Fix new NativePointer(number) for 32-bit BE architectures (e.g. ARM BE8).
This union which overlaps gpointer and int64_t fails to allow extraction of
gpointer when architecture is 32-bit big endian. For that combination padding
is required, which this patch adds only for that combination.
Fix new NativePointer(number) for 32-bit BE architectures (e.g. ARM BE8).
This union which overlaps gpointer and int64_t fails to allow extraction of gpointer when architecture is 32-bit big endian. For that combination padding is required, which this patch adds only for that combination.