NE10/FFT/complex-non-power-of-2 NEON

ARM 64-bit (Cortex-A57) complex forward float LLVM 3.5 Time in ms | |kiss|opus|pffft|NE10| | C| C| NEON|NEON| 60| 129| 113| NA| 44| 120| 148| 127| NA| 49| 240| 151| 128| 55| 47| 480| 169| 142| 60| 55| 960| 183| 149| 65| 58| 1920| 193| 167| 71| 66| 3840| 217| 175| 76| 71| SNR > 100dB ARM 64-bit (Cortex-A53) complex forward float LLVM 3.5 Time in ms | |kiss|opus|pffft|NE10| | C| C| NEON|NEON| 60| 295| 311| NA| 72| 120| 368| 375| NA| 79| 240| 345| 342| 104| 77| 480| 415| 407| 115| 87| 960| 406| 378| 121| 95| 1920| 476| 441| 138| 113| 3840| 497| 424| 161| 126| SNR > 100dB ARM 32-bit (Cortex-A9) complex forward float LLVM 3.5 Time in ms | |kiss|opus|pffft|NE10| | C| C| NEON|NEON| 60| 224| 211| NA| 98| 120| 265| 245| NA| 104| 240| 262| 240| 130| 106| 480| 302| 274| 150| 122| 960| 305| 271| 162| 153| 1920| 369| 356| 230| 206| 3840| 415| 440| 282| 239| SNR > 100dB Change-Id: If9418041b01eed49dbdc8d6a18dd03f2c5684da8
projectNe10 · Dec 18, 2014 · 20b1896 · 20b1896
1 parent 46b1da1
commit 20b1896
Show file tree

Hide file tree

Showing 21 changed files with 1,946 additions and 316 deletions.
diff --git a/CMakeLists.txt b/CMakeLists.txt
@@ -26,7 +26,7 @@
 #
 cmake_minimum_required(VERSION 2.6)
 
-project(NE10 C ASM)
+project(NE10 C CXX ASM)
 
 option(NE10_BUILD_SHARED "Build NE10 shared libraries" OFF)
 option(NE10_BUILD_STATIC "Build NE10 static libraries" ON)
@@ -83,9 +83,6 @@ if(ANDROID_PLATFORM)
     ${ANDROID_TOOLCHAIN_PATH}/arm-linux-androideabi-as
     ${ANDROID_TOOLCHAIN_PATH}/arm-linux-androideabi-ar
     ${ANDROID_TOOLCHAIN_PATH}/arm-linux-androideabi-ranlib")
-    if(ANDROID_PLATFORM AND ANDROID_DEMO)
-      add_subdirectory(android/NE10Demo/jni)
-    endif()
 elseif(GNULINUX_PLATFORM)
     set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -mthumb-interwork -mthumb -march=armv7-a -mfpu=vfp3")
     set(CMAKE_ASM_FLAGS "${CMAKE_C_FLAGS} -mthumb-interwork -mthumb -march=armv7-a -mfpu=neon")
@@ -102,11 +99,10 @@ elseif(IOS_PLATFORM)
 
     string(REPLACE ";" "" CMAKE_C_FLAGS ${CMAKE_C_FLAGS})
     string(REPLACE ";" "" CMAKE_ASM_FLAGS ${CMAKE_ASM_FLAGS})
-    if(IOS_DEMO)
-        add_subdirectory(ios)
-    endif()
 endif()
 
+set(CMAKE_CXX_FLAGS "${CMAKE_ASM_FLAGS}")
+
 # The NE10 library.
 add_subdirectory(modules)
 
@@ -117,3 +113,11 @@ endif()
 if(NE10_BUILD_UNIT_TEST)
     add_subdirectory(test)
 endif()
+
+if(ANDROID_PLATFORM AND ANDROID_DEMO)
+    add_subdirectory(android/NE10Demo/jni)
+endif()
+
+if(IOS_PLATFORM AND IOS_DEMO)
+    add_subdirectory(ios)
+endif()
diff --git a/inc/NE10_dsp.h b/inc/NE10_dsp.h
@@ -46,6 +46,8 @@ extern "C" {
     /* fft functions*/
 
     /* function pointers*/
+    extern ne10_fft_cfg_float32_t (*ne10_fft_alloc_c2c_float32) (ne10_int32_t nfft);
+
     extern void (*ne10_fft_c2c_1d_float32) (ne10_fft_cpx_float32_t *fout,
                                             ne10_fft_cpx_float32_t *fin,
                                             ne10_fft_cfg_float32_t cfg,
@@ -92,7 +94,6 @@ extern "C" {
                                           ne10_int32_t scaled_flag);
 
     /* init functions*/
-    extern ne10_fft_cfg_float32_t ne10_fft_alloc_c2c_float32 (ne10_int32_t nfft);
     extern ne10_fft_cfg_int32_t ne10_fft_alloc_c2c_int32 (ne10_int32_t nfft);
     extern ne10_fft_cfg_int16_t ne10_fft_alloc_c2c_int16 (ne10_int32_t nfft);
 
@@ -101,6 +102,8 @@ extern "C" {
     extern ne10_fft_r2c_cfg_int16_t ne10_fft_alloc_r2c_int16 (ne10_int32_t nfft);
 
     /* C version*/
+    extern ne10_fft_cfg_float32_t ne10_fft_alloc_c2c_float32_c (ne10_int32_t nfft);
+
     extern void ne10_fft_c2c_1d_float32_c (ne10_fft_cpx_float32_t *fout,
                                            ne10_fft_cpx_float32_t *fin,
                                            ne10_fft_cfg_float32_t cfg,
@@ -148,6 +151,8 @@ extern "C" {
 
 
     /* NEON version*/
+    extern ne10_fft_cfg_float32_t ne10_fft_alloc_c2c_float32_neon (ne10_int32_t nfft);
+
     extern void ne10_fft_c2c_1d_float32_neon (ne10_fft_cpx_float32_t *fout,
             ne10_fft_cpx_float32_t *fin,
             ne10_fft_cfg_float32_t cfg,

diff --git a/inc/NE10_types.h b/inc/NE10_types.h
@@ -223,6 +223,7 @@ typedef struct
     ne10_int32_t *factors;
     ne10_fft_cpx_float32_t *twiddles;
     ne10_fft_cpx_float32_t *buffer;
+    ne10_fft_cpx_float32_t *last_twiddles;
 } ne10_fft_state_float32_t;
 
 typedef ne10_fft_state_float32_t* ne10_fft_cfg_float32_t;

diff --git a/modules/CMakeLists.txt b/modules/CMakeLists.txt
@@ -190,6 +190,7 @@ if(NE10_ENABLE_DSP)
         ${PROJECT_SOURCE_DIR}/modules/dsp/NE10_fft_float32.neon.c
         ${PROJECT_SOURCE_DIR}/modules/dsp/NE10_fft_int32.neon.c
         ${PROJECT_SOURCE_DIR}/modules/dsp/NE10_fft_int16.neon.c
+        ${PROJECT_SOURCE_DIR}/modules/dsp/NE10_fft_generic_float32.neonintrinsic.cpp
         #${PROJECT_SOURCE_DIR}/modules/dsp/NE10_rfft_float32.neonintrinsic.c
         #${PROJECT_SOURCE_DIR}/modules/dsp/NE10_fft_float32.neonintrinsic.c
         #${PROJECT_SOURCE_DIR}/modules/dsp/NE10_fft_int32.neonintrinsic.c