ftrace: make work with new ring buffer

This patch ports ftrace over to the new ring buffer.

Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

kernel/trace/trace.h

@@ -5,6 +5,7 @@
 #include <asm/atomic.h>
 #include <linux/sched.h>
 #include <linux/clocksource.h>
+#include <linux/ring_buffer.h>
 #include <linux/mmiotrace.h>
 #include <linux/ftrace.h>
 
@@ -102,7 +103,6 @@ struct trace_field {
 	char			flags;
 	char			preempt_count;
 	int			pid;
-	cycle_t			t;
 	union {
 		struct ftrace_entry		fn;
 		struct ctx_switch_entry		ctx;
@@ -139,16 +139,9 @@ struct trace_entry {
  * the trace, etc.)
  */
 struct trace_array_cpu {
-	struct list_head	trace_pages;
 	atomic_t		disabled;
-	raw_spinlock_t		lock;
-	struct lock_class_key	lock_key;
 
 	/* these fields get copied into max-trace: */
-	unsigned		trace_head_idx;
-	unsigned		trace_tail_idx;
-	void			*trace_head; /* producer */
-	void			*trace_tail; /* consumer */
 	unsigned long		trace_idx;
 	unsigned long		overrun;
 	unsigned long		saved_latency;
@@ -172,6 +165,7 @@ struct trace_iterator;
  * They have on/off state as well:
  */
 struct trace_array {
+	struct ring_buffer	*buffer;
 	unsigned long		entries;
 	long			ctrl;
 	int			cpu;
@@ -219,27 +213,21 @@ struct trace_iterator {
 	struct trace_array	*tr;
 	struct tracer		*trace;
 	void			*private;
-	long			last_overrun[NR_CPUS];
+	struct ring_buffer_iter	*buffer_iter[NR_CPUS];
-	long			overrun[NR_CPUS];
 
 	/* The below is zeroed out in pipe_read */
 	struct trace_seq	seq;
 	struct trace_entry	*ent;
 	int			cpu;
-
+	u64			ts;
-	struct trace_entry	*prev_ent;
-	int			prev_cpu;
 
 	unsigned long		iter_flags;
 	loff_t			pos;
-	unsigned long		next_idx[NR_CPUS];
-	struct list_head	*next_page[NR_CPUS];
-	unsigned		next_page_idx[NR_CPUS];
 	long			idx;
 };
 
 void trace_wake_up(void);
-void tracing_reset(struct trace_array_cpu *data);
+void tracing_reset(struct trace_array *tr, int cpu);
 int tracing_open_generic(struct inode *inode, struct file *filp);
 struct dentry *tracing_init_dentry(void);
 void init_tracer_sysprof_debugfs(struct dentry *d_tracer);

kernel/trace/trace_boot.c

@@ -34,7 +34,7 @@ static void boot_trace_init(struct trace_array *tr)
 	trace_boot_enabled = 0;
 
 	for_each_cpu_mask(cpu, cpu_possible_map)
-		tracing_reset(tr->data[cpu]);
+		tracing_reset(tr, cpu);
 }
 
 static void boot_trace_ctrl_update(struct trace_array *tr)
@@ -74,6 +74,7 @@ struct tracer boot_tracer __read_mostly =
 
 void trace_boot(struct boot_trace *it)
 {
+	struct ring_buffer_event *event;
 	struct trace_entry *entry;
 	struct trace_array_cpu *data;
 	unsigned long irq_flags;
@@ -85,17 +86,18 @@ void trace_boot(struct boot_trace *it)
 	preempt_disable();
 	data = tr->data[smp_processor_id()];
 
-	raw_local_irq_save(irq_flags);
+	event = ring_buffer_lock_reserve(tr->buffer, sizeof(*entry),
-	__raw_spin_lock(&data->lock);
+					 &irq_flags);
-
+	if (!event)
-	entry = tracing_get_trace_entry(tr, data);
+		goto out;
+	entry	= ring_buffer_event_data(event);
 	tracing_generic_entry_update(entry, 0);
 	entry->type = TRACE_BOOT;
 	entry->field.initcall = *it;
+	ring_buffer_unlock_commit(tr->buffer, event, irq_flags);
 
-	__raw_spin_unlock(&data->lock);
-	raw_local_irq_restore(irq_flags);
 	trace_wake_up();
 
+ out:
 	preempt_enable();
 }

kernel/trace/trace_functions.c

@@ -23,7 +23,7 @@ static void function_reset(struct trace_array *tr)
 	tr->time_start = ftrace_now(tr->cpu);
 
 	for_each_online_cpu(cpu)
-		tracing_reset(tr->data[cpu]);
+		tracing_reset(tr, cpu);
 }
 
 static void start_function_trace(struct trace_array *tr)

kernel/trace/trace_irqsoff.c

@@ -173,7 +173,7 @@ check_critical_timing(struct trace_array *tr,
 out:
 	data->critical_sequence = max_sequence;
 	data->preempt_timestamp = ftrace_now(cpu);
-	tracing_reset(data);
+	tracing_reset(tr, cpu);
 	trace_function(tr, data, CALLER_ADDR0, parent_ip, flags);
 }
 
@@ -203,7 +203,7 @@ start_critical_timing(unsigned long ip, unsigned long parent_ip)
 	data->critical_sequence = max_sequence;
 	data->preempt_timestamp = ftrace_now(cpu);
 	data->critical_start = parent_ip ? : ip;
-	tracing_reset(data);
+	tracing_reset(tr, cpu);
 
 	local_save_flags(flags);
 
@@ -234,7 +234,7 @@ stop_critical_timing(unsigned long ip, unsigned long parent_ip)
 
 	data = tr->data[cpu];
 
-	if (unlikely(!data) || unlikely(!head_page(data)) ||
+	if (unlikely(!data) ||
 	    !data->critical_start || atomic_read(&data->disabled))
 		return;
 

kernel/trace/trace_mmiotrace.c

@@ -27,7 +27,7 @@ static void mmio_reset_data(struct trace_array *tr)
 	tr->time_start = ftrace_now(tr->cpu);
 
 	for_each_online_cpu(cpu)
-		tracing_reset(tr->data[cpu]);
+		tracing_reset(tr, cpu);
 }
 
 static void mmio_trace_init(struct trace_array *tr)
@@ -130,10 +130,14 @@ static unsigned long count_overruns(struct trace_iterator *iter)
 {
 	int cpu;
 	unsigned long cnt = 0;
+/* FIXME: */
+#if 0
 	for_each_online_cpu(cpu) {
 		cnt += iter->overrun[cpu];
 		iter->overrun[cpu] = 0;
 	}
+#endif
+	(void)cpu;
 	return cnt;
 }
 
@@ -176,7 +180,7 @@ static int mmio_print_rw(struct trace_iterator *iter)
 	struct trace_entry *entry = iter->ent;
 	struct mmiotrace_rw *rw	= &entry->field.mmiorw;
 	struct trace_seq *s	= &iter->seq;
-	unsigned long long t	= ns2usecs(entry->field.t);
+	unsigned long long t	= ns2usecs(iter->ts);
 	unsigned long usec_rem	= do_div(t, 1000000ULL);
 	unsigned secs		= (unsigned long)t;
 	int ret = 1;
@@ -218,7 +222,7 @@ static int mmio_print_map(struct trace_iterator *iter)
 	struct trace_entry *entry = iter->ent;
 	struct mmiotrace_map *m	= &entry->field.mmiomap;
 	struct trace_seq *s	= &iter->seq;
-	unsigned long long t	= ns2usecs(entry->field.t);
+	unsigned long long t	= ns2usecs(iter->ts);
 	unsigned long usec_rem	= do_div(t, 1000000ULL);
 	unsigned secs		= (unsigned long)t;
 	int ret = 1;
@@ -250,7 +254,7 @@ static int mmio_print_mark(struct trace_iterator *iter)
 	struct trace_entry *entry = iter->ent;
 	const char *msg		= entry->field.print.buf;
 	struct trace_seq *s	= &iter->seq;
-	unsigned long long t	= ns2usecs(entry->field.t);
+	unsigned long long t	= ns2usecs(iter->ts);
 	unsigned long usec_rem	= do_div(t, 1000000ULL);
 	unsigned secs		= (unsigned long)t;
 	int ret;
@@ -303,19 +307,19 @@ static void __trace_mmiotrace_rw(struct trace_array *tr,
 				struct trace_array_cpu *data,
 				struct mmiotrace_rw *rw)
 {
+	struct ring_buffer_event *event;
 	struct trace_entry *entry;
 	unsigned long irq_flags;
 
-	raw_local_irq_save(irq_flags);
+	event	= ring_buffer_lock_reserve(tr->buffer, sizeof(*entry),
-	__raw_spin_lock(&data->lock);
+					   &irq_flags);
-
+	if (!event)
-	entry				= tracing_get_trace_entry(tr, data);
+		return;
+	entry	= ring_buffer_event_data(event);
 	tracing_generic_entry_update(entry, 0);
 	entry->type			= TRACE_MMIO_RW;
 	entry->field.mmiorw		= *rw;
-
+	ring_buffer_unlock_commit(tr->buffer, event, irq_flags);
-	__raw_spin_unlock(&data->lock);
-	raw_local_irq_restore(irq_flags);
 
 	trace_wake_up();
 }
@@ -331,19 +335,19 @@ static void __trace_mmiotrace_map(struct trace_array *tr,
 				struct trace_array_cpu *data,
 				struct mmiotrace_map *map)
 {
+	struct ring_buffer_event *event;
 	struct trace_entry *entry;
 	unsigned long irq_flags;
 
-	raw_local_irq_save(irq_flags);
+	event	= ring_buffer_lock_reserve(tr->buffer, sizeof(*entry),
-	__raw_spin_lock(&data->lock);
+					   &irq_flags);
-
+	if (!event)
-	entry				= tracing_get_trace_entry(tr, data);
+		return;
+	entry	= ring_buffer_event_data(event);
 	tracing_generic_entry_update(entry, 0);
 	entry->type			= TRACE_MMIO_MAP;
 	entry->field.mmiomap		= *map;
-
+	ring_buffer_unlock_commit(tr->buffer, event, irq_flags);
-	__raw_spin_unlock(&data->lock);
-	raw_local_irq_restore(irq_flags);
 
 	trace_wake_up();
 }

kernel/trace/trace_nop.c

@@ -30,7 +30,7 @@ static void nop_trace_init(struct trace_array *tr)
 	ctx_trace = tr;
 
 	for_each_online_cpu(cpu)
-		tracing_reset(tr->data[cpu]);
+		tracing_reset(tr, cpu);
 
 	if (tr->ctrl)
 		start_nop_trace(tr);

kernel/trace/trace_sched_switch.c

@@ -81,7 +81,7 @@ static void sched_switch_reset(struct trace_array *tr)
 	tr->time_start = ftrace_now(tr->cpu);
 
 	for_each_online_cpu(cpu)
-		tracing_reset(tr->data[cpu]);
+		tracing_reset(tr, cpu);
 }
 
 static int tracing_sched_register(void)

kernel/trace/trace_sched_wakeup.c

@@ -191,7 +191,7 @@ static void __wakeup_reset(struct trace_array *tr)
 
 	for_each_possible_cpu(cpu) {
 		data = tr->data[cpu];
-		tracing_reset(data);
+		tracing_reset(tr, cpu);
 	}
 
 	wakeup_cpu = -1;

kernel/trace/trace_selftest.c

@@ -18,58 +18,20 @@ static inline int trace_valid_entry(struct trace_entry *entry)
 	return 0;
 }
 
-static int
+static int trace_test_buffer_cpu(struct trace_array *tr, int cpu)
-trace_test_buffer_cpu(struct trace_array *tr, struct trace_array_cpu *data)
 {
-	struct trace_entry *entries;
+	struct ring_buffer_event *event;
-	struct page *page;
+	struct trace_entry *entry;
-	int idx = 0;
-	int i;
 
-	BUG_ON(list_empty(&data->trace_pages));
+	while ((event = ring_buffer_consume(tr->buffer, cpu, NULL))) {
-	page = list_entry(data->trace_pages.next, struct page, lru);
+		entry = ring_buffer_event_data(event);
-	entries = page_address(page);
 
-	check_pages(data);
+		if (!trace_valid_entry(entry)) {
-	if (head_page(data) != entries)
-		goto failed;
-
-	/*
-	 * The starting trace buffer always has valid elements,
-	 * if any element exists.
-	 */
-	entries = head_page(data);
-
-	for (i = 0; i < tr->entries; i++) {
-
-		if (i < data->trace_idx && !trace_valid_entry(&entries[idx])) {
 			printk(KERN_CONT ".. invalid entry %d ",
-				entries[idx].type);
+				entry->type);
 			goto failed;
 		}
-
-		idx++;
-		if (idx >= ENTRIES_PER_PAGE) {
-			page = virt_to_page(entries);
-			if (page->lru.next == &data->trace_pages) {
-				if (i != tr->entries - 1) {
-					printk(KERN_CONT ".. entries buffer mismatch");
-					goto failed;
-				}
-			} else {
-				page = list_entry(page->lru.next, struct page, lru);
-				entries = page_address(page);
-			}
-			idx = 0;
-		}
 	}
-
-	page = virt_to_page(entries);
-	if (page->lru.next != &data->trace_pages) {
-		printk(KERN_CONT ".. too many entries");
-		goto failed;
-	}
-
 	return 0;
 
  failed:
@@ -91,13 +53,11 @@ static int trace_test_buffer(struct trace_array *tr, unsigned long *count)
 	/* Don't allow flipping of max traces now */
 	raw_local_irq_save(flags);
 	__raw_spin_lock(&ftrace_max_lock);
+
+	cnt = ring_buffer_entries(tr->buffer);
+
 	for_each_possible_cpu(cpu) {
-		if (!head_page(tr->data[cpu]))
+		ret = trace_test_buffer_cpu(tr, cpu);
-			continue;
-
-		cnt += tr->data[cpu]->trace_idx;
-
-		ret = trace_test_buffer_cpu(tr, tr->data[cpu]);
 		if (ret)
 			break;
 	}

kernel/trace/trace_sysprof.c

@@ -241,7 +241,7 @@ static void stack_reset(struct trace_array *tr)
 	tr->time_start = ftrace_now(tr->cpu);
 
 	for_each_online_cpu(cpu)
-		tracing_reset(tr->data[cpu]);
+		tracing_reset(tr, cpu);
 }
 
 static void start_stack_trace(struct trace_array *tr)