Commit Graph

58 Commits

Author SHA1 Message Date
Henrik Gramner ab43beefab x86inc: Drop SECTION_TEXT macro
The .text section is already 16-byte aligned by default on all supported
platforms so `SECTION_TEXT` isn't any different from `SECTION .text`.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
2015-08-11 11:12:01 +02:00
James Almer d68c05380c x86: check for AV_CPU_FLAG_AVXSLOW where useful
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
2015-05-31 12:07:11 +02:00
Kieran Kunhya 96fda42a8f vf_interlace: get rid of useless loads
Signed-off-by: Anton Khirnov <anton@khirnov.net>
2014-11-27 13:57:50 +01:00
Michael Niedermayer ca5c3ff909 vf_interlace: x86: improve asm performance
4775 decicycles -> 3688 decicycles
2014-11-25 02:00:06 +00:00
Kieran Kunhya 2e1704059a vf_interlace: Add SIMD for lowpass filter
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
2014-11-15 00:35:31 +01:00
Diego Biurrun 01c5779f56 x86: Drop some unnecessary YASM ifdefs
Dead code elimination is enough to avoid undefined references in these cases.
2014-04-04 19:08:05 +02:00
Robert Krüger d8e763fda7 vf_yadif: Relicense from GPL to LGPL
All copyright holders have agreed to the relicensing.
2014-01-14 00:04:59 +01:00
Diego Biurrun ed1a11ed52 gradfun: x86: Factor out common code for some gradfun_filter_line() variants 2013-10-31 16:34:18 +01:00
Diego Biurrun ee80cf741a avfilter: x86: K&R formatting cosmetics 2013-10-31 12:15:54 +01:00
Daniel Kang 0e73049416 avfilter: x86: Port gradfun filter optimizations to yasm
Signed-off-by: Diego Biurrun <diego@biurrun.de>
2013-10-23 14:50:27 +02:00
Diego Biurrun f6633c55a3 avfilter: Fix typo in Loren's email address 2013-10-23 10:25:14 +02:00
Diego Biurrun 3ac7fa81b2 Consistently use "cpu_flags" as variable/parameter name for CPU flags 2013-07-18 00:31:35 +02:00
James Darnley b0ef0ae776 yadif: restore speed of the C filtering code
Always use the special filter for the first and last 3 columns (only).

Changes made in 64ed397 slowed the filter to just under 3/4 of what it
was.  This commit restores the speed while maintaining identical output.

For reference, on my Athlon64:
1733222 decicycles in old
2358563 decicycles in new
1727558 decicycles in this

Signed-off-by: Anton Khirnov <anton@khirnov.net>
2013-05-14 09:23:55 +02:00
Diego Biurrun 6e9f8d6a7d x86: vf_yadif: Remove stray dsputil_mmx #include 2013-05-08 18:18:23 +02:00
Diego Biurrun 093804a93c avfilter: Add av_cold attributes to init/uninit functions 2013-05-04 21:10:05 +02:00
Diego Biurrun c1ad70c3cb x86: Move some conditional code around to avoid unused variable warnings 2013-04-22 17:50:02 +02:00
Clément Bœsch 1ae44c87c9 lavfi/gradfun: remove rounding to match C and SSE code.
There is no noticable benefit for such precision.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
2013-03-28 07:59:29 +01:00
Clément Bœsch 38a2f88d39 lavfi/gradfun: fix dithering in MMX code.
Current dithering only uses the first 4 instead of the whole 8 random values.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
2013-03-28 07:59:18 +01:00
Clément Bœsch 2d66fc543b lavfi/gradfun: fix rounding in MMX code.
Current code divides before increasing precision.

Also reduce upper bound for strength from 255 to 64.  This will prevent
an overflow in the SSSE3 and MMX filter_line code: delta is expressed as
an u16 being shifted by 2 to the left. If it overflows, having a
strength not above 64 will make sure that m is set to 0 (making the
m*m*delta >> 14 expression void).

A value above 64 should not make any sense unless gradfun is used as
a blur filter.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
2013-03-28 07:59:04 +01:00
Loren Merritt 5b3c1aecb2 hqdn3d: Fix out of array read in LOWPASS
CC:libav-stable@libav.org
Signed-off-by: Anton Khirnov <anton@khirnov.net>
2013-03-13 09:14:59 +01:00
Anton Khirnov 64ed397635 vf_yadif: fix out-of line reads
Some changes in the border pixels, visually indistinguishable.
2013-02-15 16:08:33 +01:00
Anton Khirnov 99162f8d46 vf_yadif: silence a warning.
clang says:
libavfilter/vf_yadif.c:192:28: warning: incompatible pointer types assigning to
'void (*)(uint8_t *, uint8_t *, uint8_t *, uint8_t *, int, int, int, int, int)'
from 'void (uint16_t *, uint16_t *, uint16_t *, uint16_t *, int, int, int, int, int)'
2013-02-06 10:21:51 +01:00
Diego Biurrun e66240f22e avfilter: x86: consistent filenames for filter optimizations 2013-02-04 15:00:47 +01:00
Diego Biurrun 76d90125cd vf_hqdn3d: x86: Add proper arch optimization initialization 2013-02-01 13:11:45 +01:00
Daniel Kang 67360ccd51 yadif: x86: fix build for compilers without aligned stack
Manually load registers to avoid using 8 registers on x86_32 with
compilers that do not align the stack (e.g. MSVC).

Signed-off-by: Diego Biurrun <diego@biurrun.de>
2013-01-14 09:51:52 +01:00
Daniel Kang 899157b308 yadif: Port inline assembly to yasm
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
2013-01-09 18:41:02 +01:00
Justin Ruggles b30a363331 x86: af_volume: add SSE2/SSSE3/AVX-optimized s32 volume scaling 2012-12-05 11:23:37 -05:00
Justin Ruggles f96f1e06a4 x86: af_volume: add SSE2-optimized s16 volume scaling 2012-12-05 11:23:37 -05:00
Diego Biurrun d8eda37080 x86: mmx2 ---> mmxext in function names 2012-10-31 17:53:57 +01:00
Diego Biurrun 04581c8c77 x86: yasm: Use complete source path for macro helper %includes
This is more consistent with the way we handle C #includes and
it simplifies the build system.
2012-10-31 00:37:42 +01:00
Diego Biurrun 6860b4081d x86: include x86inc.asm in x86util.asm
This is necessary to allow refactoring some x86util macros with cpuflags.
2012-10-31 00:37:42 +01:00
Diego Biurrun f6c38c5f4e avfilter: call x86 init functions under if (ARCH_X86), not if (HAVE_MMX) 2012-10-12 19:58:51 +02:00
Diego Biurrun 17337f54c0 x86: Split inline and external assembly #ifdefs 2012-08-31 01:53:25 +02:00
Diego Biurrun cdaec0b240 avfilter: x86: Use more precise compile template names 2012-08-30 18:51:51 +02:00
Loren Merritt 7a1944b907 vf_hqdn3d: x86 asm
13% faster on penryn, 16% on sandybridge, 15% on bulldozer
Not simd; a compiler should have generated this, but gcc didn't.
2012-08-26 10:49:14 +00:00
Martin Storsjö 1d9c2dc89a Don't include common.h from avutil.h
Signed-off-by: Martin Storsjö <martin@martin.st>
2012-08-15 22:32:06 +03:00
Mans Rullgard 480178a295 x86: yadif: fix asm with suncc
Under some circumstances, suncc will use a single register for the
address of all memory operands, inserting lea instructions loading
the correct address prior to each memory operand being used in the
code. In the yadif code, the branch in the asm block bypasses such
an lea instruction, causing an incorrect address to be used in the
following load.

This patch replaces the tmpX arrays with a single array and uses a
register operand to hold its address. Although this prevents using
offsets from the stack pointer to access these locations, the code
still builds as 32-bit PIC even with old compilers.

Signed-off-by: Mans Rullgard <mans@mansr.com>
2012-08-13 14:51:52 +01:00
Mans Rullgard c318626ce2 x86: rename libavutil/x86_cpu.h to libavutil/x86/asm.h
This puts x86-specific things in the x86/ subdirectory where they
belong.

Signed-off-by: Mans Rullgard <mans@mansr.com>
2012-08-09 00:58:20 +01:00
Diego Biurrun 239fdf1b4a x86: build: replace mmx2 by mmxext
Refactoring mmx2/mmxext YASM code with cpuflags will force renames.
So switching to a consistent naming scheme beforehand is sensible.
The name "mmxext" is more official and widespread and also the name
of the CPU flag, as reported e.g. by the Linux kernel.
2012-08-03 22:51:05 +02:00
Diego Biurrun d1505db067 x86: yadif: Mark mmxext optimizations as such
The yadif mmx optimizations contain the pmaxsw and pmaxub mmxext
instructions, causing sigills on CPUs that do not support mmxext.
2012-07-31 13:54:06 +02:00
Ronald S. Bultje 3124886554 lavfi: place x86 inline assembly under HAVE_INLINE_ASM.
This allows compiling this code using compilers that do not understand
gcc-style inline assembly.
2012-07-22 19:44:06 -07:00
Mans Rullgard fb93e61e2b x86: lavfi: fix gradfun/yadif build with mmx/sse disabled
These functions are defined conditionally so any uses need to have
preprocessor guards.

Signed-off-by: Mans Rullgard <mans@mansr.com>
2012-07-03 14:14:13 +01:00
Mans Rullgard a87b17f328 vf_yadif: move x86 init code to x86/yadif.c
Signed-off-by: Mans Rullgard <mans@mansr.com>
2012-07-02 01:16:37 +01:00
Mans Rullgard 2f0accf103 vf_gradfun: move x86 init code to x86/gradfun.c
Signed-off-by: Mans Rullgard <mans@mansr.com>
2012-07-02 01:16:37 +01:00
Janne Grunau 456d3e4fb5 yadif: specify array size outside DECLARE_ALIGNED 2012-02-09 21:28:37 +01:00
Diego Biurrun 0bba26466f cosmetics: Delete empty lines at end of file. 2012-02-09 12:26:45 +01:00
Loren Merritt 082768f0b1 vf_gradfun: relicense x86 asm to LGPL
Actually I gave permission for LGPL long ago, but the original import
failed to update the license header.
2011-07-04 00:01:47 +00:00
Michael Niedermayer 233ed6b314 Change yadif to not use out of picture lines.
Fixes issue2272.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
2011-04-03 08:46:19 +02:00
Mans Rullgard 2912e87a6c Replace FFmpeg with Libav in licence headers
Signed-off-by: Mans Rullgard <mans@mansr.com>
2011-03-19 13:33:20 +00:00
Janne Grunau 2c3589bfda consolidate .gitignore patters into a single file
Signed-off-by: Janne Grunau <janne-ffmpeg@jannau.net>
2011-01-18 21:32:05 +01:00