Am Montag, 4. September 2017 23:55:56 UTC+2 schrieb Evgeny Mandrikov:
> Hi,
> 
> 
> 
> On Monday, September 4, 2017 at 7:56:45 AM UTC+2, [email protected] wrote:
> 
>  With Eclemma 2.3.4 (from http://download.eclipselab.org/eclemma/trunk/update)
> 
> 
> Please use officially released version - latest is 3.0.0, and was released 
> after 2.3.3.
> 
>  
> i have several unit tests (about 800) that cause the error "Error while 
> dumping coverage data (code 5013)" when i run code coverage with Eclemma. 
> 
> 
>  
> this is the stack trace:
> java.net.SocketException: Connection reset
>         at java.net.SocketInputStream.read(Unknown Source)
>         at java.net.SocketInputStream.read(Unknown Source)
>         at java.net.SocketInputStream.read(Unknown Source)
>         at java.io.DataInputStream.readByte(Unknown Source)
>         at 
> org.jacoco.core.data.ExecutionDataReader.read(ExecutionDataReader.java:85)
>         at 
> com.mountainminds.eclemma.internal.core.MemoryExecutionDataSource.readFrom(MemoryExecutionDataSource.java:68)
>         at 
> com.mountainminds.eclemma.internal.core.launching.AgentServer.run(AgentServer.java:114)
>         at org.eclipse.core.internal.jobs.Worker.run(Worker.java:55)
> If i run code coverage on each package seperately, everything works fine. I 
> also tried Eclemma 3.0.0 from the Eclipse marketplace. But i got the same 
> error (with different line numbers in stack trace). I use the Eclipse 
> built-in launcher. It seems that Eclemma has a problem, when it has to 
> examine a larger number of tests.
> 
> 
> Just few moments ago I executed EclEmma 3.0.0 in Eclipse Oxygen 4.7.0 on a 
> test suite that contains more than 2000 tests covering more than 500 classes 
> without any problems.
> 
> The only known to me example of similar problem: is that one of tests causes 
> main thread to be in interrupted state - see 
> https://github.com/jacoco/eclemma/issues/64 and associated bug in JDK - 
> https://bugs.openjdk.java.net/browse/JDK-8154017
> 
> 
> Please check for this and if this is the case, then please fix your code 
> since there is nothing what we can do with such JDK bug.
> And if this is not your case and as shown above it doesn't seem to be related 
> to amount of tests, then how we can reproduce your problem?
> 
> 
> 
> 
> Regards,
> Evgeny

Hi Evgeny,

thanks for your reply. Like i wrote before, if i run the tests separately, 
there's no problem. So this doesn't seem to me, that there's a problem 
concerning the interrupted state in the test code.

Hmm, yes, to reproduce the problem is not easy. Maybe there's a kind of 
dependency between some tests. But that's hard to find out. Ok, i will see, if 
i can get any further information that could be interesting for you.

Regards
Stephan

-- 
You received this message because you are subscribed to the Google Groups 
"JaCoCo and EclEmma Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jacoco/360415ae-0eee-4fcf-8480-58933feeed7f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to