Merge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/…

…linux/kernel/git/tip/linux-2.6-tip

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  perf tools: Fix sample size bit operations
  perf tools: Fix ommitted mmap data update on remap
  watchdog: Change the default timeout and configure nmi watchdog period based on watchdog_thresh
  watchdog: Disable watchdog when thresh is zero
  watchdog: Only disable/enable watchdog if neccessary
  watchdog: Fix rounding bug in get_sample_period()
  perf tools: Propagate event parse error handling
  perf tools: Robustify dynamic sample content fetch
  perf tools: Pre-check sample size before parsing
  perf tools: Move evlist sample helpers to evlist area
  perf tools: Remove junk code in mmap size handling
  perf tools: Check we are able to read the event size on mmap

arch/x86/kernel/apic/hw_nmi.c

@@ -19,9 +19,9 @@
 #include <linux/delay.h>
 
 #ifdef CONFIG_HARDLOCKUP_DETECTOR
-u64 hw_nmi_get_sample_period(void)
+u64 hw_nmi_get_sample_period(int watchdog_thresh)
 {
-	return (u64)(cpu_khz) * 1000 * 60;
+	return (u64)(cpu_khz) * 1000 * watchdog_thresh;
 }
 #endif
 

include/linux/nmi.h

@@ -45,11 +45,12 @@ static inline bool trigger_all_cpu_backtrace(void)
 
 #ifdef CONFIG_LOCKUP_DETECTOR
 int hw_nmi_is_cpu_stuck(struct pt_regs *);
-u64 hw_nmi_get_sample_period(void);
+u64 hw_nmi_get_sample_period(int watchdog_thresh);
 extern int watchdog_enabled;
+extern int watchdog_thresh;
 struct ctl_table;
-extern int proc_dowatchdog_enabled(struct ctl_table *, int ,
-			void __user *, size_t *, loff_t *);
+extern int proc_dowatchdog(struct ctl_table *, int ,
+			   void __user *, size_t *, loff_t *);
 #endif
 
 #endif

include/linux/sched.h

@@ -315,7 +315,6 @@ extern int proc_dowatchdog_thresh(struct ctl_table *table, int write,
 				  void __user *buffer,
 				  size_t *lenp, loff_t *ppos);
 extern unsigned int  softlockup_panic;
-extern int softlockup_thresh;
 void lockup_detector_init(void);
 #else
 static inline void touch_softlockup_watchdog(void)

kernel/sysctl.c

@@ -730,14 +730,16 @@ static struct ctl_table kern_table[] = {
 		.data           = &watchdog_enabled,
 		.maxlen         = sizeof (int),
 		.mode           = 0644,
-		.proc_handler   = proc_dowatchdog_enabled,
+		.proc_handler   = proc_dowatchdog,
+		.extra1		= &zero,
+		.extra2		= &one,
 	},
 	{
 		.procname	= "watchdog_thresh",
-		.data		= &softlockup_thresh,
+		.data		= &watchdog_thresh,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= proc_dowatchdog_thresh,
+		.proc_handler	= proc_dowatchdog,
 		.extra1		= &neg_one,
 		.extra2		= &sixty,
 	},
@@ -755,7 +757,9 @@ static struct ctl_table kern_table[] = {
 		.data           = &watchdog_enabled,
 		.maxlen         = sizeof (int),
 		.mode           = 0644,
-		.proc_handler   = proc_dowatchdog_enabled,
+		.proc_handler   = proc_dowatchdog,
+		.extra1		= &zero,
+		.extra2		= &one,
 	},
 #endif
 #if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86)

kernel/watchdog.c

@@ -28,7 +28,7 @@
 #include <linux/perf_event.h>
 
 int watchdog_enabled = 1;
-int __read_mostly softlockup_thresh = 60;
+int __read_mostly watchdog_thresh = 10;
 
 static DEFINE_PER_CPU(unsigned long, watchdog_touch_ts);
 static DEFINE_PER_CPU(struct task_struct *, softlockup_watchdog);
@@ -91,6 +91,17 @@ static int __init nosoftlockup_setup(char *str)
 __setup("nosoftlockup", nosoftlockup_setup);
 /*  */
 
+/*
+ * Hard-lockup warnings should be triggered after just a few seconds. Soft-
+ * lockups can have false positives under extreme conditions. So we generally
+ * want a higher threshold for soft lockups than for hard lockups. So we couple
+ * the thresholds with a factor: we make the soft threshold twice the amount of
+ * time the hard threshold is.
+ */
+static int get_softlockup_thresh()
+{
+	return watchdog_thresh * 2;
+}
 
 /*
  * Returns seconds, approximately.  We don't need nanosecond
@@ -105,12 +116,12 @@ static unsigned long get_timestamp(int this_cpu)
 static unsigned long get_sample_period(void)
 {
 	/*
-	 * convert softlockup_thresh from seconds to ns
+	 * convert watchdog_thresh from seconds to ns
 	 * the divide by 5 is to give hrtimer 5 chances to
 	 * increment before the hardlockup detector generates
 	 * a warning
 	 */
-	return softlockup_thresh / 5 * NSEC_PER_SEC;
+	return get_softlockup_thresh() * (NSEC_PER_SEC / 5);
 }
 
 /* Commands for resetting the watchdog */
@@ -182,7 +193,7 @@ static int is_softlockup(unsigned long touch_ts)
 	unsigned long now = get_timestamp(smp_processor_id());
 
 	/* Warn about unreasonable delays: */
-	if (time_after(now, touch_ts + softlockup_thresh))
+	if (time_after(now, touch_ts + get_softlockup_thresh()))
 		return now - touch_ts;
 
 	return 0;
@@ -359,7 +370,7 @@ static int watchdog_nmi_enable(int cpu)
 
 	/* Try to register using hardware perf events */
 	wd_attr = &wd_hw_attr;
-	wd_attr->sample_period = hw_nmi_get_sample_period();
+	wd_attr->sample_period = hw_nmi_get_sample_period(watchdog_thresh);
 	event = perf_event_create_kernel_counter(wd_attr, cpu, NULL, watchdog_overflow_callback);
 	if (!IS_ERR(event)) {
 		printk(KERN_INFO "NMI watchdog enabled, takes one hw-pmu counter.\n");
@@ -501,28 +512,25 @@ static void watchdog_disable_all_cpus(void)
 /* sysctl functions */
 #ifdef CONFIG_SYSCTL
 /*
- * proc handler for /proc/sys/kernel/nmi_watchdog
+ * proc handler for /proc/sys/kernel/nmi_watchdog,watchdog_thresh
  */
 
-int proc_dowatchdog_enabled(struct ctl_table *table, int write,
-		     void __user *buffer, size_t *length, loff_t *ppos)
+int proc_dowatchdog(struct ctl_table *table, int write,
+		    void __user *buffer, size_t *lenp, loff_t *ppos)
 {
-	proc_dointvec(table, write, buffer, length, ppos);
+	int ret;
 
-	if (write) {
-		if (watchdog_enabled)
-			watchdog_enable_all_cpus();
-		else
-			watchdog_disable_all_cpus();
-	}
-	return 0;
-}
+	ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
+	if (ret || !write)
+		goto out;
 
-int proc_dowatchdog_thresh(struct ctl_table *table, int write,
-			     void __user *buffer,
-			     size_t *lenp, loff_t *ppos)
-{
-	return proc_dointvec_minmax(table, write, buffer, lenp, ppos);
+	if (watchdog_enabled && watchdog_thresh)
+		watchdog_enable_all_cpus();
+	else
+		watchdog_disable_all_cpus();
+
+out:
+	return ret;
 }
 #endif /* CONFIG_SYSCTL */
 

tools/perf/builtin-test.c

@@ -474,6 +474,7 @@ static int test__basic_mmap(void)
 	unsigned int nr_events[nsyscalls],
 		     expected_nr_events[nsyscalls], i, j;
 	struct perf_evsel *evsels[nsyscalls], *evsel;
+	int sample_size = perf_sample_size(attr.sample_type);
 
 	for (i = 0; i < nsyscalls; ++i) {
 		char name[64];
@@ -558,7 +559,13 @@ static int test__basic_mmap(void)
 			goto out_munmap;
 		}
 
-		perf_event__parse_sample(event, attr.sample_type, false, &sample);
+		err = perf_event__parse_sample(event, attr.sample_type, sample_size,
+					       false, &sample);
+		if (err) {
+			pr_err("Can't parse sample, err = %d\n", err);
+			goto out_munmap;
+		}
+
 		evsel = perf_evlist__id2evsel(evlist, sample.id);
 		if (evsel == NULL) {
 			pr_debug("event with id %" PRIu64

tools/perf/builtin-top.c

@@ -805,9 +805,14 @@ static void perf_session__mmap_read_idx(struct perf_session *self, int idx)
 {
 	struct perf_sample sample;
 	union perf_event *event;
+	int ret;
 
 	while ((event = perf_evlist__mmap_read(top.evlist, idx)) != NULL) {
-		perf_session__parse_sample(self, event, &sample);
+		ret = perf_session__parse_sample(self, event, &sample);
+		if (ret) {
+			pr_err("Can't parse sample, err = %d\n", ret);
+			continue;
+		}
 
 		if (event->header.type == PERF_RECORD_SAMPLE)
 			perf_event__process_sample(event, &sample, self);

tools/perf/util/event.c

@@ -9,21 +9,21 @@
 #include "thread_map.h"
 
 static const char *perf_event__names[] = {
-	[0]			 = "TOTAL",
-	[PERF_RECORD_MMAP]	 = "MMAP",
-	[PERF_RECORD_LOST]	 = "LOST",
-	[PERF_RECORD_COMM]	 = "COMM",
-	[PERF_RECORD_EXIT]	 = "EXIT",
-	[PERF_RECORD_THROTTLE]	 = "THROTTLE",
-	[PERF_RECORD_UNTHROTTLE] = "UNTHROTTLE",
-	[PERF_RECORD_FORK]	 = "FORK",
-	[PERF_RECORD_READ]	 = "READ",
-	[PERF_RECORD_SAMPLE]	 = "SAMPLE",
-	[PERF_RECORD_HEADER_ATTR]	 = "ATTR",
-	[PERF_RECORD_HEADER_EVENT_TYPE]	 = "EVENT_TYPE",
-	[PERF_RECORD_HEADER_TRACING_DATA]	 = "TRACING_DATA",
-	[PERF_RECORD_HEADER_BUILD_ID]	 = "BUILD_ID",
-	[PERF_RECORD_FINISHED_ROUND]	 = "FINISHED_ROUND",
+	[0]					= "TOTAL",
+	[PERF_RECORD_MMAP]			= "MMAP",
+	[PERF_RECORD_LOST]			= "LOST",
+	[PERF_RECORD_COMM]			= "COMM",
+	[PERF_RECORD_EXIT]			= "EXIT",
+	[PERF_RECORD_THROTTLE]			= "THROTTLE",
+	[PERF_RECORD_UNTHROTTLE]		= "UNTHROTTLE",
+	[PERF_RECORD_FORK]			= "FORK",
+	[PERF_RECORD_READ]			= "READ",
+	[PERF_RECORD_SAMPLE]			= "SAMPLE",
+	[PERF_RECORD_HEADER_ATTR]		= "ATTR",
+	[PERF_RECORD_HEADER_EVENT_TYPE]		= "EVENT_TYPE",
+	[PERF_RECORD_HEADER_TRACING_DATA]	= "TRACING_DATA",
+	[PERF_RECORD_HEADER_BUILD_ID]		= "BUILD_ID",
+	[PERF_RECORD_FINISHED_ROUND]		= "FINISHED_ROUND",
 };
 
 const char *perf_event__name(unsigned int id)
@@ -35,6 +35,22 @@ const char *perf_event__name(unsigned int id)
 	return perf_event__names[id];
 }
 
+int perf_sample_size(u64 sample_type)
+{
+	u64 mask = sample_type & PERF_SAMPLE_MASK;
+	int size = 0;
+	int i;
+
+	for (i = 0; i < 64; i++) {
+		if (mask & (1UL << i))
+			size++;
+	}
+
+	size *= sizeof(u64);
+
+	return size;
+}
+
 static struct perf_sample synth_sample = {
 	.pid	   = -1,
 	.tid	   = -1,

tools/perf/util/event.h

@@ -56,6 +56,13 @@ struct read_event {
 	u64 id;
 };
 
+
+#define PERF_SAMPLE_MASK				\
+	(PERF_SAMPLE_IP | PERF_SAMPLE_TID |		\
+	 PERF_SAMPLE_TIME | PERF_SAMPLE_ADDR |		\
+	PERF_SAMPLE_ID | PERF_SAMPLE_STREAM_ID |	\
+	 PERF_SAMPLE_CPU | PERF_SAMPLE_PERIOD)
+
 struct sample_event {
 	struct perf_event_header        header;
 	u64 array[];
@@ -75,6 +82,8 @@ struct perf_sample {
 	struct ip_callchain *callchain;
 };
 
+int perf_sample_size(u64 sample_type);
+
 #define BUILD_ID_SIZE 20
 
 struct build_id_event {
@@ -178,6 +187,7 @@ int perf_event__preprocess_sample(const union perf_event *self,
 const char *perf_event__name(unsigned int id);
 
 int perf_event__parse_sample(const union perf_event *event, u64 type,
-			     bool sample_id_all, struct perf_sample *sample);
+			     int sample_size, bool sample_id_all,
+			     struct perf_sample *sample);
 
 #endif /* __PERF_RECORD_H */

tools/perf/util/evlist.c

@@ -459,3 +459,34 @@ int perf_evlist__set_filters(struct perf_evlist *evlist)
 
 	return 0;
 }
+
+u64 perf_evlist__sample_type(struct perf_evlist *evlist)
+{
+	struct perf_evsel *pos;
+	u64 type = 0;
+
+	list_for_each_entry(pos, &evlist->entries, node) {
+		if (!type)
+			type = pos->attr.sample_type;
+		else if (type != pos->attr.sample_type)
+			die("non matching sample_type");
+	}
+
+	return type;
+}
+
+bool perf_evlist__sample_id_all(const struct perf_evlist *evlist)
+{
+	bool value = false, first = true;
+	struct perf_evsel *pos;
+
+	list_for_each_entry(pos, &evlist->entries, node) {
+		if (first) {
+			value = pos->attr.sample_id_all;
+			first = false;
+		} else if (value != pos->attr.sample_id_all)
+			die("non matching sample_id_all");
+	}
+
+	return value;
+}

tools/perf/util/evlist.h

@@ -66,4 +66,7 @@ int perf_evlist__create_maps(struct perf_evlist *evlist, pid_t target_pid,
 void perf_evlist__delete_maps(struct perf_evlist *evlist);
 int perf_evlist__set_filters(struct perf_evlist *evlist);
 
+u64 perf_evlist__sample_type(struct perf_evlist *evlist);
+bool perf_evlist__sample_id_all(const struct perf_evlist *evlist);
+
 #endif /* __PERF_EVLIST_H */