1 // SPDX-License-Identifier: GPL-2.0 2 #include <linux/kernel.h> 3 #include <linux/errno.h> 4 #include <linux/file.h> 5 #include <linux/slab.h> 6 #include <linux/net.h> 7 #include <linux/compat.h> 8 #include <net/compat.h> 9 #include <linux/io_uring.h> 10 11 #include <uapi/linux/io_uring.h> 12 13 #include "io_uring.h" 14 #include "kbuf.h" 15 #include "alloc_cache.h" 16 #include "net.h" 17 #include "notif.h" 18 #include "rsrc.h" 19 20 #if defined(CONFIG_NET) 21 struct io_shutdown { 22 struct file *file; 23 int how; 24 }; 25 26 struct io_accept { 27 struct file *file; 28 struct sockaddr __user *addr; 29 int __user *addr_len; 30 int flags; 31 u32 file_slot; 32 unsigned long nofile; 33 }; 34 35 struct io_socket { 36 struct file *file; 37 int domain; 38 int type; 39 int protocol; 40 int flags; 41 u32 file_slot; 42 unsigned long nofile; 43 }; 44 45 struct io_connect { 46 struct file *file; 47 struct sockaddr __user *addr; 48 int addr_len; 49 bool in_progress; 50 bool seen_econnaborted; 51 }; 52 53 struct io_sr_msg { 54 struct file *file; 55 union { 56 struct compat_msghdr __user *umsg_compat; 57 struct user_msghdr __user *umsg; 58 void __user *buf; 59 }; 60 unsigned len; 61 unsigned done_io; 62 unsigned msg_flags; 63 unsigned nr_multishot_loops; 64 u16 flags; 65 /* initialised and used only by !msg send variants */ 66 u16 addr_len; 67 u16 buf_group; 68 void __user *addr; 69 void __user *msg_control; 70 /* used only for send zerocopy */ 71 struct io_kiocb *notif; 72 }; 73 74 /* 75 * Number of times we'll try and do receives if there's more data. If we 76 * exceed this limit, then add us to the back of the queue and retry from 77 * there. This helps fairness between flooding clients. 78 */ 79 #define MULTISHOT_MAX_RETRY 32 80 81 int io_shutdown_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) 82 { 83 struct io_shutdown *shutdown = io_kiocb_to_cmd(req, struct io_shutdown); 84 85 if (unlikely(sqe->off || sqe->addr || sqe->rw_flags || 86 sqe->buf_index || sqe->splice_fd_in)) 87 return -EINVAL; 88 89 shutdown->how = READ_ONCE(sqe->len); 90 req->flags |= REQ_F_FORCE_ASYNC; 91 return 0; 92 } 93 94 int io_shutdown(struct io_kiocb *req, unsigned int issue_flags) 95 { 96 struct io_shutdown *shutdown = io_kiocb_to_cmd(req, struct io_shutdown); 97 struct socket *sock; 98 int ret; 99 100 WARN_ON_ONCE(issue_flags & IO_URING_F_NONBLOCK); 101 102 sock = sock_from_file(req->file); 103 if (unlikely(!sock)) 104 return -ENOTSOCK; 105 106 ret = __sys_shutdown_sock(sock, shutdown->how); 107 io_req_set_res(req, ret, 0); 108 return IOU_OK; 109 } 110 111 static bool io_net_retry(struct socket *sock, int flags) 112 { 113 if (!(flags & MSG_WAITALL)) 114 return false; 115 return sock->type == SOCK_STREAM || sock->type == SOCK_SEQPACKET; 116 } 117 118 static void io_netmsg_recycle(struct io_kiocb *req, unsigned int issue_flags) 119 { 120 struct io_async_msghdr *hdr = req->async_data; 121 122 if (!req_has_async_data(req) || issue_flags & IO_URING_F_UNLOCKED) 123 return; 124 125 /* Let normal cleanup path reap it if we fail adding to the cache */ 126 if (io_alloc_cache_put(&req->ctx->netmsg_cache, &hdr->cache)) { 127 req->async_data = NULL; 128 req->flags &= ~REQ_F_ASYNC_DATA; 129 } 130 } 131 132 static struct io_async_msghdr *io_msg_alloc_async(struct io_kiocb *req, 133 unsigned int issue_flags) 134 { 135 struct io_ring_ctx *ctx = req->ctx; 136 struct io_cache_entry *entry; 137 struct io_async_msghdr *hdr; 138 139 if (!(issue_flags & IO_URING_F_UNLOCKED)) { 140 entry = io_alloc_cache_get(&ctx->netmsg_cache); 141 if (entry) { 142 hdr = container_of(entry, struct io_async_msghdr, cache); 143 hdr->free_iov = NULL; 144 req->flags |= REQ_F_ASYNC_DATA; 145 req->async_data = hdr; 146 return hdr; 147 } 148 } 149 150 if (!io_alloc_async_data(req)) { 151 hdr = req->async_data; 152 hdr->free_iov = NULL; 153 return hdr; 154 } 155 return NULL; 156 } 157 158 static inline struct io_async_msghdr *io_msg_alloc_async_prep(struct io_kiocb *req) 159 { 160 /* ->prep_async is always called from the submission context */ 161 return io_msg_alloc_async(req, 0); 162 } 163 164 static int io_setup_async_msg(struct io_kiocb *req, 165 struct io_async_msghdr *kmsg, 166 unsigned int issue_flags) 167 { 168 struct io_async_msghdr *async_msg; 169 170 if (req_has_async_data(req)) 171 return -EAGAIN; 172 async_msg = io_msg_alloc_async(req, issue_flags); 173 if (!async_msg) { 174 kfree(kmsg->free_iov); 175 return -ENOMEM; 176 } 177 req->flags |= REQ_F_NEED_CLEANUP; 178 memcpy(async_msg, kmsg, sizeof(*kmsg)); 179 if (async_msg->msg.msg_name) 180 async_msg->msg.msg_name = &async_msg->addr; 181 182 if ((req->flags & REQ_F_BUFFER_SELECT) && !async_msg->msg.msg_iter.nr_segs) 183 return -EAGAIN; 184 185 /* if were using fast_iov, set it to the new one */ 186 if (iter_is_iovec(&kmsg->msg.msg_iter) && !kmsg->free_iov) { 187 size_t fast_idx = iter_iov(&kmsg->msg.msg_iter) - kmsg->fast_iov; 188 async_msg->msg.msg_iter.__iov = &async_msg->fast_iov[fast_idx]; 189 } 190 191 return -EAGAIN; 192 } 193 194 #ifdef CONFIG_COMPAT 195 static int io_compat_msg_copy_hdr(struct io_kiocb *req, 196 struct io_async_msghdr *iomsg, 197 struct compat_msghdr *msg, int ddir) 198 { 199 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 200 struct compat_iovec __user *uiov; 201 int ret; 202 203 if (copy_from_user(msg, sr->umsg_compat, sizeof(*msg))) 204 return -EFAULT; 205 206 uiov = compat_ptr(msg->msg_iov); 207 if (req->flags & REQ_F_BUFFER_SELECT) { 208 compat_ssize_t clen; 209 210 iomsg->free_iov = NULL; 211 if (msg->msg_iovlen == 0) { 212 sr->len = 0; 213 } else if (msg->msg_iovlen > 1) { 214 return -EINVAL; 215 } else { 216 if (!access_ok(uiov, sizeof(*uiov))) 217 return -EFAULT; 218 if (__get_user(clen, &uiov->iov_len)) 219 return -EFAULT; 220 if (clen < 0) 221 return -EINVAL; 222 sr->len = clen; 223 } 224 225 return 0; 226 } 227 228 iomsg->free_iov = iomsg->fast_iov; 229 ret = __import_iovec(ddir, (struct iovec __user *)uiov, msg->msg_iovlen, 230 UIO_FASTIOV, &iomsg->free_iov, 231 &iomsg->msg.msg_iter, true); 232 if (unlikely(ret < 0)) 233 return ret; 234 235 return 0; 236 } 237 #endif 238 239 static int io_msg_copy_hdr(struct io_kiocb *req, struct io_async_msghdr *iomsg, 240 struct user_msghdr *msg, int ddir) 241 { 242 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 243 int ret; 244 245 if (!user_access_begin(sr->umsg, sizeof(*sr->umsg))) 246 return -EFAULT; 247 248 ret = -EFAULT; 249 unsafe_get_user(msg->msg_name, &sr->umsg->msg_name, ua_end); 250 unsafe_get_user(msg->msg_namelen, &sr->umsg->msg_namelen, ua_end); 251 unsafe_get_user(msg->msg_iov, &sr->umsg->msg_iov, ua_end); 252 unsafe_get_user(msg->msg_iovlen, &sr->umsg->msg_iovlen, ua_end); 253 unsafe_get_user(msg->msg_control, &sr->umsg->msg_control, ua_end); 254 unsafe_get_user(msg->msg_controllen, &sr->umsg->msg_controllen, ua_end); 255 msg->msg_flags = 0; 256 257 if (req->flags & REQ_F_BUFFER_SELECT) { 258 if (msg->msg_iovlen == 0) { 259 sr->len = iomsg->fast_iov[0].iov_len = 0; 260 iomsg->fast_iov[0].iov_base = NULL; 261 iomsg->free_iov = NULL; 262 } else if (msg->msg_iovlen > 1) { 263 ret = -EINVAL; 264 goto ua_end; 265 } else { 266 /* we only need the length for provided buffers */ 267 if (!access_ok(&msg->msg_iov[0].iov_len, sizeof(__kernel_size_t))) 268 goto ua_end; 269 unsafe_get_user(iomsg->fast_iov[0].iov_len, 270 &msg->msg_iov[0].iov_len, ua_end); 271 sr->len = iomsg->fast_iov[0].iov_len; 272 iomsg->free_iov = NULL; 273 } 274 ret = 0; 275 ua_end: 276 user_access_end(); 277 return ret; 278 } 279 280 user_access_end(); 281 iomsg->free_iov = iomsg->fast_iov; 282 ret = __import_iovec(ddir, msg->msg_iov, msg->msg_iovlen, UIO_FASTIOV, 283 &iomsg->free_iov, &iomsg->msg.msg_iter, false); 284 if (unlikely(ret < 0)) 285 return ret; 286 287 return 0; 288 } 289 290 static int io_sendmsg_copy_hdr(struct io_kiocb *req, 291 struct io_async_msghdr *iomsg) 292 { 293 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 294 struct user_msghdr msg; 295 int ret; 296 297 iomsg->msg.msg_name = &iomsg->addr; 298 iomsg->msg.msg_iter.nr_segs = 0; 299 300 #ifdef CONFIG_COMPAT 301 if (unlikely(req->ctx->compat)) { 302 struct compat_msghdr cmsg; 303 304 ret = io_compat_msg_copy_hdr(req, iomsg, &cmsg, ITER_SOURCE); 305 if (unlikely(ret)) 306 return ret; 307 308 return __get_compat_msghdr(&iomsg->msg, &cmsg, NULL); 309 } 310 #endif 311 312 ret = io_msg_copy_hdr(req, iomsg, &msg, ITER_SOURCE); 313 if (unlikely(ret)) 314 return ret; 315 316 ret = __copy_msghdr(&iomsg->msg, &msg, NULL); 317 318 /* save msg_control as sys_sendmsg() overwrites it */ 319 sr->msg_control = iomsg->msg.msg_control_user; 320 return ret; 321 } 322 323 int io_send_prep_async(struct io_kiocb *req) 324 { 325 struct io_sr_msg *zc = io_kiocb_to_cmd(req, struct io_sr_msg); 326 struct io_async_msghdr *io; 327 int ret; 328 329 if (!zc->addr || req_has_async_data(req)) 330 return 0; 331 io = io_msg_alloc_async_prep(req); 332 if (!io) 333 return -ENOMEM; 334 ret = move_addr_to_kernel(zc->addr, zc->addr_len, &io->addr); 335 return ret; 336 } 337 338 static int io_setup_async_addr(struct io_kiocb *req, 339 struct sockaddr_storage *addr_storage, 340 unsigned int issue_flags) 341 { 342 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 343 struct io_async_msghdr *io; 344 345 if (!sr->addr || req_has_async_data(req)) 346 return -EAGAIN; 347 io = io_msg_alloc_async(req, issue_flags); 348 if (!io) 349 return -ENOMEM; 350 memcpy(&io->addr, addr_storage, sizeof(io->addr)); 351 return -EAGAIN; 352 } 353 354 int io_sendmsg_prep_async(struct io_kiocb *req) 355 { 356 int ret; 357 358 if (!io_msg_alloc_async_prep(req)) 359 return -ENOMEM; 360 ret = io_sendmsg_copy_hdr(req, req->async_data); 361 if (!ret) 362 req->flags |= REQ_F_NEED_CLEANUP; 363 return ret; 364 } 365 366 void io_sendmsg_recvmsg_cleanup(struct io_kiocb *req) 367 { 368 struct io_async_msghdr *io = req->async_data; 369 370 kfree(io->free_iov); 371 } 372 373 int io_sendmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) 374 { 375 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 376 377 sr->done_io = 0; 378 379 if (req->opcode == IORING_OP_SEND) { 380 if (READ_ONCE(sqe->__pad3[0])) 381 return -EINVAL; 382 sr->addr = u64_to_user_ptr(READ_ONCE(sqe->addr2)); 383 sr->addr_len = READ_ONCE(sqe->addr_len); 384 } else if (sqe->addr2 || sqe->file_index) { 385 return -EINVAL; 386 } 387 388 sr->umsg = u64_to_user_ptr(READ_ONCE(sqe->addr)); 389 sr->len = READ_ONCE(sqe->len); 390 sr->flags = READ_ONCE(sqe->ioprio); 391 if (sr->flags & ~IORING_RECVSEND_POLL_FIRST) 392 return -EINVAL; 393 sr->msg_flags = READ_ONCE(sqe->msg_flags) | MSG_NOSIGNAL; 394 if (sr->msg_flags & MSG_DONTWAIT) 395 req->flags |= REQ_F_NOWAIT; 396 397 #ifdef CONFIG_COMPAT 398 if (req->ctx->compat) 399 sr->msg_flags |= MSG_CMSG_COMPAT; 400 #endif 401 return 0; 402 } 403 404 static void io_req_msg_cleanup(struct io_kiocb *req, 405 struct io_async_msghdr *kmsg, 406 unsigned int issue_flags) 407 { 408 req->flags &= ~REQ_F_NEED_CLEANUP; 409 /* fast path, check for non-NULL to avoid function call */ 410 if (kmsg->free_iov) 411 kfree(kmsg->free_iov); 412 io_netmsg_recycle(req, issue_flags); 413 } 414 415 int io_sendmsg(struct io_kiocb *req, unsigned int issue_flags) 416 { 417 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 418 struct io_async_msghdr iomsg, *kmsg; 419 struct socket *sock; 420 unsigned flags; 421 int min_ret = 0; 422 int ret; 423 424 sock = sock_from_file(req->file); 425 if (unlikely(!sock)) 426 return -ENOTSOCK; 427 428 if (req_has_async_data(req)) { 429 kmsg = req->async_data; 430 kmsg->msg.msg_control_user = sr->msg_control; 431 } else { 432 ret = io_sendmsg_copy_hdr(req, &iomsg); 433 if (ret) 434 return ret; 435 kmsg = &iomsg; 436 } 437 438 if (!(req->flags & REQ_F_POLLED) && 439 (sr->flags & IORING_RECVSEND_POLL_FIRST)) 440 return io_setup_async_msg(req, kmsg, issue_flags); 441 442 flags = sr->msg_flags; 443 if (issue_flags & IO_URING_F_NONBLOCK) 444 flags |= MSG_DONTWAIT; 445 if (flags & MSG_WAITALL) 446 min_ret = iov_iter_count(&kmsg->msg.msg_iter); 447 448 ret = __sys_sendmsg_sock(sock, &kmsg->msg, flags); 449 450 if (ret < min_ret) { 451 if (ret == -EAGAIN && (issue_flags & IO_URING_F_NONBLOCK)) 452 return io_setup_async_msg(req, kmsg, issue_flags); 453 if (ret > 0 && io_net_retry(sock, flags)) { 454 kmsg->msg.msg_controllen = 0; 455 kmsg->msg.msg_control = NULL; 456 sr->done_io += ret; 457 req->flags |= REQ_F_BL_NO_RECYCLE; 458 return io_setup_async_msg(req, kmsg, issue_flags); 459 } 460 if (ret == -ERESTARTSYS) 461 ret = -EINTR; 462 req_set_fail(req); 463 } 464 io_req_msg_cleanup(req, kmsg, issue_flags); 465 if (ret >= 0) 466 ret += sr->done_io; 467 else if (sr->done_io) 468 ret = sr->done_io; 469 io_req_set_res(req, ret, 0); 470 return IOU_OK; 471 } 472 473 int io_send(struct io_kiocb *req, unsigned int issue_flags) 474 { 475 struct sockaddr_storage __address; 476 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 477 struct msghdr msg; 478 struct socket *sock; 479 unsigned flags; 480 int min_ret = 0; 481 int ret; 482 483 msg.msg_name = NULL; 484 msg.msg_control = NULL; 485 msg.msg_controllen = 0; 486 msg.msg_namelen = 0; 487 msg.msg_ubuf = NULL; 488 489 if (sr->addr) { 490 if (req_has_async_data(req)) { 491 struct io_async_msghdr *io = req->async_data; 492 493 msg.msg_name = &io->addr; 494 } else { 495 ret = move_addr_to_kernel(sr->addr, sr->addr_len, &__address); 496 if (unlikely(ret < 0)) 497 return ret; 498 msg.msg_name = (struct sockaddr *)&__address; 499 } 500 msg.msg_namelen = sr->addr_len; 501 } 502 503 if (!(req->flags & REQ_F_POLLED) && 504 (sr->flags & IORING_RECVSEND_POLL_FIRST)) 505 return io_setup_async_addr(req, &__address, issue_flags); 506 507 sock = sock_from_file(req->file); 508 if (unlikely(!sock)) 509 return -ENOTSOCK; 510 511 ret = import_ubuf(ITER_SOURCE, sr->buf, sr->len, &msg.msg_iter); 512 if (unlikely(ret)) 513 return ret; 514 515 flags = sr->msg_flags; 516 if (issue_flags & IO_URING_F_NONBLOCK) 517 flags |= MSG_DONTWAIT; 518 if (flags & MSG_WAITALL) 519 min_ret = iov_iter_count(&msg.msg_iter); 520 521 flags &= ~MSG_INTERNAL_SENDMSG_FLAGS; 522 msg.msg_flags = flags; 523 ret = sock_sendmsg(sock, &msg); 524 if (ret < min_ret) { 525 if (ret == -EAGAIN && (issue_flags & IO_URING_F_NONBLOCK)) 526 return io_setup_async_addr(req, &__address, issue_flags); 527 528 if (ret > 0 && io_net_retry(sock, flags)) { 529 sr->len -= ret; 530 sr->buf += ret; 531 sr->done_io += ret; 532 req->flags |= REQ_F_BL_NO_RECYCLE; 533 return io_setup_async_addr(req, &__address, issue_flags); 534 } 535 if (ret == -ERESTARTSYS) 536 ret = -EINTR; 537 req_set_fail(req); 538 } 539 if (ret >= 0) 540 ret += sr->done_io; 541 else if (sr->done_io) 542 ret = sr->done_io; 543 io_req_set_res(req, ret, 0); 544 return IOU_OK; 545 } 546 547 static int io_recvmsg_mshot_prep(struct io_kiocb *req, 548 struct io_async_msghdr *iomsg, 549 int namelen, size_t controllen) 550 { 551 if ((req->flags & (REQ_F_APOLL_MULTISHOT|REQ_F_BUFFER_SELECT)) == 552 (REQ_F_APOLL_MULTISHOT|REQ_F_BUFFER_SELECT)) { 553 int hdr; 554 555 if (unlikely(namelen < 0)) 556 return -EOVERFLOW; 557 if (check_add_overflow(sizeof(struct io_uring_recvmsg_out), 558 namelen, &hdr)) 559 return -EOVERFLOW; 560 if (check_add_overflow(hdr, controllen, &hdr)) 561 return -EOVERFLOW; 562 563 iomsg->namelen = namelen; 564 iomsg->controllen = controllen; 565 return 0; 566 } 567 568 return 0; 569 } 570 571 static int io_recvmsg_copy_hdr(struct io_kiocb *req, 572 struct io_async_msghdr *iomsg) 573 { 574 struct user_msghdr msg; 575 int ret; 576 577 iomsg->msg.msg_name = &iomsg->addr; 578 iomsg->msg.msg_iter.nr_segs = 0; 579 580 #ifdef CONFIG_COMPAT 581 if (unlikely(req->ctx->compat)) { 582 struct compat_msghdr cmsg; 583 584 ret = io_compat_msg_copy_hdr(req, iomsg, &cmsg, ITER_DEST); 585 if (unlikely(ret)) 586 return ret; 587 588 ret = __get_compat_msghdr(&iomsg->msg, &cmsg, &iomsg->uaddr); 589 if (unlikely(ret)) 590 return ret; 591 592 return io_recvmsg_mshot_prep(req, iomsg, cmsg.msg_namelen, 593 cmsg.msg_controllen); 594 } 595 #endif 596 597 ret = io_msg_copy_hdr(req, iomsg, &msg, ITER_DEST); 598 if (unlikely(ret)) 599 return ret; 600 601 ret = __copy_msghdr(&iomsg->msg, &msg, &iomsg->uaddr); 602 if (unlikely(ret)) 603 return ret; 604 605 return io_recvmsg_mshot_prep(req, iomsg, msg.msg_namelen, 606 msg.msg_controllen); 607 } 608 609 int io_recvmsg_prep_async(struct io_kiocb *req) 610 { 611 struct io_async_msghdr *iomsg; 612 int ret; 613 614 if (!io_msg_alloc_async_prep(req)) 615 return -ENOMEM; 616 iomsg = req->async_data; 617 ret = io_recvmsg_copy_hdr(req, iomsg); 618 if (!ret) 619 req->flags |= REQ_F_NEED_CLEANUP; 620 return ret; 621 } 622 623 #define RECVMSG_FLAGS (IORING_RECVSEND_POLL_FIRST | IORING_RECV_MULTISHOT) 624 625 int io_recvmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) 626 { 627 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 628 629 sr->done_io = 0; 630 631 if (unlikely(sqe->file_index || sqe->addr2)) 632 return -EINVAL; 633 634 sr->umsg = u64_to_user_ptr(READ_ONCE(sqe->addr)); 635 sr->len = READ_ONCE(sqe->len); 636 sr->flags = READ_ONCE(sqe->ioprio); 637 if (sr->flags & ~(RECVMSG_FLAGS)) 638 return -EINVAL; 639 sr->msg_flags = READ_ONCE(sqe->msg_flags); 640 if (sr->msg_flags & MSG_DONTWAIT) 641 req->flags |= REQ_F_NOWAIT; 642 if (sr->msg_flags & MSG_ERRQUEUE) 643 req->flags |= REQ_F_CLEAR_POLLIN; 644 if (sr->flags & IORING_RECV_MULTISHOT) { 645 if (!(req->flags & REQ_F_BUFFER_SELECT)) 646 return -EINVAL; 647 if (sr->msg_flags & MSG_WAITALL) 648 return -EINVAL; 649 if (req->opcode == IORING_OP_RECV && sr->len) 650 return -EINVAL; 651 req->flags |= REQ_F_APOLL_MULTISHOT; 652 /* 653 * Store the buffer group for this multishot receive separately, 654 * as if we end up doing an io-wq based issue that selects a 655 * buffer, it has to be committed immediately and that will 656 * clear ->buf_list. This means we lose the link to the buffer 657 * list, and the eventual buffer put on completion then cannot 658 * restore it. 659 */ 660 sr->buf_group = req->buf_index; 661 } 662 663 #ifdef CONFIG_COMPAT 664 if (req->ctx->compat) 665 sr->msg_flags |= MSG_CMSG_COMPAT; 666 #endif 667 sr->nr_multishot_loops = 0; 668 return 0; 669 } 670 671 static inline void io_recv_prep_retry(struct io_kiocb *req) 672 { 673 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 674 675 req->flags &= ~REQ_F_BL_EMPTY; 676 sr->done_io = 0; 677 sr->len = 0; /* get from the provided buffer */ 678 req->buf_index = sr->buf_group; 679 } 680 681 /* 682 * Finishes io_recv and io_recvmsg. 683 * 684 * Returns true if it is actually finished, or false if it should run 685 * again (for multishot). 686 */ 687 static inline bool io_recv_finish(struct io_kiocb *req, int *ret, 688 struct msghdr *msg, bool mshot_finished, 689 unsigned issue_flags) 690 { 691 unsigned int cflags; 692 693 cflags = io_put_kbuf(req, issue_flags); 694 if (msg->msg_inq > 0) 695 cflags |= IORING_CQE_F_SOCK_NONEMPTY; 696 697 /* 698 * Fill CQE for this receive and see if we should keep trying to 699 * receive from this socket. 700 */ 701 if ((req->flags & REQ_F_APOLL_MULTISHOT) && !mshot_finished && 702 io_fill_cqe_req_aux(req, issue_flags & IO_URING_F_COMPLETE_DEFER, 703 *ret, cflags | IORING_CQE_F_MORE)) { 704 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 705 int mshot_retry_ret = IOU_ISSUE_SKIP_COMPLETE; 706 707 io_recv_prep_retry(req); 708 /* Known not-empty or unknown state, retry */ 709 if (cflags & IORING_CQE_F_SOCK_NONEMPTY || msg->msg_inq < 0) { 710 if (sr->nr_multishot_loops++ < MULTISHOT_MAX_RETRY) 711 return false; 712 /* mshot retries exceeded, force a requeue */ 713 sr->nr_multishot_loops = 0; 714 mshot_retry_ret = IOU_REQUEUE; 715 } 716 if (issue_flags & IO_URING_F_MULTISHOT) 717 *ret = mshot_retry_ret; 718 else 719 *ret = -EAGAIN; 720 return true; 721 } 722 723 /* Finish the request / stop multishot. */ 724 io_req_set_res(req, *ret, cflags); 725 726 if (issue_flags & IO_URING_F_MULTISHOT) 727 *ret = IOU_STOP_MULTISHOT; 728 else 729 *ret = IOU_OK; 730 return true; 731 } 732 733 static int io_recvmsg_prep_multishot(struct io_async_msghdr *kmsg, 734 struct io_sr_msg *sr, void __user **buf, 735 size_t *len) 736 { 737 unsigned long ubuf = (unsigned long) *buf; 738 unsigned long hdr; 739 740 hdr = sizeof(struct io_uring_recvmsg_out) + kmsg->namelen + 741 kmsg->controllen; 742 if (*len < hdr) 743 return -EFAULT; 744 745 if (kmsg->controllen) { 746 unsigned long control = ubuf + hdr - kmsg->controllen; 747 748 kmsg->msg.msg_control_user = (void __user *) control; 749 kmsg->msg.msg_controllen = kmsg->controllen; 750 } 751 752 sr->buf = *buf; /* stash for later copy */ 753 *buf = (void __user *) (ubuf + hdr); 754 kmsg->payloadlen = *len = *len - hdr; 755 return 0; 756 } 757 758 struct io_recvmsg_multishot_hdr { 759 struct io_uring_recvmsg_out msg; 760 struct sockaddr_storage addr; 761 }; 762 763 static int io_recvmsg_multishot(struct socket *sock, struct io_sr_msg *io, 764 struct io_async_msghdr *kmsg, 765 unsigned int flags, bool *finished) 766 { 767 int err; 768 int copy_len; 769 struct io_recvmsg_multishot_hdr hdr; 770 771 if (kmsg->namelen) 772 kmsg->msg.msg_name = &hdr.addr; 773 kmsg->msg.msg_flags = flags & (MSG_CMSG_CLOEXEC|MSG_CMSG_COMPAT); 774 kmsg->msg.msg_namelen = 0; 775 776 if (sock->file->f_flags & O_NONBLOCK) 777 flags |= MSG_DONTWAIT; 778 779 err = sock_recvmsg(sock, &kmsg->msg, flags); 780 *finished = err <= 0; 781 if (err < 0) 782 return err; 783 784 hdr.msg = (struct io_uring_recvmsg_out) { 785 .controllen = kmsg->controllen - kmsg->msg.msg_controllen, 786 .flags = kmsg->msg.msg_flags & ~MSG_CMSG_COMPAT 787 }; 788 789 hdr.msg.payloadlen = err; 790 if (err > kmsg->payloadlen) 791 err = kmsg->payloadlen; 792 793 copy_len = sizeof(struct io_uring_recvmsg_out); 794 if (kmsg->msg.msg_namelen > kmsg->namelen) 795 copy_len += kmsg->namelen; 796 else 797 copy_len += kmsg->msg.msg_namelen; 798 799 /* 800 * "fromlen shall refer to the value before truncation.." 801 * 1003.1g 802 */ 803 hdr.msg.namelen = kmsg->msg.msg_namelen; 804 805 /* ensure that there is no gap between hdr and sockaddr_storage */ 806 BUILD_BUG_ON(offsetof(struct io_recvmsg_multishot_hdr, addr) != 807 sizeof(struct io_uring_recvmsg_out)); 808 if (copy_to_user(io->buf, &hdr, copy_len)) { 809 *finished = true; 810 return -EFAULT; 811 } 812 813 return sizeof(struct io_uring_recvmsg_out) + kmsg->namelen + 814 kmsg->controllen + err; 815 } 816 817 int io_recvmsg(struct io_kiocb *req, unsigned int issue_flags) 818 { 819 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 820 struct io_async_msghdr iomsg, *kmsg; 821 struct socket *sock; 822 unsigned flags; 823 int ret, min_ret = 0; 824 bool force_nonblock = issue_flags & IO_URING_F_NONBLOCK; 825 bool mshot_finished = true; 826 827 sock = sock_from_file(req->file); 828 if (unlikely(!sock)) 829 return -ENOTSOCK; 830 831 if (req_has_async_data(req)) { 832 kmsg = req->async_data; 833 } else { 834 ret = io_recvmsg_copy_hdr(req, &iomsg); 835 if (ret) 836 return ret; 837 kmsg = &iomsg; 838 } 839 840 if (!(req->flags & REQ_F_POLLED) && 841 (sr->flags & IORING_RECVSEND_POLL_FIRST)) 842 return io_setup_async_msg(req, kmsg, issue_flags); 843 844 flags = sr->msg_flags; 845 if (force_nonblock) 846 flags |= MSG_DONTWAIT; 847 848 retry_multishot: 849 if (io_do_buffer_select(req)) { 850 void __user *buf; 851 size_t len = sr->len; 852 853 buf = io_buffer_select(req, &len, issue_flags); 854 if (!buf) 855 return -ENOBUFS; 856 857 if (req->flags & REQ_F_APOLL_MULTISHOT) { 858 ret = io_recvmsg_prep_multishot(kmsg, sr, &buf, &len); 859 if (ret) { 860 io_kbuf_recycle(req, issue_flags); 861 return ret; 862 } 863 } 864 865 iov_iter_ubuf(&kmsg->msg.msg_iter, ITER_DEST, buf, len); 866 } 867 868 kmsg->msg.msg_get_inq = 1; 869 kmsg->msg.msg_inq = -1; 870 if (req->flags & REQ_F_APOLL_MULTISHOT) { 871 ret = io_recvmsg_multishot(sock, sr, kmsg, flags, 872 &mshot_finished); 873 } else { 874 /* disable partial retry for recvmsg with cmsg attached */ 875 if (flags & MSG_WAITALL && !kmsg->msg.msg_controllen) 876 min_ret = iov_iter_count(&kmsg->msg.msg_iter); 877 878 ret = __sys_recvmsg_sock(sock, &kmsg->msg, sr->umsg, 879 kmsg->uaddr, flags); 880 } 881 882 if (ret < min_ret) { 883 if (ret == -EAGAIN && force_nonblock) { 884 ret = io_setup_async_msg(req, kmsg, issue_flags); 885 if (ret == -EAGAIN && (issue_flags & IO_URING_F_MULTISHOT)) { 886 io_kbuf_recycle(req, issue_flags); 887 return IOU_ISSUE_SKIP_COMPLETE; 888 } 889 return ret; 890 } 891 if (ret > 0 && io_net_retry(sock, flags)) { 892 sr->done_io += ret; 893 req->flags |= REQ_F_BL_NO_RECYCLE; 894 return io_setup_async_msg(req, kmsg, issue_flags); 895 } 896 if (ret == -ERESTARTSYS) 897 ret = -EINTR; 898 req_set_fail(req); 899 } else if ((flags & MSG_WAITALL) && (kmsg->msg.msg_flags & (MSG_TRUNC | MSG_CTRUNC))) { 900 req_set_fail(req); 901 } 902 903 if (ret > 0) 904 ret += sr->done_io; 905 else if (sr->done_io) 906 ret = sr->done_io; 907 else 908 io_kbuf_recycle(req, issue_flags); 909 910 if (!io_recv_finish(req, &ret, &kmsg->msg, mshot_finished, issue_flags)) 911 goto retry_multishot; 912 913 if (mshot_finished) 914 io_req_msg_cleanup(req, kmsg, issue_flags); 915 else if (ret == -EAGAIN) 916 return io_setup_async_msg(req, kmsg, issue_flags); 917 918 return ret; 919 } 920 921 int io_recv(struct io_kiocb *req, unsigned int issue_flags) 922 { 923 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 924 struct msghdr msg; 925 struct socket *sock; 926 unsigned flags; 927 int ret, min_ret = 0; 928 bool force_nonblock = issue_flags & IO_URING_F_NONBLOCK; 929 size_t len = sr->len; 930 931 if (!(req->flags & REQ_F_POLLED) && 932 (sr->flags & IORING_RECVSEND_POLL_FIRST)) 933 return -EAGAIN; 934 935 sock = sock_from_file(req->file); 936 if (unlikely(!sock)) 937 return -ENOTSOCK; 938 939 msg.msg_name = NULL; 940 msg.msg_namelen = 0; 941 msg.msg_control = NULL; 942 msg.msg_get_inq = 1; 943 msg.msg_controllen = 0; 944 msg.msg_iocb = NULL; 945 msg.msg_ubuf = NULL; 946 947 flags = sr->msg_flags; 948 if (force_nonblock) 949 flags |= MSG_DONTWAIT; 950 951 retry_multishot: 952 if (io_do_buffer_select(req)) { 953 void __user *buf; 954 955 buf = io_buffer_select(req, &len, issue_flags); 956 if (!buf) 957 return -ENOBUFS; 958 sr->buf = buf; 959 sr->len = len; 960 } 961 962 ret = import_ubuf(ITER_DEST, sr->buf, len, &msg.msg_iter); 963 if (unlikely(ret)) 964 goto out_free; 965 966 msg.msg_inq = -1; 967 msg.msg_flags = 0; 968 969 if (flags & MSG_WAITALL) 970 min_ret = iov_iter_count(&msg.msg_iter); 971 972 ret = sock_recvmsg(sock, &msg, flags); 973 if (ret < min_ret) { 974 if (ret == -EAGAIN && force_nonblock) { 975 if (issue_flags & IO_URING_F_MULTISHOT) { 976 io_kbuf_recycle(req, issue_flags); 977 return IOU_ISSUE_SKIP_COMPLETE; 978 } 979 980 return -EAGAIN; 981 } 982 if (ret > 0 && io_net_retry(sock, flags)) { 983 sr->len -= ret; 984 sr->buf += ret; 985 sr->done_io += ret; 986 req->flags |= REQ_F_BL_NO_RECYCLE; 987 return -EAGAIN; 988 } 989 if (ret == -ERESTARTSYS) 990 ret = -EINTR; 991 req_set_fail(req); 992 } else if ((flags & MSG_WAITALL) && (msg.msg_flags & (MSG_TRUNC | MSG_CTRUNC))) { 993 out_free: 994 req_set_fail(req); 995 } 996 997 if (ret > 0) 998 ret += sr->done_io; 999 else if (sr->done_io) 1000 ret = sr->done_io; 1001 else 1002 io_kbuf_recycle(req, issue_flags); 1003 1004 if (!io_recv_finish(req, &ret, &msg, ret <= 0, issue_flags)) 1005 goto retry_multishot; 1006 1007 return ret; 1008 } 1009 1010 void io_send_zc_cleanup(struct io_kiocb *req) 1011 { 1012 struct io_sr_msg *zc = io_kiocb_to_cmd(req, struct io_sr_msg); 1013 struct io_async_msghdr *io; 1014 1015 if (req_has_async_data(req)) { 1016 io = req->async_data; 1017 /* might be ->fast_iov if *msg_copy_hdr failed */ 1018 if (io->free_iov != io->fast_iov) 1019 kfree(io->free_iov); 1020 } 1021 if (zc->notif) { 1022 io_notif_flush(zc->notif); 1023 zc->notif = NULL; 1024 } 1025 } 1026 1027 #define IO_ZC_FLAGS_COMMON (IORING_RECVSEND_POLL_FIRST | IORING_RECVSEND_FIXED_BUF) 1028 #define IO_ZC_FLAGS_VALID (IO_ZC_FLAGS_COMMON | IORING_SEND_ZC_REPORT_USAGE) 1029 1030 int io_send_zc_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) 1031 { 1032 struct io_sr_msg *zc = io_kiocb_to_cmd(req, struct io_sr_msg); 1033 struct io_ring_ctx *ctx = req->ctx; 1034 struct io_kiocb *notif; 1035 1036 zc->done_io = 0; 1037 1038 if (unlikely(READ_ONCE(sqe->__pad2[0]) || READ_ONCE(sqe->addr3))) 1039 return -EINVAL; 1040 /* we don't support IOSQE_CQE_SKIP_SUCCESS just yet */ 1041 if (req->flags & REQ_F_CQE_SKIP) 1042 return -EINVAL; 1043 1044 notif = zc->notif = io_alloc_notif(ctx); 1045 if (!notif) 1046 return -ENOMEM; 1047 notif->cqe.user_data = req->cqe.user_data; 1048 notif->cqe.res = 0; 1049 notif->cqe.flags = IORING_CQE_F_NOTIF; 1050 req->flags |= REQ_F_NEED_CLEANUP; 1051 1052 zc->flags = READ_ONCE(sqe->ioprio); 1053 if (unlikely(zc->flags & ~IO_ZC_FLAGS_COMMON)) { 1054 if (zc->flags & ~IO_ZC_FLAGS_VALID) 1055 return -EINVAL; 1056 if (zc->flags & IORING_SEND_ZC_REPORT_USAGE) { 1057 io_notif_set_extended(notif); 1058 io_notif_to_data(notif)->zc_report = true; 1059 } 1060 } 1061 1062 if (zc->flags & IORING_RECVSEND_FIXED_BUF) { 1063 unsigned idx = READ_ONCE(sqe->buf_index); 1064 1065 if (unlikely(idx >= ctx->nr_user_bufs)) 1066 return -EFAULT; 1067 idx = array_index_nospec(idx, ctx->nr_user_bufs); 1068 req->imu = READ_ONCE(ctx->user_bufs[idx]); 1069 io_req_set_rsrc_node(notif, ctx, 0); 1070 } 1071 1072 if (req->opcode == IORING_OP_SEND_ZC) { 1073 if (READ_ONCE(sqe->__pad3[0])) 1074 return -EINVAL; 1075 zc->addr = u64_to_user_ptr(READ_ONCE(sqe->addr2)); 1076 zc->addr_len = READ_ONCE(sqe->addr_len); 1077 } else { 1078 if (unlikely(sqe->addr2 || sqe->file_index)) 1079 return -EINVAL; 1080 if (unlikely(zc->flags & IORING_RECVSEND_FIXED_BUF)) 1081 return -EINVAL; 1082 } 1083 1084 zc->buf = u64_to_user_ptr(READ_ONCE(sqe->addr)); 1085 zc->len = READ_ONCE(sqe->len); 1086 zc->msg_flags = READ_ONCE(sqe->msg_flags) | MSG_NOSIGNAL; 1087 if (zc->msg_flags & MSG_DONTWAIT) 1088 req->flags |= REQ_F_NOWAIT; 1089 1090 #ifdef CONFIG_COMPAT 1091 if (req->ctx->compat) 1092 zc->msg_flags |= MSG_CMSG_COMPAT; 1093 #endif 1094 return 0; 1095 } 1096 1097 static int io_sg_from_iter_iovec(struct sock *sk, struct sk_buff *skb, 1098 struct iov_iter *from, size_t length) 1099 { 1100 skb_zcopy_downgrade_managed(skb); 1101 return __zerocopy_sg_from_iter(NULL, sk, skb, from, length); 1102 } 1103 1104 static int io_sg_from_iter(struct sock *sk, struct sk_buff *skb, 1105 struct iov_iter *from, size_t length) 1106 { 1107 struct skb_shared_info *shinfo = skb_shinfo(skb); 1108 int frag = shinfo->nr_frags; 1109 int ret = 0; 1110 struct bvec_iter bi; 1111 ssize_t copied = 0; 1112 unsigned long truesize = 0; 1113 1114 if (!frag) 1115 shinfo->flags |= SKBFL_MANAGED_FRAG_REFS; 1116 else if (unlikely(!skb_zcopy_managed(skb))) 1117 return __zerocopy_sg_from_iter(NULL, sk, skb, from, length); 1118 1119 bi.bi_size = min(from->count, length); 1120 bi.bi_bvec_done = from->iov_offset; 1121 bi.bi_idx = 0; 1122 1123 while (bi.bi_size && frag < MAX_SKB_FRAGS) { 1124 struct bio_vec v = mp_bvec_iter_bvec(from->bvec, bi); 1125 1126 copied += v.bv_len; 1127 truesize += PAGE_ALIGN(v.bv_len + v.bv_offset); 1128 __skb_fill_page_desc_noacc(shinfo, frag++, v.bv_page, 1129 v.bv_offset, v.bv_len); 1130 bvec_iter_advance_single(from->bvec, &bi, v.bv_len); 1131 } 1132 if (bi.bi_size) 1133 ret = -EMSGSIZE; 1134 1135 shinfo->nr_frags = frag; 1136 from->bvec += bi.bi_idx; 1137 from->nr_segs -= bi.bi_idx; 1138 from->count -= copied; 1139 from->iov_offset = bi.bi_bvec_done; 1140 1141 skb->data_len += copied; 1142 skb->len += copied; 1143 skb->truesize += truesize; 1144 1145 if (sk && sk->sk_type == SOCK_STREAM) { 1146 sk_wmem_queued_add(sk, truesize); 1147 if (!skb_zcopy_pure(skb)) 1148 sk_mem_charge(sk, truesize); 1149 } else { 1150 refcount_add(truesize, &skb->sk->sk_wmem_alloc); 1151 } 1152 return ret; 1153 } 1154 1155 int io_send_zc(struct io_kiocb *req, unsigned int issue_flags) 1156 { 1157 struct sockaddr_storage __address; 1158 struct io_sr_msg *zc = io_kiocb_to_cmd(req, struct io_sr_msg); 1159 struct msghdr msg; 1160 struct socket *sock; 1161 unsigned msg_flags; 1162 int ret, min_ret = 0; 1163 1164 sock = sock_from_file(req->file); 1165 if (unlikely(!sock)) 1166 return -ENOTSOCK; 1167 if (!test_bit(SOCK_SUPPORT_ZC, &sock->flags)) 1168 return -EOPNOTSUPP; 1169 1170 msg.msg_name = NULL; 1171 msg.msg_control = NULL; 1172 msg.msg_controllen = 0; 1173 msg.msg_namelen = 0; 1174 1175 if (zc->addr) { 1176 if (req_has_async_data(req)) { 1177 struct io_async_msghdr *io = req->async_data; 1178 1179 msg.msg_name = &io->addr; 1180 } else { 1181 ret = move_addr_to_kernel(zc->addr, zc->addr_len, &__address); 1182 if (unlikely(ret < 0)) 1183 return ret; 1184 msg.msg_name = (struct sockaddr *)&__address; 1185 } 1186 msg.msg_namelen = zc->addr_len; 1187 } 1188 1189 if (!(req->flags & REQ_F_POLLED) && 1190 (zc->flags & IORING_RECVSEND_POLL_FIRST)) 1191 return io_setup_async_addr(req, &__address, issue_flags); 1192 1193 if (zc->flags & IORING_RECVSEND_FIXED_BUF) { 1194 ret = io_import_fixed(ITER_SOURCE, &msg.msg_iter, req->imu, 1195 (u64)(uintptr_t)zc->buf, zc->len); 1196 if (unlikely(ret)) 1197 return ret; 1198 msg.sg_from_iter = io_sg_from_iter; 1199 } else { 1200 io_notif_set_extended(zc->notif); 1201 ret = import_ubuf(ITER_SOURCE, zc->buf, zc->len, &msg.msg_iter); 1202 if (unlikely(ret)) 1203 return ret; 1204 ret = io_notif_account_mem(zc->notif, zc->len); 1205 if (unlikely(ret)) 1206 return ret; 1207 msg.sg_from_iter = io_sg_from_iter_iovec; 1208 } 1209 1210 msg_flags = zc->msg_flags | MSG_ZEROCOPY; 1211 if (issue_flags & IO_URING_F_NONBLOCK) 1212 msg_flags |= MSG_DONTWAIT; 1213 if (msg_flags & MSG_WAITALL) 1214 min_ret = iov_iter_count(&msg.msg_iter); 1215 msg_flags &= ~MSG_INTERNAL_SENDMSG_FLAGS; 1216 1217 msg.msg_flags = msg_flags; 1218 msg.msg_ubuf = &io_notif_to_data(zc->notif)->uarg; 1219 ret = sock_sendmsg(sock, &msg); 1220 1221 if (unlikely(ret < min_ret)) { 1222 if (ret == -EAGAIN && (issue_flags & IO_URING_F_NONBLOCK)) 1223 return io_setup_async_addr(req, &__address, issue_flags); 1224 1225 if (ret > 0 && io_net_retry(sock, msg.msg_flags)) { 1226 zc->len -= ret; 1227 zc->buf += ret; 1228 zc->done_io += ret; 1229 req->flags |= REQ_F_BL_NO_RECYCLE; 1230 return io_setup_async_addr(req, &__address, issue_flags); 1231 } 1232 if (ret == -ERESTARTSYS) 1233 ret = -EINTR; 1234 req_set_fail(req); 1235 } 1236 1237 if (ret >= 0) 1238 ret += zc->done_io; 1239 else if (zc->done_io) 1240 ret = zc->done_io; 1241 1242 /* 1243 * If we're in io-wq we can't rely on tw ordering guarantees, defer 1244 * flushing notif to io_send_zc_cleanup() 1245 */ 1246 if (!(issue_flags & IO_URING_F_UNLOCKED)) { 1247 io_notif_flush(zc->notif); 1248 req->flags &= ~REQ_F_NEED_CLEANUP; 1249 } 1250 io_req_set_res(req, ret, IORING_CQE_F_MORE); 1251 return IOU_OK; 1252 } 1253 1254 int io_sendmsg_zc(struct io_kiocb *req, unsigned int issue_flags) 1255 { 1256 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 1257 struct io_async_msghdr iomsg, *kmsg; 1258 struct socket *sock; 1259 unsigned flags; 1260 int ret, min_ret = 0; 1261 1262 io_notif_set_extended(sr->notif); 1263 1264 sock = sock_from_file(req->file); 1265 if (unlikely(!sock)) 1266 return -ENOTSOCK; 1267 if (!test_bit(SOCK_SUPPORT_ZC, &sock->flags)) 1268 return -EOPNOTSUPP; 1269 1270 if (req_has_async_data(req)) { 1271 kmsg = req->async_data; 1272 } else { 1273 ret = io_sendmsg_copy_hdr(req, &iomsg); 1274 if (ret) 1275 return ret; 1276 kmsg = &iomsg; 1277 } 1278 1279 if (!(req->flags & REQ_F_POLLED) && 1280 (sr->flags & IORING_RECVSEND_POLL_FIRST)) 1281 return io_setup_async_msg(req, kmsg, issue_flags); 1282 1283 flags = sr->msg_flags | MSG_ZEROCOPY; 1284 if (issue_flags & IO_URING_F_NONBLOCK) 1285 flags |= MSG_DONTWAIT; 1286 if (flags & MSG_WAITALL) 1287 min_ret = iov_iter_count(&kmsg->msg.msg_iter); 1288 1289 kmsg->msg.msg_ubuf = &io_notif_to_data(sr->notif)->uarg; 1290 kmsg->msg.sg_from_iter = io_sg_from_iter_iovec; 1291 ret = __sys_sendmsg_sock(sock, &kmsg->msg, flags); 1292 1293 if (unlikely(ret < min_ret)) { 1294 if (ret == -EAGAIN && (issue_flags & IO_URING_F_NONBLOCK)) 1295 return io_setup_async_msg(req, kmsg, issue_flags); 1296 1297 if (ret > 0 && io_net_retry(sock, flags)) { 1298 sr->done_io += ret; 1299 req->flags |= REQ_F_BL_NO_RECYCLE; 1300 return io_setup_async_msg(req, kmsg, issue_flags); 1301 } 1302 if (ret == -ERESTARTSYS) 1303 ret = -EINTR; 1304 req_set_fail(req); 1305 } 1306 /* fast path, check for non-NULL to avoid function call */ 1307 if (kmsg->free_iov) { 1308 kfree(kmsg->free_iov); 1309 kmsg->free_iov = NULL; 1310 } 1311 1312 io_netmsg_recycle(req, issue_flags); 1313 if (ret >= 0) 1314 ret += sr->done_io; 1315 else if (sr->done_io) 1316 ret = sr->done_io; 1317 1318 /* 1319 * If we're in io-wq we can't rely on tw ordering guarantees, defer 1320 * flushing notif to io_send_zc_cleanup() 1321 */ 1322 if (!(issue_flags & IO_URING_F_UNLOCKED)) { 1323 io_notif_flush(sr->notif); 1324 req->flags &= ~REQ_F_NEED_CLEANUP; 1325 } 1326 io_req_set_res(req, ret, IORING_CQE_F_MORE); 1327 return IOU_OK; 1328 } 1329 1330 void io_sendrecv_fail(struct io_kiocb *req) 1331 { 1332 struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg); 1333 1334 if (sr->done_io) 1335 req->cqe.res = sr->done_io; 1336 1337 if ((req->flags & REQ_F_NEED_CLEANUP) && 1338 (req->opcode == IORING_OP_SEND_ZC || req->opcode == IORING_OP_SENDMSG_ZC)) 1339 req->cqe.flags |= IORING_CQE_F_MORE; 1340 } 1341 1342 int io_accept_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) 1343 { 1344 struct io_accept *accept = io_kiocb_to_cmd(req, struct io_accept); 1345 unsigned flags; 1346 1347 if (sqe->len || sqe->buf_index) 1348 return -EINVAL; 1349 1350 accept->addr = u64_to_user_ptr(READ_ONCE(sqe->addr)); 1351 accept->addr_len = u64_to_user_ptr(READ_ONCE(sqe->addr2)); 1352 accept->flags = READ_ONCE(sqe->accept_flags); 1353 accept->nofile = rlimit(RLIMIT_NOFILE); 1354 flags = READ_ONCE(sqe->ioprio); 1355 if (flags & ~IORING_ACCEPT_MULTISHOT) 1356 return -EINVAL; 1357 1358 accept->file_slot = READ_ONCE(sqe->file_index); 1359 if (accept->file_slot) { 1360 if (accept->flags & SOCK_CLOEXEC) 1361 return -EINVAL; 1362 if (flags & IORING_ACCEPT_MULTISHOT && 1363 accept->file_slot != IORING_FILE_INDEX_ALLOC) 1364 return -EINVAL; 1365 } 1366 if (accept->flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK)) 1367 return -EINVAL; 1368 if (SOCK_NONBLOCK != O_NONBLOCK && (accept->flags & SOCK_NONBLOCK)) 1369 accept->flags = (accept->flags & ~SOCK_NONBLOCK) | O_NONBLOCK; 1370 if (flags & IORING_ACCEPT_MULTISHOT) 1371 req->flags |= REQ_F_APOLL_MULTISHOT; 1372 return 0; 1373 } 1374 1375 int io_accept(struct io_kiocb *req, unsigned int issue_flags) 1376 { 1377 struct io_accept *accept = io_kiocb_to_cmd(req, struct io_accept); 1378 bool force_nonblock = issue_flags & IO_URING_F_NONBLOCK; 1379 unsigned int file_flags = force_nonblock ? O_NONBLOCK : 0; 1380 bool fixed = !!accept->file_slot; 1381 struct file *file; 1382 int ret, fd; 1383 1384 retry: 1385 if (!fixed) { 1386 fd = __get_unused_fd_flags(accept->flags, accept->nofile); 1387 if (unlikely(fd < 0)) 1388 return fd; 1389 } 1390 file = do_accept(req->file, file_flags, accept->addr, accept->addr_len, 1391 accept->flags); 1392 if (IS_ERR(file)) { 1393 if (!fixed) 1394 put_unused_fd(fd); 1395 ret = PTR_ERR(file); 1396 if (ret == -EAGAIN && force_nonblock) { 1397 /* 1398 * if it's multishot and polled, we don't need to 1399 * return EAGAIN to arm the poll infra since it 1400 * has already been done 1401 */ 1402 if (issue_flags & IO_URING_F_MULTISHOT) 1403 return IOU_ISSUE_SKIP_COMPLETE; 1404 return ret; 1405 } 1406 if (ret == -ERESTARTSYS) 1407 ret = -EINTR; 1408 req_set_fail(req); 1409 } else if (!fixed) { 1410 fd_install(fd, file); 1411 ret = fd; 1412 } else { 1413 ret = io_fixed_fd_install(req, issue_flags, file, 1414 accept->file_slot); 1415 } 1416 1417 if (!(req->flags & REQ_F_APOLL_MULTISHOT)) { 1418 io_req_set_res(req, ret, 0); 1419 return IOU_OK; 1420 } 1421 1422 if (ret < 0) 1423 return ret; 1424 if (io_fill_cqe_req_aux(req, issue_flags & IO_URING_F_COMPLETE_DEFER, 1425 ret, IORING_CQE_F_MORE)) 1426 goto retry; 1427 1428 io_req_set_res(req, ret, 0); 1429 return IOU_STOP_MULTISHOT; 1430 } 1431 1432 int io_socket_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) 1433 { 1434 struct io_socket *sock = io_kiocb_to_cmd(req, struct io_socket); 1435 1436 if (sqe->addr || sqe->rw_flags || sqe->buf_index) 1437 return -EINVAL; 1438 1439 sock->domain = READ_ONCE(sqe->fd); 1440 sock->type = READ_ONCE(sqe->off); 1441 sock->protocol = READ_ONCE(sqe->len); 1442 sock->file_slot = READ_ONCE(sqe->file_index); 1443 sock->nofile = rlimit(RLIMIT_NOFILE); 1444 1445 sock->flags = sock->type & ~SOCK_TYPE_MASK; 1446 if (sock->file_slot && (sock->flags & SOCK_CLOEXEC)) 1447 return -EINVAL; 1448 if (sock->flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK)) 1449 return -EINVAL; 1450 return 0; 1451 } 1452 1453 int io_socket(struct io_kiocb *req, unsigned int issue_flags) 1454 { 1455 struct io_socket *sock = io_kiocb_to_cmd(req, struct io_socket); 1456 bool fixed = !!sock->file_slot; 1457 struct file *file; 1458 int ret, fd; 1459 1460 if (!fixed) { 1461 fd = __get_unused_fd_flags(sock->flags, sock->nofile); 1462 if (unlikely(fd < 0)) 1463 return fd; 1464 } 1465 file = __sys_socket_file(sock->domain, sock->type, sock->protocol); 1466 if (IS_ERR(file)) { 1467 if (!fixed) 1468 put_unused_fd(fd); 1469 ret = PTR_ERR(file); 1470 if (ret == -EAGAIN && (issue_flags & IO_URING_F_NONBLOCK)) 1471 return -EAGAIN; 1472 if (ret == -ERESTARTSYS) 1473 ret = -EINTR; 1474 req_set_fail(req); 1475 } else if (!fixed) { 1476 fd_install(fd, file); 1477 ret = fd; 1478 } else { 1479 ret = io_fixed_fd_install(req, issue_flags, file, 1480 sock->file_slot); 1481 } 1482 io_req_set_res(req, ret, 0); 1483 return IOU_OK; 1484 } 1485 1486 int io_connect_prep_async(struct io_kiocb *req) 1487 { 1488 struct io_async_connect *io = req->async_data; 1489 struct io_connect *conn = io_kiocb_to_cmd(req, struct io_connect); 1490 1491 return move_addr_to_kernel(conn->addr, conn->addr_len, &io->address); 1492 } 1493 1494 int io_connect_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) 1495 { 1496 struct io_connect *conn = io_kiocb_to_cmd(req, struct io_connect); 1497 1498 if (sqe->len || sqe->buf_index || sqe->rw_flags || sqe->splice_fd_in) 1499 return -EINVAL; 1500 1501 conn->addr = u64_to_user_ptr(READ_ONCE(sqe->addr)); 1502 conn->addr_len = READ_ONCE(sqe->addr2); 1503 conn->in_progress = conn->seen_econnaborted = false; 1504 return 0; 1505 } 1506 1507 int io_connect(struct io_kiocb *req, unsigned int issue_flags) 1508 { 1509 struct io_connect *connect = io_kiocb_to_cmd(req, struct io_connect); 1510 struct io_async_connect __io, *io; 1511 unsigned file_flags; 1512 int ret; 1513 bool force_nonblock = issue_flags & IO_URING_F_NONBLOCK; 1514 1515 if (req_has_async_data(req)) { 1516 io = req->async_data; 1517 } else { 1518 ret = move_addr_to_kernel(connect->addr, 1519 connect->addr_len, 1520 &__io.address); 1521 if (ret) 1522 goto out; 1523 io = &__io; 1524 } 1525 1526 file_flags = force_nonblock ? O_NONBLOCK : 0; 1527 1528 ret = __sys_connect_file(req->file, &io->address, 1529 connect->addr_len, file_flags); 1530 if ((ret == -EAGAIN || ret == -EINPROGRESS || ret == -ECONNABORTED) 1531 && force_nonblock) { 1532 if (ret == -EINPROGRESS) { 1533 connect->in_progress = true; 1534 } else if (ret == -ECONNABORTED) { 1535 if (connect->seen_econnaborted) 1536 goto out; 1537 connect->seen_econnaborted = true; 1538 } 1539 if (req_has_async_data(req)) 1540 return -EAGAIN; 1541 if (io_alloc_async_data(req)) { 1542 ret = -ENOMEM; 1543 goto out; 1544 } 1545 memcpy(req->async_data, &__io, sizeof(__io)); 1546 return -EAGAIN; 1547 } 1548 if (connect->in_progress) { 1549 /* 1550 * At least bluetooth will return -EBADFD on a re-connect 1551 * attempt, and it's (supposedly) also valid to get -EISCONN 1552 * which means the previous result is good. For both of these, 1553 * grab the sock_error() and use that for the completion. 1554 */ 1555 if (ret == -EBADFD || ret == -EISCONN) 1556 ret = sock_error(sock_from_file(req->file)->sk); 1557 } 1558 if (ret == -ERESTARTSYS) 1559 ret = -EINTR; 1560 out: 1561 if (ret < 0) 1562 req_set_fail(req); 1563 io_req_set_res(req, ret, 0); 1564 return IOU_OK; 1565 } 1566 1567 void io_netmsg_cache_free(struct io_cache_entry *entry) 1568 { 1569 kfree(container_of(entry, struct io_async_msghdr, cache)); 1570 } 1571 #endif 1572