Bug with real / tuples?

Matthew Fluet fluet@CS.Cornell.EDU
Sun, 30 Sep 2001 11:50:22 -0400 (EDT)


Thanks for the bug report.  It is a bug in the native codegen; you can
compile with -native false and get the correct behavior.  I'm looking into
it.

> Hello!
> 
> 
> I've tried compiling some stuff with the new mlton (on Linux
> navtive). And I think I've hit a bug - it probably has to
> do with reals and tuples (but it could be something else of course...).
> 
> I have made a really small program (attached) where the problem occurs.
> 
> Short explanation of the program:
> 
> translateXYZ creates a 4x4 homogenious translation matrix.
> transformPoint transforms a 3D point using the matrix.
> Point3DToString converts a 3D point to a string :)
> 
> When I do
> 
> mlton testMatrix.sml
> ./testMatrix
> 
> I get an output something like this:
> 
> {x = inf, y = inf, z = inf}
> 
> On SML/NJ it correctly becomes:
> 
> {x = 3.0, y = 5.0, z = 7.0}
> 
> In my "real" code I have also seen ~inf and nan
> occur on mlton during similar matrix calculations.
> 
> I'm using my patched version of the 20010806 release
> (which still seems to be the latest official release).
> 
> 
> Cheers
> -- 
> http://www.HardcoreProcessing.com