Add a basic exact linear macro. Ends up being the same as enum, but making it explicit will make ou…

base/metrics/histogram_macros.h

Index: base/metrics/histogram_macros.h
diff --git a/base/metrics/histogram_macros.h b/base/metrics/histogram_macros.h
index 9cb21ec563a2ed54e45cdd885b4d2501a84ad985..e768712f63546e135c1d063352f3b741feb4fad4 100644
--- a/base/metrics/histogram_macros.h
+++ b/base/metrics/histogram_macros.h
@@ -53,17 +53,27 @@
             base::HistogramBase::kUmaTargetedHistogramFlag))
 
 //------------------------------------------------------------------------------
-// Percentage histograms.
+// Linear histograms.
 
 // All of these macros must be called with |name| as a runtime constant.
 
+// Used for capturing integer data with a linear bucketing scheme. This can be
+// used when you want the exact value of some small numeric count, with a max of
+// 100 or less. If you need to capture a range of greater than 100, we recommend
+// the use of the COUNT histograms below.
+
+// Sample usage:
+//   UMA_HISTOGRAM_EXACT_LINEAR("Histogram.Linear", count, 10);
+#define UMA_HISTOGRAM_EXACT_LINEAR(name, sample, value_max)                    \
+    UMA_HISTOGRAM_ENUMERATION(name, sample, value_max)

[Mark P, 2016/10/03 17:42:29]

Should we make this value_max+1 to better detect corruption (corrupted data will
fall into the overflow bucket, which will not overlap a real bucket)?

[rkaplow, 2016/10/03 19:19:44]

sure, that sounds good, although we do want to do a better job of this more
systematically (as per the bug I filed). Made the adjustment here though

+
+

[Mark P, 2016/10/03 17:42:29]

optional nit: remove one black line

 // Used for capturing basic percentages. This will be 100 buckets of size 1.
 
 // Sample usage:
 //   UMA_HISTOGRAM_PERCENTAGE("Histogram.Percent", percent_as_int);
-#define UMA_HISTOGRAM_PERCENTAGE(name, under_one_hundred)                      \
-    UMA_HISTOGRAM_ENUMERATION(name, under_one_hundred, 101)
-
+#define UMA_HISTOGRAM_PERCENTAGE(name, percent_as_int)                         \
+    UMA_HISTOGRAM_ENUMERATION(name, percent_as_int, 101)
 
 //------------------------------------------------------------------------------
 // Count histograms. These are used for collecting numeric data. Note that we
@@ -117,7 +127,6 @@
         name, sample, min, max, bucket_count,                                  \
         base::HistogramBase::kUmaTargetedHistogramFlag)
 
-
 //------------------------------------------------------------------------------
 // Timing histograms. These are used for collecting timing data (generally
 // latencies).