1/*
2 * Copyright (c) 2000-2013 Apple Inc. All rights reserved.
3 *
4 * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
5 *
6 * This file contains Original Code and/or Modifications of Original Code
7 * as defined in and that are subject to the Apple Public Source License
8 * Version 2.0 (the 'License'). You may not use this file except in
9 * compliance with the License. The rights granted to you under the License
10 * may not be used to create, or enable the creation or redistribution of,
11 * unlawful or unlicensed copies of an Apple operating system, or to
12 * circumvent, violate, or enable the circumvention or violation of, any
13 * terms of an Apple operating system software license agreement.
14 *
15 * Please obtain a copy of the License at
16 * http://www.opensource.apple.com/apsl/ and read it before using this file.
17 *
18 * The Original Code and all software distributed under the License are
19 * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
20 * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
21 * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
22 * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
23 * Please see the License for the specific language governing rights and
24 * limitations under the License.
25 *
26 * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
27 *
28 *
29 * Copyright (c) 1982, 1986, 1989, 1991, 1992, 1993
30 *	The Regents of the University of California.  All rights reserved.
31 * (c) UNIX System Laboratories, Inc.
32 * All or some portions of this file are derived from material licensed
33 * to the University of California by American Telephone and Telegraph
34 * Co. or Unix System Laboratories, Inc. and are reproduced herein with
35 * the permission of UNIX System Laboratories, Inc.
36 *
37 * Redistribution and use in source and binary forms, with or without
38 * modification, are permitted provided that the following conditions
39 * are met:
40 * 1. Redistributions of source code must retain the above copyright
41 *    notice, this list of conditions and the following disclaimer.
42 * 2. Redistributions in binary form must reproduce the above copyright
43 *    notice, this list of conditions and the following disclaimer in the
44 *    documentation and/or other materials provided with the distribution.
45 * 3. All advertising materials mentioning features or use of this software
46 *    must display the following acknowledgement:
47 *	This product includes software developed by the University of
48 *	California, Berkeley and its contributors.
49 * 4. Neither the name of the University nor the names of its contributors
50 *    may be used to endorse or promote products derived from this software
51 *    without specific prior written permission.
52 *
53 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
54 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
55 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
56 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
57 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
58 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
59 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
60 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
61 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
62 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
63 * SUCH DAMAGE.
64 *
65 *	@(#)init_main.c	8.16 (Berkeley) 5/14/95
66 */
67
68/*
69 *
70 * Mach Operating System
71 * Copyright (c) 1987 Carnegie-Mellon University
72 * All rights reserved.  The CMU software License Agreement specifies
73 * the terms and conditions for use and redistribution.
74 */
75/*
76 * NOTICE: This file was modified by McAfee Research in 2004 to introduce
77 * support for mandatory and extensible security protections.  This notice
78 * is included in support of clause 2.2 (b) of the Apple Public License,
79 * Version 2.0.
80 */
81
82#include <sys/param.h>
83#include <sys/filedesc.h>
84#include <sys/kernel.h>
85#include <sys/mount_internal.h>
86#include <sys/proc_internal.h>
87#include <sys/kauth.h>
88#include <sys/systm.h>
89#include <sys/vnode_internal.h>
90#include <sys/conf.h>
91#include <sys/buf_internal.h>
92#include <sys/clist.h>
93#include <sys/user.h>
94#include <sys/time.h>
95#include <sys/systm.h>
96#include <sys/mman.h>
97
98#include <security/audit/audit.h>
99
100#include <sys/malloc.h>
101#include <sys/dkstat.h>
102#include <sys/codesign.h>
103
104#include <kern/startup.h>
105#include <kern/thread.h>
106#include <kern/task.h>
107#include <kern/ast.h>
108#include <kern/kalloc.h>
109#include <mach/mach_host.h>
110
111#include <mach/vm_param.h>
112
113#include <vm/vm_map.h>
114#include <vm/vm_kern.h>
115
116#include <sys/ux_exception.h>	/* for ux_exception_port */
117
118#include <sys/reboot.h>
119#include <mach/exception_types.h>
120#include <dev/busvar.h>			/* for pseudo_inits */
121#include <sys/kdebug.h>
122
123#include <mach/mach_types.h>
124#include <mach/vm_prot.h>
125#include <mach/semaphore.h>
126#include <mach/sync_policy.h>
127#include <kern/clock.h>
128#include <mach/kern_return.h>
129#include <mach/thread_act.h>		/* for thread_resume() */
130#include <mach/task.h>			/* for task_set_exception_ports() */
131#include <sys/ux_exception.h>		/* for ux_handler() */
132#include <sys/ubc_internal.h>		/* for ubc_init() */
133#include <sys/mcache.h>			/* for mcache_init() */
134#include <sys/mbuf.h>			/* for mbinit() */
135#include <sys/event.h>			/* for knote_init() */
136#include <sys/kern_memorystatus.h>	/* for memorystatus_init() */
137#include <sys/aio_kern.h>		/* for aio_init() */
138#include <sys/semaphore.h>		/* for psem_cache_init() */
139#include <net/dlil.h>			/* for dlil_init() */
140#include <net/kpi_protocol.h>		/* for proto_kpi_init() */
141#include <net/iptap.h>			/* for iptap_init() */
142#include <sys/pipe.h>			/* for pipeinit() */
143#include <sys/socketvar.h>		/* for socketinit() */
144#include <sys/protosw.h>		/* for domaininit() */
145#include <kern/sched_prim.h>		/* for thread_wakeup() */
146#include <net/if_ether.h>		/* for ether_family_init() */
147#include <net/if_gif.h>			/* for gif_init() */
148#include <vm/vm_protos.h>		/* for vnode_pager_bootstrap() */
149#include <miscfs/devfs/devfsdefs.h>	/* for devfs_kernel_mount() */
150#include <mach/host_priv.h>		/* for host_set_exception_ports() */
151#include <kern/host.h>			/* for host_priv_self() */
152#include <vm/vm_kern.h>			/* for kmem_suballoc() */
153#include <sys/semaphore.h>		/* for psem_lock_init() */
154#include <sys/msgbuf.h>			/* for log_setsize() */
155#include <sys/tty.h>			/* for tty_init() */
156#include <sys/proc_uuid_policy.h>	/* proc_uuid_policy_init() */
157#include <netinet/flow_divert.h>	/* flow_divert_init() */
158#include <net/if_utun.h>		/* for utun_register_control() */
159#include <net/if_ipsec.h>       /* for ipsec_register_control() */
160#include <net/net_str_id.h>		/* for net_str_id_init() */
161#include <net/netsrc.h>			/* for netsrc_init() */
162#include <net/ntstat.h>			/* for nstat_init() */
163#include <kern/assert.h>		/* for assert() */
164#include <sys/kern_overrides.h>		/* for init_system_override() */
165
166#include <net/init.h>
167
168#if CONFIG_MACF
169#include <security/mac_framework.h>
170#include <security/mac_internal.h>	/* mac_init_bsd() */
171#include <security/mac_mach_internal.h>	/* mac_update_task_label() */
172#endif
173
174#include <machine/exec.h>
175
176#if NFSCLIENT
177#include <sys/netboot.h>
178#endif
179
180#if CONFIG_IMAGEBOOT
181#include <sys/imageboot.h>
182#endif
183
184#if PFLOG
185#include <net/if_pflog.h>
186#endif
187
188#include <pexpert/pexpert.h>
189#include <machine/pal_routines.h>
190#include <console/video_console.h>
191
192void * get_user_regs(thread_t);		/* XXX kludge for <machine/thread.h> */
193void IOKitInitializeTime(void);		/* XXX */
194void IOSleep(unsigned int);		/* XXX */
195void loopattach(void);			/* XXX */
196
197const char    copyright[] =
198"Copyright (c) 1982, 1986, 1989, 1991, 1993\n\t"
199"The Regents of the University of California. "
200"All rights reserved.\n\n";
201
202/* Components of the first process -- never freed. */
203struct	proc proc0;
204struct	session session0;
205struct	pgrp pgrp0;
206struct	filedesc filedesc0;
207struct	plimit limit0;
208struct	pstats pstats0;
209struct	sigacts sigacts0;
210proc_t kernproc;
211proc_t initproc;
212
213long tk_cancc;
214long tk_nin;
215long tk_nout;
216long tk_rawcc;
217
218int lock_trace = 0;
219/* Global variables to make pstat happy. We do swapping differently */
220int nswdev, nswap;
221int nswapmap;
222void *swapmap;
223struct swdevt swdevt[1];
224
225dev_t	rootdev;		/* device of the root */
226dev_t	dumpdev;		/* device to take dumps on */
227long	dumplo;			/* offset into dumpdev */
228long	hostid;
229char	hostname[MAXHOSTNAMELEN];
230int		hostnamelen;
231char	domainname[MAXDOMNAMELEN];
232int		domainnamelen;
233
234char rootdevice[16]; 	/* hfs device names have at least 9 chars */
235
236#if  KMEMSTATS
237struct	kmemstats kmemstats[M_LAST];
238#endif
239
240struct	vnode *rootvp;
241int boothowto = RB_DEBUG;
242
243extern kern_return_t IOFindBSDRoot(char *, unsigned int, dev_t *, u_int32_t *);
244extern void IOSecureBSDRoot(const char * rootName);
245extern kern_return_t IOKitBSDInit(void );
246extern void kminit(void);
247extern void file_lock_init(void);
248extern void kmeminit(void);
249extern void bsd_bufferinit(void);
250extern void throttle_init(void);
251
252extern int serverperfmode;
253extern int ncl;
254
255vm_map_t	bsd_pageable_map;
256vm_map_t	mb_map;
257
258static  int bsd_simul_execs;
259static int bsd_pageable_map_size;
260__private_extern__ int execargs_cache_size = 0;
261__private_extern__ int execargs_free_count = 0;
262__private_extern__ vm_offset_t * execargs_cache = NULL;
263
264void bsd_exec_setup(int);
265
266__private_extern__ int bootarg_vnode_cache_defeat = 0;
267
268/*
269 * Prevent kernel-based ASLR from being used, for testing.
270 */
271#if DEVELOPMENT || DEBUG
272__private_extern__ int bootarg_disable_aslr = 0;
273#endif
274
275int	cmask = CMASK;
276extern int customnbuf;
277
278void bsd_init(void);
279kern_return_t bsd_autoconf(void);
280void bsd_utaskbootstrap(void);
281
282static void parse_bsd_args(void);
283extern task_t bsd_init_task;
284extern char    init_task_failure_data[];
285extern void time_zone_slock_init(void);
286extern void select_wait_queue_init(void);
287static void process_name(const char *, proc_t);
288
289static void setconf(void);
290
291funnel_t *kernel_flock;
292
293#if SYSV_SHM
294extern void sysv_shm_lock_init(void);
295#endif
296#if SYSV_SEM
297extern void sysv_sem_lock_init(void);
298#endif
299#if SYSV_MSG
300extern void sysv_msg_lock_init(void);
301#endif
302
303#if !defined(SECURE_KERNEL)
304/* kmem access not enabled by default; can be changed with boot-args */
305/* We don't need to keep this symbol around in RELEASE kernel */
306int setup_kmem = 0;
307#endif
308
309#if CONFIG_MACF
310#if defined (__i386__) || defined (__x86_64__)
311/* MACF policy_check configuration flags; see policy_check.c for details */
312int policy_check_flags = 0;
313
314extern int check_policy_init(int);
315#endif
316#endif	/* CONFIG_MACF */
317
318/* If we are using CONFIG_DTRACE */
319#if CONFIG_DTRACE
320	extern void dtrace_postinit(void);
321#endif
322
323/*
324 * Initialization code.
325 * Called from cold start routine as
326 * soon as a stack and segmentation
327 * have been established.
328 * Functions:
329 *	turn on clock
330 *	hand craft 0th process
331 *	call all initialization routines
332 *  hand craft 1st user process
333 */
334
335/*
336 *	Sets the name for the given task.
337 */
338static void
339process_name(const char *s, proc_t p)
340{
341	size_t length = strlen(s);
342
343	bcopy(s, p->p_comm,
344		length >= sizeof(p->p_comm) ? sizeof(p->p_comm) :
345			length + 1);
346}
347
348/* To allow these values to be patched, they're globals here */
349#include <machine/vmparam.h>
350struct rlimit vm_initial_limit_stack = { DFLSSIZ, MAXSSIZ - PAGE_SIZE };
351struct rlimit vm_initial_limit_data = { DFLDSIZ, MAXDSIZ };
352struct rlimit vm_initial_limit_core = { DFLCSIZ, MAXCSIZ };
353
354extern thread_t	cloneproc(task_t, proc_t, int, int);
355extern int 	(*mountroot)(void);
356
357lck_grp_t * proc_lck_grp;
358lck_grp_t * proc_slock_grp;
359lck_grp_t * proc_fdmlock_grp;
360lck_grp_t * proc_mlock_grp;
361lck_grp_attr_t * proc_lck_grp_attr;
362lck_attr_t * proc_lck_attr;
363lck_mtx_t * proc_list_mlock;
364lck_mtx_t * proc_klist_mlock;
365
366extern lck_mtx_t * execargs_cache_lock;
367
368/* hook called after root is mounted XXX temporary hack */
369void (*mountroot_post_hook)(void);
370void (*unmountroot_pre_hook)(void);
371
372/*
373 * This function is called very early on in the Mach startup, from the
374 * function start_kernel_threads() in osfmk/kern/startup.c.  It's called
375 * in the context of the current (startup) task using a call to the
376 * function kernel_thread_create() to jump into start_kernel_threads().
377 * Internally, kernel_thread_create() calls thread_create_internal(),
378 * which calls uthread_alloc().  The function of uthread_alloc() is
379 * normally to allocate a uthread structure, and fill out the uu_sigmask,
380 * uu_context fields.  It skips filling these out in the case of the "task"
381 * being "kernel_task", because the order of operation is inverted.  To
382 * account for that, we need to manually fill in at least the contents
383 * of the uu_context.vc_ucred field so that the uthread structure can be
384 * used like any other.
385 */
386
387extern void IOServicePublishResource(const char *, boolean_t);
388
389void
390bsd_init(void)
391{
392	struct uthread *ut;
393	unsigned int i;
394	struct vfs_context context;
395	kern_return_t	ret;
396	struct ucred temp_cred;
397	struct posix_cred temp_pcred;
398#if NFSCLIENT || CONFIG_IMAGEBOOT
399	boolean_t       netboot = FALSE;
400#endif
401
402#define bsd_init_kprintf(x...) /* kprintf("bsd_init: " x) */
403
404	throttle_init();
405
406	kernel_flock = funnel_alloc(KERNEL_FUNNEL);
407	if (kernel_flock == (funnel_t *)0 ) {
408		panic("bsd_init: Failed to allocate kernel funnel");
409	}
410
411	printf(copyright);
412
413	bsd_init_kprintf("calling kmeminit\n");
414	kmeminit();
415
416	bsd_init_kprintf("calling parse_bsd_args\n");
417	parse_bsd_args();
418
419	/* Initialize kauth subsystem before instancing the first credential */
420	bsd_init_kprintf("calling kauth_init\n");
421	kauth_init();
422
423	/* Initialize process and pgrp structures. */
424	bsd_init_kprintf("calling procinit\n");
425	procinit();
426
427	/* Initialize the ttys (MUST be before kminit()/bsd_autoconf()!)*/
428	tty_init();
429
430	kernproc = &proc0;	/* implicitly bzero'ed */
431
432	/* kernel_task->proc = kernproc; */
433	set_bsdtask_info(kernel_task,(void *)kernproc);
434
435	/* give kernproc a name */
436	bsd_init_kprintf("calling process_name\n");
437	process_name("kernel_task", kernproc);
438
439	/* allocate proc lock group attribute and group */
440	bsd_init_kprintf("calling lck_grp_attr_alloc_init\n");
441	proc_lck_grp_attr= lck_grp_attr_alloc_init();
442
443	proc_lck_grp = lck_grp_alloc_init("proc",  proc_lck_grp_attr);
444#if CONFIG_FINE_LOCK_GROUPS
445	proc_slock_grp = lck_grp_alloc_init("proc-slock",  proc_lck_grp_attr);
446	proc_fdmlock_grp = lck_grp_alloc_init("proc-fdmlock",  proc_lck_grp_attr);
447	proc_mlock_grp = lck_grp_alloc_init("proc-mlock",  proc_lck_grp_attr);
448#endif
449	/* Allocate proc lock attribute */
450	proc_lck_attr = lck_attr_alloc_init();
451#if 0
452#if __PROC_INTERNAL_DEBUG
453	lck_attr_setdebug(proc_lck_attr);
454#endif
455#endif
456
457#if CONFIG_FINE_LOCK_GROUPS
458	proc_list_mlock = lck_mtx_alloc_init(proc_mlock_grp, proc_lck_attr);
459	proc_klist_mlock = lck_mtx_alloc_init(proc_mlock_grp, proc_lck_attr);
460	lck_mtx_init(&kernproc->p_mlock, proc_mlock_grp, proc_lck_attr);
461	lck_mtx_init(&kernproc->p_fdmlock, proc_fdmlock_grp, proc_lck_attr);
462	lck_spin_init(&kernproc->p_slock, proc_slock_grp, proc_lck_attr);
463#else
464	proc_list_mlock = lck_mtx_alloc_init(proc_lck_grp, proc_lck_attr);
465	proc_klist_mlock = lck_mtx_alloc_init(proc_lck_grp, proc_lck_attr);
466	lck_mtx_init(&kernproc->p_mlock, proc_lck_grp, proc_lck_attr);
467	lck_mtx_init(&kernproc->p_fdmlock, proc_lck_grp, proc_lck_attr);
468	lck_spin_init(&kernproc->p_slock, proc_lck_grp, proc_lck_attr);
469#endif
470
471	assert(bsd_simul_execs != 0);
472	execargs_cache_lock = lck_mtx_alloc_init(proc_lck_grp, proc_lck_attr);
473	execargs_cache_size = bsd_simul_execs;
474	execargs_free_count = bsd_simul_execs;
475	execargs_cache = (vm_offset_t *)kalloc(bsd_simul_execs * sizeof(vm_offset_t));
476	bzero(execargs_cache, bsd_simul_execs * sizeof(vm_offset_t));
477
478	if (current_task() != kernel_task)
479		printf("bsd_init: We have a problem, "
480				"current task is not kernel task\n");
481
482	bsd_init_kprintf("calling get_bsdthread_info\n");
483	ut = (uthread_t)get_bsdthread_info(current_thread());
484
485#if CONFIG_MACF
486	/*
487	 * Initialize the MAC Framework
488	 */
489	mac_policy_initbsd();
490	kernproc->p_mac_enforce = 0;
491
492#if defined (__i386__) || defined (__x86_64__)
493	/*
494	 * We currently only support this on i386/x86_64, as that is the
495	 * only lock code we have instrumented so far.
496	 */
497	check_policy_init(policy_check_flags);
498#endif
499#endif /* MAC */
500
501	/* Initialize System Override call */
502	init_system_override();
503
504	/*
505	 * Create process 0.
506	 */
507	proc_list_lock();
508	LIST_INSERT_HEAD(&allproc, kernproc, p_list);
509	kernproc->p_pgrp = &pgrp0;
510	LIST_INSERT_HEAD(PGRPHASH(0), &pgrp0, pg_hash);
511	LIST_INIT(&pgrp0.pg_members);
512#ifdef CONFIG_FINE_LOCK_GROUPS
513	lck_mtx_init(&pgrp0.pg_mlock, proc_mlock_grp, proc_lck_attr);
514#else
515	lck_mtx_init(&pgrp0.pg_mlock, proc_lck_grp, proc_lck_attr);
516#endif
517	/* There is no other bsd thread this point and is safe without pgrp lock */
518	LIST_INSERT_HEAD(&pgrp0.pg_members, kernproc, p_pglist);
519	kernproc->p_listflag |= P_LIST_INPGRP;
520	kernproc->p_pgrpid = 0;
521	kernproc->p_uniqueid = 0;
522
523	pgrp0.pg_session = &session0;
524	pgrp0.pg_membercnt = 1;
525
526	session0.s_count = 1;
527	session0.s_leader = kernproc;
528	session0.s_listflags = 0;
529#ifdef CONFIG_FINE_LOCK_GROUPS
530	lck_mtx_init(&session0.s_mlock, proc_mlock_grp, proc_lck_attr);
531#else
532	lck_mtx_init(&session0.s_mlock, proc_lck_grp, proc_lck_attr);
533#endif
534	LIST_INSERT_HEAD(SESSHASH(0), &session0, s_hash);
535	proc_list_unlock();
536
537#if CONFIG_LCTX
538	kernproc->p_lctx = NULL;
539#endif
540
541	kernproc->task = kernel_task;
542
543	kernproc->p_stat = SRUN;
544	kernproc->p_flag = P_SYSTEM;
545	kernproc->p_lflag = 0;
546	kernproc->p_ladvflag = 0;
547
548#if DEVELOPMENT || DEBUG
549	if (bootarg_disable_aslr)
550		kernproc->p_flag |= P_DISABLE_ASLR;
551#endif
552
553	kernproc->p_nice = NZERO;
554	kernproc->p_pptr = kernproc;
555
556	TAILQ_INIT(&kernproc->p_uthlist);
557	TAILQ_INSERT_TAIL(&kernproc->p_uthlist, ut, uu_list);
558
559	kernproc->sigwait = FALSE;
560	kernproc->sigwait_thread = THREAD_NULL;
561	kernproc->exit_thread = THREAD_NULL;
562	kernproc->p_csflags = CS_VALID;
563
564	/*
565	 * Create credential.  This also Initializes the audit information.
566	 */
567	bsd_init_kprintf("calling bzero\n");
568	bzero(&temp_cred, sizeof(temp_cred));
569	bzero(&temp_pcred, sizeof(temp_pcred));
570	temp_pcred.cr_ngroups = 1;
571	/* kern_proc, shouldn't call up to DS for group membership */
572	temp_pcred.cr_flags = CRF_NOMEMBERD;
573	temp_cred.cr_audit.as_aia_p = audit_default_aia_p;
574
575	bsd_init_kprintf("calling kauth_cred_create\n");
576	/*
577	 * We have to label the temp cred before we create from it to
578	 * properly set cr_ngroups, or the create will fail.
579	 */
580	posix_cred_label(&temp_cred, &temp_pcred);
581	kernproc->p_ucred = kauth_cred_create(&temp_cred);
582
583	/* update cred on proc */
584	PROC_UPDATE_CREDS_ONPROC(kernproc);
585
586	/* give the (already exisiting) initial thread a reference on it */
587	bsd_init_kprintf("calling kauth_cred_ref\n");
588	kauth_cred_ref(kernproc->p_ucred);
589	ut->uu_context.vc_ucred = kernproc->p_ucred;
590	ut->uu_context.vc_thread = current_thread();
591
592	TAILQ_INIT(&kernproc->p_aio_activeq);
593	TAILQ_INIT(&kernproc->p_aio_doneq);
594	kernproc->p_aio_total_count = 0;
595	kernproc->p_aio_active_count = 0;
596
597	bsd_init_kprintf("calling file_lock_init\n");
598	file_lock_init();
599
600#if CONFIG_MACF
601	mac_cred_label_associate_kernel(kernproc->p_ucred);
602	mac_task_label_update_cred (kernproc->p_ucred, (struct task *) kernproc->task);
603#endif
604
605	/* Create the file descriptor table. */
606	kernproc->p_fd = &filedesc0;
607	filedesc0.fd_cmask = cmask;
608	filedesc0.fd_knlistsize = -1;
609	filedesc0.fd_knlist = NULL;
610	filedesc0.fd_knhash = NULL;
611	filedesc0.fd_knhashmask = 0;
612
613	/* Create the limits structures. */
614	kernproc->p_limit = &limit0;
615	for (i = 0; i < sizeof(kernproc->p_rlimit)/sizeof(kernproc->p_rlimit[0]); i++)
616		limit0.pl_rlimit[i].rlim_cur =
617			limit0.pl_rlimit[i].rlim_max = RLIM_INFINITY;
618	limit0.pl_rlimit[RLIMIT_NOFILE].rlim_cur = NOFILE;
619	limit0.pl_rlimit[RLIMIT_NPROC].rlim_cur = maxprocperuid;
620	limit0.pl_rlimit[RLIMIT_NPROC].rlim_max = maxproc;
621	limit0.pl_rlimit[RLIMIT_STACK] = vm_initial_limit_stack;
622	limit0.pl_rlimit[RLIMIT_DATA] = vm_initial_limit_data;
623	limit0.pl_rlimit[RLIMIT_CORE] = vm_initial_limit_core;
624	limit0.pl_refcnt = 1;
625
626	kernproc->p_stats = &pstats0;
627	kernproc->p_sigacts = &sigacts0;
628
629	/*
630	 * Charge root for two  processes: init and mach_init.
631	 */
632	bsd_init_kprintf("calling chgproccnt\n");
633	(void)chgproccnt(0, 1);
634
635	/*
636	 *	Allocate a kernel submap for pageable memory
637	 *	for temporary copying (execve()).
638	 */
639	{
640		vm_offset_t	minimum;
641
642		bsd_init_kprintf("calling kmem_suballoc\n");
643		assert(bsd_pageable_map_size != 0);
644		ret = kmem_suballoc(kernel_map,
645				&minimum,
646				(vm_size_t)bsd_pageable_map_size,
647				TRUE,
648				VM_FLAGS_ANYWHERE,
649				&bsd_pageable_map);
650		if (ret != KERN_SUCCESS)
651			panic("bsd_init: Failed to allocate bsd pageable map");
652	}
653
654	/*
655	 * Initialize buffers and hash links for buffers
656	 *
657	 * SIDE EFFECT: Starts a thread for bcleanbuf_thread(), so must
658	 *		happen after a credential has been associated with
659	 *		the kernel task.
660	 */
661	bsd_init_kprintf("calling bsd_bufferinit\n");
662	bsd_bufferinit();
663
664	/* Initialize the execve() semaphore */
665	bsd_init_kprintf("calling semaphore_create\n");
666
667	if (ret != KERN_SUCCESS)
668		panic("bsd_init: Failed to create execve semaphore");
669
670	/*
671	 * Initialize the calendar.
672	 */
673	bsd_init_kprintf("calling IOKitInitializeTime\n");
674	IOKitInitializeTime();
675
676	bsd_init_kprintf("calling ubc_init\n");
677	ubc_init();
678
679	/*
680	 * Initialize device-switches.
681	 */
682	bsd_init_kprintf("calling devsw_init() \n");
683	devsw_init();
684
685	/* Initialize the file systems. */
686	bsd_init_kprintf("calling vfsinit\n");
687	vfsinit();
688
689#if CONFIG_PROC_UUID_POLICY
690	/* Initial proc_uuid_policy subsystem */
691	bsd_init_kprintf("calling proc_uuid_policy_init()\n");
692	proc_uuid_policy_init();
693#endif
694
695#if SOCKETS
696	/* Initialize per-CPU cache allocator */
697	mcache_init();
698
699	/* Initialize mbuf's. */
700	bsd_init_kprintf("calling mbinit\n");
701	mbinit();
702	net_str_id_init(); /* for mbuf tags */
703#endif /* SOCKETS */
704
705	/*
706	 * Initializes security event auditing.
707	 * XXX: Should/could this occur later?
708	 */
709#if CONFIG_AUDIT
710	bsd_init_kprintf("calling audit_init\n");
711 	audit_init();
712#endif
713
714	/* Initialize kqueues */
715	bsd_init_kprintf("calling knote_init\n");
716	knote_init();
717
718	/* Initialize for async IO */
719	bsd_init_kprintf("calling aio_init\n");
720	aio_init();
721
722	/* Initialize pipes */
723	bsd_init_kprintf("calling pipeinit\n");
724	pipeinit();
725
726	/* Initialize SysV shm subsystem locks; the subsystem proper is
727	 * initialized through a sysctl.
728	 */
729#if SYSV_SHM
730	bsd_init_kprintf("calling sysv_shm_lock_init\n");
731	sysv_shm_lock_init();
732#endif
733#if SYSV_SEM
734	bsd_init_kprintf("calling sysv_sem_lock_init\n");
735	sysv_sem_lock_init();
736#endif
737#if SYSV_MSG
738	bsd_init_kprintf("sysv_msg_lock_init\n");
739	sysv_msg_lock_init();
740#endif
741	bsd_init_kprintf("calling pshm_lock_init\n");
742	pshm_lock_init();
743	bsd_init_kprintf("calling psem_lock_init\n");
744	psem_lock_init();
745
746	pthread_init();
747	/* POSIX Shm and Sem */
748	bsd_init_kprintf("calling pshm_cache_init\n");
749	pshm_cache_init();
750	bsd_init_kprintf("calling psem_cache_init\n");
751	psem_cache_init();
752	bsd_init_kprintf("calling time_zone_slock_init\n");
753	time_zone_slock_init();
754	bsd_init_kprintf("calling select_wait_queue_init\n");
755	select_wait_queue_init();
756
757	/*
758	 * Initialize protocols.  Block reception of incoming packets
759	 * until everything is ready.
760	 */
761	bsd_init_kprintf("calling sysctl_register_fixed\n");
762	sysctl_register_fixed();
763	bsd_init_kprintf("calling sysctl_mib_init\n");
764	sysctl_mib_init();
765#if NETWORKING
766	bsd_init_kprintf("calling dlil_init\n");
767	dlil_init();
768	bsd_init_kprintf("calling proto_kpi_init\n");
769	proto_kpi_init();
770#endif /* NETWORKING */
771#if SOCKETS
772	bsd_init_kprintf("calling socketinit\n");
773	socketinit();
774	bsd_init_kprintf("calling domaininit\n");
775	domaininit();
776	iptap_init();
777#if FLOW_DIVERT
778	flow_divert_init();
779#endif	/* FLOW_DIVERT */
780#endif /* SOCKETS */
781
782	kernproc->p_fd->fd_cdir = NULL;
783	kernproc->p_fd->fd_rdir = NULL;
784
785#if CONFIG_FREEZE
786#ifndef CONFIG_MEMORYSTATUS
787    #error "CONFIG_FREEZE defined without matching CONFIG_MEMORYSTATUS"
788#endif
789	/* Initialise background freezing */
790	bsd_init_kprintf("calling memorystatus_freeze_init\n");
791	memorystatus_freeze_init();
792#endif
793
794#if CONFIG_MEMORYSTATUS
795	/* Initialize kernel memory status notifications */
796	bsd_init_kprintf("calling memorystatus_init\n");
797	memorystatus_init();
798#endif /* CONFIG_MEMORYSTATUS */
799
800#ifdef GPROF
801	/* Initialize kernel profiling. */
802	kmstartup();
803#endif
804
805	bsd_init_kprintf("calling bsd_autoconf\n");
806	bsd_autoconf();
807
808#if CONFIG_DTRACE
809	dtrace_postinit();
810#endif
811
812	/*
813	 * We attach the loopback interface *way* down here to ensure
814	 * it happens after autoconf(), otherwise it becomes the
815	 * "primary" interface.
816	 */
817#include <loop.h>
818#if NLOOP > 0
819	bsd_init_kprintf("calling loopattach\n");
820	loopattach();			/* XXX */
821#endif
822#if NGIF
823	/* Initialize gif interface (after lo0) */
824	gif_init();
825#endif
826
827#if PFLOG
828	/* Initialize packet filter log interface */
829	pfloginit();
830#endif /* PFLOG */
831
832#if NETHER > 0
833	/* Register the built-in dlil ethernet interface family */
834	bsd_init_kprintf("calling ether_family_init\n");
835	ether_family_init();
836#endif /* ETHER */
837
838#if NETWORKING
839	/* Call any kext code that wants to run just after network init */
840	bsd_init_kprintf("calling net_init_run\n");
841	net_init_run();
842
843	/* register user tunnel kernel control handler */
844	utun_register_control();
845#if IPSEC
846	ipsec_register_control();
847#endif /* IPSEC */
848	netsrc_init();
849	nstat_init();
850#endif /* NETWORKING */
851
852	bsd_init_kprintf("calling vnode_pager_bootstrap\n");
853	vnode_pager_bootstrap();
854#if 0
855	/* XXX Hack for early debug stop */
856	printf("\nabout to sleep for 10 seconds\n");
857	IOSleep( 10 * 1000 );
858	/* Debugger("hello"); */
859#endif
860
861	bsd_init_kprintf("calling inittodr\n");
862	inittodr(0);
863
864	/* Mount the root file system. */
865	while( TRUE) {
866		int err;
867
868		bsd_init_kprintf("calling setconf\n");
869		setconf();
870#if NFSCLIENT
871		netboot = (mountroot == netboot_mountroot);
872#endif
873
874		bsd_init_kprintf("vfs_mountroot\n");
875		if (0 == (err = vfs_mountroot()))
876			break;
877		rootdevice[0] = '\0';
878#if NFSCLIENT
879		if (netboot) {
880			PE_display_icon( 0, "noroot");  /* XXX a netboot-specific icon would be nicer */
881			vc_progress_set(FALSE, 0);
882			for (i=1; 1; i*=2) {
883				printf("bsd_init: failed to mount network root, error %d, %s\n",
884					err, PE_boot_args());
885				printf("We are hanging here...\n");
886				IOSleep(i*60*1000);
887			}
888			/*NOTREACHED*/
889		}
890#endif
891		printf("cannot mount root, errno = %d\n", err);
892		boothowto |= RB_ASKNAME;
893	}
894
895	IOSecureBSDRoot(rootdevice);
896
897	context.vc_thread = current_thread();
898	context.vc_ucred = kernproc->p_ucred;
899	mountlist.tqh_first->mnt_flag |= MNT_ROOTFS;
900
901	bsd_init_kprintf("calling VFS_ROOT\n");
902	/* Get the vnode for '/'.  Set fdp->fd_fd.fd_cdir to reference it. */
903	if (VFS_ROOT(mountlist.tqh_first, &rootvnode, &context))
904		panic("bsd_init: cannot find root vnode: %s", PE_boot_args());
905	rootvnode->v_flag |= VROOT;
906	(void)vnode_ref(rootvnode);
907	(void)vnode_put(rootvnode);
908	filedesc0.fd_cdir = rootvnode;
909
910#if NFSCLIENT
911	if (netboot) {
912		int err;
913
914		netboot = TRUE;
915		/* post mount setup */
916		if ((err = netboot_setup()) != 0) {
917			PE_display_icon( 0, "noroot");  /* XXX a netboot-specific icon would be nicer */
918			vc_progress_set(FALSE, 0);
919			for (i=1; 1; i*=2) {
920				printf("bsd_init: NetBoot could not find root, error %d: %s\n",
921					err, PE_boot_args());
922				printf("We are hanging here...\n");
923				IOSleep(i*60*1000);
924			}
925			/*NOTREACHED*/
926		}
927	}
928#endif
929
930
931#if CONFIG_IMAGEBOOT
932	/*
933	 * See if a system disk image is present. If so, mount it and
934	 * switch the root vnode to point to it
935	 */
936	if (netboot == FALSE && imageboot_needed()) {
937		/*
938		 * An image was found.  No turning back: we're booted
939		 * with a kernel from the disk image.
940		 */
941		imageboot_setup();
942	}
943#endif /* CONFIG_IMAGEBOOT */
944
945	/* set initial time; all other resource data is  already zero'ed */
946	microtime_with_abstime(&kernproc->p_start, &kernproc->p_stats->ps_start);
947
948#if DEVFS
949	{
950	    char mounthere[] = "/dev";	/* !const because of internal casting */
951
952	    bsd_init_kprintf("calling devfs_kernel_mount\n");
953	    devfs_kernel_mount(mounthere);
954	}
955#endif /* DEVFS */
956
957	/* Initialize signal state for process 0. */
958	bsd_init_kprintf("calling siginit\n");
959	siginit(kernproc);
960
961	bsd_init_kprintf("calling bsd_utaskbootstrap\n");
962	bsd_utaskbootstrap();
963
964#if defined(__LP64__)
965	kernproc->p_flag |= P_LP64;
966#endif
967
968	pal_kernel_announce();
969
970	bsd_init_kprintf("calling mountroot_post_hook\n");
971
972	/* invoke post-root-mount hook */
973	if (mountroot_post_hook != NULL)
974		mountroot_post_hook();
975
976#if 0 /* not yet */
977	consider_zone_gc(FALSE);
978#endif
979
980	bsd_init_kprintf("done\n");
981
982}
983
984/* Called with kernel funnel held */
985void
986bsdinit_task(void)
987{
988	proc_t p = current_proc();
989	struct uthread *ut;
990	thread_t thread;
991
992	process_name("init", p);
993
994	ux_handler_init();
995
996	thread = current_thread();
997	(void) host_set_exception_ports(host_priv_self(),
998					EXC_MASK_ALL & ~(EXC_MASK_RPC_ALERT),//pilotfish (shark) needs this port
999					(mach_port_t) ux_exception_port,
1000					EXCEPTION_DEFAULT| MACH_EXCEPTION_CODES,
1001					0);
1002
1003	ut = (uthread_t)get_bsdthread_info(thread);
1004
1005	bsd_init_task = get_threadtask(thread);
1006	init_task_failure_data[0] = 0;
1007
1008#if CONFIG_MACF
1009	mac_cred_label_associate_user(p->p_ucred);
1010	mac_task_label_update_cred (p->p_ucred, (struct task *) p->task);
1011#endif
1012	load_init_program(p);
1013	lock_trace = 1;
1014}
1015
1016kern_return_t
1017bsd_autoconf(void)
1018{
1019	kprintf("bsd_autoconf: calling kminit\n");
1020	kminit();
1021
1022	/*
1023	 * Early startup for bsd pseudodevices.
1024	 */
1025	{
1026	    struct pseudo_init *pi;
1027
1028	    for (pi = pseudo_inits; pi->ps_func; pi++)
1029		(*pi->ps_func) (pi->ps_count);
1030	}
1031
1032	return( IOKitBSDInit());
1033}
1034
1035
1036#include <sys/disklabel.h>  /* for MAXPARTITIONS */
1037
1038static void
1039setconf(void)
1040{
1041	u_int32_t	flags;
1042	kern_return_t	err;
1043
1044	/*
1045	 * calls into IOKit can generate networking registrations
1046	 * which needs to be under network funnel. Right thing to do
1047	 * here is to drop the funnel alltogether and regrab it afterwards
1048	 */
1049	err = IOFindBSDRoot(rootdevice, sizeof(rootdevice), &rootdev, &flags);
1050	if( err) {
1051		printf("setconf: IOFindBSDRoot returned an error (%d);"
1052			"setting rootdevice to 'sd0a'.\n", err); /* XXX DEBUG TEMP */
1053		rootdev = makedev( 6, 0 );
1054		strlcpy(rootdevice, "sd0a", sizeof(rootdevice));
1055		flags = 0;
1056	}
1057
1058#if NFSCLIENT
1059	if( flags & 1 ) {
1060		/* network device */
1061		mountroot = netboot_mountroot;
1062	} else {
1063#endif
1064		/* otherwise have vfs determine root filesystem */
1065		mountroot = NULL;
1066#if NFSCLIENT
1067	}
1068#endif
1069
1070}
1071
1072void
1073bsd_utaskbootstrap(void)
1074{
1075	thread_t thread;
1076	struct uthread *ut;
1077
1078	/*
1079	 * Clone the bootstrap process from the kernel process, without
1080	 * inheriting either task characteristics or memory from the kernel;
1081	 */
1082	thread = cloneproc(TASK_NULL, kernproc, FALSE, TRUE);
1083
1084	/* Hold the reference as it will be dropped during shutdown */
1085	initproc = proc_find(1);
1086#if __PROC_INTERNAL_DEBUG
1087	if (initproc == PROC_NULL)
1088		panic("bsd_utaskbootstrap: initproc not set\n");
1089#endif
1090	/*
1091	 * Since we aren't going back out the normal way to our parent,
1092	 * we have to drop the transition locks explicitly.
1093	 */
1094	proc_signalend(initproc, 0);
1095	proc_transend(initproc, 0);
1096
1097	ut = (struct uthread *)get_bsdthread_info(thread);
1098	ut->uu_sigmask = 0;
1099	act_set_astbsd(thread);
1100	(void) thread_resume(thread);
1101}
1102
1103static void
1104parse_bsd_args(void)
1105{
1106	char namep[16];
1107	int msgbuf;
1108
1109	if ( PE_parse_boot_argn("-s", namep, sizeof (namep)))
1110		boothowto |= RB_SINGLE;
1111
1112	if (PE_parse_boot_argn("-b", namep, sizeof (namep)))
1113		boothowto |= RB_NOBOOTRC;
1114
1115	if (PE_parse_boot_argn("-x", namep, sizeof (namep))) /* safe boot */
1116		boothowto |= RB_SAFEBOOT;
1117
1118	/* disable vnode_cache_is_authorized() by setting vnode_cache_defeat */
1119	if (PE_parse_boot_argn("-vnode_cache_defeat", namep, sizeof (namep)))
1120		bootarg_vnode_cache_defeat = 1;
1121
1122#if DEVELOPMENT || DEBUG
1123	if (PE_parse_boot_argn("-disable_aslr", namep, sizeof (namep)))
1124		bootarg_disable_aslr = 1;
1125#endif
1126
1127	PE_parse_boot_argn("ncl", &ncl, sizeof (ncl));
1128	if (PE_parse_boot_argn("nbuf", &max_nbuf_headers,
1129				sizeof (max_nbuf_headers))) {
1130		customnbuf = 1;
1131	}
1132#if !defined(SECURE_KERNEL)
1133	PE_parse_boot_argn("kmem", &setup_kmem, sizeof (setup_kmem));
1134#endif
1135
1136#if CONFIG_MACF
1137#if defined (__i386__) || defined (__x86_64__)
1138	PE_parse_boot_argn("policy_check", &policy_check_flags, sizeof (policy_check_flags));
1139#endif
1140#endif	/* CONFIG_MACF */
1141
1142	if (PE_parse_boot_argn("msgbuf", &msgbuf, sizeof (msgbuf))) {
1143		log_setsize(msgbuf);
1144	}
1145
1146	if (PE_parse_boot_argn("-novfscache", namep, sizeof(namep))) {
1147		nc_disabled = 1;
1148	}
1149
1150}
1151
1152void
1153bsd_exec_setup(int scale)
1154{
1155
1156	switch (scale) {
1157		case 0:
1158		case 1:
1159			bsd_simul_execs = BSD_SIMUL_EXECS;
1160			break;
1161		case 2:
1162		case 3:
1163			bsd_simul_execs = 65;
1164			break;
1165		case 4:
1166		case 5:
1167			bsd_simul_execs = 129;
1168			break;
1169		case 6:
1170		case 7:
1171			bsd_simul_execs = 257;
1172			break;
1173		default:
1174			bsd_simul_execs = 513;
1175			break;
1176
1177	}
1178	bsd_pageable_map_size = (bsd_simul_execs * BSD_PAGEABLE_SIZE_PER_EXEC);
1179}
1180
1181#if !NFSCLIENT
1182int
1183netboot_root(void);
1184
1185int
1186netboot_root(void)
1187{
1188	return(0);
1189}
1190#endif
1191