Get the Code: Checkout and Build Chromium DevTools front-end

In order to make changes to DevTools frontend, build, run, test, and submit changes, several workflows exist. Having depot_tools set up is a common prerequisite.

Standalone checkout

As a standalone project, Chrome DevTools frontend can be checked out and built independently from Chromium. The main advantage is not having to check out and build Chromium.

However, to run layout tests, you need to use the chromium checkout or integrated checkout.

Checking out source

To check out the source for DevTools frontend only, follow these steps:

mkdir devtools
cd devtools
fetch devtools-frontend

Build

To build, follow these steps:

cd devtools-frontend
gclient sync
npm run build

The resulting build artifacts can be found in out/Default/gen/front_end.

The build tools generally assume Default as the target (and out/Default as the build directory). You can pass -t <name> (or --target=<name>) to use a different target. For example

npm run build -- -t Debug

will build in out/Debug instead of out/Default. If the directory doesn‘t exist, it’ll automatically create and initialize it.

You can disable type checking (via TypeScript) by using the devtools_skip_typecheck argument in your GN configuration. This uses esbuild instead of tsc to compile the TypeScript files and generally results in much shorter build times.

Additionally, we now bundle files together by default in all builds, which has a build time cost. If you want an even fast fast build, you might want to opt out of bundling by setting devtools_bundle to false

gn gen out/fast-build --args="devtools_skip_typecheck=true devtools_bundle=false"

and use npm run build -- -t fast-build to build this target (you can of course also just change the Default target to skip bundling and type checking).

Rebuilding automatically

You can use

npm run build -- --watch

to have the build script watch for changes in source files and automatically trigger rebuilds as necessary.

Linux system limits

The watch mode uses inotify by default on Linux to monitor directories for changes. It‘s fairly common to encounter a system limit on the number of files you can monitor. For example, Ubuntu Lucid’s (64bit) inotify limit is set to 8192.

You can get your current inotify file watch limit by executing:

cat /proc/sys/fs/inotify/max_user_watches

When this limit is not enough to monitor all files inside a directory, the limit must be increased for the watch mode to work properly. You can set a new limit temporary with:

sudo sysctl fs.inotify.max_user_watches=524288
sudo sysctl -p

If you like to make your limit permanent, use:

echo fs.inotify.max_user_watches=524288 | sudo tee -a /etc/sysctl.conf
sudo sysctl -p

You may also need to pay attention to the values of max_queued_events and max_user_instances if you encounter any errors.

Update to latest

To update to latest tip of tree version:

git fetch origin; git checkout origin/main  # or, alternatively: git rebase-update
gclient sync

Out of sync dependencies and cross-repo changes

The revisions of git dependencies must always be in sync between the entry in DEPS and the git submodule. PRESUBMIT will reject CLs that try to submit changes to one but not the other. It can happen that dependencies go out of sync for three main reasons:

  1. The developer attempted a manual roll by only updating the DEPS file (which was the process before migrating to git submodules, see below),
  2. after switching branches or checking out new commit the developer didn't run gclient sync, or
  3. they are working across repositories including changes in both.

In the first case, follow the manual roll process. In the second case, running gclient sync is necessary. If the changes to the submodule versions were already added to any commits (this happens when commits were created using git add -A, for example), it's necessary to unstage them (for example using git checkout -p origin/main). The latter also applies in the third case: Create a CL excluding the dependency changes and a separate CL with a proper roll.

Run in a pre-built Chromium

You can run a build of DevTools frontend in a pre-built Chrome or Chromium in order to avoid the expensive build. Options:

  • Use the downloaded Chrome for Testing binary in third_party/chrome.
  • Use the latest Chrome Canary. This includes any DevTools features that are only available in regular Chrome builds (is_official_build + is_chrome_branded), such as GenAI-related features.

Using npm start script (recommended)

With Chromium 136, we added (back) a start script that can be used to easily launch DevTools with pre-built Chrome for Testing or Chrome Canary. It'll also take care of automatically enabling/disabling experimental features that are actively being worked on. Use

npm start

to build DevTools front-end in out/Default (you can change this to out/foo by passing --target=foo if needed), and open Chrome for Testing (in third_party/chrome) with the custom DevTools front-end. This will also monitor the source files for changes while Chrome is running and automatically trigger a rebuild whenever source files change.

By default, npm start will automatically open DevTools for every new tab, you can use

npm start -- --no-open

to disable this behavior. You can also use

npm start -- --browser=canary

to run in Chrome Canary instead of Chrome for Testing; this requires you to install Chrome Canary manually first (Googlers can install google-chrome-canary on gLinux). And finally use

npm start -- http://www.example.com

to automatically open http://www.example.com in the newly spawned Chrome tab. Use

npm start -- --verbose

to enable verbose logging, which among other things, also prints all output from Chrome to the terminal, which is otherwise suppressed.

Controlling the feature set

By default npm start will enable a bunch of experimental features (related to DevTools) that are considered ready for teamfood. To also enable experimental features that aren't yet considered sufficiently stable to enable them by default for the team, run:

# Long version
npm start -- --unstable-features

# Short version
npm start -- -u

Just like with Chrome itself, you can also control the set of enabled and disabled features using

npm start -- --enable-features=DevToolsWellKnown
npm start -- --disable-features=DevToolsWellKnown --enable-features=DevToolsFreestyler:multimodal/true

which you can use to override the default feature set.

Remote debugging

The npm start command also supports launching Chrome for remote debugging via

npm start -- --remote-debugging-port=9222

or

npm start -- --browser=canary --remote-debugging-port=9222 --user-data-dir=\`mktemp -d`

Note that you have to also pass the --user-data-dir and point it to a non-standard profile directory (a freshly created temporary directory in this example) for security reason when using any Chrome version except for Chrome for Testing. This article explains the reasons behind it.

Running from file system

This works with Chromium 79 or later. (Requires brew install coreutils on Mac.)

To run on Mac:

<path-to-devtools-frontend>./third_party/chrome/chrome-mac/Google\ Chrome\ for\ Testing.app/Contents/MacOS/Google\ Chrome\ for\ Testing --disable-infobars --disable-features=MediaRouter --custom-devtools-frontend=file://$(realpath out/Default/gen/front_end) --use-mock-keychain

To run on Linux:

<path-to-devtools-frontend>./third_party/chrome/chrome-linux/chrome --disable-infobars --custom-devtools-frontend=file://$(realpath out/Default/gen/front_end)

To run on Windows:

<path-to-devtools-frontend>\third_party\chrome\chrome-win\chrome.exe --disable-infobars --custom-devtools-frontend="<path-to-devtools-frontend>\out\Default\gen\front_end"

Note that $(realpath out/Default/gen/front_end) expands to the absolute path to build artifacts for DevTools frontend.

Open DevTools via F12 or Ctrl+Shift+J on Windows/Linux or Cmd+Option+I on Mac.

If you get errors along the line of Uncaught TypeError: Cannot read property 'setInspectedTabId' you probably specified an incorrect path - the path has to be absolute. On Mac and Linux, the file url will start with three slashes: file:///Users/....

Tip: You can inspect DevTools with DevTools by undocking DevTools and then opening a second instance of DevTools (see keyboard shortcut above).

Tip: On Windows it is possible the browser will re-attach to an existing session without applying command arguments. Make sure that there are no active Chrome sessions running.

Running from remote URL

This works with Chromium 85 or later.

Serve the content of out/Default/gen/front_end on a web server, e.g. via python -m http.server.

Then point to that web server when starting Chromium, for example:

<path-to-devtools-frontend>/third_party/chrome/chrome-<platform>/chrome --disable-infobars --custom-devtools-frontend=http://localhost:8000/

Open DevTools via F12 or Ctrl+Shift+J on Windows/Linux or Cmd+Option+I on Mac.

Running in hosted mode

Serve the content of out/Default/gen/front_end on a web server, e.g. via python3 -m http.server 8000.

Then start Chrome, allowing for accesses from the web server:

<path-to-devtools-frontend>/third_party/chrome/chrome-<platform>/chrome --disable-infobars --remote-debugging-port=9222 --remote-allow-origins=http://localhost:8000 about:blank

Get the list of pages together with their DevTools frontend URLs:

$ curl http://localhost:9222/json -s | grep '\(url\|devtoolsFrontend\)'
   "devtoolsFrontendUrl": "/devtools/inspector.html?ws=localhost:9222/devtools/page/BADADD4E55BADADD4E55BADADD4E5511",
   "url": "about:blank",

In a regular Chrome tab, go to the URL http://localhost:8000/inspector.html?ws=<web-socket-url>, where <web-socket-url> should be replaced by your desired DevTools web socket URL (from devtoolsFrontendUrl). For example, for "devtoolsFrontendUrl": "/devtools/inspector.html?ws=localhost:9222/devtools/page/BADADD4E55BADADD4E55BADADD4E5511", you could run the hosted DevTools with the following command:

$ google-chrome http://localhost:8000/inspector.html?ws=localhost:9222/devtools/page/BADADD4E55BADADD4E55BADADD4E5511

Chromium checkout

You can also work on the DevTools front-end within a full Chromium checkout. This workflow is particularly useful if you are working on a feature or a bug that spans back-end (Chromium C++ code) and front-end (DevTools TypeScript code), but it‘s also useful for Chromies that need to make small patches to DevTools front-end and don’t want to go through the process of setting up a dedicated devtools-frontend standalone checkout.

Checking out source

Follow instructions to check out Chromium. The DevTools front-end code is located inside the third_party/devtools-frontend/src/ folder (after running gclient sync).

Build

The instructions on the Chromium page apply as well here. In particular use

autoninja -C out/Default chrome

to build Chromium with the bundled DevTools front-end. You can also use

autoninja -C out/Default devtools_frontend_resources

to only build the DevTools front-end resources, which can afterwards be found in the out/Default/gen/third_party/devtools-frontend/src/front_end folder.

Run

Launch Chromium with the bundled DevTools front-end using

out/Default/chrome

or if you are only iterating on a small set of changes to the DevTools front-end and used the devtools_frontend_resources build target, you can run Chrome with the generated DevTools front-end artifacts using

out/Default/chrome --custom-devtools-frontend=file://$(realpath out/Default/gen/third_party/devtools-frontend/src/front_end)

afterwards, which can be quite a bit faster than building and linking the full Chromium binary.

Testing

To run the test suite, use npm test from within the DevTools front-end folder:

cd third_party/devtools-frontend/src
npm test

Juggling the git submodules

Working on DevTools within a Chromium checkout means working across two separate repositories (at the same time), which can be a bit tricky. Especially if you are working on a change that spans across the boundary of the front-end and the back-end, you‘ll need to cook two separate CLs. There are several ways to go about this, and in here, we’ll outline one somewhat well-lit path. You start by creating a branch in both Chromium and DevTools:

git new-branch my-change-backend
pushd third_party/devtools-frontend/src
git new-branch my-change-frontend
popd

Now you go about developing your patch, and commit individually to Chromium and DevTools repositories. When you're done, you can upload the changes individually using git cl upload:

git cl upload
pushd third_party/devtools-frontend/src
git cl upload
popd

The tricky part is to get the checkout back into a well-defined state, which can be accomplished using:

git checkout main
git -C third_party/devtools-frontend/src checkout \
  `gclient getdep -r src/third_party/devtools-frontend/src`
gclient sync

When you need to rebase your changes, also make sure to run the above commands and afterwards rebase the changes in the repositories separately:

git checkout my-change-backend
git rebase
pushd third_party/devtools-frontend/src
git checkout my-change-frontend
git rebase
popd

Integrated checkout

This solution is experimental, please report any trouble that you run into!

The integrated workflow offers the best of both worlds, and allows for working on both Chromium and DevTools frontend side-by-side.

A full Chromium checkout is a pre-requisite for the following steps.

Untrack the existing devtools-frontend submodule

First, you need to untrack the existing devtools-frontend submodule in the chromium checkout. This ensures that devtools isn't dragged along whenever you update your chromium dependencies.

In chromium/src, run gclient sync to make sure you have installed all required submodules.

gclient sync

Then, disable gclient sync for DevTools frontend inside of Chromium by editing .gclient config. From chromium/src/, run

vim "$(gclient root)/.gclient"

In the custom_deps section, insert this line:

"src/third_party/devtools-frontend/src": None,

For the integrated checkout, create a single gclient project that automatically gclient sync's all dependencies for both repositories. After checking out chromium, modify the .gclient file for chromium/src to add the DevTools project:

solutions = [
  {
    # Chromium src project
    "name": "src",
    "url": "https://chromium.googlesource.com/chromium/src.git",
    "custom_deps": {
      "src/third_party/devtools-frontend/src": None,
    },
  },
  {
    # devtools-frontend project
    "name": "devtools-frontend",
    "managed": False,
    "url": "https://chromium.googlesource.com/devtools/devtools-frontend.git",
  }
]

Do not run gclient sync now, first create the symlink. In the same directory as the .gclient file, run:

ln -s src/third_party/devtools-frontend/src devtools-frontend

If you did run gclient sync first, remove the devtools-frontend directory and start over.

Run gclient sync after creating the link to fetch the dependencies for the standalone checkout.