[Inductor UT][Windows][XPU] Fix Inductor UT on XPU Windows. (#146481)

This PR fixed all the inductor UT failures for XPU backend on Windows we found in local machine(Due to resource constraints, we have not yet set up a Windows CI pipeline online.)

Pull Request resolved: https://github.com/pytorch/pytorch/pull/146481
Approved by: https://github.com/jansel, https://github.com/EikanWang
ghstack dependencies: #147347

test/inductor/test_debug_trace.py

@@ -11,6 +11,7 @@ from pathlib import Path
 import torch
 from torch._inductor import config, test_operators
 from torch._inductor.utils import fresh_inductor_cache
+from torch.testing._internal.common_utils import skipIfWindows
 from torch.testing._internal.inductor_utils import GPU_TYPE, HAS_GPU
 from torch.testing._internal.logging_utils import multiple_logs_to_string
 
@@ -230,6 +231,8 @@ op2.node.kernel = extern_kernels.mm""",
         # intentionally only cleanup on success so debugging test is easier
         shutil.rmtree(filename)
 
+    # AOT compiler have not supported windows yet.
+    @skipIfWindows
     def test_debug_printer_const(self):
         """Test that having a const example_input does not break the debug printer."""
 

test/inductor/test_extension_backend.py

@@ -7,6 +7,7 @@ import torch
 import torch._dynamo
 import torch.utils.cpp_extension
 from torch._C import FileCheck
+from torch.testing._internal.common_utils import skipIfWindows
 
 
 try:
@@ -90,9 +91,9 @@ class BaseExtensionBackendTests(TestCase):
 
         torch.testing._internal.common_utils.remove_cpp_extensions_build_root()
 
+        cls.lock.release()
         if os.path.exists(cls.lock_file):
             os.remove(cls.lock_file)
-        cls.lock.release()
 
     def setUp(self):
         torch._dynamo.reset()
@@ -114,6 +115,7 @@ class BaseExtensionBackendTests(TestCase):
 
 @unittest.skipIf(IS_FBCODE, "cpp_extension doesn't work in fbcode right now")
 class ExtensionBackendTests(BaseExtensionBackendTests):
+    @skipIfWindows
     def test_open_device_registration(self):
         torch.utils.rename_privateuse1_backend("extension_device")
         torch._register_device_module("extension_device", self.module)

test/inductor/test_indexing.py

@@ -16,6 +16,7 @@ from torch._inductor.utils import run_and_get_triton_code
 from torch.testing._internal.common_utils import (
     instantiate_parametrized_tests,
     IS_MACOS,
+    IS_WINDOWS,
     parametrize,
 )
 from torch.testing._internal.inductor_utils import GPU_TYPE, HAS_CPU, HAS_GPU
@@ -30,7 +31,7 @@ from torch.utils._sympy.functions import (
 
 
 # int64_t is long long on MacOS, but long on 64-bit Linux
-LONG_SUFFIX = "LL" if IS_MACOS else "L"
+LONG_SUFFIX = "LL" if IS_MACOS or IS_WINDOWS else "L"
 DO_PERF_TEST = os.environ.get("DO_PERF_TEST") == "1"
 
 

test/inductor/test_torchinductor.py

@@ -4136,9 +4136,9 @@ class CommonTemplate:
             x = torch.sum(x.view(int(x.shape[0] / 6), 6), dim=1)
             return torch.gather(x, 0, torch.trunc(y).to(torch.int64))
 
-        x1 = torch.randn(30)
+        x1 = torch.randn(30, device=self.device)
-        x2 = torch.randn(36)
+        x2 = torch.randn(36, device=self.device)
-        y = torch.ones(1, dtype=torch.float64)
+        y = torch.ones(1, dtype=torch.float64, device=self.device)
 
         self.assertEqual(torch.compile(fn)(x1, y), fn(x1, y))
         self.assertEqual(torch.compile(fn)(x2, y), fn(x2, y))

test/inductor/test_triton_kernels.py

@@ -22,6 +22,7 @@ from torch.testing._internal import common_utils
 from torch.testing._internal.common_utils import (
     parametrize,
     skipIfRocm,
+    skipIfWindows,
     skipIfXpu,
     TEST_WITH_ROCM,
 )
@@ -3362,6 +3363,7 @@ class CustomOpTests(torch._inductor.test_case.TestCase):
         gm = make_fx(f, tracing_mode=tracing_mode)(x, x)
         self.assertEqual(gm(x, x), x + x)
 
+    @skipIfWindows(msg="AOTI/Cpp_Wrapper have not enabled on Windows")
     @requires_gpu
     @patch.object(torch._inductor.config, "cpp_wrapper", True)
     @patch.object(torch._inductor.config, "triton.autotune_at_compile_time", True)

test/inductor/test_xpu_basic.py

@@ -2,20 +2,10 @@
 import importlib
 import os
 import sys
-import unittest
 
 import torch
-from torch.testing._internal.common_utils import IS_CI, IS_WINDOWS
 
 
-if IS_WINDOWS and IS_CI:
-    sys.stderr.write(
-        "Windows CI does not have necessary dependencies for test_xpu_basic yet\n"
-    )
-    if __name__ == "__main__":
-        sys.exit(0)
-    raise unittest.SkipTest("requires sympy/functorch/filelock")
-
 importlib.import_module("filelock")
 
 pytorch_test_dir = os.path.dirname(os.path.dirname(os.path.realpath(__file__)))

torch/_inductor/codegen/cpp_wrapper_cpu.py

@@ -172,7 +172,7 @@ class CppWrapperCpu(PythonWrapperCodegen):
                 from torch._inductor.codecache import CppWrapperCodeCache
 
                 cpp_wrapper_src = (
-                '''
+                r'''
                 """
             )
 

torch/_inductor/triton_bundler.py

@@ -1,15 +1,17 @@
 import dataclasses
 import logging
 import os
+import shutil
 import uuid
 from pathlib import Path
 from typing import Optional
 
 from torch._dynamo.utils import counters, dynamo_timed, set_feature_use
 from torch._utils_internal import justknobs_check
+from torch.utils._filelock import FileLock
 
 from .runtime.runtime_utils import triton_cache_dir
-from .utils import GPU_KERNEL_BIN_EXTS
+from .utils import _IS_WINDOWS, GPU_KERNEL_BIN_EXTS
 
 
 log = logging.getLogger(__name__)
@@ -238,7 +240,7 @@ class TritonBundler:
                     )
                     continue
 
-                Path(directory).mkdir(parents=True, exist_ok=True)
+                Path(basedir).mkdir(parents=True, exist_ok=True)
 
                 # Random ID to avoid any collisions
                 rnd_id = str(uuid.uuid4())
@@ -260,6 +262,14 @@ class TritonBundler:
                         # Each kernel has bunch of files like .cubin(for cuda), spv(for xpu), .json, .ttir
                         # Just append one of them without the extension
                         kernel_names.append(Path(artifact.filename).stem)
-                # Atomic on POSIX systems
+
-                os.replace(tmp_dir, directory)
+                if _IS_WINDOWS:
+                    with FileLock(directory + ".lock"):
+                        if os.path.exists(directory):
+                            shutil.rmtree(directory)
+                        os.replace(tmp_dir, directory)
+                else:
+                    # Atomic on POSIX systems
+                    os.replace(tmp_dir, directory)
+
             return TritonBundlerMetadata(kernel_names)