[XNNPACK] Reject addmm with non-default alpha or beta

torch.addmm(input, mat1, mat2, beta=b, alpha=a) computes
b * input + a * (mat1 @ mat2), but the XNNPACK lowering of addmm ignores
alpha and beta. AddmmConfig accepted such an addmm whenever its weight and
bias were static params, so the delegate silently computed
mat1 @ mat2 + input. It now rejects them, and the addmm runs on the portable
kernels.

This change was authored with Claude Code.

Author: JakeStevens

backends/xnnpack/partition/config/gemm_configs.py

@@ -479,6 +479,12 @@ def __init__(self, **kwargs):
         self.src_partitions = None
         self.linear_modules = [torch.nn.functional.linear, torch.nn.Linear]
 
+    def check_constraints(self, node: torch.fx.Node, ep: ExportedProgram) -> bool:
+        if node.kwargs.get("alpha", 1) != 1 or node.kwargs.get("beta", 1) != 1:
+            why(node, "XNNPACK does not support addmm with alpha or beta")
+            return False
+        return super().check_constraints(node, ep)
+
     def _get_weight_deps(
         self, node: torch.fx.Node, ep: ExportedProgram, precision: ConfigPrecisionType
     ) -> Tuple[bool, List[torch.fx.Node]]:

backends/xnnpack/test/ops/test_linear.py

@@ -782,6 +782,22 @@ def test_fp32_addmm(self):
             uses_bias=True,
         )
 
+    def test_fp32_addmm_with_alpha_and_beta(self):
+        class ScaledAddMMModule(AddMMModule):
+            def forward(self, x):
+                return torch.addmm(self.bias, x, self.mat, alpha=2, beta=3)
+
+        # XNNPACK has no alpha or beta, so the addmm stays on portable.
+        (
+            Tester(ScaledAddMMModule(8, 4).eval(), (torch.randn(2, 8),))
+            .export()
+            .to_edge_transform_and_lower()
+            .check(["executorch_exir_dialects_edge__ops_aten_addmm_default"])
+            .to_executorch()
+            .serialize()
+            .run_method_and_compare_outputs()
+        )
+
     def test_fp32_linear_view_weight(self):
         class ViewWeightLinear(torch.nn.Module):
             def __init__(self):