perf top: Fall back to cpu-clock-tick hrtimer sampling if no cycle counter available

On architectures/CPUs without PMU support but with perfcounters
enabled 'perf top' currently fails because it cannot create a
cycle based hw-perfcounter.

Fall back to the cpu-clock-tick sw-perfcounter in this case, which
is hrtimer based and will always work (as long as perfcounters
is enabled).

Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

tools/perf/builtin-top.c

@@ -527,17 +527,13 @@ static void mmap_read(void)
 	}
 }
 
-static int __cmd_top(void)
+int nr_poll;
+int group_fd;
+
+static void start_counter(int i, int counter)
 {
 	struct perf_counter_attr *attr;
-	pthread_t thread;
-	int i, counter, group_fd, nr_poll = 0;
 	unsigned int cpu;
-	int ret;
-
-	for (i = 0; i < nr_cpus; i++) {
-		group_fd = -1;
-		for (counter = 0; counter < nr_counters; counter++) {
 
 	cpu = profile_cpu;
 	if (target_pid == -1 && profile_cpu == -1)
@@ -548,14 +544,30 @@ static int __cmd_top(void)
 	attr->sample_type	= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
 	attr->freq		= freq;
 
+try_again:
 	fd[i][counter] = sys_perf_counter_open(attr, target_pid, cpu, group_fd, 0);
+
 	if (fd[i][counter] < 0) {
 		int err = errno;
 
-				error("syscall returned with %d (%s)\n",
+		error("sys_perf_counter_open() syscall returned with %d (%s)\n",
 			fd[i][counter], strerror(err));
+
 		if (err == EPERM)
-					printf("Are you root?\n");
+			die(" No permission - are you root?\n");
+		/*
+		 * If it's cycles then fall back to hrtimer
+		 * based cpu-clock-tick sw counter, which
+		 * is always available even if no PMU support:
+		 */
+		if (attr->type == PERF_TYPE_HARDWARE
+			&& attr->config == PERF_COUNT_CPU_CYCLES) {
+
+			warning(" ... trying to fall back to cpu-clock-ticks\n");
+			attr->type = PERF_TYPE_SOFTWARE;
+			attr->config = PERF_COUNT_CPU_CLOCK;
+			goto try_again;
+		}
 		exit(-1);
 	}
 	assert(fd[i][counter] >= 0);
@@ -578,7 +590,18 @@ static int __cmd_top(void)
 			PROT_READ, MAP_SHARED, fd[i][counter], 0);
 	if (mmap_array[i][counter].base == MAP_FAILED)
 		die("failed to mmap with %d (%s)\n", errno, strerror(errno));
-		}
+}
+
+static int __cmd_top(void)
+{
+	pthread_t thread;
+	int i, counter;
+	int ret;
+
+	for (i = 0; i < nr_cpus; i++) {
+		group_fd = -1;
+		for (counter = 0; counter < nr_counters; counter++)
+			start_counter(i, counter);
 	}
 
 	/* Wait for a minimal set of events before starting the snapshot */

tools/perf/util/usage.c

@@ -9,29 +9,29 @@ static void report(const char *prefix, const char *err, va_list params)
 {
 	char msg[1024];
 	vsnprintf(msg, sizeof(msg), err, params);
-	fprintf(stderr, "%s%s\n", prefix, msg);
+	fprintf(stderr, " %s%s\n", prefix, msg);
 }
 
 static NORETURN void usage_builtin(const char *err)
 {
-	fprintf(stderr, "\n usage: %s\n", err);
+	fprintf(stderr, "\n Usage: %s\n", err);
 	exit(129);
 }
 
 static NORETURN void die_builtin(const char *err, va_list params)
 {
-	report("fatal: ", err, params);
+	report(" Fatal: ", err, params);
 	exit(128);
 }
 
 static void error_builtin(const char *err, va_list params)
 {
-	report("error: ", err, params);
+	report(" Error: ", err, params);
 }
 
 static void warn_builtin(const char *warn, va_list params)
 {
-	report("warning: ", warn, params);
+	report(" Warning: ", warn, params);
 }
 
 /* If we are in a dlopen()ed .so write to a global variable would segfault