2 * Copyright (c) 2000-2004 Apple Computer, Inc. All rights reserved.
4 * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
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.
15 * Please obtain a copy of the License at
16 * http://www.opensource.apple.com/apsl/ and read it before using this file.
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.
26 * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
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.
37 * Redistribution and use in source and binary forms, with or without
38 * modification, are permitted provided that the following conditions
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.
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
65 * @(#)init_main.c 8.16 (Berkeley) 5/14/95
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.
76 #include <sys/param.h>
77 #include <sys/filedesc.h>
78 #include <sys/kernel.h>
79 #include <sys/mount_internal.h>
80 #include <sys/proc_internal.h>
81 #include <sys/kauth.h>
82 #include <sys/systm.h>
83 #include <sys/vnode_internal.h>
85 #include <sys/buf_internal.h>
86 #include <sys/clist.h>
89 #include <sys/systm.h>
92 #include <bsm/audit_kernel.h>
94 #include <sys/malloc.h>
95 #include <sys/dkstat.h>
97 #include <kern/startup.h>
98 #include <kern/thread.h>
99 #include <kern/task.h>
100 #include <kern/ast.h>
102 #include <mach/vm_param.h>
104 #include <vm/vm_map.h>
105 #include <vm/vm_kern.h>
107 #include <sys/ux_exception.h> /* for ux_exception_port */
109 #include <sys/reboot.h>
110 #include <mach/exception_types.h>
111 #include <dev/busvar.h> /* for pseudo_inits */
112 #include <sys/kdebug.h>
114 #include <mach/mach_types.h>
115 #include <mach/vm_prot.h>
116 #include <mach/semaphore.h>
117 #include <mach/sync_policy.h>
118 #include <kern/clock.h>
119 #include <mach/kern_return.h>
121 #include <mach/shared_memory_server.h>
122 #include <vm/vm_shared_memory_server.h>
124 #include <net/init.h>
126 #include <machine/exec.h>
128 extern int app_profile
; /* on/off switch for pre-heat cache */
131 "Copyright (c) 1982, 1986, 1989, 1991, 1993\n\t"
132 "The Regents of the University of California. "
133 "All rights reserved.\n\n";
135 extern void ux_handler();
137 /* Components of the first process -- never freed. */
139 struct session session0
;
141 struct filedesc filedesc0
;
142 struct plimit limit0
;
143 struct pstats pstats0
;
144 struct sigacts sigacts0
;
145 struct proc
*kernproc
, *initproc
;
153 /* Global variables to make pstat happy. We do swapping differently */
157 struct swdevt swdevt
[1];
159 dev_t rootdev
; /* device of the root */
160 dev_t dumpdev
; /* device to take dumps on */
161 long dumplo
; /* offset into dumpdev */
163 char hostname
[MAXHOSTNAMELEN
];
165 char domainname
[MAXDOMNAMELEN
];
168 struct exec_archhandler exec_archhandler_ppc
= {
169 .path
= "/usr/libexec/oah/translate",
172 struct exec_archhandler exec_archhandler_ppc
;
173 #endif /* __i386__ */
175 char rootdevice
[16]; /* hfs device names have at least 9 chars */
178 struct kmemstats kmemstats
[M_LAST
];
181 int lbolt
; /* awoken once a second */
182 struct vnode
*rootvp
;
183 int boothowto
= RB_DEBUG
;
185 #define BSD_PAGABLE_MAP_SIZE (16 * 512 * 1024)
186 vm_map_t bsd_pageable_map
;
188 semaphore_t execve_semaphore
;
191 extern int customnbuf
;
193 int parse_bsd_args(void);
194 extern int bsd_hardclockinit
;
195 extern task_t bsd_init_task
;
196 extern char init_task_failure_data
[];
197 extern void time_zone_slock_init(void);
198 static void process_name(char *, struct proc
*);
200 static void setconf(void);
202 funnel_t
*kernel_flock
;
204 extern void sysv_shm_lock_init(void);
205 extern void sysv_sem_lock_init(void);
206 extern void sysv_msg_lock_init(void);
207 extern void pshm_lock_init();
208 extern void psem_lock_init();
209 extern int maxprocperuid
;
211 /* kmem access not enabled by default; can be changed with boot-args */
214 extern void stackshot_lock_init();
217 * Initialization code.
218 * Called from cold start routine as
219 * soon as a stack and segmentation
220 * have been established.
223 * hand craft 0th process
224 * call all initialization routines
225 * hand craft 1st user process
229 * Sets the name for the given task.
236 int length
= strlen(s
);
239 length
>= sizeof(p
->p_comm
) ? sizeof(p
->p_comm
) :
243 /* To allow these values to be patched, they're globals here */
244 #include <machine/vmparam.h>
245 struct rlimit vm_initial_limit_stack
= { DFLSSIZ
, MAXSSIZ
};
246 struct rlimit vm_initial_limit_data
= { DFLDSIZ
, MAXDSIZ
};
247 struct rlimit vm_initial_limit_core
= { DFLCSIZ
, MAXCSIZ
};
249 extern thread_t
cloneproc(struct proc
*, int);
250 extern int (*mountroot
)(void);
251 extern int netboot_mountroot(); /* netboot.c */
252 extern int netboot_setup(struct proc
* p
);
254 lck_grp_t
* proc_lck_grp
;
255 lck_grp_attr_t
* proc_lck_grp_attr
;
256 lck_attr_t
* proc_lck_attr
;
258 /* hook called after root is mounted XXX temporary hack */
259 void (*mountroot_post_hook
)(void);
262 * This function is called very early on in the Mach startup, from the
263 * function start_kernel_threads() in osfmk/kern/startup.c. It's called
264 * in the context of the current (startup) task using a call to the
265 * function kernel_thread_create() to jump into start_kernel_threads().
266 * Internally, kernel_thread_create() calls thread_create_internal(),
267 * which calls uthread_alloc(). The function of uthread_alloc() is
268 * normally to allocate a uthread structure, and fill out the uu_sigmask,
269 * uu_act, and uu_ucred fields. It skips filling these out in the case
270 * of the "task" being "kernel_task", because the order of operation is
271 * inverted. To account for that, we need to manually fill in at least
272 * the uu_cred field so that the uthread structure can be used like any
278 register struct proc
*p
;
280 extern kauth_cred_t rootcred
;
285 struct vfs_context context
;
286 void lightning_bolt(void );
288 boolean_t funnel_state
;
289 struct ucred temp_cred
;
290 extern void file_lock_init(void);
292 kernel_flock
= funnel_alloc(KERNEL_FUNNEL
);
293 if (kernel_flock
== (funnel_t
*)0 ) {
294 panic("bsd_init: Failed to allocate kernel funnel");
297 funnel_state
= thread_funnel_set(kernel_flock
, TRUE
);
305 /* Initialize the uthread zone */
306 //uthread_zone_init(); /* XXX redundant: previous uthread_alloc() */
308 /* Initialize kauth subsystem before instancing the first credential */
311 /* Initialize process and pgrp structures. */
318 /* kernel_task->proc = kernproc; */
319 set_bsdtask_info(kernel_task
,(void *)p
);
322 /* give kernproc a name */
323 process_name("kernel_task", p
);
325 /* allocate proc lock group attribute and group */
326 proc_lck_grp_attr
= lck_grp_attr_alloc_init();
328 proc_lck_grp
= lck_grp_alloc_init("proc", proc_lck_grp_attr
);
330 /* Allocate proc lock attribute */
331 proc_lck_attr
= lck_attr_alloc_init();
333 lck_mtx_init(&p
->p_mlock
, proc_lck_grp
, proc_lck_attr
);
334 lck_mtx_init(&p
->p_fdmlock
, proc_lck_grp
, proc_lck_attr
);
336 if (current_task() != kernel_task
)
337 printf("bsd_init: We have a problem, "
338 "current task is not kernel task\n");
340 ut
= (uthread_t
)get_bsdthread_info(current_thread());
345 LIST_INSERT_HEAD(&allproc
, p
, p_list
);
347 LIST_INSERT_HEAD(PGRPHASH(0), &pgrp0
, pg_hash
);
348 LIST_INIT(&pgrp0
.pg_members
);
349 LIST_INSERT_HEAD(&pgrp0
.pg_members
, p
, p_pglist
);
351 pgrp0
.pg_session
= &session0
;
352 session0
.s_count
= 1;
353 session0
.s_leader
= p
;
355 p
->task
= kernel_task
;
358 p
->p_flag
= P_SYSTEM
;
361 lockinit(&p
->signal_lock
, PVM
, "signal", 0, 0);
362 TAILQ_INIT(&p
->p_uthlist
);
364 p
->sigwait_thread
= THREAD_NULL
;
365 p
->exit_thread
= THREAD_NULL
;
368 * Create credential. This also Initializes the audit information.
369 * XXX It is not clear what the initial values should be for audit ID,
370 * XXX session ID, etc..
372 bzero(&temp_cred
, sizeof(temp_cred
));
373 temp_cred
.cr_ngroups
= 1;
375 p
->p_ucred
= kauth_cred_create(&temp_cred
);
377 /* give the (already exisiting) initial thread a reference on it */
378 kauth_cred_ref(p
->p_ucred
);
379 ut
->uu_ucred
= p
->p_ucred
;
381 TAILQ_INIT(&p
->aio_activeq
);
382 TAILQ_INIT(&p
->aio_doneq
);
383 p
->aio_active_count
= 0;
384 p
->aio_done_count
= 0;
388 /* Create the file descriptor table. */
389 filedesc0
.fd_refcnt
= 1+1; /* +1 so shutdown will not _FREE_ZONE */
390 p
->p_fd
= &filedesc0
;
391 filedesc0
.fd_cmask
= cmask
;
392 filedesc0
.fd_knlistsize
= -1;
393 filedesc0
.fd_knlist
= NULL
;
394 filedesc0
.fd_knhash
= NULL
;
395 filedesc0
.fd_knhashmask
= 0;
397 /* Create the limits structures. */
398 p
->p_limit
= &limit0
;
399 for (i
= 0; i
< sizeof(p
->p_rlimit
)/sizeof(p
->p_rlimit
[0]); i
++)
400 limit0
.pl_rlimit
[i
].rlim_cur
=
401 limit0
.pl_rlimit
[i
].rlim_max
= RLIM_INFINITY
;
402 limit0
.pl_rlimit
[RLIMIT_NOFILE
].rlim_cur
= NOFILE
;
403 limit0
.pl_rlimit
[RLIMIT_NPROC
].rlim_cur
= maxprocperuid
;
404 limit0
.pl_rlimit
[RLIMIT_NPROC
].rlim_max
= maxproc
;
405 limit0
.pl_rlimit
[RLIMIT_STACK
] = vm_initial_limit_stack
;
406 limit0
.pl_rlimit
[RLIMIT_DATA
] = vm_initial_limit_data
;
407 limit0
.pl_rlimit
[RLIMIT_CORE
] = vm_initial_limit_core
;
410 p
->p_stats
= &pstats0
;
411 p
->p_sigacts
= &sigacts0
;
414 * Charge root for two processes: init and mach_init.
416 (void)chgproccnt(0, 1);
419 * Allocate a kernel submap for pageable memory
420 * for temporary copying (execve()).
425 ret
= kmem_suballoc(kernel_map
,
427 (vm_size_t
)BSD_PAGABLE_MAP_SIZE
,
431 if (ret
!= KERN_SUCCESS
)
432 panic("bsd_init: Failed to allocate bsd pageable map");
436 * Initialize buffers and hash links for buffers
438 * SIDE EFFECT: Starts a thread for bcleanbuf_thread(), so must
439 * happen after a credential has been associated with
444 /* Initialize the execve() semaphore */
445 ret
= semaphore_create(kernel_task
, &execve_semaphore
,
446 SYNC_POLICY_FIFO
, (BSD_PAGABLE_MAP_SIZE
/ NCARGS
));
447 if (ret
!= KERN_SUCCESS
)
448 panic("bsd_init: Failed to create execve semaphore");
451 * Initialize the calendar.
457 /* Initialize the file systems. */
460 /* Initialize mbuf's. */
464 * Initializes security event auditing.
465 * XXX: Should/could this occur later?
469 /* Initialize kqueues */
472 /* Initialize for async IO */
475 /* Initialize pipes */
478 /* Initialize SysV shm subsystem locks; the subsystem proper is
479 * initialized through a sysctl.
481 sysv_shm_lock_init();
482 sysv_sem_lock_init();
483 sysv_msg_lock_init();
487 /* POSIX Shm and Sem */
490 time_zone_slock_init();
492 /* Stack snapshot facility lock */
493 stackshot_lock_init();
495 * Initialize protocols. Block reception of incoming packets
496 * until everything is ready.
498 sysctl_register_fixed();
505 p
->p_fd
->fd_cdir
= NULL
;
506 p
->p_fd
->fd_rdir
= NULL
;
509 /* Initialize kernel profiling. */
513 /* kick off timeout driven events by calling first time */
514 thread_wakeup(&lbolt
);
515 timeout((void (*)(void *))lightning_bolt
, 0, hz
);
520 * We attach the loopback interface *way* down here to ensure
521 * it happens after autoconf(), otherwise it becomes the
522 * "primary" interface.
526 loopattach(); /* XXX */
529 /* Register the built-in dlil ethernet interface family */
532 /* Call any kext code that wants to run just after network init */
535 vnode_pager_bootstrap();
537 /* XXX Hack for early debug stop */
538 printf("\nabout to sleep for 10 seconds\n");
539 IOSleep( 10 * 1000 );
540 /* Debugger("hello"); */
545 /* Mount the root file system. */
550 bsd_hardclockinit
= -1; /* start ticking */
552 if (0 == (err
= vfs_mountroot()))
555 if (mountroot
== netboot_mountroot
) {
556 printf("cannot mount network root, errno = %d\n", err
);
558 if (0 == (err
= vfs_mountroot()))
562 printf("cannot mount root, errno = %d\n", err
);
563 boothowto
|= RB_ASKNAME
;
567 context
.vc_ucred
= p
->p_ucred
;
568 mountlist
.tqh_first
->mnt_flag
|= MNT_ROOTFS
;
570 /* Get the vnode for '/'. Set fdp->fd_fd.fd_cdir to reference it. */
571 if (VFS_ROOT(mountlist
.tqh_first
, &rootvnode
, &context
))
572 panic("bsd_init: cannot find root vnode");
573 rootvnode
->v_flag
|= VROOT
;
574 (void)vnode_ref(rootvnode
);
575 (void)vnode_put(rootvnode
);
576 filedesc0
.fd_cdir
= rootvnode
;
579 if (mountroot
== netboot_mountroot
) {
581 /* post mount setup */
582 if (err
= netboot_setup(p
)) {
583 panic("bsd_init: NetBoot could not find root, %d", err
);
589 microtime(&p
->p_stats
->p_start
);
590 p
->p_rtime
.tv_sec
= p
->p_rtime
.tv_usec
= 0;
594 extern void devfs_kernel_mount(char * str
);
596 devfs_kernel_mount("/dev");
600 /* Initialize signal state for process 0. */
603 bsd_utaskbootstrap();
606 // this should be done after the root filesystem is mounted
607 error
= set_archhandler(p
, CPU_TYPE_POWERPC
);
609 exec_archhandler_ppc
.path
[0] = 0;
612 /* invoke post-root-mount hook */
613 if (mountroot_post_hook
!= NULL
)
614 mountroot_post_hook();
616 (void) thread_funnel_set(kernel_flock
, funnel_state
);
619 /* Called with kernel funnel held */
623 struct proc
*p
= current_proc();
627 shared_region_mapping_t system_region
;
629 process_name("init", p
);
633 th_act
= current_thread();
634 (void) host_set_exception_ports(host_priv_self(),
635 EXC_MASK_ALL
& ~(EXC_MASK_SYSCALL
|
636 EXC_MASK_MACH_SYSCALL
|
639 EXCEPTION_DEFAULT
, 0);
641 (void) task_set_exception_ports(get_threadtask(th_act
),
642 EXC_MASK_ALL
& ~(EXC_MASK_SYSCALL
|
643 EXC_MASK_MACH_SYSCALL
|
646 EXCEPTION_DEFAULT
, 0);
651 ut
= (uthread_t
)get_bsdthread_info(th_act
);
653 bsd_hardclockinit
= 1; /* Start bsd hardclock */
654 bsd_init_task
= get_threadtask(th_act
);
655 init_task_failure_data
[0] = 0;
656 system_region
= lookup_default_shared_region(ENV_DEFAULT_ROOT
, cpu_type());
657 if (system_region
== NULL
) {
658 shared_file_boot_time_init(ENV_DEFAULT_ROOT
, cpu_type());
660 vm_set_shared_region(get_threadtask(th_act
), system_region
);
662 load_init_program(p
);
663 /* turn on app-profiling i.e. pre-heating */
671 boolean_t funnel_state
;
672 extern void klogwakeup(void);
674 funnel_state
= thread_funnel_set(kernel_flock
, TRUE
);
676 thread_wakeup(&lbolt
);
677 timeout(lightning_bolt
,0,hz
);
680 (void) thread_funnel_set(kernel_flock
, FALSE
);
685 extern kern_return_t
IOKitBSDInit( void );
690 * Early startup for bsd pseudodevices.
693 struct pseudo_init
*pi
;
695 for (pi
= pseudo_inits
; pi
->ps_func
; pi
++)
696 (*pi
->ps_func
) (pi
->ps_count
);
699 return( IOKitBSDInit());
703 #include <sys/disklabel.h> /* for MAXPARTITIONS */
708 extern kern_return_t
IOFindBSDRoot( char * rootName
,
709 dev_t
* root
, u_int32_t
* flags
);
714 * calls into IOKit can generate networking registrations
715 * which needs to be under network funnel. Right thing to do
716 * here is to drop the funnel alltogether and regrab it afterwards
718 thread_funnel_set(kernel_flock
, FALSE
);
719 err
= IOFindBSDRoot( rootdevice
, &rootdev
, &flags
);
720 thread_funnel_set(kernel_flock
, TRUE
);
722 printf("setconf: IOFindBSDRoot returned an error (%d);"
723 "setting rootdevice to 'sd0a'.\n", err
); /* XXX DEBUG TEMP */
724 rootdev
= makedev( 6, 0 );
725 strcpy( rootdevice
, "sd0a" );
732 mountroot
= netboot_mountroot
;
735 /* otherwise have vfs determine root filesystem */
748 th_act
= cloneproc(kernproc
, 0);
750 /* Set the launch time for init */
751 microtime(&initproc
->p_stats
->p_start
);
753 ut
= (struct uthread
*)get_bsdthread_info(th_act
);
755 act_set_astbsd(th_act
);
756 (void) thread_resume(th_act
);
761 extern char init_args
[];
763 extern int boothowto
;
769 if (PE_parse_boot_arg("-s", namep
)) {
770 boothowto
|= RB_SINGLE
;
771 len
= strlen(init_args
);
773 strcat(init_args
," -s");
775 strcat(init_args
,"-s");
778 if (PE_parse_boot_arg("-b", namep
)) {
779 boothowto
|= RB_NOBOOTRC
;
780 len
= strlen(init_args
);
782 strcat(init_args
," -b");
784 strcat(init_args
,"-b");
787 if (PE_parse_boot_arg("-F", namep
)) {
788 len
= strlen(init_args
);
790 strcat(init_args
," -F");
792 strcat(init_args
,"-F");
795 if (PE_parse_boot_arg("-v", namep
)) {
796 len
= strlen(init_args
);
798 strcat(init_args
," -v");
800 strcat(init_args
,"-v");
803 if (PE_parse_boot_arg("-x", namep
)) { /* safe boot */
804 len
= strlen(init_args
);
806 strcat(init_args
," -x");
808 strcat(init_args
,"-x");
811 if (PE_parse_boot_arg("-d", namep
)) {
812 len
= strlen(init_args
);
814 strcat(init_args
," -d");
816 strcat(init_args
,"-d");
819 PE_parse_boot_arg("srv", &srv
);
820 PE_parse_boot_arg("ncl", &ncl
);
821 if (PE_parse_boot_arg("nbuf", &max_nbuf_headers
))
823 PE_parse_boot_arg("kmem", &setup_kmem
);