1// SPDX-License-Identifier: GPL-2.0 2/* Multipath TCP 3 * 4 * Copyright (c) 2020, Red Hat, Inc. 5 */ 6 7#define pr_fmt(fmt) "MPTCP: " fmt 8 9#include <linux/inet.h> 10#include <linux/kernel.h> 11#include <net/tcp.h> 12#include <net/inet_common.h> 13#include <net/netns/generic.h> 14#include <net/mptcp.h> 15#include <net/genetlink.h> 16#include <uapi/linux/mptcp.h> 17 18#include "protocol.h" 19#include "mib.h" 20 21/* forward declaration */ 22static struct genl_family mptcp_genl_family; 23 24static int pm_nl_pernet_id; 25 26struct mptcp_pm_add_entry { 27 struct list_head list; 28 struct mptcp_addr_info addr; 29 u8 retrans_times; 30 struct timer_list add_timer; 31 struct mptcp_sock *sock; 32}; 33 34struct pm_nl_pernet { 35 /* protects pernet updates */ 36 spinlock_t lock; 37 struct list_head local_addr_list; 38 unsigned int addrs; 39 unsigned int stale_loss_cnt; 40 unsigned int add_addr_signal_max; 41 unsigned int add_addr_accept_max; 42 unsigned int local_addr_max; 43 unsigned int subflows_max; 44 unsigned int next_id; 45 DECLARE_BITMAP(id_bitmap, MPTCP_PM_MAX_ADDR_ID + 1); 46}; 47 48#define MPTCP_PM_ADDR_MAX 8 49#define ADD_ADDR_RETRANS_MAX 3 50 51static struct pm_nl_pernet *pm_nl_get_pernet(const struct net *net) 52{ 53 return net_generic(net, pm_nl_pernet_id); 54} 55 56static struct pm_nl_pernet * 57pm_nl_get_pernet_from_msk(const struct mptcp_sock *msk) 58{ 59 return pm_nl_get_pernet(sock_net((struct sock *)msk)); 60} 61 62bool mptcp_addresses_equal(const struct mptcp_addr_info *a, 63 const struct mptcp_addr_info *b, bool use_port) 64{ 65 bool addr_equals = false; 66 67 if (a->family == b->family) { 68 if (a->family == AF_INET) 69 addr_equals = a->addr.s_addr == b->addr.s_addr; 70#if IS_ENABLED(CONFIG_MPTCP_IPV6) 71 else 72 addr_equals = !ipv6_addr_cmp(&a->addr6, &b->addr6); 73 } else if (a->family == AF_INET) { 74 if (ipv6_addr_v4mapped(&b->addr6)) 75 addr_equals = a->addr.s_addr == b->addr6.s6_addr32[3]; 76 } else if (b->family == AF_INET) { 77 if (ipv6_addr_v4mapped(&a->addr6)) 78 addr_equals = a->addr6.s6_addr32[3] == b->addr.s_addr; 79#endif 80 } 81 82 if (!addr_equals) 83 return false; 84 if (!use_port) 85 return true; 86 87 return a->port == b->port; 88} 89 90void mptcp_local_address(const struct sock_common *skc, struct mptcp_addr_info *addr) 91{ 92 addr->family = skc->skc_family; 93 addr->port = htons(skc->skc_num); 94 if (addr->family == AF_INET) 95 addr->addr.s_addr = skc->skc_rcv_saddr; 96#if IS_ENABLED(CONFIG_MPTCP_IPV6) 97 else if (addr->family == AF_INET6) 98 addr->addr6 = skc->skc_v6_rcv_saddr; 99#endif 100} 101 102static void remote_address(const struct sock_common *skc, 103 struct mptcp_addr_info *addr) 104{ 105 addr->family = skc->skc_family; 106 addr->port = skc->skc_dport; 107 if (addr->family == AF_INET) 108 addr->addr.s_addr = skc->skc_daddr; 109#if IS_ENABLED(CONFIG_MPTCP_IPV6) 110 else if (addr->family == AF_INET6) 111 addr->addr6 = skc->skc_v6_daddr; 112#endif 113} 114 115static bool lookup_subflow_by_saddr(const struct list_head *list, 116 const struct mptcp_addr_info *saddr) 117{ 118 struct mptcp_subflow_context *subflow; 119 struct mptcp_addr_info cur; 120 struct sock_common *skc; 121 122 list_for_each_entry(subflow, list, node) { 123 skc = (struct sock_common *)mptcp_subflow_tcp_sock(subflow); 124 125 mptcp_local_address(skc, &cur); 126 if (mptcp_addresses_equal(&cur, saddr, saddr->port)) 127 return true; 128 } 129 130 return false; 131} 132 133static bool lookup_subflow_by_daddr(const struct list_head *list, 134 const struct mptcp_addr_info *daddr) 135{ 136 struct mptcp_subflow_context *subflow; 137 struct mptcp_addr_info cur; 138 struct sock_common *skc; 139 140 list_for_each_entry(subflow, list, node) { 141 skc = (struct sock_common *)mptcp_subflow_tcp_sock(subflow); 142 143 remote_address(skc, &cur); 144 if (mptcp_addresses_equal(&cur, daddr, daddr->port)) 145 return true; 146 } 147 148 return false; 149} 150 151static struct mptcp_pm_addr_entry * 152select_local_address(const struct pm_nl_pernet *pernet, 153 const struct mptcp_sock *msk) 154{ 155 struct mptcp_pm_addr_entry *entry, *ret = NULL; 156 157 msk_owned_by_me(msk); 158 159 rcu_read_lock(); 160 list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) { 161 if (!(entry->flags & MPTCP_PM_ADDR_FLAG_SUBFLOW)) 162 continue; 163 164 if (!test_bit(entry->addr.id, msk->pm.id_avail_bitmap)) 165 continue; 166 167 ret = entry; 168 break; 169 } 170 rcu_read_unlock(); 171 return ret; 172} 173 174static struct mptcp_pm_addr_entry * 175select_signal_address(struct pm_nl_pernet *pernet, const struct mptcp_sock *msk) 176{ 177 struct mptcp_pm_addr_entry *entry, *ret = NULL; 178 179 rcu_read_lock(); 180 /* do not keep any additional per socket state, just signal 181 * the address list in order. 182 * Note: removal from the local address list during the msk life-cycle 183 * can lead to additional addresses not being announced. 184 */ 185 list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) { 186 if (!test_bit(entry->addr.id, msk->pm.id_avail_bitmap)) 187 continue; 188 189 if (!(entry->flags & MPTCP_PM_ADDR_FLAG_SIGNAL)) 190 continue; 191 192 ret = entry; 193 break; 194 } 195 rcu_read_unlock(); 196 return ret; 197} 198 199unsigned int mptcp_pm_get_add_addr_signal_max(const struct mptcp_sock *msk) 200{ 201 const struct pm_nl_pernet *pernet = pm_nl_get_pernet_from_msk(msk); 202 203 return READ_ONCE(pernet->add_addr_signal_max); 204} 205EXPORT_SYMBOL_GPL(mptcp_pm_get_add_addr_signal_max); 206 207unsigned int mptcp_pm_get_add_addr_accept_max(const struct mptcp_sock *msk) 208{ 209 struct pm_nl_pernet *pernet = pm_nl_get_pernet_from_msk(msk); 210 211 return READ_ONCE(pernet->add_addr_accept_max); 212} 213EXPORT_SYMBOL_GPL(mptcp_pm_get_add_addr_accept_max); 214 215unsigned int mptcp_pm_get_subflows_max(const struct mptcp_sock *msk) 216{ 217 struct pm_nl_pernet *pernet = pm_nl_get_pernet_from_msk(msk); 218 219 return READ_ONCE(pernet->subflows_max); 220} 221EXPORT_SYMBOL_GPL(mptcp_pm_get_subflows_max); 222 223unsigned int mptcp_pm_get_local_addr_max(const struct mptcp_sock *msk) 224{ 225 struct pm_nl_pernet *pernet = pm_nl_get_pernet_from_msk(msk); 226 227 return READ_ONCE(pernet->local_addr_max); 228} 229EXPORT_SYMBOL_GPL(mptcp_pm_get_local_addr_max); 230 231bool mptcp_pm_nl_check_work_pending(struct mptcp_sock *msk) 232{ 233 struct pm_nl_pernet *pernet = pm_nl_get_pernet_from_msk(msk); 234 235 if (msk->pm.subflows == mptcp_pm_get_subflows_max(msk) || 236 (find_next_and_bit(pernet->id_bitmap, msk->pm.id_avail_bitmap, 237 MPTCP_PM_MAX_ADDR_ID + 1, 0) == MPTCP_PM_MAX_ADDR_ID + 1)) { 238 WRITE_ONCE(msk->pm.work_pending, false); 239 return false; 240 } 241 return true; 242} 243 244struct mptcp_pm_add_entry * 245mptcp_lookup_anno_list_by_saddr(const struct mptcp_sock *msk, 246 const struct mptcp_addr_info *addr) 247{ 248 struct mptcp_pm_add_entry *entry; 249 250 lockdep_assert_held(&msk->pm.lock); 251 252 list_for_each_entry(entry, &msk->pm.anno_list, list) { 253 if (mptcp_addresses_equal(&entry->addr, addr, true)) 254 return entry; 255 } 256 257 return NULL; 258} 259 260bool mptcp_pm_sport_in_anno_list(struct mptcp_sock *msk, const struct sock *sk) 261{ 262 struct mptcp_pm_add_entry *entry; 263 struct mptcp_addr_info saddr; 264 bool ret = false; 265 266 mptcp_local_address((struct sock_common *)sk, &saddr); 267 268 spin_lock_bh(&msk->pm.lock); 269 list_for_each_entry(entry, &msk->pm.anno_list, list) { 270 if (mptcp_addresses_equal(&entry->addr, &saddr, true)) { 271 ret = true; 272 goto out; 273 } 274 } 275 276out: 277 spin_unlock_bh(&msk->pm.lock); 278 return ret; 279} 280 281static void mptcp_pm_add_timer(struct timer_list *timer) 282{ 283 struct mptcp_pm_add_entry *entry = from_timer(entry, timer, add_timer); 284 struct mptcp_sock *msk = entry->sock; 285 struct sock *sk = (struct sock *)msk; 286 287 pr_debug("msk=%p", msk); 288 289 if (!msk) 290 return; 291 292 if (inet_sk_state_load(sk) == TCP_CLOSE) 293 return; 294 295 if (!entry->addr.id) 296 return; 297 298 if (mptcp_pm_should_add_signal_addr(msk)) { 299 sk_reset_timer(sk, timer, jiffies + TCP_RTO_MAX / 8); 300 goto out; 301 } 302 303 spin_lock_bh(&msk->pm.lock); 304 305 if (!mptcp_pm_should_add_signal_addr(msk)) { 306 pr_debug("retransmit ADD_ADDR id=%d", entry->addr.id); 307 mptcp_pm_announce_addr(msk, &entry->addr, false); 308 mptcp_pm_add_addr_send_ack(msk); 309 entry->retrans_times++; 310 } 311 312 if (entry->retrans_times < ADD_ADDR_RETRANS_MAX) 313 sk_reset_timer(sk, timer, 314 jiffies + mptcp_get_add_addr_timeout(sock_net(sk))); 315 316 spin_unlock_bh(&msk->pm.lock); 317 318 if (entry->retrans_times == ADD_ADDR_RETRANS_MAX) 319 mptcp_pm_subflow_established(msk); 320 321out: 322 __sock_put(sk); 323} 324 325struct mptcp_pm_add_entry * 326mptcp_pm_del_add_timer(struct mptcp_sock *msk, 327 const struct mptcp_addr_info *addr, bool check_id) 328{ 329 struct mptcp_pm_add_entry *entry; 330 struct sock *sk = (struct sock *)msk; 331 332 spin_lock_bh(&msk->pm.lock); 333 entry = mptcp_lookup_anno_list_by_saddr(msk, addr); 334 if (entry && (!check_id || entry->addr.id == addr->id)) 335 entry->retrans_times = ADD_ADDR_RETRANS_MAX; 336 spin_unlock_bh(&msk->pm.lock); 337 338 if (entry && (!check_id || entry->addr.id == addr->id)) 339 sk_stop_timer_sync(sk, &entry->add_timer); 340 341 return entry; 342} 343 344bool mptcp_pm_alloc_anno_list(struct mptcp_sock *msk, 345 const struct mptcp_addr_info *addr) 346{ 347 struct mptcp_pm_add_entry *add_entry = NULL; 348 struct sock *sk = (struct sock *)msk; 349 struct net *net = sock_net(sk); 350 351 lockdep_assert_held(&msk->pm.lock); 352 353 add_entry = mptcp_lookup_anno_list_by_saddr(msk, addr); 354 355 if (add_entry) { 356 if (mptcp_pm_is_kernel(msk)) 357 return false; 358 359 sk_reset_timer(sk, &add_entry->add_timer, 360 jiffies + mptcp_get_add_addr_timeout(net)); 361 return true; 362 } 363 364 add_entry = kmalloc(sizeof(*add_entry), GFP_ATOMIC); 365 if (!add_entry) 366 return false; 367 368 list_add(&add_entry->list, &msk->pm.anno_list); 369 370 add_entry->addr = *addr; 371 add_entry->sock = msk; 372 add_entry->retrans_times = 0; 373 374 timer_setup(&add_entry->add_timer, mptcp_pm_add_timer, 0); 375 sk_reset_timer(sk, &add_entry->add_timer, 376 jiffies + mptcp_get_add_addr_timeout(net)); 377 378 return true; 379} 380 381void mptcp_pm_free_anno_list(struct mptcp_sock *msk) 382{ 383 struct mptcp_pm_add_entry *entry, *tmp; 384 struct sock *sk = (struct sock *)msk; 385 LIST_HEAD(free_list); 386 387 pr_debug("msk=%p", msk); 388 389 spin_lock_bh(&msk->pm.lock); 390 list_splice_init(&msk->pm.anno_list, &free_list); 391 spin_unlock_bh(&msk->pm.lock); 392 393 list_for_each_entry_safe(entry, tmp, &free_list, list) { 394 sk_stop_timer_sync(sk, &entry->add_timer); 395 kfree(entry); 396 } 397} 398 399/* Fill all the remote addresses into the array addrs[], 400 * and return the array size. 401 */ 402static unsigned int fill_remote_addresses_vec(struct mptcp_sock *msk, 403 struct mptcp_addr_info *local, 404 bool fullmesh, 405 struct mptcp_addr_info *addrs) 406{ 407 bool deny_id0 = READ_ONCE(msk->pm.remote_deny_join_id0); 408 struct sock *sk = (struct sock *)msk, *ssk; 409 struct mptcp_subflow_context *subflow; 410 struct mptcp_addr_info remote = { 0 }; 411 unsigned int subflows_max; 412 int i = 0; 413 414 subflows_max = mptcp_pm_get_subflows_max(msk); 415 remote_address((struct sock_common *)sk, &remote); 416 417 /* Non-fullmesh endpoint, fill in the single entry 418 * corresponding to the primary MPC subflow remote address 419 */ 420 if (!fullmesh) { 421 if (deny_id0) 422 return 0; 423 424 if (!mptcp_pm_addr_families_match(sk, local, &remote)) 425 return 0; 426 427 msk->pm.subflows++; 428 addrs[i++] = remote; 429 } else { 430 DECLARE_BITMAP(unavail_id, MPTCP_PM_MAX_ADDR_ID + 1); 431 432 /* Forbid creation of new subflows matching existing 433 * ones, possibly already created by incoming ADD_ADDR 434 */ 435 bitmap_zero(unavail_id, MPTCP_PM_MAX_ADDR_ID + 1); 436 mptcp_for_each_subflow(msk, subflow) 437 if (READ_ONCE(subflow->local_id) == local->id) 438 __set_bit(subflow->remote_id, unavail_id); 439 440 mptcp_for_each_subflow(msk, subflow) { 441 ssk = mptcp_subflow_tcp_sock(subflow); 442 remote_address((struct sock_common *)ssk, &addrs[i]); 443 addrs[i].id = READ_ONCE(subflow->remote_id); 444 if (deny_id0 && !addrs[i].id) 445 continue; 446 447 if (test_bit(addrs[i].id, unavail_id)) 448 continue; 449 450 if (!mptcp_pm_addr_families_match(sk, local, &addrs[i])) 451 continue; 452 453 if (msk->pm.subflows < subflows_max) { 454 /* forbid creating multiple address towards 455 * this id 456 */ 457 __set_bit(addrs[i].id, unavail_id); 458 msk->pm.subflows++; 459 i++; 460 } 461 } 462 } 463 464 return i; 465} 466 467static void __mptcp_pm_send_ack(struct mptcp_sock *msk, struct mptcp_subflow_context *subflow, 468 bool prio, bool backup) 469{ 470 struct sock *ssk = mptcp_subflow_tcp_sock(subflow); 471 bool slow; 472 473 pr_debug("send ack for %s", 474 prio ? "mp_prio" : (mptcp_pm_should_add_signal(msk) ? "add_addr" : "rm_addr")); 475 476 slow = lock_sock_fast(ssk); 477 if (prio) { 478 subflow->send_mp_prio = 1; 479 subflow->backup = backup; 480 subflow->request_bkup = backup; 481 } 482 483 __mptcp_subflow_send_ack(ssk); 484 unlock_sock_fast(ssk, slow); 485} 486 487static void mptcp_pm_send_ack(struct mptcp_sock *msk, struct mptcp_subflow_context *subflow, 488 bool prio, bool backup) 489{ 490 spin_unlock_bh(&msk->pm.lock); 491 __mptcp_pm_send_ack(msk, subflow, prio, backup); 492 spin_lock_bh(&msk->pm.lock); 493} 494 495static struct mptcp_pm_addr_entry * 496__lookup_addr_by_id(struct pm_nl_pernet *pernet, unsigned int id) 497{ 498 struct mptcp_pm_addr_entry *entry; 499 500 list_for_each_entry(entry, &pernet->local_addr_list, list) { 501 if (entry->addr.id == id) 502 return entry; 503 } 504 return NULL; 505} 506 507static struct mptcp_pm_addr_entry * 508__lookup_addr(struct pm_nl_pernet *pernet, const struct mptcp_addr_info *info, 509 bool lookup_by_id) 510{ 511 struct mptcp_pm_addr_entry *entry; 512 513 list_for_each_entry(entry, &pernet->local_addr_list, list) { 514 if ((!lookup_by_id && 515 mptcp_addresses_equal(&entry->addr, info, entry->addr.port)) || 516 (lookup_by_id && entry->addr.id == info->id)) 517 return entry; 518 } 519 return NULL; 520} 521 522static void mptcp_pm_create_subflow_or_signal_addr(struct mptcp_sock *msk) 523{ 524 struct sock *sk = (struct sock *)msk; 525 struct mptcp_pm_addr_entry *local; 526 unsigned int add_addr_signal_max; 527 unsigned int local_addr_max; 528 struct pm_nl_pernet *pernet; 529 unsigned int subflows_max; 530 531 pernet = pm_nl_get_pernet(sock_net(sk)); 532 533 add_addr_signal_max = mptcp_pm_get_add_addr_signal_max(msk); 534 local_addr_max = mptcp_pm_get_local_addr_max(msk); 535 subflows_max = mptcp_pm_get_subflows_max(msk); 536 537 /* do lazy endpoint usage accounting for the MPC subflows */ 538 if (unlikely(!(msk->pm.status & BIT(MPTCP_PM_MPC_ENDPOINT_ACCOUNTED))) && msk->first) { 539 struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(msk->first); 540 struct mptcp_pm_addr_entry *entry; 541 struct mptcp_addr_info mpc_addr; 542 bool backup = false; 543 544 mptcp_local_address((struct sock_common *)msk->first, &mpc_addr); 545 rcu_read_lock(); 546 entry = __lookup_addr(pernet, &mpc_addr, false); 547 if (entry) { 548 __clear_bit(entry->addr.id, msk->pm.id_avail_bitmap); 549 msk->mpc_endpoint_id = entry->addr.id; 550 backup = !!(entry->flags & MPTCP_PM_ADDR_FLAG_BACKUP); 551 } 552 rcu_read_unlock(); 553 554 if (backup) 555 mptcp_pm_send_ack(msk, subflow, true, backup); 556 557 msk->pm.status |= BIT(MPTCP_PM_MPC_ENDPOINT_ACCOUNTED); 558 } 559 560 pr_debug("local %d:%d signal %d:%d subflows %d:%d\n", 561 msk->pm.local_addr_used, local_addr_max, 562 msk->pm.add_addr_signaled, add_addr_signal_max, 563 msk->pm.subflows, subflows_max); 564 565 /* check first for announce */ 566 if (msk->pm.add_addr_signaled < add_addr_signal_max) { 567 local = select_signal_address(pernet, msk); 568 569 /* due to racing events on both ends we can reach here while 570 * previous add address is still running: if we invoke now 571 * mptcp_pm_announce_addr(), that will fail and the 572 * corresponding id will be marked as used. 573 * Instead let the PM machinery reschedule us when the 574 * current address announce will be completed. 575 */ 576 if (msk->pm.addr_signal & BIT(MPTCP_ADD_ADDR_SIGNAL)) 577 return; 578 579 if (local) { 580 if (mptcp_pm_alloc_anno_list(msk, &local->addr)) { 581 __clear_bit(local->addr.id, msk->pm.id_avail_bitmap); 582 msk->pm.add_addr_signaled++; 583 mptcp_pm_announce_addr(msk, &local->addr, false); 584 mptcp_pm_nl_addr_send_ack(msk); 585 } 586 } 587 } 588 589 /* check if should create a new subflow */ 590 while (msk->pm.local_addr_used < local_addr_max && 591 msk->pm.subflows < subflows_max) { 592 struct mptcp_addr_info addrs[MPTCP_PM_ADDR_MAX]; 593 bool fullmesh; 594 int i, nr; 595 596 local = select_local_address(pernet, msk); 597 if (!local) 598 break; 599 600 fullmesh = !!(local->flags & MPTCP_PM_ADDR_FLAG_FULLMESH); 601 602 msk->pm.local_addr_used++; 603 __clear_bit(local->addr.id, msk->pm.id_avail_bitmap); 604 nr = fill_remote_addresses_vec(msk, &local->addr, fullmesh, addrs); 605 if (nr == 0) 606 continue; 607 608 spin_unlock_bh(&msk->pm.lock); 609 for (i = 0; i < nr; i++) 610 __mptcp_subflow_connect(sk, &local->addr, &addrs[i]); 611 spin_lock_bh(&msk->pm.lock); 612 } 613 mptcp_pm_nl_check_work_pending(msk); 614} 615 616static void mptcp_pm_nl_fully_established(struct mptcp_sock *msk) 617{ 618 mptcp_pm_create_subflow_or_signal_addr(msk); 619} 620 621static void mptcp_pm_nl_subflow_established(struct mptcp_sock *msk) 622{ 623 mptcp_pm_create_subflow_or_signal_addr(msk); 624} 625 626/* Fill all the local addresses into the array addrs[], 627 * and return the array size. 628 */ 629static unsigned int fill_local_addresses_vec(struct mptcp_sock *msk, 630 struct mptcp_addr_info *remote, 631 struct mptcp_addr_info *addrs) 632{ 633 struct sock *sk = (struct sock *)msk; 634 struct mptcp_pm_addr_entry *entry; 635 struct pm_nl_pernet *pernet; 636 unsigned int subflows_max; 637 int i = 0; 638 639 pernet = pm_nl_get_pernet_from_msk(msk); 640 subflows_max = mptcp_pm_get_subflows_max(msk); 641 642 rcu_read_lock(); 643 list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) { 644 if (!(entry->flags & MPTCP_PM_ADDR_FLAG_FULLMESH)) 645 continue; 646 647 if (!mptcp_pm_addr_families_match(sk, &entry->addr, remote)) 648 continue; 649 650 if (msk->pm.subflows < subflows_max) { 651 msk->pm.subflows++; 652 addrs[i++] = entry->addr; 653 } 654 } 655 rcu_read_unlock(); 656 657 /* If the array is empty, fill in the single 658 * 'IPADDRANY' local address 659 */ 660 if (!i) { 661 struct mptcp_addr_info local; 662 663 memset(&local, 0, sizeof(local)); 664 local.family = 665#if IS_ENABLED(CONFIG_MPTCP_IPV6) 666 remote->family == AF_INET6 && 667 ipv6_addr_v4mapped(&remote->addr6) ? AF_INET : 668#endif 669 remote->family; 670 671 if (!mptcp_pm_addr_families_match(sk, &local, remote)) 672 return 0; 673 674 msk->pm.subflows++; 675 addrs[i++] = local; 676 } 677 678 return i; 679} 680 681static void mptcp_pm_nl_add_addr_received(struct mptcp_sock *msk) 682{ 683 struct mptcp_addr_info addrs[MPTCP_PM_ADDR_MAX]; 684 struct sock *sk = (struct sock *)msk; 685 unsigned int add_addr_accept_max; 686 struct mptcp_addr_info remote; 687 unsigned int subflows_max; 688 int i, nr; 689 690 add_addr_accept_max = mptcp_pm_get_add_addr_accept_max(msk); 691 subflows_max = mptcp_pm_get_subflows_max(msk); 692 693 pr_debug("accepted %d:%d remote family %d", 694 msk->pm.add_addr_accepted, add_addr_accept_max, 695 msk->pm.remote.family); 696 697 remote = msk->pm.remote; 698 mptcp_pm_announce_addr(msk, &remote, true); 699 mptcp_pm_nl_addr_send_ack(msk); 700 701 if (lookup_subflow_by_daddr(&msk->conn_list, &remote)) 702 return; 703 704 /* pick id 0 port, if none is provided the remote address */ 705 if (!remote.port) 706 remote.port = sk->sk_dport; 707 708 /* connect to the specified remote address, using whatever 709 * local address the routing configuration will pick. 710 */ 711 nr = fill_local_addresses_vec(msk, &remote, addrs); 712 if (nr == 0) 713 return; 714 715 msk->pm.add_addr_accepted++; 716 if (msk->pm.add_addr_accepted >= add_addr_accept_max || 717 msk->pm.subflows >= subflows_max) 718 WRITE_ONCE(msk->pm.accept_addr, false); 719 720 spin_unlock_bh(&msk->pm.lock); 721 for (i = 0; i < nr; i++) 722 __mptcp_subflow_connect(sk, &addrs[i], &remote); 723 spin_lock_bh(&msk->pm.lock); 724} 725 726void mptcp_pm_nl_addr_send_ack(struct mptcp_sock *msk) 727{ 728 struct mptcp_subflow_context *subflow; 729 730 msk_owned_by_me(msk); 731 lockdep_assert_held(&msk->pm.lock); 732 733 if (!mptcp_pm_should_add_signal(msk) && 734 !mptcp_pm_should_rm_signal(msk)) 735 return; 736 737 subflow = list_first_entry_or_null(&msk->conn_list, typeof(*subflow), node); 738 if (subflow) 739 mptcp_pm_send_ack(msk, subflow, false, false); 740} 741 742int mptcp_pm_nl_mp_prio_send_ack(struct mptcp_sock *msk, 743 struct mptcp_addr_info *addr, 744 struct mptcp_addr_info *rem, 745 u8 bkup) 746{ 747 struct mptcp_subflow_context *subflow; 748 749 pr_debug("bkup=%d", bkup); 750 751 mptcp_for_each_subflow(msk, subflow) { 752 struct sock *ssk = mptcp_subflow_tcp_sock(subflow); 753 struct mptcp_addr_info local, remote; 754 755 mptcp_local_address((struct sock_common *)ssk, &local); 756 if (!mptcp_addresses_equal(&local, addr, addr->port)) 757 continue; 758 759 if (rem && rem->family != AF_UNSPEC) { 760 remote_address((struct sock_common *)ssk, &remote); 761 if (!mptcp_addresses_equal(&remote, rem, rem->port)) 762 continue; 763 } 764 765 __mptcp_pm_send_ack(msk, subflow, true, bkup); 766 return 0; 767 } 768 769 return -EINVAL; 770} 771 772static bool mptcp_local_id_match(const struct mptcp_sock *msk, u8 local_id, u8 id) 773{ 774 return local_id == id || (!local_id && msk->mpc_endpoint_id == id); 775} 776 777static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk, 778 const struct mptcp_rm_list *rm_list, 779 enum linux_mptcp_mib_field rm_type) 780{ 781 struct mptcp_subflow_context *subflow, *tmp; 782 struct sock *sk = (struct sock *)msk; 783 u8 i; 784 785 pr_debug("%s rm_list_nr %d", 786 rm_type == MPTCP_MIB_RMADDR ? "address" : "subflow", rm_list->nr); 787 788 msk_owned_by_me(msk); 789 790 if (sk->sk_state == TCP_LISTEN) 791 return; 792 793 if (!rm_list->nr) 794 return; 795 796 if (list_empty(&msk->conn_list)) 797 return; 798 799 for (i = 0; i < rm_list->nr; i++) { 800 u8 rm_id = rm_list->ids[i]; 801 bool removed = false; 802 803 mptcp_for_each_subflow_safe(msk, subflow, tmp) { 804 struct sock *ssk = mptcp_subflow_tcp_sock(subflow); 805 u8 remote_id = READ_ONCE(subflow->remote_id); 806 int how = RCV_SHUTDOWN | SEND_SHUTDOWN; 807 u8 id = subflow_get_local_id(subflow); 808 809 if (rm_type == MPTCP_MIB_RMADDR && remote_id != rm_id) 810 continue; 811 if (rm_type == MPTCP_MIB_RMSUBFLOW && !mptcp_local_id_match(msk, id, rm_id)) 812 continue; 813 814 pr_debug(" -> %s rm_list_ids[%d]=%u local_id=%u remote_id=%u mpc_id=%u", 815 rm_type == MPTCP_MIB_RMADDR ? "address" : "subflow", 816 i, rm_id, id, remote_id, msk->mpc_endpoint_id); 817 spin_unlock_bh(&msk->pm.lock); 818 mptcp_subflow_shutdown(sk, ssk, how); 819 820 /* the following takes care of updating the subflows counter */ 821 mptcp_close_ssk(sk, ssk, subflow); 822 spin_lock_bh(&msk->pm.lock); 823 824 removed = true; 825 __MPTCP_INC_STATS(sock_net(sk), rm_type); 826 } 827 if (rm_type == MPTCP_MIB_RMSUBFLOW) 828 __set_bit(rm_id ? rm_id : msk->mpc_endpoint_id, msk->pm.id_avail_bitmap); 829 if (!removed) 830 continue; 831 832 if (!mptcp_pm_is_kernel(msk)) 833 continue; 834 835 if (rm_type == MPTCP_MIB_RMADDR) { 836 msk->pm.add_addr_accepted--; 837 WRITE_ONCE(msk->pm.accept_addr, true); 838 } else if (rm_type == MPTCP_MIB_RMSUBFLOW) { 839 msk->pm.local_addr_used--; 840 } 841 } 842} 843 844static void mptcp_pm_nl_rm_addr_received(struct mptcp_sock *msk) 845{ 846 mptcp_pm_nl_rm_addr_or_subflow(msk, &msk->pm.rm_list_rx, MPTCP_MIB_RMADDR); 847} 848 849void mptcp_pm_nl_rm_subflow_received(struct mptcp_sock *msk, 850 const struct mptcp_rm_list *rm_list) 851{ 852 mptcp_pm_nl_rm_addr_or_subflow(msk, rm_list, MPTCP_MIB_RMSUBFLOW); 853} 854 855void mptcp_pm_nl_work(struct mptcp_sock *msk) 856{ 857 struct mptcp_pm_data *pm = &msk->pm; 858 859 msk_owned_by_me(msk); 860 861 if (!(pm->status & MPTCP_PM_WORK_MASK)) 862 return; 863 864 spin_lock_bh(&msk->pm.lock); 865 866 pr_debug("msk=%p status=%x", msk, pm->status); 867 if (pm->status & BIT(MPTCP_PM_ADD_ADDR_RECEIVED)) { 868 pm->status &= ~BIT(MPTCP_PM_ADD_ADDR_RECEIVED); 869 mptcp_pm_nl_add_addr_received(msk); 870 } 871 if (pm->status & BIT(MPTCP_PM_ADD_ADDR_SEND_ACK)) { 872 pm->status &= ~BIT(MPTCP_PM_ADD_ADDR_SEND_ACK); 873 mptcp_pm_nl_addr_send_ack(msk); 874 } 875 if (pm->status & BIT(MPTCP_PM_RM_ADDR_RECEIVED)) { 876 pm->status &= ~BIT(MPTCP_PM_RM_ADDR_RECEIVED); 877 mptcp_pm_nl_rm_addr_received(msk); 878 } 879 if (pm->status & BIT(MPTCP_PM_ESTABLISHED)) { 880 pm->status &= ~BIT(MPTCP_PM_ESTABLISHED); 881 mptcp_pm_nl_fully_established(msk); 882 } 883 if (pm->status & BIT(MPTCP_PM_SUBFLOW_ESTABLISHED)) { 884 pm->status &= ~BIT(MPTCP_PM_SUBFLOW_ESTABLISHED); 885 mptcp_pm_nl_subflow_established(msk); 886 } 887 888 spin_unlock_bh(&msk->pm.lock); 889} 890 891static bool address_use_port(struct mptcp_pm_addr_entry *entry) 892{ 893 return (entry->flags & 894 (MPTCP_PM_ADDR_FLAG_SIGNAL | MPTCP_PM_ADDR_FLAG_SUBFLOW)) == 895 MPTCP_PM_ADDR_FLAG_SIGNAL; 896} 897 898/* caller must ensure the RCU grace period is already elapsed */ 899static void __mptcp_pm_release_addr_entry(struct mptcp_pm_addr_entry *entry) 900{ 901 if (entry->lsk) 902 sock_release(entry->lsk); 903 kfree(entry); 904} 905 906static int mptcp_pm_nl_append_new_local_addr(struct pm_nl_pernet *pernet, 907 struct mptcp_pm_addr_entry *entry, 908 bool needs_id) 909{ 910 struct mptcp_pm_addr_entry *cur, *del_entry = NULL; 911 unsigned int addr_max; 912 int ret = -EINVAL; 913 914 spin_lock_bh(&pernet->lock); 915 /* to keep the code simple, don't do IDR-like allocation for address ID, 916 * just bail when we exceed limits 917 */ 918 if (pernet->next_id == MPTCP_PM_MAX_ADDR_ID) 919 pernet->next_id = 1; 920 if (pernet->addrs >= MPTCP_PM_ADDR_MAX) { 921 ret = -ERANGE; 922 goto out; 923 } 924 if (test_bit(entry->addr.id, pernet->id_bitmap)) { 925 ret = -EBUSY; 926 goto out; 927 } 928 929 /* do not insert duplicate address, differentiate on port only 930 * singled addresses 931 */ 932 if (!address_use_port(entry)) 933 entry->addr.port = 0; 934 list_for_each_entry(cur, &pernet->local_addr_list, list) { 935 if (mptcp_addresses_equal(&cur->addr, &entry->addr, 936 cur->addr.port || entry->addr.port)) { 937 /* allow replacing the exiting endpoint only if such 938 * endpoint is an implicit one and the user-space 939 * did not provide an endpoint id 940 */ 941 if (!(cur->flags & MPTCP_PM_ADDR_FLAG_IMPLICIT)) { 942 ret = -EEXIST; 943 goto out; 944 } 945 if (entry->addr.id) 946 goto out; 947 948 pernet->addrs--; 949 entry->addr.id = cur->addr.id; 950 list_del_rcu(&cur->list); 951 del_entry = cur; 952 break; 953 } 954 } 955 956 if (!entry->addr.id && needs_id) { 957find_next: 958 entry->addr.id = find_next_zero_bit(pernet->id_bitmap, 959 MPTCP_PM_MAX_ADDR_ID + 1, 960 pernet->next_id); 961 if (!entry->addr.id && pernet->next_id != 1) { 962 pernet->next_id = 1; 963 goto find_next; 964 } 965 } 966 967 if (!entry->addr.id && needs_id) 968 goto out; 969 970 __set_bit(entry->addr.id, pernet->id_bitmap); 971 if (entry->addr.id > pernet->next_id) 972 pernet->next_id = entry->addr.id; 973 974 if (entry->flags & MPTCP_PM_ADDR_FLAG_SIGNAL) { 975 addr_max = pernet->add_addr_signal_max; 976 WRITE_ONCE(pernet->add_addr_signal_max, addr_max + 1); 977 } 978 if (entry->flags & MPTCP_PM_ADDR_FLAG_SUBFLOW) { 979 addr_max = pernet->local_addr_max; 980 WRITE_ONCE(pernet->local_addr_max, addr_max + 1); 981 } 982 983 pernet->addrs++; 984 if (!entry->addr.port) 985 list_add_tail_rcu(&entry->list, &pernet->local_addr_list); 986 else 987 list_add_rcu(&entry->list, &pernet->local_addr_list); 988 ret = entry->addr.id; 989 990out: 991 spin_unlock_bh(&pernet->lock); 992 993 /* just replaced an existing entry, free it */ 994 if (del_entry) { 995 synchronize_rcu(); 996 __mptcp_pm_release_addr_entry(del_entry); 997 } 998 return ret; 999} 1000 1001static struct lock_class_key mptcp_slock_keys[2]; 1002static struct lock_class_key mptcp_keys[2]; 1003 1004static int mptcp_pm_nl_create_listen_socket(struct sock *sk, 1005 struct mptcp_pm_addr_entry *entry) 1006{ 1007 bool is_ipv6 = sk->sk_family == AF_INET6; 1008 int addrlen = sizeof(struct sockaddr_in); 1009 struct sockaddr_storage addr; 1010 struct sock *newsk, *ssk; 1011 int backlog = 1024; 1012 int err; 1013 1014 err = sock_create_kern(sock_net(sk), entry->addr.family, 1015 SOCK_STREAM, IPPROTO_MPTCP, &entry->lsk); 1016 if (err) 1017 return err; 1018 1019 newsk = entry->lsk->sk; 1020 if (!newsk) 1021 return -EINVAL; 1022 1023 /* The subflow socket lock is acquired in a nested to the msk one 1024 * in several places, even by the TCP stack, and this msk is a kernel 1025 * socket: lockdep complains. Instead of propagating the _nested 1026 * modifiers in several places, re-init the lock class for the msk 1027 * socket to an mptcp specific one. 1028 */ 1029 sock_lock_init_class_and_name(newsk, 1030 is_ipv6 ? "mlock-AF_INET6" : "mlock-AF_INET", 1031 &mptcp_slock_keys[is_ipv6], 1032 is_ipv6 ? "msk_lock-AF_INET6" : "msk_lock-AF_INET", 1033 &mptcp_keys[is_ipv6]); 1034 1035 lock_sock(newsk); 1036 ssk = __mptcp_nmpc_sk(mptcp_sk(newsk)); 1037 release_sock(newsk); 1038 if (IS_ERR(ssk)) 1039 return PTR_ERR(ssk); 1040 1041 mptcp_info2sockaddr(&entry->addr, &addr, entry->addr.family); 1042#if IS_ENABLED(CONFIG_MPTCP_IPV6) 1043 if (entry->addr.family == AF_INET6) 1044 addrlen = sizeof(struct sockaddr_in6); 1045#endif 1046 if (ssk->sk_family == AF_INET) 1047 err = inet_bind_sk(ssk, (struct sockaddr *)&addr, addrlen); 1048#if IS_ENABLED(CONFIG_MPTCP_IPV6) 1049 else if (ssk->sk_family == AF_INET6) 1050 err = inet6_bind_sk(ssk, (struct sockaddr *)&addr, addrlen); 1051#endif 1052 if (err) 1053 return err; 1054 1055 /* We don't use mptcp_set_state() here because it needs to be called 1056 * under the msk socket lock. For the moment, that will not bring 1057 * anything more than only calling inet_sk_state_store(), because the 1058 * old status is known (TCP_CLOSE). 1059 */ 1060 inet_sk_state_store(newsk, TCP_LISTEN); 1061 lock_sock(ssk); 1062 err = __inet_listen_sk(ssk, backlog); 1063 if (!err) 1064 mptcp_event_pm_listener(ssk, MPTCP_EVENT_LISTENER_CREATED); 1065 release_sock(ssk); 1066 return err; 1067} 1068 1069int mptcp_pm_nl_get_local_id(struct mptcp_sock *msk, struct mptcp_addr_info *skc) 1070{ 1071 struct mptcp_pm_addr_entry *entry; 1072 struct pm_nl_pernet *pernet; 1073 int ret = -1; 1074 1075 pernet = pm_nl_get_pernet_from_msk(msk); 1076 1077 rcu_read_lock(); 1078 list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) { 1079 if (mptcp_addresses_equal(&entry->addr, skc, entry->addr.port)) { 1080 ret = entry->addr.id; 1081 break; 1082 } 1083 } 1084 rcu_read_unlock(); 1085 if (ret >= 0) 1086 return ret; 1087 1088 /* address not found, add to local list */ 1089 entry = kmalloc(sizeof(*entry), GFP_ATOMIC); 1090 if (!entry) 1091 return -ENOMEM; 1092 1093 entry->addr = *skc; 1094 entry->addr.id = 0; 1095 entry->addr.port = 0; 1096 entry->ifindex = 0; 1097 entry->flags = MPTCP_PM_ADDR_FLAG_IMPLICIT; 1098 entry->lsk = NULL; 1099 ret = mptcp_pm_nl_append_new_local_addr(pernet, entry, true); 1100 if (ret < 0) 1101 kfree(entry); 1102 1103 return ret; 1104} 1105 1106#define MPTCP_PM_CMD_GRP_OFFSET 0 1107#define MPTCP_PM_EV_GRP_OFFSET 1 1108 1109static const struct genl_multicast_group mptcp_pm_mcgrps[] = { 1110 [MPTCP_PM_CMD_GRP_OFFSET] = { .name = MPTCP_PM_CMD_GRP_NAME, }, 1111 [MPTCP_PM_EV_GRP_OFFSET] = { .name = MPTCP_PM_EV_GRP_NAME, 1112 .flags = GENL_UNS_ADMIN_PERM, 1113 }, 1114}; 1115 1116static const struct nla_policy 1117mptcp_pm_addr_policy[MPTCP_PM_ADDR_ATTR_MAX + 1] = { 1118 [MPTCP_PM_ADDR_ATTR_FAMILY] = { .type = NLA_U16, }, 1119 [MPTCP_PM_ADDR_ATTR_ID] = { .type = NLA_U8, }, 1120 [MPTCP_PM_ADDR_ATTR_ADDR4] = { .type = NLA_U32, }, 1121 [MPTCP_PM_ADDR_ATTR_ADDR6] = 1122 NLA_POLICY_EXACT_LEN(sizeof(struct in6_addr)), 1123 [MPTCP_PM_ADDR_ATTR_PORT] = { .type = NLA_U16 }, 1124 [MPTCP_PM_ADDR_ATTR_FLAGS] = { .type = NLA_U32 }, 1125 [MPTCP_PM_ADDR_ATTR_IF_IDX] = { .type = NLA_S32 }, 1126}; 1127 1128static const struct nla_policy mptcp_pm_policy[MPTCP_PM_ATTR_MAX + 1] = { 1129 [MPTCP_PM_ATTR_ADDR] = 1130 NLA_POLICY_NESTED(mptcp_pm_addr_policy), 1131 [MPTCP_PM_ATTR_RCV_ADD_ADDRS] = { .type = NLA_U32, }, 1132 [MPTCP_PM_ATTR_SUBFLOWS] = { .type = NLA_U32, }, 1133 [MPTCP_PM_ATTR_TOKEN] = { .type = NLA_U32, }, 1134 [MPTCP_PM_ATTR_LOC_ID] = { .type = NLA_U8, }, 1135 [MPTCP_PM_ATTR_ADDR_REMOTE] = 1136 NLA_POLICY_NESTED(mptcp_pm_addr_policy), 1137}; 1138 1139void mptcp_pm_nl_subflow_chk_stale(const struct mptcp_sock *msk, struct sock *ssk) 1140{ 1141 struct mptcp_subflow_context *iter, *subflow = mptcp_subflow_ctx(ssk); 1142 struct sock *sk = (struct sock *)msk; 1143 unsigned int active_max_loss_cnt; 1144 struct net *net = sock_net(sk); 1145 unsigned int stale_loss_cnt; 1146 bool slow; 1147 1148 stale_loss_cnt = mptcp_stale_loss_cnt(net); 1149 if (subflow->stale || !stale_loss_cnt || subflow->stale_count <= stale_loss_cnt) 1150 return; 1151 1152 /* look for another available subflow not in loss state */ 1153 active_max_loss_cnt = max_t(int, stale_loss_cnt - 1, 1); 1154 mptcp_for_each_subflow(msk, iter) { 1155 if (iter != subflow && mptcp_subflow_active(iter) && 1156 iter->stale_count < active_max_loss_cnt) { 1157 /* we have some alternatives, try to mark this subflow as idle ...*/ 1158 slow = lock_sock_fast(ssk); 1159 if (!tcp_rtx_and_write_queues_empty(ssk)) { 1160 subflow->stale = 1; 1161 __mptcp_retransmit_pending_data(sk); 1162 MPTCP_INC_STATS(net, MPTCP_MIB_SUBFLOWSTALE); 1163 } 1164 unlock_sock_fast(ssk, slow); 1165 1166 /* always try to push the pending data regardless of re-injections: 1167 * we can possibly use backup subflows now, and subflow selection 1168 * is cheap under the msk socket lock 1169 */ 1170 __mptcp_push_pending(sk, 0); 1171 return; 1172 } 1173 } 1174} 1175 1176static int mptcp_pm_family_to_addr(int family) 1177{ 1178#if IS_ENABLED(CONFIG_MPTCP_IPV6) 1179 if (family == AF_INET6) 1180 return MPTCP_PM_ADDR_ATTR_ADDR6; 1181#endif 1182 return MPTCP_PM_ADDR_ATTR_ADDR4; 1183} 1184 1185static int mptcp_pm_parse_pm_addr_attr(struct nlattr *tb[], 1186 const struct nlattr *attr, 1187 struct genl_info *info, 1188 struct mptcp_addr_info *addr, 1189 bool require_family) 1190{ 1191 int err, addr_addr; 1192 1193 if (!attr) { 1194 GENL_SET_ERR_MSG(info, "missing address info"); 1195 return -EINVAL; 1196 } 1197 1198 /* no validation needed - was already done via nested policy */ 1199 err = nla_parse_nested_deprecated(tb, MPTCP_PM_ADDR_ATTR_MAX, attr, 1200 mptcp_pm_addr_policy, info->extack); 1201 if (err) 1202 return err; 1203 1204 if (tb[MPTCP_PM_ADDR_ATTR_ID]) 1205 addr->id = nla_get_u8(tb[MPTCP_PM_ADDR_ATTR_ID]); 1206 1207 if (!tb[MPTCP_PM_ADDR_ATTR_FAMILY]) { 1208 if (!require_family) 1209 return 0; 1210 1211 NL_SET_ERR_MSG_ATTR(info->extack, attr, 1212 "missing family"); 1213 return -EINVAL; 1214 } 1215 1216 addr->family = nla_get_u16(tb[MPTCP_PM_ADDR_ATTR_FAMILY]); 1217 if (addr->family != AF_INET 1218#if IS_ENABLED(CONFIG_MPTCP_IPV6) 1219 && addr->family != AF_INET6 1220#endif 1221 ) { 1222 NL_SET_ERR_MSG_ATTR(info->extack, attr, 1223 "unknown address family"); 1224 return -EINVAL; 1225 } 1226 addr_addr = mptcp_pm_family_to_addr(addr->family); 1227 if (!tb[addr_addr]) { 1228 NL_SET_ERR_MSG_ATTR(info->extack, attr, 1229 "missing address data"); 1230 return -EINVAL; 1231 } 1232 1233#if IS_ENABLED(CONFIG_MPTCP_IPV6) 1234 if (addr->family == AF_INET6) 1235 addr->addr6 = nla_get_in6_addr(tb[addr_addr]); 1236 else 1237#endif 1238 addr->addr.s_addr = nla_get_in_addr(tb[addr_addr]); 1239 1240 if (tb[MPTCP_PM_ADDR_ATTR_PORT]) 1241 addr->port = htons(nla_get_u16(tb[MPTCP_PM_ADDR_ATTR_PORT])); 1242 1243 return 0; 1244} 1245 1246int mptcp_pm_parse_addr(struct nlattr *attr, struct genl_info *info, 1247 struct mptcp_addr_info *addr) 1248{ 1249 struct nlattr *tb[MPTCP_PM_ADDR_ATTR_MAX + 1]; 1250 1251 memset(addr, 0, sizeof(*addr)); 1252 1253 return mptcp_pm_parse_pm_addr_attr(tb, attr, info, addr, true); 1254} 1255 1256int mptcp_pm_parse_entry(struct nlattr *attr, struct genl_info *info, 1257 bool require_family, 1258 struct mptcp_pm_addr_entry *entry) 1259{ 1260 struct nlattr *tb[MPTCP_PM_ADDR_ATTR_MAX + 1]; 1261 int err; 1262 1263 memset(entry, 0, sizeof(*entry)); 1264 1265 err = mptcp_pm_parse_pm_addr_attr(tb, attr, info, &entry->addr, require_family); 1266 if (err) 1267 return err; 1268 1269 if (tb[MPTCP_PM_ADDR_ATTR_IF_IDX]) { 1270 u32 val = nla_get_s32(tb[MPTCP_PM_ADDR_ATTR_IF_IDX]); 1271 1272 entry->ifindex = val; 1273 } 1274 1275 if (tb[MPTCP_PM_ADDR_ATTR_FLAGS]) 1276 entry->flags = nla_get_u32(tb[MPTCP_PM_ADDR_ATTR_FLAGS]); 1277 1278 if (tb[MPTCP_PM_ADDR_ATTR_PORT]) 1279 entry->addr.port = htons(nla_get_u16(tb[MPTCP_PM_ADDR_ATTR_PORT])); 1280 1281 return 0; 1282} 1283 1284static struct pm_nl_pernet *genl_info_pm_nl(struct genl_info *info) 1285{ 1286 return pm_nl_get_pernet(genl_info_net(info)); 1287} 1288 1289static int mptcp_nl_add_subflow_or_signal_addr(struct net *net) 1290{ 1291 struct mptcp_sock *msk; 1292 long s_slot = 0, s_num = 0; 1293 1294 while ((msk = mptcp_token_iter_next(net, &s_slot, &s_num)) != NULL) { 1295 struct sock *sk = (struct sock *)msk; 1296 1297 if (!READ_ONCE(msk->fully_established) || 1298 mptcp_pm_is_userspace(msk)) 1299 goto next; 1300 1301 lock_sock(sk); 1302 spin_lock_bh(&msk->pm.lock); 1303 mptcp_pm_create_subflow_or_signal_addr(msk); 1304 spin_unlock_bh(&msk->pm.lock); 1305 release_sock(sk); 1306 1307next: 1308 sock_put(sk); 1309 cond_resched(); 1310 } 1311 1312 return 0; 1313} 1314 1315static bool mptcp_pm_has_addr_attr_id(const struct nlattr *attr, 1316 struct genl_info *info) 1317{ 1318 struct nlattr *tb[MPTCP_PM_ADDR_ATTR_MAX + 1]; 1319 1320 if (!nla_parse_nested_deprecated(tb, MPTCP_PM_ADDR_ATTR_MAX, attr, 1321 mptcp_pm_addr_policy, info->extack) && 1322 tb[MPTCP_PM_ADDR_ATTR_ID]) 1323 return true; 1324 return false; 1325} 1326 1327static int mptcp_nl_cmd_add_addr(struct sk_buff *skb, struct genl_info *info) 1328{ 1329 struct nlattr *attr = info->attrs[MPTCP_PM_ATTR_ADDR]; 1330 struct pm_nl_pernet *pernet = genl_info_pm_nl(info); 1331 struct mptcp_pm_addr_entry addr, *entry; 1332 int ret; 1333 1334 ret = mptcp_pm_parse_entry(attr, info, true, &addr); 1335 if (ret < 0) 1336 return ret; 1337 1338 if (addr.addr.port && !(addr.flags & MPTCP_PM_ADDR_FLAG_SIGNAL)) { 1339 GENL_SET_ERR_MSG(info, "flags must have signal when using port"); 1340 return -EINVAL; 1341 } 1342 1343 if (addr.flags & MPTCP_PM_ADDR_FLAG_SIGNAL && 1344 addr.flags & MPTCP_PM_ADDR_FLAG_FULLMESH) { 1345 GENL_SET_ERR_MSG(info, "flags mustn't have both signal and fullmesh"); 1346 return -EINVAL; 1347 } 1348 1349 if (addr.flags & MPTCP_PM_ADDR_FLAG_IMPLICIT) { 1350 GENL_SET_ERR_MSG(info, "can't create IMPLICIT endpoint"); 1351 return -EINVAL; 1352 } 1353 1354 entry = kzalloc(sizeof(*entry), GFP_KERNEL_ACCOUNT); 1355 if (!entry) { 1356 GENL_SET_ERR_MSG(info, "can't allocate addr"); 1357 return -ENOMEM; 1358 } 1359 1360 *entry = addr; 1361 if (entry->addr.port) { 1362 ret = mptcp_pm_nl_create_listen_socket(skb->sk, entry); 1363 if (ret) { 1364 GENL_SET_ERR_MSG_FMT(info, "create listen socket error: %d", ret); 1365 goto out_free; 1366 } 1367 } 1368 ret = mptcp_pm_nl_append_new_local_addr(pernet, entry, 1369 !mptcp_pm_has_addr_attr_id(attr, info)); 1370 if (ret < 0) { 1371 GENL_SET_ERR_MSG_FMT(info, "too many addresses or duplicate one: %d", ret); 1372 goto out_free; 1373 } 1374 1375 mptcp_nl_add_subflow_or_signal_addr(sock_net(skb->sk)); 1376 return 0; 1377 1378out_free: 1379 __mptcp_pm_release_addr_entry(entry); 1380 return ret; 1381} 1382 1383int mptcp_pm_nl_get_flags_and_ifindex_by_id(struct mptcp_sock *msk, unsigned int id, 1384 u8 *flags, int *ifindex) 1385{ 1386 struct mptcp_pm_addr_entry *entry; 1387 struct sock *sk = (struct sock *)msk; 1388 struct net *net = sock_net(sk); 1389 1390 rcu_read_lock(); 1391 entry = __lookup_addr_by_id(pm_nl_get_pernet(net), id); 1392 if (entry) { 1393 *flags = entry->flags; 1394 *ifindex = entry->ifindex; 1395 } 1396 rcu_read_unlock(); 1397 1398 return 0; 1399} 1400 1401static bool remove_anno_list_by_saddr(struct mptcp_sock *msk, 1402 const struct mptcp_addr_info *addr) 1403{ 1404 struct mptcp_pm_add_entry *entry; 1405 1406 entry = mptcp_pm_del_add_timer(msk, addr, false); 1407 if (entry) { 1408 list_del(&entry->list); 1409 kfree(entry); 1410 return true; 1411 } 1412 1413 return false; 1414} 1415 1416static bool mptcp_pm_remove_anno_addr(struct mptcp_sock *msk, 1417 const struct mptcp_addr_info *addr, 1418 bool force) 1419{ 1420 struct mptcp_rm_list list = { .nr = 0 }; 1421 bool ret; 1422 1423 list.ids[list.nr++] = addr->id; 1424 1425 ret = remove_anno_list_by_saddr(msk, addr); 1426 if (ret || force) { 1427 spin_lock_bh(&msk->pm.lock); 1428 mptcp_pm_remove_addr(msk, &list); 1429 spin_unlock_bh(&msk->pm.lock); 1430 } 1431 return ret; 1432} 1433 1434static int mptcp_nl_remove_subflow_and_signal_addr(struct net *net, 1435 const struct mptcp_pm_addr_entry *entry) 1436{ 1437 const struct mptcp_addr_info *addr = &entry->addr; 1438 struct mptcp_rm_list list = { .nr = 0 }; 1439 long s_slot = 0, s_num = 0; 1440 struct mptcp_sock *msk; 1441 1442 pr_debug("remove_id=%d", addr->id); 1443 1444 list.ids[list.nr++] = addr->id; 1445 1446 while ((msk = mptcp_token_iter_next(net, &s_slot, &s_num)) != NULL) { 1447 struct sock *sk = (struct sock *)msk; 1448 bool remove_subflow; 1449 1450 if (mptcp_pm_is_userspace(msk)) 1451 goto next; 1452 1453 if (list_empty(&msk->conn_list)) { 1454 mptcp_pm_remove_anno_addr(msk, addr, false); 1455 goto next; 1456 } 1457 1458 lock_sock(sk); 1459 remove_subflow = lookup_subflow_by_saddr(&msk->conn_list, addr); 1460 mptcp_pm_remove_anno_addr(msk, addr, remove_subflow && 1461 !(entry->flags & MPTCP_PM_ADDR_FLAG_IMPLICIT)); 1462 if (remove_subflow) 1463 mptcp_pm_remove_subflow(msk, &list); 1464 release_sock(sk); 1465 1466next: 1467 sock_put(sk); 1468 cond_resched(); 1469 } 1470 1471 return 0; 1472} 1473 1474static int mptcp_nl_remove_id_zero_address(struct net *net, 1475 struct mptcp_addr_info *addr) 1476{ 1477 struct mptcp_rm_list list = { .nr = 0 }; 1478 long s_slot = 0, s_num = 0; 1479 struct mptcp_sock *msk; 1480 1481 list.ids[list.nr++] = 0; 1482 1483 while ((msk = mptcp_token_iter_next(net, &s_slot, &s_num)) != NULL) { 1484 struct sock *sk = (struct sock *)msk; 1485 struct mptcp_addr_info msk_local; 1486 1487 if (list_empty(&msk->conn_list) || mptcp_pm_is_userspace(msk)) 1488 goto next; 1489 1490 mptcp_local_address((struct sock_common *)msk, &msk_local); 1491 if (!mptcp_addresses_equal(&msk_local, addr, addr->port)) 1492 goto next; 1493 1494 lock_sock(sk); 1495 spin_lock_bh(&msk->pm.lock); 1496 mptcp_pm_remove_addr(msk, &list); 1497 mptcp_pm_nl_rm_subflow_received(msk, &list); 1498 spin_unlock_bh(&msk->pm.lock); 1499 release_sock(sk); 1500 1501next: 1502 sock_put(sk); 1503 cond_resched(); 1504 } 1505 1506 return 0; 1507} 1508 1509static int mptcp_nl_cmd_del_addr(struct sk_buff *skb, struct genl_info *info) 1510{ 1511 struct nlattr *attr = info->attrs[MPTCP_PM_ATTR_ADDR]; 1512 struct pm_nl_pernet *pernet = genl_info_pm_nl(info); 1513 struct mptcp_pm_addr_entry addr, *entry; 1514 unsigned int addr_max; 1515 int ret; 1516 1517 ret = mptcp_pm_parse_entry(attr, info, false, &addr); 1518 if (ret < 0) 1519 return ret; 1520 1521 /* the zero id address is special: the first address used by the msk 1522 * always gets such an id, so different subflows can have different zero 1523 * id addresses. Additionally zero id is not accounted for in id_bitmap. 1524 * Let's use an 'mptcp_rm_list' instead of the common remove code. 1525 */ 1526 if (addr.addr.id == 0) 1527 return mptcp_nl_remove_id_zero_address(sock_net(skb->sk), &addr.addr); 1528 1529 spin_lock_bh(&pernet->lock); 1530 entry = __lookup_addr_by_id(pernet, addr.addr.id); 1531 if (!entry) { 1532 GENL_SET_ERR_MSG(info, "address not found"); 1533 spin_unlock_bh(&pernet->lock); 1534 return -EINVAL; 1535 } 1536 if (entry->flags & MPTCP_PM_ADDR_FLAG_SIGNAL) { 1537 addr_max = pernet->add_addr_signal_max; 1538 WRITE_ONCE(pernet->add_addr_signal_max, addr_max - 1); 1539 } 1540 if (entry->flags & MPTCP_PM_ADDR_FLAG_SUBFLOW) { 1541 addr_max = pernet->local_addr_max; 1542 WRITE_ONCE(pernet->local_addr_max, addr_max - 1); 1543 } 1544 1545 pernet->addrs--; 1546 list_del_rcu(&entry->list); 1547 __clear_bit(entry->addr.id, pernet->id_bitmap); 1548 spin_unlock_bh(&pernet->lock); 1549 1550 mptcp_nl_remove_subflow_and_signal_addr(sock_net(skb->sk), entry); 1551 synchronize_rcu(); 1552 __mptcp_pm_release_addr_entry(entry); 1553 1554 return ret; 1555} 1556 1557void mptcp_pm_remove_addrs(struct mptcp_sock *msk, struct list_head *rm_list) 1558{ 1559 struct mptcp_rm_list alist = { .nr = 0 }; 1560 struct mptcp_pm_addr_entry *entry; 1561 1562 list_for_each_entry(entry, rm_list, list) { 1563 if ((remove_anno_list_by_saddr(msk, &entry->addr) || 1564 lookup_subflow_by_saddr(&msk->conn_list, &entry->addr)) && 1565 alist.nr < MPTCP_RM_IDS_MAX) 1566 alist.ids[alist.nr++] = entry->addr.id; 1567 } 1568 1569 if (alist.nr) { 1570 spin_lock_bh(&msk->pm.lock); 1571 mptcp_pm_remove_addr(msk, &alist); 1572 spin_unlock_bh(&msk->pm.lock); 1573 } 1574} 1575 1576void mptcp_pm_remove_addrs_and_subflows(struct mptcp_sock *msk, 1577 struct list_head *rm_list) 1578{ 1579 struct mptcp_rm_list alist = { .nr = 0 }, slist = { .nr = 0 }; 1580 struct mptcp_pm_addr_entry *entry; 1581 1582 list_for_each_entry(entry, rm_list, list) { 1583 if (lookup_subflow_by_saddr(&msk->conn_list, &entry->addr) && 1584 slist.nr < MPTCP_RM_IDS_MAX) 1585 slist.ids[slist.nr++] = entry->addr.id; 1586 1587 if (remove_anno_list_by_saddr(msk, &entry->addr) && 1588 alist.nr < MPTCP_RM_IDS_MAX) 1589 alist.ids[alist.nr++] = entry->addr.id; 1590 } 1591 1592 if (alist.nr) { 1593 spin_lock_bh(&msk->pm.lock); 1594 mptcp_pm_remove_addr(msk, &alist); 1595 spin_unlock_bh(&msk->pm.lock); 1596 } 1597 if (slist.nr) 1598 mptcp_pm_remove_subflow(msk, &slist); 1599} 1600 1601static void mptcp_nl_remove_addrs_list(struct net *net, 1602 struct list_head *rm_list) 1603{ 1604 long s_slot = 0, s_num = 0; 1605 struct mptcp_sock *msk; 1606 1607 if (list_empty(rm_list)) 1608 return; 1609 1610 while ((msk = mptcp_token_iter_next(net, &s_slot, &s_num)) != NULL) { 1611 struct sock *sk = (struct sock *)msk; 1612 1613 if (!mptcp_pm_is_userspace(msk)) { 1614 lock_sock(sk); 1615 mptcp_pm_remove_addrs_and_subflows(msk, rm_list); 1616 release_sock(sk); 1617 } 1618 1619 sock_put(sk); 1620 cond_resched(); 1621 } 1622} 1623 1624/* caller must ensure the RCU grace period is already elapsed */ 1625static void __flush_addrs(struct list_head *list) 1626{ 1627 while (!list_empty(list)) { 1628 struct mptcp_pm_addr_entry *cur; 1629 1630 cur = list_entry(list->next, 1631 struct mptcp_pm_addr_entry, list); 1632 list_del_rcu(&cur->list); 1633 __mptcp_pm_release_addr_entry(cur); 1634 } 1635} 1636 1637static void __reset_counters(struct pm_nl_pernet *pernet) 1638{ 1639 WRITE_ONCE(pernet->add_addr_signal_max, 0); 1640 WRITE_ONCE(pernet->add_addr_accept_max, 0); 1641 WRITE_ONCE(pernet->local_addr_max, 0); 1642 pernet->addrs = 0; 1643} 1644 1645static int mptcp_nl_cmd_flush_addrs(struct sk_buff *skb, struct genl_info *info) 1646{ 1647 struct pm_nl_pernet *pernet = genl_info_pm_nl(info); 1648 LIST_HEAD(free_list); 1649 1650 spin_lock_bh(&pernet->lock); 1651 list_splice_init(&pernet->local_addr_list, &free_list); 1652 __reset_counters(pernet); 1653 pernet->next_id = 1; 1654 bitmap_zero(pernet->id_bitmap, MPTCP_PM_MAX_ADDR_ID + 1); 1655 spin_unlock_bh(&pernet->lock); 1656 mptcp_nl_remove_addrs_list(sock_net(skb->sk), &free_list); 1657 synchronize_rcu(); 1658 __flush_addrs(&free_list); 1659 return 0; 1660} 1661 1662static int mptcp_nl_fill_addr(struct sk_buff *skb, 1663 struct mptcp_pm_addr_entry *entry) 1664{ 1665 struct mptcp_addr_info *addr = &entry->addr; 1666 struct nlattr *attr; 1667 1668 attr = nla_nest_start(skb, MPTCP_PM_ATTR_ADDR); 1669 if (!attr) 1670 return -EMSGSIZE; 1671 1672 if (nla_put_u16(skb, MPTCP_PM_ADDR_ATTR_FAMILY, addr->family)) 1673 goto nla_put_failure; 1674 if (nla_put_u16(skb, MPTCP_PM_ADDR_ATTR_PORT, ntohs(addr->port))) 1675 goto nla_put_failure; 1676 if (nla_put_u8(skb, MPTCP_PM_ADDR_ATTR_ID, addr->id)) 1677 goto nla_put_failure; 1678 if (nla_put_u32(skb, MPTCP_PM_ADDR_ATTR_FLAGS, entry->flags)) 1679 goto nla_put_failure; 1680 if (entry->ifindex && 1681 nla_put_s32(skb, MPTCP_PM_ADDR_ATTR_IF_IDX, entry->ifindex)) 1682 goto nla_put_failure; 1683 1684 if (addr->family == AF_INET && 1685 nla_put_in_addr(skb, MPTCP_PM_ADDR_ATTR_ADDR4, 1686 addr->addr.s_addr)) 1687 goto nla_put_failure; 1688#if IS_ENABLED(CONFIG_MPTCP_IPV6) 1689 else if (addr->family == AF_INET6 && 1690 nla_put_in6_addr(skb, MPTCP_PM_ADDR_ATTR_ADDR6, &addr->addr6)) 1691 goto nla_put_failure; 1692#endif 1693 nla_nest_end(skb, attr); 1694 return 0; 1695 1696nla_put_failure: 1697 nla_nest_cancel(skb, attr); 1698 return -EMSGSIZE; 1699} 1700 1701static int mptcp_nl_cmd_get_addr(struct sk_buff *skb, struct genl_info *info) 1702{ 1703 struct nlattr *attr = info->attrs[MPTCP_PM_ATTR_ADDR]; 1704 struct pm_nl_pernet *pernet = genl_info_pm_nl(info); 1705 struct mptcp_pm_addr_entry addr, *entry; 1706 struct sk_buff *msg; 1707 void *reply; 1708 int ret; 1709 1710 ret = mptcp_pm_parse_entry(attr, info, false, &addr); 1711 if (ret < 0) 1712 return ret; 1713 1714 msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL); 1715 if (!msg) 1716 return -ENOMEM; 1717 1718 reply = genlmsg_put_reply(msg, info, &mptcp_genl_family, 0, 1719 info->genlhdr->cmd); 1720 if (!reply) { 1721 GENL_SET_ERR_MSG(info, "not enough space in Netlink message"); 1722 ret = -EMSGSIZE; 1723 goto fail; 1724 } 1725 1726 spin_lock_bh(&pernet->lock); 1727 entry = __lookup_addr_by_id(pernet, addr.addr.id); 1728 if (!entry) { 1729 GENL_SET_ERR_MSG(info, "address not found"); 1730 ret = -EINVAL; 1731 goto unlock_fail; 1732 } 1733 1734 ret = mptcp_nl_fill_addr(msg, entry); 1735 if (ret) 1736 goto unlock_fail; 1737 1738 genlmsg_end(msg, reply); 1739 ret = genlmsg_reply(msg, info); 1740 spin_unlock_bh(&pernet->lock); 1741 return ret; 1742 1743unlock_fail: 1744 spin_unlock_bh(&pernet->lock); 1745 1746fail: 1747 nlmsg_free(msg); 1748 return ret; 1749} 1750 1751static int mptcp_nl_cmd_dump_addrs(struct sk_buff *msg, 1752 struct netlink_callback *cb) 1753{ 1754 struct net *net = sock_net(msg->sk); 1755 struct mptcp_pm_addr_entry *entry; 1756 struct pm_nl_pernet *pernet; 1757 int id = cb->args[0]; 1758 void *hdr; 1759 int i; 1760 1761 pernet = pm_nl_get_pernet(net); 1762 1763 spin_lock_bh(&pernet->lock); 1764 for (i = id; i < MPTCP_PM_MAX_ADDR_ID + 1; i++) { 1765 if (test_bit(i, pernet->id_bitmap)) { 1766 entry = __lookup_addr_by_id(pernet, i); 1767 if (!entry) 1768 break; 1769 1770 if (entry->addr.id <= id) 1771 continue; 1772 1773 hdr = genlmsg_put(msg, NETLINK_CB(cb->skb).portid, 1774 cb->nlh->nlmsg_seq, &mptcp_genl_family, 1775 NLM_F_MULTI, MPTCP_PM_CMD_GET_ADDR); 1776 if (!hdr) 1777 break; 1778 1779 if (mptcp_nl_fill_addr(msg, entry) < 0) { 1780 genlmsg_cancel(msg, hdr); 1781 break; 1782 } 1783 1784 id = entry->addr.id; 1785 genlmsg_end(msg, hdr); 1786 } 1787 } 1788 spin_unlock_bh(&pernet->lock); 1789 1790 cb->args[0] = id; 1791 return msg->len; 1792} 1793 1794static int parse_limit(struct genl_info *info, int id, unsigned int *limit) 1795{ 1796 struct nlattr *attr = info->attrs[id]; 1797 1798 if (!attr) 1799 return 0; 1800 1801 *limit = nla_get_u32(attr); 1802 if (*limit > MPTCP_PM_ADDR_MAX) { 1803 GENL_SET_ERR_MSG(info, "limit greater than maximum"); 1804 return -EINVAL; 1805 } 1806 return 0; 1807} 1808 1809static int 1810mptcp_nl_cmd_set_limits(struct sk_buff *skb, struct genl_info *info) 1811{ 1812 struct pm_nl_pernet *pernet = genl_info_pm_nl(info); 1813 unsigned int rcv_addrs, subflows; 1814 int ret; 1815 1816 spin_lock_bh(&pernet->lock); 1817 rcv_addrs = pernet->add_addr_accept_max; 1818 ret = parse_limit(info, MPTCP_PM_ATTR_RCV_ADD_ADDRS, &rcv_addrs); 1819 if (ret) 1820 goto unlock; 1821 1822 subflows = pernet->subflows_max; 1823 ret = parse_limit(info, MPTCP_PM_ATTR_SUBFLOWS, &subflows); 1824 if (ret) 1825 goto unlock; 1826 1827 WRITE_ONCE(pernet->add_addr_accept_max, rcv_addrs); 1828 WRITE_ONCE(pernet->subflows_max, subflows); 1829 1830unlock: 1831 spin_unlock_bh(&pernet->lock); 1832 return ret; 1833} 1834 1835static int 1836mptcp_nl_cmd_get_limits(struct sk_buff *skb, struct genl_info *info) 1837{ 1838 struct pm_nl_pernet *pernet = genl_info_pm_nl(info); 1839 struct sk_buff *msg; 1840 void *reply; 1841 1842 msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL); 1843 if (!msg) 1844 return -ENOMEM; 1845 1846 reply = genlmsg_put_reply(msg, info, &mptcp_genl_family, 0, 1847 MPTCP_PM_CMD_GET_LIMITS); 1848 if (!reply) 1849 goto fail; 1850 1851 if (nla_put_u32(msg, MPTCP_PM_ATTR_RCV_ADD_ADDRS, 1852 READ_ONCE(pernet->add_addr_accept_max))) 1853 goto fail; 1854 1855 if (nla_put_u32(msg, MPTCP_PM_ATTR_SUBFLOWS, 1856 READ_ONCE(pernet->subflows_max))) 1857 goto fail; 1858 1859 genlmsg_end(msg, reply); 1860 return genlmsg_reply(msg, info); 1861 1862fail: 1863 GENL_SET_ERR_MSG(info, "not enough space in Netlink message"); 1864 nlmsg_free(msg); 1865 return -EMSGSIZE; 1866} 1867 1868static void mptcp_pm_nl_fullmesh(struct mptcp_sock *msk, 1869 struct mptcp_addr_info *addr) 1870{ 1871 struct mptcp_rm_list list = { .nr = 0 }; 1872 1873 list.ids[list.nr++] = addr->id; 1874 1875 spin_lock_bh(&msk->pm.lock); 1876 mptcp_pm_nl_rm_subflow_received(msk, &list); 1877 mptcp_pm_create_subflow_or_signal_addr(msk); 1878 spin_unlock_bh(&msk->pm.lock); 1879} 1880 1881static int mptcp_nl_set_flags(struct net *net, 1882 struct mptcp_addr_info *addr, 1883 u8 bkup, u8 changed) 1884{ 1885 long s_slot = 0, s_num = 0; 1886 struct mptcp_sock *msk; 1887 int ret = -EINVAL; 1888 1889 while ((msk = mptcp_token_iter_next(net, &s_slot, &s_num)) != NULL) { 1890 struct sock *sk = (struct sock *)msk; 1891 1892 if (list_empty(&msk->conn_list) || mptcp_pm_is_userspace(msk)) 1893 goto next; 1894 1895 lock_sock(sk); 1896 if (changed & MPTCP_PM_ADDR_FLAG_BACKUP) 1897 ret = mptcp_pm_nl_mp_prio_send_ack(msk, addr, NULL, bkup); 1898 if (changed & MPTCP_PM_ADDR_FLAG_FULLMESH) 1899 mptcp_pm_nl_fullmesh(msk, addr); 1900 release_sock(sk); 1901 1902next: 1903 sock_put(sk); 1904 cond_resched(); 1905 } 1906 1907 return ret; 1908} 1909 1910int mptcp_pm_nl_set_flags(struct net *net, struct mptcp_pm_addr_entry *addr, u8 bkup) 1911{ 1912 struct pm_nl_pernet *pernet = pm_nl_get_pernet(net); 1913 u8 changed, mask = MPTCP_PM_ADDR_FLAG_BACKUP | 1914 MPTCP_PM_ADDR_FLAG_FULLMESH; 1915 struct mptcp_pm_addr_entry *entry; 1916 u8 lookup_by_id = 0; 1917 1918 if (addr->addr.family == AF_UNSPEC) { 1919 lookup_by_id = 1; 1920 if (!addr->addr.id) 1921 return -EOPNOTSUPP; 1922 } 1923 1924 spin_lock_bh(&pernet->lock); 1925 entry = __lookup_addr(pernet, &addr->addr, lookup_by_id); 1926 if (!entry) { 1927 spin_unlock_bh(&pernet->lock); 1928 return -EINVAL; 1929 } 1930 if ((addr->flags & MPTCP_PM_ADDR_FLAG_FULLMESH) && 1931 (entry->flags & MPTCP_PM_ADDR_FLAG_SIGNAL)) { 1932 spin_unlock_bh(&pernet->lock); 1933 return -EINVAL; 1934 } 1935 1936 changed = (addr->flags ^ entry->flags) & mask; 1937 entry->flags = (entry->flags & ~mask) | (addr->flags & mask); 1938 *addr = *entry; 1939 spin_unlock_bh(&pernet->lock); 1940 1941 mptcp_nl_set_flags(net, &addr->addr, bkup, changed); 1942 return 0; 1943} 1944 1945static int mptcp_nl_cmd_set_flags(struct sk_buff *skb, struct genl_info *info) 1946{ 1947 struct mptcp_pm_addr_entry remote = { .addr = { .family = AF_UNSPEC }, }; 1948 struct mptcp_pm_addr_entry addr = { .addr = { .family = AF_UNSPEC }, }; 1949 struct nlattr *attr_rem = info->attrs[MPTCP_PM_ATTR_ADDR_REMOTE]; 1950 struct nlattr *token = info->attrs[MPTCP_PM_ATTR_TOKEN]; 1951 struct nlattr *attr = info->attrs[MPTCP_PM_ATTR_ADDR]; 1952 struct net *net = sock_net(skb->sk); 1953 u8 bkup = 0; 1954 int ret; 1955 1956 ret = mptcp_pm_parse_entry(attr, info, false, &addr); 1957 if (ret < 0) 1958 return ret; 1959 1960 if (attr_rem) { 1961 ret = mptcp_pm_parse_entry(attr_rem, info, false, &remote); 1962 if (ret < 0) 1963 return ret; 1964 } 1965 1966 if (addr.flags & MPTCP_PM_ADDR_FLAG_BACKUP) 1967 bkup = 1; 1968 1969 return mptcp_pm_set_flags(net, token, &addr, &remote, bkup); 1970} 1971 1972static void mptcp_nl_mcast_send(struct net *net, struct sk_buff *nlskb, gfp_t gfp) 1973{ 1974 genlmsg_multicast_netns(&mptcp_genl_family, net, 1975 nlskb, 0, MPTCP_PM_EV_GRP_OFFSET, gfp); 1976} 1977 1978bool mptcp_userspace_pm_active(const struct mptcp_sock *msk) 1979{ 1980 return genl_has_listeners(&mptcp_genl_family, 1981 sock_net((const struct sock *)msk), 1982 MPTCP_PM_EV_GRP_OFFSET); 1983} 1984 1985static int mptcp_event_add_subflow(struct sk_buff *skb, const struct sock *ssk) 1986{ 1987 const struct inet_sock *issk = inet_sk(ssk); 1988 const struct mptcp_subflow_context *sf; 1989 1990 if (nla_put_u16(skb, MPTCP_ATTR_FAMILY, ssk->sk_family)) 1991 return -EMSGSIZE; 1992 1993 switch (ssk->sk_family) { 1994 case AF_INET: 1995 if (nla_put_in_addr(skb, MPTCP_ATTR_SADDR4, issk->inet_saddr)) 1996 return -EMSGSIZE; 1997 if (nla_put_in_addr(skb, MPTCP_ATTR_DADDR4, issk->inet_daddr)) 1998 return -EMSGSIZE; 1999 break; 2000#if IS_ENABLED(CONFIG_MPTCP_IPV6) 2001 case AF_INET6: { 2002 const struct ipv6_pinfo *np = inet6_sk(ssk); 2003 2004 if (nla_put_in6_addr(skb, MPTCP_ATTR_SADDR6, &np->saddr)) 2005 return -EMSGSIZE; 2006 if (nla_put_in6_addr(skb, MPTCP_ATTR_DADDR6, &ssk->sk_v6_daddr)) 2007 return -EMSGSIZE; 2008 break; 2009 } 2010#endif 2011 default: 2012 WARN_ON_ONCE(1); 2013 return -EMSGSIZE; 2014 } 2015 2016 if (nla_put_be16(skb, MPTCP_ATTR_SPORT, issk->inet_sport)) 2017 return -EMSGSIZE; 2018 if (nla_put_be16(skb, MPTCP_ATTR_DPORT, issk->inet_dport)) 2019 return -EMSGSIZE; 2020 2021 sf = mptcp_subflow_ctx(ssk); 2022 if (WARN_ON_ONCE(!sf)) 2023 return -EINVAL; 2024 2025 if (nla_put_u8(skb, MPTCP_ATTR_LOC_ID, subflow_get_local_id(sf))) 2026 return -EMSGSIZE; 2027 2028 if (nla_put_u8(skb, MPTCP_ATTR_REM_ID, sf->remote_id)) 2029 return -EMSGSIZE; 2030 2031 return 0; 2032} 2033 2034static int mptcp_event_put_token_and_ssk(struct sk_buff *skb, 2035 const struct mptcp_sock *msk, 2036 const struct sock *ssk) 2037{ 2038 const struct sock *sk = (const struct sock *)msk; 2039 const struct mptcp_subflow_context *sf; 2040 u8 sk_err; 2041 2042 if (nla_put_u32(skb, MPTCP_ATTR_TOKEN, msk->token)) 2043 return -EMSGSIZE; 2044 2045 if (mptcp_event_add_subflow(skb, ssk)) 2046 return -EMSGSIZE; 2047 2048 sf = mptcp_subflow_ctx(ssk); 2049 if (WARN_ON_ONCE(!sf)) 2050 return -EINVAL; 2051 2052 if (nla_put_u8(skb, MPTCP_ATTR_BACKUP, sf->backup)) 2053 return -EMSGSIZE; 2054 2055 if (ssk->sk_bound_dev_if && 2056 nla_put_s32(skb, MPTCP_ATTR_IF_IDX, ssk->sk_bound_dev_if)) 2057 return -EMSGSIZE; 2058 2059 sk_err = READ_ONCE(ssk->sk_err); 2060 if (sk_err && sk->sk_state == TCP_ESTABLISHED && 2061 nla_put_u8(skb, MPTCP_ATTR_ERROR, sk_err)) 2062 return -EMSGSIZE; 2063 2064 return 0; 2065} 2066 2067static int mptcp_event_sub_established(struct sk_buff *skb, 2068 const struct mptcp_sock *msk, 2069 const struct sock *ssk) 2070{ 2071 return mptcp_event_put_token_and_ssk(skb, msk, ssk); 2072} 2073 2074static int mptcp_event_sub_closed(struct sk_buff *skb, 2075 const struct mptcp_sock *msk, 2076 const struct sock *ssk) 2077{ 2078 const struct mptcp_subflow_context *sf; 2079 2080 if (mptcp_event_put_token_and_ssk(skb, msk, ssk)) 2081 return -EMSGSIZE; 2082 2083 sf = mptcp_subflow_ctx(ssk); 2084 if (!sf->reset_seen) 2085 return 0; 2086 2087 if (nla_put_u32(skb, MPTCP_ATTR_RESET_REASON, sf->reset_reason)) 2088 return -EMSGSIZE; 2089 2090 if (nla_put_u32(skb, MPTCP_ATTR_RESET_FLAGS, sf->reset_transient)) 2091 return -EMSGSIZE; 2092 2093 return 0; 2094} 2095 2096static int mptcp_event_created(struct sk_buff *skb, 2097 const struct mptcp_sock *msk, 2098 const struct sock *ssk) 2099{ 2100 int err = nla_put_u32(skb, MPTCP_ATTR_TOKEN, msk->token); 2101 2102 if (err) 2103 return err; 2104 2105 if (nla_put_u8(skb, MPTCP_ATTR_SERVER_SIDE, READ_ONCE(msk->pm.server_side))) 2106 return -EMSGSIZE; 2107 2108 return mptcp_event_add_subflow(skb, ssk); 2109} 2110 2111void mptcp_event_addr_removed(const struct mptcp_sock *msk, uint8_t id) 2112{ 2113 struct net *net = sock_net((const struct sock *)msk); 2114 struct nlmsghdr *nlh; 2115 struct sk_buff *skb; 2116 2117 if (!genl_has_listeners(&mptcp_genl_family, net, MPTCP_PM_EV_GRP_OFFSET)) 2118 return; 2119 2120 skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC); 2121 if (!skb) 2122 return; 2123 2124 nlh = genlmsg_put(skb, 0, 0, &mptcp_genl_family, 0, MPTCP_EVENT_REMOVED); 2125 if (!nlh) 2126 goto nla_put_failure; 2127 2128 if (nla_put_u32(skb, MPTCP_ATTR_TOKEN, msk->token)) 2129 goto nla_put_failure; 2130 2131 if (nla_put_u8(skb, MPTCP_ATTR_REM_ID, id)) 2132 goto nla_put_failure; 2133 2134 genlmsg_end(skb, nlh); 2135 mptcp_nl_mcast_send(net, skb, GFP_ATOMIC); 2136 return; 2137 2138nla_put_failure: 2139 nlmsg_free(skb); 2140} 2141 2142void mptcp_event_addr_announced(const struct sock *ssk, 2143 const struct mptcp_addr_info *info) 2144{ 2145 struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk); 2146 struct mptcp_sock *msk = mptcp_sk(subflow->conn); 2147 struct net *net = sock_net(ssk); 2148 struct nlmsghdr *nlh; 2149 struct sk_buff *skb; 2150 2151 if (!genl_has_listeners(&mptcp_genl_family, net, MPTCP_PM_EV_GRP_OFFSET)) 2152 return; 2153 2154 skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC); 2155 if (!skb) 2156 return; 2157 2158 nlh = genlmsg_put(skb, 0, 0, &mptcp_genl_family, 0, 2159 MPTCP_EVENT_ANNOUNCED); 2160 if (!nlh) 2161 goto nla_put_failure; 2162 2163 if (nla_put_u32(skb, MPTCP_ATTR_TOKEN, msk->token)) 2164 goto nla_put_failure; 2165 2166 if (nla_put_u8(skb, MPTCP_ATTR_REM_ID, info->id)) 2167 goto nla_put_failure; 2168 2169 if (nla_put_be16(skb, MPTCP_ATTR_DPORT, 2170 info->port == 0 ? 2171 inet_sk(ssk)->inet_dport : 2172 info->port)) 2173 goto nla_put_failure; 2174 2175 switch (info->family) { 2176 case AF_INET: 2177 if (nla_put_in_addr(skb, MPTCP_ATTR_DADDR4, info->addr.s_addr)) 2178 goto nla_put_failure; 2179 break; 2180#if IS_ENABLED(CONFIG_MPTCP_IPV6) 2181 case AF_INET6: 2182 if (nla_put_in6_addr(skb, MPTCP_ATTR_DADDR6, &info->addr6)) 2183 goto nla_put_failure; 2184 break; 2185#endif 2186 default: 2187 WARN_ON_ONCE(1); 2188 goto nla_put_failure; 2189 } 2190 2191 genlmsg_end(skb, nlh); 2192 mptcp_nl_mcast_send(net, skb, GFP_ATOMIC); 2193 return; 2194 2195nla_put_failure: 2196 nlmsg_free(skb); 2197} 2198 2199void mptcp_event_pm_listener(const struct sock *ssk, 2200 enum mptcp_event_type event) 2201{ 2202 const struct inet_sock *issk = inet_sk(ssk); 2203 struct net *net = sock_net(ssk); 2204 struct nlmsghdr *nlh; 2205 struct sk_buff *skb; 2206 2207 if (!genl_has_listeners(&mptcp_genl_family, net, MPTCP_PM_EV_GRP_OFFSET)) 2208 return; 2209 2210 skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL); 2211 if (!skb) 2212 return; 2213 2214 nlh = genlmsg_put(skb, 0, 0, &mptcp_genl_family, 0, event); 2215 if (!nlh) 2216 goto nla_put_failure; 2217 2218 if (nla_put_u16(skb, MPTCP_ATTR_FAMILY, ssk->sk_family)) 2219 goto nla_put_failure; 2220 2221 if (nla_put_be16(skb, MPTCP_ATTR_SPORT, issk->inet_sport)) 2222 goto nla_put_failure; 2223 2224 switch (ssk->sk_family) { 2225 case AF_INET: 2226 if (nla_put_in_addr(skb, MPTCP_ATTR_SADDR4, issk->inet_saddr)) 2227 goto nla_put_failure; 2228 break; 2229#if IS_ENABLED(CONFIG_MPTCP_IPV6) 2230 case AF_INET6: { 2231 const struct ipv6_pinfo *np = inet6_sk(ssk); 2232 2233 if (nla_put_in6_addr(skb, MPTCP_ATTR_SADDR6, &np->saddr)) 2234 goto nla_put_failure; 2235 break; 2236 } 2237#endif 2238 default: 2239 WARN_ON_ONCE(1); 2240 goto nla_put_failure; 2241 } 2242 2243 genlmsg_end(skb, nlh); 2244 mptcp_nl_mcast_send(net, skb, GFP_KERNEL); 2245 return; 2246 2247nla_put_failure: 2248 nlmsg_free(skb); 2249} 2250 2251void mptcp_event(enum mptcp_event_type type, const struct mptcp_sock *msk, 2252 const struct sock *ssk, gfp_t gfp) 2253{ 2254 struct net *net = sock_net((const struct sock *)msk); 2255 struct nlmsghdr *nlh; 2256 struct sk_buff *skb; 2257 2258 if (!genl_has_listeners(&mptcp_genl_family, net, MPTCP_PM_EV_GRP_OFFSET)) 2259 return; 2260 2261 skb = nlmsg_new(NLMSG_DEFAULT_SIZE, gfp); 2262 if (!skb) 2263 return; 2264 2265 nlh = genlmsg_put(skb, 0, 0, &mptcp_genl_family, 0, type); 2266 if (!nlh) 2267 goto nla_put_failure; 2268 2269 switch (type) { 2270 case MPTCP_EVENT_UNSPEC: 2271 WARN_ON_ONCE(1); 2272 break; 2273 case MPTCP_EVENT_CREATED: 2274 case MPTCP_EVENT_ESTABLISHED: 2275 if (mptcp_event_created(skb, msk, ssk) < 0) 2276 goto nla_put_failure; 2277 break; 2278 case MPTCP_EVENT_CLOSED: 2279 if (nla_put_u32(skb, MPTCP_ATTR_TOKEN, msk->token) < 0) 2280 goto nla_put_failure; 2281 break; 2282 case MPTCP_EVENT_ANNOUNCED: 2283 case MPTCP_EVENT_REMOVED: 2284 /* call mptcp_event_addr_announced()/removed instead */ 2285 WARN_ON_ONCE(1); 2286 break; 2287 case MPTCP_EVENT_SUB_ESTABLISHED: 2288 case MPTCP_EVENT_SUB_PRIORITY: 2289 if (mptcp_event_sub_established(skb, msk, ssk) < 0) 2290 goto nla_put_failure; 2291 break; 2292 case MPTCP_EVENT_SUB_CLOSED: 2293 if (mptcp_event_sub_closed(skb, msk, ssk) < 0) 2294 goto nla_put_failure; 2295 break; 2296 case MPTCP_EVENT_LISTENER_CREATED: 2297 case MPTCP_EVENT_LISTENER_CLOSED: 2298 break; 2299 } 2300 2301 genlmsg_end(skb, nlh); 2302 mptcp_nl_mcast_send(net, skb, gfp); 2303 return; 2304 2305nla_put_failure: 2306 nlmsg_free(skb); 2307} 2308 2309static const struct genl_small_ops mptcp_pm_ops[] = { 2310 { 2311 .cmd = MPTCP_PM_CMD_ADD_ADDR, 2312 .doit = mptcp_nl_cmd_add_addr, 2313 .flags = GENL_UNS_ADMIN_PERM, 2314 }, 2315 { 2316 .cmd = MPTCP_PM_CMD_DEL_ADDR, 2317 .doit = mptcp_nl_cmd_del_addr, 2318 .flags = GENL_UNS_ADMIN_PERM, 2319 }, 2320 { 2321 .cmd = MPTCP_PM_CMD_FLUSH_ADDRS, 2322 .doit = mptcp_nl_cmd_flush_addrs, 2323 .flags = GENL_UNS_ADMIN_PERM, 2324 }, 2325 { 2326 .cmd = MPTCP_PM_CMD_GET_ADDR, 2327 .doit = mptcp_nl_cmd_get_addr, 2328 .dumpit = mptcp_nl_cmd_dump_addrs, 2329 }, 2330 { 2331 .cmd = MPTCP_PM_CMD_SET_LIMITS, 2332 .doit = mptcp_nl_cmd_set_limits, 2333 .flags = GENL_UNS_ADMIN_PERM, 2334 }, 2335 { 2336 .cmd = MPTCP_PM_CMD_GET_LIMITS, 2337 .doit = mptcp_nl_cmd_get_limits, 2338 }, 2339 { 2340 .cmd = MPTCP_PM_CMD_SET_FLAGS, 2341 .doit = mptcp_nl_cmd_set_flags, 2342 .flags = GENL_UNS_ADMIN_PERM, 2343 }, 2344 { 2345 .cmd = MPTCP_PM_CMD_ANNOUNCE, 2346 .doit = mptcp_nl_cmd_announce, 2347 .flags = GENL_UNS_ADMIN_PERM, 2348 }, 2349 { 2350 .cmd = MPTCP_PM_CMD_REMOVE, 2351 .doit = mptcp_nl_cmd_remove, 2352 .flags = GENL_UNS_ADMIN_PERM, 2353 }, 2354 { 2355 .cmd = MPTCP_PM_CMD_SUBFLOW_CREATE, 2356 .doit = mptcp_nl_cmd_sf_create, 2357 .flags = GENL_UNS_ADMIN_PERM, 2358 }, 2359 { 2360 .cmd = MPTCP_PM_CMD_SUBFLOW_DESTROY, 2361 .doit = mptcp_nl_cmd_sf_destroy, 2362 .flags = GENL_UNS_ADMIN_PERM, 2363 }, 2364}; 2365 2366static struct genl_family mptcp_genl_family __ro_after_init = { 2367 .name = MPTCP_PM_NAME, 2368 .version = MPTCP_PM_VER, 2369 .maxattr = MPTCP_PM_ATTR_MAX, 2370 .policy = mptcp_pm_policy, 2371 .netnsok = true, 2372 .module = THIS_MODULE, 2373 .small_ops = mptcp_pm_ops, 2374 .n_small_ops = ARRAY_SIZE(mptcp_pm_ops), 2375 .resv_start_op = MPTCP_PM_CMD_SUBFLOW_DESTROY + 1, 2376 .mcgrps = mptcp_pm_mcgrps, 2377 .n_mcgrps = ARRAY_SIZE(mptcp_pm_mcgrps), 2378}; 2379 2380static int __net_init pm_nl_init_net(struct net *net) 2381{ 2382 struct pm_nl_pernet *pernet = pm_nl_get_pernet(net); 2383 2384 INIT_LIST_HEAD_RCU(&pernet->local_addr_list); 2385 2386 /* Cit. 2 subflows ought to be enough for anybody. */ 2387 pernet->subflows_max = 2; 2388 pernet->next_id = 1; 2389 pernet->stale_loss_cnt = 4; 2390 spin_lock_init(&pernet->lock); 2391 2392 /* No need to initialize other pernet fields, the struct is zeroed at 2393 * allocation time. 2394 */ 2395 2396 return 0; 2397} 2398 2399static void __net_exit pm_nl_exit_net(struct list_head *net_list) 2400{ 2401 struct net *net; 2402 2403 list_for_each_entry(net, net_list, exit_list) { 2404 struct pm_nl_pernet *pernet = pm_nl_get_pernet(net); 2405 2406 /* net is removed from namespace list, can't race with 2407 * other modifiers, also netns core already waited for a 2408 * RCU grace period. 2409 */ 2410 __flush_addrs(&pernet->local_addr_list); 2411 } 2412} 2413 2414static struct pernet_operations mptcp_pm_pernet_ops = { 2415 .init = pm_nl_init_net, 2416 .exit_batch = pm_nl_exit_net, 2417 .id = &pm_nl_pernet_id, 2418 .size = sizeof(struct pm_nl_pernet), 2419}; 2420 2421void __init mptcp_pm_nl_init(void) 2422{ 2423 if (register_pernet_subsys(&mptcp_pm_pernet_ops) < 0) 2424 panic("Failed to register MPTCP PM pernet subsystem.\n"); 2425 2426 if (genl_register_family(&mptcp_genl_family)) 2427 panic("Failed to register MPTCP PM netlink family\n"); 2428} 2429