]> git.saurik.com Git - apple/xnu.git/blobdiff - bsd/kern/kdebug.c
xnu-1228.5.18.tar.gz
[apple/xnu.git] / bsd / kern / kdebug.c
index 381b74fe2f83a531e2ce374837b57d1e6b1f062b..952d2b87c3f7639d334b0f062ce4a48b81360f33 100644 (file)
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2000-2004 Apple Computer, Inc. All rights reserved.
+ * Copyright (c) 2000-2006 Apple Computer, Inc. All rights reserved.
  *
  * @Apple_LICENSE_HEADER_START@
  * 
  * License for the specific language governing rights and limitations
  * under the License.
  * 
- * @APPLE_LICENSE_HEADER_END@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
  */
 
 #include <machine/spl.h>
 
 #include <sys/errno.h>
 #include <sys/param.h>
+#include <sys/systm.h>
 #include <sys/proc_internal.h>
 #include <sys/vm.h>
 #include <sys/sysctl.h>
 
 #include <kern/thread.h>
 #include <kern/task.h>
+#include <kern/debug.h>
 #include <vm/vm_kern.h>
 #include <sys/lock.h>
 
+#include <sys/malloc.h>
+#include <sys/kauth.h>
+
+#include <mach/mach_host.h>            /* for host_info() */
+#include <libkern/OSAtomic.h>
+
+/* XXX should have prototypes, but Mach does not provide one */
+void task_act_iterate_wth_args(task_t, void(*)(thread_t, void *), void *);
+int cpu_number(void);  /* XXX <machine/...> include path broken */
+
+/* XXX should probably be static, but it's debugging code... */
+int kdbg_read(user_addr_t, size_t *);
+void kdbg_control_chud(int, void *);
+int kdbg_control(int *, u_int, user_addr_t, size_t *);
+int kdbg_getentropy (user_addr_t, size_t *, int);
+int kdbg_readmap(user_addr_t, size_t *);
+int kdbg_getreg(kd_regtype *);
+int kdbg_setreg(kd_regtype *);
+int kdbg_setrtcdec(kd_regtype *);
+int kdbg_setpidex(kd_regtype *);
+int kdbg_setpid(kd_regtype *);
+void kdbg_mapinit(void);
+int kdbg_reinit(void);
+int kdbg_bootstrap(void);
+
+static int create_buffers(void);
+static void delete_buffers(void);
+
+extern void IOSleep(int);
+
+#ifdef ppc
+extern uint32_t maxDec;
+#endif
+
 /* trace enable status */
 unsigned int kdebug_enable = 0;
 
@@ -58,12 +94,27 @@ unsigned int      kd_entropy_buftomem = 0;
 
 unsigned int kdebug_slowcheck=SLOW_NOLOG;
 
-/* kd_buf kd_buffer[kd_bufsize/sizeof(kd_buf)]; */
-kd_buf * kd_bufptr;
-unsigned int kd_buftomem=0;
-kd_buf * kd_buffer=0;
-kd_buf * kd_buflast;
-kd_buf * kd_readlast;
+unsigned int kd_cpus;
+
+struct kd_bufinfo {
+        kd_buf * kd_stop;
+        kd_buf * kd_bufptr;
+        kd_buf * kd_buffer;
+        kd_buf * kd_buflast;
+        kd_buf * kd_readlast;
+        int      kd_wrapped;   /* plus, the global flag KDBG_WRAPPED is set if one of the buffers has wrapped */
+        uint64_t kd_prev_timebase;
+        int     kd_pad[24];    /* pad out to 128 bytes so that no cache line is shared between CPUs */
+
+};
+
+struct kd_bufinfo *kdbip = NULL;
+
+#define KDCOPYBUF_COUNT        1024
+#define KDCOPYBUF_SIZE (KDCOPYBUF_COUNT * sizeof(kd_buf))
+kd_buf *kdcopybuf = NULL;
+
+
 unsigned int nkdbufs = 8192;
 unsigned int kd_bufsize = 0;
 unsigned int kdebug_flags = 0;
@@ -74,17 +125,28 @@ unsigned int kdlog_value2=0;
 unsigned int kdlog_value3=0;
 unsigned int kdlog_value4=0;
 
-unsigned long long kd_prev_timebase = 0LL;
+static lck_mtx_t  * kd_trace_mtx_sysctl;
+static lck_grp_t  * kd_trace_mtx_sysctl_grp;
+static lck_attr_t * kd_trace_mtx_sysctl_attr;
+static lck_grp_attr_t   *kd_trace_mtx_sysctl_grp_attr;
+
+static lck_grp_t       *stackshot_subsys_lck_grp;
+static lck_grp_attr_t  *stackshot_subsys_lck_grp_attr;
+static lck_attr_t      *stackshot_subsys_lck_attr;
+static lck_mtx_t        stackshot_subsys_mutex;
+
+void *stackshot_snapbuf = NULL;
+
+int
+stack_snapshot2(pid_t pid, user_addr_t tracebuf, uint32_t tracebuf_size, uint32_t options, register_t *retval);
 
-static lck_mtx_t  * kd_trace_mtx;
-static lck_grp_t  * kd_trace_mtx_grp;
-static lck_attr_t * kd_trace_mtx_attr;
-static lck_grp_attr_t   *kd_trace_mtx_grp_attr;
+extern void
+kdp_snapshot_preflight(int pid, void  *tracebuf, uint32_t tracebuf_size, uint32_t options);
 
-static lck_spin_t * kd_trace_lock;
-static lck_grp_t  * kd_trace_lock_grp;
-static lck_attr_t * kd_trace_lock_attr;
-static lck_grp_attr_t   *kd_trace_lock_grp_attr;
+extern int
+kdp_stack_snapshot_geterror(void);
+extern unsigned int
+kdp_stack_snapshot_bytes_traced(void);
 
 kd_threadmap *kd_mapptr = 0;
 unsigned int kd_mapsize = 0;
@@ -95,14 +157,10 @@ pid_t global_state_pid = -1;       /* Used to control exclusive use of kd_buffer
 
 #define DBG_FUNC_MASK 0xfffffffc
 
-#ifdef ppc
-extern natural_t rtclock_decrementer_min;
-#endif /* ppc */
-
 /* task to string structure */
 struct tts
 {
-  task_t   *task;            /* from procs task */
+  task_t    task;            /* from procs task */
   pid_t     pid;             /* from procs p_pid  */
   char      task_comm[20];   /* from procs p_comm */
 };
@@ -126,12 +184,11 @@ typedef void (*kd_chudhook_fn) (unsigned int debugid, unsigned int arg1,
 
 kd_chudhook_fn kdebug_chudhook = 0;   /* pointer to CHUD toolkit function */
 
+__private_extern__ void stackshot_lock_init( void ) __attribute__((section("__TEXT, initcode")));
 
 /* Support syscall SYS_kdebug_trace */
-kdebug_trace(p, uap, retval)
-     struct proc *p;
-     struct kdebug_trace_args *uap;
-     register_t *retval;
+int
+kdebug_trace(__unused struct proc *p, struct kdebug_trace_args *uap, __unused register_t *retval)
 {
     if ( (kdebug_enable == 0) )
         return(EINVAL);
@@ -140,31 +197,93 @@ kdebug_trace(p, uap, retval)
     return(0);
 }
 
+static int
+create_buffers(void)
+{
+        unsigned int   cpu, i;
+       int     nentries;
+
+       nentries = nkdbufs / kd_cpus;
+       nkdbufs = nentries * kd_cpus;
 
-void
-kernel_debug(debugid, arg1, arg2, arg3, arg4, arg5)
-unsigned int debugid, arg1, arg2, arg3, arg4, arg5;
+       kd_bufsize = nentries * sizeof(kd_buf);
+
+       bzero((char *)kdbip, sizeof(struct kd_bufinfo) * kd_cpus);
+
+       if (kdcopybuf == 0) {
+               if (kmem_alloc(kernel_map, (unsigned int *)&kdcopybuf, (vm_size_t)KDCOPYBUF_SIZE) != KERN_SUCCESS)
+                       return(ENOMEM);
+       }
+       for (cpu = 0; cpu < kd_cpus; cpu++) {
+               if (kmem_alloc(kernel_map, (unsigned int *)&kdbip[cpu].kd_buffer, kd_bufsize) != KERN_SUCCESS)
+                       break;
+       }
+       if (cpu < kd_cpus) {
+               for (i = 0; i < cpu; i++)
+                       kmem_free(kernel_map, (vm_offset_t)kdbip[i].kd_buffer, kd_bufsize);
+               kd_bufsize = 0;
+
+               kmem_free(kernel_map, (vm_offset_t)kdcopybuf, KDCOPYBUF_SIZE);
+               kdcopybuf = NULL;
+               
+               return(ENOMEM);
+       }
+       for (cpu = 0; cpu < kd_cpus; cpu++) {
+               kdbip[cpu].kd_bufptr = kdbip[cpu].kd_buffer;
+               kdbip[cpu].kd_buflast = &kdbip[cpu].kd_bufptr[nentries];
+               kdbip[cpu].kd_readlast = kdbip[cpu].kd_bufptr;
+       }
+       kdebug_flags |= KDBG_BUFINIT;
+
+       return(0);
+}
+
+
+static void
+delete_buffers(void)
 {
+        unsigned int   cpu;
+
+       if (kd_bufsize && (kdebug_flags & KDBG_BUFINIT)) {
+               for (cpu = 0; cpu < kd_cpus; cpu++)
+                       kmem_free(kernel_map, (vm_offset_t)kdbip[cpu].kd_buffer, kd_bufsize);
+               kd_bufsize = 0;
+       }
+       if (kdcopybuf) {
+               kmem_free(kernel_map, (vm_offset_t)kdcopybuf, KDCOPYBUF_SIZE);
+               kdcopybuf = NULL;
+       }
+       kdebug_flags &= ~KDBG_BUFINIT;
+}
+
+
+static void
+kernel_debug_internal(unsigned int debugid, unsigned int arg1, unsigned int arg2, unsigned int arg3,
+                     unsigned int arg4, unsigned int arg5, int entropy_flag)
+{
+       int s;
        kd_buf * kd;
        struct proc *curproc;
-       int      s;
        unsigned long long now;
+       int cpu;
+       
+       s = ml_set_interrupts_enabled(FALSE);
 
+       now = mach_absolute_time() & KDBG_TIMESTAMP_MASK;
+       cpu = cpu_number();
 
        if (kdebug_enable & KDEBUG_ENABLE_CHUD) {
            if (kdebug_chudhook)
                kdebug_chudhook(debugid, arg1, arg2, arg3, arg4, arg5);
 
            if ( !(kdebug_enable & (KDEBUG_ENABLE_ENTROPY | KDEBUG_ENABLE_TRACE)))
-               return;
+               goto out;
        }
-       s = ml_set_interrupts_enabled(FALSE);
-       lck_spin_lock(kd_trace_lock);
 
        if (kdebug_slowcheck == 0)
            goto record_trace;
 
-       if (kdebug_enable & KDEBUG_ENABLE_ENTROPY)
+       if (entropy_flag && (kdebug_enable & KDEBUG_ENABLE_ENTROPY))
          {
            if (kd_entropy_indx < kd_entropy_count)
              {
@@ -181,46 +300,30 @@ unsigned int debugid, arg1, arg2, arg3, arg4, arg5;
          }
 
        if ( (kdebug_slowcheck & SLOW_NOLOG) )
-         {
-           lck_spin_unlock(kd_trace_lock);
-           ml_set_interrupts_enabled(s);
-           return;
-         }
+           goto out;
        
        if (kdebug_flags & KDBG_PIDCHECK)
          {
            /* If kdebug flag is not set for current proc, return  */
            curproc = current_proc();
-           if ((curproc && !(curproc->p_flag & P_KDEBUG)) &&
+           if ((curproc && !(curproc->p_kdebug)) &&
                ((debugid&0xffff0000) != (MACHDBG_CODE(DBG_MACH_SCHED, 0) | DBG_FUNC_NONE)))
-             {
-               lck_spin_unlock(kd_trace_lock);
-               ml_set_interrupts_enabled(s);
-               return;
-             }
+               goto out;
          }
        else if (kdebug_flags & KDBG_PIDEXCLUDE)
          {
            /* If kdebug flag is set for current proc, return  */
            curproc = current_proc();
-           if ((curproc && (curproc->p_flag & P_KDEBUG)) &&
+           if ((curproc && curproc->p_kdebug) &&
                ((debugid&0xffff0000) != (MACHDBG_CODE(DBG_MACH_SCHED, 0) | DBG_FUNC_NONE)))
-             {
-               lck_spin_unlock(kd_trace_lock);
-               ml_set_interrupts_enabled(s);
-               return;
-             }
+               goto out;
          }
 
        if (kdebug_flags & KDBG_RANGECHECK)
          {
-           if ((debugid < kdlog_beg) || (debugid >= kdlog_end) 
-               && (debugid >> 24 != DBG_TRACE))
-             {
-               lck_spin_unlock(kd_trace_lock);
-               ml_set_interrupts_enabled(s);
-               return;
-             }
+           if ((debugid < kdlog_beg)
+               || ((debugid >= kdlog_end) && (debugid >> 24 != DBG_TRACE)))
+               goto out;
          }
        else if (kdebug_flags & KDBG_VALCHECK)
          {
@@ -229,234 +332,114 @@ unsigned int debugid, arg1, arg2, arg3, arg4, arg5;
                (debugid & DBG_FUNC_MASK) != kdlog_value3 &&
                (debugid & DBG_FUNC_MASK) != kdlog_value4 &&
                (debugid >> 24 != DBG_TRACE))
-             {
-               lck_spin_unlock(kd_trace_lock);
-               ml_set_interrupts_enabled(s);
-               return;
-             }
+               goto out;
          }
 
 record_trace:
-       kd = kd_bufptr;
+       kd = kdbip[cpu].kd_bufptr;
        kd->debugid = debugid;
        kd->arg1 = arg1;
        kd->arg2 = arg2;
        kd->arg3 = arg3;
        kd->arg4 = arg4;
-       kd->arg5 = (int)current_thread();
+       kd->arg5 = arg5;
                  
-       now = mach_absolute_time() & KDBG_TIMESTAMP_MASK;
-
-       /* Watch for out of order timestamps */ 
-
-       if (now < kd_prev_timebase)
+       /*
+        * Watch for out of order timestamps
+        */     
+       if (now < kdbip[cpu].kd_prev_timebase)
          {
-           now = ++kd_prev_timebase & KDBG_TIMESTAMP_MASK;
+           /*
+            * if so, just store the previous timestamp + a cycle
+            */
+           now = ++kdbip[cpu].kd_prev_timebase & KDBG_TIMESTAMP_MASK;
          }
        else
          {
-           /* Then just store the previous timestamp */
-           kd_prev_timebase = now;
+           kdbip[cpu].kd_prev_timebase = now;
          }
-       kd->timestamp = now | (((uint64_t)cpu_number()) << KDBG_CPU_SHIFT);
+       kd->timestamp = now | (((uint64_t)cpu) << KDBG_CPU_SHIFT);
 
-       kd_bufptr++;
+       kdbip[cpu].kd_bufptr++;
 
-       if (kd_bufptr >= kd_buflast)
-               kd_bufptr = kd_buffer;
-       if (kd_bufptr == kd_readlast) {
+       if (kdbip[cpu].kd_bufptr >= kdbip[cpu].kd_buflast)
+               kdbip[cpu].kd_bufptr = kdbip[cpu].kd_buffer;
+
+       if (kdbip[cpu].kd_bufptr == kdbip[cpu].kd_readlast) {
                if (kdebug_flags & KDBG_NOWRAP)
                        kdebug_slowcheck |= SLOW_NOLOG;
+               kdbip[cpu].kd_wrapped = 1;
                kdebug_flags |= KDBG_WRAPPED;
        }
-       lck_spin_unlock(kd_trace_lock);
+
+out:
        ml_set_interrupts_enabled(s);
 }
 
 void
-kernel_debug1(debugid, arg1, arg2, arg3, arg4, arg5)
-unsigned int debugid, arg1, arg2, arg3, arg4, arg5;
+kernel_debug(unsigned int debugid, unsigned int arg1, unsigned int arg2, unsigned int arg3,
+                     unsigned int arg4, __unused unsigned int arg5)
 {
-       kd_buf * kd;
-       struct proc *curproc;
-       int      s;
-       unsigned long long now;
-
-       if (kdebug_enable & KDEBUG_ENABLE_CHUD) {
-           if (kdebug_chudhook)
-               (void)kdebug_chudhook(debugid, arg1, arg2, arg3, arg4, arg5);
-
-           if ( !(kdebug_enable & (KDEBUG_ENABLE_ENTROPY | KDEBUG_ENABLE_TRACE)))
-               return;
-       }
-       s = ml_set_interrupts_enabled(FALSE);
-       lck_spin_lock(kd_trace_lock);
-
-       if (kdebug_slowcheck == 0)
-           goto record_trace1;
+       kernel_debug_internal(debugid, arg1, arg2, arg3, arg4, (int)current_thread(), 1);
+}
 
-       if ( (kdebug_slowcheck & SLOW_NOLOG) )
-         {
-           lck_spin_unlock(kd_trace_lock);
-           ml_set_interrupts_enabled(s);
-           return;
-         }
+void
+kernel_debug1(unsigned int debugid, unsigned int arg1, unsigned int arg2, unsigned int arg3,
+                     unsigned int arg4, unsigned int arg5)
+{
+       kernel_debug_internal(debugid, arg1, arg2, arg3, arg4, arg5, 0);
+}
 
-       if (kdebug_flags & KDBG_PIDCHECK)
-         {
-           /* If kdebug flag is not set for current proc, return  */
-           curproc = current_proc();
-           if ((curproc && !(curproc->p_flag & P_KDEBUG)) &&
-               ((debugid&0xffff0000) != (MACHDBG_CODE(DBG_MACH_SCHED, 0) | DBG_FUNC_NONE)))
-             {
-               lck_spin_unlock(kd_trace_lock);
-               ml_set_interrupts_enabled(s);
-               return;
-             }
-         }
-       else if (kdebug_flags & KDBG_PIDEXCLUDE)
-         {
-           /* If kdebug flag is set for current proc, return  */
-           curproc = current_proc();
-           if ((curproc && (curproc->p_flag & P_KDEBUG)) &&
-               ((debugid&0xffff0000) != (MACHDBG_CODE(DBG_MACH_SCHED, 0) | DBG_FUNC_NONE)))
-             {
-               lck_spin_unlock(kd_trace_lock);
-               ml_set_interrupts_enabled(s);
-               return;
-             }
-         }
+static void
+kdbg_lock_init(void)
+{
+       host_basic_info_data_t hinfo;
+       mach_msg_type_number_t count = HOST_BASIC_INFO_COUNT;
 
-       if (kdebug_flags & KDBG_RANGECHECK)
-         {
-           if ((debugid < kdlog_beg) || (debugid >= kdlog_end)
-               && (debugid >> 24 != DBG_TRACE))
-             {
-               lck_spin_unlock(kd_trace_lock);
-               ml_set_interrupts_enabled(s);
+       if (kdebug_flags & KDBG_LOCKINIT)
                return;
-             }
-         }
-       else if (kdebug_flags & KDBG_VALCHECK)
-         {
-           if ((debugid & DBG_FUNC_MASK) != kdlog_value1 &&
-               (debugid & DBG_FUNC_MASK) != kdlog_value2 &&
-               (debugid & DBG_FUNC_MASK) != kdlog_value3 &&
-               (debugid & DBG_FUNC_MASK) != kdlog_value4 &&
-               (debugid >> 24 != DBG_TRACE))
-             {
-               lck_spin_unlock(kd_trace_lock);
-               ml_set_interrupts_enabled(s);
-               return;
-             }
-         }
-
-record_trace1:
-       kd = kd_bufptr;
-       kd->debugid = debugid;
-       kd->arg1 = arg1;
-       kd->arg2 = arg2;
-       kd->arg3 = arg3;
-       kd->arg4 = arg4;
-       kd->arg5 = arg5;
-
-       now = mach_absolute_time() & KDBG_TIMESTAMP_MASK;
-
-       /* Watch for out of order timestamps */ 
-
-       if (now < kd_prev_timebase)
-         {
-           now = ++kd_prev_timebase & KDBG_TIMESTAMP_MASK;
-         }
-       else
-         {
-           /* Then just store the previous timestamp */
-           kd_prev_timebase = now;
-         }
-       kd->timestamp = now | (((uint64_t)cpu_number()) << KDBG_CPU_SHIFT);
-
-       kd_bufptr++;
-
-       if (kd_bufptr >= kd_buflast)
-               kd_bufptr = kd_buffer;
-       if (kd_bufptr == kd_readlast) {
-               if (kdebug_flags & KDBG_NOWRAP)
-                       kdebug_slowcheck |= SLOW_NOLOG;
-               kdebug_flags |= KDBG_WRAPPED;
-       }
-       lck_spin_unlock(kd_trace_lock);
-       ml_set_interrupts_enabled(s);
-}
 
+       /* get the number of cpus and cache it */
+#define BSD_HOST 1
+       host_info((host_t)BSD_HOST, HOST_BASIC_INFO, (host_info_t)&hinfo, &count);
+       kd_cpus = hinfo.physical_cpu_max;
 
-static void
-kdbg_lock_init()
-{
+       if (kmem_alloc(kernel_map, (unsigned int *)&kdbip,
+                      sizeof(struct kd_bufinfo) * kd_cpus) != KERN_SUCCESS)
+               return;
 
-        if (kdebug_flags & KDBG_LOCKINIT)
-                return;
-        /*
+       /*
         * allocate lock group attribute and group
         */
-        kd_trace_lock_grp_attr = lck_grp_attr_alloc_init();
-       //lck_grp_attr_setstat(kd_trace_lock_grp_attr);
-       kd_trace_lock_grp = lck_grp_alloc_init("kdebug", kd_trace_lock_grp_attr);
-               
-        kd_trace_mtx_grp_attr = lck_grp_attr_alloc_init();
-       //lck_grp_attr_setstat(kd_trace_mtx_grp_attr);
-       kd_trace_mtx_grp = lck_grp_alloc_init("kdebug", kd_trace_mtx_grp_attr);
+       kd_trace_mtx_sysctl_grp_attr = lck_grp_attr_alloc_init();
+       kd_trace_mtx_sysctl_grp = lck_grp_alloc_init("kdebug", kd_trace_mtx_sysctl_grp_attr);
                
        /*
         * allocate the lock attribute
         */
-       kd_trace_lock_attr = lck_attr_alloc_init();
-       //lck_attr_setdebug(kd_trace_lock_attr);
-
-       kd_trace_mtx_attr = lck_attr_alloc_init();
-       //lck_attr_setdebug(kd_trace_mtx_attr);
+       kd_trace_mtx_sysctl_attr = lck_attr_alloc_init();
 
 
        /*
         * allocate and initialize spin lock and mutex
         */
-       kd_trace_lock  = lck_spin_alloc_init(kd_trace_lock_grp, kd_trace_lock_attr);
-       kd_trace_mtx   = lck_mtx_alloc_init(kd_trace_mtx_grp, kd_trace_mtx_attr);
+       kd_trace_mtx_sysctl = lck_mtx_alloc_init(kd_trace_mtx_sysctl_grp, kd_trace_mtx_sysctl_attr);
 
        kdebug_flags |= KDBG_LOCKINIT;
 }
 
 
 int
-kdbg_bootstrap()
+kdbg_bootstrap(void)
 {
+        kdebug_flags &= ~KDBG_WRAPPED;
 
-       kd_bufsize = nkdbufs * sizeof(kd_buf);
-
-       if (kmem_alloc(kernel_map, &kd_buftomem,
-                             (vm_size_t)kd_bufsize) == KERN_SUCCESS) 
-           kd_buffer = (kd_buf *) kd_buftomem;
-       else
-           kd_buffer= (kd_buf *) 0;
-       kdebug_flags &= ~KDBG_WRAPPED;
-
-       if (kd_buffer) {
-               kdebug_flags |= (KDBG_INIT | KDBG_BUFINIT);
-               kd_bufptr = kd_buffer;
-               kd_buflast = &kd_bufptr[nkdbufs];
-               kd_readlast = kd_bufptr;
-               kd_prev_timebase = 0LL;
-               return(0);
-       } else {
-               kd_bufsize=0;
-               kdebug_flags &= ~(KDBG_INIT | KDBG_BUFINIT);
-               return(EINVAL);
-       }
-       
+       return (create_buffers());
 }
 
-kdbg_reinit()
+int
+kdbg_reinit(void)
 {
-    int s;
     int ret=0;
 
     /*
@@ -464,17 +447,18 @@ kdbg_reinit()
      * First make sure we're not in
      * the middle of cutting a trace
      */
-    s = ml_set_interrupts_enabled(FALSE);
-    lck_spin_lock(kd_trace_lock);
 
     kdebug_enable &= ~KDEBUG_ENABLE_TRACE;
     kdebug_slowcheck |= SLOW_NOLOG;
 
-    lck_spin_unlock(kd_trace_lock);
-    ml_set_interrupts_enabled(s);
+    /*
+     * make sure the SLOW_NOLOG is seen
+     * by everyone that might be trying
+     * to cut a trace..
+     */
+    IOSleep(100);
 
-    if ((kdebug_flags & KDBG_INIT) && (kdebug_flags & KDBG_BUFINIT) && kd_bufsize && kd_buffer)
-        kmem_free(kernel_map, (vm_offset_t)kd_buffer, kd_bufsize);
+    delete_buffers();
 
     if ((kdebug_flags & KDBG_MAPINIT) && kd_mapsize && kd_mapptr)
       {
@@ -485,12 +469,13 @@ kdbg_reinit()
        kd_mapcount = 0;
       }  
 
-    ret= kdbg_bootstrap();
+    ret = kdbg_bootstrap();
 
     return(ret);
 }
 
-void kdbg_trace_data(struct proc *proc, long *arg_pid)
+void
+kdbg_trace_data(struct proc *proc, long *arg_pid)
 {
     if (!proc)
         *arg_pid = 0;
@@ -501,13 +486,12 @@ void kdbg_trace_data(struct proc *proc, long *arg_pid)
 }
 
 
-void kdbg_trace_string(struct proc *proc, long *arg1, long *arg2, long *arg3, long *arg4)
+void
+kdbg_trace_string(struct proc *proc, long *arg1, long *arg2, long *arg3, long *arg4)
 {
-    int i;
     char *dbg_nameptr; 
     int dbg_namelen;
     long dbg_parms[4];
-
     if (!proc)
       {
        *arg1 = 0;
@@ -525,15 +509,10 @@ void kdbg_trace_string(struct proc *proc, long *arg1, long *arg2, long *arg3, lo
     dbg_parms[2]=0L;
     dbg_parms[3]=0L;
   
-    if(dbg_namelen > sizeof(dbg_parms))
+    if(dbg_namelen > (int)sizeof(dbg_parms))
       dbg_namelen = sizeof(dbg_parms);
     
-    for(i=0;dbg_namelen > 0; i++)
-      {
-       dbg_parms[i]=*(long*)dbg_nameptr;
-       dbg_nameptr += sizeof(long);
-       dbg_namelen -= sizeof(long);
-      }
+    strncpy((char *)dbg_parms, dbg_nameptr, dbg_namelen);
 
     *arg1=dbg_parms[0];
     *arg2=dbg_parms[1];
@@ -542,17 +521,17 @@ void kdbg_trace_string(struct proc *proc, long *arg1, long *arg2, long *arg3, lo
 }
 
 static void
-kdbg_resolve_map(thread_t th_act, krt_t *t)
+kdbg_resolve_map(thread_t th_act, void *opaque)
 {
   kd_threadmap *mapptr;
+  krt_t *t = (krt_t *)opaque;
 
   if(t->count < t->maxcount)
     {
       mapptr=&t->map[t->count];
       mapptr->thread  = (unsigned int)th_act;
-      (void) strncpy (mapptr->command, t->atts->task_comm,
-                     sizeof(t->atts->task_comm)-1);
-      mapptr->command[sizeof(t->atts->task_comm)-1] = '\0';
+      (void) strlcpy (mapptr->command, t->atts->task_comm,
+                     sizeof(t->atts->task_comm));
 
       /*
        Some kernel threads have no associated pid.
@@ -567,7 +546,8 @@ kdbg_resolve_map(thread_t th_act, krt_t *t)
     }
 }
 
-void kdbg_mapinit()
+void
+kdbg_mapinit(void)
 {
        struct proc *p;
        struct krt akrt;
@@ -581,6 +561,9 @@ void kdbg_mapinit()
         if (kdebug_flags & KDBG_MAPINIT)
          return;
 
+       /* need to use PROC_SCANPROCLIST with proc_iterate */
+       proc_list_lock();
+
        /* Calculate the sizes of map buffers*/
        for (p = allproc.lh_first, kd_mapcount=0, tts_count=0; p; 
             p = p->p_list.le_next)
@@ -589,6 +572,8 @@ void kdbg_mapinit()
            tts_count++;
          }
 
+       proc_list_unlock();
+
        /*
         * The proc count could change during buffer allocation,
         * so introduce a small fudge factor to bump up the
@@ -626,20 +611,26 @@ void kdbg_mapinit()
         */
 
        if (tts_mapptr) {
+               /* should use proc_iterate */
+               proc_list_lock();
+
                for (p = allproc.lh_first, i=0; p && i < tts_count; 
                     p = p->p_list.le_next) {
-                       if (p->p_flag & P_WEXIT)
+                       if (p->p_lflag & P_LEXIT)
                                continue;
 
                        if (p->task) {
                                task_reference(p->task);
                                tts_mapptr[i].task = p->task;
                                tts_mapptr[i].pid  = p->p_pid;
-                               (void)strncpy(&tts_mapptr[i].task_comm, p->p_comm, sizeof(tts_mapptr[i].task_comm) - 1);
+                               (void)strlcpy(tts_mapptr[i].task_comm, p->p_comm, sizeof(tts_mapptr[i].task_comm));
                                i++;
                        }
                }
                tts_count = i;
+
+               proc_list_unlock();
+
        }
 
 
@@ -664,34 +655,31 @@ void kdbg_mapinit()
 static void
 kdbg_clear(void)
 {
-        int s;
-
         /*
         * Clean up the trace buffer
         * First make sure we're not in
         * the middle of cutting a trace
         */
-       s = ml_set_interrupts_enabled(FALSE);
-       lck_spin_lock(kd_trace_lock);
 
        kdebug_enable &= ~KDEBUG_ENABLE_TRACE;
        kdebug_slowcheck = SLOW_NOLOG;
 
+       /*
+        * make sure the SLOW_NOLOG is seen
+        * by everyone that might be trying
+        * to cut a trace..
+        */
+       IOSleep(100);
+
        if (kdebug_enable & KDEBUG_ENABLE_ENTROPY)
                kdebug_slowcheck |= SLOW_ENTROPY;
 
-       lck_spin_unlock(kd_trace_lock);
-       ml_set_interrupts_enabled(s);
-
         global_state_pid = -1;
-       kdebug_flags &= ~KDBG_BUFINIT;
        kdebug_flags &= (unsigned int)~KDBG_CKTYPES;
        kdebug_flags &= ~(KDBG_NOWRAP | KDBG_RANGECHECK | KDBG_VALCHECK);
        kdebug_flags &= ~(KDBG_PIDCHECK | KDBG_PIDEXCLUDE);
-       kmem_free(kernel_map, (vm_offset_t)kd_buffer, kd_bufsize);
-       kd_buffer = (kd_buf *)0;
-       kd_bufsize = 0;
-       kd_prev_timebase = 0LL;
+
+       delete_buffers();
 
        /* Clean up the thread map buffer */
        kdebug_flags &= ~KDBG_MAPINIT;
@@ -701,6 +689,7 @@ kdbg_clear(void)
        kd_mapcount = 0;
 }
 
+int
 kdbg_setpid(kd_regtype *kdr)
 {
   pid_t pid;
@@ -712,7 +701,7 @@ kdbg_setpid(kd_regtype *kdr)
 
   if (pid > 0)
     {
-      if ((p = pfind(pid)) == NULL)
+      if ((p = proc_find(pid)) == NULL)
        ret = ESRCH;
       else
        {
@@ -722,14 +711,15 @@ kdbg_setpid(kd_regtype *kdr)
              kdebug_flags &= ~KDBG_PIDEXCLUDE;
              kdebug_slowcheck |= SLOW_CHECKS;
 
-             p->p_flag |= P_KDEBUG;
+             p->p_kdebug = 1;
            }
          else  /* turn off pid check for this pid value */
            {
              /* Don't turn off all pid checking though */
              /* kdebug_flags &= ~KDBG_PIDCHECK;*/   
-             p->p_flag &= ~P_KDEBUG;
+             p->p_kdebug = 0;
            }
+       proc_rele(p);
        }
     }
   else
@@ -738,6 +728,7 @@ kdbg_setpid(kd_regtype *kdr)
 }
 
 /* This is for pid exclusion in the trace buffer */
+int
 kdbg_setpidex(kd_regtype *kdr)
 {
   pid_t pid;
@@ -749,7 +740,7 @@ kdbg_setpidex(kd_regtype *kdr)
 
   if (pid > 0)
     {
-      if ((p = pfind(pid)) == NULL)
+      if ((p = proc_find(pid)) == NULL)
        ret = ESRCH;
       else
        {
@@ -759,14 +750,15 @@ kdbg_setpidex(kd_regtype *kdr)
              kdebug_flags &= ~KDBG_PIDCHECK;
              kdebug_slowcheck |= SLOW_CHECKS;
 
-             p->p_flag |= P_KDEBUG;
+             p->p_kdebug = 1;
            }
          else  /* turn off pid exclusion for this pid value */
            {
              /* Don't turn off all pid exclusion though */
              /* kdebug_flags &= ~KDBG_PIDEXCLUDE;*/   
-             p->p_flag &= ~P_KDEBUG;
+             p->p_kdebug = 0;
            }
+       proc_rele(p);
        }
     }
   else
@@ -774,7 +766,8 @@ kdbg_setpidex(kd_regtype *kdr)
   return(ret);
 }
 
-/* This is for setting a minimum decrementer value */
+/* This is for setting a maximum decrementer value */
+int
 kdbg_setrtcdec(kd_regtype *kdr)
 {
   int ret=0;
@@ -783,21 +776,23 @@ kdbg_setrtcdec(kd_regtype *kdr)
   decval = (natural_t)kdr->value1;
 
   if (decval && decval < KDBG_MINRTCDEC)
-      ret = EINVAL;
+       ret = EINVAL;
 #ifdef ppc
-  else
-      rtclock_decrementer_min = decval;
+       else {
+               maxDec = decval ? decval : 0x7FFFFFFF;  /* Set or reset the max decrementer */
+       }
 #else
-  else
-    ret = ENOTSUP;
+       else
+               ret = ENOTSUP;
 #endif /* ppc */
 
   return(ret);
 }
 
+int
 kdbg_setreg(kd_regtype * kdr)
 {
-       int i,j, ret=0;
+       int ret=0;
        unsigned int val_1, val_2, val;
        switch (kdr->type) {
        
@@ -858,11 +853,13 @@ kdbg_setreg(kd_regtype * kdr)
        return(ret);
 }
 
-kdbg_getreg(kd_regtype * kdr)
+int
+kdbg_getreg(__unused kd_regtype * kdr)
 {
+#if 0  
        int i,j, ret=0;
        unsigned int val_1, val_2, val;
-#if 0  
+
        switch (kdr->type) {
        case KDBG_CLASSTYPE :
                val_1 = (kdr->value1 & 0xff);
@@ -906,7 +903,7 @@ kdbg_readmap(user_addr_t buffer, size_t *number)
 {
   int avail = *number;
   int ret = 0;
-  int count = 0;
+  unsigned int count = 0;
 
   count = avail/sizeof (kd_threadmap);
 
@@ -945,7 +942,6 @@ kdbg_getentropy (user_addr_t buffer, size_t *number, int ms_timeout)
 {
   int avail = *number;
   int ret = 0;
-  int count = 0;     /* The number of timestamp entries that will fill buffer */
 
   if (kd_entropy_buffer)
     return(EBUSY);
@@ -1005,6 +1001,22 @@ kdbg_getentropy (user_addr_t buffer, size_t *number, int ms_timeout)
 }
 
 
+static void
+kdbg_set_nkdbufs(unsigned int value)
+{
+        /*
+        * We allow a maximum buffer size of 25% of either ram or max mapped address, whichever is smaller
+        * 'value' is the desired number of trace entries
+        */
+        unsigned int max_entries = (sane_size/4) / sizeof(kd_buf);
+
+       if (value <= max_entries)
+               nkdbufs = value;
+       else
+               nkdbufs = max_entries;
+}
+
+
 /*
  * This function is provided for the CHUD toolkit only.
  *    int val:
@@ -1014,7 +1026,8 @@ kdbg_getentropy (user_addr_t buffer, size_t *number, int ms_timeout)
  *        address of the enabled kdebug_chudhook function
 */
 
-void kdbg_control_chud(int val, void *fn)
+void
+kdbg_control_chud(int val, void *fn)
 {
         if (val) {
                 /* enable chudhook */
@@ -1029,11 +1042,11 @@ void kdbg_control_chud(int val, void *fn)
 }
 
        
-kdbg_control(int *name, u_int namelen, user_addr_t where, size_t *sizep)
+int
+kdbg_control(int *name, __unused u_int namelen, user_addr_t where, size_t *sizep)
 {
-    int ret=0;
-       int size=*sizep;
-       int max_entries;
+        int ret=0;
+       size_t size=*sizep;
        unsigned int value = name[1];
        kd_regtype kd_Reg;
        kbufinfo_t kd_bufinfo;
@@ -1042,7 +1055,11 @@ kdbg_control(int *name, u_int namelen, user_addr_t where, size_t *sizep)
 
 
        kdbg_lock_init();
-       lck_mtx_lock(kd_trace_mtx);
+
+       if ( !(kdebug_flags & KDBG_LOCKINIT))
+               return(ENOMEM);
+
+       lck_mtx_lock(kd_trace_mtx_sysctl);
 
        if (name[0] == KERN_KDGETBUF) {
            /* 
@@ -1054,7 +1071,7 @@ kdbg_control(int *name, u_int namelen, user_addr_t where, size_t *sizep)
                 * There is not enough room to return even
                 * the first element of the info structure.
                 */
-               lck_mtx_unlock(kd_trace_mtx);
+               lck_mtx_unlock(kd_trace_mtx_sysctl);
 
                return(EINVAL);
            }
@@ -1073,7 +1090,7 @@ kdbg_control(int *name, u_int namelen, user_addr_t where, size_t *sizep)
                     * Provide all the info we have
                    */
                if (copyout (&kd_bufinfo, where, sizeof(kd_bufinfo))) {
-                       lck_mtx_unlock(kd_trace_mtx);
+                       lck_mtx_unlock(kd_trace_mtx_sysctl);
 
                 return(EINVAL);
                    }
@@ -1084,12 +1101,12 @@ kdbg_control(int *name, u_int namelen, user_addr_t where, size_t *sizep)
                         * as much info as there is room for.
                         */
                if (copyout (&kd_bufinfo, where, size)) {
-                       lck_mtx_unlock(kd_trace_mtx);
+                       lck_mtx_unlock(kd_trace_mtx_sysctl);
 
                        return(EINVAL);
                    }
            }
-           lck_mtx_unlock(kd_trace_mtx);
+           lck_mtx_unlock(kd_trace_mtx_sysctl);
 
            return(0);
        } else if (name[0] == KERN_KDGETENTROPY) {
@@ -1097,22 +1114,22 @@ kdbg_control(int *name, u_int namelen, user_addr_t where, size_t *sizep)
                ret = EBUSY;
            else
                ret = kdbg_getentropy(where, sizep, value);
-           lck_mtx_unlock(kd_trace_mtx);
+           lck_mtx_unlock(kd_trace_mtx_sysctl);
 
            return (ret);
        }
        
-       if (curproc = current_proc())
+       if ((curproc = current_proc()) != NULL)
            curpid = curproc->p_pid;
        else {
-           lck_mtx_unlock(kd_trace_mtx);
+           lck_mtx_unlock(kd_trace_mtx_sysctl);
 
            return (ESRCH);
        }
         if (global_state_pid == -1)
            global_state_pid = curpid;
        else if (global_state_pid != curpid) {
-           if ((p = pfind(global_state_pid)) == NULL) {
+           if ((p = proc_find(global_state_pid)) == NULL) {
                /*
                 * The global pid no longer exists
                 */
@@ -1121,7 +1138,8 @@ kdbg_control(int *name, u_int namelen, user_addr_t where, size_t *sizep)
                /*
                 * The global pid exists, deny this request
                 */
-               lck_mtx_unlock(kd_trace_mtx);
+               proc_rele(p);
+               lck_mtx_unlock(kd_trace_mtx_sysctl);
 
                return(EBUSY);
            }
@@ -1145,6 +1163,8 @@ kdbg_control(int *name, u_int namelen, user_addr_t where, size_t *sizep)
                          ret=EINVAL;
                          break;
                        }
+                     kdbg_mapinit();
+
                      kdebug_enable |= KDEBUG_ENABLE_TRACE;
                      kdebug_slowcheck &= ~SLOW_NOLOG;
                    }
@@ -1153,16 +1173,9 @@ kdbg_control(int *name, u_int namelen, user_addr_t where, size_t *sizep)
                      kdebug_enable &= ~KDEBUG_ENABLE_TRACE;
                      kdebug_slowcheck |= SLOW_NOLOG;
                    }
-                 kdbg_mapinit();
                  break;
                case KERN_KDSETBUF:
-                 /* We allow a maximum buffer size of 25% of either ram or max mapped address, whichever is smaller */
-                 /* 'value' is the desired number of trace entries */
-                       max_entries = (sane_size/4) / sizeof(kd_buf);
-                       if (value <= max_entries)
-                               nkdbufs = value;
-                       else
-                         nkdbufs = max_entries;
+                       kdbg_set_nkdbufs(value);
                        break;
                case KERN_KDSETUP:
                        ret=kdbg_reinit();
@@ -1234,109 +1247,329 @@ kdbg_control(int *name, u_int namelen, user_addr_t where, size_t *sizep)
                default:
                        ret= EINVAL;
        }
-       lck_mtx_unlock(kd_trace_mtx);
+       lck_mtx_unlock(kd_trace_mtx_sysctl);
 
        return(ret);
 }
 
+
+/*
+ * This code can run concurrently with kernel_debug_internal()
+ * without the need of any locks, because all reads of kd_bufptr[i],
+ * which get modified by kernel_debug_internal(), are safe.
+ */
+int
 kdbg_read(user_addr_t buffer, size_t *number)
 {
-int avail=*number;
-int count=0;
-int copycount=0;
-int totalcount=0;
-int s;
-unsigned int my_kdebug_flags;
-kd_buf * my_kd_bufptr;
+       unsigned int count;
+       unsigned int cpu;
+       int mincpu;
+       uint64_t  mintime, t, last_wrap_time;
+       int last_wrap_cpu;
+       int error = 0;
+       kd_buf *tempbuf;
+       uint32_t tempbuf_count;
+       uint32_t tempbuf_number;
+       unsigned int old_kdebug_flags, new_kdebug_flags;
+       unsigned int old_kdebug_slowcheck, new_kdebug_slowcheck;
+       boolean_t first_event = TRUE;
+
+       count = *number/sizeof(kd_buf);
+       *number = 0;
+
+       if (count == 0 || !(kdebug_flags & KDBG_BUFINIT) || kdcopybuf == 0)
+               return EINVAL;
 
-       s = ml_set_interrupts_enabled(FALSE);
-       lck_spin_lock(kd_trace_lock);
+       /*
+        * because we hold kd_trace_mtx_sysctl, no other control threads can 
+        * be playing with kdebug_flags... the code that cuts new events could
+        * be running, but it only reads kdebug_flags, it doesn't write it..
+        * use an OSCompareAndSwap to make sure the other processors see the 
+        * change of state immediately, not to protect against 2 threads racing to update it
+        */
+       old_kdebug_slowcheck = kdebug_slowcheck;
+       do {
+               old_kdebug_flags = kdebug_flags;
+               new_kdebug_flags = old_kdebug_flags & ~KDBG_WRAPPED;
+               new_kdebug_flags |= KDBG_NOWRAP;
+       } while ( !OSCompareAndSwap((UInt32)old_kdebug_flags, (UInt32)new_kdebug_flags, (UInt32 *)&kdebug_flags));
+
+       last_wrap_time = 0;
+       last_wrap_cpu  = -1;
+
+       for (cpu = 0; cpu < kd_cpus; cpu++) {
+               kd_buf *cur_bufptr;
+               
+               if ((cur_bufptr = kdbip[cpu].kd_bufptr) >= kdbip[cpu].kd_buflast)
+                       cur_bufptr = kdbip[cpu].kd_buffer;
 
-       my_kdebug_flags = kdebug_flags;
-       my_kd_bufptr = kd_bufptr;
+               if (kdbip[cpu].kd_wrapped) {
+                       kdbip[cpu].kd_wrapped = 0;
+                       kdbip[cpu].kd_readlast = cur_bufptr;
+                       kdbip[cpu].kd_stop = cur_bufptr;
 
-       lck_spin_unlock(kd_trace_lock);
-       ml_set_interrupts_enabled(s);
+                       if (kd_cpus > 1 && ((cur_bufptr->timestamp & KDBG_TIMESTAMP_MASK) > last_wrap_time)) {
+                               last_wrap_time = cur_bufptr->timestamp & KDBG_TIMESTAMP_MASK;
+                               last_wrap_cpu = cpu;
+                       }
+               } else {
+                       if (kdbip[cpu].kd_readlast == cur_bufptr)
+                               kdbip[cpu].kd_stop = 0;
+                       else
+                               kdbip[cpu].kd_stop = cur_bufptr;
+               }
+       }
+       if (count > nkdbufs)
+               count = nkdbufs;
 
-       count = avail/sizeof(kd_buf);
+       if ((tempbuf_count = count) > KDCOPYBUF_COUNT)
+               tempbuf_count = KDCOPYBUF_COUNT;
 
-       if (count) {
-               if ((my_kdebug_flags & KDBG_BUFINIT) && kd_bufsize && kd_buffer) {
-                       if (count > nkdbufs)
-                               count = nkdbufs;
-                       
-                       if (!(my_kdebug_flags & KDBG_WRAPPED)) {
-                               if (my_kd_bufptr == kd_readlast) {
-                                       *number = 0;
-                                       return(0);
-                               }       
-                               if (my_kd_bufptr > kd_readlast) {
-                                       copycount = my_kd_bufptr - kd_readlast;
-                                       if (copycount > count)
-                                               copycount = count;
-
-                                       if (copyout(kd_readlast, buffer, copycount * sizeof(kd_buf))) {
-                                               *number = 0;
-                                               return(EINVAL);
-                                       }
-                                       kd_readlast += copycount;
-                                       *number = copycount;
-                                       return(0);
+       if (last_wrap_cpu == -1)
+               first_event = FALSE;
+
+       while (count) {
+               tempbuf = kdcopybuf;
+               tempbuf_number = 0;
+
+               while (tempbuf_count) {
+                       mintime = 0xffffffffffffffffULL; /* all actual timestamps are below */
+                       mincpu = -1;
+
+                       for (cpu = 0; cpu < kd_cpus; cpu++) {
+                               if (kdbip[cpu].kd_stop == 0)     /* empty buffer */
+                                       continue;
+                               t = kdbip[cpu].kd_readlast[0].timestamp & KDBG_TIMESTAMP_MASK;
+
+                               if (t < mintime) {
+                                       mintime = t;
+                                       mincpu = cpu;
                                }
                        }
-                       if ( (my_kdebug_flags & KDBG_WRAPPED) ) {
-                               /* Note that by setting kd_readlast equal to my_kd_bufptr,
-                                * we now treat the kd_buffer read the same as if we weren't
-                                * wrapped and my_kd_bufptr was less than kd_readlast.
+                       if (mincpu < 0)
+                               /*
+                                * all buffers ran empty early
                                 */
-                               kd_readlast = my_kd_bufptr;
-                               kdebug_flags &= ~KDBG_WRAPPED;
+                               break;
+
+                        if (first_event == TRUE) {
+                               /*
+                                * make sure we leave room for the
+                                * LAST_WRAPPER event we inject
+                                * by throwing away the first event
+                                * it's better to lose that one
+                                * than the last one
+                                */     
+                               first_event = FALSE;
+
+                               kdbip[mincpu].kd_readlast++;
+
+                               if (kdbip[mincpu].kd_readlast == kdbip[mincpu].kd_buflast)
+                                       kdbip[mincpu].kd_readlast = kdbip[mincpu].kd_buffer;
+                               if (kdbip[mincpu].kd_readlast == kdbip[mincpu].kd_stop)
+                                       kdbip[mincpu].kd_stop = 0;
+
+                               continue;
                        }
-                       /*
-                        * first copyout from readlast to end of kd_buffer
-                        */
-                       copycount = kd_buflast - kd_readlast;
-                       if (copycount > count)
-                               copycount = count;
-                       if (copyout(kd_readlast, buffer, copycount * sizeof(kd_buf))) {
-                               *number = 0;
-                               return(EINVAL);
-                       }
-                       buffer += (copycount * sizeof(kd_buf));
-                       count -= copycount;
-                       totalcount = copycount;
-                       kd_readlast += copycount;
-
-                       if (kd_readlast == kd_buflast)
-                               kd_readlast = kd_buffer;
-                       if (count == 0) {
-                               *number = totalcount;
-                               return(0);
+                       if (last_wrap_cpu == mincpu) {
+                               tempbuf->debugid = MISCDBG_CODE(DBG_BUFFER, 0) | DBG_FUNC_NONE;
+                               tempbuf->arg1 = kd_bufsize / sizeof(kd_buf);
+                               tempbuf->arg2 = kd_cpus;
+                               tempbuf->arg3 = 0;
+                               tempbuf->arg4 = 0;
+                               tempbuf->arg5 = (int)current_thread();
+                 
+                               tempbuf->timestamp = last_wrap_time | (((uint64_t)last_wrap_cpu) << KDBG_CPU_SHIFT);
+
+                               tempbuf++;
+
+                               last_wrap_cpu = -1;
+
+                       } else {
+                               *(tempbuf++) = kdbip[mincpu].kd_readlast[0];
+
+                               kdbip[mincpu].kd_readlast++;
+
+                               if (kdbip[mincpu].kd_readlast == kdbip[mincpu].kd_buflast)
+                                       kdbip[mincpu].kd_readlast = kdbip[mincpu].kd_buffer;
+                               if (kdbip[mincpu].kd_readlast == kdbip[mincpu].kd_stop)
+                                       kdbip[mincpu].kd_stop = 0;
                        }
-                       /* second copyout from top of kd_buffer to bufptr */
-                       copycount = my_kd_bufptr - kd_readlast;
-                       if (copycount > count)
-                               copycount = count;
-                       if (copycount == 0) {
-                               *number = totalcount;
-                               return(0);
+                       tempbuf_count--;
+                       tempbuf_number++;
+               }
+               if (tempbuf_number) {
+                       if ((error = copyout(kdcopybuf, buffer, tempbuf_number * sizeof(kd_buf)))) {
+                               *number = 0;
+                               error = EINVAL;
+                               break;
                        }
-                       if (copyout(kd_readlast, buffer, copycount * sizeof(kd_buf)))
-                               return(EINVAL);
-
-                       kd_readlast += copycount;
-                       totalcount += copycount;
-                       *number = totalcount;
-                       return(0);
+                       count   -= tempbuf_number;
+                       *number += tempbuf_number;
+                       buffer  += (tempbuf_number * sizeof(kd_buf));
+               }
+               if (tempbuf_count)
+                      /*
+                       * all trace buffers are empty
+                       */
+                       break;
 
-               } /* end if KDBG_BUFINIT */             
-       } /* end if count */
-       return (EINVAL);
+               if ((tempbuf_count = count) > KDCOPYBUF_COUNT)
+                       tempbuf_count = KDCOPYBUF_COUNT;
+       }
+       if ( !(old_kdebug_flags & KDBG_NOWRAP)) {
+               do {
+                       old_kdebug_flags = kdebug_flags;
+                       new_kdebug_flags = old_kdebug_flags & ~KDBG_NOWRAP;
+               } while ( !OSCompareAndSwap((UInt32)old_kdebug_flags, (UInt32)new_kdebug_flags, (UInt32 *)&kdebug_flags));
+
+               if ( !(old_kdebug_slowcheck & SLOW_NOLOG)) {
+                       do {
+                               old_kdebug_slowcheck = kdebug_slowcheck;
+                               new_kdebug_slowcheck = old_kdebug_slowcheck & ~SLOW_NOLOG;
+                       } while ( !OSCompareAndSwap((UInt32)old_kdebug_slowcheck, (UInt32)new_kdebug_slowcheck, (UInt32 *)&kdebug_slowcheck));
+               }
+       }
+       return (error);
 }
 
+
 unsigned char *getProcName(struct proc *proc);
 unsigned char *getProcName(struct proc *proc) {
 
        return (unsigned char *) &proc->p_comm; /* Return pointer to the proc name */
 
 }
+
+#define STACKSHOT_SUBSYS_LOCK() lck_mtx_lock(&stackshot_subsys_mutex)
+#define STACKSHOT_SUBSYS_UNLOCK() lck_mtx_unlock(&stackshot_subsys_mutex)
+#ifdef __i386__
+#define TRAP_DEBUGGER __asm__ volatile("int3");
+#endif
+#ifdef __ppc__
+#define TRAP_DEBUGGER __asm__ volatile("tw 4,r3,r3");
+#endif
+
+#define SANE_TRACEBUF_SIZE 2*1024*1024
+
+/* Initialize the mutex governing access to the stack snapshot subsystem */
+__private_extern__ void
+stackshot_lock_init( void )
+{
+       stackshot_subsys_lck_grp_attr = lck_grp_attr_alloc_init();
+
+       stackshot_subsys_lck_grp = lck_grp_alloc_init("stackshot_subsys_lock", stackshot_subsys_lck_grp_attr);
+
+       stackshot_subsys_lck_attr = lck_attr_alloc_init();
+
+       lck_mtx_init(&stackshot_subsys_mutex, stackshot_subsys_lck_grp, stackshot_subsys_lck_attr);
+}
+
+/*
+ * stack_snapshot:   Obtains a coherent set of stack traces for all threads
+ *                  on the system, tracing both kernel and user stacks
+ *                  where available. Uses machine specific trace routines
+ *                  for ppc, ppc64 and x86.
+ * Inputs:          uap->pid - process id of process to be traced, or -1
+ *                  for the entire system
+ *                  uap->tracebuf - address of the user space destination
+ *                  buffer 
+ *                  uap->tracebuf_size - size of the user space trace buffer
+ *                  uap->options - various options, including the maximum
+ *                  number of frames to trace.
+ * Outputs:         EPERM if the caller is not privileged
+ *                  EINVAL if the supplied trace buffer isn't sanely sized
+ *                  ENOMEM if we don't have enough memory to satisfy the
+ *                  request
+ *                  ENOENT if the target pid isn't found
+ *                  ENOSPC if the supplied buffer is insufficient
+ *                  *retval contains the number of bytes traced, if successful
+ *                  and -1 otherwise. If the request failed due to
+ *                  tracebuffer exhaustion, we copyout as much as possible.
+ */
+int
+stack_snapshot(struct proc *p, register struct stack_snapshot_args *uap, register_t *retval) {
+       int error = 0;
+
+       if ((error = suser(kauth_cred_get(), &p->p_acflag)))
+                return(error);
+
+       return stack_snapshot2(uap->pid, uap->tracebuf, uap->tracebuf_size,
+           uap->options, retval);
+}
+
+int
+stack_snapshot2(pid_t pid, user_addr_t tracebuf, uint32_t tracebuf_size, uint32_t options, register_t *retval)
+{
+       int error = 0;
+       unsigned bytesTraced = 0;
+
+       *retval = -1;
+/* Serialize tracing */        
+       STACKSHOT_SUBSYS_LOCK();
+       
+       if ((tracebuf_size <= 0) || (tracebuf_size > SANE_TRACEBUF_SIZE)) {
+               error = EINVAL;
+               goto error_exit;
+       }
+
+       MALLOC(stackshot_snapbuf, void *, tracebuf_size, M_TEMP, M_WAITOK);
+
+       if (stackshot_snapbuf == NULL) {
+               error = ENOMEM;
+               goto error_exit;
+       }
+/* Preload trace parameters*/  
+       kdp_snapshot_preflight(pid, stackshot_snapbuf, tracebuf_size, options);
+
+/* Trap to the debugger to obtain a coherent stack snapshot; this populates
+ * the trace buffer
+ */
+       if (panic_active()) {
+               error = ENOMEM;
+               goto error_exit;
+       }
+
+       TRAP_DEBUGGER;
+
+       bytesTraced = kdp_stack_snapshot_bytes_traced();
+                       
+       if (bytesTraced > 0) {
+               if ((error = copyout(stackshot_snapbuf, tracebuf,
+                       ((bytesTraced < tracebuf_size) ?
+                           bytesTraced : tracebuf_size))))
+                       goto error_exit;
+               *retval = bytesTraced;
+       }
+       else {
+               error = ENOENT;
+               goto error_exit;
+       }
+
+       error = kdp_stack_snapshot_geterror();
+       if (error == -1) {
+               error = ENOSPC;
+               *retval = -1;
+               goto error_exit;
+       }
+
+error_exit:
+       if (stackshot_snapbuf != NULL)
+               FREE(stackshot_snapbuf, M_TEMP);
+       stackshot_snapbuf = NULL;
+       STACKSHOT_SUBSYS_UNLOCK();
+       return error;
+}
+
+void
+start_kern_tracing(unsigned int new_nkdbufs) {
+       if (!new_nkdbufs)
+               return;
+       kdbg_set_nkdbufs(new_nkdbufs);
+       kdbg_lock_init();
+       kdbg_reinit();
+       kdebug_enable |= KDEBUG_ENABLE_TRACE;
+       kdebug_slowcheck &= ~SLOW_NOLOG;
+       kdbg_mapinit();
+       printf("kernel tracing started\n");
+}