Merge remote-tracking branch 'khirnov/release/0.7' into release/0.8

* khirnov/release/0.7: (64 commits)
  rv34: Check for invalid slice offsets
  rv34: Fix potential overreads
  rv34: Avoid NULL dereference on corrupted bitstream
  rv10: Reject slices that does not have the same type as the first one
  lavf: Fix context pointer in av_open_input_stream when avformat_open_input fails
  oggdec: fix out of bound write in the ogg demuxer
  Fixed size given to init_get_bits().
  smacker: fix a few off by 1 errors
  Check for invalid VLC value in smacker decoder.
  Check and propagate errors when VLC trees cannot be built in smacker decoder.
  Fixed off by one packet size allocation in the smacker demuxer.
  Check for invalid packet size in the smacker demuxer.
  ape demuxer: fix segfault on memory allocation failure.
  xan: Add some buffer checks (cherry picked from commit 0872bb23b4bd2d94a8ba91070f706d1bc1c3ced8)
  Fixed size given to init_get_bits() in xan decoder. (cherry picked from commit 393d5031c6aaaf8c2dda4eb5d676974c349fae85)
  smacker demuxer: handle possible av_realloc() failure.
  Fixed segfault with wavpack decoder on corrupted decorrelation terms sub-blocks.
  cljr: init_get_bits size in bits instead of bytes (cherry picked from commit 0c1f5b93d9b97c4cc3684ba91a040e90bfc760d2)
  indeo2: fail if input buffer too small (cherry picked from commit b7ce4f1d1c3add86ece7ca595ea6c4a10b471055)
  indeo2: init_get_bits size in bits instead of bytes (cherry picked from commit 68ca330cbd479111db9cb7649d7530ad59f04cc8)
  ...

Conflicts:
	ffmpeg.c
	libavdevice/alsa-audio.h
	libavformat/gxf.c
	libswscale/x86/swscale_template.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

15 files changed, 110 insertions, 53 deletions

diff --git a/libavcodec/aacdec.c b/libavcodec/aacdec.c
index cb8760801a..6cc0cec0f0 100644
--- a/libavcodec/aacdec.c
+++ b/libavcodec/aacdec.c
@@ -1755,12 +1755,10 @@ static void windowing_and_mdct_ltp(AACContext *ac, float *out,
     } else {
         memset(in, 0, 448 * sizeof(float));
         ac->dsp.vector_fmul(in + 448, in + 448, swindow_prev, 128);
-        memcpy(in + 576, in + 576, 448 * sizeof(float));
     }
     if (ics->window_sequence[0] != LONG_START_SEQUENCE) {
         ac->dsp.vector_fmul_reverse(in + 1024, in + 1024, lwindow, 1024);
     } else {
-        memcpy(in + 1024, in + 1024, 448 * sizeof(float));
         ac->dsp.vector_fmul_reverse(in + 1024 + 448, in + 1024 + 448, swindow, 128);
         memset(in + 1024 + 576, 0, 448 * sizeof(float));
     }
@@ -2078,7 +2076,7 @@ static int aac_decode_frame_int(AVCodecContext *avctx, void *data,
     ChannelElement *che = NULL, *che_prev = NULL;
     enum RawDataBlockType elem_type, elem_type_prev = TYPE_END;
     int err, elem_id, data_size_tmp;
-    int samples = 0, multiplier;
+    int samples = 0, multiplier, audio_found = 0;
 
     if (show_bits(gb, 12) == 0xfff) {
         if (parse_adts_frame_header(ac, gb) < 0) {
@@ -2109,10 +2107,12 @@ static int aac_decode_frame_int(AVCodecContext *avctx, void *data,
 
         case TYPE_SCE:
             err = decode_ics(ac, &che->ch[0], gb, 0, 0);
+            audio_found = 1;
             break;
 
         case TYPE_CPE:
             err = decode_cpe(ac, gb, che);
+            audio_found = 1;
             break;
 
         case TYPE_CCE:
@@ -2121,6 +2121,7 @@ static int aac_decode_frame_int(AVCodecContext *avctx, void *data,
 
         case TYPE_LFE:
             err = decode_ics(ac, &che->ch[0], gb, 0, 0);
+            audio_found = 1;
             break;
 
         case TYPE_DSE:
@@ -2197,7 +2198,7 @@ static int aac_decode_frame_int(AVCodecContext *avctx, void *data,
                                                    samples, avctx->channels);
     }
 
-    if (ac->output_configured)
+    if (ac->output_configured && audio_found)
         ac->output_configured = OC_LOCKED;
 
     return 0;
diff --git a/libavcodec/aacps.c b/libavcodec/aacps.c
index fc124d1972..3cfd18d959 100644
--- a/libavcodec/aacps.c
+++ b/libavcodec/aacps.c
@@ -813,14 +813,17 @@ static void stereo_processing(PSContext *ps, float (*l)[32][2], float (*r)[32][2
     const float (*H_LUT)[8][4] = (PS_BASELINE || ps->icc_mode < 3) ? HA : HB;
 
     //Remapping
-    memcpy(H11[0][0], H11[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H11[0][0][0]));
-    memcpy(H11[1][0], H11[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H11[1][0][0]));
-    memcpy(H12[0][0], H12[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H12[0][0][0]));
-    memcpy(H12[1][0], H12[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H12[1][0][0]));
-    memcpy(H21[0][0], H21[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H21[0][0][0]));
-    memcpy(H21[1][0], H21[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H21[1][0][0]));
-    memcpy(H22[0][0], H22[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H22[0][0][0]));
-    memcpy(H22[1][0], H22[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H22[1][0][0]));
+    if (ps->num_env_old) {
+        memcpy(H11[0][0], H11[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H11[0][0][0]));
+        memcpy(H11[1][0], H11[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H11[1][0][0]));
+        memcpy(H12[0][0], H12[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H12[0][0][0]));
+        memcpy(H12[1][0], H12[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H12[1][0][0]));
+        memcpy(H21[0][0], H21[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H21[0][0][0]));
+        memcpy(H21[1][0], H21[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H21[1][0][0]));
+        memcpy(H22[0][0], H22[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H22[0][0][0]));
+        memcpy(H22[1][0], H22[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H22[1][0][0]));
+    }
+
     if (is34) {
         remap34(&iid_mapped, ps->iid_par, ps->nr_iid_par, ps->num_env, 1);
         remap34(&icc_mapped, ps->icc_par, ps->nr_icc_par, ps->num_env, 1);
diff --git a/libavcodec/arm/fft_fixed_neon.S b/libavcodec/arm/fft_fixed_neon.S
index bd6c853ec8..4e4b94a8e3 100644
--- a/libavcodec/arm/fft_fixed_neon.S
+++ b/libavcodec/arm/fft_fixed_neon.S
@@ -56,7 +56,7 @@
         vhsub.s16       \r0, \d0, \d1           @ t3, t4, t8, t7
         vhsub.s16       \r1, \d1, \d0
         vhadd.s16       \d0, \d0, \d1           @ t1, t2, t6, t5
-        vmov.i64        \d1, #0xffff<<32
+        vmov.i64        \d1, #0xffff00000000
         vbit            \r0, \r1, \d1
         vrev64.16       \r1, \r0                @ t7, t8, t4, t3
         vtrn.32         \r0, \r1                @ t3, t4, t7, t8
diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
index 7c8228fb04..6048f20881 100644
--- a/libavcodec/avcodec.h
+++ b/libavcodec/avcodec.h
@@ -2606,7 +2606,6 @@ typedef struct AVCodecContext {
 
     /**
      * Bits per sample/pixel of internal libavcodec pixel/sample format.
-     * This field is applicable only when sample_fmt is AV_SAMPLE_FMT_S32.
      * - encoding: set by user.
      * - decoding: set by libavcodec.
      */
diff --git a/libavcodec/dca.c b/libavcodec/dca.c
index 4b445bc743..5166fc5337 100644
--- a/libavcodec/dca.c
+++ b/libavcodec/dca.c
@@ -1650,6 +1650,7 @@ static int dca_decode_frame(AVCodecContext * avctx,
     //set AVCodec values with parsed data
     avctx->sample_rate = s->sample_rate;
     avctx->bit_rate = s->bit_rate;
+    avctx->frame_size = s->sample_blocks * 32;
 
     s->profile = FF_PROFILE_DTS;
 
diff --git a/libavcodec/h264.c b/libavcodec/h264.c
index 5841b1f399..f227ddfedf 100644
--- a/libavcodec/h264.c
+++ b/libavcodec/h264.c
@@ -1173,7 +1173,10 @@ static int decode_update_thread_context(AVCodecContext *dst, const AVCodecContex
         memcpy(&h->s + 1, &h1->s + 1, sizeof(H264Context) - sizeof(MpegEncContext)); //copy all fields after MpegEnc
         memset(h->sps_buffers, 0, sizeof(h->sps_buffers));
         memset(h->pps_buffers, 0, sizeof(h->pps_buffers));
-        ff_h264_alloc_tables(h);
+        if (ff_h264_alloc_tables(h) < 0) {
+            av_log(dst, AV_LOG_ERROR, "Could not allocate memory for h264\n");
+            return AVERROR(ENOMEM);
+        }
         context_init(h);
 
         for(i=0; i<2; i++){
@@ -1856,15 +1859,30 @@ static av_always_inline void hl_decode_mb_internal(H264Context *h, int simple, i
                     tmp_y[j] = get_bits(&gb, bit_depth);
             }
             if(simple || !CONFIG_GRAY || !(s->flags&CODEC_FLAG_GRAY)){
-                for (i = 0; i < 8; i++) {
-                    uint16_t *tmp_cb = (uint16_t*)(dest_cb + i*uvlinesize);
-                    for (j = 0; j < 8; j++)
-                        tmp_cb[j] = get_bits(&gb, bit_depth);
-                }
-                for (i = 0; i < 8; i++) {
-                    uint16_t *tmp_cr = (uint16_t*)(dest_cr + i*uvlinesize);
-                    for (j = 0; j < 8; j++)
-                        tmp_cr[j] = get_bits(&gb, bit_depth);
+                if (!h->sps.chroma_format_idc) {
+                    for (i = 0; i < 8; i++) {
+                        uint16_t *tmp_cb = (uint16_t*)(dest_cb + i*uvlinesize);
+                        for (j = 0; j < 8; j++) {
+                            tmp_cb[j] = 1 << (bit_depth - 1);
+                        }
+                    }
+                    for (i = 0; i < 8; i++) {
+                        uint16_t *tmp_cr = (uint16_t*)(dest_cr + i*uvlinesize);
+                        for (j = 0; j < 8; j++) {
+                            tmp_cr[j] = 1 << (bit_depth - 1);
+                        }
+                    }
+                } else {
+                    for (i = 0; i < 8; i++) {
+                        uint16_t *tmp_cb = (uint16_t*)(dest_cb + i*uvlinesize);
+                        for (j = 0; j < 8; j++)
+                            tmp_cb[j] = get_bits(&gb, bit_depth);
+                    }
+                    for (i = 0; i < 8; i++) {
+                        uint16_t *tmp_cr = (uint16_t*)(dest_cr + i*uvlinesize);
+                        for (j = 0; j < 8; j++)
+                            tmp_cr[j] = get_bits(&gb, bit_depth);
+                    }
                 }
             }
         } else {
@@ -1872,9 +1890,16 @@ static av_always_inline void hl_decode_mb_internal(H264Context *h, int simple, i
                 memcpy(dest_y + i*  linesize, h->mb       + i*8, 16);
             }
             if(simple || !CONFIG_GRAY || !(s->flags&CODEC_FLAG_GRAY)){
-                for (i=0; i<8; i++) {
-                    memcpy(dest_cb+ i*uvlinesize, h->mb + 128 + i*4,  8);
-                    memcpy(dest_cr+ i*uvlinesize, h->mb + 160 + i*4,  8);
+                if (!h->sps.chroma_format_idc) {
+                    for (i = 0; i < 8; i++) {
+                        memset(dest_cb + i*uvlinesize, 128, 8);
+                        memset(dest_cr + i*uvlinesize, 128, 8);
+                    }
+                } else {
+                    for (i = 0; i < 8; i++) {
+                        memcpy(dest_cb + i*uvlinesize, h->mb + 128 + i*4,  8);
+                        memcpy(dest_cr + i*uvlinesize, h->mb + 160 + i*4,  8);
+                    }
                 }
             }
         }
@@ -2206,15 +2231,17 @@ static void implicit_weight_table(H264Context *h, int field){
     for(ref0=ref_start; ref0 < ref_count0; ref0++){
         int poc0 = h->ref_list[0][ref0].poc;
         for(ref1=ref_start; ref1 < ref_count1; ref1++){
-            int poc1 = h->ref_list[1][ref1].poc;
-            int td = av_clip(poc1 - poc0, -128, 127);
-            int w= 32;
-            if(td){
-                int tb = av_clip(cur_poc - poc0, -128, 127);
-                int tx = (16384 + (FFABS(td) >> 1)) / td;
-                int dist_scale_factor = (tb*tx + 32) >> 8;
-                if(dist_scale_factor >= -64 && dist_scale_factor <= 128)
-                    w = 64 - dist_scale_factor;
+            int w = 32;
+            if (!h->ref_list[0][ref0].long_ref && !h->ref_list[1][ref1].long_ref) {
+                int poc1 = h->ref_list[1][ref1].poc;
+                int td = av_clip(poc1 - poc0, -128, 127);
+                if(td){
+                    int tb = av_clip(cur_poc - poc0, -128, 127);
+                    int tx = (16384 + (FFABS(td) >> 1)) / td;
+                    int dist_scale_factor = (tb*tx + 32) >> 8;
+                    if(dist_scale_factor >= -64 && dist_scale_factor <= 128)
+                        w = 64 - dist_scale_factor;
+                }
             }
             if(field<0){
                 h->implicit_weight[ref0][ref1][0]=
@@ -2646,7 +2673,10 @@ static int decode_slice_header(H264Context *h, H264Context *h0){
         h->prev_interlaced_frame = 1;
 
         init_scan_tables(h);
-        ff_h264_alloc_tables(h);
+        if (ff_h264_alloc_tables(h) < 0) {
+            av_log(h->s.avctx, AV_LOG_ERROR, "Could not allocate memory for h264\n");
+            return AVERROR(ENOMEM);
+        }
 
         if (!HAVE_THREADS || !(s->avctx->active_thread_type&FF_THREAD_SLICE)) {
             if (context_init(h) < 0) {
diff --git a/libavcodec/h264_refs.c b/libavcodec/h264_refs.c
index 9554201522..83a5564e4c 100644
--- a/libavcodec/h264_refs.c
+++ b/libavcodec/h264_refs.c
@@ -678,7 +678,7 @@ int ff_h264_decode_ref_pic_marking(H264Context *h, GetBitContext *gb){
                 }
                 if(opcode==MMCO_SHORT2LONG || opcode==MMCO_LONG2UNUSED || opcode==MMCO_LONG || opcode==MMCO_SET_MAX_LONG){
                     unsigned int long_arg= get_ue_golomb_31(gb);
-                    if(long_arg >= 32 || (long_arg >= 16 && !(opcode == MMCO_LONG2UNUSED && FIELD_PICTURE))){
+                    if(long_arg >= 32 || (long_arg >= 16 && !(opcode == MMCO_SET_MAX_LONG && long_arg == 16) && !(opcode == MMCO_LONG2UNUSED && FIELD_PICTURE))){
                         av_log(h->s.avctx, AV_LOG_ERROR, "illegal long ref in memory management control operation %d\n", opcode);
                         return -1;
                     }
diff --git a/libavcodec/kgv1dec.c b/libavcodec/kgv1dec.c
index c364cfc593..3e3922774e 100644
--- a/libavcodec/kgv1dec.c
+++ b/libavcodec/kgv1dec.c
@@ -174,6 +174,5 @@ AVCodec ff_kgv1_decoder = {
     NULL,
     decode_end,
     decode_frame,
-    .max_lowres = 1,
     .long_name = NULL_IF_CONFIG_SMALL("Kega Game Video"),
 };
diff --git a/libavcodec/mpegvideo.c b/libavcodec/mpegvideo.c
index c3c3e15033..d819cc083f 100644
--- a/libavcodec/mpegvideo.c
+++ b/libavcodec/mpegvideo.c
@@ -285,9 +285,10 @@ int ff_alloc_picture(MpegEncContext *s, Picture *pic, int shared){
         }
 
         FF_ALLOCZ_OR_GOTO(s->avctx, pic->mbskip_table , mb_array_size * sizeof(uint8_t)+2, fail) //the +2 is for the slice end check
-        FF_ALLOCZ_OR_GOTO(s->avctx, pic->qscale_table , mb_array_size * sizeof(uint8_t)  , fail)
+        FF_ALLOCZ_OR_GOTO(s->avctx, pic->qscale_table_base , (big_mb_num + s->mb_stride) * sizeof(uint8_t)  , fail)
         FF_ALLOCZ_OR_GOTO(s->avctx, pic->mb_type_base , (big_mb_num + s->mb_stride) * sizeof(uint32_t), fail)
         pic->mb_type= pic->mb_type_base + 2*s->mb_stride+1;
+        pic->qscale_table = pic->qscale_table_base + 2*s->mb_stride + 1;
         if(s->out_format == FMT_H264){
             for(i=0; i<2; i++){
                 FF_ALLOCZ_OR_GOTO(s->avctx, pic->motion_val_base[i], 2 * (b4_array_size+4)  * sizeof(int16_t), fail)
@@ -339,7 +340,7 @@ static void free_picture(MpegEncContext *s, Picture *pic){
     av_freep(&pic->mc_mb_var);
     av_freep(&pic->mb_mean);
     av_freep(&pic->mbskip_table);
-    av_freep(&pic->qscale_table);
+    av_freep(&pic->qscale_table_base);
     av_freep(&pic->mb_type_base);
     av_freep(&pic->dct_coeff);
     av_freep(&pic->pan_scan);
diff --git a/libavcodec/mpegvideo.h b/libavcodec/mpegvideo.h
index a0ff354a08..0b233d0a20 100644
--- a/libavcodec/mpegvideo.h
+++ b/libavcodec/mpegvideo.h
@@ -88,6 +88,7 @@ typedef struct Picture{
      * halfpel luma planes.
      */
     uint8_t *interpolated[3];
+    int8_t *qscale_table_base;
     int16_t (*motion_val_base[2])[2];
     uint32_t *mb_type_base;
 #define MB_TYPE_INTRA MB_TYPE_INTRA4x4 //default mb_type if there is just one type
diff --git a/libavcodec/options.c b/libavcodec/options.c
index b6ad5d8763..4845857eff 100644
--- a/libavcodec/options.c
+++ b/libavcodec/options.c
@@ -446,10 +446,9 @@ static const AVOption options[]={
 {"lpc_passes", "deprecated, use flac-specific options", OFFSET(lpc_passes), FF_OPT_TYPE_INT, {.dbl = -1 }, INT_MIN, INT_MAX, A|E},
 #endif
 {"slices", "number of slices, used in parallelized decoding", OFFSET(slices), FF_OPT_TYPE_INT, {.dbl = 0 }, 0, INT_MAX, V|E},
-{"thread_type", "select multithreading type", OFFSET(thread_type), FF_OPT_TYPE_INT, {.dbl = FF_THREAD_SLICE|FF_THREAD_FRAME }, 0, INT_MAX, V|E|D, "thread_type"},
+{"thread_type", "select multithreading type", OFFSET(thread_type), FF_OPT_TYPE_FLAGS, {.dbl = FF_THREAD_SLICE|FF_THREAD_FRAME }, 0, INT_MAX, V|E|D, "thread_type"},
 {"slice", NULL, 0, FF_OPT_TYPE_CONST, {.dbl = FF_THREAD_SLICE }, INT_MIN, INT_MAX, V|E|D, "thread_type"},
 {"frame", NULL, 0, FF_OPT_TYPE_CONST, {.dbl = FF_THREAD_FRAME }, INT_MIN, INT_MAX, V|E|D, "thread_type"},
-{"vbv_delay", "initial buffer fill time in periods of 27Mhz clock", 0, FF_OPT_TYPE_INT64, {.dbl = 0 }, 0, INT64_MAX},
 {"audio_service_type", "audio service type", OFFSET(audio_service_type), FF_OPT_TYPE_INT, {.dbl = AV_AUDIO_SERVICE_TYPE_MAIN }, 0, AV_AUDIO_SERVICE_TYPE_NB-1, A|E, "audio_service_type"},
 {"ma", "Main Audio Service", 0, FF_OPT_TYPE_CONST, {.dbl = AV_AUDIO_SERVICE_TYPE_MAIN },              INT_MIN, INT_MAX, A|E, "audio_service_type"},
 {"ef", "Effects",            0, FF_OPT_TYPE_CONST, {.dbl = AV_AUDIO_SERVICE_TYPE_EFFECTS },           INT_MIN, INT_MAX, A|E, "audio_service_type"},
diff --git a/libavcodec/rv30.c b/libavcodec/rv30.c
index 935866da9a..600656b466 100644
--- a/libavcodec/rv30.c
+++ b/libavcodec/rv30.c
@@ -261,6 +261,7 @@ static av_cold int rv30_decode_init(AVCodecContext *avctx)
     if(avctx->extradata_size - 8 < (r->rpr - 1) * 2){
         av_log(avctx, AV_LOG_ERROR, "Insufficient extradata - need at least %d bytes, got %d\n",
                6 + r->rpr * 2, avctx->extradata_size);
+        return AVERROR(EINVAL);
     }
     r->parse_slice_header = rv30_parse_slice_header;
     r->decode_intra_types = rv30_decode_intra_types;
diff --git a/libavcodec/vc1.c b/libavcodec/vc1.c
index d6b48b46d0..90130a0b25 100644
--- a/libavcodec/vc1.c
+++ b/libavcodec/vc1.c
@@ -485,8 +485,8 @@ static int decode_sequence_header_adv(VC1Context *v, GetBitContext *gb)
         if(ar && ar < 14){
             v->s.avctx->sample_aspect_ratio = ff_vc1_pixel_aspect[ar];
         }else if(ar == 15){
-            w = get_bits(gb, 8);
-            h = get_bits(gb, 8);
+            w = get_bits(gb, 8) + 1;
+            h = get_bits(gb, 8) + 1;
             v->s.avctx->sample_aspect_ratio = (AVRational){w, h};
         }
         av_log(v->s.avctx, AV_LOG_DEBUG, "Aspect: %i:%i\n", v->s.avctx->sample_aspect_ratio.num, v->s.avctx->sample_aspect_ratio.den);
diff --git a/libavcodec/vc1dec.c b/libavcodec/vc1dec.c
index e82fba2fcb..d9ff6a22f3 100644
--- a/libavcodec/vc1dec.c
+++ b/libavcodec/vc1dec.c
@@ -243,7 +243,7 @@ static void vc1_loop_filter_iblk(VC1Context *v, int pq)
     }
     v->vc1dsp.vc1_v_loop_filter16(s->dest[0] + 8*s->linesize, s->linesize, pq);
 
-    if (s->mb_y == s->mb_height-1) {
+    if (s->mb_y == s->end_mb_y-1) {
         if (s->mb_x) {
             v->vc1dsp.vc1_h_loop_filter16(s->dest[0], s->linesize, pq);
             v->vc1dsp.vc1_h_loop_filter8(s->dest[1], s->uvlinesize, pq);
@@ -295,7 +295,7 @@ static void vc1_loop_filter_iblk_delayed(VC1Context *v, int pq)
             v->vc1dsp.vc1_v_loop_filter16(s->dest[0] - 8 * s->linesize, s->linesize, pq);
         }
 
-        if (s->mb_y == s->mb_height) {
+        if (s->mb_y == s->end_mb_y) {
             if (s->mb_x) {
                 if (s->mb_x >= 2)
                     v->vc1dsp.vc1_h_loop_filter16(s->dest[0] - 16 * s->linesize - 16, s->linesize, pq);
@@ -2330,7 +2330,7 @@ static av_always_inline void vc1_apply_p_v_loop_filter(VC1Context *v, int block_
     } else {
         dst      = s->dest[0] + (block_num & 1) * 8 + ((block_num & 2) * 4 - 8) * linesize;
     }
-    if (s->mb_y != s->mb_height || block_num < 2) {
+    if (s->mb_y != s->end_mb_y || block_num < 2) {
         int16_t (*mv)[2];
         int mv_stride;
 
@@ -3020,7 +3020,7 @@ static void vc1_decode_i_blocks_adv(VC1Context *v)
         s->mb_x = 0;
         ff_init_block_index(s);
         memset(&s->coded_block[s->block_index[0]-s->b8_stride], 0,
-               s->b8_stride * sizeof(*s->coded_block));
+               (1 + s->b8_stride) * sizeof(*s->coded_block));
     }
     for(; s->mb_y < s->end_mb_y; s->mb_y++) {
         s->mb_x = 0;
@@ -3096,7 +3096,7 @@ static void vc1_decode_i_blocks_adv(VC1Context *v)
         if(v->s.loop_filter) vc1_loop_filter_iblk_delayed(v, v->pq);
     }
     if (v->s.loop_filter)
-        ff_draw_horiz_band(s, (s->mb_height-1)*16, 16);
+        ff_draw_horiz_band(s, (s->end_mb_y-1)*16, 16);
     ff_er_add_slice(s, 0, s->start_mb_y, s->mb_width - 1, s->end_mb_y - 1, (AC_END|DC_END|MV_END));
 }
 
@@ -3219,7 +3219,7 @@ static void vc1_decode_b_blocks(VC1Context *v)
         s->first_slice_line = 0;
     }
     if (v->s.loop_filter)
-        ff_draw_horiz_band(s, (s->mb_height-1)*16, 16);
+        ff_draw_horiz_band(s, (s->end_mb_y-1)*16, 16);
     ff_er_add_slice(s, 0, s->start_mb_y, s->mb_width - 1, s->end_mb_y - 1, (AC_END|DC_END|MV_END));
 }
 
@@ -3227,9 +3227,9 @@ static void vc1_decode_skip_blocks(VC1Context *v)
 {
     MpegEncContext *s = &v->s;
 
-    ff_er_add_slice(s, 0, 0, s->mb_width - 1, s->mb_height - 1, (AC_END|DC_END|MV_END));
+    ff_er_add_slice(s, 0, s->start_mb_y, s->mb_width - 1, s->end_mb_y - 1, (AC_END|DC_END|MV_END));
     s->first_slice_line = 1;
-    for(s->mb_y = 0; s->mb_y < s->mb_height; s->mb_y++) {
+    for(s->mb_y = s->start_mb_y; s->mb_y < s->end_mb_y; s->mb_y++) {
         s->mb_x = 0;
         ff_init_block_index(s);
         ff_update_block_index(s);
diff --git a/libavcodec/vp3.c b/libavcodec/vp3.c
index 148f1179e3..b9af998bc5 100644
--- a/libavcodec/vp3.c
+++ b/libavcodec/vp3.c
@@ -2321,6 +2321,26 @@ static av_cold int theora_decode_init(AVCodecContext *avctx)
     return vp3_decode_init(avctx);
 }
 
+static void vp3_decode_flush(AVCodecContext *avctx)
+{
+    Vp3DecodeContext *s = avctx->priv_data;
+
+    if (s->golden_frame.data[0]) {
+        if (s->golden_frame.data[0] == s->last_frame.data[0])
+            memset(&s->last_frame, 0, sizeof(AVFrame));
+        if (s->current_frame.data[0] == s->golden_frame.data[0])
+            memset(&s->current_frame, 0, sizeof(AVFrame));
+        ff_thread_release_buffer(avctx, &s->golden_frame);
+    }
+    if (s->last_frame.data[0]) {
+        if (s->current_frame.data[0] == s->last_frame.data[0])
+            memset(&s->current_frame, 0, sizeof(AVFrame));
+        ff_thread_release_buffer(avctx, &s->last_frame);
+    }
+    if (s->current_frame.data[0])
+        ff_thread_release_buffer(avctx, &s->current_frame);
+}
+
 AVCodec ff_theora_decoder = {
     "theora",
     AVMEDIA_TYPE_VIDEO,
@@ -2332,6 +2352,7 @@ AVCodec ff_theora_decoder = {
     vp3_decode_frame,
     CODEC_CAP_DR1 | CODEC_CAP_DRAW_HORIZ_BAND | CODEC_CAP_FRAME_THREADS,
     NULL,
+    .flush = vp3_decode_flush,
     .long_name = NULL_IF_CONFIG_SMALL("Theora"),
     .update_thread_context = ONLY_IF_THREADS_ENABLED(vp3_update_thread_context)
 };
@@ -2348,6 +2369,7 @@ AVCodec ff_vp3_decoder = {
     vp3_decode_frame,
     CODEC_CAP_DR1 | CODEC_CAP_DRAW_HORIZ_BAND | CODEC_CAP_FRAME_THREADS,
     NULL,
+    .flush = vp3_decode_flush,
     .long_name = NULL_IF_CONFIG_SMALL("On2 VP3"),
     .update_thread_context = ONLY_IF_THREADS_ENABLED(vp3_update_thread_context)
 };