uipc_socket.c revision 1.83.2.8
1 /* $NetBSD: uipc_socket.c,v 1.83.2.8 2005/03/04 16:52:02 skrll Exp $ */ 2 3 /*- 4 * Copyright (c) 2002 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. 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 * 3. All advertising materials mentioning features or use of this software 19 * must display the following acknowledgement: 20 * This product includes software developed by the NetBSD 21 * Foundation, Inc. and its contributors. 22 * 4. Neither the name of The NetBSD Foundation nor the names of its 23 * contributors may be used to endorse or promote products derived 24 * from this software without specific prior written permission. 25 * 26 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 27 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 28 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 29 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 30 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 31 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 32 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 33 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 34 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 35 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 36 * POSSIBILITY OF SUCH DAMAGE. 37 */ 38 39 /* 40 * Copyright (c) 1982, 1986, 1988, 1990, 1993 41 * The Regents of the University of California. All rights reserved. 42 * 43 * Redistribution and use in source and binary forms, with or without 44 * modification, are permitted provided that the following conditions 45 * are met: 46 * 1. Redistributions of source code must retain the above copyright 47 * notice, this list of conditions and the following disclaimer. 48 * 2. Redistributions in binary form must reproduce the above copyright 49 * notice, this list of conditions and the following disclaimer in the 50 * documentation and/or other materials provided with the distribution. 51 * 3. Neither the name of the University nor the names of its contributors 52 * may be used to endorse or promote products derived from this software 53 * without specific prior written permission. 54 * 55 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 56 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 57 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 58 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 59 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 60 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 61 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 62 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 63 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 64 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 65 * SUCH DAMAGE. 66 * 67 * @(#)uipc_socket.c 8.6 (Berkeley) 5/2/95 68 */ 69 70 #include <sys/cdefs.h> 71 __KERNEL_RCSID(0, "$NetBSD: uipc_socket.c,v 1.83.2.8 2005/03/04 16:52:02 skrll Exp $"); 72 73 #include "opt_sock_counters.h" 74 #include "opt_sosend_loan.h" 75 #include "opt_mbuftrace.h" 76 #include "opt_somaxkva.h" 77 78 #include <sys/param.h> 79 #include <sys/systm.h> 80 #include <sys/proc.h> 81 #include <sys/file.h> 82 #include <sys/malloc.h> 83 #include <sys/mbuf.h> 84 #include <sys/domain.h> 85 #include <sys/kernel.h> 86 #include <sys/protosw.h> 87 #include <sys/socket.h> 88 #include <sys/socketvar.h> 89 #include <sys/signalvar.h> 90 #include <sys/resourcevar.h> 91 #include <sys/pool.h> 92 #include <sys/event.h> 93 #include <sys/poll.h> 94 95 #include <uvm/uvm.h> 96 97 POOL_INIT(socket_pool, sizeof(struct socket), 0, 0, 0, "sockpl", NULL); 98 99 MALLOC_DEFINE(M_SOOPTS, "soopts", "socket options"); 100 MALLOC_DEFINE(M_SONAME, "soname", "socket name"); 101 102 extern int somaxconn; /* patchable (XXX sysctl) */ 103 int somaxconn = SOMAXCONN; 104 105 #ifdef SOSEND_COUNTERS 106 #include <sys/device.h> 107 108 struct evcnt sosend_loan_big = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 109 NULL, "sosend", "loan big"); 110 struct evcnt sosend_copy_big = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 111 NULL, "sosend", "copy big"); 112 struct evcnt sosend_copy_small = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 113 NULL, "sosend", "copy small"); 114 struct evcnt sosend_kvalimit = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, 115 NULL, "sosend", "kva limit"); 116 117 #define SOSEND_COUNTER_INCR(ev) (ev)->ev_count++ 118 119 EVCNT_ATTACH_STATIC(sosend_loan_big); 120 EVCNT_ATTACH_STATIC(sosend_copy_big); 121 EVCNT_ATTACH_STATIC(sosend_copy_small); 122 EVCNT_ATTACH_STATIC(sosend_kvalimit); 123 #else 124 125 #define SOSEND_COUNTER_INCR(ev) /* nothing */ 126 127 #endif /* SOSEND_COUNTERS */ 128 129 void 130 soinit(void) 131 { 132 133 /* Set the initial adjusted socket buffer size. */ 134 if (sb_max_set(sb_max)) 135 panic("bad initial sb_max value: %lu\n", sb_max); 136 137 } 138 139 #ifdef SOSEND_NO_LOAN 140 int use_sosend_loan = 0; 141 #else 142 int use_sosend_loan = 1; 143 #endif 144 145 struct simplelock so_pendfree_slock = SIMPLELOCK_INITIALIZER; 146 struct mbuf *so_pendfree; 147 148 #ifndef SOMAXKVA 149 #define SOMAXKVA (16 * 1024 * 1024) 150 #endif 151 int somaxkva = SOMAXKVA; 152 int socurkva; 153 int sokvawaiters; 154 155 #define SOCK_LOAN_THRESH 4096 156 #define SOCK_LOAN_CHUNK 65536 157 158 static size_t sodopendfree(struct socket *); 159 static size_t sodopendfreel(struct socket *); 160 static __inline vsize_t sokvareserve(struct socket *, vsize_t); 161 static __inline void sokvaunreserve(vsize_t); 162 163 static __inline vsize_t 164 sokvareserve(struct socket *so, vsize_t len) 165 { 166 int s; 167 int error; 168 169 s = splvm(); 170 simple_lock(&so_pendfree_slock); 171 while (socurkva + len > somaxkva) { 172 size_t freed; 173 174 /* 175 * try to do pendfree. 176 */ 177 178 freed = sodopendfreel(so); 179 180 /* 181 * if some kva was freed, try again. 182 */ 183 184 if (freed) 185 continue; 186 187 SOSEND_COUNTER_INCR(&sosend_kvalimit); 188 sokvawaiters++; 189 error = ltsleep(&socurkva, PVM | PCATCH, "sokva", 0, 190 &so_pendfree_slock); 191 sokvawaiters--; 192 if (error) { 193 len = 0; 194 break; 195 } 196 } 197 socurkva += len; 198 simple_unlock(&so_pendfree_slock); 199 splx(s); 200 return len; 201 } 202 203 static __inline void 204 sokvaunreserve(vsize_t len) 205 { 206 int s; 207 208 s = splvm(); 209 simple_lock(&so_pendfree_slock); 210 socurkva -= len; 211 if (sokvawaiters) 212 wakeup(&socurkva); 213 simple_unlock(&so_pendfree_slock); 214 splx(s); 215 } 216 217 /* 218 * sokvaalloc: allocate kva for loan. 219 */ 220 221 vaddr_t 222 sokvaalloc(vsize_t len, struct socket *so) 223 { 224 vaddr_t lva; 225 226 /* 227 * reserve kva. 228 */ 229 230 if (sokvareserve(so, len) == 0) 231 return 0; 232 233 /* 234 * allocate kva. 235 */ 236 237 lva = uvm_km_valloc_wait(kernel_map, len); 238 if (lva == 0) { 239 sokvaunreserve(len); 240 return (0); 241 } 242 243 return lva; 244 } 245 246 /* 247 * sokvafree: free kva for loan. 248 */ 249 250 void 251 sokvafree(vaddr_t sva, vsize_t len) 252 { 253 254 /* 255 * free kva. 256 */ 257 258 uvm_km_free(kernel_map, sva, len); 259 260 /* 261 * unreserve kva. 262 */ 263 264 sokvaunreserve(len); 265 } 266 267 static void 268 sodoloanfree(struct vm_page **pgs, caddr_t buf, size_t size) 269 { 270 vaddr_t va, sva, eva; 271 vsize_t len; 272 paddr_t pa; 273 int i, npgs; 274 275 eva = round_page((vaddr_t) buf + size); 276 sva = trunc_page((vaddr_t) buf); 277 len = eva - sva; 278 npgs = len >> PAGE_SHIFT; 279 280 if (__predict_false(pgs == NULL)) { 281 pgs = alloca(npgs * sizeof(*pgs)); 282 283 for (i = 0, va = sva; va < eva; i++, va += PAGE_SIZE) { 284 if (pmap_extract(pmap_kernel(), va, &pa) == FALSE) 285 panic("sodoloanfree: va 0x%lx not mapped", va); 286 pgs[i] = PHYS_TO_VM_PAGE(pa); 287 } 288 } 289 290 pmap_kremove(sva, len); 291 pmap_update(pmap_kernel()); 292 uvm_unloan(pgs, npgs, UVM_LOAN_TOPAGE); 293 sokvafree(sva, len); 294 } 295 296 static size_t 297 sodopendfree(struct socket *so) 298 { 299 int s; 300 size_t rv; 301 302 s = splvm(); 303 simple_lock(&so_pendfree_slock); 304 rv = sodopendfreel(so); 305 simple_unlock(&so_pendfree_slock); 306 splx(s); 307 308 return rv; 309 } 310 311 /* 312 * sodopendfreel: free mbufs on "pendfree" list. 313 * unlock and relock so_pendfree_slock when freeing mbufs. 314 * 315 * => called with so_pendfree_slock held. 316 * => called at splvm. 317 */ 318 319 static size_t 320 sodopendfreel(struct socket *so) 321 { 322 size_t rv = 0; 323 324 LOCK_ASSERT(simple_lock_held(&so_pendfree_slock)); 325 326 for (;;) { 327 struct mbuf *m; 328 struct mbuf *next; 329 330 m = so_pendfree; 331 if (m == NULL) 332 break; 333 so_pendfree = NULL; 334 simple_unlock(&so_pendfree_slock); 335 /* XXX splx */ 336 337 for (; m != NULL; m = next) { 338 next = m->m_next; 339 340 rv += m->m_ext.ext_size; 341 sodoloanfree((m->m_flags & M_EXT_PAGES) ? 342 m->m_ext.ext_pgs : NULL, m->m_ext.ext_buf, 343 m->m_ext.ext_size); 344 pool_cache_put(&mbpool_cache, m); 345 } 346 347 /* XXX splvm */ 348 simple_lock(&so_pendfree_slock); 349 } 350 351 return (rv); 352 } 353 354 void 355 soloanfree(struct mbuf *m, caddr_t buf, size_t size, void *arg) 356 { 357 int s; 358 359 if (m == NULL) { 360 361 /* 362 * called from MEXTREMOVE. 363 */ 364 365 sodoloanfree(NULL, buf, size); 366 return; 367 } 368 369 /* 370 * postpone freeing mbuf. 371 * 372 * we can't do it in interrupt context 373 * because we need to put kva back to kernel_map. 374 */ 375 376 s = splvm(); 377 simple_lock(&so_pendfree_slock); 378 m->m_next = so_pendfree; 379 so_pendfree = m; 380 if (sokvawaiters) 381 wakeup(&socurkva); 382 simple_unlock(&so_pendfree_slock); 383 splx(s); 384 } 385 386 static long 387 sosend_loan(struct socket *so, struct uio *uio, struct mbuf *m, long space) 388 { 389 struct iovec *iov = uio->uio_iov; 390 vaddr_t sva, eva; 391 vsize_t len; 392 vaddr_t lva, va; 393 int npgs, i, error; 394 395 if (uio->uio_segflg != UIO_USERSPACE) 396 return (0); 397 398 if (iov->iov_len < (size_t) space) 399 space = iov->iov_len; 400 if (space > SOCK_LOAN_CHUNK) 401 space = SOCK_LOAN_CHUNK; 402 403 eva = round_page((vaddr_t) iov->iov_base + space); 404 sva = trunc_page((vaddr_t) iov->iov_base); 405 len = eva - sva; 406 npgs = len >> PAGE_SHIFT; 407 408 /* XXX KDASSERT */ 409 KASSERT(npgs <= M_EXT_MAXPAGES); 410 KASSERT(uio->uio_lwp != NULL); 411 412 lva = sokvaalloc(len, so); 413 if (lva == 0) 414 return 0; 415 416 error = uvm_loan(&uio->uio_lwp->l_proc->p_vmspace->vm_map, sva, len, 417 m->m_ext.ext_pgs, UVM_LOAN_TOPAGE); 418 if (error) { 419 sokvafree(lva, len); 420 return (0); 421 } 422 423 for (i = 0, va = lva; i < npgs; i++, va += PAGE_SIZE) 424 pmap_kenter_pa(va, VM_PAGE_TO_PHYS(m->m_ext.ext_pgs[i]), 425 VM_PROT_READ); 426 pmap_update(pmap_kernel()); 427 428 lva += (vaddr_t) iov->iov_base & PAGE_MASK; 429 430 MEXTADD(m, (caddr_t) lva, space, M_MBUF, soloanfree, so); 431 m->m_flags |= M_EXT_PAGES | M_EXT_ROMAP; 432 433 uio->uio_resid -= space; 434 /* uio_offset not updated, not set/used for write(2) */ 435 uio->uio_iov->iov_base = (caddr_t) uio->uio_iov->iov_base + space; 436 uio->uio_iov->iov_len -= space; 437 if (uio->uio_iov->iov_len == 0) { 438 uio->uio_iov++; 439 uio->uio_iovcnt--; 440 } 441 442 return (space); 443 } 444 445 /* 446 * Socket operation routines. 447 * These routines are called by the routines in 448 * sys_socket.c or from a system process, and 449 * implement the semantics of socket operations by 450 * switching out to the protocol specific routines. 451 */ 452 /*ARGSUSED*/ 453 int 454 socreate(int dom, struct socket **aso, int type, int proto, struct lwp *l) 455 { 456 const struct protosw *prp; 457 struct socket *so; 458 struct proc *p; 459 int error, s; 460 461 p = l->l_proc; 462 if (proto) 463 prp = pffindproto(dom, proto, type); 464 else 465 prp = pffindtype(dom, type); 466 if (prp == 0 || prp->pr_usrreq == 0) 467 return (EPROTONOSUPPORT); 468 if (prp->pr_type != type) 469 return (EPROTOTYPE); 470 s = splsoftnet(); 471 so = pool_get(&socket_pool, PR_WAITOK); 472 memset((caddr_t)so, 0, sizeof(*so)); 473 TAILQ_INIT(&so->so_q0); 474 TAILQ_INIT(&so->so_q); 475 so->so_type = type; 476 so->so_proto = prp; 477 so->so_send = sosend; 478 so->so_receive = soreceive; 479 #ifdef MBUFTRACE 480 so->so_rcv.sb_mowner = &prp->pr_domain->dom_mowner; 481 so->so_snd.sb_mowner = &prp->pr_domain->dom_mowner; 482 so->so_mowner = &prp->pr_domain->dom_mowner; 483 #endif 484 if (p != 0) 485 so->so_uid = p->p_ucred->cr_uid; 486 else 487 so->so_uid = UID_MAX; 488 error = (*prp->pr_usrreq)(so, PRU_ATTACH, (struct mbuf *)0, 489 (struct mbuf *)(long)proto, (struct mbuf *)0, l); 490 if (error) { 491 so->so_state |= SS_NOFDREF; 492 sofree(so); 493 splx(s); 494 return (error); 495 } 496 splx(s); 497 *aso = so; 498 return (0); 499 } 500 501 int 502 sobind(struct socket *so, struct mbuf *nam, struct lwp *l) 503 { 504 int s, error; 505 506 s = splsoftnet(); 507 error = (*so->so_proto->pr_usrreq)(so, PRU_BIND, (struct mbuf *)0, 508 nam, (struct mbuf *)0, l); 509 splx(s); 510 return (error); 511 } 512 513 int 514 solisten(struct socket *so, int backlog) 515 { 516 int s, error; 517 518 s = splsoftnet(); 519 error = (*so->so_proto->pr_usrreq)(so, PRU_LISTEN, (struct mbuf *)0, 520 (struct mbuf *)0, (struct mbuf *)0, (struct lwp *)0); 521 if (error) { 522 splx(s); 523 return (error); 524 } 525 if (TAILQ_EMPTY(&so->so_q)) 526 so->so_options |= SO_ACCEPTCONN; 527 if (backlog < 0) 528 backlog = 0; 529 so->so_qlimit = min(backlog, somaxconn); 530 splx(s); 531 return (0); 532 } 533 534 void 535 sofree(struct socket *so) 536 { 537 538 if (so->so_pcb || (so->so_state & SS_NOFDREF) == 0) 539 return; 540 if (so->so_head) { 541 /* 542 * We must not decommission a socket that's on the accept(2) 543 * queue. If we do, then accept(2) may hang after select(2) 544 * indicated that the listening socket was ready. 545 */ 546 if (!soqremque(so, 0)) 547 return; 548 } 549 if (so->so_rcv.sb_hiwat) 550 (void)chgsbsize(so->so_uid, &so->so_rcv.sb_hiwat, 0, 551 RLIM_INFINITY); 552 if (so->so_snd.sb_hiwat) 553 (void)chgsbsize(so->so_uid, &so->so_snd.sb_hiwat, 0, 554 RLIM_INFINITY); 555 sbrelease(&so->so_snd, so); 556 sorflush(so); 557 pool_put(&socket_pool, so); 558 } 559 560 /* 561 * Close a socket on last file table reference removal. 562 * Initiate disconnect if connected. 563 * Free socket when disconnect complete. 564 */ 565 int 566 soclose(struct socket *so) 567 { 568 struct socket *so2; 569 int s, error; 570 571 error = 0; 572 s = splsoftnet(); /* conservative */ 573 if (so->so_options & SO_ACCEPTCONN) { 574 while ((so2 = TAILQ_FIRST(&so->so_q0)) != 0) { 575 (void) soqremque(so2, 0); 576 (void) soabort(so2); 577 } 578 while ((so2 = TAILQ_FIRST(&so->so_q)) != 0) { 579 (void) soqremque(so2, 1); 580 (void) soabort(so2); 581 } 582 } 583 if (so->so_pcb == 0) 584 goto discard; 585 if (so->so_state & SS_ISCONNECTED) { 586 if ((so->so_state & SS_ISDISCONNECTING) == 0) { 587 error = sodisconnect(so); 588 if (error) 589 goto drop; 590 } 591 if (so->so_options & SO_LINGER) { 592 if ((so->so_state & SS_ISDISCONNECTING) && 593 (so->so_state & SS_NBIO)) 594 goto drop; 595 while (so->so_state & SS_ISCONNECTED) { 596 error = tsleep((caddr_t)&so->so_timeo, 597 PSOCK | PCATCH, netcls, 598 so->so_linger * hz); 599 if (error) 600 break; 601 } 602 } 603 } 604 drop: 605 if (so->so_pcb) { 606 int error2 = (*so->so_proto->pr_usrreq)(so, PRU_DETACH, 607 (struct mbuf *)0, (struct mbuf *)0, (struct mbuf *)0, 608 (struct lwp *)0); 609 if (error == 0) 610 error = error2; 611 } 612 discard: 613 if (so->so_state & SS_NOFDREF) 614 panic("soclose: NOFDREF"); 615 so->so_state |= SS_NOFDREF; 616 sofree(so); 617 splx(s); 618 return (error); 619 } 620 621 /* 622 * Must be called at splsoftnet... 623 */ 624 int 625 soabort(struct socket *so) 626 { 627 628 return (*so->so_proto->pr_usrreq)(so, PRU_ABORT, (struct mbuf *)0, 629 (struct mbuf *)0, (struct mbuf *)0, (struct lwp *)0); 630 } 631 632 int 633 soaccept(struct socket *so, struct mbuf *nam) 634 { 635 int s, error; 636 637 error = 0; 638 s = splsoftnet(); 639 if ((so->so_state & SS_NOFDREF) == 0) 640 panic("soaccept: !NOFDREF"); 641 so->so_state &= ~SS_NOFDREF; 642 if ((so->so_state & SS_ISDISCONNECTED) == 0 || 643 (so->so_proto->pr_flags & PR_ABRTACPTDIS) == 0) 644 error = (*so->so_proto->pr_usrreq)(so, PRU_ACCEPT, 645 (struct mbuf *)0, nam, (struct mbuf *)0, (struct lwp *)0); 646 else 647 error = ECONNABORTED; 648 649 splx(s); 650 return (error); 651 } 652 653 int 654 soconnect(struct socket *so, struct mbuf *nam, struct lwp *l) 655 { 656 int s, error; 657 658 if (so->so_options & SO_ACCEPTCONN) 659 return (EOPNOTSUPP); 660 s = splsoftnet(); 661 /* 662 * If protocol is connection-based, can only connect once. 663 * Otherwise, if connected, try to disconnect first. 664 * This allows user to disconnect by connecting to, e.g., 665 * a null address. 666 */ 667 if (so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING) && 668 ((so->so_proto->pr_flags & PR_CONNREQUIRED) || 669 (error = sodisconnect(so)))) 670 error = EISCONN; 671 else 672 error = (*so->so_proto->pr_usrreq)(so, PRU_CONNECT, 673 (struct mbuf *)0, nam, (struct mbuf *)0, l); 674 splx(s); 675 return (error); 676 } 677 678 int 679 soconnect2(struct socket *so1, struct socket *so2) 680 { 681 int s, error; 682 683 s = splsoftnet(); 684 error = (*so1->so_proto->pr_usrreq)(so1, PRU_CONNECT2, 685 (struct mbuf *)0, (struct mbuf *)so2, (struct mbuf *)0, 686 (struct lwp *)0); 687 splx(s); 688 return (error); 689 } 690 691 int 692 sodisconnect(struct socket *so) 693 { 694 int s, error; 695 696 s = splsoftnet(); 697 if ((so->so_state & SS_ISCONNECTED) == 0) { 698 error = ENOTCONN; 699 goto bad; 700 } 701 if (so->so_state & SS_ISDISCONNECTING) { 702 error = EALREADY; 703 goto bad; 704 } 705 error = (*so->so_proto->pr_usrreq)(so, PRU_DISCONNECT, 706 (struct mbuf *)0, (struct mbuf *)0, (struct mbuf *)0, 707 (struct lwp *)0); 708 bad: 709 splx(s); 710 sodopendfree(so); 711 return (error); 712 } 713 714 #define SBLOCKWAIT(f) (((f) & MSG_DONTWAIT) ? M_NOWAIT : M_WAITOK) 715 /* 716 * Send on a socket. 717 * If send must go all at once and message is larger than 718 * send buffering, then hard error. 719 * Lock against other senders. 720 * If must go all at once and not enough room now, then 721 * inform user that this would block and do nothing. 722 * Otherwise, if nonblocking, send as much as possible. 723 * The data to be sent is described by "uio" if nonzero, 724 * otherwise by the mbuf chain "top" (which must be null 725 * if uio is not). Data provided in mbuf chain must be small 726 * enough to send all at once. 727 * 728 * Returns nonzero on error, timeout or signal; callers 729 * must check for short counts if EINTR/ERESTART are returned. 730 * Data and control buffers are freed on return. 731 */ 732 int 733 sosend(struct socket *so, struct mbuf *addr, struct uio *uio, struct mbuf *top, 734 struct mbuf *control, int flags, struct lwp *l) 735 { 736 struct mbuf **mp, *m; 737 struct proc *p; 738 long space, len, resid, clen, mlen; 739 int error, s, dontroute, atomic; 740 741 p = l->l_proc; 742 sodopendfree(so); 743 744 clen = 0; 745 atomic = sosendallatonce(so) || top; 746 if (uio) 747 resid = uio->uio_resid; 748 else 749 resid = top->m_pkthdr.len; 750 /* 751 * In theory resid should be unsigned. 752 * However, space must be signed, as it might be less than 0 753 * if we over-committed, and we must use a signed comparison 754 * of space and resid. On the other hand, a negative resid 755 * causes us to loop sending 0-length segments to the protocol. 756 */ 757 if (resid < 0) { 758 error = EINVAL; 759 goto out; 760 } 761 dontroute = 762 (flags & MSG_DONTROUTE) && (so->so_options & SO_DONTROUTE) == 0 && 763 (so->so_proto->pr_flags & PR_ATOMIC); 764 if (p) 765 p->p_stats->p_ru.ru_msgsnd++; 766 if (control) 767 clen = control->m_len; 768 #define snderr(errno) { error = errno; splx(s); goto release; } 769 770 restart: 771 if ((error = sblock(&so->so_snd, SBLOCKWAIT(flags))) != 0) 772 goto out; 773 do { 774 s = splsoftnet(); 775 if (so->so_state & SS_CANTSENDMORE) 776 snderr(EPIPE); 777 if (so->so_error) { 778 error = so->so_error; 779 so->so_error = 0; 780 splx(s); 781 goto release; 782 } 783 if ((so->so_state & SS_ISCONNECTED) == 0) { 784 if (so->so_proto->pr_flags & PR_CONNREQUIRED) { 785 if ((so->so_state & SS_ISCONFIRMING) == 0 && 786 !(resid == 0 && clen != 0)) 787 snderr(ENOTCONN); 788 } else if (addr == 0) 789 snderr(EDESTADDRREQ); 790 } 791 space = sbspace(&so->so_snd); 792 if (flags & MSG_OOB) 793 space += 1024; 794 if ((atomic && resid > so->so_snd.sb_hiwat) || 795 clen > so->so_snd.sb_hiwat) 796 snderr(EMSGSIZE); 797 if (space < resid + clen && 798 (atomic || space < so->so_snd.sb_lowat || space < clen)) { 799 if (so->so_state & SS_NBIO) 800 snderr(EWOULDBLOCK); 801 sbunlock(&so->so_snd); 802 error = sbwait(&so->so_snd); 803 splx(s); 804 if (error) 805 goto out; 806 goto restart; 807 } 808 splx(s); 809 mp = ⊤ 810 space -= clen; 811 do { 812 if (uio == NULL) { 813 /* 814 * Data is prepackaged in "top". 815 */ 816 resid = 0; 817 if (flags & MSG_EOR) 818 top->m_flags |= M_EOR; 819 } else do { 820 if (top == 0) { 821 m = m_gethdr(M_WAIT, MT_DATA); 822 mlen = MHLEN; 823 m->m_pkthdr.len = 0; 824 m->m_pkthdr.rcvif = (struct ifnet *)0; 825 } else { 826 m = m_get(M_WAIT, MT_DATA); 827 mlen = MLEN; 828 } 829 MCLAIM(m, so->so_snd.sb_mowner); 830 if (use_sosend_loan && 831 uio->uio_iov->iov_len >= SOCK_LOAN_THRESH && 832 space >= SOCK_LOAN_THRESH && 833 (len = sosend_loan(so, uio, m, 834 space)) != 0) { 835 SOSEND_COUNTER_INCR(&sosend_loan_big); 836 space -= len; 837 goto have_data; 838 } 839 if (resid >= MINCLSIZE && space >= MCLBYTES) { 840 SOSEND_COUNTER_INCR(&sosend_copy_big); 841 m_clget(m, M_WAIT); 842 if ((m->m_flags & M_EXT) == 0) 843 goto nopages; 844 mlen = MCLBYTES; 845 if (atomic && top == 0) { 846 len = lmin(MCLBYTES - max_hdr, 847 resid); 848 m->m_data += max_hdr; 849 } else 850 len = lmin(MCLBYTES, resid); 851 space -= len; 852 } else { 853 nopages: 854 SOSEND_COUNTER_INCR(&sosend_copy_small); 855 len = lmin(lmin(mlen, resid), space); 856 space -= len; 857 /* 858 * For datagram protocols, leave room 859 * for protocol headers in first mbuf. 860 */ 861 if (atomic && top == 0 && len < mlen) 862 MH_ALIGN(m, len); 863 } 864 error = uiomove(mtod(m, caddr_t), (int)len, 865 uio); 866 have_data: 867 resid = uio->uio_resid; 868 m->m_len = len; 869 *mp = m; 870 top->m_pkthdr.len += len; 871 if (error) 872 goto release; 873 mp = &m->m_next; 874 if (resid <= 0) { 875 if (flags & MSG_EOR) 876 top->m_flags |= M_EOR; 877 break; 878 } 879 } while (space > 0 && atomic); 880 881 s = splsoftnet(); 882 883 if (so->so_state & SS_CANTSENDMORE) 884 snderr(EPIPE); 885 886 if (dontroute) 887 so->so_options |= SO_DONTROUTE; 888 if (resid > 0) 889 so->so_state |= SS_MORETOCOME; 890 error = (*so->so_proto->pr_usrreq)(so, 891 (flags & MSG_OOB) ? PRU_SENDOOB : PRU_SEND, 892 top, addr, control, curlwp); /* XXX */ 893 if (dontroute) 894 so->so_options &= ~SO_DONTROUTE; 895 if (resid > 0) 896 so->so_state &= ~SS_MORETOCOME; 897 splx(s); 898 899 clen = 0; 900 control = 0; 901 top = 0; 902 mp = ⊤ 903 if (error) 904 goto release; 905 } while (resid && space > 0); 906 } while (resid); 907 908 release: 909 sbunlock(&so->so_snd); 910 out: 911 if (top) 912 m_freem(top); 913 if (control) 914 m_freem(control); 915 return (error); 916 } 917 918 /* 919 * Implement receive operations on a socket. 920 * We depend on the way that records are added to the sockbuf 921 * by sbappend*. In particular, each record (mbufs linked through m_next) 922 * must begin with an address if the protocol so specifies, 923 * followed by an optional mbuf or mbufs containing ancillary data, 924 * and then zero or more mbufs of data. 925 * In order to avoid blocking network interrupts for the entire time here, 926 * we splx() while doing the actual copy to user space. 927 * Although the sockbuf is locked, new data may still be appended, 928 * and thus we must maintain consistency of the sockbuf during that time. 929 * 930 * The caller may receive the data as a single mbuf chain by supplying 931 * an mbuf **mp0 for use in returning the chain. The uio is then used 932 * only for the count in uio_resid. 933 */ 934 int 935 soreceive(struct socket *so, struct mbuf **paddr, struct uio *uio, 936 struct mbuf **mp0, struct mbuf **controlp, int *flagsp) 937 { 938 struct lwp *l; 939 struct mbuf *m, **mp; 940 int flags, len, error, s, offset, moff, type, orig_resid; 941 const struct protosw *pr; 942 struct mbuf *nextrecord; 943 int mbuf_removed = 0; 944 945 pr = so->so_proto; 946 mp = mp0; 947 type = 0; 948 orig_resid = uio->uio_resid; 949 l = uio->uio_lwp; 950 951 if (paddr) 952 *paddr = 0; 953 if (controlp) 954 *controlp = 0; 955 if (flagsp) 956 flags = *flagsp &~ MSG_EOR; 957 else 958 flags = 0; 959 960 if ((flags & MSG_DONTWAIT) == 0) 961 sodopendfree(so); 962 963 if (flags & MSG_OOB) { 964 m = m_get(M_WAIT, MT_DATA); 965 error = (*pr->pr_usrreq)(so, PRU_RCVOOB, m, 966 (struct mbuf *)(long)(flags & MSG_PEEK), 967 (struct mbuf *)0, l); 968 if (error) 969 goto bad; 970 do { 971 error = uiomove(mtod(m, caddr_t), 972 (int) min(uio->uio_resid, m->m_len), uio); 973 m = m_free(m); 974 } while (uio->uio_resid && error == 0 && m); 975 bad: 976 if (m) 977 m_freem(m); 978 return (error); 979 } 980 if (mp) 981 *mp = (struct mbuf *)0; 982 if (so->so_state & SS_ISCONFIRMING && uio->uio_resid) 983 (*pr->pr_usrreq)(so, PRU_RCVD, (struct mbuf *)0, 984 (struct mbuf *)0, (struct mbuf *)0, l); 985 986 restart: 987 if ((error = sblock(&so->so_rcv, SBLOCKWAIT(flags))) != 0) 988 return (error); 989 s = splsoftnet(); 990 991 m = so->so_rcv.sb_mb; 992 /* 993 * If we have less data than requested, block awaiting more 994 * (subject to any timeout) if: 995 * 1. the current count is less than the low water mark, 996 * 2. MSG_WAITALL is set, and it is possible to do the entire 997 * receive operation at once if we block (resid <= hiwat), or 998 * 3. MSG_DONTWAIT is not set. 999 * If MSG_WAITALL is set but resid is larger than the receive buffer, 1000 * we have to do the receive in sections, and thus risk returning 1001 * a short count if a timeout or signal occurs after we start. 1002 */ 1003 if (m == 0 || (((flags & MSG_DONTWAIT) == 0 && 1004 so->so_rcv.sb_cc < uio->uio_resid) && 1005 (so->so_rcv.sb_cc < so->so_rcv.sb_lowat || 1006 ((flags & MSG_WAITALL) && uio->uio_resid <= so->so_rcv.sb_hiwat)) && 1007 m->m_nextpkt == 0 && (pr->pr_flags & PR_ATOMIC) == 0)) { 1008 #ifdef DIAGNOSTIC 1009 if (m == 0 && so->so_rcv.sb_cc) 1010 panic("receive 1"); 1011 #endif 1012 if (so->so_error) { 1013 if (m) 1014 goto dontblock; 1015 error = so->so_error; 1016 if ((flags & MSG_PEEK) == 0) 1017 so->so_error = 0; 1018 goto release; 1019 } 1020 if (so->so_state & SS_CANTRCVMORE) { 1021 if (m) 1022 goto dontblock; 1023 else 1024 goto release; 1025 } 1026 for (; m; m = m->m_next) 1027 if (m->m_type == MT_OOBDATA || (m->m_flags & M_EOR)) { 1028 m = so->so_rcv.sb_mb; 1029 goto dontblock; 1030 } 1031 if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 && 1032 (so->so_proto->pr_flags & PR_CONNREQUIRED)) { 1033 error = ENOTCONN; 1034 goto release; 1035 } 1036 if (uio->uio_resid == 0) 1037 goto release; 1038 if ((so->so_state & SS_NBIO) || (flags & MSG_DONTWAIT)) { 1039 error = EWOULDBLOCK; 1040 goto release; 1041 } 1042 SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 1"); 1043 SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 1"); 1044 sbunlock(&so->so_rcv); 1045 error = sbwait(&so->so_rcv); 1046 splx(s); 1047 if (error) 1048 return (error); 1049 goto restart; 1050 } 1051 dontblock: 1052 /* 1053 * On entry here, m points to the first record of the socket buffer. 1054 * While we process the initial mbufs containing address and control 1055 * info, we save a copy of m->m_nextpkt into nextrecord. 1056 */ 1057 if (l) 1058 l->l_proc->p_stats->p_ru.ru_msgrcv++; 1059 KASSERT(m == so->so_rcv.sb_mb); 1060 SBLASTRECORDCHK(&so->so_rcv, "soreceive 1"); 1061 SBLASTMBUFCHK(&so->so_rcv, "soreceive 1"); 1062 nextrecord = m->m_nextpkt; 1063 if (pr->pr_flags & PR_ADDR) { 1064 #ifdef DIAGNOSTIC 1065 if (m->m_type != MT_SONAME) 1066 panic("receive 1a"); 1067 #endif 1068 orig_resid = 0; 1069 if (flags & MSG_PEEK) { 1070 if (paddr) 1071 *paddr = m_copy(m, 0, m->m_len); 1072 m = m->m_next; 1073 } else { 1074 sbfree(&so->so_rcv, m); 1075 mbuf_removed = 1; 1076 if (paddr) { 1077 *paddr = m; 1078 so->so_rcv.sb_mb = m->m_next; 1079 m->m_next = 0; 1080 m = so->so_rcv.sb_mb; 1081 } else { 1082 MFREE(m, so->so_rcv.sb_mb); 1083 m = so->so_rcv.sb_mb; 1084 } 1085 } 1086 } 1087 while (m && m->m_type == MT_CONTROL && error == 0) { 1088 if (flags & MSG_PEEK) { 1089 if (controlp) 1090 *controlp = m_copy(m, 0, m->m_len); 1091 m = m->m_next; 1092 } else { 1093 sbfree(&so->so_rcv, m); 1094 mbuf_removed = 1; 1095 if (controlp) { 1096 struct domain *dom = pr->pr_domain; 1097 if (dom->dom_externalize && l && 1098 mtod(m, struct cmsghdr *)->cmsg_type == 1099 SCM_RIGHTS) 1100 error = (*dom->dom_externalize)(m, l); 1101 *controlp = m; 1102 so->so_rcv.sb_mb = m->m_next; 1103 m->m_next = 0; 1104 m = so->so_rcv.sb_mb; 1105 } else { 1106 /* 1107 * Dispose of any SCM_RIGHTS message that went 1108 * through the read path rather than recv. 1109 */ 1110 if (pr->pr_domain->dom_dispose && 1111 mtod(m, struct cmsghdr *)->cmsg_type == SCM_RIGHTS) 1112 (*pr->pr_domain->dom_dispose)(m); 1113 MFREE(m, so->so_rcv.sb_mb); 1114 m = so->so_rcv.sb_mb; 1115 } 1116 } 1117 if (controlp) { 1118 orig_resid = 0; 1119 controlp = &(*controlp)->m_next; 1120 } 1121 } 1122 1123 /* 1124 * If m is non-NULL, we have some data to read. From now on, 1125 * make sure to keep sb_lastrecord consistent when working on 1126 * the last packet on the chain (nextrecord == NULL) and we 1127 * change m->m_nextpkt. 1128 */ 1129 if (m) { 1130 if ((flags & MSG_PEEK) == 0) { 1131 m->m_nextpkt = nextrecord; 1132 /* 1133 * If nextrecord == NULL (this is a single chain), 1134 * then sb_lastrecord may not be valid here if m 1135 * was changed earlier. 1136 */ 1137 if (nextrecord == NULL) { 1138 KASSERT(so->so_rcv.sb_mb == m); 1139 so->so_rcv.sb_lastrecord = m; 1140 } 1141 } 1142 type = m->m_type; 1143 if (type == MT_OOBDATA) 1144 flags |= MSG_OOB; 1145 } else { 1146 if ((flags & MSG_PEEK) == 0) { 1147 KASSERT(so->so_rcv.sb_mb == m); 1148 so->so_rcv.sb_mb = nextrecord; 1149 SB_EMPTY_FIXUP(&so->so_rcv); 1150 } 1151 } 1152 SBLASTRECORDCHK(&so->so_rcv, "soreceive 2"); 1153 SBLASTMBUFCHK(&so->so_rcv, "soreceive 2"); 1154 1155 moff = 0; 1156 offset = 0; 1157 while (m && uio->uio_resid > 0 && error == 0) { 1158 if (m->m_type == MT_OOBDATA) { 1159 if (type != MT_OOBDATA) 1160 break; 1161 } else if (type == MT_OOBDATA) 1162 break; 1163 #ifdef DIAGNOSTIC 1164 else if (m->m_type != MT_DATA && m->m_type != MT_HEADER) 1165 panic("receive 3"); 1166 #endif 1167 so->so_state &= ~SS_RCVATMARK; 1168 len = uio->uio_resid; 1169 if (so->so_oobmark && len > so->so_oobmark - offset) 1170 len = so->so_oobmark - offset; 1171 if (len > m->m_len - moff) 1172 len = m->m_len - moff; 1173 /* 1174 * If mp is set, just pass back the mbufs. 1175 * Otherwise copy them out via the uio, then free. 1176 * Sockbuf must be consistent here (points to current mbuf, 1177 * it points to next record) when we drop priority; 1178 * we must note any additions to the sockbuf when we 1179 * block interrupts again. 1180 */ 1181 if (mp == 0) { 1182 SBLASTRECORDCHK(&so->so_rcv, "soreceive uiomove"); 1183 SBLASTMBUFCHK(&so->so_rcv, "soreceive uiomove"); 1184 splx(s); 1185 error = uiomove(mtod(m, caddr_t) + moff, (int)len, uio); 1186 s = splsoftnet(); 1187 if (error) { 1188 /* 1189 * If any part of the record has been removed 1190 * (such as the MT_SONAME mbuf, which will 1191 * happen when PR_ADDR, and thus also 1192 * PR_ATOMIC, is set), then drop the entire 1193 * record to maintain the atomicity of the 1194 * receive operation. 1195 * 1196 * This avoids a later panic("receive 1a") 1197 * when compiled with DIAGNOSTIC. 1198 */ 1199 if (m && mbuf_removed 1200 && (pr->pr_flags & PR_ATOMIC)) 1201 (void) sbdroprecord(&so->so_rcv); 1202 1203 goto release; 1204 } 1205 } else 1206 uio->uio_resid -= len; 1207 if (len == m->m_len - moff) { 1208 if (m->m_flags & M_EOR) 1209 flags |= MSG_EOR; 1210 if (flags & MSG_PEEK) { 1211 m = m->m_next; 1212 moff = 0; 1213 } else { 1214 nextrecord = m->m_nextpkt; 1215 sbfree(&so->so_rcv, m); 1216 if (mp) { 1217 *mp = m; 1218 mp = &m->m_next; 1219 so->so_rcv.sb_mb = m = m->m_next; 1220 *mp = (struct mbuf *)0; 1221 } else { 1222 MFREE(m, so->so_rcv.sb_mb); 1223 m = so->so_rcv.sb_mb; 1224 } 1225 /* 1226 * If m != NULL, we also know that 1227 * so->so_rcv.sb_mb != NULL. 1228 */ 1229 KASSERT(so->so_rcv.sb_mb == m); 1230 if (m) { 1231 m->m_nextpkt = nextrecord; 1232 if (nextrecord == NULL) 1233 so->so_rcv.sb_lastrecord = m; 1234 } else { 1235 so->so_rcv.sb_mb = nextrecord; 1236 SB_EMPTY_FIXUP(&so->so_rcv); 1237 } 1238 SBLASTRECORDCHK(&so->so_rcv, "soreceive 3"); 1239 SBLASTMBUFCHK(&so->so_rcv, "soreceive 3"); 1240 } 1241 } else { 1242 if (flags & MSG_PEEK) 1243 moff += len; 1244 else { 1245 if (mp) 1246 *mp = m_copym(m, 0, len, M_WAIT); 1247 m->m_data += len; 1248 m->m_len -= len; 1249 so->so_rcv.sb_cc -= len; 1250 } 1251 } 1252 if (so->so_oobmark) { 1253 if ((flags & MSG_PEEK) == 0) { 1254 so->so_oobmark -= len; 1255 if (so->so_oobmark == 0) { 1256 so->so_state |= SS_RCVATMARK; 1257 break; 1258 } 1259 } else { 1260 offset += len; 1261 if (offset == so->so_oobmark) 1262 break; 1263 } 1264 } 1265 if (flags & MSG_EOR) 1266 break; 1267 /* 1268 * If the MSG_WAITALL flag is set (for non-atomic socket), 1269 * we must not quit until "uio->uio_resid == 0" or an error 1270 * termination. If a signal/timeout occurs, return 1271 * with a short count but without error. 1272 * Keep sockbuf locked against other readers. 1273 */ 1274 while (flags & MSG_WAITALL && m == 0 && uio->uio_resid > 0 && 1275 !sosendallatonce(so) && !nextrecord) { 1276 if (so->so_error || so->so_state & SS_CANTRCVMORE) 1277 break; 1278 /* 1279 * If we are peeking and the socket receive buffer is 1280 * full, stop since we can't get more data to peek at. 1281 */ 1282 if ((flags & MSG_PEEK) && sbspace(&so->so_rcv) <= 0) 1283 break; 1284 /* 1285 * If we've drained the socket buffer, tell the 1286 * protocol in case it needs to do something to 1287 * get it filled again. 1288 */ 1289 if ((pr->pr_flags & PR_WANTRCVD) && so->so_pcb) 1290 (*pr->pr_usrreq)(so, PRU_RCVD, 1291 (struct mbuf *)0, 1292 (struct mbuf *)(long)flags, 1293 (struct mbuf *)0, l); 1294 SBLASTRECORDCHK(&so->so_rcv, "soreceive sbwait 2"); 1295 SBLASTMBUFCHK(&so->so_rcv, "soreceive sbwait 2"); 1296 error = sbwait(&so->so_rcv); 1297 if (error) { 1298 sbunlock(&so->so_rcv); 1299 splx(s); 1300 return (0); 1301 } 1302 if ((m = so->so_rcv.sb_mb) != NULL) 1303 nextrecord = m->m_nextpkt; 1304 } 1305 } 1306 1307 if (m && pr->pr_flags & PR_ATOMIC) { 1308 flags |= MSG_TRUNC; 1309 if ((flags & MSG_PEEK) == 0) 1310 (void) sbdroprecord(&so->so_rcv); 1311 } 1312 if ((flags & MSG_PEEK) == 0) { 1313 if (m == 0) { 1314 /* 1315 * First part is an inline SB_EMPTY_FIXUP(). Second 1316 * part makes sure sb_lastrecord is up-to-date if 1317 * there is still data in the socket buffer. 1318 */ 1319 so->so_rcv.sb_mb = nextrecord; 1320 if (so->so_rcv.sb_mb == NULL) { 1321 so->so_rcv.sb_mbtail = NULL; 1322 so->so_rcv.sb_lastrecord = NULL; 1323 } else if (nextrecord->m_nextpkt == NULL) 1324 so->so_rcv.sb_lastrecord = nextrecord; 1325 } 1326 SBLASTRECORDCHK(&so->so_rcv, "soreceive 4"); 1327 SBLASTMBUFCHK(&so->so_rcv, "soreceive 4"); 1328 if (pr->pr_flags & PR_WANTRCVD && so->so_pcb) 1329 (*pr->pr_usrreq)(so, PRU_RCVD, (struct mbuf *)0, 1330 (struct mbuf *)(long)flags, (struct mbuf *)0, l); 1331 } 1332 if (orig_resid == uio->uio_resid && orig_resid && 1333 (flags & MSG_EOR) == 0 && (so->so_state & SS_CANTRCVMORE) == 0) { 1334 sbunlock(&so->so_rcv); 1335 splx(s); 1336 goto restart; 1337 } 1338 1339 if (flagsp) 1340 *flagsp |= flags; 1341 release: 1342 sbunlock(&so->so_rcv); 1343 splx(s); 1344 return (error); 1345 } 1346 1347 int 1348 soshutdown(struct socket *so, int how) 1349 { 1350 const struct protosw *pr; 1351 1352 pr = so->so_proto; 1353 if (!(how == SHUT_RD || how == SHUT_WR || how == SHUT_RDWR)) 1354 return (EINVAL); 1355 1356 if (how == SHUT_RD || how == SHUT_RDWR) 1357 sorflush(so); 1358 if (how == SHUT_WR || how == SHUT_RDWR) 1359 return (*pr->pr_usrreq)(so, PRU_SHUTDOWN, (struct mbuf *)0, 1360 (struct mbuf *)0, (struct mbuf *)0, (struct lwp *)0); 1361 return (0); 1362 } 1363 1364 void 1365 sorflush(struct socket *so) 1366 { 1367 struct sockbuf *sb, asb; 1368 const struct protosw *pr; 1369 int s; 1370 1371 sb = &so->so_rcv; 1372 pr = so->so_proto; 1373 sb->sb_flags |= SB_NOINTR; 1374 (void) sblock(sb, M_WAITOK); 1375 s = splnet(); 1376 socantrcvmore(so); 1377 sbunlock(sb); 1378 asb = *sb; 1379 /* 1380 * Clear most of the sockbuf structure, but leave some of the 1381 * fields valid. 1382 */ 1383 memset(&sb->sb_startzero, 0, 1384 sizeof(*sb) - offsetof(struct sockbuf, sb_startzero)); 1385 splx(s); 1386 if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose) 1387 (*pr->pr_domain->dom_dispose)(asb.sb_mb); 1388 sbrelease(&asb, so); 1389 } 1390 1391 int 1392 sosetopt(struct socket *so, int level, int optname, struct mbuf *m0) 1393 { 1394 int error; 1395 struct mbuf *m; 1396 1397 error = 0; 1398 m = m0; 1399 if (level != SOL_SOCKET) { 1400 if (so->so_proto && so->so_proto->pr_ctloutput) 1401 return ((*so->so_proto->pr_ctloutput) 1402 (PRCO_SETOPT, so, level, optname, &m0)); 1403 error = ENOPROTOOPT; 1404 } else { 1405 switch (optname) { 1406 1407 case SO_LINGER: 1408 if (m == NULL || m->m_len != sizeof(struct linger)) { 1409 error = EINVAL; 1410 goto bad; 1411 } 1412 so->so_linger = mtod(m, struct linger *)->l_linger; 1413 /* fall thru... */ 1414 1415 case SO_DEBUG: 1416 case SO_KEEPALIVE: 1417 case SO_DONTROUTE: 1418 case SO_USELOOPBACK: 1419 case SO_BROADCAST: 1420 case SO_REUSEADDR: 1421 case SO_REUSEPORT: 1422 case SO_OOBINLINE: 1423 case SO_TIMESTAMP: 1424 if (m == NULL || m->m_len < sizeof(int)) { 1425 error = EINVAL; 1426 goto bad; 1427 } 1428 if (*mtod(m, int *)) 1429 so->so_options |= optname; 1430 else 1431 so->so_options &= ~optname; 1432 break; 1433 1434 case SO_SNDBUF: 1435 case SO_RCVBUF: 1436 case SO_SNDLOWAT: 1437 case SO_RCVLOWAT: 1438 { 1439 int optval; 1440 1441 if (m == NULL || m->m_len < sizeof(int)) { 1442 error = EINVAL; 1443 goto bad; 1444 } 1445 1446 /* 1447 * Values < 1 make no sense for any of these 1448 * options, so disallow them. 1449 */ 1450 optval = *mtod(m, int *); 1451 if (optval < 1) { 1452 error = EINVAL; 1453 goto bad; 1454 } 1455 1456 switch (optname) { 1457 1458 case SO_SNDBUF: 1459 case SO_RCVBUF: 1460 if (sbreserve(optname == SO_SNDBUF ? 1461 &so->so_snd : &so->so_rcv, 1462 (u_long) optval, so) == 0) { 1463 error = ENOBUFS; 1464 goto bad; 1465 } 1466 break; 1467 1468 /* 1469 * Make sure the low-water is never greater than 1470 * the high-water. 1471 */ 1472 case SO_SNDLOWAT: 1473 so->so_snd.sb_lowat = 1474 (optval > so->so_snd.sb_hiwat) ? 1475 so->so_snd.sb_hiwat : optval; 1476 break; 1477 case SO_RCVLOWAT: 1478 so->so_rcv.sb_lowat = 1479 (optval > so->so_rcv.sb_hiwat) ? 1480 so->so_rcv.sb_hiwat : optval; 1481 break; 1482 } 1483 break; 1484 } 1485 1486 case SO_SNDTIMEO: 1487 case SO_RCVTIMEO: 1488 { 1489 struct timeval *tv; 1490 int val; 1491 1492 if (m == NULL || m->m_len < sizeof(*tv)) { 1493 error = EINVAL; 1494 goto bad; 1495 } 1496 tv = mtod(m, struct timeval *); 1497 if (tv->tv_sec > (INT_MAX - tv->tv_usec / tick) / hz) { 1498 error = EDOM; 1499 goto bad; 1500 } 1501 val = tv->tv_sec * hz + tv->tv_usec / tick; 1502 if (val == 0 && tv->tv_usec != 0) 1503 val = 1; 1504 1505 switch (optname) { 1506 1507 case SO_SNDTIMEO: 1508 so->so_snd.sb_timeo = val; 1509 break; 1510 case SO_RCVTIMEO: 1511 so->so_rcv.sb_timeo = val; 1512 break; 1513 } 1514 break; 1515 } 1516 1517 default: 1518 error = ENOPROTOOPT; 1519 break; 1520 } 1521 if (error == 0 && so->so_proto && so->so_proto->pr_ctloutput) { 1522 (void) ((*so->so_proto->pr_ctloutput) 1523 (PRCO_SETOPT, so, level, optname, &m0)); 1524 m = NULL; /* freed by protocol */ 1525 } 1526 } 1527 bad: 1528 if (m) 1529 (void) m_free(m); 1530 return (error); 1531 } 1532 1533 int 1534 sogetopt(struct socket *so, int level, int optname, struct mbuf **mp) 1535 { 1536 struct mbuf *m; 1537 1538 if (level != SOL_SOCKET) { 1539 if (so->so_proto && so->so_proto->pr_ctloutput) { 1540 return ((*so->so_proto->pr_ctloutput) 1541 (PRCO_GETOPT, so, level, optname, mp)); 1542 } else 1543 return (ENOPROTOOPT); 1544 } else { 1545 m = m_get(M_WAIT, MT_SOOPTS); 1546 m->m_len = sizeof(int); 1547 1548 switch (optname) { 1549 1550 case SO_LINGER: 1551 m->m_len = sizeof(struct linger); 1552 mtod(m, struct linger *)->l_onoff = 1553 so->so_options & SO_LINGER; 1554 mtod(m, struct linger *)->l_linger = so->so_linger; 1555 break; 1556 1557 case SO_USELOOPBACK: 1558 case SO_DONTROUTE: 1559 case SO_DEBUG: 1560 case SO_KEEPALIVE: 1561 case SO_REUSEADDR: 1562 case SO_REUSEPORT: 1563 case SO_BROADCAST: 1564 case SO_OOBINLINE: 1565 case SO_TIMESTAMP: 1566 *mtod(m, int *) = so->so_options & optname; 1567 break; 1568 1569 case SO_TYPE: 1570 *mtod(m, int *) = so->so_type; 1571 break; 1572 1573 case SO_ERROR: 1574 *mtod(m, int *) = so->so_error; 1575 so->so_error = 0; 1576 break; 1577 1578 case SO_SNDBUF: 1579 *mtod(m, int *) = so->so_snd.sb_hiwat; 1580 break; 1581 1582 case SO_RCVBUF: 1583 *mtod(m, int *) = so->so_rcv.sb_hiwat; 1584 break; 1585 1586 case SO_SNDLOWAT: 1587 *mtod(m, int *) = so->so_snd.sb_lowat; 1588 break; 1589 1590 case SO_RCVLOWAT: 1591 *mtod(m, int *) = so->so_rcv.sb_lowat; 1592 break; 1593 1594 case SO_SNDTIMEO: 1595 case SO_RCVTIMEO: 1596 { 1597 int val = (optname == SO_SNDTIMEO ? 1598 so->so_snd.sb_timeo : so->so_rcv.sb_timeo); 1599 1600 m->m_len = sizeof(struct timeval); 1601 mtod(m, struct timeval *)->tv_sec = val / hz; 1602 mtod(m, struct timeval *)->tv_usec = 1603 (val % hz) * tick; 1604 break; 1605 } 1606 1607 case SO_OVERFLOWED: 1608 *mtod(m, int *) = so->so_rcv.sb_overflowed; 1609 break; 1610 1611 default: 1612 (void)m_free(m); 1613 return (ENOPROTOOPT); 1614 } 1615 *mp = m; 1616 return (0); 1617 } 1618 } 1619 1620 void 1621 sohasoutofband(struct socket *so) 1622 { 1623 fownsignal(so->so_pgid, SIGURG, POLL_PRI, POLLPRI|POLLRDBAND, so); 1624 selwakeup(&so->so_rcv.sb_sel); 1625 } 1626 1627 static void 1628 filt_sordetach(struct knote *kn) 1629 { 1630 struct socket *so; 1631 1632 so = (struct socket *)kn->kn_fp->f_data; 1633 SLIST_REMOVE(&so->so_rcv.sb_sel.sel_klist, kn, knote, kn_selnext); 1634 if (SLIST_EMPTY(&so->so_rcv.sb_sel.sel_klist)) 1635 so->so_rcv.sb_flags &= ~SB_KNOTE; 1636 } 1637 1638 /*ARGSUSED*/ 1639 static int 1640 filt_soread(struct knote *kn, long hint) 1641 { 1642 struct socket *so; 1643 1644 so = (struct socket *)kn->kn_fp->f_data; 1645 kn->kn_data = so->so_rcv.sb_cc; 1646 if (so->so_state & SS_CANTRCVMORE) { 1647 kn->kn_flags |= EV_EOF; 1648 kn->kn_fflags = so->so_error; 1649 return (1); 1650 } 1651 if (so->so_error) /* temporary udp error */ 1652 return (1); 1653 if (kn->kn_sfflags & NOTE_LOWAT) 1654 return (kn->kn_data >= kn->kn_sdata); 1655 return (kn->kn_data >= so->so_rcv.sb_lowat); 1656 } 1657 1658 static void 1659 filt_sowdetach(struct knote *kn) 1660 { 1661 struct socket *so; 1662 1663 so = (struct socket *)kn->kn_fp->f_data; 1664 SLIST_REMOVE(&so->so_snd.sb_sel.sel_klist, kn, knote, kn_selnext); 1665 if (SLIST_EMPTY(&so->so_snd.sb_sel.sel_klist)) 1666 so->so_snd.sb_flags &= ~SB_KNOTE; 1667 } 1668 1669 /*ARGSUSED*/ 1670 static int 1671 filt_sowrite(struct knote *kn, long hint) 1672 { 1673 struct socket *so; 1674 1675 so = (struct socket *)kn->kn_fp->f_data; 1676 kn->kn_data = sbspace(&so->so_snd); 1677 if (so->so_state & SS_CANTSENDMORE) { 1678 kn->kn_flags |= EV_EOF; 1679 kn->kn_fflags = so->so_error; 1680 return (1); 1681 } 1682 if (so->so_error) /* temporary udp error */ 1683 return (1); 1684 if (((so->so_state & SS_ISCONNECTED) == 0) && 1685 (so->so_proto->pr_flags & PR_CONNREQUIRED)) 1686 return (0); 1687 if (kn->kn_sfflags & NOTE_LOWAT) 1688 return (kn->kn_data >= kn->kn_sdata); 1689 return (kn->kn_data >= so->so_snd.sb_lowat); 1690 } 1691 1692 /*ARGSUSED*/ 1693 static int 1694 filt_solisten(struct knote *kn, long hint) 1695 { 1696 struct socket *so; 1697 1698 so = (struct socket *)kn->kn_fp->f_data; 1699 1700 /* 1701 * Set kn_data to number of incoming connections, not 1702 * counting partial (incomplete) connections. 1703 */ 1704 kn->kn_data = so->so_qlen; 1705 return (kn->kn_data > 0); 1706 } 1707 1708 static const struct filterops solisten_filtops = 1709 { 1, NULL, filt_sordetach, filt_solisten }; 1710 static const struct filterops soread_filtops = 1711 { 1, NULL, filt_sordetach, filt_soread }; 1712 static const struct filterops sowrite_filtops = 1713 { 1, NULL, filt_sowdetach, filt_sowrite }; 1714 1715 int 1716 soo_kqfilter(struct file *fp, struct knote *kn) 1717 { 1718 struct socket *so; 1719 struct sockbuf *sb; 1720 1721 so = (struct socket *)kn->kn_fp->f_data; 1722 switch (kn->kn_filter) { 1723 case EVFILT_READ: 1724 if (so->so_options & SO_ACCEPTCONN) 1725 kn->kn_fop = &solisten_filtops; 1726 else 1727 kn->kn_fop = &soread_filtops; 1728 sb = &so->so_rcv; 1729 break; 1730 case EVFILT_WRITE: 1731 kn->kn_fop = &sowrite_filtops; 1732 sb = &so->so_snd; 1733 break; 1734 default: 1735 return (1); 1736 } 1737 SLIST_INSERT_HEAD(&sb->sb_sel.sel_klist, kn, kn_selnext); 1738 sb->sb_flags |= SB_KNOTE; 1739 return (0); 1740 } 1741 1742 #include <sys/sysctl.h> 1743 1744 static int sysctl_kern_somaxkva(SYSCTLFN_PROTO); 1745 1746 /* 1747 * sysctl helper routine for kern.somaxkva. ensures that the given 1748 * value is not too small. 1749 * (XXX should we maybe make sure it's not too large as well?) 1750 */ 1751 static int 1752 sysctl_kern_somaxkva(SYSCTLFN_ARGS) 1753 { 1754 int error, new_somaxkva; 1755 struct sysctlnode node; 1756 int s; 1757 1758 new_somaxkva = somaxkva; 1759 node = *rnode; 1760 node.sysctl_data = &new_somaxkva; 1761 error = sysctl_lookup(SYSCTLFN_CALL(&node)); 1762 if (error || newp == NULL) 1763 return (error); 1764 1765 if (new_somaxkva < (16 * 1024 * 1024)) /* sanity */ 1766 return (EINVAL); 1767 1768 s = splvm(); 1769 simple_lock(&so_pendfree_slock); 1770 somaxkva = new_somaxkva; 1771 wakeup(&socurkva); 1772 simple_unlock(&so_pendfree_slock); 1773 splx(s); 1774 1775 return (error); 1776 } 1777 1778 SYSCTL_SETUP(sysctl_kern_somaxkva_setup, "sysctl kern.somaxkva setup") 1779 { 1780 1781 sysctl_createv(clog, 0, NULL, NULL, 1782 CTLFLAG_PERMANENT, 1783 CTLTYPE_NODE, "kern", NULL, 1784 NULL, 0, NULL, 0, 1785 CTL_KERN, CTL_EOL); 1786 1787 sysctl_createv(clog, 0, NULL, NULL, 1788 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 1789 CTLTYPE_INT, "somaxkva", 1790 SYSCTL_DESCR("Maximum amount of kernel memory to be " 1791 "used for socket buffers"), 1792 sysctl_kern_somaxkva, 0, NULL, 0, 1793 CTL_KERN, KERN_SOMAXKVA, CTL_EOL); 1794 } 1795
Indexes created Sat Sep 20 22:09:52 GMT 2025