Expand dict API to allow dtype strings (#8323)

Summary:

Allow users to specify dtypes using strings, instead of kernel keys.

Note: this generates a 'fake' kernel key with no dim order data. I think this is OK, as I don't think ExecuTorch uses the kernel key for anything besides dtype.

https://www.internalfb.com/code/fbsource/[c02ea3c66de17a60b2f0e65462a03ccf5cba9804]/fbcode/caffe2/torchgen/selective_build/selector.py?lines=49-52

Reviewed By: larryliu0820

Differential Revision: D69335674

Author: lucylq

codegen/tools/gen_oplist.py

@@ -189,6 +189,23 @@ def _dump_yaml(
         )
 
 
+def create_kernel_key(maybe_kernel_key: str) -> str:
+    # It is a kernel key.
+    if maybe_kernel_key.lstrip().startswith("v1"):
+        return maybe_kernel_key
+    # It is a dtype.
+    else:
+        # Generate a kernel key based on the dtype provided.
+        # Note: no dim order is included in this kernel key.
+        # For a description of the kernel key format, see
+        # executorch/blob/main/runtime/kernel/operator_registry.h#L97-L123
+        try:
+            dtype = ScalarType[maybe_kernel_key]
+            return "v1/" + str(dtype.value) + ";"
+        except KeyError:
+            raise Exception(f"Unknown dtype: {maybe_kernel_key}")
+
+
 def gen_oplist(
     output_path: str,
     model_file_path: Optional[str] = None,
@@ -223,7 +240,11 @@ def gen_oplist(
         ops_and_metadata = json.loads(ops_dict)
         for op, metadata in ops_and_metadata.items():
             op_set.update({op})
-            op_metadata = metadata if len(metadata) > 0 else ["default"]
+            op_metadata = (
+                [create_kernel_key(x) for x in metadata]
+                if len(metadata) > 0
+                else ["default"]
+            )
             et_kernel_metadata = merge_et_kernel_metadata(
                 et_kernel_metadata, {op: op_metadata}
             )

codegen/tools/test/test_gen_oplist.py

@@ -13,6 +13,7 @@
 
 import executorch.codegen.tools.gen_oplist as gen_oplist
 import yaml
+from executorch.codegen.tools.gen_oplist import ScalarType
 
 
 class TestGenOpList(unittest.TestCase):
@@ -89,7 +90,7 @@ def test_gen_op_list_with_root_ops_and_dtypes(
     ) -> None:
         output_path = os.path.join(self.temp_dir.name, "output.yaml")
         ops_dict = {
-            "aten::add": ["v1/3;0,1|3;0,1|3;0,1|3;0,1", "v1/6;0,1|6;0,1|6;0,1|6;0,1"],
+            "aten::add": ["v1/3;0,1|3;0,1|3;0,1|3;0,1", ScalarType.Float.name],
             "aten::mul": [],
         }
         args = [
@@ -104,7 +105,7 @@ def test_gen_op_list_with_root_ops_and_dtypes(
             {
                 "aten::add": [
                     "v1/3;0,1|3;0,1|3;0,1|3;0,1",
-                    "v1/6;0,1|6;0,1|6;0,1|6;0,1",
+                    "v1/6;",
                 ],
                 "aten::mul": ["default"],
             },

codegen/tools/test/test_gen_selected_op_variants.py

@@ -12,7 +12,7 @@
 import expecttest
 
 
-class TestGenSelectedMobileOpsHeader(expecttest.TestCase):
+class TestGenSelectedOpVariants(expecttest.TestCase):
     def setUp(self):
         self.temp_dir = tempfile.TemporaryDirectory()
         self.addCleanup(self.temp_dir.cleanup)
@@ -84,7 +84,79 @@ def test_generates_correct_header(self) -> None:
             )
 
 
-class TestGenSelectedMobileOpsHeader_Empty(expecttest.TestCase):
+class TestGenSelectedOpVariants_UsingDtypeString(expecttest.TestCase):
+    def setUp(self):
+        self.temp_dir = tempfile.TemporaryDirectory()
+        self.addCleanup(self.temp_dir.cleanup)
+        self.selected_ops_yaml = os.path.join(
+            self.temp_dir.name, "selected_operators.yaml"
+        )
+        with open(self.selected_ops_yaml, "w") as f:
+            f.write(
+                """
+include_all_non_op_selectives: False
+include_all_operators: False
+debug_info:
+  - model1@v100
+  - model2@v50
+operators:
+  aten::add:
+    is_root_operator: Yes
+    is_used_for_training: Yes
+    include_all_overloads: No
+  aten::add.int:
+    is_root_operator: No
+    is_used_for_training: No
+    include_all_overloads: Yes
+kernel_metadata: {}
+et_kernel_metadata:
+  aten::add.out:
+    # A list of different kernel keys (tensors with dtype-enum/dim-order) combinations used in model
+      - v1/6; # Float
+      - v1/3; # Int
+  aten::mul.out:
+      - v1/6; # Float
+  aten::sub.out:
+      - default
+build_features: []
+custom_classes: []
+            """
+            )
+
+    def tearDown(self):
+        self.temp_dir.cleanup()
+
+    def test_generates_correct_header(self) -> None:
+        gen_selected_op_variants.write_selected_op_variants(
+            os.path.join(self.temp_dir.name, "selected_operators.yaml"),
+            self.temp_dir.name,
+        )
+        with open(
+            os.path.join(self.temp_dir.name, "selected_op_variants.h"), "r"
+        ) as result:
+            self.assertExpectedInline(
+                result.read(),
+                """#pragma once
+/**
+ * Generated by executorch/codegen/tools/gen_selected_op_variants.py
+ */
+
+inline constexpr bool should_include_kernel_dtype(
+  const char *operator_name,
+  executorch::aten::ScalarType scalar_type
+) {
+  return ((executorch::aten::string_view(operator_name).compare("add.out") == 0)
+        && (scalar_type == executorch::aten::ScalarType::Float || scalar_type == executorch::aten::ScalarType::Int))
+ || ((executorch::aten::string_view(operator_name).compare("mul.out") == 0)
+        && (scalar_type == executorch::aten::ScalarType::Float))
+ || ((executorch::aten::string_view(operator_name).compare("sub.out") == 0)
+        && (true));
+}
+""",
+            )
+
+
+class TestGenSelectedOpVariants_Empty(expecttest.TestCase):
     def setUp(self):
         self.temp_dir = tempfile.TemporaryDirectory()
         self.addCleanup(self.temp_dir.cleanup)

examples/selective_build/targets.bzl

@@ -1,5 +1,5 @@
 load("@fbsource//xplat/executorch/build:runtime_wrapper.bzl", "get_oss_build_kwargs", "is_xplat", "runtime")
-load("@fbsource//xplat/executorch/codegen:codegen.bzl", "et_operator_library", "executorch_generated_lib")
+load("@fbsource//xplat/executorch/codegen:codegen.bzl", "et_operator_library", "executorch_generated_lib", "ScalarType")
 
 def define_common_targets():
     """Defines targets that should be shared between fbcode and xplat.
@@ -49,7 +49,9 @@ def define_common_targets():
     et_operator_library(
         name = "select_ops_in_dict",
         ops_dict = {
-            "aten::add.out": ["v1/3;0,1", "v1/6;0,1"],  # int, float
+            # 1. Use kernel key, generated with a model, or
+            # 2. Specify the dtype, from executorch/codegen/codegen.bzl
+            "aten::add.out": ["v1/3;0,1", ScalarType("Float")],  # int, float
             "aten::mm.out": [],  # all dtypes
         },
     )

examples/selective_build/test_selective_build.sh

@@ -66,7 +66,6 @@ test_buck2_select_ops_in_dict() {
     # select ops and their dtypes using the dictionary API.
     $BUCK run //examples/selective_build:selective_build_test \
         --config=executorch.select_ops=dict \
-        --config=executorch.dtype_selective_build_lib=//examples/selective_build:select_ops_in_dict_lib \
         -- --model_path=./add_mul.pte
 
     echo "Removing add_mul.pte"

shim/xplat/executorch/codegen/codegen.bzl

@@ -42,6 +42,39 @@ CUSTOM_OPS_SCHEMA_REGISTRATION_SOURCES = [
     "RegisterSchema.cpp",
 ]
 
+ScalarType = enum(
+    "Byte",
+    "Char",
+    "Short",
+    "Int",
+    "Long",
+    "Half",
+    "Float",
+    "Double",
+    "ComplexHalf",
+    "ComplexFloat",
+    "ComplexDouble",
+    "Bool",
+    "QInt8",
+    "QUInt8",
+    "QInt32",
+    "BFloat16",
+    "QUInt4x2",
+    "QUInt2x4",
+    "Bits1x8",
+    "Bits2x4",
+    "Bits4x2",
+    "Bits8",
+    "Bits16",
+    "Float8_e5m2",
+    "Float8_e4m3fn",
+    "Float8_e5m2fnuz",
+    "Float8_e4m3fnuz",
+    "UInt16",
+    "UInt32",
+    "Uint64",
+)
+
 # Hide the dependency to caffe2 internally.
 def et_operator_library(
         name,