News for hackers. Stuff that matters.

A comment on Harlan, new Lispy language for GPU programming. Parent comment

Re:Harlan, new Lispy language for GPU programming by iskander on Monday July 01, 2013 @02:12PM (#5972528)
How do you support higher order procedures? Do you do control flow analysis & defunctionalization or are these "real" higher order functions?
[ Parent | on HN ]
Re:Harlan, new Lispy language for GPU programming by eholk on Monday July 01, 2013 @05:30PM (#5973696)
It's basically defunctionalization. It's hard to get access to function pointers on the GPU to do "real" higher order functions.
[ Parent | on HN ]
Re:Harlan, new Lispy language for GPU programming by seanmcdirmid on Monday July 01, 2013 @07:50PM (#5974591)
What version of CUDA are you talking about. I though the latest version supported indirect function pointers (finally).
[ Parent | on HN ]
Re:Harlan, new Lispy language for GPU programming by iskander on Monday July 01, 2013 @10:52PM (#5975516)
That's good news! Did that change with Kepler? Any links to more info?
[ Parent | on HN ]
Re:Harlan, new Lispy language for GPU programming by iskander on Monday July 01, 2013 @07:12PM (#5974359)
I wish they'd at least give us a computed goto...but alas, every switch is destined for predicated branching.
[ Parent | on HN ]