Merge pull request #28 from omlins/lr-miniapps

GPU method override for `CUDA.pow()`

Project.toml

@@ -1,11 +1,11 @@
 name = "ParallelStencil"
 uuid = "94395366-693c-11ea-3b26-d9b7aac5d958"
 authors = ["Samuel Omlin", "Ludovic Räss"]
-version = "0.5.3"
+version = "0.5.4"
 
 [compat]
 julia = "1"
-CUDA = "1, 2, 3"
+CUDA = "3"
 MacroTools = "0.5"
 
 [deps]

miniapps/HydroMech2D.jl

@@ -3,13 +3,8 @@ using ParallelStencil
 using ParallelStencil.FiniteDifferences2D
 @static if USE_GPU
  @init_parallel_stencil(CUDA, Float64, 2)
- macro pow(args...) esc(:(CUDA.pow($(args...)))) end
- macro tanh(args...) esc(:(CUDA.tanh($(args...)))) end
 else
  @init_parallel_stencil(Threads, Float64, 2)
- pow(x,y) = x^y
- macro pow(args...) esc(:(pow($(args...)))) end
- macro tanh(args...) esc(:(Base.tanh($(args...)))) end
 end
 using Plots, Printf, Statistics, LinearAlgebra
 
@@ -20,8 +15,8 @@ using Plots, Printf, Statistics, LinearAlgebra
 end
 
 @parallel function compute_params_∇!(EtaC::Data.Array, K_muf::Data.Array, Rog::Data.Array, ∇V::Data.Array, ∇qD::Data.Array, Phi::Data.Array, Pf::Data.Array, Pt::Data.Array, Vx::Data.Array, Vy::Data.Array, qDx::Data.Array, qDy::Data.Array, μs::Data.Number, η2μs::Data.Number, R::Data.Number, λPe::Data.Number, k_μf0::Data.Number, ϕ0::Data.Number, nperm::Data.Number, θ_e::Data.Number, θ_k::Data.Number, ρfg::Data.Number, ρsg::Data.Number, ρgBG::Data.Number, dx::Data.Number, dy::Data.Number)
- @all(EtaC) = (1.0-θ_e)*@all(EtaC) + θ_e*( μs/@all(Phi)*η2μs*(1.0+0.5*(1.0/R-1.0)*(1.0+@tanh((@all(Pf)-@all(Pt))/λPe))) )
- @all(K_muf) = (1.0-θ_k)*@all(K_muf) + θ_k*( k_μf0*@pow((@all(Phi)/ϕ0), nperm) )
+ @all(EtaC) = (1.0-θ_e)*@all(EtaC) + θ_e*( μs/@all(Phi)*η2μs*(1.0+0.5*(1.0/R-1.0)*(1.0+tanh((@all(Pf)-@all(Pt))/λPe))) )
+ @all(K_muf) = (1.0-θ_k)*@all(K_muf) + θ_k*( k_μf0 * (@all(Phi)/ϕ0)^nperm )
  @all(Rog) = ρfg*@all(Phi) + ρsg*(1.0-@all(Phi)) - ρgBG
  @all(∇V) = @d_xa(Vx)/dx + @d_ya(Vy)/dy
  @all(∇qD) = @d_xa(qDx)/dx + @d_ya(qDy)/dy

miniapps/scalar_porowaves2D.jl

@@ -3,16 +3,13 @@ using ParallelStencil
 using ParallelStencil.FiniteDifferences2D
 @static if USE_GPU
  @init_parallel_stencil(CUDA, Float64, 2)
- macro pow(args...) esc(:(CUDA.pow($(args...)))) end
 else
  @init_parallel_stencil(Threads, Float64, 2)
- pow(x,y) = x^y
- macro pow(args...) esc(:(pow($(args...)))) end
 end
 using Plots, Printf, Statistics
 
 @parallel function compute_nonlin!(K_mu::Data.Array, EtaR::Data.Array, Phi::Data.Array, Eta::Data.Array, k_μ0::Data.Number, n::Data.Number)
- @all(K_mu) = k_μ0*@pow(@all(Phi), n)
+ @all(K_mu) = k_μ0 * @all(Phi)^n
  @all(EtaR) = @all(Eta) 
  return
 end