hevc_deblock: Fix compilation with nasm
[ffmpeg.git] / libavcodec / svq1enc.c
index d1a55bd..f49f487 100644 (file)
  *   http://www.pcisys.net/~melanson/codecs/
  */
 
-
 #include "avcodec.h"
-#include "dsputil.h"
+#include "hpeldsp.h"
+#include "me_cmp.h"
 #include "mpegvideo.h"
 #include "h263.h"
 #include "internal.h"
-
+#include "mpegutils.h"
 #include "svq1.h"
+#include "svq1enc.h"
 #include "svq1enc_cb.h"
 
 #undef NDEBUG
 #include <assert.h>
 
-
-typedef struct SVQ1Context {
-    MpegEncContext m; // needed for motion estimation, should not be used for anything else, the idea is to make the motion estimation eventually independent of MpegEncContext, so this will be removed then (FIXME/XXX)
-    AVCodecContext *avctx;
-    DSPContext dsp;
-    AVFrame picture;
-    AVFrame current_picture;
-    AVFrame last_picture;
-    PutBitContext pb;
-    GetBitContext gb;
-
-    PutBitContext reorder_pb[6]; //why ooh why this sick breadth first order, everything is slower and more complex
-
-    int frame_width;
-    int frame_height;
-
-    /* Y plane block dimensions */
-    int y_block_width;
-    int y_block_height;
-
-    /* U & V plane (C planes) block dimensions */
-    int c_block_width;
-    int c_block_height;
-
-    uint16_t *mb_type;
-    uint32_t *dummy;
-    int16_t (*motion_val8[3])[2];
-    int16_t (*motion_val16[3])[2];
-
-    int64_t rd_total;
-
-    uint8_t *scratchbuf;
-} SVQ1Context;
-
-static void svq1_write_header(SVQ1Context *s, int frame_type)
+static void svq1_write_header(SVQ1EncContext *s, int frame_type)
 {
     int i;
 
@@ -87,21 +54,19 @@ static void svq1_write_header(SVQ1Context *s, int frame_type)
     put_bits(&s->pb, 2, frame_type - 1);
 
     if (frame_type == AV_PICTURE_TYPE_I) {
-
         /* no checksum since frame code is 0x20 */
-
         /* no embedded string either */
-
         /* output 5 unknown bits (2 + 2 + 1) */
         put_bits(&s->pb, 5, 2); /* 2 needed by quicktime decoder */
 
-        i= ff_match_2uint16(ff_svq1_frame_size_table, FF_ARRAY_ELEMS(ff_svq1_frame_size_table), s->frame_width, s->frame_height);
+        i = ff_match_2uint16(ff_svq1_frame_size_table,
+                             FF_ARRAY_ELEMS(ff_svq1_frame_size_table),
+                             s->frame_width, s->frame_height);
         put_bits(&s->pb, 3, i);
 
-        if (i == 7)
-        {
-                put_bits(&s->pb, 12, s->frame_width);
-                put_bits(&s->pb, 12, s->frame_height);
+        if (i == 7) {
+            put_bits(&s->pb, 12, s->frame_width);
+            put_bits(&s->pb, 12, s->frame_height);
         }
     }
 
@@ -109,391 +74,488 @@ static void svq1_write_header(SVQ1Context *s, int frame_type)
     put_bits(&s->pb, 2, 0);
 }
 
-
-#define QUALITY_THRESHOLD 100
+#define QUALITY_THRESHOLD    100
 #define THRESHOLD_MULTIPLIER 0.6
 
-static int encode_block(SVQ1Context *s, uint8_t *src, uint8_t *ref, uint8_t *decoded, int stride, int level, int threshold, int lambda, int intra){
+static int ssd_int8_vs_int16_c(const int8_t *pix1, const int16_t *pix2,
+                               int size)
+{
+    int score = 0, i;
+
+    for (i = 0; i < size; i++)
+        score += (pix1[i] - pix2[i]) * (pix1[i] - pix2[i]);
+    return score;
+}
+
+static int encode_block(SVQ1EncContext *s, uint8_t *src, uint8_t *ref,
+                        uint8_t *decoded, int stride, int level,
+                        int threshold, int lambda, int intra)
+{
     int count, y, x, i, j, split, best_mean, best_score, best_count;
     int best_vector[6];
-    int block_sum[7]= {0, 0, 0, 0, 0, 0};
-    int w= 2<<((level+2)>>1);
-    int h= 2<<((level+1)>>1);
-    int size=w*h;
+    int block_sum[7] = { 0, 0, 0, 0, 0, 0 };
+    int w            = 2 << (level + 2 >> 1);
+    int h            = 2 << (level + 1 >> 1);
+    int size         = w * h;
     int16_t block[7][256];
     const int8_t *codebook_sum, *codebook;
-    const uint16_t (*mean_vlc)[2];
-    const uint8_t (*multistage_vlc)[2];
-
-    best_score=0;
-    //FIXME optimize, this doenst need to be done multiple times
-    if(intra){
-        codebook_sum= svq1_intra_codebook_sum[level];
-        codebook= ff_svq1_intra_codebooks[level];
-        mean_vlc= ff_svq1_intra_mean_vlc;
-        multistage_vlc= ff_svq1_intra_multistage_vlc[level];
-        for(y=0; y<h; y++){
-            for(x=0; x<w; x++){
-                int v= src[x + y*stride];
-                block[0][x + w*y]= v;
-                best_score += v*v;
-                block_sum[0] += v;
+    const uint16_t(*mean_vlc)[2];
+    const uint8_t(*multistage_vlc)[2];
+
+    best_score = 0;
+    // FIXME: Optimize, this does not need to be done multiple times.
+    if (intra) {
+        codebook_sum   = svq1_intra_codebook_sum[level];
+        codebook       = ff_svq1_intra_codebooks[level];
+        mean_vlc       = ff_svq1_intra_mean_vlc;
+        multistage_vlc = ff_svq1_intra_multistage_vlc[level];
+        for (y = 0; y < h; y++) {
+            for (x = 0; x < w; x++) {
+                int v = src[x + y * stride];
+                block[0][x + w * y] = v;
+                best_score         += v * v;
+                block_sum[0]       += v;
             }
         }
-    }else{
-        codebook_sum= svq1_inter_codebook_sum[level];
-        codebook= ff_svq1_inter_codebooks[level];
-        mean_vlc= ff_svq1_inter_mean_vlc + 256;
-        multistage_vlc= ff_svq1_inter_multistage_vlc[level];
-        for(y=0; y<h; y++){
-            for(x=0; x<w; x++){
-                int v= src[x + y*stride] - ref[x + y*stride];
-                block[0][x + w*y]= v;
-                best_score += v*v;
-                block_sum[0] += v;
+    } else {
+        codebook_sum   = svq1_inter_codebook_sum[level];
+        codebook       = ff_svq1_inter_codebooks[level];
+        mean_vlc       = ff_svq1_inter_mean_vlc + 256;
+        multistage_vlc = ff_svq1_inter_multistage_vlc[level];
+        for (y = 0; y < h; y++) {
+            for (x = 0; x < w; x++) {
+                int v = src[x + y * stride] - ref[x + y * stride];
+                block[0][x + w * y] = v;
+                best_score         += v * v;
+                block_sum[0]       += v;
             }
         }
     }
 
-    best_count=0;
-    best_score -= (int)(((unsigned)block_sum[0]*block_sum[0])>>(level+3));
-    best_mean= (block_sum[0] + (size>>1)) >> (level+3);
+    best_count  = 0;
+    best_score -= (int)((unsigned)block_sum[0] * block_sum[0] >> (level + 3));
+    best_mean   = block_sum[0] + (size >> 1) >> (level + 3);
 
-    if(level<4){
-        for(count=1; count<7; count++){
-            int best_vector_score= INT_MAX;
-            int best_vector_sum=-999, best_vector_mean=-999;
-            const int stage= count-1;
+    if (level < 4) {
+        for (count = 1; count < 7; count++) {
+            int best_vector_score = INT_MAX;
+            int best_vector_sum   = -999, best_vector_mean = -999;
+            const int stage       = count - 1;
             const int8_t *vector;
 
-            for(i=0; i<16; i++){
-                int sum= codebook_sum[stage*16 + i];
+            for (i = 0; i < 16; i++) {
+                int sum = codebook_sum[stage * 16 + i];
                 int sqr, diff, score;
 
-                vector = codebook + stage*size*16 + i*size;
-                sqr = s->dsp.ssd_int8_vs_int16(vector, block[stage], size);
-                diff= block_sum[stage] - sum;
-                score= sqr - ((diff*(int64_t)diff)>>(level+3)); //FIXME 64bit slooow
-                if(score < best_vector_score){
-                    int mean= (diff + (size>>1)) >> (level+3);
-                    assert(mean >-300 && mean<300);
-                    mean= av_clip(mean, intra?0:-256, 255);
-                    best_vector_score= score;
-                    best_vector[stage]= i;
-                    best_vector_sum= sum;
-                    best_vector_mean= mean;
+                vector = codebook + stage * size * 16 + i * size;
+                sqr    = s->ssd_int8_vs_int16(vector, block[stage], size);
+                diff   = block_sum[stage] - sum;
+                score  = sqr - (diff * (int64_t)diff >> (level + 3)); // FIXME: 64bit slooow
+                if (score < best_vector_score) {
+                    int mean = diff + (size >> 1) >> (level + 3);
+                    assert(mean > -300 && mean < 300);
+                    mean               = av_clip(mean, intra ? 0 : -256, 255);
+                    best_vector_score  = score;
+                    best_vector[stage] = i;
+                    best_vector_sum    = sum;
+                    best_vector_mean   = mean;
                 }
             }
             assert(best_vector_mean != -999);
-            vector= codebook + stage*size*16 + best_vector[stage]*size;
-            for(j=0; j<size; j++){
-                block[stage+1][j] = block[stage][j] - vector[j];
-            }
-            block_sum[stage+1]= block_sum[stage] - best_vector_sum;
-            best_vector_score +=
-                lambda*(+ 1 + 4*count
-                        + multistage_vlc[1+count][1]
-                        + mean_vlc[best_vector_mean][1]);
-
-            if(best_vector_score < best_score){
-                best_score= best_vector_score;
-                best_count= count;
-                best_mean= best_vector_mean;
+            vector = codebook + stage * size * 16 + best_vector[stage] * size;
+            for (j = 0; j < size; j++)
+                block[stage + 1][j] = block[stage][j] - vector[j];
+            block_sum[stage + 1] = block_sum[stage] - best_vector_sum;
+            best_vector_score   += lambda *
+                                   (+1 + 4 * count +
+                                    multistage_vlc[1 + count][1]
+                                    + mean_vlc[best_vector_mean][1]);
+
+            if (best_vector_score < best_score) {
+                best_score = best_vector_score;
+                best_count = count;
+                best_mean  = best_vector_mean;
             }
         }
     }
 
-    split=0;
-    if(best_score > threshold && level){
-        int score=0;
-        int offset= (level&1) ? stride*h/2 : w/2;
+    split = 0;
+    if (best_score > threshold && level) {
+        int score  = 0;
+        int offset = level & 1 ? stride * h / 2 : w / 2;
         PutBitContext backup[6];
 
-        for(i=level-1; i>=0; i--){
-            backup[i]= s->reorder_pb[i];
-        }
-        score += encode_block(s, src         , ref         , decoded         , stride, level-1, threshold>>1, lambda, intra);
-        score += encode_block(s, src + offset, ref + offset, decoded + offset, stride, level-1, threshold>>1, lambda, intra);
+        for (i = level - 1; i >= 0; i--)
+            backup[i] = s->reorder_pb[i];
+        score += encode_block(s, src, ref, decoded, stride, level - 1,
+                              threshold >> 1, lambda, intra);
+        score += encode_block(s, src + offset, ref + offset, decoded + offset,
+                              stride, level - 1, threshold >> 1, lambda, intra);
         score += lambda;
 
-        if(score < best_score){
-            best_score= score;
-            split=1;
-        }else{
-            for(i=level-1; i>=0; i--){
-                s->reorder_pb[i]= backup[i];
-            }
+        if (score < best_score) {
+            best_score = score;
+            split      = 1;
+        } else {
+            for (i = level - 1; i >= 0; i--)
+                s->reorder_pb[i] = backup[i];
         }
     }
     if (level > 0)
         put_bits(&s->reorder_pb[level], 1, split);
 
-    if(!split){
-        assert((best_mean >= 0 && best_mean<256) || !intra);
-        assert(best_mean >= -256 && best_mean<256);
-        assert(best_count >=0 && best_count<7);
-        assert(level<4 || best_count==0);
+    if (!split) {
+        assert(best_mean >= 0 && best_mean < 256 || !intra);
+        assert(best_mean >= -256 && best_mean < 256);
+        assert(best_count >= 0 && best_count < 7);
+        assert(level < 4 || best_count == 0);
 
         /* output the encoding */
         put_bits(&s->reorder_pb[level],
-            multistage_vlc[1 + best_count][1],
-            multistage_vlc[1 + best_count][0]);
+                 multistage_vlc[1 + best_count][1],
+                 multistage_vlc[1 + best_count][0]);
         put_bits(&s->reorder_pb[level], mean_vlc[best_mean][1],
-            mean_vlc[best_mean][0]);
+                 mean_vlc[best_mean][0]);
 
-        for (i = 0; i < best_count; i++){
-            assert(best_vector[i]>=0 && best_vector[i]<16);
+        for (i = 0; i < best_count; i++) {
+            assert(best_vector[i] >= 0 && best_vector[i] < 16);
             put_bits(&s->reorder_pb[level], 4, best_vector[i]);
         }
 
-        for(y=0; y<h; y++){
-            for(x=0; x<w; x++){
-                decoded[x + y*stride]= src[x + y*stride] - block[best_count][x + w*y] + best_mean;
-            }
-        }
+        for (y = 0; y < h; y++)
+            for (x = 0; x < w; x++)
+                decoded[x + y * stride] = src[x + y * stride] -
+                                          block[best_count][x + w * y] +
+                                          best_mean;
     }
 
     return best_score;
 }
 
-
-static int svq1_encode_plane(SVQ1Context *s, int plane, unsigned char *src_plane, unsigned char *ref_plane, unsigned char *decoded_plane,
-    int width, int height, int src_stride, int stride)
+static int svq1_encode_plane(SVQ1EncContext *s, int plane,
+                             unsigned char *src_plane,
+                             unsigned char *ref_plane,
+                             unsigned char *decoded_plane,
+                             int width, int height, int src_stride, int stride)
 {
+    const AVFrame *f = s->avctx->coded_frame;
     int x, y;
     int i;
     int block_width, block_height;
     int level;
     int threshold[6];
-    uint8_t *src = s->scratchbuf + stride * 16;
-    const int lambda= (s->picture.quality*s->picture.quality) >> (2*FF_LAMBDA_SHIFT);
+    uint8_t *src     = s->scratchbuf + stride * 16;
+    const int lambda = (f->quality * f->quality) >>
+                       (2 * FF_LAMBDA_SHIFT);
 
     /* figure out the acceptable level thresholds in advance */
     threshold[5] = QUALITY_THRESHOLD;
     for (level = 4; level >= 0; level--)
         threshold[level] = threshold[level + 1] * THRESHOLD_MULTIPLIER;
 
-    block_width = (width + 15) / 16;
+    block_width  = (width  + 15) / 16;
     block_height = (height + 15) / 16;
 
-    if(s->picture.pict_type == AV_PICTURE_TYPE_P){
-        s->m.avctx= s->avctx;
-        s->m.current_picture_ptr= &s->m.current_picture;
-        s->m.last_picture_ptr   = &s->m.last_picture;
-        s->m.last_picture.f.data[0] = ref_plane;
-        s->m.linesize=
-        s->m.last_picture.f.linesize[0] =
-        s->m.new_picture.f.linesize[0] =
-        s->m.current_picture.f.linesize[0] = stride;
-        s->m.width= width;
-        s->m.height= height;
-        s->m.mb_width= block_width;
-        s->m.mb_height= block_height;
-        s->m.mb_stride= s->m.mb_width+1;
-        s->m.b8_stride= 2*s->m.mb_width+1;
-        s->m.f_code=1;
-        s->m.pict_type= s->picture.pict_type;
-        s->m.me_method= s->avctx->me_method;
-        s->m.me.scene_change_score=0;
-        s->m.flags= s->avctx->flags;
-//        s->m.out_format = FMT_H263;
-//        s->m.unrestricted_mv= 1;
-
-        s->m.lambda= s->picture.quality;
-        s->m.qscale= (s->m.lambda*139 + FF_LAMBDA_SCALE*64) >> (FF_LAMBDA_SHIFT + 7);
-        s->m.lambda2= (s->m.lambda*s->m.lambda + FF_LAMBDA_SCALE/2) >> FF_LAMBDA_SHIFT;
-
-        if(!s->motion_val8[plane]){
-            s->motion_val8 [plane]= av_mallocz((s->m.b8_stride*block_height*2 + 2)*2*sizeof(int16_t));
-            s->motion_val16[plane]= av_mallocz((s->m.mb_stride*(block_height + 2) + 1)*2*sizeof(int16_t));
+    if (f->pict_type == AV_PICTURE_TYPE_P) {
+        s->m.avctx                         = s->avctx;
+        s->m.current_picture_ptr           = &s->m.current_picture;
+        s->m.last_picture_ptr              = &s->m.last_picture;
+        s->m.last_picture.f->data[0]        = ref_plane;
+        s->m.linesize                      =
+        s->m.last_picture.f->linesize[0]    =
+        s->m.new_picture.f->linesize[0]     =
+        s->m.current_picture.f->linesize[0] = stride;
+        s->m.width                         = width;
+        s->m.height                        = height;
+        s->m.mb_width                      = block_width;
+        s->m.mb_height                     = block_height;
+        s->m.mb_stride                     = s->m.mb_width + 1;
+        s->m.b8_stride                     = 2 * s->m.mb_width + 1;
+        s->m.f_code                        = 1;
+        s->m.pict_type                     = f->pict_type;
+        s->m.me_method                     = s->avctx->me_method;
+        s->m.me.scene_change_score         = 0;
+        s->m.flags                         = s->avctx->flags;
+        // s->m.out_format                    = FMT_H263;
+        // s->m.unrestricted_mv               = 1;
+        s->m.lambda                        = f->quality;
+        s->m.qscale                        = s->m.lambda * 139 +
+                                             FF_LAMBDA_SCALE * 64 >>
+                                             FF_LAMBDA_SHIFT + 7;
+        s->m.lambda2                       = s->m.lambda * s->m.lambda +
+                                             FF_LAMBDA_SCALE / 2 >>
+                                             FF_LAMBDA_SHIFT;
+
+        if (!s->motion_val8[plane]) {
+            s->motion_val8[plane]  = av_mallocz((s->m.b8_stride *
+                                                 block_height * 2 + 2) *
+                                                2 * sizeof(int16_t));
+            s->motion_val16[plane] = av_mallocz((s->m.mb_stride *
+                                                 (block_height + 2) + 1) *
+                                                2 * sizeof(int16_t));
         }
 
-        s->m.mb_type= s->mb_type;
+        s->m.mb_type = s->mb_type;
 
-        //dummies, to avoid segfaults
-        s->m.current_picture.mb_mean=   (uint8_t *)s->dummy;
-        s->m.current_picture.mb_var=    (uint16_t*)s->dummy;
-        s->m.current_picture.mc_mb_var= (uint16_t*)s->dummy;
-        s->m.current_picture.f.mb_type = s->dummy;
+        // dummies, to avoid segfaults
+        s->m.current_picture.mb_mean   = (uint8_t *)s->dummy;
+        s->m.current_picture.mb_var    = (uint16_t *)s->dummy;
+        s->m.current_picture.mc_mb_var = (uint16_t *)s->dummy;
+        s->m.current_picture.mb_type = s->dummy;
 
-        s->m.current_picture.f.motion_val[0] = s->motion_val8[plane] + 2;
-        s->m.p_mv_table= s->motion_val16[plane] + s->m.mb_stride + 1;
-        s->m.dsp= s->dsp; //move
+        s->m.current_picture.motion_val[0]   = s->motion_val8[plane] + 2;
+        s->m.p_mv_table                      = s->motion_val16[plane] +
+                                               s->m.mb_stride + 1;
+        s->m.mecc                            = s->mecc; // move
         ff_init_me(&s->m);
 
-        s->m.me.dia_size= s->avctx->dia_size;
-        s->m.first_slice_line=1;
+        s->m.me.dia_size      = s->avctx->dia_size;
+        s->m.first_slice_line = 1;
         for (y = 0; y < block_height; y++) {
-            s->m.new_picture.f.data[0] = src - y*16*stride; //ugly
-            s->m.mb_y= y;
-
-            for(i=0; i<16 && i + 16*y<height; i++){
-                memcpy(&src[i*stride], &src_plane[(i+16*y)*src_stride], width);
-                for(x=width; x<16*block_width; x++)
-                    src[i*stride+x]= src[i*stride+x-1];
+            s->m.new_picture.f->data[0] = src - y * 16 * stride; // ugly
+            s->m.mb_y                  = y;
+
+            for (i = 0; i < 16 && i + 16 * y < height; i++) {
+                memcpy(&src[i * stride], &src_plane[(i + 16 * y) * src_stride],
+                       width);
+                for (x = width; x < 16 * block_width; x++)
+                    src[i * stride + x] = src[i * stride + x - 1];
             }
-            for(; i<16 && i + 16*y<16*block_height; i++)
-                memcpy(&src[i*stride], &src[(i-1)*stride], 16*block_width);
+            for (; i < 16 && i + 16 * y < 16 * block_height; i++)
+                memcpy(&src[i * stride], &src[(i - 1) * stride],
+                       16 * block_width);
 
             for (x = 0; x < block_width; x++) {
-                s->m.mb_x= x;
+                s->m.mb_x = x;
                 ff_init_block_index(&s->m);
                 ff_update_block_index(&s->m);
 
                 ff_estimate_p_frame_motion(&s->m, x, y);
             }
-            s->m.first_slice_line=0;
+            s->m.first_slice_line = 0;
         }
 
         ff_fix_long_p_mvs(&s->m);
-        ff_fix_long_mvs(&s->m, NULL, 0, s->m.p_mv_table, s->m.f_code, CANDIDATE_MB_TYPE_INTER, 0);
+        ff_fix_long_mvs(&s->m, NULL, 0, s->m.p_mv_table, s->m.f_code,
+                        CANDIDATE_MB_TYPE_INTER, 0);
     }
 
-    s->m.first_slice_line=1;
+    s->m.first_slice_line = 1;
     for (y = 0; y < block_height; y++) {
-        for(i=0; i<16 && i + 16*y<height; i++){
-            memcpy(&src[i*stride], &src_plane[(i+16*y)*src_stride], width);
-            for(x=width; x<16*block_width; x++)
-                src[i*stride+x]= src[i*stride+x-1];
+        for (i = 0; i < 16 && i + 16 * y < height; i++) {
+            memcpy(&src[i * stride], &src_plane[(i + 16 * y) * src_stride],
+                   width);
+            for (x = width; x < 16 * block_width; x++)
+                src[i * stride + x] = src[i * stride + x - 1];
         }
-        for(; i<16 && i + 16*y<16*block_height; i++)
-            memcpy(&src[i*stride], &src[(i-1)*stride], 16*block_width);
+        for (; i < 16 && i + 16 * y < 16 * block_height; i++)
+            memcpy(&src[i * stride], &src[(i - 1) * stride], 16 * block_width);
 
-        s->m.mb_y= y;
+        s->m.mb_y = y;
         for (x = 0; x < block_width; x++) {
-            uint8_t reorder_buffer[3][6][7*32];
+            uint8_t reorder_buffer[3][6][7 * 32];
             int count[3][6];
-            int offset = y * 16 * stride + x * 16;
-            uint8_t *decoded= decoded_plane + offset;
-            uint8_t *ref= ref_plane + offset;
-            int score[4]={0,0,0,0}, best;
-            uint8_t *temp = s->scratchbuf;
-
-            if(s->pb.buf_end - s->pb.buf - (put_bits_count(&s->pb)>>3) < 3000){ //FIXME check size
+            int offset       = y * 16 * stride + x * 16;
+            uint8_t *decoded = decoded_plane + offset;
+            uint8_t *ref     = ref_plane + offset;
+            int score[4]     = { 0, 0, 0, 0 }, best;
+            uint8_t *temp    = s->scratchbuf;
+
+            if (s->pb.buf_end - s->pb.buf -
+                (put_bits_count(&s->pb) >> 3) < 3000) { // FIXME: check size
                 av_log(s->avctx, AV_LOG_ERROR, "encoded frame too large\n");
                 return -1;
             }
 
-            s->m.mb_x= x;
+            s->m.mb_x = x;
             ff_init_block_index(&s->m);
             ff_update_block_index(&s->m);
 
-            if(s->picture.pict_type == AV_PICTURE_TYPE_I || (s->m.mb_type[x + y*s->m.mb_stride]&CANDIDATE_MB_TYPE_INTRA)){
-                for(i=0; i<6; i++){
-                    init_put_bits(&s->reorder_pb[i], reorder_buffer[0][i], 7*32);
-                }
-                if(s->picture.pict_type == AV_PICTURE_TYPE_P){
-                    const uint8_t *vlc= ff_svq1_block_type_vlc[SVQ1_BLOCK_INTRA];
+            if (f->pict_type == AV_PICTURE_TYPE_I ||
+                (s->m.mb_type[x + y * s->m.mb_stride] &
+                 CANDIDATE_MB_TYPE_INTRA)) {
+                for (i = 0; i < 6; i++)
+                    init_put_bits(&s->reorder_pb[i], reorder_buffer[0][i],
+                                  7 * 32);
+                if (f->pict_type == AV_PICTURE_TYPE_P) {
+                    const uint8_t *vlc = ff_svq1_block_type_vlc[SVQ1_BLOCK_INTRA];
                     put_bits(&s->reorder_pb[5], vlc[1], vlc[0]);
-                    score[0]= vlc[1]*lambda;
+                    score[0] = vlc[1] * lambda;
                 }
-                score[0]+= encode_block(s, src+16*x, NULL, temp, stride, 5, 64, lambda, 1);
-                for(i=0; i<6; i++){
-                    count[0][i]= put_bits_count(&s->reorder_pb[i]);
+                score[0] += encode_block(s, src + 16 * x, NULL, temp, stride,
+                                         5, 64, lambda, 1);
+                for (i = 0; i < 6; i++) {
+                    count[0][i] = put_bits_count(&s->reorder_pb[i]);
                     flush_put_bits(&s->reorder_pb[i]);
                 }
-            }else
-                score[0]= INT_MAX;
+            } else
+                score[0] = INT_MAX;
 
-            best=0;
+            best = 0;
 
-            if(s->picture.pict_type == AV_PICTURE_TYPE_P){
-                const uint8_t *vlc= ff_svq1_block_type_vlc[SVQ1_BLOCK_INTER];
+            if (f->pict_type == AV_PICTURE_TYPE_P) {
+                const uint8_t *vlc = ff_svq1_block_type_vlc[SVQ1_BLOCK_INTER];
                 int mx, my, pred_x, pred_y, dxy;
                 int16_t *motion_ptr;
 
-                motion_ptr= ff_h263_pred_motion(&s->m, 0, 0, &pred_x, &pred_y);
-                if(s->m.mb_type[x + y*s->m.mb_stride]&CANDIDATE_MB_TYPE_INTER){
-                    for(i=0; i<6; i++)
-                        init_put_bits(&s->reorder_pb[i], reorder_buffer[1][i], 7*32);
+                motion_ptr = ff_h263_pred_motion(&s->m, 0, 0, &pred_x, &pred_y);
+                if (s->m.mb_type[x + y * s->m.mb_stride] &
+                    CANDIDATE_MB_TYPE_INTER) {
+                    for (i = 0; i < 6; i++)
+                        init_put_bits(&s->reorder_pb[i], reorder_buffer[1][i],
+                                      7 * 32);
 
                     put_bits(&s->reorder_pb[5], vlc[1], vlc[0]);
 
-                    s->m.pb= s->reorder_pb[5];
-                    mx= motion_ptr[0];
-                    my= motion_ptr[1];
-                    assert(mx>=-32 && mx<=31);
-                    assert(my>=-32 && my<=31);
-                    assert(pred_x>=-32 && pred_x<=31);
-                    assert(pred_y>=-32 && pred_y<=31);
+                    s->m.pb = s->reorder_pb[5];
+                    mx      = motion_ptr[0];
+                    my      = motion_ptr[1];
+                    assert(mx     >= -32 && mx     <= 31);
+                    assert(my     >= -32 && my     <= 31);
+                    assert(pred_x >= -32 && pred_x <= 31);
+                    assert(pred_y >= -32 && pred_y <= 31);
                     ff_h263_encode_motion(&s->m, mx - pred_x, 1);
                     ff_h263_encode_motion(&s->m, my - pred_y, 1);
-                    s->reorder_pb[5]= s->m.pb;
-                    score[1] += lambda*put_bits_count(&s->reorder_pb[5]);
-
-                    dxy= (mx&1) + 2*(my&1);
-
-                    s->dsp.put_pixels_tab[0][dxy](temp+16, ref + (mx>>1) + stride*(my>>1), stride, 16);
-
-                    score[1]+= encode_block(s, src+16*x, temp+16, decoded, stride, 5, 64, lambda, 0);
-                    best= score[1] <= score[0];
-
-                    vlc= ff_svq1_block_type_vlc[SVQ1_BLOCK_SKIP];
-                    score[2]= s->dsp.sse[0](NULL, src+16*x, ref, stride, 16);
-                    score[2]+= vlc[1]*lambda;
-                    if(score[2] < score[best] && mx==0 && my==0){
-                        best=2;
-                        s->dsp.put_pixels_tab[0][0](decoded, ref, stride, 16);
-                        for(i=0; i<6; i++){
-                            count[2][i]=0;
-                        }
+                    s->reorder_pb[5] = s->m.pb;
+                    score[1]        += lambda * put_bits_count(&s->reorder_pb[5]);
+
+                    dxy = (mx & 1) + 2 * (my & 1);
+
+                    s->hdsp.put_pixels_tab[0][dxy](temp + 16,
+                                                   ref + (mx >> 1) +
+                                                   stride * (my >> 1),
+                                                   stride, 16);
+
+                    score[1] += encode_block(s, src + 16 * x, temp + 16,
+                                             decoded, stride, 5, 64, lambda, 0);
+                    best      = score[1] <= score[0];
+
+                    vlc       = ff_svq1_block_type_vlc[SVQ1_BLOCK_SKIP];
+                    score[2]  = s->mecc.sse[0](NULL, src + 16 * x, ref,
+                                               stride, 16);
+                    score[2] += vlc[1] * lambda;
+                    if (score[2] < score[best] && mx == 0 && my == 0) {
+                        best = 2;
+                        s->hdsp.put_pixels_tab[0][0](decoded, ref, stride, 16);
+                        for (i = 0; i < 6; i++)
+                            count[2][i] = 0;
                         put_bits(&s->pb, vlc[1], vlc[0]);
                     }
                 }
 
-                if(best==1){
-                    for(i=0; i<6; i++){
-                        count[1][i]= put_bits_count(&s->reorder_pb[i]);
+                if (best == 1) {
+                    for (i = 0; i < 6; i++) {
+                        count[1][i] = put_bits_count(&s->reorder_pb[i]);
                         flush_put_bits(&s->reorder_pb[i]);
                     }
-                }else{
-                    motion_ptr[0                 ] = motion_ptr[1                 ]=
-                    motion_ptr[2                 ] = motion_ptr[3                 ]=
-                    motion_ptr[0+2*s->m.b8_stride] = motion_ptr[1+2*s->m.b8_stride]=
-                    motion_ptr[2+2*s->m.b8_stride] = motion_ptr[3+2*s->m.b8_stride]=0;
+                } else {
+                    motion_ptr[0]                      =
+                    motion_ptr[1]                      =
+                    motion_ptr[2]                      =
+                    motion_ptr[3]                      =
+                    motion_ptr[0 + 2 * s->m.b8_stride] =
+                    motion_ptr[1 + 2 * s->m.b8_stride] =
+                    motion_ptr[2 + 2 * s->m.b8_stride] =
+                    motion_ptr[3 + 2 * s->m.b8_stride] = 0;
                 }
             }
 
             s->rd_total += score[best];
 
-            for(i=5; i>=0; i--){
-                avpriv_copy_bits(&s->pb, reorder_buffer[best][i], count[best][i]);
-            }
-            if(best==0){
-                s->dsp.put_pixels_tab[0][0](decoded, temp, stride, 16);
-            }
+            for (i = 5; i >= 0; i--)
+                avpriv_copy_bits(&s->pb, reorder_buffer[best][i],
+                                 count[best][i]);
+            if (best == 0)
+                s->hdsp.put_pixels_tab[0][0](decoded, temp, stride, 16);
         }
-        s->m.first_slice_line=0;
+        s->m.first_slice_line = 0;
     }
     return 0;
 }
 
-static av_cold int svq1_encode_init(AVCodecContext *avctx)
+static av_cold int svq1_encode_end(AVCodecContext *avctx)
 {
-    SVQ1Context * const s = avctx->priv_data;
+    SVQ1EncContext *const s = avctx->priv_data;
+    int i;
+
+    av_log(avctx, AV_LOG_DEBUG, "RD: %f\n",
+           s->rd_total / (double)(avctx->width * avctx->height *
+                                  avctx->frame_number));
 
-    ff_dsputil_init(&s->dsp, avctx);
-    avctx->coded_frame = &s->picture;
+    s->m.mb_type = NULL;
+    ff_mpv_common_end(&s->m);
 
-    s->frame_width = avctx->width;
+    av_freep(&s->m.me.scratchpad);
+    av_freep(&s->m.me.map);
+    av_freep(&s->m.me.score_map);
+    av_freep(&s->mb_type);
+    av_freep(&s->dummy);
+    av_freep(&s->scratchbuf);
+
+    for (i = 0; i < 3; i++) {
+        av_freep(&s->motion_val8[i]);
+        av_freep(&s->motion_val16[i]);
+    }
+
+    av_frame_free(&s->current_picture);
+    av_frame_free(&s->last_picture);
+    av_frame_free(&avctx->coded_frame);
+
+    return 0;
+}
+
+static av_cold int svq1_encode_init(AVCodecContext *avctx)
+{
+    SVQ1EncContext *const s = avctx->priv_data;
+    int ret;
+
+    ff_hpeldsp_init(&s->hdsp, avctx->flags);
+    ff_me_cmp_init(&s->mecc, avctx);
+    ff_mpegvideoencdsp_init(&s->m.mpvencdsp, avctx);
+
+    avctx->coded_frame = av_frame_alloc();
+    s->current_picture = av_frame_alloc();
+    s->last_picture    = av_frame_alloc();
+    if (!avctx->coded_frame || !s->current_picture || !s->last_picture) {
+        svq1_encode_end(avctx);
+        return AVERROR(ENOMEM);
+    }
+
+    s->frame_width  = avctx->width;
     s->frame_height = avctx->height;
 
-    s->y_block_width = (s->frame_width + 15) / 16;
+    s->y_block_width  = (s->frame_width  + 15) / 16;
     s->y_block_height = (s->frame_height + 15) / 16;
 
-    s->c_block_width = (s->frame_width / 4 + 15) / 16;
+    s->c_block_width  = (s->frame_width  / 4 + 15) / 16;
     s->c_block_height = (s->frame_height / 4 + 15) / 16;
 
-    s->avctx= avctx;
-    s->m.avctx= avctx;
+    s->avctx               = avctx;
+    s->m.avctx             = avctx;
+
+    if ((ret = ff_mpv_common_init(&s->m)) < 0) {
+        svq1_encode_end(avctx);
+        return ret;
+    }
+
     s->m.picture_structure = PICT_FRAME;
-    s->m.me.temp      =
-    s->m.me.scratchpad= av_mallocz((avctx->width+64)*2*16*2*sizeof(uint8_t));
-    s->m.me.map       = av_mallocz(ME_MAP_SIZE*sizeof(uint32_t));
-    s->m.me.score_map = av_mallocz(ME_MAP_SIZE*sizeof(uint32_t));
-    s->mb_type        = av_mallocz((s->y_block_width+1)*s->y_block_height*sizeof(int16_t));
-    s->dummy          = av_mallocz((s->y_block_width+1)*s->y_block_height*sizeof(int32_t));
-    ff_h263_encode_init(&s->m); //mv_penalty
+    s->m.me.temp           =
+    s->m.me.scratchpad     = av_mallocz((avctx->width + 64) *
+                                        2 * 16 * 2 * sizeof(uint8_t));
+    s->m.me.map            = av_mallocz(ME_MAP_SIZE * sizeof(uint32_t));
+    s->m.me.score_map      = av_mallocz(ME_MAP_SIZE * sizeof(uint32_t));
+    s->mb_type             = av_mallocz((s->y_block_width + 1) *
+                                        s->y_block_height * sizeof(int16_t));
+    s->dummy               = av_mallocz((s->y_block_width + 1) *
+                                        s->y_block_height * sizeof(int32_t));
+    s->ssd_int8_vs_int16   = ssd_int8_vs_int16_c;
+
+    if (ARCH_PPC)
+        ff_svq1enc_init_ppc(s);
+    if (ARCH_X86)
+        ff_svq1enc_init_x86(s);
+
+    ff_h263_encode_init(&s->m); // mv_penalty
 
     return 0;
 }
@@ -501,49 +563,61 @@ static av_cold int svq1_encode_init(AVCodecContext *avctx)
 static int svq1_encode_frame(AVCodecContext *avctx, AVPacket *pkt,
                              const AVFrame *pict, int *got_packet)
 {
-    SVQ1Context * const s = avctx->priv_data;
-    AVFrame * const p = &s->picture;
-    AVFrame temp;
+    SVQ1EncContext *const s = avctx->priv_data;
+    AVFrame *const p        = avctx->coded_frame;
     int i, ret;
 
     if (!pkt->data &&
-        (ret = av_new_packet(pkt, s->y_block_width*s->y_block_height*MAX_MB_BYTES*3 + FF_MIN_BUFFER_SIZE) < 0)) {
+        (ret = av_new_packet(pkt, s->y_block_width * s->y_block_height *
+                             MAX_MB_BYTES * 3 + FF_MIN_BUFFER_SIZE)) < 0) {
         av_log(avctx, AV_LOG_ERROR, "Error getting output packet.\n");
         return ret;
     }
 
-    if(avctx->pix_fmt != PIX_FMT_YUV410P){
+    if (avctx->pix_fmt != AV_PIX_FMT_YUV410P) {
         av_log(avctx, AV_LOG_ERROR, "unsupported pixel format\n");
         return -1;
     }
 
-    if(!s->current_picture.data[0]){
-        avctx->get_buffer(avctx, &s->current_picture);
-        avctx->get_buffer(avctx, &s->last_picture);
-        s->scratchbuf = av_malloc(s->current_picture.linesize[0] * 16 * 2);
+    if (!s->current_picture->data[0]) {
+        ret = ff_get_buffer(avctx, s->current_picture, 0);
+        if (ret < 0)
+            return ret;
+    }
+    if (!s->last_picture->data[0]) {
+        ret = ff_get_buffer(avctx, s->last_picture, 0);
+        if (ret < 0)
+            return ret;
+    }
+    if (!s->scratchbuf) {
+        s->scratchbuf = av_malloc(s->current_picture->linesize[0] * 16 * 2);
+        if (!s->scratchbuf)
+            return AVERROR(ENOMEM);
     }
 
-    temp= s->current_picture;
-    s->current_picture= s->last_picture;
-    s->last_picture= temp;
+    FFSWAP(AVFrame*, s->current_picture, s->last_picture);
 
     init_put_bits(&s->pb, pkt->data, pkt->size);
 
-    *p = *pict;
-    p->pict_type = avctx->gop_size && avctx->frame_number % avctx->gop_size ? AV_PICTURE_TYPE_P : AV_PICTURE_TYPE_I;
+    p->pict_type = avctx->gop_size && avctx->frame_number % avctx->gop_size ?
+                   AV_PICTURE_TYPE_P : AV_PICTURE_TYPE_I;
     p->key_frame = p->pict_type == AV_PICTURE_TYPE_I;
+    p->quality   = pict->quality;
 
     svq1_write_header(s, p->pict_type);
-    for(i=0; i<3; i++){
-        if(svq1_encode_plane(s, i,
-            s->picture.data[i], s->last_picture.data[i], s->current_picture.data[i],
-            s->frame_width / (i?4:1), s->frame_height / (i?4:1),
-            s->picture.linesize[i], s->current_picture.linesize[i]) < 0)
-                return -1;
-    }
-
-//    avpriv_align_put_bits(&s->pb);
-    while(put_bits_count(&s->pb) & 31)
+    for (i = 0; i < 3; i++)
+        if (svq1_encode_plane(s, i,
+                              pict->data[i],
+                              s->last_picture->data[i],
+                              s->current_picture->data[i],
+                              s->frame_width  / (i ? 4 : 1),
+                              s->frame_height / (i ? 4 : 1),
+                              pict->linesize[i],
+                              s->current_picture->linesize[i]) < 0)
+            return -1;
+
+    // avpriv_align_put_bits(&s->pb);
+    while (put_bits_count(&s->pb) & 31)
         put_bits(&s->pb, 1, 0);
 
     flush_put_bits(&s->pb);
@@ -556,37 +630,15 @@ static int svq1_encode_frame(AVCodecContext *avctx, AVPacket *pkt,
     return 0;
 }
 
-static av_cold int svq1_encode_end(AVCodecContext *avctx)
-{
-    SVQ1Context * const s = avctx->priv_data;
-    int i;
-
-    av_log(avctx, AV_LOG_DEBUG, "RD: %f\n", s->rd_total/(double)(avctx->width*avctx->height*avctx->frame_number));
-
-    av_freep(&s->m.me.scratchpad);
-    av_freep(&s->m.me.map);
-    av_freep(&s->m.me.score_map);
-    av_freep(&s->mb_type);
-    av_freep(&s->dummy);
-    av_freep(&s->scratchbuf);
-
-    for(i=0; i<3; i++){
-        av_freep(&s->motion_val8[i]);
-        av_freep(&s->motion_val16[i]);
-    }
-
-    return 0;
-}
-
-
 AVCodec ff_svq1_encoder = {
     .name           = "svq1",
+    .long_name      = NULL_IF_CONFIG_SMALL("Sorenson Vector Quantizer 1 / Sorenson Video 1 / SVQ1"),
     .type           = AVMEDIA_TYPE_VIDEO,
     .id             = AV_CODEC_ID_SVQ1,
-    .priv_data_size = sizeof(SVQ1Context),
+    .priv_data_size = sizeof(SVQ1EncContext),
     .init           = svq1_encode_init,
     .encode2        = svq1_encode_frame,
     .close          = svq1_encode_end,
-    .pix_fmts       = (const enum PixelFormat[]){ PIX_FMT_YUV410P, PIX_FMT_NONE },
-    .long_name      = NULL_IF_CONFIG_SMALL("Sorenson Vector Quantizer 1 / Sorenson Video 1 / SVQ1"),
+    .pix_fmts       = (const enum AVPixelFormat[]) { AV_PIX_FMT_YUV410P,
+                                                     AV_PIX_FMT_NONE },
 };