Android App 运行 进入 java.lang.OutOfMemoryError 而 EDF 文件解析

Android App running into java.lang.OutOfMemoryError while EDF file parsing

我搜索了类似的主题,但找不到与 EDF 文件和内存不足错误相关的任何内容。

所以,这是我的问题: 我正在 Android studio 中开发一个 Android 应用程序,它可以浏览文件、选择“.edf”文件并对其进行解析。但对于某些文件,它会抛出以下错误(使用堆栈跟踪):

I/xample.styleap: Starting a blocking GC Alloc
    Starting a blocking GC Alloc
I/xample.styleap: Alloc concurrent copying GC freed 15804(1107KB) AllocSpace objects, 5(108KB) LOS objects, 86% free, 1955KB/13MB, paused 148us total 44.577ms
    Starting a blocking GC Alloc
I/xample.styleap: Alloc young concurrent copying GC freed 0(15KB) AllocSpace objects, 0(0B) LOS objects, 86% free, 1955KB/13MB, paused 120us total 16.603ms
    Starting a blocking GC Alloc
I/xample.styleap: Alloc concurrent copying GC freed 2019(110KB) AllocSpace objects, 0(0B) LOS objects, 86% free, 1844KB/13MB, paused 83us total 25.049ms
    Forcing collection of SoftReferences for 488MB allocation
    Starting a blocking GC Alloc
I/xample.styleap: Alloc concurrent copying GC freed 2053(69KB) AllocSpace objects, 0(0B) LOS objects, 87% free, 1790KB/13MB, paused 75us total 22.140ms
W/xample.styleap: Throwing OutOfMemoryError "Failed to allocate a 512000012 byte allocation with 12582912 free bytes and 190MB until OOM, target footprint 14416192, growth limit 201326592" (VmSize 1212600 kB)
I/xample.styleap: Starting a blocking GC Alloc
    Starting a blocking GC Alloc
I/xample.styleap: Alloc young concurrent copying GC freed 4(31KB) AllocSpace objects, 0(0B) LOS objects, 87% free, 1774KB/13MB, paused 79us total 15.057ms
I/xample.styleap: Starting a blocking GC Alloc
I/xample.styleap: Alloc concurrent copying GC freed 3(16KB) AllocSpace objects, 0(0B) LOS objects, 87% free, 1774KB/13MB, paused 83us total 23.495ms
    Forcing collection of SoftReferences for 488MB allocation
    Starting a blocking GC Alloc
I/xample.styleap: Alloc concurrent copying GC freed 50(17KB) AllocSpace objects, 0(0B) LOS objects, 87% free, 1788KB/13MB, paused 73us total 22.532ms
W/xample.styleap: Throwing OutOfMemoryError "Failed to allocate a 512000016 byte allocation with 12582912 free bytes and 190MB until OOM, target footprint 14414784, growth limit 201326592" (VmSize 1212600 kB)
D/AndroidRuntime: Shutting down VM
E/AndroidRuntime: FATAL EXCEPTION: main
    Process: com.example.styleapp, PID: 30041
    java.lang.OutOfMemoryError: Failed to allocate a 512000016 byte allocation with 12582912 free bytes and 190MB until OOM, target footprint 14414784, growth limit 201326592
        at br.unb.biologiaanimal.edf.EDFReader.readRecords(EDFReader.java:106)
        at br.unb.biologiaanimal.edf.EDFReader.<init>(EDFReader.java:25)
        at br.unb.biologiaanimal.edf.EDF.<init>(EDF.java:24)
        at com.example.styleapp.MainActivity.runEDFfile(MainActivity.java:100)
        at com.example.styleapp.MainActivity.access[=13=]0(MainActivity.java:24)
        at com.example.styleapp.MainActivity.onMenuItemClick(MainActivity.java:82)
        at androidx.appcompat.view.menu.MenuItemImpl.invoke(MenuItemImpl.java:154)
        at androidx.appcompat.view.menu.MenuBuilder.performItemAction(MenuBuilder.java:991)
        at androidx.appcompat.view.menu.MenuBuilder.performItemAction(MenuBuilder.java:981)
        at androidx.appcompat.widget.ActionMenuView.invokeItem(ActionMenuView.java:625)
        at androidx.appcompat.view.menu.ActionMenuItemView.onClick(ActionMenuItemView.java:151)
        at android.view.View.performClick(View.java:7140)
        at android.view.View.performClickInternal(View.java:7117)
        at android.view.View.access00(View.java:801)
        at android.view.View$PerformClick.run(View.java:27355)
        at android.os.Handler.handleCallback(Handler.java:883)
        at android.os.Handler.dispatchMessage(Handler.java:100)
        at android.os.Looper.loop(Looper.java:214)
        at android.app.ActivityThread.main(ActivityThread.java:7410)
        at java.lang.reflect.Method.invoke(Native Method)
        at com.android.internal.os.RuntimeInit$MethodAndArgsCaller.run(RuntimeInit.java:492)
        at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:935)
W/System: A resource failed to call close.

我正在为 java 使用 EDF 解析器库:https://github.com/lab-neuro-comp/EDF

我的源代码(MainActivity.java):

import br.unb.biologiaanimal.edf.EDF;
EDF edf = new EDF(filepath);

使用特定文件路径调用 EDF 构造函数时失败。 在许多地方,据说由于 JVM/Android 或堆大小中的 VM 不足而导致失败。但我想我也已经处理好了:

里面studio.vmoptions

-Xmx2048m

我想这指定了 Android Studio 的 JVM VM 大小。 我还在打印 Java 虚拟机将尝试使用的最大内存量:

Runtime rt = Runtime.getRuntime();
long maxMemory = rt.maxMemory();

这会打印

201326592

这意味着大约有 201MB 可用。

那么当它成功加载一个4.2MB的EDF文件时,加载一个1.9MB的EDF文件失败(内存不足)的原因是什么?

由于 EDF 解析库的逻辑错误,它失败了。 EDF class 方法查找名为 "duration" 的 header(原因尚不清楚)并将其用于后续算术运算,其中使用 header 值分配数组内存等. 有时持续时间值是浮动的,这会导致错误的算术运算,结果值也会变得非常大。在那里它无法分配内存。 这是代码片段:

private void readRecords(FileInputStream stream)
    throws IOException
    {
        records = new HashMap();
        int[] numberSamples = getNumberSamples();
        int[] sampling = new int[numberSignals];
        int duration = paramToInt("duration");
        int dataRecords = paramToInt("datarecords");
        byte[][] recordList = new byte[numberSignals][5];
        String[] labels = getLabels();
        byte[] buffer = null;

        // Initalizing variables
        for (int i = 0; i < numberSignals; ++i)
        {
            sampling[i] = duration * numberSamples[i];
            recordList[i] = null;
        }

        // Reading records
        for (int d = 0; d < dataRecords; ++d)
        {
            for (int i = 0; i < numberSignals; ++i)
            {
                duration = 2 * sampling[i];
                buffer = new byte[duration];
                stream.read(buffer);
                // TODO Make this insertion faster. Maybe a queue?
                recordList[i] = (recordList[i] == null)?
                    buffer :
                    EDFUtil.insert(recordList[i], buffer);
            }
        }

        // Packing records
        for (int i = 0; i < numberSignals; ++i)
        {
            String label = labels[i].trim();
            records.put(label, recordList[i]);
        }
    }