GPU compilation of #25(CuArrays.CuKernelState, CUDAnative.CuDeviceArray{ForwardDiff.Dual{Nothing,Float32,2},2,CUDAnative.AS.Global}, Base.Broadcast.Broadcasted{Nothing,Tuple{Base.OneTo{Int64},Base.OneTo{Int64}},Zygote.var"#1596#1599"{typeof(-)},Tuple{Base.Broadcast.Extruded{CUDAnative.CuDeviceArray{Float32,2,CUDAnative.AS.Global},Tuple{Bool,Bool},Tuple{Int64,Int64}},Base.Broadcast.Extruded{DiffEqArray{Float32,2,Array{CuArrays.CuArray{Float32,1,Nothing},1},StepRangeLen{Float32,Float64,Float64}},Tuple{Bool,Bool},Tuple{Int64,Int64}}}}) failed
KernelError: recursion is currently not supported

Try inspecting the generated code with any of the @device_code_... macros.

Stacktrace:
 [1] _nextind_str at strings/string.jl:140
 [2] nextind at strings/string.jl:136
 [3] _nextind_str at strings/string.jl:140
 [4] _split at strings/util.jl:325
 [5] env_override_minlevel at logging.jl:419
 [6] current_logger_for_env at logging.jl:383
 [7] assertscalar at /root/.julia/packages/GPUArrays/1wgPO/src/indexing.jl:13
 [8] #25 at /root/.julia/packages/GPUArrays/1wgPO/src/broadcast.jl:49
(::CUDAnative.var"#hook_emit_function#94"{CUDAnative.CompilerJob,Array{Core.MethodInstance,1}})(::Core.MethodInstance, ::Core.CodeInfo, ::UInt64) at irgen.jl:102
compile_method_instance(::CUDAnative.CompilerJob, ::Core.MethodInstance, ::UInt64) at irgen.jl:153
macro expansion at TimerOutput.jl:228 [inlined]
irgen(::CUDAnative.CompilerJob, ::Core.MethodInstance, ::UInt64) at irgen.jl:167
macro expansion at TimerOutput.jl:228 [inlined]
macro expansion at driver.jl:99 [inlined]
macro expansion at TimerOutput.jl:228 [inlined]
#codegen#151(::Bool, ::Bool, ::Bool, ::Bool, ::Bool, ::typeof(CUDAnative.codegen), ::Symbol, ::CUDAnative.CompilerJob) at driver.jl:98
#codegen at none:0 [inlined]
#compile#150(::Bool, ::Bool, ::Bool, ::Bool, ::Bool, ::typeof(CUDAnative.compile), ::Symbol, ::CUDAnative.CompilerJob) at driver.jl:47
#compile#149 at none:0 [inlined]
#compile at none:0 [inlined]
#compile at none:0 [inlined]
macro expansion at execution.jl:393 [inlined]
#cufunction#195(::Nothing, ::Base.Iterators.Pairs{Union{},Union{},Tuple{},NamedTuple{,Tuple{}}}, ::typeof(CUDAnative.cufunction), ::GPUArrays.var"#25#26", ::Type{Tuple{CuArrays.CuKernelState,CUDAnative.CuDeviceArray{ForwardDiff.Dual{Nothing,Float32,2},2,CUDAnative.AS.Global},Base.Broadcast.Broadcasted{Nothing,Tuple{Base.OneTo{Int64},Base.OneTo{Int64}},Zygote.var"#1596#1599"{typeof(-)},Tuple{Base.Broadcast.Extruded{CUDAnative.CuDeviceArray{Float32,2,CUDAnative.AS.Global},Tuple{Bool,Bool},Tuple{Int64,Int64}},Base.Broadcast.Extruded{DiffEqArray{Float32,2,Array{CuArrays.CuArray{Float32,1,Nothing},1},StepRangeLen{Float32,Float64,Float64}},Tuple{Bool,Bool},Tuple{Int64,Int64}}}}}}) at execution.jl:360
cufunction(::Function, ::Type) at execution.jl:360
macro expansion at execution.jl:179 [inlined]
macro expansion at gcutils.jl:91 [inlined]
macro expansion at execution.jl:176 [inlined]
_gpu_call(::CuArrays.CuArrayBackend, ::Function, ::CuArrays.CuArray{ForwardDiff.Dual{Nothing,Float32,2},2,Nothing}, ::Tuple{CuArrays.CuArray{ForwardDiff.Dual{Nothing,Float32,2},2,Nothing},Base.Broadcast.Broadcasted{Nothing,Tuple{Base.OneTo{Int64},Base.OneTo{Int64}},Zygote.var"#1596#1599"{typeof(-)},Tuple{Base.Broadcast.Extruded{CuArrays.CuArray{Float32,2,Nothing},Tuple{Bool,Bool},Tuple{Int64,Int64}},Base.Broadcast.Extruded{DiffEqArray{Float32,2,Array{CuArrays.CuArray{Float32,1,Nothing},1},StepRangeLen{Float32,Float64,Float64}},Tuple{Bool,Bool},Tuple{Int64,Int64}}}}}, ::Tuple{Tuple{Int64},Tuple{Int64}}) at gpuarray_interface.jl:62
gpu_call(::Function, ::CuArrays.CuArray{ForwardDiff.Dual{Nothing,Float32,2},2,Nothing}, ::Tuple{CuArrays.CuArray{ForwardDiff.Dual{Nothing,Float32,2},2,Nothing},Base.Broadcast.Broadcasted{Nothing,Tuple{Base.OneTo{Int64},Base.OneTo{Int64}},Zygote.var"#1596#1599"{typeof(-)},Tuple{Base.Broadcast.Extruded{CuArrays.CuArray{Float32,2,Nothing},Tuple{Bool,Bool},Tuple{Int64,Int64}},Base.Broadcast.Extruded{DiffEqArray{Float32,2,Array{CuArrays.CuArray{Float32,1,Nothing},1},StepRangeLen{Float32,Float64,Float64}},Tuple{Bool,Bool},Tuple{Int64,Int64}}}}}, ::Int64) at abstract_gpu_interface.jl:151
gpu_call at abstract_gpu_interface.jl:128 [inlined]
copyto! at broadcast.jl:48 [inlined]
copyto! at broadcast.jl:863 [inlined]
copy(::Base.Broadcast.Broadcasted{Base.Broadcast.ArrayStyle{CuArrays.CuArray},Tuple{Base.OneTo{Int64},Base.OneTo{Int64}},Zygote.var"#1596#1599"{typeof(-)},Tuple{CuArrays.CuArray{Float32,2,Nothing},DiffEqArray{Float32,2,Array{CuArrays.CuArray{Float32,1,Nothing},1},StepRangeLen{Float32,Float64,Float64}}}}) at broadcast.jl:839
materialize at broadcast.jl:819 [inlined]
broadcast_forward at broadcast.jl:168 [inlined]
adjoint at broadcast.jl:178 [inlined]
_pullback at adjoint.jl:47 [inlined]
adjoint at lib.jl:153 [inlined]
_pullback at adjoint.jl:47 [inlined]
broadcasted at broadcast.jl:1237 [inlined]
_pullback(::Zygote.Context, ::typeof(Base.Broadcast.broadcasted), ::typeof(-), ::CuArrays.CuArray{Float32,...