| Commit message (Expand) | Author | Age | Files | Lines |
* | get_pixels_sse2, ~+12% performance compared to mmx | Baptiste Coudurier | 2008-10-09 | 1 | -0/+35 |
* | Add a bunch of const qualifiers to function arguments. | Diego Biurrun | 2008-10-07 | 1 | -12/+17 |
* | Sync cosmetics from upstream. | Diego Biurrun | 2008-10-07 | 1 | -0/+3 |
* | Declare ff_pw_53 and ff_pw_18 as assembler constants. | Diego Pettenò | 2008-10-04 | 1 | -2/+2 |
* | Remove duplication of TRANSPOSE8 macro, also removing some redundant load/stores | David Conrad | 2008-09-18 | 1 | -58/+2 |
* | Make a separate macro for writing an entire block from registers to memory | David Conrad | 2008-09-18 | 1 | -9/+12 |
* | Factorize SSE2_(Row|Column)_IDCT into one macro | David Conrad | 2008-09-18 | 1 | -116/+19 |
* | Cosmetics: remove excessive blank lines | David Conrad | 2008-09-17 | 1 | -143/+0 |
* | Ensure MMX/SSE2 VP3 IDCT selection isn't disabled when only Theora is enabled | David Conrad | 2008-09-17 | 1 | -1/+1 |
* | Cosmetics: reindent | David Conrad | 2008-09-17 | 2 | -9/+9 |
* | Rewrite SSE2 VP3 IDCT in inline asm | David Conrad | 2008-09-17 | 1 | -283/+273 |
* | Rewrite MMX VP3 IDCT in inline asm | David Conrad | 2008-09-17 | 1 | -202/+185 |
* | MMX/SSE2 VP3 IDCT are bitexact now that the dequantization matrices are permu... | David Conrad | 2008-09-17 | 1 | -3/+1 |
* | Globally rename the header inclusion guard names. | Stefano Sabatini | 2008-08-31 | 7 | -21/+21 |
* | Use ff_vp3_idct_data in vp3dsp_mmx.c rather than duplicating it | David Conrad | 2008-08-31 | 3 | -24/+2 |
* | Declare ff_vp3_idct_data to be uint16_t | David Conrad | 2008-08-31 | 1 | -1/+1 |
* | Don't declare SSE vp3 idct data static, so it can be used in the mmx version | David Conrad | 2008-08-31 | 1 | -2/+2 |
* | Remove unused dequantization code from SSE VP3 IDCT | David Conrad | 2008-08-30 | 1 | -343/+0 |
* | Use ff_pw_8 in MMX/SSE VP3 IDCT | David Conrad | 2008-08-30 | 2 | -16/+5 |
* | Let ff_pw_8 be used as an SSE constant | David Conrad | 2008-08-30 | 2 | -2/+2 |
* | Add explicit (int) cast to i386 optimized MUL* macros. | Vladimir Voroshilov | 2008-08-26 | 1 | -3/+3 |
* | stricter constraints of asm() blocks | Alexis Ballier | 2008-08-24 | 1 | -6/+6 |
* | file which should have been added in r14749 | Loren Merritt | 2008-08-14 | 1 | -0/+92 |
* | missing prototype | Loren Merritt | 2008-08-14 | 1 | -0/+1 |
* | gcc chokes on the 7 registers needed for float_to_int16_interleave6 (even ins... | Loren Merritt | 2008-08-14 | 1 | -43/+19 |
* | gcc chokes on xmm constraints, so pessimize int32_to_float_fmul_scalar_sse a ... | Loren Merritt | 2008-08-14 | 1 | -2/+2 |
* | special case 6 channel version of float_to_int16_interleave | Loren Merritt | 2008-08-13 | 1 | -5/+51 |
* | simd int->float | Loren Merritt | 2008-08-13 | 1 | -0/+46 |
* | simd downmix | Loren Merritt | 2008-08-13 | 1 | -0/+100 |
* | cosmetics in dsp init | Loren Merritt | 2008-08-12 | 1 | -6/+4 |
* | remove mdct tmp buffer | Loren Merritt | 2008-08-12 | 2 | -4/+2 |
* | optimize imdct_half: | Loren Merritt | 2008-08-12 | 2 | -283/+191 |
* | split-radix FFT | Loren Merritt | 2008-08-12 | 4 | -309/+523 |
* | import yasm macros from x264 | Loren Merritt | 2008-08-11 | 1 | -0/+546 |
* | Mark add_png_paeth_prediction_* functions which are only used within this file | Uoti Urpala | 2008-08-02 | 1 | -1/+1 |
* | Fix h264_loop_filter_strength_mmx2() so it works with PAFF. | Michael Niedermayer | 2008-07-19 | 1 | -1/+11 |
* | float_to_int16_interleave: change src to an array of pointers instead of assu... | Loren Merritt | 2008-07-16 | 1 | -10/+10 |
* | 10l, float_to_int16_interleave_sse/3dnow wrote the wrong samples | Loren Merritt | 2008-07-15 | 1 | -2/+2 |
* | exploit mdct symmetry | Loren Merritt | 2008-07-13 | 3 | -22/+161 |
* | simplify vorbis windowing | Loren Merritt | 2008-07-13 | 1 | -1/+120 |
* | SSE2 optimizations for Monkey's Audio decoder vector functions | Kostya Shishkov | 2008-07-11 | 1 | -0/+78 |
* | Make the function prototype visible to comply with C99 inline. | Alexander Strange | 2008-07-09 | 1 | -0/+1 |
* | float_to_int16_sse2() | Michael Niedermayer | 2008-07-09 | 1 | -0/+20 |
* | Make LOAD4/STORE4 macros more generic. | Victor Pollex | 2008-07-08 | 2 | -18/+18 |
* | 1 c-asm loop less and 1x unroll of float_to_int16_sse() | Michael Niedermayer | 2008-07-07 | 1 | -12/+19 |
* | Fix x86-64 | Michael Niedermayer | 2008-07-07 | 1 | -2/+2 |
* | dont use C-asm loops and unroll once float_to_int16_3dnow() | Michael Niedermayer | 2008-07-07 | 1 | -12/+19 |
* | Realign newlines. | Alexander Strange | 2008-06-28 | 1 | -62/+62 |
* | Use MANGLE() instead of memory operands to read globals. | Alexander Strange | 2008-06-28 | 1 | -10/+10 |
* | Fix add_bytes_mmx and add_bytes_l2_mmx for w < 16 | Reimar Döffinger | 2008-06-22 | 1 | -2/+6 |