Fixed
Status Update
No update yet.
Comments
vi...@google.com
vi...@google.com
I am also on Catalina if it helps.
ma...@marcardar.com
I'm also affected but I'm on Windows10. can confirm it on version 4.0 Canary 9
ma...@marcardar.com
Happening on junit tests for me as well.
Android Studio 3.6 RC 1
Build #AI-192.7142.36.36.6071332, built on December 13, 2019
Runtime version: 1.8.0_212-release-1586-b4-5784211 x86_64
VM: OpenJDK 64-Bit Server VM by JetBrains s.r.o
macOS 10.15.2
GC: ParNew, ConcurrentMarkSweep
Memory: 1237M
Cores: 16
Registry: ide.new.welcome.screen.force=true, ide.tooltip.initialDelay=600
Non-Bundled Plugins: IdeaVIM, org.jetbrains.kotlin, com.squareup.sqldelight
Android Studio 3.6 RC 1
Build #AI-192.7142.36.36.6071332, built on December 13, 2019
Runtime version: 1.8.0_212-release-1586-b4-5784211 x86_64
VM: OpenJDK 64-Bit Server VM by JetBrains s.r.o
macOS 10.15.2
GC: ParNew, ConcurrentMarkSweep
Memory: 1237M
Cores: 16
Registry: ide.new.welcome.screen.force=true, ide.tooltip.initialDelay=600
Non-Bundled Plugins: IdeaVIM, org.jetbrains.kotlin, com.squareup.sqldelight
ma...@marcardar.com
I'm not yet able to repro this using MvRx project.
Any additional info any of you have to help us repro this would be useful. Such as:
- Is this happening to Java as well as Kotlin tests?
- How did you start the test run exactly? Through clicking the run button at the top bar, or the run button in the gutter next to the test class/method name, or through context menu in the project view?
- Please attach idea.log of when this reproes for you.
Thanks!
Any additional info any of you have to help us repro this would be useful. Such as:
- Is this happening to Java as well as Kotlin tests?
- How did you start the test run exactly? Through clicking the run button at the top bar, or the run button in the gutter next to the test class/method name, or through context menu in the project view?
- Please attach idea.log of when this reproes for you.
Thanks!
ma...@marcardar.com
1) I ran StateStoreTest by clicking the arrow next to the class name. Tests passed
2) I changed the expected value for the first assert in testSubscribeNotCalledForNoop from 1 to 2
3) I reran the tests with control + r. Tests passed
4) I reran them again and they correctly failed (because I changed the assert)
I also ran YourKit. Out of the last 2 runs, the 2nd to last ran with out of date sources and incorrectly passed. The last run incorrectly failed.
2) I changed the expected value for the first assert in testSubscribeNotCalledForNoop from 1 to 2
3) I reran the tests with control + r. Tests passed
4) I reran them again and they correctly failed (because I changed the assert)
I also ran YourKit. Out of the last 2 runs, the 2nd to last ran with out of date sources and incorrectly passed. The last run incorrectly failed.
vi...@google.com
Thanks, I'm able to repro this now.
ma...@marcardar.com
Any update or workaround here? It's extremely error prone to work when you don't know what version of the source Android Studio is using.
da...@google.com
This seems to be an issue where the Gradle plugin does not see up to date sources, maybe due to VFS freshness. Jerome, anyone in Build team can help take a look?
da...@google.com
The issue seems to be that the IDE is not saving files to disc. After changing assert (expecting it to fail), build output contains:
"> Task :mvrx:compileDebugUnitTestSources UP-TO-DATE"
This means that the file state did not change on disk (AGP&Gradle do not use Intellij VFS).
"> Task :mvrx:compileDebugUnitTestSources UP-TO-DATE"
This means that the file state did not change on disk (AGP&Gradle do not use Intellij VFS).
ma...@marcardar.com
The issue is in GradleBuildInvoker.executeTasks(Request) method. Because "request.isWaitForCompletion()" is true and the current thread is not a dispatch thread, we invoke "executor.queueAndWaitForCompletion()" without invoking "myDocumentManager.saveAllDocuments()" first. This means that we do not flush editor state to disc before Gradle is invoked. Consequently, build is using stale files.
Raluca, should we always invoke "myDocumentManager.saveAllDocuments()"?
Raluca, should we always invoke "myDocumentManager.saveAllDocuments()"?
da...@google.com
I think so, yes. It used to be the case, before Change I3b29b1a3bfc. And also, looking at IntelliJ's own handling of Gradle-based projects (ExternalSystemUtil), their incantation before calling Gradle is:
TransactionGuard.getInstance().assertWriteSafeContext(ModalityState.defaultModalityState());
ApplicationManager.getApplication().invokeAndWait(FileDocumentManager.getInstance()::saveAllDocuments);
da...@google.com
Thanks for confirming that. Bradley can take care of fixing this.
ma...@marcardar.com
Thanks for fixing this! Will it be cherry-picked into 4.0?
ma...@marcardar.com
Yes, it should be in 4.0 Beta 3. Thanks for the report. Please let me know if you still encounter any issues.
da...@google.com
Yes, it will be in our upcoming 4.0 Beta release likely 4.0 Beta 3.
ma...@marcardar.com
Amazing, thank you!
da...@google.com
Thank you! What's the status for AndroidStudio 3.6?
ma...@marcardar.com
This should be in the next point release (if there is one, 3.6.2).
da...@google.com
Still actual in latest version of AS
ro...@gmail.com
#20, can you provide more details with steps and a reproducable test project maybe ?
da...@google.com
Issue still exists in Android Studio Chipmunk | 2021.2.1 Patch 1
ku...@gmail.com
Same here.
Android Studio Chipmunk | 2021.2.1 Patch 1
Build #AI-212.5712.43.2112.8609683, built on May 19, 2022
Runtime version: 11.0.12+0-b1504.28-7817840 x86_64
VM: OpenJDK 64-Bit Server VM by JetBrains s.r.o.
macOS 12.5
GC: G1 Young Generation, G1 Old Generation
Memory: 2560M
Cores: 8
Registry: external.system.auto.import.disabled=true
Non-Bundled Plugins: org.jetbrains.kotlin (212-1.7.10-release-333-AS5457.46)
Android Studio Chipmunk | 2021.2.1 Patch 1
Build #AI-212.5712.43.2112.8609683, built on May 19, 2022
Runtime version: 11.0.12+0-b1504.28-7817840 x86_64
VM: OpenJDK 64-Bit Server VM by JetBrains s.r.o.
macOS 12.5
GC: G1 Young Generation, G1 Old Generation
Memory: 2560M
Cores: 8
Registry: external.system.auto.import.disabled=true
Non-Bundled Plugins: org.jetbrains.kotlin (212-1.7.10-release-333-AS5457.46)
da...@google.com
I am with you.
Build: AI-212.5712.43.2112.8609683, 202205181650,
AI-212.5712.43.2112.8609683, JRE 11.0.12+0-b1504.28-7817840x64 JetBrains s.r.o., OS Linux(amd64) v5.15.0-46-generic, screens 2560.0x1440.0
AS: Chipmunk | 2021.2.1 Patch 1; Kotlin plugin: 212-1.7.10-release-333-AS5457.46; Android Gradle Plugin: 7.2.1; Gradle: 7.3.3; Gradle JDK: version 11.0.12; NDK: from local.properties: (not specified), latest from SDK: (not found); LLDB: pinned revision 3.1 not found, latest from SDK: (package not found); CMake: from local.properties: (not specified), latest from SDK: (not found), from PATH: (not found)Source: send_feedback_icon
Description
Version used: 1.1.0
All modules has:
buildFeatures {
dataBinding = true
viewBinding = true
}
How to reproduce: Create 2 modules >>> create layout1 in module1 and create layout2 in module2 >>> Create Fragment1 in module1 with view which will be including layout1, layout2 >>> try to get binding >>> layout1 will work becouse it is in the same module as Fragment1, but layout2 will not work becouse of wrong binding generation
My example:
Authentication module has dependency to common module.
base_button.xml is in common module.
login_input_component.xml is in Authentication module so there is no problem.
But this error occures on loginButton binding.
It tries to find BaseButtonBinding in authentication module even if base_button.xml is in another module.
Cannot access class 'com.application.authentication.databinding.BaseButtonBinding'. Check your module classpath for missing or conflicting dependencies