| Safe Haskell | None |
|---|---|
| Language | Haskell2010 |
Futhark.Pass.ExtractKernels.StreamKernel
Synopsis
- segThreadCapped :: MonadFreshNames m => MkSegLevel GPU m
- streamRed :: (MonadFreshNames m, HasScope GPU m) => MkSegLevel GPU m -> Pat Type -> SubExp -> Commutativity -> Lambda GPU -> Lambda GPU -> [SubExp] -> [VName] -> m (Stms GPU)
- streamMap :: (MonadFreshNames m, HasScope GPU m) => MkSegLevel GPU m -> [String] -> [PatElem Type] -> SubExp -> Commutativity -> Lambda GPU -> [SubExp] -> [VName] -> m ((SubExp, [VName]), Stms GPU)
Documentation
segThreadCapped :: MonadFreshNames m => MkSegLevel GPU m Source #
Like segThread, but cap the thread count to the input size.
This is more efficient for small kernels, e.g. summing a small
array.