Buildbot/Talos/Tests: Difference between revisions

Shouldn't have re-added what was deleted
(Sync with :rwood's changes from January 18, 2018)
(Shouldn't have re-added what was deleted)
Line 19: Line 19:
* [https://wiki.mozilla.org/Buildbot/Talos/Tests#tresize tresize]
* [https://wiki.mozilla.org/Buildbot/Talos/Tests#tresize tresize]
* [https://wiki.mozilla.org/Buildbot/Talos/Tests#sessionrestore.2Fsessionrestore_no_auto_restore.2Fsessionrestore_many_windows sessionrestore[_no_auto_restore/_many_windows]]
* [https://wiki.mozilla.org/Buildbot/Talos/Tests#sessionrestore.2Fsessionrestore_no_auto_restore.2Fsessionrestore_many_windows sessionrestore[_no_auto_restore/_many_windows]]
* [https://wiki.mozilla.org/Buildbot/Talos/Tests#CanvasMark tcanvasmark] (this is the one test that is a startup test but unique as we run 5 cycles and report a series of numbers)


=== Page Load Tests ===
=== Page Load Tests ===
Line 61: Line 60:


Tests which are imported benchmarks are:
Tests which are imported benchmarks are:
* [https://wiki.mozilla.org/Buildbot/Talos/Tests#CanvasMark Canvasmark]
* [https://wiki.mozilla.org/Buildbot/Talos/Tests#V8.2C_version_7 V8 Version 7]
* [https://wiki.mozilla.org/Buildbot/Talos/Tests#V8.2C_version_7 V8 Version 7]
* [https://wiki.mozilla.org/Buildbot/Talos/Tests#kraken Kraken]
* [https://wiki.mozilla.org/Buildbot/Talos/Tests#kraken Kraken]
Line 243: Line 241:
* None listed yet. If you fix a regression for this test and have some tips to share, this is a good place for them.
* None listed yet. If you fix a regression for this test and have some tips to share, this is a good place for them.


=== CanvasMark ===
* contact: :snorp, :milan, :jmaher
* source: https://github.com/kevinroast/CanvasMark
* type: [https://wiki.mozilla.org/Buildbot/Talos#Page_Load_Tests PageLoader]
* data: 5 cycles of the entire benchmark, each subtest will have 5 data points (see below)
* summarization:
** subtest: [https://wiki.mozilla.org/Buildbot/Talos/Data#ignore_first ignore first] data point, then take the [https://wiki.mozilla.org/Buildbot/Talos/Data#median median] of the remaining 4; [https://dxr.mozilla.org/mozilla-central/source/testing/talos/talos/test.py#l522 source: test.py]
** score: [https://wiki.mozilla.org/Buildbot/Talos/Data#Canvasmark_metric canvasmark metric] on all summarized subtests
* '''Higher is better'''
{| cellspacing="1" cellpadding="1" border="1"
|-
| Talos test name
| Description
|-
|tcanvasmark
|
|}
These tests run the third-party [http://www.kevs3d.co.uk/dev/canvasmark/ CanvasMark] benchmark suite, which measures the browser's ability to render a variety of canvas animations at a smooth framerate as the scenes grow more complex.
Results are a score "based on the length of time the browser was able to maintain the test scene at greater than 30 FPS, multiplied by a weighting for the complexity of each test type".
{| class="mw-collapsible mw-collapsed wikitable"
! Example data
|-
| <pre>
1;Asteroids - Vectors;704;672;713;738;796
2;Asteroids - Bitmaps- shapes- text;1105;1139;1002;1122;1050
3;Asteroids - Shapes- shadows- blending;740;731;775;725;732
4;Arena5 - Vectors- shadows- bitmaps- text;841;679;865;942;873
5;Plasma - Maths- canvas shapes;583;583;557;604;632
6;3D Rendering - Maths- polygons- image transforms;340;445;374;425;425
7;Pixel blur - Math- getImageData- putImageData;1409;1408;1415;1508;1335
</pre>
|}
==== Possible regression causes ====
* None listed yet. If you fix a regression for this test and have some tips to share, this is a good place for them.
=== cpstartup ===
=== cpstartup ===
* contact: :mconley
* contact: :mconley
160

edits