97e5fac389
new file: lossless_neon.c speedup is ~5% gcc 4.6.3 seems to be doing some sub-optimal things here, storing register on stack using 'vstmia' and such. Looks similar to gcc.gnu.org/bugzilla/show_bug.cgi?id=51509 I've tried adding -fno-split-wide-types and it does help the generated assembly. But the overall speed gets worse with this flag. We should only compile lossless_neon.c with it -> urk. Change-Id: I2ccc0929f5ef9dfb0105960e65c0b79b5f18d3b0
107 lines
2.6 KiB
Makefile
107 lines
2.6 KiB
Makefile
LOCAL_PATH := $(call my-dir)
|
|
|
|
WEBP_CFLAGS := -Wall -DANDROID -DHAVE_MALLOC_H -DHAVE_PTHREAD -DWEBP_USE_THREAD
|
|
|
|
ifeq ($(APP_OPTIM),release)
|
|
WEBP_CFLAGS += -finline-functions -ffast-math \
|
|
-ffunction-sections -fdata-sections
|
|
ifeq ($(findstring clang,$(NDK_TOOLCHAIN_VERSION)),)
|
|
WEBP_CFLAGS += -frename-registers -s
|
|
endif
|
|
endif
|
|
|
|
include $(CLEAR_VARS)
|
|
|
|
LOCAL_SRC_FILES := \
|
|
src/dec/alpha.c \
|
|
src/dec/buffer.c \
|
|
src/dec/frame.c \
|
|
src/dec/idec.c \
|
|
src/dec/io.c \
|
|
src/dec/layer.c \
|
|
src/dec/quant.c \
|
|
src/dec/tree.c \
|
|
src/dec/vp8.c \
|
|
src/dec/vp8l.c \
|
|
src/dec/webp.c \
|
|
src/dsp/cpu.c \
|
|
src/dsp/dec.c \
|
|
src/dsp/dec_clip_tables.c \
|
|
src/dsp/dec_mips32.c \
|
|
src/dsp/dec_sse2.c \
|
|
src/dsp/enc.c \
|
|
src/dsp/enc_sse2.c \
|
|
src/dsp/lossless.c \
|
|
src/dsp/lossless_sse2.c \
|
|
src/dsp/upsampling.c \
|
|
src/dsp/upsampling_mips32.c \
|
|
src/dsp/upsampling_sse2.c \
|
|
src/dsp/yuv.c \
|
|
src/enc/alpha.c \
|
|
src/enc/analysis.c \
|
|
src/enc/backward_references.c \
|
|
src/enc/config.c \
|
|
src/enc/cost.c \
|
|
src/enc/filter.c \
|
|
src/enc/frame.c \
|
|
src/enc/histogram.c \
|
|
src/enc/iterator.c \
|
|
src/enc/layer.c \
|
|
src/enc/picture.c \
|
|
src/enc/quant.c \
|
|
src/enc/syntax.c \
|
|
src/enc/token.c \
|
|
src/enc/tree.c \
|
|
src/enc/vp8l.c \
|
|
src/enc/webpenc.c \
|
|
src/utils/alpha_processing.c \
|
|
src/utils/bit_reader.c \
|
|
src/utils/bit_writer.c \
|
|
src/utils/color_cache.c \
|
|
src/utils/filters.c \
|
|
src/utils/huffman.c \
|
|
src/utils/huffman_encode.c \
|
|
src/utils/quant_levels.c \
|
|
src/utils/quant_levels_dec.c \
|
|
src/utils/random.c \
|
|
src/utils/rescaler.c \
|
|
src/utils/thread.c \
|
|
src/utils/utils.c \
|
|
|
|
LOCAL_CFLAGS := $(WEBP_CFLAGS)
|
|
LOCAL_C_INCLUDES += $(LOCAL_PATH)/src
|
|
|
|
# prefer arm over thumb mode for performance gains
|
|
LOCAL_ARM_MODE := arm
|
|
|
|
ifneq ($(findstring armeabi-v7a, $(TARGET_ARCH_ABI)),)
|
|
# Setting LOCAL_ARM_NEON will enable -mfpu=neon which may cause illegal
|
|
# instructions to be generated for armv7a code. Instead target the neon code
|
|
# specifically.
|
|
LOCAL_SRC_FILES += src/dsp/dec_neon.c.neon
|
|
LOCAL_SRC_FILES += src/dsp/enc_neon.c.neon
|
|
LOCAL_SRC_FILES += src/dsp/lossless_neon.c.neon
|
|
LOCAL_SRC_FILES += src/dsp/upsampling_neon.c.neon
|
|
endif
|
|
LOCAL_STATIC_LIBRARIES := cpufeatures
|
|
|
|
LOCAL_MODULE := webp
|
|
|
|
include $(BUILD_STATIC_LIBRARY)
|
|
|
|
include $(CLEAR_VARS)
|
|
|
|
LOCAL_SRC_FILES := \
|
|
examples/dwebp.c \
|
|
examples/example_util.c \
|
|
|
|
LOCAL_CFLAGS := $(WEBP_CFLAGS)
|
|
LOCAL_C_INCLUDES := $(LOCAL_PATH)/src
|
|
LOCAL_STATIC_LIBRARIES := webp
|
|
|
|
LOCAL_MODULE := dwebp
|
|
|
|
include $(BUILD_EXECUTABLE)
|
|
|
|
$(call import-module,android/cpufeatures)
|