[PHI decoupling] remove framework/data_type.h from phi (#47776)

* remove framework/data_type.h from phi

* fix CI fail: map proto::VarType to phi::DataType

* refactor code to add more detailed comments

paddle/phi/core/utils/data_type.h

@@ -14,6 +14,7 @@ limitations under the License. */
 
 #pragma once
 #include <iostream>
+#include <map>
 #include <string>
 #include <typeindex>
 
@@ -23,6 +24,23 @@ limitations under the License. */
 
 namespace phi {
 
+// Here we can't depend on the fluid proto::VarType, so we use the dtype enum
+// value directly. See also `assign_value_sig.cc`.
+// proto::VarType::INT16 -> 1  -> phi::DataType::INT16
+// proto::VarType::INT32 -> 2  -> phi::DataType::INT32
+// proto::VarType::INT64 -> 3  -> phi::DataType::INT64
+// proto::VarType::FP16 ->  4  -> phi::DataType::FLOAT16
+// proto::VarType::FP32 ->  5  -> phi::DataType::FLOAT32
+// proto::VarType::FP64 ->  6  -> phi::DataType::FLOAT64
+// proto::VarType::UINT8 -> 20 -> phi::DataType::UINT8
+static std::map<int, phi::DataType> var_type_map{{1, phi::DataType::INT16},
+                                                 {2, phi::DataType::INT32},
+                                                 {3, phi::DataType::INT64},
+                                                 {4, phi::DataType::FLOAT16},
+                                                 {5, phi::DataType::FLOAT32},
+                                                 {6, phi::DataType::FLOAT64},
+                                                 {20, phi::DataType::UINT8}};
+
 #define _PhiForEachDataTypeHelper_(callback, cpp_type, data_type) \
   callback(cpp_type, data_type);
 

paddle/phi/kernels/cpu/arg_min_max_kernel.cc

@@ -17,6 +17,7 @@
 #include "paddle/phi/backends/cpu/cpu_context.h"
 #include "paddle/phi/core/ddim.h"
 #include "paddle/phi/core/kernel_registry.h"
+#include "paddle/phi/core/utils/data_type.h"
 #include "paddle/phi/kernels/funcs/eigen/common.h"
 #include "paddle/phi/kernels/funcs/math_function.h"
 
@@ -141,15 +142,14 @@ void ArgMinMaxKernel(const Context& dev_ctx,
                      int dtype,
                      DenseTensor* out) {
   if (dtype < 0) {
-    paddle::framework::VisitDataTypeTiny(
-        static_cast<paddle::framework::proto::VarType::Type>(
-            paddle::framework::proto::VarType::INT64),
+    phi::VisitDataTypeTiny(
+        phi::DataType::INT64,
         VisitDataArgMinMaxFunctor<Context, T, EnumArgMinMaxValue>(
             dev_ctx, x, axis.to<int64_t>(), keepdims, flatten, out));
     return;
   }
-  paddle::framework::VisitDataTypeTiny(
-      static_cast<paddle::framework::proto::VarType::Type>(dtype),
+  phi::VisitDataTypeTiny(
+      var_type_map[dtype],
       VisitDataArgMinMaxFunctor<Context, T, EnumArgMinMaxValue>(
           dev_ctx, x, axis.to<int64_t>(), keepdims, flatten, out));
 }

paddle/phi/kernels/cpu/cumprod_grad_kernel.cc

@@ -23,7 +23,7 @@
 #include "paddle/phi/kernels/funcs/for_range.h"
 
 // NOTE(@xiongkun): use of IsComplex<>
-#include "paddle/fluid/framework/data_type.h"
+#include "paddle/phi/core/utils/data_type.h"
 
 namespace phi {
 template <typename T, typename Context>
@@ -51,7 +51,7 @@ void CumprodGradKernel(const Context& dev_ctx,
   const T* out_data_deal;
   Allocator::AllocationPtr x_conj;
   Allocator::AllocationPtr out_conj;
-  if (paddle::framework::IsComplex<T>::value) {
+  if (phi::IsComplexType(x.dtype())) {
     x_conj = const_cast<Allocator&>(dev_ctx.GetAllocator())
                  .Allocate(numel * sizeof(T));
     auto* x_data_conj = reinterpret_cast<T*>(x_conj->ptr());

paddle/phi/kernels/cpu/unique_consecutive_kernel.cc

@@ -18,8 +18,7 @@
 #include "paddle/phi/backends/cpu/cpu_context.h"
 #include "paddle/phi/core/errors.h"
 #include "paddle/phi/core/kernel_registry.h"
-
-#include "paddle/fluid/framework/data_type.h"
+#include "paddle/phi/core/utils/data_type.h"
 
 namespace phi {
 
@@ -33,8 +32,8 @@ void UniqueConsecutiveKernel(const Context& dev_ctx,
                              DenseTensor* out,
                              DenseTensor* index,
                              DenseTensor* counts) {
-  auto data_type = static_cast<paddle::framework::proto::VarType::Type>(dtype);
-  if (data_type == paddle::framework::proto::VarType::INT32) {
+  auto data_type = var_type_map[dtype];
+  if (data_type == phi::DataType::INT32) {
     PADDLE_ENFORCE_LE(
         x.numel(),
         INT_MAX,
@@ -46,13 +45,13 @@ void UniqueConsecutiveKernel(const Context& dev_ctx,
   }
 
   if (axis.empty()) {
-    paddle::framework::VisitDataTypeTiny(
+    phi::VisitDataTypeTiny(
         data_type,
         UniqueConsecutiveFlattenedTensorFunctor<Context, T>(
             dev_ctx, x, out, return_inverse, return_counts, index, counts));
   } else {
     int valid_axis = axis[0];
-    paddle::framework::VisitDataTypeTiny(
+    phi::VisitDataTypeTiny(
         data_type,
         UniqueConsecutiveDimFunctor<Context, T>(dev_ctx,
                                                 x,

paddle/phi/kernels/funcs/math_function.cc

@@ -26,10 +26,10 @@ limitations under the License. */
 #include <utility>
 #include <vector>
 
-#include "paddle/fluid/framework/data_type.h"
 #include "paddle/fluid/platform/bfloat16.h"
 #include "paddle/fluid/platform/float16.h"
 #include "paddle/phi/backends/cpu/cpu_context.h"
+#include "paddle/phi/common/data_type.h"
 #include "paddle/phi/kernels/funcs/eigen/common.h"
 #include "paddle/phi/kernels/funcs/math_function_impl.h"
 #include "unsupported/Eigen/CXX11/Tensor"

paddle/phi/kernels/funcs/math_function.cu

@@ -14,12 +14,12 @@ limitations under the License. */
 #include <algorithm>
 #include <vector>
 
-#include "paddle/fluid/framework/data_type.h"
 #include "paddle/fluid/memory/malloc.h"
 #include "paddle/fluid/memory/memcpy.h"
 #include "paddle/fluid/platform/bfloat16.h"
 #include "paddle/fluid/platform/float16.h"
 #include "paddle/phi/backends/gpu/gpu_context.h"
+#include "paddle/phi/common/data_type.h"
 #include "paddle/phi/kernels/funcs/blas/blas.h"
 #include "paddle/phi/kernels/funcs/math_function.h"
 #include "paddle/phi/kernels/funcs/math_function_impl.h"

paddle/phi/kernels/funcs/math_function_impl.h

@@ -16,7 +16,7 @@ limitations under the License. */
 #include <memory>
 #include <vector>
 
-#include "paddle/fluid/framework/data_type.h"
+#include "paddle/phi/common/data_type.h"
 #include "paddle/phi/kernels/funcs/math_function.h"
 
 namespace phi {

paddle/phi/kernels/gpu/arg_min_max_kernel.cu

@@ -28,9 +28,8 @@ namespace cub = hipcub;
 #endif
 #include <limits>
 
-#include "paddle/fluid/framework/data_type.h"
 #include "paddle/phi/core/ddim.h"
-
+#include "paddle/phi/core/utils/data_type.h"
 namespace phi {
 
 namespace {  // NOLINT
@@ -209,15 +208,14 @@ void ArgMinMaxOpCUDAKernel(const Context& dev_ctx,
                            int dtype,
                            DenseTensor* out) {
   if (dtype < 0) {
-    paddle::framework::VisitDataTypeTiny(
-        static_cast<paddle::framework::proto::VarType::Type>(
-            paddle::framework::proto::VarType::INT64),
+    phi::VisitDataTypeTiny(
+        phi::DataType::INT64,
         VisitDataCudaArgMinMaxFunctor<Context, T, Reducer>(
             dev_ctx, x, axis.to<int64_t>(), keepdims, flatten, out));
     return;
   }
-  paddle::framework::VisitDataTypeTiny(
-      static_cast<paddle::framework::proto::VarType::Type>(dtype),
+  phi::VisitDataTypeTiny(
+      var_type_map[dtype],
       VisitDataCudaArgMinMaxFunctor<Context, T, Reducer>(
           dev_ctx, x, axis.to<int64_t>(), keepdims, flatten, out));
 }

paddle/phi/kernels/gpu/cumprod_grad_kernel.cu

@@ -24,7 +24,7 @@
 #include "paddle/phi/kernels/funcs/elementwise_functor.h"
 #include "paddle/phi/kernels/funcs/for_range.h"
 // NOTE(@xiongkun): use of IsComplex<>
-#include "paddle/fluid/framework/data_type.h"
+#include "paddle/phi/core/utils/data_type.h"
 
 namespace phi {
 
@@ -152,7 +152,7 @@ void CumprodGradKernel(const Context &dev_ctx,
   const T *y_data_deal;
   Allocator::AllocationPtr x_conj;
   Allocator::AllocationPtr y_conj;
-  if (paddle::framework::IsComplex<T>::value) {
+  if (phi::IsComplexType(x.dtype())) {
     x_conj = const_cast<Allocator &>(dev_ctx.GetAllocator())
                  .Allocate(numel * sizeof(T));
     auto *x_data_conj = reinterpret_cast<T *>(x_conj->ptr());

paddle/phi/kernels/gpu/unique_consecutive_kernel.cu

@@ -21,8 +21,6 @@
 #include "paddle/phi/core/errors.h"
 #include "paddle/phi/core/kernel_registry.h"
 
-#include "paddle/fluid/framework/data_type.h"
-
 namespace phi {
 
 template <typename T, typename Context>
@@ -35,8 +33,8 @@ void UniqueConsecutiveKernel(const Context& dev_ctx,
                              DenseTensor* out,
                              DenseTensor* index,
                              DenseTensor* counts) {
-  auto data_type = static_cast<paddle::framework::proto::VarType::Type>(dtype);
-  if (data_type == paddle::framework::proto::VarType::INT32) {
+  auto data_type = var_type_map[dtype];
+  if (data_type == phi::DataType::INT32) {
     PADDLE_ENFORCE_LE(
         x.numel() + 1,
         INT_MAX,
@@ -49,14 +47,14 @@ void UniqueConsecutiveKernel(const Context& dev_ctx,
 
   // if 'axis' is not required, flatten the Tensor.
   if (axis.empty()) {
-    paddle::framework::VisitDataTypeTiny(
+    phi::VisitDataTypeTiny(
         data_type,
         UniqueConsecutiveFlattenedCUDAFunctor<Context, T>(
             dev_ctx, x, out, return_inverse, return_counts, index, counts));
   } else {
     // 'axis' is required.
     int valid_axis = axis[0];
-    paddle::framework::VisitDataTypeTiny(
+    phi::VisitDataTypeTiny(
         data_type,
         UniqueConsecutiveDimsCUDAFunctor<Context, T>(dev_ctx,
                                                      x,

paddle/phi/kernels/impl/isclose_kernel_impl.h

@@ -16,7 +16,6 @@
 #include <cmath>
 #include <string>
 
-#include "paddle/fluid/framework/data_type.h"
 #include "paddle/phi/backends/cpu/cpu_context.h"
 #include "paddle/phi/backends/gpu/gpu_context.h"
 #include "paddle/phi/common/data_type.h"

paddle/phi/kernels/xpu/arg_min_max_kernel.cc

@@ -17,6 +17,7 @@
 #include "paddle/phi/backends/xpu/xpu_context.h"
 #include "paddle/phi/core/ddim.h"
 #include "paddle/phi/core/kernel_registry.h"
+#include "paddle/phi/core/utils/data_type.h"
 
 namespace phi {