Mark RecursiveFactorization Enzyme tests broken (Enzyme >= 0.13.155 regression)

test/nopre/enzyme.jl

@@ -42,25 +42,32 @@ _ff = (
 )
 _ff(copy(A), copy(b1))
 
-Enzyme.autodiff(
-    Reverse,
-    (
-        x,
-        y,
-    ) -> f(
-        x,
-        y;
-        alg = LinearSolve.DefaultLinearSolver(LinearSolve.DefaultAlgorithmChoice.LUFactorization)
-    ),
-    Duplicated(copy(A), dA),
-    Duplicated(copy(b1), db1)
-)
+# Enzyme >= 0.13.155 crashes compiling any primal containing
+# RecursiveFactorization's explicit-SIMD kernels (vector GEP MethodError in
+# abs_typeof, https://github.com/EnzymeAD/Enzyme.jl/issues/3164). With
+# RecursiveFactorization loaded, DefaultLinearSolver compiles the RFLU branch,
+# so this site is affected. The @test_broken flips to an unexpected-pass error
+# when a fixed Enzyme release lands — unbreak this and the other #3164 sites.
+@test_broken begin
+    Enzyme.autodiff(
+        Reverse,
+        (
+            x,
+            y,
+        ) -> f(
+            x,
+            y;
+            alg = LinearSolve.DefaultLinearSolver(LinearSolve.DefaultAlgorithmChoice.LUFactorization)
+        ),
+        Duplicated(copy(A), dA),
+        Duplicated(copy(b1), db1)
+    )
 
-dA2 = ForwardDiff.gradient(x -> f(x, eltype(x).(b1)), copy(A))
-db12 = ForwardDiff.gradient(x -> f(eltype(x).(A), x), copy(b1))
+    dA2 = ForwardDiff.gradient(x -> f(x, eltype(x).(b1)), copy(A))
+    db12 = ForwardDiff.gradient(x -> f(eltype(x).(A), x), copy(b1))
 
-@test dA ≈ dA2
-@test db1 ≈ db12
+    dA ≈ dA2 && db1 ≈ db12
+end
 
 A = rand(n, n);
 dA = zeros(n, n);
@@ -167,14 +174,16 @@ f2(A, b1, b2)
 dA = zeros(n, n);
 db1 = zeros(n);
 db2 = zeros(n);
-Enzyme.autodiff(
-    Reverse, f2, Duplicated(copy(A), dA),
-    Duplicated(copy(b1), db1), Duplicated(copy(b2), db2)
-)
+# Broken by the Enzyme >= 0.13.155 RecursiveFactorization regression
+# (https://github.com/EnzymeAD/Enzyme.jl/issues/3164)
+@test_broken begin
+    Enzyme.autodiff(
+        Reverse, f2, Duplicated(copy(A), dA),
+        Duplicated(copy(b1), db1), Duplicated(copy(b2), db2)
+    )
 
-@test dA ≈ dA2
-@test db1 ≈ db12
-@test db2 ≈ db22
+    dA ≈ dA2 && db1 ≈ db12 && db2 ≈ db22
+end
 
 function f3(A, b1, b2; alg = KrylovJL_GMRES())
     prob = LinearProblem(A, b1)
@@ -245,6 +254,25 @@ end
         LUFactorization(),
         RFLUFactorization(),    # KrylovJL_GMRES(), fails
     )
+    # Forward mode over RFLU also hits the Enzyme >= 0.13.155 regression
+    # (https://github.com/EnzymeAD/Enzyme.jl/issues/3164)
+    if alg isa RFLUFactorization
+        @test_broken begin
+            en_jac = map(onehot(b1)) do db1
+                return only(
+                    Enzyme.autodiff(
+                        set_runtime_activity(Forward), fnice,
+                        Const(A), Duplicated(b1, db1), Const(alg)
+                    )
+                )
+            end |> collect
+            fd_jac = FiniteDiff.finite_difference_jacobian(b -> fnice(A, b, alg), b1) |>
+                vec
+            isapprox(en_jac, fd_jac, rtol = 1.0e-4)
+        end
+        continue
+    end
+
     fb_closure = b -> fnice(A, b, alg)
 
     fd_jac = FiniteDiff.finite_difference_jacobian(fb_closure, b1) |> vec