Merge branch 'tracing/ftrace' into tracing/core

Merge reason: this mini-topic had outstanding problems that delayed
              its merge, so it does not fast-forward.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

Documentation/kernel-parameters.txt

@@ -751,12 +751,25 @@ and is between 256 and 4096 characters. It is defined in the file
 			ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
 
 	ftrace=[tracer]
-			[ftrace] will set and start the specified tracer
+			[FTRACE] will set and start the specified tracer
 			as early as possible in order to facilitate early
 			boot debugging.
 
 	ftrace_dump_on_oops
-			[ftrace] will dump the trace buffers on oops.
+			[FTRACE] will dump the trace buffers on oops.
+
+	ftrace_filter=[function-list]
+			[FTRACE] Limit the functions traced by the function
+			tracer at boot up. function-list is a comma separated
+			list of functions. This list can be changed at run
+			time by the set_ftrace_filter file in the debugfs
+			tracing directory. 
+
+	ftrace_notrace=[function-list]
+			[FTRACE] Do not trace the functions specified in
+			function-list. This list can be changed at run time
+			by the set_ftrace_notrace file in the debugfs
+			tracing directory.
 
 	gamecon.map[2|3]=
 			[HW,JOY] Multisystem joystick and NES/SNES/PSX pad

include/linux/ftrace_event.h

@@ -51,6 +51,7 @@ struct trace_iterator {
 	int			cpu_file;
 	struct mutex		mutex;
 	struct ring_buffer_iter	*buffer_iter[NR_CPUS];
+	unsigned long		iter_flags;
 
 	/* The below is zeroed out in pipe_read */
 	struct trace_seq	seq;
@@ -58,7 +59,6 @@ struct trace_iterator {
 	int			cpu;
 	u64			ts;
 
-	unsigned long		iter_flags;
 	loff_t			pos;
 	long			idx;
 

include/trace/events/irq.h

@@ -7,18 +7,18 @@
 #undef TRACE_SYSTEM
 #define TRACE_SYSTEM irq
 
-#define softirq_name(sirq) { sirq, #sirq }
-#define show_softirq_name(val)						\
-	__print_symbolic(val,						\
-			 softirq_name(HI_SOFTIRQ),			\
-			 softirq_name(TIMER_SOFTIRQ),			\
-			 softirq_name(NET_TX_SOFTIRQ),			\
-			 softirq_name(NET_RX_SOFTIRQ),			\
-			 softirq_name(BLOCK_SOFTIRQ),			\
-			 softirq_name(TASKLET_SOFTIRQ),			\
-			 softirq_name(SCHED_SOFTIRQ),			\
-			 softirq_name(HRTIMER_SOFTIRQ),			\
-			 softirq_name(RCU_SOFTIRQ))
+#define softirq_name(sirq) { sirq##_SOFTIRQ, #sirq }
+#define show_softirq_name(val)			\
+	__print_symbolic(val,			\
+			 softirq_name(HI),	\
+			 softirq_name(TIMER),	\
+			 softirq_name(NET_TX),	\
+			 softirq_name(NET_RX),	\
+			 softirq_name(BLOCK),	\
+			 softirq_name(TASKLET),	\
+			 softirq_name(SCHED),	\
+			 softirq_name(HRTIMER),	\
+			 softirq_name(RCU))
 
 /**
  * irq_handler_entry - called immediately before the irq action handler

include/trace/ftrace.h

@@ -18,14 +18,17 @@
 
 #include <linux/ftrace_event.h>
 
+#undef __field
+#define __field(type, item)		type	item;
+
 #undef __array
 #define __array(type, item, len)	type	item[len];
 
-#undef __field
-#define __field(type, item)		type	item;
+#undef __dynamic_array
+#define __dynamic_array(type, item, len) unsigned short __data_loc_##item;
 
 #undef __string
-#define __string(item, src)		unsigned short	__str_loc_##item;
+#define __string(item, src) __dynamic_array(char, item, -1)
 
 #undef TP_STRUCT__entry
 #define TP_STRUCT__entry(args...) args
@@ -35,7 +38,7 @@
 	struct ftrace_raw_##name {				\
 		struct trace_entry	ent;			\
 		tstruct						\
-		char			__str_data[0];		\
+		char			__data[0];		\
 	};							\
 	static struct ftrace_event_call event_##name
 
@@ -47,30 +50,31 @@
  *
  * Include the following:
  *
- * struct ftrace_str_offsets_<call> {
- *	int				<str1>;
- *	int				<str2>;
+ * struct ftrace_data_offsets_<call> {
+ *	int				<item1>;
+ *	int				<item2>;
  *	[...]
  * };
  *
- * The __string() macro will create each int <str>, this is to
- * keep the offset of each string from the beggining of the event
- * once we perform the strlen() of the src strings.
- *
+ * The __dynamic_array() macro will create each int <item>, this is
+ * to keep the offset of each array from the beginning of the event.
  */
 
+#undef __field
+#define __field(type, item);
+
 #undef __array
 #define __array(type, item, len)
 
-#undef __field
-#define __field(type, item);
+#undef __dynamic_array
+#define __dynamic_array(type, item, len)	int item;
 
 #undef __string
-#define __string(item, src)	int item;
+#define __string(item, src) __dynamic_array(char, item, -1)
 
 #undef TRACE_EVENT
 #define TRACE_EVENT(call, proto, args, tstruct, assign, print)		\
-	struct ftrace_str_offsets_##call {				\
+	struct ftrace_data_offsets_##call {				\
 		tstruct;						\
 	};
 
@@ -119,8 +123,12 @@
 #undef TP_printk
 #define TP_printk(fmt, args...) fmt "\n", args
 
+#undef __get_dynamic_array
+#define __get_dynamic_array(field)	\
+		((void *)__entry + __entry->__data_loc_##field)
+
 #undef __get_str
-#define __get_str(field)	((char *)__entry + __entry->__str_loc_##field)
+#define __get_str(field) (char *)__get_dynamic_array(field)
 
 #undef __print_flags
 #define __print_flags(flag, delim, flag_array...)			\
@@ -207,16 +215,19 @@ ftrace_raw_output_##call(struct trace_iterator *iter, int flags)	\
 	if (!ret)							\
 		return 0;
 
-#undef __string
-#define __string(item, src)						       \
-	ret = trace_seq_printf(s, "\tfield: __str_loc " #item ";\t"	       \
-			       "offset:%u;tsize:%u;\n",			       \
+#undef __dynamic_array
+#define __dynamic_array(type, item, len)				       \
+	ret = trace_seq_printf(s, "\tfield:__data_loc " #item ";\t"	       \
+			       "offset:%u;\tsize:%u;\n",		       \
 			       (unsigned int)offsetof(typeof(field),	       \
-					__str_loc_##item),		       \
-			       (unsigned int)sizeof(field.__str_loc_##item));  \
+					__data_loc_##item),		       \
+			       (unsigned int)sizeof(field.__data_loc_##item)); \
 	if (!ret)							       \
 		return 0;
 
+#undef __string
+#define __string(item, src) __dynamic_array(char, item, -1)
+
 #undef __entry
 #define __entry REC
 
@@ -260,11 +271,14 @@ ftrace_format_##call(struct trace_seq *s)				\
 	if (ret)							\
 		return ret;
 
+#undef __dynamic_array
+#define __dynamic_array(type, item, len)				       \
+	ret = trace_define_field(event_call, "__data_loc" "[" #type "]", #item,\
+				offsetof(typeof(field), __data_loc_##item),    \
+				 sizeof(field.__data_loc_##item), 0);
+
 #undef __string
-#define __string(item, src)						       \
-	ret = trace_define_field(event_call, "__str_loc", #item,	       \
-				offsetof(typeof(field), __str_loc_##item),     \
-				 sizeof(field.__str_loc_##item), 0);
+#define __string(item, src) __dynamic_array(char, item, -1)
 
 #undef TRACE_EVENT
 #define TRACE_EVENT(call, proto, args, tstruct, func, print)		\
@@ -288,6 +302,43 @@ ftrace_define_fields_##call(void)					\
 
 #include TRACE_INCLUDE(TRACE_INCLUDE_FILE)
 
+/*
+ * remember the offset of each array from the beginning of the event.
+ */
+
+#undef __entry
+#define __entry entry
+
+#undef __field
+#define __field(type, item)
+
+#undef __array
+#define __array(type, item, len)
+
+#undef __dynamic_array
+#define __dynamic_array(type, item, len)				\
+	__data_offsets->item = __data_size +				\
+			       offsetof(typeof(*entry), __data);	\
+	__data_size += (len) * sizeof(type);
+
+#undef __string
+#define __string(item, src) __dynamic_array(char, item, strlen(src) + 1)       \
+
+#undef TRACE_EVENT
+#define TRACE_EVENT(call, proto, args, tstruct, assign, print)		\
+static inline int ftrace_get_offsets_##call(				\
+	struct ftrace_data_offsets_##call *__data_offsets, proto)       \
+{									\
+	int __data_size = 0;						\
+	struct ftrace_raw_##call __maybe_unused *entry;			\
+									\
+	tstruct;							\
+									\
+	return __data_size;						\
+}
+
+#include TRACE_INCLUDE(TRACE_INCLUDE_FILE)
+
 /*
  * Stage 4 of the trace events.
  *
@@ -432,15 +483,15 @@ static void ftrace_profile_disable_##call(struct ftrace_event_call *event_call)\
 #undef __array
 #define __array(type, item, len)
 
+#undef __dynamic_array
+#define __dynamic_array(type, item, len)				\
+	__entry->__data_loc_##item = __data_offsets.item;
+
 #undef __string
-#define __string(item, src)						\
-	__str_offsets.item = __str_size +				\
-			     offsetof(typeof(*entry), __str_data);	\
-	__str_size += strlen(src) + 1;
+#define __string(item, src) __dynamic_array(char, item, -1)       	\
 
 #undef __assign_str
 #define __assign_str(dst, src)						\
-	__entry->__str_loc_##dst = __str_offsets.dst;			\
 	strcpy(__get_str(dst), src);
 
 #undef TRACE_EVENT
@@ -451,27 +502,30 @@ static struct ftrace_event_call event_##call;				\
 									\
 static void ftrace_raw_event_##call(proto)				\
 {									\
-	struct ftrace_str_offsets_##call __maybe_unused __str_offsets;	\
+	struct ftrace_data_offsets_##call __maybe_unused __data_offsets;\
 	struct ftrace_event_call *event_call = &event_##call;		\
 	struct ring_buffer_event *event;				\
 	struct ftrace_raw_##call *entry;				\
 	unsigned long irq_flags;					\
-	int __str_size = 0;						\
+	int __data_size;						\
 	int pc;								\
 									\
 	local_save_flags(irq_flags);					\
 	pc = preempt_count();						\
 									\
-	tstruct;							\
+	__data_size = ftrace_get_offsets_##call(&__data_offsets, args); \
 									\
 	event = trace_current_buffer_lock_reserve(event_##call.id,	\
-				 sizeof(struct ftrace_raw_##call) + __str_size,\
+				 sizeof(*entry) + __data_size,		\
 				 irq_flags, pc);			\
 	if (!event)							\
 		return;							\
 	entry	= ring_buffer_event_data(event);			\
 									\
-	assign;								\
+									\
+	tstruct								\
+									\
+	{ assign; }							\
 									\
 	if (!filter_current_check_discard(event_call, entry, event))	\
 		trace_nowake_buffer_unlock_commit(event, irq_flags, pc); \

kernel/trace/Kconfig

@@ -56,6 +56,13 @@ config CONTEXT_SWITCH_TRACER
 	select MARKERS
 	bool
 
+# All tracer options should select GENERIC_TRACER. For those options that are
+# enabled by all tracers (context switch and event tracer) they select TRACING.
+# This allows those options to appear when no other tracer is selected. But the
+# options do not appear when something else selects it. We need the two options
+# GENERIC_TRACER and TRACING to avoid circular dependencies to accomplish the
+# hidding of the automatic options options.
+
 config TRACING
 	bool
 	select DEBUG_FS
@@ -66,6 +73,10 @@ config TRACING
 	select BINARY_PRINTF
 	select EVENT_TRACING
 
+config GENERIC_TRACER
+	bool
+	select TRACING
+
 #
 # Minimum requirements an architecture has to meet for us to
 # be able to offer generic tracing facilities:
@@ -95,7 +106,7 @@ config FUNCTION_TRACER
 	depends on HAVE_FUNCTION_TRACER
 	select FRAME_POINTER
 	select KALLSYMS
-	select TRACING
+	select GENERIC_TRACER
 	select CONTEXT_SWITCH_TRACER
 	help
 	  Enable the kernel to trace every kernel function. This is done
@@ -126,7 +137,7 @@ config IRQSOFF_TRACER
 	depends on TRACE_IRQFLAGS_SUPPORT
 	depends on GENERIC_TIME
 	select TRACE_IRQFLAGS
-	select TRACING
+	select GENERIC_TRACER
 	select TRACER_MAX_TRACE
 	help
 	  This option measures the time spent in irqs-off critical
@@ -147,7 +158,7 @@ config PREEMPT_TRACER
 	default n
 	depends on GENERIC_TIME
 	depends on PREEMPT
-	select TRACING
+	select GENERIC_TRACER
 	select TRACER_MAX_TRACE
 	help
 	  This option measures the time spent in preemption off critical
@@ -166,7 +177,7 @@ config PREEMPT_TRACER
 config SYSPROF_TRACER
 	bool "Sysprof Tracer"
 	depends on X86
-	select TRACING
+	select GENERIC_TRACER
 	select CONTEXT_SWITCH_TRACER
 	help
 	  This tracer provides the trace needed by the 'Sysprof' userspace
@@ -174,44 +185,33 @@ config SYSPROF_TRACER
 
 config SCHED_TRACER
 	bool "Scheduling Latency Tracer"
-	select TRACING
+	select GENERIC_TRACER
 	select CONTEXT_SWITCH_TRACER
 	select TRACER_MAX_TRACE
 	help
 	  This tracer tracks the latency of the highest priority task
 	  to be scheduled in, starting from the point it has woken up.
 
-config ENABLE_CONTEXT_SWITCH_TRACER
-	bool "Trace process context switches"
-	select TRACING
-	select CONTEXT_SWITCH_TRACER
-	help
-	  This tracer gets called from the context switch and records
-	  all switching of tasks.
-
-config ENABLE_EVENT_TRACING
-	bool "Trace various events in the kernel"
+config ENABLE_DEFAULT_TRACERS
+	bool "Trace process context switches and events"
+	depends on !GENERIC_TRACER
 	select TRACING
 	help
 	  This tracer hooks to various trace points in the kernel
 	  allowing the user to pick and choose which trace point they
-	  want to trace.
-
-	  Note, all tracers enable event tracing. This option is
-	  only a convenience to enable event tracing when no other
-	  tracers are selected.
+	  want to trace. It also includes the sched_switch tracer plugin.
 
 config FTRACE_SYSCALLS
 	bool "Trace syscalls"
 	depends on HAVE_FTRACE_SYSCALLS
-	select TRACING
+	select GENERIC_TRACER
 	select KALLSYMS
 	help
 	  Basic tracer to catch the syscall entry and exit events.
 
 config BOOT_TRACER
 	bool "Trace boot initcalls"
-	select TRACING
+	select GENERIC_TRACER
 	select CONTEXT_SWITCH_TRACER
 	help
 	  This tracer helps developers to optimize boot times: it records
@@ -228,7 +228,7 @@ config BOOT_TRACER
 
 config TRACE_BRANCH_PROFILING
 	bool
-	select TRACING
+	select GENERIC_TRACER
 
 choice
 	prompt "Branch Profiling"
@@ -308,7 +308,7 @@ config BRANCH_TRACER
 config POWER_TRACER
 	bool "Trace power consumption behavior"
 	depends on X86
-	select TRACING
+	select GENERIC_TRACER
 	help
 	  This tracer helps developers to analyze and optimize the kernels
 	  power management decisions, specifically the C-state and P-state
@@ -342,14 +342,14 @@ config STACK_TRACER
 config HW_BRANCH_TRACER
 	depends on HAVE_HW_BRANCH_TRACER
 	bool "Trace hw branches"
-	select TRACING
+	select GENERIC_TRACER
 	help
 	  This tracer records all branches on the system in a circular
 	  buffer giving access to the last N branches for each cpu.
 
 config KMEMTRACE
 	bool "Trace SLAB allocations"
-	select TRACING
+	select GENERIC_TRACER
 	help
 	  kmemtrace provides tracing for slab allocator functions, such as
 	  kmalloc, kfree, kmem_cache_alloc, kmem_cache_free etc.. Collected
@@ -369,7 +369,7 @@ config KMEMTRACE
 
 config WORKQUEUE_TRACER
 	bool "Trace workqueues"
-	select TRACING
+	select GENERIC_TRACER
 	help
 	  The workqueue tracer provides some statistical informations
           about each cpu workqueue thread such as the number of the
@@ -385,7 +385,7 @@ config BLK_DEV_IO_TRACE
 	select RELAY
 	select DEBUG_FS
 	select TRACEPOINTS
-	select TRACING
+	select GENERIC_TRACER
 	select STACKTRACE
 	help
 	  Say Y here if you want to be able to trace the block layer actions
@@ -446,7 +446,7 @@ config FTRACE_SELFTEST
 
 config FTRACE_STARTUP_TEST
 	bool "Perform a startup test on ftrace"
-	depends on TRACING
+	depends on GENERIC_TRACER
 	select FTRACE_SELFTEST
 	help
 	  This option performs a series of startup tests on ftrace. On bootup
@@ -457,7 +457,7 @@ config FTRACE_STARTUP_TEST
 config MMIOTRACE
 	bool "Memory mapped IO tracing"
 	depends on HAVE_MMIOTRACE_SUPPORT && PCI
-	select TRACING
+	select GENERIC_TRACER
 	help
 	  Mmiotrace traces Memory Mapped I/O access and is meant for
 	  debugging and reverse engineering. It is called from the ioremap

kernel/trace/ftrace.c

@@ -32,6 +32,7 @@
 #include <trace/events/sched.h>
 
 #include <asm/ftrace.h>
+#include <asm/setup.h>
 
 #include "trace_output.h"
 #include "trace_stat.h"
@@ -598,7 +599,7 @@ function_profile_call(unsigned long ip, unsigned long parent_ip)
 	local_irq_save(flags);
 
 	stat = &__get_cpu_var(ftrace_profile_stats);
-	if (!stat->hash)
+	if (!stat->hash || !ftrace_profile_enabled)
 		goto out;
 
 	rec = ftrace_find_profiled_func(stat, ip);
@@ -629,7 +630,7 @@ static void profile_graph_return(struct ftrace_graph_ret *trace)
 
 	local_irq_save(flags);
 	stat = &__get_cpu_var(ftrace_profile_stats);
-	if (!stat->hash)
+	if (!stat->hash || !ftrace_profile_enabled)
 		goto out;
 
 	calltime = trace->rettime - trace->calltime;
@@ -723,6 +724,10 @@ ftrace_profile_write(struct file *filp, const char __user *ubuf,
 			ftrace_profile_enabled = 1;
 		} else {
 			ftrace_profile_enabled = 0;
+			/*
+			 * unregister_ftrace_profiler calls stop_machine
+			 * so this acts like an synchronize_sched.
+			 */
 			unregister_ftrace_profiler();
 		}
 	}
@@ -2369,6 +2374,45 @@ void ftrace_set_notrace(unsigned char *buf, int len, int reset)
 	ftrace_set_regex(buf, len, reset, 0);
 }
 
+/*
+ * command line interface to allow users to set filters on boot up.
+ */
+#define FTRACE_FILTER_SIZE		COMMAND_LINE_SIZE
+static char ftrace_notrace_buf[FTRACE_FILTER_SIZE] __initdata;
+static char ftrace_filter_buf[FTRACE_FILTER_SIZE] __initdata;
+
+static int __init set_ftrace_notrace(char *str)
+{
+	strncpy(ftrace_notrace_buf, str, FTRACE_FILTER_SIZE);
+	return 1;
+}
+__setup("ftrace_notrace=", set_ftrace_notrace);
+
+static int __init set_ftrace_filter(char *str)
+{
+	strncpy(ftrace_filter_buf, str, FTRACE_FILTER_SIZE);
+	return 1;
+}
+__setup("ftrace_filter=", set_ftrace_filter);
+
+static void __init set_ftrace_early_filter(char *buf, int enable)
+{
+	char *func;
+
+	while (buf) {
+		func = strsep(&buf, ",");
+		ftrace_set_regex(func, strlen(func), 0, enable);
+	}
+}
+
+static void __init set_ftrace_early_filters(void)
+{
+	if (ftrace_filter_buf[0])
+		set_ftrace_early_filter(ftrace_filter_buf, 1);
+	if (ftrace_notrace_buf[0])
+		set_ftrace_early_filter(ftrace_notrace_buf, 0);
+}
+
 static int
 ftrace_regex_release(struct inode *inode, struct file *file, int enable)
 {
@@ -2829,6 +2873,8 @@ void __init ftrace_init(void)
 	if (ret)
 		pr_warning("Failed to register trace ftrace module notifier\n");
 
+	set_ftrace_early_filters();
+
 	return;
  failed:
 	ftrace_disabled = 1;

kernel/trace/trace.c

@@ -2826,6 +2826,9 @@ static int tracing_open_pipe(struct inode *inode, struct file *filp)
 	/* trace pipe does not show start of buffer */
 	cpumask_setall(iter->started);
 
+	if (trace_flags & TRACE_ITER_LATENCY_FMT)
+		iter->iter_flags |= TRACE_FILE_LAT_FMT;
+
 	iter->cpu_file = cpu_file;
 	iter->tr = &global_trace;
 	mutex_init(&iter->mutex);

kernel/trace/trace_events_filter.c

@@ -478,12 +478,12 @@ enum {
 
 static int is_string_field(const char *type)
 {
+	if (strstr(type, "__data_loc") && strstr(type, "char"))
+		return FILTER_DYN_STRING;
+
 	if (strchr(type, '[') && strstr(type, "char"))
 		return FILTER_STATIC_STRING;
 
-	if (!strcmp(type, "__str_loc"))
-		return FILTER_DYN_STRING;
-
 	return 0;
 }
 

kernel/trace/trace_output.c

@@ -17,6 +17,7 @@
 static DECLARE_RWSEM(trace_event_mutex);
 
 DEFINE_PER_CPU(struct trace_seq, ftrace_event_seq);
+EXPORT_PER_CPU_SYMBOL(ftrace_event_seq);
 
 static struct hlist_head event_hash[EVENT_HASHSIZE] __read_mostly;
 
@@ -250,6 +251,7 @@ ftrace_print_flags_seq(struct trace_seq *p, const char *delim,
 
 	return p->buffer;
 }
+EXPORT_SYMBOL(ftrace_print_flags_seq);
 
 const char *
 ftrace_print_symbols_seq(struct trace_seq *p, unsigned long val,
@@ -275,6 +277,7 @@ ftrace_print_symbols_seq(struct trace_seq *p, unsigned long val,
 
 	return p->buffer;
 }
+EXPORT_SYMBOL(ftrace_print_symbols_seq);
 
 #ifdef CONFIG_KRETPROBES
 static inline const char *kretprobed(const char *name)