I am getting a primitive error #9 (Unknown floating point exception) when calling the following Platform Function with a Signalling NaN. Can anyone from Instantiations confirm whether this is an expected behaviour, and if so whether there is anything I can do to prevent it?
<c: int32 'gcirpc32':GciFloatKind double>
The purpose of the function is to return an integer categorizing the argument. It isn't helpful when it yields a primitive error.
I have also tried it with <c: int32 'gcirpc32':GciFloatKind float>, and the same result.
I found the following information in the Programmer's Guide, but it doesn't really address this topic.
"float, double, extended
If used as a parameter type, the parameter must be a Float. The parameter is converted to the specified size of float."
At this point, I am guessing that the "conversion" to a C double is acting on the "signalling" aspect. I do not see this problem with Quiet NaNs, nor with Infinity values.
In case it matters, the exact representation of the Float is FF F0 00 00 00 00 00 01.
--
You received this message because you are subscribed to the Google Groups "VA Smalltalk" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
[hidden email].
To post to this group, send email to
[hidden email].
Visit this group at
http://groups.google.com/group/va-smalltalk.
For more options, visit
https://groups.google.com/groups/opt_out.
Locked
