Re-using array buffers + multiple Float32Arrays passed to Emscripten's HEAP

[Pawel Misiurski]

Hi I'm completely new to Emscripten and I have a problem which I posted on stack overflow, but it's a very niche thing I'm trying to find someone who tried this already.

http://stackoverflow.com/questions/31550198/emscripten-uncaught-rangeerror-source-is-too-large-multiple-float32arrays

It looks like once an array has been added to Emscripten's Heap it contains all it's data(only in the buffer).Therefore when trying to send it again to this function I'm getting "Source is too large" error. After 2 days maybe I'm missing something or maybe my code is completely wrong? I only want to add that it is not caused by my C++ code because even after commenting out the function call I'm still getting the error.

Thanks for help and in the meantime I'm coming back to pulling my hair out.

[Alon Zakai]

Where does data2 come from? If you trace that back, you should be able to see why it is the wrong array or size.

--
You received this message because you are subscribed to the Google Groups "emscripten-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to emscripten-disc...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Pawel Misiurski]

Thanks for your quick answer Alon! Data2 has at first normal size: 64 bytes, but sometimes returned value "return" is being again reused as a parameter. So when I use debugger to see sizes of arraybuffers all data.buffer.byteLength, data2.buffer.byteLength and targetdata.buffer.byteLength return 64 bytes. But when I check result.buffer.byteLength it returns exactly 16777216 bytes.

So this line:  var result = new Float32Array(targetHeap.buffer, targetHeap.byteOffset, 16);
produces a result with value as expected:
[0.000001385850055157789, -1, 7.756424391658356e-10, 0, -1, -0.000001385850055157789, 0.0000013868001360606286, 0, -0.0000013868001360606286, -7.775643462437642e-10, -1, 0, 0, 0, 0, 1]

but after checking with debugger in the console result.buffer.byteLength the length is 16777216 bytes, same as targetHeap.buffer. Now it's impossible to feed it back into this function(so it's buffer will be used in this line: dataHeap2.set( new Uint8Array(data2.buffer)  ).

Is it an error of Float32Array that it doesn't return a buffer of byteLength of 64 or is it because Uint8Array should atomatically truncate input? Or is it because the design of my function is wrong?

Following "Uint8Array should atomatically truncate input" I made a change to limit amount of data coming to Uint8Array by adding data2.byteOffset and nDataBytes parameters:
dataHeap2.set( new Uint8Array(data2.buffer, data2.byteOffset, nDataBytes) );

what's interesting is that the game is now running, but the screen is white. At least it's not crashing. Something is still very wrong but it looks like tons of calculations are happening now. Only wrong.


The returned value after checking it with debugger() on first run

You received this message because you are subscribed to a topic in the Google Groups "emscripten-discuss" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/emscripten-discuss/JvDRhCWpY2c/unsubscribe.
To unsubscribe from this group and all its topics, send an email to emscripten-disc...@googlegroups.com.

[Alon Zakai]

If you tell it to give you a 16-byte Float32Array by passing a length of 16, you get a 16-byte FloatArray. But it is a view on the same ArrayBuffer, which can be much larger. So the .buffer of that view will potentially be larger than 16.