ldiv! ambiguous for fully blocked problem

[termi-official]

MWE on Julia 1.9.3 with latest release

julia> mortar(reshape([spdiagm([1.0,1.0]), spzeros(2,2), spzeros(2,2), spdiagm([1.0,1.0])], 2,2)) \ Vector(mortar([[1.0,1.0],[1.0,1.0]]))
4-element Vector{Float64}:
 1.0
 1.0
 1.0
 1.0

julia> mortar(reshape([spdiagm([1.0,1.0]), spzeros(2,2), spzeros(2,2), spdiagm([1.0,1.0])], 2,2)) \ mortar([[1.0,1.0],[1.0,1.0]])
ERROR: MethodError: ldiv!(::LinearAlgebra.QRCompactWY{Float64, Matrix{Float64}, Matrix{Float64}}, ::PseudoBlockVector{Float64, Vector{Float64}, Tuple{BlockedUnitRange{Vector{Int64}}}}) is ambiguous.

Candidates:
  ldiv!(A::LinearAlgebra.QRCompactWY{T, M, C} where {M<:AbstractMatrix{T}, C<:AbstractMatrix{T}}, b::AbstractVector{T}) where T<:Union{Float32, Float64, ComplexF32, ComplexF64}
    @ LinearAlgebra ~/Tools/julia-1.9.3/share/julia/stdlib/v1.9/LinearAlgebra/src/qr.jl:873
  ldiv!(A::LinearAlgebra.QRCompactWY, x::ArrayLayouts.LayoutVector; kwds...)
    @ ArrayLayouts ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:153
  ldiv!(A::LinearAlgebra.Factorization, x::ArrayLayouts.LayoutVector; kwds...)
    @ ArrayLayouts ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:150

Possible fix, define
  ldiv!(::LinearAlgebra.QRCompactWY{T, M, C} where {M<:AbstractMatrix{T}, C<:AbstractMatrix{T}}, ::ArrayLayouts.LayoutVector{T}) where T<:Union{Float32, Float64, ComplexF32, ComplexF64}

Stacktrace:
  [1] ldiv!(Y::PseudoBlockVector{Float64, Vector{Float64}, Tuple{BlockedUnitRange{Vector{Int64}}}}, A::LinearAlgebra.QRCompactWY{Float64, Matrix{Float64}, Matrix{Float64}}, B::BlockVector{Float64, Vector{Vector{Float64}}, Tuple{BlockedUnitRange{Vector{Int64}}}})
    @ LinearAlgebra ~/Tools/julia-1.9.3/share/julia/stdlib/v1.9/LinearAlgebra/src/factorization.jl:126
  [2] #_ldiv!#17
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:84 [inlined]
  [3] _ldiv!
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:84 [inlined]
  [4] copyto!(dest::PseudoBlockVector{Float64, Vector{Float64}, Tuple{BlockedUnitRange{Vector{Int64}}}}, M::ArrayLayouts.Ldiv{ArrayLayouts.QRCompactWYLayout{ArrayLayouts.DenseColumnMajor, ArrayLayouts.DenseColumnMajor}, BlockArrays.BlockLayout{ArrayLayouts.DenseColumnMajor, ArrayLayouts.DenseColumnMajor}, LinearAlgebra.QRCompactWY{Float64, Matrix{Float64}, Matrix{Float64}}, BlockVector{Float64, Vector{Vector{Float64}}, Tuple{BlockedUnitRange{Vector{Int64}}}}}; kwds::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}})
    @ ArrayLayouts ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:113
  [5] copyto!
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:113 [inlined]
  [6] #ldiv!#24
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:104 [inlined]
  [7] ldiv!
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:104 [inlined]
  [8] #_ldiv!#16
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:83 [inlined]
  [9] _ldiv!
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:83 [inlined]
 [10] #copyto!#27
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:113 [inlined]
 [11] copyto!
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:113 [inlined]
 [12] #copy#12
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:21 [inlined]
 [13] copy
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:21 [inlined]
 [14] #materialize#13
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:22 [inlined]
 [15] materialize
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:22 [inlined]
 [16] #ldiv#20
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:98 [inlined]
 [17] ldiv
    @ ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:98 [inlined]
 [18] \(A::BlockMatrix{Float64, Matrix{SparseMatrixCSC{Float64, Int64}}, Tuple{BlockedUnitRange{Vector{Int64}}, BlockedUnitRange{Vector{Int64}}}}, x::BlockVector{Float64, Vector{Vector{Float64}}, Tuple{BlockedUnitRange{Vector{Int64}}}}; kwds::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}})
    @ ArrayLayouts ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:199
 [19] \(A::BlockMatrix{Float64, Matrix{SparseMatrixCSC{Float64, Int64}}, Tuple{BlockedUnitRange{Vector{Int64}}, BlockedUnitRange{Vector{Int64}}}}, x::BlockVector{Float64, Vector{Vector{Float64}}, Tuple{BlockedUnitRange{Vector{Int64}}}})
    @ ArrayLayouts ~/.julia/packages/ArrayLayouts/RwIbZ/src/ldiv.jl:199
 [20] top-level scope
    @ REPL[17]:1

[dlfivefifty]

This is actually an issue in ArrayLayouts.jl.

[termi-official]

Oh, sorry. Didn't noticed. Will open an issue there.

[dlfivefifty]

No worries! I maintain both so its fine.

I think we just need to add another case here:

https://github.com/JuliaLinearAlgebra/ArrayLayouts.jl/blob/d44e944e3621153f558e919fdeccf6d7c0604266/src/ldiv.jl#L153C26-L153C26

ie

LinearAlgebra.ldiv!(A::LinearAlgebra.QRCompactWY{T}, x::$Typ{T}; kwds...) where T<:BlasFloat = ArrayLayouts.ldiv!(A,x; kwds...)

If you felt like making a PR and checking it works that would be helpful

[termi-official]

Well, after taking a look into this I think this is a bit harder to fix. Aqua found 2815 ambiguities for this package in total. I picked random samples and they look legit. The ambiguity check is currently disabled ( I think due to the problem with the dependencies in Aqua ambiguitiy check), but this issue can be circumvented by adding the test separately.

Aqua.test_ambiguities(ArrayLayouts) # Must be separate for now, see https://github.com/JuliaTesting/Aqua.jl/issues/77

I will try again to see if I can come up with something.

[termi-official]

Okay let us go with the hotfix for now. Trying to fix more ambiguities just made thigns worse.

Edit: Okay, I diretly ran into more dispatches after fixing this. Will include the fixes in the PR.