Merge #978

978: Extend and fix Ref broadcasting with StencilCoefs r=charleskawczynski a=charleskawczynski This PR extends and fixes `Ref` broadcasting with StencilCoefs (#963), solution was basically the same as the one applied in #945. Closes #963. I tried applying only `copyto!`, and only `apply_stencil!`, but that didn't work so I applied to all three methods (which does work). Co-authored-by: Charles Kawczynski <kawczynski.charles@gmail.com>
CliMA · Oct 4, 2022 · f2eb09c · f2eb09c
2 parents 64b5023 + 9779510
commit f2eb09c
Show file tree

Hide file tree

Showing 8 changed files with 556 additions and 497 deletions.
diff --git a/.buildkite/pipeline.yml b/.buildkite/pipeline.yml
@@ -118,6 +118,7 @@ steps:
       - "julia -O0 --color=yes --check-bounds=yes --project=test test/Operators/finitedifference/implicit_stencils.jl --float_type Float32"
     agents:
       config: cpu
+      slurm_mem: 10GB
       queue: central
       slurm_ntasks: 1
 
@@ -127,6 +128,7 @@ steps:
       - "julia -O0 --color=yes --check-bounds=yes --project=test test/Operators/finitedifference/implicit_stencils.jl --float_type Float64"
     agents:
       config: cpu
+      slurm_mem: 10GB
       queue: central
       slurm_ntasks: 1
 
@@ -775,6 +777,7 @@ steps:
       - "examples/hybrid/sphere/output/deformation_flow/*"
     agents:
       config: cpu
+      slurm_mem: 20GB
       queue: central
       slurm_ntasks: 1
 
@@ -786,6 +789,7 @@ steps:
       - "examples/hybrid/sphere/output/hadley_circulation/*"
     agents:
       config: cpu
+      slurm_mem: 20GB
       queue: central
       slurm_ntasks: 1
 
@@ -813,6 +817,7 @@ steps:
       TEST_NAME: "sphere/baroclinic_wave_rhoe"
     agents:
       config: cpu
+      slurm_mem: 10GB
       queue: central
       slurm_ntasks: 1
 
@@ -827,6 +832,7 @@ steps:
       CLIMACORE_DISTRIBUTED: "MPI"
     agents:
       config: cpu
+      slurm_mem: 20GB
       queue: central
       slurm_ntasks: 2
 
@@ -841,6 +847,7 @@ steps:
       TEST_NAME: "sphere/baroclinic_wave_rhotheta"
     agents:
       config: cpu
+      slurm_mem: 10GB
       queue: central
       slurm_ntasks: 1
 
@@ -874,6 +881,7 @@ steps:
       TEST_NAME: "sphere/balanced_flow_rhoe"
     agents:
       config: cpu
+      slurm_mem: 20GB
       queue: central
       slurm_ntasks: 1
 
@@ -900,6 +908,7 @@ steps:
       TEST_NAME: "sphere/held_suarez_rhoe"
     agents:
       config: cpu
+      slurm_mem: 10GB
       queue: central
       slurm_ntasks: 1
 
@@ -914,6 +923,7 @@ steps:
       FLOAT_TYPE: "Float64"
     agents:
       config: cpu
+      slurm_mem: 20GB
       queue: central
       slurm_ntasks: 1
 
@@ -940,6 +950,7 @@ steps:
       TEST_NAME: "sphere/held_suarez_rhoe_int"
     agents:
       config: cpu
+      slurm_mem: 10GB
       queue: central
       slurm_ntasks: 1
 
@@ -967,6 +978,7 @@ steps:
       Z_STRETCH: "true"
     agents:
       config: cpu
+      slurm_mem: 10GB
       queue: central
       slurm_ntasks: 1
 

diff --git a/Project.toml b/Project.toml
@@ -1,7 +1,7 @@
 name = "ClimaCore"
 uuid = "d414da3d-4745-48bb-8d80-42e94e092884"
 authors = ["CliMA Contributors <clima-software@caltech.edu>"]
-version = "0.10.15"
+version = "0.10.16"
 
 [deps]
 Adapt = "79e6a3ab-5dfb-504d-930d-738a2a938a0e"

diff --git a/docs/Manifest.toml b/docs/Manifest.toml
@@ -205,7 +205,7 @@ version = "0.3.0"
 deps = ["Adapt", "BlockArrays", "CUDA", "ClimaComms", "CubedSphere", "DataStructures", "DiffEqBase", "DocStringExtensions", "ForwardDiff", "GaussQuadrature", "GilbertCurves", "HDF5", "InteractiveUtils", "IntervalSets", "LinearAlgebra", "PkgVersion", "RecursiveArrayTools", "RootSolvers", "Rotations", "SparseArrays", "Static", "StaticArrays", "Statistics", "UnPack"]
 path = ".."
 uuid = "d414da3d-4745-48bb-8d80-42e94e092884"
-version = "0.10.15"
+version = "0.10.16"
 
 [[deps.ClimaCoreMakie]]
 deps = ["ClimaCore", "Makie"]

diff --git a/examples/Manifest.toml b/examples/Manifest.toml
@@ -170,7 +170,7 @@ version = "0.3.2"
 deps = ["Adapt", "BlockArrays", "CUDA", "ClimaComms", "CubedSphere", "DataStructures", "DiffEqBase", "DocStringExtensions", "ForwardDiff", "GaussQuadrature", "GilbertCurves", "HDF5", "InteractiveUtils", "IntervalSets", "LinearAlgebra", "PkgVersion", "RecursiveArrayTools", "RootSolvers", "Rotations", "SparseArrays", "Static", "StaticArrays", "Statistics", "UnPack"]
 path = ".."
 uuid = "d414da3d-4745-48bb-8d80-42e94e092884"
-version = "0.10.15"
+version = "0.10.16"
 
 [[deps.ClimaCorePlots]]
 deps = ["ClimaCore", "RecipesBase", "StaticArrays", "TriplotBase"]
@@ -529,10 +529,10 @@ uuid = "61eb1bfa-7361-4325-ad38-22787b887f55"
 version = "0.16.4"
 
 [[deps.GR]]
-deps = ["Base64", "DelimitedFiles", "GR_jll", "HTTP", "JSON", "Libdl", "LinearAlgebra", "Pkg", "Printf", "Random", "RelocatableFolders", "Serialization", "Sockets", "Test", "UUIDs"]
-git-tree-sha1 = "0ac6f27e784059c68b987f42b909ade0bcfabe69"
+deps = ["Base64", "DelimitedFiles", "GR_jll", "HTTP", "JSON", "Libdl", "LinearAlgebra", "Pkg", "Preferences", "Printf", "Random", "Serialization", "Sockets", "Test", "UUIDs"]
+git-tree-sha1 = "859e9cfa91d21ed87e8ea9f2998c94df1de12045"
 uuid = "28b8d3ca-fb5f-59d9-8090-bfdbd6d07a71"
-version = "0.68.0"
+version = "0.69.2"
 
 [[deps.GR_jll]]
 deps = ["Artifacts", "Bzip2_jll", "Cairo_jll", "FFMPEG_jll", "Fontconfig_jll", "GLFW_jll", "JLLWrappers", "JpegTurbo_jll", "Libdl", "Libtiff_jll", "Pixman_jll", "Pkg", "Qt5Base_jll", "Zlib_jll", "libpng_jll"]
@@ -671,9 +671,9 @@ version = "1.0.0"
 
 [[deps.JLD2]]
 deps = ["FileIO", "MacroTools", "Mmap", "OrderedCollections", "Pkg", "Printf", "Reexport", "TranscodingStreams", "UUIDs"]
-git-tree-sha1 = "6c38bbe47948f74d63434abed68bdfc8d2c46b99"
+git-tree-sha1 = "0d0ad913e827d13c5e88a73f9333d7e33c424576"
 uuid = "033835bb-8acc-5ee8-8aae-3f567f8a3819"
-version = "0.4.23"
+version = "0.4.24"
 
 [[deps.JLFzf]]
 deps = ["Pipe", "REPL", "Random", "fzf_jll"]
@@ -1146,9 +1146,9 @@ version = "1.3.1"
 
 [[deps.Plots]]
 deps = ["Base64", "Contour", "Dates", "Downloads", "FFMPEG", "FixedPointNumbers", "GR", "JLFzf", "JSON", "LaTeXStrings", "Latexify", "LinearAlgebra", "Measures", "NaNMath", "Pkg", "PlotThemes", "PlotUtils", "Printf", "REPL", "Random", "RecipesBase", "RecipesPipeline", "Reexport", "RelocatableFolders", "Requires", "Scratch", "Showoff", "SnoopPrecompile", "SparseArrays", "Statistics", "StatsBase", "UUIDs", "UnicodeFun", "Unzip"]
-git-tree-sha1 = "284a353a34a352a95fca1d61ea28a0d48feaf273"
+git-tree-sha1 = "f60a3090028cdf16b33a62f97eaedf67a6509824"
 uuid = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
-version = "1.34.4"
+version = "1.35.0"
 
 [[deps.Polyester]]
 deps = ["ArrayInterface", "BitTwiddlingConvenienceFunctions", "CPUSummary", "IfElse", "ManualMemory", "PolyesterWeave", "Requires", "Static", "StrideArraysCore", "ThreadingUtilities"]
@@ -1230,9 +1230,9 @@ version = "1.3.0"
 
 [[deps.RecipesPipeline]]
 deps = ["Dates", "NaNMath", "PlotUtils", "RecipesBase"]
-git-tree-sha1 = "e7eac76a958f8664f2718508435d058168c7953d"
+git-tree-sha1 = "017f217e647cf20b0081b9be938b78c3443356a0"
 uuid = "01d81517-befc-4cb6-b9ec-a95719d0359c"
-version = "0.6.3"
+version = "0.6.6"
 
 [[deps.RecursiveArrayTools]]
 deps = ["Adapt", "ArrayInterfaceCore", "ArrayInterfaceStaticArraysCore", "ChainRulesCore", "DocStringExtensions", "FillArrays", "GPUArraysCore", "IteratorInterfaceExtensions", "LinearAlgebra", "RecipesBase", "StaticArraysCore", "Statistics", "Tables", "ZygoteRules"]

diff --git a/perf/Manifest.toml b/perf/Manifest.toml
@@ -181,7 +181,7 @@ version = "0.3.2"
 deps = ["Adapt", "BlockArrays", "CUDA", "ClimaComms", "CubedSphere", "DataStructures", "DiffEqBase", "DocStringExtensions", "ForwardDiff", "GaussQuadrature", "GilbertCurves", "HDF5", "InteractiveUtils", "IntervalSets", "LinearAlgebra", "PkgVersion", "RecursiveArrayTools", "RootSolvers", "Rotations", "SparseArrays", "Static", "StaticArrays", "Statistics", "UnPack"]
 path = ".."
 uuid = "d414da3d-4745-48bb-8d80-42e94e092884"
-version = "0.10.15"
+version = "0.10.16"
 
 [[deps.ClimaCorePlots]]
 deps = ["ClimaCore", "RecipesBase", "StaticArrays", "TriplotBase"]
@@ -585,10 +585,10 @@ uuid = "61eb1bfa-7361-4325-ad38-22787b887f55"
 version = "0.16.4"
 
 [[deps.GR]]
-deps = ["Base64", "DelimitedFiles", "GR_jll", "HTTP", "JSON", "Libdl", "LinearAlgebra", "Pkg", "Printf", "Random", "RelocatableFolders", "Serialization", "Sockets", "Test", "UUIDs"]
-git-tree-sha1 = "0ac6f27e784059c68b987f42b909ade0bcfabe69"
+deps = ["Base64", "DelimitedFiles", "GR_jll", "HTTP", "JSON", "Libdl", "LinearAlgebra", "Pkg", "Preferences", "Printf", "Random", "Serialization", "Sockets", "Test", "UUIDs"]
+git-tree-sha1 = "859e9cfa91d21ed87e8ea9f2998c94df1de12045"
 uuid = "28b8d3ca-fb5f-59d9-8090-bfdbd6d07a71"
-version = "0.68.0"
+version = "0.69.2"
 
 [[deps.GR_jll]]
 deps = ["Artifacts", "Bzip2_jll", "Cairo_jll", "FFMPEG_jll", "Fontconfig_jll", "GLFW_jll", "JLLWrappers", "JpegTurbo_jll", "Libdl", "Libtiff_jll", "Pixman_jll", "Pkg", "Qt5Base_jll", "Zlib_jll", "libpng_jll"]
@@ -738,9 +738,9 @@ version = "0.6.5"
 
 [[deps.JLD2]]
 deps = ["FileIO", "MacroTools", "Mmap", "OrderedCollections", "Pkg", "Printf", "Reexport", "TranscodingStreams", "UUIDs"]
-git-tree-sha1 = "6c38bbe47948f74d63434abed68bdfc8d2c46b99"
+git-tree-sha1 = "0d0ad913e827d13c5e88a73f9333d7e33c424576"
 uuid = "033835bb-8acc-5ee8-8aae-3f567f8a3819"
-version = "0.4.23"
+version = "0.4.24"
 
 [[deps.JLFzf]]
 deps = ["Pipe", "REPL", "Random", "fzf_jll"]
@@ -1231,9 +1231,9 @@ version = "1.3.1"
 
 [[deps.Plots]]
 deps = ["Base64", "Contour", "Dates", "Downloads", "FFMPEG", "FixedPointNumbers", "GR", "JLFzf", "JSON", "LaTeXStrings", "Latexify", "LinearAlgebra", "Measures", "NaNMath", "Pkg", "PlotThemes", "PlotUtils", "Printf", "REPL", "Random", "RecipesBase", "RecipesPipeline", "Reexport", "RelocatableFolders", "Requires", "Scratch", "Showoff", "SnoopPrecompile", "SparseArrays", "Statistics", "StatsBase", "UUIDs", "UnicodeFun", "Unzip"]
-git-tree-sha1 = "284a353a34a352a95fca1d61ea28a0d48feaf273"
+git-tree-sha1 = "f60a3090028cdf16b33a62f97eaedf67a6509824"
 uuid = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
-version = "1.34.4"
+version = "1.35.0"
 
 [[deps.Polyester]]
 deps = ["ArrayInterface", "BitTwiddlingConvenienceFunctions", "CPUSummary", "IfElse", "ManualMemory", "PolyesterWeave", "Requires", "Static", "StrideArraysCore", "ThreadingUtilities"]
@@ -1343,9 +1343,9 @@ version = "1.3.0"
 
 [[deps.RecipesPipeline]]
 deps = ["Dates", "NaNMath", "PlotUtils", "RecipesBase"]
-git-tree-sha1 = "e7eac76a958f8664f2718508435d058168c7953d"
+git-tree-sha1 = "017f217e647cf20b0081b9be938b78c3443356a0"
 uuid = "01d81517-befc-4cb6-b9ec-a95719d0359c"
-version = "0.6.3"
+version = "0.6.6"
 
 [[deps.RecursiveArrayTools]]
 deps = ["Adapt", "ArrayInterfaceCore", "ArrayInterfaceStaticArraysCore", "ChainRulesCore", "DocStringExtensions", "FillArrays", "GPUArraysCore", "IteratorInterfaceExtensions", "LinearAlgebra", "RecipesBase", "StaticArraysCore", "Statistics", "Tables", "ZygoteRules"]

diff --git a/src/Operators/finitedifference.jl b/src/Operators/finitedifference.jl
@@ -3277,7 +3277,7 @@ function Base.similar(
     return Field(Eltype, sp)
 end
 
-function _serial_copyto!(
+@inline function _serial_copyto!(
     field_out::Field,
     bc::Base.Broadcast.Broadcasted{S},
     Ni::Int,
@@ -3290,7 +3290,7 @@ function _serial_copyto!(
     return field_out
 end
 
-function _threaded_copyto!(
+@inline function _threaded_copyto!(
     field_out::Field,
     bc::Base.Broadcast.Broadcasted{S},
     Ni::Int,
@@ -3307,7 +3307,7 @@ function _threaded_copyto!(
     return field_out
 end
 
-function Base.copyto!(
+@inline function Base.copyto!(
     field_out::Field,
     bc::Base.Broadcast.Broadcasted{S},
 ) where {S <: AbstractStencilStyle}
@@ -3320,7 +3320,7 @@ function Base.copyto!(
     return _serial_copyto!(field_out, bc, Ni, Nj, Nh)
 end
 
-function apply_stencil!(field_out, bc, hidx)
+Base.@propagate_inbounds function apply_stencil!(field_out, bc, hidx)
     space = axes(bc)
     if Topologies.isperiodic(Spaces.vertical_topology(space))
         @inbounds for idx in

diff --git a/test/Fields/field_opt.jl b/test/Fields/field_opt.jl
@@ -187,7 +187,11 @@ end
 
         allocs_test_Ref_with_compose!(S, ∂ᶠ𝕄ₜ∂ᶜρ, ∂ᶜρₜ∂ᶠ𝕄)
         p = @allocated allocs_test_Ref_with_compose!(S, ∂ᶠ𝕄ₜ∂ᶜρ, ∂ᶜρₜ∂ᶠ𝕄)
-        @test_broken p == 0
+        @test p == 0
+
+        allocs_test_Ref_with_compose_column!(S, ∂ᶠ𝕄ₜ∂ᶜρ, ∂ᶜρₜ∂ᶠ𝕄)
+        p = @allocated allocs_test_Ref_with_compose_column!(S, ∂ᶠ𝕄ₜ∂ᶜρ, ∂ᶜρₜ∂ᶠ𝕄)
+        @test p == 0
     end
 end
 nothing