generic-library/vpx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
2d4ef37507
vpx/vp8/encoder/x86/sad_mmx.asm

430 lines
11 KiB
NASM
Raw Normal View History

Initial WebM release
2010-05-18 17:58:33 +02:00
;
Use WebM in copyright notice for consistency Changes 'The VP8 project' to 'The WebM project', for consistency with other webmproject.org repositories. Fixes issue #97. Change-Id: I37c13ed5fbdb9d334ceef71c6350e9febed9bbba
2010-09-09 14:16:39 +02:00
; Copyright (c) 2010 The WebM project authors. All Rights Reserved.
Initial WebM release
2010-05-18 17:58:33 +02:00
;
cosmetics: trim trailing whitespace When the license headers were updated, they accidentally contained trailing whitespace, so unfortunately we have to touch all the files again. Change-Id: I236c05fade06589e417179c0444cb39b09e4200d
2010-06-18 18:39:21 +02:00
; Use of this source code is governed by a BSD-style license
LICENSE: update with latest text Change-Id: Ieebea089095d9073b3a94932791099f614ce120c
2010-06-04 22:19:40 +02:00
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
cosmetics: trim trailing whitespace When the license headers were updated, they accidentally contained trailing whitespace, so unfortunately we have to touch all the files again. Change-Id: I236c05fade06589e417179c0444cb39b09e4200d
2010-06-18 18:39:21 +02:00
; in the file PATENTS. All contributing project authors may
LICENSE: update with latest text Change-Id: Ieebea089095d9073b3a94932791099f614ce120c
2010-06-04 22:19:40 +02:00
; be found in the AUTHORS file in the root of the source tree.
Initial WebM release
2010-05-18 17:58:33 +02:00
;
%include "vpx_ports/x86_abi_support.asm"
global sym(vp8_sad16x16_mmx)
global sym(vp8_sad8x16_mmx)
global sym(vp8_sad8x8_mmx)
global sym(vp8_sad4x4_mmx)
global sym(vp8_sad16x8_mmx)
%idefine QWORD
;unsigned int vp8_sad16x16_mmx(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride)
sym(vp8_sad16x16_mmx):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
lea rcx, [rsi+rax*8]
lea rcx, [rcx+rax*8]
pxor mm7, mm7
pxor mm6, mm6
x16x16sad_mmx_loop:
movq mm0, QWORD PTR [rsi]
movq mm2, QWORD PTR [rsi+8]
movq mm1, QWORD PTR [rdi]
movq mm3, QWORD PTR [rdi+8]
movq mm4, mm0
movq mm5, mm2
psubusb mm0, mm1
psubusb mm1, mm4
psubusb mm2, mm3
psubusb mm3, mm5
por mm0, mm1
por mm2, mm3
movq mm1, mm0
movq mm3, mm2
punpcklbw mm0, mm6
punpcklbw mm2, mm6
punpckhbw mm1, mm6
punpckhbw mm3, mm6
paddw mm0, mm2
paddw mm1, mm3
lea rsi, [rsi+rax]
add rdi, rdx
paddw mm7, mm0
paddw mm7, mm1
cmp rsi, rcx
jne x16x16sad_mmx_loop
movq mm0, mm7
punpcklwd mm0, mm6
punpckhwd mm7, mm6
paddw mm0, mm7
movq mm7, mm0
psrlq mm0, 32
paddw mm7, mm0
movd rax, mm7
pop rdi
pop rsi
mov rsp, rbp
; begin epilog
UNSHADOW_ARGS
pop rbp
ret
;unsigned int vp8_sad8x16_mmx(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride)
sym(vp8_sad8x16_mmx):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
lea rcx, [rsi+rax*8]
lea rcx, [rcx+rax*8]
pxor mm7, mm7
pxor mm6, mm6
x8x16sad_mmx_loop:
movq mm0, QWORD PTR [rsi]
movq mm1, QWORD PTR [rdi]
movq mm2, mm0
psubusb mm0, mm1
psubusb mm1, mm2
por mm0, mm1
movq mm2, mm0
punpcklbw mm0, mm6
punpckhbw mm2, mm6
lea rsi, [rsi+rax]
add rdi, rdx
paddw mm7, mm0
paddw mm7, mm2
cmp rsi, rcx
jne x8x16sad_mmx_loop
movq mm0, mm7
punpcklwd mm0, mm6
punpckhwd mm7, mm6
paddw mm0, mm7
movq mm7, mm0
psrlq mm0, 32
paddw mm7, mm0
movd rax, mm7
pop rdi
pop rsi
mov rsp, rbp
; begin epilog
UNSHADOW_ARGS
pop rbp
ret
;unsigned int vp8_sad8x8_mmx(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride)
sym(vp8_sad8x8_mmx):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
lea rcx, [rsi+rax*8]
pxor mm7, mm7
pxor mm6, mm6
x8x8sad_mmx_loop:
movq mm0, QWORD PTR [rsi]
movq mm1, QWORD PTR [rdi]
movq mm2, mm0
psubusb mm0, mm1
psubusb mm1, mm2
por mm0, mm1
movq mm2, mm0
punpcklbw mm0, mm6
punpckhbw mm2, mm6
paddw mm0, mm2
lea rsi, [rsi+rax]
add rdi, rdx
paddw mm7, mm0
cmp rsi, rcx
jne x8x8sad_mmx_loop
movq mm0, mm7
punpcklwd mm0, mm6
punpckhwd mm7, mm6
paddw mm0, mm7
movq mm7, mm0
psrlq mm0, 32
paddw mm7, mm0
movd rax, mm7
pop rdi
pop rsi
mov rsp, rbp
; begin epilog
UNSHADOW_ARGS
pop rbp
ret
;unsigned int vp8_sad4x4_mmx(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride)
sym(vp8_sad4x4_mmx):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
movd mm0, QWORD PTR [rsi]
movd mm1, QWORD PTR [rdi]
movd mm2, QWORD PTR [rsi+rax]
movd mm3, QWORD PTR [rdi+rdx]
punpcklbw mm0, mm2
punpcklbw mm1, mm3
movq mm2, mm0
psubusb mm0, mm1
psubusb mm1, mm2
por mm0, mm1
movq mm2, mm0
pxor mm3, mm3
punpcklbw mm0, mm3
punpckhbw mm2, mm3
paddw mm0, mm2
lea rsi, [rsi+rax*2]
lea rdi, [rdi+rdx*2]
movd mm4, QWORD PTR [rsi]
movd mm5, QWORD PTR [rdi]
movd mm6, QWORD PTR [rsi+rax]
movd mm7, QWORD PTR [rdi+rdx]
punpcklbw mm4, mm6
punpcklbw mm5, mm7
movq mm6, mm4
psubusb mm4, mm5
psubusb mm5, mm6
por mm4, mm5
movq mm5, mm4
punpcklbw mm4, mm3
punpckhbw mm5, mm3
paddw mm4, mm5
paddw mm0, mm4
movq mm1, mm0
punpcklwd mm0, mm3
punpckhwd mm1, mm3
paddw mm0, mm1
movq mm1, mm0
psrlq mm0, 32
paddw mm0, mm1
movd rax, mm0
pop rdi
pop rsi
mov rsp, rbp
; begin epilog
UNSHADOW_ARGS
pop rbp
ret
;unsigned int vp8_sad16x8_mmx(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride)
sym(vp8_sad16x8_mmx):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
lea rcx, [rsi+rax*8]
pxor mm7, mm7
pxor mm6, mm6
x16x8sad_mmx_loop:
movq mm0, [rsi]
movq mm1, [rdi]
movq mm2, [rsi+8]
movq mm3, [rdi+8]
movq mm4, mm0
movq mm5, mm2
psubusb mm0, mm1
psubusb mm1, mm4
psubusb mm2, mm3
psubusb mm3, mm5
por mm0, mm1
por mm2, mm3
movq mm1, mm0
movq mm3, mm2
punpcklbw mm0, mm6
punpckhbw mm1, mm6
punpcklbw mm2, mm6
punpckhbw mm3, mm6
paddw mm0, mm2
paddw mm1, mm3
paddw mm0, mm1
lea rsi, [rsi+rax]
add rdi, rdx
paddw mm7, mm0
cmp rsi, rcx
jne x16x8sad_mmx_loop
movq mm0, mm7
punpcklwd mm0, mm6
punpckhwd mm7, mm6
paddw mm0, mm7
movq mm7, mm0
psrlq mm0, 32
paddw mm7, mm0
movd rax, mm7
pop rdi
pop rsi
mov rsp, rbp
; begin epilog
UNSHADOW_ARGS
pop rbp
ret
Reference in New Issue Copy Permalink