From: Ramiro Polla <ramiro.polla@gmail.com>
To: ffmpeg-devel@ffmpeg.org
Subject: Re: [FFmpeg-devel] [PATCH 3/4] swscale/x86: add sse4 {lum, chr}ConvertRange
Date: Fri, 7 Jun 2024 19:38:26 +0200
Message-ID: <CALweWgAwdiXYXB5X=ZSa9MW0NJg1u4u2opNjMw+HK61Vw2KKVA@mail.gmail.com> (raw)
In-Reply-To: <20240607140543.130761-3-ramiro.polla@gmail.com>
[-- Attachment #1: Type: text/plain, Size: 10195 bytes --]
On Fri, Jun 7, 2024 at 4:05 PM Ramiro Polla <ramiro.polla@gmail.com> wrote:
>
> chrRangeFromJpeg_8_c: 19.9
> chrRangeFromJpeg_8_sse4: 16.2
> chrRangeFromJpeg_24_c: 60.7
> chrRangeFromJpeg_24_sse4: 28.9
> chrRangeFromJpeg_128_c: 325.7
> chrRangeFromJpeg_128_sse4: 160.2
> chrRangeFromJpeg_144_c: 364.2
> chrRangeFromJpeg_144_sse4: 194.9
> chrRangeFromJpeg_256_c: 630.7
> chrRangeFromJpeg_256_sse4: 337.4
> chrRangeFromJpeg_512_c: 1240.4
> chrRangeFromJpeg_512_sse4: 668.4
> chrRangeToJpeg_8_c: 37.7
> chrRangeToJpeg_8_sse4: 19.7
> chrRangeToJpeg_24_c: 114.7
> chrRangeToJpeg_24_sse4: 30.2
> chrRangeToJpeg_128_c: 636.4
> chrRangeToJpeg_128_sse4: 161.7
> chrRangeToJpeg_144_c: 715.7
> chrRangeToJpeg_144_sse4: 272.9
> chrRangeToJpeg_256_c: 1256.7
> chrRangeToJpeg_256_sse4: 341.9
> chrRangeToJpeg_512_c: 2498.7
> chrRangeToJpeg_512_sse4: 668.4
> lumRangeFromJpeg_8_c: 11.7
> lumRangeFromJpeg_8_sse4: 12.4
> lumRangeFromJpeg_24_c: 36.9
> lumRangeFromJpeg_24_sse4: 17.7
> lumRangeFromJpeg_128_c: 228.4
> lumRangeFromJpeg_128_sse4: 85.2
> lumRangeFromJpeg_144_c: 272.9
> lumRangeFromJpeg_144_sse4: 96.9
> lumRangeFromJpeg_256_c: 463.4
> lumRangeFromJpeg_256_sse4: 183.9
> lumRangeFromJpeg_512_c: 879.9
> lumRangeFromJpeg_512_sse4: 355.9
> lumRangeToJpeg_8_c: 17.7
> lumRangeToJpeg_8_sse4: 15.4
> lumRangeToJpeg_24_c: 56.2
> lumRangeToJpeg_24_sse4: 18.4
> lumRangeToJpeg_128_c: 331.4
> lumRangeToJpeg_128_sse4: 84.4
> lumRangeToJpeg_144_c: 375.2
> lumRangeToJpeg_144_sse4: 96.9
> lumRangeToJpeg_256_c: 649.7
> lumRangeToJpeg_256_sse4: 184.4
> lumRangeToJpeg_512_c: 1281.9
> lumRangeToJpeg_512_sse4: 355.9
> ---
> libswscale/swscale_internal.h | 1 +
> libswscale/utils.c | 2 +
> libswscale/x86/Makefile | 1 +
> libswscale/x86/range_convert.asm | 100 +++++++++++++++++++++++++++++++
> libswscale/x86/swscale.c | 36 +++++++++++
> 5 files changed, 140 insertions(+)
> create mode 100644 libswscale/x86/range_convert.asm
>
> diff --git a/libswscale/swscale_internal.h b/libswscale/swscale_internal.h
> index d4b0c3cee2..92f6105443 100644
> --- a/libswscale/swscale_internal.h
> +++ b/libswscale/swscale_internal.h
> @@ -698,6 +698,7 @@ void ff_updateMMXDitherTables(SwsContext *c, int dstY);
>
> av_cold void ff_sws_init_range_convert(SwsContext *c);
> av_cold void ff_sws_init_range_convert_loongarch(SwsContext *c);
> +av_cold void ff_sws_init_range_convert_x86(SwsContext *c);
>
> SwsFunc ff_yuv2rgb_init_x86(SwsContext *c);
> SwsFunc ff_yuv2rgb_init_ppc(SwsContext *c);
> diff --git a/libswscale/utils.c b/libswscale/utils.c
> index 476a24fea5..8dfa57b5ff 100644
> --- a/libswscale/utils.c
> +++ b/libswscale/utils.c
> @@ -1082,6 +1082,8 @@ int sws_setColorspaceDetails(struct SwsContext *c, const int inv_table[4],
> ff_sws_init_range_convert(c);
> #if ARCH_LOONGARCH64
> ff_sws_init_range_convert_loongarch(c);
> +#elif ARCH_X86
> + ff_sws_init_range_convert_x86(c);
> #endif
> }
>
> diff --git a/libswscale/x86/Makefile b/libswscale/x86/Makefile
> index 68391494be..f00154941d 100644
> --- a/libswscale/x86/Makefile
> +++ b/libswscale/x86/Makefile
> @@ -12,6 +12,7 @@ X86ASM-OBJS += x86/input.o \
> x86/output.o \
> x86/scale.o \
> x86/scale_avx2.o \
> + x86/range_convert.o \
> x86/rgb_2_rgb.o \
> x86/yuv_2_rgb.o \
> x86/yuv2yuvX.o \
> diff --git a/libswscale/x86/range_convert.asm b/libswscale/x86/range_convert.asm
> new file mode 100644
> index 0000000000..333265fb65
> --- /dev/null
> +++ b/libswscale/x86/range_convert.asm
> @@ -0,0 +1,100 @@
> +;******************************************************************************
> +;* Copyright (c) 2024 Ramiro Polla
> +;*
> +;* 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"
> +
> +; NOTE: there is no need to clamp the input when converting to jpeg range
> +; (like we do in the C code) because packssdw will saturate the output.
> +
> +;-----------------------------------------------------------------------------
> +; lumConvertRange
> +;
> +; void ff_lumRangeToJpeg_<opt>(int16_t *dst, int width);
> +; void ff_lumRangeFromJpeg_<opt>(int16_t *dst, int width);
> +;
> +;-----------------------------------------------------------------------------
> +
> +%macro LUMCONVERTRANGE 4
> +SECTION_RODATA
> +mult_%1: times 4 dd %2
> +offset_%1: times 4 dd %3
> +SECTION .text
> +cglobal %1, 2, 3, 3, dst, width, x
> + movsxdifnidn widthq, widthd
> + xor xq, xq
> + mova m1, [mult_%1]
> + mova m2, [offset_%1]
> +.loop:
> + pmovsxwd m0, [dstq+xq*2]
> + pmulld m0, m1
> + paddd m0, m2
> + psrad m0, %4
> + packssdw m0, m0
> + movh [dstq+xq*2], m0
> + add xq, mmsize / 4
> + cmp xd, widthd
> + jl .loop
> + RET
> +%endmacro
> +
> +;-----------------------------------------------------------------------------
> +; chrConvertRange
> +;
> +; void ff_chrRangeToJpeg_<opt>(int16_t *dstU, int16_t *dstV, int width);
> +; void ff_chrRangeFromJpeg_<opt>(int16_t *dstU, int16_t *dstV, int width);
> +;
> +;-----------------------------------------------------------------------------
> +
> +%macro CHRCONVERTRANGE 4
> +SECTION_RODATA
> +mult_%1: times 4 dd %2
> +offset_%1: times 4 dd %3
> +SECTION .text
> +cglobal %1, 3, 4, 4, dstU, dstV, width, x
> + movsxdifnidn widthq, widthd
> + xor xq, xq
> + mova m1, [mult_%1]
> + mova m2, [offset_%1]
> +.loop:
> + pmovsxwd m0, [dstUq+xq*2]
> + pmulld m0, m1
> + paddd m0, m2
> + psrad m0, %4
> + packssdw m0, m0
> + movh [dstUq+xq*2], m0
> + pmovsxwd m0, [dstVq+xq*2]
> + pmulld m0, m1
> + paddd m0, m2
> + psrad m0, %4
> + packssdw m0, m0
> + movh [dstVq+xq*2], m0
> + add xq, mmsize / 4
> + cmp xd, widthd
> + jl .loop
> + RET
> +%endmacro
> +
> +%if ARCH_X86_64
> +INIT_XMM sse4
> +LUMCONVERTRANGE lumRangeToJpeg, 19077, -39057361, 14
> +CHRCONVERTRANGE chrRangeToJpeg, 4663, -9289992, 12
> +LUMCONVERTRANGE lumRangeFromJpeg, 14071, 33561947, 14
> +CHRCONVERTRANGE chrRangeFromJpeg, 1799, 4081085, 11
> +%endif
> diff --git a/libswscale/x86/swscale.c b/libswscale/x86/swscale.c
> index fff8bb4396..c5ddfb5605 100644
> --- a/libswscale/x86/swscale.c
> +++ b/libswscale/x86/swscale.c
> @@ -447,6 +447,38 @@ INPUT_PLANAR_RGB_UV_ALL_DECL(avx2);
> INPUT_PLANAR_RGB_A_ALL_DECL(avx2);
> #endif
>
> +#if ARCH_X86_64
> +#define RANGE_CONVERT_FUNCS(opt) do { \
> + if (c->dstBpc <= 14) { \
> + if (c->srcRange) { \
> + c->lumConvertRange = ff_lumRangeFromJpeg_ ##opt; \
> + c->chrConvertRange = ff_chrRangeFromJpeg_ ##opt; \
> + } else { \
> + c->lumConvertRange = ff_lumRangeToJpeg_ ##opt; \
> + c->chrConvertRange = ff_chrRangeToJpeg_ ##opt; \
> + } \
> + } \
> +} while (0)
> +
> +#define RANGE_CONVERT_FUNCS_DECL(opt) \
> +void ff_lumRangeFromJpeg_ ##opt(int16_t *dst, int width); \
> +void ff_chrRangeFromJpeg_ ##opt(int16_t *dstU, int16_t *dstV, int width); \
> +void ff_lumRangeToJpeg_ ##opt(int16_t *dst, int width); \
> +void ff_chrRangeToJpeg_ ##opt(int16_t *dstU, int16_t *dstV, int width); \
> +
> +RANGE_CONVERT_FUNCS_DECL(sse4);
> +
> +av_cold void ff_sws_init_range_convert_x86(SwsContext *c)
> +{
> + if (c->srcRange != c->dstRange && !isAnyRGB(c->dstFormat)) {
> + int cpu_flags = av_get_cpu_flags();
> + if (EXTERNAL_SSE4(cpu_flags)) {
> + RANGE_CONVERT_FUNCS(sse4);
> + }
> + }
> +}
> +#endif
> +
> av_cold void ff_sws_init_swscale_x86(SwsContext *c)
> {
> int cpu_flags = av_get_cpu_flags();
> @@ -805,4 +837,8 @@ switch(c->dstBpc){ \
> }
>
> #endif
> +
> +#if ARCH_X86_64
> + ff_sws_init_range_convert_x86(c);
> +#endif
> }
> --
> 2.30.2
>
Attached version is a little bit different, moving the consts out of
the macro (so they can be reused by avx2) and processing twice the
amount of data per loop.
[-- Attachment #2: 0001-swscale-x86-add-sse4-lum-chr-ConvertRange.patch --]
[-- Type: text/x-patch, Size: 10363 bytes --]
From b8f72b1c4c8393becea9962378af6d7dffabbce2 Mon Sep 17 00:00:00 2001
From: Ramiro Polla <ramiro.polla@gmail.com>
Date: Thu, 6 Jun 2024 18:33:34 +0200
Subject: [PATCH] swscale/x86: add sse4 {lum,chr}ConvertRange
chrRangeFromJpeg_8_c: 19.9
chrRangeFromJpeg_8_sse4: 16.2
chrRangeFromJpeg_24_c: 60.7
chrRangeFromJpeg_24_sse4: 28.9
chrRangeFromJpeg_128_c: 325.7
chrRangeFromJpeg_128_sse4: 160.2
chrRangeFromJpeg_144_c: 364.2
chrRangeFromJpeg_144_sse4: 194.9
chrRangeFromJpeg_256_c: 630.7
chrRangeFromJpeg_256_sse4: 337.4
chrRangeFromJpeg_512_c: 1240.4
chrRangeFromJpeg_512_sse4: 668.4
chrRangeToJpeg_8_c: 37.7
chrRangeToJpeg_8_sse4: 19.7
chrRangeToJpeg_24_c: 114.7
chrRangeToJpeg_24_sse4: 30.2
chrRangeToJpeg_128_c: 636.4
chrRangeToJpeg_128_sse4: 161.7
chrRangeToJpeg_144_c: 715.7
chrRangeToJpeg_144_sse4: 272.9
chrRangeToJpeg_256_c: 1256.7
chrRangeToJpeg_256_sse4: 341.9
chrRangeToJpeg_512_c: 2498.7
chrRangeToJpeg_512_sse4: 668.4
lumRangeFromJpeg_8_c: 11.7
lumRangeFromJpeg_8_sse4: 12.4
lumRangeFromJpeg_24_c: 36.9
lumRangeFromJpeg_24_sse4: 17.7
lumRangeFromJpeg_128_c: 228.4
lumRangeFromJpeg_128_sse4: 85.2
lumRangeFromJpeg_144_c: 272.9
lumRangeFromJpeg_144_sse4: 96.9
lumRangeFromJpeg_256_c: 463.4
lumRangeFromJpeg_256_sse4: 183.9
lumRangeFromJpeg_512_c: 879.9
lumRangeFromJpeg_512_sse4: 355.9
lumRangeToJpeg_8_c: 17.7
lumRangeToJpeg_8_sse4: 15.4
lumRangeToJpeg_24_c: 56.2
lumRangeToJpeg_24_sse4: 18.4
lumRangeToJpeg_128_c: 331.4
lumRangeToJpeg_128_sse4: 84.4
lumRangeToJpeg_144_c: 375.2
lumRangeToJpeg_144_sse4: 96.9
lumRangeToJpeg_256_c: 649.7
lumRangeToJpeg_256_sse4: 184.4
lumRangeToJpeg_512_c: 1281.9
lumRangeToJpeg_512_sse4: 355.9
---
libswscale/swscale_internal.h | 1 +
libswscale/utils.c | 2 +
libswscale/x86/Makefile | 1 +
libswscale/x86/range_convert.asm | 130 +++++++++++++++++++++++++++++++
libswscale/x86/swscale.c | 36 +++++++++
5 files changed, 170 insertions(+)
create mode 100644 libswscale/x86/range_convert.asm
diff --git a/libswscale/swscale_internal.h b/libswscale/swscale_internal.h
index d4b0c3cee2..92f6105443 100644
--- a/libswscale/swscale_internal.h
+++ b/libswscale/swscale_internal.h
@@ -698,6 +698,7 @@ void ff_updateMMXDitherTables(SwsContext *c, int dstY);
av_cold void ff_sws_init_range_convert(SwsContext *c);
av_cold void ff_sws_init_range_convert_loongarch(SwsContext *c);
+av_cold void ff_sws_init_range_convert_x86(SwsContext *c);
SwsFunc ff_yuv2rgb_init_x86(SwsContext *c);
SwsFunc ff_yuv2rgb_init_ppc(SwsContext *c);
diff --git a/libswscale/utils.c b/libswscale/utils.c
index 476a24fea5..8dfa57b5ff 100644
--- a/libswscale/utils.c
+++ b/libswscale/utils.c
@@ -1082,6 +1082,8 @@ int sws_setColorspaceDetails(struct SwsContext *c, const int inv_table[4],
ff_sws_init_range_convert(c);
#if ARCH_LOONGARCH64
ff_sws_init_range_convert_loongarch(c);
+#elif ARCH_X86
+ ff_sws_init_range_convert_x86(c);
#endif
}
diff --git a/libswscale/x86/Makefile b/libswscale/x86/Makefile
index 68391494be..f00154941d 100644
--- a/libswscale/x86/Makefile
+++ b/libswscale/x86/Makefile
@@ -12,6 +12,7 @@ X86ASM-OBJS += x86/input.o \
x86/output.o \
x86/scale.o \
x86/scale_avx2.o \
+ x86/range_convert.o \
x86/rgb_2_rgb.o \
x86/yuv_2_rgb.o \
x86/yuv2yuvX.o \
diff --git a/libswscale/x86/range_convert.asm b/libswscale/x86/range_convert.asm
new file mode 100644
index 0000000000..13983a386b
--- /dev/null
+++ b/libswscale/x86/range_convert.asm
@@ -0,0 +1,130 @@
+;******************************************************************************
+;* Copyright (c) 2024 Ramiro Polla
+;*
+;* 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
+
+chr_to_mult: times 4 dd 4663
+chr_to_offset: times 4 dd -9289992
+%define chr_to_shift 12
+
+chr_from_mult: times 4 dd 1799
+chr_from_offset: times 4 dd 4081085
+%define chr_from_shift 11
+
+lum_to_mult: times 4 dd 19077
+lum_to_offset: times 4 dd -39057361
+%define lum_to_shift 14
+
+lum_from_mult: times 4 dd 14071
+lum_from_offset: times 4 dd 33561947
+%define lum_from_shift 14
+
+SECTION .text
+
+; NOTE: there is no need to clamp the input when converting to jpeg range
+; (like we do in the C code) because packssdw will saturate the output.
+
+;-----------------------------------------------------------------------------
+; lumConvertRange
+;
+; void ff_lumRangeToJpeg_<opt>(int16_t *dst, int width);
+; void ff_lumRangeFromJpeg_<opt>(int16_t *dst, int width);
+;
+;-----------------------------------------------------------------------------
+
+%macro LUMCONVERTRANGE 4
+cglobal %1, 2, 3, 3, dst, width, x
+ movsxdifnidn widthq, widthd
+ xor xq, xq
+ mova m4, [%2]
+ mova m5, [%3]
+.loop:
+ pmovsxwd m0, [dstq+xq*2]
+ pmovsxwd m1, [dstq+xq*2+mmsize/2]
+ pmulld m0, m4
+ pmulld m1, m4
+ paddd m0, m5
+ paddd m1, m5
+ psrad m0, %4
+ psrad m1, %4
+ packssdw m0, m0
+ packssdw m1, m1
+ movq [dstq+xq*2], m0
+ movq [dstq+xq*2+mmsize/2], m1
+ add xq, mmsize / 2
+ cmp xd, widthd
+ jl .loop
+ RET
+%endmacro
+
+;-----------------------------------------------------------------------------
+; chrConvertRange
+;
+; void ff_chrRangeToJpeg_<opt>(int16_t *dstU, int16_t *dstV, int width);
+; void ff_chrRangeFromJpeg_<opt>(int16_t *dstU, int16_t *dstV, int width);
+;
+;-----------------------------------------------------------------------------
+
+%macro CHRCONVERTRANGE 4
+cglobal %1, 3, 4, 4, dstU, dstV, width, x
+ movsxdifnidn widthq, widthd
+ xor xq, xq
+ mova m4, [%2]
+ mova m5, [%3]
+.loop:
+ pmovsxwd m0, [dstUq+xq*2]
+ pmovsxwd m1, [dstUq+xq*2+mmsize/2]
+ pmovsxwd m2, [dstVq+xq*2]
+ pmovsxwd m3, [dstVq+xq*2+mmsize/2]
+ pmulld m0, m4
+ pmulld m1, m4
+ pmulld m2, m4
+ pmulld m3, m4
+ paddd m0, m5
+ paddd m1, m5
+ paddd m2, m5
+ paddd m3, m5
+ psrad m0, %4
+ psrad m1, %4
+ psrad m2, %4
+ psrad m3, %4
+ packssdw m0, m0
+ packssdw m1, m1
+ packssdw m2, m2
+ packssdw m3, m3
+ movq [dstUq+xq*2], m0
+ movq [dstUq+xq*2+mmsize/2], m1
+ movq [dstVq+xq*2], m2
+ movq [dstVq+xq*2+mmsize/2], m3
+ add xq, mmsize / 2
+ cmp xd, widthd
+ jl .loop
+ RET
+%endmacro
+
+%if ARCH_X86_64
+INIT_XMM sse4
+LUMCONVERTRANGE lumRangeToJpeg, lum_to_mult, lum_to_offset, lum_to_shift
+CHRCONVERTRANGE chrRangeToJpeg, chr_to_mult, chr_to_offset, chr_to_shift
+LUMCONVERTRANGE lumRangeFromJpeg, lum_from_mult, lum_from_offset, lum_from_shift
+CHRCONVERTRANGE chrRangeFromJpeg, chr_from_mult, chr_from_offset, chr_from_shift
+%endif
diff --git a/libswscale/x86/swscale.c b/libswscale/x86/swscale.c
index fff8bb4396..c5ddfb5605 100644
--- a/libswscale/x86/swscale.c
+++ b/libswscale/x86/swscale.c
@@ -447,6 +447,38 @@ INPUT_PLANAR_RGB_UV_ALL_DECL(avx2);
INPUT_PLANAR_RGB_A_ALL_DECL(avx2);
#endif
+#if ARCH_X86_64
+#define RANGE_CONVERT_FUNCS(opt) do { \
+ if (c->dstBpc <= 14) { \
+ if (c->srcRange) { \
+ c->lumConvertRange = ff_lumRangeFromJpeg_ ##opt; \
+ c->chrConvertRange = ff_chrRangeFromJpeg_ ##opt; \
+ } else { \
+ c->lumConvertRange = ff_lumRangeToJpeg_ ##opt; \
+ c->chrConvertRange = ff_chrRangeToJpeg_ ##opt; \
+ } \
+ } \
+} while (0)
+
+#define RANGE_CONVERT_FUNCS_DECL(opt) \
+void ff_lumRangeFromJpeg_ ##opt(int16_t *dst, int width); \
+void ff_chrRangeFromJpeg_ ##opt(int16_t *dstU, int16_t *dstV, int width); \
+void ff_lumRangeToJpeg_ ##opt(int16_t *dst, int width); \
+void ff_chrRangeToJpeg_ ##opt(int16_t *dstU, int16_t *dstV, int width); \
+
+RANGE_CONVERT_FUNCS_DECL(sse4);
+
+av_cold void ff_sws_init_range_convert_x86(SwsContext *c)
+{
+ if (c->srcRange != c->dstRange && !isAnyRGB(c->dstFormat)) {
+ int cpu_flags = av_get_cpu_flags();
+ if (EXTERNAL_SSE4(cpu_flags)) {
+ RANGE_CONVERT_FUNCS(sse4);
+ }
+ }
+}
+#endif
+
av_cold void ff_sws_init_swscale_x86(SwsContext *c)
{
int cpu_flags = av_get_cpu_flags();
@@ -805,4 +837,8 @@ switch(c->dstBpc){ \
}
#endif
+
+#if ARCH_X86_64
+ ff_sws_init_range_convert_x86(c);
+#endif
}
--
2.30.2
[-- Attachment #3: Type: text/plain, Size: 251 bytes --]
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
next prev parent reply other threads:[~2024-06-07 17:38 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-07 14:05 [FFmpeg-devel] [PATCH 1/4] tests/checkasm: cosmetics, one object per line in Makefile Ramiro Polla
2024-06-07 14:05 ` [FFmpeg-devel] [PATCH 2/4] checkasm: add tests for {lum, chr}ConvertRange Ramiro Polla
2024-06-07 14:05 ` [FFmpeg-devel] [PATCH 3/4] swscale/x86: add sse4 " Ramiro Polla
2024-06-07 17:38 ` Ramiro Polla [this message]
2024-06-07 14:05 ` [FFmpeg-devel] [PATCH 4/4] swscale/aarch64: add neon " Ramiro Polla
2024-06-10 11:56 ` Martin Storsjö
2024-06-11 12:33 ` Ramiro Polla
2024-06-07 18:45 ` [FFmpeg-devel] [PATCH 1/4] tests/checkasm: cosmetics, one object per line in Makefile Andreas Rheinhardt
2024-06-07 19:09 ` Ramiro Polla
2024-06-07 19:12 ` Andreas Rheinhardt
2024-06-07 19:47 ` Ramiro Polla
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CALweWgAwdiXYXB5X=ZSa9MW0NJg1u4u2opNjMw+HK61Vw2KKVA@mail.gmail.com' \
--to=ramiro.polla@gmail.com \
--cc=ffmpeg-devel@ffmpeg.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
This inbox may be cloned and mirrored by anyone:
git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git
# If you have public-inbox 1.1+ installed, you may
# initialize and index your mirror using the following commands:
public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
ffmpegdev@gitmailbox.com
public-inbox-index ffmpegdev
Example config snippet for mirrors.
AGPL code for this site: git clone https://public-inbox.org/public-inbox.git