Alternative discretizations #48

JordiManyer · 2025-07-03T18:36:27Z

Adds support and solvers for the H1-H1 and HDiv-HDiv discretizations of the MHD equations.

Requires:

Gridap#polytopal
GridapDistributed#polytopal
GridapSolvers#develop

codecov-commenter · 2025-07-09T10:36:09Z

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

❌ Patch coverage is 0% with 913 lines in your changes missing coverage. Please review.
✅ Project coverage is 0.00%. Comparing base (41193d6) to head (1370e08).
⚠️ Report is 108 commits behind head on master.

Files with missing lines	Patch %	Lines
src/weakforms.jl	0.00%	400 Missing ⚠️
src/Solvers/gmg.jl	0.00%	189 Missing ⚠️
src/fespaces.jl	0.00%	80 Missing ⚠️
src/parameters.jl	0.00%	67 Missing ⚠️
src/Applications/cavity.jl	0.00%	58 Missing ⚠️
src/gridap_extras.jl	0.00%	28 Missing ⚠️
src/Solvers/h1h1blocks.jl	0.00%	26 Missing ⚠️
src/geometry.jl	0.00%	23 Missing ⚠️
src/Applications/hunt.jl	0.00%	20 Missing ⚠️
src/Meshers/hunt_mesher.jl	0.00%	11 Missing ⚠️
... and 4 more
❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@           Coverage Diff            @@
##           master     #48     +/-   ##
========================================
  Coverage    0.00%   0.00%             
========================================
  Files          19      27      +8     
  Lines        2220    3396   +1176     
========================================
- Misses       2220    3396   +1176

Flag	Coverage Δ
unittests	`0.00% <0.00%> (ø)`

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

…elop

principejavier · 2025-07-30T15:11:02Z

I'm getting an error with GMG in Hunt problem. The command

mpirun -n 4 julia --project=. -e 'using GridapPETSc; using SparseMatricesCSR; using SparseArrays;  using GridapMHD:hunt;
  hunt(
    nc=(4,4),
    np=(2,2),
    backend=:mpi,
    L=1.0,
    B=(0.,50.,0.),
    debug=false,
    vtk=true,
    title="hunt-H1H1-gmg",
    solver= Dict(
       :solver => :h1h1blocks,
       :matrix_type    => SparseMatrixCSC{Float64,Int},
       :vector_type    => Vector{Float64},
       :block_solvers  => [:gmg,:julia,:julia],
       :petsc_options => "-ksp_error_if_not_converged true -ksp_converged_reason"
       ),
    fluid_disc = :Qk_dPkm1,
    current_disc = :H1,
    ranks_per_level = [(2,2,1),(2,2,1)],
    ζᵤ = 10.0
  )
'

gives this output

...
[7.805e+00 s in MAIN] fe_spaces
┌ Error: 
│   exception =
│    ArgumentError: Collection is empty, must contain exactly 1 element
│    Stacktrace:
│      [1] _only
│        @ ./iterators.jl:1550 [inlined]
│      [2] only
│        @ ./iterators.jl:1545 [inlined]
│      [3] #1
│        @ ~/.julia/packages/GridapSolvers/gqvqX/src/PatchBasedSmoothers/PatchTransferOperators.jl:25 [inlined]
│      [4] iterate
│        @ ./generator.jl:48 [inlined]
│      [5] _collect(c::Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, itr::Base.Generator{Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, GridapSolvers.PatchBasedSmoothers.var"#1#4"{Gridap.Arrays.Table{Int32, Vector{Int32}, Vector{Int32}}, GridapSolvers.PatchBasedSmoothers.var"#is_interior#3"{Vector{Int32}}}}, ::Base.EltypeUnknown, isz::Base.HasShape{1})
│        @ Base ./array.jl:800
│      [6] collect_similar(cont::Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, itr::Base.Generator{Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, GridapSolvers.PatchBasedSmoothers.var"#1#4"{Gridap.Arrays.Table{Int32, Vector{Int32}, Vector{Int32}}, GridapSolvers.PatchBasedSmoothers.var"#is_interior#3"{Vector{Int32}}}})
│        @ Base ./array.jl:709
│      [7] map(f::Function, A::Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}})
│        @ Base ./abstractarray.jl:3371
│      [8] CoarsePatchTopology(model::Gridap.Adaptivity.AdaptedDiscreteModel{3, 3, Gridap.Geometry.CartesianDiscreteModel{3, Float64, GridapMHD.Meshers.var"#map1#19"{Float64, Float64, Float64}}, Gridap.Geometry.CartesianDiscreteModel{3, Float64, GridapMHD.Meshers.var"#map1#19"{Float64, Float64, Float64}}, Gridap.Adaptivity.AdaptivityGlue{Gridap.Adaptivity.RefinementGlue, 3, Vector{Vector{Int64}}, Vector{Int64}, FillArrays.Fill{Gridap.Adaptivity.RefinementRule{Gridap.ReferenceFEs.ExtrusionPolytope{3}}, 1, Tuple{Base.OneTo{Int64}}}, Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, FillArrays.Fill{Bool, 1, Tuple{Base.OneTo{Int64}}}}}, coarse_ids::Vector{Int32})
│        @ GridapSolvers.PatchBasedSmoothers ~/.julia/packages/GridapSolvers/gqvqX/src/PatchBasedSmoothers/PatchTransferOperators.jl:23
│      [9] #6
│        @ ~/.julia/packages/GridapSolvers/gqvqX/src/PatchBasedSmoothers/PatchTransferOperators.jl:36 [inlined]
...

@JordiManyer am I doing something wrong? Full log:

Details

mpirun -n 4 julia --project=. -e 'using GridapPETSc; using SparseMatricesCSR; using SparseArrays;  using GridapMHD:hunt;
  hunt(
    nc=(4,4),
    np=(2,2),
    backend=:mpi,
    L=1.0,
    B=(0.,50.,0.),
    debug=false,
    vtk=true,
    title="hunt-H1H1-gmg",
    solver= Dict(
       :solver => :h1h1blocks,
       :matrix_type    => SparseMatrixCSC{Float64,Int},
       :vector_type    => Vector{Float64},
       :block_solvers  => [:gmg,:julia,:julia],
       :petsc_options => "-ksp_error_if_not_converged true -ksp_converged_reason"
       ),
    fluid_disc = :Qk_dPkm1,
    current_disc = :H1,
    ranks_per_level = [(2,2,1),(2,2,1)],
    ζᵤ = 10.0
  )
'
No protocol specified
┌ Info: 2-level CartesianModelHierarchy:
│   > Level 1: 8x8x3 CartesianDescriptor distributed in 2x2x1 ranks
└   > Level 2: 4x4x3 CartesianDescriptor distributed in 2x2x1 ranks
[8.786e+00 s in MAIN] pre_process
┌ Info: Parameters :: 
│   > x0 => zero 
│   > debug => false 
│   > multigrid => { 
│     > tests => { 
│     } 
│     > trials => { 
│     } 
│     > num_refs_coarse => 0 
│   } 
│   > jac_assemble => false 
│   > bcs => { 
│     > φ => { 
│       > values => 0.0 
│       > tags => conducting 
│     } 
│     > j => { 
│       > values => (0.0, 0.0, 0.0) 
│       > tags => insulating 
│     } 
│     > u => { 
│       > values => (0.0, 0.0, 0.0) 
│       > tags => noslip 
│     } 
│   } 
│   > solve => true 
│   > solid => nothing 
│   > fespaces => { 
│     > φ_conformity => H1 
│     > fluid_disc => Qk_dPkm1 
│     > φ_constraint => nothing 
│     > u_conformity => H1 
│     > p_constraint => nothing 
│     > order_u => 2 
│     > order_φ => 2 
│     > j_conformity => L2 
│     > k => 2 
│     > order_p => 1 
│     > rt_scaling => nothing 
│     > q => 5 
│     > current_disc => H1 
│     > p_conformity => L2 
│     > order_j => 2 
│     > formulation => H1H1 
│   } 
│   > solver => { 
│     > petsc_options => -ksp_error_if_not_converged true -ksp_converged_reason 
│     > rtol => 1.0e-10 
│     > atol => 1.0e-8 
│     > solver => h1h1blocks 
│     > niter => 20 
│     > niter_ls => 15 
│   } 
│   > fluid => { 
│     > α => 1.0 
│     > B => (0.0, 1.0, 0.0) 
│     > domain => nothing 
│     > μ => 10.0 
│     > σ => 1.0 
│     > convection => newton 
│     > ζⱼ => 0.0 
│     > g => (0.0, 0.0, 0.0) 
│     > β => 1.0 
│     > ζᵤ => 10.0 
│     > γ => 2500.0 
│     > f => (0.0, 0.0, 1.0) 
│     > divg => 0.0 
│   } 
│   > check_valid => true 
│   > continuation => nothing 
│   > transient => nothing 
└   > res_assemble => false 
[7.805e+00 s in MAIN] fe_spaces
┌ Error: 
│   exception =
│    ArgumentError: Collection is empty, must contain exactly 1 element
│    Stacktrace:
│      [1] _only
│        @ ./iterators.jl:1550 [inlined]
│      [2] only
│        @ ./iterators.jl:1545 [inlined]
│      [3] #1
│        @ ~/.julia/packages/GridapSolvers/gqvqX/src/PatchBasedSmoothers/PatchTransferOperators.jl:25 [inlined]
│      [4] iterate
│        @ ./generator.jl:48 [inlined]
│      [5] _collect(c::Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, itr::Base.Generator{Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, GridapSolvers.PatchBasedSmoothers.var"#1#4"{Gridap.Arrays.Table{Int32, Vector{Int32}, Vector{Int32}}, GridapSolvers.PatchBasedSmoothers.var"#is_interior#3"{Vector{Int32}}}}, ::Base.EltypeUnknown, isz::Base.HasShape{1})
│        @ Base ./array.jl:800
│      [6] collect_similar(cont::Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, itr::Base.Generator{Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, GridapSolvers.PatchBasedSmoothers.var"#1#4"{Gridap.Arrays.Table{Int32, Vector{Int32}, Vector{Int32}}, GridapSolvers.PatchBasedSmoothers.var"#is_interior#3"{Vector{Int32}}}})
│        @ Base ./array.jl:709
│      [7] map(f::Function, A::Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}})
│        @ Base ./abstractarray.jl:3371
│      [8] CoarsePatchTopology(model::Gridap.Adaptivity.AdaptedDiscreteModel{3, 3, Gridap.Geometry.CartesianDiscreteModel{3, Float64, GridapMHD.Meshers.var"#map1#19"{Float64, Float64, Float64}}, Gridap.Geometry.CartesianDiscreteModel{3, Float64, GridapMHD.Meshers.var"#map1#19"{Float64, Float64, Float64}}, Gridap.Adaptivity.AdaptivityGlue{Gridap.Adaptivity.RefinementGlue, 3, Vector{Vector{Int64}}, Vector{Int64}, FillArrays.Fill{Gridap.Adaptivity.RefinementRule{Gridap.ReferenceFEs.ExtrusionPolytope{3}}, 1, Tuple{Base.OneTo{Int64}}}, Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, FillArrays.Fill{Bool, 1, Tuple{Base.OneTo{Int64}}}}}, coarse_ids::Vector{Int32})
│        @ GridapSolvers.PatchBasedSmoothers ~/.julia/packages/GridapSolvers/gqvqX/src/PatchBasedSmoothers/PatchTransferOperators.jl:23
│      [9] #6
│        @ ~/.julia/packages/GridapSolvers/gqvqX/src/PatchBasedSmoothers/PatchTransferOperators.jl:36 [inlined]
│     [10] map(::GridapSolvers.PatchBasedSmoothers.var"#6#7", ::PartitionedArrays.MPIArray{Gridap.Adaptivity.AdaptedDiscreteModel{3, 3, Gridap.Geometry.CartesianDiscreteModel{3, Float64, GridapMHD.Meshers.var"#map1#19"{Float64, Float64, Float64}}, Gridap.Geometry.CartesianDiscreteModel{3, Float64, GridapMHD.Meshers.var"#map1#19"{Float64, Float64, Float64}}, Gridap.Adaptivity.AdaptivityGlue{Gridap.Adaptivity.RefinementGlue, 3, Vector{Vector{Int64}}, Vector{Int64}, FillArrays.Fill{Gridap.Adaptivity.RefinementRule{Gridap.ReferenceFEs.ExtrusionPolytope{3}}, 1, Tuple{Base.OneTo{Int64}}}, Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, FillArrays.Fill{Bool, 1, Tuple{Base.OneTo{Int64}}}}}, 1}, ::PartitionedArrays.MPIArray{PartitionedArrays.PermutedLocalIndices{PartitionedArrays.LocalIndicesWithConstantBlockSize{3}}, 1})
│        @ PartitionedArrays ~/.julia/packages/PartitionedArrays/py6uo/src/mpi_array.jl:229
│     [11] CoarsePatchTopology(model::GridapDistributed.GenericDistributedDiscreteModel{3, 3, PartitionedArrays.MPIArray{Gridap.Adaptivity.AdaptedDiscreteModel{3, 3, Gridap.Geometry.CartesianDiscreteModel{3, Float64, GridapMHD.Meshers.var"#map1#19"{Float64, Float64, Float64}}, Gridap.Geometry.CartesianDiscreteModel{3, Float64, GridapMHD.Meshers.var"#map1#19"{Float64, Float64, Float64}}, Gridap.Adaptivity.AdaptivityGlue{Gridap.Adaptivity.RefinementGlue, 3, Vector{Vector{Int64}}, Vector{Int64}, FillArrays.Fill{Gridap.Adaptivity.RefinementRule{Gridap.ReferenceFEs.ExtrusionPolytope{3}}, 1, Tuple{Base.OneTo{Int64}}}, Gridap.Arrays.Table{Int64, Vector{Int64}, Vector{Int64}}, FillArrays.Fill{Bool, 1, Tuple{Base.OneTo{Int64}}}}}, 1}, Vector{PartitionedArrays.PRange}, GridapDistributed.DistributedAdaptedDiscreteModelCache{GridapDistributed.DistributedCartesianDescriptor{PartitionedArrays.MPIArray{Int64, 1}, Tuple{Int64, Int64, Int64}, Gridap.Geometry.CartesianDescriptor{3, Float64, GridapMHD.Meshers.var"#map1#19"{Float64, Float64, Float64}}}, GridapDistributed.DistributedCartesianDescriptor{PartitionedArrays.MPIArray{Int64, 1}, Tuple{Int64, Int64, Int64}, Gridap.Geometry.CartesianDescriptor{3, Float64, GridapMHD.Meshers.var"#map1#19"{Float64, Float64, Float64}}}, PartitionedArrays.PRange{PartitionedArrays.MPIArray{PartitionedArrays.PermutedLocalIndices{PartitionedArrays.LocalIndicesWithConstantBlockSize{3}}, 1}}}})
│        @ GridapSolvers.PatchBasedSmoothers ~/.julia/packages/GridapSolvers/gqvqX/src/PatchBasedSmoothers/PatchTransferOperators.jl:35
│     [12] (::GridapMHD.var"#73#74"{GridapSolvers.MultilevelTools.FESpaceHierarchy{Vector{GridapSolvers.MultilevelTools.FESpaceHierarchyLevel{A, Nothing} where A}, Vector{Union{GridapDistributed.MPIVoidVector{Int64}, PartitionedArrays.MPIArray{Int64, 1}}}, GridapSolvers.MultilevelTools.FESpaceHierarchyLevel{A, Nothing} where A}, GridapMHD.var"#weakform#76"{Dict{Symbol, Any}}})(lev::Int64)
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/Solvers/gmg.jl:66
│     [13] (::GridapSolvers.MultilevelTools.var"#11#12"{GridapMHD.var"#73#74"{GridapSolvers.MultilevelTools.FESpaceHierarchy{Vector{GridapSolvers.MultilevelTools.FESpaceHierarchyLevel{A, Nothing} where A}, Vector{Union{GridapDistributed.MPIVoidVector{Int64}, PartitionedArrays.MPIArray{Int64, 1}}}, GridapSolvers.MultilevelTools.FESpaceHierarchyLevel{A, Nothing} where A}, GridapMHD.var"#weakform#76"{Dict{Symbol, Any}}}})(ranks::PartitionedArrays.MPIArray{Int64, 1}, ai::Int64)
│        @ GridapSolvers.MultilevelTools ~/.julia/packages/GridapSolvers/gqvqX/src/MultilevelTools/HierarchicalArrays.jl:109
│     [14] #4
│        @ ./generator.jl:37 [inlined]
│     [15] iterate
│        @ ./generator.jl:48 [inlined]
│     [16] collect(itr::Base.Generator{Base.Iterators.Zip{Tuple{SubArray{Union{GridapDistributed.MPIVoidVector{Int64}, PartitionedArrays.MPIArray{Int64, 1}}, 1, Vector{Union{GridapDistributed.MPIVoidVector{Int64}, PartitionedArrays.MPIArray{Int64, 1}}}, Tuple{UnitRange{Int64}}, true}, SubArray{Int64, 1, LinearIndices{1, Tuple{Base.OneTo{Int64}}}, Tuple{UnitRange{Int64}}, true}}}, Base.var"#4#5"{GridapSolvers.MultilevelTools.var"#11#12"{GridapMHD.var"#73#74"{GridapSolvers.MultilevelTools.FESpaceHierarchy{Vector{GridapSolvers.MultilevelTools.FESpaceHierarchyLevel{A, Nothing} where A}, Vector{Union{GridapDistributed.MPIVoidVector{Int64}, PartitionedArrays.MPIArray{Int64, 1}}}, GridapSolvers.MultilevelTools.FESpaceHierarchyLevel{A, Nothing} where A}, GridapMHD.var"#weakform#76"{Dict{Symbol, Any}}}}}})
│        @ Base ./array.jl:780
│     [17] map
│        @ ./abstractarray.jl:3495 [inlined]
│     [18] map(f::GridapMHD.var"#73#74"{GridapSolvers.MultilevelTools.FESpaceHierarchy{Vector{GridapSolvers.MultilevelTools.FESpaceHierarchyLevel{A, Nothing} where A}, Vector{Union{GridapDistributed.MPIVoidVector{Int64}, PartitionedArrays.MPIArray{Int64, 1}}}, GridapSolvers.MultilevelTools.FESpaceHierarchyLevel{A, Nothing} where A}, GridapMHD.var"#weakform#76"{Dict{Symbol, Any}}}, a::GridapSolvers.MultilevelTools.HierarchicalArray{Int64, SubArray{Int64, 1, LinearIndices{1, Tuple{Base.OneTo{Int64}}}, Tuple{UnitRange{Int64}}, true}, SubArray{Union{GridapDistributed.MPIVoidVector{Int64}, PartitionedArrays.MPIArray{Int64, 1}}, 1, Vector{Union{GridapDistributed.MPIVoidVector{Int64}, PartitionedArrays.MPIArray{Int64, 1}}}, Tuple{UnitRange{Int64}}, true}})
│        @ GridapSolvers.MultilevelTools ~/.julia/packages/GridapSolvers/gqvqX/src/MultilevelTools/HierarchicalArrays.jl:107
│     [19] gmg_patch_prolongations(tests::GridapSolvers.MultilevelTools.FESpaceHierarchy{Vector{GridapSolvers.MultilevelTools.FESpaceHierarchyLevel{A, Nothing} where A}, Vector{Union{GridapDistributed.MPIVoidVector{Int64}, PartitionedArrays.MPIArray{Int64, 1}}}, GridapSolvers.MultilevelTools.FESpaceHierarchyLevel{A, Nothing} where A}, weakform::GridapMHD.var"#weakform#76"{Dict{Symbol, Any}})
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/Solvers/gmg.jl:62
│     [20] gmg_solver(::Val{:H1H1}, ::Val{:h1h1blocks}, params::Dict{Symbol, Any})
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/Solvers/gmg.jl:94
│     [21] get_block_solver(::Val{:gmg}, params::Dict{Symbol, Any})
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/Solvers/gmg.jl:5
│     [22] (::GridapMHD.var"#97#99"{Dict{Symbol, Any}})(s::Symbol)
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/Solvers/h1h1blocks.jl:15
│     [23] iterate
│        @ ./generator.jl:48 [inlined]
│     [24] _collect(c::Vector{Symbol}, itr::Base.Generator{Vector{Symbol}, GridapMHD.var"#97#99"{Dict{Symbol, Any}}}, ::Base.EltypeUnknown, isz::Base.HasShape{1})
│        @ Base ./array.jl:800
│     [25] collect_similar(cont::Vector{Symbol}, itr::Base.Generator{Vector{Symbol}, GridapMHD.var"#97#99"{Dict{Symbol, Any}}})
│        @ Base ./array.jl:709
│     [26] map(f::Function, A::Vector{Symbol})
│        @ Base ./abstractarray.jl:3371
│     [27] H1H1BlockSolver(op::Gridap.FESpaces.FEOperatorFromWeakForm, params::Dict{Symbol, Any})
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/Solvers/h1h1blocks.jl:15
│     [28] _solver(::Val{:h1h1blocks}, op::Gridap.FESpaces.FEOperatorFromWeakForm, params::Dict{Symbol, Any})
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/main.jl:190
│     [29] _solver(op::Gridap.FESpaces.FEOperatorFromWeakForm, params::Dict{Symbol, Any})
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/main.jl:174
│     [30] main(_params::Dict{Symbol, Any}; output::Dict{Symbol, Any})
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/main.jl:147
│     [31] (::GridapMHD.var"#111#112"{Dict{Symbol, Any}})()
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/Applications/hunt.jl:201
│     [32] with(f::GridapMHD.var"#111#112"{Dict{Symbol, Any}}; kwargs::@Kwargs{args::Vector{SubString{String}}})
│        @ GridapPETSc ~/.julia/packages/GridapPETSc/iYLxh/src/Environment.jl:38
│     [33] _hunt(; distribute::PartitionedArrays.var"#88#89"{MPI.Comm, Bool}, rank_partition::Tuple{Int64, Int64}, nc::Tuple{Int64, Int64}, ν::Float64, ρ::Float64, σ::Float64, B::Tuple{Float64, Float64, Float64}, f::Tuple{Float64, Float64, Float64}, ζᵤ::Float64, ζⱼ::Float64, μ::Int64, L::Float64, u0::Float64, B0::Float64, σw1::Float64, σw2::Float64, tw::Float64, order::Int64, order_j::Int64, nsums::Int64, vtk::Bool, title::String, path::String, debug::Bool, res_assemble::Bool, jac_assemble::Bool, solve::Bool, solver::Dict{Symbol, Any}, formulation::Symbol, initial_value::Symbol, rt_scaling::Bool, verbose::Bool, BL_adapted::Bool, kmap_x::Int64, kmap_y::Int64, ranks_per_level::Vector{Tuple{Int64, Int64, Int64}}, adaptivity_method::Int64, fluid_disc::Symbol, current_disc::Symbol)
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/Applications/hunt.jl:200
│     [34] _hunt
│        @ ~/Prog/GridapMHD.jl/src/Applications/hunt.jl:41 [inlined]
│     [35] #105
│        @ ~/Prog/GridapMHD.jl/src/Applications/hunt.jl:23 [inlined]
│     [36] with_mpi(f::GridapMHD.var"#105#108"{Tuple{Int64, Int64}, String, @Kwargs{nc::Tuple{Int64, Int64}, L::Float64, B::Tuple{Float64, Float64, Float64}, debug::Bool, vtk::Bool, solver::Dict{Symbol, Any}, fluid_disc::Symbol, current_disc::Symbol, ranks_per_level::Vector{Tuple{Int64, Int64, Int64}}, ζᵤ::Float64}, String}; comm::MPI.Comm, duplicate_comm::Bool)
│        @ PartitionedArrays ~/.julia/packages/PartitionedArrays/py6uo/src/mpi_array.jl:73
│     [37] with_mpi
│        @ ~/.julia/packages/PartitionedArrays/py6uo/src/mpi_array.jl:64 [inlined]
│     [38] hunt(; backend::Symbol, np::Tuple{Int64, Int64}, title::String, nruns::Int64, path::String, kwargs::@Kwargs{nc::Tuple{Int64, Int64}, L::Float64, B::Tuple{Float64, Float64, Float64}, debug::Bool, vtk::Bool, solver::Dict{Symbol, Any}, fluid_disc::Symbol, current_disc::Symbol, ranks_per_level::Vector{Tuple{Int64, Int64, Int64}}, ζᵤ::Float64})
│        @ GridapMHD ~/Prog/GridapMHD.jl/src/Applications/hunt.jl:22
│     [39] top-level scope
│        @ none:2
│     [40] eval
│        @ ./boot.jl:430 [inlined]
│     [41] exec_options(opts::Base.JLOptions)
│        @ Base ./client.jl:296
│     [42] _start()
│        @ Base ./client.jl:531
└ @ PartitionedArrays ~/.julia/packages/PartitionedArrays/py6uo/src/mpi_array.jl:75
--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD
with errorcode 1.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------

JordiManyer · 2025-07-30T22:22:43Z

Can you update all branches and try again?

principejavier · 2025-07-30T22:37:30Z

Already done, cavity works. Could periodic mesh be an issue?

BTW, cavity only works with PETSc 3.23.4 but not with 3.15.4, seems to be a requirement of GridapSolvers, I remember you mentioned it. Could it be interesting to add a message elsewhere? Let me know whether you want me to open an issue about this, I could still reproduce the error I get with 3.15.4.

JordiManyer · 2025-07-30T23:01:00Z

Already done, cavity works. Could periodic mesh be an issue?

Ok yeah, the periodic meshes thing makes sense. I'll have a look.

BTW, cavity only works with PETSc 3.23.4 but not with 3.15.4, seems to be a requirement of GridapSolvers, I remember you mentioned it. Could it be interesting to add a message elsewhere? Let me know whether you want me to open an issue about this, I could still reproduce the error I get with 3.15.4.

Has nothing to do with GridapSolvers, it has to do with GridapPETSc. Basically the old version only worked with old versions of PETSc. I did changes to make it work with newer versions of PETSc, but it made old versions break. I have a new PR open in GridapPETSc, with some nice changes to how memory is handled. Can you try to use that to see if it makes a difference? Otherwise I can just add a comment in GridapPETSc's documentation, I don't think it's a big deal since one can simply update PETSc.

JordiManyer · 2025-07-30T23:28:49Z

This works for me:

using Gridap
using Gridap.Geometry, Gridap.Arrays

using GridapSolvers
using GridapSolvers: PatchBasedSmoothers  
using PartitionedArrays, GridapDistributed

parts = (2,2,1)
ranks = with_debug() do distribute
  distribute(LinearIndices((prod(parts),)))
end

cmodel = CartesianDiscreteModel((0,1,0,1),(5,5);isperiodic=(true,true))
cmodel = CartesianDiscreteModel(ranks,parts,(0,1,0,1,0,1),(4,4,3);isperiodic=(false,false,true))
model = Gridap.Adaptivity.refine(cmodel)

glue = Gridap.Adaptivity.get_adaptivity_glue(model)
ptopo = PatchBasedSmoothers.CoarsePatchTopology(model);

so I think you did not update all your branches like I suggested. Can you do go into your environment and update all your packages?

principejavier · 2025-07-31T16:59:43Z

Confirmed. Hunt with isotropic refinement (0a1dad9) runs.

…elop

…ding mg tests

…precompilation

…elop

JordiManyer and others added 12 commits July 3, 2025 15:44

Started trying to optimise integration

bd878da

Reactivated transient

71cd10f

Added H1-H1 and HDiv-HDiv discretizations

52e6017

FESpaces and solver for H1H1 formulation

31b08af

Minor

46109a4

bugfixes

167a079

H1-H1 block solver working

1226c77

Minor changes and fixes to params

6dedece

Fixed Hunt tests

bedf25d

Fixed all tests

7cc08d1

Fixed MPI tests

cec5e38

Yet another fix

3e086ef

JordiManyer and others added 12 commits July 10, 2025 11:51

First working GMG version for H1H1

590f7da

Minor

42efda6

Started implementing Hdiv-HDiv solver

26ec7fc

More progress

c7c0297

Added solid tags to cavity

5dd9448

Fluid + Solid working in Cavity

6e5b80d

Minor

30ee133

More changes to make hdiv-hdiv work

40a4946

Minor

a457a78

Minor

ca44119

Minor fix in Hunt output (scaling)

9ec2a5e

Merge branch 'develop' of github.com:gridapapps/GridapMHD.jl into dev…

624bdda

…elop

JordiManyer and others added 29 commits August 4, 2025 15:59

Minor bugfix in HDiv velocity terms

dedc537

Merge branch 'develop' of github.com:gridapapps/GridapMHD.jl into dev…

96f6abc

…elop

Minor changes and fixes

bfee25a

Added block jacobi smoothers

5cff314

Merge branch 'develop' of github.com:gridapapps/GridapMHD.jl into dev…

62e660a

…elop

Changed Julia version in CI to 1.11

b730ce6

Minor fix to mg params

e6d4f4f

Changed default maxiter in FGMRES of H1H1 precond

c9c953d

Modified (temporarily) MPI CI workflow to add branches that permit ad…

dd1391f

…ding mg tests

Added mpi tests

5c6d68a

Fixed MPI CI workflow

cfccedf

Minor fixes to Hunt

3e4ceb9

Added Hunt gmg tests

5495853

Minor fix to cavity

9bd48fc

Added a log function to save solver history in output dict

ce80a52

Changed default max it in solvers

0928f7d

Added a test for gmg parameter dependence

9f99ed7

Minor fix to gmg_par_dep

51b34e6

yet another minor fix gmg_par_dep

13b8a04

gmg_par_dep without mass

b8d4424

gmg_par_dep wip

b6c1e82

gmg_par_dep including vxB.uxB

714e8c8

gmg_par_der with a nonlinear solver

4a46f30

Implemented block precond in gmg_par_dep

aad3457

gmg_par_dep with the same problem of GridapMHD

e207221

Layer adapted mesh and several options in gmg_par_dep.jl

75f3fc1

Changed phi order in H1 formulation

596770a

Compat restricted gmsh_jll whose latest build triggers errors during …

0b5a2f1

…precompilation

Merge branch 'develop' of github.com:gridapapps/GridapMHD.jl into dev…

1370e08

…elop

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Alternative discretizations #48

Alternative discretizations #48

Uh oh!

JordiManyer commented Jul 3, 2025 •

edited

Loading

Uh oh!

codecov-commenter commented Jul 9, 2025 •

edited

Loading

Uh oh!

principejavier commented Jul 30, 2025

Uh oh!

JordiManyer commented Jul 30, 2025

Uh oh!

principejavier commented Jul 30, 2025 •

edited

Loading

Uh oh!

JordiManyer commented Jul 30, 2025

Uh oh!

JordiManyer commented Jul 30, 2025

Uh oh!

principejavier commented Jul 31, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Alternative discretizations #48

Are you sure you want to change the base?

Alternative discretizations #48

Uh oh!

Conversation

JordiManyer commented Jul 3, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codecov-commenter commented Jul 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

principejavier commented Jul 30, 2025

Uh oh!

JordiManyer commented Jul 30, 2025

Uh oh!

principejavier commented Jul 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

JordiManyer commented Jul 30, 2025

Uh oh!

JordiManyer commented Jul 30, 2025

Uh oh!

principejavier commented Jul 31, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

JordiManyer commented Jul 3, 2025 •

edited

Loading

codecov-commenter commented Jul 9, 2025 •

edited

Loading

principejavier commented Jul 30, 2025 •

edited

Loading