grpc
/
grpc-java
mirror of https://github.com/grpc/grpc-java.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

core: prevent data race in ManagedChannelOrphanWrapper (#6854) 

Data race was detected internally when
ManagedChannelOrphanWrapper.shutdown() was called concurrently:

WARNING: ThreadSanitizer: data race (pid=5009)
  Write of size 8 at 0x7fd2f7f37530 by thread T49:
    #0 java.lang.ref.Reference.clear()V Reference.java:265
    #1 io.grpc.internal.ManagedChannelOrphanWrapper$ManagedChannelReference.clearInternal()V ManagedChannelOrphanWrapper.java:118
    #2 io.grpc.internal.ManagedChannelOrphanWrapper$ManagedChannelReference.clear()V ManagedChannelOrphanWrapper.java:110
    #3 io.grpc.internal.ManagedChannelOrphanWrapper.shutdown()Lio/grpc/ManagedChannel; ManagedChannelOrphanWrapper.java:58
    (stacktrace redacted)

  Previous write of size 8 at 0x7fd2f7f37530 by thread T45 (mutexes: write M267260296638793720, write M267541771615505864, write M267823246592216728, write M267260296898451984, write M267541771875162784, write M267823246851873416):
    #0 java.lang.ref.Reference.clear()V Reference.java:265
    #1 io.grpc.internal.ManagedChannelOrphanWrapper$ManagedChannelReference.clearInternal()V ManagedChannelOrphanWrapper.java:118
    #2 io.grpc.internal.ManagedChannelOrphanWrapper$ManagedChannelReference.clear()V ManagedChannelOrphanWrapper.java:110
    #3 io.grpc.internal.ManagedChannelOrphanWrapper.shutdown()Lio/grpc/ManagedChannel; ManagedChannelOrphanWrapper.java:58
    (stacktrace redacted)
This commit is contained in:
Kun Zhang 2020-03-23 18:02:00 -07:00 committed by GitHub
parent 0ba9d3e2d1
commit 2e3ad1de25
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 14 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
core/src/main/java/io/grpc/internal/ManagedChannelOrphanWrapper.java
View File

@ -24,6 +24,7 @@ import java.lang.ref.SoftReference;
import java.lang.ref.WeakReference;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ConcurrentMap;
import java.util.concurrent.atomic.AtomicBoolean;
import java.util.logging.Level;
import java.util.logging.LogRecord;
import java.util.logging.Logger;
@ -54,15 +55,13 @@ final class ManagedChannelOrphanWrapper extends ForwardingManagedChannel {
@Override
public ManagedChannel shutdown() {
phantom.shutdown = true;
phantom.clear();
phantom.clearSafely();
return super.shutdown();
}
@Override
public ManagedChannel shutdownNow() {
phantom.shutdown = true;
phantom.clear();
phantom.clearSafely();
return super.shutdownNow();
}
@ -81,7 +80,7 @@ final class ManagedChannelOrphanWrapper extends ForwardingManagedChannel {
private final String channelStr;
private final Reference<RuntimeException> allocationSite;
private volatile boolean shutdown;
private final AtomicBoolean shutdown = new AtomicBoolean();
ManagedChannelReference(
ManagedChannelOrphanWrapper orphanable,
@ -113,6 +112,15 @@ final class ManagedChannelOrphanWrapper extends ForwardingManagedChannel {
cleanQueue(refqueue);
}
/**
* Safe to call concurrently.
*/
private void clearSafely() {
if (!shutdown.getAndSet(true)) {
clear();
}
}
// avoid reentrancy
private void clearInternal() {
super.clear();
@ -135,7 +143,7 @@ final class ManagedChannelOrphanWrapper extends ForwardingManagedChannel {
while ((ref = (ManagedChannelReference) refqueue.poll()) != null) {
RuntimeException maybeAllocationSite = ref.allocationSite.get();
ref.clearInternal(); // technically the reference is gone already.
if (!ref.shutdown) {
if (!ref.shutdown.get()) {
orphanedChannels++;
Level level = Level.SEVERE;
if (logger.isLoggable(level)) {