1 /* $NetBSD: kern_fork.c,v 1.123 2005/12/11 12:24:29 christos Exp $ */ 2 3 /*- 4 * Copyright (c) 1999, 2001, 2004 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 the Numerical Aerospace Simulation Facility, 9 * NASA Ames Research Center. 10 * This code is derived from software contributed to The NetBSD Foundation 11 * by Charles M. Hannum. 12 * 13 * Redistribution and use in source and binary forms, with or without 14 * modification, are permitted provided that the following conditions 15 * are met: 16 * 1. Redistributions of source code must retain the above copyright 17 * notice, this list of conditions and the following disclaimer. 18 * 2. Redistributions in binary form must reproduce the above copyright 19 * notice, this list of conditions and the following disclaimer in the 20 * documentation and/or other materials provided with the distribution. 21 * 3. All advertising materials mentioning features or use of this software 22 * must display the following acknowledgement: 23 * This product includes software developed by the NetBSD 24 * Foundation, Inc. and its contributors. 25 * 4. Neither the name of The NetBSD Foundation nor the names of its 26 * contributors may be used to endorse or promote products derived 27 * from this software without specific prior written permission. 28 * 29 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 30 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 31 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 32 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 33 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 34 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 35 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 36 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 37 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 38 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 39 * POSSIBILITY OF SUCH DAMAGE. 40 */ 41 42 /* 43 * Copyright (c) 1982, 1986, 1989, 1991, 1993 44 * The Regents of the University of California. All rights reserved. 45 * (c) UNIX System Laboratories, Inc. 46 * All or some portions of this file are derived from material licensed 47 * to the University of California by American Telephone and Telegraph 48 * Co. or Unix System Laboratories, Inc. and are reproduced herein with 49 * the permission of UNIX System Laboratories, Inc. 50 * 51 * Redistribution and use in source and binary forms, with or without 52 * modification, are permitted provided that the following conditions 53 * are met: 54 * 1. Redistributions of source code must retain the above copyright 55 * notice, this list of conditions and the following disclaimer. 56 * 2. Redistributions in binary form must reproduce the above copyright 57 * notice, this list of conditions and the following disclaimer in the 58 * documentation and/or other materials provided with the distribution. 59 * 3. Neither the name of the University nor the names of its contributors 60 * may be used to endorse or promote products derived from this software 61 * without specific prior written permission. 62 * 63 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 64 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 65 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 66 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 67 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 68 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 69 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 70 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 71 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 72 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 73 * SUCH DAMAGE. 74 * 75 * @(#)kern_fork.c 8.8 (Berkeley) 2/14/95 76 */ 77 78 #include <sys/cdefs.h> 79 __KERNEL_RCSID(0, "$NetBSD: kern_fork.c,v 1.123 2005/12/11 12:24:29 christos Exp $"); 80 81 #include "opt_ktrace.h" 82 #include "opt_systrace.h" 83 #include "opt_multiprocessor.h" 84 85 #include <sys/param.h> 86 #include <sys/systm.h> 87 #include <sys/filedesc.h> 88 #include <sys/kernel.h> 89 #include <sys/malloc.h> 90 #include <sys/pool.h> 91 #include <sys/mount.h> 92 #include <sys/proc.h> 93 #include <sys/ras.h> 94 #include <sys/resourcevar.h> 95 #include <sys/vnode.h> 96 #include <sys/file.h> 97 #include <sys/acct.h> 98 #include <sys/ktrace.h> 99 #include <sys/vmmeter.h> 100 #include <sys/sched.h> 101 #include <sys/signalvar.h> 102 #include <sys/systrace.h> 103 104 #include <sys/sa.h> 105 #include <sys/syscallargs.h> 106 107 #include <uvm/uvm_extern.h> 108 109 110 int nprocs = 1; /* process 0 */ 111 112 /* 113 * Number of ticks to sleep if fork() would fail due to process hitting 114 * limits. Exported in miliseconds to userland via sysctl. 115 */ 116 int forkfsleep = 0; 117 118 /*ARGSUSED*/ 119 int 120 sys_fork(struct lwp *l, void *v, register_t *retval) 121 { 122 123 return (fork1(l, 0, SIGCHLD, NULL, 0, NULL, NULL, retval, NULL)); 124 } 125 126 /* 127 * vfork(2) system call compatible with 4.4BSD (i.e. BSD with Mach VM). 128 * Address space is not shared, but parent is blocked until child exit. 129 */ 130 /*ARGSUSED*/ 131 int 132 sys_vfork(struct lwp *l, void *v, register_t *retval) 133 { 134 135 return (fork1(l, FORK_PPWAIT, SIGCHLD, NULL, 0, NULL, NULL, 136 retval, NULL)); 137 } 138 139 /* 140 * New vfork(2) system call for NetBSD, which implements original 3BSD vfork(2) 141 * semantics. Address space is shared, and parent is blocked until child exit. 142 */ 143 /*ARGSUSED*/ 144 int 145 sys___vfork14(struct lwp *l, void *v, register_t *retval) 146 { 147 148 return (fork1(l, FORK_PPWAIT|FORK_SHAREVM, SIGCHLD, NULL, 0, 149 NULL, NULL, retval, NULL)); 150 } 151 152 /* 153 * Linux-compatible __clone(2) system call. 154 */ 155 int 156 sys___clone(struct lwp *l, void *v, register_t *retval) 157 { 158 struct sys___clone_args /* { 159 syscallarg(int) flags; 160 syscallarg(void *) stack; 161 } */ *uap = v; 162 int flags, sig; 163 164 /* 165 * We don't support the CLONE_PID or CLONE_PTRACE flags. 166 */ 167 if (SCARG(uap, flags) & (CLONE_PID|CLONE_PTRACE)) 168 return (EINVAL); 169 170 /* 171 * Linux enforces CLONE_VM with CLONE_SIGHAND, do same. 172 */ 173 if (SCARG(uap, flags) & CLONE_SIGHAND 174 && (SCARG(uap, flags) & CLONE_VM) == 0) 175 return (EINVAL); 176 177 flags = 0; 178 179 if (SCARG(uap, flags) & CLONE_VM) 180 flags |= FORK_SHAREVM; 181 if (SCARG(uap, flags) & CLONE_FS) 182 flags |= FORK_SHARECWD; 183 if (SCARG(uap, flags) & CLONE_FILES) 184 flags |= FORK_SHAREFILES; 185 if (SCARG(uap, flags) & CLONE_SIGHAND) 186 flags |= FORK_SHARESIGS; 187 if (SCARG(uap, flags) & CLONE_VFORK) 188 flags |= FORK_PPWAIT; 189 190 sig = SCARG(uap, flags) & CLONE_CSIGNAL; 191 if (sig < 0 || sig >= _NSIG) 192 return (EINVAL); 193 194 /* 195 * Note that the Linux API does not provide a portable way of 196 * specifying the stack area; the caller must know if the stack 197 * grows up or down. So, we pass a stack size of 0, so that the 198 * code that makes this adjustment is a noop. 199 */ 200 return (fork1(l, flags, sig, SCARG(uap, stack), 0, 201 NULL, NULL, retval, NULL)); 202 } 203 204 /* print the 'table full' message once per 10 seconds */ 205 struct timeval fork_tfmrate = { 10, 0 }; 206 207 int 208 fork1(struct lwp *l1, int flags, int exitsig, void *stack, size_t stacksize, 209 void (*func)(void *), void *arg, register_t *retval, 210 struct proc **rnewprocp) 211 { 212 struct proc *p1, *p2, *parent; 213 uid_t uid; 214 struct lwp *l2; 215 int count, s; 216 vaddr_t uaddr; 217 boolean_t inmem; 218 219 /* 220 * Although process entries are dynamically created, we still keep 221 * a global limit on the maximum number we will create. Don't allow 222 * a nonprivileged user to use the last few processes; don't let root 223 * exceed the limit. The variable nprocs is the current number of 224 * processes, maxproc is the limit. 225 */ 226 p1 = l1->l_proc; 227 uid = p1->p_cred->p_ruid; 228 if (__predict_false((nprocs >= maxproc - 5 && uid != 0) || 229 nprocs >= maxproc)) { 230 static struct timeval lasttfm; 231 232 if (ratecheck(&lasttfm, &fork_tfmrate)) 233 tablefull("proc", "increase kern.maxproc or NPROC"); 234 if (forkfsleep) 235 (void)tsleep(&nprocs, PUSER, "forkmx", forkfsleep); 236 return (EAGAIN); 237 } 238 nprocs++; 239 240 /* 241 * Increment the count of procs running with this uid. Don't allow 242 * a nonprivileged user to exceed their current limit. 243 */ 244 count = chgproccnt(uid, 1); 245 if (__predict_false(uid != 0 && count > 246 p1->p_rlimit[RLIMIT_NPROC].rlim_cur)) { 247 (void)chgproccnt(uid, -1); 248 nprocs--; 249 if (forkfsleep) 250 (void)tsleep(&nprocs, PUSER, "forkulim", forkfsleep); 251 return (EAGAIN); 252 } 253 254 /* 255 * Allocate virtual address space for the U-area now, while it 256 * is still easy to abort the fork operation if we're out of 257 * kernel virtual address space. The actual U-area pages will 258 * be allocated and wired in uvm_fork() if needed. 259 */ 260 261 inmem = uvm_uarea_alloc(&uaddr); 262 if (__predict_false(uaddr == 0)) { 263 (void)chgproccnt(uid, -1); 264 nprocs--; 265 return (ENOMEM); 266 } 267 268 /* 269 * We are now committed to the fork. From here on, we may 270 * block on resources, but resource allocation may NOT fail. 271 */ 272 273 /* Allocate new proc. */ 274 p2 = proc_alloc(); 275 276 /* 277 * Make a proc table entry for the new process. 278 * Start by zeroing the section of proc that is zero-initialized, 279 * then copy the section that is copied directly from the parent. 280 */ 281 memset(&p2->p_startzero, 0, 282 (unsigned) ((caddr_t)&p2->p_endzero - (caddr_t)&p2->p_startzero)); 283 memcpy(&p2->p_startcopy, &p1->p_startcopy, 284 (unsigned) ((caddr_t)&p2->p_endcopy - (caddr_t)&p2->p_startcopy)); 285 286 simple_lock_init(&p2->p_sigctx.ps_silock); 287 CIRCLEQ_INIT(&p2->p_sigctx.ps_siginfo); 288 simple_lock_init(&p2->p_lock); 289 LIST_INIT(&p2->p_lwps); 290 291 /* 292 * Duplicate sub-structures as needed. 293 * Increase reference counts on shared objects. 294 * The p_stats and p_sigacts substructs are set in uvm_fork(). 295 * Inherit flags we want to keep. The flags related to SIGCHLD 296 * handling are important in order to keep a consistent behaviour 297 * for the child after the fork. 298 */ 299 p2->p_flag = p1->p_flag & (P_SUGID | P_STOPFORK | P_STOPEXEC | 300 P_NOCLDSTOP | P_NOCLDWAIT | P_CLDSIGIGN); 301 p2->p_emul = p1->p_emul; 302 p2->p_execsw = p1->p_execsw; 303 304 if (p1->p_flag & P_PROFIL) 305 startprofclock(p2); 306 p2->p_cred = pool_get(&pcred_pool, PR_WAITOK); 307 memcpy(p2->p_cred, p1->p_cred, sizeof(*p2->p_cred)); 308 p2->p_cred->p_refcnt = 1; 309 crhold(p1->p_ucred); 310 311 LIST_INIT(&p2->p_raslist); 312 #if defined(__HAVE_RAS) 313 ras_fork(p1, p2); 314 #endif 315 316 /* bump references to the text vnode (for procfs) */ 317 p2->p_textvp = p1->p_textvp; 318 if (p2->p_textvp) 319 VREF(p2->p_textvp); 320 321 if (flags & FORK_SHAREFILES) 322 fdshare(p1, p2); 323 else if (flags & FORK_CLEANFILES) 324 p2->p_fd = fdinit(p1); 325 else 326 p2->p_fd = fdcopy(p1); 327 328 if (flags & FORK_SHARECWD) 329 cwdshare(p1, p2); 330 else 331 p2->p_cwdi = cwdinit(p1); 332 333 /* 334 * If p_limit is still copy-on-write, bump refcnt, 335 * otherwise get a copy that won't be modified. 336 * (If PL_SHAREMOD is clear, the structure is shared 337 * copy-on-write.) 338 */ 339 if (p1->p_limit->p_lflags & PL_SHAREMOD) 340 p2->p_limit = limcopy(p1->p_limit); 341 else { 342 simple_lock(&p1->p_limit->p_slock); 343 p1->p_limit->p_refcnt++; 344 simple_unlock(&p1->p_limit->p_slock); 345 p2->p_limit = p1->p_limit; 346 } 347 348 if (p1->p_session->s_ttyvp != NULL && p1->p_flag & P_CONTROLT) 349 p2->p_flag |= P_CONTROLT; 350 if (flags & FORK_PPWAIT) 351 p2->p_flag |= P_PPWAIT; 352 parent = (flags & FORK_NOWAIT) ? initproc : p1; 353 p2->p_pptr = parent; 354 LIST_INIT(&p2->p_children); 355 356 s = proclist_lock_write(); 357 LIST_INSERT_AFTER(p1, p2, p_pglist); 358 LIST_INSERT_HEAD(&parent->p_children, p2, p_sibling); 359 proclist_unlock_write(s); 360 361 #ifdef KTRACE 362 /* 363 * Copy traceflag and tracefile if enabled. 364 * If not inherited, these were zeroed above. 365 */ 366 if (p1->p_traceflag & KTRFAC_INHERIT) { 367 p2->p_traceflag = p1->p_traceflag; 368 if ((p2->p_tracep = p1->p_tracep) != NULL) 369 ktradref(p2); 370 } 371 #endif 372 373 scheduler_fork_hook(p1, p2); 374 375 /* 376 * Create signal actions for the child process. 377 */ 378 sigactsinit(p2, p1, flags & FORK_SHARESIGS); 379 380 /* 381 * p_stats. 382 * Copy parts of p_stats, and zero out the rest. 383 */ 384 p2->p_stats = pstatscopy(p1->p_stats); 385 386 /* 387 * If emulation has process fork hook, call it now. 388 */ 389 if (p2->p_emul->e_proc_fork) 390 (*p2->p_emul->e_proc_fork)(p2, p1, flags); 391 392 /* 393 * ...and finally, any other random fork hooks that subsystems 394 * might have registered. 395 */ 396 doforkhooks(p2, p1); 397 398 /* 399 * This begins the section where we must prevent the parent 400 * from being swapped. 401 */ 402 PHOLD(l1); 403 404 uvm_proc_fork(p1, p2, (flags & FORK_SHAREVM) ? TRUE : FALSE); 405 406 /* 407 * Finish creating the child process. 408 * It will return through a different path later. 409 */ 410 newlwp(l1, p2, uaddr, inmem, 0, stack, stacksize, 411 (func != NULL) ? func : child_return, 412 arg, &l2); 413 414 /* Now safe for scheduler to see child process */ 415 s = proclist_lock_write(); 416 p2->p_exitsig = exitsig; /* signal for parent on exit */ 417 LIST_INSERT_HEAD(&allproc, p2, p_list); 418 proclist_unlock_write(s); 419 420 #ifdef SYSTRACE 421 /* Tell systrace what's happening. */ 422 if (ISSET(p1->p_flag, P_SYSTRACE)) 423 systrace_sys_fork(p1, p2); 424 #endif 425 426 #ifdef __HAVE_SYSCALL_INTERN 427 (*p2->p_emul->e_syscall_intern)(p2); 428 #endif 429 430 /* 431 * Make child runnable, set start time, and add to run queue 432 * except if the parent requested the child to start in SSTOP state. 433 */ 434 SCHED_LOCK(s); 435 p2->p_stats->p_start = time; 436 p2->p_acflag = AFORK; 437 if (p1->p_flag & P_STOPFORK) { 438 p2->p_nrlwps = 0; 439 p1->p_nstopchild++; 440 p2->p_stat = SSTOP; 441 l2->l_stat = LSSTOP; 442 } else { 443 p2->p_nrlwps = 1; 444 p2->p_stat = SACTIVE; 445 l2->l_stat = LSRUN; 446 setrunqueue(l2); 447 } 448 SCHED_UNLOCK(s); 449 450 /* 451 * Now can be swapped. 452 */ 453 PRELE(l1); 454 455 /* 456 * Notify any interested parties about the new process. 457 */ 458 KNOTE(&p1->p_klist, NOTE_FORK | p2->p_pid); 459 460 /* 461 * Update stats now that we know the fork was successful. 462 */ 463 uvmexp.forks++; 464 if (flags & FORK_PPWAIT) 465 uvmexp.forks_ppwait++; 466 if (flags & FORK_SHAREVM) 467 uvmexp.forks_sharevm++; 468 469 /* 470 * Pass a pointer to the new process to the caller. 471 */ 472 if (rnewprocp != NULL) 473 *rnewprocp = p2; 474 475 #ifdef KTRACE 476 if (KTRPOINT(p2, KTR_EMUL)) 477 p2->p_traceflag |= KTRFAC_TRC_EMUL; 478 #endif 479 480 /* 481 * Preserve synchronization semantics of vfork. If waiting for 482 * child to exec or exit, set P_PPWAIT on child, and sleep on our 483 * proc (in case of exit). 484 */ 485 if (flags & FORK_PPWAIT) 486 while (p2->p_flag & P_PPWAIT) 487 tsleep(p1, PWAIT, "ppwait", 0); 488 489 /* 490 * Return child pid to parent process, 491 * marking us as parent via retval[1]. 492 */ 493 if (retval != NULL) { 494 retval[0] = p2->p_pid; 495 retval[1] = 0; 496 } 497 498 return (0); 499 } 500 501 #if defined(MULTIPROCESSOR) 502 /* 503 * XXX This is a slight hack to get newly-formed processes to 504 * XXX acquire the kernel lock as soon as they run. 505 */ 506 void 507 proc_trampoline_mp(void) 508 { 509 struct lwp *l; 510 511 l = curlwp; 512 513 SCHED_ASSERT_UNLOCKED(); 514 KERNEL_PROC_LOCK(l); 515 } 516 #endif 517