Notable changes:
* Allows macrobenchmarks to run via `flutter run`.
* Splits macrobenchmarking between "orchestration logic" and "app
serving" (served on separate ports on the same machine to keep the
scheme consistent with the `flutter build` path).
* Adds an intercepted entrypoint for web benchmarks. We can't pass flags
to the app since it's not supported, so I've hard-coded the
orchestration server's port.
* Adding logic to connect to an existing Chrome debugger instance (vs
spawning one) for benchmarks.
<!-- start_original_pr_link -->
Reverts: flutter/flutter#165749
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: matanlurey
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: Still passing command-line arguments from recipes
that have no effect but cause the runner to crash.
<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: matanlurey
<!-- end_original_pr_author -->
<!-- start_reviewers -->
Reviewed By: {jtmcdole}
<!-- end_reviewers -->
<!-- start_revert_body -->
This change reverts the following previous change:
Partial re-land of https://github.com/flutter/flutter/pull/165628:
- Fixes the mistake that the `Cocoon` class did things that well, were
not specific to Cocoon.
- Renamed to `MetricsResultWriter`, as that is all it does now.
---
Closes https://github.com/flutter/flutter/issues/165618.
The `devicelab/bin/test_runner.dart upload-metrics` command use to have
_two_ responsibilities:
- Well, upload test **metrics** (benchmarks) to Skia Perf (it still does
that)
- Upload test **status** to Cocoon (it did until
https://github.com/flutter/flutter/pull/165614)
As https://github.com/flutter/flutter/pull/165614 proved, this API
predated the current LUCI setup, where Cocoon itself receives task
status updates from LUCI, and it turns out this entire time, DeviceLab
was making (at best) NOP calls, and at worst, causing crashes and
corrupt data (https://github.com/flutter/flutter/issues/165610).
<!-- end_revert_body -->
Co-authored-by: auto-submit[bot] <flutter-engprod-team@google.com>
<!-- start_original_pr_link -->
Reverts: flutter/flutter#165628
<!-- end_original_pr_link -->
<!-- start_initiating_author -->
Initiated by: jtmcdole
<!-- end_initiating_author -->
<!-- start_revert_reason -->
Reason for reverting: breaking tree:
https://ci.chromium.org/ui/p/flutter/builders/prod/Linux_mokey%20backdrop_filter_perf__e2e_summary/3833/overview
<!-- end_revert_reason -->
<!-- start_original_pr_author -->
Original PR Author: matanlurey
<!-- end_original_pr_author -->
<!-- start_reviewers -->
Reviewed By: {jtmcdole, yjbanov}
<!-- end_reviewers -->
<!-- start_revert_body -->
This change reverts the following previous change:
🚫 **Do not merge** until
https://flutter-review.googlesource.com/c/recipes/+/64220/1 is merged.
---
Closes https://github.com/flutter/flutter/issues/165618.
The `devicelab/bin/test_runner.dart upload-metrics` command use to have
_two_ responsibilities:
- Well, upload test **metrics** (benchmarks) to Skia Perf (it still does
that)
- Upload test **status** to Cocoon (it did until
https://github.com/flutter/flutter/pull/165614)
As https://github.com/flutter/flutter/pull/165614 proved, this API
predated the current LUCI setup, where Cocoon itself receives task
status updates from LUCI, and it turns out this entire time, DeviceLab
was making (at best) NOP calls, and at worst, causing crashes and
corrupt data (https://github.com/flutter/flutter/issues/165610).
In other words, this is removing entirely dead/unused code (though the
recipes have to be updated first).
/cc @jason-simmons as I need Jason's help reviewing the recipes change
and want to provide context.
<!-- end_revert_body -->
Co-authored-by: auto-submit[bot] <flutter-engprod-team@google.com>
This auto-formats all *.dart files in the repository outside of the
`engine` subdirectory and enforces that these files stay formatted with
a presubmit check.
**Reviewers:** Please carefully review all the commits except for the
one titled "formatted". The "formatted" commit was auto-generated by
running `dev/tools/format.sh -a -f`. The other commits were hand-crafted
to prepare the repo for the formatting change. I recommend reviewing the
commits one-by-one via the "Commits" tab and avoiding Github's "Files
changed" tab as it will likely slow down your browser because of the
size of this PR.
---------
Co-authored-by: Kate Lovett <katelovett@google.com>
Co-authored-by: LongCatIsLooong <31859944+LongCatIsLooong@users.noreply.github.com>
Fix the `--ab` option in the benchmark harness:
- Make `--local-engine-host` optional. The web engine doesn't need it, so it doesn't build it. But the tool crashes by failing to find it.
- Disable icon tree shaking because `--ab` runs against local engine build, whose Dart kernel version is out of sync with the framework, which crashes the const finder.
Reverts: flutter/flutter#154061
Initiated by: QuncCccccc
Reason for reverting: might be the reason that cause Framework tree to red
Original PR Author: bartekpacia
Reviewed By: {gmackall, reidbaker}
This change reverts the following previous change:
This PR resolves#151166
This pull request aims to improve code readability, based on feedback gathered in a recent design doc.
<br>
There are two factors that hugely impact how easy it is to understand a piece of code: **verbosity** and **complexity**.
Reducing **verbosity** is important, because boilerplate makes a project more difficult to navigate. It also has a tendency to make one's eyes gloss over, and subtle typos/bugs become more likely to slip through.
Reducing **complexity** makes the code more accessible to more people. This is especially important for open-source projects like Flutter, where the code is read by those who make contributions, as well as others who read through source code as they debug their own projects.
<hr>
<br>
The following examples show how pattern-matching might affect these two factors:
<details> <summary><h3>Example 1 (GOOD)</h3> [click to expand]</summary>
```dart
if (ancestor case InheritedElement(:final InheritedTheme widget)) {
themes.add(widget);
}
```
Without using patterns, this might expand to
```dart
if (ancestor is InheritedElement) {
final InheritedWidget widget = ancestor.widget;
if (widget is InheritedTheme) {
themes.add(widget);
}
}
```
Had `ancestor` been a non-local variable, it would need to be "converted" as well:
```dart
final Element ancestor = this.ancestor;
if (ancestor is InheritedElement) {
final InheritedWidget inheritedWidget = ancestor.widget;
if (widget is InheritedTheme) {
themes.add(theme);
}
}
```
</details>
<details> <summary><h3>Example 2 (BAD) </h3> [click to expand]</summary>
```dart
if (widget case PreferredSizeWidget(preferredSize: Size(:final double height))) {
return height;
}
```
Assuming `widget` is a non-local variable, this would expand to:
```dart
final Widget widget = this.widget;
if (widget is PreferredSizeWidget) {
return widget.preferredSize.height;
}
```
<br>
</details>
In both of the examples above, an `if-case` statement simultaneously verifies that an object meets the specified criteria and performs a variable assignment accordingly.
But there are some differences: Example 2 uses a more deeply-nested pattern than Example 1 but makes fewer useful checks.
**Example 1:**
- checks that `ancestor` is an `InheritedElement`
- checks that the inherited element's `widget` is an `InheritedTheme`
**Example 2:**
- checks that `widget` is a `PreferredSizeWidget`
(every `PreferredSizeWidget` has a `size` field, and every `Size` has a `height` field)
<br>
<hr>
I feel hesitant to try presenting a set of cut-and-dry rules as to which scenarios should/shouldn't use pattern-matching, since there are an abundance of different types of patterns, and an abundance of different places where they might be used.
But hopefully the conversations we've had recently will help us converge toward a common intuition of how pattern-matching can best be utilized for improved readability.
<br><br>
- resolves https://github.com/flutter/flutter/issues/152313
- Design Doc: [flutter.dev/go/dart-patterns](https://flutter.dev/go/dart-patterns)
Prior to this fix, `_TaskRunner.run` would immediately cleanup the
keep-alive port once the task completed, which would result in the
isolate shutting down as soon as the task result was returned from
`ext.cocoonRunTask` callback in the form of a
`ServiceExtensionResponse`. Since the service extension response is
processed by the service isolate, it was possible for the VM to start
shutting down before the service isolate could send the task result data
back to the task runner.
This change introduces a new service extension,
`ext.cocoonTaskResultReceived`, that the task runner invokes after it
receives the task result from `ext.cocoonRunTask`, notifying the task
process that it can close the keep-alive port and shutdown.
Fixes https://github.com/flutter/flutter/issues/155475
Reverts: flutter/flutter#152487
Initiated by: gmackall
Reason for reverting: I forgot that I need to override the compileSdkVersion in the AGP 8.0 [instance of this test](ef9cd32f5a/dev/devicelab/bin/tasks/android_java17_dependency_smoke_tests.dart (L19))
Original PR Author: gmackall
Reviewed By: {reidbaker}
This change reverts the following previous change:
Updates `compileSdk`, `targetSdk`, and `ndk` versions (former 2 to latest, latter to the version of the ndk we are hosting on CIPD).
Summary of changes:
- Updates mentioned template values
- `compileSdk` 35 requires AGP 8.0+, so updated to 8.1 in many places.
- This also necessitated Gradle upgrades in most places
- This also necessitated moving the `package` xml attribute to the AGP `namespace` field in a couple places (test + template).
- Some tests use the output of `flutter create` but then use intentionally lower AGP versions. [I downgraded the `compileSdk` in these tests.](fee34fd61a)
- [Stopped lockfile generation](82324a2570) script from hitting the `hello_world` example because it uses `.kts` gradle files.
- One test needed [some Gradle options we had already added to templates](6aa187b4b6).
Updates `compileSdk`, `targetSdk`, and `ndk` versions (former 2 to latest, latter to the version of the ndk we are hosting on CIPD).
Summary of changes:
- Updates mentioned template values
- `compileSdk` 35 requires AGP 8.0+, so updated to 8.1 in many places.
- This also necessitated Gradle upgrades in most places
- This also necessitated moving the `package` xml attribute to the AGP `namespace` field in a couple places (test + template).
- Some tests use the output of `flutter create` but then use intentionally lower AGP versions. [I downgraded the `compileSdk` in these tests.](fee34fd61a)
- [Stopped lockfile generation](82324a2570) script from hitting the `hello_world` example because it uses `.kts` gradle files.
- One test needed [some Gradle options we had already added to templates](6aa187b4b6).
As in https://github.com/flutter/flutter/pull/152696, It's not clear
that DDS initialization is necessary for any of these tests.
In particular, from the comments on the tool flag:
```
'It may be necessary to disable this when attaching to an application with '
'an existing DDS instance (e.g., attaching to an application currently '
'connected to by "flutter run"), or when running certain tests.\n'
'Disabling this feature may degrade IDE functionality if a DDS instance is '
'not already connected to the target application.'
```
These tests are unlikely to need IDE functionality. I have initially
modified the Linux_android_emu tests in the ci.yaml so that they run in
presubmit to verify that this change is safe. Many other postsubmit-only
tests that will now be passed `--no-dds` require physical devices, and
if there is an issue in post-submit, this change will need to be
reverted.
Sets up tests that verify we can build a fresh counter app across our Gradle/AGP/Kotlin support range.
Post submit only, because the suite takes ~30 minutes to run, and I expect it to be _somewhat_ rare that we break only one of these versions (and therefore it doesn't get caught by existing presubmits).
Reverts: flutter/flutter#152049
Initiated by: cbracken
Reason for reverting: iOS builds failing in post-submit
Original PR Author: loic-sharma
Reviewed By: {jmagman}
This change reverts the following previous change:
Changes:
1. Enables Swift Package Manager by default on the main/master channel
2. Fixes tests that fail if Swift Package Manager is enabled
Corresponding docs change: https://github.com/flutter/website/pull/10938
Addresses https://github.com/flutter/flutter/issues/151567
Fixes https://github.com/flutter/flutter/issues/150093
New tests added to cover that we at least pass the arguments we expect to adb.
The test for #150093 is not ideal in that it does not verify the behavior of a failed process but instead ensures we set the parameter that contains the behavior we want.
devicelab code and tests are not setup to enable fake process or fake output from stdin/stderr and hang if adb or no hardware are present.
This PR is step 12 in the journey to solve issue #136139 and make the
entire Flutter repo more readable.
Most of it involves implementing switch expressions, and there's also a
few other random things that I wanted to clean up a bit.
Originally, my aim was just to refactor (as per usual), but while messing around with the `TableBorder.symmetric` constructor, I realized that `borderRadius` was missing!
This pull request makes a few class constructors more efficient, and it fixes#144277 by adding the missing parameter.
<br>
This pull request fixes#143803 by taking advantage of Dart's null-aware operators.
And unlike `switch` expressions ([9 PRs](https://github.com/flutter/flutter/pull/143634) and counting), the Flutter codebase is already fantastic when it comes to null-aware coding. After refactoring the entire repo, all the changes involving `?.` and `??` can fit into a single pull request.
When the Dart VM is not found within 10 minutes in CI on CoreDevices (iOS 17+), stop the app and upload the logs from DerivedData. The app has to be stopped first since the logs are not put in DerivedData until it's stopped.
Also, rearranged some logic to have CoreDevice have its own function for Dart VM url discovery.
Debugging for https://github.com/flutter/flutter/issues/142448.
Support for FFI calls with `@Native external` functions through Native assets on Android. This enables bundling native code without any build-system boilerplate code.
For more info see:
* https://github.com/flutter/flutter/issues/129757
### Implementation details for Android.
Mainly follows the design of the previous PRs.
For Android, we detect the compilers inside the NDK inside SDK.
And bundling of the assets is done by the flutter.groovy file.
The `minSdkVersion` is propagated from the flutter.groovy file as well.
The NDK is not part of `flutter doctor`, and users can omit it if no native assets have to be build.
However, if any native assets must be built, flutter throws a tool exit if the NDK is not installed.
Add 2 app is not part of this PR yet, instead `flutter build aar` will tool exit if there are any native assets.
