/* * Copyright (c) 2010 The WebM project authors. All Rights Reserved. * * Use of this source code is governed by a BSD-style license * that can be found in the LICENSE file in the root of the source * tree. An additional intellectual property rights grant can be found * in the file PATENTS. All contributing project authors may * be found in the AUTHORS file in the root of the source tree. */ #include <assert.h> #include <emmintrin.h> // SSE2 #include "./vpx_config.h" #include "./vpx_dsp_rtcd.h" #include "vpx_ports/mem.h" #include "vpx_dsp/x86/mem_sse2.h" static INLINE unsigned int add32x4_sse2(__m128i val) { … } unsigned int vpx_get_mb_ss_sse2(const int16_t *src_ptr) { … } static INLINE __m128i load4x2_sse2(const uint8_t *const p, const int stride) { … } static INLINE void variance_kernel_sse2(const __m128i src_ptr, const __m128i ref_ptr, __m128i *const sse, __m128i *const sum) { … } // Can handle 128 pixels' diff sum (such as 8x16 or 16x8) // Slightly faster than variance_final_256_pel_sse2() static INLINE void variance_final_128_pel_sse2(__m128i vsse, __m128i vsum, unsigned int *const sse, int *const sum) { … } // Can handle 256 pixels' diff sum (such as 16x16) static INLINE void variance_final_256_pel_sse2(__m128i vsse, __m128i vsum, unsigned int *const sse, int *const sum) { … } // Can handle 512 pixels' diff sum (such as 16x32 or 32x16) static INLINE void variance_final_512_pel_sse2(__m128i vsse, __m128i vsum, unsigned int *const sse, int *const sum) { … } static INLINE __m128i sum_to_32bit_sse2(const __m128i sum) { … } // Can handle 1024 pixels' diff sum (such as 32x32) static INLINE int sum_final_sse2(const __m128i sum) { … } static INLINE void variance4_sse2(const uint8_t *src_ptr, const int src_stride, const uint8_t *ref_ptr, const int ref_stride, const int h, __m128i *const sse, __m128i *const sum) { … } static INLINE void variance8_sse2(const uint8_t *src_ptr, const int src_stride, const uint8_t *ref_ptr, const int ref_stride, const int h, __m128i *const sse, __m128i *const sum) { … } static INLINE void variance16_kernel_sse2(const uint8_t *const src_ptr, const uint8_t *const ref_ptr, __m128i *const sse, __m128i *const sum) { … } static INLINE void variance16_sse2(const uint8_t *src_ptr, const int src_stride, const uint8_t *ref_ptr, const int ref_stride, const int h, __m128i *const sse, __m128i *const sum) { … } static INLINE void variance32_sse2(const uint8_t *src_ptr, const int src_stride, const uint8_t *ref_ptr, const int ref_stride, const int h, __m128i *const sse, __m128i *const sum) { … } static INLINE void variance64_sse2(const uint8_t *src_ptr, const int src_stride, const uint8_t *ref_ptr, const int ref_stride, const int h, __m128i *const sse, __m128i *const sum) { … } void vpx_get8x8var_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum) { … } void vpx_get16x16var_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum) { … } unsigned int vpx_variance4x4_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance4x8_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance8x4_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance8x8_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance8x16_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance16x8_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance16x16_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance16x32_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance32x16_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance32x32_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance32x64_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance64x32_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_variance64x64_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_mse8x8_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_mse8x16_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_mse16x8_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } unsigned int vpx_mse16x16_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse) { … } // The 2 unused parameters are place holders for PIC enabled build. // These definitions are for functions defined in subpel_variance.asm #define DECL … #define DECLS … DECLS(sse2, sse2); DECLS(ssse3, ssse3); #undef DECLS #undef DECL #define FN … #define FNS … FNS FNS #undef FNS #undef FN // The 2 unused parameters are place holders for PIC enabled build. #define DECL … #define DECLS … DECLS(sse2, sse2); DECLS(ssse3, ssse3); #undef DECL #undef DECLS #define FN … #define FNS … FNS FNS #undef FNS #undef FN
Codebase Browser
chromium