Loading...
1/*
2 * Copyright (c) 2006 Oracle. All rights reserved.
3 *
4 * This software is available to you under a choice of one of two
5 * licenses. You may choose to be licensed under the terms of the GNU
6 * General Public License (GPL) Version 2, available from the file
7 * COPYING in the main directory of this source tree, or the
8 * OpenIB.org BSD license below:
9 *
10 * Redistribution and use in source and binary forms, with or
11 * without modification, are permitted provided that the following
12 * conditions are met:
13 *
14 * - Redistributions of source code must retain the above
15 * copyright notice, this list of conditions and the following
16 * disclaimer.
17 *
18 * - Redistributions in binary form must reproduce the above
19 * copyright notice, this list of conditions and the following
20 * disclaimer in the documentation and/or other materials
21 * provided with the distribution.
22 *
23 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
24 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
25 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
26 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
27 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
28 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
29 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
30 * SOFTWARE.
31 *
32 */
33#include <linux/kernel.h>
34#include <linux/in.h>
35#include <linux/device.h>
36#include <linux/dmapool.h>
37#include <linux/ratelimit.h>
38
39#include "rds.h"
40#include "ib.h"
41
42static char *rds_ib_wc_status_strings[] = {
43#define RDS_IB_WC_STATUS_STR(foo) \
44 [IB_WC_##foo] = __stringify(IB_WC_##foo)
45 RDS_IB_WC_STATUS_STR(SUCCESS),
46 RDS_IB_WC_STATUS_STR(LOC_LEN_ERR),
47 RDS_IB_WC_STATUS_STR(LOC_QP_OP_ERR),
48 RDS_IB_WC_STATUS_STR(LOC_EEC_OP_ERR),
49 RDS_IB_WC_STATUS_STR(LOC_PROT_ERR),
50 RDS_IB_WC_STATUS_STR(WR_FLUSH_ERR),
51 RDS_IB_WC_STATUS_STR(MW_BIND_ERR),
52 RDS_IB_WC_STATUS_STR(BAD_RESP_ERR),
53 RDS_IB_WC_STATUS_STR(LOC_ACCESS_ERR),
54 RDS_IB_WC_STATUS_STR(REM_INV_REQ_ERR),
55 RDS_IB_WC_STATUS_STR(REM_ACCESS_ERR),
56 RDS_IB_WC_STATUS_STR(REM_OP_ERR),
57 RDS_IB_WC_STATUS_STR(RETRY_EXC_ERR),
58 RDS_IB_WC_STATUS_STR(RNR_RETRY_EXC_ERR),
59 RDS_IB_WC_STATUS_STR(LOC_RDD_VIOL_ERR),
60 RDS_IB_WC_STATUS_STR(REM_INV_RD_REQ_ERR),
61 RDS_IB_WC_STATUS_STR(REM_ABORT_ERR),
62 RDS_IB_WC_STATUS_STR(INV_EECN_ERR),
63 RDS_IB_WC_STATUS_STR(INV_EEC_STATE_ERR),
64 RDS_IB_WC_STATUS_STR(FATAL_ERR),
65 RDS_IB_WC_STATUS_STR(RESP_TIMEOUT_ERR),
66 RDS_IB_WC_STATUS_STR(GENERAL_ERR),
67#undef RDS_IB_WC_STATUS_STR
68};
69
70char *rds_ib_wc_status_str(enum ib_wc_status status)
71{
72 return rds_str_array(rds_ib_wc_status_strings,
73 ARRAY_SIZE(rds_ib_wc_status_strings), status);
74}
75
76/*
77 * Convert IB-specific error message to RDS error message and call core
78 * completion handler.
79 */
80static void rds_ib_send_complete(struct rds_message *rm,
81 int wc_status,
82 void (*complete)(struct rds_message *rm, int status))
83{
84 int notify_status;
85
86 switch (wc_status) {
87 case IB_WC_WR_FLUSH_ERR:
88 return;
89
90 case IB_WC_SUCCESS:
91 notify_status = RDS_RDMA_SUCCESS;
92 break;
93
94 case IB_WC_REM_ACCESS_ERR:
95 notify_status = RDS_RDMA_REMOTE_ERROR;
96 break;
97
98 default:
99 notify_status = RDS_RDMA_OTHER_ERROR;
100 break;
101 }
102 complete(rm, notify_status);
103}
104
105static void rds_ib_send_unmap_data(struct rds_ib_connection *ic,
106 struct rm_data_op *op,
107 int wc_status)
108{
109 if (op->op_nents)
110 ib_dma_unmap_sg(ic->i_cm_id->device,
111 op->op_sg, op->op_nents,
112 DMA_TO_DEVICE);
113}
114
115static void rds_ib_send_unmap_rdma(struct rds_ib_connection *ic,
116 struct rm_rdma_op *op,
117 int wc_status)
118{
119 if (op->op_mapped) {
120 ib_dma_unmap_sg(ic->i_cm_id->device,
121 op->op_sg, op->op_nents,
122 op->op_write ? DMA_TO_DEVICE : DMA_FROM_DEVICE);
123 op->op_mapped = 0;
124 }
125
126 /* If the user asked for a completion notification on this
127 * message, we can implement three different semantics:
128 * 1. Notify when we received the ACK on the RDS message
129 * that was queued with the RDMA. This provides reliable
130 * notification of RDMA status at the expense of a one-way
131 * packet delay.
132 * 2. Notify when the IB stack gives us the completion event for
133 * the RDMA operation.
134 * 3. Notify when the IB stack gives us the completion event for
135 * the accompanying RDS messages.
136 * Here, we implement approach #3. To implement approach #2,
137 * we would need to take an event for the rdma WR. To implement #1,
138 * don't call rds_rdma_send_complete at all, and fall back to the notify
139 * handling in the ACK processing code.
140 *
141 * Note: There's no need to explicitly sync any RDMA buffers using
142 * ib_dma_sync_sg_for_cpu - the completion for the RDMA
143 * operation itself unmapped the RDMA buffers, which takes care
144 * of synching.
145 */
146 rds_ib_send_complete(container_of(op, struct rds_message, rdma),
147 wc_status, rds_rdma_send_complete);
148
149 if (op->op_write)
150 rds_stats_add(s_send_rdma_bytes, op->op_bytes);
151 else
152 rds_stats_add(s_recv_rdma_bytes, op->op_bytes);
153}
154
155static void rds_ib_send_unmap_atomic(struct rds_ib_connection *ic,
156 struct rm_atomic_op *op,
157 int wc_status)
158{
159 /* unmap atomic recvbuf */
160 if (op->op_mapped) {
161 ib_dma_unmap_sg(ic->i_cm_id->device, op->op_sg, 1,
162 DMA_FROM_DEVICE);
163 op->op_mapped = 0;
164 }
165
166 rds_ib_send_complete(container_of(op, struct rds_message, atomic),
167 wc_status, rds_atomic_send_complete);
168
169 if (op->op_type == RDS_ATOMIC_TYPE_CSWP)
170 rds_ib_stats_inc(s_ib_atomic_cswp);
171 else
172 rds_ib_stats_inc(s_ib_atomic_fadd);
173}
174
175/*
176 * Unmap the resources associated with a struct send_work.
177 *
178 * Returns the rm for no good reason other than it is unobtainable
179 * other than by switching on wr.opcode, currently, and the caller,
180 * the event handler, needs it.
181 */
182static struct rds_message *rds_ib_send_unmap_op(struct rds_ib_connection *ic,
183 struct rds_ib_send_work *send,
184 int wc_status)
185{
186 struct rds_message *rm = NULL;
187
188 /* In the error case, wc.opcode sometimes contains garbage */
189 switch (send->s_wr.opcode) {
190 case IB_WR_SEND:
191 if (send->s_op) {
192 rm = container_of(send->s_op, struct rds_message, data);
193 rds_ib_send_unmap_data(ic, send->s_op, wc_status);
194 }
195 break;
196 case IB_WR_RDMA_WRITE:
197 case IB_WR_RDMA_READ:
198 if (send->s_op) {
199 rm = container_of(send->s_op, struct rds_message, rdma);
200 rds_ib_send_unmap_rdma(ic, send->s_op, wc_status);
201 }
202 break;
203 case IB_WR_ATOMIC_FETCH_AND_ADD:
204 case IB_WR_ATOMIC_CMP_AND_SWP:
205 if (send->s_op) {
206 rm = container_of(send->s_op, struct rds_message, atomic);
207 rds_ib_send_unmap_atomic(ic, send->s_op, wc_status);
208 }
209 break;
210 default:
211 printk_ratelimited(KERN_NOTICE
212 "RDS/IB: %s: unexpected opcode 0x%x in WR!\n",
213 __func__, send->s_wr.opcode);
214 break;
215 }
216
217 send->s_wr.opcode = 0xdead;
218
219 return rm;
220}
221
222void rds_ib_send_init_ring(struct rds_ib_connection *ic)
223{
224 struct rds_ib_send_work *send;
225 u32 i;
226
227 for (i = 0, send = ic->i_sends; i < ic->i_send_ring.w_nr; i++, send++) {
228 struct ib_sge *sge;
229
230 send->s_op = NULL;
231
232 send->s_wr.wr_id = i;
233 send->s_wr.sg_list = send->s_sge;
234 send->s_wr.ex.imm_data = 0;
235
236 sge = &send->s_sge[0];
237 sge->addr = ic->i_send_hdrs_dma + (i * sizeof(struct rds_header));
238 sge->length = sizeof(struct rds_header);
239 sge->lkey = ic->i_mr->lkey;
240
241 send->s_sge[1].lkey = ic->i_mr->lkey;
242 }
243}
244
245void rds_ib_send_clear_ring(struct rds_ib_connection *ic)
246{
247 struct rds_ib_send_work *send;
248 u32 i;
249
250 for (i = 0, send = ic->i_sends; i < ic->i_send_ring.w_nr; i++, send++) {
251 if (send->s_op && send->s_wr.opcode != 0xdead)
252 rds_ib_send_unmap_op(ic, send, IB_WC_WR_FLUSH_ERR);
253 }
254}
255
256/*
257 * The only fast path caller always has a non-zero nr, so we don't
258 * bother testing nr before performing the atomic sub.
259 */
260static void rds_ib_sub_signaled(struct rds_ib_connection *ic, int nr)
261{
262 if ((atomic_sub_return(nr, &ic->i_signaled_sends) == 0) &&
263 waitqueue_active(&rds_ib_ring_empty_wait))
264 wake_up(&rds_ib_ring_empty_wait);
265 BUG_ON(atomic_read(&ic->i_signaled_sends) < 0);
266}
267
268/*
269 * The _oldest/_free ring operations here race cleanly with the alloc/unalloc
270 * operations performed in the send path. As the sender allocs and potentially
271 * unallocs the next free entry in the ring it doesn't alter which is
272 * the next to be freed, which is what this is concerned with.
273 */
274void rds_ib_send_cq_comp_handler(struct ib_cq *cq, void *context)
275{
276 struct rds_connection *conn = context;
277 struct rds_ib_connection *ic = conn->c_transport_data;
278 struct rds_message *rm = NULL;
279 struct ib_wc wc;
280 struct rds_ib_send_work *send;
281 u32 completed;
282 u32 oldest;
283 u32 i = 0;
284 int ret;
285 int nr_sig = 0;
286
287 rdsdebug("cq %p conn %p\n", cq, conn);
288 rds_ib_stats_inc(s_ib_tx_cq_call);
289 ret = ib_req_notify_cq(cq, IB_CQ_NEXT_COMP);
290 if (ret)
291 rdsdebug("ib_req_notify_cq send failed: %d\n", ret);
292
293 while (ib_poll_cq(cq, 1, &wc) > 0) {
294 rdsdebug("wc wr_id 0x%llx status %u (%s) byte_len %u imm_data %u\n",
295 (unsigned long long)wc.wr_id, wc.status,
296 rds_ib_wc_status_str(wc.status), wc.byte_len,
297 be32_to_cpu(wc.ex.imm_data));
298 rds_ib_stats_inc(s_ib_tx_cq_event);
299
300 if (wc.wr_id == RDS_IB_ACK_WR_ID) {
301 if (ic->i_ack_queued + HZ/2 < jiffies)
302 rds_ib_stats_inc(s_ib_tx_stalled);
303 rds_ib_ack_send_complete(ic);
304 continue;
305 }
306
307 oldest = rds_ib_ring_oldest(&ic->i_send_ring);
308
309 completed = rds_ib_ring_completed(&ic->i_send_ring, wc.wr_id, oldest);
310
311 for (i = 0; i < completed; i++) {
312 send = &ic->i_sends[oldest];
313 if (send->s_wr.send_flags & IB_SEND_SIGNALED)
314 nr_sig++;
315
316 rm = rds_ib_send_unmap_op(ic, send, wc.status);
317
318 if (send->s_queued + HZ/2 < jiffies)
319 rds_ib_stats_inc(s_ib_tx_stalled);
320
321 if (send->s_op) {
322 if (send->s_op == rm->m_final_op) {
323 /* If anyone waited for this message to get flushed out, wake
324 * them up now */
325 rds_message_unmapped(rm);
326 }
327 rds_message_put(rm);
328 send->s_op = NULL;
329 }
330
331 oldest = (oldest + 1) % ic->i_send_ring.w_nr;
332 }
333
334 rds_ib_ring_free(&ic->i_send_ring, completed);
335 rds_ib_sub_signaled(ic, nr_sig);
336 nr_sig = 0;
337
338 if (test_and_clear_bit(RDS_LL_SEND_FULL, &conn->c_flags) ||
339 test_bit(0, &conn->c_map_queued))
340 queue_delayed_work(rds_wq, &conn->c_send_w, 0);
341
342 /* We expect errors as the qp is drained during shutdown */
343 if (wc.status != IB_WC_SUCCESS && rds_conn_up(conn)) {
344 rds_ib_conn_error(conn, "send completion on %pI4 had status "
345 "%u (%s), disconnecting and reconnecting\n",
346 &conn->c_faddr, wc.status,
347 rds_ib_wc_status_str(wc.status));
348 }
349 }
350}
351
352/*
353 * This is the main function for allocating credits when sending
354 * messages.
355 *
356 * Conceptually, we have two counters:
357 * - send credits: this tells us how many WRs we're allowed
358 * to submit without overruning the receiver's queue. For
359 * each SEND WR we post, we decrement this by one.
360 *
361 * - posted credits: this tells us how many WRs we recently
362 * posted to the receive queue. This value is transferred
363 * to the peer as a "credit update" in a RDS header field.
364 * Every time we transmit credits to the peer, we subtract
365 * the amount of transferred credits from this counter.
366 *
367 * It is essential that we avoid situations where both sides have
368 * exhausted their send credits, and are unable to send new credits
369 * to the peer. We achieve this by requiring that we send at least
370 * one credit update to the peer before exhausting our credits.
371 * When new credits arrive, we subtract one credit that is withheld
372 * until we've posted new buffers and are ready to transmit these
373 * credits (see rds_ib_send_add_credits below).
374 *
375 * The RDS send code is essentially single-threaded; rds_send_xmit
376 * sets RDS_IN_XMIT to ensure exclusive access to the send ring.
377 * However, the ACK sending code is independent and can race with
378 * message SENDs.
379 *
380 * In the send path, we need to update the counters for send credits
381 * and the counter of posted buffers atomically - when we use the
382 * last available credit, we cannot allow another thread to race us
383 * and grab the posted credits counter. Hence, we have to use a
384 * spinlock to protect the credit counter, or use atomics.
385 *
386 * Spinlocks shared between the send and the receive path are bad,
387 * because they create unnecessary delays. An early implementation
388 * using a spinlock showed a 5% degradation in throughput at some
389 * loads.
390 *
391 * This implementation avoids spinlocks completely, putting both
392 * counters into a single atomic, and updating that atomic using
393 * atomic_add (in the receive path, when receiving fresh credits),
394 * and using atomic_cmpxchg when updating the two counters.
395 */
396int rds_ib_send_grab_credits(struct rds_ib_connection *ic,
397 u32 wanted, u32 *adv_credits, int need_posted, int max_posted)
398{
399 unsigned int avail, posted, got = 0, advertise;
400 long oldval, newval;
401
402 *adv_credits = 0;
403 if (!ic->i_flowctl)
404 return wanted;
405
406try_again:
407 advertise = 0;
408 oldval = newval = atomic_read(&ic->i_credits);
409 posted = IB_GET_POST_CREDITS(oldval);
410 avail = IB_GET_SEND_CREDITS(oldval);
411
412 rdsdebug("rds_ib_send_grab_credits(%u): credits=%u posted=%u\n",
413 wanted, avail, posted);
414
415 /* The last credit must be used to send a credit update. */
416 if (avail && !posted)
417 avail--;
418
419 if (avail < wanted) {
420 struct rds_connection *conn = ic->i_cm_id->context;
421
422 /* Oops, there aren't that many credits left! */
423 set_bit(RDS_LL_SEND_FULL, &conn->c_flags);
424 got = avail;
425 } else {
426 /* Sometimes you get what you want, lalala. */
427 got = wanted;
428 }
429 newval -= IB_SET_SEND_CREDITS(got);
430
431 /*
432 * If need_posted is non-zero, then the caller wants
433 * the posted regardless of whether any send credits are
434 * available.
435 */
436 if (posted && (got || need_posted)) {
437 advertise = min_t(unsigned int, posted, max_posted);
438 newval -= IB_SET_POST_CREDITS(advertise);
439 }
440
441 /* Finally bill everything */
442 if (atomic_cmpxchg(&ic->i_credits, oldval, newval) != oldval)
443 goto try_again;
444
445 *adv_credits = advertise;
446 return got;
447}
448
449void rds_ib_send_add_credits(struct rds_connection *conn, unsigned int credits)
450{
451 struct rds_ib_connection *ic = conn->c_transport_data;
452
453 if (credits == 0)
454 return;
455
456 rdsdebug("rds_ib_send_add_credits(%u): current=%u%s\n",
457 credits,
458 IB_GET_SEND_CREDITS(atomic_read(&ic->i_credits)),
459 test_bit(RDS_LL_SEND_FULL, &conn->c_flags) ? ", ll_send_full" : "");
460
461 atomic_add(IB_SET_SEND_CREDITS(credits), &ic->i_credits);
462 if (test_and_clear_bit(RDS_LL_SEND_FULL, &conn->c_flags))
463 queue_delayed_work(rds_wq, &conn->c_send_w, 0);
464
465 WARN_ON(IB_GET_SEND_CREDITS(credits) >= 16384);
466
467 rds_ib_stats_inc(s_ib_rx_credit_updates);
468}
469
470void rds_ib_advertise_credits(struct rds_connection *conn, unsigned int posted)
471{
472 struct rds_ib_connection *ic = conn->c_transport_data;
473
474 if (posted == 0)
475 return;
476
477 atomic_add(IB_SET_POST_CREDITS(posted), &ic->i_credits);
478
479 /* Decide whether to send an update to the peer now.
480 * If we would send a credit update for every single buffer we
481 * post, we would end up with an ACK storm (ACK arrives,
482 * consumes buffer, we refill the ring, send ACK to remote
483 * advertising the newly posted buffer... ad inf)
484 *
485 * Performance pretty much depends on how often we send
486 * credit updates - too frequent updates mean lots of ACKs.
487 * Too infrequent updates, and the peer will run out of
488 * credits and has to throttle.
489 * For the time being, 16 seems to be a good compromise.
490 */
491 if (IB_GET_POST_CREDITS(atomic_read(&ic->i_credits)) >= 16)
492 set_bit(IB_ACK_REQUESTED, &ic->i_ack_flags);
493}
494
495static inline int rds_ib_set_wr_signal_state(struct rds_ib_connection *ic,
496 struct rds_ib_send_work *send,
497 bool notify)
498{
499 /*
500 * We want to delay signaling completions just enough to get
501 * the batching benefits but not so much that we create dead time
502 * on the wire.
503 */
504 if (ic->i_unsignaled_wrs-- == 0 || notify) {
505 ic->i_unsignaled_wrs = rds_ib_sysctl_max_unsig_wrs;
506 send->s_wr.send_flags |= IB_SEND_SIGNALED;
507 return 1;
508 }
509 return 0;
510}
511
512/*
513 * This can be called multiple times for a given message. The first time
514 * we see a message we map its scatterlist into the IB device so that
515 * we can provide that mapped address to the IB scatter gather entries
516 * in the IB work requests. We translate the scatterlist into a series
517 * of work requests that fragment the message. These work requests complete
518 * in order so we pass ownership of the message to the completion handler
519 * once we send the final fragment.
520 *
521 * The RDS core uses the c_send_lock to only enter this function once
522 * per connection. This makes sure that the tx ring alloc/unalloc pairs
523 * don't get out of sync and confuse the ring.
524 */
525int rds_ib_xmit(struct rds_connection *conn, struct rds_message *rm,
526 unsigned int hdr_off, unsigned int sg, unsigned int off)
527{
528 struct rds_ib_connection *ic = conn->c_transport_data;
529 struct ib_device *dev = ic->i_cm_id->device;
530 struct rds_ib_send_work *send = NULL;
531 struct rds_ib_send_work *first;
532 struct rds_ib_send_work *prev;
533 struct ib_send_wr *failed_wr;
534 struct scatterlist *scat;
535 u32 pos;
536 u32 i;
537 u32 work_alloc;
538 u32 credit_alloc = 0;
539 u32 posted;
540 u32 adv_credits = 0;
541 int send_flags = 0;
542 int bytes_sent = 0;
543 int ret;
544 int flow_controlled = 0;
545 int nr_sig = 0;
546
547 BUG_ON(off % RDS_FRAG_SIZE);
548 BUG_ON(hdr_off != 0 && hdr_off != sizeof(struct rds_header));
549
550 /* Do not send cong updates to IB loopback */
551 if (conn->c_loopback
552 && rm->m_inc.i_hdr.h_flags & RDS_FLAG_CONG_BITMAP) {
553 rds_cong_map_updated(conn->c_fcong, ~(u64) 0);
554 scat = &rm->data.op_sg[sg];
555 ret = sizeof(struct rds_header) + RDS_CONG_MAP_BYTES;
556 ret = min_t(int, ret, scat->length - conn->c_xmit_data_off);
557 return ret;
558 }
559
560 /* FIXME we may overallocate here */
561 if (be32_to_cpu(rm->m_inc.i_hdr.h_len) == 0)
562 i = 1;
563 else
564 i = ceil(be32_to_cpu(rm->m_inc.i_hdr.h_len), RDS_FRAG_SIZE);
565
566 work_alloc = rds_ib_ring_alloc(&ic->i_send_ring, i, &pos);
567 if (work_alloc == 0) {
568 set_bit(RDS_LL_SEND_FULL, &conn->c_flags);
569 rds_ib_stats_inc(s_ib_tx_ring_full);
570 ret = -ENOMEM;
571 goto out;
572 }
573
574 if (ic->i_flowctl) {
575 credit_alloc = rds_ib_send_grab_credits(ic, work_alloc, &posted, 0, RDS_MAX_ADV_CREDIT);
576 adv_credits += posted;
577 if (credit_alloc < work_alloc) {
578 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc - credit_alloc);
579 work_alloc = credit_alloc;
580 flow_controlled = 1;
581 }
582 if (work_alloc == 0) {
583 set_bit(RDS_LL_SEND_FULL, &conn->c_flags);
584 rds_ib_stats_inc(s_ib_tx_throttle);
585 ret = -ENOMEM;
586 goto out;
587 }
588 }
589
590 /* map the message the first time we see it */
591 if (!ic->i_data_op) {
592 if (rm->data.op_nents) {
593 rm->data.op_count = ib_dma_map_sg(dev,
594 rm->data.op_sg,
595 rm->data.op_nents,
596 DMA_TO_DEVICE);
597 rdsdebug("ic %p mapping rm %p: %d\n", ic, rm, rm->data.op_count);
598 if (rm->data.op_count == 0) {
599 rds_ib_stats_inc(s_ib_tx_sg_mapping_failure);
600 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
601 ret = -ENOMEM; /* XXX ? */
602 goto out;
603 }
604 } else {
605 rm->data.op_count = 0;
606 }
607
608 rds_message_addref(rm);
609 ic->i_data_op = &rm->data;
610
611 /* Finalize the header */
612 if (test_bit(RDS_MSG_ACK_REQUIRED, &rm->m_flags))
613 rm->m_inc.i_hdr.h_flags |= RDS_FLAG_ACK_REQUIRED;
614 if (test_bit(RDS_MSG_RETRANSMITTED, &rm->m_flags))
615 rm->m_inc.i_hdr.h_flags |= RDS_FLAG_RETRANSMITTED;
616
617 /* If it has a RDMA op, tell the peer we did it. This is
618 * used by the peer to release use-once RDMA MRs. */
619 if (rm->rdma.op_active) {
620 struct rds_ext_header_rdma ext_hdr;
621
622 ext_hdr.h_rdma_rkey = cpu_to_be32(rm->rdma.op_rkey);
623 rds_message_add_extension(&rm->m_inc.i_hdr,
624 RDS_EXTHDR_RDMA, &ext_hdr, sizeof(ext_hdr));
625 }
626 if (rm->m_rdma_cookie) {
627 rds_message_add_rdma_dest_extension(&rm->m_inc.i_hdr,
628 rds_rdma_cookie_key(rm->m_rdma_cookie),
629 rds_rdma_cookie_offset(rm->m_rdma_cookie));
630 }
631
632 /* Note - rds_ib_piggyb_ack clears the ACK_REQUIRED bit, so
633 * we should not do this unless we have a chance of at least
634 * sticking the header into the send ring. Which is why we
635 * should call rds_ib_ring_alloc first. */
636 rm->m_inc.i_hdr.h_ack = cpu_to_be64(rds_ib_piggyb_ack(ic));
637 rds_message_make_checksum(&rm->m_inc.i_hdr);
638
639 /*
640 * Update adv_credits since we reset the ACK_REQUIRED bit.
641 */
642 if (ic->i_flowctl) {
643 rds_ib_send_grab_credits(ic, 0, &posted, 1, RDS_MAX_ADV_CREDIT - adv_credits);
644 adv_credits += posted;
645 BUG_ON(adv_credits > 255);
646 }
647 }
648
649 /* Sometimes you want to put a fence between an RDMA
650 * READ and the following SEND.
651 * We could either do this all the time
652 * or when requested by the user. Right now, we let
653 * the application choose.
654 */
655 if (rm->rdma.op_active && rm->rdma.op_fence)
656 send_flags = IB_SEND_FENCE;
657
658 /* Each frag gets a header. Msgs may be 0 bytes */
659 send = &ic->i_sends[pos];
660 first = send;
661 prev = NULL;
662 scat = &ic->i_data_op->op_sg[sg];
663 i = 0;
664 do {
665 unsigned int len = 0;
666
667 /* Set up the header */
668 send->s_wr.send_flags = send_flags;
669 send->s_wr.opcode = IB_WR_SEND;
670 send->s_wr.num_sge = 1;
671 send->s_wr.next = NULL;
672 send->s_queued = jiffies;
673 send->s_op = NULL;
674
675 send->s_sge[0].addr = ic->i_send_hdrs_dma
676 + (pos * sizeof(struct rds_header));
677 send->s_sge[0].length = sizeof(struct rds_header);
678
679 memcpy(&ic->i_send_hdrs[pos], &rm->m_inc.i_hdr, sizeof(struct rds_header));
680
681 /* Set up the data, if present */
682 if (i < work_alloc
683 && scat != &rm->data.op_sg[rm->data.op_count]) {
684 len = min(RDS_FRAG_SIZE, ib_sg_dma_len(dev, scat) - off);
685 send->s_wr.num_sge = 2;
686
687 send->s_sge[1].addr = ib_sg_dma_address(dev, scat) + off;
688 send->s_sge[1].length = len;
689
690 bytes_sent += len;
691 off += len;
692 if (off == ib_sg_dma_len(dev, scat)) {
693 scat++;
694 off = 0;
695 }
696 }
697
698 rds_ib_set_wr_signal_state(ic, send, 0);
699
700 /*
701 * Always signal the last one if we're stopping due to flow control.
702 */
703 if (ic->i_flowctl && flow_controlled && i == (work_alloc-1))
704 send->s_wr.send_flags |= IB_SEND_SIGNALED | IB_SEND_SOLICITED;
705
706 if (send->s_wr.send_flags & IB_SEND_SIGNALED)
707 nr_sig++;
708
709 rdsdebug("send %p wr %p num_sge %u next %p\n", send,
710 &send->s_wr, send->s_wr.num_sge, send->s_wr.next);
711
712 if (ic->i_flowctl && adv_credits) {
713 struct rds_header *hdr = &ic->i_send_hdrs[pos];
714
715 /* add credit and redo the header checksum */
716 hdr->h_credit = adv_credits;
717 rds_message_make_checksum(hdr);
718 adv_credits = 0;
719 rds_ib_stats_inc(s_ib_tx_credit_updates);
720 }
721
722 if (prev)
723 prev->s_wr.next = &send->s_wr;
724 prev = send;
725
726 pos = (pos + 1) % ic->i_send_ring.w_nr;
727 send = &ic->i_sends[pos];
728 i++;
729
730 } while (i < work_alloc
731 && scat != &rm->data.op_sg[rm->data.op_count]);
732
733 /* Account the RDS header in the number of bytes we sent, but just once.
734 * The caller has no concept of fragmentation. */
735 if (hdr_off == 0)
736 bytes_sent += sizeof(struct rds_header);
737
738 /* if we finished the message then send completion owns it */
739 if (scat == &rm->data.op_sg[rm->data.op_count]) {
740 prev->s_op = ic->i_data_op;
741 prev->s_wr.send_flags |= IB_SEND_SOLICITED;
742 ic->i_data_op = NULL;
743 }
744
745 /* Put back wrs & credits we didn't use */
746 if (i < work_alloc) {
747 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc - i);
748 work_alloc = i;
749 }
750 if (ic->i_flowctl && i < credit_alloc)
751 rds_ib_send_add_credits(conn, credit_alloc - i);
752
753 if (nr_sig)
754 atomic_add(nr_sig, &ic->i_signaled_sends);
755
756 /* XXX need to worry about failed_wr and partial sends. */
757 failed_wr = &first->s_wr;
758 ret = ib_post_send(ic->i_cm_id->qp, &first->s_wr, &failed_wr);
759 rdsdebug("ic %p first %p (wr %p) ret %d wr %p\n", ic,
760 first, &first->s_wr, ret, failed_wr);
761 BUG_ON(failed_wr != &first->s_wr);
762 if (ret) {
763 printk(KERN_WARNING "RDS/IB: ib_post_send to %pI4 "
764 "returned %d\n", &conn->c_faddr, ret);
765 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
766 rds_ib_sub_signaled(ic, nr_sig);
767 if (prev->s_op) {
768 ic->i_data_op = prev->s_op;
769 prev->s_op = NULL;
770 }
771
772 rds_ib_conn_error(ic->conn, "ib_post_send failed\n");
773 goto out;
774 }
775
776 ret = bytes_sent;
777out:
778 BUG_ON(adv_credits);
779 return ret;
780}
781
782/*
783 * Issue atomic operation.
784 * A simplified version of the rdma case, we always map 1 SG, and
785 * only 8 bytes, for the return value from the atomic operation.
786 */
787int rds_ib_xmit_atomic(struct rds_connection *conn, struct rm_atomic_op *op)
788{
789 struct rds_ib_connection *ic = conn->c_transport_data;
790 struct rds_ib_send_work *send = NULL;
791 struct ib_send_wr *failed_wr;
792 struct rds_ib_device *rds_ibdev;
793 u32 pos;
794 u32 work_alloc;
795 int ret;
796 int nr_sig = 0;
797
798 rds_ibdev = ib_get_client_data(ic->i_cm_id->device, &rds_ib_client);
799
800 work_alloc = rds_ib_ring_alloc(&ic->i_send_ring, 1, &pos);
801 if (work_alloc != 1) {
802 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
803 rds_ib_stats_inc(s_ib_tx_ring_full);
804 ret = -ENOMEM;
805 goto out;
806 }
807
808 /* address of send request in ring */
809 send = &ic->i_sends[pos];
810 send->s_queued = jiffies;
811
812 if (op->op_type == RDS_ATOMIC_TYPE_CSWP) {
813 send->s_wr.opcode = IB_WR_MASKED_ATOMIC_CMP_AND_SWP;
814 send->s_wr.wr.atomic.compare_add = op->op_m_cswp.compare;
815 send->s_wr.wr.atomic.swap = op->op_m_cswp.swap;
816 send->s_wr.wr.atomic.compare_add_mask = op->op_m_cswp.compare_mask;
817 send->s_wr.wr.atomic.swap_mask = op->op_m_cswp.swap_mask;
818 } else { /* FADD */
819 send->s_wr.opcode = IB_WR_MASKED_ATOMIC_FETCH_AND_ADD;
820 send->s_wr.wr.atomic.compare_add = op->op_m_fadd.add;
821 send->s_wr.wr.atomic.swap = 0;
822 send->s_wr.wr.atomic.compare_add_mask = op->op_m_fadd.nocarry_mask;
823 send->s_wr.wr.atomic.swap_mask = 0;
824 }
825 nr_sig = rds_ib_set_wr_signal_state(ic, send, op->op_notify);
826 send->s_wr.num_sge = 1;
827 send->s_wr.next = NULL;
828 send->s_wr.wr.atomic.remote_addr = op->op_remote_addr;
829 send->s_wr.wr.atomic.rkey = op->op_rkey;
830 send->s_op = op;
831 rds_message_addref(container_of(send->s_op, struct rds_message, atomic));
832
833 /* map 8 byte retval buffer to the device */
834 ret = ib_dma_map_sg(ic->i_cm_id->device, op->op_sg, 1, DMA_FROM_DEVICE);
835 rdsdebug("ic %p mapping atomic op %p. mapped %d pg\n", ic, op, ret);
836 if (ret != 1) {
837 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
838 rds_ib_stats_inc(s_ib_tx_sg_mapping_failure);
839 ret = -ENOMEM; /* XXX ? */
840 goto out;
841 }
842
843 /* Convert our struct scatterlist to struct ib_sge */
844 send->s_sge[0].addr = ib_sg_dma_address(ic->i_cm_id->device, op->op_sg);
845 send->s_sge[0].length = ib_sg_dma_len(ic->i_cm_id->device, op->op_sg);
846 send->s_sge[0].lkey = ic->i_mr->lkey;
847
848 rdsdebug("rva %Lx rpa %Lx len %u\n", op->op_remote_addr,
849 send->s_sge[0].addr, send->s_sge[0].length);
850
851 if (nr_sig)
852 atomic_add(nr_sig, &ic->i_signaled_sends);
853
854 failed_wr = &send->s_wr;
855 ret = ib_post_send(ic->i_cm_id->qp, &send->s_wr, &failed_wr);
856 rdsdebug("ic %p send %p (wr %p) ret %d wr %p\n", ic,
857 send, &send->s_wr, ret, failed_wr);
858 BUG_ON(failed_wr != &send->s_wr);
859 if (ret) {
860 printk(KERN_WARNING "RDS/IB: atomic ib_post_send to %pI4 "
861 "returned %d\n", &conn->c_faddr, ret);
862 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
863 rds_ib_sub_signaled(ic, nr_sig);
864 goto out;
865 }
866
867 if (unlikely(failed_wr != &send->s_wr)) {
868 printk(KERN_WARNING "RDS/IB: atomic ib_post_send() rc=%d, but failed_wqe updated!\n", ret);
869 BUG_ON(failed_wr != &send->s_wr);
870 }
871
872out:
873 return ret;
874}
875
876int rds_ib_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op)
877{
878 struct rds_ib_connection *ic = conn->c_transport_data;
879 struct rds_ib_send_work *send = NULL;
880 struct rds_ib_send_work *first;
881 struct rds_ib_send_work *prev;
882 struct ib_send_wr *failed_wr;
883 struct scatterlist *scat;
884 unsigned long len;
885 u64 remote_addr = op->op_remote_addr;
886 u32 max_sge = ic->rds_ibdev->max_sge;
887 u32 pos;
888 u32 work_alloc;
889 u32 i;
890 u32 j;
891 int sent;
892 int ret;
893 int num_sge;
894 int nr_sig = 0;
895
896 /* map the op the first time we see it */
897 if (!op->op_mapped) {
898 op->op_count = ib_dma_map_sg(ic->i_cm_id->device,
899 op->op_sg, op->op_nents, (op->op_write) ?
900 DMA_TO_DEVICE : DMA_FROM_DEVICE);
901 rdsdebug("ic %p mapping op %p: %d\n", ic, op, op->op_count);
902 if (op->op_count == 0) {
903 rds_ib_stats_inc(s_ib_tx_sg_mapping_failure);
904 ret = -ENOMEM; /* XXX ? */
905 goto out;
906 }
907
908 op->op_mapped = 1;
909 }
910
911 /*
912 * Instead of knowing how to return a partial rdma read/write we insist that there
913 * be enough work requests to send the entire message.
914 */
915 i = ceil(op->op_count, max_sge);
916
917 work_alloc = rds_ib_ring_alloc(&ic->i_send_ring, i, &pos);
918 if (work_alloc != i) {
919 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
920 rds_ib_stats_inc(s_ib_tx_ring_full);
921 ret = -ENOMEM;
922 goto out;
923 }
924
925 send = &ic->i_sends[pos];
926 first = send;
927 prev = NULL;
928 scat = &op->op_sg[0];
929 sent = 0;
930 num_sge = op->op_count;
931
932 for (i = 0; i < work_alloc && scat != &op->op_sg[op->op_count]; i++) {
933 send->s_wr.send_flags = 0;
934 send->s_queued = jiffies;
935 send->s_op = NULL;
936
937 nr_sig += rds_ib_set_wr_signal_state(ic, send, op->op_notify);
938
939 send->s_wr.opcode = op->op_write ? IB_WR_RDMA_WRITE : IB_WR_RDMA_READ;
940 send->s_wr.wr.rdma.remote_addr = remote_addr;
941 send->s_wr.wr.rdma.rkey = op->op_rkey;
942
943 if (num_sge > max_sge) {
944 send->s_wr.num_sge = max_sge;
945 num_sge -= max_sge;
946 } else {
947 send->s_wr.num_sge = num_sge;
948 }
949
950 send->s_wr.next = NULL;
951
952 if (prev)
953 prev->s_wr.next = &send->s_wr;
954
955 for (j = 0; j < send->s_wr.num_sge && scat != &op->op_sg[op->op_count]; j++) {
956 len = ib_sg_dma_len(ic->i_cm_id->device, scat);
957 send->s_sge[j].addr =
958 ib_sg_dma_address(ic->i_cm_id->device, scat);
959 send->s_sge[j].length = len;
960 send->s_sge[j].lkey = ic->i_mr->lkey;
961
962 sent += len;
963 rdsdebug("ic %p sent %d remote_addr %llu\n", ic, sent, remote_addr);
964
965 remote_addr += len;
966 scat++;
967 }
968
969 rdsdebug("send %p wr %p num_sge %u next %p\n", send,
970 &send->s_wr, send->s_wr.num_sge, send->s_wr.next);
971
972 prev = send;
973 if (++send == &ic->i_sends[ic->i_send_ring.w_nr])
974 send = ic->i_sends;
975 }
976
977 /* give a reference to the last op */
978 if (scat == &op->op_sg[op->op_count]) {
979 prev->s_op = op;
980 rds_message_addref(container_of(op, struct rds_message, rdma));
981 }
982
983 if (i < work_alloc) {
984 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc - i);
985 work_alloc = i;
986 }
987
988 if (nr_sig)
989 atomic_add(nr_sig, &ic->i_signaled_sends);
990
991 failed_wr = &first->s_wr;
992 ret = ib_post_send(ic->i_cm_id->qp, &first->s_wr, &failed_wr);
993 rdsdebug("ic %p first %p (wr %p) ret %d wr %p\n", ic,
994 first, &first->s_wr, ret, failed_wr);
995 BUG_ON(failed_wr != &first->s_wr);
996 if (ret) {
997 printk(KERN_WARNING "RDS/IB: rdma ib_post_send to %pI4 "
998 "returned %d\n", &conn->c_faddr, ret);
999 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
1000 rds_ib_sub_signaled(ic, nr_sig);
1001 goto out;
1002 }
1003
1004 if (unlikely(failed_wr != &first->s_wr)) {
1005 printk(KERN_WARNING "RDS/IB: ib_post_send() rc=%d, but failed_wqe updated!\n", ret);
1006 BUG_ON(failed_wr != &first->s_wr);
1007 }
1008
1009
1010out:
1011 return ret;
1012}
1013
1014void rds_ib_xmit_complete(struct rds_connection *conn)
1015{
1016 struct rds_ib_connection *ic = conn->c_transport_data;
1017
1018 /* We may have a pending ACK or window update we were unable
1019 * to send previously (due to flow control). Try again. */
1020 rds_ib_attempt_ack(ic);
1021}
1/*
2 * Copyright (c) 2006 Oracle. All rights reserved.
3 *
4 * This software is available to you under a choice of one of two
5 * licenses. You may choose to be licensed under the terms of the GNU
6 * General Public License (GPL) Version 2, available from the file
7 * COPYING in the main directory of this source tree, or the
8 * OpenIB.org BSD license below:
9 *
10 * Redistribution and use in source and binary forms, with or
11 * without modification, are permitted provided that the following
12 * conditions are met:
13 *
14 * - Redistributions of source code must retain the above
15 * copyright notice, this list of conditions and the following
16 * disclaimer.
17 *
18 * - Redistributions in binary form must reproduce the above
19 * copyright notice, this list of conditions and the following
20 * disclaimer in the documentation and/or other materials
21 * provided with the distribution.
22 *
23 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
24 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
25 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
26 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
27 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
28 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
29 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
30 * SOFTWARE.
31 *
32 */
33#include <linux/kernel.h>
34#include <linux/in.h>
35#include <linux/device.h>
36#include <linux/dmapool.h>
37#include <linux/ratelimit.h>
38
39#include "rds.h"
40#include "ib.h"
41
42/*
43 * Convert IB-specific error message to RDS error message and call core
44 * completion handler.
45 */
46static void rds_ib_send_complete(struct rds_message *rm,
47 int wc_status,
48 void (*complete)(struct rds_message *rm, int status))
49{
50 int notify_status;
51
52 switch (wc_status) {
53 case IB_WC_WR_FLUSH_ERR:
54 return;
55
56 case IB_WC_SUCCESS:
57 notify_status = RDS_RDMA_SUCCESS;
58 break;
59
60 case IB_WC_REM_ACCESS_ERR:
61 notify_status = RDS_RDMA_REMOTE_ERROR;
62 break;
63
64 default:
65 notify_status = RDS_RDMA_OTHER_ERROR;
66 break;
67 }
68 complete(rm, notify_status);
69}
70
71static void rds_ib_send_unmap_data(struct rds_ib_connection *ic,
72 struct rm_data_op *op,
73 int wc_status)
74{
75 if (op->op_nents)
76 ib_dma_unmap_sg(ic->i_cm_id->device,
77 op->op_sg, op->op_nents,
78 DMA_TO_DEVICE);
79}
80
81static void rds_ib_send_unmap_rdma(struct rds_ib_connection *ic,
82 struct rm_rdma_op *op,
83 int wc_status)
84{
85 if (op->op_mapped) {
86 ib_dma_unmap_sg(ic->i_cm_id->device,
87 op->op_sg, op->op_nents,
88 op->op_write ? DMA_TO_DEVICE : DMA_FROM_DEVICE);
89 op->op_mapped = 0;
90 }
91
92 /* If the user asked for a completion notification on this
93 * message, we can implement three different semantics:
94 * 1. Notify when we received the ACK on the RDS message
95 * that was queued with the RDMA. This provides reliable
96 * notification of RDMA status at the expense of a one-way
97 * packet delay.
98 * 2. Notify when the IB stack gives us the completion event for
99 * the RDMA operation.
100 * 3. Notify when the IB stack gives us the completion event for
101 * the accompanying RDS messages.
102 * Here, we implement approach #3. To implement approach #2,
103 * we would need to take an event for the rdma WR. To implement #1,
104 * don't call rds_rdma_send_complete at all, and fall back to the notify
105 * handling in the ACK processing code.
106 *
107 * Note: There's no need to explicitly sync any RDMA buffers using
108 * ib_dma_sync_sg_for_cpu - the completion for the RDMA
109 * operation itself unmapped the RDMA buffers, which takes care
110 * of synching.
111 */
112 rds_ib_send_complete(container_of(op, struct rds_message, rdma),
113 wc_status, rds_rdma_send_complete);
114
115 if (op->op_write)
116 rds_stats_add(s_send_rdma_bytes, op->op_bytes);
117 else
118 rds_stats_add(s_recv_rdma_bytes, op->op_bytes);
119}
120
121static void rds_ib_send_unmap_atomic(struct rds_ib_connection *ic,
122 struct rm_atomic_op *op,
123 int wc_status)
124{
125 /* unmap atomic recvbuf */
126 if (op->op_mapped) {
127 ib_dma_unmap_sg(ic->i_cm_id->device, op->op_sg, 1,
128 DMA_FROM_DEVICE);
129 op->op_mapped = 0;
130 }
131
132 rds_ib_send_complete(container_of(op, struct rds_message, atomic),
133 wc_status, rds_atomic_send_complete);
134
135 if (op->op_type == RDS_ATOMIC_TYPE_CSWP)
136 rds_ib_stats_inc(s_ib_atomic_cswp);
137 else
138 rds_ib_stats_inc(s_ib_atomic_fadd);
139}
140
141/*
142 * Unmap the resources associated with a struct send_work.
143 *
144 * Returns the rm for no good reason other than it is unobtainable
145 * other than by switching on wr.opcode, currently, and the caller,
146 * the event handler, needs it.
147 */
148static struct rds_message *rds_ib_send_unmap_op(struct rds_ib_connection *ic,
149 struct rds_ib_send_work *send,
150 int wc_status)
151{
152 struct rds_message *rm = NULL;
153
154 /* In the error case, wc.opcode sometimes contains garbage */
155 switch (send->s_wr.opcode) {
156 case IB_WR_SEND:
157 if (send->s_op) {
158 rm = container_of(send->s_op, struct rds_message, data);
159 rds_ib_send_unmap_data(ic, send->s_op, wc_status);
160 }
161 break;
162 case IB_WR_RDMA_WRITE:
163 case IB_WR_RDMA_READ:
164 if (send->s_op) {
165 rm = container_of(send->s_op, struct rds_message, rdma);
166 rds_ib_send_unmap_rdma(ic, send->s_op, wc_status);
167 }
168 break;
169 case IB_WR_ATOMIC_FETCH_AND_ADD:
170 case IB_WR_ATOMIC_CMP_AND_SWP:
171 if (send->s_op) {
172 rm = container_of(send->s_op, struct rds_message, atomic);
173 rds_ib_send_unmap_atomic(ic, send->s_op, wc_status);
174 }
175 break;
176 default:
177 printk_ratelimited(KERN_NOTICE
178 "RDS/IB: %s: unexpected opcode 0x%x in WR!\n",
179 __func__, send->s_wr.opcode);
180 break;
181 }
182
183 send->s_wr.opcode = 0xdead;
184
185 return rm;
186}
187
188void rds_ib_send_init_ring(struct rds_ib_connection *ic)
189{
190 struct rds_ib_send_work *send;
191 u32 i;
192
193 for (i = 0, send = ic->i_sends; i < ic->i_send_ring.w_nr; i++, send++) {
194 struct ib_sge *sge;
195
196 send->s_op = NULL;
197
198 send->s_wr.wr_id = i;
199 send->s_wr.sg_list = send->s_sge;
200 send->s_wr.ex.imm_data = 0;
201
202 sge = &send->s_sge[0];
203 sge->addr = ic->i_send_hdrs_dma + (i * sizeof(struct rds_header));
204 sge->length = sizeof(struct rds_header);
205 sge->lkey = ic->i_pd->local_dma_lkey;
206
207 send->s_sge[1].lkey = ic->i_pd->local_dma_lkey;
208 }
209}
210
211void rds_ib_send_clear_ring(struct rds_ib_connection *ic)
212{
213 struct rds_ib_send_work *send;
214 u32 i;
215
216 for (i = 0, send = ic->i_sends; i < ic->i_send_ring.w_nr; i++, send++) {
217 if (send->s_op && send->s_wr.opcode != 0xdead)
218 rds_ib_send_unmap_op(ic, send, IB_WC_WR_FLUSH_ERR);
219 }
220}
221
222/*
223 * The only fast path caller always has a non-zero nr, so we don't
224 * bother testing nr before performing the atomic sub.
225 */
226static void rds_ib_sub_signaled(struct rds_ib_connection *ic, int nr)
227{
228 if ((atomic_sub_return(nr, &ic->i_signaled_sends) == 0) &&
229 waitqueue_active(&rds_ib_ring_empty_wait))
230 wake_up(&rds_ib_ring_empty_wait);
231 BUG_ON(atomic_read(&ic->i_signaled_sends) < 0);
232}
233
234/*
235 * The _oldest/_free ring operations here race cleanly with the alloc/unalloc
236 * operations performed in the send path. As the sender allocs and potentially
237 * unallocs the next free entry in the ring it doesn't alter which is
238 * the next to be freed, which is what this is concerned with.
239 */
240void rds_ib_send_cqe_handler(struct rds_ib_connection *ic, struct ib_wc *wc)
241{
242 struct rds_message *rm = NULL;
243 struct rds_connection *conn = ic->conn;
244 struct rds_ib_send_work *send;
245 u32 completed;
246 u32 oldest;
247 u32 i = 0;
248 int nr_sig = 0;
249
250
251 rdsdebug("wc wr_id 0x%llx status %u (%s) byte_len %u imm_data %u\n",
252 (unsigned long long)wc->wr_id, wc->status,
253 ib_wc_status_msg(wc->status), wc->byte_len,
254 be32_to_cpu(wc->ex.imm_data));
255 rds_ib_stats_inc(s_ib_tx_cq_event);
256
257 if (wc->wr_id == RDS_IB_ACK_WR_ID) {
258 if (time_after(jiffies, ic->i_ack_queued + HZ / 2))
259 rds_ib_stats_inc(s_ib_tx_stalled);
260 rds_ib_ack_send_complete(ic);
261 return;
262 }
263
264 oldest = rds_ib_ring_oldest(&ic->i_send_ring);
265
266 completed = rds_ib_ring_completed(&ic->i_send_ring, wc->wr_id, oldest);
267
268 for (i = 0; i < completed; i++) {
269 send = &ic->i_sends[oldest];
270 if (send->s_wr.send_flags & IB_SEND_SIGNALED)
271 nr_sig++;
272
273 rm = rds_ib_send_unmap_op(ic, send, wc->status);
274
275 if (time_after(jiffies, send->s_queued + HZ / 2))
276 rds_ib_stats_inc(s_ib_tx_stalled);
277
278 if (send->s_op) {
279 if (send->s_op == rm->m_final_op) {
280 /* If anyone waited for this message to get
281 * flushed out, wake them up now
282 */
283 rds_message_unmapped(rm);
284 }
285 rds_message_put(rm);
286 send->s_op = NULL;
287 }
288
289 oldest = (oldest + 1) % ic->i_send_ring.w_nr;
290 }
291
292 rds_ib_ring_free(&ic->i_send_ring, completed);
293 rds_ib_sub_signaled(ic, nr_sig);
294 nr_sig = 0;
295
296 if (test_and_clear_bit(RDS_LL_SEND_FULL, &conn->c_flags) ||
297 test_bit(0, &conn->c_map_queued))
298 queue_delayed_work(rds_wq, &conn->c_send_w, 0);
299
300 /* We expect errors as the qp is drained during shutdown */
301 if (wc->status != IB_WC_SUCCESS && rds_conn_up(conn)) {
302 rds_ib_conn_error(conn, "send completion on %pI4 had status %u (%s), disconnecting and reconnecting\n",
303 &conn->c_faddr, wc->status,
304 ib_wc_status_msg(wc->status));
305 }
306}
307
308/*
309 * This is the main function for allocating credits when sending
310 * messages.
311 *
312 * Conceptually, we have two counters:
313 * - send credits: this tells us how many WRs we're allowed
314 * to submit without overruning the receiver's queue. For
315 * each SEND WR we post, we decrement this by one.
316 *
317 * - posted credits: this tells us how many WRs we recently
318 * posted to the receive queue. This value is transferred
319 * to the peer as a "credit update" in a RDS header field.
320 * Every time we transmit credits to the peer, we subtract
321 * the amount of transferred credits from this counter.
322 *
323 * It is essential that we avoid situations where both sides have
324 * exhausted their send credits, and are unable to send new credits
325 * to the peer. We achieve this by requiring that we send at least
326 * one credit update to the peer before exhausting our credits.
327 * When new credits arrive, we subtract one credit that is withheld
328 * until we've posted new buffers and are ready to transmit these
329 * credits (see rds_ib_send_add_credits below).
330 *
331 * The RDS send code is essentially single-threaded; rds_send_xmit
332 * sets RDS_IN_XMIT to ensure exclusive access to the send ring.
333 * However, the ACK sending code is independent and can race with
334 * message SENDs.
335 *
336 * In the send path, we need to update the counters for send credits
337 * and the counter of posted buffers atomically - when we use the
338 * last available credit, we cannot allow another thread to race us
339 * and grab the posted credits counter. Hence, we have to use a
340 * spinlock to protect the credit counter, or use atomics.
341 *
342 * Spinlocks shared between the send and the receive path are bad,
343 * because they create unnecessary delays. An early implementation
344 * using a spinlock showed a 5% degradation in throughput at some
345 * loads.
346 *
347 * This implementation avoids spinlocks completely, putting both
348 * counters into a single atomic, and updating that atomic using
349 * atomic_add (in the receive path, when receiving fresh credits),
350 * and using atomic_cmpxchg when updating the two counters.
351 */
352int rds_ib_send_grab_credits(struct rds_ib_connection *ic,
353 u32 wanted, u32 *adv_credits, int need_posted, int max_posted)
354{
355 unsigned int avail, posted, got = 0, advertise;
356 long oldval, newval;
357
358 *adv_credits = 0;
359 if (!ic->i_flowctl)
360 return wanted;
361
362try_again:
363 advertise = 0;
364 oldval = newval = atomic_read(&ic->i_credits);
365 posted = IB_GET_POST_CREDITS(oldval);
366 avail = IB_GET_SEND_CREDITS(oldval);
367
368 rdsdebug("wanted=%u credits=%u posted=%u\n",
369 wanted, avail, posted);
370
371 /* The last credit must be used to send a credit update. */
372 if (avail && !posted)
373 avail--;
374
375 if (avail < wanted) {
376 struct rds_connection *conn = ic->i_cm_id->context;
377
378 /* Oops, there aren't that many credits left! */
379 set_bit(RDS_LL_SEND_FULL, &conn->c_flags);
380 got = avail;
381 } else {
382 /* Sometimes you get what you want, lalala. */
383 got = wanted;
384 }
385 newval -= IB_SET_SEND_CREDITS(got);
386
387 /*
388 * If need_posted is non-zero, then the caller wants
389 * the posted regardless of whether any send credits are
390 * available.
391 */
392 if (posted && (got || need_posted)) {
393 advertise = min_t(unsigned int, posted, max_posted);
394 newval -= IB_SET_POST_CREDITS(advertise);
395 }
396
397 /* Finally bill everything */
398 if (atomic_cmpxchg(&ic->i_credits, oldval, newval) != oldval)
399 goto try_again;
400
401 *adv_credits = advertise;
402 return got;
403}
404
405void rds_ib_send_add_credits(struct rds_connection *conn, unsigned int credits)
406{
407 struct rds_ib_connection *ic = conn->c_transport_data;
408
409 if (credits == 0)
410 return;
411
412 rdsdebug("credits=%u current=%u%s\n",
413 credits,
414 IB_GET_SEND_CREDITS(atomic_read(&ic->i_credits)),
415 test_bit(RDS_LL_SEND_FULL, &conn->c_flags) ? ", ll_send_full" : "");
416
417 atomic_add(IB_SET_SEND_CREDITS(credits), &ic->i_credits);
418 if (test_and_clear_bit(RDS_LL_SEND_FULL, &conn->c_flags))
419 queue_delayed_work(rds_wq, &conn->c_send_w, 0);
420
421 WARN_ON(IB_GET_SEND_CREDITS(credits) >= 16384);
422
423 rds_ib_stats_inc(s_ib_rx_credit_updates);
424}
425
426void rds_ib_advertise_credits(struct rds_connection *conn, unsigned int posted)
427{
428 struct rds_ib_connection *ic = conn->c_transport_data;
429
430 if (posted == 0)
431 return;
432
433 atomic_add(IB_SET_POST_CREDITS(posted), &ic->i_credits);
434
435 /* Decide whether to send an update to the peer now.
436 * If we would send a credit update for every single buffer we
437 * post, we would end up with an ACK storm (ACK arrives,
438 * consumes buffer, we refill the ring, send ACK to remote
439 * advertising the newly posted buffer... ad inf)
440 *
441 * Performance pretty much depends on how often we send
442 * credit updates - too frequent updates mean lots of ACKs.
443 * Too infrequent updates, and the peer will run out of
444 * credits and has to throttle.
445 * For the time being, 16 seems to be a good compromise.
446 */
447 if (IB_GET_POST_CREDITS(atomic_read(&ic->i_credits)) >= 16)
448 set_bit(IB_ACK_REQUESTED, &ic->i_ack_flags);
449}
450
451static inline int rds_ib_set_wr_signal_state(struct rds_ib_connection *ic,
452 struct rds_ib_send_work *send,
453 bool notify)
454{
455 /*
456 * We want to delay signaling completions just enough to get
457 * the batching benefits but not so much that we create dead time
458 * on the wire.
459 */
460 if (ic->i_unsignaled_wrs-- == 0 || notify) {
461 ic->i_unsignaled_wrs = rds_ib_sysctl_max_unsig_wrs;
462 send->s_wr.send_flags |= IB_SEND_SIGNALED;
463 return 1;
464 }
465 return 0;
466}
467
468/*
469 * This can be called multiple times for a given message. The first time
470 * we see a message we map its scatterlist into the IB device so that
471 * we can provide that mapped address to the IB scatter gather entries
472 * in the IB work requests. We translate the scatterlist into a series
473 * of work requests that fragment the message. These work requests complete
474 * in order so we pass ownership of the message to the completion handler
475 * once we send the final fragment.
476 *
477 * The RDS core uses the c_send_lock to only enter this function once
478 * per connection. This makes sure that the tx ring alloc/unalloc pairs
479 * don't get out of sync and confuse the ring.
480 */
481int rds_ib_xmit(struct rds_connection *conn, struct rds_message *rm,
482 unsigned int hdr_off, unsigned int sg, unsigned int off)
483{
484 struct rds_ib_connection *ic = conn->c_transport_data;
485 struct ib_device *dev = ic->i_cm_id->device;
486 struct rds_ib_send_work *send = NULL;
487 struct rds_ib_send_work *first;
488 struct rds_ib_send_work *prev;
489 struct ib_send_wr *failed_wr;
490 struct scatterlist *scat;
491 u32 pos;
492 u32 i;
493 u32 work_alloc;
494 u32 credit_alloc = 0;
495 u32 posted;
496 u32 adv_credits = 0;
497 int send_flags = 0;
498 int bytes_sent = 0;
499 int ret;
500 int flow_controlled = 0;
501 int nr_sig = 0;
502
503 BUG_ON(off % RDS_FRAG_SIZE);
504 BUG_ON(hdr_off != 0 && hdr_off != sizeof(struct rds_header));
505
506 /* Do not send cong updates to IB loopback */
507 if (conn->c_loopback
508 && rm->m_inc.i_hdr.h_flags & RDS_FLAG_CONG_BITMAP) {
509 rds_cong_map_updated(conn->c_fcong, ~(u64) 0);
510 scat = &rm->data.op_sg[sg];
511 ret = max_t(int, RDS_CONG_MAP_BYTES, scat->length);
512 return sizeof(struct rds_header) + ret;
513 }
514
515 /* FIXME we may overallocate here */
516 if (be32_to_cpu(rm->m_inc.i_hdr.h_len) == 0)
517 i = 1;
518 else
519 i = ceil(be32_to_cpu(rm->m_inc.i_hdr.h_len), RDS_FRAG_SIZE);
520
521 work_alloc = rds_ib_ring_alloc(&ic->i_send_ring, i, &pos);
522 if (work_alloc == 0) {
523 set_bit(RDS_LL_SEND_FULL, &conn->c_flags);
524 rds_ib_stats_inc(s_ib_tx_ring_full);
525 ret = -ENOMEM;
526 goto out;
527 }
528
529 if (ic->i_flowctl) {
530 credit_alloc = rds_ib_send_grab_credits(ic, work_alloc, &posted, 0, RDS_MAX_ADV_CREDIT);
531 adv_credits += posted;
532 if (credit_alloc < work_alloc) {
533 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc - credit_alloc);
534 work_alloc = credit_alloc;
535 flow_controlled = 1;
536 }
537 if (work_alloc == 0) {
538 set_bit(RDS_LL_SEND_FULL, &conn->c_flags);
539 rds_ib_stats_inc(s_ib_tx_throttle);
540 ret = -ENOMEM;
541 goto out;
542 }
543 }
544
545 /* map the message the first time we see it */
546 if (!ic->i_data_op) {
547 if (rm->data.op_nents) {
548 rm->data.op_count = ib_dma_map_sg(dev,
549 rm->data.op_sg,
550 rm->data.op_nents,
551 DMA_TO_DEVICE);
552 rdsdebug("ic %p mapping rm %p: %d\n", ic, rm, rm->data.op_count);
553 if (rm->data.op_count == 0) {
554 rds_ib_stats_inc(s_ib_tx_sg_mapping_failure);
555 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
556 ret = -ENOMEM; /* XXX ? */
557 goto out;
558 }
559 } else {
560 rm->data.op_count = 0;
561 }
562
563 rds_message_addref(rm);
564 rm->data.op_dmasg = 0;
565 rm->data.op_dmaoff = 0;
566 ic->i_data_op = &rm->data;
567
568 /* Finalize the header */
569 if (test_bit(RDS_MSG_ACK_REQUIRED, &rm->m_flags))
570 rm->m_inc.i_hdr.h_flags |= RDS_FLAG_ACK_REQUIRED;
571 if (test_bit(RDS_MSG_RETRANSMITTED, &rm->m_flags))
572 rm->m_inc.i_hdr.h_flags |= RDS_FLAG_RETRANSMITTED;
573
574 /* If it has a RDMA op, tell the peer we did it. This is
575 * used by the peer to release use-once RDMA MRs. */
576 if (rm->rdma.op_active) {
577 struct rds_ext_header_rdma ext_hdr;
578
579 ext_hdr.h_rdma_rkey = cpu_to_be32(rm->rdma.op_rkey);
580 rds_message_add_extension(&rm->m_inc.i_hdr,
581 RDS_EXTHDR_RDMA, &ext_hdr, sizeof(ext_hdr));
582 }
583 if (rm->m_rdma_cookie) {
584 rds_message_add_rdma_dest_extension(&rm->m_inc.i_hdr,
585 rds_rdma_cookie_key(rm->m_rdma_cookie),
586 rds_rdma_cookie_offset(rm->m_rdma_cookie));
587 }
588
589 /* Note - rds_ib_piggyb_ack clears the ACK_REQUIRED bit, so
590 * we should not do this unless we have a chance of at least
591 * sticking the header into the send ring. Which is why we
592 * should call rds_ib_ring_alloc first. */
593 rm->m_inc.i_hdr.h_ack = cpu_to_be64(rds_ib_piggyb_ack(ic));
594 rds_message_make_checksum(&rm->m_inc.i_hdr);
595
596 /*
597 * Update adv_credits since we reset the ACK_REQUIRED bit.
598 */
599 if (ic->i_flowctl) {
600 rds_ib_send_grab_credits(ic, 0, &posted, 1, RDS_MAX_ADV_CREDIT - adv_credits);
601 adv_credits += posted;
602 BUG_ON(adv_credits > 255);
603 }
604 }
605
606 /* Sometimes you want to put a fence between an RDMA
607 * READ and the following SEND.
608 * We could either do this all the time
609 * or when requested by the user. Right now, we let
610 * the application choose.
611 */
612 if (rm->rdma.op_active && rm->rdma.op_fence)
613 send_flags = IB_SEND_FENCE;
614
615 /* Each frag gets a header. Msgs may be 0 bytes */
616 send = &ic->i_sends[pos];
617 first = send;
618 prev = NULL;
619 scat = &ic->i_data_op->op_sg[rm->data.op_dmasg];
620 i = 0;
621 do {
622 unsigned int len = 0;
623
624 /* Set up the header */
625 send->s_wr.send_flags = send_flags;
626 send->s_wr.opcode = IB_WR_SEND;
627 send->s_wr.num_sge = 1;
628 send->s_wr.next = NULL;
629 send->s_queued = jiffies;
630 send->s_op = NULL;
631
632 send->s_sge[0].addr = ic->i_send_hdrs_dma
633 + (pos * sizeof(struct rds_header));
634 send->s_sge[0].length = sizeof(struct rds_header);
635
636 memcpy(&ic->i_send_hdrs[pos], &rm->m_inc.i_hdr, sizeof(struct rds_header));
637
638 /* Set up the data, if present */
639 if (i < work_alloc
640 && scat != &rm->data.op_sg[rm->data.op_count]) {
641 len = min(RDS_FRAG_SIZE,
642 ib_sg_dma_len(dev, scat) - rm->data.op_dmaoff);
643 send->s_wr.num_sge = 2;
644
645 send->s_sge[1].addr = ib_sg_dma_address(dev, scat);
646 send->s_sge[1].addr += rm->data.op_dmaoff;
647 send->s_sge[1].length = len;
648
649 bytes_sent += len;
650 rm->data.op_dmaoff += len;
651 if (rm->data.op_dmaoff == ib_sg_dma_len(dev, scat)) {
652 scat++;
653 rm->data.op_dmasg++;
654 rm->data.op_dmaoff = 0;
655 }
656 }
657
658 rds_ib_set_wr_signal_state(ic, send, 0);
659
660 /*
661 * Always signal the last one if we're stopping due to flow control.
662 */
663 if (ic->i_flowctl && flow_controlled && i == (work_alloc-1))
664 send->s_wr.send_flags |= IB_SEND_SIGNALED | IB_SEND_SOLICITED;
665
666 if (send->s_wr.send_flags & IB_SEND_SIGNALED)
667 nr_sig++;
668
669 rdsdebug("send %p wr %p num_sge %u next %p\n", send,
670 &send->s_wr, send->s_wr.num_sge, send->s_wr.next);
671
672 if (ic->i_flowctl && adv_credits) {
673 struct rds_header *hdr = &ic->i_send_hdrs[pos];
674
675 /* add credit and redo the header checksum */
676 hdr->h_credit = adv_credits;
677 rds_message_make_checksum(hdr);
678 adv_credits = 0;
679 rds_ib_stats_inc(s_ib_tx_credit_updates);
680 }
681
682 if (prev)
683 prev->s_wr.next = &send->s_wr;
684 prev = send;
685
686 pos = (pos + 1) % ic->i_send_ring.w_nr;
687 send = &ic->i_sends[pos];
688 i++;
689
690 } while (i < work_alloc
691 && scat != &rm->data.op_sg[rm->data.op_count]);
692
693 /* Account the RDS header in the number of bytes we sent, but just once.
694 * The caller has no concept of fragmentation. */
695 if (hdr_off == 0)
696 bytes_sent += sizeof(struct rds_header);
697
698 /* if we finished the message then send completion owns it */
699 if (scat == &rm->data.op_sg[rm->data.op_count]) {
700 prev->s_op = ic->i_data_op;
701 prev->s_wr.send_flags |= IB_SEND_SOLICITED;
702 if (!(prev->s_wr.send_flags & IB_SEND_SIGNALED)) {
703 ic->i_unsignaled_wrs = rds_ib_sysctl_max_unsig_wrs;
704 prev->s_wr.send_flags |= IB_SEND_SIGNALED;
705 nr_sig++;
706 }
707 ic->i_data_op = NULL;
708 }
709
710 /* Put back wrs & credits we didn't use */
711 if (i < work_alloc) {
712 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc - i);
713 work_alloc = i;
714 }
715 if (ic->i_flowctl && i < credit_alloc)
716 rds_ib_send_add_credits(conn, credit_alloc - i);
717
718 if (nr_sig)
719 atomic_add(nr_sig, &ic->i_signaled_sends);
720
721 /* XXX need to worry about failed_wr and partial sends. */
722 failed_wr = &first->s_wr;
723 ret = ib_post_send(ic->i_cm_id->qp, &first->s_wr, &failed_wr);
724 rdsdebug("ic %p first %p (wr %p) ret %d wr %p\n", ic,
725 first, &first->s_wr, ret, failed_wr);
726 BUG_ON(failed_wr != &first->s_wr);
727 if (ret) {
728 printk(KERN_WARNING "RDS/IB: ib_post_send to %pI4 "
729 "returned %d\n", &conn->c_faddr, ret);
730 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
731 rds_ib_sub_signaled(ic, nr_sig);
732 if (prev->s_op) {
733 ic->i_data_op = prev->s_op;
734 prev->s_op = NULL;
735 }
736
737 rds_ib_conn_error(ic->conn, "ib_post_send failed\n");
738 goto out;
739 }
740
741 ret = bytes_sent;
742out:
743 BUG_ON(adv_credits);
744 return ret;
745}
746
747/*
748 * Issue atomic operation.
749 * A simplified version of the rdma case, we always map 1 SG, and
750 * only 8 bytes, for the return value from the atomic operation.
751 */
752int rds_ib_xmit_atomic(struct rds_connection *conn, struct rm_atomic_op *op)
753{
754 struct rds_ib_connection *ic = conn->c_transport_data;
755 struct rds_ib_send_work *send = NULL;
756 struct ib_send_wr *failed_wr;
757 struct rds_ib_device *rds_ibdev;
758 u32 pos;
759 u32 work_alloc;
760 int ret;
761 int nr_sig = 0;
762
763 rds_ibdev = ib_get_client_data(ic->i_cm_id->device, &rds_ib_client);
764
765 work_alloc = rds_ib_ring_alloc(&ic->i_send_ring, 1, &pos);
766 if (work_alloc != 1) {
767 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
768 rds_ib_stats_inc(s_ib_tx_ring_full);
769 ret = -ENOMEM;
770 goto out;
771 }
772
773 /* address of send request in ring */
774 send = &ic->i_sends[pos];
775 send->s_queued = jiffies;
776
777 if (op->op_type == RDS_ATOMIC_TYPE_CSWP) {
778 send->s_atomic_wr.wr.opcode = IB_WR_MASKED_ATOMIC_CMP_AND_SWP;
779 send->s_atomic_wr.compare_add = op->op_m_cswp.compare;
780 send->s_atomic_wr.swap = op->op_m_cswp.swap;
781 send->s_atomic_wr.compare_add_mask = op->op_m_cswp.compare_mask;
782 send->s_atomic_wr.swap_mask = op->op_m_cswp.swap_mask;
783 } else { /* FADD */
784 send->s_atomic_wr.wr.opcode = IB_WR_MASKED_ATOMIC_FETCH_AND_ADD;
785 send->s_atomic_wr.compare_add = op->op_m_fadd.add;
786 send->s_atomic_wr.swap = 0;
787 send->s_atomic_wr.compare_add_mask = op->op_m_fadd.nocarry_mask;
788 send->s_atomic_wr.swap_mask = 0;
789 }
790 nr_sig = rds_ib_set_wr_signal_state(ic, send, op->op_notify);
791 send->s_atomic_wr.wr.num_sge = 1;
792 send->s_atomic_wr.wr.next = NULL;
793 send->s_atomic_wr.remote_addr = op->op_remote_addr;
794 send->s_atomic_wr.rkey = op->op_rkey;
795 send->s_op = op;
796 rds_message_addref(container_of(send->s_op, struct rds_message, atomic));
797
798 /* map 8 byte retval buffer to the device */
799 ret = ib_dma_map_sg(ic->i_cm_id->device, op->op_sg, 1, DMA_FROM_DEVICE);
800 rdsdebug("ic %p mapping atomic op %p. mapped %d pg\n", ic, op, ret);
801 if (ret != 1) {
802 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
803 rds_ib_stats_inc(s_ib_tx_sg_mapping_failure);
804 ret = -ENOMEM; /* XXX ? */
805 goto out;
806 }
807
808 /* Convert our struct scatterlist to struct ib_sge */
809 send->s_sge[0].addr = ib_sg_dma_address(ic->i_cm_id->device, op->op_sg);
810 send->s_sge[0].length = ib_sg_dma_len(ic->i_cm_id->device, op->op_sg);
811 send->s_sge[0].lkey = ic->i_pd->local_dma_lkey;
812
813 rdsdebug("rva %Lx rpa %Lx len %u\n", op->op_remote_addr,
814 send->s_sge[0].addr, send->s_sge[0].length);
815
816 if (nr_sig)
817 atomic_add(nr_sig, &ic->i_signaled_sends);
818
819 failed_wr = &send->s_atomic_wr.wr;
820 ret = ib_post_send(ic->i_cm_id->qp, &send->s_atomic_wr.wr, &failed_wr);
821 rdsdebug("ic %p send %p (wr %p) ret %d wr %p\n", ic,
822 send, &send->s_atomic_wr, ret, failed_wr);
823 BUG_ON(failed_wr != &send->s_atomic_wr.wr);
824 if (ret) {
825 printk(KERN_WARNING "RDS/IB: atomic ib_post_send to %pI4 "
826 "returned %d\n", &conn->c_faddr, ret);
827 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
828 rds_ib_sub_signaled(ic, nr_sig);
829 goto out;
830 }
831
832 if (unlikely(failed_wr != &send->s_atomic_wr.wr)) {
833 printk(KERN_WARNING "RDS/IB: atomic ib_post_send() rc=%d, but failed_wqe updated!\n", ret);
834 BUG_ON(failed_wr != &send->s_atomic_wr.wr);
835 }
836
837out:
838 return ret;
839}
840
841int rds_ib_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op)
842{
843 struct rds_ib_connection *ic = conn->c_transport_data;
844 struct rds_ib_send_work *send = NULL;
845 struct rds_ib_send_work *first;
846 struct rds_ib_send_work *prev;
847 struct ib_send_wr *failed_wr;
848 struct scatterlist *scat;
849 unsigned long len;
850 u64 remote_addr = op->op_remote_addr;
851 u32 max_sge = ic->rds_ibdev->max_sge;
852 u32 pos;
853 u32 work_alloc;
854 u32 i;
855 u32 j;
856 int sent;
857 int ret;
858 int num_sge;
859 int nr_sig = 0;
860
861 /* map the op the first time we see it */
862 if (!op->op_mapped) {
863 op->op_count = ib_dma_map_sg(ic->i_cm_id->device,
864 op->op_sg, op->op_nents, (op->op_write) ?
865 DMA_TO_DEVICE : DMA_FROM_DEVICE);
866 rdsdebug("ic %p mapping op %p: %d\n", ic, op, op->op_count);
867 if (op->op_count == 0) {
868 rds_ib_stats_inc(s_ib_tx_sg_mapping_failure);
869 ret = -ENOMEM; /* XXX ? */
870 goto out;
871 }
872
873 op->op_mapped = 1;
874 }
875
876 /*
877 * Instead of knowing how to return a partial rdma read/write we insist that there
878 * be enough work requests to send the entire message.
879 */
880 i = ceil(op->op_count, max_sge);
881
882 work_alloc = rds_ib_ring_alloc(&ic->i_send_ring, i, &pos);
883 if (work_alloc != i) {
884 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
885 rds_ib_stats_inc(s_ib_tx_ring_full);
886 ret = -ENOMEM;
887 goto out;
888 }
889
890 send = &ic->i_sends[pos];
891 first = send;
892 prev = NULL;
893 scat = &op->op_sg[0];
894 sent = 0;
895 num_sge = op->op_count;
896
897 for (i = 0; i < work_alloc && scat != &op->op_sg[op->op_count]; i++) {
898 send->s_wr.send_flags = 0;
899 send->s_queued = jiffies;
900 send->s_op = NULL;
901
902 nr_sig += rds_ib_set_wr_signal_state(ic, send, op->op_notify);
903
904 send->s_wr.opcode = op->op_write ? IB_WR_RDMA_WRITE : IB_WR_RDMA_READ;
905 send->s_rdma_wr.remote_addr = remote_addr;
906 send->s_rdma_wr.rkey = op->op_rkey;
907
908 if (num_sge > max_sge) {
909 send->s_rdma_wr.wr.num_sge = max_sge;
910 num_sge -= max_sge;
911 } else {
912 send->s_rdma_wr.wr.num_sge = num_sge;
913 }
914
915 send->s_rdma_wr.wr.next = NULL;
916
917 if (prev)
918 prev->s_rdma_wr.wr.next = &send->s_rdma_wr.wr;
919
920 for (j = 0; j < send->s_rdma_wr.wr.num_sge &&
921 scat != &op->op_sg[op->op_count]; j++) {
922 len = ib_sg_dma_len(ic->i_cm_id->device, scat);
923 send->s_sge[j].addr =
924 ib_sg_dma_address(ic->i_cm_id->device, scat);
925 send->s_sge[j].length = len;
926 send->s_sge[j].lkey = ic->i_pd->local_dma_lkey;
927
928 sent += len;
929 rdsdebug("ic %p sent %d remote_addr %llu\n", ic, sent, remote_addr);
930
931 remote_addr += len;
932 scat++;
933 }
934
935 rdsdebug("send %p wr %p num_sge %u next %p\n", send,
936 &send->s_rdma_wr.wr,
937 send->s_rdma_wr.wr.num_sge,
938 send->s_rdma_wr.wr.next);
939
940 prev = send;
941 if (++send == &ic->i_sends[ic->i_send_ring.w_nr])
942 send = ic->i_sends;
943 }
944
945 /* give a reference to the last op */
946 if (scat == &op->op_sg[op->op_count]) {
947 prev->s_op = op;
948 rds_message_addref(container_of(op, struct rds_message, rdma));
949 }
950
951 if (i < work_alloc) {
952 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc - i);
953 work_alloc = i;
954 }
955
956 if (nr_sig)
957 atomic_add(nr_sig, &ic->i_signaled_sends);
958
959 failed_wr = &first->s_rdma_wr.wr;
960 ret = ib_post_send(ic->i_cm_id->qp, &first->s_rdma_wr.wr, &failed_wr);
961 rdsdebug("ic %p first %p (wr %p) ret %d wr %p\n", ic,
962 first, &first->s_rdma_wr.wr, ret, failed_wr);
963 BUG_ON(failed_wr != &first->s_rdma_wr.wr);
964 if (ret) {
965 printk(KERN_WARNING "RDS/IB: rdma ib_post_send to %pI4 "
966 "returned %d\n", &conn->c_faddr, ret);
967 rds_ib_ring_unalloc(&ic->i_send_ring, work_alloc);
968 rds_ib_sub_signaled(ic, nr_sig);
969 goto out;
970 }
971
972 if (unlikely(failed_wr != &first->s_rdma_wr.wr)) {
973 printk(KERN_WARNING "RDS/IB: ib_post_send() rc=%d, but failed_wqe updated!\n", ret);
974 BUG_ON(failed_wr != &first->s_rdma_wr.wr);
975 }
976
977
978out:
979 return ret;
980}
981
982void rds_ib_xmit_complete(struct rds_connection *conn)
983{
984 struct rds_ib_connection *ic = conn->c_transport_data;
985
986 /* We may have a pending ACK or window update we were unable
987 * to send previously (due to flow control). Try again. */
988 rds_ib_attempt_ack(ic);
989}