uipc_socket.c revision 1.160.2.3
1 /* $NetBSD: uipc_socket.c,v 1.160.2.3 2009/09/16 13:38:01 yamt Exp $ */ 2 3 /*- 4 * Copyright (c) 2002, 2007, 2008, 2009 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Jason R. Thorpe of Wasabi Systems, Inc, and by Andrew Doran. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 19 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 20 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 21 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 22 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 23 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 24 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 25 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 26 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 27 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 28 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 29 * POSSIBILITY OF SUCH DAMAGE. 30 */ 31 32 /* 33 * Copyright (c) 2004 The FreeBSD Foundation 34 * Copyright (c) 2004 Robert Watson 35 * Copyright (c) 1982, 1986, 1988, 1990, 1993 36 * The Regents of the University of California. All rights reserved. 37 * 38 * Redistribution and use in source and binary forms, with or without 39 * modification, are permitted provided that the following conditions 40 * are met: 41 * 1. Redistributions of source code must retain the above copyright 42 * notice, this list of conditions and the following disclaimer. 43 * 2. Redistributions in binary form must reproduce the above copyright 44 * notice, this list of conditions and the following disclaimer in the 45 * documentation and/or other materials provided with the distribution. 46 * 3. Neither the name of the University nor the names of its contributors 47 * may be used to endorse or promote products derived from this software 48 * without specific prior written permission. 49 * 50 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 51 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 52 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 53 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 54 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 55 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 56 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 57 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 58 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 59 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 60 * SUCH DAMAGE. 61 * 62 * @(#)uipc_socket.c 8.6 (Berkeley) 5/2/95 63 */ 64 65 #include <sys/cdefs.h> 66 __KERNEL_RCSID(0, "$NetBSD: uipc_socket.c,v 1.160.2.3 2009/09/16 13:38:01 yamt Exp $"); 67 68 #include "opt_compat_netbsd.h" 69 #include "opt_sock_counters.h" 70 #include "opt_sosend_loan.h" 71 #include "opt_mbuftrace.h" 72 #include "opt_somaxkva.h" 73 #include "opt_multiprocessor.h" /* XXX */ 74 75 #include <sys/param.h> 76 #include <sys/systm.h> 77 #include <sys/proc.h> 78 #include <sys/file.h> 79 #include <sys/filedesc.h> 80 #include <sys/kmem.h> 81 #include <sys/mbuf.h> 82 #include <sys/domain.h> 83 #include <sys/kernel.h> 84 #include <sys/protosw.h> 85 #include <sys/socket.h> 86 #include <sys/socketvar.h> 87 #include <sys/signalvar.h> 88 #include <sys/resourcevar.h> 89 #include <sys/uidinfo.h> 90 #include <sys/event.h> 91 #include <sys/poll.h> 92 #include <sys/kauth.h> 93 #include <sys/mutex.h> 94 #include <sys/condvar.h> 95 96 #ifdef COMPAT_50 97 #include <compat/sys/time.h> 98 #include <compat/sys/socket.h> 99 #endif 100 101 #include <uvm/uvm.h> 102 103 MALLOC_DEFINE(M_SOOPTS, "soopts", "socket options"); 104 MALLOC_DEFINE(M_SONAME, "soname", "socket name"); 105 106 extern const struct fileops socketops; 107 108 extern int somaxconn; /* patchable (XXX sysctl) */ 109 int somaxconn = SOMAXCONN; 110 kmutex_t *softnet_lock; 111 112 #ifdef SOSEND_COUNTERS 113 #include <sys/device.h> 114 115 static struct evcnt sosend_loan_big = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 116 NULL, "sosend", "loan big"); 117 static struct evcnt sosend_copy_big = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 118 NULL, "sosend", "copy big"); 119 static struct evcnt sosend_copy_small = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 120 NULL, "sosend", "copy small"); 121 static struct evcnt sosend_kvalimit = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 122 NULL, "sosend", "kva limit"); 123 124 #define SOSEND_COUNTER_INCR(ev) (ev)->ev_count++ 125 126 EVCNT_ATTACH_STATIC(sosend_loan_big); 127 EVCNT_ATTACH_STATIC(sosend_copy_big); 128 EVCNT_ATTACH_STATIC(sosend_copy_small); 129 EVCNT_ATTACH_STATIC(sosend_kvalimit); 130 #else 131 132 #define SOSEND_COUNTER_INCR(ev) /* nothing */ 133 134 #endif /* SOSEND_COUNTERS */ 135 136 static struct callback_entry sokva_reclaimerentry; 137 138 #if defined(SOSEND_NO_LOAN) || defined(MULTIPROCESSOR) 139 int sock_loan_thresh = -1; 140 #else 141 int sock_loan_thresh = 4096; 142 #endif 143 144 static kmutex_t so_pendfree_lock; 145 static struct mbuf *so_pendfree; 146 147 #ifndef SOMAXKVA 148 #define SOMAXKVA (16 * 1024 * 1024) 149 #endif 150 int somaxkva = SOMAXKVA; 151 static int socurkva; 152 static kcondvar_t socurkva_cv; 153 154 #define SOCK_LOAN_CHUNK 65536 155 156 static size_t sodopendfree(void); 157 static size_t sodopendfreel(void); 158 159 static void sysctl_kern_somaxkva_setup(void); 160 static struct sysctllog *socket_sysctllog; 161 162 static vsize_t 163 sokvareserve(struct socket *so, vsize_t len) 164 { 165 int error; 166 167 mutex_enter(&so_pendfree_lock); 168 while (socurkva + len > somaxkva) { 169 size_t freed; 170 171 /* 172 * try to do pendfree. 173 */ 174 175 freed = sodopendfreel(); 176 177 /* 178 * if some kva was freed, try again. 179 */ 180 181 if (freed) 182 continue; 183 184 SOSEND_COUNTER_INCR(&sosend_kvalimit); 185 error = cv_wait_sig(&socurkva_cv, &so_pendfree_lock); 186 if (error) { 187 len = 0; 188 break; 189 } 190 } 191 socurkva += len; 192 mutex_exit(&so_pendfree_lock); 193 return len; 194 } 195 196 static void 197 sokvaunreserve(vsize_t len) 198 { 199 200 mutex_enter(&so_pendfree_lock); 201 socurkva -= len; 202 cv_broadcast(&socurkva_cv); 203 mutex_exit(&so_pendfree_lock); 204 } 205 206 /* 207 * sokvaalloc: allocate kva for loan. 208 */ 209 210 vaddr_t 211 sokvaalloc(vsize_t len, struct socket *so) 212 { 213 vaddr_t lva; 214 215 /* 216 * reserve kva. 217 */ 218 219 if (sokvareserve(so, len) == 0) 220 return 0; 221 222 /* 223 * allocate kva. 224 */ 225 226 lva = uvm_km_alloc(kernel_map, len, 0, UVM_KMF_VAONLY | UVM_KMF_WAITVA); 227 if (lva == 0) { 228 sokvaunreserve(len); 229 return (0); 230 } 231 232 return lva; 233 } 234 235 /* 236 * sokvafree: free kva for loan. 237 */ 238 239 void 240 sokvafree(vaddr_t sva, vsize_t len) 241 { 242 243 /* 244 * free kva. 245 */ 246 247 uvm_km_free(kernel_map, sva, len, UVM_KMF_VAONLY); 248 249 /* 250 * unreserve kva. 251 */ 252 253 sokvaunreserve(len); 254 } 255 256 static void 257 sodoloanfree(struct vm_page **pgs, void *buf, size_t size) 258 { 259 vaddr_t sva, eva; 260 vsize_t len; 261 int npgs; 262 263 KASSERT(pgs != NULL); 264 265 eva = round_page((vaddr_t) buf + size); 266 sva = trunc_page((vaddr_t) buf); 267 len = eva - sva; 268 npgs = len >> PAGE_SHIFT; 269 270 pmap_kremove(sva, len); 271 pmap_update(pmap_kernel()); 272 uvm_unloan(pgs, npgs, UVM_LOAN_TOPAGE); 273 sokvafree(sva, len); 274 } 275 276 static size_t 277 sodopendfree(void) 278 { 279 size_t rv; 280 281 if (__predict_true(so_pendfree == NULL)) 282 return 0; 283 284 mutex_enter(&so_pendfree_lock); 285 rv = sodopendfreel(); 286 mutex_exit(&so_pendfree_lock); 287 288 return rv; 289 } 290 291 /* 292 * sodopendfreel: free mbufs on "pendfree" list. 293 * unlock and relock so_pendfree_lock when freeing mbufs. 294 * 295 * => called with so_pendfree_lock held. 296 */ 297 298 static size_t 299 sodopendfreel(void) 300 { 301 struct mbuf *m, *next; 302 size_t rv = 0; 303 304 KASSERT(mutex_owned(&so_pendfree_lock)); 305 306 while (so_pendfree != NULL) { 307 m = so_pendfree; 308 so_pendfree = NULL; 309 mutex_exit(&so_pendfree_lock); 310 311 for (; m != NULL; m = next) { 312 next = m->m_next; 313 KASSERT((~m->m_flags & (M_EXT|M_EXT_PAGES)) == 0); 314 KASSERT(m->m_ext.ext_refcnt == 0); 315 316 rv += m->m_ext.ext_size; 317 sodoloanfree(m->m_ext.ext_pgs, m->m_ext.ext_buf, 318 m->m_ext.ext_size); 319 pool_cache_put(mb_cache, m); 320 } 321 322 mutex_enter(&so_pendfree_lock); 323 } 324 325 return (rv); 326 } 327 328 void 329 soloanfree(struct mbuf *m, void *buf, size_t size, void *arg) 330 { 331 332 KASSERT(m != NULL); 333 334 /* 335 * postpone freeing mbuf. 336 * 337 * we can't do it in interrupt context 338 * because we need to put kva back to kernel_map. 339 */ 340 341 mutex_enter(&so_pendfree_lock); 342 m->m_next = so_pendfree; 343 so_pendfree = m; 344 cv_broadcast(&socurkva_cv); 345 mutex_exit(&so_pendfree_lock); 346 } 347 348 static long 349 sosend_loan(struct socket *so, struct uio *uio, struct mbuf *m, long space) 350 { 351 struct iovec *iov = uio->uio_iov; 352 vaddr_t sva, eva; 353 vsize_t len; 354 vaddr_t lva; 355 int npgs, error; 356 vaddr_t va; 357 int i; 358 359 if (VMSPACE_IS_KERNEL_P(uio->uio_vmspace)) 360 return (0); 361 362 if (iov->iov_len < (size_t) space) 363 space = iov->iov_len; 364 if (space > SOCK_LOAN_CHUNK) 365 space = SOCK_LOAN_CHUNK; 366 367 eva = round_page((vaddr_t) iov->iov_base + space); 368 sva = trunc_page((vaddr_t) iov->iov_base); 369 len = eva - sva; 370 npgs = len >> PAGE_SHIFT; 371 372 KASSERT(npgs <= M_EXT_MAXPAGES); 373 374 lva = sokvaalloc(len, so); 375 if (lva == 0) 376 return 0; 377 378 error = uvm_loan(&uio->uio_vmspace->vm_map, sva, len, 379 m->m_ext.ext_pgs, UVM_LOAN_TOPAGE); 380 if (error) { 381 sokvafree(lva, len); 382 return (0); 383 } 384 385 for (i = 0, va = lva; i < npgs; i++, va += PAGE_SIZE) 386 pmap_kenter_pa(va, VM_PAGE_TO_PHYS(m->m_ext.ext_pgs[i]), 387 VM_PROT_READ); 388 pmap_update(pmap_kernel()); 389 390 lva += (vaddr_t) iov->iov_base & PAGE_MASK; 391 392 MEXTADD(m, (void *) lva, space, M_MBUF, soloanfree, so); 393 m->m_flags |= M_EXT_PAGES | M_EXT_ROMAP; 394 395 uio->uio_resid -= space; 396 /* uio_offset not updated, not set/used for write(2) */ 397 uio->uio_iov->iov_base = (char *)uio->uio_iov->iov_base + space; 398 uio->uio_iov->iov_len -= space; 399 if (uio->uio_iov->iov_len == 0) { 400 uio->uio_iov++; 401 uio->uio_iovcnt--; 402 } 403 404 return (space); 405 } 406 407 static int 408 sokva_reclaim_callback(struct callback_entry *ce, void *obj, void *arg) 409 { 410 411 KASSERT(ce == &sokva_reclaimerentry); 412 KASSERT(obj == NULL); 413 414 sodopendfree(); 415 if (!vm_map_starved_p(kernel_map)) { 416 return CALLBACK_CHAIN_ABORT; 417 } 418 return CALLBACK_CHAIN_CONTINUE; 419 } 420 421 struct mbuf * 422 getsombuf(struct socket *so, int type) 423 { 424 struct mbuf *m; 425 426 m = m_get(M_WAIT, type); 427 MCLAIM(m, so->so_mowner); 428 return m; 429 } 430 431 void 432 soinit(void) 433 { 434 435 sysctl_kern_somaxkva_setup(); 436 437 mutex_init(&so_pendfree_lock, MUTEX_DEFAULT, IPL_VM); 438 softnet_lock = mutex_obj_alloc(MUTEX_DEFAULT, IPL_NONE); 439 cv_init(&socurkva_cv, "sokva"); 440 soinit2(); 441 442 /* Set the initial adjusted socket buffer size. */ 443 if (sb_max_set(sb_max)) 444 panic("bad initial sb_max value: %lu", sb_max); 445 446 callback_register(&vm_map_to_kernel(kernel_map)->vmk_reclaim_callback, 447 &sokva_reclaimerentry, NULL, sokva_reclaim_callback); 448 } 449 450 /* 451 * Socket operation routines. 452 * These routines are called by the routines in 453 * sys_socket.c or from a system process, and 454 * implement the semantics of socket operations by 455 * switching out to the protocol specific routines. 456 */ 457 /*ARGSUSED*/ 458 int 459 socreate(int dom, struct socket **aso, int type, int proto, struct lwp *l, 460 struct socket *lockso) 461 { 462 const struct protosw *prp; 463 struct socket *so; 464 uid_t uid; 465 int error; 466 kmutex_t *lock; 467 468 error = kauth_authorize_network(l->l_cred, KAUTH_NETWORK_SOCKET, 469 KAUTH_REQ_NETWORK_SOCKET_OPEN, KAUTH_ARG(dom), KAUTH_ARG(type), 470 KAUTH_ARG(proto)); 471 if (error != 0) 472 return error; 473 474 if (proto) 475 prp = pffindproto(dom, proto, type); 476 else 477 prp = pffindtype(dom, type); 478 if (prp == NULL) { 479 /* no support for domain */ 480 if (pffinddomain(dom) == 0) 481 return EAFNOSUPPORT; 482 /* no support for socket type */ 483 if (proto == 0 && type != 0) 484 return EPROTOTYPE; 485 return EPROTONOSUPPORT; 486 } 487 if (prp->pr_usrreq == NULL) 488 return EPROTONOSUPPORT; 489 if (prp->pr_type != type) 490 return EPROTOTYPE; 491 492 so = soget(true); 493 so->so_type = type; 494 so->so_proto = prp; 495 so->so_send = sosend; 496 so->so_receive = soreceive; 497 #ifdef MBUFTRACE 498 so->so_rcv.sb_mowner = &prp->pr_domain->dom_mowner; 499 so->so_snd.sb_mowner = &prp->pr_domain->dom_mowner; 500 so->so_mowner = &prp->pr_domain->dom_mowner; 501 #endif 502 uid = kauth_cred_geteuid(l->l_cred); 503 so->so_uidinfo = uid_find(uid); 504 so->so_egid = kauth_cred_getegid(l->l_cred); 505 so->so_cpid = l->l_proc->p_pid; 506 if (lockso != NULL) { 507 /* Caller wants us to share a lock. */ 508 lock = lockso->so_lock; 509 so->so_lock = lock; 510 mutex_obj_hold(lock); 511 mutex_enter(lock); 512 } else { 513 /* Lock assigned and taken during PRU_ATTACH. */ 514 } 515 error = (*prp->pr_usrreq)(so, PRU_ATTACH, NULL, 516 (struct mbuf *)(long)proto, NULL, l); 517 KASSERT(solocked(so)); 518 if (error != 0) { 519 so->so_state |= SS_NOFDREF; 520 sofree(so); 521 return error; 522 } 523 sounlock(so); 524 *aso = so; 525 return 0; 526 } 527 528 /* On success, write file descriptor to fdout and return zero. On 529 * failure, return non-zero; *fdout will be undefined. 530 */ 531 int 532 fsocreate(int domain, struct socket **sop, int type, int protocol, 533 struct lwp *l, int *fdout) 534 { 535 struct socket *so; 536 struct file *fp; 537 int fd, error; 538 539 if ((error = fd_allocfile(&fp, &fd)) != 0) 540 return (error); 541 fp->f_flag = FREAD|FWRITE; 542 fp->f_type = DTYPE_SOCKET; 543 fp->f_ops = &socketops; 544 error = socreate(domain, &so, type, protocol, l, NULL); 545 if (error != 0) { 546 fd_abort(curproc, fp, fd); 547 } else { 548 if (sop != NULL) 549 *sop = so; 550 fp->f_data = so; 551 fd_affix(curproc, fp, fd); 552 *fdout = fd; 553 } 554 return error; 555 } 556 557 int 558 sofamily(const struct socket *so) 559 { 560 const struct protosw *pr; 561 const struct domain *dom; 562 563 if ((pr = so->so_proto) == NULL) 564 return AF_UNSPEC; 565 if ((dom = pr->pr_domain) == NULL) 566 return AF_UNSPEC; 567 return dom->dom_family; 568 } 569 570 int 571 sobind(struct socket *so, struct mbuf *nam, struct lwp *l) 572 { 573 int error; 574 575 solock(so); 576 error = (*so->so_proto->pr_usrreq)(so, PRU_BIND, NULL, nam, NULL, l); 577 sounlock(so); 578 return error; 579 } 580 581 int 582 solisten(struct socket *so, int backlog, struct lwp *l) 583 { 584 int error; 585 586 solock(so); 587 if ((so->so_state & (SS_ISCONNECTED | SS_ISCONNECTING | 588 SS_ISDISCONNECTING)) != 0) { 589 sounlock(so); 590 return (EOPNOTSUPP); 591 } 592 error = (*so->so_proto->pr_usrreq)(so, PRU_LISTEN, NULL, 593 NULL, NULL, l); 594 if (error != 0) { 595 sounlock(so); 596 return error; 597 } 598 if (TAILQ_EMPTY(&so->so_q)) 599 so->so_options |= SO_ACCEPTCONN; 600 if (backlog < 0) 601 backlog = 0; 602 so->so_qlimit = min(backlog, somaxconn); 603 sounlock(so); 604 return 0; 605 } 606 607 void 608 sofree(struct socket *so) 609 { 610 u_int refs; 611 612 KASSERT(solocked(so)); 613 614 if (so->so_pcb || (so->so_state & SS_NOFDREF) == 0) { 615 sounlock(so); 616 return; 617 } 618 if (so->so_head) { 619 /* 620 * We must not decommission a socket that's on the accept(2) 621 * queue. If we do, then accept(2) may hang after select(2) 622 * indicated that the listening socket was ready. 623 */ 624 if (!soqremque(so, 0)) { 625 sounlock(so); 626 return; 627 } 628 } 629 if (so->so_rcv.sb_hiwat) 630 (void)chgsbsize(so->so_uidinfo, &so->so_rcv.sb_hiwat, 0, 631 RLIM_INFINITY); 632 if (so->so_snd.sb_hiwat) 633 (void)chgsbsize(so->so_uidinfo, &so->so_snd.sb_hiwat, 0, 634 RLIM_INFINITY); 635 sbrelease(&so->so_snd, so); 636 KASSERT(!cv_has_waiters(&so->so_cv)); 637 KASSERT(!cv_has_waiters(&so->so_rcv.sb_cv)); 638 KASSERT(!cv_has_waiters(&so->so_snd.sb_cv)); 639 sorflush(so); 640 refs = so->so_aborting; /* XXX */ 641 /* Remove acccept filter if one is present. */ 642 if (so->so_accf != NULL) 643 (void)accept_filt_clear(so); 644 sounlock(so); 645 if (refs == 0) /* XXX */ 646 soput(so); 647 } 648 649 /* 650 * Close a socket on last file table reference removal. 651 * Initiate disconnect if connected. 652 * Free socket when disconnect complete. 653 */ 654 int 655 soclose(struct socket *so) 656 { 657 struct socket *so2; 658 int error; 659 int error2; 660 661 error = 0; 662 solock(so); 663 if (so->so_options & SO_ACCEPTCONN) { 664 for (;;) { 665 if ((so2 = TAILQ_FIRST(&so->so_q0)) != 0) { 666 KASSERT(solocked2(so, so2)); 667 (void) soqremque(so2, 0); 668 /* soabort drops the lock. */ 669 (void) soabort(so2); 670 solock(so); 671 continue; 672 } 673 if ((so2 = TAILQ_FIRST(&so->so_q)) != 0) { 674 KASSERT(solocked2(so, so2)); 675 (void) soqremque(so2, 1); 676 /* soabort drops the lock. */ 677 (void) soabort(so2); 678 solock(so); 679 continue; 680 } 681 break; 682 } 683 } 684 if (so->so_pcb == 0) 685 goto discard; 686 if (so->so_state & SS_ISCONNECTED) { 687 if ((so->so_state & SS_ISDISCONNECTING) == 0) { 688 error = sodisconnect(so); 689 if (error) 690 goto drop; 691 } 692 if (so->so_options & SO_LINGER) { 693 if ((so->so_state & SS_ISDISCONNECTING) && so->so_nbio) 694 goto drop; 695 while (so->so_state & SS_ISCONNECTED) { 696 error = sowait(so, true, so->so_linger * hz); 697 if (error) 698 break; 699 } 700 } 701 } 702 drop: 703 if (so->so_pcb) { 704 error2 = (*so->so_proto->pr_usrreq)(so, PRU_DETACH, 705 NULL, NULL, NULL, NULL); 706 if (error == 0) 707 error = error2; 708 } 709 discard: 710 if (so->so_state & SS_NOFDREF) 711 panic("soclose: NOFDREF"); 712 so->so_state |= SS_NOFDREF; 713 sofree(so); 714 return (error); 715 } 716 717 /* 718 * Must be called with the socket locked.. Will return with it unlocked. 719 */ 720 int 721 soabort(struct socket *so) 722 { 723 u_int refs; 724 int error; 725 726 KASSERT(solocked(so)); 727 KASSERT(so->so_head == NULL); 728 729 so->so_aborting++; /* XXX */ 730 error = (*so->so_proto->pr_usrreq)(so, PRU_ABORT, NULL, 731 NULL, NULL, NULL); 732 refs = --so->so_aborting; /* XXX */ 733 if (error || (refs == 0)) { 734 sofree(so); 735 } else { 736 sounlock(so); 737 } 738 return error; 739 } 740 741 int 742 soaccept(struct socket *so, struct mbuf *nam) 743 { 744 int error; 745 746 KASSERT(solocked(so)); 747 748 error = 0; 749 if ((so->so_state & SS_NOFDREF) == 0) 750 panic("soaccept: !NOFDREF"); 751 so->so_state &= ~SS_NOFDREF; 752 if ((so->so_state & SS_ISDISCONNECTED) == 0 || 753 (so->so_proto->pr_flags & PR_ABRTACPTDIS) == 0) 754 error = (*so->so_proto->pr_usrreq)(so, PRU_ACCEPT, 755 NULL, nam, NULL, NULL); 756 else 757 error = ECONNABORTED; 758 759 return (error); 760 } 761 762 int 763 soconnect(struct socket *so, struct mbuf *nam, struct lwp *l) 764 { 765 int error; 766 767 KASSERT(solocked(so)); 768 769 if (so->so_options & SO_ACCEPTCONN) 770 return (EOPNOTSUPP); 771 /* 772 * If protocol is connection-based, can only connect once. 773 * Otherwise, if connected, try to disconnect first. 774 * This allows user to disconnect by connecting to, e.g., 775 * a null address. 776 */ 777 if (so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING) && 778 ((so->so_proto->pr_flags & PR_CONNREQUIRED) || 779 (error = sodisconnect(so)))) 780 error = EISCONN; 781 else 782 error = (*so->so_proto->pr_usrreq)(so, PRU_CONNECT, 783 NULL, nam, NULL, l); 784 return (error); 785 } 786 787 int 788 soconnect2(struct socket *so1, struct socket *so2) 789 { 790 int error; 791 792 KASSERT(solocked2(so1, so2)); 793 794 error = (*so1->so_proto->pr_usrreq)(so1, PRU_CONNECT2, 795 NULL, (struct mbuf *)so2, NULL, NULL); 796 return (error); 797 } 798 799 int 800 sodisconnect(struct socket *so) 801 { 802 int error; 803 804 KASSERT(solocked(so)); 805 806 if ((so->so_state & SS_ISCONNECTED) == 0) { 807 error = ENOTCONN; 808 } else if (so->so_state & SS_ISDISCONNECTING) { 809 error = EALREADY; 810 } else { 811 error = (*so->so_proto->pr_usrreq)(so, PRU_DISCONNECT, 812 NULL, NULL, NULL, NULL); 813 } 814 sodopendfree(); 815 return (error); 816 } 817 818 #define SBLOCKWAIT(f) (((f) & MSG_DONTWAIT) ? M_NOWAIT : M_WAITOK) 819 /* 820 * Send on a socket. 821 * If send must go all at once and message is larger than 822 * send buffering, then hard error. 823 * Lock against other senders. 824 * If must go all at once and not enough room now, then 825 * inform user that this would block and do nothing. 826 * Otherwise, if nonblocking, send as much as possible. 827 * The data to be sent is described by "uio" if nonzero, 828 * otherwise by the mbuf chain "top" (which must be null 829 * if uio is not). Data provided in mbuf chain must be small 830 * enough to send all at once. 831 * 832 * Returns nonzero on error, timeout or signal; callers 833 * must check for short counts if EINTR/ERESTART are returned. 834 * Data and control buffers are freed on return. 835 */ 836 int 837 sosend(struct socket *so, struct mbuf *addr, struct uio *uio, struct mbuf *top, 838 struct mbuf *control, int flags, struct lwp *l) 839 { 840 struct mbuf **mp, *m; 841 struct proc *p; 842 long space, len, resid, clen, mlen; 843 int error, s, dontroute, atomic; 844 845 p = l->l_proc; 846 sodopendfree(); 847 clen = 0; 848 849 /* 850 * solock() provides atomicity of access. splsoftnet() prevents 851 * protocol processing soft interrupts from interrupting us and 852 * blocking (expensive). 853 */ 854 s = splsoftnet(); 855 solock(so); 856 atomic = sosendallatonce(so) || top; 857 if (uio) 858 resid = uio->uio_resid; 859 else 860 resid = top->m_pkthdr.len; 861 /* 862 * In theory resid should be unsigned. 863 * However, space must be signed, as it might be less than 0 864 * if we over-committed, and we must use a signed comparison 865 * of space and resid. On the other hand, a negative resid 866 * causes us to loop sending 0-length segments to the protocol. 867 */ 868 if (resid < 0) { 869 error = EINVAL; 870 goto out; 871 } 872 dontroute = 873 (flags & MSG_DONTROUTE) && (so->so_options & SO_DONTROUTE) == 0 && 874 (so->so_proto->pr_flags & PR_ATOMIC); 875 l->l_ru.ru_msgsnd++; 876 if (control) 877 clen = control->m_len; 878 restart: 879 if ((error = sblock(&so->so_snd, SBLOCKWAIT(flags))) != 0) 880 goto out; 881 do { 882 if (so->so_state & SS_CANTSENDMORE) { 883 error = EPIPE; 884 goto release; 885 } 886 if (so->so_error) { 887 error = so->so_error; 888 so->so_error = 0; 889 goto release; 890 } 891 if ((so->so_state & SS_ISCONNECTED) == 0) { 892 if (so->so_proto->pr_flags & PR_CONNREQUIRED) { 893 if ((so->so_state & SS_ISCONFIRMING) == 0 && 894 !(resid == 0 && clen != 0)) { 895 error = ENOTCONN; 896 goto release; 897 } 898 } else if (addr == 0) { 899 error = EDESTADDRREQ; 900 goto release; 901 } 902 } 903 space = sbspace(&so->so_snd); 904 if (flags & MSG_OOB) 905 space += 1024; 906 if ((atomic && resid > so->so_snd.sb_hiwat) || 907 clen > so->so_snd.sb_hiwat) { 908 error = EMSGSIZE; 909 goto release; 910 } 911 if (space < resid + clen && 912 (atomic || space < so->so_snd.sb_lowat || space < clen)) { 913 if (so->so_nbio) { 914 error = EWOULDBLOCK; 915 goto release; 916 } 917 sbunlock(&so->so_snd); 918 error = sbwait(&so->so_snd); 919 if (error) 920 goto out; 921 goto restart; 922 } 923 mp = ⊤ 924 space -= clen; 925 do { 926 if (uio == NULL) { 927 /* 928 * Data is prepackaged in "top". 929 */ 930 resid = 0; 931 if (flags & MSG_EOR) 932 top->m_flags |= M_EOR; 933 } else do { 934 sounlock(so); 935 splx(s); 936 if (top == NULL) { 937 m = m_gethdr(M_WAIT, MT_DATA); 938 mlen = MHLEN; 939 m->m_pkthdr.len = 0; 940 m->m_pkthdr.rcvif = NULL; 941 } else { 942 m = m_get(M_WAIT, MT_DATA); 943 mlen = MLEN; 944 } 945 MCLAIM(m, so->so_snd.sb_mowner); 946 if (sock_loan_thresh >= 0 && 947 uio->uio_iov->iov_len >= sock_loan_thresh && 948 space >= sock_loan_thresh && 949 (len = sosend_loan(so, uio, m, 950 space)) != 0) { 951 SOSEND_COUNTER_INCR(&sosend_loan_big); 952 space -= len; 953 goto have_data; 954 } 955 if (resid >= MINCLSIZE && space >= MCLBYTES) { 956 SOSEND_COUNTER_INCR(&sosend_copy_big); 957 m_clget(m, M_WAIT); 958 if ((m->m_flags & M_EXT) == 0) 959 goto nopages; 960 mlen = MCLBYTES; 961 if (atomic && top == 0) { 962 len = lmin(MCLBYTES - max_hdr, 963 resid); 964 m->m_data += max_hdr; 965 } else 966 len = lmin(MCLBYTES, resid); 967 space -= len; 968 } else { 969 nopages: 970 SOSEND_COUNTER_INCR(&sosend_copy_small); 971 len = lmin(lmin(mlen, resid), space); 972 space -= len; 973 /* 974 * For datagram protocols, leave room 975 * for protocol headers in first mbuf. 976 */ 977 if (atomic && top == 0 && len < mlen) 978 MH_ALIGN(m, len); 979 } 980 error = uiomove(mtod(m, void *), (int)len, uio); 981 have_data: 982 resid = uio->uio_resid; 983 m->m_len = len; 984 *mp = m; 985 top->m_pkthdr.len += len; 986 s = splsoftnet(); 987 solock(so); 988 if (error != 0) 989 goto release; 990 mp = &m->m_next; 991 if (resid <= 0) { 992 if (flags & MSG_EOR) 993 top->m_flags |= M_EOR; 994 break; 995 } 996 } while (space > 0 && atomic); 997 998 if (so->so_state & SS_CANTSENDMORE) { 999 error = EPIPE; 1000 goto release; 1001 } 1002 if (dontroute) 1003 so->so_options |= SO_DONTROUTE; 1004 if (resid > 0) 1005 so->so_state |= SS_MORETOCOME; 1006 error = (*so->so_proto->pr_usrreq)(so, 1007 (flags & MSG_OOB) ? PRU_SENDOOB : PRU_SEND, 1008 top, addr, control, curlwp); 1009 if (dontroute) 1010 so->so_options &= ~SO_DONTROUTE; 1011 if (resid > 0) 1012 so->so_state &= ~SS_MORETOCOME; 1013 clen = 0; 1014 control = NULL; 1015 top = NULL; 1016 mp = ⊤ 1017 if (error != 0) 1018 goto release; 1019 } while (resid && space > 0); 1020 } while (resid); 1021 1022 release: 1023 sbunlock(&so->so_snd); 1024 out: 1025 sounlock(so); 1026 splx(s); 1027 if (top) 1028 m_freem(top); 1029 if (control) 1030 m_freem(control); 1031 return (error); 1032 } 1033 1034 /* 1035 * Following replacement or removal of the first mbuf on the first 1036 * mbuf chain of a socket buffer, push necessary state changes back 1037 * into the socket buffer so that other consumers see the values 1038 * consistently. 'nextrecord' is the callers locally stored value of 1039 * the original value of sb->sb_mb->m_nextpkt which must be restored 1040 * when the lead mbuf changes. NOTE: 'nextrecord' may be NULL. 1041 */ 1042 static void 1043 sbsync(struct sockbuf *sb, struct mbuf *nextrecord) 1044 { 1045 1046 KASSERT(solocked(sb->sb_so)); 1047 1048 /* 1049 * First, update for the new value of nextrecord. If necessary, 1050 * make it the first record. 1051 */ 1052 if (sb->sb_mb != NULL) 1053 sb->sb_mb->m_nextpkt = nextrecord; 1054 else 1055 sb->sb_mb = nextrecord; 1056 1057 /* 1058 * Now update any dependent socket buffer fields to reflect 1059 * the new state. This is an inline of SB_EMPTY_FIXUP, with 1060 * the addition of a second clause that takes care of the 1061 * case where sb_mb has been updated, but remains the last 1062 * record. 1063 */ 1064 if (sb->sb_mb == NULL) { 1065 sb->sb_mbtail = NULL; 1066 sb->sb_lastrecord = NULL; 1067 } else if (sb->sb_mb->m_nextpkt == NULL) 1068 sb->sb_lastrecord = sb->sb_mb; 1069 } 1070 1071 /* 1072 * Implement receive operations on a socket. 1073 * We depend on the way that records are added to the sockbuf 1074 * by sbappend*. In particular, each record (mbufs linked through m_next) 1075 * must begin with an address if the protocol so specifies, 1076 * followed by an optional mbuf or mbufs containing ancillary data, 1077 * and then zero or more mbufs of data. 1078 * In order to avoid blocking network interrupts for the entire time here, 1079 * we splx() while doing the actual copy to user space. 1080 * Although the sockbuf is locked, new data may still be appended, 1081 * and thus we must maintain consistency of the sockbuf during that time. 1082 * 1083 * The caller may receive the data as a single mbuf chain by supplying 1084 * an mbuf **mp0 for use in returning the chain. The uio is then used 1085 * only for the count in uio_resid. 1086 */ 1087 int 1088 soreceive(struct socket *so, struct mbuf **paddr, struct uio *uio, 1089 struct mbuf **mp0, struct mbuf **controlp, int *flagsp) 1090 { 1091 struct lwp *l = curlwp; 1092 struct mbuf *m, **mp, *mt; 1093 int atomic, flags, len, error, s, offset, moff, type, orig_resid; 1094 const struct protosw *pr; 1095 struct mbuf *nextrecord; 1096 int mbuf_removed = 0; 1097 const struct domain *dom; 1098 1099 pr = so->so_proto; 1100 atomic = pr->pr_flags & PR_ATOMIC; 1101 dom = pr->pr_domain; 1102 mp = mp0; 1103 type = 0; 1104 orig_resid = uio->uio_resid; 1105 1106 if (paddr != NULL) 1107 *paddr = NULL; 1108 if (controlp != NULL) 1109 *controlp = NULL; 1110 if (flagsp != NULL) 1111 flags = *flagsp &~ MSG_EOR; 1112 else 1113 flags = 0; 1114 1115 if ((flags & MSG_DONTWAIT) == 0) 1116 sodopendfree(); 1117 1118 if (flags & MSG_OOB) { 1119 m = m_get(M_WAIT, MT_DATA); 1120 solock(so); 1121 error = (*pr->pr_usrreq)(so, PRU_RCVOOB, m, 1122 (struct mbuf *)(long)(flags & MSG_PEEK), NULL, l); 1123 sounlock(so); 1124 if (error) 1125 goto bad; 1126 do { 1127 error = uiomove(mtod(m, void *), 1128 (int) min(uio->uio_resid, m->m_len), uio); 1129 m = m_free(m); 1130 } while (uio->uio_resid > 0 && error == 0 && m); 1131 bad: 1132 if (m != NULL) 1133 m_freem(m); 1134 return error; 1135 } 1136 if (mp != NULL) 1137 *mp = NULL; 1138 1139 /* 1140 * solock() provides atomicity of access. splsoftnet() prevents 1141 * protocol processing soft interrupts from interrupting us and 1142 * blocking (expensive). 1143 */ 1144 s = splsoftnet(); 1145 solock(so); 1146 if (so->so_state & SS_ISCONFIRMING && uio->uio_resid) 1147 (*pr->pr_usrreq)(so, PRU_RCVD, NULL, NULL, NULL, l); 1148 1149 restart: 1150 if ((error = sblock(&so->so_rcv, SBLOCKWAIT(flags))) != 0) { 1151 sounlock(so); 1152 splx(s); 1153 return error; 1154 } 1155 1156 m = so->so_rcv.sb_mb; 1157 /* 1158 * If we have less data than requested, block awaiting more 1159 * (subject to any timeout) if: 1160 * 1. the current count is less than the low water mark, 1161 * 2. MSG_WAITALL is set, and it is possible to do the entire 1162 * receive operation at once if we block (resid <= hiwat), or 1163 * 3. MSG_DONTWAIT is not set. 1164 * If MSG_WAITALL is set but resid is larger than the receive buffer, 1165 * we have to do the receive in sections, and thus risk returning 1166 * a short count if a timeout or signal occurs after we start. 1167 */ 1168 if (m == NULL || 1169 ((flags & MSG_DONTWAIT) == 0 && 1170 so->so_rcv.sb_cc < uio->uio_resid && 1171 (so->so_rcv.sb_cc < so->so_rcv.sb_lowat || 1172 ((flags & MSG_WAITALL) && 1173 uio->uio_resid <= so->so_rcv.sb_hiwat)) && 1174 m->m_nextpkt == NULL && !atomic)) { 1175 #ifdef DIAGNOSTIC 1176 if (m == NULL && so->so_rcv.sb_cc) 1177 panic("receive 1"); 1178 #endif 1179 if (so->so_error) { 1180 if (m != NULL) 1181 goto dontblock; 1182 error = so->so_error; 1183 if ((flags & MSG_PEEK) == 0) 1184 so->so_error = 0; 1185 goto release; 1186 } 1187 if (so->so_state & SS_CANTRCVMORE) { 1188 if (m != NULL) 1189 goto dontblock; 1190 else 1191 goto release; 1192 } 1193 for (; m != NULL; m = m->m_next) 1194 if (m->m_type == MT_OOBDATA || (m->m_flags & M_EOR)) { 1195 m = so->so_rcv.sb_mb; 1196 goto dontblock; 1197 } 1198 if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 && 1199 (so->so_proto->pr_flags & PR_CONNREQUIRED)) { 1200 error = ENOTCONN; 1201 goto release; 1202 } 1203 if (uio->uio_resid == 0) 1204 goto release; 1205 if (so->so_nbio || (flags & MSG_DONTWAIT)) { 1206 error = EWOULDBLOCK; 1207 goto release; 1208 } 1209 SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 1"); 1210 SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 1"); 1211 sbunlock(&so->so_rcv); 1212 error = sbwait(&so->so_rcv); 1213 if (error != 0) { 1214 sounlock(so); 1215 splx(s); 1216 return error; 1217 } 1218 goto restart; 1219 } 1220 dontblock: 1221 /* 1222 * On entry here, m points to the first record of the socket buffer. 1223 * From this point onward, we maintain 'nextrecord' as a cache of the 1224 * pointer to the next record in the socket buffer. We must keep the 1225 * various socket buffer pointers and local stack versions of the 1226 * pointers in sync, pushing out modifications before dropping the 1227 * socket lock, and re-reading them when picking it up. 1228 * 1229 * Otherwise, we will race with the network stack appending new data 1230 * or records onto the socket buffer by using inconsistent/stale 1231 * versions of the field, possibly resulting in socket buffer 1232 * corruption. 1233 * 1234 * By holding the high-level sblock(), we prevent simultaneous 1235 * readers from pulling off the front of the socket buffer. 1236 */ 1237 if (l != NULL) 1238 l->l_ru.ru_msgrcv++; 1239 KASSERT(m == so->so_rcv.sb_mb); 1240 SBLASTRECORDCHK(&so->so_rcv, "soreceive 1"); 1241 SBLASTMBUFCHK(&so->so_rcv, "soreceive 1"); 1242 nextrecord = m->m_nextpkt; 1243 if (pr->pr_flags & PR_ADDR) { 1244 #ifdef DIAGNOSTIC 1245 if (m->m_type != MT_SONAME) 1246 panic("receive 1a"); 1247 #endif 1248 orig_resid = 0; 1249 if (flags & MSG_PEEK) { 1250 if (paddr) 1251 *paddr = m_copy(m, 0, m->m_len); 1252 m = m->m_next; 1253 } else { 1254 sbfree(&so->so_rcv, m); 1255 mbuf_removed = 1; 1256 if (paddr != NULL) { 1257 *paddr = m; 1258 so->so_rcv.sb_mb = m->m_next; 1259 m->m_next = NULL; 1260 m = so->so_rcv.sb_mb; 1261 } else { 1262 MFREE(m, so->so_rcv.sb_mb); 1263 m = so->so_rcv.sb_mb; 1264 } 1265 sbsync(&so->so_rcv, nextrecord); 1266 } 1267 } 1268 1269 /* 1270 * Process one or more MT_CONTROL mbufs present before any data mbufs 1271 * in the first mbuf chain on the socket buffer. If MSG_PEEK, we 1272 * just copy the data; if !MSG_PEEK, we call into the protocol to 1273 * perform externalization (or freeing if controlp == NULL). 1274 */ 1275 if (__predict_false(m != NULL && m->m_type == MT_CONTROL)) { 1276 struct mbuf *cm = NULL, *cmn; 1277 struct mbuf **cme = &cm; 1278 1279 do { 1280 if (flags & MSG_PEEK) { 1281 if (controlp != NULL) { 1282 *controlp = m_copy(m, 0, m->m_len); 1283 controlp = &(*controlp)->m_next; 1284 } 1285 m = m->m_next; 1286 } else { 1287 sbfree(&so->so_rcv, m); 1288 so->so_rcv.sb_mb = m->m_next; 1289 m->m_next = NULL; 1290 *cme = m; 1291 cme = &(*cme)->m_next; 1292 m = so->so_rcv.sb_mb; 1293 } 1294 } while (m != NULL && m->m_type == MT_CONTROL); 1295 if ((flags & MSG_PEEK) == 0) 1296 sbsync(&so->so_rcv, nextrecord); 1297 for (; cm != NULL; cm = cmn) { 1298 cmn = cm->m_next; 1299 cm->m_next = NULL; 1300 type = mtod(cm, struct cmsghdr *)->cmsg_type; 1301 if (controlp != NULL) { 1302 if (dom->dom_externalize != NULL && 1303 type == SCM_RIGHTS) { 1304 sounlock(so); 1305 splx(s); 1306 error = (*dom->dom_externalize)(cm, l); 1307 s = splsoftnet(); 1308 solock(so); 1309 } 1310 *controlp = cm; 1311 while (*controlp != NULL) 1312 controlp = &(*controlp)->m_next; 1313 } else { 1314 /* 1315 * Dispose of any SCM_RIGHTS message that went 1316 * through the read path rather than recv. 1317 */ 1318 if (dom->dom_dispose != NULL && 1319 type == SCM_RIGHTS) { 1320 sounlock(so); 1321 (*dom->dom_dispose)(cm); 1322 solock(so); 1323 } 1324 m_freem(cm); 1325 } 1326 } 1327 if (m != NULL) 1328 nextrecord = so->so_rcv.sb_mb->m_nextpkt; 1329 else 1330 nextrecord = so->so_rcv.sb_mb; 1331 orig_resid = 0; 1332 } 1333 1334 /* If m is non-NULL, we have some data to read. */ 1335 if (__predict_true(m != NULL)) { 1336 type = m->m_type; 1337 if (type == MT_OOBDATA) 1338 flags |= MSG_OOB; 1339 } 1340 SBLASTRECORDCHK(&so->so_rcv, "soreceive 2"); 1341 SBLASTMBUFCHK(&so->so_rcv, "soreceive 2"); 1342 1343 moff = 0; 1344 offset = 0; 1345 while (m != NULL && uio->uio_resid > 0 && error == 0) { 1346 if (m->m_type == MT_OOBDATA) { 1347 if (type != MT_OOBDATA) 1348 break; 1349 } else if (type == MT_OOBDATA) 1350 break; 1351 #ifdef DIAGNOSTIC 1352 else if (m->m_type != MT_DATA && m->m_type != MT_HEADER) 1353 panic("receive 3"); 1354 #endif 1355 so->so_state &= ~SS_RCVATMARK; 1356 len = uio->uio_resid; 1357 if (so->so_oobmark && len > so->so_oobmark - offset) 1358 len = so->so_oobmark - offset; 1359 if (len > m->m_len - moff) 1360 len = m->m_len - moff; 1361 /* 1362 * If mp is set, just pass back the mbufs. 1363 * Otherwise copy them out via the uio, then free. 1364 * Sockbuf must be consistent here (points to current mbuf, 1365 * it points to next record) when we drop priority; 1366 * we must note any additions to the sockbuf when we 1367 * block interrupts again. 1368 */ 1369 if (mp == NULL) { 1370 SBLASTRECORDCHK(&so->so_rcv, "soreceive uiomove"); 1371 SBLASTMBUFCHK(&so->so_rcv, "soreceive uiomove"); 1372 sounlock(so); 1373 splx(s); 1374 error = uiomove(mtod(m, char *) + moff, (int)len, uio); 1375 s = splsoftnet(); 1376 solock(so); 1377 if (error != 0) { 1378 /* 1379 * If any part of the record has been removed 1380 * (such as the MT_SONAME mbuf, which will 1381 * happen when PR_ADDR, and thus also 1382 * PR_ATOMIC, is set), then drop the entire 1383 * record to maintain the atomicity of the 1384 * receive operation. 1385 * 1386 * This avoids a later panic("receive 1a") 1387 * when compiled with DIAGNOSTIC. 1388 */ 1389 if (m && mbuf_removed && atomic) 1390 (void) sbdroprecord(&so->so_rcv); 1391 1392 goto release; 1393 } 1394 } else 1395 uio->uio_resid -= len; 1396 if (len == m->m_len - moff) { 1397 if (m->m_flags & M_EOR) 1398 flags |= MSG_EOR; 1399 if (flags & MSG_PEEK) { 1400 m = m->m_next; 1401 moff = 0; 1402 } else { 1403 nextrecord = m->m_nextpkt; 1404 sbfree(&so->so_rcv, m); 1405 if (mp) { 1406 *mp = m; 1407 mp = &m->m_next; 1408 so->so_rcv.sb_mb = m = m->m_next; 1409 *mp = NULL; 1410 } else { 1411 MFREE(m, so->so_rcv.sb_mb); 1412 m = so->so_rcv.sb_mb; 1413 } 1414 /* 1415 * If m != NULL, we also know that 1416 * so->so_rcv.sb_mb != NULL. 1417 */ 1418 KASSERT(so->so_rcv.sb_mb == m); 1419 if (m) { 1420 m->m_nextpkt = nextrecord; 1421 if (nextrecord == NULL) 1422 so->so_rcv.sb_lastrecord = m; 1423 } else { 1424 so->so_rcv.sb_mb = nextrecord; 1425 SB_EMPTY_FIXUP(&so->so_rcv); 1426 } 1427 SBLASTRECORDCHK(&so->so_rcv, "soreceive 3"); 1428 SBLASTMBUFCHK(&so->so_rcv, "soreceive 3"); 1429 } 1430 } else if (flags & MSG_PEEK) 1431 moff += len; 1432 else { 1433 if (mp != NULL) { 1434 mt = m_copym(m, 0, len, M_NOWAIT); 1435 if (__predict_false(mt == NULL)) { 1436 sounlock(so); 1437 mt = m_copym(m, 0, len, M_WAIT); 1438 solock(so); 1439 } 1440 *mp = mt; 1441 } 1442 m->m_data += len; 1443 m->m_len -= len; 1444 so->so_rcv.sb_cc -= len; 1445 } 1446 if (so->so_oobmark) { 1447 if ((flags & MSG_PEEK) == 0) { 1448 so->so_oobmark -= len; 1449 if (so->so_oobmark == 0) { 1450 so->so_state |= SS_RCVATMARK; 1451 break; 1452 } 1453 } else { 1454 offset += len; 1455 if (offset == so->so_oobmark) 1456 break; 1457 } 1458 } 1459 if (flags & MSG_EOR) 1460 break; 1461 /* 1462 * If the MSG_WAITALL flag is set (for non-atomic socket), 1463 * we must not quit until "uio->uio_resid == 0" or an error 1464 * termination. If a signal/timeout occurs, return 1465 * with a short count but without error. 1466 * Keep sockbuf locked against other readers. 1467 */ 1468 while (flags & MSG_WAITALL && m == NULL && uio->uio_resid > 0 && 1469 !sosendallatonce(so) && !nextrecord) { 1470 if (so->so_error || so->so_state & SS_CANTRCVMORE) 1471 break; 1472 /* 1473 * If we are peeking and the socket receive buffer is 1474 * full, stop since we can't get more data to peek at. 1475 */ 1476 if ((flags & MSG_PEEK) && sbspace(&so->so_rcv) <= 0) 1477 break; 1478 /* 1479 * If we've drained the socket buffer, tell the 1480 * protocol in case it needs to do something to 1481 * get it filled again. 1482 */ 1483 if ((pr->pr_flags & PR_WANTRCVD) && so->so_pcb) 1484 (*pr->pr_usrreq)(so, PRU_RCVD, 1485 NULL, (struct mbuf *)(long)flags, NULL, l); 1486 SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 2"); 1487 SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 2"); 1488 error = sbwait(&so->so_rcv); 1489 if (error != 0) { 1490 sbunlock(&so->so_rcv); 1491 sounlock(so); 1492 splx(s); 1493 return 0; 1494 } 1495 if ((m = so->so_rcv.sb_mb) != NULL) 1496 nextrecord = m->m_nextpkt; 1497 } 1498 } 1499 1500 if (m && atomic) { 1501 flags |= MSG_TRUNC; 1502 if ((flags & MSG_PEEK) == 0) 1503 (void) sbdroprecord(&so->so_rcv); 1504 } 1505 if ((flags & MSG_PEEK) == 0) { 1506 if (m == NULL) { 1507 /* 1508 * First part is an inline SB_EMPTY_FIXUP(). Second 1509 * part makes sure sb_lastrecord is up-to-date if 1510 * there is still data in the socket buffer. 1511 */ 1512 so->so_rcv.sb_mb = nextrecord; 1513 if (so->so_rcv.sb_mb == NULL) { 1514 so->so_rcv.sb_mbtail = NULL; 1515 so->so_rcv.sb_lastrecord = NULL; 1516 } else if (nextrecord->m_nextpkt == NULL) 1517 so->so_rcv.sb_lastrecord = nextrecord; 1518 } 1519 SBLASTRECORDCHK(&so->so_rcv, "soreceive 4"); 1520 SBLASTMBUFCHK(&so->so_rcv, "soreceive 4"); 1521 if (pr->pr_flags & PR_WANTRCVD && so->so_pcb) 1522 (*pr->pr_usrreq)(so, PRU_RCVD, NULL, 1523 (struct mbuf *)(long)flags, NULL, l); 1524 } 1525 if (orig_resid == uio->uio_resid && orig_resid && 1526 (flags & MSG_EOR) == 0 && (so->so_state & SS_CANTRCVMORE) == 0) { 1527 sbunlock(&so->so_rcv); 1528 goto restart; 1529 } 1530 1531 if (flagsp != NULL) 1532 *flagsp |= flags; 1533 release: 1534 sbunlock(&so->so_rcv); 1535 sounlock(so); 1536 splx(s); 1537 return error; 1538 } 1539 1540 int 1541 soshutdown(struct socket *so, int how) 1542 { 1543 const struct protosw *pr; 1544 int error; 1545 1546 KASSERT(solocked(so)); 1547 1548 pr = so->so_proto; 1549 if (!(how == SHUT_RD || how == SHUT_WR || how == SHUT_RDWR)) 1550 return (EINVAL); 1551 1552 if (how == SHUT_RD || how == SHUT_RDWR) { 1553 sorflush(so); 1554 error = 0; 1555 } 1556 if (how == SHUT_WR || how == SHUT_RDWR) 1557 error = (*pr->pr_usrreq)(so, PRU_SHUTDOWN, NULL, 1558 NULL, NULL, NULL); 1559 1560 return error; 1561 } 1562 1563 int 1564 sodrain(struct socket *so) 1565 { 1566 int error; 1567 1568 solock(so); 1569 so->so_state |= SS_ISDRAINING; 1570 cv_broadcast(&so->so_cv); 1571 error = soshutdown(so, SHUT_RDWR); 1572 sounlock(so); 1573 1574 return error; 1575 } 1576 1577 void 1578 sorflush(struct socket *so) 1579 { 1580 struct sockbuf *sb, asb; 1581 const struct protosw *pr; 1582 1583 KASSERT(solocked(so)); 1584 1585 sb = &so->so_rcv; 1586 pr = so->so_proto; 1587 socantrcvmore(so); 1588 sb->sb_flags |= SB_NOINTR; 1589 (void )sblock(sb, M_WAITOK); 1590 sbunlock(sb); 1591 asb = *sb; 1592 /* 1593 * Clear most of the sockbuf structure, but leave some of the 1594 * fields valid. 1595 */ 1596 memset(&sb->sb_startzero, 0, 1597 sizeof(*sb) - offsetof(struct sockbuf, sb_startzero)); 1598 if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose) { 1599 sounlock(so); 1600 (*pr->pr_domain->dom_dispose)(asb.sb_mb); 1601 solock(so); 1602 } 1603 sbrelease(&asb, so); 1604 } 1605 1606 /* 1607 * internal set SOL_SOCKET options 1608 */ 1609 static int 1610 sosetopt1(struct socket *so, const struct sockopt *sopt) 1611 { 1612 int error = EINVAL, optval, opt; 1613 struct linger l; 1614 struct timeval tv; 1615 1616 switch ((opt = sopt->sopt_name)) { 1617 1618 case SO_ACCEPTFILTER: 1619 error = accept_filt_setopt(so, sopt); 1620 KASSERT(solocked(so)); 1621 break; 1622 1623 case SO_LINGER: 1624 error = sockopt_get(sopt, &l, sizeof(l)); 1625 solock(so); 1626 if (error) 1627 break; 1628 if (l.l_linger < 0 || l.l_linger > USHRT_MAX || 1629 l.l_linger > (INT_MAX / hz)) { 1630 error = EDOM; 1631 break; 1632 } 1633 so->so_linger = l.l_linger; 1634 if (l.l_onoff) 1635 so->so_options |= SO_LINGER; 1636 else 1637 so->so_options &= ~SO_LINGER; 1638 break; 1639 1640 case SO_DEBUG: 1641 case SO_KEEPALIVE: 1642 case SO_DONTROUTE: 1643 case SO_USELOOPBACK: 1644 case SO_BROADCAST: 1645 case SO_REUSEADDR: 1646 case SO_REUSEPORT: 1647 case SO_OOBINLINE: 1648 case SO_TIMESTAMP: 1649 #ifdef SO_OTIMESTAMP 1650 case SO_OTIMESTAMP: 1651 #endif 1652 error = sockopt_getint(sopt, &optval); 1653 solock(so); 1654 if (error) 1655 break; 1656 if (optval) 1657 so->so_options |= opt; 1658 else 1659 so->so_options &= ~opt; 1660 break; 1661 1662 case SO_SNDBUF: 1663 case SO_RCVBUF: 1664 case SO_SNDLOWAT: 1665 case SO_RCVLOWAT: 1666 error = sockopt_getint(sopt, &optval); 1667 solock(so); 1668 if (error) 1669 break; 1670 1671 /* 1672 * Values < 1 make no sense for any of these 1673 * options, so disallow them. 1674 */ 1675 if (optval < 1) { 1676 error = EINVAL; 1677 break; 1678 } 1679 1680 switch (opt) { 1681 case SO_SNDBUF: 1682 if (sbreserve(&so->so_snd, (u_long)optval, so) == 0) { 1683 error = ENOBUFS; 1684 break; 1685 } 1686 so->so_snd.sb_flags &= ~SB_AUTOSIZE; 1687 break; 1688 1689 case SO_RCVBUF: 1690 if (sbreserve(&so->so_rcv, (u_long)optval, so) == 0) { 1691 error = ENOBUFS; 1692 break; 1693 } 1694 so->so_rcv.sb_flags &= ~SB_AUTOSIZE; 1695 break; 1696 1697 /* 1698 * Make sure the low-water is never greater than 1699 * the high-water. 1700 */ 1701 case SO_SNDLOWAT: 1702 if (optval > so->so_snd.sb_hiwat) 1703 optval = so->so_snd.sb_hiwat; 1704 1705 so->so_snd.sb_lowat = optval; 1706 break; 1707 1708 case SO_RCVLOWAT: 1709 if (optval > so->so_rcv.sb_hiwat) 1710 optval = so->so_rcv.sb_hiwat; 1711 1712 so->so_rcv.sb_lowat = optval; 1713 break; 1714 } 1715 break; 1716 1717 #ifdef COMPAT_50 1718 case SO_OSNDTIMEO: 1719 case SO_ORCVTIMEO: { 1720 struct timeval50 otv; 1721 error = sockopt_get(sopt, &otv, sizeof(otv)); 1722 if (error) { 1723 solock(so); 1724 break; 1725 } 1726 timeval50_to_timeval(&otv, &tv); 1727 opt = opt == SO_OSNDTIMEO ? SO_SNDTIMEO : SO_RCVTIMEO; 1728 error = 0; 1729 /*FALLTHROUGH*/ 1730 } 1731 #endif /* COMPAT_50 */ 1732 1733 case SO_SNDTIMEO: 1734 case SO_RCVTIMEO: 1735 if (error) 1736 error = sockopt_get(sopt, &tv, sizeof(tv)); 1737 solock(so); 1738 if (error) 1739 break; 1740 1741 if (tv.tv_sec > (INT_MAX - tv.tv_usec / tick) / hz) { 1742 error = EDOM; 1743 break; 1744 } 1745 1746 optval = tv.tv_sec * hz + tv.tv_usec / tick; 1747 if (optval == 0 && tv.tv_usec != 0) 1748 optval = 1; 1749 1750 switch (opt) { 1751 case SO_SNDTIMEO: 1752 so->so_snd.sb_timeo = optval; 1753 break; 1754 case SO_RCVTIMEO: 1755 so->so_rcv.sb_timeo = optval; 1756 break; 1757 } 1758 break; 1759 1760 default: 1761 solock(so); 1762 error = ENOPROTOOPT; 1763 break; 1764 } 1765 KASSERT(solocked(so)); 1766 return error; 1767 } 1768 1769 int 1770 sosetopt(struct socket *so, struct sockopt *sopt) 1771 { 1772 int error, prerr; 1773 1774 if (sopt->sopt_level == SOL_SOCKET) { 1775 error = sosetopt1(so, sopt); 1776 KASSERT(solocked(so)); 1777 } else { 1778 error = ENOPROTOOPT; 1779 solock(so); 1780 } 1781 1782 if ((error == 0 || error == ENOPROTOOPT) && 1783 so->so_proto != NULL && so->so_proto->pr_ctloutput != NULL) { 1784 /* give the protocol stack a shot */ 1785 prerr = (*so->so_proto->pr_ctloutput)(PRCO_SETOPT, so, sopt); 1786 if (prerr == 0) 1787 error = 0; 1788 else if (prerr != ENOPROTOOPT) 1789 error = prerr; 1790 } 1791 sounlock(so); 1792 return error; 1793 } 1794 1795 /* 1796 * so_setsockopt() is a wrapper providing a sockopt structure for sosetopt() 1797 */ 1798 int 1799 so_setsockopt(struct lwp *l, struct socket *so, int level, int name, 1800 const void *val, size_t valsize) 1801 { 1802 struct sockopt sopt; 1803 int error; 1804 1805 KASSERT(valsize == 0 || val != NULL); 1806 1807 sockopt_init(&sopt, level, name, valsize); 1808 sockopt_set(&sopt, val, valsize); 1809 1810 error = sosetopt(so, &sopt); 1811 1812 sockopt_destroy(&sopt); 1813 1814 return error; 1815 } 1816 1817 /* 1818 * internal get SOL_SOCKET options 1819 */ 1820 static int 1821 sogetopt1(struct socket *so, struct sockopt *sopt) 1822 { 1823 int error, optval, opt; 1824 struct linger l; 1825 struct timeval tv; 1826 1827 switch ((opt = sopt->sopt_name)) { 1828 1829 case SO_ACCEPTFILTER: 1830 error = accept_filt_getopt(so, sopt); 1831 break; 1832 1833 case SO_LINGER: 1834 l.l_onoff = (so->so_options & SO_LINGER) ? 1 : 0; 1835 l.l_linger = so->so_linger; 1836 1837 error = sockopt_set(sopt, &l, sizeof(l)); 1838 break; 1839 1840 case SO_USELOOPBACK: 1841 case SO_DONTROUTE: 1842 case SO_DEBUG: 1843 case SO_KEEPALIVE: 1844 case SO_REUSEADDR: 1845 case SO_REUSEPORT: 1846 case SO_BROADCAST: 1847 case SO_OOBINLINE: 1848 case SO_TIMESTAMP: 1849 #ifdef SO_OTIMESTAMP 1850 case SO_OTIMESTAMP: 1851 #endif 1852 error = sockopt_setint(sopt, (so->so_options & opt) ? 1 : 0); 1853 break; 1854 1855 case SO_TYPE: 1856 error = sockopt_setint(sopt, so->so_type); 1857 break; 1858 1859 case SO_ERROR: 1860 error = sockopt_setint(sopt, so->so_error); 1861 so->so_error = 0; 1862 break; 1863 1864 case SO_SNDBUF: 1865 error = sockopt_setint(sopt, so->so_snd.sb_hiwat); 1866 break; 1867 1868 case SO_RCVBUF: 1869 error = sockopt_setint(sopt, so->so_rcv.sb_hiwat); 1870 break; 1871 1872 case SO_SNDLOWAT: 1873 error = sockopt_setint(sopt, so->so_snd.sb_lowat); 1874 break; 1875 1876 case SO_RCVLOWAT: 1877 error = sockopt_setint(sopt, so->so_rcv.sb_lowat); 1878 break; 1879 1880 #ifdef COMPAT_50 1881 case SO_OSNDTIMEO: 1882 case SO_ORCVTIMEO: { 1883 struct timeval50 otv; 1884 1885 optval = (opt == SO_OSNDTIMEO ? 1886 so->so_snd.sb_timeo : so->so_rcv.sb_timeo); 1887 1888 otv.tv_sec = optval / hz; 1889 otv.tv_usec = (optval % hz) * tick; 1890 1891 error = sockopt_set(sopt, &otv, sizeof(otv)); 1892 break; 1893 } 1894 #endif /* COMPAT_50 */ 1895 1896 case SO_SNDTIMEO: 1897 case SO_RCVTIMEO: 1898 optval = (opt == SO_SNDTIMEO ? 1899 so->so_snd.sb_timeo : so->so_rcv.sb_timeo); 1900 1901 tv.tv_sec = optval / hz; 1902 tv.tv_usec = (optval % hz) * tick; 1903 1904 error = sockopt_set(sopt, &tv, sizeof(tv)); 1905 break; 1906 1907 case SO_OVERFLOWED: 1908 error = sockopt_setint(sopt, so->so_rcv.sb_overflowed); 1909 break; 1910 1911 default: 1912 error = ENOPROTOOPT; 1913 break; 1914 } 1915 1916 return (error); 1917 } 1918 1919 int 1920 sogetopt(struct socket *so, struct sockopt *sopt) 1921 { 1922 int error; 1923 1924 solock(so); 1925 if (sopt->sopt_level != SOL_SOCKET) { 1926 if (so->so_proto && so->so_proto->pr_ctloutput) { 1927 error = ((*so->so_proto->pr_ctloutput) 1928 (PRCO_GETOPT, so, sopt)); 1929 } else 1930 error = (ENOPROTOOPT); 1931 } else { 1932 error = sogetopt1(so, sopt); 1933 } 1934 sounlock(so); 1935 return (error); 1936 } 1937 1938 /* 1939 * alloc sockopt data buffer buffer 1940 * - will be released at destroy 1941 */ 1942 static int 1943 sockopt_alloc(struct sockopt *sopt, size_t len, km_flag_t kmflag) 1944 { 1945 1946 KASSERT(sopt->sopt_size == 0); 1947 1948 if (len > sizeof(sopt->sopt_buf)) { 1949 sopt->sopt_data = kmem_zalloc(len, kmflag); 1950 if (sopt->sopt_data == NULL) 1951 return ENOMEM; 1952 } else 1953 sopt->sopt_data = sopt->sopt_buf; 1954 1955 sopt->sopt_size = len; 1956 return 0; 1957 } 1958 1959 /* 1960 * initialise sockopt storage 1961 * - MAY sleep during allocation 1962 */ 1963 void 1964 sockopt_init(struct sockopt *sopt, int level, int name, size_t size) 1965 { 1966 1967 memset(sopt, 0, sizeof(*sopt)); 1968 1969 sopt->sopt_level = level; 1970 sopt->sopt_name = name; 1971 (void)sockopt_alloc(sopt, size, KM_SLEEP); 1972 } 1973 1974 /* 1975 * destroy sockopt storage 1976 * - will release any held memory references 1977 */ 1978 void 1979 sockopt_destroy(struct sockopt *sopt) 1980 { 1981 1982 if (sopt->sopt_data != sopt->sopt_buf) 1983 kmem_free(sopt->sopt_data, sopt->sopt_size); 1984 1985 memset(sopt, 0, sizeof(*sopt)); 1986 } 1987 1988 /* 1989 * set sockopt value 1990 * - value is copied into sockopt 1991 * - memory is allocated when necessary, will not sleep 1992 */ 1993 int 1994 sockopt_set(struct sockopt *sopt, const void *buf, size_t len) 1995 { 1996 int error; 1997 1998 if (sopt->sopt_size == 0) { 1999 error = sockopt_alloc(sopt, len, KM_NOSLEEP); 2000 if (error) 2001 return error; 2002 } 2003 2004 KASSERT(sopt->sopt_size == len); 2005 memcpy(sopt->sopt_data, buf, len); 2006 return 0; 2007 } 2008 2009 /* 2010 * common case of set sockopt integer value 2011 */ 2012 int 2013 sockopt_setint(struct sockopt *sopt, int val) 2014 { 2015 2016 return sockopt_set(sopt, &val, sizeof(int)); 2017 } 2018 2019 /* 2020 * get sockopt value 2021 * - correct size must be given 2022 */ 2023 int 2024 sockopt_get(const struct sockopt *sopt, void *buf, size_t len) 2025 { 2026 2027 if (sopt->sopt_size != len) 2028 return EINVAL; 2029 2030 memcpy(buf, sopt->sopt_data, len); 2031 return 0; 2032 } 2033 2034 /* 2035 * common case of get sockopt integer value 2036 */ 2037 int 2038 sockopt_getint(const struct sockopt *sopt, int *valp) 2039 { 2040 2041 return sockopt_get(sopt, valp, sizeof(int)); 2042 } 2043 2044 /* 2045 * set sockopt value from mbuf 2046 * - ONLY for legacy code 2047 * - mbuf is released by sockopt 2048 * - will not sleep 2049 */ 2050 int 2051 sockopt_setmbuf(struct sockopt *sopt, struct mbuf *m) 2052 { 2053 size_t len; 2054 int error; 2055 2056 len = m_length(m); 2057 2058 if (sopt->sopt_size == 0) { 2059 error = sockopt_alloc(sopt, len, KM_NOSLEEP); 2060 if (error) 2061 return error; 2062 } 2063 2064 KASSERT(sopt->sopt_size == len); 2065 m_copydata(m, 0, len, sopt->sopt_data); 2066 m_freem(m); 2067 2068 return 0; 2069 } 2070 2071 /* 2072 * get sockopt value into mbuf 2073 * - ONLY for legacy code 2074 * - mbuf to be released by the caller 2075 * - will not sleep 2076 */ 2077 struct mbuf * 2078 sockopt_getmbuf(const struct sockopt *sopt) 2079 { 2080 struct mbuf *m; 2081 2082 if (sopt->sopt_size > MCLBYTES) 2083 return NULL; 2084 2085 m = m_get(M_DONTWAIT, MT_SOOPTS); 2086 if (m == NULL) 2087 return NULL; 2088 2089 if (sopt->sopt_size > MLEN) { 2090 MCLGET(m, M_DONTWAIT); 2091 if ((m->m_flags & M_EXT) == 0) { 2092 m_free(m); 2093 return NULL; 2094 } 2095 } 2096 2097 memcpy(mtod(m, void *), sopt->sopt_data, sopt->sopt_size); 2098 m->m_len = sopt->sopt_size; 2099 2100 return m; 2101 } 2102 2103 void 2104 sohasoutofband(struct socket *so) 2105 { 2106 2107 fownsignal(so->so_pgid, SIGURG, POLL_PRI, POLLPRI|POLLRDBAND, so); 2108 selnotify(&so->so_rcv.sb_sel, POLLPRI | POLLRDBAND, NOTE_SUBMIT); 2109 } 2110 2111 static void 2112 filt_sordetach(struct knote *kn) 2113 { 2114 struct socket *so; 2115 2116 so = ((file_t *)kn->kn_obj)->f_data; 2117 solock(so); 2118 SLIST_REMOVE(&so->so_rcv.sb_sel.sel_klist, kn, knote, kn_selnext); 2119 if (SLIST_EMPTY(&so->so_rcv.sb_sel.sel_klist)) 2120 so->so_rcv.sb_flags &= ~SB_KNOTE; 2121 sounlock(so); 2122 } 2123 2124 /*ARGSUSED*/ 2125 static int 2126 filt_soread(struct knote *kn, long hint) 2127 { 2128 struct socket *so; 2129 int rv; 2130 2131 so = ((file_t *)kn->kn_obj)->f_data; 2132 if (hint != NOTE_SUBMIT) 2133 solock(so); 2134 kn->kn_data = so->so_rcv.sb_cc; 2135 if (so->so_state & SS_CANTRCVMORE) { 2136 kn->kn_flags |= EV_EOF; 2137 kn->kn_fflags = so->so_error; 2138 rv = 1; 2139 } else if (so->so_error) /* temporary udp error */ 2140 rv = 1; 2141 else if (kn->kn_sfflags & NOTE_LOWAT) 2142 rv = (kn->kn_data >= kn->kn_sdata); 2143 else 2144 rv = (kn->kn_data >= so->so_rcv.sb_lowat); 2145 if (hint != NOTE_SUBMIT) 2146 sounlock(so); 2147 return rv; 2148 } 2149 2150 static void 2151 filt_sowdetach(struct knote *kn) 2152 { 2153 struct socket *so; 2154 2155 so = ((file_t *)kn->kn_obj)->f_data; 2156 solock(so); 2157 SLIST_REMOVE(&so->so_snd.sb_sel.sel_klist, kn, knote, kn_selnext); 2158 if (SLIST_EMPTY(&so->so_snd.sb_sel.sel_klist)) 2159 so->so_snd.sb_flags &= ~SB_KNOTE; 2160 sounlock(so); 2161 } 2162 2163 /*ARGSUSED*/ 2164 static int 2165 filt_sowrite(struct knote *kn, long hint) 2166 { 2167 struct socket *so; 2168 int rv; 2169 2170 so = ((file_t *)kn->kn_obj)->f_data; 2171 if (hint != NOTE_SUBMIT) 2172 solock(so); 2173 kn->kn_data = sbspace(&so->so_snd); 2174 if (so->so_state & SS_CANTSENDMORE) { 2175 kn->kn_flags |= EV_EOF; 2176 kn->kn_fflags = so->so_error; 2177 rv = 1; 2178 } else if (so->so_error) /* temporary udp error */ 2179 rv = 1; 2180 else if (((so->so_state & SS_ISCONNECTED) == 0) && 2181 (so->so_proto->pr_flags & PR_CONNREQUIRED)) 2182 rv = 0; 2183 else if (kn->kn_sfflags & NOTE_LOWAT) 2184 rv = (kn->kn_data >= kn->kn_sdata); 2185 else 2186 rv = (kn->kn_data >= so->so_snd.sb_lowat); 2187 if (hint != NOTE_SUBMIT) 2188 sounlock(so); 2189 return rv; 2190 } 2191 2192 /*ARGSUSED*/ 2193 static int 2194 filt_solisten(struct knote *kn, long hint) 2195 { 2196 struct socket *so; 2197 int rv; 2198 2199 so = ((file_t *)kn->kn_obj)->f_data; 2200 2201 /* 2202 * Set kn_data to number of incoming connections, not 2203 * counting partial (incomplete) connections. 2204 */ 2205 if (hint != NOTE_SUBMIT) 2206 solock(so); 2207 kn->kn_data = so->so_qlen; 2208 rv = (kn->kn_data > 0); 2209 if (hint != NOTE_SUBMIT) 2210 sounlock(so); 2211 return rv; 2212 } 2213 2214 static const struct filterops solisten_filtops = 2215 { 1, NULL, filt_sordetach, filt_solisten }; 2216 static const struct filterops soread_filtops = 2217 { 1, NULL, filt_sordetach, filt_soread }; 2218 static const struct filterops sowrite_filtops = 2219 { 1, NULL, filt_sowdetach, filt_sowrite }; 2220 2221 int 2222 soo_kqfilter(struct file *fp, struct knote *kn) 2223 { 2224 struct socket *so; 2225 struct sockbuf *sb; 2226 2227 so = ((file_t *)kn->kn_obj)->f_data; 2228 solock(so); 2229 switch (kn->kn_filter) { 2230 case EVFILT_READ: 2231 if (so->so_options & SO_ACCEPTCONN) 2232 kn->kn_fop = &solisten_filtops; 2233 else 2234 kn->kn_fop = &soread_filtops; 2235 sb = &so->so_rcv; 2236 break; 2237 case EVFILT_WRITE: 2238 kn->kn_fop = &sowrite_filtops; 2239 sb = &so->so_snd; 2240 break; 2241 default: 2242 sounlock(so); 2243 return (EINVAL); 2244 } 2245 SLIST_INSERT_HEAD(&sb->sb_sel.sel_klist, kn, kn_selnext); 2246 sb->sb_flags |= SB_KNOTE; 2247 sounlock(so); 2248 return (0); 2249 } 2250 2251 static int 2252 sodopoll(struct socket *so, int events) 2253 { 2254 int revents; 2255 2256 revents = 0; 2257 2258 if (events & (POLLIN | POLLRDNORM)) 2259 if (soreadable(so)) 2260 revents |= events & (POLLIN | POLLRDNORM); 2261 2262 if (events & (POLLOUT | POLLWRNORM)) 2263 if (sowritable(so)) 2264 revents |= events & (POLLOUT | POLLWRNORM); 2265 2266 if (events & (POLLPRI | POLLRDBAND)) 2267 if (so->so_oobmark || (so->so_state & SS_RCVATMARK)) 2268 revents |= events & (POLLPRI | POLLRDBAND); 2269 2270 return revents; 2271 } 2272 2273 int 2274 sopoll(struct socket *so, int events) 2275 { 2276 int revents = 0; 2277 2278 #ifndef DIAGNOSTIC 2279 /* 2280 * Do a quick, unlocked check in expectation that the socket 2281 * will be ready for I/O. Don't do this check if DIAGNOSTIC, 2282 * as the solocked() assertions will fail. 2283 */ 2284 if ((revents = sodopoll(so, events)) != 0) 2285 return revents; 2286 #endif 2287 2288 solock(so); 2289 if ((revents = sodopoll(so, events)) == 0) { 2290 if (events & (POLLIN | POLLPRI | POLLRDNORM | POLLRDBAND)) { 2291 selrecord(curlwp, &so->so_rcv.sb_sel); 2292 so->so_rcv.sb_flags |= SB_NOTIFY; 2293 } 2294 2295 if (events & (POLLOUT | POLLWRNORM)) { 2296 selrecord(curlwp, &so->so_snd.sb_sel); 2297 so->so_snd.sb_flags |= SB_NOTIFY; 2298 } 2299 } 2300 sounlock(so); 2301 2302 return revents; 2303 } 2304 2305 2306 #include <sys/sysctl.h> 2307 2308 static int sysctl_kern_somaxkva(SYSCTLFN_PROTO); 2309 2310 /* 2311 * sysctl helper routine for kern.somaxkva. ensures that the given 2312 * value is not too small. 2313 * (XXX should we maybe make sure it's not too large as well?) 2314 */ 2315 static int 2316 sysctl_kern_somaxkva(SYSCTLFN_ARGS) 2317 { 2318 int error, new_somaxkva; 2319 struct sysctlnode node; 2320 2321 new_somaxkva = somaxkva; 2322 node = *rnode; 2323 node.sysctl_data = &new_somaxkva; 2324 error = sysctl_lookup(SYSCTLFN_CALL(&node)); 2325 if (error || newp == NULL) 2326 return (error); 2327 2328 if (new_somaxkva < (16 * 1024 * 1024)) /* sanity */ 2329 return (EINVAL); 2330 2331 mutex_enter(&so_pendfree_lock); 2332 somaxkva = new_somaxkva; 2333 cv_broadcast(&socurkva_cv); 2334 mutex_exit(&so_pendfree_lock); 2335 2336 return (error); 2337 } 2338 2339 static void 2340 sysctl_kern_somaxkva_setup(void) 2341 { 2342 2343 KASSERT(socket_sysctllog == NULL); 2344 sysctl_createv(&socket_sysctllog, 0, NULL, NULL, 2345 CTLFLAG_PERMANENT, 2346 CTLTYPE_NODE, "kern", NULL, 2347 NULL, 0, NULL, 0, 2348 CTL_KERN, CTL_EOL); 2349 2350 sysctl_createv(&socket_sysctllog, 0, NULL, NULL, 2351 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 2352 CTLTYPE_INT, "somaxkva", 2353 SYSCTL_DESCR("Maximum amount of kernel memory to be " 2354 "used for socket buffers"), 2355 sysctl_kern_somaxkva, 0, NULL, 0, 2356 CTL_KERN, KERN_SOMAXKVA, CTL_EOL); 2357 } 2358
Indexes created Sat Sep 20 22:09:52 GMT 2025