data: cat SGEMM files from 3 repos into cat_files/

siboehm/SGEMM_CUDA (19 files):
  siboehm_sgemm.cu, siboehm_runner.cu, siboehm_runner.cuh, siboehm_kernels.cuh
  siboehm_cuBLAS_sgemm.cu, siboehm_simplest_kernel.cu, siboehm_CMakeLists.txt
  siboehm_{1_naive..12_kernel_double_buffering}.cuh

wangzyon/NVIDIA_SGEMM_PRACTICE (12 files):
  wangzyon_sgemm.cu, wangzyon_utils.cu, wangzyon_utils.cuh, wangzyon_kernel.cuh
  wangzyon_CMakeLists.txt, wangzyon_kernel_{1..7}.cuh

edtallison/sgemm-cuda (19 files):
  edtallison_sgemm.cu, edtallison_runner.cu, edtallison_runner.cuh
  edtallison_kernels.cuh, edtallison_cuBLAS_sgemm.cu, edtallison_simplest_kernel.cu
  edtallison_CMakeLists.txt, edtallison_{01_naive..12_kernel_double_buffering}.cuh

cat_files/ total: 25 → 75 files

cat_files/wangzyon_CMakeLists.txt

@@ -0,0 +1,36 @@
+cmake_minimum_required(VERSION 3.0)
+project(NVIDIA_SGEMM_PRACTICE)
+
+# gcc/g++编译参数说明：
+# -O1~3编译器优化选项的4个级别，-O1默认，级别越大优化效果越好，但编译时间越长;
+# -std=c++11，采用C++11标准编译
+set(CMAKE_CXX_FLAGS "-O3 -std=c++11")
+
+# nvcc编译参数说明：
+# -g:主机代码添加调试信息;
+# -G:设备代码产生调试信息，将会禁用大多数编译器优化，造成设备代码运行缓慢;
+# -Xptxas -dlcm=ca启用L1缓存，-Xptxas -dlcm=cg关闭L1缓存
+
+# set(CUDA_NVCC_FLAGS -g;-G;-Xptxas;-dlcm=ca)
+# set(CUDA_NVCC_FLAGS -Xptxas;-dlcm=cg)
+set(CUDA_NVCC_FLAGS -arch=compute_70;-code=compute_70)
+
+# 若FIND CUDA ERROR，在~/.bashrc中添加配置环境变量和动态库路径
+# CUDA_HOME=/usr/local/cuda
+# export PATH=$CUDA_HOME/bin:$PATH
+# export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH
+find_package(CUDA REQUIRED)
+
+# 配置头文件搜索路径
+include_directories(${CUDA_INCLUDE_DIRS})
+include_directories(${PROJECT_SOURCE_DIR}/src)
+# 配置待编译的源文件路径
+aux_source_directory(${PROJECT_SOURCE_DIR}/src SRC)
+# 可执行文件输出路径
+set(EXECUTABLE_OUTPUT_PATH ${PROJECT_SOURCE_DIR})
+# 生成可执行文件
+CUDA_ADD_EXECUTABLE(sgemm sgemm.cu ${SRC})
+
+# link cudart cublas
+target_link_libraries(sgemm ${CUDA_LIBRARIES} ${CUDA_cublas_LIBRARY})
+