Callback TerminateSteadyState() does not work with AutoVern7/9 and Rodas4/Rodas5P

[jo-ap]

Describe the bug 🐞

Using TerminateSteadyState() as a callback to solve for an ODE problem with AutoVern7 or AutoVern9 and Rodas4 or Rodas5P (which is suggested by https://docs.sciml.ai/DiffEqDocs/stable/solvers/ode_solve/) fails.
I traced the problem to allDerivPass, which calls DiffEqBase.get_du!. There, out is updated to integrator.fsallast, which is not set. Therefore, get_du! returns nothing for the RHS and this results in errors further down.

Expected behavior

The callback TerminateSteadyState() should work with AutoVern7 or AutoVern9 together with Rodas4 or Rodas5P. I guess this boils down to DiffEqBase.get_du! returning the RHS for the integrator correctly.
If the error is to be expected, this should be documented and a warning thrown when calling solve.

Minimal Reproducible Example 👇

using DifferentialEquations

# Rosenzweig-MacArthur Predator-Prey Model
function f!(du, u, p, t)
  x, y = u 
  α, β, δ, γ, η, ρ = p
  du[1] = ρ*x*(1-x) - α*(η + β)*x*y/(δ + γ*x)
  du[2] = -η*y + γ*(η + β)*x*y/(δ + γ*x)
end

u0 = [0.1,0.1]
p = [0.1, 0.2, 1.5, 0.4, 0.2, 0.3]

prob = ODEProblem(f!, u0, (0.0, 1000.0), p)

# error comes from integrator.fsallast not being set when calling DiffEqBase.get_du! in allDerivPass
# occurs for all combinations of AutoVern7/AutoVern9 and Rodas4/Rodas5P
solve(prob, AutoVern9(Rodas4()); callback = TerminateSteadyState())

Error & Stacktrace ⚠️

julia> solve(prob, AutoVern9(Rodas4()); callback = TerminateSteadyState())
ERROR: MethodError: Cannot `convert` an object of type Nothing to an object of type Float64
The function `convert` exists, but no method is defined for this combination of argument types.

Closest candidates are:
  convert(::Type{T}, ::VectorizationBase.AbstractSIMD) where T<:Union{Bool, Float16, Float32, Float64, Int16, Int32, Int64, Int8, UInt16, UInt32, UInt64, UInt8, SIMDTypes.Bit}
   @ VectorizationBase ~/.julia/packages/VectorizationBase/LqJbS/src/base_defs.jl:201
  convert(::Type{<:Real}, ::T) where T<:SparseConnectivityTracer.AbstractTracer
   @ SparseConnectivityTracer ~/.julia/packages/SparseConnectivityTracer/hOFV1/src/overloads/conversion.jl:10
  convert(::Type{T}, ::T) where T
   @ Base Base.jl:126
  ...

Stacktrace:
  [1] fill!(dest::Vector{Float64}, x::Nothing)
    @ Base ./array.jl:327
  [2] copyto!
    @ ./broadcast.jl:928 [inlined]
  [3] materialize!
    @ ./broadcast.jl:878 [inlined]
  [4] materialize!
    @ ./broadcast.jl:875 [inlined]
  [5] get_du!
    @ ~/.julia/packages/OrdinaryDiffEqCore/YWZVL/src/integrators/integrator_interface.jl:82 [inlined]
  [6] allDerivPass(integrator::OrdinaryDiffEqCore.ODEIntegrator{…}, abstol::Float64, reltol::Float64, min_t::Nothing)
    @ DiffEqCallbacks ~/.julia/packages/DiffEqCallbacks/n5zrr/src/terminatesteadystate.jl:7
  [7] (::DiffEqCallbacks.var"#104#106"{…})(u::Vector{…}, t::Float64, integrator::OrdinaryDiffEqCore.ODEIntegrator{…})
    @ DiffEqCallbacks ~/.julia/packages/DiffEqCallbacks/n5zrr/src/terminatesteadystate.jl:68
  [8] apply_discrete_callback!
    @ ~/.julia/packages/DiffEqBase/ODi5x/src/callbacks.jl:611 [inlined]
  [9] handle_callbacks!
    @ ~/.julia/packages/OrdinaryDiffEqCore/YWZVL/src/integrators/integrator_utils.jl:355 [inlined]
 [10] _loopfooter!(integrator::OrdinaryDiffEqCore.ODEIntegrator{…})
    @ OrdinaryDiffEqCore ~/.julia/packages/OrdinaryDiffEqCore/YWZVL/src/integrators/integrator_utils.jl:243
 [11] loopfooter!
    @ ~/.julia/packages/OrdinaryDiffEqCore/YWZVL/src/integrators/integrator_utils.jl:207 [inlined]
 [12] solve!(integrator::OrdinaryDiffEqCore.ODEIntegrator{…})
    @ OrdinaryDiffEqCore ~/.julia/packages/OrdinaryDiffEqCore/YWZVL/src/solve.jl:552
 [13] #__solve#75
    @ ~/.julia/packages/OrdinaryDiffEqCore/YWZVL/src/solve.jl:7 [inlined]
 [14] __solve
    @ ~/.julia/packages/OrdinaryDiffEqCore/YWZVL/src/solve.jl:1 [inlined]
 [15] #solve_call#44
    @ ~/.julia/packages/DiffEqBase/ODi5x/src/solve.jl:612 [inlined]
 [16] solve_call
    @ ~/.julia/packages/DiffEqBase/ODi5x/src/solve.jl:569 [inlined]
 [17] #solve_up#53
    @ ~/.julia/packages/DiffEqBase/ODi5x/src/solve.jl:1092 [inlined]
 [18] solve_up
    @ ~/.julia/packages/DiffEqBase/ODi5x/src/solve.jl:1078 [inlined]
 [19] #solve#51
    @ ~/.julia/packages/DiffEqBase/ODi5x/src/solve.jl:1015 [inlined]
 [20] top-level scope
    @ REPL[2]:1
Some type information was truncated. Use `show(err)` to see complete types.

Environment (please complete the following information):

julia> import Pkg; Pkg.status()
Status `~/.julia/environments/v1.11/Project.toml`
  [31a5f54b] Debugger v0.7.10
  [0c46a032] DifferentialEquations v7.14.0

julia> versioninfo()
Julia Version 1.11.0
Commit 501a4f25c2b (2024-10-07 11:40 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 4 × Intel(R) Core(TM) i5-6300U CPU @ 2.40GHz
  WORD_SIZE: 64
  LLVM: libLLVM-16.0.6 (ORCJIT, skylake)
Threads: 1 default, 0 interactive, 1 GC (on 4 virtual cores)

[jo-ap]

I came here via the docs and just now noticed that it is only the repo for the documentation. Closed it here and reported it again: SciML/DifferentialEquations.jl#1058.