ERROR: a InexactError was thrown during kernel execution

[Franc-Z]

ERROR: a InexactError was thrown during kernel execution on thread (658, 1, 1) in block (2, 1, 1).
Inexact conversion
Stacktrace not available, run Julia on debug level 2 for more details (by passing -g2 to the executable).

ERROR: CUDA.KernelException(CUDA.CuDevice(0))
Stacktrace:
[1] check_exceptions()
@ CUDA ~/.julia/packages/CUDA/L1qZp/src/compiler/exceptions.jl:39
[2] device_synchronize(; blocking::Bool, spin::Bool)
@ CUDA ~/.julia/packages/CUDA/L1qZp/lib/cudadrv/synchronization.jl:191
[3] device_synchronize
@ ~/.julia/packages/CUDA/L1qZp/lib/cudadrv/synchronization.jl:178 [inlined]
[4] checked_cuModuleLoadDataEx(_module::Base.RefValue{…}, image::Ptr{…}, numOptions::Int64, options::Vector{…}, optionValues::Vector{…})
@ CUDA ~/.julia/packages/CUDA/L1qZp/lib/cudadrv/module.jl:18
[5] CUDA.CuModule(data::Vector{UInt8}, options::Dict{CUDA.CUjit_option_enum, Any})
@ CUDA ~/.julia/packages/CUDA/L1qZp/lib/cudadrv/module.jl:60
[6] CuModule
@ ~/.julia/packages/CUDA/L1qZp/lib/cudadrv/module.jl:49 [inlined]
[7] link(job::GPUCompiler.CompilerJob, compiled::@NamedTuple{image::Vector{UInt8}, entry::String})
@ CUDA ~/.julia/packages/CUDA/L1qZp/src/compiler/compilation.jl:409
[8] actual_compilation(cache::Dict{…}, src::Core.MethodInstance, world::UInt64, cfg::GPUCompiler.CompilerConfig{…}, compiler::typeof(CUDA.compile), linker::typeof(CUDA.link))
@ GPUCompiler ~/.julia/packages/GPUCompiler/Nxf8r/src/execution.jl:262
[9] cached_compilation(cache::Dict{…}, src::Core.MethodInstance, cfg::GPUCompiler.CompilerConfig{…}, compiler::Function, linker::Function)
@ GPUCompiler ~/.julia/packages/GPUCompiler/Nxf8r/src/execution.jl:151
[10] macro expansion
@ ~/.julia/packages/CUDA/L1qZp/src/compiler/execution.jl:380 [inlined]
[11] macro expansion
@ ./lock.jl:273 [inlined]
[12] cufunction(f::typeof(MadNLPGPU.gpu__force_lower_triangular_kernel!), tt::Type{…}; kwargs::@kwargs{…})
@ CUDA ~/.julia/packages/CUDA/L1qZp/src/compiler/execution.jl:375
[13] macro expansion
@ ~/.julia/packages/CUDA/L1qZp/src/compiler/execution.jl:112 [inlined]
[14] (::KernelAbstractions.Kernel{…})(::CUDA.CuArray{…}, ::Vararg{…}; ndrange::Int64, workgroupsize::Nothing)
@ CUDA.CUDAKernels ~/.julia/packages/CUDA/L1qZp/src/CUDAKernels.jl:103
[15] Kernel
@ ~/.julia/packages/CUDA/L1qZp/src/CUDAKernels.jl:89 [inlined]
[16] force_lower_triangular!
@ ~/.julia/packages/MadNLPGPU/v1OVP/src/KKT/sparse.jl:271 [inlined]
[17] create_kkt_system(::Type{…}, cb::MadNLP.SparseCallback{…}, ind_cons::@NamedTuple{…}, linear_solver::Type{…}; opt_linear_solver::MadNLPGPU.CudssSolverOptions, hessian_approximation::Type)
@ MadNLP ~/.julia/packages/MadNLP/6x6Eg/src/KKT/Sparse/condensed.jl:80
[18] MadNLP.MadNLPSolver(nlp::PortfolioNLPModelCUDA{…}; kwargs::@kwargs{…})
@ MadNLP ~/.julia/packages/MadNLP/6x6Eg/src/IPM/IPM.jl:155
[19] MadNLPSolver
@ ~/.julia/packages/MadNLP/6x6Eg/src/IPM/IPM.jl:115 [inlined]
[20] run_mean_risk_optimization(lambda_risk::Float64, w_max::Float64, w_min::Float64)
@ Main /nvtest/GPU-Accelerated_Portfolio_Optimization/Mean-Risk-Barra-Sparse/mean_std_gpu_barra.jl:511
[21] run_mean_risk_optimization()
@ Main /nvtest/GPU-Accelerated_Portfolio_Optimization/Mean-Risk-Barra-Sparse/mean_std_gpu_barra.jl:383
[22] top-level scope
@ /nvtest/GPU-Accelerated_Portfolio_Optimization/Mean-Risk-Barra-Sparse/mean_std_gpu_barra.jl:580
Some type information was truncated. Use show(err) to see complete types.

[sshin23]

Could you send the code? Thank you!