101 lines
3.6 KiB
Java
Raw Normal View History

/*
* ******************************************************************************
* *
* *
* * This program and the accompanying materials are made available under the
* * terms of the Apache License, Version 2.0 which is available at
* * https://www.apache.org/licenses/LICENSE-2.0.
* *
* * See the NOTICE file distributed with this work for additional
* * information regarding copyright ownership.
* * Unless required by applicable law or agreed to in writing, software
* * distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
* * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
* * License for the specific language governing permissions and limitations
* * under the License.
* *
* * SPDX-License-Identifier: Apache-2.0
* *****************************************************************************
*/
2019-06-06 15:21:15 +03:00
package org.nd4j.jita.constant;
Refactor packages to fix split package issues (#411) * Refactor nd4j-common: org.nd4j.* -> org.nd4j.common.* Signed-off-by: Alex Black <blacka101@gmail.com> * Fix CUDA (missed nd4j-common package refactoring changes) Signed-off-by: Alex Black <blacka101@gmail.com> * nd4j-kryo: org.nd4j -> org.nd4j.kryo Signed-off-by: Alex Black <blacka101@gmail.com> * Fix nd4j-common for deeplearning4j-cuda Signed-off-by: Alex Black <blacka101@gmail.com> * nd4j-grppc-client: org.nd4j.graph -> org.nd4j.remote.grpc Signed-off-by: Alex Black <blacka101@gmail.com> * deeplearning4j-common: org.deeplearning4.* -> org.deeplearning4j.common.* Signed-off-by: Alex Black <blacka101@gmail.com> * deeplearning4j-core: org.deeplearning4j.* -> org.deeplearning.core.* Signed-off-by: Alex Black <blacka101@gmail.com> * deeplearning4j-cuda: org.deeplearning4j.nn.layers.* -> org.deeplearning4j.cuda.* Signed-off-by: Alex Black <blacka101@gmail.com> * Import fixes Signed-off-by: Alex Black <blacka101@gmail.com> * deeplearning4j-nlp-*: org.deeplearning4.text.* -> org.deeplearning4j.nlp.(language).* Signed-off-by: Alex Black <blacka101@gmail.com> * deeplearning4j-ui-model: org.deeplearning4j.ui -> org.deeplearning4j.ui.model Signed-off-by: Alex Black <blacka101@gmail.com> * datavec-spark-inference-{server/model/client}: org.datavec.spark.transform -> org.datavec.spark.inference.{server/model/client} Signed-off-by: Alex Black <blacka101@gmail.com> * datavec-jdbc: org.datavec.api -> org.datavec.jdbc Signed-off-by: Alex Black <blacka101@gmail.com> * Delete org.deeplearning4j.datasets.iterator.impl.MultiDataSetIteratorAdapter in favor of (essentially identical) org.nd4j.linalg.dataset.adapter.MultiDataSetIteratorAdapter Signed-off-by: Alex Black <blacka101@gmail.com> * ND4S fixes Signed-off-by: Alex Black <blacka101@gmail.com> * Fixes Signed-off-by: Alex Black <blacka101@gmail.com> * nd4j-common-tests: org.nd4j.* -> org.nd4j.common.tests Signed-off-by: Alex Black <blacka101@gmail.com> * Trigger CI Signed-off-by: Alex Black <blacka101@gmail.com> * Fixes Signed-off-by: Alex Black <blacka101@gmail.com> * #8878 Ignore CUDA tests on modules with 'nd4j-native under cuda' issue Signed-off-by: Alex Black <blacka101@gmail.com> * Fix bad imports in tests Signed-off-by: Alex Black <blacka101@gmail.com> * Add ignore on test (already failing) due to #8882 Signed-off-by: Alex Black <blacka101@gmail.com> * Import fixes Signed-off-by: Alex Black <blacka101@gmail.com> * Additional import fixes Signed-off-by: Alex Black <blacka101@gmail.com>
2020-04-29 11:19:26 +10:00
import org.nd4j.common.primitives.Pair;
2019-06-06 15:21:15 +03:00
import org.nd4j.linalg.api.buffer.DataBuffer;
import org.nd4j.linalg.api.shape.LongShapeDescriptor;
2019-06-06 15:21:15 +03:00
import org.nd4j.linalg.api.shape.ShapeDescriptor;
import org.nd4j.linalg.factory.Nd4j;
import java.util.ArrayList;
import java.util.List;
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.CopyOnWriteArrayList;
/**
* This class implements storage singleton, to guarantee constant buffers persistence
*
* @author raver119@gmail.com
*/
public class ConstantProtector {
2022-10-21 15:19:32 +02:00
private static final ConstantProtector ourInstance = new ConstantProtector();
2019-06-06 15:21:15 +03:00
public static ConstantProtector getInstance() {
return ourInstance;
}
2022-10-21 15:19:32 +02:00
private final List<DataBuffer> protectorLegacy = new CopyOnWriteArrayList<>();
2019-06-06 15:21:15 +03:00
private List<Pair<DataBuffer, long[]>> protector = new CopyOnWriteArrayList<>();
private List<Map<LongShapeDescriptor, Pair<DataBuffer, long[]>>> deviceCache = new ArrayList<>();
private ConstantProtector() {
purgeProtector();
}
public void purgeProtector() {
protector = new CopyOnWriteArrayList<>();
deviceCache = new ArrayList<>();
int numDevices = Nd4j.getAffinityManager().getNumberOfDevices();
for (int i = 0; i < numDevices; i++) {
deviceCache.add(i, new ConcurrentHashMap<LongShapeDescriptor, Pair<DataBuffer, long[]>>());
}
}
public void persistDataBuffer(DataBuffer buffer) {
protectorLegacy.add(buffer);
}
public void persistDataBuffer(Pair<DataBuffer, long[]> buffer) {
protector.add(buffer);
}
public void persistDataBuffer(int deviceId, ShapeDescriptor descriptor, Pair<DataBuffer, long[]> buffer) {
deviceCache.get(deviceId).put(LongShapeDescriptor.fromShapeDescriptor(descriptor), buffer);
}
public void persistDataBuffer(int deviceId, LongShapeDescriptor descriptor, Pair<DataBuffer, long[]> buffer) {
deviceCache.get(deviceId).put(descriptor, buffer);
}
public Pair<DataBuffer, long[]> getDataBuffer(int deviceId, ShapeDescriptor descriptor) {
return deviceCache.get(deviceId).get(LongShapeDescriptor.fromShapeDescriptor(descriptor));
}
public Pair<DataBuffer, long[]> getDataBuffer(int deviceId, LongShapeDescriptor descriptor) {
return deviceCache.get(deviceId).get(descriptor);
}
public boolean containsDataBuffer(int deviceId, ShapeDescriptor descriptor) {
return deviceCache.get(deviceId).containsKey(LongShapeDescriptor.fromShapeDescriptor(descriptor));
}
public boolean containsDataBuffer(int deviceId, LongShapeDescriptor descriptor) {
return deviceCache.get(deviceId).containsKey(descriptor);
}
}