Revert "build flash-attn whl (PaddlePaddle#33)" (PaddlePaddle#39)

This reverts commit 4b554d0.
kircle888 · Apr 19, 2024 · ce8a8fe · ce8a8fe
1 parent 5341f98
commit ce8a8fe
Show file tree

Hide file tree

Showing 6 changed files with 10 additions and 215 deletions.
diff --git a/csrc/CMakeLists.txt b/csrc/CMakeLists.txt
@@ -1,24 +1,15 @@
 cmake_minimum_required(VERSION 3.9 FATAL_ERROR)
 project(flash-attention LANGUAGES CXX CUDA)
-set(CMAKE_CXX_STANDARD 17)
-set(CMAKE_CXX_STANDARD_REQUIRED ON)
-
 
 find_package(Git QUIET REQUIRED)
 
 execute_process(COMMAND ${GIT_EXECUTABLE} submodule update --init --recursive
                 WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}
                 RESULT_VARIABLE GIT_SUBMOD_RESULT)
 
-#cmake -DWITH_ADVANCED=ON
-if (WITH_ADVANCED)
-  add_compile_definitions(PADDLE_WITH_ADVANCED)
-endif()
-
 add_definitions("-DFLASH_ATTN_WITH_TORCH=0")
 
 set(CUTLASS_3_DIR ${CMAKE_CURRENT_SOURCE_DIR}/cutlass)
-set(BINARY_DIR ${CMAKE_BINARY_DIR})
 
 set(FA2_SOURCES_CU
      flash_attn/src/cuda_utils.cu
@@ -64,7 +55,6 @@ target_include_directories(flashattn PRIVATE
     flash_attn
     ${CUTLASS_3_DIR}/include)
 
-if (WITH_ADVANCED)
 set(FA1_SOURCES_CU
     flash_attn_with_bias_and_mask/flash_attn_with_bias_mask.cu
     flash_attn_with_bias_and_mask/src/cuda_utils.cu
@@ -75,12 +65,6 @@ set(FA1_SOURCES_CU
     flash_attn_with_bias_and_mask/src/fmha_bwd_with_mask_bias_hdim64.cu
     flash_attn_with_bias_and_mask/src/fmha_bwd_with_mask_bias_hdim128.cu
     flash_attn_with_bias_and_mask/src/utils.cu)
-else()
-set(FA1_SOURCES_CU
-    flash_attn_with_bias_and_mask/flash_attn_with_bias_mask.cu
-    flash_attn_with_bias_and_mask/src/cuda_utils.cu
-    flash_attn_with_bias_and_mask/src/utils.cu)
-endif()
 
 add_library(flashattn_with_bias_mask STATIC
     flash_attn_with_bias_and_mask/
@@ -99,14 +83,18 @@ target_link_libraries(flashattn flashattn_with_bias_mask)
 
 add_dependencies(flashattn flashattn_with_bias_mask)
 
-set(NVCC_ARCH_BIN 80 CACHE STRING "CUDA architectures")
 
-message("NVCC_ARCH_BIN is set to: ${NVCC_ARCH_BIN}")
+if (NOT DEFINED NVCC_ARCH_BIN)
+  message(FATAL_ERROR "NVCC_ARCH_BIN is not defined.")
+endif()
+
+if (NVCC_ARCH_BIN STREQUAL "")
+  message(FATAL_ERROR "NVCC_ARCH_BIN is not set.")
+endif()
 
 STRING(REPLACE "-" ";" FA_NVCC_ARCH_BIN ${NVCC_ARCH_BIN})
 
 set(FA_GENCODE_OPTION "SHELL:")
-
 foreach(arch ${FA_NVCC_ARCH_BIN})
    if(${arch} GREATER_EQUAL 80)
      set(FA_GENCODE_OPTION "${FA_GENCODE_OPTION} -gencode arch=compute_${arch},code=sm_${arch}")
@@ -143,35 +131,7 @@ target_compile_options(flashattn_with_bias_mask PRIVATE $<$<COMPILE_LANGUAGE:CUD
   "${FA_GENCODE_OPTION}"
   >)
 
-
 INSTALL(TARGETS flashattn
         LIBRARY DESTINATION "lib")
 
 INSTALL(FILES capi/flash_attn.h DESTINATION "include")
-
-if (WITH_ADVANCED)
-  if(WIN32)
-    set(target_output_name "flashattn")
-  else()
-    set(target_output_name "libflashattn")
-  endif()
-  set_target_properties(flashattn PROPERTIES
-    OUTPUT_NAME ${target_output_name}_advanced
-    PREFIX ""
-  )
-
-  configure_file(${CMAKE_SOURCE_DIR}/env_dict.py.in ${CMAKE_SOURCE_DIR}/env_dict.py @ONLY)
-  set_target_properties(flashattn PROPERTIES
-      LIBRARY_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/paddle_flash_attn/
-  )
-  add_custom_target(build_whl
-      COMMAND ${CMAKE_COMMAND} -E env python ${CMAKE_SOURCE_DIR}/setup.py bdist_wheel
-      WORKING_DIRECTORY ${CMAKE_BINARY_DIR}
-      DEPENDS flashattn
-      COMMENT "Running build wheel"
-  )
-
-  add_custom_target(default_target DEPENDS build_whl)
-
-  set_property(DIRECTORY PROPERTY DEFAULT_TARGET default_target)
-endif()
diff --git a/csrc/env_dict.py.in b/csrc/env_dict.py.in
diff --git a/csrc/flash_attn/src/flash_bwd_launch_template.h b/csrc/flash_attn/src/flash_bwd_launch_template.h
@@ -68,8 +68,8 @@ void run_flash_bwd_seqk_parallel(Flash_bwd_params &params, cudaStream_t stream,
     BOOL_SWITCH(params.is_causal, IsCausalConst, [&] {
         BOOL_SWITCH(is_even_MN, IsEvenMNConst, [&] {
             BOOL_SWITCH(is_even_K, IsEvenKConst, [&] {
-                BOOL_SWITCH_ADVANCED(is_attn_mask, Is_attn_mask, [&] {
-                    BOOL_SWITCH_ADVANCED(is_deterministic, Is_deterministic, [&] {
+                BOOL_SWITCH(is_attn_mask, Is_attn_mask, [&] {
+                    BOOL_SWITCH(is_deterministic, Is_deterministic, [&] {
                         auto kernel = &flash_bwd_dq_dk_dv_loop_seqk_parallel_kernel<Kernel_traits, Is_dropout, IsCausalConst, IsEvenMNConst, IsEvenKConst, Is_attn_mask && !IsCausalConst, Is_deterministic>;
                         // auto kernel = &flash_bwd_dq_dk_dv_loop_seqk_parallel_kernel<Kernel_traits, Is_dropout, IsCausalConst, IsEvenMNConst, true>;
                         if (smem_size_dq_dk_dv >= 48 * 1024)  {

diff --git a/csrc/flash_attn/src/flash_fwd_launch_template.h b/csrc/flash_attn/src/flash_fwd_launch_template.h
@@ -40,7 +40,7 @@ void run_flash_fwd(Flash_fwd_params &params, cudaStream_t stream) {
     BOOL_SWITCH(is_even_N, IsEvenNConst, [&] {
         BOOL_SWITCH(is_even_K, IsEvenKConst, [&] {
             BOOL_SWITCH(return_softmax, ReturnSoftmaxConst, [&] {
-                BOOL_SWITCH_ADVANCED(is_attn_mask, Is_attn_mask, [&] {
+                BOOL_SWITCH(is_attn_mask, Is_attn_mask, [&] {
                     BOOL_SWITCH(is_equal_qk, Is_equal_seq_qk, [&] {
                         // Will only return softmax if dropout, to reduce compilation time.
                         auto kernel = &flash_fwd_kernel<Kernel_traits, Is_dropout, Is_causal, IsEvenNConst, IsEvenKConst, ReturnSoftmaxConst && Is_dropout, Is_attn_mask && !Is_causal, Is_equal_seq_qk>;

diff --git a/csrc/flash_attn/src/static_switch.h b/csrc/flash_attn/src/static_switch.h
@@ -25,25 +25,6 @@
     }                                           \
   }()
 
-#ifdef PADDLE_WITH_ADVANCED
-#define BOOL_SWITCH_ADVANCED(COND, CONST_NAME, ...) \
-  [&] {                                             \
-    if (COND) {                                     \
-      constexpr static bool CONST_NAME = true;      \
-      return __VA_ARGS__();                         \
-    } else {                                        \
-      constexpr static bool CONST_NAME = false;     \
-      return __VA_ARGS__();                         \
-    }                                               \
-  }()
-#else
-#define BOOL_SWITCH_ADVANCED(COND, CONST_NAME, ...) \
-  [&] {                                             \
-    constexpr static bool CONST_NAME = false;       \
-    return __VA_ARGS__();                           \
-  }()
-#endif
-
 #define FP16_SWITCH(COND, ...)               \
   [&] {                                      \
     if (COND) {                              \

diff --git a/csrc/setup.py b/csrc/setup.py