How to enable debug on Eclipse IDE

[MLobatoR]

Hi,

I'm getting the following error message but I couldn't find a way to enable debug on eclipse:

2 errors generated.
[TornadoVM-OCL-JNI] ERROR : clBuildProgram -> Returned: -11
�[31m[Bailout] Running the sequential implementation. Enable --debug to see the reason.�[0m

Could you help me please?

Thank you in advance
ML

[jjfumero]

Hi @MLobatoR . This error is due to a problem with the generated OpenCL code. Is it possible to share a test case with the Java code that you are trying to run?

[jjfumero]

I forgot to mention. You can enable the print kernel option (--printKernel) to check the OpenCL-generated kernel.

[MLobatoR]

@jjfumero
I was not able to come up with a simple example. I'm going to try to explain what I did. I've been commenting the code in order to know which line is causing the error. The first thing I noticed is that if I comment any while cycle I have and a for loop using data that I transferred to device on first execution the error disappears (obviously the code is not doing what it needs to be done).

Here's an example of a while cycle:

protected static int getHourOfDay(int currentCost) {
	int cost = currentCost;
	while(cost > dayMillis) {
		cost = cost - dayMillis;
	}

	return cost;
}

And a little example of a for loop that does not work:

for(int destinationNode = Va[originNode]; destinationNode < Va[originNode + 1]; destinationNode++) {
			

}

Even doing nothing the I got the error. These to pieces of the code ran under another for with the annotation parallel.

Va is transferred to device on first execution, and is a read-only, it never changes.
dayMillis is a static final int and current cost changes while the algorithm runs.

Can you help me or do you need more information?

Thank you in advance

[MLobatoR]

@jjfumero,

Here is some log I found generated by Tornado:

<kernel>:873:1: error: extraneous closing brace ('}')
}  //  kernel
^

Is this helpful in some way?

[MLobatoR]

@jjfumero,

Another example that throw the error

protected static void initPathCost(Matrix2DInt r, VectorInt t, int iSize) {
		for(int i = 0; i < iSize; i++) {
			for(int j = 0; j < t.getLength(); j++) {
				r.set(i, j, Integer.MAX_VALUE);
			}
		}
}

r is transferred to host, t is ransferred to device on first execution, since is only data, iSize is a constant that never changes. When I comment the line "r.set(i, j, Integer.MAX_VALUE);" the error stops.

[jjfumero]

ok, that's useful. This is an error in our codegen. We will investigate this issue and let you know.

[mairooni]

Hi @MLobatoR. We are working on solving this issue, however, we could not reproduce it for the function you provided (initPathCost). Could you provide the code of your application that is generating this error or give us another example?

[MLobatoR]

Hi @mairooni,

I'm sorry to keep you waiting but here's a dummy code where I have the error:

public class test3 {

	public static void main(String[] args) {

		
		List<Integer> targetsList = Arrays.asList(1, 2, 3, 4, 5, 6, 7);
		int[] targetsArray = targetsList.stream().mapToInt(Integer::intValue).toArray();
		
		VectorInt targets = new VectorInt(targetsArray);
		
		Matrix2DInt ptCost = new Matrix2DInt(1, 7);
		
		int[] m = new int[8];
		m[0] = 1000;
		m[1] = 1000;
		m[2] = 1000;
		m[3] = 1000;
		m[4] = 1000;
		m[5] = 1000;
		m[6] = 1000;
		m[7] = 1000;
		
		int[][] c = new int[1][8];
		for(int i = 0; i < 8; i++) {
			c[0][i] = 1000;
		}
		
		Matrix2DInt tc = new Matrix2DInt(c);
		
		int[] v = new int[9];
		v[0] = 0;
		v[1] = 1;
		v[2] = 2;
		v[3] = 4;
		v[4] = 5;
		v[5] = 6;
		v[6] = 7;
		v[7] = 8;
		
		int[] ea = new int[8];
		ea[0] = 1;
		ea[1] = 2;
		ea[2] = 3;
		ea[3] = 6;
		ea[4] = 4;
		ea[5] = 5;
		ea[6] = 2;
		ea[7] = 7;
		
		int[] es = new int[8];
		es[0] = 0;
		es[1] = 1;
		es[2] = 2;
		es[3] = 2;
		es[4] = 3;
		es[5] = 4;
		es[6] = 5;
		es[7] = 6;
		
		TaskGraph morpheus = new TaskGraph("morpheus");
		morpheus.transferToDevice(DataTransferMode.FIRST_EXECUTION, ea, es, v, m, 1, tc, targets);
		morpheus.transferToHost(DataTransferMode.EVERY_EXECUTION, ptCost);
		
		morpheus.task("t", test3::compute, ea, es, v, m, 1, 0, targets, ptCost);
		
		// Create an immutable task-graph
        ImmutableTaskGraph immutableTaskGraph = morpheus.snapshot();
        
        // Create an execution plan from an immutable task-graph
        TornadoExecutionPlan executionPlan = new TornadoExecutionPlan(immutableTaskGraph);

        // Execute the execution plan
        executionPlan.execute();
		
	}
	
	protected static void compute(int[] ea, int[] es, int[] v, int[] m, int s, int sr, VectorInt targets, Matrix2DInt pathCost) {
		Matrix2DInt ec = new Matrix2DInt(s, ea.length);
		Matrix2DInt mark = new Matrix2DInt(s, ea.length);
		
		Matrix2DInt results = new Matrix2DInt(s, ea.length);
		
		init(es, v.length, m, s, sr, ec, mark, results);
		
		initPathCost(pathCost, targets, s);
	}
	
	protected static void init(int[] es, int size, int[] m, int s, int sr, Matrix2DInt ec, Matrix2DInt mark, Matrix2DInt results) {
		for(int st = 0; st < s; st++) {
			for(int e = 0; e < es.length; e++) {
				int sn = es[e];
				if(sn == sr) {
					int cost = m[e];
					ec.set(st, e, cost);
					results.set(st, e, cost);

					if(sr == size - 1) {
						mark.set(st, e, 0);
					} else {
						mark.set(st, e, 1);
					}

				} else {
					mark.set(st, e, 0);
					ec.set(st, e, Integer.MAX_VALUE);
					results.set(st, e, Integer.MAX_VALUE);
				}
			}
		}
	}
	
	protected static void initPathCost(Matrix2DInt pathCost, VectorInt targets, int s) {
		for(int st = 0; st < s; st++) {
			for(int t = 0; t < targets.getLength(); t++) {
				pathCost.set(st, t, Integer.MAX_VALUE);
			}
		}
	}
}

Hope this helps to solve the issue.

Thank you,
ML

[mairooni]

Hi @MLobatoR, thank you for providing more details. We identified the issue, a fix will be provided soon. Nevertheless, we noticed that you are using an older version of TornadoVM. We highly encourage you to use the latest version, as this will include all the latest bug fixes. In the current version of TornadoVM, we are using off-heap data types instead of primitive arrays. You can find a guide for migrating your current application to the latest version of TornadoVM here: Migrating from v0.15.2 to v1.0.
As an example, this is what the code you provided would look like, using the latest TornadoVM API:

public class test3 {
    public static void main(String[] args) {
            IntArray targetsArray = IntArray.fromElements(1, 2, 3, 4, 5, 6, 7);

            VectorInt targets = new VectorInt(targetsArray);

            Matrix2DInt ptCost = new Matrix2DInt(1, 8);

            IntArray m = IntArray.fromElements(1000, 1000, 1000, 1000, 1000, 1000, 1000, 1000);

            int[][] c = new int[1][8];
            for (int i = 0; i < 8; i++) {
                c[0][i] = 1000;
            }

            Matrix2DInt tc = new Matrix2DInt(c);

            IntArray v = IntArray.fromElements(0, 1, 2, 3, 4, 5, 6, 7, 8);

            IntArray ea = IntArray.fromElements(1, 2, 3, 6, 4, 5, 2, 7);

            IntArray es = IntArray.fromElements(0, 1, 2, 2, 3, 4, 5, 6);

            TaskGraph morpheus = new TaskGraph("morpheus");
            morpheus.transferToDevice(DataTransferMode.FIRST_EXECUTION, ea, es, v, m, 1, tc, targets);
            morpheus.transferToHost(DataTransferMode.EVERY_EXECUTION, ptCost);

            morpheus.task("t", test3::compute, ea, es, v, m, 1, 0, targets, ptCost);

            // Create an immutable task-graph
            ImmutableTaskGraph immutableTaskGraph = morpheus.snapshot();

            // Create an execution plan from an immutable task-graph
            TornadoExecutionPlan executionPlan = new TornadoExecutionPlan(immutableTaskGraph);

            // Execute the execution plan
            executionPlan.execute();
    }
    
    protected static void compute(IntArray ea, IntArray es, IntArray v, IntArray m, int s, int sr, VectorInt targets, Matrix2DInt pathCost) {
        Matrix2DInt ec = new Matrix2DInt(s, ea.getSize());
        Matrix2DInt mark = new Matrix2DInt(s, ea.getSize());

        Matrix2DInt results = new Matrix2DInt(s, ea.getSize());

        init(es, v.getSize(), m, s, sr, ec, mark, results);

        initPathCost(pathCost, targets, s);
    }

    protected static void init(IntArray es, int size, IntArray m, int s, int sr, Matrix2DInt ec, Matrix2DInt mark, Matrix2DInt results) {
        for (int st = 0; st < s; st++) {
            for (int e = 0; e < es.getSize(); e++) {
                int sn = es.get(e);
                if (sn == sr) {
                    int cost = m.get(e);
                    ec.set(st, e, cost);
                    results.set(st, e, cost);
                    int cond = size - 1;
                    if (sr == cond) {
                        mark.set(st, e, 0);
                    } else {
                        mark.set(st, e, 1);
                    }

                } else {
                    mark.set(st, e, 0);
                    ec.set(st, e, Integer.MAX_VALUE);
                    results.set(st, e, Integer.MAX_VALUE);
                }
            }
        }
    }

    protected static void initPathCost(Matrix2DInt pathCost, VectorInt targets, int s) {
        for (int st = 0; st < s; st++) {
            for (int t = 0; t < targets.getLength(); t++) {
                pathCost.set(st, t, Integer.MAX_VALUE);
            }
        }
    }
    
}

Additionally, I would recommend some rewriting of your application to take advantage of the full capabilities of TornadoVM.
Specifically, first of all, instead of passing the compute method to the TaskGraph, it would be better to create a TaskGraph with two tasks, one for the method init and one for the initPathCosts, as all the compute method does is to invoke the two aforementioned methods. Usually, what you want to pass to the TaskGraph is the computation that will offloaded for acceleration, so it would make more sense this way. With this change, the TaskGraph would look like so:

public static void main(String[] args) {
     ....
      Matrix2DInt ec = new Matrix2DInt(1, ea.getSize());
      Matrix2DInt mark = new Matrix2DInt(1, ea.getSize());

      Matrix2DInt results = new Matrix2DInt(1, ea.getSize());

      TaskGraph morpheus = new TaskGraph("morpheus");
      
      morpheus.transferToDevice(DataTransferMode.FIRST_EXECUTION, ea, es, v, m, ec, mark, results, ptCost, targets);
      morpheus.transferToHost(DataTransferMode.EVERY_EXECUTION, ptCost);
      morpheus.task("t0", test3::init, es, v.getSize(), m, 1, 0, ec, mark, results);
      morpheus.task("t1", test3::initPathCost, ptCost, targets, 1);
      
      ....
  }

Secondly, I recommend you use the @Parallel annotation before the index initialization of your for-loops. This annotation is used by the compiler to parallelize the computation among the parallel GPU threads:
E.g. :

protected static void initPathCost(Matrix2DInt pathCost, VectorInt targets, int s) {
       for (@Parallel int st = 0; st < s; st++) {
         ....

protected static void init(IntArray es, int size, IntArray m, int s, int sr, Matrix2DInt ec, Matrix2DInt mark, Matrix2DInt results) {
        for (@Parallel int st = 0; st < s; st++) {
            for (@Parallel int e = 0; e < es.getSize(); e++) {
            ....         

I hope this helps! Let us know if something is not clear. We will notify you in this thread as soon as the fix is pushed.

[MLobatoR]

@mairooni,

Thank you so much for your help. I will be waiting for your fix.

:)

[mairooni]

Hi @MLobatoR, the fix has now been pushed. Have a look and let us know if there are any other issues.

[MLobatoR]

Hi @mairooni,

I updated TornadoVM version but I'm strugling with the following error when I try to run my code:

java.lang.reflect.InaccessibleObjectException: Unable to make static void java.lang.invoke.LambdaForm$MH/0x000001c7d36b13f0.invokeExact_MT(java.lang.Object,java.lang.Object,java.lang.Object,java.lang.Object,java.lang.Object,java.lang.Object,java.lang.Object,java.lang.Object,int,java.lang.Object,int,java.lang.Object,int,java.lang.Object,java.lang.Object) accessible: module java.base does not "opens java.lang.invoke" to module tornado.runtime
	at java.base/java.lang.reflect.AccessibleObject.throwInaccessibleObjectException(AccessibleObject.java:391)
	at java.base/java.lang.reflect.AccessibleObject.checkCanSetAccessible(AccessibleObject.java:367)
	at java.base/java.lang.reflect.AccessibleObject.checkCanSetAccessible(AccessibleObject.java:315)
	at java.base/java.lang.reflect.Method.checkCanSetAccessible(Method.java:203)
	at java.base/java.lang.reflect.Method.setAccessible(Method.java:197)
	at tornado.runtime@1.0.1-dev/uk.ac.manchester.tornado.runtime.analyzer.TaskUtils.callGetMethod(TaskUtils.java:125)
	at tornado.runtime@1.0.1-dev/uk.ac.manchester.tornado.runtime.analyzer.TaskUtils.resolveMethodHandle(TaskUtils.java:191)
	at tornado.runtime@1.0.1-dev/uk.ac.manchester.tornado.runtime.tasks.TornadoTaskGraph.addTask(TornadoTaskGraph.java:2030)
	at tornado.api@1.0.1-dev/uk.ac.manchester.tornado.api.TaskGraph.task(TaskGraph.java:512)
	at com.widescope.algorithm.gpu.NeoOneToManyAlgorithmGPU.run(NeoOneToManyAlgorithmGPU.java:40)
	at com.widescope.algorithm.gpu.NeoOneToManyAlgorithmGPUTest.run(NeoOneToManyAlgorithmGPUTest.java:97)
	at com.widescope.algorithm.gpu.NeoOneToManyAlgorithmGPUTest.given_AGraph_When_modeIsFastestWithoutTraffic_Then_calculateCostForEveryNode(NeoOneToManyAlgorithmGPUTest.java:57)
	at java.base/jdk.internal.reflect.DirectMethodHandleAccessor.invoke(DirectMethodHandleAccessor.java:103)
	at java.base/java.lang.reflect.Method.invoke(Method.java:580)
	at org.junit.platform.commons.util.ReflectionUtils.invokeMethod(ReflectionUtils.java:728)
	at org.junit.jupiter.engine.execution.MethodInvocation.proceed(MethodInvocation.java:60)
	at org.junit.jupiter.engine.execution.InvocationInterceptorChain$ValidatingInvocation.proceed(InvocationInterceptorChain.java:131)
	at org.junit.jupiter.engine.extension.TimeoutExtension.intercept(TimeoutExtension.java:156)
	at org.junit.jupiter.engine.extension.TimeoutExtension.interceptTestableMethod(TimeoutExtension.java:147)
	at org.junit.jupiter.engine.extension.TimeoutExtension.interceptTestMethod(TimeoutExtension.java:86)
	at org.junit.jupiter.engine.execution.InterceptingExecutableInvoker$ReflectiveInterceptorCall.lambda$ofVoidMethod$0(InterceptingExecutableInvoker.java:103)
	at org.junit.jupiter.engine.execution.InterceptingExecutableInvoker.lambda$invoke$0(InterceptingExecutableInvoker.java:93)
	at org.junit.jupiter.engine.execution.InvocationInterceptorChain$InterceptedInvocation.proceed(InvocationInterceptorChain.java:106)
	at org.junit.jupiter.engine.execution.InvocationInterceptorChain.proceed(InvocationInterceptorChain.java:64)
	at org.junit.jupiter.engine.execution.InvocationInterceptorChain.chainAndInvoke(InvocationInterceptorChain.java:45)
	at org.junit.jupiter.engine.execution.InvocationInterceptorChain.invoke(InvocationInterceptorChain.java:37)
	at org.junit.jupiter.engine.execution.InterceptingExecutableInvoker.invoke(InterceptingExecutableInvoker.java:92)
	at org.junit.jupiter.engine.execution.InterceptingExecutableInvoker.invoke(InterceptingExecutableInvoker.java:86)
	at org.junit.jupiter.engine.descriptor.TestMethodTestDescriptor.lambda$invokeTestMethod$7(TestMethodTestDescriptor.java:218)
	at org.junit.platform.engine.support.hierarchical.ThrowableCollector.execute(ThrowableCollector.java:73)
	at org.junit.jupiter.engine.descriptor.TestMethodTestDescriptor.invokeTestMethod(TestMethodTestDescriptor.java:214)
	at org.junit.jupiter.engine.descriptor.TestMethodTestDescriptor.execute(TestMethodTestDescriptor.java:139)
	at org.junit.jupiter.engine.descriptor.TestMethodTestDescriptor.execute(TestMethodTestDescriptor.java:69)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.lambda$executeRecursively$6(NodeTestTask.java:151)
	at org.junit.platform.engine.support.hierarchical.ThrowableCollector.execute(ThrowableCollector.java:73)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.lambda$executeRecursively$8(NodeTestTask.java:141)
	at org.junit.platform.engine.support.hierarchical.Node.around(Node.java:137)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.lambda$executeRecursively$9(NodeTestTask.java:139)
	at org.junit.platform.engine.support.hierarchical.ThrowableCollector.execute(ThrowableCollector.java:73)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.executeRecursively(NodeTestTask.java:138)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.execute(NodeTestTask.java:95)
	at java.base/java.util.ArrayList.forEach(ArrayList.java:1596)
	at org.junit.platform.engine.support.hierarchical.SameThreadHierarchicalTestExecutorService.invokeAll(SameThreadHierarchicalTestExecutorService.java:41)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.lambda$executeRecursively$6(NodeTestTask.java:155)
	at org.junit.platform.engine.support.hierarchical.ThrowableCollector.execute(ThrowableCollector.java:73)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.lambda$executeRecursively$8(NodeTestTask.java:141)
	at org.junit.platform.engine.support.hierarchical.Node.around(Node.java:137)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.lambda$executeRecursively$9(NodeTestTask.java:139)
	at org.junit.platform.engine.support.hierarchical.ThrowableCollector.execute(ThrowableCollector.java:73)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.executeRecursively(NodeTestTask.java:138)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.execute(NodeTestTask.java:95)
	at java.base/java.util.ArrayList.forEach(ArrayList.java:1596)
	at org.junit.platform.engine.support.hierarchical.SameThreadHierarchicalTestExecutorService.invokeAll(SameThreadHierarchicalTestExecutorService.java:41)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.lambda$executeRecursively$6(NodeTestTask.java:155)
	at org.junit.platform.engine.support.hierarchical.ThrowableCollector.execute(ThrowableCollector.java:73)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.lambda$executeRecursively$8(NodeTestTask.java:141)
	at org.junit.platform.engine.support.hierarchical.Node.around(Node.java:137)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.lambda$executeRecursively$9(NodeTestTask.java:139)
	at org.junit.platform.engine.support.hierarchical.ThrowableCollector.execute(ThrowableCollector.java:73)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.executeRecursively(NodeTestTask.java:138)
	at org.junit.platform.engine.support.hierarchical.NodeTestTask.execute(NodeTestTask.java:95)
	at org.junit.platform.engine.support.hierarchical.SameThreadHierarchicalTestExecutorService.submit(SameThreadHierarchicalTestExecutorService.java:35)
	at org.junit.platform.engine.support.hierarchical.HierarchicalTestExecutor.execute(HierarchicalTestExecutor.java:57)
	at org.junit.platform.engine.support.hierarchical.HierarchicalTestEngine.execute(HierarchicalTestEngine.java:54)
	at org.junit.platform.launcher.core.DefaultLauncher.execute(DefaultLauncher.java:229)
	at org.junit.platform.launcher.core.DefaultLauncher.lambda$execute$6(DefaultLauncher.java:197)
	at org.junit.platform.launcher.core.DefaultLauncher.withInterceptedStreams(DefaultLauncher.java:211)
	at org.junit.platform.launcher.core.DefaultLauncher.execute(DefaultLauncher.java:191)
	at org.junit.platform.launcher.core.DefaultLauncher.execute(DefaultLauncher.java:137)
	at org.eclipse.jdt.internal.junit5.runner.JUnit5TestReference.run(JUnit5TestReference.java:89)
	at org.eclipse.jdt.internal.junit.runner.TestExecution.run(TestExecution.java:41)
	at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.runTests(RemoteTestRunner.java:541)
	at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.runTests(RemoteTestRunner.java:763)
	at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.run(RemoteTestRunner.java:463)
	at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.main(RemoteTestRunner.java:209)

My VM arguments are the following (copied after running command --printJavaFlags):
-server -XX:-UseCompressedOops -XX:+UnlockExperimentalVMOptions -XX:+EnableJVMCI -XX:-UseCompressedClassPointers --enable-preview -Djava.library.path=C:\msys64\home\Miriam\TornadoVM\bin\sdk\lib --module-path C:\msys64\home\Miriam\TornadoVM\bin\sdk\share\java\tornado -Dtornado.load.api.implementation=uk.ac.manchester.tornado.runtime.tasks.TornadoTaskGraph -Dtornado.load.runtime.implementation=uk.ac.manchester.tornado.runtime.TornadoCoreRuntime -Dtornado.load.tornado.implementation=uk.ac.manchester.tornado.runtime.common.Tornado -Dtornado.load.device.implementation.opencl=uk.ac.manchester.tornado.drivers.opencl.runtime.OCLDeviceFactory -Dtornado.load.device.implementation.ptx=uk.ac.manchester.tornado.drivers.ptx.runtime.PTXDeviceFactory -Dtornado.load.device.implementation.spirv=uk.ac.manchester.tornado.drivers.spirv.runtime.SPIRVDeviceFactory -Dtornado.load.annotation.implementation=uk.ac.manchester.tornado.annotation.ASMClassVisitor -Dtornado.load.annotation.parallel=uk.ac.manchester.tornado.api.annotations.Parallel -XX:+UseParallelGC @C:\msys64\home\Miriam\TornadoVM\bin\sdk\etc\exportLists\common-exports @C:\msys64\home\Miriam\TornadoVM\bin\sdk\etc\exportLists\opencl-exports --add-modules ALL-SYSTEM,tornado.runtime,tornado.annotation,tornado.drivers.common,tornado.drivers.opencl

I'm running with JDK 21.

When I run your tests they run fine, so I'm guessing that the installation went well and this must be something about the VM arguments.

Thank you for your help,
ML

[MLobatoR]

Hi @jjfumero @mairooni,

Can either of you help me with this? I'm not able to use TornadoVM since the upgrade..

Thank you,
ML

[jjfumero]

HI @MLobatoR , It seems you trigger an unsupported feature. Is it possible for you to share the test-case you are running?

[jjfumero]

Are you using reflection in your code-base?