1 |
;/************************************************************************** |
;/***************************************************************************** |
2 |
; * |
; * |
3 |
; * XVID MPEG-4 VIDEO CODEC |
; * XVID MPEG-4 VIDEO CODEC |
4 |
; * xmm 8x8 block-based halfpel interpolation |
; * - mmx 8x8 block-based halfpel interpolation - |
5 |
|
; * |
6 |
|
; * Copyright(C) 2002 Michael Militzer <isibaar@xvid.org> |
7 |
|
; * 2002 Pascal Massimino <skal@planet-d.net> |
8 |
; * |
; * |
9 |
; * This program is free software; you can redistribute it and/or modify |
; * This program is free software; you can redistribute it and/or modify |
10 |
; * it under the terms of the GNU General Public License as published by |
; * it under the terms of the GNU General Public License as published by |
18 |
; * |
; * |
19 |
; * You should have received a copy of the GNU General Public License |
; * You should have received a copy of the GNU General Public License |
20 |
; * along with this program; if not, write to the Free Software |
; * along with this program; if not, write to the Free Software |
21 |
; * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. |
; * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA |
|
; * |
|
|
; *************************************************************************/ |
|
|
|
|
|
;/************************************************************************** |
|
|
; * |
|
|
; * History: |
|
|
; * |
|
|
; * 04.06.2002 rewrote some funcs, mostly XMM. -Skal- |
|
|
; * Heavily tuned for overlap and AGI-stalls avoidance |
|
|
; * 04.02.2002 initial version (Isibaar) |
|
22 |
; * |
; * |
23 |
; *************************************************************************/ |
; ****************************************************************************/ |
24 |
|
|
25 |
|
BITS 32 |
|
bits 32 |
|
26 |
|
|
27 |
%macro cglobal 1 |
%macro cglobal 1 |
28 |
%ifdef PREFIX |
%ifdef PREFIX |
33 |
%endif |
%endif |
34 |
%endmacro |
%endmacro |
35 |
|
|
36 |
section .data |
;============================================================================= |
37 |
|
; Read only data |
38 |
|
;============================================================================= |
39 |
|
|
40 |
align 16 |
%ifdef FORMAT_COFF |
41 |
|
SECTION .rodata data |
42 |
|
%else |
43 |
|
SECTION .rodata data align=16 |
44 |
|
%endif |
45 |
|
|
46 |
mmx_one |
ALIGN 16 |
47 |
|
mmx_one: |
48 |
times 8 db 1 |
times 8 db 1 |
49 |
|
|
50 |
section .text |
SECTION .text |
51 |
|
|
52 |
cglobal interpolate8x8_halfpel_h_xmm |
cglobal interpolate8x8_halfpel_h_xmm |
53 |
cglobal interpolate8x8_halfpel_v_xmm |
cglobal interpolate8x8_halfpel_v_xmm |
92 |
movq [ecx+edx], mm1 |
movq [ecx+edx], mm1 |
93 |
%endmacro |
%endmacro |
94 |
|
|
95 |
align 16 |
ALIGN 16 |
96 |
interpolate8x8_halfpel_h_xmm: |
interpolate8x8_halfpel_h_xmm: |
97 |
|
|
98 |
mov eax, [esp+16]; rounding |
mov eax, [esp+16]; rounding |
162 |
movq [ecx+edx], mm1 |
movq [ecx+edx], mm1 |
163 |
%endmacro |
%endmacro |
164 |
|
|
165 |
align 16 |
ALIGN 16 |
166 |
interpolate8x8_halfpel_v_xmm: |
interpolate8x8_halfpel_v_xmm: |
167 |
|
|
168 |
mov eax, [esp+16]; rounding |
mov eax, [esp+16]; rounding |
172 |
mov edx, [esp+12] ; stride |
mov edx, [esp+12] ; stride |
173 |
|
|
174 |
; we process 2 line at a time |
; we process 2 line at a time |
|
|
|
175 |
jnz near .rounding1 |
jnz near .rounding1 |
176 |
|
|
177 |
COPY_V_SSE_RND0 |
COPY_V_SSE_RND0 |
298 |
movq [ecx], mm0 |
movq [ecx], mm0 |
299 |
%endmacro |
%endmacro |
300 |
|
|
301 |
align 16 |
ALIGN 16 |
302 |
interpolate8x8_halfpel_hv_xmm: |
interpolate8x8_halfpel_hv_xmm: |
303 |
mov eax, [esp+16] ; rounding |
mov eax, [esp+16] ; rounding |
304 |
mov ecx, [esp+ 4] ; Dst |
mov ecx, [esp+ 4] ; Dst |