# Cuda Comment
cuda_comment: >
   From the template it looks like you are installing **TensorFlow** (TF) prebuilt binaries:
      * For TF-GPU - See point 1
      * For TF-CPU - See point 2
   -----------------------------------------------------------------------------------------------
   
   **1. Installing **TensorFlow-GPU** (TF) prebuilt binaries**
   
   
   Make sure you are using compatible TF and CUDA versions.

import org.gradle.api.internal.resolve.ProjectModelResolver;
import org.gradle.internal.service.Provides;
import org.gradle.internal.service.ServiceRegistrationProvider;
import org.gradle.nativeplatform.internal.prebuilt.PrebuiltLibraryBinaryLocator;

import java.util.ArrayList;
import java.util.List;

public class NativeDependencyResolverServices implements ServiceRegistrationProvider {
    @Provides

    finalizedBy 'largeNativeBuildPrebuilt'
}

performanceTest.registerTestProject("largeNativeBuildPrebuilt", GradleBuild) {
    buildFile = file("$largeNativeBuild.outputDirectory/prebuilt/util/build.gradle")
    dir = file("$largeNativeBuild.outputDirectory/prebuilt/util")
    tasks = ['build']
}

performanceTest.registerTestProject("archivePerformanceProject", Copy) {
    into "build/$name"
    from "src/templates/$name"
}

                implementation files('a.thing', additionalFile.output)
            }
        """
    }

    def "task input file collection can include project dependencies, external dependencies, prebuilt file dependencies and task output file dependencies"() {
        def configurationCache = newConfigurationCacheFixture()

        setupBuildWithEachDependencyType()
        taskTypeLogsInputFileCollectionContent()

H, tkunic, Tod, vihanjain, Yan Facai (颜发才), Yin Li, Yong Tang, Yukun Chen,
Yusuke Yamada

# Release 1.6.0

## Breaking Changes

*   Prebuilt binaries are now built against CUDA 9.0 and cuDNN 7.
*   Prebuilt binaries will use AVX instructions. This may break TF on older
    CPUs.

## Major Features And Improvements

Incremental Builds are a great optimization that helps avoid work already done.
If a developer continuously changes a single file, there is likely no need to rebuild all the other files in the project.

However, what happens when the same developer switches to a new branch created last week?
The files are rebuilt, even though the developer is building something that has been built before.

This is where a *build cache* is helpful.

Class <org.gradle.nativeplatform.internal.prebuilt.DefaultPrebuiltLibraries> is not annotated (directly or via its package) with @org.gradle.api.NonNullApi in (DefaultPrebuiltLibraries.java:0)
Class <org.gradle.nativeplatform.internal.prebuilt.DefaultPrebuiltLibrary> is not annotated (directly or via its package) with @org.gradle.api.NonNullApi in (DefaultPrebuiltLibrary.java:0)

However, what happens when the same developer decides to switch to a new branch created last week?
The files are rebuilt, even though the developer is building something that has been built before.

This is where a Build Cache is helpful. The cache stores previous build results, and greatly reduces the need to rebuild things when they have already been built locally.

        }
        FileLock fileLock = lockManager.lock(lockTarget, lockOptions, cacheDisplayName);
        try {
            boolean rebuild = initializationAction.requiresInitialization(fileLock);
            if (rebuild) {
                Exception latestException = null;
                for (int tries = 0; rebuild && tries < 3; tries++) {
                    fileLock.close();
                    fileLock = null;

	// DeletedClusters contains list of all subset clusters for the deleted DR or updated DR.
	// When clusters are rebuilt, we rebuild the subset clusters as well. So, we know what
	// subset clusters are really needed. So if deleted cluster is not rebuilt, then it is really deleted.
	builtClusters := sets.New[string]()
	for _, c := range clusters {
		builtClusters.Insert(c.Name)
	}