/* $NetBSD: init_main.c,v 1.279 2006/10/08 04:28:44 thorpej Exp $ */ /* * Copyright (c) 1982, 1986, 1989, 1991, 1992, 1993 * The Regents of the University of California. All rights reserved. * (c) UNIX System Laboratories, Inc. * All or some portions of this file are derived from material licensed * to the University of California by American Telephone and Telegraph * Co. or Unix System Laboratories, Inc. and are reproduced herein with * the permission of UNIX System Laboratories, Inc. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * @(#)init_main.c 8.16 (Berkeley) 5/14/95 */ /* * Copyright (c) 1995 Christopher G. Demetriou. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed by the University of * California, Berkeley and its contributors. * 4. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * @(#)init_main.c 8.16 (Berkeley) 5/14/95 */ #include __KERNEL_RCSID(0, "$NetBSD: init_main.c,v 1.279 2006/10/08 04:28:44 thorpej Exp $"); #include "opt_ipsec.h" #include "opt_kcont.h" #include "opt_multiprocessor.h" #include "opt_ntp.h" #include "opt_pipe.h" #include "opt_posix.h" #include "opt_syscall_debug.h" #include "opt_sysv.h" #include "opt_fileassoc.h" #include "rnd.h" #include "veriexec.h" #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #ifdef FAST_IPSEC #include #endif #ifdef SYSVSHM #include #endif #ifdef SYSVSEM #include #endif #ifdef SYSVMSG #include #endif #ifdef P1003_1B_SEMAPHORE #include #endif #include #include #if NRND > 0 #include #endif #ifndef PIPE_SOCKETPAIR #include #endif #ifdef LKM #include #endif #if NVERIEXEC > 0 #include #endif /* NVERIEXEC > 0 */ #include #include #include #include #include #ifdef FILEASSOC #include #endif /* FILEASSOC */ #include #include #include #include #include #include #include #include #include extern struct proc proc0; extern struct lwp lwp0; extern struct cwdinfo cwdi0; #ifndef curlwp struct lwp *curlwp = &lwp0; #endif struct proc *initproc; struct vnode *rootvp, *swapdev_vp; int boothowto; int cold = 1; /* still working on startup */ struct timeval boottime; /* time at system startup - will only follow settime deltas */ time_t rootfstime; /* recorded root fs time, if known */ volatile int start_init_exec; /* semaphore for start_init() */ static void check_console(struct lwp *l); static void start_init(void *); void main(void); /* * System startup; initialize the world, create process 0, mount root * filesystem, and fork to create init and pagedaemon. Most of the * hard work is done in the lower-level initialization routines including * startup(), which does memory initialization and autoconfiguration. */ void main(void) { #ifdef __HAVE_TIMECOUNTER struct timeval time; #endif struct lwp *l; struct proc *p; struct pdevinit *pdev; int s, error; extern struct pdevinit pdevinit[]; extern void schedcpu(void *); #ifdef NVNODE_IMPLICIT int usevnodes; #endif /* * Initialize the current LWP pointer (curlwp) before * any possible traps/probes to simplify trap processing. */ l = &lwp0; curlwp = l; l->l_cpu = curcpu(); l->l_proc = &proc0; l->l_lid = 1; /* * Attempt to find console and initialize * in case of early panic or other messages. */ consinit(); KERNEL_LOCK_INIT(); uvm_init(); kmem_init(); /* Do machine-dependent initialization. */ cpu_startup(); /* Initialize callouts. */ callout_startup(); /* Initialize the buffer cache */ bufinit(); /* * Initialize mbuf's. Do this now because we might attempt to * allocate mbufs or mbuf clusters during autoconfiguration. */ mbinit(); /* Initialize sockets. */ soinit(); #ifdef KCONT /* Initialize kcont. */ kcont_init(); #endif /* * The following things must be done before autoconfiguration. */ evcnt_init(); /* initialize event counters */ #if NRND > 0 rnd_init(); /* initialize RNG */ #endif /* Initialize the sysctl subsystem. */ sysctl_init(); /* Initialize process and pgrp structures. */ procinit(); lwpinit(); /* Initialize signal-related data structures. */ signal_init(); /* Create process 0 (the swapper). */ proc0_init(); /* * Charge root for one process. */ (void)chgproccnt(0, 1); rqinit(); /* Initialize the file systems. */ #ifdef NVNODE_IMPLICIT /* * If maximum number of vnodes in namei vnode cache is not explicitly * defined in kernel config, adjust the number such as we use roughly * 1.0% of memory for vnode cache (but not less than NVNODE vnodes). */ usevnodes = (ptoa((unsigned)physmem) / 100) / sizeof(struct vnode); if (usevnodes > desiredvnodes) desiredvnodes = usevnodes; #endif vfsinit(); #ifdef __HAVE_TIMECOUNTER inittimecounter(); ntp_init(); #endif /* __HAVE_TIMECOUNTER */ /* Initialize kauth. */ kauth_init(); /* Configure the system hardware. This will enable interrupts. */ configure(); ubc_init(); /* must be after autoconfig */ /* Lock the kernel on behalf of proc0. */ KERNEL_PROC_LOCK(l); #ifdef SYSVSHM /* Initialize System V style shared memory. */ shminit(); #endif #ifdef SYSVSEM /* Initialize System V style semaphores. */ seminit(); #endif #ifdef SYSVMSG /* Initialize System V style message queues. */ msginit(); #endif #ifdef P1003_1B_SEMAPHORE /* Initialize posix semaphores */ ksem_init(); #endif /* Initialize default security model. */ secmodel_start(); #ifdef FILEASSOC fileassoc_init(); #endif /* FILEASSOC */ #if NVERIEXEC > 0 /* * Initialise the fingerprint operations vectors before * fingerprints can be loaded. */ veriexec_init_fp_ops(); #endif /* NVERIEXEC > 0 */ /* Attach pseudo-devices. */ for (pdev = pdevinit; pdev->pdev_attach != NULL; pdev++) (*pdev->pdev_attach)(pdev->pdev_count); #ifdef FAST_IPSEC /* Attach network crypto subsystem */ ipsec_attach(); #endif /* * Initialize protocols. Block reception of incoming packets * until everything is ready. */ s = splnet(); ifinit(); domaininit(); if_attachdomain(); splx(s); #ifdef GPROF /* Initialize kernel profiling. */ kmstartup(); #endif /* Initialize system accouting. */ acct_init(); /* Kick off timeout driven events by calling first time. */ schedcpu(NULL); /* * Create process 1 (init(8)). We do this now, as Unix has * historically had init be process 1, and changing this would * probably upset a lot of people. * * Note that process 1 won't immediately exec init(8), but will * wait for us to inform it that the root file system has been * mounted. */ if (fork1(l, 0, SIGCHLD, NULL, 0, start_init, NULL, NULL, &initproc)) panic("fork init"); /* * Create any kernel threads who's creation was deferred because * initproc had not yet been created. */ kthread_run_deferred_queue(); /* * Now that device driver threads have been created, wait for * them to finish any deferred autoconfiguration. Note we don't * need to lock this semaphore, since we haven't booted any * secondary processors, yet. */ while (config_pending) (void) tsleep(&config_pending, PWAIT, "cfpend", 0); /* * Finalize configuration now that all real devices have been * found. This needs to be done before the root device is * selected, since finalization may create the root device. */ config_finalize(); /* * Now that autoconfiguration has completed, we can determine * the root and dump devices. */ cpu_rootconf(); cpu_dumpconf(); /* Mount the root file system. */ do { domountroothook(); if ((error = vfs_mountroot())) { printf("cannot mount root, error = %d\n", error); boothowto |= RB_ASKNAME; setroot(root_device, (rootdev != NODEV) ? DISKPART(rootdev) : 0); } } while (error != 0); mountroothook_destroy(); /* * Initialise the time-of-day clock, passing the time recorded * in the root filesystem (if any) for use by systems that * don't have a non-volatile time-of-day device. */ inittodr(rootfstime); CIRCLEQ_FIRST(&mountlist)->mnt_flag |= MNT_ROOTFS; CIRCLEQ_FIRST(&mountlist)->mnt_op->vfs_refcount++; /* * Get the vnode for '/'. Set filedesc0.fd_fd.fd_cdir to * reference it. */ error = VFS_ROOT(CIRCLEQ_FIRST(&mountlist), &rootvnode); if (error) panic("cannot find root vnode, error=%d", error); cwdi0.cwdi_cdir = rootvnode; VREF(cwdi0.cwdi_cdir); VOP_UNLOCK(rootvnode, 0); cwdi0.cwdi_rdir = NULL; /* * Now that root is mounted, we can fixup initproc's CWD * info. All other processes are kthreads, which merely * share proc0's CWD info. */ initproc->p_cwdi->cwdi_cdir = rootvnode; VREF(initproc->p_cwdi->cwdi_cdir); initproc->p_cwdi->cwdi_rdir = NULL; /* * Now can look at time, having had a chance to verify the time * from the file system. Reset p->p_rtime as it may have been * munched in mi_switch() after the time got set. */ proclist_lock_read(); s = splsched(); #ifdef __HAVE_TIMECOUNTER getmicrotime(&time); #else mono_time = time; #endif boottime = time; LIST_FOREACH(p, &allproc, p_list) { KASSERT((p->p_flag & P_MARKER) == 0); p->p_stats->p_start = time; LIST_FOREACH(l, &p->p_lwps, l_sibling) { if (l->l_cpu != NULL) l->l_cpu->ci_schedstate.spc_runtime = time; } p->p_rtime.tv_sec = p->p_rtime.tv_usec = 0; } splx(s); proclist_unlock_read(); /* Create the pageout daemon kernel thread. */ uvm_swap_init(); if (kthread_create1(uvm_pageout, NULL, NULL, "pagedaemon")) panic("fork pagedaemon"); /* Create the filesystem syncer kernel thread. */ if (kthread_create1(sched_sync, NULL, NULL, "ioflush")) panic("fork syncer"); /* Create the aiodone daemon kernel thread. */ if (kthread_create1(uvm_aiodone_daemon, NULL, &uvm.aiodoned_proc, "aiodoned")) panic("fork aiodoned"); #if defined(MULTIPROCESSOR) /* Boot the secondary processors. */ cpu_boot_secondary_processors(); #endif /* Initialize exec structures */ exec_init(1); /* * Okay, now we can let init(8) exec! It's off to userland! */ start_init_exec = 1; wakeup(&start_init_exec); /* The scheduler is an infinite loop. */ uvm_scheduler(); /* NOTREACHED */ } void setrootfstime(time_t t) { rootfstime = t; } static void check_console(struct lwp *l) { struct nameidata nd; int error; NDINIT(&nd, LOOKUP, FOLLOW, UIO_SYSSPACE, "/dev/console", l); error = namei(&nd); if (error == 0) vrele(nd.ni_vp); else if (error == ENOENT) printf("warning: no /dev/console\n"); else printf("warning: lookup /dev/console: error %d\n", error); } /* * List of paths to try when searching for "init". */ static const char *initpaths[] = { "/sbin/init", "/sbin/oinit", "/sbin/init.bak", NULL, }; /* * Start the initial user process; try exec'ing each pathname in "initpaths". * The program is invoked with one argument containing the boot flags. */ static void start_init(void *arg) { struct lwp *l = arg; struct proc *p = l->l_proc; vaddr_t addr; struct sys_execve_args /* { syscallarg(const char *) path; syscallarg(char * const *) argp; syscallarg(char * const *) envp; } */ args; int options, i, error; register_t retval[2]; char flags[4], *flagsp; const char *path, *slash; char *ucp, **uap, *arg0, *arg1 = NULL; char ipath[129]; int ipx, len; /* * Now in process 1. */ strncpy(p->p_comm, "init", MAXCOMLEN); /* * Wait for main() to tell us that it's safe to exec. */ while (start_init_exec == 0) (void) tsleep(&start_init_exec, PWAIT, "initexec", 0); /* * This is not the right way to do this. We really should * hand-craft a descriptor onto /dev/console to hand to init, * but that's a _lot_ more work, and the benefit from this easy * hack makes up for the "good is the enemy of the best" effect. */ check_console(l); /* * Need just enough stack to hold the faked-up "execve()" arguments. */ addr = (vaddr_t)STACK_ALLOC(USRSTACK, PAGE_SIZE); if (uvm_map(&p->p_vmspace->vm_map, &addr, PAGE_SIZE, NULL, UVM_UNKNOWN_OFFSET, 0, UVM_MAPFLAG(UVM_PROT_ALL, UVM_PROT_ALL, UVM_INH_COPY, UVM_ADV_NORMAL, UVM_FLAG_FIXED|UVM_FLAG_OVERLAY|UVM_FLAG_COPYONW)) != 0) panic("init: couldn't allocate argument space"); p->p_vmspace->vm_maxsaddr = (caddr_t)STACK_MAX(addr, PAGE_SIZE); ipx = 0; while (1) { if (boothowto & RB_ASKNAME) { printf("init path"); if (initpaths[ipx]) printf(" (default %s)", initpaths[ipx]); printf(": "); len = cngetsn(ipath, sizeof(ipath)-1); if (len == 0) { if (initpaths[ipx]) path = initpaths[ipx++]; else continue; } else { ipath[len] = '\0'; path = ipath; } } else { if ((path = initpaths[ipx++]) == NULL) break; } ucp = (char *)USRSTACK; /* * Construct the boot flag argument. */ flagsp = flags; *flagsp++ = '-'; options = 0; if (boothowto & RB_SINGLE) { *flagsp++ = 's'; options = 1; } #ifdef notyet if (boothowto & RB_FASTBOOT) { *flagsp++ = 'f'; options = 1; } #endif /* * Move out the flags (arg 1), if necessary. */ if (options != 0) { *flagsp++ = '\0'; i = flagsp - flags; #ifdef DEBUG printf("init: copying out flags `%s' %d\n", flags, i); #endif arg1 = STACK_ALLOC(ucp, i); ucp = STACK_MAX(arg1, i); (void)copyout((caddr_t)flags, arg1, i); } /* * Move out the file name (also arg 0). */ i = strlen(path) + 1; #ifdef DEBUG printf("init: copying out path `%s' %d\n", path, i); #else if (boothowto & RB_ASKNAME || path != initpaths[0]) printf("init: trying %s\n", path); #endif arg0 = STACK_ALLOC(ucp, i); ucp = STACK_MAX(arg0, i); (void)copyout(path, arg0, i); /* * Move out the arg pointers. */ ucp = (caddr_t)STACK_ALIGN(ucp, ALIGNBYTES); uap = (char **)STACK_ALLOC(ucp, sizeof(char *) * 3); SCARG(&args, path) = arg0; SCARG(&args, argp) = uap; SCARG(&args, envp) = NULL; slash = strrchr(path, '/'); if (slash) (void)suword((caddr_t)uap++, (long)arg0 + (slash + 1 - path)); else (void)suword((caddr_t)uap++, (long)arg0); if (options != 0) (void)suword((caddr_t)uap++, (long)arg1); (void)suword((caddr_t)uap++, 0); /* terminator */ /* * Now try to exec the program. If can't for any reason * other than it doesn't exist, complain. */ error = sys_execve(l, &args, retval); if (error == 0 || error == EJUSTRETURN) { KERNEL_PROC_UNLOCK(l); return; } printf("exec %s: error %d\n", path, error); } printf("init: not found\n"); panic("no init"); }