110796739a
x86: hpeldsp: Move avg_pixels8_x2_mmx() out of hpeldsp_rnd_template.c
...
The function is only instantiated once, so there is no point
in keeping it in a template file.
2013-05-06 11:02:08 +02:00
dc1b328d0d
x86: hpeldsp: Only compile MMX hpeldsp code if MMX is enabled
2013-05-06 11:02:08 +02:00
9e5e76ef9e
x86: More specific ifdefs for dsputil/hpeldsp init functions
2013-05-06 11:02:07 +02:00
6fee1b90ce
avcodec: Add av_cold attributes to init functions missing them
2013-05-04 21:09:45 +02:00
a5f8873620
silly typo fixes
2013-05-03 18:26:12 +02:00
5a97469a4f
x86: sbrdsp: Implement SSE2 qmf_deint_bfly
...
Sandybridge: 47 cycles
Having a loop counter is a 7 cycle gain.
Unrolling is another 7 cycle gain.
Working in reverse scan is another 6 cycles.
Signed-off-by: Diego Biurrun <diego@biurrun.de >
2013-05-03 18:23:14 +02:00
bf7c3c6b15
x86: dsputil: Move cavs and vc1-specific functions where they belong
2013-05-02 11:45:37 +02:00
9328062321
x86: dsputil: Move avg_pixels16_mmx() out of rnd_template.c
...
The function does not do any rounding, so there is no point in
keeping it in a round template file.
2013-05-02 11:45:37 +02:00
9c112a6158
x86: dsputil: Move avg_pixels8_mmx() out of rnd_template.c
...
The function is only instantiated once, so there is no point
in keeping it in a template file.
2013-05-02 11:45:37 +02:00
9b3a04d306
x86: Move duplicated put_pixels{8|16}_mmx functions into their own file
2013-05-02 11:16:45 +02:00
f2e9d44a57
x86: Drop unnecessary ff_ name prefixes from static functions
2013-04-30 16:02:03 +02:00
643e433bf7
mpegaudiosp: More consistent names for ppc/x86 optimization files
2013-04-30 12:19:43 +02:00
97c56ad796
x86: dsputil: Remove a set of pointless #ifs around function declarations
2013-04-30 01:42:32 +02:00
85f2f82af6
x86: dsputil: cosmetics: Group ff_{avg|put}_pixels16_mmxext() declarations
2013-04-30 01:41:05 +02:00
20784aa678
x86: hpeldsp: Remove unused macro definitions
2013-04-29 15:57:00 +02:00
7c00e9d8ae
x86: ac3dsp: Remove 3dnow version of ff_ac3_extract_exponents
...
The function requires increasing the fuzz factor for the ac3/eac3 encode
tests and even so makes fate fail. It only provides a slight encoding
speedup for legacy CPUs that do not support SS2. Thus its benefit is not
worth the trouble it creates and fixing it would be a waste of time.
2013-04-26 21:06:52 +02:00
74685f6783
x86: Rename dsputil_rnd_template.c to rnd_template.c
...
This makes it less confusing when this template is shared both by
dsputil and by hpeldsp.
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-04-25 23:03:09 +03:00
486f76f029
x86: Get rid of duplication between *_rnd_template.c
...
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-04-23 23:30:17 +03:00
6a8561dbd7
x86: Factorize duplicated inline assembly snippets
...
Signed-off-by: Diego Biurrun <diego@biurrun.de >
2013-04-23 15:07:31 +02:00
c1ad70c3cb
x86: Move some conditional code around to avoid unused variable warnings
2013-04-22 17:50:02 +02:00
b4ad7c54c8
x86: cavs: Refactor duplicate dspfunc macro
2013-04-22 12:05:09 +02:00
78fa0bd0f7
x86: cavs: Put mmx-specific code into its own init function
...
Before, this code was labeled as mmxext and enabled both for the
3dnow and the mmxext case.
2013-04-22 10:42:50 +02:00
311a592dfc
x86: Remove some duplicate function declarations
2013-04-22 02:29:57 +02:00
b71a0507b0
x86: Remove unused inline asm instruction defines
...
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-04-20 00:44:54 +03:00
8db00081a3
x86: hpeldsp: Move half-pel assembly from dsputil to hpeldsp
...
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-04-19 23:18:53 +03:00
015821229f
vp3: Use full transpose for all IDCTs
...
This way, the special IDCT permutations are no longer needed. This
is similar to how H264 does it, and removes the dsputil dependency
imposed by the scantable code.
Also remove the unused type == 0 cases from the plain C version
of the idct.
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-04-15 12:32:05 +03:00
c46819f229
x86: Move constants to the only place where they are used
...
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-04-15 12:17:39 +03:00
a3cb865310
x86: dsputil: Move some ifdefs to avoid unused variable warnings
2013-04-12 09:36:47 +02:00
2004c7c8f7
x86: dsputil: cosmetics: Remove two pointless variable indirections
2013-04-12 09:36:47 +02:00
c51a3a5bd9
x86: dsputil: Refactor some ff_{avg|put}_pixels function declarations
2013-04-12 09:36:46 +02:00
e027032fc6
x86: dsputil: ff_h263_*_loop_filter declarations to a more suitable place
2013-04-12 09:36:46 +02:00
a89c05500f
x86: h264qpel: int --> ptrdiff_t for some line_size parameters
2013-04-12 09:30:12 +02:00
ac9362c5d9
Move misplaced file author information where it belongs
2013-04-11 02:42:11 +02:00
b93b27edb0
dsputil: Make dsputil selectable
...
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-04-10 11:04:05 +03:00
62844c3fd6
h264: Integrate clear_blocks calls with IDCT
...
The non-intra-pcm branch in hl_decode_mb (simple, 8bpp) goes from 700
to 672 cycles, and the complete loop of decode_mb_cabac and hl_decode_mb
(in the decode_slice loop) goes from 1759 to 1733 cycles on the clip
tested (cathedral), i.e. almost 30 cycles per mb faster.
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-04-10 11:03:06 +03:00
610b18e2e3
x86: qpel: Move fullpel and l2 functions to a separate file
...
This way, they can be shared between mpeg4qpel and h264qpel without
requiring either one to be compiled unconditionally.
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-04-08 12:38:33 +03:00
f4b0d12f5b
x86: sbrdsp: Implement SSE neg_odd_64
...
Timing on Arrandale:
C SSE
Win32: 57 44
Win64: 47 38
Unrolling and not storing mask both save some cycles.
Signed-off-by: Diego Biurrun <diego@biurrun.de >
2013-04-05 22:47:04 +02:00
b6649ab503
cosmetics: Remove unnecessary extern keywords from function declarations
2013-03-27 14:21:45 +01:00
a2acadd058
x86: vc1dsp: Fix indentation
...
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-03-26 15:49:42 +02:00
e5c2794a71
x86: consistently use unaligned movs in the unaligned bswap
...
Fixes fate errors in asv1, ffvhuff and huffyuv on x86_32.
2013-03-25 12:11:11 +01:00
285ff14413
x86: Change a missed occurrance of int to ptrdiff_t for strides
...
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-03-24 12:06:53 +02:00
352dbdb96c
x86: Remove win64 xmm clobbering wrappers for the now removed avcodec_encode_video function
...
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-03-23 23:37:27 +02:00
a8b6015823
dsputil: convert remaining functions to use ptrdiff_t strides
...
Signed-off-by: Luca Barbato <lu_zero@gentoo.org >
2013-03-12 18:26:42 +01:00
e8c52271c4
Revert "Move H264/QPEL specific asm from dsputil.asm to h264_qpel_*.asm."
...
This reverts commit f90ff772e771155d7
2013-02-28 21:50:02 +01:00
ebc701993f
x86: dsputil: Drop some unused function #defines
2013-02-26 23:36:24 +01:00
845cfc92f9
x86: dsputil: Drop aliasing of ff_put_pixels8_mmx to ff_put_pixels8_mmxext
...
The external assembly function uses mmxext instructions and should not be
masqueraded as an mmx-only function. Instead, use the mmx-only inline
assembly function.
2013-02-26 23:36:24 +01:00
096cc11ec1
x86: vc1dsp: Move ff_avg_vc1_mspel_mc00_mmxext out of dsputil_mmx.c
2013-02-26 23:36:24 +01:00
31a23a0dc6
x86: dsputil_mmx: Remove leftover inline assembly fragments
...
These became unused in 71155d7bmartin@martin.st >
2013-02-27 00:17:05 +02:00
c242bbd8b6
Remove unnecessary dsputil.h #includes
2013-02-26 00:51:34 +01:00
311443f6c7
x86: h264: Don't use redzone in AVX h264_deblock on Win64
...
This fixes crashes in chromium on win64 on machines with AVX
(crashes that apparently aren't triggered by fate).
Signed-off-by: Martin Storsjö <martin@martin.st >
2013-02-21 15:02:16 +02:00
Diego Biurrun
Christophe Gisquet
Martin Storsjö
Ronald S. Bultje
Janne Grunau
Luca Barbato
Matt Wolenetz