nfsd/create race fixes, infrastructure
[safe/jmp/linux-2.6] / kernel / trace / ftrace.c
index f6d9af3..2f32969 100644 (file)
 #include <linux/stop_machine.h>
 #include <linux/clocksource.h>
 #include <linux/kallsyms.h>
-#include <linux/kthread.h>
+#include <linux/seq_file.h>
+#include <linux/debugfs.h>
 #include <linux/hardirq.h>
+#include <linux/kthread.h>
+#include <linux/uaccess.h>
+#include <linux/kprobes.h>
 #include <linux/ftrace.h>
-#include <linux/module.h>
 #include <linux/sysctl.h>
-#include <linux/hash.h>
+#include <linux/ctype.h>
 #include <linux/list.h>
 
+#include <asm/ftrace.h>
+
 #include "trace.h"
 
-#ifdef CONFIG_DYNAMIC_FTRACE
-# define FTRACE_ENABLED_INIT 1
-#else
-# define FTRACE_ENABLED_INIT 0
-#endif
+#define FTRACE_WARN_ON(cond)                   \
+       do {                                    \
+               if (WARN_ON(cond))              \
+                       ftrace_kill();          \
+       } while (0)
+
+#define FTRACE_WARN_ON_ONCE(cond)              \
+       do {                                    \
+               if (WARN_ON_ONCE(cond))         \
+                       ftrace_kill();          \
+       } while (0)
+
+/* ftrace_enabled is a method to turn ftrace on or off */
+int ftrace_enabled __read_mostly;
+static int last_ftrace_enabled;
 
-int ftrace_enabled = FTRACE_ENABLED_INIT;
-static int last_ftrace_enabled = FTRACE_ENABLED_INIT;
+/* set when tracing only a pid */
+struct pid *ftrace_pid_trace;
+static struct pid * const ftrace_swapper_pid = &init_struct_pid;
+
+/* Quick disabling of function tracer. */
+int function_trace_stop;
+
+/*
+ * ftrace_disabled is set when an anomaly is discovered.
+ * ftrace_disabled is much stronger than ftrace_enabled.
+ */
+static int ftrace_disabled __read_mostly;
 
 static DEFINE_SPINLOCK(ftrace_lock);
 static DEFINE_MUTEX(ftrace_sysctl_lock);
+static DEFINE_MUTEX(ftrace_start_lock);
 
 static struct ftrace_ops ftrace_list_end __read_mostly =
 {
@@ -45,11 +71,10 @@ static struct ftrace_ops ftrace_list_end __read_mostly =
 
 static struct ftrace_ops *ftrace_list __read_mostly = &ftrace_list_end;
 ftrace_func_t ftrace_trace_function __read_mostly = ftrace_stub;
+ftrace_func_t __ftrace_trace_function __read_mostly = ftrace_stub;
+ftrace_func_t ftrace_pid_function __read_mostly = ftrace_stub;
 
-/* mcount is defined per arch in assembly */
-EXPORT_SYMBOL(mcount);
-
-notrace void ftrace_list_func(unsigned long ip, unsigned long parent_ip)
+static void ftrace_list_func(unsigned long ip, unsigned long parent_ip)
 {
        struct ftrace_ops *op = ftrace_list;
 
@@ -64,6 +89,21 @@ notrace void ftrace_list_func(unsigned long ip, unsigned long parent_ip)
        };
 }
 
+static void ftrace_pid_func(unsigned long ip, unsigned long parent_ip)
+{
+       if (!test_tsk_trace_trace(current))
+               return;
+
+       ftrace_pid_function(ip, parent_ip);
+}
+
+static void set_ftrace_pid_function(ftrace_func_t func)
+{
+       /* do not set ftrace_pid_function to itself! */
+       if (func != ftrace_pid_func)
+               ftrace_pid_function = func;
+}
+
 /**
  * clear_ftrace_function - reset the ftrace function
  *
@@ -73,11 +113,27 @@ notrace void ftrace_list_func(unsigned long ip, unsigned long parent_ip)
 void clear_ftrace_function(void)
 {
        ftrace_trace_function = ftrace_stub;
+       __ftrace_trace_function = ftrace_stub;
+       ftrace_pid_function = ftrace_stub;
+}
+
+#ifndef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST
+/*
+ * For those archs that do not test ftrace_trace_stop in their
+ * mcount call site, we need to do it from C.
+ */
+static void ftrace_test_stop_func(unsigned long ip, unsigned long parent_ip)
+{
+       if (function_trace_stop)
+               return;
+
+       __ftrace_trace_function(ip, parent_ip);
 }
+#endif
 
-static int notrace __register_ftrace_function(struct ftrace_ops *ops)
+static int __register_ftrace_function(struct ftrace_ops *ops)
 {
-       /* Should never be called by interrupts */
+       /* should not be called from interrupt context */
        spin_lock(&ftrace_lock);
 
        ops->next = ftrace_list;
@@ -91,14 +147,28 @@ static int notrace __register_ftrace_function(struct ftrace_ops *ops)
        ftrace_list = ops;
 
        if (ftrace_enabled) {
+               ftrace_func_t func;
+
+               if (ops->next == &ftrace_list_end)
+                       func = ops->func;
+               else
+                       func = ftrace_list_func;
+
+               if (ftrace_pid_trace) {
+                       set_ftrace_pid_function(func);
+                       func = ftrace_pid_func;
+               }
+
                /*
                 * For one func, simply call it directly.
                 * For more than one func, call the chain.
                 */
-               if (ops->next == &ftrace_list_end)
-                       ftrace_trace_function = ops->func;
-               else
-                       ftrace_trace_function = ftrace_list_func;
+#ifdef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST
+               ftrace_trace_function = func;
+#else
+               __ftrace_trace_function = func;
+               ftrace_trace_function = ftrace_test_stop_func;
+#endif
        }
 
        spin_unlock(&ftrace_lock);
@@ -106,11 +176,12 @@ static int notrace __register_ftrace_function(struct ftrace_ops *ops)
        return 0;
 }
 
-static int notrace __unregister_ftrace_function(struct ftrace_ops *ops)
+static int __unregister_ftrace_function(struct ftrace_ops *ops)
 {
        struct ftrace_ops **p;
        int ret = 0;
 
+       /* should not be called from interrupt context */
        spin_lock(&ftrace_lock);
 
        /*
@@ -136,9 +207,19 @@ static int notrace __unregister_ftrace_function(struct ftrace_ops *ops)
 
        if (ftrace_enabled) {
                /* If we only have one func left, then call that directly */
-               if (ftrace_list == &ftrace_list_end ||
-                   ftrace_list->next == &ftrace_list_end)
-                       ftrace_trace_function = ftrace_list->func;
+               if (ftrace_list->next == &ftrace_list_end) {
+                       ftrace_func_t func = ftrace_list->func;
+
+                       if (ftrace_pid_trace) {
+                               set_ftrace_pid_function(func);
+                               func = ftrace_pid_func;
+                       }
+#ifdef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST
+                       ftrace_trace_function = func;
+#else
+                       __ftrace_trace_function = func;
+#endif
+               }
        }
 
  out:
@@ -147,20 +228,70 @@ static int notrace __unregister_ftrace_function(struct ftrace_ops *ops)
        return ret;
 }
 
+static void ftrace_update_pid_func(void)
+{
+       ftrace_func_t func;
+
+       /* should not be called from interrupt context */
+       spin_lock(&ftrace_lock);
+
+       if (ftrace_trace_function == ftrace_stub)
+               goto out;
+
+       func = ftrace_trace_function;
+
+       if (ftrace_pid_trace) {
+               set_ftrace_pid_function(func);
+               func = ftrace_pid_func;
+       } else {
+               if (func == ftrace_pid_func)
+                       func = ftrace_pid_function;
+       }
+
+#ifdef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST
+       ftrace_trace_function = func;
+#else
+       __ftrace_trace_function = func;
+#endif
+
+ out:
+       spin_unlock(&ftrace_lock);
+}
+
 #ifdef CONFIG_DYNAMIC_FTRACE
+#ifndef CONFIG_FTRACE_MCOUNT_RECORD
+# error Dynamic ftrace depends on MCOUNT_RECORD
+#endif
+
+/*
+ * Since MCOUNT_ADDR may point to mcount itself, we do not want
+ * to get it confused by reading a reference in the code as we
+ * are parsing on objcopy output of text. Use a variable for
+ * it instead.
+ */
+static unsigned long mcount_addr = MCOUNT_ADDR;
+
+enum {
+       FTRACE_ENABLE_CALLS             = (1 << 0),
+       FTRACE_DISABLE_CALLS            = (1 << 1),
+       FTRACE_UPDATE_TRACE_FUNC        = (1 << 2),
+       FTRACE_ENABLE_MCOUNT            = (1 << 3),
+       FTRACE_DISABLE_MCOUNT           = (1 << 4),
+       FTRACE_START_FUNC_RET           = (1 << 5),
+       FTRACE_STOP_FUNC_RET            = (1 << 6),
+};
 
-static struct hlist_head ftrace_hash[FTRACE_HASHSIZE];
+static int ftrace_filtered;
 
-static DEFINE_PER_CPU(int, ftrace_shutdown_disable_cpu);
+static LIST_HEAD(ftrace_new_addrs);
 
-static DEFINE_SPINLOCK(ftrace_shutdown_lock);
-static DEFINE_MUTEX(ftraced_lock);
+static DEFINE_MUTEX(ftrace_regex_lock);
 
 struct ftrace_page {
        struct ftrace_page      *next;
-       int                     index;
+       unsigned long           index;
        struct dyn_ftrace       records[];
-} __attribute__((packed));
+};
 
 #define ENTRIES_PER_PAGE \
   ((PAGE_SIZE - sizeof(struct ftrace_page)) / sizeof(struct dyn_ftrace))
@@ -171,355 +302,440 @@ struct ftrace_page {
 static struct ftrace_page      *ftrace_pages_start;
 static struct ftrace_page      *ftrace_pages;
 
-static int ftraced_trigger;
-static int ftraced_suspend;
+static struct dyn_ftrace *ftrace_free_records;
+
 
-static int ftrace_record_suspend;
+#ifdef CONFIG_KPROBES
 
-static inline int
-notrace ftrace_ip_in_hash(unsigned long ip, unsigned long key)
+static int frozen_record_count;
+
+static inline void freeze_record(struct dyn_ftrace *rec)
 {
-       struct dyn_ftrace *p;
-       struct hlist_node *t;
-       int found = 0;
+       if (!(rec->flags & FTRACE_FL_FROZEN)) {
+               rec->flags |= FTRACE_FL_FROZEN;
+               frozen_record_count++;
+       }
+}
 
-       hlist_for_each_entry(p, t, &ftrace_hash[key], node) {
-               if (p->ip == ip) {
-                       found = 1;
-                       break;
-               }
+static inline void unfreeze_record(struct dyn_ftrace *rec)
+{
+       if (rec->flags & FTRACE_FL_FROZEN) {
+               rec->flags &= ~FTRACE_FL_FROZEN;
+               frozen_record_count--;
        }
+}
 
-       return found;
+static inline int record_frozen(struct dyn_ftrace *rec)
+{
+       return rec->flags & FTRACE_FL_FROZEN;
 }
+#else
+# define freeze_record(rec)                    ({ 0; })
+# define unfreeze_record(rec)                  ({ 0; })
+# define record_frozen(rec)                    ({ 0; })
+#endif /* CONFIG_KPROBES */
 
-static inline void notrace
-ftrace_add_hash(struct dyn_ftrace *node, unsigned long key)
+static void ftrace_free_rec(struct dyn_ftrace *rec)
 {
-       hlist_add_head(&node->node, &ftrace_hash[key]);
+       rec->ip = (unsigned long)ftrace_free_records;
+       ftrace_free_records = rec;
+       rec->flags |= FTRACE_FL_FREE;
 }
 
-static notrace struct dyn_ftrace *ftrace_alloc_shutdown_node(unsigned long ip)
+void ftrace_release(void *start, unsigned long size)
 {
-       /* If this was already converted, skip it */
-       if (ftrace_ip_converted(ip))
-               return NULL;
+       struct dyn_ftrace *rec;
+       struct ftrace_page *pg;
+       unsigned long s = (unsigned long)start;
+       unsigned long e = s + size;
+       int i;
 
-       if (ftrace_pages->index == ENTRIES_PER_PAGE) {
-               if (!ftrace_pages->next)
+       if (ftrace_disabled || !start)
+               return;
+
+       /* should not be called from interrupt context */
+       spin_lock(&ftrace_lock);
+
+       for (pg = ftrace_pages_start; pg; pg = pg->next) {
+               for (i = 0; i < pg->index; i++) {
+                       rec = &pg->records[i];
+
+                       if ((rec->ip >= s) && (rec->ip < e))
+                               ftrace_free_rec(rec);
+               }
+       }
+       spin_unlock(&ftrace_lock);
+}
+
+static struct dyn_ftrace *ftrace_alloc_dyn_node(unsigned long ip)
+{
+       struct dyn_ftrace *rec;
+
+       /* First check for freed records */
+       if (ftrace_free_records) {
+               rec = ftrace_free_records;
+
+               if (unlikely(!(rec->flags & FTRACE_FL_FREE))) {
+                       FTRACE_WARN_ON_ONCE(1);
+                       ftrace_free_records = NULL;
                        return NULL;
+               }
+
+               ftrace_free_records = (void *)rec->ip;
+               memset(rec, 0, sizeof(*rec));
+               return rec;
+       }
+
+       if (ftrace_pages->index == ENTRIES_PER_PAGE) {
+               if (!ftrace_pages->next) {
+                       /* allocate another page */
+                       ftrace_pages->next =
+                               (void *)get_zeroed_page(GFP_KERNEL);
+                       if (!ftrace_pages->next)
+                               return NULL;
+               }
                ftrace_pages = ftrace_pages->next;
        }
 
        return &ftrace_pages->records[ftrace_pages->index++];
 }
 
-static void notrace
-ftrace_record_ip(unsigned long ip, unsigned long parent_ip)
+static struct dyn_ftrace *
+ftrace_record_ip(unsigned long ip)
 {
-       struct dyn_ftrace *node;
-       unsigned long flags;
-       unsigned long key;
-       int resched;
-       int atomic;
+       struct dyn_ftrace *rec;
 
-       resched = need_resched();
-       preempt_disable_notrace();
+       if (ftrace_disabled)
+               return NULL;
 
-       /* We simply need to protect against recursion */
-       __get_cpu_var(ftrace_shutdown_disable_cpu)++;
-       if (__get_cpu_var(ftrace_shutdown_disable_cpu) != 1)
-               goto out;
+       rec = ftrace_alloc_dyn_node(ip);
+       if (!rec)
+               return NULL;
 
-       if (unlikely(ftrace_record_suspend))
-               goto out;
+       rec->ip = ip;
 
-       key = hash_long(ip, FTRACE_HASHBITS);
+       list_add(&rec->list, &ftrace_new_addrs);
 
-       WARN_ON_ONCE(key >= FTRACE_HASHSIZE);
+       return rec;
+}
+
+static void print_ip_ins(const char *fmt, unsigned char *p)
+{
+       int i;
+
+       printk(KERN_CONT "%s", fmt);
+
+       for (i = 0; i < MCOUNT_INSN_SIZE; i++)
+               printk(KERN_CONT "%s%02x", i ? ":" : "", p[i]);
+}
+
+static void ftrace_bug(int failed, unsigned long ip)
+{
+       switch (failed) {
+       case -EFAULT:
+               FTRACE_WARN_ON_ONCE(1);
+               pr_info("ftrace faulted on modifying ");
+               print_ip_sym(ip);
+               break;
+       case -EINVAL:
+               FTRACE_WARN_ON_ONCE(1);
+               pr_info("ftrace failed to modify ");
+               print_ip_sym(ip);
+               print_ip_ins(" actual: ", (unsigned char *)ip);
+               printk(KERN_CONT "\n");
+               break;
+       case -EPERM:
+               FTRACE_WARN_ON_ONCE(1);
+               pr_info("ftrace faulted on writing ");
+               print_ip_sym(ip);
+               break;
+       default:
+               FTRACE_WARN_ON_ONCE(1);
+               pr_info("ftrace faulted on unknown error ");
+               print_ip_sym(ip);
+       }
+}
 
-       if (ftrace_ip_in_hash(ip, key))
-               goto out;
 
-       atomic = irqs_disabled();
+static int
+__ftrace_replace_code(struct dyn_ftrace *rec, int enable)
+{
+       unsigned long ip, fl;
+       unsigned long ftrace_addr;
 
-       spin_lock_irqsave(&ftrace_shutdown_lock, flags);
+       ftrace_addr = (unsigned long)ftrace_caller;
 
-       /* This ip may have hit the hash before the lock */
-       if (ftrace_ip_in_hash(ip, key))
-               goto out_unlock;
+       ip = rec->ip;
 
        /*
-        * There's a slight race that the ftraced will update the
-        * hash and reset here. The arch alloc is responsible
-        * for seeing if the IP has already changed, and if
-        * it has, the alloc will fail.
+        * If this record is not to be traced and
+        * it is not enabled then do nothing.
+        *
+        * If this record is not to be traced and
+        * it is enabled then disabled it.
+        *
         */
-       node = ftrace_alloc_shutdown_node(ip);
-       if (!node)
-               goto out_unlock;
+       if (rec->flags & FTRACE_FL_NOTRACE) {
+               if (rec->flags & FTRACE_FL_ENABLED)
+                       rec->flags &= ~FTRACE_FL_ENABLED;
+               else
+                       return 0;
 
-       node->ip = ip;
+       } else if (ftrace_filtered && enable) {
+               /*
+                * Filtering is on:
+                */
 
-       ftrace_add_hash(node, key);
+               fl = rec->flags & (FTRACE_FL_FILTER | FTRACE_FL_ENABLED);
 
-       ftraced_trigger = 1;
+               /* Record is filtered and enabled, do nothing */
+               if (fl == (FTRACE_FL_FILTER | FTRACE_FL_ENABLED))
+                       return 0;
 
- out_unlock:
-       spin_unlock_irqrestore(&ftrace_shutdown_lock, flags);
- out:
-       __get_cpu_var(ftrace_shutdown_disable_cpu)--;
+               /* Record is not filtered and is not enabled do nothing */
+               if (!fl)
+                       return 0;
 
-       /* prevent recursion with scheduler */
-       if (resched)
-               preempt_enable_no_resched_notrace();
-       else
-               preempt_enable_notrace();
-}
+               /* Record is not filtered but enabled, disable it */
+               if (fl == FTRACE_FL_ENABLED)
+                       rec->flags &= ~FTRACE_FL_ENABLED;
+               else
+               /* Otherwise record is filtered but not enabled, enable it */
+                       rec->flags |= FTRACE_FL_ENABLED;
+       } else {
+               /* Disable or not filtered */
 
-static struct ftrace_ops ftrace_shutdown_ops __read_mostly =
-{
-       .func = ftrace_record_ip,
-};
+               if (enable) {
+                       /* if record is enabled, do nothing */
+                       if (rec->flags & FTRACE_FL_ENABLED)
+                               return 0;
+
+                       rec->flags |= FTRACE_FL_ENABLED;
+
+               } else {
+
+                       /* if record is not enabled do nothing */
+                       if (!(rec->flags & FTRACE_FL_ENABLED))
+                               return 0;
+
+                       rec->flags &= ~FTRACE_FL_ENABLED;
+               }
+       }
 
-#define MCOUNT_ADDR ((long)(&mcount))
+       if (rec->flags & FTRACE_FL_ENABLED)
+               return ftrace_make_call(rec, ftrace_addr);
+       else
+               return ftrace_make_nop(NULL, rec, ftrace_addr);
+}
 
-static void notrace ftrace_replace_code(int saved)
+static void ftrace_replace_code(int enable)
 {
-       unsigned char *new = NULL, *old = NULL;
+       int i, failed;
        struct dyn_ftrace *rec;
        struct ftrace_page *pg;
-       unsigned long ip;
-       int failed;
-       int i;
-
-       if (saved)
-               old = ftrace_nop_replace();
-       else
-               new = ftrace_nop_replace();
 
        for (pg = ftrace_pages_start; pg; pg = pg->next) {
                for (i = 0; i < pg->index; i++) {
                        rec = &pg->records[i];
 
-                       /* don't modify code that has already faulted */
-                       if (rec->flags & FTRACE_FL_FAILED)
+                       /*
+                        * Skip over free records and records that have
+                        * failed.
+                        */
+                       if (rec->flags & FTRACE_FL_FREE ||
+                           rec->flags & FTRACE_FL_FAILED)
                                continue;
 
-                       ip = rec->ip;
-
-                       if (saved)
-                               new = ftrace_call_replace(ip, MCOUNT_ADDR);
-                       else
-                               old = ftrace_call_replace(ip, MCOUNT_ADDR);
+                       /* ignore updates to this record's mcount site */
+                       if (get_kprobe((void *)rec->ip)) {
+                               freeze_record(rec);
+                               continue;
+                       } else {
+                               unfreeze_record(rec);
+                       }
 
-                       failed = ftrace_modify_code(ip, old, new);
-                       if (failed)
+                       failed = __ftrace_replace_code(rec, enable);
+                       if (failed && (rec->flags & FTRACE_FL_CONVERTED)) {
                                rec->flags |= FTRACE_FL_FAILED;
+                               if ((system_state == SYSTEM_BOOTING) ||
+                                   !core_kernel_text(rec->ip)) {
+                                       ftrace_free_rec(rec);
+                               } else
+                                       ftrace_bug(failed, rec->ip);
+                       }
                }
        }
 }
 
-static notrace void ftrace_startup_code(void)
+static int
+ftrace_code_disable(struct module *mod, struct dyn_ftrace *rec)
 {
-       ftrace_replace_code(1);
-}
+       unsigned long ip;
+       int ret;
 
-static notrace void ftrace_shutdown_code(void)
-{
-       ftrace_replace_code(0);
+       ip = rec->ip;
+
+       ret = ftrace_make_nop(mod, rec, mcount_addr);
+       if (ret) {
+               ftrace_bug(ret, ip);
+               rec->flags |= FTRACE_FL_FAILED;
+               return 0;
+       }
+       return 1;
 }
 
-static notrace void ftrace_shutdown_replenish(void)
+static int __ftrace_modify_code(void *data)
 {
-       if (ftrace_pages->next)
-               return;
+       int *command = data;
 
-       /* allocate another page */
-       ftrace_pages->next = (void *)get_zeroed_page(GFP_KERNEL);
-}
+       if (*command & FTRACE_ENABLE_CALLS)
+               ftrace_replace_code(1);
+       else if (*command & FTRACE_DISABLE_CALLS)
+               ftrace_replace_code(0);
 
-static int notrace __ftrace_modify_code(void *data)
-{
-       void (*func)(void) = data;
+       if (*command & FTRACE_UPDATE_TRACE_FUNC)
+               ftrace_update_ftrace_func(ftrace_trace_function);
+
+       if (*command & FTRACE_START_FUNC_RET)
+               ftrace_enable_ftrace_graph_caller();
+       else if (*command & FTRACE_STOP_FUNC_RET)
+               ftrace_disable_ftrace_graph_caller();
 
-       func();
        return 0;
 }
 
-static notrace void
-ftrace_code_disable(struct dyn_ftrace *rec, unsigned long addr)
+static void ftrace_run_update_code(int command)
 {
-       unsigned long ip;
-       unsigned char *nop, *call;
-       int failed;
+       stop_machine(__ftrace_modify_code, &command, NULL);
+}
 
-       ip = rec->ip;
+static ftrace_func_t saved_ftrace_func;
+static int ftrace_start_up;
 
-       nop = ftrace_nop_replace();
-       call = ftrace_call_replace(ip, addr);
+static void ftrace_startup_enable(int command)
+{
+       if (saved_ftrace_func != ftrace_trace_function) {
+               saved_ftrace_func = ftrace_trace_function;
+               command |= FTRACE_UPDATE_TRACE_FUNC;
+       }
 
-       failed = ftrace_modify_code(ip, call, nop);
-       if (failed)
-               rec->flags |= FTRACE_FL_FAILED;
-}
+       if (!command || !ftrace_enabled)
+               return;
 
-static void notrace ftrace_run_startup_code(void)
-{
-       stop_machine_run(__ftrace_modify_code, ftrace_startup_code, NR_CPUS);
+       ftrace_run_update_code(command);
 }
 
-static void notrace ftrace_run_shutdown_code(void)
+static void ftrace_startup(int command)
 {
-       stop_machine_run(__ftrace_modify_code, ftrace_shutdown_code, NR_CPUS);
+       if (unlikely(ftrace_disabled))
+               return;
+
+       mutex_lock(&ftrace_start_lock);
+       ftrace_start_up++;
+       command |= FTRACE_ENABLE_CALLS;
+
+       ftrace_startup_enable(command);
+
+       mutex_unlock(&ftrace_start_lock);
 }
 
-static void notrace ftrace_startup(void)
+static void ftrace_shutdown(int command)
 {
-       mutex_lock(&ftraced_lock);
-       ftraced_suspend++;
-       if (ftraced_suspend != 1)
+       if (unlikely(ftrace_disabled))
+               return;
+
+       mutex_lock(&ftrace_start_lock);
+       ftrace_start_up--;
+       if (!ftrace_start_up)
+               command |= FTRACE_DISABLE_CALLS;
+
+       if (saved_ftrace_func != ftrace_trace_function) {
+               saved_ftrace_func = ftrace_trace_function;
+               command |= FTRACE_UPDATE_TRACE_FUNC;
+       }
+
+       if (!command || !ftrace_enabled)
                goto out;
-       __unregister_ftrace_function(&ftrace_shutdown_ops);
 
-       if (ftrace_enabled)
-               ftrace_run_startup_code();
+       ftrace_run_update_code(command);
  out:
-       mutex_unlock(&ftraced_lock);
+       mutex_unlock(&ftrace_start_lock);
 }
 
-static void notrace ftrace_shutdown(void)
+static void ftrace_startup_sysctl(void)
 {
-       mutex_lock(&ftraced_lock);
-       ftraced_suspend--;
-       if (ftraced_suspend)
-               goto out;
+       int command = FTRACE_ENABLE_MCOUNT;
+
+       if (unlikely(ftrace_disabled))
+               return;
 
-       if (ftrace_enabled)
-               ftrace_run_shutdown_code();
+       mutex_lock(&ftrace_start_lock);
+       /* Force update next time */
+       saved_ftrace_func = NULL;
+       /* ftrace_start_up is true if we want ftrace running */
+       if (ftrace_start_up)
+               command |= FTRACE_ENABLE_CALLS;
 
-       __register_ftrace_function(&ftrace_shutdown_ops);
- out:
-       mutex_unlock(&ftraced_lock);
+       ftrace_run_update_code(command);
+       mutex_unlock(&ftrace_start_lock);
 }
 
-static void notrace ftrace_startup_sysctl(void)
+static void ftrace_shutdown_sysctl(void)
 {
-       mutex_lock(&ftraced_lock);
-       /* ftraced_suspend is true if we want ftrace running */
-       if (ftraced_suspend)
-               ftrace_run_startup_code();
-       mutex_unlock(&ftraced_lock);
-}
+       int command = FTRACE_DISABLE_MCOUNT;
 
-static void notrace ftrace_shutdown_sysctl(void)
-{
-       mutex_lock(&ftraced_lock);
-       /* ftraced_suspend is true if ftrace is running */
-       if (ftraced_suspend)
-               ftrace_run_shutdown_code();
-       mutex_unlock(&ftraced_lock);
+       if (unlikely(ftrace_disabled))
+               return;
+
+       mutex_lock(&ftrace_start_lock);
+       /* ftrace_start_up is true if ftrace is running */
+       if (ftrace_start_up)
+               command |= FTRACE_DISABLE_CALLS;
+
+       ftrace_run_update_code(command);
+       mutex_unlock(&ftrace_start_lock);
 }
 
 static cycle_t         ftrace_update_time;
 static unsigned long   ftrace_update_cnt;
 unsigned long          ftrace_update_tot_cnt;
 
-static int notrace __ftrace_update_code(void *ignore)
+static int ftrace_update_code(struct module *mod)
 {
-       struct dyn_ftrace *p;
-       struct hlist_head head;
-       struct hlist_node *t;
+       struct dyn_ftrace *p, *t;
        cycle_t start, stop;
-       int i;
 
-       /* Don't be calling ftrace ops now */
-       __unregister_ftrace_function(&ftrace_shutdown_ops);
-
-       start = now(raw_smp_processor_id());
+       start = ftrace_now(raw_smp_processor_id());
        ftrace_update_cnt = 0;
 
-       /* No locks needed, the machine is stopped! */
-       for (i = 0; i < FTRACE_HASHSIZE; i++) {
-               if (hlist_empty(&ftrace_hash[i]))
-                       continue;
+       list_for_each_entry_safe(p, t, &ftrace_new_addrs, list) {
 
-               head = ftrace_hash[i];
-               INIT_HLIST_HEAD(&ftrace_hash[i]);
+               /* If something went wrong, bail without enabling anything */
+               if (unlikely(ftrace_disabled))
+                       return -1;
 
-               /* all CPUS are stopped, we are safe to modify code */
-               hlist_for_each_entry(p, t, &head, node) {
-                       ftrace_code_disable(p, MCOUNT_ADDR);
-                       ftrace_update_cnt++;
-               }
+               list_del_init(&p->list);
 
+               /* convert record (i.e, patch mcount-call with NOP) */
+               if (ftrace_code_disable(mod, p)) {
+                       p->flags |= FTRACE_FL_CONVERTED;
+                       ftrace_update_cnt++;
+               } else
+                       ftrace_free_rec(p);
        }
 
-       stop = now(raw_smp_processor_id());
+       stop = ftrace_now(raw_smp_processor_id());
        ftrace_update_time = stop - start;
        ftrace_update_tot_cnt += ftrace_update_cnt;
 
-       __register_ftrace_function(&ftrace_shutdown_ops);
-
-       return 0;
-}
-
-static void notrace ftrace_update_code(void)
-{
-       stop_machine_run(__ftrace_update_code, NULL, NR_CPUS);
-}
-
-static int notrace ftraced(void *ignore)
-{
-       unsigned long usecs;
-
-       set_current_state(TASK_INTERRUPTIBLE);
-
-       while (!kthread_should_stop()) {
-
-               /* check once a second */
-               schedule_timeout(HZ);
-
-               mutex_lock(&ftrace_sysctl_lock);
-               mutex_lock(&ftraced_lock);
-               if (ftrace_enabled && ftraced_trigger && !ftraced_suspend) {
-                       ftrace_record_suspend++;
-                       ftrace_update_code();
-                       usecs = nsecs_to_usecs(ftrace_update_time);
-                       if (ftrace_update_tot_cnt > 100000) {
-                               ftrace_update_tot_cnt = 0;
-                               pr_info("hm, dftrace overflow: %lu change%s"
-                                        " (%lu total) in %lu usec%s\n",
-                                       ftrace_update_cnt,
-                                       ftrace_update_cnt != 1 ? "s" : "",
-                                       ftrace_update_tot_cnt,
-                                       usecs, usecs != 1 ? "s" : "");
-                               WARN_ON_ONCE(1);
-                       }
-                       ftraced_trigger = 0;
-                       ftrace_record_suspend--;
-               }
-               mutex_unlock(&ftraced_lock);
-               mutex_unlock(&ftrace_sysctl_lock);
-
-               ftrace_shutdown_replenish();
-
-               set_current_state(TASK_INTERRUPTIBLE);
-       }
-       __set_current_state(TASK_RUNNING);
        return 0;
 }
 
-static int __init ftrace_dyn_table_alloc(void)
+static int __init ftrace_dyn_table_alloc(unsigned long num_to_init)
 {
        struct ftrace_page *pg;
        int cnt;
        int i;
-       int ret;
-
-       ret = ftrace_dyn_arch_init();
-       if (ret)
-               return ret;
 
        /* allocate a few pages */
        ftrace_pages_start = (void *)get_zeroed_page(GFP_KERNEL);
@@ -542,7 +758,9 @@ static int __init ftrace_dyn_table_alloc(void)
 
        pg = ftrace_pages = ftrace_pages_start;
 
-       cnt = NR_TO_INIT / ENTRIES_PER_PAGE;
+       cnt = num_to_init / ENTRIES_PER_PAGE;
+       pr_info("ftrace: allocating %ld entries in %d pages\n",
+               num_to_init, cnt + 1);
 
        for (i = 0; i < cnt; i++) {
                pg->next = (void *)get_zeroed_page(GFP_KERNEL);
@@ -557,54 +775,1130 @@ static int __init ftrace_dyn_table_alloc(void)
        return 0;
 }
 
-static int __init notrace ftrace_shutdown_init(void)
-{
-       struct task_struct *p;
-       int ret;
+enum {
+       FTRACE_ITER_FILTER      = (1 << 0),
+       FTRACE_ITER_CONT        = (1 << 1),
+       FTRACE_ITER_NOTRACE     = (1 << 2),
+       FTRACE_ITER_FAILURES    = (1 << 3),
+};
 
-       ret = ftrace_dyn_table_alloc();
-       if (ret)
-               return ret;
+#define FTRACE_BUFF_MAX (KSYM_SYMBOL_LEN+4) /* room for wildcards */
 
-       p = kthread_run(ftraced, NULL, "ftraced");
-       if (IS_ERR(p))
-               return -1;
+struct ftrace_iterator {
+       struct ftrace_page      *pg;
+       unsigned                idx;
+       unsigned                flags;
+       unsigned char           buffer[FTRACE_BUFF_MAX+1];
+       unsigned                buffer_idx;
+       unsigned                filtered;
+};
 
-       __register_ftrace_function(&ftrace_shutdown_ops);
+static void *
+t_next(struct seq_file *m, void *v, loff_t *pos)
+{
+       struct ftrace_iterator *iter = m->private;
+       struct dyn_ftrace *rec = NULL;
 
-       return 0;
-}
+       (*pos)++;
 
-core_initcall(ftrace_shutdown_init);
-#else
-# define ftrace_startup()        do { } while (0)
-# define ftrace_shutdown()       do { } while (0)
-# define ftrace_startup_sysctl()  do { } while (0)
-# define ftrace_shutdown_sysctl() do { } while (0)
-#endif /* CONFIG_DYNAMIC_FTRACE */
+       /* should not be called from interrupt context */
+       spin_lock(&ftrace_lock);
+ retry:
+       if (iter->idx >= iter->pg->index) {
+               if (iter->pg->next) {
+                       iter->pg = iter->pg->next;
+                       iter->idx = 0;
+                       goto retry;
+               } else {
+                       iter->idx = -1;
+               }
+       } else {
+               rec = &iter->pg->records[iter->idx++];
+               if ((rec->flags & FTRACE_FL_FREE) ||
 
-/**
- * register_ftrace_function - register a function for profiling
- * @ops - ops structure that holds the function for profiling.
- *
- * Register a function to be called by all functions in the
- * kernel.
- *
- * Note: @ops->func and all the functions it calls must be labeled
- *       with "notrace", otherwise it will go into a
- *       recursive loop.
- */
-int register_ftrace_function(struct ftrace_ops *ops)
-{
-       int ret;
+                   (!(iter->flags & FTRACE_ITER_FAILURES) &&
+                    (rec->flags & FTRACE_FL_FAILED)) ||
 
-       mutex_lock(&ftrace_sysctl_lock);
-       ftrace_startup();
+                   ((iter->flags & FTRACE_ITER_FAILURES) &&
+                    !(rec->flags & FTRACE_FL_FAILED)) ||
 
-       ret = __register_ftrace_function(ops);
-       mutex_unlock(&ftrace_sysctl_lock);
+                   ((iter->flags & FTRACE_ITER_FILTER) &&
+                    !(rec->flags & FTRACE_FL_FILTER)) ||
 
-       return ret;
+                   ((iter->flags & FTRACE_ITER_NOTRACE) &&
+                    !(rec->flags & FTRACE_FL_NOTRACE))) {
+                       rec = NULL;
+                       goto retry;
+               }
+       }
+       spin_unlock(&ftrace_lock);
+
+       return rec;
+}
+
+static void *t_start(struct seq_file *m, loff_t *pos)
+{
+       struct ftrace_iterator *iter = m->private;
+       void *p = NULL;
+
+       if (*pos > 0) {
+               if (iter->idx < 0)
+                       return p;
+               (*pos)--;
+               iter->idx--;
+       }
+
+       p = t_next(m, p, pos);
+
+       return p;
+}
+
+static void t_stop(struct seq_file *m, void *p)
+{
+}
+
+static int t_show(struct seq_file *m, void *v)
+{
+       struct dyn_ftrace *rec = v;
+       char str[KSYM_SYMBOL_LEN];
+
+       if (!rec)
+               return 0;
+
+       kallsyms_lookup(rec->ip, NULL, NULL, NULL, str);
+
+       seq_printf(m, "%s\n", str);
+
+       return 0;
+}
+
+static struct seq_operations show_ftrace_seq_ops = {
+       .start = t_start,
+       .next = t_next,
+       .stop = t_stop,
+       .show = t_show,
+};
+
+static int
+ftrace_avail_open(struct inode *inode, struct file *file)
+{
+       struct ftrace_iterator *iter;
+       int ret;
+
+       if (unlikely(ftrace_disabled))
+               return -ENODEV;
+
+       iter = kzalloc(sizeof(*iter), GFP_KERNEL);
+       if (!iter)
+               return -ENOMEM;
+
+       iter->pg = ftrace_pages_start;
+
+       ret = seq_open(file, &show_ftrace_seq_ops);
+       if (!ret) {
+               struct seq_file *m = file->private_data;
+
+               m->private = iter;
+       } else {
+               kfree(iter);
+       }
+
+       return ret;
+}
+
+int ftrace_avail_release(struct inode *inode, struct file *file)
+{
+       struct seq_file *m = (struct seq_file *)file->private_data;
+       struct ftrace_iterator *iter = m->private;
+
+       seq_release(inode, file);
+       kfree(iter);
+
+       return 0;
+}
+
+static int
+ftrace_failures_open(struct inode *inode, struct file *file)
+{
+       int ret;
+       struct seq_file *m;
+       struct ftrace_iterator *iter;
+
+       ret = ftrace_avail_open(inode, file);
+       if (!ret) {
+               m = (struct seq_file *)file->private_data;
+               iter = (struct ftrace_iterator *)m->private;
+               iter->flags = FTRACE_ITER_FAILURES;
+       }
+
+       return ret;
+}
+
+
+static void ftrace_filter_reset(int enable)
+{
+       struct ftrace_page *pg;
+       struct dyn_ftrace *rec;
+       unsigned long type = enable ? FTRACE_FL_FILTER : FTRACE_FL_NOTRACE;
+       unsigned i;
+
+       /* should not be called from interrupt context */
+       spin_lock(&ftrace_lock);
+       if (enable)
+               ftrace_filtered = 0;
+       pg = ftrace_pages_start;
+       while (pg) {
+               for (i = 0; i < pg->index; i++) {
+                       rec = &pg->records[i];
+                       if (rec->flags & FTRACE_FL_FAILED)
+                               continue;
+                       rec->flags &= ~type;
+               }
+               pg = pg->next;
+       }
+       spin_unlock(&ftrace_lock);
+}
+
+static int
+ftrace_regex_open(struct inode *inode, struct file *file, int enable)
+{
+       struct ftrace_iterator *iter;
+       int ret = 0;
+
+       if (unlikely(ftrace_disabled))
+               return -ENODEV;
+
+       iter = kzalloc(sizeof(*iter), GFP_KERNEL);
+       if (!iter)
+               return -ENOMEM;
+
+       mutex_lock(&ftrace_regex_lock);
+       if ((file->f_mode & FMODE_WRITE) &&
+           !(file->f_flags & O_APPEND))
+               ftrace_filter_reset(enable);
+
+       if (file->f_mode & FMODE_READ) {
+               iter->pg = ftrace_pages_start;
+               iter->flags = enable ? FTRACE_ITER_FILTER :
+                       FTRACE_ITER_NOTRACE;
+
+               ret = seq_open(file, &show_ftrace_seq_ops);
+               if (!ret) {
+                       struct seq_file *m = file->private_data;
+                       m->private = iter;
+               } else
+                       kfree(iter);
+       } else
+               file->private_data = iter;
+       mutex_unlock(&ftrace_regex_lock);
+
+       return ret;
+}
+
+static int
+ftrace_filter_open(struct inode *inode, struct file *file)
+{
+       return ftrace_regex_open(inode, file, 1);
+}
+
+static int
+ftrace_notrace_open(struct inode *inode, struct file *file)
+{
+       return ftrace_regex_open(inode, file, 0);
+}
+
+static ssize_t
+ftrace_regex_read(struct file *file, char __user *ubuf,
+                      size_t cnt, loff_t *ppos)
+{
+       if (file->f_mode & FMODE_READ)
+               return seq_read(file, ubuf, cnt, ppos);
+       else
+               return -EPERM;
+}
+
+static loff_t
+ftrace_regex_lseek(struct file *file, loff_t offset, int origin)
+{
+       loff_t ret;
+
+       if (file->f_mode & FMODE_READ)
+               ret = seq_lseek(file, offset, origin);
+       else
+               file->f_pos = ret = 1;
+
+       return ret;
+}
+
+enum {
+       MATCH_FULL,
+       MATCH_FRONT_ONLY,
+       MATCH_MIDDLE_ONLY,
+       MATCH_END_ONLY,
+};
+
+static void
+ftrace_match(unsigned char *buff, int len, int enable)
+{
+       char str[KSYM_SYMBOL_LEN];
+       char *search = NULL;
+       struct ftrace_page *pg;
+       struct dyn_ftrace *rec;
+       int type = MATCH_FULL;
+       unsigned long flag = enable ? FTRACE_FL_FILTER : FTRACE_FL_NOTRACE;
+       unsigned i, match = 0, search_len = 0;
+       int not = 0;
+
+       if (buff[0] == '!') {
+               not = 1;
+               buff++;
+               len--;
+       }
+
+       for (i = 0; i < len; i++) {
+               if (buff[i] == '*') {
+                       if (!i) {
+                               search = buff + i + 1;
+                               type = MATCH_END_ONLY;
+                               search_len = len - (i + 1);
+                       } else {
+                               if (type == MATCH_END_ONLY) {
+                                       type = MATCH_MIDDLE_ONLY;
+                               } else {
+                                       match = i;
+                                       type = MATCH_FRONT_ONLY;
+                               }
+                               buff[i] = 0;
+                               break;
+                       }
+               }
+       }
+
+       /* should not be called from interrupt context */
+       spin_lock(&ftrace_lock);
+       if (enable)
+               ftrace_filtered = 1;
+       pg = ftrace_pages_start;
+       while (pg) {
+               for (i = 0; i < pg->index; i++) {
+                       int matched = 0;
+                       char *ptr;
+
+                       rec = &pg->records[i];
+                       if (rec->flags & FTRACE_FL_FAILED)
+                               continue;
+                       kallsyms_lookup(rec->ip, NULL, NULL, NULL, str);
+                       switch (type) {
+                       case MATCH_FULL:
+                               if (strcmp(str, buff) == 0)
+                                       matched = 1;
+                               break;
+                       case MATCH_FRONT_ONLY:
+                               if (memcmp(str, buff, match) == 0)
+                                       matched = 1;
+                               break;
+                       case MATCH_MIDDLE_ONLY:
+                               if (strstr(str, search))
+                                       matched = 1;
+                               break;
+                       case MATCH_END_ONLY:
+                               ptr = strstr(str, search);
+                               if (ptr && (ptr[search_len] == 0))
+                                       matched = 1;
+                               break;
+                       }
+                       if (matched) {
+                               if (not)
+                                       rec->flags &= ~flag;
+                               else
+                                       rec->flags |= flag;
+                       }
+               }
+               pg = pg->next;
+       }
+       spin_unlock(&ftrace_lock);
+}
+
+static ssize_t
+ftrace_regex_write(struct file *file, const char __user *ubuf,
+                  size_t cnt, loff_t *ppos, int enable)
+{
+       struct ftrace_iterator *iter;
+       char ch;
+       size_t read = 0;
+       ssize_t ret;
+
+       if (!cnt || cnt < 0)
+               return 0;
+
+       mutex_lock(&ftrace_regex_lock);
+
+       if (file->f_mode & FMODE_READ) {
+               struct seq_file *m = file->private_data;
+               iter = m->private;
+       } else
+               iter = file->private_data;
+
+       if (!*ppos) {
+               iter->flags &= ~FTRACE_ITER_CONT;
+               iter->buffer_idx = 0;
+       }
+
+       ret = get_user(ch, ubuf++);
+       if (ret)
+               goto out;
+       read++;
+       cnt--;
+
+       if (!(iter->flags & ~FTRACE_ITER_CONT)) {
+               /* skip white space */
+               while (cnt && isspace(ch)) {
+                       ret = get_user(ch, ubuf++);
+                       if (ret)
+                               goto out;
+                       read++;
+                       cnt--;
+               }
+
+               if (isspace(ch)) {
+                       file->f_pos += read;
+                       ret = read;
+                       goto out;
+               }
+
+               iter->buffer_idx = 0;
+       }
+
+       while (cnt && !isspace(ch)) {
+               if (iter->buffer_idx < FTRACE_BUFF_MAX)
+                       iter->buffer[iter->buffer_idx++] = ch;
+               else {
+                       ret = -EINVAL;
+                       goto out;
+               }
+               ret = get_user(ch, ubuf++);
+               if (ret)
+                       goto out;
+               read++;
+               cnt--;
+       }
+
+       if (isspace(ch)) {
+               iter->filtered++;
+               iter->buffer[iter->buffer_idx] = 0;
+               ftrace_match(iter->buffer, iter->buffer_idx, enable);
+               iter->buffer_idx = 0;
+       } else
+               iter->flags |= FTRACE_ITER_CONT;
+
+
+       file->f_pos += read;
+
+       ret = read;
+ out:
+       mutex_unlock(&ftrace_regex_lock);
+
+       return ret;
+}
+
+static ssize_t
+ftrace_filter_write(struct file *file, const char __user *ubuf,
+                   size_t cnt, loff_t *ppos)
+{
+       return ftrace_regex_write(file, ubuf, cnt, ppos, 1);
+}
+
+static ssize_t
+ftrace_notrace_write(struct file *file, const char __user *ubuf,
+                    size_t cnt, loff_t *ppos)
+{
+       return ftrace_regex_write(file, ubuf, cnt, ppos, 0);
+}
+
+static void
+ftrace_set_regex(unsigned char *buf, int len, int reset, int enable)
+{
+       if (unlikely(ftrace_disabled))
+               return;
+
+       mutex_lock(&ftrace_regex_lock);
+       if (reset)
+               ftrace_filter_reset(enable);
+       if (buf)
+               ftrace_match(buf, len, enable);
+       mutex_unlock(&ftrace_regex_lock);
+}
+
+/**
+ * ftrace_set_filter - set a function to filter on in ftrace
+ * @buf - the string that holds the function filter text.
+ * @len - the length of the string.
+ * @reset - non zero to reset all filters before applying this filter.
+ *
+ * Filters denote which functions should be enabled when tracing is enabled.
+ * If @buf is NULL and reset is set, all functions will be enabled for tracing.
+ */
+void ftrace_set_filter(unsigned char *buf, int len, int reset)
+{
+       ftrace_set_regex(buf, len, reset, 1);
+}
+
+/**
+ * ftrace_set_notrace - set a function to not trace in ftrace
+ * @buf - the string that holds the function notrace text.
+ * @len - the length of the string.
+ * @reset - non zero to reset all filters before applying this filter.
+ *
+ * Notrace Filters denote which functions should not be enabled when tracing
+ * is enabled. If @buf is NULL and reset is set, all functions will be enabled
+ * for tracing.
+ */
+void ftrace_set_notrace(unsigned char *buf, int len, int reset)
+{
+       ftrace_set_regex(buf, len, reset, 0);
+}
+
+static int
+ftrace_regex_release(struct inode *inode, struct file *file, int enable)
+{
+       struct seq_file *m = (struct seq_file *)file->private_data;
+       struct ftrace_iterator *iter;
+
+       mutex_lock(&ftrace_regex_lock);
+       if (file->f_mode & FMODE_READ) {
+               iter = m->private;
+
+               seq_release(inode, file);
+       } else
+               iter = file->private_data;
+
+       if (iter->buffer_idx) {
+               iter->filtered++;
+               iter->buffer[iter->buffer_idx] = 0;
+               ftrace_match(iter->buffer, iter->buffer_idx, enable);
+       }
+
+       mutex_lock(&ftrace_sysctl_lock);
+       mutex_lock(&ftrace_start_lock);
+       if (ftrace_start_up && ftrace_enabled)
+               ftrace_run_update_code(FTRACE_ENABLE_CALLS);
+       mutex_unlock(&ftrace_start_lock);
+       mutex_unlock(&ftrace_sysctl_lock);
+
+       kfree(iter);
+       mutex_unlock(&ftrace_regex_lock);
+       return 0;
+}
+
+static int
+ftrace_filter_release(struct inode *inode, struct file *file)
+{
+       return ftrace_regex_release(inode, file, 1);
+}
+
+static int
+ftrace_notrace_release(struct inode *inode, struct file *file)
+{
+       return ftrace_regex_release(inode, file, 0);
+}
+
+static struct file_operations ftrace_avail_fops = {
+       .open = ftrace_avail_open,
+       .read = seq_read,
+       .llseek = seq_lseek,
+       .release = ftrace_avail_release,
+};
+
+static struct file_operations ftrace_failures_fops = {
+       .open = ftrace_failures_open,
+       .read = seq_read,
+       .llseek = seq_lseek,
+       .release = ftrace_avail_release,
+};
+
+static struct file_operations ftrace_filter_fops = {
+       .open = ftrace_filter_open,
+       .read = ftrace_regex_read,
+       .write = ftrace_filter_write,
+       .llseek = ftrace_regex_lseek,
+       .release = ftrace_filter_release,
+};
+
+static struct file_operations ftrace_notrace_fops = {
+       .open = ftrace_notrace_open,
+       .read = ftrace_regex_read,
+       .write = ftrace_notrace_write,
+       .llseek = ftrace_regex_lseek,
+       .release = ftrace_notrace_release,
+};
+
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+
+static DEFINE_MUTEX(graph_lock);
+
+int ftrace_graph_count;
+unsigned long ftrace_graph_funcs[FTRACE_GRAPH_MAX_FUNCS] __read_mostly;
+
+static void *
+g_next(struct seq_file *m, void *v, loff_t *pos)
+{
+       unsigned long *array = m->private;
+       int index = *pos;
+
+       (*pos)++;
+
+       if (index >= ftrace_graph_count)
+               return NULL;
+
+       return &array[index];
+}
+
+static void *g_start(struct seq_file *m, loff_t *pos)
+{
+       void *p = NULL;
+
+       mutex_lock(&graph_lock);
+
+       p = g_next(m, p, pos);
+
+       return p;
+}
+
+static void g_stop(struct seq_file *m, void *p)
+{
+       mutex_unlock(&graph_lock);
+}
+
+static int g_show(struct seq_file *m, void *v)
+{
+       unsigned long *ptr = v;
+       char str[KSYM_SYMBOL_LEN];
+
+       if (!ptr)
+               return 0;
+
+       kallsyms_lookup(*ptr, NULL, NULL, NULL, str);
+
+       seq_printf(m, "%s\n", str);
+
+       return 0;
+}
+
+static struct seq_operations ftrace_graph_seq_ops = {
+       .start = g_start,
+       .next = g_next,
+       .stop = g_stop,
+       .show = g_show,
+};
+
+static int
+ftrace_graph_open(struct inode *inode, struct file *file)
+{
+       int ret = 0;
+
+       if (unlikely(ftrace_disabled))
+               return -ENODEV;
+
+       mutex_lock(&graph_lock);
+       if ((file->f_mode & FMODE_WRITE) &&
+           !(file->f_flags & O_APPEND)) {
+               ftrace_graph_count = 0;
+               memset(ftrace_graph_funcs, 0, sizeof(ftrace_graph_funcs));
+       }
+
+       if (file->f_mode & FMODE_READ) {
+               ret = seq_open(file, &ftrace_graph_seq_ops);
+               if (!ret) {
+                       struct seq_file *m = file->private_data;
+                       m->private = ftrace_graph_funcs;
+               }
+       } else
+               file->private_data = ftrace_graph_funcs;
+       mutex_unlock(&graph_lock);
+
+       return ret;
+}
+
+static ssize_t
+ftrace_graph_read(struct file *file, char __user *ubuf,
+                      size_t cnt, loff_t *ppos)
+{
+       if (file->f_mode & FMODE_READ)
+               return seq_read(file, ubuf, cnt, ppos);
+       else
+               return -EPERM;
+}
+
+static int
+ftrace_set_func(unsigned long *array, int idx, char *buffer)
+{
+       char str[KSYM_SYMBOL_LEN];
+       struct dyn_ftrace *rec;
+       struct ftrace_page *pg;
+       int found = 0;
+       int i, j;
+
+       if (ftrace_disabled)
+               return -ENODEV;
+
+       /* should not be called from interrupt context */
+       spin_lock(&ftrace_lock);
+
+       for (pg = ftrace_pages_start; pg; pg = pg->next) {
+               for (i = 0; i < pg->index; i++) {
+                       rec = &pg->records[i];
+
+                       if (rec->flags & (FTRACE_FL_FAILED | FTRACE_FL_FREE))
+                               continue;
+
+                       kallsyms_lookup(rec->ip, NULL, NULL, NULL, str);
+                       if (strcmp(str, buffer) == 0) {
+                               found = 1;
+                               for (j = 0; j < idx; j++)
+                                       if (array[j] == rec->ip) {
+                                               found = 0;
+                                               break;
+                                       }
+                               if (found)
+                                       array[idx] = rec->ip;
+                               break;
+                       }
+               }
+       }
+       spin_unlock(&ftrace_lock);
+
+       return found ? 0 : -EINVAL;
+}
+
+static ssize_t
+ftrace_graph_write(struct file *file, const char __user *ubuf,
+                  size_t cnt, loff_t *ppos)
+{
+       unsigned char buffer[FTRACE_BUFF_MAX+1];
+       unsigned long *array;
+       size_t read = 0;
+       ssize_t ret;
+       int index = 0;
+       char ch;
+
+       if (!cnt || cnt < 0)
+               return 0;
+
+       mutex_lock(&graph_lock);
+
+       if (ftrace_graph_count >= FTRACE_GRAPH_MAX_FUNCS) {
+               ret = -EBUSY;
+               goto out;
+       }
+
+       if (file->f_mode & FMODE_READ) {
+               struct seq_file *m = file->private_data;
+               array = m->private;
+       } else
+               array = file->private_data;
+
+       ret = get_user(ch, ubuf++);
+       if (ret)
+               goto out;
+       read++;
+       cnt--;
+
+       /* skip white space */
+       while (cnt && isspace(ch)) {
+               ret = get_user(ch, ubuf++);
+               if (ret)
+                       goto out;
+               read++;
+               cnt--;
+       }
+
+       if (isspace(ch)) {
+               *ppos += read;
+               ret = read;
+               goto out;
+       }
+
+       while (cnt && !isspace(ch)) {
+               if (index < FTRACE_BUFF_MAX)
+                       buffer[index++] = ch;
+               else {
+                       ret = -EINVAL;
+                       goto out;
+               }
+               ret = get_user(ch, ubuf++);
+               if (ret)
+                       goto out;
+               read++;
+               cnt--;
+       }
+       buffer[index] = 0;
+
+       /* we allow only one at a time */
+       ret = ftrace_set_func(array, ftrace_graph_count, buffer);
+       if (ret)
+               goto out;
+
+       ftrace_graph_count++;
+
+       file->f_pos += read;
+
+       ret = read;
+ out:
+       mutex_unlock(&graph_lock);
+
+       return ret;
+}
+
+static const struct file_operations ftrace_graph_fops = {
+       .open = ftrace_graph_open,
+       .read = ftrace_graph_read,
+       .write = ftrace_graph_write,
+};
+#endif /* CONFIG_FUNCTION_GRAPH_TRACER */
+
+static __init int ftrace_init_dyn_debugfs(struct dentry *d_tracer)
+{
+       struct dentry *entry;
+
+       entry = debugfs_create_file("available_filter_functions", 0444,
+                                   d_tracer, NULL, &ftrace_avail_fops);
+       if (!entry)
+               pr_warning("Could not create debugfs "
+                          "'available_filter_functions' entry\n");
+
+       entry = debugfs_create_file("failures", 0444,
+                                   d_tracer, NULL, &ftrace_failures_fops);
+       if (!entry)
+               pr_warning("Could not create debugfs 'failures' entry\n");
+
+       entry = debugfs_create_file("set_ftrace_filter", 0644, d_tracer,
+                                   NULL, &ftrace_filter_fops);
+       if (!entry)
+               pr_warning("Could not create debugfs "
+                          "'set_ftrace_filter' entry\n");
+
+       entry = debugfs_create_file("set_ftrace_notrace", 0644, d_tracer,
+                                   NULL, &ftrace_notrace_fops);
+       if (!entry)
+               pr_warning("Could not create debugfs "
+                          "'set_ftrace_notrace' entry\n");
+
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+       entry = debugfs_create_file("set_graph_function", 0444, d_tracer,
+                                   NULL,
+                                   &ftrace_graph_fops);
+       if (!entry)
+               pr_warning("Could not create debugfs "
+                          "'set_graph_function' entry\n");
+#endif /* CONFIG_FUNCTION_GRAPH_TRACER */
+
+       return 0;
+}
+
+static int ftrace_convert_nops(struct module *mod,
+                              unsigned long *start,
+                              unsigned long *end)
+{
+       unsigned long *p;
+       unsigned long addr;
+       unsigned long flags;
+
+       mutex_lock(&ftrace_start_lock);
+       p = start;
+       while (p < end) {
+               addr = ftrace_call_adjust(*p++);
+               /*
+                * Some architecture linkers will pad between
+                * the different mcount_loc sections of different
+                * object files to satisfy alignments.
+                * Skip any NULL pointers.
+                */
+               if (!addr)
+                       continue;
+               ftrace_record_ip(addr);
+       }
+
+       /* disable interrupts to prevent kstop machine */
+       local_irq_save(flags);
+       ftrace_update_code(mod);
+       local_irq_restore(flags);
+       mutex_unlock(&ftrace_start_lock);
+
+       return 0;
+}
+
+void ftrace_init_module(struct module *mod,
+                       unsigned long *start, unsigned long *end)
+{
+       if (ftrace_disabled || start == end)
+               return;
+       ftrace_convert_nops(mod, start, end);
+}
+
+extern unsigned long __start_mcount_loc[];
+extern unsigned long __stop_mcount_loc[];
+
+void __init ftrace_init(void)
+{
+       unsigned long count, addr, flags;
+       int ret;
+
+       /* Keep the ftrace pointer to the stub */
+       addr = (unsigned long)ftrace_stub;
+
+       local_irq_save(flags);
+       ftrace_dyn_arch_init(&addr);
+       local_irq_restore(flags);
+
+       /* ftrace_dyn_arch_init places the return code in addr */
+       if (addr)
+               goto failed;
+
+       count = __stop_mcount_loc - __start_mcount_loc;
+
+       ret = ftrace_dyn_table_alloc(count);
+       if (ret)
+               goto failed;
+
+       last_ftrace_enabled = ftrace_enabled = 1;
+
+       ret = ftrace_convert_nops(NULL,
+                                 __start_mcount_loc,
+                                 __stop_mcount_loc);
+
+       return;
+ failed:
+       ftrace_disabled = 1;
+}
+
+#else
+
+static int __init ftrace_nodyn_init(void)
+{
+       ftrace_enabled = 1;
+       return 0;
+}
+device_initcall(ftrace_nodyn_init);
+
+static inline int ftrace_init_dyn_debugfs(struct dentry *d_tracer) { return 0; }
+static inline void ftrace_startup_enable(int command) { }
+/* Keep as macros so we do not need to define the commands */
+# define ftrace_startup(command)       do { } while (0)
+# define ftrace_shutdown(command)      do { } while (0)
+# define ftrace_startup_sysctl()       do { } while (0)
+# define ftrace_shutdown_sysctl()      do { } while (0)
+#endif /* CONFIG_DYNAMIC_FTRACE */
+
+static ssize_t
+ftrace_pid_read(struct file *file, char __user *ubuf,
+                      size_t cnt, loff_t *ppos)
+{
+       char buf[64];
+       int r;
+
+       if (ftrace_pid_trace == ftrace_swapper_pid)
+               r = sprintf(buf, "swapper tasks\n");
+       else if (ftrace_pid_trace)
+               r = sprintf(buf, "%u\n", pid_nr(ftrace_pid_trace));
+       else
+               r = sprintf(buf, "no pid\n");
+
+       return simple_read_from_buffer(ubuf, cnt, ppos, buf, r);
+}
+
+static void clear_ftrace_swapper(void)
+{
+       struct task_struct *p;
+       int cpu;
+
+       get_online_cpus();
+       for_each_online_cpu(cpu) {
+               p = idle_task(cpu);
+               clear_tsk_trace_trace(p);
+       }
+       put_online_cpus();
+}
+
+static void set_ftrace_swapper(void)
+{
+       struct task_struct *p;
+       int cpu;
+
+       get_online_cpus();
+       for_each_online_cpu(cpu) {
+               p = idle_task(cpu);
+               set_tsk_trace_trace(p);
+       }
+       put_online_cpus();
+}
+
+static void clear_ftrace_pid(struct pid *pid)
+{
+       struct task_struct *p;
+
+       do_each_pid_task(pid, PIDTYPE_PID, p) {
+               clear_tsk_trace_trace(p);
+       } while_each_pid_task(pid, PIDTYPE_PID, p);
+       put_pid(pid);
+}
+
+static void set_ftrace_pid(struct pid *pid)
+{
+       struct task_struct *p;
+
+       do_each_pid_task(pid, PIDTYPE_PID, p) {
+               set_tsk_trace_trace(p);
+       } while_each_pid_task(pid, PIDTYPE_PID, p);
+}
+
+static void clear_ftrace_pid_task(struct pid **pid)
+{
+       if (*pid == ftrace_swapper_pid)
+               clear_ftrace_swapper();
+       else
+               clear_ftrace_pid(*pid);
+
+       *pid = NULL;
+}
+
+static void set_ftrace_pid_task(struct pid *pid)
+{
+       if (pid == ftrace_swapper_pid)
+               set_ftrace_swapper();
+       else
+               set_ftrace_pid(pid);
+}
+
+static ssize_t
+ftrace_pid_write(struct file *filp, const char __user *ubuf,
+                  size_t cnt, loff_t *ppos)
+{
+       struct pid *pid;
+       char buf[64];
+       long val;
+       int ret;
+
+       if (cnt >= sizeof(buf))
+               return -EINVAL;
+
+       if (copy_from_user(&buf, ubuf, cnt))
+               return -EFAULT;
+
+       buf[cnt] = 0;
+
+       ret = strict_strtol(buf, 10, &val);
+       if (ret < 0)
+               return ret;
+
+       mutex_lock(&ftrace_start_lock);
+       if (val < 0) {
+               /* disable pid tracing */
+               if (!ftrace_pid_trace)
+                       goto out;
+
+               clear_ftrace_pid_task(&ftrace_pid_trace);
+
+       } else {
+               /* swapper task is special */
+               if (!val) {
+                       pid = ftrace_swapper_pid;
+                       if (pid == ftrace_pid_trace)
+                               goto out;
+               } else {
+                       pid = find_get_pid(val);
+
+                       if (pid == ftrace_pid_trace) {
+                               put_pid(pid);
+                               goto out;
+                       }
+               }
+
+               if (ftrace_pid_trace)
+                       clear_ftrace_pid_task(&ftrace_pid_trace);
+
+               if (!pid)
+                       goto out;
+
+               ftrace_pid_trace = pid;
+
+               set_ftrace_pid_task(ftrace_pid_trace);
+       }
+
+       /* update the function call */
+       ftrace_update_pid_func();
+       ftrace_startup_enable(0);
+
+ out:
+       mutex_unlock(&ftrace_start_lock);
+
+       return cnt;
+}
+
+static struct file_operations ftrace_pid_fops = {
+       .read = ftrace_pid_read,
+       .write = ftrace_pid_write,
+};
+
+static __init int ftrace_init_debugfs(void)
+{
+       struct dentry *d_tracer;
+       struct dentry *entry;
+
+       d_tracer = tracing_init_dentry();
+       if (!d_tracer)
+               return 0;
+
+       ftrace_init_dyn_debugfs(d_tracer);
+
+       entry = debugfs_create_file("set_ftrace_pid", 0644, d_tracer,
+                                   NULL, &ftrace_pid_fops);
+       if (!entry)
+               pr_warning("Could not create debugfs "
+                          "'set_ftrace_pid' entry\n");
+       return 0;
+}
+
+fs_initcall(ftrace_init_debugfs);
+
+/**
+ * ftrace_kill - kill ftrace
+ *
+ * This function should be used by panic code. It stops ftrace
+ * but in a not so nice way. If you need to simply kill ftrace
+ * from a non-atomic section, use ftrace_kill.
+ */
+void ftrace_kill(void)
+{
+       ftrace_disabled = 1;
+       ftrace_enabled = 0;
+       clear_ftrace_function();
+}
+
+/**
+ * register_ftrace_function - register a function for profiling
+ * @ops - ops structure that holds the function for profiling.
+ *
+ * Register a function to be called by all functions in the
+ * kernel.
+ *
+ * Note: @ops->func and all the functions it calls must be labeled
+ *       with "notrace", otherwise it will go into a
+ *       recursive loop.
+ */
+int register_ftrace_function(struct ftrace_ops *ops)
+{
+       int ret;
+
+       if (unlikely(ftrace_disabled))
+               return -1;
+
+       mutex_lock(&ftrace_sysctl_lock);
+
+       ret = __register_ftrace_function(ops);
+       ftrace_startup(0);
+
+       mutex_unlock(&ftrace_sysctl_lock);
+       return ret;
 }
 
 /**
@@ -619,25 +1913,25 @@ int unregister_ftrace_function(struct ftrace_ops *ops)
 
        mutex_lock(&ftrace_sysctl_lock);
        ret = __unregister_ftrace_function(ops);
-
-       if (ftrace_list == &ftrace_list_end)
-               ftrace_shutdown();
-
+       ftrace_shutdown(0);
        mutex_unlock(&ftrace_sysctl_lock);
 
        return ret;
 }
 
-notrace int
+int
 ftrace_enable_sysctl(struct ctl_table *table, int write,
-                    struct file *filp, void __user *buffer, size_t *lenp,
+                    struct file *file, void __user *buffer, size_t *lenp,
                     loff_t *ppos)
 {
        int ret;
 
+       if (unlikely(ftrace_disabled))
+               return -ENODEV;
+
        mutex_lock(&ftrace_sysctl_lock);
 
-       ret  = proc_dointvec(table, write, filp, buffer, lenp, ppos);
+       ret  = proc_dointvec(table, write, file, buffer, lenp, ppos);
 
        if (ret || !write || (last_ftrace_enabled == ftrace_enabled))
                goto out;
@@ -667,3 +1961,154 @@ ftrace_enable_sysctl(struct ctl_table *table, int write,
        mutex_unlock(&ftrace_sysctl_lock);
        return ret;
 }
+
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+
+static atomic_t ftrace_graph_active;
+
+int ftrace_graph_entry_stub(struct ftrace_graph_ent *trace)
+{
+       return 0;
+}
+
+/* The callbacks that hook a function */
+trace_func_graph_ret_t ftrace_graph_return =
+                       (trace_func_graph_ret_t)ftrace_stub;
+trace_func_graph_ent_t ftrace_graph_entry = ftrace_graph_entry_stub;
+
+/* Try to assign a return stack array on FTRACE_RETSTACK_ALLOC_SIZE tasks. */
+static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list)
+{
+       int i;
+       int ret = 0;
+       unsigned long flags;
+       int start = 0, end = FTRACE_RETSTACK_ALLOC_SIZE;
+       struct task_struct *g, *t;
+
+       for (i = 0; i < FTRACE_RETSTACK_ALLOC_SIZE; i++) {
+               ret_stack_list[i] = kmalloc(FTRACE_RETFUNC_DEPTH
+                                       * sizeof(struct ftrace_ret_stack),
+                                       GFP_KERNEL);
+               if (!ret_stack_list[i]) {
+                       start = 0;
+                       end = i;
+                       ret = -ENOMEM;
+                       goto free;
+               }
+       }
+
+       read_lock_irqsave(&tasklist_lock, flags);
+       do_each_thread(g, t) {
+               if (start == end) {
+                       ret = -EAGAIN;
+                       goto unlock;
+               }
+
+               if (t->ret_stack == NULL) {
+                       t->curr_ret_stack = -1;
+                       /* Make sure IRQs see the -1 first: */
+                       barrier();
+                       t->ret_stack = ret_stack_list[start++];
+                       atomic_set(&t->tracing_graph_pause, 0);
+                       atomic_set(&t->trace_overrun, 0);
+               }
+       } while_each_thread(g, t);
+
+unlock:
+       read_unlock_irqrestore(&tasklist_lock, flags);
+free:
+       for (i = start; i < end; i++)
+               kfree(ret_stack_list[i]);
+       return ret;
+}
+
+/* Allocate a return stack for each task */
+static int start_graph_tracing(void)
+{
+       struct ftrace_ret_stack **ret_stack_list;
+       int ret;
+
+       ret_stack_list = kmalloc(FTRACE_RETSTACK_ALLOC_SIZE *
+                               sizeof(struct ftrace_ret_stack *),
+                               GFP_KERNEL);
+
+       if (!ret_stack_list)
+               return -ENOMEM;
+
+       do {
+               ret = alloc_retstack_tasklist(ret_stack_list);
+       } while (ret == -EAGAIN);
+
+       kfree(ret_stack_list);
+       return ret;
+}
+
+int register_ftrace_graph(trace_func_graph_ret_t retfunc,
+                       trace_func_graph_ent_t entryfunc)
+{
+       int ret = 0;
+
+       mutex_lock(&ftrace_sysctl_lock);
+
+       atomic_inc(&ftrace_graph_active);
+       ret = start_graph_tracing();
+       if (ret) {
+               atomic_dec(&ftrace_graph_active);
+               goto out;
+       }
+
+       ftrace_graph_return = retfunc;
+       ftrace_graph_entry = entryfunc;
+
+       ftrace_startup(FTRACE_START_FUNC_RET);
+
+out:
+       mutex_unlock(&ftrace_sysctl_lock);
+       return ret;
+}
+
+void unregister_ftrace_graph(void)
+{
+       mutex_lock(&ftrace_sysctl_lock);
+
+       atomic_dec(&ftrace_graph_active);
+       ftrace_graph_return = (trace_func_graph_ret_t)ftrace_stub;
+       ftrace_graph_entry = ftrace_graph_entry_stub;
+       ftrace_shutdown(FTRACE_STOP_FUNC_RET);
+
+       mutex_unlock(&ftrace_sysctl_lock);
+}
+
+/* Allocate a return stack for newly created task */
+void ftrace_graph_init_task(struct task_struct *t)
+{
+       if (atomic_read(&ftrace_graph_active)) {
+               t->ret_stack = kmalloc(FTRACE_RETFUNC_DEPTH
+                               * sizeof(struct ftrace_ret_stack),
+                               GFP_KERNEL);
+               if (!t->ret_stack)
+                       return;
+               t->curr_ret_stack = -1;
+               atomic_set(&t->tracing_graph_pause, 0);
+               atomic_set(&t->trace_overrun, 0);
+       } else
+               t->ret_stack = NULL;
+}
+
+void ftrace_graph_exit_task(struct task_struct *t)
+{
+       struct ftrace_ret_stack *ret_stack = t->ret_stack;
+
+       t->ret_stack = NULL;
+       /* NULL must become visible to IRQs before we free it: */
+       barrier();
+
+       kfree(ret_stack);
+}
+
+void ftrace_graph_stop(void)
+{
+       ftrace_stop();
+}
+#endif
+