ffmpeg/libavcodec/x86/hevc_sao_10bit.asm
Lynne bbe95f7353
x86: replace explicit REP_RETs with RETs
From x86inc:
> On AMD cpus <=K10, an ordinary ret is slow if it immediately follows either
> a branch or a branch target. So switch to a 2-byte form of ret in that case.
> We can automatically detect "follows a branch", but not a branch target.
> (SSSE3 is a sufficient condition to know that your cpu doesn't have this problem.)

x86inc can automatically determine whether to use REP_RET rather than
REP in most of these cases, so impact is minimal. Additionally, a few
REP_RETs were used unnecessary, despite the return being nowhere near a
branch.

The only CPUs affected were AMD K10s, made between 2007 and 2011, 16
years ago and 12 years ago, respectively.

In the future, everyone involved with x86inc should consider dropping
REP_RETs altogether.
2023-02-01 04:23:55 +01:00

371 lines
10 KiB
NASM

;******************************************************************************
;* SIMD optimized SAO functions for HEVC 10/12bit decoding
;*
;* Copyright (c) 2013 Pierre-Edouard LEPERE
;* Copyright (c) 2014 James Almer
;*
;* This file is part of FFmpeg.
;*
;* FFmpeg is free software; you can redistribute it and/or
;* modify it under the terms of the GNU Lesser General Public
;* License as published by the Free Software Foundation; either
;* version 2.1 of the License, or (at your option) any later version.
;*
;* FFmpeg 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
;* Lesser General Public License for more details.
;*
;* You should have received a copy of the GNU Lesser General Public
;* License along with FFmpeg; if not, write to the Free Software
;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
;******************************************************************************
%include "libavutil/x86/x86util.asm"
SECTION_RODATA 32
pw_m2: times 16 dw -2
pw_mask10: times 16 dw 0x03FF
pw_mask12: times 16 dw 0x0FFF
pb_eo: db -1, 0, 1, 0, 0, -1, 0, 1, -1, -1, 1, 1, 1, -1, -1, 1
cextern pw_m1
cextern pw_1
cextern pw_2
SECTION .text
;******************************************************************************
;SAO Band Filter
;******************************************************************************
%macro HEVC_SAO_BAND_FILTER_INIT 1
and leftq, 31
movd xm0, leftd
add leftq, 1
and leftq, 31
movd xm1, leftd
add leftq, 1
and leftq, 31
movd xm2, leftd
add leftq, 1
and leftq, 31
movd xm3, leftd
SPLATW m0, xm0
SPLATW m1, xm1
SPLATW m2, xm2
SPLATW m3, xm3
%if mmsize > 16
SPLATW m4, [offsetq + 2]
SPLATW m5, [offsetq + 4]
SPLATW m6, [offsetq + 6]
SPLATW m7, [offsetq + 8]
%else
movq m7, [offsetq + 2]
SPLATW m4, m7, 0
SPLATW m5, m7, 1
SPLATW m6, m7, 2
SPLATW m7, m7, 3
%endif
%if ARCH_X86_64
mova m13, [pw_mask %+ %1]
pxor m14, m14
%else ; ARCH_X86_32
mova [rsp+mmsize*0], m0
mova [rsp+mmsize*1], m1
mova [rsp+mmsize*2], m2
mova [rsp+mmsize*3], m3
mova [rsp+mmsize*4], m4
mova [rsp+mmsize*5], m5
mova [rsp+mmsize*6], m6
mova m1, [pw_mask %+ %1]
pxor m0, m0
%define m14 m0
%define m13 m1
%define m9 m2
%define m8 m3
%endif ; ARCH
DEFINE_ARGS dst, src, dststride, srcstride, offset, height
mov heightd, r7m
%endmacro
;void ff_hevc_sao_band_filter_<width>_<depth>_<opt>(uint8_t *_dst, const uint8_t *_src, ptrdiff_t _stride_dst, ptrdiff_t _stride_src,
; int16_t *sao_offset_val, int sao_left_class, int width, int height);
%macro HEVC_SAO_BAND_FILTER 3
cglobal hevc_sao_band_filter_%2_%1, 6, 6, 15, 7*mmsize*ARCH_X86_32, dst, src, dststride, srcstride, offset, left
HEVC_SAO_BAND_FILTER_INIT %1
align 16
.loop:
%assign i 0
%assign j 0
%rep %3
%assign k 8+(j&1)
%assign l 9-(j&1)
mova m %+ k, [srcq + i]
psraw m %+ l, m %+ k, %1-5
%if ARCH_X86_64
pcmpeqw m10, m %+ l, m0
pcmpeqw m11, m %+ l, m1
pcmpeqw m12, m %+ l, m2
pcmpeqw m %+ l, m3
pand m10, m4
pand m11, m5
pand m12, m6
pand m %+ l, m7
por m10, m11
por m12, m %+ l
por m10, m12
paddw m %+ k, m10
%else ; ARCH_X86_32
pcmpeqw m4, m %+ l, [rsp+mmsize*0]
pcmpeqw m5, m %+ l, [rsp+mmsize*1]
pcmpeqw m6, m %+ l, [rsp+mmsize*2]
pcmpeqw m %+ l, [rsp+mmsize*3]
pand m4, [rsp+mmsize*4]
pand m5, [rsp+mmsize*5]
pand m6, [rsp+mmsize*6]
pand m %+ l, m7
por m4, m5
por m6, m %+ l
por m4, m6
paddw m %+ k, m4
%endif ; ARCH
CLIPW m %+ k, m14, m13
mova [dstq + i], m %+ k
%assign i i+mmsize
%assign j j+1
%endrep
add dstq, dststrideq
add srcq, srcstrideq
dec heightd
jg .loop
RET
%endmacro
%macro HEVC_SAO_BAND_FILTER_FUNCS 0
HEVC_SAO_BAND_FILTER 10, 8, 1
HEVC_SAO_BAND_FILTER 10, 16, 2
HEVC_SAO_BAND_FILTER 10, 32, 4
HEVC_SAO_BAND_FILTER 10, 48, 6
HEVC_SAO_BAND_FILTER 10, 64, 8
HEVC_SAO_BAND_FILTER 12, 8, 1
HEVC_SAO_BAND_FILTER 12, 16, 2
HEVC_SAO_BAND_FILTER 12, 32, 4
HEVC_SAO_BAND_FILTER 12, 48, 6
HEVC_SAO_BAND_FILTER 12, 64, 8
%endmacro
INIT_XMM sse2
HEVC_SAO_BAND_FILTER_FUNCS
INIT_XMM avx
HEVC_SAO_BAND_FILTER_FUNCS
%if HAVE_AVX2_EXTERNAL
INIT_XMM avx2
HEVC_SAO_BAND_FILTER 10, 8, 1
INIT_YMM avx2
HEVC_SAO_BAND_FILTER 10, 16, 1
HEVC_SAO_BAND_FILTER 10, 32, 2
HEVC_SAO_BAND_FILTER 10, 48, 3
HEVC_SAO_BAND_FILTER 10, 64, 4
INIT_XMM avx2
HEVC_SAO_BAND_FILTER 12, 8, 1
INIT_YMM avx2
HEVC_SAO_BAND_FILTER 12, 16, 1
HEVC_SAO_BAND_FILTER 12, 32, 2
HEVC_SAO_BAND_FILTER 12, 48, 3
HEVC_SAO_BAND_FILTER 12, 64, 4
%endif
;******************************************************************************
;SAO Edge Filter
;******************************************************************************
%define MAX_PB_SIZE 64
%define PADDING_SIZE 64 ; AV_INPUT_BUFFER_PADDING_SIZE
%define EDGE_SRCSTRIDE 2 * MAX_PB_SIZE + PADDING_SIZE
%macro PMINUW 4
%if cpuflag(sse4)
pminuw %1, %2, %3
%else
psubusw %4, %2, %3
psubw %1, %2, %4
%endif
%endmacro
%macro HEVC_SAO_EDGE_FILTER_INIT 0
%if WIN64
movsxd eoq, dword eom
%elif ARCH_X86_64
movsxd eoq, eod
%else
mov eoq, r4m
%endif
lea tmp2q, [pb_eo]
movsx a_strideq, byte [tmp2q+eoq*4+1]
movsx b_strideq, byte [tmp2q+eoq*4+3]
imul a_strideq, EDGE_SRCSTRIDE >> 1
imul b_strideq, EDGE_SRCSTRIDE >> 1
movsx tmpq, byte [tmp2q+eoq*4]
add a_strideq, tmpq
movsx tmpq, byte [tmp2q+eoq*4+2]
add b_strideq, tmpq
%endmacro
;void ff_hevc_sao_edge_filter_<width>_<depth>_<opt>(uint8_t *_dst, uint8_t *_src, ptrdiff_t stride_dst, int16_t *sao_offset_val,
; int eo, int width, int height);
%macro HEVC_SAO_EDGE_FILTER 3
%if ARCH_X86_64
cglobal hevc_sao_edge_filter_%2_%1, 4, 9, 16, dst, src, dststride, offset, eo, a_stride, b_stride, height, tmp
%define tmp2q heightq
HEVC_SAO_EDGE_FILTER_INIT
mov heightd, r6m
add a_strideq, a_strideq
add b_strideq, b_strideq
%else ; ARCH_X86_32
cglobal hevc_sao_edge_filter_%2_%1, 1, 6, 8, 5*mmsize, dst, src, dststride, a_stride, b_stride, height
%define eoq srcq
%define tmpq heightq
%define tmp2q dststrideq
%define offsetq heightq
%define m8 m1
%define m9 m2
%define m10 m3
%define m11 m4
%define m12 m5
HEVC_SAO_EDGE_FILTER_INIT
mov srcq, srcm
mov offsetq, r3m
mov dststrideq, dststridem
add a_strideq, a_strideq
add b_strideq, b_strideq
%endif ; ARCH
%if mmsize > 16
SPLATW m8, [offsetq+2]
SPLATW m9, [offsetq+4]
SPLATW m10, [offsetq+0]
SPLATW m11, [offsetq+6]
SPLATW m12, [offsetq+8]
%else
movq m10, [offsetq+0]
movd m12, [offsetq+6]
SPLATW m8, xm10, 1
SPLATW m9, xm10, 2
SPLATW m10, xm10, 0
SPLATW m11, xm12, 0
SPLATW m12, xm12, 1
%endif
pxor m0, m0
%if ARCH_X86_64
mova m13, [pw_m1]
mova m14, [pw_1]
mova m15, [pw_2]
%else
mov heightd, r6m
mova [rsp+mmsize*0], m8
mova [rsp+mmsize*1], m9
mova [rsp+mmsize*2], m10
mova [rsp+mmsize*3], m11
mova [rsp+mmsize*4], m12
%endif
align 16
.loop:
%assign i 0
%rep %3
mova m1, [srcq + i]
movu m2, [srcq+a_strideq + i]
movu m3, [srcq+b_strideq + i]
PMINUW m4, m1, m2, m6
PMINUW m5, m1, m3, m7
pcmpeqw m2, m4
pcmpeqw m3, m5
pcmpeqw m4, m1
pcmpeqw m5, m1
psubw m4, m2
psubw m5, m3
paddw m4, m5
pcmpeqw m2, m4, [pw_m2]
%if ARCH_X86_64
pcmpeqw m3, m4, m13
pcmpeqw m5, m4, m0
pcmpeqw m6, m4, m14
pcmpeqw m7, m4, m15
pand m2, m8
pand m3, m9
pand m5, m10
pand m6, m11
pand m7, m12
%else
pcmpeqw m3, m4, [pw_m1]
pcmpeqw m5, m4, m0
pcmpeqw m6, m4, [pw_1]
pcmpeqw m7, m4, [pw_2]
pand m2, [rsp+mmsize*0]
pand m3, [rsp+mmsize*1]
pand m5, [rsp+mmsize*2]
pand m6, [rsp+mmsize*3]
pand m7, [rsp+mmsize*4]
%endif
paddw m2, m3
paddw m5, m6
paddw m2, m7
paddw m2, m1
paddw m2, m5
CLIPW m2, m0, [pw_mask %+ %1]
mova [dstq + i], m2
%assign i i+mmsize
%endrep
add dstq, dststrideq
add srcq, EDGE_SRCSTRIDE
dec heightd
jg .loop
RET
%endmacro
INIT_XMM sse2
HEVC_SAO_EDGE_FILTER 10, 8, 1
HEVC_SAO_EDGE_FILTER 10, 16, 2
HEVC_SAO_EDGE_FILTER 10, 32, 4
HEVC_SAO_EDGE_FILTER 10, 48, 6
HEVC_SAO_EDGE_FILTER 10, 64, 8
HEVC_SAO_EDGE_FILTER 12, 8, 1
HEVC_SAO_EDGE_FILTER 12, 16, 2
HEVC_SAO_EDGE_FILTER 12, 32, 4
HEVC_SAO_EDGE_FILTER 12, 48, 6
HEVC_SAO_EDGE_FILTER 12, 64, 8
%if HAVE_AVX2_EXTERNAL
INIT_XMM avx2
HEVC_SAO_EDGE_FILTER 10, 8, 1
INIT_YMM avx2
HEVC_SAO_EDGE_FILTER 10, 16, 1
HEVC_SAO_EDGE_FILTER 10, 32, 2
HEVC_SAO_EDGE_FILTER 10, 48, 3
HEVC_SAO_EDGE_FILTER 10, 64, 4
INIT_XMM avx2
HEVC_SAO_EDGE_FILTER 12, 8, 1
INIT_YMM avx2
HEVC_SAO_EDGE_FILTER 12, 16, 1
HEVC_SAO_EDGE_FILTER 12, 32, 2
HEVC_SAO_EDGE_FILTER 12, 48, 3
HEVC_SAO_EDGE_FILTER 12, 64, 4
%endif