[Bug 92] New: threading is broken (most visible on SMP)

http://sablevm.org/bugs/show_bug.cgi?id=92

           Summary: threading is broken (most visible on SMP)
           Product: SableVM
           Version: staging
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: default
        AssignedTo: [email protected]
        ReportedBy: [email protected]
         QAContact: [email protected]


log into an SMP machine, like tofu.cs.mcgill.ca or magic.cs.mcgill.ca

compile ThreadStarter.java

$ for num in `seq 1 10000`; do sablevm -Y ThreadStarter; done

get the following four errors, plus core dumps as well.

error 1:

sablevm: thread.c:1104: _svmf_exit_object_monitor: Assertion
`_svmf_lockword_get_thinlock_id
((*(current->contention.requester.jobject))-> lockword) ==
env->thread.thinlock_id' failed.
-bash: line 23: 28147 Aborted                 sablevm -Y ThreadStarter

error 2:

java.lang.NullPointerException
   at IncrementRunnable.run (ThreadStarter.java:51)
   at ThreadStarter.main (ThreadStarter.java:38)
   at java.lang.VirtualMachine.invokeMain (VirtualMachine.java)
   at java.lang.VirtualMachine.main (VirtualMachine.java:108)

error 3:

Exception in thread "Thread-1" java.lang.NullPointerException
   at IncrementRunnable.run (ThreadStarter.java:51)
   at java.lang.Thread.run (Thread.java:705)
   at java.lang.VMThread.callRun (VMThread.java:119)
   at java.lang.Thread.callRun (Thread.java:391)
   at java.lang.VirtualMachine.runThread (VirtualMachine.java:137)

error 4:

sablevm: global_refs.c:57: _svmh_new_native_global: Assertion
`native_global->next->previous == ((void *)0)' failed.
-bash: line 23: 15932 Aborted                 sablevm -Y ThreadStarter

--------- begin ThreadStarter.java -------------
/* attempts two different ways to start threads.
   
   note that none of:

     a) way 1; way 1;
     b) way 2; way 2;
     c) way 2; way 1;
 
   were observed to fail, with 100+ runs, whereas

     d) way 1; way 2;

   fails about 5 percent of the time on
   tofu.cs.mcgill.ca (2P).

   (if you make things more complex by adding another
    Runnable implementation, it fails more often, but
    this was the most stripped down I could get it)
*/

public class ThreadStarter {
    
    public ThreadStarter() {
    }    
    
    public static void main(String[] argv) {
        
        Runnable myRunnable;
        Thread myThread;  

        /* way 1 */
        myRunnable = new IncrementRunnable();
        myThread = new Thread(myRunnable);
        myThread.start();

        /* way 2 */
        myRunnable = new IncrementRunnable();
        myRunnable.run();
    }
}

class IncrementRunnable implements Runnable {

    private int k = 0;

    public IncrementRunnable() {
    }

    public void run() {
        for(int i=0;i<100;i++) {
            k++;
        }
    }
}
--------------- end ThreadStarter.java -------------------



------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
You are the QA contact for the bug, or are watching the QA contact.