xref: /netbsd-src/sys/kern/kern_fork.c (revision ed75d7a867996c84cfa88e3b8906816277e957f7)
1 /*	$NetBSD: kern_fork.c,v 1.218 2020/01/29 15:47:52 ad Exp $	*/
2 
3 /*-
4  * Copyright (c) 1999, 2001, 2004, 2006, 2007, 2008, 2019
5  *     The NetBSD Foundation, Inc.
6  * All rights reserved.
7  *
8  * This code is derived from software contributed to The NetBSD Foundation
9  * by Jason R. Thorpe of the Numerical Aerospace Simulation Facility,
10  * NASA Ames Research Center, by Charles M. Hannum, and by Andrew Doran.
11  *
12  * Redistribution and use in source and binary forms, with or without
13  * modification, are permitted provided that the following conditions
14  * are met:
15  * 1. Redistributions of source code must retain the above copyright
16  *    notice, this list of conditions and the following disclaimer.
17  * 2. Redistributions in binary form must reproduce the above copyright
18  *    notice, this list of conditions and the following disclaimer in the
19  *    documentation and/or other materials provided with the distribution.
20  *
21  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
22  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
23  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
24  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
25  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
26  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
27  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
28  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
29  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
30  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
31  * POSSIBILITY OF SUCH DAMAGE.
32  */
33 
34 /*
35  * Copyright (c) 1982, 1986, 1989, 1991, 1993
36  *	The Regents of the University of California.  All rights reserved.
37  * (c) UNIX System Laboratories, Inc.
38  * All or some portions of this file are derived from material licensed
39  * to the University of California by American Telephone and Telegraph
40  * Co. or Unix System Laboratories, Inc. and are reproduced herein with
41  * the permission of UNIX System Laboratories, Inc.
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  *	@(#)kern_fork.c	8.8 (Berkeley) 2/14/95
68  */
69 
70 #include <sys/cdefs.h>
71 __KERNEL_RCSID(0, "$NetBSD: kern_fork.c,v 1.218 2020/01/29 15:47:52 ad Exp $");
72 
73 #include "opt_ktrace.h"
74 #include "opt_dtrace.h"
75 
76 #include <sys/param.h>
77 #include <sys/systm.h>
78 #include <sys/filedesc.h>
79 #include <sys/kernel.h>
80 #include <sys/pool.h>
81 #include <sys/mount.h>
82 #include <sys/proc.h>
83 #include <sys/ras.h>
84 #include <sys/resourcevar.h>
85 #include <sys/vnode.h>
86 #include <sys/file.h>
87 #include <sys/acct.h>
88 #include <sys/ktrace.h>
89 #include <sys/sched.h>
90 #include <sys/signalvar.h>
91 #include <sys/syscall.h>
92 #include <sys/kauth.h>
93 #include <sys/atomic.h>
94 #include <sys/syscallargs.h>
95 #include <sys/uidinfo.h>
96 #include <sys/sdt.h>
97 #include <sys/ptrace.h>
98 
99 /*
100  * DTrace SDT provider definitions
101  */
102 SDT_PROVIDER_DECLARE(proc);
103 SDT_PROBE_DEFINE3(proc, kernel, , create,
104     "struct proc *", /* new process */
105     "struct proc *", /* parent process */
106     "int" /* flags */);
107 
108 u_int	nprocs __cacheline_aligned = 1;		/* process 0 */
109 
110 /*
111  * Number of ticks to sleep if fork() would fail due to process hitting
112  * limits. Exported in miliseconds to userland via sysctl.
113  */
114 int	forkfsleep = 0;
115 
116 int
117 sys_fork(struct lwp *l, const void *v, register_t *retval)
118 {
119 
120 	return fork1(l, 0, SIGCHLD, NULL, 0, NULL, NULL, retval);
121 }
122 
123 /*
124  * vfork(2) system call compatible with 4.4BSD (i.e. BSD with Mach VM).
125  * Address space is not shared, but parent is blocked until child exit.
126  */
127 int
128 sys_vfork(struct lwp *l, const void *v, register_t *retval)
129 {
130 
131 	return fork1(l, FORK_PPWAIT, SIGCHLD, NULL, 0, NULL, NULL,
132 	    retval);
133 }
134 
135 /*
136  * New vfork(2) system call for NetBSD, which implements original 3BSD vfork(2)
137  * semantics.  Address space is shared, and parent is blocked until child exit.
138  */
139 int
140 sys___vfork14(struct lwp *l, const void *v, register_t *retval)
141 {
142 
143 	return fork1(l, FORK_PPWAIT|FORK_SHAREVM, SIGCHLD, NULL, 0,
144 	    NULL, NULL, retval);
145 }
146 
147 /*
148  * Linux-compatible __clone(2) system call.
149  */
150 int
151 sys___clone(struct lwp *l, const struct sys___clone_args *uap,
152     register_t *retval)
153 {
154 	/* {
155 		syscallarg(int) flags;
156 		syscallarg(void *) stack;
157 	} */
158 	int flags, sig;
159 
160 	/*
161 	 * We don't support the CLONE_PID or CLONE_PTRACE flags.
162 	 */
163 	if (SCARG(uap, flags) & (CLONE_PID|CLONE_PTRACE))
164 		return EINVAL;
165 
166 	/*
167 	 * Linux enforces CLONE_VM with CLONE_SIGHAND, do same.
168 	 */
169 	if (SCARG(uap, flags) & CLONE_SIGHAND
170 	    && (SCARG(uap, flags) & CLONE_VM) == 0)
171 		return EINVAL;
172 
173 	flags = 0;
174 
175 	if (SCARG(uap, flags) & CLONE_VM)
176 		flags |= FORK_SHAREVM;
177 	if (SCARG(uap, flags) & CLONE_FS)
178 		flags |= FORK_SHARECWD;
179 	if (SCARG(uap, flags) & CLONE_FILES)
180 		flags |= FORK_SHAREFILES;
181 	if (SCARG(uap, flags) & CLONE_SIGHAND)
182 		flags |= FORK_SHARESIGS;
183 	if (SCARG(uap, flags) & CLONE_VFORK)
184 		flags |= FORK_PPWAIT;
185 
186 	sig = SCARG(uap, flags) & CLONE_CSIGNAL;
187 	if (sig < 0 || sig >= _NSIG)
188 		return EINVAL;
189 
190 	/*
191 	 * Note that the Linux API does not provide a portable way of
192 	 * specifying the stack area; the caller must know if the stack
193 	 * grows up or down.  So, we pass a stack size of 0, so that the
194 	 * code that makes this adjustment is a noop.
195 	 */
196 	return fork1(l, flags, sig, SCARG(uap, stack), 0,
197 	    NULL, NULL, retval);
198 }
199 
200 /*
201  * Print the 'table full' message once per 10 seconds.
202  */
203 static struct timeval fork_tfmrate = { 10, 0 };
204 
205 /*
206  * Check if a process is traced and shall inform about FORK events.
207  */
208 static inline bool
209 tracefork(struct proc *p, int flags)
210 {
211 
212 	return (p->p_slflag & (PSL_TRACEFORK|PSL_TRACED)) ==
213 	    (PSL_TRACEFORK|PSL_TRACED) && (flags & FORK_PPWAIT) == 0;
214 }
215 
216 /*
217  * Check if a process is traced and shall inform about VFORK events.
218  */
219 static inline bool
220 tracevfork(struct proc *p, int flags)
221 {
222 
223 	return (p->p_slflag & (PSL_TRACEVFORK|PSL_TRACED)) ==
224 	    (PSL_TRACEVFORK|PSL_TRACED) && (flags & FORK_PPWAIT) != 0;
225 }
226 
227 /*
228  * Check if a process is traced and shall inform about VFORK_DONE events.
229  */
230 static inline bool
231 tracevforkdone(struct proc *p, int flags)
232 {
233 
234 	return (p->p_slflag & (PSL_TRACEVFORK_DONE|PSL_TRACED)) ==
235 	    (PSL_TRACEVFORK_DONE|PSL_TRACED) && (flags & FORK_PPWAIT);
236 }
237 
238 /*
239  * General fork call.  Note that another LWP in the process may call exec()
240  * or exit() while we are forking.  It's safe to continue here, because
241  * neither operation will complete until all LWPs have exited the process.
242  */
243 int
244 fork1(struct lwp *l1, int flags, int exitsig, void *stack, size_t stacksize,
245     void (*func)(void *), void *arg, register_t *retval)
246 {
247 	struct proc	*p1, *p2, *parent;
248 	struct plimit   *p1_lim;
249 	uid_t		uid;
250 	struct lwp	*l2;
251 	int		count;
252 	vaddr_t		uaddr;
253 	int		tnprocs;
254 	int		error = 0;
255 
256 	p1 = l1->l_proc;
257 	uid = kauth_cred_getuid(l1->l_cred);
258 	tnprocs = atomic_inc_uint_nv(&nprocs);
259 
260 	/*
261 	 * Although process entries are dynamically created, we still keep
262 	 * a global limit on the maximum number we will create.
263 	 */
264 	if (__predict_false(tnprocs >= maxproc))
265 		error = -1;
266 	else
267 		error = kauth_authorize_process(l1->l_cred,
268 		    KAUTH_PROCESS_FORK, p1, KAUTH_ARG(tnprocs), NULL, NULL);
269 
270 	if (error) {
271 		static struct timeval lasttfm;
272 		atomic_dec_uint(&nprocs);
273 		if (ratecheck(&lasttfm, &fork_tfmrate))
274 			tablefull("proc", "increase kern.maxproc or NPROC");
275 		if (forkfsleep)
276 			kpause("forkmx", false, forkfsleep, NULL);
277 		return EAGAIN;
278 	}
279 
280 	/*
281 	 * Enforce limits.
282 	 */
283 	count = chgproccnt(uid, 1);
284 	if (__predict_false(count > p1->p_rlimit[RLIMIT_NPROC].rlim_cur)) {
285 		if (kauth_authorize_process(l1->l_cred, KAUTH_PROCESS_RLIMIT,
286 		    p1, KAUTH_ARG(KAUTH_REQ_PROCESS_RLIMIT_BYPASS),
287 		    &p1->p_rlimit[RLIMIT_NPROC], KAUTH_ARG(RLIMIT_NPROC)) != 0) {
288 			(void)chgproccnt(uid, -1);
289 			atomic_dec_uint(&nprocs);
290 			if (forkfsleep)
291 				kpause("forkulim", false, forkfsleep, NULL);
292 			return EAGAIN;
293 		}
294 	}
295 
296 	/*
297 	 * Allocate virtual address space for the U-area now, while it
298 	 * is still easy to abort the fork operation if we're out of
299 	 * kernel virtual address space.
300 	 */
301 	uaddr = uvm_uarea_alloc();
302 	if (__predict_false(uaddr == 0)) {
303 		(void)chgproccnt(uid, -1);
304 		atomic_dec_uint(&nprocs);
305 		return ENOMEM;
306 	}
307 
308 	/*
309 	 * We are now committed to the fork.  From here on, we may
310 	 * block on resources, but resource allocation may NOT fail.
311 	 */
312 
313 	/* Allocate new proc. */
314 	p2 = proc_alloc();
315 
316 	/*
317 	 * Make a proc table entry for the new process.
318 	 * Start by zeroing the section of proc that is zero-initialized,
319 	 * then copy the section that is copied directly from the parent.
320 	 */
321 	memset(&p2->p_startzero, 0,
322 	    (unsigned) ((char *)&p2->p_endzero - (char *)&p2->p_startzero));
323 	memcpy(&p2->p_startcopy, &p1->p_startcopy,
324 	    (unsigned) ((char *)&p2->p_endcopy - (char *)&p2->p_startcopy));
325 
326 	TAILQ_INIT(&p2->p_sigpend.sp_info);
327 
328 	LIST_INIT(&p2->p_lwps);
329 	LIST_INIT(&p2->p_sigwaiters);
330 	radix_tree_init_tree(&p2->p_lwptree);
331 
332 	/*
333 	 * Duplicate sub-structures as needed.
334 	 * Increase reference counts on shared objects.
335 	 * Inherit flags we want to keep.  The flags related to SIGCHLD
336 	 * handling are important in order to keep a consistent behaviour
337 	 * for the child after the fork.  If we are a 32-bit process, the
338 	 * child will be too.
339 	 */
340 	p2->p_flag =
341 	    p1->p_flag & (PK_SUGID | PK_NOCLDWAIT | PK_CLDSIGIGN | PK_32);
342 	p2->p_emul = p1->p_emul;
343 	p2->p_execsw = p1->p_execsw;
344 
345 	if (flags & FORK_SYSTEM) {
346 		/*
347 		 * Mark it as a system process.  Set P_NOCLDWAIT so that
348 		 * children are reparented to init(8) when they exit.
349 		 * init(8) can easily wait them out for us.
350 		 */
351 		p2->p_flag |= (PK_SYSTEM | PK_NOCLDWAIT);
352 	}
353 
354 	mutex_init(&p2->p_stmutex, MUTEX_DEFAULT, IPL_HIGH);
355 	mutex_init(&p2->p_auxlock, MUTEX_DEFAULT, IPL_NONE);
356 	rw_init(&p2->p_reflock);
357 	rw_init(&p2->p_treelock);
358 	cv_init(&p2->p_waitcv, "wait");
359 	cv_init(&p2->p_lwpcv, "lwpwait");
360 
361 	/*
362 	 * Share a lock between the processes if they are to share signal
363 	 * state: we must synchronize access to it.
364 	 */
365 	if (flags & FORK_SHARESIGS) {
366 		p2->p_lock = p1->p_lock;
367 		mutex_obj_hold(p1->p_lock);
368 	} else
369 		p2->p_lock = mutex_obj_alloc(MUTEX_DEFAULT, IPL_NONE);
370 
371 	kauth_proc_fork(p1, p2);
372 
373 	p2->p_raslist = NULL;
374 #if defined(__HAVE_RAS)
375 	ras_fork(p1, p2);
376 #endif
377 
378 	/* bump references to the text vnode (for procfs) */
379 	p2->p_textvp = p1->p_textvp;
380 	if (p2->p_textvp)
381 		vref(p2->p_textvp);
382 	if (p1->p_path)
383 		p2->p_path = kmem_strdupsize(p1->p_path, NULL, KM_SLEEP);
384 	else
385 		p2->p_path = NULL;
386 
387 	if (flags & FORK_SHAREFILES)
388 		fd_share(p2);
389 	else if (flags & FORK_CLEANFILES)
390 		p2->p_fd = fd_init(NULL);
391 	else
392 		p2->p_fd = fd_copy();
393 
394 	/* XXX racy */
395 	p2->p_mqueue_cnt = p1->p_mqueue_cnt;
396 
397 	if (flags & FORK_SHARECWD)
398 		cwdshare(p2);
399 	else
400 		p2->p_cwdi = cwdinit();
401 
402 	/*
403 	 * Note: p_limit (rlimit stuff) is copy-on-write, so normally
404 	 * we just need increase pl_refcnt.
405 	 */
406 	p1_lim = p1->p_limit;
407 	if (!p1_lim->pl_writeable) {
408 		lim_addref(p1_lim);
409 		p2->p_limit = p1_lim;
410 	} else {
411 		p2->p_limit = lim_copy(p1_lim);
412 	}
413 
414 	if (flags & FORK_PPWAIT) {
415 		/* Mark ourselves as waiting for a child. */
416 		p2->p_lflag = PL_PPWAIT;
417 		l1->l_vforkwaiting = true;
418 		p2->p_vforklwp = l1;
419 	} else {
420 		p2->p_lflag = 0;
421 		l1->l_vforkwaiting = false;
422 	}
423 	p2->p_sflag = 0;
424 	p2->p_slflag = 0;
425 	parent = (flags & FORK_NOWAIT) ? initproc : p1;
426 	p2->p_pptr = parent;
427 	p2->p_ppid = parent->p_pid;
428 	LIST_INIT(&p2->p_children);
429 
430 	p2->p_aio = NULL;
431 
432 #ifdef KTRACE
433 	/*
434 	 * Copy traceflag and tracefile if enabled.
435 	 * If not inherited, these were zeroed above.
436 	 */
437 	if (p1->p_traceflag & KTRFAC_INHERIT) {
438 		mutex_enter(&ktrace_lock);
439 		p2->p_traceflag = p1->p_traceflag;
440 		if ((p2->p_tracep = p1->p_tracep) != NULL)
441 			ktradref(p2);
442 		mutex_exit(&ktrace_lock);
443 	}
444 #endif
445 
446 	/*
447 	 * Create signal actions for the child process.
448 	 */
449 	p2->p_sigacts = sigactsinit(p1, flags & FORK_SHARESIGS);
450 	mutex_enter(p1->p_lock);
451 	p2->p_sflag |=
452 	    (p1->p_sflag & (PS_STOPFORK | PS_STOPEXEC | PS_NOCLDSTOP));
453 	sched_proc_fork(p1, p2);
454 	mutex_exit(p1->p_lock);
455 
456 	p2->p_stflag = p1->p_stflag;
457 
458 	/*
459 	 * p_stats.
460 	 * Copy parts of p_stats, and zero out the rest.
461 	 */
462 	p2->p_stats = pstatscopy(p1->p_stats);
463 
464 	/*
465 	 * Set up the new process address space.
466 	 */
467 	uvm_proc_fork(p1, p2, (flags & FORK_SHAREVM) ? true : false);
468 
469 	/*
470 	 * Finish creating the child process.
471 	 * It will return through a different path later.
472 	 */
473 	lwp_create(l1, p2, uaddr, (flags & FORK_PPWAIT) ? LWP_VFORK : 0,
474 	    stack, stacksize, (func != NULL) ? func : child_return, arg, &l2,
475 	    l1->l_class, &l1->l_sigmask, &l1->l_sigstk);
476 
477 	/*
478 	 * Inherit l_private from the parent.
479 	 * Note that we cannot use lwp_setprivate() here since that
480 	 * also sets the CPU TLS register, which is incorrect if the
481 	 * process has changed that without letting the kernel know.
482 	 */
483 	l2->l_private = l1->l_private;
484 
485 	/*
486 	 * If emulation has a process fork hook, call it now.
487 	 */
488 	if (p2->p_emul->e_proc_fork)
489 		(*p2->p_emul->e_proc_fork)(p2, l1, flags);
490 
491 	/*
492 	 * ...and finally, any other random fork hooks that subsystems
493 	 * might have registered.
494 	 */
495 	doforkhooks(p2, p1);
496 
497 	SDT_PROBE(proc, kernel, , create, p2, p1, flags, 0, 0);
498 
499 	/*
500 	 * It's now safe for the scheduler and other processes to see the
501 	 * child process.
502 	 */
503 	mutex_enter(proc_lock);
504 
505 	if (p1->p_session->s_ttyvp != NULL && p1->p_lflag & PL_CONTROLT)
506 		p2->p_lflag |= PL_CONTROLT;
507 
508 	LIST_INSERT_HEAD(&parent->p_children, p2, p_sibling);
509 	p2->p_exitsig = exitsig;		/* signal for parent on exit */
510 
511 	/*
512 	 * Trace fork(2) and vfork(2)-like events on demand in a debugger.
513 	 */
514 	if (tracefork(p1, flags) || tracevfork(p1, flags))
515 		proc_changeparent(p2, p1->p_pptr);
516 
517 	LIST_INSERT_AFTER(p1, p2, p_pglist);
518 	LIST_INSERT_HEAD(&allproc, p2, p_list);
519 
520 	p2->p_trace_enabled = trace_is_enabled(p2);
521 #ifdef __HAVE_SYSCALL_INTERN
522 	(*p2->p_emul->e_syscall_intern)(p2);
523 #endif
524 
525 	/*
526 	 * Update stats now that we know the fork was successful.
527 	 */
528 	KPREEMPT_DISABLE(l1);
529 	CPU_COUNT(CPU_COUNT_FORKS, 1);
530 	if (flags & FORK_PPWAIT)
531 		CPU_COUNT(CPU_COUNT_FORKS_PPWAIT, 1);
532 	if (flags & FORK_SHAREVM)
533 		CPU_COUNT(CPU_COUNT_FORKS_SHAREVM, 1);
534 	KPREEMPT_ENABLE(l1);
535 
536 	if (ktrpoint(KTR_EMUL))
537 		p2->p_traceflag |= KTRFAC_TRC_EMUL;
538 
539 	/*
540 	 * Notify any interested parties about the new process.
541 	 */
542 	if (!SLIST_EMPTY(&p1->p_klist)) {
543 		mutex_exit(proc_lock);
544 		KNOTE(&p1->p_klist, NOTE_FORK | p2->p_pid);
545 		mutex_enter(proc_lock);
546 	}
547 
548 	/*
549 	 * Make child runnable, set start time, and add to run queue except
550 	 * if the parent requested the child to start in SSTOP state.
551 	 */
552 	mutex_enter(p2->p_lock);
553 
554 	/*
555 	 * Start profiling.
556 	 */
557 	if ((p2->p_stflag & PST_PROFIL) != 0) {
558 		mutex_spin_enter(&p2->p_stmutex);
559 		startprofclock(p2);
560 		mutex_spin_exit(&p2->p_stmutex);
561 	}
562 
563 	getmicrotime(&p2->p_stats->p_start);
564 	p2->p_acflag = AFORK;
565 	lwp_lock(l2);
566 	KASSERT(p2->p_nrlwps == 1);
567 	KASSERT(l2->l_stat == LSIDL);
568 	if (p2->p_sflag & PS_STOPFORK) {
569 		p2->p_nrlwps = 0;
570 		p2->p_stat = SSTOP;
571 		p2->p_waited = 0;
572 		p1->p_nstopchild++;
573 		l2->l_stat = LSSTOP;
574 		KASSERT(l2->l_wchan == NULL);
575 		lwp_unlock(l2);
576 	} else {
577 		p2->p_nrlwps = 1;
578 		p2->p_stat = SACTIVE;
579 		setrunnable(l2);
580 		/* LWP now unlocked */
581 	}
582 
583 	/*
584 	 * Return child pid to parent process,
585 	 * marking us as parent via retval[1].
586 	 */
587 	if (retval != NULL) {
588 		retval[0] = p2->p_pid;
589 		retval[1] = 0;
590 	}
591 
592 	mutex_exit(p2->p_lock);
593 
594 	/*
595 	 * Let the parent know that we are tracing its child.
596 	 */
597 	if (tracefork(p1, flags) || tracevfork(p1, flags)) {
598 		mutex_enter(p1->p_lock);
599 		eventswitch(TRAP_CHLD,
600 		    tracefork(p1, flags) ? PTRACE_FORK : PTRACE_VFORK,
601 		    retval[0]);
602 		mutex_enter(proc_lock);
603 	}
604 
605 	/*
606 	 * Preserve synchronization semantics of vfork.  If waiting for
607 	 * child to exec or exit, sleep until it clears p_vforkwaiting.
608 	 */
609 	while (l1->l_vforkwaiting)
610 		cv_wait(&l1->l_waitcv, proc_lock);
611 
612 	/*
613 	 * Let the parent know that we are tracing its child.
614 	 */
615 	if (tracevforkdone(p1, flags)) {
616 		mutex_enter(p1->p_lock);
617 		eventswitch(TRAP_CHLD, PTRACE_VFORK_DONE, retval[0]);
618 	} else
619 		mutex_exit(proc_lock);
620 
621 	return 0;
622 }
623 
624 /*
625  * MI code executed in each newly spawned process before returning to userland.
626  */
627 void
628 child_return(void *arg)
629 {
630 	struct lwp *l = arg;
631 	struct proc *p = l->l_proc;
632 
633 	if (p->p_slflag & PSL_TRACED) {
634 		/* Paranoid check */
635 		mutex_enter(proc_lock);
636 		if (!(p->p_slflag & PSL_TRACED)) {
637 			mutex_exit(proc_lock);
638 			goto my_tracer_is_gone;
639 		}
640 		mutex_enter(p->p_lock);
641 		eventswitch(TRAP_CHLD,
642 		    ISSET(p->p_lflag, PL_PPWAIT) ? PTRACE_VFORK : PTRACE_FORK,
643 		    p->p_opptr->p_pid);
644 	}
645 
646 my_tracer_is_gone:
647 	md_child_return(l);
648 
649 	/*
650 	 * Return SYS_fork for all fork types, including vfork(2) and clone(2).
651 	 *
652 	 * This approach simplifies the code and avoids extra locking.
653 	 */
654 	ktrsysret(SYS_fork, 0, 0);
655 }
656