2015-01-19 03:56:48 +05:30
|
|
|
OBJS-$(CONFIG_EQ_FILTER) += x86/vf_eq.o
|
2014-12-26 15:37:54 -03:00
|
|
|
OBJS-$(CONFIG_FSPP_FILTER) += x86/vf_fspp_init.o
|
2013-10-21 20:37:46 -04:00
|
|
|
OBJS-$(CONFIG_GRADFUN_FILTER) += x86/vf_gradfun_init.o
|
2013-01-22 01:39:37 +01:00
|
|
|
OBJS-$(CONFIG_HQDN3D_FILTER) += x86/vf_hqdn3d_init.o
|
2014-09-03 11:02:32 +02:00
|
|
|
OBJS-$(CONFIG_IDET_FILTER) += x86/vf_idet_init.o
|
2014-11-11 18:43:42 +00:00
|
|
|
OBJS-$(CONFIG_INTERLACE_FILTER) += x86/vf_interlace_init.o
|
2014-10-16 22:24:42 -03:00
|
|
|
OBJS-$(CONFIG_NOISE_FILTER) += x86/vf_noise.o
|
2015-01-09 16:51:13 -03:00
|
|
|
OBJS-$(CONFIG_PP7_FILTER) += x86/vf_pp7_init.o
|
2013-07-08 12:42:53 +00:00
|
|
|
OBJS-$(CONFIG_PULLUP_FILTER) += x86/vf_pullup_init.o
|
2013-05-11 12:03:38 +02:00
|
|
|
OBJS-$(CONFIG_SPP_FILTER) += x86/vf_spp.o
|
2014-11-15 03:49:37 +01:00
|
|
|
OBJS-$(CONFIG_TINTERLACE_FILTER) += x86/vf_tinterlace_init.o
|
2012-09-23 14:49:26 -04:00
|
|
|
OBJS-$(CONFIG_VOLUME_FILTER) += x86/af_volume_init.o
|
2014-01-04 13:49:38 +01:00
|
|
|
OBJS-$(CONFIG_YADIF_FILTER) += x86/vf_yadif_init.o
|
2012-08-29 19:37:14 +02:00
|
|
|
|
2014-12-26 15:37:54 -03:00
|
|
|
YASM-OBJS-$(CONFIG_FSPP_FILTER) += x86/vf_fspp.o
|
2013-10-21 20:37:46 -04:00
|
|
|
YASM-OBJS-$(CONFIG_GRADFUN_FILTER) += x86/vf_gradfun.o
|
2013-02-01 13:14:31 +01:00
|
|
|
YASM-OBJS-$(CONFIG_HQDN3D_FILTER) += x86/vf_hqdn3d.o
|
2014-09-03 11:02:32 +02:00
|
|
|
YASM-OBJS-$(CONFIG_IDET_FILTER) += x86/vf_idet.o
|
2014-11-11 18:43:42 +00:00
|
|
|
YASM-OBJS-$(CONFIG_INTERLACE_FILTER) += x86/vf_interlace.o
|
2015-01-09 16:51:13 -03:00
|
|
|
YASM-OBJS-$(CONFIG_PP7_FILTER) += x86/vf_pp7.o
|
2013-07-08 12:42:53 +00:00
|
|
|
YASM-OBJS-$(CONFIG_PULLUP_FILTER) += x86/vf_pullup.o
|
2014-11-15 03:49:37 +01:00
|
|
|
YASM-OBJS-$(CONFIG_TINTERLACE_FILTER) += x86/vf_interlace.o
|
2012-09-23 14:49:26 -04:00
|
|
|
YASM-OBJS-$(CONFIG_VOLUME_FILTER) += x86/af_volume.o
|
yadif: x86 assembly for 9 to 14-bit samples
These smaller samples do not need to be unpacked to double words
allowing the code to process more pixels every iteration (still 2 in MMX
but 6 in SSE2). It also avoids emulating the missing double word
instructions on older instruction sets.
Like with the previous code for 16-bit samples this has been tested on
an Athlon64 and a Core2Quad.
Athlon64:
1809275 decicycles in C, 32718 runs, 50 skips
911675 decicycles in mmx, 32727 runs, 41 skips, 2.0x faster
495284 decicycles in sse2, 32747 runs, 21 skips, 3.7x faster
Core2Quad:
921363 decicycles in C, 32756 runs, 12 skips
486537 decicycles in mmx, 32764 runs, 4 skips, 1.9x faster
293296 decicycles in sse2, 32759 runs, 9 skips, 3.1x faster
284910 decicycles in ssse3, 32759 runs, 9 skips, 3.2x faster
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2013-03-16 21:42:24 +01:00
|
|
|
YASM-OBJS-$(CONFIG_YADIF_FILTER) += x86/vf_yadif.o x86/yadif-16.o x86/yadif-10.o
|
