uipc_socket.c revision 1.221.2.1
1 /* $NetBSD: uipc_socket.c,v 1.221.2.1 2014/08/10 06:55:58 tls 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 /* 66 * Socket operation routines. 67 * 68 * These routines are called by the routines in sys_socket.c or from a 69 * system process, and implement the semantics of socket operations by 70 * switching out to the protocol specific routines. 71 */ 72 73 #include <sys/cdefs.h> 74 __KERNEL_RCSID(0, "$NetBSD: uipc_socket.c,v 1.221.2.1 2014/08/10 06:55:58 tls Exp $"); 75 76 #include "opt_compat_netbsd.h" 77 #include "opt_sock_counters.h" 78 #include "opt_sosend_loan.h" 79 #include "opt_mbuftrace.h" 80 #include "opt_somaxkva.h" 81 #include "opt_multiprocessor.h" /* XXX */ 82 83 #include <sys/param.h> 84 #include <sys/systm.h> 85 #include <sys/proc.h> 86 #include <sys/file.h> 87 #include <sys/filedesc.h> 88 #include <sys/kmem.h> 89 #include <sys/mbuf.h> 90 #include <sys/domain.h> 91 #include <sys/kernel.h> 92 #include <sys/protosw.h> 93 #include <sys/socket.h> 94 #include <sys/socketvar.h> 95 #include <sys/signalvar.h> 96 #include <sys/resourcevar.h> 97 #include <sys/uidinfo.h> 98 #include <sys/event.h> 99 #include <sys/poll.h> 100 #include <sys/kauth.h> 101 #include <sys/mutex.h> 102 #include <sys/condvar.h> 103 #include <sys/kthread.h> 104 105 #ifdef COMPAT_50 106 #include <compat/sys/time.h> 107 #include <compat/sys/socket.h> 108 #endif 109 110 #include <uvm/uvm_extern.h> 111 #include <uvm/uvm_loan.h> 112 #include <uvm/uvm_page.h> 113 114 MALLOC_DEFINE(M_SONAME, "soname", "socket name"); 115 116 extern const struct fileops socketops; 117 118 extern int somaxconn; /* patchable (XXX sysctl) */ 119 int somaxconn = SOMAXCONN; 120 kmutex_t *softnet_lock; 121 122 #ifdef SOSEND_COUNTERS 123 #include <sys/device.h> 124 125 static struct evcnt sosend_loan_big = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 126 NULL, "sosend", "loan big"); 127 static struct evcnt sosend_copy_big = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 128 NULL, "sosend", "copy big"); 129 static struct evcnt sosend_copy_small = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 130 NULL, "sosend", "copy small"); 131 static struct evcnt sosend_kvalimit = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 132 NULL, "sosend", "kva limit"); 133 134 #define SOSEND_COUNTER_INCR(ev) (ev)->ev_count++ 135 136 EVCNT_ATTACH_STATIC(sosend_loan_big); 137 EVCNT_ATTACH_STATIC(sosend_copy_big); 138 EVCNT_ATTACH_STATIC(sosend_copy_small); 139 EVCNT_ATTACH_STATIC(sosend_kvalimit); 140 #else 141 142 #define SOSEND_COUNTER_INCR(ev) /* nothing */ 143 144 #endif /* SOSEND_COUNTERS */ 145 146 #if defined(SOSEND_NO_LOAN) || defined(MULTIPROCESSOR) 147 int sock_loan_thresh = -1; 148 #else 149 int sock_loan_thresh = 4096; 150 #endif 151 152 static kmutex_t so_pendfree_lock; 153 static struct mbuf *so_pendfree = NULL; 154 155 #ifndef SOMAXKVA 156 #define SOMAXKVA (16 * 1024 * 1024) 157 #endif 158 int somaxkva = SOMAXKVA; 159 static int socurkva; 160 static kcondvar_t socurkva_cv; 161 162 static kauth_listener_t socket_listener; 163 164 #define SOCK_LOAN_CHUNK 65536 165 166 static void sopendfree_thread(void *); 167 static kcondvar_t pendfree_thread_cv; 168 static lwp_t *sopendfree_lwp; 169 170 static void sysctl_kern_socket_setup(void); 171 static struct sysctllog *socket_sysctllog; 172 173 static vsize_t 174 sokvareserve(struct socket *so, vsize_t len) 175 { 176 int error; 177 178 mutex_enter(&so_pendfree_lock); 179 while (socurkva + len > somaxkva) { 180 SOSEND_COUNTER_INCR(&sosend_kvalimit); 181 error = cv_wait_sig(&socurkva_cv, &so_pendfree_lock); 182 if (error) { 183 len = 0; 184 break; 185 } 186 } 187 socurkva += len; 188 mutex_exit(&so_pendfree_lock); 189 return len; 190 } 191 192 static void 193 sokvaunreserve(vsize_t len) 194 { 195 196 mutex_enter(&so_pendfree_lock); 197 socurkva -= len; 198 cv_broadcast(&socurkva_cv); 199 mutex_exit(&so_pendfree_lock); 200 } 201 202 /* 203 * sokvaalloc: allocate kva for loan. 204 */ 205 206 vaddr_t 207 sokvaalloc(vaddr_t sva, vsize_t len, struct socket *so) 208 { 209 vaddr_t lva; 210 211 /* 212 * reserve kva. 213 */ 214 215 if (sokvareserve(so, len) == 0) 216 return 0; 217 218 /* 219 * allocate kva. 220 */ 221 222 lva = uvm_km_alloc(kernel_map, len, atop(sva) & uvmexp.colormask, 223 UVM_KMF_COLORMATCH | UVM_KMF_VAONLY | UVM_KMF_WAITVA); 224 if (lva == 0) { 225 sokvaunreserve(len); 226 return (0); 227 } 228 229 return lva; 230 } 231 232 /* 233 * sokvafree: free kva for loan. 234 */ 235 236 void 237 sokvafree(vaddr_t sva, vsize_t len) 238 { 239 240 /* 241 * free kva. 242 */ 243 244 uvm_km_free(kernel_map, sva, len, UVM_KMF_VAONLY); 245 246 /* 247 * unreserve kva. 248 */ 249 250 sokvaunreserve(len); 251 } 252 253 static void 254 sodoloanfree(struct vm_page **pgs, void *buf, size_t size) 255 { 256 vaddr_t sva, eva; 257 vsize_t len; 258 int npgs; 259 260 KASSERT(pgs != NULL); 261 262 eva = round_page((vaddr_t) buf + size); 263 sva = trunc_page((vaddr_t) buf); 264 len = eva - sva; 265 npgs = len >> PAGE_SHIFT; 266 267 pmap_kremove(sva, len); 268 pmap_update(pmap_kernel()); 269 uvm_unloan(pgs, npgs, UVM_LOAN_TOPAGE); 270 sokvafree(sva, len); 271 } 272 273 /* 274 * sopendfree_thread: free mbufs on "pendfree" list. 275 * unlock and relock so_pendfree_lock when freeing mbufs. 276 */ 277 278 static void 279 sopendfree_thread(void *v) 280 { 281 struct mbuf *m, *next; 282 size_t rv; 283 284 mutex_enter(&so_pendfree_lock); 285 286 for (;;) { 287 rv = 0; 288 while (so_pendfree != NULL) { 289 m = so_pendfree; 290 so_pendfree = NULL; 291 mutex_exit(&so_pendfree_lock); 292 293 for (; m != NULL; m = next) { 294 next = m->m_next; 295 KASSERT((~m->m_flags & (M_EXT|M_EXT_PAGES)) == 0); 296 KASSERT(m->m_ext.ext_refcnt == 0); 297 298 rv += m->m_ext.ext_size; 299 sodoloanfree(m->m_ext.ext_pgs, m->m_ext.ext_buf, 300 m->m_ext.ext_size); 301 pool_cache_put(mb_cache, m); 302 } 303 304 mutex_enter(&so_pendfree_lock); 305 } 306 if (rv) 307 cv_broadcast(&socurkva_cv); 308 cv_wait(&pendfree_thread_cv, &so_pendfree_lock); 309 } 310 panic("sopendfree_thread"); 311 /* NOTREACHED */ 312 } 313 314 void 315 soloanfree(struct mbuf *m, void *buf, size_t size, void *arg) 316 { 317 318 KASSERT(m != NULL); 319 320 /* 321 * postpone freeing mbuf. 322 * 323 * we can't do it in interrupt context 324 * because we need to put kva back to kernel_map. 325 */ 326 327 mutex_enter(&so_pendfree_lock); 328 m->m_next = so_pendfree; 329 so_pendfree = m; 330 cv_signal(&pendfree_thread_cv); 331 mutex_exit(&so_pendfree_lock); 332 } 333 334 static long 335 sosend_loan(struct socket *so, struct uio *uio, struct mbuf *m, long space) 336 { 337 struct iovec *iov = uio->uio_iov; 338 vaddr_t sva, eva; 339 vsize_t len; 340 vaddr_t lva; 341 int npgs, error; 342 vaddr_t va; 343 int i; 344 345 if (VMSPACE_IS_KERNEL_P(uio->uio_vmspace)) 346 return (0); 347 348 if (iov->iov_len < (size_t) space) 349 space = iov->iov_len; 350 if (space > SOCK_LOAN_CHUNK) 351 space = SOCK_LOAN_CHUNK; 352 353 eva = round_page((vaddr_t) iov->iov_base + space); 354 sva = trunc_page((vaddr_t) iov->iov_base); 355 len = eva - sva; 356 npgs = len >> PAGE_SHIFT; 357 358 KASSERT(npgs <= M_EXT_MAXPAGES); 359 360 lva = sokvaalloc(sva, len, so); 361 if (lva == 0) 362 return 0; 363 364 error = uvm_loan(&uio->uio_vmspace->vm_map, sva, len, 365 m->m_ext.ext_pgs, UVM_LOAN_TOPAGE); 366 if (error) { 367 sokvafree(lva, len); 368 return (0); 369 } 370 371 for (i = 0, va = lva; i < npgs; i++, va += PAGE_SIZE) 372 pmap_kenter_pa(va, VM_PAGE_TO_PHYS(m->m_ext.ext_pgs[i]), 373 VM_PROT_READ, 0); 374 pmap_update(pmap_kernel()); 375 376 lva += (vaddr_t) iov->iov_base & PAGE_MASK; 377 378 MEXTADD(m, (void *) lva, space, M_MBUF, soloanfree, so); 379 m->m_flags |= M_EXT_PAGES | M_EXT_ROMAP; 380 381 uio->uio_resid -= space; 382 /* uio_offset not updated, not set/used for write(2) */ 383 uio->uio_iov->iov_base = (char *)uio->uio_iov->iov_base + space; 384 uio->uio_iov->iov_len -= space; 385 if (uio->uio_iov->iov_len == 0) { 386 uio->uio_iov++; 387 uio->uio_iovcnt--; 388 } 389 390 return (space); 391 } 392 393 struct mbuf * 394 getsombuf(struct socket *so, int type) 395 { 396 struct mbuf *m; 397 398 m = m_get(M_WAIT, type); 399 MCLAIM(m, so->so_mowner); 400 return m; 401 } 402 403 static int 404 socket_listener_cb(kauth_cred_t cred, kauth_action_t action, void *cookie, 405 void *arg0, void *arg1, void *arg2, void *arg3) 406 { 407 int result; 408 enum kauth_network_req req; 409 410 result = KAUTH_RESULT_DEFER; 411 req = (enum kauth_network_req)arg0; 412 413 if ((action != KAUTH_NETWORK_SOCKET) && 414 (action != KAUTH_NETWORK_BIND)) 415 return result; 416 417 switch (req) { 418 case KAUTH_REQ_NETWORK_BIND_PORT: 419 result = KAUTH_RESULT_ALLOW; 420 break; 421 422 case KAUTH_REQ_NETWORK_SOCKET_DROP: { 423 /* Normal users can only drop their own connections. */ 424 struct socket *so = (struct socket *)arg1; 425 426 if (so->so_cred && proc_uidmatch(cred, so->so_cred) == 0) 427 result = KAUTH_RESULT_ALLOW; 428 429 break; 430 } 431 432 case KAUTH_REQ_NETWORK_SOCKET_OPEN: 433 /* We allow "raw" routing/bluetooth sockets to anyone. */ 434 if ((u_long)arg1 == PF_ROUTE || (u_long)arg1 == PF_OROUTE 435 || (u_long)arg1 == PF_BLUETOOTH) { 436 result = KAUTH_RESULT_ALLOW; 437 } else { 438 /* Privileged, let secmodel handle this. */ 439 if ((u_long)arg2 == SOCK_RAW) 440 break; 441 } 442 443 result = KAUTH_RESULT_ALLOW; 444 445 break; 446 447 case KAUTH_REQ_NETWORK_SOCKET_CANSEE: 448 result = KAUTH_RESULT_ALLOW; 449 450 break; 451 452 default: 453 break; 454 } 455 456 return result; 457 } 458 459 void 460 soinit(void) 461 { 462 463 sysctl_kern_socket_setup(); 464 465 mutex_init(&so_pendfree_lock, MUTEX_DEFAULT, IPL_VM); 466 softnet_lock = mutex_obj_alloc(MUTEX_DEFAULT, IPL_NONE); 467 cv_init(&socurkva_cv, "sokva"); 468 cv_init(&pendfree_thread_cv, "sopendfr"); 469 soinit2(); 470 471 /* Set the initial adjusted socket buffer size. */ 472 if (sb_max_set(sb_max)) 473 panic("bad initial sb_max value: %lu", sb_max); 474 475 socket_listener = kauth_listen_scope(KAUTH_SCOPE_NETWORK, 476 socket_listener_cb, NULL); 477 } 478 479 void 480 soinit1(void) 481 { 482 int error = kthread_create(PRI_NONE, KTHREAD_MPSAFE, NULL, 483 sopendfree_thread, NULL, &sopendfree_lwp, "sopendfree"); 484 if (error) 485 panic("soinit1 %d", error); 486 } 487 488 /* 489 * socreate: create a new socket of the specified type and the protocol. 490 * 491 * => Caller may specify another socket for lock sharing (must not be held). 492 * => Returns the new socket without lock held. 493 */ 494 int 495 socreate(int dom, struct socket **aso, int type, int proto, struct lwp *l, 496 struct socket *lockso) 497 { 498 const struct protosw *prp; 499 struct socket *so; 500 uid_t uid; 501 int error; 502 kmutex_t *lock; 503 504 error = kauth_authorize_network(l->l_cred, KAUTH_NETWORK_SOCKET, 505 KAUTH_REQ_NETWORK_SOCKET_OPEN, KAUTH_ARG(dom), KAUTH_ARG(type), 506 KAUTH_ARG(proto)); 507 if (error != 0) 508 return error; 509 510 if (proto) 511 prp = pffindproto(dom, proto, type); 512 else 513 prp = pffindtype(dom, type); 514 if (prp == NULL) { 515 /* no support for domain */ 516 if (pffinddomain(dom) == 0) 517 return EAFNOSUPPORT; 518 /* no support for socket type */ 519 if (proto == 0 && type != 0) 520 return EPROTOTYPE; 521 return EPROTONOSUPPORT; 522 } 523 if (prp->pr_usrreqs == NULL) 524 return EPROTONOSUPPORT; 525 if (prp->pr_type != type) 526 return EPROTOTYPE; 527 528 so = soget(true); 529 so->so_type = type; 530 so->so_proto = prp; 531 so->so_send = sosend; 532 so->so_receive = soreceive; 533 #ifdef MBUFTRACE 534 so->so_rcv.sb_mowner = &prp->pr_domain->dom_mowner; 535 so->so_snd.sb_mowner = &prp->pr_domain->dom_mowner; 536 so->so_mowner = &prp->pr_domain->dom_mowner; 537 #endif 538 uid = kauth_cred_geteuid(l->l_cred); 539 so->so_uidinfo = uid_find(uid); 540 so->so_cpid = l->l_proc->p_pid; 541 542 /* 543 * Lock assigned and taken during PCB attach, unless we share 544 * the lock with another socket, e.g. socketpair(2) case. 545 */ 546 if (lockso) { 547 lock = lockso->so_lock; 548 so->so_lock = lock; 549 mutex_obj_hold(lock); 550 mutex_enter(lock); 551 } 552 553 /* Attach the PCB (returns with the socket lock held). */ 554 error = (*prp->pr_usrreqs->pr_attach)(so, proto); 555 KASSERT(solocked(so)); 556 557 if (error) { 558 KASSERT(so->so_pcb == NULL); 559 so->so_state |= SS_NOFDREF; 560 sofree(so); 561 return error; 562 } 563 so->so_cred = kauth_cred_dup(l->l_cred); 564 sounlock(so); 565 566 *aso = so; 567 return 0; 568 } 569 570 /* 571 * fsocreate: create a socket and a file descriptor associated with it. 572 * 573 * => On success, write file descriptor to fdout and return zero. 574 * => On failure, return non-zero; *fdout will be undefined. 575 */ 576 int 577 fsocreate(int domain, struct socket **sop, int type, int proto, int *fdout) 578 { 579 lwp_t *l = curlwp; 580 int error, fd, flags; 581 struct socket *so; 582 struct file *fp; 583 584 if ((error = fd_allocfile(&fp, &fd)) != 0) { 585 return error; 586 } 587 flags = type & SOCK_FLAGS_MASK; 588 fd_set_exclose(l, fd, (flags & SOCK_CLOEXEC) != 0); 589 fp->f_flag = FREAD|FWRITE|((flags & SOCK_NONBLOCK) ? FNONBLOCK : 0)| 590 ((flags & SOCK_NOSIGPIPE) ? FNOSIGPIPE : 0); 591 fp->f_type = DTYPE_SOCKET; 592 fp->f_ops = &socketops; 593 594 type &= ~SOCK_FLAGS_MASK; 595 error = socreate(domain, &so, type, proto, l, NULL); 596 if (error) { 597 fd_abort(curproc, fp, fd); 598 return error; 599 } 600 if (flags & SOCK_NONBLOCK) { 601 so->so_state |= SS_NBIO; 602 } 603 fp->f_data = so; 604 fd_affix(curproc, fp, fd); 605 606 if (sop != NULL) { 607 *sop = so; 608 } 609 *fdout = fd; 610 return error; 611 } 612 613 int 614 sofamily(const struct socket *so) 615 { 616 const struct protosw *pr; 617 const struct domain *dom; 618 619 if ((pr = so->so_proto) == NULL) 620 return AF_UNSPEC; 621 if ((dom = pr->pr_domain) == NULL) 622 return AF_UNSPEC; 623 return dom->dom_family; 624 } 625 626 int 627 sobind(struct socket *so, struct mbuf *nam, struct lwp *l) 628 { 629 int error; 630 631 solock(so); 632 error = (*so->so_proto->pr_usrreqs->pr_bind)(so, nam, l); 633 sounlock(so); 634 return error; 635 } 636 637 int 638 solisten(struct socket *so, int backlog, struct lwp *l) 639 { 640 int error; 641 642 solock(so); 643 if ((so->so_state & (SS_ISCONNECTED | SS_ISCONNECTING | 644 SS_ISDISCONNECTING)) != 0) { 645 sounlock(so); 646 return EINVAL; 647 } 648 error = (*so->so_proto->pr_usrreqs->pr_listen)(so, l); 649 if (error != 0) { 650 sounlock(so); 651 return error; 652 } 653 if (TAILQ_EMPTY(&so->so_q)) 654 so->so_options |= SO_ACCEPTCONN; 655 if (backlog < 0) 656 backlog = 0; 657 so->so_qlimit = min(backlog, somaxconn); 658 sounlock(so); 659 return 0; 660 } 661 662 void 663 sofree(struct socket *so) 664 { 665 u_int refs; 666 667 KASSERT(solocked(so)); 668 669 if (so->so_pcb || (so->so_state & SS_NOFDREF) == 0) { 670 sounlock(so); 671 return; 672 } 673 if (so->so_head) { 674 /* 675 * We must not decommission a socket that's on the accept(2) 676 * queue. If we do, then accept(2) may hang after select(2) 677 * indicated that the listening socket was ready. 678 */ 679 if (!soqremque(so, 0)) { 680 sounlock(so); 681 return; 682 } 683 } 684 if (so->so_rcv.sb_hiwat) 685 (void)chgsbsize(so->so_uidinfo, &so->so_rcv.sb_hiwat, 0, 686 RLIM_INFINITY); 687 if (so->so_snd.sb_hiwat) 688 (void)chgsbsize(so->so_uidinfo, &so->so_snd.sb_hiwat, 0, 689 RLIM_INFINITY); 690 sbrelease(&so->so_snd, so); 691 KASSERT(!cv_has_waiters(&so->so_cv)); 692 KASSERT(!cv_has_waiters(&so->so_rcv.sb_cv)); 693 KASSERT(!cv_has_waiters(&so->so_snd.sb_cv)); 694 sorflush(so); 695 refs = so->so_aborting; /* XXX */ 696 /* Remove acccept filter if one is present. */ 697 if (so->so_accf != NULL) 698 (void)accept_filt_clear(so); 699 sounlock(so); 700 if (refs == 0) /* XXX */ 701 soput(so); 702 } 703 704 /* 705 * soclose: close a socket on last file table reference removal. 706 * Initiate disconnect if connected. Free socket when disconnect complete. 707 */ 708 int 709 soclose(struct socket *so) 710 { 711 struct socket *so2; 712 int error = 0; 713 714 solock(so); 715 if (so->so_options & SO_ACCEPTCONN) { 716 for (;;) { 717 if ((so2 = TAILQ_FIRST(&so->so_q0)) != 0) { 718 KASSERT(solocked2(so, so2)); 719 (void) soqremque(so2, 0); 720 /* soabort drops the lock. */ 721 (void) soabort(so2); 722 solock(so); 723 continue; 724 } 725 if ((so2 = TAILQ_FIRST(&so->so_q)) != 0) { 726 KASSERT(solocked2(so, so2)); 727 (void) soqremque(so2, 1); 728 /* soabort drops the lock. */ 729 (void) soabort(so2); 730 solock(so); 731 continue; 732 } 733 break; 734 } 735 } 736 if (so->so_pcb == NULL) 737 goto discard; 738 if (so->so_state & SS_ISCONNECTED) { 739 if ((so->so_state & SS_ISDISCONNECTING) == 0) { 740 error = sodisconnect(so); 741 if (error) 742 goto drop; 743 } 744 if (so->so_options & SO_LINGER) { 745 if ((so->so_state & (SS_ISDISCONNECTING|SS_NBIO)) == 746 (SS_ISDISCONNECTING|SS_NBIO)) 747 goto drop; 748 while (so->so_state & SS_ISCONNECTED) { 749 error = sowait(so, true, so->so_linger * hz); 750 if (error) 751 break; 752 } 753 } 754 } 755 drop: 756 if (so->so_pcb) { 757 KASSERT(solocked(so)); 758 (*so->so_proto->pr_usrreqs->pr_detach)(so); 759 } 760 discard: 761 KASSERT((so->so_state & SS_NOFDREF) == 0); 762 kauth_cred_free(so->so_cred); 763 so->so_state |= SS_NOFDREF; 764 sofree(so); 765 return error; 766 } 767 768 /* 769 * Must be called with the socket locked.. Will return with it unlocked. 770 */ 771 int 772 soabort(struct socket *so) 773 { 774 u_int refs; 775 int error; 776 777 KASSERT(solocked(so)); 778 KASSERT(so->so_head == NULL); 779 780 so->so_aborting++; /* XXX */ 781 error = (*so->so_proto->pr_usrreqs->pr_abort)(so); 782 refs = --so->so_aborting; /* XXX */ 783 if (error || (refs == 0)) { 784 sofree(so); 785 } else { 786 sounlock(so); 787 } 788 return error; 789 } 790 791 int 792 soaccept(struct socket *so, struct mbuf *nam) 793 { 794 int error; 795 796 KASSERT(solocked(so)); 797 KASSERT((so->so_state & SS_NOFDREF) != 0); 798 799 so->so_state &= ~SS_NOFDREF; 800 if ((so->so_state & SS_ISDISCONNECTED) == 0 || 801 (so->so_proto->pr_flags & PR_ABRTACPTDIS) == 0) 802 error = (*so->so_proto->pr_usrreqs->pr_accept)(so, nam); 803 else 804 error = ECONNABORTED; 805 806 return error; 807 } 808 809 int 810 soconnect(struct socket *so, struct mbuf *nam, struct lwp *l) 811 { 812 int error; 813 814 KASSERT(solocked(so)); 815 816 if (so->so_options & SO_ACCEPTCONN) 817 return EOPNOTSUPP; 818 /* 819 * If protocol is connection-based, can only connect once. 820 * Otherwise, if connected, try to disconnect first. 821 * This allows user to disconnect by connecting to, e.g., 822 * a null address. 823 */ 824 if (so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING) && 825 ((so->so_proto->pr_flags & PR_CONNREQUIRED) || 826 (error = sodisconnect(so)))) 827 error = EISCONN; 828 else 829 error = (*so->so_proto->pr_usrreqs->pr_connect)(so, nam, l); 830 831 return error; 832 } 833 834 int 835 soconnect2(struct socket *so1, struct socket *so2) 836 { 837 KASSERT(solocked2(so1, so2)); 838 839 return (*so1->so_proto->pr_usrreqs->pr_connect2)(so1, so2); 840 } 841 842 int 843 sodisconnect(struct socket *so) 844 { 845 int error; 846 847 KASSERT(solocked(so)); 848 849 if ((so->so_state & SS_ISCONNECTED) == 0) { 850 error = ENOTCONN; 851 } else if (so->so_state & SS_ISDISCONNECTING) { 852 error = EALREADY; 853 } else { 854 error = (*so->so_proto->pr_usrreqs->pr_disconnect)(so); 855 } 856 return (error); 857 } 858 859 #define SBLOCKWAIT(f) (((f) & MSG_DONTWAIT) ? M_NOWAIT : M_WAITOK) 860 /* 861 * Send on a socket. 862 * If send must go all at once and message is larger than 863 * send buffering, then hard error. 864 * Lock against other senders. 865 * If must go all at once and not enough room now, then 866 * inform user that this would block and do nothing. 867 * Otherwise, if nonblocking, send as much as possible. 868 * The data to be sent is described by "uio" if nonzero, 869 * otherwise by the mbuf chain "top" (which must be null 870 * if uio is not). Data provided in mbuf chain must be small 871 * enough to send all at once. 872 * 873 * Returns nonzero on error, timeout or signal; callers 874 * must check for short counts if EINTR/ERESTART are returned. 875 * Data and control buffers are freed on return. 876 */ 877 int 878 sosend(struct socket *so, struct mbuf *addr, struct uio *uio, struct mbuf *top, 879 struct mbuf *control, int flags, struct lwp *l) 880 { 881 struct mbuf **mp, *m; 882 long space, len, resid, clen, mlen; 883 int error, s, dontroute, atomic; 884 short wakeup_state = 0; 885 886 clen = 0; 887 888 /* 889 * solock() provides atomicity of access. splsoftnet() prevents 890 * protocol processing soft interrupts from interrupting us and 891 * blocking (expensive). 892 */ 893 s = splsoftnet(); 894 solock(so); 895 atomic = sosendallatonce(so) || top; 896 if (uio) 897 resid = uio->uio_resid; 898 else 899 resid = top->m_pkthdr.len; 900 /* 901 * In theory resid should be unsigned. 902 * However, space must be signed, as it might be less than 0 903 * if we over-committed, and we must use a signed comparison 904 * of space and resid. On the other hand, a negative resid 905 * causes us to loop sending 0-length segments to the protocol. 906 */ 907 if (resid < 0) { 908 error = EINVAL; 909 goto out; 910 } 911 dontroute = 912 (flags & MSG_DONTROUTE) && (so->so_options & SO_DONTROUTE) == 0 && 913 (so->so_proto->pr_flags & PR_ATOMIC); 914 l->l_ru.ru_msgsnd++; 915 if (control) 916 clen = control->m_len; 917 restart: 918 if ((error = sblock(&so->so_snd, SBLOCKWAIT(flags))) != 0) 919 goto out; 920 do { 921 if (so->so_state & SS_CANTSENDMORE) { 922 error = EPIPE; 923 goto release; 924 } 925 if (so->so_error) { 926 error = so->so_error; 927 so->so_error = 0; 928 goto release; 929 } 930 if ((so->so_state & SS_ISCONNECTED) == 0) { 931 if (so->so_proto->pr_flags & PR_CONNREQUIRED) { 932 if (resid || clen == 0) { 933 error = ENOTCONN; 934 goto release; 935 } 936 } else if (addr == 0) { 937 error = EDESTADDRREQ; 938 goto release; 939 } 940 } 941 space = sbspace(&so->so_snd); 942 if (flags & MSG_OOB) 943 space += 1024; 944 if ((atomic && resid > so->so_snd.sb_hiwat) || 945 clen > so->so_snd.sb_hiwat) { 946 error = EMSGSIZE; 947 goto release; 948 } 949 if (space < resid + clen && 950 (atomic || space < so->so_snd.sb_lowat || space < clen)) { 951 if ((so->so_state & SS_NBIO) || (flags & MSG_NBIO)) { 952 error = EWOULDBLOCK; 953 goto release; 954 } 955 sbunlock(&so->so_snd); 956 if (wakeup_state & SS_RESTARTSYS) { 957 error = ERESTART; 958 goto out; 959 } 960 error = sbwait(&so->so_snd); 961 if (error) 962 goto out; 963 wakeup_state = so->so_state; 964 goto restart; 965 } 966 wakeup_state = 0; 967 mp = ⊤ 968 space -= clen; 969 do { 970 if (uio == NULL) { 971 /* 972 * Data is prepackaged in "top". 973 */ 974 resid = 0; 975 if (flags & MSG_EOR) 976 top->m_flags |= M_EOR; 977 } else do { 978 sounlock(so); 979 splx(s); 980 if (top == NULL) { 981 m = m_gethdr(M_WAIT, MT_DATA); 982 mlen = MHLEN; 983 m->m_pkthdr.len = 0; 984 m->m_pkthdr.rcvif = NULL; 985 } else { 986 m = m_get(M_WAIT, MT_DATA); 987 mlen = MLEN; 988 } 989 MCLAIM(m, so->so_snd.sb_mowner); 990 if (sock_loan_thresh >= 0 && 991 uio->uio_iov->iov_len >= sock_loan_thresh && 992 space >= sock_loan_thresh && 993 (len = sosend_loan(so, uio, m, 994 space)) != 0) { 995 SOSEND_COUNTER_INCR(&sosend_loan_big); 996 space -= len; 997 goto have_data; 998 } 999 if (resid >= MINCLSIZE && space >= MCLBYTES) { 1000 SOSEND_COUNTER_INCR(&sosend_copy_big); 1001 m_clget(m, M_DONTWAIT); 1002 if ((m->m_flags & M_EXT) == 0) 1003 goto nopages; 1004 mlen = MCLBYTES; 1005 if (atomic && top == 0) { 1006 len = lmin(MCLBYTES - max_hdr, 1007 resid); 1008 m->m_data += max_hdr; 1009 } else 1010 len = lmin(MCLBYTES, resid); 1011 space -= len; 1012 } else { 1013 nopages: 1014 SOSEND_COUNTER_INCR(&sosend_copy_small); 1015 len = lmin(lmin(mlen, resid), space); 1016 space -= len; 1017 /* 1018 * For datagram protocols, leave room 1019 * for protocol headers in first mbuf. 1020 */ 1021 if (atomic && top == 0 && len < mlen) 1022 MH_ALIGN(m, len); 1023 } 1024 error = uiomove(mtod(m, void *), (int)len, uio); 1025 have_data: 1026 resid = uio->uio_resid; 1027 m->m_len = len; 1028 *mp = m; 1029 top->m_pkthdr.len += len; 1030 s = splsoftnet(); 1031 solock(so); 1032 if (error != 0) 1033 goto release; 1034 mp = &m->m_next; 1035 if (resid <= 0) { 1036 if (flags & MSG_EOR) 1037 top->m_flags |= M_EOR; 1038 break; 1039 } 1040 } while (space > 0 && atomic); 1041 1042 if (so->so_state & SS_CANTSENDMORE) { 1043 error = EPIPE; 1044 goto release; 1045 } 1046 if (dontroute) 1047 so->so_options |= SO_DONTROUTE; 1048 if (resid > 0) 1049 so->so_state |= SS_MORETOCOME; 1050 if (flags & MSG_OOB) 1051 error = (*so->so_proto->pr_usrreqs->pr_sendoob)(so, 1052 top, control); 1053 else 1054 error = (*so->so_proto->pr_usrreqs->pr_send)(so, 1055 top, addr, control, l); 1056 if (dontroute) 1057 so->so_options &= ~SO_DONTROUTE; 1058 if (resid > 0) 1059 so->so_state &= ~SS_MORETOCOME; 1060 clen = 0; 1061 control = NULL; 1062 top = NULL; 1063 mp = ⊤ 1064 if (error != 0) 1065 goto release; 1066 } while (resid && space > 0); 1067 } while (resid); 1068 1069 release: 1070 sbunlock(&so->so_snd); 1071 out: 1072 sounlock(so); 1073 splx(s); 1074 if (top) 1075 m_freem(top); 1076 if (control) 1077 m_freem(control); 1078 return (error); 1079 } 1080 1081 /* 1082 * Following replacement or removal of the first mbuf on the first 1083 * mbuf chain of a socket buffer, push necessary state changes back 1084 * into the socket buffer so that other consumers see the values 1085 * consistently. 'nextrecord' is the callers locally stored value of 1086 * the original value of sb->sb_mb->m_nextpkt which must be restored 1087 * when the lead mbuf changes. NOTE: 'nextrecord' may be NULL. 1088 */ 1089 static void 1090 sbsync(struct sockbuf *sb, struct mbuf *nextrecord) 1091 { 1092 1093 KASSERT(solocked(sb->sb_so)); 1094 1095 /* 1096 * First, update for the new value of nextrecord. If necessary, 1097 * make it the first record. 1098 */ 1099 if (sb->sb_mb != NULL) 1100 sb->sb_mb->m_nextpkt = nextrecord; 1101 else 1102 sb->sb_mb = nextrecord; 1103 1104 /* 1105 * Now update any dependent socket buffer fields to reflect 1106 * the new state. This is an inline of SB_EMPTY_FIXUP, with 1107 * the addition of a second clause that takes care of the 1108 * case where sb_mb has been updated, but remains the last 1109 * record. 1110 */ 1111 if (sb->sb_mb == NULL) { 1112 sb->sb_mbtail = NULL; 1113 sb->sb_lastrecord = NULL; 1114 } else if (sb->sb_mb->m_nextpkt == NULL) 1115 sb->sb_lastrecord = sb->sb_mb; 1116 } 1117 1118 /* 1119 * Implement receive operations on a socket. 1120 * We depend on the way that records are added to the sockbuf 1121 * by sbappend*. In particular, each record (mbufs linked through m_next) 1122 * must begin with an address if the protocol so specifies, 1123 * followed by an optional mbuf or mbufs containing ancillary data, 1124 * and then zero or more mbufs of data. 1125 * In order to avoid blocking network interrupts for the entire time here, 1126 * we splx() while doing the actual copy to user space. 1127 * Although the sockbuf is locked, new data may still be appended, 1128 * and thus we must maintain consistency of the sockbuf during that time. 1129 * 1130 * The caller may receive the data as a single mbuf chain by supplying 1131 * an mbuf **mp0 for use in returning the chain. The uio is then used 1132 * only for the count in uio_resid. 1133 */ 1134 int 1135 soreceive(struct socket *so, struct mbuf **paddr, struct uio *uio, 1136 struct mbuf **mp0, struct mbuf **controlp, int *flagsp) 1137 { 1138 struct lwp *l = curlwp; 1139 struct mbuf *m, **mp, *mt; 1140 size_t len, offset, moff, orig_resid; 1141 int atomic, flags, error, s, type; 1142 const struct protosw *pr; 1143 struct mbuf *nextrecord; 1144 int mbuf_removed = 0; 1145 const struct domain *dom; 1146 short wakeup_state = 0; 1147 1148 pr = so->so_proto; 1149 atomic = pr->pr_flags & PR_ATOMIC; 1150 dom = pr->pr_domain; 1151 mp = mp0; 1152 type = 0; 1153 orig_resid = uio->uio_resid; 1154 1155 if (paddr != NULL) 1156 *paddr = NULL; 1157 if (controlp != NULL) 1158 *controlp = NULL; 1159 if (flagsp != NULL) 1160 flags = *flagsp &~ MSG_EOR; 1161 else 1162 flags = 0; 1163 1164 if (flags & MSG_OOB) { 1165 m = m_get(M_WAIT, MT_DATA); 1166 solock(so); 1167 error = (*pr->pr_usrreqs->pr_recvoob)(so, m, flags & MSG_PEEK); 1168 sounlock(so); 1169 if (error) 1170 goto bad; 1171 do { 1172 error = uiomove(mtod(m, void *), 1173 MIN(uio->uio_resid, m->m_len), uio); 1174 m = m_free(m); 1175 } while (uio->uio_resid > 0 && error == 0 && m); 1176 bad: 1177 if (m != NULL) 1178 m_freem(m); 1179 return error; 1180 } 1181 if (mp != NULL) 1182 *mp = NULL; 1183 1184 /* 1185 * solock() provides atomicity of access. splsoftnet() prevents 1186 * protocol processing soft interrupts from interrupting us and 1187 * blocking (expensive). 1188 */ 1189 s = splsoftnet(); 1190 solock(so); 1191 restart: 1192 if ((error = sblock(&so->so_rcv, SBLOCKWAIT(flags))) != 0) { 1193 sounlock(so); 1194 splx(s); 1195 return error; 1196 } 1197 1198 m = so->so_rcv.sb_mb; 1199 /* 1200 * If we have less data than requested, block awaiting more 1201 * (subject to any timeout) if: 1202 * 1. the current count is less than the low water mark, 1203 * 2. MSG_WAITALL is set, and it is possible to do the entire 1204 * receive operation at once if we block (resid <= hiwat), or 1205 * 3. MSG_DONTWAIT is not set. 1206 * If MSG_WAITALL is set but resid is larger than the receive buffer, 1207 * we have to do the receive in sections, and thus risk returning 1208 * a short count if a timeout or signal occurs after we start. 1209 */ 1210 if (m == NULL || 1211 ((flags & MSG_DONTWAIT) == 0 && 1212 so->so_rcv.sb_cc < uio->uio_resid && 1213 (so->so_rcv.sb_cc < so->so_rcv.sb_lowat || 1214 ((flags & MSG_WAITALL) && 1215 uio->uio_resid <= so->so_rcv.sb_hiwat)) && 1216 m->m_nextpkt == NULL && !atomic)) { 1217 #ifdef DIAGNOSTIC 1218 if (m == NULL && so->so_rcv.sb_cc) 1219 panic("receive 1"); 1220 #endif 1221 if (so->so_error) { 1222 if (m != NULL) 1223 goto dontblock; 1224 error = so->so_error; 1225 if ((flags & MSG_PEEK) == 0) 1226 so->so_error = 0; 1227 goto release; 1228 } 1229 if (so->so_state & SS_CANTRCVMORE) { 1230 if (m != NULL) 1231 goto dontblock; 1232 else 1233 goto release; 1234 } 1235 for (; m != NULL; m = m->m_next) 1236 if (m->m_type == MT_OOBDATA || (m->m_flags & M_EOR)) { 1237 m = so->so_rcv.sb_mb; 1238 goto dontblock; 1239 } 1240 if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 && 1241 (so->so_proto->pr_flags & PR_CONNREQUIRED)) { 1242 error = ENOTCONN; 1243 goto release; 1244 } 1245 if (uio->uio_resid == 0) 1246 goto release; 1247 if ((so->so_state & SS_NBIO) || 1248 (flags & (MSG_DONTWAIT|MSG_NBIO))) { 1249 error = EWOULDBLOCK; 1250 goto release; 1251 } 1252 SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 1"); 1253 SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 1"); 1254 sbunlock(&so->so_rcv); 1255 if (wakeup_state & SS_RESTARTSYS) 1256 error = ERESTART; 1257 else 1258 error = sbwait(&so->so_rcv); 1259 if (error != 0) { 1260 sounlock(so); 1261 splx(s); 1262 return error; 1263 } 1264 wakeup_state = so->so_state; 1265 goto restart; 1266 } 1267 dontblock: 1268 /* 1269 * On entry here, m points to the first record of the socket buffer. 1270 * From this point onward, we maintain 'nextrecord' as a cache of the 1271 * pointer to the next record in the socket buffer. We must keep the 1272 * various socket buffer pointers and local stack versions of the 1273 * pointers in sync, pushing out modifications before dropping the 1274 * socket lock, and re-reading them when picking it up. 1275 * 1276 * Otherwise, we will race with the network stack appending new data 1277 * or records onto the socket buffer by using inconsistent/stale 1278 * versions of the field, possibly resulting in socket buffer 1279 * corruption. 1280 * 1281 * By holding the high-level sblock(), we prevent simultaneous 1282 * readers from pulling off the front of the socket buffer. 1283 */ 1284 if (l != NULL) 1285 l->l_ru.ru_msgrcv++; 1286 KASSERT(m == so->so_rcv.sb_mb); 1287 SBLASTRECORDCHK(&so->so_rcv, "soreceive 1"); 1288 SBLASTMBUFCHK(&so->so_rcv, "soreceive 1"); 1289 nextrecord = m->m_nextpkt; 1290 if (pr->pr_flags & PR_ADDR) { 1291 #ifdef DIAGNOSTIC 1292 if (m->m_type != MT_SONAME) 1293 panic("receive 1a"); 1294 #endif 1295 orig_resid = 0; 1296 if (flags & MSG_PEEK) { 1297 if (paddr) 1298 *paddr = m_copy(m, 0, m->m_len); 1299 m = m->m_next; 1300 } else { 1301 sbfree(&so->so_rcv, m); 1302 mbuf_removed = 1; 1303 if (paddr != NULL) { 1304 *paddr = m; 1305 so->so_rcv.sb_mb = m->m_next; 1306 m->m_next = NULL; 1307 m = so->so_rcv.sb_mb; 1308 } else { 1309 MFREE(m, so->so_rcv.sb_mb); 1310 m = so->so_rcv.sb_mb; 1311 } 1312 sbsync(&so->so_rcv, nextrecord); 1313 } 1314 } 1315 1316 /* 1317 * Process one or more MT_CONTROL mbufs present before any data mbufs 1318 * in the first mbuf chain on the socket buffer. If MSG_PEEK, we 1319 * just copy the data; if !MSG_PEEK, we call into the protocol to 1320 * perform externalization (or freeing if controlp == NULL). 1321 */ 1322 if (__predict_false(m != NULL && m->m_type == MT_CONTROL)) { 1323 struct mbuf *cm = NULL, *cmn; 1324 struct mbuf **cme = &cm; 1325 1326 do { 1327 if (flags & MSG_PEEK) { 1328 if (controlp != NULL) { 1329 *controlp = m_copy(m, 0, m->m_len); 1330 controlp = &(*controlp)->m_next; 1331 } 1332 m = m->m_next; 1333 } else { 1334 sbfree(&so->so_rcv, m); 1335 so->so_rcv.sb_mb = m->m_next; 1336 m->m_next = NULL; 1337 *cme = m; 1338 cme = &(*cme)->m_next; 1339 m = so->so_rcv.sb_mb; 1340 } 1341 } while (m != NULL && m->m_type == MT_CONTROL); 1342 if ((flags & MSG_PEEK) == 0) 1343 sbsync(&so->so_rcv, nextrecord); 1344 for (; cm != NULL; cm = cmn) { 1345 cmn = cm->m_next; 1346 cm->m_next = NULL; 1347 type = mtod(cm, struct cmsghdr *)->cmsg_type; 1348 if (controlp != NULL) { 1349 if (dom->dom_externalize != NULL && 1350 type == SCM_RIGHTS) { 1351 sounlock(so); 1352 splx(s); 1353 error = (*dom->dom_externalize)(cm, l, 1354 (flags & MSG_CMSG_CLOEXEC) ? 1355 O_CLOEXEC : 0); 1356 s = splsoftnet(); 1357 solock(so); 1358 } 1359 *controlp = cm; 1360 while (*controlp != NULL) 1361 controlp = &(*controlp)->m_next; 1362 } else { 1363 /* 1364 * Dispose of any SCM_RIGHTS message that went 1365 * through the read path rather than recv. 1366 */ 1367 if (dom->dom_dispose != NULL && 1368 type == SCM_RIGHTS) { 1369 sounlock(so); 1370 (*dom->dom_dispose)(cm); 1371 solock(so); 1372 } 1373 m_freem(cm); 1374 } 1375 } 1376 if (m != NULL) 1377 nextrecord = so->so_rcv.sb_mb->m_nextpkt; 1378 else 1379 nextrecord = so->so_rcv.sb_mb; 1380 orig_resid = 0; 1381 } 1382 1383 /* If m is non-NULL, we have some data to read. */ 1384 if (__predict_true(m != NULL)) { 1385 type = m->m_type; 1386 if (type == MT_OOBDATA) 1387 flags |= MSG_OOB; 1388 } 1389 SBLASTRECORDCHK(&so->so_rcv, "soreceive 2"); 1390 SBLASTMBUFCHK(&so->so_rcv, "soreceive 2"); 1391 1392 moff = 0; 1393 offset = 0; 1394 while (m != NULL && uio->uio_resid > 0 && error == 0) { 1395 if (m->m_type == MT_OOBDATA) { 1396 if (type != MT_OOBDATA) 1397 break; 1398 } else if (type == MT_OOBDATA) 1399 break; 1400 #ifdef DIAGNOSTIC 1401 else if (m->m_type != MT_DATA && m->m_type != MT_HEADER) 1402 panic("receive 3"); 1403 #endif 1404 so->so_state &= ~SS_RCVATMARK; 1405 wakeup_state = 0; 1406 len = uio->uio_resid; 1407 if (so->so_oobmark && len > so->so_oobmark - offset) 1408 len = so->so_oobmark - offset; 1409 if (len > m->m_len - moff) 1410 len = m->m_len - moff; 1411 /* 1412 * If mp is set, just pass back the mbufs. 1413 * Otherwise copy them out via the uio, then free. 1414 * Sockbuf must be consistent here (points to current mbuf, 1415 * it points to next record) when we drop priority; 1416 * we must note any additions to the sockbuf when we 1417 * block interrupts again. 1418 */ 1419 if (mp == NULL) { 1420 SBLASTRECORDCHK(&so->so_rcv, "soreceive uiomove"); 1421 SBLASTMBUFCHK(&so->so_rcv, "soreceive uiomove"); 1422 sounlock(so); 1423 splx(s); 1424 error = uiomove(mtod(m, char *) + moff, len, uio); 1425 s = splsoftnet(); 1426 solock(so); 1427 if (error != 0) { 1428 /* 1429 * If any part of the record has been removed 1430 * (such as the MT_SONAME mbuf, which will 1431 * happen when PR_ADDR, and thus also 1432 * PR_ATOMIC, is set), then drop the entire 1433 * record to maintain the atomicity of the 1434 * receive operation. 1435 * 1436 * This avoids a later panic("receive 1a") 1437 * when compiled with DIAGNOSTIC. 1438 */ 1439 if (m && mbuf_removed && atomic) 1440 (void) sbdroprecord(&so->so_rcv); 1441 1442 goto release; 1443 } 1444 } else 1445 uio->uio_resid -= len; 1446 if (len == m->m_len - moff) { 1447 if (m->m_flags & M_EOR) 1448 flags |= MSG_EOR; 1449 if (flags & MSG_PEEK) { 1450 m = m->m_next; 1451 moff = 0; 1452 } else { 1453 nextrecord = m->m_nextpkt; 1454 sbfree(&so->so_rcv, m); 1455 if (mp) { 1456 *mp = m; 1457 mp = &m->m_next; 1458 so->so_rcv.sb_mb = m = m->m_next; 1459 *mp = NULL; 1460 } else { 1461 MFREE(m, so->so_rcv.sb_mb); 1462 m = so->so_rcv.sb_mb; 1463 } 1464 /* 1465 * If m != NULL, we also know that 1466 * so->so_rcv.sb_mb != NULL. 1467 */ 1468 KASSERT(so->so_rcv.sb_mb == m); 1469 if (m) { 1470 m->m_nextpkt = nextrecord; 1471 if (nextrecord == NULL) 1472 so->so_rcv.sb_lastrecord = m; 1473 } else { 1474 so->so_rcv.sb_mb = nextrecord; 1475 SB_EMPTY_FIXUP(&so->so_rcv); 1476 } 1477 SBLASTRECORDCHK(&so->so_rcv, "soreceive 3"); 1478 SBLASTMBUFCHK(&so->so_rcv, "soreceive 3"); 1479 } 1480 } else if (flags & MSG_PEEK) 1481 moff += len; 1482 else { 1483 if (mp != NULL) { 1484 mt = m_copym(m, 0, len, M_NOWAIT); 1485 if (__predict_false(mt == NULL)) { 1486 sounlock(so); 1487 mt = m_copym(m, 0, len, M_WAIT); 1488 solock(so); 1489 } 1490 *mp = mt; 1491 } 1492 m->m_data += len; 1493 m->m_len -= len; 1494 so->so_rcv.sb_cc -= len; 1495 } 1496 if (so->so_oobmark) { 1497 if ((flags & MSG_PEEK) == 0) { 1498 so->so_oobmark -= len; 1499 if (so->so_oobmark == 0) { 1500 so->so_state |= SS_RCVATMARK; 1501 break; 1502 } 1503 } else { 1504 offset += len; 1505 if (offset == so->so_oobmark) 1506 break; 1507 } 1508 } 1509 if (flags & MSG_EOR) 1510 break; 1511 /* 1512 * If the MSG_WAITALL flag is set (for non-atomic socket), 1513 * we must not quit until "uio->uio_resid == 0" or an error 1514 * termination. If a signal/timeout occurs, return 1515 * with a short count but without error. 1516 * Keep sockbuf locked against other readers. 1517 */ 1518 while (flags & MSG_WAITALL && m == NULL && uio->uio_resid > 0 && 1519 !sosendallatonce(so) && !nextrecord) { 1520 if (so->so_error || so->so_state & SS_CANTRCVMORE) 1521 break; 1522 /* 1523 * If we are peeking and the socket receive buffer is 1524 * full, stop since we can't get more data to peek at. 1525 */ 1526 if ((flags & MSG_PEEK) && sbspace(&so->so_rcv) <= 0) 1527 break; 1528 /* 1529 * If we've drained the socket buffer, tell the 1530 * protocol in case it needs to do something to 1531 * get it filled again. 1532 */ 1533 if ((pr->pr_flags & PR_WANTRCVD) && so->so_pcb) 1534 (*pr->pr_usrreqs->pr_rcvd)(so, flags, l); 1535 SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 2"); 1536 SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 2"); 1537 if (wakeup_state & SS_RESTARTSYS) 1538 error = ERESTART; 1539 else 1540 error = sbwait(&so->so_rcv); 1541 if (error != 0) { 1542 sbunlock(&so->so_rcv); 1543 sounlock(so); 1544 splx(s); 1545 return 0; 1546 } 1547 if ((m = so->so_rcv.sb_mb) != NULL) 1548 nextrecord = m->m_nextpkt; 1549 wakeup_state = so->so_state; 1550 } 1551 } 1552 1553 if (m && atomic) { 1554 flags |= MSG_TRUNC; 1555 if ((flags & MSG_PEEK) == 0) 1556 (void) sbdroprecord(&so->so_rcv); 1557 } 1558 if ((flags & MSG_PEEK) == 0) { 1559 if (m == NULL) { 1560 /* 1561 * First part is an inline SB_EMPTY_FIXUP(). Second 1562 * part makes sure sb_lastrecord is up-to-date if 1563 * there is still data in the socket buffer. 1564 */ 1565 so->so_rcv.sb_mb = nextrecord; 1566 if (so->so_rcv.sb_mb == NULL) { 1567 so->so_rcv.sb_mbtail = NULL; 1568 so->so_rcv.sb_lastrecord = NULL; 1569 } else if (nextrecord->m_nextpkt == NULL) 1570 so->so_rcv.sb_lastrecord = nextrecord; 1571 } 1572 SBLASTRECORDCHK(&so->so_rcv, "soreceive 4"); 1573 SBLASTMBUFCHK(&so->so_rcv, "soreceive 4"); 1574 if (pr->pr_flags & PR_WANTRCVD && so->so_pcb) 1575 (*pr->pr_usrreqs->pr_rcvd)(so, flags, l); 1576 } 1577 if (orig_resid == uio->uio_resid && orig_resid && 1578 (flags & MSG_EOR) == 0 && (so->so_state & SS_CANTRCVMORE) == 0) { 1579 sbunlock(&so->so_rcv); 1580 goto restart; 1581 } 1582 1583 if (flagsp != NULL) 1584 *flagsp |= flags; 1585 release: 1586 sbunlock(&so->so_rcv); 1587 sounlock(so); 1588 splx(s); 1589 return error; 1590 } 1591 1592 int 1593 soshutdown(struct socket *so, int how) 1594 { 1595 const struct protosw *pr; 1596 int error; 1597 1598 KASSERT(solocked(so)); 1599 1600 pr = so->so_proto; 1601 if (!(how == SHUT_RD || how == SHUT_WR || how == SHUT_RDWR)) 1602 return (EINVAL); 1603 1604 if (how == SHUT_RD || how == SHUT_RDWR) { 1605 sorflush(so); 1606 error = 0; 1607 } 1608 if (how == SHUT_WR || how == SHUT_RDWR) 1609 error = (*pr->pr_usrreqs->pr_shutdown)(so); 1610 1611 return error; 1612 } 1613 1614 void 1615 sorestart(struct socket *so) 1616 { 1617 /* 1618 * An application has called close() on an fd on which another 1619 * of its threads has called a socket system call. 1620 * Mark this and wake everyone up, and code that would block again 1621 * instead returns ERESTART. 1622 * On system call re-entry the fd is validated and EBADF returned. 1623 * Any other fd will block again on the 2nd syscall. 1624 */ 1625 solock(so); 1626 so->so_state |= SS_RESTARTSYS; 1627 cv_broadcast(&so->so_cv); 1628 cv_broadcast(&so->so_snd.sb_cv); 1629 cv_broadcast(&so->so_rcv.sb_cv); 1630 sounlock(so); 1631 } 1632 1633 void 1634 sorflush(struct socket *so) 1635 { 1636 struct sockbuf *sb, asb; 1637 const struct protosw *pr; 1638 1639 KASSERT(solocked(so)); 1640 1641 sb = &so->so_rcv; 1642 pr = so->so_proto; 1643 socantrcvmore(so); 1644 sb->sb_flags |= SB_NOINTR; 1645 (void )sblock(sb, M_WAITOK); 1646 sbunlock(sb); 1647 asb = *sb; 1648 /* 1649 * Clear most of the sockbuf structure, but leave some of the 1650 * fields valid. 1651 */ 1652 memset(&sb->sb_startzero, 0, 1653 sizeof(*sb) - offsetof(struct sockbuf, sb_startzero)); 1654 if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose) { 1655 sounlock(so); 1656 (*pr->pr_domain->dom_dispose)(asb.sb_mb); 1657 solock(so); 1658 } 1659 sbrelease(&asb, so); 1660 } 1661 1662 /* 1663 * internal set SOL_SOCKET options 1664 */ 1665 static int 1666 sosetopt1(struct socket *so, const struct sockopt *sopt) 1667 { 1668 int error = EINVAL, opt; 1669 int optval = 0; /* XXX: gcc */ 1670 struct linger l; 1671 struct timeval tv; 1672 1673 switch ((opt = sopt->sopt_name)) { 1674 1675 case SO_ACCEPTFILTER: 1676 error = accept_filt_setopt(so, sopt); 1677 KASSERT(solocked(so)); 1678 break; 1679 1680 case SO_LINGER: 1681 error = sockopt_get(sopt, &l, sizeof(l)); 1682 solock(so); 1683 if (error) 1684 break; 1685 if (l.l_linger < 0 || l.l_linger > USHRT_MAX || 1686 l.l_linger > (INT_MAX / hz)) { 1687 error = EDOM; 1688 break; 1689 } 1690 so->so_linger = l.l_linger; 1691 if (l.l_onoff) 1692 so->so_options |= SO_LINGER; 1693 else 1694 so->so_options &= ~SO_LINGER; 1695 break; 1696 1697 case SO_DEBUG: 1698 case SO_KEEPALIVE: 1699 case SO_DONTROUTE: 1700 case SO_USELOOPBACK: 1701 case SO_BROADCAST: 1702 case SO_REUSEADDR: 1703 case SO_REUSEPORT: 1704 case SO_OOBINLINE: 1705 case SO_TIMESTAMP: 1706 case SO_NOSIGPIPE: 1707 #ifdef SO_OTIMESTAMP 1708 case SO_OTIMESTAMP: 1709 #endif 1710 error = sockopt_getint(sopt, &optval); 1711 solock(so); 1712 if (error) 1713 break; 1714 if (optval) 1715 so->so_options |= opt; 1716 else 1717 so->so_options &= ~opt; 1718 break; 1719 1720 case SO_SNDBUF: 1721 case SO_RCVBUF: 1722 case SO_SNDLOWAT: 1723 case SO_RCVLOWAT: 1724 error = sockopt_getint(sopt, &optval); 1725 solock(so); 1726 if (error) 1727 break; 1728 1729 /* 1730 * Values < 1 make no sense for any of these 1731 * options, so disallow them. 1732 */ 1733 if (optval < 1) { 1734 error = EINVAL; 1735 break; 1736 } 1737 1738 switch (opt) { 1739 case SO_SNDBUF: 1740 if (sbreserve(&so->so_snd, (u_long)optval, so) == 0) { 1741 error = ENOBUFS; 1742 break; 1743 } 1744 so->so_snd.sb_flags &= ~SB_AUTOSIZE; 1745 break; 1746 1747 case SO_RCVBUF: 1748 if (sbreserve(&so->so_rcv, (u_long)optval, so) == 0) { 1749 error = ENOBUFS; 1750 break; 1751 } 1752 so->so_rcv.sb_flags &= ~SB_AUTOSIZE; 1753 break; 1754 1755 /* 1756 * Make sure the low-water is never greater than 1757 * the high-water. 1758 */ 1759 case SO_SNDLOWAT: 1760 if (optval > so->so_snd.sb_hiwat) 1761 optval = so->so_snd.sb_hiwat; 1762 1763 so->so_snd.sb_lowat = optval; 1764 break; 1765 1766 case SO_RCVLOWAT: 1767 if (optval > so->so_rcv.sb_hiwat) 1768 optval = so->so_rcv.sb_hiwat; 1769 1770 so->so_rcv.sb_lowat = optval; 1771 break; 1772 } 1773 break; 1774 1775 #ifdef COMPAT_50 1776 case SO_OSNDTIMEO: 1777 case SO_ORCVTIMEO: { 1778 struct timeval50 otv; 1779 error = sockopt_get(sopt, &otv, sizeof(otv)); 1780 if (error) { 1781 solock(so); 1782 break; 1783 } 1784 timeval50_to_timeval(&otv, &tv); 1785 opt = opt == SO_OSNDTIMEO ? SO_SNDTIMEO : SO_RCVTIMEO; 1786 error = 0; 1787 /*FALLTHROUGH*/ 1788 } 1789 #endif /* COMPAT_50 */ 1790 1791 case SO_SNDTIMEO: 1792 case SO_RCVTIMEO: 1793 if (error) 1794 error = sockopt_get(sopt, &tv, sizeof(tv)); 1795 solock(so); 1796 if (error) 1797 break; 1798 1799 if (tv.tv_sec > (INT_MAX - tv.tv_usec / tick) / hz) { 1800 error = EDOM; 1801 break; 1802 } 1803 1804 optval = tv.tv_sec * hz + tv.tv_usec / tick; 1805 if (optval == 0 && tv.tv_usec != 0) 1806 optval = 1; 1807 1808 switch (opt) { 1809 case SO_SNDTIMEO: 1810 so->so_snd.sb_timeo = optval; 1811 break; 1812 case SO_RCVTIMEO: 1813 so->so_rcv.sb_timeo = optval; 1814 break; 1815 } 1816 break; 1817 1818 default: 1819 solock(so); 1820 error = ENOPROTOOPT; 1821 break; 1822 } 1823 KASSERT(solocked(so)); 1824 return error; 1825 } 1826 1827 int 1828 sosetopt(struct socket *so, struct sockopt *sopt) 1829 { 1830 int error, prerr; 1831 1832 if (sopt->sopt_level == SOL_SOCKET) { 1833 error = sosetopt1(so, sopt); 1834 KASSERT(solocked(so)); 1835 } else { 1836 error = ENOPROTOOPT; 1837 solock(so); 1838 } 1839 1840 if ((error == 0 || error == ENOPROTOOPT) && 1841 so->so_proto != NULL && so->so_proto->pr_ctloutput != NULL) { 1842 /* give the protocol stack a shot */ 1843 prerr = (*so->so_proto->pr_ctloutput)(PRCO_SETOPT, so, sopt); 1844 if (prerr == 0) 1845 error = 0; 1846 else if (prerr != ENOPROTOOPT) 1847 error = prerr; 1848 } 1849 sounlock(so); 1850 return error; 1851 } 1852 1853 /* 1854 * so_setsockopt() is a wrapper providing a sockopt structure for sosetopt() 1855 */ 1856 int 1857 so_setsockopt(struct lwp *l, struct socket *so, int level, int name, 1858 const void *val, size_t valsize) 1859 { 1860 struct sockopt sopt; 1861 int error; 1862 1863 KASSERT(valsize == 0 || val != NULL); 1864 1865 sockopt_init(&sopt, level, name, valsize); 1866 sockopt_set(&sopt, val, valsize); 1867 1868 error = sosetopt(so, &sopt); 1869 1870 sockopt_destroy(&sopt); 1871 1872 return error; 1873 } 1874 1875 /* 1876 * internal get SOL_SOCKET options 1877 */ 1878 static int 1879 sogetopt1(struct socket *so, struct sockopt *sopt) 1880 { 1881 int error, optval, opt; 1882 struct linger l; 1883 struct timeval tv; 1884 1885 switch ((opt = sopt->sopt_name)) { 1886 1887 case SO_ACCEPTFILTER: 1888 error = accept_filt_getopt(so, sopt); 1889 break; 1890 1891 case SO_LINGER: 1892 l.l_onoff = (so->so_options & SO_LINGER) ? 1 : 0; 1893 l.l_linger = so->so_linger; 1894 1895 error = sockopt_set(sopt, &l, sizeof(l)); 1896 break; 1897 1898 case SO_USELOOPBACK: 1899 case SO_DONTROUTE: 1900 case SO_DEBUG: 1901 case SO_KEEPALIVE: 1902 case SO_REUSEADDR: 1903 case SO_REUSEPORT: 1904 case SO_BROADCAST: 1905 case SO_OOBINLINE: 1906 case SO_TIMESTAMP: 1907 case SO_NOSIGPIPE: 1908 #ifdef SO_OTIMESTAMP 1909 case SO_OTIMESTAMP: 1910 #endif 1911 case SO_ACCEPTCONN: 1912 error = sockopt_setint(sopt, (so->so_options & opt) ? 1 : 0); 1913 break; 1914 1915 case SO_TYPE: 1916 error = sockopt_setint(sopt, so->so_type); 1917 break; 1918 1919 case SO_ERROR: 1920 error = sockopt_setint(sopt, so->so_error); 1921 so->so_error = 0; 1922 break; 1923 1924 case SO_SNDBUF: 1925 error = sockopt_setint(sopt, so->so_snd.sb_hiwat); 1926 break; 1927 1928 case SO_RCVBUF: 1929 error = sockopt_setint(sopt, so->so_rcv.sb_hiwat); 1930 break; 1931 1932 case SO_SNDLOWAT: 1933 error = sockopt_setint(sopt, so->so_snd.sb_lowat); 1934 break; 1935 1936 case SO_RCVLOWAT: 1937 error = sockopt_setint(sopt, so->so_rcv.sb_lowat); 1938 break; 1939 1940 #ifdef COMPAT_50 1941 case SO_OSNDTIMEO: 1942 case SO_ORCVTIMEO: { 1943 struct timeval50 otv; 1944 1945 optval = (opt == SO_OSNDTIMEO ? 1946 so->so_snd.sb_timeo : so->so_rcv.sb_timeo); 1947 1948 otv.tv_sec = optval / hz; 1949 otv.tv_usec = (optval % hz) * tick; 1950 1951 error = sockopt_set(sopt, &otv, sizeof(otv)); 1952 break; 1953 } 1954 #endif /* COMPAT_50 */ 1955 1956 case SO_SNDTIMEO: 1957 case SO_RCVTIMEO: 1958 optval = (opt == SO_SNDTIMEO ? 1959 so->so_snd.sb_timeo : so->so_rcv.sb_timeo); 1960 1961 tv.tv_sec = optval / hz; 1962 tv.tv_usec = (optval % hz) * tick; 1963 1964 error = sockopt_set(sopt, &tv, sizeof(tv)); 1965 break; 1966 1967 case SO_OVERFLOWED: 1968 error = sockopt_setint(sopt, so->so_rcv.sb_overflowed); 1969 break; 1970 1971 default: 1972 error = ENOPROTOOPT; 1973 break; 1974 } 1975 1976 return (error); 1977 } 1978 1979 int 1980 sogetopt(struct socket *so, struct sockopt *sopt) 1981 { 1982 int error; 1983 1984 solock(so); 1985 if (sopt->sopt_level != SOL_SOCKET) { 1986 if (so->so_proto && so->so_proto->pr_ctloutput) { 1987 error = ((*so->so_proto->pr_ctloutput) 1988 (PRCO_GETOPT, so, sopt)); 1989 } else 1990 error = (ENOPROTOOPT); 1991 } else { 1992 error = sogetopt1(so, sopt); 1993 } 1994 sounlock(so); 1995 return (error); 1996 } 1997 1998 /* 1999 * alloc sockopt data buffer buffer 2000 * - will be released at destroy 2001 */ 2002 static int 2003 sockopt_alloc(struct sockopt *sopt, size_t len, km_flag_t kmflag) 2004 { 2005 2006 KASSERT(sopt->sopt_size == 0); 2007 2008 if (len > sizeof(sopt->sopt_buf)) { 2009 sopt->sopt_data = kmem_zalloc(len, kmflag); 2010 if (sopt->sopt_data == NULL) 2011 return ENOMEM; 2012 } else 2013 sopt->sopt_data = sopt->sopt_buf; 2014 2015 sopt->sopt_size = len; 2016 return 0; 2017 } 2018 2019 /* 2020 * initialise sockopt storage 2021 * - MAY sleep during allocation 2022 */ 2023 void 2024 sockopt_init(struct sockopt *sopt, int level, int name, size_t size) 2025 { 2026 2027 memset(sopt, 0, sizeof(*sopt)); 2028 2029 sopt->sopt_level = level; 2030 sopt->sopt_name = name; 2031 (void)sockopt_alloc(sopt, size, KM_SLEEP); 2032 } 2033 2034 /* 2035 * destroy sockopt storage 2036 * - will release any held memory references 2037 */ 2038 void 2039 sockopt_destroy(struct sockopt *sopt) 2040 { 2041 2042 if (sopt->sopt_data != sopt->sopt_buf) 2043 kmem_free(sopt->sopt_data, sopt->sopt_size); 2044 2045 memset(sopt, 0, sizeof(*sopt)); 2046 } 2047 2048 /* 2049 * set sockopt value 2050 * - value is copied into sockopt 2051 * - memory is allocated when necessary, will not sleep 2052 */ 2053 int 2054 sockopt_set(struct sockopt *sopt, const void *buf, size_t len) 2055 { 2056 int error; 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 memcpy(sopt->sopt_data, buf, len); 2066 return 0; 2067 } 2068 2069 /* 2070 * common case of set sockopt integer value 2071 */ 2072 int 2073 sockopt_setint(struct sockopt *sopt, int val) 2074 { 2075 2076 return sockopt_set(sopt, &val, sizeof(int)); 2077 } 2078 2079 /* 2080 * get sockopt value 2081 * - correct size must be given 2082 */ 2083 int 2084 sockopt_get(const struct sockopt *sopt, void *buf, size_t len) 2085 { 2086 2087 if (sopt->sopt_size != len) 2088 return EINVAL; 2089 2090 memcpy(buf, sopt->sopt_data, len); 2091 return 0; 2092 } 2093 2094 /* 2095 * common case of get sockopt integer value 2096 */ 2097 int 2098 sockopt_getint(const struct sockopt *sopt, int *valp) 2099 { 2100 2101 return sockopt_get(sopt, valp, sizeof(int)); 2102 } 2103 2104 /* 2105 * set sockopt value from mbuf 2106 * - ONLY for legacy code 2107 * - mbuf is released by sockopt 2108 * - will not sleep 2109 */ 2110 int 2111 sockopt_setmbuf(struct sockopt *sopt, struct mbuf *m) 2112 { 2113 size_t len; 2114 int error; 2115 2116 len = m_length(m); 2117 2118 if (sopt->sopt_size == 0) { 2119 error = sockopt_alloc(sopt, len, KM_NOSLEEP); 2120 if (error) 2121 return error; 2122 } 2123 2124 KASSERT(sopt->sopt_size == len); 2125 m_copydata(m, 0, len, sopt->sopt_data); 2126 m_freem(m); 2127 2128 return 0; 2129 } 2130 2131 /* 2132 * get sockopt value into mbuf 2133 * - ONLY for legacy code 2134 * - mbuf to be released by the caller 2135 * - will not sleep 2136 */ 2137 struct mbuf * 2138 sockopt_getmbuf(const struct sockopt *sopt) 2139 { 2140 struct mbuf *m; 2141 2142 if (sopt->sopt_size > MCLBYTES) 2143 return NULL; 2144 2145 m = m_get(M_DONTWAIT, MT_SOOPTS); 2146 if (m == NULL) 2147 return NULL; 2148 2149 if (sopt->sopt_size > MLEN) { 2150 MCLGET(m, M_DONTWAIT); 2151 if ((m->m_flags & M_EXT) == 0) { 2152 m_free(m); 2153 return NULL; 2154 } 2155 } 2156 2157 memcpy(mtod(m, void *), sopt->sopt_data, sopt->sopt_size); 2158 m->m_len = sopt->sopt_size; 2159 2160 return m; 2161 } 2162 2163 void 2164 sohasoutofband(struct socket *so) 2165 { 2166 2167 fownsignal(so->so_pgid, SIGURG, POLL_PRI, POLLPRI|POLLRDBAND, so); 2168 selnotify(&so->so_rcv.sb_sel, POLLPRI | POLLRDBAND, NOTE_SUBMIT); 2169 } 2170 2171 static void 2172 filt_sordetach(struct knote *kn) 2173 { 2174 struct socket *so; 2175 2176 so = ((file_t *)kn->kn_obj)->f_data; 2177 solock(so); 2178 SLIST_REMOVE(&so->so_rcv.sb_sel.sel_klist, kn, knote, kn_selnext); 2179 if (SLIST_EMPTY(&so->so_rcv.sb_sel.sel_klist)) 2180 so->so_rcv.sb_flags &= ~SB_KNOTE; 2181 sounlock(so); 2182 } 2183 2184 /*ARGSUSED*/ 2185 static int 2186 filt_soread(struct knote *kn, long hint) 2187 { 2188 struct socket *so; 2189 int rv; 2190 2191 so = ((file_t *)kn->kn_obj)->f_data; 2192 if (hint != NOTE_SUBMIT) 2193 solock(so); 2194 kn->kn_data = so->so_rcv.sb_cc; 2195 if (so->so_state & SS_CANTRCVMORE) { 2196 kn->kn_flags |= EV_EOF; 2197 kn->kn_fflags = so->so_error; 2198 rv = 1; 2199 } else if (so->so_error) /* temporary udp error */ 2200 rv = 1; 2201 else if (kn->kn_sfflags & NOTE_LOWAT) 2202 rv = (kn->kn_data >= kn->kn_sdata); 2203 else 2204 rv = (kn->kn_data >= so->so_rcv.sb_lowat); 2205 if (hint != NOTE_SUBMIT) 2206 sounlock(so); 2207 return rv; 2208 } 2209 2210 static void 2211 filt_sowdetach(struct knote *kn) 2212 { 2213 struct socket *so; 2214 2215 so = ((file_t *)kn->kn_obj)->f_data; 2216 solock(so); 2217 SLIST_REMOVE(&so->so_snd.sb_sel.sel_klist, kn, knote, kn_selnext); 2218 if (SLIST_EMPTY(&so->so_snd.sb_sel.sel_klist)) 2219 so->so_snd.sb_flags &= ~SB_KNOTE; 2220 sounlock(so); 2221 } 2222 2223 /*ARGSUSED*/ 2224 static int 2225 filt_sowrite(struct knote *kn, long hint) 2226 { 2227 struct socket *so; 2228 int rv; 2229 2230 so = ((file_t *)kn->kn_obj)->f_data; 2231 if (hint != NOTE_SUBMIT) 2232 solock(so); 2233 kn->kn_data = sbspace(&so->so_snd); 2234 if (so->so_state & SS_CANTSENDMORE) { 2235 kn->kn_flags |= EV_EOF; 2236 kn->kn_fflags = so->so_error; 2237 rv = 1; 2238 } else if (so->so_error) /* temporary udp error */ 2239 rv = 1; 2240 else if (((so->so_state & SS_ISCONNECTED) == 0) && 2241 (so->so_proto->pr_flags & PR_CONNREQUIRED)) 2242 rv = 0; 2243 else if (kn->kn_sfflags & NOTE_LOWAT) 2244 rv = (kn->kn_data >= kn->kn_sdata); 2245 else 2246 rv = (kn->kn_data >= so->so_snd.sb_lowat); 2247 if (hint != NOTE_SUBMIT) 2248 sounlock(so); 2249 return rv; 2250 } 2251 2252 /*ARGSUSED*/ 2253 static int 2254 filt_solisten(struct knote *kn, long hint) 2255 { 2256 struct socket *so; 2257 int rv; 2258 2259 so = ((file_t *)kn->kn_obj)->f_data; 2260 2261 /* 2262 * Set kn_data to number of incoming connections, not 2263 * counting partial (incomplete) connections. 2264 */ 2265 if (hint != NOTE_SUBMIT) 2266 solock(so); 2267 kn->kn_data = so->so_qlen; 2268 rv = (kn->kn_data > 0); 2269 if (hint != NOTE_SUBMIT) 2270 sounlock(so); 2271 return rv; 2272 } 2273 2274 static const struct filterops solisten_filtops = 2275 { 1, NULL, filt_sordetach, filt_solisten }; 2276 static const struct filterops soread_filtops = 2277 { 1, NULL, filt_sordetach, filt_soread }; 2278 static const struct filterops sowrite_filtops = 2279 { 1, NULL, filt_sowdetach, filt_sowrite }; 2280 2281 int 2282 soo_kqfilter(struct file *fp, struct knote *kn) 2283 { 2284 struct socket *so; 2285 struct sockbuf *sb; 2286 2287 so = ((file_t *)kn->kn_obj)->f_data; 2288 solock(so); 2289 switch (kn->kn_filter) { 2290 case EVFILT_READ: 2291 if (so->so_options & SO_ACCEPTCONN) 2292 kn->kn_fop = &solisten_filtops; 2293 else 2294 kn->kn_fop = &soread_filtops; 2295 sb = &so->so_rcv; 2296 break; 2297 case EVFILT_WRITE: 2298 kn->kn_fop = &sowrite_filtops; 2299 sb = &so->so_snd; 2300 break; 2301 default: 2302 sounlock(so); 2303 return (EINVAL); 2304 } 2305 SLIST_INSERT_HEAD(&sb->sb_sel.sel_klist, kn, kn_selnext); 2306 sb->sb_flags |= SB_KNOTE; 2307 sounlock(so); 2308 return (0); 2309 } 2310 2311 static int 2312 sodopoll(struct socket *so, int events) 2313 { 2314 int revents; 2315 2316 revents = 0; 2317 2318 if (events & (POLLIN | POLLRDNORM)) 2319 if (soreadable(so)) 2320 revents |= events & (POLLIN | POLLRDNORM); 2321 2322 if (events & (POLLOUT | POLLWRNORM)) 2323 if (sowritable(so)) 2324 revents |= events & (POLLOUT | POLLWRNORM); 2325 2326 if (events & (POLLPRI | POLLRDBAND)) 2327 if (so->so_oobmark || (so->so_state & SS_RCVATMARK)) 2328 revents |= events & (POLLPRI | POLLRDBAND); 2329 2330 return revents; 2331 } 2332 2333 int 2334 sopoll(struct socket *so, int events) 2335 { 2336 int revents = 0; 2337 2338 #ifndef DIAGNOSTIC 2339 /* 2340 * Do a quick, unlocked check in expectation that the socket 2341 * will be ready for I/O. Don't do this check if DIAGNOSTIC, 2342 * as the solocked() assertions will fail. 2343 */ 2344 if ((revents = sodopoll(so, events)) != 0) 2345 return revents; 2346 #endif 2347 2348 solock(so); 2349 if ((revents = sodopoll(so, events)) == 0) { 2350 if (events & (POLLIN | POLLPRI | POLLRDNORM | POLLRDBAND)) { 2351 selrecord(curlwp, &so->so_rcv.sb_sel); 2352 so->so_rcv.sb_flags |= SB_NOTIFY; 2353 } 2354 2355 if (events & (POLLOUT | POLLWRNORM)) { 2356 selrecord(curlwp, &so->so_snd.sb_sel); 2357 so->so_snd.sb_flags |= SB_NOTIFY; 2358 } 2359 } 2360 sounlock(so); 2361 2362 return revents; 2363 } 2364 2365 2366 #include <sys/sysctl.h> 2367 2368 static int sysctl_kern_somaxkva(SYSCTLFN_PROTO); 2369 static int sysctl_kern_sbmax(SYSCTLFN_PROTO); 2370 2371 /* 2372 * sysctl helper routine for kern.somaxkva. ensures that the given 2373 * value is not too small. 2374 * (XXX should we maybe make sure it's not too large as well?) 2375 */ 2376 static int 2377 sysctl_kern_somaxkva(SYSCTLFN_ARGS) 2378 { 2379 int error, new_somaxkva; 2380 struct sysctlnode node; 2381 2382 new_somaxkva = somaxkva; 2383 node = *rnode; 2384 node.sysctl_data = &new_somaxkva; 2385 error = sysctl_lookup(SYSCTLFN_CALL(&node)); 2386 if (error || newp == NULL) 2387 return (error); 2388 2389 if (new_somaxkva < (16 * 1024 * 1024)) /* sanity */ 2390 return (EINVAL); 2391 2392 mutex_enter(&so_pendfree_lock); 2393 somaxkva = new_somaxkva; 2394 cv_broadcast(&socurkva_cv); 2395 mutex_exit(&so_pendfree_lock); 2396 2397 return (error); 2398 } 2399 2400 /* 2401 * sysctl helper routine for kern.sbmax. Basically just ensures that 2402 * any new value is not too small. 2403 */ 2404 static int 2405 sysctl_kern_sbmax(SYSCTLFN_ARGS) 2406 { 2407 int error, new_sbmax; 2408 struct sysctlnode node; 2409 2410 new_sbmax = sb_max; 2411 node = *rnode; 2412 node.sysctl_data = &new_sbmax; 2413 error = sysctl_lookup(SYSCTLFN_CALL(&node)); 2414 if (error || newp == NULL) 2415 return (error); 2416 2417 KERNEL_LOCK(1, NULL); 2418 error = sb_max_set(new_sbmax); 2419 KERNEL_UNLOCK_ONE(NULL); 2420 2421 return (error); 2422 } 2423 2424 static void 2425 sysctl_kern_socket_setup(void) 2426 { 2427 2428 KASSERT(socket_sysctllog == NULL); 2429 2430 sysctl_createv(&socket_sysctllog, 0, NULL, NULL, 2431 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 2432 CTLTYPE_INT, "somaxkva", 2433 SYSCTL_DESCR("Maximum amount of kernel memory to be " 2434 "used for socket buffers"), 2435 sysctl_kern_somaxkva, 0, NULL, 0, 2436 CTL_KERN, KERN_SOMAXKVA, CTL_EOL); 2437 2438 sysctl_createv(&socket_sysctllog, 0, NULL, NULL, 2439 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 2440 CTLTYPE_INT, "sbmax", 2441 SYSCTL_DESCR("Maximum socket buffer size"), 2442 sysctl_kern_sbmax, 0, NULL, 0, 2443 CTL_KERN, KERN_SBMAX, CTL_EOL); 2444 } 2445
Indexes created Sat Sep 20 22:09:52 GMT 2025