Skip to content

Latest commit

 

History

History
691 lines (559 loc) · 29.6 KB

web_tests.md

File metadata and controls

691 lines (559 loc) · 29.6 KB
Raw

Web Tests (formerly known as "Layout Tests" or "LayoutTests")

Web tests are used by Blink to test many components, including but not limited to layout and rendering. In general, web tests involve loading pages in a test renderer (content_shell) and comparing the rendered output or JavaScript output against an expected output file.

This document covers running and debugging existing web tests. See the Writing Web Tests documentation if you find yourself writing web tests.

Note that we changed the term "layout tests" to "web tests". Please assume these terms mean the identical stuff. We also call it as "WebKit tests" and "WebKit layout tests".

"Web platform tests" (WPT) are the preferred form of web tests and are located at web_tests/external/wpt. Tests that should work across browsers go there. Other directories are for Chrome-specific tests only.

Note: if you are looking for a guide for the Web Platform Test, you should read "Web platform tests" (WPT). This document does not cover WPT specific features/behaviors.

Note: if you are looking for a guide for running the Web Platform Tests with Chrome, Chrome Android or WebView, you should read "Running Web Platform Tests with run_wpt_tests.py".

[TOC]

Running Web Tests

Supported Platforms

  • Linux
  • MacOS
  • Windows
  • Fuchsia

Android is not supported.

Initial Setup

Before you can run the web tests, you need to build the blink_tests target to get content_shell and all of the other needed binaries.

autoninja -C out/Default blink_tests

On Mac, you probably want to strip the content_shell binary before starting the tests. If you don't, you'll have 5-10 running concurrently, all stuck being examined by the OS crash reporter. This may cause other failures like timeouts where they normally don't occur.

strip ./out/Default/Content\ Shell.app/Contents/MacOS/Content\ Shell

Running the Tests

The test runner script is in third_party/blink/tools/run_web_tests.py.

To specify which build directory to use (e.g. out/Default, etc.) you should pass the -t or --target parameter. If no directory is specified, out/Release will be used. To use the built-in out/Default, use:

third_party/blink/tools/run_web_tests.py -t Default

*** promo

  • Windows users need to use third_party\blink\tools\run_web_tests.bat instead.
  • Linux users should not use testing/xvfb.py; run_web_tests.py manages Xvfb itself.

Tests marked as [ Skip ] in TestExpectations won't be run by default, generally because they cause some intractable tool error. To force one of them to be run, either rename that file or specify the skipped test on the command line (see below) or in a file specified with --test-list (however, --skip=always can make the tests marked as [ Skip ] always skipped). Read the Web Test Expectations documentation to learn more about TestExpectations and related files.

*** promo Currently only the tests listed in Default.txt are run on the Fuchsia bots, since running all web tests takes too long on Fuchshia. Most developers focus their Blink testing on Linux. We rely on the fact that the Linux and Fuchsia behavior is nearly identical for scenarios outside those covered by the smoke tests.

*** promo Similar to Fuchsia's case, the tests listed in [Mac.txt] (../../third_party/blink/web_tests/TestLists/Mac.txt) are run on older mac version bots. By doing this we reduced the resources needed to run the tests. This relies on the fact that the majority of web tests will behavior similarly on different mac versions.

To run only some of the tests, specify their directories or filenames as arguments to run_web_tests.py relative to the web test directory (src/third_party/blink/web_tests). For example, to run the fast form tests, use:

third_party/blink/tools/run_web_tests.py fast/forms

Or you could use the following shorthand:

third_party/blink/tools/run_web_tests.py fast/fo\*

*** promo Example: To run the web tests with a debug build of content_shell, but only test the SVG tests and run pixel tests, you would run:

third_party/blink/tools/run_web_tests.py -t Default svg

As a final quick-but-less-robust alternative, you can also just use the content_shell executable to run specific tests by using (example on Windows):

out\Default\content_shell.exe --run-web-tests <url>|<full_test_source_path>|<relative_test_path>

as in:

out\Default\content_shell.exe --run-web-tests \
    c:\chrome\src\third_party\blink\web_tests\fast\forms\001.html

or

out\Default\content_shell.exe --run-web-tests fast\forms\001.html

but this requires a manual diff against expected results, because the shell doesn't do it for you. It also just dumps the text result only (as the dump of pixels and audio binary data is not human readable). See Running Web Tests Using the Content Shell for more details of running content_shell.

To see a complete list of arguments supported, run:

third_party/blink/tools/run_web_tests.py --help

*** note Linux Note: We try to match the Windows render tree output exactly by matching font metrics and widget metrics. If there's a difference in the render tree output, we should see if we can avoid rebaselining by improving our font metrics. For additional information on Linux web tests, please see docs/web_tests_linux.md.

*** note Mac Note: While the tests are running, a bunch of Appearance settings are overridden for you so the right type of scroll bars, colors, etc. are used. Your main display's "Color Profile" is also changed to make sure color correction by ColorSync matches what is expected in the pixel tests. The change is noticeable, how much depends on the normal level of correction for your display. The tests do their best to restore your setting when done, but if you're left in the wrong state, you can manually reset it by going to System Preferences → Displays → Color and selecting the "right" value.

Test Harness Options

This script has a lot of command line flags. You can pass --help to the script to see a full list of options. A few of the most useful options are below:

Option Meaning
--debug Run the debug build of the test shell (default is release). Equivalent to -t Debug
--nocheck-sys-deps Don't check system dependencies; this allows faster iteration.
--verbose Produce more verbose output, including a list of tests that pass.
--reset-results Overwrite the current baselines (-expected.{png|txt|wav} files) with actual results, or create new baselines if there are no existing baselines.
--fully-parallel Run tests in parallel using as many child processes as the system has cores.
--driver-logging Print C++ logs (LOG(WARNING), etc).

Success and Failure

A test succeeds when its output matches the pre-defined expected results. If any tests fail, the test script will place the actual generated results, along with a diff of the actual and expected results, into src/out/Default/layout-test-results/, and by default launch a browser with a summary and link to the results/diffs.

The expected results for tests are in the src/third_party/blink/web_tests/platform or alongside their respective tests.

*** note Tests which use testharness.js do not have expected result files if all test cases pass.

A test that runs but produces the wrong output is marked as "failed", one that causes the test shell to crash is marked as "crashed", and one that takes longer than a certain amount of time to complete is aborted and marked as "timed out". A row of dots in the script's output indicates one or more tests that passed.

Test expectations

The TestExpectations file (and related files) contains the list of all known web test failures. See the Web Test Expectations documentation for more on this.

Testing Runtime Flags

There are two ways to run web tests with additional command-line arguments:

--flag-specific

third_party/blink/tools/run_web_tests.py --flag-specific=blocking-repaint

It requires that web_tests/FlagSpecificConfig contains an entry like:

{
  "name": "blocking-repaint",
  "args": ["--blocking-repaint", "--another-flag"]
}

This tells the test harness to pass --blocking-repaint --another-flag to the content_shell binary.

It will also look for flag-specific expectations in web_tests/FlagExpectations/blocking-repaint, if this file exists. The suppressions in this file override the main TestExpectations files. However, [ Slow ] in either flag-specific expectations or base expectations is always merged into the used expectations.

It will also look for baselines in web_tests/flag-specific/blocking-repaint. The baselines in this directory override the fallback baselines.

*** note BUILD.gn assumes flag-specific builders always runs on linux bots, so flag-specific test expectations and baselines are only downloaded to linux bots. If you need run flag-specific builders on other platforms, please update BUILD.gn to download flag-specific related data to that platform.

You can also use --additional-driver-flag to specify additional command-line arguments to content_shell, but the test harness won't use any flag-specific test expectations or baselines.

Virtual test suites

A virtual test suite can be defined in web_tests/VirtualTestSuites, to run a subset of web tests with additional flags, with virtual/<prefix>/... in their paths. The tests can be virtual tests that map to real base tests (directories or files) whose paths match any of the specified bases, or any real tests under web_tests/virtual/<prefix>/ directory. For example, you could test a (hypothetical) new mode for repainting using the following virtual test suite:

{
  "prefix": "blocking_repaint",
  "platforms": ["Linux", "Mac", "Win"],
  "bases": ["compositing", "fast/repaint"],
  "args": ["--blocking-repaint"]
}

This will create new "virtual" tests of the form virtual/blocking_repaint/compositing/... and virtual/blocking_repaint/fast/repaint/... which correspond to the files under web_tests/compositing and web_tests/fast/repaint, respectively, and pass --blocking-repaint to content_shell when they are run.

Note that you can run the tests with the following command line:

third_party/blink/tools/run_web_tests.py virtual/blocking_repaint/compositing \
  virtual/blocking_repaint/fast/repaint

These virtual tests exist in addition to the original compositing/... and fast/repaint/... tests. They can have their own expectations in web_tests/TestExpectations, and their own baselines. The test harness will use the non-virtual expectations and baselines as a fallback. If a virtual test has its own expectations, they will override all non-virtual expectations. Otherwise the non-virtual expectations will be used. However, [ Slow ] in either virtual or non-virtual expectations is always merged into the used expectations. If a virtual test is expected to pass while the non-virtual test is expected to fail, you need to add an explicit [ Pass ] entry for the virtual test.

This will also let any real tests under web_tests/virtual/blocking_repaint directory run with the --blocking-repaint flag.

The "platforms" configuration can be used to skip tests on some platforms. If a virtual test suites uses more than 5% of total test time, we should consider to skip the test suites on some platforms.

The "prefix" value should be unique. Multiple directories with the same flags should be listed in the same "bases" list. The "bases" list can be empty, in case that we just want to run the real tests under virtual/<prefix> with the flags without creating any virtual tests.

A virtual test suite can have an optional exclusive_tests field to specify all (with "ALL") or a subset of bases tests that will be exclusively run under this virtual suite. The specified base tests will be skipped. Corresponding virtual tests under other virtual suites that don't specify the tests in their exclusive_tests list will be skipped, too. For example (unrelated fields are omitted):

{
  "prefix": "v1",
  "bases": ["a"],
}
{
  "prefix": "v2",
  "bases": ["a/a1", "a/a2"],
  "exclusive_tests": "ALL",
}
{
  "prefix": "v3",
  "bases": ["a"],
  "exclusive_tests": ["a/a1"],
}

Suppose there are directories a/a1, a/a2 and a/a3, we will run the following tests:

Suite a/a1 a/a2 a/a3
base skipped skipped run
virtual/v1 skipped skipped run
virtual/v2 run run n/a
virtual/v3 run skipped run

In a similar manner, a virtual test suite can also have an optional skip_base_tests field to specify all (with "ALL") or a subset of bases tests that will be run under this virtual while the base tests will be skipped. This will not affect other virtual suites.

{
  "prefix": "v1",
  "bases": ["a/a1"],
}
{
  "prefix": "v2",
  "bases": ["a/a1"],
  "skip_base_tests": "ALL",
}

Suppose there are directories a/a1 and a/a2 we will run the following tests:

Suite a/a1 a/a2
base skipped run
virtual/v1 run n/a
virtual/v2 run n/a

Choosing between flag-specific and virtual test suite

For flags whose implementation is still in progress, flag-specific expectations and virtual test suites represent two alternative strategies for testing both the enabled code path and non-enabled code path. They are preferred to only setting a runtime enabled feature to status: "test" if the feature has substantially different code path from production because the latter would cause loss of test coverage of the production code path.

Consider the following when choosing between virtual test suites and flag-specific suites:

  • The waterfall builders and try bots will run all virtual test suites in addition to the non-virtual tests. Conversely, a flag-specific configuration won't automatically cause the bots to test your flag - if you want bot coverage without virtual test suites, you will need to follow these instructions.

  • Due to the above, virtual test suites incur a performance penalty for the commit queue and the continuous build infrastructure. This is exacerbated by the need to restart content_shell whenever flags change, which limits parallelism. Therefore, you should avoid adding large numbers of virtual test suites. They are well suited to running a subset of tests that are directly related to the feature, but they don't scale to flags that make deep architectural changes that potentially impact all of the tests.

  • Note that using wildcards in virtual test path names (e.g. virtual/blocking_repaint/fast/repaint/*) is not supported in run_web_tests.py command line , but you can still use virtual/blocking_repaint to run all real and virtual tests in the suite or virtual/blocking_repaint/fast/repaint/dir to run real or virtual tests in the suite under a specific directory.

*** note We can run a virtual test with additional flags. Both the virtual args and the additional flags will be applied. The fallback order of baselines and expectations will be: 1) flag-specific virtual, 2) non-flag-specific virtual, 3) flag-specific base, 4) non-flag-specific base

Running a New Flag-Specific Suite in CQ/CI

Assuming you have already created a FlagSpecificConfig entry:

  1. File a resource request (internal docs) for increased capacity in the chromium.tests swarming pool and wait for approval.
  2. Define a new dedicated Buildbot test suite with --flag-specific and possibly other special configurations (e.g., fewer shards).
  3. Add the Buildbot suite to the relevant *-blink-rel builder's composition suite first (example).
  4. Add the flag-specific step name to the relevant builder in builders.json. rebaseline-cl and the WPT importer will now create baselines for that suite.
  5. Rebaseline the new suite and add any necessary suppressions under FlagExpectations/.
  6. Enable the flag-specific suite for CQ/CI by adding the Buildbot suite to the desired builder. This could be an existing CQ builder like linux-rel or a dedicated builder like linux-blink-web-tests-force-accessibility-rel, which has customized location filters.

Tracking Test Failures

All bugs, associated with web test failures must have the Test-Layout label. Depending on how much you know about the bug, assign the status accordingly:

  • Unconfirmed -- You aren't sure if this is a simple rebaseline, possible duplicate of an existing bug, or a real failure
  • Untriaged -- Confirmed but unsure of priority or root cause.
  • Available -- You know the root cause of the issue.
  • Assigned or Started -- You will fix this issue.

When creating a new web test bug, please set the following properties:

  • Components: a sub-component of Blink
  • OS: All (or whichever OS the failure is on)
  • Priority: 2 (1 if it's a crash)
  • Type: Bug
  • Labels: Test-Layout

You can also use the Layout Test Failure template, which pre-sets these labels for you.

Debugging Web Tests

After the web tests run, you should get a summary of tests that pass or fail. If something fails unexpectedly (a new regression), you will get a content_shell window with a summary of the unexpected failures. Or you might have a failing test in mind to investigate. In any case, here are some steps and tips for finding the problem.

  • Take a look at the result. Sometimes tests just need to be rebaselined (see below) to account for changes introduced in your patch.
    • Load the test into a trunk Chrome or content_shell build and look at its result. (For tests in the http/ directory, start the http server first. See above. Navigate to http://localhost:8000/ and proceed from there.) The best tests describe what they're looking for, but not all do, and sometimes things they're not explicitly testing are still broken. Compare it to Safari, Firefox, and IE if necessary to see if it's correct. If you're still not sure, find the person who knows the most about it and ask.
    • Some tests only work properly in content_shell, not Chrome, because they rely on extra APIs exposed there.
    • Some tests only work properly when they're run in the web-test framework, not when they're loaded into content_shell directly. The test should mention that in its visible text, but not all do. So try that too. See "Running the tests", above.
  • If you think the test is correct, confirm your suspicion by looking at the diffs between the expected result and the actual one.
    • Make sure that the diffs reported aren't important. Small differences in spacing or box sizes are often unimportant, especially around fonts and form controls. Differences in wording of JS error messages are also usually acceptable.
    • third_party/blink/tools/run_web_tests.py path/to/your/test.html produces a page listing all test results. Those which fail their expectations will include links to the expected result, actual result, and diff. These results are saved to $root_build_dir/layout-test-results.
      • Alternatively the --results-directory=path/for/output/ option allows you to specify an alternative directory for the output to be saved to.
    • If you're still sure it's correct, rebaseline the test (see below). Otherwise...
  • If you're lucky, your test is one that runs properly when you navigate to it in content_shell normally. In that case, build the Debug content_shell project, fire it up in your favorite debugger, and load the test file either from a file: URL.
    • You'll probably be starting and stopping the content_shell a lot. In VS, to save navigating to the test every time, you can set the URL to your test (file: or http:) as the command argument in the Debugging section of the content_shell project Properties.
    • If your test contains a JS call, DOM manipulation, or other distinctive piece of code that you think is failing, search for that in the Chrome solution. That's a good place to put a starting breakpoint to start tracking down the issue.
    • Otherwise, you're running in a standard message loop just like in Chrome. If you have no other information, set a breakpoint on page load.
  • If your test only works in full web-test mode, or if you find it simpler to debug without all the overhead of an interactive session, start the content_shell with the command-line flag --run-web-tests, followed by the URL (file: or http:) to your test. More information about running web tests in content_shell can be found here.
    • In VS, you can do this in the Debugging section of the content_shell project Properties.
    • Now you're running with exactly the same API, theme, and other setup that the web tests use.
    • Again, if your test contains a JS call, DOM manipulation, or other distinctive piece of code that you think is failing, search for that in the Chrome solution. That's a good place to put a starting breakpoint to start tracking down the issue.
    • If you can't find any better place to set a breakpoint, start at the TestShell::RunFileTest() call in content_shell_main.cc, or at shell->LoadURL() within RunFileTest() in content_shell_win.cc.
  • Debug as usual. Once you've gotten this far, the failing web test is just a (hopefully) reduced test case that exposes a problem.

Debugging HTTP Tests

Note: HTTP Tests mean tests under web_tests/http/tests/, which is a subset of WebKit Layout Tests originated suite. If you want to debug WPT's HTTP behavior, you should read "Web platform tests" instead.

To run the server manually to reproduce/debug a failure:

third_party/blink/tools/run_blink_httpd.py

The web tests are served from http://127.0.0.1:8000/. For example, to run the test web_tests/http/tests/serviceworker/chromium/service-worker-allowed.html, navigate to http://127.0.0.1:8000/serviceworker/chromium/service-worker-allowed.html. Some tests behave differently if you go to 127.0.0.1 vs. localhost, so use 127.0.0.1.

To kill the server, hit any key on the terminal where run_blink_httpd.py is running, use taskkill or the Task Manager on Windows, or killall or Activity Monitor on macOS.

The test server sets up an alias to the web_tests/resources directory. For example, in HTTP tests, you can access the testing framework using src="/js-test-resources/js-test.js".

Tips

Check https://test-results.appspot.com/ to see how a test did in the most recent ~100 builds on each builder (as long as the page is being updated regularly).

A timeout will often also be a text mismatch, since the wrapper script kills the content_shell before it has a chance to finish. The exception is if the test finishes loading properly, but somehow hangs before it outputs the bit of text that tells the wrapper it's done.

Why might a test fail (or crash, or timeout) on buildbot, but pass on your local machine?

  • If the test finishes locally but is slow, more than 10 seconds or so, that would be why it's called a timeout on the bot.
  • Otherwise, try running it as part of a set of tests; it's possible that a test one or two (or ten) before this one is corrupting something that makes this one fail.
  • If it consistently works locally, make sure your environment looks like the one on the bot (look at the top of the stdio for the webkit_tests step to see all the environment variables and so on).
  • If none of that helps, and you have access to the bot itself, you may have to log in there and see if you can reproduce the problem manually.

Debugging DevTools Tests

  • Do one of the following:
    • Option A) Run from the chromium/src folder: third_party/blink/tools/run_web_tests.py --additional-driver-flag='--remote-debugging-port=9222' --additional-driver-flag='--remote-allow-origins=*' --additional-driver-flag='--debug-devtools' --timeout-ms=6000000
    • Option B) If you need to debug an http/tests/inspector test, start httpd as described above. Then, run content_shell: out/Default/content_shell --remote-debugging-port=9222 --additional-driver-flag='--remote-allow-origins=*' --additional-driver-flag='--debug-devtools' --run-web-tests http://127.0.0.1:8000/path/to/test.html
  • Open http://localhost:9222 in a stable/beta/canary Chrome, click the single link to open the devtools with the test loaded.
  • In the loaded devtools, set any required breakpoints and execute test() in the console to actually start the test.

NOTE: If the test is an html file, this means it's a legacy test so you need to add:

  • Add window.debugTest = true; to your test code as follows:

    window.debugTest = true;
function test() {
  /* TEST CODE */
}

Reproducing flaky inspector protocol tests

https://crrev.com/c/5318502 implemented logging for inspector-protocol tests. With this CL for each test in stderr you should see Chrome DevTools Protocol messages that the test and the browser exchanged.

You can use this log to reproduce the failure or timeout locally.

  • Prepare a log file and ensure each line contains one protocol message in the JSON format. Strip any prefixes or non-protocol messages from the original log.

  • Make sure your local test file version matches the version that produced the log file.

  • Run the test using the log file:

    third_party/blink/tools/run_web_tests.py -t Release \
 --additional-driver-flag="--inspector-protocol-log=/path/to/log.txt" \
 http/tests/inspector-protocol/network/url-fragment.js

Bisecting Regressions

You can use git bisect to find which commit broke (or fixed!) a web test in a fully automated way. Unlike bisect-builds.py, which downloads pre-built Chromium binaries, git bisect operates on your local checkout, so it can run tests with content_shell.

Bisecting can take several hours, but since it is fully automated you can leave it running overnight and view the results the next day.

To set up an automated bisect of a web test regression, create a script like this:

#!/bin/bash

# Exit code 125 tells git bisect to skip the revision.
gclient sync || exit 125
autoninja -C out/Debug -j100 blink_tests || exit 125

third_party/blink/tools/run_web_tests.py -t Debug \
  --no-show-results --no-retry-failures \
  path/to/web/test.html

Modify the out directory, ninja args, and test name as appropriate, and save the script in ~/checkrev.sh. Then run:

chmod u+x ~/checkrev.sh  # mark script as executable
git bisect start <badrev> <goodrev>
git bisect run ~/checkrev.sh
git bisect reset  # quit the bisect session

Rebaselining Web Tests

See How to rebaseline.

Known Issues

See bugs with the component Blink>Infra for issues related to Blink tools, include the web test runner.

  • If QuickTime is not installed, the plugin tests fast/dom/object-embed-plugin-scripting.html and plugins/embed-attributes-setting.html are expected to fail.
  • Fluent scrollbar rendering has some tweaks to geometry and behavior that are just for web tests. These are described in the Fluent Scrollbars Visual Spec under "Special rendering - Web tests". We'd like to remove them eventually (crbug.com/382298324).