<!--
Copyright 2020 The Chromium Authors
Use of this source code is governed by a BSD-style license that can be
found in the LICENSE file.
-->
<!--
This file is used to generate a comprehensive list of Renderer histograms
along with a detailed description for each histogram.
For best practices on writing histogram descriptions, see
https://chromium.googlesource.com/chromium/src.git/+/HEAD/tools/metrics/histograms/README.md
Please follow the instructions in the OWNERS file in this directory to find a
reviewer. If no OWNERS file exists, please consider signing up at
go/reviewing-metrics (Googlers only), as all subdirectories are expected to
have an OWNERS file. As a last resort you can send the CL to
[email protected].
-->
<histogram-configuration>
<histograms>
<variants name="DelegatedInkRenderer">
<variant name="OS" summary="using the native OS Delegated Ink API"/>
<variant name="Skia"
summary="using the Skia implementation for Delegated Ink"/>
</variants>
<variants name="RendererHostedContentType">
<variant name="" summary="all renderers"/>
<variant name=".Empty"
summary="renderers that never hosted a frame or a worker"/>
<variant name=".Extension"
summary="renderers that hosted an extension background page"/>
<variant name=".MainFrame"
summary="renderers that hosted a navigated main frame"/>
<variant name=".Speculative"
summary="speculative renderers, i.e. renderers that hosted frames but
in which no navigation commit occured"/>
<variant name=".Subframe_Ad"
summary="renderers that hosted a navigated subframe (tagged as an ad),
but no main frame"/>
<variant name=".Subframe_NoAd"
summary="renderers that hosted a navigated subframe (not tagged as an
ad), but no main frame"/>
<variant name=".Worker"
summary="renderers that hosted a worker (service, dedicated or shared),
but no navigated frame"/>
</variants>
<histogram name="Renderer.BrowserLaunchToRunLoopStart" units="ms"
expires_after="2025-01-05">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Time (in milliseconds) between when browser launches renderer and renderer
message loop starts.
</summary>
</histogram>
<histogram
name="Renderer.Core.Timing.Performance.SwapsPerPerformanceEntryInsertion"
units="swaps" expires_after="2024-12-30">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Records the number of 'swaps' performed when inserting a single new
Performance Entry into its respective buffer. Recorded each time a new
performance entry is inserted into a sorted PerformanceEntryVector. Note
that this does not include singleton entries such as NavigationTiming or
PaintTiming.
</summary>
</histogram>
<histogram name="Renderer.CriticalFonts.BlockingResourcesLoadTime" units="ms"
expires_after="2025-03-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Time (in milliseconds) from timeOrigin until all the render blocking
resources were loaded in pages that have preloaded fonts. The metric will be
reported once per page load, once all the render blocking resources and
pending preloaded fonts have finished loading.
</summary>
</histogram>
<histogram name="Renderer.CriticalFonts.CriticalFontDelay" units="ms"
expires_after="2025-03-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Time difference (in milliseconds) between `PreloadedFontLoadTime` and
`BlockingResourcesLoadTime`, when the preloaded fonts took longer to load
than the render blocking resources, and otherwise 0. This represents the
time preloaded fonts would block rendering, if we were to consider them as
critical.
</summary>
</histogram>
<histogram name="Renderer.CriticalFonts.PreloadedFontsLoadTime" units="ms"
expires_after="2025-03-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Time (in milliseconds) from timeOrigin until all the preloaded fonts were
loaded. The metric will be reported once per page load, once all the render
blocking resources and pending preloaded fonts have finished loading.
</summary>
</histogram>
<histogram
name="Renderer.DelegatedInkTrail.LatencyImprovement.{Renderer}.{Prediction}"
units="ms" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<improvement direction="HIGHER_IS_BETTER"/>
<summary>
The latency improvement provided by the delegated ink trail API
{Prediction}, in milliseconds {Renderer}. This latency improvement is the
difference between the timestamp of the pointerevent provided to the API,
and the final point of the trail that is drawn. Points are forwarded from
the browser to viz when this API is being used, so viz will have points
ahead of what the app provided the API, and they are used to draw the trail.
This metric measures how much closer the trail is to the pointer, in time,
compared to the last API drawn point. The bigger the histogram's value, the
closer Delegated Ink got us to the last pointer position.
</summary>
<token key="Renderer" variants="DelegatedInkRenderer"/>
<token key="Prediction">
<variant name="WithoutPrediction" summary="without point prediction"/>
<variant name="WithPrediction" summary="using point draw prediction"/>
</token>
</histogram>
<histogram
name="Renderer.DelegatedInkTrail.LatencyImprovementWithPrediction.Experiment{Number}"
units="ms" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The latency improvement provided by predicting extra delegated ink points
when using the delegated ink trail API with SkiaRenderer in milliseconds.
The improvement is the difference between the timestamp of the final real
point and the final predicted point.
Part of a series of predicted point experimentation histograms. This
experiment predicts {Number} in the future.
</summary>
<token key="Number">
<variant name="0" summary="1 predicted point, 12ms"/>
<variant name="1" summary="2 predicted points, 6ms and 12ms"/>
<variant name="2" summary="1 predicted point, 6ms"/>
<variant name="3" summary="2 predicted points, 3ms and 6ms"/>
</token>
</histogram>
<histogram name="Renderer.DelegatedInkTrail.Prediction.WrongDirection"
enum="BooleanDirection" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Boolean that indicates whether the prediction is in the same direction as
the real trajectory at the same prediction time. True means the direction is
not the same (wrong direction).
Temporarily not logged in favor of
Renderer.DelegatedInkTrail.PredictionExperiment*.WrongDirection since 02-26.
Will be reenabled in M95, after experimentation.
</summary>
</histogram>
<histogram name="Renderer.DelegatedInkTrail.Prediction.{Direction}Prediction"
units="pixels" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
{Direction}shoot distance in pixels between a predicted point and the
corresponding interpolated real point that occurred at the prediction time.
Temporarily not logged in favor of
Renderer.DelegatedInkTrail.PredictionExperiment*.{Direction}Prediction since
02-26. Will be reenabled in M95, after experimentation.
</summary>
<token key="Direction">
<variant name="Over"/>
<variant name="Under"/>
</token>
</histogram>
<histogram name="Renderer.DelegatedInkTrail.Prediction.{Type}Jitter"
units="pixels" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Euclidean distance in pixels between two successive variations of delegated
ink trail deltas which are the difference beween a predicted delegated ink
point and its corresponding interpolated real position that occurred at
{Type} time.
Temporarily not logged in favor of
Renderer.DelegatedInkTrail.PredictionExperiment*.{Type}Jitter since 02-26.
Will be reenabled in M95, after experimentation.
</summary>
<token key="Type">
<variant name="Prediction" summary="prediction"/>
<variant name="Visual" summary="frame"/>
</token>
</histogram>
<histogram
name="Renderer.DelegatedInkTrail.PredictionExperiment{Number}.Frame{ScoreType}"
units="pixels" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Distance in pixels between a predicted point and the corresponding
interpolated real point that occurred at the prediction time.
The difference between {ScoreType} and Frame{ScoreType} is that the former
uses prediction time and the later uses frame time.
Part of a series of predicted point experimentation histograms. This
experiment predicts {Number} in the future.
</summary>
<token key="Number">
<variant name="0" summary="1 predicted point, 12ms"/>
<variant name="1" summary="2 predicted points, 6ms and 12ms"/>
<variant name="2" summary="1 predicted point, 6ms"/>
<variant name="3" summary="2 predicted points, 3ms and 6ms"/>
</token>
<token key="ScoreType">
<variant name="OverPrediction" summary="Positive score values"/>
<variant name="PredictionScore" summary="Score abolute value"/>
<variant name="UnderPrediction" summary="Negative score values"/>
</token>
</histogram>
<histogram
name="Renderer.DelegatedInkTrail.PredictionExperiment{Number}.WrongDirection"
enum="BooleanDirection" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Whether the prediction is in the same direction as the real trajectory at
the same prediction time.
Part of a series of predicted point experimentation histograms. This
experiment predicts {Number} in the future.
</summary>
<token key="Number">
<variant name="0" summary="1 predicted point, 12ms"/>
<variant name="1" summary="2 predicted points, 6ms and 12ms"/>
<variant name="2" summary="1 predicted point, 6ms"/>
<variant name="3" summary="2 predicted points, 3ms and 6ms"/>
</token>
</histogram>
<histogram
name="Renderer.DelegatedInkTrail.PredictionExperiment{Number}.{ScoreType}"
units="pixels" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Distance in pixels between a predicted point and the corresponding
interpolated real point that occurred at the prediction time.
Part of a series of predicted point experimentation histograms. This
experiment predicts {Number} in the future.
</summary>
<token key="Number">
<variant name="0" summary="1 predicted point, 12ms"/>
<variant name="1" summary="2 predicted points, 6ms and 12ms"/>
<variant name="2" summary="1 predicted point, 6ms"/>
<variant name="3" summary="2 predicted points, 3ms and 6ms"/>
</token>
<token key="ScoreType">
<variant name="OverPrediction" summary="Positive score values"/>
<variant name="PredictionScore" summary="Score abolute value"/>
<variant name="UnderPrediction" summary="Negative score values"/>
</token>
</histogram>
<histogram
name="Renderer.DelegatedInkTrail.PredictionExperiment{Number}.{Type}Jitter"
units="pixels" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Euclidean distance in pixels between two successive variations of delegated
ink trail deltas which are the difference beween a predicted delegated ink
point and its corresponding interpolated real position that occurred at
{Type} time.
Part of a series of predicted point experimentation histograms. This
experiment predicts {Number} in the future.
</summary>
<token key="Number">
<variant name="0" summary="1 predicted point, 12ms"/>
<variant name="1" summary="2 predicted points, 6ms and 12ms"/>
<variant name="2" summary="1 predicted point, 6ms"/>
<variant name="3" summary="2 predicted points, 3ms and 6ms"/>
</token>
<token key="Type">
<variant name="Prediction" summary="prediction"/>
<variant name="Visual" summary="frame"/>
</token>
</histogram>
<histogram name="Renderer.DelegatedInkTrail.{Renderer}.OutstandingPointsToDraw"
units="count" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Amount of points that are drawn per frame by the Delegated Ink API
{Renderer}. This metric can be used as a measurement of how much improvement
using the Delegated Ink API by telling us how many points ahead the
Delegated Ink API is from the JS API.
</summary>
<token key="Renderer" variants="DelegatedInkRenderer"/>
</histogram>
<histogram
name="Renderer.DelegatedInkTrail.{Renderer}.TimeFromDelegatedInkToApiPaint"
units="ms" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<improvement direction="HIGHER_IS_BETTER"/>
<summary>
Time in milliseconds between the first moment a point is drawn in the screen
via the Delegated Ink API using the {Renderer} and the call to
updateInkTrailStartPoint is made. This histogram assumes that the point
represented by the metadata is painted in the same frame that the metadata
is reported in. Note that not all points received as metadata will match a
point in the trail. Thus, not all metadata points will generate a histogram
fire.
</summary>
<token key="Renderer" variants="DelegatedInkRenderer"/>
</histogram>
<histogram name="Renderer.DelegatedInkTrail.{Renderer}.TimeToDrawPointsMillis"
units="ms" expires_after="2025-08-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<improvement direction="LOWER_IS_BETTER"/>
<summary>
Span of time in milliseconds between the creation of pointer events in the
browser process and sending the draw operation to the OS {Renderer}.
</summary>
<token key="Renderer" variants="DelegatedInkRenderer"/>
</histogram>
<histogram name="Renderer.Font.PrimaryFont.DomContentLoaded" units="ms"
expires_after="2025-01-05">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The time spent acquiring the primary font consumed in the main tread before
DomContentLoaded. On Windows, this includes IPC roundtrip to browser process
and DirectWriteFontProxy and DirectWrite calls.
</summary>
</histogram>
<histogram name="Renderer.Font.PrimaryFont.DomContentLoaded.Style" units="ms"
expires_after="2025-01-05">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The time spent acquiring the primary font consumed in style recalc, in the
main tread before DomContentLoaded. On Windows, this includes IPC roundtrip
to browser process and DirectWriteFontProxy and DirectWrite calls.
</summary>
</histogram>
<histogram name="Renderer.Font.PrimaryFont.FCP" units="ms"
expires_after="2025-01-05">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The time spent acquiring the primary font consumed in the main tread before
FCP. On Windows, this includes IPC roundtrip to browser process and
DirectWriteFontProxy and DirectWrite calls.
</summary>
</histogram>
<histogram name="Renderer.Font.PrimaryFont.FCP.Style" units="ms"
expires_after="2025-01-05">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The time spent acquiring the primary font consumed in style recalc, in the
main tread before FCP. On Windows, this includes IPC roundtrip to browser
process and DirectWriteFontProxy and DirectWrite calls.
</summary>
</histogram>
<histogram name="Renderer.Font.SystemFallback.DomContentLoaded" units="ms"
expires_after="2024-09-01">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The time spent finding the system fallback font consumed in the main tread
before DomContentLoaded. On Windows, this includes IPC roundtrip to browser
process and DirectWriteFontProxy and DirectWrite calls.
</summary>
</histogram>
<histogram name="Renderer.Font.SystemFallback.FCP" units="ms"
expires_after="2025-01-05">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The time spent finding the system fallback font consumed in the main tread
before FCP. On Windows, this includes IPC roundtrip to browser process and
DirectWriteFontProxy and DirectWrite calls.
</summary>
</histogram>
<histogram name="Renderer.Images.HasOverfetchedCappedPixels" enum="Boolean"
expires_after="2025-03-03">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Boolean indicating that OverfetchedCappedPixels was a positive integer.
Reported whenever an image is loaded and painted.
</summary>
</histogram>
<histogram name="Renderer.Images.HasOverfetchedPixels" enum="Boolean"
expires_after="2025-03-03">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Boolean indicating that OverfetchedPixels was a positive integer. Reported
whenever an image is loaded and painted.
</summary>
</histogram>
<histogram name="Renderer.Images.HasSizesAttributeMiss" enum="Boolean"
expires_after="2025-03-03">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Boolean indicating that SizesAttributeMiss was a positive integer. Only set
when the sizes attribute is present. Reported whenever an image is loaded
and painted.
</summary>
</histogram>
<histogram name="Renderer.Images.OverfetchedCappedPixels" units="px^2"
expires_after="2025-03-03">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The amount of HTMLImageElement image pixels squared (width * height) that
were fetched by the renderer, but eventually weren't needed, as the layout
dimensions of the image corrected to the screen's device pixel ratio (up to
a predefined cap) were smaller than the image pixels fetched. Reported
whenever an image is loaded and painted.
</summary>
</histogram>
<histogram name="Renderer.Images.OverfetchedPixels" units="px^2"
expires_after="2025-03-03">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The amount of HTMLImageElement image pixels squared (width * height) that
were fetched by the renderer, but eventually weren't needed, as the layout
dimensions of the image corrected to the screen's device pixel ratio were
smaller than the image pixels fetched. Reported whenever an image is loaded
and painted.
</summary>
</histogram>
<histogram name="Renderer.Images.SizesAttributeMiss" units="px"
expires_after="2025-03-03">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The amount of HTMLImageElement pixels by which the eventual layout width was
smaller than the width provided by the sizes attribute.
</summary>
</histogram>
<histogram name="Renderer.Layout.TextWrapBalance" units="ms"
expires_after="2024-12-30">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>The time spent to balance lines of paragraphs.</summary>
</histogram>
<histogram name="Renderer.Layout.TextWrapPretty" units="ms"
expires_after="2024-12-30">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>The time spent to optimize line breaks.</summary>
</histogram>
<histogram name="Renderer.PaintPreview.Capture.MainFrameBlinkCaptureDuration"
units="ms" expires_after="2025-02-16">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Records the number of milliseconds spent blocking Blink's main thread while
capturing the main frame. Reported whenever an image is loaded and painted.
</summary>
</histogram>
<histogram name="Renderer.PaintPreview.Capture.MainFrameSuccess"
enum="BooleanSuccess" expires_after="2024-12-30">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Records a boolean indicating whether a capture for the main frame succeeded.
</summary>
</histogram>
<histogram name="Renderer.PaintPreview.Capture.SubframeBlinkCaptureDuration"
units="ms" expires_after="2024-12-30">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Records the number of milliseconds spent blocking Blink's main thread while
capturing a subframe.
</summary>
</histogram>
<histogram name="Renderer.PaintPreview.Capture.SubframeSuccess"
enum="BooleanSuccess" expires_after="2024-12-30">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Records a boolean indicating whether a capture for a subframe succeeded.
</summary>
</histogram>
<histogram name="Renderer.Preload.UnusedResource" enum="BooleanSuccess"
expires_after="2025-02-16">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Counts the number of unused preloads per ResourceType.
Renderer.Preload.UnusedResource2.LinkPreload is an equivalent metric that
records as enum values.
Warning: an unexpected reorder happened in Apr 2021
(https://crrev.com/c/2798954). Please do not trust older reports.
</summary>
</histogram>
<histogram name="Renderer.Preload.UnusedResource2{Trigger}" enum="ResourceType"
expires_after="2025-02-02">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Counts the number of unused preloads created {Trigger} per ResourceType.
Unlike Renderer.Preload.UnusedResource, this records enum value as it is,
not BooleanSuccess.
This is recorded after the document completes.
</summary>
<token key="Trigger">
<variant name="" summary="regardless of whether link preload or not"/>
<variant name=".LinkPreload" summary="by link preload"/>
<variant name=".NoLinkPreload" summary="by not link preload"/>
</token>
</histogram>
<histogram name="Renderer.Preload.UnusedResourceCount" units="count"
expires_after="2025-02-10">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Counts the total number of unused preloads.
This is recorded after the document completes.
</summary>
</histogram>
<histogram name="Renderer.ProcessLifetime3{RendererHostedContentType}"
units="ms" expires_after="2025-02-02">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The time elapsed between the creation and destruction of a renderer.
Recorded for {RendererHostedContentType}.
</summary>
<token key="RendererHostedContentType" variants="RendererHostedContentType"/>
</histogram>
<histogram name="Renderer.ReduceSubresourceResponseIPC.DidNotifyBrowser"
enum="Boolean" expires_after="2025-01-05">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The histogram is being used to track the number of messages
(content::mojom::FrameHost::SubresourceResponseStarted IPCs) being sent by
the RenderFrameImpl to notify RenderFrameHostImpl of starting subresource
responses, versus the number of times it was deemed unnecessary due to the
kReduceSubresourceResponseStartedIPC feature.
</summary>
</histogram>
<histogram name="Renderer.RenderThreadImpl.Init" units="ms"
expires_after="2025-01-26">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
The time it takes RenderThreadImpl::Init() to run, recorded for each
renderer process.
</summary>
</histogram>
<histogram name="RendererScheduler.QueueingDuration.{Priority}Priority"
units="microseconds" expires_after="2025-01-05">
<owner>[email protected]</owner>
<owner>[email protected]</owner>
<summary>
Time between when a new task is added to a TaskQueue with priority
k{Priority}Priority, and when the task begins execution. For delayed (i.e.
timer-based) tasks, the initial timestamp is based on the ideal starting
time, rather than the time when the event loop notices that the timer has
expired.
Note: Not recorded on ~5% of Windows machines with low-resolution clocks.
</summary>
<token key="Priority">
<variant name="BestEffort"/>
<variant name="Control"/>
<variant name="ExtremelyHigh"/>
<variant name="High"/>
<variant name="HighContinuation"/>
<variant name="Highest"/>
<variant name="Low"/>
<variant name="LowContinaution"/>
<variant name="Normal"/>
<variant name="NormalContinuation"/>
<variant name="VeryHigh"/>
</token>
</histogram>
<histogram name="RendererScheduler.RendererMainThreadLoad5" units="%"
expires_after="2025-02-02">
<owner>[email protected]</owner>
<summary>
Renderer main thread load (percentage of time spent in tasks), reported in
one second chunks.
See http://bit.ly/chromium-renderer-main-thread-load-metric for details.
This metric is emitted when the renderer main thread task is completed or
renderer is backgrounded or foregrounded, at most once per second per
renderer amortized.
</summary>
</histogram>
<histogram name="RendererSyncIPC.ElapsedTime" units="ms" expires_after="M85">
<owner>[email protected]</owner>
<summary>
Roundtrip times for synchronous IPC calls from the renderer to browser.
</summary>
</histogram>
</histograms>
</histogram-configuration>
Codebase Browser
chromium