bsd/kern/bsd_init.c

   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/content_filter.h>         /* for cfil_init() */
 159 #include <net/necp.h>                   /* for necp_init() */
 160 #include <net/packet_mangler.h>         /* for pkt_mnglr_init() */
 161 #include <net/if_utun.h>                /* for utun_register_control() */
 162 #include <net/if_ipsec.h>       /* for ipsec_register_control() */
 163 #include <net/net_str_id.h>             /* for net_str_id_init() */
 164 #include <net/netsrc.h>                 /* for netsrc_init() */
 165 #include <net/ntstat.h>                 /* for nstat_init() */
 166 #include <netinet/tcp_cc.h>                     /* for tcp_cc_init() */
 167 #include <kern/assert.h>                /* for assert() */
 168 #include <sys/kern_overrides.h>         /* for init_system_override() */
 169
 170 #include <net/init.h>
 171
 172 #if CONFIG_MACF
 173 #include <security/mac_framework.h>
 174 #include <security/mac_internal.h>      /* mac_init_bsd() */
 175 #include <security/mac_mach_internal.h> /* mac_update_task_label() */
 176 #endif
 177
 178 #include <machine/exec.h>
 179
 180 #if NFSCLIENT
 181 #include <sys/netboot.h>
 182 #endif
 183
 184 #if CONFIG_IMAGEBOOT
 185 #include <sys/imageboot.h>
 186 #endif
 187
 188 #if PFLOG
 189 #include <net/if_pflog.h>
 190 #endif
 191
 192 #include <pexpert/pexpert.h>
 193 #include <machine/pal_routines.h>
 194 #include <console/video_console.h>
 195
 196 void * get_user_regs(thread_t);         /* XXX kludge for <machine/thread.h> */
 197 void IOKitInitializeTime(void);         /* XXX */
 198 void IOSleep(unsigned int);             /* XXX */
 199 void loopattach(void);                  /* XXX */
 200
 201 const char    copyright[] =
 202 "Copyright (c) 1982, 1986, 1989, 1991, 1993\n\t"
 203 "The Regents of the University of California. "
 204 "All rights reserved.\n\n";
 205
 206 /* Components of the first process -- never freed. */
 207 struct  proc proc0;
 208 struct  session session0;
 209 struct  pgrp pgrp0;
 210 struct  filedesc filedesc0;
 211 struct  plimit limit0;
 212 struct  pstats pstats0;
 213 struct  sigacts sigacts0;
 214 proc_t kernproc;
 215 proc_t initproc;
 216
 217 long tk_cancc;
 218 long tk_nin;
 219 long tk_nout;
 220 long tk_rawcc;
 221
 222 int lock_trace = 0;
 223 /* Global variables to make pstat happy. We do swapping differently */
 224 int nswdev, nswap;
 225 int nswapmap;
 226 void *swapmap;
 227 struct swdevt swdevt[1];
 228
 229 dev_t   rootdev;                /* device of the root */
 230 dev_t   dumpdev;                /* device to take dumps on */
 231 long    dumplo;                 /* offset into dumpdev */
 232 long    hostid;
 233 char    hostname[MAXHOSTNAMELEN];
 234 int             hostnamelen;
 235 char    domainname[MAXDOMNAMELEN];
 236 int             domainnamelen;
 237
 238 char rootdevice[16];    /* hfs device names have at least 9 chars */
 239
 240 #if  KMEMSTATS
 241 struct  kmemstats kmemstats[M_LAST];
 242 #endif
 243
 244 struct  vnode *rootvp;
 245 int boothowto = RB_DEBUG;
 246
 247 extern kern_return_t IOFindBSDRoot(char *, unsigned int, dev_t *, u_int32_t *);
 248 extern void IOSecureBSDRoot(const char * rootName);
 249 extern kern_return_t IOKitBSDInit(void );
 250 extern void kminit(void);
 251 extern void file_lock_init(void);
 252 extern void kmeminit(void);
 253 extern void bsd_bufferinit(void);
 254 extern void throttle_init(void);
 255 extern void macx_init(void);
 256 extern void acct_init(void);
 257
 258 extern int serverperfmode;
 259 extern int ncl;
 260
 261 vm_map_t        bsd_pageable_map;
 262 vm_map_t        mb_map;
 263
 264 static  int bsd_simul_execs;
 265 static int bsd_pageable_map_size;
 266 __private_extern__ int execargs_cache_size = 0;
 267 __private_extern__ int execargs_free_count = 0;
 268 __private_extern__ vm_offset_t * execargs_cache = NULL;
 269
 270 void bsd_exec_setup(int);
 271
 272 __private_extern__ int bootarg_vnode_cache_defeat = 0;
 273
 274 /*
 275  * Prevent kernel-based ASLR from being used, for testing.
 276  */
 277 #if DEVELOPMENT || DEBUG
 278 __private_extern__ int bootarg_disable_aslr = 0;
 279 #endif
 280
 281 int     cmask = CMASK;
 282 extern int customnbuf;
 283
 284 void bsd_init(void);
 285 kern_return_t bsd_autoconf(void);
 286 void bsd_utaskbootstrap(void);
 287
 288 static void parse_bsd_args(void);
 289 extern task_t bsd_init_task;
 290 extern boolean_t init_task_died;
 291 extern char    init_task_failure_data[];
 292 #if CONFIG_DEV_KMEM
 293 extern void dev_kmem_init(void);
 294 #endif
 295 extern void time_zone_slock_init(void);
 296 extern void select_wait_queue_init(void);
 297 static void process_name(const char *, proc_t);
 298
 299 static void setconf(void);
 300
 301 #if SYSV_SHM
 302 extern void sysv_shm_lock_init(void);
 303 #endif
 304 #if SYSV_SEM
 305 extern void sysv_sem_lock_init(void);
 306 #endif
 307 #if SYSV_MSG
 308 extern void sysv_msg_lock_init(void);
 309 #endif
 310
 311 #if CONFIG_MACF
 312 #if defined (__i386__) || defined (__x86_64__)
 313 /* MACF policy_check configuration flags; see policy_check.c for details */
 314 int policy_check_flags = 0;
 315
 316 extern int check_policy_init(int);
 317 #endif
 318 #endif  /* CONFIG_MACF */
 319
 320 /* If we are using CONFIG_DTRACE */
 321 #if CONFIG_DTRACE
 322         extern void dtrace_postinit(void);
 323 #endif
 324
 325 /*
 326  * Initialization code.
 327  * Called from cold start routine as
 328  * soon as a stack and segmentation
 329  * have been established.
 330  * Functions:
 331  *      turn on clock
 332  *      hand craft 0th process
 333  *      call all initialization routines
 334  *  hand craft 1st user process
 335  */
 336
 337 /*
 338  *      Sets the name for the given task.
 339  */
 340 static void
 341 process_name(const char *s, proc_t p)
 342 {
 343         size_t length = strlen(s);
 344
 345         bcopy(s, p->p_comm,
 346                 length >= sizeof(p->p_comm) ? sizeof(p->p_comm) :
 347                         length + 1);
 348 }
 349
 350 /* To allow these values to be patched, they're globals here */
 351 #include <machine/vmparam.h>
 352 struct rlimit vm_initial_limit_stack = { DFLSSIZ, MAXSSIZ - PAGE_MAX_SIZE };
 353 struct rlimit vm_initial_limit_data = { DFLDSIZ, MAXDSIZ };
 354 struct rlimit vm_initial_limit_core = { DFLCSIZ, MAXCSIZ };
 355
 356 extern thread_t cloneproc(task_t, coalition_t, proc_t, int, int);
 357 extern int      (*mountroot)(void);
 358
 359 lck_grp_t * proc_lck_grp;
 360 lck_grp_t * proc_slock_grp;
 361 lck_grp_t * proc_fdmlock_grp;
 362 lck_grp_t * proc_mlock_grp;
 363 lck_grp_attr_t * proc_lck_grp_attr;
 364 lck_attr_t * proc_lck_attr;
 365 lck_mtx_t * proc_list_mlock;
 366 lck_mtx_t * proc_klist_mlock;
 367
 368 extern lck_mtx_t * execargs_cache_lock;
 369
 370 /* hook called after root is mounted XXX temporary hack */
 371 void (*mountroot_post_hook)(void);
 372 void (*unmountroot_pre_hook)(void);
 373
 374 /*
 375  * This function is called very early on in the Mach startup, from the
 376  * function start_kernel_threads() in osfmk/kern/startup.c.  It's called
 377  * in the context of the current (startup) task using a call to the
 378  * function kernel_thread_create() to jump into start_kernel_threads().
 379  * Internally, kernel_thread_create() calls thread_create_internal(),
 380  * which calls uthread_alloc().  The function of uthread_alloc() is
 381  * normally to allocate a uthread structure, and fill out the uu_sigmask,
 382  * uu_context fields.  It skips filling these out in the case of the "task"
 383  * being "kernel_task", because the order of operation is inverted.  To
 384  * account for that, we need to manually fill in at least the contents
 385  * of the uu_context.vc_ucred field so that the uthread structure can be
 386  * used like any other.
 387  */
 388
 389 void
 390 bsd_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         printf(copyright);
 407
 408         bsd_init_kprintf("calling kmeminit\n");
 409         kmeminit();
 410
 411         bsd_init_kprintf("calling parse_bsd_args\n");
 412         parse_bsd_args();
 413
 414 #if CONFIG_DEV_KMEM
 415         bsd_init_kprintf("calling dev_kmem_init\n");
 416         dev_kmem_init();
 417 #endif
 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 #endif
 603
 604         /* Create the file descriptor table. */
 605         kernproc->p_fd = &filedesc0;
 606         filedesc0.fd_cmask = cmask;
 607         filedesc0.fd_knlistsize = -1;
 608         filedesc0.fd_knlist = NULL;
 609         filedesc0.fd_knhash = NULL;
 610         filedesc0.fd_knhashmask = 0;
 611
 612         /* Create the limits structures. */
 613         kernproc->p_limit = &limit0;
 614         for (i = 0; i < sizeof(kernproc->p_rlimit)/sizeof(kernproc->p_rlimit[0]); i++)
 615                 limit0.pl_rlimit[i].rlim_cur =
 616                         limit0.pl_rlimit[i].rlim_max = RLIM_INFINITY;
 617         limit0.pl_rlimit[RLIMIT_NOFILE].rlim_cur = NOFILE;
 618         limit0.pl_rlimit[RLIMIT_NPROC].rlim_cur = maxprocperuid;
 619         limit0.pl_rlimit[RLIMIT_NPROC].rlim_max = maxproc;
 620         limit0.pl_rlimit[RLIMIT_STACK] = vm_initial_limit_stack;
 621         limit0.pl_rlimit[RLIMIT_DATA] = vm_initial_limit_data;
 622         limit0.pl_rlimit[RLIMIT_CORE] = vm_initial_limit_core;
 623         limit0.pl_refcnt = 1;
 624
 625         kernproc->p_stats = &pstats0;
 626         kernproc->p_sigacts = &sigacts0;
 627
 628         /*
 629          * Charge root for one process: launchd.
 630          */
 631         bsd_init_kprintf("calling chgproccnt\n");
 632         (void)chgproccnt(0, 1);
 633
 634         /*
 635          *      Allocate a kernel submap for pageable memory
 636          *      for temporary copying (execve()).
 637          */
 638         {
 639                 vm_offset_t     minimum;
 640
 641                 bsd_init_kprintf("calling kmem_suballoc\n");
 642                 assert(bsd_pageable_map_size != 0);
 643                 ret = kmem_suballoc(kernel_map,
 644                                 &minimum,
 645                                 (vm_size_t)bsd_pageable_map_size,
 646                                 TRUE,
 647                                 VM_FLAGS_ANYWHERE,
 648                                 &bsd_pageable_map);
 649                 if (ret != KERN_SUCCESS)
 650                         panic("bsd_init: Failed to allocate bsd pageable map");
 651         }
 652
 653         /*
 654          * Initialize buffers and hash links for buffers
 655          *
 656          * SIDE EFFECT: Starts a thread for bcleanbuf_thread(), so must
 657          *              happen after a credential has been associated with
 658          *              the kernel task.
 659          */
 660         bsd_init_kprintf("calling bsd_bufferinit\n");
 661         bsd_bufferinit();
 662
 663         /* Initialize the execve() semaphore */
 664         bsd_init_kprintf("calling semaphore_create\n");
 665
 666         if (ret != KERN_SUCCESS)
 667                 panic("bsd_init: Failed to create execve semaphore");
 668
 669         /*
 670          * Initialize the calendar.
 671          */
 672         bsd_init_kprintf("calling IOKitInitializeTime\n");
 673         IOKitInitializeTime();
 674
 675         bsd_init_kprintf("calling ubc_init\n");
 676         ubc_init();
 677
 678         /*
 679          * Initialize device-switches.
 680          */
 681         bsd_init_kprintf("calling devsw_init() \n");
 682         devsw_init();
 683
 684         /* Initialize the file systems. */
 685         bsd_init_kprintf("calling vfsinit\n");
 686         vfsinit();
 687
 688 #if CONFIG_PROC_UUID_POLICY
 689         /* Initial proc_uuid_policy subsystem */
 690         bsd_init_kprintf("calling proc_uuid_policy_init()\n");
 691         proc_uuid_policy_init();
 692 #endif
 693
 694 #if SOCKETS
 695         /* Initialize per-CPU cache allocator */
 696         mcache_init();
 697
 698         /* Initialize mbuf's. */
 699         bsd_init_kprintf("calling mbinit\n");
 700         mbinit();
 701         net_str_id_init(); /* for mbuf tags */
 702 #endif /* SOCKETS */
 703
 704         /*
 705          * Initializes security event auditing.
 706          * XXX: Should/could this occur later?
 707          */
 708 #if CONFIG_AUDIT
 709         bsd_init_kprintf("calling audit_init\n");
 710         audit_init();
 711 #endif
 712
 713         /* Initialize kqueues */
 714         bsd_init_kprintf("calling knote_init\n");
 715         knote_init();
 716
 717         /* Initialize for async IO */
 718         bsd_init_kprintf("calling aio_init\n");
 719         aio_init();
 720
 721         /* Initialize pipes */
 722         bsd_init_kprintf("calling pipeinit\n");
 723         pipeinit();
 724
 725         /* Initialize SysV shm subsystem locks; the subsystem proper is
 726          * initialized through a sysctl.
 727          */
 728 #if SYSV_SHM
 729         bsd_init_kprintf("calling sysv_shm_lock_init\n");
 730         sysv_shm_lock_init();
 731 #endif
 732 #if SYSV_SEM
 733         bsd_init_kprintf("calling sysv_sem_lock_init\n");
 734         sysv_sem_lock_init();
 735 #endif
 736 #if SYSV_MSG
 737         bsd_init_kprintf("sysv_msg_lock_init\n");
 738         sysv_msg_lock_init();
 739 #endif
 740         bsd_init_kprintf("calling pshm_lock_init\n");
 741         pshm_lock_init();
 742         bsd_init_kprintf("calling psem_lock_init\n");
 743         psem_lock_init();
 744
 745         pthread_init();
 746         /* POSIX Shm and Sem */
 747         bsd_init_kprintf("calling pshm_cache_init\n");
 748         pshm_cache_init();
 749         bsd_init_kprintf("calling psem_cache_init\n");
 750         psem_cache_init();
 751         bsd_init_kprintf("calling time_zone_slock_init\n");
 752         time_zone_slock_init();
 753         bsd_init_kprintf("calling select_wait_queue_init\n");
 754         select_wait_queue_init();
 755
 756         /*
 757          * Initialize protocols.  Block reception of incoming packets
 758          * until everything is ready.
 759          */
 760         bsd_init_kprintf("calling sysctl_register_fixed\n");
 761         sysctl_register_fixed();
 762         bsd_init_kprintf("calling sysctl_mib_init\n");
 763         sysctl_mib_init();
 764 #if NETWORKING
 765         bsd_init_kprintf("calling dlil_init\n");
 766         dlil_init();
 767         bsd_init_kprintf("calling proto_kpi_init\n");
 768         proto_kpi_init();
 769 #endif /* NETWORKING */
 770 #if SOCKETS
 771         bsd_init_kprintf("calling socketinit\n");
 772         socketinit();
 773         bsd_init_kprintf("calling domaininit\n");
 774         domaininit();
 775         iptap_init();
 776 #if FLOW_DIVERT
 777         flow_divert_init();
 778 #endif  /* FLOW_DIVERT */
 779 #endif /* SOCKETS */
 780
 781         kernproc->p_fd->fd_cdir = NULL;
 782         kernproc->p_fd->fd_rdir = NULL;
 783
 784 #if CONFIG_FREEZE
 785 #ifndef CONFIG_MEMORYSTATUS
 786     #error "CONFIG_FREEZE defined without matching CONFIG_MEMORYSTATUS"
 787 #endif
 788         /* Initialise background freezing */
 789         bsd_init_kprintf("calling memorystatus_freeze_init\n");
 790         memorystatus_freeze_init();
 791 #endif
 792
 793 #if CONFIG_MEMORYSTATUS
 794         /* Initialize kernel memory status notifications */
 795         bsd_init_kprintf("calling memorystatus_init\n");
 796         memorystatus_init();
 797 #endif /* CONFIG_MEMORYSTATUS */
 798
 799         bsd_init_kprintf("calling macx_init\n");
 800         macx_init();
 801
 802         bsd_init_kprintf("calling acct_init\n");
 803         acct_init();
 804
 805 #ifdef GPROF
 806         /* Initialize kernel profiling. */
 807         kmstartup();
 808 #endif
 809
 810         bsd_init_kprintf("calling bsd_autoconf\n");
 811         bsd_autoconf();
 812
 813 #if CONFIG_DTRACE
 814         dtrace_postinit();
 815 #endif
 816
 817         /*
 818          * We attach the loopback interface *way* down here to ensure
 819          * it happens after autoconf(), otherwise it becomes the
 820          * "primary" interface.
 821          */
 822 #include <loop.h>
 823 #if NLOOP > 0
 824         bsd_init_kprintf("calling loopattach\n");
 825         loopattach();                   /* XXX */
 826 #endif
 827 #if NGIF
 828         /* Initialize gif interface (after lo0) */
 829         gif_init();
 830 #endif
 831
 832 #if PFLOG
 833         /* Initialize packet filter log interface */
 834         pfloginit();
 835 #endif /* PFLOG */
 836
 837 #if NETHER > 0
 838         /* Register the built-in dlil ethernet interface family */
 839         bsd_init_kprintf("calling ether_family_init\n");
 840         ether_family_init();
 841 #endif /* ETHER */
 842
 843 #if NETWORKING
 844         /* Call any kext code that wants to run just after network init */
 845         bsd_init_kprintf("calling net_init_run\n");
 846         net_init_run();
 847
 848 #if CONTENT_FILTER
 849         cfil_init();
 850 #endif
 851
 852 #if PACKET_MANGLER
 853         pkt_mnglr_init();
 854 #endif
 855
 856 #if NECP
 857         /* Initialize Network Extension Control Policies */
 858         necp_init();
 859 #endif
 860
 861         /* register user tunnel kernel control handler */
 862         utun_register_control();
 863 #if IPSEC
 864         ipsec_register_control();
 865 #endif /* IPSEC */
 866         netsrc_init();
 867         nstat_init();
 868         tcp_cc_init();
 869 #endif /* NETWORKING */
 870
 871         bsd_init_kprintf("calling vnode_pager_bootstrap\n");
 872         vnode_pager_bootstrap();
 873
 874         bsd_init_kprintf("calling inittodr\n");
 875         inittodr(0);
 876
 877         /* Mount the root file system. */
 878         while( TRUE) {
 879                 int err;
 880
 881                 bsd_init_kprintf("calling setconf\n");
 882                 setconf();
 883 #if NFSCLIENT
 884                 netboot = (mountroot == netboot_mountroot);
 885 #endif
 886
 887                 bsd_init_kprintf("vfs_mountroot\n");
 888                 if (0 == (err = vfs_mountroot()))
 889                         break;
 890                 rootdevice[0] = '\0';
 891 #if NFSCLIENT
 892                 if (netboot) {
 893                         PE_display_icon( 0, "noroot");  /* XXX a netboot-specific icon would be nicer */
 894                         vc_progress_set(FALSE, 0);
 895                         for (i=1; 1; i*=2) {
 896                                 printf("bsd_init: failed to mount network root, error %d, %s\n",
 897                                         err, PE_boot_args());
 898                                 printf("We are hanging here...\n");
 899                                 IOSleep(i*60*1000);
 900                         }
 901                         /*NOTREACHED*/
 902                 }
 903 #endif
 904                 printf("cannot mount root, errno = %d\n", err);
 905                 boothowto |= RB_ASKNAME;
 906         }
 907
 908         IOSecureBSDRoot(rootdevice);
 909
 910         context.vc_thread = current_thread();
 911         context.vc_ucred = kernproc->p_ucred;
 912         mountlist.tqh_first->mnt_flag |= MNT_ROOTFS;
 913
 914         bsd_init_kprintf("calling VFS_ROOT\n");
 915         /* Get the vnode for '/'.  Set fdp->fd_fd.fd_cdir to reference it. */
 916         if (VFS_ROOT(mountlist.tqh_first, &rootvnode, &context))
 917                 panic("bsd_init: cannot find root vnode: %s", PE_boot_args());
 918         rootvnode->v_flag |= VROOT;
 919         (void)vnode_ref(rootvnode);
 920         (void)vnode_put(rootvnode);
 921         filedesc0.fd_cdir = rootvnode;
 922
 923 #if NFSCLIENT
 924         if (netboot) {
 925                 int err;
 926
 927                 netboot = TRUE;
 928                 /* post mount setup */
 929                 if ((err = netboot_setup()) != 0) {
 930                         PE_display_icon( 0, "noroot");  /* XXX a netboot-specific icon would be nicer */
 931                         vc_progress_set(FALSE, 0);
 932                         for (i=1; 1; i*=2) {
 933                                 printf("bsd_init: NetBoot could not find root, error %d: %s\n",
 934                                         err, PE_boot_args());
 935                                 printf("We are hanging here...\n");
 936                                 IOSleep(i*60*1000);
 937                         }
 938                         /*NOTREACHED*/
 939                 }
 940         }
 941 #endif
 942
 943
 944 #if CONFIG_IMAGEBOOT
 945         /*
 946          * See if a system disk image is present. If so, mount it and
 947          * switch the root vnode to point to it
 948          */
 949         if (netboot == FALSE && imageboot_needed()) {
 950                 /*
 951                  * An image was found.  No turning back: we're booted
 952                  * with a kernel from the disk image.
 953                  */
 954                 imageboot_setup();
 955         }
 956 #endif /* CONFIG_IMAGEBOOT */
 957
 958         /* set initial time; all other resource data is  already zero'ed */
 959         microtime_with_abstime(&kernproc->p_start, &kernproc->p_stats->ps_start);
 960
 961 #if DEVFS
 962         {
 963             char mounthere[] = "/dev";  /* !const because of internal casting */
 964
 965             bsd_init_kprintf("calling devfs_kernel_mount\n");
 966             devfs_kernel_mount(mounthere);
 967         }
 968 #endif /* DEVFS */
 969
 970         /* Initialize signal state for process 0. */
 971         bsd_init_kprintf("calling siginit\n");
 972         siginit(kernproc);
 973
 974         bsd_init_kprintf("calling bsd_utaskbootstrap\n");
 975         bsd_utaskbootstrap();
 976
 977 #if defined(__LP64__)
 978         kernproc->p_flag |= P_LP64;
 979 #endif
 980
 981         pal_kernel_announce();
 982
 983         bsd_init_kprintf("calling mountroot_post_hook\n");
 984
 985         /* invoke post-root-mount hook */
 986         if (mountroot_post_hook != NULL)
 987                 mountroot_post_hook();
 988
 989 #if 0 /* not yet */
 990         consider_zone_gc(FALSE);
 991 #endif
 992
 993         bsd_init_kprintf("done\n");
 994 }
 995
 996 void
 997 bsdinit_task(void)
 998 {
 999         proc_t p = current_proc();
1000         struct uthread *ut;
1001         thread_t thread;
1002
1003         process_name("init", p);
1004
1005         ux_handler_init();
1006
1007         thread = current_thread();
1008         (void) host_set_exception_ports(host_priv_self(),
1009                                         EXC_MASK_ALL & ~(EXC_MASK_RPC_ALERT),//pilotfish (shark) needs this port
1010                                         (mach_port_t) ux_exception_port,
1011                                         EXCEPTION_DEFAULT| MACH_EXCEPTION_CODES,
1012                                         0);
1013
1014         ut = (uthread_t)get_bsdthread_info(thread);
1015
1016         bsd_init_task = get_threadtask(thread);
1017         init_task_died = FALSE;
1018         init_task_failure_data[0] = 0;
1019
1020 #if CONFIG_MACF
1021         mac_cred_label_associate_user(p->p_ucred);
1022 #endif
1023         load_init_program(p);
1024         lock_trace = 1;
1025 }
1026
1027 kern_return_t
1028 bsd_autoconf(void)
1029 {
1030         kprintf("bsd_autoconf: calling kminit\n");
1031         kminit();
1032
1033         /*
1034          * Early startup for bsd pseudodevices.
1035          */
1036         {
1037             struct pseudo_init *pi;
1038
1039             for (pi = pseudo_inits; pi->ps_func; pi++)
1040                 (*pi->ps_func) (pi->ps_count);
1041         }
1042
1043         return( IOKitBSDInit());
1044 }
1045
1046
1047 #include <sys/disklabel.h>  /* for MAXPARTITIONS */
1048
1049 static void
1050 setconf(void)
1051 {
1052         u_int32_t       flags;
1053         kern_return_t   err;
1054
1055         err = IOFindBSDRoot(rootdevice, sizeof(rootdevice), &rootdev, &flags);
1056         if( err) {
1057                 printf("setconf: IOFindBSDRoot returned an error (%d);"
1058                         "setting rootdevice to 'sd0a'.\n", err); /* XXX DEBUG TEMP */
1059                 rootdev = makedev( 6, 0 );
1060                 strlcpy(rootdevice, "sd0a", sizeof(rootdevice));
1061                 flags = 0;
1062         }
1063
1064 #if NFSCLIENT
1065         if( flags & 1 ) {
1066                 /* network device */
1067                 mountroot = netboot_mountroot;
1068         } else {
1069 #endif
1070                 /* otherwise have vfs determine root filesystem */
1071                 mountroot = NULL;
1072 #if NFSCLIENT
1073         }
1074 #endif
1075
1076 }
1077
1078 void
1079 bsd_utaskbootstrap(void)
1080 {
1081         thread_t thread;
1082         struct uthread *ut;
1083
1084         /*
1085          * Clone the bootstrap process from the kernel process, without
1086          * inheriting either task characteristics or memory from the kernel;
1087          */
1088         thread = cloneproc(TASK_NULL, COALITION_NULL, kernproc, FALSE, TRUE);
1089
1090         /* Hold the reference as it will be dropped during shutdown */
1091         initproc = proc_find(1);
1092 #if __PROC_INTERNAL_DEBUG
1093         if (initproc == PROC_NULL)
1094                 panic("bsd_utaskbootstrap: initproc not set\n");
1095 #endif
1096         /*
1097          * Since we aren't going back out the normal way to our parent,
1098          * we have to drop the transition locks explicitly.
1099          */
1100         proc_signalend(initproc, 0);
1101         proc_transend(initproc, 0);
1102
1103         ut = (struct uthread *)get_bsdthread_info(thread);
1104         ut->uu_sigmask = 0;
1105         act_set_astbsd(thread);
1106         (void) thread_resume(thread);
1107 }
1108
1109 static void
1110 parse_bsd_args(void)
1111 {
1112         char namep[16];
1113         int msgbuf;
1114
1115         if ( PE_parse_boot_argn("-s", namep, sizeof (namep)))
1116                 boothowto |= RB_SINGLE;
1117
1118         if (PE_parse_boot_argn("-b", namep, sizeof (namep)))
1119                 boothowto |= RB_NOBOOTRC;
1120
1121         if (PE_parse_boot_argn("-x", namep, sizeof (namep))) /* safe boot */
1122                 boothowto |= RB_SAFEBOOT;
1123
1124
1125         /* disable vnode_cache_is_authorized() by setting vnode_cache_defeat */
1126         if (PE_parse_boot_argn("-vnode_cache_defeat", namep, sizeof (namep)))
1127                 bootarg_vnode_cache_defeat = 1;
1128
1129 #if DEVELOPMENT || DEBUG
1130         if (PE_parse_boot_argn("-disable_aslr", namep, sizeof (namep)))
1131                 bootarg_disable_aslr = 1;
1132 #endif
1133
1134         PE_parse_boot_argn("ncl", &ncl, sizeof (ncl));
1135         if (PE_parse_boot_argn("nbuf", &max_nbuf_headers,
1136                                 sizeof (max_nbuf_headers))) {
1137                 customnbuf = 1;
1138         }
1139
1140 #if CONFIG_MACF
1141 #if defined (__i386__) || defined (__x86_64__)
1142         PE_parse_boot_argn("policy_check", &policy_check_flags, sizeof (policy_check_flags));
1143 #endif
1144 #endif  /* CONFIG_MACF */
1145
1146         if (PE_parse_boot_argn("msgbuf", &msgbuf, sizeof (msgbuf))) {
1147                 log_setsize(msgbuf);
1148         }
1149
1150         if (PE_parse_boot_argn("-novfscache", namep, sizeof(namep))) {
1151                 nc_disabled = 1;
1152         }
1153 }
1154
1155 void
1156 bsd_exec_setup(int scale)
1157 {
1158
1159         switch (scale) {
1160                 case 0:
1161                 case 1:
1162                         bsd_simul_execs = BSD_SIMUL_EXECS;
1163                         break;
1164                 case 2:
1165                 case 3:
1166                         bsd_simul_execs = 65;
1167                         break;
1168                 case 4:
1169                 case 5:
1170                         bsd_simul_execs = 129;
1171                         break;
1172                 case 6:
1173                 case 7:
1174                         bsd_simul_execs = 257;
1175                         break;
1176                 default:
1177                         bsd_simul_execs = 513;
1178                         break;
1179
1180         }
1181         bsd_pageable_map_size = (bsd_simul_execs * BSD_PAGEABLE_SIZE_PER_EXEC);
1182 }
1183
1184 #if !NFSCLIENT
1185 int
1186 netboot_root(void);
1187
1188 int
1189 netboot_root(void)
1190 {
1191         return(0);
1192 }
1193 #endif