Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | x86: Drop some unnecessary YASM ifdefs | Diego Biurrun | 2014-04-04 | 1 | -2/+0 |
| | | | | Dead code elimination is enough to avoid undefined references in these cases. | ||||
* | x86/arm: Add clobber tests to libavresample | Martin Storsjö | 2014-01-13 | 2 | -0/+33 |
| | | | | Signed-off-by: Martin Storsjö <martin@martin.st> | ||||
* | x86inc: Remove our FMA4 support | Derek Buitenhuis | 2013-10-14 | 1 | -9/+9 |
| | | | | | | | | This is so we can sync to x264's version of FMA4 support. This partialy reverts commit 79687079a97a039c325ab79d7a95920d800b791f. Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com> | ||||
* | avresample/x86: Switch operand order for mulps | Derek Buitenhuis | 2013-10-14 | 1 | -1/+1 |
| | | | | | | | | With the forthcoming VEX instruction emulation, mulps must have only the third operand point to memory, as this is what vmulps expects. Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com> | ||||
* | Consistently use "cpu_flags" as variable/parameter name for CPU flags | Diego Biurrun | 2013-07-18 | 3 | -24/+24 |
| | |||||
* | cosmetics: Remove unnecessary extern keywords from function declarations | Diego Biurrun | 2013-03-27 | 3 | -167/+166 |
| | |||||
* | x86: lavr: add SSE2/AVX dither_int_to_float() | Justin Ruggles | 2013-01-08 | 2 | -0/+86 |
| | |||||
* | x86: lavr: add SSE2 quantize() for dithering | Justin Ruggles | 2013-01-08 | 3 | -0/+94 |
| | |||||
* | x86: lavr: use the x86inc.asm automatic stack alignment in mixing functions | Justin Ruggles | 2013-01-05 | 1 | -26/+9 |
| | | | | CC:libav-stable@libav.org | ||||
* | x86: lavr: fix stack allocation for 7 and 8 channel downmixing on x86-32 | Ronald S. Bultje | 2012-11-17 | 1 | -1/+9 |
| | | | | Fixes crashes on Win32 and stack overruns on x86-32 in general. | ||||
* | build: Drop AVX assembly ifdefs | Diego Biurrun | 2012-11-11 | 2 | -42/+0 |
| | | | | An assembler able to cope with AVX instructions is now required. | ||||
* | x86: PALIGNR: port to cpuflags | Diego Biurrun | 2012-11-09 | 1 | -6/+0 |
| | |||||
* | x86: avresample: Add missing colons to assembly labels | Diego Biurrun | 2012-11-06 | 1 | -2/+2 |
| | | | | YASM accepts labels without colons, but NASM issues warnings. | ||||
* | x86: yasm: Use complete source path for macro helper %includes | Diego Biurrun | 2012-10-31 | 2 | -2/+2 |
| | | | | | This is more consistent with the way we handle C #includes and it simplifies the build system. | ||||
* | x86: include x86inc.asm in x86util.asm | Diego Biurrun | 2012-10-31 | 2 | -2/+0 |
| | | | | This is necessary to allow refactoring some x86util macros with cpuflags. | ||||
* | lavr: handle clipping in the float to s32 conversion | Justin Ruggles | 2012-10-13 | 1 | -1/+7 |
| | | | | | | | | | | | We cannot clip to INT_MAX because that value cannot be exactly represented by a float value and ends up overflowing during conversion anyway. We need to use a slightly smaller float value, which ends up with slightly inaccurate results for samples which clip or nearly clip, but it is close enough. Using doubles as intermediates in the conversion would be more accurate, but it takes about twice as much time. Signed-off-by: Luca Barbato <lu_zero@gentoo.org> | ||||
* | x86: Replace checks for CPU extensions and flags by convenience macros | Diego Biurrun | 2012-09-08 | 2 | -17/+17 |
| | | | | | This separates code relying on inline from that relying on external assembly and fixes instances where the coalesced check was incorrect. | ||||
* | x86: Split inline and external assembly #ifdefs | Diego Biurrun | 2012-08-31 | 2 | -22/+22 |
| | |||||
* | x86: cosmetics: Comment some #endifs for better readability | Diego Biurrun | 2012-08-30 | 1 | -1/+1 |
| | |||||
* | lavr: x86: optimized 6-channel flt to fltp conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+72 |
| | |||||
* | lavr: x86: optimized 2-channel flt to fltp conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+41 |
| | |||||
* | lavr: x86: optimized 6-channel flt to s16p conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+87 |
| | |||||
* | lavr: x86: optimized 2-channel flt to s16p conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+55 |
| | |||||
* | lavr: x86: optimized 6-channel s16 to fltp conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+98 |
| | |||||
* | lavr: x86: optimized 2-channel s16 to fltp conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+48 |
| | |||||
* | lavr: x86: optimized 6-channel s16 to s16p conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+74 |
| | |||||
* | lavr: x86: optimized 2-channel s16 to s16p conversion | Justin Ruggles | 2012-08-23 | 3 | -0/+71 |
| | |||||
* | lavr: x86: optimized 2-channel fltp to flt conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+44 |
| | |||||
* | lavr: x86: optimized 6-channel fltp to s16 conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+129 |
| | |||||
* | lavr: x86: optimized 2-channel fltp to s16 conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+52 |
| | |||||
* | lavr: x86: optimized 6-channel s16p to flt conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+121 |
| | |||||
* | lavr: x86: optimized 2-channel s16p to flt conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+58 |
| | |||||
* | lavr: x86: optimized 6-channel s16p to s16 conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+137 |
| | |||||
* | lavr: x86: optimized 2-channel s16p to s16 conversion | Justin Ruggles | 2012-08-23 | 2 | -0/+50 |
| | |||||
* | x86: add colons after labels | Mans Rullgard | 2012-08-07 | 1 | -1/+1 |
| | | | | | | nasm prints a warning if the colon is missing. Signed-off-by: Mans Rullgard <mans@mansr.com> | ||||
* | lavr: x86: improve non-SSE4 version of S16_TO_S32_SX macro | Justin Ruggles | 2012-07-27 | 1 | -1/+2 |
| | | | | | Removes a false dependency on existing contents of the 2nd dst register, giving better performance for OOE. | ||||
* | lavr: add x86-optimized mixing functions | Justin Ruggles | 2012-07-27 | 2 | -0/+423 |
| | | | | | Adds optimized functions for mixing 3 through 8 input channels to 1 and 2 output channels in fltp or s16p format with flt coeffs. | ||||
* | x86inc: automatically insert vzeroupper for YMM functions. | Ronald S. Bultje | 2012-07-26 | 2 | -20/+0 |
| | |||||
* | lavr: x86: add missing vzeroupper in ff_mix_1_to_2_fltp_flt() | Justin Ruggles | 2012-07-25 | 1 | -0/+5 |
| | |||||
* | lavr: x86: fix ff_conv_fltp_to_flt_6ch function prototypes | Justin Ruggles | 2012-06-26 | 1 | -3/+6 |
| | | | | | | Changed to match the number of parameters in conv_func_interleave(), which is how they are called. The change isn't strictly necessary because the 4th parameter is not used, but the code is clearer if they match. | ||||
* | lavr: x86: merge some branches | Justin Ruggles | 2012-06-25 | 1 | -8/+4 |
| | |||||
* | lavr: Add x86-optimized function for flt to s32 conversion | Justin Ruggles | 2012-06-18 | 2 | -0/+49 |
| | |||||
* | lavr: Add x86-optimized function for flt to s16 conversion | Justin Ruggles | 2012-06-18 | 2 | -0/+37 |
| | |||||
* | lavr: Add x86-optimized functions for s32 to flt conversion | Justin Ruggles | 2012-06-18 | 2 | -0/+46 |
| | |||||
* | lavr: Add x86-optimized functions for s32 to s16 conversion | Justin Ruggles | 2012-06-18 | 2 | -0/+47 |
| | |||||
* | lavr: Add x86-optimized functions for s16 to flt conversion | Justin Ruggles | 2012-06-18 | 2 | -0/+45 |
| | |||||
* | lavr: Add x86-optimized function for s16 to s32 conversion | Justin Ruggles | 2012-06-18 | 2 | -0/+29 |
| | |||||
* | lavr: add x86-optimized functions for mixing 1-to-2 s16p with flt coeffs | Justin Ruggles | 2012-06-18 | 2 | -0/+60 |
| | |||||
* | lavr: add x86-optimized functions for mixing 1-to-2 fltp with flt coeffs | Justin Ruggles | 2012-06-18 | 2 | -0/+43 |
| | |||||
* | lavr: add x86-optimized function for mixing 2 to 1 s16p with q8 coeffs | Justin Ruggles | 2012-05-29 | 2 | -0/+46 |
| |