Type Stability of IpoptProblem by hfytr · Pull Request #532 · jump-dev/Ipopt.jl

hfytr · 2026-04-02T00:18:39Z

I have been trying to use JuliaC.jl with Ipopt.jl. This requires fixing #518. (JuliaC also doesn't work with #519).

odow · 2026-04-02T01:45:52Z

The current suggestion is to use PackageCompiler, not JuliaC:

We test it with JuMP and Ipopt: https://github.com/jump-dev/JuMP.jl/blob/master/test/PackageCompiler/build_app.jl

There are a number of trade-offs to consider here, including whether it is worth making a breaking change.

hfytr · 2026-04-02T01:56:51Z

Thanks @odow. I'd like to take advantage of juliac trimming capability. If I make the pr backward compatible can it be merged? Also, what is the breaking change?

odow · 2026-04-02T02:30:38Z

I understand the motivation for wanting to use JuliaC, but proper support is likely years rather than months away.

Adding new type parameters can be breaking if anyone dispatched on Type{IpoptProblem}. There's also a cost to highly parameterised types in error messages and stacktraces and precompilation.

Ideally we'd also merge something like this only if there are also tests that run JuliaC so we can confirm it works.

I guess the fact that there are multiple open PRs shows that there's not a simple quick fix 😄

odow · 2026-04-02T02:32:55Z


-mutable struct IpoptProblem
+mutable struct IntermediateCallbackWrapper
+    f::Function


Why is this one okay? Is it just that you didn't test JuliaC with an intermediate callback?

This should be fixed in latest commit.

I still don't understand: why is the non-type-stable f::Function okay here, but it wasn't okay in IpoptProblem for eval_f::Function etc?

codecov · 2026-04-02T18:40:48Z

Codecov Report

❌ Patch coverage is 87.50000% with 3 lines in your changes missing coverage. Please review.
✅ Project coverage is 99.75%. Comparing base (1f6a40c) to head (a2fc4e1).

Files with missing lines	Patch %	Lines
src/C_wrapper.jl	87.50%	3 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##            master     #532      +/-   ##
===========================================
- Coverage   100.00%   99.75%   -0.25%     
===========================================
  Files            5        5              
  Lines         1191     1200       +9     
===========================================
+ Hits          1191     1197       +6     
- Misses           0        3       +3

☔ 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.

odow · 2026-04-02T19:59:48Z

I'd need to double check, but I think the other consequence of this is that precompilation would now be specific to the model:

Ipopt.jl/ext/IpoptMathOptInterfaceExt/IpoptMathOptInterfaceExt.jl

Lines 21 to 77 in 1f6a40c

    
           PrecompileTools.@setup_workload begin 
        
               PrecompileTools.@compile_workload begin 
        
                   model = MOI.Utilities.CachingOptimizer( 
        
                       MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}()), 
        
                       MOI.instantiate(Optimizer; with_bridge_type = Float64), 
        
                   ) 
        
                   # We don't want to advertise this option, but it's required so that 
        
                   # we don't print the banner during precompilation. 
        
                   MOI.set(model, MOI.RawOptimizerAttribute("sb"), "yes") 
        
                   MOI.set(model, MOI.Silent(), true) 
        
                   x = MOI.add_variables(model, 3) 
        
                   MOI.supports(model, MOI.VariableName(), typeof(x[1])) 
        
                   MOI.set(model, MOI.VariableName(), x[1], "x1") 
        
                   MOI.set(model, MOI.VariablePrimalStart(), x[1], 0.0) 
        
                   for F in (MOI.VariableIndex, MOI.ScalarAffineFunction{Float64}) 
        
                       MOI.supports_constraint(model, F, MOI.GreaterThan{Float64}) 
        
                       MOI.supports_constraint(model, F, MOI.LessThan{Float64}) 
        
                       MOI.supports_constraint(model, F, MOI.EqualTo{Float64}) 
        
                       # These return false, but it doesn't matter 
        
                       MOI.supports_constraint(model, F, MOI.ZeroOne) 
        
                       MOI.supports_constraint(model, F, MOI.Integer) 
        
                   end 
        
                   MOI.add_constraint(model, x[1], MOI.GreaterThan(0.0)) 
        
                   MOI.add_constraint(model, x[2], MOI.LessThan(0.0)) 
        
                   MOI.add_constraint(model, x[3], MOI.EqualTo(0.0)) 
        
                   f = 1.0 * x[1] + x[2] + x[3] 
        
                   c1 = MOI.add_constraint(model, f, MOI.GreaterThan(0.0)) 
        
                   MOI.set(model, MOI.ConstraintName(), c1, "c1") 
        
                   MOI.supports(model, MOI.ConstraintName(), typeof(c1)) 
        
                   MOI.add_constraint(model, f, MOI.LessThan(0.0)) 
        
                   MOI.add_constraint(model, f, MOI.EqualTo(0.0)) 
        
                   y, _ = MOI.add_constrained_variables(model, MOI.Nonnegatives(2)) 
        
                   MOI.set(model, MOI.ObjectiveSense(), MOI.MAX_SENSE) 
        
                   f = MOI.ScalarNonlinearFunction( 
        
                       :+, 
        
                       Any[MOI.ScalarNonlinearFunction(:sin, Any[x[i]]) for i in 1:3], 
        
                   ) 
        
                   MOI.supports(model, MOI.ObjectiveFunction{typeof(f)}()) 
        
                   MOI.set(model, MOI.ObjectiveFunction{typeof(f)}(), f) 
        
                   MOI.add_constraint(model, f, MOI.EqualTo(0.0)) 
        
                   MOI.optimize!(model) 
        
                   MOI.get(model, MOI.TerminationStatus()) 
        
                   MOI.get(model, MOI.PrimalStatus()) 
        
                   MOI.get(model, MOI.DualStatus()) 
        
                   MOI.get(model, MOI.VariablePrimal(), x) 
        
                   # We put these after `optimize!` so that the error is thrown on add, 
        
                   # not on optimize! 
        
                   try 
        
                       MOI.add_constraint(model, x[1], MOI.ZeroOne()) 
        
                   catch 
        
                   end 
        
                   try 
        
                       MOI.add_constraint(model, x[1], MOI.Integer()) 
        
                   catch 
        
                   end 
        
               end 
        
           end

So this would hurt the startup times of all users not using JuliaC.

odow · 2026-04-02T22:59:51Z

Precompilation doesn't seem to be affected:

Before

(tmp) pkg> st
Status `/private/tmp/Project.toml`
  [b6b21f68] Ipopt v1.14.1
  [4076af6c] JuMP v1.30.0

julia> using JuMP, Ipopt

julia> @time begin
           model = Model(Ipopt.Optimizer)
           set_silent(model)
           @variable(model, x)
           @variable(model, y)
           @objective(model, Min, (1 - x)^2 + 100 * (y - x^2)^2)
           optimize!(model)
           assert_is_solved_and_feasible(model)
       end

******************************************************************************
This program contains Ipopt, a library for large-scale nonlinear optimization.
 Ipopt is released as open source code under the Eclipse Public License (EPL).
         For more information visit https://github.com/coin-or/Ipopt
******************************************************************************

  2.261278 seconds (11.55 M allocations: 556.102 MiB, 6.35% gc time, 99.85% compilation time: 1% of which was recompilation)

After

(ip2) pkg> st
Status `/private/tmp/ip2/Project.toml`
  [b6b21f68] Ipopt v1.14.1 `~/git/jump-dev/Ipopt`
  [4076af6c] JuMP v1.30.0

julia> using JuMP, Ipopt

julia> @time begin
           model = Model(Ipopt.Optimizer)
           set_silent(model)
           @variable(model, x)
           @variable(model, y)
           @objective(model, Min, (1 - x)^2 + 100 * (y - x^2)^2)
           optimize!(model)
           assert_is_solved_and_feasible(model)
       end

******************************************************************************
This program contains Ipopt, a library for large-scale nonlinear optimization.
 Ipopt is released as open source code under the Eclipse Public License (EPL).
         For more information visit https://github.com/coin-or/Ipopt
******************************************************************************

  2.262386 seconds (11.55 M allocations: 556.119 MiB, 7.09% gc time, 99.85% compilation time: 1% of which was recompilation)

odow · 2026-04-02T23:19:10Z

If we are to consider making a breaking change by adding type parameters, I wonder if we're better off doing something like:

abstract type AbstractOracle end

struct HS071 <: AbstractOracle end

function Ipopt.eval_f(::HS071, x::Vector{Float64})
    return x[1] * x[4] * (x[1] + x[2] + x[3]) + x[3]
end

Ipopt.eval_grad_f()
Ipopt.eval_g()
Ipopt.eval_jac_g()
Ipopt.has_eval_h(::HS071) = true
Ipopt.eval_h()

prob = Ipopt.CreateIpoptProblem(n, x_L, x_U, m, g_L, g_U, 8, 10, HS071())

struct IpoptProblem{O<:AbstractOracle}
    oracle::O
end

This would be much more "Julian" in its approach. It would be completely type stable. It avoids the intermediate callback issue, because you could have has_intermediate_cb(::HS071). The oracles are more testable in isolation, and it also solves #514 because the data could be part of the oracle.

odow · 2026-04-03T02:16:21Z

Take a look at the alternative in #535.

There's now a merge conflict because I made some changes inspired by some of your diffs.

hfytr added 2 commits April 1, 2026 19:15

remove use of Function

c02a6e3

reenable setting intermediate callback

5f36077

hfytr mentioned this pull request Apr 2, 2026

Use new Ipopt.jl interface. JuliaSmoothOptimizers/NLPModelsIpopt.jl#155

Draft

odow reviewed Apr 2, 2026

View reviewed changes

hfytr added 2 commits April 2, 2026 09:30

c_wrapper fixes

c01b081

bool vs cint fix

a2fc4e1

This was referenced Apr 2, 2026

Change to using Ptr{IpoptProblem} to give callbacks more type information #533

Merged

Add a CI job to test JuliaC #534

Merged

odow mentioned this pull request Apr 3, 2026

Add AbstractOracle to make a type-stable interface #535

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Type Stability of IpoptProblem#532

Type Stability of IpoptProblem#532
hfytr wants to merge 4 commits intojump-dev:masterfrom
hfytr:master

hfytr commented Apr 2, 2026

Uh oh!

odow commented Apr 2, 2026

Uh oh!

hfytr commented Apr 2, 2026

Uh oh!

odow commented Apr 2, 2026

Uh oh!

odow Apr 2, 2026

Uh oh!

hfytr Apr 2, 2026

Uh oh!

odow Apr 2, 2026

Uh oh!

codecov Bot commented Apr 2, 2026 •

edited

Loading

Uh oh!

odow commented Apr 2, 2026

Uh oh!

odow commented Apr 2, 2026

Uh oh!

odow commented Apr 2, 2026

Uh oh!

odow commented Apr 3, 2026

Uh oh!

Reviewers

Assignees

Labels

Milestone

Development

Uh oh!

2 participants

Conversation

hfytr commented Apr 2, 2026

Uh oh!

odow commented Apr 2, 2026

Uh oh!

hfytr commented Apr 2, 2026

Uh oh!

odow commented Apr 2, 2026

Uh oh!

odow Apr 2, 2026

Choose a reason for hiding this comment

Uh oh!

hfytr Apr 2, 2026

Choose a reason for hiding this comment

Uh oh!

odow Apr 2, 2026

Choose a reason for hiding this comment

Uh oh!

codecov Bot commented Apr 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

odow commented Apr 2, 2026

Uh oh!

odow commented Apr 2, 2026

Uh oh!

odow commented Apr 2, 2026

Uh oh!

odow commented Apr 3, 2026

Uh oh!

Reviewers

Assignees

Labels

Milestone

Development

Uh oh!

2 participants

codecov Bot commented Apr 2, 2026 •

edited

Loading