mirror of https://github.com/tongzx/nt5src
You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
1981 lines
46 KiB
1981 lines
46 KiB
#include "stdafx.h"
|
|
#pragma hdrstop
|
|
|
|
/*
|
|
* jchuff.c
|
|
*
|
|
* Copyright (C) 1991-1996, Thomas G. Lane.
|
|
* This file is part of the Independent JPEG Group's software.
|
|
* For conditions of distribution and use, see the accompanying README file.
|
|
*
|
|
* This file contains Huffman entropy encoding routines.
|
|
*
|
|
* Much of the complexity here has to do with supporting output suspension.
|
|
* If the data destination module demands suspension, we want to be able to
|
|
* back up to the start of the current MCU. To do this, we copy state
|
|
* variables into local working storage, and update them back to the
|
|
* permanent JPEG objects only upon successful completion of an MCU.
|
|
*/
|
|
|
|
|
|
// MMx Optimisation disabled 5/29/97 Gromit Bug 4375 -Tiling error. - ajais.
|
|
|
|
|
|
#pragma warning( disable : 4799 )
|
|
|
|
#define JPEG_INTERNALS
|
|
#include "jinclude.h"
|
|
#include "jpeglib.h"
|
|
#include "jchuff.h" /* Declarations shared with jcphuff.c */
|
|
|
|
/* Expanded entropy encoder object for Huffman encoding.
|
|
*
|
|
* The savable_state subrecord contains fields that change within an MCU,
|
|
* but must not be updated permanently until we complete the MCU.
|
|
*/
|
|
typedef struct
|
|
{
|
|
|
|
__int64 put_buffer_64; //mmx bit-accumulation buffer
|
|
INT32 put_buffer; /* current bit-accumulation buffer */
|
|
int put_bits; /* # of bits now in it */
|
|
int last_dc_val[MAX_COMPS_IN_SCAN]; /* last DC coef for each component */
|
|
|
|
} savable_state;
|
|
|
|
/* This macro is to work around compilers with missing or broken
|
|
* structure assignment. You'll need to fix this code if you have
|
|
* such a compiler and you change MAX_COMPS_IN_SCAN.
|
|
*/
|
|
|
|
//#ifndef NO_STRUCT_ASSIGN
|
|
//#define ASSIGN_STATE(dest,src) ((dest) = (src))
|
|
//#else
|
|
// pull out the assignments to put_buffer and put_bits since they are implentation dependent
|
|
#if MAX_COMPS_IN_SCAN == 4
|
|
#define ASSIGN_STATE(dest,src) \
|
|
((dest).last_dc_val[0] = (src).last_dc_val[0], \
|
|
(dest).last_dc_val[1] = (src).last_dc_val[1], \
|
|
(dest).last_dc_val[2] = (src).last_dc_val[2], \
|
|
(dest).last_dc_val[3] = (src).last_dc_val[3])
|
|
/*((dest).put_buffer = (src).put_buffer, \
|
|
(dest).put_bits = (src).put_bits, */
|
|
#endif
|
|
//#endif
|
|
|
|
|
|
typedef struct
|
|
{
|
|
|
|
struct jpeg_entropy_encoder pub; /* public fields */
|
|
|
|
savable_state saved; /* Bit buffer & DC state at start of MCU */
|
|
|
|
/* These fields are NOT loaded into local working state. */
|
|
unsigned int restarts_to_go; /* MCUs left in this restart interval */
|
|
int next_restart_num; /* next restart number to write (0-7) */
|
|
|
|
/* Pointers to derived tables (these workspaces have image lifespan) */
|
|
c_derived_tbl * dc_derived_tbls[NUM_HUFF_TBLS];
|
|
c_derived_tbl * ac_derived_tbls[NUM_HUFF_TBLS];
|
|
|
|
#ifdef ENTROPY_OPT_SUPPORTED /* Statistics tables for optimization */
|
|
long * dc_count_ptrs[NUM_HUFF_TBLS];
|
|
long * ac_count_ptrs[NUM_HUFF_TBLS];
|
|
#endif
|
|
|
|
} huff_entropy_encoder;
|
|
|
|
typedef huff_entropy_encoder * huff_entropy_ptr;
|
|
|
|
/* Working state while writing an MCU.
|
|
* This struct contains all the fields that are needed by subroutines.
|
|
*/
|
|
|
|
typedef struct
|
|
{
|
|
// make the next two variables global for easy access in mmx version
|
|
// JOCTET * next_output_byte; /* => next byte to write in buffer */
|
|
// size_t free_in_buffer; /* # of byte spaces remaining in buffer */
|
|
// savable_state cur; /* Current bit buffer & DC state */
|
|
// flatten (instantiate) savable state here
|
|
__int64 put_buffer_64; // mmx bit accumulation buffer
|
|
INT32 put_buffer; /* current bit-accumulation buffer */
|
|
int put_bits; /* # of bits now in it */
|
|
int last_dc_val[MAX_COMPS_IN_SCAN]; /* last DC coef for each component */
|
|
j_compress_ptr cinfo; /* dump_buffer needs access to this */
|
|
|
|
} working_state;
|
|
|
|
//global vaiables
|
|
__int64 put_buffer_64;
|
|
|
|
// INT32 put_buffer;
|
|
int put_bits;
|
|
JOCTET * next_output_byte; /* => next byte to write in buffer */
|
|
size_t free_in_buffer; /* # of byte spaces remaining in buffer */
|
|
boolean mmx_cpu=1;
|
|
|
|
|
|
/* Forward declarations */
|
|
METHODDEF(boolean) encode_mcu_huff JPP((j_compress_ptr cinfo,
|
|
JBLOCKROW *MCU_data));
|
|
METHODDEF(void) finish_pass_huff JPP((j_compress_ptr cinfo));
|
|
|
|
#ifdef ENTROPY_OPT_SUPPORTED
|
|
METHODDEF(boolean) encode_mcu_gather JPP((j_compress_ptr cinfo,
|
|
JBLOCKROW *MCU_data));
|
|
METHODDEF(void) finish_pass_gather JPP((j_compress_ptr cinfo));
|
|
#endif
|
|
|
|
void countZeros(int *indexBlock,short *coefBlock,short *outBlock,int *lastZeros,int *numElements);
|
|
|
|
boolean emit_bits_fast (working_state * state, unsigned int code, int bsize, int only1);
|
|
//extern boolean emit_bits (working_state * state, unsigned int code, int size);
|
|
|
|
boolean encode_one_block_fast (working_state * state, JCOEFPTR block, int last_dc_val,
|
|
c_derived_tbl *dctbl, c_derived_tbl *actbl);
|
|
//extern boolean encode_one_block (working_state * state, JCOEFPTR block, int last_dc_val,
|
|
// c_derived_tbl *dctbl, c_derived_tbl *actbl);
|
|
|
|
/*
|
|
* Initialize for a Huffman-compressed scan.
|
|
* If gather_statistics is TRUE, we do not output anything during the scan,
|
|
* just count the Huffman symbols used and generate Huffman code tables.
|
|
*/
|
|
|
|
METHODDEF(void)
|
|
start_pass_huff (j_compress_ptr cinfo, boolean gather_statistics)
|
|
{
|
|
huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
|
|
int ci, dctbl, actbl;
|
|
jpeg_component_info * compptr;
|
|
|
|
if (gather_statistics) {
|
|
#ifdef ENTROPY_OPT_SUPPORTED
|
|
entropy->pub.encode_mcu = encode_mcu_gather;
|
|
entropy->pub.finish_pass = finish_pass_gather;
|
|
#else
|
|
ERREXIT(cinfo, JERR_NOT_COMPILED);
|
|
#endif
|
|
} else {
|
|
entropy->pub.encode_mcu = encode_mcu_huff;
|
|
entropy->pub.finish_pass = finish_pass_huff;
|
|
}
|
|
|
|
for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
|
|
compptr = cinfo->cur_comp_info[ci];
|
|
dctbl = compptr->dc_tbl_no;
|
|
actbl = compptr->ac_tbl_no;
|
|
/* Make sure requested tables are present */
|
|
/* (In gather mode, tables need not be allocated yet) */
|
|
if (dctbl < 0 || dctbl >= NUM_HUFF_TBLS ||
|
|
(cinfo->dc_huff_tbl_ptrs[dctbl] == NULL && !gather_statistics))
|
|
ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, dctbl);
|
|
if (actbl < 0 || actbl >= NUM_HUFF_TBLS ||
|
|
(cinfo->ac_huff_tbl_ptrs[actbl] == NULL && !gather_statistics))
|
|
ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, actbl);
|
|
if (gather_statistics) {
|
|
#ifdef ENTROPY_OPT_SUPPORTED
|
|
/* Allocate and zero the statistics tables */
|
|
/* Note that jpeg_gen_optimal_table expects 257 entries in each table! */
|
|
if (entropy->dc_count_ptrs[dctbl] == NULL)
|
|
entropy->dc_count_ptrs[dctbl] = (long *)
|
|
(*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
|
|
257 * SIZEOF(long));
|
|
MEMZERO(entropy->dc_count_ptrs[dctbl], 257 * SIZEOF(long));
|
|
if (entropy->ac_count_ptrs[actbl] == NULL)
|
|
entropy->ac_count_ptrs[actbl] = (long *)
|
|
(*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
|
|
257 * SIZEOF(long));
|
|
MEMZERO(entropy->ac_count_ptrs[actbl], 257 * SIZEOF(long));
|
|
#endif
|
|
} else {
|
|
/* Compute derived values for Huffman tables */
|
|
/* We may do this more than once for a table, but it's not expensive */
|
|
jpeg_make_c_derived_tbl(cinfo, cinfo->dc_huff_tbl_ptrs[dctbl],
|
|
& entropy->dc_derived_tbls[dctbl]);
|
|
jpeg_make_c_derived_tbl(cinfo, cinfo->ac_huff_tbl_ptrs[actbl],
|
|
& entropy->ac_derived_tbls[actbl]);
|
|
}
|
|
/* Initialize DC predictions to 0 */
|
|
entropy->saved.last_dc_val[ci] = 0;
|
|
}
|
|
|
|
/* Initialize bit buffer to empty */
|
|
entropy->saved.put_buffer_64 = 0;
|
|
entropy->saved.put_buffer = 0;
|
|
entropy->saved.put_bits = 0;
|
|
|
|
/* Initialize restart stuff */
|
|
entropy->restarts_to_go = cinfo->restart_interval;
|
|
entropy->next_restart_num = 0;
|
|
}
|
|
|
|
|
|
/*
|
|
* Compute the derived values for a Huffman table.
|
|
* Note this is also used by jcphuff.c.
|
|
*/
|
|
|
|
GLOBAL(void)
|
|
jpeg_make_c_derived_tbl (j_compress_ptr cinfo, JHUFF_TBL * htbl,
|
|
c_derived_tbl ** pdtbl)
|
|
{
|
|
c_derived_tbl *dtbl;
|
|
int p, i, l, lastp, si;
|
|
char huffsize[257];
|
|
unsigned int huffcode[257];
|
|
unsigned int code;
|
|
|
|
/* Allocate a workspace if we haven't already done so. */
|
|
if (*pdtbl == NULL)
|
|
*pdtbl = (c_derived_tbl *)
|
|
(*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
|
|
SIZEOF(c_derived_tbl));
|
|
dtbl = *pdtbl;
|
|
|
|
/* Figure C.1: make table of Huffman code length for each symbol */
|
|
/* Note that this is in code-length order. */
|
|
|
|
p = 0;
|
|
for (l = 1; l <= 16; l++) {
|
|
for (i = 1; i <= (int) htbl->bits[l]; i++)
|
|
huffsize[p++] = (char) l;
|
|
}
|
|
huffsize[p] = 0;
|
|
lastp = p;
|
|
|
|
/* Figure C.2: generate the codes themselves */
|
|
/* Note that this is in code-length order. */
|
|
|
|
code = 0;
|
|
si = huffsize[0];
|
|
p = 0;
|
|
while (huffsize[p]) {
|
|
while (((int) huffsize[p]) == si) {
|
|
huffcode[p++] = code;
|
|
code++;
|
|
}
|
|
code <<= 1;
|
|
si++;
|
|
}
|
|
|
|
/* Figure C.3: generate encoding tables */
|
|
/* These are code and size indexed by symbol value */
|
|
|
|
/* Set any codeless symbols to have code length 0;
|
|
* this allows emit_bits to detect any attempt to emit such symbols.
|
|
*/
|
|
MEMZERO(dtbl->ehufsi, SIZEOF(dtbl->ehufsi));
|
|
|
|
for (p = 0; p < lastp; p++) {
|
|
dtbl->ehufco[htbl->huffval[p]] = huffcode[p];
|
|
dtbl->ehufsi[htbl->huffval[p]] = huffsize[p];
|
|
}
|
|
}
|
|
|
|
|
|
/* Outputting bytes to the file */
|
|
|
|
/* Emit a byte, taking 'action' if must suspend. */
|
|
#define emit_byte(state,val,action) \
|
|
{ *next_output_byte++ = (JOCTET) (val); \
|
|
if (--free_in_buffer == 0) \
|
|
if (! dump_buffer(state)) \
|
|
{ action; } }
|
|
|
|
|
|
GLOBAL(boolean)
|
|
dump_buffer (working_state * state)
|
|
/* Empty the output buffer; return TRUE if successful, FALSE if must suspend */
|
|
{
|
|
struct jpeg_destination_mgr * dest = state->cinfo->dest;
|
|
|
|
if (! (*dest->empty_output_buffer) (state->cinfo))
|
|
return FALSE;
|
|
/* After a successful buffer dump, must reset buffer pointers */
|
|
next_output_byte = dest->next_output_byte;
|
|
free_in_buffer = dest->free_in_buffer;
|
|
return TRUE;
|
|
}
|
|
|
|
|
|
/* Outputting bits to the file */
|
|
|
|
/* Only the right 24 bits of put_buffer are used; the valid bits are
|
|
* left-justified in this part. At most 16 bits can be passed to emit_bits
|
|
* in one call, and we never retain more than 7 bits in put_buffer
|
|
* between calls, so 24 bits are sufficient.
|
|
*/
|
|
|
|
//INLINE
|
|
LOCAL(boolean)
|
|
emit_bits (working_state * state, unsigned int code, int size)
|
|
/* Emit some bits; return TRUE if successful, FALSE if must suspend */
|
|
{
|
|
/* This routine is heavily used, so it's worth coding tightly. */
|
|
register INT32 put_buffer = (INT32) code;
|
|
register int put_bits = state->put_bits;
|
|
|
|
/* if size is 0, caller used an invalid Huffman table entry */
|
|
if (size == 0)
|
|
ERREXIT(state->cinfo, JERR_HUFF_MISSING_CODE);
|
|
|
|
put_buffer &= (((INT32) 1)<<size) - 1; /* mask off any extra bits in code */
|
|
|
|
put_bits += size; /* new number of bits in buffer */
|
|
|
|
put_buffer <<= 24 - put_bits; /* align incoming bits */
|
|
|
|
put_buffer |= state->put_buffer; /* and merge with old buffer contents */
|
|
|
|
while (put_bits >= 8) {
|
|
int c = (int) ((put_buffer >> 16) & 0xFF);
|
|
|
|
emit_byte(state, c, return FALSE);
|
|
if (c == 0xFF) { /* need to stuff a zero byte? */
|
|
emit_byte(state, 0, return FALSE);
|
|
}
|
|
put_buffer <<= 8;
|
|
put_bits -= 8;
|
|
}
|
|
|
|
state->put_buffer = put_buffer; /* update state variables */
|
|
state->put_bits = put_bits;
|
|
|
|
return TRUE;
|
|
}
|
|
|
|
|
|
//This is a routine to dump whatever is in put_buffer out - I salvaged it from another
|
|
//routine so there is some dead-code as-used.
|
|
|
|
//As flush-bits is not called frequently, there should not be much overhead to this code . . .
|
|
//MJB
|
|
|
|
//
|
|
// Need to add #ifdef for Alpha port
|
|
//
|
|
#if defined (_X86_)
|
|
void flush_bit_buffer_64()
|
|
{
|
|
// byte-align previous bits if any
|
|
__asm{
|
|
|
|
mov ebx,[put_bits]
|
|
|
|
mov eax,[next_output_byte]
|
|
|
|
test ebx,ebx
|
|
je no_ser_buf_data
|
|
|
|
|
|
movq mm0,[put_buffer_64]
|
|
pxor mm2,mm2
|
|
|
|
dump_loop: movq mm1,mm0
|
|
psrlq mm0,56
|
|
|
|
movd ecx,mm0
|
|
movq mm0,mm1
|
|
|
|
mov (byte ptr[eax]),cl
|
|
inc eax
|
|
|
|
cmp ecx,0xFF
|
|
jne not_ff
|
|
|
|
mov (byte ptr[eax]),0x00
|
|
dec [free_in_buffer]
|
|
|
|
inc eax
|
|
nop
|
|
not_ff:
|
|
dec [free_in_buffer]
|
|
psllq mm0,8
|
|
|
|
sub ebx,8
|
|
jg dump_loop
|
|
|
|
mov [put_bits],0
|
|
//mov [eb_ptr],eax
|
|
|
|
movq [put_buffer_64],mm2
|
|
//emms
|
|
|
|
no_ser_buf_data:
|
|
|
|
mov [next_output_byte],eax
|
|
|
|
}
|
|
}
|
|
#endif // #ifdef (_X86_)
|
|
|
|
LOCAL(boolean)
|
|
flush_bits (working_state * state)
|
|
{
|
|
//
|
|
// Need to add #ifdef for Alpha port
|
|
//
|
|
#if defined (_X86_)
|
|
if (0)//vfMMXMachine)
|
|
{
|
|
//if (! emit_bits(state, 0x7F, 7)) /* fill any partial byte with ones */
|
|
if (! emit_bits_fast(state, 0x7F, 7, 1)) /* fill any partial byte with ones */
|
|
return FALSE;
|
|
if (put_bits)
|
|
{
|
|
flush_bit_buffer_64(); // New stuff to write the last, few bits . . .
|
|
}
|
|
|
|
}
|
|
else
|
|
#endif
|
|
{
|
|
if (! emit_bits(state, 0x7F, 7)) /* fill any partial byte with ones */
|
|
return FALSE;
|
|
}
|
|
|
|
state->put_buffer_64 = 0; /* and reset bit-buffer to empty */
|
|
state->put_buffer = 0; /* and reset bit-buffer to empty */
|
|
state->put_bits = 0;
|
|
return TRUE;
|
|
}
|
|
|
|
|
|
/* Encode a single block's worth of coefficients */
|
|
|
|
LOCAL(boolean)
|
|
encode_one_block (working_state * state, JCOEFPTR block, int last_dc_val,
|
|
c_derived_tbl *dctbl, c_derived_tbl *actbl)
|
|
{
|
|
register int temp, temp2;
|
|
register int nbits;
|
|
register int k, r, i;
|
|
|
|
/* Encode the DC coefficient difference per section F.1.2.1 */
|
|
|
|
temp = temp2 = block[0] - last_dc_val;
|
|
|
|
if (temp < 0)
|
|
{
|
|
temp = -temp; /* temp is abs value of input */
|
|
/* For a negative input, want temp2 = bitwise complement of abs(input) */
|
|
/* This code assumes we are on a two's complement machine */
|
|
temp2--;
|
|
}
|
|
|
|
/* Find the number of bits needed for the magnitude of the coefficient */
|
|
nbits = 0;
|
|
while (temp)
|
|
{
|
|
nbits++;
|
|
temp >>= 1;
|
|
}
|
|
|
|
/* Emit the Huffman-coded symbol for the number of bits */
|
|
if (! emit_bits(state, dctbl->ehufco[nbits], dctbl->ehufsi[nbits]))
|
|
return FALSE;
|
|
|
|
/* Emit that number of bits of the value, if positive, */
|
|
/* or the complement of its magnitude, if negative. */
|
|
if (nbits) /* emit_bits rejects calls with size 0 */
|
|
if (! emit_bits(state, (unsigned int) temp2, nbits))
|
|
return FALSE;
|
|
|
|
/* Encode the AC coefficients per section F.1.2.2 */
|
|
|
|
r = 0; /* r = run length of zeros */
|
|
|
|
for (k = 1; k < DCTSIZE2; k++)
|
|
{
|
|
if ((temp = block[jpeg_natural_order[k]]) == 0)
|
|
{
|
|
r++;
|
|
}
|
|
else
|
|
{
|
|
/* if run length > 15, must emit special run-length-16 codes (0xF0) */
|
|
while (r > 15)
|
|
{
|
|
if (! emit_bits(state, actbl->ehufco[0xF0], actbl->ehufsi[0xF0]))
|
|
return FALSE;
|
|
r -= 16;
|
|
}
|
|
|
|
temp2 = temp;
|
|
|
|
if (temp < 0)
|
|
{
|
|
temp = -temp; /* temp is abs value of input */
|
|
/* This code assumes we are on a two's complement machine */
|
|
temp2--;
|
|
}
|
|
|
|
/* Find the number of bits needed for the magnitude of the coefficient */
|
|
nbits = 1; /* there must be at least one 1 bit */
|
|
while ((temp >>= 1))
|
|
nbits++;
|
|
|
|
/* Emit Huffman symbol for run length / number of bits */
|
|
i = (r << 4) + nbits;
|
|
if (! emit_bits(state, actbl->ehufco[i], actbl->ehufsi[i]))
|
|
return FALSE;
|
|
|
|
/* Emit that number of bits of the value, if positive, */
|
|
/* or the complement of its magnitude, if negative. */
|
|
if (! emit_bits(state, (unsigned int) temp2, nbits))
|
|
return FALSE;
|
|
|
|
r = 0;
|
|
}
|
|
}
|
|
|
|
/* If the last coef(s) were zero, emit an end-of-block code */
|
|
if (r > 0)
|
|
if (! emit_bits(state, actbl->ehufco[0], actbl->ehufsi[0]))
|
|
return FALSE;
|
|
|
|
return TRUE;
|
|
}
|
|
|
|
|
|
/*
|
|
* Emit a restart marker & resynchronize predictions.
|
|
*/
|
|
|
|
LOCAL(boolean)
|
|
emit_restart (working_state * state, int restart_num)
|
|
{
|
|
int ci;
|
|
|
|
if (! flush_bits(state))
|
|
return FALSE;
|
|
|
|
emit_byte(state, 0xFF, return FALSE);
|
|
emit_byte(state, JPEG_RST0 + restart_num, return FALSE);
|
|
|
|
/* Re-initialize DC predictions to 0 */
|
|
for (ci = 0; ci < state->cinfo->comps_in_scan; ci++)
|
|
state->last_dc_val[ci] = 0;
|
|
|
|
/* The restart counter is not updated until we successfully write the MCU. */
|
|
|
|
return TRUE;
|
|
}
|
|
|
|
|
|
/*
|
|
* Encode and output one MCU's worth of Huffman-compressed coefficients.
|
|
*/
|
|
|
|
METHODDEF(boolean)
|
|
encode_mcu_huff (j_compress_ptr cinfo, JBLOCKROW *MCU_data)
|
|
{
|
|
huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
|
|
working_state state;
|
|
int blkn, ci;
|
|
jpeg_component_info * compptr;
|
|
|
|
/* Load up working state */
|
|
next_output_byte = cinfo->dest->next_output_byte;
|
|
free_in_buffer = cinfo->dest->free_in_buffer;
|
|
if (0)//vfMMXMachine)
|
|
{
|
|
state.put_buffer_64=entropy->saved.put_buffer_64;
|
|
}
|
|
else
|
|
{
|
|
state.put_buffer=entropy->saved.put_buffer;
|
|
}
|
|
state.put_bits=entropy->saved.put_bits;
|
|
ASSIGN_STATE(state, entropy->saved);
|
|
state.cinfo = cinfo;
|
|
|
|
/* Emit restart marker if needed */
|
|
if (cinfo->restart_interval)
|
|
{
|
|
if (entropy->restarts_to_go == 0)
|
|
if (! emit_restart(&state, entropy->next_restart_num))
|
|
return FALSE;
|
|
}
|
|
|
|
/* Encode the MCU data blocks */
|
|
for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++)
|
|
{
|
|
ci = cinfo->MCU_membership[blkn];
|
|
compptr = cinfo->cur_comp_info[ci];
|
|
|
|
//
|
|
// Need to add #ifdef for Alpha port
|
|
//
|
|
#if defined (_X86_)
|
|
if (0)//vfMMXMachine)
|
|
{
|
|
if (! encode_one_block_fast(&state,
|
|
MCU_data[blkn][0], state.last_dc_val[ci],
|
|
entropy->dc_derived_tbls[compptr->dc_tbl_no],
|
|
entropy->ac_derived_tbls[compptr->ac_tbl_no]))
|
|
return FALSE;
|
|
}
|
|
else
|
|
#endif
|
|
{
|
|
if (! encode_one_block(&state,
|
|
MCU_data[blkn][0], state.last_dc_val[ci],
|
|
entropy->dc_derived_tbls[compptr->dc_tbl_no],
|
|
entropy->ac_derived_tbls[compptr->ac_tbl_no]))
|
|
return FALSE;
|
|
}
|
|
/* Update last_dc_val */
|
|
state.last_dc_val[ci] = MCU_data[blkn][0][0];
|
|
}
|
|
|
|
/* Completed MCU, so update state */
|
|
cinfo->dest->next_output_byte = next_output_byte;
|
|
cinfo->dest->free_in_buffer = free_in_buffer;
|
|
|
|
if (0)//vfMMXMachine)
|
|
{
|
|
entropy->saved.put_buffer_64=state.put_buffer_64;
|
|
}
|
|
else
|
|
{
|
|
entropy->saved.put_buffer=state.put_buffer;
|
|
}
|
|
|
|
entropy->saved.put_bits=state.put_bits;
|
|
|
|
ASSIGN_STATE(entropy->saved, state);
|
|
|
|
/* Update restart-interval state too */
|
|
if (cinfo->restart_interval)
|
|
{
|
|
if (entropy->restarts_to_go == 0)
|
|
{
|
|
entropy->restarts_to_go = cinfo->restart_interval;
|
|
entropy->next_restart_num++;
|
|
entropy->next_restart_num &= 7;
|
|
}
|
|
entropy->restarts_to_go--;
|
|
}
|
|
|
|
//
|
|
// Need to add #ifdef for Alpha port
|
|
//
|
|
#if defined (_X86_)
|
|
if (0)//vfMMXMachine)
|
|
{
|
|
__asm emms
|
|
}
|
|
#endif
|
|
|
|
return TRUE;
|
|
}
|
|
|
|
|
|
/*
|
|
* Finish up at the end of a Huffman-compressed scan.
|
|
*/
|
|
|
|
METHODDEF(void)
|
|
finish_pass_huff (j_compress_ptr cinfo)
|
|
{
|
|
huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
|
|
working_state state;
|
|
|
|
/* Load up working state ... flush_bits needs it */
|
|
next_output_byte = cinfo->dest->next_output_byte;
|
|
free_in_buffer = cinfo->dest->free_in_buffer;
|
|
|
|
if (0)//vfMMXMachine)
|
|
{
|
|
state.put_buffer_64=entropy->saved.put_buffer_64;
|
|
}
|
|
else
|
|
{
|
|
state.put_buffer=entropy->saved.put_buffer;
|
|
}
|
|
|
|
state.put_bits=entropy->saved.put_bits;
|
|
ASSIGN_STATE(state, entropy->saved);
|
|
state.cinfo = cinfo;
|
|
|
|
/* Flush out the last data */
|
|
if (!flush_bits(&state))
|
|
ERREXIT(cinfo, JERR_CANT_SUSPEND);
|
|
|
|
/* Update state */
|
|
cinfo->dest->next_output_byte = next_output_byte;
|
|
cinfo->dest->free_in_buffer = free_in_buffer;
|
|
|
|
if (0)//vfMMXMachine)
|
|
{
|
|
entropy->saved.put_buffer_64=state.put_buffer_64;
|
|
}
|
|
else
|
|
{
|
|
entropy->saved.put_buffer=state.put_buffer;
|
|
}
|
|
entropy->saved.put_bits=state.put_bits;
|
|
ASSIGN_STATE(entropy->saved, state);
|
|
|
|
//
|
|
// Need to add #ifdef for Alpha port
|
|
//
|
|
#if defined (_X86_)
|
|
if (0)//vfMMXMachine)
|
|
{
|
|
__asm emms
|
|
}
|
|
#endif
|
|
}
|
|
|
|
|
|
/*
|
|
* Huffman coding optimization.
|
|
*
|
|
* This actually is optimization, in the sense that we find the best possible
|
|
* Huffman table(s) for the given data. We first scan the supplied data and
|
|
* count the number of uses of each symbol that is to be Huffman-coded.
|
|
* (This process must agree with the code above.) Then we build an
|
|
* optimal Huffman coding tree for the observed counts.
|
|
*
|
|
* The JPEG standard requires Huffman codes to be no more than 16 bits long.
|
|
* If some symbols have a very small but nonzero probability, the Huffman tree
|
|
* must be adjusted to meet the code length restriction. We currently use
|
|
* the adjustment method suggested in the JPEG spec. This method is *not*
|
|
* optimal; it may not choose the best possible limited-length code. But
|
|
* since the symbols involved are infrequently used, it's not clear that
|
|
* going to extra trouble is worthwhile.
|
|
*/
|
|
|
|
#ifdef ENTROPY_OPT_SUPPORTED
|
|
|
|
|
|
/* Process a single block's worth of coefficients */
|
|
|
|
LOCAL(void)
|
|
htest_one_block (JCOEFPTR block, int last_dc_val,
|
|
long dc_counts[], long ac_counts[])
|
|
{
|
|
register int temp;
|
|
register int nbits;
|
|
register int k, r;
|
|
|
|
/* Encode the DC coefficient difference per section F.1.2.1 */
|
|
|
|
temp = block[0] - last_dc_val;
|
|
if (temp < 0)
|
|
temp = -temp;
|
|
|
|
/* Find the number of bits needed for the magnitude of the coefficient */
|
|
nbits = 0;
|
|
while (temp)
|
|
{
|
|
nbits++;
|
|
temp >>= 1;
|
|
}
|
|
|
|
/* Count the Huffman symbol for the number of bits */
|
|
dc_counts[nbits]++;
|
|
|
|
/* Encode the AC coefficients per section F.1.2.2 */
|
|
|
|
r = 0; /* r = run length of zeros */
|
|
|
|
for (k = 1; k < DCTSIZE2; k++)
|
|
{
|
|
if ((temp = block[jpeg_natural_order[k]]) == 0)
|
|
{
|
|
r++;
|
|
}
|
|
else
|
|
{
|
|
/* if run length > 15, must emit special run-length-16 codes (0xF0) */
|
|
while (r > 15)
|
|
{
|
|
ac_counts[0xF0]++;
|
|
r -= 16;
|
|
}
|
|
|
|
/* Find the number of bits needed for the magnitude of the coefficient */
|
|
if (temp < 0) temp = -temp;
|
|
|
|
/* Find the number of bits needed for the magnitude of the coefficient */
|
|
nbits = 1; /* there must be at least one 1 bit */
|
|
while ((temp >>= 1)) nbits++;
|
|
|
|
/* Count Huffman symbol for run length / number of bits */
|
|
ac_counts[(r << 4) + nbits]++;
|
|
|
|
r = 0;
|
|
}
|
|
}
|
|
|
|
/* If the last coef(s) were zero, emit an end-of-block code */
|
|
if (r > 0)
|
|
ac_counts[0]++;
|
|
}
|
|
|
|
|
|
/*
|
|
* Trial-encode one MCU's worth of Huffman-compressed coefficients.
|
|
* No data is actually output, so no suspension return is possible.
|
|
*/
|
|
|
|
METHODDEF(boolean)
|
|
encode_mcu_gather (j_compress_ptr cinfo, JBLOCKROW *MCU_data)
|
|
{
|
|
huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
|
|
int blkn, ci;
|
|
jpeg_component_info * compptr;
|
|
|
|
/* Take care of restart intervals if needed */
|
|
if (cinfo->restart_interval) {
|
|
if (entropy->restarts_to_go == 0) {
|
|
/* Re-initialize DC predictions to 0 */
|
|
for (ci = 0; ci < cinfo->comps_in_scan; ci++)
|
|
entropy->saved.last_dc_val[ci] = 0;
|
|
/* Update restart state */
|
|
entropy->restarts_to_go = cinfo->restart_interval;
|
|
}
|
|
entropy->restarts_to_go--;
|
|
}
|
|
|
|
for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
|
|
ci = cinfo->MCU_membership[blkn];
|
|
compptr = cinfo->cur_comp_info[ci];
|
|
htest_one_block(MCU_data[blkn][0], entropy->saved.last_dc_val[ci],
|
|
entropy->dc_count_ptrs[compptr->dc_tbl_no],
|
|
entropy->ac_count_ptrs[compptr->ac_tbl_no]);
|
|
entropy->saved.last_dc_val[ci] = MCU_data[blkn][0][0];
|
|
}
|
|
|
|
return TRUE;
|
|
}
|
|
|
|
|
|
/*
|
|
* Generate the optimal coding for the given counts, fill htbl.
|
|
* Note this is also used by jcphuff.c.
|
|
*/
|
|
|
|
GLOBAL(void)
|
|
jpeg_gen_optimal_table (j_compress_ptr cinfo, JHUFF_TBL * htbl, long freq[])
|
|
{
|
|
#define MAX_CLEN 32 /* assumed maximum initial code length */
|
|
UINT8 bits[MAX_CLEN+1]; /* bits[k] = # of symbols with code length k */
|
|
int codesize[257]; /* codesize[k] = code length of symbol k */
|
|
int others[257]; /* next symbol in current branch of tree */
|
|
int c1, c2;
|
|
int p, i, j;
|
|
long v;
|
|
|
|
/* This algorithm is explained in section K.2 of the JPEG standard */
|
|
|
|
MEMZERO(bits, SIZEOF(bits));
|
|
MEMZERO(codesize, SIZEOF(codesize));
|
|
for (i = 0; i < 257; i++)
|
|
others[i] = -1; /* init links to empty */
|
|
|
|
freq[256] = 1; /* make sure there is a nonzero count */
|
|
/* Including the pseudo-symbol 256 in the Huffman procedure guarantees
|
|
* that no real symbol is given code-value of all ones, because 256
|
|
* will be placed in the largest codeword category.
|
|
*/
|
|
|
|
/* Huffman's basic algorithm to assign optimal code lengths to symbols */
|
|
|
|
for (;;) {
|
|
/* Find the smallest nonzero frequency, set c1 = its symbol */
|
|
/* In case of ties, take the larger symbol number */
|
|
c1 = -1;
|
|
v = 1000000000L;
|
|
for (i = 0; i <= 256; i++) {
|
|
if (freq[i] && freq[i] <= v) {
|
|
v = freq[i];
|
|
c1 = i;
|
|
}
|
|
}
|
|
|
|
/* Find the next smallest nonzero frequency, set c2 = its symbol */
|
|
/* In case of ties, take the larger symbol number */
|
|
c2 = -1;
|
|
v = 1000000000L;
|
|
for (i = 0; i <= 256; i++) {
|
|
if (freq[i] && freq[i] <= v && i != c1) {
|
|
v = freq[i];
|
|
c2 = i;
|
|
}
|
|
}
|
|
|
|
/* Done if we've merged everything into one frequency */
|
|
if (c2 < 0)
|
|
break;
|
|
|
|
/* Else merge the two counts/trees */
|
|
freq[c1] += freq[c2];
|
|
freq[c2] = 0;
|
|
|
|
/* Increment the codesize of everything in c1's tree branch */
|
|
codesize[c1]++;
|
|
while (others[c1] >= 0) {
|
|
c1 = others[c1];
|
|
codesize[c1]++;
|
|
}
|
|
|
|
others[c1] = c2; /* chain c2 onto c1's tree branch */
|
|
|
|
/* Increment the codesize of everything in c2's tree branch */
|
|
codesize[c2]++;
|
|
while (others[c2] >= 0) {
|
|
c2 = others[c2];
|
|
codesize[c2]++;
|
|
}
|
|
}
|
|
|
|
/* Now count the number of symbols of each code length */
|
|
for (i = 0; i <= 256; i++) {
|
|
if (codesize[i]) {
|
|
/* The JPEG standard seems to think that this can't happen, */
|
|
/* but I'm paranoid... */
|
|
if (codesize[i] > MAX_CLEN)
|
|
ERREXIT(cinfo, JERR_HUFF_CLEN_OVERFLOW);
|
|
|
|
bits[codesize[i]]++;
|
|
}
|
|
}
|
|
|
|
/* JPEG doesn't allow symbols with code lengths over 16 bits, so if the pure
|
|
* Huffman procedure assigned any such lengths, we must adjust the coding.
|
|
* Here is what the JPEG spec says about how this next bit works:
|
|
* Since symbols are paired for the longest Huffman code, the symbols are
|
|
* removed from this length category two at a time. The prefix for the pair
|
|
* (which is one bit shorter) is allocated to one of the pair; then,
|
|
* skipping the BITS entry for that prefix length, a code word from the next
|
|
* shortest nonzero BITS entry is converted into a prefix for two code words
|
|
* one bit longer.
|
|
*/
|
|
|
|
for (i = MAX_CLEN; i > 16; i--) {
|
|
while (bits[i] > 0) {
|
|
j = i - 2; /* find length of new prefix to be used */
|
|
while (bits[j] == 0)
|
|
j--;
|
|
|
|
bits[i] -= 2; /* remove two symbols */
|
|
bits[i-1]++; /* one goes in this length */
|
|
bits[j+1] += 2; /* two new symbols in this length */
|
|
bits[j]--; /* symbol of this length is now a prefix */
|
|
}
|
|
}
|
|
|
|
/* Remove the count for the pseudo-symbol 256 from the largest codelength */
|
|
while (bits[i] == 0) /* find largest codelength still in use */
|
|
i--;
|
|
bits[i]--;
|
|
|
|
/* Return final symbol counts (only for lengths 0..16) */
|
|
MEMCOPY(htbl->bits, bits, SIZEOF(htbl->bits));
|
|
|
|
/* Return a list of the symbols sorted by code length */
|
|
/* It's not real clear to me why we don't need to consider the codelength
|
|
* changes made above, but the JPEG spec seems to think this works.
|
|
*/
|
|
p = 0;
|
|
for (i = 1; i <= MAX_CLEN; i++) {
|
|
for (j = 0; j <= 255; j++) {
|
|
if (codesize[j] == i) {
|
|
htbl->huffval[p] = (UINT8) j;
|
|
p++;
|
|
}
|
|
}
|
|
}
|
|
|
|
/* Set sent_table FALSE so updated table will be written to JPEG file. */
|
|
htbl->sent_table = FALSE;
|
|
}
|
|
|
|
|
|
/*
|
|
* Finish up a statistics-gathering pass and create the new Huffman tables.
|
|
*/
|
|
|
|
METHODDEF(void)
|
|
finish_pass_gather (j_compress_ptr cinfo)
|
|
{
|
|
huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy;
|
|
int ci, dctbl, actbl;
|
|
jpeg_component_info * compptr;
|
|
JHUFF_TBL **htblptr;
|
|
boolean did_dc[NUM_HUFF_TBLS];
|
|
boolean did_ac[NUM_HUFF_TBLS];
|
|
|
|
/* It's important not to apply jpeg_gen_optimal_table more than once
|
|
* per table, because it clobbers the input frequency counts!
|
|
*/
|
|
MEMZERO(did_dc, SIZEOF(did_dc));
|
|
MEMZERO(did_ac, SIZEOF(did_ac));
|
|
|
|
for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
|
|
compptr = cinfo->cur_comp_info[ci];
|
|
dctbl = compptr->dc_tbl_no;
|
|
actbl = compptr->ac_tbl_no;
|
|
if (! did_dc[dctbl]) {
|
|
htblptr = & cinfo->dc_huff_tbl_ptrs[dctbl];
|
|
if (*htblptr == NULL)
|
|
*htblptr = jpeg_alloc_huff_table((j_common_ptr) cinfo);
|
|
jpeg_gen_optimal_table(cinfo, *htblptr, entropy->dc_count_ptrs[dctbl]);
|
|
did_dc[dctbl] = TRUE;
|
|
}
|
|
if (! did_ac[actbl]) {
|
|
htblptr = & cinfo->ac_huff_tbl_ptrs[actbl];
|
|
if (*htblptr == NULL)
|
|
*htblptr = jpeg_alloc_huff_table((j_common_ptr) cinfo);
|
|
jpeg_gen_optimal_table(cinfo, *htblptr, entropy->ac_count_ptrs[actbl]);
|
|
did_ac[actbl] = TRUE;
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
#endif /* ENTROPY_OPT_SUPPORTED */
|
|
|
|
|
|
/*
|
|
* Module initialization routine for Huffman entropy encoding.
|
|
*/
|
|
|
|
GLOBAL(void)
|
|
jinit_huff_encoder (j_compress_ptr cinfo)
|
|
{
|
|
huff_entropy_ptr entropy;
|
|
int i;
|
|
|
|
entropy = (huff_entropy_ptr)
|
|
(*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
|
|
SIZEOF(huff_entropy_encoder));
|
|
cinfo->entropy = (struct jpeg_entropy_encoder *) entropy;
|
|
entropy->pub.start_pass = start_pass_huff;
|
|
|
|
/* Mark tables unallocated */
|
|
for (i = 0; i < NUM_HUFF_TBLS; i++) {
|
|
entropy->dc_derived_tbls[i] = entropy->ac_derived_tbls[i] = NULL;
|
|
#ifdef ENTROPY_OPT_SUPPORTED
|
|
entropy->dc_count_ptrs[i] = entropy->ac_count_ptrs[i] = NULL;
|
|
#endif
|
|
}
|
|
}
|
|
|
|
//mark buxton's new emit_bits:
|
|
|
|
static unsigned int onlynbits[] = {
|
|
|
|
0x00000000, 0x00000001, 0x00000003, 0x00000007, 0x000000F, 0x0000001F,
|
|
0x0000003F, 0x0000007F, 0x000000FF, 0x000001FF, 0x000003FF, 0x000007FF,
|
|
0x00000FFF, 0x00001FFF, 0x00003FFF, 0x00007FFF, 0x0000FFFF, 0x0001FFFF,
|
|
0x0003FFFF, 0x0007FFFF, 0x000FFFFF, 0x001FFFFF, 0x003FFFFF, 0x007FFFFF,
|
|
0x00FFFFFF, 0x01FFFFFF, 0x03FFFFFF, 0x07FFFFFF, 0x0FFFFFFF, 0x1FFFFFFF,
|
|
0x3FFFFFFF, 0x7FFFFFFF, 0xFFFFFFFF
|
|
};
|
|
|
|
|
|
//
|
|
// Need to add #ifdef for Alpha port
|
|
//
|
|
#if defined (_X86_)
|
|
|
|
GLOBAL(boolean)
|
|
emit_bits_fast (working_state * state, unsigned int code, int bsize, int only1){
|
|
// Emit some bits; return TRUE if successful, FALSE if must suspend
|
|
// This routine is heavily used, so it's worth coding tightly.
|
|
//unsigned int put_buffer = code;
|
|
|
|
// int put_bits = state->put_bits;
|
|
unsigned c;
|
|
|
|
__asm{
|
|
mov edx,64
|
|
mov esi,[put_bits]
|
|
|
|
add esi,dword ptr[bsize]
|
|
|
|
mov [put_bits],esi
|
|
sub edx,esi
|
|
|
|
mov ebx,[free_in_buffer]
|
|
|
|
movd mm3,edx
|
|
mov edx,[next_output_byte]
|
|
|
|
|
|
movd mm0,[code]
|
|
|
|
movq mm7,[put_buffer_64]
|
|
psllq mm0,mm3 //put_buffer <<= 64-put_bits;
|
|
|
|
por mm7,mm0
|
|
cmp [only1],0
|
|
|
|
movq mm0,mm7
|
|
jne got_FF
|
|
|
|
cmp ebx,8
|
|
jng got_FF
|
|
|
|
cmp esi,32
|
|
jng buffer_not_full
|
|
|
|
// test [next_output_byte],0x3
|
|
// jnz byte_write
|
|
//test to see if the data is on a 4-byte boundary. If not, don't use the
|
|
//integer write.
|
|
//integer_write: //Write 32 bits.
|
|
|
|
movq mm1,mm5
|
|
psrlq mm0,32
|
|
|
|
pcmpeqb mm1,mm0
|
|
sub ebx,4
|
|
|
|
movd eax,mm1
|
|
movq mm2,mm0 // | - | - | - | - | D | C | B | A | MM2
|
|
|
|
test eax,eax
|
|
jne got_FF
|
|
// big-endian data
|
|
psrlq mm0,8 // | - | - | - | - | - | D | C | B | MM0
|
|
mov [free_in_buffer],ebx
|
|
|
|
punpcklbw mm0,mm2 // | - | - | C | D | B | C | A | B | MM0
|
|
add edx,4
|
|
|
|
pslld mm0,16 // | C | D | B | C | A | B | - | - | MM0
|
|
sub esi,32
|
|
|
|
psrad mm0,16 //have to use this pair because packssdw expects 16-byte _signed_ data.
|
|
psllq mm7,32
|
|
|
|
packssdw mm0,mm0 // | - | - | - | - | C | D | A | B | MM0
|
|
mov [put_bits],esi
|
|
|
|
movq mm2,mm0 // | - | - | - | - | C | D | A | B | MM2
|
|
movq [put_buffer_64],mm7
|
|
|
|
psrlq mm0,16 // | - | - | - | - | - | - | C | D | MM0
|
|
mov [next_output_byte],edx
|
|
|
|
punpcklwd mm0,mm2 // | - | - | - | - | C | B | C | D | MM0
|
|
|
|
movd [edx-4],mm0
|
|
nop
|
|
|
|
}
|
|
return TRUE;
|
|
|
|
got_FF: //only if an FF was returned.
|
|
while (put_bits >=8) {
|
|
|
|
__asm{
|
|
movq mm4,mm7
|
|
psrlq mm7,56
|
|
|
|
movd [c],mm7
|
|
psllq mm4,8
|
|
|
|
movq mm7,mm4
|
|
sub [put_bits],8
|
|
}
|
|
|
|
emit_byte(state, c, return FALSE);
|
|
//emit_byte_fast(c);
|
|
if (c == 0xFF) emit_byte(state, 0, return FALSE);
|
|
//if (c==0xFF) emit_byte_fast(0);
|
|
//put_bits -= 8;
|
|
}
|
|
|
|
buffer_not_full:
|
|
__asm movq [put_buffer_64],mm7
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
/* Encode a single block's worth of coefficients */
|
|
|
|
GLOBAL(boolean)
|
|
encode_one_block_fast (working_state * state, JCOEFPTR block, int last_dc_val,
|
|
c_derived_tbl *dctbl, c_derived_tbl *actbl)
|
|
{
|
|
JCOEF temp ;
|
|
int nbits;
|
|
int k, i ,j;
|
|
/*unsigned <-Buxton BUG??*/ int l;
|
|
int lastzeros = 0, numElements = 0 ;
|
|
|
|
short dummy_outblock[192+4] ; // 64 values, 64 corresponding zero & bit counts
|
|
short *outblock;
|
|
outblock= (short *)(((unsigned int)dummy_outblock+7)&0xFFFFFFF8);
|
|
|
|
|
|
// Encode the DC coefficient difference per section F.1.2.1
|
|
//starttimer();
|
|
|
|
temp = block[0] ;
|
|
block[0] = (JCOEF) (block[0] - last_dc_val) ;
|
|
countZeros((int *)jpeg_natural_order, block, outblock, &lastzeros, &numElements) ;
|
|
//formatting for frequently used MMX registers in emit_bits_fast
|
|
__asm{
|
|
//provide two constants: mm5 = 0xFFFF|FFFF|FFFF|FFFF
|
|
// mm6 = 0x0000|FFFF|0000|FFFF
|
|
pcmpeqb mm5,mm5
|
|
pxor mm0,mm0
|
|
movq mm6,mm5
|
|
punpcklwd mm6,mm0
|
|
|
|
}
|
|
if(block[0] == 0) {
|
|
// Emit the Huffman-coded symbol for the number of bits
|
|
//changed to only1 - dshade
|
|
if (! emit_bits_fast(state, dctbl->ehufco[0], dctbl->ehufsi[0],1)) return FALSE;
|
|
|
|
outblock[64] -- ;
|
|
k=0;
|
|
} else {
|
|
nbits = outblock[128] ;
|
|
k=1;
|
|
//changed to only1 - dshade
|
|
if (!emit_bits_fast(state,dctbl->ehufco[nbits]<<nbits | ((unsigned int) outblock[0]&onlynbits[nbits]),nbits + dctbl->ehufsi[nbits],1)) return FALSE;
|
|
}
|
|
block[0] = temp ; // get the original value of block[0] back in.
|
|
|
|
if( (numElements == 0) || (numElements == 1) ) {
|
|
lastzeros = 63 ; // DC element handled outside of the AC loop below
|
|
}
|
|
|
|
|
|
|
|
// Encode the AC coefficients per section F.1.2.2
|
|
|
|
for (; k < numElements; k++) {
|
|
|
|
l = outblock[64+k];//<<4;
|
|
//store frequently used lookups:
|
|
j = outblock[128+k];
|
|
// if run length > 15, must emit special run-length-16 codes (0xF0)
|
|
while (l > 15/*240*/) {
|
|
// changed to only1 - dshade
|
|
if (! emit_bits_fast(state, actbl->ehufco[0xF0], actbl->ehufsi[0xF0],1))
|
|
return FALSE;
|
|
l -= 16; //256;
|
|
}
|
|
|
|
// if (l < 0) // dshade
|
|
// l = 0; // dshade
|
|
|
|
l = l << 4;
|
|
|
|
i = l + j;
|
|
|
|
//hufandval = actbl->ehufco[i]<<j | ((unsigned int) outblock[k]&onlynbits[j]);
|
|
//hufandvallen = j + actbl->ehufsi[i];
|
|
|
|
//changed to only1 - dshade
|
|
if (!emit_bits_fast(state,actbl->ehufco[i]<<j | ((unsigned int) outblock[k]&onlynbits[j]),j + actbl->ehufsi[i],1)) return FALSE;
|
|
|
|
}
|
|
|
|
// If the last coef(s) were zero, emit an end-of-block code
|
|
if (lastzeros > 0)
|
|
{ //changed to only1 - dshade
|
|
if (! emit_bits_fast(state, actbl->ehufco[0], actbl->ehufsi[0],1))
|
|
return FALSE;
|
|
}
|
|
|
|
//cumulative_time += stoptimer();
|
|
|
|
_asm emms // dshade
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
//used above as:
|
|
//countZeros((int *)jpeg_natural_order, block, outblock, &lastzeros, &numElements) ;
|
|
__int32 jmpswitch;
|
|
|
|
void countZeros(
|
|
int *dwindexBlock,
|
|
short *dwcoefBlock,
|
|
short *dwoutBlock,
|
|
int *dwlastZeros,
|
|
int *dwnumElem
|
|
){
|
|
|
|
static __int64 const_1 = 0x0001000100010001;
|
|
static __int64 const_2 = 0x0002000200020002;
|
|
static __int64 const_3 = 0x0003000300030003;
|
|
static __int64 const_4 = 0x0004000400040004;
|
|
static __int64 const_8 = 0x0008000800080008;
|
|
static __int64 const_15 = 0x000f000f000f000f;
|
|
static __int64 const_255 = 0x00ff00ff00ff00ff;
|
|
|
|
//#define sizLOCALS 8
|
|
|
|
//_countZeros proc USES eax ebx ecx edx esi edi
|
|
|
|
// Move all paramters to be based on esp
|
|
// Must REMEMBER NOT to use the stack for any push/pops
|
|
|
|
// the following are the parameters passed into the routine.
|
|
//#define dwindexBlock dword ptr [esp+32+sizLOCALS] // 32 bit elements
|
|
//#define dwcoefBlock dword ptr [esp+36+sizLOCALS] // 16 bit elements
|
|
//#define dwoutBlock dword ptr [esp+40+sizLOCALS] // 16 bit elements
|
|
//#define dwlastZeros dword ptr [esp+44+sizLOCALS] // address of a 32 bit element
|
|
//#define dwnumElem dword ptr [esp+48+sizLOCALS] // number of non-zero elements
|
|
// dwlastZeros stores the number of trailing zero values.
|
|
|
|
//;;;;; LOCALS :;;;;;;;;;;;;;;;
|
|
__int32 locdwoutBlock;
|
|
__int32 locdwZeroCount;
|
|
__int32 loopctr;
|
|
// these are used as scratchpad registers.
|
|
// right now a new local has been added on an as needed
|
|
// basis. There's potential for reducing the number of locals.
|
|
//#define locdwoutBlock dword ptr [esp+0]
|
|
//#define locdwZeroCount dword ptr [esp+4]
|
|
|
|
//;;;;; END OF LOCALS ;;;;;;;;;;;;;;;
|
|
__asm{
|
|
//sub esp, sizLOCALS
|
|
|
|
|
|
mov esi, dwindexBlock ; load the input array pointer
|
|
mov edi, dwcoefBlock
|
|
|
|
mov eax, dwoutBlock
|
|
nop //************************;;
|
|
|
|
mov locdwoutBlock, eax
|
|
nop //************************;;
|
|
|
|
mov dword ptr[loopctr], 10h // loop count of 16 : four elements handled per loop
|
|
mov locdwZeroCount, 0h // initialize zero counter to 0
|
|
|
|
CountZeroLoop:
|
|
// align the zigzag elements of inblock into MMX register, four words
|
|
// at a time.
|
|
|
|
// get index for next four elements in coeff array from the zigzag array
|
|
mov eax, [esi]
|
|
mov ebx, [esi+4]
|
|
|
|
mov ecx, [esi+8]
|
|
mov edx, [esi+12]
|
|
|
|
// get the next four coeff. words
|
|
|
|
mov eax, [edi+2*eax]
|
|
mov ebx, [edi+2*ebx]
|
|
|
|
mov ecx, [edi+2*ecx]
|
|
mov edx, [edi+2*edx]
|
|
|
|
// pack first two words in eax (first word in LS 16 bits)
|
|
shl ebx, 16
|
|
and eax, 0ffffh
|
|
|
|
// pack next two words in ecx (third word in LS 16 bits)
|
|
shl edx, 16
|
|
and ecx, 0ffffh
|
|
|
|
or eax, ebx
|
|
or ecx, edx
|
|
|
|
mov ebx, eax
|
|
or eax, ecx // check to see if all 4 elems. are zero
|
|
|
|
cmp eax, 0h
|
|
jz caseAllZeros
|
|
|
|
movd mm0, ebx // move LS two words into mm0
|
|
pxor mm2, mm2 // initialize mm2 to zero
|
|
|
|
movd mm1, ecx // move MS two words into mm1
|
|
nop //************************;;
|
|
|
|
pcmpeqw mm0, mm2
|
|
pcmpeqw mm1, mm2
|
|
|
|
movq mm3,mm0
|
|
por mm0,mm1
|
|
|
|
movd eax,mm0
|
|
pcmpeqw mm2,mm2
|
|
|
|
cmp eax,0h
|
|
jz caseNoZeros
|
|
|
|
movd eax,mm3
|
|
pandn mm1,mm2
|
|
|
|
movd edx,mm1
|
|
not eax
|
|
|
|
and eax,0x00020001
|
|
and edx,0x00080004
|
|
|
|
or eax,edx
|
|
add esi,16
|
|
|
|
mov edx,eax
|
|
shr eax,16
|
|
|
|
or eax,edx
|
|
mov edx, locdwZeroCount
|
|
|
|
and eax,0xFFFF
|
|
nop
|
|
|
|
dec eax
|
|
|
|
lea eax,[JmpTable+eax*8]
|
|
jmp eax
|
|
|
|
|
|
JmpTable:
|
|
jmp case0001
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case0010
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case0011
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case0100
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case0101
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case0110
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case0111
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case1000
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case1001
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case1010
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case1011
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case1100
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case1101
|
|
nop
|
|
nop
|
|
nop
|
|
jmp case1110
|
|
nop
|
|
nop
|
|
nop
|
|
jmp caseNoZeros
|
|
|
|
|
|
caseAllZeros:
|
|
|
|
add locdwZeroCount, 4
|
|
add esi, 16
|
|
|
|
dec [loopctr] // decrement loop counter
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
|
|
caseNoZeros:
|
|
|
|
mov eax, locdwoutBlock
|
|
mov edx, locdwZeroCount
|
|
|
|
mov locdwZeroCount, 0h
|
|
add esi, 16 // esi points to a 32 bit quantity
|
|
|
|
mov [eax], ebx // store the LS two words
|
|
mov [eax+4], ecx // store the MS two words
|
|
|
|
add locdwoutBlock, 8
|
|
mov [eax+128], edx
|
|
|
|
mov dword ptr [eax+132], 0
|
|
nop //************************;;
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
// case0000:
|
|
// this case is taken care of by caseAllZero
|
|
|
|
case0001:
|
|
|
|
mov eax, locdwoutBlock
|
|
mov locdwZeroCount, 3
|
|
|
|
mov [eax], bx // store the LS word
|
|
mov [eax+128], dx //; store the corresponding zero count
|
|
|
|
add locdwoutBlock, 2
|
|
nop //************************;;
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
case0010:
|
|
|
|
mov eax, locdwoutBlock
|
|
mov locdwZeroCount, 2
|
|
|
|
shr ebx, 16 // get the MS word into LS 16 bits
|
|
add edx, 1 // increment zero count
|
|
|
|
mov [eax+128], dx // store the corresponding zero count
|
|
nop //************************;;
|
|
|
|
mov [eax], bx // store the LS word
|
|
add locdwoutBlock, 2
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
case0011:
|
|
|
|
mov eax, locdwoutBlock
|
|
mov locdwZeroCount, 2
|
|
|
|
mov [eax], ebx // store the LS word
|
|
mov [eax+128], edx // store the corresponding zero count
|
|
|
|
add locdwoutBlock, 4
|
|
nop //************************;;
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
case0100:
|
|
|
|
mov eax, locdwoutBlock
|
|
add edx, 2
|
|
|
|
mov [eax], cx // store the LS word within MS DWORD
|
|
mov [eax+128], dx // store the corresponding zero count
|
|
|
|
add locdwoutBlock, 2
|
|
mov locdwZeroCount, 1
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
case0101:
|
|
|
|
mov eax, locdwoutBlock
|
|
or edx, 10000h // zero count is 1 for second word
|
|
|
|
mov [eax], bx // store the LS word within LS DWORD
|
|
mov [eax+2], cx // store the LS word within MS DWORD
|
|
|
|
mov [eax+128], edx // store the corresponding zero count
|
|
nop //************************;;
|
|
|
|
add locdwoutBlock, 4
|
|
mov locdwZeroCount, 1
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
case0110:
|
|
|
|
mov eax, locdwoutBlock
|
|
add edx, 1 // zero count is incremented for first word
|
|
|
|
shr ebx, 16 // move the word to be written into LS 16 bits
|
|
mov [eax], bx // store the LS word within LS DWORD
|
|
|
|
mov [eax+2], cx // store the LS word within MS DWORD
|
|
add locdwoutBlock, 4
|
|
|
|
mov [eax+128], edx // store the corresponding zero count
|
|
mov locdwZeroCount, 1
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
case0111:
|
|
|
|
mov eax, locdwoutBlock
|
|
nop
|
|
|
|
add locdwoutBlock, 6
|
|
mov locdwZeroCount, 1
|
|
|
|
mov [eax], ebx // store the LS word within LS DWORD
|
|
mov [eax+4], cx // store the LS word within MS DWORD
|
|
|
|
mov [eax+128], edx // store the corresponding zero count
|
|
mov word ptr [eax+132], 0 // zerocount of 0 for third word
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
case1000:
|
|
|
|
mov eax, locdwoutBlock
|
|
add edx, 3
|
|
|
|
|
|
shr ecx, 16
|
|
nop //************************;;
|
|
|
|
mov [eax], cx // store the LS word within MS DWORD
|
|
mov [eax+128], dx // store the corresponding zero count
|
|
|
|
add locdwoutBlock, 2
|
|
mov locdwZeroCount, 0
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
case1001:
|
|
|
|
mov eax, locdwoutBlock
|
|
shr ecx, 16 // word 3 into LS bits
|
|
|
|
or edx, 00020000h //// zero count of two for MS word
|
|
mov [eax], bx // store the LS word within MS DWORD
|
|
|
|
mov [eax+2], cx // store the LS word within MS DWORD
|
|
add locdwoutBlock, 4
|
|
|
|
mov [eax+128], edx // store the corresponding zero count
|
|
mov locdwZeroCount, 0
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
case1010:
|
|
|
|
mov eax, locdwoutBlock
|
|
nop
|
|
|
|
add edx, 1 // increment zero count
|
|
shr ecx, 16 // word 3 into LS bits
|
|
|
|
shr ebx, 16 // word 2 into LS bits
|
|
or edx, 00010000h // zero count of two for MS word
|
|
|
|
mov [eax], bx // store the LS word within MS DWORD
|
|
mov [eax+2], cx // store the LS word within MS DWORD
|
|
|
|
mov [eax+128], edx // store the corresponding zero count
|
|
//add esi, 16 // esi points to a 32 bit quantity
|
|
|
|
add locdwoutBlock, 4
|
|
mov locdwZeroCount, 0
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
|
|
case1011:
|
|
|
|
|
|
mov eax, locdwoutBlock
|
|
shr ecx, 16 // word 3 into LS bits
|
|
|
|
mov [eax], ebx // store the LS DWORD
|
|
mov [eax+4], cx // store the LS word within MS DWORD
|
|
|
|
mov [eax+128], edx // store the corresponding zero count
|
|
mov word ptr [eax+132], 1
|
|
|
|
add locdwoutBlock, 6
|
|
mov locdwZeroCount, 0
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
|
|
case1100:
|
|
|
|
mov eax, locdwoutBlock
|
|
add edx, 2 // add 2 to zeroc count
|
|
|
|
mov [eax], ecx // store the LS DWORD
|
|
mov [eax+128], edx // store the corresponding zero count
|
|
|
|
add locdwoutBlock, 4
|
|
mov locdwZeroCount, 0
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
case1101:
|
|
|
|
mov eax, locdwoutBlock
|
|
nop ////************************////
|
|
|
|
mov [eax], bx // store the LS DWORD
|
|
mov [eax+128], dx // store the corresponding zero count
|
|
|
|
mov [eax+2], ecx
|
|
mov dword ptr [eax+130], 1 // zero count of 1 for 2nd word
|
|
|
|
add locdwoutBlock, 6
|
|
mov locdwZeroCount, 0
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
case1110:
|
|
|
|
mov eax, locdwoutBlock
|
|
shr ebx, 16 // get word 1 in LS word
|
|
|
|
add dx, 1 // add 1 to zerocount
|
|
nop ////************************////
|
|
|
|
mov [eax], bx // store the LS DWORD
|
|
mov [eax+128], dx // store the corresponding zero count
|
|
|
|
mov [eax+2], ecx
|
|
mov dword ptr [eax+130], 0 // zero count of 0 for 2nd & 3rd word
|
|
|
|
add locdwoutBlock, 6
|
|
mov locdwZeroCount, 0
|
|
|
|
dec [loopctr]
|
|
jnz CountZeroLoop
|
|
|
|
jmp AllDone
|
|
|
|
|
|
|
|
// case1111:
|
|
// this case is handled by caseNoZeros
|
|
|
|
|
|
////////////////////////////////////////////////////////////////////////////////////////////////////////////
|
|
|
|
AllDone:
|
|
|
|
// at this point all zero counting is done
|
|
// now get the number of non-zero elements and round to the
|
|
// nearest multiple of four in the +infinity direction
|
|
|
|
mov eax, locdwoutBlock
|
|
|
|
sub eax, dwoutBlock // how many non-zero elements did you write
|
|
mov ebx, dwnumElem // address to store number of non-zero elements
|
|
|
|
//*** small bug fix by dshade 4/8/97 to eliminate case where loop count went negative below
|
|
//*** all changes have //*** by them
|
|
mov edx, eax //*** make a copy before shifting
|
|
|
|
shr eax, 1 // a/c for each element being 2 bytes
|
|
|
|
//// round the number of elements to nearest multiple of four
|
|
mov [ebx], eax
|
|
//add eax, 4
|
|
add edx, 7 //*** add get an even multiple of 8
|
|
|
|
and edx, 0x1f8 //*** edx holds the number of writes rounded up to 8
|
|
//and eax, 0fch // get the LS 2 bits to be zero
|
|
mov esi, dwoutBlock
|
|
|
|
//// This loop should count the number of bits needed to represent
|
|
//// the input value. It handles four inputs in each iteration
|
|
|
|
CountBitsLoop:
|
|
|
|
movq mm0, [esi] // get the first four input data
|
|
pxor mm7, mm7 // clear mm7
|
|
|
|
movq mm1, mm0
|
|
pcmpgtw mm0, mm7 // is input number positive ?
|
|
|
|
movq mm3, mm1
|
|
pand mm1, mm0 // original number if greater than 0, else 0
|
|
|
|
psubw mm7, mm3 // 0 minus input number
|
|
movq mm2, mm0
|
|
|
|
psubw mm3, qword ptr [const_1] // decrement input
|
|
pandn mm0, mm7 // if number < 0 then (- input), else 0
|
|
|
|
por mm0, mm1 // abs(input)
|
|
pandn mm2, mm3 // input minus 1 if input < 0, else 0
|
|
|
|
por mm2, mm1 // same as input, if input positive, else input minus 1.
|
|
nop ////************************////
|
|
|
|
movq mm3, mm0
|
|
movq mm1, mm0
|
|
|
|
pcmpgtw mm1, qword ptr [const_255] // split the 16bit value across bit 8 (256)
|
|
psrlw mm0, 8 // get MS 8 bits into LS byte
|
|
|
|
movq [esi], mm2 // store input (if it's +ve), else store 1s comp. of input
|
|
movq mm2, mm1
|
|
|
|
pand mm1, qword ptr [const_8] // value > 255 implies need for 8 bits, else zero
|
|
pand mm0, mm2 // sift the ones greater than 255, else zeros
|
|
|
|
pandn mm2, mm3 // sift the ones less than 256, else zeros
|
|
movq mm5, mm0
|
|
|
|
por mm0, mm2 // get reqd. portions of data in LS 8 bits
|
|
por mm5, mm2 // copy of above instruction
|
|
|
|
pcmpgtw mm5, qword ptr [const_15]
|
|
movq mm4, mm0
|
|
|
|
movq mm3, mm5
|
|
psrlw mm0, 4
|
|
|
|
pand mm5, qword ptr [const_4]
|
|
pand mm0, mm3
|
|
|
|
movq mm2, mm0
|
|
pandn mm3, mm4
|
|
|
|
por mm0, mm3
|
|
por mm2, mm3
|
|
|
|
pcmpgtw mm2, qword ptr [const_3]
|
|
movq mm4, mm0
|
|
|
|
movq mm3, mm2
|
|
psrlw mm0, 2
|
|
|
|
pand mm2, qword ptr [const_2]
|
|
pand mm0, mm3
|
|
|
|
movq mm6, mm0
|
|
pandn mm3, mm4
|
|
|
|
por mm0, mm3
|
|
por mm6, mm3
|
|
|
|
pcmpgtw mm6, qword ptr [const_1]
|
|
movq mm4, mm0
|
|
|
|
movq mm3, mm6
|
|
psrlw mm0, 1
|
|
|
|
pand mm6, qword ptr [const_1]
|
|
pand mm0, mm3
|
|
|
|
por mm1, mm5
|
|
pandn mm3, mm4
|
|
|
|
por mm6, mm2
|
|
por mm0, mm3
|
|
|
|
por mm1, mm6
|
|
nop ////************************////
|
|
|
|
paddw mm0, mm1
|
|
nop ////************************////
|
|
|
|
movq [esi+256], mm0
|
|
nop ////************************////
|
|
|
|
add esi, 8
|
|
nop ////************************////
|
|
|
|
//sub eax, 4
|
|
sub edx, 8 //*** decrement byte count
|
|
jg CountBitsLoop //*** changed loop conditions to break out if not positive
|
|
|
|
|
|
|
|
|
|
mov eax, locdwZeroCount
|
|
mov ebx, dwlastZeros
|
|
|
|
//emms
|
|
//nop ////************************////
|
|
|
|
mov [ebx], eax
|
|
//add esp, sizLOCALS
|
|
|
|
|
|
}
|
|
return;
|
|
}
|
|
|
|
#endif // #define (_X86_)
|