| Commit message (Expand) | Author | Age | Files | Lines |
* | Replace complicated and currently broken manual alignment code by | Reimar Döffinger | 2007-08-30 | 1 | -3/+1 |
* | typo | Michael Niedermayer | 2007-08-27 | 1 | -1/+1 |
* | CONFIG_7REGS has been renamed to HAVE_7REGS | Ramiro Polla | 2007-08-26 | 2 | -4/+4 |
* | workaround gcc bug, untested as my gcc is not complaining | Michael Niedermayer | 2007-08-26 | 2 | -0/+6 |
* | optimize the first vertical lifting step, this also prevents another | Michael Niedermayer | 2007-08-26 | 1 | -25/+25 |
* | optimize 1st horizontal lifting step | Michael Niedermayer | 2007-08-26 | 1 | -34/+22 |
* | typo | Michael Niedermayer | 2007-08-26 | 1 | -1/+1 |
* | get rid of totally senseless "m" + read in register we have enough | Michael Niedermayer | 2007-08-26 | 1 | -72/+52 |
* | simplify senselessly complex addressing | Michael Niedermayer | 2007-08-26 | 1 | -27/+28 |
* | cosmetics | Michael Niedermayer | 2007-08-26 | 1 | -36/+27 |
* | avoid overflow in the 3rd lifting step, this now needs mmx2 at minimum | Michael Niedermayer | 2007-08-26 | 2 | -30/+77 |
* | avoid an overflow in the 1 horizontal lifting step | Michael Niedermayer | 2007-08-25 | 1 | -12/+12 |
* | prevent one overflow in the first vertical lifting step | Michael Niedermayer | 2007-08-25 | 1 | -6/+6 |
* | update mmx code to latest snow changes | Michael Niedermayer | 2007-08-25 | 2 | -238/+210 |
* | use 16bit IDWT (a SIMD implementation of it should be >2x faster then with | Michael Niedermayer | 2007-08-25 | 1 | -0/+2 |
* | Change rounding of the horizontal DWT to match the vertical one. | Michael Niedermayer | 2007-08-21 | 1 | -27/+17 |
* | remove code which become unused by the previous changes | Michael Niedermayer | 2007-08-21 | 1 | -9/+0 |
* | Simplify and optimize the 4th vertical lifting step of the SSE2 code (untested) | Michael Niedermayer | 2007-08-21 | 1 | -2/+1 |
* | Simplify and optimize the 4th vertical lifting step of the MMX code | Michael Niedermayer | 2007-08-21 | 1 | -2/+1 |
* | Simplify and speedup code, reduce needed headroom by 2 bits in the 3rd | Michael Niedermayer | 2007-08-20 | 1 | -5/+4 |
* | simplify, speedup and reduce needed headroom by 2 bits in the 3rd | Michael Niedermayer | 2007-08-20 | 1 | -5/+4 |
* | replace <<1 by add for SSE2 (untested) | Michael Niedermayer | 2007-08-20 | 1 | -2/+2 |
* | replace <<1 by add | Michael Niedermayer | 2007-08-20 | 1 | -2/+2 |
* | slightly change horizontal lift3 so it needs 1 bit less headroom | Michael Niedermayer | 2007-08-20 | 1 | -6/+6 |
* | remove idiotc double subtraction from the sse2 code (untested, no sse2 here) | Michael Niedermayer | 2007-08-20 | 1 | -18/+15 |
* | fixme note | Michael Niedermayer | 2007-08-20 | 1 | -1/+1 |
* | and of course the unneeded double subtractions were blindly put in the | Michael Niedermayer | 2007-08-20 | 1 | -18/+15 |
* | help some gcc version to optimize out those functions | Aurelien Jacobs | 2007-07-24 | 1 | -0/+4 |
* | cosmetics: indentation | Aurelien Jacobs | 2007-07-10 | 1 | -2/+2 |
* | Avoid linking with h263.c functions when the relevant codecs | Aurelien Jacobs | 2007-07-10 | 1 | -0/+3 |
* | make arguments to ssd_int8_vs_int16() const | Måns Rullgård | 2007-07-08 | 1 | -1/+1 |
* | typos | Diego Biurrun | 2007-07-05 | 2 | -2/+2 |
* | license header consistency cosmetics | Diego Biurrun | 2007-07-05 | 1 | -1/+1 |
* | Group all copyright and author notices together. | Diego Biurrun | 2007-07-05 | 6 | -66/+56 |
* | add a comment to indicate which #endif belong to which #define | Guillaume Poirier | 2007-06-17 | 1 | -1/+1 |
* | add multiple inclusion guards to headers | Måns Rullgård | 2007-06-17 | 1 | -0/+4 |
* | correct type for {put,avg}_h264_chroma_mc2_mmx2 | Måns Rullgård | 2007-06-16 | 1 | -2/+2 |
* | misc typo fixes | Diego Biurrun | 2007-06-12 | 2 | -3/+3 |
* | misc spelling fixes | Diego Biurrun | 2007-06-12 | 5 | -14/+15 |
* | work around issues with the old version of Gnu Assembler shipped on | Graham Booker | 2007-06-08 | 1 | -8/+8 |
* | Preprocessor #defines starting with __ are reserved for the system. | Diego Biurrun | 2007-06-07 | 1 | -1/+1 |
* | Use FFmpeg-specific CPU feature definitions. | Ronald S. Bultje | 2007-06-07 | 1 | -17/+17 |
* | rename attribute_unused to av_unused and moves its declaration to common.h | Carl Eugen Hoyos | 2007-05-30 | 1 | -5/+5 |
* | Remove incorrect comment; MMX2 is preferred over 3DNow! on Athlon | Zuxy Meng | 2007-05-20 | 1 | -1/+0 |
* | 3DNow! and SSSE3 optimization to QNS DSP functions; use pmulhrw/pmulhrsw inst... | Zuxy Meng | 2007-05-18 | 2 | -80/+170 |
* | better separation of vp3dsp functions from dsputil_mmx.c | Aurelien Jacobs | 2007-05-16 | 5 | -24/+89 |
* | Add libavcodec to compiler include flags in order to simplify header | Ronald S. Bultje | 2007-05-16 | 15 | -21/+21 |
* | Make vp3dsp*.c compilation optional. | Panagiotis Issaris | 2007-05-14 | 1 | -1/+2 |
* | Change some leftover __attribute__((unused)) and __attribute__((used)) to | Reimar Döffinger | 2007-05-14 | 1 | -2/+2 |
* | Faster SSE FFT/MDCT, patch by Zuxy Meng %zuxy P meng A gmail P com% | Zuxy Meng | 2007-05-13 | 1 | -14/+66 |