vc1dsp: introduce cases for 8x8 and 16x16

This allows further unrolling the DSP implementation where possible.

x86 and ARM DSP modified by simply moving the multiple calls from vc1dec
to the DSP code. Decoding improvements should only occurs because of the
compiler actually able to unroll more.

Decoding time: ~8.80s -> 8.64s (ie around 2%)

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

1 files changed, 55 insertions, 50 deletions

diff --git a/libavcodec/arm/vc1dsp_init_neon.c b/libavcodec/arm/vc1dsp_init_neon.c
index f562344b85..47ae34b056 100644
--- a/libavcodec/arm/vc1dsp_init_neon.c
+++ b/libavcodec/arm/vc1dsp_init_neon.c
@@ -35,40 +35,38 @@ void ff_vc1_inv_trans_4x4_dc_neon(uint8_t *dest, int linesize, int16_t *block);
 void ff_put_pixels8x8_neon(uint8_t *block, const uint8_t *pixels,
                            ptrdiff_t line_size, int rnd);
 
-void ff_put_vc1_mspel_mc10_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-void ff_put_vc1_mspel_mc20_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-void ff_put_vc1_mspel_mc30_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-
-void ff_put_vc1_mspel_mc01_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-void ff_put_vc1_mspel_mc02_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-void ff_put_vc1_mspel_mc03_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-
-void ff_put_vc1_mspel_mc11_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-void ff_put_vc1_mspel_mc12_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-void ff_put_vc1_mspel_mc13_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-
-void ff_put_vc1_mspel_mc21_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-void ff_put_vc1_mspel_mc22_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-void ff_put_vc1_mspel_mc23_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-
-void ff_put_vc1_mspel_mc31_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-void ff_put_vc1_mspel_mc32_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
-void ff_put_vc1_mspel_mc33_neon(uint8_t *dst, const uint8_t *src,
-                                ptrdiff_t stride, int rnd);
+#define DECL_PUT(X, Y) \
+void ff_put_vc1_mspel_mc##X##Y##_neon(uint8_t *dst, const uint8_t *src, \
+                                      ptrdiff_t stride, int rnd); \
+void ff_put_vc1_mspel_mc##X##Y##_16_neon(uint8_t *dst, const uint8_t *src, \
+                                         ptrdiff_t stride, int rnd) \
+{ \
+  ff_put_vc1_mspel_mc##X##Y##_neon(dst+0, src+0, stride, rnd); \
+  ff_put_vc1_mspel_mc##X##Y##_neon(dst+8, src+8, stride, rnd); \
+  dst += 8*stride; src += 8*stride; \
+  ff_put_vc1_mspel_mc##X##Y##_neon(dst+0, src+0, stride, rnd); \
+  ff_put_vc1_mspel_mc##X##Y##_neon(dst+8, src+8, stride, rnd); \
+}
+
+DECL_PUT(1, 0)
+DECL_PUT(2, 0)
+DECL_PUT(3, 0)
+
+DECL_PUT(0, 1)
+DECL_PUT(0, 2)
+DECL_PUT(0, 3)
+
+DECL_PUT(1, 1)
+DECL_PUT(1, 2)
+DECL_PUT(1, 3)
+
+DECL_PUT(2, 1)
+DECL_PUT(2, 2)
+DECL_PUT(2, 3)
+
+DECL_PUT(3, 1)
+DECL_PUT(3, 2)
+DECL_PUT(3, 3)
 
 void ff_put_vc1_chroma_mc8_neon(uint8_t *dst, uint8_t *src, int stride, int h,
                                 int x, int y);
@@ -79,6 +77,10 @@ void ff_put_vc1_chroma_mc4_neon(uint8_t *dst, uint8_t *src, int stride, int h,
 void ff_avg_vc1_chroma_mc4_neon(uint8_t *dst, uint8_t *src, int stride, int h,
                                 int x, int y);
 
+#define FN_ASSIGN(X, Y) \
+    dsp->put_vc1_mspel_pixels_tab[0][X+4*Y] = ff_put_vc1_mspel_mc##X##Y##_16_neon;
+    dsp->put_vc1_mspel_pixels_tab[1][X+4*Y] = ff_put_vc1_mspel_mc##X##Y##_neon
+
 av_cold void ff_vc1dsp_init_neon(VC1DSPContext *dsp)
 {
     dsp->vc1_inv_trans_8x8 = ff_vc1_inv_trans_8x8_neon;
@@ -90,22 +92,25 @@ av_cold void ff_vc1dsp_init_neon(VC1DSPContext *dsp)
     dsp->vc1_inv_trans_8x4_dc = ff_vc1_inv_trans_8x4_dc_neon;
     dsp->vc1_inv_trans_4x4_dc = ff_vc1_inv_trans_4x4_dc_neon;
 
-    dsp->put_vc1_mspel_pixels_tab[ 0] = ff_put_pixels8x8_neon;
-    dsp->put_vc1_mspel_pixels_tab[ 1] = ff_put_vc1_mspel_mc10_neon;
-    dsp->put_vc1_mspel_pixels_tab[ 2] = ff_put_vc1_mspel_mc20_neon;
-    dsp->put_vc1_mspel_pixels_tab[ 3] = ff_put_vc1_mspel_mc30_neon;
-    dsp->put_vc1_mspel_pixels_tab[ 4] = ff_put_vc1_mspel_mc01_neon;
-    dsp->put_vc1_mspel_pixels_tab[ 5] = ff_put_vc1_mspel_mc11_neon;
-    dsp->put_vc1_mspel_pixels_tab[ 6] = ff_put_vc1_mspel_mc21_neon;
-    dsp->put_vc1_mspel_pixels_tab[ 7] = ff_put_vc1_mspel_mc31_neon;
-    dsp->put_vc1_mspel_pixels_tab[ 8] = ff_put_vc1_mspel_mc02_neon;
-    dsp->put_vc1_mspel_pixels_tab[ 9] = ff_put_vc1_mspel_mc12_neon;
-    dsp->put_vc1_mspel_pixels_tab[10] = ff_put_vc1_mspel_mc22_neon;
-    dsp->put_vc1_mspel_pixels_tab[11] = ff_put_vc1_mspel_mc32_neon;
-    dsp->put_vc1_mspel_pixels_tab[12] = ff_put_vc1_mspel_mc03_neon;
-    dsp->put_vc1_mspel_pixels_tab[13] = ff_put_vc1_mspel_mc13_neon;
-    dsp->put_vc1_mspel_pixels_tab[14] = ff_put_vc1_mspel_mc23_neon;
-    dsp->put_vc1_mspel_pixels_tab[15] = ff_put_vc1_mspel_mc33_neon;
+    dsp->put_vc1_mspel_pixels_tab[0][ 0] = ff_put_pixels16x16_neon;
+    dsp->put_vc1_mspel_pixels_tab[1][ 0] = ff_put_pixels8x8_neon;
+    FN_ASSIGN(2, 0);
+    FN_ASSIGN(3, 0);
+
+    FN_ASSIGN(0, 1);
+    FN_ASSIGN(1, 1);
+    FN_ASSIGN(2, 1);
+    FN_ASSIGN(3, 1);
+
+    FN_ASSIGN(0, 2);
+    FN_ASSIGN(1, 2);
+    FN_ASSIGN(2, 2);
+    FN_ASSIGN(3, 2);
+
+    FN_ASSIGN(0, 3);
+    FN_ASSIGN(1, 3);
+    FN_ASSIGN(2, 3);
+    FN_ASSIGN(3, 3);
 
     dsp->put_no_rnd_vc1_chroma_pixels_tab[0] = ff_put_vc1_chroma_mc8_neon;
     dsp->avg_no_rnd_vc1_chroma_pixels_tab[0] = ff_avg_vc1_chroma_mc8_neon;