Add canvas context type histogram

Add canvas context type histogram

Add histogram for tracking which names are being used to create canvas
rendering contexts.

BUG=112928
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=215952

[Mark P, 2013-08-02 14:46:13]

https://codereview.chromium.org/21835003/diff/1/tools/metrics/histograms/hist...
File tools/metrics/histograms/histograms.xml (right):

https://codereview.chromium.org/21835003/diff/1/tools/metrics/histograms/hist...
tools/metrics/histograms/histograms.xml:1090: The context type passed into the
getContext() canvas function. Used to track
This sounds icky.  Most histograms are emitted on a clearly defined basis: on
startup, on profile load, on navigation, when the user does X, when this network
event occurs, etc.  It's not obvious to me what causes a function like
getContext() to be called.  I can easily imagine another programmer to add
another call to a function named getContext() rather than, say, pass the context
down through six layers of function parameters.  I can imagine if you have
multiple calls in close proximity that some may be optimized away b by the
compiler.

Can you please make the histogram emit at a more definitive time?  (or argue all
my worries are inappropriate)

[Sami, 2013-08-02 15:30:39]

https://codereview.chromium.org/21835003/diff/1/tools/metrics/histograms/hist...
File tools/metrics/histograms/histograms.xml (right):

https://codereview.chromium.org/21835003/diff/1/tools/metrics/histograms/hist...
tools/metrics/histograms/histograms.xml:1090: The context type passed into the
getContext() canvas function. Used to track
On 2013/08/02 14:46:14, Mark P wrote:
> This sounds icky.  Most histograms are emitted on a clearly defined basis: on
> startup, on profile load, on navigation, when the user does X, when this
network
> event occurs, etc.  It's not obvious to me what causes a function like
> getContext() to be called.  I can easily imagine another programmer to add
> another call to a function named getContext() rather than, say, pass the
context
> down through six layers of function parameters.  I can imagine if you have
> multiple calls in close proximity that some may be optimized away b by the
> compiler.
> 
> Can you please make the histogram emit at a more definitive time?  (or argue
all
> my worries are inappropriate)

You're right, my first stab of the description wasn't very accurate. The object
here is to record which context type name is used to create each canvas
rendering context. For example, for a 2D rendering context we'll record the name
"2d", while a 3D context can use a set of aliases like "webkit-3d", "webgl" and
"experimental-webgl". This happens at most once per <canvas> element on the
page.

Here's the blink-side patch that records the data if it helps:
https://codereview.chromium.org/21861002/

[Mark P, 2013-08-02 15:40:58]

lgtm

https://codereview.chromium.org/21835003/diff/8001/tools/metrics/histograms/h...
File tools/metrics/histograms/histograms.xml (right):

https://codereview.chromium.org/21835003/diff/8001/tools/metrics/histograms/h...
tools/metrics/histograms/histograms.xml:1090: Records the context type names
used to create canvas rendering contexts.
for my edification (don't revise the comment):
Is this recorded (in effect) once only when the canvas is created?  Does
scrolling/resizing/etc. cause this to be emitted?

[Sami, 2013-08-02 16:12:51]

On 2013/08/02 15:40:58, Mark P wrote:
> lgtm
> 
>
https://codereview.chromium.org/21835003/diff/8001/tools/metrics/histograms/h...
> File tools/metrics/histograms/histograms.xml (right):
> 
>
https://codereview.chromium.org/21835003/diff/8001/tools/metrics/histograms/h...
> tools/metrics/histograms/histograms.xml:1090: Records the context type names
> used to create canvas rendering contexts.
> for my edification (don't revise the comment):
> Is this recorded (in effect) once only when the canvas is created?  Does
> scrolling/resizing/etc. cause this to be emitted?

Thanks. Yes, this only happens when the canvas rendering context is created for
the first time. Since context creation is generally a heavy-weight operation,
sites aren't expected to do it frequently. Resizing the canvas or scrolling have
no effect on the context and thus won't cause histogram events.

[commit-bot: I haz the power, 2013-08-06 16:12:05]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/21835003/8001

[commit-bot: I haz the power, 2013-08-06 16:12:10]

Failed to apply patch for tools/metrics/histograms/histograms.xml:
While running patch -p1 --forward --force --no-backup-if-mismatch;
  patching file tools/metrics/histograms/histograms.xml
  Hunk #2 succeeded at 15760 (offset 207 lines).
  Hunk #3 succeeded at 21135 with fuzz 1 (offset 296 lines).
  Hunk #4 FAILED at 20853.
  1 out of 4 hunks FAILED -- saving rejects to file
tools/metrics/histograms/histograms.xml.rej

Patch:       tools/metrics/histograms/histograms.xml
Index: tools/metrics/histograms/histograms.xml
diff --git a/tools/metrics/histograms/histograms.xml
b/tools/metrics/histograms/histograms.xml
index
fecd8a04c4665a07428dc687595e45e57840bd57..884e3c60c3ee59fbff98f723a68ba9fdcea617b0
100644
--- a/tools/metrics/histograms/histograms.xml
+++ b/tools/metrics/histograms/histograms.xml
@@ -1085,6 +1085,12 @@ other types of suffix sets.
   </summary>
 </histogram>
 
+<histogram name="Canvas.ContextType" enum="CanvasContextType">
+  <summary>
+    Records the context type names used to create canvas rendering contexts.
+  </summary>
+</histogram>
+
 <histogram name="Cellular.ActivationFailure">
   <summary>
     The count of cellular device activation failures (Chrome OS).
@@ -15547,6 +15553,13 @@ other types of suffix sets.
   <int value="1" label="Wiped out"/>
 </enum>
 
+<enum name="CanvasContextType" type="int">
+  <int value="0" label="2d"/>
+  <int value="1" label="webkit-3d"/>
+  <int value="2" label="experimental-webgl"/>
+  <int value="3" label="webgl"/>
+</enum>
+
 <enum name="ChannelLayout" type="int">
   <int value="0" label="CHANNEL_LAYOUT_NONE"/>
   <int value="1" label="CHANNEL_LAYOUT_UNSUPPORTED"/>
@@ -20826,6 +20839,12 @@ other types of suffix sets.
   <int value="2" label="Dismiss"/>
 </enum>
 
+<enum name="SimpleCache.FileDescriptorLimitStatus" type="int">
+  <int value="0" label="Unsupported"/>
+  <int value="1" label="Supported but failed"/>
+  <int value="2" label="Succeeded"/>
+</enum>
+
 <enum name="SimpleCacheHeaderSizeChange" type="int">
   <int value="0" label="Written for the first time"/>
   <int value="1" label="Rewritten with same size"/>
@@ -20834,12 +20853,6 @@ other types of suffix sets.
   <int value="4" label="Unexpected header stream write"/>
 </enum>
 
-<enum name="SimpleCache.FileDescriptorLimitStatus" type="int">
-  <int value="0" label="Unsupported"/>
-  <int value="1" label="Supported but failed"/>
-  <int value="2" label="Succeeded"/>
-</enum>
-
 <enum name="SimpleCacheIndexInitializeMethod" type="int">
   <int value="0" label="Directory Scan"/>
   <int value="1" label="Index File"/>

[commit-bot: I haz the power, 2013-08-06 16:23:59]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/21835003/16001

[commit-bot: I haz the power, 2013-08-06 17:25:19]

Retried try job too often on mac_rel for step(s) remoting_unittests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=mac_rel&nu...

[commit-bot: I haz the power, 2013-08-06 17:38:48]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/21835003/16001

[commit-bot: I haz the power, 2013-08-06 19:40:39]

Change committed as 215952