- 26 Mar, 2014 31 commits
-
-
Ben Avison authored
Profiling results for overall audio decode and the mlp_filter_channel(_arm) function in particular are as follows: Before After Mean StdDev Mean StdDev Confidence Change 6:2 total 380.4 22.0 370.8 17.0 87.4% +2.6% (insignificant) 6:2 function 60.7 7.2 36.6 8.1 100.0% +65.8% 8:2 total 357.0 17.5 343.2 19.0 97.8% +4.0% (insignificant) 8:2 function 60.3 8.8 37.3 3.8 100.0% +61.8% 6:6 total 717.2 23.2 658.4 15.7 100.0% +8.9% 6:6 function 140.4 12.9 81.5 9.2 100.0% +72.4% 8:8 total 981.9 16.2 896.2 24.5 100.0% +9.6% 8:8 function 193.4 15.0 103.3 11.5 100.0% +87.2% Experiments with adding preload instructions to this function yielded no useful benefit, so these have not been included. The assembly version has also been tested with a fuzz tester to ensure that any combinations of inputs not exercised by my available test streams still generate mathematically identical results to the C version. Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
wm4 authored
The memory allocation for f->diffs was freed multiple times in some corner cases. Simplify the code so that this doesn't happen. Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
Fixes h264_mp4toannexb_bsf_failure.mkv Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* qatar/master: x86: hpeldsp: Keep all rnd_template instantiations in hpeldsp_init Conflicts: libavcodec/x86/rnd_mmx.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit '55d7f26e': hpeldsp_template: Move content to hpeldsp Conflicts: libavcodec/hpeldsp_template.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit '09d4389d': hpeldsp_template: Drop av_unused attribute from *_no_rnd_pixels16_8_c functions Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit '92ba9651': dsputil: Move draw_edges and clear_block* out of dsputil_template Conflicts: libavcodec/dsputil.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit 'da5be235': dsputil: Move RV40-specific bits into rv40dsp Conflicts: libavcodec/dsputil.c libavcodec/rv40dsp.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit '8011ac91': hpeldsp_template: Detemplatize the code Conflicts: libavcodec/hpeldsp_template.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit '2c01ad8b': dsputil_template: Detemplatize the code Conflicts: libavcodec/dsputil.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit 'aba70bb5': Add missing headers to make template files compile (more) standalone Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit 'e7373585': dsputil_template: Move bits that are used templatized into separate file Conflicts: libavcodec/dsputil_template.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit 'd3c3c166': dsputil: Move hpel_template #include out of dsputil_template Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit 'd0aabeab': x86: h264_qpel: Fix typo in CALL_2X_PIXELS macro invocation See: c8246d37Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Diego Biurrun authored
There is no point in having a separate file just for the instantiation that provides the public functions.
-
Diego Biurrun authored
There is no point in having this separate; it is not used as a template.
-
Diego Biurrun authored
-
Diego Biurrun authored
The functions are not used templatized.
-
Diego Biurrun authored
-
Diego Biurrun authored
The indirection makes no sense without multiple instantiation.
-
Diego Biurrun authored
The indirection makes no sense without multiple instantiation.
-
Diego Biurrun authored
-
Diego Biurrun authored
This allows detemplatizing the bits that are not instantiated twice.
-
Diego Biurrun authored
Multiple inclusion makes no sense as it is only used in the 8-bit case.
-
Diego Biurrun authored
This fixes FATE with mmxext CPUFLAGS set.
-
Michael Niedermayer authored
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
This only uses the first slice, improvement here is welcome analyzing all slices the trivial way would interfere with threads Fixes Ticket3185 Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
- 25 Mar, 2014 9 commits
-
-
Michael Niedermayer authored
* qatar/master: avcodec: add missing includes Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit 'aa499568': avconv: More descriptive message about framedrop Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
Michael Niedermayer authored
* commit '6ee55c7b': vf_transpose: K&R formatting cosmetics Conflicts: libavfilter/vf_transpose.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
-
James Almer authored
AV_CPU_FLAG_AVX is enabled at this point only if there's OS support. Signed-off-by: James Almer <jamrial@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
rogerdpack authored
Signed-off-by: rogerdpack <rogerpack2005@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
rogerdpack authored
Signed-off-by: rogerdpack <rogerpack2005@gmail.com> Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
-
Carl Eugen Hoyos authored
-
Carl Eugen Hoyos authored
Fixes ticket #3495.
-
Øyvind Kolås authored
Both of these dithering methods are from http://pippin.gimp.org/a_dither/ for GIF they can be considered better than bayer (provides more gray-levels), and spatial stability - often more than twice as good compression and less visual flicker than error diffusion methods (the methods also avoids error-shadow artifacts of diffusion dithers). These methods are similar to blue/green noise type dither masks; but are simple enough to generate their mask on the fly. They are still research work in progress; though more expensive to generate masks (which can be used in a LUT) like 'void and cluster' and similar methods will yield superior results
-