[Temp] Add subtiling to FlexAttention

Differential Revision: D86172337

.ci/pytorch/test.sh

@@ -337,7 +337,7 @@ test_python() {
 
 test_python_smoke() {
   # Smoke tests for H100/B200
-  time python test/run_test.py --include test_matmul_cuda test_scaled_matmul_cuda inductor/test_fp8 inductor/test_max_autotune inductor/test_cutedsl_grouped_mm $PYTHON_TEST_EXTRA_OPTION --upload-artifacts-while-running
+  time python test/run_test.py --include test_matmul_cuda test_scaled_matmul_cuda inductor/test_fp8 inductor/test_max_autotune $PYTHON_TEST_EXTRA_OPTION --upload-artifacts-while-running
   assert_git_not_dirty
 }
 

.github/labeler.yml

@@ -138,7 +138,8 @@
 - test/test_matmul_cuda.py
 - test/test_scaled_matmul_cuda.py
 - test/inductor/test_fp8.py
-- aten/src/ATen/native/cuda/Blas.cpp
+- aten/src/ATen/native/cuda/*Blas.cpp
+- aten/src/ATen/cuda/CUDA*Blas.*
 - torch/**/*cublas*
 - torch/_inductor/kernel/mm.py
 - test/inductor/test_max_autotune.py
@@ -148,7 +149,8 @@
 - test/test_matmul_cuda.py
 - test/test_scaled_matmul_cuda.py
 - test/inductor/test_fp8.py
-- aten/src/ATen/native/cuda/Blas.cpp
+- aten/src/ATen/native/cuda/*Blas.cpp
+- aten/src/ATen/cuda/CUDA*Blas.*
 - torch/**/*cublas*
 - torch/_inductor/kernel/mm.py
 - test/inductor/test_max_autotune.py
@@ -158,7 +160,8 @@
 - test/test_matmul_cuda.py
 - test/test_scaled_matmul_cuda.py
 - test/inductor/test_fp8.py
-- aten/src/ATen/native/cuda/Blas.cpp
+- aten/src/ATen/native/cuda/*Blas.cpp
+- aten/src/ATen/cuda/CUDA*Blas.*
 - torch/_inductor/kernel/mm.py
 - test/inductor/test_max_autotune.py
 - third_party/fbgemm

.gitignore

@@ -127,7 +127,6 @@ torch/test/
 torch/utils/benchmark/utils/valgrind_wrapper/callgrind.h
 torch/utils/benchmark/utils/valgrind_wrapper/valgrind.h
 torch/version.py
-torch/_inductor/kernel/vendored_templates/*
 minifier_launcher.py
 aten/src/ATen/native/transformers/hip/flash_attn/ck/fmha_fwd_d*
 aten/src/ATen/native/transformers/hip/flash_attn/ck/fmha_bwd_d*

CODEOWNERS

@@ -210,8 +210,12 @@ torch/backends/cudnn/ @eqy @syed-ahmed @Aidyn-A
 /test/inductor/test_flex_attention.py @drisspg
 /test/inductor/test_flex_decoding.py @drisspg
 
-# Low Precision GEMMs
+# Low Precision & Grouped GEMMs
 /aten/src/ATen/native/cuda/Blas.cpp @drisspg @slayton58
+/aten/src/ATen/native/cuda/GroupedBlas.cpp @drisspg @slayton58
+/aten/src/ATen/native/cuda/ScaledBlas.cpp @drisspg @slayton58
 /aten/src/ATen/cuda/CUDABlas.cpp @drisspg @slayton58
 /aten/src/ATen/cuda/CUDABlas.h @drisspg @slayton58
+/aten/src/ATen/cuda/CUDAScaledBlas.cpp @drisspg @slayton58
+/aten/src/ATen/cuda/CUDAScaledBlas.h @drisspg @slayton58
 /test/test_scaled_matmul_cuda.py @drisspg @slayton58

aten/src/ATen/native/native_functions.yaml

@@ -2803,7 +2803,7 @@
 - func: floor_divide.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
   device_check: NoCheck   # TensorIterator
   dispatch:
-    CPU, CUDA, MPS: floor_divide_out
+    CPU, CUDA, MPS, MTIA: floor_divide_out
     SparseCPU, SparseCUDA, SparseMPS: floor_divide_out_sparse_zerodim
 
 - func: floor_divide.Scalar(Tensor self, Scalar other) -> Tensor

c10/test/build.bzl

@@ -66,6 +66,15 @@ def define_targets(rules):
         ],
     )
 
+    rules.cc_test(
+        name = "util/nofatal_test",
+        srcs = ["util/nofatal_test.cpp"],
+        deps = [
+            "//c10/util:base",
+            "@com_google_googletest//:gtest_main",
+        ],
+    )
+
     rules.cc_test(
         name = "util/ssize_test",
         srcs = ["util/ssize_test.cpp"],

c10/test/util/nofatal_test.cpp

@@ -0,0 +1,53 @@
+#include <gtest/gtest.h>
+
+#include <c10/util/Exception.h>
+#include <c10/util/Logging.h>
+
+namespace {
+template <typename T>
+inline void expectThrowsEq(T&& fn, const char* expected_msg) {
+  try {
+    std::forward<T>(fn)();
+  } catch (const c10::Error& e) {
+    EXPECT_TRUE(
+        std::string(e.what_without_backtrace()).find(expected_msg) !=
+        std::string::npos);
+    return;
+  }
+  ADD_FAILURE() << "Expected to throw exception with message \"" << expected_msg
+                << "\" but didn't throw";
+}
+} // namespace
+
+TEST(NofatalTest, TorchCheckComparisons) {
+  // quick make sure that no-op works as expected
+  TORCH_CHECK_EQ(1, 1) << "i am a silly message " << 1;
+  expectThrowsEq(
+      []() { TORCH_CHECK_EQ(1, 2) << "i am a silly message " << 1; },
+      "Check failed: 1 == 2 (1 vs. 2). i am a silly message 1");
+  expectThrowsEq(
+      []() { TORCH_CHECK_NE(2, 2); }, "Check failed: 2 != 2 (2 vs. 2).");
+  expectThrowsEq(
+      []() { TORCH_CHECK_LT(2, 2); }, "Check failed: 2 < 2 (2 vs. 2).");
+  expectThrowsEq(
+      []() { TORCH_CHECK_LE(3, 2); }, "Check failed: 3 <= 2 (3 vs. 2).");
+  expectThrowsEq(
+      []() { TORCH_CHECK_GT(2, 2); }, "Check failed: 2 > 2 (2 vs. 2).");
+  expectThrowsEq(
+      []() { TORCH_CHECK_GE(2, 3); }, "Check failed: 2 >= 3 (2 vs. 3).");
+  expectThrowsEq(
+      []() {
+        void* p = nullptr;
+        TORCH_CHECK_NOTNULL(p);
+      },
+      "Check failed: 'p' must be non NULL.");
+
+#if GTEST_HAS_DEATH_TEST
+#ifndef NDEBUG
+  // if dbg build, DCHECK should result in deth
+  EXPECT_DEATH(TORCH_DCHECK_EQ(1, 2), "Check failed");
+#else
+  TORCH_DCHECK_EQ(1, 2); // no-op
+#endif
+#endif // GTEST_HAS_DEATH_TEST
+}

c10/util/Exception.h

@@ -702,6 +702,98 @@ namespace c10::detail {
 #define TORCH_CHECK_ARG(cond, argN, ...) \
   TORCH_CHECK(cond, "invalid argument ", argN, ": ", __VA_ARGS__)
 
+#ifndef FATAL_IF
+#ifdef C10_USE_GLOG
+#define FATAL_IF(condition)                                              \
+  condition ? (void)0                                                    \
+            : ::c10::LoggerVoidify() &                                   \
+          ::c10::MessageLogger(__FILE__, __LINE__, ::google::GLOG_FATAL) \
+              .stream()
+#else
+#define FATAL_IF(condition)            \
+  condition ? (void)0                  \
+            : ::c10::LoggerVoidify() & \
+          ::c10::MessageLogger(__FILE__, __LINE__, ::c10::GLOG_FATAL).stream()
+#endif
+#endif
+
+#ifndef NON_FATAL_IF
+#ifdef C10_USE_GLOG
+#define NON_FATAL_IF(condition)                                \
+  condition ? (void)0                                          \
+            : ::c10::LoggerVoidify() &                         \
+          ::c10::MessageLogger(                                \
+              __FILE__, __LINE__, ::google::GLOG_FATAL, false) \
+              .stream()
+#else
+#define NON_FATAL_IF(condition)                                              \
+  condition ? (void)0                                                        \
+            : ::c10::LoggerVoidify() &                                       \
+          ::c10::MessageLogger(__FILE__, __LINE__, ::c10::GLOG_FATAL, false) \
+              .stream()
+#endif
+#endif
+
+// Binary comparison check macros
+#define TORCH_CHECK_OP(val1, val2, op)                                      \
+  NON_FATAL_IF(((val1)op(val2)))                                            \
+      << "Check failed: " #val1 " " #op " " #val2 " (" << (val1) << " vs. " \
+      << (val2) << "). "
+
+#define TORCH_DCHECK_OP(val1, val2, op)                                       \
+  FATAL_IF(((val1)op(val2))) << "Check failed: " #val1 " " #op " " #val2 " (" \
+                             << (val1) << " vs. " << (val2) << "). "
+
+#define TORCH_CHECK_EQ(val1, val2) TORCH_CHECK_OP(val1, val2, ==)
+#define TORCH_CHECK_NE(val1, val2) TORCH_CHECK_OP(val1, val2, !=)
+#define TORCH_CHECK_LE(val1, val2) TORCH_CHECK_OP(val1, val2, <=)
+#define TORCH_CHECK_LT(val1, val2) TORCH_CHECK_OP(val1, val2, <)
+#define TORCH_CHECK_GE(val1, val2) TORCH_CHECK_OP(val1, val2, >=)
+#define TORCH_CHECK_GT(val1, val2) TORCH_CHECK_OP(val1, val2, >)
+
+// Debug versions of TORCH_CHECK_OP macros
+#ifndef NDEBUG
+#define TORCH_DCHECK_EQ(val1, val2) TORCH_DCHECK_OP(val1, val2, ==)
+#define TORCH_DCHECK_NE(val1, val2) TORCH_DCHECK_OP(val1, val2, !=)
+#define TORCH_DCHECK_LE(val1, val2) TORCH_DCHECK_OP(val1, val2, <=)
+#define TORCH_DCHECK_LT(val1, val2) TORCH_DCHECK_OP(val1, val2, <)
+#define TORCH_DCHECK_GE(val1, val2) TORCH_DCHECK_OP(val1, val2, >=)
+#define TORCH_DCHECK_GT(val1, val2) TORCH_DCHECK_OP(val1, val2, >)
+#else // !NDEBUG
+// Optimized versions - generate no code
+#define TORCH_DCHECK_EQ(val1, val2) \
+  while (false)                     \
+  TORCH_DCHECK_OP(val1, val2, ==)
+#define TORCH_DCHECK_NE(val1, val2) \
+  while (false)                     \
+  TORCH_DCHECK_OP(val1, val2, !=)
+#define TORCH_DCHECK_LE(val1, val2) \
+  while (false)                     \
+  TORCH_DCHECK_OP(val1, val2, <=)
+#define TORCH_DCHECK_LT(val1, val2) \
+  while (false)                     \
+  TORCH_DCHECK_OP(val1, val2, <)
+#define TORCH_DCHECK_GE(val1, val2) \
+  while (false)                     \
+  TORCH_DCHECK_OP(val1, val2, >=)
+#define TORCH_DCHECK_GT(val1, val2) \
+  while (false)                     \
+  TORCH_DCHECK_OP(val1, val2, >)
+#endif // NDEBUG
+
+// Null pointer check macro
+#define TORCH_CHECK_NOTNULL(val) \
+  ::c10::CheckNotNull(__FILE__, __LINE__, #val, (val), false)
+
+#ifndef NDEBUG
+#define TORCH_DCHECK_NOTNULL(val) \
+  ::c10::CheckNotNull(__FILE__, __LINE__, #val, (val), true)
+#else // !NDEBUG
+#define TORCH_DCHECK_NOTNULL(val) \
+  while (false)                   \
+  TORCH_CHECK_NOTNULL(val)
+#endif // NDEBUG
+
 // ----------------------------------------------------------------------------
 // Deprecated macros
 // ----------------------------------------------------------------------------

c10/util/Logging.cpp

@@ -291,6 +291,32 @@ namespace c10 {
 using fLB::FLAGS_logtostderr;
 using fLI::FLAGS_minloglevel;
 using fLI::FLAGS_v;
+
+MessageLogger::MessageLogger(
+    const char* file,
+    int line,
+    int severity,
+    bool exit_on_fatal)
+    : stream_(), severity_(severity), exit_on_fatal_(exit_on_fatal) {}
+
+MessageLogger::~MessageLogger() noexcept(false) {
+  if (severity_ == ::google::GLOG_FATAL) {
+    DealWithFatal();
+  }
+}
+
+std::stringstream& MessageLogger::stream() {
+  return stream_;
+}
+
+void MessageLogger::DealWithFatal() {
+  if (exit_on_fatal_) {
+    LOG(FATAL) << stream_.str();
+  } else {
+    throw c10::Error(stream_.str(), nullptr, nullptr);
+  }
+}
+
 } // namespace c10
 
 C10_DEFINE_int(
@@ -412,17 +438,16 @@ void ShowLogInfoToStderr() {
   FLAGS_caffe2_log_level = GLOG_INFO;
 }
 
-MessageLogger::MessageLogger(const char* file, int line, int severity)
-    : severity_(severity) {
+MessageLogger::MessageLogger(
+    const char* file,
+    int line,
+    int severity,
+    bool exit_on_fatal)
+    : severity_(severity), exit_on_fatal_(exit_on_fatal) {
   if (severity_ < FLAGS_caffe2_log_level) {
     // Nothing needs to be logged.
     return;
   }
-#ifdef ANDROID
-  tag_ = "native";
-#else // !ANDROID
-  tag_ = "";
-#endif // ANDROID
 
   time_t rawtime = 0;
   time(&rawtime);
@@ -458,7 +483,7 @@ MessageLogger::MessageLogger(const char* file, int line, int severity)
 }
 
 // Output the contents of the stream to the proper channel on destruction.
-MessageLogger::~MessageLogger() {
+MessageLogger::~MessageLogger() noexcept(false) {
   if (severity_ < FLAGS_caffe2_log_level) {
     // Nothing needs to be logged.
     return;
@@ -498,6 +523,18 @@ MessageLogger::~MessageLogger() {
   }
 }
 
+std::stringstream& MessageLogger::stream() {
+  return stream_;
+}
+
+void MessageLogger::DealWithFatal() {
+  if (exit_on_fatal_) {
+    abort();
+  } else {
+    throw c10::Error(stream_.str(), nullptr, nullptr);
+  }
+}
+
 } // namespace c10
 
 #endif // !C10_USE_GLOG

c10/util/logging_common.h

@@ -0,0 +1,74 @@
+#ifndef C10_UTIL_LOGGING_COMMON_H_
+#define C10_UTIL_LOGGING_COMMON_H_
+
+#include <c10/macros/Export.h>
+#include <sstream>
+
+namespace c10 {
+
+// MessageLogger that throws exceptions instead of aborting (glog version)
+// or logs and may abort (non-glog version).
+class C10_API MessageLogger {
+ public:
+  MessageLogger(
+      const char* file,
+      int line,
+      int severity,
+      bool exit_on_fatal = true);
+  ~MessageLogger() noexcept(false);
+
+  // Return the stream associated with the logger object.
+  std::stringstream& stream();
+
+ private:
+  // When there is a fatal log, and fatal == true, we abort
+  // otherwise, we throw.
+  void DealWithFatal();
+
+#if defined(ANDROID) && !defined(C10_USE_GLOG)
+  const char* tag_{"native"};
+#endif
+  std::stringstream stream_;
+  int severity_;
+  bool exit_on_fatal_;
+};
+
+// This class is used to explicitly ignore values in the conditional
+// logging macros. This avoids compiler warnings like "value computed
+// is not used" and "statement has no effect".
+class C10_API LoggerVoidify {
+ public:
+  LoggerVoidify() = default;
+  // This has to be an operator with a precedence lower than << but
+  // higher than ?:
+  void operator&(const std::ostream& s [[maybe_unused]]) {}
+};
+
+// Forward declarations for CheckNotNull functions
+template <typename T>
+T& CheckNotNullCommon(
+    const char* file,
+    int line,
+    const char* names,
+    T& t,
+    bool fatal = true);
+
+template <typename T>
+T* CheckNotNull(
+    const char* file,
+    int line,
+    const char* names,
+    T* t,
+    bool fatal = true);
+
+template <typename T>
+T& CheckNotNull(
+    const char* file,
+    int line,
+    const char* names,
+    T& t,
+    bool fatal = true);
+
+} // namespace c10
+
+#endif // C10_UTIL_LOGGING_COMMON_H_

c10/util/logging_is_google_glog.h

@@ -47,57 +47,53 @@ INSTANTIATE_FOR_CONTAINER(set)
 
 #endif
 
+#include <c10/util/logging_common.h>
 #include <glog/logging.h>
 
-// Additional macros on top of glog
-#define TORCH_CHECK_EQ(val1, val2) CHECK_EQ(val1, val2)
-#define TORCH_CHECK_NE(val1, val2) CHECK_NE(val1, val2)
-#define TORCH_CHECK_LE(val1, val2) CHECK_LE(val1, val2)
-#define TORCH_CHECK_LT(val1, val2) CHECK_LT(val1, val2)
-#define TORCH_CHECK_GE(val1, val2) CHECK_GE(val1, val2)
-#define TORCH_CHECK_GT(val1, val2) CHECK_GT(val1, val2)
+namespace c10 {
 
-#ifndef NDEBUG
-#define TORCH_DCHECK_EQ(val1, val2) DCHECK_EQ(val1, val2)
-#define TORCH_DCHECK_NE(val1, val2) DCHECK_NE(val1, val2)
-#define TORCH_DCHECK_LE(val1, val2) DCHECK_LE(val1, val2)
-#define TORCH_DCHECK_LT(val1, val2) DCHECK_LT(val1, val2)
-#define TORCH_DCHECK_GE(val1, val2) DCHECK_GE(val1, val2)
-#define TORCH_DCHECK_GT(val1, val2) DCHECK_GT(val1, val2)
-#else // !NDEBUG
-// These versions generate no code in optimized mode.
-#define TORCH_DCHECK_EQ(val1, val2) \
-  while (false)                     \
-  DCHECK_EQ(val1, val2)
-#define TORCH_DCHECK_NE(val1, val2) \
-  while (false)                     \
-  DCHECK_NE(val1, val2)
-#define TORCH_DCHECK_LE(val1, val2) \
-  while (false)                     \
-  DCHECK_LE(val1, val2)
-#define TORCH_DCHECK_LT(val1, val2) \
-  while (false)                     \
-  DCHECK_LT(val1, val2)
-#define TORCH_DCHECK_GE(val1, val2) \
-  while (false)                     \
-  DCHECK_GE(val1, val2)
-#define TORCH_DCHECK_GT(val1, val2) \
-  while (false)                     \
-  DCHECK_GT(val1, val2)
-#endif // NDEBUG
+[[noreturn]] void ThrowEnforceNotMet(
+    const char* file,
+    const int line,
+    const char* condition,
+    const std::string& msg,
+    const void* caller);
 
-// Check that a pointer is not null.
-#define TORCH_CHECK_NOTNULL(val) CHECK_NOTNULL(val)
+template <typename T>
+T& CheckNotNullCommon(
+    const char* file,
+    int line,
+    const char* names,
+    T& t,
+    bool fatal) {
+  if (t == nullptr) {
+    MessageLogger(file, line, ::google::GLOG_FATAL, fatal).stream()
+        << "Check failed: '" << names << "' must be non NULL. ";
+  }
+  return t;
+}
 
-#ifndef NDEBUG
-// Debug only version of TORCH_CHECK_NOTNULL
-#define TORCH_DCHECK_NOTNULL(val) DCHECK_NOTNULL(val)
-#else // !NDEBUG
-// Optimized version - generates no code.
-#define TORCH_DCHECK_NOTNULL(val) \
-  while (false)                   \
-  DCHECK_NOTNULL(val)
-#endif // NDEBUG
+template <typename T>
+T* CheckNotNull(
+    const char* file,
+    int line,
+    const char* names,
+    T* t,
+    bool fatal) {
+  return CheckNotNullCommon(file, line, names, t, fatal);
+}
+
+template <typename T>
+T& CheckNotNull(
+    const char* file,
+    int line,
+    const char* names,
+    T& t,
+    bool fatal) {
+  return CheckNotNullCommon(file, line, names, t, fatal);
+}
+
+} // namespace c10
 
 // Log with source location information override (to be used in generic
 // warning/error handlers implemented as functions, not macros)

c10/util/logging_is_not_google_glog.h

@@ -13,6 +13,7 @@
 #include <vector>
 
 #include <c10/util/Flags.h>
+#include <c10/util/logging_common.h>
 
 const char CAFFE2_SEVERITY_PREFIX[] = "FEWIV";
 
@@ -24,61 +25,40 @@ const int GLOG_ERROR = 2;
 const int GLOG_WARNING = 1;
 const int GLOG_INFO = 0;
 
-class C10_API MessageLogger {
- public:
-  MessageLogger(const char* file, int line, int severity);
-  ~MessageLogger();
-  // Return the stream associated with the logger object.
-  std::stringstream& stream() {
-    return stream_;
-  }
-
- private:
-  // When there is a fatal log, we simply abort.
-  void DealWithFatal() {
-    abort();
-  }
-
-  const char* tag_;
-  std::stringstream stream_;
-  int severity_;
-};
-
-// This class is used to explicitly ignore values in the conditional
-// logging macros.  This avoids compiler warnings like "value computed
-// is not used" and "statement has no effect".
-class C10_API LoggerVoidify {
- public:
-  LoggerVoidify() = default;
-  // This has to be an operator with a precedence lower than << but
-  // higher than ?:
-  void operator&(const std::ostream& s [[maybe_unused]]) {}
-};
-
-// Log a message and terminate.
-template <class T>
-void LogMessageFatal(const char* file, int line, const T& message) {
-  MessageLogger(file, line, GLOG_FATAL).stream() << message;
-}
-
 // Helpers for TORCH_CHECK_NOTNULL(). Two are necessary to support both raw
 // pointers and smart pointers.
 template <typename T>
-T& CheckNotNullCommon(const char* file, int line, const char* names, T& t) {
+T& CheckNotNullCommon(
+    const char* file,
+    int line,
+    const char* names,
+    T& t,
+    bool fatal) {
   if (t == nullptr) {
-    LogMessageFatal(file, line, std::string(names));
+    MessageLogger(file, line, GLOG_FATAL, fatal).stream()
+        << "Check failed: '" << names << "' must be non NULL. ";
   }
   return t;
 }
 
 template <typename T>
-T* CheckNotNull(const char* file, int line, const char* names, T* t) {
-  return CheckNotNullCommon(file, line, names, t);
+T* CheckNotNull(
+    const char* file,
+    int line,
+    const char* names,
+    T* t,
+    bool fatal) {
+  return CheckNotNullCommon(file, line, names, t, fatal);
 }
 
 template <typename T>
-T& CheckNotNull(const char* file, int line, const char* names, T& t) {
-  return CheckNotNullCommon(file, line, names, t);
+T& CheckNotNull(
+    const char* file,
+    int line,
+    const char* names,
+    T& t,
+    bool fatal) {
+  return CheckNotNullCommon(file, line, names, t, fatal);
 }
 } // namespace c10
 
@@ -136,65 +116,6 @@ static_assert(
           ::c10::MessageLogger(__FILE__, __LINE__, ::c10::GLOG_##n).stream()
 #endif // NDEBUG
 
-#define TORCH_CHECK_OP(val1, val2, op)                                        \
-  FATAL_IF(((val1)op(val2))) << "Check failed: " #val1 " " #op " " #val2 " (" \
-                             << (val1) << " vs. " << (val2) << ") "
-
-// TORCH_CHECK_OP macro definitions
-#define TORCH_CHECK_EQ(val1, val2) TORCH_CHECK_OP(val1, val2, ==)
-#define TORCH_CHECK_NE(val1, val2) TORCH_CHECK_OP(val1, val2, !=)
-#define TORCH_CHECK_LE(val1, val2) TORCH_CHECK_OP(val1, val2, <=)
-#define TORCH_CHECK_LT(val1, val2) TORCH_CHECK_OP(val1, val2, <)
-#define TORCH_CHECK_GE(val1, val2) TORCH_CHECK_OP(val1, val2, >=)
-#define TORCH_CHECK_GT(val1, val2) TORCH_CHECK_OP(val1, val2, >)
-
-#ifndef NDEBUG
-// Debug only versions of TORCH_CHECK_OP macros.
-#define TORCH_DCHECK_EQ(val1, val2) TORCH_CHECK_OP(val1, val2, ==)
-#define TORCH_DCHECK_NE(val1, val2) TORCH_CHECK_OP(val1, val2, !=)
-#define TORCH_DCHECK_LE(val1, val2) TORCH_CHECK_OP(val1, val2, <=)
-#define TORCH_DCHECK_LT(val1, val2) TORCH_CHECK_OP(val1, val2, <)
-#define TORCH_DCHECK_GE(val1, val2) TORCH_CHECK_OP(val1, val2, >=)
-#define TORCH_DCHECK_GT(val1, val2) TORCH_CHECK_OP(val1, val2, >)
-#else // !NDEBUG
-// These versions generate no code in optimized mode.
-#define TORCH_DCHECK_EQ(val1, val2) \
-  while (false)                     \
-  TORCH_CHECK_OP(val1, val2, ==)
-#define TORCH_DCHECK_NE(val1, val2) \
-  while (false)                     \
-  TORCH_CHECK_OP(val1, val2, !=)
-#define TORCH_DCHECK_LE(val1, val2) \
-  while (false)                     \
-  TORCH_CHECK_OP(val1, val2, <=)
-#define TORCH_DCHECK_LT(val1, val2) \
-  while (false)                     \
-  TORCH_CHECK_OP(val1, val2, <)
-#define TORCH_DCHECK_GE(val1, val2) \
-  while (false)                     \
-  TORCH_CHECK_OP(val1, val2, >=)
-#define TORCH_DCHECK_GT(val1, val2) \
-  while (false)                     \
-  TORCH_CHECK_OP(val1, val2, >)
-#endif // NDEBUG
-
-// Check that a pointer is not null.
-#define TORCH_CHECK_NOTNULL(val) \
-  ::c10::CheckNotNull(           \
-      __FILE__, __LINE__, "Check failed: '" #val "' Must be non NULL", (val))
-
-#ifndef NDEBUG
-// Debug only version of TORCH_CHECK_NOTNULL
-#define TORCH_DCHECK_NOTNULL(val) \
-  ::c10::CheckNotNull(            \
-      __FILE__, __LINE__, "Check failed: '" #val "' Must be non NULL", (val))
-#else // !NDEBUG
-// Optimized version - generates no code.
-#define TORCH_DCHECK_NOTNULL(val) \
-  while (false)                   \
-  TORCH_CHECK_NOTNULL(val)
-#endif // NDEBUG
-
 // ---------------------- Support for std objects --------------------------
 // These are adapted from glog to support a limited set of logging capability
 // for STL objects.

setup.py

@@ -630,37 +630,6 @@ def mirror_files_into_torchgen() -> None:
         raise RuntimeError("Check the file paths in `mirror_files_into_torchgen()`")
 
 
-def mirror_inductor_external_kernels() -> None:
-    """
-    Copy external kernels into Inductor so they are importable.
-    """
-    paths = [
-        (
-            CWD / "torch/_inductor/kernel/vendored_templates/cutedsl_grouped_gemm.py",
-            CWD
-            / "third_party/cutlass/examples/python/CuTeDSL/blackwell/grouped_gemm.py",
-        ),
-    ]
-    for new_path, orig_path in paths:
-        # Create the dirs involved in new_path if they don't exist
-        if not new_path.exists():
-            new_path.parent.mkdir(parents=True, exist_ok=True)
-
-        # Copy the files from the orig location to the new location
-        if orig_path.is_file():
-            shutil.copyfile(orig_path, new_path)
-            continue
-        if orig_path.is_dir():
-            if new_path.exists():
-                # copytree fails if the tree exists already, so remove it.
-                shutil.rmtree(new_path)
-            shutil.copytree(orig_path, new_path)
-            continue
-        raise RuntimeError(
-            "Check the file paths in `mirror_inductor_external_kernels()`"
-        )
-
-
 # ATTENTION: THIS IS AI SLOP
 def extract_variant_from_version(version: str) -> str:
     """Extract variant from version string, defaulting to 'cpu'."""
@@ -1647,8 +1616,6 @@ def main() -> None:
     if RUN_BUILD_DEPS:
         build_deps()
 
-    mirror_inductor_external_kernels()
-
     (
         ext_modules,
         cmdclass,
@@ -1682,7 +1649,6 @@ def main() -> None:
         "_inductor/codegen/aoti_runtime/*.cpp",
         "_inductor/script.ld",
         "_inductor/kernel/flex/templates/*.jinja",
-        "_inductor/kernel/templates/*.jinja",
         "_export/serde/*.yaml",
         "_export/serde/*.thrift",
         "share/cmake/ATen/*.cmake",

test/distributed/_composable/test_replicate_with_fsdp.py

@@ -76,7 +76,7 @@ class ReplicateTest(MultiProcessTestCase):
             store=dist.FileStore(self.file_name, self.world_size),
         )
 
-    @skip_if_lt_x_gpu(2)
+    @skip_if_lt_x_gpu(4)
     def test_replicate_transformer(self):
         """
         This tests that replicate works on a transformer model with fully_shard and replicate layers
@@ -126,7 +126,7 @@ class ReplicateTest(MultiProcessTestCase):
                 for parameter in layer.parameters():
                     self.assertEqual(parameter.placements, (Shard(dim=0),))
 
-    @skip_if_lt_x_gpu(2)
+    @skip_if_lt_x_gpu(4)
     def test_replicate_transformer_managed_modules(self):
         """
         This tests that replicate managed modules works properly. In this test we use a Transformer Module with 3 layers,
@@ -178,7 +178,7 @@ class ReplicateTest(MultiProcessTestCase):
         replicate_model = replicate(replicate_model)
         self.assertEqual(len(_get_managed_modules((replicate_model,))), 21)
 
-    @skip_if_lt_x_gpu(2)
+    @skip_if_lt_x_gpu(4)
     def test_replicate_tp_device_mesh(self):
         """
         This tests that a user can pass in a device mesh to replicate a module
@@ -206,7 +206,7 @@ class ReplicateTest(MultiProcessTestCase):
                 self.assertEqual(parameter.device_mesh.shape, (2,))
                 self.assertEqual(parameter.placements, (Replicate(),))
 
-    @skip_if_lt_x_gpu(2)
+    @skip_if_lt_x_gpu(4)
     def test_train_replicate_fsdp(self):
         """
         Tests that replicate_model has the same behavior as original model when training
@@ -253,7 +253,7 @@ class ReplicateTest(MultiProcessTestCase):
             self.assertEqual(replicate_loss, loss)
             check_sharded_parity(self, model, replicate_model)
 
-    @skip_if_lt_x_gpu(2)
+    @skip_if_lt_x_gpu(4)
     def test_train_parity_2d_mlp(self):
         """
         Verifies when a device mesh is passed in, the model has the same behavior as the original model when training

test/distributed/checkpoint/test_dtensor_resharding.py

@@ -299,7 +299,7 @@ class TestDTensorReshardMeshChange(DTensorTestBase):
 
     @with_comms
     @with_temp_dir
-    @skip_if_lt_x_gpu(2)
+    @skip_if_lt_x_gpu(4)
     def test_dtensor_checkpoint_with_uneven_shards(self) -> None:
         """
         Saving a dtensor with uneven shards.
@@ -436,6 +436,7 @@ class TestCheckpointableReshard(DTensorTestBase):
 
     @with_comms
     @with_temp_dir
+    @skip_if_lt_x_gpu(4)
     def test_uneven_reshard_with_checkpointable_api(self) -> None:
         """
         Saves a 1d distributed tensor that has shards with uneven sizes using Checkpointable API.
@@ -498,6 +499,7 @@ class TestCheckpointableReshard(DTensorTestBase):
 
     @with_comms
     @with_temp_dir
+    @skip_if_lt_x_gpu(4)
     def test_uneven_reshard_with_dtensor_shards_wrapper_api(self) -> None:
         """
         Saves a 1d distributed tensor that has shards with uneven sizes using Checkpointable API.

test/distributed/checkpoint/test_state_dict.py

@@ -886,7 +886,7 @@ class TestStateDict(DTensorTestBase, VerifyStateDictMixin):
             self.assertEqual(cpu_model_value, meta_model_value)
 
     @with_comms
-    @skip_if_lt_x_gpu(2)
+    @skip_if_lt_x_gpu(4)
     def test_setting_meta_device_model_broadcasting_and_memory(self) -> None:
         # This test verifies that we can set model state dict by a meta device model
         # With the correlated changes in state_dict, meta device model should be accepted

test/distributed/pipelining/test_schedule_multiproc.py

@@ -39,6 +39,7 @@ from torch.nn.modules.loss import MSELoss
 from torch.testing._internal.common_distributed import (
     MultiProcContinuousTest,
     requires_accelerator_dist_backend,
+    skip_if_lt_x_gpu,
 )
 from torch.testing._internal.common_utils import (
     check_leaked_tensors,
@@ -231,6 +232,7 @@ class ScheduleTest(MultiProcContinuousTest):
         not TEST_MULTIACCELERATOR, f"{backend} test requires 2+ GPUs"
     )
     @parametrize("ScheduleClass", [_ScheduleForwardOnly])
+    @skip_if_lt_x_gpu(4)
     def test_forward_only(self, ScheduleClass):
         mod, mod_ref, x, _, _ = setup_models_and_data(self.config)
         x_clone = x.clone()
@@ -274,6 +276,7 @@ class ScheduleTest(MultiProcContinuousTest):
             ScheduleInterleavedZeroBubble,
         ],
     )
+    @skip_if_lt_x_gpu(4)
     def test_eval_inference_mode(self, ScheduleClass):
         num_microbatches = 4
         if ScheduleClass in [
@@ -351,6 +354,7 @@ class ScheduleTest(MultiProcContinuousTest):
             ScheduleInterleavedZeroBubble,
         ],
     )
+    @skip_if_lt_x_gpu(4)
     def test_return_output(self, ScheduleClass):
         num_microbatches = 4
         if ScheduleClass in [
@@ -406,6 +410,7 @@ class ScheduleTest(MultiProcContinuousTest):
         not TEST_MULTIACCELERATOR, f"{backend} test requires 2+ GPUs"
     )
     @parametrize("ScheduleClass", [ScheduleGPipe, Schedule1F1B])
+    @skip_if_lt_x_gpu(4)
     def test_multi_iter(self, ScheduleClass):
         mod, _, x, target, loss_fn = setup_models_and_data(self.config)
         chunks = 4
@@ -429,6 +434,7 @@ class ScheduleTest(MultiProcContinuousTest):
         not TEST_MULTIACCELERATOR, f"{backend} test requires 2+ GPUs"
     )
     @parametrize("ScheduleClass", [ScheduleGPipe, Schedule1F1B])
+    @skip_if_lt_x_gpu(4)
     def test_kwargs_with_tracer(self, ScheduleClass):
         mod = ModelWithKwargs(d_hid, splits=self.world_size)
         mod.to(self.device)
@@ -481,6 +487,7 @@ class ScheduleTest(MultiProcContinuousTest):
         not TEST_MULTIACCELERATOR, f"{backend} test requires 2+ GPUs"
     )
     @parametrize("ScheduleClass", [ScheduleGPipe, Schedule1F1B])
+    @skip_if_lt_x_gpu(4)
     def test_grad_with_tracer(self, ScheduleClass):
         mod, ref_mod, x, target, loss_fn = setup_models_and_data(self.config)
 
@@ -523,6 +530,7 @@ class ScheduleTest(MultiProcContinuousTest):
     )
     @parametrize("ScheduleClass", [ScheduleGPipe, Schedule1F1B])
     @parametrize("shape_inference", [True, False])
+    @skip_if_lt_x_gpu(4)
     def test_grad_with_manual(self, ScheduleClass, shape_inference):
         mod, ref_mod, x, target, loss_fn = setup_models_and_data(self.config)
 
@@ -586,6 +594,7 @@ class ScheduleTest(MultiProcContinuousTest):
             ScheduleInterleavedZeroBubble,
         ],
     )
+    @skip_if_lt_x_gpu(4)
     def test_grad_with_manual_interleaved(self, ScheduleClass):
         stages_per_rank = 2
         n_stages = stages_per_rank * self.world_size
@@ -650,6 +659,7 @@ class ScheduleTest(MultiProcContinuousTest):
         not TEST_MULTIACCELERATOR, f"{backend} test requires 2+ GPUs"
     )
     @parametrize("ScheduleClass", [ScheduleInterleavedZeroBubble])
+    @skip_if_lt_x_gpu(4)
     def test_schedule_with_weight_update_mlp_e2e(self, ScheduleClass):
         stages_per_rank = 2
         n_stages = stages_per_rank * self.world_size
@@ -736,6 +746,7 @@ class ScheduleTest(MultiProcContinuousTest):
         "schedule_class",
         [ScheduleZBVZeroBubble, ScheduleDualPipeV],
     )
+    @skip_if_lt_x_gpu(4)
     def test_v_shape_schedules(self, schedule_class):
         n_stages = 8
         rank_stages = {0: [0, 7], 1: [1, 6], 2: [2, 5], 3: [3, 4]}
@@ -780,6 +791,7 @@ class ScheduleTest(MultiProcContinuousTest):
     @skip_but_pass_in_sandcastle_if(
         not TEST_MULTIACCELERATOR, f"{backend} test requires 2+ GPUs"
     )
+    @skip_if_lt_x_gpu(4)
     def test_custom_function_callback(self):
         """Test the custom function callback functionality with _PipelineScheduleRuntime."""
         n_stages = 8
@@ -979,6 +991,7 @@ class ScheduleTest(MultiProcContinuousTest):
         "ScheduleClass",
         [ScheduleInterleavedZeroBubble, ScheduleInterleaved1F1B],
     )
+    @skip_if_lt_x_gpu(4)
     def test_zero_bubble_with_model_kwargs(self, ScheduleClass):
         stages_per_rank = 2
         n_stages = stages_per_rank * self.world_size
@@ -1072,6 +1085,7 @@ class CustomSchedulesTest(MultiProcContinuousTest):
         "schedule_class",
         [ScheduleVShaped, ScheduleUnbalanced],
     )
+    @skip_if_lt_x_gpu(4)
     def test_non_symmetric_stage_ids(self, schedule_class):
         n_stages = schedule_class.n_stages
         rank_stages = schedule_class.rank_stages
@@ -1121,6 +1135,7 @@ class CustomSchedulesTest(MultiProcContinuousTest):
         not TEST_MULTIACCELERATOR, f"{backend} test requires 2+ GPUs"
     )
     @parametrize("ScheduleClass", [ScheduleWithReorderedB])
+    @skip_if_lt_x_gpu(4)
     def test_pipeline_schedule_runtime_custom_sched(self, ScheduleClass):
         n_stages = 2
         stages_per_rank = 1
@@ -1181,6 +1196,7 @@ class CustomSchedulesTest(MultiProcContinuousTest):
         not TEST_MULTIACCELERATOR, f"{backend} test requires 2+ GPUs"
     )
     @parametrize("ScheduleClass", [ScheduleWithW])
+    @skip_if_lt_x_gpu(4)
     def test_schedule_with_native_zero_bubble(self, ScheduleClass):
         n_stages = ScheduleClass.n_stages
         num_microbatches = ScheduleClass.num_microbatches

test/distributed/tensor/test_math_ops.py

@@ -26,6 +26,7 @@ from torch.distributed.tensor.parallel import (
     RowwiseParallel,
     SequenceParallel,
 )
+from torch.testing._internal.common_distributed import skip_if_lt_x_gpu
 from torch.testing._internal.common_utils import run_tests
 from torch.testing._internal.distributed._tensor.common_dtensor import (
     create_local_tensor_test_class,
@@ -764,6 +765,7 @@ class DistMathOpsTest(DTensorTestBase):
         self.assertEqual(grad1_norm.device_mesh, mesh_y)
 
     @with_comms
+    @skip_if_lt_x_gpu(4)
     def test_foreach_add_different_mesh(self):
         mesh_shape = (2, self.world_size // 2)
         mesh_2d = init_device_mesh(

test/distributed/tensor/test_op_strategy.py

@@ -577,7 +577,7 @@ class DistTensorReplicateStrategyRegistrationTest(DTensorTestBase):
                 self.assertEqual(
                     comm_mode.get_comm_counts(),
                     {
-                        torch.ops.c10d_functional.all_gather_into_tensor: 4,
+                        torch.ops.c10d_functional.all_gather_into_tensor: self.world_size,
                     },
                 )
                 expected_cost = [

test/distributed/test_functional_api.py

@@ -485,7 +485,7 @@ elif TEST_XPU:
 def exit_if_lt_x_accelerators(x):
     if torch.accelerator.is_available():
         if torch.accelerator.device_count() < x:
-            sys.exit(TEST_SKIPS[f"multi-accelerator-{x}"].exit_code)
+            sys.exit(TEST_SKIPS[f"multi-gpu-{x}"].exit_code)
 
 
 def with_comms(func=None):

test/dynamo/test_decorators.py

@@ -2109,6 +2109,52 @@ Detected recompile when torch.compile stance is 'fail_on_recompile'. filename: '
         with self.assertRaises(Unsupported):
             outer_f2(inp)
 
+    def test_disable_recursive_flags(self):
+        class SimpleLinear(torch.nn.Module):
+            def __init__(self) -> None:
+                super().__init__()
+                self.layer0 = torch.nn.Linear(4, 4)
+
+            def forward(self, inp):
+                return self.layer0(torch.sigmoid(inp))
+
+        class SimpleModel(torch.nn.Module):
+            def __init__(self) -> None:
+                super().__init__()
+                self.layer0 = SimpleLinear()
+                self.layer1 = torch.nn.Linear(4, 4)
+
+            def forward(self, inp):
+                z = self.layer0(torch.sin(inp))
+                return self.layer1(z)
+
+        for recursive_flag in [True, False]:
+            model = SimpleModel()
+            other_model = SimpleModel()
+
+            model.forward = torch._dynamo.disable(
+                model.forward,
+                recursive=recursive_flag,
+            )
+            self.assertEqual(
+                torch._dynamo.is_dynamo_disable_recursive(model.forward),
+                recursive_flag,
+            )
+
+            other_model = torch._dynamo.disable(other_model, recursive=recursive_flag)
+            self.assertEqual(
+                torch._dynamo.is_dynamo_disable_recursive(
+                    other_model.forward
+                    if isinstance(other_model, torch.nn.Module)
+                    else other_model
+                ),
+                recursive_flag,
+            )
+
+            # check the model is compilable
+            torch.compile(model)
+            torch.compile(other_model)
+
 
 if __name__ == "__main__":
     from torch._dynamo.test_case import run_tests

test/dynamo/test_error_messages.py

@@ -422,34 +422,41 @@ from user code:
         import optree
 
         @torch.compile(backend="eager")
-        def fn(x):
-            d = {"a": 1}
-            optree.tree_flatten_with_path(d)
-            return torch.sin(x)
-
-        def post_munge(s):
-            s = re.sub(
-                r"optree\.\S*\.flatten_with_path",
-                "optree.<path>.flatten_with_path",
-                s,
-            )
-            return re.sub(
-                r"qualname: \S*flatten_with_path",
-                "qualname: <path>.flatten_with_path",
-                s,
+        def fn1(x):
+            tree = {"a": x, "b": (x - 1, 2 * x)}
+            sin, cos = optree.tree_transpose_map(
+                lambda t: (torch.sin(t), torch.cos(t)),
+                tree,
             )
+            return sin, cos
 
-        fn(torch.randn(4))
-        self.assertEqual(len(counters["graph_break"]), 1)
+        fn1(torch.randn(4))
+        self.assertEqual(len(counters["graph_break"]), 0)
+
+        @torch.compile(backend="eager")
+        def fn2(x):
+            spec = optree.treespec_deque([])
+            return spec, x
+
+        fn2(torch.randn(4))
+        self.assertGreaterEqual(len(counters["graph_break"]), 1)
         first_graph_break = next(iter(counters["graph_break"].keys()))
+
+        def post_munge(string):
+            return re.sub(
+                r"(optree\.|qualname: )\S*(\.make_from_collection)",
+                r"\1<path>\2",
+                string,
+            )
+
         self.assertExpectedInline(
             post_munge(first_graph_break),
             """\
 Attempted to call function marked as skipped
-  Explanation: Dynamo cannot trace optree C/C++ function optree.<path>.flatten_with_path.
+  Explanation: Dynamo cannot trace optree C/C++ function optree.<path>.make_from_collection.
   Hint: Consider using torch.utils._pytree - https://github.com/pytorch/pytorch/blob/main/torch/utils/_pytree.py
 
-  Developer debug context: module: optree._C, qualname: <path>.flatten_with_path, skip reason: <missing reason>
+  Developer debug context: module: optree._C, qualname: <path>.make_from_collection, skip reason: <missing reason>
 
  For more details about this graph break, please visit: https://meta-pytorch.github.io/compile-graph-break-site/gb/gb0007.html""",
         )

test/inductor/test_cpu_repro.py

@@ -3278,6 +3278,15 @@ class CPUReproTests(TestCase):
         metrics.reset()
         self.common(fn, (x,))
 
+    def test_softmax_with_zero_dim(self):
+        def fn(x):
+            x = torch.softmax(x, 0)
+            return x
+
+        x = torch.rand([], dtype=torch.bfloat16)
+        metrics.reset()
+        self.common(fn, (x,))
+
     @config.patch({"fx_graph_cache": False, "fx_graph_remote_cache": False})
     def test_local_buffer_in_outer_loop_fusion(self):
         def fn(x):

test/inductor/test_cutedsl_grouped_mm.py

@@ -1,154 +0,0 @@
-# Owner(s): ["module: inductor"]
-
-
-import unittest
-
-import torch
-from torch import Tensor
-from torch._inductor import config
-from torch._inductor.codegen.cuda.cuda_env import is_datacenter_blackwell_arch
-from torch._inductor.test_case import run_tests, TestCase as InductorTestCase
-from torch._inductor.utils import ensure_cute_available
-from torch.testing._internal.common_utils import (
-    instantiate_parametrized_tests,
-    parametrize,
-)
-
-
-@unittest.skipIf(
-    not (ensure_cute_available() and is_datacenter_blackwell_arch()),
-    "CuTeDSL library or Blackwell device not available",
-)
-@instantiate_parametrized_tests
-class TestCuTeDSLGroupedGemm(InductorTestCase):
-    def _get_inputs(
-        self,
-        group_size: int,
-        M_hint: int,
-        K: int,
-        N: int,
-        device: str,
-        dtype: torch.dtype,
-        alignment: int = 16,
-    ) -> tuple[Tensor, Tensor, Tensor]:
-        # --- Random, tile-aligned M sizes ---
-        M_sizes = (
-            torch.randint(1, (M_hint // alignment) + 1, (group_size,), dtype=torch.int)
-            * alignment
-        )
-
-        M_total = torch.sum(M_sizes).item()
-
-        # --- Construct input tensors ---
-        A = torch.randn(int(M_total), K, dtype=dtype, device=device) * 0.1
-        B = torch.randn((group_size, K, N), dtype=dtype, device=device) * 0.01
-
-        # --- Build offsets (no leading zero, strictly increasing) ---
-        offsets = torch.cumsum(M_sizes, dim=0).to(dtype=torch.int32, device=device)
-
-        return (A, B, offsets)
-
-    @parametrize("group_size", (2, 8))
-    @parametrize("M_hint", (256, 1024))
-    @parametrize("K", (64, 128))
-    @parametrize("N", (128, 256))
-    def test_grouped_gemm_basic(self, group_size: int, M_hint: int, K: int, N: int):
-        device = "cuda"
-        dtype = torch.bfloat16
-
-        A, B, offsets = self._get_inputs(group_size, M_hint, K, N, device, dtype)
-
-        def grouped_gemm_fn(A_packed, B_batched, offs):
-            return torch._grouped_mm(A_packed, B_batched, offs=offs)
-
-        # Eager execution
-        c_eager = grouped_gemm_fn(A, B, offsets)
-
-        # Test with Cute backend
-        with config.patch(
-            {
-                "max_autotune": True,
-                "max_autotune_gemm_backends": "CUTEDSL",
-                "test_configs.autotune_choice_name_regex": "cutedsl",
-                "autotune_fallback_to_aten": False,
-            }
-        ):
-            grouped_gemm_compiled = torch.compile(
-                grouped_gemm_fn, backend="inductor", dynamic=False
-            )
-            c_compiled = grouped_gemm_compiled(A, B, offsets)
-
-        self.assertEqual(c_eager.dtype, dtype)
-        self.assertEqual(c_compiled.dtype, dtype)
-        torch.testing.assert_close(c_eager, c_compiled)
-
-    @parametrize("layout_A", ("contiguous", "offset", "padded", "view"))
-    @parametrize("layout_B", ("contiguous", "broadcasted"))
-    def test_grouped_gemm_assorted_layouts(
-        self,
-        layout_A: str,
-        layout_B: str,
-    ):
-        device = "cuda"
-        dtype = torch.bfloat16
-
-        G, K, N = 8, 64, 128
-        M_sizes = [128] * G
-        sum_M = sum(M_sizes)
-        offsets = torch.tensor(
-            [sum(M_sizes[: i + 1]) for i in range(G)], dtype=torch.int32, device=device
-        )
-
-        A_base = torch.randn(sum_M, K, device=device, dtype=dtype)
-        A = A_base
-
-        if layout_A == "offset":
-            # allocate bigger buffer than needed, use nonzero storage offset
-            storage = torch.randn(sum_M * K + 512, device=device, dtype=dtype)
-            offset = 128  # skip first 128 elements
-            A = torch.as_strided(storage[offset:], (sum_M, K), (K, 1))
-        elif layout_A == "padded":
-            # simulate row pitch > K (row_stride = K + pad)
-            row_pitch = K + 8
-            storage = torch.randn(sum_M * row_pitch, device=device, dtype=dtype)
-            A = torch.as_strided(storage, (sum_M, K), (row_pitch, 1))
-        elif layout_A == "view":
-            A_storage = torch.randn(sum_M * K, device=device, dtype=dtype)
-            A = A_storage.view(sum_M, K)
-            assert A._base is not None
-            assert A.shape == (sum_M, K)
-
-        B = torch.randn((G, K, N), dtype=dtype, device=device) * 0.01
-
-        if layout_B == "broadcasted":
-            # Broadcast B across groups (zero stride along G)
-            B = B[0].expand(G, K, N)
-            assert B.stride(0) == 0
-
-        def grouped_gemm_fn(A_packed, B_batched, offs):
-            return torch._grouped_mm(A_packed, B_batched, offs=offs)
-
-        # --- eager ---
-        c_eager = grouped_gemm_fn(A, B, offsets)
-
-        # --- compiled (CUTE backend) ---
-        with config.patch(
-            {
-                "max_autotune": True,
-                "max_autotune_gemm_backends": "CUTEDSL",
-                "test_configs.autotune_choice_name_regex": "cutedsl",
-                "autotune_fallback_to_aten": False,
-            }
-        ):
-            grouped_gemm_compiled = torch.compile(
-                grouped_gemm_fn, backend="inductor", dynamic=False
-            )
-            c_compiled = grouped_gemm_compiled(A, B, offsets)
-
-        self.assertEqual(c_eager.dtype, dtype)
-        self.assertEqual(c_compiled.dtype, dtype)
-        torch.testing.assert_close(c_eager, c_compiled)
-
-
-if __name__ == "__main__":
-    run_tests()

test/inductor/test_max_autotune.py

@@ -1913,6 +1913,29 @@ class TestMaxAutotune(TestCase):
             # Check that contiguous transform was used
             FileCheck().check("contiguous_mm").run(code[0])
 
+    @unittest.skipIf(config.cpp_wrapper, "out_dtype override not supported for AOTI")
+    @unittest.skipIf(TEST_WITH_ROCM, "out_dtype override only available on NVIDIA")
+    def test_bmm_out_dtype(self):
+        def f(a, b):
+            return torch.bmm(a, b, out_dtype=torch.float32)
+
+        a = torch.randn(2, 3, 4, device=GPU_TYPE, dtype=torch.float16)
+        b = torch.randn(2, 4, 5, device=GPU_TYPE, dtype=torch.float16)
+        with config.patch(
+            max_autotune=True,
+            max_autotune_gemm_backends="TRITON",
+        ):
+            compiled_f = torch.compile(f)
+            with self.assertRaisesRegex(
+                torch._inductor.exc.InductorError,
+                r"LoweringException: NoValidChoicesError: No choices to select",
+            ):
+                out, code = run_and_get_code(compiled_f, a, b)
+
+        compiled_f = torch.compile(f)
+        out, code = run_and_get_code(compiled_f, a, b)
+        FileCheck().check("extern_kernels.bmm_dtype").run(code[0])
+
     def test_triton_template_generated_code_cache_key(self):
         generate_and_load_args = len(
             inspect.signature(

test/inductor/test_pattern_matcher.py

@@ -1217,6 +1217,43 @@ class TestPatternMatcher(TestCase):
         _, (code) = run_and_get_code(fn2, args[0], args[1], args[2])
         FileCheck().check_not("extern_kernels.addmm(").run(code[0])
 
+    def test_addmm_alpha_beta_with_pointwise(self):
+        # Test that addmm with alpha/beta != 1 is unfused correctly with pointwise ops
+        # See https://github.com/pytorch/pytorch/issues/167313
+        x = torch.rand(2, device=GPU_TYPE)
+        a = torch.rand(2, 3, device=GPU_TYPE)
+        b = torch.rand(3, 2, device=GPU_TYPE)
+
+        def f(x, a, b):
+            return torch.nn.functional.relu(torch.addmm(x, a, b, alpha=0.8, beta=0.2))
+
+        fc = torch.compile(f)
+
+        expected = f(x, a, b)
+        actual = fc(x, a, b)
+
+        # The compiled version should produce the same result as eager
+        torch.testing.assert_close(actual, expected)
+
+        # Verify that addmm is unfused (should not use extern_kernels.addmm)
+        # The pattern should be replaced with beta * x + alpha * (a @ b)
+        _, (code) = run_and_get_code(fc, x, a, b)
+        FileCheck().check_not("extern_kernels.addmm(").run(code[0])
+
+        # Test with alpha=1, beta=1 (default) - should also unfuse
+        def f_default(x, a, b):
+            return torch.nn.functional.relu(torch.addmm(x, a, b))
+
+        fc_default = torch.compile(f_default)
+        expected_default = f_default(x, a, b)
+        actual_default = fc_default(x, a, b)
+
+        torch.testing.assert_close(actual_default, expected_default)
+
+        # Should unfuse and not use extern_kernels.addmm
+        _, (code) = run_and_get_code(fc_default, x, a, b)
+        FileCheck().check_not("extern_kernels.addmm(").run(code[0])
+
     def test_serialized_patterns_up_to_date(self):
         import torch.utils._pytree as pytree
         from torch._inductor.fx_passes import joint_graph

test/test_cuda.py

@@ -7486,7 +7486,7 @@ class TestFXMemoryProfiler(TestCase):
         return fx_frames
 
     @unittest.skipIf(not torch.cuda.is_available(), "CUDA not available")
-    @torch._dynamo.config.patch("enrich_profiler_metadata", True)
+    @torch.fx.experimental._config.patch("enrich_profiler_metadata", True)
     def test_fx_memory_profiler_augmentation(self):
         """Test that memory snapshots are augmented with FX debug information."""
 

test/test_fx.py

@@ -4251,7 +4251,7 @@ def forward(self, args_list: List[torch.Tensor]){maybe_return_annotation}:
 
     @unittest.skipIf(not torch.cuda.is_available(), "CUDA not available")
     @skipIfRocm
-    @torch._dynamo.config.patch("enrich_profiler_metadata", True)
+    @torch.fx.experimental._config.patch("enrich_profiler_metadata", True)
     def test_profiler_stack_trace_augmentation(self):
         """
         Test that map_recorded_events_to_aten_ops_with_stack_trace correctly
@@ -4307,7 +4307,7 @@ event=cudaLaunchKernel node=addmm_1 stack_trace=x = self.linear2(x)"""
 
     @unittest.skipIf(not torch.cuda.is_available(), "CUDA not available")
     @skipIfRocm
-    @torch._dynamo.config.patch("enrich_profiler_metadata", True)
+    @torch.fx.experimental._config.patch("enrich_profiler_metadata", True)
     def test_profiler_multiple_modules(self):
         """
         Test that multiple compiled modules under the same profiler session
@@ -4351,7 +4351,7 @@ event=cudaLaunchKernel node=sub stack_trace=return x - 1"""
 
     @unittest.skipIf(not torch.cuda.is_available(), "CUDA not available")
     @skipIfRocm
-    @torch._dynamo.config.patch("enrich_profiler_metadata", True)
+    @torch.fx.experimental._config.patch("enrich_profiler_metadata", True)
     def test_profiler_nested_graph_modules(self):
         """
         Test that nested graph modules (e.g., graph modules calling subgraphs)

tools/stats/upload_test_stats.py

@@ -165,21 +165,6 @@ def get_tests(workflow_run_id: int, workflow_run_attempt: int) -> list[dict[str,
         return flattened
 
 
-def get_tests_for_circleci(
-    workflow_run_id: int, workflow_run_attempt: int
-) -> list[dict[str, Any]]:
-    # Parse the reports and transform them to JSON
-    test_cases = []
-    for xml_report in Path(".").glob("**/test/test-reports/**/*.xml"):
-        test_cases.extend(
-            parse_xml_report(
-                "testcase", xml_report, workflow_run_id, workflow_run_attempt
-            )
-        )
-
-    return test_cases
-
-
 def summarize_test_cases(test_cases: list[dict[str, Any]]) -> list[dict[str, Any]]:
     """Group test cases by classname, file, and job_id. We perform the aggregation
     manually instead of using the `test-suite` XML tag because xmlrunner does
@@ -258,21 +243,11 @@ if __name__ == "__main__":
         required=True,
         help="Head repository of the workflow",
     )
-    parser.add_argument(
-        "--circleci",
-        action="store_true",
-        help="If this is being run through circleci",
-    )
     args = parser.parse_args()
 
     print(f"Workflow id is: {args.workflow_run_id}")
 
-    if args.circleci:
-        test_cases = get_tests_for_circleci(
-            args.workflow_run_id, args.workflow_run_attempt
-        )
-    else:
-        test_cases = get_tests(args.workflow_run_id, args.workflow_run_attempt)
+    test_cases = get_tests(args.workflow_run_id, args.workflow_run_attempt)
 
     # Flush stdout so that any errors in the upload show up last in the logs.
     sys.stdout.flush()
@@ -313,12 +288,4 @@ if __name__ == "__main__":
             remove_nan_inf(test_cases),
         )
 
-    # Part of an experiment to see if we can handle all the data as is
-    upload_workflow_stats_to_s3(
-        args.workflow_run_id,
-        args.workflow_run_attempt,
-        "all_test_runs",
-        remove_nan_inf(test_cases),
-    )
-
     upload_additional_info(args.workflow_run_id, args.workflow_run_attempt, test_cases)

torch/_dynamo/__init__.py

@@ -32,6 +32,7 @@ from .decorators import (
     error_on_graph_break,
     forbid_in_graph,
     graph_break,
+    is_dynamo_disable_recursive,
     mark_dynamic,
     mark_static,
     mark_static_address,
@@ -87,6 +88,7 @@ __all__ = [
     "forbid_in_graph",
     "graph_break",
     "is_compiling",
+    "is_dynamo_disable_recursive",
     "list_backends",
     "lookup_backend",
     "mark_dynamic",

torch/_dynamo/codegen.py

@@ -15,7 +15,7 @@ import dataclasses
 import re
 import sys
 import types
-from collections import Counter
+from collections import Counter, deque
 from collections.abc import Callable, Iterable
 from typing import Any, Optional, TYPE_CHECKING, Union
 
@@ -597,32 +597,35 @@ class PyCodegen:
 
         graphargs = self.tx.output.graphargs
 
-        seen_sources: OrderedSet[Source] = OrderedSet()
-
-        def collect_temp_source(source: Source) -> None:
-            if source in seen_sources:
-                # This source is used at least twice, so it can be reused
-                self.mark_source_temp(source)
-                # Dont trace source further. This prevents us from marking too
-                # many nodes as temp sources.
-                return
-
-            seen_sources.add(source)
-
+        def extract_nested_sources(source: Source) -> list[Source]:
+            nested_sources: list[Source] = []
             if isinstance(source, ChainedSource):
-                collect_temp_source(source.base)
-
+                nested_sources.append(source.base)
             if isinstance(source, DictGetItemSource) and isinstance(
                 source.index, Source
             ):
-                collect_temp_source(source.index)
+                nested_sources.append(source.index)
+            return nested_sources
+
+        def collect_temp_sources(sources: deque[Source], codegen: PyCodegen) -> None:
+            seen_sources: OrderedSet[Source] = OrderedSet()
+            while sources:
+                current_source = sources.popleft()
+                if current_source in seen_sources:
+                    # This source is used at least twice, so it can be reused
+                    codegen.mark_source_temp(current_source)
+                    # Dont trace source further. This prevents us from marking too
+                    # many nodes as temp sources.
+                    continue
+                seen_sources.add(current_source)
+                sources.extend(extract_nested_sources(current_source))
 
         # Collect all the sources that are used more than once, so that we can
         # generate tmp variables in the generated pre-graph bytecode. This
         # essentially implements CSE.
-        for arg in graphargs:
-            if arg.source is not None:
-                collect_temp_source(arg.source)
+        collect_temp_sources(
+            deque([arg.source for arg in graphargs if arg.source is not None]), self
+        )
 
         cm_var = None
         if config.record_runtime_overhead:

torch/_dynamo/config.py

@@ -740,11 +740,8 @@ enable_aot_compile = False
 # HACK: this is for testing custom ops profiling only
 _custom_ops_profile: Optional[Any] = None
 
-# Experimental: If True, graph module will register fx metadata during recompile()
-enrich_profiler_metadata: bool = Config(  # type: ignore[var-annotated]
-    default=False,
-    env_name_default="TORCH_ENRICH_RPOFILER_STACK_TRACE",
-)
+# Deprecated! Please use the config in torch/fx/experimental/_config instead.
+enrich_profiler_metadata: bool = False
 
 if TYPE_CHECKING:
     from torch.utils._config_typing import *  # noqa: F401, F403

torch/_dynamo/convert_frame.py

@@ -828,6 +828,7 @@ def trace_frame(
             raise
         finally:
             tracer.output.call_cleanup_hooks()
+            tracer.f_locals = {}
 
     try:
         run_tracer()

torch/_dynamo/decorators.py

@@ -96,6 +96,7 @@ def disable(fn=None, recursive=True, *, reason=None, wrapping=True):  # type: ig
             nonrecursive_disable_wrapper._torchdynamo_disable = True  # type: ignore[attr-defined]
             nonrecursive_disable_wrapper._torchdynamo_disable_msg = reason  # type: ignore[attr-defined]
             nonrecursive_disable_wrapper._torchdynamo_orig_callable = fn  # type: ignore[attr-defined]
+            nonrecursive_disable_wrapper._torchdynamo_disable_recursive = False  # type: ignore[attr-defined]
             # pyrefly: ignore [bad-return]
             return nonrecursive_disable_wrapper
 
@@ -1023,3 +1024,13 @@ def error_on_graph_break(
     The default value of torch.compile's `error_on_graph_break` setting is False.
     """
     return ErrorOnGraphBreakDecoratorContextManager(error_on_graph_break)
+
+
+def is_dynamo_disable_recursive(method: Callable[[Any], Any]) -> Optional[bool]:
+    """
+    Check if a method is marked as `dynamo_disable` recursively. It returns:
+    - True if disable(recursive=True)
+    - False if disable(recursive=False)
+    - None if method is not a disable decorator
+    """
+    return getattr(method, "_torchdynamo_disable_recursive", None)

torch/_dynamo/eval_frame.py

@@ -1155,6 +1155,8 @@ class DisableContext(_TorchDynamoContext):
         # of decorators.
         _fn._torchdynamo_orig_callable = fn  # type: ignore[attr-defined]
 
+        _fn._torchdynamo_disable_recursive = True  # type: ignore[attr-defined]
+
         return _fn
 
     def __reduce__(self) -> tuple[type[DisableContext], tuple[Any, ...]]:

torch/_dynamo/polyfills/loader.py

@@ -4,6 +4,8 @@
 import importlib
 from typing import TYPE_CHECKING
 
+import torch.utils._pytree as python_pytree
+
 from .. import polyfills, trace_rules
 
 
@@ -19,12 +21,14 @@ POLYFILLED_MODULE_NAMES: tuple[str, ...] = (
     "itertools",
     "operator",
     "os",
-    "pytree",
     "struct",
     "sys",
     "fx",
     "tensor",
 )
+if python_pytree._cxx_pytree_dynamo_traceable:
+    POLYFILLED_MODULE_NAMES += ("pytree",)
+
 POLYFILLED_MODULES: tuple["ModuleType", ...] = tuple(
     importlib.import_module(f".{submodule}", package=polyfills.__name__)
     for submodule in POLYFILLED_MODULE_NAMES

torch/_dynamo/variables/builtin.py

@@ -1991,7 +1991,7 @@ class BuiltinVariable(VariableTracker):
             # If the object implements a __getitem__ method, iter(...) will call obj.__getitem__()
             # with an integer argument starting at 0, until __getitem__ raises IndexError
             ret = variables.UserFunctionVariable(
-                polyfills.builtins.iter_  # type: ignore[arg-type]
+                polyfills.builtins.iter_
             ).call_function(tx, [obj, *args], {})
 
             if args:

torch/_dynamo/variables/iter.py

@@ -590,7 +590,7 @@ class FilterVariable(IteratorVariable):
             else:
                 res = self.fn.call_function(tx, [item], {})
             pred_res = variables.UserFunctionVariable(
-                polyfills.predicate  # type: ignore[arg-type]
+                polyfills.predicate
             ).call_function(tx, [res], {})
             if pred_res.as_python_constant():
                 return item

torch/_dynamo/variables/lists.py

@@ -1498,7 +1498,6 @@ class NamedTupleVariable(TupleVariable):
                     variables.UserDefinedClassVariable(self.tuple_cls),
                 )
             elif isinstance(method, staticmethod):
-                # pyrefly: ignore[bad-argument-type]
                 return UserFunctionVariable(method.__func__)
             elif inspect.isfunction(method):
                 return UserMethodVariable(method, self)

torch/_dynamo/variables/torch.py

@@ -472,12 +472,7 @@ class TorchCtxManagerClassVariable(BaseTorchVariable):
             )
         elif self.value is torch.nn.attention.sdpa_kernel.__wrapped__:  # type: ignore[attr-defined]
             name_to_arg_map = bind_args_cached(
-                # pyrefly: ignore[bad-argument-type]
-                self.value,
-                tx,
-                self.source,
-                args,
-                kwargs,
+                self.value, tx, self.source, args, kwargs
             )
             backends = name_to_arg_map["backends"].as_python_constant()
             set_priority = name_to_arg_map["set_priority"].as_python_constant()
@@ -1434,7 +1429,7 @@ class TorchInGraphFunctionVariable(BaseTorchVariable):
             packed_input_vt = TupleVariable.build(
                 tx, (TupleVariable.build(tx, args), ConstDictVariable.build(tx, kwargs))
             )
-            out_vt = variables.UserFunctionVariable(tree_flatten).call_function(  # type: ignore[arg-type]
+            out_vt = variables.UserFunctionVariable(tree_flatten).call_function(
                 tx, [packed_input_vt], {}
             )
             assert isinstance(out_vt, TupleVariable) and len(out_vt.items) == 2

torch/_inductor/config.py

@@ -550,10 +550,6 @@ max_autotune_flex_search_space: Literal["DEFAULT", "EXHAUSTIVE"] = os.environ.ge
     "TORCHINDUCTOR_MAX_AUTOTUNE_FLEX_SEARCH_SPACE", "DEFAULT"
 ).upper()  # type: ignore[assignment]
 
-cutedsl_enable_autotuning: bool = (
-    os.environ.get("CUTEDSL_ENABLE_AUTOTUNING", "0") == "1"
-)
-
 # DEPRECATED. This setting is ignored.
 autotune_fallback_to_aten = False
 

torch/_inductor/cpp_builder.py

@@ -913,6 +913,10 @@ def _get_optimization_cflags(
             if not config.is_fbcode():
                 if platform.machine() == "ppc64le":
                     cflags.append("mcpu=native")
+                elif platform.machine() == "riscv64":
+                    cflags.append("march=rv64gc")
+                elif platform.machine() == "riscv32":
+                    cflags.append("march=rv32gc")
                 else:
                     cflags.append("march=native")
 

torch/_inductor/fx_passes/joint_graph.py

@@ -893,6 +893,9 @@ def _other_is_broadcasted_in_dim(match):
     if isinstance(dim, int):
         dim = (dim,)
 
+    if any(d >= len(other_shape) for d in dim):
+        return False
+
     return all(statically_known_true(other_shape[d] == 1) for d in dim)
 
 

torch/_inductor/fx_passes/post_grad.py

@@ -1516,17 +1516,29 @@ def should_prefer_unfused_addmm(match):
 
 
 @register_graph_pattern(
-    CallFunction(aten.addmm, KeywordArg("inp"), Arg(), Arg()),
+    CallFunction(
+        aten.addmm,
+        KeywordArg("inp"),
+        Arg(),
+        Arg(),
+        beta=KeywordArg("beta"),
+        alpha=KeywordArg("alpha"),
+    ),
     # pyrefly: ignore [bad-argument-type]
     pass_dict=pass_patterns[2],
     extra_check=should_prefer_unfused_addmm,
 )
-def unfuse_bias_add_to_pointwise(match: Match, mat1, mat2, *, inp):
-    def repl(inp, x1, x2):
-        return x1 @ x2 + inp
+def unfuse_bias_add_to_pointwise(match: Match, mat1, mat2, *, inp, alpha, beta):
+    def repl(inp, x1, x2, alpha, beta):
+        mm_result = x1 @ x2
+        if alpha != 1:
+            mm_result = alpha * mm_result
+        if beta != 1:
+            inp = beta * inp
+        return inp + mm_result
 
     # pyrefly: ignore [bad-argument-type]
-    match.replace_by_example(repl, [inp, mat1, mat2])
+    match.replace_by_example(repl, [inp, mat1, mat2, alpha, beta])
 
 
 def is_valid_addmm_fusion(match):

torch/_inductor/kernel/flex/templates/common.py.jinja

@@ -100,7 +100,21 @@ def forward_block_mn(
     # m_ij
     l_i = l_i * alpha + tl.sum(p, 1)
     # # -- scale and update acc --
-    acc = acc * alpha[:, None]
+
+    {# acc = acc * alpha[:, None] #}
+    # Epilogue Subtile
+    acc0, acc1 = tl.split(
+        acc.reshape(BLOCK_M, 2, V_HEAD_DIM_ROUNDED//2).permute((0, 2, 1))
+    )
+
+    acc0 = acc0 * alpha[:, None]
+    acc1 = acc1 * alpha[:, None]
+
+    acc = (
+        tl.join(acc0, acc1)
+        .permute(0, 2, 1)
+        .reshape(BLOCK_M, V_HEAD_DIM_ROUNDED)
+    )
     {%- if USE_TMA %}
     v = tl.load_tensor_descriptor(
         desc_v,

torch/_inductor/kernel/mm_common.py

@@ -1,8 +1,6 @@
 # mypy: allow-untyped-defs
 import logging
 from collections.abc import Sequence
-from functools import partial
-from pathlib import Path
 from typing import Any
 
 import torch
@@ -14,7 +12,6 @@ from torch.fx.experimental.symbolic_shapes import has_free_unbacked_symbols
 from .. import config
 from ..codegen.wrapper import PythonWrapperCodegen
 from ..ir import _IntLike, Layout, TensorBox
-from ..utils import load_template
 
 
 log = logging.getLogger(__name__)
@@ -257,7 +254,3 @@ def is_batch_stride_largest_or_zero(mat1, mat2, layout) -> bool:
             return False
 
     return True
-
-
-_KERNEL_TEMPLATE_DIR = Path(__file__).parent / "templates"
-load_kernel_template = partial(load_template, template_dir=_KERNEL_TEMPLATE_DIR)

torch/_inductor/kernel/mm_grouped.py

@@ -1,13 +1,11 @@
 # mypy: allow-untyped-defs
 import logging
-from dataclasses import asdict, dataclass
+from dataclasses import dataclass
 from typing import Any, Optional
 
 import torch
 from torch._dynamo.utils import counters
-from torch._inductor.codegen.cutedsl.cutedsl_template import CuteDSLTemplate
 from torch._inductor.runtime.triton_compat import tl
-from torch._inductor.template_heuristics.cutedsl import get_groupgemm_configs
 from torch._inductor.virtualized import V
 from torch.utils._triton import has_triton
 
@@ -24,13 +22,11 @@ from ..utils import (
     get_num_sms,
     has_free_symbols,
     use_aten_gemm_kernels,
-    use_blackwell_cutedsl_grouped_mm,
     use_triton_template,
 )
 from .mm_common import (
     _is_static_problem,
     check_supported_striding,
-    load_kernel_template,
     persistent_grouped_mm_grid,
 )
 
@@ -517,11 +513,6 @@ triton_scaled_grouped_mm_template = TritonTemplate(
     source=triton_grouped_mm_source,
 )
 
-cutedsl_grouped_mm_template = CuteDSLTemplate(
-    name="grouped_gemm_cutedsl",
-    source=load_kernel_template("cutedsl_mm_grouped"),
-)
-
 
 def grouped_mm_args(
     mat1: TensorBox,
@@ -723,44 +714,43 @@ def _tuned_grouped_mm_common(
     # Checking only for the equality of corresponding dims of
     # multiplicands here, relying on meta function checks for
     # everything else.
-    if len(m1_size) == 2:
-        if len(m2_size) == 2:
-            m, k1 = m1_size
-            k2, _ = m2_size
-            # pyrefly: ignore [missing-attribute]
-            g = offs.get_size()[0]
-            V.graph.sizevars.check_equals(k1, k2)
-            a_is_2d, b_is_2d = True, True
-        else:
-            # pyrefly: ignore [missing-attribute]
-            g1 = offs.layout.size[0]
-            m, k1 = m1_size
-            g2, k2, _ = m2_size
-            g = V.graph.sizevars.check_equals_and_simplify(g1, g2)
-            V.graph.sizevars.check_equals(k1, k2)
-            a_is_2d, b_is_2d = True, False
-    else:
-        if len(m2_size) == 2:
-            # pyrefly: ignore [missing-attribute]
-            g1 = offs.layout.size[0]
-            g2, m, k1 = m1_size
-            k2, _ = m2_size
-            g = V.graph.sizevars.check_equals_and_simplify(g1, g2)
-            V.graph.sizevars.check_equals(k1, k2)
-            a_is_2d, b_is_2d = False, True
-        else:
-            g1, m, k1 = m1_size
-            g2, k2, _ = m2_size
-            g = V.graph.sizevars.check_equals_and_simplify(g1, g2)
-            V.graph.sizevars.check_equals(k1, k2)
-            a_is_2d, b_is_2d = False, False
-
     if (
         is_nonzero
         and use_triton_template(layout)
         and can_use_triton_kernel(mat_a, mat_b, offs, bias, scale_result)
     ):
         scaled = scale_a is not None
+        if len(m1_size) == 2:
+            if len(m2_size) == 2:
+                m, k1 = m1_size
+                k2, _ = m2_size
+                # pyrefly: ignore [missing-attribute]
+                g = offs.get_size()[0]
+                V.graph.sizevars.check_equals(k1, k2)
+                a_is_2d, b_is_2d = True, True
+            else:
+                # pyrefly: ignore [missing-attribute]
+                g1 = offs.layout.size[0]
+                m, k1 = m1_size
+                g2, k2, _ = m2_size
+                g = V.graph.sizevars.check_equals_and_simplify(g1, g2)
+                V.graph.sizevars.check_equals(k1, k2)
+                a_is_2d, b_is_2d = True, False
+        else:
+            if len(m2_size) == 2:
+                # pyrefly: ignore [missing-attribute]
+                g1 = offs.layout.size[0]
+                g2, m, k1 = m1_size
+                k2, _ = m2_size
+                g = V.graph.sizevars.check_equals_and_simplify(g1, g2)
+                V.graph.sizevars.check_equals(k1, k2)
+                a_is_2d, b_is_2d = False, True
+            else:
+                g1, m, k1 = m1_size
+                g2, k2, _ = m2_size
+                g = V.graph.sizevars.check_equals_and_simplify(g1, g2)
+                V.graph.sizevars.check_equals(k1, k2)
+                a_is_2d, b_is_2d = False, False
 
         a_is_k_major = mat_a.get_stride()[-1] == 1
         b_is_k_major = mat_b.get_stride()[-2] == 1
@@ -798,22 +788,6 @@ def _tuned_grouped_mm_common(
                 **config.kwargs,
             )
 
-    if use_blackwell_cutedsl_grouped_mm(
-        mat_a, mat_b, layout, a_is_2d, b_is_2d, offs, bias, scale_result
-    ):
-        for config in get_groupgemm_configs():
-            kwargs = dict(
-                ACC_DTYPE="cutlass.Float32",
-            )
-
-            cutedsl_grouped_mm_template.maybe_append_choice(
-                choices,
-                input_nodes=input_nodes,
-                layout=layout,
-                **kwargs,
-                **asdict(config),
-            )
-
     input_gen_fns = {
         4: lambda x: create_offsets(
             x, m1_size, m2_size, offs.get_size() if offs is not None else None

torch/_inductor/kernel/templates/cutedsl_mm_grouped.py.jinja

@@ -1,333 +0,0 @@
-import functools
-from torch._inductor.runtime.runtime_utils import ceildiv
-from cutlass.utils import TensorMapUpdateMode
-{{gen_defines()}}
-# ---- Import GroupedGemm implementation, copied on PyTorch build from Cutlass repository: cutlass/examples/python/CuTeDSL/blackwell/grouped_gemm.py ----
-from torch._inductor.kernel.vendored_templates.cutedsl_grouped_gemm import (
-    GroupedGemmKernel,
-)
-
-
-# Note about caching:
-# Each instantiated CuTeDSL grouped GEMM kernel file generated by Inductor
-# maintains its own local caching system. At this stage, all compile-time
-# constexprs (e.g., TILE_M, TILE_N, CLUSTER_M/N, USE_2_CTA) and the kernel
-# name itself ({{kernel_name}}) are permanently baked into the file, so they
-# do not need to be included in any cache key.
-#
-# The caching mechanism is split into two levels:
-#
-#   1. prep_cache
-#      Caches the compiled executor for build_group_ptrs_from_bases(). This
-#      kernel depends only on the tensor shapes, strides, and dtypes of A/B/C,
-#      and can therefore be safely reused across runs with different group
-#      partitioning (`offs`).
-#
-#   2. gemm_cache
-#      Caches the compiled Grouped GEMM executor. Its key extends the prep
-#      cache key with hardware- and grid-specific parameters:
-#      (prep_cache_key, max_active_clusters, total_num_clusters).
-#      This is necessary because different `offs` tensors can change the
-#      per-group problem sizes and thus alter `total_num_clusters`, which in
-#      turn changes the grid shape and persistent scheduler configuration.
-#      Kernels compiled for one grid cannot be safely reused for another.
-#
-#
-# Additionally, note the @lru_cache decorator on get_hardware_info(). Empirically,
-# hw.get_max_active_clusters() triggers significant MLIR recompilation overhead,
-# despite depending only on the GPU type. We cache this function to mitigate
-# redundant recompiles even when shape/stride/dtype cache misses force kernel
-# regeneration. A follow-up study will investigate the root cause.
-
-prep_cache = {}
-gemm_cache = {}
-
-
-@functools.lru_cache
-def get_hardware_info():
-    hw = cutlass.utils.HardwareInfo()
-    sm_count = hw.get_max_active_clusters(1)
-    max_active_clusters = hw.get_max_active_clusters(CLUSTER_M * CLUSTER_N)
-
-    return (sm_count, max_active_clusters)
-
-
-def get_prep_cache_key(input_a, input_b, output):
-    """
-    Returns a tuple key for caching the preprocessing kernel executor based on kernel name,
-    shapes, strides, and dtypes of input/output tensors.
-    """
-    return (
-        tuple(input_a.shape),
-        tuple(input_a.stride()),
-        input_a.dtype,
-        tuple(input_b.shape),
-        tuple(input_b.stride()),
-        input_b.dtype,
-        tuple(output.shape),
-        tuple(output.stride()),
-        output.dtype,
-    )
-
-
-def get_gemm_cache_key(prep_cache_key, max_active_clusters, total_num_clusters):
-    """
-    Returns a tuple key for caching the gemm kernel executor by extending the
-    prep cache key with hardware- and grid-specific parameters.
-    """
-    return (
-        prep_cache_key,
-        max_active_clusters,
-        total_num_clusters,
-    )
-
-
-@cute.kernel
-def build_group_ptrs_from_bases_kernel(
-    base_A_u64: cutlass.Int64,  # device addr of input_a (bytes)
-    base_B_u64: cutlass.Int64,  # device addr of input_b (bytes)
-    base_C_u64: cutlass.Int64,  # device addr of Output (bytes)
-    offs: cute.Tensor,  # [G], cutlass.Int32/64 cumulative
-    K: cutlass.Constexpr,
-    N: cutlass.Constexpr,
-    sizeof_element: cutlass.Int32,  # bytes
-    # -------- STRIDES (in ELEMENTS) --------
-    stride_A_m_elems: cutlass.Constexpr,  # A.stride(0)
-    stride_A_k_elems: cutlass.Constexpr,  # A.stride(1)
-    stride_B0_elems: cutlass.Constexpr,  # B.stride(0)
-    stride_Bk_elems: cutlass.Constexpr,  # B.stride(1)
-    stride_Bn_elems: cutlass.Constexpr,  # B.stride(2)
-    stride_C_m_elems: cutlass.Constexpr,  # C.stride(0)
-    stride_C_n_elems: cutlass.Constexpr,  # C.stride(1)
-    # -------- OUTPUTS --------
-    out_ptrs: cute.Tensor,  # [G,3] cutlass.Int64: (A_ptr, B_ptr, C_ptr)
-    out_problem: cute.Tensor,  # [G,4] cutlass.Int32: (m_g, n, k, 1)
-    out_strides_abc: cute.Tensor,  # [G,3,2] cutlass.Int32 [[A_m,A_k],[B_n,B_k],[C_m,C_n]]
-):
-    tidx, _, _ = cute.arch.thread_idx()
-    g = tidx
-
-    m_beg_i32 = 0
-    if g > 0:
-        m_beg_i32 = offs[g - 1]
-    m_end_i32 = offs[g]
-    m_g_i32 = m_end_i32 - m_beg_i32
-
-    a_byte_off = (
-        cutlass.Int64(m_beg_i32) * stride_A_m_elems * cutlass.Int64(sizeof_element)
-    )
-    c_byte_off = (
-        cutlass.Int64(m_beg_i32) * stride_C_m_elems * cutlass.Int64(sizeof_element)
-    )
-    b_byte_off = cutlass.Int64(g) * stride_B0_elems * cutlass.Int64(sizeof_element)
-
-    # ---- pointers ----
-    out_ptrs[g, 0] = base_A_u64 + a_byte_off
-    out_ptrs[g, 1] = base_B_u64 + b_byte_off
-    out_ptrs[g, 2] = base_C_u64 + c_byte_off
-
-    # ---- (m, n, k, 1) ----
-    out_problem[g, 0] = m_g_i32
-    out_problem[g, 1] = N
-    out_problem[g, 2] = K
-    out_problem[g, 3] = cutlass.Int32(1)
-
-    # ---- strides ----
-    out_strides_abc[g, 0, 0] = cutlass.Int32(stride_A_m_elems)
-    out_strides_abc[g, 0, 1] = cutlass.Int32(stride_A_k_elems)
-    out_strides_abc[g, 1, 0] = cutlass.Int32(stride_Bn_elems)
-    out_strides_abc[g, 1, 1] = cutlass.Int32(stride_Bk_elems)
-    out_strides_abc[g, 2, 0] = cutlass.Int32(stride_C_m_elems)
-    out_strides_abc[g, 2, 1] = cutlass.Int32(stride_C_n_elems)
-
-
-@cute.jit
-def launch_build_group_ptrs_from_bases(
-    base_A_u64: cutlass.Int64,
-    base_B_u64: cutlass.Int64,
-    base_C_u64: cutlass.Int64,
-    offs: cute.Tensor,
-    G: cutlass.Constexpr,
-    K: cutlass.Constexpr,
-    N: cutlass.Constexpr,
-    sizeof_element: cutlass.Constexpr,
-    stride_A_m_elems: cutlass.Constexpr,
-    stride_A_k_elems: cutlass.Constexpr,
-    stride_B0_elems: cutlass.Constexpr,
-    stride_Bk_elems: cutlass.Constexpr,
-    stride_Bn_elems: cutlass.Constexpr,
-    stride_C_m_elems: cutlass.Constexpr,
-    stride_C_n_elems: cutlass.Constexpr,
-    out_ptrs: cute.Tensor,  # [G,3] cutlass.Int64
-    out_problem: cute.Tensor,  # [G,4] cutlass.Int32
-    out_strides_abc: cute.Tensor,  # [3,2] cutlass.Int32
-    stream: cuda.CUstream,
-):
-    build_group_ptrs_from_bases_kernel(
-        base_A_u64,
-        base_B_u64,
-        base_C_u64,
-        offs,
-        K,
-        N,
-        sizeof_element,
-        stride_A_m_elems,
-        stride_A_k_elems,
-        stride_B0_elems,
-        stride_Bk_elems,
-        stride_Bn_elems,
-        stride_C_m_elems,
-        stride_C_n_elems,
-        out_ptrs,
-        out_problem,
-        out_strides_abc,
-    ).launch(grid=(1, 1, 1), block=(G, 1, 1), stream=stream)
-
-
-{{def_kernel("input_a", "input_b", "input_a_offs")}}
-    stream = cuda.CUstream(stream)
-
-    input_b = input_b.transpose(1, 2)
-
-    sumM, K = input_a.shape
-    G, N, Kb = input_b.shape
-
-    dev = input_a.device
-
-    base_A_u64 = int(input_a.data_ptr())
-    base_B_u64 = int(input_b.data_ptr())
-    base_C_u64 = int({{get_output()}}.data_ptr())
-
-    ptrs_t = torch.empty((G, 3), device=dev, dtype=torch.int64)
-    probs_t = torch.empty((G, 4), device=dev, dtype=torch.int32)
-    strides_t = torch.empty((G, 3, 2), device=dev, dtype=torch.int32)
-    ptrs = from_dlpack(ptrs_t)
-    probs = from_dlpack(probs_t)
-    strides = from_dlpack(strides_t)
-
-    prep_cache_key = get_prep_cache_key(input_a, input_b, {{get_output()}})
-    prep_executor = prep_cache.get(prep_cache_key)
-
-    if prep_executor is None:
-        sizeof_element = int(input_a.element_size())
-        sA_m, sA_k = map(int, input_a.stride())
-        sB_0, sB_n, sB_k = map(int, input_b.stride())
-        sC_m, sC_n = map(int, {{get_output()}}.stride())
-
-        prep_executor = cute.compile(
-            launch_build_group_ptrs_from_bases,
-            base_A_u64=base_A_u64,
-            base_B_u64=base_B_u64,
-            base_C_u64=base_C_u64,
-            offs=from_dlpack(input_a_offs),
-            G=int(G),
-            K=int(K),
-            N=int(N),
-            sizeof_element=sizeof_element,
-            stride_A_m_elems=sA_m,
-            stride_A_k_elems=sA_k,
-            stride_B0_elems=sB_0,
-            stride_Bk_elems=sB_k,
-            stride_Bn_elems=sB_n,
-            stride_C_m_elems=sC_m,
-            stride_C_n_elems=sC_n,
-            out_ptrs=ptrs,
-            out_problem=probs,
-            out_strides_abc=strides,
-            stream=stream,
-        )
-
-        prep_cache[prep_cache_key] = prep_executor
-
-    prep_executor(
-        base_A_u64=base_A_u64,
-        base_B_u64=base_B_u64,
-        base_C_u64=base_C_u64,
-        offs=from_dlpack(input_a_offs),
-        out_ptrs=ptrs,
-        out_problem=probs,
-        out_strides_abc=strides,
-        stream=stream,
-    )
-
-    # --- Tensormap workspace per SM ---
-    num_tensormap_buffers, max_active_clusters = get_hardware_info()
-    tensormap_shape = (
-        num_tensormap_buffers,
-        GroupedGemmKernel.num_tensormaps,
-        GroupedGemmKernel.bytes_per_tensormap // 8,
-    )
-    tensormap_workspace_t = torch.empty(tensormap_shape, device=dev, dtype=torch.int64)
-    tensormap_workspace = from_dlpack(tensormap_workspace_t)
-
-    # --- Total clusters ---
-    def compute_total_num_clusters(
-        problem_sizes_mnkl,
-        cluster_tile_shape_mn,
-    ):
-        total_num_clusters = 0
-        for m, n, _, _ in problem_sizes_mnkl:
-            num_clusters_mn = tuple(
-                ceildiv(x, y) for x, y in zip((m, n), cluster_tile_shape_mn)
-            )
-            total_num_clusters += functools.reduce(lambda x, y: x * y, num_clusters_mn)
-        return total_num_clusters
-
-    # Compute cluster tile shape
-    def compute_cluster_tile_shape(
-        mma_tiler_mn,
-        cluster_shape_mn,
-        use_2cta_instrs,
-    ):
-        cta_tile_shape_mn = list(mma_tiler_mn)
-        if use_2cta_instrs:
-            cta_tile_shape_mn[0] = cta_tile_shape_mn[0] // 2
-        return tuple(x * y for x, y in zip(cta_tile_shape_mn, cluster_shape_mn))
-
-    cluster_tile_shape_mn = compute_cluster_tile_shape(
-        (TILE_M, TILE_N), (CLUSTER_M, CLUSTER_N), bool(USE_2_CTA)
-    )
-
-    total_num_clusters = int(compute_total_num_clusters(probs_t, cluster_tile_shape_mn))
-
-    gemm_cache_key = get_gemm_cache_key(
-        prep_cache_key, max_active_clusters, total_num_clusters
-    )
-    gemm_executor = gemm_cache.get(gemm_cache_key)
-
-    if gemm_executor is None:
-        grouped_gemm = GroupedGemmKernel(
-            acc_dtype=ACC_DTYPE,
-            use_2cta_instrs=USE_2_CTA,
-            mma_tiler_mn=(TILE_M, TILE_N),
-            cluster_shape_mn=(CLUSTER_M, CLUSTER_N),
-            tensormap_update_mode=TENSORMAP_UPDATE_MODE,
-        )
-
-        gemm_executor = cute.compile(
-            grouped_gemm,
-            from_dlpack(input_a.unsqueeze(-1), assumed_align=16),
-            from_dlpack(input_b[0].unsqueeze(-1), assumed_align=16),
-            from_dlpack({{get_output()}}.unsqueeze(-1), assumed_align=16),
-            G,
-            probs,
-            strides,
-            ptrs,
-            total_num_clusters,
-            tensormap_workspace,
-            max_active_clusters,
-            stream,
-        )
-
-        gemm_cache[gemm_cache_key] = gemm_executor
-
-    gemm_executor(
-        from_dlpack(input_a.unsqueeze(-1), assumed_align=16),
-        from_dlpack(input_b[0].unsqueeze(-1), assumed_align=16),
-        from_dlpack({{get_output()}}.unsqueeze(-1), assumed_align=16),
-        probs,
-        strides,
-        ptrs,
-        tensormap_workspace,
-        stream,
-    )

torch/_inductor/template_heuristics/cutedsl.py

@@ -1,141 +0,0 @@
-from dataclasses import dataclass
-from enum import auto, Enum
-from itertools import product
-
-import torch._inductor.config as config
-
-
-class TensorMapUpdateMode(Enum):
-    """Enum mirroring cutlass.utils.TensorMapUpdateMode to decouple this file from a cutlass dependency."""
-
-    SMEM = auto()
-    GMEM = auto()
-
-
-@dataclass(frozen=True)
-class CuTeGemmConfig:
-    TILE_M: int = 128
-    TILE_N: int = 192
-    CLUSTER_M: int = 2
-    CLUSTER_N: int = 1
-    USE_2_CTA: bool = False
-    TENSORMAP_UPDATE_MODE: TensorMapUpdateMode = TensorMapUpdateMode.SMEM
-
-
-def get_exhaustive_groupgemm_configs() -> list[CuTeGemmConfig]:
-    """
-    Returns the exhaustive configuration set for the Blackwell CuTeDSL Grouped GEMM kernel.
-    For information regarding valid config sets, see:
-    https://github.com/NVIDIA/cutlass/blob/main/examples/python/CuTeDSL/blackwell/grouped_gemm.py
-    """
-
-    # Tile_n is always the same regardless of 2cta
-    tile_n_vals = [32, 64, 96, 128, 160, 192, 224, 256]
-
-    # Valid clusters
-    clusters_no_2cta = [
-        (1, 1),
-        (1, 2),
-        (1, 4),
-        (1, 8),
-        (1, 16),
-        (2, 1),
-        (2, 2),
-        (2, 4),
-        (2, 8),
-        (4, 1),
-        (4, 2),
-        (4, 4),
-        (8, 1),
-        (8, 2),
-        (16, 1),
-    ]
-    clusters_2cta = [
-        (2, 1),
-        (2, 2),
-        (2, 4),
-        (2, 8),
-        (4, 1),
-        (4, 2),
-        (4, 4),
-        (8, 1),
-        (8, 2),
-        (16, 1),
-    ]
-
-    configs: list[CuTeGemmConfig] = []
-
-    for use_2cta, cluster_set, tile_m_range in [
-        (False, clusters_no_2cta, [64, 128]),
-        (True, clusters_2cta, [128, 256]),
-    ]:
-        for tensormap_update_mode, tile_m, tile_n, (cluster_m, cluster_n) in product(
-            [TensorMapUpdateMode.SMEM, TensorMapUpdateMode.GMEM],
-            tile_m_range,
-            tile_n_vals,
-            cluster_set,
-        ):
-            configs.append(
-                CuTeGemmConfig(
-                    tile_m,
-                    tile_n,
-                    cluster_m,
-                    cluster_n,
-                    USE_2_CTA=use_2cta,
-                    TENSORMAP_UPDATE_MODE=tensormap_update_mode,
-                )
-            )
-
-    return configs
-
-
-def get_default_groupgemm_configs() -> list[CuTeGemmConfig]:
-    """
-    Returns the default configuration set for the Blackwell CuTeDSL Grouped GEMM kernel.
-    """
-
-    config_tuples = [
-        (128, 256, 2, 1, False, TensorMapUpdateMode.SMEM),
-        (256, 160, 2, 1, True, TensorMapUpdateMode.GMEM),
-        (256, 256, 2, 1, True, TensorMapUpdateMode.GMEM),
-        (64, 32, 1, 1, False, TensorMapUpdateMode.GMEM),
-        (64, 256, 1, 2, False, TensorMapUpdateMode.SMEM),
-        (128, 256, 1, 2, False, TensorMapUpdateMode.SMEM),
-        (256, 256, 2, 2, True, TensorMapUpdateMode.GMEM),
-        (128, 256, 1, 2, False, TensorMapUpdateMode.GMEM),
-        (64, 32, 1, 1, False, TensorMapUpdateMode.SMEM),
-        (256, 256, 2, 1, True, TensorMapUpdateMode.SMEM),
-        (128, 256, 1, 1, False, TensorMapUpdateMode.GMEM),
-        (256, 256, 8, 1, True, TensorMapUpdateMode.GMEM),
-        (64, 32, 1, 2, False, TensorMapUpdateMode.SMEM),
-        (256, 192, 2, 1, True, TensorMapUpdateMode.GMEM),
-        (256, 256, 2, 2, True, TensorMapUpdateMode.SMEM),
-        (128, 96, 1, 2, False, TensorMapUpdateMode.SMEM),
-        (64, 192, 1, 1, False, TensorMapUpdateMode.SMEM),
-        (64, 64, 1, 1, False, TensorMapUpdateMode.GMEM),
-        (64, 192, 1, 1, False, TensorMapUpdateMode.GMEM),
-        (128, 64, 1, 1, False, TensorMapUpdateMode.GMEM),
-        (64, 160, 1, 1, False, TensorMapUpdateMode.GMEM),
-        (64, 256, 1, 1, False, TensorMapUpdateMode.GMEM),
-    ]
-
-    return [CuTeGemmConfig(*args) for args in config_tuples]
-
-
-def get_groupgemm_configs() -> list[CuTeGemmConfig]:
-    """
-    Returns the configuration set for the Blackwell CuTeDSL Grouped GEMM kernel.
-
-    Note: CuTeDSL autotuning is still experimental — enabling it may trigger kernel launch failures
-    or unstable results. By default, autotuning is disabled and we return only
-    a single baseline config.
-    """
-    if (
-        config.cutedsl_enable_autotuning
-        and config.max_autotune_gemm_search_space == "EXHAUSTIVE"
-    ):
-        return get_exhaustive_groupgemm_configs()
-    elif config.cutedsl_enable_autotuning:
-        return get_default_groupgemm_configs()
-    else:
-        return [get_default_groupgemm_configs()[0]]

torch/_inductor/utils.py

@@ -1911,84 +1911,6 @@ def use_triton_blackwell_tma_template(
     return has_triton_tensor_descriptor_host_tma() and is_datacenter_blackwell_arch()
 
 
-@functools.lru_cache(maxsize=1)
-def ensure_cute_available() -> bool:
-    """Check if CuTeDSL is importable; cache the result for reuse.
-
-    Call ensure_cute_available.cache_clear() after installing CuTeDSL
-    in the same interpreter to retry the import.
-    """
-    try:
-        return importlib.util.find_spec("cutlass.cute") is not None
-    except ImportError:
-        return False
-
-
-def use_blackwell_cutedsl_grouped_mm(
-    mat_a: Any,
-    mat_b: Any,
-    layout: Layout,
-    a_is_2d: bool,
-    b_is_2d: bool,
-    offs: Optional[Any],
-    bias: Optional[Any],
-    scale_result: Optional[Any],
-) -> bool:
-    """
-    Returns True if we can use the blackwell kernel for grouped mm.
-    Required conditions:
-        1. CuTeDSL backend is enabled
-        2. CuTeDSL is available
-        3. We are on a blackwell arch
-        4. The dtype is bf16
-        5. Max autotune or max autotune gemm is enabled
-        6. A, B, and the output are 16B aligned
-        7. We are not using dynamic shapes
-        8. A is 2d
-        9. B is 3d
-        10. Offsets are provided
-        11. Bias and Scale are not provided
-    """
-    if not ensure_cute_available():
-        return False
-
-    if not _use_autotune_backend("CUTEDSL"):
-        return False
-
-    from .codegen.cuda.cuda_env import is_datacenter_blackwell_arch
-
-    if not is_gpu(layout.device.type):
-        return False
-
-    if not is_datacenter_blackwell_arch():
-        return False
-
-    layout_dtypes = [torch.bfloat16]
-    if not _use_template_for_gpu(layout, layout_dtypes):
-        return False
-
-    if not (config.max_autotune or config.max_autotune_gemm):
-        return False
-
-    # Checks for 16B ptr and stride alignment
-    if not can_use_tma(mat_a, mat_b, output_layout=layout):
-        return False
-
-    if any(is_dynamic(x) for x in [mat_a, mat_b]):
-        return False
-
-    if not a_is_2d or b_is_2d:
-        return False
-
-    if offs is None:
-        return False
-
-    if bias is not None or scale_result is not None:
-        return False
-
-    return True
-
-
 def use_cutlass_template(layout: Layout, m: int, n: int, k: int) -> bool:
     from .virtualized import V
 

torch/distributed/pipelining/schedules.py

@@ -1485,6 +1485,7 @@ class PipelineScheduleMulti(_PipelineSchedule):
         output_merge_spec: Optional[Union[dict[str, Any], tuple[Any]]] = None,
         use_full_backward: Optional[bool] = None,
         scale_grads: bool = True,
+        backward_requires_autograd: bool = True,
     ):
         # Init parent
         super().__init__(
@@ -1517,6 +1518,11 @@ class PipelineScheduleMulti(_PipelineSchedule):
         # This will be set during init of derived schedules
         self.pipeline_order: dict[int, list[Optional[_Action]]] = {}
 
+        # When using a custom backward function, we may or may not need autograd to be used
+        # for the backward pass. This flag is used to determine whether or torch.is_grad_enabled()
+        # check should be performed before the step function.
+        self._backward_requires_autograd = backward_requires_autograd
+
         if use_full_backward is not None:
             logger.warning(
                 "Deprecation warning: 'use_full_backward' is no longer supported. "
@@ -1609,7 +1615,11 @@ class PipelineScheduleMulti(_PipelineSchedule):
         losses: a list to store the losses for each microbatch.
         return_outputs: whether to return the outputs from the last stage.
         """
-        if self._has_backward and not torch.is_grad_enabled():
+        if (
+            self._has_backward
+            and self._backward_requires_autograd
+            and not torch.is_grad_enabled()
+        ):
             raise RuntimeError(
                 "step() requires gradients to be enabled for backward computation; "
                 "it should not be used under torch.no_grad() context. "
@@ -1891,7 +1901,7 @@ class _PipelineScheduleRuntime(PipelineScheduleMulti):
         Args:
             computation_type: The computation type for which to register the custom function
             custom_function: The function to execute when this computation type is encountered.
-                Must have signature: (stage: _PipelineStageBase, mb_index: int, *args, **kwargs) -> None
+                Must have signature: (action: _Action, ctx: _PipelineContext) -> None
         """
         # Ensure that the computation type is valid
         if computation_type not in (
@@ -1900,10 +1910,13 @@ class _PipelineScheduleRuntime(PipelineScheduleMulti):
             BACKWARD_INPUT,
             BACKWARD_WEIGHT,
             OVERLAP_F_B,
+            UNSHARD,
+            RESHARD,
+            REDUCE_GRAD,
         ):
             raise ValueError(
                 f"Invalid computation type {computation_type}. Only FORWARD, FULL_BACKWARD, \
-BACKWARD_INPUT, BACKWARD_WEIGHT, and OVERLAP_F_B are supported."
+                BACKWARD_INPUT, BACKWARD_WEIGHT, OVERLAP_F_B, UNSHARD, RESHARD and REDUCE_GRAD are supported."
             )
 
         # Check if computation_type is already registered
@@ -2296,6 +2309,7 @@ class ScheduleLoopedBFS(_PipelineScheduleRuntime):
         loss_fn: Optional[Union[Callable, _Loss]] = None,
         output_merge_spec: Optional[Union[dict[str, Any], tuple[Any]]] = None,
         scale_grads: bool = True,
+        backward_requires_autograd: bool = True,
     ):
         super().__init__(
             stages=stages,
@@ -2303,6 +2317,7 @@ class ScheduleLoopedBFS(_PipelineScheduleRuntime):
             loss_fn=loss_fn,
             output_merge_spec=output_merge_spec,
             scale_grads=scale_grads,
+            backward_requires_autograd=backward_requires_autograd,
         )
 
         # 1. Create the pipeline_order (all ranks do this calculation)
@@ -2510,6 +2525,7 @@ class ScheduleInterleaved1F1B(_PipelineScheduleRuntime):
         kwargs_chunk_spec: Optional[dict[str, TensorChunkSpec]] = None,
         output_merge_spec: Optional[Union[dict[str, Any], tuple[Any]]] = None,
         scale_grads: bool = True,
+        backward_requires_autograd: bool = True,
     ):
         self.pp_group_size = stages[0].group_size
         super().__init__(
@@ -2520,6 +2536,7 @@ class ScheduleInterleaved1F1B(_PipelineScheduleRuntime):
             kwargs_chunk_spec=kwargs_chunk_spec,
             output_merge_spec=output_merge_spec,
             scale_grads=scale_grads,
+            backward_requires_autograd=backward_requires_autograd,
         )
         self.n_local_stages = len(stages)
         self.rank = stages[0].group_rank
@@ -2622,6 +2639,7 @@ class ScheduleInterleavedZeroBubble(_PipelineScheduleRuntime):
         kwargs_chunk_spec: Optional[dict[str, TensorChunkSpec]] = None,
         output_merge_spec: Optional[Union[dict[str, Any], tuple[Any]]] = None,
         scale_grads: bool = True,
+        backward_requires_autograd: bool = True,
     ):
         # TODO: we dont support input/weight backward split with torch.compile
         _check_torch_compile_compatibility(stages, self.__class__.__name__)
@@ -2634,6 +2652,7 @@ class ScheduleInterleavedZeroBubble(_PipelineScheduleRuntime):
             kwargs_chunk_spec=kwargs_chunk_spec,
             output_merge_spec=output_merge_spec,
             scale_grads=scale_grads,
+            backward_requires_autograd=backward_requires_autograd,
         )
         self.n_local_stages = len(stages)
         self.rank = stages[0].group_rank
@@ -2819,6 +2838,7 @@ class ScheduleZBVZeroBubble(_PipelineScheduleRuntime):
         kwargs_chunk_spec: Optional[dict[str, TensorChunkSpec]] = None,
         output_merge_spec: Optional[Union[dict[str, Any], tuple[Any]]] = None,
         scale_grads: bool = True,
+        backward_requires_autograd: bool = True,
     ):
         # TODO: we dont support input/weight backward split with torch.compile
         _check_torch_compile_compatibility(stages, self.__class__.__name__)
@@ -2831,6 +2851,7 @@ class ScheduleZBVZeroBubble(_PipelineScheduleRuntime):
             kwargs_chunk_spec=kwargs_chunk_spec,
             output_merge_spec=output_merge_spec,
             scale_grads=scale_grads,
+            backward_requires_autograd=backward_requires_autograd,
         )
         self.stage_index_to_group_rank = generate_stage_to_rank_mapping(
             self.pp_group_size, self._num_stages, style="v"
@@ -2995,6 +3016,7 @@ class ScheduleDualPipeV(_PipelineScheduleRuntime):
         kwargs_chunk_spec: Optional[dict[str, TensorChunkSpec]] = None,
         output_merge_spec: Optional[Union[dict[str, Any], tuple[Any]]] = None,
         scale_grads: bool = True,
+        backward_requires_autograd: bool = True,
     ):
         # TODO: we dont support input/weight backward split with torch.compile
         _check_torch_compile_compatibility(stages, self.__class__.__name__)
@@ -3007,6 +3029,7 @@ class ScheduleDualPipeV(_PipelineScheduleRuntime):
             kwargs_chunk_spec=kwargs_chunk_spec,
             output_merge_spec=output_merge_spec,
             scale_grads=scale_grads,
+            backward_requires_autograd=backward_requires_autograd,
         )
         self.stage_index_to_group_rank = generate_stage_to_rank_mapping(
             self.pp_group_size, self._num_stages, style="v"

torch/fx/experimental/_config.py

@@ -2,6 +2,8 @@ import os
 import sys
 from typing import Optional
 
+from torch.utils._config_module import Config, install_config_module
+
 
 # [@compile_ignored: debug] Fails hard instead of graph breaking on guard on data dependent errors.
 no_data_dependent_graph_break = (
@@ -100,7 +102,11 @@ backed_size_oblivious = False
 # Skip dtype check in meta registrations. Only used for systems that does its own dtype checking.
 skip_dtype_check_in_meta_registrations = False
 
-from torch.utils._config_module import install_config_module
+# Experimental: If True, graph module will register fx metadata during recompile()
+enrich_profiler_metadata: bool = Config(  # type: ignore[var-annotated]
+    default=False,
+    env_name_default="TORCH_ENRICH_RPOFILER_STACK_TRACE",
+)
 
 
 install_config_module(sys.modules[__name__])

torch/fx/graph_module.py

@@ -20,6 +20,7 @@ from torch.nn.modules.module import _addindent
 from torch.package import Importer, PackageExporter, PackageImporter, sys_importer
 
 from ._compatibility import compatibility
+from .experimental import _config as fx_experimental_config
 from .graph import (
     _BoxedCodeGen,
     _custom_builtins,
@@ -858,14 +859,15 @@ class {module_name}(torch.nn.Module):
         called after editing the contained ``graph``, otherwise the generated
         code of this ``GraphModule`` will be out of date.
         """
+        # Do not import anything inside recompile, it might slow down the
+        # function and cause perf regression. Import outside of the method instead.
         if isinstance(self._graph._codegen, _PyTreeCodeGen):
             self._in_spec = self._graph._codegen.pytree_info.in_spec
             self._out_spec = self._graph._codegen.pytree_info.out_spec
 
-        from torch._dynamo import config as dynamo_config
-
         python_code = self._graph.python_code(
-            root_module="self", record_func=dynamo_config.enrich_profiler_metadata
+            root_module="self",
+            record_func=fx_experimental_config.enrich_profiler_metadata,
         )
         self._code = python_code.src
         self._lineno_map = python_code._lineno_map
@@ -874,7 +876,7 @@ class {module_name}(torch.nn.Module):
         cls = type(self)
         co_fields = self._graph._co_fields if hasattr(self._graph, "_co_fields") else {}
 
-        if dynamo_config.enrich_profiler_metadata:
+        if fx_experimental_config.enrich_profiler_metadata:
             # Generate metadata and register for profiler augmentation
             node_metadata: dict[int, dict[str, Any]] = {}
             for i, node in enumerate(self._graph.nodes):

torch/testing/_internal/common_distributed.py

@@ -1771,6 +1771,22 @@ class MultiProcContinuousTest(TestCase):
                 cls._run_test_given_id(test_id)
                 completion_queue.put(test_id)
             except BaseException as ex:  # noqa: B036
+                if isinstance(ex, SystemExit):
+                    # Get exit code from the process
+                    exit_code = getattr(ex, "code", None)
+
+                    # Look up exit code in TEST_SKIPS to see if it is a valid skip
+                    skip_entry = next(
+                        (v for v in TEST_SKIPS.values() if v.exit_code == exit_code),
+                        None,
+                    )
+
+                    # If we found an entry, we want to skip the test and the object back to the main process
+                    if skip_entry:
+                        completion_queue.put(unittest.SkipTest(skip_entry.message))
+                        # Skip exception handling below, move to main thread for processing the skip
+                        continue
+
                 raised_exception = True
                 # Send the exception and stack trace back to the dispatcher
                 exc_info = sys.exc_info()
@@ -1892,6 +1908,8 @@ class MultiProcContinuousTest(TestCase):
                 # Wait for the workers to finish the test
                 for i, completion_queue in enumerate(self.completion_queues):
                     rv = completion_queue.get()
+                    if isinstance(rv, unittest.SkipTest):
+                        raise rv
                     if isinstance(rv, BaseException):
                         # Hit an exception, re-raise it in the main process.
                         logger.warning(

torch/testing/_internal/common_utils.py

@@ -114,8 +114,6 @@ class ProfilingMode(Enum):
     PROFILING = 3
 
 # Set by parse_cmd_line_args() if called
-CI_FUNCTORCH_ROOT = ""
-CI_PT_ROOT = ""
 CI_TEST_PREFIX = ""
 DISABLED_TESTS_FILE = ""
 GRAPH_EXECUTOR : Optional[ProfilingMode] = None
@@ -959,8 +957,6 @@ def _get_test_report_path():
     return os.path.join('test-reports', test_source)
 
 def parse_cmd_line_args():
-    global CI_FUNCTORCH_ROOT
-    global CI_PT_ROOT
     global CI_TEST_PREFIX
     global DISABLED_TESTS_FILE
     global GRAPH_EXECUTOR
@@ -1039,10 +1035,8 @@ def parse_cmd_line_args():
 
     set_rng_seed()
 
-# CI Prefix path used only on CI environment
+    # CI Prefix path used only on CI environment
     CI_TEST_PREFIX = str(Path(os.getcwd()))
-    CI_PT_ROOT = str(Path(os.getcwd()).parent)
-    CI_FUNCTORCH_ROOT = str(os.path.join(Path(os.getcwd()).parent, "functorch"))
 
 def wait_for_process(p, timeout=None):
     try: