Claire Alvis
Claire Alvis
We do a lot of code transformations, and it may be that we end up with something that isn't well-typed. It might be worth adding another typechecking pass towards the...
I experienced this problem trying to get Harlan installed on my Arch desktop, and it looks like Will is experiencing the same problems with an Ubuntu install. Our compilation of...
During optimization/fusing, we have to walk the kernel body to replace references to get-global-id when the dimension of the argument changes. It would be helpful if we could just refer...
In particular, this means adding an `async-kernel` form. This may not be necessary. However, the language has enough information that we can probably generate asynchronous code for kernels by default,...
It'd be nice to be able to specify stencils with special syntax in a kernel. For example, we could do something like this: ``` (kernel ([(stencil (a b c) (d...
Harlan should be memory safe, but it's not if we don't check to make sure array accesses are in bounds. We currently check (some) vector-refs, but we do not check...
We want to mark tests that are expected to fail at runtime, by design. While we're at it, we should have a compile-fail testspec to make sure we reject all...
This would be good if we start getting more users. We could probably do it under MinG32.
We're planning on enforcing that kernels may not write to free variables, only locals. This means we can get away with using read-only memory for arguments and free variables, and...