]> git.saurik.com Git - apple/xnu.git/blobdiff - bsd/kern/kern_shutdown.c
xnu-1228.15.4.tar.gz
[apple/xnu.git] / bsd / kern / kern_shutdown.c
index 45697a58db7c60164b3a29d0a48055b4a0598aaa..4b59526ed4471c1f7831c5048c2a61b41d3f849c 100644 (file)
@@ -1,23 +1,29 @@
 /*
- * Copyright (c) 2000 Apple Computer, Inc. All rights reserved.
+ * Copyright (c) 2000-2006 Apple Computer, Inc. All rights reserved.
  *
- * @APPLE_LICENSE_HEADER_START@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
  * 
- * The contents of this file constitute Original Code as defined in and
- * are subject to the Apple Public Source License Version 1.1 (the
- * "License").  You may not use this file except in compliance with the
- * License.  Please obtain a copy of the License at
- * http://www.apple.com/publicsource and read it before using this file.
+ * 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
+ * compliance with the License. The rights granted to you under the License
+ * may not be used to create, or enable the creation or redistribution of,
+ * 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.
  * 
- * This Original Code and all software distributed under the License are
- * distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER
+ * 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,
  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT.  Please see the
- * License for the specific language governing rights and limitations
- * under the License.
+ * 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_LICENSE_HEADER_END@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
  */
 /*
  *     File:   bsd/kern/kern_shutdown.c
 #include <sys/systm.h>
 #include <sys/kernel.h>
 #include <sys/vm.h>
-#include <sys/proc.h>
+#include <sys/proc_internal.h>
 #include <sys/user.h>
-#include <sys/buf.h>
 #include <sys/reboot.h>
 #include <sys/conf.h>
-#include <sys/vnode.h>
-#include <sys/file.h>
+#include <sys/vnode_internal.h>
+#include <sys/file_internal.h>
 #include <sys/clist.h>
 #include <sys/callout.h>
 #include <sys/mbuf.h>
 #include <vm/vm_kern.h>
 #include <mach/vm_param.h>
 #include <sys/filedesc.h>
+#include <mach/host_priv.h>
 #include <mach/host_reboot.h>
 
+#include <bsm/audit_kernel.h>
+
+#include <kern/sched_prim.h>           /* for thread_block() */
+#include <kern/host.h>                 /* for host_priv_self() */
+#include <net/if_var.h>                        /* for if_down_all() */
+#include <sys/buf_internal.h>          /* for count_busy_buffers() */
+#include <sys/mount_internal.h>                /* for vfs_unmountall() */
+#include <mach/task.h>                 /* for task_suspend() */
+#include <sys/sysproto.h>              /* abused for sync() */
+#include <kern/clock.h>                        /* for delay_for_interval() */
+
+/* XXX should be in a header file somewhere, but isn't */
+extern void md_prepare_for_shutdown(int, int, char *);
+
 int    waittime = -1;
+static int shutting_down = 0;
+
+static void proc_shutdown(void);
+int in_shutdown(void);
+
+extern void IOSystemShutdownNotification(void);
+
+struct sd_filterargs{
+       int delayterm;
+       int shutdownstate;
+};
+
+
+struct sd_iterargs {
+       int signo;      /* the signal to be posted */
+       int setsdstate;  /* shutdown state to be set */
+};
+
+static int sd_filt1(proc_t, void *);
+static int sd_filt2(proc_t, void *);
+static int  sd_callback1(proc_t p, void * arg);
+static int  sd_callback2(proc_t p, void * arg);
+static int  sd_callback3(proc_t p, void * arg);
 
 void
-boot(paniced, howto, command)
-       int paniced, howto;
-       char *command;
+boot(int paniced, int howto, char *command)
 {
-       register int i;
-       int s;
        struct proc *p = current_proc();        /* XXX */
        int hostboot_option=0;
        int funnel_state;
 
-       static void proc_shutdown();
-    extern void md_prepare_for_shutdown(int paniced, int howto, char * command);
-
        funnel_state = thread_funnel_set(kernel_flock, TRUE);
 
+       /*
+       * Temporary hack to notify the power management root domain
+       * that the system will shut down.
+       */
+       IOSystemShutdownNotification();
+
+       shutting_down = 1;
+           
        md_prepare_for_shutdown(paniced, howto, command);
 
        if ((howto&RB_NOSYNC)==0 && waittime < 0) {
@@ -91,26 +135,32 @@ boot(paniced, howto, command)
                /* handle live procs (deallocate their root and current directories). */                
                proc_shutdown();
 
+#if AUDIT
+               audit_shutdown();
+#endif
+
                sync(p, (void *)NULL, (int *)NULL);
 
-               /* Release vnodes from the VM object cache */    
-               ubc_unmountall();
+               /*
+                * Now that all processes have been terminated and system is
+                * sync'ed up, suspend init
+                */
 
-               IOSleep( 1 * 1000 );
+               if (initproc && p != initproc)
+                       task_suspend(initproc->task);
 
                /*
                 * Unmount filesystems
                 */
-               if (panicstr == 0)
-                       vfs_unmountall();
+               vfs_unmountall();
 
                /* Wait for the buffer cache to clean remaining dirty buffers */
-               for (iter = 0; iter < 20; iter++) {
+               for (iter = 0; iter < 100; iter++) {
                        nbusy = count_busy_buffers();
                        if (nbusy == 0)
                                break;
                        printf("%d ", nbusy);
-                       IOSleep( 4 * nbusy );
+                       delay_for_interval( 1 * nbusy, 1000 * 1000);
                }
                if (nbusy)
                        printf("giving up\n");
@@ -118,12 +168,14 @@ boot(paniced, howto, command)
                        printf("done\n");
        }
 
+#if NETWORKING
        /*
         * Can't just use an splnet() here to disable the network
         * because that will lock out softints which the disk
         * drivers depend on to finish DMAs.
         */
        if_down_all();
+#endif /* NETWORKING */
 
        if (howto & RB_POWERDOWN)
                hostboot_option = HOST_REBOOT_HALT;
@@ -132,14 +184,123 @@ boot(paniced, howto, command)
        if (paniced == RB_PANIC)
                hostboot_option = HOST_REBOOT_HALT;
 
-       if (hostboot_option == HOST_REBOOT_HALT)
-               IOSleep( 1 * 1000 );
+    if (howto & RB_UPSDELAY) {
+        hostboot_option = HOST_REBOOT_UPSDELAY;
+    }
 
        host_reboot(host_priv_self(), hostboot_option);
 
        thread_funnel_set(kernel_flock, FALSE);
 }
 
+static int
+sd_filt1(proc_t p, void * args)
+{
+       proc_t self = current_proc();
+       struct sd_filterargs * sf = (struct sd_filterargs *)args;
+       int delayterm = sf-> delayterm;
+       int shutdownstate = sf->shutdownstate;
+
+       if (((p->p_flag&P_SYSTEM) != 0) || (p->p_ppid == 0) 
+               ||(p == self) || (p->p_stat == SZOMB) 
+               || (p->p_shutdownstate != shutdownstate) 
+               ||((delayterm == 0) && ((p->p_lflag& P_LDELAYTERM) == P_LDELAYTERM))
+               || ((p->p_sigcatch & sigmask(SIGTERM))== 0)) {
+                       return(0);
+               }
+        else 
+                return(1);
+}
+
+
+static int  
+sd_callback1(proc_t p, void * args)
+{
+       struct sd_iterargs * sd = (struct sd_iterargs *)args;
+       int signo = sd->signo;
+       int setsdstate = sd->setsdstate;
+
+       proc_lock(p);
+       p->p_shutdownstate = setsdstate;
+       if (p->p_stat != SZOMB) {
+               proc_unlock(p);
+               psignal(p, signo);
+       } else
+               proc_unlock(p);
+       return(PROC_RETURNED);
+
+}
+
+static int
+sd_filt2(proc_t p, void * args)
+{
+       proc_t self = current_proc();
+       struct sd_filterargs * sf = (struct sd_filterargs *)args;
+       int delayterm = sf-> delayterm;
+       int shutdownstate = sf->shutdownstate;
+
+       if (((p->p_flag&P_SYSTEM) != 0) || (p->p_ppid == 0) 
+               ||(p == self) || (p->p_stat == SZOMB) 
+               || (p->p_shutdownstate == shutdownstate) 
+               ||((delayterm == 0) && ((p->p_lflag& P_LDELAYTERM) == P_LDELAYTERM))) {
+                       return(0);
+               }
+        else
+                return(1);
+}
+
+static int  
+sd_callback2(proc_t p, void * args)
+{
+       struct sd_iterargs * sd = (struct sd_iterargs *)args;
+       int signo = sd->signo;
+       int setsdstate = sd->setsdstate;
+
+       proc_lock(p);
+       p->p_shutdownstate = setsdstate;
+       if (p->p_stat != SZOMB) {
+               proc_unlock(p);
+               psignal(p, signo);
+       } else
+               proc_unlock(p);
+
+       return(PROC_RETURNED);
+
+}
+
+static int  
+sd_callback3(proc_t p, void * args)
+{
+       struct sd_iterargs * sd = (struct sd_iterargs *)args;
+       int setsdstate = sd->setsdstate;
+
+       proc_lock(p);
+       p->p_shutdownstate = setsdstate;
+       if (p->p_stat != SZOMB) {
+              /*
+               * NOTE: following code ignores sig_lock and plays
+               * with exit_thread correctly.  This is OK unless we
+               * are a multiprocessor, in which case I do not
+               * understand the sig_lock.  This needs to be fixed.
+               * XXX
+               */
+               if (p->exit_thread) {   /* someone already doing it */
+                       proc_unlock(p);
+                       /* give him a chance */
+                       thread_block(THREAD_CONTINUE_NULL);
+               } else {
+                       p->exit_thread = current_thread();
+                       printf(".");
+                       proc_unlock(p);
+                       exit1(p, 1, (int *)NULL);
+               }
+       } else
+               proc_unlock(p);
+
+       return(PROC_RETURNED);
+}
+
+
 /*
  * proc_shutdown()
  *
@@ -153,11 +314,13 @@ boot(paniced, howto, command)
  */
 
 static void
-proc_shutdown()
+proc_shutdown(void)
 {
        struct proc     *p, *self;
-       struct vnode    **cdirp, **rdirp, *vp;
-       int             restart, i, TERM_catch;
+       int             i, TERM_catch;
+       int delayterm = 0;
+       struct sd_filterargs sfargs;
+       struct sd_iterargs sdargs;
 
        /*
         *      Kill as many procs as we can.  (Except ourself...)
@@ -165,23 +328,29 @@ proc_shutdown()
        self = (struct proc *)current_proc();
        
        /*
-        * Suspend /etc/init
+        * Signal the init with SIGTERM so that he does not launch
+        * new processes 
         */
-       p = pfind(1);
-       if (p && p != self)
-               task_suspend(p->task);          /* stop init */
+       p = proc_find(1);
+       if (p && p != self) {
+               psignal(p, SIGTERM);
+       }
+       proc_rele(p);
 
        printf("Killing all processes ");
 
+sigterm_loop:
        /*
         * send SIGTERM to those procs interested in catching one
         */
-       for (p = allproc.lh_first; p; p = p->p_list.le_next) {
-               if (((p->p_flag&P_SYSTEM) == 0) && (p->p_pptr->p_pid != 0) && (p != self)) {
-                       if (p->p_sigcatch & sigmask(SIGTERM))
-                               psignal(p, SIGTERM);
-               }
-       }
+       sfargs.delayterm = delayterm;
+       sfargs.shutdownstate = 0;
+       sdargs.signo = SIGTERM;
+       sdargs.setsdstate = 1;
+
+       /* post a SIGTERM to all that catch SIGTERM and not marked for delay */
+       proc_rebootscan(sd_callback1, (void *)&sdargs, sd_filt1, (void *)&sfargs);
+
        /*
         * now wait for up to 30 seconds to allow those procs catching SIGTERM
         * to digest it
@@ -193,36 +362,72 @@ proc_shutdown()
                 * and then check to see if the tasks that were sent a
                 * SIGTERM have exited
                 */
-               IOSleep(100);   
+               delay_for_interval(100, 1000 * 1000);
                TERM_catch = 0;
 
-               for (p = allproc.lh_first; p; p = p->p_list.le_next) {
-                       if (((p->p_flag&P_SYSTEM) == 0) && (p->p_pptr->p_pid != 0) && (p != self)) {
-                               if (p->p_sigcatch & sigmask(SIGTERM))
-                                       TERM_catch++;
+
+               proc_list_lock();
+
+               for (p = allproc.lh_first; p; p = p->p_list.le_next) {
+                       if (p->p_shutdownstate == 1) {
+                               TERM_catch++;
                        }
                }
+
+               proc_list_unlock();
+
                if (TERM_catch == 0)
                        break;
        }
+       if (TERM_catch) {
+               /*
+                * log the names of the unresponsive tasks
+                */
+
+
+               proc_list_lock();
+
+               for (p = allproc.lh_first; p; p = p->p_list.le_next) {
+                       if (p->p_shutdownstate == 1) {
+                                 printf("%s[%d]: didn't act on SIGTERM\n", p->p_comm, p->p_pid);
+                       }
+               }
+
+               proc_list_unlock();
+
+               delay_for_interval(1000 * 5, 1000 * 1000);
+       }
 
        /*
         * send a SIGKILL to all the procs still hanging around
         */
-       for (p = allproc.lh_first; p; p = p->p_list.le_next) {
-               if (((p->p_flag&P_SYSTEM) == 0) && (p->p_pptr->p_pid != 0) && (p != self))
-                       psignal(p, SIGKILL);
-       }
+       sfargs.delayterm = delayterm;
+       sfargs.shutdownstate = 2;
+       sdargs.signo = SIGKILL;
+       sdargs.setsdstate = 2;
+
+       /* post a SIGTERM to all that catch SIGTERM and not marked for delay */
+       proc_rebootscan(sd_callback2, (void *)&sdargs, sd_filt2, (void *)&sfargs);
+
        /*
         * wait for up to 60 seconds to allow these procs to exit normally
+        *
+        * History:     The delay interval was changed from 100 to 200
+        *              for NFS requests in particular.
         */
        for (i = 0; i < 300; i++) {
-               IOSleep(200);  /* double the time from 100 to 200 for NFS requests in particular */
+               delay_for_interval(200, 1000 * 1000);
+
+
+               proc_list_lock();
 
                for (p = allproc.lh_first; p; p = p->p_list.le_next) {
-                       if (((p->p_flag&P_SYSTEM) == 0) && (p->p_pptr->p_pid != 0) && (p != self))
+                               if (p->p_shutdownstate == 2)
                                break;
                }
+
+               proc_list_unlock();
+
                if (!p)
                        break;
        }
@@ -230,53 +435,30 @@ proc_shutdown()
        /*
         * if we still have procs that haven't exited, then brute force 'em
         */
-       p = allproc.lh_first;
-       while (p) {
-               if ((p->p_flag&P_SYSTEM) || (p->p_pptr->p_pid == 0) || (p == self)) {
-                       p = p->p_list.le_next;
-               }
-               else {
-                       /*
-                        * NOTE: following code ignores sig_lock and plays
-                        * with exit_thread correctly.  This is OK unless we
-                        * are a multiprocessor, in which case I do not
-                        * understand the sig_lock.  This needs to be fixed.
-                        * XXX
-                        */
-                       if (p->exit_thread) {   /* someone already doing it */
-                                               /* give him a chance */
-                               thread_block(THREAD_CONTINUE_NULL);
-                       }
-                       else {
-                               p->exit_thread = current_thread();
-                               printf(".");
-                               exit1(p, 1, (int *)NULL);
-                       }
-                       p = allproc.lh_first;
-               }
-       }
+       sfargs.delayterm = delayterm;
+       sfargs.shutdownstate = 3;
+       sdargs.signo = 0;
+       sdargs.setsdstate = 3;
+
+       /* post a SIGTERM to all that catch SIGTERM and not marked for delay */
+       proc_rebootscan(sd_callback3, (void *)&sdargs, sd_filt2, (void *)&sfargs);
        printf("\n");
-       /*
-        *      Forcibly free resources of what's left.
-        */
-       p = allproc.lh_first;
-       while (p) {
-       /*
-        * Close open files and release open-file table.
-        * This may block!
-        */
-#ifdef notyet
-       /* panics on reboot due to "zfree: non-allocated memory in collectable zone" message */
-       fdfree(p);
-#endif /* notyet */
-       p = p->p_list.le_next;
+
+       /* Now start the termination of processes that are marked for delayed termn */
+       if (delayterm == 0) {
+               delayterm = 1;
+               goto  sigterm_loop;
        }
-       /* Wait for the reaper thread to run, and clean up what we have done 
-        * before we proceed with the hardcore shutdown. This reduces the race
-        * between kill_tasks and the reaper thread.
-        */
-       /* thread_wakeup(&reaper_queue); */
-       /*      IOSleep( 1 * 1000);      */
+       /* drop the ref on initproc */
+       proc_rele(initproc);
        printf("continuing\n");
 }
 
+/*
+ * Check whether the system has begun its shutdown sequence. 
+ */
+int
+in_shutdown(void)
+{
+       return shutting_down;
+}