mc-a.asm

DM642 H.264 codec DM642 H.264 codec DM642 H.264 codec DM642 H.264 codec

;*****************************************************************************
;* mc.asm: h264 encoder library
;*****************************************************************************
;* Copyright (C) 2003 x264 project
;* $Id: mc.asm,v 1.3 2004/06/18 01:59:58 chenm001 Exp $
;*
;* Authors: Min Chen <chenm001.163.com> (converted to nasm)
;*          Laurent Aimar <[e-mail]> (init algorithm)
;*
;* This program is free software; you can redistribute it and/or modify
;* it under the terms of the GNU General Public License as published by
;* the Free Software Foundation; either version 2 of the License, or
;* (at your option) any later version.
;*
;* This program is distributed in the hope that it will be useful,
;* but WITHOUT ANY WARRANTY; without even the implied warranty of
;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
;* GNU General Public License for more details.
;*
;* You should have received a copy of the GNU General Public License
;* along with this program; if not, write to the Free Software
;* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111, USA.
;*****************************************************************************

;*****************************************************************************
;*                                                                           *
;*  Revision history:                                                        *
;*                                                                           *
;*  2004.05.17 portab mc_copy_w4/8/16 (CM)                                   *
;*                                                                           *
;*****************************************************************************

BITS 32

;=============================================================================
; Macros and other preprocessor constants
;=============================================================================

%include "i386inc.asm"

;=============================================================================
; Constants
;=============================================================================

SECTION_RODATA

ALIGN 16
pw_4:  times 4 dw  4
pw_8:  times 4 dw  8
pw_32: times 4 dw 32
pw_64: times 4 dw 64

;=============================================================================
; Code
;=============================================================================

SECTION .text

cglobal x264_pixel_avg_w4_mmxext
cglobal x264_pixel_avg_w8_mmxext
cglobal x264_pixel_avg_w16_mmxext
cglobal x264_pixel_avg_w16_sse2

cglobal x264_pixel_avg_weight_4x4_mmxext
cglobal x264_pixel_avg_weight_w8_mmxext
cglobal x264_pixel_avg_weight_w16_mmxext

cglobal x264_mc_copy_w4_mmx
cglobal x264_mc_copy_w8_mmx
cglobal x264_mc_copy_w16_mmx
cglobal x264_mc_copy_w16_sse2

cglobal x264_mc_chroma_mmxext

;=============================================================================
; pixel avg
;=============================================================================

ALIGN 16
;-----------------------------------------------------------------------------
; void x264_pixel_avg_w4_mmxext( uint8_t *dst,  int i_dst_stride,
;                                uint8_t *src1, int i_src1_stride,
;                                uint8_t *src2, int i_src2_stride,
;                                int i_height );
;-----------------------------------------------------------------------------
x264_pixel_avg_w4_mmxext:
    push        ebp
    push        ebx
    push        esi
    push        edi

    mov         edi, [esp+20]       ; dst
    mov         ebx, [esp+28]       ; src1
    mov         ecx, [esp+36]       ; src2
    mov         esi, [esp+24]       ; i_dst_stride
    mov         eax, [esp+32]       ; i_src1_stride
    mov         edx, [esp+40]       ; i_src2_stride
    mov         ebp, [esp+44]       ; i_height
ALIGN 4
.height_loop    
    movd        mm0, [ebx]
    pavgb       mm0, [ecx]
    movd        mm1, [ebx+eax]
    pavgb       mm1, [ecx+edx]
    movd        [edi], mm0
    movd        [edi+esi], mm1
    dec         ebp
    dec         ebp
    lea         ebx, [ebx+eax*2]
    lea         ecx, [ecx+edx*2]
    lea         edi, [edi+esi*2]
    jne         .height_loop

    pop         edi
    pop         esi
    pop         ebx
    pop         ebp
    ret

                          

ALIGN 16
;-----------------------------------------------------------------------------
; void x264_pixel_avg_w8_mmxext( uint8_t *dst,  int i_dst_stride,
;                                uint8_t *src1, int i_src1_stride,
;                                uint8_t *src2, int i_src2_stride,
;                                int i_height );
;-----------------------------------------------------------------------------
x264_pixel_avg_w8_mmxext:
    push        ebp
    push        ebx
    push        esi
    push        edi

    mov         edi, [esp+20]       ; dst
    mov         ebx, [esp+28]       ; src1
    mov         ecx, [esp+36]       ; src2
    mov         esi, [esp+24]       ; i_dst_stride
    mov         eax, [esp+32]       ; i_src1_stride
    mov         edx, [esp+40]       ; i_src2_stride
    mov         ebp, [esp+44]       ; i_height
ALIGN 4
.height_loop    
    movq        mm0, [ebx]
    pavgb       mm0, [ecx]
    movq        [edi], mm0
    dec         ebp
    lea         ebx, [ebx+eax]
    lea         ecx, [ecx+edx]
    lea         edi, [edi+esi]
    jne         .height_loop

    pop         edi
    pop         esi
    pop         ebx
    pop         ebp
    ret



ALIGN 16
;-----------------------------------------------------------------------------
; void x264_pixel_avg_w16_mmxext( uint8_t *dst,  int i_dst_stride,
;                                 uint8_t *src1, int i_src1_stride,
;                                 uint8_t *src2, int i_src2_stride,
;                                 int i_height );
;-----------------------------------------------------------------------------
x264_pixel_avg_w16_mmxext:
    push        ebp
    push        ebx
    push        esi
    push        edi

    mov         edi, [esp+20]       ; dst
    mov         ebx, [esp+28]       ; src1
    mov         ecx, [esp+36]       ; src2
    mov         esi, [esp+24]       ; i_dst_stride
    mov         eax, [esp+32]       ; i_src1_stride
    mov         edx, [esp+40]       ; i_src2_stride
    mov         ebp, [esp+44]       ; i_height
ALIGN 4
.height_loop    
    movq        mm0, [ebx  ]
    movq        mm1, [ebx+8]
    pavgb       mm0, [ecx  ]
    pavgb       mm1, [ecx+8]
    movq        [edi  ], mm0
    movq        [edi+8], mm1
    dec         ebp
    lea         ebx, [ebx+eax]
    lea         ecx, [ecx+edx]
    lea         edi, [edi+esi]
    jne         .height_loop

    pop         edi
    pop         esi
    pop         ebx
    pop         ebp
    ret

ALIGN 16
;-----------------------------------------------------------------------------
; void x264_pixel_avg_w16_sse2( uint8_t *dst,  int i_dst_stride,
;                               uint8_t *src1, int i_src1_stride,
;                               uint8_t *src2, int i_src2_stride,
;                               int i_height );
;-----------------------------------------------------------------------------
x264_pixel_avg_w16_sse2:
    push        ebp
    push        ebx
    push        esi
    push        edi

    mov         edi, [esp+20]       ; dst
    mov         ebx, [esp+28]       ; src1
    mov         ecx, [esp+36]       ; src2
    mov         esi, [esp+24]       ; i_dst_stride
    mov         eax, [esp+32]       ; i_src1_stride
    mov         edx, [esp+40]       ; i_src2_stride
    mov         ebp, [esp+44]       ; i_height
ALIGN 4
.height_loop    
    movdqu      xmm0, [ebx]
    pavgb       xmm0, [ecx]
    movdqu      [edi], xmm0

    dec         ebp
    lea         ebx, [ebx+eax]
    lea         ecx, [ecx+edx]
    lea         edi, [edi+esi]
    jne         .height_loop

    pop         edi
    pop         esi
    pop         ebx
    pop         ebp
    ret


;=============================================================================
; weighted prediction
;=============================================================================
; implicit bipred only:
; assumes log2_denom = 5, offset = 0, weight1 + weight2 = 64

%macro BIWEIGHT_4P_MMX 2
    movd      mm0, %1
    movd      mm1, %2
    punpcklbw mm0, mm7
    punpcklbw mm1, mm7
    pmullw    mm0, mm4
    pmullw    mm1, mm5
    paddw     mm0, mm1
    paddw     mm0, mm6
    psraw     mm0, 6
    pmaxsw    mm0, mm7
    packuswb  mm0, mm0
    movd      %1,  mm0
%endmacro

%macro BIWEIGHT_START_MMX 0
    push    edi
    push    esi
    picpush ebx
    picgetgot ebx
    mov     edi, [picesp+12]    ; dst
    mov     esi, [picesp+16]    ; i_dst
    mov     edx, [picesp+20]    ; src
    mov     ecx, [picesp+24]    ; i_src

    pshufw  mm4, [picesp+28], 0  ; weight_dst
    movq    mm5, [pw_64 GOT_ebx]
    psubw   mm5, mm4             ; weight_src
    movq    mm6, [pw_32 GOT_ebx] ; rounding
    pxor    mm7, mm7
%endmacro
%macro BIWEIGHT_END_MMX 0
    picpop  ebx
    pop     esi
    pop     edi
    ret
%endmacro

ALIGN 16
;-----------------------------------------------------------------------------
;   int __cdecl x264_pixel_avg_weight_w16_mmxext( uint8_t *, int, uint8_t *, int, int, int )
;-----------------------------------------------------------------------------
x264_pixel_avg_weight_w16_mmxext:
    BIWEIGHT_START_MMX
    mov     eax, [picesp+32] ; i_height
    ALIGN 4
    .height_loop