count, xprt->sc_max_sge);
}

/*
 * Use RDMA_READ to read data from the advertised client buffer into the
 * XDR stream starting at rq_arg.head[0].iov_base.
 * Each chunk in the array
 * contains the following fields:
 * discrim      - '1', This isn't used for data placement
 * position     - The xdr stream offset (the same for every chunk)
 * handle       - RMR for client memory region
 * length       - data transfer length
 * offset       - 64 bit tagged offset in remote memory region
 *
 * On our side, we need to read into a pagelist. The first page immediately
 * follows the RPC header.
 *
 * This function returns:
 * 0 - No error and no read-list found.
 *
 * 1 - Successful read-list processing. The data is not yet in
 * the pagelist and therefore the RPC request must be deferred. The
 * I/O completion will enqueue the transport again and
 * svc_rdma_recvfrom will complete the request.
 *
 * <0 - Error processing/posting read-list.
 *
 * NOTE: The ctxt must not be touched after the last WR has been posted
 * because the I/O completion processing may occur on another
 * processor and free / modify the context. Ne touche pas!
 */
static int rdma_read_xdr(struct svcxprt_rdma *xprt,
			 struct rpcrdma_msg *rmsgp,
			 struct svc_rqst *rqstp,
			 struct svc_rdma_op_ctxt *hdr_ctxt)
{
	struct ib_send_wr read_wr;
	struct ib_send_wr inv_wr;
	int err = 0;
	int ch_no;
	int ch_count;
	int byte_count;
	int sge_count;
	u64 sgl_offset;
	struct rpcrdma_read_chunk *ch;
	struct svc_rdma_op_ctxt *ctxt = NULL;
	struct svc_rdma_req_map *rpl_map;
	struct svc_rdma_req_map *chl_map;

	/* If no read list is present, return 0 */
	ch = svc_rdma_get_read_chunk(rmsgp);
	if (!ch)
		return 0;

	svc_rdma_rcl_chunk_counts(ch, &ch_count, &byte_count);
	if (ch_count > RPCSVC_MAXPAGES)
		return -EINVAL;

	/* Allocate temporary reply and chunk maps */
	rpl_map = svc_rdma_get_req_map();
	chl_map = svc_rdma_get_req_map();

	if (!xprt->sc_frmr_pg_list_len)
		sge_count = map_read_chunks(xprt, rqstp, hdr_ctxt, rmsgp,
					    rpl_map, chl_map, ch_count,
					    byte_count);
	else
		sge_count = fast_reg_read_chunks(xprt, rqstp, hdr_ctxt, rmsgp,
						 rpl_map, chl_map, ch_count,
						 byte_count);
	if (sge_count < 0) {
		err = -EIO;
		goto out;
	}

	sgl_offset = 0;
	ch_no = 0;

	for (ch = (struct rpcrdma_read_chunk *)&rmsgp->rm_body.rm_chunks[0];
	     ch->rc_discrim != 0; ch++, ch_no++) {
next_sge:
		ctxt = svc_rdma_get_context(xprt);
		ctxt->direction = DMA_FROM_DEVICE;
		ctxt->frmr = hdr_ctxt->frmr;
		ctxt->read_hdr = NULL;
		clear_bit(RDMACTXT_F_LAST_CTXT, &ctxt->flags);
		clear_bit(RDMACTXT_F_FAST_UNREG, &ctxt->flags);

		/* Prepare READ WR */
		memset(&read_wr, 0, sizeof read_wr);
		read_wr.wr_id = (unsigned long)ctxt;
		read_wr.opcode = IB_WR_RDMA_READ;
		ctxt->wr_op = read_wr.opcode;
		read_wr.send_flags = IB_SEND_SIGNALED;
		read_wr.wr.rdma.rkey = ch->rc_target.rs_handle;
		read_wr.wr.rdma.remote_addr =
			get_unaligned(&(ch->rc_target.rs_offset)) +
			sgl_offset;
		read_wr.sg_list = ctxt->sge;
		read_wr.num_sge =
			rdma_read_max_sge(xprt, chl_map->ch[ch_no].count);
		err = rdma_set_ctxt_sge(xprt, ctxt, hdr_ctxt->frmr,
					&rpl_map->sge[chl_map->ch[ch_no].start],
					&sgl_offset,
					read_wr.num_sge);
		if (err) {
			svc_rdma_unmap_dma(ctxt);
			svc_rdma_put_context(ctxt, 0);
			goto out;
		}
		if (((ch+1)->rc_discrim == 0) &&
		    (read_wr.num_sge == chl_map->ch[ch_no].count)) {
			/*
			 * Mark the last RDMA_READ with a bit to
			 * indicate all RPC data has been fetched from
			 * the client and the RPC needs to be enqueued.
			 */
			set_bit(RDMACTXT_F_LAST_CTXT, &ctxt->flags);
			if (hdr_ctxt->frmr) {
				set_bit(RDMACTXT_F_FAST_UNREG, &ctxt->flags);
				/*
				 * Invalidate the local MR used to map the data
				 * sink.
				 */
				if (xprt->sc_dev_caps &
				    SVCRDMA_DEVCAP_READ_W_INV) {
					read_wr.opcode =
						IB_WR_RDMA_READ_WITH_INV;
					ctxt->wr_op = read_wr.opcode;
					read_wr.ex.invalidate_rkey =
						ctxt->frmr->mr->lkey;
				} else {
					/* Prepare INVALIDATE WR */
					memset(&inv_wr, 0, sizeof inv_wr);
					inv_wr.opcode = IB_WR_LOCAL_INV;
					inv_wr.send_flags = IB_SEND_SIGNALED;
					inv_wr.ex.invalidate_rkey =
						hdr_ctxt->frmr->mr->lkey;
					read_wr.next = &inv_wr;
				}
			}
			ctxt->read_hdr = hdr_ctxt;
		}
		/* Post the read */
		err = svc_rdma_send(xprt, &read_wr);
		if (err) {
			printk(KERN_ERR "svcrdma: Error %d posting RDMA_READ\n",
			       err);
			set_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags);
			svc_rdma_put_context(ctxt, 0);
			goto out;
		}
		atomic_inc(&rdma_stat_read);

		if (read_wr.num_sge < chl_map->ch[ch_no].count) {
			chl_map->ch[ch_no].count -= read_wr.num_sge;
			chl_map->ch[ch_no].start += read_wr.num_sge;
			goto next_sge;
		}
		sgl_offset = 0;
		err = 1;
	}

 out:
	svc_rdma_put_req_map(rpl_map);
	svc_rdma_put_req_map(chl_map);

	/* Detach arg pages. svc_recv will replenish them */
	for (ch_no = 0; &rqstp->rq_pages[ch_no] < rqstp->rq_respages; ch_no++)
		rqstp->rq_pages[ch_no] = NULL;

	/*
	 * Detach res pages. svc_release must see a resused count of
	 * zero or it will attempt to put them.
	 */
	while (rqstp->rq_resused)
		rqstp->rq_respages[--rqstp->rq_resused] = NULL;

	return err;
}

static int rdma_read_complete(struct svc_rqst *rqstp,
			      struct svc_rdma_op_ctxt *head)
{
	int page_no;
	int ret;

	BUG_ON(!head);

	/* Copy RPC pages */
	for (page_no = 0; page_no < head->count; page_no++) {
		put_page(rqstp->rq_pages[page_no]);
		rqstp->rq_pages[page_no] = head->pages[page_no];
	}
	/* Point rq_arg.pages past header */
	rqstp->rq_arg.pages = &rqstp->rq_pages[head->hdr_count];
	rqstp->rq_arg.page_len = head->arg.page_len;
	rqstp->rq_arg.page_base = head->arg.page_base;

	/* rq_respages starts after the last arg page */
	rqstp->rq_respages = &rqstp->rq_arg.pages[page_no];
	rqstp->rq_resused = 0;

	/* Rebuild rq_arg head and tail. */
	rqstp->rq_arg.head[0] = head->arg.head[0];
	rqstp->rq_arg.tail[0] = head->arg.tail[0];
	rqstp->rq_arg.len = head->arg.len;
	rqstp->rq_arg.buflen = head->arg.buflen;

	/* Free the context */
	svc_rdma_put_context(head, 0);

	/* XXX: What should this be? */
	rqstp->rq_prot = IPPROTO_MAX;
	svc_xprt_copy_addrs(rqstp, rqstp->rq_xprt);

	ret = rqstp->rq_arg.head[0].iov_len
		+ rqstp->rq_arg.page_len
		+ rqstp->rq_arg.tail[0].iov_len;
	dprintk("svcrdma: deferred read ret=%d, rq_arg.len =%d, "
		"rq_arg.head[0].iov_base=%p, rq_arg.head[0].iov_len = %zd\n",
		ret, rqstp->rq_arg.len,	rqstp->rq_arg.head[0].iov_base,
		rqstp->rq_arg.head[0].iov_len);

	svc_xprt_received(rqstp->rq_xprt);
	return ret;
}

/*
 * Set up the rqstp thread context to point to the RQ buffer. If
 * necessary, pull additional data from the client with an RDMA_READ
 * request.
 */
int svc_rdma_recvfrom(struct svc_rqst *rqstp)
{
	struct svc_xprt *xprt = rqstp->rq_xprt;
	struct svcxprt_rdma *rdma_xprt =
		container_of(xprt, struct svcxprt_rdma, sc_xprt);
	struct svc_rdma_op_ctxt *ctxt = NULL;
	struct rpcrdma_msg *rmsgp;
	int ret = 0;
	int len;

	dprintk("svcrdma: rqstp=%p\n", rqstp);

	spin_lock_bh(&rdma_xprt->sc_rq_dto_lock);
	if (!list_empty(&rdma_xprt->sc_read_complete_q)) {
		ctxt = list_entry(rdma_xprt->sc_read_complete_q.next,
				  struct svc_rdma_op_ctxt,
				  dto_q);
		list_del_init(&ctxt->dto_q);
	}
	if (ctxt) {
		spin_unlock_bh(&rdma_xprt->sc_rq_dto_lock);
		return rdma_read_complete(rqstp, ctxt);
	}

	if (!list_empty(&rdma_xprt->sc_rq_dto_q)) {
		ctxt = list_entry(rdma_xprt->sc_rq_dto_q.next,
				  struct svc_rdma_op_ctxt,
				  dto_q);
		list_del_init(&ctxt->dto_q);
	} else {
		atomic_inc(&rdma_stat_rq_starve);
		clear_bit(XPT_DATA, &xprt->xpt_flags);
		ctxt = NULL;
	}
	spin_unlock_bh(&rdma_xprt->sc_rq_dto_lock);
	if (!ctxt) {
		/* This is the EAGAIN path. The svc_recv routine will
		 * return -EAGAIN, the nfsd thread will go to call into
		 * svc_recv again and we shouldn't be on the active
		 * transport list
		 */
		if (test_bit(XPT_CLOSE, &xprt->xpt_flags))
			goto close_out;

		BUG_ON(ret);
		goto out;
	}
	dprintk("svcrdma: processing ctxt=%p on xprt=%p, rqstp=%p, status=%d\n",
		ctxt, rdma_xprt, rqstp, ctxt->wc_status);
	BUG_ON(ctxt->wc_status != IB_WC_SUCCESS);
	atomic_inc(&rdma_stat_recv);

	/* Build up the XDR from the receive buffers. */
	rdma_build_arg_xdr(rqstp, ctxt, ctxt->byte_len);

	/* Decode the RDMA header. */
	len = svc_rdma_xdr_decode_req(&rmsgp, rqstp);
	rqstp->rq_xprt_hlen = len;

	/* If the request is invalid, reply with an error */
	if (len < 0) {
		if (len == -ENOSYS)
			svc_rdma_send_error(rdma_xprt, rmsgp, ERR_VERS);
		goto close_out;
	}

	/* Read read-list data. */
	ret = rdma_read_xdr(rdma_xprt, rmsgp, rqstp, ctxt);
	if (ret > 0) {
		/* read-list posted, defer until data received from client. */
		goto defer;
	}
	if (ret < 0) {
		/* Post of read-list failed, free context. */
		svc_rdma_put_context(ctxt, 1);
		return 0;
	}

	ret = rqstp->rq_arg.head[0].iov_len
		+ rqstp->rq_arg.page_len
		+ rqstp->rq_arg.tail[0].iov_len;
	svc_rdma_put_context(ctxt, 0);
 out:
	dprintk("svcrdma: ret = %d, rq_arg.len =%d, "
		"rq_arg.head[0].iov_base=%p, rq_arg.head[0].iov_len = %zd\n",
		ret, rqstp->rq_arg.len,
		rqstp->rq_arg.head[0].iov_base,
		rqstp->rq_arg.head[0].iov_len);
	rqstp->rq_prot = IPPROTO_MAX;
	svc_xprt_copy_addrs(rqstp, xprt);
	svc_xprt_received(xprt);
	return ret;

 close_out:
	if (ctxt)
		svc_rdma_put_context(ctxt, 1);
	dprintk("svcrdma: transport %p is closing\n", xprt);
	/*
	 * Set the close bit and enqueue it. svc_recv will see the
	 * close bit and call svc_xprt_delete
	 */
	set_bit(XPT_CLOSE, &xprt->xpt_flags);
defer:
	svc_xprt_received(xprt);
	return 0;
}