support cuda11. test=develop (#6545)

Author: jiweibo

cmake/cuda.cmake

@@ -6,11 +6,9 @@ if(WITH_CUDA_FP16)
   add_definitions("-DCUDA_WITH_FP16")
 endif()
 
-set(paddle_known_gpu_archs "30 35 50 52 60 61 70")
-set(paddle_known_gpu_archs7 "30 35 50 52")
-set(paddle_known_gpu_archs8 "30 35 50 52 53 60 61 62")
-set(paddle_known_gpu_archs9 "30 35 50 52 53 60 61 62 70")
-set(paddle_known_gpu_archs10 "30 35 50 52 53 60 61 62 70 72 75")
+set(paddle_known_gpu_archs "35 50 52 60 61 70 75 80")
+set(paddle_known_gpu_archs10 "35 50 52 60 61 70 75")
+set(paddle_known_gpu_archs11 "52 60 61 70 75 80")
 
 ######################################################################################
 # A function for automatic detection of GPUs installed  (if autodetection is enabled)
@@ -65,8 +63,8 @@ endfunction()
 #   select_nvcc_arch_flags(out_variable)
 function(select_nvcc_arch_flags out_variable)
   # List of arch names
-  set(archs_names "Kepler" "Maxwell" "Pascal" "Volta" "Turing" "All" "Manual")
-  set(archs_name_default "All")
+  set(archs_names "Kepler" "Maxwell" "Pascal" "Volta" "Turing" "Ampere" "All" "Manual")
+  set(archs_name_default "Auto")
   list(APPEND archs_names "Auto")
 
   # set CUDA_ARCH_NAME strings (so it will be seen as dropbox in CMake-Gui)
@@ -82,7 +80,7 @@ function(select_nvcc_arch_flags out_variable)
 
   if(${CUDA_ARCH_NAME} STREQUAL "Manual")
     set(CUDA_ARCH_BIN ${paddle_known_gpu_archs} CACHE STRING "Specify 'real' GPU architectures to build binaries for, BIN(PTX) format is supported")
-    set(CUDA_ARCH_PTX "50"                     CACHE STRING "Specify 'virtual' PTX architectures to build PTX intermediate code for")
+    set(CUDA_ARCH_PTX ""                     CACHE STRING "Specify 'virtual' PTX architectures to build PTX intermediate code for")
     mark_as_advanced(CUDA_ARCH_BIN CUDA_ARCH_PTX)
   else()
     unset(CUDA_ARCH_BIN CACHE)
@@ -99,6 +97,8 @@ function(select_nvcc_arch_flags out_variable)
     set(cuda_arch_bin "70")
   elseif(${CUDA_ARCH_NAME} STREQUAL "Turing")
     set(cuda_arch_bin "75")
+  elseif(${CUDA_ARCH_NAME} STREQUAL "Ampere")
+    set(cuda_arch_bin "80")
   elseif(${CUDA_ARCH_NAME} STREQUAL "All")
     set(cuda_arch_bin ${paddle_known_gpu_archs})
   elseif(${CUDA_ARCH_NAME} STREQUAL "Auto")
@@ -143,32 +143,24 @@ function(select_nvcc_arch_flags out_variable)
 endfunction()
 
 message(STATUS "CUDA detected: " ${CUDA_VERSION})
-if (${CUDA_VERSION} LESS 7.0)
-  set(paddle_known_gpu_archs ${paddle_known_gpu_archs})
-  add_definitions("-DPADDLE_CUDA_BINVER=\"60\"")
-elseif (${CUDA_VERSION} LESS 8.0) # CUDA 7.x
-  set(paddle_known_gpu_archs ${paddle_known_gpu_archs7})
-  list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
-  list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
-  add_definitions("-DPADDLE_CUDA_BINVER=\"70\"")
-elseif (${CUDA_VERSION} LESS 9.0) # CUDA 8.x
-  set(paddle_known_gpu_archs ${paddle_known_gpu_archs8})
+if (${CUDA_VERSION} LESS 11.0) # CUDA 10.x
+  set(paddle_known_gpu_archs ${paddle_known_gpu_archs10})
   list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
   list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
-  # CUDA 8 may complain that sm_20 is no longer supported. Suppress the
-  # warning for now.
   list(APPEND CUDA_NVCC_FLAGS "-Wno-deprecated-gpu-targets")
-  add_definitions("-DPADDLE_CUDA_BINVER=\"80\"")
-elseif (${CUDA_VERSION} LESS 10.0) # CUDA 9.x
-  set(paddle_known_gpu_archs ${paddle_known_gpu_archs9})
+  add_definitions("-DPADDLE_CUDA_BINVER=\"100\"")
+elseif (${CUDA_VERSION} LESS 11.2) # CUDA 11.0/11.1
+  set(paddle_known_gpu_archs ${paddle_known_gpu_archs11})
   list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
   list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
-  add_definitions("-DPADDLE_CUDA_BINVER=\"90\"")
-elseif (${CUDA_VERSION} LESS 11.0) # CUDA 10.x
-  set(paddle_known_gpu_archs ${paddle_known_gpu_archs10})
+  list(APPEND CUDA_NVCC_FLAGS "-Wno-deprecated-gpu-targets")
+  add_definitions("-DPADDLE_CUDA_BINVER=\"110\"")
+elseif (${CUDA_VERSION} LESS 12.0) # CUDA 11.2+
+  set(paddle_known_gpu_archs "${paddle_known_gpu_archs11} 86")
   list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
   list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
-  add_definitions("-DPADDLE_CUDA_BINVER=\"100\"")
+  list(APPEND CUDA_NVCC_FLAGS "-Wno-deprecated-gpu-targets")
+  add_definitions("-DPADDLE_CUDA_BINVER=\"110\"")
 endif()
 
 if (CUDA_WITH_FP16)