[internal] jvm/java: ensure JDK downloaded in one process

[tdyas]

Motivation

As described in #12293, multiple Coursier invocations were downloading the JDK and triggering a race condition in Coursier's locking that caused flakiness in tests.

Solution

This PR mitigates the issue by isolating JDK download to a single Process. The new JdkSetup type provides rules with the command to obtain the location of the JDK so they may query Coursier for JAVA_HOME. This has the benefit of still downloading in remote execution, but providing some guarantee that there will be a single download.

[stuhood]

A tricky bit about this in the context of remote execution is that the JDK selection process isn't guaranteed to run on the same machine as the compile.

So rather than having the jdk_setup expose java_home as a property, it should probably expose it as a command prefix, which we can hope (or know, locally) will hit the cached pre-selected JDK rather than re-fetching it.

[tdyas]

So rather than having the jdk_setup expose java_home as a property, it should probably expose it as a command prefix, which we can hope (or know, locally) will hit the cached pre-selected JDK rather than re-fetching it.

That would require using coursier to invoke the JDK every time?

[stuhood]

Correct. But it appears to take a few milliseconds, because the coursier binary is native.

Also, that time is moot with nailgun: I'll probably get to that after hours this week.

[tdyas]

Okay implemented calling into coursier's binary each time.

[tdyas]

Note: With remote execution, the user will likely have to choose "system" JDK since there is no guarantee that one remote executor will share the same JDK cache as an executor used for a subsequent remote execution request. We shouldn't attempt to solve the remote execution case right now.

[stuhood]

Thanks!

[tdyas]

There is still a failure when downloading even with this PR:

ValueError: Failed to determine JAVA_HOME for JDK system: Downloading https://github.com/shyiko/jabba/raw/master/index.json
E           Downloaded https://github.com/shyiko/jabba/raw/master/index.json
E           Downloading https://github.com/AdoptOpenJDK/openjdk8-binaries/releases/download/jdk8u292-b10/OpenJDK8U-jdk_x64_linux_hotspot_8u292b10.tar.gz
E           Still downloading:
E           https://github.com/AdoptOpenJDK/openjdk8-binaries/releases/download/jdk8u292-b10/OpenJDK8U-jdk_x64_linux_hotspot_8u292b10.tar.gz (26.83 %, 27640420 / 103026380)
E           
E           Downloaded https://github.com/AdoptOpenJDK/openjdk8-binaries/releases/download/jdk8u292-b10/OpenJDK8U-jdk_x64_linux_hotspot_8u292b10.tar.gz
E           Extracting
E             /home/runner/.cache/coursier/v1/https/github.com/AdoptOpenJDK/openjdk8-binaries/releases/download/jdk8u292-b10/OpenJDK8U-jdk_x64_linux_hotspot_8u292b10.tar.gz
E           in
E             /home/runner/.cache/coursier/jvm/adopt@1.8.0-292
E           Extraction failed: java.nio.file.FileSystemException: /home/runner/.cache/coursier/jvm/.adopt@1.8.0-292.part/jdk8u292-b10 -> /home/runner/.cache/coursier/jvm/adopt@1.8.0-292: Directory not empty
E           Exception in thread "main" java.nio.file.FileSystemException: /home/runner/.cache/coursier/jvm/.adopt@1.8.0-292.part/jdk8u292-b10 -> /home/runner/.cache/coursier/jvm/adopt@1.8.0-292: Directory not empty
E           	at sun.nio.fs.UnixCopyFile.move(UnixCopyFile.java:417)
E           	at sun.nio.fs.UnixFileSystemProvider.move(UnixFileSystemProvider.java:267)
E           	at java.nio.file.Files.move(Files.java:1421)
E           	at coursier.jvm.JvmCache.$anonfun$tryExtract$1(JvmCache.scala:66)
E           	at coursier.jvm.JvmCache.$anonfun$withLockFor$1(JvmCache.scala:267)
E           	at coursier.cache.CacheLocks$.loop$1(CacheLocks.scala:72)
E           	at coursier.cache.CacheLocks$.withLockOr(CacheLocks.scala:98)
E           	at coursier.jvm.JvmCache.withLockFor(JvmCache.scala:267)
E           	at coursier.jvm.JvmCache.tryExtract(JvmCache.scala:48)
E           	at coursier.jvm.JvmCache.$anonfun$get$9(JvmCache.scala:140)
E           	at coursier.util.Task$.wrap(Task.scala:84)
E           	at coursier.util.Task$.$anonfun$delay$2(Task.scala:49)
E           	at scala.concurrent.Future$.$anonfun$apply$1(Future.scala:659)
E           	at scala.util.Success.$anonfun$map$1(Try.scala:255)
E           	at scala.util.Success.map(Try.scala:213)
E           	at scala.concurrent.Future.$anonfun$map$1(Future.scala:292)
E           	at scala.concurrent.impl.Promise.liftedTree1$1(Promise.scala:33)
E           	at scala.concurrent.impl.Promise.$anonfun$transform$1(Promise.scala:33)
E           	at scala.concurrent.impl.CallbackRunnable.run(Promise.scala:64)
E           	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1128)
E           	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:628)
E           	at java.lang.Thread.run(Thread.java:834)
E           	at com.oracle.svm.core.thread.JavaThreads.threadStartRoutine(JavaThreads.java:517)
E           	at com.oracle.svm.core.posix.thread.PosixJavaThreads.pthreadStartRoutine(PosixJavaThreads.java:193)

https://github.com/pantsbuild/pants/runs/3669182720?check_suite_focus=true#step:11:534

#12977 adds a JDK into the GitHub Actions CI so there is always a "system" JDK available.

[stuhood]

I noticed while rebasing #12982 that this line was lost: it's important when using the system JDK, because it isn't stable, and the fingerprint computation can't be cached. I doubt it's related to the current flakiness, but fyi.