1 // SPDX-License-Identifier: GPL-2.0 2 /* 3 * Shared Memory Communications over RDMA (SMC-R) and RoCE 4 * 5 * Manage send buffer. 6 * Producer: 7 * Copy user space data into send buffer, if send buffer space available. 8 * Consumer: 9 * Trigger RDMA write into RMBE of peer and send CDC, if RMBE space available. 10 * 11 * Copyright IBM Corp. 2016 12 * 13 * Author(s): Ursula Braun <ubraun@linux.vnet.ibm.com> 14 */ 15 16 #include <linux/net.h> 17 #include <linux/rcupdate.h> 18 #include <linux/workqueue.h> 19 #include <linux/sched/signal.h> 20 21 #include <net/sock.h> 22 #include <net/tcp.h> 23 24 #include "smc.h" 25 #include "smc_wr.h" 26 #include "smc_cdc.h" 27 #include "smc_close.h" 28 #include "smc_ism.h" 29 #include "smc_tx.h" 30 #include "smc_stats.h" 31 #include "smc_tracepoint.h" 32 33 #define SMC_TX_WORK_DELAY 0 34 #define SMC_TX_CORK_DELAY (HZ >> 2) /* 250 ms */ 35 36 /***************************** sndbuf producer *******************************/ 37 38 /* callback implementation for sk.sk_write_space() 39 * to wakeup sndbuf producers that blocked with smc_tx_wait(). 40 * called under sk_socket lock. 41 */ 42 static void smc_tx_write_space(struct sock *sk) 43 { 44 struct socket *sock = sk->sk_socket; 45 struct smc_sock *smc = smc_sk(sk); 46 struct socket_wq *wq; 47 48 /* similar to sk_stream_write_space */ 49 if (atomic_read(&smc->conn.sndbuf_space) && sock) { 50 if (test_bit(SOCK_NOSPACE, &sock->flags)) 51 SMC_STAT_RMB_TX_FULL(smc, !smc->conn.lnk); 52 clear_bit(SOCK_NOSPACE, &sock->flags); 53 rcu_read_lock(); 54 wq = rcu_dereference(sk->sk_wq); 55 if (skwq_has_sleeper(wq)) 56 wake_up_interruptible_poll(&wq->wait, 57 EPOLLOUT | EPOLLWRNORM | 58 EPOLLWRBAND); 59 if (wq && wq->fasync_list && !(sk->sk_shutdown & SEND_SHUTDOWN)) 60 sock_wake_async(wq, SOCK_WAKE_SPACE, POLL_OUT); 61 rcu_read_unlock(); 62 } 63 } 64 65 /* Wakeup sndbuf producers that blocked with smc_tx_wait(). 66 * Cf. tcp_data_snd_check()=>tcp_check_space()=>tcp_new_space(). 67 */ 68 void smc_tx_sndbuf_nonfull(struct smc_sock *smc) 69 { 70 if (smc->sk.sk_socket && 71 test_bit(SOCK_NOSPACE, &smc->sk.sk_socket->flags)) 72 smc->sk.sk_write_space(&smc->sk); 73 } 74 75 /* blocks sndbuf producer until at least one byte of free space available 76 * or urgent Byte was consumed 77 */ 78 static int smc_tx_wait(struct smc_sock *smc, int flags) 79 { 80 DEFINE_WAIT_FUNC(wait, woken_wake_function); 81 struct smc_connection *conn = &smc->conn; 82 struct sock *sk = &smc->sk; 83 long timeo; 84 int rc = 0; 85 86 /* similar to sk_stream_wait_memory */ 87 timeo = sock_sndtimeo(sk, flags & MSG_DONTWAIT); 88 add_wait_queue(sk_sleep(sk), &wait); 89 while (1) { 90 sk_set_bit(SOCKWQ_ASYNC_NOSPACE, sk); 91 if (sk->sk_err || 92 (sk->sk_shutdown & SEND_SHUTDOWN) || 93 conn->killed || 94 conn->local_tx_ctrl.conn_state_flags.peer_done_writing) { 95 rc = -EPIPE; 96 break; 97 } 98 if (smc_cdc_rxed_any_close(conn)) { 99 rc = -ECONNRESET; 100 break; 101 } 102 if (!timeo) { 103 /* ensure EPOLLOUT is subsequently generated */ 104 set_bit(SOCK_NOSPACE, &sk->sk_socket->flags); 105 rc = -EAGAIN; 106 break; 107 } 108 if (signal_pending(current)) { 109 rc = sock_intr_errno(timeo); 110 break; 111 } 112 sk_clear_bit(SOCKWQ_ASYNC_NOSPACE, sk); 113 if (atomic_read(&conn->sndbuf_space) && !conn->urg_tx_pend) 114 break; /* at least 1 byte of free & no urgent data */ 115 set_bit(SOCK_NOSPACE, &sk->sk_socket->flags); 116 sk_wait_event(sk, &timeo, 117 sk->sk_err || 118 (sk->sk_shutdown & SEND_SHUTDOWN) || 119 smc_cdc_rxed_any_close(conn) || 120 (atomic_read(&conn->sndbuf_space) && 121 !conn->urg_tx_pend), 122 &wait); 123 } 124 remove_wait_queue(sk_sleep(sk), &wait); 125 return rc; 126 } 127 128 static bool smc_tx_is_corked(struct smc_sock *smc) 129 { 130 struct tcp_sock *tp = tcp_sk(smc->clcsock->sk); 131 132 return (tp->nonagle & TCP_NAGLE_CORK) ? true : false; 133 } 134 135 /* sndbuf producer: main API called by socket layer. 136 * called under sock lock. 137 */ 138 int smc_tx_sendmsg(struct smc_sock *smc, struct msghdr *msg, size_t len) 139 { 140 size_t copylen, send_done = 0, send_remaining = len; 141 size_t chunk_len, chunk_off, chunk_len_sum; 142 struct smc_connection *conn = &smc->conn; 143 union smc_host_cursor prep; 144 struct sock *sk = &smc->sk; 145 char *sndbuf_base; 146 int tx_cnt_prep; 147 int writespace; 148 int rc, chunk; 149 150 /* This should be in poll */ 151 sk_clear_bit(SOCKWQ_ASYNC_NOSPACE, sk); 152 153 if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN)) { 154 rc = -EPIPE; 155 goto out_err; 156 } 157 158 if (sk->sk_state == SMC_INIT) 159 return -ENOTCONN; 160 161 if (len > conn->sndbuf_desc->len) 162 SMC_STAT_RMB_TX_SIZE_SMALL(smc, !conn->lnk); 163 164 if (len > conn->peer_rmbe_size) 165 SMC_STAT_RMB_TX_PEER_SIZE_SMALL(smc, !conn->lnk); 166 167 if (msg->msg_flags & MSG_OOB) 168 SMC_STAT_INC(smc, urg_data_cnt); 169 170 while (msg_data_left(msg)) { 171 if (smc->sk.sk_shutdown & SEND_SHUTDOWN || 172 (smc->sk.sk_err == ECONNABORTED) || 173 conn->killed) 174 return -EPIPE; 175 if (smc_cdc_rxed_any_close(conn)) 176 return send_done ?: -ECONNRESET; 177 178 if (msg->msg_flags & MSG_OOB) 179 conn->local_tx_ctrl.prod_flags.urg_data_pending = 1; 180 181 if (!atomic_read(&conn->sndbuf_space) || conn->urg_tx_pend) { 182 if (send_done) 183 return send_done; 184 rc = smc_tx_wait(smc, msg->msg_flags); 185 if (rc) 186 goto out_err; 187 continue; 188 } 189 190 /* initialize variables for 1st iteration of subsequent loop */ 191 /* could be just 1 byte, even after smc_tx_wait above */ 192 writespace = atomic_read(&conn->sndbuf_space); 193 /* not more than what user space asked for */ 194 copylen = min_t(size_t, send_remaining, writespace); 195 /* determine start of sndbuf */ 196 sndbuf_base = conn->sndbuf_desc->cpu_addr; 197 smc_curs_copy(&prep, &conn->tx_curs_prep, conn); 198 tx_cnt_prep = prep.count; 199 /* determine chunks where to write into sndbuf */ 200 /* either unwrapped case, or 1st chunk of wrapped case */ 201 chunk_len = min_t(size_t, copylen, conn->sndbuf_desc->len - 202 tx_cnt_prep); 203 chunk_len_sum = chunk_len; 204 chunk_off = tx_cnt_prep; 205 smc_sndbuf_sync_sg_for_cpu(conn); 206 for (chunk = 0; chunk < 2; chunk++) { 207 rc = memcpy_from_msg(sndbuf_base + chunk_off, 208 msg, chunk_len); 209 if (rc) { 210 smc_sndbuf_sync_sg_for_device(conn); 211 if (send_done) 212 return send_done; 213 goto out_err; 214 } 215 send_done += chunk_len; 216 send_remaining -= chunk_len; 217 218 if (chunk_len_sum == copylen) 219 break; /* either on 1st or 2nd iteration */ 220 /* prepare next (== 2nd) iteration */ 221 chunk_len = copylen - chunk_len; /* remainder */ 222 chunk_len_sum += chunk_len; 223 chunk_off = 0; /* modulo offset in send ring buffer */ 224 } 225 smc_sndbuf_sync_sg_for_device(conn); 226 /* update cursors */ 227 smc_curs_add(conn->sndbuf_desc->len, &prep, copylen); 228 smc_curs_copy(&conn->tx_curs_prep, &prep, conn); 229 /* increased in send tasklet smc_cdc_tx_handler() */ 230 smp_mb__before_atomic(); 231 atomic_sub(copylen, &conn->sndbuf_space); 232 /* guarantee 0 <= sndbuf_space <= sndbuf_desc->len */ 233 smp_mb__after_atomic(); 234 /* since we just produced more new data into sndbuf, 235 * trigger sndbuf consumer: RDMA write into peer RMBE and CDC 236 */ 237 if ((msg->msg_flags & MSG_OOB) && !send_remaining) 238 conn->urg_tx_pend = true; 239 if ((msg->msg_flags & MSG_MORE || smc_tx_is_corked(smc)) && 240 (atomic_read(&conn->sndbuf_space) > 241 (conn->sndbuf_desc->len >> 1))) 242 /* for a corked socket defer the RDMA writes if there 243 * is still sufficient sndbuf_space available 244 */ 245 queue_delayed_work(conn->lgr->tx_wq, &conn->tx_work, 246 SMC_TX_CORK_DELAY); 247 else 248 smc_tx_sndbuf_nonempty(conn); 249 250 trace_smc_tx_sendmsg(smc, copylen); 251 } /* while (msg_data_left(msg)) */ 252 253 return send_done; 254 255 out_err: 256 rc = sk_stream_error(sk, msg->msg_flags, rc); 257 /* make sure we wake any epoll edge trigger waiter */ 258 if (unlikely(rc == -EAGAIN)) 259 sk->sk_write_space(sk); 260 return rc; 261 } 262 263 /***************************** sndbuf consumer *******************************/ 264 265 /* sndbuf consumer: actual data transfer of one target chunk with ISM write */ 266 int smcd_tx_ism_write(struct smc_connection *conn, void *data, size_t len, 267 u32 offset, int signal) 268 { 269 struct smc_ism_position pos; 270 int rc; 271 272 memset(&pos, 0, sizeof(pos)); 273 pos.token = conn->peer_token; 274 pos.index = conn->peer_rmbe_idx; 275 pos.offset = conn->tx_off + offset; 276 pos.signal = signal; 277 rc = smc_ism_write(conn->lgr->smcd, &pos, data, len); 278 if (rc) 279 conn->local_tx_ctrl.conn_state_flags.peer_conn_abort = 1; 280 return rc; 281 } 282 283 /* sndbuf consumer: actual data transfer of one target chunk with RDMA write */ 284 static int smc_tx_rdma_write(struct smc_connection *conn, int peer_rmbe_offset, 285 int num_sges, struct ib_rdma_wr *rdma_wr) 286 { 287 struct smc_link_group *lgr = conn->lgr; 288 struct smc_link *link = conn->lnk; 289 int rc; 290 291 rdma_wr->wr.wr_id = smc_wr_tx_get_next_wr_id(link); 292 rdma_wr->wr.num_sge = num_sges; 293 rdma_wr->remote_addr = 294 lgr->rtokens[conn->rtoken_idx][link->link_idx].dma_addr + 295 /* RMBE within RMB */ 296 conn->tx_off + 297 /* offset within RMBE */ 298 peer_rmbe_offset; 299 rdma_wr->rkey = lgr->rtokens[conn->rtoken_idx][link->link_idx].rkey; 300 rc = ib_post_send(link->roce_qp, &rdma_wr->wr, NULL); 301 if (rc) 302 smcr_link_down_cond_sched(link); 303 return rc; 304 } 305 306 /* sndbuf consumer */ 307 static inline void smc_tx_advance_cursors(struct smc_connection *conn, 308 union smc_host_cursor *prod, 309 union smc_host_cursor *sent, 310 size_t len) 311 { 312 smc_curs_add(conn->peer_rmbe_size, prod, len); 313 /* increased in recv tasklet smc_cdc_msg_rcv() */ 314 smp_mb__before_atomic(); 315 /* data in flight reduces usable snd_wnd */ 316 atomic_sub(len, &conn->peer_rmbe_space); 317 /* guarantee 0 <= peer_rmbe_space <= peer_rmbe_size */ 318 smp_mb__after_atomic(); 319 smc_curs_add(conn->sndbuf_desc->len, sent, len); 320 } 321 322 /* SMC-R helper for smc_tx_rdma_writes() */ 323 static int smcr_tx_rdma_writes(struct smc_connection *conn, size_t len, 324 size_t src_off, size_t src_len, 325 size_t dst_off, size_t dst_len, 326 struct smc_rdma_wr *wr_rdma_buf) 327 { 328 struct smc_link *link = conn->lnk; 329 330 dma_addr_t dma_addr = 331 sg_dma_address(conn->sndbuf_desc->sgt[link->link_idx].sgl); 332 int src_len_sum = src_len, dst_len_sum = dst_len; 333 int sent_count = src_off; 334 int srcchunk, dstchunk; 335 int num_sges; 336 int rc; 337 338 for (dstchunk = 0; dstchunk < 2; dstchunk++) { 339 struct ib_sge *sge = 340 wr_rdma_buf->wr_tx_rdma[dstchunk].wr.sg_list; 341 342 num_sges = 0; 343 for (srcchunk = 0; srcchunk < 2; srcchunk++) { 344 sge[srcchunk].addr = dma_addr + src_off; 345 sge[srcchunk].length = src_len; 346 num_sges++; 347 348 src_off += src_len; 349 if (src_off >= conn->sndbuf_desc->len) 350 src_off -= conn->sndbuf_desc->len; 351 /* modulo in send ring */ 352 if (src_len_sum == dst_len) 353 break; /* either on 1st or 2nd iteration */ 354 /* prepare next (== 2nd) iteration */ 355 src_len = dst_len - src_len; /* remainder */ 356 src_len_sum += src_len; 357 } 358 rc = smc_tx_rdma_write(conn, dst_off, num_sges, 359 &wr_rdma_buf->wr_tx_rdma[dstchunk]); 360 if (rc) 361 return rc; 362 if (dst_len_sum == len) 363 break; /* either on 1st or 2nd iteration */ 364 /* prepare next (== 2nd) iteration */ 365 dst_off = 0; /* modulo offset in RMBE ring buffer */ 366 dst_len = len - dst_len; /* remainder */ 367 dst_len_sum += dst_len; 368 src_len = min_t(int, dst_len, conn->sndbuf_desc->len - 369 sent_count); 370 src_len_sum = src_len; 371 } 372 return 0; 373 } 374 375 /* SMC-D helper for smc_tx_rdma_writes() */ 376 static int smcd_tx_rdma_writes(struct smc_connection *conn, size_t len, 377 size_t src_off, size_t src_len, 378 size_t dst_off, size_t dst_len) 379 { 380 int src_len_sum = src_len, dst_len_sum = dst_len; 381 int srcchunk, dstchunk; 382 int rc; 383 384 for (dstchunk = 0; dstchunk < 2; dstchunk++) { 385 for (srcchunk = 0; srcchunk < 2; srcchunk++) { 386 void *data = conn->sndbuf_desc->cpu_addr + src_off; 387 388 rc = smcd_tx_ism_write(conn, data, src_len, dst_off + 389 sizeof(struct smcd_cdc_msg), 0); 390 if (rc) 391 return rc; 392 dst_off += src_len; 393 src_off += src_len; 394 if (src_off >= conn->sndbuf_desc->len) 395 src_off -= conn->sndbuf_desc->len; 396 /* modulo in send ring */ 397 if (src_len_sum == dst_len) 398 break; /* either on 1st or 2nd iteration */ 399 /* prepare next (== 2nd) iteration */ 400 src_len = dst_len - src_len; /* remainder */ 401 src_len_sum += src_len; 402 } 403 if (dst_len_sum == len) 404 break; /* either on 1st or 2nd iteration */ 405 /* prepare next (== 2nd) iteration */ 406 dst_off = 0; /* modulo offset in RMBE ring buffer */ 407 dst_len = len - dst_len; /* remainder */ 408 dst_len_sum += dst_len; 409 src_len = min_t(int, dst_len, conn->sndbuf_desc->len - src_off); 410 src_len_sum = src_len; 411 } 412 return 0; 413 } 414 415 /* sndbuf consumer: prepare all necessary (src&dst) chunks of data transmit; 416 * usable snd_wnd as max transmit 417 */ 418 static int smc_tx_rdma_writes(struct smc_connection *conn, 419 struct smc_rdma_wr *wr_rdma_buf) 420 { 421 size_t len, src_len, dst_off, dst_len; /* current chunk values */ 422 union smc_host_cursor sent, prep, prod, cons; 423 struct smc_cdc_producer_flags *pflags; 424 int to_send, rmbespace; 425 int rc; 426 427 /* source: sndbuf */ 428 smc_curs_copy(&sent, &conn->tx_curs_sent, conn); 429 smc_curs_copy(&prep, &conn->tx_curs_prep, conn); 430 /* cf. wmem_alloc - (snd_max - snd_una) */ 431 to_send = smc_curs_diff(conn->sndbuf_desc->len, &sent, &prep); 432 if (to_send <= 0) 433 return 0; 434 435 /* destination: RMBE */ 436 /* cf. snd_wnd */ 437 rmbespace = atomic_read(&conn->peer_rmbe_space); 438 if (rmbespace <= 0) { 439 struct smc_sock *smc = container_of(conn, struct smc_sock, 440 conn); 441 SMC_STAT_RMB_TX_PEER_FULL(smc, !conn->lnk); 442 return 0; 443 } 444 smc_curs_copy(&prod, &conn->local_tx_ctrl.prod, conn); 445 smc_curs_copy(&cons, &conn->local_rx_ctrl.cons, conn); 446 447 /* if usable snd_wnd closes ask peer to advertise once it opens again */ 448 pflags = &conn->local_tx_ctrl.prod_flags; 449 pflags->write_blocked = (to_send >= rmbespace); 450 /* cf. usable snd_wnd */ 451 len = min(to_send, rmbespace); 452 453 /* initialize variables for first iteration of subsequent nested loop */ 454 dst_off = prod.count; 455 if (prod.wrap == cons.wrap) { 456 /* the filled destination area is unwrapped, 457 * hence the available free destination space is wrapped 458 * and we need 2 destination chunks of sum len; start with 1st 459 * which is limited by what's available in sndbuf 460 */ 461 dst_len = min_t(size_t, 462 conn->peer_rmbe_size - prod.count, len); 463 } else { 464 /* the filled destination area is wrapped, 465 * hence the available free destination space is unwrapped 466 * and we need a single destination chunk of entire len 467 */ 468 dst_len = len; 469 } 470 /* dst_len determines the maximum src_len */ 471 if (sent.count + dst_len <= conn->sndbuf_desc->len) { 472 /* unwrapped src case: single chunk of entire dst_len */ 473 src_len = dst_len; 474 } else { 475 /* wrapped src case: 2 chunks of sum dst_len; start with 1st: */ 476 src_len = conn->sndbuf_desc->len - sent.count; 477 } 478 479 if (conn->lgr->is_smcd) 480 rc = smcd_tx_rdma_writes(conn, len, sent.count, src_len, 481 dst_off, dst_len); 482 else 483 rc = smcr_tx_rdma_writes(conn, len, sent.count, src_len, 484 dst_off, dst_len, wr_rdma_buf); 485 if (rc) 486 return rc; 487 488 if (conn->urg_tx_pend && len == to_send) 489 pflags->urg_data_present = 1; 490 smc_tx_advance_cursors(conn, &prod, &sent, len); 491 /* update connection's cursors with advanced local cursors */ 492 smc_curs_copy(&conn->local_tx_ctrl.prod, &prod, conn); 493 /* dst: peer RMBE */ 494 smc_curs_copy(&conn->tx_curs_sent, &sent, conn);/* src: local sndbuf */ 495 496 return 0; 497 } 498 499 /* Wakeup sndbuf consumers from any context (IRQ or process) 500 * since there is more data to transmit; usable snd_wnd as max transmit 501 */ 502 static int smcr_tx_sndbuf_nonempty(struct smc_connection *conn) 503 { 504 struct smc_cdc_producer_flags *pflags = &conn->local_tx_ctrl.prod_flags; 505 struct smc_link *link = conn->lnk; 506 struct smc_rdma_wr *wr_rdma_buf; 507 struct smc_cdc_tx_pend *pend; 508 struct smc_wr_buf *wr_buf; 509 int rc; 510 511 if (!link || !smc_wr_tx_link_hold(link)) 512 return -ENOLINK; 513 rc = smc_cdc_get_free_slot(conn, link, &wr_buf, &wr_rdma_buf, &pend); 514 if (rc < 0) { 515 smc_wr_tx_link_put(link); 516 if (rc == -EBUSY) { 517 struct smc_sock *smc = 518 container_of(conn, struct smc_sock, conn); 519 520 if (smc->sk.sk_err == ECONNABORTED) 521 return sock_error(&smc->sk); 522 if (conn->killed) 523 return -EPIPE; 524 rc = 0; 525 mod_delayed_work(conn->lgr->tx_wq, &conn->tx_work, 526 SMC_TX_WORK_DELAY); 527 } 528 return rc; 529 } 530 531 spin_lock_bh(&conn->send_lock); 532 if (link != conn->lnk) { 533 /* link of connection changed, tx_work will restart */ 534 smc_wr_tx_put_slot(link, 535 (struct smc_wr_tx_pend_priv *)pend); 536 rc = -ENOLINK; 537 goto out_unlock; 538 } 539 if (!pflags->urg_data_present) { 540 rc = smc_tx_rdma_writes(conn, wr_rdma_buf); 541 if (rc) { 542 smc_wr_tx_put_slot(link, 543 (struct smc_wr_tx_pend_priv *)pend); 544 goto out_unlock; 545 } 546 } 547 548 rc = smc_cdc_msg_send(conn, wr_buf, pend); 549 if (!rc && pflags->urg_data_present) { 550 pflags->urg_data_pending = 0; 551 pflags->urg_data_present = 0; 552 } 553 554 out_unlock: 555 spin_unlock_bh(&conn->send_lock); 556 smc_wr_tx_link_put(link); 557 return rc; 558 } 559 560 static int smcd_tx_sndbuf_nonempty(struct smc_connection *conn) 561 { 562 struct smc_cdc_producer_flags *pflags = &conn->local_tx_ctrl.prod_flags; 563 int rc = 0; 564 565 spin_lock_bh(&conn->send_lock); 566 if (!pflags->urg_data_present) 567 rc = smc_tx_rdma_writes(conn, NULL); 568 if (!rc) 569 rc = smcd_cdc_msg_send(conn); 570 571 if (!rc && pflags->urg_data_present) { 572 pflags->urg_data_pending = 0; 573 pflags->urg_data_present = 0; 574 } 575 spin_unlock_bh(&conn->send_lock); 576 return rc; 577 } 578 579 int smc_tx_sndbuf_nonempty(struct smc_connection *conn) 580 { 581 int rc; 582 583 if (conn->killed || 584 conn->local_rx_ctrl.conn_state_flags.peer_conn_abort) 585 return -EPIPE; /* connection being aborted */ 586 if (conn->lgr->is_smcd) 587 rc = smcd_tx_sndbuf_nonempty(conn); 588 else 589 rc = smcr_tx_sndbuf_nonempty(conn); 590 591 if (!rc) { 592 /* trigger socket release if connection is closing */ 593 struct smc_sock *smc = container_of(conn, struct smc_sock, 594 conn); 595 smc_close_wake_tx_prepared(smc); 596 } 597 return rc; 598 } 599 600 /* Wakeup sndbuf consumers from process context 601 * since there is more data to transmit 602 */ 603 void smc_tx_work(struct work_struct *work) 604 { 605 struct smc_connection *conn = container_of(to_delayed_work(work), 606 struct smc_connection, 607 tx_work); 608 struct smc_sock *smc = container_of(conn, struct smc_sock, conn); 609 int rc; 610 611 lock_sock(&smc->sk); 612 if (smc->sk.sk_err) 613 goto out; 614 615 rc = smc_tx_sndbuf_nonempty(conn); 616 if (!rc && conn->local_rx_ctrl.prod_flags.write_blocked && 617 !atomic_read(&conn->bytes_to_rcv)) 618 conn->local_rx_ctrl.prod_flags.write_blocked = 0; 619 620 out: 621 release_sock(&smc->sk); 622 } 623 624 void smc_tx_consumer_update(struct smc_connection *conn, bool force) 625 { 626 union smc_host_cursor cfed, cons, prod; 627 int sender_free = conn->rmb_desc->len; 628 int to_confirm; 629 630 smc_curs_copy(&cons, &conn->local_tx_ctrl.cons, conn); 631 smc_curs_copy(&cfed, &conn->rx_curs_confirmed, conn); 632 to_confirm = smc_curs_diff(conn->rmb_desc->len, &cfed, &cons); 633 if (to_confirm > conn->rmbe_update_limit) { 634 smc_curs_copy(&prod, &conn->local_rx_ctrl.prod, conn); 635 sender_free = conn->rmb_desc->len - 636 smc_curs_diff_large(conn->rmb_desc->len, 637 &cfed, &prod); 638 } 639 640 if (conn->local_rx_ctrl.prod_flags.cons_curs_upd_req || 641 force || 642 ((to_confirm > conn->rmbe_update_limit) && 643 ((sender_free <= (conn->rmb_desc->len / 2)) || 644 conn->local_rx_ctrl.prod_flags.write_blocked))) { 645 if (conn->killed || 646 conn->local_rx_ctrl.conn_state_flags.peer_conn_abort) 647 return; 648 if ((smc_cdc_get_slot_and_msg_send(conn) < 0) && 649 !conn->killed) { 650 queue_delayed_work(conn->lgr->tx_wq, &conn->tx_work, 651 SMC_TX_WORK_DELAY); 652 return; 653 } 654 } 655 if (conn->local_rx_ctrl.prod_flags.write_blocked && 656 !atomic_read(&conn->bytes_to_rcv)) 657 conn->local_rx_ctrl.prod_flags.write_blocked = 0; 658 } 659 660 /***************************** send initialize *******************************/ 661 662 /* Initialize send properties on connection establishment. NB: not __init! */ 663 void smc_tx_init(struct smc_sock *smc) 664 { 665 smc->sk.sk_write_space = smc_tx_write_space; 666 } 667