i#6938 sched migrate: Add lock contention stats

Adds lock contention counters for mutex_dbg_owned::lock() in
non-NDEBUG builds.  While it's not there for release builds, the
results in non-release should still be indicative of general lock
behavior.

Prints the stats for sched_lock_ with the other scheduler stats.

Sample results on schedule_stats on a threadsig trace show a lot of
contention even with only 3 cores:

```
$ clients/bin64/drmemtrace_launcher -indir ../build_x64_dbg_tests/drmemtrace.threadsig.5* -core_sharded -cores 3 -tool schedule_stats -verbose 1
[scheduler] Schedule lock acquired     :   2196602
[scheduler] Schedule lock contended    :    257580
```

Issue: #6938

clients/drcachesim/common/mutex_dbg_owned.h

@@ -50,9 +50,18 @@ class mutex_dbg_owned {
     void
     lock()
     {
+#ifdef NDEBUG
         lock_.lock();
-#ifndef NDEBUG
+#else
+        bool contended = true;
+        if (lock_.try_lock())
+            contended = false;
+        else
+            lock_.lock();
         owner_ = std::this_thread::get_id();
+        ++count_acquired_;
+        if (contended)
+            ++count_contended_;
 #endif
     }
     bool
@@ -76,6 +85,8 @@ class mutex_dbg_owned {
 #endif
         lock_.unlock();
     }
+
+#ifndef NDEBUG
     // This query should only be called when the lock is required to be held
     // as it is racy when the lock is not held.
     bool
@@ -84,9 +95,27 @@ class mutex_dbg_owned {
         return owner_ == std::this_thread::get_id();
     }
 
+    // These statistics only count lock(): they do *not* count try_lock()
+    // (we could count try_lock with std::atomic on count_contended).
+    int64_t
+    get_count_acquired()
+    {
+        return count_acquired_;
+    }
+    int64_t
+    get_count_contended()
+    {
+        return count_contended_;
+    }
+#endif
+
 private:
     std::mutex lock_;
+#ifndef NDEBUG
     std::thread::id owner_;
+    int64_t count_acquired_ = 0;
+    int64_t count_contended_ = 0;
+#endif
 };
 
 } // namespace drmemtrace

clients/drcachesim/scheduler/scheduler.cpp

@@ -687,6 +687,10 @@ scheduler_tmpl_t<RecordType, ReaderType>::~scheduler_tmpl_t()
         VPRINT(this, 1, "  %-25s: %9" PRId64 "\n", "Migrations",
                outputs_[i].stats[memtrace_stream_t::SCHED_STAT_MIGRATIONS]);
     }
+    VPRINT(this, 1, "%-27s: %9" PRId64 "\n", "Schedule lock acquired",
+           sched_lock_.get_count_acquired());
+    VPRINT(this, 1, "%-27s: %9" PRId64 "\n", "Schedule lock contended",
+           sched_lock_.get_count_contended());
 }
 
 template <typename RecordType, typename ReaderType>