From e8b9e1e2942355660e246cf4b9a18ce1a095419f Mon Sep 17 00:00:00 2001
From: Michael Niedermayer <michaelni@gmx.at>
Date: Thu, 5 Jul 2012 01:32:13 +0200
Subject: [PATCH] dsputil: extend things to 12 and 14 bits

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
---
 libavcodec/dsputil.c          | 28 ++++++++++++++++++++++++++++
 libavcodec/dsputil.h          |  6 ++++++
 libavcodec/dsputil_template.c | 28 +++++++++++++++++++---------
 3 files changed, 53 insertions(+), 9 deletions(-)

diff --git a/libavcodec/dsputil.c b/libavcodec/dsputil.c
index 9b32dff117..05140e32ea 100644
--- a/libavcodec/dsputil.c
+++ b/libavcodec/dsputil.c
@@ -43,6 +43,7 @@
 uint8_t ff_cropTbl[256 + 2 * MAX_NEG_CROP] = {0, };
 uint32_t ff_squareTbl[512] = {0, };
 
+#define pixeltmp int16_t
 #define BIT_DEPTH 9
 #include "dsputil_template.c"
 #undef BIT_DEPTH
@@ -51,8 +52,21 @@ uint32_t ff_squareTbl[512] = {0, };
 #include "dsputil_template.c"
 #undef BIT_DEPTH
 
+#undef pixeltmp
+#define pixeltmp int32_t
+#define BIT_DEPTH 12
+#include "dsputil_template.c"
+#undef BIT_DEPTH
+
+#define BIT_DEPTH 14
+#include "dsputil_template.c"
+#undef BIT_DEPTH
+
+#undef pixeltmp
+#define pixeltmp int16_t
 #define BIT_DEPTH 8
 #include "dsputil_template.c"
+#undef pixeltmp
 
 // 0x7f7f7f7f or 0x7f7f7f7f7f7f7f7f or whatever, depending on the cpu's native arithmetic size
 #define pb_7f (~0UL/255 * 0x7f)
@@ -3147,6 +3161,20 @@ av_cold void ff_dsputil_init(DSPContext* c, AVCodecContext *avctx)
             BIT_DEPTH_FUNCS(10, _16);
         }
         break;
+    case 12:
+        if (c->dct_bits == 32) {
+            BIT_DEPTH_FUNCS(12, _32);
+        } else {
+            BIT_DEPTH_FUNCS(12, _16);
+        }
+        break;
+    case 14:
+        if (c->dct_bits == 32) {
+            BIT_DEPTH_FUNCS(14, _32);
+        } else {
+            BIT_DEPTH_FUNCS(14, _16);
+        }
+        break;
     default:
         av_assert0(avctx->bits_per_raw_sample<=8 || avctx->codec_type != AVMEDIA_TYPE_VIDEO);
         BIT_DEPTH_FUNCS(8, _16);
diff --git a/libavcodec/dsputil.h b/libavcodec/dsputil.h
index c1397980f9..af1104825f 100644
--- a/libavcodec/dsputil.h
+++ b/libavcodec/dsputil.h
@@ -72,6 +72,8 @@ void ff_h264_chroma_dc_dequant_idct_ ## depth ## _c(DCTELEM *block, int qmul);
 H264_IDCT( 8)
 H264_IDCT( 9)
 H264_IDCT(10)
+H264_IDCT(12)
+H264_IDCT(14)
 
 void ff_svq3_luma_dc_dequant_idct_c(DCTELEM *output, DCTELEM *input, int qp);
 void ff_svq3_add_idct_c(uint8_t *dst, DCTELEM *block, int stride, int qp, int dc);
@@ -98,6 +100,8 @@ void ff_avg_pixels16x16_ ## depth ## _c(uint8_t *dst, uint8_t *src, int stride);
 PUTAVG_PIXELS( 8)
 PUTAVG_PIXELS( 9)
 PUTAVG_PIXELS(10)
+PUTAVG_PIXELS(12)
+PUTAVG_PIXELS(14)
 
 #define ff_put_pixels8x8_c ff_put_pixels8x8_8_c
 #define ff_avg_pixels8x8_c ff_avg_pixels8x8_8_c
@@ -211,6 +215,8 @@ void ff_emulated_edge_mc_ ## depth (uint8_t *buf, const uint8_t *src, int linesi
 EMULATED_EDGE(8)
 EMULATED_EDGE(9)
 EMULATED_EDGE(10)
+EMULATED_EDGE(12)
+EMULATED_EDGE(14)
 
 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);
diff --git a/libavcodec/dsputil_template.c b/libavcodec/dsputil_template.c
index 85d4fec7dc..72d8d4ae59 100644
--- a/libavcodec/dsputil_template.c
+++ b/libavcodec/dsputil_template.c
@@ -829,7 +829,7 @@ static av_unused void FUNC(OPNAME ## h264_qpel2_v_lowpass)(uint8_t *p_dst, uint8
     }\
 }\
 \
-static av_unused void FUNC(OPNAME ## h264_qpel2_hv_lowpass)(uint8_t *p_dst, int16_t *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
+static av_unused void FUNC(OPNAME ## h264_qpel2_hv_lowpass)(uint8_t *p_dst, pixeltmp *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
     const int h=2;\
     const int w=2;\
     const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
@@ -910,7 +910,7 @@ static void FUNC(OPNAME ## h264_qpel4_v_lowpass)(uint8_t *p_dst, uint8_t *p_src,
     }\
 }\
 \
-static void FUNC(OPNAME ## h264_qpel4_hv_lowpass)(uint8_t *p_dst, int16_t *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
+static void FUNC(OPNAME ## h264_qpel4_hv_lowpass)(uint8_t *p_dst, pixeltmp *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
     const int h=4;\
     const int w=4;\
     const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
@@ -1010,7 +1010,7 @@ static void FUNC(OPNAME ## h264_qpel8_v_lowpass)(uint8_t *p_dst, uint8_t *p_src,
     }\
 }\
 \
-static void FUNC(OPNAME ## h264_qpel8_hv_lowpass)(uint8_t *p_dst, int16_t *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
+static void FUNC(OPNAME ## h264_qpel8_hv_lowpass)(uint8_t *p_dst, pixeltmp *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
     const int h=8;\
     const int w=8;\
     const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
@@ -1081,7 +1081,7 @@ static void FUNC(OPNAME ## h264_qpel16_h_lowpass)(uint8_t *dst, uint8_t *src, in
     FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\
 }\
 \
-static void FUNC(OPNAME ## h264_qpel16_hv_lowpass)(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride){\
+static void FUNC(OPNAME ## h264_qpel16_hv_lowpass)(uint8_t *dst, pixeltmp *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride){\
     FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst                , tmp  , src                , dstStride, tmpStride, srcStride);\
     FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst+8*sizeof(pixel), tmp+8, src+8*sizeof(pixel), dstStride, tmpStride, srcStride);\
     src += 8*srcStride;\
@@ -1181,12 +1181,12 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc33)(uint8_t *dst, uint8_t *s
 }\
 \
 static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc22)(uint8_t *dst, uint8_t *src, int stride){\
-    int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
+    pixeltmp tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
     FUNC(OPNAME ## h264_qpel ## SIZE ## _hv_lowpass)(dst, tmp, src, stride, SIZE*sizeof(pixel), stride);\
 }\
 \
 static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc21)(uint8_t *dst, uint8_t *src, int stride){\
-    int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
+    pixeltmp tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
     uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\
     uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\
     FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(halfH, src, SIZE*sizeof(pixel), stride);\
@@ -1195,7 +1195,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc21)(uint8_t *dst, uint8_t *s
 }\
 \
 static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc23)(uint8_t *dst, uint8_t *src, int stride){\
-    int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
+    pixeltmp tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
     uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\
     uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\
     FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(halfH, src + stride, SIZE*sizeof(pixel), stride);\
@@ -1206,7 +1206,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc23)(uint8_t *dst, uint8_t *s
 static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc12)(uint8_t *dst, uint8_t *src, int stride){\
     uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
     uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
+    pixeltmp tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
     uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\
     uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\
     FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel),  stride, SIZE + 5);\
@@ -1218,7 +1218,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc12)(uint8_t *dst, uint8_t *s
 static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc32)(uint8_t *dst, uint8_t *src, int stride){\
     uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
     uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
+    pixeltmp tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
     uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\
     uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\
     FUNC(copy_block ## SIZE )(full, src - stride*2 + sizeof(pixel), SIZE*sizeof(pixel),  stride, SIZE + 5);\
@@ -1263,6 +1263,16 @@ H264_MC(avg_, 16)
 #   define avg_h264_qpel8_mc00_10_c  ff_avg_pixels8x8_10_c
 #   define put_h264_qpel16_mc00_10_c ff_put_pixels16x16_10_c
 #   define avg_h264_qpel16_mc00_10_c ff_avg_pixels16x16_10_c
+#elif BIT_DEPTH == 12
+#   define put_h264_qpel8_mc00_12_c  ff_put_pixels8x8_12_c
+#   define avg_h264_qpel8_mc00_12_c  ff_avg_pixels8x8_12_c
+#   define put_h264_qpel16_mc00_12_c ff_put_pixels16x16_12_c
+#   define avg_h264_qpel16_mc00_12_c ff_avg_pixels16x16_12_c
+#elif BIT_DEPTH == 14
+#   define put_h264_qpel8_mc00_14_c  ff_put_pixels8x8_14_c
+#   define avg_h264_qpel8_mc00_14_c  ff_avg_pixels8x8_14_c
+#   define put_h264_qpel16_mc00_14_c ff_put_pixels16x16_14_c
+#   define avg_h264_qpel16_mc00_14_c ff_avg_pixels16x16_14_c
 #endif
 
 void FUNCC(ff_put_pixels8x8)(uint8_t *dst, uint8_t *src, int stride) {