Merge remote-tracking branch 'qatar/master'

* qatar/master:
  ARM: ac3: update ff_ac3_extract_exponents_neon per 8b7b2d6
  ARM: NEON optimised vector_clip_int32()
  swscale: disable full_chroma_int when converting to non-24/32bpp RGB.
  suggest to use av_get_bytes_per_sample() in av_get_bits_per_sample_format() doxy
  ffmpeg: use av_get_bytes_per_sample() in place of av_get_bits_per_sample_fmt()
  put_bits: remove ALT_BITSTREAM_WRITER
  put_bits: always use intreadwrite.h macros
  libavformat: Add an example how to use the metadata API
  doxygen: Prefer member groups over grouping into modules
  doxygen: be more permissive when searching for API examples
  avformat: doxify the Metadata API
  lavf: restore old behavior for custom AVIOContex with an AVFMT_NOFILE format.
  lavf: use the correct pointer in av_open_input_stream().
  avidec: infer absolute vs relative index from first packet

Conflicts:
	libavformat/Makefile
	libavformat/avidec.c

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

22 files changed, 159 insertions, 263 deletions

diff --git a/libavcodec/aac.h b/libavcodec/aac.h
index f089ee9a29..477acb5d4e 100644
--- a/libavcodec/aac.h
+++ b/libavcodec/aac.h
@@ -258,7 +258,7 @@ typedef struct {
     DynamicRangeControl che_drc;
 
     /**
-     * @defgroup elements Channel element related data.
+     * @name Channel element related data
      * @{
      */
     enum ChannelPosition che_pos[4][MAX_ELEM_ID]; /**< channel element channel mapping with the
@@ -270,14 +270,15 @@ typedef struct {
     /** @} */
 
     /**
-     * @defgroup temporary aligned temporary buffers (We do not want to have these on the stack.)
+     * @name temporary aligned temporary buffers
+     * (We do not want to have these on the stack.)
      * @{
      */
     DECLARE_ALIGNED(32, float, buf_mdct)[1024];
     /** @} */
 
     /**
-     * @defgroup tables   Computed / set up during initialization.
+     * @name Computed / set up during initialization
      * @{
      */
     FFTContext mdct;
@@ -289,7 +290,7 @@ typedef struct {
     /** @} */
 
     /**
-     * @defgroup output   Members used for output interleaving.
+     * @name Members used for output interleaving
      * @{
      */
     float *output_data[MAX_CHANNELS];                 ///< Points to each element's 'ret' buffer (PCM output).
diff --git a/libavcodec/ac3.h b/libavcodec/ac3.h
index a53817e1b0..8e8482f7c5 100644
--- a/libavcodec/ac3.h
+++ b/libavcodec/ac3.h
@@ -94,7 +94,7 @@ typedef struct AC3BitAllocParameters {
  * Coded AC-3 header values up to the lfeon element, plus derived values.
  */
 typedef struct {
-    /** @defgroup coded Coded elements
+    /** @name Coded elements
      * @{
      */
     uint16_t sync_word;
@@ -112,7 +112,7 @@ typedef struct {
     int num_blocks;                         ///< number of audio blocks
     /** @} */
 
-    /** @defgroup derived Derived values
+    /** @name Derived values
      * @{
      */
     uint8_t sr_shift;
diff --git a/libavcodec/ac3dec.h b/libavcodec/ac3dec.h
index 377e5154d7..4b3e7e3d7e 100644
--- a/libavcodec/ac3dec.h
+++ b/libavcodec/ac3dec.h
@@ -69,7 +69,7 @@ typedef struct {
     AVCodecContext *avctx;                  ///< parent context
     GetBitContext gbc;                      ///< bitstream reader
 
-///@defgroup bsi bit stream information
+///@name Bit stream information
 ///@{
     int frame_type;                         ///< frame type                             (strmtyp)
     int substreamid;                        ///< substream identification
@@ -87,7 +87,7 @@ typedef struct {
     int eac3;                               ///< indicates if current frame is E-AC-3
 ///@}
 
-///@defgroup audfrm frame syntax parameters
+///@name Frame syntax parameters
     int snr_offset_strategy;                ///< SNR offset strategy                    (snroffststr)
     int block_switch_syntax;                ///< block switch syntax enabled            (blkswe)
     int dither_flag_syntax;                 ///< dither flag syntax enabled             (dithflage)
@@ -97,7 +97,7 @@ typedef struct {
     int skip_syntax;                        ///< skip field syntax enabled              (skipflde)
  ///@}
 
-///@defgroup cpl standard coupling
+///@name Standard coupling
     int cpl_in_use[AC3_MAX_BLOCKS];         ///< coupling in use                        (cplinu)
     int cpl_strategy_exists[AC3_MAX_BLOCKS];///< coupling strategy exists               (cplstre)
     int channel_in_cpl[AC3_MAX_CHANNELS];   ///< channel in coupling                    (chincpl)
@@ -110,7 +110,7 @@ typedef struct {
     int cpl_coords[AC3_MAX_CHANNELS][AC3_MAX_CPL_BANDS]; ///< coupling coordinates      (cplco)
 ///@}
 
-///@defgroup spx spectral extension
+///@name Spectral extension
 ///@{
     int spx_in_use;                             ///< spectral extension in use              (spxinu)
     uint8_t channel_uses_spx[AC3_MAX_CHANNELS]; ///< channel uses spectral extension        (chinspx)
@@ -126,12 +126,12 @@ typedef struct {
     float spx_signal_blend[AC3_MAX_CHANNELS][SPX_MAX_BANDS];///< spx signal blending factor (sblendfact)
 ///@}
 
-///@defgroup aht adaptive hybrid transform
+///@name Adaptive hybrid transform
     int channel_uses_aht[AC3_MAX_CHANNELS];                         ///< channel AHT in use (chahtinu)
     int pre_mantissa[AC3_MAX_CHANNELS][AC3_MAX_COEFS][AC3_MAX_BLOCKS];  ///< pre-IDCT mantissas
 ///@}
 
-///@defgroup channel channel
+///@name Channel
     int fbw_channels;                           ///< number of full-bandwidth channels
     int channels;                               ///< number of total channels
     int lfe_ch;                                 ///< index of LFE channel
@@ -141,27 +141,27 @@ typedef struct {
     int out_channels;                           ///< number of output channels
 ///@}
 
-///@defgroup dynrng dynamic range
+///@name Dynamic range
     float dynamic_range[2];                 ///< dynamic range
 ///@}
 
-///@defgroup bandwidth bandwidth
+///@name Bandwidth
     int start_freq[AC3_MAX_CHANNELS];       ///< start frequency bin                    (strtmant)
     int end_freq[AC3_MAX_CHANNELS];         ///< end frequency bin                      (endmant)
 ///@}
 
-///@defgroup rematrixing rematrixing
+///@name Rematrixing
     int num_rematrixing_bands;              ///< number of rematrixing bands            (nrematbnd)
     int rematrixing_flags[4];               ///< rematrixing flags                      (rematflg)
 ///@}
 
-///@defgroup exponents exponents
+///@name Exponents
     int num_exp_groups[AC3_MAX_CHANNELS];           ///< Number of exponent groups      (nexpgrp)
     int8_t dexps[AC3_MAX_CHANNELS][AC3_MAX_COEFS];  ///< decoded exponents
     int exp_strategy[AC3_MAX_BLOCKS][AC3_MAX_CHANNELS]; ///< exponent strategies        (expstr)
 ///@}
 
-///@defgroup bitalloc bit allocation
+///@name Bit allocation
     AC3BitAllocParameters bit_alloc_params;         ///< bit allocation parameters
     int first_cpl_leak;                             ///< first coupling leak state      (firstcplleak)
     int snr_offset[AC3_MAX_CHANNELS];               ///< signal-to-noise ratio offsets  (snroffst)
@@ -177,25 +177,25 @@ typedef struct {
     uint8_t dba_values[AC3_MAX_CHANNELS][8];        ///< delta values for each segment
 ///@}
 
-///@defgroup dithering zero-mantissa dithering
+///@name Zero-mantissa dithering
     int dither_flag[AC3_MAX_CHANNELS];      ///< dither flags                           (dithflg)
     AVLFG dith_state;                       ///< for dither generation
 ///@}
 
-///@defgroup imdct IMDCT
+///@name IMDCT
     int block_switch[AC3_MAX_CHANNELS];     ///< block switch flags                     (blksw)
     FFTContext imdct_512;                   ///< for 512 sample IMDCT
     FFTContext imdct_256;                   ///< for 256 sample IMDCT
 ///@}
 
-///@defgroup opt optimization
+///@name Optimization
     DSPContext dsp;                         ///< for optimization
     AC3DSPContext ac3dsp;
     FmtConvertContext fmt_conv;             ///< optimized conversion functions
     float mul_bias;                         ///< scaling for float_to_int16 conversion
 ///@}
 
-///@defgroup arrays aligned arrays
+///@name Aligned arrays
     DECLARE_ALIGNED(16, int,   fixed_coeffs)[AC3_MAX_CHANNELS][AC3_MAX_COEFS];       ///< fixed-point transform coefficients
     DECLARE_ALIGNED(32, float, transform_coeffs)[AC3_MAX_CHANNELS][AC3_MAX_COEFS];   ///< transform coefficients
     DECLARE_ALIGNED(32, float, delay)[AC3_MAX_CHANNELS][AC3_BLOCK_SIZE];             ///< delay - added to the next block
diff --git a/libavcodec/amrnbdec.c b/libavcodec/amrnbdec.c
index e0b6332065..b8d826e139 100644
--- a/libavcodec/amrnbdec.c
+++ b/libavcodec/amrnbdec.c
@@ -204,7 +204,7 @@ static enum Mode unpack_bitstream(AMRContext *p, const uint8_t *buf,
 }
 
 
-/// @defgroup amr_lpc_decoding AMR pitch LPC coefficient decoding functions
+/// @name AMR pitch LPC coefficient decoding functions
 /// @{
 
 /**
@@ -341,7 +341,7 @@ static void lsf2lsp_3(AMRContext *p)
 /// @}
 
 
-/// @defgroup amr_pitch_vector_decoding AMR pitch vector decoding functions
+/// @name AMR pitch vector decoding functions
 /// @{
 
 /**
@@ -403,7 +403,7 @@ static void decode_pitch_vector(AMRContext *p,
 /// @}
 
 
-/// @defgroup amr_algebraic_code_book AMR algebraic code book (fixed) vector decoding functions
+/// @name AMR algebraic code book (fixed) vector decoding functions
 /// @{
 
 /**
@@ -547,7 +547,7 @@ static void pitch_sharpening(AMRContext *p, int subframe, enum Mode mode,
 /// @}
 
 
-/// @defgroup amr_gain_decoding AMR gain decoding functions
+/// @name AMR gain decoding functions
 /// @{
 
 /**
@@ -633,7 +633,7 @@ static void decode_gains(AMRContext *p, const AMRNBSubframe *amr_subframe,
 /// @}
 
 
-/// @defgroup amr_pre_processing AMR pre-processing functions
+/// @name AMR preprocessing functions
 /// @{
 
 /**
@@ -751,7 +751,7 @@ static const float *anti_sparseness(AMRContext *p, AMRFixed *fixed_sparse,
 /// @}
 
 
-/// @defgroup amr_synthesis AMR synthesis functions
+/// @name AMR synthesis functions
 /// @{
 
 /**
@@ -812,7 +812,7 @@ static int synthesis(AMRContext *p, float *lpc,
 /// @}
 
 
-/// @defgroup amr_update AMR update functions
+/// @name AMR update functions
 /// @{
 
 /**
@@ -837,7 +837,7 @@ static void update_state(AMRContext *p)
 /// @}
 
 
-/// @defgroup amr_postproc AMR Post processing functions
+/// @name AMR Postprocessing functions
 /// @{
 
 /**
diff --git a/libavcodec/apedec.c b/libavcodec/apedec.c
index 05498777b4..f036c4a1d7 100644
--- a/libavcodec/apedec.c
+++ b/libavcodec/apedec.c
@@ -217,7 +217,7 @@ static av_cold int ape_decode_close(AVCodecContext * avctx)
 }
 
 /**
- * @defgroup rangecoder APE range decoder
+ * @name APE range decoding functions
  * @{
  */
 
diff --git a/libavcodec/arm/ac3dsp_neon.S b/libavcodec/arm/ac3dsp_neon.S
index fdf1deabc9..e97197c27a 100644
--- a/libavcodec/arm/ac3dsp_neon.S
+++ b/libavcodec/arm/ac3dsp_neon.S
@@ -95,19 +95,14 @@ function ff_float_to_fixed24_neon, export=1
 endfunc
 
 function ff_ac3_extract_exponents_neon, export=1
-        vmov.i32        q14, #24
         vmov.i32        q15, #8
 1:
-        vld1.32         {q0},     [r1,:128]
+        vld1.32         {q0},     [r1,:128]!
         vabs.s32        q1,  q0
         vclz.i32        q3,  q1
         vsub.i32        q3,  q3,  q15
-        vcge.s32        q2,  q3,  q14
-        vbit            q3,  q14, q2
-        vbic            q0,  q0,  q2
         vmovn.i32       d6,  q3
         vmovn.i16       d6,  q3
-        vst1.32         {q0},     [r1,:128]!
         vst1.32         {d6[0]},  [r0,:32]!
         subs            r2,  r2,  #4
         bgt             1b
diff --git a/libavcodec/arm/dsputil_init_neon.c b/libavcodec/arm/dsputil_init_neon.c
index 6faf3dc8d0..c818b33fef 100644
--- a/libavcodec/arm/dsputil_init_neon.c
+++ b/libavcodec/arm/dsputil_init_neon.c
@@ -160,6 +160,8 @@ void ff_vector_fmul_add_neon(float *dst, const float *src0, const float *src1,
 
 void ff_vector_clipf_neon(float *dst, const float *src, float min, float max,
                           int len);
+void ff_vector_clip_int32_neon(int32_t *dst, const int32_t *src, int32_t min,
+                               int32_t max, unsigned int len);
 
 void ff_vorbis_inverse_coupling_neon(float *mag, float *ang, int blocksize);
 
@@ -316,6 +318,7 @@ void ff_dsputil_init_neon(DSPContext *c, AVCodecContext *avctx)
     c->vector_fmul_reverse        = ff_vector_fmul_reverse_neon;
     c->vector_fmul_add            = ff_vector_fmul_add_neon;
     c->vector_clipf               = ff_vector_clipf_neon;
+    c->vector_clip_int32          = ff_vector_clip_int32_neon;
 
     c->vector_fmul_sv_scalar[0] = ff_vector_fmul_sv_scalar_2_neon;
     c->vector_fmul_sv_scalar[1] = ff_vector_fmul_sv_scalar_4_neon;
diff --git a/libavcodec/arm/dsputil_neon.S b/libavcodec/arm/dsputil_neon.S
index 2147658af6..250a32af6e 100644
--- a/libavcodec/arm/dsputil_neon.S
+++ b/libavcodec/arm/dsputil_neon.S
@@ -815,3 +815,19 @@ function ff_apply_window_int16_neon, export=1
 
         pop             {r4,pc}
 endfunc
+
+function ff_vector_clip_int32_neon, export=1
+        vdup.32         q0,  r2
+        vdup.32         q1,  r3
+        ldr             r2,  [sp]
+1:
+        vld1.32         {q2-q3},  [r1,:128]!
+        vmin.s32        q2,  q2,  q1
+        vmin.s32        q3,  q3,  q1
+        vmax.s32        q2,  q2,  q0
+        vmax.s32        q3,  q3,  q0
+        vst1.32         {q2-q3},  [r0,:128]!
+        subs            r2,  r2,  #8
+        bgt             1b
+        bx              lr
+endfunc
diff --git a/libavcodec/ass.h b/libavcodec/ass.h
index 74ef61b776..efff44d6ed 100644
--- a/libavcodec/ass.h
+++ b/libavcodec/ass.h
@@ -25,8 +25,7 @@
 #include "avcodec.h"
 
 /**
- * Default values for ASS style.
- * @defgroup ass_default
+ * @name Default values for ASS style
  * @{
  */
 #define ASS_DEFAULT_FONT        "Arial"
diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
index 4cd928dc83..1eb10ab41d 100644
--- a/libavcodec/avcodec.h
+++ b/libavcodec/avcodec.h
@@ -2535,7 +2535,7 @@ typedef struct AVCodecContext {
 
 #if FF_API_FLAC_GLOBAL_OPTS
     /**
-     * @defgroup flac_opts FLAC options
+     * @name FLAC options
      * @deprecated Use FLAC encoder private options instead.
      * @{
      */
@@ -2957,7 +2957,7 @@ typedef struct AVCodec {
     const AVProfile *profiles;              ///< array of recognized profiles, or NULL if unknown, array is terminated by {FF_PROFILE_UNKNOWN}
 
     /**
-     * @defgroup framethreading Frame-level threading support functions.
+     * @name Frame-level threading support functions
      * @{
      */
     /**
@@ -3915,7 +3915,7 @@ int av_get_bits_per_sample(enum CodecID codec_id);
 
 #if FF_API_OLD_SAMPLE_FMT
 /**
- * @deprecated Use av_get_bits_per_sample_fmt() instead.
+ * @deprecated Use av_get_bytes_per_sample() instead.
  */
 attribute_deprecated
 int av_get_bits_per_sample_format(enum AVSampleFormat sample_fmt);
diff --git a/libavcodec/bitstream.c b/libavcodec/bitstream.c
index f0fa9652c6..b164ecdd11 100644
--- a/libavcodec/bitstream.c
+++ b/libavcodec/bitstream.c
@@ -43,11 +43,7 @@ const uint8_t ff_log2_run[41]={
 
 void align_put_bits(PutBitContext *s)
 {
-#ifdef ALT_BITSTREAM_WRITER
-    put_bits(s,(  - s->index) & 7,0);
-#else
     put_bits(s,s->bit_left & 7,0);
-#endif
 }
 
 void ff_put_string(PutBitContext *pb, const char *string, int terminate_string)
diff --git a/libavcodec/h264.h b/libavcodec/h264.h
index 7bad91b215..4188ad922c 100644
--- a/libavcodec/h264.h
+++ b/libavcodec/h264.h
@@ -507,7 +507,7 @@ typedef struct H264Context{
     int          cabac_init_idc;
 
     /**
-     * @defgroup multithreading Members for slice based multithreading
+     * @name Members for slice based multithreading
      * @{
      */
     struct H264Context *thread_context[MAX_THREADS];
diff --git a/libavcodec/mpegvideo_enc.c b/libavcodec/mpegvideo_enc.c
index bb7fdc735f..888d8130e9 100644
--- a/libavcodec/mpegvideo_enc.c
+++ b/libavcodec/mpegvideo_enc.c
@@ -2170,9 +2170,7 @@ static int encode_thread(AVCodecContext *c, void *arg){
                         int d= 100 / s->avctx->error_rate;
                         if(r % d == 0){
                             current_packet_size=0;
-#ifndef ALT_BITSTREAM_WRITER
                             s->pb.buf_ptr= s->ptr_lastgob;
-#endif
                             assert(put_bits_ptr(&s->pb) == s->ptr_lastgob);
                         }
                     }
diff --git a/libavcodec/put_bits.h b/libavcodec/put_bits.h
index 79016912d5..a3fc5f16f1 100644
--- a/libavcodec/put_bits.h
+++ b/libavcodec/put_bits.h
@@ -36,19 +36,10 @@
 #include "mathops.h"
 #include "config.h"
 
-//#define ALT_BITSTREAM_WRITER
-//#define ALIGNED_BITSTREAM_WRITER
-
-/* buf and buf_end must be present and used by every alternative writer. */
 typedef struct PutBitContext {
-#ifdef ALT_BITSTREAM_WRITER
-    uint8_t *buf, *buf_end;
-    int index;
-#else
     uint32_t bit_buf;
     int bit_left;
     uint8_t *buf, *buf_ptr, *buf_end;
-#endif
     int size_in_bits;
 } PutBitContext;
 
@@ -68,15 +59,9 @@ static inline void init_put_bits(PutBitContext *s, uint8_t *buffer, int buffer_s
     s->size_in_bits= 8*buffer_size;
     s->buf = buffer;
     s->buf_end = s->buf + buffer_size;
-#ifdef ALT_BITSTREAM_WRITER
-    s->index=0;
-    ((uint32_t*)(s->buf))[0]=0;
-//    memset(buffer, 0, buffer_size);
-#else
     s->buf_ptr = s->buf;
     s->bit_left=32;
     s->bit_buf=0;
-#endif
 }
 
 /**
@@ -84,11 +69,7 @@ static inline void init_put_bits(PutBitContext *s, uint8_t *buffer, int buffer_s
  */
 static inline int put_bits_count(PutBitContext *s)
 {
-#ifdef ALT_BITSTREAM_WRITER
-    return s->index;
-#else
     return (s->buf_ptr - s->buf) * 8 + 32 - s->bit_left;
-#endif
 }
 
 /**
@@ -96,9 +77,6 @@ static inline int put_bits_count(PutBitContext *s)
  */
 static inline void flush_put_bits(PutBitContext *s)
 {
-#ifdef ALT_BITSTREAM_WRITER
-    align_put_bits(s);
-#else
 #ifndef BITSTREAM_WRITER_LE
     s->bit_buf<<= s->bit_left;
 #endif
@@ -115,10 +93,9 @@ static inline void flush_put_bits(PutBitContext *s)
     }
     s->bit_left=32;
     s->bit_buf=0;
-#endif
 }
 
-#if defined(ALT_BITSTREAM_WRITER) || defined(BITSTREAM_WRITER_LE)
+#ifdef BITSTREAM_WRITER_LE
 #define align_put_bits align_put_bits_unsupported_here
 #define ff_put_string ff_put_string_unsupported_here
 #define ff_copy_bits ff_copy_bits_unsupported_here
@@ -148,7 +125,6 @@ void ff_copy_bits(PutBitContext *pb, const uint8_t *src, int length);
  * Use put_bits32 to write 32 bits.
  */
 static inline void put_bits(PutBitContext *s, int n, unsigned int value)
-#ifndef ALT_BITSTREAM_WRITER
 {
     unsigned int bit_buf;
     int bit_left;
@@ -164,12 +140,7 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
 #ifdef BITSTREAM_WRITER_LE
     bit_buf |= value << (32 - bit_left);
     if (n >= bit_left) {
-#if !HAVE_FAST_UNALIGNED
-        if (3 & (intptr_t) s->buf_ptr) {
-            AV_WL32(s->buf_ptr, bit_buf);
-        } else
-#endif
-        *(uint32_t *)s->buf_ptr = av_le2ne32(bit_buf);
+        AV_WL32(s->buf_ptr, bit_buf);
         s->buf_ptr+=4;
         bit_buf = (bit_left==32)?0:value >> bit_left;
         bit_left+=32;
@@ -182,12 +153,7 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
     } else {
         bit_buf<<=bit_left;
         bit_buf |= value >> (n - bit_left);
-#if !HAVE_FAST_UNALIGNED
-        if (3 & (intptr_t) s->buf_ptr) {
-            AV_WB32(s->buf_ptr, bit_buf);
-        } else
-#endif
-        *(uint32_t *)s->buf_ptr = av_be2ne32(bit_buf);
+        AV_WB32(s->buf_ptr, bit_buf);
         //printf("bitbuf = %08x\n", bit_buf);
         s->buf_ptr+=4;
         bit_left+=32 - n;
@@ -198,70 +164,6 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
     s->bit_buf = bit_buf;
     s->bit_left = bit_left;
 }
-#else  /* ALT_BITSTREAM_WRITER defined */
-{
-#    ifdef ALIGNED_BITSTREAM_WRITER
-#        if ARCH_X86
-    __asm__ volatile(
-        "movl %0, %%ecx                 \n\t"
-        "xorl %%eax, %%eax              \n\t"
-        "shrdl %%cl, %1, %%eax          \n\t"
-        "shrl %%cl, %1                  \n\t"
-        "movl %0, %%ecx                 \n\t"
-        "shrl $3, %%ecx                 \n\t"
-        "andl $0xFFFFFFFC, %%ecx        \n\t"
-        "bswapl %1                      \n\t"
-        "orl %1, (%2, %%ecx)            \n\t"
-        "bswapl %%eax                   \n\t"
-        "addl %3, %0                    \n\t"
-        "movl %%eax, 4(%2, %%ecx)       \n\t"
-        : "=&r" (s->index), "=&r" (value)
-        : "r" (s->buf), "r" (n), "0" (s->index), "1" (value<<(-n))
-        : "%eax", "%ecx"
-    );
-#        else
-    int index= s->index;
-    uint32_t *ptr= ((uint32_t *)s->buf)+(index>>5);
-
-    value<<= 32-n;
-
-    ptr[0] |= av_be2ne32(value>>(index&31));
-    ptr[1]  = av_be2ne32(value<<(32-(index&31)));
-//if(n>24) printf("%d %d\n", n, value);
-    index+= n;
-    s->index= index;
-#        endif
-#    else //ALIGNED_BITSTREAM_WRITER
-#        if ARCH_X86
-    __asm__ volatile(
-        "movl $7, %%ecx                 \n\t"
-        "andl %0, %%ecx                 \n\t"
-        "addl %3, %%ecx                 \n\t"
-        "negl %%ecx                     \n\t"
-        "shll %%cl, %1                  \n\t"
-        "bswapl %1                      \n\t"
-        "movl %0, %%ecx                 \n\t"
-        "shrl $3, %%ecx                 \n\t"
-        "orl %1, (%%ecx, %2)            \n\t"
-        "addl %3, %0                    \n\t"
-        "movl $0, 4(%%ecx, %2)          \n\t"
-        : "=&r" (s->index), "=&r" (value)
-        : "r" (s->buf), "r" (n), "0" (s->index), "1" (value)
-        : "%ecx"
-    );
-#        else
-    int index= s->index;
-    uint32_t *ptr= (uint32_t*)(((uint8_t *)s->buf)+(index>>3));
-
-    ptr[0] |= av_be2ne32(value<<(32-n-(index&7) ));
-    ptr[1] = 0;
-//if(n>24) printf("%d %d\n", n, value);
-    index+= n;
-    s->index= index;
-#        endif
-#    endif //!ALIGNED_BITSTREAM_WRITER
-}
-#endif
 
 static inline void put_sbits(PutBitContext *pb, int n, int32_t value)
 {
@@ -292,11 +194,7 @@ static void av_unused put_bits32(PutBitContext *s, uint32_t value)
  */
 static inline uint8_t* put_bits_ptr(PutBitContext *s)
 {
-#ifdef ALT_BITSTREAM_WRITER
-        return s->buf + (s->index>>3);
-#else
         return s->buf_ptr;
-#endif
 }
 
 /**
@@ -306,13 +204,8 @@ static inline uint8_t* put_bits_ptr(PutBitContext *s)
 static inline void skip_put_bytes(PutBitContext *s, int n)
 {
         assert((put_bits_count(s)&7)==0);
-#ifdef ALT_BITSTREAM_WRITER
-        FIXME may need some cleaning of the buffer
-        s->index += n<<3;
-#else
         assert(s->bit_left==32);
         s->buf_ptr += n;
-#endif
 }
 
 /**
@@ -322,13 +215,9 @@ static inline void skip_put_bytes(PutBitContext *s, int n)
  */
 static inline void skip_put_bits(PutBitContext *s, int n)
 {
-#ifdef ALT_BITSTREAM_WRITER
-    s->index += n;
-#else
     s->bit_left -= n;
     s->buf_ptr-= 4*(s->bit_left>>5);
     s->bit_left &= 31;
-#endif
 }
 
 /**
diff --git a/libavcodec/qcelpdata.h b/libavcodec/qcelpdata.h
index 82cc61d7ea..e71ee9fdb7 100644
--- a/libavcodec/qcelpdata.h
+++ b/libavcodec/qcelpdata.h
@@ -38,14 +38,14 @@
  * QCELP unpacked data frame
  */
 typedef struct {
-/// @defgroup qcelp_codebook_parameters QCELP excitation codebook parameters
+/// @name QCELP excitation codebook parameters
 /// @{
     uint8_t cbsign[16]; ///!< sign of the codebook gain for each codebook subframe
     uint8_t cbgain[16]; ///!< unsigned codebook gain for each codebook subframe
     uint8_t cindex[16]; ///!< codebook index for each codebook subframe
 /// @}
 
-/// @defgroup qcelp_pitch_parameters QCELP pitch prediction parameters
+/// @name QCELP pitch prediction parameters
 /// @{
     uint8_t plag[4];    ///!< pitch lag for each pitch subframe
     uint8_t pfrac[4];   ///!< fractional pitch lag for each pitch subframe
diff --git a/libavcodec/rv34.c b/libavcodec/rv34.c
index ae9ae1084f..241132e092 100644
--- a/libavcodec/rv34.c
+++ b/libavcodec/rv34.c
@@ -62,8 +62,10 @@ static const int rv34_mb_type_to_lavc[12] = {
 
 static RV34VLC intra_vlcs[NUM_INTRA_TABLES], inter_vlcs[NUM_INTER_TABLES];
 
+static int rv34_decode_mv(RV34DecContext *r, int block_type);
+
 /**
- * @defgroup vlc RV30/40 VLC generating functions
+ * @name RV30/40 VLC generating functions
  * @{
  */
 
@@ -171,7 +173,7 @@ static av_cold void rv34_init_tables(void)
 
 
 /**
- * @defgroup transform RV30/40 inverse transform functions
+ * @name RV30/40 inverse transform functions
  * @{
  */
 
@@ -246,7 +248,7 @@ static void rv34_inv_transform_noround(DCTELEM *block){
 
 
 /**
- * @defgroup block RV30/40 4x4 block decoding functions
+ * @name RV30/40 4x4 block decoding functions
  * @{
  */
 
@@ -393,7 +395,7 @@ static inline void rv34_dequant4x4_16x16(DCTELEM *block, int Qdc, int Q)
 
 
 /**
- * @defgroup rv3040_bitstream RV30/40 bitstream parsing
+ * @name RV30/40 bitstream parsing
  * @{
  */
 
@@ -432,10 +434,76 @@ static inline int rv34_decode_dquant(GetBitContext *gb, int quant)
         return get_bits(gb, 5);
 }
 
+/**
+ * Decode macroblock header and return CBP in case of success, -1 otherwise.
+ */
+static int rv34_decode_mb_header(RV34DecContext *r, int8_t *intra_types)
+{
+    MpegEncContext *s = &r->s;
+    GetBitContext *gb = &s->gb;
+    int mb_pos = s->mb_x + s->mb_y * s->mb_stride;
+    int i, t;
+
+    if(!r->si.type){
+        r->is16 = get_bits1(gb);
+        if(!r->is16 && !r->rv30){
+            if(!get_bits1(gb))
+                av_log(s->avctx, AV_LOG_ERROR, "Need DQUANT\n");
+        }
+        s->current_picture_ptr->mb_type[mb_pos] = r->is16 ? MB_TYPE_INTRA16x16 : MB_TYPE_INTRA;
+        r->block_type = r->is16 ? RV34_MB_TYPE_INTRA16x16 : RV34_MB_TYPE_INTRA;
+    }else{
+        r->block_type = r->decode_mb_info(r);
+        if(r->block_type == -1)
+            return -1;
+        s->current_picture_ptr->mb_type[mb_pos] = rv34_mb_type_to_lavc[r->block_type];
+        r->mb_type[mb_pos] = r->block_type;
+        if(r->block_type == RV34_MB_SKIP){
+            if(s->pict_type == AV_PICTURE_TYPE_P)
+                r->mb_type[mb_pos] = RV34_MB_P_16x16;
+            if(s->pict_type == AV_PICTURE_TYPE_B)
+                r->mb_type[mb_pos] = RV34_MB_B_DIRECT;
+        }
+        r->is16 = !!IS_INTRA16x16(s->current_picture_ptr->mb_type[mb_pos]);
+        rv34_decode_mv(r, r->block_type);
+        if(r->block_type == RV34_MB_SKIP){
+            fill_rectangle(intra_types, 4, 4, r->intra_types_stride, 0, sizeof(intra_types[0]));
+            return 0;
+        }
+        r->chroma_vlc = 1;
+        r->luma_vlc   = 0;
+    }
+    if(IS_INTRA(s->current_picture_ptr->mb_type[mb_pos])){
+        if(r->is16){
+            t = get_bits(gb, 2);
+            fill_rectangle(intra_types, 4, 4, r->intra_types_stride, t, sizeof(intra_types[0]));
+            r->luma_vlc   = 2;
+        }else{
+            if(r->decode_intra_types(r, gb, intra_types) < 0)
+                return -1;
+            r->luma_vlc   = 1;
+        }
+        r->chroma_vlc = 0;
+        r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 0);
+    }else{
+        for(i = 0; i < 16; i++)
+            intra_types[(i & 3) + (i>>2) * r->intra_types_stride] = 0;
+        r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 1);
+        if(r->mb_type[mb_pos] == RV34_MB_P_MIX16x16){
+            r->is16 = 1;
+            r->chroma_vlc = 1;
+            r->luma_vlc   = 2;
+            r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 0);
+        }
+    }
+
+    return rv34_decode_cbp(gb, r->cur_vlcs, r->is16);
+}
+
 /** @} */ //bitstream functions
 
 /**
- * @defgroup mv motion vector related code (prediction, reconstruction, motion compensation)
+ * @name motion vector related code (prediction, reconstruction, motion compensation)
  * @{
  */
 
@@ -885,7 +953,7 @@ static int rv34_decode_mv(RV34DecContext *r, int block_type)
 /** @} */ // mv group
 
 /**
- * @defgroup recons Macroblock reconstruction functions
+ * @name Macroblock reconstruction functions
  * @{
  */
 /** mapping of RV30/40 intra prediction types to standard H.264 types */
@@ -1027,79 +1095,6 @@ static void rv34_output_macroblock(RV34DecContext *r, int8_t *intra_types, int c
     }
 }
 
-/** @} */ // recons group
-
-/**
- * @addtogroup bitstream
- * Decode macroblock header and return CBP in case of success, -1 otherwise.
- */
-static int rv34_decode_mb_header(RV34DecContext *r, int8_t *intra_types)
-{
-    MpegEncContext *s = &r->s;
-    GetBitContext *gb = &s->gb;
-    int mb_pos = s->mb_x + s->mb_y * s->mb_stride;
-    int i, t;
-
-    if(!r->si.type){
-        r->is16 = get_bits1(gb);
-        if(!r->is16 && !r->rv30){
-            if(!get_bits1(gb))
-                av_log(s->avctx, AV_LOG_ERROR, "Need DQUANT\n");
-        }
-        s->current_picture_ptr->mb_type[mb_pos] = r->is16 ? MB_TYPE_INTRA16x16 : MB_TYPE_INTRA;
-        r->block_type = r->is16 ? RV34_MB_TYPE_INTRA16x16 : RV34_MB_TYPE_INTRA;
-    }else{
-        r->block_type = r->decode_mb_info(r);
-        if(r->block_type == -1)
-            return -1;
-        s->current_picture_ptr->mb_type[mb_pos] = rv34_mb_type_to_lavc[r->block_type];
-        r->mb_type[mb_pos] = r->block_type;
-        if(r->block_type == RV34_MB_SKIP){
-            if(s->pict_type == AV_PICTURE_TYPE_P)
-                r->mb_type[mb_pos] = RV34_MB_P_16x16;
-            if(s->pict_type == AV_PICTURE_TYPE_B)
-                r->mb_type[mb_pos] = RV34_MB_B_DIRECT;
-        }
-        r->is16 = !!IS_INTRA16x16(s->current_picture_ptr->mb_type[mb_pos]);
-        rv34_decode_mv(r, r->block_type);
-        if(r->block_type == RV34_MB_SKIP){
-            fill_rectangle(intra_types, 4, 4, r->intra_types_stride, 0, sizeof(intra_types[0]));
-            return 0;
-        }
-        r->chroma_vlc = 1;
-        r->luma_vlc   = 0;
-    }
-    if(IS_INTRA(s->current_picture_ptr->mb_type[mb_pos])){
-        if(r->is16){
-            t = get_bits(gb, 2);
-            fill_rectangle(intra_types, 4, 4, r->intra_types_stride, t, sizeof(intra_types[0]));
-            r->luma_vlc   = 2;
-        }else{
-            if(r->decode_intra_types(r, gb, intra_types) < 0)
-                return -1;
-            r->luma_vlc   = 1;
-        }
-        r->chroma_vlc = 0;
-        r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 0);
-    }else{
-        for(i = 0; i < 16; i++)
-            intra_types[(i & 3) + (i>>2) * r->intra_types_stride] = 0;
-        r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 1);
-        if(r->mb_type[mb_pos] == RV34_MB_P_MIX16x16){
-            r->is16 = 1;
-            r->chroma_vlc = 1;
-            r->luma_vlc   = 2;
-            r->cur_vlcs = choose_vlc_set(r->si.quant, r->si.vlc_set, 0);
-        }
-    }
-
-    return rv34_decode_cbp(gb, r->cur_vlcs, r->is16);
-}
-
-/**
- * @addtogroup recons
- * @{
- */
 /**
  * mask for retrieving all bits in coded block pattern
  * corresponding to one 8x8 block
@@ -1109,6 +1104,8 @@ static int rv34_decode_mb_header(RV34DecContext *r, int8_t *intra_types)
 #define U_CBP_MASK 0x0F0000
 #define V_CBP_MASK 0xF00000
 
+/** @} */ // recons group
+
 
 static void rv34_apply_differences(RV34DecContext *r, int cbp)
 {
diff --git a/libavcodec/rv40data.h b/libavcodec/rv40data.h
index 7912919ba1..436afa84e2 100644
--- a/libavcodec/rv40data.h
+++ b/libavcodec/rv40data.h
@@ -65,7 +65,7 @@ static const uint8_t rv40_luma_dc_quant[2][32] = {
 };
 
 /**
- * @defgroup loopfilter coefficients used by the RV40 loop filter
+ * @name Coefficients used by the RV40 loop filter
  * @{
  */
 /**
diff --git a/libavcodec/sbr.h b/libavcodec/sbr.h
index 82a996fdc6..69c847ac15 100644
--- a/libavcodec/sbr.h
+++ b/libavcodec/sbr.h
@@ -42,7 +42,7 @@ typedef struct {
     uint8_t bs_xover_band;
 
     /**
-     * @defgroup bs_header_extra_1     Variables associated with bs_header_extra_1
+     * @name Variables associated with bs_header_extra_1
      * @{
      */
     uint8_t bs_freq_scale;
@@ -58,7 +58,7 @@ typedef struct {
  */
 typedef struct {
     /**
-     * @defgroup aac_bitstream     Main bitstream data variables
+     * @name Main bitstream data variables
      * @{
      */
     unsigned           bs_frame_class;
@@ -74,7 +74,7 @@ typedef struct {
     /** @} */
 
     /**
-     * @defgroup state         State variables
+     * @name State variables
      * @{
      */
     DECLARE_ALIGNED(16, float, synthesis_filterbank_samples)[SBR_SYNTHESIS_BUF_SIZE];
@@ -116,7 +116,7 @@ typedef struct {
     SpectrumParameters spectrum_params;
     int                bs_amp_res_header;
     /**
-     * @defgroup bs_header_extra_2     variables associated with bs_header_extra_2
+     * @name Variables associated with bs_header_extra_2
      * @{
      */
     unsigned           bs_limiter_bands;
diff --git a/libavcodec/tableprint.h b/libavcodec/tableprint.h
index d3e4dd956f..e126a72afb 100644
--- a/libavcodec/tableprint.h
+++ b/libavcodec/tableprint.h
@@ -56,7 +56,7 @@ void write_##type##_2d_array(const void *arg, int len, int len2)\
 }
 
 /**
- * @defgroup printfuncs Predefined functions for printing tables
+ * @name Predefined functions for printing tables
  *
  * \{
  */
diff --git a/libavcodec/vc1.c b/libavcodec/vc1.c
index e062a35cc1..d6b48b46d0 100644
--- a/libavcodec/vc1.c
+++ b/libavcodec/vc1.c
@@ -40,7 +40,7 @@
 
 /***********************************************************************/
 /**
- * @defgroup vc1bitplane VC-1 Bitplane decoding
+ * @name VC-1 Bitplane decoding
  * @see 8.7, p56
  * @{
  */
diff --git a/libavcodec/vc1dec.c b/libavcodec/vc1dec.c
index 852c874ee1..e82fba2fcb 100644
--- a/libavcodec/vc1dec.c
+++ b/libavcodec/vc1dec.c
@@ -137,7 +137,7 @@ static int vc1_init_common(VC1Context *v)
 
 /***********************************************************************/
 /**
- * @defgroup vc1bitplane VC-1 Bitplane decoding
+ * @name VC-1 Bitplane decoding
  * @see 8.7, p56
  * @{
  */
@@ -774,7 +774,7 @@ static void vc1_mc_4mv_chroma(VC1Context *v)
 
 /***********************************************************************/
 /**
- * @defgroup vc1block VC-1 Block-level functions
+ * @name VC-1 Block-level functions
  * @see 7.1.4, p91 and 8.1.1.7, p(1)04
  * @{
  */
@@ -1512,7 +1512,7 @@ static inline int vc1_pred_dc(MpegEncContext *s, int overlap, int pq, int n,
 /** @} */ // Block group
 
 /**
- * @defgroup vc1_std_mb VC1 Macroblock-level functions in Simple/Main Profiles
+ * @name VC1 Macroblock-level functions in Simple/Main Profiles
  * @see 7.1.4, p91 and 8.1.1.7, p(1)04
  * @{
  */
diff --git a/libavcodec/wmavoice.c b/libavcodec/wmavoice.c
index 3604eac782..002c529a30 100644
--- a/libavcodec/wmavoice.c
+++ b/libavcodec/wmavoice.c
@@ -128,9 +128,7 @@ static const struct frame_type_desc {
  */
 typedef struct {
     /**
-     * @defgroup struct_global Global values
-     * Global values, specified in the stream header / extradata or used
-     * all over.
+     * @name Global values specified in the stream header / extradata or used all over.
      * @{
      */
     GetBitContext gb;             ///< packet bitreader. During decoder init,
@@ -182,8 +180,9 @@ typedef struct {
 
     /**
      * @}
-     * @defgroup struct_packet Packet values
-     * Packet values, specified in the packet header or related to a packet.
+     *
+     * @name Packet values specified in the packet header or related to a packet.
+     *
      * A packet is considered to be a single unit of data provided to this
      * decoder by the demuxer.
      * @{
@@ -213,7 +212,8 @@ typedef struct {
 
     /**
      * @}
-     * @defgroup struct_frame Frame and superframe values
+     *
+     * @name Frame and superframe values
      * Superframe and frame data - these can change from frame to frame,
      * although some of them do in that case serve as a cache / history for
      * the next frame or superframe.
@@ -256,7 +256,9 @@ typedef struct {
     float synth_history[MAX_LSPS]; ///< see #excitation_history
     /**
      * @}
-     * @defgroup post_filter Postfilter values
+     *
+     * @name Postfilter values
+     *
      * Variables used for postfilter implementation, mostly history for
      * smoothing and so on, and context variables for FFT/iFFT.
      * @{
@@ -432,7 +434,7 @@ static av_cold int wmavoice_decode_init(AVCodecContext *ctx)
 }
 
 /**
- * @defgroup postfilter Postfilter functions
+ * @name Postfilter functions
  * Postfilter functions (gain control, wiener denoise filter, DC filter,
  * kalman smoothening, plus surrounding code to wrap it)
  * @{
@@ -825,7 +827,7 @@ static void dequant_lsps(double *lsps, int num,
 }
 
 /**
- * @defgroup lsp_dequant LSP dequantization routines
+ * @name LSP dequantization routines
  * LSP dequantization routines, for 10/16LSPs and independent/residual coding.
  * @note we assume enough bits are available, caller should check.
  * lsp10i() consumes 24 bits; lsp10r() consumes an additional 24 bits;
@@ -969,7 +971,7 @@ static void dequant_lsp16r(GetBitContext *gb,
 
 /**
  * @}
- * @defgroup aw Pitch-adaptive window coding functions
+ * @name Pitch-adaptive window coding functions
  * The next few functions are for pitch-adaptive window coding.
  * @{
  */