B4J Question Canvas Snapshot Crash

[RobertK]

To follow up on

Research Resuts: Calculating 3D Land Surface Area

Hello everyone, Those of you working with GIS might be interested in the following recent research paper describing 3D land surface area calculations implemented in B4J. https://www.researchgate.net/publication/355179797 Many thanks to you Erel, and to members of the Forum, for providing a...

www.b4x.com

Sorry I have been away longer than expected.

Regarding your question about whether Fast3D is open source, Erel, I first need to sort out some outstanding issues. Foremost among these is the Canvas Snapshot crash first reported in https://www.b4x.com/android/forum/threads/canvas-snapshot-size-limit.127859/

The maximum size of the canvas is now kept well below the upper limit, however Snapshot still crashes erratically with the following java.lang.NullPointerException:

java.lang.NullPointerException
    at javafx.graphics/com.sun.prism.d3d.D3DTexture.getContext(D3DTexture.java:84)
    at javafx.graphics/com.sun.prism.d3d.D3DTexture.update(D3DTexture.java:207)
    at javafx.graphics/com.sun.prism.d3d.D3DTexture.update(D3DTexture.java:151)
    at javafx.graphics/com.sun.prism.impl.BaseContext.flushMask(BaseContext.java:109)
    at javafx.graphics/com.sun.prism.impl.BaseContext.drawQuads(BaseContext.java:118)
    at javafx.graphics/com.sun.prism.impl.VertexBuffer.flush(VertexBuffer.java:98)
    at javafx.graphics/com.sun.prism.impl.BaseContext.flushVertexBuffer(BaseContext.java:101)
    at javafx.graphics/com.sun.prism.impl.ps.BaseShaderContext.setRenderTarget(BaseShaderContext.java:746)
    at javafx.graphics/com.sun.prism.impl.BaseContext.setRenderTarget(BaseContext.java:131)
    at javafx.graphics/com.sun.prism.impl.BaseGraphics.<init>(BaseGraphics.java:106)
    at javafx.graphics/com.sun.prism.impl.ps.BaseShaderGraphics.<init>(BaseShaderGraphics.java:86)
    at javafx.graphics/com.sun.prism.d3d.D3DGraphics.<init>(D3DGraphics.java:42)
    at javafx.graphics/com.sun.prism.d3d.D3DGraphics.create(D3DGraphics.java:65)
    at javafx.graphics/com.sun.prism.d3d.D3DRTTexture.createGraphics(D3DRTTexture.java:77)
    at javafx.graphics/com.sun.prism.d3d.D3DResourceFactory.createRTTexture(D3DResourceFactory.java:316)
    at javafx.graphics/com.sun.prism.d3d.D3DResourceFactory.createRTTexture(D3DResourceFactory.java:270)
    at javafx.graphics/com.sun.prism.d3d.D3DResourceFactory.createRTTexture(D3DResourceFactory.java:63)
    at javafx.graphics/com.sun.scenario.effect.impl.prism.ps.PPSDrawable.create(PPSDrawable.java:59)
    at javafx.graphics/com.sun.scenario.effect.impl.prism.ps.PPSRenderer.createCompatibleImage(PPSRenderer.java:163)
    at javafx.graphics/com.sun.scenario.effect.impl.prism.ps.PPSRenderer.createCompatibleImage(PPSRenderer.java:67)
    at javafx.graphics/com.sun.scenario.effect.impl.ImagePool.checkOut(ImagePool.java:183)
    at javafx.graphics/com.sun.scenario.effect.impl.Renderer.getCompatibleImage(Renderer.java:116)
    at javafx.graphics/com.sun.scenario.effect.impl.prism.ps.PPSRenderer.getCompatibleImage(PPSRenderer.java:168)
    at javafx.graphics/com.sun.scenario.effect.impl.prism.ps.PPSTwoSamplerPeer.filterImpl(PPSTwoSamplerPeer.java:62)
    at javafx.graphics/com.sun.scenario.effect.impl.prism.ps.PPSEffectPeer.filter(PPSEffectPeer.java:54)
    at javafx.graphics/com.sun.scenario.effect.CoreEffect.filterImageDatas(CoreEffect.java:106)
    at javafx.graphics/com.sun.scenario.effect.Blend.filterImageDatas(Blend.java:37)
    at javafx.graphics/com.sun.scenario.effect.FilterEffect.filter(FilterEffect.java:195)
    at javafx.graphics/com.sun.javafx.sg.prism.NGCanvas.applyEffectOnAintoC(NGCanvas.java:738)
    at javafx.graphics/com.sun.javafx.sg.prism.NGCanvas.blendAthruBintoC(NGCanvas.java:768)
    at javafx.graphics/com.sun.javafx.sg.prism.NGCanvas.renderClip(NGCanvas.java:724)
    at javafx.graphics/com.sun.javafx.sg.prism.NGCanvas.initClip(NGCanvas.java:711)
    at javafx.graphics/com.sun.javafx.sg.prism.NGCanvas.renderStream(NGCanvas.java:1056)
    at javafx.graphics/com.sun.javafx.sg.prism.NGCanvas.renderContent(NGCanvas.java:607)
    at javafx.graphics/com.sun.javafx.sg.prism.NGNode.doRender(NGNode.java:2072)
    at javafx.graphics/com.sun.javafx.sg.prism.NGNode.render(NGNode.java:1964)
    at javafx.graphics/com.sun.javafx.tk.quantum.QuantumToolkit$5.draw(QuantumToolkit.java:1411)
    at javafx.graphics/com.sun.javafx.tk.quantum.QuantumToolkit$5.run(QuantumToolkit.java:1447)
    at java.base/java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:515)
    at java.base/java.util.concurrent.FutureTask.runAndReset(FutureTask.java:305)
    at javafx.graphics/com.sun.javafx.tk.RenderJob.run(RenderJob.java:58)
    at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1128)
    at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:628)
    at javafx.graphics/com.sun.javafx.tk.quantum.QuantumRenderer$PipelineRunnable.run(QuantumRenderer.java:125)
    at java.base/java.lang.Thread.run(Thread.java:834)

This would appear to correspond to a OpenJDK bug reported here

Loading...

bugs.openjdk.java.net

Can anyone suggest workarounds?

Many thanks!

 

[Erel]

It crashes the app, right?

Unless you are able to find what triggers this error, your best option is to catch it: https://www.b4x.com/android/forum/threads/unhandled-exceptions.77506/#content

 

[RobertK]

Thanks for your quick reply, Erel.

No, the error does not crash the app right away, but once the error has occurred all subsequent calls to Canvas.Shapshot throw the same error, as described in the JDK Bug Report:

“After this moment, every GUI rendering causes a NullPointerException (triggered by the method BaseContext.flushMask()), meaning an endless loop causing a GUI freeze.”

After 129 times, the app crashes.

Finding exactly what triggers the error is not easy. The Canvas.Shapshot statement is inside nested loops and is executed very frequently. Sometimes it throws the error almost immediately, at other times only after hundreds of thousands of calls. It does not seem to depend on the canvas size nor the number of vertices in the clipping polygon applied to the canvas using ClipPath. My code does trap the error when it occurs but is unable to resume its calculations because Canvas.Shapshot has stopped working.

As a temporary patch, I have implemented a monitor app which runs as an independent process. When the main app traps the error it raises a flag (in a file) before terminating. The monitor app detects the flag and restarts the main app which continues its calculations where it left off. This works but is not a satisfactory long term solution.

I have also tried or considered the following:

1) Allocating additional memory using -Xmx10G
It does not help.

Would using -Dprism.maxvram=10G be more effective?
But this is still not a satisfactory long term solution, crashes would eventually occur when processing large datasets.

2) Forcing garbage collection at regular intervals as outlined here

Garbage collection

Hello For those of you interested in GC here is a class I have in a B4J server: Sub Class_Globals Private tmrHook As Timer Private gcTime As Long Private freq_gc As Int = 1 End Sub Public Sub Initialize ' // 3 seconds...

www.b4x.com

It does not help.

Is there a way to force the specific cleanup referred to in the JDK Bug Report?

“If the BaseResourcePool runs out of available space (maxSize is reached), it performs a cleanup to remove dead/unused textures and free some space.”

3) Do you think that switching from Open JDK 11 to Oracle Java 8 might help?

Any other ideas?

Thanks.

 

[Erel]

Do you think that switching from Open JDK 11 to Oracle Java 8 might help?

Worth testing.

Also worth testing with Open JDK 14: https://www.b4x.com/android/forum/threads/b4j-v8-30-beta-is-available-for-download.117877/#content

 

[RobertK]

Thanks for your latest suggestions, Erel.

Unfortunately Open JDK 14 does not help, it has the same bug. I have not yet tried Oracle Java 8.

In the meantime, allocating additional memory with -Dprism.maxvram allows me to process large datasets. Perhaps this can tide me over until they fix the bug. Or am I being too optimistic? The original bug report is dated 2017, last updated 2018. How long do such fixes usually take?

 

[Erel]

Or am I being too optimistic?

The original bug report is dated 2017, last updated 2018. How long do such fixes usually take?

It is impossible to say.