Commit message style guide

The Git project provides general guidelines which include how to compose commit messages. When writing a commit message, follow these guidelines:

General guidelines

  • Add an empty line between the capitalized first line (summary of the commit) and the additional detailed description.

  • If possible, limit the first line to 50 characters and the detailed description to 72 characters per line.

    • The detailed description is recommended but can be omitted if the change is well explained by the first line.
    • The 50 characters limit is not a hard limit. Especially when using multiple tags it can be hard to formulate a meaningful summary in 50 characters.

  • Make use of issue tracker integration, but not the in first line.

  • Use the imperative mood to summarize the commit. For example, "Fix memory leak in file.cc", not "I fixed a memory leak the file.cc".

  • Do not reference relative points in time, private URLs, individuals, private API keys, passwords, user names, etc.

Add required tag

The required [tag] in the first line helps readers of the commit tell what subject your change is about. The format is simply a keyword between brackets, for example, [docs]. The keyword has no specific meaning, but should help readers identify the subject easily. More specific tags or multiple tags can also be used to specify more fine-grained subjects, for example,[docs][fidl]. The following example shows required tags in the commit message subject:

[parent][component] Update component in Fuchsia

Write the details of a commit message here.

Bug: <issue-tracker-ID>

Test: Added test X.

You can view the commit history of the files you've edited to check for the tags used previously. See these examples:

Commit message tags are required. If the subject of a commit message doesn't include tags, Gerrit flags your change with Needs Label: Commit-Message-has-tags.

Add paragraph

The paragraph underneath the header line describes the change in better detail. Ensure the reason and intention of the change are clear: for example:

[docs] Adding Fuchsia Commit message style guide

This change centralizes all commit message style guide into one style
guide. It also removes duplicate content from existing pages and points
to the new style guide instead.

Change-Id: I307e5b24df4273661d22c52c81038de50600c76c

Add bug

If you want Fuchsia Gerrit to know what issue this change is associated with, you need to add the Bug: <issue-tracker-ID> line. To associate multiple issues with a change, list each bug in a separate line. For example:

[parent][component] Update component in Fuchsia

Write the details of a commit message here.

Bug: 82657
Bug: 82658

Test: Added test X.

The difference between Bug: and Fixed: is that Fixed: automatically closes the issue for you once your change is submitted, whereas Bug: only comments on your issue once submitted. If you have multiple changes attached to your issue, use the Bug: tag for all the changes up until the final change. Use Fixed: on the final change, so that the issue is closed.

Indicate multiple steps

When executing a change that requires multiple steps across various repositories (for instance, to soft transition APIs defined in one repository and used in others), indicate multiple steps by referencing the last step taken and the next step taken.

This enables reviewers and those looking at the log to understand and navigate the totality of the change. When possible, it is encouraged to provide all steps to complete the migration in each commit log (but that may be impractical in some cases).

Here's an example of a commit message with multiple steps:

[fidl][go] Support for flexible enums (1/3)

Step 1 of 3. Adds support for flexible enums to fidlgen_go:

* For all enums, emit `IsUnknown()` method indicating whether the
  value is unknown (or known). While relevant only for flexible enums,
  it is possible to construct unknown strict enums using type casts.
* Emit an internal method `I_EnumIsStrict` indicating whether the enum
  is strict or flexible. This method is read by the runtime, when
  creating enum marshaler.
* For flexible enums, we generate a default unknown placeholder

Step 2: I1102f244aa5ab4545fab21218c1da90be08604ec
Step 3: If0a047a4db804a183e984676217b31e17b4af0ea

Test: fx test fidl_go_conformance at If0a047a4db804a183e984676217b31e17b4af0ea

Change-Id: Id71eb879e4d7dfabe228cc7b4e2fedb7f52db7b7

Add tests and multiply line

The Test: line is necessary to indicate what type of test to run to make sure your change is working. You can add multiple different tests in this line, for example, fx test setui_service_tests, setui_client_interface_test. You can also add explanations of what tests you added below. If you did not add or modify tests, you can specify None:, with an explanation of why it doesn't need to be tested, for example, None: documentation change only.

If you added new tests, you can get deflake runs automatically by adding the Multiply: line with the test to run multiple times.

The following example shows Test: and Multiply: in the commit message:

[SetUI] Correct internal items' visibility Part II

This CL marks some internal items with pub(crate), pub(super) or leaves
as private. Related CL: fxr/535942

Bug: 72941
Test: fx test -o setui_service_tests setui_client_interface_test
sample-setui-config-test setting-service-config-test
Multiply: setui_service_tests
Multiply: setui_client_interface_test

Change-Id: I67e061edee1e81a6875bf26b752ba5687c4ced71

If the testing instructions are complex, create an issue and provide a link to that issue in the change description. If the change doesn't intend to change behavior, indicate that fact in the commit message.

In some cases, certain behavior changes cannot be tested because Fuchsia lacks some particular piece of infrastructure. If so, create an issue in the tracker about the necessary infrastructure support and provide the bug number in the change description, in addition to describing how the change is tested manually, for example:

Test: Manually tested that [...]. Automated testing needs US-XXXX.

Developers are responsible for high-quality automated testing of their code. Reviewers are responsible for pushing back on changes that do not include sufficient tests. See Fuchsia testability rubrics for more information on how to introduce testable and tested code in the Fuchsia project.

For more commit message options, see the Commit message options guide.

Add a buffer line before Change-Id

The Change-Id gets added automatically when you commit your changes. You may want to add a buffer line between your other lines and the Change-Id, in case you use git commit --amend, so that Gerrit doesn't parse the Change-Id as a regular line, and add a second ID to it.

For more specific details, see the Git interpret trailer rules.

Use Change-Id to refer to related changes

To reference another Gerrit change in a commit message, always use the Change-Id.

Using the Change-Id is preferred since:

The git SHA is only known after a change is merged, and while guidance could be given to use the Change-Id in one case, and the git SHA in the other, prefer uniform guidance. Furthermore, you cannot reference other repositories using the git SHA.

The link to the change is assigned by Gerrit, and is not part of the persistent history of the repository. Should the review mechanism change, the Change-Id will continue to be part of the recorded history, whereas the change's number will not. There are also rare occurrences where change numbers may be lost, for example, due to re-indexing issues.

For instance, to refer to the change that added RFC-0042, use I32b966810d21a249647887fa45b61720ad01714c, and not the git SHA 5d40ee8c42d1b0e4d8b690786da12a0a947c1aaa or the link to the change, https://fuchsia-review.googlesource.com/c/fuchsia/+/284569.