Class Dune::Amg::AMGCPR< M, X, S, PI, A >: drop the smoother template parameter and replace with dynamic construction
Class Opm::AbstractISTLSolver< TypeTag >: We should remove the use of setResidual() and setMatrix() methods and instead use prepare() with the SparseMatrixAdapter and Vector directly. This would simplify the interface and reduce the number of required method calls before solving the system.
Class Opm::DiscreteFractureModel< TypeTag >: So far, the discrete fracture model only works for 2D grids and without energy. Also only the Darcy velocity model is supported for the fractures.
Member Opm::gpuistl::detail::cublasSafeCall (cublasStatus_t error, const std::string_view &expression, const std::string_view &filename, const std::string_view &functionName, size_t lineNumber): Refactor to use std::source_location once we shift to C++20
Member Opm::gpuistl::detail::cublasWarnIfError (cublasStatus_t error, const std::string_view &expression, const std::string_view &filename, const std::string_view &functionName, size_t lineNumber): Refactor to use std::source_location once we shift to C++20
Member Opm::gpuistl::detail::cudaSafeCall (cudaError_t error, const std::string_view &expression, const std::string_view &filename, const std::string_view &functionName, size_t lineNumber): Refactor to use std::source_location once we shift to C++20
Member Opm::gpuistl::detail::cudaWarnIfError (cudaError_t error, const std::string_view &expression, const std::string_view &filename, const std::string_view &functionName, size_t lineNumber): Refactor to use std::source_location once we shift to C++20
Member Opm::gpuistl::detail::cusparseSafeCall (cusparseStatus_t error, const std::string_view &expression, const std::string_view &filename, const std::string_view &functionName, size_t lineNumber): Refactor to use std::source_location once we shift to C++20
Member Opm::gpuistl::detail::cusparseWarnIfError (cusparseStatus_t error, const std::string_view &expression, const std::string_view &filename, const std::string_view &functionName, size_t lineNumber): Refactor to use std::source_location once we shift to C++20
Member Opm::gpuistl::detail::getCublasErrorMessage (cublasStatus_t error, const std::string_view &expression, const std::string_view &filename, const std::string_view &functionName, size_t lineNumber): Refactor to use std::source_location once we shift to C++20
Member Opm::gpuistl::detail::getCudaErrorMessage (cudaError_t error, const std::string_view &expression, const std::string_view &filename, const std::string_view &functionName, size_t lineNumber): Refactor to use std::source_location once we shift to C++20
Member Opm::gpuistl::detail::getCusparseErrorMessage (cusparseStatus_t error, const std::string_view &expression, const std::string_view &filename, const std::string_view &functionName, size_t lineNumber): Refactor to use std::source_location once we shift to C++20
Member Opm::gpuistl::detail::to_int (std::size_t s): This can be done for more generic types, but then it is probably wise to wait for C++20's cmp-functions
Member Opm::gpuistl::detail::to_size_t (int i): This can be done for more generic types, but then it is probably wise to wait for C++20's cmp-functions
Class Opm::gpuistl::GpuDILU< CPUMatrixT, X, Y, l >: Remove the reliance on CPUMatrix. We should be able to use the GPU matrix type directly.
Class Opm::gpuistl::OpmGpuILU0< CPUMatrixT, X, Y, l >: Remove the reliance on CPUMatrix. We should be able to use the GPU matrix type directly.
Member Opm::gpuistl::SolverAdapter< Operator, UnderlyingSolver, X >::SolverAdapter (Operator &op, Dune::ScalarProduct< X > &sp, std::shared_ptr< Dune::Preconditioner< X, X > > prec, scalar_real_type reduction, int maxit, int verbose, const Comm &comm): Use a std::forward in this function
Class Opm::Linear::VertexBorderListFromGrid< GridView, VertexMapper >: implement this class generically. For this, it must be possible to query the mapper whether it contains entities of a given codimension without the need to hand it an actual entity.
Member Opm::ParallelWellInfo< Scalar >::sumPerfValues (It begin, It end) const: cater for overlap later. Currently only owner
Member Opm::StandardWellPrimaryVariables< FluidSystem, Indices >::Bhp: : We should have indices for the well equations and well primary variables separately.