2017-01-26 18:03:08 +02:00
|
|
|
OBJS-$(CONFIG_AFIR_FILTER) += x86/af_afir_init.o
|
2015-10-02 17:22:42 +02:00
|
|
|
OBJS-$(CONFIG_BLEND_FILTER) += x86/vf_blend_init.o
|
2016-03-13 11:06:21 +02:00
|
|
|
OBJS-$(CONFIG_BWDIF_FILTER) += x86/vf_bwdif_init.o
|
2016-04-06 20:09:08 +02:00
|
|
|
OBJS-$(CONFIG_COLORSPACE_FILTER) += x86/colorspacedsp_init.o
|
2015-01-19 00:26:48 +02:00
|
|
|
OBJS-$(CONFIG_EQ_FILTER) += x86/vf_eq.o
|
2014-12-26 20:37:54 +02:00
|
|
|
OBJS-$(CONFIG_FSPP_FILTER) += x86/vf_fspp_init.o
|
2013-10-22 03:37:46 +03:00
|
|
|
OBJS-$(CONFIG_GRADFUN_FILTER) += x86/vf_gradfun_init.o
|
avfilter/vf_framerate: add SIMD functions for frame blending
Blend function speedups on x86_64 Core i5 4460:
ffmpeg -f lavfi -i allyuv -vf framerate=60:threads=1 -f null none
C: 447548411 decicycles in Blend, 2048 runs, 0 skips
SSSE3: 130020087 decicycles in Blend, 2048 runs, 0 skips
AVX2: 128508221 decicycles in Blend, 2048 runs, 0 skips
ffmpeg -f lavfi -i allyuv -vf format=yuv420p12,framerate=60:threads=1 -f null none
C: 228932745 decicycles in Blend, 2048 runs, 0 skips
SSE4: 123357781 decicycles in Blend, 2048 runs, 0 skips
AVX2: 121215353 decicycles in Blend, 2048 runs, 0 skips
Signed-off-by: Marton Balint <cus@passwd.hu>
2018-01-08 02:05:45 +02:00
|
|
|
OBJS-$(CONFIG_FRAMERATE_FILTER) += x86/vf_framerate_init.o
|
2017-12-01 21:56:45 +02:00
|
|
|
OBJS-$(CONFIG_HFLIP_FILTER) += x86/vf_hflip_init.o
|
2013-01-22 03:39:37 +03:00
|
|
|
OBJS-$(CONFIG_HQDN3D_FILTER) += x86/vf_hqdn3d_init.o
|
2014-09-03 12:02:32 +03:00
|
|
|
OBJS-$(CONFIG_IDET_FILTER) += x86/vf_idet_init.o
|
2014-11-11 20:43:42 +02:00
|
|
|
OBJS-$(CONFIG_INTERLACE_FILTER) += x86/vf_interlace_init.o
|
2017-07-03 17:42:03 +02:00
|
|
|
OBJS-$(CONFIG_LIMITER_FILTER) += x86/vf_limiter_init.o
|
2015-09-30 23:00:14 +02:00
|
|
|
OBJS-$(CONFIG_MASKEDMERGE_FILTER) += x86/vf_maskedmerge_init.o
|
2014-10-17 04:24:42 +03:00
|
|
|
OBJS-$(CONFIG_NOISE_FILTER) += x86/vf_noise.o
|
2015-01-09 21:51:13 +02:00
|
|
|
OBJS-$(CONFIG_PP7_FILTER) += x86/vf_pp7_init.o
|
2015-07-12 12:44:39 +02:00
|
|
|
OBJS-$(CONFIG_PSNR_FILTER) += x86/vf_psnr_init.o
|
2013-07-08 15:42:53 +03:00
|
|
|
OBJS-$(CONFIG_PULLUP_FILTER) += x86/vf_pullup_init.o
|
avfilter/vf_removegrain: add x86 and x86_64 SSE2 functions
Speed of all modes increased by a factor between 7.4 and 19.8 largely depending
on whether bytes are unpacked into words. Modes 2, 3, and 4 have been sped-up
by a factor of 43 (thanks quick sort!)
All modes are available on x86_64 but only modes 1, 10, 11, 12, 13, 14, 19, 20,
21, and 22 are available on x86 due to the number of SIMD registers used.
With a contribution from James Almer <jamrial@gmail.com>
2015-07-15 01:48:47 +02:00
|
|
|
OBJS-$(CONFIG_REMOVEGRAIN_FILTER) += x86/vf_removegrain_init.o
|
2016-06-04 09:33:05 +02:00
|
|
|
OBJS-$(CONFIG_SHOWCQT_FILTER) += x86/avf_showcqt_init.o
|
2013-05-11 13:03:38 +03:00
|
|
|
OBJS-$(CONFIG_SPP_FILTER) += x86/vf_spp.o
|
2015-07-13 01:33:06 +02:00
|
|
|
OBJS-$(CONFIG_SSIM_FILTER) += x86/vf_ssim_init.o
|
2015-10-04 11:34:03 +02:00
|
|
|
OBJS-$(CONFIG_STEREO3D_FILTER) += x86/vf_stereo3d_init.o
|
2015-10-02 17:22:42 +02:00
|
|
|
OBJS-$(CONFIG_TBLEND_FILTER) += x86/vf_blend_init.o
|
2017-11-12 20:11:51 +02:00
|
|
|
OBJS-$(CONFIG_THRESHOLD_FILTER) += x86/vf_threshold_init.o
|
2014-11-15 04:49:37 +02:00
|
|
|
OBJS-$(CONFIG_TINTERLACE_FILTER) += x86/vf_tinterlace_init.o
|
2012-09-23 21:49:26 +03:00
|
|
|
OBJS-$(CONFIG_VOLUME_FILTER) += x86/af_volume_init.o
|
2015-10-07 21:03:16 +02:00
|
|
|
OBJS-$(CONFIG_W3FDIF_FILTER) += x86/vf_w3fdif_init.o
|
2014-01-04 15:49:38 +03:00
|
|
|
OBJS-$(CONFIG_YADIF_FILTER) += x86/vf_yadif_init.o
|
2012-08-29 20:37:14 +03:00
|
|
|
|
2016-10-08 16:18:33 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_AFIR_FILTER) += x86/af_afir.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_BLEND_FILTER) += x86/vf_blend.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_BWDIF_FILTER) += x86/vf_bwdif.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_COLORSPACE_FILTER) += x86/colorspacedsp.o
|
avfilter/vf_framerate: add SIMD functions for frame blending
Blend function speedups on x86_64 Core i5 4460:
ffmpeg -f lavfi -i allyuv -vf framerate=60:threads=1 -f null none
C: 447548411 decicycles in Blend, 2048 runs, 0 skips
SSSE3: 130020087 decicycles in Blend, 2048 runs, 0 skips
AVX2: 128508221 decicycles in Blend, 2048 runs, 0 skips
ffmpeg -f lavfi -i allyuv -vf format=yuv420p12,framerate=60:threads=1 -f null none
C: 228932745 decicycles in Blend, 2048 runs, 0 skips
SSE4: 123357781 decicycles in Blend, 2048 runs, 0 skips
AVX2: 121215353 decicycles in Blend, 2048 runs, 0 skips
Signed-off-by: Marton Balint <cus@passwd.hu>
2018-01-08 02:05:45 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_FRAMERATE_FILTER) += x86/vf_framerate.o
|
2016-10-08 16:18:33 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_FSPP_FILTER) += x86/vf_fspp.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_GRADFUN_FILTER) += x86/vf_gradfun.o
|
2017-12-01 21:56:45 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_HFLIP_FILTER) += x86/vf_hflip.o
|
2016-10-08 16:18:33 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_HQDN3D_FILTER) += x86/vf_hqdn3d.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_IDET_FILTER) += x86/vf_idet.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_INTERLACE_FILTER) += x86/vf_interlace.o
|
2017-07-03 17:42:03 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_LIMITER_FILTER) += x86/vf_limiter.o
|
2016-10-08 16:18:33 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_MASKEDMERGE_FILTER) += x86/vf_maskedmerge.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_PP7_FILTER) += x86/vf_pp7.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_PSNR_FILTER) += x86/vf_psnr.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_PULLUP_FILTER) += x86/vf_pullup.o
|
avfilter/vf_removegrain: add x86 and x86_64 SSE2 functions
Speed of all modes increased by a factor between 7.4 and 19.8 largely depending
on whether bytes are unpacked into words. Modes 2, 3, and 4 have been sped-up
by a factor of 43 (thanks quick sort!)
All modes are available on x86_64 but only modes 1, 10, 11, 12, 13, 14, 19, 20,
21, and 22 are available on x86 due to the number of SIMD registers used.
With a contribution from James Almer <jamrial@gmail.com>
2015-07-15 01:48:47 +02:00
|
|
|
ifdef CONFIG_GPL
|
2016-10-08 16:18:33 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_REMOVEGRAIN_FILTER) += x86/vf_removegrain.o
|
avfilter/vf_removegrain: add x86 and x86_64 SSE2 functions
Speed of all modes increased by a factor between 7.4 and 19.8 largely depending
on whether bytes are unpacked into words. Modes 2, 3, and 4 have been sped-up
by a factor of 43 (thanks quick sort!)
All modes are available on x86_64 but only modes 1, 10, 11, 12, 13, 14, 19, 20,
21, and 22 are available on x86 due to the number of SIMD registers used.
With a contribution from James Almer <jamrial@gmail.com>
2015-07-15 01:48:47 +02:00
|
|
|
endif
|
2016-10-08 16:18:33 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_SHOWCQT_FILTER) += x86/avf_showcqt.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_SSIM_FILTER) += x86/vf_ssim.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_STEREO3D_FILTER) += x86/vf_stereo3d.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_TBLEND_FILTER) += x86/vf_blend.o
|
2017-11-12 20:11:51 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_THRESHOLD_FILTER) += x86/vf_threshold.o
|
2016-10-08 16:18:33 +02:00
|
|
|
X86ASM-OBJS-$(CONFIG_TINTERLACE_FILTER) += x86/vf_interlace.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_VOLUME_FILTER) += x86/af_volume.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_W3FDIF_FILTER) += x86/vf_w3fdif.o
|
|
|
|
X86ASM-OBJS-$(CONFIG_YADIF_FILTER) += x86/vf_yadif.o x86/yadif-16.o x86/yadif-10.o
|