Move tensor utilities out of sparse_ops_utils.h (pytorch#3019)

Summary: Pull Request resolved: pytorch#3019 X-link: facebookresearch/FBGEMM#115 - Move tensor utilities out of sparse_ops_utils.h Reviewed By: spcyppt Differential Revision: D61572323 fbshipit-source-id: 3eb8f2ce7746064d2a2fc445e62aec52d67287bf
spcyppt · Aug 21, 2024 · 2e190b4 · 2e190b4
1 parent d3af64c
commit 2e190b4
Show file tree

Hide file tree

Showing 37 changed files with 339 additions and 325 deletions.
diff --git a/fbgemm_gpu/codegen/inference/embedding_forward_quantized_cpu_template.cpp b/fbgemm_gpu/codegen/inference/embedding_forward_quantized_cpu_template.cpp
@@ -17,7 +17,7 @@
 #include "fbgemm_gpu/utils/dispatch_macros.h"
 #include "fbgemm_gpu/embedding_common.h"
 #include "fbgemm/FbgemmEmbedding.h"
-#include "fbgemm_gpu/sparse_ops_utils.h"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 
 #if defined(__x86_64__) || defined(__i386__) || (defined(_MSC_VER) && (defined(_M_X64) || defined(_M_IX86)))
 #include <immintrin.h>

diff --git a/fbgemm_gpu/codegen/training/backward/embedding_backward_split_host_cpu_template.cpp b/fbgemm_gpu/codegen/training/backward/embedding_backward_split_host_cpu_template.cpp
@@ -13,7 +13,7 @@
 
 #include "fbgemm_gpu/embedding_forward_split_cpu.h"
 #include "fbgemm_gpu/embedding_common.h"
-#include "fbgemm_gpu/sparse_ops_utils.h"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 
 using Tensor = at::Tensor;
 using namespace fbgemm_gpu;

diff --git a/fbgemm_gpu/codegen/training/backward/embedding_backward_split_indice_weights_template.cu b/fbgemm_gpu/codegen/training/backward/embedding_backward_split_indice_weights_template.cu
@@ -17,9 +17,9 @@
 ////////////////////////////////////////////////////////////////////////////////
 // Required for op registrations
 ////////////////////////////////////////////////////////////////////////////////
-#include "fbgemm_gpu/utils/ops_utils.h"
 #include "fbgemm_gpu/embedding_forward_template_helpers.cuh"
-#include "fbgemm_gpu/sparse_ops_utils.h"
+#include "fbgemm_gpu/utils/ops_utils.h"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 
 using Tensor = at::Tensor;
 using namespace fbgemm_gpu;

diff --git a/fbgemm_gpu/codegen/training/backward/embedding_backward_split_meta_template.cpp b/fbgemm_gpu/codegen/training/backward/embedding_backward_split_meta_template.cpp
@@ -29,7 +29,7 @@
 ////////////////////////////////////////////////////////////////////////////////
 #include <torch/library.h>
 #include "fbgemm_gpu/embedding_common.h"
-#include "fbgemm_gpu/sparse_ops_utils.h"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 #include "fbgemm_gpu/split_embeddings_utils.cuh"
 
 using namespace fbgemm_gpu;

diff --git a/fbgemm_gpu/codegen/training/forward/embedding_forward_split_meta_template.cpp b/fbgemm_gpu/codegen/training/forward/embedding_forward_split_meta_template.cpp
@@ -26,9 +26,9 @@
 ////////////////////////////////////////////////////////////////////////////////
 // Required for op registrations
 ////////////////////////////////////////////////////////////////////////////////
-#include "fbgemm_gpu/utils/ops_utils.h"
-#include "fbgemm_gpu/sparse_ops_utils.h"
 #include "fbgemm_gpu/embedding_common.h"
+#include "fbgemm_gpu/utils/ops_utils.h"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 
 using namespace fbgemm_gpu;
 using Tensor = at::Tensor;

diff --git a/fbgemm_gpu/codegen/training/index_select/batch_index_select_dim0_host.cpp b/fbgemm_gpu/codegen/training/index_select/batch_index_select_dim0_host.cpp
@@ -14,8 +14,8 @@
 
 #include "fbgemm_gpu/embedding_common.h"
 #include "fbgemm_gpu/sparse_ops.h"
-#include "fbgemm_gpu/sparse_ops_utils.h"
 #include "fbgemm_gpu/utils/ops_utils.h"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 
 using Tensor = at::Tensor;
 using namespace fbgemm_gpu;

diff --git a/fbgemm_gpu/codegen/utils/embedding_bounds_check_host_cpu.cpp b/fbgemm_gpu/codegen/utils/embedding_bounds_check_host_cpu.cpp
@@ -11,9 +11,9 @@
 #include <ATen/core/op_registration/op_registration.h>
 #include <torch/script.h>
 #include "fbgemm_gpu/embedding_common.h"
-#include "fbgemm_gpu/sparse_ops_utils.h"
 #include "fbgemm_gpu/utils/dispatch_macros.h"
 #include "fbgemm_gpu/utils/ops_utils.h"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 
 using Tensor = at::Tensor;
 using namespace fbgemm_gpu;

diff --git a/fbgemm_gpu/include/fbgemm_gpu/embedding_backward_template_helpers.cuh b/fbgemm_gpu/include/fbgemm_gpu/embedding_backward_template_helpers.cuh
@@ -24,11 +24,11 @@
 #include <mutex>
 
 #include "fbgemm_gpu/embedding_common.h"
-#include "fbgemm_gpu/sparse_ops_utils.h"
 #include "fbgemm_gpu/utils/dispatch_macros.h"
 #include "fbgemm_gpu/utils/find_qparams.cuh"
 #include "fbgemm_gpu/utils/fixed_divisor.cuh"
 #include "fbgemm_gpu/utils/shared_memory.cuh"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 #include "fbgemm_gpu/utils/vec4.cuh"
 #include "fbgemm_gpu/utils/weight_row.cuh"
 

diff --git a/fbgemm_gpu/include/fbgemm_gpu/embedding_forward_template_helpers.cuh b/fbgemm_gpu/include/fbgemm_gpu/embedding_forward_template_helpers.cuh
@@ -31,11 +31,11 @@
 #include <mutex>
 
 #include "fbgemm_gpu/embedding_common.h"
-#include "fbgemm_gpu/sparse_ops_utils.h"
 #include "fbgemm_gpu/utils/dispatch_macros.h"
 #include "fbgemm_gpu/utils/find_qparams.cuh"
 #include "fbgemm_gpu/utils/fixed_divisor.cuh"
 #include "fbgemm_gpu/utils/tensor_accessor.h"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 #include "fbgemm_gpu/utils/vec4.cuh"
 #include "fbgemm_gpu/utils/vec4acc.cuh"
 #include "fbgemm_gpu/utils/vecn.cuh"

diff --git a/fbgemm_gpu/include/fbgemm_gpu/embedding_inplace_update.h b/fbgemm_gpu/include/fbgemm_gpu/embedding_inplace_update.h
@@ -13,7 +13,6 @@
 #include <torch/torch.h>
 
 #include "fbgemm_gpu/embedding_common.h"
-#include "fbgemm_gpu/sparse_ops_utils.h"
 
 using Tensor = at::Tensor;
 

diff --git a/fbgemm_gpu/include/fbgemm_gpu/permute_multi_embedding_function.h b/fbgemm_gpu/include/fbgemm_gpu/permute_multi_embedding_function.h
@@ -13,9 +13,9 @@
 #include <torch/csrc/api/include/torch/types.h>
 #include <torch/csrc/autograd/custom_function.h>
 
-#include "fbgemm_gpu/sparse_ops_utils.h"
 #include "fbgemm_gpu/utils/dispatch_macros.h"
 #include "fbgemm_gpu/utils/ops_utils.h"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 
 namespace fbgemm_gpu {
 

diff --git a/fbgemm_gpu/include/fbgemm_gpu/permute_pooled_embedding_ops.h b/fbgemm_gpu/include/fbgemm_gpu/permute_pooled_embedding_ops.h
@@ -12,8 +12,8 @@
 #include <torch/csrc/api/include/torch/types.h>
 #include <torch/csrc/autograd/custom_function.h>
 #include "fbgemm_gpu/permute_pooled_embedding_function.h"
-#include "fbgemm_gpu/sparse_ops_utils.h"
 #include "fbgemm_gpu/utils/ops_utils.h"
+#include "fbgemm_gpu/utils/tensor_utils.h"
 
 /// @defgroup permute-pooled-embs-gpu Permute Pooled Embeddings Operators (CUDA)
 /// @defgroup permute-pooled-embs-cpu Permute Pooled Embeddings Operators (CPU)