blob: 60c005726f56b40433a8a2d778dfd1ece3e3c12d [file] [log] [blame]
// Copyright 2016 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#ifndef BASE_METRICS_HISTOGRAM_FUNCTIONS_H_
#define BASE_METRICS_HISTOGRAM_FUNCTIONS_H_
#include "base/metrics/histogram.h"
#include "base/metrics/histogram_base.h"
#include "base/time/time.h"
// Functions for recording metrics.
//
// For best practices on deciding when to emit to a histogram and what form
// the histogram should take, see
// https://chromium.googlesource.com/chromium/src.git/+/HEAD/tools/metrics/histograms/README.md
// Functions for recording UMA histograms. These can be used for cases
// when the histogram name is generated at runtime. The functionality is
// equivalent to macros defined in histogram_macros.h but allowing non-constant
// histogram names. These functions are slower compared to their macro
// equivalent because the histogram objects are not cached between calls.
// So, these shouldn't be used in performance critical code.
namespace base {
// For histograms with linear buckets.
// 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:
// base::UmaHistogramExactLinear("Histogram.Linear", some_value, 10);
BASE_EXPORT void UmaHistogramExactLinear(const std::string& name,
int sample,
int value_max);
// For adding a sample to an enumerated histogram.
// Sample usage:
// // These values are persisted to logs. Entries should not be renumbered and
// // numeric values should never be reused.
// enum class MyEnum {
// FIRST_VALUE = 0,
// SECOND_VALUE = 1,
// ...
// FINAL_VALUE = N,
// COUNT
// };
// base::UmaHistogramEnumeration("My.Enumeration",
// MyEnum::SOME_VALUE, MyEnum::COUNT);
//
// Note: The value in |sample| must be strictly less than |enum_size|.
template <typename T>
void UmaHistogramEnumeration(const std::string& name, T sample, T enum_size) {
static_assert(std::is_enum<T>::value,
"Non enum passed to UmaHistogramEnumeration");
DCHECK_LE(static_cast<uintmax_t>(enum_size), static_cast<uintmax_t>(INT_MAX));
DCHECK_LT(static_cast<uintmax_t>(sample), static_cast<uintmax_t>(enum_size));
return UmaHistogramExactLinear(name, static_cast<int>(sample),
static_cast<int>(enum_size));
}
// Same as above, but uses T::kMaxValue as the inclusive maximum value of the
// enum.
template <typename T>
void UmaHistogramEnumeration(const std::string& name, T sample) {
static_assert(std::is_enum<T>::value,
"Non enum passed to UmaHistogramEnumeration");
DCHECK_LE(static_cast<uintmax_t>(T::kMaxValue),
static_cast<uintmax_t>(INT_MAX) - 1);
DCHECK_LE(static_cast<uintmax_t>(sample),
static_cast<uintmax_t>(T::kMaxValue));
return UmaHistogramExactLinear(name, static_cast<int>(sample),
static_cast<int>(T::kMaxValue) + 1);
}
// For adding boolean sample to histogram.
// Sample usage:
// base::UmaHistogramBoolean("My.Boolean", true)
BASE_EXPORT void UmaHistogramBoolean(const std::string& name, bool sample);
// For adding histogram with percent.
// Percents are integer between 1 and 100.
// Sample usage:
// base::UmaHistogramPercentage("My.Percent", 69)
BASE_EXPORT void UmaHistogramPercentage(const std::string& name, int percent);
// For adding counts histogram.
// Sample usage:
// base::UmaHistogramCustomCounts("My.Counts", some_value, 1, 600, 30)
BASE_EXPORT void UmaHistogramCustomCounts(const std::string& name,
int sample,
int min,
int max,
int buckets);
// Counts specialization for maximum counts 100, 1000, 10k, 100k, 1M and 10M.
BASE_EXPORT void UmaHistogramCounts100(const std::string& name, int sample);
BASE_EXPORT void UmaHistogramCounts1000(const std::string& name, int sample);
BASE_EXPORT void UmaHistogramCounts10000(const std::string& name, int sample);
BASE_EXPORT void UmaHistogramCounts100000(const std::string& name, int sample);
BASE_EXPORT void UmaHistogramCounts1M(const std::string& name, int sample);
BASE_EXPORT void UmaHistogramCounts10M(const std::string& name, int sample);
// For histograms storing times.
BASE_EXPORT void UmaHistogramCustomTimes(const std::string& name,
TimeDelta sample,
TimeDelta min,
TimeDelta max,
int buckets);
// For short timings from 1 ms up to 10 seconds (50 buckets).
BASE_EXPORT void UmaHistogramTimes(const std::string& name, TimeDelta sample);
// For medium timings up to 3 minutes (50 buckets).
BASE_EXPORT void UmaHistogramMediumTimes(const std::string& name,
TimeDelta sample);
// For time intervals up to 1 hr (50 buckets).
BASE_EXPORT void UmaHistogramLongTimes(const std::string& name,
TimeDelta sample);
// For recording memory related histograms.
// Used to measure common KB-granularity memory stats. Range is up to 500M.
BASE_EXPORT void UmaHistogramMemoryKB(const std::string& name, int sample);
// Used to measure common MB-granularity memory stats. Range is up to ~1G.
BASE_EXPORT void UmaHistogramMemoryMB(const std::string& name, int sample);
// Used to measure common MB-granularity memory stats. Range is up to ~64G.
BASE_EXPORT void UmaHistogramMemoryLargeMB(const std::string& name, int sample);
// For recording sparse histograms.
// The |sample| can be a negative or non-negative number.
//
// Sparse histograms are well suited for recording counts of exact sample values
// that are sparsely distributed over a relatively large range, in cases where
// ultra-fast performance is not critical. For instance, Sqlite.Version.* are
// sparse because for any given database, there's going to be exactly one
// version logged.
//
// Performance:
// ------------
// Sparse histograms are typically more memory-efficient but less time-efficient
// than other histograms. Essentially, they sparse histograms use a map rather
// than a vector for their backing storage; they also require lock acquisition
// to increment a sample, whereas other histogram do not. Hence, each increment
// operation is a bit slower than for other histograms. But, if the data is
// sparse, then they use less memory client-side, because they allocate buckets
// on demand rather than preallocating.
//
// Data size:
// ----------
// Note that server-side, we still need to load all buckets, across all users,
// at once. Thus, please avoid exploding such histograms, i.e. uploading many
// many distinct values to the server (across all users). Concretely, keep the
// number of distinct values <= 100 ideally, definitely <= 1000. If you have no
// guarantees on the range of your data, use clamping, e.g.:
// UmaHistogramSparse("MyHistogram", ClampToRange(value, 0, 200));
BASE_EXPORT void UmaHistogramSparse(const std::string& name, int sample);
} // namespace base
#endif // BASE_METRICS_HISTOGRAM_FUNCTIONS_H_