Merge branch 'perf/core' into perf/probes

Conflicts:
	tools/perf/Makefile

Merge reason:

 - fix the conflict
 - pick up the pr_*() infrastructure to queue up dependent patch
Signed-off-by: NIngo Molnar <mingo@elte.hu>

Documentation/kernel-parameters.txt

@@ -779,6 +779,13 @@ and is between 256 and 4096 characters. It is defined in the file
 			by the set_ftrace_notrace file in the debugfs
 			tracing directory.
 
+	ftrace_graph_filter=[function-list]
+			[FTRACE] Limit the top level callers functions traced
+			by the function graph tracer at boot up.
+			function-list is a comma separated list of functions
+			that can be changed at run time by the
+			set_graph_function file in the debugfs tracing directory.
+
 	gamecon.map[2|3]=
 			[HW,JOY] Multisystem joystick and NES/SNES/PSX pad
 			support via parallel port (up to 5 devices per port)

Documentation/trace/ftrace-design.txt

@@ -213,10 +213,19 @@ If you can't trace NMI functions, then skip this option.
 <details to be filled>
 
 
-HAVE_FTRACE_SYSCALLS
+HAVE_SYSCALL_TRACEPOINTS
 ---------------------
 
-<details to be filled>
+You need very few things to get the syscalls tracing in an arch.
+
+- Have a NR_syscalls variable in <asm/unistd.h> that provides the number
+  of syscalls supported by the arch.
+- Implement arch_syscall_addr() that resolves a syscall address from a
+  syscall number.
+- Support the TIF_SYSCALL_TRACEPOINT thread flags
+- Put the trace_sys_enter() and trace_sys_exit() tracepoints calls from ptrace
+  in the ptrace syscalls tracing path.
+- Tag this arch as HAVE_SYSCALL_TRACEPOINTS.
 
 
 HAVE_FTRACE_MCOUNT_RECORD

arch/s390/kernel/ftrace.c

@@ -203,73 +203,10 @@ unsigned long prepare_ftrace_return(unsigned long ip, unsigned long parent)
 
 #ifdef CONFIG_FTRACE_SYSCALLS
 
-extern unsigned long __start_syscalls_metadata[];
-extern unsigned long __stop_syscalls_metadata[];
 extern unsigned int sys_call_table[];
 
-static struct syscall_metadata **syscalls_metadata;
-
-struct syscall_metadata *syscall_nr_to_meta(int nr)
-{
-	if (!syscalls_metadata || nr >= NR_syscalls || nr < 0)
-		return NULL;
-
-	return syscalls_metadata[nr];
-}
-
-int syscall_name_to_nr(char *name)
-{
-	int i;
-
-	if (!syscalls_metadata)
-		return -1;
-	for (i = 0; i < NR_syscalls; i++)
-		if (syscalls_metadata[i])
-			if (!strcmp(syscalls_metadata[i]->name, name))
-				return i;
-	return -1;
-}
-
-void set_syscall_enter_id(int num, int id)
-{
-	syscalls_metadata[num]->enter_id = id;
-}
-
-void set_syscall_exit_id(int num, int id)
+unsigned long __init arch_syscall_addr(int nr)
 {
-	syscalls_metadata[num]->exit_id = id;
-}
-
-static struct syscall_metadata *find_syscall_meta(unsigned long syscall)
-{
-	struct syscall_metadata *start;
-	struct syscall_metadata *stop;
-	char str[KSYM_SYMBOL_LEN];
-
-	start = (struct syscall_metadata *)__start_syscalls_metadata;
-	stop = (struct syscall_metadata *)__stop_syscalls_metadata;
-	kallsyms_lookup(syscall, NULL, NULL, NULL, str);
-
-	for ( ; start < stop; start++) {
-		if (start->name && !strcmp(start->name + 3, str + 3))
-			return start;
-	}
-	return NULL;
-}
-
-static int __init arch_init_ftrace_syscalls(void)
-{
-	struct syscall_metadata *meta;
-	int i;
-	syscalls_metadata = kzalloc(sizeof(*syscalls_metadata) * NR_syscalls,
-				    GFP_KERNEL);
-	if (!syscalls_metadata)
-		return -ENOMEM;
-	for (i = 0; i < NR_syscalls; i++) {
-		meta = find_syscall_meta((unsigned long)sys_call_table[i]);
-		syscalls_metadata[i] = meta;
-	}
-	return 0;
+	return (unsigned long)sys_call_table[nr];
 }
-arch_initcall(arch_init_ftrace_syscalls);
 #endif

arch/x86/include/asm/perf_event.h

@@ -28,9 +28,20 @@
  */
 #define ARCH_PERFMON_EVENT_MASK				    0xffff
 
+/*
+ * filter mask to validate fixed counter events.
+ * the following filters disqualify for fixed counters:
+ *  - inv
+ *  - edge
+ *  - cnt-mask
+ *  The other filters are supported by fixed counters.
+ *  The any-thread option is supported starting with v3.
+ */
+#define ARCH_PERFMON_EVENT_FILTER_MASK			0xff840000
+
 #define ARCH_PERFMON_UNHALTED_CORE_CYCLES_SEL		      0x3c
 #define ARCH_PERFMON_UNHALTED_CORE_CYCLES_UMASK		(0x00 << 8)
-#define ARCH_PERFMON_UNHALTED_CORE_CYCLES_INDEX 		 0
+#define ARCH_PERFMON_UNHALTED_CORE_CYCLES_INDEX			 0
 #define ARCH_PERFMON_UNHALTED_CORE_CYCLES_PRESENT \
 		(1 << (ARCH_PERFMON_UNHALTED_CORE_CYCLES_INDEX))
 

arch/x86/kernel/cpu/perf_event.c

@@ -77,6 +77,18 @@ struct cpu_hw_events {
 	struct debug_store	*ds;
 };
 
+struct event_constraint {
+	unsigned long	idxmsk[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
+	int		code;
+};
+
+#define EVENT_CONSTRAINT(c, m) { .code = (c), .idxmsk[0] = (m) }
+#define EVENT_CONSTRAINT_END  { .code = 0, .idxmsk[0] = 0 }
+
+#define for_each_event_constraint(e, c) \
+	for ((e) = (c); (e)->idxmsk[0]; (e)++)
+
+
 /*
  * struct x86_pmu - generic x86 pmu
  */
@@ -102,6 +114,8 @@ struct x86_pmu {
 	u64		intel_ctrl;
 	void		(*enable_bts)(u64 config);
 	void		(*disable_bts)(void);
+	int		(*get_event_idx)(struct cpu_hw_events *cpuc,
+					 struct hw_perf_event *hwc);
 };
 
 static struct x86_pmu x86_pmu __read_mostly;
@@ -110,6 +124,8 @@ static DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events) = {
 	.enabled = 1,
 };
 
+static const struct event_constraint *event_constraints;
+
 /*
  * Not sure about some of these
  */
@@ -155,6 +171,16 @@ static u64 p6_pmu_raw_event(u64 hw_event)
 	return hw_event & P6_EVNTSEL_MASK;
 }
 
+static const struct event_constraint intel_p6_event_constraints[] =
+{
+	EVENT_CONSTRAINT(0xc1, 0x1),	/* FLOPS */
+	EVENT_CONSTRAINT(0x10, 0x1),	/* FP_COMP_OPS_EXE */
+	EVENT_CONSTRAINT(0x11, 0x1),	/* FP_ASSIST */
+	EVENT_CONSTRAINT(0x12, 0x2),	/* MUL */
+	EVENT_CONSTRAINT(0x13, 0x2),	/* DIV */
+	EVENT_CONSTRAINT(0x14, 0x1),	/* CYCLES_DIV_BUSY */
+	EVENT_CONSTRAINT_END
+};
 
 /*
  * Intel PerfMon v3. Used on Core2 and later.
@@ -170,6 +196,35 @@ static const u64 intel_perfmon_event_map[] =
   [PERF_COUNT_HW_BUS_CYCLES]		= 0x013c,
 };
 
+static const struct event_constraint intel_core_event_constraints[] =
+{
+	EVENT_CONSTRAINT(0x10, 0x1),	/* FP_COMP_OPS_EXE */
+	EVENT_CONSTRAINT(0x11, 0x2),	/* FP_ASSIST */
+	EVENT_CONSTRAINT(0x12, 0x2),	/* MUL */
+	EVENT_CONSTRAINT(0x13, 0x2),	/* DIV */
+	EVENT_CONSTRAINT(0x14, 0x1),	/* CYCLES_DIV_BUSY */
+	EVENT_CONSTRAINT(0x18, 0x1),	/* IDLE_DURING_DIV */
+	EVENT_CONSTRAINT(0x19, 0x2),	/* DELAYED_BYPASS */
+	EVENT_CONSTRAINT(0xa1, 0x1),	/* RS_UOPS_DISPATCH_CYCLES */
+	EVENT_CONSTRAINT(0xcb, 0x1),	/* MEM_LOAD_RETIRED */
+	EVENT_CONSTRAINT_END
+};
+
+static const struct event_constraint intel_nehalem_event_constraints[] =
+{
+	EVENT_CONSTRAINT(0x40, 0x3),	/* L1D_CACHE_LD */
+	EVENT_CONSTRAINT(0x41, 0x3),	/* L1D_CACHE_ST */
+	EVENT_CONSTRAINT(0x42, 0x3),	/* L1D_CACHE_LOCK */
+	EVENT_CONSTRAINT(0x43, 0x3),	/* L1D_ALL_REF */
+	EVENT_CONSTRAINT(0x4e, 0x3),	/* L1D_PREFETCH */
+	EVENT_CONSTRAINT(0x4c, 0x3),	/* LOAD_HIT_PRE */
+	EVENT_CONSTRAINT(0x51, 0x3),	/* L1D */
+	EVENT_CONSTRAINT(0x52, 0x3),	/* L1D_CACHE_PREFETCH_LOCK_FB_HIT */
+	EVENT_CONSTRAINT(0x53, 0x3),	/* L1D_CACHE_LOCK_FB_HIT */
+	EVENT_CONSTRAINT(0xc5, 0x3),	/* CACHE_LOCK_CYCLES */
+	EVENT_CONSTRAINT_END
+};
+
 static u64 intel_pmu_event_map(int hw_event)
 {
 	return intel_perfmon_event_map[hw_event];
@@ -469,7 +524,7 @@ static u64 intel_pmu_raw_event(u64 hw_event)
 #define CORE_EVNTSEL_UNIT_MASK		0x0000FF00ULL
 #define CORE_EVNTSEL_EDGE_MASK		0x00040000ULL
 #define CORE_EVNTSEL_INV_MASK		0x00800000ULL
-#define CORE_EVNTSEL_REG_MASK	0xFF000000ULL
+#define CORE_EVNTSEL_REG_MASK		0xFF000000ULL
 
 #define CORE_EVNTSEL_MASK		\
 	(CORE_EVNTSEL_EVENT_MASK |	\
@@ -932,6 +987,8 @@ static int __hw_perf_event_init(struct perf_event *event)
 	 */
 	hwc->config = ARCH_PERFMON_EVENTSEL_INT;
 
+	hwc->idx = -1;
+
 	/*
 	 * Count user and OS events unless requested not to.
 	 */
@@ -1334,8 +1391,7 @@ static void amd_pmu_enable_event(struct hw_perf_event *hwc, int idx)
 		x86_pmu_enable_event(hwc, idx);
 }
 
-static int
-fixed_mode_idx(struct perf_event *event, struct hw_perf_event *hwc)
+static int fixed_mode_idx(struct hw_perf_event *hwc)
 {
 	unsigned int hw_event;
 
@@ -1349,6 +1405,12 @@ fixed_mode_idx(struct perf_event *event, struct hw_perf_event *hwc)
 	if (!x86_pmu.num_events_fixed)
 		return -1;
 
+	/*
+	 * fixed counters do not take all possible filters
+	 */
+	if (hwc->config & ARCH_PERFMON_EVENT_FILTER_MASK)
+		return -1;
+
 	if (unlikely(hw_event == x86_pmu.event_map(PERF_COUNT_HW_INSTRUCTIONS)))
 		return X86_PMC_IDX_FIXED_INSTRUCTIONS;
 	if (unlikely(hw_event == x86_pmu.event_map(PERF_COUNT_HW_CPU_CYCLES)))
@@ -1360,22 +1422,57 @@ fixed_mode_idx(struct perf_event *event, struct hw_perf_event *hwc)
 }
 
 /*
- * Find a PMC slot for the freshly enabled / scheduled in event:
+ * generic counter allocator: get next free counter
  */
-static int x86_pmu_enable(struct perf_event *event)
+static int
+gen_get_event_idx(struct cpu_hw_events *cpuc, struct hw_perf_event *hwc)
+{
+	int idx;
+
+	idx = find_first_zero_bit(cpuc->used_mask, x86_pmu.num_events);
+	return idx == x86_pmu.num_events ? -1 : idx;
+}
+
+/*
+ * intel-specific counter allocator: check event constraints
+ */
+static int
+intel_get_event_idx(struct cpu_hw_events *cpuc, struct hw_perf_event *hwc)
+{
+	const struct event_constraint *event_constraint;
+	int i, code;
+
+	if (!event_constraints)
+		goto skip;
+
+	code = hwc->config & CORE_EVNTSEL_EVENT_MASK;
+
+	for_each_event_constraint(event_constraint, event_constraints) {
+		if (code == event_constraint->code) {
+			for_each_bit(i, event_constraint->idxmsk, X86_PMC_IDX_MAX) {
+				if (!test_and_set_bit(i, cpuc->used_mask))
+					return i;
+			}
+			return -1;
+		}
+	}
+skip:
+	return gen_get_event_idx(cpuc, hwc);
+}
+
+static int
+x86_schedule_event(struct cpu_hw_events *cpuc, struct hw_perf_event *hwc)
 {
-	struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
-	struct hw_perf_event *hwc = &event->hw;
 	int idx;
 
-	idx = fixed_mode_idx(event, hwc);
+	idx = fixed_mode_idx(hwc);
 	if (idx == X86_PMC_IDX_FIXED_BTS) {
 		/* BTS is already occupied. */
 		if (test_and_set_bit(idx, cpuc->used_mask))
 			return -EAGAIN;
 
 		hwc->config_base	= 0;
-		hwc->event_base	= 0;
+		hwc->event_base		= 0;
 		hwc->idx		= idx;
 	} else if (idx >= 0) {
 		/*
@@ -1396,20 +1493,35 @@ static int x86_pmu_enable(struct perf_event *event)
 	} else {
 		idx = hwc->idx;
 		/* Try to get the previous generic event again */
-		if (test_and_set_bit(idx, cpuc->used_mask)) {
+		if (idx == -1 || test_and_set_bit(idx, cpuc->used_mask)) {
 try_generic:
-			idx = find_first_zero_bit(cpuc->used_mask,
-						  x86_pmu.num_events);
-			if (idx == x86_pmu.num_events)
+			idx = x86_pmu.get_event_idx(cpuc, hwc);
+			if (idx == -1)
 				return -EAGAIN;
 
 			set_bit(idx, cpuc->used_mask);
 			hwc->idx = idx;
 		}
-		hwc->config_base  = x86_pmu.eventsel;
-		hwc->event_base = x86_pmu.perfctr;
+		hwc->config_base = x86_pmu.eventsel;
+		hwc->event_base  = x86_pmu.perfctr;
 	}
 
+	return idx;
+}
+
+/*
+ * Find a PMC slot for the freshly enabled / scheduled in event:
+ */
+static int x86_pmu_enable(struct perf_event *event)
+{
+	struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
+	struct hw_perf_event *hwc = &event->hw;
+	int idx;
+
+	idx = x86_schedule_event(cpuc, hwc);
+	if (idx < 0)
+		return idx;
+
 	perf_events_lapic_init();
 
 	x86_pmu.disable(hwc, idx);
@@ -1877,6 +1989,7 @@ static struct x86_pmu p6_pmu = {
 	 */
 	.event_bits		= 32,
 	.event_mask		= (1ULL << 32) - 1,
+	.get_event_idx		= intel_get_event_idx,
 };
 
 static struct x86_pmu intel_pmu = {
@@ -1900,6 +2013,7 @@ static struct x86_pmu intel_pmu = {
 	.max_period		= (1ULL << 31) - 1,
 	.enable_bts		= intel_pmu_enable_bts,
 	.disable_bts		= intel_pmu_disable_bts,
+	.get_event_idx		= intel_get_event_idx,
 };
 
 static struct x86_pmu amd_pmu = {
@@ -1920,6 +2034,7 @@ static struct x86_pmu amd_pmu = {
 	.apic			= 1,
 	/* use highest bit to detect overflow */
 	.max_period		= (1ULL << 47) - 1,
+	.get_event_idx		= gen_get_event_idx,
 };
 
 static int p6_pmu_init(void)
@@ -1932,10 +2047,12 @@ static int p6_pmu_init(void)
 	case 7:
 	case 8:
 	case 11: /* Pentium III */
+		event_constraints = intel_p6_event_constraints;
 		break;
 	case 9:
 	case 13:
 		/* Pentium M */
+		event_constraints = intel_p6_event_constraints;
 		break;
 	default:
 		pr_cont("unsupported p6 CPU model %d ",
@@ -2007,12 +2124,14 @@ static int intel_pmu_init(void)
 		       sizeof(hw_cache_event_ids));
 
 		pr_cont("Core2 events, ");
+		event_constraints = intel_core_event_constraints;
 		break;
 	default:
 	case 26:
 		memcpy(hw_cache_event_ids, nehalem_hw_cache_event_ids,
 		       sizeof(hw_cache_event_ids));
 
+		event_constraints = intel_nehalem_event_constraints;
 		pr_cont("Nehalem/Corei7 events, ");
 		break;
 	case 28:
@@ -2105,11 +2224,47 @@ static const struct pmu pmu = {
 	.unthrottle	= x86_pmu_unthrottle,
 };
 
+static int
+validate_event(struct cpu_hw_events *cpuc, struct perf_event *event)
+{
+	struct hw_perf_event fake_event = event->hw;
+
+	if (event->pmu != &pmu)
+		return 0;
+
+	return x86_schedule_event(cpuc, &fake_event);
+}
+
+static int validate_group(struct perf_event *event)
+{
+	struct perf_event *sibling, *leader = event->group_leader;
+	struct cpu_hw_events fake_pmu;
+
+	memset(&fake_pmu, 0, sizeof(fake_pmu));
+
+	if (!validate_event(&fake_pmu, leader))
+		return -ENOSPC;
+
+	list_for_each_entry(sibling, &leader->sibling_list, group_entry) {
+		if (!validate_event(&fake_pmu, sibling))
+			return -ENOSPC;
+	}
+
+	if (!validate_event(&fake_pmu, event))
+		return -ENOSPC;
+
+	return 0;
+}
+
 const struct pmu *hw_perf_event_init(struct perf_event *event)
 {
 	int err;
 
 	err = __hw_perf_event_init(event);
+	if (!err) {
+		if (event->group_leader != event)
+			err = validate_group(event);
+	}
 	if (err) {
 		if (event->destroy)
 			event->destroy(event);

arch/x86/kernel/entry_32.S

@@ -1209,17 +1209,14 @@ END(ftrace_graph_caller)
 
 .globl return_to_handler
 return_to_handler:
-	pushl $0
 	pushl %eax
-	pushl %ecx
 	pushl %edx
 	movl %ebp, %eax
 	call ftrace_return_to_handler
-	movl %eax, 0xc(%esp)
+	movl %eax, %ecx
 	popl %edx
-	popl %ecx
 	popl %eax
-	ret
+	jmp *%ecx
 #endif
 
 .section .rodata,"a"

arch/x86/kernel/entry_64.S

@@ -155,11 +155,11 @@ GLOBAL(return_to_handler)
 
 	call ftrace_return_to_handler
 
-	movq %rax, 16(%rsp)
+	movq %rax, %rdi
 	movq 8(%rsp), %rdx
 	movq (%rsp), %rax
-	addq $16, %rsp
-	retq
+	addq $24, %rsp
+	jmp *%rdi
 #endif
 
 

arch/x86/kernel/ftrace.c

@@ -9,6 +9,8 @@
  * the dangers of modifying code on the run.
  */
 
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
 #include <linux/spinlock.h>
 #include <linux/hardirq.h>
 #include <linux/uaccess.h>
@@ -336,15 +338,15 @@ int __init ftrace_dyn_arch_init(void *data)
 
 	switch (faulted) {
 	case 0:
-		pr_info("ftrace: converting mcount calls to 0f 1f 44 00 00\n");
+		pr_info("converting mcount calls to 0f 1f 44 00 00\n");
 		memcpy(ftrace_nop, ftrace_test_p6nop, MCOUNT_INSN_SIZE);
 		break;
 	case 1:
-		pr_info("ftrace: converting mcount calls to 66 66 66 66 90\n");
+		pr_info("converting mcount calls to 66 66 66 66 90\n");
 		memcpy(ftrace_nop, ftrace_test_nop5, MCOUNT_INSN_SIZE);
 		break;
 	case 2:
-		pr_info("ftrace: converting mcount calls to jmp . + 5\n");
+		pr_info("converting mcount calls to jmp . + 5\n");
 		memcpy(ftrace_nop, ftrace_test_jmp, MCOUNT_INSN_SIZE);
 		break;
 	}
@@ -468,82 +470,10 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr,
 
 #ifdef CONFIG_FTRACE_SYSCALLS
 
-extern unsigned long __start_syscalls_metadata[];
-extern unsigned long __stop_syscalls_metadata[];
 extern unsigned long *sys_call_table;
 
-static struct syscall_metadata **syscalls_metadata;
-
-static struct syscall_metadata *find_syscall_meta(unsigned long *syscall)
-{
-	struct syscall_metadata *start;
-	struct syscall_metadata *stop;
-	char str[KSYM_SYMBOL_LEN];
-
-
-	start = (struct syscall_metadata *)__start_syscalls_metadata;
-	stop = (struct syscall_metadata *)__stop_syscalls_metadata;
-	kallsyms_lookup((unsigned long) syscall, NULL, NULL, NULL, str);
-
-	for ( ; start < stop; start++) {
-		if (start->name && !strcmp(start->name, str))
-			return start;
-	}
-	return NULL;
-}
-
-struct syscall_metadata *syscall_nr_to_meta(int nr)
-{
-	if (!syscalls_metadata || nr >= NR_syscalls || nr < 0)
-		return NULL;
-
-	return syscalls_metadata[nr];
-}
-
-int syscall_name_to_nr(char *name)
+unsigned long __init arch_syscall_addr(int nr)
 {
-	int i;
-
-	if (!syscalls_metadata)
-		return -1;
-
-	for (i = 0; i < NR_syscalls; i++) {
-		if (syscalls_metadata[i]) {
-			if (!strcmp(syscalls_metadata[i]->name, name))
-				return i;
-		}
-	}
-	return -1;
-}
-
-void set_syscall_enter_id(int num, int id)
-{
-	syscalls_metadata[num]->enter_id = id;
-}
-
-void set_syscall_exit_id(int num, int id)
-{
-	syscalls_metadata[num]->exit_id = id;
-}
-
-static int __init arch_init_ftrace_syscalls(void)
-{
-	int i;
-	struct syscall_metadata *meta;
-	unsigned long **psys_syscall_table = &sys_call_table;
-
-	syscalls_metadata = kzalloc(sizeof(*syscalls_metadata) *
-					NR_syscalls, GFP_KERNEL);
-	if (!syscalls_metadata) {
-		WARN_ON(1);
-		return -ENOMEM;
-	}
-
-	for (i = 0; i < NR_syscalls; i++) {
-		meta = find_syscall_meta(psys_syscall_table[i]);
-		syscalls_metadata[i] = meta;
-	}
-	return 0;
+	return (unsigned long)(&sys_call_table)[nr];
 }
-arch_initcall(arch_init_ftrace_syscalls);
 #endif

arch/x86/mm/testmmiotrace.c

 /*
  * Written by Pekka Paalanen, 2008-2009 <pq@iki.fi>
  */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
 #include <linux/module.h>
 #include <linux/io.h>
 #include <linux/mmiotrace.h>
 
-#define MODULE_NAME "testmmiotrace"
-
 static unsigned long mmio_address;
 module_param(mmio_address, ulong, 0);
 MODULE_PARM_DESC(mmio_address, " Start address of the mapping of 16 kB "
@@ -30,7 +31,7 @@ static unsigned v32(unsigned i)
 static void do_write_test(void __iomem *p)
 {
 	unsigned int i;
-	pr_info(MODULE_NAME ": write test.\n");
+	pr_info("write test.\n");
 	mmiotrace_printk("Write test.\n");
 
 	for (i = 0; i < 256; i++)
@@ -47,7 +48,7 @@ static void do_read_test(void __iomem *p)
 {
 	unsigned int i;
 	unsigned errs[3] = { 0 };
-	pr_info(MODULE_NAME ": read test.\n");
+	pr_info("read test.\n");
 	mmiotrace_printk("Read test.\n");
 
 	for (i = 0; i < 256; i++)
@@ -68,7 +69,7 @@ static void do_read_test(void __iomem *p)
 
 static void do_read_far_test(void __iomem *p)
 {
-	pr_info(MODULE_NAME ": read far test.\n");
+	pr_info("read far test.\n");
 	mmiotrace_printk("Read far test.\n");
 
 	ioread32(p + read_far);
@@ -78,7 +79,7 @@ static void do_test(unsigned long size)
 {
 	void __iomem *p = ioremap_nocache(mmio_address, size);
 	if (!p) {
-		pr_err(MODULE_NAME ": could not ioremap, aborting.\n");
+		pr_err("could not ioremap, aborting.\n");
 		return;
 	}
 	mmiotrace_printk("ioremap returned %p.\n", p);
@@ -94,24 +95,22 @@ static int __init init(void)
 	unsigned long size = (read_far) ? (8 << 20) : (16 << 10);
 
 	if (mmio_address == 0) {
-		pr_err(MODULE_NAME ": you have to use the module argument "
-							"mmio_address.\n");
-		pr_err(MODULE_NAME ": DO NOT LOAD THIS MODULE UNLESS"
-				" YOU REALLY KNOW WHAT YOU ARE DOING!\n");
+		pr_err("you have to use the module argument mmio_address.\n");
+		pr_err("DO NOT LOAD THIS MODULE UNLESS YOU REALLY KNOW WHAT YOU ARE DOING!\n");
 		return -ENXIO;
 	}
 
-	pr_warning(MODULE_NAME ": WARNING: mapping %lu kB @ 0x%08lx in PCI "
-		"address space, and writing 16 kB of rubbish in there.\n",
-		 size >> 10, mmio_address);
+	pr_warning("WARNING: mapping %lu kB @ 0x%08lx in PCI address space, "
+		   "and writing 16 kB of rubbish in there.\n",
+		   size >> 10, mmio_address);
 	do_test(size);
-	pr_info(MODULE_NAME ": All done.\n");
+	pr_info("All done.\n");
 	return 0;
 }
 
 static void __exit cleanup(void)
 {
-	pr_debug(MODULE_NAME ": unloaded.\n");
+	pr_debug("unloaded.\n");
 }
 
 module_init(init);

include/linux/ftrace_event.h

@@ -144,7 +144,7 @@ extern char			*trace_profile_buf_nmi;
 #define MAX_FILTER_STR_VAL	256	/* Should handle KSYM_SYMBOL_LEN */
 
 extern void destroy_preds(struct ftrace_event_call *call);
-extern int filter_match_preds(struct ftrace_event_call *call, void *rec);
+extern int filter_match_preds(struct event_filter *filter, void *rec);
 extern int filter_current_check_discard(struct ring_buffer *buffer,
 					struct ftrace_event_call *call,
 					void *rec,
@@ -187,4 +187,13 @@ do {									\
 		__trace_printk(ip, fmt, ##args);			\
 } while (0)
 
+#ifdef CONFIG_EVENT_PROFILE
+struct perf_event;
+extern int ftrace_profile_enable(int event_id);
+extern void ftrace_profile_disable(int event_id);
+extern int ftrace_profile_set_filter(struct perf_event *event, int event_id,
+				     char *filter_str);
+extern void ftrace_profile_free_filter(struct perf_event *event);
+#endif
+
 #endif /* _LINUX_FTRACE_EVENT_H */

include/linux/perf_counter.h

@@ -225,6 +225,7 @@ struct perf_counter_attr {
 #define PERF_COUNTER_IOC_RESET		_IO ('$', 3)
 #define PERF_COUNTER_IOC_PERIOD		_IOW('$', 4, u64)
 #define PERF_COUNTER_IOC_SET_OUTPUT	_IO ('$', 5)
+#define PERF_COUNTER_IOC_SET_FILTER	_IOW('$', 6, char *)
 
 enum perf_counter_ioc_flags {
 	PERF_IOC_FLAG_GROUP		= 1U << 0,

include/linux/perf_event.h

@@ -221,6 +221,7 @@ struct perf_event_attr {
 #define PERF_EVENT_IOC_RESET		_IO ('$', 3)
 #define PERF_EVENT_IOC_PERIOD		_IOW('$', 4, u64)
 #define PERF_EVENT_IOC_SET_OUTPUT	_IO ('$', 5)
+#define PERF_EVENT_IOC_SET_FILTER	_IOW('$', 6, char *)
 
 enum perf_event_ioc_flags {
 	PERF_IOC_FLAG_GROUP		= 1U << 0,
@@ -633,7 +634,12 @@ struct perf_event {
 
 	struct pid_namespace		*ns;
 	u64				id;
+
+#ifdef CONFIG_EVENT_PROFILE
+	struct event_filter		*filter;
 #endif
+
+#endif /* CONFIG_PERF_EVENTS */
 };
 
 /**

include/linux/smp_lock.h

@@ -24,8 +24,21 @@ static inline int reacquire_kernel_lock(struct task_struct *task)
 	return 0;
 }
 
-extern void __lockfunc lock_kernel(void)	__acquires(kernel_lock);
-extern void __lockfunc unlock_kernel(void)	__releases(kernel_lock);
+extern void __lockfunc
+_lock_kernel(const char *func, const char *file, int line)
+__acquires(kernel_lock);
+
+extern void __lockfunc
+_unlock_kernel(const char *func, const char *file, int line)
+__releases(kernel_lock);
+
+#define lock_kernel() do {					\
+	_lock_kernel(__func__, __FILE__, __LINE__);		\
+} while (0)
+
+#define unlock_kernel()	do {					\
+	_unlock_kernel(__func__, __FILE__, __LINE__);		\
+} while (0)
 
 /*
  * Various legacy drivers don't really need the BKL in a specific
@@ -41,8 +54,8 @@ static inline void cycle_kernel_lock(void)
 
 #else
 
-#define lock_kernel()				do { } while(0)
-#define unlock_kernel()				do { } while(0)
+#define lock_kernel()
+#define unlock_kernel()
 #define release_kernel_lock(task)		do { } while(0)
 #define cycle_kernel_lock()			do { } while(0)
 #define reacquire_kernel_lock(task)		0

include/trace/events/bkl.h

+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM bkl
+
+#if !defined(_TRACE_BKL_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_BKL_H
+
+#include <linux/tracepoint.h>
+
+TRACE_EVENT(lock_kernel,
+
+	TP_PROTO(const char *func, const char *file, int line),
+
+	TP_ARGS(func, file, line),
+
+	TP_STRUCT__entry(
+		__field(	int,		depth			)
+		__field_ext(	const char *,	func, FILTER_PTR_STRING	)
+		__field_ext(	const char *,	file, FILTER_PTR_STRING	)
+		__field(	int,		line			)
+	),
+
+	TP_fast_assign(
+		/* We want to record the lock_depth after lock is acquired */
+		__entry->depth = current->lock_depth + 1;
+		__entry->func = func;
+		__entry->file = file;
+		__entry->line = line;
+	),
+
+	TP_printk("depth=%d file:line=%s:%d func=%s()", __entry->depth,
+		  __entry->file, __entry->line, __entry->func)
+);
+
+TRACE_EVENT(unlock_kernel,
+
+	TP_PROTO(const char *func, const char *file, int line),
+
+	TP_ARGS(func, file, line),
+
+	TP_STRUCT__entry(
+		__field(int,		depth		)
+		__field(const char *,	func		)
+		__field(const char *,	file		)
+		__field(int,		line		)
+	),
+
+	TP_fast_assign(
+		__entry->depth = current->lock_depth;
+		__entry->func = func;
+		__entry->file = file;
+		__entry->line = line;
+	),
+
+	TP_printk("depth=%d file:line=%s:%d func=%s()", __entry->depth,
+		  __entry->file, __entry->line, __entry->func)
+);
+
+#endif /* _TRACE_BKL_H */
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>

include/trace/events/irq.h

@@ -48,7 +48,7 @@ TRACE_EVENT(irq_handler_entry,
 		__assign_str(name, action->name);
 	),
 
-	TP_printk("irq=%d handler=%s", __entry->irq, __get_str(name))
+	TP_printk("irq=%d name=%s", __entry->irq, __get_str(name))
 );
 
 /**
@@ -78,7 +78,7 @@ TRACE_EVENT(irq_handler_exit,
 		__entry->ret	= ret;
 	),
 
-	TP_printk("irq=%d return=%s",
+	TP_printk("irq=%d ret=%s",
 		  __entry->irq, __entry->ret ? "handled" : "unhandled")
 );
 
@@ -107,7 +107,7 @@ TRACE_EVENT(softirq_entry,
 		__entry->vec = (int)(h - vec);
 	),
 
-	TP_printk("softirq=%d action=%s", __entry->vec,
+	TP_printk("vec=%d [action=%s]", __entry->vec,
 		  show_softirq_name(__entry->vec))
 );
 
@@ -136,7 +136,7 @@ TRACE_EVENT(softirq_exit,
 		__entry->vec = (int)(h - vec);
 	),
 
-	TP_printk("softirq=%d action=%s", __entry->vec,
+	TP_printk("vec=%d [action=%s]", __entry->vec,
 		  show_softirq_name(__entry->vec))
 );
 

include/trace/events/power.h

@@ -16,8 +16,6 @@ enum {
 };
 #endif
 
-
-
 TRACE_EVENT(power_start,
 
 	TP_PROTO(unsigned int type, unsigned int state),

include/trace/events/sched.h

@@ -26,7 +26,7 @@ TRACE_EVENT(sched_kthread_stop,
 		__entry->pid	= t->pid;
 	),
 
-	TP_printk("task %s:%d", __entry->comm, __entry->pid)
+	TP_printk("comm=%s pid=%d", __entry->comm, __entry->pid)
 );
 
 /*
@@ -46,7 +46,7 @@ TRACE_EVENT(sched_kthread_stop_ret,
 		__entry->ret	= ret;
 	),
 
-	TP_printk("ret %d", __entry->ret)
+	TP_printk("ret=%d", __entry->ret)
 );
 
 /*
@@ -73,7 +73,7 @@ TRACE_EVENT(sched_wait_task,
 		__entry->prio	= p->prio;
 	),
 
-	TP_printk("task %s:%d [%d]",
+	TP_printk("comm=%s pid=%d prio=%d",
 		  __entry->comm, __entry->pid, __entry->prio)
 );
 
@@ -94,7 +94,7 @@ TRACE_EVENT(sched_wakeup,
 		__field(	pid_t,	pid			)
 		__field(	int,	prio			)
 		__field(	int,	success			)
-		__field(	int,	cpu			)
+		__field(	int,	target_cpu		)
 	),
 
 	TP_fast_assign(
@@ -102,12 +102,12 @@ TRACE_EVENT(sched_wakeup,
 		__entry->pid		= p->pid;
 		__entry->prio		= p->prio;
 		__entry->success	= success;
-		__entry->cpu		= task_cpu(p);
+		__entry->target_cpu	= task_cpu(p);
 	),
 
-	TP_printk("task %s:%d [%d] success=%d [%03d]",
+	TP_printk("comm=%s pid=%d prio=%d success=%d target_cpu=%03d",
 		  __entry->comm, __entry->pid, __entry->prio,
-		  __entry->success, __entry->cpu)
+		  __entry->success, __entry->target_cpu)
 );
 
 /*
@@ -127,7 +127,7 @@ TRACE_EVENT(sched_wakeup_new,
 		__field(	pid_t,	pid			)
 		__field(	int,	prio			)
 		__field(	int,	success			)
-		__field(	int,	cpu			)
+		__field(	int,	target_cpu		)
 	),
 
 	TP_fast_assign(
@@ -135,12 +135,12 @@ TRACE_EVENT(sched_wakeup_new,
 		__entry->pid		= p->pid;
 		__entry->prio		= p->prio;
 		__entry->success	= success;
-		__entry->cpu		= task_cpu(p);
+		__entry->target_cpu	= task_cpu(p);
 	),
 
-	TP_printk("task %s:%d [%d] success=%d [%03d]",
+	TP_printk("comm=%s pid=%d prio=%d success=%d target_cpu=%03d",
 		  __entry->comm, __entry->pid, __entry->prio,
-		  __entry->success, __entry->cpu)
+		  __entry->success, __entry->target_cpu)
 );
 
 /*
@@ -176,7 +176,7 @@ TRACE_EVENT(sched_switch,
 		__entry->next_prio	= next->prio;
 	),
 
-	TP_printk("task %s:%d [%d] (%s) ==> %s:%d [%d]",
+	TP_printk("prev_comm=%s prev_pid=%d prev_prio=%d prev_state=%s ==> next_comm=%s next_pid=%d next_prio=%d",
 		__entry->prev_comm, __entry->prev_pid, __entry->prev_prio,
 		__entry->prev_state ?
 		  __print_flags(__entry->prev_state, "|",
@@ -211,7 +211,7 @@ TRACE_EVENT(sched_migrate_task,
 		__entry->dest_cpu	= dest_cpu;
 	),
 
-	TP_printk("task %s:%d [%d] from: %d  to: %d",
+	TP_printk("comm=%s pid=%d prio=%d orig_cpu=%d dest_cpu=%d",
 		  __entry->comm, __entry->pid, __entry->prio,
 		  __entry->orig_cpu, __entry->dest_cpu)
 );
@@ -237,7 +237,7 @@ TRACE_EVENT(sched_process_free,
 		__entry->prio		= p->prio;
 	),
 
-	TP_printk("task %s:%d [%d]",
+	TP_printk("comm=%s pid=%d prio=%d",
 		  __entry->comm, __entry->pid, __entry->prio)
 );
 
@@ -262,7 +262,7 @@ TRACE_EVENT(sched_process_exit,
 		__entry->prio		= p->prio;
 	),
 
-	TP_printk("task %s:%d [%d]",
+	TP_printk("comm=%s pid=%d prio=%d",
 		  __entry->comm, __entry->pid, __entry->prio)
 );
 
@@ -287,7 +287,7 @@ TRACE_EVENT(sched_process_wait,
 		__entry->prio		= current->prio;
 	),
 
-	TP_printk("task %s:%d [%d]",
+	TP_printk("comm=%s pid=%d prio=%d",
 		  __entry->comm, __entry->pid, __entry->prio)
 );
 
@@ -314,7 +314,7 @@ TRACE_EVENT(sched_process_fork,
 		__entry->child_pid	= child->pid;
 	),
 
-	TP_printk("parent %s:%d  child %s:%d",
+	TP_printk("comm=%s pid=%d child_comm=%s child_pid=%d",
 		__entry->parent_comm, __entry->parent_pid,
 		__entry->child_comm, __entry->child_pid)
 );
@@ -340,7 +340,7 @@ TRACE_EVENT(sched_signal_send,
 		__entry->sig	= sig;
 	),
 
-	TP_printk("sig: %d  task %s:%d",
+	TP_printk("sig=%d comm=%s pid=%d",
 		  __entry->sig, __entry->comm, __entry->pid)
 );
 
@@ -374,7 +374,7 @@ TRACE_EVENT(sched_stat_wait,
 		__perf_count(delay);
 	),
 
-	TP_printk("task: %s:%d wait: %Lu [ns]",
+	TP_printk("comm=%s pid=%d delay=%Lu [ns]",
 			__entry->comm, __entry->pid,
 			(unsigned long long)__entry->delay)
 );
@@ -406,7 +406,7 @@ TRACE_EVENT(sched_stat_runtime,
 		__perf_count(runtime);
 	),
 
-	TP_printk("task: %s:%d runtime: %Lu [ns], vruntime: %Lu [ns]",
+	TP_printk("comm=%s pid=%d runtime=%Lu [ns] vruntime=%Lu [ns]",
 			__entry->comm, __entry->pid,
 			(unsigned long long)__entry->runtime,
 			(unsigned long long)__entry->vruntime)
@@ -437,7 +437,7 @@ TRACE_EVENT(sched_stat_sleep,
 		__perf_count(delay);
 	),
 
-	TP_printk("task: %s:%d sleep: %Lu [ns]",
+	TP_printk("comm=%s pid=%d delay=%Lu [ns]",
 			__entry->comm, __entry->pid,
 			(unsigned long long)__entry->delay)
 );
@@ -467,7 +467,7 @@ TRACE_EVENT(sched_stat_iowait,
 		__perf_count(delay);
 	),
 
-	TP_printk("task: %s:%d iowait: %Lu [ns]",
+	TP_printk("comm=%s pid=%d delay=%Lu [ns]",
 			__entry->comm, __entry->pid,
 			(unsigned long long)__entry->delay)
 );

include/trace/events/timer.h

@@ -26,7 +26,7 @@ TRACE_EVENT(timer_init,
 		__entry->timer	= timer;
 	),
 
-	TP_printk("timer %p", __entry->timer)
+	TP_printk("timer=%p", __entry->timer)
 );
 
 /**
@@ -54,7 +54,7 @@ TRACE_EVENT(timer_start,
 		__entry->now		= jiffies;
 	),
 
-	TP_printk("timer %p: func %pf, expires %lu, timeout %ld",
+	TP_printk("timer=%p function=%pf expires=%lu [timeout=%ld]",
 		  __entry->timer, __entry->function, __entry->expires,
 		  (long)__entry->expires - __entry->now)
 );
@@ -81,7 +81,7 @@ TRACE_EVENT(timer_expire_entry,
 		__entry->now		= jiffies;
 	),
 
-	TP_printk("timer %p: now %lu", __entry->timer, __entry->now)
+	TP_printk("timer=%p now=%lu", __entry->timer, __entry->now)
 );
 
 /**
@@ -108,7 +108,7 @@ TRACE_EVENT(timer_expire_exit,
 		__entry->timer	= timer;
 	),
 
-	TP_printk("timer %p", __entry->timer)
+	TP_printk("timer=%p", __entry->timer)
 );
 
 /**
@@ -129,7 +129,7 @@ TRACE_EVENT(timer_cancel,
 		__entry->timer	= timer;
 	),
 
-	TP_printk("timer %p", __entry->timer)
+	TP_printk("timer=%p", __entry->timer)
 );
 
 /**
@@ -140,24 +140,24 @@ TRACE_EVENT(timer_cancel,
  */
 TRACE_EVENT(hrtimer_init,
 
-	TP_PROTO(struct hrtimer *timer, clockid_t clockid,
+	TP_PROTO(struct hrtimer *hrtimer, clockid_t clockid,
 		 enum hrtimer_mode mode),
 
-	TP_ARGS(timer, clockid, mode),
+	TP_ARGS(hrtimer, clockid, mode),
 
 	TP_STRUCT__entry(
-		__field( void *,		timer		)
+		__field( void *,		hrtimer		)
 		__field( clockid_t,		clockid		)
 		__field( enum hrtimer_mode,	mode		)
 	),
 
 	TP_fast_assign(
-		__entry->timer		= timer;
+		__entry->hrtimer	= hrtimer;
 		__entry->clockid	= clockid;
 		__entry->mode		= mode;
 	),
 
-	TP_printk("hrtimer %p, clockid %s, mode %s", __entry->timer,
+	TP_printk("hrtimer=%p clockid=%s mode=%s", __entry->hrtimer,
 		  __entry->clockid == CLOCK_REALTIME ?
 			"CLOCK_REALTIME" : "CLOCK_MONOTONIC",
 		  __entry->mode == HRTIMER_MODE_ABS ?
@@ -170,26 +170,26 @@ TRACE_EVENT(hrtimer_init,
  */
 TRACE_EVENT(hrtimer_start,
 
-	TP_PROTO(struct hrtimer *timer),
+	TP_PROTO(struct hrtimer *hrtimer),
 
-	TP_ARGS(timer),
+	TP_ARGS(hrtimer),
 
 	TP_STRUCT__entry(
-		__field( void *,	timer		)
+		__field( void *,	hrtimer		)
 		__field( void *,	function	)
 		__field( s64,		expires		)
 		__field( s64,		softexpires	)
 	),
 
 	TP_fast_assign(
-		__entry->timer		= timer;
-		__entry->function	= timer->function;
-		__entry->expires	= hrtimer_get_expires(timer).tv64;
-		__entry->softexpires	= hrtimer_get_softexpires(timer).tv64;
+		__entry->hrtimer	= hrtimer;
+		__entry->function	= hrtimer->function;
+		__entry->expires	= hrtimer_get_expires(hrtimer).tv64;
+		__entry->softexpires	= hrtimer_get_softexpires(hrtimer).tv64;
 	),
 
-	TP_printk("hrtimer %p, func %pf, expires %llu, softexpires %llu",
-		  __entry->timer, __entry->function,
+	TP_printk("hrtimer=%p function=%pf expires=%llu softexpires=%llu",
+		  __entry->hrtimer, __entry->function,
 		  (unsigned long long)ktime_to_ns((ktime_t) {
 				  .tv64 = __entry->expires }),
 		  (unsigned long long)ktime_to_ns((ktime_t) {
@@ -206,23 +206,22 @@ TRACE_EVENT(hrtimer_start,
  */
 TRACE_EVENT(hrtimer_expire_entry,
 
-	TP_PROTO(struct hrtimer *timer, ktime_t *now),
+	TP_PROTO(struct hrtimer *hrtimer, ktime_t *now),
 
-	TP_ARGS(timer, now),
+	TP_ARGS(hrtimer, now),
 
 	TP_STRUCT__entry(
-		__field( void *,	timer	)
+		__field( void *,	hrtimer	)
 		__field( s64,		now	)
 	),
 
 	TP_fast_assign(
-		__entry->timer	= timer;
-		__entry->now	= now->tv64;
+		__entry->hrtimer	= hrtimer;
+		__entry->now		= now->tv64;
 	),
 
-	TP_printk("hrtimer %p, now %llu", __entry->timer,
-		  (unsigned long long)ktime_to_ns((ktime_t) {
-				  .tv64 = __entry->now }))
+	TP_printk("hrtimer=%p now=%llu", __entry->hrtimer,
+		  (unsigned long long)ktime_to_ns((ktime_t) { .tv64 = __entry->now }))
  );
 
 /**
@@ -234,40 +233,40 @@ TRACE_EVENT(hrtimer_expire_entry,
  */
 TRACE_EVENT(hrtimer_expire_exit,
 
-	TP_PROTO(struct hrtimer *timer),
+	TP_PROTO(struct hrtimer *hrtimer),
 
-	TP_ARGS(timer),
+	TP_ARGS(hrtimer),
 
 	TP_STRUCT__entry(
-		__field( void *,	timer	)
+		__field( void *,	hrtimer	)
 	),
 
 	TP_fast_assign(
-		__entry->timer	= timer;
+		__entry->hrtimer	= hrtimer;
 	),
 
-	TP_printk("hrtimer %p", __entry->timer)
+	TP_printk("hrtimer=%p", __entry->hrtimer)
 );
 
 /**
  * hrtimer_cancel - called when the hrtimer is canceled
- * @timer:	pointer to struct hrtimer
+ * @hrtimer:	pointer to struct hrtimer
  */
 TRACE_EVENT(hrtimer_cancel,
 
-	TP_PROTO(struct hrtimer *timer),
+	TP_PROTO(struct hrtimer *hrtimer),
 
-	TP_ARGS(timer),
+	TP_ARGS(hrtimer),
 
 	TP_STRUCT__entry(
-		__field( void *,	timer	)
+		__field( void *,	hrtimer	)
 	),
 
 	TP_fast_assign(
-		__entry->timer	= timer;
+		__entry->hrtimer	= hrtimer;
 	),
 
-	TP_printk("hrtimer %p", __entry->timer)
+	TP_printk("hrtimer=%p", __entry->hrtimer)
 );
 
 /**
@@ -302,7 +301,7 @@ TRACE_EVENT(itimer_state,
 		__entry->interval_usec	= value->it_interval.tv_usec;
 	),
 
-	TP_printk("which %d, expires %lu, it_value %lu.%lu, it_interval %lu.%lu",
+	TP_printk("which=%d expires=%lu it_value=%lu.%lu it_interval=%lu.%lu",
 		  __entry->which, __entry->expires,
 		  __entry->value_sec, __entry->value_usec,
 		  __entry->interval_sec, __entry->interval_usec)
@@ -332,7 +331,7 @@ TRACE_EVENT(itimer_expire,
 		__entry->pid	= pid_nr(pid);
 	),
 
-	    TP_printk("which %d, pid %d, now %lu", __entry->which,
+	    TP_printk("which=%d pid=%d now=%lu", __entry->which,
 		      (int) __entry->pid, __entry->now)
 );
 

include/trace/ftrace.h

@@ -120,9 +120,10 @@
 #undef __field
 #define __field(type, item)					\
 	ret = trace_seq_printf(s, "\tfield:" #type " " #item ";\t"	\
-			       "offset:%u;\tsize:%u;\n",		\
+			       "offset:%u;\tsize:%u;\tsigned:%u;\n",	\
 			       (unsigned int)offsetof(typeof(field), item), \
-			       (unsigned int)sizeof(field.item));	\
+			       (unsigned int)sizeof(field.item),	\
+			       (unsigned int)is_signed_type(type));	\
 	if (!ret)							\
 		return 0;
 
@@ -132,19 +133,21 @@
 #undef __array
 #define __array(type, item, len)						\
 	ret = trace_seq_printf(s, "\tfield:" #type " " #item "[" #len "];\t"	\
-			       "offset:%u;\tsize:%u;\n",		\
+			       "offset:%u;\tsize:%u;\tsigned:%u;\n",	\
 			       (unsigned int)offsetof(typeof(field), item), \
-			       (unsigned int)sizeof(field.item));	\
+			       (unsigned int)sizeof(field.item),	\
+			       (unsigned int)is_signed_type(type));	\
 	if (!ret)							\
 		return 0;
 
 #undef __dynamic_array
 #define __dynamic_array(type, item, len)				       \
 	ret = trace_seq_printf(s, "\tfield:__data_loc " #type "[] " #item ";\t"\
-			       "offset:%u;\tsize:%u;\n",		       \
+			       "offset:%u;\tsize:%u;\tsigned:%u;\n",	       \
 			       (unsigned int)offsetof(typeof(field),	       \
 					__data_loc_##item),		       \
-			       (unsigned int)sizeof(field.__data_loc_##item)); \
+			       (unsigned int)sizeof(field.__data_loc_##item), \
+			       (unsigned int)is_signed_type(type));	\
 	if (!ret)							       \
 		return 0;
 

include/trace/syscall.h

@@ -33,7 +33,7 @@ struct syscall_metadata {
 };
 
 #ifdef CONFIG_FTRACE_SYSCALLS
-extern struct syscall_metadata *syscall_nr_to_meta(int nr);
+extern unsigned long arch_syscall_addr(int nr);
 extern int syscall_name_to_nr(char *name);
 void set_syscall_enter_id(int num, int id);
 void set_syscall_exit_id(int num, int id);

kernel/perf_event.c

@@ -28,6 +28,7 @@
 #include <linux/anon_inodes.h>
 #include <linux/kernel_stat.h>
 #include <linux/perf_event.h>
+#include <linux/ftrace_event.h>
 
 #include <asm/irq_regs.h>
 
@@ -1355,7 +1356,7 @@ static void perf_ctx_adjust_freq(struct perf_event_context *ctx)
 	u64 interrupts, freq;
 
 	spin_lock(&ctx->lock);
-	list_for_each_entry(event, &ctx->group_list, group_entry) {
+	list_for_each_entry_rcu(event, &ctx->event_list, event_entry) {
 		if (event->state != PERF_EVENT_STATE_ACTIVE)
 			continue;
 
@@ -1658,6 +1659,8 @@ static struct perf_event_context *find_get_context(pid_t pid, int cpu)
 	return ERR_PTR(err);
 }
 
+static void perf_event_free_filter(struct perf_event *event);
+
 static void free_event_rcu(struct rcu_head *head)
 {
 	struct perf_event *event;
@@ -1665,6 +1668,7 @@ static void free_event_rcu(struct rcu_head *head)
 	event = container_of(head, struct perf_event, rcu_head);
 	if (event->ns)
 		put_pid_ns(event->ns);
+	perf_event_free_filter(event);
 	kfree(event);
 }
 
@@ -1974,7 +1978,8 @@ static int perf_event_period(struct perf_event *event, u64 __user *arg)
 	return ret;
 }
 
-int perf_event_set_output(struct perf_event *event, int output_fd);
+static int perf_event_set_output(struct perf_event *event, int output_fd);
+static int perf_event_set_filter(struct perf_event *event, void __user *arg);
 
 static long perf_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 {
@@ -2002,6 +2007,9 @@ static long perf_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	case PERF_EVENT_IOC_SET_OUTPUT:
 		return perf_event_set_output(event, arg);
 
+	case PERF_EVENT_IOC_SET_FILTER:
+		return perf_event_set_filter(event, (void __user *)arg);
+
 	default:
 		return -ENOTTY;
 	}
@@ -3806,9 +3814,14 @@ static int perf_swevent_is_counting(struct perf_event *event)
 	return 1;
 }
 
+static int perf_tp_event_match(struct perf_event *event,
+				struct perf_sample_data *data);
+
 static int perf_swevent_match(struct perf_event *event,
 				enum perf_type_id type,
-				u32 event_id, struct pt_regs *regs)
+				u32 event_id,
+				struct perf_sample_data *data,
+				struct pt_regs *regs)
 {
 	if (!perf_swevent_is_counting(event))
 		return 0;
@@ -3826,6 +3839,10 @@ static int perf_swevent_match(struct perf_event *event,
 			return 0;
 	}
 
+	if (event->attr.type == PERF_TYPE_TRACEPOINT &&
+	    !perf_tp_event_match(event, data))
+		return 0;
+
 	return 1;
 }
 
@@ -3842,7 +3859,7 @@ static void perf_swevent_ctx_event(struct perf_event_context *ctx,
 
 	rcu_read_lock();
 	list_for_each_entry_rcu(event, &ctx->event_list, event_entry) {
-		if (perf_swevent_match(event, type, event_id, regs))
+		if (perf_swevent_match(event, type, event_id, data, regs))
 			perf_swevent_add(event, nr, nmi, data, regs);
 	}
 	rcu_read_unlock();
@@ -4086,6 +4103,7 @@ static const struct pmu perf_ops_task_clock = {
 };
 
 #ifdef CONFIG_EVENT_PROFILE
+
 void perf_tp_event(int event_id, u64 addr, u64 count, void *record,
 			  int entry_size)
 {
@@ -4109,8 +4127,15 @@ void perf_tp_event(int event_id, u64 addr, u64 count, void *record,
 }
 EXPORT_SYMBOL_GPL(perf_tp_event);
 
-extern int ftrace_profile_enable(int);
-extern void ftrace_profile_disable(int);
+static int perf_tp_event_match(struct perf_event *event,
+				struct perf_sample_data *data)
+{
+	void *record = data->raw->data;
+
+	if (likely(!event->filter) || filter_match_preds(event->filter, record))
+		return 1;
+	return 0;
+}
 
 static void tp_perf_event_destroy(struct perf_event *event)
 {
@@ -4135,12 +4160,53 @@ static const struct pmu *tp_perf_event_init(struct perf_event *event)
 
 	return &perf_ops_generic;
 }
+
+static int perf_event_set_filter(struct perf_event *event, void __user *arg)
+{
+	char *filter_str;
+	int ret;
+
+	if (event->attr.type != PERF_TYPE_TRACEPOINT)
+		return -EINVAL;
+
+	filter_str = strndup_user(arg, PAGE_SIZE);
+	if (IS_ERR(filter_str))
+		return PTR_ERR(filter_str);
+
+	ret = ftrace_profile_set_filter(event, event->attr.config, filter_str);
+
+	kfree(filter_str);
+	return ret;
+}
+
+static void perf_event_free_filter(struct perf_event *event)
+{
+	ftrace_profile_free_filter(event);
+}
+
 #else
+
+static int perf_tp_event_match(struct perf_event *event,
+				struct perf_sample_data *data)
+{
+	return 1;
+}
+
 static const struct pmu *tp_perf_event_init(struct perf_event *event)
 {
 	return NULL;
 }
-#endif
+
+static int perf_event_set_filter(struct perf_event *event, void __user *arg)
+{
+	return -ENOENT;
+}
+
+static void perf_event_free_filter(struct perf_event *event)
+{
+}
+
+#endif /* CONFIG_EVENT_PROFILE */
 
 atomic_t perf_swevent_enabled[PERF_COUNT_SW_MAX];
 
@@ -4394,7 +4460,7 @@ static int perf_copy_attr(struct perf_event_attr __user *uattr,
 	goto out;
 }
 
-int perf_event_set_output(struct perf_event *event, int output_fd)
+static int perf_event_set_output(struct perf_event *event, int output_fd)
 {
 	struct perf_event *output_event = NULL;
 	struct file *output_file = NULL;

kernel/trace/ftrace.c

@@ -60,6 +60,13 @@ static int last_ftrace_enabled;
 /* Quick disabling of function tracer. */
 int function_trace_stop;
 
+/* List for set_ftrace_pid's pids. */
+LIST_HEAD(ftrace_pids);
+struct ftrace_pid {
+	struct list_head list;
+	struct pid *pid;
+};
+
 /*
  * ftrace_disabled is set when an anomaly is discovered.
  * ftrace_disabled is much stronger than ftrace_enabled.
@@ -78,6 +85,10 @@ ftrace_func_t ftrace_trace_function __read_mostly = ftrace_stub;
 ftrace_func_t __ftrace_trace_function __read_mostly = ftrace_stub;
 ftrace_func_t ftrace_pid_function __read_mostly = ftrace_stub;
 
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+static int ftrace_set_func(unsigned long *array, int *idx, char *buffer);
+#endif
+
 static void ftrace_list_func(unsigned long ip, unsigned long parent_ip)
 {
 	struct ftrace_ops *op = ftrace_list;
@@ -155,7 +166,7 @@ static int __register_ftrace_function(struct ftrace_ops *ops)
 		else
 			func = ftrace_list_func;
 
-		if (ftrace_pid_trace) {
+		if (!list_empty(&ftrace_pids)) {
 			set_ftrace_pid_function(func);
 			func = ftrace_pid_func;
 		}
@@ -203,7 +214,7 @@ static int __unregister_ftrace_function(struct ftrace_ops *ops)
 		if (ftrace_list->next == &ftrace_list_end) {
 			ftrace_func_t func = ftrace_list->func;
 
-			if (ftrace_pid_trace) {
+			if (!list_empty(&ftrace_pids)) {
 				set_ftrace_pid_function(func);
 				func = ftrace_pid_func;
 			}
@@ -231,7 +242,7 @@ static void ftrace_update_pid_func(void)
 	func = __ftrace_trace_function;
 #endif
 
-	if (ftrace_pid_trace) {
+	if (!list_empty(&ftrace_pids)) {
 		set_ftrace_pid_function(func);
 		func = ftrace_pid_func;
 	} else {
@@ -821,8 +832,6 @@ static __init void ftrace_profile_debugfs(struct dentry *d_tracer)
 }
 #endif /* CONFIG_FUNCTION_PROFILER */
 
-/* set when tracing only a pid */
-struct pid *ftrace_pid_trace;
 static struct pid * const ftrace_swapper_pid = &init_struct_pid;
 
 #ifdef CONFIG_DYNAMIC_FTRACE
@@ -1261,12 +1270,34 @@ static int ftrace_update_code(struct module *mod)
 		ftrace_new_addrs = p->newlist;
 		p->flags = 0L;
 
-		/* convert record (i.e, patch mcount-call with NOP) */
-		if (ftrace_code_disable(mod, p)) {
-			p->flags |= FTRACE_FL_CONVERTED;
-			ftrace_update_cnt++;
-		} else
+		/*
+		 * Do the initial record convertion from mcount jump
+		 * to the NOP instructions.
+		 */
+		if (!ftrace_code_disable(mod, p)) {
 			ftrace_free_rec(p);
+			continue;
+		}
+
+		p->flags |= FTRACE_FL_CONVERTED;
+		ftrace_update_cnt++;
+
+		/*
+		 * If the tracing is enabled, go ahead and enable the record.
+		 *
+		 * The reason not to enable the record immediatelly is the
+		 * inherent check of ftrace_make_nop/ftrace_make_call for
+		 * correct previous instructions.  Making first the NOP
+		 * conversion puts the module to the correct state, thus
+		 * passing the ftrace_make_call check.
+		 */
+		if (ftrace_start_up) {
+			int failed = __ftrace_replace_code(p, 1);
+			if (failed) {
+				ftrace_bug(failed, p->ip);
+				ftrace_free_rec(p);
+			}
+		}
 	}
 
 	stop = ftrace_now(raw_smp_processor_id());
@@ -1656,60 +1687,6 @@ ftrace_regex_lseek(struct file *file, loff_t offset, int origin)
 	return ret;
 }
 
-enum {
-	MATCH_FULL,
-	MATCH_FRONT_ONLY,
-	MATCH_MIDDLE_ONLY,
-	MATCH_END_ONLY,
-};
-
-/*
- * (static function - no need for kernel doc)
- *
- * Pass in a buffer containing a glob and this function will
- * set search to point to the search part of the buffer and
- * return the type of search it is (see enum above).
- * This does modify buff.
- *
- * Returns enum type.
- *  search returns the pointer to use for comparison.
- *  not returns 1 if buff started with a '!'
- *     0 otherwise.
- */
-static int
-ftrace_setup_glob(char *buff, int len, char **search, int *not)
-{
-	int type = MATCH_FULL;
-	int i;
-
-	if (buff[0] == '!') {
-		*not = 1;
-		buff++;
-		len--;
-	} else
-		*not = 0;
-
-	*search = buff;
-
-	for (i = 0; i < len; i++) {
-		if (buff[i] == '*') {
-			if (!i) {
-				*search = buff + 1;
-				type = MATCH_END_ONLY;
-			} else {
-				if (type == MATCH_END_ONLY)
-					type = MATCH_MIDDLE_ONLY;
-				else
-					type = MATCH_FRONT_ONLY;
-				buff[i] = 0;
-				break;
-			}
-		}
-	}
-
-	return type;
-}
-
 static int ftrace_match(char *str, char *regex, int len, int type)
 {
 	int matched = 0;
@@ -1758,7 +1735,7 @@ static void ftrace_match_records(char *buff, int len, int enable)
 	int not;
 
 	flag = enable ? FTRACE_FL_FILTER : FTRACE_FL_NOTRACE;
-	type = ftrace_setup_glob(buff, len, &search, &not);
+	type = filter_parse_regex(buff, len, &search, &not);
 
 	search_len = strlen(search);
 
@@ -1826,7 +1803,7 @@ static void ftrace_match_module_records(char *buff, char *mod, int enable)
 	}
 
 	if (strlen(buff)) {
-		type = ftrace_setup_glob(buff, strlen(buff), &search, &not);
+		type = filter_parse_regex(buff, strlen(buff), &search, &not);
 		search_len = strlen(search);
 	}
 
@@ -1991,7 +1968,7 @@ register_ftrace_function_probe(char *glob, struct ftrace_probe_ops *ops,
 	int count = 0;
 	char *search;
 
-	type = ftrace_setup_glob(glob, strlen(glob), &search, &not);
+	type = filter_parse_regex(glob, strlen(glob), &search, &not);
 	len = strlen(search);
 
 	/* we do not support '!' for function probes */
@@ -2068,7 +2045,7 @@ __unregister_ftrace_function_probe(char *glob, struct ftrace_probe_ops *ops,
 	else if (glob) {
 		int not;
 
-		type = ftrace_setup_glob(glob, strlen(glob), &search, &not);
+		type = filter_parse_regex(glob, strlen(glob), &search, &not);
 		len = strlen(search);
 
 		/* we do not support '!' for function probes */
@@ -2297,6 +2274,7 @@ void ftrace_set_notrace(unsigned char *buf, int len, int reset)
 #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 char ftrace_graph_buf[FTRACE_FILTER_SIZE] __initdata;
 
 static int __init set_ftrace_notrace(char *str)
 {
@@ -2312,6 +2290,31 @@ static int __init set_ftrace_filter(char *str)
 }
 __setup("ftrace_filter=", set_ftrace_filter);
 
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+static int __init set_graph_function(char *str)
+{
+	strlcpy(ftrace_graph_buf, str, FTRACE_FILTER_SIZE);
+	return 1;
+}
+__setup("ftrace_graph_filter=", set_graph_function);
+
+static void __init set_ftrace_early_graph(char *buf)
+{
+	int ret;
+	char *func;
+
+	while (buf) {
+		func = strsep(&buf, ",");
+		/* we allow only one expression at a time */
+		ret = ftrace_set_func(ftrace_graph_funcs, &ftrace_graph_count,
+				      func);
+		if (ret)
+			printk(KERN_DEBUG "ftrace: function %s not "
+					  "traceable\n", func);
+	}
+}
+#endif /* CONFIG_FUNCTION_GRAPH_TRACER */
+
 static void __init set_ftrace_early_filter(char *buf, int enable)
 {
 	char *func;
@@ -2328,6 +2331,10 @@ static void __init set_ftrace_early_filters(void)
 		set_ftrace_early_filter(ftrace_filter_buf, 1);
 	if (ftrace_notrace_buf[0])
 		set_ftrace_early_filter(ftrace_notrace_buf, 0);
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+	if (ftrace_graph_buf[0])
+		set_ftrace_early_graph(ftrace_graph_buf);
+#endif /* CONFIG_FUNCTION_GRAPH_TRACER */
 }
 
 static int
@@ -2513,7 +2520,7 @@ ftrace_set_func(unsigned long *array, int *idx, char *buffer)
 		return -ENODEV;
 
 	/* decode regex */
-	type = ftrace_setup_glob(buffer, strlen(buffer), &search, &not);
+	type = filter_parse_regex(buffer, strlen(buffer), &search, &not);
 	if (not)
 		return -EINVAL;
 
@@ -2624,7 +2631,7 @@ static __init int ftrace_init_dyn_debugfs(struct dentry *d_tracer)
 	return 0;
 }
 
-static int ftrace_convert_nops(struct module *mod,
+static int ftrace_process_locs(struct module *mod,
 			       unsigned long *start,
 			       unsigned long *end)
 {
@@ -2684,7 +2691,7 @@ static void ftrace_init_module(struct module *mod,
 {
 	if (ftrace_disabled || start == end)
 		return;
-	ftrace_convert_nops(mod, start, end);
+	ftrace_process_locs(mod, start, end);
 }
 
 static int ftrace_module_notify(struct notifier_block *self,
@@ -2745,7 +2752,7 @@ void __init ftrace_init(void)
 
 	last_ftrace_enabled = ftrace_enabled = 1;
 
-	ret = ftrace_convert_nops(NULL,
+	ret = ftrace_process_locs(NULL,
 				  __start_mcount_loc,
 				  __stop_mcount_loc);
 
@@ -2778,23 +2785,6 @@ static inline void ftrace_startup_enable(int command) { }
 # define ftrace_shutdown_sysctl()	do { } while (0)
 #endif /* CONFIG_DYNAMIC_FTRACE */
 
-static ssize_t
-ftrace_pid_read(struct file *file, char __user *ubuf,
-		       size_t cnt, loff_t *ppos)
-{
-	char buf[64];
-	int r;
-
-	if (ftrace_pid_trace == ftrace_swapper_pid)
-		r = sprintf(buf, "swapper tasks\n");
-	else if (ftrace_pid_trace)
-		r = sprintf(buf, "%u\n", pid_vnr(ftrace_pid_trace));
-	else
-		r = sprintf(buf, "no pid\n");
-
-	return simple_read_from_buffer(ubuf, cnt, ppos, buf, r);
-}
-
 static void clear_ftrace_swapper(void)
 {
 	struct task_struct *p;
@@ -2845,14 +2835,12 @@ static void set_ftrace_pid(struct pid *pid)
 	rcu_read_unlock();
 }
 
-static void clear_ftrace_pid_task(struct pid **pid)
+static void clear_ftrace_pid_task(struct pid *pid)
 {
-	if (*pid == ftrace_swapper_pid)
+	if (pid == ftrace_swapper_pid)
 		clear_ftrace_swapper();
 	else
-		clear_ftrace_pid(*pid);
-
-	*pid = NULL;
+		clear_ftrace_pid(pid);
 }
 
 static void set_ftrace_pid_task(struct pid *pid)
@@ -2863,11 +2851,140 @@ static void set_ftrace_pid_task(struct pid *pid)
 		set_ftrace_pid(pid);
 }
 
+static int ftrace_pid_add(int p)
+{
+	struct pid *pid;
+	struct ftrace_pid *fpid;
+	int ret = -EINVAL;
+
+	mutex_lock(&ftrace_lock);
+
+	if (!p)
+		pid = ftrace_swapper_pid;
+	else
+		pid = find_get_pid(p);
+
+	if (!pid)
+		goto out;
+
+	ret = 0;
+
+	list_for_each_entry(fpid, &ftrace_pids, list)
+		if (fpid->pid == pid)
+			goto out_put;
+
+	ret = -ENOMEM;
+
+	fpid = kmalloc(sizeof(*fpid), GFP_KERNEL);
+	if (!fpid)
+		goto out_put;
+
+	list_add(&fpid->list, &ftrace_pids);
+	fpid->pid = pid;
+
+	set_ftrace_pid_task(pid);
+
+	ftrace_update_pid_func();
+	ftrace_startup_enable(0);
+
+	mutex_unlock(&ftrace_lock);
+	return 0;
+
+out_put:
+	if (pid != ftrace_swapper_pid)
+		put_pid(pid);
+
+out:
+	mutex_unlock(&ftrace_lock);
+	return ret;
+}
+
+static void ftrace_pid_reset(void)
+{
+	struct ftrace_pid *fpid, *safe;
+
+	mutex_lock(&ftrace_lock);
+	list_for_each_entry_safe(fpid, safe, &ftrace_pids, list) {
+		struct pid *pid = fpid->pid;
+
+		clear_ftrace_pid_task(pid);
+
+		list_del(&fpid->list);
+		kfree(fpid);
+	}
+
+	ftrace_update_pid_func();
+	ftrace_startup_enable(0);
+
+	mutex_unlock(&ftrace_lock);
+}
+
+static void *fpid_start(struct seq_file *m, loff_t *pos)
+{
+	mutex_lock(&ftrace_lock);
+
+	if (list_empty(&ftrace_pids) && (!*pos))
+		return (void *) 1;
+
+	return seq_list_start(&ftrace_pids, *pos);
+}
+
+static void *fpid_next(struct seq_file *m, void *v, loff_t *pos)
+{
+	if (v == (void *)1)
+		return NULL;
+
+	return seq_list_next(v, &ftrace_pids, pos);
+}
+
+static void fpid_stop(struct seq_file *m, void *p)
+{
+	mutex_unlock(&ftrace_lock);
+}
+
+static int fpid_show(struct seq_file *m, void *v)
+{
+	const struct ftrace_pid *fpid = list_entry(v, struct ftrace_pid, list);
+
+	if (v == (void *)1) {
+		seq_printf(m, "no pid\n");
+		return 0;
+	}
+
+	if (fpid->pid == ftrace_swapper_pid)
+		seq_printf(m, "swapper tasks\n");
+	else
+		seq_printf(m, "%u\n", pid_vnr(fpid->pid));
+
+	return 0;
+}
+
+static const struct seq_operations ftrace_pid_sops = {
+	.start = fpid_start,
+	.next = fpid_next,
+	.stop = fpid_stop,
+	.show = fpid_show,
+};
+
+static int
+ftrace_pid_open(struct inode *inode, struct file *file)
+{
+	int ret = 0;
+
+	if ((file->f_mode & FMODE_WRITE) &&
+	    (file->f_flags & O_TRUNC))
+		ftrace_pid_reset();
+
+	if (file->f_mode & FMODE_READ)
+		ret = seq_open(file, &ftrace_pid_sops);
+
+	return ret;
+}
+
 static ssize_t
 ftrace_pid_write(struct file *filp, const char __user *ubuf,
 		   size_t cnt, loff_t *ppos)
 {
-	struct pid *pid;
 	char buf[64];
 	long val;
 	int ret;
@@ -2880,57 +2997,38 @@ ftrace_pid_write(struct file *filp, const char __user *ubuf,
 
 	buf[cnt] = 0;
 
+	/*
+	 * Allow "echo > set_ftrace_pid" or "echo -n '' > set_ftrace_pid"
+	 * to clean the filter quietly.
+	 */
+	strstrip(buf);
+	if (strlen(buf) == 0)
+		return 1;
+
 	ret = strict_strtol(buf, 10, &val);
 	if (ret < 0)
 		return ret;
 
-	mutex_lock(&ftrace_lock);
-	if (val < 0) {
-		/* disable pid tracing */
-		if (!ftrace_pid_trace)
-			goto out;
-
-		clear_ftrace_pid_task(&ftrace_pid_trace);
-
-	} else {
-		/* swapper task is special */
-		if (!val) {
-			pid = ftrace_swapper_pid;
-			if (pid == ftrace_pid_trace)
-				goto out;
-		} else {
-			pid = find_get_pid(val);
+	ret = ftrace_pid_add(val);
 
-			if (pid == ftrace_pid_trace) {
-				put_pid(pid);
-				goto out;
-			}
-		}
-
-		if (ftrace_pid_trace)
-			clear_ftrace_pid_task(&ftrace_pid_trace);
-
-		if (!pid)
-			goto out;
-
-		ftrace_pid_trace = pid;
-
-		set_ftrace_pid_task(ftrace_pid_trace);
-	}
-
-	/* update the function call */
-	ftrace_update_pid_func();
-	ftrace_startup_enable(0);
+	return ret ? ret : cnt;
+}
 
- out:
-	mutex_unlock(&ftrace_lock);
+static int
+ftrace_pid_release(struct inode *inode, struct file *file)
+{
+	if (file->f_mode & FMODE_READ)
+		seq_release(inode, file);
 
-	return cnt;
+	return 0;
 }
 
 static const struct file_operations ftrace_pid_fops = {
-	.read = ftrace_pid_read,
-	.write = ftrace_pid_write,
+	.open		= ftrace_pid_open,
+	.write		= ftrace_pid_write,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= ftrace_pid_release,
 };
 
 static __init int ftrace_init_debugfs(void)

kernel/trace/ring_buffer.c

@@ -397,18 +397,21 @@ int ring_buffer_print_page_header(struct trace_seq *s)
 	int ret;
 
 	ret = trace_seq_printf(s, "\tfield: u64 timestamp;\t"
-			       "offset:0;\tsize:%u;\n",
-			       (unsigned int)sizeof(field.time_stamp));
+			       "offset:0;\tsize:%u;\tsigned:%u;\n",
+			       (unsigned int)sizeof(field.time_stamp),
+			       (unsigned int)is_signed_type(u64));
 
 	ret = trace_seq_printf(s, "\tfield: local_t commit;\t"
-			       "offset:%u;\tsize:%u;\n",
+			       "offset:%u;\tsize:%u;\tsigned:%u;\n",
 			       (unsigned int)offsetof(typeof(field), commit),
-			       (unsigned int)sizeof(field.commit));
+			       (unsigned int)sizeof(field.commit),
+			       (unsigned int)is_signed_type(long));
 
 	ret = trace_seq_printf(s, "\tfield: char data;\t"
-			       "offset:%u;\tsize:%u;\n",
+			       "offset:%u;\tsize:%u;\tsigned:%u;\n",
 			       (unsigned int)offsetof(typeof(field), data),
-			       (unsigned int)BUF_PAGE_SIZE);
+			       (unsigned int)BUF_PAGE_SIZE,
+			       (unsigned int)is_signed_type(char));
 
 	return ret;
 }

kernel/trace/trace.c

@@ -129,7 +129,7 @@ static int tracing_set_tracer(const char *buf);
 static char bootup_tracer_buf[MAX_TRACER_SIZE] __initdata;
 static char *default_bootup_tracer;
 
-static int __init set_ftrace(char *str)
+static int __init set_cmdline_ftrace(char *str)
 {
 	strncpy(bootup_tracer_buf, str, MAX_TRACER_SIZE);
 	default_bootup_tracer = bootup_tracer_buf;
@@ -137,7 +137,7 @@ static int __init set_ftrace(char *str)
 	ring_buffer_expanded = 1;
 	return 1;
 }
-__setup("ftrace=", set_ftrace);
+__setup("ftrace=", set_cmdline_ftrace);
 
 static int __init set_ftrace_dump_on_oops(char *str)
 {

kernel/trace/trace.h

@@ -506,10 +506,6 @@ static inline int ftrace_graph_addr(unsigned long addr)
 	return 0;
 }
 #else
-static inline int ftrace_trace_addr(unsigned long addr)
-{
-	return 1;
-}
 static inline int ftrace_graph_addr(unsigned long addr)
 {
 	return 1;
@@ -523,12 +519,12 @@ print_graph_function(struct trace_iterator *iter)
 }
 #endif /* CONFIG_FUNCTION_GRAPH_TRACER */
 
-extern struct pid *ftrace_pid_trace;
+extern struct list_head ftrace_pids;
 
 #ifdef CONFIG_FUNCTION_TRACER
 static inline int ftrace_trace_task(struct task_struct *task)
 {
-	if (!ftrace_pid_trace)
+	if (list_empty(&ftrace_pids))
 		return 1;
 
 	return test_tsk_trace_trace(task);
@@ -710,7 +706,6 @@ struct event_filter {
 	int			n_preds;
 	struct filter_pred	**preds;
 	char			*filter_string;
-	bool			no_reset;
 };
 
 struct event_subsystem {
@@ -722,22 +717,40 @@ struct event_subsystem {
 };
 
 struct filter_pred;
+struct regex;
 
 typedef int (*filter_pred_fn_t) (struct filter_pred *pred, void *event,
 				 int val1, int val2);
 
+typedef int (*regex_match_func)(char *str, struct regex *r, int len);
+
+enum regex_type {
+	MATCH_FULL = 0,
+	MATCH_FRONT_ONLY,
+	MATCH_MIDDLE_ONLY,
+	MATCH_END_ONLY,
+};
+
+struct regex {
+	char			pattern[MAX_FILTER_STR_VAL];
+	int			len;
+	int			field_len;
+	regex_match_func	match;
+};
+
 struct filter_pred {
-	filter_pred_fn_t fn;
-	u64 val;
-	char str_val[MAX_FILTER_STR_VAL];
-	int str_len;
-	char *field_name;
-	int offset;
-	int not;
-	int op;
-	int pop_n;
+	filter_pred_fn_t 	fn;
+	u64 			val;
+	struct regex		regex;
+	char 			*field_name;
+	int 			offset;
+	int 			not;
+	int 			op;
+	int 			pop_n;
 };
 
+extern enum regex_type
+filter_parse_regex(char *buff, int len, char **search, int *not);
 extern void print_event_filter(struct ftrace_event_call *call,
 			       struct trace_seq *s);
 extern int apply_event_filter(struct ftrace_event_call *call,
@@ -753,7 +766,8 @@ filter_check_discard(struct ftrace_event_call *call, void *rec,
 		     struct ring_buffer *buffer,
 		     struct ring_buffer_event *event)
 {
-	if (unlikely(call->filter_active) && !filter_match_preds(call, rec)) {
+	if (unlikely(call->filter_active) &&
+	    !filter_match_preds(call->filter, rec)) {
 		ring_buffer_discard_commit(buffer, event);
 		return 1;
 	}

kernel/trace/trace_events.c

@@ -503,7 +503,7 @@ extern char *__bad_type_size(void);
 #define FIELD(type, name)						\
 	sizeof(type) != sizeof(field.name) ? __bad_type_size() :	\
 	#type, "common_" #name, offsetof(typeof(field), name),		\
-		sizeof(field.name)
+		sizeof(field.name), is_signed_type(type)
 
 static int trace_write_header(struct trace_seq *s)
 {
@@ -511,17 +511,17 @@ static int trace_write_header(struct trace_seq *s)
 
 	/* struct trace_entry */
 	return trace_seq_printf(s,
-				"\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n"
-				"\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n"
-				"\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n"
-				"\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n"
-				"\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n"
-				"\n",
-				FIELD(unsigned short, type),
-				FIELD(unsigned char, flags),
-				FIELD(unsigned char, preempt_count),
-				FIELD(int, pid),
-				FIELD(int, lock_depth));
+			"\tfield:%s %s;\toffset:%zu;\tsize:%zu;\tsigned:%u;\n"
+			"\tfield:%s %s;\toffset:%zu;\tsize:%zu;\tsigned:%u;\n"
+			"\tfield:%s %s;\toffset:%zu;\tsize:%zu;\tsigned:%u;\n"
+			"\tfield:%s %s;\toffset:%zu;\tsize:%zu;\tsigned:%u;\n"
+			"\tfield:%s %s;\toffset:%zu;\tsize:%zu;\tsigned:%u;\n"
+			"\n",
+			FIELD(unsigned short, type),
+			FIELD(unsigned char, flags),
+			FIELD(unsigned char, preempt_count),
+			FIELD(int, pid),
+			FIELD(int, lock_depth));
 }
 
 static ssize_t
@@ -874,9 +874,9 @@ event_subsystem_dir(const char *name, struct dentry *d_events)
 			   "'%s/filter' entry\n", name);
 	}
 
-	entry = trace_create_file("enable", 0644, system->entry,
-				  (void *)system->name,
-				  &ftrace_system_enable_fops);
+	trace_create_file("enable", 0644, system->entry,
+			  (void *)system->name,
+			  &ftrace_system_enable_fops);
 
 	return system->entry;
 }
@@ -888,7 +888,6 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events,
 		 const struct file_operations *filter,
 		 const struct file_operations *format)
 {
-	struct dentry *entry;
 	int ret;
 
 	/*
@@ -906,12 +905,12 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events,
 	}
 
 	if (call->regfunc)
-		entry = trace_create_file("enable", 0644, call->dir, call,
-					  enable);
+		trace_create_file("enable", 0644, call->dir, call,
+				  enable);
 
 	if (call->id && call->profile_enable)
-		entry = trace_create_file("id", 0444, call->dir, call,
-					  id);
+		trace_create_file("id", 0444, call->dir, call,
+		 		  id);
 
 	if (call->define_fields) {
 		ret = call->define_fields(call);
@@ -920,16 +919,16 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events,
 				   " events/%s\n", call->name);
 			return ret;
 		}
-		entry = trace_create_file("filter", 0644, call->dir, call,
-					  filter);
+		trace_create_file("filter", 0644, call->dir, call,
+				  filter);
 	}
 
 	/* A trace may not want to export its format */
 	if (!call->show_format)
 		return 0;
 
-	entry = trace_create_file("format", 0444, call->dir, call,
-				  format);
+	trace_create_file("format", 0444, call->dir, call,
+			  format);
 
 	return 0;
 }

kernel/trace/trace_events_filter.c

@@ -18,11 +18,10 @@
  * Copyright (C) 2009 Tom Zanussi <tzanussi@gmail.com>
  */
 
-#include <linux/debugfs.h>
-#include <linux/uaccess.h>
 #include <linux/module.h>
 #include <linux/ctype.h>
 #include <linux/mutex.h>
+#include <linux/perf_event.h>
 
 #include "trace.h"
 #include "trace_output.h"
@@ -31,6 +30,7 @@ enum filter_op_ids
 {
 	OP_OR,
 	OP_AND,
+	OP_GLOB,
 	OP_NE,
 	OP_EQ,
 	OP_LT,
@@ -48,16 +48,17 @@ struct filter_op {
 };
 
 static struct filter_op filter_ops[] = {
-	{ OP_OR, "||", 1 },
-	{ OP_AND, "&&", 2 },
-	{ OP_NE, "!=", 4 },
-	{ OP_EQ, "==", 4 },
-	{ OP_LT, "<", 5 },
-	{ OP_LE, "<=", 5 },
-	{ OP_GT, ">", 5 },
-	{ OP_GE, ">=", 5 },
-	{ OP_NONE, "OP_NONE", 0 },
-	{ OP_OPEN_PAREN, "(", 0 },
+	{ OP_OR,	"||",		1 },
+	{ OP_AND,	"&&",		2 },
+	{ OP_GLOB,	"~",		4 },
+	{ OP_NE,	"!=",		4 },
+	{ OP_EQ,	"==",		4 },
+	{ OP_LT,	"<",		5 },
+	{ OP_LE,	"<=",		5 },
+	{ OP_GT,	">",		5 },
+	{ OP_GE,	">=",		5 },
+	{ OP_NONE,	"OP_NONE",	0 },
+	{ OP_OPEN_PAREN, "(",		0 },
 };
 
 enum {
@@ -197,9 +198,9 @@ static int filter_pred_string(struct filter_pred *pred, void *event,
 	char *addr = (char *)(event + pred->offset);
 	int cmp, match;
 
-	cmp = strncmp(addr, pred->str_val, pred->str_len);
+	cmp = pred->regex.match(addr, &pred->regex, pred->regex.field_len);
 
-	match = (!cmp) ^ pred->not;
+	match = cmp ^ pred->not;
 
 	return match;
 }
@@ -211,9 +212,9 @@ static int filter_pred_pchar(struct filter_pred *pred, void *event,
 	char **addr = (char **)(event + pred->offset);
 	int cmp, match;
 
-	cmp = strncmp(*addr, pred->str_val, pred->str_len);
+	cmp = pred->regex.match(*addr, &pred->regex, pred->regex.field_len);
 
-	match = (!cmp) ^ pred->not;
+	match = cmp ^ pred->not;
 
 	return match;
 }
@@ -237,9 +238,9 @@ static int filter_pred_strloc(struct filter_pred *pred, void *event,
 	char *addr = (char *)(event + str_loc);
 	int cmp, match;
 
-	cmp = strncmp(addr, pred->str_val, str_len);
+	cmp = pred->regex.match(addr, &pred->regex, str_len);
 
-	match = (!cmp) ^ pred->not;
+	match = cmp ^ pred->not;
 
 	return match;
 }
@@ -250,10 +251,121 @@ static int filter_pred_none(struct filter_pred *pred, void *event,
 	return 0;
 }
 
+/* Basic regex callbacks */
+static int regex_match_full(char *str, struct regex *r, int len)
+{
+	if (strncmp(str, r->pattern, len) == 0)
+		return 1;
+	return 0;
+}
+
+static int regex_match_front(char *str, struct regex *r, int len)
+{
+	if (strncmp(str, r->pattern, len) == 0)
+		return 1;
+	return 0;
+}
+
+static int regex_match_middle(char *str, struct regex *r, int len)
+{
+	if (strstr(str, r->pattern))
+		return 1;
+	return 0;
+}
+
+static int regex_match_end(char *str, struct regex *r, int len)
+{
+	char *ptr = strstr(str, r->pattern);
+
+	if (ptr && (ptr[r->len] == 0))
+		return 1;
+	return 0;
+}
+
+/**
+ * filter_parse_regex - parse a basic regex
+ * @buff:   the raw regex
+ * @len:    length of the regex
+ * @search: will point to the beginning of the string to compare
+ * @not:    tell whether the match will have to be inverted
+ *
+ * This passes in a buffer containing a regex and this function will
+ * set search to point to the search part of the buffer and
+ * return the type of search it is (see enum above).
+ * This does modify buff.
+ *
+ * Returns enum type.
+ *  search returns the pointer to use for comparison.
+ *  not returns 1 if buff started with a '!'
+ *     0 otherwise.
+ */
+enum regex_type filter_parse_regex(char *buff, int len, char **search, int *not)
+{
+	int type = MATCH_FULL;
+	int i;
+
+	if (buff[0] == '!') {
+		*not = 1;
+		buff++;
+		len--;
+	} else
+		*not = 0;
+
+	*search = buff;
+
+	for (i = 0; i < len; i++) {
+		if (buff[i] == '*') {
+			if (!i) {
+				*search = buff + 1;
+				type = MATCH_END_ONLY;
+			} else {
+				if (type == MATCH_END_ONLY)
+					type = MATCH_MIDDLE_ONLY;
+				else
+					type = MATCH_FRONT_ONLY;
+				buff[i] = 0;
+				break;
+			}
+		}
+	}
+
+	return type;
+}
+
+static void filter_build_regex(struct filter_pred *pred)
+{
+	struct regex *r = &pred->regex;
+	char *search;
+	enum regex_type type = MATCH_FULL;
+	int not = 0;
+
+	if (pred->op == OP_GLOB) {
+		type = filter_parse_regex(r->pattern, r->len, &search, &not);
+		r->len = strlen(search);
+		memmove(r->pattern, search, r->len+1);
+	}
+
+	switch (type) {
+	case MATCH_FULL:
+		r->match = regex_match_full;
+		break;
+	case MATCH_FRONT_ONLY:
+		r->match = regex_match_front;
+		break;
+	case MATCH_MIDDLE_ONLY:
+		r->match = regex_match_middle;
+		break;
+	case MATCH_END_ONLY:
+		r->match = regex_match_end;
+		break;
+	}
+
+	pred->not ^= not;
+}
+
 /* return 1 if event matches, 0 otherwise (discard) */
-int filter_match_preds(struct ftrace_event_call *call, void *rec)
+int filter_match_preds(struct event_filter *filter, void *rec)
 {
-	struct event_filter *filter = call->filter;
 	int match, top = 0, val1 = 0, val2 = 0;
 	int stack[MAX_FILTER_PRED];
 	struct filter_pred *pred;
@@ -396,7 +508,7 @@ static void filter_clear_pred(struct filter_pred *pred)
 {
 	kfree(pred->field_name);
 	pred->field_name = NULL;
-	pred->str_len = 0;
+	pred->regex.len = 0;
 }
 
 static int filter_set_pred(struct filter_pred *dest,
@@ -426,9 +538,8 @@ static void filter_disable_preds(struct ftrace_event_call *call)
 		filter->preds[i]->fn = filter_pred_none;
 }
 
-void destroy_preds(struct ftrace_event_call *call)
+static void __free_preds(struct event_filter *filter)
 {
-	struct event_filter *filter = call->filter;
 	int i;
 
 	if (!filter)
@@ -441,21 +552,24 @@ void destroy_preds(struct ftrace_event_call *call)
 	kfree(filter->preds);
 	kfree(filter->filter_string);
 	kfree(filter);
+}
+
+void destroy_preds(struct ftrace_event_call *call)
+{
+	__free_preds(call->filter);
 	call->filter = NULL;
+	call->filter_active = 0;
 }
 
-static int init_preds(struct ftrace_event_call *call)
+static struct event_filter *__alloc_preds(void)
 {
 	struct event_filter *filter;
 	struct filter_pred *pred;
 	int i;
 
-	if (call->filter)
-		return 0;
-
-	filter = call->filter = kzalloc(sizeof(*filter), GFP_KERNEL);
-	if (!call->filter)
-		return -ENOMEM;
+	filter = kzalloc(sizeof(*filter), GFP_KERNEL);
+	if (!filter)
+		return ERR_PTR(-ENOMEM);
 
 	filter->n_preds = 0;
 
@@ -471,12 +585,24 @@ static int init_preds(struct ftrace_event_call *call)
 		filter->preds[i] = pred;
 	}
 
-	return 0;
+	return filter;
 
 oom:
-	destroy_preds(call);
+	__free_preds(filter);
+	return ERR_PTR(-ENOMEM);
+}
+
+static int init_preds(struct ftrace_event_call *call)
+{
+	if (call->filter)
+		return 0;
 
-	return -ENOMEM;
+	call->filter_active = 0;
+	call->filter = __alloc_preds();
+	if (IS_ERR(call->filter))
+		return PTR_ERR(call->filter);
+
+	return 0;
 }
 
 static int init_subsystem_preds(struct event_subsystem *system)
@@ -499,14 +625,7 @@ static int init_subsystem_preds(struct event_subsystem *system)
 	return 0;
 }
 
-enum {
-	FILTER_DISABLE_ALL,
-	FILTER_INIT_NO_RESET,
-	FILTER_SKIP_NO_RESET,
-};
-
-static void filter_free_subsystem_preds(struct event_subsystem *system,
-					int flag)
+static void filter_free_subsystem_preds(struct event_subsystem *system)
 {
 	struct ftrace_event_call *call;
 
@@ -517,14 +636,6 @@ static void filter_free_subsystem_preds(struct event_subsystem *system,
 		if (strcmp(call->system, system->name) != 0)
 			continue;
 
-		if (flag == FILTER_INIT_NO_RESET) {
-			call->filter->no_reset = false;
-			continue;
-		}
-
-		if (flag == FILTER_SKIP_NO_RESET && call->filter->no_reset)
-			continue;
-
 		filter_disable_preds(call);
 		remove_filter_string(call->filter);
 	}
@@ -532,10 +643,10 @@ static void filter_free_subsystem_preds(struct event_subsystem *system,
 
 static int filter_add_pred_fn(struct filter_parse_state *ps,
 			      struct ftrace_event_call *call,
+			      struct event_filter *filter,
 			      struct filter_pred *pred,
 			      filter_pred_fn_t fn)
 {
-	struct event_filter *filter = call->filter;
 	int idx, err;
 
 	if (filter->n_preds == MAX_FILTER_PRED) {
@@ -550,7 +661,6 @@ static int filter_add_pred_fn(struct filter_parse_state *ps,
 		return err;
 
 	filter->n_preds++;
-	call->filter_active = 1;
 
 	return 0;
 }
@@ -575,7 +685,10 @@ static bool is_string_field(struct ftrace_event_field *field)
 
 static int is_legal_op(struct ftrace_event_field *field, int op)
 {
-	if (is_string_field(field) && (op != OP_EQ && op != OP_NE))
+	if (is_string_field(field) &&
+	    (op != OP_EQ && op != OP_NE && op != OP_GLOB))
+		return 0;
+	if (!is_string_field(field) && op == OP_GLOB)
 		return 0;
 
 	return 1;
@@ -626,6 +739,7 @@ static filter_pred_fn_t select_comparison_fn(int op, int field_size,
 
 static int filter_add_pred(struct filter_parse_state *ps,
 			   struct ftrace_event_call *call,
+			   struct event_filter *filter,
 			   struct filter_pred *pred,
 			   bool dry_run)
 {
@@ -660,21 +774,22 @@ static int filter_add_pred(struct filter_parse_state *ps,
 	}
 
 	if (is_string_field(field)) {
-		pred->str_len = field->size;
+		filter_build_regex(pred);
 
-		if (field->filter_type == FILTER_STATIC_STRING)
+		if (field->filter_type == FILTER_STATIC_STRING) {
 			fn = filter_pred_string;
-		else if (field->filter_type == FILTER_DYN_STRING)
+			pred->regex.field_len = field->size;
+		} else if (field->filter_type == FILTER_DYN_STRING)
 			fn = filter_pred_strloc;
 		else {
 			fn = filter_pred_pchar;
-			pred->str_len = strlen(pred->str_val);
+			pred->regex.field_len = strlen(pred->regex.pattern);
 		}
 	} else {
 		if (field->is_signed)
-			ret = strict_strtoll(pred->str_val, 0, &val);
+			ret = strict_strtoll(pred->regex.pattern, 0, &val);
 		else
-			ret = strict_strtoull(pred->str_val, 0, &val);
+			ret = strict_strtoull(pred->regex.pattern, 0, &val);
 		if (ret) {
 			parse_error(ps, FILT_ERR_ILLEGAL_INTVAL, 0);
 			return -EINVAL;
@@ -694,45 +809,7 @@ static int filter_add_pred(struct filter_parse_state *ps,
 
 add_pred_fn:
 	if (!dry_run)
-		return filter_add_pred_fn(ps, call, pred, fn);
-	return 0;
-}
-
-static int filter_add_subsystem_pred(struct filter_parse_state *ps,
-				     struct event_subsystem *system,
-				     struct filter_pred *pred,
-				     char *filter_string,
-				     bool dry_run)
-{
-	struct ftrace_event_call *call;
-	int err = 0;
-	bool fail = true;
-
-	list_for_each_entry(call, &ftrace_events, list) {
-
-		if (!call->define_fields)
-			continue;
-
-		if (strcmp(call->system, system->name))
-			continue;
-
-		if (call->filter->no_reset)
-			continue;
-
-		err = filter_add_pred(ps, call, pred, dry_run);
-		if (err)
-			call->filter->no_reset = true;
-		else
-			fail = false;
-
-		if (!dry_run)
-			replace_filter_string(call->filter, filter_string);
-	}
-
-	if (fail) {
-		parse_error(ps, FILT_ERR_BAD_SUBSYS_FILTER, 0);
-		return err;
-	}
+		return filter_add_pred_fn(ps, call, filter, pred, fn);
 	return 0;
 }
 
@@ -1045,8 +1122,8 @@ static struct filter_pred *create_pred(int op, char *operand1, char *operand2)
 		return NULL;
 	}
 
-	strcpy(pred->str_val, operand2);
-	pred->str_len = strlen(operand2);
+	strcpy(pred->regex.pattern, operand2);
+	pred->regex.len = strlen(pred->regex.pattern);
 
 	pred->op = op;
 
@@ -1090,8 +1167,8 @@ static int check_preds(struct filter_parse_state *ps)
 	return 0;
 }
 
-static int replace_preds(struct event_subsystem *system,
-			 struct ftrace_event_call *call,
+static int replace_preds(struct ftrace_event_call *call,
+			 struct event_filter *filter,
 			 struct filter_parse_state *ps,
 			 char *filter_string,
 			 bool dry_run)
@@ -1138,11 +1215,7 @@ static int replace_preds(struct event_subsystem *system,
 add_pred:
 		if (!pred)
 			return -ENOMEM;
-		if (call)
-			err = filter_add_pred(ps, call, pred, false);
-		else
-			err = filter_add_subsystem_pred(ps, system, pred,
-						filter_string, dry_run);
+		err = filter_add_pred(ps, call, filter, pred, dry_run);
 		filter_free_pred(pred);
 		if (err)
 			return err;
@@ -1153,10 +1226,50 @@ static int replace_preds(struct event_subsystem *system,
 	return 0;
 }
 
-int apply_event_filter(struct ftrace_event_call *call, char *filter_string)
+static int replace_system_preds(struct event_subsystem *system,
+				struct filter_parse_state *ps,
+				char *filter_string)
 {
+	struct event_filter *filter = system->filter;
+	struct ftrace_event_call *call;
+	bool fail = true;
 	int err;
 
+	list_for_each_entry(call, &ftrace_events, list) {
+
+		if (!call->define_fields)
+			continue;
+
+		if (strcmp(call->system, system->name) != 0)
+			continue;
+
+		/* try to see if the filter can be applied */
+		err = replace_preds(call, filter, ps, filter_string, true);
+		if (err)
+			continue;
+
+		/* really apply the filter */
+		filter_disable_preds(call);
+		err = replace_preds(call, filter, ps, filter_string, false);
+		if (err)
+			filter_disable_preds(call);
+		else {
+			call->filter_active = 1;
+			replace_filter_string(filter, filter_string);
+		}
+		fail = false;
+	}
+
+	if (fail) {
+		parse_error(ps, FILT_ERR_BAD_SUBSYS_FILTER, 0);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+int apply_event_filter(struct ftrace_event_call *call, char *filter_string)
+{
+	int err;
 	struct filter_parse_state *ps;
 
 	mutex_lock(&event_mutex);
@@ -1168,8 +1281,7 @@ int apply_event_filter(struct ftrace_event_call *call, char *filter_string)
 	if (!strcmp(strstrip(filter_string), "0")) {
 		filter_disable_preds(call);
 		remove_filter_string(call->filter);
-		mutex_unlock(&event_mutex);
-		return 0;
+		goto out_unlock;
 	}
 
 	err = -ENOMEM;
@@ -1187,10 +1299,11 @@ int apply_event_filter(struct ftrace_event_call *call, char *filter_string)
 		goto out;
 	}
 
-	err = replace_preds(NULL, call, ps, filter_string, false);
+	err = replace_preds(call, call->filter, ps, filter_string, false);
 	if (err)
 		append_filter_err(ps, call->filter);
-
+	else
+		call->filter_active = 1;
 out:
 	filter_opstack_clear(ps);
 	postfix_clear(ps);
@@ -1205,7 +1318,6 @@ int apply_subsystem_event_filter(struct event_subsystem *system,
 				 char *filter_string)
 {
 	int err;
-
 	struct filter_parse_state *ps;
 
 	mutex_lock(&event_mutex);
@@ -1215,10 +1327,9 @@ int apply_subsystem_event_filter(struct event_subsystem *system,
 		goto out_unlock;
 
 	if (!strcmp(strstrip(filter_string), "0")) {
-		filter_free_subsystem_preds(system, FILTER_DISABLE_ALL);
+		filter_free_subsystem_preds(system);
 		remove_filter_string(system->filter);
-		mutex_unlock(&event_mutex);
-		return 0;
+		goto out_unlock;
 	}
 
 	err = -ENOMEM;
@@ -1235,31 +1346,87 @@ int apply_subsystem_event_filter(struct event_subsystem *system,
 		goto out;
 	}
 
-	filter_free_subsystem_preds(system, FILTER_INIT_NO_RESET);
-
-	/* try to see the filter can be applied to which events */
-	err = replace_preds(system, NULL, ps, filter_string, true);
-	if (err) {
+	err = replace_system_preds(system, ps, filter_string);
+	if (err)
 		append_filter_err(ps, system->filter);
-		goto out;
+
+out:
+	filter_opstack_clear(ps);
+	postfix_clear(ps);
+	kfree(ps);
+out_unlock:
+	mutex_unlock(&event_mutex);
+
+	return err;
+}
+
+#ifdef CONFIG_EVENT_PROFILE
+
+void ftrace_profile_free_filter(struct perf_event *event)
+{
+	struct event_filter *filter = event->filter;
+
+	event->filter = NULL;
+	__free_preds(filter);
+}
+
+int ftrace_profile_set_filter(struct perf_event *event, int event_id,
+			      char *filter_str)
+{
+	int err;
+	struct event_filter *filter;
+	struct filter_parse_state *ps;
+	struct ftrace_event_call *call = NULL;
+
+	mutex_lock(&event_mutex);
+
+	list_for_each_entry(call, &ftrace_events, list) {
+		if (call->id == event_id)
+			break;
 	}
 
-	filter_free_subsystem_preds(system, FILTER_SKIP_NO_RESET);
+	err = -EINVAL;
+	if (!call)
+		goto out_unlock;
 
-	/* really apply the filter to the events */
-	err = replace_preds(system, NULL, ps, filter_string, false);
-	if (err) {
-		append_filter_err(ps, system->filter);
-		filter_free_subsystem_preds(system, 2);
+	err = -EEXIST;
+	if (event->filter)
+		goto out_unlock;
+
+	filter = __alloc_preds();
+	if (IS_ERR(filter)) {
+		err = PTR_ERR(filter);
+		goto out_unlock;
 	}
 
-out:
+	err = -ENOMEM;
+	ps = kzalloc(sizeof(*ps), GFP_KERNEL);
+	if (!ps)
+		goto free_preds;
+
+	parse_init(ps, filter_ops, filter_str);
+	err = filter_parse(ps);
+	if (err)
+		goto free_ps;
+
+	err = replace_preds(call, filter, ps, filter_str, false);
+	if (!err)
+		event->filter = filter;
+
+free_ps:
 	filter_opstack_clear(ps);
 	postfix_clear(ps);
 	kfree(ps);
+
+free_preds:
+	if (err)
+		__free_preds(filter);
+
 out_unlock:
 	mutex_unlock(&event_mutex);
 
 	return err;
 }
 
+#endif /* CONFIG_EVENT_PROFILE */
+

kernel/trace/trace_export.c

@@ -66,44 +66,47 @@ static void __used ____ftrace_check_##name(void)		\
 #undef __field
 #define __field(type, item)						\
 	ret = trace_seq_printf(s, "\tfield:" #type " " #item ";\t"	\
-			       "offset:%zu;\tsize:%zu;\n",		\
+			       "offset:%zu;\tsize:%zu;\tsigned:%u;\n",	\
 			       offsetof(typeof(field), item),		\
-			       sizeof(field.item));			\
+			       sizeof(field.item), is_signed_type(type)); \
 	if (!ret)							\
 		return 0;
 
 #undef __field_desc
 #define __field_desc(type, container, item)				\
 	ret = trace_seq_printf(s, "\tfield:" #type " " #item ";\t"	\
-			       "offset:%zu;\tsize:%zu;\n",		\
+			       "offset:%zu;\tsize:%zu;\tsigned:%u;\n",	\
 			       offsetof(typeof(field), container.item),	\
-			       sizeof(field.container.item));		\
+			       sizeof(field.container.item),		\
+			       is_signed_type(type));			\
 	if (!ret)							\
 		return 0;
 
 #undef __array
 #define __array(type, item, len)					\
 	ret = trace_seq_printf(s, "\tfield:" #type " " #item "[" #len "];\t" \
-			       "offset:%zu;\tsize:%zu;\n",		\
-			       offsetof(typeof(field), item),	\
-			       sizeof(field.item));		\
+			       "offset:%zu;\tsize:%zu;\tsigned:%u;\n",	\
+			       offsetof(typeof(field), item),		\
+			       sizeof(field.item), is_signed_type(type)); \
 	if (!ret)							\
 		return 0;
 
 #undef __array_desc
 #define __array_desc(type, container, item, len)			\
 	ret = trace_seq_printf(s, "\tfield:" #type " " #item "[" #len "];\t" \
-			       "offset:%zu;\tsize:%zu;\n",		\
+			       "offset:%zu;\tsize:%zu;\tsigned:%u;\n",	\
 			       offsetof(typeof(field), container.item),	\
-			       sizeof(field.container.item));		\
+			       sizeof(field.container.item),		\
+			       is_signed_type(type));			\
 	if (!ret)							\
 		return 0;
 
 #undef __dynamic_array
 #define __dynamic_array(type, item)					\
 	ret = trace_seq_printf(s, "\tfield:" #type " " #item ";\t"	\
-			       "offset:%zu;\tsize:0;\n",		\
-			       offsetof(typeof(field), item));		\
+			       "offset:%zu;\tsize:0;\tsigned:%u;\n",	\
+			       offsetof(typeof(field), item),		\
+			       is_signed_type(type));			\
 	if (!ret)							\
 		return 0;
 

kernel/trace/trace_syscalls.c

@@ -14,6 +14,69 @@ static int sys_refcount_exit;
 static DECLARE_BITMAP(enabled_enter_syscalls, NR_syscalls);
 static DECLARE_BITMAP(enabled_exit_syscalls, NR_syscalls);
 
+extern unsigned long __start_syscalls_metadata[];
+extern unsigned long __stop_syscalls_metadata[];
+
+static struct syscall_metadata **syscalls_metadata;
+
+static struct syscall_metadata *find_syscall_meta(unsigned long syscall)
+{
+	struct syscall_metadata *start;
+	struct syscall_metadata *stop;
+	char str[KSYM_SYMBOL_LEN];
+
+
+	start = (struct syscall_metadata *)__start_syscalls_metadata;
+	stop = (struct syscall_metadata *)__stop_syscalls_metadata;
+	kallsyms_lookup(syscall, NULL, NULL, NULL, str);
+
+	for ( ; start < stop; start++) {
+		/*
+		 * Only compare after the "sys" prefix. Archs that use
+		 * syscall wrappers may have syscalls symbols aliases prefixed
+		 * with "SyS" instead of "sys", leading to an unwanted
+		 * mismatch.
+		 */
+		if (start->name && !strcmp(start->name + 3, str + 3))
+			return start;
+	}
+	return NULL;
+}
+
+static struct syscall_metadata *syscall_nr_to_meta(int nr)
+{
+	if (!syscalls_metadata || nr >= NR_syscalls || nr < 0)
+		return NULL;
+
+	return syscalls_metadata[nr];
+}
+
+int syscall_name_to_nr(char *name)
+{
+	int i;
+
+	if (!syscalls_metadata)
+		return -1;
+
+	for (i = 0; i < NR_syscalls; i++) {
+		if (syscalls_metadata[i]) {
+			if (!strcmp(syscalls_metadata[i]->name, name))
+				return i;
+		}
+	}
+	return -1;
+}
+
+void set_syscall_enter_id(int num, int id)
+{
+	syscalls_metadata[num]->enter_id = id;
+}
+
+void set_syscall_exit_id(int num, int id)
+{
+	syscalls_metadata[num]->exit_id = id;
+}
+
 enum print_line_t
 print_syscall_enter(struct trace_iterator *iter, int flags)
 {
@@ -103,7 +166,8 @@ extern char *__bad_type_size(void);
 #define SYSCALL_FIELD(type, name)					\
 	sizeof(type) != sizeof(trace.name) ?				\
 		__bad_type_size() :					\
-		#type, #name, offsetof(typeof(trace), name), sizeof(trace.name)
+		#type, #name, offsetof(typeof(trace), name),		\
+		sizeof(trace.name), is_signed_type(type)
 
 int syscall_enter_format(struct ftrace_event_call *call, struct trace_seq *s)
 {
@@ -120,7 +184,8 @@ int syscall_enter_format(struct ftrace_event_call *call, struct trace_seq *s)
 	if (!entry)
 		return 0;
 
-	ret = trace_seq_printf(s, "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n",
+	ret = trace_seq_printf(s, "\tfield:%s %s;\toffset:%zu;\tsize:%zu;"
+			       "\tsigned:%u;\n",
 			       SYSCALL_FIELD(int, nr));
 	if (!ret)
 		return 0;
@@ -130,8 +195,10 @@ int syscall_enter_format(struct ftrace_event_call *call, struct trace_seq *s)
 				        entry->args[i]);
 		if (!ret)
 			return 0;
-		ret = trace_seq_printf(s, "\toffset:%d;\tsize:%zu;\n", offset,
-				       sizeof(unsigned long));
+		ret = trace_seq_printf(s, "\toffset:%d;\tsize:%zu;"
+				       "\tsigned:%u;\n", offset,
+				       sizeof(unsigned long),
+				       is_signed_type(unsigned long));
 		if (!ret)
 			return 0;
 		offset += sizeof(unsigned long);
@@ -163,8 +230,10 @@ int syscall_exit_format(struct ftrace_event_call *call, struct trace_seq *s)
 	struct syscall_trace_exit trace;
 
 	ret = trace_seq_printf(s,
-			       "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n"
-			       "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n",
+			       "\tfield:%s %s;\toffset:%zu;\tsize:%zu;"
+			       "\tsigned:%u;\n"
+			       "\tfield:%s %s;\toffset:%zu;\tsize:%zu;"
+			       "\tsigned:%u;\n",
 			       SYSCALL_FIELD(int, nr),
 			       SYSCALL_FIELD(long, ret));
 	if (!ret)
@@ -212,7 +281,7 @@ int syscall_exit_define_fields(struct ftrace_event_call *call)
 	if (ret)
 		return ret;
 
-	ret = trace_define_field(call, SYSCALL_FIELD(long, ret), 0,
+	ret = trace_define_field(call, SYSCALL_FIELD(long, ret),
 				 FILTER_OTHER);
 
 	return ret;
@@ -375,6 +444,29 @@ struct trace_event event_syscall_exit = {
 	.trace			= print_syscall_exit,
 };
 
+int __init init_ftrace_syscalls(void)
+{
+	struct syscall_metadata *meta;
+	unsigned long addr;
+	int i;
+
+	syscalls_metadata = kzalloc(sizeof(*syscalls_metadata) *
+					NR_syscalls, GFP_KERNEL);
+	if (!syscalls_metadata) {
+		WARN_ON(1);
+		return -ENOMEM;
+	}
+
+	for (i = 0; i < NR_syscalls; i++) {
+		addr = arch_syscall_addr(i);
+		meta = find_syscall_meta(addr);
+		syscalls_metadata[i] = meta;
+	}
+
+	return 0;
+}
+core_initcall(init_ftrace_syscalls);
+
 #ifdef CONFIG_EVENT_PROFILE
 
 static DECLARE_BITMAP(enabled_prof_enter_syscalls, NR_syscalls);

lib/kernel_lock.c

@@ -5,10 +5,13 @@
  * relegated to obsolescence, but used by various less
  * important (or lazy) subsystems.
  */
-#include <linux/smp_lock.h>
 #include <linux/module.h>
 #include <linux/kallsyms.h>
 #include <linux/semaphore.h>
+#include <linux/smp_lock.h>
+
+#define CREATE_TRACE_POINTS
+#include <trace/events/bkl.h>
 
 /*
  * The 'big kernel lock'
@@ -113,21 +116,26 @@ static inline void __unlock_kernel(void)
  * This cannot happen asynchronously, so we only need to
  * worry about other CPU's.
  */
-void __lockfunc lock_kernel(void)
+void __lockfunc _lock_kernel(const char *func, const char *file, int line)
 {
-	int depth = current->lock_depth+1;
+	int depth = current->lock_depth + 1;
+
+	trace_lock_kernel(func, file, line);
+
 	if (likely(!depth))
 		__lock_kernel();
 	current->lock_depth = depth;
 }
 
-void __lockfunc unlock_kernel(void)
+void __lockfunc _unlock_kernel(const char *func, const char *file, int line)
 {
 	BUG_ON(current->lock_depth < 0);
 	if (likely(--current->lock_depth < 0))
 		__unlock_kernel();
+
+	trace_unlock_kernel(func, file, line);
 }
 
-EXPORT_SYMBOL(lock_kernel);
-EXPORT_SYMBOL(unlock_kernel);
+EXPORT_SYMBOL(_lock_kernel);
+EXPORT_SYMBOL(_unlock_kernel);
 

scripts/recordmcount.pl

@@ -119,6 +119,7 @@ my %text_sections = (
      ".sched.text" => 1,
      ".spinlock.text" => 1,
      ".irqentry.text" => 1,
+     ".text.unlikely" => 1,
 );
 
 $objdump = "objdump" if ((length $objdump) == 0);

tools/perf/Documentation/perf-timechart.txt

@@ -31,9 +31,12 @@ OPTIONS
 -w::
 --width=::
         Select the width of the SVG file (default: 1000)
--p::
+-P::
 --power-only::
         Only output the CPU power section of the diagram
+-p::
+--process::
+        Select the processes to display, by name or PID
 
 
 SEE ALSO

tools/perf/Makefile

@@ -201,7 +201,14 @@ EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wold-style-definition
 EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstrict-prototypes
 EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wdeclaration-after-statement
 
-CFLAGS = $(MBITS) -ggdb3 -Wall -Wextra -std=gnu99 -Werror -O6 -fstack-protector-all -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS)
+ifeq ("$(origin DEBUG)", "command line")
+  PERF_DEBUG = $(DEBUG)
+endif
+ifndef PERF_DEBUG
+  CFLAGS_OPTIMIZE = -O6
+endif
+
+CFLAGS = $(MBITS) -ggdb3 -Wall -Wextra -std=gnu99 -Werror $(CFLAGS_OPTIMIZE) -fstack-protector-all -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS)
 LDFLAGS = -lpthread -lrt -lelf -lm
 ALL_CFLAGS = $(CFLAGS)
 ALL_LDFLAGS = $(LDFLAGS)
@@ -329,8 +336,26 @@ LIB_H += ../../include/linux/perf_event.h
 LIB_H += ../../include/linux/rbtree.h
 LIB_H += ../../include/linux/list.h
 LIB_H += ../../include/linux/stringify.h
+LIB_H += util/include/linux/bitmap.h
+LIB_H += util/include/linux/bitops.h
+LIB_H += util/include/linux/compiler.h
+LIB_H += util/include/linux/ctype.h
+LIB_H += util/include/linux/kernel.h
 LIB_H += util/include/linux/list.h
+LIB_H += util/include/linux/module.h
+LIB_H += util/include/linux/poison.h
+LIB_H += util/include/linux/prefetch.h
+LIB_H += util/include/linux/rbtree.h
+LIB_H += util/include/linux/string.h
+LIB_H += util/include/linux/types.h
+LIB_H += util/include/asm/asm-offsets.h
+LIB_H += util/include/asm/bitops.h
+LIB_H += util/include/asm/byteorder.h
+LIB_H += util/include/asm/swab.h
+LIB_H += util/include/asm/system.h
+LIB_H += util/include/asm/uaccess.h
 LIB_H += perf.h
+LIB_H += util/event.h
 LIB_H += util/types.h
 LIB_H += util/levenshtein.h
 LIB_H += util/parse-options.h
@@ -344,9 +369,12 @@ LIB_H += util/strlist.h
 LIB_H += util/run-command.h
 LIB_H += util/sigchain.h
 LIB_H += util/symbol.h
-LIB_H += util/module.h
 LIB_H += util/color.h
 LIB_H += util/values.h
+LIB_H += util/sort.h
+LIB_H += util/hist.h
+LIB_H += util/thread.h
+LIB_H += util/data_map.h
 
 LIB_OBJS += util/abspath.o
 LIB_OBJS += util/alias.o
@@ -360,6 +388,9 @@ LIB_OBJS += util/parse-options.o
 LIB_OBJS += util/parse-events.o
 LIB_OBJS += util/path.o
 LIB_OBJS += util/rbtree.o
+LIB_OBJS += util/bitmap.o
+LIB_OBJS += util/hweight.o
+LIB_OBJS += util/find_next_bit.o
 LIB_OBJS += util/run-command.o
 LIB_OBJS += util/quote.o
 LIB_OBJS += util/strbuf.o
@@ -369,7 +400,6 @@ LIB_OBJS += util/usage.o
 LIB_OBJS += util/wrapper.o
 LIB_OBJS += util/sigchain.o
 LIB_OBJS += util/symbol.o
-LIB_OBJS += util/module.o
 LIB_OBJS += util/color.o
 LIB_OBJS += util/pager.o
 LIB_OBJS += util/header.o
@@ -382,6 +412,9 @@ LIB_OBJS += util/trace-event-parse.o
 LIB_OBJS += util/trace-event-read.o
 LIB_OBJS += util/trace-event-info.o
 LIB_OBJS += util/svghelper.o
+LIB_OBJS += util/sort.o
+LIB_OBJS += util/hist.o
+LIB_OBJS += util/data_map.o
 
 BUILTIN_OBJS += builtin-annotate.o
 BUILTIN_OBJS += builtin-help.o
@@ -424,8 +457,12 @@ ifeq ($(uname_S),Darwin)
 	PTHREAD_LIBS =
 endif
 
+ifneq ($(shell sh -c "(echo '\#include <gnu/libc-version.h>'; echo 'int main(void) { const char * version = gnu_get_libc_version(); return (long)version; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
+	msg := $(error No gnu/libc-version.h found, please install glibc-dev[el]);
+endif
+
 ifneq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ_MMAP, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
-	msg := $(error No libelf.h/libelf found, please install libelf-dev/elfutils-libelf-devel and glibc-dev[el]);
+	msg := $(error No libelf.h/libelf found, please install libelf-dev/elfutils-libelf-devel);
 endif
 
 ifneq ($(shell sh -c "(echo '\#include <libdwarf/dwarf.h>'; echo '\#include <libdwarf/libdwarf.h>'; echo 'int main(void) { Dwarf_Debug dbg; Dwarf_Error err; Dwarf_Ranges *rng; dwarf_init(0, DW_DLC_READ, 0, 0, &dbg, &err); dwarf_get_ranges(dbg, 0, &rng, 0, 0, &err); return (long)dbg; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -ldwarf -lelf -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
@@ -795,6 +832,19 @@ util/config.o: util/config.c PERF-CFLAGS
 util/rbtree.o: ../../lib/rbtree.c PERF-CFLAGS
 	$(QUIET_CC)$(CC) -o util/rbtree.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
 
+# some perf warning policies can't fit to lib/bitmap.c, eg: it warns about variable shadowing
+# from <string.h> that comes from kernel headers wrapping.
+KBITMAP_FLAGS=`echo $(ALL_CFLAGS) | sed s/-Wshadow// | sed s/-Wswitch-default// | sed s/-Wextra//`
+
+util/bitmap.o: ../../lib/bitmap.c PERF-CFLAGS
+	$(QUIET_CC)$(CC) -o util/bitmap.o -c $(KBITMAP_FLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
+
+util/hweight.o: ../../lib/hweight.c PERF-CFLAGS
+	$(QUIET_CC)$(CC) -o util/hweight.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
+
+util/find_next_bit.o: ../../lib/find_next_bit.c PERF-CFLAGS
+	$(QUIET_CC)$(CC) -o util/find_next_bit.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
+
 perf-%$X: %.o $(PERFLIBS)
 	$(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ $(ALL_LDFLAGS) $(filter %.o,$^) $(LIBS)
 

tools/perf/builtin-record.c

@@ -17,55 +17,51 @@
 #include "util/header.h"
 #include "util/event.h"
 #include "util/debug.h"
-#include "util/trace-event.h"
 
 #include <unistd.h>
 #include <sched.h>
 
-#define ALIGN(x, a)		__ALIGN_MASK(x, (typeof(x))(a)-1)
-#define __ALIGN_MASK(x, mask)	(((x)+(mask))&~(mask))
-
 static int			fd[MAX_NR_CPUS][MAX_COUNTERS];
 
-static long			default_interval		= 100000;
+static long			default_interval		=      0;
 
-static int			nr_cpus				= 0;
+static int			nr_cpus				=      0;
 static unsigned int		page_size;
-static unsigned int		mmap_pages			= 128;
-static int			freq				= 0;
+static unsigned int		mmap_pages			=    128;
+static int			freq				=   1000;
 static int			output;
 static const char		*output_name			= "perf.data";
-static int			group				= 0;
-static unsigned int		realtime_prio			= 0;
-static int			raw_samples			= 0;
-static int			system_wide			= 0;
-static int			profile_cpu			= -1;
-static pid_t			target_pid			= -1;
-static pid_t			child_pid			= -1;
-static int			inherit				= 1;
-static int			force				= 0;
-static int			append_file			= 0;
-static int			call_graph			= 0;
-static int			inherit_stat			= 0;
-static int			no_samples			= 0;
-static int			sample_address			= 0;
-static int			multiplex			= 0;
-static int			multiplex_fd			= -1;
-
-static long			samples;
+static int			group				=      0;
+static unsigned int		realtime_prio			=      0;
+static int			raw_samples			=      0;
+static int			system_wide			=      0;
+static int			profile_cpu			=     -1;
+static pid_t			target_pid			=     -1;
+static pid_t			child_pid			=     -1;
+static int			inherit				=      1;
+static int			force				=      0;
+static int			append_file			=      0;
+static int			call_graph			=      0;
+static int			inherit_stat			=      0;
+static int			no_samples			=      0;
+static int			sample_address			=      0;
+static int			multiplex			=      0;
+static int			multiplex_fd			=     -1;
+
+static long			samples				=      0;
 static struct timeval		last_read;
 static struct timeval		this_read;
 
-static u64			bytes_written;
+static u64			bytes_written			=      0;
 
 static struct pollfd		event_array[MAX_NR_CPUS * MAX_COUNTERS];
 
-static int			nr_poll;
-static int			nr_cpu;
+static int			nr_poll				=      0;
+static int			nr_cpu				=      0;
 
-static int			file_new = 1;
+static int			file_new			=      1;
 
-struct perf_header		*header;
+struct perf_header		*header				=   NULL;
 
 struct mmap_data {
 	int			counter;
@@ -375,9 +371,11 @@ static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int n
 
 static void create_counter(int counter, int cpu, pid_t pid)
 {
+	char *filter = filters[counter];
 	struct perf_event_attr *attr = attrs + counter;
 	struct perf_header_attr *h_attr;
 	int track = !counter; /* only the first counter needs these */
+	int ret;
 	struct {
 		u64 count;
 		u64 time_enabled;
@@ -480,7 +478,6 @@ static void create_counter(int counter, int cpu, pid_t pid)
 		multiplex_fd = fd[nr_cpu][counter];
 
 	if (multiplex && fd[nr_cpu][counter] != multiplex_fd) {
-		int ret;
 
 		ret = ioctl(fd[nr_cpu][counter], PERF_EVENT_IOC_SET_OUTPUT, multiplex_fd);
 		assert(ret != -1);
@@ -500,6 +497,16 @@ static void create_counter(int counter, int cpu, pid_t pid)
 		}
 	}
 
+	if (filter != NULL) {
+		ret = ioctl(fd[nr_cpu][counter],
+			    PERF_EVENT_IOC_SET_FILTER, filter);
+		if (ret) {
+			error("failed to set filter with %d (%s)\n", errno,
+			      strerror(errno));
+			exit(-1);
+		}
+	}
+
 	ioctl(fd[nr_cpu][counter], PERF_EVENT_IOC_ENABLE);
 }
 
@@ -566,17 +573,17 @@ static int __cmd_record(int argc, const char **argv)
 	else
 		header = perf_header__new();
 
-
 	if (raw_samples) {
-		read_tracing_data(attrs, nr_counters);
+		perf_header__feat_trace_info(header);
 	} else {
 		for (i = 0; i < nr_counters; i++) {
 			if (attrs[i].sample_type & PERF_SAMPLE_RAW) {
-				read_tracing_data(attrs, nr_counters);
+				perf_header__feat_trace_info(header);
 				break;
 			}
 		}
 	}
+
 	atexit(atexit_header);
 
 	if (!system_wide) {
@@ -623,7 +630,7 @@ static int __cmd_record(int argc, const char **argv)
 
 		param.sched_priority = realtime_prio;
 		if (sched_setscheduler(0, SCHED_FIFO, &param)) {
-			printf("Could not set realtime priority.\n");
+			pr_err("Could not set realtime priority.\n");
 			exit(-1);
 		}
 	}
@@ -677,6 +684,8 @@ static const struct option options[] = {
 	OPT_CALLBACK('e', "event", NULL, "event",
 		     "event selector. use 'perf list' to list available events",
 		     parse_events),
+	OPT_CALLBACK(0, "filter", NULL, "filter",
+		     "event filter", parse_filter),
 	OPT_INTEGER('p', "pid", &target_pid,
 		    "record events on existing pid"),
 	OPT_INTEGER('r', "realtime", &realtime_prio,
@@ -731,6 +740,18 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 		attrs[0].config = PERF_COUNT_HW_CPU_CYCLES;
 	}
 
+	/*
+	 * User specified count overrides default frequency.
+	 */
+	if (default_interval)
+		freq = 0;
+	else if (freq) {
+		default_interval = freq;
+	} else {
+		fprintf(stderr, "frequency and count are zero, aborting\n");
+		exit(EXIT_FAILURE);
+	}
+
 	for (counter = 0; counter < nr_counters; counter++) {
 		if (attrs[counter].sample_period)
 			continue;

tools/perf/builtin-sched.c

@@ -11,6 +11,7 @@
 #include "util/trace-event.h"
 
 #include "util/debug.h"
+#include "util/data_map.h"
 
 #include <sys/types.h>
 #include <sys/prctl.h>
@@ -20,26 +21,23 @@
 #include <math.h>
 
 static char			const *input_name = "perf.data";
-static int			input;
-static unsigned long		page_size;
-static unsigned long		mmap_window = 32;
 
 static unsigned long		total_comm = 0;
 
-static struct rb_root		threads;
-static struct thread		*last_match;
-
 static struct perf_header	*header;
 static u64			sample_type;
 
 static char			default_sort_order[] = "avg, max, switch, runtime";
 static char			*sort_order = default_sort_order;
 
+static int			profile_cpu = -1;
+
+static char			*cwd;
+static int			cwdlen;
+
 #define PR_SET_NAME		15               /* Set process name */
 #define MAX_CPUS		4096
 
-#define BUG_ON(x)		assert(!(x))
-
 static u64			run_measurement_overhead;
 static u64			sleep_measurement_overhead;
 
@@ -74,6 +72,7 @@ enum sched_event_type {
 	SCHED_EVENT_RUN,
 	SCHED_EVENT_SLEEP,
 	SCHED_EVENT_WAKEUP,
+	SCHED_EVENT_MIGRATION,
 };
 
 struct sched_atom {
@@ -398,6 +397,8 @@ process_sched_event(struct task_desc *this_task __used, struct sched_atom *atom)
 				ret = sem_post(atom->wait_sem);
 			BUG_ON(ret);
 			break;
+		case SCHED_EVENT_MIGRATION:
+			break;
 		default:
 			BUG_ON(1);
 	}
@@ -635,9 +636,7 @@ static void test_calibrations(void)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread;
-
-	thread = threads__findnew(event->comm.pid, &threads, &last_match);
+	struct thread *thread = threads__findnew(event->comm.tid);
 
 	dump_printf("%p [%p]: perf_event_comm: %s:%d\n",
 		(void *)(offset + head),
@@ -745,6 +744,22 @@ struct trace_fork_event {
 	u32 child_pid;
 };
 
+struct trace_migrate_task_event {
+	u32 size;
+
+	u16 common_type;
+	u8 common_flags;
+	u8 common_preempt_count;
+	u32 common_pid;
+	u32 common_tgid;
+
+	char comm[16];
+	u32 pid;
+
+	u32 prio;
+	u32 cpu;
+};
+
 struct trace_sched_handler {
 	void (*switch_event)(struct trace_switch_event *,
 			     struct event *,
@@ -769,6 +784,12 @@ struct trace_sched_handler {
 			   int cpu,
 			   u64 timestamp,
 			   struct thread *thread);
+
+	void (*migrate_task_event)(struct trace_migrate_task_event *,
+			   struct event *,
+			   int cpu,
+			   u64 timestamp,
+			   struct thread *thread);
 };
 
 
@@ -1058,8 +1079,8 @@ latency_switch_event(struct trace_switch_event *switch_event,
 		die("hm, delta: %Ld < 0 ?\n", delta);
 
 
-	sched_out = threads__findnew(switch_event->prev_pid, &threads, &last_match);
-	sched_in = threads__findnew(switch_event->next_pid, &threads, &last_match);
+	sched_out = threads__findnew(switch_event->prev_pid);
+	sched_in = threads__findnew(switch_event->next_pid);
 
 	out_events = thread_atoms_search(&atom_root, sched_out, &cmp_pid);
 	if (!out_events) {
@@ -1092,13 +1113,10 @@ latency_runtime_event(struct trace_runtime_event *runtime_event,
 		     u64 timestamp,
 		     struct thread *this_thread __used)
 {
-	struct work_atoms *atoms;
-	struct thread *thread;
+	struct thread *thread = threads__findnew(runtime_event->pid);
+	struct work_atoms *atoms = thread_atoms_search(&atom_root, thread, &cmp_pid);
 
 	BUG_ON(cpu >= MAX_CPUS || cpu < 0);
-
-	thread = threads__findnew(runtime_event->pid, &threads, &last_match);
-	atoms = thread_atoms_search(&atom_root, thread, &cmp_pid);
 	if (!atoms) {
 		thread_atoms_insert(thread);
 		atoms = thread_atoms_search(&atom_root, thread, &cmp_pid);
@@ -1125,7 +1143,7 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event,
 	if (!wakeup_event->success)
 		return;
 
-	wakee = threads__findnew(wakeup_event->pid, &threads, &last_match);
+	wakee = threads__findnew(wakeup_event->pid);
 	atoms = thread_atoms_search(&atom_root, wakee, &cmp_pid);
 	if (!atoms) {
 		thread_atoms_insert(wakee);
@@ -1139,7 +1157,12 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event,
 
 	atom = list_entry(atoms->work_list.prev, struct work_atom, list);
 
-	if (atom->state != THREAD_SLEEPING)
+	/*
+	 * You WILL be missing events if you've recorded only
+	 * one CPU, or are only looking at only one, so don't
+	 * make useless noise.
+	 */
+	if (profile_cpu == -1 && atom->state != THREAD_SLEEPING)
 		nr_state_machine_bugs++;
 
 	nr_timestamps++;
@@ -1152,11 +1175,51 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event,
 	atom->wake_up_time = timestamp;
 }
 
+static void
+latency_migrate_task_event(struct trace_migrate_task_event *migrate_task_event,
+		     struct event *__event __used,
+		     int cpu __used,
+		     u64 timestamp,
+		     struct thread *thread __used)
+{
+	struct work_atoms *atoms;
+	struct work_atom *atom;
+	struct thread *migrant;
+
+	/*
+	 * Only need to worry about migration when profiling one CPU.
+	 */
+	if (profile_cpu == -1)
+		return;
+
+	migrant = threads__findnew(migrate_task_event->pid);
+	atoms = thread_atoms_search(&atom_root, migrant, &cmp_pid);
+	if (!atoms) {
+		thread_atoms_insert(migrant);
+		register_pid(migrant->pid, migrant->comm);
+		atoms = thread_atoms_search(&atom_root, migrant, &cmp_pid);
+		if (!atoms)
+			die("migration-event: Internal tree error");
+		add_sched_out_event(atoms, 'R', timestamp);
+	}
+
+	BUG_ON(list_empty(&atoms->work_list));
+
+	atom = list_entry(atoms->work_list.prev, struct work_atom, list);
+	atom->sched_in_time = atom->sched_out_time = atom->wake_up_time = timestamp;
+
+	nr_timestamps++;
+
+	if (atom->sched_out_time > timestamp)
+		nr_unordered_timestamps++;
+}
+
 static struct trace_sched_handler lat_ops  = {
 	.wakeup_event		= latency_wakeup_event,
 	.switch_event		= latency_switch_event,
 	.runtime_event		= latency_runtime_event,
 	.fork_event		= latency_fork_event,
+	.migrate_task_event	= latency_migrate_task_event,
 };
 
 static void output_lat_thread(struct work_atoms *work_list)
@@ -1385,8 +1448,8 @@ map_switch_event(struct trace_switch_event *switch_event,
 		die("hm, delta: %Ld < 0 ?\n", delta);
 
 
-	sched_out = threads__findnew(switch_event->prev_pid, &threads, &last_match);
-	sched_in = threads__findnew(switch_event->next_pid, &threads, &last_match);
+	sched_out = threads__findnew(switch_event->prev_pid);
+	sched_in = threads__findnew(switch_event->next_pid);
 
 	curr_thread[this_cpu] = sched_in;
 
@@ -1516,6 +1579,26 @@ process_sched_exit_event(struct event *event,
 		printf("sched_exit event %p\n", event);
 }
 
+static void
+process_sched_migrate_task_event(struct raw_event_sample *raw,
+			   struct event *event,
+			   int cpu __used,
+			   u64 timestamp __used,
+			   struct thread *thread __used)
+{
+	struct trace_migrate_task_event migrate_task_event;
+
+	FILL_COMMON_FIELDS(migrate_task_event, event, raw->data);
+
+	FILL_ARRAY(migrate_task_event, comm, event, raw->data);
+	FILL_FIELD(migrate_task_event, pid, event, raw->data);
+	FILL_FIELD(migrate_task_event, prio, event, raw->data);
+	FILL_FIELD(migrate_task_event, cpu, event, raw->data);
+
+	if (trace_handler->migrate_task_event)
+		trace_handler->migrate_task_event(&migrate_task_event, event, cpu, timestamp, thread);
+}
+
 static void
 process_raw_event(event_t *raw_event __used, void *more_data,
 		  int cpu, u64 timestamp, struct thread *thread)
@@ -1539,23 +1622,24 @@ process_raw_event(event_t *raw_event __used, void *more_data,
 		process_sched_fork_event(raw, event, cpu, timestamp, thread);
 	if (!strcmp(event->name, "sched_process_exit"))
 		process_sched_exit_event(event, cpu, timestamp, thread);
+	if (!strcmp(event->name, "sched_migrate_task"))
+		process_sched_migrate_task_event(raw, event, cpu, timestamp, thread);
 }
 
 static int
 process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	char level;
-	int show = 0;
-	struct dso *dso = NULL;
 	struct thread *thread;
 	u64 ip = event->ip.ip;
 	u64 timestamp = -1;
 	u32 cpu = -1;
 	u64 period = 1;
 	void *more_data = event->ip.__more_data;
-	int cpumode;
 
-	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+	if (!(sample_type & PERF_SAMPLE_RAW))
+		return 0;
+
+	thread = threads__findnew(event->ip.pid);
 
 	if (sample_type & PERF_SAMPLE_TIME) {
 		timestamp = *(u64 *)more_data;
@@ -1581,169 +1665,60 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 		(void *)(long)ip,
 		(long long)period);
 
-	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
-
 	if (thread == NULL) {
-		eprintf("problem processing %d event, skipping it.\n",
-			event->header.type);
+		pr_debug("problem processing %d event, skipping it.\n",
+			 event->header.type);
 		return -1;
 	}
 
-	cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
-
-	if (cpumode == PERF_RECORD_MISC_KERNEL) {
-		show = SHOW_KERNEL;
-		level = 'k';
-
-		dso = kernel_dso;
-
-		dump_printf(" ...... dso: %s\n", dso->name);
-
-	} else if (cpumode == PERF_RECORD_MISC_USER) {
-
-		show = SHOW_USER;
-		level = '.';
-
-	} else {
-		show = SHOW_HV;
-		level = 'H';
-
-		dso = hypervisor_dso;
+	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
 
-		dump_printf(" ...... dso: [hypervisor]\n");
-	}
+	if (profile_cpu != -1 && profile_cpu != (int) cpu)
+		return 0;
 
-	if (sample_type & PERF_SAMPLE_RAW)
-		process_raw_event(event, more_data, cpu, timestamp, thread);
+	process_raw_event(event, more_data, cpu, timestamp, thread);
 
 	return 0;
 }
 
 static int
-process_event(event_t *event, unsigned long offset, unsigned long head)
+process_lost_event(event_t *event __used,
+		   unsigned long offset __used,
+		   unsigned long head __used)
 {
-	trace_event(event);
+	nr_lost_chunks++;
+	nr_lost_events += event->lost.lost;
 
-	nr_events++;
-	switch (event->header.type) {
-	case PERF_RECORD_MMAP:
-		return 0;
-	case PERF_RECORD_LOST:
-		nr_lost_chunks++;
-		nr_lost_events += event->lost.lost;
-		return 0;
-
-	case PERF_RECORD_COMM:
-		return process_comm_event(event, offset, head);
-
-	case PERF_RECORD_EXIT ... PERF_RECORD_READ:
-		return 0;
+	return 0;
+}
 
-	case PERF_RECORD_SAMPLE:
-		return process_sample_event(event, offset, head);
+static int sample_type_check(u64 type)
+{
+	sample_type = type;
 
-	case PERF_RECORD_MAX:
-	default:
+	if (!(sample_type & PERF_SAMPLE_RAW)) {
+		fprintf(stderr,
+			"No trace sample to read. Did you call perf record "
+			"without -R?");
 		return -1;
 	}
 
 	return 0;
 }
 
+static struct perf_file_handler file_handler = {
+	.process_sample_event	= process_sample_event,
+	.process_comm_event	= process_comm_event,
+	.process_lost_event	= process_lost_event,
+	.sample_type_check	= sample_type_check,
+};
+
 static int read_events(void)
 {
-	int ret, rc = EXIT_FAILURE;
-	unsigned long offset = 0;
-	unsigned long head = 0;
-	struct stat perf_stat;
-	event_t *event;
-	uint32_t size;
-	char *buf;
-
-	trace_report();
-	register_idle_thread(&threads, &last_match);
-
-	input = open(input_name, O_RDONLY);
-	if (input < 0) {
-		perror("failed to open file");
-		exit(-1);
-	}
-
-	ret = fstat(input, &perf_stat);
-	if (ret < 0) {
-		perror("failed to stat file");
-		exit(-1);
-	}
-
-	if (!perf_stat.st_size) {
-		fprintf(stderr, "zero-sized file, nothing to do!\n");
-		exit(0);
-	}
-	header = perf_header__read(input);
-	head = header->data_offset;
-	sample_type = perf_header__sample_type(header);
-
-	if (!(sample_type & PERF_SAMPLE_RAW))
-		die("No trace sample to read. Did you call perf record "
-		    "without -R?");
-
-	if (load_kernel() < 0) {
-		perror("failed to load kernel symbols");
-		return EXIT_FAILURE;
-	}
-
-remap:
-	buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ,
-			   MAP_SHARED, input, offset);
-	if (buf == MAP_FAILED) {
-		perror("failed to mmap file");
-		exit(-1);
-	}
-
-more:
-	event = (event_t *)(buf + head);
-
-	size = event->header.size;
-	if (!size)
-		size = 8;
-
-	if (head + event->header.size >= page_size * mmap_window) {
-		unsigned long shift = page_size * (head / page_size);
-		int res;
-
-		res = munmap(buf, page_size * mmap_window);
-		assert(res == 0);
-
-		offset += shift;
-		head -= shift;
-		goto remap;
-	}
-
-	size = event->header.size;
-
-
-	if (!size || process_event(event, offset, head) < 0) {
-
-		/*
-		 * assume we lost track of the stream, check alignment, and
-		 * increment a single u64 in the hope to catch on again 'soon'.
-		 */
-
-		if (unlikely(head & 7))
-			head &= ~7ULL;
-
-		size = 8;
-	}
-
-	head += size;
-
-	if (offset + head < (unsigned long)perf_stat.st_size)
-		goto more;
-
-	rc = EXIT_SUCCESS;
-	close(input);
+	register_idle_thread();
+	register_perf_file_handler(&file_handler);
 
-	return rc;
+	return mmap_dispatch_perf_file(&header, input_name, 0, 0, &cwdlen, &cwd);
 }
 
 static void print_bad_events(void)
@@ -1883,6 +1858,8 @@ static const struct option latency_options[] = {
 		   "sort by key(s): runtime, switch, avg, max"),
 	OPT_BOOLEAN('v', "verbose", &verbose,
 		    "be more verbose (show symbol address, etc)"),
+	OPT_INTEGER('C', "CPU", &profile_cpu,
+		    "CPU to profile on"),
 	OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
 		    "dump raw trace in ASCII"),
 	OPT_END()
@@ -1961,7 +1938,6 @@ static int __cmd_record(int argc, const char **argv)
 int cmd_sched(int argc, const char **argv, const char *prefix __used)
 {
 	symbol__init();
-	page_size = getpagesize();
 
 	argc = parse_options(argc, argv, sched_options, sched_usage,
 			     PARSE_OPT_STOP_AT_NON_OPTION);

tools/perf/builtin-stat.c

@@ -50,15 +50,17 @@
 
 static struct perf_event_attr default_attrs[] = {
 
-  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK	},
-  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES},
-  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CPU_MIGRATIONS	},
-  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_PAGE_FAULTS	},
-
-  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES	},
-  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_INSTRUCTIONS	},
-  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_REFERENCES},
-  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_MISSES	},
+  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK		},
+  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES	},
+  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CPU_MIGRATIONS		},
+  { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_PAGE_FAULTS		},
+
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES		},
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_INSTRUCTIONS		},
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_INSTRUCTIONS	},
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_MISSES		},
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_REFERENCES	},
+  { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_MISSES		},
 
 };
 
@@ -125,6 +127,7 @@ struct stats			event_res_stats[MAX_COUNTERS][3];
 struct stats			runtime_nsecs_stats;
 struct stats			walltime_nsecs_stats;
 struct stats			runtime_cycles_stats;
+struct stats			runtime_branches_stats;
 
 #define MATCH_EVENT(t, c, counter)			\
 	(attrs[counter].type == PERF_TYPE_##t &&	\
@@ -235,6 +238,8 @@ static void read_counter(int counter)
 		update_stats(&runtime_nsecs_stats, count[0]);
 	if (MATCH_EVENT(HARDWARE, HW_CPU_CYCLES, counter))
 		update_stats(&runtime_cycles_stats, count[0]);
+	if (MATCH_EVENT(HARDWARE, HW_BRANCH_INSTRUCTIONS, counter))
+		update_stats(&runtime_branches_stats, count[0]);
 }
 
 static int run_perf_stat(int argc __used, const char **argv)
@@ -352,6 +357,14 @@ static void abs_printout(int counter, double avg)
 			ratio = avg / total;
 
 		fprintf(stderr, " # %10.3f IPC  ", ratio);
+	} else if (MATCH_EVENT(HARDWARE, HW_BRANCH_MISSES, counter)) {
+		total = avg_stats(&runtime_branches_stats);
+
+		if (total)
+			ratio = avg * 100 / total;
+
+		fprintf(stderr, " # %10.3f %%    ", ratio);
+
 	} else {
 		total = avg_stats(&runtime_nsecs_stats);
 

tools/perf/perf.c

@@ -89,8 +89,8 @@ static int handle_options(const char*** argv, int* argc, int* envchanged)
 		/*
 		 * Check remaining flags.
 		 */
-		if (!prefixcmp(cmd, "--exec-path")) {
-			cmd += 11;
+		if (!prefixcmp(cmd, CMD_EXEC_PATH)) {
+			cmd += strlen(CMD_EXEC_PATH);
 			if (*cmd == '=')
 				perf_set_argv_exec_path(cmd + 1);
 			else {
@@ -117,8 +117,8 @@ static int handle_options(const char*** argv, int* argc, int* envchanged)
 			(*argv)++;
 			(*argc)--;
 			handled++;
-		} else if (!prefixcmp(cmd, "--perf-dir=")) {
-			setenv(PERF_DIR_ENVIRONMENT, cmd + 10, 1);
+		} else if (!prefixcmp(cmd, CMD_PERF_DIR)) {
+			setenv(PERF_DIR_ENVIRONMENT, cmd + strlen(CMD_PERF_DIR), 1);
 			if (envchanged)
 				*envchanged = 1;
 		} else if (!strcmp(cmd, "--work-tree")) {
@@ -131,8 +131,8 @@ static int handle_options(const char*** argv, int* argc, int* envchanged)
 				*envchanged = 1;
 			(*argv)++;
 			(*argc)--;
-		} else if (!prefixcmp(cmd, "--work-tree=")) {
-			setenv(PERF_WORK_TREE_ENVIRONMENT, cmd + 12, 1);
+		} else if (!prefixcmp(cmd, CMD_WORK_TREE)) {
+			setenv(PERF_WORK_TREE_ENVIRONMENT, cmd + strlen(CMD_WORK_TREE), 1);
 			if (envchanged)
 				*envchanged = 1;
 		} else if (!strcmp(cmd, "--debugfs-dir")) {
@@ -146,8 +146,8 @@ static int handle_options(const char*** argv, int* argc, int* envchanged)
 				*envchanged = 1;
 			(*argv)++;
 			(*argc)--;
-		} else if (!prefixcmp(cmd, "--debugfs-dir=")) {
-			strncpy(debugfs_mntpt, cmd + 14, MAXPATHLEN);
+		} else if (!prefixcmp(cmd, CMD_DEBUGFS_DIR)) {
+			strncpy(debugfs_mntpt, cmd + strlen(CMD_DEBUGFS_DIR), MAXPATHLEN);
 			debugfs_mntpt[MAXPATHLEN - 1] = '\0';
 			if (envchanged)
 				*envchanged = 1;

tools/perf/util/PERF-VERSION-GEN

 #!/bin/sh
 
 GVF=PERF-VERSION-FILE
-DEF_VER=v0.0.1.PERF
+DEF_VER=v0.0.2.PERF
 
 LF='
 '

tools/perf/util/cache.h

-#ifndef CACHE_H
-#define CACHE_H
+#ifndef __PERF_CACHE_H
+#define __PERF_CACHE_H
 
 #include "util.h"
 #include "strbuf.h"
 #include "../perf.h"
 
+#define CMD_EXEC_PATH "--exec-path"
+#define CMD_PERF_DIR "--perf-dir="
+#define CMD_WORK_TREE "--work-tree="
+#define CMD_DEBUGFS_DIR "--debugfs-dir="
+
 #define PERF_DIR_ENVIRONMENT "PERF_DIR"
 #define PERF_WORK_TREE_ENVIRONMENT "PERF_WORK_TREE"
 #define DEFAULT_PERF_DIR_ENVIRONMENT ".perf"
@@ -117,4 +122,4 @@ extern char *perf_pathdup(const char *fmt, ...)
 
 extern size_t strlcpy(char *dest, const char *src, size_t size);
 
-#endif /* CACHE_H */
+#endif /* __PERF_CACHE_H */

tools/perf/util/callchain.c

@@ -206,7 +206,7 @@ fill_node(struct callchain_node *node, struct ip_callchain *chain,
 	}
 	node->val_nr = chain->nr - start;
 	if (!node->val_nr)
-		printf("Warning: empty node in callchain tree\n");
+		pr_warning("Warning: empty node in callchain tree\n");
 }
 
 static void

tools/perf/util/callchain.h

@@ -58,4 +58,4 @@ static inline u64 cumul_hits(struct callchain_node *node)
 int register_callchain_param(struct callchain_param *param);
 void append_chain(struct callchain_node *root, struct ip_callchain *chain,
 		  struct symbol **syms);
-#endif
+#endif	/* __PERF_CALLCHAIN_H */

tools/perf/util/color.h

-#ifndef COLOR_H
-#define COLOR_H
+#ifndef __PERF_COLOR_H
+#define __PERF_COLOR_H
 
 /* "\033[1;38;5;2xx;48;5;2xxm\0" is 23 bytes */
 #define COLOR_MAXLEN 24
@@ -39,4 +39,4 @@ int color_fwrite_lines(FILE *fp, const char *color, size_t count, const char *bu
 int percent_color_fprintf(FILE *fp, const char *fmt, double percent);
 const char *get_percent_color(double percent);
 
-#endif /* COLOR_H */
+#endif /* __PERF_COLOR_H */

tools/perf/util/data_map.h

+#ifndef __PERF_DATAMAP_H
+#define __PERF_DATAMAP_H
+
+#include "event.h"
+#include "header.h"
+
+typedef int (*event_type_handler_t)(event_t *, unsigned long, unsigned long);
+
+struct perf_file_handler {
+	event_type_handler_t	process_sample_event;
+	event_type_handler_t	process_mmap_event;
+	event_type_handler_t	process_comm_event;
+	event_type_handler_t	process_fork_event;
+	event_type_handler_t	process_exit_event;
+	event_type_handler_t	process_lost_event;
+	event_type_handler_t	process_read_event;
+	event_type_handler_t	process_throttle_event;
+	event_type_handler_t	process_unthrottle_event;
+	int			(*sample_type_check)(u64 sample_type);
+	unsigned long		total_unknown;
+};
+
+void register_perf_file_handler(struct perf_file_handler *handler);
+int mmap_dispatch_perf_file(struct perf_header **pheader,
+			    const char *input_name,
+			    int force,
+			    int full_paths,
+			    int *cwdlen,
+			    char **cwd);
+
+#endif

tools/perf/util/debug.c

@@ -13,12 +13,12 @@
 int verbose = 0;
 int dump_trace = 0;
 
-int eprintf(const char *fmt, ...)
+int eprintf(int level, const char *fmt, ...)
 {
 	va_list args;
 	int ret = 0;
 
-	if (verbose) {
+	if (verbose >= level) {
 		va_start(args, fmt);
 		ret = vfprintf(stderr, fmt, args);
 		va_end(args);

tools/perf/util/debug.h

 /* For debugging general purposes */
+#ifndef __PERF_DEBUG_H
+#define __PERF_DEBUG_H
 
 extern int verbose;
 extern int dump_trace;
 
-int eprintf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
+int eprintf(int level,
+	    const char *fmt, ...) __attribute__((format(printf, 2, 3)));
 int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
 void trace_event(event_t *event);
+
+#endif	/* __PERF_DEBUG_H */

tools/perf/util/exec_cmd.h

-#ifndef PERF_EXEC_CMD_H
-#define PERF_EXEC_CMD_H
+#ifndef __PERF_EXEC_CMD_H
+#define __PERF_EXEC_CMD_H
 
 extern void perf_set_argv_exec_path(const char *exec_path);
 extern const char *perf_extract_argv0_path(const char *path);
@@ -10,4 +10,4 @@ extern int execv_perf_cmd(const char **argv); /* NULL terminated */
 extern int execl_perf_cmd(const char *cmd, ...);
 extern const char *system_path(const char *path);
 
-#endif /* PERF_EXEC_CMD_H */
+#endif /* __PERF_EXEC_CMD_H */

tools/perf/util/help.h

-#ifndef HELP_H
-#define HELP_H
+#ifndef __PERF_HELP_H
+#define __PERF_HELP_H
 
 struct cmdnames {
 	size_t alloc;
@@ -26,4 +26,4 @@ int is_in_cmdlist(struct cmdnames *c, const char *s);
 void list_commands(const char *title, struct cmdnames *main_cmds,
 		   struct cmdnames *other_cmds);
 
-#endif /* HELP_H */
+#endif /* __PERF_HELP_H */

tools/perf/util/include/asm/asm-offsets.h

+/* stub */

tools/perf/util/include/asm/bitops.h

+#ifndef _PERF_ASM_BITOPS_H_
+#define _PERF_ASM_BITOPS_H_
+
+#include <sys/types.h>
+#include "../../types.h"
+#include <linux/compiler.h>
+
+/* CHECKME: Not sure both always match */
+#define BITS_PER_LONG	__WORDSIZE
+
+#include "../../../../include/asm-generic/bitops/__fls.h"
+#include "../../../../include/asm-generic/bitops/fls.h"
+#include "../../../../include/asm-generic/bitops/fls64.h"
+#include "../../../../include/asm-generic/bitops/__ffs.h"
+#include "../../../../include/asm-generic/bitops/ffz.h"
+#include "../../../../include/asm-generic/bitops/hweight.h"
+
+#endif

tools/perf/util/include/asm/byteorder.h

+#include <asm/types.h>
+#include "../../../../include/linux/swab.h"

tools/perf/util/include/asm/swab.h

+/* stub */

tools/perf/util/include/linux/bitmap.h

+#include "../../../../include/linux/bitmap.h"
+#include "../../../../include/asm-generic/bitops/find.h"

tools/perf/util/include/linux/ctype.h

+#include "../../../../include/linux/ctype.h"

tools/perf/util/include/linux/string.h

+#include <string.h>