lavc: move SANE_NB_CHANNELS to internal.h and use it in the PCM decoders

[ffmpeg] / libavcodec / dsputil.h

diff --git a/libavcodec/dsputil.h b/libavcodec/dsputil.h
index c3a787c5f3f59d0886f26c0c717a69a6531bf81d..f48aa96017d2cf7017fd831967e3f3e278445eba 100644 (file)
--- a/libavcodec/dsputil.h
+++ b/libavcodec/dsputil.h
@@ -49,7 +49,7 @@ void ff_j_rev_dct (DCTELEM *data);
 void ff_wmv2_idct_c(DCTELEM *data);
 
 void ff_fdct_mmx(DCTELEM *block);
-void ff_fdct_mmx2(DCTELEM *block);
+void ff_fdct_mmxext(DCTELEM *block);
 void ff_fdct_sse2(DCTELEM *block);
 
 #define H264_IDCT(depth) \
@@ -101,9 +101,6 @@ PUTAVG_PIXELS(10)
 #define ff_put_pixels16x16_c ff_put_pixels16x16_8_c
 #define ff_avg_pixels16x16_c ff_avg_pixels16x16_8_c
 
-/* EA functions */
-void ff_ea_idct_put_c(uint8_t *dest, int linesize, DCTELEM *block);
-
 /* RV40 functions */
 void ff_put_rv40_qpel16_mc33_c(uint8_t *dst, uint8_t *src, int stride);
 void ff_avg_rv40_qpel16_mc33_c(uint8_t *dst, uint8_t *src, int stride);
@@ -200,10 +197,6 @@ EMULATED_EDGE(8)
 EMULATED_EDGE(9)
 EMULATED_EDGE(10)
 
-void ff_add_pixels_clamped_c(const DCTELEM *block, uint8_t *dest, int linesize);
-void ff_put_pixels_clamped_c(const DCTELEM *block, uint8_t *dest, int linesize);
-void ff_put_signed_pixels_clamped_c(const DCTELEM *block, uint8_t *dest, int linesize);
-
 /**
  * DSPContext.
  */
@@ -379,12 +372,8 @@ typedef struct DSPContext {
 
     void (*h261_loop_filter)(uint8_t *src, int stride);
 
-    void (*x8_v_loop_filter)(uint8_t *src, int stride, int qscale);
-    void (*x8_h_loop_filter)(uint8_t *src, int stride, int qscale);
-
     /* assume len is a multiple of 4, and arrays are 16-byte aligned */
     void (*vorbis_inverse_coupling)(float *mag, float *ang, int blocksize);
-    void (*ac3_downmix)(float (*samples)[256], float (*matrix)[2], int out_ch, int in_ch, int len);
     /* assume len is a multiple of 16, and arrays are 32-byte aligned */
     void (*vector_fmul_reverse)(float *dst, const float *src0, const float *src1, int len);
     /* assume len is a multiple of 8, and src arrays are 16-byte aligned */
@@ -490,17 +479,6 @@ typedef struct DSPContext {
 
     void (*shrink[4])(uint8_t *dst, int dst_wrap, const uint8_t *src, int src_wrap, int width, int height);
 
-    /* mlp/truehd functions */
-    void (*mlp_filter_channel)(int32_t *state, const int32_t *coeff,
-                               int firorder, int iirorder,
-                               unsigned int filter_shift, int32_t mask, int blocksize,
-                               int32_t *sample_buffer);
-
-    /* intrax8 functions */
-    void (*x8_spatial_compensation[12])(uint8_t *src , uint8_t *dst, int linesize);
-    void (*x8_setup_spatial_compensation)(uint8_t *src, uint8_t *dst, int linesize,
-           int * range, int * sum,  int edges);
-
     /**
      * Calculate scalar product of two vectors.
      * @param len length of vectors, should be multiple of 16
@@ -552,6 +530,17 @@ void ff_dsputil_init(DSPContext* p, AVCodecContext *avctx);
 
 int ff_check_alignment(void);
 
+/**
+ * Return the scalar product of two vectors.
+ *
+ * @param v1  first input vector
+ * @param v2  first input vector
+ * @param len number of elements
+ *
+ * @return sum of elementwise products
+ */
+float ff_scalarproduct_float_c(const float *v1, const float *v2, int len);
+
 /**
  * permute block according to permuatation.
  * @param last last non zero element in scantable order
@@ -610,39 +599,39 @@ static inline int get_penalty_factor(int lambda, int lambda2, int type){
 void ff_dsputil_init_alpha(DSPContext* c, AVCodecContext *avctx);
 void ff_dsputil_init_arm(DSPContext* c, AVCodecContext *avctx);
 void ff_dsputil_init_bfin(DSPContext* c, AVCodecContext *avctx);
-void ff_dsputil_init_mmi(DSPContext* c, AVCodecContext *avctx);
 void ff_dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx);
 void ff_dsputil_init_ppc(DSPContext* c, AVCodecContext *avctx);
 void ff_dsputil_init_sh4(DSPContext* c, AVCodecContext *avctx);
 void ff_dsputil_init_vis(DSPContext* c, AVCodecContext *avctx);
 
 void ff_dsputil_init_dwt(DSPContext *c);
-void ff_intrax8dsp_init(DSPContext* c, AVCodecContext *avctx);
-void ff_mlp_init(DSPContext* c, AVCodecContext *avctx);
-void ff_mlp_init_x86(DSPContext* c, AVCodecContext *avctx);
 
-#if (ARCH_ARM && HAVE_NEON) || ARCH_PPC || HAVE_MMI || HAVE_MMX
+#if (ARCH_ARM && HAVE_NEON) || ARCH_PPC || HAVE_MMX
 #   define STRIDE_ALIGN 16
 #else
 #   define STRIDE_ALIGN 8
 #endif
 
+// Some broken preprocessors need a second expansion
+// to be forced to tokenize __VA_ARGS__
+#define E(x) x
+
 #define LOCAL_ALIGNED_A(a, t, v, s, o, ...)             \
     uint8_t la_##v[sizeof(t s o) + (a)];                \
     t (*v) o = (void *)FFALIGN((uintptr_t)la_##v, a)
 
 #define LOCAL_ALIGNED_D(a, t, v, s, o, ...) DECLARE_ALIGNED(a, t, v) s o
 
-#define LOCAL_ALIGNED(a, t, v, ...) LOCAL_ALIGNED_A(a, t, v, __VA_ARGS__,,)
+#define LOCAL_ALIGNED(a, t, v, ...) E(LOCAL_ALIGNED_A(a, t, v, __VA_ARGS__,,))
 
 #if HAVE_LOCAL_ALIGNED_8
-#   define LOCAL_ALIGNED_8(t, v, ...) LOCAL_ALIGNED_D(8, t, v, __VA_ARGS__,,)
+#   define LOCAL_ALIGNED_8(t, v, ...) E(LOCAL_ALIGNED_D(8, t, v, __VA_ARGS__,,))
 #else
 #   define LOCAL_ALIGNED_8(t, v, ...) LOCAL_ALIGNED(8, t, v, __VA_ARGS__)
 #endif
 
 #if HAVE_LOCAL_ALIGNED_16
-#   define LOCAL_ALIGNED_16(t, v, ...) LOCAL_ALIGNED_D(16, t, v, __VA_ARGS__,,)
+#   define LOCAL_ALIGNED_16(t, v, ...) E(LOCAL_ALIGNED_D(16, t, v, __VA_ARGS__,,))
 #else
 #   define LOCAL_ALIGNED_16(t, v, ...) LOCAL_ALIGNED(16, t, v, __VA_ARGS__)
 #endif
@@ -673,7 +662,7 @@ static inline void copy_block2(uint8_t *dst, const uint8_t *src, int dstStride,
     int i;
     for(i=0; i<h; i++)
     {
-        AV_WN16(dst   , AV_RN16(src   ));
+        AV_COPY16U(dst, src);
         dst+=dstStride;
         src+=srcStride;
     }
@@ -684,7 +673,7 @@ static inline void copy_block4(uint8_t *dst, const uint8_t *src, int dstStride,
     int i;
     for(i=0; i<h; i++)
     {
-        AV_WN32(dst   , AV_RN32(src   ));
+        AV_COPY32U(dst, src);
         dst+=dstStride;
         src+=srcStride;
     }
@@ -695,8 +684,7 @@ static inline void copy_block8(uint8_t *dst, const uint8_t *src, int dstStride,
     int i;
     for(i=0; i<h; i++)
     {
-        AV_WN32(dst   , AV_RN32(src   ));
-        AV_WN32(dst+4 , AV_RN32(src+4 ));
+        AV_COPY64U(dst, src);
         dst+=dstStride;
         src+=srcStride;
     }
@@ -707,8 +695,7 @@ static inline void copy_block9(uint8_t *dst, const uint8_t *src, int dstStride,
     int i;
     for(i=0; i<h; i++)
     {
-        AV_WN32(dst   , AV_RN32(src   ));
-        AV_WN32(dst+4 , AV_RN32(src+4 ));
+        AV_COPY64U(dst, src);
         dst[8]= src[8];
         dst+=dstStride;
         src+=srcStride;
@@ -720,10 +707,7 @@ static inline void copy_block16(uint8_t *dst, const uint8_t *src, int dstStride,
     int i;
     for(i=0; i<h; i++)
     {
-        AV_WN32(dst   , AV_RN32(src   ));
-        AV_WN32(dst+4 , AV_RN32(src+4 ));
-        AV_WN32(dst+8 , AV_RN32(src+8 ));
-        AV_WN32(dst+12, AV_RN32(src+12));
+        AV_COPY128U(dst, src);
         dst+=dstStride;
         src+=srcStride;
     }
@@ -734,10 +718,7 @@ static inline void copy_block17(uint8_t *dst, const uint8_t *src, int dstStride,
     int i;
     for(i=0; i<h; i++)
     {
-        AV_WN32(dst   , AV_RN32(src   ));
-        AV_WN32(dst+4 , AV_RN32(src+4 ));
-        AV_WN32(dst+8 , AV_RN32(src+8 ));
-        AV_WN32(dst+12, AV_RN32(src+12));
+        AV_COPY128U(dst, src);
         dst[16]= src[16];
         dst+=dstStride;
         src+=srcStride;