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>

5 files changed, 73 insertions, 5 deletions

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