/* *  Copyright 2013 The LibYuv 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 "libyuv/scale.h" #include <assert.h>#include <string.h> #include "libyuv/cpu_id.h"#include "libyuv/planar_functions.h"  // For CopyARGB#include "libyuv/row.h"#include "libyuv/scale_row.h" #ifdef __cplusplusnamespace libyuv {extern "C" {#endif static __inline int Abs(int v) {  return v >= 0 ? v : -v;} // CPU agnostic row functionsvoid ScaleRowDown2_C(const uint8_t* src_ptr,                     ptrdiff_t src_stride,                     uint8_t* dst,                     int dst_width) {  int x;  (void)src_stride;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = src_ptr[1];    dst[1] = src_ptr[3];    dst += 2;    src_ptr += 4;  }  if (dst_width & 1) {    dst[0] = src_ptr[1];  }} void ScaleRowDown2_16_C(const uint16_t* src_ptr,                        ptrdiff_t src_stride,                        uint16_t* dst,                        int dst_width) {  int x;  (void)src_stride;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = src_ptr[1];    dst[1] = src_ptr[3];    dst += 2;    src_ptr += 4;  }  if (dst_width & 1) {    dst[0] = src_ptr[1];  }} void ScaleRowDown2Linear_C(const uint8_t* src_ptr,                           ptrdiff_t src_stride,                           uint8_t* dst,                           int dst_width) {  const uint8_t* s = src_ptr;  int x;  (void)src_stride;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = (s[0] + s[1] + 1) >> 1;    dst[1] = (s[2] + s[3] + 1) >> 1;    dst += 2;    s += 4;  }  if (dst_width & 1) {    dst[0] = (s[0] + s[1] + 1) >> 1;  }} void ScaleRowDown2Linear_16_C(const uint16_t* src_ptr,                              ptrdiff_t src_stride,                              uint16_t* dst,                              int dst_width) {  const uint16_t* s = src_ptr;  int x;  (void)src_stride;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = (s[0] + s[1] + 1) >> 1;    dst[1] = (s[2] + s[3] + 1) >> 1;    dst += 2;    s += 4;  }  if (dst_width & 1) {    dst[0] = (s[0] + s[1] + 1) >> 1;  }} void ScaleRowDown2Box_C(const uint8_t* src_ptr,                        ptrdiff_t src_stride,                        uint8_t* dst,                        int dst_width) {  const uint8_t* s = src_ptr;  const uint8_t* t = src_ptr + src_stride;  int x;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = (s[0] + s[1] + t[0] + t[1] + 2) >> 2;    dst[1] = (s[2] + s[3] + t[2] + t[3] + 2) >> 2;    dst += 2;    s += 4;    t += 4;  }  if (dst_width & 1) {    dst[0] = (s[0] + s[1] + t[0] + t[1] + 2) >> 2;  }} void ScaleRowDown2Box_Odd_C(const uint8_t* src_ptr,                            ptrdiff_t src_stride,                            uint8_t* dst,                            int dst_width) {  const uint8_t* s = src_ptr;  const uint8_t* t = src_ptr + src_stride;  int x;  dst_width -= 1;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = (s[0] + s[1] + t[0] + t[1] + 2) >> 2;    dst[1] = (s[2] + s[3] + t[2] + t[3] + 2) >> 2;    dst += 2;    s += 4;    t += 4;  }  if (dst_width & 1) {    dst[0] = (s[0] + s[1] + t[0] + t[1] + 2) >> 2;    dst += 1;    s += 2;    t += 2;  }  dst[0] = (s[0] + t[0] + 1) >> 1;} void ScaleRowDown2Box_16_C(const uint16_t* src_ptr,                           ptrdiff_t src_stride,                           uint16_t* dst,                           int dst_width) {  const uint16_t* s = src_ptr;  const uint16_t* t = src_ptr + src_stride;  int x;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = (s[0] + s[1] + t[0] + t[1] + 2) >> 2;    dst[1] = (s[2] + s[3] + t[2] + t[3] + 2) >> 2;    dst += 2;    s += 4;    t += 4;  }  if (dst_width & 1) {    dst[0] = (s[0] + s[1] + t[0] + t[1] + 2) >> 2;  }} void ScaleRowDown4_C(const uint8_t* src_ptr,                     ptrdiff_t src_stride,                     uint8_t* dst,                     int dst_width) {  int x;  (void)src_stride;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = src_ptr[2];    dst[1] = src_ptr[6];    dst += 2;    src_ptr += 8;  }  if (dst_width & 1) {    dst[0] = src_ptr[2];  }} void ScaleRowDown4_16_C(const uint16_t* src_ptr,                        ptrdiff_t src_stride,                        uint16_t* dst,                        int dst_width) {  int x;  (void)src_stride;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = src_ptr[2];    dst[1] = src_ptr[6];    dst += 2;    src_ptr += 8;  }  if (dst_width & 1) {    dst[0] = src_ptr[2];  }} void ScaleRowDown4Box_C(const uint8_t* src_ptr,                        ptrdiff_t src_stride,                        uint8_t* dst,                        int dst_width) {  intptr_t stride = src_stride;  int x;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = (src_ptr[0] + src_ptr[1] + src_ptr[2] + src_ptr[3] +              src_ptr[stride + 0] + src_ptr[stride + 1] + src_ptr[stride + 2] +              src_ptr[stride + 3] + src_ptr[stride * 2 + 0] +              src_ptr[stride * 2 + 1] + src_ptr[stride * 2 + 2] +              src_ptr[stride * 2 + 3] + src_ptr[stride * 3 + 0] +              src_ptr[stride * 3 + 1] + src_ptr[stride * 3 + 2] +              src_ptr[stride * 3 + 3] + 8) >>             4;    dst[1] = (src_ptr[4] + src_ptr[5] + src_ptr[6] + src_ptr[7] +              src_ptr[stride + 4] + src_ptr[stride + 5] + src_ptr[stride + 6] +              src_ptr[stride + 7] + src_ptr[stride * 2 + 4] +              src_ptr[stride * 2 + 5] + src_ptr[stride * 2 + 6] +              src_ptr[stride * 2 + 7] + src_ptr[stride * 3 + 4] +              src_ptr[stride * 3 + 5] + src_ptr[stride * 3 + 6] +              src_ptr[stride * 3 + 7] + 8) >>             4;    dst += 2;    src_ptr += 8;  }  if (dst_width & 1) {    dst[0] = (src_ptr[0] + src_ptr[1] + src_ptr[2] + src_ptr[3] +              src_ptr[stride + 0] + src_ptr[stride + 1] + src_ptr[stride + 2] +              src_ptr[stride + 3] + src_ptr[stride * 2 + 0] +              src_ptr[stride * 2 + 1] + src_ptr[stride * 2 + 2] +              src_ptr[stride * 2 + 3] + src_ptr[stride * 3 + 0] +              src_ptr[stride * 3 + 1] + src_ptr[stride * 3 + 2] +              src_ptr[stride * 3 + 3] + 8) >>             4;  }} void ScaleRowDown4Box_16_C(const uint16_t* src_ptr,                           ptrdiff_t src_stride,                           uint16_t* dst,                           int dst_width) {  intptr_t stride = src_stride;  int x;  for (x = 0; x < dst_width - 1; x += 2) {    dst[0] = (src_ptr[0] + src_ptr[1] + src_ptr[2] + src_ptr[3] +              src_ptr[stride + 0] + src_ptr[stride + 1] + src_ptr[stride + 2] +              src_ptr[stride + 3] + src_ptr[stride * 2 + 0] +              src_ptr[stride * 2 + 1] + src_ptr[stride * 2 + 2] +              src_ptr[stride * 2 + 3] + src_ptr[stride * 3 + 0] +              src_ptr[stride * 3 + 1] + src_ptr[stride * 3 + 2] +              src_ptr[stride * 3 + 3] + 8) >>             4;    dst[1] = (src_ptr[4] + src_ptr[5] + src_ptr[6] + src_ptr[7] +              src_ptr[stride + 4] + src_ptr[stride + 5] + src_ptr[stride + 6] +              src_ptr[stride + 7] + src_ptr[stride * 2 + 4] +              src_ptr[stride * 2 + 5] + src_ptr[stride * 2 + 6] +              src_ptr[stride * 2 + 7] + src_ptr[stride * 3 + 4] +              src_ptr[stride * 3 + 5] + src_ptr[stride * 3 + 6] +              src_ptr[stride * 3 + 7] + 8) >>             4;    dst += 2;    src_ptr += 8;  }  if (dst_width & 1) {    dst[0] = (src_ptr[0] + src_ptr[1] + src_ptr[2] + src_ptr[3] +              src_ptr[stride + 0] + src_ptr[stride + 1] + src_ptr[stride + 2] +              src_ptr[stride + 3] + src_ptr[stride * 2 + 0] +              src_ptr[stride * 2 + 1] + src_ptr[stride * 2 + 2] +              src_ptr[stride * 2 + 3] + src_ptr[stride * 3 + 0] +              src_ptr[stride * 3 + 1] + src_ptr[stride * 3 + 2] +              src_ptr[stride * 3 + 3] + 8) >>             4;  }} void ScaleRowDown34_C(const uint8_t* src_ptr,                      ptrdiff_t src_stride,                      uint8_t* dst,                      int dst_width) {  int x;  (void)src_stride;  assert((dst_width % 3 == 0) && (dst_width > 0));  for (x = 0; x < dst_width; x += 3) {    dst[0] = src_ptr[0];    dst[1] = src_ptr[1];    dst[2] = src_ptr[3];    dst += 3;    src_ptr += 4;  }} void ScaleRowDown34_16_C(const uint16_t* src_ptr,                         ptrdiff_t src_stride,                         uint16_t* dst,                         int dst_width) {  int x;  (void)src_stride;  assert((dst_width % 3 == 0) && (dst_width > 0));  for (x = 0; x < dst_width; x += 3) {    dst[0] = src_ptr[0];    dst[1] = src_ptr[1];    dst[2] = src_ptr[3];    dst += 3;    src_ptr += 4;  }} // Filter rows 0 and 1 together, 3 : 1void ScaleRowDown34_0_Box_C(const uint8_t* src_ptr,                            ptrdiff_t src_stride,                            uint8_t* d,                            int dst_width) {  const uint8_t* s = src_ptr;  const uint8_t* t = src_ptr + src_stride;  int x;  assert((dst_width % 3 == 0) && (dst_width > 0));  for (x = 0; x < dst_width; x += 3) {    uint8_t a0 = (s[0] * 3 + s[1] * 1 + 2) >> 2;    uint8_t a1 = (s[1] * 1 + s[2] * 1 + 1) >> 1;    uint8_t a2 = (s[2] * 1 + s[3] * 3 + 2) >> 2;    uint8_t b0 = (t[0] * 3 + t[1] * 1 + 2) >> 2;    uint8_t b1 = (t[1] * 1 + t[2] * 1 + 1) >> 1;    uint8_t b2 = (t[2] * 1 + t[3] * 3 + 2) >> 2;    d[0] = (a0 * 3 + b0 + 2) >> 2;    d[1] = (a1 * 3 + b1 + 2) >> 2;    d[2] = (a2 * 3 + b2 + 2) >> 2;    d += 3;    s += 4;    t += 4;  }} void ScaleRowDown34_0_Box_16_C(const uint16_t* src_ptr,                               ptrdiff_t src_stride,                               uint16_t* d,                               int dst_width) {  const uint16_t* s = src_ptr;  const uint16_t* t = src_ptr + src_stride;  int x;  assert((dst_width % 3 == 0) && (dst_width > 0));  for (x = 0; x < dst_width; x += 3) {    uint16_t a0 = (s[0] * 3 + s[1] * 1 + 2) >> 2;    uint16_t a1 = (s[1] * 1 + s[2] * 1 + 1) >> 1;    uint16_t a2 = (s[2] * 1 + s[3] * 3 + 2) >> 2;    uint16_t b0 = (t[0] * 3 + t[1] * 1 + 2) >> 2;    uint16_t b1 = (t[1] * 1 + t[2] * 1 + 1) >> 1;    uint16_t b2 = (t[2] * 1 + t[3] * 3 + 2) >> 2;    d[0] = (a0 * 3 + b0 + 2) >> 2;    d[1] = (a1 * 3 + b1 + 2) >> 2;    d[2] = (a2 * 3 + b2 + 2) >> 2;    d += 3;    s += 4;    t += 4;  }} // Filter rows 1 and 2 together, 1 : 1void ScaleRowDown34_1_Box_C(const uint8_t* src_ptr,                            ptrdiff_t src_stride,                            uint8_t* d,                            int dst_width) {  const uint8_t* s = src_ptr;  const uint8_t* t = src_ptr + src_stride;  int x;  assert((dst_width % 3 == 0) && (dst_width > 0));  for (x = 0; x < dst_width; x += 3) {    uint8_t a0 = (s[0] * 3 + s[1] * 1 + 2) >> 2;    uint8_t a1 = (s[1] * 1 + s[2] * 1 + 1) >> 1;    uint8_t a2 = (s[2] * 1 + s[3] * 3 + 2) >> 2;    uint8_t b0 = (t[0] * 3 + t[1] * 1 + 2) >> 2;    uint8_t b1 = (t[1] * 1 + t[2] * 1 + 1) >> 1;    uint8_t b2 = (t[2] * 1 + t[3] * 3 + 2) >> 2;    d[0] = (a0 + b0 + 1) >> 1;    d[1] = (a1 + b1 + 1) >> 1;    d[2] = (a2 + b2 + 1) >> 1;    d += 3;    s += 4;    t += 4;  }} void ScaleRowDown34_1_Box_16_C(const uint16_t* src_ptr,                               ptrdiff_t src_stride,                               uint16_t* d,                               int dst_width) {  const uint16_t* s = src_ptr;  const uint16_t* t = src_ptr + src_stride;  int x;  assert((dst_width % 3 == 0) && (dst_width > 0));  for (x = 0; x < dst_width; x += 3) {    uint16_t a0 = (s[0] * 3 + s[1] * 1 + 2) >> 2;    uint16_t a1 = (s[1] * 1 + s[2] * 1 + 1) >> 1;    uint16_t a2 = (s[2] * 1 + s[3] * 3 + 2) >> 2;    uint16_t b0 = (t[0] * 3 + t[1] * 1 + 2) >> 2;    uint16_t b1 = (t[1] * 1 + t[2] * 1 + 1) >> 1;    uint16_t b2 = (t[2] * 1 + t[3] * 3 + 2) >> 2;    d[0] = (a0 + b0 + 1) >> 1;    d[1] = (a1 + b1 + 1) >> 1;    d[2] = (a2 + b2 + 1) >> 1;    d += 3;    s += 4;    t += 4;  }} // Scales a single row of pixels using point sampling.void ScaleCols_C(uint8_t* dst_ptr,                 const uint8_t* src_ptr,                 int dst_width,                 int x,                 int dx) {  int j;  for (j = 0; j < dst_width - 1; j += 2) {    dst_ptr[0] = src_ptr[x >> 16];    x += dx;    dst_ptr[1] = src_ptr[x >> 16];    x += dx;    dst_ptr += 2;  }  if (dst_width & 1) {    dst_ptr[0] = src_ptr[x >> 16];  }} void ScaleCols_16_C(uint16_t* dst_ptr,                    const uint16_t* src_ptr,                    int dst_width,                    int x,                    int dx) {  int j;  for (j = 0; j < dst_width - 1; j += 2) {    dst_ptr[0] = src_ptr[x >> 16];    x += dx;    dst_ptr[1] = src_ptr[x >> 16];    x += dx;    dst_ptr += 2;  }  if (dst_width & 1) {    dst_ptr[0] = src_ptr[x >> 16];  }} // Scales a single row of pixels up by 2x using point sampling.void ScaleColsUp2_C(uint8_t* dst_ptr,                    const uint8_t* src_ptr,                    int dst_width,                    int x,                    int dx) {  int j;  (void)x;  (void)dx;  for (j = 0; j < dst_width - 1; j += 2) {    dst_ptr[1] = dst_ptr[0] = src_ptr[0];    src_ptr += 1;    dst_ptr += 2;  }  if (dst_width & 1) {    dst_ptr[0] = src_ptr[0];  }} void ScaleColsUp2_16_C(uint16_t* dst_ptr,                       const uint16_t* src_ptr,                       int dst_width,                       int x,                       int dx) {  int j;  (void)x;  (void)dx;  for (j = 0; j < dst_width - 1; j += 2) {    dst_ptr[1] = dst_ptr[0] = src_ptr[0];    src_ptr += 1;    dst_ptr += 2;  }  if (dst_width & 1) {    dst_ptr[0] = src_ptr[0];  }} // (1-f)a + fb can be replaced with a + f(b-a)#if defined(__arm__) || defined(__aarch64__)#define BLENDER(a, b, f) \  (uint8_t)((int)(a) + ((((int)((f)) * ((int)(b) - (int)(a))) + 0x8000) >> 16))#else// Intel uses 7 bit math with rounding.#define BLENDER(a, b, f) \  (uint8_t)((int)(a) + (((int)((f) >> 9) * ((int)(b) - (int)(a)) + 0x40) >> 7))#endif void ScaleFilterCols_C(uint8_t* dst_ptr,                       const uint8_t* src_ptr,                       int dst_width,                       int x,                       int dx) {  int j;  for (j = 0; j < dst_width - 1; j += 2) {    int xi = x >> 16;    int a = src_ptr[xi];    int b = src_ptr[xi + 1];    dst_ptr[0] = BLENDER(a, b, x & 0xffff);    x += dx;    xi = x >> 16;    a = src_ptr[xi];

Lines 1–500 of 1323. Load more