Connect SparseHistogram with the rest of stats system

base/metrics/histogram_snapshot_manager.cc

Index: base/metrics/histogram_snapshot_manager.cc
diff --git a/base/metrics/histogram_snapshot_manager.cc b/base/metrics/histogram_snapshot_manager.cc
index ad054b1003c07e4d2cb862b8f29b9c640b2f8fdd..ec685fe424a2bd7f7e659b9a536b4beac331e730 100644
--- a/base/metrics/histogram_snapshot_manager.cc
+++ b/base/metrics/histogram_snapshot_manager.cc
@@ -25,7 +25,7 @@ HistogramSnapshotManager::~HistogramSnapshotManager() {
   STLDeleteValues(&logged_samples_);
 }
 
-void HistogramSnapshotManager::PrepareDeltas(Histogram::Flags flag_to_set,
+void HistogramSnapshotManager::PrepareDeltas(HistogramBase::Flags flag_to_set,
                                              bool record_only_uma) {
   StatisticsRecorder::Histograms histograms;
   StatisticsRecorder::GetHistograms(&histograms);
@@ -40,7 +40,7 @@ void HistogramSnapshotManager::PrepareDeltas(Histogram::Flags flag_to_set,
   }
 }
 
-void HistogramSnapshotManager::PrepareDelta(const Histogram& histogram) {
+void HistogramSnapshotManager::PrepareDelta(const HistogramBase& histogram) {
   DCHECK(histogram_flattener_);
 
   // Get up-to-date snapshot of sample stats.
@@ -52,13 +52,13 @@ void HistogramSnapshotManager::PrepareDelta(const Histogram& histogram) {
   // Crash if we detect that our histograms have been overwritten.  This may be
   // a fair distance from the memory smasher, but we hope to correlate these
   // crashes with other events, such as plugins, or usage patterns, etc.
-  if (Histogram::BUCKET_ORDER_ERROR & corruption) {
+  if (BUCKET_ORDER_ERROR & corruption) {
     // The checksum should have caught this, so crash separately if it didn't.
-    CHECK_NE(0, Histogram::RANGE_CHECKSUM_ERROR & corruption);
+    CHECK_NE(0, RANGE_CHECKSUM_ERROR & corruption);
     CHECK(false);  // Crash for the bucket order corruption.
   }
   // Checksum corruption might not have caused order corruption.
-  CHECK_EQ(0, Histogram::RANGE_CHECKSUM_ERROR & corruption);
+  CHECK_EQ(0, RANGE_CHECKSUM_ERROR & corruption);
 
   // Note, at this point corruption can only be COUNT_HIGH_ERROR or
   // COUNT_LOW_ERROR and they never arise together, so we don't need to extract
@@ -67,14 +67,14 @@ void HistogramSnapshotManager::PrepareDelta(const Histogram& histogram) {
     DLOG(ERROR) << "Histogram: " << histogram_name
                 << " has data corruption: " << corruption;
     histogram_flattener_->InconsistencyDetected(
-        static_cast<Histogram::Inconsistencies>(corruption));
+        static_cast<HistogramInconsistency>(corruption));

[Ilya Sherman, 2013/02/21 05:36:06]

Seems like if you're willing to cast here, you might as well just have the
return type for HistogramBase::FindCorruption be HistogramInconsistency.

[kaiwang, 2013/02/27 04:39:42]

the value FindCorruption returns can be a combination of HistogramInconsistency.
As the comment above mentioned, in this case, only one type of inconsistency can
happen, hence it's directly casted.

     // Don't record corrupt data to metrics services.
     int old_corruption = inconsistencies_[histogram_name];
     if (old_corruption == (corruption | old_corruption))
       return;  // We've already seen this corruption for this histogram.
     inconsistencies_[histogram_name] |= corruption;
     histogram_flattener_->UniqueInconsistencyDetected(
-        static_cast<Histogram::Inconsistencies>(corruption));
+        static_cast<HistogramInconsistency>(corruption));
     return;
   }