Non 32-bit float bitcasting functions

Vulkan
1

Hi,

Are there any 16 or 64-bit equivalents to the floatBitsToUint and uintBitsToFloat functions?
I’m mainly working with GLSL that compiles to SPIR-V through shaderc, so I was wondering if there is a way to access bitcasting instructions for this purpose.

Thank you

2

halfBitsToUInt16 and uint16BitsToHalf, resp. doubleBitsToUint64 and uint64BitsToDouble?

3

Oh, I wasn’t awere these exist. Thank you very much!

On a related note, I had to replace halfBitsToUInt16 with float16BitsToUint16 when using float16_t. I suppose this is due to stricter type conversion rules.

4

Weird. Perhaps the function names differ depending on which extension introduces the arithmetic types…