]> git.saurik.com Git - apple/xnu.git/blobdiff - bsd/kern/bsd_init.c
xnu-7195.101.1.tar.gz
[apple/xnu.git] / bsd / kern / bsd_init.c
index 862a0557aa0f586d35ee4f1ad5f123f871384f95..6dd42a5f0e71dcc7ecfd2fd43b297ad7d5e0a9cd 100644 (file)
@@ -1,8 +1,8 @@
 /*
- * Copyright (c) 2000-2004 Apple Computer, Inc. All rights reserved.
+ * Copyright (c) 2000-2020 Apple Inc. All rights reserved.
  *
  * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
- * 
+ *
  * This file contains Original Code and/or Modifications of Original Code
  * as defined in and that are subject to the Apple Public Source License
  * Version 2.0 (the 'License'). You may not use this file except in
  * unlawful or unlicensed copies of an Apple operating system, or to
  * circumvent, violate, or enable the circumvention or violation of, any
  * terms of an Apple operating system software license agreement.
- * 
+ *
  * Please obtain a copy of the License at
  * http://www.opensource.apple.com/apsl/ and read it before using this file.
- * 
+ *
  * The Original Code and all software distributed under the License are
  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  * Please see the License for the specific language governing rights and
  * limitations under the License.
- * 
+ *
  * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
- */
-/*
+ *
+ *
  * Copyright (c) 1982, 1986, 1989, 1991, 1992, 1993
  *     The Regents of the University of California.  All rights reserved.
  * (c) UNIX System Laboratories, Inc.
  *     @(#)init_main.c 8.16 (Berkeley) 5/14/95
  */
 
-/* 
+/*
  *
  * Mach Operating System
  * Copyright (c) 1987 Carnegie-Mellon University
  * All rights reserved.  The CMU software License Agreement specifies
  * the terms and conditions for use and redistribution.
  */
+/*
+ * NOTICE: This file was modified by McAfee Research in 2004 to introduce
+ * support for mandatory and extensible security protections.  This notice
+ * is included in support of clause 2.2 (b) of the Apple Public License,
+ * Version 2.0.
+ */
 
 #include <sys/param.h>
 #include <sys/filedesc.h>
 #include <sys/vnode_internal.h>
 #include <sys/conf.h>
 #include <sys/buf_internal.h>
-#include <sys/clist.h>
 #include <sys/user.h>
 #include <sys/time.h>
 #include <sys/systm.h>
 #include <sys/mman.h>
+#include <sys/kasl.h>
 
-#include <bsm/audit_kernel.h>
+#include <security/audit/audit.h>
 
 #include <sys/malloc.h>
 #include <sys/dkstat.h>
+#include <sys/codesign.h>
 
 #include <kern/startup.h>
 #include <kern/thread.h>
 #include <kern/task.h>
 #include <kern/ast.h>
+#include <kern/zalloc.h>
+#include <kern/ux_handler.h>            /* for ux_handler_setup() */
 
 #include <mach/vm_param.h>
 
 #include <vm/vm_map.h>
 #include <vm/vm_kern.h>
 
-#include <sys/ux_exception.h>  /* for ux_exception_port */
-
 #include <sys/reboot.h>
-#include <mach/exception_types.h>
-#include <dev/busvar.h>                        /* for pseudo_inits */
+#include <dev/busvar.h>                 /* for pseudo_inits */
 #include <sys/kdebug.h>
+#include <sys/monotonic.h>
 
 #include <mach/mach_types.h>
 #include <mach/vm_prot.h>
 #include <mach/semaphore.h>
 #include <mach/sync_policy.h>
 #include <kern/clock.h>
+#include <sys/csr.h>
 #include <mach/kern_return.h>
-
-#include <mach/shared_memory_server.h>
-#include <vm/vm_shared_memory_server.h>
+#include <mach/thread_act.h>            /* for thread_resume() */
+#include <sys/mcache.h>                 /* for mcache_init() */
+#include <sys/mbuf.h>                   /* for mbinit() */
+#include <sys/event.h>                  /* for knote_init() */
+#include <sys/eventhandler.h>           /* for eventhandler_init() */
+#include <sys/kern_memorystatus.h>      /* for memorystatus_init() */
+#include <sys/kern_memorystatus_freeze.h> /* for memorystatus_freeze_init() */
+#include <sys/aio_kern.h>               /* for aio_init() */
+#include <sys/semaphore.h>              /* for psem_cache_init() */
+#include <net/dlil.h>                   /* for dlil_init() */
+#include <net/kpi_protocol.h>           /* for proto_kpi_init() */
+#include <net/iptap.h>                  /* for iptap_init() */
+#include <sys/socketvar.h>              /* for socketinit() */
+#include <sys/protosw.h>                /* for domaininit() */
+#include <kern/sched_prim.h>            /* for thread_wakeup() */
+#include <net/if_ether.h>               /* for ether_family_init() */
+#include <net/if_gif.h>                 /* for gif_init() */
+#include <miscfs/devfs/devfsdefs.h>     /* for devfs_kernel_mount() */
+#include <vm/vm_kern.h>                 /* for kmem_suballoc() */
+#include <sys/msgbuf.h>                 /* for log_setsize() */
+#include <sys/proc_uuid_policy.h>       /* proc_uuid_policy_init() */
+#include <netinet/flow_divert.h>        /* flow_divert_init() */
+#include <net/content_filter.h>         /* for cfil_init() */
+#include <net/necp.h>                   /* for necp_init() */
+#include <net/network_agent.h>          /* for netagent_init() */
+#include <net/packet_mangler.h>         /* for pkt_mnglr_init() */
+#include <net/if_utun.h>                /* for utun_register_control() */
+#include <net/if_ipsec.h>               /* for ipsec_register_control() */
+#include <net/net_str_id.h>             /* for net_str_id_init() */
+#include <net/netsrc.h>                 /* for netsrc_init() */
+#include <net/ntstat.h>                 /* for nstat_init() */
+#include <netinet/tcp_cc.h>                     /* for tcp_cc_init() */
+#include <netinet/mptcp_var.h>          /* for mptcp_control_register() */
+#include <net/nwk_wq.h>                 /* for nwk_wq_init */
+#include <net/restricted_in_port.h> /* for restricted_in_port_init() */
+#include <kern/assert.h>                /* for assert() */
+#include <sys/kern_overrides.h>         /* for init_system_override() */
+#include <sys/lockf.h>                  /* for lf_init() */
+#include <sys/fsctl.h>
 
 #include <net/init.h>
 
+#if CONFIG_MACF
+#include <security/mac_framework.h>
+#include <security/mac_internal.h>      /* mac_init_bsd() */
+#include <security/mac_mach_internal.h> /* mac_update_task_label() */
+#endif
+
 #include <machine/exec.h>
 
-extern int app_profile;                /* on/off switch for pre-heat cache */
+#if CONFIG_NETBOOT
+#include <sys/netboot.h>
+#endif
+
+#if CONFIG_IMAGEBOOT
+#include <sys/imageboot.h>
+#endif
+
+#if PFLOG
+#include <net/if_pflog.h>
+#endif
+
+
+#include <pexpert/pexpert.h>
+#include <machine/pal_routines.h>
+#include <console/video_console.h>
+
+#if CONFIG_XNUPOST
+#include <tests/xnupost.h>
+#endif
 
-char    copyright[] =
-"Copyright (c) 1982, 1986, 1989, 1991, 1993\n\t"
-"The Regents of the University of California. "
-"All rights reserved.\n\n";
+void * get_user_regs(thread_t);         /* XXX kludge for <machine/thread.h> */
+void IOKitInitializeTime(void);         /* XXX */
+void IOSleep(unsigned int);             /* XXX */
+void loopattach(void);                  /* XXX */
 
-extern void    ux_handler();
+const char *const copyright =
+    "Copyright (c) 1982, 1986, 1989, 1991, 1993\n\t"
+    "The Regents of the University of California. "
+    "All rights reserved.\n\n";
 
 /* Components of the first process -- never freed. */
-struct proc proc0;
-struct session session0;
-struct pgrp pgrp0;
-struct filedesc filedesc0;
-struct plimit limit0;
-struct pstats pstats0;
-struct sigacts sigacts0;
-struct proc *kernproc, *initproc;
+struct  proc proc0 = { .p_comm = "kernel_task", .p_name = "kernel_task" };
+struct  session session0;
+struct  pgrp pgrp0;
+struct  filedesc filedesc0;
+struct  plimit limit0;
+struct  pstats pstats0;
+struct  sigacts sigacts0;
+SECURITY_READ_ONLY_LATE(proc_t) kernproc = &proc0;
+proc_t XNU_PTRAUTH_SIGNED_PTR("initproc") initproc;
 
 long tk_cancc;
 long tk_nin;
@@ -156,62 +229,145 @@ int nswapmap;
 void *swapmap;
 struct swdevt swdevt[1];
 
-dev_t  rootdev;                /* device of the root */
-dev_t  dumpdev;                /* device to take dumps on */
-long   dumplo;                 /* offset into dumpdev */
-long   hostid;
-char   hostname[MAXHOSTNAMELEN];
-int            hostnamelen;
-char   domainname[MAXDOMNAMELEN];
-int            domainnamelen;
-#if __i386__
-struct exec_archhandler exec_archhandler_ppc = {
-       .path = "/usr/libexec/oah/translate",
-};
-#else /* __i386__ */
-struct exec_archhandler exec_archhandler_ppc;
-#endif /* __i386__ */
-
-char rootdevice[16];   /* hfs device names have at least 9 chars */
-
-#ifdef  KMEMSTATS
-struct kmemstats kmemstats[M_LAST];
-#endif
-
-int    lbolt;                          /* awoken once a second */
-struct vnode *rootvp;
-int boothowto = RB_DEBUG;
-
-#define        BSD_PAGABLE_MAP_SIZE    (16 * 512 * 1024)
-vm_map_t       bsd_pageable_map;
-vm_map_t       mb_map;
-semaphore_t execve_semaphore;
-
-int    cmask = CMASK;
+static LCK_GRP_DECLARE(hostname_lck_grp, "hostname");
+LCK_MTX_DECLARE(hostname_lock, &hostname_lck_grp);
+LCK_MTX_DECLARE(domainname_lock, &hostname_lck_grp);
+
+dev_t   rootdev;                /* device of the root */
+dev_t   dumpdev;                /* device to take dumps on */
+long    dumplo;                 /* offset into dumpdev */
+long    hostid;
+char    hostname[MAXHOSTNAMELEN];
+char    domainname[MAXDOMNAMELEN];
+char    rootdevice[DEVMAXNAMESIZE];
+
+struct  vnode *rootvp;
+bool rootvp_is_ssd = false;
+int boothowto;
+int minimalboot = 0;
+#if CONFIG_DARKBOOT
+int darkboot = 0;
+#endif
+
+#if __arm64__
+int legacy_footprint_entitlement_mode = LEGACY_FOOTPRINT_ENTITLEMENT_IGNORE;
+#endif /* __arm64__ */
+
+#if PROC_REF_DEBUG
+__private_extern__ int proc_ref_tracking_disabled = 0; /* disable panics on leaked proc refs across syscall boundary */
+#endif
+
+extern kern_return_t IOFindBSDRoot(char *, unsigned int, dev_t *, u_int32_t *);
+extern void IOSecureBSDRoot(const char * rootName);
+extern kern_return_t IOKitBSDInit(void );
+extern boolean_t IOSetRecoveryBoot(bsd_bootfail_mode_t, uuid_t, boolean_t);
+extern void kminit(void);
+extern void bsd_bufferinit(void);
+extern void oslog_setsize(int size);
+extern void throttle_init(void);
+
+#if CONFIG_LOCKERBOOT
+#define LOCKER_PROTOBOOT_MOUNT "/protoboot"
+
+const char kernel_protoboot_mount[] = LOCKER_PROTOBOOT_MOUNT;
+extern int mount_locker_protoboot(const char *fsname, const char *mntpoint,
+    const char *pbdevpath);
+#endif
+
+extern int ncl;
+#if DEVELOPMENT || DEBUG
+extern int syscallfilter_disable;
+#endif // DEVELOPMENT || DEBUG
+
+vm_map_t        bsd_pageable_map;
+vm_map_t        mb_map;
+
+static  int bsd_simul_execs;
+static int bsd_pageable_map_size;
+__private_extern__ int execargs_cache_size = 0;
+__private_extern__ int execargs_free_count = 0;
+__private_extern__ vm_offset_t * execargs_cache = NULL;
+
+void bsd_exec_setup(int);
+
+__private_extern__ int bootarg_execfailurereports = 0;
+
+#if __x86_64__
+__private_extern__ int bootarg_no32exec = 1;
+#endif
+__private_extern__ int bootarg_vnode_cache_defeat = 0;
+
+#if CONFIG_JETSAM && (DEVELOPMENT || DEBUG)
+__private_extern__ int bootarg_no_vnode_jetsam = 0;
+#endif /* CONFIG_JETSAM && (DEVELOPMENT || DEBUG) */
+
+__private_extern__ int bootarg_no_vnode_drain = 0;
+
+/*
+ * Prevent kernel-based ASLR from being used, for testing.
+ */
+#if DEVELOPMENT || DEBUG
+__private_extern__ int bootarg_disable_aslr = 0;
+#endif
+
+
+/*
+ * Allow an alternate dyld to be used for testing.
+ */
+
+#if DEVELOPMENT || DEBUG
+char dyld_alt_path[MAXPATHLEN];
+int use_alt_dyld = 0;
+extern uint64_t dyld_flags;
+#endif
+
+int     cmask = CMASK;
 extern int customnbuf;
 
-int parse_bsd_args(void);
-extern int bsd_hardclockinit;
-extern task_t bsd_init_task;
-extern char    init_task_failure_data[];
-extern void time_zone_slock_init(void);
-static void process_name(char *, struct proc *);
+kern_return_t bsd_autoconf(void);
+void bsd_utaskbootstrap(void);
+
+static void parse_bsd_args(void);
+#if CONFIG_DEV_KMEM
+extern void dev_kmem_init(void);
+#endif
+extern void select_waitq_init(void);
+static void process_name(const char *, proc_t);
 
 static void setconf(void);
 
-funnel_t *kernel_flock;
+#if CONFIG_BASESYSTEMROOT
+static int bsd_find_basesystem_dmg(char *bsdmgpath_out, bool *rooted_dmg, bool *skip_signature_check);
+static boolean_t bsdmgroot_bootable(void);
+#endif // CONFIG_BASESYSTEMROOT
+
+static bool bsd_rooted_ramdisk(void);
 
+#if SYSV_SHM
 extern void sysv_shm_lock_init(void);
+#endif
+#if SYSV_SEM
 extern void sysv_sem_lock_init(void);
+#endif
+#if SYSV_MSG
 extern void sysv_msg_lock_init(void);
-extern void pshm_lock_init();
-extern void psem_lock_init();
-extern int maxprocperuid;
+#endif
 
-/* kmem access not enabled by default; can be changed with boot-args */
-int setup_kmem = 0;
+extern void ulock_initialize(void);
 
-extern void stackshot_lock_init();
+#if CONFIG_MACF
+#if defined (__i386__) || defined (__x86_64__)
+/* MACF policy_check configuration flags; see policy_check.c for details */
+int policy_check_flags = 0;
+
+extern int check_policy_init(int);
+#endif
+#endif  /* CONFIG_MACF */
+
+/* If we are using CONFIG_DTRACE */
+#if CONFIG_DTRACE
+extern void dtrace_postinit(void);
+#endif
 
 /*
  * Initialization code.
@@ -229,34 +385,90 @@ extern void stackshot_lock_init();
  *     Sets the name for the given task.
  */
 static void
-process_name(s, p)
-       char            *s;
-       struct proc *p;
+process_name(const char *s, proc_t p)
 {
-       int             length = strlen(s);
-
-       bcopy(s, p->p_comm,
-               length >= sizeof(p->p_comm) ? sizeof(p->p_comm) :
-                       length + 1);
+       strlcpy(p->p_comm, s, sizeof(p->p_comm));
+       strlcpy(p->p_name, s, sizeof(p->p_name));
 }
 
 /* To allow these values to be patched, they're globals here */
 #include <machine/vmparam.h>
-struct rlimit vm_initial_limit_stack = { DFLSSIZ, MAXSSIZ };
-struct rlimit vm_initial_limit_data = { DFLDSIZ, MAXDSIZ };
-struct rlimit vm_initial_limit_core = { DFLCSIZ, MAXCSIZ };
+struct rlimit vm_initial_limit_stack = { .rlim_cur = DFLSSIZ, .rlim_max = MAXSSIZ - PAGE_MAX_SIZE };
+struct rlimit vm_initial_limit_data = { .rlim_cur = DFLDSIZ, .rlim_max = MAXDSIZ };
+struct rlimit vm_initial_limit_core = { .rlim_cur = DFLCSIZ, .rlim_max = MAXCSIZ };
+
+extern struct os_refgrp rlimit_refgrp;
 
-extern thread_t        cloneproc(struct proc *, int);
-extern int     (*mountroot)(void);
-extern int     netboot_mountroot();    /* netboot.c */
-extern int     netboot_setup(struct proc * p);
+extern thread_t cloneproc(task_t, coalition_t, proc_t, int, int);
+extern int      (*mountroot)(void);
 
-lck_grp_t * proc_lck_grp;
-lck_grp_attr_t * proc_lck_grp_attr;
-lck_attr_t * proc_lck_attr;
+LCK_ATTR_DECLARE(proc_lck_attr, 0, 0);
+LCK_GRP_DECLARE(proc_lck_grp, "proc");
+LCK_GRP_DECLARE(proc_slock_grp, "proc-slock");
+LCK_GRP_DECLARE(proc_fdmlock_grp, "proc-fdmlock");
+LCK_GRP_DECLARE(proc_mlock_grp, "proc-mlock");
+LCK_GRP_DECLARE(proc_ucred_mlock_grp, "proc-ucred-mlock");
+LCK_GRP_DECLARE(proc_dirslock_grp, "proc-dirslock");
+LCK_GRP_DECLARE(proc_kqhashlock_grp, "proc-kqhashlock");
+LCK_GRP_DECLARE(proc_knhashlock_grp, "proc-knhashlock");
 
+
+LCK_MTX_DECLARE_ATTR(proc_list_mlock, &proc_mlock_grp, &proc_lck_attr);
+
+#if XNU_TARGET_OS_OSX
 /* hook called after root is mounted XXX temporary hack */
 void (*mountroot_post_hook)(void);
+void (*unmountroot_pre_hook)(void);
+#endif
+void set_rootvnode(vnode_t);
+
+extern lck_rw_t rootvnode_rw_lock;
+
+/* called with an iocount and usecount on new_rootvnode */
+void
+set_rootvnode(vnode_t new_rootvnode)
+{
+       mount_t new_mount = (new_rootvnode != NULL) ? new_rootvnode->v_mount : NULL;
+       vnode_t new_devvp = (new_mount != NULL) ? new_mount->mnt_devvp : NULL;
+       vnode_t old_rootvnode = rootvnode;
+
+       new_rootvnode->v_flag |= VROOT;
+       rootvp = new_devvp;
+       rootvnode = new_rootvnode;
+       filedesc0.fd_cdir = new_rootvnode;
+       if (new_devvp != NULL) {
+               rootdev = vnode_specrdev(new_devvp);
+       } else if (new_mount != NULL) {
+               rootdev = vfs_statfs(new_mount)->f_fsid.val[0];  /* like ATTR_CMN_DEVID */
+       } else {
+               rootdev = NODEV;
+       }
+
+       if (old_rootvnode) {
+               vnode_rele(old_rootvnode);
+       }
+}
+
+#define RAMDEV "md0"
+
+bool
+bsd_rooted_ramdisk(void)
+{
+       bool is_ramdisk = false;
+       char *dev_path = zalloc(ZV_NAMEI);
+       if (dev_path == NULL) {
+               panic("failed to allocate devpath string! \n");
+       }
+
+       if (PE_parse_boot_argn("rd", dev_path, MAXPATHLEN)) {
+               if (strncmp(dev_path, RAMDEV, strlen(RAMDEV)) == 0) {
+                       is_ramdisk = true;
+               }
+       }
+
+       zfree(ZV_NAMEI, dev_path);
+       return is_ramdisk;
+}
 
 /*
  * This function is called very early on in the Mach startup, from the
@@ -266,153 +478,236 @@ void (*mountroot_post_hook)(void);
  * Internally, kernel_thread_create() calls thread_create_internal(),
  * which calls uthread_alloc().  The function of uthread_alloc() is
  * normally to allocate a uthread structure, and fill out the uu_sigmask,
- * uu_act, and uu_ucred fields.  It skips filling these out in the case
- * of the "task" being "kernel_task", because the order of operation is
- * inverted.  To account for that, we need to manually fill in at least
- * the uu_cred field so that the uthread structure can be used like any
- * other.
+ * uu_context fields.  It skips filling these out in the case of the "task"
+ * being "kernel_task", because the order of operation is inverted.  To
+ * account for that, we need to manually fill in at least the contents
+ * of the uu_context.vc_ucred field so that the uthread structure can be
+ * used like any other.
  */
+
 void
-bsd_init()
+bsd_init(void)
 {
-       register struct proc *p;
        struct uthread *ut;
-       extern kauth_cred_t rootcred;
-       register int i;
-       int s;
-       int error;
-       thread_t        th;
+       unsigned int i;
        struct vfs_context context;
-       void            lightning_bolt(void );
-       kern_return_t   ret;
-       boolean_t funnel_state;
+       kern_return_t   ret;
        struct ucred temp_cred;
-       extern void file_lock_init(void);
+       struct posix_cred temp_pcred;
+       vnode_t init_rootvnode = NULLVP;
+#if CONFIG_NETBOOT || CONFIG_IMAGEBOOT
+       boolean_t       netboot = FALSE;
+#endif
+#if CONFIG_LOCKERBOOT
+       vnode_t pbvn = NULLVP;
+       mount_t pbmnt = NULL;
+       char *pbdevp = NULL;
+       char pbdevpath[64];
+       char pbfsname[MFSNAMELEN];
+       const char *slash_dev = NULL;
+#endif
 
-       kernel_flock = funnel_alloc(KERNEL_FUNNEL);
-       if (kernel_flock == (funnel_t *)0 ) {
-               panic("bsd_init: Failed to allocate kernel funnel");
-       }
-        
-       funnel_state = thread_funnel_set(kernel_flock, TRUE);
+#define DEBUG_BSDINIT 0
+
+#if DEBUG_BSDINIT
+#define bsd_init_kprintf(x, ...) kprintf("bsd_init: " x, ## __VA_ARGS__)
+#else
+#define bsd_init_kprintf(x, ...)
+#endif
+
+       throttle_init();
 
        printf(copyright);
-       
-       kmeminit();
-       
+
+       bsd_init_kprintf("calling parse_bsd_args\n");
        parse_bsd_args();
 
-       /* Initialize the uthread zone */
-       //uthread_zone_init();  /* XXX redundant: previous uthread_alloc() */
+#if CONFIG_DEV_KMEM
+       bsd_init_kprintf("calling dev_kmem_init\n");
+       dev_kmem_init();
+#endif
 
        /* Initialize kauth subsystem before instancing the first credential */
+       bsd_init_kprintf("calling kauth_init\n");
        kauth_init();
 
        /* Initialize process and pgrp structures. */
+       bsd_init_kprintf("calling procinit\n");
        procinit();
 
-       kernproc = &proc0;
-
-       p = kernproc;
-
        /* kernel_task->proc = kernproc; */
-       set_bsdtask_info(kernel_task,(void *)p);
-       p->p_pid = 0;
+       set_bsdtask_info(kernel_task, (void *)kernproc);
 
        /* give kernproc a name */
-       process_name("kernel_task", p);
-
-       /* allocate proc lock group attribute and group */
-       proc_lck_grp_attr= lck_grp_attr_alloc_init();
-       
-       proc_lck_grp = lck_grp_alloc_init("proc",  proc_lck_grp_attr);
+       bsd_init_kprintf("calling process_name\n");
+       process_name("kernel_task", kernproc);
 
        /* Allocate proc lock attribute */
-       proc_lck_attr = lck_attr_alloc_init();
 
-       lck_mtx_init(&p->p_mlock, proc_lck_grp, proc_lck_attr);
-       lck_mtx_init(&p->p_fdmlock, proc_lck_grp, proc_lck_attr);
+       lck_mtx_init(&kernproc->p_mlock, &proc_mlock_grp, &proc_lck_attr);
+       lck_mtx_init(&kernproc->p_fdmlock, &proc_fdmlock_grp, &proc_lck_attr);
+       lck_mtx_init(&kernproc->p_ucred_mlock, &proc_ucred_mlock_grp, &proc_lck_attr);
+       lck_spin_init(&kernproc->p_slock, &proc_slock_grp, &proc_lck_attr);
+       lck_rw_init(&kernproc->p_dirs_lock, &proc_dirslock_grp, &proc_lck_attr);
 
-       if (current_task() != kernel_task)
+       assert(bsd_simul_execs != 0);
+       execargs_cache_size = bsd_simul_execs;
+       execargs_free_count = bsd_simul_execs;
+       execargs_cache = zalloc_permanent(bsd_simul_execs * sizeof(vm_offset_t),
+           ZALIGN(vm_offset_t));
+
+       if (current_task() != kernel_task) {
                printf("bsd_init: We have a problem, "
-                               "current task is not kernel task\n");
-       
+                   "current task is not kernel task\n");
+       }
+
+       bsd_init_kprintf("calling get_bsdthread_info\n");
        ut = (uthread_t)get_bsdthread_info(current_thread());
 
+#if CONFIG_MACF
+       /*
+        * Initialize the MAC Framework
+        */
+       mac_policy_initbsd();
+
+#if defined (__i386__) || defined (__x86_64__)
+       /*
+        * We currently only support this on i386/x86_64, as that is the
+        * only lock code we have instrumented so far.
+        */
+       check_policy_init(policy_check_flags);
+#endif
+#endif /* MAC */
+
+       ulock_initialize();
+
        /*
         * Create process 0.
         */
-       LIST_INSERT_HEAD(&allproc, p, p_list);
-       p->p_pgrp = &pgrp0;
+       proc_list_lock();
+       LIST_INSERT_HEAD(&allproc, kernproc, p_list);
+       kernproc->p_pgrp = &pgrp0;
        LIST_INSERT_HEAD(PGRPHASH(0), &pgrp0, pg_hash);
        LIST_INIT(&pgrp0.pg_members);
-       LIST_INSERT_HEAD(&pgrp0.pg_members, p, p_pglist);
+       lck_mtx_init(&pgrp0.pg_mlock, &proc_mlock_grp, &proc_lck_attr);
+       /* There is no other bsd thread this point and is safe without pgrp lock */
+       LIST_INSERT_HEAD(&pgrp0.pg_members, kernproc, p_pglist);
+       kernproc->p_listflag |= P_LIST_INPGRP;
+       kernproc->p_pgrpid = 0;
+       kernproc->p_uniqueid = 0;
 
        pgrp0.pg_session = &session0;
+       pgrp0.pg_membercnt = 1;
+
        session0.s_count = 1;
-       session0.s_leader = p;
-
-       p->task = kernel_task;
-       
-       p->p_stat = SRUN;
-       p->p_flag = P_SYSTEM;
-       p->p_nice = NZERO;
-       p->p_pptr = p;
-       lockinit(&p->signal_lock, PVM, "signal", 0, 0);
-       TAILQ_INIT(&p->p_uthlist);
-       p->sigwait = FALSE;
-       p->sigwait_thread = THREAD_NULL;
-       p->exit_thread = THREAD_NULL;
+       session0.s_leader = kernproc;
+       session0.s_listflags = 0;
+       lck_mtx_init(&session0.s_mlock, &proc_mlock_grp, &proc_lck_attr);
+       LIST_INSERT_HEAD(SESSHASH(0), &session0, s_hash);
+       proc_list_unlock();
+
+#if CONFIG_PERSONAS
+       kernproc->p_persona = NULL;
+#endif
+
+       kernproc->task = kernel_task;
+
+       kernproc->p_stat = SRUN;
+       kernproc->p_flag = P_SYSTEM;
+       kernproc->p_lflag = 0;
+       kernproc->p_ladvflag = 0;
+
+#if defined(__LP64__)
+       kernproc->p_flag |= P_LP64;
+#endif
+
+#if DEVELOPMENT || DEBUG
+       if (bootarg_disable_aslr) {
+               kernproc->p_flag |= P_DISABLE_ASLR;
+       }
+#endif
+
+       kernproc->p_nice = NZERO;
+       kernproc->p_pptr = kernproc;
+
+       TAILQ_INIT(&kernproc->p_uthlist);
+       TAILQ_INSERT_TAIL(&kernproc->p_uthlist, ut, uu_list);
+
+       kernproc->sigwait = FALSE;
+       kernproc->sigwait_thread = THREAD_NULL;
+       kernproc->exit_thread = THREAD_NULL;
+       kernproc->p_csflags = CS_VALID;
 
        /*
         * Create credential.  This also Initializes the audit information.
-        * XXX It is not clear what the initial values should be for audit ID,
-        * XXX session ID, etc..
         */
+       bsd_init_kprintf("calling bzero\n");
        bzero(&temp_cred, sizeof(temp_cred));
-       temp_cred.cr_ngroups = 1;
+       bzero(&temp_pcred, sizeof(temp_pcred));
+       temp_pcred.cr_ngroups = 1;
+       /* kern_proc, shouldn't call up to DS for group membership */
+       temp_pcred.cr_flags = CRF_NOMEMBERD;
+       temp_cred.cr_audit.as_aia_p = audit_default_aia_p;
+
+       bsd_init_kprintf("calling kauth_cred_create\n");
+       /*
+        * We have to label the temp cred before we create from it to
+        * properly set cr_ngroups, or the create will fail.
+        */
+       posix_cred_label(&temp_cred, &temp_pcred);
+       kernproc->p_ucred = kauth_cred_create(&temp_cred);
 
-       p->p_ucred = kauth_cred_create(&temp_cred); 
+       /* update cred on proc */
+       PROC_UPDATE_CREDS_ONPROC(kernproc);
 
        /* give the (already exisiting) initial thread a reference on it */
-       kauth_cred_ref(p->p_ucred);
-       ut->uu_ucred = p->p_ucred;
-       
-       TAILQ_INIT(&p->aio_activeq);
-       TAILQ_INIT(&p->aio_doneq);
-       p->aio_active_count = 0;
-       p->aio_done_count = 0;
+       bsd_init_kprintf("calling kauth_cred_ref\n");
+       kauth_cred_ref(kernproc->p_ucred);
+       ut->uu_context.vc_ucred = kernproc->p_ucred;
+       ut->uu_context.vc_thread = current_thread();
 
-       file_lock_init();
+       vfs_set_context_kernel(&ut->uu_context);
+
+       TAILQ_INIT(&kernproc->p_aio_activeq);
+       TAILQ_INIT(&kernproc->p_aio_doneq);
+       kernproc->p_aio_total_count = 0;
+
+#if CONFIG_MACF
+       mac_cred_label_associate_kernel(kernproc->p_ucred);
+#endif
 
        /* Create the file descriptor table. */
-       filedesc0.fd_refcnt = 1+1;      /* +1 so shutdown will not _FREE_ZONE */
-       p->p_fd = &filedesc0;
-       filedesc0.fd_cmask = cmask;
-       filedesc0.fd_knlistsize = -1;
+       kernproc->p_fd = &filedesc0;
+       filedesc0.fd_cmask = (mode_t)cmask;
+       filedesc0.fd_knlistsize = 0;
        filedesc0.fd_knlist = NULL;
        filedesc0.fd_knhash = NULL;
        filedesc0.fd_knhashmask = 0;
+       lck_mtx_init(&filedesc0.fd_kqhashlock, &proc_kqhashlock_grp, &proc_lck_attr);
+       lck_mtx_init(&filedesc0.fd_knhashlock, &proc_knhashlock_grp, &proc_lck_attr);
 
        /* Create the limits structures. */
-       p->p_limit = &limit0;
-       for (i = 0; i < sizeof(p->p_rlimit)/sizeof(p->p_rlimit[0]); i++)
-               limit0.pl_rlimit[i].rlim_cur = 
-                       limit0.pl_rlimit[i].rlim_max = RLIM_INFINITY;
+       kernproc->p_limit = &limit0;
+       for (i = 0; i < sizeof(kernproc->p_limit->pl_rlimit) / sizeof(kernproc->p_limit->pl_rlimit[0]); i++) {
+               limit0.pl_rlimit[i].rlim_cur =
+                   limit0.pl_rlimit[i].rlim_max = RLIM_INFINITY;
+       }
        limit0.pl_rlimit[RLIMIT_NOFILE].rlim_cur = NOFILE;
        limit0.pl_rlimit[RLIMIT_NPROC].rlim_cur = maxprocperuid;
        limit0.pl_rlimit[RLIMIT_NPROC].rlim_max = maxproc;
        limit0.pl_rlimit[RLIMIT_STACK] = vm_initial_limit_stack;
        limit0.pl_rlimit[RLIMIT_DATA] = vm_initial_limit_data;
        limit0.pl_rlimit[RLIMIT_CORE] = vm_initial_limit_core;
-       limit0.p_refcnt = 1;
+       os_ref_init_count(&limit0.pl_refcnt, &rlimit_refgrp, 1);
 
-       p->p_stats = &pstats0;
-       p->p_sigacts = &sigacts0;
+       kernproc->p_stats = &pstats0;
+       kernproc->p_sigacts = &sigacts0;
+       kernproc->p_subsystem_root_path = NULL;
 
        /*
-        * Charge root for two  processes: init and mach_init.
+        * Charge root for one process: launchd.
         */
+       bsd_init_kprintf("calling chgproccnt\n");
        (void)chgproccnt(0, 1);
 
        /*
@@ -420,16 +715,21 @@ bsd_init()
         *      for temporary copying (execve()).
         */
        {
-               vm_offset_t     min;
+               vm_offset_t     minimum;
 
+               bsd_init_kprintf("calling kmem_suballoc\n");
+               assert(bsd_pageable_map_size != 0);
                ret = kmem_suballoc(kernel_map,
-                               &min,
-                               (vm_size_t)BSD_PAGABLE_MAP_SIZE,
-                               TRUE,
-                               VM_FLAGS_ANYWHERE,
-                               &bsd_pageable_map);
-               if (ret != KERN_SUCCESS) 
+                   &minimum,
+                   (vm_size_t)bsd_pageable_map_size,
+                   TRUE,
+                   VM_FLAGS_ANYWHERE,
+                   VM_MAP_KERNEL_FLAGS_NONE,
+                   VM_KERN_MEMORY_BSD,
+                   &bsd_pageable_map);
+               if (ret != KERN_SUCCESS) {
                        panic("bsd_init: Failed to allocate bsd pageable map");
+               }
        }
 
        /*
@@ -439,83 +739,127 @@ bsd_init()
         *              happen after a credential has been associated with
         *              the kernel task.
         */
+       bsd_init_kprintf("calling bsd_bufferinit\n");
        bsd_bufferinit();
 
-       /* Initialize the execve() semaphore */
-       ret = semaphore_create(kernel_task, &execve_semaphore,
-                       SYNC_POLICY_FIFO, (BSD_PAGABLE_MAP_SIZE / NCARGS));
-       if (ret != KERN_SUCCESS)
-               panic("bsd_init: Failed to create execve semaphore");
-
        /*
         * Initialize the calendar.
         */
-       IOKitResetTime();
-
-       ubc_init();
+       bsd_init_kprintf("calling IOKitInitializeTime\n");
+       IOKitInitializeTime();
 
        /* Initialize the file systems. */
+       bsd_init_kprintf("calling vfsinit\n");
        vfsinit();
 
+#if CONFIG_PROC_UUID_POLICY
+       /* Initial proc_uuid_policy subsystem */
+       bsd_init_kprintf("calling proc_uuid_policy_init()\n");
+       proc_uuid_policy_init();
+#endif
+
+#if SOCKETS
+       /* Initialize per-CPU cache allocator */
+       mcache_init();
+
        /* Initialize mbuf's. */
+       bsd_init_kprintf("calling mbinit\n");
        mbinit();
+       net_str_id_init(); /* for mbuf tags */
+       restricted_in_port_init();
+#endif /* SOCKETS */
 
        /*
         * Initializes security event auditing.
         * XXX: Should/could this occur later?
         */
-       audit_init();  
+#if CONFIG_AUDIT
+       bsd_init_kprintf("calling audit_init\n");
+       audit_init();
+#endif
 
        /* Initialize kqueues */
+       bsd_init_kprintf("calling knote_init\n");
        knote_init();
 
+       /* Initialize event handler */
+       bsd_init_kprintf("calling eventhandler_init\n");
+       eventhandler_init();
+
        /* Initialize for async IO */
+       bsd_init_kprintf("calling aio_init\n");
        aio_init();
 
-       /* Initialize pipes */
-       pipeinit();
-
-       /* Initialize SysV shm subsystem locks; the subsystem proper is
-        * initialized through a sysctl.
-        */
-       sysv_shm_lock_init();
-       sysv_sem_lock_init();
-       sysv_msg_lock_init();
-       pshm_lock_init();
-       psem_lock_init();
-
+       pthread_init();
        /* POSIX Shm and Sem */
+       bsd_init_kprintf("calling pshm_cache_init\n");
        pshm_cache_init();
+       bsd_init_kprintf("calling psem_cache_init\n");
        psem_cache_init();
-       time_zone_slock_init();
+       bsd_init_kprintf("calling select_waitq_init\n");
+       select_waitq_init();
 
-       /* Stack snapshot facility lock */
-       stackshot_lock_init();
        /*
         * Initialize protocols.  Block reception of incoming packets
         * until everything is ready.
         */
-       sysctl_register_fixed(); 
-       sysctl_mib_init();
+#if NETWORKING
+       bsd_init_kprintf("calling nwk_wq_init\n");
+       nwk_wq_init();
+       bsd_init_kprintf("calling dlil_init\n");
        dlil_init();
+       bsd_init_kprintf("calling proto_kpi_init\n");
        proto_kpi_init();
+#endif /* NETWORKING */
+#if SOCKETS
+       bsd_init_kprintf("calling socketinit\n");
        socketinit();
+       bsd_init_kprintf("calling domaininit\n");
        domaininit();
-
-       p->p_fd->fd_cdir = NULL;
-       p->p_fd->fd_rdir = NULL;
-
-#ifdef GPROF
-       /* Initialize kernel profiling. */
-       kmstartup();
+       iptap_init();
+#if FLOW_DIVERT
+       flow_divert_init();
+#endif  /* FLOW_DIVERT */
+#endif /* SOCKETS */
+#if NETWORKING
+#if NECP
+       /* Initialize Network Extension Control Policies */
+       necp_init();
+#endif
+       netagent_init();
+#endif /* NETWORKING */
+       kernproc->p_fd->fd_cdir = NULL;
+       kernproc->p_fd->fd_rdir = NULL;
+
+#if defined (__x86_64__) && (DEBUG || DEVELOPMENT)
+       hvg_bsd_init();
+#endif /* DEBUG || DEVELOPMENT */
+
+#if CONFIG_FREEZE
+#ifndef CONFIG_MEMORYSTATUS
+    #error "CONFIG_FREEZE defined without matching CONFIG_MEMORYSTATUS"
+#endif
+       /* Initialise background freezing */
+       bsd_init_kprintf("calling memorystatus_freeze_init\n");
+       memorystatus_freeze_init();
 #endif
 
-       /* kick off timeout driven events by calling first time */
-       thread_wakeup(&lbolt);
-       timeout((void (*)(void *))lightning_bolt, 0, hz);
+#if CONFIG_MEMORYSTATUS
+       /* Initialize kernel memory status notifications */
+       bsd_init_kprintf("calling memorystatus_init\n");
+       memorystatus_init();
+#endif /* CONFIG_MEMORYSTATUS */
 
+       bsd_init_kprintf("calling sysctl_mib_init\n");
+       sysctl_mib_init();
+
+       bsd_init_kprintf("calling bsd_autoconf\n");
        bsd_autoconf();
 
+#if CONFIG_DTRACE
+       dtrace_postinit();
+#endif
+
        /*
         * We attach the loopback interface *way* down here to ensure
         * it happens after autoconf(), otherwise it becomes the
@@ -523,180 +867,346 @@ bsd_init()
         */
 #include <loop.h>
 #if NLOOP > 0
-       loopattach();                   /* XXX */
+       bsd_init_kprintf("calling loopattach\n");
+       loopattach();                   /* XXX */
+#endif
+#if NGIF
+       /* Initialize gif interface (after lo0) */
+       gif_init();
 #endif
-        
-        /* Register the built-in dlil ethernet interface family */
+
+#if PFLOG
+       /* Initialize packet filter log interface */
+       pfloginit();
+#endif /* PFLOG */
+
+#if NETHER > 0
+       /* Register the built-in dlil ethernet interface family */
+       bsd_init_kprintf("calling ether_family_init\n");
        ether_family_init();
+#endif /* ETHER */
 
-       /* Call any kext code that wants to run just after network init */
-       net_init_run();
+#if NETWORKING
+#if CONTENT_FILTER
+       cfil_init();
+#endif
 
-       vnode_pager_bootstrap();
-#if 0
-       /* XXX Hack for early debug stop */
-       printf("\nabout to sleep for 10 seconds\n");
-       IOSleep( 10 * 1000 );
-       /* Debugger("hello"); */
+#if PACKET_MANGLER
+       pkt_mnglr_init();
 #endif
 
+       /*
+        * Register subsystems with kernel control handlers
+        */
+       utun_register_control();
+#if IPSEC
+       ipsec_register_control();
+#endif /* IPSEC */
+       netsrc_init();
+       nstat_init();
+       tcp_cc_init();
+#if MPTCP
+       mptcp_control_register();
+#endif /* MPTCP */
+
+       /*
+        * The the networking stack is now initialized so it is a good time to call
+        * the clients that are waiting for the networking stack to be usable.
+        */
+       bsd_init_kprintf("calling net_init_run\n");
+       net_init_run();
+#endif /* NETWORKING */
+
+       bsd_init_kprintf("calling inittodr\n");
        inittodr(0);
 
        /* Mount the root file system. */
-       whileTRUE) {
+       while (TRUE) {
                int err;
 
+               bsd_init_kprintf("calling setconf\n");
                setconf();
-               bsd_hardclockinit = -1; /* start ticking */
+#if CONFIG_NETBOOT
+               netboot = (mountroot == netboot_mountroot);
+#endif
 
-               if (0 == (err = vfs_mountroot()))
+               bsd_init_kprintf("vfs_mountroot\n");
+               if (0 == (err = vfs_mountroot())) {
                        break;
-#if NFSCLIENT
-               if (mountroot == netboot_mountroot) {
-                       printf("cannot mount network root, errno = %d\n", err);
-                       mountroot = NULL;
-                       if (0 == (err = vfs_mountroot()))
-                               break;
+               }
+               rootdevice[0] = '\0';
+#if CONFIG_NETBOOT
+               if (netboot) {
+                       PE_display_icon( 0, "noroot");  /* XXX a netboot-specific icon would be nicer */
+                       vc_progress_set(FALSE, 0);
+                       for (i = 1; 1; i *= 2) {
+                               printf("bsd_init: failed to mount network root, error %d, %s\n",
+                                   err, PE_boot_args());
+                               printf("We are hanging here...\n");
+                               IOSleep(i * 60 * 1000);
+                       }
+                       /*NOTREACHED*/
                }
 #endif
                printf("cannot mount root, errno = %d\n", err);
                boothowto |= RB_ASKNAME;
        }
 
-       context.vc_proc = p;
-       context.vc_ucred = p->p_ucred;
+       IOSecureBSDRoot(rootdevice);
+
+       context.vc_thread = current_thread();
+       context.vc_ucred = kernproc->p_ucred;
        mountlist.tqh_first->mnt_flag |= MNT_ROOTFS;
 
+       bsd_init_kprintf("calling VFS_ROOT\n");
        /* Get the vnode for '/'.  Set fdp->fd_fd.fd_cdir to reference it. */
-       if (VFS_ROOT(mountlist.tqh_first, &rootvnode, &context))
-               panic("bsd_init: cannot find root vnode");
-       rootvnode->v_flag |= VROOT;
-       (void)vnode_ref(rootvnode);
-       (void)vnode_put(rootvnode);
-       filedesc0.fd_cdir = rootvnode;
-
-#if NFSCLIENT
-       if (mountroot == netboot_mountroot) {
+       if (VFS_ROOT(mountlist.tqh_first, &init_rootvnode, &context)) {
+               panic("bsd_init: cannot find root vnode: %s", PE_boot_args());
+       }
+       (void)vnode_ref(init_rootvnode);
+       (void)vnode_put(init_rootvnode);
+
+       lck_rw_lock_exclusive(&rootvnode_rw_lock);
+       set_rootvnode(init_rootvnode);
+       lck_rw_unlock_exclusive(&rootvnode_rw_lock);
+       init_rootvnode = NULLVP;  /* use rootvnode after this point */
+
+
+       if (!bsd_rooted_ramdisk()) {
+#if CONFIG_IMAGEBOOT
+#if XNU_TARGET_OS_OSX && defined(__arm64__)
+               /* Apple Silicon MacOS */
+               if (!imageboot_desired()) {
+                       /* enforce sealedness */
+                       int autherr = VNOP_IOCTL(rootvnode, FSIOC_KERNEL_ROOTAUTH, NULL, 0, vfs_context_kernel());
+                       if (autherr) {
+                               panic("rootvp not authenticated after mounting \n");
+                       }
+               }
+#endif // TARGET_OS_OSX && arm64
+#endif // config_imageboot
+               /* Otherwise, noop */
+       }
+
+
+#if CONFIG_NETBOOT
+       if (netboot) {
                int err;
+
+               netboot = TRUE;
                /* post mount setup */
-               if (err = netboot_setup(p)) {
-                       panic("bsd_init: NetBoot could not find root, %d", err);
+               if ((err = netboot_setup()) != 0) {
+                       PE_display_icon( 0, "noroot");  /* XXX a netboot-specific icon would be nicer */
+                       vc_progress_set(FALSE, 0);
+                       for (i = 1; 1; i *= 2) {
+                               printf("bsd_init: NetBoot could not find root, error %d: %s\n",
+                                   err, PE_boot_args());
+                               printf("We are hanging here...\n");
+                               IOSleep(i * 60 * 1000);
+                       }
+                       /*NOTREACHED*/
                }
        }
 #endif
-       
 
-       microtime(&p->p_stats->p_start);
-       p->p_rtime.tv_sec = p->p_rtime.tv_usec = 0;
+
+#if CONFIG_IMAGEBOOT
+#if CONFIG_LOCKERBOOT
+       /*
+        * Stash the protoboot vnode, mount, filesystem name, and device name for
+        * later use. Note that the mount-from name may not have the "/dev/"
+        * component, so we must sniff out this condition and add it as needed.
+        */
+       pbvn = rootvnode;
+       pbmnt = pbvn->v_mount;
+       pbdevp = vfs_statfs(pbmnt)->f_mntfromname;
+       slash_dev = strnstr(pbdevp, "/dev/", strlen(pbdevp));
+       if (slash_dev) {
+               /*
+                * If the old root is a snapshot mount, it will have the form:
+                *
+                *     com.apple.os.update-<boot manifest hash>@<dev node path>
+                *
+                * So we just search the mntfromname for any occurrence of "/dev/" and
+                * grab that as the device path. The image boot code needs a dev node to
+                * do the re-mount, so we cannot directly mount the snapshot as the
+                * protoboot volume currently.
+                */
+               strlcpy(pbdevpath, slash_dev, sizeof(pbdevpath));
+       } else {
+               snprintf(pbdevpath, sizeof(pbdevpath), "/dev/%s", pbdevp);
+       }
+
+       bsd_init_kprintf("protoboot mount-from: %s\n", pbdevp);
+       bsd_init_kprintf("protoboot dev path: %s\n", pbdevpath);
+
+       strlcpy(pbfsname, pbmnt->mnt_vtable->vfc_name, sizeof(pbfsname));
+#endif
+
+       /*
+        * See if a system disk image is present. If so, mount it and
+        * switch the root vnode to point to it
+        */
+       imageboot_type_t imageboot_type = imageboot_needed();
+       if (netboot == FALSE && imageboot_type) {
+               /*
+                * An image was found.  No turning back: we're booted
+                * with a kernel from the disk image.
+                */
+               bsd_init_kprintf("doing image boot: type = %d\n", imageboot_type);
+               imageboot_setup(imageboot_type);
+       }
+
+#if CONFIG_LOCKERBOOT
+       if (imageboot_type == IMAGEBOOT_LOCKER) {
+               bsd_init_kprintf("booting from locker\n");
+               if (vnode_tag(rootvnode) != VT_LOCKERFS) {
+                       panic("root filesystem not a locker: fsname = %s",
+                           rootvnode->v_mount->mnt_vtable->vfc_name);
+               }
+       }
+#endif /* CONFIG_LOCKERBOOT */
+#endif /* CONFIG_IMAGEBOOT */
+
+       /* set initial time; all other resource data is  already zero'ed */
+       microtime_with_abstime(&kernproc->p_start, &kernproc->p_stats->ps_start);
 
 #if DEVFS
        {
-           extern void devfs_kernel_mount(char * str);
-           
-           devfs_kernel_mount("/dev");
+               char mounthere[] = "/dev"; /* !const because of internal casting */
+
+               bsd_init_kprintf("calling devfs_kernel_mount\n");
+               devfs_kernel_mount(mounthere);
        }
 #endif /* DEVFS */
-       
+
+#if CONFIG_BASESYSTEMROOT
+#if CONFIG_IMAGEBOOT
+       if (bsdmgroot_bootable()) {
+               int error;
+               bool rooted_dmg = false;
+               bool skip_signature_check = false;
+
+               printf("trying to find and mount BaseSystem dmg as root volume\n");
+#if DEVELOPMENT || DEBUG
+               printf("(set boot-arg -nobsdmgroot to avoid this)\n");
+#endif // DEVELOPMENT || DEBUG
+
+               char *dmgpath = NULL;
+               dmgpath = zalloc_flags(ZV_NAMEI, Z_ZERO | Z_WAITOK);
+               if (dmgpath == NULL) {
+                       panic("%s: M_NAMEI zone exhausted", __FUNCTION__);
+               }
+
+               error = bsd_find_basesystem_dmg(dmgpath, &rooted_dmg, &skip_signature_check);
+               if (error) {
+                       bsd_init_kprintf("failed to to find BaseSystem dmg: error = %d\n", error);
+               } else {
+                       PE_parse_boot_argn("bsdmgpath", dmgpath, sizeof(dmgpath));
+
+                       bsd_init_kprintf("found BaseSystem dmg at: %s\n", dmgpath);
+
+                       error = imageboot_pivot_image(dmgpath, IMAGEBOOT_DMG, "/System/Volumes/BaseSystem", "System/Volumes/macOS", rooted_dmg, skip_signature_check);
+                       if (error) {
+                               bsd_init_kprintf("couldn't mount BaseSystem dmg: error = %d", error);
+                       }
+               }
+               zfree(ZV_NAMEI, dmgpath);
+       }
+#else /* CONFIG_IMAGEBOOT */
+#error CONFIG_BASESYSTEMROOT requires CONFIG_IMAGEBOOT
+#endif /* CONFIG_IMAGEBOOT */
+#endif /* CONFIG_BASESYSTEMROOT */
+
+#if CONFIG_LOCKERBOOT
+       /*
+        * We need to wait until devfs is up before remounting the protoboot volume
+        * within the locker so that it can have a real devfs vnode backing it.
+        */
+       if (imageboot_type == IMAGEBOOT_LOCKER) {
+               bsd_init_kprintf("re-mounting protoboot volume\n");
+               int error = mount_locker_protoboot(pbfsname, LOCKER_PROTOBOOT_MOUNT,
+                   pbdevpath);
+               if (error) {
+                       panic("failed to mount protoboot volume: dev path = %s, error = %d",
+                           pbdevpath, error);
+               }
+       }
+#endif /* CONFIG_LOCKERBOOT */
+
        /* Initialize signal state for process 0. */
-       siginit(p);
+       bsd_init_kprintf("calling siginit\n");
+       siginit(kernproc);
 
+       bsd_init_kprintf("calling bsd_utaskbootstrap\n");
        bsd_utaskbootstrap();
 
-#if __i386__
-       // this should be done after the root filesystem is mounted
-       error = set_archhandler(p, CPU_TYPE_POWERPC);
-       if (error)
-               exec_archhandler_ppc.path[0] = 0;
-#endif 
-       
+       pal_kernel_announce();
+
+       bsd_init_kprintf("calling mountroot_post_hook\n");
+
+#if XNU_TARGET_OS_OSX
        /* invoke post-root-mount hook */
-       if (mountroot_post_hook != NULL)
+       if (mountroot_post_hook != NULL) {
                mountroot_post_hook();
-       
-       (void) thread_funnel_set(kernel_flock, funnel_state);
+       }
+#endif
+
+#if 0 /* not yet */
+       consider_zone_gc(FALSE);
+#endif
+
+       bsd_init_kprintf("done\n");
 }
 
-/* Called with kernel funnel held */
 void
 bsdinit_task(void)
 {
-       struct proc *p = current_proc();
-       struct uthread *ut;
-       kern_return_t   kr;
-       thread_t th_act;
-       shared_region_mapping_t system_region;
+       proc_t p = current_proc();
 
        process_name("init", p);
 
-       ux_handler_init();
-
-       th_act = current_thread();
-       (void) host_set_exception_ports(host_priv_self(),
-                                       EXC_MASK_ALL & ~(EXC_MASK_SYSCALL |
-                                                        EXC_MASK_MACH_SYSCALL |
-                                                        EXC_MASK_RPC_ALERT),
-                                       ux_exception_port,
-                                       EXCEPTION_DEFAULT, 0);
-
-       (void) task_set_exception_ports(get_threadtask(th_act),
-                                       EXC_MASK_ALL & ~(EXC_MASK_SYSCALL |
-                                                        EXC_MASK_MACH_SYSCALL |
-                                                        EXC_MASK_RPC_ALERT),
-                                       ux_exception_port,
-                                       EXCEPTION_DEFAULT, 0);
+       /* Set up exception-to-signal reflection */
+       ux_handler_setup();
 
+#if CONFIG_MACF
+       mac_cred_label_associate_user(p->p_ucred);
+#endif
 
+       vm_init_before_launchd();
 
+#if CONFIG_XNUPOST
+       int result = bsd_list_tests();
+       result = bsd_do_post();
+       if (result != 0) {
+               panic("bsd_do_post: Tests failed with result = 0x%08x\n", result);
+       }
+#endif
 
-       ut = (uthread_t)get_bsdthread_info(th_act);
+       bsd_init_kprintf("bsd_do_post - done");
 
-       bsd_hardclockinit = 1;  /* Start bsd hardclock */
-       bsd_init_task = get_threadtask(th_act);
-       init_task_failure_data[0] = 0;
-       system_region = lookup_default_shared_region(ENV_DEFAULT_ROOT, cpu_type());
-        if (system_region == NULL) {
-               shared_file_boot_time_init(ENV_DEFAULT_ROOT, cpu_type());
-       } else {
-               vm_set_shared_region(get_threadtask(th_act), system_region);
-       }
        load_init_program(p);
-       /* turn on app-profiling i.e. pre-heating */
-       app_profile = 1;
        lock_trace = 1;
 }
 
-void
-lightning_bolt()
-{                      
-       boolean_t       funnel_state;
-       extern void klogwakeup(void);
-
-       funnel_state = thread_funnel_set(kernel_flock, TRUE);
-
-       thread_wakeup(&lbolt);
-       timeout(lightning_bolt,0,hz);
-       klogwakeup();
-
-       (void) thread_funnel_set(kernel_flock, FALSE);
-}
-
-bsd_autoconf()
+kern_return_t
+bsd_autoconf(void)
 {
-       extern kern_return_t IOKitBSDInit( void );
-
+       kprintf("bsd_autoconf: calling kminit\n");
        kminit();
 
-       /* 
+       /*
         * Early startup for bsd pseudodevices.
         */
        {
-           struct pseudo_init *pi;
-       
-           for (pi = pseudo_inits; pi->ps_func; pi++)
-               (*pi->ps_func) (pi->ps_count);
+               struct pseudo_init *pi;
+
+               for (pi = pseudo_inits; pi->ps_func; pi++) {
+                       (*pi->ps_func)(pi->ps_count);
+               }
        }
 
-       return( IOKitBSDInit());
+       return IOKitBSDInit();
 }
 
 
@@ -704,131 +1214,553 @@ bsd_autoconf()
 
 static void
 setconf(void)
-{      
-       extern kern_return_t IOFindBSDRoot( char * rootName,
-                               dev_t * root, u_int32_t * flags );
-       u_int32_t       flags;
-       kern_return_t   err;
+{
+       u_int32_t       flags;
+       kern_return_t   err;
 
-       /*
-        * calls into IOKit can generate networking registrations
-        * which needs to be under network funnel. Right thing to do
-        * here is to drop the funnel alltogether and regrab it afterwards
-        */
-       thread_funnel_set(kernel_flock, FALSE);
-       err = IOFindBSDRoot( rootdevice, &rootdev, &flags );
-       thread_funnel_set(kernel_flock, TRUE);
-       if( err) {
+       err = IOFindBSDRoot(rootdevice, sizeof(rootdevice), &rootdev, &flags);
+       if (err) {
                printf("setconf: IOFindBSDRoot returned an error (%d);"
-                       "setting rootdevice to 'sd0a'.\n", err); /* XXX DEBUG TEMP */
+                   "setting rootdevice to 'sd0a'.\n", err);     /* XXX DEBUG TEMP */
                rootdev = makedev( 6, 0 );
-               strcpy( rootdevice, "sd0a" );
+               strlcpy(rootdevice, "sd0a", sizeof(rootdevice));
                flags = 0;
        }
 
-#if NFSCLIENT
-       if( flags & 1 ) {
+#if CONFIG_NETBOOT
+       if (flags & 1) {
                /* network device */
                mountroot = netboot_mountroot;
        } else {
 #endif
-               /* otherwise have vfs determine root filesystem */
-               mountroot = NULL;
-#if NFSCLIENT
-       }
+       /* otherwise have vfs determine root filesystem */
+       mountroot = NULL;
+#if CONFIG_NETBOOT
+}
 #endif
+}
 
+/*
+ * Boot into the flavor of Recovery dictated by `mode`.
+ */
+boolean_t
+bsd_boot_to_recovery(bsd_bootfail_mode_t mode, uuid_t volume_uuid, boolean_t reboot)
+{
+       return IOSetRecoveryBoot(mode, volume_uuid, reboot);
 }
 
-bsd_utaskbootstrap()
+void
+bsd_utaskbootstrap(void)
 {
-       thread_t th_act;
+       thread_t thread;
        struct uthread *ut;
 
-       th_act = cloneproc(kernproc, 0);
-       initproc = pfind(1);                            
-       /* Set the launch time for init */
-       microtime(&initproc->p_stats->p_start);
+       /*
+        * Clone the bootstrap process from the kernel process, without
+        * inheriting either task characteristics or memory from the kernel;
+        */
+       thread = cloneproc(TASK_NULL, COALITION_NULL, kernproc, FALSE, TRUE);
+
+       /* Hold the reference as it will be dropped during shutdown */
+       initproc = proc_find(1);
+#if __PROC_INTERNAL_DEBUG
+       if (initproc == PROC_NULL) {
+               panic("bsd_utaskbootstrap: initproc not set\n");
+       }
+#endif
+
+       zalloc_first_proc_made();
+
+       /*
+        * Since we aren't going back out the normal way to our parent,
+        * we have to drop the transition locks explicitly.
+        */
+       proc_signalend(initproc, 0);
+       proc_transend(initproc, 0);
 
-       ut = (struct uthread *)get_bsdthread_info(th_act);
+       ut = (struct uthread *)get_bsdthread_info(thread);
        ut->uu_sigmask = 0;
-       act_set_astbsd(th_act);
-       (void) thread_resume(th_act);
+       act_set_astbsd(thread);
+       task_clear_return_wait(get_threadtask(thread), TCRW_CLEAR_ALL_WAIT);
 }
 
-parse_bsd_args()
+static void
+parse_bsd_args(void)
 {
-       extern char init_args[];
-       char    namep[16];
-       extern int boothowto;
-       extern int srv;
-       extern int ncl;
-
-       int len;
+       char namep[48];
+       int msgbuf;
 
-       if (PE_parse_boot_arg("-s", namep)) {
+       if (PE_parse_boot_argn("-s", namep, sizeof(namep))) {
                boothowto |= RB_SINGLE;
-               len = strlen(init_args);
-               if(len != 0)
-                       strcat(init_args," -s");
-               else
-                       strcat(init_args,"-s");
        }
 
-       if (PE_parse_boot_arg("-b", namep)) {
-               boothowto |= RB_NOBOOTRC;
-               len = strlen(init_args);
-               if(len != 0)
-                       strcat(init_args," -b");
-               else
-                       strcat(init_args,"-b");
+       if (PE_parse_boot_argn("-x", namep, sizeof(namep))) { /* safe boot */
+               boothowto |= RB_SAFEBOOT;
        }
 
-       if (PE_parse_boot_arg("-F", namep)) {
-               len = strlen(init_args);
-               if(len != 0)
-                       strcat(init_args," -F");
-               else
-                       strcat(init_args,"-F");
+       if (PE_parse_boot_argn("-minimalboot", namep, sizeof(namep))) {
+               /*
+                * -minimalboot indicates that we want userspace to be bootstrapped to a
+                * minimal environment.  What constitutes minimal is up to the bootstrap
+                * process.
+                */
+               minimalboot = 1;
        }
 
-       if (PE_parse_boot_arg("-v", namep)) {
-               len = strlen(init_args);
-               if(len != 0)
-                       strcat(init_args," -v");
-               else
-                       strcat(init_args,"-v");
+#if __x86_64__
+       int no32exec;
+
+       /* disable 32 bit grading */
+       if (PE_parse_boot_argn("no32exec", &no32exec, sizeof(no32exec))) {
+               bootarg_no32exec = !!no32exec;
+       }
+#endif
+
+       int execfailure_crashreports;
+       /* enable crash reports on various exec failures */
+       if (PE_parse_boot_argn("execfailurecrashes", &execfailure_crashreports, sizeof(execfailure_crashreports))) {
+               bootarg_execfailurereports = !!execfailure_crashreports;
        }
 
-       if (PE_parse_boot_arg("-x", namep)) { /* safe boot */
-               len = strlen(init_args);
-               if(len != 0)
-                       strcat(init_args," -x");
-               else
-                       strcat(init_args,"-x");
+       /* disable vnode_cache_is_authorized() by setting vnode_cache_defeat */
+       if (PE_parse_boot_argn("-vnode_cache_defeat", namep, sizeof(namep))) {
+               bootarg_vnode_cache_defeat = 1;
        }
 
-       if (PE_parse_boot_arg("-d", namep)) {
-               len = strlen(init_args);
-               if(len != 0)
-                       strcat(init_args," -d");
-               else
-                       strcat(init_args,"-d");
+#if DEVELOPMENT || DEBUG
+       if (PE_parse_boot_argn("-disable_aslr", namep, sizeof(namep))) {
+               bootarg_disable_aslr = 1;
        }
+#endif
+
 
-       PE_parse_boot_arg("srv", &srv);
-       PE_parse_boot_arg("ncl", &ncl);
-       if (PE_parse_boot_arg("nbuf", &max_nbuf_headers)) 
+
+       PE_parse_boot_argn("ncl", &ncl, sizeof(ncl));
+       if (PE_parse_boot_argn("nbuf", &max_nbuf_headers,
+           sizeof(max_nbuf_headers))) {
                customnbuf = 1;
-       PE_parse_boot_arg("kmem", &setup_kmem);
+       }
 
-       return 0;
+#if CONFIG_MACF
+#if defined (__i386__) || defined (__x86_64__)
+       PE_parse_boot_argn("policy_check", &policy_check_flags, sizeof(policy_check_flags));
+#endif
+#endif  /* CONFIG_MACF */
+
+       if (PE_parse_boot_argn("msgbuf", &msgbuf, sizeof(msgbuf))) {
+               log_setsize(msgbuf);
+               oslog_setsize(msgbuf);
+       }
+
+       if (PE_parse_boot_argn("-novfscache", namep, sizeof(namep))) {
+               nc_disabled = 1;
+       }
+
+#if CONFIG_JETSAM && (DEVELOPMENT || DEBUG)
+       if (PE_parse_boot_argn("-no_vnode_jetsam", namep, sizeof(namep))) {
+               bootarg_no_vnode_jetsam = 1;
+       }
+#endif /* CONFIG_JETSAM && (DEVELOPMENT || DEBUG) */
+
+       if (PE_parse_boot_argn("-no_vnode_drain", namep, sizeof(namep))) {
+               bootarg_no_vnode_drain = 1;
+       }
+
+#if CONFIG_DARKBOOT
+       /*
+        * The darkboot flag is specified by the bootloader and is stored in
+        * boot_args->bootFlags. This flag is available starting revision 2.
+        */
+       boot_args *args = (boot_args *) PE_state.bootArgs;
+       if ((args != NULL) && (args->Revision >= kBootArgsRevision2)) {
+               darkboot = (args->bootFlags & kBootFlagsDarkBoot) ? 1 : 0;
+       } else {
+               darkboot = 0;
+       }
+#endif
+
+#if PROC_REF_DEBUG
+       if (PE_parse_boot_argn("-disable_procref_tracking", namep, sizeof(namep))) {
+               proc_ref_tracking_disabled = 1;
+       }
+#endif
+
+       PE_parse_boot_argn("sigrestrict", &sigrestrict_arg, sizeof(sigrestrict_arg));
+
+#if DEVELOPMENT || DEBUG
+       if (PE_parse_boot_argn("-no_sigsys", namep, sizeof(namep))) {
+               send_sigsys = false;
+       }
+
+       if (PE_parse_boot_argn("alt-dyld", dyld_alt_path, sizeof(dyld_alt_path))) {
+               if (strlen(dyld_alt_path) > 0) {
+                       use_alt_dyld = 1;
+               }
+       }
+       PE_parse_boot_argn("dyld_flags", &dyld_flags, sizeof(dyld_flags));
+
+       if (PE_parse_boot_argn("-disable_syscallfilter", &namep, sizeof(namep))) {
+               syscallfilter_disable = 1;
+       }
+
+#if __arm64__
+       if (PE_parse_boot_argn("legacy_footprint_entitlement_mode", &legacy_footprint_entitlement_mode, sizeof(legacy_footprint_entitlement_mode))) {
+               /*
+                * legacy_footprint_entitlement_mode specifies the behavior we want associated
+                * with the entitlement. The supported modes are:
+                *
+                * LEGACY_FOOTPRINT_ENTITLEMENT_IGNORE:
+                *      Indicates that we want every process to have the memory accounting
+                *      that is available in iOS 12.0 and beyond.
+                *
+                * LEGACY_FOOTPRINT_ENTITLEMENT_IOS11_ACCT:
+                *      Indicates that for every process that has the 'legacy footprint entitlement',
+                *      we want to give it the old iOS 11.0 accounting behavior which accounted some
+                *      of the process's memory to the kernel.
+                *
+                * LEGACY_FOOTPRINT_ENTITLEMENT_LIMIT_INCREASE:
+                *      Indicates that for every process that has the 'legacy footprint entitlement',
+                *      we want it to have a higher memory limit which will help them acclimate to the
+                *      iOS 12.0 (& beyond) accounting behavior that does the right accounting.
+                *      The bonus added to the system-wide task limit to calculate this higher memory limit
+                *      is available in legacy_footprint_bonus_mb.
+                */
+
+               if (legacy_footprint_entitlement_mode < LEGACY_FOOTPRINT_ENTITLEMENT_IGNORE ||
+                   legacy_footprint_entitlement_mode > LEGACY_FOOTPRINT_ENTITLEMENT_LIMIT_INCREASE) {
+                       legacy_footprint_entitlement_mode = LEGACY_FOOTPRINT_ENTITLEMENT_LIMIT_INCREASE;
+               }
+       }
+#endif /* __arm64__ */
+#endif /* DEVELOPMENT || DEBUG */
+}
+
+#if CONFIG_BASESYSTEMROOT
+
+extern const char* IOGetBootUUID(void);
+extern const char* IOGetApfsPrebootUUID(void);
+
+// Get the UUID of the Preboot (and Recovery) folder associated with the
+// current boot volume, if applicable. The meaning of the UUID can be
+// filesystem-dependent and not all kinds of boots will have a UUID.
+// If available, the string will be returned. It does not need to be
+// deallocate. (Future: if we need to return the string as a copy that the
+// caller must free, we'll introduce a new functcion for that.)
+// NULL will be returned if the current boot has no applicable Preboot UUID.
+static
+const char *
+get_preboot_uuid(void)
+{
+       const char *maybe_uuid_string;
+
+       // try IOGetApfsPrebootUUID
+       maybe_uuid_string = IOGetApfsPrebootUUID();
+       if (maybe_uuid_string) {
+               uuid_t maybe_uuid;
+               int error = uuid_parse(maybe_uuid_string, maybe_uuid);
+               if (error == 0) {
+                       return maybe_uuid_string;
+               }
+       }
+
+       // try IOGetBootUUID
+       maybe_uuid_string = IOGetBootUUID();
+       if (maybe_uuid_string) {
+               uuid_t maybe_uuid;
+               int error = uuid_parse(maybe_uuid_string, maybe_uuid);
+               if (error == 0) {
+                       return maybe_uuid_string;
+               }
+       }
+
+       // didn't find it
+       return NULL;
+}
+
+#if defined(__arm64__)
+extern const char *IOGetBootObjectsPath(void);
+#endif
+
+// Find the BaseSystem.dmg to be used as the initial root volume during certain
+// kinds of boots.
+// This may mount volumes and lookup vnodes.
+// The DEVELOPMENT kernel will look for BaseSystem.rooted.dmg first.
+// If it returns 0 (no error), then it also writes the absolute path to the
+// BaseSystem.dmg into its argument (which must be a char[MAXPATHLEN]).
+static
+int
+bsd_find_basesystem_dmg(char *bsdmgpath_out, bool *rooted_dmg, bool *skip_signature_check)
+{
+       int error;
+       size_t len;
+       char *dmgbasepath;
+       char *dmgpath;
+       bool allow_rooted_dmg = false;
+
+       dmgbasepath = zalloc_flags(ZV_NAMEI, Z_ZERO | Z_WAITOK);
+       dmgpath = zalloc_flags(ZV_NAMEI, Z_ZERO | Z_WAITOK);
+       vnode_t imagevp = NULLVP;
+
+#if DEVELOPMENT || DEBUG
+       allow_rooted_dmg = true;
+#endif
+
+       //must provide output bool
+       if (rooted_dmg && skip_signature_check) {
+               *rooted_dmg = false;
+               *skip_signature_check = false;
+       } else {
+               error = EINVAL;
+               goto done;
+       }
+
+       error = vfs_mount_recovery();
+       if (error) {
+               goto done;
+       }
+
+       len = strlcpy(dmgbasepath, "/System/Volumes/Recovery/", MAXPATHLEN);
+       if (len > MAXPATHLEN) {
+               error = ENAMETOOLONG;
+               goto done;
+       }
+
+       if (csr_check(CSR_ALLOW_ANY_RECOVERY_OS) == 0) {
+               *skip_signature_check = true;
+               allow_rooted_dmg = true;
+       }
+
+#if defined(__arm64__)
+       const char *boot_obj_path = IOGetBootObjectsPath();
+       if (boot_obj_path) {
+               if (boot_obj_path[0] == '/') {
+                       dmgbasepath[len - 1] = '\0';
+               }
+
+               len = strlcat(dmgbasepath, boot_obj_path, MAXPATHLEN);
+               if (len > MAXPATHLEN) {
+                       error = ENAMETOOLONG;
+                       goto done;
+               }
+
+               len = strlcat(dmgbasepath, "/usr/standalone/firmware/", MAXPATHLEN);
+               if (len > MAXPATHLEN) {
+                       error = ENAMETOOLONG;
+                       goto done;
+               }
+
+               if (allow_rooted_dmg) {
+                       len = strlcpy(dmgpath, dmgbasepath, MAXPATHLEN);
+                       if (len > MAXPATHLEN) {
+                               error = ENAMETOOLONG;
+                               goto done;
+                       }
+
+                       len = strlcat(dmgpath, "arm64eBaseSystem.rooted.dmg", MAXPATHLEN);
+                       if (len > MAXPATHLEN) {
+                               error = ENAMETOOLONG;
+                               goto done;
+                       }
+
+                       error = vnode_lookup(dmgpath, 0, &imagevp, vfs_context_kernel());
+                       if (error == 0) {
+                               *rooted_dmg = true;
+                               *skip_signature_check = true;
+                               goto done;
+                       }
+                       memset(dmgpath, 0, MAXPATHLEN);
+               }
+
+               len = strlcpy(dmgpath, dmgbasepath, MAXPATHLEN);
+               if (len > MAXPATHLEN) {
+                       error = ENAMETOOLONG;
+                       goto done;
+               }
+
+               len = strlcat(dmgpath, "arm64eBaseSystem.dmg", MAXPATHLEN);
+               if (len > MAXPATHLEN) {
+                       error = ENAMETOOLONG;
+                       goto done;
+               }
+
+               error = vnode_lookup(dmgpath, 0, &imagevp, vfs_context_kernel());
+               if (error == 0) {
+                       goto done;
+               }
+               memset(dmgpath, 0, MAXPATHLEN);
+               dmgbasepath[strlen("/System/Volumes/Recovery/")] = '\0';
+       }
+#endif // __arm64__
+
+       const char *preboot_uuid = get_preboot_uuid();
+       if (preboot_uuid == NULL) {
+               // no preboot? bail out
+               return EINVAL;
+       }
+
+       len = strlcat(dmgbasepath, preboot_uuid, MAXPATHLEN);
+       if (len > MAXPATHLEN) {
+               error = ENAMETOOLONG;
+               goto done;
+       }
+
+       if (allow_rooted_dmg) {
+               // Try BaseSystem.rooted.dmg
+               len = strlcpy(dmgpath, dmgbasepath, MAXPATHLEN);
+               if (len > MAXPATHLEN) {
+                       error = ENAMETOOLONG;
+                       goto done;
+               }
+
+               len = strlcat(dmgpath, "/BaseSystem.rooted.dmg", MAXPATHLEN);
+               if (len > MAXPATHLEN) {
+                       error = ENAMETOOLONG;
+                       goto done;
+               }
+
+               error = vnode_lookup(dmgpath, 0, &imagevp, vfs_context_kernel());
+               if (error == 0) {
+                       // we found it! success!
+                       *rooted_dmg = true;
+                       *skip_signature_check = true;
+                       goto done;
+               }
+       }
+
+       // Try BaseSystem.dmg
+       len = strlcpy(dmgpath, dmgbasepath, MAXPATHLEN);
+       if (len > MAXPATHLEN) {
+               error = ENAMETOOLONG;
+               goto done;
+       }
+
+       len = strlcat(dmgpath, "/BaseSystem.dmg", MAXPATHLEN);
+       if (len > MAXPATHLEN) {
+               error = ENAMETOOLONG;
+               goto done;
+       }
+
+       error = vnode_lookup(dmgpath, 0, &imagevp, vfs_context_kernel());
+       if (error == 0) {
+               // success!
+               goto done;
+       }
+
+done:
+       if (error == 0) {
+               strlcpy(bsdmgpath_out, dmgpath, MAXPATHLEN);
+       } else {
+               bsd_init_kprintf("%s: error %d\n", __func__, error);
+       }
+       if (imagevp != NULLVP) {
+               vnode_put(imagevp);
+       }
+       zfree(ZV_NAMEI, dmgpath);
+       zfree(ZV_NAMEI, dmgbasepath);
+       return error;
+}
+
+static boolean_t
+bsdmgroot_bootable(void)
+{
+#if defined(__arm64__)
+#define BSDMGROOT_DEFAULT true
+#else
+#define BSDMGROOT_DEFAULT false
+#endif
+
+       boolean_t resolved = BSDMGROOT_DEFAULT;
+
+       boolean_t boot_arg_bsdmgroot = false;
+       boolean_t boot_arg_nobsdmgroot = false;
+       int error;
+       mount_t mp;
+       boolean_t root_part_of_volume_group = false;
+       struct vfs_attr vfsattr;
+
+       mp = rootvnode->v_mount;
+       VFSATTR_INIT(&vfsattr);
+       VFSATTR_WANTED(&vfsattr, f_capabilities);
+
+       boot_arg_bsdmgroot = PE_parse_boot_argn("-bsdmgroot", NULL, 0);
+       boot_arg_nobsdmgroot = PE_parse_boot_argn("-nobsdmgroot", NULL, 0);
+
+       error = vfs_getattr(mp, &vfsattr, vfs_context_kernel());
+       if (!error && VFSATTR_IS_SUPPORTED(&vfsattr, f_capabilities)) {
+               if ((vfsattr.f_capabilities.capabilities[VOL_CAPABILITIES_FORMAT] & VOL_CAP_FMT_VOL_GROUPS) &&
+                   (vfsattr.f_capabilities.valid[VOL_CAPABILITIES_FORMAT] & VOL_CAP_FMT_VOL_GROUPS)) {
+                       root_part_of_volume_group = true;
+               }
+       }
+
+       boolean_t singleuser = (boothowto & RB_SINGLE) != 0;
+
+       // Start with the #defined default above.
+       // If booting to single-user mode, default to false, because single-
+       // user mode inside the BaseSystem is probably not what's wanted.
+       // If the 'yes' boot-arg is set, we'll allow that even in single-user
+       // mode, we'll assume you know what you're doing.
+       // The 'no' boot-arg overpowers the 'yes' boot-arg.
+       // In any case, we will not attempt to root from BaseSystem if the
+       // original (booter-chosen) root volume isn't in a volume group.
+       // This is just out of an abundance of caution: if the boot environment
+       // seems to be "something other than a standard install",
+       // we'll be conservative in messing with the root volume.
+
+       if (singleuser) {
+               resolved = false;
+       }
+
+       if (boot_arg_bsdmgroot) {
+               resolved = true;
+       }
+
+       if (boot_arg_nobsdmgroot) {
+               resolved = false;
+       }
+
+       if (!root_part_of_volume_group) {
+               resolved = false;
+       }
+
+       return resolved;
 }
+#endif // CONFIG_BASESYSTEMROOT
 
-#if !NFSCLIENT
-int 
+void
+bsd_exec_setup(int scale)
+{
+       switch (scale) {
+       case 0:
+       case 1:
+               bsd_simul_execs = BSD_SIMUL_EXECS;
+               break;
+       case 2:
+       case 3:
+               bsd_simul_execs = 65;
+               break;
+       case 4:
+       case 5:
+               bsd_simul_execs = 129;
+               break;
+       case 6:
+       case 7:
+               bsd_simul_execs = 257;
+               break;
+       default:
+               bsd_simul_execs = 513;
+               break;
+       }
+       bsd_pageable_map_size = (bsd_simul_execs * BSD_PAGEABLE_SIZE_PER_EXEC);
+}
+
+#if !CONFIG_NETBOOT
+int
+netboot_root(void);
+
+int
 netboot_root(void)
 {
-       return(0);
+       return 0;
 }
 #endif