kprobes: Unify kretprobe trampoline handlers and make kretprobe lockless

index 9be1bff4f586..c6a913e608b7 100644
@@ -187,10 +187,38 @@ static inline int kprobes_built_in(void)
 	return 1;
 }
 
+extern struct kprobe kprobe_busy;
+void kprobe_busy_begin(void);
+void kprobe_busy_end(void);
+
 #ifdef CONFIG_KRETPROBES
 extern void arch_prepare_kretprobe(struct kretprobe_instance *ri,
 				   struct pt_regs *regs);
 extern int arch_trampoline_kprobe(struct kprobe *p);
+
+/* If the trampoline handler called from a kprobe, use this version */
+unsigned long __kretprobe_trampoline_handler(struct pt_regs *regs,
+				void *trampoline_address,
+				void *frame_pointer);
+
+static nokprobe_inline
+unsigned long kretprobe_trampoline_handler(struct pt_regs *regs,
+				void *trampoline_address,
+				void *frame_pointer)
+{
+	unsigned long ret;
+	/*
+	 * Set a dummy kprobe for avoiding kretprobe recursion.
+	 * Since kretprobe never runs in kprobe handler, any kprobe must not
+	 * be running at this point.
+	 */
+	kprobe_busy_begin();
+	ret = __kretprobe_trampoline_handler(regs, trampoline_address, frame_pointer);
+	kprobe_busy_end();
+
+	return ret;
+}
+
 #else /* CONFIG_KRETPROBES */
 static inline void arch_prepare_kretprobe(struct kretprobe *rp,
 					struct pt_regs *regs)

@@ -354,10 +382,6 @@ static inline struct kprobe_ctlblk *get_kprobe_ctlblk(void)
 	return this_cpu_ptr(&kprobe_ctlblk);
 }
 
-extern struct kprobe kprobe_busy;
-void kprobe_busy_begin(void);
-void kprobe_busy_end(void);
-
 kprobe_opcode_t *kprobe_lookup_name(const char *name, unsigned int offset);
 int register_kprobe(struct kprobe *p);
 void unregister_kprobe(struct kprobe *p);

102 lines not on this page · raw message

index fdadc37d72af..882b95313ad6 100644
@@ -767,124 +767,22 @@ asm(
 NOKPROBE_SYMBOL(kretprobe_trampoline);
 STACK_FRAME_NON_STANDARD(kretprobe_trampoline);
 
+
 /*
  * Called from kretprobe_trampoline
  */
 __used __visible void *trampoline_handler(struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline;
-	kprobe_opcode_t *correct_ret_addr = NULL;
-	void *frame_pointer;
-	bool skipped = false;
-
-	/*
-	 * Set a dummy kprobe for avoiding kretprobe recursion.
-	 * Since kretprobe never run in kprobe handler, kprobe must not
-	 * be running at this point.
-	 */
-	kprobe_busy_begin();
-
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
 	/* fixup registers */
 	regs->cs = __KERNEL_CS;
 #ifdef CONFIG_X86_32
 	regs->cs |= get_kernel_rpl();
 	regs->gs = 0;
 #endif
-	/* We use pt_regs->sp for return address holder. */
-	frame_pointer = &regs->sp;
-	regs->ip = trampoline_address;
+	regs->ip = (unsigned long)&kretprobe_trampoline;
 	regs->orig_ax = ~0UL;
 
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because multiple functions in the call path have
-	 * return probes installed on them, and/or more than one
-	 * return probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always pushed into the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *	 function, the (chronologically) first instance's ret_addr
-	 *	 will be the real return address, and all the rest will
-	 *	 point to kretprobe_trampoline.
-	 */
-	hlist_for_each_entry(ri, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-		/*
-		 * Return probes must be pushed on this hash list correct
-		 * order (same as return order) so that it can be popped
-		 * correctly. However, if we find it is pushed it incorrect
-		 * order, this means we find a function which should not be
-		 * probed, because the wrong order entry is pushed on the
-		 * path of processing other kretprobe itself.
-		 */
-		if (ri->fp != frame_pointer) {
-			if (!skipped)
-				pr_warn("kretprobe is stacked incorrectly. Trying to fixup.\n");
-			skipped = true;
-			continue;
-		}
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		if (skipped)
-			pr_warn("%ps must be blacklisted because of incorrect kretprobe order\n",
-				ri->rp->kp.addr);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-
-	correct_ret_addr = ri->ret_addr;
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-		if (ri->fp != frame_pointer)
-			continue;
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		if (ri->rp && ri->rp->handler) {
-			__this_cpu_write(current_kprobe, &ri->rp->kp);
-			ri->ret_addr = correct_ret_addr;
-			ri->rp->handler(ri, regs);
-			__this_cpu_write(current_kprobe, &kprobe_busy);
-		}
-
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_hash_unlock(current, &flags);
-
-	kprobe_busy_end();
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
-	return (void *)orig_ret_address;
+	return (void *)kretprobe_trampoline_handler(regs, &kretprobe_trampoline, &regs->sp);
 }
 NOKPROBE_SYMBOL(trampoline_handler);
 

index feefa2055eba..a9653117ca0d 100644
@@ -413,87 +413,15 @@ void __naked __kprobes kretprobe_trampoline(void)
 /* Called from kretprobe_trampoline */
 static __used __kprobes void *trampoline_handler(struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline;
-	kprobe_opcode_t *correct_ret_addr = NULL;
-
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because multiple functions in the call path have
-	 * a return probe installed on them, and/or more than one return
-	 * probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always inserted at the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *       function, the first instance's ret_addr will point to the
-	 *       real return address, and all the rest will point to
-	 *       kretprobe_trampoline
-	 */
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-
-	correct_ret_addr = ri->ret_addr;
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		if (ri->rp && ri->rp->handler) {
-			__this_cpu_write(current_kprobe, &ri->rp->kp);
-			get_kprobe_ctlblk()->kprobe_status = KPROBE_HIT_ACTIVE;
-			ri->ret_addr = correct_ret_addr;
-			ri->rp->handler(ri, regs);
-			__this_cpu_write(current_kprobe, NULL);
-		}
-
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
-
-	return (void *)orig_ret_address;
+	return (void *)kretprobe_trampoline_handler(regs, &kretprobe_trampoline,
+						    (void *)regs->ARM_fp);
 }
 
 void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 				      struct pt_regs *regs)
 {
 	ri->ret_addr = (kprobe_opcode_t *)regs->ARM_lr;
+	ri->fp = (void *)regs->ARM_fp;
 
 	/* Replace the return addr with trampoline addr. */
 	regs->ARM_lr = (unsigned long)&kretprobe_trampoline;

index 5290f17a4d80..deba738142ed 100644
@@ -464,87 +464,15 @@ int __init arch_populate_kprobe_blacklist(void)
 
 void __kprobes __used *trampoline_probe_handler(struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address =
-		(unsigned long)&kretprobe_trampoline;
-	kprobe_opcode_t *correct_ret_addr = NULL;
-
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because multiple functions in the call path have
-	 * return probes installed on them, and/or more than one
-	 * return probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always pushed into the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *	 function, the (chronologically) first instance's ret_addr
-	 *	 will be the real return address, and all the rest will
-	 *	 point to kretprobe_trampoline.
-	 */
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-
-	correct_ret_addr = ri->ret_addr;
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		if (ri->rp && ri->rp->handler) {
-			__this_cpu_write(current_kprobe, &ri->rp->kp);
-			get_kprobe_ctlblk()->kprobe_status = KPROBE_HIT_ACTIVE;
-			ri->ret_addr = correct_ret_addr;
-			ri->rp->handler(ri, regs);
-			__this_cpu_write(current_kprobe, NULL);
-		}
-
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
-	return (void *)orig_ret_address;
+	return (void *)kretprobe_trampoline_handler(regs, &kretprobe_trampoline,
+					(void *)kernel_stack_pointer(regs));
 }
 
 void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 				      struct pt_regs *regs)
 {
 	ri->ret_addr = (kprobe_opcode_t *)regs->regs[30];
+	ri->fp = (void *)kernel_stack_pointer(regs);
 
 	/* replace return addr (x30) with trampoline */
 	regs->regs[30] = (long)&kretprobe_trampoline;

index 7d3efe83cba7..cabef45f11df 100644
@@ -388,6 +388,7 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 {
 
 	ri->ret_addr = (kprobe_opcode_t *) regs->blink;
+	ri->fp = NULL;
 
 	/* Replace the return addr with trampoline addr */
 	regs->blink = (unsigned long)&kretprobe_trampoline;

@@ -396,58 +397,7 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 static int __kprobes trampoline_probe_handler(struct kprobe *p,
 					      struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline;
-
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because an multiple functions in the call path
-	 * have a return probe installed on them, and/or more than one return
-	 * return probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always inserted at the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *       function, the first instance's ret_addr will point to the
-	 *       real return address, and all the rest will point to
-	 *       kretprobe_trampoline
-	 */
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		if (ri->rp && ri->rp->handler)
-			ri->rp->handler(ri, regs);
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address) {
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-		}
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-	regs->ret = orig_ret_address;
-
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
+	regs->ret = __kretprobe_trampoline_handler(regs, &kretprobe_trampoline, NULL);
 
 	/* By returning a non zero value, we are telling the kprobe handler
 	 * that we don't want the post_handler to run

index f0f733b7ac5a..589f090f48b9 100644
@@ -404,87 +404,14 @@ int __init arch_populate_kprobe_blacklist(void)
 
 void __kprobes __used *trampoline_probe_handler(struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address =
-		(unsigned long)&kretprobe_trampoline;
-	kprobe_opcode_t *correct_ret_addr = NULL;
-
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because multiple functions in the call path have
-	 * return probes installed on them, and/or more than one
-	 * return probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always pushed into the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *	 function, the (chronologically) first instance's ret_addr
-	 *	 will be the real return address, and all the rest will
-	 *	 point to kretprobe_trampoline.
-	 */
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-
-	correct_ret_addr = ri->ret_addr;
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		if (ri->rp && ri->rp->handler) {
-			__this_cpu_write(current_kprobe, &ri->rp->kp);
-			get_kprobe_ctlblk()->kprobe_status = KPROBE_HIT_ACTIVE;
-			ri->ret_addr = correct_ret_addr;
-			ri->rp->handler(ri, regs);
-			__this_cpu_write(current_kprobe, NULL);
-		}
-
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
-	return (void *)orig_ret_address;
+	return (void *)kretprobe_trampoline_handler(regs, &kretprobe_trampoline, NULL);
 }
 
 void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 				      struct pt_regs *regs)
 {
 	ri->ret_addr = (kprobe_opcode_t *)regs->lr;
+	ri->fp = NULL;
 	regs->lr = (unsigned long) &kretprobe_trampoline;
 }
 

index 7a7df944d798..fc1ff8a4d7de 100644
@@ -396,83 +396,9 @@ static void kretprobe_trampoline(void)
 {
 }
 
-/*
- * At this point the target function has been tricked into
- * returning into our trampoline.  Lookup the associated instance
- * and then:
- *    - call the handler function
- *    - cleanup by marking the instance as unused
- *    - long jump back to the original return address
- */
 int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address =
-		((struct fnptr *)kretprobe_trampoline)->ip;
-
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because an multiple functions in the call path
-	 * have a return probe installed on them, and/or more than one return
-	 * return probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always inserted at the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *       function, the first instance's ret_addr will point to the
-	 *       real return address, and all the rest will point to
-	 *       kretprobe_trampoline
-	 */
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	regs->cr_iip = orig_ret_address;
-
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		if (ri->rp && ri->rp->handler)
-			ri->rp->handler(ri, regs);
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
+	regs->cr_iip = __kretprobe_trampoline_handler(regs, kretprobe_trampoline, NULL);
 	/*
 	 * By returning a non-zero value, we are telling
 	 * kprobe_handler() that we don't want the post_handler

@@ -485,6 +411,7 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 				      struct pt_regs *regs)
 {
 	ri->ret_addr = (kprobe_opcode_t *)regs->b0;
+	ri->fp = NULL;
 
 	/* Replace the return addr with trampoline addr */
 	regs->b0 = ((struct fnptr *)kretprobe_trampoline)->ip;

index d043c2f897fc..54dfba8fa77c 100644
@@ -477,6 +477,7 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 				      struct pt_regs *regs)
 {
 	ri->ret_addr = (kprobe_opcode_t *) regs->regs[31];
+	ri->fp = NULL;
 
 	/* Replace the return addr with trampoline addr */
 	regs->regs[31] = (unsigned long)kretprobe_trampoline;

@@ -488,57 +489,8 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 static int __kprobes trampoline_probe_handler(struct kprobe *p,
 						struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address = (unsigned long)kretprobe_trampoline;
-
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because an multiple functions in the call path
-	 * have a return probe installed on them, and/or more than one return
-	 * return probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always inserted at the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *	 function, the first instance's ret_addr will point to the
-	 *	 real return address, and all the rest will point to
-	 *	 kretprobe_trampoline
-	 */
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		if (ri->rp && ri->rp->handler)
-			ri->rp->handler(ri, regs);
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-	instruction_pointer(regs) = orig_ret_address;
-
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
+	instruction_pointer(regs) = __kretprobe_trampoline_handler(regs,
+						kretprobe_trampoline, NULL);
 	/*
 	 * By returning a non-zero value, we are telling
 	 * kprobe_handler() that we don't want the post_handler

index 77ec51818916..6d21a515eea5 100644
@@ -191,80 +191,11 @@ static struct kprobe trampoline_p = {
 static int __kprobes trampoline_probe_handler(struct kprobe *p,
 					      struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address = (unsigned long)trampoline_p.addr;
-	kprobe_opcode_t *correct_ret_addr = NULL;
-
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because multiple functions in the call path have
-	 * a return probe installed on them, and/or more than one return
-	 * probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always inserted at the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *       function, the first instance's ret_addr will point to the
-	 *       real return address, and all the rest will point to
-	 *       kretprobe_trampoline
-	 */
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-
-	correct_ret_addr = ri->ret_addr;
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		if (ri->rp && ri->rp->handler) {
-			__this_cpu_write(current_kprobe, &ri->rp->kp);
-			get_kprobe_ctlblk()->kprobe_status = KPROBE_HIT_ACTIVE;
-			ri->ret_addr = correct_ret_addr;
-			ri->rp->handler(ri, regs);
-			__this_cpu_write(current_kprobe, NULL);
-		}
-
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
+	unsigned long orig_ret_address;
 
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
+	orig_ret_address = __kretprobe_trampoline_handler(regs, trampoline_p.addr, NULL);
 	instruction_pointer_set(regs, orig_ret_address);
+
 	return 1;
 }
 

@@ -272,6 +203,7 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 				      struct pt_regs *regs)
 {
 	ri->ret_addr = (kprobe_opcode_t *)regs->gr[2];
+	ri->fp = NULL;
 
 	/* Replace the return addr with trampoline addr. */
 	regs->gr[2] = (unsigned long)trampoline_p.addr;

index 6ab9b4d037c3..01ab2163659e 100644
@@ -218,6 +218,7 @@ bool arch_kprobe_on_func_entry(unsigned long offset)
 void arch_prepare_kretprobe(struct kretprobe_instance *ri, struct pt_regs *regs)
 {
 	ri->ret_addr = (kprobe_opcode_t *)regs->link;
+	ri->fp = NULL;
 
 	/* Replace the return addr with trampoline addr */
 	regs->link = (unsigned long)kretprobe_trampoline;

@@ -396,50 +397,9 @@ asm(".global kretprobe_trampoline\n"
  */
 static int trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address =(unsigned long)&kretprobe_trampoline;
-
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because an multiple functions in the call path
-	 * have a return probe installed on them, and/or more than one return
-	 * return probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always inserted at the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *       function, the first instance's ret_addr will point to the
-	 *       real return address, and all the rest will point to
-	 *       kretprobe_trampoline
-	 */
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		if (ri->rp && ri->rp->handler)
-			ri->rp->handler(ri, regs);
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
+	unsigned long orig_ret_address;
 
+	orig_ret_address = __kretprobe_trampoline_handler(regs, &kretprobe_trampoline, NULL);
 	/*
 	 * We get here through one of two paths:
 	 * 1. by taking a trap -> kprobe_handler() -> here

@@ -458,13 +418,6 @@ static int trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
 	regs->nip = orig_ret_address - 4;
 	regs->link = orig_ret_address;
 
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
-
 	return 0;
 }
 NOKPROBE_SYMBOL(trampoline_probe_handler);

index d2a71d872638..fc30e799bd84 100644
@@ -228,6 +228,7 @@ NOKPROBE_SYMBOL(pop_kprobe);
 void arch_prepare_kretprobe(struct kretprobe_instance *ri, struct pt_regs *regs)
 {
 	ri->ret_addr = (kprobe_opcode_t *) regs->gprs[14];
+	ri->fp = NULL;
 
 	/* Replace the return addr with trampoline addr */
 	regs->gprs[14] = (unsigned long) &kretprobe_trampoline;

@@ -331,83 +332,7 @@ static void __used kretprobe_trampoline_holder(void)
  */
 static int trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address;
-	unsigned long trampoline_address;
-	kprobe_opcode_t *correct_ret_addr;
-
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because an multiple functions in the call path
-	 * have a return probe installed on them, and/or more than one return
-	 * return probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always inserted at the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *	 function, the first instance's ret_addr will point to the
-	 *	 real return address, and all the rest will point to
-	 *	 kretprobe_trampoline
-	 */
-	ri = NULL;
-	orig_ret_address = 0;
-	correct_ret_addr = NULL;
-	trampoline_address = (unsigned long) &kretprobe_trampoline;
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long) ri->ret_addr;
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-
-	correct_ret_addr = ri->ret_addr;
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		orig_ret_address = (unsigned long) ri->ret_addr;
-
-		if (ri->rp && ri->rp->handler) {
-			ri->ret_addr = correct_ret_addr;
-			ri->rp->handler(ri, regs);
-		}
-
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	regs->psw.addr = orig_ret_address;
-
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
+	regs->psw.addr = __kretprobe_trampoline_handler(regs, &kretprobe_trampoline, NULL);
 	/*
 	 * By returning a non-zero value, we are telling
 	 * kprobe_handler() that we don't want the post_handler

index 318296f48f1a..756100b01e84 100644
@@ -204,6 +204,7 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 				      struct pt_regs *regs)
 {
 	ri->ret_addr = (kprobe_opcode_t *) regs->pr;
+	ri->fp = NULL;
 
 	/* Replace the return addr with trampoline addr */
 	regs->pr = (unsigned long)kretprobe_trampoline;

@@ -302,62 +303,9 @@ static void __used kretprobe_trampoline_holder(void)
  */
 int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline;
+	regs->pc = __kretprobe_trampoline_handler(regs, &kretprobe_trampoline, NULL);
 
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because an multiple functions in the call path
-	 * have a return probe installed on them, and/or more then one return
-	 * return probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always inserted at the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *       function, the first instance's ret_addr will point to the
-	 *       real return address, and all the rest will point to
-	 *       kretprobe_trampoline
-	 */
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		if (ri->rp && ri->rp->handler) {
-			__this_cpu_write(current_kprobe, &ri->rp->kp);
-			ri->rp->handler(ri, regs);
-			__this_cpu_write(current_kprobe, NULL);
-		}
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
-
-	regs->pc = orig_ret_address;
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
-
-	return orig_ret_address;
+	return 1;
 }
 
 static int __kprobes post_kprobe_handler(struct pt_regs *regs)

index dfbca2470536..217c21a6986a 100644
@@ -453,6 +453,7 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 				      struct pt_regs *regs)
 {
 	ri->ret_addr = (kprobe_opcode_t *)(regs->u_regs[UREG_RETPC] + 8);
+	ri->fp = NULL;
 
 	/* Replace the return addr with trampoline addr */
 	regs->u_regs[UREG_RETPC] =

@@ -465,58 +466,12 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 static int __kprobes trampoline_probe_handler(struct kprobe *p,
 					      struct pt_regs *regs)
 {
-	struct kretprobe_instance *ri = NULL;
-	struct hlist_head *head, empty_rp;
-	struct hlist_node *tmp;
-	unsigned long flags, orig_ret_address = 0;
-	unsigned long trampoline_address =(unsigned long)&kretprobe_trampoline;
+	unsigned long orig_ret_address = 0;
 
-	INIT_HLIST_HEAD(&empty_rp);
-	kretprobe_hash_lock(current, &head, &flags);
-
-	/*
-	 * It is possible to have multiple instances associated with a given
-	 * task either because an multiple functions in the call path
-	 * have a return probe installed on them, and/or more than one return
-	 * return probe was registered for a target function.
-	 *
-	 * We can handle this because:
-	 *     - instances are always inserted at the head of the list
-	 *     - when multiple return probes are registered for the same
-	 *       function, the first instance's ret_addr will point to the
-	 *       real return address, and all the rest will point to
-	 *       kretprobe_trampoline
-	 */
-	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
-		if (ri->task != current)
-			/* another task is sharing our hash bucket */
-			continue;
-
-		if (ri->rp && ri->rp->handler)
-			ri->rp->handler(ri, regs);
-
-		orig_ret_address = (unsigned long)ri->ret_addr;
-		recycle_rp_inst(ri, &empty_rp);
-
-		if (orig_ret_address != trampoline_address)
-			/*
-			 * This is the real return address. Any other
-			 * instances associated with this task are for
-			 * other calls deeper on the call stack
-			 */
-			break;
-	}
-
-	kretprobe_assert(ri, orig_ret_address, trampoline_address);
+	orig_ret_address = __kretprobe_trampoline_handler(regs, &kretprobe_trampoline, NULL);
 	regs->tpc = orig_ret_address;
 	regs->tnpc = orig_ret_address + 4;
 
-	kretprobe_hash_unlock(current, &flags);
-
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
 	/*
 	 * By returning a non-zero value, we are telling
 	 * kprobe_handler() that we don't want the post_handler

index 46510e5000ff..c8de76d230e3 100644
@@ -1359,7 +1359,8 @@ static void cleanup_rp_inst(struct kretprobe *rp)
 	struct hlist_node *next;
 	struct hlist_head *head;
 
-	/* No race here */
+	/* To avoid recursive kretprobe by NMI, set kprobe busy here */
+	kprobe_busy_begin();
 	for (hash = 0; hash < KPROBE_TABLE_SIZE; hash++) {
 		kretprobe_table_lock(hash, &flags);
 		head = &kretprobe_inst_table[hash];

@@ -1369,6 +1370,8 @@ static void cleanup_rp_inst(struct kretprobe *rp)
 		}
 		kretprobe_table_unlock(hash, &flags);
 	}
+	kprobe_busy_end();
+
 	free_rp_inst(rp);
 }
 NOKPROBE_SYMBOL(cleanup_rp_inst);

@@ -2033,17 +2036,6 @@ static int pre_handler_kretprobe(struct kprobe *p, struct pt_regs *regs)
 	unsigned long hash, flags = 0;
 	struct kretprobe_instance *ri;
 
-	/*
-	 * To avoid deadlocks, prohibit return probing in NMI contexts,
-	 * just skip the probe and increase the (inexact) 'nmissed'
-	 * statistical counter, so that the user is informed that
-	 * something happened:
-	 */
-	if (unlikely(in_nmi())) {
-		rp->nmissed++;
-		return 0;
-	}
-
 	/* TODO: consider to only swap the RA after the last pre_handler fired */
 	hash = hash_ptr(current, KPROBE_HASH_BITS);
 	raw_spin_lock_irqsave(&rp->lock, flags);

index c6a913e608b7..663be8debf25 100644
@@ -156,7 +156,10 @@ struct kretprobe {
 };
 
 struct kretprobe_instance {
-	struct hlist_node hlist;
+	union {
+		struct hlist_node hlist;
+		struct rcu_head rcu;
+	};
 	struct kretprobe *rp;
 	kprobe_opcode_t *ret_addr;
 	struct task_struct *task;

@@ -395,7 +398,6 @@ int register_kretprobes(struct kretprobe **rps, int num);
 void unregister_kretprobes(struct kretprobe **rps, int num);
 
 void kprobe_flush_task(struct task_struct *tk);
-void recycle_rp_inst(struct kretprobe_instance *ri, struct hlist_head *head);
 
 int disable_kprobe(struct kprobe *kp);
 int enable_kprobe(struct kprobe *kp);

index c8de76d230e3..807d4429e8a2 100644
@@ -1223,8 +1223,7 @@ void kprobes_inc_nmissed_count(struct kprobe *p)
 }
 NOKPROBE_SYMBOL(kprobes_inc_nmissed_count);
 
-void recycle_rp_inst(struct kretprobe_instance *ri,
-		     struct hlist_head *head)
+static void recycle_rp_inst(struct kretprobe_instance *ri)
 {
 	struct kretprobe *rp = ri->rp;
 

@@ -1236,8 +1235,7 @@ void recycle_rp_inst(struct kretprobe_instance *ri,
 		hlist_add_head(&ri->hlist, &rp->free_instances);
 		raw_spin_unlock(&rp->lock);
 	} else
-		/* Unregistering */
-		hlist_add_head(&ri->hlist, head);
+		kfree_rcu(ri, rcu);
 }
 NOKPROBE_SYMBOL(recycle_rp_inst);
 

@@ -1313,7 +1311,7 @@ void kprobe_busy_end(void)
 void kprobe_flush_task(struct task_struct *tk)
 {
 	struct kretprobe_instance *ri;
-	struct hlist_head *head, empty_rp;
+	struct hlist_head *head;
 	struct hlist_node *tmp;
 	unsigned long hash, flags = 0;
 

@@ -1323,19 +1321,14 @@ void kprobe_flush_task(struct task_struct *tk)
 
 	kprobe_busy_begin();
 
-	INIT_HLIST_HEAD(&empty_rp);
 	hash = hash_ptr(tk, KPROBE_HASH_BITS);
 	head = &kretprobe_inst_table[hash];
 	kretprobe_table_lock(hash, &flags);
 	hlist_for_each_entry_safe(ri, tmp, head, hlist) {
 		if (ri->task == tk)
-			recycle_rp_inst(ri, &empty_rp);
+			recycle_rp_inst(ri);
 	}
 	kretprobe_table_unlock(hash, &flags);
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
 
 	kprobe_busy_end();
 }

@@ -1936,13 +1929,12 @@ unsigned long __kretprobe_trampoline_handler(struct pt_regs *regs,
 					     void *frame_pointer)
 {
 	struct kretprobe_instance *ri = NULL, *last = NULL;
-	struct hlist_head *head, empty_rp;
+	struct hlist_head *head;
 	struct hlist_node *tmp;
 	unsigned long flags;
 	kprobe_opcode_t *correct_ret_addr = NULL;
 	bool skipped = false;
 
-	INIT_HLIST_HEAD(&empty_rp);
 	kretprobe_hash_lock(current, &head, &flags);
 
 	/*

@@ -2009,7 +2001,7 @@ unsigned long __kretprobe_trampoline_handler(struct pt_regs *regs,
 			__this_cpu_write(current_kprobe, &kprobe_busy);
 		}
 
-		recycle_rp_inst(ri, &empty_rp);
+		recycle_rp_inst(ri);
 
 		if (ri == last)
 			break;

@@ -2017,11 +2009,6 @@ unsigned long __kretprobe_trampoline_handler(struct pt_regs *regs,
 
 	kretprobe_hash_unlock(current, &flags);
 
-	hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
-		hlist_del(&ri->hlist);
-		kfree(ri);
-	}
-
 	return (unsigned long)correct_ret_addr;
 }
 NOKPROBE_SYMBOL(__kretprobe_trampoline_handler)

index 663be8debf25..9c880c8a4e80 100644
@@ -190,7 +190,6 @@ static inline int kprobes_built_in(void)
 	return 1;
 }
 
-extern struct kprobe kprobe_busy;
 void kprobe_busy_begin(void);
 void kprobe_busy_end(void);
 

@@ -235,16 +234,6 @@ static inline int arch_trampoline_kprobe(struct kprobe *p)
 
 extern struct kretprobe_blackpoint kretprobe_blacklist[];
 
-static inline void kretprobe_assert(struct kretprobe_instance *ri,
-	unsigned long orig_ret_address, unsigned long trampoline_address)
-{
-	if (!orig_ret_address || (orig_ret_address == trampoline_address)) {
-		printk("kretprobe BUG!: Processing kretprobe %p @ %p\n",
-				ri->rp, ri->rp->kp.addr);
-		BUG();
-	}
-}
-
 #ifdef CONFIG_KPROBES_SANITY_TEST
 extern int init_test_probes(void);
 #else

@@ -364,10 +353,6 @@ int arch_check_ftrace_location(struct kprobe *p);
 
 /* Get the kprobe at this addr (if any) - called with preemption disabled */
 struct kprobe *get_kprobe(void *addr);
-void kretprobe_hash_lock(struct task_struct *tsk,
-			 struct hlist_head **head, unsigned long *flags);
-void kretprobe_hash_unlock(struct task_struct *tsk, unsigned long *flags);
-struct hlist_head * kretprobe_inst_table_head(struct task_struct *tsk);
 
 /* kprobe_running() will just return the current_kprobe on this CPU */
 static inline struct kprobe *kprobe_running(void)

index 807d4429e8a2..d0b4b7e89fa6 100644
@@ -1239,7 +1239,7 @@ static void recycle_rp_inst(struct kretprobe_instance *ri)
 }
 NOKPROBE_SYMBOL(recycle_rp_inst);
 
-void kretprobe_hash_lock(struct task_struct *tsk,
+static void kretprobe_hash_lock(struct task_struct *tsk,
 			 struct hlist_head **head, unsigned long *flags)
 __acquires(hlist_lock)
 {

@@ -1261,7 +1261,7 @@ __acquires(hlist_lock)
 }
 NOKPROBE_SYMBOL(kretprobe_table_lock);
 
-void kretprobe_hash_unlock(struct task_struct *tsk,
+static void kretprobe_hash_unlock(struct task_struct *tsk,
 			   unsigned long *flags)
 __releases(hlist_lock)
 {

@@ -1282,7 +1282,7 @@ __releases(hlist_lock)
 }
 NOKPROBE_SYMBOL(kretprobe_table_unlock);
 
-struct kprobe kprobe_busy = {
+static struct kprobe kprobe_busy = {
 	.addr = (void *) get_kprobe,
 };
 

@@ -1983,8 +1983,7 @@ unsigned long __kretprobe_trampoline_handler(struct pt_regs *regs,
 			break;
 	}
 
-	kretprobe_assert(ri, (unsigned long)correct_ret_addr,
-			     (unsigned long)trampoline_address);
+	BUG_ON(!correct_ret_addr || (correct_ret_addr == trampoline_address));
 	last = ri;
 
 	hlist_for_each_entry_safe(ri, tmp, head, hlist) {

index 0b2fe55e6194..0e851b925c8c 100644
@@ -357,12 +357,6 @@ void i915_request_retire_upto(struct i915_request *rq)
 	} while (i915_request_retire(tmp) && tmp != rq);
 }
 
-static void __llist_add(struct llist_node *node, struct llist_head *head)
-{
-	node->next = head->first;
-	head->first = node;
-}
-
 static struct i915_request * const *
 __engine_active(struct intel_engine_cs *engine)
 {

index 2e9c7215882b..c17893dcc591 100644
@@ -197,6 +197,16 @@ static inline struct llist_node *llist_next(struct llist_node *node)
 extern bool llist_add_batch(struct llist_node *new_first,
 			    struct llist_node *new_last,
 			    struct llist_head *head);
+
+static inline bool __llist_add_batch(struct llist_node *new_first,
+				     struct llist_node *new_last,
+				     struct llist_head *head)
+{
+	new_last->next = head->first;
+	head->first = new_first;
+	return new_last->next == NULL;
+}
+
 /**
  * llist_add - add a new entry
  * @new:	new entry to be added

@@ -209,6 +219,11 @@ static inline bool llist_add(struct llist_node *new, struct llist_head *head)
 	return llist_add_batch(new, new, head);
 }
 
+static inline bool __llist_add(struct llist_node *new, struct llist_head *head)
+{
+	return __llist_add_batch(new, new, head);
+}
+
 /**
  * llist_del_all - delete all entries from lock-less list
  * @head:	the head of lock-less list to delete all entries

index 2d95dc3f4644..13b0db2d0be2 100644
@@ -3006,15 +3006,14 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
  */
 bool try_invoke_on_locked_down_task(struct task_struct *p, bool (*func)(struct task_struct *t, void *arg), void *arg)
 {
-	bool ret = false;
 	struct rq_flags rf;
+	bool ret = false;
 	struct rq *rq;
 
-	lockdep_assert_irqs_enabled();
-	raw_spin_lock_irq(&p->pi_lock);
+	raw_spin_lock_irqsave(&p->pi_lock, rf.flags);
 	if (p->on_rq) {
 		rq = __task_rq_lock(p, &rf);
-		if (task_rq(p) == rq)
+		if (task_rq(p) == rq && rq->curr != p)
 			ret = func(p, arg);
 		rq_unlock(rq, &rf);
 	} else {

@@ -3028,7 +3027,7 @@ bool try_invoke_on_locked_down_task(struct task_struct *p, bool (*func)(struct t
 				ret = func(p, arg);
 		}
 	}
-	raw_spin_unlock_irq(&p->pi_lock);
+	raw_spin_unlock_irqrestore(&p->pi_lock, rf.flags);
 	return ret;
 }
 

index 8471a0f..a814028 100644
@@ -2988,7 +2988,7 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
 
 /**
  * try_invoke_on_locked_down_task - Invoke a function on task in fixed state
- * @p: Process for which the function is to be invoked.
+ * @p: Process for which the function is to be invoked, can be @current.
  * @func: Function to invoke.
  * @arg: Argument to function.
  *

@@ -3006,12 +3006,11 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
  */
 bool try_invoke_on_locked_down_task(struct task_struct *p, bool (*func)(struct task_struct *t, void *arg), void *arg)
 {
-	bool ret = false;
 	struct rq_flags rf;
+	bool ret = false;
 	struct rq *rq;
 
-	lockdep_assert_irqs_enabled();
-	raw_spin_lock_irq(&p->pi_lock);
+	raw_spin_lock_irqsave(&p->pi_lock, rf.flags);
 	if (p->on_rq) {
 		rq = __task_rq_lock(p, &rf);
 		if (task_rq(p) == rq)

@@ -3028,7 +3027,7 @@ bool try_invoke_on_locked_down_task(struct task_struct *p, bool (*func)(struct t
 				ret = func(p, arg);
 		}
 	}
-	raw_spin_unlock_irq(&p->pi_lock);
+	raw_spin_unlock_irqrestore(&p->pi_lock, rf.flags);
 	return ret;
 }
 

index 9c880c8a4e80..a30cccb07f21 100644
@@ -157,6 +157,7 @@ struct kretprobe {
 
 struct kretprobe_instance {
 	union {
+		struct llist_node llist;
 		struct hlist_node hlist;
 		struct rcu_head rcu;
 	};

index 93ecd930efd3..0f2532f052a9 100644
@@ -1315,6 +1315,10 @@ struct task_struct {
 	struct callback_head		mce_kill_me;
 #endif
 
+#ifdef CONFIG_KRETPROBES
+	struct llist_head               kretprobe_instances;
+#endif
+
 	/*
 	 * New fields for task_struct should be added above here, so that
 	 * they are included in the randomized portion of task_struct.

index 4d32190861bd..2ff5cceb0732 100644
@@ -2161,6 +2161,10 @@ static __latent_entropy struct task_struct *copy_process(
 	INIT_LIST_HEAD(&p->thread_group);
 	p->task_works = NULL;
 
+#ifdef CONFIG_KRETPROBES
+	p->kretprobe_instances.first = NULL;
+#endif
+
 	/*
 	 * Ensure that the cgroup subsystem policies allow the new process to be
 	 * forked. It should be noted the the new process's css_set can be changed

338 lines not on this page · raw message

Index: linux-2.6/kernel/kprobes.c
===================================================================
@@ -1275,9 +1275,7 @@ void kprobe_flush_task(struct task_struc
 
 	kprobe_busy_begin();
 
-	node = current->kretprobe_instances.first;
-	current->kretprobe_instances.first = NULL;
-
+	node == __llist_del_all(&tk->kretprobe_instances);
 	while (node) {
 		ri = container_of(node, struct kretprobe_instance, llist);
 		node = node->next;

@@ -1871,6 +1869,7 @@ unsigned long __kretprobe_trampoline_han
 	struct llist_node *first, *node;
 	struct kretprobe *rp;
 
+	/* Find all nodes for this frame. */
 	first = node = current->kretprobe_instances.first;
 	while (node) {
 		ri = container_of(node, struct kretprobe_instance, llist);

@@ -1900,7 +1899,7 @@ unsigned long __kretprobe_trampoline_han
 	while (first) {
 		ri = container_of(first, struct kretprobe_instance, llist);
 		rp = get_kretprobe(ri);
-		node = first->next;
+		first = first->next;
 
 		if (rp && rp->handler) {
 			__this_cpu_write(current_kprobe, &rp->kp);

@@ -1910,8 +1909,6 @@ unsigned long __kretprobe_trampoline_han
 		}
 
 		recycle_rp_inst(ri);
-
-		first = node;
 	}
 
 	return (unsigned long)correct_ret_addr;

Index: linux-2.6/include/linux/llist.h
===================================================================
@@ -237,6 +237,14 @@ static inline struct llist_node *llist_d
 	return xchg(&head->first, NULL);
 }
 
+static inline struct llist_node *__llist_del_all(struct llist_head *head)
+{
+	struct llist_node *first = head->first;
+
+	head->first = NULL;
+	return first;
+}
+
 extern struct llist_node *llist_del_first(struct llist_head *head);
 
 struct llist_node *llist_reverse_order(struct llist_node *head);

index a30cccb07f21..d7cdae2d8f2e 100644
@@ -27,6 +27,7 @@
 #include <linux/rcupdate.h>
 #include <linux/mutex.h>
 #include <linux/ftrace.h>
+#include <linux/refcount.h>
 #include <asm/kprobes.h>
 
 #ifdef CONFIG_KPROBES

@@ -144,6 +145,11 @@ static inline int kprobe_ftrace(struct kprobe *p)
  * ignored, due to maxactive being too low.
  *
  */
+struct kretprobe_holder {
+	struct kretprobe	*rp;
+	refcount_t		ref;
+};
+
 struct kretprobe {
 	struct kprobe kp;
 	kretprobe_handler_t handler;

@@ -152,6 +158,7 @@ struct kretprobe {
 	int nmissed;
 	size_t data_size;
 	struct hlist_head free_instances;
+	struct kretprobe_holder *rph;
 	raw_spinlock_t lock;
 };
 

@@ -161,9 +168,8 @@ struct kretprobe_instance {
 		struct hlist_node hlist;
 		struct rcu_head rcu;
 	};
-	struct kretprobe *rp;
+	struct kretprobe_holder *rph;
 	kprobe_opcode_t *ret_addr;
-	struct task_struct *task;
 	void *fp;
 	char data[];
 };

@@ -222,6 +228,13 @@ unsigned long kretprobe_trampoline_handler(struct pt_regs *regs,
 	return ret;
 }
 
+static nokprobe_inline struct kretprobe *get_kretprobe(struct kretprobe_instance *ri)
+{
+	/* rph->rp can be updated by unregister_kretprobe() on other cpu */
+	smp_rmb();
+	return ri->rph->rp;
+}
+
 #else /* CONFIG_KRETPROBES */
 static inline void arch_prepare_kretprobe(struct kretprobe *rp,
 					struct pt_regs *regs)

192 lines not on this page · raw message

index aefb6065b508..07baf6f6cecc 100644
@@ -1714,7 +1714,8 @@ NOKPROBE_SYMBOL(kprobe_dispatcher);
 static int
 kretprobe_dispatcher(struct kretprobe_instance *ri, struct pt_regs *regs)
 {
-	struct trace_kprobe *tk = container_of(ri->rp, struct trace_kprobe, rp);
+	struct kretprobe *rp = get_kretprobe(ri);
+	struct trace_kprobe *tk = container_of(rp, struct trace_kprobe, rp);
 
 	raw_cpu_inc(*tk->nhit);
 

index b6cac6e9bb70..f732741ad7c7 100644
@@ -199,7 +199,7 @@ static __always_inline int arch_atomic_cmpxchg(atomic_t *v, int old, int new)
 
 static __always_inline bool arch_atomic_try_cmpxchg(atomic_t *v, int *old, int new)
 {
-	return try_cmpxchg(&v->counter, old, new);
+	return arch_try_cmpxchg(&v->counter, old, new);
 }
 #define arch_atomic_try_cmpxchg arch_atomic_try_cmpxchg
 

index 809bd010a751..7886d0578fc9 100644
@@ -187,7 +187,7 @@ static inline s64 arch_atomic64_cmpxchg(atomic64_t *v, s64 old, s64 new)
 
 static __always_inline bool arch_atomic64_try_cmpxchg(atomic64_t *v, s64 *old, s64 new)
 {
-	return try_cmpxchg(&v->counter, old, new);
+	return arch_try_cmpxchg(&v->counter, old, new);
 }
 #define arch_atomic64_try_cmpxchg arch_atomic64_try_cmpxchg
 

index a8bfac131256..4d4ec5cbdc51 100644
@@ -221,7 +221,7 @@ extern void __add_wrong_size(void)
 #define __try_cmpxchg(ptr, pold, new, size)				\
 	__raw_try_cmpxchg((ptr), (pold), (new), (size), LOCK_PREFIX)
 
-#define try_cmpxchg(ptr, pold, new) 					\
+#define arch_try_cmpxchg(ptr, pold, new) 				\
 	__try_cmpxchg((ptr), (pold), (new), sizeof(*(ptr)))
 
 /*

278 lines not on this page · raw message

117 lines not on this page · raw message

117 lines not on this page · raw message

new file mode 100644
index 000000000000..fc1842b96469
@@ -0,0 +1,129 @@
+/* SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause */
+#ifndef FREELIST_H
+#define FREELIST_H
+
+#include <linux/atomic.h>
+
+/*
+ * Copyright: [email protected]
+ *
+ * A simple CAS-based lock-free free list. Not the fastest thing in the world
+ * under heavy contention, but simple and correct (assuming nodes are never
+ * freed until after the free list is destroyed), and fairly speedy under low
+ * contention.
+ *
+ * Adapted from: https://moodycamel.com/blog/2014/solving-the-aba-problem-for-lock-free-free-lists
+ */
+
+struct freelist_node {
+	atomic_t		refs;
+	struct freelist_node	*next;
+};
+
+struct freelist_head {
+	struct freelist_node	*head;
+};
+
+#define REFS_ON_FREELIST 0x80000000
+#define REFS_MASK	 0x7FFFFFFF
+
+static inline void __freelist_add(struct freelist_node *node, struct freelist_head *list)
+{
+	/*
+	 * Since the refcount is zero, and nobody can increase it once it's
+	 * zero (except us, and we run only one copy of this method per node at
+	 * a time, i.e. the single thread case), then we know we can safely
+	 * change the next pointer of the node; however, once the refcount is
+	 * back above zero, then other threads could increase it (happens under
+	 * heavy contention, when the refcount goes to zero in between a load
+	 * and a refcount increment of a node in try_get, then back up to
+	 * something non-zero, then the refcount increment is done by the other
+	 * thread) -- so if the CAS to add the node to the actual list fails,
+	 * decrese the refcount and leave the add operation to the next thread
+	 * who puts the refcount back to zero (which could be us, hence the
+	 * loop).
+	 */
+	struct freelist_node *head = READ_ONCE(list->head);
+
+	for (;;) {
+		WRITE_ONCE(node->next, head);
+		atomic_set_release(&node->refs, 1);
+
+		if (!try_cmpxchg_release(&list->head, &head, node)) {
+			/*
+			 * Hmm, the add failed, but we can only try again when
+			 * the refcount goes back to zero.
+			 */
+			if (atomic_fetch_add_release(REFS_ON_FREELIST - 1, &node->refs) == 1)
+				continue;
+		}
+		return;
+	}
+}
+
+static inline void freelist_add(struct freelist_node *node, struct freelist_head *list)
+{
+	/*
+	 * We know that the should-be-on-freelist bit is 0 at this point, so
+	 * it's safe to set it using a fetch_add.
+	 */
+	if (!atomic_fetch_add_release(REFS_ON_FREELIST, &node->refs)) {
+		/*
+		 * Oh look! We were the last ones referencing this node, and we
+		 * know we want to add it to the free list, so let's do it!
+		 */
+		__freelist_add(node, list);
+	}
+}
+
+static inline struct freelist_node *freelist_try_get(struct freelist_head *list)
+{
+	struct freelist_node *prev, *next, *head = smp_load_acquire(&list->head);
+	unsigned int refs;
+
+	while (head) {
+		prev = head;
+		refs = atomic_read(&head->refs);
+		if ((refs & REFS_MASK) == 0 ||
+		    !atomic_try_cmpxchg_acquire(&head->refs, &refs, refs+1)) {
+			head = smp_load_acquire(&list->head);
+			continue;
+		}
+
+		/*
+		 * Good, reference count has been incremented (it wasn't at
+		 * zero), which means we can read the next and not worry about
+		 * it changing between now and the time we do the CAS.
+		 */
+		next = READ_ONCE(head->next);
+		if (try_cmpxchg_acquire(&list->head, &head, next)) {
+			/*
+			 * Yay, got the node. This means it was on the list,
+			 * which means should-be-on-freelist must be false no
+			 * matter the refcount (because nobody else knows it's
+			 * been taken off yet, it can't have been put back on).
+			 */
+			WARN_ON_ONCE(atomic_read(&head->refs) & REFS_ON_FREELIST);
+
+			/*
+			 * Decrease refcount twice, once for our ref, and once
+			 * for the list's ref.
+			 */
+			atomic_fetch_add(-2, &head->refs);
+
+			return head;
+		}
+
+		/*
+		 * OK, the head must have changed on us, but we still need to decrement
+		 * the refcount we increased.
+		 */
+		refs = atomic_fetch_add(-1, &prev->refs);
+		if (refs == REFS_ON_FREELIST + 1)
+			__freelist_add(prev, list);
+	}
+
+	return NULL;
+}
+
+#endif /* FREELIST_H */

index d7cdae2d8f2e..aa330bfefefd 100644
@@ -28,6 +28,7 @@
 #include <linux/mutex.h>
 #include <linux/ftrace.h>
 #include <linux/refcount.h>
+#include <linux/freelist.h>
 #include <asm/kprobes.h>
 
 #ifdef CONFIG_KPROBES

@@ -157,17 +158,16 @@ struct kretprobe {
 	int maxactive;
 	int nmissed;
 	size_t data_size;
-	struct hlist_head free_instances;
+	struct freelist_head freelist;
 	struct kretprobe_holder *rph;
-	raw_spinlock_t lock;
 };
 
 struct kretprobe_instance {
 	union {
-		struct llist_node llist;
-		struct hlist_node hlist;
+		struct freelist_node freelist;
 		struct rcu_head rcu;
 	};
+	struct llist_node llist;
 	struct kretprobe_holder *rph;
 	kprobe_opcode_t *ret_addr;
 	void *fp;

94 lines not on this page · raw message