GDL on ARM64 machines ( can be YOUR Mac )

[GillesDuvert]

ARM64 architecture does not handle floating-point to integer conversions identically as INTEL-based computers. See e.g., https://embeddeduse.com/2013/08/25/casting-a-negative-float-to-an-unsigned-int
Would somebody having access to a running IDL on a M1/M2 Mac check if it keeps a consistent behaviour, i.e., that float-to-integer conversion is done identically as on intel64 machines?

To recall, a simple check

IDL> z=ulong(2)
IDL> z[0]=-1.0 ; NOTE: needs a FLOAT 
IDL> z
  4294967295

return for ARM64 based machines would be

IDL> z
                     0

if no special care was taken.