Skip to contentSkip to conversation

rtla/osnoise: Better report when histogram is empty

PATCH In review Luis Claudio R. Goncalves 3 participants 3 messages Jun 12, 2024 – Jun 21, 2024

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 rtla/osnoise: Better report when histogram is empty

When osnoise hist does not observe any samples above the threshold, no entries are recorded and the final report shows empty entries for the usual statistics (count, min, max, avg):

[~]# osnoise hist -d 5s -T 500
# RTLA osnoise histogram
# Time unit is microseconds (us)
# Duration: 0 00:00:05
Index
over:
count:
min:
avg:
max:

That could lead users to confusing interpretations of the results.

A simple solution is to report 0 for count and the statistics, making it clear that no noise (above the defined threshold) was observed:

    [~]# osnoise hist -d 5s -T 500
    # RTLA osnoise histogram
    # Time unit is microseconds (us)
    # Duration:   0 00:00:05
    Index
    over: 0
    count: 0
    min: 0
    avg: 0
    max: 0
---
 tools/tracing/rtla/src/osnoise_hist.c | 15 +++++++++++++++
 1 file changed, 15 insertions(+)
tools/tracing/rtla/src/osnoise_hist.c +15 -0
index 7be17d09f7e85..214e2c93fde01 100644
@@ -374,6 +374,7 @@ osnoise_print_stats(struct osnoise_hist_params *params, struct osnoise_tool *too
 {
 	struct osnoise_hist_data *data = tool->data;
 	struct trace_instance *trace = &tool->trace;
+	int has_samples = 0;
 	int bucket, cpu;
 	int total;
 
@@ -402,11 +403,25 @@ osnoise_print_stats(struct osnoise_hist_params *params, struct osnoise_tool *too
 			continue;
 		}
 
+		/* There are samples above the threshold */
+		has_samples = 1;
 		trace_seq_printf(trace->seq, "\n");
 		trace_seq_do_printf(trace->seq);
 		trace_seq_reset(trace->seq);
 	}
 
+	/*
+	 * If no samples were recorded, skip calculations, print zeroed statistics
+	 * and return.
+	 */
+	if (!has_samples) {
+		trace_seq_reset(trace->seq);
+		trace_seq_printf(trace->seq, "over: 0\ncount: 0\nmin: 0\navg: 0\nmax: 0\n");
+		trace_seq_do_printf(trace->seq);
+		trace_seq_reset(trace->seq);
+		return;
+	}
+
 	if (!params->no_index)
 		trace_seq_printf(trace->seq, "over: ");
 
Reviewed-by on v1 1/1
#1 lore raw
Quoted 50 lines · Luis Claudio R. Goncalves

When osnoise hist does not observe any samples above the threshold, no entries are recorded and the final report shows empty entries for the usual statistics (count, min, max, avg):

[~]# osnoise hist -d 5s -T 500
# RTLA osnoise histogram
# Time unit is microseconds (us)
# Duration: 0 00:00:05
Index
over:
count:
min:
avg:
max:

That could lead users to confusing interpretations of the results.

A simple solution is to report 0 for count and the statistics, making it clear that no noise (above the defined threshold) was observed:

    [~]# osnoise hist -d 5s -T 500
    # RTLA osnoise histogram
    # Time unit is microseconds (us)
    # Duration:   0 00:00:05
    Index
    over: 0
    count: 0
    min: 0
    avg: 0
    max: 0
---
 tools/tracing/rtla/src/osnoise_hist.c | 15 +++++++++++++++
 1 file changed, 15 insertions(+)
tools/tracing/rtla/src/osnoise_hist.c +1 -0
index 7be17d09f7e85..214e2c93fde01 100644
@@ -374,6 +374,7 @@ osnoise_print_stats(struct osnoise_hist_params *params, struct osnoise_tool *too
 {
 	struct osnoise_hist_data *data = tool->data;
 	struct trace_instance *trace = &tool->trace;
+	int has_samples = 0;
 	int bucket, cpu;
 	int total;
 

tools/tracing/rtla/src/osnoise_hist.c osnoise_print_stats(struct osnoise_hist_params *params, struct osnoise_tool *too · line 406

Quoted 4 lines
tools/tracing/rtla/src/osnoise_hist.c +1 -0
@@ -402,11 +403,25 @@ osnoise_print_stats(struct osnoise_hist_params *params, struct osnoise_tool *too
 			continue;
 		}
 
+		/* There are samples above the threshold */

IMHO The comment isn't needed because the variable had_samples is descriptive, but it's not a big deal either

Quoted 14 lines
tools/tracing/rtla/src/osnoise_hist.c +6 -0
+		has_samples = 1;
 		trace_seq_printf(trace->seq, "\n");
 		trace_seq_do_printf(trace->seq);
 		trace_seq_reset(trace->seq);
 	}
 
+	/*
+	 * If no samples were recorded, skip calculations, print zeroed statistics
+	 * and return.
+	 */
+	if (!has_samples) {

tools/tracing/rtla/src/osnoise_hist.c osnoise_print_stats(struct osnoise_hist_params *params, struct osnoise_tool *too · line 427

Quoted 10 lines
tools/tracing/rtla/src/osnoise_hist.c +7 -0
+		trace_seq_reset(trace->seq);
+		trace_seq_printf(trace->seq, "over: 0\ncount: 0\nmin: 0\navg: 0\nmax: 0\n");
+		trace_seq_do_printf(trace->seq);
+		trace_seq_reset(trace->seq);
+		return;
+	}
+
 	if (!params->no_index)
 		trace_seq_printf(trace->seq, "over: ");
 
#2 lore raw
Quoted 3 lines · John Kacur

tools/tracing/rtla/src/osnoise_hist.c osnoise_print_stats(struct osnoise_hist_params *params, struct osnoise_tool *too · line 406

Quoted 1 line
tools/tracing/rtla/src/osnoise_hist.c +1 -0
+		/* There are samples above the threshold */

IMHO The comment isn't needed because the variable had_samples is descriptive, but it's not a big deal either

Yeah, I would not do it too, but it is fine.

I am adding it to my queue.

Thanks Luis and John!
-- Daniel