add fallback for older algorithms that assumes square

Author: ChrisRackauckas

lib/LinearSolvePardiso/src/LinearSolvePardiso.jl

@@ -17,8 +17,9 @@ LinearSolve.needs_concrete_A(alg::PardisoJL) = true
 
 # TODO schur complement functionality
 
-function LinearSolve.init_cacheval(alg::PardisoJL, A, b, u, Pl, Pr, maxiters, abstol,
-                                   reltol, verbose)
+function LinearSolve.init_cacheval(alg::PardisoJL, A, b, u, Pl, Pr, maxiters::Int, abstol,
+                                   reltol, verbose::Bool,
+                                   assumptions::LinearSolve.OperatorAssumptions)
     @unpack nprocs, solver_type, matrix_type, iparm, dparm = alg
     A = convert(AbstractMatrix, A)
 

src/default.jl

@@ -1,3 +1,7 @@
+# Legacy fallback
+# For SciML algorithms already using `defaultalg`, all assume square matrix.
+defaultalg(A, b) = defaultalg(A, b, OperatorAssumptions(Val(true)))
+
 function defaultalg(A::DiffEqArrayOperator, b, assumptions::OperatorAssumptions)
     defaultalg(A.A, b, assumptions)
 end