Browse Source

Merge remote-tracking branch 'qatar/master'

* qatar/master: (25 commits)
  rv40dsp x86: MMX/MMX2/3DNow/SSE2/SSSE3 implementations of MC
  ape: Use unsigned integer maths
  arm: dsputil: fix overreads in put/avg_pixels functions
  h264: K&R formatting cosmetics for header files (part II/II)
  h264: K&R formatting cosmetics for header files (part I/II)
  rtmp: Implement check bandwidth notification.
  rtmp: Support 'rtmp_swfurl', an option which specifies the URL of the SWF player.
  rtmp: Support 'rtmp_flashver', an option which overrides the version of the Flash plugin.
  rtmp: Support 'rtmp_tcurl', an option which overrides the URL of the target stream.
  cmdutils: Add fallback case to switch in check_stream_specifier().
  sctp: be consistent with socket option level
  configure: Add _XOPEN_SOURCE=600 to Solaris preprocessor flags.
  vcr1enc: drop pointless empty encode_init() wrapper function
  vcr1: drop pointless write-only AVCodecContext member from VCR1Context
  vcr1: group encoder code together to save #ifdefs
  vcr1: cosmetics: K&R prettyprinting, typos, parentheses, dead code, comments
  mov: make one comment slightly more specific
  lavr: replace the SSE version of ff_conv_fltp_to_flt_6ch() with SSE4 and AVX
  lavfi: move audio-related functions to a separate file.
  lavfi: remove some audio-related function from public API.
  ...

Conflicts:
	cmdutils.c
	libavcodec/h264.h
	libavcodec/h264_mvpred.h
	libavcodec/vcr1.c
	libavfilter/avfilter.c
	libavfilter/avfilter.h
	libavfilter/defaults.c
	libavfilter/internal.h

Merged-by: Michael Niedermayer <michaelni@gmx.at>
Michael Niedermayer 13 years ago
parent
commit
015903294c
10 changed files with 1262 additions and 1067 deletions
  1. 2 1
      cmdutils.c
  2. 1 1
      configure
  3. 2 2
      libavcodec/apedec.c
  4. 92 0
      libavcodec/arm/dsputil_neon.S
  5. 260 236
      libavcodec/h264.h
  6. 507 462
      libavcodec/h264_mvpred.h
  7. 186 183
      libavcodec/h264data.h
  8. 65 31
      libavcodec/h264dsp.h
  9. 45 36
      libavcodec/h264pred.h
  10. 102 115
      libavcodec/vcr1.c

+ 2 - 1
cmdutils.c

@@ -38,6 +38,7 @@
 #if CONFIG_POSTPROC
 #include "libpostproc/postprocess.h"
 #endif
+#include "libavutil/avassert.h"
 #include "libavutil/avstring.h"
 #include "libavutil/mathematics.h"
 #include "libavutil/parseutils.h"
@@ -1075,7 +1076,7 @@ int check_stream_specifier(AVFormatContext *s, AVStream *st, const char *spec)
         case 's': type = AVMEDIA_TYPE_SUBTITLE;   break;
         case 'd': type = AVMEDIA_TYPE_DATA;       break;
         case 't': type = AVMEDIA_TYPE_ATTACHMENT; break;
-        default: abort(); // never reached, silence warning
+        default:  av_assert0(0);
         }
         if (type != st->codec->codec_type)
             return 0;

+ 1 - 1
configure

@@ -2662,7 +2662,7 @@ case $target_os in
         SHFLAGS='-shared -Wl,-h,$$(@F)'
         enabled x86 && SHFLAGS="-mimpure-text $SHFLAGS"
         network_extralibs="-lsocket -lnsl"
-        add_cppflags -D__EXTENSIONS__
+        add_cppflags -D__EXTENSIONS__ -D_XOPEN_SOURCE=600
         # When using suncc to build, the Solaris linker will mark
         # an executable with each instruction set encountered by
         # the Solaris assembler.  As our libraries contain their own

+ 2 - 2
libavcodec/apedec.c

@@ -393,7 +393,7 @@ static inline int range_get_symbol(APEContext *ctx,
 }
 /** @} */ // group rangecoder
 
-static inline void update_rice(APERice *rice, int x)
+static inline void update_rice(APERice *rice, unsigned int x)
 {
     int lim = rice->k ? (1 << (rice->k + 4)) : 0;
     rice->ksum += ((x + 1) / 2) - ((rice->ksum + 16) >> 5);
@@ -406,7 +406,7 @@ static inline void update_rice(APERice *rice, int x)
 
 static inline int ape_decode_value(APEContext *ctx, APERice *rice)
 {
-    int x, overflow;
+    unsigned int x, overflow;
 
     if (ctx->fileversion < 3990) {
         int tmpk;

+ 92 - 0
libavcodec/arm/dsputil_neon.S

@@ -95,6 +95,7 @@ endfunc
 .endm
 
 .macro  pixels16_y2     rnd=1, avg=0
+        sub             r3,  r3,  #2
         vld1.64         {q0},     [r1], r2
         vld1.64         {q1},     [r1], r2
 1:      subs            r3,  r3,  #2
@@ -114,10 +115,25 @@ endfunc
         vst1.64         {q2},     [r0,:128], r2
         vst1.64         {q3},     [r0,:128], r2
         bne             1b
+
+        avg             q2,  q0,  q1
+        vld1.64         {q0},     [r1], r2
+        avg             q3,  q0,  q1
+  .if \avg
+        vld1.8          {q8},     [r0,:128], r2
+        vld1.8          {q9},     [r0,:128]
+        vrhadd.u8       q2,  q2,  q8
+        vrhadd.u8       q3,  q3,  q9
+        sub             r0,  r0,  r2
+  .endif
+        vst1.64         {q2},     [r0,:128], r2
+        vst1.64         {q3},     [r0,:128], r2
+
         bx              lr
 .endm
 
 .macro  pixels16_xy2    rnd=1, avg=0
+        sub             r3,  r3,  #2
         vld1.64         {d0-d2},  [r1], r2
         vld1.64         {d4-d6},  [r1], r2
   .ifeq \rnd
@@ -173,6 +189,42 @@ endfunc
         vaddl.u8        q11, d3,  d5
         vst1.64         {q15},    [r0,:128], r2
         bgt             1b
+
+        vld1.64         {d0-d2},  [r1], r2
+        vadd.u16        q12, q8,  q9
+  .ifeq \rnd
+        vadd.u16        q12, q12, q13
+  .endif
+        vext.8          q15, q0,  q1,  #1
+        vadd.u16        q1 , q10, q11
+        shrn            d28, q12, #2
+  .ifeq \rnd
+        vadd.u16        q1,  q1,  q13
+  .endif
+        shrn            d29, q1,  #2
+  .if \avg
+        vld1.8          {q8},     [r0,:128]
+        vrhadd.u8       q14, q14, q8
+  .endif
+        vaddl.u8        q8,  d0,  d30
+        vaddl.u8        q10, d1,  d31
+        vst1.64         {q14},    [r0,:128], r2
+        vadd.u16        q12, q8,  q9
+  .ifeq \rnd
+        vadd.u16        q12, q12, q13
+  .endif
+        vadd.u16        q0,  q10, q11
+        shrn            d30, q12, #2
+  .ifeq \rnd
+        vadd.u16        q0,  q0,  q13
+  .endif
+        shrn            d31, q0,  #2
+  .if \avg
+        vld1.8          {q9},     [r0,:128]
+        vrhadd.u8       q15, q15, q9
+  .endif
+        vst1.64         {q15},    [r0,:128], r2
+
         bx              lr
 .endm
 
@@ -228,6 +280,7 @@ endfunc
 .endm
 
 .macro  pixels8_y2      rnd=1, avg=0
+        sub             r3,  r3,  #2
         vld1.64         {d0},     [r1], r2
         vld1.64         {d1},     [r1], r2
 1:      subs            r3,  r3,  #2
@@ -246,10 +299,24 @@ endfunc
         vst1.64         {d4},     [r0,:64], r2
         vst1.64         {d5},     [r0,:64], r2
         bne             1b
+
+        avg             d4,  d0,  d1
+        vld1.64         {d0},     [r1], r2
+        avg             d5,  d0,  d1
+  .if \avg
+        vld1.8          {d2},     [r0,:64], r2
+        vld1.8          {d3},     [r0,:64]
+        vrhadd.u8       q2,  q2,  q1
+        sub             r0,  r0,  r2
+  .endif
+        vst1.64         {d4},     [r0,:64], r2
+        vst1.64         {d5},     [r0,:64], r2
+
         bx              lr
 .endm
 
 .macro  pixels8_xy2     rnd=1, avg=0
+        sub             r3,  r3,  #2
         vld1.64         {q0},     [r1], r2
         vld1.64         {q1},     [r1], r2
   .ifeq \rnd
@@ -291,6 +358,31 @@ endfunc
         vaddl.u8        q9,  d2,  d6
         vst1.64         {d7},     [r0,:64], r2
         bgt             1b
+
+        vld1.64         {q0},     [r1], r2
+        vadd.u16        q10, q8,  q9
+        vext.8          d4,  d0,  d1,  #1
+  .ifeq \rnd
+        vadd.u16        q10, q10, q11
+  .endif
+        vaddl.u8        q8,  d0,  d4
+        shrn            d5,  q10, #2
+        vadd.u16        q10, q8,  q9
+  .if \avg
+        vld1.8          {d7},     [r0,:64]
+        vrhadd.u8       d5,  d5,  d7
+  .endif
+  .ifeq \rnd
+        vadd.u16        q10, q10, q11
+  .endif
+        vst1.64         {d5},     [r0,:64], r2
+        shrn            d7,  q10, #2
+  .if \avg
+        vld1.8          {d5},     [r0,:64]
+        vrhadd.u8       d7,  d7,  d5
+  .endif
+        vst1.64         {d7},     [r0,:64], r2
+
         bx              lr
 .endm
 

+ 260 - 236
libavcodec/h264.h

@@ -37,14 +37,14 @@
 #include "rectangle.h"
 
 #define interlaced_dct interlaced_dct_is_a_bad_name
-#define mb_intra mb_intra_is_not_initialized_see_mb_type
+#define mb_intra       mb_intra_is_not_initialized_see_mb_type
 
-#define MAX_SPS_COUNT 32
-#define MAX_PPS_COUNT 256
+#define MAX_SPS_COUNT          32
+#define MAX_PPS_COUNT         256
 
-#define MAX_MMCO_COUNT 66
+#define MAX_MMCO_COUNT         66
 
-#define MAX_DELAYED_PIC_COUNT 16
+#define MAX_DELAYED_PIC_COUNT  16
 
 #define MAX_MBPAIR_SIZE (256*1024) // a tighter bound could be calculated if someone cares about a few bytes
 
@@ -61,25 +61,25 @@
 #define MAX_SLICES 16
 
 #ifdef ALLOW_INTERLACE
-#define MB_MBAFF h->mb_mbaff
-#define MB_FIELD h->mb_field_decoding_flag
+#define MB_MBAFF    h->mb_mbaff
+#define MB_FIELD    h->mb_field_decoding_flag
 #define FRAME_MBAFF h->mb_aff_frame
 #define FIELD_PICTURE (s->picture_structure != PICT_FRAME)
 #define LEFT_MBS 2
-#define LTOP 0
-#define LBOT 1
-#define LEFT(i) (i)
+#define LTOP     0
+#define LBOT     1
+#define LEFT(i)  (i)
 #else
-#define MB_MBAFF 0
-#define MB_FIELD 0
-#define FRAME_MBAFF 0
+#define MB_MBAFF      0
+#define MB_FIELD      0
+#define FRAME_MBAFF   0
 #define FIELD_PICTURE 0
 #undef  IS_INTERLACED
 #define IS_INTERLACED(mb_type) 0
 #define LEFT_MBS 1
-#define LTOP 0
-#define LBOT 0
-#define LEFT(i) 0
+#define LTOP     0
+#define LBOT     0
+#define LEFT(i)  0
 #endif
 #define FIELD_OR_MBAFF_PICTURE (FRAME_MBAFF || FIELD_PICTURE)
 
@@ -91,9 +91,9 @@
 #define CHROMA422 (h->sps.chroma_format_idc == 2)
 #define CHROMA444 (h->sps.chroma_format_idc == 3)
 
-#define EXTENDED_SAR          255
+#define EXTENDED_SAR       255
 
-#define MB_TYPE_REF0       MB_TYPE_ACPRED //dirty but it fits in 16 bit
+#define MB_TYPE_REF0       MB_TYPE_ACPRED // dirty but it fits in 16 bit
 #define MB_TYPE_8x8DCT     0x01000000
 #define IS_REF0(a)         ((a) & MB_TYPE_REF0)
 #define IS_8x8DCT(a)       ((a) & MB_TYPE_8x8DCT)
@@ -108,7 +108,7 @@
 
 /* NAL unit types */
 enum {
-    NAL_SLICE=1,
+    NAL_SLICE = 1,
     NAL_DPA,
     NAL_DPB,
     NAL_DPC,
@@ -121,17 +121,17 @@ enum {
     NAL_END_STREAM,
     NAL_FILLER_DATA,
     NAL_SPS_EXT,
-    NAL_AUXILIARY_SLICE=19
+    NAL_AUXILIARY_SLICE = 19
 };
 
 /**
  * SEI message types
  */
 typedef enum {
-    SEI_BUFFERING_PERIOD             =  0, ///< buffering period (H.264, D.1.1)
-    SEI_TYPE_PIC_TIMING              =  1, ///< picture timing
-    SEI_TYPE_USER_DATA_UNREGISTERED  =  5, ///< unregistered user data
-    SEI_TYPE_RECOVERY_POINT          =  6  ///< recovery point (frame # to decoder sync)
+    SEI_BUFFERING_PERIOD            = 0,   ///< buffering period (H.264, D.1.1)
+    SEI_TYPE_PIC_TIMING             = 1,   ///< picture timing
+    SEI_TYPE_USER_DATA_UNREGISTERED = 5,   ///< unregistered user data
+    SEI_TYPE_RECOVERY_POINT         = 6    ///< recovery point (frame # to decoder sync)
 } SEI_Type;
 
 /**
@@ -152,8 +152,7 @@ typedef enum {
 /**
  * Sequence parameter set
  */
-typedef struct SPS{
-
+typedef struct SPS {
     int profile_idc;
     int level_idc;
     int chroma_format_idc;
@@ -170,9 +169,9 @@ typedef struct SPS{
     int mb_width;                      ///< pic_width_in_mbs_minus1 + 1
     int mb_height;                     ///< pic_height_in_map_units_minus1 + 1
     int frame_mbs_only_flag;
-    int mb_aff;                        ///<mb_adaptive_frame_field_flag
+    int mb_aff;                        ///< mb_adaptive_frame_field_flag
     int direct_8x8_inference_flag;
-    int crop;                   ///< frame_cropping_flag
+    int crop;                          ///< frame_cropping_flag
     unsigned int crop_left;            ///< frame_cropping_rect_left_offset
     unsigned int crop_right;           ///< frame_cropping_rect_right_offset
     unsigned int crop_top;             ///< frame_cropping_rect_top_offset
@@ -189,7 +188,7 @@ typedef struct SPS{
     uint32_t num_units_in_tick;
     uint32_t time_scale;
     int fixed_frame_rate_flag;
-    short offset_for_ref_frame[256]; //FIXME dyn aloc?
+    short offset_for_ref_frame[256]; // FIXME dyn aloc?
     int bitstream_restriction_flag;
     int num_reorder_frames;
     int scaling_matrix_present;
@@ -199,20 +198,20 @@ typedef struct SPS{
     int vcl_hrd_parameters_present_flag;
     int pic_struct_present_flag;
     int time_offset_length;
-    int cpb_cnt;                       ///< See H.264 E.1.2
-    int initial_cpb_removal_delay_length; ///< initial_cpb_removal_delay_length_minus1 +1
-    int cpb_removal_delay_length;      ///< cpb_removal_delay_length_minus1 + 1
-    int dpb_output_delay_length;       ///< dpb_output_delay_length_minus1 + 1
-    int bit_depth_luma;                ///< bit_depth_luma_minus8 + 8
-    int bit_depth_chroma;              ///< bit_depth_chroma_minus8 + 8
-    int residual_color_transform_flag; ///< residual_colour_transform_flag
-    int constraint_set_flags;          ///< constraint_set[0-3]_flag
-}SPS;
+    int cpb_cnt;                          ///< See H.264 E.1.2
+    int initial_cpb_removal_delay_length; ///< initial_cpb_removal_delay_length_minus1 + 1
+    int cpb_removal_delay_length;         ///< cpb_removal_delay_length_minus1 + 1
+    int dpb_output_delay_length;          ///< dpb_output_delay_length_minus1 + 1
+    int bit_depth_luma;                   ///< bit_depth_luma_minus8 + 8
+    int bit_depth_chroma;                 ///< bit_depth_chroma_minus8 + 8
+    int residual_color_transform_flag;    ///< residual_colour_transform_flag
+    int constraint_set_flags;             ///< constraint_set[0-3]_flag
+} SPS;
 
 /**
  * Picture parameter set
  */
-typedef struct PPS{
+typedef struct PPS {
     unsigned int sps_id;
     int cabac;                  ///< entropy_coding_mode_flag
     int pic_order_present;      ///< pic_order_present_flag
@@ -225,20 +224,20 @@ typedef struct PPS{
     int init_qs;                ///< pic_init_qs_minus26 + 26
     int chroma_qp_index_offset[2];
     int deblocking_filter_parameters_present; ///< deblocking_filter_parameters_present_flag
-    int constrained_intra_pred; ///< constrained_intra_pred_flag
-    int redundant_pic_cnt_present; ///< redundant_pic_cnt_present_flag
-    int transform_8x8_mode;     ///< transform_8x8_mode_flag
+    int constrained_intra_pred;     ///< constrained_intra_pred_flag
+    int redundant_pic_cnt_present;  ///< redundant_pic_cnt_present_flag
+    int transform_8x8_mode;         ///< transform_8x8_mode_flag
     uint8_t scaling_matrix4[6][16];
     uint8_t scaling_matrix8[6][64];
     uint8_t chroma_qp_table[2][QP_MAX_NUM+1];  ///< pre-scaled (with chroma_qp_index_offset) version of qp_table
     int chroma_qp_diff;
-}PPS;
+} PPS;
 
 /**
  * Memory management control operation opcode.
  */
-typedef enum MMCOOpcode{
-    MMCO_END=0,
+typedef enum MMCOOpcode {
+    MMCO_END = 0,
     MMCO_SHORT2UNUSED,
     MMCO_LONG2UNUSED,
     MMCO_SHORT2LONG,
@@ -250,7 +249,7 @@ typedef enum MMCOOpcode{
 /**
  * Memory management control operation.
  */
-typedef struct MMCO{
+typedef struct MMCO {
     MMCOOpcode opcode;
     int short_pic_num;  ///< pic_num without wrapping (pic_num & max_pic_num)
     int long_arg;       ///< index, pic_num, or num long refs depending on opcode
@@ -259,18 +258,18 @@ typedef struct MMCO{
 /**
  * H264Context
  */
-typedef struct H264Context{
+typedef struct H264Context {
     MpegEncContext s;
     H264DSPContext h264dsp;
     int pixel_shift;    ///< 0 for 8-bit H264, 1 for high-bit-depth H264
-    int chroma_qp[2]; //QPc
+    int chroma_qp[2];   // QPc
 
     int qp_thresh;      ///< QP threshold to skip loopfilter
 
     int prev_mb_skipped;
     int next_mb_skipped;
 
-    //prediction stuff
+    // prediction stuff
     int chroma_pred_mode;
     int intra16x16_pred_mode;
 
@@ -284,32 +283,32 @@ typedef struct H264Context{
     int topright_type;
     int left_type[LEFT_MBS];
 
-    const uint8_t * left_block;
+    const uint8_t *left_block;
     int topleft_partition;
 
-    int8_t intra4x4_pred_mode_cache[5*8];
-    int8_t (*intra4x4_pred_mode);
+    int8_t intra4x4_pred_mode_cache[5 * 8];
+    int8_t(*intra4x4_pred_mode);
     H264PredContext hpc;
     unsigned int topleft_samples_available;
     unsigned int top_samples_available;
     unsigned int topright_samples_available;
     unsigned int left_samples_available;
-    uint8_t (*top_borders[2])[(16*3)*2];
+    uint8_t (*top_borders[2])[(16 * 3) * 2];
 
     /**
      * non zero coeff count cache.
      * is 64 if not available.
      */
-    DECLARE_ALIGNED(8, uint8_t, non_zero_count_cache)[15*8];
+    DECLARE_ALIGNED(8, uint8_t, non_zero_count_cache)[15 * 8];
 
     uint8_t (*non_zero_count)[48];
 
     /**
      * Motion vector cache.
      */
-    DECLARE_ALIGNED(16, int16_t, mv_cache)[2][5*8][2];
-    DECLARE_ALIGNED(8, int8_t, ref_cache)[2][5*8];
-#define LIST_NOT_USED -1 //FIXME rename?
+    DECLARE_ALIGNED(16, int16_t, mv_cache)[2][5 * 8][2];
+    DECLARE_ALIGNED(8, int8_t, ref_cache)[2][5 * 8];
+#define LIST_NOT_USED -1 // FIXME rename?
 #define PART_NOT_AVAILABLE -2
 
     /**
@@ -321,13 +320,13 @@ typedef struct H264Context{
      * block_offset[ 0..23] for frame macroblocks
      * block_offset[24..47] for field macroblocks
      */
-    int block_offset[2*(16*3)];
+    int block_offset[2 * (16 * 3)];
 
-    uint32_t *mb2b_xy; //FIXME are these 4 a good idea?
+    uint32_t *mb2b_xy;  // FIXME are these 4 a good idea?
     uint32_t *mb2br_xy;
-    int b_stride; //FIXME use s->b4_stride
+    int b_stride;       // FIXME use s->b4_stride
 
-    int mb_linesize;   ///< may be equal to s->linesize or s->linesize*2, for mbaff
+    int mb_linesize;    ///< may be equal to s->linesize or s->linesize * 2, for mbaff
     int mb_uvlinesize;
 
     int emu_edge_width;
@@ -338,32 +337,32 @@ typedef struct H264Context{
     /**
      * current pps
      */
-    PPS pps; //FIXME move to Picture perhaps? (->no) do we need that?
+    PPS pps; // FIXME move to Picture perhaps? (->no) do we need that?
 
-    uint32_t dequant4_buffer[6][QP_MAX_NUM+1][16]; //FIXME should these be moved down?
-    uint32_t dequant8_buffer[6][QP_MAX_NUM+1][64];
-    uint32_t (*dequant4_coeff[6])[16];
-    uint32_t (*dequant8_coeff[6])[64];
+    uint32_t dequant4_buffer[6][QP_MAX_NUM + 1][16]; // FIXME should these be moved down?
+    uint32_t dequant8_buffer[6][QP_MAX_NUM + 1][64];
+    uint32_t(*dequant4_coeff[6])[16];
+    uint32_t(*dequant8_coeff[6])[64];
 
     int slice_num;
-    uint16_t *slice_table;     ///< slice_table_base + 2*mb_stride + 1
+    uint16_t *slice_table;      ///< slice_table_base + 2*mb_stride + 1
     int slice_type;
-    int slice_type_nos;        ///< S free slice type (SI/SP are remapped to I/P)
+    int slice_type_nos;         ///< S free slice type (SI/SP are remapped to I/P)
     int slice_type_fixed;
 
-    //interlacing specific flags
+    // interlacing specific flags
     int mb_aff_frame;
     int mb_field_decoding_flag;
-    int mb_mbaff;              ///< mb_aff_frame && mb_field_decoding_flag
+    int mb_mbaff;               ///< mb_aff_frame && mb_field_decoding_flag
 
     DECLARE_ALIGNED(8, uint16_t, sub_mb_type)[4];
 
-    //Weighted pred stuff
+    // Weighted pred stuff
     int use_weight;
     int use_weight_chroma;
     int luma_log2_weight_denom;
     int chroma_log2_weight_denom;
-    //The following 2 can be changed to int8_t but that causes 10cpu cycles speedloss
+    // The following 2 can be changed to int8_t but that causes 10cpu cycles speedloss
     int luma_weight[48][2][2];
     int chroma_weight[48][2][2][2];
     int implicit_weight[48][48][2];
@@ -373,48 +372,48 @@ typedef struct H264Context{
     int col_fieldoff;
     int dist_scale_factor[16];
     int dist_scale_factor_field[2][32];
-    int map_col_to_list0[2][16+32];
-    int map_col_to_list0_field[2][2][16+32];
+    int map_col_to_list0[2][16 + 32];
+    int map_col_to_list0_field[2][2][16 + 32];
 
     /**
      * num_ref_idx_l0/1_active_minus1 + 1
      */
-    uint8_t *list_counts;            ///< Array of list_count per MB specifying the slice type
-    unsigned int ref_count[2];   ///< counts frames or fields, depending on current mb mode
+    unsigned int ref_count[2];          ///< counts frames or fields, depending on current mb mode
     unsigned int list_count;
-    Picture ref_list[2][48];         /**< 0..15: frame refs, 16..47: mbaff field refs.
-                                          Reordered version of default_ref_list
-                                          according to picture reordering in slice header */
-    int ref2frm[MAX_SLICES][2][64];  ///< reference to frame number lists, used in the loop filter, the first 2 are for -2,-1
+    uint8_t *list_counts;               ///< Array of list_count per MB specifying the slice type
+    Picture ref_list[2][48];            /**< 0..15: frame refs, 16..47: mbaff field refs.
+                                         *   Reordered version of default_ref_list
+                                         *   according to picture reordering in slice header */
+    int ref2frm[MAX_SLICES][2][64];     ///< reference to frame number lists, used in the loop filter, the first 2 are for -2,-1
 
-    //data partitioning
+    // data partitioning
     GetBitContext intra_gb;
     GetBitContext inter_gb;
     GetBitContext *intra_gb_ptr;
     GetBitContext *inter_gb_ptr;
 
-    DECLARE_ALIGNED(16, DCTELEM, mb)[16*48*2]; ///< as a dct coeffecient is int32_t in high depth, we need to reserve twice the space.
-    DECLARE_ALIGNED(16, DCTELEM, mb_luma_dc)[3][16*2];
-    DCTELEM mb_padding[256*2];        ///< as mb is addressed by scantable[i] and scantable is uint8_t we can either check that i is not too large or ensure that there is some unused stuff after mb
+    DECLARE_ALIGNED(16, DCTELEM, mb)[16 * 48 * 2]; ///< as a dct coeffecient is int32_t in high depth, we need to reserve twice the space.
+    DECLARE_ALIGNED(16, DCTELEM, mb_luma_dc)[3][16 * 2];
+    DCTELEM mb_padding[256 * 2];        ///< as mb is addressed by scantable[i] and scantable is uint8_t we can either check that i is not too large or ensure that there is some unused stuff after mb
 
     /**
      * Cabac
      */
     CABACContext cabac;
-    uint8_t      cabac_state[1024];
+    uint8_t cabac_state[1024];
 
-    /* 0x100 -> non null luma_dc, 0x80/0x40 -> non null chroma_dc (cb/cr), 0x?0 -> chroma_cbp(0,1,2), 0x0? luma_cbp */
-    uint16_t     *cbp_table;
+    /* 0x100 -> non null luma_dc, 0x80/0x40 -> non null chroma_dc (cb/cr), 0x?0 -> chroma_cbp(0, 1, 2), 0x0? luma_cbp */
+    uint16_t *cbp_table;
     int cbp;
     int top_cbp;
     int left_cbp;
     /* chroma_pred_mode for i4x4 or i16x16, else 0 */
-    uint8_t     *chroma_pred_mode_table;
-    int         last_qscale_diff;
-    uint8_t     (*mvd_table[2])[2];
-    DECLARE_ALIGNED(16, uint8_t, mvd_cache)[2][5*8][2];
-    uint8_t     *direct_table;
-    uint8_t     direct_cache[5*8];
+    uint8_t *chroma_pred_mode_table;
+    int last_qscale_diff;
+    uint8_t (*mvd_table[2])[2];
+    DECLARE_ALIGNED(16, uint8_t, mvd_cache)[2][5 * 8][2];
+    uint8_t *direct_table;
+    uint8_t direct_cache[5 * 8];
 
     uint8_t zigzag_scan[16];
     uint8_t zigzag_scan8x8[64];
@@ -435,13 +434,13 @@ typedef struct H264Context{
 
     int is_complex;
 
-    //deblock
-    int deblocking_filter;         ///< disable_deblocking_filter_idc with 1<->0
+    // deblock
+    int deblocking_filter;          ///< disable_deblocking_filter_idc with 1 <-> 0
     int slice_alpha_c0_offset;
     int slice_beta_offset;
 
-//=============================================================
-    //Things below are not used in the MB or more inner code
+    // =============================================================
+    // Things below are not used in the MB or more inner code
 
     int nal_ref_idc;
     int nal_unit_type;
@@ -451,37 +450,36 @@ typedef struct H264Context{
     /**
      * Used to parse AVC variant of h264
      */
-    int is_avc; ///< this flag is != 0 if codec is avc1
-    int nal_length_size; ///< Number of bytes used for nal length (1, 2 or 4)
-    int got_first; ///< this flag is != 0 if we've parsed a frame
+    int is_avc;           ///< this flag is != 0 if codec is avc1
+    int nal_length_size;  ///< Number of bytes used for nal length (1, 2 or 4)
+    int got_first;        ///< this flag is != 0 if we've parsed a frame
 
     SPS *sps_buffers[MAX_SPS_COUNT];
     PPS *pps_buffers[MAX_PPS_COUNT];
 
-    int dequant_coeff_pps;     ///< reinit tables when pps changes
+    int dequant_coeff_pps;      ///< reinit tables when pps changes
 
     uint16_t *slice_table_base;
 
-
-    //POC stuff
+    // POC stuff
     int poc_lsb;
     int poc_msb;
     int delta_poc_bottom;
     int delta_poc[2];
     int frame_num;
-    int prev_poc_msb;             ///< poc_msb of the last reference pic for POC type 0
-    int prev_poc_lsb;             ///< poc_lsb of the last reference pic for POC type 0
-    int frame_num_offset;         ///< for POC type 2
-    int prev_frame_num_offset;    ///< for POC type 2
-    int prev_frame_num;           ///< frame_num of the last pic for POC type 1/2
+    int prev_poc_msb;           ///< poc_msb of the last reference pic for POC type 0
+    int prev_poc_lsb;           ///< poc_lsb of the last reference pic for POC type 0
+    int frame_num_offset;       ///< for POC type 2
+    int prev_frame_num_offset;  ///< for POC type 2
+    int prev_frame_num;         ///< frame_num of the last pic for POC type 1/2
 
     /**
-     * frame_num for frames or 2*frame_num+1 for field pics.
+     * frame_num for frames or 2 * frame_num + 1 for field pics.
      */
     int curr_pic_num;
 
     /**
-     * max_frame_num or 2*max_frame_num for field pics.
+     * max_frame_num or 2 * max_frame_num for field pics.
      */
     int max_pic_num;
 
@@ -490,7 +488,7 @@ typedef struct H264Context{
     Picture *short_ref[32];
     Picture *long_ref[32];
     Picture default_ref_list[2][32]; ///< base reference list for all slices of a coded picture
-    Picture *delayed_pic[MAX_DELAYED_PIC_COUNT+2]; //FIXME size?
+    Picture *delayed_pic[MAX_DELAYED_PIC_COUNT + 2]; // FIXME size?
     int last_pocs[MAX_DELAYED_PIC_COUNT];
     Picture *next_output_pic;
     int outputed_poc;
@@ -503,10 +501,10 @@ typedef struct H264Context{
     int mmco_index;
     int mmco_reset;
 
-    int long_ref_count;  ///< number of actual long term references
-    int short_ref_count; ///< number of actual short term references
+    int long_ref_count;     ///< number of actual long term references
+    int short_ref_count;    ///< number of actual short term references
 
-    int          cabac_init_idc;
+    int cabac_init_idc;
 
     /**
      * @name Members for slice based multithreading
@@ -582,12 +580,12 @@ typedef struct H264Context{
      */
     int recovery_frame;
 
-    int luma_weight_flag[2];   ///< 7.4.3.2 luma_weight_lX_flag
-    int chroma_weight_flag[2]; ///< 7.4.3.2 chroma_weight_lX_flag
+    int luma_weight_flag[2];    ///< 7.4.3.2 luma_weight_lX_flag
+    int chroma_weight_flag[2];  ///< 7.4.3.2 chroma_weight_lX_flag
 
     // Timestamp stuff
-    int sei_buffering_period_present;  ///< Buffering period SEI flag
-    int initial_cpb_removal_delay[32]; ///< Initial timestamps for CPBs
+    int sei_buffering_period_present;   ///< Buffering period SEI flag
+    int initial_cpb_removal_delay[32];  ///< Initial timestamps for CPBs
 
     int cur_chroma_format_idc;
 
@@ -598,10 +596,9 @@ typedef struct H264Context{
     uint8_t parse_history[4];
     int parse_history_count;
     int parse_last_mb;
-}H264Context;
-
+} H264Context;
 
-extern const uint8_t ff_h264_chroma_qp[5][QP_MAX_NUM+1]; ///< One chroma qp table for each possible bit depth (8-12).
+extern const uint8_t ff_h264_chroma_qp[5][QP_MAX_NUM + 1]; ///< One chroma qp table for each possible bit depth (8-12).
 extern const uint16_t ff_h264_mb_sizes[4];
 
 /**
@@ -628,13 +625,16 @@ int ff_h264_decode_picture_parameter_set(H264Context *h, int bit_length);
  * Decode a network abstraction layer unit.
  * @param consumed is the number of bytes used as input
  * @param length is the length of the array
- * @param dst_length is the number of decoded bytes FIXME here or a decode rbsp tailing?
+ * @param dst_length is the number of decoded bytes FIXME here
+ *                   or a decode rbsp tailing?
  * @return decoded bytes, might be src+1 if no escapes
  */
-const uint8_t *ff_h264_decode_nal(H264Context *h, const uint8_t *src, int *dst_length, int *consumed, int length);
+const uint8_t *ff_h264_decode_nal(H264Context *h, const uint8_t *src,
+                                  int *dst_length, int *consumed, int length);
 
 /**
- * Free any data that may have been allocated in the H264 context like SPS, PPS etc.
+ * Free any data that may have been allocated in the H264 context
+ * like SPS, PPS etc.
  */
 av_cold void ff_h264_free_context(H264Context *h);
 
@@ -667,12 +667,16 @@ int ff_h264_decode_ref_pic_marking(H264Context *h, GetBitContext *gb);
 
 void ff_generate_sliding_window_mmcos(H264Context *h);
 
-
 /**
- * Check if the top & left blocks are available if needed & change the dc mode so it only uses the available blocks.
+ * Check if the top & left blocks are available if needed & change the
+ * dc mode so it only uses the available blocks.
  */
 int ff_h264_check_intra4x4_pred_mode(H264Context *h);
 
+/**
+ * Check if the top & left blocks are available if needed & change the
+ * dc mode so it only uses the available blocks.
+ */
 int ff_h264_check_intra_pred_mode(H264Context *h, int mode, int is_chroma);
 
 void ff_h264_hl_decode_mb(H264Context *h);
@@ -683,24 +687,28 @@ av_cold void ff_h264_decode_init_vlc(void);
 
 /**
  * Decode a macroblock
- * @return 0 if OK, ER_AC_ERROR / ER_DC_ERROR / ER_MV_ERROR if an error is noticed
+ * @return 0 if OK, ER_AC_ERROR / ER_DC_ERROR / ER_MV_ERROR on error
  */
 int ff_h264_decode_mb_cavlc(H264Context *h);
 
 /**
  * Decode a CABAC coded macroblock
- * @return 0 if OK, ER_AC_ERROR / ER_DC_ERROR / ER_MV_ERROR if an error is noticed
+ * @return 0 if OK, ER_AC_ERROR / ER_DC_ERROR / ER_MV_ERROR on error
  */
 int ff_h264_decode_mb_cabac(H264Context *h);
 
 void ff_h264_init_cabac_states(H264Context *h);
 
-void ff_h264_direct_dist_scale_factor(H264Context * const h);
-void ff_h264_direct_ref_list_init(H264Context * const h);
-void ff_h264_pred_direct_motion(H264Context * const h, int *mb_type);
+void ff_h264_direct_dist_scale_factor(H264Context *const h);
+void ff_h264_direct_ref_list_init(H264Context *const h);
+void ff_h264_pred_direct_motion(H264Context *const h, int *mb_type);
 
-void ff_h264_filter_mb_fast( H264Context *h, int mb_x, int mb_y, uint8_t *img_y, uint8_t *img_cb, uint8_t *img_cr, unsigned int linesize, unsigned int uvlinesize);
-void ff_h264_filter_mb( H264Context *h, int mb_x, int mb_y, uint8_t *img_y, uint8_t *img_cb, uint8_t *img_cr, unsigned int linesize, unsigned int uvlinesize);
+void ff_h264_filter_mb_fast(H264Context *h, int mb_x, int mb_y,
+                            uint8_t *img_y, uint8_t *img_cb, uint8_t *img_cr,
+                            unsigned int linesize, unsigned int uvlinesize);
+void ff_h264_filter_mb(H264Context *h, int mb_x, int mb_y,
+                       uint8_t *img_y, uint8_t *img_cb, uint8_t *img_cr,
+                       unsigned int linesize, unsigned int uvlinesize);
 
 /**
  * Reset SEI values at the beginning of the frame.
@@ -709,16 +717,15 @@ void ff_h264_filter_mb( H264Context *h, int mb_x, int mb_y, uint8_t *img_y, uint
  */
 void ff_h264_reset_sei(H264Context *h);
 
-
 /*
-o-o o-o
- / / /
-o-o o-o
- ,---'
-o-o o-o
- / / /
-o-o o-o
-*/
+ * o-o o-o
+ *  / / /
+ * o-o o-o
+ *  ,---'
+ * o-o o-o
+ *  / / /
+ * o-o o-o
+ */
 
 /* Scan8 organization:
  *    0 1 2 3 4 5 6 7
@@ -743,156 +750,173 @@ o-o o-o
 #define LUMA_DC_BLOCK_INDEX   48
 #define CHROMA_DC_BLOCK_INDEX 49
 
-//This table must be here because scan8[constant] must be known at compiletime
-static const uint8_t scan8[16*3 + 3]={
- 4+ 1*8, 5+ 1*8, 4+ 2*8, 5+ 2*8,
- 6+ 1*8, 7+ 1*8, 6+ 2*8, 7+ 2*8,
- 4+ 3*8, 5+ 3*8, 4+ 4*8, 5+ 4*8,
- 6+ 3*8, 7+ 3*8, 6+ 4*8, 7+ 4*8,
- 4+ 6*8, 5+ 6*8, 4+ 7*8, 5+ 7*8,
- 6+ 6*8, 7+ 6*8, 6+ 7*8, 7+ 7*8,
- 4+ 8*8, 5+ 8*8, 4+ 9*8, 5+ 9*8,
- 6+ 8*8, 7+ 8*8, 6+ 9*8, 7+ 9*8,
- 4+11*8, 5+11*8, 4+12*8, 5+12*8,
- 6+11*8, 7+11*8, 6+12*8, 7+12*8,
- 4+13*8, 5+13*8, 4+14*8, 5+14*8,
- 6+13*8, 7+13*8, 6+14*8, 7+14*8,
- 0+ 0*8, 0+ 5*8, 0+10*8
+// This table must be here because scan8[constant] must be known at compiletime
+static const uint8_t scan8[16 * 3 + 3] = {
+    4 +  1 * 8, 5 +  1 * 8, 4 +  2 * 8, 5 +  2 * 8,
+    6 +  1 * 8, 7 +  1 * 8, 6 +  2 * 8, 7 +  2 * 8,
+    4 +  3 * 8, 5 +  3 * 8, 4 +  4 * 8, 5 +  4 * 8,
+    6 +  3 * 8, 7 +  3 * 8, 6 +  4 * 8, 7 +  4 * 8,
+    4 +  6 * 8, 5 +  6 * 8, 4 +  7 * 8, 5 +  7 * 8,
+    6 +  6 * 8, 7 +  6 * 8, 6 +  7 * 8, 7 +  7 * 8,
+    4 +  8 * 8, 5 +  8 * 8, 4 +  9 * 8, 5 +  9 * 8,
+    6 +  8 * 8, 7 +  8 * 8, 6 +  9 * 8, 7 +  9 * 8,
+    4 + 11 * 8, 5 + 11 * 8, 4 + 12 * 8, 5 + 12 * 8,
+    6 + 11 * 8, 7 + 11 * 8, 6 + 12 * 8, 7 + 12 * 8,
+    4 + 13 * 8, 5 + 13 * 8, 4 + 14 * 8, 5 + 14 * 8,
+    6 + 13 * 8, 7 + 13 * 8, 6 + 14 * 8, 7 + 14 * 8,
+    0 +  0 * 8, 0 +  5 * 8, 0 + 10 * 8
 };
 
-static av_always_inline uint32_t pack16to32(int a, int b){
+static av_always_inline uint32_t pack16to32(int a, int b)
+{
 #if HAVE_BIGENDIAN
-   return (b&0xFFFF) + (a<<16);
+    return (b & 0xFFFF) + (a << 16);
 #else
-   return (a&0xFFFF) + (b<<16);
+    return (a & 0xFFFF) + (b << 16);
 #endif
 }
 
-static av_always_inline uint16_t pack8to16(int a, int b){
+static av_always_inline uint16_t pack8to16(int a, int b)
+{
 #if HAVE_BIGENDIAN
-   return (b&0xFF) + (a<<8);
+    return (b & 0xFF) + (a << 8);
 #else
-   return (a&0xFF) + (b<<8);
+    return (a & 0xFF) + (b << 8);
 #endif
 }
 
 /**
  * Get the chroma qp.
  */
-static av_always_inline int get_chroma_qp(H264Context *h, int t, int qscale){
+static av_always_inline int get_chroma_qp(H264Context *h, int t, int qscale)
+{
     return h->pps.chroma_qp_table[t][qscale];
 }
 
 /**
  * Get the predicted intra4x4 prediction mode.
  */
-static av_always_inline int pred_intra_mode(H264Context *h, int n){
-    const int index8= scan8[n];
-    const int left= h->intra4x4_pred_mode_cache[index8 - 1];
-    const int top = h->intra4x4_pred_mode_cache[index8 - 8];
-    const int min= FFMIN(left, top);
+static av_always_inline int pred_intra_mode(H264Context *h, int n)
+{
+    const int index8 = scan8[n];
+    const int left   = h->intra4x4_pred_mode_cache[index8 - 1];
+    const int top    = h->intra4x4_pred_mode_cache[index8 - 8];
+    const int min    = FFMIN(left, top);
 
-    tprintf(h->s.avctx, "mode:%d %d min:%d\n", left ,top, min);
+    tprintf(h->s.avctx, "mode:%d %d min:%d\n", left, top, min);
 
-    if(min<0) return DC_PRED;
-    else      return min;
+    if (min < 0)
+        return DC_PRED;
+    else
+        return min;
 }
 
-static av_always_inline void write_back_intra_pred_mode(H264Context *h){
-    int8_t *i4x4= h->intra4x4_pred_mode + h->mb2br_xy[h->mb_xy];
-    int8_t *i4x4_cache= h->intra4x4_pred_mode_cache;
+static av_always_inline void write_back_intra_pred_mode(H264Context *h)
+{
+    int8_t *i4x4       = h->intra4x4_pred_mode + h->mb2br_xy[h->mb_xy];
+    int8_t *i4x4_cache = h->intra4x4_pred_mode_cache;
 
-    AV_COPY32(i4x4, i4x4_cache + 4 + 8*4);
-    i4x4[4]= i4x4_cache[7+8*3];
-    i4x4[5]= i4x4_cache[7+8*2];
-    i4x4[6]= i4x4_cache[7+8*1];
+    AV_COPY32(i4x4, i4x4_cache + 4 + 8 * 4);
+    i4x4[4] = i4x4_cache[7 + 8 * 3];
+    i4x4[5] = i4x4_cache[7 + 8 * 2];
+    i4x4[6] = i4x4_cache[7 + 8 * 1];
 }
 
-static av_always_inline void write_back_non_zero_count(H264Context *h){
-    const int mb_xy= h->mb_xy;
-    uint8_t *nnz = h->non_zero_count[mb_xy];
+static av_always_inline void write_back_non_zero_count(H264Context *h)
+{
+    const int mb_xy    = h->mb_xy;
+    uint8_t *nnz       = h->non_zero_count[mb_xy];
     uint8_t *nnz_cache = h->non_zero_count_cache;
 
-    AV_COPY32(&nnz[ 0], &nnz_cache[4+8* 1]);
-    AV_COPY32(&nnz[ 4], &nnz_cache[4+8* 2]);
-    AV_COPY32(&nnz[ 8], &nnz_cache[4+8* 3]);
-    AV_COPY32(&nnz[12], &nnz_cache[4+8* 4]);
-    AV_COPY32(&nnz[16], &nnz_cache[4+8* 6]);
-    AV_COPY32(&nnz[20], &nnz_cache[4+8* 7]);
-    AV_COPY32(&nnz[32], &nnz_cache[4+8*11]);
-    AV_COPY32(&nnz[36], &nnz_cache[4+8*12]);
-
-    if(!h->s.chroma_y_shift){
-        AV_COPY32(&nnz[24], &nnz_cache[4+8* 8]);
-        AV_COPY32(&nnz[28], &nnz_cache[4+8* 9]);
-        AV_COPY32(&nnz[40], &nnz_cache[4+8*13]);
-        AV_COPY32(&nnz[44], &nnz_cache[4+8*14]);
+    AV_COPY32(&nnz[ 0], &nnz_cache[4 + 8 * 1]);
+    AV_COPY32(&nnz[ 4], &nnz_cache[4 + 8 * 2]);
+    AV_COPY32(&nnz[ 8], &nnz_cache[4 + 8 * 3]);
+    AV_COPY32(&nnz[12], &nnz_cache[4 + 8 * 4]);
+    AV_COPY32(&nnz[16], &nnz_cache[4 + 8 * 6]);
+    AV_COPY32(&nnz[20], &nnz_cache[4 + 8 * 7]);
+    AV_COPY32(&nnz[32], &nnz_cache[4 + 8 * 11]);
+    AV_COPY32(&nnz[36], &nnz_cache[4 + 8 * 12]);
+
+    if (!h->s.chroma_y_shift) {
+        AV_COPY32(&nnz[24], &nnz_cache[4 + 8 * 8]);
+        AV_COPY32(&nnz[28], &nnz_cache[4 + 8 * 9]);
+        AV_COPY32(&nnz[40], &nnz_cache[4 + 8 * 13]);
+        AV_COPY32(&nnz[44], &nnz_cache[4 + 8 * 14]);
     }
 }
 
-static av_always_inline void write_back_motion_list(H264Context *h, MpegEncContext * const s, int b_stride,
-                                                    int b_xy, int b8_xy, int mb_type, int list )
+static av_always_inline void write_back_motion_list(H264Context *h,
+                                                    MpegEncContext *const s,
+                                                    int b_stride,
+                                                    int b_xy, int b8_xy,
+                                                    int mb_type, int list)
 {
-    int16_t (*mv_dst)[2] = &s->current_picture.f.motion_val[list][b_xy];
-    int16_t (*mv_src)[2] = &h->mv_cache[list][scan8[0]];
-    AV_COPY128(mv_dst + 0*b_stride, mv_src + 8*0);
-    AV_COPY128(mv_dst + 1*b_stride, mv_src + 8*1);
-    AV_COPY128(mv_dst + 2*b_stride, mv_src + 8*2);
-    AV_COPY128(mv_dst + 3*b_stride, mv_src + 8*3);
-    if( CABAC ) {
-        uint8_t (*mvd_dst)[2] = &h->mvd_table[list][FMO ? 8*h->mb_xy : h->mb2br_xy[h->mb_xy]];
-        uint8_t (*mvd_src)[2] = &h->mvd_cache[list][scan8[0]];
-        if(IS_SKIP(mb_type))
+    int16_t(*mv_dst)[2] = &s->current_picture.f.motion_val[list][b_xy];
+    int16_t(*mv_src)[2] = &h->mv_cache[list][scan8[0]];
+    AV_COPY128(mv_dst + 0 * b_stride, mv_src + 8 * 0);
+    AV_COPY128(mv_dst + 1 * b_stride, mv_src + 8 * 1);
+    AV_COPY128(mv_dst + 2 * b_stride, mv_src + 8 * 2);
+    AV_COPY128(mv_dst + 3 * b_stride, mv_src + 8 * 3);
+    if (CABAC) {
+        uint8_t (*mvd_dst)[2] = &h->mvd_table[list][FMO ? 8 * h->mb_xy
+                                                        : h->mb2br_xy[h->mb_xy]];
+        uint8_t(*mvd_src)[2]  = &h->mvd_cache[list][scan8[0]];
+        if (IS_SKIP(mb_type)) {
             AV_ZERO128(mvd_dst);
-        else{
-            AV_COPY64(mvd_dst, mvd_src + 8*3);
-            AV_COPY16(mvd_dst + 3 + 3, mvd_src + 3 + 8*0);
-            AV_COPY16(mvd_dst + 3 + 2, mvd_src + 3 + 8*1);
-            AV_COPY16(mvd_dst + 3 + 1, mvd_src + 3 + 8*2);
+        } else {
+            AV_COPY64(mvd_dst, mvd_src + 8 * 3);
+            AV_COPY16(mvd_dst + 3 + 3, mvd_src + 3 + 8 * 0);
+            AV_COPY16(mvd_dst + 3 + 2, mvd_src + 3 + 8 * 1);
+            AV_COPY16(mvd_dst + 3 + 1, mvd_src + 3 + 8 * 2);
         }
     }
 
     {
         int8_t *ref_index = &s->current_picture.f.ref_index[list][b8_xy];
         int8_t *ref_cache = h->ref_cache[list];
-        ref_index[0+0*2]= ref_cache[scan8[0]];
-        ref_index[1+0*2]= ref_cache[scan8[4]];
-        ref_index[0+1*2]= ref_cache[scan8[8]];
-        ref_index[1+1*2]= ref_cache[scan8[12]];
+        ref_index[0 + 0 * 2] = ref_cache[scan8[0]];
+        ref_index[1 + 0 * 2] = ref_cache[scan8[4]];
+        ref_index[0 + 1 * 2] = ref_cache[scan8[8]];
+        ref_index[1 + 1 * 2] = ref_cache[scan8[12]];
     }
 }
 
-static av_always_inline void write_back_motion(H264Context *h, int mb_type){
-    MpegEncContext * const s = &h->s;
-    const int b_stride = h->b_stride;
-    const int b_xy = 4*s->mb_x + 4*s->mb_y*h->b_stride; //try mb2b(8)_xy
-    const int b8_xy= 4*h->mb_xy;
+static av_always_inline void write_back_motion(H264Context *h, int mb_type)
+{
+    MpegEncContext *const s = &h->s;
+    const int b_stride      = h->b_stride;
+    const int b_xy  = 4 * s->mb_x + 4 * s->mb_y * h->b_stride; // try mb2b(8)_xy
+    const int b8_xy = 4 * h->mb_xy;
 
-    if(USES_LIST(mb_type, 0)){
+    if (USES_LIST(mb_type, 0)) {
         write_back_motion_list(h, s, b_stride, b_xy, b8_xy, mb_type, 0);
-    }else{
+    } else {
         fill_rectangle(&s->current_picture.f.ref_index[0][b8_xy],
                        2, 2, 2, (uint8_t)LIST_NOT_USED, 1);
     }
-    if(USES_LIST(mb_type, 1)){
+    if (USES_LIST(mb_type, 1))
         write_back_motion_list(h, s, b_stride, b_xy, b8_xy, mb_type, 1);
-    }
 
-    if(h->slice_type_nos == AV_PICTURE_TYPE_B && CABAC){
-        if(IS_8X8(mb_type)){
-            uint8_t *direct_table = &h->direct_table[4*h->mb_xy];
-            direct_table[1] = h->sub_mb_type[1]>>1;
-            direct_table[2] = h->sub_mb_type[2]>>1;
-            direct_table[3] = h->sub_mb_type[3]>>1;
+    if (h->slice_type_nos == AV_PICTURE_TYPE_B && CABAC) {
+        if (IS_8X8(mb_type)) {
+            uint8_t *direct_table = &h->direct_table[4 * h->mb_xy];
+            direct_table[1] = h->sub_mb_type[1] >> 1;
+            direct_table[2] = h->sub_mb_type[2] >> 1;
+            direct_table[3] = h->sub_mb_type[3] >> 1;
         }
     }
 }
 
-static av_always_inline int get_dct8x8_allowed(H264Context *h){
-    if(h->sps.direct_8x8_inference_flag)
-        return !(AV_RN64A(h->sub_mb_type) & ((MB_TYPE_16x8|MB_TYPE_8x16|MB_TYPE_8x8                )*0x0001000100010001ULL));
+static av_always_inline int get_dct8x8_allowed(H264Context *h)
+{
+    if (h->sps.direct_8x8_inference_flag)
+        return !(AV_RN64A(h->sub_mb_type) &
+                 ((MB_TYPE_16x8 | MB_TYPE_8x16 | MB_TYPE_8x8) *
+                  0x0001000100010001ULL));
     else
-        return !(AV_RN64A(h->sub_mb_type) & ((MB_TYPE_16x8|MB_TYPE_8x16|MB_TYPE_8x8|MB_TYPE_DIRECT2)*0x0001000100010001ULL));
+        return !(AV_RN64A(h->sub_mb_type) &
+                 ((MB_TYPE_16x8 | MB_TYPE_8x16 | MB_TYPE_8x8 | MB_TYPE_DIRECT2) *
+                  0x0001000100010001ULL));
 }
 
 #endif /* AVCODEC_H264_H */

File diff suppressed because it is too large
+ 507 - 462
libavcodec/h264_mvpred.h


+ 186 - 183
libavcodec/h264data.h

@@ -30,240 +30,243 @@
 #define AVCODEC_H264DATA_H
 
 #include <stdint.h>
+
 #include "libavutil/rational.h"
 #include "mpegvideo.h"
 #include "h264.h"
 
+static const uint8_t golomb_to_pict_type[5] = {
+    AV_PICTURE_TYPE_P, AV_PICTURE_TYPE_B, AV_PICTURE_TYPE_I,
+    AV_PICTURE_TYPE_SP, AV_PICTURE_TYPE_SI
+};
 
-static const uint8_t golomb_to_pict_type[5]=
-{AV_PICTURE_TYPE_P, AV_PICTURE_TYPE_B, AV_PICTURE_TYPE_I, AV_PICTURE_TYPE_SP, AV_PICTURE_TYPE_SI};
-
-static const uint8_t golomb_to_intra4x4_cbp[48]={
- 47, 31, 15,  0, 23, 27, 29, 30,  7, 11, 13, 14, 39, 43, 45, 46,
- 16,  3,  5, 10, 12, 19, 21, 26, 28, 35, 37, 42, 44,  1,  2,  4,
-  8, 17, 18, 20, 24,  6,  9, 22, 25, 32, 33, 34, 36, 40, 38, 41
+static const uint8_t golomb_to_intra4x4_cbp[48] = {
+    47, 31, 15, 0,  23, 27, 29, 30, 7,  11, 13, 14, 39, 43, 45, 46,
+    16, 3,  5,  10, 12, 19, 21, 26, 28, 35, 37, 42, 44, 1,  2,  4,
+    8,  17, 18, 20, 24, 6,  9,  22, 25, 32, 33, 34, 36, 40, 38, 41
 };
 
-static const uint8_t golomb_to_inter_cbp[48]={
-  0, 16,  1,  2,  4,  8, 32,  3,  5, 10, 12, 15, 47,  7, 11, 13,
- 14,  6,  9, 31, 35, 37, 42, 44, 33, 34, 36, 40, 39, 43, 45, 46,
- 17, 18, 20, 24, 19, 21, 26, 28, 23, 27, 29, 30, 22, 25, 38, 41
+static const uint8_t golomb_to_inter_cbp[48] = {
+    0,  16, 1,  2,  4,  8,  32, 3,  5,  10, 12, 15, 47, 7,  11, 13,
+    14, 6,  9,  31, 35, 37, 42, 44, 33, 34, 36, 40, 39, 43, 45, 46,
+    17, 18, 20, 24, 19, 21, 26, 28, 23, 27, 29, 30, 22, 25, 38, 41
 };
 
-static const uint8_t zigzag_scan[16]={
- 0+0*4, 1+0*4, 0+1*4, 0+2*4,
- 1+1*4, 2+0*4, 3+0*4, 2+1*4,
- 1+2*4, 0+3*4, 1+3*4, 2+2*4,
- 3+1*4, 3+2*4, 2+3*4, 3+3*4,
+static const uint8_t zigzag_scan[16] = {
+    0 + 0 * 4, 1 + 0 * 4, 0 + 1 * 4, 0 + 2 * 4,
+    1 + 1 * 4, 2 + 0 * 4, 3 + 0 * 4, 2 + 1 * 4,
+    1 + 2 * 4, 0 + 3 * 4, 1 + 3 * 4, 2 + 2 * 4,
+    3 + 1 * 4, 3 + 2 * 4, 2 + 3 * 4, 3 + 3 * 4,
 };
 
-static const uint8_t field_scan[16]={
- 0+0*4, 0+1*4, 1+0*4, 0+2*4,
- 0+3*4, 1+1*4, 1+2*4, 1+3*4,
- 2+0*4, 2+1*4, 2+2*4, 2+3*4,
- 3+0*4, 3+1*4, 3+2*4, 3+3*4,
+static const uint8_t field_scan[16] = {
+    0 + 0 * 4, 0 + 1 * 4, 1 + 0 * 4, 0 + 2 * 4,
+    0 + 3 * 4, 1 + 1 * 4, 1 + 2 * 4, 1 + 3 * 4,
+    2 + 0 * 4, 2 + 1 * 4, 2 + 2 * 4, 2 + 3 * 4,
+    3 + 0 * 4, 3 + 1 * 4, 3 + 2 * 4, 3 + 3 * 4,
 };
 
-static const uint8_t luma_dc_zigzag_scan[16]={
- 0*16 + 0*64, 1*16 + 0*64, 2*16 + 0*64, 0*16 + 2*64,
- 3*16 + 0*64, 0*16 + 1*64, 1*16 + 1*64, 2*16 + 1*64,
- 1*16 + 2*64, 2*16 + 2*64, 3*16 + 2*64, 0*16 + 3*64,
- 3*16 + 1*64, 1*16 + 3*64, 2*16 + 3*64, 3*16 + 3*64,
+static const uint8_t luma_dc_zigzag_scan[16] = {
+    0 * 16 + 0 * 64, 1 * 16 + 0 * 64, 2 * 16 + 0 * 64, 0 * 16 + 2 * 64,
+    3 * 16 + 0 * 64, 0 * 16 + 1 * 64, 1 * 16 + 1 * 64, 2 * 16 + 1 * 64,
+    1 * 16 + 2 * 64, 2 * 16 + 2 * 64, 3 * 16 + 2 * 64, 0 * 16 + 3 * 64,
+    3 * 16 + 1 * 64, 1 * 16 + 3 * 64, 2 * 16 + 3 * 64, 3 * 16 + 3 * 64,
 };
 
-static const uint8_t luma_dc_field_scan[16]={
- 0*16 + 0*64, 2*16 + 0*64, 1*16 + 0*64, 0*16 + 2*64,
- 2*16 + 2*64, 3*16 + 0*64, 1*16 + 2*64, 3*16 + 2*64,
- 0*16 + 1*64, 2*16 + 1*64, 0*16 + 3*64, 2*16 + 3*64,
- 1*16 + 1*64, 3*16 + 1*64, 1*16 + 3*64, 3*16 + 3*64,
+static const uint8_t luma_dc_field_scan[16] = {
+    0 * 16 + 0 * 64, 2 * 16 + 0 * 64, 1 * 16 + 0 * 64, 0 * 16 + 2 * 64,
+    2 * 16 + 2 * 64, 3 * 16 + 0 * 64, 1 * 16 + 2 * 64, 3 * 16 + 2 * 64,
+    0 * 16 + 1 * 64, 2 * 16 + 1 * 64, 0 * 16 + 3 * 64, 2 * 16 + 3 * 64,
+    1 * 16 + 1 * 64, 3 * 16 + 1 * 64, 1 * 16 + 3 * 64, 3 * 16 + 3 * 64,
 };
 
-static const uint8_t chroma_dc_scan[4]={
- (0+0*2)*16, (1+0*2)*16,
- (0+1*2)*16, (1+1*2)*16,
+static const uint8_t chroma_dc_scan[4] = {
+    (0 + 0 * 2) * 16, (1 + 0 * 2) * 16,
+    (0 + 1 * 2) * 16, (1 + 1 * 2) * 16,
 };
 
-static const uint8_t chroma422_dc_scan[8]={
- (0+0*2)*16, (0+1*2)*16,
- (1+0*2)*16, (0+2*2)*16,
- (0+3*2)*16, (1+1*2)*16,
- (1+2*2)*16, (1+3*2)*16,
+static const uint8_t chroma422_dc_scan[8] = {
+    (0 + 0 * 2) * 16, (0 + 1 * 2) * 16,
+    (1 + 0 * 2) * 16, (0 + 2 * 2) * 16,
+    (0 + 3 * 2) * 16, (1 + 1 * 2) * 16,
+    (1 + 2 * 2) * 16, (1 + 3 * 2) * 16,
 };
 
 // zigzag_scan8x8_cavlc[i] = zigzag_scan8x8[(i/4) + 16*(i%4)]
-static const uint8_t zigzag_scan8x8_cavlc[64]={
- 0+0*8, 1+1*8, 1+2*8, 2+2*8,
- 4+1*8, 0+5*8, 3+3*8, 7+0*8,
- 3+4*8, 1+7*8, 5+3*8, 6+3*8,
- 2+7*8, 6+4*8, 5+6*8, 7+5*8,
- 1+0*8, 2+0*8, 0+3*8, 3+1*8,
- 3+2*8, 0+6*8, 4+2*8, 6+1*8,
- 2+5*8, 2+6*8, 6+2*8, 5+4*8,
- 3+7*8, 7+3*8, 4+7*8, 7+6*8,
- 0+1*8, 3+0*8, 0+4*8, 4+0*8,
- 2+3*8, 1+5*8, 5+1*8, 5+2*8,
- 1+6*8, 3+5*8, 7+1*8, 4+5*8,
- 4+6*8, 7+4*8, 5+7*8, 6+7*8,
- 0+2*8, 2+1*8, 1+3*8, 5+0*8,
- 1+4*8, 2+4*8, 6+0*8, 4+3*8,
- 0+7*8, 4+4*8, 7+2*8, 3+6*8,
- 5+5*8, 6+5*8, 6+6*8, 7+7*8,
+static const uint8_t zigzag_scan8x8_cavlc[64] = {
+    0 + 0 * 8, 1 + 1 * 8, 1 + 2 * 8, 2 + 2 * 8,
+    4 + 1 * 8, 0 + 5 * 8, 3 + 3 * 8, 7 + 0 * 8,
+    3 + 4 * 8, 1 + 7 * 8, 5 + 3 * 8, 6 + 3 * 8,
+    2 + 7 * 8, 6 + 4 * 8, 5 + 6 * 8, 7 + 5 * 8,
+    1 + 0 * 8, 2 + 0 * 8, 0 + 3 * 8, 3 + 1 * 8,
+    3 + 2 * 8, 0 + 6 * 8, 4 + 2 * 8, 6 + 1 * 8,
+    2 + 5 * 8, 2 + 6 * 8, 6 + 2 * 8, 5 + 4 * 8,
+    3 + 7 * 8, 7 + 3 * 8, 4 + 7 * 8, 7 + 6 * 8,
+    0 + 1 * 8, 3 + 0 * 8, 0 + 4 * 8, 4 + 0 * 8,
+    2 + 3 * 8, 1 + 5 * 8, 5 + 1 * 8, 5 + 2 * 8,
+    1 + 6 * 8, 3 + 5 * 8, 7 + 1 * 8, 4 + 5 * 8,
+    4 + 6 * 8, 7 + 4 * 8, 5 + 7 * 8, 6 + 7 * 8,
+    0 + 2 * 8, 2 + 1 * 8, 1 + 3 * 8, 5 + 0 * 8,
+    1 + 4 * 8, 2 + 4 * 8, 6 + 0 * 8, 4 + 3 * 8,
+    0 + 7 * 8, 4 + 4 * 8, 7 + 2 * 8, 3 + 6 * 8,
+    5 + 5 * 8, 6 + 5 * 8, 6 + 6 * 8, 7 + 7 * 8,
 };
 
-static const uint8_t field_scan8x8[64]={
- 0+0*8, 0+1*8, 0+2*8, 1+0*8,
- 1+1*8, 0+3*8, 0+4*8, 1+2*8,
- 2+0*8, 1+3*8, 0+5*8, 0+6*8,
- 0+7*8, 1+4*8, 2+1*8, 3+0*8,
- 2+2*8, 1+5*8, 1+6*8, 1+7*8,
- 2+3*8, 3+1*8, 4+0*8, 3+2*8,
- 2+4*8, 2+5*8, 2+6*8, 2+7*8,
- 3+3*8, 4+1*8, 5+0*8, 4+2*8,
- 3+4*8, 3+5*8, 3+6*8, 3+7*8,
- 4+3*8, 5+1*8, 6+0*8, 5+2*8,
- 4+4*8, 4+5*8, 4+6*8, 4+7*8,
- 5+3*8, 6+1*8, 6+2*8, 5+4*8,
- 5+5*8, 5+6*8, 5+7*8, 6+3*8,
- 7+0*8, 7+1*8, 6+4*8, 6+5*8,
- 6+6*8, 6+7*8, 7+2*8, 7+3*8,
- 7+4*8, 7+5*8, 7+6*8, 7+7*8,
+static const uint8_t field_scan8x8[64] = {
+    0 + 0 * 8, 0 + 1 * 8, 0 + 2 * 8, 1 + 0 * 8,
+    1 + 1 * 8, 0 + 3 * 8, 0 + 4 * 8, 1 + 2 * 8,
+    2 + 0 * 8, 1 + 3 * 8, 0 + 5 * 8, 0 + 6 * 8,
+    0 + 7 * 8, 1 + 4 * 8, 2 + 1 * 8, 3 + 0 * 8,
+    2 + 2 * 8, 1 + 5 * 8, 1 + 6 * 8, 1 + 7 * 8,
+    2 + 3 * 8, 3 + 1 * 8, 4 + 0 * 8, 3 + 2 * 8,
+    2 + 4 * 8, 2 + 5 * 8, 2 + 6 * 8, 2 + 7 * 8,
+    3 + 3 * 8, 4 + 1 * 8, 5 + 0 * 8, 4 + 2 * 8,
+    3 + 4 * 8, 3 + 5 * 8, 3 + 6 * 8, 3 + 7 * 8,
+    4 + 3 * 8, 5 + 1 * 8, 6 + 0 * 8, 5 + 2 * 8,
+    4 + 4 * 8, 4 + 5 * 8, 4 + 6 * 8, 4 + 7 * 8,
+    5 + 3 * 8, 6 + 1 * 8, 6 + 2 * 8, 5 + 4 * 8,
+    5 + 5 * 8, 5 + 6 * 8, 5 + 7 * 8, 6 + 3 * 8,
+    7 + 0 * 8, 7 + 1 * 8, 6 + 4 * 8, 6 + 5 * 8,
+    6 + 6 * 8, 6 + 7 * 8, 7 + 2 * 8, 7 + 3 * 8,
+    7 + 4 * 8, 7 + 5 * 8, 7 + 6 * 8, 7 + 7 * 8,
 };
 
-static const uint8_t field_scan8x8_cavlc[64]={
- 0+0*8, 1+1*8, 2+0*8, 0+7*8,
- 2+2*8, 2+3*8, 2+4*8, 3+3*8,
- 3+4*8, 4+3*8, 4+4*8, 5+3*8,
- 5+5*8, 7+0*8, 6+6*8, 7+4*8,
- 0+1*8, 0+3*8, 1+3*8, 1+4*8,
- 1+5*8, 3+1*8, 2+5*8, 4+1*8,
- 3+5*8, 5+1*8, 4+5*8, 6+1*8,
- 5+6*8, 7+1*8, 6+7*8, 7+5*8,
- 0+2*8, 0+4*8, 0+5*8, 2+1*8,
- 1+6*8, 4+0*8, 2+6*8, 5+0*8,
- 3+6*8, 6+0*8, 4+6*8, 6+2*8,
- 5+7*8, 6+4*8, 7+2*8, 7+6*8,
- 1+0*8, 1+2*8, 0+6*8, 3+0*8,
- 1+7*8, 3+2*8, 2+7*8, 4+2*8,
- 3+7*8, 5+2*8, 4+7*8, 5+4*8,
- 6+3*8, 6+5*8, 7+3*8, 7+7*8,
+static const uint8_t field_scan8x8_cavlc[64] = {
+    0 + 0 * 8, 1 + 1 * 8, 2 + 0 * 8, 0 + 7 * 8,
+    2 + 2 * 8, 2 + 3 * 8, 2 + 4 * 8, 3 + 3 * 8,
+    3 + 4 * 8, 4 + 3 * 8, 4 + 4 * 8, 5 + 3 * 8,
+    5 + 5 * 8, 7 + 0 * 8, 6 + 6 * 8, 7 + 4 * 8,
+    0 + 1 * 8, 0 + 3 * 8, 1 + 3 * 8, 1 + 4 * 8,
+    1 + 5 * 8, 3 + 1 * 8, 2 + 5 * 8, 4 + 1 * 8,
+    3 + 5 * 8, 5 + 1 * 8, 4 + 5 * 8, 6 + 1 * 8,
+    5 + 6 * 8, 7 + 1 * 8, 6 + 7 * 8, 7 + 5 * 8,
+    0 + 2 * 8, 0 + 4 * 8, 0 + 5 * 8, 2 + 1 * 8,
+    1 + 6 * 8, 4 + 0 * 8, 2 + 6 * 8, 5 + 0 * 8,
+    3 + 6 * 8, 6 + 0 * 8, 4 + 6 * 8, 6 + 2 * 8,
+    5 + 7 * 8, 6 + 4 * 8, 7 + 2 * 8, 7 + 6 * 8,
+    1 + 0 * 8, 1 + 2 * 8, 0 + 6 * 8, 3 + 0 * 8,
+    1 + 7 * 8, 3 + 2 * 8, 2 + 7 * 8, 4 + 2 * 8,
+    3 + 7 * 8, 5 + 2 * 8, 4 + 7 * 8, 5 + 4 * 8,
+    6 + 3 * 8, 6 + 5 * 8, 7 + 3 * 8, 7 + 7 * 8,
 };
 
-typedef struct IMbInfo{
+typedef struct IMbInfo {
     uint16_t type;
     uint8_t pred_mode;
     uint8_t cbp;
 } IMbInfo;
 
-static const IMbInfo i_mb_type_info[26]={
-{MB_TYPE_INTRA4x4  , -1, -1},
-{MB_TYPE_INTRA16x16,  2,  0},
-{MB_TYPE_INTRA16x16,  1,  0},
-{MB_TYPE_INTRA16x16,  0,  0},
-{MB_TYPE_INTRA16x16,  3,  0},
-{MB_TYPE_INTRA16x16,  2,  16},
-{MB_TYPE_INTRA16x16,  1,  16},
-{MB_TYPE_INTRA16x16,  0,  16},
-{MB_TYPE_INTRA16x16,  3,  16},
-{MB_TYPE_INTRA16x16,  2,  32},
-{MB_TYPE_INTRA16x16,  1,  32},
-{MB_TYPE_INTRA16x16,  0,  32},
-{MB_TYPE_INTRA16x16,  3,  32},
-{MB_TYPE_INTRA16x16,  2,  15+0},
-{MB_TYPE_INTRA16x16,  1,  15+0},
-{MB_TYPE_INTRA16x16,  0,  15+0},
-{MB_TYPE_INTRA16x16,  3,  15+0},
-{MB_TYPE_INTRA16x16,  2,  15+16},
-{MB_TYPE_INTRA16x16,  1,  15+16},
-{MB_TYPE_INTRA16x16,  0,  15+16},
-{MB_TYPE_INTRA16x16,  3,  15+16},
-{MB_TYPE_INTRA16x16,  2,  15+32},
-{MB_TYPE_INTRA16x16,  1,  15+32},
-{MB_TYPE_INTRA16x16,  0,  15+32},
-{MB_TYPE_INTRA16x16,  3,  15+32},
-{MB_TYPE_INTRA_PCM , -1, -1},
+static const IMbInfo i_mb_type_info[26] = {
+    { MB_TYPE_INTRA4x4,  -1,  -1 },
+    { MB_TYPE_INTRA16x16, 2,   0 },
+    { MB_TYPE_INTRA16x16, 1,   0 },
+    { MB_TYPE_INTRA16x16, 0,   0 },
+    { MB_TYPE_INTRA16x16, 3,   0 },
+    { MB_TYPE_INTRA16x16, 2,  16 },
+    { MB_TYPE_INTRA16x16, 1,  16 },
+    { MB_TYPE_INTRA16x16, 0,  16 },
+    { MB_TYPE_INTRA16x16, 3,  16 },
+    { MB_TYPE_INTRA16x16, 2,  32 },
+    { MB_TYPE_INTRA16x16, 1,  32 },
+    { MB_TYPE_INTRA16x16, 0,  32 },
+    { MB_TYPE_INTRA16x16, 3,  32 },
+    { MB_TYPE_INTRA16x16, 2,  15 +  0 },
+    { MB_TYPE_INTRA16x16, 1,  15 +  0 },
+    { MB_TYPE_INTRA16x16, 0,  15 +  0 },
+    { MB_TYPE_INTRA16x16, 3,  15 +  0 },
+    { MB_TYPE_INTRA16x16, 2,  15 + 16 },
+    { MB_TYPE_INTRA16x16, 1,  15 + 16 },
+    { MB_TYPE_INTRA16x16, 0,  15 + 16 },
+    { MB_TYPE_INTRA16x16, 3,  15 + 16 },
+    { MB_TYPE_INTRA16x16, 2,  15 + 32 },
+    { MB_TYPE_INTRA16x16, 1,  15 + 32 },
+    { MB_TYPE_INTRA16x16, 0,  15 + 32 },
+    { MB_TYPE_INTRA16x16, 3,  15 + 32 },
+    { MB_TYPE_INTRA_PCM,  -1, -1 },
 };
 
-typedef struct PMbInfo{
+typedef struct PMbInfo {
     uint16_t type;
     uint8_t partition_count;
 } PMbInfo;
 
-static const PMbInfo p_mb_type_info[5]={
-{MB_TYPE_16x16|MB_TYPE_P0L0             , 1},
-{MB_TYPE_16x8 |MB_TYPE_P0L0|MB_TYPE_P1L0, 2},
-{MB_TYPE_8x16 |MB_TYPE_P0L0|MB_TYPE_P1L0, 2},
-{MB_TYPE_8x8  |MB_TYPE_P0L0|MB_TYPE_P1L0, 4},
-{MB_TYPE_8x8  |MB_TYPE_P0L0|MB_TYPE_P1L0|MB_TYPE_REF0, 4},
+static const PMbInfo p_mb_type_info[5] = {
+    { MB_TYPE_16x16 | MB_TYPE_P0L0,                               1 },
+    { MB_TYPE_16x8  | MB_TYPE_P0L0 | MB_TYPE_P1L0,                2 },
+    { MB_TYPE_8x16  | MB_TYPE_P0L0 | MB_TYPE_P1L0,                2 },
+    { MB_TYPE_8x8   | MB_TYPE_P0L0 | MB_TYPE_P1L0,                4 },
+    { MB_TYPE_8x8   | MB_TYPE_P0L0 | MB_TYPE_P1L0 | MB_TYPE_REF0, 4 },
 };
 
-static const PMbInfo p_sub_mb_type_info[4]={
-{MB_TYPE_16x16|MB_TYPE_P0L0             , 1},
-{MB_TYPE_16x8 |MB_TYPE_P0L0             , 2},
-{MB_TYPE_8x16 |MB_TYPE_P0L0             , 2},
-{MB_TYPE_8x8  |MB_TYPE_P0L0             , 4},
+static const PMbInfo p_sub_mb_type_info[4] = {
+    { MB_TYPE_16x16 | MB_TYPE_P0L0, 1 },
+    { MB_TYPE_16x8  | MB_TYPE_P0L0, 2 },
+    { MB_TYPE_8x16  | MB_TYPE_P0L0, 2 },
+    { MB_TYPE_8x8   | MB_TYPE_P0L0, 4 },
 };
 
-static const PMbInfo b_mb_type_info[23]={
-{MB_TYPE_DIRECT2|MB_TYPE_L0L1                                      , 1, },
-{MB_TYPE_16x16|MB_TYPE_P0L0                                       , 1, },
-{MB_TYPE_16x16             |MB_TYPE_P0L1                          , 1, },
-{MB_TYPE_16x16|MB_TYPE_P0L0|MB_TYPE_P0L1                          , 1, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0             |MB_TYPE_P1L0             , 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0             |MB_TYPE_P1L0             , 2, },
-{MB_TYPE_16x8              |MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16              |MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0                          |MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0                          |MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8              |MB_TYPE_P0L1|MB_TYPE_P1L0             , 2, },
-{MB_TYPE_8x16              |MB_TYPE_P0L1|MB_TYPE_P1L0             , 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0             |MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0             |MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8              |MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16              |MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0             , 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0             , 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0|MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0|MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x8  |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 4, },
+static const PMbInfo b_mb_type_info[23] = {
+    { MB_TYPE_DIRECT2 | MB_TYPE_L0L1,                                              1, },
+    { MB_TYPE_16x16   | MB_TYPE_P0L0,                                              1, },
+    { MB_TYPE_16x16   | MB_TYPE_P0L1,                                              1, },
+    { MB_TYPE_16x16   | MB_TYPE_P0L0 | MB_TYPE_P0L1,                               1, },
+    { MB_TYPE_16x8    | MB_TYPE_P0L0 | MB_TYPE_P1L0,                               2, },
+    { MB_TYPE_8x16    | MB_TYPE_P0L0 | MB_TYPE_P1L0,                               2, },
+    { MB_TYPE_16x8    | MB_TYPE_P0L1 | MB_TYPE_P1L1,                               2, },
+    { MB_TYPE_8x16    | MB_TYPE_P0L1 | MB_TYPE_P1L1,                               2, },
+    { MB_TYPE_16x8    | MB_TYPE_P0L0 | MB_TYPE_P1L1,                               2, },
+    { MB_TYPE_8x16    | MB_TYPE_P0L0 | MB_TYPE_P1L1,                               2, },
+    { MB_TYPE_16x8    | MB_TYPE_P0L1 | MB_TYPE_P1L0,                               2, },
+    { MB_TYPE_8x16    | MB_TYPE_P0L1 | MB_TYPE_P1L0,                               2, },
+    { MB_TYPE_16x8    | MB_TYPE_P0L0 | MB_TYPE_P1L0 | MB_TYPE_P1L1,                2, },
+    { MB_TYPE_8x16    | MB_TYPE_P0L0 | MB_TYPE_P1L0 | MB_TYPE_P1L1,                2, },
+    { MB_TYPE_16x8    | MB_TYPE_P0L1 | MB_TYPE_P1L0 | MB_TYPE_P1L1,                2, },
+    { MB_TYPE_8x16    | MB_TYPE_P0L1 | MB_TYPE_P1L0 | MB_TYPE_P1L1,                2, },
+    { MB_TYPE_16x8    | MB_TYPE_P0L0 | MB_TYPE_P0L1 | MB_TYPE_P1L0,                2, },
+    { MB_TYPE_8x16    | MB_TYPE_P0L0 | MB_TYPE_P0L1 | MB_TYPE_P1L0,                2, },
+    { MB_TYPE_16x8    | MB_TYPE_P0L0 | MB_TYPE_P0L1 | MB_TYPE_P1L1,                2, },
+    { MB_TYPE_8x16    | MB_TYPE_P0L0 | MB_TYPE_P0L1 | MB_TYPE_P1L1,                2, },
+    { MB_TYPE_16x8    | MB_TYPE_P0L0 | MB_TYPE_P0L1 | MB_TYPE_P1L0 | MB_TYPE_P1L1, 2, },
+    { MB_TYPE_8x16    | MB_TYPE_P0L0 | MB_TYPE_P0L1 | MB_TYPE_P1L0 | MB_TYPE_P1L1, 2, },
+    { MB_TYPE_8x8     | MB_TYPE_P0L0 | MB_TYPE_P0L1 | MB_TYPE_P1L0 | MB_TYPE_P1L1, 4, },
 };
 
-static const PMbInfo b_sub_mb_type_info[13]={
-{MB_TYPE_DIRECT2                                                   , 1, },
-{MB_TYPE_16x16|MB_TYPE_P0L0                                       , 1, },
-{MB_TYPE_16x16             |MB_TYPE_P0L1                          , 1, },
-{MB_TYPE_16x16|MB_TYPE_P0L0|MB_TYPE_P0L1                          , 1, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0             |MB_TYPE_P1L0             , 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0             |MB_TYPE_P1L0             , 2, },
-{MB_TYPE_16x8              |MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16              |MB_TYPE_P0L1             |MB_TYPE_P1L1, 2, },
-{MB_TYPE_16x8 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x16 |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 2, },
-{MB_TYPE_8x8  |MB_TYPE_P0L0             |MB_TYPE_P1L0             , 4, },
-{MB_TYPE_8x8               |MB_TYPE_P0L1             |MB_TYPE_P1L1, 4, },
-{MB_TYPE_8x8  |MB_TYPE_P0L0|MB_TYPE_P0L1|MB_TYPE_P1L0|MB_TYPE_P1L1, 4, },
+static const PMbInfo b_sub_mb_type_info[13] = {
+    { MB_TYPE_DIRECT2,                                                           1, },
+    { MB_TYPE_16x16 | MB_TYPE_P0L0,                                              1, },
+    { MB_TYPE_16x16 | MB_TYPE_P0L1,                                              1, },
+    { MB_TYPE_16x16 | MB_TYPE_P0L0 | MB_TYPE_P0L1,                               1, },
+    { MB_TYPE_16x8  | MB_TYPE_P0L0 | MB_TYPE_P1L0,                               2, },
+    { MB_TYPE_8x16  | MB_TYPE_P0L0 | MB_TYPE_P1L0,                               2, },
+    { MB_TYPE_16x8  | MB_TYPE_P0L1 | MB_TYPE_P1L1,                               2, },
+    { MB_TYPE_8x16  | MB_TYPE_P0L1 | MB_TYPE_P1L1,                               2, },
+    { MB_TYPE_16x8  | MB_TYPE_P0L0 | MB_TYPE_P0L1 | MB_TYPE_P1L0 | MB_TYPE_P1L1, 2, },
+    { MB_TYPE_8x16  | MB_TYPE_P0L0 | MB_TYPE_P0L1 | MB_TYPE_P1L0 | MB_TYPE_P1L1, 2, },
+    { MB_TYPE_8x8   | MB_TYPE_P0L0 | MB_TYPE_P1L0,                               4, },
+    { MB_TYPE_8x8   | MB_TYPE_P0L1 | MB_TYPE_P1L1,                               4, },
+    { MB_TYPE_8x8   | MB_TYPE_P0L0 | MB_TYPE_P0L1 | MB_TYPE_P1L0 | MB_TYPE_P1L1, 4, },
 };
 
-static const uint8_t dequant4_coeff_init[6][3]={
-  {10,13,16},
-  {11,14,18},
-  {13,16,20},
-  {14,18,23},
-  {16,20,25},
-  {18,23,29},
+static const uint8_t dequant4_coeff_init[6][3] = {
+    { 10, 13, 16 },
+    { 11, 14, 18 },
+    { 13, 16, 20 },
+    { 14, 18, 23 },
+    { 16, 20, 25 },
+    { 18, 23, 29 },
 };
 
 static const uint8_t dequant8_coeff_init_scan[16] = {
-  0,3,4,3, 3,1,5,1, 4,5,2,5, 3,1,5,1
+    0, 3, 4, 3, 3, 1, 5, 1, 4, 5, 2, 5, 3, 1, 5, 1
 };
-static const uint8_t dequant8_coeff_init[6][6]={
-  {20,18,32,19,25,24},
-  {22,19,35,21,28,26},
-  {26,23,42,24,33,31},
-  {28,25,45,26,35,33},
-  {32,28,51,30,40,38},
-  {36,32,58,34,46,43},
+
+static const uint8_t dequant8_coeff_init[6][6] = {
+    { 20, 18, 32, 19, 25, 24 },
+    { 22, 19, 35, 21, 28, 26 },
+    { 26, 23, 42, 24, 33, 31 },
+    { 28, 25, 45, 26, 35, 33 },
+    { 32, 28, 51, 30, 40, 38 },
+    { 36, 32, 58, 34, 46, 43 },
 };
 
 #endif /* AVCODEC_H264DATA_H */

+ 65 - 31
libavcodec/h264dsp.h

@@ -28,56 +28,90 @@
 #define AVCODEC_H264DSP_H
 
 #include <stdint.h>
+
 #include "dsputil.h"
 
 typedef void (*h264_weight_func)(uint8_t *block, int stride, int height,
                                  int log2_denom, int weight, int offset);
-typedef void (*h264_biweight_func)(uint8_t *dst, uint8_t *src, int stride, int height,
-                                   int log2_denom, int weightd, int weights, int offset);
+typedef void (*h264_biweight_func)(uint8_t *dst, uint8_t *src,
+                                   int stride, int height, int log2_denom,
+                                   int weightd, int weights, int offset);
 
 /**
  * Context for storing H.264 DSP functions
  */
-typedef struct H264DSPContext{
+typedef struct H264DSPContext {
     /* weighted MC */
     h264_weight_func weight_h264_pixels_tab[4];
     h264_biweight_func biweight_h264_pixels_tab[4];
 
     /* loop filter */
-    void (*h264_v_loop_filter_luma)(uint8_t *pix/*align 16*/, int stride, int alpha, int beta, int8_t *tc0);
-    void (*h264_h_loop_filter_luma)(uint8_t *pix/*align 4 */, int stride, int alpha, int beta, int8_t *tc0);
-    void (*h264_h_loop_filter_luma_mbaff)(uint8_t *pix/*align 16*/, int stride, int alpha, int beta, int8_t *tc0);
+    void (*h264_v_loop_filter_luma)(uint8_t *pix /*align 16*/, int stride,
+                                    int alpha, int beta, int8_t *tc0);
+    void (*h264_h_loop_filter_luma)(uint8_t *pix /*align 4 */, int stride,
+                                    int alpha, int beta, int8_t *tc0);
+    void (*h264_h_loop_filter_luma_mbaff)(uint8_t *pix /*align 16*/, int stride,
+                                          int alpha, int beta, int8_t *tc0);
     /* v/h_loop_filter_luma_intra: align 16 */
-    void (*h264_v_loop_filter_luma_intra)(uint8_t *pix, int stride, int alpha, int beta);
-    void (*h264_h_loop_filter_luma_intra)(uint8_t *pix, int stride, int alpha, int beta);
-    void (*h264_h_loop_filter_luma_mbaff_intra)(uint8_t *pix/*align 16*/, int stride, int alpha, int beta);
-    void (*h264_v_loop_filter_chroma)(uint8_t *pix/*align 8*/, int stride, int alpha, int beta, int8_t *tc0);
-    void (*h264_h_loop_filter_chroma)(uint8_t *pix/*align 4*/, int stride, int alpha, int beta, int8_t *tc0);
-    void (*h264_h_loop_filter_chroma_mbaff)(uint8_t *pix/*align 8*/, int stride, int alpha, int beta, int8_t *tc0);
-    void (*h264_v_loop_filter_chroma_intra)(uint8_t *pix/*align 8*/, int stride, int alpha, int beta);
-    void (*h264_h_loop_filter_chroma_intra)(uint8_t *pix/*align 8*/, int stride, int alpha, int beta);
-    void (*h264_h_loop_filter_chroma_mbaff_intra)(uint8_t *pix/*align 8*/, int stride, int alpha, int beta);
+    void (*h264_v_loop_filter_luma_intra)(uint8_t *pix, int stride,
+                                          int alpha, int beta);
+    void (*h264_h_loop_filter_luma_intra)(uint8_t *pix, int stride,
+                                          int alpha, int beta);
+    void (*h264_h_loop_filter_luma_mbaff_intra)(uint8_t *pix /*align 16*/,
+                                                int stride, int alpha, int beta);
+    void (*h264_v_loop_filter_chroma)(uint8_t *pix /*align 8*/, int stride,
+                                      int alpha, int beta, int8_t *tc0);
+    void (*h264_h_loop_filter_chroma)(uint8_t *pix /*align 4*/, int stride,
+                                      int alpha, int beta, int8_t *tc0);
+    void (*h264_h_loop_filter_chroma_mbaff)(uint8_t *pix /*align 8*/,
+                                            int stride, int alpha, int beta,
+                                            int8_t *tc0);
+    void (*h264_v_loop_filter_chroma_intra)(uint8_t *pix /*align 8*/,
+                                            int stride, int alpha, int beta);
+    void (*h264_h_loop_filter_chroma_intra)(uint8_t *pix /*align 8*/,
+                                            int stride, int alpha, int beta);
+    void (*h264_h_loop_filter_chroma_mbaff_intra)(uint8_t *pix /*align 8*/,
+                                                  int stride, int alpha, int beta);
     // h264_loop_filter_strength: simd only. the C version is inlined in h264.c
-    void (*h264_loop_filter_strength)(int16_t bS[2][4][4], uint8_t nnz[40], int8_t ref[2][40], int16_t mv[2][40][2],
-                                      int bidir, int edges, int step, int mask_mv0, int mask_mv1, int field);
+    void (*h264_loop_filter_strength)(int16_t bS[2][4][4], uint8_t nnz[40],
+                                      int8_t ref[2][40], int16_t mv[2][40][2],
+                                      int bidir, int edges, int step,
+                                      int mask_mv0, int mask_mv1, int field);
 
     /* IDCT */
-    void (*h264_idct_add)(uint8_t *dst/*align 4*/, DCTELEM *block/*align 16*/, int stride);
-    void (*h264_idct8_add)(uint8_t *dst/*align 8*/, DCTELEM *block/*align 16*/, int stride);
-    void (*h264_idct_dc_add)(uint8_t *dst/*align 4*/, DCTELEM *block/*align 16*/, int stride);
-    void (*h264_idct8_dc_add)(uint8_t *dst/*align 8*/, DCTELEM *block/*align 16*/, int stride);
+    void (*h264_idct_add)(uint8_t *dst /*align 4*/,
+                          DCTELEM *block /*align 16*/, int stride);
+    void (*h264_idct8_add)(uint8_t *dst /*align 8*/,
+                           DCTELEM *block /*align 16*/, int stride);
+    void (*h264_idct_dc_add)(uint8_t *dst /*align 4*/,
+                             DCTELEM *block /*align 16*/, int stride);
+    void (*h264_idct8_dc_add)(uint8_t *dst /*align 8*/,
+                              DCTELEM *block /*align 16*/, int stride);
 
-    void (*h264_idct_add16)(uint8_t *dst/*align 16*/, const int *blockoffset, DCTELEM *block/*align 16*/, int stride, const uint8_t nnzc[15*8]);
-    void (*h264_idct8_add4)(uint8_t *dst/*align 16*/, const int *blockoffset, DCTELEM *block/*align 16*/, int stride, const uint8_t nnzc[15*8]);
-    void (*h264_idct_add8)(uint8_t **dst/*align 16*/, const int *blockoffset, DCTELEM *block/*align 16*/, int stride, const uint8_t nnzc[15*8]);
-    void (*h264_idct_add16intra)(uint8_t *dst/*align 16*/, const int *blockoffset, DCTELEM *block/*align 16*/, int stride, const uint8_t nnzc[15*8]);
-    void (*h264_luma_dc_dequant_idct)(DCTELEM *output, DCTELEM *input/*align 16*/, int qmul);
+    void (*h264_idct_add16)(uint8_t *dst /*align 16*/, const int *blockoffset,
+                            DCTELEM *block /*align 16*/, int stride,
+                            const uint8_t nnzc[15 * 8]);
+    void (*h264_idct8_add4)(uint8_t *dst /*align 16*/, const int *blockoffset,
+                            DCTELEM *block /*align 16*/, int stride,
+                            const uint8_t nnzc[15 * 8]);
+    void (*h264_idct_add8)(uint8_t **dst /*align 16*/, const int *blockoffset,
+                           DCTELEM *block /*align 16*/, int stride,
+                           const uint8_t nnzc[15 * 8]);
+    void (*h264_idct_add16intra)(uint8_t *dst /*align 16*/, const int *blockoffset,
+                                 DCTELEM *block /*align 16*/,
+                                 int stride, const uint8_t nnzc[15 * 8]);
+    void (*h264_luma_dc_dequant_idct)(DCTELEM *output,
+                                      DCTELEM *input /*align 16*/, int qmul);
     void (*h264_chroma_dc_dequant_idct)(DCTELEM *block, int qmul);
-}H264DSPContext;
+} H264DSPContext;
 
-void ff_h264dsp_init(H264DSPContext *c, const int bit_depth, const int chroma_format_idc);
-void ff_h264dsp_init_arm(H264DSPContext *c, const int bit_depth, const int chroma_format_idc);
-void ff_h264dsp_init_ppc(H264DSPContext *c, const int bit_depth, const int chroma_format_idc);
-void ff_h264dsp_init_x86(H264DSPContext *c, const int bit_depth, const int chroma_format_idc);
+void ff_h264dsp_init(H264DSPContext *c, const int bit_depth,
+                     const int chroma_format_idc);
+void ff_h264dsp_init_arm(H264DSPContext *c, const int bit_depth,
+                         const int chroma_format_idc);
+void ff_h264dsp_init_ppc(H264DSPContext *c, const int bit_depth,
+                         const int chroma_format_idc);
+void ff_h264dsp_init_x86(H264DSPContext *c, const int bit_depth,
+                         const int chroma_format_idc);
 
 #endif /* AVCODEC_H264DSP_H */

+ 45 - 36
libavcodec/h264pred.h

@@ -35,18 +35,18 @@
  * Prediction types
  */
 //@{
-#define VERT_PRED             0
-#define HOR_PRED              1
-#define DC_PRED               2
-#define DIAG_DOWN_LEFT_PRED   3
-#define DIAG_DOWN_RIGHT_PRED  4
-#define VERT_RIGHT_PRED       5
-#define HOR_DOWN_PRED         6
-#define VERT_LEFT_PRED        7
-#define HOR_UP_PRED           8
+#define VERT_PRED              0
+#define HOR_PRED               1
+#define DC_PRED                2
+#define DIAG_DOWN_LEFT_PRED    3
+#define DIAG_DOWN_RIGHT_PRED   4
+#define VERT_RIGHT_PRED        5
+#define HOR_DOWN_PRED          6
+#define VERT_LEFT_PRED         7
+#define HOR_UP_PRED            8
 
 // DC edge (not for VP8)
-#define LEFT_DC_PRED          9
+#define LEFT_DC_PRED           9
 #define TOP_DC_PRED           10
 #define DC_128_PRED           11
 
@@ -56,7 +56,7 @@
 #define VERT_LEFT_PRED_RV40_NODOWN        14
 
 // VP8 specific
-#define TM_VP8_PRED           9     ///< "True Motion", used instead of plane
+#define TM_VP8_PRED            9    ///< "True Motion", used instead of plane
 #define VERT_VP8_PRED         10    ///< for VP8, #VERT_PRED is the average of
                                     ///< (left col+cur col x2+right col) / 4;
                                     ///< this is the "unaveraged" one
@@ -65,44 +65,53 @@
 #define DC_127_PRED           12
 #define DC_129_PRED           13
 
-#define DC_PRED8x8            0
-#define HOR_PRED8x8           1
-#define VERT_PRED8x8          2
-#define PLANE_PRED8x8         3
+#define DC_PRED8x8             0
+#define HOR_PRED8x8            1
+#define VERT_PRED8x8           2
+#define PLANE_PRED8x8          3
 
 // DC edge
-#define LEFT_DC_PRED8x8       4
-#define TOP_DC_PRED8x8        5
-#define DC_128_PRED8x8        6
+#define LEFT_DC_PRED8x8        4
+#define TOP_DC_PRED8x8         5
+#define DC_128_PRED8x8         6
 
 // H264/SVQ3 (8x8) specific
-#define ALZHEIMER_DC_L0T_PRED8x8 7
-#define ALZHEIMER_DC_0LT_PRED8x8 8
-#define ALZHEIMER_DC_L00_PRED8x8 9
+#define ALZHEIMER_DC_L0T_PRED8x8  7
+#define ALZHEIMER_DC_0LT_PRED8x8  8
+#define ALZHEIMER_DC_L00_PRED8x8  9
 #define ALZHEIMER_DC_0L0_PRED8x8 10
 
 // VP8 specific
-#define DC_127_PRED8x8        7
-#define DC_129_PRED8x8        8
+#define DC_127_PRED8x8         7
+#define DC_129_PRED8x8         8
 //@}
 
 /**
  * Context for storing H.264 prediction functions
  */
-typedef struct H264PredContext{
-    void (*pred4x4  [9+3+3])(uint8_t *src, const uint8_t *topright, int stride);//FIXME move to dsp?
-    void (*pred8x8l [9+3])(uint8_t *src, int topleft, int topright, int stride);
-    void (*pred8x8  [4+3+4])(uint8_t *src, int stride);
-    void (*pred16x16[4+3+2])(uint8_t *src, int stride);
+typedef struct H264PredContext {
+    void(*pred4x4[9 + 3 + 3])(uint8_t *src, const uint8_t *topright, int stride); //FIXME move to dsp?
+    void(*pred8x8l[9 + 3])(uint8_t *src, int topleft, int topright, int stride);
+    void(*pred8x8[4 + 3 + 4])(uint8_t *src, int stride);
+    void(*pred16x16[4 + 3 + 2])(uint8_t *src, int stride);
 
-    void (*pred4x4_add  [2])(uint8_t *pix/*align  4*/, const DCTELEM *block/*align 16*/, int stride);
-    void (*pred8x8l_add [2])(uint8_t *pix/*align  8*/, const DCTELEM *block/*align 16*/, int stride);
-    void (*pred8x8_add  [3])(uint8_t *pix/*align  8*/, const int *block_offset, const DCTELEM *block/*align 16*/, int stride);
-    void (*pred16x16_add[3])(uint8_t *pix/*align 16*/, const int *block_offset, const DCTELEM *block/*align 16*/, int stride);
-}H264PredContext;
+    void(*pred4x4_add[2])(uint8_t *pix /*align  4*/,
+                          const DCTELEM *block /*align 16*/, int stride);
+    void(*pred8x8l_add[2])(uint8_t *pix /*align  8*/,
+                           const DCTELEM *block /*align 16*/, int stride);
+    void(*pred8x8_add[3])(uint8_t *pix /*align  8*/,
+                          const int *block_offset,
+                          const DCTELEM *block /*align 16*/, int stride);
+    void(*pred16x16_add[3])(uint8_t *pix /*align 16*/,
+                            const int *block_offset,
+                            const DCTELEM *block /*align 16*/, int stride);
+} H264PredContext;
 
-void ff_h264_pred_init(H264PredContext *h, int codec_id, const int bit_depth, const int chroma_format_idc);
-void ff_h264_pred_init_arm(H264PredContext *h, int codec_id, const int bit_depth, const int chroma_format_idc);
-void ff_h264_pred_init_x86(H264PredContext *h, int codec_id, const int bit_depth, const int chroma_format_idc);
+void ff_h264_pred_init(H264PredContext *h, int codec_id,
+                       const int bit_depth, const int chroma_format_idc);
+void ff_h264_pred_init_arm(H264PredContext *h, int codec_id,
+                           const int bit_depth, const int chroma_format_idc);
+void ff_h264_pred_init_x86(H264PredContext *h, int codec_id,
+                           const int bit_depth, const int chroma_format_idc);
 
 #endif /* AVCODEC_H264PRED_H */

+ 102 - 115
libavcodec/vcr1.c

@@ -21,39 +21,57 @@
 
 /**
  * @file
- * ati vcr1 codec.
+ * ATI VCR1 codec
  */
 
 #include "avcodec.h"
 #include "dsputil.h"
 
-//#undef NDEBUG
-//#include <assert.h>
-
-/* Disable the encoder. */
-#undef CONFIG_VCR1_ENCODER
-#define CONFIG_VCR1_ENCODER 0
-
-typedef struct VCR1Context{
-    AVCodecContext *avctx;
+typedef struct VCR1Context {
     AVFrame picture;
     int delta[16];
     int offset[4];
 } VCR1Context;
 
-static int decode_frame(AVCodecContext *avctx,
-                        void *data, int *data_size,
-                        AVPacket *avpkt)
+static av_cold void common_init(AVCodecContext *avctx)
+{
+    VCR1Context *const a = avctx->priv_data;
+
+    avctx->coded_frame = &a->picture;
+    avcodec_get_frame_defaults(&a->picture);
+}
+
+static av_cold int decode_init(AVCodecContext *avctx)
+{
+    common_init(avctx);
+
+    avctx->pix_fmt = PIX_FMT_YUV410P;
+
+    return 0;
+}
+
+static av_cold int decode_end(AVCodecContext *avctx)
+{
+    VCR1Context *s = avctx->priv_data;
+
+    if (s->picture.data[0])
+        avctx->release_buffer(avctx, &s->picture);
+
+    return 0;
+}
+
+static int decode_frame(AVCodecContext *avctx, void *data,
+                        int *data_size, AVPacket *avpkt)
 {
-    const uint8_t *buf = avpkt->data;
-    int buf_size = avpkt->size;
-    VCR1Context * const a = avctx->priv_data;
-    AVFrame *picture = data;
-    AVFrame * const p = &a->picture;
-    const uint8_t *bytestream= buf;
+    const uint8_t *buf        = avpkt->data;
+    int buf_size              = avpkt->size;
+    VCR1Context *const a      = avctx->priv_data;
+    AVFrame *picture          = data;
+    AVFrame *const p          = &a->picture;
+    const uint8_t *bytestream = buf;
     int i, x, y;
 
-    if(p->data[0])
+    if (p->data[0])
         avctx->release_buffer(avctx, p);
 
     if(buf_size < 16 + avctx->height + avctx->width*avctx->height*5/8){
@@ -61,57 +79,57 @@ static int decode_frame(AVCodecContext *avctx,
         return AVERROR(EINVAL);
     }
 
-    p->reference= 0;
-    if(avctx->get_buffer(avctx, p) < 0){
+    p->reference = 0;
+    if (avctx->get_buffer(avctx, p) < 0) {
         av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
         return -1;
     }
-    p->pict_type= AV_PICTURE_TYPE_I;
-    p->key_frame= 1;
+    p->pict_type = AV_PICTURE_TYPE_I;
+    p->key_frame = 1;
 
-    for(i=0; i<16; i++){
-        a->delta[i]= *(bytestream++);
+    for (i = 0; i < 16; i++) {
+        a->delta[i] = *bytestream++;
         bytestream++;
     }
 
-    for(y=0; y<avctx->height; y++){
+    for (y = 0; y < avctx->height; y++) {
         int offset;
-        uint8_t *luma= &a->picture.data[0][ y*a->picture.linesize[0] ];
+        uint8_t *luma = &a->picture.data[0][y * a->picture.linesize[0]];
 
-        if((y&3) == 0){
-            uint8_t *cb= &a->picture.data[1][ (y>>2)*a->picture.linesize[1] ];
-            uint8_t *cr= &a->picture.data[2][ (y>>2)*a->picture.linesize[2] ];
+        if ((y & 3) == 0) {
+            uint8_t *cb = &a->picture.data[1][(y >> 2) * a->picture.linesize[1]];
+            uint8_t *cr = &a->picture.data[2][(y >> 2) * a->picture.linesize[2]];
 
-            for(i=0; i<4; i++)
-                a->offset[i]= *(bytestream++);
+            for (i = 0; i < 4; i++)
+                a->offset[i] = *bytestream++;
 
-            offset= a->offset[0] - a->delta[ bytestream[2]&0xF ];
-            for(x=0; x<avctx->width; x+=4){
-                luma[0]=( offset += a->delta[ bytestream[2]&0xF ]);
-                luma[1]=( offset += a->delta[ bytestream[2]>>4  ]);
-                luma[2]=( offset += a->delta[ bytestream[0]&0xF ]);
-                luma[3]=( offset += a->delta[ bytestream[0]>>4  ]);
-                luma += 4;
+            offset = a->offset[0] - a->delta[bytestream[2] & 0xF];
+            for (x = 0; x < avctx->width; x += 4) {
+                luma[0]     = offset += a->delta[bytestream[2] & 0xF];
+                luma[1]     = offset += a->delta[bytestream[2] >>  4];
+                luma[2]     = offset += a->delta[bytestream[0] & 0xF];
+                luma[3]     = offset += a->delta[bytestream[0] >>  4];
+                luma       += 4;
 
-                *(cb++) = bytestream[3];
-                *(cr++) = bytestream[1];
+                *cb++       = bytestream[3];
+                *cr++       = bytestream[1];
 
-                bytestream+= 4;
+                bytestream += 4;
             }
-        }else{
-            offset= a->offset[y&3] - a->delta[ bytestream[2]&0xF ];
-
-            for(x=0; x<avctx->width; x+=8){
-                luma[0]=( offset += a->delta[ bytestream[2]&0xF ]);
-                luma[1]=( offset += a->delta[ bytestream[2]>>4  ]);
-                luma[2]=( offset += a->delta[ bytestream[3]&0xF ]);
-                luma[3]=( offset += a->delta[ bytestream[3]>>4  ]);
-                luma[4]=( offset += a->delta[ bytestream[0]&0xF ]);
-                luma[5]=( offset += a->delta[ bytestream[0]>>4  ]);
-                luma[6]=( offset += a->delta[ bytestream[1]&0xF ]);
-                luma[7]=( offset += a->delta[ bytestream[1]>>4  ]);
-                luma += 8;
-                bytestream+= 4;
+        } else {
+            offset = a->offset[y & 3] - a->delta[bytestream[2] & 0xF];
+
+            for (x = 0; x < avctx->width; x += 8) {
+                luma[0]     = offset += a->delta[bytestream[2] & 0xF];
+                luma[1]     = offset += a->delta[bytestream[2] >>  4];
+                luma[2]     = offset += a->delta[bytestream[3] & 0xF];
+                luma[3]     = offset += a->delta[bytestream[3] >>  4];
+                luma[4]     = offset += a->delta[bytestream[0] & 0xF];
+                luma[5]     = offset += a->delta[bytestream[0] >>  4];
+                luma[6]     = offset += a->delta[bytestream[1] & 0xF];
+                luma[7]     = offset += a->delta[bytestream[1] >>  4];
+                luma       += 8;
+                bytestream += 4;
             }
         }
     }
@@ -122,62 +140,6 @@ static int decode_frame(AVCodecContext *avctx,
     return buf_size;
 }
 
-#if CONFIG_VCR1_ENCODER
-static int encode_frame(AVCodecContext *avctx, unsigned char *buf, int buf_size, void *data){
-    VCR1Context * const a = avctx->priv_data;
-    AVFrame *pict = data;
-    AVFrame * const p = &a->picture;
-    int size;
-
-    *p = *pict;
-    p->pict_type= AV_PICTURE_TYPE_I;
-    p->key_frame= 1;
-
-    avpriv_align_put_bits(&a->pb);
-    while(get_bit_count(&a->pb)&31)
-        put_bits(&a->pb, 8, 0);
-
-    size= get_bit_count(&a->pb)/32;
-
-    return size*4;
-}
-#endif
-
-static av_cold void common_init(AVCodecContext *avctx){
-    VCR1Context * const a = avctx->priv_data;
-
-    avctx->coded_frame = &a->picture;
-    avcodec_get_frame_defaults(&a->picture);
-    a->avctx= avctx;
-}
-
-static av_cold int decode_init(AVCodecContext *avctx){
-
-    common_init(avctx);
-
-    avctx->pix_fmt= PIX_FMT_YUV410P;
-
-    return 0;
-}
-
-static av_cold int decode_end(AVCodecContext *avctx){
-    VCR1Context *s = avctx->priv_data;
-
-    if (s->picture.data[0])
-        avctx->release_buffer(avctx, &s->picture);
-
-    return 0;
-}
-
-#if CONFIG_VCR1_ENCODER
-static av_cold int encode_init(AVCodecContext *avctx){
-
-    common_init(avctx);
-
-    return 0;
-}
-#endif
-
 AVCodec ff_vcr1_decoder = {
     .name           = "vcr1",
     .type           = AVMEDIA_TYPE_VIDEO,
@@ -190,14 +152,39 @@ AVCodec ff_vcr1_decoder = {
     .long_name      = NULL_IF_CONFIG_SMALL("ATI VCR1"),
 };
 
+/* Disable the encoder. */
+#undef CONFIG_VCR1_ENCODER
+#define CONFIG_VCR1_ENCODER 0
+
 #if CONFIG_VCR1_ENCODER
+static int encode_frame(AVCodecContext *avctx, unsigned char *buf,
+                        int buf_size, void *data)
+{
+    VCR1Context *const a = avctx->priv_data;
+    AVFrame *pict        = data;
+    AVFrame *const p     = &a->picture;
+    int size;
+
+    *p           = *pict;
+    p->pict_type = AV_PICTURE_TYPE_I;
+    p->key_frame = 1;
+
+    avpriv_align_put_bits(&a->pb);
+    while (get_bit_count(&a->pb) & 31)
+        put_bits(&a->pb, 8, 0);
+
+    size = get_bit_count(&a->pb) / 32;
+
+    return size * 4;
+}
+
 AVCodec ff_vcr1_encoder = {
     .name           = "vcr1",
     .type           = AVMEDIA_TYPE_VIDEO,
     .id             = CODEC_ID_VCR1,
     .priv_data_size = sizeof(VCR1Context),
-    .init           = encode_init,
+    .init           = common_init,
     .encode         = encode_frame,
     .long_name      = NULL_IF_CONFIG_SMALL("ATI VCR1"),
 };
-#endif
+#endif /* CONFIG_VCR1_ENCODER */

Some files were not shown because too many files changed in this diff