[FFmpeg-devel] [PATCH v2 1/5] lavc/vp9dsp: R-V V mc avg

flow gg hlefthleft at gmail.com
Sun May 26 15:27:55 EEST 2024


Hi, maybe we can prioritize this revert:
https://git.ffmpeg.org/gitweb/ffmpeg.git/commit/0c1304ae11b0361ede055ee8ffc6e83529468c73
Using [PATCH v2 1/5] lavc/vp9dsp: R-V V mc avg to avoid conflicts with
other patches.

flow gg <hlefthleft at gmail.com> 于2024年5月24日周五 14:13写道:

> I want to update the VP9 bilin load, just like you did with VP8, but it
> seems like this patch([PATCH v2 1/5] lavc/vp9dsp: R-V V mc avg) doesn't
> merge the current updates here but merges the previous version instead, so
> the subsequent patches will have conflicts.
>
> flow gg <hlefthleft at gmail.com> 于2024年5月22日周三 01:15写道:
>
>> > Please put commas between operands.
>> > This should probably be ff_avg_vp9 or something slightly more specific.
>>
>> Updated here.
>>
>> <uk7b at foxmail.com> 于2024年5月22日周三 01:14写道:
>>
>>> From: sunyuechi <sunyuechi at iscas.ac.cn>
>>>
>>> C908:
>>> vp9_avg4_8bpp_c: 1.2
>>> vp9_avg4_8bpp_rvv_i64: 1.0
>>> vp9_avg8_8bpp_c: 3.7
>>> vp9_avg8_8bpp_rvv_i64: 1.5
>>> vp9_avg16_8bpp_c: 14.7
>>> vp9_avg16_8bpp_rvv_i64: 3.5
>>> vp9_avg32_8bpp_c: 57.7
>>> vp9_avg32_8bpp_rvv_i64: 10.0
>>> vp9_avg64_8bpp_c: 229.0
>>> vp9_avg64_8bpp_rvv_i64: 31.7
>>> ---
>>>  libavcodec/riscv/Makefile      |  3 +-
>>>  libavcodec/riscv/vp9_mc_rvv.S  | 58 ++++++++++++++++++++++++++++++++++
>>>  libavcodec/riscv/vp9dsp.h      |  4 +--
>>>  libavcodec/riscv/vp9dsp_init.c | 18 +++++++++++
>>>  4 files changed, 80 insertions(+), 3 deletions(-)
>>>  create mode 100644 libavcodec/riscv/vp9_mc_rvv.S
>>>
>>> diff --git a/libavcodec/riscv/Makefile b/libavcodec/riscv/Makefile
>>> index 07d5c2915d..67e198d754 100644
>>> --- a/libavcodec/riscv/Makefile
>>> +++ b/libavcodec/riscv/Makefile
>>> @@ -69,6 +69,7 @@ RVV-OBJS-$(CONFIG_VP8DSP) += riscv/vp8dsp_rvv.o
>>>  OBJS-$(CONFIG_VP9_DECODER) += riscv/vp9dsp_init.o
>>>  RV-OBJS-$(CONFIG_VP9_DECODER) += riscv/vp9_intra_rvi.o \
>>>                                   riscv/vp9_mc_rvi.o
>>> -RVV-OBJS-$(CONFIG_VP9_DECODER) += riscv/vp9_intra_rvv.o
>>> +RVV-OBJS-$(CONFIG_VP9_DECODER) += riscv/vp9_intra_rvv.o \
>>> +                                  riscv/vp9_mc_rvv.o
>>>  OBJS-$(CONFIG_VORBIS_DECODER) += riscv/vorbisdsp_init.o
>>>  RVV-OBJS-$(CONFIG_VORBIS_DECODER) += riscv/vorbisdsp_rvv.o
>>> diff --git a/libavcodec/riscv/vp9_mc_rvv.S
>>> b/libavcodec/riscv/vp9_mc_rvv.S
>>> new file mode 100644
>>> index 0000000000..7cb38ec94a
>>> --- /dev/null
>>> +++ b/libavcodec/riscv/vp9_mc_rvv.S
>>> @@ -0,0 +1,58 @@
>>> +/*
>>> + * Copyright (c) 2024 Institue of Software Chinese Academy of Sciences
>>> (ISCAS).
>>> + *
>>> + * 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/riscv/asm.S"
>>> +
>>> +.macro vsetvlstatic8 len an maxlen mn=m4
>>> +.if \len == 4
>>> +        vsetivli        zero, \len, e8, mf4, ta, ma
>>> +.elseif \len == 8
>>> +        vsetivli        zero, \len, e8, mf2, ta, ma
>>> +.elseif \len == 16
>>> +        vsetivli        zero, \len, e8, m1, ta, ma
>>> +.elseif \len == 32
>>> +        li              \an, \len
>>> +        vsetvli         zero, \an, e8, m2, ta, ma
>>> +.elseif \len == 64
>>> +        li              \an, \maxlen
>>> +        vsetvli         zero, \an, e8, \mn, ta, ma
>>> +.endif
>>> +.endm
>>> +
>>> +.macro copy_avg len
>>> +func ff_vp9_avg\len\()_rvv, zve32x
>>> +        csrwi           vxrm, 0
>>> +        vsetvlstatic8   \len, t0, 64
>>> +1:
>>> +        vle8.v          v8, (a2)
>>> +        vle8.v          v16, (a0)
>>> +        vaaddu.vv       v8, v8, v16
>>> +        addi            a4, a4, -1
>>> +        vse8.v          v8, (a0)
>>> +        add             a2, a2, a3
>>> +        add             a0, a0, a1
>>> +        bnez            a4, 1b
>>> +        ret
>>> +endfunc
>>> +.endm
>>> +
>>> +.irp len, 64, 32, 16, 8, 4
>>> +        copy_avg \len
>>> +.endr
>>> diff --git a/libavcodec/riscv/vp9dsp.h b/libavcodec/riscv/vp9dsp.h
>>> index 79330b4968..ff8431591c 100644
>>> --- a/libavcodec/riscv/vp9dsp.h
>>> +++ b/libavcodec/riscv/vp9dsp.h
>>> @@ -138,11 +138,11 @@ void ff_avg_bilin_##SIZE##hv_rvv(uint8_t *dst,
>>> ptrdiff_t dststride,        \
>>>                                   int h, int mx, int my);
>>>
>>>  #define VP9_COPY_AVG_RISCV_RVV_FUNC(SIZE)                           \
>>> -void ff_copy##SIZE##_rvv(uint8_t *dst, ptrdiff_t dststride,        \
>>> +void ff_vp9_copy##SIZE##_rvv(uint8_t *dst, ptrdiff_t dststride,    \
>>>                           const uint8_t *src, ptrdiff_t srcstride,  \
>>>                           int h, int mx, int my);                   \
>>>                                                                     \
>>> -void ff_avg##SIZE##_rvv(uint8_t *dst, ptrdiff_t dststride,         \
>>> +void ff_vp9_avg##SIZE##_rvv(uint8_t *dst, ptrdiff_t dststride,     \
>>>                          const uint8_t *src, ptrdiff_t srcstride,   \
>>>                          int h, int mx, int my);
>>>
>>> diff --git a/libavcodec/riscv/vp9dsp_init.c
>>> b/libavcodec/riscv/vp9dsp_init.c
>>> index ab99294d44..454dcd963f 100644
>>> --- a/libavcodec/riscv/vp9dsp_init.c
>>> +++ b/libavcodec/riscv/vp9dsp_init.c
>>> @@ -48,6 +48,24 @@ static av_cold void
>>> vp9dsp_mc_init_riscv(VP9DSPContext *dsp, int bpp)
>>>      }
>>>  # endif
>>>
>>> +#if HAVE_RVV
>>> +    if (bpp == 8 && (flags & AV_CPU_FLAG_RVV_I32) &&
>>> ff_rv_vlen_least(128)) {
>>> +
>>> +#define init_fpel(idx1, sz)                                           \
>>> +    dsp->mc[idx1][FILTER_8TAP_SMOOTH ][1][0][0] =
>>> ff_vp9_avg##sz##_rvv;  \
>>> +    dsp->mc[idx1][FILTER_8TAP_REGULAR][1][0][0] =
>>> ff_vp9_avg##sz##_rvv;  \
>>> +    dsp->mc[idx1][FILTER_8TAP_SHARP  ][1][0][0] =
>>> ff_vp9_avg##sz##_rvv;  \
>>> +    dsp->mc[idx1][FILTER_BILINEAR    ][1][0][0] = ff_vp9_avg##sz##_rvv
>>> +
>>> +    init_fpel(0, 64);
>>> +    init_fpel(1, 32);
>>> +    init_fpel(2, 16);
>>> +    init_fpel(3, 8);
>>> +    init_fpel(4, 4);
>>> +
>>> +#undef init_fpel
>>> +    }
>>> +#endif
>>>  #endif
>>>  }
>>>
>>> --
>>> 2.45.1
>>>
>>> _______________________________________________
>>> ffmpeg-devel mailing list
>>> ffmpeg-devel at ffmpeg.org
>>> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>>>
>>> To unsubscribe, visit link above, or email
>>> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".
>>>
>>


More information about the ffmpeg-devel mailing list