Commit 18188c24 authored by Christophe Massiot's avatar Christophe Massiot

* 32-bit aligned bitstream (not that fast).

parent 647cca0e
......@@ -3,7 +3,7 @@
* Collection of useful common types and macros definitions
*****************************************************************************
* Copyright (C) 1998, 1999, 2000 VideoLAN
* $Id: common.h,v 1.25 2001/01/18 05:13:22 sam Exp $
* $Id: common.h,v 1.26 2001/01/18 17:40:06 massiot Exp $
*
* Authors: Samuel Hocevar <sam@via.ecp.fr>
* Vincent Seguin <seguin@via.ecp.fr>
......@@ -155,31 +155,6 @@ struct macroblock_s;
#define MIN(a, b) ( ((a) < (b)) ? (a) : (b) )
#endif
/*
* This is stolen from the livid source who stole it from the kernel
*/
#if defined(SYS_BEOS)
# define swab32(x) B_BENDIAN_TO_HOST_INT32(x)
#else
# ifdef WORDS_BIG_ENDIAN
# define swab32(x) (x)
# else
# if defined (HAVE_X86_BSWAP)
static __inline__ const u32 __i386_swab32( u32 x )
{
__asm__("bswap %0" : "=r" (x) : "0" (x));
return x;
}
# define swab32(x) __i386_swab32(x)
# else
# define swab32(x) \
( ( (u32)(((u8*)&x)[0]) << 24 ) | ( (u32)(((u8*)&x)[1]) << 16 ) |\
( (u32)(((u8*)&x)[2]) << 8 ) | ( (u32)(((u8*)&x)[3])) )
# endif
# endif
#endif
/* MSB (big endian)/LSB (little endian) conversions - network order is always
* MSB, and should be used for both network communications and files. Note that
* byte orders other than little and big endians are not supported, but only
......@@ -204,6 +179,7 @@ static __inline__ const u32 __i386_swab32( u32 x )
#endif
/* Macros with automatic casts */
#define U32_AT(p) ( swab32 ( *( (u32 *)(p) ) ) )
#define U16_AT(p) ( ntohs ( *( (u16 *)(p) ) ) )
#define U64_AT(p) ( ntoh64 ( *( (u64 *)(p) ) ) )
#define U32_AT(p) ( ntoh32 ( *( (u32 *)(p) ) ) )
#define U16_AT(p) ( ntoh16 ( *( (u16 *)(p) ) ) )
......@@ -2,7 +2,7 @@
* input_ext-dec.h: structures exported to the VideoLAN decoders
*****************************************************************************
* Copyright (C) 1999, 2000 VideoLAN
* $Id: input_ext-dec.h,v 1.14 2001/01/13 12:57:19 sam Exp $
* $Id: input_ext-dec.h,v 1.15 2001/01/18 17:40:06 massiot Exp $
*
* Authors:
*
......@@ -113,7 +113,7 @@ typedef struct decoder_fifo_s
* This type describes a bit fifo used to store bits while working with the
* input stream at the bit level.
*****************************************************************************/
typedef u32 WORD_TYPE; /* only u32 is supported at the moment */
typedef u32 WORD_TYPE;
typedef struct bit_fifo_s
{
......@@ -172,123 +172,88 @@ typedef struct bit_stream_s
*****************************************************************************/
/*
* Philosophy of the first implementation : the bit buffer is first filled by
* NeedBits, then the buffer can be read via p_bit_stream->fifo.buffer, and
* unnecessary bits are dumped with a DumpBits() call.
* DISCUSSION : How to use the bit_stream structures
*
* sizeof(WORD_TYPE) (usually 32) bits are read at the same time, thus
* minimizing the number of p_byte changes.
* Bits are read via GetBits() or ShowBits.
*
* XXX : Be aware that if, in the forthcoming functions, i_bits > 24,
* the data have to be already aligned on an 8-bit boundary, or wrong
* results will be returned. Use RealignBits() if unsure.
*/
#if (WORD_TYPE == u32)
# define WORD_AT U32_AT
#elif (WORD_TYPE == u64)
# define WORD_AT U64_AT
#else
# error Unsupported WORD_TYPE
#endif
/*****************************************************************************
* GetByte : reads the next byte in the input stream (PRIVATE)
* Protoypes from input_ext-dec.c
*****************************************************************************/
static __inline__ byte_t _GetByte( bit_stream_t * p_bit_stream )
{
/* Are there some bytes left in the current data packet ? */
/* could change this test to have a if (! (bytes--)) instead */
if ( p_bit_stream->p_byte >= p_bit_stream->p_end )
{
/* no, switch to next data packet */
p_bit_stream->pf_next_data_packet( p_bit_stream );
}
return( *(p_bit_stream->p_byte++) );
}
u32 UnalignedShowBits( struct bit_stream_s *, unsigned int );
void UnalignedRemoveBits( struct bit_stream_s * );
u32 UnalignedGetBits( struct bit_stream_s *, unsigned int );
/*****************************************************************************
* NeedBits : reads i_bits new bits in the bit stream and stores them in the
* bit buffer
*****************************************************************************
* - i_bits must be less or equal 32 !
* - There is something important to notice with that function : if the number
* of bits available in the bit buffer when calling NeedBits() is greater than
* 24 (i_available > 24) but less than the number of needed bits
* (i_available < i_bits), the byte returned by GetByte() will be shifted with
* a negative value and the number of bits available in the bit buffer will be
* set to more than 32 !
* AlignWord : fill in the bit buffer so that the byte pointer be aligned
* on a word boundary (XXX: there must be at least sizeof(WORD_TYPE) - 1
* empty bytes in the bit buffer)
*****************************************************************************/
static __inline__ void NeedBits( bit_stream_t * p_bit_stream, int i_bits )
static __inline__ void AlignWord( bit_stream_t * p_bit_stream )
{
while ( p_bit_stream->fifo.i_available < i_bits )
while( (p_bit_stream->p_byte - p_bit_stream->p_data->p_buffer)
& (sizeof(WORD_TYPE) - 1) )
{
p_bit_stream->fifo.buffer |= ((WORD_TYPE)_GetByte( p_bit_stream ))
<< (sizeof(WORD_TYPE) - 8
- p_bit_stream->fifo.i_available);
p_bit_stream->fifo.i_available += 8;
if( p_bit_stream->p_byte < p_bit_stream->p_end )
{
p_bit_stream->fifo.buffer |= *(p_bit_stream->p_byte++)
<< (8 * sizeof(WORD_TYPE) - 8
- p_bit_stream->fifo.i_available);
p_bit_stream->fifo.i_available += 8;
}
else
{
p_bit_stream->pf_next_data_packet( p_bit_stream );
p_bit_stream->fifo.buffer |= *(p_bit_stream->p_byte++)
<< (8 * sizeof(WORD_TYPE) - 8
- p_bit_stream->fifo.i_available);
p_bit_stream->fifo.i_available += 8;
}
}
}
/*****************************************************************************
* DumpBits : removes i_bits bits from the bit buffer
*****************************************************************************
* - i_bits <= i_available
* - i_bits < 32 (because (u32 << 32) <=> (u32 = u32))
*****************************************************************************/
static __inline__ void DumpBits( bit_stream_t * p_bit_stream, int i_bits )
{
p_bit_stream->fifo.buffer <<= i_bits;
p_bit_stream->fifo.i_available -= i_bits;
}
/*
* Philosophy of the second implementation : WORD_LENGTH (usually 32) bits
* are read at the same time, thus minimizing the number of p_byte changes.
* Bits are read via GetBits() or ShowBits. This is slightly faster. Be
* aware that if, in the forthcoming functions, i_bits > 24, the data have to
* be already aligned on an 8-bit boundary, or wrong results will be
* returned.
*/
#if (WORD_TYPE != u32)
# error Not supported word
#endif
/*****************************************************************************
* ShowBits : return i_bits bits from the bit stream
*****************************************************************************/
static __inline__ WORD_TYPE _ShowWord( bit_stream_t * p_bit_stream )
{
if( p_bit_stream->p_byte <= p_bit_stream->p_end - sizeof(WORD_TYPE) )
{
return( swab32( *((WORD_TYPE *)p_bit_stream->p_byte) ) );
}
p_bit_stream->pf_next_data_packet( p_bit_stream );
return( swab32( *((WORD_TYPE *)p_bit_stream->p_byte) ) );
}
static __inline__ WORD_TYPE ShowBits( bit_stream_t * p_bit_stream, int i_bits )
static __inline__ u32 ShowBits( bit_stream_t * p_bit_stream,
unsigned int i_bits )
{
if( p_bit_stream->fifo.i_available >= i_bits )
{
return( p_bit_stream->fifo.buffer >> (8 * sizeof(WORD_TYPE) - i_bits) );
}
return( (p_bit_stream->fifo.buffer |
(_ShowWord( p_bit_stream ) >> p_bit_stream->fifo.i_available))
>> (8 * sizeof(WORD_TYPE) - i_bits) );
}
/*****************************************************************************
* GetWord : returns the next word to be read (PRIVATE)
*****************************************************************************/
static __inline__ WORD_TYPE _GetWord( bit_stream_t * p_bit_stream )
{
if( p_bit_stream->p_byte <= p_bit_stream->p_end - sizeof(WORD_TYPE) )
{
return( swab32( *(((WORD_TYPE *)p_bit_stream->p_byte)++) ) );
}
else
{
p_bit_stream->pf_next_data_packet( p_bit_stream );
return( swab32( *(((WORD_TYPE *)p_bit_stream->p_byte)++) ) );
return( (p_bit_stream->fifo.buffer |
(WORD_AT( p_bit_stream->p_byte )
>> p_bit_stream->fifo.i_available))
>> (8 * sizeof(WORD_TYPE) - i_bits) );
}
return UnalignedShowBits( p_bit_stream, i_bits );
}
/*****************************************************************************
* RemoveBits : removes i_bits bits from the bit buffer
* XXX: do not use for 32 bits, see RemoveBits32
*****************************************************************************/
static __inline__ void RemoveBits( bit_stream_t * p_bit_stream, int i_bits )
static __inline__ void RemoveBits( bit_stream_t * p_bit_stream,
unsigned int i_bits )
{
p_bit_stream->fifo.i_available -= i_bits;
......@@ -297,78 +262,122 @@ static __inline__ void RemoveBits( bit_stream_t * p_bit_stream, int i_bits )
p_bit_stream->fifo.buffer <<= i_bits;
return;
}
p_bit_stream->fifo.buffer = _GetWord( p_bit_stream )
<< ( -p_bit_stream->fifo.i_available );
p_bit_stream->fifo.i_available += sizeof(WORD_TYPE) * 8;
if( p_bit_stream->p_byte <= p_bit_stream->p_end - sizeof(WORD_TYPE) )
{
p_bit_stream->fifo.buffer = WORD_AT( p_bit_stream->p_byte )
<< ( -p_bit_stream->fifo.i_available );
((WORD_TYPE *)p_bit_stream->p_byte)++;
p_bit_stream->fifo.i_available += sizeof(WORD_TYPE) * 8;
return;
}
UnalignedRemoveBits( p_bit_stream );
}
/*****************************************************************************
* RemoveBits32 : removes 32 bits from the bit buffer (and as a side effect,
* refill it)
*****************************************************************************/
#if (WORD_TYPE == u32)
static __inline__ void RemoveBits32( bit_stream_t * p_bit_stream )
{
if( p_bit_stream->fifo.i_available )
if( p_bit_stream->p_byte <= p_bit_stream->p_end - sizeof(WORD_TYPE) )
{
p_bit_stream->fifo.buffer = _GetWord( p_bit_stream )
if( p_bit_stream->fifo.i_available )
{
p_bit_stream->fifo.buffer = WORD_AT( p_bit_stream->p_byte )
<< (32 - p_bit_stream->fifo.i_available);
((WORD_TYPE *)p_bit_stream->p_byte)++;
return;
}
((WORD_TYPE *)p_bit_stream->p_byte)++;
return;
}
else
{
_GetWord( p_bit_stream );
}
p_bit_stream->fifo.i_available -= 32;
UnalignedRemoveBits( p_bit_stream );
}
#else
# define RemoveBits32( p_bit_stream ) RemoveBits( p_bit_stream, 32 )
#endif
/*****************************************************************************
* GetBits : returns i_bits bits from the bit stream and removes them
* XXX: do not use for 32 bits, see GetBits32
*****************************************************************************/
static __inline__ WORD_TYPE GetBits( bit_stream_t * p_bit_stream, int i_bits )
static __inline__ u32 GetBits( bit_stream_t * p_bit_stream,
unsigned int i_bits )
{
u32 i_result;
p_bit_stream->fifo.i_available -= i_bits;
if( p_bit_stream->fifo.i_available >= 0 )
if( p_bit_stream->fifo.i_available >= i_bits )
{
i_result = p_bit_stream->fifo.buffer >> (8 * sizeof(WORD_TYPE) - i_bits);
p_bit_stream->fifo.i_available -= i_bits;
i_result = p_bit_stream->fifo.buffer
>> (8 * sizeof(WORD_TYPE) - i_bits);
p_bit_stream->fifo.buffer <<= i_bits;
return( i_result );
}
i_result = p_bit_stream->fifo.buffer >> (8 * sizeof(WORD_TYPE) - i_bits);
p_bit_stream->fifo.buffer = _GetWord( p_bit_stream );
i_result |= p_bit_stream->fifo.buffer
>> (8 * sizeof(WORD_TYPE)
if( p_bit_stream->p_byte <= p_bit_stream->p_end - sizeof(WORD_TYPE) )
{
p_bit_stream->fifo.i_available -= i_bits;
i_result = p_bit_stream->fifo.buffer
>> (8 * sizeof(WORD_TYPE) - i_bits);
p_bit_stream->fifo.buffer = WORD_AT( p_bit_stream->p_byte );
((WORD_TYPE *)p_bit_stream->p_byte)++;
i_result |= p_bit_stream->fifo.buffer
>> (8 * sizeof(WORD_TYPE)
+ p_bit_stream->fifo.i_available);
p_bit_stream->fifo.buffer <<= ( -p_bit_stream->fifo.i_available );
p_bit_stream->fifo.i_available += sizeof(WORD_TYPE) * 8;
p_bit_stream->fifo.buffer <<= ( -p_bit_stream->fifo.i_available );
p_bit_stream->fifo.i_available += sizeof(WORD_TYPE) * 8;
return( i_result );
}
return( i_result );
return UnalignedGetBits( p_bit_stream, i_bits );
}
/*****************************************************************************
* GetBits32 : returns 32 bits from the bit stream and removes them
*****************************************************************************/
static __inline__ WORD_TYPE GetBits32( bit_stream_t * p_bit_stream )
#if (WORD_TYPE == u32)
static __inline__ u32 GetBits32( bit_stream_t * p_bit_stream )
{
WORD_TYPE i_result;
u32 i_result;
if( p_bit_stream->fifo.i_available )
if( p_bit_stream->fifo.i_available == 32 )
{
p_bit_stream->fifo.i_available = 0;
i_result = p_bit_stream->fifo.buffer;
p_bit_stream->fifo.buffer = _GetWord( p_bit_stream );
i_result |= p_bit_stream->fifo.buffer
>> (p_bit_stream->fifo.i_available);
p_bit_stream->fifo.buffer <<= (8 * sizeof(WORD_TYPE)
- p_bit_stream->fifo.i_available);
p_bit_stream->fifo.buffer = 0;
return( i_result );
}
else
if( p_bit_stream->p_byte <= p_bit_stream->p_end - sizeof(WORD_TYPE) )
{
return( _GetWord( p_bit_stream ) );
if( p_bit_stream->fifo.i_available )
{
i_result = p_bit_stream->fifo.buffer;
p_bit_stream->fifo.buffer = WORD_AT( p_bit_stream->p_byte );
((WORD_TYPE *)p_bit_stream->p_byte)++;
i_result |= p_bit_stream->fifo.buffer
>> (p_bit_stream->fifo.i_available);
p_bit_stream->fifo.buffer <<= (32 - p_bit_stream->fifo.i_available);
return( i_result );
}
i_result = WORD_AT( p_bit_stream->p_byte );
((WORD_TYPE *)p_bit_stream->p_byte)++;
return( i_result );
}
return UnalignedGetBits( p_bit_stream, 32 );
}
#else
# define GetBits32( p_bit_stream ) GetBits( p_bit_stream, 32 )
#endif
/*****************************************************************************
* RealignBits : realigns the bit buffer on an 8-bit boundary
......@@ -380,11 +389,6 @@ static __inline__ void RealignBits( bit_stream_t * p_bit_stream )
}
/*
* Philosophy of the third implementation : the decoder asks for n bytes,
* and we will copy them in its buffer.
*/
/*****************************************************************************
* GetChunk : reads a large chunk of data
*****************************************************************************
......@@ -397,6 +401,15 @@ static __inline__ void GetChunk( bit_stream_t * p_bit_stream,
{
ptrdiff_t i_available;
if( p_bit_stream->fifo.i_available )
{
*((WORD_TYPE *)p_buffer) = WORD_AT( p_bit_stream->fifo.buffer );
p_buffer += p_bit_stream->fifo.i_available >> 3;
i_buf_len -= p_bit_stream->fifo.i_available >> 3;
p_bit_stream->fifo.buffer = 0;
p_bit_stream->fifo.i_available = 0;
}
if( (i_available = p_bit_stream->p_end - p_bit_stream->p_byte)
>= i_buf_len )
{
......@@ -422,6 +435,43 @@ static __inline__ void GetChunk( bit_stream_t * p_bit_stream,
p_bit_stream->p_byte += i_buf_len;
}
}
if( p_bit_stream->p_byte <= p_bit_stream->p_end - sizeof(WORD_TYPE) )
{
AlignWord( p_bit_stream );
}
}
/*
* The following functions are now deprecated.
*/
static __inline__ byte_t _GetByte( bit_stream_t * p_bit_stream )
{
if ( p_bit_stream->p_byte >= p_bit_stream->p_end )
{
p_bit_stream->pf_next_data_packet( p_bit_stream );
}
return( *(p_bit_stream->p_byte++) );
}
static __inline__ void NeedBits( bit_stream_t * p_bit_stream, int i_bits )
{
while ( p_bit_stream->fifo.i_available < i_bits )
{
p_bit_stream->fifo.buffer |= ((WORD_TYPE)_GetByte( p_bit_stream ))
<< (8 * sizeof(WORD_TYPE) - 8
- p_bit_stream->fifo.i_available);
p_bit_stream->fifo.i_available += 8;
}
}
static __inline__ void DumpBits( bit_stream_t * p_bit_stream, int i_bits )
{
p_bit_stream->fifo.buffer <<= i_bits;
p_bit_stream->fifo.i_available -= i_bits;
}
......
......@@ -2,7 +2,7 @@
* input.h: structures of the input not exported to other modules
*****************************************************************************
* Copyright (C) 1999, 2000 VideoLAN
* $Id: input.h,v 1.10 2001/01/15 08:07:31 sam Exp $
* $Id: input.h,v 1.11 2001/01/18 17:40:06 massiot Exp $
*
* Authors:
*
......@@ -118,10 +118,6 @@ static __inline__ void input_NullPacket( input_thread_t * p_input,
return;
}
/* XXX FIXME SARASS TODO: remove the following one-liner kludge when
* we have bitstream IV, we won't need it anymore */
((WORD_TYPE *)p_pad_data->p_payload_start)++;
memset( p_pad_data->p_buffer, 0, PADDING_PACKET_SIZE );
p_pad_data->b_discard_payload = 1;
p_pes = p_es->p_pes;
......
......@@ -72,16 +72,9 @@ void InitBitstream( bit_stream_t * p_bit_stream, decoder_fifo_t * p_fifo )
*****************************************************************************/
void NextDataPacket( bit_stream_t * p_bit_stream )
{
WORD_TYPE buffer_left;
ptrdiff_t i_bytes_left;
decoder_fifo_t * p_fifo = p_bit_stream->p_decoder_fifo;
boolean_t b_new_pes;
/* Put the remaining bytes (not aligned on a word boundary) in a
* temporary buffer. */
i_bytes_left = p_bit_stream->p_end - p_bit_stream->p_byte;
buffer_left = *((WORD_TYPE *)p_bit_stream->p_end - 1);
/* We are looking for the next data packet that contains real data,
* and not just a PES header */
do
......@@ -133,8 +126,176 @@ void NextDataPacket( bit_stream_t * p_bit_stream )
{
p_bit_stream->pf_bitstream_callback( p_bit_stream, b_new_pes );
}
}
void OldKludge( bit_stream_t * p_bit_stream )
{
WORD_TYPE buffer_left;
ptrdiff_t i_bytes_left;
/* Put the remaining bytes (not aligned on a word boundary) in a
* temporary buffer. */
i_bytes_left = p_bit_stream->p_end - p_bit_stream->p_byte;
buffer_left = *((WORD_TYPE *)p_bit_stream->p_end - 1);
p_bit_stream->pf_next_data_packet( p_bit_stream );
/* Copy remaining bits of the previous packet */
*((WORD_TYPE *)p_bit_stream->p_byte - 1) = buffer_left;
p_bit_stream->p_byte -= i_bytes_left;
}
/*****************************************************************************
* UnalignedShowBits : return i_bits bits from the bit stream, even when
* not aligned on a word boundary
*****************************************************************************/
u32 UnalignedShowBits( bit_stream_t * p_bit_stream, unsigned int i_bits )
{
/* We just fill in the bit buffer. */
while( p_bit_stream->fifo.i_available < i_bits )
{
if( p_bit_stream->p_byte < p_bit_stream->p_end )
{
p_bit_stream->fifo.buffer |= *(p_bit_stream->p_byte++)
<< (8 * sizeof(WORD_TYPE) - 8
- p_bit_stream->fifo.i_available);
p_bit_stream->fifo.i_available += 8;
}
else
{
p_bit_stream->pf_next_data_packet( p_bit_stream );
p_bit_stream->fifo.buffer |= *(p_bit_stream->p_byte++)
<< (8 * sizeof(WORD_TYPE) - 8
- p_bit_stream->fifo.i_available);
p_bit_stream->fifo.i_available += 8;
}
}
return( p_bit_stream->fifo.buffer >> (8 * sizeof(WORD_TYPE) - i_bits) );
}
/*****************************************************************************
* UnalignedGetBits : returns i_bits bits from the bit stream and removes
* them from the buffer, even when the bit stream is not aligned on a word
* boundary
*****************************************************************************/
u32 UnalignedGetBits( bit_stream_t * p_bit_stream, unsigned int i_bits )
{
u32 i_result;
i_result = p_bit_stream->fifo.buffer
>> (8 * sizeof(WORD_TYPE) - i_bits);
i_bits -= p_bit_stream->fifo.i_available;
/* Gather missing bytes. */
while( i_bits >= 8 )
{
if( p_bit_stream->p_byte < p_bit_stream->p_end )
{
i_result |= *(p_bit_stream->p_byte++) << (i_bits - 8);
i_bits -= 8;
}
else
{
p_bit_stream->pf_next_data_packet( p_bit_stream );
i_result |= *(p_bit_stream->p_byte++) << (i_bits - 8);
i_bits -= 8;
}
}
/* Gather missing bits. */
if( i_bits > 0 )
{
unsigned int i_tmp = 8 - i_bits;
if( p_bit_stream->p_byte < p_bit_stream->p_end )
{
i_result |= *p_bit_stream->p_byte >> i_tmp;
p_bit_stream->fifo.buffer = *(p_bit_stream->p_byte++)
<< ( sizeof(WORD_TYPE) * 8 - i_tmp );
p_bit_stream->fifo.i_available = i_tmp;
}
else
{
p_bit_stream->pf_next_data_packet( p_bit_stream );
i_result |= *p_bit_stream->p_byte >> i_tmp;
p_bit_stream->fifo.buffer = *(p_bit_stream->p_byte++)
<< ( sizeof(WORD_TYPE) * 8 - i_tmp );
p_bit_stream->fifo.i_available = i_tmp;
}
}
else
{
p_bit_stream->fifo.i_available = 0;
p_bit_stream->fifo.buffer = 0;
}
if( p_bit_stream->p_byte <= p_bit_stream->p_end - sizeof(WORD_TYPE) )
{
/* Get aligned on a word boundary. Otherwise it is safer
* to do it the next time.
* NB : we _will_ get aligned, because we have at most
* sizeof(WORD_TYPE) - 1 bytes to store, and at least
* sizeof(WORD_TYPE) - 1 empty bytes in the bit buffer. */
AlignWord( p_bit_stream );
}
return( i_result );
}
/*****************************************************************************
* UnalignedRemoveBits : removes i_bits (== -i_available) from the bit
* buffer, even when the bit stream is not aligned on a word boundary
*****************************************************************************/
void UnalignedRemoveBits( bit_stream_t * p_bit_stream )
{
/* First remove all unnecessary bytes. */
while( p_bit_stream->fifo.i_available <= -8 )
{
if( p_bit_stream->p_byte < p_bit_stream->p_end )
{
p_bit_stream->p_byte++;
p_bit_stream->fifo.i_available += 8;
}
else
{
p_bit_stream->pf_next_data_packet( p_bit_stream );
p_bit_stream->p_byte++;
p_bit_stream->fifo.i_available += 8;
}
}
/* Remove unnecessary bits. */
if( p_bit_stream->fifo.i_available < 0 )
{
if( p_bit_stream->p_byte < p_bit_stream->p_end )
{
p_bit_stream->fifo.buffer = *(p_bit_stream->p_byte++)
<< ( sizeof(WORD_TYPE) * 8 - 8
- p_bit_stream->fifo.i_available );
p_bit_stream->fifo.i_available += 8;
}
else
{
p_bit_stream->pf_next_data_packet( p_bit_stream );
p_bit_stream->fifo.buffer = *(p_bit_stream->p_byte++)
<< ( sizeof(WORD_TYPE) * 8 - 8
- p_bit_stream->fifo.i_available );
p_bit_stream->fifo.i_available += 8;
}
}
else
{
p_bit_stream->fifo.buffer = 0;
}
if( p_bit_stream->p_byte <= p_bit_stream->p_end - sizeof(WORD_TYPE) )
{
/* Get aligned on a word boundary. Otherwise it is safer
* to do it the next time.
* NB : we _will_ get aligned, because we have at most
* sizeof(WORD_TYPE) - 1 bytes to store, and at least
* sizeof(WORD_TYPE) - 1 empty bytes in the bit buffer. */
AlignWord( p_bit_stream );
}
}
......@@ -2,7 +2,7 @@
* vpar_synchro.c : frame dropping routines
*****************************************************************************
* Copyright (C) 1999, 2000 VideoLAN
* $Id: vpar_synchro.c,v 1.78 2001/01/18 05:13:23 sam Exp $
* $Id: vpar_synchro.c,v 1.79 2001/01/18 17:40:06 massiot Exp $
*
* Authors: Christophe Massiot <massiot@via.ecp.fr>
* Samuel Hocevar <sam@via.ecp.fr>
......@@ -222,7 +222,8 @@ boolean_t vpar_SynchroChoose( vpar_thread_t * p_vpar, int i_coding_type,
+ tau_yuv)
#define S p_vpar->synchro
/* VPAR_SYNCHRO_DEFAULT */
mtime_t now, pts, period, tau_yuv;
mtime_t now, period, tau_yuv;
mtime_t pts = 0;
boolean_t b_decode = 0;
#ifdef DEBUG_VPAR
char p_date[MSTRTIME_MAX_SIZE];
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment