google · dmah42 · Apr 9, 2019 · Feb 11, 2019 · Feb 11, 2019 · Feb 16, 2019
diff --git a/README.md b/README.md
@@ -428,6 +428,49 @@ BENCHMARK(BM_test)->Range(8, 8<<10)->UseRealTime();
 
 Without `UseRealTime`, CPU time is used by default.
 
+## CPU timers
+By default, the CPU timer only measures the time spent by the main thread.
+If the benchmark itself uses threads internally, this measurement may not
+be what you are looking for. Instead, there is a way to measure the total
+CPU usage of the process, by all the threads.
+
+```c++
+void callee(int i);
+
+static void MyMain(int size) {
+#pragma omp parallel for
+  for(int i = 0; i < size; i++)
+    callee(i);
+}
+
+static void BM_OpenMP(benchmark::State& state) {
+  for (auto _ : state)
+    MyMain(state.range(0);
+}
+
+// Measure the time spent by the main thread, use it to decide for how long to
+// run the benchmark loop. Depending on the internal implementation detail may
+// measure to anywhere from near-zero (the overhead spent before/after work
+// handoff to worker thread[s]) to the whole single-thread time.
+BENCHMARK(BM_OpenMP)->Range(8, 8<<10);
+
+// Measure the user-visible time, the wall clock (literally, the time that
+// has passed on the clock on the wall), use it to decide for how long to
+// run the benchmark loop. This will always be meaningful, an will match the
+// time spent by the main thread in single-threaded case, in general decreasing
+// with the number of internal threads doing the work.
+BENCHMARK(BM_OpenMP)->Range(8, 8<<10)->UseRealTime();
+
+// Measure the total CPU consumption, use it to decide for how long to
+// run the benchmark loop. This will always measure to no less than the
+// time spent by the main thread in single-threaded case.
+BENCHMARK(BM_OpenMP)->Range(8, 8<<10)->MeasureProcessCPUTime();
+
+// A mixture of the last two. Measure the total CPU consumption, but use the
+// wall clock to decide for how long to run the benchmark loop.
+BENCHMARK(BM_OpenMP)->Range(8, 8<<10)->MeasureProcessCPUTime()->UseRealTime();
+```
+
 ## Controlling timers
 Normally, the entire duration of the work loop (`for (auto _ : state) {}`)
 is measured. But sometimes, it is nessesary to do some work inside of

diff --git a/include/benchmark/benchmark.h b/include/benchmark/benchmark.h
@@ -874,11 +874,18 @@ class Benchmark {
   // Same as ReportAggregatesOnly(), but applies to display reporter only.
   Benchmark* DisplayAggregatesOnly(bool value = true);
 
-  // If a particular benchmark is I/O bound, runs multiple threads internally or
-  // if for some reason CPU timings are not representative, call this method. If
-  // called, the elapsed time will be used to control how many iterations are
-  // run, and in the printing of items/second or MB/seconds values.  If not
-  // called, the cpu time used by the benchmark will be used.
+  // By default, the CPU time is measured only for the main thread, which may
+  // be unrepresentative if the benchmark uses threads internally. If called,
+  // the total CPU time spent by all the threads will be measured instead.
+  // By default, the only the main thread CPU time will be measured.
+  Benchmark* MeasureProcessCPUTime();
+
+  // If a particular benchmark should use the Wall clock instead of the CPU time
+  // (be it either the CPU time of the main thread only (default), or the
+  // total CPU usage of the benchmark), call this method. If called, the elapsed
+  // (wall) time will be used to control how many iterations are run, and in the
+  // printing of items/second or MB/seconds values.
+  // If not called, the CPU time used by the benchmark will be used.
   Benchmark* UseRealTime();
 
   // If a benchmark must measure time manually (e.g. if GPU execution time is
@@ -952,6 +959,7 @@ class Benchmark {
   double min_time_;
   size_t iterations_;
   int repetitions_;
+  bool measure_process_cpu_time_;
   bool use_real_time_;
   bool use_manual_time_;
   BigO complexity_;

diff --git a/src/benchmark_api_internal.h b/src/benchmark_api_internal.h
@@ -22,6 +22,7 @@ struct BenchmarkInstance {
   std::vector<int64_t> arg;
   TimeUnit time_unit;
   int range_multiplier;
+  bool measure_process_cpu_time;
   bool use_real_time;
   bool use_manual_time;
   BigO complexity;

diff --git a/src/benchmark_register.cc b/src/benchmark_register.cc
@@ -165,6 +165,7 @@ bool BenchmarkFamilies::FindBenchmarks(
         instance.min_time = family->min_time_;
         instance.iterations = family->iterations_;
         instance.repetitions = family->repetitions_;
+        instance.measure_process_cpu_time = family->measure_process_cpu_time_;
         instance.use_real_time = family->use_real_time_;
         instance.use_manual_time = family->use_manual_time_;
         instance.complexity = family->complexity_;
@@ -202,10 +203,20 @@ bool BenchmarkFamilies::FindBenchmarks(
           instance.name.repetitions =
               StrFormat("repeats:%d", family->repetitions_);
 
+        if (family->measure_process_cpu_time_) {
+          instance.name.time_type = "process_time";
+        }
+
         if (family->use_manual_time_) {
-          instance.name.time_type = "manual_time";
+          if (!instance.name.time_type.empty()) {
+            instance.name.time_type += '/';
+          }
+          instance.name.time_type += "manual_time";
         } else if (family->use_real_time_) {
-          instance.name.time_type = "real_time";
+          if (!instance.name.time_type.empty()) {
+            instance.name.time_type += '/';
+          }
+          instance.name.time_type += "real_time";
         }
 
         // Add the number of threads used to the name
@@ -252,6 +263,7 @@ Benchmark::Benchmark(const char* name)
       min_time_(0),
       iterations_(0),
       repetitions_(0),
+      measure_process_cpu_time_(false),
       use_real_time_(false),
       use_manual_time_(false),
       complexity_(oNone),
@@ -398,6 +410,12 @@ Benchmark* Benchmark::DisplayAggregatesOnly(bool value) {
   return this;
 }
 
+Benchmark* Benchmark::MeasureProcessCPUTime() {
+  // Can be used together with UseRealTime() / UseManualTime().
+  measure_process_cpu_time_ = true;
+  return this;
+}
+
 Benchmark* Benchmark::UseRealTime() {
   CHECK(!use_manual_time_)
       << "Cannot set UseRealTime and UseManualTime simultaneously.";

diff --git a/src/benchmark_runner.cc b/src/benchmark_runner.cc
@@ -111,7 +111,10 @@ BenchmarkReporter::Run CreateRunReport(
 // Adds the stats collected for the thread into *total.
 void RunInThread(const BenchmarkInstance* b, size_t iters, int thread_id,
                  ThreadManager* manager) {
-  internal::ThreadTimer timer;
+  internal::ThreadTimer timer(
+      b->measure_process_cpu_time
+          ? internal::ThreadTimer::CreateProcessCpuTime()
+          : internal::ThreadTimer::Create());
   State st = b->Run(iters, thread_id, &timer, manager);
   CHECK(st.iterations() >= st.max_iterations)
       << "Benchmark returned before State::KeepRunning() returned false!";
@@ -226,6 +229,8 @@ class BenchmarkRunner {
     // Adjust real/manual time stats since they were reported per thread.
     i.results.real_time_used /= b.threads;
     i.results.manual_time_used /= b.threads;
+    // If we were measuring whole-process CPU usage, adjust the CPU time too.
+    if (b.measure_process_cpu_time) i.results.cpu_time_used /= b.threads;
 
     VLOG(2) << "Ran in " << i.results.cpu_time_used << "/"
             << i.results.real_time_used << "\n";

diff --git a/src/thread_timer.h b/src/thread_timer.h
@@ -8,14 +8,22 @@ namespace benchmark {
 namespace internal {
 
 class ThreadTimer {
+  explicit ThreadTimer(bool measure_process_cpu_time_)
+      : measure_process_cpu_time(measure_process_cpu_time_) {}
+
  public:
-  ThreadTimer() = default;
+  static ThreadTimer Create() {
+    return ThreadTimer(/*measure_process_cpu_time_=*/false);
+  }
+  static ThreadTimer CreateProcessCpuTime() {
+    return ThreadTimer(/*measure_process_cpu_time_=*/true);
+  }
 
   // Called by each thread
   void StartTimer() {
     running_ = true;
     start_real_time_ = ChronoClockNow();
-    start_cpu_time_ = ThreadCPUUsage();
+    start_cpu_time_ = ReadCpuTimerOfChoice();
   }
 
   // Called by each thread
@@ -25,7 +33,8 @@ class ThreadTimer {
     real_time_used_ += ChronoClockNow() - start_real_time_;
     // Floating point error can result in the subtraction producing a negative
     // time. Guard against that.
-    cpu_time_used_ += std::max<double>(ThreadCPUUsage() - start_cpu_time_, 0);
+    cpu_time_used_ +=
+        std::max<double>(ReadCpuTimerOfChoice() - start_cpu_time_, 0);
   }
 
   // Called by each thread
@@ -52,6 +61,14 @@ class ThreadTimer {
   }
 
  private:
+  double ReadCpuTimerOfChoice() const {
+    if (measure_process_cpu_time) return ProcessCPUUsage();
+    return ThreadCPUUsage();
+  }
+
+  // should the thread, or the process, time be measured?
+  const bool measure_process_cpu_time;
+
   bool running_ = false;        // Is the timer running
   double start_real_time_ = 0;  // If running_
   double start_cpu_time_ = 0;   // If running_

diff --git a/test/CMakeLists.txt b/test/CMakeLists.txt
@@ -125,6 +125,9 @@ add_test(templated_fixture_test templated_fixture_test --benchmark_min_time=0.01
 compile_output_test(user_counters_test)
 add_test(user_counters_test user_counters_test --benchmark_min_time=0.01)
 
+compile_output_test(internal_threading_test)
+add_test(internal_threading_test internal_threading_test --benchmark_min_time=0.01)
+
 compile_output_test(report_aggregates_only_test)
 add_test(report_aggregates_only_test report_aggregates_only_test --benchmark_min_time=0.01)