From cd93ed81bc4e238fea0a04942fd5cfe7ff834827 Mon Sep 17 00:00:00 2001
From: Joshua Deng <joshuadeng@meta.com>
Date: Thu, 30 May 2024 09:45:10 -0700
Subject: [PATCH] update cmake for dense TBE VBE support (#2641)

Summary:

follow up to https://github.com/pytorch/FBGEMM/pull/2620

Reviewed By: sryap

Differential Revision: D57938517
---
 fbgemm_gpu/FbgemmGpu.cmake | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/fbgemm_gpu/FbgemmGpu.cmake b/fbgemm_gpu/FbgemmGpu.cmake
index f41f976284..b394800817 100644
--- a/fbgemm_gpu/FbgemmGpu.cmake
+++ b/fbgemm_gpu/FbgemmGpu.cmake
@@ -90,7 +90,8 @@ set(GPU_OPTIMIZERS ${COMMON_OPTIMIZERS} ${GPU_ONLY_OPTIMIZERS})
 set(VBE_OPTIMIZERS
     rowwise_adagrad
     rowwise_adagrad_with_counter
-    sgd)
+    sgd
+    dense)
 
 # Optimizers with the GWD support
 set(GWD_OPTIMIZERS
@@ -151,6 +152,8 @@ set(gen_gpu_kernel_source_files
     "gen_embedding_backward_dense_indice_weights_codegen_cuda.cu"
     "gen_embedding_backward_split_indice_weights_codegen_cuda.cu"
     "gen_embedding_backward_ssd_indice_weights_codegen_cuda.cu"
+    "gen_embedding_forward_dense_weighted_vbe_codegen_cuda.cu"
+    "gen_embedding_forward_dense_unweighted_vbe_codegen_cuda.cu"
     "gen_embedding_forward_split_weighted_vbe_codegen_cuda.cu"
     "gen_embedding_forward_split_unweighted_vbe_codegen_cuda.cu"
     "gen_batch_index_select_dim0_forward_codegen_cuda.cu"
@@ -203,7 +206,9 @@ endforeach()
 foreach(wdesc weighted unweighted)
   list(APPEND gen_gpu_kernel_source_files
       "gen_embedding_forward_split_${wdesc}_vbe_kernel.cu"
-      "gen_embedding_backward_split_${wdesc}_vbe_device_kernel.cuh")
+      "gen_embedding_backward_split_${wdesc}_vbe_device_kernel.cuh"
+      "gen_embedding_forward_dense_${wdesc}_vbe_kernel.cu")
+
 endforeach()
 
 # Generate GWD files
@@ -261,6 +266,9 @@ foreach(optimizer ${ALL_OPTIMIZERS})
     "gen_embedding_backward_split_${optimizer}_pt2_cuda_wrapper.cpp")
 endforeach()
 
+list(APPEND gen_gpu_host_source_files
+    "gen_embedding_backward_split_dense.cpp")
+
 foreach(optimizer ${CPU_OPTIMIZERS})
   list(APPEND gen_cpu_source_files
     "gen_embedding_backward_${optimizer}_split_cpu.cpp")
@@ -462,7 +470,6 @@ set(fbgemm_gpu_sources_static_cpu
 if(NOT FBGEMM_CPU_ONLY)
   list(APPEND fbgemm_gpu_sources_static_cpu
     codegen/inference/embedding_forward_quantized_host.cpp
-    codegen/training/backward/embedding_backward_dense_host.cpp
     codegen/utils/embedding_bounds_check_host.cpp
     src/memory_utils/memory_utils.cpp
     src/memory_utils/memory_utils_ops.cpp