Merge remote-tracking branch 'qatar/master'

* qatar/master: (23 commits)
  avconv: Reformat s16 volume adjustment.
  ARM: NEON optimised vector_fmac_scalar()
  dca: use vector_fmac_scalar from dsputil
  dsputil: add vector_fmac_scalar()
  latmenc: Fix private options
  vf_unsharp: store hsub/vsub in the filter context
  vf_unsharp: adopt a more natural order of params in apply_unsharp()
  vf_unsharp: rename method "unsharpen" to "apply_unsharp"
  vf_scale: apply the same transform to the aspect during init that is applied per frame
  vf_pad: fix "vsub" variable value computation
  vf_scale: add a "sar" variable
  lavfi: fix realloc size computation in avfilter_add_format()
  vsrc_color: use internal timebase
  lavfi: fix signature for avfilter_graph_parse() and avfilter_graph_config()
  graphparser: prefer void * over AVClass * for log contexts
  avfiltergraph: use meaningful error codes
  avconv: Initialize return value for codec copy path.
  fate: use 'run' helper for seek-test
  fate: remove seek-mpeg2reuse test
  Fix memory (re)allocation in matroskadec.c, related to MSVR-11-0080.
  ...

Conflicts:
	doc/filters.texi
	libavfilter/avfilter.h
	libavfilter/avfiltergraph.c
	libavfilter/avfiltergraph.h
	libavfilter/graphparser.c
	libavfilter/vf_scale.c
	libavfilter/vsrc_color.c

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

13 files changed, 203 insertions, 27 deletions

diff --git a/avconv.c b/avconv.c
index 7bb1d312a4..acb497b759 100644
--- a/avconv.c
+++ b/avconv.c
@@ -153,7 +153,7 @@ static uint8_t *audio_buf;
 static uint8_t *audio_out;
 static unsigned int allocated_audio_out_size, allocated_audio_buf_size;
 
-static short *samples;
+static void *samples;
 
 #define DEFAULT_PASS_LOGFILENAME_PREFIX "av2pass"
 
@@ -1586,7 +1586,7 @@ static int output_packet(InputStream *ist, int ist_index,
 {
     AVFormatContext *os;
     OutputStream *ost;
-    int ret, i;
+    int ret = 0, i;
     int got_output;
     void *buffer_to_free = NULL;
     static unsigned int samples_size= 0;
@@ -1641,8 +1641,8 @@ static int output_packet(InputStream *ist, int ist_index,
         if (ist->decoding_needed) {
             switch(ist->st->codec->codec_type) {
             case AVMEDIA_TYPE_AUDIO:{
-                if(pkt && samples_size < FFMAX(pkt->size*sizeof(*samples), AVCODEC_MAX_AUDIO_FRAME_SIZE)) {
-                    samples_size = FFMAX(pkt->size*sizeof(*samples), AVCODEC_MAX_AUDIO_FRAME_SIZE);
+                if(pkt && samples_size < FFMAX(pkt->size * bps, AVCODEC_MAX_AUDIO_FRAME_SIZE)) {
+                    samples_size = FFMAX(pkt->size * bps, AVCODEC_MAX_AUDIO_FRAME_SIZE);
                     av_free(samples);
                     samples= av_malloc(samples_size);
                 }
@@ -1731,11 +1731,57 @@ static int output_packet(InputStream *ist, int ist_index,
         // preprocess audio (volume)
         if (ist->st->codec->codec_type == AVMEDIA_TYPE_AUDIO) {
             if (audio_volume != 256) {
-                short *volp;
-                volp = samples;
-                for(i=0;i<(decoded_data_size / sizeof(short));i++) {
-                    int v = ((*volp) * audio_volume + 128) >> 8;
-                    *volp++ = av_clip_int16(v);
+                switch (ist->st->codec->sample_fmt) {
+                case AV_SAMPLE_FMT_U8:
+                {
+                    uint8_t *volp = samples;
+                    for (i = 0; i < (decoded_data_size / sizeof(*volp)); i++) {
+                        int v = (((*volp - 128) * audio_volume + 128) >> 8) + 128;
+                        *volp++ = av_clip_uint8(v);
+                    }
+                    break;
+                }
+                case AV_SAMPLE_FMT_S16:
+                {
+                    int16_t *volp = samples;
+                    for (i = 0; i < (decoded_data_size / sizeof(*volp)); i++) {
+                        int v = ((*volp) * audio_volume + 128) >> 8;
+                        *volp++ = av_clip_int16(v);
+                    }
+                    break;
+                }
+                case AV_SAMPLE_FMT_S32:
+                {
+                    int32_t *volp = samples;
+                    for (i = 0; i < (decoded_data_size / sizeof(*volp)); i++) {
+                        int64_t v = (((int64_t)*volp * audio_volume + 128) >> 8);
+                        *volp++ = av_clipl_int32(v);
+                    }
+                    break;
+                }
+                case AV_SAMPLE_FMT_FLT:
+                {
+                    float *volp = samples;
+                    float scale = audio_volume / 256.f;
+                    for (i = 0; i < (decoded_data_size / sizeof(*volp)); i++) {
+                        *volp++ *= scale;
+                    }
+                    break;
+                }
+                case AV_SAMPLE_FMT_DBL:
+                {
+                    double *volp = samples;
+                    double scale = audio_volume / 256.;
+                    for (i = 0; i < (decoded_data_size / sizeof(*volp)); i++) {
+                        *volp++ *= scale;
+                    }
+                    break;
+                }
+                default:
+                    av_log(NULL, AV_LOG_FATAL,
+                           "Audio volume adjustment on sample format %s is not supported.\n",
+                           av_get_sample_fmt_name(ist->st->codec->sample_fmt));
+                    exit_program(1);
                 }
             }
         }
diff --git a/doc/filters.texi b/doc/filters.texi
index b40dc5d7b7..bbd4e228af 100644
--- a/doc/filters.texi
+++ b/doc/filters.texi
@@ -1687,6 +1687,9 @@ input sample aspect ratio
 @item dar
 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
 
+@item sar
+input sample aspect ratio
+
 @item hsub, vsub
 horizontal and vertical chroma subsample values. For example for the
 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
diff --git a/ffmpeg.c b/ffmpeg.c
index 0ced8f8e0b..b250f0a944 100644
--- a/ffmpeg.c
+++ b/ffmpeg.c
@@ -160,7 +160,7 @@ static uint8_t *audio_buf;
 static uint8_t *audio_out;
 static unsigned int allocated_audio_out_size, allocated_audio_buf_size;
 
-static short *samples;
+static void *samples;
 static uint8_t *input_tmp= NULL;
 
 #define DEFAULT_PASS_LOGFILENAME_PREFIX "ffmpeg2pass"
@@ -1596,7 +1596,7 @@ static int output_packet(InputStream *ist, int ist_index,
 {
     AVFormatContext *os;
     OutputStream *ost;
-    int ret, i;
+    int ret = 0, i;
     int got_output;
     void *buffer_to_free = NULL;
     static unsigned int samples_size= 0;
@@ -1651,8 +1651,8 @@ static int output_packet(InputStream *ist, int ist_index,
         if (ist->decoding_needed) {
             switch(ist->st->codec->codec_type) {
             case AVMEDIA_TYPE_AUDIO:{
-                if(pkt && samples_size < FFMAX(pkt->size*sizeof(*samples), AVCODEC_MAX_AUDIO_FRAME_SIZE)) {
-                    samples_size = FFMAX(pkt->size*sizeof(*samples), AVCODEC_MAX_AUDIO_FRAME_SIZE);
+                if(pkt && samples_size < FFMAX(pkt->size * bps, AVCODEC_MAX_AUDIO_FRAME_SIZE)) {
+                    samples_size = FFMAX(pkt->size * bps, AVCODEC_MAX_AUDIO_FRAME_SIZE);
                     av_free(samples);
                     samples= av_malloc(samples_size);
                 }
@@ -1758,11 +1758,57 @@ static int output_packet(InputStream *ist, int ist_index,
         // preprocess audio (volume)
         if (ist->st->codec->codec_type == AVMEDIA_TYPE_AUDIO) {
             if (audio_volume != 256) {
-                short *volp;
-                volp = samples;
-                for(i=0;i<(decoded_data_size / sizeof(short));i++) {
-                    int v = ((*volp) * audio_volume + 128) >> 8;
-                    *volp++ = av_clip_int16(v);
+                switch (ist->st->codec->sample_fmt) {
+                case AV_SAMPLE_FMT_U8:
+                {
+                    uint8_t *volp = samples;
+                    for (i = 0; i < (decoded_data_size / sizeof(*volp)); i++) {
+                        int v = (((*volp - 128) * audio_volume + 128) >> 8) + 128;
+                        *volp++ = av_clip_uint8(v);
+                    }
+                    break;
+                }
+                case AV_SAMPLE_FMT_S16:
+                {
+                    int16_t *volp = samples;
+                    for (i = 0; i < (decoded_data_size / sizeof(*volp)); i++) {
+                        int v = ((*volp) * audio_volume + 128) >> 8;
+                        *volp++ = av_clip_int16(v);
+                    }
+                    break;
+                }
+                case AV_SAMPLE_FMT_S32:
+                {
+                    int32_t *volp = samples;
+                    for (i = 0; i < (decoded_data_size / sizeof(*volp)); i++) {
+                        int64_t v = (((int64_t)*volp * audio_volume + 128) >> 8);
+                        *volp++ = av_clipl_int32(v);
+                    }
+                    break;
+                }
+                case AV_SAMPLE_FMT_FLT:
+                {
+                    float *volp = samples;
+                    float scale = audio_volume / 256.f;
+                    for (i = 0; i < (decoded_data_size / sizeof(*volp)); i++) {
+                        *volp++ *= scale;
+                    }
+                    break;
+                }
+                case AV_SAMPLE_FMT_DBL:
+                {
+                    double *volp = samples;
+                    double scale = audio_volume / 256.;
+                    for (i = 0; i < (decoded_data_size / sizeof(*volp)); i++) {
+                        *volp++ *= scale;
+                    }
+                    break;
+                }
+                default:
+                    av_log(NULL, AV_LOG_FATAL,
+                           "Audio volume adjustment on sample format %s is not supported.\n",
+                           av_get_sample_fmt_name(ist->st->codec->sample_fmt));
+                    exit_program(1);
                 }
             }
         }
diff --git a/libavcodec/arm/dsputil_init_neon.c b/libavcodec/arm/dsputil_init_neon.c
index 15536d0bd2..ddc9d640f8 100644
--- a/libavcodec/arm/dsputil_init_neon.c
+++ b/libavcodec/arm/dsputil_init_neon.c
@@ -143,6 +143,8 @@ void ff_vector_fmul_window_neon(float *dst, const float *src0,
                                 const float *src1, const float *win, int len);
 void ff_vector_fmul_scalar_neon(float *dst, const float *src, float mul,
                                 int len);
+void ff_vector_fmac_scalar_neon(float *dst, const float *src, float mul,
+                                int len);
 void ff_butterflies_float_neon(float *v1, float *v2, int len);
 float ff_scalarproduct_float_neon(const float *v1, const float *v2, int len);
 void ff_vector_fmul_reverse_neon(float *dst, const float *src0,
@@ -305,6 +307,7 @@ void ff_dsputil_init_neon(DSPContext *c, AVCodecContext *avctx)
     c->vector_fmul                = ff_vector_fmul_neon;
     c->vector_fmul_window         = ff_vector_fmul_window_neon;
     c->vector_fmul_scalar         = ff_vector_fmul_scalar_neon;
+    c->vector_fmac_scalar         = ff_vector_fmac_scalar_neon;
     c->butterflies_float          = ff_butterflies_float_neon;
     c->scalarproduct_float        = ff_scalarproduct_float_neon;
     c->vector_fmul_reverse        = ff_vector_fmul_reverse_neon;
diff --git a/libavcodec/arm/dsputil_neon.S b/libavcodec/arm/dsputil_neon.S
index 94a7a8cb75..1574ad6496 100644
--- a/libavcodec/arm/dsputil_neon.S
+++ b/libavcodec/arm/dsputil_neon.S
@@ -587,6 +587,54 @@ NOVFP   vdup.32         q8,  r2
         .unreq          len
 endfunc
 
+function ff_vector_fmac_scalar_neon, export=1
+VFP     len .req r2
+VFP     acc .req r3
+NOVFP   len .req r3
+NOVFP   acc .req r2
+VFP     vdup.32         q15, d0[0]
+NOVFP   vdup.32         q15, r2
+        bics            r12, len, #15
+        mov             acc, r0
+        beq             3f
+        vld1.32         {q0},     [r1,:128]!
+        vld1.32         {q8},     [acc,:128]!
+        vld1.32         {q1},     [r1,:128]!
+        vld1.32         {q9},     [acc,:128]!
+1:      vmla.f32        q8,  q0,  q15
+        vld1.32         {q2},     [r1,:128]!
+        vld1.32         {q10},    [acc,:128]!
+        vmla.f32        q9,  q1,  q15
+        vld1.32         {q3},     [r1,:128]!
+        vld1.32         {q11},    [acc,:128]!
+        vmla.f32        q10, q2,  q15
+        vst1.32         {q8},     [r0,:128]!
+        vmla.f32        q11, q3,  q15
+        vst1.32         {q9},     [r0,:128]!
+        subs            r12, r12, #16
+        beq             2f
+        vld1.32         {q0},     [r1,:128]!
+        vld1.32         {q8},     [acc,:128]!
+        vst1.32         {q10},    [r0,:128]!
+        vld1.32         {q1},     [r1,:128]!
+        vld1.32         {q9},     [acc,:128]!
+        vst1.32         {q11},    [r0,:128]!
+        b               1b
+2:      vst1.32         {q10},    [r0,:128]!
+        vst1.32         {q11},    [r0,:128]!
+        ands            len, len, #15
+        it              eq
+        bxeq            lr
+3:      vld1.32         {q0},     [r1,:128]!
+        vld1.32         {q8},     [acc,:128]!
+        vmla.f32        q8,  q0,  q15
+        vst1.32         {q8},     [r0,:128]!
+        subs            len, len, #4
+        bgt             3b
+        bx              lr
+        .unreq          len
+endfunc
+
 function ff_butterflies_float_neon, export=1
 1:      vld1.32         {q0},[r0,:128]
         vld1.32         {q1},[r1,:128]
diff --git a/libavcodec/dca.c b/libavcodec/dca.c
index e11439f939..762821c3dc 100644
--- a/libavcodec/dca.c
+++ b/libavcodec/dca.c
@@ -1833,11 +1833,8 @@ static int dca_decode_frame(AVCodecContext * avctx,
             float* back_chan = s->samples + s->channel_order_tab[s->xch_base_channel] * 256;
             float* lt_chan   = s->samples + s->channel_order_tab[s->xch_base_channel - 2] * 256;
             float* rt_chan   = s->samples + s->channel_order_tab[s->xch_base_channel - 1] * 256;
-            int j;
-            for(j = 0; j < 256; ++j) {
-                lt_chan[j] -= back_chan[j] * M_SQRT1_2;
-                rt_chan[j] -= back_chan[j] * M_SQRT1_2;
-            }
+            s->dsp.vector_fmac_scalar(lt_chan, back_chan, -M_SQRT1_2, 256);
+            s->dsp.vector_fmac_scalar(rt_chan, back_chan, -M_SQRT1_2, 256);
         }
 
         if (avctx->sample_fmt == AV_SAMPLE_FMT_FLT) {
diff --git a/libavcodec/dsputil.c b/libavcodec/dsputil.c
index c64a39d250..ebce93039f 100644
--- a/libavcodec/dsputil.c
+++ b/libavcodec/dsputil.c
@@ -2443,6 +2443,14 @@ static void vector_fmul_scalar_c(float *dst, const float *src, float mul,
         dst[i] = src[i] * mul;
 }
 
+static void vector_fmac_scalar_c(float *dst, const float *src, float mul,
+                                 int len)
+{
+    int i;
+    for (i = 0; i < len; i++)
+        dst[i] += src[i] * mul;
+}
+
 static void butterflies_float_c(float *restrict v1, float *restrict v2,
                                 int len)
 {
@@ -2978,6 +2986,7 @@ av_cold void dsputil_init(DSPContext* c, AVCodecContext *avctx)
     c->scalarproduct_float = scalarproduct_float_c;
     c->butterflies_float = butterflies_float_c;
     c->vector_fmul_scalar = vector_fmul_scalar_c;
+    c->vector_fmac_scalar = vector_fmac_scalar_c;
 
     c->shrink[0]= av_image_copy_plane;
     c->shrink[1]= ff_shrink22;
diff --git a/libavcodec/dsputil.h b/libavcodec/dsputil.h
index 07ef196185..057c41cca5 100644
--- a/libavcodec/dsputil.h
+++ b/libavcodec/dsputil.h
@@ -424,6 +424,17 @@ typedef struct DSPContext {
     void (*vector_fmul_scalar)(float *dst, const float *src, float mul,
                                int len);
     /**
+     * Multiply a vector of floats by a scalar float and add to
+     * destination vector.  Source and destination vectors must
+     * overlap exactly or not at all.
+     * @param dst result vector, 16-byte aligned
+     * @param src input vector, 16-byte aligned
+     * @param mul scalar value
+     * @param len length of vector, multiple of 4
+     */
+    void (*vector_fmac_scalar)(float *dst, const float *src, float mul,
+                               int len);
+    /**
      * Calculate the scalar product of two vectors of floats.
      * @param v1  first vector, 16-byte aligned
      * @param v2  second vector, 16-byte aligned
diff --git a/libavfilter/avfilter.h b/libavfilter/avfilter.h
index 19c26b5245..5c73256601 100644
--- a/libavfilter/avfilter.h
+++ b/libavfilter/avfilter.h
@@ -30,7 +30,7 @@
 
 #define LIBAVFILTER_VERSION_MAJOR  2
 #define LIBAVFILTER_VERSION_MINOR 43
-#define LIBAVFILTER_VERSION_MICRO  5
+#define LIBAVFILTER_VERSION_MICRO  6
 
 #define LIBAVFILTER_VERSION_INT AV_VERSION_INT(LIBAVFILTER_VERSION_MAJOR, \
                                                LIBAVFILTER_VERSION_MINOR, \
diff --git a/libavfilter/avfiltergraph.h b/libavfilter/avfiltergraph.h
index 3062b04fd8..76fb8aed43 100644
--- a/libavfilter/avfiltergraph.h
+++ b/libavfilter/avfiltergraph.h
@@ -171,4 +171,5 @@ int avfilter_graph_send_command(AVFilterGraph *graph, const char *target, const
 int avfilter_graph_queue_command(AVFilterGraph *graph, const char *target, const char *cmd, const char *arg, int flags, double ts);
 
 
+
 #endif /* AVFILTER_AVFILTERGRAPH_H */
diff --git a/libavformat/latmenc.c b/libavformat/latmenc.c
index 9e85ee5672..9299cc3a7b 100644
--- a/libavformat/latmenc.c
+++ b/libavformat/latmenc.c
@@ -28,6 +28,7 @@
 #include "rawenc.h"
 
 typedef struct {
+    AVClass *av_class;
     int off;
     int channel_conf;
     int object_type;
diff --git a/libavformat/matroskadec.c b/libavformat/matroskadec.c
index c05a0b5653..dccf589177 100644
--- a/libavformat/matroskadec.c
+++ b/libavformat/matroskadec.c
@@ -967,6 +967,7 @@ static int matroska_decode_buffer(uint8_t** buf, int* buf_size,
     uint8_t* data = *buf;
     int isize = *buf_size;
     uint8_t* pkt_data = NULL;
+    uint8_t* newpktdata;
     int pkt_size = isize;
     int result = 0;
     int olen;
@@ -996,7 +997,12 @@ static int matroska_decode_buffer(uint8_t** buf, int* buf_size,
         zstream.avail_in = isize;
         do {
             pkt_size *= 3;
-            pkt_data = av_realloc(pkt_data, pkt_size);
+            newpktdata = av_realloc(pkt_data, pkt_size);
+            if (!newpktdata) {
+                inflateEnd(&zstream);
+                goto failed;
+            }
+            pkt_data = newpktdata;
             zstream.avail_out = pkt_size - zstream.total_out;
             zstream.next_out = pkt_data + zstream.total_out;
             if (pkt_data) {
@@ -1020,7 +1026,12 @@ static int matroska_decode_buffer(uint8_t** buf, int* buf_size,
         bzstream.avail_in = isize;
         do {
             pkt_size *= 3;
-            pkt_data = av_realloc(pkt_data, pkt_size);
+            newpktdata = av_realloc(pkt_data, pkt_size);
+            if (!newpktdata) {
+                BZ2_bzDecompressEnd(&bzstream);
+                goto failed;
+            }
+            pkt_data = newpktdata;
             bzstream.avail_out = pkt_size - bzstream.total_out_lo32;
             bzstream.next_out = pkt_data + bzstream.total_out_lo32;
             if (pkt_data) {
diff --git a/tests/fate-run.sh b/tests/fate-run.sh
index a09c13ab47..a2aa7cbb9e 100755
--- a/tests/fate-run.sh
+++ b/tests/fate-run.sh
@@ -105,7 +105,7 @@ seektest(){
                  file=$(echo tests/data/$d/$file)
                  ;;
     esac
-    $target_exec $target_path/libavformat/seek-test $target_path/$file
+    run libavformat/seek-test $target_path/$file
 }
 
 mkdir -p "$outdir"