Should Tulip catch SQDExceptions thrown by LDLFactorizations?

Specific Domains Optimization (Mathematical)
1

@mtanneau Sometimes MathOptInterface.optimize!(::Tulip.Optimizer{BigFloat}) throws an LDLFactorizations.SQDException:

ERROR: LoadError: LDLFactorizations.SQDException("LDL' factorization was not computed or failed")
Stacktrace:
  [1] ldiv!(LDL::LDLFactorizations.LDLFactorization{BigFloat, Int64, Int64, Int64}, b::Vector{BigFloat})
    @ LDLFactorizations ~/.julia/packages/LDLFactorizations/AIqZZ/src/LDLFactorizations.jl:808 [inlined]
  [2] solve!(dx::Vector{BigFloat}, dy::Vector{BigFloat}, kkt::Tulip.KKT.TlpLDLFactorizations.LDLFactSolver{BigFloat, Tulip.KKT.K2}, ξp::Vector{BigFloat}, ξd::Vector{BigFloat})
    @ Tulip.KKT.TlpLDLFactorizations ~/.julia/packages/Tulip/SRdil/src/KKT/LDLFactorizations/ldlfact.jl:131
  [3] compute_step!(hsd::Tulip.HSD{BigFloat, Vector{…}, BitVector, SparseArrays.SparseMatrixCSC{…}, Tulip.KKT.TlpLDLFactorizations.LDLFactSolver{…}}, params::Tulip.IPMOptions{BigFloat})
    @ Tulip ~/.julia/packages/Tulip/SRdil/src/IPM/HSD/step.jl:62
  [4] macro expansion
    @ Tulip ~/.julia/packages/TimerOutputs/RsWnF/src/TimerOutput.jl:237 [inlined]
  [5] macro expansion
    @ Tulip ~/.julia/packages/Tulip/SRdil/src/IPM/HSD/HSD.jl:320 [inlined]
  [6] macro expansion
    @ Tulip ~/.julia/packages/TimerOutputs/RsWnF/src/TimerOutput.jl:237 [inlined]
  [7] ipm_optimize!(hsd::Tulip.HSD{BigFloat, Vector{…}, BitVector, SparseArrays.SparseMatrixCSC{…}, Tulip.KKT.TlpLDLFactorizations.LDLFactSolver{…}}, params::Tulip.IPMOptions{BigFloat})
    @ Tulip ~/.julia/packages/Tulip/SRdil/src/IPM/HSD/HSD.jl:256
  [8] optimize!(model::Tulip.Model{BigFloat})
    @ Tulip ~/.julia/packages/Tulip/SRdil/src/model.jl:135
  [9] macro expansion
    @ Tulip ./timing.jl:393 [inlined]
 [10] optimize!(m::Tulip.Optimizer{BigFloat})
    @ Tulip ~/.julia/packages/Tulip/SRdil/src/Interfaces/MOI/MOI_wrapper.jl:178 [inlined]

This is with Tulip v0.9.5 and LDLFactorizations v0.10.0.

Is this a Tulip bug? If so, do you want a reproducer, or some specific debugging information?

If this isn’t a bug, what does it indicate?