The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/af8ba97cd1a395bb097493d819cef8ea1888d447

commit af8ba97cd1a395bb097493d819cef8ea1888d447
Author: alancutter <alancutter@chromium.org>
Date: Wed Apr 13 09:21:12 2016

Fix Javascript bugs in animation LayoutTests

These tests were hitting unintended exceptions in their test code.
This patch corrects their errors.

BUG= 602886 

Review URL: https://codereview.chromium.org/1882383002

Cr-Commit-Position: refs/heads/master@{#386946}

[modify] https://crrev.com/af8ba97cd1a395bb097493d819cef8ea1888d447/third_party/WebKit/LayoutTests/animations/responsive/color-responsive.html
[modify] https://crrev.com/af8ba97cd1a395bb097493d819cef8ea1888d447/third_party/WebKit/LayoutTests/animations/svg-attribute-composition/svg-mode-composition.html
[modify] https://crrev.com/af8ba97cd1a395bb097493d819cef8ea1888d447/third_party/WebKit/LayoutTests/animations/svg-attribute-composition/svg-targetX-targetY-composition.html

Blink>ToolsTest renamed to Blink>Infra

Update: Uploaded a patch some time ago to make run-webkit-tests treat console warnings and errors as failures.
https://codereview.chromium.org/1885703003

Ran into problems with expected warnings that included nonces that differed with each run.

Hi Alan -- so this is still an issue, but it's not as easy to fix as originally hoped, since there a lots of tests that have console error messages before loading testharness.js, right?

And if we want to fix this, then we might first want to change all those tests that have console errors before testharness.js is loaded?

There are many tests where console error are intentional. For most of them we can just make expectation files, others have randomised output due to nonce tokens in the URL and are much harder to write expectations for. I should have kept a record of which tests they were, from memory I think they were in the http directory.

My guess is that progress on this patch is blocked on making run-webkit-tests use a seeded random function.

This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue.

Sorry for the inconvenience if the bug really should have been left as Available.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

I created https://jsbin.com/yisiwol/edit?html,output to test this, and unsurprisingly found that the test passed.

It's important that the behavior when running tests is as similar as possible in upstream wpt and in blink, so we'd want to discuss this in upstream first. alancutter@, if you still think this is important, can you file an issue at https://github.com/web-platform-tests/wpt/issues? Thanks!