]> git.saurik.com Git - apple/xnu.git/blobdiff - bsd/kern/kern_sig.c
xnu-3789.70.16.tar.gz
[apple/xnu.git] / bsd / kern / kern_sig.c
index 1492505f3f6de93ccd81dfafe8e2c10e3c13a345..5d38f292ed97459ae279dbe462a841435ef5e4e0 100644 (file)
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2000-2001 Apple Computer, Inc. All rights reserved.
+ * Copyright (c) 1995-2016 Apple Inc. All rights reserved.
  *
  * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
  * 
@@ -25,7 +25,6 @@
  * 
  * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
  */
-/* Copyright (c) 1995-1998 Apple Computer, Inc. All Rights Reserved */
 /*
  * Copyright (c) 1982, 1986, 1989, 1991, 1993
  *     The Regents of the University of California.  All rights reserved.
  *
  *     @(#)kern_sig.c  8.7 (Berkeley) 4/18/94
  */
+/*
+ * NOTICE: This file was modified by SPARTA, Inc. in 2005 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.
+ */
 
 #define        SIGPROP         /* include signal properties table */
 #include <sys/param.h>
 #include <sys/kernel.h>
 #include <sys/wait.h>
 #include <sys/signalvar.h>
-#if KTRACE
-#include <sys/ktrace.h>
-#endif
 #include <sys/syslog.h>
 #include <sys/stat.h>
 #include <sys/lock.h>
 #include <sys/kdebug.h>
+#include <sys/reason.h>
 
 #include <sys/mount.h>
 #include <sys/sysproto.h>
 
-#include <bsm/audit_kernel.h>
+#include <security/audit/audit.h>
 
 #include <machine/spl.h>
 
 #include <sys/vm.h>
 #include <sys/user.h>          /* for coredump */
 #include <kern/ast.h>          /* for APC support */
-#include <kern/lock.h>
+#include <kern/kalloc.h>
 #include <kern/task.h>         /* extern void   *get_bsdtask_info(task_t); */
 #include <kern/thread.h>
 #include <kern/sched_prim.h>
 #include <kern/thread_call.h>
+#include <kern/policy_internal.h>
+
 #include <mach/exception.h>
 #include <mach/task.h>
 #include <mach/thread_act.h>
+#include <libkern/OSAtomic.h>
+
+#include <sys/sdt.h>
+#include <sys/codesign.h>
 
 /*
  * Missing prototypes that Mach should export
  * +++
  */
 extern int thread_enable_fpe(thread_t act, int onoff);
-extern void unix_syscall_return(int error);
 extern thread_t        port_name_to_thread(mach_port_name_t port_name);
-extern kern_return_t check_actforsig(task_t task, thread_t thread, int setast);
 extern kern_return_t get_signalact(task_t , thread_t *, int);
-extern boolean_t thread_should_abort(thread_t);
 extern unsigned int get_useraddr(void);
+extern boolean_t task_did_exec(task_t task);
+extern boolean_t task_is_exec_copy(task_t task);
 
 /*
  * ---
  */
 
-extern void doexception(int exc, int code, int sub);
+extern void doexception(int exc, mach_exception_code_t code, 
+               mach_exception_subcode_t sub);
 
-void stop(struct proc *p);
-int cansignal(struct proc *, kauth_cred_t, struct proc *, int);
-int killpg1(struct proc *, int, int, int);
-void sigexit_locked(struct proc *, int);
-int setsigvec(struct proc *, int, struct __user_sigaction *);
-void exit1(struct proc *, int, int *);
-void psignal_uthread(thread_t, int);
+static void stop(proc_t, proc_t);
+int cansignal(proc_t, kauth_cred_t, proc_t, int, int);
+int killpg1(proc_t, int, int, int, int);
 kern_return_t do_bsdexception(int, int, int);
 void __posix_sem_syscall_return(kern_return_t);
+char *proc_name_address(void *p);
 
 /* implementations in osfmk/kern/sync_sema.c. We do not want port.h in this scope, so void * them  */
-kern_return_t semaphore_timedwait_signal_trap_internal(void *, void *,time_t, int32_t, void (*)(int));
-kern_return_t semaphore_timedwait_trap_internal(void *, time_t, int32_t, void (*)(int));
-kern_return_t semaphore_wait_signal_trap_internal(void *, void *, void (*)(int));
-kern_return_t semaphore_wait_trap_internal(void *, void (*)(int));
+kern_return_t semaphore_timedwait_signal_trap_internal(mach_port_name_t, mach_port_name_t, unsigned int, clock_res_t, void (*)(kern_return_t));
+kern_return_t semaphore_timedwait_trap_internal(mach_port_name_t, unsigned int, clock_res_t, void (*)(kern_return_t));
+kern_return_t semaphore_wait_signal_trap_internal(mach_port_name_t, mach_port_name_t, void (*)(kern_return_t));
+kern_return_t semaphore_wait_trap_internal(mach_port_name_t, void (*)(kern_return_t));
 
 static int     filt_sigattach(struct knote *kn);
 static void    filt_sigdetach(struct knote *kn);
 static int     filt_signal(struct knote *kn, long hint);
+static int     filt_signaltouch(struct knote *kn, struct kevent_internal_s *kev);
+static int     filt_signalprocess(struct knote *kn, struct filt_process_s *data, struct kevent_internal_s *kev);
+
+struct filterops sig_filtops = {
+        .f_attach = filt_sigattach,
+        .f_detach = filt_sigdetach,
+        .f_event = filt_signal,
+        .f_touch = filt_signaltouch,
+        .f_process = filt_signalprocess,
+};
+
+/* structures  and fns for killpg1 iterartion callback and filters */
+struct killpg1_filtargs {
+       int  posix;
+       proc_t cp;
+};
+
+struct killpg1_iterargs {
+       proc_t cp;
+       kauth_cred_t uc;
+       int signum;
+       int * nfoundp;
+       int zombie;
+};
+
+static int killpg1_filt(proc_t p, void * arg);
+static int killpg1_pgrpfilt(proc_t p, __unused void * arg);
+static int killpg1_callback(proc_t p, void * arg);
 
-struct filterops sig_filtops =
-       { 0, filt_sigattach, filt_sigdetach, filt_signal };
+static int pgsignal_filt(proc_t p, void * arg);
+static int pgsignal_callback(proc_t p, void * arg);
+static kern_return_t get_signalthread(proc_t, int, thread_t *);
 
 
+/* flags for psignal_internal */
+#define PSIG_LOCKED     0x1
+#define PSIG_VFORK      0x2
+#define PSIG_THREAD     0x4
+#define PSIG_TRY_THREAD 0x8
+
+static os_reason_t build_signal_reason(int signum, const char *procname);
+static void psignal_internal(proc_t p, task_t task, thread_t thread, int flavor, int signum, os_reason_t signal_reason);
+
 /*
  * NOTE: Source and target may *NOT* overlap! (target is smaller)
  */
 static void
-sigaltstack_64to32(struct user_sigaltstack *in, struct sigaltstack *out)
+sigaltstack_kern_to_user32(struct kern_sigaltstack *in, struct user32_sigaltstack *out)
+{
+       out->ss_sp          = CAST_DOWN_EXPLICIT(user32_addr_t, in->ss_sp);
+       out->ss_size    = CAST_DOWN_EXPLICIT(user32_size_t, in->ss_size);
+       out->ss_flags   = in->ss_flags;
+}
+
+static void
+sigaltstack_kern_to_user64(struct kern_sigaltstack *in, struct user64_sigaltstack *out)
 {
-       out->ss_sp      = CAST_DOWN(void *,in->ss_sp);
+       out->ss_sp          = in->ss_sp;
        out->ss_size    = in->ss_size;
        out->ss_flags   = in->ss_flags;
 }
@@ -168,24 +220,39 @@ sigaltstack_64to32(struct user_sigaltstack *in, struct sigaltstack *out)
  * the beginning.
  */
 static void
-sigaltstack_32to64(struct sigaltstack *in, struct user_sigaltstack *out)
+sigaltstack_user32_to_kern(struct user32_sigaltstack *in, struct kern_sigaltstack *out)
+{
+       out->ss_flags   = in->ss_flags;
+       out->ss_size    = in->ss_size;
+       out->ss_sp              = CAST_USER_ADDR_T(in->ss_sp);
+}
+static void
+sigaltstack_user64_to_kern(struct user64_sigaltstack *in, struct kern_sigaltstack *out)
 {
        out->ss_flags   = in->ss_flags;
        out->ss_size    = in->ss_size;
-       out->ss_sp      = CAST_USER_ADDR_T(in->ss_sp);
+       out->ss_sp              = in->ss_sp;
 }
 
 static void
-sigaction_64to32(struct user_sigaction *in, struct sigaction *out)
+sigaction_kern_to_user32(struct kern_sigaction *in, struct user32_sigaction *out)
+{
+       /* This assumes 32 bit __sa_handler is of type sig_t */
+       out->__sigaction_u.__sa_handler = CAST_DOWN_EXPLICIT(user32_addr_t,in->__sigaction_u.__sa_handler);
+       out->sa_mask = in->sa_mask;
+       out->sa_flags = in->sa_flags;
+}
+static void
+sigaction_kern_to_user64(struct kern_sigaction *in, struct user64_sigaction *out)
 {
        /* This assumes 32 bit __sa_handler is of type sig_t */
-       out->__sigaction_u.__sa_handler = CAST_DOWN(sig_t,in->__sigaction_u.__sa_handler);
+       out->__sigaction_u.__sa_handler = in->__sigaction_u.__sa_handler;
        out->sa_mask = in->sa_mask;
        out->sa_flags = in->sa_flags;
 }
 
 static void
-__sigaction_32to64(struct __sigaction *in, struct __user_sigaction *out)
+__sigaction_user32_to_kern(struct __user32_sigaction *in, struct __kern_sigaction *out)
 {
        out->__sigaction_u.__sa_handler = CAST_USER_ADDR_T(in->__sigaction_u.__sa_handler);
        out->sa_tramp = CAST_USER_ADDR_T(in->sa_tramp);
@@ -193,6 +260,14 @@ __sigaction_32to64(struct __sigaction *in, struct __user_sigaction *out)
        out->sa_flags = in->sa_flags;
 }
 
+static void
+__sigaction_user64_to_kern(struct __user64_sigaction *in, struct __kern_sigaction *out)
+{
+       out->__sigaction_u.__sa_handler = in->__sigaction_u.__sa_handler;
+       out->sa_tramp = in->sa_tramp;
+       out->sa_mask = in->sa_mask;
+       out->sa_flags = in->sa_flags;
+}
 
 #if SIGNAL_DEBUG
 void ram_printf(int);
@@ -206,150 +281,185 @@ ram_printf(int x)
 }
 #endif /* SIGNAL_DEBUG */
 
-int
-signal_lock(struct proc *p)
-{
-int error = 0;
-#if DIAGNOSTIC
-#if SIGNAL_DEBUG
-#ifdef __ppc__
-        {
-            int register sp, *fp, numsaved; 
-            __asm__ volatile("mr %0,r1" : "=r" (sp));
-
-            fp = (int *)*((int *)sp);
-            for (numsaved = 0; numsaved < 3; numsaved++) {
-                p->lockpc[numsaved] = fp[2];
-                if ((int)fp <= 0)
-                        break;
-                fp = (int *)*fp;
-            }
-        }
-#endif /* __ppc__ */       
-#endif /* SIGNAL_DEBUG */
-#endif /* DIAGNOSTIC */
-
-siglock_retry:
-       error = lockmgr((struct lock__bsd__ *)&p->signal_lock[0], LK_EXCLUSIVE, 0, (struct proc *)0);
-       if (error == EINTR)
-               goto siglock_retry;
-       return(error);
-}
-
-int
-signal_unlock(struct proc *p)
-{
-#if DIAGNOSTIC
-#if SIGNAL_DEBUG
-#ifdef __ppc__
-        {
-            int register sp, *fp, numsaved; 
-            __asm__ volatile("mr %0,r1" : "=r" (sp));
-
-            fp = (int *)*((int *)sp);
-            for (numsaved = 0; numsaved < 3; numsaved++) {
-                p->unlockpc[numsaved] = fp[2];
-                if ((int)fp <= 0)
-                        break;
-                fp = (int *)*fp;
-            }
-        }
-#endif /* __ppc__ */       
-#endif /* SIGNAL_DEBUG */
-#endif /* DIAGNOSTIC */
-
-       /* TBD:  check p last arg */
-       return(lockmgr((struct lock__bsd__ *)&p->signal_lock[0], LK_RELEASE, (simple_lock_t)0, (struct proc *)0));
-}
 
 void
-signal_setast(sig_actthread)
-thread_t sig_actthread;
+signal_setast(thread_t sig_actthread)
 {
        act_set_astbsd(sig_actthread);
 }
 
 /*
  * Can process p, with ucred uc, send the signal signum to process q?
+ * uc is refcounted  by the caller so internal fileds can be used safely
+ * when called with zombie arg, list lock is held
  */
 int
-cansignal(p, uc, q, signum)
-       struct proc *p;
-       kauth_cred_t uc;
-       struct proc *q;
-       int signum;
+cansignal(proc_t p, kauth_cred_t uc, proc_t q, int signum, int zombie)
 {
+       kauth_cred_t my_cred;
+       struct session * p_sessp = SESSION_NULL;
+       struct session * q_sessp = SESSION_NULL;
+#if CONFIG_MACF
+       int error;
+
+       error = mac_proc_check_signal(p, q, signum);
+       if (error)
+               return (0);
+#endif
+
        /* you can signal yourself */
        if (p == q)
                return(1);
 
+       /* you can't send launchd SIGKILL, even if root */
+       if (signum == SIGKILL && q == initproc)
+               return(0);
+
        if (!suser(uc, NULL))
                return (1);             /* root can always signal */
 
-       if (signum == SIGCONT && q->p_session == p->p_session)
+       if (zombie == 0)
+               proc_list_lock();
+       if (p->p_pgrp != PGRP_NULL)
+               p_sessp = p->p_pgrp->pg_session;
+       if (q->p_pgrp != PGRP_NULL)
+               q_sessp = q->p_pgrp->pg_session;
+
+       if (signum == SIGCONT && q_sessp == p_sessp) {
+               if (zombie == 0)
+                       proc_list_unlock();
                return (1);             /* SIGCONT in session */
+       }
+
+       if (zombie == 0) 
+               proc_list_unlock();
 
        /*
-        * Using kill(), only certain signals can be sent to setugid
-        * child processes
+        * If the real or effective UID of the sender matches the real
+        * or saved UID of the target, permit the signal to
+        * be sent.
         */
-       if (q->p_flag & P_SUGID) {
-               switch (signum) {
-               case 0:
-               case SIGKILL:
-               case SIGINT:
-               case SIGTERM:
-               case SIGSTOP:
-               case SIGTTIN:
-               case SIGTTOU:
-               case SIGTSTP:
-               case SIGHUP:
-               case SIGUSR1:
-               case SIGUSR2:
-                       if (uc->cr_ruid == q->p_ucred->cr_ruid ||
-                           kauth_cred_getuid(uc) == q->p_ucred->cr_ruid ||
-                           uc->cr_ruid == kauth_cred_getuid(q->p_ucred) ||
-                           kauth_cred_getuid(uc) == kauth_cred_getuid(q->p_ucred))
-                               return (1);
-               }
-               return (0);
+       if (zombie == 0)
+               my_cred = kauth_cred_proc_ref(q);
+       else
+               my_cred = proc_ucred(q);
+
+       if (kauth_cred_getruid(uc) == kauth_cred_getruid(my_cred) ||
+           kauth_cred_getruid(uc) == kauth_cred_getsvuid(my_cred) ||
+           kauth_cred_getuid(uc) == kauth_cred_getruid(my_cred) ||
+           kauth_cred_getuid(uc) == kauth_cred_getsvuid(my_cred)) {
+               if (zombie == 0)
+                       kauth_cred_unref(&my_cred);
+               return (1);
        }
 
-       /* XXX
-        * because the P_SUGID test exists, this has extra tests which
-        * could be removed.
-        */
-       if (uc->cr_ruid == q->p_ucred->cr_ruid ||
-           uc->cr_ruid == q->p_ucred->cr_svuid ||
-           kauth_cred_getuid(uc) == q->p_ucred->cr_ruid ||
-           kauth_cred_getuid(uc) == q->p_ucred->cr_svuid ||
-           uc->cr_ruid == kauth_cred_getuid(q->p_ucred) ||
-           kauth_cred_getuid(uc) == kauth_cred_getuid(q->p_ucred))
-               return (1);
+       if (zombie == 0)
+               kauth_cred_unref(&my_cred);
+
        return (0);
 }
 
+/*
+ * <rdar://problem/21952708> Some signals can be restricted from being handled,
+ * forcing the default action for that signal. This behavior applies only to
+ * non-root (EUID != 0) processes, and is configured with the "sigrestrict=x"
+ * bootarg:
+ *
+ *   0 (default): Disallow use of restricted signals. Trying to register a handler
+ *             returns ENOTSUP, which userspace may use to take special action (e.g. abort).
+ *   1: As above, but return EINVAL. Restricted signals behave similarly to SIGKILL.
+ *   2: Usual POSIX semantics.
+ */
+unsigned sigrestrict_arg = 0;
+
+#if PLATFORM_WatchOS
+static int
+sigrestrictmask(void)
+{
+       if (kauth_getuid() != 0 && sigrestrict_arg != 2) {
+               return SIGRESTRICTMASK;
+       }
+       return 0;
+}
+
+static int
+signal_is_restricted(proc_t p, int signum)
+{
+       if (sigmask(signum) & sigrestrictmask()) {
+               if (sigrestrict_arg == 0 &&
+                               task_get_apptype(p->task) == TASK_APPTYPE_APP_DEFAULT) {
+                       return ENOTSUP;
+               } else {
+                       return EINVAL;
+               }
+       }
+       return 0;
+}
+
+#else
 
+static inline int
+signal_is_restricted(proc_t p, int signum)
+{
+       (void)p;
+       (void)signum;
+       return 0;
+}
+#endif /* !PLATFORM_WatchOS */
+
+/*
+ * Returns:    0                       Success
+ *             EINVAL
+ *     copyout:EFAULT
+ *     copyin:EFAULT
+ *
+ * Notes:      Uses current thread as a parameter to inform PPC to enable
+ *             FPU exceptions via setsigvec(); this operation is not proxy
+ *             safe!
+ */
 /* ARGSUSED */
 int
-sigaction(struct proc *p, register struct sigaction_args *uap, __unused register_t *retval)
+sigaction(proc_t p, struct sigaction_args *uap, __unused int32_t *retval)
 {
-       struct user_sigaction vec;
-       struct __user_sigaction __vec;
+       struct kern_sigaction vec;
+       struct __kern_sigaction __vec;
 
-       struct user_sigaction *sa = &vec;
-       register struct sigacts *ps = p->p_sigacts;
+       struct kern_sigaction *sa = &vec;
+       struct sigacts *ps = p->p_sigacts;
 
-       register int signum;
+       int signum;
        int bit, error=0;
 
        signum = uap->signum;
        if (signum <= 0 || signum >= NSIG ||
-           signum == SIGKILL || signum == SIGSTOP)
+                       signum == SIGKILL || signum == SIGSTOP)
                return (EINVAL);
 
+       if (uap->nsa) {
+               if (IS_64BIT_PROCESS(p)) {
+                       struct __user64_sigaction       __vec64;
+                       error = copyin(uap->nsa, &__vec64, sizeof(__vec64));
+                       __sigaction_user64_to_kern(&__vec64, &__vec);
+               } else {
+                       struct __user32_sigaction       __vec32;
+                       error = copyin(uap->nsa, &__vec32, sizeof(__vec32));
+                       __sigaction_user32_to_kern(&__vec32, &__vec);
+               }
+               if (error)
+                       return (error);
+               __vec.sa_flags &= SA_USERSPACE_MASK; /* Only pass on valid sa_flags */
+
+               if ((__vec.sa_flags & SA_SIGINFO) || __vec.sa_handler != SIG_DFL) {
+                       if ((error = signal_is_restricted(p, signum))) {
+                               if (error == ENOTSUP) {
+                                       printf("%s(%d): denied attempt to register action for signal %d\n",
+                                                       proc_name_address(p), proc_pid(p), signum);
+                               }
+                               return error;
+                       }
+               }
+       }
+
        if (uap->osa) {
                sa->sa_handler = ps->ps_sigact[signum];
                sa->sa_mask = ps->ps_catchmask[signum];
@@ -371,47 +481,45 @@ sigaction(struct proc *p, register struct sigaction_args *uap, __unused register
                        sa->sa_flags |= SA_NOCLDWAIT;
 
                if (IS_64BIT_PROCESS(p)) {
-                       error = copyout(sa, uap->osa, sizeof(struct user_sigaction));
+                       struct user64_sigaction vec64;
+                       sigaction_kern_to_user64(sa, &vec64);
+                       error = copyout(&vec64, uap->osa, sizeof(vec64));
                } else {
-                       struct sigaction vec32;
-                       sigaction_64to32(sa, &vec32);
-                       error = copyout(&vec32, uap->osa, sizeof(struct sigaction));
+                       struct user32_sigaction vec32;
+                       sigaction_kern_to_user32(sa, &vec32);
+                       error = copyout(&vec32, uap->osa, sizeof(vec32));
                }
                if (error)
                        return (error);
        }
+
        if (uap->nsa) {
-               if (IS_64BIT_PROCESS(p)) {
-                       error = copyin(uap->nsa, &__vec, sizeof(struct __user_sigaction));
-               } else {
-                       struct __sigaction __vec32;
-                       error = copyin(uap->nsa, &__vec32, sizeof(struct __sigaction));
-                       __sigaction_32to64(&__vec32, &__vec);
-               }
-               if (error)
-                       return (error);
-               error = setsigvec(p, signum, &__vec);
+               error = setsigvec(p, current_thread(), signum, &__vec, FALSE);
        }
+
        return (error);
 }
 
 /* Routines to manipulate bits on all threads */
 int
-clear_procsiglist(struct proc *p,  int bit)
+clear_procsiglist(proc_t p,  int bit, boolean_t in_signalstart)
 {
        struct uthread * uth;
        thread_t thact;
 
-       signal_lock(p);
+       proc_lock(p);
+       if (!in_signalstart)
+               proc_signalstart(p, 1);
 
-       if ((p->p_flag & P_INVFORK) && p->p_vforkact) {
+       if ((p->p_lflag & P_LINVFORK) && p->p_vforkact) {
                thact = p->p_vforkact;  
                uth = (struct uthread *)get_bsdthread_info(thact);
                if (uth) {
                        uth->uu_siglist &= ~bit;
                }
-               p->p_siglist &= ~bit;
-               signal_unlock(p);
+               if (!in_signalstart)
+                       proc_signalend(p, 1);
+               proc_unlock(p);
                return(0);
        } 
 
@@ -419,93 +527,122 @@ clear_procsiglist(struct proc *p,  int bit)
                uth->uu_siglist &= ~bit;
        }
        p->p_siglist &= ~bit;
-       signal_unlock(p);
+       if (!in_signalstart)
+               proc_signalend(p, 1);
+       proc_unlock(p);
+
        return(0);
 }
 
 
 static int
-unblock_procsigmask(struct proc *p,  int bit)
+unblock_procsigmask(proc_t p,  int bit)
 {
        struct uthread * uth;
        thread_t thact;
 
-       signal_lock(p);
-       if ((p->p_flag & P_INVFORK) && p->p_vforkact) {
+       proc_lock(p);
+       proc_signalstart(p, 1);
+
+       if ((p->p_lflag & P_LINVFORK) && p->p_vforkact) {
                thact = p->p_vforkact;  
                uth = (struct uthread *)get_bsdthread_info(thact);
                if (uth) {
                        uth->uu_sigmask &= ~bit;
                }
                p->p_sigmask &= ~bit;
-               signal_unlock(p);
+               proc_signalend(p, 1);
+               proc_unlock(p);
                return(0);
        } 
        TAILQ_FOREACH(uth, &p->p_uthlist, uu_list) {
                uth->uu_sigmask &= ~bit;
        }
        p->p_sigmask &= ~bit;
-       signal_unlock(p);
+
+       proc_signalend(p, 1);
+       proc_unlock(p);
        return(0);
 }
 
-
 static int
-block_procsigmask(struct proc *p,  int bit)
+block_procsigmask(proc_t p,  int bit)
 {
        struct uthread * uth;
        thread_t thact;
 
-       signal_lock(p);
-       if ((p->p_flag & P_INVFORK) && p->p_vforkact) {
+       proc_lock(p);
+       proc_signalstart(p, 1);
+
+       if ((p->p_lflag & P_LINVFORK) && p->p_vforkact) {
                thact = p->p_vforkact;  
                uth = (struct uthread *)get_bsdthread_info(thact);
                if (uth) {
                        uth->uu_sigmask |= bit;
                }
                p->p_sigmask |=  bit;
-               signal_unlock(p);
+               proc_signalend(p, 1);
+               proc_unlock(p);
                return(0);
        } 
        TAILQ_FOREACH(uth, &p->p_uthlist, uu_list) {
                uth->uu_sigmask |= bit;
        }
        p->p_sigmask |=  bit;
-       signal_unlock(p);
+
+       proc_signalend(p, 1);
+       proc_unlock(p);
        return(0);
 }
 
 int
-set_procsigmask(struct proc *p,  int bit)
+set_procsigmask(proc_t p,  int bit)
 {
        struct uthread * uth;
        thread_t thact;
 
-       signal_lock(p);
-       if ((p->p_flag & P_INVFORK) && p->p_vforkact) {
+       proc_lock(p);
+       proc_signalstart(p, 1);
+
+       if ((p->p_lflag & P_LINVFORK) && p->p_vforkact) {
                thact = p->p_vforkact;  
                uth = (struct uthread *)get_bsdthread_info(thact);
                if (uth) {
                        uth->uu_sigmask = bit;
                }
                p->p_sigmask =  bit;
-               signal_unlock(p);
+               proc_signalend(p, 1);
+               proc_unlock(p);
                return(0);
        } 
        TAILQ_FOREACH(uth, &p->p_uthlist, uu_list) {
                uth->uu_sigmask = bit;
        }
        p->p_sigmask =  bit;
-       signal_unlock(p);
+       proc_signalend(p, 1);
+       proc_unlock(p);
+
        return(0);
 }
 
 /* XXX should be static? */
+/*
+ * Notes:      The thread parameter is used in the PPC case to select the
+ *             thread on which the floating point exception will be enabled
+ *             or disabled.  We can't simply take current_thread(), since
+ *             this is called from posix_spawn() on the not currently running
+ *             process/thread pair.
+ *
+ *             We mark thread as unused to alow compilation without warning
+ *             on non-PPC platforms.
+ */
 int
-setsigvec(struct proc *p, int signum, struct __user_sigaction *sa)
+setsigvec(proc_t p, __unused thread_t thread, int signum, struct __kern_sigaction *sa, boolean_t in_sigstart)
 {
-       register struct sigacts *ps = p->p_sigacts;
-       register int bit;
+       struct sigacts *ps = p->p_sigacts;
+       int bit;
+
+       assert(signum < NSIG);
 
        if ((signum == SIGKILL || signum == SIGSTOP) &&
                sa->sa_handler != SIG_DFL)
@@ -547,23 +684,15 @@ setsigvec(struct proc *p, int signum, struct __user_sigaction *sa)
                ps->ps_signodefer &= ~bit;
        if (signum == SIGCHLD) {
                if (sa->sa_flags & SA_NOCLDSTOP)
-                       p->p_flag |= P_NOCLDSTOP;
+                       OSBitOrAtomic(P_NOCLDSTOP, &p->p_flag);
                else
-                       p->p_flag &= ~P_NOCLDSTOP;
+                       OSBitAndAtomic(~((uint32_t)P_NOCLDSTOP), &p->p_flag);
                if ((sa->sa_flags & SA_NOCLDWAIT) || (sa->sa_handler == SIG_IGN))
-                       p->p_flag |= P_NOCLDWAIT;
+                       OSBitOrAtomic(P_NOCLDWAIT, &p->p_flag);
                else
-                       p->p_flag &= ~P_NOCLDWAIT;
+                       OSBitAndAtomic(~((uint32_t)P_NOCLDWAIT), &p->p_flag);
        }
 
-#ifdef __ppc__ 
-       if (signum == SIGFPE) {
-               if (sa->sa_handler == SIG_DFL || sa->sa_handler == SIG_IGN) 
-                       thread_enable_fpe(current_thread(), 0);
-               else
-                       thread_enable_fpe(current_thread(), 1);
-       }
-#endif  /* __ppc__ */
        /*
         * Set bit in p_sigignore for signals that are set to SIG_IGN,
         * and for signals set to SIG_DFL where the default is to ignore.
@@ -573,7 +702,7 @@ setsigvec(struct proc *p, int signum, struct __user_sigaction *sa)
        if (sa->sa_handler == SIG_IGN ||
            (sigprop[signum] & SA_IGNORE && sa->sa_handler == SIG_DFL)) {
 
-               clear_procsiglist(p, bit);
+               clear_procsiglist(p, bit, in_sigstart);
                if (signum != SIGCONT)
                        p->p_sigignore |= bit;  /* easier in psignal */
                p->p_sigcatch &= ~bit;
@@ -592,12 +721,11 @@ setsigvec(struct proc *p, int signum, struct __user_sigaction *sa)
  * set to ignore signals that are ignored by default.
  */
 void
-siginit(p)
-       struct proc *p;
+siginit(proc_t p)
 {
-       register int i;
+       int i;
 
-       for (i = 0; i < NSIG; i++)
+       for (i = 1; i < NSIG; i++)
                if (sigprop[i] & SA_IGNORE && i != SIGCONT)
                        p->p_sigignore |= sigmask(i);
 }
@@ -606,14 +734,26 @@ siginit(p)
  * Reset signals for an exec of the specified process.
  */
 void
-execsigs(p, thr_act)
-       register struct proc *p;
-       register thread_t thr_act;
+execsigs(proc_t p, thread_t thread)
 {
-       register struct sigacts *ps = p->p_sigacts;
-       register int nc, mask;
+       struct sigacts *ps = p->p_sigacts;
+       int nc, mask;
        struct uthread *ut;
 
+       ut = (struct uthread *)get_bsdthread_info(thread);
+
+       /*
+        * transfer saved signal states from the process
+        * back to the current thread.
+        *
+        * NOTE: We do this without the process locked,
+        * because we are guaranteed to be single-threaded
+        * by this point in exec and the p_siglist is
+        * only accessed by threads inside the process.
+        */
+       ut->uu_siglist |= p->p_siglist;
+       p->p_siglist = 0;
+
        /*
         * Reset caught signals.  Held signals remain held
         * through p_sigmask (unless they were caught,
@@ -626,23 +766,22 @@ execsigs(p, thr_act)
                if (sigprop[nc] & SA_IGNORE) {
                        if (nc != SIGCONT)
                                p->p_sigignore |= mask;
-                       if (thr_act){
-                               ut = (struct uthread *)get_bsdthread_info(thr_act);
-                               ut->uu_siglist &= ~mask;
-                               p->p_siglist &= ~mask;
-                       } else
-                               clear_procsiglist(p, mask);
+                       ut->uu_siglist &= ~mask;
                }
                ps->ps_sigact[nc] = SIG_DFL;
        }
+
        /*
         * Reset stack state to the user stack.
         * Clear set of signals caught on the signal stack.
         */
-       ps->ps_sigstk.ss_flags = SA_DISABLE;
-       ps->ps_sigstk.ss_size = 0;
-       ps->ps_sigstk.ss_sp = USER_ADDR_NULL;
-       ps->ps_flags = 0;
+       /* thread */
+       ut->uu_sigstk.ss_flags = SA_DISABLE;
+       ut->uu_sigstk.ss_size = 0;
+       ut->uu_sigstk.ss_sp = USER_ADDR_NULL;
+       ut->uu_flag &= ~UT_ALTSTACK;
+       /* process */
+       ps->ps_sigonstack = 0;
 }
 
 /*
@@ -652,7 +791,7 @@ execsigs(p, thr_act)
  * the library stub does the rest.
  */
 int
-sigprocmask(register struct proc *p, struct sigprocmask_args *uap, __unused register_t *retval)
+sigprocmask(proc_t p, struct sigprocmask_args *uap, __unused int32_t *retval)
 {
        int error = 0;
        sigset_t oldmask, nmask;
@@ -697,7 +836,7 @@ out:
 }
 
 int
-sigpending(__unused struct proc *p, register struct sigpending_args *uap, __unused register_t *retval)
+sigpending(__unused proc_t p, struct sigpending_args *uap, __unused int32_t *retval)
 {
        struct uthread *ut;
        sigset_t pendlist;
@@ -710,7 +849,6 @@ sigpending(__unused struct proc *p, register struct sigpending_args *uap, __unus
        return(0);
 }
 
-
 /*
  * Suspend process until signal, providing mask to be set
  * in the meantime.  Note nonstandard calling convention:
@@ -721,11 +859,18 @@ static int
 sigcontinue(__unused int error)
 {
 //     struct uthread *ut = get_bsdthread_info(current_thread());
-  unix_syscall_return(EINTR);
+       unix_syscall_return(EINTR);
+}
+
+int
+sigsuspend(proc_t p, struct sigsuspend_args *uap, int32_t *retval)
+{
+       __pthread_testcancel(1);
+       return(sigsuspend_nocancel(p, (struct sigsuspend_nocancel_args *)uap, retval));
 }
 
 int
-sigsuspend(register struct proc *p, struct sigsuspend_args *uap, __unused register_t *retval)
+sigsuspend_nocancel(proc_t p, struct sigsuspend_nocancel_args *uap, __unused int32_t *retval)
 {
        struct uthread *ut;
 
@@ -748,29 +893,47 @@ sigsuspend(register struct proc *p, struct sigsuspend_args *uap, __unused regist
 
 
 int
-__disable_threadsignal(struct proc *p, 
-                                          __unused register struct __disable_threadsignal_args *uap, 
-                                          __unused register_t *retval)
+__disable_threadsignal(__unused proc_t p,
+                      __unused struct __disable_threadsignal_args *uap,
+                      __unused int32_t *retval)
 {
        struct uthread *uth;
 
        uth = (struct uthread *)get_bsdthread_info(current_thread());
 
        /* No longer valid to have any signal delivered */
-       signal_lock(p);
-       uth->uu_flag |= UT_NO_SIGMASK;
-       signal_unlock(p);
+       uth->uu_flag |= (UT_NO_SIGMASK | UT_CANCELDISABLE);
 
        return(0);
 
 }
 
+void
+__pthread_testcancel(int presyscall)
+{
+
+       thread_t self = current_thread();
+       struct uthread * uthread;
+
+       uthread = (struct uthread *)get_bsdthread_info(self);
+
+       
+       uthread->uu_flag &= ~UT_NOTCANCELPT;
+
+       if ((uthread->uu_flag & (UT_CANCELDISABLE | UT_CANCEL | UT_CANCELED)) == UT_CANCEL) {
+               if(presyscall != 0) {
+                       unix_syscall_return(EINTR);
+                       /* NOTREACHED */
+               } else 
+                       thread_abort_safely(self);
+       }
+}
+
+
 
 int
-__pthread_markcancel(p, uap, retval)
-       struct proc *p;
-       register struct __pthread_markcancel_args *uap;
-       register_t *retval;
+__pthread_markcancel(__unused proc_t p,
+       struct __pthread_markcancel_args *uap, __unused int32_t *retval)
 {
        thread_act_t target_act;
        int error = 0;
@@ -784,7 +947,7 @@ __pthread_markcancel(p, uap, retval)
        uth = (struct uthread *)get_bsdthread_info(target_act);
 
        /* if the thread is in vfork do not cancel */
-       if ((uth->uu_flag & (P_VFORK | UT_CANCEL | UT_CANCELED )) == 0) {
+       if ((uth->uu_flag & (UT_VFORK | UT_CANCEL | UT_CANCELED )) == 0) {
                uth->uu_flag |= (UT_CANCEL | UT_NO_SIGMASK);
                if (((uth->uu_flag & UT_NOTCANCELPT) == 0) 
                        && ((uth->uu_flag & UT_CANCELDISABLE) == 0))
@@ -801,17 +964,15 @@ __pthread_markcancel(p, uap, retval)
  * if action = 2; Disable the cancel handling
  */
 int
-__pthread_canceled(p, uap, retval)
-       struct proc *p;
-       register struct __pthread_canceled_args *uap;
-       register_t *retval;
+__pthread_canceled(__unused proc_t p,
+       struct __pthread_canceled_args *uap, __unused int32_t *retval)
 {
-       thread_act_t thr_act;
+       thread_act_t thread;
        struct uthread *uth;
        int action = uap->action;
 
-       thr_act = current_act();
-       uth = (struct uthread *)get_bsdthread_info(thr_act);
+       thread = current_thread();
+       uth = (struct uthread *)get_bsdthread_info(thread);
 
        switch (action) {
                case 1:
@@ -833,6 +994,7 @@ __pthread_canceled(p, uap, retval)
        return(EINVAL);
 }
 
+__attribute__((noreturn))
 void
 __posix_sem_syscall_return(kern_return_t kern_result) 
 {
@@ -850,50 +1012,184 @@ __posix_sem_syscall_return(kern_return_t kern_result)
        /* does not return */
 }
 
-
+#if OLD_SEMWAIT_SIGNAL
+/*
+ * Returns:    0                       Success
+ *             EINTR
+ *             ETIMEDOUT
+ *             EINVAL
+ *      EFAULT if timespec is NULL
+ */
 int
-__semwait_signal(p, uap, retval)
-       struct proc *p;
-       register struct __semwait_signal_args *uap;
-       register_t *retval;
+__old_semwait_signal(proc_t p, struct __old_semwait_signal_args *uap,
+                     int32_t *retval)
 {
+       __pthread_testcancel(0);
+       return(__old_semwait_signal_nocancel(p, (struct __old_semwait_signal_nocancel_args *)uap, retval));
+}
 
+int
+__old_semwait_signal_nocancel(proc_t p, struct __old_semwait_signal_nocancel_args *uap,
+                              __unused int32_t *retval)
+{
+       
        kern_return_t kern_result;
+       int error;
        mach_timespec_t then;
        struct timespec now;
-
+       struct user_timespec ts;
+       boolean_t truncated_timeout = FALSE;
+       
        if(uap->timeout) {
-
+               
+               if (IS_64BIT_PROCESS(p)) {
+                       struct user64_timespec ts64;
+                       error = copyin(uap->ts, &ts64, sizeof(ts64));
+                       ts.tv_sec = ts64.tv_sec;
+                       ts.tv_nsec = ts64.tv_nsec;
+               } else {
+                       struct user32_timespec ts32;
+                       error = copyin(uap->ts, &ts32, sizeof(ts32));
+                       ts.tv_sec = ts32.tv_sec;
+                       ts.tv_nsec = ts32.tv_nsec;
+               }
+               
+               if (error) {
+                       return error;
+               }
+               
+               if ((ts.tv_sec & 0xFFFFFFFF00000000ULL) != 0) {
+                       ts.tv_sec = 0xFFFFFFFF;
+                       ts.tv_nsec = 0;
+                       truncated_timeout = TRUE;
+               }
+               
                if (uap->relative) {
-                       then.tv_sec = uap->tv_sec;
-                       then.tv_nsec = uap->tv_nsec;
+                       then.tv_sec = ts.tv_sec;
+                       then.tv_nsec = ts.tv_nsec;
                } else {
                        nanotime(&now);
-                       then.tv_sec = uap->tv_sec - now.tv_sec;
-                       then.tv_nsec = uap->tv_nsec - now.tv_nsec;
-                       if (then.tv_nsec < 0) {
-                               then.tv_nsec += NSEC_PER_SEC;
-                               then.tv_sec--; 
+                       
+                       /* if time has elapsed, set time to null timepsec to bailout rightaway */
+                       if (now.tv_sec == ts.tv_sec ?
+                               now.tv_nsec > ts.tv_nsec :
+                               now.tv_sec > ts.tv_sec) {
+                               then.tv_sec = 0;
+                               then.tv_nsec = 0;
+                       } else {
+                               then.tv_sec = ts.tv_sec - now.tv_sec;
+                               then.tv_nsec = ts.tv_nsec - now.tv_nsec;
+                               if (then.tv_nsec < 0) {
+                                       then.tv_nsec += NSEC_PER_SEC;
+                                       then.tv_sec--; 
+                               }
                        }
                }
-
-               if (uap->mutex_sem == (void *)NULL)
-                       kern_result = semaphore_timedwait_trap_internal(uap->cond_sem, then.tv_sec, then.tv_nsec, __posix_sem_syscall_return);
+               
+               if (uap->mutex_sem == 0)
+                       kern_result = semaphore_timedwait_trap_internal((mach_port_name_t)uap->cond_sem, then.tv_sec, then.tv_nsec, __posix_sem_syscall_return);
                else
                        kern_result = semaphore_timedwait_signal_trap_internal(uap->cond_sem, uap->mutex_sem, then.tv_sec, then.tv_nsec, __posix_sem_syscall_return);
-
+               
        } else {
-
-               if (uap->mutex_sem == (void *)NULL)
+               
+               if (uap->mutex_sem == 0)
                        kern_result = semaphore_wait_trap_internal(uap->cond_sem, __posix_sem_syscall_return);
                else
-
+                       
                        kern_result = semaphore_wait_signal_trap_internal(uap->cond_sem, uap->mutex_sem, __posix_sem_syscall_return);
        }
-
-out:
-       if (kern_result == KERN_SUCCESS)
+       
+       if (kern_result == KERN_SUCCESS && !truncated_timeout)
                return(0);
+       else if (kern_result == KERN_SUCCESS && truncated_timeout)
+               return(EINTR); /* simulate an exceptional condition because Mach doesn't support a longer timeout */
+       else if (kern_result == KERN_ABORTED)
+               return(EINTR);
+       else if (kern_result == KERN_OPERATION_TIMED_OUT) 
+               return(ETIMEDOUT);
+       else
+               return(EINVAL);
+}
+#endif /* OLD_SEMWAIT_SIGNAL*/
+
+/*
+ * Returns:    0                       Success
+ *             EINTR
+ *             ETIMEDOUT
+ *             EINVAL
+ *      EFAULT if timespec is NULL
+ */
+int
+__semwait_signal(proc_t p, struct __semwait_signal_args *uap,
+                     int32_t *retval)
+{
+       __pthread_testcancel(0);
+       return(__semwait_signal_nocancel(p, (struct __semwait_signal_nocancel_args *)uap, retval));
+}
+
+int
+__semwait_signal_nocancel(__unused proc_t p, struct __semwait_signal_nocancel_args *uap,
+                              __unused int32_t *retval)
+{
+       
+       kern_return_t kern_result;
+       mach_timespec_t then;
+       struct timespec now;
+       struct user_timespec ts;
+        boolean_t truncated_timeout = FALSE;
+       
+       if(uap->timeout) {
+                
+               ts.tv_sec = uap->tv_sec;
+                ts.tv_nsec = uap->tv_nsec;
+
+                if ((ts.tv_sec & 0xFFFFFFFF00000000ULL) != 0) {
+                        ts.tv_sec = 0xFFFFFFFF;
+                        ts.tv_nsec = 0;
+                        truncated_timeout = TRUE;
+                }              
+               
+               if (uap->relative) {
+                       then.tv_sec = ts.tv_sec;
+                       then.tv_nsec = ts.tv_nsec;
+               } else {
+                       nanotime(&now);
+
+                        /* if time has elapsed, set time to null timepsec to bailout rightaway */
+                        if (now.tv_sec == ts.tv_sec ?
+                                now.tv_nsec > ts.tv_nsec :
+                                now.tv_sec > ts.tv_sec) {
+                                then.tv_sec = 0;
+                                then.tv_nsec = 0;
+                        } else {
+                                then.tv_sec = ts.tv_sec - now.tv_sec;
+                                then.tv_nsec = ts.tv_nsec - now.tv_nsec;
+                                if (then.tv_nsec < 0) {
+                                        then.tv_nsec += NSEC_PER_SEC;
+                                        then.tv_sec--;
+                                }
+                        }
+               }
+                                       
+               if (uap->mutex_sem == 0)
+                       kern_result = semaphore_timedwait_trap_internal((mach_port_name_t)uap->cond_sem, then.tv_sec, then.tv_nsec, __posix_sem_syscall_return);
+               else
+                       kern_result = semaphore_timedwait_signal_trap_internal(uap->cond_sem, uap->mutex_sem, then.tv_sec, then.tv_nsec, __posix_sem_syscall_return);
+               
+       } else {
+               
+               if (uap->mutex_sem == 0)
+                       kern_result = semaphore_wait_trap_internal(uap->cond_sem, __posix_sem_syscall_return);
+               else
+                       
+                       kern_result = semaphore_wait_signal_trap_internal(uap->cond_sem, uap->mutex_sem, __posix_sem_syscall_return);
+       }
+       
+       if (kern_result == KERN_SUCCESS && !truncated_timeout)
+                return(0);
+        else if (kern_result == KERN_SUCCESS && truncated_timeout)
+                return(EINTR); /* simulate an exceptional condition because Mach doesn't support a longer timeout */
        else if (kern_result == KERN_ABORTED)
                return(EINTR);
        else if (kern_result == KERN_OPERATION_TIMED_OUT) 
@@ -904,9 +1200,8 @@ out:
 
 
 int 
-__pthread_kill(__unused struct proc *p, 
-                          register struct __pthread_kill_args *uap, 
-                          __unused register_t *retval) 
+__pthread_kill(__unused proc_t p, struct __pthread_kill_args *uap,
+              __unused int32_t *retval) 
 {
        thread_t target_act;
        int error = 0;
@@ -938,9 +1233,8 @@ out:
 
 
 int 
-pthread_sigmask(__unused register struct proc *p, 
-                               register struct pthread_sigmask_args *uap, 
-                               __unused register_t *retval)
+__pthread_sigmask(__unused proc_t p, struct __pthread_sigmask_args *uap,
+                 __unused int32_t *retval)
 {
        user_addr_t set = uap->set;
        user_addr_t oset = uap->oset;
@@ -987,9 +1281,21 @@ out:
        return(error);
 }
 
+/*
+ * Returns:    0                       Success
+ *             EINVAL
+ *     copyin:EFAULT
+ *     copyout:EFAULT
+ */
+int 
+__sigwait(proc_t p, struct __sigwait_args *uap, int32_t *retval)
+{
+       __pthread_testcancel(1);
+       return(__sigwait_nocancel(p, (struct __sigwait_nocancel_args *)uap, retval));
+}
 
 int 
-sigwait(register struct proc *p, register struct sigwait_args *uap, __unused register_t *retval)
+__sigwait_nocancel(proc_t p, struct __sigwait_nocancel_args *uap, __unused int32_t *retval)
 {
        struct uthread *ut;
        struct uthread *uth;
@@ -1013,18 +1319,20 @@ sigwait(register struct proc *p, register struct sigwait_args *uap, __unused reg
        if (siglist == 0)
                return(EINVAL);
 
-       signal_lock(p);
-       if ((p->p_flag & P_INVFORK) && p->p_vforkact) {
-               signal_unlock(p);
+       proc_lock(p);
+       if ((p->p_lflag & P_LINVFORK) && p->p_vforkact) {
+               proc_unlock(p);
                return(EINVAL);
        } else {
+               proc_signalstart(p, 1);
                TAILQ_FOREACH(uth, &p->p_uthlist, uu_list) {
                        if ( (sigw = uth->uu_siglist & siglist) ) {
                                break;
                        }
                }
+               proc_signalend(p, 1);
        }
-       signal_unlock(p);
+
        if (sigw) {
                /* The signal was pending on a thread */
                goto sigwait1;
@@ -1036,17 +1344,21 @@ sigwait(register struct proc *p, register struct sigwait_args *uap, __unused reg
         * save it here and mark the sigacts structure
         * to indicate this.
         */
+       uth = ut;               /* wait for it to be delivered to us */
        ut->uu_oldmask = ut->uu_sigmask;
        ut->uu_flag |= UT_SAS_OLDMASK;
-       if (siglist == (sigset_t)0)
+       if (siglist == (sigset_t)0) {
+               proc_unlock(p);
                return(EINVAL);
+       }
        /* SIGKILL and SIGSTOP are not maskable as well */
        ut->uu_sigmask = ~(siglist|sigcantmask);
        ut->uu_sigwait = siglist; 
+
        /* No Continuations for now */
-       error =  tsleep((caddr_t)&ut->uu_sigwait, PPAUSE|PCATCH, "pause", 0);
+       error =  msleep((caddr_t)&ut->uu_sigwait, &p->p_mlock, PPAUSE|PCATCH, "pause", 0);
 
-       if ((error == EINTR) || (error == ERESTART))
+       if (error == ERESTART)
                error = 0;
 
        sigw = (ut->uu_sigwait & siglist);
@@ -1059,33 +1371,47 @@ sigwait1:
                signum = ffs((unsigned int)sigw);
                if (!signum)
                        panic("sigwait with no signal wakeup");
-               ut->uu_siglist &= ~(sigmask(signum));
+               /* Clear the pending signal in the thread it was delivered */
+               uth->uu_siglist &= ~(sigmask(signum));
+
+#if CONFIG_DTRACE
+               DTRACE_PROC2(signal__clear, int, signum, siginfo_t *, &(ut->t_dtrace_siginfo));
+#endif
+
+               proc_unlock(p);
                if (uap->sig != USER_ADDR_NULL)
                                error = copyout(&signum, uap->sig, sizeof(int));
-       }
+       } else
+               proc_unlock(p);
 
        return(error);
 
 }
 
-
 int
-sigaltstack(struct proc *p, register struct sigaltstack_args *uap, __unused register_t *retval)
+sigaltstack(__unused proc_t p, struct sigaltstack_args *uap, __unused int32_t *retval)
 {
-       struct sigacts *psp;
-       struct user_sigaltstack ss;
+       struct kern_sigaltstack ss;
+       struct kern_sigaltstack *pstk;
        int error;
+       struct uthread *uth;
+       int onstack;
+
+       uth = (struct uthread *)get_bsdthread_info(current_thread());
 
-       psp = p->p_sigacts;
-       if ((psp->ps_flags & SAS_ALTSTACK) == 0)
-               psp->ps_sigstk.ss_flags |= SA_DISABLE;
+       pstk = &uth->uu_sigstk;
+       if ((uth->uu_flag & UT_ALTSTACK) == 0)
+               uth->uu_sigstk.ss_flags |= SA_DISABLE;
+       onstack = pstk->ss_flags & SA_ONSTACK;
        if (uap->oss) {
                if (IS_64BIT_PROCESS(p)) {
-                       error = copyout(&psp->ps_sigstk, uap->oss, sizeof(struct user_sigaltstack));
+                       struct user64_sigaltstack ss64;
+                       sigaltstack_kern_to_user64(pstk, &ss64);                        
+                       error = copyout(&ss64, uap->oss, sizeof(ss64));
                } else {
-                       struct sigaltstack ss32;
-                       sigaltstack_64to32(&psp->ps_sigstk, &ss32);
-                       error = copyout(&ss32, uap->oss, sizeof(struct sigaltstack));
+                       struct user32_sigaltstack ss32;
+                       sigaltstack_kern_to_user32(pstk, &ss32);                        
+                       error = copyout(&ss32, uap->oss, sizeof(ss32));
                }
                if (error)
                        return (error);
@@ -1093,11 +1419,13 @@ sigaltstack(struct proc *p, register struct sigaltstack_args *uap, __unused regi
        if (uap->nss == USER_ADDR_NULL)
                return (0);
        if (IS_64BIT_PROCESS(p)) {
-               error = copyin(uap->nss, &ss, sizeof(struct user_sigaltstack));
+               struct user64_sigaltstack ss64;
+               error = copyin(uap->nss, &ss64, sizeof(ss64));
+               sigaltstack_user64_to_kern(&ss64, &ss);
        } else {
-               struct sigaltstack ss32;
-               error = copyin(uap->nss, &ss32, sizeof(struct sigaltstack));
-               sigaltstack_32to64(&ss32,&ss);
+               struct user32_sigaltstack ss32;
+               error = copyin(uap->nss, &ss32, sizeof(ss32));
+               sigaltstack_user32_to_kern(&ss32, &ss);
        }
        if (error)
                return (error);
@@ -1106,26 +1434,30 @@ sigaltstack(struct proc *p, register struct sigaltstack_args *uap, __unused regi
        }
 
        if (ss.ss_flags & SA_DISABLE) {
-               if (psp->ps_sigstk.ss_flags & SA_ONSTACK)
+               /* if we are here we are not in the signal handler ;so no need to check */
+               if (uth->uu_sigstk.ss_flags & SA_ONSTACK)
                        return (EINVAL);
-               psp->ps_flags &= ~SAS_ALTSTACK;
-               psp->ps_sigstk.ss_flags = ss.ss_flags;
+               uth->uu_flag &= ~UT_ALTSTACK;
+               uth->uu_sigstk.ss_flags = ss.ss_flags;
                return (0);
        }
+       if (onstack)
+               return (EPERM);
 /* The older stacksize was 8K, enforce that one so no compat problems */
 #define OLDMINSIGSTKSZ 8*1024
        if (ss.ss_size < OLDMINSIGSTKSZ)
                return (ENOMEM);
-       psp->ps_flags |= SAS_ALTSTACK;
-       psp->ps_sigstk= ss;
+       uth->uu_flag |= UT_ALTSTACK;
+       uth->uu_sigstk= ss;
        return (0);
 }
 
 int
-kill(struct proc *cp, struct kill_args *uap, __unused register_t *retval)
+kill(proc_t cp, struct kill_args *uap, __unused int32_t *retval)
 {
-       register struct proc *p;
+       proc_t p;
        kauth_cred_t uc = kauth_cred_get();
+       int posix = uap->posix;         /* !0 if posix behaviour desired */
 
        AUDIT_ARG(pid, uap->pid);
        AUDIT_ARG(signum, uap->signum);
@@ -1134,7 +1466,7 @@ kill(struct proc *cp, struct kill_args *uap, __unused register_t *retval)
                return (EINVAL);
        if (uap->pid > 0) {
                /* kill single process */
-               if ((p = proc_findref(uap->pid)) == NULL) {
+               if ((p = proc_find(uap->pid)) == NULL) {
                        if ((p = pzfind(uap->pid)) != NULL) {
                                /*
                                 * IEEE Std 1003.1-2001: return success
@@ -1145,772 +1477,839 @@ kill(struct proc *cp, struct kill_args *uap, __unused register_t *retval)
                        return (ESRCH);
                }
                AUDIT_ARG(process, p);
-               if (!cansignal(cp, uc, p, uap->signum)) {
-                       proc_dropref(p);
+               if (!cansignal(cp, uc, p, uap->signum, 0)) {
+                       proc_rele(p);
                        return(EPERM);
                }
                if (uap->signum)
                        psignal(p, uap->signum);
-               proc_dropref(p);
+               proc_rele(p);
                return (0);
        }
        switch (uap->pid) {
        case -1:                /* broadcast signal */
-               return (killpg1(cp, uap->signum, 0, 1));
+               return (killpg1(cp, uap->signum, 0, 1, posix));
        case 0:                 /* signal own process group */
-               return (killpg1(cp, uap->signum, 0, 0));
+               return (killpg1(cp, uap->signum, 0, 0, posix));
        default:                /* negative explicit process group */
-               return (killpg1(cp, uap->signum, -(uap->pid), 0));
+               return (killpg1(cp, uap->signum, -(uap->pid), 0, posix));
        }
        /* NOTREACHED */
 }
 
-
-/*
- * Common code for kill process group/broadcast kill.
- * cp is calling process.
- */
-int
-killpg1(cp, signum, pgid, all)
-       register struct proc *cp;
-       int signum, pgid, all;
+os_reason_t
+build_userspace_exit_reason(uint32_t reason_namespace, uint64_t reason_code, user_addr_t payload, uint32_t payload_size,
+                                       user_addr_t reason_string, uint64_t reason_flags)
 {
-       register struct proc *p;
-       kauth_cred_t uc = cp->p_ucred;
-       struct pgrp *pgrp;
-       int nfound = 0;
-       
-       if (all) {
-               /* 
-                * broadcast 
-                */
-               for (p = allproc.lh_first; p != 0; p = p->p_list.le_next) {
-                       if (p->p_pid <= 1 || p->p_flag & P_SYSTEM || 
-                           p == cp || !cansignal(cp, uc, p, signum))
-                               continue;
-                       nfound++;
-                       if (signum)
-                               psignal(p, signum);
+       os_reason_t exit_reason = OS_REASON_NULL;
+
+       int error = 0;
+       int num_items_to_copy = 0;
+       uint32_t user_data_to_copy = 0;
+       char *reason_user_desc = NULL;
+       size_t reason_user_desc_len = 0;
+
+       exit_reason = os_reason_create(reason_namespace, reason_code);
+       if (exit_reason == OS_REASON_NULL) {
+               printf("build_userspace_exit_reason: failed to allocate exit reason\n");
+               return exit_reason;
+       }
+
+       exit_reason->osr_flags |= OS_REASON_FLAG_FROM_USERSPACE;
+
+       /*
+        * Only apply flags that are allowed to be passed from userspace.
+        */
+       exit_reason->osr_flags |= (reason_flags & OS_REASON_FLAG_MASK_ALLOWED_FROM_USER);
+       if ((reason_flags & OS_REASON_FLAG_MASK_ALLOWED_FROM_USER) != reason_flags) {
+               printf("build_userspace_exit_reason: illegal flags passed from userspace (some masked off) 0x%llx, ns: %u, code 0x%llx\n",
+                       reason_flags, reason_namespace, reason_code);
+       }
+
+       if (!(exit_reason->osr_flags & OS_REASON_FLAG_NO_CRASH_REPORT)) {
+               exit_reason->osr_flags |= OS_REASON_FLAG_GENERATE_CRASH_REPORT;
+       }
+
+       if (payload != USER_ADDR_NULL) {
+               if (payload_size == 0) {
+                       printf("build_userspace_exit_reason: exit reason with namespace %u, nonzero payload but zero length\n",
+                                       reason_namespace);
+                       exit_reason->osr_flags |= OS_REASON_FLAG_BAD_PARAMS;
+                       payload = USER_ADDR_NULL;
+               } else {
+                       num_items_to_copy++;
+
+                       if (payload_size > EXIT_REASON_PAYLOAD_MAX_LEN) {
+                               exit_reason->osr_flags |= OS_REASON_FLAG_PAYLOAD_TRUNCATED;
+                               payload_size = EXIT_REASON_PAYLOAD_MAX_LEN;
+                       }
+
+                       user_data_to_copy += payload_size;
                }
-       } else {
-               if (pgid == 0)          
-                       /* 
-                        * zero pgid means send to my process group.
-                        */
-                       pgrp = cp->p_pgrp;
-               else {
-                       pgrp = pgfind(pgid);
-                       if (pgrp == NULL)
-                               return (ESRCH);
+       }
+
+       if (reason_string != USER_ADDR_NULL) {
+               reason_user_desc = (char *) kalloc(EXIT_REASON_USER_DESC_MAX_LEN);
+
+               if (reason_user_desc != NULL) {
+                       error = copyinstr(reason_string, (void *) reason_user_desc,
+                                               EXIT_REASON_USER_DESC_MAX_LEN, &reason_user_desc_len);
+
+                       if (error == 0) {
+                               num_items_to_copy++;
+                               user_data_to_copy += reason_user_desc_len;
+                       } else if (error == ENAMETOOLONG) {
+                               num_items_to_copy++;
+                               reason_user_desc[EXIT_REASON_USER_DESC_MAX_LEN - 1] = '\0';
+                               user_data_to_copy += reason_user_desc_len;
+                       } else {
+                               exit_reason->osr_flags |= OS_REASON_FLAG_FAILED_DATA_COPYIN;
+                               kfree(reason_user_desc, EXIT_REASON_USER_DESC_MAX_LEN);
+                               reason_user_desc = NULL;
+                               reason_user_desc_len = 0;
+                       }
                }
-               for (p = pgrp->pg_members.lh_first; p != 0;
-                    p = p->p_pglist.le_next) {
-                       if (p->p_pid <= 1 || p->p_flag & P_SYSTEM ||
-                           p->p_stat == SZOMB ||
-                           !cansignal(cp, uc, p, signum))
-                               continue;
-                       nfound++;
-                       if (signum)
-                               psignal(p, signum);
+       }
+
+       if (num_items_to_copy != 0) {
+               uint32_t reason_buffer_size_estimate = 0;
+               mach_vm_address_t data_addr = 0;
+
+               reason_buffer_size_estimate = kcdata_estimate_required_buffer_size(num_items_to_copy, user_data_to_copy);
+
+               error = os_reason_alloc_buffer(exit_reason, reason_buffer_size_estimate);
+               if (error != 0) {
+                       printf("build_userspace_exit_reason: failed to allocate signal reason buffer\n");
+                       goto out_failed_copyin;
+               }
+
+               if (reason_user_desc != NULL && reason_user_desc_len != 0) {
+                       if (KERN_SUCCESS == kcdata_get_memory_addr(&exit_reason->osr_kcd_descriptor,
+                                               EXIT_REASON_USER_DESC,
+                                               reason_user_desc_len,
+                                               &data_addr)) {
+
+                               kcdata_memcpy(&exit_reason->osr_kcd_descriptor, (mach_vm_address_t) data_addr,
+                                               reason_user_desc, reason_user_desc_len);
+                       } else {
+                               printf("build_userspace_exit_reason: failed to allocate space for reason string\n");
+                               goto out_failed_copyin;
+                       }
+               }
+
+               if (payload != USER_ADDR_NULL) {
+                       if (KERN_SUCCESS ==
+                                       kcdata_get_memory_addr(&exit_reason->osr_kcd_descriptor,
+                                               EXIT_REASON_USER_PAYLOAD,
+                                               payload_size,
+                                               &data_addr)) {
+                               error = copyin(payload, (void *) data_addr, payload_size);
+                               if (error) {
+                                       printf("build_userspace_exit_reason: failed to copy in payload data with error %d\n", error);
+                                       goto out_failed_copyin;
+                               }
+                       } else {
+                               printf("build_userspace_exit_reason: failed to allocate space for payload data\n");
+                               goto out_failed_copyin;
+                       }
                }
        }
-       return (nfound ? 0 : ESRCH);
-}
 
-/*
- * Send a signal to a process group.
- */
-void
-gsignal(pgid, signum)
-       int pgid, signum;
-{
-       struct pgrp *pgrp;
+       if (reason_user_desc != NULL) {
+               kfree(reason_user_desc, EXIT_REASON_USER_DESC_MAX_LEN);
+               reason_user_desc = NULL;
+               reason_user_desc_len = 0;
+       }
 
-       if (pgid && (pgrp = pgfind(pgid)))
-               pgsignal(pgrp, signum, 0);
-}
+       return exit_reason;
 
-/*
- * Send a signal to a process group.  If checktty is 1,
- * limit to members which have a controlling terminal.
- */
-void
-pgsignal(pgrp, signum, checkctty)
-       struct pgrp *pgrp;
-       int signum, checkctty;
-{
-       register struct proc *p;
+out_failed_copyin:
 
-       if (pgrp)
-               for (p = pgrp->pg_members.lh_first; p != 0;
-                    p = p->p_pglist.le_next)
-                       if (checkctty == 0 || p->p_flag & P_CONTROLT)
-                               psignal(p, signum);
+       if (reason_user_desc != NULL) {
+               kfree(reason_user_desc, EXIT_REASON_USER_DESC_MAX_LEN);
+               reason_user_desc = NULL;
+               reason_user_desc_len = 0;
+       }
+
+       exit_reason->osr_flags |= OS_REASON_FLAG_FAILED_DATA_COPYIN;
+       os_reason_alloc_buffer(exit_reason, 0);
+       return exit_reason;
 }
 
-/*
- * Send signal to a backgrounded process blocked due to tty access
- * In FreeBSD, the backgrounded process wakes up every second and
- * discovers whether it is foregounded or not. In our case, we block 
- * the thread in tsleep as we want to avoid storm of processes as well 
- * as the suspend is only at AST level
- */
-void
-tty_pgsignal(pgrp, signum)
-       struct pgrp *pgrp;
-       int signum;
+static int
+terminate_with_payload_internal(struct proc *cur_proc, int target_pid, uint32_t reason_namespace,
+                               uint64_t reason_code, user_addr_t payload, uint32_t payload_size,
+                               user_addr_t reason_string, uint64_t reason_flags)
 {
-       register struct proc *p;
+       proc_t target_proc = PROC_NULL;
+       kauth_cred_t cur_cred = kauth_cred_get();
 
-       if (pgrp)
-               for (p = pgrp->pg_members.lh_first; p != 0;
-                    p = p->p_pglist.le_next)
-                       if ((p->p_flag & P_TTYSLEEP) && (p->p_flag & P_CONTROLT))
-                               psignal(p, signum);
-}
+       os_reason_t signal_reason = OS_REASON_NULL;
 
-/*
- * Send a signal caused by a trap to a specific thread.
- */
-void
-threadsignal(thread_t sig_actthread, int signum, u_long code)
-{
-       register struct uthread *uth;
-       register struct task * sig_task;
-       register struct proc *p ;
-       int mask;
+       AUDIT_ARG(pid, target_pid);
+       if ((target_pid <= 0)) {
+               return EINVAL;
+       }
 
-       if ((u_int)signum >= NSIG || signum == 0)
-               return;
+       target_proc = proc_find(target_pid);
+       if (target_proc == PROC_NULL) {
+               return ESRCH;
+       }
 
-       mask = sigmask(signum);
-       if ((mask & threadmask) == 0)
-               return;
-       sig_task = get_threadtask(sig_actthread);
-       p = (struct proc *)(get_bsdtask_info(sig_task));
+       AUDIT_ARG(process, target_proc);
 
-       uth = get_bsdthread_info(sig_actthread);
-       if (uth && (uth->uu_flag & UT_VFORK))
-               p = uth->uu_proc;
+       if (!cansignal(cur_proc, cur_cred, target_proc, SIGKILL, 0)) {
+               proc_rele(target_proc);
+               return EPERM;
+       }
 
-       if (!(p->p_flag & P_TRACED) && (p->p_sigignore & mask))
-               return;
+       KERNEL_DEBUG_CONSTANT(BSDDBG_CODE(DBG_BSD_PROC, BSD_PROC_EXITREASON_CREATE) | DBG_FUNC_NONE,
+                                       target_proc->p_pid, reason_namespace,
+                                       reason_code, 0, 0);
 
-       uth->uu_siglist |= mask;
-       p->p_siglist |= mask;   /* just for lame ones looking here */
-       uth->uu_code = code;
-       /* mark on process as well */
-       signal_setast(sig_actthread);
+       signal_reason = build_userspace_exit_reason(reason_namespace, reason_code, payload, payload_size,
+                                                       reason_string, reason_flags);
+
+       if (target_pid == cur_proc->p_pid) {
+               /*
+                * psignal_thread_with_reason() will pend a SIGKILL on the specified thread or
+                * return if the thread and/or task are already terminating. Either way, the
+                * current thread won't return to userspace.
+                */
+               psignal_thread_with_reason(target_proc, current_thread(), SIGKILL, signal_reason);
+       } else {
+               psignal_with_reason(target_proc, SIGKILL, signal_reason);
+       }
+
+       proc_rele(target_proc);
+
+       return 0;
 }
 
-void
-psignal(p, signum)
-       register struct proc *p;
-       register int signum;
+int
+terminate_with_payload(struct proc *cur_proc, struct terminate_with_payload_args *args,
+                               __unused int32_t *retval)
 {
-       psignal_lock(p, signum, 1);
+       return terminate_with_payload_internal(cur_proc, args->pid, args->reason_namespace, args->reason_code, args->payload,
+                                               args->payload_size, args->reason_string, args->reason_flags);
 }
 
-void
-psignal_vfork(struct proc *p, task_t new_task, thread_t thr_act, int signum)
+static int
+killpg1_filt(proc_t p, void * arg)
 {
-       register int prop;
-       register sig_t action;
-       int mask;
-       struct uthread *uth;
+       struct killpg1_filtargs * kfargp = (struct killpg1_filtargs *)arg;
+       proc_t cp = kfargp->cp;
+       int posix = kfargp->posix;
 
-       if ((u_int)signum >= NSIG || signum == 0)
-               panic("psignal signal number");
-       mask = sigmask(signum);
-       prop = sigprop[signum];
 
-#if SIGNAL_DEBUG
-        if(rdebug_proc && (p == rdebug_proc)) {
-                ram_printf(3);
-        }
-#endif /* SIGNAL_DEBUG */
+       if (p->p_pid <= 1 || p->p_flag & P_SYSTEM ||
+               (!posix && p == cp))
+               return(0);
+       else
+               return(1);
+}
 
-       if ((new_task == TASK_NULL) || (thr_act == (thread_t)NULL)  || is_kerneltask(new_task))
-               return;
 
+static int
+killpg1_pgrpfilt(proc_t p, __unused void * arg)
+{
+        if (p->p_pid <= 1 || p->p_flag & P_SYSTEM ||
+                (p->p_stat == SZOMB))
+                return(0);
+        else
+                return(1);
+}
 
-       uth = get_bsdthread_info(thr_act);
-       signal_lock(p);
 
-       /*
-        * proc is traced, always give parent a chance.
-        */
-       action = SIG_DFL;
 
-       if (p->p_nice > NZERO && action == SIG_DFL && (prop & SA_KILL) &&
-               (p->p_flag & P_TRACED) == 0)
-               p->p_nice = NZERO;
+static int
+killpg1_callback(proc_t p, void * arg)
+{
+       struct killpg1_iterargs * kargp = (struct killpg1_iterargs *)arg;
+        proc_t cp = kargp->cp;
+        kauth_cred_t uc = kargp->uc;   /* refcounted by the caller safe to use internal fields */
+        int signum = kargp->signum;
+       int * nfoundp = kargp->nfoundp;
+       int n;
+       int zombie = 0;
+       int error = 0;
 
-       if (prop & SA_CONT) {
-               p->p_siglist &= ~stopsigmask;
-               uth->uu_siglist &= ~stopsigmask;
-       }
+       if ((kargp->zombie != 0) && ((p->p_listflag & P_LIST_EXITED) == P_LIST_EXITED))
+               zombie = 1;
 
-       if (prop & SA_STOP) {
-               /*
-                * If sending a tty stop signal to a member of an orphaned
-                * process group, discard the signal here if the action
-                * is default; don't stop the process below if sleeping,
-                * and don't clear any pending SIGCONT.
-                */
-               if (prop & SA_TTYSTOP && p->p_pgrp->pg_jobc == 0 &&
-                       action == SIG_DFL)
-                       goto psigout;
-               uth->uu_siglist &= ~contsigmask;
-               p->p_siglist &= ~contsigmask;
-       }
-       uth->uu_siglist |= mask;
-       p->p_siglist |= mask;   /* just for lame ones looking here */
+       if (zombie != 0) {
+               proc_list_lock();
+               error = cansignal(cp, uc, p, signum, zombie);
+               proc_list_unlock();
        
-       /* Deliver signal to the activation passed in */
-       act_set_astbsd(thr_act);
-
-       /*
-        *      SIGKILL priority twiddling moved here from above because
-        *      it needs sig_thread.  Could merge it into large switch
-        *      below if we didn't care about priority for tracing
-        *      as SIGKILL's action is always SIG_DFL.
-        */
-       if ((signum == SIGKILL) && (p->p_nice > NZERO)) {
-               p->p_nice = NZERO;
-       }
+               if (error != 0 && nfoundp != NULL) {
+                       n = *nfoundp;
+                       *nfoundp = n+1;
+               }
+       } else {
+               if (cansignal(cp, uc, p, signum, 0) == 0)
+                       return(PROC_RETURNED);
 
-       /*
-        *      This Process is traced - wake it up (if not already
-        *      stopped) so that it can discover the signal in
-        *      issig() and stop for the parent.
-        */
-         if (p->p_flag & P_TRACED) {
-               if (p->p_stat != SSTOP)
-                       goto run;
-               else
-                       goto psigout;
-       }
-run:
-       /*
-        * If we're being traced (possibly because someone attached us
-        * while we were stopped), check for a signal from the debugger.
-        */
-       if (p->p_stat == SSTOP) {
-               if ((p->p_flag & P_TRACED) != 0 && p->p_xstat != 0) {
-                       uth->uu_siglist |= sigmask(p->p_xstat); 
-                       p->p_siglist |= mask;   /* just for lame ones looking here */
+               if (nfoundp != NULL) {
+                       n = *nfoundp;
+                       *nfoundp = n+1;
                }
+               if (signum != 0)
+                       psignal(p, signum);
        }
 
-       /*
-        * setrunnable(p) in BSD
-        */
-       p->p_stat = SRUN;
-
-psigout:
-       signal_unlock(p);
+       return(PROC_RETURNED);
 }
 
-static thread_t
-get_signalthread(struct proc *p, int signum)
+/*
+ * Common code for kill process group/broadcast kill.
+ * cp is calling process.
+ */
+int
+killpg1(proc_t cp, int signum, int pgid, int all, int posix)
 {
-       struct uthread *uth;
-       thread_t thr_act;
-       sigset_t mask = sigmask(signum);
-       thread_t sig_thread_act;
-       struct task * sig_task = p->task;
-       kern_return_t kret;
+       kauth_cred_t uc;
+       struct pgrp *pgrp;
+       int nfound = 0;
+       struct killpg1_iterargs karg;
+       struct killpg1_filtargs kfarg;
+       int error = 0;
        
-       if ((p->p_flag & P_INVFORK) && p->p_vforkact) {
-               sig_thread_act = p->p_vforkact; 
-               kret = check_actforsig(sig_task, sig_thread_act, 1);
-               if (kret == KERN_SUCCESS) 
-                       return(sig_thread_act);
-               else
-                       return(THREAD_NULL);
-       } 
+       uc = kauth_cred_proc_ref(cp);
+       if (all) {
+               /* 
+                * broadcast 
+                */
+               kfarg.posix = posix;
+               kfarg.cp = cp;
 
-       TAILQ_FOREACH(uth, &p->p_uthlist, uu_list) {
-               if(((uth->uu_flag & UT_NO_SIGMASK)== 0) && 
-                       (((uth->uu_sigmask & mask) == 0) || (uth->uu_sigwait & mask))) {
-                       if (check_actforsig(p->task, uth->uu_act, 1) == KERN_SUCCESS)
-                               return(uth->uu_act);
+               karg.cp = cp;
+               karg.uc = uc;
+               karg.nfoundp = &nfound;
+               karg.signum = signum;
+               karg.zombie = 1;
+
+               proc_iterate((PROC_ALLPROCLIST | PROC_ZOMBPROCLIST), killpg1_callback, &karg, killpg1_filt, (void *)&kfarg);
+
+       } else {
+               if (pgid == 0) {
+                       /* 
+                        * zero pgid means send to my process group.
+                        */
+                       pgrp = proc_pgrp(cp);
+                } else {
+                       pgrp = pgfind(pgid);
+                       if (pgrp == NULL) {
+                               error = ESRCH;
+                               goto out;
+                       }
                }
+
+                karg.nfoundp = &nfound;
+                karg.uc = uc;
+                karg.signum = signum;
+               karg.cp = cp;
+               karg.zombie = 0;
+
+
+               /* PGRP_DROPREF drops the pgrp refernce */
+               pgrp_iterate(pgrp, PGRP_DROPREF, killpg1_callback, &karg,
+                       killpg1_pgrpfilt, NULL);
        }
-       if (get_signalact(p->task, &thr_act, 1) == KERN_SUCCESS) {
-               return(thr_act);
+       error =  (nfound ? 0 : (posix ? EPERM : ESRCH));
+out:
+       kauth_cred_unref(&uc);
+       return (error);
+}
+
+
+/*
+ * Send a signal to a process group.
+ */
+void
+gsignal(int pgid, int signum)
+{
+       struct pgrp *pgrp;
+
+       if (pgid && (pgrp = pgfind(pgid))) {
+               pgsignal(pgrp, signum, 0);
+               pg_rele(pgrp);
        }
+}
+
+/*
+ * Send a signal to a process group.  If checkctty is 1,
+ * limit to members which have a controlling terminal.
+ */
 
-       return(THREAD_NULL);
+static int
+pgsignal_filt(proc_t p, void * arg)
+{
+       int checkctty = *(int*)arg;
+
+       if ((checkctty == 0) || p->p_flag & P_CONTROLT)
+               return(1);
+       else 
+               return(0);
+}
+
+
+static int
+pgsignal_callback(proc_t p, void * arg)
+{
+        int  signum = *(int*)arg;
+
+       psignal(p, signum);
+       return(PROC_RETURNED);
 }
 
+
+void
+pgsignal(struct pgrp *pgrp, int signum, int checkctty)
+{
+       if (pgrp != PGRP_NULL) {
+               pgrp_iterate(pgrp, 0, pgsignal_callback, &signum, pgsignal_filt, &checkctty);
+       }
+}
+
+
+void
+tty_pgsignal(struct tty *tp, int signum, int checkctty)
+{
+       struct pgrp * pg;
+
+       pg = tty_pgrp(tp);
+       if (pg != PGRP_NULL) {
+               pgrp_iterate(pg, 0, pgsignal_callback, &signum, pgsignal_filt, &checkctty);
+               pg_rele(pg);
+       }
+}
 /*
- * Send the signal to the process.  If the signal has an action, the action
- * is usually performed by the target process rather than the caller; we add
- * the signal to the set of pending signals for the process.
- *
- * Exceptions:
- *   o When a stop signal is sent to a sleeping process that takes the
- *     default action, the process is stopped without awakening it.
- *   o SIGCONT restarts stopped processes (or puts them back to sleep)
- *     regardless of the signal action (eg, blocked or ignored).
- *
- * Other ignored signals are discarded immediately.
+ * Send a signal caused by a trap to a specific thread.
  */
 void
-psignal_lock(p, signum, withlock)
-       register struct proc *p;
-       register int signum;
-       register int withlock;
-{
-       register int s, prop;
-       register sig_t action;
-       thread_t        sig_thread_act;
-       register task_t         sig_task;
-       int mask;
+threadsignal(thread_t sig_actthread, int signum, mach_exception_code_t code, boolean_t set_exitreason)
+{
        struct uthread *uth;
-       boolean_t funnel_state = FALSE;
-       int sw_funnel = 0;
+       struct task * sig_task;
+       proc_t p;
+       int mask;
 
        if ((u_int)signum >= NSIG || signum == 0)
-               panic("psignal signal number");
-       mask = sigmask(signum);
-       prop = sigprop[signum];
-
-#if SIGNAL_DEBUG
-        if(rdebug_proc && (p == rdebug_proc)) {
-                ram_printf(3);
-        }
-#endif /* SIGNAL_DEBUG */
+               return;
 
-       if (thread_funnel_get() == (funnel_t *)0) {
-               sw_funnel = 1;
-               funnel_state = thread_funnel_set(kernel_flock, TRUE);
-       }
-       /*
-        *      We will need the task pointer later.  Grab it now to
-        *      check for a zombie process.  Also don't send signals
-        *      to kernel internal tasks.
-        */
-       if (((sig_task = p->task) == TASK_NULL)  || is_kerneltask(sig_task)) {
-               if (sw_funnel)
-                       thread_funnel_set(kernel_flock, funnel_state);
+       mask = sigmask(signum);
+       if ((mask & threadmask) == 0)
                return;
-       }
+       sig_task = get_threadtask(sig_actthread);
+       p = (proc_t)(get_bsdtask_info(sig_task));
 
-        s = splhigh();
-        KNOTE(&p->p_klist, NOTE_SIGNAL | signum);
-        splx(s);
+       uth = get_bsdthread_info(sig_actthread);
+       if (uth->uu_flag & UT_VFORK)
+               p = uth->uu_proc;
 
-       /*
-        * do not send signals to the process that has the thread
-        * doing a reboot(). Not doing so will mark that thread aborted
-        * and can cause IO failures wich will cause data loss.
-        */
-       if (ISSET(p->p_flag, P_REBOOT)) {
-               if (sw_funnel)
-                       thread_funnel_set(kernel_flock, funnel_state);
+       proc_lock(p);
+       if (!(p->p_lflag & P_LTRACED) && (p->p_sigignore & mask)) {
+               proc_unlock(p);
                return;
        }
 
-       if (withlock)
-               signal_lock(p);
+       uth->uu_siglist |= mask;
+       uth->uu_code = code;
 
-       /*
-        *      Deliver the signal to the first thread in the task. This
-        *      allows single threaded applications which use signals to
-        *      be able to be linked with multithreaded libraries.  We have
-        *      an implicit reference to the current thread, but need
-        *      an explicit one otherwise.  The thread reference keeps
-        *      the corresponding task data structures around too.  This
-        *      reference is released by thread_deallocate.
-        */
-       
-       if (((p->p_flag & P_TRACED) == 0) && (p->p_sigignore & mask))
-               goto psigout;
+       /* Attempt to establish whether the signal will be fatal (mirrors logic in psignal_internal()) */
+       if (set_exitreason && ((p->p_lflag & P_LTRACED) || (!(uth->uu_sigwait & mask)
+               && !(uth->uu_sigmask & mask) && !(p->p_sigcatch & mask))) &&
+               !(mask & stopsigmask) && !(mask & contsigmask)) {
 
-       /* If successful return with ast set */
-       sig_thread_act = get_signalthread(p, signum);
+               if (uth->uu_exit_reason == OS_REASON_NULL) {
+                       KERNEL_DEBUG_CONSTANT(BSDDBG_CODE(DBG_BSD_PROC, BSD_PROC_EXITREASON_CREATE) | DBG_FUNC_NONE,
+                                                               p->p_pid, OS_REASON_SIGNAL, signum, 0, 0);
 
-       if (sig_thread_act == THREAD_NULL) {
-               /* XXXX FIXME
-                * if it is sigkill, may be we should
-                * inject a thread to terminate
-                */
-#if SIGNAL_DEBUG
-                       ram_printf(1);
-#endif /* SIGNAL_DEBUG */
-               goto psigout;
-       }
+                       os_reason_t signal_reason = build_signal_reason(signum, "exc handler");
 
-       uth = get_bsdthread_info(sig_thread_act);
+                       set_thread_exit_reason(sig_actthread, signal_reason, TRUE);
 
-       /*
-        * If proc is traced, always give parent a chance.
-        */
-       if (p->p_flag & P_TRACED)
-               action = SIG_DFL;
-       else {
-               /*
-                * If the signal is being ignored,
-                * then we forget about it immediately.
-                * (Note: we don't set SIGCONT in p_sigignore,
-                * and if it is set to SIG_IGN,
-                * action will be SIG_DFL here.)
-                */
-               if (p->p_sigignore & mask)
-                       goto psigout;
-               /* sigwait takes precedence */
-               if (uth->uu_sigwait & mask)
-                       action = KERN_SIG_WAIT;
-               else if (uth->uu_sigmask & mask)
-                       action = KERN_SIG_HOLD;
-               else if (p->p_sigcatch & mask)
-                       action = KERN_SIG_CATCH;
-               else
-                       action = SIG_DFL;
+                       /* We dropped/consumed the reference in set_thread_exit_reason() */
+                       signal_reason = OS_REASON_NULL;
+               }
        }
 
-       if (p->p_nice > NZERO && action == SIG_DFL && (prop & SA_KILL) &&
-               (p->p_flag & P_TRACED) == 0)
-               p->p_nice = NZERO;
+       proc_unlock(p);
 
-       if (prop & SA_CONT) {
-               uth->uu_siglist &= ~stopsigmask;
-               p->p_siglist &= ~stopsigmask;
-       }
-
-       if (prop & SA_STOP) {
-               /*
-                * If sending a tty stop signal to a member of an orphaned
-                * process group, discard the signal here if the action
-                * is default; don't stop the process below if sleeping,
-                * and don't clear any pending SIGCONT.
-                */
-               if (prop & SA_TTYSTOP && p->p_pgrp->pg_jobc == 0 &&
-                       action == SIG_DFL)
-                       goto psigout;
-               uth->uu_siglist &= ~contsigmask;
-               p->p_siglist &= ~contsigmask;
-       }
-       uth->uu_siglist |= mask;
-       p->p_siglist |= mask;   /* just for lame ones looking here */
+       /* mark on process as well */
+       signal_setast(sig_actthread);
+}
 
-       
-       /*
-        * Defer further processing for signals which are held,
-        * except that stopped processes must be continued by SIGCONT.
-        */
-       if (action == KERN_SIG_HOLD && ((prop & SA_CONT) == 0 || p->p_stat != SSTOP)) {
-               goto psigout;
-       }
-       /*
-        *      SIGKILL priority twiddling moved here from above because
-        *      it needs sig_thread.  Could merge it into large switch
-        *      below if we didn't care about priority for tracing
-        *      as SIGKILL's action is always SIG_DFL.
-        */
-       if ((signum == SIGKILL) && (p->p_nice > NZERO)) {
-               p->p_nice = NZERO;
-       }
+void
+set_thread_exit_reason(void *th, void *reason, boolean_t proc_locked)
+{
+       struct uthread *targ_uth = get_bsdthread_info(th);
+       struct task *targ_task = NULL;
+       proc_t targ_proc = NULL;
 
-       /*
-        *      Process is traced - wake it up (if not already
-        *      stopped) so that it can discover the signal in
-        *      issig() and stop for the parent.
-        */
-       if (p->p_flag & P_TRACED) {
-               if (p->p_stat != SSTOP)
-                       goto run;
-               else
-                       goto psigout;
-       }
+       os_reason_t exit_reason = (os_reason_t)reason;
 
-       if (action == KERN_SIG_WAIT) {
-               uth->uu_sigwait = mask;
-               uth->uu_siglist &= ~mask;
-               p->p_siglist &= ~mask;
-               wakeup(&uth->uu_sigwait);
-               /* if it is SIGCONT resume whole process */
-               if (prop & SA_CONT) {
-                       p->p_flag |= P_CONTINUED;
-                       (void) task_resume(sig_task);
-               }
-               goto psigout;
+       if (exit_reason == OS_REASON_NULL)
+               return;
+
+       if (!proc_locked) {
+               targ_task = get_threadtask(th);
+               targ_proc = (proc_t)(get_bsdtask_info(targ_task));
+
+               proc_lock(targ_proc);
        }
 
-       if (action != SIG_DFL) {
-               /*
-                *      User wants to catch the signal.
-                *      Wake up the thread, but don't un-suspend it
-                *      (except for SIGCONT).
-                */
-               if (prop & SA_CONT) {
-                       if (p->p_flag & P_TTYSLEEP) {
-                               p->p_flag &= ~P_TTYSLEEP;
-                               wakeup(&p->p_siglist);
-                       } else {
-                               p->p_flag |= P_CONTINUED;
-                               (void) task_resume(sig_task);
-                       }
-                       p->p_stat = SRUN;
-               } else if (p->p_stat == SSTOP)
-                       goto psigout;
-               goto run;
+       if (targ_uth->uu_exit_reason == OS_REASON_NULL) {
+               targ_uth->uu_exit_reason = exit_reason;
        } else {
-               /*      Default action - varies */
-               if (mask & stopsigmask) {
-                       /*
-                        * These are the signals which by default
-                        * stop a process.
-                        *
-                        * Don't clog system with children of init
-                        * stopped from the keyboard.
-                        */
-                       if (!(prop & SA_STOP) && p->p_pptr == initproc) {
-                               psignal_lock(p, SIGKILL, 0);
-                               uth->uu_siglist &= ~mask;
-                               p->p_siglist &= ~mask;
-                               goto psigout;
-                       }
-                        
-                       /*
-                        *      Stop the task
-                        *      if task hasn't already been stopped by
-                        *      a signal.
-                        */
-                       uth->uu_siglist &= ~mask;
-                       p->p_siglist &= ~mask;
-                       if (p->p_stat != SSTOP) {
-                               p->p_xstat = signum;
-                               stop(p);
-                               if ((p->p_pptr->p_flag & P_NOCLDSTOP) == 0) {
-                                       struct proc *pp = p->p_pptr;
+               /* The caller expects that we drop a reference on the exit reason */
+               os_reason_free(exit_reason);
+       }
 
-                                       pp->si_pid = p->p_pid;
-                                       pp->si_status = p->p_xstat;
-                                       pp->si_code = CLD_STOPPED;
-                                       pp->si_uid = p->p_ucred->cr_ruid;
-                                       psignal(pp, SIGCHLD);
-                               }
+       if (!proc_locked) {
+               assert(targ_proc != NULL);
+               proc_unlock(targ_proc);
+       }
+}
+
+/*
+ * get_signalthread
+ *
+ * Picks an appropriate thread from a process to target with a signal.
+ *
+ * Called with proc locked.
+ * Returns thread with BSD ast set.
+ *
+ * We attempt to deliver a proc-wide signal to the first thread in the task.
+ * This allows single threaded applications which use signals to
+ * be able to be linked with multithreaded libraries.
+ */
+static kern_return_t
+get_signalthread(proc_t p, int signum, thread_t * thr)
+{
+       struct uthread *uth;
+       sigset_t mask = sigmask(signum);
+       thread_t sig_thread;
+       struct task * sig_task = p->task;
+       kern_return_t kret;
+
+       *thr = THREAD_NULL;
+
+       if ((p->p_lflag & P_LINVFORK) && p->p_vforkact) {
+               sig_thread = p->p_vforkact;     
+               kret = check_actforsig(sig_task, sig_thread, 1);
+               if (kret == KERN_SUCCESS)  {
+                       *thr = sig_thread;
+                       return(KERN_SUCCESS);
+               }else
+                       return(KERN_FAILURE);
+       }
+
+       TAILQ_FOREACH(uth, &p->p_uthlist, uu_list) {
+               if(((uth->uu_flag & UT_NO_SIGMASK)== 0) && 
+                       (((uth->uu_sigmask & mask) == 0) || (uth->uu_sigwait & mask))) {
+                       if (check_actforsig(p->task, uth->uu_context.vc_thread, 1) == KERN_SUCCESS) {
+                               *thr = uth->uu_context.vc_thread;
+                               return(KERN_SUCCESS);
                        }
-                       goto psigout;
                }
+       }
+       if (get_signalact(p->task, thr, 1) == KERN_SUCCESS) {
+               return(KERN_SUCCESS);
+       }
 
-               switch (signum) {
-                       /*
-                        * Signals ignored by default have been dealt
-                        * with already, since their bits are on in
-                        * p_sigignore.
-                        */
-
-               case SIGKILL:
-                       /*
-                        * Kill signal always sets process running and
-                        * unsuspends it.
-                        */
-                       /*
-                        *      Process will be running after 'run'
-                        */
-                       p->p_stat = SRUN;
+       return(KERN_FAILURE);
+}
 
-                       thread_abort(sig_thread_act);
+static os_reason_t
+build_signal_reason(int signum, const char *procname)
+{
+       os_reason_t signal_reason = OS_REASON_NULL;
+       proc_t sender_proc = current_proc();
+       uint32_t reason_buffer_size_estimate = 0, proc_name_length = 0;
+       const char *default_sender_procname = "unknown";
+       mach_vm_address_t data_addr;
+       int ret;
 
-                       goto psigout;
+       signal_reason = os_reason_create(OS_REASON_SIGNAL, signum);
+       if (signal_reason == OS_REASON_NULL) {
+               printf("build_signal_reason: unable to allocate signal reason structure.\n");
+               return signal_reason;
+       }
 
-               case SIGCONT:
-                       /*
-                        * Let the process run.  If it's sleeping on an
-                        * event, it remains so.
-                        */
-                       if (p->p_flag & P_TTYSLEEP) {
-                               p->p_flag &= ~P_TTYSLEEP;
-                               wakeup(&p->p_siglist);
-                       } else {
-                               p->p_flag |= P_CONTINUED;
-                               (void) task_resume(sig_task);
-                       }
-                       uth->uu_siglist &= ~mask;
-                       p->p_siglist &= ~mask;
-                       p->p_stat = SRUN;
+       reason_buffer_size_estimate = kcdata_estimate_required_buffer_size(2, sizeof(sender_proc->p_name) +
+                                                                               sizeof(sender_proc->p_pid));
 
-                       goto psigout;
+       ret = os_reason_alloc_buffer_noblock(signal_reason, reason_buffer_size_estimate);
+       if (ret != 0) {
+               printf("build_signal_reason: unable to allocate signal reason buffer.\n");
+               return signal_reason;
+       }
 
-               default:
-                       /*
-                        * All other signals wake up the process, but don't
-                        * resume it.
-                        */
-                       if (p->p_stat == SSTOP)
-                               goto psigout;
-                       goto run;
+       if (KERN_SUCCESS == kcdata_get_memory_addr(&signal_reason->osr_kcd_descriptor, KCDATA_TYPE_PID,
+                                                       sizeof(sender_proc->p_pid), &data_addr)) {
+               kcdata_memcpy(&signal_reason->osr_kcd_descriptor, data_addr, &sender_proc->p_pid,
+                                       sizeof(sender_proc->p_pid));
+       } else {
+               printf("build_signal_reason: exceeded space in signal reason buf, unable to log PID\n");
+       }
+
+       proc_name_length = sizeof(sender_proc->p_name);
+       if (KERN_SUCCESS == kcdata_get_memory_addr(&signal_reason->osr_kcd_descriptor, KCDATA_TYPE_PROCNAME,
+                                                       proc_name_length, &data_addr)) {
+               if (procname) {
+                       char truncated_procname[proc_name_length];
+                       strncpy((char *) &truncated_procname, procname, proc_name_length);
+                       truncated_procname[proc_name_length - 1] = '\0';
+
+                       kcdata_memcpy(&signal_reason->osr_kcd_descriptor, data_addr, truncated_procname,
+                                       strlen((char *) &truncated_procname));
+               } else if (*sender_proc->p_name) {
+                       kcdata_memcpy(&signal_reason->osr_kcd_descriptor, data_addr, &sender_proc->p_name,
+                                       sizeof(sender_proc->p_name));
+               } else {
+                       kcdata_memcpy(&signal_reason->osr_kcd_descriptor, data_addr, &default_sender_procname,
+                                       strlen(default_sender_procname) + 1);
                }
-       }
-       /*NOTREACHED*/
-run:
-       /*
-        * If we're being traced (possibly because someone attached us
-        * while we were stopped), check for a signal from the debugger.
-        */
-       if (p->p_stat == SSTOP) {
-               if ((p->p_flag & P_TRACED) != 0 && p->p_xstat != 0)
-                       uth->uu_siglist |= sigmask(p->p_xstat); 
        } else {
-               /*
-                * setrunnable(p) in BSD and
-                * Wake up the thread if it is interruptible.
-                */
-               p->p_stat = SRUN;
-               thread_abort_safely(sig_thread_act);
+               printf("build_signal_reason: exceeded space in signal reason buf, unable to log procname\n");
        }
-psigout:
-       if (withlock) 
-               signal_unlock(p);
-       if (sw_funnel)
-                       thread_funnel_set(kernel_flock, funnel_state);
-}
 
+       return signal_reason;
+}
 
-/* psignal_lock(p, signum, withlock ) */
-void
-psignal_uthread(thr_act, signum)
-       thread_t thr_act;
-       int signum;
+/*
+ * Send the signal to the process.  If the signal has an action, the action
+ * is usually performed by the target process rather than the caller; we add
+ * the signal to the set of pending signals for the process.
+ *
+ * Always drops a reference on a signal_reason if one is provided, whether via
+ * passing it to a thread or deallocating directly.
+ *
+ * Exceptions:
+ *   o When a stop signal is sent to a sleeping process that takes the
+ *     default action, the process is stopped without awakening it.
+ *   o SIGCONT restarts stopped processes (or puts them back to sleep)
+ *     regardless of the signal action (eg, blocked or ignored).
+ *
+ * Other ignored signals are discarded immediately.
+ */
+static void
+psignal_internal(proc_t p, task_t task, thread_t thread, int flavor, int signum, os_reason_t signal_reason)
 {
-       struct proc *p;
-       register int prop;
-       register sig_t action;
-       thread_t        sig_thread_act;
-       register task_t         sig_task;
-       int mask;
-       struct uthread *uth;
-       kern_return_t kret;
-       int error = 0;
+       int prop;
+       user_addr_t action = USER_ADDR_NULL;
+       proc_t                  sig_proc;
+       thread_t                sig_thread;
+       task_t                  sig_task;
+       int                     mask;
+       struct uthread          *uth;
+       kern_return_t           kret;
+       uid_t                   r_uid;
+       proc_t                  pp;
+       kauth_cred_t            my_cred;
+       char                    *launchd_exit_reason_desc = NULL;
+       boolean_t               update_thread_policy = FALSE;
 
-       p = (struct proc *)get_bsdtask_info(get_threadtask(thr_act));
        if ((u_int)signum >= NSIG || signum == 0)
-               panic("Invalid signal number in psignal_uthread"); 
+               panic("psignal: bad signal number %d", signum);
+
        mask = sigmask(signum);
        prop = sigprop[signum];
 
 #if SIGNAL_DEBUG
-        if(rdebug_proc && (p == rdebug_proc)) {
+        if(rdebug_proc && (p != PROC_NULL) && (p == rdebug_proc)) {
                 ram_printf(3);
         }
 #endif /* SIGNAL_DEBUG */
 
+       /* catch unexpected initproc kills early for easier debuggging */
+       if (signum == SIGKILL && p == initproc) {
+               if (signal_reason == NULL) {
+                       panic_plain("unexpected SIGKILL of %s %s (no reason provided)",
+                           (p->p_name[0] != '\0' ? p->p_name : "initproc"),
+                           ((p->p_csflags & CS_KILLED) ? "(CS_KILLED)" : ""));
+               } else {
+                       launchd_exit_reason_desc = launchd_exit_reason_get_string_desc(signal_reason);
+                       panic_plain("unexpected SIGKILL of %s %s with reason -- namespace %d code 0x%llx description %." LAUNCHD_PANIC_REASON_STRING_MAXLEN "s",
+                           (p->p_name[0] != '\0' ? p->p_name : "initproc"),
+                           ((p->p_csflags & CS_KILLED) ? "(CS_KILLED)" : ""),
+                           signal_reason->osr_namespace, signal_reason->osr_code,
+                           launchd_exit_reason_desc ? launchd_exit_reason_desc : "none");
+               }
+       }
+
        /*
         *      We will need the task pointer later.  Grab it now to
         *      check for a zombie process.  Also don't send signals
         *      to kernel internal tasks.
         */
-       if (((sig_task = p->task) == TASK_NULL)  || is_kerneltask(sig_task)) {
+       if (flavor & PSIG_VFORK) {
+               sig_task = task;
+               sig_thread = thread;
+               sig_proc = p;
+       } else if (flavor & PSIG_THREAD) {
+               sig_task = get_threadtask(thread);
+               sig_thread = thread;
+               sig_proc = (proc_t)get_bsdtask_info(sig_task);
+       } else if (flavor & PSIG_TRY_THREAD) {
+               assert((thread == current_thread()) && (p == current_proc()));
+               sig_task = p->task;
+               sig_thread = thread;
+               sig_proc = p;
+       } else {
+               sig_task = p->task;
+               sig_thread = THREAD_NULL;
+               sig_proc = p;
+       }
+
+       if ((sig_task == TASK_NULL) || is_kerneltask(sig_task)) {
+               os_reason_free(signal_reason);
                return;
        }
 
-       sig_thread_act = thr_act;
        /*
         * do not send signals to the process that has the thread
         * doing a reboot(). Not doing so will mark that thread aborted
-        * and can cause IO failures wich will cause data loss.
+        * and can cause IO failures wich will cause data loss.  There's
+        * also no need to send a signal to a process that is in the middle
+        * of being torn down.
         */
-       if (ISSET(p->p_flag, P_REBOOT)) {
+       if (ISSET(sig_proc->p_flag, P_REBOOT) || ISSET(sig_proc->p_lflag, P_LEXIT)) {
+               DTRACE_PROC3(signal__discard, thread_t, sig_thread, proc_t, sig_proc, int, signum);
+               os_reason_free(signal_reason);
                return;
        }
 
-       signal_lock(p);
+       if( (flavor & (PSIG_VFORK | PSIG_THREAD)) == 0) {
+               proc_knote(sig_proc, NOTE_SIGNAL | signum);
+       }
+
+       if ((flavor & PSIG_LOCKED)== 0)
+               proc_signalstart(sig_proc, 0);
+
+       /* Don't send signals to a process that has ignored them. */
+       if (((flavor & PSIG_VFORK) == 0) && ((sig_proc->p_lflag & P_LTRACED) == 0) && (sig_proc->p_sigignore & mask)) {
+               DTRACE_PROC3(signal__discard, thread_t, sig_thread, proc_t, sig_proc, int, signum);
+               goto sigout_unlocked;
+       }
 
        /*
-        *      Deliver the signal to the first thread in the task. This
-        *      allows single threaded applications which use signals to
-        *      be able to be linked with multithreaded libraries.  We have
-        *      an implicit reference to the current thread, but need
-        *      an explicit one otherwise.  The thread reference keeps
-        *      the corresponding task data structures around too.  This
-        *      reference is released by thread_deallocate.
+        * The proc_lock prevents the targeted thread from being deallocated
+        * or handling the signal until we're done signaling it.
+        *
+        * Once the proc_lock is dropped, we have no guarantee the thread or uthread exists anymore.
+        *
+        * XXX: What if the thread goes inactive after the thread passes bsd ast point?
         */
-       
-       if (((p->p_flag & P_TRACED) == 0) && (p->p_sigignore & mask))
-               goto puthout;
+       proc_lock(sig_proc);
 
-       kret = check_actforsig(sig_task, sig_thread_act, 1);
+       if (flavor & PSIG_VFORK) {
+               action = SIG_DFL;
+               act_set_astbsd(sig_thread);
+               kret = KERN_SUCCESS;
+       } else if (flavor & PSIG_TRY_THREAD) {
+               uth = get_bsdthread_info(sig_thread);
+               if (((uth->uu_flag & UT_NO_SIGMASK) == 0) &&
+                               (((uth->uu_sigmask & mask) == 0) || (uth->uu_sigwait & mask)) &&
+                               ((kret = check_actforsig(sig_proc->task, sig_thread, 1)) == KERN_SUCCESS)) {
+                       /* deliver to specified thread */
+               } else {
+                       /* deliver to any willing thread */
+                       kret = get_signalthread(sig_proc, signum, &sig_thread);
+               }
+       } else if (flavor & PSIG_THREAD) {
+               /* If successful return with ast set */
+               kret = check_actforsig(sig_task, sig_thread, 1);
+       } else {
+               /* If successful return with ast set */
+               kret = get_signalthread(sig_proc, signum, &sig_thread);
+       }
 
        if (kret != KERN_SUCCESS) {
-               error = EINVAL;
-               goto puthout;
+               DTRACE_PROC3(signal__discard, thread_t, sig_thread, proc_t, sig_proc, int, signum);
+               proc_unlock(sig_proc);
+               goto sigout_unlocked;
        }
 
-
-       uth = get_bsdthread_info(sig_thread_act);
+       uth = get_bsdthread_info(sig_thread);
 
        /*
         * If proc is traced, always give parent a chance.
         */
-       if (p->p_flag & P_TRACED)
-               action = SIG_DFL;
-       else {
-               /*
-                * If the signal is being ignored,
-                * then we forget about it immediately.
-                * (Note: we don't set SIGCONT in p_sigignore,
-                * and if it is set to SIG_IGN,
-                * action will be SIG_DFL here.)
-                */
-               if (p->p_sigignore & mask)
-                       goto puthout;
-               /* sigwait takes precedence */
-               if (uth->uu_sigwait & mask)
-                       action = KERN_SIG_WAIT;
-               else if (uth->uu_sigmask & mask)
-                       action = KERN_SIG_HOLD;
-               else if (p->p_sigcatch & mask)
-                       action = KERN_SIG_CATCH;
-               else
+
+       if ((flavor & PSIG_VFORK) == 0) {
+               if (sig_proc->p_lflag & P_LTRACED)
                        action = SIG_DFL;
+               else {
+                       /*
+                        * If the signal is being ignored,
+                        * then we forget about it immediately.
+                        * (Note: we don't set SIGCONT in p_sigignore,
+                        * and if it is set to SIG_IGN,
+                        * action will be SIG_DFL here.)
+                        */
+                       if (sig_proc->p_sigignore & mask)
+                               goto sigout_locked;
+
+                       if (uth->uu_sigwait & mask)
+                               action = KERN_SIG_WAIT;
+                       else if (uth->uu_sigmask & mask)
+                               action = KERN_SIG_HOLD;
+                       else if (sig_proc->p_sigcatch & mask)
+                               action = KERN_SIG_CATCH;
+                       else
+                               action = SIG_DFL;
+               }
        }
 
-       if (p->p_nice > NZERO && action == SIG_DFL && (prop & SA_KILL) &&
-               (p->p_flag & P_TRACED) == 0)
-               p->p_nice = NZERO;
+       /* TODO: p_nice isn't hooked up to the scheduler... */
+       if (sig_proc->p_nice > NZERO && action == SIG_DFL && (prop & SA_KILL) &&
+               (sig_proc->p_lflag & P_LTRACED) == 0)
+                       sig_proc->p_nice = NZERO;
 
-       if (prop & SA_CONT) {
+       if (prop & SA_CONT)
                uth->uu_siglist &= ~stopsigmask;
-               p->p_siglist &= ~stopsigmask;
-       }
 
        if (prop & SA_STOP) {
+               struct pgrp *pg;
                /*
                 * If sending a tty stop signal to a member of an orphaned
                 * process group, discard the signal here if the action
                 * is default; don't stop the process below if sleeping,
                 * and don't clear any pending SIGCONT.
                 */
-               if (prop & SA_TTYSTOP && p->p_pgrp->pg_jobc == 0 &&
-                       action == SIG_DFL)
-                       goto puthout;
+               pg = proc_pgrp(sig_proc);
+               if (prop & SA_TTYSTOP && pg->pg_jobc == 0 &&
+                       action == SIG_DFL) {
+                       pg_rele(pg);
+                       goto sigout_locked;
+               }
+               pg_rele(pg);
                uth->uu_siglist &= ~contsigmask;
-               p->p_siglist &= ~contsigmask;
        }
+
        uth->uu_siglist |= mask;
-       p->p_siglist |= mask;   /* just for lame ones looking here */
 
        /*
         * Defer further processing for signals which are held,
         * except that stopped processes must be continued by SIGCONT.
         */
-       if (action == KERN_SIG_HOLD && ((prop & SA_CONT) == 0 || p->p_stat != SSTOP))
-               goto puthout;
-               
+       /* vfork will not go thru as action is SIG_DFL */
+       if ((action == KERN_SIG_HOLD) && ((prop & SA_CONT) == 0 || sig_proc->p_stat != SSTOP))
+               goto sigout_locked;
+
        /*
         *      SIGKILL priority twiddling moved here from above because
         *      it needs sig_thread.  Could merge it into large switch
         *      below if we didn't care about priority for tracing
         *      as SIGKILL's action is always SIG_DFL.
+        *
+        *      TODO: p_nice isn't hooked up to the scheduler...
         */
-       if ((signum == SIGKILL) && (p->p_nice > NZERO)) {
-               p->p_nice = NZERO;
+       if ((signum == SIGKILL) && (sig_proc->p_nice > NZERO)) {
+               sig_proc->p_nice = NZERO;
        }
 
        /*
@@ -1918,24 +2317,41 @@ psignal_uthread(thr_act, signum)
         *      stopped) so that it can discover the signal in
         *      issig() and stop for the parent.
         */
-       if (p->p_flag & P_TRACED) {
-               if (p->p_stat != SSTOP)
-                       goto psurun;
+       if (sig_proc->p_lflag & P_LTRACED) {
+               if (sig_proc->p_stat != SSTOP)
+                       goto runlocked;
                else
-                       goto puthout;
+                       goto sigout_locked;
        }
 
+       if ((flavor & PSIG_VFORK) != 0)
+               goto runlocked;
+
        if (action == KERN_SIG_WAIT) {
+#if CONFIG_DTRACE
+               /*
+                * DTrace proc signal-clear returns a siginfo_t. Collect the needed info.
+                */
+               r_uid = kauth_getruid(); /* per thread credential; protected by our thread context */
+
+               bzero((caddr_t)&(uth->t_dtrace_siginfo), sizeof(uth->t_dtrace_siginfo));
+
+               uth->t_dtrace_siginfo.si_signo = signum;
+               uth->t_dtrace_siginfo.si_pid = current_proc()->p_pid;
+               uth->t_dtrace_siginfo.si_status = W_EXITCODE(signum, 0);
+               uth->t_dtrace_siginfo.si_uid = r_uid;
+               uth->t_dtrace_siginfo.si_code = 0;
+#endif
                uth->uu_sigwait = mask;
                uth->uu_siglist &= ~mask;
-               p->p_siglist &= ~mask;
                wakeup(&uth->uu_sigwait);
                /* if it is SIGCONT resume whole process */
                if (prop & SA_CONT) {
-                       p->p_flag |= P_CONTINUED;
-                       (void) task_resume(sig_task);
+                       OSBitOrAtomic(P_CONTINUED, &sig_proc->p_flag);
+                       sig_proc->p_contproc = current_proc()->p_pid;
+                       (void) task_resume_internal(sig_task);
                }
-               goto puthout;
+               goto sigout_locked;
        }
 
        if (action != SIG_DFL) {
@@ -1945,13 +2361,38 @@ psignal_uthread(thr_act, signum)
                 *      (except for SIGCONT).
                 */
                if (prop & SA_CONT) {
-                       p->p_flag |= P_CONTINUED;
-                       (void) task_resume(sig_task);
+                       OSBitOrAtomic(P_CONTINUED, &sig_proc->p_flag);
+                       (void) task_resume_internal(sig_task);
+                       sig_proc->p_stat = SRUN;
+               }  else if (sig_proc->p_stat == SSTOP) {
+                       goto sigout_locked;
+               }
+               /*
+                * Fill out siginfo structure information to pass to the
+                * signalled process/thread sigaction handler, when it
+                * wakes up.  si_code is 0 because this is an ordinary
+                * signal, not a SIGCHLD, and so si_status is the signal
+                * number itself, instead of the child process exit status.
+                * We shift this left because it will be shifted right before
+                * it is passed to user space.  kind of ugly to use W_EXITCODE
+                * this way, but it beats defining a new macro.
+                *
+                * Note:        Avoid the SIGCHLD recursion case!
+                */
+               if (signum != SIGCHLD) {
+                       r_uid = kauth_getruid();
+
+                       sig_proc->si_pid = current_proc()->p_pid;
+                       sig_proc->si_status = W_EXITCODE(signum, 0);
+                       sig_proc->si_uid = r_uid;
+                       sig_proc->si_code = 0;
                }
-               goto psurun;
+
+               goto runlocked;
        } else {
                /*      Default action - varies */
                if (mask & stopsigmask) {
+                       assert(signal_reason == NULL);
                        /*
                         * These are the signals which by default
                         * stop a process.
@@ -1959,36 +2400,64 @@ psignal_uthread(thr_act, signum)
                         * Don't clog system with children of init
                         * stopped from the keyboard.
                         */
-                       if (!(prop & SA_STOP) && p->p_pptr == initproc) {
-                               psignal_lock(p, SIGKILL, 0);
+                       if (!(prop & SA_STOP) && sig_proc->p_pptr == initproc) {
                                uth->uu_siglist &= ~mask;
-                               p->p_siglist &= ~mask;
-                               goto puthout;
+                               proc_unlock(sig_proc);
+                               /* siglock still locked, proc_lock not locked */
+                               psignal_locked(sig_proc, SIGKILL);
+                               goto sigout_unlocked;
                        }
-                        
+
                        /*
                         *      Stop the task
                         *      if task hasn't already been stopped by
                         *      a signal.
                         */
                        uth->uu_siglist &= ~mask;
-                       p->p_siglist &= ~mask;
-                       if (p->p_stat != SSTOP) {
-                               p->p_xstat = signum;
-                               if ((p->p_pptr->p_flag & P_NOCLDSTOP) == 0) {
-                                       struct proc *pp = p->p_pptr;
-
-                                       pp->si_pid = p->p_pid;
-                                       pp->si_status = p->p_xstat;
+                       if (sig_proc->p_stat != SSTOP) {
+                               sig_proc->p_xstat = signum;
+                               sig_proc->p_stat = SSTOP;
+                               OSBitAndAtomic(~((uint32_t)P_CONTINUED), &sig_proc->p_flag);
+                               sig_proc->p_lflag &= ~P_LWAITED;
+                               proc_unlock(sig_proc);
+
+                               pp = proc_parentholdref(sig_proc);
+                               stop(sig_proc, pp);
+                               if (( pp != PROC_NULL) && ((pp->p_flag & P_NOCLDSTOP) == 0)) {
+
+                                       my_cred = kauth_cred_proc_ref(sig_proc);
+                                       r_uid = kauth_cred_getruid(my_cred);
+                                       kauth_cred_unref(&my_cred);
+
+                                       proc_lock(sig_proc);
+                                       pp->si_pid = sig_proc->p_pid;
+                                       /*
+                                        * POSIX: sigaction for a stopped child
+                                        * when sent to the parent must set the
+                                        * child's signal number into si_status.
+                                        */
+                                       if (signum != SIGSTOP)
+                                               pp->si_status = WEXITSTATUS(sig_proc->p_xstat);
+                                       else
+                                               pp->si_status = W_EXITCODE(signum, signum);
                                        pp->si_code = CLD_STOPPED;
-                                       pp->si_uid = p->p_ucred->cr_ruid;
+                                       pp->si_uid = r_uid;
+                                       proc_unlock(sig_proc);
+
                                        psignal(pp, SIGCHLD);
                                }
-                               stop(p);
+                               if (pp != PROC_NULL) {
+                                       proc_parentdropref(pp, 0);
+                               }
+
+                               goto sigout_unlocked;
                        }
-                       goto puthout;
+
+                       goto sigout_locked;
                }
 
+               DTRACE_PROC3(signal__send, thread_t, sig_thread, proc_t, p, int, signum);
+
                switch (signum) {
                        /*
                         * Signals ignored by default have been dealt
@@ -2004,101 +2473,196 @@ psignal_uthread(thr_act, signum)
                        /*
                         *      Process will be running after 'run'
                         */
-                       p->p_stat = SRUN;
+                       sig_proc->p_stat = SRUN;
+                       /*
+                        * In scenarios where suspend/resume are racing
+                        * the signal we are missing AST_BSD by the time
+                        * we get here, set again to avoid races. This
+                        * was the scenario with spindump enabled shutdowns.
+                        * We would need to cover this approp down the line.
+                        */
+                       act_set_astbsd(sig_thread);
+                       kret = thread_abort(sig_thread);
+                       update_thread_policy = (kret == KERN_SUCCESS);
+
+                       if (uth->uu_exit_reason == OS_REASON_NULL) {
+                               if (signal_reason == OS_REASON_NULL) {
+                                       KERNEL_DEBUG_CONSTANT(BSDDBG_CODE(DBG_BSD_PROC, BSD_PROC_EXITREASON_CREATE) | DBG_FUNC_NONE,
+                                                               sig_proc->p_pid, OS_REASON_SIGNAL, signum, 0, 0);
+
+                                       signal_reason = build_signal_reason(signum, NULL);
+                               }
 
-                       thread_abort(sig_thread_act);
+                               os_reason_ref(signal_reason);
+                               set_thread_exit_reason(sig_thread, signal_reason, TRUE);
+                       }
 
-                       goto puthout;
+                       goto sigout_locked;
 
                case SIGCONT:
                        /*
                         * Let the process run.  If it's sleeping on an
                         * event, it remains so.
                         */
-                       if (p->p_flag & P_TTYSLEEP) {
-                               p->p_flag &= ~P_TTYSLEEP;
-                               wakeup(&p->p_siglist);
-                       } else {
-                               p->p_flag |= P_CONTINUED;
-                               (void) task_resume(sig_task);
+                       assert(signal_reason == NULL);
+                       OSBitOrAtomic(P_CONTINUED, &sig_proc->p_flag);
+                       sig_proc->p_contproc = sig_proc->p_pid;
+
+                       (void) task_resume_internal(sig_task);
+
+                       /*
+                        * When processing a SIGCONT, we need to check
+                        * to see if there are signals pending that
+                        * were not delivered because we had been
+                        * previously stopped.  If that's the case,
+                        * we need to thread_abort_safely() to trigger
+                        * interruption of the current system call to
+                        * cause their handlers to fire.  If it's only
+                        * the SIGCONT, then don't wake up.
+                        */
+                       if (((flavor & (PSIG_VFORK|PSIG_THREAD)) == 0) && (((uth->uu_siglist & ~uth->uu_sigmask) & ~sig_proc->p_sigignore) & ~mask)) {
+                               uth->uu_siglist &= ~mask;
+                               sig_proc->p_stat = SRUN;
+                               goto runlocked;
                        }
+
                        uth->uu_siglist &= ~mask;
-                       p->p_siglist &= ~mask;
-                       p->p_stat = SRUN;
-                       goto puthout;
+                       sig_proc->p_stat = SRUN;
+                       goto sigout_locked;
 
                default:
+                       /*
+                        * A signal which has a default action of killing
+                        * the process, and for which there is no handler,
+                        * needs to act like SIGKILL
+                        */
+                       if (((flavor & (PSIG_VFORK|PSIG_THREAD)) == 0) && (action == SIG_DFL) && (prop & SA_KILL)) {
+                               sig_proc->p_stat = SRUN;
+                               kret = thread_abort(sig_thread);
+                               update_thread_policy = (kret == KERN_SUCCESS);
+
+                               if (uth->uu_exit_reason == OS_REASON_NULL) {
+                                       if (signal_reason == OS_REASON_NULL) {
+                                               KERNEL_DEBUG_CONSTANT(BSDDBG_CODE(DBG_BSD_PROC, BSD_PROC_EXITREASON_CREATE) | DBG_FUNC_NONE,
+                                                                       sig_proc->p_pid, OS_REASON_SIGNAL, signum, 0, 0);
+
+                                               signal_reason = build_signal_reason(signum, NULL);
+                                       }
+
+                                       os_reason_ref(signal_reason);
+                                       set_thread_exit_reason(sig_thread, signal_reason, TRUE);
+                               }
+
+                               goto sigout_locked;
+                       }
+
                        /*
                         * All other signals wake up the process, but don't
                         * resume it.
                         */
-                       goto psurun;
+                       if (sig_proc->p_stat == SSTOP) {
+                               goto sigout_locked;
+                       }
+                       goto runlocked;
                }
        }
        /*NOTREACHED*/
-psurun:
+
+runlocked:
        /*
         * If we're being traced (possibly because someone attached us
         * while we were stopped), check for a signal from the debugger.
         */
-       if (p->p_stat == SSTOP) {
-               if ((p->p_flag & P_TRACED) != 0 && p->p_xstat != 0) {
-                       uth->uu_siglist |= sigmask(p->p_xstat); 
-                       p->p_siglist |= sigmask(p->p_xstat); 
+       if (sig_proc->p_stat == SSTOP) {
+               if ((sig_proc->p_lflag & P_LTRACED) != 0 && sig_proc->p_xstat != 0)
+                       uth->uu_siglist |= sigmask(sig_proc->p_xstat);
+
+               if ((flavor & PSIG_VFORK) != 0) {
+                       sig_proc->p_stat = SRUN;
                }
        } else {
                /*
                 * setrunnable(p) in BSD and
                 * Wake up the thread if it is interruptible.
                 */
-               p->p_stat = SRUN;
-               thread_abort_safely(sig_thread_act);
+               sig_proc->p_stat = SRUN;
+               if ((flavor & PSIG_VFORK) == 0)
+                       thread_abort_safely(sig_thread);
+       }
+
+sigout_locked:
+       if (update_thread_policy) {
+               /*
+                * Update the thread policy to heading to terminate, increase priority if
+                * necessary. This needs to be done before we drop the proc lock because the
+                * thread can take the fatal signal once it's dropped.
+                */
+               proc_set_thread_policy(sig_thread, TASK_POLICY_ATTRIBUTE, TASK_POLICY_TERMINATED, TASK_POLICY_ENABLE);
        }
 
-puthout:
-               signal_unlock(p);
+       proc_unlock(sig_proc);
+
+sigout_unlocked:
+       os_reason_free(signal_reason);
+       if ((flavor & PSIG_LOCKED)== 0) {
+               proc_signalend(sig_proc, 0);
+       }
 }
 
+void
+psignal(proc_t p, int signum)
+{
+       psignal_internal(p, NULL, NULL, 0, signum, NULL);
+}
 
-__inline__ void
-sig_lock_to_exit(struct proc *p)
+void
+psignal_with_reason(proc_t p, int signum, struct os_reason *signal_reason)
 {
-       thread_t        self = current_thread();
+       psignal_internal(p, NULL, NULL, 0, signum, signal_reason);
+}
 
-       p->exit_thread = self;
-       (void) task_suspend(p->task);
+void
+psignal_locked(proc_t p, int signum)
+{
+       psignal_internal(p, NULL, NULL, PSIG_LOCKED, signum, NULL);
 }
 
-__inline__ int
-sig_try_locked(struct proc *p)
+void
+psignal_vfork_with_reason(proc_t p, task_t new_task, thread_t thread, int signum, struct os_reason *signal_reason)
 {
-       thread_t        self = current_thread();
+       psignal_internal(p, new_task, thread, PSIG_VFORK, signum, signal_reason);
+}
 
-       while (p->sigwait || p->exit_thread) {
-               if (p->exit_thread) {
-                       if (p->exit_thread != self) {
-                               /*
-                                * Already exiting - no signals.
-                                */
-                               thread_abort(self);
-                       }
-                       return(0);
-               }
-               if(assert_wait_possible()) {
-                       assert_wait((caddr_t)&p->sigwait_thread, 
-                                       (THREAD_INTERRUPTIBLE));
-               }
-               signal_unlock(p);
-               thread_block(THREAD_CONTINUE_NULL);
-               signal_lock(p);
-               if (thread_should_abort(self)) {
-                       /*
-                        * Terminate request - clean up.
-                        */
-                       return -1;
-               }
-       }
-       return 1;
+
+void
+psignal_vfork(proc_t p, task_t new_task, thread_t thread, int signum)
+{
+       psignal_internal(p, new_task, thread, PSIG_VFORK, signum, NULL);
+}
+
+void
+psignal_uthread(thread_t thread, int signum)
+{
+       psignal_internal(PROC_NULL, TASK_NULL, thread, PSIG_THREAD, signum, NULL);
+}
+
+/* same as psignal(), but prefer delivery to 'thread' if possible */
+void
+psignal_try_thread(proc_t p, thread_t thread, int signum)
+{
+       psignal_internal(p, NULL, thread, PSIG_TRY_THREAD, signum, NULL);
+}
+
+void
+psignal_try_thread_with_reason(proc_t p, thread_t thread, int signum, struct os_reason *signal_reason)
+{
+       psignal_internal(p, TASK_NULL, thread, PSIG_TRY_THREAD, signum, signal_reason);
+}
+
+void
+psignal_thread_with_reason(proc_t p, thread_t thread, int signum, struct os_reason *signal_reason)
+{
+       psignal_internal(p, TASK_NULL, thread, PSIG_THREAD, signum, signal_reason);
 }
 
 /*
@@ -2114,140 +2678,147 @@ sig_try_locked(struct proc *p)
  *             postsig(signum);
  */
 int
-issignal(p)
-       register struct proc *p;
+issignal_locked(proc_t p)
 {
-       register int signum, mask, prop, sigbits;
+       int signum, mask, prop, sigbits;
        thread_t cur_act;
        struct uthread * ut;
-       struct proc *pp;
+       proc_t pp;
+       kauth_cred_t my_cred;
+       int retval = 0;
+       uid_t r_uid;
 
        cur_act = current_thread();
 
 #if SIGNAL_DEBUG
-        if(rdebug_proc && (p == rdebug_proc)) {
-                ram_printf(3);
-        }
+       if(rdebug_proc && (p == rdebug_proc)) {
+               ram_printf(3);
+       }
 #endif /* SIGNAL_DEBUG */
-       signal_lock(p);
 
        /*
         * Try to grab the signal lock.
         */
        if (sig_try_locked(p) <= 0) {
-               signal_unlock(p);
-               return (0);
+               return 0;
        }
 
+       proc_signalstart(p, 1);
+
        ut = get_bsdthread_info(cur_act);
-       for(;;) {
-               sigbits = ut->uu_siglist  & ~ut->uu_sigmask;
+       for (;;) {
+               sigbits = ut->uu_siglist & ~ut->uu_sigmask;
 
-               if (p->p_flag & P_PPWAIT)
+               if (p->p_lflag & P_LPPWAIT)
                        sigbits &= ~stopsigmask;
                if (sigbits == 0) {             /* no signal to send */
-                       signal_unlock(p);
-                       return (0);
+                       retval = 0;
+                       goto out;
                }
+
                signum = ffs((long)sigbits);
                mask = sigmask(signum);
                prop = sigprop[signum];
 
                /*
                 * We should see pending but ignored signals
-                * only if P_TRACED was on when they were posted.
+                * only if P_LTRACED was on when they were posted.
                 */
-               if (mask & p->p_sigignore && (p->p_flag & P_TRACED) == 0) {
-                       ut->uu_siglist &= ~mask;                /* take the signal! */
-                       p->p_siglist &= ~mask;          /* take the signal! */
+               if (mask & p->p_sigignore && (p->p_lflag & P_LTRACED) == 0) {
+                       ut->uu_siglist &= ~mask;
                        continue;
                }
-               if (p->p_flag & P_TRACED && (p->p_flag & P_PPWAIT) == 0)  {
-                       register task_t task;
+
+               if (p->p_lflag & P_LTRACED && (p->p_lflag & P_LPPWAIT) == 0)  {
                        /*
-                        * If traced, always stop, and stay
-                        * stopped until released by the debugger.
+                        * If traced, deliver the signal to the debugger, and wait to be
+                        * released.
                         */
-                       /* ptrace debugging */
+                       task_t  task;
                        p->p_xstat = signum;
-                       pp = p->p_pptr;
-                       if (p->p_flag & P_SIGEXC) {
+
+                       if (p->p_lflag & P_LSIGEXC) {
                                p->sigwait = TRUE;
                                p->sigwait_thread = cur_act;
                                p->p_stat = SSTOP;
-                               p->p_flag &= ~(P_WAITED|P_CONTINUED);
-                               ut->uu_siglist &= ~mask;        /* clear the old signal */
-                               p->p_siglist &= ~mask;  /* clear the old signal */
-                               signal_unlock(p);
+                               OSBitAndAtomic(~((uint32_t)P_CONTINUED), &p->p_flag);
+                               p->p_lflag &= ~P_LWAITED;
+                               ut->uu_siglist &= ~mask; /* clear the current signal from the pending list */
+                               proc_signalend(p, 1);
+                               proc_unlock(p);
                                do_bsdexception(EXC_SOFTWARE, EXC_SOFT_SIGNAL, signum);
-                               signal_lock(p);
+                               proc_lock(p);
+                               proc_signalstart(p, 1);
                        } else {
-//                             panic("Unsupportef gdb option \n");;
-                               pp->si_pid = p->p_pid;
-                               pp->si_status = p->p_xstat;
-                               pp->si_code = CLD_TRAPPED;
-                               pp->si_uid = p->p_ucred->cr_ruid;
-                               psignal(pp, SIGCHLD);
+                               proc_unlock(p);
+                               my_cred = kauth_cred_proc_ref(p);
+                               r_uid = kauth_cred_getruid(my_cred);
+                               kauth_cred_unref(&my_cred);
+
+                               pp = proc_parentholdref(p);
+                               if (pp != PROC_NULL) {
+                                       proc_lock(pp);
+
+                                       pp->si_pid = p->p_pid;
+                                       pp->si_status = p->p_xstat;
+                                       pp->si_code = CLD_TRAPPED;
+                                       pp->si_uid = r_uid;
+
+                                       proc_unlock(pp);
+                               }
+
                                /*
                                *       XXX Have to really stop for debuggers;
                                *       XXX stop() doesn't do the right thing.
-                               *       XXX Inline the task_suspend because we
-                               *       XXX have to diddle Unix state in the
-                               *       XXX middle of it.
                                */
                                task = p->task;
-                               task_hold(task);
+                               task_suspend_internal(task);
+
+                               proc_lock(p);
                                p->sigwait = TRUE;
                                p->sigwait_thread = cur_act;
                                p->p_stat = SSTOP;
-                               p->p_flag &= ~(P_WAITED|P_CONTINUED);
-                               ut->uu_siglist &= ~mask;        /* clear the old signal */
-                               p->p_siglist &= ~mask;  /* clear the old signal */
+                               OSBitAndAtomic(~((uint32_t)P_CONTINUED), &p->p_flag);
+                               p->p_lflag &= ~P_LWAITED;
+                               ut->uu_siglist &= ~mask;
+
+                               proc_signalend(p, 1);
+                               proc_unlock(p);
+
+                               if (pp != PROC_NULL) {
+                                       psignal(pp, SIGCHLD);
+                                       proc_list_lock();
+                                       wakeup((caddr_t)pp);
+                                       proc_parentdropref(pp, 1);
+                                       proc_list_unlock();
+                               }
 
-                               wakeup((caddr_t)p->p_pptr);
-                               signal_unlock(p);
                                assert_wait((caddr_t)&p->sigwait, (THREAD_INTERRUPTIBLE));
                                thread_block(THREAD_CONTINUE_NULL);
-                               signal_lock(p);
+                               proc_lock(p);
+                               proc_signalstart(p, 1);
                        }
 
                        p->sigwait = FALSE;
                        p->sigwait_thread = NULL;
                        wakeup((caddr_t)&p->sigwait_thread);
 
-                       /*
-                        * This code is to detect when gdb is killed
-                        * even as the traced program is attached.
-                        * pgsignal would get the SIGKILL to traced program
-                        * That's what we are trying to see (I hope)
-                        */
-                       if (ut->uu_siglist & sigmask(SIGKILL)) {
+                       if (signum == SIGKILL || ut->uu_siglist & sigmask(SIGKILL)) {
                                /*
-                                * Wait event may still be outstanding;
-                                * clear it, since sig_lock_to_exit will
-                                * wait.
+                                * Deliver a pending sigkill even if it's not the current signal.
+                                * Necessary for PT_KILL, which should not be delivered to the
+                                * debugger, but we can't differentiate it from any other KILL.
                                 */
-                               clear_wait(current_thread(), THREAD_INTERRUPTED);
-                               sig_lock_to_exit(p);
-                               /*
-                               * Since this thread will be resumed
-                               * to allow the current syscall to
-                               * be completed, must save u_qsave
-                               * before calling exit().  (Since exit()
-                               * calls closef() which can trash u_qsave.)
-                               */
-                               signal_unlock(p);
-                               exit1(p,signum, (int *)NULL);
-                               return(0);
+                               signum = SIGKILL;
+                               goto deliver_sig;
                        }
 
-                       /*
-                        *      We may have to quit
-                        */
+                       /* We may have to quit. */
                        if (thread_should_abort(current_thread())) {
-                               signal_unlock(p);
-                               return(0);
+                               retval = 0;
+                               goto out;
                        }
+
                        /*
                         * If parent wants us to take the signal,
                         * then it will leave it in p->p_xstat;
@@ -2256,13 +2827,13 @@ issignal(p)
                        signum = p->p_xstat;
                        if (signum == 0)
                                continue;
+
                        /*
                         * Put the new signal into p_siglist.  If the
                         * signal is being masked, look for other signals.
                         */
                        mask = sigmask(signum);
                        ut->uu_siglist |= mask;
-                       p->p_siglist |= mask;   /* just for lame ones looking here */
                        if (ut->uu_sigmask & mask)
                                continue;
                }
@@ -2274,23 +2845,8 @@ issignal(p)
                 */
 
                switch ((long)p->p_sigacts->ps_sigact[signum]) {
-               
+
                case (long)SIG_DFL:
-                       /*
-                        * Don't take default actions on system processes.
-                        */
-                       if (p->p_pptr->p_pid == 0) {
-#if DIAGNOSTIC
-                               /*
-                                * Are you sure you want to ignore SIGSEGV
-                                * in init? XXX
-                                */
-                               printf("Process (pid %d) got signal %d\n",
-                                       p->p_pid, signum);
-#endif
-                               break;                          /* == ignore */
-                       }
-                       
                        /*
                         * If there is a pending stop signal to process
                         * with default action, stop here,
@@ -2299,36 +2855,55 @@ issignal(p)
                         * process group, ignore tty stop signals.
                         */
                        if (prop & SA_STOP) {
-                               if (p->p_flag & P_TRACED ||
-                                       (p->p_pgrp->pg_jobc == 0 &&
-                                       prop & SA_TTYSTOP))
-                                       break;  /* == ignore */
+                               struct pgrp * pg;
+
+                               proc_unlock(p);
+                               pg = proc_pgrp(p);
+                               if (p->p_lflag & P_LTRACED ||
+                                       (pg->pg_jobc == 0 &&
+                                       prop & SA_TTYSTOP)) {
+                                       proc_lock(p);
+                                       pg_rele(pg);
+                                       break; /* ignore signal */
+                               }
+                               pg_rele(pg);
                                if (p->p_stat != SSTOP) {
+                                       proc_lock(p);
                                        p->p_xstat = signum;
-                                       stop(p);
-                                       if ((p->p_pptr->p_flag & P_NOCLDSTOP) == 0) {
-                                               pp = p->p_pptr;
+                                       p->p_stat = SSTOP;
+                                       p->p_lflag &= ~P_LWAITED;
+                                       proc_unlock(p);
+
+                                       pp = proc_parentholdref(p);
+                                       stop(p, pp);
+                                       if ((pp != PROC_NULL) && ((pp->p_flag & P_NOCLDSTOP) == 0)) {
+                                               my_cred = kauth_cred_proc_ref(p);
+                                               r_uid = kauth_cred_getruid(my_cred);
+                                               kauth_cred_unref(&my_cred);
+
+                                               proc_lock(pp);
                                                pp->si_pid = p->p_pid;
-                                               pp->si_status = p->p_xstat;
+                                               pp->si_status = WEXITSTATUS(p->p_xstat);
                                                pp->si_code = CLD_STOPPED;
-                                               pp->si_uid = p->p_ucred->cr_ruid;
+                                               pp->si_uid = r_uid;
+                                               proc_unlock(pp);
+
                                                psignal(pp, SIGCHLD);
                                        }
+                                       if (pp != PROC_NULL)
+                                               proc_parentdropref(pp, 0);
                                }
+                               proc_lock(p);
                                break;
                        } else if (prop & SA_IGNORE) {
                                /*
                                 * Except for SIGCONT, shouldn't get here.
                                 * Default action is to ignore; drop it.
                                 */
-                               break;          /* == ignore */
+                               break; /* ignore signal */
                        } else {
-                               ut->uu_siglist &= ~mask;        /* take the signal! */
-                               p->p_siglist &= ~mask;  /* take the signal! */
-                               signal_unlock(p);
-                               return (signum);
+                               goto deliver_sig;
                        }
-                       /*NOTREACHED*/
 
                case (long)SIG_IGN:
                        /*
@@ -2337,32 +2912,36 @@ issignal(p)
                         * than SIGCONT, unless process is traced.
                         */
                        if ((prop & SA_CONT) == 0 &&
-                               (p->p_flag & P_TRACED) == 0)
+                               (p->p_lflag & P_LTRACED) == 0)
                                printf("issignal\n");
-                       break;          /* == ignore */
+                       break; /* ignore signal */
 
                default:
-                       /*
-                        * This signal has an action, let
-                        * postsig() process it.
-                        */
-                       ut->uu_siglist &= ~mask;                /* take the signal! */
-                       p->p_siglist &= ~mask;          /* take the signal! */
-                       signal_unlock(p);
-                       return (signum);
-               }
-               ut->uu_siglist &= ~mask;                /* take the signal! */
-               p->p_siglist &= ~mask;          /* take the signal! */
+                       /* This signal has an action - deliver it. */
+                       goto deliver_sig;
                }
+
+               /* If we dropped through, the signal was ignored - remove it from pending list. */
+               ut->uu_siglist &= ~mask;
+
+       } /* for(;;) */
+
        /* NOTREACHED */
+
+deliver_sig:
+       ut->uu_siglist &= ~mask;
+       retval = signum;
+
+out:
+       proc_signalend(p, 1);
+       return retval;
 }
 
 /* called from _sleep */
 int
-CURSIG(p)
-    register struct proc *p;
+CURSIG(proc_t p)
 {
-       register int signum, mask, prop, sigbits;
+       int signum, mask, prop, sigbits;
        thread_t cur_act;
        struct uthread * ut;
        int retnum = 0;
@@ -2375,13 +2954,13 @@ CURSIG(p)
        if (ut->uu_siglist == 0)
                return (0);
 
-       if (((ut->uu_siglist & ~ut->uu_sigmask) == 0) && ((p->p_flag & P_TRACED) == 0))
+       if (((ut->uu_siglist & ~ut->uu_sigmask) == 0) && ((p->p_lflag & P_LTRACED) == 0))
                return (0);
 
        sigbits = ut->uu_siglist & ~ut->uu_sigmask;
 
        for(;;) {
-               if (p->p_flag & P_PPWAIT)
+               if (p->p_lflag & P_LPPWAIT)
                        sigbits &= ~stopsigmask;
                if (sigbits == 0) {             /* no signal to send */
                        return (retnum);
@@ -2390,22 +2969,17 @@ CURSIG(p)
                signum = ffs((long)sigbits);
                mask = sigmask(signum);
                prop = sigprop[signum];
+               sigbits &= ~mask;               /* take the signal out */
 
                /*
                 * We should see pending but ignored signals
-                * only if P_TRACED was on when they were posted.
+                * only if P_LTRACED was on when they were posted.
                 */
-               if (mask & p->p_sigignore && (p->p_flag & P_TRACED) == 0) {
+               if (mask & p->p_sigignore && (p->p_lflag & P_LTRACED) == 0) {
                        continue;
                }
-               if (p->p_flag & P_TRACED && (p->p_flag & P_PPWAIT) == 0) {
-                       /*
-                        * Put the new signal into p_siglist.  If the
-                        * signal is being masked, look for other signals.
-                        */
-                       mask = sigmask(signum);
-                       if (ut->uu_sigmask & mask)
-                               continue;
+
+               if (p->p_lflag & P_LTRACED && (p->p_lflag & P_LPPWAIT) == 0) {
                        return(signum);
                }
 
@@ -2418,21 +2992,6 @@ CURSIG(p)
                switch ((long)p->p_sigacts->ps_sigact[signum]) {
                
                case (long)SIG_DFL:
-                       /*
-                        * Don't take default actions on system processes.
-                        */
-                       if (p->p_pptr->p_pid == 0) {
-#if DIAGNOSTIC
-                               /*
-                                * Are you sure you want to ignore SIGSEGV
-                                * in init? XXX
-                                */
-                               printf("Process (pid %d) got signal %d\n",
-                                       p->p_pid, signum);
-#endif
-                               break;                          /* == ignore */
-                       }
-                       
                        /*
                         * If there is a pending stop signal to process
                         * with default action, stop here,
@@ -2441,10 +3000,17 @@ CURSIG(p)
                         * process group, ignore tty stop signals.
                         */
                        if (prop & SA_STOP) {
-                               if (p->p_flag & P_TRACED ||
-                                       (p->p_pgrp->pg_jobc == 0 &&
-                                       prop & SA_TTYSTOP))
+                               struct pgrp *pg;
+
+                               pg = proc_pgrp(p);
+
+                               if (p->p_lflag & P_LTRACED ||
+                                       (pg->pg_jobc == 0 &&
+                                       prop & SA_TTYSTOP)) {
+                                       pg_rele(pg);
                                        break;  /* == ignore */
+                               }
+                               pg_rele(pg);
                                retnum = signum;
                                break;
                        } else if (prop & SA_IGNORE) {
@@ -2465,7 +3031,7 @@ CURSIG(p)
                         * than SIGCONT, unless process is traced.
                         */
                        if ((prop & SA_CONT) == 0 &&
-                               (p->p_flag & P_TRACED) == 0)
+                               (p->p_lflag & P_LTRACED) == 0)
                                printf("issignal\n");
                        break;          /* == ignore */
 
@@ -2476,7 +3042,6 @@ CURSIG(p)
                         */
                        return (signum);
                }
-               sigbits &= ~mask;               /* take the signal! */
        }
        /* NOTREACHED */
 }
@@ -2486,15 +3051,16 @@ CURSIG(p)
  * via wakeup.  Signals are handled elsewhere.  The process must not be
  * on the run queue.
  */
-void
-stop(p)
-       register struct proc *p;
+static void
+stop(proc_t p, proc_t parent)
 {
-       p->p_stat = SSTOP;
-       p->p_flag &= ~(P_WAITED|P_CONTINUED);
-       if (p->p_pptr->p_stat != SSTOP)
-               wakeup((caddr_t)p->p_pptr);
-       (void) task_suspend(p->task);   /*XXX*/
+       OSBitAndAtomic(~((uint32_t)P_CONTINUED), &p->p_flag);
+       if ((parent != PROC_NULL) && (parent->p_stat != SSTOP)) {
+               proc_list_lock();
+               wakeup((caddr_t)parent);
+               proc_list_unlock();
+       }
+       (void) task_suspend_internal(p->task);
 }
 
 /*
@@ -2502,12 +3068,12 @@ stop(p)
  * from the current set of pending signals.
  */
 void
-postsig(int signum)
+postsig_locked(int signum)
 {
-       struct proc *p = current_proc();
+       proc_t p = current_proc();
        struct sigacts *ps = p->p_sigacts;
        user_addr_t catcher;
-       u_long code;
+       uint32_t code;
        int mask, returnmask;
        struct uthread * ut;
 
@@ -2521,36 +3087,73 @@ postsig(int signum)
                panic("psig not on master");
 #endif
 
-       signal_lock(p);
        /*
         * Try to grab the signal lock.
         */
        if (sig_try_locked(p) <= 0) {
-               signal_unlock(p);
                return;
        }
 
+       proc_signalstart(p, 1);
+
        ut = (struct uthread *)get_bsdthread_info(current_thread());
        mask = sigmask(signum);
        ut->uu_siglist &= ~mask;
-       p->p_siglist &= ~mask;
        catcher = ps->ps_sigact[signum];
-#if KTRACE
-       //LP64: catcher argument is a 64 bit user space handler address
-       if (KTRPOINT(p, KTR_PSIG))
-               ktrpsig(p->p_tracep,
-                   signum, CAST_DOWN(void *,catcher), ut->uu_flag & UT_SAS_OLDMASK ?
-                   &ut->uu_oldmask : &ut->uu_sigmask, 0);
-#endif
        if (catcher == SIG_DFL) {
                /*
                 * Default catcher, where the default is to kill
                 * the process.  (Other cases were ignored above.)
                 */
-               /* called with signal_lock() held */
-               sigexit_locked(p, signum);
+               sig_lock_to_exit(p);
+               p->p_acflag |= AXSIG;
+               if (sigprop[signum] & SA_CORE) {
+                       p->p_sigacts->ps_sig = signum;
+                       proc_signalend(p, 1);
+                       proc_unlock(p);
+#if CONFIG_COREDUMP
+                       if (coredump(p, 0, 0) == 0)
+                               signum |= WCOREFLAG;
+#endif
+               } else  {
+                       proc_signalend(p, 1);
+                       proc_unlock(p);
+               }
+               
+#if CONFIG_DTRACE
+               bzero((caddr_t)&(ut->t_dtrace_siginfo), sizeof(ut->t_dtrace_siginfo));
+
+               ut->t_dtrace_siginfo.si_signo = signum;
+               ut->t_dtrace_siginfo.si_pid = p->si_pid;
+               ut->t_dtrace_siginfo.si_uid = p->si_uid;
+               ut->t_dtrace_siginfo.si_status = WEXITSTATUS(p->si_status);
+
+               /* Fire DTrace proc:::fault probe when signal is generated by hardware. */
+               switch (signum) {
+               case SIGILL: case SIGBUS: case SIGSEGV: case SIGFPE: case SIGTRAP:
+                       DTRACE_PROC2(fault, int, (int)(ut->uu_code), siginfo_t *, &(ut->t_dtrace_siginfo));
+                       break;
+               default:
+                       break;
+               }
+               
+
+               DTRACE_PROC3(signal__handle, int, signum, siginfo_t *, &(ut->t_dtrace_siginfo),
+                                       void (*)(void), SIG_DFL);
+#endif
+
+               KERNEL_DEBUG_CONSTANT(BSDDBG_CODE(DBG_BSD_PROC, BSD_PROC_FRCEXIT) | DBG_FUNC_NONE,
+                                             p->p_pid, W_EXITCODE(0, signum), 3, 0, 0);
+
+               /*
+                * exit_with_reason() will consume a reference to the thread's exit reason, so we take another
+                * reference for the thread. This reference will be destroyed in uthread_cleanup().
+                */
+               os_reason_ref(ut->uu_exit_reason);
+               exit_with_reason(p, W_EXITCODE(0, signum), (int *)NULL, TRUE, TRUE, 0, ut->uu_exit_reason);
+
+               proc_lock(p);
                return;
-               /* NOTREACHED */
        } else {
                /*
                 * If we get here, the signal must be caught.
@@ -2560,6 +3163,7 @@ postsig(int signum)
                        log(LOG_WARNING,
                                "postsig: processing masked or ignored signal\n");
 #endif
+
                /*
                 * Set the new mask value and also defer further
                 * occurences of this signal.
@@ -2585,12 +3189,6 @@ postsig(int signum)
                        ps->ps_siginfo &= ~mask;
                        ps->ps_signodefer &= ~mask;
                }
-#ifdef __ppc__
-               /* Needs to disable to run in user mode */
-               if (signum == SIGFPE) {
-                       thread_enable_fpe(current_thread(), 0);
-               }
-#endif  /* __ppc__ */
 
                if (ps->ps_sig != signum) {
                        code = 0;
@@ -2598,69 +3196,61 @@ postsig(int signum)
                        code = ps->ps_code;
                        ps->ps_code = 0;
                }
-               p->p_stats->p_ru.ru_nsignals++;
+               OSIncrementAtomicLong(&p->p_stats->p_ru.ru_nsignals);
                sendsig(p, catcher, signum, returnmask, code);
        }
-       signal_unlock(p);
+       proc_signalend(p, 1);
 }
 
 /*
- * Force the current process to exit with the specified signal, dumping core
- * if appropriate.  We bypass the normal tests for masked and caught signals,
- * allowing unrecoverable failures to terminate the process without changing
- * signal state.  Mark the accounting record with the signal termination.
- * If dumping core, save the signal number for the debugger.  Calls exit and
- * does not return.
+ * Attach a signal knote to the list of knotes for this process.
+ *
+ * Signal knotes share the knote list with proc knotes.  This
+ * could be avoided by using a signal-specific knote list, but
+ * probably isn't worth the trouble.
  */
- /* called with signal lock */
-void
-sigexit_locked(p, signum)
-       register struct proc *p;
-       int signum;
-{
-
-       sig_lock_to_exit(p);
-       p->p_acflag |= AXSIG;
-       if (sigprop[signum] & SA_CORE) {
-               p->p_sigacts->ps_sig = signum;
-               signal_unlock(p);
-               if (coredump(p) == 0)
-                       signum |= WCOREFLAG;
-       } else 
-               signal_unlock(p);
-               
-       exit1(p, W_EXITCODE(0, signum), (int *)NULL);
-       /* NOTREACHED */
-}
-
 
 static int
 filt_sigattach(struct knote *kn)
 {
-       struct proc *p = current_proc();
+       proc_t p = current_proc();  /* can attach only to oneself */
+
+       proc_klist_lock();
 
        kn->kn_ptr.p_proc = p;
-       kn->kn_flags |= EV_CLEAR;               /* automatically set */
 
-       /* XXX lock the proc here while adding to the list? */
        KNOTE_ATTACH(&p->p_klist, kn);
 
+       proc_klist_unlock();
+
+       /* edge-triggered events can't have fired before we attached */
        return (0);
 }
 
+/*
+ * remove the knote from the process list, if it hasn't already
+ * been removed by exit processing.  
+ */
+          
 static void
 filt_sigdetach(struct knote *kn)
 {
-       struct proc *p = kn->kn_ptr.p_proc;
+       proc_t p = kn->kn_ptr.p_proc;
 
+       proc_klist_lock();
+       kn->kn_ptr.p_proc = NULL;
        KNOTE_DETACH(&p->p_klist, kn);
+       proc_klist_unlock();
 }
 
 /*
- * signal knotes are shared with proc knotes, so we apply a mask to 
- * the hint in order to differentiate them from process hints.  This
- * could be avoided by using a signal-specific knote list, but probably
- * isn't worth the trouble.
+ * Post an event to the signal filter.  Because we share the same list
+ * as process knotes, we have to filter out and handle only signal events.
+ *
+ * We assume that we process fdfree() before we post the NOTE_EXIT for
+ * a process during exit.  Therefore, since signal filters can only be
+ * set up "in-process", we should have already torn down the kqueue
+ * hosting the EVFILT_SIGNAL knote and should never see NOTE_EXIT.
  */
 static int
 filt_signal(struct knote *kn, long hint)
@@ -2671,109 +3261,214 @@ filt_signal(struct knote *kn, long hint)
 
                if (kn->kn_id == (unsigned int)hint)
                        kn->kn_data++;
+       } else if (hint & NOTE_EXIT) {
+               panic("filt_signal: detected NOTE_EXIT event");
        }
+
        return (kn->kn_data != 0);
 }
 
+static int
+filt_signaltouch(
+       struct knote *kn,
+       struct kevent_internal_s *kev)
+{
+#pragma unused(kev)
+
+       int res;
+
+       proc_klist_lock();
+
+       if ((kn->kn_status & KN_UDATA_SPECIFIC) == 0)
+               kn->kn_udata = kev->udata;
+       /* 
+        * No data to save - 
+        * just capture if it is already fired
+        */
+       res = (kn->kn_data > 0);
+
+       proc_klist_unlock();
+
+       return res;
+}
+
+static int
+filt_signalprocess(
+       struct knote *kn,
+       __unused struct filt_process_s *data,
+       struct kevent_internal_s *kev)
+{
+       proc_klist_lock();
+
+       if (kn->kn_data == 0) {
+               proc_klist_unlock();
+               return 0;
+       }
+
+       /*
+        * Snapshot the event data.
+        * All signal events are EV_CLEAR, so
+        * add that and clear out the data field.
+        */
+       *kev = kn->kn_kevent;
+       kev->flags |= EV_CLEAR;
+       kn->kn_data = 0;
+
+       proc_klist_unlock();
+       return 1;
+}
 
 void
-bsd_ast(thread_t thr_act)
+bsd_ast(thread_t thread)
 {
-       struct proc *p = current_proc();
-       struct uthread *ut = get_bsdthread_info(thr_act);
+       proc_t p = current_proc();
+       struct uthread *ut = get_bsdthread_info(thread);
        int     signum;
        user_addr_t pc;
-       boolean_t funnel_state;
        static int bsd_init_done = 0;
 
        if (p == NULL)
                return;
 
-       funnel_state = thread_funnel_set(kernel_flock, TRUE);
+       /* don't run bsd ast on exec copy or exec'ed tasks */
+       if (task_did_exec(current_task()) || task_is_exec_copy(current_task())) {
+               return;
+       }
 
        if ((p->p_flag & P_OWEUPC) && (p->p_flag & P_PROFIL)) {
                pc = get_useraddr();
                addupc_task(p, pc, 1);
-               p->p_flag &= ~P_OWEUPC;
+               OSBitAndAtomic(~((uint32_t)P_OWEUPC), &p->p_flag);
        }
 
-       if (CHECK_SIGNALS(p, current_thread(), ut)) {
-               while ( (signum = issignal(p)) )
-                       postsig(signum);
+       if (timerisset(&p->p_vtimer_user.it_value)) {
+               uint32_t        microsecs;
+
+               task_vtimer_update(p->task, TASK_VTIMER_USER, &microsecs);
+
+               if (!itimerdecr(p, &p->p_vtimer_user, microsecs)) {
+                       if (timerisset(&p->p_vtimer_user.it_value))
+                               task_vtimer_set(p->task, TASK_VTIMER_USER);
+                       else
+                               task_vtimer_clear(p->task, TASK_VTIMER_USER);
+
+                       psignal_try_thread(p, thread, SIGVTALRM);
+               }
        }
-       if (!bsd_init_done) {
-               bsd_init_done = 1;
-               bsdinit_task();
+
+       if (timerisset(&p->p_vtimer_prof.it_value)) {
+               uint32_t        microsecs;
+
+               task_vtimer_update(p->task, TASK_VTIMER_PROF, &microsecs);
+
+               if (!itimerdecr(p, &p->p_vtimer_prof, microsecs)) {
+                       if (timerisset(&p->p_vtimer_prof.it_value))
+                               task_vtimer_set(p->task, TASK_VTIMER_PROF);
+                       else
+                               task_vtimer_clear(p->task, TASK_VTIMER_PROF);
+
+                       psignal_try_thread(p, thread, SIGPROF);
+               }
        }
 
-       (void) thread_funnel_set(kernel_flock, FALSE);
-}
+       if (timerisset(&p->p_rlim_cpu)) {
+               struct timeval          tv;
 
-/*
- * Follwing routines are called using callout from bsd_hardclock
- * so that psignals are called in a thread context and are funneled
- */
-void
-psignal_vtalarm(struct proc *p)
-{
-       boolean_t funnel_state;
+               task_vtimer_update(p->task, TASK_VTIMER_RLIM, (uint32_t *) &tv.tv_usec);
 
-       if (p == NULL)
-               return;
-       funnel_state = thread_funnel_set(kernel_flock, TRUE);
-       psignal_lock(p, SIGVTALRM, 1);
-       (void) thread_funnel_set(kernel_flock, FALSE);
-}
+               proc_spinlock(p);
+               if (p->p_rlim_cpu.tv_sec > 0 || p->p_rlim_cpu.tv_usec > tv.tv_usec) {
+                       tv.tv_sec = 0;
+                       timersub(&p->p_rlim_cpu, &tv, &p->p_rlim_cpu);
+                       proc_spinunlock(p);
+               } else {
 
-void
-psignal_xcpu(struct proc *p)
-{
-       boolean_t funnel_state;
+                       timerclear(&p->p_rlim_cpu);
+                       proc_spinunlock(p);
 
-       if (p == NULL)
-               return;
-       funnel_state = thread_funnel_set(kernel_flock, TRUE);
-       psignal_lock(p, SIGXCPU, 1);
-       (void) thread_funnel_set(kernel_flock, FALSE);
-}
+                       task_vtimer_clear(p->task, TASK_VTIMER_RLIM);
 
-void
-psignal_sigprof(struct proc *p)
-{
-       boolean_t funnel_state;
+                       psignal_try_thread(p, thread, SIGXCPU);
+               }
+       }
+
+#if CONFIG_DTRACE
+       if (ut->t_dtrace_sig) {
+           uint8_t dt_action_sig = ut->t_dtrace_sig;
+           ut->t_dtrace_sig = 0;
+           psignal(p, dt_action_sig);
+       }
+
+       if (ut->t_dtrace_stop) {
+               ut->t_dtrace_stop = 0;
+               proc_lock(p);
+               p->p_dtrace_stop = 1;
+               proc_unlock(p);
+               (void)task_suspend_internal(p->task);
+       }
+
+       if (ut->t_dtrace_resumepid) {
+               proc_t resumeproc = proc_find(ut->t_dtrace_resumepid);
+               ut->t_dtrace_resumepid = 0;
+               if (resumeproc != PROC_NULL) {
+                       proc_lock(resumeproc);
+                       /* We only act on processes stopped by dtrace */
+                       if (resumeproc->p_dtrace_stop) {
+                               resumeproc->p_dtrace_stop = 0;
+                               proc_unlock(resumeproc);
+                               task_resume_internal(resumeproc->task);
+                       }
+                       else {
+                               proc_unlock(resumeproc);
+                       }
+                       proc_rele(resumeproc);
+               }
+       }
+                   
+#endif /* CONFIG_DTRACE */
+
+       proc_lock(p);
+       if (CHECK_SIGNALS(p, current_thread(), ut)) {
+               while ( (signum = issignal_locked(p)) )
+                       postsig_locked(signum);
+       }
+       proc_unlock(p);
+
+       if (!bsd_init_done) {
+               bsd_init_done = 1;
+               bsdinit_task();
+       }
 
-       if (p == NULL)
-               return;
-       funnel_state = thread_funnel_set(kernel_flock, TRUE);
-       psignal_lock(p, SIGPROF, 1);
-       (void) thread_funnel_set(kernel_flock, FALSE);
 }
 
-/* ptrace set runnalbe */
+/* ptrace set runnable */
 void
-pt_setrunnable(struct proc *p)
+pt_setrunnable(proc_t p)
 {
-task_t task;
+       task_t task;
 
        task = p->task;
 
-       if (p->p_flag & P_TRACED) {
+       if (p->p_lflag & P_LTRACED) {
+               proc_lock(p);
                p->p_stat = SRUN;
+               proc_unlock(p);
                if (p->sigwait) {
                        wakeup((caddr_t)&(p->sigwait));
-                       task_release(task);
+                       if ((p->p_lflag & P_LSIGEXC) == 0) {    // 5878479
+                               task_release(task);
+                       }
                }
        }
 }
 
-
 kern_return_t
 do_bsdexception(
            int exc,
            int code,
            int sub)
 {
-       exception_data_type_t   codes[EXCEPTION_CODE_MAX];
+       mach_exception_data_type_t   codes[EXCEPTION_CODE_MAX];
 
        codes[0] = code;        
        codes[1] = sub;
@@ -2781,28 +3476,19 @@ do_bsdexception(
 }
 
 int
-proc_pendingsignals(struct proc *p, sigset_t mask)
+proc_pendingsignals(proc_t p, sigset_t mask)
 {
        struct uthread * uth;
        thread_t th;
        sigset_t bits = 0;
-       int error;
 
+       proc_lock(p);
        /* If the process is in proc exit return no signal info */
-       if (p->p_lflag & P_LPEXIT) 
-               return(0);
-
-       /* duplicate the signal lock code to enable recursion; as exit
-        * holds the lock too long. All this code is being reworked
-        * this is just a workaround for regressions till new code
-        * arrives.
-        */
-ppend_retry:
-               error = lockmgr((struct lock__bsd__ *)&p->signal_lock[0], (LK_EXCLUSIVE | LK_CANRECURSE), 0, (struct proc *)0);
-               if (error == EINTR)
-                       goto ppend_retry;
+       if (p->p_lflag & P_LPEXIT)  {
+               goto out;
+       }
 
-       if ((p->p_flag & P_INVFORK) && p->p_vforkact) {
+       if ((p->p_lflag & P_LINVFORK) && p->p_vforkact) {
                th = p->p_vforkact;     
                uth = (struct uthread *)get_bsdthread_info(th);
                if (uth) {
@@ -2816,7 +3502,7 @@ ppend_retry:
                bits |= (((uth->uu_siglist & ~uth->uu_sigmask) & ~p->p_sigignore) & mask);
        }
 out:
-       signal_unlock(p);
+       proc_unlock(p);
        return(bits);
 }
 
@@ -2826,11 +3512,105 @@ thread_issignal(proc_t p, thread_t th, sigset_t mask)
        struct uthread * uth;
        sigset_t  bits=0;
 
+       proc_lock(p);
+       uth = (struct uthread *)get_bsdthread_info(th);
+       if (uth) {
+               bits = (((uth->uu_siglist & ~uth->uu_sigmask) & ~p->p_sigignore) & mask);
+       }
+       proc_unlock(p);
+       return(bits);
+}
 
-               uth = (struct uthread *)get_bsdthread_info(th);
-               if (uth) {
-                       bits = (((uth->uu_siglist & ~uth->uu_sigmask) & ~p->p_sigignore) & mask);
-               }
-               return(bits);
+/*
+ * Allow external reads of the sigprop array.
+ */
+int
+hassigprop(int sig, int prop)
+{
+       return (sigprop[sig] & prop);
+}
+
+void
+pgsigio(pid_t pgid, int sig)
+{ 
+       proc_t p = PROC_NULL;
+
+       if (pgid < 0) 
+               gsignal(-(pgid), sig);
+
+       else if (pgid > 0 && (p = proc_find(pgid)) != 0) 
+               psignal(p, sig);
+       if (p != PROC_NULL)
+               proc_rele(p);
+}
+
+void
+proc_signalstart(proc_t p, int locked)
+{
+       if (!locked)
+               proc_lock(p);
+       
+       if(p->p_signalholder == current_thread())
+               panic("proc_signalstart: thread attempting to signal a process for which it holds the signal lock");    
+       
+       p->p_sigwaitcnt++;
+       while ((p->p_lflag & P_LINSIGNAL) == P_LINSIGNAL)
+               msleep(&p->p_sigmask, &p->p_mlock, 0, "proc_signstart", NULL);
+       p->p_sigwaitcnt--;
+
+       p->p_lflag |= P_LINSIGNAL;
+       p->p_signalholder = current_thread();
+       if (!locked)
+               proc_unlock(p);
 }
 
+void
+proc_signalend(proc_t p, int locked)
+{
+       if (!locked)
+               proc_lock(p);
+       p->p_lflag &= ~P_LINSIGNAL;
+
+       if (p->p_sigwaitcnt > 0)
+               wakeup(&p->p_sigmask);
+
+       p->p_signalholder = NULL;
+       if (!locked)
+               proc_unlock(p);
+}
+
+void
+sig_lock_to_exit(proc_t p)
+{
+       thread_t        self = current_thread();
+
+       p->exit_thread = self;
+       proc_unlock(p);
+
+       task_hold(p->task);
+       task_wait(p->task, FALSE);
+
+       proc_lock(p);
+}
+
+int
+sig_try_locked(proc_t p)
+{
+       thread_t        self = current_thread();
+
+       while (p->sigwait || p->exit_thread) {
+               if (p->exit_thread) {
+                       return(0);
+               }
+               msleep((caddr_t)&p->sigwait_thread, &p->p_mlock, PCATCH | PDROP, 0, 0);
+               if (thread_should_abort(self)) {
+                       /*
+                        * Terminate request - clean up.
+                        */
+                       proc_lock(p);
+                       return -1;
+               }
+               proc_lock(p);
+       }
+       return 1;
+}