Details
Bug
Closed
Major
Description
Hi,
following up on our discussion about XStream's thread safety on the mailing list, I found a problem when calling createObjectOutputStream() and toXML() concurrently. I'll attach a minimal example, see below for the stack trace.
According to your explanations about the thread safety that XStream guarantees, this seems a clear bug.
Thanks!
Exception in thread "Thread-4154" java.lang.NullPointerException
at com.sun.xml.internal.stream.writers.UTF8OutputStreamWriter.write(UTF8OutputStreamWriter.java:94)
at com.sun.xml.internal.stream.writers.UTF8OutputStreamWriter.write(UTF8OutputStreamWriter.java:129)
at com.sun.xml.internal.stream.writers.XMLStreamWriterImpl.writeStartDocument(XMLStreamWriterImpl.java:1109)
at com.thoughtworks.xstream.io.xml.StaxWriter.<init>(StaxWriter.java:74)
at com.thoughtworks.xstream.io.xml.StaxDriver.createStaxWriter(StaxDriver.java:141)
at com.thoughtworks.xstream.io.xml.StaxDriver.createStaxWriter(StaxDriver.java:145)
at com.thoughtworks.xstream.io.xml.StaxDriver.createWriter(StaxDriver.java:129)
at com.thoughtworks.xstream.XStream.createObjectOutputStream(XStream.java:1575)
at XStreamTest6$1.run(XStreamTest6.java:24)
at java.lang.Thread.run(Thread.java:619)
Activity
Tried it now again two times, once with current Java 6 and once with current Java 7, no problem.
============= %< =================
$ java -version
java version "1.7.0"
Java(TM) SE Runtime Environment (build 1.7.0-b147)
Java HotSpot(TM) 64-Bit Server VM (build 21.0-b17, mixed mode)
$ java -version
java version "1.6.0_27"
Java(TM) SE Runtime Environment (build 1.6.0_27-b07)
Java HotSpot(TM) 64-Bit Server VM (build 20.2-b06, mixed mode)
$ uname -a
Linux floh 2.6.39-gentoo-r3 #1 SMP Sat Jul 23 16:28:48 CEST 2011 x86_64 Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz GenuineIntel GNU/Linux
============= %< =================
I even tried it with 20 parallel threads (and moved XStream initialization out of the loop) to penetrate my hex core:
private void run() throws Exception { SjsxpDriver driver = new SjsxpDriver(); final XStream xstream = new XStream(driver); class T extends Thread { public T(final boolean x) { super(new Runnable() { public void run() { try { if (x) { xstream.createObjectOutputStream(new ByteArrayOutputStream()); } else { xstream.toXML("aaa"); } } catch (IOException ioe) { throw new RuntimeException(ioe); } } }); } } T threads[] = new T[20]; for (int i = 0; i < 50000; i++ ) { boolean x = true; for (int j = threads.length; j-- > 0;) { threads[j] = new T(x ^= true); } for (T t : threads) { t.start(); } try { for (T t : threads) { t.join(); } } catch (InterruptedException ie) { } } }
No problem here.
Jörg Schaible
added a comment - Tried it now again two times, once with current Java 6 and once with current Java 7, no problem.
============= %< =================
$ java -version
java version "1.7.0"
Java(TM) SE Runtime Environment (build 1.7.0-b147)
Java HotSpot(TM) 64-Bit Server VM (build 21.0-b17, mixed mode)
$ java -version
java version "1.6.0_27"
Java(TM) SE Runtime Environment (build 1.6.0_27-b07)
Java HotSpot(TM) 64-Bit Server VM (build 20.2-b06, mixed mode)
$ uname -a
Linux floh 2.6.39-gentoo-r3 #1 SMP Sat Jul 23 16:28:48 CEST 2011 x86_64 Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz GenuineIntel GNU/Linux
============= %< =================
I even tried it with 20 parallel threads (and moved XStream initialization out of the loop) to penetrate my hex core:
private void run() throws Exception {
SjsxpDriver driver = new SjsxpDriver();
final XStream xstream = new XStream(driver);
class T extends Thread {
public T( final boolean x) {
super ( new Runnable () {
public void run() {
try {
if (x) {
xstream.createObjectOutputStream( new ByteArrayOutputStream());
} else {
xstream.toXML( "aaa" );
}
} catch (IOException ioe) {
throw new RuntimeException(ioe);
}
}
});
}
}
T threads[] = new T[20];
for ( int i = 0; i < 50000; i++ ) {
boolean x = true ;
for ( int j = threads.length; j-- > 0;) {
threads[j] = new T(x ^= true );
}
for (T t : threads) {
t.start();
}
try {
for (T t : threads) {
t.join();
}
} catch (InterruptedException ie) {
}
}
}
No problem here. I also cannot reproduce it with JDK 1.6.0_27 or JDK 1.7.0. Either the problem is due to a bug in JDK 1.6.0_07 that has been fixed, or some other change in the JDK influences the scheduling in a way that makes reproducing the bug much more difficult with the test case.
Michael Pradel
added a comment - I also cannot reproduce it with JDK 1.6.0_27 or JDK 1.7.0. Either the problem is due to a bug in JDK 1.6.0_07 that has been fixed, or some other change in the JDK influences the scheduling in a way that makes reproducing the bug much more difficult with the test case. IMHO there's nothing to fix in XStream.
Jörg Schaible
added a comment - IMHO there's nothing to fix in XStream. 
Hmm. I cannot reproduce this, your test runs through. Since the error actually happens in the JDK's StAX implementation, can you use a recent version of the JDK?
========= %< ======
$ java -version
java version "1.6.0_27"
Java(TM) SE Runtime Environment (build 1.6.0_27-b07)
Java HotSpot(TM) Server VM (build 20.2-b06, mixed mode)
$ uname -a
Linux josix 2.6.39-gentoo-r3 #2 SMP PREEMPT Fri Sep 9 10:04:29 CEST 2011 i686 Intel(R) Core(TM)2 Duo CPU T8300 @ 2.40GHz GenuineIntel GNU/Linux
========= %< ======
I'll try it with a different machine later on.