OnnxRuntimeMultithreadCrash

OnnxRuntimeMultithreadCrash.java

import java.util.HashMap;
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.ScheduledExecutorService;
import java.util.concurrent.TimeUnit;

import ai.onnxruntime.OnnxTensor;
import ai.onnxruntime.OrtEnvironment;
import ai.onnxruntime.OrtSession;
import ai.onnxruntime.OrtSession.Result;

public class OnnxRuntimeMultithreadCrash {

	private static final String SESSION_NAME = "foo";
	private static final String MODEL_FILE = "/my_model.onnx";

	// a fake AI service
	public static void main(String[] args) {
		OrtEnvironment env = OrtEnvironment.getEnvironment();
		Map<String, OrtSession> sessions = new ConcurrentHashMap<>();
		ScheduledExecutorService sched = Executors.newScheduledThreadPool(1);
		sched.scheduleWithFixedDelay(() -> {
			// management thread: periodically reloads models
			try {
				System.out.println("SWAP START");
				OrtSession session = env.createSession(MODEL_FILE);
				System.out.println("LOADED");
				OrtSession oldSession = sessions.put(SESSION_NAME, session);
				if (oldSession != null) {
					// kill the session, but it may still be in use.
					oldSession.close();
				}
				System.out.println("SWAP DONE");
			} catch (Exception e) {
				System.err.println("Failed to swap");
				e.printStackTrace();
			}
		}, 0, 10, TimeUnit.SECONDS);
		int nThreads = 4;
		ExecutorService exec = Executors.newFixedThreadPool(nThreads);
		for (int i = 0; i < nThreads; i++) {
			exec.submit(() -> {
				// server threads which do evaluations of the model
				// an infinite loop here, but in reality handling requests
				while (true) {
					try {
						OrtSession session = sessions.get(SESSION_NAME);
						if (session == null) {
							continue;
						}
						try (OnnxTensor tensor = OnnxTensor.createTensor(env,
								new float[][] { { 0, 24603, 1800000 } })) {
							Map<String, OnnxTensor> input = new HashMap<>();
							input.put("float_input", tensor);
							try (Result result = session.run(input)) {
								float[][] value = (float[][]) result.get(0).getValue();
								if (value[0][0] != 0.56f) {
									System.err.println("bad value");
								}
							}
						}
					} catch (Exception e) {
						System.err.println("Failed evaluation");
						e.printStackTrace();
					}
				}
			});
		}
		// shutdown hooks would normally be here
	}

}

output.log

SWAP START
LOADED
SWAP DONE
SWAP START
LOADED
SWAP DONE
SWAP START
LOADED
#
# A fatal error has been detected by the Java Runtime Environment:
#
#  SIGSEGV (0xb) at pc=0x000000011941afb6, pid=65753, tid=0x0000000000003e03
#
# JRE version: Java(TM) SE Runtime Environment (8.0_121-b13) (build 1.8.0_121-b13)
# Java VM: Java HotSpot(TM) 64-Bit Server VM (25.121-b13 mixed mode bsd-amd64 compressed oops)
# Problematic frame:
# C  [libonnxruntime.dylib+0x32fb6]  std::__1::__hash_const_iterator<std::__1::__hash_node<std::__1::__hash_value_type<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, onnxruntime::InferenceSession::InputDefMetaData>, void*>*> std::__1::__hash_table<std::__1::__hash_value_type<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, onnxruntime::InferenceSession::InputDefMetaData>, std::__1::__unordered_map_hasher<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, std::__1::__hash_value_type<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, onnxruntime::InferenceSession::InputDefMetaData>, std::__1::hash<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > >, true>, std::__1::__unordered_map_equal<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, std::__1::__hash_value_type<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, onnxruntime::InferenceSession::InputDefMetaData>, std::__1::equal_to<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > >, true>, std::__1::allocator<std::__1::__hash_value_type<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, onnxruntime::InferenceSession::InputDefMetaData> > >::find<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > >(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&) const+0x86
#
# Failed to write core dump. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again
#
# An error report file with more information is saved as:
# /hs_err_pid65753.log
SWAP DONE
#
# If you would like to submit a bug report, please visit:
#   http://bugreport.java.com/bugreport/crash.jsp
# The crash happened outside the Java Virtual Machine in native code.
# See problematic frame for where to report the bug.
#