Qualcomm AI Engine Direct - Support floor_divide with int input in QNN HTP backend (#14888)

### Summary
- Since QNN does not support floor_divide operations for int32 or int64
inputs, it is necessary to decompose the operation into a division using
floating-point precision, followed by applying the floor function.

### Test plan
UT added

Author: @shewu-quic

cc @cccclai @shewu-quic @haowhsu-quic @DannyYuyang-quic @cbilgin

---------

Co-authored-by: shewu <[email protected]>
(cherry picked from commit 5af73eb)

Author: 2 people

backends/qualcomm/_passes/__init__.py

@@ -18,6 +18,7 @@
 from .decompose_col_im import DecomposeColIm
 from .decompose_einsum import DecomposeEinsum
 from .decompose_expm1 import DecomposeExpM1
+from .decompose_floor_divide import DecomposeFloorDivide
 from .decompose_glu import DecomposeGlu
 from .decompose_linalg_vector_norm import DecomposeLinalgVectorNorm
 from .decompose_minmaxdim import DecomposeMinMaxDim
@@ -61,6 +62,7 @@
     DecomposeColIm,
     DecomposeEinsum,
     DecomposeExpM1,
+    DecomposeFloorDivide,
     DecomposeGlu,
     DecomposeLinalgVectorNorm,
     DecomposeMinMaxDim,

backends/qualcomm/_passes/decompose_floor_divide.py

@@ -0,0 +1,62 @@
+# Copyright (c) Qualcomm Innovation Center, Inc.
+# All rights reserved
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+import torch
+from executorch.exir.pass_base import ExportPass, PassResult
+
+from .utils import merge_decomposed_graph
+
+
+class FloorDivide(torch.nn.Module):
+    def __init__(self):
+        super().__init__()
+
+    def forward(self, x, y):
+        dtype = x.dtype
+        result = torch.div(x, y)
+        result = torch.floor(result)
+        return result.to(dtype)
+
+
+class DecomposeFloorDivide(ExportPass):
+    """
+    Decompose for math equivalent op.
+    Since QNN does not support floor_divide operations for int32 or int64 inputs,
+    it is necessary to decompose the operation into a division using floating-point precision,
+    followed by applying the floor function.
+    """
+
+    def __init__(self) -> None:
+        super().__init__()
+
+    def call(self, graph_module: torch.fx.GraphModule) -> PassResult:
+        graph = graph_module.graph
+        for node in graph.nodes:
+            model = FloorDivide()
+            if (
+                torch.ops.aten.floor_divide.default == node.target
+                and not torch.is_floating_point(node.meta["val"])
+            ):
+                decomposed_module = torch.export.export(
+                    model,
+                    (node.args[0].meta["val"], node.args[1].meta["val"]),
+                    strict=True,
+                ).module()
+                with graph.inserting_before(node):
+                    # remap is used to map original node values to new node values,
+                    # which ensures that reference to nodes are correctly updated in the new graph
+                    remap = {"x": node.args[0], "y": node.args[1]}
+                    merge_decomposed_graph(
+                        remap=remap,
+                        target_node=node,
+                        target_graph=graph,
+                        decomposed_graph_module=decomposed_module,
+                    )
+                    graph.erase_node(node)
+
+        graph.eliminate_dead_code()
+        graph_module.recompile()
+        return PassResult(graph_module, True)

backends/qualcomm/_passes/qnn_pass_manager.py

@@ -23,6 +23,7 @@
     DecomposeColIm,
     DecomposeEinsum,
     DecomposeExpM1,
+    DecomposeFloorDivide,
     DecomposeGlu,
     DecomposeLinalgVectorNorm,
     DecomposeMinMaxDim,
@@ -223,6 +224,11 @@ def transform_for_export_pipeline(
         self.add_pass(DecomposeThreshold())
         self.add_pass(DecomposeLinalgVectorNorm(quantization_capture=True))
         self.add_pass(DecomposeExpM1())
+        # DecomposeFloorDivide does not apply to the annotation pipeline,
+        # since the CPU QDQ model would reduce accuracy.
+        # We keep div and floor operations in floating-point to maintain precision.
+        # This pass is needed before to_edge pipeline to avoid mixed type for div operator with RemoveMixedTypeOperators pass.
+        self.add_pass(DecomposeFloorDivide())
         self.add_pass(DecomposeWrapWithAutocast())
         # this pass will rewrite state_dict, it needs to be accomplished before
         # to_edge_transform_and_lower

backends/qualcomm/tests/test_qnn_delegate.py

@@ -398,8 +398,8 @@ def test_qnn_backend_cumsum(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        self.lower_module_and_test_output(module, sample_input)
                         index += 1
+                        self.lower_module_and_test_output(module, sample_input)
 
     def test_qnn_backend_einsum_outer_product(self):
         module = EinsumOuterProduct()  # noqa: F405
@@ -467,8 +467,8 @@ def test_qnn_backend_element_wise_add(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        self.lower_module_and_test_output(module, sample_input)
                         index += 1
+                        self.lower_module_and_test_output(module, sample_input)
 
     def test_qnn_backend_element_wise_and(self):
         module = And(torch.tensor(1.7), torch.tensor(0.2))  # noqa: F405
@@ -506,8 +506,8 @@ def test_qnn_backend_element_wise_div(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        self.lower_module_and_test_output(module, sample_input)
                         index += 1
+                        self.lower_module_and_test_output(module, sample_input)
 
     def test_qnn_backend_element_wise_mul(self):
         test_comb = [
@@ -533,8 +533,8 @@ def test_qnn_backend_element_wise_mul(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        self.lower_module_and_test_output(module, sample_input)
                         index += 1
+                        self.lower_module_and_test_output(module, sample_input)
 
     def test_qnn_backend_element_wise_or(self):
         test_comb = [
@@ -608,8 +608,8 @@ def test_qnn_backend_element_wise_sub(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        self.lower_module_and_test_output(module, sample_input)
                         index += 1
+                        self.lower_module_and_test_output(module, sample_input)
 
     @unittest.expectedFailure
     def test_qnn_backend_elu(self):
@@ -651,10 +651,10 @@ def test_qnn_backend_expand(self):
         for module in modules:
             for sample_input in sample_inputs:
                 with self.subTest(i=index):
+                    index += 1
                     self.lower_module_and_test_output(
                         module, sample_input, passes_job=passes_job
                     )
-                    index += 1
 
     def test_qnn_backend_expm1(self):
         sample_input = (torch.randn(3, 4, 5),)
@@ -677,6 +677,21 @@ def test_qnn_backend_floor_divide(self):
             {
                 QCOM_MODULE: [FloorDiv()],  # noqa: F405
                 QCOM_SAMPLE_INPUTS: [
+                    (torch.randint(-100, 100, (10, 10)), torch.full((10, 10), 3)),
+                    (
+                        torch.randint(-100, 100, (10, 10)).float(),
+                        torch.full((10, 10), 2.5),
+                    ),
+                    (torch.randint(-1000, 1000, (10, 10)), torch.full((10, 10), 100)),
+                    (torch.tensor([10]), torch.arange(1, 5)),  # Failed
+                    (torch.arange(-10, 10), torch.tensor([2])),
+                    (torch.randint(-100, 100, (20,)), torch.full((20,), 2)),
+                    (torch.randint(-100, 100, (5, 10)), torch.full((5, 10), 2)),
+                    (torch.randint(-100, 100, (3, 4, 5)), torch.full((3, 4, 5), 2)),
+                    (
+                        torch.randint(-100, 100, (2, 3, 4, 5)),
+                        torch.full((2, 3, 4, 5), 2),
+                    ),
                     (torch.randn(2, 5, 1, 3), eps + torch.randn(2, 5, 1, 3)),
                     (torch.randn([2, 5, 1, 3]), eps + torch.randn([4, 1])),
                 ],
@@ -692,8 +707,8 @@ def test_qnn_backend_floor_divide(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        self.lower_module_and_test_output(module, sample_input)
                         index += 1
+                        self.lower_module_and_test_output(module, sample_input)
 
     def test_qnn_backend_fold(self):
         sample_input = (torch.randn(3, 512, 256),)
@@ -1136,8 +1151,8 @@ def test_qnn_backend_leaky_relu(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        self.lower_module_and_test_output(module, sample_input)
                         index += 1
+                        self.lower_module_and_test_output(module, sample_input)
 
     def test_qnn_backend_less_equal(self):
         test_comb = [
@@ -1392,8 +1407,8 @@ def test_qnn_backend_prelu(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        self.lower_module_and_test_output(module, sample_input)
                         index += 1
+                        self.lower_module_and_test_output(module, sample_input)
 
     def test_qnn_backend_relu(self):
         module = Relu()  # noqa: F405
@@ -1520,8 +1535,8 @@ def test_qnn_backend_slice_scatter(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        self.lower_module_and_test_output(module, sample_input)
                         index += 1
+                        self.lower_module_and_test_output(module, sample_input)
 
     def test_qnn_backend_stack(self):
         module = Stack()  # noqa: F405
@@ -2332,9 +2347,9 @@ def test_qnn_backend_element_wise_add(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
+                        index += 1
                         gm = self.get_qdq_module(module, sample_input)
                         self.lower_module_and_test_output(gm, sample_input)
-                        index += 1
 
     def test_qnn_backend_element_wise_and(self):
         module = And(torch.tensor(1.7), torch.tensor(0.2))  # noqa: F405
@@ -2373,9 +2388,9 @@ def test_qnn_backend_element_wise_div(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
+                        index += 1
                         gm = self.get_qdq_module(module, sample_input)
                         self.lower_module_and_test_output(gm, sample_input)
-                        index += 1
 
     def test_qnn_backend_element_wise_mul(self):
         test_comb = [
@@ -2401,9 +2416,9 @@ def test_qnn_backend_element_wise_mul(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
+                        index += 1
                         gm = self.get_qdq_module(module, sample_input)
                         self.lower_module_and_test_output(gm, sample_input)
-                        index += 1
 
     def test_qnn_backend_element_wise_or(self):
         test_comb = [
@@ -2479,9 +2494,9 @@ def test_qnn_backend_element_wise_sub(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
+                        index += 1
                         gm = self.get_qdq_module(module, sample_input)
                         self.lower_module_and_test_output(gm, sample_input)
-                        index += 1
 
     def test_qnn_backend_elu(self):
         module = Elu()  # noqa: F405
@@ -2530,11 +2545,11 @@ def test_qnn_backend_expand(self):
         for module in modules:
             for sample_input in sample_inputs:
                 with self.subTest(i=index):
+                    index += 1
                     module = self.get_qdq_module(module, sample_input)
                     self.lower_module_and_test_output(
                         module, sample_input, passes_job=passes_job
                     )
-                    index += 1
 
     def test_qnn_backend_expm1(self):
         sample_input = (torch.randn(3, 4, 5),)
@@ -2560,6 +2575,21 @@ def test_qnn_backend_floor_divide(self):
             {
                 QCOM_MODULE: [FloorDiv()],  # noqa: F405
                 QCOM_SAMPLE_INPUTS: [
+                    (torch.randint(-100, 100, (10, 10)), torch.full((10, 10), 3)),
+                    (
+                        torch.randint(-100, 100, (10, 10)).float(),
+                        torch.full((10, 10), 2.5),
+                    ),
+                    (torch.randint(-1000, 1000, (10, 10)), torch.full((10, 10), 100)),
+                    (torch.tensor([10]), torch.arange(1, 5)),
+                    (torch.arange(-10, 10), torch.tensor([2])),
+                    (torch.randint(-100, 100, (20,)), torch.full((20,), 2)),
+                    (torch.randint(-100, 100, (5, 10)), torch.full((5, 10), 2)),
+                    (torch.randint(-100, 100, (3, 4, 5)), torch.full((3, 4, 5), 2)),
+                    (
+                        torch.randint(-100, 100, (2, 3, 4, 5)),
+                        torch.full((2, 3, 4, 5), 2),
+                    ),
                     (torch.randn(2, 5, 1, 3), eps + torch.randn(2, 5, 1, 3)),
                     (torch.randn([2, 5, 1, 3]), eps + torch.randn([4, 1])),
                 ],
@@ -2575,9 +2605,12 @@ def test_qnn_backend_floor_divide(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        gm = self.get_qdq_module(module, sample_input)
-                        self.lower_module_and_test_output(gm, sample_input)
                         index += 1
+                        # Support int input cases with bypass_check=True
+                        gm = self.get_qdq_module(
+                            module, sample_input, bypass_check=True
+                        )
+                        self.lower_module_and_test_output(gm, sample_input)
 
     def test_qnn_backend_fold(self):
         sample_input = (torch.randn(3, 512, 256),)
@@ -3048,9 +3081,9 @@ def test_qnn_backend_leaky_relu(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
+                        index += 1
                         module = self.get_qdq_module(module, sample_input)
                         self.lower_module_and_test_output(module, sample_input)
-                        index += 1
 
     def test_qnn_backend_less_equal(self):
         test_comb = [
@@ -3352,9 +3385,9 @@ def test_qnn_backend_prelu(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
-                        qdq_module = self.get_qdq_module(module, sample_input)
-                        self.lower_module_and_test_output(qdq_module, sample_input)
                         index += 1
+                        module = self.get_qdq_module(module, sample_input)
+                        self.lower_module_and_test_output(module, sample_input)
 
     def test_qnn_backend_relu(self):
         module = Relu()  # noqa: F405
@@ -3504,9 +3537,9 @@ def test_qnn_backend_slice_scatter(self):
             for module in comb[QCOM_MODULE]:
                 for sample_input in comb[QCOM_SAMPLE_INPUTS]:
                     with self.subTest(i=index):
+                        index += 1
                         module = self.get_qdq_module(module, sample_input)
                         self.lower_module_and_test_output(module, sample_input)
-                        index += 1
 
     def test_qnn_backend_softmax(self):
         modules = [Softmax(dim=1), Softmax(dim=-1)]  # noqa: F405