nick piggin: change email address
[linux-2.6.git] / kernel / kprobes.c
index 16b5739..282035f 100644 (file)
 #include <linux/freezer.h>
 #include <linux/seq_file.h>
 #include <linux/debugfs.h>
+#include <linux/sysctl.h>
 #include <linux/kdebug.h>
 #include <linux/memory.h>
+#include <linux/ftrace.h>
+#include <linux/cpu.h>
 
 #include <asm-generic/sections.h>
 #include <asm/cacheflush.h>
@@ -90,6 +93,10 @@ static spinlock_t *kretprobe_table_lock_ptr(unsigned long hash)
  */
 static struct kprobe_blackpoint kprobe_blacklist[] = {
        {"preempt_schedule",},
+       {"native_get_debugreg",},
+       {"irq_entries_start",},
+       {"common_interrupt",},
+       {"mcount",},    /* mcount can be called from everywhere */
        {NULL}    /* Terminator */
 };
 
@@ -100,82 +107,74 @@ static struct kprobe_blackpoint kprobe_blacklist[] = {
  * stepping on the instruction on a vmalloced/kmalloced/data page
  * is a recipe for disaster
  */
-#define INSNS_PER_PAGE (PAGE_SIZE/(MAX_INSN_SIZE * sizeof(kprobe_opcode_t)))
-
 struct kprobe_insn_page {
-       struct hlist_node hlist;
+       struct list_head list;
        kprobe_opcode_t *insns;         /* Page of instruction slots */
-       char slot_used[INSNS_PER_PAGE];
        int nused;
        int ngarbage;
+       char slot_used[];
+};
+
+#define KPROBE_INSN_PAGE_SIZE(slots)                   \
+       (offsetof(struct kprobe_insn_page, slot_used) + \
+        (sizeof(char) * (slots)))
+
+struct kprobe_insn_cache {
+       struct list_head pages; /* list of kprobe_insn_page */
+       size_t insn_size;       /* size of instruction slot */
+       int nr_garbage;
 };
 
+static int slots_per_page(struct kprobe_insn_cache *c)
+{
+       return PAGE_SIZE/(c->insn_size * sizeof(kprobe_opcode_t));
+}
+
 enum kprobe_slot_state {
        SLOT_CLEAN = 0,
        SLOT_DIRTY = 1,
        SLOT_USED = 2,
 };
 
-static DEFINE_MUTEX(kprobe_insn_mutex);        /* Protects kprobe_insn_pages */
-static struct hlist_head kprobe_insn_pages;
-static int kprobe_garbage_slots;
-static int collect_garbage_slots(void);
-
-static int __kprobes check_safety(void)
-{
-       int ret = 0;
-#if defined(CONFIG_PREEMPT) && defined(CONFIG_FREEZER)
-       ret = freeze_processes();
-       if (ret == 0) {
-               struct task_struct *p, *q;
-               do_each_thread(p, q) {
-                       if (p != current && p->state == TASK_RUNNING &&
-                           p->pid != 0) {
-                               printk("Check failed: %s is running\n",p->comm);
-                               ret = -1;
-                               goto loop_end;
-                       }
-               } while_each_thread(p, q);
-       }
-loop_end:
-       thaw_processes();
-#else
-       synchronize_sched();
-#endif
-       return ret;
-}
+static DEFINE_MUTEX(kprobe_insn_mutex);        /* Protects kprobe_insn_slots */
+static struct kprobe_insn_cache kprobe_insn_slots = {
+       .pages = LIST_HEAD_INIT(kprobe_insn_slots.pages),
+       .insn_size = MAX_INSN_SIZE,
+       .nr_garbage = 0,
+};
+static int __kprobes collect_garbage_slots(struct kprobe_insn_cache *c);
 
 /**
  * __get_insn_slot() - Find a slot on an executable page for an instruction.
  * We allocate an executable page if there's no room on existing ones.
  */
-static kprobe_opcode_t __kprobes *__get_insn_slot(void)
+static kprobe_opcode_t __kprobes *__get_insn_slot(struct kprobe_insn_cache *c)
 {
        struct kprobe_insn_page *kip;
-       struct hlist_node *pos;
 
  retry:
-       hlist_for_each_entry(kip, pos, &kprobe_insn_pages, hlist) {
-               if (kip->nused < INSNS_PER_PAGE) {
+       list_for_each_entry(kip, &c->pages, list) {
+               if (kip->nused < slots_per_page(c)) {
                        int i;
-                       for (i = 0; i < INSNS_PER_PAGE; i++) {
+                       for (i = 0; i < slots_per_page(c); i++) {
                                if (kip->slot_used[i] == SLOT_CLEAN) {
                                        kip->slot_used[i] = SLOT_USED;
                                        kip->nused++;
-                                       return kip->insns + (i * MAX_INSN_SIZE);
+                                       return kip->insns + (i * c->insn_size);
                                }
                        }
-                       /* Surprise!  No unused slots.  Fix kip->nused. */
-                       kip->nused = INSNS_PER_PAGE;
+                       /* kip->nused is broken. Fix it. */
+                       kip->nused = slots_per_page(c);
+                       WARN_ON(1);
                }
        }
 
        /* If there are any garbage slots, collect it and try again. */
-       if (kprobe_garbage_slots && collect_garbage_slots() == 0) {
+       if (c->nr_garbage && collect_garbage_slots(c) == 0)
                goto retry;
-       }
-       /* All out of space.  Need to allocate a new page. Use slot 0. */
-       kip = kmalloc(sizeof(struct kprobe_insn_page), GFP_KERNEL);
+
+       /* All out of space.  Need to allocate a new page. */
+       kip = kmalloc(KPROBE_INSN_PAGE_SIZE(slots_per_page(c)), GFP_KERNEL);
        if (!kip)
                return NULL;
 
@@ -189,21 +188,24 @@ static kprobe_opcode_t __kprobes *__get_insn_slot(void)
                kfree(kip);
                return NULL;
        }
-       INIT_HLIST_NODE(&kip->hlist);
-       hlist_add_head(&kip->hlist, &kprobe_insn_pages);
-       memset(kip->slot_used, SLOT_CLEAN, INSNS_PER_PAGE);
+       INIT_LIST_HEAD(&kip->list);
+       memset(kip->slot_used, SLOT_CLEAN, slots_per_page(c));
        kip->slot_used[0] = SLOT_USED;
        kip->nused = 1;
        kip->ngarbage = 0;
+       list_add(&kip->list, &c->pages);
        return kip->insns;
 }
 
+
 kprobe_opcode_t __kprobes *get_insn_slot(void)
 {
-       kprobe_opcode_t *ret;
+       kprobe_opcode_t *ret = NULL;
+
        mutex_lock(&kprobe_insn_mutex);
-       ret = __get_insn_slot();
+       ret = __get_insn_slot(&kprobe_insn_slots);
        mutex_unlock(&kprobe_insn_mutex);
+
        return ret;
 }
 
@@ -219,12 +221,8 @@ static int __kprobes collect_one_slot(struct kprobe_insn_page *kip, int idx)
                 * so as not to have to set it up again the
                 * next time somebody inserts a probe.
                 */
-               hlist_del(&kip->hlist);
-               if (hlist_empty(&kprobe_insn_pages)) {
-                       INIT_HLIST_NODE(&kip->hlist);
-                       hlist_add_head(&kip->hlist,
-                                      &kprobe_insn_pages);
-               } else {
+               if (!list_is_singular(&kip->list)) {
+                       list_del(&kip->list);
                        module_free(NULL, kip->insns);
                        kfree(kip);
                }
@@ -233,55 +231,85 @@ static int __kprobes collect_one_slot(struct kprobe_insn_page *kip, int idx)
        return 0;
 }
 
-static int __kprobes collect_garbage_slots(void)
+static int __kprobes collect_garbage_slots(struct kprobe_insn_cache *c)
 {
-       struct kprobe_insn_page *kip;
-       struct hlist_node *pos, *next;
+       struct kprobe_insn_page *kip, *next;
 
-       /* Ensure no-one is preepmted on the garbages */
-       if (check_safety())
-               return -EAGAIN;
+       /* Ensure no-one is interrupted on the garbages */
+       synchronize_sched();
 
-       hlist_for_each_entry_safe(kip, pos, next, &kprobe_insn_pages, hlist) {
+       list_for_each_entry_safe(kip, next, &c->pages, list) {
                int i;
                if (kip->ngarbage == 0)
                        continue;
                kip->ngarbage = 0;      /* we will collect all garbages */
-               for (i = 0; i < INSNS_PER_PAGE; i++) {
+               for (i = 0; i < slots_per_page(c); i++) {
                        if (kip->slot_used[i] == SLOT_DIRTY &&
                            collect_one_slot(kip, i))
                                break;
                }
        }
-       kprobe_garbage_slots = 0;
+       c->nr_garbage = 0;
        return 0;
 }
 
-void __kprobes free_insn_slot(kprobe_opcode_t * slot, int dirty)
+static void __kprobes __free_insn_slot(struct kprobe_insn_cache *c,
+                                      kprobe_opcode_t *slot, int dirty)
 {
        struct kprobe_insn_page *kip;
-       struct hlist_node *pos;
 
-       mutex_lock(&kprobe_insn_mutex);
-       hlist_for_each_entry(kip, pos, &kprobe_insn_pages, hlist) {
-               if (kip->insns <= slot &&
-                   slot < kip->insns + (INSNS_PER_PAGE * MAX_INSN_SIZE)) {
-                       int i = (slot - kip->insns) / MAX_INSN_SIZE;
+       list_for_each_entry(kip, &c->pages, list) {
+               long idx = ((long)slot - (long)kip->insns) /
+                               (c->insn_size * sizeof(kprobe_opcode_t));
+               if (idx >= 0 && idx < slots_per_page(c)) {
+                       WARN_ON(kip->slot_used[idx] != SLOT_USED);
                        if (dirty) {
-                               kip->slot_used[i] = SLOT_DIRTY;
+                               kip->slot_used[idx] = SLOT_DIRTY;
                                kip->ngarbage++;
-                       } else {
-                               collect_one_slot(kip, i);
-                       }
-                       break;
+                               if (++c->nr_garbage > slots_per_page(c))
+                                       collect_garbage_slots(c);
+                       } else
+                               collect_one_slot(kip, idx);
+                       return;
                }
        }
+       /* Could not free this slot. */
+       WARN_ON(1);
+}
 
-       if (dirty && ++kprobe_garbage_slots > INSNS_PER_PAGE)
-               collect_garbage_slots();
-
+void __kprobes free_insn_slot(kprobe_opcode_t * slot, int dirty)
+{
+       mutex_lock(&kprobe_insn_mutex);
+       __free_insn_slot(&kprobe_insn_slots, slot, dirty);
        mutex_unlock(&kprobe_insn_mutex);
 }
+#ifdef CONFIG_OPTPROBES
+/* For optimized_kprobe buffer */
+static DEFINE_MUTEX(kprobe_optinsn_mutex); /* Protects kprobe_optinsn_slots */
+static struct kprobe_insn_cache kprobe_optinsn_slots = {
+       .pages = LIST_HEAD_INIT(kprobe_optinsn_slots.pages),
+       /* .insn_size is initialized later */
+       .nr_garbage = 0,
+};
+/* Get a slot for optimized_kprobe buffer */
+kprobe_opcode_t __kprobes *get_optinsn_slot(void)
+{
+       kprobe_opcode_t *ret = NULL;
+
+       mutex_lock(&kprobe_optinsn_mutex);
+       ret = __get_insn_slot(&kprobe_optinsn_slots);
+       mutex_unlock(&kprobe_optinsn_mutex);
+
+       return ret;
+}
+
+void __kprobes free_optinsn_slot(kprobe_opcode_t * slot, int dirty)
+{
+       mutex_lock(&kprobe_optinsn_mutex);
+       __free_insn_slot(&kprobe_optinsn_slots, slot, dirty);
+       mutex_unlock(&kprobe_optinsn_mutex);
+}
+#endif
 #endif
 
 /* We have preemption disabled.. so it is safe to use __ versions */
@@ -312,23 +340,401 @@ struct kprobe __kprobes *get_kprobe(void *addr)
                if (p->addr == addr)
                        return p;
        }
+
+       return NULL;
+}
+
+static int __kprobes aggr_pre_handler(struct kprobe *p, struct pt_regs *regs);
+
+/* Return true if the kprobe is an aggregator */
+static inline int kprobe_aggrprobe(struct kprobe *p)
+{
+       return p->pre_handler == aggr_pre_handler;
+}
+
+/*
+ * Keep all fields in the kprobe consistent
+ */
+static inline void copy_kprobe(struct kprobe *old_p, struct kprobe *p)
+{
+       memcpy(&p->opcode, &old_p->opcode, sizeof(kprobe_opcode_t));
+       memcpy(&p->ainsn, &old_p->ainsn, sizeof(struct arch_specific_insn));
+}
+
+#ifdef CONFIG_OPTPROBES
+/* NOTE: change this value only with kprobe_mutex held */
+static bool kprobes_allow_optimization;
+
+/*
+ * Call all pre_handler on the list, but ignores its return value.
+ * This must be called from arch-dep optimized caller.
+ */
+void __kprobes opt_pre_handler(struct kprobe *p, struct pt_regs *regs)
+{
+       struct kprobe *kp;
+
+       list_for_each_entry_rcu(kp, &p->list, list) {
+               if (kp->pre_handler && likely(!kprobe_disabled(kp))) {
+                       set_kprobe_instance(kp);
+                       kp->pre_handler(kp, regs);
+               }
+               reset_kprobe_instance();
+       }
+}
+
+/* Return true(!0) if the kprobe is ready for optimization. */
+static inline int kprobe_optready(struct kprobe *p)
+{
+       struct optimized_kprobe *op;
+
+       if (kprobe_aggrprobe(p)) {
+               op = container_of(p, struct optimized_kprobe, kp);
+               return arch_prepared_optinsn(&op->optinsn);
+       }
+
+       return 0;
+}
+
+/*
+ * Return an optimized kprobe whose optimizing code replaces
+ * instructions including addr (exclude breakpoint).
+ */
+struct kprobe *__kprobes get_optimized_kprobe(unsigned long addr)
+{
+       int i;
+       struct kprobe *p = NULL;
+       struct optimized_kprobe *op;
+
+       /* Don't check i == 0, since that is a breakpoint case. */
+       for (i = 1; !p && i < MAX_OPTIMIZED_LENGTH; i++)
+               p = get_kprobe((void *)(addr - i));
+
+       if (p && kprobe_optready(p)) {
+               op = container_of(p, struct optimized_kprobe, kp);
+               if (arch_within_optimized_kprobe(op, addr))
+                       return p;
+       }
+
        return NULL;
 }
 
+/* Optimization staging list, protected by kprobe_mutex */
+static LIST_HEAD(optimizing_list);
+
+static void kprobe_optimizer(struct work_struct *work);
+static DECLARE_DELAYED_WORK(optimizing_work, kprobe_optimizer);
+#define OPTIMIZE_DELAY 5
+
+/* Kprobe jump optimizer */
+static __kprobes void kprobe_optimizer(struct work_struct *work)
+{
+       struct optimized_kprobe *op, *tmp;
+
+       /* Lock modules while optimizing kprobes */
+       mutex_lock(&module_mutex);
+       mutex_lock(&kprobe_mutex);
+       if (kprobes_all_disarmed || !kprobes_allow_optimization)
+               goto end;
+
+       /*
+        * Wait for quiesence period to ensure all running interrupts
+        * are done. Because optprobe may modify multiple instructions
+        * there is a chance that Nth instruction is interrupted. In that
+        * case, running interrupt can return to 2nd-Nth byte of jump
+        * instruction. This wait is for avoiding it.
+        */
+       synchronize_sched();
+
+       /*
+        * The optimization/unoptimization refers online_cpus via
+        * stop_machine() and cpu-hotplug modifies online_cpus.
+        * And same time, text_mutex will be held in cpu-hotplug and here.
+        * This combination can cause a deadlock (cpu-hotplug try to lock
+        * text_mutex but stop_machine can not be done because online_cpus
+        * has been changed)
+        * To avoid this deadlock, we need to call get_online_cpus()
+        * for preventing cpu-hotplug outside of text_mutex locking.
+        */
+       get_online_cpus();
+       mutex_lock(&text_mutex);
+       list_for_each_entry_safe(op, tmp, &optimizing_list, list) {
+               WARN_ON(kprobe_disabled(&op->kp));
+               if (arch_optimize_kprobe(op) < 0)
+                       op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED;
+               list_del_init(&op->list);
+       }
+       mutex_unlock(&text_mutex);
+       put_online_cpus();
+end:
+       mutex_unlock(&kprobe_mutex);
+       mutex_unlock(&module_mutex);
+}
+
+/* Optimize kprobe if p is ready to be optimized */
+static __kprobes void optimize_kprobe(struct kprobe *p)
+{
+       struct optimized_kprobe *op;
+
+       /* Check if the kprobe is disabled or not ready for optimization. */
+       if (!kprobe_optready(p) || !kprobes_allow_optimization ||
+           (kprobe_disabled(p) || kprobes_all_disarmed))
+               return;
+
+       /* Both of break_handler and post_handler are not supported. */
+       if (p->break_handler || p->post_handler)
+               return;
+
+       op = container_of(p, struct optimized_kprobe, kp);
+
+       /* Check there is no other kprobes at the optimized instructions */
+       if (arch_check_optimized_kprobe(op) < 0)
+               return;
+
+       /* Check if it is already optimized. */
+       if (op->kp.flags & KPROBE_FLAG_OPTIMIZED)
+               return;
+
+       op->kp.flags |= KPROBE_FLAG_OPTIMIZED;
+       list_add(&op->list, &optimizing_list);
+       if (!delayed_work_pending(&optimizing_work))
+               schedule_delayed_work(&optimizing_work, OPTIMIZE_DELAY);
+}
+
+/* Unoptimize a kprobe if p is optimized */
+static __kprobes void unoptimize_kprobe(struct kprobe *p)
+{
+       struct optimized_kprobe *op;
+
+       if ((p->flags & KPROBE_FLAG_OPTIMIZED) && kprobe_aggrprobe(p)) {
+               op = container_of(p, struct optimized_kprobe, kp);
+               if (!list_empty(&op->list))
+                       /* Dequeue from the optimization queue */
+                       list_del_init(&op->list);
+               else
+                       /* Replace jump with break */
+                       arch_unoptimize_kprobe(op);
+               op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED;
+       }
+}
+
+/* Remove optimized instructions */
+static void __kprobes kill_optimized_kprobe(struct kprobe *p)
+{
+       struct optimized_kprobe *op;
+
+       op = container_of(p, struct optimized_kprobe, kp);
+       if (!list_empty(&op->list)) {
+               /* Dequeue from the optimization queue */
+               list_del_init(&op->list);
+               op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED;
+       }
+       /* Don't unoptimize, because the target code will be freed. */
+       arch_remove_optimized_kprobe(op);
+}
+
+/* Try to prepare optimized instructions */
+static __kprobes void prepare_optimized_kprobe(struct kprobe *p)
+{
+       struct optimized_kprobe *op;
+
+       op = container_of(p, struct optimized_kprobe, kp);
+       arch_prepare_optimized_kprobe(op);
+}
+
+/* Free optimized instructions and optimized_kprobe */
+static __kprobes void free_aggr_kprobe(struct kprobe *p)
+{
+       struct optimized_kprobe *op;
+
+       op = container_of(p, struct optimized_kprobe, kp);
+       arch_remove_optimized_kprobe(op);
+       kfree(op);
+}
+
+/* Allocate new optimized_kprobe and try to prepare optimized instructions */
+static __kprobes struct kprobe *alloc_aggr_kprobe(struct kprobe *p)
+{
+       struct optimized_kprobe *op;
+
+       op = kzalloc(sizeof(struct optimized_kprobe), GFP_KERNEL);
+       if (!op)
+               return NULL;
+
+       INIT_LIST_HEAD(&op->list);
+       op->kp.addr = p->addr;
+       arch_prepare_optimized_kprobe(op);
+
+       return &op->kp;
+}
+
+static void __kprobes init_aggr_kprobe(struct kprobe *ap, struct kprobe *p);
+
+/*
+ * Prepare an optimized_kprobe and optimize it
+ * NOTE: p must be a normal registered kprobe
+ */
+static __kprobes void try_to_optimize_kprobe(struct kprobe *p)
+{
+       struct kprobe *ap;
+       struct optimized_kprobe *op;
+
+       ap = alloc_aggr_kprobe(p);
+       if (!ap)
+               return;
+
+       op = container_of(ap, struct optimized_kprobe, kp);
+       if (!arch_prepared_optinsn(&op->optinsn)) {
+               /* If failed to setup optimizing, fallback to kprobe */
+               free_aggr_kprobe(ap);
+               return;
+       }
+
+       init_aggr_kprobe(ap, p);
+       optimize_kprobe(ap);
+}
+
+#ifdef CONFIG_SYSCTL
+static void __kprobes optimize_all_kprobes(void)
+{
+       struct hlist_head *head;
+       struct hlist_node *node;
+       struct kprobe *p;
+       unsigned int i;
+
+       /* If optimization is already allowed, just return */
+       if (kprobes_allow_optimization)
+               return;
+
+       kprobes_allow_optimization = true;
+       mutex_lock(&text_mutex);
+       for (i = 0; i < KPROBE_TABLE_SIZE; i++) {
+               head = &kprobe_table[i];
+               hlist_for_each_entry_rcu(p, node, head, hlist)
+                       if (!kprobe_disabled(p))
+                               optimize_kprobe(p);
+       }
+       mutex_unlock(&text_mutex);
+       printk(KERN_INFO "Kprobes globally optimized\n");
+}
+
+static void __kprobes unoptimize_all_kprobes(void)
+{
+       struct hlist_head *head;
+       struct hlist_node *node;
+       struct kprobe *p;
+       unsigned int i;
+
+       /* If optimization is already prohibited, just return */
+       if (!kprobes_allow_optimization)
+               return;
+
+       kprobes_allow_optimization = false;
+       printk(KERN_INFO "Kprobes globally unoptimized\n");
+       get_online_cpus();      /* For avoiding text_mutex deadlock */
+       mutex_lock(&text_mutex);
+       for (i = 0; i < KPROBE_TABLE_SIZE; i++) {
+               head = &kprobe_table[i];
+               hlist_for_each_entry_rcu(p, node, head, hlist) {
+                       if (!kprobe_disabled(p))
+                               unoptimize_kprobe(p);
+               }
+       }
+
+       mutex_unlock(&text_mutex);
+       put_online_cpus();
+       /* Allow all currently running kprobes to complete */
+       synchronize_sched();
+}
+
+int sysctl_kprobes_optimization;
+int proc_kprobes_optimization_handler(struct ctl_table *table, int write,
+                                     void __user *buffer, size_t *length,
+                                     loff_t *ppos)
+{
+       int ret;
+
+       mutex_lock(&kprobe_mutex);
+       sysctl_kprobes_optimization = kprobes_allow_optimization ? 1 : 0;
+       ret = proc_dointvec_minmax(table, write, buffer, length, ppos);
+
+       if (sysctl_kprobes_optimization)
+               optimize_all_kprobes();
+       else
+               unoptimize_all_kprobes();
+       mutex_unlock(&kprobe_mutex);
+
+       return ret;
+}
+#endif /* CONFIG_SYSCTL */
+
+static void __kprobes __arm_kprobe(struct kprobe *p)
+{
+       struct kprobe *old_p;
+
+       /* Check collision with other optimized kprobes */
+       old_p = get_optimized_kprobe((unsigned long)p->addr);
+       if (unlikely(old_p))
+               unoptimize_kprobe(old_p); /* Fallback to unoptimized kprobe */
+
+       arch_arm_kprobe(p);
+       optimize_kprobe(p);     /* Try to optimize (add kprobe to a list) */
+}
+
+static void __kprobes __disarm_kprobe(struct kprobe *p)
+{
+       struct kprobe *old_p;
+
+       unoptimize_kprobe(p);   /* Try to unoptimize */
+       arch_disarm_kprobe(p);
+
+       /* If another kprobe was blocked, optimize it. */
+       old_p = get_optimized_kprobe((unsigned long)p->addr);
+       if (unlikely(old_p))
+               optimize_kprobe(old_p);
+}
+
+#else /* !CONFIG_OPTPROBES */
+
+#define optimize_kprobe(p)                     do {} while (0)
+#define unoptimize_kprobe(p)                   do {} while (0)
+#define kill_optimized_kprobe(p)               do {} while (0)
+#define prepare_optimized_kprobe(p)            do {} while (0)
+#define try_to_optimize_kprobe(p)              do {} while (0)
+#define __arm_kprobe(p)                                arch_arm_kprobe(p)
+#define __disarm_kprobe(p)                     arch_disarm_kprobe(p)
+
+static __kprobes void free_aggr_kprobe(struct kprobe *p)
+{
+       kfree(p);
+}
+
+static __kprobes struct kprobe *alloc_aggr_kprobe(struct kprobe *p)
+{
+       return kzalloc(sizeof(struct kprobe), GFP_KERNEL);
+}
+#endif /* CONFIG_OPTPROBES */
+
 /* Arm a kprobe with text_mutex */
 static void __kprobes arm_kprobe(struct kprobe *kp)
 {
+       /*
+        * Here, since __arm_kprobe() doesn't use stop_machine(),
+        * this doesn't cause deadlock on text_mutex. So, we don't
+        * need get_online_cpus().
+        */
        mutex_lock(&text_mutex);
-       arch_arm_kprobe(kp);
+       __arm_kprobe(kp);
        mutex_unlock(&text_mutex);
 }
 
 /* Disarm a kprobe with text_mutex */
 static void __kprobes disarm_kprobe(struct kprobe *kp)
 {
+       get_online_cpus();      /* For avoiding text_mutex deadlock */
        mutex_lock(&text_mutex);
-       arch_disarm_kprobe(kp);
+       __disarm_kprobe(kp);
        mutex_unlock(&text_mutex);
+       put_online_cpus();
 }
 
 /*
@@ -397,7 +803,7 @@ static int __kprobes aggr_break_handler(struct kprobe *p, struct pt_regs *regs)
 void __kprobes kprobes_inc_nmissed_count(struct kprobe *p)
 {
        struct kprobe *kp;
-       if (p->pre_handler != aggr_pre_handler) {
+       if (!kprobe_aggrprobe(p)) {
                p->nmissed++;
        } else {
                list_for_each_entry_rcu(kp, &p->list, list)
@@ -521,21 +927,16 @@ static void __kprobes cleanup_rp_inst(struct kretprobe *rp)
 }
 
 /*
- * Keep all fields in the kprobe consistent
- */
-static inline void copy_kprobe(struct kprobe *old_p, struct kprobe *p)
-{
-       memcpy(&p->opcode, &old_p->opcode, sizeof(kprobe_opcode_t));
-       memcpy(&p->ainsn, &old_p->ainsn, sizeof(struct arch_specific_insn));
-}
-
-/*
 * Add the new probe to ap->list. Fail if this is the
 * second jprobe at the address - two jprobes can't coexist
 */
 static int __kprobes add_new_kprobe(struct kprobe *ap, struct kprobe *p)
 {
        BUG_ON(kprobe_gone(ap) || kprobe_gone(p));
+
+       if (p->break_handler || p->post_handler)
+               unoptimize_kprobe(ap);  /* Fall back to normal kprobe */
+
        if (p->break_handler) {
                if (ap->break_handler)
                        return -EEXIST;
@@ -550,7 +951,7 @@ static int __kprobes add_new_kprobe(struct kprobe *ap, struct kprobe *p)
                ap->flags &= ~KPROBE_FLAG_DISABLED;
                if (!kprobes_all_disarmed)
                        /* Arm the breakpoint again. */
-                       arm_kprobe(ap);
+                       __arm_kprobe(ap);
        }
        return 0;
 }
@@ -559,12 +960,13 @@ static int __kprobes add_new_kprobe(struct kprobe *ap, struct kprobe *p)
  * Fill in the required fields of the "manager kprobe". Replace the
  * earlier kprobe in the hlist with the manager kprobe
  */
-static inline void add_aggr_kprobe(struct kprobe *ap, struct kprobe *p)
+static void __kprobes init_aggr_kprobe(struct kprobe *ap, struct kprobe *p)
 {
+       /* Copy p's insn slot to ap */
        copy_kprobe(p, ap);
        flush_insn_slot(ap);
        ap->addr = p->addr;
-       ap->flags = p->flags;
+       ap->flags = p->flags & ~KPROBE_FLAG_OPTIMIZED;
        ap->pre_handler = aggr_pre_handler;
        ap->fault_handler = aggr_fault_handler;
        /* We don't care the kprobe which has gone. */
@@ -574,8 +976,9 @@ static inline void add_aggr_kprobe(struct kprobe *ap, struct kprobe *p)
                ap->break_handler = aggr_break_handler;
 
        INIT_LIST_HEAD(&ap->list);
-       list_add_rcu(&p->list, &ap->list);
+       INIT_HLIST_NODE(&ap->hlist);
 
+       list_add_rcu(&p->list, &ap->list);
        hlist_replace_rcu(&p->hlist, &ap->hlist);
 }
 
@@ -589,12 +992,12 @@ static int __kprobes register_aggr_kprobe(struct kprobe *old_p,
        int ret = 0;
        struct kprobe *ap = old_p;
 
-       if (old_p->pre_handler != aggr_pre_handler) {
-               /* If old_p is not an aggr_probe, create new aggr_kprobe. */
-               ap = kzalloc(sizeof(struct kprobe), GFP_KERNEL);
+       if (!kprobe_aggrprobe(old_p)) {
+               /* If old_p is not an aggr_kprobe, create new aggr_kprobe. */
+               ap = alloc_aggr_kprobe(old_p);
                if (!ap)
                        return -ENOMEM;
-               add_aggr_kprobe(ap, old_p);
+               init_aggr_kprobe(ap, old_p);
        }
 
        if (kprobe_gone(ap)) {
@@ -613,6 +1016,9 @@ static int __kprobes register_aggr_kprobe(struct kprobe *old_p,
                         */
                        return ret;
 
+               /* Prepare optimized instructions if possible. */
+               prepare_optimized_kprobe(ap);
+
                /*
                 * Clear gone flag to prevent allocating new slot again, and
                 * set disabled flag because it is not armed yet.
@@ -621,6 +1027,7 @@ static int __kprobes register_aggr_kprobe(struct kprobe *old_p,
                            | KPROBE_FLAG_DISABLED;
        }
 
+       /* Copy ap's insn slot to p */
        copy_kprobe(ap, p);
        return add_new_kprobe(ap, p);
 }
@@ -681,6 +1088,40 @@ static kprobe_opcode_t __kprobes *kprobe_addr(struct kprobe *p)
        return (kprobe_opcode_t *)(((char *)addr) + p->offset);
 }
 
+/* Check passed kprobe is valid and return kprobe in kprobe_table. */
+static struct kprobe * __kprobes __get_valid_kprobe(struct kprobe *p)
+{
+       struct kprobe *old_p, *list_p;
+
+       old_p = get_kprobe(p->addr);
+       if (unlikely(!old_p))
+               return NULL;
+
+       if (p != old_p) {
+               list_for_each_entry_rcu(list_p, &old_p->list, list)
+                       if (list_p == p)
+                       /* kprobe p is a valid probe */
+                               goto valid;
+               return NULL;
+       }
+valid:
+       return old_p;
+}
+
+/* Return error if the kprobe is being re-registered */
+static inline int check_kprobe_rereg(struct kprobe *p)
+{
+       int ret = 0;
+       struct kprobe *old_p;
+
+       mutex_lock(&kprobe_mutex);
+       old_p = __get_valid_kprobe(p);
+       if (old_p)
+               ret = -EINVAL;
+       mutex_unlock(&kprobe_mutex);
+       return ret;
+}
+
 int __kprobes register_kprobe(struct kprobe *p)
 {
        int ret = 0;
@@ -693,9 +1134,14 @@ int __kprobes register_kprobe(struct kprobe *p)
                return -EINVAL;
        p->addr = addr;
 
+       ret = check_kprobe_rereg(p);
+       if (ret)
+               return ret;
+
        preempt_disable();
-       if (!__kernel_text_address((unsigned long) p->addr) ||
-           in_kprobes_functions((unsigned long) p->addr)) {
+       if (!kernel_text_address((unsigned long) p->addr) ||
+           in_kprobes_functions((unsigned long) p->addr) ||
+           ftrace_text_reserved(p->addr, p->addr)) {
                preempt_enable();
                return -EINVAL;
        }
@@ -732,27 +1178,34 @@ int __kprobes register_kprobe(struct kprobe *p)
        p->nmissed = 0;
        INIT_LIST_HEAD(&p->list);
        mutex_lock(&kprobe_mutex);
+
+       get_online_cpus();      /* For avoiding text_mutex deadlock. */
+       mutex_lock(&text_mutex);
+
        old_p = get_kprobe(p->addr);
        if (old_p) {
+               /* Since this may unoptimize old_p, locking text_mutex. */
                ret = register_aggr_kprobe(old_p, p);
                goto out;
        }
 
-       mutex_lock(&text_mutex);
        ret = arch_prepare_kprobe(p);
        if (ret)
-               goto out_unlock_text;
+               goto out;
 
        INIT_HLIST_NODE(&p->hlist);
        hlist_add_head_rcu(&p->hlist,
                       &kprobe_table[hash_ptr(p->addr, KPROBE_HASH_BITS)]);
 
        if (!kprobes_all_disarmed && !kprobe_disabled(p))
-               arch_arm_kprobe(p);
+               __arm_kprobe(p);
+
+       /* Try to optimize kprobe */
+       try_to_optimize_kprobe(p);
 
-out_unlock_text:
-       mutex_unlock(&text_mutex);
 out:
+       mutex_unlock(&text_mutex);
+       put_online_cpus();
        mutex_unlock(&kprobe_mutex);
 
        if (probed_mod)
@@ -762,26 +1215,6 @@ out:
 }
 EXPORT_SYMBOL_GPL(register_kprobe);
 
-/* Check passed kprobe is valid and return kprobe in kprobe_table. */
-static struct kprobe * __kprobes __get_valid_kprobe(struct kprobe *p)
-{
-       struct kprobe *old_p, *list_p;
-
-       old_p = get_kprobe(p->addr);
-       if (unlikely(!old_p))
-               return NULL;
-
-       if (p != old_p) {
-               list_for_each_entry_rcu(list_p, &old_p->list, list)
-                       if (list_p == p)
-                       /* kprobe p is a valid probe */
-                               goto valid;
-               return NULL;
-       }
-valid:
-       return old_p;
-}
-
 /*
  * Unregister a kprobe without a scheduler synchronization.
  */
@@ -794,7 +1227,7 @@ static int __kprobes __unregister_kprobe_top(struct kprobe *p)
                return -EINVAL;
 
        if (old_p == p ||
-           (old_p->pre_handler == aggr_pre_handler &&
+           (kprobe_aggrprobe(old_p) &&
             list_is_singular(&old_p->list))) {
                /*
                 * Only probe on the hash list. Disarm only if kprobes are
@@ -802,7 +1235,7 @@ static int __kprobes __unregister_kprobe_top(struct kprobe *p)
                 * already have been removed. We save on flushing icache.
                 */
                if (!kprobes_all_disarmed && !kprobe_disabled(old_p))
-                       disarm_kprobe(p);
+                       disarm_kprobe(old_p);
                hlist_del_rcu(&old_p->hlist);
        } else {
                if (p->break_handler && !kprobe_gone(p))
@@ -818,8 +1251,13 @@ noclean:
                list_del_rcu(&p->list);
                if (!kprobe_disabled(old_p)) {
                        try_to_disable_aggr_kprobe(old_p);
-                       if (!kprobes_all_disarmed && kprobe_disabled(old_p))
-                               disarm_kprobe(old_p);
+                       if (!kprobes_all_disarmed) {
+                               if (kprobe_disabled(old_p))
+                                       disarm_kprobe(old_p);
+                               else
+                                       /* Try to optimize this probe again */
+                                       optimize_kprobe(old_p);
+                       }
                }
        }
        return 0;
@@ -836,7 +1274,7 @@ static void __kprobes __unregister_kprobe_bottom(struct kprobe *p)
                old_p = list_entry(p->list.next, struct kprobe, list);
                list_del(&p->list);
                arch_remove_kprobe(old_p);
-               kfree(old_p);
+               free_aggr_kprobe(old_p);
        }
 }
 
@@ -1022,9 +1460,9 @@ int __kprobes register_kretprobe(struct kretprobe *rp)
        /* Pre-allocate memory for max kretprobe instances */
        if (rp->maxactive <= 0) {
 #ifdef CONFIG_PREEMPT
-               rp->maxactive = max(10, 2 * NR_CPUS);
+               rp->maxactive = max_t(unsigned int, 10, 2*num_possible_cpus());
 #else
-               rp->maxactive = NR_CPUS;
+               rp->maxactive = num_possible_cpus();
 #endif
        }
        spin_lock_init(&rp->lock);
@@ -1132,7 +1570,7 @@ static void __kprobes kill_kprobe(struct kprobe *p)
        struct kprobe *kp;
 
        p->flags |= KPROBE_FLAG_GONE;
-       if (p->pre_handler == aggr_pre_handler) {
+       if (kprobe_aggrprobe(p)) {
                /*
                 * If this is an aggr_kprobe, we have to list all the
                 * chained probes and mark them GONE.
@@ -1141,6 +1579,7 @@ static void __kprobes kill_kprobe(struct kprobe *p)
                        kp->flags |= KPROBE_FLAG_GONE;
                p->post_handler = NULL;
                p->break_handler = NULL;
+               kill_optimized_kprobe(p);
        }
        /*
         * Here, we can remove insn_slot safely, because no thread calls
@@ -1149,6 +1588,79 @@ static void __kprobes kill_kprobe(struct kprobe *p)
        arch_remove_kprobe(p);
 }
 
+/* Disable one kprobe */
+int __kprobes disable_kprobe(struct kprobe *kp)
+{
+       int ret = 0;
+       struct kprobe *p;
+
+       mutex_lock(&kprobe_mutex);
+
+       /* Check whether specified probe is valid. */
+       p = __get_valid_kprobe(kp);
+       if (unlikely(p == NULL)) {
+               ret = -EINVAL;
+               goto out;
+       }
+
+       /* If the probe is already disabled (or gone), just return */
+       if (kprobe_disabled(kp))
+               goto out;
+
+       kp->flags |= KPROBE_FLAG_DISABLED;
+       if (p != kp)
+               /* When kp != p, p is always enabled. */
+               try_to_disable_aggr_kprobe(p);
+
+       if (!kprobes_all_disarmed && kprobe_disabled(p))
+               disarm_kprobe(p);
+out:
+       mutex_unlock(&kprobe_mutex);
+       return ret;
+}
+EXPORT_SYMBOL_GPL(disable_kprobe);
+
+/* Enable one kprobe */
+int __kprobes enable_kprobe(struct kprobe *kp)
+{
+       int ret = 0;
+       struct kprobe *p;
+
+       mutex_lock(&kprobe_mutex);
+
+       /* Check whether specified probe is valid. */
+       p = __get_valid_kprobe(kp);
+       if (unlikely(p == NULL)) {
+               ret = -EINVAL;
+               goto out;
+       }
+
+       if (kprobe_gone(kp)) {
+               /* This kprobe has gone, we couldn't enable it. */
+               ret = -EINVAL;
+               goto out;
+       }
+
+       if (p != kp)
+               kp->flags &= ~KPROBE_FLAG_DISABLED;
+
+       if (!kprobes_all_disarmed && kprobe_disabled(p)) {
+               p->flags &= ~KPROBE_FLAG_DISABLED;
+               arm_kprobe(p);
+       }
+out:
+       mutex_unlock(&kprobe_mutex);
+       return ret;
+}
+EXPORT_SYMBOL_GPL(enable_kprobe);
+
+void __kprobes dump_kprobe(struct kprobe *kp)
+{
+       printk(KERN_WARNING "Dumping kprobe:\n");
+       printk(KERN_WARNING "Name: %s\nAddress: %p\nOffset: %x\n",
+              kp->symbol_name, kp->addr, kp->offset);
+}
+
 /* Module notifier call back, checking kprobes on the module */
 static int __kprobes kprobes_module_callback(struct notifier_block *nb,
                                             unsigned long val, void *data)
@@ -1243,6 +1755,15 @@ static int __init init_kprobes(void)
                }
        }
 
+#if defined(CONFIG_OPTPROBES)
+#if defined(__ARCH_WANT_KPROBES_INSN_SLOT)
+       /* Init kprobe_optinsn_slots */
+       kprobe_optinsn_slots.insn_size = MAX_OPTINSN_SIZE;
+#endif
+       /* By default, kprobes can be optimized */
+       kprobes_allow_optimization = true;
+#endif
+
        /* By default, kprobes are armed */
        kprobes_all_disarmed = false;
 
@@ -1261,7 +1782,7 @@ static int __init init_kprobes(void)
 
 #ifdef CONFIG_DEBUG_FS
 static void __kprobes report_probe(struct seq_file *pi, struct kprobe *p,
-               const char *sym, int offset,char *modname)
+               const char *sym, int offset, char *modname, struct kprobe *pp)
 {
        char *kprobe_type;
 
@@ -1271,19 +1792,21 @@ static void __kprobes report_probe(struct seq_file *pi, struct kprobe *p,
                kprobe_type = "j";
        else
                kprobe_type = "k";
+
        if (sym)
-               seq_printf(pi, "%p  %s  %s+0x%x  %s %s%s\n",
+               seq_printf(pi, "%p  %s  %s+0x%x  %s ",
                        p->addr, kprobe_type, sym, offset,
-                       (modname ? modname : " "),
-                       (kprobe_gone(p) ? "[GONE]" : ""),
-                       ((kprobe_disabled(p) && !kprobe_gone(p)) ?
-                        "[DISABLED]" : ""));
+                       (modname ? modname : " "));
        else
-               seq_printf(pi, "%p  %s  %p %s%s\n",
-                       p->addr, kprobe_type, p->addr,
-                       (kprobe_gone(p) ? "[GONE]" : ""),
-                       ((kprobe_disabled(p) && !kprobe_gone(p)) ?
-                        "[DISABLED]" : ""));
+               seq_printf(pi, "%p  %s  %p ",
+                       p->addr, kprobe_type, p->addr);
+
+       if (!pp)
+               pp = p;
+       seq_printf(pi, "%s%s%s\n",
+               (kprobe_gone(p) ? "[GONE]" : ""),
+               ((kprobe_disabled(p) && !kprobe_gone(p)) ?  "[DISABLED]" : ""),
+               (kprobe_optimized(pp) ? "[OPTIMIZED]" : ""));
 }
 
 static void __kprobes *kprobe_seq_start(struct seq_file *f, loff_t *pos)
@@ -1319,17 +1842,17 @@ static int __kprobes show_kprobe_addr(struct seq_file *pi, void *v)
        hlist_for_each_entry_rcu(p, node, head, hlist) {
                sym = kallsyms_lookup((unsigned long)p->addr, NULL,
                                        &offset, &modname, namebuf);
-               if (p->pre_handler == aggr_pre_handler) {
+               if (kprobe_aggrprobe(p)) {
                        list_for_each_entry_rcu(kp, &p->list, list)
-                               report_probe(pi, kp, sym, offset, modname);
+                               report_probe(pi, kp, sym, offset, modname, p);
                } else
-                       report_probe(pi, p, sym, offset, modname);
+                       report_probe(pi, p, sym, offset, modname, NULL);
        }
        preempt_enable();
        return 0;
 }
 
-static struct seq_operations kprobes_seq_ops = {
+static const struct seq_operations kprobes_seq_ops = {
        .start = kprobe_seq_start,
        .next  = kprobe_seq_next,
        .stop  = kprobe_seq_stop,
@@ -1341,78 +1864,13 @@ static int __kprobes kprobes_open(struct inode *inode, struct file *filp)
        return seq_open(filp, &kprobes_seq_ops);
 }
 
-static struct file_operations debugfs_kprobes_operations = {
+static const struct file_operations debugfs_kprobes_operations = {
        .open           = kprobes_open,
        .read           = seq_read,
        .llseek         = seq_lseek,
        .release        = seq_release,
 };
 
-/* Disable one kprobe */
-int __kprobes disable_kprobe(struct kprobe *kp)
-{
-       int ret = 0;
-       struct kprobe *p;
-
-       mutex_lock(&kprobe_mutex);
-
-       /* Check whether specified probe is valid. */
-       p = __get_valid_kprobe(kp);
-       if (unlikely(p == NULL)) {
-               ret = -EINVAL;
-               goto out;
-       }
-
-       /* If the probe is already disabled (or gone), just return */
-       if (kprobe_disabled(kp))
-               goto out;
-
-       kp->flags |= KPROBE_FLAG_DISABLED;
-       if (p != kp)
-               /* When kp != p, p is always enabled. */
-               try_to_disable_aggr_kprobe(p);
-
-       if (!kprobes_all_disarmed && kprobe_disabled(p))
-               disarm_kprobe(p);
-out:
-       mutex_unlock(&kprobe_mutex);
-       return ret;
-}
-EXPORT_SYMBOL_GPL(disable_kprobe);
-
-/* Enable one kprobe */
-int __kprobes enable_kprobe(struct kprobe *kp)
-{
-       int ret = 0;
-       struct kprobe *p;
-
-       mutex_lock(&kprobe_mutex);
-
-       /* Check whether specified probe is valid. */
-       p = __get_valid_kprobe(kp);
-       if (unlikely(p == NULL)) {
-               ret = -EINVAL;
-               goto out;
-       }
-
-       if (kprobe_gone(kp)) {
-               /* This kprobe has gone, we couldn't enable it. */
-               ret = -EINVAL;
-               goto out;
-       }
-
-       if (!kprobes_all_disarmed && kprobe_disabled(p))
-               arm_kprobe(p);
-
-       p->flags &= ~KPROBE_FLAG_DISABLED;
-       if (p != kp)
-               kp->flags &= ~KPROBE_FLAG_DISABLED;
-out:
-       mutex_unlock(&kprobe_mutex);
-       return ret;
-}
-EXPORT_SYMBOL_GPL(enable_kprobe);
-
 static void __kprobes arm_all_kprobes(void)
 {
        struct hlist_head *head;
@@ -1426,12 +1884,13 @@ static void __kprobes arm_all_kprobes(void)
        if (!kprobes_all_disarmed)
                goto already_enabled;
 
+       /* Arming kprobes doesn't optimize kprobe itself */
        mutex_lock(&text_mutex);
        for (i = 0; i < KPROBE_TABLE_SIZE; i++) {
                head = &kprobe_table[i];
                hlist_for_each_entry_rcu(p, node, head, hlist)
                        if (!kprobe_disabled(p))
-                               arch_arm_kprobe(p);
+                               __arm_kprobe(p);
        }
        mutex_unlock(&text_mutex);
 
@@ -1458,16 +1917,23 @@ static void __kprobes disarm_all_kprobes(void)
 
        kprobes_all_disarmed = true;
        printk(KERN_INFO "Kprobes globally disabled\n");
+
+       /*
+        * Here we call get_online_cpus() for avoiding text_mutex deadlock,
+        * because disarming may also unoptimize kprobes.
+        */
+       get_online_cpus();
        mutex_lock(&text_mutex);
        for (i = 0; i < KPROBE_TABLE_SIZE; i++) {
                head = &kprobe_table[i];
                hlist_for_each_entry_rcu(p, node, head, hlist) {
                        if (!arch_trampoline_kprobe(p) && !kprobe_disabled(p))
-                               arch_disarm_kprobe(p);
+                               __disarm_kprobe(p);
                }
        }
 
        mutex_unlock(&text_mutex);
+       put_online_cpus();
        mutex_unlock(&kprobe_mutex);
        /* Allow all currently running kprobes to complete */
        synchronize_sched();
@@ -1523,7 +1989,7 @@ static ssize_t write_enabled_file_bool(struct file *file,
        return count;
 }
 
-static struct file_operations fops_kp = {
+static const struct file_operations fops_kp = {
        .read =         read_enabled_file_bool,
        .write =        write_enabled_file_bool,
 };