Merge branch 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (106 commits)
  perf kvm: Fix copy & paste error in description
  perf script: Kill script_spec__delete
  perf top: Fix a memory leak
  perf stat: Introduce get_ratio_color() helper
  perf session: Remove impossible condition check
  perf tools: Fix feature-bits rework fallout, remove unused variable
  perf script: Add generic perl handler to process events
  perf tools: Use for_each_set_bit() to iterate over feature flags
  perf tools: Unify handling of features when writing feature section
  perf report: Accept fifos as input file
  perf tools: Moving code in some files
  perf tools: Fix out-of-bound access to struct perf_session
  perf tools: Continue processing header on unknown features
  perf tools: Improve macros for struct feature_ops
  perf: builtin-record: Document and check that mmap_pages must be a power of two.
  perf: builtin-record: Provide advice if mmap'ing fails with EPERM.
  perf tools: Fix truncated annotation
  perf script: look up thread using tid instead of pid
  perf tools: Look up thread names for system wide profiling
  perf tools: Fix comm for processes with named threads
  ...

Documentation/kernel-parameters.txt

@@ -1885,6 +1885,11 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
 			arch_perfmon: [X86] Force use of architectural
 				perfmon on Intel CPUs instead of the
 				CPU specific event set.
+			timer: [X86] Force use of architectural NMI
+				timer mode (see also oprofile.timer
+				for generic hr timer mode)
+				[s390] Force legacy basic mode sampling
+                                (report cpu_type "timer")
 
 	oops=panic	Always panic on oopses. Default is to just kill the
 			process, but there is a small probability of

Documentation/trace/events.txt

@@ -191,8 +191,6 @@ And for string fields they are:
 
 Currently, only exact string matches are supported.
 
-Currently, the maximum number of predicates in a filter is 16.
-
 5.2 Setting filters
 -------------------
 

arch/Kconfig

@@ -30,6 +30,10 @@ config OPROFILE_EVENT_MULTIPLEX
 config HAVE_OPROFILE
 	bool
 
+config OPROFILE_NMI_TIMER
+	def_bool y
+	depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI
+
 config KPROBES
 	bool "Kprobes"
 	depends on MODULES

arch/s390/oprofile/hwsampler.c

@@ -22,6 +22,7 @@
 #include <asm/irq.h>
 
 #include "hwsampler.h"
+#include "op_counter.h"
 
 #define MAX_NUM_SDB 511
 #define MIN_NUM_SDB 1
@@ -896,6 +897,8 @@ static void add_samples_to_oprofile(unsigned int cpu, unsigned long *sdbt,
 		if (sample_data_ptr->P == 1) {
 			/* userspace sample */
 			unsigned int pid = sample_data_ptr->prim_asn;
+			if (!counter_config.user)
+				goto skip_sample;
 			rcu_read_lock();
 			tsk = pid_task(find_vpid(pid), PIDTYPE_PID);
 			if (tsk)
@@ -903,6 +906,8 @@ static void add_samples_to_oprofile(unsigned int cpu, unsigned long *sdbt,
 			rcu_read_unlock();
 		} else {
 			/* kernelspace sample */
+			if (!counter_config.kernel)
+				goto skip_sample;
 			regs = task_pt_regs(current);
 		}
 
@@ -910,7 +915,7 @@ static void add_samples_to_oprofile(unsigned int cpu, unsigned long *sdbt,
 		oprofile_add_ext_hw_sample(sample_data_ptr->ia, regs, 0,
 				!sample_data_ptr->P, tsk);
 		mutex_unlock(&hws_sem);
-
+	skip_sample:
 		sample_data_ptr++;
 	}
 }

arch/s390/oprofile/init.c

@@ -2,10 +2,11 @@
  * arch/s390/oprofile/init.c
  *
  * S390 Version
- *   Copyright (C) 2003 IBM Deutschland Entwicklung GmbH, IBM Corporation
+ *   Copyright (C) 2002-2011 IBM Deutschland Entwicklung GmbH, IBM Corporation
  *   Author(s): Thomas Spatzier (tspat@de.ibm.com)
  *   Author(s): Mahesh Salgaonkar (mahesh@linux.vnet.ibm.com)
  *   Author(s): Heinz Graalfs (graalfs@linux.vnet.ibm.com)
+ *   Author(s): Andreas Krebbel (krebbel@linux.vnet.ibm.com)
  *
  * @remark Copyright 2002-2011 OProfile authors
  */
@@ -14,6 +15,8 @@
 #include <linux/init.h>
 #include <linux/errno.h>
 #include <linux/fs.h>
+#include <linux/module.h>
+#include <asm/processor.h>
 
 #include "../../../drivers/oprofile/oprof.h"
 
@@ -22,6 +25,7 @@ extern void s390_backtrace(struct pt_regs * const regs, unsigned int depth);
 #ifdef CONFIG_64BIT
 
 #include "hwsampler.h"
+#include "op_counter.h"
 
 #define DEFAULT_INTERVAL	4127518
 
@@ -35,16 +39,41 @@ static unsigned long oprofile_max_interval;
 static unsigned long oprofile_sdbt_blocks = DEFAULT_SDBT_BLOCKS;
 static unsigned long oprofile_sdb_blocks = DEFAULT_SDB_BLOCKS;
 
-static int hwsampler_file;
+static int hwsampler_enabled;
 static int hwsampler_running;	/* start_mutex must be held to change */
+static int hwsampler_available;
 
 static struct oprofile_operations timer_ops;
 
+struct op_counter_config counter_config;
+
+enum __force_cpu_type {
+	reserved = 0,		/* do not force */
+	timer,
+};
+static int force_cpu_type;
+
+static int set_cpu_type(const char *str, struct kernel_param *kp)
+{
+	if (!strcmp(str, "timer")) {
+		force_cpu_type = timer;
+		printk(KERN_INFO "oprofile: forcing timer to be returned "
+		                 "as cpu type\n");
+	} else {
+		force_cpu_type = 0;
+	}
+
+	return 0;
+}
+module_param_call(cpu_type, set_cpu_type, NULL, NULL, 0);
+MODULE_PARM_DESC(cpu_type, "Force legacy basic mode sampling"
+		           "(report cpu_type \"timer\"");
+
 static int oprofile_hwsampler_start(void)
 {
 	int retval;
 
-	hwsampler_running = hwsampler_file;
+	hwsampler_running = hwsampler_enabled;
 
 	if (!hwsampler_running)
 		return timer_ops.start();
@@ -72,10 +101,16 @@ static void oprofile_hwsampler_stop(void)
 	return;
 }
 
+/*
+ * File ops used for:
+ * /dev/oprofile/0/enabled
+ * /dev/oprofile/hwsampling/hwsampler  (cpu_type = timer)
+ */
+
 static ssize_t hwsampler_read(struct file *file, char __user *buf,
 		size_t count, loff_t *offset)
 {
-	return oprofilefs_ulong_to_user(hwsampler_file, buf, count, offset);
+	return oprofilefs_ulong_to_user(hwsampler_enabled, buf, count, offset);
 }
 
 static ssize_t hwsampler_write(struct file *file, char const __user *buf,
@@ -91,6 +126,9 @@ static ssize_t hwsampler_write(struct file *file, char const __user *buf,
 	if (retval <= 0)
 		return retval;
 
+	if (val != 0 && val != 1)
+		return -EINVAL;
+
 	if (oprofile_started)
 		/*
 		 * save to do without locking as we set
@@ -99,7 +137,7 @@ static ssize_t hwsampler_write(struct file *file, char const __user *buf,
 		 */
 		return -EBUSY;
 
-	hwsampler_file = val;
+	hwsampler_enabled = val;
 
 	return count;
 }
@@ -109,38 +147,311 @@ static const struct file_operations hwsampler_fops = {
 	.write		= hwsampler_write,
 };
 
+/*
+ * File ops used for:
+ * /dev/oprofile/0/count
+ * /dev/oprofile/hwsampling/hw_interval  (cpu_type = timer)
+ *
+ * Make sure that the value is within the hardware range.
+ */
+
+static ssize_t hw_interval_read(struct file *file, char __user *buf,
+				size_t count, loff_t *offset)
+{
+	return oprofilefs_ulong_to_user(oprofile_hw_interval, buf,
+					count, offset);
+}
+
+static ssize_t hw_interval_write(struct file *file, char const __user *buf,
+				 size_t count, loff_t *offset)
+{
+	unsigned long val;
+	int retval;
+
+	if (*offset)
+		return -EINVAL;
+	retval = oprofilefs_ulong_from_user(&val, buf, count);
+	if (retval)
+		return retval;
+	if (val < oprofile_min_interval)
+		oprofile_hw_interval = oprofile_min_interval;
+	else if (val > oprofile_max_interval)
+		oprofile_hw_interval = oprofile_max_interval;
+	else
+		oprofile_hw_interval = val;
+
+	return count;
+}
+
+static const struct file_operations hw_interval_fops = {
+	.read		= hw_interval_read,
+	.write		= hw_interval_write,
+};
+
+/*
+ * File ops used for:
+ * /dev/oprofile/0/event
+ * Only a single event with number 0 is supported with this counter.
+ *
+ * /dev/oprofile/0/unit_mask
+ * This is a dummy file needed by the user space tools.
+ * No value other than 0 is accepted or returned.
+ */
+
+static ssize_t hwsampler_zero_read(struct file *file, char __user *buf,
+				    size_t count, loff_t *offset)
+{
+	return oprofilefs_ulong_to_user(0, buf, count, offset);
+}
+
+static ssize_t hwsampler_zero_write(struct file *file, char const __user *buf,
+				     size_t count, loff_t *offset)
+{
+	unsigned long val;
+	int retval;
+
+	if (*offset)
+		return -EINVAL;
+
+	retval = oprofilefs_ulong_from_user(&val, buf, count);
+	if (retval)
+		return retval;
+	if (val != 0)
+		return -EINVAL;
+	return count;
+}
+
+static const struct file_operations zero_fops = {
+	.read		= hwsampler_zero_read,
+	.write		= hwsampler_zero_write,
+};
+
+/* /dev/oprofile/0/kernel file ops.  */
+
+static ssize_t hwsampler_kernel_read(struct file *file, char __user *buf,
+				     size_t count, loff_t *offset)
+{
+	return oprofilefs_ulong_to_user(counter_config.kernel,
+					buf, count, offset);
+}
+
+static ssize_t hwsampler_kernel_write(struct file *file, char const __user *buf,
+				      size_t count, loff_t *offset)
+{
+	unsigned long val;
+	int retval;
+
+	if (*offset)
+		return -EINVAL;
+
+	retval = oprofilefs_ulong_from_user(&val, buf, count);
+	if (retval)
+		return retval;
+
+	if (val != 0 && val != 1)
+		return -EINVAL;
+
+	counter_config.kernel = val;
+
+	return count;
+}
+
+static const struct file_operations kernel_fops = {
+	.read		= hwsampler_kernel_read,
+	.write		= hwsampler_kernel_write,
+};
+
+/* /dev/oprofile/0/user file ops. */
+
+static ssize_t hwsampler_user_read(struct file *file, char __user *buf,
+				   size_t count, loff_t *offset)
+{
+	return oprofilefs_ulong_to_user(counter_config.user,
+					buf, count, offset);
+}
+
+static ssize_t hwsampler_user_write(struct file *file, char const __user *buf,
+				    size_t count, loff_t *offset)
+{
+	unsigned long val;
+	int retval;
+
+	if (*offset)
+		return -EINVAL;
+
+	retval = oprofilefs_ulong_from_user(&val, buf, count);
+	if (retval)
+		return retval;
+
+	if (val != 0 && val != 1)
+		return -EINVAL;
+
+	counter_config.user = val;
+
+	return count;
+}
+
+static const struct file_operations user_fops = {
+	.read		= hwsampler_user_read,
+	.write		= hwsampler_user_write,
+};
+
+
+/*
+ * File ops used for: /dev/oprofile/timer/enabled
+ * The value always has to be the inverted value of hwsampler_enabled. So
+ * no separate variable is created. That way we do not need locking.
+ */
+
+static ssize_t timer_enabled_read(struct file *file, char __user *buf,
+				  size_t count, loff_t *offset)
+{
+	return oprofilefs_ulong_to_user(!hwsampler_enabled, buf, count, offset);
+}
+
+static ssize_t timer_enabled_write(struct file *file, char const __user *buf,
+				   size_t count, loff_t *offset)
+{
+	unsigned long val;
+	int retval;
+
+	if (*offset)
+		return -EINVAL;
+
+	retval = oprofilefs_ulong_from_user(&val, buf, count);
+	if (retval)
+		return retval;
+
+	if (val != 0 && val != 1)
+		return -EINVAL;
+
+	/* Timer cannot be disabled without having hardware sampling.  */
+	if (val == 0 && !hwsampler_available)
+		return -EINVAL;
+
+	if (oprofile_started)
+		/*
+		 * save to do without locking as we set
+		 * hwsampler_running in start() when start_mutex is
+		 * held
+		 */
+		return -EBUSY;
+
+	hwsampler_enabled = !val;
+
+	return count;
+}
+
+static const struct file_operations timer_enabled_fops = {
+	.read		= timer_enabled_read,
+	.write		= timer_enabled_write,
+};
+
+
 static int oprofile_create_hwsampling_files(struct super_block *sb,
-						struct dentry *root)
+					    struct dentry *root)
 {
-	struct dentry *hw_dir;
+	struct dentry *dir;
+
+	dir = oprofilefs_mkdir(sb, root, "timer");
+	if (!dir)
+		return -EINVAL;
+
+	oprofilefs_create_file(sb, dir, "enabled", &timer_enabled_fops);
+
+	if (!hwsampler_available)
+		return 0;
 
 	/* reinitialize default values */
-	hwsampler_file = 1;
+	hwsampler_enabled = 1;
+	counter_config.kernel = 1;
+	counter_config.user = 1;
 
-	hw_dir = oprofilefs_mkdir(sb, root, "hwsampling");
-	if (!hw_dir)
-		return -EINVAL;
+	if (!force_cpu_type) {
+		/*
+		 * Create the counter file system.  A single virtual
+		 * counter is created which can be used to
+		 * enable/disable hardware sampling dynamically from
+		 * user space.  The user space will configure a single
+		 * counter with a single event.  The value of 'event'
+		 * and 'unit_mask' are not evaluated by the kernel code
+		 * and can only be set to 0.
+		 */
+
+		dir = oprofilefs_mkdir(sb, root, "0");
+		if (!dir)
+			return -EINVAL;
 
-	oprofilefs_create_file(sb, hw_dir, "hwsampler", &hwsampler_fops);
-	oprofilefs_create_ulong(sb, hw_dir, "hw_interval",
-				&oprofile_hw_interval);
-	oprofilefs_create_ro_ulong(sb, hw_dir, "hw_min_interval",
-				&oprofile_min_interval);
-	oprofilefs_create_ro_ulong(sb, hw_dir, "hw_max_interval",
-				&oprofile_max_interval);
-	oprofilefs_create_ulong(sb, hw_dir, "hw_sdbt_blocks",
-				&oprofile_sdbt_blocks);
+		oprofilefs_create_file(sb, dir, "enabled", &hwsampler_fops);
+		oprofilefs_create_file(sb, dir, "event", &zero_fops);
+		oprofilefs_create_file(sb, dir, "count", &hw_interval_fops);
+		oprofilefs_create_file(sb, dir, "unit_mask", &zero_fops);
+		oprofilefs_create_file(sb, dir, "kernel", &kernel_fops);
+		oprofilefs_create_file(sb, dir, "user", &user_fops);
+		oprofilefs_create_ulong(sb, dir, "hw_sdbt_blocks",
+					&oprofile_sdbt_blocks);
 
+	} else {
+		/*
+		 * Hardware sampling can be used but the cpu_type is
+		 * forced to timer in order to deal with legacy user
+		 * space tools.  The /dev/oprofile/hwsampling fs is
+		 * provided in that case.
+		 */
+		dir = oprofilefs_mkdir(sb, root, "hwsampling");
+		if (!dir)
+			return -EINVAL;
+
+		oprofilefs_create_file(sb, dir, "hwsampler",
+				       &hwsampler_fops);
+		oprofilefs_create_file(sb, dir, "hw_interval",
+				       &hw_interval_fops);
+		oprofilefs_create_ro_ulong(sb, dir, "hw_min_interval",
+					   &oprofile_min_interval);
+		oprofilefs_create_ro_ulong(sb, dir, "hw_max_interval",
+					   &oprofile_max_interval);
+		oprofilefs_create_ulong(sb, dir, "hw_sdbt_blocks",
+					&oprofile_sdbt_blocks);
+	}
 	return 0;
 }
 
 static int oprofile_hwsampler_init(struct oprofile_operations *ops)
 {
+	/*
+	 * Initialize the timer mode infrastructure as well in order
+	 * to be able to switch back dynamically.  oprofile_timer_init
+	 * is not supposed to fail.
+	 */
+	if (oprofile_timer_init(ops))
+		BUG();
+
+	memcpy(&timer_ops, ops, sizeof(timer_ops));
+	ops->create_files = oprofile_create_hwsampling_files;
+
+	/*
+	 * If the user space tools do not support newer cpu types,
+	 * the force_cpu_type module parameter
+	 * can be used to always return \"timer\" as cpu type.
+	 */
+	if (force_cpu_type != timer) {
+		struct cpuid id;
+
+		get_cpu_id (&id);
+
+		switch (id.machine) {
+		case 0x2097: case 0x2098: ops->cpu_type = "s390/z10"; break;
+		case 0x2817: case 0x2818: ops->cpu_type = "s390/z196"; break;
+		default: return -ENODEV;
+		}
+	}
+
 	if (hwsampler_setup())
 		return -ENODEV;
 
 	/*
-	 * create hwsampler files only if hwsampler_setup() succeeds.
+	 * Query the range for the sampling interval from the
+	 * hardware.
 	 */
 	oprofile_min_interval = hwsampler_query_min_interval();
 	if (oprofile_min_interval == 0)
@@ -155,23 +466,17 @@ static int oprofile_hwsampler_init(struct oprofile_operations *ops)
 	if (oprofile_hw_interval > oprofile_max_interval)
 		oprofile_hw_interval = oprofile_max_interval;
 
-	if (oprofile_timer_init(ops))
-		return -ENODEV;
-
-	printk(KERN_INFO "oprofile: using hardware sampling\n");
-
-	memcpy(&timer_ops, ops, sizeof(timer_ops));
+	printk(KERN_INFO "oprofile: System z hardware sampling "
+	       "facility found.\n");
 
 	ops->start = oprofile_hwsampler_start;
 	ops->stop = oprofile_hwsampler_stop;
-	ops->create_files = oprofile_create_hwsampling_files;
 
 	return 0;
 }
 
 static void oprofile_hwsampler_exit(void)
 {
-	oprofile_timer_exit();
 	hwsampler_shutdown();
 }
 
@@ -182,7 +487,15 @@ int __init oprofile_arch_init(struct oprofile_operations *ops)
 	ops->backtrace = s390_backtrace;
 
 #ifdef CONFIG_64BIT
-	return oprofile_hwsampler_init(ops);
+
+	/*
+	 * -ENODEV is not reported to the caller.  The module itself
+         * will use the timer mode sampling as fallback and this is
+         * always available.
+	 */
+	hwsampler_available = oprofile_hwsampler_init(ops) == 0;
+
+	return 0;
 #else
 	return -ENODEV;
 #endif

arch/s390/oprofile/op_counter.h

@@ -0,0 +1,23 @@
+/**
+ * arch/s390/oprofile/op_counter.h
+ *
+ *   Copyright (C) 2011 IBM Deutschland Entwicklung GmbH, IBM Corporation
+ *   Author(s): Andreas Krebbel (krebbel@linux.vnet.ibm.com)
+ *
+ * @remark Copyright 2011 OProfile authors
+ */
+
+#ifndef OP_COUNTER_H
+#define OP_COUNTER_H
+
+struct op_counter_config {
+	/* `enabled' maps to the hwsampler_file variable.  */
+	/* `count' maps to the oprofile_hw_interval variable.  */
+	/* `event' and `unit_mask' are unused. */
+	unsigned long kernel;
+	unsigned long user;
+};
+
+extern struct op_counter_config counter_config;
+
+#endif /* OP_COUNTER_H */

arch/x86/include/asm/insn.h

@@ -137,6 +137,13 @@ static inline int insn_is_avx(struct insn *insn)
 	return (insn->vex_prefix.value != 0);
 }
 
+/* Ensure this instruction is decoded completely */
+static inline int insn_complete(struct insn *insn)
+{
+	return insn->opcode.got && insn->modrm.got && insn->sib.got &&
+		insn->displacement.got && insn->immediate.got;
+}
+
 static inline insn_byte_t insn_vex_m_bits(struct insn *insn)
 {
 	if (insn->vex_prefix.nbytes == 2)	/* 2 bytes VEX */

arch/x86/include/asm/perf_event.h

@@ -57,6 +57,7 @@
 		(1 << (ARCH_PERFMON_UNHALTED_CORE_CYCLES_INDEX))
 
 #define ARCH_PERFMON_BRANCH_MISSES_RETIRED		6
+#define ARCH_PERFMON_EVENTS_COUNT			7
 
 /*
  * Intel "Architectural Performance Monitoring" CPUID
@@ -72,6 +73,19 @@ union cpuid10_eax {
 	unsigned int full;
 };
 
+union cpuid10_ebx {
+	struct {
+		unsigned int no_unhalted_core_cycles:1;
+		unsigned int no_instructions_retired:1;
+		unsigned int no_unhalted_reference_cycles:1;
+		unsigned int no_llc_reference:1;
+		unsigned int no_llc_misses:1;
+		unsigned int no_branch_instruction_retired:1;
+		unsigned int no_branch_misses_retired:1;
+	} split;
+	unsigned int full;
+};
+
 union cpuid10_edx {
 	struct {
 		unsigned int num_counters_fixed:5;
@@ -81,6 +95,15 @@ union cpuid10_edx {
 	unsigned int full;
 };
 
+struct x86_pmu_capability {
+	int		version;
+	int		num_counters_gp;
+	int		num_counters_fixed;
+	int		bit_width_gp;
+	int		bit_width_fixed;
+	unsigned int	events_mask;
+	int		events_mask_len;
+};
 
 /*
  * Fixed-purpose performance events:
@@ -89,23 +112,24 @@ union cpuid10_edx {
 /*
  * All 3 fixed-mode PMCs are configured via this single MSR:
  */
-#define MSR_ARCH_PERFMON_FIXED_CTR_CTRL			0x38d
+#define MSR_ARCH_PERFMON_FIXED_CTR_CTRL	0x38d
 
 /*
  * The counts are available in three separate MSRs:
  */
 
 /* Instr_Retired.Any: */
-#define MSR_ARCH_PERFMON_FIXED_CTR0			0x309
-#define X86_PMC_IDX_FIXED_INSTRUCTIONS			(X86_PMC_IDX_FIXED + 0)
+#define MSR_ARCH_PERFMON_FIXED_CTR0	0x309
+#define X86_PMC_IDX_FIXED_INSTRUCTIONS	(X86_PMC_IDX_FIXED + 0)
 
 /* CPU_CLK_Unhalted.Core: */
-#define MSR_ARCH_PERFMON_FIXED_CTR1			0x30a
-#define X86_PMC_IDX_FIXED_CPU_CYCLES			(X86_PMC_IDX_FIXED + 1)
+#define MSR_ARCH_PERFMON_FIXED_CTR1	0x30a
+#define X86_PMC_IDX_FIXED_CPU_CYCLES	(X86_PMC_IDX_FIXED + 1)
 
 /* CPU_CLK_Unhalted.Ref: */
-#define MSR_ARCH_PERFMON_FIXED_CTR2			0x30b
-#define X86_PMC_IDX_FIXED_BUS_CYCLES			(X86_PMC_IDX_FIXED + 2)
+#define MSR_ARCH_PERFMON_FIXED_CTR2	0x30b
+#define X86_PMC_IDX_FIXED_REF_CYCLES	(X86_PMC_IDX_FIXED + 2)
+#define X86_PMC_MSK_FIXED_REF_CYCLES	(1ULL << X86_PMC_IDX_FIXED_REF_CYCLES)
 
 /*
  * We model BTS tracing as another fixed-mode PMC.
@@ -202,6 +226,7 @@ struct perf_guest_switch_msr {
 };
 
 extern struct perf_guest_switch_msr *perf_guest_get_msrs(int *nr);
+extern void perf_get_x86_pmu_capability(struct x86_pmu_capability *cap);
 #else
 static inline perf_guest_switch_msr *perf_guest_get_msrs(int *nr)
 {
@@ -209,6 +234,11 @@ static inline perf_guest_switch_msr *perf_guest_get_msrs(int *nr)
 	return NULL;
 }
 
+static inline void perf_get_x86_pmu_capability(struct x86_pmu_capability *cap)
+{
+	memset(cap, 0, sizeof(*cap));
+}
+
 static inline void perf_events_lapic_init(void)	{ }
 #endif
 

arch/x86/kernel/cpu/perf_event.c

@@ -484,18 +484,195 @@ static inline int is_x86_event(struct perf_event *event)
 	return event->pmu == &pmu;
 }
 
+/*
+ * Event scheduler state:
+ *
+ * Assign events iterating over all events and counters, beginning
+ * with events with least weights first. Keep the current iterator
+ * state in struct sched_state.
+ */
+struct sched_state {
+	int	weight;
+	int	event;		/* event index */
+	int	counter;	/* counter index */
+	int	unassigned;	/* number of events to be assigned left */
+	unsigned long used[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
+};
+
+/* Total max is X86_PMC_IDX_MAX, but we are O(n!) limited */
+#define	SCHED_STATES_MAX	2
+
+struct perf_sched {
+	int			max_weight;
+	int			max_events;
+	struct event_constraint	**constraints;
+	struct sched_state	state;
+	int			saved_states;
+	struct sched_state	saved[SCHED_STATES_MAX];
+};
+
+/*
+ * Initialize interator that runs through all events and counters.
+ */
+static void perf_sched_init(struct perf_sched *sched, struct event_constraint **c,
+			    int num, int wmin, int wmax)
+{
+	int idx;
+
+	memset(sched, 0, sizeof(*sched));
+	sched->max_events	= num;
+	sched->max_weight	= wmax;
+	sched->constraints	= c;
+
+	for (idx = 0; idx < num; idx++) {
+		if (c[idx]->weight == wmin)
+			break;
+	}
+
+	sched->state.event	= idx;		/* start with min weight */
+	sched->state.weight	= wmin;
+	sched->state.unassigned	= num;
+}
+
+static void perf_sched_save_state(struct perf_sched *sched)
+{
+	if (WARN_ON_ONCE(sched->saved_states >= SCHED_STATES_MAX))
+		return;
+
+	sched->saved[sched->saved_states] = sched->state;
+	sched->saved_states++;
+}
+
+static bool perf_sched_restore_state(struct perf_sched *sched)
+{
+	if (!sched->saved_states)
+		return false;
+
+	sched->saved_states--;
+	sched->state = sched->saved[sched->saved_states];
+
+	/* continue with next counter: */
+	clear_bit(sched->state.counter++, sched->state.used);
+
+	return true;
+}
+
+/*
+ * Select a counter for the current event to schedule. Return true on
+ * success.
+ */
+static bool __perf_sched_find_counter(struct perf_sched *sched)
+{
+	struct event_constraint *c;
+	int idx;
+
+	if (!sched->state.unassigned)
+		return false;
+
+	if (sched->state.event >= sched->max_events)
+		return false;
+
+	c = sched->constraints[sched->state.event];
+
+	/* Prefer fixed purpose counters */
+	if (x86_pmu.num_counters_fixed) {
+		idx = X86_PMC_IDX_FIXED;
+		for_each_set_bit_cont(idx, c->idxmsk, X86_PMC_IDX_MAX) {
+			if (!__test_and_set_bit(idx, sched->state.used))
+				goto done;
+		}
+	}
+	/* Grab the first unused counter starting with idx */
+	idx = sched->state.counter;
+	for_each_set_bit_cont(idx, c->idxmsk, X86_PMC_IDX_FIXED) {
+		if (!__test_and_set_bit(idx, sched->state.used))
+			goto done;
+	}
+
+	return false;
+
+done:
+	sched->state.counter = idx;
+
+	if (c->overlap)
+		perf_sched_save_state(sched);
+
+	return true;
+}
+
+static bool perf_sched_find_counter(struct perf_sched *sched)
+{
+	while (!__perf_sched_find_counter(sched)) {
+		if (!perf_sched_restore_state(sched))
+			return false;
+	}
+
+	return true;
+}
+
+/*
+ * Go through all unassigned events and find the next one to schedule.
+ * Take events with the least weight first. Return true on success.
+ */
+static bool perf_sched_next_event(struct perf_sched *sched)
+{
+	struct event_constraint *c;
+
+	if (!sched->state.unassigned || !--sched->state.unassigned)
+		return false;
+
+	do {
+		/* next event */
+		sched->state.event++;
+		if (sched->state.event >= sched->max_events) {
+			/* next weight */
+			sched->state.event = 0;
+			sched->state.weight++;
+			if (sched->state.weight > sched->max_weight)
+				return false;
+		}
+		c = sched->constraints[sched->state.event];
+	} while (c->weight != sched->state.weight);
+
+	sched->state.counter = 0;	/* start with first counter */
+
+	return true;
+}
+
+/*
+ * Assign a counter for each event.
+ */
+static int perf_assign_events(struct event_constraint **constraints, int n,
+			      int wmin, int wmax, int *assign)
+{
+	struct perf_sched sched;
+
+	perf_sched_init(&sched, constraints, n, wmin, wmax);
+
+	do {
+		if (!perf_sched_find_counter(&sched))
+			break;	/* failed */
+		if (assign)
+			assign[sched.state.event] = sched.state.counter;
+	} while (perf_sched_next_event(&sched));
+
+	return sched.state.unassigned;
+}
+
 int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign)
 {
 	struct event_constraint *c, *constraints[X86_PMC_IDX_MAX];
 	unsigned long used_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
-	int i, j, w, wmax, num = 0;
+	int i, wmin, wmax, num = 0;
 	struct hw_perf_event *hwc;
 
 	bitmap_zero(used_mask, X86_PMC_IDX_MAX);
 
-	for (i = 0; i < n; i++) {
+	for (i = 0, wmin = X86_PMC_IDX_MAX, wmax = 0; i < n; i++) {
 		c = x86_pmu.get_event_constraints(cpuc, cpuc->event_list[i]);
 		constraints[i] = c;
+		wmin = min(wmin, c->weight);
+		wmax = max(wmax, c->weight);
 	}
 
 	/*
@@ -521,59 +698,11 @@ int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign)
 		if (assign)
 			assign[i] = hwc->idx;
 	}
-	if (i == n)
-		goto done;
-
-	/*
-	 * begin slow path
-	 */
-
-	bitmap_zero(used_mask, X86_PMC_IDX_MAX);
 
-	/*
-	 * weight = number of possible counters
-	 *
-	 * 1    = most constrained, only works on one counter
-	 * wmax = least constrained, works on any counter
-	 *
-	 * assign events to counters starting with most
-	 * constrained events.
-	 */
-	wmax = x86_pmu.num_counters;
+	/* slow path */
+	if (i != n)
+		num = perf_assign_events(constraints, n, wmin, wmax, assign);
 
-	/*
-	 * when fixed event counters are present,
-	 * wmax is incremented by 1 to account
-	 * for one more choice
-	 */
-	if (x86_pmu.num_counters_fixed)
-		wmax++;
-
-	for (w = 1, num = n; num && w <= wmax; w++) {
-		/* for each event */
-		for (i = 0; num && i < n; i++) {
-			c = constraints[i];
-			hwc = &cpuc->event_list[i]->hw;
-
-			if (c->weight != w)
-				continue;
-
-			for_each_set_bit(j, c->idxmsk, X86_PMC_IDX_MAX) {
-				if (!test_bit(j, used_mask))
-					break;
-			}
-
-			if (j == X86_PMC_IDX_MAX)
-				break;
-
-			__set_bit(j, used_mask);
-
-			if (assign)
-				assign[i] = j;
-			num--;
-		}
-	}
-done:
 	/*
 	 * scheduling failed or is just a simulation,
 	 * free resources if necessary
@@ -1119,6 +1248,7 @@ static void __init pmu_check_apic(void)
 
 static int __init init_hw_perf_events(void)
 {
+	struct x86_pmu_quirk *quirk;
 	struct event_constraint *c;
 	int err;
 
@@ -1147,8 +1277,8 @@ static int __init init_hw_perf_events(void)
 
 	pr_cont("%s PMU driver.\n", x86_pmu.name);
 
-	if (x86_pmu.quirks)
-		x86_pmu.quirks();
+	for (quirk = x86_pmu.quirks; quirk; quirk = quirk->next)
+		quirk->func();
 
 	if (x86_pmu.num_counters > X86_PMC_MAX_GENERIC) {
 		WARN(1, KERN_ERR "hw perf events %d > max(%d), clipping!",
@@ -1171,12 +1301,18 @@ static int __init init_hw_perf_events(void)
 
 	unconstrained = (struct event_constraint)
 		__EVENT_CONSTRAINT(0, (1ULL << x86_pmu.num_counters) - 1,
-				   0, x86_pmu.num_counters);
+				   0, x86_pmu.num_counters, 0);
 
 	if (x86_pmu.event_constraints) {
+		/*
+		 * event on fixed counter2 (REF_CYCLES) only works on this
+		 * counter, so do not extend mask to generic counters
+		 */
 		for_each_event_constraint(c, x86_pmu.event_constraints) {
-			if (c->cmask != X86_RAW_EVENT_MASK)
+			if (c->cmask != X86_RAW_EVENT_MASK
+			    || c->idxmsk64 == X86_PMC_MSK_FIXED_REF_CYCLES) {
 				continue;
+			}
 
 			c->idxmsk64 |= (1ULL << x86_pmu.num_counters) - 1;
 			c->weight += x86_pmu.num_counters;
@@ -1566,3 +1702,15 @@ unsigned long perf_misc_flags(struct pt_regs *regs)
 
 	return misc;
 }
+
+void perf_get_x86_pmu_capability(struct x86_pmu_capability *cap)
+{
+	cap->version		= x86_pmu.version;
+	cap->num_counters_gp	= x86_pmu.num_counters;
+	cap->num_counters_fixed	= x86_pmu.num_counters_fixed;
+	cap->bit_width_gp	= x86_pmu.cntval_bits;
+	cap->bit_width_fixed	= x86_pmu.cntval_bits;
+	cap->events_mask	= (unsigned int)x86_pmu.events_maskl;
+	cap->events_mask_len	= x86_pmu.events_mask_len;
+}
+EXPORT_SYMBOL_GPL(perf_get_x86_pmu_capability);

arch/x86/kernel/cpu/perf_event.h

@@ -45,6 +45,7 @@ struct event_constraint {
 	u64	code;
 	u64	cmask;
 	int	weight;
+	int	overlap;
 };
 
 struct amd_nb {
@@ -151,15 +152,40 @@ struct cpu_hw_events {
 	void				*kfree_on_online;
 };
 
-#define __EVENT_CONSTRAINT(c, n, m, w) {\
+#define __EVENT_CONSTRAINT(c, n, m, w, o) {\
 	{ .idxmsk64 = (n) },		\
 	.code = (c),			\
 	.cmask = (m),			\
 	.weight = (w),			\
+	.overlap = (o),			\
 }
 
 #define EVENT_CONSTRAINT(c, n, m)	\
-	__EVENT_CONSTRAINT(c, n, m, HWEIGHT(n))
+	__EVENT_CONSTRAINT(c, n, m, HWEIGHT(n), 0)
+
+/*
+ * The overlap flag marks event constraints with overlapping counter
+ * masks. This is the case if the counter mask of such an event is not
+ * a subset of any other counter mask of a constraint with an equal or
+ * higher weight, e.g.:
+ *
+ *  c_overlaps = EVENT_CONSTRAINT_OVERLAP(0, 0x09, 0);
+ *  c_another1 = EVENT_CONSTRAINT(0, 0x07, 0);
+ *  c_another2 = EVENT_CONSTRAINT(0, 0x38, 0);
+ *
+ * The event scheduler may not select the correct counter in the first
+ * cycle because it needs to know which subsequent events will be
+ * scheduled. It may fail to schedule the events then. So we set the
+ * overlap flag for such constraints to give the scheduler a hint which
+ * events to select for counter rescheduling.
+ *
+ * Care must be taken as the rescheduling algorithm is O(n!) which
+ * will increase scheduling cycles for an over-commited system
+ * dramatically.  The number of such EVENT_CONSTRAINT_OVERLAP() macros
+ * and its counter masks must be kept at a minimum.
+ */
+#define EVENT_CONSTRAINT_OVERLAP(c, n, m)	\
+	__EVENT_CONSTRAINT(c, n, m, HWEIGHT(n), 1)
 
 /*
  * Constraint on the Event code.
@@ -235,6 +261,11 @@ union perf_capabilities {
 	u64	capabilities;
 };
 
+struct x86_pmu_quirk {
+	struct x86_pmu_quirk *next;
+	void (*func)(void);
+};
+
 /*
  * struct x86_pmu - generic x86 pmu
  */
@@ -259,6 +290,11 @@ struct x86_pmu {
 	int		num_counters_fixed;
 	int		cntval_bits;
 	u64		cntval_mask;
+	union {
+			unsigned long events_maskl;
+			unsigned long events_mask[BITS_TO_LONGS(ARCH_PERFMON_EVENTS_COUNT)];
+	};
+	int		events_mask_len;
 	int		apic;
 	u64		max_period;
 	struct event_constraint *
@@ -268,7 +304,7 @@ struct x86_pmu {
 	void		(*put_event_constraints)(struct cpu_hw_events *cpuc,
 						 struct perf_event *event);
 	struct event_constraint *event_constraints;
-	void		(*quirks)(void);
+	struct x86_pmu_quirk *quirks;
 	int		perfctr_second_write;
 
 	int		(*cpu_prepare)(int cpu);
@@ -309,6 +345,15 @@ struct x86_pmu {
 	struct perf_guest_switch_msr *(*guest_get_msrs)(int *nr);
 };
 
+#define x86_add_quirk(func_)						\
+do {									\
+	static struct x86_pmu_quirk __quirk __initdata = {		\
+		.func = func_,						\
+	};								\
+	__quirk.next = x86_pmu.quirks;					\
+	x86_pmu.quirks = &__quirk;					\
+} while (0)
+
 #define ERF_NO_HT_SHARING	1
 #define ERF_HAS_RSP_1		2
 

arch/x86/kernel/cpu/perf_event_amd.c

@@ -492,7 +492,7 @@ static __initconst const struct x86_pmu amd_pmu = {
 static struct event_constraint amd_f15_PMC0  = EVENT_CONSTRAINT(0, 0x01, 0);
 static struct event_constraint amd_f15_PMC20 = EVENT_CONSTRAINT(0, 0x07, 0);
 static struct event_constraint amd_f15_PMC3  = EVENT_CONSTRAINT(0, 0x08, 0);
-static struct event_constraint amd_f15_PMC30 = EVENT_CONSTRAINT(0, 0x09, 0);
+static struct event_constraint amd_f15_PMC30 = EVENT_CONSTRAINT_OVERLAP(0, 0x09, 0);
 static struct event_constraint amd_f15_PMC50 = EVENT_CONSTRAINT(0, 0x3F, 0);
 static struct event_constraint amd_f15_PMC53 = EVENT_CONSTRAINT(0, 0x38, 0);
 

arch/x86/kernel/cpu/perf_event_intel.c

@@ -28,6 +28,7 @@ static u64 intel_perfmon_event_map[PERF_COUNT_HW_MAX] __read_mostly =
   [PERF_COUNT_HW_BRANCH_INSTRUCTIONS]	= 0x00c4,
   [PERF_COUNT_HW_BRANCH_MISSES]		= 0x00c5,
   [PERF_COUNT_HW_BUS_CYCLES]		= 0x013c,
+  [PERF_COUNT_HW_REF_CPU_CYCLES]	= 0x0300, /* pseudo-encoding */
 };
 
 static struct event_constraint intel_core_event_constraints[] __read_mostly =
@@ -45,12 +46,7 @@ static struct event_constraint intel_core2_event_constraints[] __read_mostly =
 {
 	FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */
 	FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */
-	/*
-	 * Core2 has Fixed Counter 2 listed as CPU_CLK_UNHALTED.REF and event
-	 * 0x013c as CPU_CLK_UNHALTED.BUS and specifies there is a fixed
-	 * ratio between these counters.
-	 */
-	/* FIXED_EVENT_CONSTRAINT(0x013c, 2),  CPU_CLK_UNHALTED.REF */
+	FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */
 	INTEL_EVENT_CONSTRAINT(0x10, 0x1), /* FP_COMP_OPS_EXE */
 	INTEL_EVENT_CONSTRAINT(0x11, 0x2), /* FP_ASSIST */
 	INTEL_EVENT_CONSTRAINT(0x12, 0x2), /* MUL */
@@ -68,7 +64,7 @@ static struct event_constraint intel_nehalem_event_constraints[] __read_mostly =
 {
 	FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */
 	FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */
-	/* FIXED_EVENT_CONSTRAINT(0x013c, 2), CPU_CLK_UNHALTED.REF */
+	FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */
 	INTEL_EVENT_CONSTRAINT(0x40, 0x3), /* L1D_CACHE_LD */
 	INTEL_EVENT_CONSTRAINT(0x41, 0x3), /* L1D_CACHE_ST */
 	INTEL_EVENT_CONSTRAINT(0x42, 0x3), /* L1D_CACHE_LOCK */
@@ -90,7 +86,7 @@ static struct event_constraint intel_westmere_event_constraints[] __read_mostly
 {
 	FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */
 	FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */
-	/* FIXED_EVENT_CONSTRAINT(0x013c, 2), CPU_CLK_UNHALTED.REF */
+	FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */
 	INTEL_EVENT_CONSTRAINT(0x51, 0x3), /* L1D */
 	INTEL_EVENT_CONSTRAINT(0x60, 0x1), /* OFFCORE_REQUESTS_OUTSTANDING */
 	INTEL_EVENT_CONSTRAINT(0x63, 0x3), /* CACHE_LOCK_CYCLES */
@@ -102,7 +98,7 @@ static struct event_constraint intel_snb_event_constraints[] __read_mostly =
 {
 	FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */
 	FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */
-	/* FIXED_EVENT_CONSTRAINT(0x013c, 2), CPU_CLK_UNHALTED.REF */
+	FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */
 	INTEL_EVENT_CONSTRAINT(0x48, 0x4), /* L1D_PEND_MISS.PENDING */
 	INTEL_UEVENT_CONSTRAINT(0x01c0, 0x2), /* INST_RETIRED.PREC_DIST */
 	INTEL_EVENT_CONSTRAINT(0xcd, 0x8), /* MEM_TRANS_RETIRED.LOAD_LATENCY */
@@ -125,7 +121,7 @@ static struct event_constraint intel_gen_event_constraints[] __read_mostly =
 {
 	FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */
 	FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */
-	/* FIXED_EVENT_CONSTRAINT(0x013c, 2), CPU_CLK_UNHALTED.REF */
+	FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */
 	EVENT_CONSTRAINT_END
 };
 
@@ -1519,7 +1515,7 @@ static __initconst const struct x86_pmu intel_pmu = {
 	.guest_get_msrs		= intel_guest_get_msrs,
 };
 
-static void intel_clovertown_quirks(void)
+static __init void intel_clovertown_quirk(void)
 {
 	/*
 	 * PEBS is unreliable due to:
@@ -1545,19 +1541,60 @@ static void intel_clovertown_quirks(void)
 	x86_pmu.pebs_constraints = NULL;
 }
 
-static void intel_sandybridge_quirks(void)
+static __init void intel_sandybridge_quirk(void)
 {
 	printk(KERN_WARNING "PEBS disabled due to CPU errata.\n");
 	x86_pmu.pebs = 0;
 	x86_pmu.pebs_constraints = NULL;
 }
 
+static const struct { int id; char *name; } intel_arch_events_map[] __initconst = {
+	{ PERF_COUNT_HW_CPU_CYCLES, "cpu cycles" },
+	{ PERF_COUNT_HW_INSTRUCTIONS, "instructions" },
+	{ PERF_COUNT_HW_BUS_CYCLES, "bus cycles" },
+	{ PERF_COUNT_HW_CACHE_REFERENCES, "cache references" },
+	{ PERF_COUNT_HW_CACHE_MISSES, "cache misses" },
+	{ PERF_COUNT_HW_BRANCH_INSTRUCTIONS, "branch instructions" },
+	{ PERF_COUNT_HW_BRANCH_MISSES, "branch misses" },
+};
+
+static __init void intel_arch_events_quirk(void)
+{
+	int bit;
+
+	/* disable event that reported as not presend by cpuid */
+	for_each_set_bit(bit, x86_pmu.events_mask, ARRAY_SIZE(intel_arch_events_map)) {
+		intel_perfmon_event_map[intel_arch_events_map[bit].id] = 0;
+		printk(KERN_WARNING "CPUID marked event: \'%s\' unavailable\n",
+				intel_arch_events_map[bit].name);
+	}
+}
+
+static __init void intel_nehalem_quirk(void)
+{
+	union cpuid10_ebx ebx;
+
+	ebx.full = x86_pmu.events_maskl;
+	if (ebx.split.no_branch_misses_retired) {
+		/*
+		 * Erratum AAJ80 detected, we work it around by using
+		 * the BR_MISP_EXEC.ANY event. This will over-count
+		 * branch-misses, but it's still much better than the
+		 * architectural event which is often completely bogus:
+		 */
+		intel_perfmon_event_map[PERF_COUNT_HW_BRANCH_MISSES] = 0x7f89;
+		ebx.split.no_branch_misses_retired = 0;
+		x86_pmu.events_maskl = ebx.full;
+		printk(KERN_INFO "CPU erratum AAJ80 worked around\n");
+	}
+}
+
 __init int intel_pmu_init(void)
 {
 	union cpuid10_edx edx;
 	union cpuid10_eax eax;
+	union cpuid10_ebx ebx;
 	unsigned int unused;
-	unsigned int ebx;
 	int version;
 
 	if (!cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) {
@@ -1574,8 +1611,8 @@ __init int intel_pmu_init(void)
 	 * Check whether the Architectural PerfMon supports
 	 * Branch Misses Retired hw_event or not.
 	 */
-	cpuid(10, &eax.full, &ebx, &unused, &edx.full);
-	if (eax.split.mask_length <= ARCH_PERFMON_BRANCH_MISSES_RETIRED)
+	cpuid(10, &eax.full, &ebx.full, &unused, &edx.full);
+	if (eax.split.mask_length < ARCH_PERFMON_EVENTS_COUNT)
 		return -ENODEV;
 
 	version = eax.split.version_id;
@@ -1589,6 +1626,9 @@ __init int intel_pmu_init(void)
 	x86_pmu.cntval_bits		= eax.split.bit_width;
 	x86_pmu.cntval_mask		= (1ULL << eax.split.bit_width) - 1;
 
+	x86_pmu.events_maskl		= ebx.full;
+	x86_pmu.events_mask_len		= eax.split.mask_length;
+
 	/*
 	 * Quirk: v2 perfmon does not report fixed-purpose events, so
 	 * assume at least 3 events:
@@ -1608,6 +1648,8 @@ __init int intel_pmu_init(void)
 
 	intel_ds_init();
 
+	x86_add_quirk(intel_arch_events_quirk); /* Install first, so it runs last */
+
 	/*
 	 * Install the hw-cache-events table:
 	 */
@@ -1617,7 +1659,7 @@ __init int intel_pmu_init(void)
 		break;
 
 	case 15: /* original 65 nm celeron/pentium/core2/xeon, "Merom"/"Conroe" */
-		x86_pmu.quirks = intel_clovertown_quirks;
+		x86_add_quirk(intel_clovertown_quirk);
 	case 22: /* single-core 65 nm celeron/core2solo "Merom-L"/"Conroe-L" */
 	case 23: /* current 45 nm celeron/core2/xeon "Penryn"/"Wolfdale" */
 	case 29: /* six-core 45 nm xeon "Dunnington" */
@@ -1651,17 +1693,8 @@ __init int intel_pmu_init(void)
 		/* UOPS_EXECUTED.CORE_ACTIVE_CYCLES,c=1,i=1 */
 		intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = 0x1803fb1;
 
-		if (ebx & 0x40) {
-			/*
-			 * Erratum AAJ80 detected, we work it around by using
-			 * the BR_MISP_EXEC.ANY event. This will over-count
-			 * branch-misses, but it's still much better than the
-			 * architectural event which is often completely bogus:
-			 */
-			intel_perfmon_event_map[PERF_COUNT_HW_BRANCH_MISSES] = 0x7f89;
+		x86_add_quirk(intel_nehalem_quirk);
 
-			pr_cont("erratum AAJ80 worked around, ");
-		}
 		pr_cont("Nehalem events, ");
 		break;
 
@@ -1701,7 +1734,7 @@ __init int intel_pmu_init(void)
 		break;
 
 	case 42: /* SandyBridge */
-		x86_pmu.quirks = intel_sandybridge_quirks;
+		x86_add_quirk(intel_sandybridge_quirk);
 	case 45: /* SandyBridge, "Romely-EP" */
 		memcpy(hw_cache_event_ids, snb_hw_cache_event_ids,
 		       sizeof(hw_cache_event_ids));
@@ -1738,5 +1771,6 @@ __init int intel_pmu_init(void)
 			break;
 		}
 	}
+
 	return 0;
 }

arch/x86/kernel/jump_label.c

@@ -50,7 +50,7 @@ void arch_jump_label_transform(struct jump_entry *entry,
 	put_online_cpus();
 }
 
-void arch_jump_label_transform_static(struct jump_entry *entry,
+__init_or_module void arch_jump_label_transform_static(struct jump_entry *entry,
 				      enum jump_label_type type)
 {
 	__jump_label_transform(entry, type, text_poke_early);

arch/x86/lib/inat.c

@@ -82,9 +82,16 @@ insn_attr_t inat_get_avx_attribute(insn_byte_t opcode, insn_byte_t vex_m,
 	const insn_attr_t *table;
 	if (vex_m > X86_VEX_M_MAX || vex_p > INAT_LSTPFX_MAX)
 		return 0;
-	table = inat_avx_tables[vex_m][vex_p];
+	/* At first, this checks the master table */
+	table = inat_avx_tables[vex_m][0];
 	if (!table)
 		return 0;
+	if (!inat_is_group(table[opcode]) && vex_p) {
+		/* If this is not a group, get attribute directly */
+		table = inat_avx_tables[vex_m][vex_p];
+		if (!table)
+			return 0;
+	}
 	return table[opcode];
 }
 

arch/x86/lib/insn.c

@@ -202,7 +202,7 @@ void insn_get_opcode(struct insn *insn)
 		m = insn_vex_m_bits(insn);
 		p = insn_vex_p_bits(insn);
 		insn->attr = inat_get_avx_attribute(op, m, p);
-		if (!inat_accept_vex(insn->attr))
+		if (!inat_accept_vex(insn->attr) && !inat_is_group(insn->attr))
 			insn->attr = 0;	/* This instruction is bad */
 		goto end;	/* VEX has only 1 byte for opcode */
 	}
@@ -249,6 +249,8 @@ void insn_get_modrm(struct insn *insn)
 			pfx = insn_last_prefix(insn);
 			insn->attr = inat_get_group_attribute(mod, pfx,
 							      insn->attr);
+			if (insn_is_avx(insn) && !inat_accept_vex(insn->attr))
+				insn->attr = 0;	/* This is bad */
 		}
 	}
 

arch/x86/lib/x86-opcode-map.txt

@@ -1,5 +1,11 @@
 # x86 Opcode Maps
 #
+# This is (mostly) based on following documentations.
+# - Intel(R) 64 and IA-32 Architectures Software Developer's Manual Vol.2
+#   (#325383-040US, October 2011)
+# - Intel(R) Advanced Vector Extensions Programming Reference
+#   (#319433-011,JUNE 2011).
+#
 #<Opcode maps>
 # Table: table-name
 # Referrer: escaped-name
@@ -15,10 +21,13 @@
 # EndTable
 #
 # AVX Superscripts
-#  (VEX): this opcode can accept VEX prefix.
-#  (oVEX): this opcode requires VEX prefix.
-#  (o128): this opcode only supports 128bit VEX.
-#  (o256): this opcode only supports 256bit VEX.
+#  (v): this opcode requires VEX prefix.
+#  (v1): this opcode only supports 128bit VEX.
+#
+# Last Prefix Superscripts
+#  - (66): the last prefix is 0x66
+#  - (F3): the last prefix is 0xF3
+#  - (F2): the last prefix is 0xF2
 #
 
 Table: one byte opcode
@@ -199,8 +208,8 @@ a0: MOV AL,Ob
 a1: MOV rAX,Ov
 a2: MOV Ob,AL
 a3: MOV Ov,rAX
-a4: MOVS/B Xb,Yb
-a5: MOVS/W/D/Q Xv,Yv
+a4: MOVS/B Yb,Xb
+a5: MOVS/W/D/Q Yv,Xv
 a6: CMPS/B Xb,Yb
 a7: CMPS/W/D Xv,Yv
 a8: TEST AL,Ib
@@ -233,8 +242,8 @@ c0: Grp2 Eb,Ib (1A)
 c1: Grp2 Ev,Ib (1A)
 c2: RETN Iw (f64)
 c3: RETN
-c4: LES Gz,Mp (i64) | 3bytes-VEX (Prefix)
-c5: LDS Gz,Mp (i64) | 2bytes-VEX (Prefix)
+c4: LES Gz,Mp (i64) | VEX+2byte (Prefix)
+c5: LDS Gz,Mp (i64) | VEX+1byte (Prefix)
 c6: Grp11 Eb,Ib (1A)
 c7: Grp11 Ev,Iz (1A)
 c8: ENTER Iw,Ib
@@ -320,14 +329,19 @@ AVXcode: 1
 # 3DNow! uses the last imm byte as opcode extension.
 0f: 3DNow! Pq,Qq,Ib
 # 0x0f 0x10-0x1f
-10: movups Vps,Wps (VEX) | movss Vss,Wss (F3),(VEX),(o128) | movupd Vpd,Wpd (66),(VEX) | movsd Vsd,Wsd (F2),(VEX),(o128)
-11: movups Wps,Vps (VEX) | movss Wss,Vss (F3),(VEX),(o128) | movupd Wpd,Vpd (66),(VEX) | movsd Wsd,Vsd (F2),(VEX),(o128)
-12: movlps Vq,Mq (VEX),(o128) | movlpd Vq,Mq (66),(VEX),(o128) | movhlps Vq,Uq (VEX),(o128) | movddup Vq,Wq (F2),(VEX) | movsldup Vq,Wq (F3),(VEX)
-13: mpvlps Mq,Vq (VEX),(o128) | movlpd Mq,Vq (66),(VEX),(o128)
-14: unpcklps Vps,Wq (VEX) | unpcklpd Vpd,Wq (66),(VEX)
-15: unpckhps Vps,Wq (VEX) | unpckhpd Vpd,Wq (66),(VEX)
-16: movhps Vq,Mq (VEX),(o128) | movhpd Vq,Mq (66),(VEX),(o128) | movlsps Vq,Uq (VEX),(o128) | movshdup Vq,Wq (F3),(VEX)
-17: movhps Mq,Vq (VEX),(o128) | movhpd Mq,Vq (66),(VEX),(o128)
+# NOTE: According to Intel SDM opcode map, vmovups and vmovupd has no operands
+# but it actually has operands. And also, vmovss and vmovsd only accept 128bit.
+# MOVSS/MOVSD has too many forms(3) on SDM. This map just shows a typical form.
+# Many AVX instructions lack v1 superscript, according to Intel AVX-Prgramming
+# Reference A.1
+10: vmovups Vps,Wps | vmovupd Vpd,Wpd (66) | vmovss Vx,Hx,Wss (F3),(v1) | vmovsd Vx,Hx,Wsd (F2),(v1)
+11: vmovups Wps,Vps | vmovupd Wpd,Vpd (66) | vmovss Wss,Hx,Vss (F3),(v1) | vmovsd Wsd,Hx,Vsd (F2),(v1)
+12: vmovlps Vq,Hq,Mq (v1) | vmovhlps Vq,Hq,Uq (v1) | vmovlpd Vq,Hq,Mq (66),(v1) | vmovsldup Vx,Wx (F3) | vmovddup Vx,Wx (F2)
+13: vmovlps Mq,Vq (v1) | vmovlpd Mq,Vq (66),(v1)
+14: vunpcklps Vx,Hx,Wx | vunpcklpd Vx,Hx,Wx (66)
+15: vunpckhps Vx,Hx,Wx | vunpckhpd Vx,Hx,Wx (66)
+16: vmovhps Vdq,Hq,Mq (v1) | vmovlhps Vdq,Hq,Uq (v1) | vmovhpd Vdq,Hq,Mq (66),(v1) | vmovshdup Vx,Wx (F3)
+17: vmovhps Mq,Vq (v1) | vmovhpd Mq,Vq (66),(v1)
 18: Grp16 (1A)
 19:
 1a:
@@ -345,14 +359,14 @@ AVXcode: 1
 25:
 26:
 27:
-28: movaps Vps,Wps (VEX) | movapd Vpd,Wpd (66),(VEX)
-29: movaps Wps,Vps (VEX) | movapd Wpd,Vpd (66),(VEX)
-2a: cvtpi2ps Vps,Qpi | cvtsi2ss Vss,Ed/q (F3),(VEX),(o128) | cvtpi2pd Vpd,Qpi (66) | cvtsi2sd Vsd,Ed/q (F2),(VEX),(o128)
-2b: movntps Mps,Vps (VEX) | movntpd Mpd,Vpd (66),(VEX)
-2c: cvttps2pi Ppi,Wps | cvttss2si  Gd/q,Wss (F3),(VEX),(o128) | cvttpd2pi Ppi,Wpd (66) | cvttsd2si Gd/q,Wsd (F2),(VEX),(o128)
-2d: cvtps2pi Ppi,Wps | cvtss2si Gd/q,Wss (F3),(VEX),(o128) | cvtpd2pi Qpi,Wpd (66) | cvtsd2si Gd/q,Wsd (F2),(VEX),(o128)
-2e: ucomiss Vss,Wss (VEX),(o128) | ucomisd  Vsd,Wsd (66),(VEX),(o128)
-2f: comiss Vss,Wss (VEX),(o128) | comisd  Vsd,Wsd (66),(VEX),(o128)
+28: vmovaps Vps,Wps | vmovapd Vpd,Wpd (66)
+29: vmovaps Wps,Vps | vmovapd Wpd,Vpd (66)
+2a: cvtpi2ps Vps,Qpi | cvtpi2pd Vpd,Qpi (66) | vcvtsi2ss Vss,Hss,Ey (F3),(v1) | vcvtsi2sd Vsd,Hsd,Ey (F2),(v1)
+2b: vmovntps Mps,Vps | vmovntpd Mpd,Vpd (66)
+2c: cvttps2pi Ppi,Wps | cvttpd2pi Ppi,Wpd (66) | vcvttss2si Gy,Wss (F3),(v1) | vcvttsd2si Gy,Wsd (F2),(v1)
+2d: cvtps2pi Ppi,Wps | cvtpd2pi Qpi,Wpd (66) | vcvtss2si Gy,Wss (F3),(v1) | vcvtsd2si Gy,Wsd (F2),(v1)
+2e: vucomiss Vss,Wss (v1) | vucomisd  Vsd,Wsd (66),(v1)
+2f: vcomiss Vss,Wss (v1) | vcomisd  Vsd,Wsd (66),(v1)
 # 0x0f 0x30-0x3f
 30: WRMSR
 31: RDTSC
@@ -388,65 +402,66 @@ AVXcode: 1
 4e: CMOVLE/NG Gv,Ev
 4f: CMOVNLE/G Gv,Ev
 # 0x0f 0x50-0x5f
-50: movmskps Gd/q,Ups (VEX) | movmskpd Gd/q,Upd (66),(VEX)
-51: sqrtps Vps,Wps (VEX) | sqrtss Vss,Wss (F3),(VEX),(o128) | sqrtpd Vpd,Wpd (66),(VEX) | sqrtsd Vsd,Wsd (F2),(VEX),(o128)
-52: rsqrtps Vps,Wps (VEX) | rsqrtss Vss,Wss (F3),(VEX),(o128)
-53: rcpps Vps,Wps (VEX) | rcpss Vss,Wss (F3),(VEX),(o128)
-54: andps Vps,Wps (VEX) | andpd Vpd,Wpd (66),(VEX)
-55: andnps Vps,Wps (VEX) | andnpd Vpd,Wpd (66),(VEX)
-56: orps Vps,Wps (VEX) | orpd Vpd,Wpd (66),(VEX)
-57: xorps Vps,Wps (VEX) | xorpd Vpd,Wpd (66),(VEX)
-58: addps Vps,Wps (VEX) | addss Vss,Wss (F3),(VEX),(o128) | addpd Vpd,Wpd (66),(VEX) | addsd Vsd,Wsd (F2),(VEX),(o128)
-59: mulps Vps,Wps (VEX) | mulss Vss,Wss (F3),(VEX),(o128) | mulpd Vpd,Wpd (66),(VEX) | mulsd Vsd,Wsd (F2),(VEX),(o128)
-5a: cvtps2pd Vpd,Wps (VEX) | cvtss2sd Vsd,Wss (F3),(VEX),(o128) | cvtpd2ps Vps,Wpd (66),(VEX) | cvtsd2ss Vsd,Wsd (F2),(VEX),(o128)
-5b: cvtdq2ps Vps,Wdq (VEX) | cvtps2dq Vdq,Wps (66),(VEX) | cvttps2dq Vdq,Wps (F3),(VEX)
-5c: subps Vps,Wps (VEX) | subss Vss,Wss (F3),(VEX),(o128) | subpd Vpd,Wpd (66),(VEX) | subsd Vsd,Wsd (F2),(VEX),(o128)
-5d: minps Vps,Wps (VEX) | minss Vss,Wss (F3),(VEX),(o128) | minpd Vpd,Wpd (66),(VEX) | minsd Vsd,Wsd (F2),(VEX),(o128)
-5e: divps Vps,Wps (VEX) | divss Vss,Wss (F3),(VEX),(o128) | divpd Vpd,Wpd (66),(VEX) | divsd Vsd,Wsd (F2),(VEX),(o128)
-5f: maxps Vps,Wps (VEX) | maxss Vss,Wss (F3),(VEX),(o128) | maxpd Vpd,Wpd (66),(VEX) | maxsd Vsd,Wsd (F2),(VEX),(o128)
+50: vmovmskps Gy,Ups | vmovmskpd Gy,Upd (66)
+51: vsqrtps Vps,Wps | vsqrtpd Vpd,Wpd (66) | vsqrtss Vss,Hss,Wss (F3),(v1) | vsqrtsd Vsd,Hsd,Wsd (F2),(v1)
+52: vrsqrtps Vps,Wps | vrsqrtss Vss,Hss,Wss (F3),(v1)
+53: vrcpps Vps,Wps | vrcpss Vss,Hss,Wss (F3),(v1)
+54: vandps Vps,Hps,Wps | vandpd Vpd,Hpd,Wpd (66)
+55: vandnps Vps,Hps,Wps | vandnpd Vpd,Hpd,Wpd (66)
+56: vorps Vps,Hps,Wps | vorpd Vpd,Hpd,Wpd (66)
+57: vxorps Vps,Hps,Wps | vxorpd Vpd,Hpd,Wpd (66)
+58: vaddps Vps,Hps,Wps | vaddpd Vpd,Hpd,Wpd (66) | vaddss Vss,Hss,Wss (F3),(v1) | vaddsd Vsd,Hsd,Wsd (F2),(v1)
+59: vmulps Vps,Hps,Wps | vmulpd Vpd,Hpd,Wpd (66) | vmulss Vss,Hss,Wss (F3),(v1) | vmulsd Vsd,Hsd,Wsd (F2),(v1)
+5a: vcvtps2pd Vpd,Wps | vcvtpd2ps Vps,Wpd (66) | vcvtss2sd Vsd,Hx,Wss (F3),(v1) | vcvtsd2ss Vss,Hx,Wsd (F2),(v1)
+5b: vcvtdq2ps Vps,Wdq | vcvtps2dq Vdq,Wps (66) | vcvttps2dq Vdq,Wps (F3)
+5c: vsubps Vps,Hps,Wps | vsubpd Vpd,Hpd,Wpd (66) | vsubss Vss,Hss,Wss (F3),(v1) | vsubsd Vsd,Hsd,Wsd (F2),(v1)
+5d: vminps Vps,Hps,Wps | vminpd Vpd,Hpd,Wpd (66) | vminss Vss,Hss,Wss (F3),(v1) | vminsd Vsd,Hsd,Wsd (F2),(v1)
+5e: vdivps Vps,Hps,Wps | vdivpd Vpd,Hpd,Wpd (66) | vdivss Vss,Hss,Wss (F3),(v1) | vdivsd Vsd,Hsd,Wsd (F2),(v1)
+5f: vmaxps Vps,Hps,Wps | vmaxpd Vpd,Hpd,Wpd (66) | vmaxss Vss,Hss,Wss (F3),(v1) | vmaxsd Vsd,Hsd,Wsd (F2),(v1)
 # 0x0f 0x60-0x6f
-60: punpcklbw Pq,Qd | punpcklbw Vdq,Wdq (66),(VEX),(o128)
-61: punpcklwd Pq,Qd | punpcklwd Vdq,Wdq (66),(VEX),(o128)
-62: punpckldq Pq,Qd | punpckldq Vdq,Wdq (66),(VEX),(o128)
-63: packsswb Pq,Qq | packsswb Vdq,Wdq (66),(VEX),(o128)
-64: pcmpgtb Pq,Qq | pcmpgtb Vdq,Wdq (66),(VEX),(o128)
-65: pcmpgtw Pq,Qq | pcmpgtw Vdq,Wdq (66),(VEX),(o128)
-66: pcmpgtd Pq,Qq | pcmpgtd Vdq,Wdq (66),(VEX),(o128)
-67: packuswb Pq,Qq | packuswb Vdq,Wdq (66),(VEX),(o128)
-68: punpckhbw Pq,Qd | punpckhbw Vdq,Wdq (66),(VEX),(o128)
-69: punpckhwd Pq,Qd | punpckhwd Vdq,Wdq (66),(VEX),(o128)
-6a: punpckhdq Pq,Qd | punpckhdq Vdq,Wdq (66),(VEX),(o128)
-6b: packssdw Pq,Qd | packssdw Vdq,Wdq (66),(VEX),(o128)
-6c: punpcklqdq Vdq,Wdq (66),(VEX),(o128)
-6d: punpckhqdq Vdq,Wdq (66),(VEX),(o128)
-6e: movd/q/ Pd,Ed/q | movd/q Vdq,Ed/q (66),(VEX),(o128)
-6f: movq Pq,Qq | movdqa Vdq,Wdq (66),(VEX) | movdqu Vdq,Wdq (F3),(VEX)
+60: punpcklbw Pq,Qd | vpunpcklbw Vx,Hx,Wx (66),(v1)
+61: punpcklwd Pq,Qd | vpunpcklwd Vx,Hx,Wx (66),(v1)
+62: punpckldq Pq,Qd | vpunpckldq Vx,Hx,Wx (66),(v1)
+63: packsswb Pq,Qq | vpacksswb Vx,Hx,Wx (66),(v1)
+64: pcmpgtb Pq,Qq | vpcmpgtb Vx,Hx,Wx (66),(v1)
+65: pcmpgtw Pq,Qq | vpcmpgtw Vx,Hx,Wx (66),(v1)
+66: pcmpgtd Pq,Qq | vpcmpgtd Vx,Hx,Wx (66),(v1)
+67: packuswb Pq,Qq | vpackuswb Vx,Hx,Wx (66),(v1)
+68: punpckhbw Pq,Qd | vpunpckhbw Vx,Hx,Wx (66),(v1)
+69: punpckhwd Pq,Qd | vpunpckhwd Vx,Hx,Wx (66),(v1)
+6a: punpckhdq Pq,Qd | vpunpckhdq Vx,Hx,Wx (66),(v1)
+6b: packssdw Pq,Qd | vpackssdw Vx,Hx,Wx (66),(v1)
+6c: vpunpcklqdq Vx,Hx,Wx (66),(v1)
+6d: vpunpckhqdq Vx,Hx,Wx (66),(v1)
+6e: movd/q Pd,Ey | vmovd/q Vy,Ey (66),(v1)
+6f: movq Pq,Qq | vmovdqa Vx,Wx (66) | vmovdqu Vx,Wx (F3)
 # 0x0f 0x70-0x7f
-70: pshufw Pq,Qq,Ib | pshufd Vdq,Wdq,Ib (66),(VEX),(o128) | pshufhw Vdq,Wdq,Ib (F3),(VEX),(o128) | pshuflw VdqWdq,Ib (F2),(VEX),(o128)
+70: pshufw Pq,Qq,Ib | vpshufd Vx,Wx,Ib (66),(v1) | vpshufhw Vx,Wx,Ib (F3),(v1) | vpshuflw Vx,Wx,Ib (F2),(v1)
 71: Grp12 (1A)
 72: Grp13 (1A)
 73: Grp14 (1A)
-74: pcmpeqb Pq,Qq | pcmpeqb Vdq,Wdq (66),(VEX),(o128)
-75: pcmpeqw Pq,Qq | pcmpeqw Vdq,Wdq (66),(VEX),(o128)
-76: pcmpeqd Pq,Qq | pcmpeqd Vdq,Wdq (66),(VEX),(o128)
-77: emms/vzeroupper/vzeroall (VEX)
-78: VMREAD Ed/q,Gd/q
-79: VMWRITE Gd/q,Ed/q
+74: pcmpeqb Pq,Qq | vpcmpeqb Vx,Hx,Wx (66),(v1)
+75: pcmpeqw Pq,Qq | vpcmpeqw Vx,Hx,Wx (66),(v1)
+76: pcmpeqd Pq,Qq | vpcmpeqd Vx,Hx,Wx (66),(v1)
+# Note: Remove (v), because vzeroall and vzeroupper becomes emms without VEX.
+77: emms | vzeroupper | vzeroall
+78: VMREAD Ey,Gy
+79: VMWRITE Gy,Ey
 7a:
 7b:
-7c: haddps Vps,Wps (F2),(VEX) | haddpd Vpd,Wpd (66),(VEX)
-7d: hsubps Vps,Wps (F2),(VEX) | hsubpd Vpd,Wpd (66),(VEX)
-7e: movd/q Ed/q,Pd | movd/q Ed/q,Vdq (66),(VEX),(o128) | movq Vq,Wq (F3),(VEX),(o128)
-7f: movq Qq,Pq | movdqa Wdq,Vdq (66),(VEX) | movdqu Wdq,Vdq (F3),(VEX)
+7c: vhaddpd Vpd,Hpd,Wpd (66) | vhaddps Vps,Hps,Wps (F2)
+7d: vhsubpd Vpd,Hpd,Wpd (66) | vhsubps Vps,Hps,Wps (F2)
+7e: movd/q Ey,Pd | vmovd/q Ey,Vy (66),(v1) | vmovq Vq,Wq (F3),(v1)
+7f: movq Qq,Pq | vmovdqa Wx,Vx (66) | vmovdqu Wx,Vx (F3)
 # 0x0f 0x80-0x8f
 80: JO Jz (f64)
 81: JNO Jz (f64)
-82: JB/JNAE/JC Jz (f64)
-83: JNB/JAE/JNC Jz (f64)
-84: JZ/JE Jz (f64)
-85: JNZ/JNE Jz (f64)
+82: JB/JC/JNAE Jz (f64)
+83: JAE/JNB/JNC Jz (f64)
+84: JE/JZ Jz (f64)
+85: JNE/JNZ Jz (f64)
 86: JBE/JNA Jz (f64)
-87: JNBE/JA Jz (f64)
+87: JA/JNBE Jz (f64)
 88: JS Jz (f64)
 89: JNS Jz (f64)
 8a: JP/JPE Jz (f64)
@@ -502,18 +517,18 @@ b8: JMPE | POPCNT Gv,Ev (F3)
 b9: Grp10 (1A)
 ba: Grp8 Ev,Ib (1A)
 bb: BTC Ev,Gv
-bc: BSF Gv,Ev
-bd: BSR Gv,Ev
+bc: BSF Gv,Ev | TZCNT Gv,Ev (F3)
+bd: BSR Gv,Ev | LZCNT Gv,Ev (F3)
 be: MOVSX Gv,Eb
 bf: MOVSX Gv,Ew
 # 0x0f 0xc0-0xcf
 c0: XADD Eb,Gb
 c1: XADD Ev,Gv
-c2: cmpps Vps,Wps,Ib (VEX) | cmpss Vss,Wss,Ib (F3),(VEX),(o128) | cmppd Vpd,Wpd,Ib (66),(VEX) | cmpsd Vsd,Wsd,Ib (F2),(VEX)
-c3: movnti Md/q,Gd/q
-c4: pinsrw Pq,Rd/q/Mw,Ib | pinsrw Vdq,Rd/q/Mw,Ib (66),(VEX),(o128)
-c5: pextrw Gd,Nq,Ib | pextrw Gd,Udq,Ib (66),(VEX),(o128)
-c6: shufps Vps,Wps,Ib (VEX) | shufpd Vpd,Wpd,Ib (66),(VEX)
+c2: vcmpps Vps,Hps,Wps,Ib | vcmppd Vpd,Hpd,Wpd,Ib (66) | vcmpss Vss,Hss,Wss,Ib (F3),(v1) | vcmpsd Vsd,Hsd,Wsd,Ib (F2),(v1)
+c3: movnti My,Gy
+c4: pinsrw Pq,Ry/Mw,Ib | vpinsrw Vdq,Hdq,Ry/Mw,Ib (66),(v1)
+c5: pextrw Gd,Nq,Ib | vpextrw Gd,Udq,Ib (66),(v1)
+c6: vshufps Vps,Hps,Wps,Ib | vshufpd Vpd,Hpd,Wpd,Ib (66)
 c7: Grp9 (1A)
 c8: BSWAP RAX/EAX/R8/R8D
 c9: BSWAP RCX/ECX/R9/R9D
@@ -524,55 +539,55 @@ cd: BSWAP RBP/EBP/R13/R13D
 ce: BSWAP RSI/ESI/R14/R14D
 cf: BSWAP RDI/EDI/R15/R15D
 # 0x0f 0xd0-0xdf
-d0: addsubps Vps,Wps (F2),(VEX) | addsubpd Vpd,Wpd (66),(VEX)
-d1: psrlw Pq,Qq | psrlw Vdq,Wdq (66),(VEX),(o128)
-d2: psrld Pq,Qq | psrld Vdq,Wdq (66),(VEX),(o128)
-d3: psrlq Pq,Qq | psrlq Vdq,Wdq (66),(VEX),(o128)
-d4: paddq Pq,Qq | paddq Vdq,Wdq (66),(VEX),(o128)
-d5: pmullw Pq,Qq | pmullw Vdq,Wdq (66),(VEX),(o128)
-d6: movq Wq,Vq (66),(VEX),(o128) | movq2dq Vdq,Nq (F3) | movdq2q Pq,Uq (F2)
-d7: pmovmskb Gd,Nq | pmovmskb Gd,Udq (66),(VEX),(o128)
-d8: psubusb Pq,Qq | psubusb Vdq,Wdq (66),(VEX),(o128)
-d9: psubusw Pq,Qq | psubusw Vdq,Wdq (66),(VEX),(o128)
-da: pminub Pq,Qq | pminub Vdq,Wdq (66),(VEX),(o128)
-db: pand Pq,Qq | pand Vdq,Wdq (66),(VEX),(o128)
-dc: paddusb Pq,Qq | paddusb Vdq,Wdq (66),(VEX),(o128)
-dd: paddusw Pq,Qq | paddusw Vdq,Wdq (66),(VEX),(o128)
-de: pmaxub Pq,Qq | pmaxub Vdq,Wdq (66),(VEX),(o128)
-df: pandn Pq,Qq | pandn Vdq,Wdq (66),(VEX),(o128)
+d0: vaddsubpd Vpd,Hpd,Wpd (66) | vaddsubps Vps,Hps,Wps (F2)
+d1: psrlw Pq,Qq | vpsrlw Vx,Hx,Wx (66),(v1)
+d2: psrld Pq,Qq | vpsrld Vx,Hx,Wx (66),(v1)
+d3: psrlq Pq,Qq | vpsrlq Vx,Hx,Wx (66),(v1)
+d4: paddq Pq,Qq | vpaddq Vx,Hx,Wx (66),(v1)
+d5: pmullw Pq,Qq | vpmullw Vx,Hx,Wx (66),(v1)
+d6: vmovq Wq,Vq (66),(v1) | movq2dq Vdq,Nq (F3) | movdq2q Pq,Uq (F2)
+d7: pmovmskb Gd,Nq | vpmovmskb Gd,Ux (66),(v1)
+d8: psubusb Pq,Qq | vpsubusb Vx,Hx,Wx (66),(v1)
+d9: psubusw Pq,Qq | vpsubusw Vx,Hx,Wx (66),(v1)
+da: pminub Pq,Qq | vpminub Vx,Hx,Wx (66),(v1)
+db: pand Pq,Qq | vpand Vx,Hx,Wx (66),(v1)
+dc: paddusb Pq,Qq | vpaddusb Vx,Hx,Wx (66),(v1)
+dd: paddusw Pq,Qq | vpaddusw Vx,Hx,Wx (66),(v1)
+de: pmaxub Pq,Qq | vpmaxub Vx,Hx,Wx (66),(v1)
+df: pandn Pq,Qq | vpandn Vx,Hx,Wx (66),(v1)
 # 0x0f 0xe0-0xef
-e0: pavgb Pq,Qq | pavgb Vdq,Wdq (66),(VEX),(o128)
-e1: psraw Pq,Qq | psraw Vdq,Wdq (66),(VEX),(o128)
-e2: psrad Pq,Qq | psrad Vdq,Wdq (66),(VEX),(o128)
-e3: pavgw Pq,Qq | pavgw Vdq,Wdq (66),(VEX),(o128)
-e4: pmulhuw Pq,Qq | pmulhuw Vdq,Wdq (66),(VEX),(o128)
-e5: pmulhw Pq,Qq | pmulhw Vdq,Wdq (66),(VEX),(o128)
-e6: cvtpd2dq Vdq,Wpd (F2),(VEX) | cvttpd2dq Vdq,Wpd (66),(VEX) | cvtdq2pd Vpd,Wdq (F3),(VEX)
-e7: movntq Mq,Pq | movntdq Mdq,Vdq (66),(VEX)
-e8: psubsb Pq,Qq | psubsb Vdq,Wdq (66),(VEX),(o128)
-e9: psubsw Pq,Qq | psubsw Vdq,Wdq (66),(VEX),(o128)
-ea: pminsw Pq,Qq | pminsw Vdq,Wdq (66),(VEX),(o128)
-eb: por Pq,Qq | por Vdq,Wdq (66),(VEX),(o128)
-ec: paddsb Pq,Qq | paddsb Vdq,Wdq (66),(VEX),(o128)
-ed: paddsw Pq,Qq | paddsw Vdq,Wdq (66),(VEX),(o128)
-ee: pmaxsw Pq,Qq | pmaxsw Vdq,Wdq (66),(VEX),(o128)
-ef: pxor Pq,Qq | pxor Vdq,Wdq (66),(VEX),(o128)
+e0: pavgb Pq,Qq | vpavgb Vx,Hx,Wx (66),(v1)
+e1: psraw Pq,Qq | vpsraw Vx,Hx,Wx (66),(v1)
+e2: psrad Pq,Qq | vpsrad Vx,Hx,Wx (66),(v1)
+e3: pavgw Pq,Qq | vpavgw Vx,Hx,Wx (66),(v1)
+e4: pmulhuw Pq,Qq | vpmulhuw Vx,Hx,Wx (66),(v1)
+e5: pmulhw Pq,Qq | vpmulhw Vx,Hx,Wx (66),(v1)
+e6: vcvttpd2dq Vx,Wpd (66) | vcvtdq2pd Vx,Wdq (F3) | vcvtpd2dq Vx,Wpd (F2)
+e7: movntq Mq,Pq | vmovntdq Mx,Vx (66)
+e8: psubsb Pq,Qq | vpsubsb Vx,Hx,Wx (66),(v1)
+e9: psubsw Pq,Qq | vpsubsw Vx,Hx,Wx (66),(v1)
+ea: pminsw Pq,Qq | vpminsw Vx,Hx,Wx (66),(v1)
+eb: por Pq,Qq | vpor Vx,Hx,Wx (66),(v1)
+ec: paddsb Pq,Qq | vpaddsb Vx,Hx,Wx (66),(v1)
+ed: paddsw Pq,Qq | vpaddsw Vx,Hx,Wx (66),(v1)
+ee: pmaxsw Pq,Qq | vpmaxsw Vx,Hx,Wx (66),(v1)
+ef: pxor Pq,Qq | vpxor Vx,Hx,Wx (66),(v1)
 # 0x0f 0xf0-0xff
-f0: lddqu Vdq,Mdq (F2),(VEX)
-f1: psllw Pq,Qq | psllw Vdq,Wdq (66),(VEX),(o128)
-f2: pslld Pq,Qq | pslld Vdq,Wdq (66),(VEX),(o128)
-f3: psllq Pq,Qq | psllq Vdq,Wdq (66),(VEX),(o128)
-f4: pmuludq Pq,Qq | pmuludq Vdq,Wdq (66),(VEX),(o128)
-f5: pmaddwd Pq,Qq | pmaddwd Vdq,Wdq (66),(VEX),(o128)
-f6: psadbw Pq,Qq | psadbw Vdq,Wdq (66),(VEX),(o128)
-f7: maskmovq Pq,Nq | maskmovdqu Vdq,Udq (66),(VEX),(o128)
-f8: psubb Pq,Qq | psubb Vdq,Wdq (66),(VEX),(o128)
-f9: psubw Pq,Qq | psubw Vdq,Wdq (66),(VEX),(o128)
-fa: psubd Pq,Qq | psubd Vdq,Wdq (66),(VEX),(o128)
-fb: psubq Pq,Qq | psubq Vdq,Wdq (66),(VEX),(o128)
-fc: paddb Pq,Qq | paddb Vdq,Wdq (66),(VEX),(o128)
-fd: paddw Pq,Qq | paddw Vdq,Wdq (66),(VEX),(o128)
-fe: paddd Pq,Qq | paddd Vdq,Wdq (66),(VEX),(o128)
+f0: vlddqu Vx,Mx (F2)
+f1: psllw Pq,Qq | vpsllw Vx,Hx,Wx (66),(v1)
+f2: pslld Pq,Qq | vpslld Vx,Hx,Wx (66),(v1)
+f3: psllq Pq,Qq | vpsllq Vx,Hx,Wx (66),(v1)
+f4: pmuludq Pq,Qq | vpmuludq Vx,Hx,Wx (66),(v1)
+f5: pmaddwd Pq,Qq | vpmaddwd Vx,Hx,Wx (66),(v1)
+f6: psadbw Pq,Qq | vpsadbw Vx,Hx,Wx (66),(v1)
+f7: maskmovq Pq,Nq | vmaskmovdqu Vx,Ux (66),(v1)
+f8: psubb Pq,Qq | vpsubb Vx,Hx,Wx (66),(v1)
+f9: psubw Pq,Qq | vpsubw Vx,Hx,Wx (66),(v1)
+fa: psubd Pq,Qq | vpsubd Vx,Hx,Wx (66),(v1)
+fb: psubq Pq,Qq | vpsubq Vx,Hx,Wx (66),(v1)
+fc: paddb Pq,Qq | vpaddb Vx,Hx,Wx (66),(v1)
+fd: paddw Pq,Qq | vpaddw Vx,Hx,Wx (66),(v1)
+fe: paddd Pq,Qq | vpaddd Vx,Hx,Wx (66),(v1)
 ff:
 EndTable
 
@@ -580,155 +595,193 @@ Table: 3-byte opcode 1 (0x0f 0x38)
 Referrer: 3-byte escape 1
 AVXcode: 2
 # 0x0f 0x38 0x00-0x0f
-00: pshufb Pq,Qq | pshufb Vdq,Wdq (66),(VEX),(o128)
-01: phaddw Pq,Qq | phaddw Vdq,Wdq (66),(VEX),(o128)
-02: phaddd Pq,Qq | phaddd Vdq,Wdq (66),(VEX),(o128)
-03: phaddsw Pq,Qq | phaddsw Vdq,Wdq (66),(VEX),(o128)
-04: pmaddubsw Pq,Qq | pmaddubsw Vdq,Wdq (66),(VEX),(o128)
-05: phsubw Pq,Qq | phsubw Vdq,Wdq (66),(VEX),(o128)
-06: phsubd Pq,Qq | phsubd Vdq,Wdq (66),(VEX),(o128)
-07: phsubsw Pq,Qq | phsubsw Vdq,Wdq (66),(VEX),(o128)
-08: psignb Pq,Qq | psignb Vdq,Wdq (66),(VEX),(o128)
-09: psignw Pq,Qq | psignw Vdq,Wdq (66),(VEX),(o128)
-0a: psignd Pq,Qq | psignd Vdq,Wdq (66),(VEX),(o128)
-0b: pmulhrsw Pq,Qq | pmulhrsw Vdq,Wdq (66),(VEX),(o128)
-0c: Vpermilps /r (66),(oVEX)
-0d: Vpermilpd /r (66),(oVEX)
-0e: vtestps /r (66),(oVEX)
-0f: vtestpd /r (66),(oVEX)
+00: pshufb Pq,Qq | vpshufb Vx,Hx,Wx (66),(v1)
+01: phaddw Pq,Qq | vphaddw Vx,Hx,Wx (66),(v1)
+02: phaddd Pq,Qq | vphaddd Vx,Hx,Wx (66),(v1)
+03: phaddsw Pq,Qq | vphaddsw Vx,Hx,Wx (66),(v1)
+04: pmaddubsw Pq,Qq | vpmaddubsw Vx,Hx,Wx (66),(v1)
+05: phsubw Pq,Qq | vphsubw Vx,Hx,Wx (66),(v1)
+06: phsubd Pq,Qq | vphsubd Vx,Hx,Wx (66),(v1)
+07: phsubsw Pq,Qq | vphsubsw Vx,Hx,Wx (66),(v1)
+08: psignb Pq,Qq | vpsignb Vx,Hx,Wx (66),(v1)
+09: psignw Pq,Qq | vpsignw Vx,Hx,Wx (66),(v1)
+0a: psignd Pq,Qq | vpsignd Vx,Hx,Wx (66),(v1)
+0b: pmulhrsw Pq,Qq | vpmulhrsw Vx,Hx,Wx (66),(v1)
+0c: vpermilps Vx,Hx,Wx (66),(v)
+0d: vpermilpd Vx,Hx,Wx (66),(v)
+0e: vtestps Vx,Wx (66),(v)
+0f: vtestpd Vx,Wx (66),(v)
 # 0x0f 0x38 0x10-0x1f
 10: pblendvb Vdq,Wdq (66)
 11:
 12:
-13:
+13: vcvtph2ps Vx,Wx,Ib (66),(v)
 14: blendvps Vdq,Wdq (66)
 15: blendvpd Vdq,Wdq (66)
-16:
-17: ptest Vdq,Wdq (66),(VEX)
-18: vbroadcastss /r (66),(oVEX)
-19: vbroadcastsd /r (66),(oVEX),(o256)
-1a: vbroadcastf128 /r (66),(oVEX),(o256)
+16: vpermps Vqq,Hqq,Wqq (66),(v)
+17: vptest Vx,Wx (66)
+18: vbroadcastss Vx,Wd (66),(v)
+19: vbroadcastsd Vqq,Wq (66),(v)
+1a: vbroadcastf128 Vqq,Mdq (66),(v)
 1b:
-1c: pabsb Pq,Qq | pabsb Vdq,Wdq (66),(VEX),(o128)
-1d: pabsw Pq,Qq | pabsw Vdq,Wdq (66),(VEX),(o128)
-1e: pabsd Pq,Qq | pabsd Vdq,Wdq (66),(VEX),(o128)
+1c: pabsb Pq,Qq | vpabsb Vx,Wx (66),(v1)
+1d: pabsw Pq,Qq | vpabsw Vx,Wx (66),(v1)
+1e: pabsd Pq,Qq | vpabsd Vx,Wx (66),(v1)
 1f:
 # 0x0f 0x38 0x20-0x2f
-20: pmovsxbw Vdq,Udq/Mq (66),(VEX),(o128)
-21: pmovsxbd Vdq,Udq/Md (66),(VEX),(o128)
-22: pmovsxbq Vdq,Udq/Mw (66),(VEX),(o128)
-23: pmovsxwd Vdq,Udq/Mq (66),(VEX),(o128)
-24: pmovsxwq Vdq,Udq/Md (66),(VEX),(o128)
-25: pmovsxdq Vdq,Udq/Mq (66),(VEX),(o128)
+20: vpmovsxbw Vx,Ux/Mq (66),(v1)
+21: vpmovsxbd Vx,Ux/Md (66),(v1)
+22: vpmovsxbq Vx,Ux/Mw (66),(v1)
+23: vpmovsxwd Vx,Ux/Mq (66),(v1)
+24: vpmovsxwq Vx,Ux/Md (66),(v1)
+25: vpmovsxdq Vx,Ux/Mq (66),(v1)
 26:
 27:
-28: pmuldq Vdq,Wdq (66),(VEX),(o128)
-29: pcmpeqq Vdq,Wdq (66),(VEX),(o128)
-2a: movntdqa Vdq,Mdq (66),(VEX),(o128)
-2b: packusdw Vdq,Wdq (66),(VEX),(o128)
-2c: vmaskmovps(ld) /r (66),(oVEX)
-2d: vmaskmovpd(ld) /r (66),(oVEX)
-2e: vmaskmovps(st) /r (66),(oVEX)
-2f: vmaskmovpd(st) /r (66),(oVEX)
+28: vpmuldq Vx,Hx,Wx (66),(v1)
+29: vpcmpeqq Vx,Hx,Wx (66),(v1)
+2a: vmovntdqa Vx,Mx (66),(v1)
+2b: vpackusdw Vx,Hx,Wx (66),(v1)
+2c: vmaskmovps Vx,Hx,Mx (66),(v)
+2d: vmaskmovpd Vx,Hx,Mx (66),(v)
+2e: vmaskmovps Mx,Hx,Vx (66),(v)
+2f: vmaskmovpd Mx,Hx,Vx (66),(v)
 # 0x0f 0x38 0x30-0x3f
-30: pmovzxbw Vdq,Udq/Mq (66),(VEX),(o128)
-31: pmovzxbd Vdq,Udq/Md (66),(VEX),(o128)
-32: pmovzxbq Vdq,Udq/Mw (66),(VEX),(o128)
-33: pmovzxwd Vdq,Udq/Mq (66),(VEX),(o128)
-34: pmovzxwq Vdq,Udq/Md (66),(VEX),(o128)
-35: pmovzxdq Vdq,Udq/Mq (66),(VEX),(o128)
-36:
-37: pcmpgtq Vdq,Wdq (66),(VEX),(o128)
-38: pminsb Vdq,Wdq (66),(VEX),(o128)
-39: pminsd Vdq,Wdq (66),(VEX),(o128)
-3a: pminuw Vdq,Wdq (66),(VEX),(o128)
-3b: pminud Vdq,Wdq (66),(VEX),(o128)
-3c: pmaxsb Vdq,Wdq (66),(VEX),(o128)
-3d: pmaxsd Vdq,Wdq (66),(VEX),(o128)
-3e: pmaxuw Vdq,Wdq (66),(VEX),(o128)
-3f: pmaxud Vdq,Wdq (66),(VEX),(o128)
+30: vpmovzxbw Vx,Ux/Mq (66),(v1)
+31: vpmovzxbd Vx,Ux/Md (66),(v1)
+32: vpmovzxbq Vx,Ux/Mw (66),(v1)
+33: vpmovzxwd Vx,Ux/Mq (66),(v1)
+34: vpmovzxwq Vx,Ux/Md (66),(v1)
+35: vpmovzxdq Vx,Ux/Mq (66),(v1)
+36: vpermd Vqq,Hqq,Wqq (66),(v)
+37: vpcmpgtq Vx,Hx,Wx (66),(v1)
+38: vpminsb Vx,Hx,Wx (66),(v1)
+39: vpminsd Vx,Hx,Wx (66),(v1)
+3a: vpminuw Vx,Hx,Wx (66),(v1)
+3b: vpminud Vx,Hx,Wx (66),(v1)
+3c: vpmaxsb Vx,Hx,Wx (66),(v1)
+3d: vpmaxsd Vx,Hx,Wx (66),(v1)
+3e: vpmaxuw Vx,Hx,Wx (66),(v1)
+3f: vpmaxud Vx,Hx,Wx (66),(v1)
 # 0x0f 0x38 0x40-0x8f
-40: pmulld Vdq,Wdq (66),(VEX),(o128)
-41: phminposuw Vdq,Wdq (66),(VEX),(o128)
-80: INVEPT Gd/q,Mdq (66)
-81: INVPID Gd/q,Mdq (66)
+40: vpmulld Vx,Hx,Wx (66),(v1)
+41: vphminposuw Vdq,Wdq (66),(v1)
+42:
+43:
+44:
+45: vpsrlvd/q Vx,Hx,Wx (66),(v)
+46: vpsravd Vx,Hx,Wx (66),(v)
+47: vpsllvd/q Vx,Hx,Wx (66),(v)
+# Skip 0x48-0x57
+58: vpbroadcastd Vx,Wx (66),(v)
+59: vpbroadcastq Vx,Wx (66),(v)
+5a: vbroadcasti128 Vqq,Mdq (66),(v)
+# Skip 0x5b-0x77
+78: vpbroadcastb Vx,Wx (66),(v)
+79: vpbroadcastw Vx,Wx (66),(v)
+# Skip 0x7a-0x7f
+80: INVEPT Gy,Mdq (66)
+81: INVPID Gy,Mdq (66)
+82: INVPCID Gy,Mdq (66)
+8c: vpmaskmovd/q Vx,Hx,Mx (66),(v)
+8e: vpmaskmovd/q Mx,Vx,Hx (66),(v)
 # 0x0f 0x38 0x90-0xbf (FMA)
-96: vfmaddsub132pd/ps /r (66),(VEX)
-97: vfmsubadd132pd/ps /r (66),(VEX)
-98: vfmadd132pd/ps /r (66),(VEX)
-99: vfmadd132sd/ss /r (66),(VEX),(o128)
-9a: vfmsub132pd/ps /r (66),(VEX)
-9b: vfmsub132sd/ss /r (66),(VEX),(o128)
-9c: vfnmadd132pd/ps /r (66),(VEX)
-9d: vfnmadd132sd/ss /r (66),(VEX),(o128)
-9e: vfnmsub132pd/ps /r (66),(VEX)
-9f: vfnmsub132sd/ss /r (66),(VEX),(o128)
-a6: vfmaddsub213pd/ps /r (66),(VEX)
-a7: vfmsubadd213pd/ps /r (66),(VEX)
-a8: vfmadd213pd/ps /r (66),(VEX)
-a9: vfmadd213sd/ss /r (66),(VEX),(o128)
-aa: vfmsub213pd/ps /r (66),(VEX)
-ab: vfmsub213sd/ss /r (66),(VEX),(o128)
-ac: vfnmadd213pd/ps /r (66),(VEX)
-ad: vfnmadd213sd/ss /r (66),(VEX),(o128)
-ae: vfnmsub213pd/ps /r (66),(VEX)
-af: vfnmsub213sd/ss /r (66),(VEX),(o128)
-b6: vfmaddsub231pd/ps /r (66),(VEX)
-b7: vfmsubadd231pd/ps /r (66),(VEX)
-b8: vfmadd231pd/ps /r (66),(VEX)
-b9: vfmadd231sd/ss /r (66),(VEX),(o128)
-ba: vfmsub231pd/ps /r (66),(VEX)
-bb: vfmsub231sd/ss /r (66),(VEX),(o128)
-bc: vfnmadd231pd/ps /r (66),(VEX)
-bd: vfnmadd231sd/ss /r (66),(VEX),(o128)
-be: vfnmsub231pd/ps /r (66),(VEX)
-bf: vfnmsub231sd/ss /r (66),(VEX),(o128)
+90: vgatherdd/q Vx,Hx,Wx (66),(v)
+91: vgatherqd/q Vx,Hx,Wx (66),(v)
+92: vgatherdps/d Vx,Hx,Wx (66),(v)
+93: vgatherqps/d Vx,Hx,Wx (66),(v)
+94:
+95:
+96: vfmaddsub132ps/d Vx,Hx,Wx (66),(v)
+97: vfmsubadd132ps/d Vx,Hx,Wx (66),(v)
+98: vfmadd132ps/d Vx,Hx,Wx (66),(v)
+99: vfmadd132ss/d Vx,Hx,Wx (66),(v),(v1)
+9a: vfmsub132ps/d Vx,Hx,Wx (66),(v)
+9b: vfmsub132ss/d Vx,Hx,Wx (66),(v),(v1)
+9c: vfnmadd132ps/d Vx,Hx,Wx (66),(v)
+9d: vfnmadd132ss/d Vx,Hx,Wx (66),(v),(v1)
+9e: vfnmsub132ps/d Vx,Hx,Wx (66),(v)
+9f: vfnmsub132ss/d Vx,Hx,Wx (66),(v),(v1)
+a6: vfmaddsub213ps/d Vx,Hx,Wx (66),(v)
+a7: vfmsubadd213ps/d Vx,Hx,Wx (66),(v)
+a8: vfmadd213ps/d Vx,Hx,Wx (66),(v)
+a9: vfmadd213ss/d Vx,Hx,Wx (66),(v),(v1)
+aa: vfmsub213ps/d Vx,Hx,Wx (66),(v)
+ab: vfmsub213ss/d Vx,Hx,Wx (66),(v),(v1)
+ac: vfnmadd213ps/d Vx,Hx,Wx (66),(v)
+ad: vfnmadd213ss/d Vx,Hx,Wx (66),(v),(v1)
+ae: vfnmsub213ps/d Vx,Hx,Wx (66),(v)
+af: vfnmsub213ss/d Vx,Hx,Wx (66),(v),(v1)
+b6: vfmaddsub231ps/d Vx,Hx,Wx (66),(v)
+b7: vfmsubadd231ps/d Vx,Hx,Wx (66),(v)
+b8: vfmadd231ps/d Vx,Hx,Wx (66),(v)
+b9: vfmadd231ss/d Vx,Hx,Wx (66),(v),(v1)
+ba: vfmsub231ps/d Vx,Hx,Wx (66),(v)
+bb: vfmsub231ss/d Vx,Hx,Wx (66),(v),(v1)
+bc: vfnmadd231ps/d Vx,Hx,Wx (66),(v)
+bd: vfnmadd231ss/d Vx,Hx,Wx (66),(v),(v1)
+be: vfnmsub231ps/d Vx,Hx,Wx (66),(v)
+bf: vfnmsub231ss/d Vx,Hx,Wx (66),(v),(v1)
 # 0x0f 0x38 0xc0-0xff
-db: aesimc Vdq,Wdq (66),(VEX),(o128)
-dc: aesenc Vdq,Wdq (66),(VEX),(o128)
-dd: aesenclast Vdq,Wdq (66),(VEX),(o128)
-de: aesdec Vdq,Wdq (66),(VEX),(o128)
-df: aesdeclast Vdq,Wdq (66),(VEX),(o128)
-f0: MOVBE Gv,Mv | CRC32 Gd,Eb (F2)
-f1: MOVBE Mv,Gv | CRC32 Gd,Ev (F2)
+db: VAESIMC Vdq,Wdq (66),(v1)
+dc: VAESENC Vdq,Hdq,Wdq (66),(v1)
+dd: VAESENCLAST Vdq,Hdq,Wdq (66),(v1)
+de: VAESDEC Vdq,Hdq,Wdq (66),(v1)
+df: VAESDECLAST Vdq,Hdq,Wdq (66),(v1)
+f0: MOVBE Gy,My | MOVBE Gw,Mw (66) | CRC32 Gd,Eb (F2)
+f1: MOVBE My,Gy | MOVBE Mw,Gw (66) | CRC32 Gd,Ey (F2)
+f3: ANDN Gy,By,Ey (v)
+f4: Grp17 (1A)
+f5: BZHI Gy,Ey,By (v) | PEXT Gy,By,Ey (F3),(v) | PDEP Gy,By,Ey (F2),(v)
+f6: MULX By,Gy,rDX,Ey (F2),(v)
+f7: BEXTR Gy,Ey,By (v) | SHLX Gy,Ey,By (66),(v) | SARX Gy,Ey,By (F3),(v) | SHRX Gy,Ey,By (F2),(v)
 EndTable
 
 Table: 3-byte opcode 2 (0x0f 0x3a)
 Referrer: 3-byte escape 2
 AVXcode: 3
 # 0x0f 0x3a 0x00-0xff
-04: vpermilps /r,Ib (66),(oVEX)
-05: vpermilpd /r,Ib (66),(oVEX)
-06: vperm2f128 /r,Ib (66),(oVEX),(o256)
-08: roundps Vdq,Wdq,Ib (66),(VEX)
-09: roundpd Vdq,Wdq,Ib (66),(VEX)
-0a: roundss Vss,Wss,Ib (66),(VEX),(o128)
-0b: roundsd Vsd,Wsd,Ib (66),(VEX),(o128)
-0c: blendps Vdq,Wdq,Ib (66),(VEX)
-0d: blendpd Vdq,Wdq,Ib (66),(VEX)
-0e: pblendw Vdq,Wdq,Ib (66),(VEX),(o128)
-0f: palignr Pq,Qq,Ib | palignr Vdq,Wdq,Ib (66),(VEX),(o128)
-14: pextrb Rd/Mb,Vdq,Ib (66),(VEX),(o128)
-15: pextrw Rd/Mw,Vdq,Ib (66),(VEX),(o128)
-16: pextrd/pextrq Ed/q,Vdq,Ib (66),(VEX),(o128)
-17: extractps Ed,Vdq,Ib (66),(VEX),(o128)
-18: vinsertf128 /r,Ib (66),(oVEX),(o256)
-19: vextractf128 /r,Ib (66),(oVEX),(o256)
-20: pinsrb Vdq,Rd/q/Mb,Ib (66),(VEX),(o128)
-21: insertps Vdq,Udq/Md,Ib (66),(VEX),(o128)
-22: pinsrd/pinsrq Vdq,Ed/q,Ib (66),(VEX),(o128)
-40: dpps Vdq,Wdq,Ib (66),(VEX)
-41: dppd Vdq,Wdq,Ib (66),(VEX),(o128)
-42: mpsadbw Vdq,Wdq,Ib (66),(VEX),(o128)
-44: pclmulq Vdq,Wdq,Ib (66),(VEX),(o128)
-4a: vblendvps /r,Ib (66),(oVEX)
-4b: vblendvpd /r,Ib (66),(oVEX)
-4c: vpblendvb /r,Ib (66),(oVEX),(o128)
-60: pcmpestrm Vdq,Wdq,Ib (66),(VEX),(o128)
-61: pcmpestri Vdq,Wdq,Ib (66),(VEX),(o128)
-62: pcmpistrm Vdq,Wdq,Ib (66),(VEX),(o128)
-63: pcmpistri Vdq,Wdq,Ib (66),(VEX),(o128)
-df: aeskeygenassist Vdq,Wdq,Ib (66),(VEX),(o128)
+00: vpermq Vqq,Wqq,Ib (66),(v)
+01: vpermpd Vqq,Wqq,Ib (66),(v)
+02: vpblendd Vx,Hx,Wx,Ib (66),(v)
+03:
+04: vpermilps Vx,Wx,Ib (66),(v)
+05: vpermilpd Vx,Wx,Ib (66),(v)
+06: vperm2f128 Vqq,Hqq,Wqq,Ib (66),(v)
+07:
+08: vroundps Vx,Wx,Ib (66)
+09: vroundpd Vx,Wx,Ib (66)
+0a: vroundss Vss,Wss,Ib (66),(v1)
+0b: vroundsd Vsd,Wsd,Ib (66),(v1)
+0c: vblendps Vx,Hx,Wx,Ib (66)
+0d: vblendpd Vx,Hx,Wx,Ib (66)
+0e: vpblendw Vx,Hx,Wx,Ib (66),(v1)
+0f: palignr Pq,Qq,Ib | vpalignr Vx,Hx,Wx,Ib (66),(v1)
+14: vpextrb Rd/Mb,Vdq,Ib (66),(v1)
+15: vpextrw Rd/Mw,Vdq,Ib (66),(v1)
+16: vpextrd/q Ey,Vdq,Ib (66),(v1)
+17: vextractps Ed,Vdq,Ib (66),(v1)
+18: vinsertf128 Vqq,Hqq,Wqq,Ib (66),(v)
+19: vextractf128 Wdq,Vqq,Ib (66),(v)
+1d: vcvtps2ph Wx,Vx,Ib (66),(v)
+20: vpinsrb Vdq,Hdq,Ry/Mb,Ib (66),(v1)
+21: vinsertps Vdq,Hdq,Udq/Md,Ib (66),(v1)
+22: vpinsrd/q Vdq,Hdq,Ey,Ib (66),(v1)
+38: vinserti128 Vqq,Hqq,Wqq,Ib (66),(v)
+39: vextracti128 Wdq,Vqq,Ib (66),(v)
+40: vdpps Vx,Hx,Wx,Ib (66)
+41: vdppd Vdq,Hdq,Wdq,Ib (66),(v1)
+42: vmpsadbw Vx,Hx,Wx,Ib (66),(v1)
+44: vpclmulqdq Vdq,Hdq,Wdq,Ib (66),(v1)
+46: vperm2i128 Vqq,Hqq,Wqq,Ib (66),(v)
+4a: vblendvps Vx,Hx,Wx,Lx (66),(v)
+4b: vblendvpd Vx,Hx,Wx,Lx (66),(v)
+4c: vpblendvb Vx,Hx,Wx,Lx (66),(v1)
+60: vpcmpestrm Vdq,Wdq,Ib (66),(v1)
+61: vpcmpestri Vdq,Wdq,Ib (66),(v1)
+62: vpcmpistrm Vdq,Wdq,Ib (66),(v1)
+63: vpcmpistri Vdq,Wdq,Ib (66),(v1)
+df: VAESKEYGEN Vdq,Wdq,Ib (66),(v1)
+f0: RORX Gy,Ey,Ib (F2),(v)
 EndTable
 
 GrpTable: Grp1
@@ -790,7 +843,7 @@ GrpTable: Grp5
 2: CALLN Ev (f64)
 3: CALLF Ep
 4: JMPN Ev (f64)
-5: JMPF Ep
+5: JMPF Mp
 6: PUSH Ev (d64)
 7:
 EndTable
@@ -807,7 +860,7 @@ EndTable
 GrpTable: Grp7
 0: SGDT Ms | VMCALL (001),(11B) | VMLAUNCH (010),(11B) | VMRESUME (011),(11B) | VMXOFF (100),(11B)
 1: SIDT Ms | MONITOR (000),(11B) | MWAIT (001)
-2: LGDT Ms | XGETBV (000),(11B) | XSETBV (001),(11B)
+2: LGDT Ms | XGETBV (000),(11B) | XSETBV (001),(11B) | VMFUNC (100),(11B)
 3: LIDT Ms
 4: SMSW Mw/Rv
 5:
@@ -824,44 +877,45 @@ EndTable
 
 GrpTable: Grp9
 1: CMPXCHG8B/16B Mq/Mdq
-6: VMPTRLD Mq | VMCLEAR Mq (66) | VMXON Mq (F3)
-7: VMPTRST Mq
+6: VMPTRLD Mq | VMCLEAR Mq (66) | VMXON Mq (F3) | RDRAND Rv (11B)
+7: VMPTRST Mq | VMPTRST Mq (F3)
 EndTable
 
 GrpTable: Grp10
 EndTable
 
 GrpTable: Grp11
+# Note: the operands are given by group opcode
 0: MOV
 EndTable
 
 GrpTable: Grp12
-2: psrlw Nq,Ib (11B) | psrlw Udq,Ib (66),(11B),(VEX),(o128)
-4: psraw Nq,Ib (11B) | psraw Udq,Ib (66),(11B),(VEX),(o128)
-6: psllw Nq,Ib (11B) | psllw Udq,Ib (66),(11B),(VEX),(o128)
+2: psrlw Nq,Ib (11B) | vpsrlw Hx,Ux,Ib (66),(11B),(v1)
+4: psraw Nq,Ib (11B) | vpsraw Hx,Ux,Ib (66),(11B),(v1)
+6: psllw Nq,Ib (11B) | vpsllw Hx,Ux,Ib (66),(11B),(v1)
 EndTable
 
 GrpTable: Grp13
-2: psrld Nq,Ib (11B) | psrld Udq,Ib (66),(11B),(VEX),(o128)
-4: psrad Nq,Ib (11B) | psrad Udq,Ib (66),(11B),(VEX),(o128)
-6: pslld Nq,Ib (11B) | pslld Udq,Ib (66),(11B),(VEX),(o128)
+2: psrld Nq,Ib (11B) | vpsrld Hx,Ux,Ib (66),(11B),(v1)
+4: psrad Nq,Ib (11B) | vpsrad Hx,Ux,Ib (66),(11B),(v1)
+6: pslld Nq,Ib (11B) | vpslld Hx,Ux,Ib (66),(11B),(v1)
 EndTable
 
 GrpTable: Grp14
-2: psrlq Nq,Ib (11B) | psrlq Udq,Ib (66),(11B),(VEX),(o128)
-3: psrldq Udq,Ib (66),(11B),(VEX),(o128)
-6: psllq Nq,Ib (11B) | psllq Udq,Ib (66),(11B),(VEX),(o128)
-7: pslldq Udq,Ib (66),(11B),(VEX),(o128)
+2: psrlq Nq,Ib (11B) | vpsrlq Hx,Ux,Ib (66),(11B),(v1)
+3: vpsrldq Hx,Ux,Ib (66),(11B),(v1)
+6: psllq Nq,Ib (11B) | vpsllq Hx,Ux,Ib (66),(11B),(v1)
+7: vpslldq Hx,Ux,Ib (66),(11B),(v1)
 EndTable
 
 GrpTable: Grp15
-0: fxsave
-1: fxstor
-2: ldmxcsr (VEX)
-3: stmxcsr (VEX)
+0: fxsave | RDFSBASE Ry (F3),(11B)
+1: fxstor | RDGSBASE Ry (F3),(11B)
+2: vldmxcsr Md (v1) | WRFSBASE Ry (F3),(11B)
+3: vstmxcsr Md (v1) | WRGSBASE Ry (F3),(11B)
 4: XSAVE
 5: XRSTOR | lfence (11B)
-6: mfence (11B)
+6: XSAVEOPT | mfence (11B)
 7: clflush | sfence (11B)
 EndTable
 
@@ -872,6 +926,12 @@ GrpTable: Grp16
 3: prefetch T2
 EndTable
 
+GrpTable: Grp17
+1: BLSR By,Ey (v)
+2: BLSMSK By,Ey (v)
+3: BLSI By,Ey (v)
+EndTable
+
 # AMD's Prefetch Group
 GrpTable: GrpP
 0: PREFETCH

arch/x86/oprofile/Makefile

@@ -4,9 +4,8 @@ DRIVER_OBJS = $(addprefix ../../../drivers/oprofile/, \
 		oprof.o cpu_buffer.o buffer_sync.o \
 		event_buffer.o oprofile_files.o \
 		oprofilefs.o oprofile_stats.o  \
-		timer_int.o )
+		timer_int.o nmi_timer_int.o )
 
 oprofile-y				:= $(DRIVER_OBJS) init.o backtrace.o
 oprofile-$(CONFIG_X86_LOCAL_APIC) 	+= nmi_int.o op_model_amd.o \
 					   op_model_ppro.o op_model_p4.o
-oprofile-$(CONFIG_X86_IO_APIC)		+= nmi_timer_int.o

arch/x86/oprofile/init.c

@@ -16,37 +16,23 @@
  * with the NMI mode driver.
  */
 
+#ifdef CONFIG_X86_LOCAL_APIC
 extern int op_nmi_init(struct oprofile_operations *ops);
-extern int op_nmi_timer_init(struct oprofile_operations *ops);
 extern void op_nmi_exit(void);
-extern void x86_backtrace(struct pt_regs * const regs, unsigned int depth);
+#else
+static int op_nmi_init(struct oprofile_operations *ops) { return -ENODEV; }
+static void op_nmi_exit(void) { }
+#endif
 
-static int nmi_timer;
+extern void x86_backtrace(struct pt_regs * const regs, unsigned int depth);
 
 int __init oprofile_arch_init(struct oprofile_operations *ops)
 {
-	int ret;
-
-	ret = -ENODEV;
-
-#ifdef CONFIG_X86_LOCAL_APIC
-	ret = op_nmi_init(ops);
-#endif
-	nmi_timer = (ret != 0);
-#ifdef CONFIG_X86_IO_APIC
-	if (nmi_timer)
-		ret = op_nmi_timer_init(ops);
-#endif
 	ops->backtrace = x86_backtrace;
-
-	return ret;
+	return op_nmi_init(ops);
 }
 
-
 void oprofile_arch_exit(void)
 {
-#ifdef CONFIG_X86_LOCAL_APIC
-	if (!nmi_timer)
-		op_nmi_exit();
-#endif
+	op_nmi_exit();
 }

arch/x86/oprofile/nmi_int.c

@@ -595,24 +595,36 @@ static int __init p4_init(char **cpu_type)
 	return 0;
 }
 
-static int force_arch_perfmon;
-static int force_cpu_type(const char *str, struct kernel_param *kp)
+enum __force_cpu_type {
+	reserved = 0,		/* do not force */
+	timer,
+	arch_perfmon,
+};
+
+static int force_cpu_type;
+
+static int set_cpu_type(const char *str, struct kernel_param *kp)
 {
-	if (!strcmp(str, "arch_perfmon")) {
-		force_arch_perfmon = 1;
+	if (!strcmp(str, "timer")) {
+		force_cpu_type = timer;
+		printk(KERN_INFO "oprofile: forcing NMI timer mode\n");
+	} else if (!strcmp(str, "arch_perfmon")) {
+		force_cpu_type = arch_perfmon;
 		printk(KERN_INFO "oprofile: forcing architectural perfmon\n");
+	} else {
+		force_cpu_type = 0;
 	}
 
 	return 0;
 }
-module_param_call(cpu_type, force_cpu_type, NULL, NULL, 0);
+module_param_call(cpu_type, set_cpu_type, NULL, NULL, 0);
 
 static int __init ppro_init(char **cpu_type)
 {
 	__u8 cpu_model = boot_cpu_data.x86_model;
 	struct op_x86_model_spec *spec = &op_ppro_spec;	/* default */
 
-	if (force_arch_perfmon && cpu_has_arch_perfmon)
+	if (force_cpu_type == arch_perfmon && cpu_has_arch_perfmon)
 		return 0;
 
 	/*
@@ -679,6 +691,9 @@ int __init op_nmi_init(struct oprofile_operations *ops)
 	if (!cpu_has_apic)
 		return -ENODEV;
 
+	if (force_cpu_type == timer)
+		return -ENODEV;
+
 	switch (vendor) {
 	case X86_VENDOR_AMD:
 		/* Needs to be at least an Athlon (or hammer in 32bit mode) */

arch/x86/oprofile/nmi_timer_int.c

@@ -1,50 +0,0 @@
-/**
- * @file nmi_timer_int.c
- *
- * @remark Copyright 2003 OProfile authors
- * @remark Read the file COPYING
- *
- * @author Zwane Mwaikambo <zwane@linuxpower.ca>
- */
-
-#include <linux/init.h>
-#include <linux/smp.h>
-#include <linux/errno.h>
-#include <linux/oprofile.h>
-#include <linux/rcupdate.h>
-#include <linux/kdebug.h>
-
-#include <asm/nmi.h>
-#include <asm/apic.h>
-#include <asm/ptrace.h>
-
-static int profile_timer_exceptions_notify(unsigned int val, struct pt_regs *regs)
-{
-	oprofile_add_sample(regs, 0);
-	return NMI_HANDLED;
-}
-
-static int timer_start(void)
-{
-	if (register_nmi_handler(NMI_LOCAL, profile_timer_exceptions_notify,
-					0, "oprofile-timer"))
-		return 1;
-	return 0;
-}
-
-
-static void timer_stop(void)
-{
-	unregister_nmi_handler(NMI_LOCAL, "oprofile-timer");
-	synchronize_sched();  /* Allow already-started NMIs to complete. */
-}
-
-
-int __init op_nmi_timer_init(struct oprofile_operations *ops)
-{
-	ops->start = timer_start;
-	ops->stop = timer_stop;
-	ops->cpu_type = "timer";
-	printk(KERN_INFO "oprofile: using NMI timer interrupt.\n");
-	return 0;
-}

arch/x86/tools/Makefile

@@ -18,14 +18,21 @@ chkobjdump = $(srctree)/arch/x86/tools/chkobjdump.awk
 quiet_cmd_posttest = TEST    $@
       cmd_posttest = ($(OBJDUMP) -v | $(AWK) -f $(chkobjdump)) || $(OBJDUMP) -d -j .text $(objtree)/vmlinux | $(AWK) -f $(distill_awk) | $(obj)/test_get_len $(posttest_64bit) $(posttest_verbose)
 
-posttest: $(obj)/test_get_len vmlinux
+quiet_cmd_sanitytest = TEST    $@
+      cmd_sanitytest = $(obj)/insn_sanity $(posttest_64bit) -m 1000000
+
+posttest: $(obj)/test_get_len vmlinux $(obj)/insn_sanity
 	$(call cmd,posttest)
+	$(call cmd,sanitytest)
 
-hostprogs-y	:= test_get_len
+hostprogs-y	+= test_get_len insn_sanity
 
 # -I needed for generated C source and C source which in the kernel tree.
 HOSTCFLAGS_test_get_len.o := -Wall -I$(objtree)/arch/x86/lib/ -I$(srctree)/arch/x86/include/ -I$(srctree)/arch/x86/lib/ -I$(srctree)/include/
 
+HOSTCFLAGS_insn_sanity.o := -Wall -I$(objtree)/arch/x86/lib/ -I$(srctree)/arch/x86/include/ -I$(srctree)/arch/x86/lib/ -I$(srctree)/include/
+
 # Dependencies are also needed.
 $(obj)/test_get_len.o: $(srctree)/arch/x86/lib/insn.c $(srctree)/arch/x86/lib/inat.c $(srctree)/arch/x86/include/asm/inat_types.h $(srctree)/arch/x86/include/asm/inat.h $(srctree)/arch/x86/include/asm/insn.h $(objtree)/arch/x86/lib/inat-tables.c
 
+$(obj)/insn_sanity.o: $(srctree)/arch/x86/lib/insn.c $(srctree)/arch/x86/lib/inat.c $(srctree)/arch/x86/include/asm/inat_types.h $(srctree)/arch/x86/include/asm/inat.h $(srctree)/arch/x86/include/asm/insn.h $(objtree)/arch/x86/lib/inat-tables.c

arch/x86/tools/gen-insn-attr-x86.awk

@@ -47,7 +47,7 @@ BEGIN {
 	sep_expr = "^\\|$"
 	group_expr = "^Grp[0-9A-Za-z]+"
 
-	imm_expr = "^[IJAO][a-z]"
+	imm_expr = "^[IJAOL][a-z]"
 	imm_flag["Ib"] = "INAT_MAKE_IMM(INAT_IMM_BYTE)"
 	imm_flag["Jb"] = "INAT_MAKE_IMM(INAT_IMM_BYTE)"
 	imm_flag["Iw"] = "INAT_MAKE_IMM(INAT_IMM_WORD)"
@@ -59,6 +59,7 @@ BEGIN {
 	imm_flag["Iv"] = "INAT_MAKE_IMM(INAT_IMM_VWORD)"
 	imm_flag["Ob"] = "INAT_MOFFSET"
 	imm_flag["Ov"] = "INAT_MOFFSET"
+	imm_flag["Lx"] = "INAT_MAKE_IMM(INAT_IMM_BYTE)"
 
 	modrm_expr = "^([CDEGMNPQRSUVW/][a-z]+|NTA|T[012])"
 	force64_expr = "\\([df]64\\)"
@@ -70,8 +71,12 @@ BEGIN {
 	lprefix3_expr = "\\(F2\\)"
 	max_lprefix = 4
 
-	vexok_expr = "\\(VEX\\)"
-	vexonly_expr = "\\(oVEX\\)"
+	# All opcodes starting with lower-case 'v' or with (v1) superscript
+	# accepts VEX prefix
+	vexok_opcode_expr = "^v.*"
+	vexok_expr = "\\(v1\\)"
+	# All opcodes with (v) superscript supports *only* VEX prefix
+	vexonly_expr = "\\(v\\)"
 
 	prefix_expr = "\\(Prefix\\)"
 	prefix_num["Operand-Size"] = "INAT_PFX_OPNDSZ"
@@ -85,8 +90,8 @@ BEGIN {
 	prefix_num["SEG=GS"] = "INAT_PFX_GS"
 	prefix_num["SEG=SS"] = "INAT_PFX_SS"
 	prefix_num["Address-Size"] = "INAT_PFX_ADDRSZ"
-	prefix_num["2bytes-VEX"] = "INAT_PFX_VEX2"
-	prefix_num["3bytes-VEX"] = "INAT_PFX_VEX3"
+	prefix_num["VEX+1byte"] = "INAT_PFX_VEX2"
+	prefix_num["VEX+2byte"] = "INAT_PFX_VEX3"
 
 	clear_vars()
 }
@@ -310,12 +315,10 @@ function convert_operands(count,opnd,       i,j,imm,mod)
 		if (match(opcode, fpu_expr))
 			flags = add_flags(flags, "INAT_MODRM")
 
-		# check VEX only code
+		# check VEX codes
 		if (match(ext, vexonly_expr))
 			flags = add_flags(flags, "INAT_VEXOK | INAT_VEXONLY")
-
-		# check VEX only code
-		if (match(ext, vexok_expr))
+		else if (match(ext, vexok_expr) || match(opcode, vexok_opcode_expr))
 			flags = add_flags(flags, "INAT_VEXOK")
 
 		# check prefixes

arch/x86/tools/insn_sanity.c

@@ -0,0 +1,275 @@
+/*
+ * x86 decoder sanity test - based on test_get_insn.c
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright (C) IBM Corporation, 2009
+ * Copyright (C) Hitachi, Ltd., 2011
+ */
+
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <assert.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+
+#define unlikely(cond) (cond)
+#define ARRAY_SIZE(a)	(sizeof(a)/sizeof(a[0]))
+
+#include <asm/insn.h>
+#include <inat.c>
+#include <insn.c>
+
+/*
+ * Test of instruction analysis against tampering.
+ * Feed random binary to instruction decoder and ensure not to
+ * access out-of-instruction-buffer.
+ */
+
+#define DEFAULT_MAX_ITER	10000
+#define INSN_NOP 0x90
+
+static const char	*prog;		/* Program name */
+static int		verbose;	/* Verbosity */
+static int		x86_64;		/* x86-64 bit mode flag */
+static unsigned int	seed;		/* Random seed */
+static unsigned long	iter_start;	/* Start of iteration number */
+static unsigned long	iter_end = DEFAULT_MAX_ITER;	/* End of iteration number */
+static FILE		*input_file;	/* Input file name */
+
+static void usage(const char *err)
+{
+	if (err)
+		fprintf(stderr, "Error: %s\n\n", err);
+	fprintf(stderr, "Usage: %s [-y|-n|-v] [-s seed[,no]] [-m max] [-i input]\n", prog);
+	fprintf(stderr, "\t-y	64bit mode\n");
+	fprintf(stderr, "\t-n	32bit mode\n");
+	fprintf(stderr, "\t-v	Verbosity(-vv dumps any decoded result)\n");
+	fprintf(stderr, "\t-s	Give a random seed (and iteration number)\n");
+	fprintf(stderr, "\t-m	Give a maximum iteration number\n");
+	fprintf(stderr, "\t-i	Give an input file with decoded binary\n");
+	exit(1);
+}
+
+static void dump_field(FILE *fp, const char *name, const char *indent,
+		       struct insn_field *field)
+{
+	fprintf(fp, "%s.%s = {\n", indent, name);
+	fprintf(fp, "%s\t.value = %d, bytes[] = {%x, %x, %x, %x},\n",
+		indent, field->value, field->bytes[0], field->bytes[1],
+		field->bytes[2], field->bytes[3]);
+	fprintf(fp, "%s\t.got = %d, .nbytes = %d},\n", indent,
+		field->got, field->nbytes);
+}
+
+static void dump_insn(FILE *fp, struct insn *insn)
+{
+	fprintf(fp, "Instruction = {\n");
+	dump_field(fp, "prefixes", "\t",	&insn->prefixes);
+	dump_field(fp, "rex_prefix", "\t",	&insn->rex_prefix);
+	dump_field(fp, "vex_prefix", "\t",	&insn->vex_prefix);
+	dump_field(fp, "opcode", "\t",		&insn->opcode);
+	dump_field(fp, "modrm", "\t",		&insn->modrm);
+	dump_field(fp, "sib", "\t",		&insn->sib);
+	dump_field(fp, "displacement", "\t",	&insn->displacement);
+	dump_field(fp, "immediate1", "\t",	&insn->immediate1);
+	dump_field(fp, "immediate2", "\t",	&insn->immediate2);
+	fprintf(fp, "\t.attr = %x, .opnd_bytes = %d, .addr_bytes = %d,\n",
+		insn->attr, insn->opnd_bytes, insn->addr_bytes);
+	fprintf(fp, "\t.length = %d, .x86_64 = %d, .kaddr = %p}\n",
+		insn->length, insn->x86_64, insn->kaddr);
+}
+
+static void dump_stream(FILE *fp, const char *msg, unsigned long nr_iter,
+			unsigned char *insn_buf, struct insn *insn)
+{
+	int i;
+
+	fprintf(fp, "%s:\n", msg);
+
+	dump_insn(fp, insn);
+
+	fprintf(fp, "You can reproduce this with below command(s);\n");
+
+	/* Input a decoded instruction sequence directly */
+	fprintf(fp, " $ echo ");
+	for (i = 0; i < MAX_INSN_SIZE; i++)
+		fprintf(fp, " %02x", insn_buf[i]);
+	fprintf(fp, " | %s -i -\n", prog);
+
+	if (!input_file) {
+		fprintf(fp, "Or \n");
+		/* Give a seed and iteration number */
+		fprintf(fp, " $ %s -s 0x%x,%lu\n", prog, seed, nr_iter);
+	}
+}
+
+static void init_random_seed(void)
+{
+	int fd;
+
+	fd = open("/dev/urandom", O_RDONLY);
+	if (fd < 0)
+		goto fail;
+
+	if (read(fd, &seed, sizeof(seed)) != sizeof(seed))
+		goto fail;
+
+	close(fd);
+	return;
+fail:
+	usage("Failed to open /dev/urandom");
+}
+
+/* Read given instruction sequence from the input file */
+static int read_next_insn(unsigned char *insn_buf)
+{
+	char buf[256]  = "", *tmp;
+	int i;
+
+	tmp = fgets(buf, ARRAY_SIZE(buf), input_file);
+	if (tmp == NULL || feof(input_file))
+		return 0;
+
+	for (i = 0; i < MAX_INSN_SIZE; i++) {
+		insn_buf[i] = (unsigned char)strtoul(tmp, &tmp, 16);
+		if (*tmp != ' ')
+			break;
+	}
+
+	return i;
+}
+
+static int generate_insn(unsigned char *insn_buf)
+{
+	int i;
+
+	if (input_file)
+		return read_next_insn(insn_buf);
+
+	/* Fills buffer with random binary up to MAX_INSN_SIZE */
+	for (i = 0; i < MAX_INSN_SIZE - 1; i += 2)
+		*(unsigned short *)(&insn_buf[i]) = random() & 0xffff;
+
+	while (i < MAX_INSN_SIZE)
+		insn_buf[i++] = random() & 0xff;
+
+	return i;
+}
+
+static void parse_args(int argc, char **argv)
+{
+	int c;
+	char *tmp = NULL;
+	int set_seed = 0;
+
+	prog = argv[0];
+	while ((c = getopt(argc, argv, "ynvs:m:i:")) != -1) {
+		switch (c) {
+		case 'y':
+			x86_64 = 1;
+			break;
+		case 'n':
+			x86_64 = 0;
+			break;
+		case 'v':
+			verbose++;
+			break;
+		case 'i':
+			if (strcmp("-", optarg) == 0)
+				input_file = stdin;
+			else
+				input_file = fopen(optarg, "r");
+			if (!input_file)
+				usage("Failed to open input file");
+			break;
+		case 's':
+			seed = (unsigned int)strtoul(optarg, &tmp, 0);
+			if (*tmp == ',') {
+				optarg = tmp + 1;
+				iter_start = strtoul(optarg, &tmp, 0);
+			}
+			if (*tmp != '\0' || tmp == optarg)
+				usage("Failed to parse seed");
+			set_seed = 1;
+			break;
+		case 'm':
+			iter_end = strtoul(optarg, &tmp, 0);
+			if (*tmp != '\0' || tmp == optarg)
+				usage("Failed to parse max_iter");
+			break;
+		default:
+			usage(NULL);
+		}
+	}
+
+	/* Check errors */
+	if (iter_end < iter_start)
+		usage("Max iteration number must be bigger than iter-num");
+
+	if (set_seed && input_file)
+		usage("Don't use input file (-i) with random seed (-s)");
+
+	/* Initialize random seed */
+	if (!input_file) {
+		if (!set_seed)	/* No seed is given */
+			init_random_seed();
+		srand(seed);
+	}
+}
+
+int main(int argc, char **argv)
+{
+	struct insn insn;
+	int insns = 0;
+	int errors = 0;
+	unsigned long i;
+	unsigned char insn_buf[MAX_INSN_SIZE * 2];
+
+	parse_args(argc, argv);
+
+	/* Prepare stop bytes with NOPs */
+	memset(insn_buf + MAX_INSN_SIZE, INSN_NOP, MAX_INSN_SIZE);
+
+	for (i = 0; i < iter_end; i++) {
+		if (generate_insn(insn_buf) <= 0)
+			break;
+
+		if (i < iter_start)	/* Skip to given iteration number */
+			continue;
+
+		/* Decode an instruction */
+		insn_init(&insn, insn_buf, x86_64);
+		insn_get_length(&insn);
+
+		if (insn.next_byte <= insn.kaddr ||
+		    insn.kaddr + MAX_INSN_SIZE < insn.next_byte) {
+			/* Access out-of-range memory */
+			dump_stream(stderr, "Error: Found an access violation", i, insn_buf, &insn);
+			errors++;
+		} else if (verbose && !insn_complete(&insn))
+			dump_stream(stdout, "Info: Found an undecodable input", i, insn_buf, &insn);
+		else if (verbose >= 2)
+			dump_insn(stdout, &insn);
+		insns++;
+	}
+
+	fprintf(stdout, "%s: decoded and checked %d %s instructions with %d errors (seed:0x%x)\n", (errors) ? "Failure" : "Success", insns, (input_file) ? "given" : "random", errors, seed);
+
+	return errors ? 1 : 0;
+}

drivers/oprofile/nmi_timer_int.c

@@ -0,0 +1,173 @@
+/**
+ * @file nmi_timer_int.c
+ *
+ * @remark Copyright 2011 Advanced Micro Devices, Inc.
+ *
+ * @author Robert Richter <robert.richter@amd.com>
+ */
+
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <linux/errno.h>
+#include <linux/oprofile.h>
+#include <linux/perf_event.h>
+
+#ifdef CONFIG_OPROFILE_NMI_TIMER
+
+static DEFINE_PER_CPU(struct perf_event *, nmi_timer_events);
+static int ctr_running;
+
+static struct perf_event_attr nmi_timer_attr = {
+	.type           = PERF_TYPE_HARDWARE,
+	.config         = PERF_COUNT_HW_CPU_CYCLES,
+	.size           = sizeof(struct perf_event_attr),
+	.pinned         = 1,
+	.disabled       = 1,
+};
+
+static void nmi_timer_callback(struct perf_event *event,
+			       struct perf_sample_data *data,
+			       struct pt_regs *regs)
+{
+	event->hw.interrupts = 0;       /* don't throttle interrupts */
+	oprofile_add_sample(regs, 0);
+}
+
+static int nmi_timer_start_cpu(int cpu)
+{
+	struct perf_event *event = per_cpu(nmi_timer_events, cpu);
+
+	if (!event) {
+		event = perf_event_create_kernel_counter(&nmi_timer_attr, cpu, NULL,
+							 nmi_timer_callback, NULL);
+		if (IS_ERR(event))
+			return PTR_ERR(event);
+		per_cpu(nmi_timer_events, cpu) = event;
+	}
+
+	if (event && ctr_running)
+		perf_event_enable(event);
+
+	return 0;
+}
+
+static void nmi_timer_stop_cpu(int cpu)
+{
+	struct perf_event *event = per_cpu(nmi_timer_events, cpu);
+
+	if (event && ctr_running)
+		perf_event_disable(event);
+}
+
+static int nmi_timer_cpu_notifier(struct notifier_block *b, unsigned long action,
+				  void *data)
+{
+	int cpu = (unsigned long)data;
+	switch (action) {
+	case CPU_DOWN_FAILED:
+	case CPU_ONLINE:
+		nmi_timer_start_cpu(cpu);
+		break;
+	case CPU_DOWN_PREPARE:
+		nmi_timer_stop_cpu(cpu);
+		break;
+	}
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block nmi_timer_cpu_nb = {
+	.notifier_call = nmi_timer_cpu_notifier
+};
+
+static int nmi_timer_start(void)
+{
+	int cpu;
+
+	get_online_cpus();
+	ctr_running = 1;
+	for_each_online_cpu(cpu)
+		nmi_timer_start_cpu(cpu);
+	put_online_cpus();
+
+	return 0;
+}
+
+static void nmi_timer_stop(void)
+{
+	int cpu;
+
+	get_online_cpus();
+	for_each_online_cpu(cpu)
+		nmi_timer_stop_cpu(cpu);
+	ctr_running = 0;
+	put_online_cpus();
+}
+
+static void nmi_timer_shutdown(void)
+{
+	struct perf_event *event;
+	int cpu;
+
+	get_online_cpus();
+	unregister_cpu_notifier(&nmi_timer_cpu_nb);
+	for_each_possible_cpu(cpu) {
+		event = per_cpu(nmi_timer_events, cpu);
+		if (!event)
+			continue;
+		perf_event_disable(event);
+		per_cpu(nmi_timer_events, cpu) = NULL;
+		perf_event_release_kernel(event);
+	}
+
+	put_online_cpus();
+}
+
+static int nmi_timer_setup(void)
+{
+	int cpu, err;
+	u64 period;
+
+	/* clock cycles per tick: */
+	period = (u64)cpu_khz * 1000;
+	do_div(period, HZ);
+	nmi_timer_attr.sample_period = period;
+
+	get_online_cpus();
+	err = register_cpu_notifier(&nmi_timer_cpu_nb);
+	if (err)
+		goto out;
+	/* can't attach events to offline cpus: */
+	for_each_online_cpu(cpu) {
+		err = nmi_timer_start_cpu(cpu);
+		if (err)
+			break;
+	}
+	if (err)
+		nmi_timer_shutdown();
+out:
+	put_online_cpus();
+	return err;
+}
+
+int __init op_nmi_timer_init(struct oprofile_operations *ops)
+{
+	int err = 0;
+
+	err = nmi_timer_setup();
+	if (err)
+		return err;
+	nmi_timer_shutdown();		/* only check, don't alloc */
+
+	ops->create_files	= NULL;
+	ops->setup		= nmi_timer_setup;
+	ops->shutdown		= nmi_timer_shutdown;
+	ops->start		= nmi_timer_start;
+	ops->stop		= nmi_timer_stop;
+	ops->cpu_type		= "timer";
+
+	printk(KERN_INFO "oprofile: using NMI timer interrupt.\n");
+
+	return 0;
+}
+
+#endif

drivers/oprofile/oprof.c

@@ -246,37 +246,31 @@ static int __init oprofile_init(void)
 	int err;
 
 	/* always init architecture to setup backtrace support */
+	timer_mode = 0;
 	err = oprofile_arch_init(&oprofile_ops);
+	if (!err) {
+		if (!timer && !oprofilefs_register())
+			return 0;
+		oprofile_arch_exit();
+	}
 
-	timer_mode = err || timer;	/* fall back to timer mode on errors */
-	if (timer_mode) {
-		if (!err)
-			oprofile_arch_exit();
+	/* setup timer mode: */
+	timer_mode = 1;
+	/* no nmi timer mode if oprofile.timer is set */
+	if (timer || op_nmi_timer_init(&oprofile_ops)) {
 		err = oprofile_timer_init(&oprofile_ops);
 		if (err)
 			return err;
 	}
 
-	err = oprofilefs_register();
-	if (!err)
-		return 0;
-
-	/* failed */
-	if (timer_mode)
-		oprofile_timer_exit();
-	else
-		oprofile_arch_exit();
-
-	return err;
+	return oprofilefs_register();
 }
 
 
 static void __exit oprofile_exit(void)
 {
 	oprofilefs_unregister();
-	if (timer_mode)
-		oprofile_timer_exit();
-	else
+	if (!timer_mode)
 		oprofile_arch_exit();
 }
 

drivers/oprofile/oprof.h

@@ -35,7 +35,15 @@ struct dentry;
 
 void oprofile_create_files(struct super_block *sb, struct dentry *root);
 int oprofile_timer_init(struct oprofile_operations *ops);
-void oprofile_timer_exit(void);
+#ifdef CONFIG_OPROFILE_NMI_TIMER
+int op_nmi_timer_init(struct oprofile_operations *ops);
+#else
+static inline int op_nmi_timer_init(struct oprofile_operations *ops)
+{
+	return -ENODEV;
+}
+#endif
+
 
 int oprofile_set_ulong(unsigned long *addr, unsigned long val);
 int oprofile_set_timeout(unsigned long time);

drivers/oprofile/timer_int.c

@@ -97,24 +97,24 @@ static struct notifier_block __refdata oprofile_cpu_notifier = {
 	.notifier_call = oprofile_cpu_notify,
 };
 
-int oprofile_timer_init(struct oprofile_operations *ops)
+static int oprofile_hrtimer_setup(void)
 {
-	int rc;
-
-	rc = register_hotcpu_notifier(&oprofile_cpu_notifier);
-	if (rc)
-		return rc;
-	ops->create_files = NULL;
-	ops->setup = NULL;
-	ops->shutdown = NULL;
-	ops->start = oprofile_hrtimer_start;
-	ops->stop = oprofile_hrtimer_stop;
-	ops->cpu_type = "timer";
-	printk(KERN_INFO "oprofile: using timer interrupt.\n");
-	return 0;
+	return register_hotcpu_notifier(&oprofile_cpu_notifier);
 }
 
-void oprofile_timer_exit(void)
+static void oprofile_hrtimer_shutdown(void)
 {
 	unregister_hotcpu_notifier(&oprofile_cpu_notifier);
 }
+
+int oprofile_timer_init(struct oprofile_operations *ops)
+{
+	ops->create_files	= NULL;
+	ops->setup		= oprofile_hrtimer_setup;
+	ops->shutdown		= oprofile_hrtimer_shutdown;
+	ops->start		= oprofile_hrtimer_start;
+	ops->stop		= oprofile_hrtimer_stop;
+	ops->cpu_type		= "timer";
+	printk(KERN_INFO "oprofile: using timer interrupt.\n");
+	return 0;
+}

include/linux/bitops.h

@@ -22,8 +22,14 @@ extern unsigned long __sw_hweight64(__u64 w);
 #include <asm/bitops.h>
 
 #define for_each_set_bit(bit, addr, size) \
-	for ((bit) = find_first_bit((addr), (size)); \
-	     (bit) < (size); \
+	for ((bit) = find_first_bit((addr), (size));		\
+	     (bit) < (size);					\
+	     (bit) = find_next_bit((addr), (size), (bit) + 1))
+
+/* same as for_each_set_bit() but use bit as value to start with */
+#define for_each_set_bit_cont(bit, addr, size) \
+	for ((bit) = find_next_bit((addr), (size), (bit));	\
+	     (bit) < (size);					\
 	     (bit) = find_next_bit((addr), (size), (bit) + 1))
 
 static __inline__ int get_bitmask_order(unsigned int count)

include/linux/jump_label.h

@@ -3,6 +3,7 @@
 
 #include <linux/types.h>
 #include <linux/compiler.h>
+#include <linux/workqueue.h>
 
 #if defined(CC_HAVE_ASM_GOTO) && defined(CONFIG_JUMP_LABEL)
 
@@ -14,6 +15,12 @@ struct jump_label_key {
 #endif
 };
 
+struct jump_label_key_deferred {
+	struct jump_label_key key;
+	unsigned long timeout;
+	struct delayed_work work;
+};
+
 # include <asm/jump_label.h>
 # define HAVE_JUMP_LABEL
 #endif	/* CC_HAVE_ASM_GOTO && CONFIG_JUMP_LABEL */
@@ -51,8 +58,11 @@ extern void arch_jump_label_transform_static(struct jump_entry *entry,
 extern int jump_label_text_reserved(void *start, void *end);
 extern void jump_label_inc(struct jump_label_key *key);
 extern void jump_label_dec(struct jump_label_key *key);
+extern void jump_label_dec_deferred(struct jump_label_key_deferred *key);
 extern bool jump_label_enabled(struct jump_label_key *key);
 extern void jump_label_apply_nops(struct module *mod);
+extern void jump_label_rate_limit(struct jump_label_key_deferred *key,
+		unsigned long rl);
 
 #else  /* !HAVE_JUMP_LABEL */
 
@@ -68,6 +78,10 @@ static __always_inline void jump_label_init(void)
 {
 }
 
+struct jump_label_key_deferred {
+	struct jump_label_key  key;
+};
+
 static __always_inline bool static_branch(struct jump_label_key *key)
 {
 	if (unlikely(atomic_read(&key->enabled)))
@@ -85,6 +99,11 @@ static inline void jump_label_dec(struct jump_label_key *key)
 	atomic_dec(&key->enabled);
 }
 
+static inline void jump_label_dec_deferred(struct jump_label_key_deferred *key)
+{
+	jump_label_dec(&key->key);
+}
+
 static inline int jump_label_text_reserved(void *start, void *end)
 {
 	return 0;
@@ -102,6 +121,14 @@ static inline int jump_label_apply_nops(struct module *mod)
 {
 	return 0;
 }
+
+static inline void jump_label_rate_limit(struct jump_label_key_deferred *key,
+		unsigned long rl)
+{
+}
 #endif	/* HAVE_JUMP_LABEL */
 
+#define jump_label_key_enabled	((struct jump_label_key){ .enabled = ATOMIC_INIT(1), })
+#define jump_label_key_disabled	((struct jump_label_key){ .enabled = ATOMIC_INIT(0), })
+
 #endif	/* _LINUX_JUMP_LABEL_H */

include/linux/perf_event.h

@@ -54,6 +54,7 @@ enum perf_hw_id {
 	PERF_COUNT_HW_BUS_CYCLES		= 6,
 	PERF_COUNT_HW_STALLED_CYCLES_FRONTEND	= 7,
 	PERF_COUNT_HW_STALLED_CYCLES_BACKEND	= 8,
+	PERF_COUNT_HW_REF_CPU_CYCLES		= 9,
 
 	PERF_COUNT_HW_MAX,			/* non-ABI */
 };
@@ -890,6 +891,7 @@ struct perf_event_context {
 	int				nr_active;
 	int				is_active;
 	int				nr_stat;
+	int				nr_freq;
 	int				rotate_disable;
 	atomic_t			refcount;
 	struct task_struct		*task;
@@ -1063,12 +1065,12 @@ perf_sw_event(u32 event_id, u64 nr, struct pt_regs *regs, u64 addr)
 	}
 }
 
-extern struct jump_label_key perf_sched_events;
+extern struct jump_label_key_deferred perf_sched_events;
 
 static inline void perf_event_task_sched_in(struct task_struct *prev,
 					    struct task_struct *task)
 {
-	if (static_branch(&perf_sched_events))
+	if (static_branch(&perf_sched_events.key))
 		__perf_event_task_sched_in(prev, task);
 }
 
@@ -1077,7 +1079,7 @@ static inline void perf_event_task_sched_out(struct task_struct *prev,
 {
 	perf_sw_event(PERF_COUNT_SW_CONTEXT_SWITCHES, 1, NULL, 0);
 
-	if (static_branch(&perf_sched_events))
+	if (static_branch(&perf_sched_events.key))
 		__perf_event_task_sched_out(prev, next);
 }
 

kernel/events/Makefile

@@ -2,5 +2,5 @@ ifdef CONFIG_FUNCTION_TRACER
 CFLAGS_REMOVE_core.o = -pg
 endif
 
-obj-y := core.o ring_buffer.o
+obj-y := core.o ring_buffer.o callchain.o
 obj-$(CONFIG_HAVE_HW_BREAKPOINT) += hw_breakpoint.o

kernel/events/callchain.c

@@ -0,0 +1,191 @@
+/*
+ * Performance events callchain code, extracted from core.c:
+ *
+ *  Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de>
+ *  Copyright (C) 2008-2011 Red Hat, Inc., Ingo Molnar
+ *  Copyright (C) 2008-2011 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com>
+ *  Copyright  ©  2009 Paul Mackerras, IBM Corp. <paulus@au1.ibm.com>
+ *
+ * For licensing details see kernel-base/COPYING
+ */
+
+#include <linux/perf_event.h>
+#include <linux/slab.h>
+#include "internal.h"
+
+struct callchain_cpus_entries {
+	struct rcu_head			rcu_head;
+	struct perf_callchain_entry	*cpu_entries[0];
+};
+
+static DEFINE_PER_CPU(int, callchain_recursion[PERF_NR_CONTEXTS]);
+static atomic_t nr_callchain_events;
+static DEFINE_MUTEX(callchain_mutex);
+static struct callchain_cpus_entries *callchain_cpus_entries;
+
+
+__weak void perf_callchain_kernel(struct perf_callchain_entry *entry,
+				  struct pt_regs *regs)
+{
+}
+
+__weak void perf_callchain_user(struct perf_callchain_entry *entry,
+				struct pt_regs *regs)
+{
+}
+
+static void release_callchain_buffers_rcu(struct rcu_head *head)
+{
+	struct callchain_cpus_entries *entries;
+	int cpu;
+
+	entries = container_of(head, struct callchain_cpus_entries, rcu_head);
+
+	for_each_possible_cpu(cpu)
+		kfree(entries->cpu_entries[cpu]);
+
+	kfree(entries);
+}
+
+static void release_callchain_buffers(void)
+{
+	struct callchain_cpus_entries *entries;
+
+	entries = callchain_cpus_entries;
+	rcu_assign_pointer(callchain_cpus_entries, NULL);
+	call_rcu(&entries->rcu_head, release_callchain_buffers_rcu);
+}
+
+static int alloc_callchain_buffers(void)
+{
+	int cpu;
+	int size;
+	struct callchain_cpus_entries *entries;
+
+	/*
+	 * We can't use the percpu allocation API for data that can be
+	 * accessed from NMI. Use a temporary manual per cpu allocation
+	 * until that gets sorted out.
+	 */
+	size = offsetof(struct callchain_cpus_entries, cpu_entries[nr_cpu_ids]);
+
+	entries = kzalloc(size, GFP_KERNEL);
+	if (!entries)
+		return -ENOMEM;
+
+	size = sizeof(struct perf_callchain_entry) * PERF_NR_CONTEXTS;
+
+	for_each_possible_cpu(cpu) {
+		entries->cpu_entries[cpu] = kmalloc_node(size, GFP_KERNEL,
+							 cpu_to_node(cpu));
+		if (!entries->cpu_entries[cpu])
+			goto fail;
+	}
+
+	rcu_assign_pointer(callchain_cpus_entries, entries);
+
+	return 0;
+
+fail:
+	for_each_possible_cpu(cpu)
+		kfree(entries->cpu_entries[cpu]);
+	kfree(entries);
+
+	return -ENOMEM;
+}
+
+int get_callchain_buffers(void)
+{
+	int err = 0;
+	int count;
+
+	mutex_lock(&callchain_mutex);
+
+	count = atomic_inc_return(&nr_callchain_events);
+	if (WARN_ON_ONCE(count < 1)) {
+		err = -EINVAL;
+		goto exit;
+	}
+
+	if (count > 1) {
+		/* If the allocation failed, give up */
+		if (!callchain_cpus_entries)
+			err = -ENOMEM;
+		goto exit;
+	}
+
+	err = alloc_callchain_buffers();
+	if (err)
+		release_callchain_buffers();
+exit:
+	mutex_unlock(&callchain_mutex);
+
+	return err;
+}
+
+void put_callchain_buffers(void)
+{
+	if (atomic_dec_and_mutex_lock(&nr_callchain_events, &callchain_mutex)) {
+		release_callchain_buffers();
+		mutex_unlock(&callchain_mutex);
+	}
+}
+
+static struct perf_callchain_entry *get_callchain_entry(int *rctx)
+{
+	int cpu;
+	struct callchain_cpus_entries *entries;
+
+	*rctx = get_recursion_context(__get_cpu_var(callchain_recursion));
+	if (*rctx == -1)
+		return NULL;
+
+	entries = rcu_dereference(callchain_cpus_entries);
+	if (!entries)
+		return NULL;
+
+	cpu = smp_processor_id();
+
+	return &entries->cpu_entries[cpu][*rctx];
+}
+
+static void
+put_callchain_entry(int rctx)
+{
+	put_recursion_context(__get_cpu_var(callchain_recursion), rctx);
+}
+
+struct perf_callchain_entry *perf_callchain(struct pt_regs *regs)
+{
+	int rctx;
+	struct perf_callchain_entry *entry;
+
+
+	entry = get_callchain_entry(&rctx);
+	if (rctx == -1)
+		return NULL;
+
+	if (!entry)
+		goto exit_put;
+
+	entry->nr = 0;
+
+	if (!user_mode(regs)) {
+		perf_callchain_store(entry, PERF_CONTEXT_KERNEL);
+		perf_callchain_kernel(entry, regs);
+		if (current->mm)
+			regs = task_pt_regs(current);
+		else
+			regs = NULL;
+	}
+
+	if (regs) {
+		perf_callchain_store(entry, PERF_CONTEXT_USER);
+		perf_callchain_user(entry, regs);
+	}
+
+exit_put:
+	put_callchain_entry(rctx);
+
+	return entry;
+}

kernel/events/core.c

@@ -128,7 +128,7 @@ enum event_type_t {
  * perf_sched_events : >0 events exist
  * perf_cgroup_events: >0 per-cpu cgroup events exist on this cpu
  */
-struct jump_label_key perf_sched_events __read_mostly;
+struct jump_label_key_deferred perf_sched_events __read_mostly;
 static DEFINE_PER_CPU(atomic_t, perf_cgroup_events);
 
 static atomic_t nr_mmap_events __read_mostly;
@@ -1130,6 +1130,8 @@ event_sched_out(struct perf_event *event,
 	if (!is_software_event(event))
 		cpuctx->active_oncpu--;
 	ctx->nr_active--;
+	if (event->attr.freq && event->attr.sample_freq)
+		ctx->nr_freq--;
 	if (event->attr.exclusive || !cpuctx->active_oncpu)
 		cpuctx->exclusive = 0;
 }
@@ -1325,6 +1327,7 @@ retry:
 	}
 	raw_spin_unlock_irq(&ctx->lock);
 }
+EXPORT_SYMBOL_GPL(perf_event_disable);
 
 static void perf_set_shadow_time(struct perf_event *event,
 				 struct perf_event_context *ctx,
@@ -1406,6 +1409,8 @@ event_sched_in(struct perf_event *event,
 	if (!is_software_event(event))
 		cpuctx->active_oncpu++;
 	ctx->nr_active++;
+	if (event->attr.freq && event->attr.sample_freq)
+		ctx->nr_freq++;
 
 	if (event->attr.exclusive)
 		cpuctx->exclusive = 1;
@@ -1662,8 +1667,7 @@ retry:
  * Note: this works for group members as well as group leaders
  * since the non-leader members' sibling_lists will be empty.
  */
-static void __perf_event_mark_enabled(struct perf_event *event,
-					struct perf_event_context *ctx)
+static void __perf_event_mark_enabled(struct perf_event *event)
 {
 	struct perf_event *sub;
 	u64 tstamp = perf_event_time(event);
@@ -1701,7 +1705,7 @@ static int __perf_event_enable(void *info)
 	 */
 	perf_cgroup_set_timestamp(current, ctx);
 
-	__perf_event_mark_enabled(event, ctx);
+	__perf_event_mark_enabled(event);
 
 	if (!event_filter_match(event)) {
 		if (is_cgroup_event(event))
@@ -1782,7 +1786,7 @@ void perf_event_enable(struct perf_event *event)
 
 retry:
 	if (!ctx->is_active) {
-		__perf_event_mark_enabled(event, ctx);
+		__perf_event_mark_enabled(event);
 		goto out;
 	}
 
@@ -1809,6 +1813,7 @@ retry:
 out:
 	raw_spin_unlock_irq(&ctx->lock);
 }
+EXPORT_SYMBOL_GPL(perf_event_enable);
 
 int perf_event_refresh(struct perf_event *event, int refresh)
 {
@@ -2327,6 +2332,9 @@ static void perf_ctx_adjust_freq(struct perf_event_context *ctx, u64 period)
 	u64 interrupts, now;
 	s64 delta;
 
+	if (!ctx->nr_freq)
+		return;
+
 	list_for_each_entry_rcu(event, &ctx->event_list, event_entry) {
 		if (event->state != PERF_EVENT_STATE_ACTIVE)
 			continue;
@@ -2382,12 +2390,14 @@ static void perf_rotate_context(struct perf_cpu_context *cpuctx)
 {
 	u64 interval = (u64)cpuctx->jiffies_interval * TICK_NSEC;
 	struct perf_event_context *ctx = NULL;
-	int rotate = 0, remove = 1;
+	int rotate = 0, remove = 1, freq = 0;
 
 	if (cpuctx->ctx.nr_events) {
 		remove = 0;
 		if (cpuctx->ctx.nr_events != cpuctx->ctx.nr_active)
 			rotate = 1;
+		if (cpuctx->ctx.nr_freq)
+			freq = 1;
 	}
 
 	ctx = cpuctx->task_ctx;
@@ -2395,33 +2405,40 @@ static void perf_rotate_context(struct perf_cpu_context *cpuctx)
 		remove = 0;
 		if (ctx->nr_events != ctx->nr_active)
 			rotate = 1;
+		if (ctx->nr_freq)
+			freq = 1;
 	}
 
+	if (!rotate && !freq)
+		goto done;
+
 	perf_ctx_lock(cpuctx, cpuctx->task_ctx);
 	perf_pmu_disable(cpuctx->ctx.pmu);
-	perf_ctx_adjust_freq(&cpuctx->ctx, interval);
-	if (ctx)
-		perf_ctx_adjust_freq(ctx, interval);
 
-	if (!rotate)
-		goto done;
+	if (freq) {
+		perf_ctx_adjust_freq(&cpuctx->ctx, interval);
+		if (ctx)
+			perf_ctx_adjust_freq(ctx, interval);
+	}
 
-	cpu_ctx_sched_out(cpuctx, EVENT_FLEXIBLE);
-	if (ctx)
-		ctx_sched_out(ctx, cpuctx, EVENT_FLEXIBLE);
+	if (rotate) {
+		cpu_ctx_sched_out(cpuctx, EVENT_FLEXIBLE);
+		if (ctx)
+			ctx_sched_out(ctx, cpuctx, EVENT_FLEXIBLE);
 
-	rotate_ctx(&cpuctx->ctx);
-	if (ctx)
-		rotate_ctx(ctx);
+		rotate_ctx(&cpuctx->ctx);
+		if (ctx)
+			rotate_ctx(ctx);
 
-	perf_event_sched_in(cpuctx, ctx, current);
+		perf_event_sched_in(cpuctx, ctx, current);
+	}
+
+	perf_pmu_enable(cpuctx->ctx.pmu);
+	perf_ctx_unlock(cpuctx, cpuctx->task_ctx);
 
 done:
 	if (remove)
 		list_del_init(&cpuctx->rotation_list);
-
-	perf_pmu_enable(cpuctx->ctx.pmu);
-	perf_ctx_unlock(cpuctx, cpuctx->task_ctx);
 }
 
 void perf_event_task_tick(void)
@@ -2448,7 +2465,7 @@ static int event_enable_on_exec(struct perf_event *event,
 	if (event->state >= PERF_EVENT_STATE_INACTIVE)
 		return 0;
 
-	__perf_event_mark_enabled(event, ctx);
+	__perf_event_mark_enabled(event);
 
 	return 1;
 }
@@ -2480,13 +2497,7 @@ static void perf_event_enable_on_exec(struct perf_event_context *ctx)
 	raw_spin_lock(&ctx->lock);
 	task_ctx_sched_out(ctx);
 
-	list_for_each_entry(event, &ctx->pinned_groups, group_entry) {
-		ret = event_enable_on_exec(event, ctx);
-		if (ret)
-			enabled = 1;
-	}
-
-	list_for_each_entry(event, &ctx->flexible_groups, group_entry) {
+	list_for_each_entry(event, &ctx->event_list, event_entry) {
 		ret = event_enable_on_exec(event, ctx);
 		if (ret)
 			enabled = 1;
@@ -2573,215 +2584,6 @@ static u64 perf_event_read(struct perf_event *event)
 	return perf_event_count(event);
 }
 
-/*
- * Callchain support
- */
-
-struct callchain_cpus_entries {
-	struct rcu_head			rcu_head;
-	struct perf_callchain_entry	*cpu_entries[0];
-};
-
-static DEFINE_PER_CPU(int, callchain_recursion[PERF_NR_CONTEXTS]);
-static atomic_t nr_callchain_events;
-static DEFINE_MUTEX(callchain_mutex);
-struct callchain_cpus_entries *callchain_cpus_entries;
-
-
-__weak void perf_callchain_kernel(struct perf_callchain_entry *entry,
-				  struct pt_regs *regs)
-{
-}
-
-__weak void perf_callchain_user(struct perf_callchain_entry *entry,
-				struct pt_regs *regs)
-{
-}
-
-static void release_callchain_buffers_rcu(struct rcu_head *head)
-{
-	struct callchain_cpus_entries *entries;
-	int cpu;
-
-	entries = container_of(head, struct callchain_cpus_entries, rcu_head);
-
-	for_each_possible_cpu(cpu)
-		kfree(entries->cpu_entries[cpu]);
-
-	kfree(entries);
-}
-
-static void release_callchain_buffers(void)
-{
-	struct callchain_cpus_entries *entries;
-
-	entries = callchain_cpus_entries;
-	rcu_assign_pointer(callchain_cpus_entries, NULL);
-	call_rcu(&entries->rcu_head, release_callchain_buffers_rcu);
-}
-
-static int alloc_callchain_buffers(void)
-{
-	int cpu;
-	int size;
-	struct callchain_cpus_entries *entries;
-
-	/*
-	 * We can't use the percpu allocation API for data that can be
-	 * accessed from NMI. Use a temporary manual per cpu allocation
-	 * until that gets sorted out.
-	 */
-	size = offsetof(struct callchain_cpus_entries, cpu_entries[nr_cpu_ids]);
-
-	entries = kzalloc(size, GFP_KERNEL);
-	if (!entries)
-		return -ENOMEM;
-
-	size = sizeof(struct perf_callchain_entry) * PERF_NR_CONTEXTS;
-
-	for_each_possible_cpu(cpu) {
-		entries->cpu_entries[cpu] = kmalloc_node(size, GFP_KERNEL,
-							 cpu_to_node(cpu));
-		if (!entries->cpu_entries[cpu])
-			goto fail;
-	}
-
-	rcu_assign_pointer(callchain_cpus_entries, entries);
-
-	return 0;
-
-fail:
-	for_each_possible_cpu(cpu)
-		kfree(entries->cpu_entries[cpu]);
-	kfree(entries);
-
-	return -ENOMEM;
-}
-
-static int get_callchain_buffers(void)
-{
-	int err = 0;
-	int count;
-
-	mutex_lock(&callchain_mutex);
-
-	count = atomic_inc_return(&nr_callchain_events);
-	if (WARN_ON_ONCE(count < 1)) {
-		err = -EINVAL;
-		goto exit;
-	}
-
-	if (count > 1) {
-		/* If the allocation failed, give up */
-		if (!callchain_cpus_entries)
-			err = -ENOMEM;
-		goto exit;
-	}
-
-	err = alloc_callchain_buffers();
-	if (err)
-		release_callchain_buffers();
-exit:
-	mutex_unlock(&callchain_mutex);
-
-	return err;
-}
-
-static void put_callchain_buffers(void)
-{
-	if (atomic_dec_and_mutex_lock(&nr_callchain_events, &callchain_mutex)) {
-		release_callchain_buffers();
-		mutex_unlock(&callchain_mutex);
-	}
-}
-
-static int get_recursion_context(int *recursion)
-{
-	int rctx;
-
-	if (in_nmi())
-		rctx = 3;
-	else if (in_irq())
-		rctx = 2;
-	else if (in_softirq())
-		rctx = 1;
-	else
-		rctx = 0;
-
-	if (recursion[rctx])
-		return -1;
-
-	recursion[rctx]++;
-	barrier();
-
-	return rctx;
-}
-
-static inline void put_recursion_context(int *recursion, int rctx)
-{
-	barrier();
-	recursion[rctx]--;
-}
-
-static struct perf_callchain_entry *get_callchain_entry(int *rctx)
-{
-	int cpu;
-	struct callchain_cpus_entries *entries;
-
-	*rctx = get_recursion_context(__get_cpu_var(callchain_recursion));
-	if (*rctx == -1)
-		return NULL;
-
-	entries = rcu_dereference(callchain_cpus_entries);
-	if (!entries)
-		return NULL;
-
-	cpu = smp_processor_id();
-
-	return &entries->cpu_entries[cpu][*rctx];
-}
-
-static void
-put_callchain_entry(int rctx)
-{
-	put_recursion_context(__get_cpu_var(callchain_recursion), rctx);
-}
-
-static struct perf_callchain_entry *perf_callchain(struct pt_regs *regs)
-{
-	int rctx;
-	struct perf_callchain_entry *entry;
-
-
-	entry = get_callchain_entry(&rctx);
-	if (rctx == -1)
-		return NULL;
-
-	if (!entry)
-		goto exit_put;
-
-	entry->nr = 0;
-
-	if (!user_mode(regs)) {
-		perf_callchain_store(entry, PERF_CONTEXT_KERNEL);
-		perf_callchain_kernel(entry, regs);
-		if (current->mm)
-			regs = task_pt_regs(current);
-		else
-			regs = NULL;
-	}
-
-	if (regs) {
-		perf_callchain_store(entry, PERF_CONTEXT_USER);
-		perf_callchain_user(entry, regs);
-	}
-
-exit_put:
-	put_callchain_entry(rctx);
-
-	return entry;
-}
-
 /*
  * Initialize the perf_event context in a task_struct:
  */
@@ -2946,7 +2748,7 @@ static void free_event(struct perf_event *event)
 
 	if (!event->parent) {
 		if (event->attach_state & PERF_ATTACH_TASK)
-			jump_label_dec(&perf_sched_events);
+			jump_label_dec_deferred(&perf_sched_events);
 		if (event->attr.mmap || event->attr.mmap_data)
 			atomic_dec(&nr_mmap_events);
 		if (event->attr.comm)
@@ -2957,7 +2759,7 @@ static void free_event(struct perf_event *event)
 			put_callchain_buffers();
 		if (is_cgroup_event(event)) {
 			atomic_dec(&per_cpu(perf_cgroup_events, event->cpu));
-			jump_label_dec(&perf_sched_events);
+			jump_label_dec_deferred(&perf_sched_events);
 		}
 	}
 
@@ -4820,7 +4622,6 @@ static void perf_swevent_overflow(struct perf_event *event, u64 overflow,
 	struct hw_perf_event *hwc = &event->hw;
 	int throttle = 0;
 
-	data->period = event->hw.last_period;
 	if (!overflow)
 		overflow = perf_swevent_set_period(event);
 
@@ -4854,6 +4655,12 @@ static void perf_swevent_event(struct perf_event *event, u64 nr,
 	if (!is_sampling_event(event))
 		return;
 
+	if ((event->attr.sample_type & PERF_SAMPLE_PERIOD) && !event->attr.freq) {
+		data->period = nr;
+		return perf_swevent_overflow(event, 1, data, regs);
+	} else
+		data->period = event->hw.last_period;
+
 	if (nr == 1 && hwc->sample_period == 1 && !event->attr.freq)
 		return perf_swevent_overflow(event, 1, data, regs);
 
@@ -5981,7 +5788,7 @@ done:
 
 	if (!event->parent) {
 		if (event->attach_state & PERF_ATTACH_TASK)
-			jump_label_inc(&perf_sched_events);
+			jump_label_inc(&perf_sched_events.key);
 		if (event->attr.mmap || event->attr.mmap_data)
 			atomic_inc(&nr_mmap_events);
 		if (event->attr.comm)
@@ -6219,7 +6026,7 @@ SYSCALL_DEFINE5(perf_event_open,
 		 * - that may need work on context switch
 		 */
 		atomic_inc(&per_cpu(perf_cgroup_events, event->cpu));
-		jump_label_inc(&perf_sched_events);
+		jump_label_inc(&perf_sched_events.key);
 	}
 
 	/*
@@ -7065,6 +6872,9 @@ void __init perf_event_init(void)
 
 	ret = init_hw_breakpoint();
 	WARN(ret, "hw_breakpoint initialization failed with: %d", ret);
+
+	/* do not patch jump label more than once per second */
+	jump_label_rate_limit(&perf_sched_events, HZ);
 }
 
 static int __init perf_event_sysfs_init(void)

kernel/events/internal.h

@@ -1,6 +1,10 @@
 #ifndef _KERNEL_EVENTS_INTERNAL_H
 #define _KERNEL_EVENTS_INTERNAL_H
 
+#include <linux/hardirq.h>
+
+/* Buffer handling */
+
 #define RING_BUFFER_WRITABLE		0x01
 
 struct ring_buffer {
@@ -67,7 +71,7 @@ static inline int page_order(struct ring_buffer *rb)
 }
 #endif
 
-static unsigned long perf_data_size(struct ring_buffer *rb)
+static inline unsigned long perf_data_size(struct ring_buffer *rb)
 {
 	return rb->nr_pages << (PAGE_SHIFT + page_order(rb));
 }
@@ -96,4 +100,37 @@ __output_copy(struct perf_output_handle *handle,
 	} while (len);
 }
 
+/* Callchain handling */
+extern struct perf_callchain_entry *perf_callchain(struct pt_regs *regs);
+extern int get_callchain_buffers(void);
+extern void put_callchain_buffers(void);
+
+static inline int get_recursion_context(int *recursion)
+{
+	int rctx;
+
+	if (in_nmi())
+		rctx = 3;
+	else if (in_irq())
+		rctx = 2;
+	else if (in_softirq())
+		rctx = 1;
+	else
+		rctx = 0;
+
+	if (recursion[rctx])
+		return -1;
+
+	recursion[rctx]++;
+	barrier();
+
+	return rctx;
+}
+
+static inline void put_recursion_context(int *recursion, int rctx)
+{
+	barrier();
+	recursion[rctx]--;
+}
+
 #endif /* _KERNEL_EVENTS_INTERNAL_H */

kernel/jump_label.c

@@ -72,15 +72,46 @@ void jump_label_inc(struct jump_label_key *key)
 	jump_label_unlock();
 }
 
-void jump_label_dec(struct jump_label_key *key)
+static void __jump_label_dec(struct jump_label_key *key,
+		unsigned long rate_limit, struct delayed_work *work)
 {
 	if (!atomic_dec_and_mutex_lock(&key->enabled, &jump_label_mutex))
 		return;
 
-	jump_label_update(key, JUMP_LABEL_DISABLE);
+	if (rate_limit) {
+		atomic_inc(&key->enabled);
+		schedule_delayed_work(work, rate_limit);
+	} else
+		jump_label_update(key, JUMP_LABEL_DISABLE);
+
 	jump_label_unlock();
 }
 
+static void jump_label_update_timeout(struct work_struct *work)
+{
+	struct jump_label_key_deferred *key =
+		container_of(work, struct jump_label_key_deferred, work.work);
+	__jump_label_dec(&key->key, 0, NULL);
+}
+
+void jump_label_dec(struct jump_label_key *key)
+{
+	__jump_label_dec(key, 0, NULL);
+}
+
+void jump_label_dec_deferred(struct jump_label_key_deferred *key)
+{
+	__jump_label_dec(&key->key, key->timeout, &key->work);
+}
+
+
+void jump_label_rate_limit(struct jump_label_key_deferred *key,
+		unsigned long rl)
+{
+	key->timeout = rl;
+	INIT_DELAYED_WORK(&key->work, jump_label_update_timeout);
+}
+
 static int addr_conflict(struct jump_entry *entry, void *start, void *end)
 {
 	if (entry->code <= (unsigned long)end &&
@@ -111,7 +142,7 @@ static int __jump_label_text_reserved(struct jump_entry *iter_start,
  * running code can override this to make the non-live update case
  * cheaper.
  */
-void __weak arch_jump_label_transform_static(struct jump_entry *entry,
+void __weak __init_or_module arch_jump_label_transform_static(struct jump_entry *entry,
 					    enum jump_label_type type)
 {
 	arch_jump_label_transform(entry, type);	
@@ -217,8 +248,13 @@ void jump_label_apply_nops(struct module *mod)
 	if (iter_start == iter_stop)
 		return;
 
-	for (iter = iter_start; iter < iter_stop; iter++)
-		arch_jump_label_transform_static(iter, JUMP_LABEL_DISABLE);
+	for (iter = iter_start; iter < iter_stop; iter++) {
+		struct jump_label_key *iterk;
+
+		iterk = (struct jump_label_key *)(unsigned long)iter->key;
+		arch_jump_label_transform_static(iter, jump_label_enabled(iterk) ?
+				JUMP_LABEL_ENABLE : JUMP_LABEL_DISABLE);
+	}
 }
 
 static int jump_label_add_module(struct module *mod)
@@ -258,8 +294,7 @@ static int jump_label_add_module(struct module *mod)
 		key->next = jlm;
 
 		if (jump_label_enabled(key))
-			__jump_label_update(key, iter, iter_stop,
-					    JUMP_LABEL_ENABLE);
+			__jump_label_update(key, iter, iter_stop, JUMP_LABEL_ENABLE);
 	}
 
 	return 0;

kernel/lockdep.c

@@ -500,36 +500,32 @@ void get_usage_chars(struct lock_class *class, char usage[LOCK_USAGE_CHARS])
 	usage[i] = '\0';
 }
 
-static int __print_lock_name(struct lock_class *class)
+static void __print_lock_name(struct lock_class *class)
 {
 	char str[KSYM_NAME_LEN];
 	const char *name;
 
-	name = class->name;
-	if (!name)
-		name = __get_key_name(class->key, str);
-
-	return printk("%s", name);
-}
-
-static void print_lock_name(struct lock_class *class)
-{
-	char str[KSYM_NAME_LEN], usage[LOCK_USAGE_CHARS];
-	const char *name;
-
-	get_usage_chars(class, usage);
-
 	name = class->name;
 	if (!name) {
 		name = __get_key_name(class->key, str);
-		printk(" (%s", name);
+		printk("%s", name);
 	} else {
-		printk(" (%s", name);
+		printk("%s", name);
 		if (class->name_version > 1)
 			printk("#%d", class->name_version);
 		if (class->subclass)
 			printk("/%d", class->subclass);
 	}
+}
+
+static void print_lock_name(struct lock_class *class)
+{
+	char usage[LOCK_USAGE_CHARS];
+
+	get_usage_chars(class, usage);
+
+	printk(" (");
+	__print_lock_name(class);
 	printk("){%s}", usage);
 }
 

kernel/trace/trace.c

@@ -338,7 +338,8 @@ static DECLARE_WAIT_QUEUE_HEAD(trace_wait);
 /* trace_flags holds trace_options default values */
 unsigned long trace_flags = TRACE_ITER_PRINT_PARENT | TRACE_ITER_PRINTK |
 	TRACE_ITER_ANNOTATE | TRACE_ITER_CONTEXT_INFO | TRACE_ITER_SLEEP_TIME |
-	TRACE_ITER_GRAPH_TIME | TRACE_ITER_RECORD_CMD | TRACE_ITER_OVERWRITE;
+	TRACE_ITER_GRAPH_TIME | TRACE_ITER_RECORD_CMD | TRACE_ITER_OVERWRITE |
+	TRACE_ITER_IRQ_INFO;
 
 static int trace_stop_count;
 static DEFINE_RAW_SPINLOCK(tracing_start_lock);
@@ -426,6 +427,7 @@ static const char *trace_options[] = {
 	"record-cmd",
 	"overwrite",
 	"disable_on_free",
+	"irq-info",
 	NULL
 };
 
@@ -1843,6 +1845,33 @@ static void s_stop(struct seq_file *m, void *p)
 	trace_event_read_unlock();
 }
 
+static void
+get_total_entries(struct trace_array *tr, unsigned long *total, unsigned long *entries)
+{
+	unsigned long count;
+	int cpu;
+
+	*total = 0;
+	*entries = 0;
+
+	for_each_tracing_cpu(cpu) {
+		count = ring_buffer_entries_cpu(tr->buffer, cpu);
+		/*
+		 * If this buffer has skipped entries, then we hold all
+		 * entries for the trace and we need to ignore the
+		 * ones before the time stamp.
+		 */
+		if (tr->data[cpu]->skipped_entries) {
+			count -= tr->data[cpu]->skipped_entries;
+			/* total is the same as the entries */
+			*total += count;
+		} else
+			*total += count +
+				ring_buffer_overrun_cpu(tr->buffer, cpu);
+		*entries += count;
+	}
+}
+
 static void print_lat_help_header(struct seq_file *m)
 {
 	seq_puts(m, "#                  _------=> CPU#            \n");
@@ -1855,12 +1884,35 @@ static void print_lat_help_header(struct seq_file *m)
 	seq_puts(m, "#     \\   /      |||||  \\    |   /           \n");
 }
 
-static void print_func_help_header(struct seq_file *m)
+static void print_event_info(struct trace_array *tr, struct seq_file *m)
+{
+	unsigned long total;
+	unsigned long entries;
+
+	get_total_entries(tr, &total, &entries);
+	seq_printf(m, "# entries-in-buffer/entries-written: %lu/%lu   #P:%d\n",
+		   entries, total, num_online_cpus());
+	seq_puts(m, "#\n");
+}
+
+static void print_func_help_header(struct trace_array *tr, struct seq_file *m)
 {
-	seq_puts(m, "#           TASK-PID    CPU#    TIMESTAMP  FUNCTION\n");
+	print_event_info(tr, m);
+	seq_puts(m, "#           TASK-PID   CPU#      TIMESTAMP  FUNCTION\n");
 	seq_puts(m, "#              | |       |          |         |\n");
 }
 
+static void print_func_help_header_irq(struct trace_array *tr, struct seq_file *m)
+{
+	print_event_info(tr, m);
+	seq_puts(m, "#                              _-----=> irqs-off\n");
+	seq_puts(m, "#                             / _----=> need-resched\n");
+	seq_puts(m, "#                            | / _---=> hardirq/softirq\n");
+	seq_puts(m, "#                            || / _--=> preempt-depth\n");
+	seq_puts(m, "#                            ||| /     delay\n");
+	seq_puts(m, "#           TASK-PID   CPU#  ||||    TIMESTAMP  FUNCTION\n");
+	seq_puts(m, "#              | |       |   ||||       |         |\n");
+}
 
 void
 print_trace_header(struct seq_file *m, struct trace_iterator *iter)
@@ -1869,32 +1921,14 @@ print_trace_header(struct seq_file *m, struct trace_iterator *iter)
 	struct trace_array *tr = iter->tr;
 	struct trace_array_cpu *data = tr->data[tr->cpu];
 	struct tracer *type = current_trace;
-	unsigned long entries = 0;
-	unsigned long total = 0;
-	unsigned long count;
+	unsigned long entries;
+	unsigned long total;
 	const char *name = "preemption";
-	int cpu;
 
 	if (type)
 		name = type->name;
 
-
-	for_each_tracing_cpu(cpu) {
-		count = ring_buffer_entries_cpu(tr->buffer, cpu);
-		/*
-		 * If this buffer has skipped entries, then we hold all
-		 * entries for the trace and we need to ignore the
-		 * ones before the time stamp.
-		 */
-		if (tr->data[cpu]->skipped_entries) {
-			count -= tr->data[cpu]->skipped_entries;
-			/* total is the same as the entries */
-			total += count;
-		} else
-			total += count +
-				ring_buffer_overrun_cpu(tr->buffer, cpu);
-		entries += count;
-	}
+	get_total_entries(tr, &total, &entries);
 
 	seq_printf(m, "# %s latency trace v1.1.5 on %s\n",
 		   name, UTS_RELEASE);
@@ -2140,6 +2174,21 @@ enum print_line_t print_trace_line(struct trace_iterator *iter)
 	return print_trace_fmt(iter);
 }
 
+void trace_latency_header(struct seq_file *m)
+{
+	struct trace_iterator *iter = m->private;
+
+	/* print nothing if the buffers are empty */
+	if (trace_empty(iter))
+		return;
+
+	if (iter->iter_flags & TRACE_FILE_LAT_FMT)
+		print_trace_header(m, iter);
+
+	if (!(trace_flags & TRACE_ITER_VERBOSE))
+		print_lat_help_header(m);
+}
+
 void trace_default_header(struct seq_file *m)
 {
 	struct trace_iterator *iter = m->private;
@@ -2155,8 +2204,12 @@ void trace_default_header(struct seq_file *m)
 		if (!(trace_flags & TRACE_ITER_VERBOSE))
 			print_lat_help_header(m);
 	} else {
-		if (!(trace_flags & TRACE_ITER_VERBOSE))
-			print_func_help_header(m);
+		if (!(trace_flags & TRACE_ITER_VERBOSE)) {
+			if (trace_flags & TRACE_ITER_IRQ_INFO)
+				print_func_help_header_irq(iter->tr, m);
+			else
+				print_func_help_header(iter->tr, m);
+		}
 	}
 }
 

kernel/trace/trace.h

@@ -370,6 +370,7 @@ void trace_graph_function(struct trace_array *tr,
 		    unsigned long ip,
 		    unsigned long parent_ip,
 		    unsigned long flags, int pc);
+void trace_latency_header(struct seq_file *m);
 void trace_default_header(struct seq_file *m);
 void print_trace_header(struct seq_file *m, struct trace_iterator *iter);
 int trace_empty(struct trace_iterator *iter);
@@ -654,6 +655,7 @@ enum trace_iterator_flags {
 	TRACE_ITER_RECORD_CMD		= 0x100000,
 	TRACE_ITER_OVERWRITE		= 0x200000,
 	TRACE_ITER_STOP_ON_FREE		= 0x400000,
+	TRACE_ITER_IRQ_INFO		= 0x800000,
 };
 
 /*

kernel/trace/trace_events_filter.c

@@ -27,6 +27,12 @@
 #include "trace.h"
 #include "trace_output.h"
 
+#define DEFAULT_SYS_FILTER_MESSAGE					\
+	"### global filter ###\n"					\
+	"# Use this to set filters for multiple events.\n"		\
+	"# Only events with the given fields will be affected.\n"	\
+	"# If no events are modified, an error message will be displayed here"
+
 enum filter_op_ids
 {
 	OP_OR,
@@ -646,7 +652,7 @@ void print_subsystem_event_filter(struct event_subsystem *system,
 	if (filter && filter->filter_string)
 		trace_seq_printf(s, "%s\n", filter->filter_string);
 	else
-		trace_seq_printf(s, "none\n");
+		trace_seq_printf(s, DEFAULT_SYS_FILTER_MESSAGE "\n");
 	mutex_unlock(&event_mutex);
 }
 
@@ -1838,7 +1844,10 @@ int apply_subsystem_event_filter(struct event_subsystem *system,
 	if (!filter)
 		goto out;
 
-	replace_filter_string(filter, filter_string);
+	/* System filters just show a default message */
+	kfree(filter->filter_string);
+	filter->filter_string = NULL;
+
 	/*
 	 * No event actually uses the system filter
 	 * we can free it without synchronize_sched().
@@ -1848,14 +1857,12 @@ int apply_subsystem_event_filter(struct event_subsystem *system,
 
 	parse_init(ps, filter_ops, filter_string);
 	err = filter_parse(ps);
-	if (err) {
-		append_filter_err(ps, system->filter);
-		goto out;
-	}
+	if (err)
+		goto err_filter;
 
 	err = replace_system_preds(system, ps, filter_string);
 	if (err)
-		append_filter_err(ps, system->filter);
+		goto err_filter;
 
 out:
 	filter_opstack_clear(ps);
@@ -1865,6 +1872,11 @@ out_unlock:
 	mutex_unlock(&event_mutex);
 
 	return err;
+
+err_filter:
+	replace_filter_string(filter, filter_string);
+	append_filter_err(ps, system->filter);
+	goto out;
 }
 
 #ifdef CONFIG_PERF_EVENTS

kernel/trace/trace_irqsoff.c

@@ -280,9 +280,20 @@ static enum print_line_t irqsoff_print_line(struct trace_iterator *iter)
 }
 
 static void irqsoff_graph_return(struct ftrace_graph_ret *trace) { }
-static void irqsoff_print_header(struct seq_file *s) { }
 static void irqsoff_trace_open(struct trace_iterator *iter) { }
 static void irqsoff_trace_close(struct trace_iterator *iter) { }
+
+#ifdef CONFIG_FUNCTION_TRACER
+static void irqsoff_print_header(struct seq_file *s)
+{
+	trace_default_header(s);
+}
+#else
+static void irqsoff_print_header(struct seq_file *s)
+{
+	trace_latency_header(s);
+}
+#endif /* CONFIG_FUNCTION_TRACER */
 #endif /* CONFIG_FUNCTION_GRAPH_TRACER */
 
 /*

kernel/trace/trace_output.c

@@ -627,11 +627,23 @@ int trace_print_context(struct trace_iterator *iter)
 	unsigned long usec_rem = do_div(t, USEC_PER_SEC);
 	unsigned long secs = (unsigned long)t;
 	char comm[TASK_COMM_LEN];
+	int ret;
 
 	trace_find_cmdline(entry->pid, comm);
 
-	return trace_seq_printf(s, "%16s-%-5d [%03d] %5lu.%06lu: ",
-				comm, entry->pid, iter->cpu, secs, usec_rem);
+	ret = trace_seq_printf(s, "%16s-%-5d [%03d] ",
+			       comm, entry->pid, iter->cpu);
+	if (!ret)
+		return 0;
+
+	if (trace_flags & TRACE_ITER_IRQ_INFO) {
+		ret = trace_print_lat_fmt(s, entry);
+		if (!ret)
+			return 0;
+	}
+
+	return trace_seq_printf(s, " %5lu.%06lu: ",
+				secs, usec_rem);
 }
 
 int trace_print_lat_context(struct trace_iterator *iter)

kernel/trace/trace_sched_wakeup.c

@@ -280,9 +280,20 @@ static enum print_line_t wakeup_print_line(struct trace_iterator *iter)
 }
 
 static void wakeup_graph_return(struct ftrace_graph_ret *trace) { }
-static void wakeup_print_header(struct seq_file *s) { }
 static void wakeup_trace_open(struct trace_iterator *iter) { }
 static void wakeup_trace_close(struct trace_iterator *iter) { }
+
+#ifdef CONFIG_FUNCTION_TRACER
+static void wakeup_print_header(struct seq_file *s)
+{
+	trace_default_header(s);
+}
+#else
+static void wakeup_print_header(struct seq_file *s)
+{
+	trace_latency_header(s);
+}
+#endif /* CONFIG_FUNCTION_TRACER */
 #endif /* CONFIG_FUNCTION_GRAPH_TRACER */
 
 /*

tools/perf/Documentation/perf-annotate.txt

@@ -22,7 +22,7 @@ OPTIONS
 -------
 -i::
 --input=::
-        Input file name. (default: perf.data)
+        Input file name. (default: perf.data unless stdin is a fifo)
 
 -d::
 --dsos=<dso[,dso...]>::
@@ -66,7 +66,7 @@ OPTIONS
 	used. This interfaces starts by centering on the line with more
 	samples, TAB/UNTAB cycles through the lines with more samples.
 
--c::
+-C::
 --cpu:: Only report samples for the list of CPUs provided. Multiple CPUs can
 	be provided as a comma-separated list with no space: 0,1. Ranges of
 	CPUs are specified with -: 0-2. Default is to report samples on all

tools/perf/Documentation/perf-buildid-list.txt

@@ -26,7 +26,7 @@ OPTIONS
         Show only DSOs with hits.
 -i::
 --input=::
-        Input file name. (default: perf.data)
+        Input file name. (default: perf.data unless stdin is a fifo)
 -f::
 --force::
 	Don't do ownership validation.

tools/perf/Documentation/perf-evlist.txt

@@ -18,7 +18,7 @@ OPTIONS
 -------
 -i::
 --input=::
-        Input file name. (default: perf.data)
+        Input file name. (default: perf.data unless stdin is a fifo)
 
 SEE ALSO
 --------

tools/perf/Documentation/perf-kmem.txt

@@ -23,7 +23,7 @@ OPTIONS
 -------
 -i <file>::
 --input=<file>::
-	Select the input file (default: perf.data)
+	Select the input file (default: perf.data unless stdin is a fifo)
 
 --caller::
 	Show per-callsite statistics

tools/perf/Documentation/perf-lock.txt

@@ -29,7 +29,7 @@ COMMON OPTIONS
 
 -i::
 --input=<file>::
-        Input file name.
+        Input file name. (default: perf.data unless stdin is a fifo)
 
 -v::
 --verbose::

tools/perf/Documentation/perf-record.txt

@@ -89,7 +89,7 @@ OPTIONS
 
 -m::
 --mmap-pages=::
-	Number of mmap data pages.
+	Number of mmap data pages. Must be a power of two.
 
 -g::
 --call-graph::

tools/perf/Documentation/perf-report.txt

@@ -19,7 +19,7 @@ OPTIONS
 -------
 -i::
 --input=::
-        Input file name. (default: perf.data)
+        Input file name. (default: perf.data unless stdin is a fifo)
 
 -v::
 --verbose::
@@ -39,7 +39,7 @@ OPTIONS
 -T::
 --threads::
 	Show per-thread event counters
--C::
+-c::
 --comms=::
 	Only consider symbols in these comms. CSV that understands
 	file://filename entries.
@@ -80,9 +80,10 @@ OPTIONS
 --dump-raw-trace::
         Dump raw trace in ASCII.
 
--g [type,min,order]::
+-g [type,min[,limit],order]::
 --call-graph::
-        Display call chains using type, min percent threshold and order.
+        Display call chains using type, min percent threshold, optional print
+	limit and order.
 	type can be either:
 	- flat: single column, linear exposure of call chains.
 	- graph: use a graph tree, displaying absolute overhead rates.
@@ -128,7 +129,7 @@ OPTIONS
 --symfs=<directory>::
         Look for files with symbols relative to this directory.
 
--c::
+-C::
 --cpu:: Only report samples for the list of CPUs provided. Multiple CPUs can
 	be provided as a comma-separated list with no space: 0,1. Ranges of
 	CPUs are specified with -: 0-2. Default is to report samples on all

tools/perf/Documentation/perf-sched.txt

@@ -40,7 +40,7 @@ OPTIONS
 -------
 -i::
 --input=<file>::
-        Input file name. (default: perf.data)
+        Input file name. (default: perf.data unless stdin is a fifo)
 
 -v::
 --verbose::

tools/perf/Documentation/perf-script.txt

@@ -106,7 +106,7 @@ OPTIONS
 
 -i::
 --input=::
-        Input file name.
+        Input file name. (default: perf.data unless stdin is a fifo)
 
 -d::
 --debug-mode::
@@ -182,12 +182,17 @@ OPTIONS
 --hide-call-graph::
         When printing symbols do not display call chain.
 
--c::
+-C::
 --cpu:: Only report samples for the list of CPUs provided. Multiple CPUs can
 	be provided as a comma-separated list with no space: 0,1. Ranges of
 	CPUs are specified with -: 0-2. Default is to report samples on all
 	CPUs.
 
+-c::
+--comms=::
+	Only display events for these comms. CSV that understands
+	file://filename entries.
+
 -I::
 --show-info::
 	Display extended information about the perf.data file. This adds

tools/perf/Documentation/perf-test.txt

@@ -8,13 +8,19 @@ perf-test - Runs sanity tests.
 SYNOPSIS
 --------
 [verse]
-'perf test <options>'
+'perf test [<options>] [{list <test-name-fragment>|[<test-name-fragments>|<test-numbers>]}]'
 
 DESCRIPTION
 -----------
 This command does assorted sanity tests, initially through linked routines but
 also will look for a directory with more tests in the form of scripts.
 
+To get a list of available tests use 'perf test list', specifying a test name
+fragment will show all tests that have it.
+
+To run just specific tests, inform test name fragments or the numbers obtained
+from 'perf test list'.
+
 OPTIONS
 -------
 -v::

tools/perf/Documentation/perf-timechart.txt

@@ -27,7 +27,7 @@ OPTIONS
         Select the output file (default: output.svg)
 -i::
 --input=::
-        Select the input file (default: perf.data)
+        Select the input file (default: perf.data unless stdin is a fifo)
 -w::
 --width=::
         Select the width of the SVG file (default: 1000)

tools/perf/Makefile

@@ -278,6 +278,7 @@ LIB_H += util/strbuf.h
 LIB_H += util/strlist.h
 LIB_H += util/strfilter.h
 LIB_H += util/svghelper.h
+LIB_H += util/tool.h
 LIB_H += util/run-command.h
 LIB_H += util/sigchain.h
 LIB_H += util/symbol.h

tools/perf/builtin-annotate.c

@@ -27,32 +27,32 @@
 #include "util/sort.h"
 #include "util/hist.h"
 #include "util/session.h"
+#include "util/tool.h"
 
 #include <linux/bitmap.h>
 
-static char		const *input_name = "perf.data";
-
-static bool		force, use_tui, use_stdio;
-
-static bool		full_paths;
-
-static bool		print_line;
-
-static const char *sym_hist_filter;
-
-static const char	*cpu_list;
-static DECLARE_BITMAP(cpu_bitmap, MAX_NR_CPUS);
+struct perf_annotate {
+	struct perf_tool tool;
+	char const *input_name;
+	bool	   force, use_tui, use_stdio;
+	bool	   full_paths;
+	bool	   print_line;
+	const char *sym_hist_filter;
+	const char *cpu_list;
+	DECLARE_BITMAP(cpu_bitmap, MAX_NR_CPUS);
+};
 
-static int perf_evlist__add_sample(struct perf_evlist *evlist,
-				   struct perf_sample *sample,
-				   struct perf_evsel *evsel,
-				   struct addr_location *al)
+static int perf_evsel__add_sample(struct perf_evsel *evsel,
+				  struct perf_sample *sample,
+				  struct addr_location *al,
+				  struct perf_annotate *ann)
 {
 	struct hist_entry *he;
 	int ret;
 
-	if (sym_hist_filter != NULL &&
-	    (al->sym == NULL || strcmp(sym_hist_filter, al->sym->name) != 0)) {
+	if (ann->sym_hist_filter != NULL &&
+	    (al->sym == NULL ||
+	     strcmp(ann->sym_hist_filter, al->sym->name) != 0)) {
 		/* We're only interested in a symbol named sym_hist_filter */
 		if (al->sym != NULL) {
 			rb_erase(&al->sym->rb_node,
@@ -69,8 +69,7 @@ static int perf_evlist__add_sample(struct perf_evlist *evlist,
 	ret = 0;
 	if (he->ms.sym != NULL) {
 		struct annotation *notes = symbol__annotation(he->ms.sym);
-		if (notes->src == NULL &&
-		    symbol__alloc_hist(he->ms.sym, evlist->nr_entries) < 0)
+		if (notes->src == NULL && symbol__alloc_hist(he->ms.sym) < 0)
 			return -ENOMEM;
 
 		ret = hist_entry__inc_addr_samples(he, evsel->idx, al->addr);
@@ -81,25 +80,26 @@ static int perf_evlist__add_sample(struct perf_evlist *evlist,
 	return ret;
 }
 
-static int process_sample_event(union perf_event *event,
+static int process_sample_event(struct perf_tool *tool,
+				union perf_event *event,
 				struct perf_sample *sample,
 				struct perf_evsel *evsel,
-				struct perf_session *session)
+				struct machine *machine)
 {
+	struct perf_annotate *ann = container_of(tool, struct perf_annotate, tool);
 	struct addr_location al;
 
-	if (perf_event__preprocess_sample(event, session, &al, sample,
+	if (perf_event__preprocess_sample(event, machine, &al, sample,
 					  symbol__annotate_init) < 0) {
 		pr_warning("problem processing %d event, skipping it.\n",
 			   event->header.type);
 		return -1;
 	}
 
-	if (cpu_list && !test_bit(sample->cpu, cpu_bitmap))
+	if (ann->cpu_list && !test_bit(sample->cpu, ann->cpu_bitmap))
 		return 0;
 
-	if (!al.filtered &&
-	    perf_evlist__add_sample(session->evlist, sample, evsel, &al)) {
+	if (!al.filtered && perf_evsel__add_sample(evsel, sample, &al, ann)) {
 		pr_warning("problem incrementing symbol count, "
 			   "skipping event\n");
 		return -1;
@@ -108,14 +108,15 @@ static int process_sample_event(union perf_event *event,
 	return 0;
 }
 
-static int hist_entry__tty_annotate(struct hist_entry *he, int evidx)
+static int hist_entry__tty_annotate(struct hist_entry *he, int evidx,
+				    struct perf_annotate *ann)
 {
 	return symbol__tty_annotate(he->ms.sym, he->ms.map, evidx,
-				    print_line, full_paths, 0, 0);
+				    ann->print_line, ann->full_paths, 0, 0);
 }
 
 static void hists__find_annotations(struct hists *self, int evidx,
-				    int nr_events)
+				    struct perf_annotate *ann)
 {
 	struct rb_node *nd = rb_first(&self->entries), *next;
 	int key = K_RIGHT;
@@ -138,8 +139,7 @@ find_next:
 		}
 
 		if (use_browser > 0) {
-			key = hist_entry__tui_annotate(he, evidx, nr_events,
-						       NULL, NULL, 0);
+			key = hist_entry__tui_annotate(he, evidx, NULL, NULL, 0);
 			switch (key) {
 			case K_RIGHT:
 				next = rb_next(nd);
@@ -154,7 +154,7 @@ find_next:
 			if (next != NULL)
 				nd = next;
 		} else {
-			hist_entry__tty_annotate(he, evidx);
+			hist_entry__tty_annotate(he, evidx, ann);
 			nd = rb_next(nd);
 			/*
 			 * Since we have a hist_entry per IP for the same
@@ -167,33 +167,26 @@ find_next:
 	}
 }
 
-static struct perf_event_ops event_ops = {
-	.sample	= process_sample_event,
-	.mmap	= perf_event__process_mmap,
-	.comm	= perf_event__process_comm,
-	.fork	= perf_event__process_task,
-	.ordered_samples = true,
-	.ordering_requires_timestamps = true,
-};
-
-static int __cmd_annotate(void)
+static int __cmd_annotate(struct perf_annotate *ann)
 {
 	int ret;
 	struct perf_session *session;
 	struct perf_evsel *pos;
 	u64 total_nr_samples;
 
-	session = perf_session__new(input_name, O_RDONLY, force, false, &event_ops);
+	session = perf_session__new(ann->input_name, O_RDONLY,
+				    ann->force, false, &ann->tool);
 	if (session == NULL)
 		return -ENOMEM;
 
-	if (cpu_list) {
-		ret = perf_session__cpu_bitmap(session, cpu_list, cpu_bitmap);
+	if (ann->cpu_list) {
+		ret = perf_session__cpu_bitmap(session, ann->cpu_list,
+					       ann->cpu_bitmap);
 		if (ret)
 			goto out_delete;
 	}
 
-	ret = perf_session__process_events(session, &event_ops);
+	ret = perf_session__process_events(session, &ann->tool);
 	if (ret)
 		goto out_delete;
 
@@ -217,13 +210,12 @@ static int __cmd_annotate(void)
 			total_nr_samples += nr_samples;
 			hists__collapse_resort(hists);
 			hists__output_resort(hists);
-			hists__find_annotations(hists, pos->idx,
-						session->evlist->nr_entries);
+			hists__find_annotations(hists, pos->idx, ann);
 		}
 	}
 
 	if (total_nr_samples == 0) {
-		ui__warning("The %s file has no samples!\n", input_name);
+		ui__warning("The %s file has no samples!\n", session->filename);
 		goto out_delete;
 	}
 out_delete:
@@ -247,29 +239,41 @@ static const char * const annotate_usage[] = {
 	NULL
 };
 
-static const struct option options[] = {
-	OPT_STRING('i', "input", &input_name, "file",
+int cmd_annotate(int argc, const char **argv, const char *prefix __used)
+{
+	struct perf_annotate annotate = {
+		.tool = {
+			.sample	= process_sample_event,
+			.mmap	= perf_event__process_mmap,
+			.comm	= perf_event__process_comm,
+			.fork	= perf_event__process_task,
+			.ordered_samples = true,
+			.ordering_requires_timestamps = true,
+		},
+	};
+	const struct option options[] = {
+	OPT_STRING('i', "input", &annotate.input_name, "file",
 		    "input file name"),
 	OPT_STRING('d', "dsos", &symbol_conf.dso_list_str, "dso[,dso...]",
 		   "only consider symbols in these dsos"),
-	OPT_STRING('s', "symbol", &sym_hist_filter, "symbol",
+	OPT_STRING('s', "symbol", &annotate.sym_hist_filter, "symbol",
 		    "symbol to annotate"),
-	OPT_BOOLEAN('f', "force", &force, "don't complain, do it"),
+	OPT_BOOLEAN('f', "force", &annotate.force, "don't complain, do it"),
 	OPT_INCR('v', "verbose", &verbose,
 		    "be more verbose (show symbol address, etc)"),
 	OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
 		    "dump raw trace in ASCII"),
-	OPT_BOOLEAN(0, "tui", &use_tui, "Use the TUI interface"),
-	OPT_BOOLEAN(0, "stdio", &use_stdio, "Use the stdio interface"),
+	OPT_BOOLEAN(0, "tui", &annotate.use_tui, "Use the TUI interface"),
+	OPT_BOOLEAN(0, "stdio", &annotate.use_stdio, "Use the stdio interface"),
 	OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name,
 		   "file", "vmlinux pathname"),
 	OPT_BOOLEAN('m', "modules", &symbol_conf.use_modules,
 		    "load module symbols - WARNING: use only with -k and LIVE kernel"),
-	OPT_BOOLEAN('l', "print-line", &print_line,
+	OPT_BOOLEAN('l', "print-line", &annotate.print_line,
 		    "print matching source lines (may be slow)"),
-	OPT_BOOLEAN('P', "full-paths", &full_paths,
+	OPT_BOOLEAN('P', "full-paths", &annotate.full_paths,
 		    "Don't shorten the displayed pathnames"),
-	OPT_STRING('c', "cpu", &cpu_list, "cpu", "list of cpus to profile"),
+	OPT_STRING('C', "cpu", &annotate.cpu_list, "cpu", "list of cpus to profile"),
 	OPT_STRING(0, "symfs", &symbol_conf.symfs, "directory",
 		   "Look for files with symbols relative to this directory"),
 	OPT_BOOLEAN(0, "source", &symbol_conf.annotate_src,
@@ -279,15 +283,13 @@ static const struct option options[] = {
 	OPT_STRING('M', "disassembler-style", &disassembler_style, "disassembler style",
 		   "Specify disassembler style (e.g. -M intel for intel syntax)"),
 	OPT_END()
-};
+	};
 
-int cmd_annotate(int argc, const char **argv, const char *prefix __used)
-{
 	argc = parse_options(argc, argv, options, annotate_usage, 0);
 
-	if (use_stdio)
+	if (annotate.use_stdio)
 		use_browser = 0;
-	else if (use_tui)
+	else if (annotate.use_tui)
 		use_browser = 1;
 
 	setup_browser(true);
@@ -308,7 +310,7 @@ int cmd_annotate(int argc, const char **argv, const char *prefix __used)
 		if (argc > 1)
 			usage_with_options(annotate_usage, options);
 
-		sym_hist_filter = argv[0];
+		annotate.sym_hist_filter = argv[0];
 	}
 
 	if (field_sep && *field_sep == '.') {
@@ -316,5 +318,5 @@ int cmd_annotate(int argc, const char **argv, const char *prefix __used)
 		return -1;
 	}
 
-	return __cmd_annotate();
+	return __cmd_annotate(&annotate);
 }

tools/perf/builtin-buildid-list.c

@@ -18,7 +18,7 @@
 
 #include <libelf.h>
 
-static char const *input_name = "perf.data";
+static const char *input_name;
 static bool force;
 static bool show_kernel;
 static bool with_hits;
@@ -39,24 +39,6 @@ static const struct option options[] = {
 	OPT_END()
 };
 
-static int perf_session__list_build_ids(void)
-{
-	struct perf_session *session;
-
-	session = perf_session__new(input_name, O_RDONLY, force, false,
-				    &build_id__mark_dso_hit_ops);
-	if (session == NULL)
-		return -1;
-
-	if (with_hits)
-		perf_session__process_events(session, &build_id__mark_dso_hit_ops);
-
-	perf_session__fprintf_dsos_buildid(session, stdout, with_hits);
-
-	perf_session__delete(session);
-	return 0;
-}
-
 static int sysfs__fprintf_build_id(FILE *fp)
 {
 	u8 kallsyms_build_id[BUILD_ID_SIZE];
@@ -85,17 +67,36 @@ static int filename__fprintf_build_id(const char *name, FILE *fp)
 	return fprintf(fp, "%s\n", sbuild_id);
 }
 
-static int __cmd_buildid_list(void)
+static int perf_session__list_build_ids(void)
 {
-	if (show_kernel)
-		return sysfs__fprintf_build_id(stdout);
+	struct perf_session *session;
 
 	elf_version(EV_CURRENT);
+
+	session = perf_session__new(input_name, O_RDONLY, force, false,
+				    &build_id__mark_dso_hit_ops);
+	if (session == NULL)
+		return -1;
+
 	/*
- 	 * See if this is an ELF file first:
- 	 */
-	if (filename__fprintf_build_id(input_name, stdout))
-		return 0;
+	 * See if this is an ELF file first:
+	 */
+	if (filename__fprintf_build_id(session->filename, stdout))
+		goto out;
+
+	if (with_hits)
+		perf_session__process_events(session, &build_id__mark_dso_hit_ops);
+
+	perf_session__fprintf_dsos_buildid(session, stdout, with_hits);
+out:
+	perf_session__delete(session);
+	return 0;
+}
+
+static int __cmd_buildid_list(void)
+{
+	if (show_kernel)
+		return sysfs__fprintf_build_id(stdout);
 
 	return perf_session__list_build_ids();
 }

tools/perf/builtin-diff.c

@@ -9,7 +9,9 @@
 #include "util/debug.h"
 #include "util/event.h"
 #include "util/hist.h"
+#include "util/evsel.h"
 #include "util/session.h"
+#include "util/tool.h"
 #include "util/sort.h"
 #include "util/symbol.h"
 #include "util/util.h"
@@ -30,14 +32,15 @@ static int hists__add_entry(struct hists *self,
 	return -ENOMEM;
 }
 
-static int diff__process_sample_event(union perf_event *event,
+static int diff__process_sample_event(struct perf_tool *tool __used,
+				      union perf_event *event,
 				      struct perf_sample *sample,
 				      struct perf_evsel *evsel __used,
-				      struct perf_session *session)
+				      struct machine *machine)
 {
 	struct addr_location al;
 
-	if (perf_event__preprocess_sample(event, session, &al, sample, NULL) < 0) {
+	if (perf_event__preprocess_sample(event, machine, &al, sample, NULL) < 0) {
 		pr_warning("problem processing %d event, skipping it.\n",
 			   event->header.type);
 		return -1;
@@ -46,16 +49,16 @@ static int diff__process_sample_event(union perf_event *event,
 	if (al.filtered || al.sym == NULL)
 		return 0;
 
-	if (hists__add_entry(&session->hists, &al, sample->period)) {
+	if (hists__add_entry(&evsel->hists, &al, sample->period)) {
 		pr_warning("problem incrementing symbol period, skipping event\n");
 		return -1;
 	}
 
-	session->hists.stats.total_period += sample->period;
+	evsel->hists.stats.total_period += sample->period;
 	return 0;
 }
 
-static struct perf_event_ops event_ops = {
+static struct perf_tool perf_diff = {
 	.sample	= diff__process_sample_event,
 	.mmap	= perf_event__process_mmap,
 	.comm	= perf_event__process_comm,
@@ -145,13 +148,13 @@ static int __cmd_diff(void)
 	int ret, i;
 	struct perf_session *session[2];
 
-	session[0] = perf_session__new(input_old, O_RDONLY, force, false, &event_ops);
-	session[1] = perf_session__new(input_new, O_RDONLY, force, false, &event_ops);
+	session[0] = perf_session__new(input_old, O_RDONLY, force, false, &perf_diff);
+	session[1] = perf_session__new(input_new, O_RDONLY, force, false, &perf_diff);
 	if (session[0] == NULL || session[1] == NULL)
 		return -ENOMEM;
 
 	for (i = 0; i < 2; ++i) {
-		ret = perf_session__process_events(session[i], &event_ops);
+		ret = perf_session__process_events(session[i], &perf_diff);
 		if (ret)
 			goto out_delete;
 	}

tools/perf/builtin-evlist.c

@@ -15,7 +15,7 @@
 #include "util/parse-options.h"
 #include "util/session.h"
 
-static char const *input_name = "perf.data";
+static const char *input_name;
 
 static int __cmd_evlist(void)
 {

tools/perf/builtin-inject.c

@@ -9,6 +9,7 @@
 
 #include "perf.h"
 #include "util/session.h"
+#include "util/tool.h"
 #include "util/debug.h"
 
 #include "util/parse-options.h"
@@ -16,8 +17,9 @@
 static char		const *input_name = "-";
 static bool		inject_build_ids;
 
-static int perf_event__repipe_synth(union perf_event *event,
-				    struct perf_session *session __used)
+static int perf_event__repipe_synth(struct perf_tool *tool __used,
+				    union perf_event *event,
+				    struct machine *machine __used)
 {
 	uint32_t size;
 	void *buf = event;
@@ -36,41 +38,70 @@ static int perf_event__repipe_synth(union perf_event *event,
 	return 0;
 }
 
-static int perf_event__repipe(union perf_event *event,
+static int perf_event__repipe_op2_synth(struct perf_tool *tool,
+					union perf_event *event,
+					struct perf_session *session __used)
+{
+	return perf_event__repipe_synth(tool, event, NULL);
+}
+
+static int perf_event__repipe_event_type_synth(struct perf_tool *tool,
+					       union perf_event *event)
+{
+	return perf_event__repipe_synth(tool, event, NULL);
+}
+
+static int perf_event__repipe_tracing_data_synth(union perf_event *event,
+						 struct perf_session *session __used)
+{
+	return perf_event__repipe_synth(NULL, event, NULL);
+}
+
+static int perf_event__repipe_attr(union perf_event *event,
+				   struct perf_evlist **pevlist __used)
+{
+	return perf_event__repipe_synth(NULL, event, NULL);
+}
+
+static int perf_event__repipe(struct perf_tool *tool,
+			      union perf_event *event,
 			      struct perf_sample *sample __used,
-			      struct perf_session *session)
+			      struct machine *machine)
 {
-	return perf_event__repipe_synth(event, session);
+	return perf_event__repipe_synth(tool, event, machine);
 }
 
-static int perf_event__repipe_sample(union perf_event *event,
+static int perf_event__repipe_sample(struct perf_tool *tool,
+				     union perf_event *event,
 			      struct perf_sample *sample __used,
 			      struct perf_evsel *evsel __used,
-			      struct perf_session *session)
+			      struct machine *machine)
 {
-	return perf_event__repipe_synth(event, session);
+	return perf_event__repipe_synth(tool, event, machine);
 }
 
-static int perf_event__repipe_mmap(union perf_event *event,
+static int perf_event__repipe_mmap(struct perf_tool *tool,
+				   union perf_event *event,
 				   struct perf_sample *sample,
-				   struct perf_session *session)
+				   struct machine *machine)
 {
 	int err;
 
-	err = perf_event__process_mmap(event, sample, session);
-	perf_event__repipe(event, sample, session);
+	err = perf_event__process_mmap(tool, event, sample, machine);
+	perf_event__repipe(tool, event, sample, machine);
 
 	return err;
 }
 
-static int perf_event__repipe_task(union perf_event *event,
+static int perf_event__repipe_task(struct perf_tool *tool,
+				   union perf_event *event,
 				   struct perf_sample *sample,
-				   struct perf_session *session)
+				   struct machine *machine)
 {
 	int err;
 
-	err = perf_event__process_task(event, sample, session);
-	perf_event__repipe(event, sample, session);
+	err = perf_event__process_task(tool, event, sample, machine);
+	perf_event__repipe(tool, event, sample, machine);
 
 	return err;
 }
@@ -80,7 +111,7 @@ static int perf_event__repipe_tracing_data(union perf_event *event,
 {
 	int err;
 
-	perf_event__repipe_synth(event, session);
+	perf_event__repipe_synth(NULL, event, NULL);
 	err = perf_event__process_tracing_data(event, session);
 
 	return err;
@@ -100,10 +131,10 @@ static int dso__read_build_id(struct dso *self)
 	return -1;
 }
 
-static int dso__inject_build_id(struct dso *self, struct perf_session *session)
+static int dso__inject_build_id(struct dso *self, struct perf_tool *tool,
+				struct machine *machine)
 {
 	u16 misc = PERF_RECORD_MISC_USER;
-	struct machine *machine;
 	int err;
 
 	if (dso__read_build_id(self) < 0) {
@@ -111,17 +142,11 @@ static int dso__inject_build_id(struct dso *self, struct perf_session *session)
 		return -1;
 	}
 
-	machine = perf_session__find_host_machine(session);
-	if (machine == NULL) {
-		pr_err("Can't find machine for session\n");
-		return -1;
-	}
-
 	if (self->kernel)
 		misc = PERF_RECORD_MISC_KERNEL;
 
-	err = perf_event__synthesize_build_id(self, misc, perf_event__repipe,
-					      machine, session);
+	err = perf_event__synthesize_build_id(tool, self, misc, perf_event__repipe,
+					      machine);
 	if (err) {
 		pr_err("Can't synthesize build_id event for %s\n", self->long_name);
 		return -1;
@@ -130,10 +155,11 @@ static int dso__inject_build_id(struct dso *self, struct perf_session *session)
 	return 0;
 }
 
-static int perf_event__inject_buildid(union perf_event *event,
+static int perf_event__inject_buildid(struct perf_tool *tool,
+				      union perf_event *event,
 				      struct perf_sample *sample,
 				      struct perf_evsel *evsel __used,
-				      struct perf_session *session)
+				      struct machine *machine)
 {
 	struct addr_location al;
 	struct thread *thread;
@@ -141,21 +167,21 @@ static int perf_event__inject_buildid(union perf_event *event,
 
 	cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
 
-	thread = perf_session__findnew(session, event->ip.pid);
+	thread = machine__findnew_thread(machine, event->ip.pid);
 	if (thread == NULL) {
 		pr_err("problem processing %d event, skipping it.\n",
 		       event->header.type);
 		goto repipe;
 	}
 
-	thread__find_addr_map(thread, session, cpumode, MAP__FUNCTION,
-			      event->ip.pid, event->ip.ip, &al);
+	thread__find_addr_map(thread, machine, cpumode, MAP__FUNCTION,
+			      event->ip.ip, &al);
 
 	if (al.map != NULL) {
 		if (!al.map->dso->hit) {
 			al.map->dso->hit = 1;
 			if (map__load(al.map, NULL) >= 0) {
-				dso__inject_build_id(al.map->dso, session);
+				dso__inject_build_id(al.map->dso, tool, machine);
 				/*
 				 * If this fails, too bad, let the other side
 				 * account this as unresolved.
@@ -168,24 +194,24 @@ static int perf_event__inject_buildid(union perf_event *event,
 	}
 
 repipe:
-	perf_event__repipe(event, sample, session);
+	perf_event__repipe(tool, event, sample, machine);
 	return 0;
 }
 
-struct perf_event_ops inject_ops = {
+struct perf_tool perf_inject = {
 	.sample		= perf_event__repipe_sample,
 	.mmap		= perf_event__repipe,
 	.comm		= perf_event__repipe,
 	.fork		= perf_event__repipe,
 	.exit		= perf_event__repipe,
 	.lost		= perf_event__repipe,
-	.read		= perf_event__repipe,
+	.read		= perf_event__repipe_sample,
 	.throttle	= perf_event__repipe,
 	.unthrottle	= perf_event__repipe,
-	.attr		= perf_event__repipe_synth,
-	.event_type 	= perf_event__repipe_synth,
-	.tracing_data 	= perf_event__repipe_synth,
-	.build_id 	= perf_event__repipe_synth,
+	.attr		= perf_event__repipe_attr,
+	.event_type	= perf_event__repipe_event_type_synth,
+	.tracing_data	= perf_event__repipe_tracing_data_synth,
+	.build_id	= perf_event__repipe_op2_synth,
 };
 
 extern volatile int session_done;
@@ -203,17 +229,17 @@ static int __cmd_inject(void)
 	signal(SIGINT, sig_handler);
 
 	if (inject_build_ids) {
-		inject_ops.sample	= perf_event__inject_buildid;
-		inject_ops.mmap		= perf_event__repipe_mmap;
-		inject_ops.fork		= perf_event__repipe_task;
-		inject_ops.tracing_data	= perf_event__repipe_tracing_data;
+		perf_inject.sample	 = perf_event__inject_buildid;
+		perf_inject.mmap	 = perf_event__repipe_mmap;
+		perf_inject.fork	 = perf_event__repipe_task;
+		perf_inject.tracing_data = perf_event__repipe_tracing_data;
 	}
 
-	session = perf_session__new(input_name, O_RDONLY, false, true, &inject_ops);
+	session = perf_session__new(input_name, O_RDONLY, false, true, &perf_inject);
 	if (session == NULL)
 		return -ENOMEM;
 
-	ret = perf_session__process_events(session, &inject_ops);
+	ret = perf_session__process_events(session, &perf_inject);
 
 	perf_session__delete(session);
 

tools/perf/builtin-kmem.c

@@ -7,6 +7,7 @@
 #include "util/thread.h"
 #include "util/header.h"
 #include "util/session.h"
+#include "util/tool.h"
 
 #include "util/parse-options.h"
 #include "util/trace-event.h"
@@ -18,7 +19,7 @@
 struct alloc_stat;
 typedef int (*sort_fn_t)(struct alloc_stat *, struct alloc_stat *);
 
-static char const		*input_name = "perf.data";
+static const char		*input_name;
 
 static int			alloc_flag;
 static int			caller_flag;
@@ -303,12 +304,13 @@ static void process_raw_event(union perf_event *raw_event __used, void *data,
 	}
 }
 
-static int process_sample_event(union perf_event *event,
+static int process_sample_event(struct perf_tool *tool __used,
+				union perf_event *event,
 				struct perf_sample *sample,
 				struct perf_evsel *evsel __used,
-				struct perf_session *session)
+				struct machine *machine)
 {
-	struct thread *thread = perf_session__findnew(session, event->ip.pid);
+	struct thread *thread = machine__findnew_thread(machine, event->ip.pid);
 
 	if (thread == NULL) {
 		pr_debug("problem processing %d event, skipping it.\n",
@@ -324,7 +326,7 @@ static int process_sample_event(union perf_event *event,
 	return 0;
 }
 
-static struct perf_event_ops event_ops = {
+static struct perf_tool perf_kmem = {
 	.sample			= process_sample_event,
 	.comm			= perf_event__process_comm,
 	.ordered_samples	= true,
@@ -483,7 +485,7 @@ static int __cmd_kmem(void)
 {
 	int err = -EINVAL;
 	struct perf_session *session = perf_session__new(input_name, O_RDONLY,
-							 0, false, &event_ops);
+							 0, false, &perf_kmem);
 	if (session == NULL)
 		return -ENOMEM;
 
@@ -494,7 +496,7 @@ static int __cmd_kmem(void)
 		goto out_delete;
 
 	setup_pager();
-	err = perf_session__process_events(session, &event_ops);
+	err = perf_session__process_events(session, &perf_kmem);
 	if (err != 0)
 		goto out_delete;
 	sort_result();

tools/perf/builtin-kvm.c

@@ -38,7 +38,7 @@ static const struct option kvm_options[] = {
 	OPT_BOOLEAN(0, "guest", &perf_guest,
 		    "Collect guest os data"),
 	OPT_BOOLEAN(0, "host", &perf_host,
-		    "Collect guest os data"),
+		    "Collect host os data"),
 	OPT_STRING(0, "guestmount", &symbol_conf.guestmount, "directory",
 		   "guest mount directory under which every guest os"
 		   " instance has a subdir"),

tools/perf/builtin-lock.c

@@ -12,6 +12,7 @@
 
 #include "util/debug.h"
 #include "util/session.h"
+#include "util/tool.h"
 
 #include <sys/types.h>
 #include <sys/prctl.h>
@@ -325,7 +326,7 @@ alloc_failed:
 	die("memory allocation failed\n");
 }
 
-static char			const *input_name = "perf.data";
+static const char *input_name;
 
 struct raw_event_sample {
 	u32			size;
@@ -845,12 +846,13 @@ static void dump_info(void)
 		die("Unknown type of information\n");
 }
 
-static int process_sample_event(union perf_event *event,
+static int process_sample_event(struct perf_tool *tool __used,
+				union perf_event *event,
 				struct perf_sample *sample,
 				struct perf_evsel *evsel __used,
-				struct perf_session *s)
+				struct machine *machine)
 {
-	struct thread *thread = perf_session__findnew(s, sample->tid);
+	struct thread *thread = machine__findnew_thread(machine, sample->tid);
 
 	if (thread == NULL) {
 		pr_debug("problem processing %d event, skipping it.\n",
@@ -863,7 +865,7 @@ static int process_sample_event(union perf_event *event,
 	return 0;
 }
 
-static struct perf_event_ops eops = {
+static struct perf_tool eops = {
 	.sample			= process_sample_event,
 	.comm			= perf_event__process_comm,
 	.ordered_samples	= true,

tools/perf/builtin-probe.c

@@ -46,7 +46,6 @@
 
 #define DEFAULT_VAR_FILTER "!__k???tab_* & !__crc_*"
 #define DEFAULT_FUNC_FILTER "!_*"
-#define MAX_PATH_LEN 256
 
 /* Session management structure */
 static struct {

tools/perf/builtin-record.c

@@ -22,6 +22,7 @@
 #include "util/evsel.h"
 #include "util/debug.h"
 #include "util/session.h"
+#include "util/tool.h"
 #include "util/symbol.h"
 #include "util/cpumap.h"
 #include "util/thread_map.h"
@@ -35,55 +36,36 @@ enum write_mode_t {
 	WRITE_APPEND
 };
 
-static u64			user_interval			= ULLONG_MAX;
-static u64			default_interval		=      0;
-
-static unsigned int		page_size;
-static unsigned int		mmap_pages			= UINT_MAX;
-static unsigned int		user_freq 			= UINT_MAX;
-static int			freq				=   1000;
-static int			output;
-static int			pipe_output			=      0;
-static const char		*output_name			= NULL;
-static bool			group				=  false;
-static int			realtime_prio			=      0;
-static bool			nodelay				=  false;
-static bool			raw_samples			=  false;
-static bool			sample_id_all_avail		=   true;
-static bool			system_wide			=  false;
-static pid_t			target_pid			=     -1;
-static pid_t			target_tid			=     -1;
-static pid_t			child_pid			=     -1;
-static bool			no_inherit			=  false;
-static enum write_mode_t	write_mode			= WRITE_FORCE;
-static bool			call_graph			=  false;
-static bool			inherit_stat			=  false;
-static bool			no_samples			=  false;
-static bool			sample_address			=  false;
-static bool			sample_time			=  false;
-static bool			no_buildid			=  false;
-static bool			no_buildid_cache		=  false;
-static struct perf_evlist	*evsel_list;
-
-static long			samples				=      0;
-static u64			bytes_written			=      0;
-
-static int			file_new			=      1;
-static off_t			post_processing_offset;
-
-static struct perf_session	*session;
-static const char		*cpu_list;
-static const char               *progname;
-
-static void advance_output(size_t size)
+struct perf_record {
+	struct perf_tool	tool;
+	struct perf_record_opts	opts;
+	u64			bytes_written;
+	const char		*output_name;
+	struct perf_evlist	*evlist;
+	struct perf_session	*session;
+	const char		*progname;
+	int			output;
+	unsigned int		page_size;
+	int			realtime_prio;
+	enum write_mode_t	write_mode;
+	bool			no_buildid;
+	bool			no_buildid_cache;
+	bool			force;
+	bool			file_new;
+	bool			append_file;
+	long			samples;
+	off_t			post_processing_offset;
+};
+
+static void advance_output(struct perf_record *rec, size_t size)
 {
-	bytes_written += size;
+	rec->bytes_written += size;
 }
 
-static void write_output(void *buf, size_t size)
+static void write_output(struct perf_record *rec, void *buf, size_t size)
 {
 	while (size) {
-		int ret = write(output, buf, size);
+		int ret = write(rec->output, buf, size);
 
 		if (ret < 0)
 			die("failed to write");
@@ -91,30 +73,33 @@ static void write_output(void *buf, size_t size)
 		size -= ret;
 		buf += ret;
 
-		bytes_written += ret;
+		rec->bytes_written += ret;
 	}
 }
 
-static int process_synthesized_event(union perf_event *event,
+static int process_synthesized_event(struct perf_tool *tool,
+				     union perf_event *event,
 				     struct perf_sample *sample __used,
-				     struct perf_session *self __used)
+				     struct machine *machine __used)
 {
-	write_output(event, event->header.size);
+	struct perf_record *rec = container_of(tool, struct perf_record, tool);
+	write_output(rec, event, event->header.size);
 	return 0;
 }
 
-static void mmap_read(struct perf_mmap *md)
+static void perf_record__mmap_read(struct perf_record *rec,
+				   struct perf_mmap *md)
 {
 	unsigned int head = perf_mmap__read_head(md);
 	unsigned int old = md->prev;
-	unsigned char *data = md->base + page_size;
+	unsigned char *data = md->base + rec->page_size;
 	unsigned long size;
 	void *buf;
 
 	if (old == head)
 		return;
 
-	samples++;
+	rec->samples++;
 
 	size = head - old;
 
@@ -123,14 +108,14 @@ static void mmap_read(struct perf_mmap *md)
 		size = md->mask + 1 - (old & md->mask);
 		old += size;
 
-		write_output(buf, size);
+		write_output(rec, buf, size);
 	}
 
 	buf = &data[old & md->mask];
 	size = head - old;
 	old += size;
 
-	write_output(buf, size);
+	write_output(rec, buf, size);
 
 	md->prev = old;
 	perf_mmap__write_tail(md, old);
@@ -149,17 +134,18 @@ static void sig_handler(int sig)
 	signr = sig;
 }
 
-static void sig_atexit(void)
+static void perf_record__sig_exit(int exit_status __used, void *arg)
 {
+	struct perf_record *rec = arg;
 	int status;
 
-	if (child_pid > 0) {
+	if (rec->evlist->workload.pid > 0) {
 		if (!child_finished)
-			kill(child_pid, SIGTERM);
+			kill(rec->evlist->workload.pid, SIGTERM);
 
 		wait(&status);
 		if (WIFSIGNALED(status))
-			psignal(WTERMSIG(status), progname);
+			psignal(WTERMSIG(status), rec->progname);
 	}
 
 	if (signr == -1 || signr == SIGUSR1)
@@ -169,78 +155,6 @@ static void sig_atexit(void)
 	kill(getpid(), signr);
 }
 
-static void config_attr(struct perf_evsel *evsel, struct perf_evlist *evlist)
-{
-	struct perf_event_attr *attr = &evsel->attr;
-	int track = !evsel->idx; /* only the first counter needs these */
-
-	attr->disabled		= 1;
-	attr->inherit		= !no_inherit;
-	attr->read_format	= PERF_FORMAT_TOTAL_TIME_ENABLED |
-				  PERF_FORMAT_TOTAL_TIME_RUNNING |
-				  PERF_FORMAT_ID;
-
-	attr->sample_type	|= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
-
-	if (evlist->nr_entries > 1)
-		attr->sample_type |= PERF_SAMPLE_ID;
-
-	/*
-	 * We default some events to a 1 default interval. But keep
-	 * it a weak assumption overridable by the user.
-	 */
-	if (!attr->sample_period || (user_freq != UINT_MAX &&
-				     user_interval != ULLONG_MAX)) {
-		if (freq) {
-			attr->sample_type	|= PERF_SAMPLE_PERIOD;
-			attr->freq		= 1;
-			attr->sample_freq	= freq;
-		} else {
-			attr->sample_period = default_interval;
-		}
-	}
-
-	if (no_samples)
-		attr->sample_freq = 0;
-
-	if (inherit_stat)
-		attr->inherit_stat = 1;
-
-	if (sample_address) {
-		attr->sample_type	|= PERF_SAMPLE_ADDR;
-		attr->mmap_data = track;
-	}
-
-	if (call_graph)
-		attr->sample_type	|= PERF_SAMPLE_CALLCHAIN;
-
-	if (system_wide)
-		attr->sample_type	|= PERF_SAMPLE_CPU;
-
-	if (sample_id_all_avail &&
-	    (sample_time || system_wide || !no_inherit || cpu_list))
-		attr->sample_type	|= PERF_SAMPLE_TIME;
-
-	if (raw_samples) {
-		attr->sample_type	|= PERF_SAMPLE_TIME;
-		attr->sample_type	|= PERF_SAMPLE_RAW;
-		attr->sample_type	|= PERF_SAMPLE_CPU;
-	}
-
-	if (nodelay) {
-		attr->watermark = 0;
-		attr->wakeup_events = 1;
-	}
-
-	attr->mmap		= track;
-	attr->comm		= track;
-
-	if (target_pid == -1 && target_tid == -1 && !system_wide) {
-		attr->disabled = 1;
-		attr->enable_on_exec = 1;
-	}
-}
-
 static bool perf_evlist__equal(struct perf_evlist *evlist,
 			       struct perf_evlist *other)
 {
@@ -260,15 +174,17 @@ static bool perf_evlist__equal(struct perf_evlist *evlist,
 	return true;
 }
 
-static void open_counters(struct perf_evlist *evlist)
+static void perf_record__open(struct perf_record *rec)
 {
 	struct perf_evsel *pos, *first;
-
-	if (evlist->cpus->map[0] < 0)
-		no_inherit = true;
+	struct perf_evlist *evlist = rec->evlist;
+	struct perf_session *session = rec->session;
+	struct perf_record_opts *opts = &rec->opts;
 
 	first = list_entry(evlist->entries.next, struct perf_evsel, node);
 
+	perf_evlist__config_attrs(evlist, opts);
+
 	list_for_each_entry(pos, &evlist->entries, node) {
 		struct perf_event_attr *attr = &pos->attr;
 		struct xyarray *group_fd = NULL;
@@ -286,29 +202,27 @@ static void open_counters(struct perf_evlist *evlist)
 		 */
 		bool time_needed = attr->sample_type & PERF_SAMPLE_TIME;
 
-		if (group && pos != first)
+		if (opts->group && pos != first)
 			group_fd = first->fd;
-
-		config_attr(pos, evlist);
 retry_sample_id:
-		attr->sample_id_all = sample_id_all_avail ? 1 : 0;
+		attr->sample_id_all = opts->sample_id_all_avail ? 1 : 0;
 try_again:
-		if (perf_evsel__open(pos, evlist->cpus, evlist->threads, group,
-				     group_fd) < 0) {
+		if (perf_evsel__open(pos, evlist->cpus, evlist->threads,
+				     opts->group, group_fd) < 0) {
 			int err = errno;
 
 			if (err == EPERM || err == EACCES) {
 				ui__error_paranoid();
 				exit(EXIT_FAILURE);
-			} else if (err ==  ENODEV && cpu_list) {
+			} else if (err ==  ENODEV && opts->cpu_list) {
 				die("No such device - did you specify"
 					" an out-of-range profile CPU?\n");
-			} else if (err == EINVAL && sample_id_all_avail) {
+			} else if (err == EINVAL && opts->sample_id_all_avail) {
 				/*
 				 * Old kernel, no attr->sample_id_type_all field
 				 */
-				sample_id_all_avail = false;
-				if (!sample_time && !raw_samples && !time_needed)
+				opts->sample_id_all_avail = false;
+				if (!opts->sample_time && !opts->raw_samples && !time_needed)
 					attr->sample_type &= ~PERF_SAMPLE_TIME;
 
 				goto retry_sample_id;
@@ -358,10 +272,20 @@ try_again:
 		exit(-1);
 	}
 
-	if (perf_evlist__mmap(evlist, mmap_pages, false) < 0)
+	if (perf_evlist__mmap(evlist, opts->mmap_pages, false) < 0) {
+		if (errno == EPERM)
+			die("Permission error mapping pages.\n"
+			    "Consider increasing "
+			    "/proc/sys/kernel/perf_event_mlock_kb,\n"
+			    "or try again with a smaller value of -m/--mmap_pages.\n"
+			    "(current value: %d)\n", opts->mmap_pages);
+		else if (!is_power_of_2(opts->mmap_pages))
+			die("--mmap_pages/-m value must be a power of two.");
+
 		die("failed to mmap with %d (%s)\n", errno, strerror(errno));
+	}
 
-	if (file_new)
+	if (rec->file_new)
 		session->evlist = evlist;
 	else {
 		if (!perf_evlist__equal(session->evlist, evlist)) {
@@ -373,29 +297,32 @@ try_again:
 	perf_session__update_sample_type(session);
 }
 
-static int process_buildids(void)
+static int process_buildids(struct perf_record *rec)
 {
-	u64 size = lseek(output, 0, SEEK_CUR);
+	u64 size = lseek(rec->output, 0, SEEK_CUR);
 
 	if (size == 0)
 		return 0;
 
-	session->fd = output;
-	return __perf_session__process_events(session, post_processing_offset,
-					      size - post_processing_offset,
+	rec->session->fd = rec->output;
+	return __perf_session__process_events(rec->session, rec->post_processing_offset,
+					      size - rec->post_processing_offset,
 					      size, &build_id__mark_dso_hit_ops);
 }
 
-static void atexit_header(void)
+static void perf_record__exit(int status __used, void *arg)
 {
-	if (!pipe_output) {
-		session->header.data_size += bytes_written;
-
-		if (!no_buildid)
-			process_buildids();
-		perf_session__write_header(session, evsel_list, output, true);
-		perf_session__delete(session);
-		perf_evlist__delete(evsel_list);
+	struct perf_record *rec = arg;
+
+	if (!rec->opts.pipe_output) {
+		rec->session->header.data_size += rec->bytes_written;
+
+		if (!rec->no_buildid)
+			process_buildids(rec);
+		perf_session__write_header(rec->session, rec->evlist,
+					   rec->output, true);
+		perf_session__delete(rec->session);
+		perf_evlist__delete(rec->evlist);
 		symbol__exit();
 	}
 }
@@ -403,7 +330,7 @@ static void atexit_header(void)
 static void perf_event__synthesize_guest_os(struct machine *machine, void *data)
 {
 	int err;
-	struct perf_session *psession = data;
+	struct perf_tool *tool = data;
 
 	if (machine__is_host(machine))
 		return;
@@ -416,8 +343,8 @@ static void perf_event__synthesize_guest_os(struct machine *machine, void *data)
 	 *method is used to avoid symbol missing when the first addr is
 	 *in module instead of in guest kernel.
 	 */
-	err = perf_event__synthesize_modules(process_synthesized_event,
-					     psession, machine);
+	err = perf_event__synthesize_modules(tool, process_synthesized_event,
+					     machine);
 	if (err < 0)
 		pr_err("Couldn't record guest kernel [%d]'s reference"
 		       " relocation symbol.\n", machine->pid);
@@ -426,12 +353,11 @@ static void perf_event__synthesize_guest_os(struct machine *machine, void *data)
 	 * We use _stext for guest kernel because guest kernel's /proc/kallsyms
 	 * have no _text sometimes.
 	 */
-	err = perf_event__synthesize_kernel_mmap(process_synthesized_event,
-						 psession, machine, "_text");
+	err = perf_event__synthesize_kernel_mmap(tool, process_synthesized_event,
+						 machine, "_text");
 	if (err < 0)
-		err = perf_event__synthesize_kernel_mmap(process_synthesized_event,
-							 psession, machine,
-							 "_stext");
+		err = perf_event__synthesize_kernel_mmap(tool, process_synthesized_event,
+							 machine, "_stext");
 	if (err < 0)
 		pr_err("Couldn't record guest kernel [%d]'s reference"
 		       " relocation symbol.\n", machine->pid);
@@ -442,73 +368,71 @@ static struct perf_event_header finished_round_event = {
 	.type = PERF_RECORD_FINISHED_ROUND,
 };
 
-static void mmap_read_all(void)
+static void perf_record__mmap_read_all(struct perf_record *rec)
 {
 	int i;
 
-	for (i = 0; i < evsel_list->nr_mmaps; i++) {
-		if (evsel_list->mmap[i].base)
-			mmap_read(&evsel_list->mmap[i]);
+	for (i = 0; i < rec->evlist->nr_mmaps; i++) {
+		if (rec->evlist->mmap[i].base)
+			perf_record__mmap_read(rec, &rec->evlist->mmap[i]);
 	}
 
-	if (perf_header__has_feat(&session->header, HEADER_TRACE_INFO))
-		write_output(&finished_round_event, sizeof(finished_round_event));
+	if (perf_header__has_feat(&rec->session->header, HEADER_TRACE_INFO))
+		write_output(rec, &finished_round_event, sizeof(finished_round_event));
 }
 
-static int __cmd_record(int argc, const char **argv)
+static int __cmd_record(struct perf_record *rec, int argc, const char **argv)
 {
 	struct stat st;
 	int flags;
-	int err;
+	int err, output;
 	unsigned long waking = 0;
-	int child_ready_pipe[2], go_pipe[2];
 	const bool forks = argc > 0;
-	char buf;
 	struct machine *machine;
+	struct perf_tool *tool = &rec->tool;
+	struct perf_record_opts *opts = &rec->opts;
+	struct perf_evlist *evsel_list = rec->evlist;
+	const char *output_name = rec->output_name;
+	struct perf_session *session;
 
-	progname = argv[0];
+	rec->progname = argv[0];
 
-	page_size = sysconf(_SC_PAGE_SIZE);
+	rec->page_size = sysconf(_SC_PAGE_SIZE);
 
-	atexit(sig_atexit);
+	on_exit(perf_record__sig_exit, rec);
 	signal(SIGCHLD, sig_handler);
 	signal(SIGINT, sig_handler);
 	signal(SIGUSR1, sig_handler);
 
-	if (forks && (pipe(child_ready_pipe) < 0 || pipe(go_pipe) < 0)) {
-		perror("failed to create pipes");
-		exit(-1);
-	}
-
 	if (!output_name) {
 		if (!fstat(STDOUT_FILENO, &st) && S_ISFIFO(st.st_mode))
-			pipe_output = 1;
+			opts->pipe_output = true;
 		else
-			output_name = "perf.data";
+			rec->output_name = output_name = "perf.data";
 	}
 	if (output_name) {
 		if (!strcmp(output_name, "-"))
-			pipe_output = 1;
+			opts->pipe_output = true;
 		else if (!stat(output_name, &st) && st.st_size) {
-			if (write_mode == WRITE_FORCE) {
+			if (rec->write_mode == WRITE_FORCE) {
 				char oldname[PATH_MAX];
 				snprintf(oldname, sizeof(oldname), "%s.old",
 					 output_name);
 				unlink(oldname);
 				rename(output_name, oldname);
 			}
-		} else if (write_mode == WRITE_APPEND) {
-			write_mode = WRITE_FORCE;
+		} else if (rec->write_mode == WRITE_APPEND) {
+			rec->write_mode = WRITE_FORCE;
 		}
 	}
 
 	flags = O_CREAT|O_RDWR;
-	if (write_mode == WRITE_APPEND)
-		file_new = 0;
+	if (rec->write_mode == WRITE_APPEND)
+		rec->file_new = 0;
 	else
 		flags |= O_TRUNC;
 
-	if (pipe_output)
+	if (opts->pipe_output)
 		output = STDOUT_FILENO;
 	else
 		output = open(output_name, flags, S_IRUSR | S_IWUSR);
@@ -517,17 +441,21 @@ static int __cmd_record(int argc, const char **argv)
 		exit(-1);
 	}
 
+	rec->output = output;
+
 	session = perf_session__new(output_name, O_WRONLY,
-				    write_mode == WRITE_FORCE, false, NULL);
+				    rec->write_mode == WRITE_FORCE, false, NULL);
 	if (session == NULL) {
 		pr_err("Not enough memory for reading perf file header\n");
 		return -1;
 	}
 
-	if (!no_buildid)
+	rec->session = session;
+
+	if (!rec->no_buildid)
 		perf_header__set_feat(&session->header, HEADER_BUILD_ID);
 
-	if (!file_new) {
+	if (!rec->file_new) {
 		err = perf_session__read_header(session, output);
 		if (err < 0)
 			goto out_delete_session;
@@ -549,94 +477,57 @@ static int __cmd_record(int argc, const char **argv)
 	perf_header__set_feat(&session->header, HEADER_NUMA_TOPOLOGY);
 	perf_header__set_feat(&session->header, HEADER_CPUID);
 
-	/* 512 kiB: default amount of unprivileged mlocked memory */
-	if (mmap_pages == UINT_MAX)
-		mmap_pages = (512 * 1024) / page_size;
-
 	if (forks) {
-		child_pid = fork();
-		if (child_pid < 0) {
-			perror("failed to fork");
-			exit(-1);
-		}
-
-		if (!child_pid) {
-			if (pipe_output)
-				dup2(2, 1);
-			close(child_ready_pipe[0]);
-			close(go_pipe[1]);
-			fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC);
-
-			/*
-			 * Do a dummy execvp to get the PLT entry resolved,
-			 * so we avoid the resolver overhead on the real
-			 * execvp call.
-			 */
-			execvp("", (char **)argv);
-
-			/*
-			 * Tell the parent we're ready to go
-			 */
-			close(child_ready_pipe[1]);
-
-			/*
-			 * Wait until the parent tells us to go.
-			 */
-			if (read(go_pipe[0], &buf, 1) == -1)
-				perror("unable to read pipe");
-
-			execvp(argv[0], (char **)argv);
-
-			perror(argv[0]);
-			kill(getppid(), SIGUSR1);
-			exit(-1);
-		}
-
-		if (!system_wide && target_tid == -1 && target_pid == -1)
-			evsel_list->threads->map[0] = child_pid;
-
-		close(child_ready_pipe[1]);
-		close(go_pipe[0]);
-		/*
-		 * wait for child to settle
-		 */
-		if (read(child_ready_pipe[0], &buf, 1) == -1) {
-			perror("unable to read pipe");
-			exit(-1);
+		err = perf_evlist__prepare_workload(evsel_list, opts, argv);
+		if (err < 0) {
+			pr_err("Couldn't run the workload!\n");
+			goto out_delete_session;
 		}
-		close(child_ready_pipe[0]);
 	}
 
-	open_counters(evsel_list);
+	perf_record__open(rec);
 
 	/*
-	 * perf_session__delete(session) will be called at atexit_header()
+	 * perf_session__delete(session) will be called at perf_record__exit()
 	 */
-	atexit(atexit_header);
+	on_exit(perf_record__exit, rec);
 
-	if (pipe_output) {
+	if (opts->pipe_output) {
 		err = perf_header__write_pipe(output);
 		if (err < 0)
 			return err;
-	} else if (file_new) {
+	} else if (rec->file_new) {
 		err = perf_session__write_header(session, evsel_list,
 						 output, false);
 		if (err < 0)
 			return err;
 	}
 
-	post_processing_offset = lseek(output, 0, SEEK_CUR);
+	if (!!rec->no_buildid
+	    && !perf_header__has_feat(&session->header, HEADER_BUILD_ID)) {
+		pr_err("Couldn't generating buildids. "
+		       "Use --no-buildid to profile anyway.\n");
+		return -1;
+	}
 
-	if (pipe_output) {
-		err = perf_session__synthesize_attrs(session,
-						     process_synthesized_event);
+	rec->post_processing_offset = lseek(output, 0, SEEK_CUR);
+
+	machine = perf_session__find_host_machine(session);
+	if (!machine) {
+		pr_err("Couldn't find native kernel information.\n");
+		return -1;
+	}
+
+	if (opts->pipe_output) {
+		err = perf_event__synthesize_attrs(tool, session,
+						   process_synthesized_event);
 		if (err < 0) {
 			pr_err("Couldn't synthesize attrs.\n");
 			return err;
 		}
 
-		err = perf_event__synthesize_event_types(process_synthesized_event,
-							 session);
+		err = perf_event__synthesize_event_types(tool, process_synthesized_event,
+							 machine);
 		if (err < 0) {
 			pr_err("Couldn't synthesize event_types.\n");
 			return err;
@@ -651,56 +542,49 @@ static int __cmd_record(int argc, const char **argv)
 			 * return this more properly and also
 			 * propagate errors that now are calling die()
 			 */
-			err = perf_event__synthesize_tracing_data(output, evsel_list,
-								  process_synthesized_event,
-								  session);
+			err = perf_event__synthesize_tracing_data(tool, output, evsel_list,
+								  process_synthesized_event);
 			if (err <= 0) {
 				pr_err("Couldn't record tracing data.\n");
 				return err;
 			}
-			advance_output(err);
+			advance_output(rec, err);
 		}
 	}
 
-	machine = perf_session__find_host_machine(session);
-	if (!machine) {
-		pr_err("Couldn't find native kernel information.\n");
-		return -1;
-	}
-
-	err = perf_event__synthesize_kernel_mmap(process_synthesized_event,
-						 session, machine, "_text");
+	err = perf_event__synthesize_kernel_mmap(tool, process_synthesized_event,
+						 machine, "_text");
 	if (err < 0)
-		err = perf_event__synthesize_kernel_mmap(process_synthesized_event,
-							 session, machine, "_stext");
+		err = perf_event__synthesize_kernel_mmap(tool, process_synthesized_event,
+							 machine, "_stext");
 	if (err < 0)
 		pr_err("Couldn't record kernel reference relocation symbol\n"
 		       "Symbol resolution may be skewed if relocation was used (e.g. kexec).\n"
 		       "Check /proc/kallsyms permission or run as root.\n");
 
-	err = perf_event__synthesize_modules(process_synthesized_event,
-					     session, machine);
+	err = perf_event__synthesize_modules(tool, process_synthesized_event,
+					     machine);
 	if (err < 0)
 		pr_err("Couldn't record kernel module information.\n"
 		       "Symbol resolution may be skewed if relocation was used (e.g. kexec).\n"
 		       "Check /proc/modules permission or run as root.\n");
 
 	if (perf_guest)
-		perf_session__process_machines(session,
+		perf_session__process_machines(session, tool,
 					       perf_event__synthesize_guest_os);
 
-	if (!system_wide)
-		perf_event__synthesize_thread_map(evsel_list->threads,
+	if (!opts->system_wide)
+		perf_event__synthesize_thread_map(tool, evsel_list->threads,
 						  process_synthesized_event,
-						  session);
+						  machine);
 	else
-		perf_event__synthesize_threads(process_synthesized_event,
-					       session);
+		perf_event__synthesize_threads(tool, process_synthesized_event,
+					       machine);
 
-	if (realtime_prio) {
+	if (rec->realtime_prio) {
 		struct sched_param param;
 
-		param.sched_priority = realtime_prio;
+		param.sched_priority = rec->realtime_prio;
 		if (sched_setscheduler(0, SCHED_FIFO, &param)) {
 			pr_err("Could not set realtime priority.\n");
 			exit(-1);
@@ -713,14 +597,14 @@ static int __cmd_record(int argc, const char **argv)
 	 * Let the child rip
 	 */
 	if (forks)
-		close(go_pipe[1]);
+		perf_evlist__start_workload(evsel_list);
 
 	for (;;) {
-		int hits = samples;
+		int hits = rec->samples;
 
-		mmap_read_all();
+		perf_record__mmap_read_all(rec);
 
-		if (hits == samples) {
+		if (hits == rec->samples) {
 			if (done)
 				break;
 			err = poll(evsel_list->pollfd, evsel_list->nr_fds, -1);
@@ -741,9 +625,9 @@ static int __cmd_record(int argc, const char **argv)
 	 */
 	fprintf(stderr,
 		"[ perf record: Captured and wrote %.3f MB %s (~%" PRIu64 " samples) ]\n",
-		(double)bytes_written / 1024.0 / 1024.0,
+		(double)rec->bytes_written / 1024.0 / 1024.0,
 		output_name,
-		bytes_written / 24);
+		rec->bytes_written / 24);
 
 	return 0;
 
@@ -758,58 +642,89 @@ static const char * const record_usage[] = {
 	NULL
 };
 
-static bool force, append_file;
+/*
+ * XXX Ideally would be local to cmd_record() and passed to a perf_record__new
+ * because we need to have access to it in perf_record__exit, that is called
+ * after cmd_record() exits, but since record_options need to be accessible to
+ * builtin-script, leave it here.
+ *
+ * At least we don't ouch it in all the other functions here directly.
+ *
+ * Just say no to tons of global variables, sigh.
+ */
+static struct perf_record record = {
+	.opts = {
+		.target_pid	     = -1,
+		.target_tid	     = -1,
+		.mmap_pages	     = UINT_MAX,
+		.user_freq	     = UINT_MAX,
+		.user_interval	     = ULLONG_MAX,
+		.freq		     = 1000,
+		.sample_id_all_avail = true,
+	},
+	.write_mode = WRITE_FORCE,
+	.file_new   = true,
+};
 
+/*
+ * XXX Will stay a global variable till we fix builtin-script.c to stop messing
+ * with it and switch to use the library functions in perf_evlist that came
+ * from builtin-record.c, i.e. use perf_record_opts,
+ * perf_evlist__prepare_workload, etc instead of fork+exec'in 'perf record',
+ * using pipes, etc.
+ */
 const struct option record_options[] = {
-	OPT_CALLBACK('e', "event", &evsel_list, "event",
+	OPT_CALLBACK('e', "event", &record.evlist, "event",
 		     "event selector. use 'perf list' to list available events",
 		     parse_events_option),
-	OPT_CALLBACK(0, "filter", &evsel_list, "filter",
+	OPT_CALLBACK(0, "filter", &record.evlist, "filter",
 		     "event filter", parse_filter),
-	OPT_INTEGER('p', "pid", &target_pid,
+	OPT_INTEGER('p', "pid", &record.opts.target_pid,
 		    "record events on existing process id"),
-	OPT_INTEGER('t', "tid", &target_tid,
+	OPT_INTEGER('t', "tid", &record.opts.target_tid,
 		    "record events on existing thread id"),
-	OPT_INTEGER('r', "realtime", &realtime_prio,
+	OPT_INTEGER('r', "realtime", &record.realtime_prio,
 		    "collect data with this RT SCHED_FIFO priority"),
-	OPT_BOOLEAN('D', "no-delay", &nodelay,
+	OPT_BOOLEAN('D', "no-delay", &record.opts.no_delay,
 		    "collect data without buffering"),
-	OPT_BOOLEAN('R', "raw-samples", &raw_samples,
+	OPT_BOOLEAN('R', "raw-samples", &record.opts.raw_samples,
 		    "collect raw sample records from all opened counters"),
-	OPT_BOOLEAN('a', "all-cpus", &system_wide,
+	OPT_BOOLEAN('a', "all-cpus", &record.opts.system_wide,
 			    "system-wide collection from all CPUs"),
-	OPT_BOOLEAN('A', "append", &append_file,
+	OPT_BOOLEAN('A', "append", &record.append_file,
 			    "append to the output file to do incremental profiling"),
-	OPT_STRING('C', "cpu", &cpu_list, "cpu",
+	OPT_STRING('C', "cpu", &record.opts.cpu_list, "cpu",
 		    "list of cpus to monitor"),
-	OPT_BOOLEAN('f', "force", &force,
+	OPT_BOOLEAN('f', "force", &record.force,
 			"overwrite existing data file (deprecated)"),
-	OPT_U64('c', "count", &user_interval, "event period to sample"),
-	OPT_STRING('o', "output", &output_name, "file",
+	OPT_U64('c', "count", &record.opts.user_interval, "event period to sample"),
+	OPT_STRING('o', "output", &record.output_name, "file",
 		    "output file name"),
-	OPT_BOOLEAN('i', "no-inherit", &no_inherit,
+	OPT_BOOLEAN('i', "no-inherit", &record.opts.no_inherit,
 		    "child tasks do not inherit counters"),
-	OPT_UINTEGER('F', "freq", &user_freq, "profile at this frequency"),
-	OPT_UINTEGER('m', "mmap-pages", &mmap_pages, "number of mmap data pages"),
-	OPT_BOOLEAN(0, "group", &group,
+	OPT_UINTEGER('F', "freq", &record.opts.user_freq, "profile at this frequency"),
+	OPT_UINTEGER('m', "mmap-pages", &record.opts.mmap_pages,
+		     "number of mmap data pages"),
+	OPT_BOOLEAN(0, "group", &record.opts.group,
 		    "put the counters into a counter group"),
-	OPT_BOOLEAN('g', "call-graph", &call_graph,
+	OPT_BOOLEAN('g', "call-graph", &record.opts.call_graph,
 		    "do call-graph (stack chain/backtrace) recording"),
 	OPT_INCR('v', "verbose", &verbose,
 		    "be more verbose (show counter open errors, etc)"),
 	OPT_BOOLEAN('q', "quiet", &quiet, "don't print any message"),
-	OPT_BOOLEAN('s', "stat", &inherit_stat,
+	OPT_BOOLEAN('s', "stat", &record.opts.inherit_stat,
 		    "per thread counts"),
-	OPT_BOOLEAN('d', "data", &sample_address,
+	OPT_BOOLEAN('d', "data", &record.opts.sample_address,
 		    "Sample addresses"),
-	OPT_BOOLEAN('T', "timestamp", &sample_time, "Sample timestamps"),
-	OPT_BOOLEAN('n', "no-samples", &no_samples,
+	OPT_BOOLEAN('T', "timestamp", &record.opts.sample_time, "Sample timestamps"),
+	OPT_BOOLEAN('P', "period", &record.opts.period, "Sample period"),
+	OPT_BOOLEAN('n', "no-samples", &record.opts.no_samples,
 		    "don't sample"),
-	OPT_BOOLEAN('N', "no-buildid-cache", &no_buildid_cache,
+	OPT_BOOLEAN('N', "no-buildid-cache", &record.no_buildid_cache,
 		    "do not update the buildid cache"),
-	OPT_BOOLEAN('B', "no-buildid", &no_buildid,
+	OPT_BOOLEAN('B', "no-buildid", &record.no_buildid,
 		    "do not collect buildids in perf.data"),
-	OPT_CALLBACK('G', "cgroup", &evsel_list, "name",
+	OPT_CALLBACK('G', "cgroup", &record.evlist, "name",
 		     "monitor event in cgroup name only",
 		     parse_cgroups),
 	OPT_END()
@@ -819,6 +734,8 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 {
 	int err = -ENOMEM;
 	struct perf_evsel *pos;
+	struct perf_evlist *evsel_list;
+	struct perf_record *rec = &record;
 
 	perf_header__set_cmdline(argc, argv);
 
@@ -826,23 +743,25 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 	if (evsel_list == NULL)
 		return -ENOMEM;
 
+	rec->evlist = evsel_list;
+
 	argc = parse_options(argc, argv, record_options, record_usage,
 			    PARSE_OPT_STOP_AT_NON_OPTION);
-	if (!argc && target_pid == -1 && target_tid == -1 &&
-		!system_wide && !cpu_list)
+	if (!argc && rec->opts.target_pid == -1 && rec->opts.target_tid == -1 &&
+		!rec->opts.system_wide && !rec->opts.cpu_list)
 		usage_with_options(record_usage, record_options);
 
-	if (force && append_file) {
+	if (rec->force && rec->append_file) {
 		fprintf(stderr, "Can't overwrite and append at the same time."
 				" You need to choose between -f and -A");
 		usage_with_options(record_usage, record_options);
-	} else if (append_file) {
-		write_mode = WRITE_APPEND;
+	} else if (rec->append_file) {
+		rec->write_mode = WRITE_APPEND;
 	} else {
-		write_mode = WRITE_FORCE;
+		rec->write_mode = WRITE_FORCE;
 	}
 
-	if (nr_cgroups && !system_wide) {
+	if (nr_cgroups && !rec->opts.system_wide) {
 		fprintf(stderr, "cgroup monitoring only available in"
 			" system-wide mode\n");
 		usage_with_options(record_usage, record_options);
@@ -860,7 +779,7 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 "If some relocation was applied (e.g. kexec) symbols may be misresolved\n"
 "even with a suitable vmlinux or kallsyms file.\n\n");
 
-	if (no_buildid_cache || no_buildid)
+	if (rec->no_buildid_cache || rec->no_buildid)
 		disable_buildid_cache();
 
 	if (evsel_list->nr_entries == 0 &&
@@ -869,43 +788,37 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 		goto out_symbol_exit;
 	}
 
-	if (target_pid != -1)
-		target_tid = target_pid;
+	if (rec->opts.target_pid != -1)
+		rec->opts.target_tid = rec->opts.target_pid;
 
-	if (perf_evlist__create_maps(evsel_list, target_pid,
-				     target_tid, cpu_list) < 0)
+	if (perf_evlist__create_maps(evsel_list, rec->opts.target_pid,
+				     rec->opts.target_tid, rec->opts.cpu_list) < 0)
 		usage_with_options(record_usage, record_options);
 
 	list_for_each_entry(pos, &evsel_list->entries, node) {
-		if (perf_evsel__alloc_fd(pos, evsel_list->cpus->nr,
-					 evsel_list->threads->nr) < 0)
-			goto out_free_fd;
 		if (perf_header__push_event(pos->attr.config, event_name(pos)))
 			goto out_free_fd;
 	}
 
-	if (perf_evlist__alloc_pollfd(evsel_list) < 0)
-		goto out_free_fd;
-
-	if (user_interval != ULLONG_MAX)
-		default_interval = user_interval;
-	if (user_freq != UINT_MAX)
-		freq = user_freq;
+	if (rec->opts.user_interval != ULLONG_MAX)
+		rec->opts.default_interval = rec->opts.user_interval;
+	if (rec->opts.user_freq != UINT_MAX)
+		rec->opts.freq = rec->opts.user_freq;
 
 	/*
 	 * User specified count overrides default frequency.
 	 */
-	if (default_interval)
-		freq = 0;
-	else if (freq) {
-		default_interval = freq;
+	if (rec->opts.default_interval)
+		rec->opts.freq = 0;
+	else if (rec->opts.freq) {
+		rec->opts.default_interval = rec->opts.freq;
 	} else {
 		fprintf(stderr, "frequency and count are zero, aborting\n");
 		err = -EINVAL;
 		goto out_free_fd;
 	}
 
-	err = __cmd_record(argc, argv);
+	err = __cmd_record(&record, argc, argv);
 out_free_fd:
 	perf_evlist__delete_maps(evsel_list);
 out_symbol_exit:

tools/perf/builtin-report.c

@@ -25,6 +25,7 @@
 #include "util/evsel.h"
 #include "util/header.h"
 #include "util/session.h"
+#include "util/tool.h"
 
 #include "util/parse-options.h"
 #include "util/parse-events.h"
@@ -35,38 +36,35 @@
 
 #include <linux/bitmap.h>
 
-static char		const *input_name = "perf.data";
-
-static bool		force, use_tui, use_stdio;
-static bool		hide_unresolved;
-static bool		dont_use_callchains;
-static bool		show_full_info;
-
-static bool		show_threads;
-static struct perf_read_values	show_threads_values;
-
-static const char	default_pretty_printing_style[] = "normal";
-static const char	*pretty_printing_style = default_pretty_printing_style;
-
-static char		callchain_default_opt[] = "fractal,0.5,callee";
-static bool		inverted_callchain;
-static symbol_filter_t	annotate_init;
-
-static const char	*cpu_list;
-static DECLARE_BITMAP(cpu_bitmap, MAX_NR_CPUS);
+struct perf_report {
+	struct perf_tool	tool;
+	struct perf_session	*session;
+	char const		*input_name;
+	bool			force, use_tui, use_stdio;
+	bool			hide_unresolved;
+	bool			dont_use_callchains;
+	bool			show_full_info;
+	bool			show_threads;
+	bool			inverted_callchain;
+	struct perf_read_values	show_threads_values;
+	const char		*pretty_printing_style;
+	symbol_filter_t		annotate_init;
+	const char		*cpu_list;
+	DECLARE_BITMAP(cpu_bitmap, MAX_NR_CPUS);
+};
 
-static int perf_session__add_hist_entry(struct perf_session *session,
-					struct addr_location *al,
-					struct perf_sample *sample,
-					struct perf_evsel *evsel)
+static int perf_evsel__add_hist_entry(struct perf_evsel *evsel,
+				      struct addr_location *al,
+				      struct perf_sample *sample,
+				      struct machine *machine)
 {
 	struct symbol *parent = NULL;
 	int err = 0;
 	struct hist_entry *he;
 
 	if ((sort__has_parent || symbol_conf.use_callchain) && sample->callchain) {
-		err = perf_session__resolve_callchain(session, al->thread,
-						      sample->callchain, &parent);
+		err = machine__resolve_callchain(machine, evsel, al->thread,
+						 sample->callchain, &parent);
 		if (err)
 			return err;
 	}
@@ -76,7 +74,8 @@ static int perf_session__add_hist_entry(struct perf_session *session,
 		return -ENOMEM;
 
 	if (symbol_conf.use_callchain) {
-		err = callchain_append(he->callchain, &session->callchain_cursor,
+		err = callchain_append(he->callchain,
+				       &evsel->hists.callchain_cursor,
 				       sample->period);
 		if (err)
 			return err;
@@ -92,8 +91,7 @@ static int perf_session__add_hist_entry(struct perf_session *session,
 		assert(evsel != NULL);
 
 		err = -ENOMEM;
-		if (notes->src == NULL &&
-		    symbol__alloc_hist(he->ms.sym, session->evlist->nr_entries) < 0)
+		if (notes->src == NULL && symbol__alloc_hist(he->ms.sym) < 0)
 			goto out;
 
 		err = hist_entry__inc_addr_samples(he, evsel->idx, al->addr);
@@ -106,30 +104,32 @@ out:
 }
 
 
-static int process_sample_event(union perf_event *event,
+static int process_sample_event(struct perf_tool *tool,
+				union perf_event *event,
 				struct perf_sample *sample,
 				struct perf_evsel *evsel,
-				struct perf_session *session)
+				struct machine *machine)
 {
+	struct perf_report *rep = container_of(tool, struct perf_report, tool);
 	struct addr_location al;
 
-	if (perf_event__preprocess_sample(event, session, &al, sample,
-					  annotate_init) < 0) {
+	if (perf_event__preprocess_sample(event, machine, &al, sample,
+					  rep->annotate_init) < 0) {
 		fprintf(stderr, "problem processing %d event, skipping it.\n",
 			event->header.type);
 		return -1;
 	}
 
-	if (al.filtered || (hide_unresolved && al.sym == NULL))
+	if (al.filtered || (rep->hide_unresolved && al.sym == NULL))
 		return 0;
 
-	if (cpu_list && !test_bit(sample->cpu, cpu_bitmap))
+	if (rep->cpu_list && !test_bit(sample->cpu, rep->cpu_bitmap))
 		return 0;
 
 	if (al.map != NULL)
 		al.map->dso->hit = 1;
 
-	if (perf_session__add_hist_entry(session, &al, sample, evsel)) {
+	if (perf_evsel__add_hist_entry(evsel, &al, sample, machine)) {
 		pr_debug("problem incrementing symbol period, skipping event\n");
 		return -1;
 	}
@@ -137,15 +137,17 @@ static int process_sample_event(union perf_event *event,
 	return 0;
 }
 
-static int process_read_event(union perf_event *event,
+static int process_read_event(struct perf_tool *tool,
+			      union perf_event *event,
 			      struct perf_sample *sample __used,
-			      struct perf_session *session)
+			      struct perf_evsel *evsel,
+			      struct machine *machine __used)
 {
-	struct perf_evsel *evsel = perf_evlist__id2evsel(session->evlist,
-							 event->read.id);
-	if (show_threads) {
+	struct perf_report *rep = container_of(tool, struct perf_report, tool);
+
+	if (rep->show_threads) {
 		const char *name = evsel ? event_name(evsel) : "unknown";
-		perf_read_values_add_value(&show_threads_values,
+		perf_read_values_add_value(&rep->show_threads_values,
 					   event->read.pid, event->read.tid,
 					   event->read.id,
 					   name,
@@ -159,8 +161,10 @@ static int process_read_event(union perf_event *event,
 	return 0;
 }
 
-static int perf_session__setup_sample_type(struct perf_session *self)
+static int perf_report__setup_sample_type(struct perf_report *rep)
 {
+	struct perf_session *self = rep->session;
+
 	if (!(self->sample_type & PERF_SAMPLE_CALLCHAIN)) {
 		if (sort__has_parent) {
 			ui__warning("Selected --sort parent, but no "
@@ -173,7 +177,8 @@ static int perf_session__setup_sample_type(struct perf_session *self)
 				    "you call 'perf record' without -g?\n");
 			return -1;
 		}
-	} else if (!dont_use_callchains && callchain_param.mode != CHAIN_NONE &&
+	} else if (!rep->dont_use_callchains &&
+		   callchain_param.mode != CHAIN_NONE &&
 		   !symbol_conf.use_callchain) {
 			symbol_conf.use_callchain = true;
 			if (callchain_register_param(&callchain_param) < 0) {
@@ -186,22 +191,6 @@ static int perf_session__setup_sample_type(struct perf_session *self)
 	return 0;
 }
 
-static struct perf_event_ops event_ops = {
-	.sample		 = process_sample_event,
-	.mmap		 = perf_event__process_mmap,
-	.comm		 = perf_event__process_comm,
-	.exit		 = perf_event__process_task,
-	.fork		 = perf_event__process_task,
-	.lost		 = perf_event__process_lost,
-	.read		 = process_read_event,
-	.attr		 = perf_event__process_attr,
-	.event_type	 = perf_event__process_event_type,
-	.tracing_data	 = perf_event__process_tracing_data,
-	.build_id	 = perf_event__process_build_id,
-	.ordered_samples = true,
-	.ordering_requires_timestamps = true,
-};
-
 extern volatile int session_done;
 
 static void sig_handler(int sig __used)
@@ -224,6 +213,7 @@ static size_t hists__fprintf_nr_sample_events(struct hists *self,
 }
 
 static int perf_evlist__tty_browse_hists(struct perf_evlist *evlist,
+					 struct perf_report *rep,
 					 const char *help)
 {
 	struct perf_evsel *pos;
@@ -241,18 +231,18 @@ static int perf_evlist__tty_browse_hists(struct perf_evlist *evlist,
 	    parent_pattern == default_parent_pattern) {
 		fprintf(stdout, "#\n# (%s)\n#\n", help);
 
-		if (show_threads) {
-			bool style = !strcmp(pretty_printing_style, "raw");
-			perf_read_values_display(stdout, &show_threads_values,
+		if (rep->show_threads) {
+			bool style = !strcmp(rep->pretty_printing_style, "raw");
+			perf_read_values_display(stdout, &rep->show_threads_values,
 						 style);
-			perf_read_values_destroy(&show_threads_values);
+			perf_read_values_destroy(&rep->show_threads_values);
 		}
 	}
 
 	return 0;
 }
 
-static int __cmd_report(void)
+static int __cmd_report(struct perf_report *rep)
 {
 	int ret = -EINVAL;
 	u64 nr_samples;
@@ -264,27 +254,31 @@ static int __cmd_report(void)
 
 	signal(SIGINT, sig_handler);
 
-	session = perf_session__new(input_name, O_RDONLY, force, false, &event_ops);
+	session = perf_session__new(rep->input_name, O_RDONLY,
+				    rep->force, false, &rep->tool);
 	if (session == NULL)
 		return -ENOMEM;
 
-	if (cpu_list) {
-		ret = perf_session__cpu_bitmap(session, cpu_list, cpu_bitmap);
+	rep->session = session;
+
+	if (rep->cpu_list) {
+		ret = perf_session__cpu_bitmap(session, rep->cpu_list,
+					       rep->cpu_bitmap);
 		if (ret)
 			goto out_delete;
 	}
 
 	if (use_browser <= 0)
-		perf_session__fprintf_info(session, stdout, show_full_info);
+		perf_session__fprintf_info(session, stdout, rep->show_full_info);
 
-	if (show_threads)
-		perf_read_values_init(&show_threads_values);
+	if (rep->show_threads)
+		perf_read_values_init(&rep->show_threads_values);
 
-	ret = perf_session__setup_sample_type(session);
+	ret = perf_report__setup_sample_type(rep);
 	if (ret)
 		goto out_delete;
 
-	ret = perf_session__process_events(session, &event_ops);
+	ret = perf_session__process_events(session, &rep->tool);
 	if (ret)
 		goto out_delete;
 
@@ -327,7 +321,7 @@ static int __cmd_report(void)
 	}
 
 	if (nr_samples == 0) {
-		ui__warning("The %s file has no samples!\n", input_name);
+		ui__warning("The %s file has no samples!\n", session->filename);
 		goto out_delete;
 	}
 
@@ -335,7 +329,7 @@ static int __cmd_report(void)
 		perf_evlist__tui_browse_hists(session->evlist, help,
 					      NULL, NULL, 0);
 	} else
-		perf_evlist__tty_browse_hists(session->evlist, help);
+		perf_evlist__tty_browse_hists(session->evlist, rep, help);
 
 out_delete:
 	/*
@@ -354,9 +348,9 @@ out_delete:
 }
 
 static int
-parse_callchain_opt(const struct option *opt __used, const char *arg,
-		    int unset)
+parse_callchain_opt(const struct option *opt, const char *arg, int unset)
 {
+	struct perf_report *rep = (struct perf_report *)opt->value;
 	char *tok, *tok2;
 	char *endptr;
 
@@ -364,7 +358,7 @@ parse_callchain_opt(const struct option *opt __used, const char *arg,
 	 * --no-call-graph
 	 */
 	if (unset) {
-		dont_use_callchains = true;
+		rep->dont_use_callchains = true;
 		return 0;
 	}
 
@@ -412,7 +406,7 @@ parse_callchain_opt(const struct option *opt __used, const char *arg,
 		goto setup;
 
 	if (tok2[0] != 'c') {
-		callchain_param.print_limit = strtod(tok2, &endptr);
+		callchain_param.print_limit = strtoul(tok2, &endptr, 0);
 		tok2 = strtok(NULL, ",");
 		if (!tok2)
 			goto setup;
@@ -433,13 +427,34 @@ setup:
 	return 0;
 }
 
-static const char * const report_usage[] = {
-	"perf report [<options>] <command>",
-	NULL
-};
-
-static const struct option options[] = {
-	OPT_STRING('i', "input", &input_name, "file",
+int cmd_report(int argc, const char **argv, const char *prefix __used)
+{
+	struct stat st;
+	char callchain_default_opt[] = "fractal,0.5,callee";
+	const char * const report_usage[] = {
+		"perf report [<options>]",
+		NULL
+	};
+	struct perf_report report = {
+		.tool = {
+			.sample		 = process_sample_event,
+			.mmap		 = perf_event__process_mmap,
+			.comm		 = perf_event__process_comm,
+			.exit		 = perf_event__process_task,
+			.fork		 = perf_event__process_task,
+			.lost		 = perf_event__process_lost,
+			.read		 = process_read_event,
+			.attr		 = perf_event__process_attr,
+			.event_type	 = perf_event__process_event_type,
+			.tracing_data	 = perf_event__process_tracing_data,
+			.build_id	 = perf_event__process_build_id,
+			.ordered_samples = true,
+			.ordering_requires_timestamps = true,
+		},
+		.pretty_printing_style	 = "normal",
+	};
+	const struct option options[] = {
+	OPT_STRING('i', "input", &report.input_name, "file",
 		    "input file name"),
 	OPT_INCR('v', "verbose", &verbose,
 		    "be more verbose (show symbol address, etc)"),
@@ -449,17 +464,18 @@ static const struct option options[] = {
 		   "file", "vmlinux pathname"),
 	OPT_STRING(0, "kallsyms", &symbol_conf.kallsyms_name,
 		   "file", "kallsyms pathname"),
-	OPT_BOOLEAN('f', "force", &force, "don't complain, do it"),
+	OPT_BOOLEAN('f', "force", &report.force, "don't complain, do it"),
 	OPT_BOOLEAN('m', "modules", &symbol_conf.use_modules,
 		    "load module symbols - WARNING: use only with -k and LIVE kernel"),
 	OPT_BOOLEAN('n', "show-nr-samples", &symbol_conf.show_nr_samples,
 		    "Show a column with the number of samples"),
-	OPT_BOOLEAN('T', "threads", &show_threads,
+	OPT_BOOLEAN('T', "threads", &report.show_threads,
 		    "Show per-thread event counters"),
-	OPT_STRING(0, "pretty", &pretty_printing_style, "key",
+	OPT_STRING(0, "pretty", &report.pretty_printing_style, "key",
 		   "pretty printing style key: normal raw"),
-	OPT_BOOLEAN(0, "tui", &use_tui, "Use the TUI interface"),
-	OPT_BOOLEAN(0, "stdio", &use_stdio, "Use the stdio interface"),
+	OPT_BOOLEAN(0, "tui", &report.use_tui, "Use the TUI interface"),
+	OPT_BOOLEAN(0, "stdio", &report.use_stdio,
+		    "Use the stdio interface"),
 	OPT_STRING('s', "sort", &sort_order, "key[,key2...]",
 		   "sort by key(s): pid, comm, dso, symbol, parent"),
 	OPT_BOOLEAN(0, "showcpuutilization", &symbol_conf.show_cpu_utilization,
@@ -468,13 +484,14 @@ static const struct option options[] = {
 		   "regex filter to identify parent, see: '--sort parent'"),
 	OPT_BOOLEAN('x', "exclude-other", &symbol_conf.exclude_other,
 		    "Only display entries with parent-match"),
-	OPT_CALLBACK_DEFAULT('g', "call-graph", NULL, "output_type,min_percent, call_order",
-		     "Display callchains using output_type (graph, flat, fractal, or none) , min percent threshold and callchain order. "
+	OPT_CALLBACK_DEFAULT('g', "call-graph", &report, "output_type,min_percent[,print_limit],call_order",
+		     "Display callchains using output_type (graph, flat, fractal, or none) , min percent threshold, optional print limit and callchain order. "
 		     "Default: fractal,0.5,callee", &parse_callchain_opt, callchain_default_opt),
-	OPT_BOOLEAN('G', "inverted", &inverted_callchain, "alias for inverted call graph"),
+	OPT_BOOLEAN('G', "inverted", &report.inverted_callchain,
+		    "alias for inverted call graph"),
 	OPT_STRING('d', "dsos", &symbol_conf.dso_list_str, "dso[,dso...]",
 		   "only consider symbols in these dsos"),
-	OPT_STRING('C', "comms", &symbol_conf.comm_list_str, "comm[,comm...]",
+	OPT_STRING('c', "comms", &symbol_conf.comm_list_str, "comm[,comm...]",
 		   "only consider symbols in these comms"),
 	OPT_STRING('S', "symbols", &symbol_conf.sym_list_str, "symbol[,symbol...]",
 		   "only consider these symbols"),
@@ -484,12 +501,13 @@ static const struct option options[] = {
 	OPT_STRING('t', "field-separator", &symbol_conf.field_sep, "separator",
 		   "separator for columns, no spaces will be added between "
 		   "columns '.' is reserved."),
-	OPT_BOOLEAN('U', "hide-unresolved", &hide_unresolved,
+	OPT_BOOLEAN('U', "hide-unresolved", &report.hide_unresolved,
 		    "Only display entries resolved to a symbol"),
 	OPT_STRING(0, "symfs", &symbol_conf.symfs, "directory",
 		    "Look for files with symbols relative to this directory"),
-	OPT_STRING('c', "cpu", &cpu_list, "cpu", "list of cpus to profile"),
-	OPT_BOOLEAN('I', "show-info", &show_full_info,
+	OPT_STRING('C', "cpu", &report.cpu_list, "cpu",
+		   "list of cpus to profile"),
+	OPT_BOOLEAN('I', "show-info", &report.show_full_info,
 		    "Display extended information about perf.data file"),
 	OPT_BOOLEAN(0, "source", &symbol_conf.annotate_src,
 		    "Interleave source code with assembly code (default)"),
@@ -500,24 +518,30 @@ static const struct option options[] = {
 	OPT_BOOLEAN(0, "show-total-period", &symbol_conf.show_total_period,
 		    "Show a column with the sum of periods"),
 	OPT_END()
-};
+	};
 
-int cmd_report(int argc, const char **argv, const char *prefix __used)
-{
 	argc = parse_options(argc, argv, options, report_usage, 0);
 
-	if (use_stdio)
+	if (report.use_stdio)
 		use_browser = 0;
-	else if (use_tui)
+	else if (report.use_tui)
 		use_browser = 1;
 
-	if (inverted_callchain)
+	if (report.inverted_callchain)
 		callchain_param.order = ORDER_CALLER;
 
-	if (strcmp(input_name, "-") != 0)
+	if (!report.input_name || !strlen(report.input_name)) {
+		if (!fstat(STDIN_FILENO, &st) && S_ISFIFO(st.st_mode))
+			report.input_name = "-";
+		else
+			report.input_name = "perf.data";
+	}
+
+	if (strcmp(report.input_name, "-") != 0)
 		setup_browser(true);
 	else
 		use_browser = 0;
+
 	/*
 	 * Only in the newt browser we are doing integrated annotation,
 	 * so don't allocate extra space that won't be used in the stdio
@@ -525,7 +549,7 @@ int cmd_report(int argc, const char **argv, const char *prefix __used)
 	 */
 	if (use_browser > 0) {
 		symbol_conf.priv_size = sizeof(struct annotation);
-		annotate_init	      = symbol__annotate_init;
+		report.annotate_init  = symbol__annotate_init;
 		/*
  		 * For searching by name on the "Browse map details".
  		 * providing it only in verbose mode not to bloat too
@@ -572,5 +596,5 @@ int cmd_report(int argc, const char **argv, const char *prefix __used)
 	sort_entry__setup_elide(&sort_comm, symbol_conf.comm_list, "comm", stdout);
 	sort_entry__setup_elide(&sort_sym, symbol_conf.sym_list, "symbol", stdout);
 
-	return __cmd_report();
+	return __cmd_report(&report);
 }

tools/perf/builtin-sched.c

@@ -2,11 +2,14 @@
 #include "perf.h"
 
 #include "util/util.h"
+#include "util/evlist.h"
 #include "util/cache.h"
+#include "util/evsel.h"
 #include "util/symbol.h"
 #include "util/thread.h"
 #include "util/header.h"
 #include "util/session.h"
+#include "util/tool.h"
 
 #include "util/parse-options.h"
 #include "util/trace-event.h"
@@ -19,7 +22,7 @@
 #include <pthread.h>
 #include <math.h>
 
-static char			const *input_name = "perf.data";
+static const char		*input_name;
 
 static char			default_sort_order[] = "avg, max, switch, runtime";
 static const char		*sort_order = default_sort_order;
@@ -723,21 +726,21 @@ struct trace_migrate_task_event {
 
 struct trace_sched_handler {
 	void (*switch_event)(struct trace_switch_event *,
-			     struct perf_session *,
+			     struct machine *,
 			     struct event *,
 			     int cpu,
 			     u64 timestamp,
 			     struct thread *thread);
 
 	void (*runtime_event)(struct trace_runtime_event *,
-			      struct perf_session *,
+			      struct machine *,
 			      struct event *,
 			      int cpu,
 			      u64 timestamp,
 			      struct thread *thread);
 
 	void (*wakeup_event)(struct trace_wakeup_event *,
-			     struct perf_session *,
+			     struct machine *,
 			     struct event *,
 			     int cpu,
 			     u64 timestamp,
@@ -750,7 +753,7 @@ struct trace_sched_handler {
 			   struct thread *thread);
 
 	void (*migrate_task_event)(struct trace_migrate_task_event *,
-			   struct perf_session *session,
+			   struct machine *machine,
 			   struct event *,
 			   int cpu,
 			   u64 timestamp,
@@ -760,7 +763,7 @@ struct trace_sched_handler {
 
 static void
 replay_wakeup_event(struct trace_wakeup_event *wakeup_event,
-		    struct perf_session *session __used,
+		    struct machine *machine __used,
 		    struct event *event,
 		    int cpu __used,
 		    u64 timestamp __used,
@@ -787,7 +790,7 @@ static u64 cpu_last_switched[MAX_CPUS];
 
 static void
 replay_switch_event(struct trace_switch_event *switch_event,
-		    struct perf_session *session __used,
+		    struct machine *machine __used,
 		    struct event *event,
 		    int cpu,
 		    u64 timestamp,
@@ -1021,7 +1024,7 @@ add_sched_in_event(struct work_atoms *atoms, u64 timestamp)
 
 static void
 latency_switch_event(struct trace_switch_event *switch_event,
-		     struct perf_session *session,
+		     struct machine *machine,
 		     struct event *event __used,
 		     int cpu,
 		     u64 timestamp,
@@ -1045,8 +1048,8 @@ latency_switch_event(struct trace_switch_event *switch_event,
 		die("hm, delta: %" PRIu64 " < 0 ?\n", delta);
 
 
-	sched_out = perf_session__findnew(session, switch_event->prev_pid);
-	sched_in = perf_session__findnew(session, switch_event->next_pid);
+	sched_out = machine__findnew_thread(machine, switch_event->prev_pid);
+	sched_in = machine__findnew_thread(machine, switch_event->next_pid);
 
 	out_events = thread_atoms_search(&atom_root, sched_out, &cmp_pid);
 	if (!out_events) {
@@ -1074,13 +1077,13 @@ latency_switch_event(struct trace_switch_event *switch_event,
 
 static void
 latency_runtime_event(struct trace_runtime_event *runtime_event,
-		     struct perf_session *session,
+		     struct machine *machine,
 		     struct event *event __used,
 		     int cpu,
 		     u64 timestamp,
 		     struct thread *this_thread __used)
 {
-	struct thread *thread = perf_session__findnew(session, runtime_event->pid);
+	struct thread *thread = machine__findnew_thread(machine, runtime_event->pid);
 	struct work_atoms *atoms = thread_atoms_search(&atom_root, thread, &cmp_pid);
 
 	BUG_ON(cpu >= MAX_CPUS || cpu < 0);
@@ -1097,7 +1100,7 @@ latency_runtime_event(struct trace_runtime_event *runtime_event,
 
 static void
 latency_wakeup_event(struct trace_wakeup_event *wakeup_event,
-		     struct perf_session *session,
+		     struct machine *machine,
 		     struct event *__event __used,
 		     int cpu __used,
 		     u64 timestamp,
@@ -1111,7 +1114,7 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event,
 	if (!wakeup_event->success)
 		return;
 
-	wakee = perf_session__findnew(session, wakeup_event->pid);
+	wakee = machine__findnew_thread(machine, wakeup_event->pid);
 	atoms = thread_atoms_search(&atom_root, wakee, &cmp_pid);
 	if (!atoms) {
 		thread_atoms_insert(wakee);
@@ -1145,7 +1148,7 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event,
 
 static void
 latency_migrate_task_event(struct trace_migrate_task_event *migrate_task_event,
-		     struct perf_session *session,
+		     struct machine *machine,
 		     struct event *__event __used,
 		     int cpu __used,
 		     u64 timestamp,
@@ -1161,7 +1164,7 @@ latency_migrate_task_event(struct trace_migrate_task_event *migrate_task_event,
 	if (profile_cpu == -1)
 		return;
 
-	migrant = perf_session__findnew(session, migrate_task_event->pid);
+	migrant = machine__findnew_thread(machine, migrate_task_event->pid);
 	atoms = thread_atoms_search(&atom_root, migrant, &cmp_pid);
 	if (!atoms) {
 		thread_atoms_insert(migrant);
@@ -1356,12 +1359,13 @@ static void sort_lat(void)
 static struct trace_sched_handler *trace_handler;
 
 static void
-process_sched_wakeup_event(void *data, struct perf_session *session,
+process_sched_wakeup_event(struct perf_tool *tool __used,
 			   struct event *event,
-			   int cpu __used,
-			   u64 timestamp __used,
-			   struct thread *thread __used)
+			   struct perf_sample *sample,
+			   struct machine *machine,
+			   struct thread *thread)
 {
+	void *data = sample->raw_data;
 	struct trace_wakeup_event wakeup_event;
 
 	FILL_COMMON_FIELDS(wakeup_event, event, data);
@@ -1373,8 +1377,8 @@ process_sched_wakeup_event(void *data, struct perf_session *session,
 	FILL_FIELD(wakeup_event, cpu, event, data);
 
 	if (trace_handler->wakeup_event)
-		trace_handler->wakeup_event(&wakeup_event, session, event,
-					    cpu, timestamp, thread);
+		trace_handler->wakeup_event(&wakeup_event, machine, event,
+					    sample->cpu, sample->time, thread);
 }
 
 /*
@@ -1392,7 +1396,7 @@ static char next_shortname2 = '0';
 
 static void
 map_switch_event(struct trace_switch_event *switch_event,
-		 struct perf_session *session,
+		 struct machine *machine,
 		 struct event *event __used,
 		 int this_cpu,
 		 u64 timestamp,
@@ -1420,8 +1424,8 @@ map_switch_event(struct trace_switch_event *switch_event,
 		die("hm, delta: %" PRIu64 " < 0 ?\n", delta);
 
 
-	sched_out = perf_session__findnew(session, switch_event->prev_pid);
-	sched_in = perf_session__findnew(session, switch_event->next_pid);
+	sched_out = machine__findnew_thread(machine, switch_event->prev_pid);
+	sched_in = machine__findnew_thread(machine, switch_event->next_pid);
 
 	curr_thread[this_cpu] = sched_in;
 
@@ -1469,14 +1473,15 @@ map_switch_event(struct trace_switch_event *switch_event,
 	}
 }
 
-
 static void
-process_sched_switch_event(void *data, struct perf_session *session,
+process_sched_switch_event(struct perf_tool *tool __used,
 			   struct event *event,
-			   int this_cpu,
-			   u64 timestamp __used,
-			   struct thread *thread __used)
+			   struct perf_sample *sample,
+			   struct machine *machine,
+			   struct thread *thread)
 {
+	int this_cpu = sample->cpu;
+	void *data = sample->raw_data;
 	struct trace_switch_event switch_event;
 
 	FILL_COMMON_FIELDS(switch_event, event, data);
@@ -1498,19 +1503,20 @@ process_sched_switch_event(void *data, struct perf_session *session,
 			nr_context_switch_bugs++;
 	}
 	if (trace_handler->switch_event)
-		trace_handler->switch_event(&switch_event, session, event,
-					    this_cpu, timestamp, thread);
+		trace_handler->switch_event(&switch_event, machine, event,
+					    this_cpu, sample->time, thread);
 
 	curr_pid[this_cpu] = switch_event.next_pid;
 }
 
 static void
-process_sched_runtime_event(void *data, struct perf_session *session,
-			   struct event *event,
-			   int cpu __used,
-			   u64 timestamp __used,
-			   struct thread *thread __used)
+process_sched_runtime_event(struct perf_tool *tool __used,
+			    struct event *event,
+			    struct perf_sample *sample,
+			    struct machine *machine,
+			    struct thread *thread)
 {
+	void *data = sample->raw_data;
 	struct trace_runtime_event runtime_event;
 
 	FILL_ARRAY(runtime_event, comm, event, data);
@@ -1519,16 +1525,18 @@ process_sched_runtime_event(void *data, struct perf_session *session,
 	FILL_FIELD(runtime_event, vruntime, event, data);
 
 	if (trace_handler->runtime_event)
-		trace_handler->runtime_event(&runtime_event, session, event, cpu, timestamp, thread);
+		trace_handler->runtime_event(&runtime_event, machine, event,
+					     sample->cpu, sample->time, thread);
 }
 
 static void
-process_sched_fork_event(void *data,
+process_sched_fork_event(struct perf_tool *tool __used,
 			 struct event *event,
-			 int cpu __used,
-			 u64 timestamp __used,
-			 struct thread *thread __used)
+			 struct perf_sample *sample,
+			 struct machine *machine __used,
+			 struct thread *thread)
 {
+	void *data = sample->raw_data;
 	struct trace_fork_event fork_event;
 
 	FILL_COMMON_FIELDS(fork_event, event, data);
@@ -1540,13 +1548,14 @@ process_sched_fork_event(void *data,
 
 	if (trace_handler->fork_event)
 		trace_handler->fork_event(&fork_event, event,
-					  cpu, timestamp, thread);
+					  sample->cpu, sample->time, thread);
 }
 
 static void
-process_sched_exit_event(struct event *event,
-			 int cpu __used,
-			 u64 timestamp __used,
+process_sched_exit_event(struct perf_tool *tool __used,
+			 struct event *event,
+			 struct perf_sample *sample __used,
+			 struct machine *machine __used,
 			 struct thread *thread __used)
 {
 	if (verbose)
@@ -1554,12 +1563,13 @@ process_sched_exit_event(struct event *event,
 }
 
 static void
-process_sched_migrate_task_event(void *data, struct perf_session *session,
-			   struct event *event,
-			   int cpu __used,
-			   u64 timestamp __used,
-			   struct thread *thread __used)
+process_sched_migrate_task_event(struct perf_tool *tool __used,
+				 struct event *event,
+				 struct perf_sample *sample,
+				 struct machine *machine,
+				 struct thread *thread)
 {
+	void *data = sample->raw_data;
 	struct trace_migrate_task_event migrate_task_event;
 
 	FILL_COMMON_FIELDS(migrate_task_event, event, data);
@@ -1570,67 +1580,47 @@ process_sched_migrate_task_event(void *data, struct perf_session *session,
 	FILL_FIELD(migrate_task_event, cpu, event, data);
 
 	if (trace_handler->migrate_task_event)
-		trace_handler->migrate_task_event(&migrate_task_event, session,
-						 event, cpu, timestamp, thread);
+		trace_handler->migrate_task_event(&migrate_task_event, machine,
+						  event, sample->cpu,
+						  sample->time, thread);
 }
 
-static void process_raw_event(union perf_event *raw_event __used,
-			      struct perf_session *session, void *data, int cpu,
-			      u64 timestamp, struct thread *thread)
-{
-	struct event *event;
-	int type;
-
-
-	type = trace_parse_common_type(data);
-	event = trace_find_event(type);
-
-	if (!strcmp(event->name, "sched_switch"))
-		process_sched_switch_event(data, session, event, cpu, timestamp, thread);
-	if (!strcmp(event->name, "sched_stat_runtime"))
-		process_sched_runtime_event(data, session, event, cpu, timestamp, thread);
-	if (!strcmp(event->name, "sched_wakeup"))
-		process_sched_wakeup_event(data, session, event, cpu, timestamp, thread);
-	if (!strcmp(event->name, "sched_wakeup_new"))
-		process_sched_wakeup_event(data, session, event, cpu, timestamp, thread);
-	if (!strcmp(event->name, "sched_process_fork"))
-		process_sched_fork_event(data, 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(data, session, event, cpu, timestamp, thread);
-}
+typedef void (*tracepoint_handler)(struct perf_tool *tool, struct event *event,
+				   struct perf_sample *sample,
+				   struct machine *machine,
+				   struct thread *thread);
 
-static int process_sample_event(union perf_event *event,
-				struct perf_sample *sample,
-				struct perf_evsel *evsel __used,
-				struct perf_session *session)
+static int perf_sched__process_tracepoint_sample(struct perf_tool *tool,
+						 union perf_event *event __used,
+						 struct perf_sample *sample,
+						 struct perf_evsel *evsel,
+						 struct machine *machine)
 {
-	struct thread *thread;
-
-	if (!(session->sample_type & PERF_SAMPLE_RAW))
-		return 0;
+	struct thread *thread = machine__findnew_thread(machine, sample->pid);
 
-	thread = perf_session__findnew(session, sample->pid);
 	if (thread == NULL) {
-		pr_debug("problem processing %d event, skipping it.\n",
-			 event->header.type);
+		pr_debug("problem processing %s event, skipping it.\n",
+			 evsel->name);
 		return -1;
 	}
 
-	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
+	evsel->hists.stats.total_period += sample->period;
+	hists__inc_nr_events(&evsel->hists, PERF_RECORD_SAMPLE);
 
-	if (profile_cpu != -1 && profile_cpu != (int)sample->cpu)
-		return 0;
+	if (evsel->handler.func != NULL) {
+		tracepoint_handler f = evsel->handler.func;
 
-	process_raw_event(event, session, sample->raw_data, sample->cpu,
-			  sample->time, thread);
+		if (evsel->handler.data == NULL)
+			evsel->handler.data = trace_find_event(evsel->attr.config);
+
+		f(tool, evsel->handler.data, sample, machine, thread);
+	}
 
 	return 0;
 }
 
-static struct perf_event_ops event_ops = {
-	.sample			= process_sample_event,
+static struct perf_tool perf_sched = {
+	.sample			= perf_sched__process_tracepoint_sample,
 	.comm			= perf_event__process_comm,
 	.lost			= perf_event__process_lost,
 	.fork			= perf_event__process_task,
@@ -1640,13 +1630,25 @@ static struct perf_event_ops event_ops = {
 static void read_events(bool destroy, struct perf_session **psession)
 {
 	int err = -EINVAL;
+	const struct perf_evsel_str_handler handlers[] = {
+		{ "sched:sched_switch",	      process_sched_switch_event, },
+		{ "sched:sched_stat_runtime", process_sched_runtime_event, },
+		{ "sched:sched_wakeup",	      process_sched_wakeup_event, },
+		{ "sched:sched_wakeup_new",   process_sched_wakeup_event, },
+		{ "sched:sched_process_fork", process_sched_fork_event, },
+		{ "sched:sched_process_exit", process_sched_exit_event, },
+		{ "sched:sched_migrate_task", process_sched_migrate_task_event, },
+	};
 	struct perf_session *session = perf_session__new(input_name, O_RDONLY,
-							 0, false, &event_ops);
+							 0, false, &perf_sched);
 	if (session == NULL)
 		die("No Memory");
 
+	err = perf_evlist__set_tracepoints_handlers_array(session->evlist, handlers);
+	assert(err == 0);
+
 	if (perf_session__has_traces(session, "record -R")) {
-		err = perf_session__process_events(session, &event_ops);
+		err = perf_session__process_events(session, &perf_sched);
 		if (err)
 			die("Failed to process events, error %d", err);
 

tools/perf/builtin-script.c

@@ -7,6 +7,7 @@
 #include "util/header.h"
 #include "util/parse-options.h"
 #include "util/session.h"
+#include "util/tool.h"
 #include "util/symbol.h"
 #include "util/thread.h"
 #include "util/trace-event.h"
@@ -23,6 +24,7 @@ static u64			nr_unordered;
 extern const struct option	record_options[];
 static bool			no_callchain;
 static bool			show_full_info;
+static bool			system_wide;
 static const char		*cpu_list;
 static DECLARE_BITMAP(cpu_bitmap, MAX_NR_CPUS);
 
@@ -315,7 +317,7 @@ static bool sample_addr_correlates_sym(struct perf_event_attr *attr)
 
 static void print_sample_addr(union perf_event *event,
 			  struct perf_sample *sample,
-			  struct perf_session *session,
+			  struct machine *machine,
 			  struct thread *thread,
 			  struct perf_event_attr *attr)
 {
@@ -328,11 +330,11 @@ static void print_sample_addr(union perf_event *event,
 	if (!sample_addr_correlates_sym(attr))
 		return;
 
-	thread__find_addr_map(thread, session, cpumode, MAP__FUNCTION,
-			      event->ip.pid, sample->addr, &al);
+	thread__find_addr_map(thread, machine, cpumode, MAP__FUNCTION,
+			      sample->addr, &al);
 	if (!al.map)
-		thread__find_addr_map(thread, session, cpumode, MAP__VARIABLE,
-				      event->ip.pid, sample->addr, &al);
+		thread__find_addr_map(thread, machine, cpumode, MAP__VARIABLE,
+				      sample->addr, &al);
 
 	al.cpu = sample->cpu;
 	al.sym = NULL;
@@ -362,7 +364,7 @@ static void print_sample_addr(union perf_event *event,
 static void process_event(union perf_event *event __unused,
 			  struct perf_sample *sample,
 			  struct perf_evsel *evsel,
-			  struct perf_session *session,
+			  struct machine *machine,
 			  struct thread *thread)
 {
 	struct perf_event_attr *attr = &evsel->attr;
@@ -377,15 +379,15 @@ static void process_event(union perf_event *event __unused,
 				  sample->raw_size);
 
 	if (PRINT_FIELD(ADDR))
-		print_sample_addr(event, sample, session, thread, attr);
+		print_sample_addr(event, sample, machine, thread, attr);
 
 	if (PRINT_FIELD(IP)) {
 		if (!symbol_conf.use_callchain)
 			printf(" ");
 		else
 			printf("\n");
-		perf_session__print_ip(event, sample, session,
-					      PRINT_FIELD(SYM), PRINT_FIELD(DSO));
+		perf_event__print_ip(event, sample, machine, evsel,
+				     PRINT_FIELD(SYM), PRINT_FIELD(DSO));
 	}
 
 	printf("\n");
@@ -432,14 +434,16 @@ static int cleanup_scripting(void)
 	return scripting_ops->stop_script();
 }
 
-static char const		*input_name = "perf.data";
+static const char *input_name;
 
-static int process_sample_event(union perf_event *event,
+static int process_sample_event(struct perf_tool *tool __used,
+				union perf_event *event,
 				struct perf_sample *sample,
 				struct perf_evsel *evsel,
-				struct perf_session *session)
+				struct machine *machine)
 {
-	struct thread *thread = perf_session__findnew(session, event->ip.pid);
+	struct addr_location al;
+	struct thread *thread = machine__findnew_thread(machine, event->ip.tid);
 
 	if (thread == NULL) {
 		pr_debug("problem processing %d event, skipping it.\n",
@@ -458,16 +462,25 @@ static int process_sample_event(union perf_event *event,
 		return 0;
 	}
 
+	if (perf_event__preprocess_sample(event, machine, &al, sample, 0) < 0) {
+		pr_err("problem processing %d event, skipping it.\n",
+		       event->header.type);
+		return -1;
+	}
+
+	if (al.filtered)
+		return 0;
+
 	if (cpu_list && !test_bit(sample->cpu, cpu_bitmap))
 		return 0;
 
-	scripting_ops->process_event(event, sample, evsel, session, thread);
+	scripting_ops->process_event(event, sample, evsel, machine, thread);
 
-	session->hists.stats.total_period += sample->period;
+	evsel->hists.stats.total_period += sample->period;
 	return 0;
 }
 
-static struct perf_event_ops event_ops = {
+static struct perf_tool perf_script = {
 	.sample		 = process_sample_event,
 	.mmap		 = perf_event__process_mmap,
 	.comm		 = perf_event__process_comm,
@@ -494,7 +507,7 @@ static int __cmd_script(struct perf_session *session)
 
 	signal(SIGINT, sig_handler);
 
-	ret = perf_session__process_events(session, &event_ops);
+	ret = perf_session__process_events(session, &perf_script);
 
 	if (debug_mode)
 		pr_err("Misordered timestamps: %" PRIu64 "\n", nr_unordered);
@@ -523,12 +536,6 @@ static struct script_spec *script_spec__new(const char *spec,
 	return s;
 }
 
-static void script_spec__delete(struct script_spec *s)
-{
-	free(s->spec);
-	free(s);
-}
-
 static void script_spec__add(struct script_spec *s)
 {
 	list_add_tail(&s->node, &script_specs);
@@ -554,16 +561,11 @@ static struct script_spec *script_spec__findnew(const char *spec,
 
 	s = script_spec__new(spec, ops);
 	if (!s)
-		goto out_delete_spec;
+		return NULL;
 
 	script_spec__add(s);
 
 	return s;
-
-out_delete_spec:
-	script_spec__delete(s);
-
-	return NULL;
 }
 
 int script_spec_register(const char *spec, struct scripting_ops *ops)
@@ -681,7 +683,8 @@ static int parse_output_fields(const struct option *opt __used,
 			type = PERF_TYPE_RAW;
 		else {
 			fprintf(stderr, "Invalid event type in field string.\n");
-			return -EINVAL;
+			rc = -EINVAL;
+			goto out;
 		}
 
 		if (output[type].user_set)
@@ -923,6 +926,24 @@ static int read_script_info(struct script_desc *desc, const char *filename)
 	return 0;
 }
 
+static char *get_script_root(struct dirent *script_dirent, const char *suffix)
+{
+	char *script_root, *str;
+
+	script_root = strdup(script_dirent->d_name);
+	if (!script_root)
+		return NULL;
+
+	str = (char *)ends_with(script_root, suffix);
+	if (!str) {
+		free(script_root);
+		return NULL;
+	}
+
+	*str = '\0';
+	return script_root;
+}
+
 static int list_available_scripts(const struct option *opt __used,
 				  const char *s __used, int unset __used)
 {
@@ -934,7 +955,6 @@ static int list_available_scripts(const struct option *opt __used,
 	struct script_desc *desc;
 	char first_half[BUFSIZ];
 	char *script_root;
-	char *str;
 
 	snprintf(scripts_path, MAXPATHLEN, "%s/scripts", perf_exec_path());
 
@@ -950,16 +970,14 @@ static int list_available_scripts(const struct option *opt __used,
 			continue;
 
 		for_each_script(lang_path, lang_dir, script_dirent, script_next) {
-			script_root = strdup(script_dirent.d_name);
-			str = (char *)ends_with(script_root, REPORT_SUFFIX);
-			if (str) {
-				*str = '\0';
+			script_root = get_script_root(&script_dirent, REPORT_SUFFIX);
+			if (script_root) {
 				desc = script_desc__findnew(script_root);
 				snprintf(script_path, MAXPATHLEN, "%s/%s",
 					 lang_path, script_dirent.d_name);
 				read_script_info(desc, script_path);
+				free(script_root);
 			}
-			free(script_root);
 		}
 	}
 
@@ -981,8 +999,7 @@ static char *get_script_path(const char *script_root, const char *suffix)
 	char script_path[MAXPATHLEN];
 	DIR *scripts_dir, *lang_dir;
 	char lang_path[MAXPATHLEN];
-	char *str, *__script_root;
-	char *path = NULL;
+	char *__script_root;
 
 	snprintf(scripts_path, MAXPATHLEN, "%s/scripts", perf_exec_path());
 
@@ -998,23 +1015,18 @@ static char *get_script_path(const char *script_root, const char *suffix)
 			continue;
 
 		for_each_script(lang_path, lang_dir, script_dirent, script_next) {
-			__script_root = strdup(script_dirent.d_name);
-			str = (char *)ends_with(__script_root, suffix);
-			if (str) {
-				*str = '\0';
-				if (strcmp(__script_root, script_root))
-					continue;
+			__script_root = get_script_root(&script_dirent, suffix);
+			if (__script_root && !strcmp(script_root, __script_root)) {
+				free(__script_root);
 				snprintf(script_path, MAXPATHLEN, "%s/%s",
 					 lang_path, script_dirent.d_name);
-				path = strdup(script_path);
-				free(__script_root);
-				break;
+				return strdup(script_path);
 			}
 			free(__script_root);
 		}
 	}
 
-	return path;
+	return NULL;
 }
 
 static bool is_top_script(const char *script_path)
@@ -1083,7 +1095,11 @@ static const struct option options[] = {
 	OPT_CALLBACK('f', "fields", NULL, "str",
 		     "comma separated output fields prepend with 'type:'. Valid types: hw,sw,trace,raw. Fields: comm,tid,pid,time,cpu,event,trace,ip,sym,dso,addr",
 		     parse_output_fields),
-	OPT_STRING('c', "cpu", &cpu_list, "cpu", "list of cpus to profile"),
+	OPT_BOOLEAN('a', "all-cpus", &system_wide,
+		     "system-wide collection from all CPUs"),
+	OPT_STRING('C', "cpu", &cpu_list, "cpu", "list of cpus to profile"),
+	OPT_STRING('c', "comms", &symbol_conf.comm_list_str, "comm[,comm...]",
+		   "only display events for these comms"),
 	OPT_BOOLEAN('I', "show-info", &show_full_info,
 		    "display extended information from perf.data file"),
 	OPT_END()
@@ -1110,7 +1126,6 @@ int cmd_script(int argc, const char **argv, const char *prefix __used)
 	struct perf_session *session;
 	char *script_path = NULL;
 	const char **__argv;
-	bool system_wide;
 	int i, j, err;
 
 	setup_scripting();
@@ -1178,15 +1193,17 @@ int cmd_script(int argc, const char **argv, const char *prefix __used)
 		}
 
 		if (!pid) {
-			system_wide = true;
 			j = 0;
 
 			dup2(live_pipe[1], 1);
 			close(live_pipe[0]);
 
-			if (!is_top_script(argv[0]))
+			if (is_top_script(argv[0])) {
+				system_wide = true;
+			} else if (!system_wide) {
 				system_wide = !have_cmd(argc - rep_args,
 							&argv[rep_args]);
+			}
 
 			__argv = malloc((argc + 6) * sizeof(const char *));
 			if (!__argv)
@@ -1234,10 +1251,11 @@ int cmd_script(int argc, const char **argv, const char *prefix __used)
 		script_path = rep_script_path;
 
 	if (script_path) {
-		system_wide = false;
 		j = 0;
 
-		if (rec_script_path)
+		if (!rec_script_path)
+			system_wide = false;
+		else if (!system_wide)
 			system_wide = !have_cmd(argc - 1, &argv[1]);
 
 		__argv = malloc((argc + 2) * sizeof(const char *));
@@ -1261,7 +1279,7 @@ int cmd_script(int argc, const char **argv, const char *prefix __used)
 	if (!script_name)
 		setup_pager();
 
-	session = perf_session__new(input_name, O_RDONLY, 0, false, &event_ops);
+	session = perf_session__new(input_name, O_RDONLY, 0, false, &perf_script);
 	if (session == NULL)
 		return -ENOMEM;
 
@@ -1287,7 +1305,7 @@ int cmd_script(int argc, const char **argv, const char *prefix __used)
 			return -1;
 		}
 
-		input = open(input_name, O_RDONLY);
+		input = open(session->filename, O_RDONLY);	/* input_name */
 		if (input < 0) {
 			perror("failed to open file");
 			exit(-1);

tools/perf/builtin-stat.c

@@ -578,6 +578,33 @@ static void nsec_printout(int cpu, struct perf_evsel *evsel, double avg)
 			avg / avg_stats(&walltime_nsecs_stats));
 }
 
+/* used for get_ratio_color() */
+enum grc_type {
+	GRC_STALLED_CYCLES_FE,
+	GRC_STALLED_CYCLES_BE,
+	GRC_CACHE_MISSES,
+	GRC_MAX_NR
+};
+
+static const char *get_ratio_color(enum grc_type type, double ratio)
+{
+	static const double grc_table[GRC_MAX_NR][3] = {
+		[GRC_STALLED_CYCLES_FE] = { 50.0, 30.0, 10.0 },
+		[GRC_STALLED_CYCLES_BE] = { 75.0, 50.0, 20.0 },
+		[GRC_CACHE_MISSES] 	= { 20.0, 10.0, 5.0 },
+	};
+	const char *color = PERF_COLOR_NORMAL;
+
+	if (ratio > grc_table[type][0])
+		color = PERF_COLOR_RED;
+	else if (ratio > grc_table[type][1])
+		color = PERF_COLOR_MAGENTA;
+	else if (ratio > grc_table[type][2])
+		color = PERF_COLOR_YELLOW;
+
+	return color;
+}
+
 static void print_stalled_cycles_frontend(int cpu, struct perf_evsel *evsel __used, double avg)
 {
 	double total, ratio = 0.0;
@@ -588,13 +615,7 @@ static void print_stalled_cycles_frontend(int cpu, struct perf_evsel *evsel __us
 	if (total)
 		ratio = avg / total * 100.0;
 
-	color = PERF_COLOR_NORMAL;
-	if (ratio > 50.0)
-		color = PERF_COLOR_RED;
-	else if (ratio > 30.0)
-		color = PERF_COLOR_MAGENTA;
-	else if (ratio > 10.0)
-		color = PERF_COLOR_YELLOW;
+	color = get_ratio_color(GRC_STALLED_CYCLES_FE, ratio);
 
 	fprintf(output, " #  ");
 	color_fprintf(output, color, "%6.2f%%", ratio);
@@ -611,13 +632,7 @@ static void print_stalled_cycles_backend(int cpu, struct perf_evsel *evsel __use
 	if (total)
 		ratio = avg / total * 100.0;
 
-	color = PERF_COLOR_NORMAL;
-	if (ratio > 75.0)
-		color = PERF_COLOR_RED;
-	else if (ratio > 50.0)
-		color = PERF_COLOR_MAGENTA;
-	else if (ratio > 20.0)
-		color = PERF_COLOR_YELLOW;
+	color = get_ratio_color(GRC_STALLED_CYCLES_BE, ratio);
 
 	fprintf(output, " #  ");
 	color_fprintf(output, color, "%6.2f%%", ratio);
@@ -634,13 +649,7 @@ static void print_branch_misses(int cpu, struct perf_evsel *evsel __used, double
 	if (total)
 		ratio = avg / total * 100.0;
 
-	color = PERF_COLOR_NORMAL;
-	if (ratio > 20.0)
-		color = PERF_COLOR_RED;
-	else if (ratio > 10.0)
-		color = PERF_COLOR_MAGENTA;
-	else if (ratio > 5.0)
-		color = PERF_COLOR_YELLOW;
+	color = get_ratio_color(GRC_CACHE_MISSES, ratio);
 
 	fprintf(output, " #  ");
 	color_fprintf(output, color, "%6.2f%%", ratio);
@@ -657,13 +666,7 @@ static void print_l1_dcache_misses(int cpu, struct perf_evsel *evsel __used, dou
 	if (total)
 		ratio = avg / total * 100.0;
 
-	color = PERF_COLOR_NORMAL;
-	if (ratio > 20.0)
-		color = PERF_COLOR_RED;
-	else if (ratio > 10.0)
-		color = PERF_COLOR_MAGENTA;
-	else if (ratio > 5.0)
-		color = PERF_COLOR_YELLOW;
+	color = get_ratio_color(GRC_CACHE_MISSES, ratio);
 
 	fprintf(output, " #  ");
 	color_fprintf(output, color, "%6.2f%%", ratio);
@@ -680,13 +683,7 @@ static void print_l1_icache_misses(int cpu, struct perf_evsel *evsel __used, dou
 	if (total)
 		ratio = avg / total * 100.0;
 
-	color = PERF_COLOR_NORMAL;
-	if (ratio > 20.0)
-		color = PERF_COLOR_RED;
-	else if (ratio > 10.0)
-		color = PERF_COLOR_MAGENTA;
-	else if (ratio > 5.0)
-		color = PERF_COLOR_YELLOW;
+	color = get_ratio_color(GRC_CACHE_MISSES, ratio);
 
 	fprintf(output, " #  ");
 	color_fprintf(output, color, "%6.2f%%", ratio);
@@ -703,13 +700,7 @@ static void print_dtlb_cache_misses(int cpu, struct perf_evsel *evsel __used, do
 	if (total)
 		ratio = avg / total * 100.0;
 
-	color = PERF_COLOR_NORMAL;
-	if (ratio > 20.0)
-		color = PERF_COLOR_RED;
-	else if (ratio > 10.0)
-		color = PERF_COLOR_MAGENTA;
-	else if (ratio > 5.0)
-		color = PERF_COLOR_YELLOW;
+	color = get_ratio_color(GRC_CACHE_MISSES, ratio);
 
 	fprintf(output, " #  ");
 	color_fprintf(output, color, "%6.2f%%", ratio);
@@ -726,13 +717,7 @@ static void print_itlb_cache_misses(int cpu, struct perf_evsel *evsel __used, do
 	if (total)
 		ratio = avg / total * 100.0;
 
-	color = PERF_COLOR_NORMAL;
-	if (ratio > 20.0)
-		color = PERF_COLOR_RED;
-	else if (ratio > 10.0)
-		color = PERF_COLOR_MAGENTA;
-	else if (ratio > 5.0)
-		color = PERF_COLOR_YELLOW;
+	color = get_ratio_color(GRC_CACHE_MISSES, ratio);
 
 	fprintf(output, " #  ");
 	color_fprintf(output, color, "%6.2f%%", ratio);
@@ -749,13 +734,7 @@ static void print_ll_cache_misses(int cpu, struct perf_evsel *evsel __used, doub
 	if (total)
 		ratio = avg / total * 100.0;
 
-	color = PERF_COLOR_NORMAL;
-	if (ratio > 20.0)
-		color = PERF_COLOR_RED;
-	else if (ratio > 10.0)
-		color = PERF_COLOR_MAGENTA;
-	else if (ratio > 5.0)
-		color = PERF_COLOR_YELLOW;
+	color = get_ratio_color(GRC_CACHE_MISSES, ratio);
 
 	fprintf(output, " #  ");
 	color_fprintf(output, color, "%6.2f%%", ratio);
@@ -1108,22 +1087,13 @@ static const struct option options[] = {
  */
 static int add_default_attributes(void)
 {
-	struct perf_evsel *pos;
-	size_t attr_nr = 0;
-	size_t c;
-
 	/* Set attrs if no event is selected and !null_run: */
 	if (null_run)
 		return 0;
 
 	if (!evsel_list->nr_entries) {
-		for (c = 0; c < ARRAY_SIZE(default_attrs); c++) {
-			pos = perf_evsel__new(default_attrs + c, c + attr_nr);
-			if (pos == NULL)
-				return -1;
-			perf_evlist__add(evsel_list, pos);
-		}
-		attr_nr += c;
+		if (perf_evlist__add_attrs_array(evsel_list, default_attrs) < 0)
+			return -1;
 	}
 
 	/* Detailed events get appended to the event list: */
@@ -1132,38 +1102,21 @@ static int add_default_attributes(void)
 		return 0;
 
 	/* Append detailed run extra attributes: */
-	for (c = 0; c < ARRAY_SIZE(detailed_attrs); c++) {
-		pos = perf_evsel__new(detailed_attrs + c, c + attr_nr);
-		if (pos == NULL)
-			return -1;
-		perf_evlist__add(evsel_list, pos);
-	}
-	attr_nr += c;
+	if (perf_evlist__add_attrs_array(evsel_list, detailed_attrs) < 0)
+		return -1;
 
 	if (detailed_run < 2)
 		return 0;
 
 	/* Append very detailed run extra attributes: */
-	for (c = 0; c < ARRAY_SIZE(very_detailed_attrs); c++) {
-		pos = perf_evsel__new(very_detailed_attrs + c, c + attr_nr);
-		if (pos == NULL)
-			return -1;
-		perf_evlist__add(evsel_list, pos);
-	}
+	if (perf_evlist__add_attrs_array(evsel_list, very_detailed_attrs) < 0)
+		return -1;
 
 	if (detailed_run < 3)
 		return 0;
 
 	/* Append very, very detailed run extra attributes: */
-	for (c = 0; c < ARRAY_SIZE(very_very_detailed_attrs); c++) {
-		pos = perf_evsel__new(very_very_detailed_attrs + c, c + attr_nr);
-		if (pos == NULL)
-			return -1;
-		perf_evlist__add(evsel_list, pos);
-	}
-
-
-	return 0;
+	return perf_evlist__add_attrs_array(evsel_list, very_very_detailed_attrs);
 }
 
 int cmd_stat(int argc, const char **argv, const char *prefix __used)
@@ -1267,8 +1220,7 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
 
 	list_for_each_entry(pos, &evsel_list->entries, node) {
 		if (perf_evsel__alloc_stat_priv(pos) < 0 ||
-		    perf_evsel__alloc_counts(pos, evsel_list->cpus->nr) < 0 ||
-		    perf_evsel__alloc_fd(pos, evsel_list->cpus->nr, evsel_list->threads->nr) < 0)
+		    perf_evsel__alloc_counts(pos, evsel_list->cpus->nr) < 0)
 			goto out_free_fd;
 	}
 

tools/perf/builtin-test.c

@@ -7,6 +7,7 @@
 
 #include "util/cache.h"
 #include "util/debug.h"
+#include "util/debugfs.h"
 #include "util/evlist.h"
 #include "util/parse-options.h"
 #include "util/parse-events.h"
@@ -14,8 +15,6 @@
 #include "util/thread_map.h"
 #include "../../include/linux/hw_breakpoint.h"
 
-static long page_size;
-
 static int vmlinux_matches_kallsyms_filter(struct map *map __used, struct symbol *sym)
 {
 	bool *visited = symbol__priv(sym);
@@ -31,6 +30,7 @@ static int test__vmlinux_matches_kallsyms(void)
 	struct map *kallsyms_map, *vmlinux_map;
 	struct machine kallsyms, vmlinux;
 	enum map_type type = MAP__FUNCTION;
+	long page_size = sysconf(_SC_PAGE_SIZE);
 	struct ref_reloc_sym ref_reloc_sym = { .name = "_stext", };
 
 	/*
@@ -247,7 +247,7 @@ static int trace_event__id(const char *evname)
 
 	if (asprintf(&filename,
 		     "%s/syscalls/%s/id",
-		     debugfs_path, evname) < 0)
+		     tracing_events_path, evname) < 0)
 		return -1;
 
 	fd = open(filename, O_RDONLY);
@@ -603,7 +603,7 @@ out_free_threads:
 
 #define TEST_ASSERT_VAL(text, cond) \
 do { \
-	if (!cond) { \
+	if (!(cond)) { \
 		pr_debug("FAILED %s:%d %s\n", __FILE__, __LINE__, text); \
 		return -1; \
 	} \
@@ -759,6 +759,103 @@ static int test__checkevent_breakpoint_w(struct perf_evlist *evlist)
 	return 0;
 }
 
+static int test__checkevent_tracepoint_modifier(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = list_entry(evlist->entries.next,
+					      struct perf_evsel, node);
+
+	TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user);
+	TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel);
+	TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv);
+	TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip);
+
+	return test__checkevent_tracepoint(evlist);
+}
+
+static int
+test__checkevent_tracepoint_multi_modifier(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel;
+
+	TEST_ASSERT_VAL("wrong number of entries", evlist->nr_entries > 1);
+
+	list_for_each_entry(evsel, &evlist->entries, node) {
+		TEST_ASSERT_VAL("wrong exclude_user",
+				!evsel->attr.exclude_user);
+		TEST_ASSERT_VAL("wrong exclude_kernel",
+				evsel->attr.exclude_kernel);
+		TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv);
+		TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip);
+	}
+
+	return test__checkevent_tracepoint_multi(evlist);
+}
+
+static int test__checkevent_raw_modifier(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = list_entry(evlist->entries.next,
+					      struct perf_evsel, node);
+
+	TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user);
+	TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel);
+	TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv);
+	TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip);
+
+	return test__checkevent_raw(evlist);
+}
+
+static int test__checkevent_numeric_modifier(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = list_entry(evlist->entries.next,
+					      struct perf_evsel, node);
+
+	TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user);
+	TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel);
+	TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv);
+	TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip);
+
+	return test__checkevent_numeric(evlist);
+}
+
+static int test__checkevent_symbolic_name_modifier(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = list_entry(evlist->entries.next,
+					      struct perf_evsel, node);
+
+	TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user);
+	TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel);
+	TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv);
+	TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip);
+
+	return test__checkevent_symbolic_name(evlist);
+}
+
+static int test__checkevent_symbolic_alias_modifier(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = list_entry(evlist->entries.next,
+					      struct perf_evsel, node);
+
+	TEST_ASSERT_VAL("wrong exclude_user", !evsel->attr.exclude_user);
+	TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel);
+	TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv);
+	TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip);
+
+	return test__checkevent_symbolic_alias(evlist);
+}
+
+static int test__checkevent_genhw_modifier(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = list_entry(evlist->entries.next,
+					      struct perf_evsel, node);
+
+	TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user);
+	TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel);
+	TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv);
+	TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip);
+
+	return test__checkevent_genhw(evlist);
+}
+
 static struct test__event_st {
 	const char *name;
 	__u32 type;
@@ -808,6 +905,34 @@ static struct test__event_st {
 		.name  = "mem:0:w",
 		.check = test__checkevent_breakpoint_w,
 	},
+	{
+		.name  = "syscalls:sys_enter_open:k",
+		.check = test__checkevent_tracepoint_modifier,
+	},
+	{
+		.name  = "syscalls:*:u",
+		.check = test__checkevent_tracepoint_multi_modifier,
+	},
+	{
+		.name  = "r1:kp",
+		.check = test__checkevent_raw_modifier,
+	},
+	{
+		.name  = "1:1:hp",
+		.check = test__checkevent_numeric_modifier,
+	},
+	{
+		.name  = "instructions:h",
+		.check = test__checkevent_symbolic_name_modifier,
+	},
+	{
+		.name  = "faults:u",
+		.check = test__checkevent_symbolic_alias_modifier,
+	},
+	{
+		.name  = "L1-dcache-load-miss:kp",
+		.check = test__checkevent_genhw_modifier,
+	},
 };
 
 #define TEST__EVENTS_CNT (sizeof(test__events) / sizeof(struct test__event_st))
@@ -841,6 +966,336 @@ static int test__parse_events(void)
 
 	return ret;
 }
+
+static int sched__get_first_possible_cpu(pid_t pid, cpu_set_t **maskp,
+					 size_t *sizep)
+{
+	cpu_set_t *mask;
+	size_t size;
+	int i, cpu = -1, nrcpus = 1024;
+realloc:
+	mask = CPU_ALLOC(nrcpus);
+	size = CPU_ALLOC_SIZE(nrcpus);
+	CPU_ZERO_S(size, mask);
+
+	if (sched_getaffinity(pid, size, mask) == -1) {
+		CPU_FREE(mask);
+		if (errno == EINVAL && nrcpus < (1024 << 8)) {
+			nrcpus = nrcpus << 2;
+			goto realloc;
+		}
+		perror("sched_getaffinity");
+			return -1;
+	}
+
+	for (i = 0; i < nrcpus; i++) {
+		if (CPU_ISSET_S(i, size, mask)) {
+			if (cpu == -1) {
+				cpu = i;
+				*maskp = mask;
+				*sizep = size;
+			} else
+				CPU_CLR_S(i, size, mask);
+		}
+	}
+
+	if (cpu == -1)
+		CPU_FREE(mask);
+
+	return cpu;
+}
+
+static int test__PERF_RECORD(void)
+{
+	struct perf_record_opts opts = {
+		.target_pid = -1,
+		.target_tid = -1,
+		.no_delay   = true,
+		.freq	    = 10,
+		.mmap_pages = 256,
+		.sample_id_all_avail = true,
+	};
+	cpu_set_t *cpu_mask = NULL;
+	size_t cpu_mask_size = 0;
+	struct perf_evlist *evlist = perf_evlist__new(NULL, NULL);
+	struct perf_evsel *evsel;
+	struct perf_sample sample;
+	const char *cmd = "sleep";
+	const char *argv[] = { cmd, "1", NULL, };
+	char *bname;
+	u64 sample_type, prev_time = 0;
+	bool found_cmd_mmap = false,
+	     found_libc_mmap = false,
+	     found_vdso_mmap = false,
+	     found_ld_mmap = false;
+	int err = -1, errs = 0, i, wakeups = 0, sample_size;
+	u32 cpu;
+	int total_events = 0, nr_events[PERF_RECORD_MAX] = { 0, };
+
+	if (evlist == NULL || argv == NULL) {
+		pr_debug("Not enough memory to create evlist\n");
+		goto out;
+	}
+
+	/*
+	 * We need at least one evsel in the evlist, use the default
+	 * one: "cycles".
+	 */
+	err = perf_evlist__add_default(evlist);
+	if (err < 0) {
+		pr_debug("Not enough memory to create evsel\n");
+		goto out_delete_evlist;
+	}
+
+	/*
+	 * Create maps of threads and cpus to monitor. In this case
+	 * we start with all threads and cpus (-1, -1) but then in
+	 * perf_evlist__prepare_workload we'll fill in the only thread
+	 * we're monitoring, the one forked there.
+	 */
+	err = perf_evlist__create_maps(evlist, opts.target_pid,
+				       opts.target_tid, opts.cpu_list);
+	if (err < 0) {
+		pr_debug("Not enough memory to create thread/cpu maps\n");
+		goto out_delete_evlist;
+	}
+
+	/*
+	 * Prepare the workload in argv[] to run, it'll fork it, and then wait
+	 * for perf_evlist__start_workload() to exec it. This is done this way
+	 * so that we have time to open the evlist (calling sys_perf_event_open
+	 * on all the fds) and then mmap them.
+	 */
+	err = perf_evlist__prepare_workload(evlist, &opts, argv);
+	if (err < 0) {
+		pr_debug("Couldn't run the workload!\n");
+		goto out_delete_evlist;
+	}
+
+	/*
+	 * Config the evsels, setting attr->comm on the first one, etc.
+	 */
+	evsel = list_entry(evlist->entries.next, struct perf_evsel, node);
+	evsel->attr.sample_type |= PERF_SAMPLE_CPU;
+	evsel->attr.sample_type |= PERF_SAMPLE_TID;
+	evsel->attr.sample_type |= PERF_SAMPLE_TIME;
+	perf_evlist__config_attrs(evlist, &opts);
+
+	err = sched__get_first_possible_cpu(evlist->workload.pid, &cpu_mask,
+					    &cpu_mask_size);
+	if (err < 0) {
+		pr_debug("sched__get_first_possible_cpu: %s\n", strerror(errno));
+		goto out_delete_evlist;
+	}
+
+	cpu = err;
+
+	/*
+	 * So that we can check perf_sample.cpu on all the samples.
+	 */
+	if (sched_setaffinity(evlist->workload.pid, cpu_mask_size, cpu_mask) < 0) {
+		pr_debug("sched_setaffinity: %s\n", strerror(errno));
+		goto out_free_cpu_mask;
+	}
+
+	/*
+	 * Call sys_perf_event_open on all the fds on all the evsels,
+	 * grouping them if asked to.
+	 */
+	err = perf_evlist__open(evlist, opts.group);
+	if (err < 0) {
+		pr_debug("perf_evlist__open: %s\n", strerror(errno));
+		goto out_delete_evlist;
+	}
+
+	/*
+	 * mmap the first fd on a given CPU and ask for events for the other
+	 * fds in the same CPU to be injected in the same mmap ring buffer
+	 * (using ioctl(PERF_EVENT_IOC_SET_OUTPUT)).
+	 */
+	err = perf_evlist__mmap(evlist, opts.mmap_pages, false);
+	if (err < 0) {
+		pr_debug("perf_evlist__mmap: %s\n", strerror(errno));
+		goto out_delete_evlist;
+	}
+
+	/*
+	 * We'll need these two to parse the PERF_SAMPLE_* fields in each
+	 * event.
+	 */
+	sample_type = perf_evlist__sample_type(evlist);
+	sample_size = __perf_evsel__sample_size(sample_type);
+
+	/*
+	 * Now that all is properly set up, enable the events, they will
+	 * count just on workload.pid, which will start...
+	 */
+	perf_evlist__enable(evlist);
+
+	/*
+	 * Now!
+	 */
+	perf_evlist__start_workload(evlist);
+
+	while (1) {
+		int before = total_events;
+
+		for (i = 0; i < evlist->nr_mmaps; i++) {
+			union perf_event *event;
+
+			while ((event = perf_evlist__mmap_read(evlist, i)) != NULL) {
+				const u32 type = event->header.type;
+				const char *name = perf_event__name(type);
+
+				++total_events;
+				if (type < PERF_RECORD_MAX)
+					nr_events[type]++;
+
+				err = perf_event__parse_sample(event, sample_type,
+							       sample_size, true,
+							       &sample, false);
+				if (err < 0) {
+					if (verbose)
+						perf_event__fprintf(event, stderr);
+					pr_debug("Couldn't parse sample\n");
+					goto out_err;
+				}
+
+				if (verbose) {
+					pr_info("%" PRIu64" %d ", sample.time, sample.cpu);
+					perf_event__fprintf(event, stderr);
+				}
+
+				if (prev_time > sample.time) {
+					pr_debug("%s going backwards in time, prev=%" PRIu64 ", curr=%" PRIu64 "\n",
+						 name, prev_time, sample.time);
+					++errs;
+				}
+
+				prev_time = sample.time;
+
+				if (sample.cpu != cpu) {
+					pr_debug("%s with unexpected cpu, expected %d, got %d\n",
+						 name, cpu, sample.cpu);
+					++errs;
+				}
+
+				if ((pid_t)sample.pid != evlist->workload.pid) {
+					pr_debug("%s with unexpected pid, expected %d, got %d\n",
+						 name, evlist->workload.pid, sample.pid);
+					++errs;
+				}
+
+				if ((pid_t)sample.tid != evlist->workload.pid) {
+					pr_debug("%s with unexpected tid, expected %d, got %d\n",
+						 name, evlist->workload.pid, sample.tid);
+					++errs;
+				}
+
+				if ((type == PERF_RECORD_COMM ||
+				     type == PERF_RECORD_MMAP ||
+				     type == PERF_RECORD_FORK ||
+				     type == PERF_RECORD_EXIT) &&
+				     (pid_t)event->comm.pid != evlist->workload.pid) {
+					pr_debug("%s with unexpected pid/tid\n", name);
+					++errs;
+				}
+
+				if ((type == PERF_RECORD_COMM ||
+				     type == PERF_RECORD_MMAP) &&
+				     event->comm.pid != event->comm.tid) {
+					pr_debug("%s with different pid/tid!\n", name);
+					++errs;
+				}
+
+				switch (type) {
+				case PERF_RECORD_COMM:
+					if (strcmp(event->comm.comm, cmd)) {
+						pr_debug("%s with unexpected comm!\n", name);
+						++errs;
+					}
+					break;
+				case PERF_RECORD_EXIT:
+					goto found_exit;
+				case PERF_RECORD_MMAP:
+					bname = strrchr(event->mmap.filename, '/');
+					if (bname != NULL) {
+						if (!found_cmd_mmap)
+							found_cmd_mmap = !strcmp(bname + 1, cmd);
+						if (!found_libc_mmap)
+							found_libc_mmap = !strncmp(bname + 1, "libc", 4);
+						if (!found_ld_mmap)
+							found_ld_mmap = !strncmp(bname + 1, "ld", 2);
+					} else if (!found_vdso_mmap)
+						found_vdso_mmap = !strcmp(event->mmap.filename, "[vdso]");
+					break;
+
+				case PERF_RECORD_SAMPLE:
+					/* Just ignore samples for now */
+					break;
+				default:
+					pr_debug("Unexpected perf_event->header.type %d!\n",
+						 type);
+					++errs;
+				}
+			}
+		}
+
+		/*
+		 * We don't use poll here because at least at 3.1 times the
+		 * PERF_RECORD_{!SAMPLE} events don't honour
+		 * perf_event_attr.wakeup_events, just PERF_EVENT_SAMPLE does.
+		 */
+		if (total_events == before && false)
+			poll(evlist->pollfd, evlist->nr_fds, -1);
+
+		sleep(1);
+		if (++wakeups > 5) {
+			pr_debug("No PERF_RECORD_EXIT event!\n");
+			break;
+		}
+	}
+
+found_exit:
+	if (nr_events[PERF_RECORD_COMM] > 1) {
+		pr_debug("Excessive number of PERF_RECORD_COMM events!\n");
+		++errs;
+	}
+
+	if (nr_events[PERF_RECORD_COMM] == 0) {
+		pr_debug("Missing PERF_RECORD_COMM for %s!\n", cmd);
+		++errs;
+	}
+
+	if (!found_cmd_mmap) {
+		pr_debug("PERF_RECORD_MMAP for %s missing!\n", cmd);
+		++errs;
+	}
+
+	if (!found_libc_mmap) {
+		pr_debug("PERF_RECORD_MMAP for %s missing!\n", "libc");
+		++errs;
+	}
+
+	if (!found_ld_mmap) {
+		pr_debug("PERF_RECORD_MMAP for %s missing!\n", "ld");
+		++errs;
+	}
+
+	if (!found_vdso_mmap) {
+		pr_debug("PERF_RECORD_MMAP for %s missing!\n", "[vdso]");
+		++errs;
+	}
+out_err:
+	perf_evlist__munmap(evlist);
+out_free_cpu_mask:
+	CPU_FREE(cpu_mask);
+out_delete_evlist:
+	perf_evlist__delete(evlist);
+out:
+	return (err < 0 || errs > 0) ? -1 : 0;
+}
+
 static struct test {
 	const char *desc;
 	int (*func)(void);
@@ -865,46 +1320,90 @@ static struct test {
 		.desc = "parse events tests",
 		.func = test__parse_events,
 	},
+	{
+		.desc = "Validate PERF_RECORD_* events & perf_sample fields",
+		.func = test__PERF_RECORD,
+	},
 	{
 		.func = NULL,
 	},
 };
 
-static int __cmd_test(void)
+static bool perf_test__matches(int curr, int argc, const char *argv[])
 {
-	int i = 0;
+	int i;
+
+	if (argc == 0)
+		return true;
 
-	page_size = sysconf(_SC_PAGE_SIZE);
+	for (i = 0; i < argc; ++i) {
+		char *end;
+		long nr = strtoul(argv[i], &end, 10);
+
+		if (*end == '\0') {
+			if (nr == curr + 1)
+				return true;
+			continue;
+		}
+
+		if (strstr(tests[curr].desc, argv[i]))
+			return true;
+	}
+
+	return false;
+}
+
+static int __cmd_test(int argc, const char *argv[])
+{
+	int i = 0;
 
 	while (tests[i].func) {
-		int err;
-		pr_info("%2d: %s:", i + 1, tests[i].desc);
+		int curr = i++, err;
+
+		if (!perf_test__matches(curr, argc, argv))
+			continue;
+
+		pr_info("%2d: %s:", i, tests[curr].desc);
 		pr_debug("\n--- start ---\n");
-		err = tests[i].func();
-		pr_debug("---- end ----\n%s:", tests[i].desc);
+		err = tests[curr].func();
+		pr_debug("---- end ----\n%s:", tests[curr].desc);
 		pr_info(" %s\n", err ? "FAILED!\n" : "Ok");
-		++i;
 	}
 
 	return 0;
 }
 
-static const char * const test_usage[] = {
-	"perf test [<options>]",
-	NULL,
-};
+static int perf_test__list(int argc, const char **argv)
+{
+	int i = 0;
+
+	while (tests[i].func) {
+		int curr = i++;
 
-static const struct option test_options[] = {
+		if (argc > 1 && !strstr(tests[curr].desc, argv[1]))
+			continue;
+
+		pr_info("%2d: %s\n", i, tests[curr].desc);
+	}
+
+	return 0;
+}
+
+int cmd_test(int argc, const char **argv, const char *prefix __used)
+{
+	const char * const test_usage[] = {
+	"perf test [<options>] [{list <test-name-fragment>|[<test-name-fragments>|<test-numbers>]}]",
+	NULL,
+	};
+	const struct option test_options[] = {
 	OPT_INTEGER('v', "verbose", &verbose,
 		    "be more verbose (show symbol address, etc)"),
 	OPT_END()
-};
+	};
 
-int cmd_test(int argc, const char **argv, const char *prefix __used)
-{
 	argc = parse_options(argc, argv, test_options, test_usage, 0);
-	if (argc)
-		usage_with_options(test_usage, test_options);
+	if (argc >= 1 && !strcmp(argv[0], "list"))
+		return perf_test__list(argc, argv);
 
 	symbol_conf.priv_size = sizeof(int);
 	symbol_conf.sort_by_name = true;
@@ -915,5 +1414,5 @@ int cmd_test(int argc, const char **argv, const char *prefix __used)
 
 	setup_pager();
 
-	return __cmd_test();
+	return __cmd_test(argc, argv);
 }

tools/perf/builtin-timechart.c

@@ -19,6 +19,7 @@
 #include "util/color.h"
 #include <linux/list.h>
 #include "util/cache.h"
+#include "util/evsel.h"
 #include <linux/rbtree.h>
 #include "util/symbol.h"
 #include "util/callchain.h"
@@ -31,13 +32,14 @@
 #include "util/event.h"
 #include "util/session.h"
 #include "util/svghelper.h"
+#include "util/tool.h"
 
 #define SUPPORT_OLD_POWER_EVENTS 1
 #define PWR_EVENT_EXIT -1
 
 
-static char		const *input_name = "perf.data";
-static char		const *output_name = "output.svg";
+static const char	*input_name;
+static const char	*output_name = "output.svg";
 
 static unsigned int	numcpus;
 static u64		min_freq;	/* Lowest CPU frequency seen */
@@ -273,25 +275,28 @@ static int cpus_cstate_state[MAX_CPUS];
 static u64 cpus_pstate_start_times[MAX_CPUS];
 static u64 cpus_pstate_state[MAX_CPUS];
 
-static int process_comm_event(union perf_event *event,
+static int process_comm_event(struct perf_tool *tool __used,
+			      union perf_event *event,
 			      struct perf_sample *sample __used,
-			      struct perf_session *session __used)
+			      struct machine *machine __used)
 {
 	pid_set_comm(event->comm.tid, event->comm.comm);
 	return 0;
 }
 
-static int process_fork_event(union perf_event *event,
+static int process_fork_event(struct perf_tool *tool __used,
+			      union perf_event *event,
 			      struct perf_sample *sample __used,
-			      struct perf_session *session __used)
+			      struct machine *machine __used)
 {
 	pid_fork(event->fork.pid, event->fork.ppid, event->fork.time);
 	return 0;
 }
 
-static int process_exit_event(union perf_event *event,
+static int process_exit_event(struct perf_tool *tool __used,
+			      union perf_event *event,
 			      struct perf_sample *sample __used,
-			      struct perf_session *session __used)
+			      struct machine *machine __used)
 {
 	pid_exit(event->fork.pid, event->fork.time);
 	return 0;
@@ -486,14 +491,15 @@ static void sched_switch(int cpu, u64 timestamp, struct trace_entry *te)
 }
 
 
-static int process_sample_event(union perf_event *event __used,
+static int process_sample_event(struct perf_tool *tool __used,
+				union perf_event *event __used,
 				struct perf_sample *sample,
-				struct perf_evsel *evsel __used,
-				struct perf_session *session)
+				struct perf_evsel *evsel,
+				struct machine *machine __used)
 {
 	struct trace_entry *te;
 
-	if (session->sample_type & PERF_SAMPLE_TIME) {
+	if (evsel->attr.sample_type & PERF_SAMPLE_TIME) {
 		if (!first_time || first_time > sample->time)
 			first_time = sample->time;
 		if (last_time < sample->time)
@@ -501,7 +507,7 @@ static int process_sample_event(union perf_event *event __used,
 	}
 
 	te = (void *)sample->raw_data;
-	if (session->sample_type & PERF_SAMPLE_RAW && sample->raw_size > 0) {
+	if ((evsel->attr.sample_type & PERF_SAMPLE_RAW) && sample->raw_size > 0) {
 		char *event_str;
 #ifdef SUPPORT_OLD_POWER_EVENTS
 		struct power_entry_old *peo;
@@ -974,7 +980,7 @@ static void write_svg_file(const char *filename)
 	svg_close();
 }
 
-static struct perf_event_ops event_ops = {
+static struct perf_tool perf_timechart = {
 	.comm			= process_comm_event,
 	.fork			= process_fork_event,
 	.exit			= process_exit_event,
@@ -985,7 +991,7 @@ static struct perf_event_ops event_ops = {
 static int __cmd_timechart(void)
 {
 	struct perf_session *session = perf_session__new(input_name, O_RDONLY,
-							 0, false, &event_ops);
+							 0, false, &perf_timechart);
 	int ret = -EINVAL;
 
 	if (session == NULL)
@@ -994,7 +1000,7 @@ static int __cmd_timechart(void)
 	if (!perf_session__has_traces(session, "timechart record"))
 		goto out_delete;
 
-	ret = perf_session__process_events(session, &event_ops);
+	ret = perf_session__process_events(session, &perf_timechart);
 	if (ret)
 		goto out_delete;
 

tools/perf/perf.c

@@ -29,8 +29,6 @@ struct pager_config {
 	int val;
 };
 
-static char debugfs_mntpt[MAXPATHLEN];
-
 static int pager_command_config(const char *var, const char *value, void *data)
 {
 	struct pager_config *c = data;
@@ -81,15 +79,6 @@ static void commit_pager_choice(void)
 	}
 }
 
-static void set_debugfs_path(void)
-{
-	char *path;
-
-	path = getenv(PERF_DEBUGFS_ENVIRONMENT);
-	snprintf(debugfs_path, MAXPATHLEN, "%s/%s", path ?: debugfs_mntpt,
-		 "tracing/events");
-}
-
 static int handle_options(const char ***argv, int *argc, int *envchanged)
 {
 	int handled = 0;
@@ -161,15 +150,14 @@ static int handle_options(const char ***argv, int *argc, int *envchanged)
 				fprintf(stderr, "No directory given for --debugfs-dir.\n");
 				usage(perf_usage_string);
 			}
-			strncpy(debugfs_mntpt, (*argv)[1], MAXPATHLEN);
-			debugfs_mntpt[MAXPATHLEN - 1] = '\0';
+			debugfs_set_path((*argv)[1]);
 			if (envchanged)
 				*envchanged = 1;
 			(*argv)++;
 			(*argc)--;
 		} else if (!prefixcmp(cmd, CMD_DEBUGFS_DIR)) {
-			strncpy(debugfs_mntpt, cmd + strlen(CMD_DEBUGFS_DIR), MAXPATHLEN);
-			debugfs_mntpt[MAXPATHLEN - 1] = '\0';
+			debugfs_set_path(cmd + strlen(CMD_DEBUGFS_DIR));
+			fprintf(stderr, "dir: %s\n", debugfs_mountpoint);
 			if (envchanged)
 				*envchanged = 1;
 		} else {
@@ -281,7 +269,6 @@ static int run_builtin(struct cmd_struct *p, int argc, const char **argv)
 	if (use_pager == -1 && p->option & USE_PAGER)
 		use_pager = 1;
 	commit_pager_choice();
-	set_debugfs_path();
 
 	status = p->fn(argc, argv, prefix);
 	exit_browser(status);
@@ -416,17 +403,6 @@ static int run_argv(int *argcp, const char ***argv)
 	return done_alias;
 }
 
-/* mini /proc/mounts parser: searching for "^blah /mount/point debugfs" */
-static void get_debugfs_mntpt(void)
-{
-	const char *path = debugfs_mount(NULL);
-
-	if (path)
-		strncpy(debugfs_mntpt, path, sizeof(debugfs_mntpt));
-	else
-		debugfs_mntpt[0] = '\0';
-}
-
 static void pthread__block_sigwinch(void)
 {
 	sigset_t set;
@@ -453,7 +429,7 @@ int main(int argc, const char **argv)
 	if (!cmd)
 		cmd = "perf-help";
 	/* get debugfs mount point from /proc/mounts */
-	get_debugfs_mntpt();
+	debugfs_mount(NULL);
 	/*
 	 * "perf-xxxx" is the same as "perf xxxx", but we obviously:
 	 *
@@ -476,7 +452,6 @@ int main(int argc, const char **argv)
 	argc--;
 	handle_options(&argv, &argc, NULL);
 	commit_pager_choice();
-	set_debugfs_path();
 	set_buildid_dir();
 
 	if (argc > 0) {

tools/perf/perf.h

@@ -185,4 +185,28 @@ extern const char perf_version_string[];
 
 void pthread__unblock_sigwinch(void);
 
+struct perf_record_opts {
+	pid_t	     target_pid;
+	pid_t	     target_tid;
+	bool	     call_graph;
+	bool	     group;
+	bool	     inherit_stat;
+	bool	     no_delay;
+	bool	     no_inherit;
+	bool	     no_samples;
+	bool	     pipe_output;
+	bool	     raw_samples;
+	bool	     sample_address;
+	bool	     sample_time;
+	bool	     sample_id_all_avail;
+	bool	     system_wide;
+	bool	     period;
+	unsigned int freq;
+	unsigned int mmap_pages;
+	unsigned int user_freq;
+	u64	     default_interval;
+	u64	     user_interval;
+	const char   *cpu_list;
+};
+
 #endif

tools/perf/util/annotate.c

@@ -25,17 +25,17 @@ int symbol__annotate_init(struct map *map __used, struct symbol *sym)
 	return 0;
 }
 
-int symbol__alloc_hist(struct symbol *sym, int nevents)
+int symbol__alloc_hist(struct symbol *sym)
 {
 	struct annotation *notes = symbol__annotation(sym);
 	size_t sizeof_sym_hist = (sizeof(struct sym_hist) +
 				  (sym->end - sym->start) * sizeof(u64));
 
-	notes->src = zalloc(sizeof(*notes->src) + nevents * sizeof_sym_hist);
+	notes->src = zalloc(sizeof(*notes->src) + symbol_conf.nr_events * sizeof_sym_hist);
 	if (notes->src == NULL)
 		return -1;
 	notes->src->sizeof_sym_hist = sizeof_sym_hist;
-	notes->src->nr_histograms   = nevents;
+	notes->src->nr_histograms   = symbol_conf.nr_events;
 	INIT_LIST_HEAD(&notes->src->source);
 	return 0;
 }
@@ -334,7 +334,7 @@ fallback:
 		 disassembler_style ? "-M " : "",
 		 disassembler_style ? disassembler_style : "",
 		 map__rip_2objdump(map, sym->start),
-		 map__rip_2objdump(map, sym->end),
+		 map__rip_2objdump(map, sym->end+1),
 		 symbol_conf.annotate_asm_raw ? "" : "--no-show-raw",
 		 symbol_conf.annotate_src ? "-S" : "",
 		 symfs_filename, filename);

tools/perf/util/annotate.h

@@ -72,7 +72,7 @@ static inline struct annotation *symbol__annotation(struct symbol *sym)
 
 int symbol__inc_addr_samples(struct symbol *sym, struct map *map,
 			     int evidx, u64 addr);
-int symbol__alloc_hist(struct symbol *sym, int nevents);
+int symbol__alloc_hist(struct symbol *sym);
 void symbol__annotate_zero_histograms(struct symbol *sym);
 
 int symbol__annotate(struct symbol *sym, struct map *map, size_t privsize);
@@ -99,8 +99,7 @@ static inline int symbol__tui_annotate(struct symbol *sym __used,
 }
 #else
 int symbol__tui_annotate(struct symbol *sym, struct map *map, int evidx,
-			 int nr_events, void(*timer)(void *arg), void *arg,
-			 int delay_secs);
+			 void(*timer)(void *arg), void *arg, int delay_secs);
 #endif
 
 extern const char	*disassembler_style;

tools/perf/util/build-id.c

@@ -13,15 +13,18 @@
 #include "symbol.h"
 #include <linux/kernel.h>
 #include "debug.h"
+#include "session.h"
+#include "tool.h"
 
-static int build_id__mark_dso_hit(union perf_event *event,
+static int build_id__mark_dso_hit(struct perf_tool *tool __used,
+				  union perf_event *event,
 				  struct perf_sample *sample __used,
 				  struct perf_evsel *evsel __used,
-				  struct perf_session *session)
+				  struct machine *machine)
 {
 	struct addr_location al;
 	u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
-	struct thread *thread = perf_session__findnew(session, event->ip.pid);
+	struct thread *thread = machine__findnew_thread(machine, event->ip.pid);
 
 	if (thread == NULL) {
 		pr_err("problem processing %d event, skipping it.\n",
@@ -29,8 +32,8 @@ static int build_id__mark_dso_hit(union perf_event *event,
 		return -1;
 	}
 
-	thread__find_addr_map(thread, session, cpumode, MAP__FUNCTION,
-			      event->ip.pid, event->ip.ip, &al);
+	thread__find_addr_map(thread, machine, cpumode, MAP__FUNCTION,
+			      event->ip.ip, &al);
 
 	if (al.map != NULL)
 		al.map->dso->hit = 1;
@@ -38,25 +41,26 @@ static int build_id__mark_dso_hit(union perf_event *event,
 	return 0;
 }
 
-static int perf_event__exit_del_thread(union perf_event *event,
+static int perf_event__exit_del_thread(struct perf_tool *tool __used,
+				       union perf_event *event,
 				       struct perf_sample *sample __used,
-				       struct perf_session *session)
+				       struct machine *machine)
 {
-	struct thread *thread = perf_session__findnew(session, event->fork.tid);
+	struct thread *thread = machine__findnew_thread(machine, event->fork.tid);
 
 	dump_printf("(%d:%d):(%d:%d)\n", event->fork.pid, event->fork.tid,
 		    event->fork.ppid, event->fork.ptid);
 
 	if (thread) {
-		rb_erase(&thread->rb_node, &session->threads);
-		session->last_match = NULL;
+		rb_erase(&thread->rb_node, &machine->threads);
+		machine->last_match = NULL;
 		thread__delete(thread);
 	}
 
 	return 0;
 }
 
-struct perf_event_ops build_id__mark_dso_hit_ops = {
+struct perf_tool build_id__mark_dso_hit_ops = {
 	.sample	= build_id__mark_dso_hit,
 	.mmap	= perf_event__process_mmap,
 	.fork	= perf_event__process_task,

tools/perf/util/build-id.h

@@ -3,7 +3,7 @@
 
 #include "session.h"
 
-extern struct perf_event_ops build_id__mark_dso_hit_ops;
+extern struct perf_tool build_id__mark_dso_hit_ops;
 
 char *dso__build_id_filename(struct dso *self, char *bf, size_t size);
 

tools/perf/util/callchain.h

@@ -101,6 +101,9 @@ int callchain_append(struct callchain_root *root,
 int callchain_merge(struct callchain_cursor *cursor,
 		    struct callchain_root *dst, struct callchain_root *src);
 
+struct ip_callchain;
+union perf_event;
+
 bool ip_callchain__valid(struct ip_callchain *chain,
 			 const union perf_event *event);
 /*

tools/perf/util/cgroup.c

@@ -3,7 +3,6 @@
 #include "parse-options.h"
 #include "evsel.h"
 #include "cgroup.h"
-#include "debugfs.h" /* MAX_PATH, STR() */
 #include "evlist.h"
 
 int nr_cgroups;
@@ -12,7 +11,7 @@ static int
 cgroupfs_find_mountpoint(char *buf, size_t maxlen)
 {
 	FILE *fp;
-	char mountpoint[MAX_PATH+1], tokens[MAX_PATH+1], type[MAX_PATH+1];
+	char mountpoint[PATH_MAX + 1], tokens[PATH_MAX + 1], type[PATH_MAX + 1];
 	char *token, *saved_ptr = NULL;
 	int found = 0;
 
@@ -25,8 +24,8 @@ cgroupfs_find_mountpoint(char *buf, size_t maxlen)
 	 * and inspect every cgroupfs mount point to find one that has
 	 * perf_event subsystem
 	 */
-	while (fscanf(fp, "%*s %"STR(MAX_PATH)"s %"STR(MAX_PATH)"s %"
-				STR(MAX_PATH)"s %*d %*d\n",
+	while (fscanf(fp, "%*s %"STR(PATH_MAX)"s %"STR(PATH_MAX)"s %"
+				STR(PATH_MAX)"s %*d %*d\n",
 				mountpoint, type, tokens) == 3) {
 
 		if (!strcmp(type, "cgroup")) {
@@ -57,15 +56,15 @@ cgroupfs_find_mountpoint(char *buf, size_t maxlen)
 
 static int open_cgroup(char *name)
 {
-	char path[MAX_PATH+1];
-	char mnt[MAX_PATH+1];
+	char path[PATH_MAX + 1];
+	char mnt[PATH_MAX + 1];
 	int fd;
 
 
-	if (cgroupfs_find_mountpoint(mnt, MAX_PATH+1))
+	if (cgroupfs_find_mountpoint(mnt, PATH_MAX + 1))
 		return -1;
 
-	snprintf(path, MAX_PATH, "%s/%s", mnt, name);
+	snprintf(path, PATH_MAX, "%s/%s", mnt, name);
 
 	fd = open(path, O_RDONLY);
 	if (fd == -1)

tools/perf/util/config.c

@@ -1,5 +1,8 @@
 /*
- * GIT - The information manager from hell
+ * config.c
+ *
+ * Helper functions for parsing config items.
+ * Originally copied from GIT source.
  *
  * Copyright (C) Linus Torvalds, 2005
  * Copyright (C) Johannes Schindelin, 2005

tools/perf/util/debugfs.c

@@ -2,8 +2,12 @@
 #include "debugfs.h"
 #include "cache.h"
 
+#include <linux/kernel.h>
+#include <sys/mount.h>
+
 static int debugfs_premounted;
-static char debugfs_mountpoint[MAX_PATH+1];
+char debugfs_mountpoint[PATH_MAX + 1] = "/sys/kernel/debug";
+char tracing_events_path[PATH_MAX + 1] = "/sys/kernel/debug/tracing/events";
 
 static const char *debugfs_known_mountpoints[] = {
 	"/sys/kernel/debug/",
@@ -62,11 +66,9 @@ const char *debugfs_find_mountpoint(void)
 	/* give up and parse /proc/mounts */
 	fp = fopen("/proc/mounts", "r");
 	if (fp == NULL)
-		die("Can't open /proc/mounts for read");
+		return NULL;
 
-	while (fscanf(fp, "%*s %"
-		      STR(MAX_PATH)
-		      "s %99s %*s %*d %*d\n",
+	while (fscanf(fp, "%*s %" STR(PATH_MAX) "s %99s %*s %*d %*d\n",
 		      debugfs_mountpoint, type) == 2) {
 		if (strcmp(type, "debugfs") == 0)
 			break;
@@ -106,6 +108,12 @@ int debugfs_valid_entry(const char *path)
 	return 0;
 }
 
+static void debugfs_set_tracing_events_path(const char *mountpoint)
+{
+	snprintf(tracing_events_path, sizeof(tracing_events_path), "%s/%s",
+		 mountpoint, "tracing/events");
+}
+
 /* mount the debugfs somewhere if it's not mounted */
 
 char *debugfs_mount(const char *mountpoint)
@@ -113,7 +121,7 @@ char *debugfs_mount(const char *mountpoint)
 	/* see if it's already mounted */
 	if (debugfs_find_mountpoint()) {
 		debugfs_premounted = 1;
-		return debugfs_mountpoint;
+		goto out;
 	}
 
 	/* if not mounted and no argument */
@@ -129,12 +137,19 @@ char *debugfs_mount(const char *mountpoint)
 		return NULL;
 
 	/* save the mountpoint */
-	strncpy(debugfs_mountpoint, mountpoint, sizeof(debugfs_mountpoint));
 	debugfs_found = 1;
-
+	strncpy(debugfs_mountpoint, mountpoint, sizeof(debugfs_mountpoint));
+out:
+	debugfs_set_tracing_events_path(debugfs_mountpoint);
 	return debugfs_mountpoint;
 }
 
+void debugfs_set_path(const char *mountpoint)
+{
+	snprintf(debugfs_mountpoint, sizeof(debugfs_mountpoint), "%s", mountpoint);
+	debugfs_set_tracing_events_path(mountpoint);
+}
+
 /* umount the debugfs */
 
 int debugfs_umount(void)
@@ -158,7 +173,7 @@ int debugfs_umount(void)
 
 int debugfs_write(const char *entry, const char *value)
 {
-	char path[MAX_PATH+1];
+	char path[PATH_MAX + 1];
 	int ret, count;
 	int fd;
 
@@ -203,7 +218,7 @@ int debugfs_write(const char *entry, const char *value)
  */
 int debugfs_read(const char *entry, char *buffer, size_t size)
 {
-	char path[MAX_PATH+1];
+	char path[PATH_MAX + 1];
 	int ret;
 	int fd;
 

tools/perf/util/debugfs.h

@@ -1,25 +1,18 @@
 #ifndef __DEBUGFS_H__
 #define __DEBUGFS_H__
 
-#include <sys/mount.h>
+const char *debugfs_find_mountpoint(void);
+int debugfs_valid_mountpoint(const char *debugfs);
+int debugfs_valid_entry(const char *path);
+char *debugfs_mount(const char *mountpoint);
+int debugfs_umount(void);
+void debugfs_set_path(const char *mountpoint);
+int debugfs_write(const char *entry, const char *value);
+int debugfs_read(const char *entry, char *buffer, size_t size);
+void debugfs_force_cleanup(void);
+int debugfs_make_path(const char *element, char *buffer, int size);
 
-#ifndef MAX_PATH
-# define MAX_PATH 256
-#endif
-
-#ifndef STR
-# define _STR(x) #x
-# define STR(x) _STR(x)
-#endif
-
-extern const char *debugfs_find_mountpoint(void);
-extern int debugfs_valid_mountpoint(const char *debugfs);
-extern int debugfs_valid_entry(const char *path);
-extern char *debugfs_mount(const char *mountpoint);
-extern int debugfs_umount(void);
-extern int debugfs_write(const char *entry, const char *value);
-extern int debugfs_read(const char *entry, char *buffer, size_t size);
-extern void debugfs_force_cleanup(void);
-extern int debugfs_make_path(const char *element, char *buffer, int size);
+extern char debugfs_mountpoint[];
+extern char tracing_events_path[];
 
 #endif /* __DEBUGFS_H__ */

tools/perf/util/event.c

@@ -1,7 +1,6 @@
 #include <linux/types.h>
 #include "event.h"
 #include "debug.h"
-#include "session.h"
 #include "sort.h"
 #include "string.h"
 #include "strlist.h"
@@ -44,36 +43,27 @@ static struct perf_sample synth_sample = {
 	.period	   = 1,
 };
 
-static pid_t perf_event__synthesize_comm(union perf_event *event, pid_t pid,
-					 int full, perf_event__handler_t process,
-					 struct perf_session *session)
+static pid_t perf_event__get_comm_tgid(pid_t pid, char *comm, size_t len)
 {
 	char filename[PATH_MAX];
 	char bf[BUFSIZ];
 	FILE *fp;
 	size_t size = 0;
-	DIR *tasks;
-	struct dirent dirent, *next;
-	pid_t tgid = 0;
+	pid_t tgid = -1;
 
 	snprintf(filename, sizeof(filename), "/proc/%d/status", pid);
 
 	fp = fopen(filename, "r");
 	if (fp == NULL) {
-out_race:
-		/*
-		 * We raced with a task exiting - just return:
-		 */
 		pr_debug("couldn't open %s\n", filename);
 		return 0;
 	}
 
-	memset(&event->comm, 0, sizeof(event->comm));
-
-	while (!event->comm.comm[0] || !event->comm.pid) {
+	while (!comm[0] || (tgid < 0)) {
 		if (fgets(bf, sizeof(bf), fp) == NULL) {
-			pr_warning("couldn't get COMM and pgid, malformed %s\n", filename);
-			goto out;
+			pr_warning("couldn't get COMM and pgid, malformed %s\n",
+				   filename);
+			break;
 		}
 
 		if (memcmp(bf, "Name:", 5) == 0) {
@@ -81,33 +71,65 @@ out_race:
 			while (*name && isspace(*name))
 				++name;
 			size = strlen(name) - 1;
-			memcpy(event->comm.comm, name, size++);
+			if (size >= len)
+				size = len - 1;
+			memcpy(comm, name, size);
+
 		} else if (memcmp(bf, "Tgid:", 5) == 0) {
 			char *tgids = bf + 5;
 			while (*tgids && isspace(*tgids))
 				++tgids;
-			tgid = event->comm.pid = atoi(tgids);
+			tgid = atoi(tgids);
 		}
 	}
 
+	fclose(fp);
+
+	return tgid;
+}
+
+static pid_t perf_event__synthesize_comm(struct perf_tool *tool,
+					 union perf_event *event, pid_t pid,
+					 int full,
+					 perf_event__handler_t process,
+					 struct machine *machine)
+{
+	char filename[PATH_MAX];
+	size_t size;
+	DIR *tasks;
+	struct dirent dirent, *next;
+	pid_t tgid;
+
+	memset(&event->comm, 0, sizeof(event->comm));
+
+	tgid = perf_event__get_comm_tgid(pid, event->comm.comm,
+					 sizeof(event->comm.comm));
+	if (tgid < 0)
+		goto out;
+
+	event->comm.pid = tgid;
 	event->comm.header.type = PERF_RECORD_COMM;
+
+	size = strlen(event->comm.comm) + 1;
 	size = ALIGN(size, sizeof(u64));
-	memset(event->comm.comm + size, 0, session->id_hdr_size);
+	memset(event->comm.comm + size, 0, machine->id_hdr_size);
 	event->comm.header.size = (sizeof(event->comm) -
 				(sizeof(event->comm.comm) - size) +
-				session->id_hdr_size);
+				machine->id_hdr_size);
 	if (!full) {
 		event->comm.tid = pid;
 
-		process(event, &synth_sample, session);
+		process(tool, event, &synth_sample, machine);
 		goto out;
 	}
 
 	snprintf(filename, sizeof(filename), "/proc/%d/task", pid);
 
 	tasks = opendir(filename);
-	if (tasks == NULL)
-		goto out_race;
+	if (tasks == NULL) {
+		pr_debug("couldn't open %s\n", filename);
+		return 0;
+	}
 
 	while (!readdir_r(tasks, &dirent, &next) && next) {
 		char *end;
@@ -115,22 +137,32 @@ out_race:
 		if (*end)
 			continue;
 
+		/* already have tgid; jut want to update the comm */
+		(void) perf_event__get_comm_tgid(pid, event->comm.comm,
+					 sizeof(event->comm.comm));
+
+		size = strlen(event->comm.comm) + 1;
+		size = ALIGN(size, sizeof(u64));
+		memset(event->comm.comm + size, 0, machine->id_hdr_size);
+		event->comm.header.size = (sizeof(event->comm) -
+					  (sizeof(event->comm.comm) - size) +
+					  machine->id_hdr_size);
+
 		event->comm.tid = pid;
 
-		process(event, &synth_sample, session);
+		process(tool, event, &synth_sample, machine);
 	}
 
 	closedir(tasks);
 out:
-	fclose(fp);
-
 	return tgid;
 }
 
-static int perf_event__synthesize_mmap_events(union perf_event *event,
+static int perf_event__synthesize_mmap_events(struct perf_tool *tool,
+					      union perf_event *event,
 					      pid_t pid, pid_t tgid,
 					      perf_event__handler_t process,
-					      struct perf_session *session)
+					      struct machine *machine)
 {
 	char filename[PATH_MAX];
 	FILE *fp;
@@ -193,12 +225,12 @@ static int perf_event__synthesize_mmap_events(union perf_event *event,
 			event->mmap.len -= event->mmap.start;
 			event->mmap.header.size = (sizeof(event->mmap) -
 					        (sizeof(event->mmap.filename) - size));
-			memset(event->mmap.filename + size, 0, session->id_hdr_size);
-			event->mmap.header.size += session->id_hdr_size;
+			memset(event->mmap.filename + size, 0, machine->id_hdr_size);
+			event->mmap.header.size += machine->id_hdr_size;
 			event->mmap.pid = tgid;
 			event->mmap.tid = pid;
 
-			process(event, &synth_sample, session);
+			process(tool, event, &synth_sample, machine);
 		}
 	}
 
@@ -206,14 +238,14 @@ static int perf_event__synthesize_mmap_events(union perf_event *event,
 	return 0;
 }
 
-int perf_event__synthesize_modules(perf_event__handler_t process,
-				   struct perf_session *session,
+int perf_event__synthesize_modules(struct perf_tool *tool,
+				   perf_event__handler_t process,
 				   struct machine *machine)
 {
 	struct rb_node *nd;
 	struct map_groups *kmaps = &machine->kmaps;
 	union perf_event *event = zalloc((sizeof(event->mmap) +
-					  session->id_hdr_size));
+					  machine->id_hdr_size));
 	if (event == NULL) {
 		pr_debug("Not enough memory synthesizing mmap event "
 			 "for kernel modules\n");
@@ -243,15 +275,15 @@ int perf_event__synthesize_modules(perf_event__handler_t process,
 		event->mmap.header.type = PERF_RECORD_MMAP;
 		event->mmap.header.size = (sizeof(event->mmap) -
 				        (sizeof(event->mmap.filename) - size));
-		memset(event->mmap.filename + size, 0, session->id_hdr_size);
-		event->mmap.header.size += session->id_hdr_size;
+		memset(event->mmap.filename + size, 0, machine->id_hdr_size);
+		event->mmap.header.size += machine->id_hdr_size;
 		event->mmap.start = pos->start;
 		event->mmap.len   = pos->end - pos->start;
 		event->mmap.pid   = machine->pid;
 
 		memcpy(event->mmap.filename, pos->dso->long_name,
 		       pos->dso->long_name_len + 1);
-		process(event, &synth_sample, session);
+		process(tool, event, &synth_sample, machine);
 	}
 
 	free(event);
@@ -260,40 +292,69 @@ int perf_event__synthesize_modules(perf_event__handler_t process,
 
 static int __event__synthesize_thread(union perf_event *comm_event,
 				      union perf_event *mmap_event,
-				      pid_t pid, perf_event__handler_t process,
-				      struct perf_session *session)
+				      pid_t pid, int full,
+					  perf_event__handler_t process,
+				      struct perf_tool *tool,
+				      struct machine *machine)
 {
-	pid_t tgid = perf_event__synthesize_comm(comm_event, pid, 1, process,
-					    session);
+	pid_t tgid = perf_event__synthesize_comm(tool, comm_event, pid, full,
+						 process, machine);
 	if (tgid == -1)
 		return -1;
-	return perf_event__synthesize_mmap_events(mmap_event, pid, tgid,
-					     process, session);
+	return perf_event__synthesize_mmap_events(tool, mmap_event, pid, tgid,
+						  process, machine);
 }
 
-int perf_event__synthesize_thread_map(struct thread_map *threads,
+int perf_event__synthesize_thread_map(struct perf_tool *tool,
+				      struct thread_map *threads,
 				      perf_event__handler_t process,
-				      struct perf_session *session)
+				      struct machine *machine)
 {
 	union perf_event *comm_event, *mmap_event;
-	int err = -1, thread;
+	int err = -1, thread, j;
 
-	comm_event = malloc(sizeof(comm_event->comm) + session->id_hdr_size);
+	comm_event = malloc(sizeof(comm_event->comm) + machine->id_hdr_size);
 	if (comm_event == NULL)
 		goto out;
 
-	mmap_event = malloc(sizeof(mmap_event->mmap) + session->id_hdr_size);
+	mmap_event = malloc(sizeof(mmap_event->mmap) + machine->id_hdr_size);
 	if (mmap_event == NULL)
 		goto out_free_comm;
 
 	err = 0;
 	for (thread = 0; thread < threads->nr; ++thread) {
 		if (__event__synthesize_thread(comm_event, mmap_event,
-					       threads->map[thread],
-					       process, session)) {
+					       threads->map[thread], 0,
+					       process, tool, machine)) {
 			err = -1;
 			break;
 		}
+
+		/*
+		 * comm.pid is set to thread group id by
+		 * perf_event__synthesize_comm
+		 */
+		if ((int) comm_event->comm.pid != threads->map[thread]) {
+			bool need_leader = true;
+
+			/* is thread group leader in thread_map? */
+			for (j = 0; j < threads->nr; ++j) {
+				if ((int) comm_event->comm.pid == threads->map[j]) {
+					need_leader = false;
+					break;
+				}
+			}
+
+			/* if not, generate events for it */
+			if (need_leader &&
+			    __event__synthesize_thread(comm_event,
+						      mmap_event,
+						      comm_event->comm.pid, 0,
+						      process, tool, machine)) {
+				err = -1;
+				break;
+			}
+		}
 	}
 	free(mmap_event);
 out_free_comm:
@@ -302,19 +363,20 @@ out:
 	return err;
 }
 
-int perf_event__synthesize_threads(perf_event__handler_t process,
-				   struct perf_session *session)
+int perf_event__synthesize_threads(struct perf_tool *tool,
+				   perf_event__handler_t process,
+				   struct machine *machine)
 {
 	DIR *proc;
 	struct dirent dirent, *next;
 	union perf_event *comm_event, *mmap_event;
 	int err = -1;
 
-	comm_event = malloc(sizeof(comm_event->comm) + session->id_hdr_size);
+	comm_event = malloc(sizeof(comm_event->comm) + machine->id_hdr_size);
 	if (comm_event == NULL)
 		goto out;
 
-	mmap_event = malloc(sizeof(mmap_event->mmap) + session->id_hdr_size);
+	mmap_event = malloc(sizeof(mmap_event->mmap) + machine->id_hdr_size);
 	if (mmap_event == NULL)
 		goto out_free_comm;
 
@@ -329,8 +391,8 @@ int perf_event__synthesize_threads(perf_event__handler_t process,
 		if (*end) /* only interested in proper numerical dirents */
 			continue;
 
-		__event__synthesize_thread(comm_event, mmap_event, pid,
-					   process, session);
+		__event__synthesize_thread(comm_event, mmap_event, pid, 1,
+					   process, tool, machine);
 	}
 
 	closedir(proc);
@@ -365,8 +427,8 @@ static int find_symbol_cb(void *arg, const char *name, char type,
 	return 1;
 }
 
-int perf_event__synthesize_kernel_mmap(perf_event__handler_t process,
-				       struct perf_session *session,
+int perf_event__synthesize_kernel_mmap(struct perf_tool *tool,
+				       perf_event__handler_t process,
 				       struct machine *machine,
 				       const char *symbol_name)
 {
@@ -383,7 +445,7 @@ int perf_event__synthesize_kernel_mmap(perf_event__handler_t process,
 	 */
 	struct process_symbol_args args = { .name = symbol_name, };
 	union perf_event *event = zalloc((sizeof(event->mmap) +
-					  session->id_hdr_size));
+					  machine->id_hdr_size));
 	if (event == NULL) {
 		pr_debug("Not enough memory synthesizing mmap event "
 			 "for kernel modules\n");
@@ -417,25 +479,32 @@ int perf_event__synthesize_kernel_mmap(perf_event__handler_t process,
 	size = ALIGN(size, sizeof(u64));
 	event->mmap.header.type = PERF_RECORD_MMAP;
 	event->mmap.header.size = (sizeof(event->mmap) -
-			(sizeof(event->mmap.filename) - size) + session->id_hdr_size);
+			(sizeof(event->mmap.filename) - size) + machine->id_hdr_size);
 	event->mmap.pgoff = args.start;
 	event->mmap.start = map->start;
 	event->mmap.len   = map->end - event->mmap.start;
 	event->mmap.pid   = machine->pid;
 
-	err = process(event, &synth_sample, session);
+	err = process(tool, event, &synth_sample, machine);
 	free(event);
 
 	return err;
 }
 
-int perf_event__process_comm(union perf_event *event,
+size_t perf_event__fprintf_comm(union perf_event *event, FILE *fp)
+{
+	return fprintf(fp, ": %s:%d\n", event->comm.comm, event->comm.tid);
+}
+
+int perf_event__process_comm(struct perf_tool *tool __used,
+			     union perf_event *event,
 			     struct perf_sample *sample __used,
-			     struct perf_session *session)
+			     struct machine *machine)
 {
-	struct thread *thread = perf_session__findnew(session, event->comm.tid);
+	struct thread *thread = machine__findnew_thread(machine, event->comm.tid);
 
-	dump_printf(": %s:%d\n", event->comm.comm, event->comm.tid);
+	if (dump_trace)
+		perf_event__fprintf_comm(event, stdout);
 
 	if (thread == NULL || thread__set_comm(thread, event->comm.comm)) {
 		dump_printf("problem processing PERF_RECORD_COMM, skipping event.\n");
@@ -445,13 +514,13 @@ int perf_event__process_comm(union perf_event *event,
 	return 0;
 }
 
-int perf_event__process_lost(union perf_event *event,
+int perf_event__process_lost(struct perf_tool *tool __used,
+			     union perf_event *event,
 			     struct perf_sample *sample __used,
-			     struct perf_session *session)
+			     struct machine *machine __used)
 {
 	dump_printf(": id:%" PRIu64 ": lost:%" PRIu64 "\n",
 		    event->lost.id, event->lost.lost);
-	session->hists.stats.total_lost += event->lost.lost;
 	return 0;
 }
 
@@ -468,21 +537,15 @@ static void perf_event__set_kernel_mmap_len(union perf_event *event,
 		maps[MAP__FUNCTION]->end = ~0ULL;
 }
 
-static int perf_event__process_kernel_mmap(union perf_event *event,
-					   struct perf_session *session)
+static int perf_event__process_kernel_mmap(struct perf_tool *tool __used,
+					   union perf_event *event,
+					   struct machine *machine)
 {
 	struct map *map;
 	char kmmap_prefix[PATH_MAX];
-	struct machine *machine;
 	enum dso_kernel_type kernel_type;
 	bool is_kernel_mmap;
 
-	machine = perf_session__findnew_machine(session, event->mmap.pid);
-	if (!machine) {
-		pr_err("Can't find id %d's machine\n", event->mmap.pid);
-		goto out_problem;
-	}
-
 	machine__mmap_name(machine, kmmap_prefix, sizeof(kmmap_prefix));
 	if (machine__is_host(machine))
 		kernel_type = DSO_TYPE_KERNEL;
@@ -549,9 +612,9 @@ static int perf_event__process_kernel_mmap(union perf_event *event,
 		 * time /proc/sys/kernel/kptr_restrict was non zero.
 		 */
 		if (event->mmap.pgoff != 0) {
-			perf_session__set_kallsyms_ref_reloc_sym(machine->vmlinux_maps,
-								 symbol_name,
-								 event->mmap.pgoff);
+			maps__set_kallsyms_ref_reloc_sym(machine->vmlinux_maps,
+							 symbol_name,
+							 event->mmap.pgoff);
 		}
 
 		if (machine__is_default_guest(machine)) {
@@ -567,32 +630,35 @@ out_problem:
 	return -1;
 }
 
-int perf_event__process_mmap(union perf_event *event,
+size_t perf_event__fprintf_mmap(union perf_event *event, FILE *fp)
+{
+	return fprintf(fp, " %d/%d: [%#" PRIx64 "(%#" PRIx64 ") @ %#" PRIx64 "]: %s\n",
+		       event->mmap.pid, event->mmap.tid, event->mmap.start,
+		       event->mmap.len, event->mmap.pgoff, event->mmap.filename);
+}
+
+int perf_event__process_mmap(struct perf_tool *tool,
+			     union perf_event *event,
 			     struct perf_sample *sample __used,
-			     struct perf_session *session)
+			     struct machine *machine)
 {
-	struct machine *machine;
 	struct thread *thread;
 	struct map *map;
 	u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
 	int ret = 0;
 
-	dump_printf(" %d/%d: [%#" PRIx64 "(%#" PRIx64 ") @ %#" PRIx64 "]: %s\n",
-			event->mmap.pid, event->mmap.tid, event->mmap.start,
-			event->mmap.len, event->mmap.pgoff, event->mmap.filename);
+	if (dump_trace)
+		perf_event__fprintf_mmap(event, stdout);
 
 	if (cpumode == PERF_RECORD_MISC_GUEST_KERNEL ||
 	    cpumode == PERF_RECORD_MISC_KERNEL) {
-		ret = perf_event__process_kernel_mmap(event, session);
+		ret = perf_event__process_kernel_mmap(tool, event, machine);
 		if (ret < 0)
 			goto out_problem;
 		return 0;
 	}
 
-	machine = perf_session__find_host_machine(session);
-	if (machine == NULL)
-		goto out_problem;
-	thread = perf_session__findnew(session, event->mmap.pid);
+	thread = machine__findnew_thread(machine, event->mmap.pid);
 	if (thread == NULL)
 		goto out_problem;
 	map = map__new(&machine->user_dsos, event->mmap.start,
@@ -610,18 +676,26 @@ out_problem:
 	return 0;
 }
 
-int perf_event__process_task(union perf_event *event,
+size_t perf_event__fprintf_task(union perf_event *event, FILE *fp)
+{
+	return fprintf(fp, "(%d:%d):(%d:%d)\n",
+		       event->fork.pid, event->fork.tid,
+		       event->fork.ppid, event->fork.ptid);
+}
+
+int perf_event__process_task(struct perf_tool *tool __used,
+			     union perf_event *event,
 			     struct perf_sample *sample __used,
-			     struct perf_session *session)
+			      struct machine *machine)
 {
-	struct thread *thread = perf_session__findnew(session, event->fork.tid);
-	struct thread *parent = perf_session__findnew(session, event->fork.ptid);
+	struct thread *thread = machine__findnew_thread(machine, event->fork.tid);
+	struct thread *parent = machine__findnew_thread(machine, event->fork.ptid);
 
-	dump_printf("(%d:%d):(%d:%d)\n", event->fork.pid, event->fork.tid,
-		    event->fork.ppid, event->fork.ptid);
+	if (dump_trace)
+		perf_event__fprintf_task(event, stdout);
 
 	if (event->header.type == PERF_RECORD_EXIT) {
-		perf_session__remove_thread(session, thread);
+		machine__remove_thread(machine, thread);
 		return 0;
 	}
 
@@ -634,22 +708,45 @@ int perf_event__process_task(union perf_event *event,
 	return 0;
 }
 
-int perf_event__process(union perf_event *event, struct perf_sample *sample,
-			struct perf_session *session)
+size_t perf_event__fprintf(union perf_event *event, FILE *fp)
+{
+	size_t ret = fprintf(fp, "PERF_RECORD_%s",
+			     perf_event__name(event->header.type));
+
+	switch (event->header.type) {
+	case PERF_RECORD_COMM:
+		ret += perf_event__fprintf_comm(event, fp);
+		break;
+	case PERF_RECORD_FORK:
+	case PERF_RECORD_EXIT:
+		ret += perf_event__fprintf_task(event, fp);
+		break;
+	case PERF_RECORD_MMAP:
+		ret += perf_event__fprintf_mmap(event, fp);
+		break;
+	default:
+		ret += fprintf(fp, "\n");
+	}
+
+	return ret;
+}
+
+int perf_event__process(struct perf_tool *tool, union perf_event *event,
+			struct perf_sample *sample, struct machine *machine)
 {
 	switch (event->header.type) {
 	case PERF_RECORD_COMM:
-		perf_event__process_comm(event, sample, session);
+		perf_event__process_comm(tool, event, sample, machine);
 		break;
 	case PERF_RECORD_MMAP:
-		perf_event__process_mmap(event, sample, session);
+		perf_event__process_mmap(tool, event, sample, machine);
 		break;
 	case PERF_RECORD_FORK:
 	case PERF_RECORD_EXIT:
-		perf_event__process_task(event, sample, session);
+		perf_event__process_task(tool, event, sample, machine);
 		break;
 	case PERF_RECORD_LOST:
-		perf_event__process_lost(event, sample, session);
+		perf_event__process_lost(tool, event, sample, machine);
 	default:
 		break;
 	}
@@ -658,36 +755,29 @@ int perf_event__process(union perf_event *event, struct perf_sample *sample,
 }
 
 void thread__find_addr_map(struct thread *self,
-			   struct perf_session *session, u8 cpumode,
-			   enum map_type type, pid_t pid, u64 addr,
+			   struct machine *machine, u8 cpumode,
+			   enum map_type type, u64 addr,
 			   struct addr_location *al)
 {
 	struct map_groups *mg = &self->mg;
-	struct machine *machine = NULL;
 
 	al->thread = self;
 	al->addr = addr;
 	al->cpumode = cpumode;
 	al->filtered = false;
 
+	if (machine == NULL) {
+		al->map = NULL;
+		return;
+	}
+
 	if (cpumode == PERF_RECORD_MISC_KERNEL && perf_host) {
 		al->level = 'k';
-		machine = perf_session__find_host_machine(session);
-		if (machine == NULL) {
-			al->map = NULL;
-			return;
-		}
 		mg = &machine->kmaps;
 	} else if (cpumode == PERF_RECORD_MISC_USER && perf_host) {
 		al->level = '.';
-		machine = perf_session__find_host_machine(session);
 	} else if (cpumode == PERF_RECORD_MISC_GUEST_KERNEL && perf_guest) {
 		al->level = 'g';
-		machine = perf_session__find_machine(session, pid);
-		if (machine == NULL) {
-			al->map = NULL;
-			return;
-		}
 		mg = &machine->kmaps;
 	} else {
 		/*
@@ -733,13 +823,12 @@ try_again:
 		al->addr = al->map->map_ip(al->map, al->addr);
 }
 
-void thread__find_addr_location(struct thread *self,
-				struct perf_session *session, u8 cpumode,
-				enum map_type type, pid_t pid, u64 addr,
+void thread__find_addr_location(struct thread *thread, struct machine *machine,
+				u8 cpumode, enum map_type type, u64 addr,
 				struct addr_location *al,
 				symbol_filter_t filter)
 {
-	thread__find_addr_map(self, session, cpumode, type, pid, addr, al);
+	thread__find_addr_map(thread, machine, cpumode, type, addr, al);
 	if (al->map != NULL)
 		al->sym = map__find_symbol(al->map, al->addr, filter);
 	else
@@ -747,13 +836,13 @@ void thread__find_addr_location(struct thread *self,
 }
 
 int perf_event__preprocess_sample(const union perf_event *event,
-				  struct perf_session *session,
+				  struct machine *machine,
 				  struct addr_location *al,
 				  struct perf_sample *sample,
 				  symbol_filter_t filter)
 {
 	u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
-	struct thread *thread = perf_session__findnew(session, event->ip.pid);
+	struct thread *thread = machine__findnew_thread(machine, event->ip.pid);
 
 	if (thread == NULL)
 		return -1;
@@ -764,18 +853,18 @@ int perf_event__preprocess_sample(const union perf_event *event,
 
 	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
 	/*
-	 * Have we already created the kernel maps for the host machine?
+	 * Have we already created the kernel maps for this machine?
 	 *
 	 * This should have happened earlier, when we processed the kernel MMAP
 	 * events, but for older perf.data files there was no such thing, so do
 	 * it now.
 	 */
 	if (cpumode == PERF_RECORD_MISC_KERNEL &&
-	    session->host_machine.vmlinux_maps[MAP__FUNCTION] == NULL)
-		machine__create_kernel_maps(&session->host_machine);
+	    machine->vmlinux_maps[MAP__FUNCTION] == NULL)
+		machine__create_kernel_maps(machine);
 
-	thread__find_addr_map(thread, session, cpumode, MAP__FUNCTION,
-			      event->ip.pid, event->ip.ip, al);
+	thread__find_addr_map(thread, machine, cpumode, MAP__FUNCTION,
+			      event->ip.ip, al);
 	dump_printf(" ...... dso: %s\n",
 		    al->map ? al->map->dso->long_name :
 			al->level == 'H' ? "[hypervisor]" : "<not found>");
@@ -783,13 +872,14 @@ int perf_event__preprocess_sample(const union perf_event *event,
 	al->cpu = sample->cpu;
 
 	if (al->map) {
+		struct dso *dso = al->map->dso;
+
 		if (symbol_conf.dso_list &&
-		    (!al->map || !al->map->dso ||
-		     !(strlist__has_entry(symbol_conf.dso_list,
-					  al->map->dso->short_name) ||
-		       (al->map->dso->short_name != al->map->dso->long_name &&
-			strlist__has_entry(symbol_conf.dso_list,
-					   al->map->dso->long_name)))))
+		    (!dso || !(strlist__has_entry(symbol_conf.dso_list,
+						  dso->short_name) ||
+			       (dso->short_name != dso->long_name &&
+				strlist__has_entry(symbol_conf.dso_list,
+						   dso->long_name)))))
 			goto out_filtered;
 
 		al->sym = map__find_symbol(al->map, al->addr, filter);

tools/perf/util/event.h

@@ -2,6 +2,7 @@
 #define __PERF_RECORD_H
 
 #include <limits.h>
+#include <stdio.h>
 
 #include "../perf.h"
 #include "map.h"
@@ -141,43 +142,54 @@ union perf_event {
 
 void perf_event__print_totals(void);
 
-struct perf_session;
+struct perf_tool;
 struct thread_map;
 
-typedef int (*perf_event__handler_synth_t)(union perf_event *event, 
-					   struct perf_session *session);
-typedef int (*perf_event__handler_t)(union perf_event *event,
+typedef int (*perf_event__handler_t)(struct perf_tool *tool,
+				     union perf_event *event,
 				     struct perf_sample *sample,
-				      struct perf_session *session);
+				     struct machine *machine);
 
-int perf_event__synthesize_thread_map(struct thread_map *threads,
+int perf_event__synthesize_thread_map(struct perf_tool *tool,
+				      struct thread_map *threads,
 				      perf_event__handler_t process,
-				      struct perf_session *session);
-int perf_event__synthesize_threads(perf_event__handler_t process,
-				   struct perf_session *session);
-int perf_event__synthesize_kernel_mmap(perf_event__handler_t process,
-				       struct perf_session *session,
+				      struct machine *machine);
+int perf_event__synthesize_threads(struct perf_tool *tool,
+				   perf_event__handler_t process,
+				   struct machine *machine);
+int perf_event__synthesize_kernel_mmap(struct perf_tool *tool,
+				       perf_event__handler_t process,
 				       struct machine *machine,
 				       const char *symbol_name);
 
-int perf_event__synthesize_modules(perf_event__handler_t process,
-				   struct perf_session *session,
+int perf_event__synthesize_modules(struct perf_tool *tool,
+				   perf_event__handler_t process,
 				   struct machine *machine);
 
-int perf_event__process_comm(union perf_event *event, struct perf_sample *sample,
-			     struct perf_session *session);
-int perf_event__process_lost(union perf_event *event, struct perf_sample *sample,
-			     struct perf_session *session);
-int perf_event__process_mmap(union perf_event *event, struct perf_sample *sample,
-			     struct perf_session *session);
-int perf_event__process_task(union perf_event *event, struct perf_sample *sample,
-			     struct perf_session *session);
-int perf_event__process(union perf_event *event, struct perf_sample *sample,
-			struct perf_session *session);
+int perf_event__process_comm(struct perf_tool *tool,
+			     union perf_event *event,
+			     struct perf_sample *sample,
+			     struct machine *machine);
+int perf_event__process_lost(struct perf_tool *tool,
+			     union perf_event *event,
+			     struct perf_sample *sample,
+			     struct machine *machine);
+int perf_event__process_mmap(struct perf_tool *tool,
+			     union perf_event *event,
+			     struct perf_sample *sample,
+			     struct machine *machine);
+int perf_event__process_task(struct perf_tool *tool,
+			     union perf_event *event,
+			     struct perf_sample *sample,
+			     struct machine *machine);
+int perf_event__process(struct perf_tool *tool,
+			union perf_event *event,
+			struct perf_sample *sample,
+			struct machine *machine);
 
 struct addr_location;
 int perf_event__preprocess_sample(const union perf_event *self,
-				  struct perf_session *session,
+				  struct machine *machine,
 				  struct addr_location *al,
 				  struct perf_sample *sample,
 				  symbol_filter_t filter);
@@ -187,5 +199,13 @@ const char *perf_event__name(unsigned int id);
 int perf_event__parse_sample(const union perf_event *event, u64 type,
 			     int sample_size, bool sample_id_all,
 			     struct perf_sample *sample, bool swapped);
+int perf_event__synthesize_sample(union perf_event *event, u64 type,
+				  const struct perf_sample *sample,
+				  bool swapped);
+
+size_t perf_event__fprintf_comm(union perf_event *event, FILE *fp);
+size_t perf_event__fprintf_mmap(union perf_event *event, FILE *fp);
+size_t perf_event__fprintf_task(union perf_event *event, FILE *fp);
+size_t perf_event__fprintf(union perf_event *event, FILE *fp);
 
 #endif /* __PERF_RECORD_H */

tools/perf/util/evlist.c

@@ -6,12 +6,16 @@
  *
  * Released under the GPL v2. (and only v2, not any later version)
  */
+#include "util.h"
+#include "debugfs.h"
 #include <poll.h>
 #include "cpumap.h"
 #include "thread_map.h"
 #include "evlist.h"
 #include "evsel.h"
-#include "util.h"
+#include <unistd.h>
+
+#include "parse-events.h"
 
 #include <sys/mman.h>
 
@@ -30,6 +34,7 @@ void perf_evlist__init(struct perf_evlist *evlist, struct cpu_map *cpus,
 		INIT_HLIST_HEAD(&evlist->heads[i]);
 	INIT_LIST_HEAD(&evlist->entries);
 	perf_evlist__set_maps(evlist, cpus, threads);
+	evlist->workload.pid = -1;
 }
 
 struct perf_evlist *perf_evlist__new(struct cpu_map *cpus,
@@ -43,6 +48,22 @@ struct perf_evlist *perf_evlist__new(struct cpu_map *cpus,
 	return evlist;
 }
 
+void perf_evlist__config_attrs(struct perf_evlist *evlist,
+			       struct perf_record_opts *opts)
+{
+	struct perf_evsel *evsel;
+
+	if (evlist->cpus->map[0] < 0)
+		opts->no_inherit = true;
+
+	list_for_each_entry(evsel, &evlist->entries, node) {
+		perf_evsel__config(evsel, opts);
+
+		if (evlist->nr_entries > 1)
+			evsel->attr.sample_type |= PERF_SAMPLE_ID;
+	}
+}
+
 static void perf_evlist__purge(struct perf_evlist *evlist)
 {
 	struct perf_evsel *pos, *n;
@@ -76,6 +97,14 @@ void perf_evlist__add(struct perf_evlist *evlist, struct perf_evsel *entry)
 	++evlist->nr_entries;
 }
 
+static void perf_evlist__splice_list_tail(struct perf_evlist *evlist,
+					  struct list_head *list,
+					  int nr_entries)
+{
+	list_splice_tail(list, &evlist->entries);
+	evlist->nr_entries += nr_entries;
+}
+
 int perf_evlist__add_default(struct perf_evlist *evlist)
 {
 	struct perf_event_attr attr = {
@@ -100,6 +129,126 @@ error:
 	return -ENOMEM;
 }
 
+int perf_evlist__add_attrs(struct perf_evlist *evlist,
+			   struct perf_event_attr *attrs, size_t nr_attrs)
+{
+	struct perf_evsel *evsel, *n;
+	LIST_HEAD(head);
+	size_t i;
+
+	for (i = 0; i < nr_attrs; i++) {
+		evsel = perf_evsel__new(attrs + i, evlist->nr_entries + i);
+		if (evsel == NULL)
+			goto out_delete_partial_list;
+		list_add_tail(&evsel->node, &head);
+	}
+
+	perf_evlist__splice_list_tail(evlist, &head, nr_attrs);
+
+	return 0;
+
+out_delete_partial_list:
+	list_for_each_entry_safe(evsel, n, &head, node)
+		perf_evsel__delete(evsel);
+	return -1;
+}
+
+static int trace_event__id(const char *evname)
+{
+	char *filename, *colon;
+	int err = -1, fd;
+
+	if (asprintf(&filename, "%s/%s/id", tracing_events_path, evname) < 0)
+		return -1;
+
+	colon = strrchr(filename, ':');
+	if (colon != NULL)
+		*colon = '/';
+
+	fd = open(filename, O_RDONLY);
+	if (fd >= 0) {
+		char id[16];
+		if (read(fd, id, sizeof(id)) > 0)
+			err = atoi(id);
+		close(fd);
+	}
+
+	free(filename);
+	return err;
+}
+
+int perf_evlist__add_tracepoints(struct perf_evlist *evlist,
+				 const char *tracepoints[],
+				 size_t nr_tracepoints)
+{
+	int err;
+	size_t i;
+	struct perf_event_attr *attrs = zalloc(nr_tracepoints * sizeof(*attrs));
+
+	if (attrs == NULL)
+		return -1;
+
+	for (i = 0; i < nr_tracepoints; i++) {
+		err = trace_event__id(tracepoints[i]);
+
+		if (err < 0)
+			goto out_free_attrs;
+
+		attrs[i].type	       = PERF_TYPE_TRACEPOINT;
+		attrs[i].config	       = err;
+	        attrs[i].sample_type   = (PERF_SAMPLE_RAW | PERF_SAMPLE_TIME |
+					  PERF_SAMPLE_CPU);
+		attrs[i].sample_period = 1;
+	}
+
+	err = perf_evlist__add_attrs(evlist, attrs, nr_tracepoints);
+out_free_attrs:
+	free(attrs);
+	return err;
+}
+
+static struct perf_evsel *
+	perf_evlist__find_tracepoint_by_id(struct perf_evlist *evlist, int id)
+{
+	struct perf_evsel *evsel;
+
+	list_for_each_entry(evsel, &evlist->entries, node) {
+		if (evsel->attr.type   == PERF_TYPE_TRACEPOINT &&
+		    (int)evsel->attr.config == id)
+			return evsel;
+	}
+
+	return NULL;
+}
+
+int perf_evlist__set_tracepoints_handlers(struct perf_evlist *evlist,
+					  const struct perf_evsel_str_handler *assocs,
+					  size_t nr_assocs)
+{
+	struct perf_evsel *evsel;
+	int err;
+	size_t i;
+
+	for (i = 0; i < nr_assocs; i++) {
+		err = trace_event__id(assocs[i].name);
+		if (err < 0)
+			goto out;
+
+		evsel = perf_evlist__find_tracepoint_by_id(evlist, err);
+		if (evsel == NULL)
+			continue;
+
+		err = -EEXIST;
+		if (evsel->handler.func != NULL)
+			goto out;
+		evsel->handler.func = assocs[i].handler;
+	}
+
+	err = 0;
+out:
+	return err;
+}
+
 void perf_evlist__disable(struct perf_evlist *evlist)
 {
 	int cpu, thread;
@@ -126,7 +275,7 @@ void perf_evlist__enable(struct perf_evlist *evlist)
 	}
 }
 
-int perf_evlist__alloc_pollfd(struct perf_evlist *evlist)
+static int perf_evlist__alloc_pollfd(struct perf_evlist *evlist)
 {
 	int nfds = evlist->cpus->nr * evlist->threads->nr * evlist->nr_entries;
 	evlist->pollfd = malloc(sizeof(struct pollfd) * nfds);
@@ -282,7 +431,7 @@ void perf_evlist__munmap(struct perf_evlist *evlist)
 	evlist->mmap = NULL;
 }
 
-int perf_evlist__alloc_mmap(struct perf_evlist *evlist)
+static int perf_evlist__alloc_mmap(struct perf_evlist *evlist)
 {
 	evlist->nr_mmaps = evlist->cpus->nr;
 	if (evlist->cpus->map[0] == -1)
@@ -298,8 +447,10 @@ static int __perf_evlist__mmap(struct perf_evlist *evlist,
 	evlist->mmap[idx].mask = mask;
 	evlist->mmap[idx].base = mmap(NULL, evlist->mmap_len, prot,
 				      MAP_SHARED, fd, 0);
-	if (evlist->mmap[idx].base == MAP_FAILED)
+	if (evlist->mmap[idx].base == MAP_FAILED) {
+		evlist->mmap[idx].base = NULL;
 		return -1;
+	}
 
 	perf_evlist__add_pollfd(evlist, fd);
 	return 0;
@@ -400,14 +551,22 @@ out_unmap:
  *
  * Using perf_evlist__read_on_cpu does this automatically.
  */
-int perf_evlist__mmap(struct perf_evlist *evlist, int pages, bool overwrite)
+int perf_evlist__mmap(struct perf_evlist *evlist, unsigned int pages,
+		      bool overwrite)
 {
 	unsigned int page_size = sysconf(_SC_PAGE_SIZE);
-	int mask = pages * page_size - 1;
 	struct perf_evsel *evsel;
 	const struct cpu_map *cpus = evlist->cpus;
 	const struct thread_map *threads = evlist->threads;
-	int prot = PROT_READ | (overwrite ? 0 : PROT_WRITE);
+	int prot = PROT_READ | (overwrite ? 0 : PROT_WRITE), mask;
+
+        /* 512 kiB: default amount of unprivileged mlocked memory */
+        if (pages == UINT_MAX)
+                pages = (512 * 1024) / page_size;
+	else if (!is_power_of_2(pages))
+		return -EINVAL;
+
+	mask = pages * page_size - 1;
 
 	if (evlist->mmap == NULL && perf_evlist__alloc_mmap(evlist) < 0)
 		return -ENOMEM;
@@ -512,6 +671,38 @@ u64 perf_evlist__sample_type(const struct perf_evlist *evlist)
 	return first->attr.sample_type;
 }
 
+u16 perf_evlist__id_hdr_size(const struct perf_evlist *evlist)
+{
+	struct perf_evsel *first;
+	struct perf_sample *data;
+	u64 sample_type;
+	u16 size = 0;
+
+	first = list_entry(evlist->entries.next, struct perf_evsel, node);
+
+	if (!first->attr.sample_id_all)
+		goto out;
+
+	sample_type = first->attr.sample_type;
+
+	if (sample_type & PERF_SAMPLE_TID)
+		size += sizeof(data->tid) * 2;
+
+       if (sample_type & PERF_SAMPLE_TIME)
+		size += sizeof(data->time);
+
+	if (sample_type & PERF_SAMPLE_ID)
+		size += sizeof(data->id);
+
+	if (sample_type & PERF_SAMPLE_STREAM_ID)
+		size += sizeof(data->stream_id);
+
+	if (sample_type & PERF_SAMPLE_CPU)
+		size += sizeof(data->cpu) * 2;
+out:
+	return size;
+}
+
 bool perf_evlist__valid_sample_id_all(const struct perf_evlist *evlist)
 {
 	struct perf_evsel *pos, *first;
@@ -569,3 +760,97 @@ out_err:
 
 	return err;
 }
+
+int perf_evlist__prepare_workload(struct perf_evlist *evlist,
+				  struct perf_record_opts *opts,
+				  const char *argv[])
+{
+	int child_ready_pipe[2], go_pipe[2];
+	char bf;
+
+	if (pipe(child_ready_pipe) < 0) {
+		perror("failed to create 'ready' pipe");
+		return -1;
+	}
+
+	if (pipe(go_pipe) < 0) {
+		perror("failed to create 'go' pipe");
+		goto out_close_ready_pipe;
+	}
+
+	evlist->workload.pid = fork();
+	if (evlist->workload.pid < 0) {
+		perror("failed to fork");
+		goto out_close_pipes;
+	}
+
+	if (!evlist->workload.pid) {
+		if (opts->pipe_output)
+			dup2(2, 1);
+
+		close(child_ready_pipe[0]);
+		close(go_pipe[1]);
+		fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC);
+
+		/*
+		 * Do a dummy execvp to get the PLT entry resolved,
+		 * so we avoid the resolver overhead on the real
+		 * execvp call.
+		 */
+		execvp("", (char **)argv);
+
+		/*
+		 * Tell the parent we're ready to go
+		 */
+		close(child_ready_pipe[1]);
+
+		/*
+		 * Wait until the parent tells us to go.
+		 */
+		if (read(go_pipe[0], &bf, 1) == -1)
+			perror("unable to read pipe");
+
+		execvp(argv[0], (char **)argv);
+
+		perror(argv[0]);
+		kill(getppid(), SIGUSR1);
+		exit(-1);
+	}
+
+	if (!opts->system_wide && opts->target_tid == -1 && opts->target_pid == -1)
+		evlist->threads->map[0] = evlist->workload.pid;
+
+	close(child_ready_pipe[1]);
+	close(go_pipe[0]);
+	/*
+	 * wait for child to settle
+	 */
+	if (read(child_ready_pipe[0], &bf, 1) == -1) {
+		perror("unable to read pipe");
+		goto out_close_pipes;
+	}
+
+	evlist->workload.cork_fd = go_pipe[1];
+	close(child_ready_pipe[0]);
+	return 0;
+
+out_close_pipes:
+	close(go_pipe[0]);
+	close(go_pipe[1]);
+out_close_ready_pipe:
+	close(child_ready_pipe[0]);
+	close(child_ready_pipe[1]);
+	return -1;
+}
+
+int perf_evlist__start_workload(struct perf_evlist *evlist)
+{
+	if (evlist->workload.cork_fd > 0) {
+		/*
+		 * Remove the cork, let it rip!
+		 */
+		return close(evlist->workload.cork_fd);
+	}
+
+	return 0;
+}

tools/perf/util/evlist.h

@@ -2,12 +2,16 @@
 #define __PERF_EVLIST_H 1
 
 #include <linux/list.h>
+#include <stdio.h>
 #include "../perf.h"
 #include "event.h"
+#include "util.h"
+#include <unistd.h>
 
 struct pollfd;
 struct thread_map;
 struct cpu_map;
+struct perf_record_opts;
 
 #define PERF_EVLIST__HLIST_BITS 8
 #define PERF_EVLIST__HLIST_SIZE (1 << PERF_EVLIST__HLIST_BITS)
@@ -19,6 +23,10 @@ struct perf_evlist {
 	int		 nr_fds;
 	int		 nr_mmaps;
 	int		 mmap_len;
+	struct {
+		int	cork_fd;
+		pid_t	pid;
+	} workload;
 	bool		 overwrite;
 	union perf_event event_copy;
 	struct perf_mmap *mmap;
@@ -28,6 +36,11 @@ struct perf_evlist {
 	struct perf_evsel *selected;
 };
 
+struct perf_evsel_str_handler {
+	const char *name;
+	void	   *handler;
+};
+
 struct perf_evsel;
 
 struct perf_evlist *perf_evlist__new(struct cpu_map *cpus,
@@ -39,11 +52,26 @@ void perf_evlist__delete(struct perf_evlist *evlist);
 
 void perf_evlist__add(struct perf_evlist *evlist, struct perf_evsel *entry);
 int perf_evlist__add_default(struct perf_evlist *evlist);
+int perf_evlist__add_attrs(struct perf_evlist *evlist,
+			   struct perf_event_attr *attrs, size_t nr_attrs);
+int perf_evlist__add_tracepoints(struct perf_evlist *evlist,
+				 const char *tracepoints[], size_t nr_tracepoints);
+int perf_evlist__set_tracepoints_handlers(struct perf_evlist *evlist,
+					  const struct perf_evsel_str_handler *assocs,
+					  size_t nr_assocs);
+
+#define perf_evlist__add_attrs_array(evlist, array) \
+	perf_evlist__add_attrs(evlist, array, ARRAY_SIZE(array))
+
+#define perf_evlist__add_tracepoints_array(evlist, array) \
+	perf_evlist__add_tracepoints(evlist, array, ARRAY_SIZE(array))
+
+#define perf_evlist__set_tracepoints_handlers_array(evlist, array) \
+	perf_evlist__set_tracepoints_handlers(evlist, array, ARRAY_SIZE(array))
 
 void perf_evlist__id_add(struct perf_evlist *evlist, struct perf_evsel *evsel,
 			 int cpu, int thread, u64 id);
 
-int perf_evlist__alloc_pollfd(struct perf_evlist *evlist);
 void perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd);
 
 struct perf_evsel *perf_evlist__id2evsel(struct perf_evlist *evlist, u64 id);
@@ -52,8 +80,16 @@ union perf_event *perf_evlist__mmap_read(struct perf_evlist *self, int idx);
 
 int perf_evlist__open(struct perf_evlist *evlist, bool group);
 
-int perf_evlist__alloc_mmap(struct perf_evlist *evlist);
-int perf_evlist__mmap(struct perf_evlist *evlist, int pages, bool overwrite);
+void perf_evlist__config_attrs(struct perf_evlist *evlist,
+			       struct perf_record_opts *opts);
+
+int perf_evlist__prepare_workload(struct perf_evlist *evlist,
+				  struct perf_record_opts *opts,
+				  const char *argv[]);
+int perf_evlist__start_workload(struct perf_evlist *evlist);
+
+int perf_evlist__mmap(struct perf_evlist *evlist, unsigned int pages,
+		      bool overwrite);
 void perf_evlist__munmap(struct perf_evlist *evlist);
 
 void perf_evlist__disable(struct perf_evlist *evlist);
@@ -77,6 +113,7 @@ int perf_evlist__set_filters(struct perf_evlist *evlist);
 
 u64 perf_evlist__sample_type(const struct perf_evlist *evlist);
 bool perf_evlist__sample_id_all(const const struct perf_evlist *evlist);
+u16 perf_evlist__id_hdr_size(const struct perf_evlist *evlist);
 
 bool perf_evlist__valid_sample_type(const struct perf_evlist *evlist);
 bool perf_evlist__valid_sample_id_all(const struct perf_evlist *evlist);

tools/perf/util/evsel.c

@@ -63,6 +63,79 @@ struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr, int idx)
 	return evsel;
 }
 
+void perf_evsel__config(struct perf_evsel *evsel, struct perf_record_opts *opts)
+{
+	struct perf_event_attr *attr = &evsel->attr;
+	int track = !evsel->idx; /* only the first counter needs these */
+
+	attr->sample_id_all = opts->sample_id_all_avail ? 1 : 0;
+	attr->inherit	    = !opts->no_inherit;
+	attr->read_format   = PERF_FORMAT_TOTAL_TIME_ENABLED |
+			      PERF_FORMAT_TOTAL_TIME_RUNNING |
+			      PERF_FORMAT_ID;
+
+	attr->sample_type  |= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
+
+	/*
+	 * We default some events to a 1 default interval. But keep
+	 * it a weak assumption overridable by the user.
+	 */
+	if (!attr->sample_period || (opts->user_freq != UINT_MAX &&
+				     opts->user_interval != ULLONG_MAX)) {
+		if (opts->freq) {
+			attr->sample_type	|= PERF_SAMPLE_PERIOD;
+			attr->freq		= 1;
+			attr->sample_freq	= opts->freq;
+		} else {
+			attr->sample_period = opts->default_interval;
+		}
+	}
+
+	if (opts->no_samples)
+		attr->sample_freq = 0;
+
+	if (opts->inherit_stat)
+		attr->inherit_stat = 1;
+
+	if (opts->sample_address) {
+		attr->sample_type	|= PERF_SAMPLE_ADDR;
+		attr->mmap_data = track;
+	}
+
+	if (opts->call_graph)
+		attr->sample_type	|= PERF_SAMPLE_CALLCHAIN;
+
+	if (opts->system_wide)
+		attr->sample_type	|= PERF_SAMPLE_CPU;
+
+	if (opts->period)
+		attr->sample_type	|= PERF_SAMPLE_PERIOD;
+
+	if (opts->sample_id_all_avail &&
+	    (opts->sample_time || opts->system_wide ||
+	     !opts->no_inherit || opts->cpu_list))
+		attr->sample_type	|= PERF_SAMPLE_TIME;
+
+	if (opts->raw_samples) {
+		attr->sample_type	|= PERF_SAMPLE_TIME;
+		attr->sample_type	|= PERF_SAMPLE_RAW;
+		attr->sample_type	|= PERF_SAMPLE_CPU;
+	}
+
+	if (opts->no_delay) {
+		attr->watermark = 0;
+		attr->wakeup_events = 1;
+	}
+
+	attr->mmap = track;
+	attr->comm = track;
+
+	if (opts->target_pid == -1 && opts->target_tid == -1 && !opts->system_wide) {
+		attr->disabled = 1;
+		attr->enable_on_exec = 1;
+	}
+}
+
 int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads)
 {
 	int cpu, thread;
@@ -387,7 +460,7 @@ int perf_event__parse_sample(const union perf_event *event, u64 type,
 		u32 val32[2];
 	} u;
 
-
+	memset(data, 0, sizeof(*data));
 	data->cpu = data->pid = data->tid = -1;
 	data->stream_id = data->id = data->time = -1ULL;
 
@@ -504,3 +577,82 @@ int perf_event__parse_sample(const union perf_event *event, u64 type,
 
 	return 0;
 }
+
+int perf_event__synthesize_sample(union perf_event *event, u64 type,
+				  const struct perf_sample *sample,
+				  bool swapped)
+{
+	u64 *array;
+
+	/*
+	 * used for cross-endian analysis. See git commit 65014ab3
+	 * for why this goofiness is needed.
+	 */
+	union {
+		u64 val64;
+		u32 val32[2];
+	} u;
+
+	array = event->sample.array;
+
+	if (type & PERF_SAMPLE_IP) {
+		event->ip.ip = sample->ip;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_TID) {
+		u.val32[0] = sample->pid;
+		u.val32[1] = sample->tid;
+		if (swapped) {
+			/*
+			 * Inverse of what is done in perf_event__parse_sample
+			 */
+			u.val32[0] = bswap_32(u.val32[0]);
+			u.val32[1] = bswap_32(u.val32[1]);
+			u.val64 = bswap_64(u.val64);
+		}
+
+		*array = u.val64;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_TIME) {
+		*array = sample->time;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_ADDR) {
+		*array = sample->addr;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_ID) {
+		*array = sample->id;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_STREAM_ID) {
+		*array = sample->stream_id;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_CPU) {
+		u.val32[0] = sample->cpu;
+		if (swapped) {
+			/*
+			 * Inverse of what is done in perf_event__parse_sample
+			 */
+			u.val32[0] = bswap_32(u.val32[0]);
+			u.val64 = bswap_64(u.val64);
+		}
+		*array = u.val64;
+		array++;
+	}
+
+	if (type & PERF_SAMPLE_PERIOD) {
+		*array = sample->period;
+		array++;
+	}
+
+	return 0;
+}

tools/perf/util/evsel.h

@@ -61,12 +61,17 @@ struct perf_evsel {
 		off_t		id_offset;
 	};
 	struct cgroup_sel	*cgrp;
+	struct {
+		void		*func;
+		void		*data;
+	} handler;
 	bool 			supported;
 };
 
 struct cpu_map;
 struct thread_map;
 struct perf_evlist;
+struct perf_record_opts;
 
 struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr, int idx);
 void perf_evsel__init(struct perf_evsel *evsel,
@@ -74,6 +79,9 @@ void perf_evsel__init(struct perf_evsel *evsel,
 void perf_evsel__exit(struct perf_evsel *evsel);
 void perf_evsel__delete(struct perf_evsel *evsel);
 
+void perf_evsel__config(struct perf_evsel *evsel,
+			struct perf_record_opts *opts);
+
 int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads);
 int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads);
 int perf_evsel__alloc_counts(struct perf_evsel *evsel, int ncpus);

tools/perf/util/header.c

@@ -8,6 +8,7 @@
 #include <stdlib.h>
 #include <linux/list.h>
 #include <linux/kernel.h>
+#include <linux/bitops.h>
 #include <sys/utsname.h>
 
 #include "evlist.h"
@@ -28,9 +29,6 @@ static struct perf_trace_event_type *events;
 static u32 header_argc;
 static const char **header_argv;
 
-static int dsos__write_buildid_table(struct perf_header *header, int fd);
-static int perf_session__cache_build_ids(struct perf_session *session);
-
 int perf_header__push_event(u64 id, const char *name)
 {
 	if (strlen(name) > MAX_EVENT_NAME)
@@ -187,6 +185,252 @@ perf_header__set_cmdline(int argc, const char **argv)
 	return 0;
 }
 
+#define dsos__for_each_with_build_id(pos, head)	\
+	list_for_each_entry(pos, head, node)	\
+		if (!pos->has_build_id)		\
+			continue;		\
+		else
+
+static int __dsos__write_buildid_table(struct list_head *head, pid_t pid,
+				u16 misc, int fd)
+{
+	struct dso *pos;
+
+	dsos__for_each_with_build_id(pos, head) {
+		int err;
+		struct build_id_event b;
+		size_t len;
+
+		if (!pos->hit)
+			continue;
+		len = pos->long_name_len + 1;
+		len = ALIGN(len, NAME_ALIGN);
+		memset(&b, 0, sizeof(b));
+		memcpy(&b.build_id, pos->build_id, sizeof(pos->build_id));
+		b.pid = pid;
+		b.header.misc = misc;
+		b.header.size = sizeof(b) + len;
+		err = do_write(fd, &b, sizeof(b));
+		if (err < 0)
+			return err;
+		err = write_padded(fd, pos->long_name,
+				   pos->long_name_len + 1, len);
+		if (err < 0)
+			return err;
+	}
+
+	return 0;
+}
+
+static int machine__write_buildid_table(struct machine *machine, int fd)
+{
+	int err;
+	u16 kmisc = PERF_RECORD_MISC_KERNEL,
+	    umisc = PERF_RECORD_MISC_USER;
+
+	if (!machine__is_host(machine)) {
+		kmisc = PERF_RECORD_MISC_GUEST_KERNEL;
+		umisc = PERF_RECORD_MISC_GUEST_USER;
+	}
+
+	err = __dsos__write_buildid_table(&machine->kernel_dsos, machine->pid,
+					  kmisc, fd);
+	if (err == 0)
+		err = __dsos__write_buildid_table(&machine->user_dsos,
+						  machine->pid, umisc, fd);
+	return err;
+}
+
+static int dsos__write_buildid_table(struct perf_header *header, int fd)
+{
+	struct perf_session *session = container_of(header,
+			struct perf_session, header);
+	struct rb_node *nd;
+	int err = machine__write_buildid_table(&session->host_machine, fd);
+
+	if (err)
+		return err;
+
+	for (nd = rb_first(&session->machines); nd; nd = rb_next(nd)) {
+		struct machine *pos = rb_entry(nd, struct machine, rb_node);
+		err = machine__write_buildid_table(pos, fd);
+		if (err)
+			break;
+	}
+	return err;
+}
+
+int build_id_cache__add_s(const char *sbuild_id, const char *debugdir,
+			  const char *name, bool is_kallsyms)
+{
+	const size_t size = PATH_MAX;
+	char *realname, *filename = zalloc(size),
+	     *linkname = zalloc(size), *targetname;
+	int len, err = -1;
+
+	if (is_kallsyms) {
+		if (symbol_conf.kptr_restrict) {
+			pr_debug("Not caching a kptr_restrict'ed /proc/kallsyms\n");
+			return 0;
+		}
+		realname = (char *)name;
+	} else
+		realname = realpath(name, NULL);
+
+	if (realname == NULL || filename == NULL || linkname == NULL)
+		goto out_free;
+
+	len = snprintf(filename, size, "%s%s%s",
+		       debugdir, is_kallsyms ? "/" : "", realname);
+	if (mkdir_p(filename, 0755))
+		goto out_free;
+
+	snprintf(filename + len, sizeof(filename) - len, "/%s", sbuild_id);
+
+	if (access(filename, F_OK)) {
+		if (is_kallsyms) {
+			 if (copyfile("/proc/kallsyms", filename))
+				goto out_free;
+		} else if (link(realname, filename) && copyfile(name, filename))
+			goto out_free;
+	}
+
+	len = snprintf(linkname, size, "%s/.build-id/%.2s",
+		       debugdir, sbuild_id);
+
+	if (access(linkname, X_OK) && mkdir_p(linkname, 0755))
+		goto out_free;
+
+	snprintf(linkname + len, size - len, "/%s", sbuild_id + 2);
+	targetname = filename + strlen(debugdir) - 5;
+	memcpy(targetname, "../..", 5);
+
+	if (symlink(targetname, linkname) == 0)
+		err = 0;
+out_free:
+	if (!is_kallsyms)
+		free(realname);
+	free(filename);
+	free(linkname);
+	return err;
+}
+
+static int build_id_cache__add_b(const u8 *build_id, size_t build_id_size,
+				 const char *name, const char *debugdir,
+				 bool is_kallsyms)
+{
+	char sbuild_id[BUILD_ID_SIZE * 2 + 1];
+
+	build_id__sprintf(build_id, build_id_size, sbuild_id);
+
+	return build_id_cache__add_s(sbuild_id, debugdir, name, is_kallsyms);
+}
+
+int build_id_cache__remove_s(const char *sbuild_id, const char *debugdir)
+{
+	const size_t size = PATH_MAX;
+	char *filename = zalloc(size),
+	     *linkname = zalloc(size);
+	int err = -1;
+
+	if (filename == NULL || linkname == NULL)
+		goto out_free;
+
+	snprintf(linkname, size, "%s/.build-id/%.2s/%s",
+		 debugdir, sbuild_id, sbuild_id + 2);
+
+	if (access(linkname, F_OK))
+		goto out_free;
+
+	if (readlink(linkname, filename, size - 1) < 0)
+		goto out_free;
+
+	if (unlink(linkname))
+		goto out_free;
+
+	/*
+	 * Since the link is relative, we must make it absolute:
+	 */
+	snprintf(linkname, size, "%s/.build-id/%.2s/%s",
+		 debugdir, sbuild_id, filename);
+
+	if (unlink(linkname))
+		goto out_free;
+
+	err = 0;
+out_free:
+	free(filename);
+	free(linkname);
+	return err;
+}
+
+static int dso__cache_build_id(struct dso *dso, const char *debugdir)
+{
+	bool is_kallsyms = dso->kernel && dso->long_name[0] != '/';
+
+	return build_id_cache__add_b(dso->build_id, sizeof(dso->build_id),
+				     dso->long_name, debugdir, is_kallsyms);
+}
+
+static int __dsos__cache_build_ids(struct list_head *head, const char *debugdir)
+{
+	struct dso *pos;
+	int err = 0;
+
+	dsos__for_each_with_build_id(pos, head)
+		if (dso__cache_build_id(pos, debugdir))
+			err = -1;
+
+	return err;
+}
+
+static int machine__cache_build_ids(struct machine *machine, const char *debugdir)
+{
+	int ret = __dsos__cache_build_ids(&machine->kernel_dsos, debugdir);
+	ret |= __dsos__cache_build_ids(&machine->user_dsos, debugdir);
+	return ret;
+}
+
+static int perf_session__cache_build_ids(struct perf_session *session)
+{
+	struct rb_node *nd;
+	int ret;
+	char debugdir[PATH_MAX];
+
+	snprintf(debugdir, sizeof(debugdir), "%s", buildid_dir);
+
+	if (mkdir(debugdir, 0755) != 0 && errno != EEXIST)
+		return -1;
+
+	ret = machine__cache_build_ids(&session->host_machine, debugdir);
+
+	for (nd = rb_first(&session->machines); nd; nd = rb_next(nd)) {
+		struct machine *pos = rb_entry(nd, struct machine, rb_node);
+		ret |= machine__cache_build_ids(pos, debugdir);
+	}
+	return ret ? -1 : 0;
+}
+
+static bool machine__read_build_ids(struct machine *machine, bool with_hits)
+{
+	bool ret = __dsos__read_build_ids(&machine->kernel_dsos, with_hits);
+	ret |= __dsos__read_build_ids(&machine->user_dsos, with_hits);
+	return ret;
+}
+
+static bool perf_session__read_build_ids(struct perf_session *session, bool with_hits)
+{
+	struct rb_node *nd;
+	bool ret = machine__read_build_ids(&session->host_machine, with_hits);
+
+	for (nd = rb_first(&session->machines); nd; nd = rb_next(nd)) {
+		struct machine *pos = rb_entry(nd, struct machine, rb_node);
+		ret |= machine__read_build_ids(pos, with_hits);
+	}
+
+	return ret;
+}
+
 static int write_trace_info(int fd, struct perf_header *h __used,
 			    struct perf_evlist *evlist)
 {
@@ -202,6 +446,9 @@ static int write_build_id(int fd, struct perf_header *h,
 
 	session = container_of(h, struct perf_session, header);
 
+	if (!perf_session__read_build_ids(session, true))
+		return -1;
+
 	err = dsos__write_buildid_table(h, fd);
 	if (err < 0) {
 		pr_debug("failed to write buildid table\n");
@@ -1065,26 +1312,30 @@ struct feature_ops {
 	bool full_only;
 };
 
-#define FEAT_OPA(n, w, p) \
-	[n] = { .name = #n, .write = w, .print = p }
-#define FEAT_OPF(n, w, p) \
-	[n] = { .name = #n, .write = w, .print = p, .full_only = true }
+#define FEAT_OPA(n, func) \
+	[n] = { .name = #n, .write = write_##func, .print = print_##func }
+#define FEAT_OPF(n, func) \
+	[n] = { .name = #n, .write = write_##func, .print = print_##func, .full_only = true }
+
+/* feature_ops not implemented: */
+#define print_trace_info		NULL
+#define print_build_id			NULL
 
 static const struct feature_ops feat_ops[HEADER_LAST_FEATURE] = {
-	FEAT_OPA(HEADER_TRACE_INFO, write_trace_info, NULL),
-	FEAT_OPA(HEADER_BUILD_ID, write_build_id, NULL),
-	FEAT_OPA(HEADER_HOSTNAME, write_hostname, print_hostname),
-	FEAT_OPA(HEADER_OSRELEASE, write_osrelease, print_osrelease),
-	FEAT_OPA(HEADER_VERSION, write_version, print_version),
-	FEAT_OPA(HEADER_ARCH, write_arch, print_arch),
-	FEAT_OPA(HEADER_NRCPUS, write_nrcpus, print_nrcpus),
-	FEAT_OPA(HEADER_CPUDESC, write_cpudesc, print_cpudesc),
-	FEAT_OPA(HEADER_CPUID, write_cpuid, print_cpuid),
-	FEAT_OPA(HEADER_TOTAL_MEM, write_total_mem, print_total_mem),
-	FEAT_OPA(HEADER_EVENT_DESC, write_event_desc, print_event_desc),
-	FEAT_OPA(HEADER_CMDLINE, write_cmdline, print_cmdline),
-	FEAT_OPF(HEADER_CPU_TOPOLOGY, write_cpu_topology, print_cpu_topology),
-	FEAT_OPF(HEADER_NUMA_TOPOLOGY, write_numa_topology, print_numa_topology),
+	FEAT_OPA(HEADER_TRACE_INFO,	trace_info),
+	FEAT_OPA(HEADER_BUILD_ID,	build_id),
+	FEAT_OPA(HEADER_HOSTNAME,	hostname),
+	FEAT_OPA(HEADER_OSRELEASE,	osrelease),
+	FEAT_OPA(HEADER_VERSION,	version),
+	FEAT_OPA(HEADER_ARCH,		arch),
+	FEAT_OPA(HEADER_NRCPUS,		nrcpus),
+	FEAT_OPA(HEADER_CPUDESC,	cpudesc),
+	FEAT_OPA(HEADER_CPUID,		cpuid),
+	FEAT_OPA(HEADER_TOTAL_MEM,	total_mem),
+	FEAT_OPA(HEADER_EVENT_DESC,	event_desc),
+	FEAT_OPA(HEADER_CMDLINE,	cmdline),
+	FEAT_OPF(HEADER_CPU_TOPOLOGY,	cpu_topology),
+	FEAT_OPF(HEADER_NUMA_TOPOLOGY,	numa_topology),
 };
 
 struct header_print_data {
@@ -1103,9 +1354,9 @@ static int perf_file_section__fprintf_info(struct perf_file_section *section,
 				"%d, continuing...\n", section->offset, feat);
 		return 0;
 	}
-	if (feat < HEADER_TRACE_INFO || feat >= HEADER_LAST_FEATURE) {
+	if (feat >= HEADER_LAST_FEATURE) {
 		pr_warning("unknown feature %d\n", feat);
-		return -1;
+		return 0;
 	}
 	if (!feat_ops[feat].print)
 		return 0;
@@ -1132,252 +1383,6 @@ int perf_header__fprintf_info(struct perf_session *session, FILE *fp, bool full)
 	return 0;
 }
 
-#define dsos__for_each_with_build_id(pos, head)	\
-	list_for_each_entry(pos, head, node)	\
-		if (!pos->has_build_id)		\
-			continue;		\
-		else
-
-static int __dsos__write_buildid_table(struct list_head *head, pid_t pid,
-				u16 misc, int fd)
-{
-	struct dso *pos;
-
-	dsos__for_each_with_build_id(pos, head) {
-		int err;
-		struct build_id_event b;
-		size_t len;
-
-		if (!pos->hit)
-			continue;
-		len = pos->long_name_len + 1;
-		len = ALIGN(len, NAME_ALIGN);
-		memset(&b, 0, sizeof(b));
-		memcpy(&b.build_id, pos->build_id, sizeof(pos->build_id));
-		b.pid = pid;
-		b.header.misc = misc;
-		b.header.size = sizeof(b) + len;
-		err = do_write(fd, &b, sizeof(b));
-		if (err < 0)
-			return err;
-		err = write_padded(fd, pos->long_name,
-				   pos->long_name_len + 1, len);
-		if (err < 0)
-			return err;
-	}
-
-	return 0;
-}
-
-static int machine__write_buildid_table(struct machine *machine, int fd)
-{
-	int err;
-	u16 kmisc = PERF_RECORD_MISC_KERNEL,
-	    umisc = PERF_RECORD_MISC_USER;
-
-	if (!machine__is_host(machine)) {
-		kmisc = PERF_RECORD_MISC_GUEST_KERNEL;
-		umisc = PERF_RECORD_MISC_GUEST_USER;
-	}
-
-	err = __dsos__write_buildid_table(&machine->kernel_dsos, machine->pid,
-					  kmisc, fd);
-	if (err == 0)
-		err = __dsos__write_buildid_table(&machine->user_dsos,
-						  machine->pid, umisc, fd);
-	return err;
-}
-
-static int dsos__write_buildid_table(struct perf_header *header, int fd)
-{
-	struct perf_session *session = container_of(header,
-			struct perf_session, header);
-	struct rb_node *nd;
-	int err = machine__write_buildid_table(&session->host_machine, fd);
-
-	if (err)
-		return err;
-
-	for (nd = rb_first(&session->machines); nd; nd = rb_next(nd)) {
-		struct machine *pos = rb_entry(nd, struct machine, rb_node);
-		err = machine__write_buildid_table(pos, fd);
-		if (err)
-			break;
-	}
-	return err;
-}
-
-int build_id_cache__add_s(const char *sbuild_id, const char *debugdir,
-			  const char *name, bool is_kallsyms)
-{
-	const size_t size = PATH_MAX;
-	char *realname, *filename = zalloc(size),
-	     *linkname = zalloc(size), *targetname;
-	int len, err = -1;
-
-	if (is_kallsyms) {
-		if (symbol_conf.kptr_restrict) {
-			pr_debug("Not caching a kptr_restrict'ed /proc/kallsyms\n");
-			return 0;
-		}
-		realname = (char *)name;
-	} else
-		realname = realpath(name, NULL);
-
-	if (realname == NULL || filename == NULL || linkname == NULL)
-		goto out_free;
-
-	len = snprintf(filename, size, "%s%s%s",
-		       debugdir, is_kallsyms ? "/" : "", realname);
-	if (mkdir_p(filename, 0755))
-		goto out_free;
-
-	snprintf(filename + len, sizeof(filename) - len, "/%s", sbuild_id);
-
-	if (access(filename, F_OK)) {
-		if (is_kallsyms) {
-			 if (copyfile("/proc/kallsyms", filename))
-				goto out_free;
-		} else if (link(realname, filename) && copyfile(name, filename))
-			goto out_free;
-	}
-
-	len = snprintf(linkname, size, "%s/.build-id/%.2s",
-		       debugdir, sbuild_id);
-
-	if (access(linkname, X_OK) && mkdir_p(linkname, 0755))
-		goto out_free;
-
-	snprintf(linkname + len, size - len, "/%s", sbuild_id + 2);
-	targetname = filename + strlen(debugdir) - 5;
-	memcpy(targetname, "../..", 5);
-
-	if (symlink(targetname, linkname) == 0)
-		err = 0;
-out_free:
-	if (!is_kallsyms)
-		free(realname);
-	free(filename);
-	free(linkname);
-	return err;
-}
-
-static int build_id_cache__add_b(const u8 *build_id, size_t build_id_size,
-				 const char *name, const char *debugdir,
-				 bool is_kallsyms)
-{
-	char sbuild_id[BUILD_ID_SIZE * 2 + 1];
-
-	build_id__sprintf(build_id, build_id_size, sbuild_id);
-
-	return build_id_cache__add_s(sbuild_id, debugdir, name, is_kallsyms);
-}
-
-int build_id_cache__remove_s(const char *sbuild_id, const char *debugdir)
-{
-	const size_t size = PATH_MAX;
-	char *filename = zalloc(size),
-	     *linkname = zalloc(size);
-	int err = -1;
-
-	if (filename == NULL || linkname == NULL)
-		goto out_free;
-
-	snprintf(linkname, size, "%s/.build-id/%.2s/%s",
-		 debugdir, sbuild_id, sbuild_id + 2);
-
-	if (access(linkname, F_OK))
-		goto out_free;
-
-	if (readlink(linkname, filename, size - 1) < 0)
-		goto out_free;
-
-	if (unlink(linkname))
-		goto out_free;
-
-	/*
-	 * Since the link is relative, we must make it absolute:
-	 */
-	snprintf(linkname, size, "%s/.build-id/%.2s/%s",
-		 debugdir, sbuild_id, filename);
-
-	if (unlink(linkname))
-		goto out_free;
-
-	err = 0;
-out_free:
-	free(filename);
-	free(linkname);
-	return err;
-}
-
-static int dso__cache_build_id(struct dso *dso, const char *debugdir)
-{
-	bool is_kallsyms = dso->kernel && dso->long_name[0] != '/';
-
-	return build_id_cache__add_b(dso->build_id, sizeof(dso->build_id),
-				     dso->long_name, debugdir, is_kallsyms);
-}
-
-static int __dsos__cache_build_ids(struct list_head *head, const char *debugdir)
-{
-	struct dso *pos;
-	int err = 0;
-
-	dsos__for_each_with_build_id(pos, head)
-		if (dso__cache_build_id(pos, debugdir))
-			err = -1;
-
-	return err;
-}
-
-static int machine__cache_build_ids(struct machine *machine, const char *debugdir)
-{
-	int ret = __dsos__cache_build_ids(&machine->kernel_dsos, debugdir);
-	ret |= __dsos__cache_build_ids(&machine->user_dsos, debugdir);
-	return ret;
-}
-
-static int perf_session__cache_build_ids(struct perf_session *session)
-{
-	struct rb_node *nd;
-	int ret;
-	char debugdir[PATH_MAX];
-
-	snprintf(debugdir, sizeof(debugdir), "%s", buildid_dir);
-
-	if (mkdir(debugdir, 0755) != 0 && errno != EEXIST)
-		return -1;
-
-	ret = machine__cache_build_ids(&session->host_machine, debugdir);
-
-	for (nd = rb_first(&session->machines); nd; nd = rb_next(nd)) {
-		struct machine *pos = rb_entry(nd, struct machine, rb_node);
-		ret |= machine__cache_build_ids(pos, debugdir);
-	}
-	return ret ? -1 : 0;
-}
-
-static bool machine__read_build_ids(struct machine *machine, bool with_hits)
-{
-	bool ret = __dsos__read_build_ids(&machine->kernel_dsos, with_hits);
-	ret |= __dsos__read_build_ids(&machine->user_dsos, with_hits);
-	return ret;
-}
-
-static bool perf_session__read_build_ids(struct perf_session *session, bool with_hits)
-{
-	struct rb_node *nd;
-	bool ret = machine__read_build_ids(&session->host_machine, with_hits);
-
-	for (nd = rb_first(&session->machines); nd; nd = rb_next(nd)) {
-		struct machine *pos = rb_entry(nd, struct machine, rb_node);
-		ret |= machine__read_build_ids(pos, with_hits);
-	}
-
-	return ret;
-}
-
 static int do_write_feat(int fd, struct perf_header *h, int type,
 			 struct perf_file_section **p,
 			 struct perf_evlist *evlist)
@@ -1386,6 +1391,8 @@ static int do_write_feat(int fd, struct perf_header *h, int type,
 	int ret = 0;
 
 	if (perf_header__has_feat(h, type)) {
+		if (!feat_ops[type].write)
+			return -1;
 
 		(*p)->offset = lseek(fd, 0, SEEK_CUR);
 
@@ -1408,18 +1415,12 @@ static int perf_header__adds_write(struct perf_header *header,
 				   struct perf_evlist *evlist, int fd)
 {
 	int nr_sections;
-	struct perf_session *session;
 	struct perf_file_section *feat_sec, *p;
 	int sec_size;
 	u64 sec_start;
+	int feat;
 	int err;
 
-	session = container_of(header, struct perf_session, header);
-
-	if (perf_header__has_feat(header, HEADER_BUILD_ID &&
-	    !perf_session__read_build_ids(session, true)))
-		perf_header__clear_feat(header, HEADER_BUILD_ID);
-
 	nr_sections = bitmap_weight(header->adds_features, HEADER_FEAT_BITS);
 	if (!nr_sections)
 		return 0;
@@ -1433,64 +1434,11 @@ static int perf_header__adds_write(struct perf_header *header,
 	sec_start = header->data_offset + header->data_size;
 	lseek(fd, sec_start + sec_size, SEEK_SET);
 
-	err = do_write_feat(fd, header, HEADER_TRACE_INFO, &p, evlist);
-	if (err)
-		goto out_free;
-
-	err = do_write_feat(fd, header, HEADER_BUILD_ID, &p, evlist);
-	if (err) {
-		perf_header__clear_feat(header, HEADER_BUILD_ID);
-		goto out_free;
+	for_each_set_bit(feat, header->adds_features, HEADER_FEAT_BITS) {
+		if (do_write_feat(fd, header, feat, &p, evlist))
+			perf_header__clear_feat(header, feat);
 	}
 
-	err = do_write_feat(fd, header, HEADER_HOSTNAME, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_HOSTNAME);
-
-	err = do_write_feat(fd, header, HEADER_OSRELEASE, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_OSRELEASE);
-
-	err = do_write_feat(fd, header, HEADER_VERSION, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_VERSION);
-
-	err = do_write_feat(fd, header, HEADER_ARCH, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_ARCH);
-
-	err = do_write_feat(fd, header, HEADER_NRCPUS, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_NRCPUS);
-
-	err = do_write_feat(fd, header, HEADER_CPUDESC, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_CPUDESC);
-
-	err = do_write_feat(fd, header, HEADER_CPUID, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_CPUID);
-
-	err = do_write_feat(fd, header, HEADER_TOTAL_MEM, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_TOTAL_MEM);
-
-	err = do_write_feat(fd, header, HEADER_CMDLINE, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_CMDLINE);
-
-	err = do_write_feat(fd, header, HEADER_EVENT_DESC, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_EVENT_DESC);
-
-	err = do_write_feat(fd, header, HEADER_CPU_TOPOLOGY, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_CPU_TOPOLOGY);
-
-	err = do_write_feat(fd, header, HEADER_NUMA_TOPOLOGY, &p, evlist);
-	if (err)
-		perf_header__clear_feat(header, HEADER_NUMA_TOPOLOGY);
-
 	lseek(fd, sec_start, SEEK_SET);
 	/*
 	 * may write more than needed due to dropped feature, but
@@ -1499,7 +1447,6 @@ static int perf_header__adds_write(struct perf_header *header,
 	err = do_write(fd, feat_sec, sec_size);
 	if (err < 0)
 		pr_debug("failed to write feature section\n");
-out_free:
 	free(feat_sec);
 	return err;
 }
@@ -1637,20 +1584,20 @@ static int perf_header__getbuffer64(struct perf_header *header,
 int perf_header__process_sections(struct perf_header *header, int fd,
 				  void *data,
 				  int (*process)(struct perf_file_section *section,
-				  struct perf_header *ph,
-				  int feat, int fd, void *data))
+						 struct perf_header *ph,
+						 int feat, int fd, void *data))
 {
-	struct perf_file_section *feat_sec;
+	struct perf_file_section *feat_sec, *sec;
 	int nr_sections;
 	int sec_size;
-	int idx = 0;
-	int err = -1, feat = 1;
+	int feat;
+	int err;
 
 	nr_sections = bitmap_weight(header->adds_features, HEADER_FEAT_BITS);
 	if (!nr_sections)
 		return 0;
 
-	feat_sec = calloc(sizeof(*feat_sec), nr_sections);
+	feat_sec = sec = calloc(sizeof(*feat_sec), nr_sections);
 	if (!feat_sec)
 		return -1;
 
@@ -1658,20 +1605,16 @@ int perf_header__process_sections(struct perf_header *header, int fd,
 
 	lseek(fd, header->data_offset + header->data_size, SEEK_SET);
 
-	if (perf_header__getbuffer64(header, fd, feat_sec, sec_size))
+	err = perf_header__getbuffer64(header, fd, feat_sec, sec_size);
+	if (err < 0)
 		goto out_free;
 
-	err = 0;
-	while (idx < nr_sections && feat < HEADER_LAST_FEATURE) {
-		if (perf_header__has_feat(header, feat)) {
-			struct perf_file_section *sec = &feat_sec[idx++];
-
-			err = process(sec, header, feat, fd, data);
-			if (err < 0)
-				break;
-		}
-		++feat;
+	for_each_set_bit(feat, header->adds_features, HEADER_LAST_FEATURE) {
+		err = process(sec++, header, feat, fd, data);
+		if (err < 0)
+			goto out_free;
 	}
+	err = 0;
 out_free:
 	free(feat_sec);
 	return err;
@@ -1906,32 +1849,21 @@ static int perf_file_section__process(struct perf_file_section *section,
 		return 0;
 	}
 
+	if (feat >= HEADER_LAST_FEATURE) {
+		pr_debug("unknown feature %d, continuing...\n", feat);
+		return 0;
+	}
+
 	switch (feat) {
 	case HEADER_TRACE_INFO:
 		trace_report(fd, false);
 		break;
-
 	case HEADER_BUILD_ID:
 		if (perf_header__read_build_ids(ph, fd, section->offset, section->size))
 			pr_debug("Failed to read buildids, continuing...\n");
 		break;
-
-	case HEADER_HOSTNAME:
-	case HEADER_OSRELEASE:
-	case HEADER_VERSION:
-	case HEADER_ARCH:
-	case HEADER_NRCPUS:
-	case HEADER_CPUDESC:
-	case HEADER_CPUID:
-	case HEADER_TOTAL_MEM:
-	case HEADER_CMDLINE:
-	case HEADER_EVENT_DESC:
-	case HEADER_CPU_TOPOLOGY:
-	case HEADER_NUMA_TOPOLOGY:
-		break;
-
 	default:
-		pr_debug("unknown feature %d, continuing...\n", feat);
+		break;
 	}
 
 	return 0;
@@ -2041,6 +1973,8 @@ int perf_session__read_header(struct perf_session *session, int fd)
 		lseek(fd, tmp, SEEK_SET);
 	}
 
+	symbol_conf.nr_events = nr_attrs;
+
 	if (f_header.event_types.size) {
 		lseek(fd, f_header.event_types.offset, SEEK_SET);
 		events = malloc(f_header.event_types.size);
@@ -2068,9 +2002,9 @@ out_delete_evlist:
 	return -ENOMEM;
 }
 
-int perf_event__synthesize_attr(struct perf_event_attr *attr, u16 ids, u64 *id,
-				perf_event__handler_t process,
-				struct perf_session *session)
+int perf_event__synthesize_attr(struct perf_tool *tool,
+				struct perf_event_attr *attr, u16 ids, u64 *id,
+				perf_event__handler_t process)
 {
 	union perf_event *ev;
 	size_t size;
@@ -2092,22 +2026,23 @@ int perf_event__synthesize_attr(struct perf_event_attr *attr, u16 ids, u64 *id,
 	ev->attr.header.type = PERF_RECORD_HEADER_ATTR;
 	ev->attr.header.size = size;
 
-	err = process(ev, NULL, session);
+	err = process(tool, ev, NULL, NULL);
 
 	free(ev);
 
 	return err;
 }
 
-int perf_session__synthesize_attrs(struct perf_session *session,
+int perf_event__synthesize_attrs(struct perf_tool *tool,
+				   struct perf_session *session,
 				   perf_event__handler_t process)
 {
 	struct perf_evsel *attr;
 	int err = 0;
 
 	list_for_each_entry(attr, &session->evlist->entries, node) {
-		err = perf_event__synthesize_attr(&attr->attr, attr->ids,
-						  attr->id, process, session);
+		err = perf_event__synthesize_attr(tool, &attr->attr, attr->ids,
+						  attr->id, process);
 		if (err) {
 			pr_debug("failed to create perf header attribute\n");
 			return err;
@@ -2118,23 +2053,23 @@ int perf_session__synthesize_attrs(struct perf_session *session,
 }
 
 int perf_event__process_attr(union perf_event *event,
-			     struct perf_session *session)
+			     struct perf_evlist **pevlist)
 {
 	unsigned int i, ids, n_ids;
 	struct perf_evsel *evsel;
+	struct perf_evlist *evlist = *pevlist;
 
-	if (session->evlist == NULL) {
-		session->evlist = perf_evlist__new(NULL, NULL);
-		if (session->evlist == NULL)
+	if (evlist == NULL) {
+		*pevlist = evlist = perf_evlist__new(NULL, NULL);
+		if (evlist == NULL)
 			return -ENOMEM;
 	}
 
-	evsel = perf_evsel__new(&event->attr.attr,
-				session->evlist->nr_entries);
+	evsel = perf_evsel__new(&event->attr.attr, evlist->nr_entries);
 	if (evsel == NULL)
 		return -ENOMEM;
 
-	perf_evlist__add(session->evlist, evsel);
+	perf_evlist__add(evlist, evsel);
 
 	ids = event->header.size;
 	ids -= (void *)&event->attr.id - (void *)event;
@@ -2148,18 +2083,16 @@ int perf_event__process_attr(union perf_event *event,
 		return -ENOMEM;
 
 	for (i = 0; i < n_ids; i++) {
-		perf_evlist__id_add(session->evlist, evsel, 0, i,
-				    event->attr.id[i]);
+		perf_evlist__id_add(evlist, evsel, 0, i, event->attr.id[i]);
 	}
 
-	perf_session__update_sample_type(session);
-
 	return 0;
 }
 
-int perf_event__synthesize_event_type(u64 event_id, char *name,
+int perf_event__synthesize_event_type(struct perf_tool *tool,
+				      u64 event_id, char *name,
 				      perf_event__handler_t process,
-				      struct perf_session *session)
+				      struct machine *machine)
 {
 	union perf_event ev;
 	size_t size = 0;
@@ -2177,13 +2110,14 @@ int perf_event__synthesize_event_type(u64 event_id, char *name,
 	ev.event_type.header.size = sizeof(ev.event_type) -
 		(sizeof(ev.event_type.event_type.name) - size);
 
-	err = process(&ev, NULL, session);
+	err = process(tool, &ev, NULL, machine);
 
 	return err;
 }
 
-int perf_event__synthesize_event_types(perf_event__handler_t process,
-				       struct perf_session *session)
+int perf_event__synthesize_event_types(struct perf_tool *tool,
+				       perf_event__handler_t process,
+				       struct machine *machine)
 {
 	struct perf_trace_event_type *type;
 	int i, err = 0;
@@ -2191,9 +2125,9 @@ int perf_event__synthesize_event_types(perf_event__handler_t process,
 	for (i = 0; i < event_count; i++) {
 		type = &events[i];
 
-		err = perf_event__synthesize_event_type(type->event_id,
+		err = perf_event__synthesize_event_type(tool, type->event_id,
 							type->name, process,
-							session);
+							machine);
 		if (err) {
 			pr_debug("failed to create perf header event type\n");
 			return err;
@@ -2203,8 +2137,8 @@ int perf_event__synthesize_event_types(perf_event__handler_t process,
 	return err;
 }
 
-int perf_event__process_event_type(union perf_event *event,
-				   struct perf_session *session __unused)
+int perf_event__process_event_type(struct perf_tool *tool __unused,
+				   union perf_event *event)
 {
 	if (perf_header__push_event(event->event_type.event_type.event_id,
 				    event->event_type.event_type.name) < 0)
@@ -2213,9 +2147,9 @@ int perf_event__process_event_type(union perf_event *event,
 	return 0;
 }
 
-int perf_event__synthesize_tracing_data(int fd, struct perf_evlist *evlist,
-					 perf_event__handler_t process,
-				   struct perf_session *session __unused)
+int perf_event__synthesize_tracing_data(struct perf_tool *tool, int fd,
+					struct perf_evlist *evlist,
+					perf_event__handler_t process)
 {
 	union perf_event ev;
 	struct tracing_data *tdata;
@@ -2246,7 +2180,7 @@ int perf_event__synthesize_tracing_data(int fd, struct perf_evlist *evlist,
 	ev.tracing_data.header.size = sizeof(ev.tracing_data);
 	ev.tracing_data.size = aligned_size;
 
-	process(&ev, NULL, session);
+	process(tool, &ev, NULL, NULL);
 
 	/*
 	 * The put function will copy all the tracing data
@@ -2288,10 +2222,10 @@ int perf_event__process_tracing_data(union perf_event *event,
 	return size_read + padding;
 }
 
-int perf_event__synthesize_build_id(struct dso *pos, u16 misc,
+int perf_event__synthesize_build_id(struct perf_tool *tool,
+				    struct dso *pos, u16 misc,
 				    perf_event__handler_t process,
-				    struct machine *machine,
-				    struct perf_session *session)
+				    struct machine *machine)
 {
 	union perf_event ev;
 	size_t len;
@@ -2311,12 +2245,13 @@ int perf_event__synthesize_build_id(struct dso *pos, u16 misc,
 	ev.build_id.header.size = sizeof(ev.build_id) + len;
 	memcpy(&ev.build_id.filename, pos->long_name, pos->long_name_len);
 
-	err = process(&ev, NULL, session);
+	err = process(tool, &ev, NULL, machine);
 
 	return err;
 }
 
-int perf_event__process_build_id(union perf_event *event,
+int perf_event__process_build_id(struct perf_tool *tool __used,
+				 union perf_event *event,
 				 struct perf_session *session)
 {
 	__event_process_build_id(&event->build_id,

tools/perf/util/header.h

@@ -10,7 +10,8 @@
 #include <linux/bitmap.h>
 
 enum {
-	HEADER_TRACE_INFO = 1,
+	HEADER_RESERVED		= 0,	/* always cleared */
+	HEADER_TRACE_INFO	= 1,
 	HEADER_BUILD_ID,
 
 	HEADER_HOSTNAME,
@@ -27,10 +28,9 @@ enum {
 	HEADER_NUMA_TOPOLOGY,
 
 	HEADER_LAST_FEATURE,
+	HEADER_FEAT_BITS	= 256,
 };
 
-#define HEADER_FEAT_BITS			256
-
 struct perf_file_section {
 	u64 offset;
 	u64 size;
@@ -68,6 +68,7 @@ struct perf_header {
 };
 
 struct perf_evlist;
+struct perf_session;
 
 int perf_session__read_header(struct perf_session *session, int fd);
 int perf_session__write_header(struct perf_session *session,
@@ -96,32 +97,36 @@ int build_id_cache__add_s(const char *sbuild_id, const char *debugdir,
 			  const char *name, bool is_kallsyms);
 int build_id_cache__remove_s(const char *sbuild_id, const char *debugdir);
 
-int perf_event__synthesize_attr(struct perf_event_attr *attr, u16 ids, u64 *id,
-				perf_event__handler_t process,
-				struct perf_session *session);
-int perf_session__synthesize_attrs(struct perf_session *session,
-				   perf_event__handler_t process);
-int perf_event__process_attr(union perf_event *event, struct perf_session *session);
+int perf_event__synthesize_attr(struct perf_tool *tool,
+				struct perf_event_attr *attr, u16 ids, u64 *id,
+				perf_event__handler_t process);
+int perf_event__synthesize_attrs(struct perf_tool *tool,
+				 struct perf_session *session,
+				 perf_event__handler_t process);
+int perf_event__process_attr(union perf_event *event, struct perf_evlist **pevlist);
 
-int perf_event__synthesize_event_type(u64 event_id, char *name,
+int perf_event__synthesize_event_type(struct perf_tool *tool,
+				      u64 event_id, char *name,
 				      perf_event__handler_t process,
-				      struct perf_session *session);
-int perf_event__synthesize_event_types(perf_event__handler_t process,
-				       struct perf_session *session);
-int perf_event__process_event_type(union perf_event *event,
-				   struct perf_session *session);
-
-int perf_event__synthesize_tracing_data(int fd, struct perf_evlist *evlist,
-					perf_event__handler_t process,
-					struct perf_session *session);
+				      struct machine *machine);
+int perf_event__synthesize_event_types(struct perf_tool *tool,
+				       perf_event__handler_t process,
+				       struct machine *machine);
+int perf_event__process_event_type(struct perf_tool *tool,
+				   union perf_event *event);
+
+int perf_event__synthesize_tracing_data(struct perf_tool *tool,
+					int fd, struct perf_evlist *evlist,
+					perf_event__handler_t process);
 int perf_event__process_tracing_data(union perf_event *event,
 				     struct perf_session *session);
 
-int perf_event__synthesize_build_id(struct dso *pos, u16 misc,
+int perf_event__synthesize_build_id(struct perf_tool *tool,
+				    struct dso *pos, u16 misc,
 				    perf_event__handler_t process,
-				    struct machine *machine,
-				    struct perf_session *session);
-int perf_event__process_build_id(union perf_event *event,
+				    struct machine *machine);
+int perf_event__process_build_id(struct perf_tool *tool,
+				 union perf_event *event,
 				 struct perf_session *session);
 
 /*

tools/perf/util/hist.h

@@ -117,7 +117,6 @@ int perf_evlist__tui_browse_hists(struct perf_evlist *evlist __used,
 
 static inline int hist_entry__tui_annotate(struct hist_entry *self __used,
 					   int evidx __used,
-					   int nr_events __used,
 					   void(*timer)(void *arg) __used,
 					   void *arg __used,
 					   int delay_secs __used)
@@ -128,7 +127,7 @@ static inline int hist_entry__tui_annotate(struct hist_entry *self __used,
 #define K_RIGHT -2
 #else
 #include "ui/keysyms.h"
-int hist_entry__tui_annotate(struct hist_entry *he, int evidx, int nr_events,
+int hist_entry__tui_annotate(struct hist_entry *he, int evidx,
 			     void(*timer)(void *arg), void *arg, int delay_secs);
 
 int perf_evlist__tui_browse_hists(struct perf_evlist *evlist, const char *help,

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

@@ -9,6 +9,17 @@
 #define BITS_PER_BYTE           8
 #define BITS_TO_LONGS(nr)       DIV_ROUND_UP(nr, BITS_PER_BYTE * sizeof(long))
 
+#define for_each_set_bit(bit, addr, size) \
+	for ((bit) = find_first_bit((addr), (size));		\
+	     (bit) < (size);					\
+	     (bit) = find_next_bit((addr), (size), (bit) + 1))
+
+/* same as for_each_set_bit() but use bit as value to start with */
+#define for_each_set_bit_cont(bit, addr, size) \
+	for ((bit) = find_next_bit((addr), (size), (bit));	\
+	     (bit) < (size);					\
+	     (bit) = find_next_bit((addr), (size), (bit) + 1))
+
 static inline void set_bit(int nr, unsigned long *addr)
 {
 	addr[nr / BITS_PER_LONG] |= 1UL << (nr % BITS_PER_LONG);
@@ -30,4 +41,111 @@ static inline unsigned long hweight_long(unsigned long w)
 	return sizeof(w) == 4 ? hweight32(w) : hweight64(w);
 }
 
+#define BITOP_WORD(nr)		((nr) / BITS_PER_LONG)
+
+/**
+ * __ffs - find first bit in word.
+ * @word: The word to search
+ *
+ * Undefined if no bit exists, so code should check against 0 first.
+ */
+static __always_inline unsigned long __ffs(unsigned long word)
+{
+	int num = 0;
+
+#if BITS_PER_LONG == 64
+	if ((word & 0xffffffff) == 0) {
+		num += 32;
+		word >>= 32;
+	}
+#endif
+	if ((word & 0xffff) == 0) {
+		num += 16;
+		word >>= 16;
+	}
+	if ((word & 0xff) == 0) {
+		num += 8;
+		word >>= 8;
+	}
+	if ((word & 0xf) == 0) {
+		num += 4;
+		word >>= 4;
+	}
+	if ((word & 0x3) == 0) {
+		num += 2;
+		word >>= 2;
+	}
+	if ((word & 0x1) == 0)
+		num += 1;
+	return num;
+}
+
+/*
+ * Find the first set bit in a memory region.
+ */
+static inline unsigned long
+find_first_bit(const unsigned long *addr, unsigned long size)
+{
+	const unsigned long *p = addr;
+	unsigned long result = 0;
+	unsigned long tmp;
+
+	while (size & ~(BITS_PER_LONG-1)) {
+		if ((tmp = *(p++)))
+			goto found;
+		result += BITS_PER_LONG;
+		size -= BITS_PER_LONG;
+	}
+	if (!size)
+		return result;
+
+	tmp = (*p) & (~0UL >> (BITS_PER_LONG - size));
+	if (tmp == 0UL)		/* Are any bits set? */
+		return result + size;	/* Nope. */
+found:
+	return result + __ffs(tmp);
+}
+
+/*
+ * Find the next set bit in a memory region.
+ */
+static inline unsigned long
+find_next_bit(const unsigned long *addr, unsigned long size, unsigned long offset)
+{
+	const unsigned long *p = addr + BITOP_WORD(offset);
+	unsigned long result = offset & ~(BITS_PER_LONG-1);
+	unsigned long tmp;
+
+	if (offset >= size)
+		return size;
+	size -= result;
+	offset %= BITS_PER_LONG;
+	if (offset) {
+		tmp = *(p++);
+		tmp &= (~0UL << offset);
+		if (size < BITS_PER_LONG)
+			goto found_first;
+		if (tmp)
+			goto found_middle;
+		size -= BITS_PER_LONG;
+		result += BITS_PER_LONG;
+	}
+	while (size & ~(BITS_PER_LONG-1)) {
+		if ((tmp = *(p++)))
+			goto found_middle;
+		result += BITS_PER_LONG;
+		size -= BITS_PER_LONG;
+	}
+	if (!size)
+		return result;
+	tmp = *p;
+
+found_first:
+	tmp &= (~0UL >> (BITS_PER_LONG - size));
+	if (tmp == 0UL)		/* Are any bits set? */
+		return result + size;	/* Nope. */
+found_middle:
+	return result + __ffs(tmp);
+}
+
 #endif

tools/perf/util/map.c

@@ -562,6 +562,10 @@ int machine__init(struct machine *self, const char *root_dir, pid_t pid)
 	INIT_LIST_HEAD(&self->user_dsos);
 	INIT_LIST_HEAD(&self->kernel_dsos);
 
+	self->threads = RB_ROOT;
+	INIT_LIST_HEAD(&self->dead_threads);
+	self->last_match = NULL;
+
 	self->kmaps.machine = self;
 	self->pid	    = pid;
 	self->root_dir      = strdup(root_dir);

tools/perf/util/map.h

@@ -18,9 +18,11 @@ enum map_type {
 extern const char *map_type__name[MAP__NR_TYPES];
 
 struct dso;
+struct ip_callchain;
 struct ref_reloc_sym;
 struct map_groups;
 struct machine;
+struct perf_evsel;
 
 struct map {
 	union {
@@ -61,7 +63,11 @@ struct map_groups {
 struct machine {
 	struct rb_node	  rb_node;
 	pid_t		  pid;
+	u16		  id_hdr_size;
 	char		  *root_dir;
+	struct rb_root	  threads;
+	struct list_head  dead_threads;
+	struct thread	  *last_match;
 	struct list_head  user_dsos;
 	struct list_head  kernel_dsos;
 	struct map_groups kmaps;
@@ -148,6 +154,13 @@ int machine__init(struct machine *self, const char *root_dir, pid_t pid);
 void machine__exit(struct machine *self);
 void machine__delete(struct machine *self);
 
+int machine__resolve_callchain(struct machine *machine,
+			       struct perf_evsel *evsel, struct thread *thread,
+			       struct ip_callchain *chain,
+			       struct symbol **parent);
+int maps__set_kallsyms_ref_reloc_sym(struct map **maps, const char *symbol_name,
+				     u64 addr);
+
 /*
  * Default guest kernel is defined by parameter --guestkallsyms
  * and --guestmodules
@@ -190,6 +203,12 @@ struct symbol *map_groups__find_symbol_by_name(struct map_groups *mg,
 					       struct map **mapp,
 					       symbol_filter_t filter);
 
+
+struct thread *machine__findnew_thread(struct machine *machine, pid_t pid);
+void machine__remove_thread(struct machine *machine, struct thread *th);
+
+size_t machine__fprintf(struct machine *machine, FILE *fp);
+
 static inline
 struct symbol *machine__find_kernel_symbol(struct machine *self,
 					   enum map_type type, u64 addr,

tools/perf/util/parse-events.c

@@ -25,8 +25,6 @@ enum event_result {
 	EVT_HANDLED_ALL
 };
 
-char debugfs_path[MAXPATHLEN];
-
 #define CHW(x) .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_##x
 #define CSW(x) .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_##x
 
@@ -40,6 +38,7 @@ static struct event_symbol event_symbols[] = {
   { CHW(BRANCH_INSTRUCTIONS),		"branch-instructions",		"branches"		},
   { CHW(BRANCH_MISSES),			"branch-misses",		""			},
   { CHW(BUS_CYCLES),			"bus-cycles",			""			},
+  { CHW(REF_CPU_CYCLES),		"ref-cycles",			""			},
 
   { CSW(CPU_CLOCK),			"cpu-clock",			""			},
   { CSW(TASK_CLOCK),			"task-clock",			""			},
@@ -70,6 +69,7 @@ static const char *hw_event_names[PERF_COUNT_HW_MAX] = {
 	"bus-cycles",
 	"stalled-cycles-frontend",
 	"stalled-cycles-backend",
+	"ref-cycles",
 };
 
 static const char *sw_event_names[PERF_COUNT_SW_MAX] = {
@@ -140,7 +140,7 @@ static int tp_event_has_id(struct dirent *sys_dir, struct dirent *evt_dir)
 	char evt_path[MAXPATHLEN];
 	int fd;
 
-	snprintf(evt_path, MAXPATHLEN, "%s/%s/%s/id", debugfs_path,
+	snprintf(evt_path, MAXPATHLEN, "%s/%s/%s/id", tracing_events_path,
 			sys_dir->d_name, evt_dir->d_name);
 	fd = open(evt_path, O_RDONLY);
 	if (fd < 0)
@@ -171,16 +171,16 @@ struct tracepoint_path *tracepoint_id_to_path(u64 config)
 	char evt_path[MAXPATHLEN];
 	char dir_path[MAXPATHLEN];
 
-	if (debugfs_valid_mountpoint(debugfs_path))
+	if (debugfs_valid_mountpoint(tracing_events_path))
 		return NULL;
 
-	sys_dir = opendir(debugfs_path);
+	sys_dir = opendir(tracing_events_path);
 	if (!sys_dir)
 		return NULL;
 
 	for_each_subsystem(sys_dir, sys_dirent, sys_next) {
 
-		snprintf(dir_path, MAXPATHLEN, "%s/%s", debugfs_path,
+		snprintf(dir_path, MAXPATHLEN, "%s/%s", tracing_events_path,
 			 sys_dirent.d_name);
 		evt_dir = opendir(dir_path);
 		if (!evt_dir)
@@ -447,7 +447,7 @@ parse_single_tracepoint_event(char *sys_name,
 	u64 id;
 	int fd;
 
-	snprintf(evt_path, MAXPATHLEN, "%s/%s/%s/id", debugfs_path,
+	snprintf(evt_path, MAXPATHLEN, "%s/%s/%s/id", tracing_events_path,
 		 sys_name, evt_name);
 
 	fd = open(evt_path, O_RDONLY);
@@ -485,7 +485,7 @@ parse_multiple_tracepoint_event(struct perf_evlist *evlist, char *sys_name,
 	struct dirent *evt_ent;
 	DIR *evt_dir;
 
-	snprintf(evt_path, MAXPATHLEN, "%s/%s", debugfs_path, sys_name);
+	snprintf(evt_path, MAXPATHLEN, "%s/%s", tracing_events_path, sys_name);
 	evt_dir = opendir(evt_path);
 
 	if (!evt_dir) {
@@ -528,7 +528,7 @@ parse_tracepoint_event(struct perf_evlist *evlist, const char **strp,
 	char sys_name[MAX_EVENT_LENGTH];
 	unsigned int sys_length, evt_length;
 
-	if (debugfs_valid_mountpoint(debugfs_path))
+	if (debugfs_valid_mountpoint(tracing_events_path))
 		return 0;
 
 	evt_name = strchr(*strp, ':');
@@ -920,10 +920,10 @@ void print_tracepoint_events(const char *subsys_glob, const char *event_glob)
 	char evt_path[MAXPATHLEN];
 	char dir_path[MAXPATHLEN];
 
-	if (debugfs_valid_mountpoint(debugfs_path))
+	if (debugfs_valid_mountpoint(tracing_events_path))
 		return;
 
-	sys_dir = opendir(debugfs_path);
+	sys_dir = opendir(tracing_events_path);
 	if (!sys_dir)
 		return;
 
@@ -932,7 +932,7 @@ void print_tracepoint_events(const char *subsys_glob, const char *event_glob)
 		    !strglobmatch(sys_dirent.d_name, subsys_glob))
 			continue;
 
-		snprintf(dir_path, MAXPATHLEN, "%s/%s", debugfs_path,
+		snprintf(dir_path, MAXPATHLEN, "%s/%s", tracing_events_path,
 			 sys_dirent.d_name);
 		evt_dir = opendir(dir_path);
 		if (!evt_dir)
@@ -964,16 +964,16 @@ int is_valid_tracepoint(const char *event_string)
 	char evt_path[MAXPATHLEN];
 	char dir_path[MAXPATHLEN];
 
-	if (debugfs_valid_mountpoint(debugfs_path))
+	if (debugfs_valid_mountpoint(tracing_events_path))
 		return 0;
 
-	sys_dir = opendir(debugfs_path);
+	sys_dir = opendir(tracing_events_path);
 	if (!sys_dir)
 		return 0;
 
 	for_each_subsystem(sys_dir, sys_dirent, sys_next) {
 
-		snprintf(dir_path, MAXPATHLEN, "%s/%s", debugfs_path,
+		snprintf(dir_path, MAXPATHLEN, "%s/%s", tracing_events_path,
 			 sys_dirent.d_name);
 		evt_dir = opendir(dir_path);
 		if (!evt_dir)

tools/perf/util/parse-events.h

@@ -39,7 +39,6 @@ void print_tracepoint_events(const char *subsys_glob, const char *event_glob);
 int print_hwcache_events(const char *event_glob);
 extern int is_valid_tracepoint(const char *event_string);
 
-extern char debugfs_path[];
 extern int valid_debugfs_mount(const char *debugfs);
 
 #endif /* __PERF_PARSE_EVENTS_H */

tools/perf/util/probe-finder.h

@@ -5,7 +5,6 @@
 #include "util.h"
 #include "probe-event.h"
 
-#define MAX_PATH_LEN		 256
 #define MAX_PROBE_BUFFER	1024
 #define MAX_PROBES		 128
 

tools/perf/util/scripting-engines/trace-event-perl.c

@@ -27,7 +27,10 @@
 
 #include "../../perf.h"
 #include "../util.h"
+#include "../thread.h"
+#include "../event.h"
 #include "../trace-event.h"
+#include "../evsel.h"
 
 #include <EXTERN.h>
 #include <perl.h>
@@ -245,11 +248,11 @@ static inline struct event *find_cache_event(int type)
 	return event;
 }
 
-static void perl_process_event(union perf_event *pevent __unused,
-			       struct perf_sample *sample,
-			       struct perf_evsel *evsel,
-			       struct perf_session *session __unused,
-			       struct thread *thread)
+static void perl_process_tracepoint(union perf_event *pevent __unused,
+				    struct perf_sample *sample,
+				    struct perf_evsel *evsel,
+				    struct machine *machine __unused,
+				    struct thread *thread)
 {
 	struct format_field *field;
 	static char handler[256];
@@ -265,6 +268,9 @@ static void perl_process_event(union perf_event *pevent __unused,
 
 	dSP;
 
+	if (evsel->attr.type != PERF_TYPE_TRACEPOINT)
+		return;
+
 	type = trace_parse_common_type(data);
 
 	event = find_cache_event(type);
@@ -332,6 +338,42 @@ static void perl_process_event(union perf_event *pevent __unused,
 	LEAVE;
 }
 
+static void perl_process_event_generic(union perf_event *pevent __unused,
+				       struct perf_sample *sample,
+				       struct perf_evsel *evsel __unused,
+				       struct machine *machine __unused,
+				       struct thread *thread __unused)
+{
+	dSP;
+
+	if (!get_cv("process_event", 0))
+		return;
+
+	ENTER;
+	SAVETMPS;
+	PUSHMARK(SP);
+	XPUSHs(sv_2mortal(newSVpvn((const char *)pevent, pevent->header.size)));
+	XPUSHs(sv_2mortal(newSVpvn((const char *)&evsel->attr, sizeof(evsel->attr))));
+	XPUSHs(sv_2mortal(newSVpvn((const char *)sample, sizeof(*sample))));
+	XPUSHs(sv_2mortal(newSVpvn((const char *)sample->raw_data, sample->raw_size)));
+	PUTBACK;
+	call_pv("process_event", G_SCALAR);
+	SPAGAIN;
+	PUTBACK;
+	FREETMPS;
+	LEAVE;
+}
+
+static void perl_process_event(union perf_event *pevent,
+			       struct perf_sample *sample,
+			       struct perf_evsel *evsel,
+			       struct machine *machine,
+			       struct thread *thread)
+{
+	perl_process_tracepoint(pevent, sample, evsel, machine, thread);
+	perl_process_event_generic(pevent, sample, evsel, machine, thread);
+}
+
 static void run_start_sub(void)
 {
 	dSP; /* access to Perl stack */
@@ -553,7 +595,28 @@ static int perl_generate_script(const char *outfile)
 	fprintf(ofp, "sub print_header\n{\n"
 		"\tmy ($event_name, $cpu, $secs, $nsecs, $pid, $comm) = @_;\n\n"
 		"\tprintf(\"%%-20s %%5u %%05u.%%09u %%8u %%-20s \",\n\t       "
-		"$event_name, $cpu, $secs, $nsecs, $pid, $comm);\n}");
+		"$event_name, $cpu, $secs, $nsecs, $pid, $comm);\n}\n");
+
+	fprintf(ofp,
+		"\n# Packed byte string args of process_event():\n"
+		"#\n"
+		"# $event:\tunion perf_event\tutil/event.h\n"
+		"# $attr:\tstruct perf_event_attr\tlinux/perf_event.h\n"
+		"# $sample:\tstruct perf_sample\tutil/event.h\n"
+		"# $raw_data:\tperf_sample->raw_data\tutil/event.h\n"
+		"\n"
+		"sub process_event\n"
+		"{\n"
+		"\tmy ($event, $attr, $sample, $raw_data) = @_;\n"
+		"\n"
+		"\tmy @event\t= unpack(\"LSS\", $event);\n"
+		"\tmy @attr\t= unpack(\"LLQQQQQLLQQ\", $attr);\n"
+		"\tmy @sample\t= unpack(\"QLLQQQQQLL\", $sample);\n"
+		"\tmy @raw_data\t= unpack(\"C*\", $raw_data);\n"
+		"\n"
+		"\tuse Data::Dumper;\n"
+		"\tprint Dumper \\@event, \\@attr, \\@sample, \\@raw_data;\n"
+		"}\n");
 
 	fclose(ofp);
 

tools/perf/util/scripting-engines/trace-event-python.c

@@ -29,6 +29,8 @@
 
 #include "../../perf.h"
 #include "../util.h"
+#include "../event.h"
+#include "../thread.h"
 #include "../trace-event.h"
 
 PyMODINIT_FUNC initperf_trace_context(void);
@@ -207,7 +209,7 @@ static inline struct event *find_cache_event(int type)
 static void python_process_event(union perf_event *pevent __unused,
 				 struct perf_sample *sample,
 				 struct perf_evsel *evsel __unused,
-				 struct perf_session *session __unused,
+				 struct machine *machine __unused,
 				 struct thread *thread)
 {
 	PyObject *handler, *retval, *context, *t, *obj, *dict = NULL;

tools/perf/util/session.c

@@ -10,6 +10,7 @@
 #include "evlist.h"
 #include "evsel.h"
 #include "session.h"
+#include "tool.h"
 #include "sort.h"
 #include "util.h"
 #include "cpumap.h"
@@ -78,39 +79,13 @@ out_close:
 	return -1;
 }
 
-static void perf_session__id_header_size(struct perf_session *session)
-{
-       struct perf_sample *data;
-       u64 sample_type = session->sample_type;
-       u16 size = 0;
-
-	if (!session->sample_id_all)
-		goto out;
-
-       if (sample_type & PERF_SAMPLE_TID)
-               size += sizeof(data->tid) * 2;
-
-       if (sample_type & PERF_SAMPLE_TIME)
-               size += sizeof(data->time);
-
-       if (sample_type & PERF_SAMPLE_ID)
-               size += sizeof(data->id);
-
-       if (sample_type & PERF_SAMPLE_STREAM_ID)
-               size += sizeof(data->stream_id);
-
-       if (sample_type & PERF_SAMPLE_CPU)
-               size += sizeof(data->cpu) * 2;
-out:
-       session->id_hdr_size = size;
-}
-
 void perf_session__update_sample_type(struct perf_session *self)
 {
 	self->sample_type = perf_evlist__sample_type(self->evlist);
 	self->sample_size = __perf_evsel__sample_size(self->sample_type);
 	self->sample_id_all = perf_evlist__sample_id_all(self->evlist);
-	perf_session__id_header_size(self);
+	self->id_hdr_size = perf_evlist__id_hdr_size(self->evlist);
+	self->host_machine.id_hdr_size = self->id_hdr_size;
 }
 
 int perf_session__create_kernel_maps(struct perf_session *self)
@@ -130,18 +105,26 @@ static void perf_session__destroy_kernel_maps(struct perf_session *self)
 
 struct perf_session *perf_session__new(const char *filename, int mode,
 				       bool force, bool repipe,
-				       struct perf_event_ops *ops)
+				       struct perf_tool *tool)
 {
-	size_t len = filename ? strlen(filename) + 1 : 0;
-	struct perf_session *self = zalloc(sizeof(*self) + len);
+	struct perf_session *self;
+	struct stat st;
+	size_t len;
+
+	if (!filename || !strlen(filename)) {
+		if (!fstat(STDIN_FILENO, &st) && S_ISFIFO(st.st_mode))
+			filename = "-";
+		else
+			filename = "perf.data";
+	}
+
+	len = strlen(filename);
+	self = zalloc(sizeof(*self) + len);
 
 	if (self == NULL)
 		goto out;
 
 	memcpy(self->filename, filename, len);
-	self->threads = RB_ROOT;
-	INIT_LIST_HEAD(&self->dead_threads);
-	self->last_match = NULL;
 	/*
 	 * On 64bit we can mmap the data file in one go. No need for tiny mmap
 	 * slices. On 32bit we use 32MB.
@@ -171,10 +154,10 @@ struct perf_session *perf_session__new(const char *filename, int mode,
 			goto out_delete;
 	}
 
-	if (ops && ops->ordering_requires_timestamps &&
-	    ops->ordered_samples && !self->sample_id_all) {
+	if (tool && tool->ordering_requires_timestamps &&
+	    tool->ordered_samples && !self->sample_id_all) {
 		dump_printf("WARNING: No sample_id_all support, falling back to unordered processing\n");
-		ops->ordered_samples = false;
+		tool->ordered_samples = false;
 	}
 
 out:
@@ -184,17 +167,22 @@ out_delete:
 	return NULL;
 }
 
-static void perf_session__delete_dead_threads(struct perf_session *self)
+static void machine__delete_dead_threads(struct machine *machine)
 {
 	struct thread *n, *t;
 
-	list_for_each_entry_safe(t, n, &self->dead_threads, node) {
+	list_for_each_entry_safe(t, n, &machine->dead_threads, node) {
 		list_del(&t->node);
 		thread__delete(t);
 	}
 }
 
-static void perf_session__delete_threads(struct perf_session *self)
+static void perf_session__delete_dead_threads(struct perf_session *session)
+{
+	machine__delete_dead_threads(&session->host_machine);
+}
+
+static void machine__delete_threads(struct machine *self)
 {
 	struct rb_node *nd = rb_first(&self->threads);
 
@@ -207,6 +195,11 @@ static void perf_session__delete_threads(struct perf_session *self)
 	}
 }
 
+static void perf_session__delete_threads(struct perf_session *session)
+{
+	machine__delete_threads(&session->host_machine);
+}
+
 void perf_session__delete(struct perf_session *self)
 {
 	perf_session__destroy_kernel_maps(self);
@@ -217,7 +210,7 @@ void perf_session__delete(struct perf_session *self)
 	free(self);
 }
 
-void perf_session__remove_thread(struct perf_session *self, struct thread *th)
+void machine__remove_thread(struct machine *self, struct thread *th)
 {
 	self->last_match = NULL;
 	rb_erase(&th->rb_node, &self->threads);
@@ -236,16 +229,16 @@ static bool symbol__match_parent_regex(struct symbol *sym)
 	return 0;
 }
 
-int perf_session__resolve_callchain(struct perf_session *self,
-				    struct thread *thread,
-				    struct ip_callchain *chain,
-				    struct symbol **parent)
+int machine__resolve_callchain(struct machine *self, struct perf_evsel *evsel,
+			       struct thread *thread,
+			       struct ip_callchain *chain,
+			       struct symbol **parent)
 {
 	u8 cpumode = PERF_RECORD_MISC_USER;
 	unsigned int i;
 	int err;
 
-	callchain_cursor_reset(&self->callchain_cursor);
+	callchain_cursor_reset(&evsel->hists.callchain_cursor);
 
 	for (i = 0; i < chain->nr; i++) {
 		u64 ip;
@@ -272,7 +265,7 @@ int perf_session__resolve_callchain(struct perf_session *self,
 
 		al.filtered = false;
 		thread__find_addr_location(thread, self, cpumode,
-				MAP__FUNCTION, thread->pid, ip, &al, NULL);
+					   MAP__FUNCTION, ip, &al, NULL);
 		if (al.sym != NULL) {
 			if (sort__has_parent && !*parent &&
 			    symbol__match_parent_regex(al.sym))
@@ -281,7 +274,7 @@ int perf_session__resolve_callchain(struct perf_session *self,
 				break;
 		}
 
-		err = callchain_cursor_append(&self->callchain_cursor,
+		err = callchain_cursor_append(&evsel->hists.callchain_cursor,
 					      ip, al.map, al.sym);
 		if (err)
 			return err;
@@ -290,75 +283,91 @@ int perf_session__resolve_callchain(struct perf_session *self,
 	return 0;
 }
 
-static int process_event_synth_stub(union perf_event *event __used,
-				    struct perf_session *session __used)
+static int process_event_synth_tracing_data_stub(union perf_event *event __used,
+						 struct perf_session *session __used)
+{
+	dump_printf(": unhandled!\n");
+	return 0;
+}
+
+static int process_event_synth_attr_stub(union perf_event *event __used,
+					 struct perf_evlist **pevlist __used)
 {
 	dump_printf(": unhandled!\n");
 	return 0;
 }
 
-static int process_event_sample_stub(union perf_event *event __used,
+static int process_event_sample_stub(struct perf_tool *tool __used,
+				     union perf_event *event __used,
 				     struct perf_sample *sample __used,
 				     struct perf_evsel *evsel __used,
-				     struct perf_session *session __used)
+				     struct machine *machine __used)
 {
 	dump_printf(": unhandled!\n");
 	return 0;
 }
 
-static int process_event_stub(union perf_event *event __used,
+static int process_event_stub(struct perf_tool *tool __used,
+			      union perf_event *event __used,
 			      struct perf_sample *sample __used,
-			      struct perf_session *session __used)
+			      struct machine *machine __used)
 {
 	dump_printf(": unhandled!\n");
 	return 0;
 }
 
-static int process_finished_round_stub(union perf_event *event __used,
-				       struct perf_session *session __used,
-				       struct perf_event_ops *ops __used)
+static int process_finished_round_stub(struct perf_tool *tool __used,
+				       union perf_event *event __used,
+				       struct perf_session *perf_session __used)
 {
 	dump_printf(": unhandled!\n");
 	return 0;
 }
 
-static int process_finished_round(union perf_event *event,
-				  struct perf_session *session,
-				  struct perf_event_ops *ops);
+static int process_event_type_stub(struct perf_tool *tool __used,
+				   union perf_event *event __used)
+{
+	dump_printf(": unhandled!\n");
+	return 0;
+}
 
-static void perf_event_ops__fill_defaults(struct perf_event_ops *handler)
+static int process_finished_round(struct perf_tool *tool,
+				  union perf_event *event,
+				  struct perf_session *session);
+
+static void perf_tool__fill_defaults(struct perf_tool *tool)
 {
-	if (handler->sample == NULL)
-		handler->sample = process_event_sample_stub;
-	if (handler->mmap == NULL)
-		handler->mmap = process_event_stub;
-	if (handler->comm == NULL)
-		handler->comm = process_event_stub;
-	if (handler->fork == NULL)
-		handler->fork = process_event_stub;
-	if (handler->exit == NULL)
-		handler->exit = process_event_stub;
-	if (handler->lost == NULL)
-		handler->lost = perf_event__process_lost;
-	if (handler->read == NULL)
-		handler->read = process_event_stub;
-	if (handler->throttle == NULL)
-		handler->throttle = process_event_stub;
-	if (handler->unthrottle == NULL)
-		handler->unthrottle = process_event_stub;
-	if (handler->attr == NULL)
-		handler->attr = process_event_synth_stub;
-	if (handler->event_type == NULL)
-		handler->event_type = process_event_synth_stub;
-	if (handler->tracing_data == NULL)
-		handler->tracing_data = process_event_synth_stub;
-	if (handler->build_id == NULL)
-		handler->build_id = process_event_synth_stub;
-	if (handler->finished_round == NULL) {
-		if (handler->ordered_samples)
-			handler->finished_round = process_finished_round;
+	if (tool->sample == NULL)
+		tool->sample = process_event_sample_stub;
+	if (tool->mmap == NULL)
+		tool->mmap = process_event_stub;
+	if (tool->comm == NULL)
+		tool->comm = process_event_stub;
+	if (tool->fork == NULL)
+		tool->fork = process_event_stub;
+	if (tool->exit == NULL)
+		tool->exit = process_event_stub;
+	if (tool->lost == NULL)
+		tool->lost = perf_event__process_lost;
+	if (tool->read == NULL)
+		tool->read = process_event_sample_stub;
+	if (tool->throttle == NULL)
+		tool->throttle = process_event_stub;
+	if (tool->unthrottle == NULL)
+		tool->unthrottle = process_event_stub;
+	if (tool->attr == NULL)
+		tool->attr = process_event_synth_attr_stub;
+	if (tool->event_type == NULL)
+		tool->event_type = process_event_type_stub;
+	if (tool->tracing_data == NULL)
+		tool->tracing_data = process_event_synth_tracing_data_stub;
+	if (tool->build_id == NULL)
+		tool->build_id = process_finished_round_stub;
+	if (tool->finished_round == NULL) {
+		if (tool->ordered_samples)
+			tool->finished_round = process_finished_round;
 		else
-			handler->finished_round = process_finished_round_stub;
+			tool->finished_round = process_finished_round_stub;
 	}
 }
 
@@ -490,11 +499,11 @@ static void perf_session_free_sample_buffers(struct perf_session *session)
 static int perf_session_deliver_event(struct perf_session *session,
 				      union perf_event *event,
 				      struct perf_sample *sample,
-				      struct perf_event_ops *ops,
+				      struct perf_tool *tool,
 				      u64 file_offset);
 
 static void flush_sample_queue(struct perf_session *s,
-			       struct perf_event_ops *ops)
+			       struct perf_tool *tool)
 {
 	struct ordered_samples *os = &s->ordered_samples;
 	struct list_head *head = &os->samples;
@@ -505,7 +514,7 @@ static void flush_sample_queue(struct perf_session *s,
 	unsigned idx = 0, progress_next = os->nr_samples / 16;
 	int ret;
 
-	if (!ops->ordered_samples || !limit)
+	if (!tool->ordered_samples || !limit)
 		return;
 
 	list_for_each_entry_safe(iter, tmp, head, list) {
@@ -516,7 +525,7 @@ static void flush_sample_queue(struct perf_session *s,
 		if (ret)
 			pr_err("Can't parse sample, err = %d\n", ret);
 		else
-			perf_session_deliver_event(s, iter->event, &sample, ops,
+			perf_session_deliver_event(s, iter->event, &sample, tool,
 						   iter->file_offset);
 
 		os->last_flush = iter->timestamp;
@@ -578,11 +587,11 @@ static void flush_sample_queue(struct perf_session *s,
  *      Flush every events below timestamp 7
  *      etc...
  */
-static int process_finished_round(union perf_event *event __used,
-				  struct perf_session *session,
-				  struct perf_event_ops *ops)
+static int process_finished_round(struct perf_tool *tool,
+				  union perf_event *event __used,
+				  struct perf_session *session)
 {
-	flush_sample_queue(session, ops);
+	flush_sample_queue(session, tool);
 	session->ordered_samples.next_flush = session->ordered_samples.max_timestamp;
 
 	return 0;
@@ -737,13 +746,26 @@ static void dump_sample(struct perf_session *session, union perf_event *event,
 		callchain__printf(sample);
 }
 
+static struct machine *
+	perf_session__find_machine_for_cpumode(struct perf_session *session,
+					       union perf_event *event)
+{
+	const u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
+
+	if (cpumode == PERF_RECORD_MISC_GUEST_KERNEL && perf_guest)
+		return perf_session__find_machine(session, event->ip.pid);
+
+	return perf_session__find_host_machine(session);
+}
+
 static int perf_session_deliver_event(struct perf_session *session,
 				      union perf_event *event,
 				      struct perf_sample *sample,
-				      struct perf_event_ops *ops,
+				      struct perf_tool *tool,
 				      u64 file_offset)
 {
 	struct perf_evsel *evsel;
+	struct machine *machine;
 
 	dump_event(session, event, file_offset, sample);
 
@@ -765,6 +787,8 @@ static int perf_session_deliver_event(struct perf_session *session,
 		hists__inc_nr_events(&evsel->hists, event->header.type);
 	}
 
+	machine = perf_session__find_machine_for_cpumode(session, event);
+
 	switch (event->header.type) {
 	case PERF_RECORD_SAMPLE:
 		dump_sample(session, event, sample);
@@ -772,23 +796,25 @@ static int perf_session_deliver_event(struct perf_session *session,
 			++session->hists.stats.nr_unknown_id;
 			return -1;
 		}
-		return ops->sample(event, sample, evsel, session);
+		return tool->sample(tool, event, sample, evsel, machine);
 	case PERF_RECORD_MMAP:
-		return ops->mmap(event, sample, session);
+		return tool->mmap(tool, event, sample, machine);
 	case PERF_RECORD_COMM:
-		return ops->comm(event, sample, session);
+		return tool->comm(tool, event, sample, machine);
 	case PERF_RECORD_FORK:
-		return ops->fork(event, sample, session);
+		return tool->fork(tool, event, sample, machine);
 	case PERF_RECORD_EXIT:
-		return ops->exit(event, sample, session);
+		return tool->exit(tool, event, sample, machine);
 	case PERF_RECORD_LOST:
-		return ops->lost(event, sample, session);
+		if (tool->lost == perf_event__process_lost)
+			session->hists.stats.total_lost += event->lost.lost;
+		return tool->lost(tool, event, sample, machine);
 	case PERF_RECORD_READ:
-		return ops->read(event, sample, session);
+		return tool->read(tool, event, sample, evsel, machine);
 	case PERF_RECORD_THROTTLE:
-		return ops->throttle(event, sample, session);
+		return tool->throttle(tool, event, sample, machine);
 	case PERF_RECORD_UNTHROTTLE:
-		return ops->unthrottle(event, sample, session);
+		return tool->unthrottle(tool, event, sample, machine);
 	default:
 		++session->hists.stats.nr_unknown_events;
 		return -1;
@@ -812,24 +838,29 @@ static int perf_session__preprocess_sample(struct perf_session *session,
 }
 
 static int perf_session__process_user_event(struct perf_session *session, union perf_event *event,
-					    struct perf_event_ops *ops, u64 file_offset)
+					    struct perf_tool *tool, u64 file_offset)
 {
+	int err;
+
 	dump_event(session, event, file_offset, NULL);
 
 	/* These events are processed right away */
 	switch (event->header.type) {
 	case PERF_RECORD_HEADER_ATTR:
-		return ops->attr(event, session);
+		err = tool->attr(event, &session->evlist);
+		if (err == 0)
+			perf_session__update_sample_type(session);
+		return err;
 	case PERF_RECORD_HEADER_EVENT_TYPE:
-		return ops->event_type(event, session);
+		return tool->event_type(tool, event);
 	case PERF_RECORD_HEADER_TRACING_DATA:
 		/* setup for reading amidst mmap */
 		lseek(session->fd, file_offset, SEEK_SET);
-		return ops->tracing_data(event, session);
+		return tool->tracing_data(event, session);
 	case PERF_RECORD_HEADER_BUILD_ID:
-		return ops->build_id(event, session);
+		return tool->build_id(tool, event, session);
 	case PERF_RECORD_FINISHED_ROUND:
-		return ops->finished_round(event, session, ops);
+		return tool->finished_round(tool, event, session);
 	default:
 		return -EINVAL;
 	}
@@ -837,7 +868,7 @@ static int perf_session__process_user_event(struct perf_session *session, union
 
 static int perf_session__process_event(struct perf_session *session,
 				       union perf_event *event,
-				       struct perf_event_ops *ops,
+				       struct perf_tool *tool,
 				       u64 file_offset)
 {
 	struct perf_sample sample;
@@ -853,7 +884,7 @@ static int perf_session__process_event(struct perf_session *session,
 	hists__inc_nr_events(&session->hists, event->header.type);
 
 	if (event->header.type >= PERF_RECORD_USER_TYPE_START)
-		return perf_session__process_user_event(session, event, ops, file_offset);
+		return perf_session__process_user_event(session, event, tool, file_offset);
 
 	/*
 	 * For all kernel events we get the sample data
@@ -866,14 +897,14 @@ static int perf_session__process_event(struct perf_session *session,
 	if (perf_session__preprocess_sample(session, event, &sample))
 		return 0;
 
-	if (ops->ordered_samples) {
+	if (tool->ordered_samples) {
 		ret = perf_session_queue_event(session, event, &sample,
 					       file_offset);
 		if (ret != -ETIME)
 			return ret;
 	}
 
-	return perf_session_deliver_event(session, event, &sample, ops,
+	return perf_session_deliver_event(session, event, &sample, tool,
 					  file_offset);
 }
 
@@ -884,6 +915,11 @@ void perf_event_header__bswap(struct perf_event_header *self)
 	self->size = bswap_16(self->size);
 }
 
+struct thread *perf_session__findnew(struct perf_session *session, pid_t pid)
+{
+	return machine__findnew_thread(&session->host_machine, pid);
+}
+
 static struct thread *perf_session__register_idle_thread(struct perf_session *self)
 {
 	struct thread *thread = perf_session__findnew(self, 0);
@@ -897,9 +933,9 @@ static struct thread *perf_session__register_idle_thread(struct perf_session *se
 }
 
 static void perf_session__warn_about_errors(const struct perf_session *session,
-					    const struct perf_event_ops *ops)
+					    const struct perf_tool *tool)
 {
-	if (ops->lost == perf_event__process_lost &&
+	if (tool->lost == perf_event__process_lost &&
 	    session->hists.stats.nr_events[PERF_RECORD_LOST] != 0) {
 		ui__warning("Processed %d events and lost %d chunks!\n\n"
 			    "Check IO/CPU overload!\n\n",
@@ -934,7 +970,7 @@ static void perf_session__warn_about_errors(const struct perf_session *session,
 volatile int session_done;
 
 static int __perf_session__process_pipe_events(struct perf_session *self,
-					       struct perf_event_ops *ops)
+					       struct perf_tool *tool)
 {
 	union perf_event event;
 	uint32_t size;
@@ -943,7 +979,7 @@ static int __perf_session__process_pipe_events(struct perf_session *self,
 	int err;
 	void *p;
 
-	perf_event_ops__fill_defaults(ops);
+	perf_tool__fill_defaults(tool);
 
 	head = 0;
 more:
@@ -979,8 +1015,7 @@ more:
 		}
 	}
 
-	if (size == 0 ||
-	    (skip = perf_session__process_event(self, &event, ops, head)) < 0) {
+	if ((skip = perf_session__process_event(self, &event, tool, head)) < 0) {
 		dump_printf("%#" PRIx64 " [%#x]: skipping unknown header type: %d\n",
 			    head, event.header.size, event.header.type);
 		/*
@@ -1003,7 +1038,7 @@ more:
 done:
 	err = 0;
 out_err:
-	perf_session__warn_about_errors(self, ops);
+	perf_session__warn_about_errors(self, tool);
 	perf_session_free_sample_buffers(self);
 	return err;
 }
@@ -1034,7 +1069,7 @@ fetch_mmaped_event(struct perf_session *session,
 
 int __perf_session__process_events(struct perf_session *session,
 				   u64 data_offset, u64 data_size,
-				   u64 file_size, struct perf_event_ops *ops)
+				   u64 file_size, struct perf_tool *tool)
 {
 	u64 head, page_offset, file_offset, file_pos, progress_next;
 	int err, mmap_prot, mmap_flags, map_idx = 0;
@@ -1043,7 +1078,7 @@ int __perf_session__process_events(struct perf_session *session,
 	union perf_event *event;
 	uint32_t size;
 
-	perf_event_ops__fill_defaults(ops);
+	perf_tool__fill_defaults(tool);
 
 	page_size = sysconf(_SC_PAGESIZE);
 
@@ -1098,7 +1133,7 @@ more:
 	size = event->header.size;
 
 	if (size == 0 ||
-	    perf_session__process_event(session, event, ops, file_pos) < 0) {
+	    perf_session__process_event(session, event, tool, file_pos) < 0) {
 		dump_printf("%#" PRIx64 " [%#x]: skipping unknown header type: %d\n",
 			    file_offset + head, event->header.size,
 			    event->header.type);
@@ -1127,15 +1162,15 @@ more:
 	err = 0;
 	/* do the final flush for ordered samples */
 	session->ordered_samples.next_flush = ULLONG_MAX;
-	flush_sample_queue(session, ops);
+	flush_sample_queue(session, tool);
 out_err:
-	perf_session__warn_about_errors(session, ops);
+	perf_session__warn_about_errors(session, tool);
 	perf_session_free_sample_buffers(session);
 	return err;
 }
 
 int perf_session__process_events(struct perf_session *self,
-				 struct perf_event_ops *ops)
+				 struct perf_tool *tool)
 {
 	int err;
 
@@ -1146,9 +1181,9 @@ int perf_session__process_events(struct perf_session *self,
 		err = __perf_session__process_events(self,
 						     self->header.data_offset,
 						     self->header.data_size,
-						     self->size, ops);
+						     self->size, tool);
 	else
-		err = __perf_session__process_pipe_events(self, ops);
+		err = __perf_session__process_pipe_events(self, tool);
 
 	return err;
 }
@@ -1163,9 +1198,8 @@ bool perf_session__has_traces(struct perf_session *self, const char *msg)
 	return true;
 }
 
-int perf_session__set_kallsyms_ref_reloc_sym(struct map **maps,
-					     const char *symbol_name,
-					     u64 addr)
+int maps__set_kallsyms_ref_reloc_sym(struct map **maps,
+				     const char *symbol_name, u64 addr)
 {
 	char *bracket;
 	enum map_type i;
@@ -1224,6 +1258,27 @@ size_t perf_session__fprintf_nr_events(struct perf_session *session, FILE *fp)
 	return ret;
 }
 
+size_t perf_session__fprintf(struct perf_session *session, FILE *fp)
+{
+	/*
+	 * FIXME: Here we have to actually print all the machines in this
+	 * session, not just the host...
+	 */
+	return machine__fprintf(&session->host_machine, fp);
+}
+
+void perf_session__remove_thread(struct perf_session *session,
+				 struct thread *th)
+{
+	/*
+	 * FIXME: This one makes no sense, we need to remove the thread from
+	 * the machine it belongs to, perf_session can have many machines, so
+	 * doing it always on ->host_machine is wrong.  Fix when auditing all
+	 * the 'perf kvm' code.
+	 */
+	machine__remove_thread(&session->host_machine, th);
+}
+
 struct perf_evsel *perf_session__find_first_evtype(struct perf_session *session,
 					      unsigned int type)
 {
@@ -1236,17 +1291,16 @@ struct perf_evsel *perf_session__find_first_evtype(struct perf_session *session,
 	return NULL;
 }
 
-void perf_session__print_ip(union perf_event *event,
-			    struct perf_sample *sample,
-			    struct perf_session *session,
-			    int print_sym, int print_dso)
+void perf_event__print_ip(union perf_event *event, struct perf_sample *sample,
+			  struct machine *machine, struct perf_evsel *evsel,
+			  int print_sym, int print_dso)
 {
 	struct addr_location al;
 	const char *symname, *dsoname;
-	struct callchain_cursor *cursor = &session->callchain_cursor;
+	struct callchain_cursor *cursor = &evsel->hists.callchain_cursor;
 	struct callchain_cursor_node *node;
 
-	if (perf_event__preprocess_sample(event, session, &al, sample,
+	if (perf_event__preprocess_sample(event, machine, &al, sample,
 					  NULL) < 0) {
 		error("problem processing %d event, skipping it.\n",
 			event->header.type);
@@ -1255,7 +1309,7 @@ void perf_session__print_ip(union perf_event *event,
 
 	if (symbol_conf.use_callchain && sample->callchain) {
 
-		if (perf_session__resolve_callchain(session, al.thread,
+		if (machine__resolve_callchain(machine, evsel, al.thread,
 						sample->callchain, NULL) != 0) {
 			if (verbose)
 				error("Failed to resolve callchain. Skipping\n");