#include "avcodec.h"
#include "dsputil.h"
-#include "mpegvideo.h"
#include "simple_idct.h"
#include "faandct.h"
#include "faanidct.h"
0x32, 0x3A, 0x36, 0x3B, 0x33, 0x3E, 0x37, 0x3F,
};
+static const uint8_t idct_sse2_row_perm[8] = {0, 4, 1, 5, 2, 6, 3, 7};
+
+void ff_init_scantable(uint8_t *permutation, ScanTable *st, const uint8_t *src_scantable){
+ int i;
+ int end;
+
+ st->scantable= src_scantable;
+
+ for(i=0; i<64; i++){
+ int j;
+ j = src_scantable[i];
+ st->permutated[i] = permutation[j];
+#ifdef ARCH_POWERPC
+ st->inverse[j] = i;
+#endif
+ }
+
+ end=-1;
+ for(i=0; i<64; i++){
+ int j;
+ j = st->permutated[i];
+ if(j>end) end=j;
+ st->raster_end[i]= end;
+ }
+}
+
static int pix_sum_c(uint8_t * pix, int line_size)
{
int s, i, j;
}
#endif
+/* draw the edges of width 'w' of an image of size width, height */
+//FIXME check that this is ok for mpeg4 interlaced
+static void draw_edges_c(uint8_t *buf, int wrap, int width, int height, int w)
+{
+ uint8_t *ptr, *last_line;
+ int i;
+
+ last_line = buf + (height - 1) * wrap;
+ for(i=0;i<w;i++) {
+ /* top and bottom */
+ memcpy(buf - (i + 1) * wrap, buf, width);
+ memcpy(last_line + (i + 1) * wrap, last_line, width);
+ }
+ /* left and right */
+ ptr = buf;
+ for(i=0;i<height;i++) {
+ memset(ptr - w, ptr[0], w);
+ memset(ptr + width, ptr[width-1], w);
+ ptr += wrap;
+ }
+ /* corners */
+ for(i=0;i<w;i++) {
+ memset(buf - (i + 1) * wrap - w, buf[0], w); /* top left */
+ memset(buf - (i + 1) * wrap + width, buf[width-1], w); /* top right */
+ memset(last_line + (i + 1) * wrap - w, last_line[0], w); /* top left */
+ memset(last_line + (i + 1) * wrap + width, last_line[width-1], w); /* top right */
+ }
+}
+
+/**
+ * Copies a rectangular area of samples to a temporary buffer and replicates the boarder samples.
+ * @param buf destination buffer
+ * @param src source buffer
+ * @param linesize number of bytes between 2 vertically adjacent samples in both the source and destination buffers
+ * @param block_w width of block
+ * @param block_h height of block
+ * @param src_x x coordinate of the top left sample of the block in the source buffer
+ * @param src_y y coordinate of the top left sample of the block in the source buffer
+ * @param w width of the source buffer
+ * @param h height of the source buffer
+ */
+void ff_emulated_edge_mc(uint8_t *buf, uint8_t *src, int linesize, int block_w, int block_h,
+ int src_x, int src_y, int w, int h){
+ int x, y;
+ int start_y, start_x, end_y, end_x;
+
+ if(src_y>= h){
+ src+= (h-1-src_y)*linesize;
+ src_y=h-1;
+ }else if(src_y<=-block_h){
+ src+= (1-block_h-src_y)*linesize;
+ src_y=1-block_h;
+ }
+ if(src_x>= w){
+ src+= (w-1-src_x);
+ src_x=w-1;
+ }else if(src_x<=-block_w){
+ src+= (1-block_w-src_x);
+ src_x=1-block_w;
+ }
+
+ start_y= FFMAX(0, -src_y);
+ start_x= FFMAX(0, -src_x);
+ end_y= FFMIN(block_h, h-src_y);
+ end_x= FFMIN(block_w, w-src_x);
+
+ // copy existing part
+ for(y=start_y; y<end_y; y++){
+ for(x=start_x; x<end_x; x++){
+ buf[x + y*linesize]= src[x + y*linesize];
+ }
+ }
+
+ //top
+ for(y=0; y<start_y; y++){
+ for(x=start_x; x<end_x; x++){
+ buf[x + y*linesize]= buf[x + start_y*linesize];
+ }
+ }
+
+ //bottom
+ for(y=end_y; y<block_h; y++){
+ for(x=start_x; x<end_x; x++){
+ buf[x + y*linesize]= buf[x + (end_y-1)*linesize];
+ }
+ }
+
+ for(y=0; y<block_h; y++){
+ //left
+ for(x=0; x<start_x; x++){
+ buf[x + y*linesize]= buf[start_x + y*linesize];
+ }
+
+ //right
+ for(x=end_x; x<block_w; x++){
+ buf[x + y*linesize]= buf[end_x - 1 + y*linesize];
+ }
+ }
+}
+
static void get_pixels_c(DCTELEM *restrict block, const uint8_t *pixels, int line_size)
{
int i;
DECLARE_ALIGNED_8 (uint64_t, aligned_bak[stride]);
DCTELEM * const temp= (DCTELEM*)aligned_temp;
uint8_t * const bak= (uint8_t*)aligned_bak;
- int i, last, run, bits, level, distoration, start_i;
+ int i, last, run, bits, level, distortion, start_i;
const int esc_length= s->ac_esc_length;
uint8_t * length;
uint8_t * last_length;
s->dsp.idct_add(bak, stride, temp);
- distoration= s->dsp.sse[1](NULL, bak, src1, stride, 8);
+ distortion= s->dsp.sse[1](NULL, bak, src1, stride, 8);
- return distoration + ((bits*s->qscale*s->qscale*109 + 64)>>7);
+ return distortion + ((bits*s->qscale*s->qscale*109 + 64)>>7);
}
static int bit8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, int stride, int h){
dst[i*step] = src0[i] * src1[i] + src2[i] + src3;
}
-void ff_float_to_int16_c(int16_t *dst, const float *src, int len){
+void ff_vector_fmul_window_c(float *dst, const float *src0, const float *src1, const float *win, float add_bias, int len){
+ int i,j;
+ dst += len;
+ win += len;
+ src0+= len;
+ for(i=-len, j=len-1; i<0; i++, j--) {
+ float s0 = src0[i];
+ float s1 = src1[j];
+ float wi = win[i];
+ float wj = win[j];
+ dst[i] = s0*wj - s1*wi + add_bias;
+ dst[j] = s0*wi + s1*wj + add_bias;
+ }
+}
+
+static av_always_inline int float_to_int16_one(const float *src){
+ int_fast32_t tmp = *(const int32_t*)src;
+ if(tmp & 0xf0000){
+ tmp = (0x43c0ffff - tmp)>>31;
+ // is this faster on some gcc/cpu combinations?
+// if(tmp > 0x43c0ffff) tmp = 0xFFFF;
+// else tmp = 0;
+ }
+ return tmp - 0x8000;
+}
+
+void ff_float_to_int16_c(int16_t *dst, const float *src, long len){
int i;
- for(i=0; i<len; i++) {
- int_fast32_t tmp = ((const int32_t*)src)[i];
- if(tmp & 0xf0000){
- tmp = (0x43c0ffff - tmp)>>31;
- // is this faster on some gcc/cpu combinations?
-// if(tmp > 0x43c0ffff) tmp = 0xFFFF;
-// else tmp = 0;
+ for(i=0; i<len; i++)
+ dst[i] = float_to_int16_one(src+i);
+}
+
+void ff_float_to_int16_interleave_c(int16_t *dst, const float **src, long len, int channels){
+ int i,j,c;
+ if(channels==2){
+ for(i=0; i<len; i++){
+ dst[2*i] = float_to_int16_one(src[0]+i);
+ dst[2*i+1] = float_to_int16_one(src[1]+i);
}
- dst[i] = tmp - 0x8000;
+ }else{
+ for(c=0; c<channels; c++)
+ for(i=0, j=c; i<len; i++, j+=channels)
+ dst[j] = float_to_int16_one(src[c]+i);
}
}
+static void add_int16_c(int16_t * v1, int16_t * v2, int order)
+{
+ while (order--)
+ *v1++ += *v2++;
+}
+
+static void sub_int16_c(int16_t * v1, int16_t * v2, int order)
+{
+ while (order--)
+ *v1++ -= *v2++;
+}
+
+static int32_t scalarproduct_int16_c(int16_t * v1, int16_t * v2, int order, int shift)
+{
+ int res = 0;
+
+ while (order--)
+ res += (*v1++ * *v2++) >> shift;
+
+ return res;
+}
+
#define W0 2048
#define W1 2841 /* 2048*sqrt (2)*cos (1*pi/16) */
#define W2 2676 /* 2048*sqrt (2)*cos (2*pi/16) */
c->biweight_h264_pixels_tab[8]= biweight_h264_pixels2x4_c;
c->biweight_h264_pixels_tab[9]= biweight_h264_pixels2x2_c;
+ c->draw_edges = draw_edges_c;
+
#ifdef CONFIG_CAVS_DECODER
ff_cavsdsp_init(c,avctx);
#endif
c->vector_fmul = vector_fmul_c;
c->vector_fmul_reverse = vector_fmul_reverse_c;
c->vector_fmul_add_add = ff_vector_fmul_add_add_c;
+ c->vector_fmul_window = ff_vector_fmul_window_c;
c->float_to_int16 = ff_float_to_int16_c;
+ c->float_to_int16_interleave = ff_float_to_int16_interleave_c;
+ c->add_int16 = add_int16_c;
+ c->sub_int16 = sub_int16_c;
+ c->scalarproduct_int16 = scalarproduct_int16_c;
c->shrink[0]= ff_img_copy_plane;
c->shrink[1]= ff_shrink22;
for(i=0; i<64; i++)
c->idct_permutation[i]= (i&0x24) | ((i&3)<<3) | ((i>>3)&3);
break;
+ case FF_SSE2_IDCT_PERM:
+ for(i=0; i<64; i++)
+ c->idct_permutation[i]= (i&0x38) | idct_sse2_row_perm[i&7];
+ break;
default:
av_log(avctx, AV_LOG_ERROR, "Internal error, IDCT permutation not set\n");
}
projects / ffmpeg / blobdiff