| Commit message (Expand) | Author | Age | Files | Lines |
* | avutil/x86util: Fix broken pre-SSE4.1 PMINSD emulation | Henrik Gramner | 2024-03-17 | 1 | -4/+0 |
* | configure: Remove av_restrict | Andreas Rheinhardt | 2024-03-15 | 2 | -7/+2 |
* | x86: Remove inline MMX assembly that clobbers the FPU state | Martin Storsjö | 2024-02-09 | 1 | -36/+0 |
* | x86/tx_init: propely indicate the extended available transform sizes | Lynne | 2024-02-09 | 1 | -9/+9 |
* | x86/tx_float: enable SIMD for sizes over 131072 | Lynne | 2024-02-07 | 1 | -2/+6 |
* | x86inc: Add REPX macro to repeat instructions/operations | Henrik Gramner | 2023-11-08 | 1 | -0/+10 |
* | avutil/x86/pixelutils: Empty MMX state in ff_pixelutils_sad_8x8_mmxext | Andreas Rheinhardt | 2023-11-04 | 1 | -0/+1 |
* | avutil/internal: Don't auto-include emms.h | Andreas Rheinhardt | 2023-09-04 | 1 | -58/+0 |
* | x86: replace explicit REP_RETs with RETs | Lynne | 2023-02-01 | 2 | -11/+11 |
* | x86/tx_float: fix stray change in 15xM FFT and replace imul->lea | Lynne | 2022-11-28 | 1 | -2/+2 |
* | lavu/tx: refactor to explicitly track and convert lookup table order | Lynne | 2022-11-24 | 1 | -21/+25 |
* | x86/tx_float: implement striding in fft_15xM | Lynne | 2022-11-24 | 1 | -16/+29 |
* | x86/tx_float_init: properly specify the supported factors of 15xM FFTs | Lynne | 2022-11-24 | 1 | -3/+3 |
* | x86/tx_float: add a standalone 15-point AVX2 transform | Lynne | 2022-11-24 | 2 | -0/+117 |
* | x86/tx_float: optimize and macro out FFT15 | Lynne | 2022-11-24 | 1 | -134/+143 |
* | lavu/fixed_dsp: add missing av_restrict qualifiers | Johannes Kauffmann | 2022-10-04 | 1 | -1/+1 |
* | x86/tx_float: enable AVX-only split-radix FFT codelets | Lynne | 2022-09-24 | 2 | -0/+10 |
* | x86/tx_float: fix some symbol names | James Almer | 2022-09-23 | 1 | -3/+3 |
* | x86/tx_float: change a condition in a preprocessor check | James Almer | 2022-09-23 | 1 | -1/+1 |
* | x86/tx_float: add missing preprocessor wrapper for AVX2 functions | James Almer | 2022-09-23 | 1 | -1/+1 |
* | x86/tx_float: generalize iMDCT | Lynne | 2022-09-23 | 2 | -29/+40 |
* | x86/tx_float: add 15xN PFA FFT AVX SIMD | Lynne | 2022-09-23 | 2 | -0/+348 |
* | x86/tx_float: adjust internal ASM call ABI again | Lynne | 2022-09-23 | 1 | -20/+8 |
* | x86/tx_float: add asm call versions of the 2pt and 4pt transforms | Lynne | 2022-09-19 | 2 | -3/+32 |
* | x86/tx_float: fully support 128bit regs in LOAD64_LUT | Lynne | 2022-09-19 | 1 | -5/+5 |
* | x86/tx_float: simplify and describe the intra-asm call convention | Lynne | 2022-09-19 | 1 | -13/+30 |
* | x86/float_dsp: use three operand form for some instructions | James Almer | 2022-09-13 | 1 | -8/+8 |
* | avutil/x86/float_dsp: add fma3 for scalarproduct | Paul B Mahol | 2022-09-13 | 2 | -0/+129 |
* | avutil/x86/intreadwrite: Add ability to detect whether MMX code is used | Andreas Rheinhardt | 2022-09-11 | 1 | -0/+2 |
* | x86/tx_float: add missing check for AVX2 | James Almer | 2022-09-06 | 1 | -1/+1 |
* | x86/tx_float: set all operands for shufps | James Almer | 2022-09-06 | 1 | -2/+2 |
* | x86/tx_float: Fix building for platforms with a symbol prefix | Martin Storsjö | 2022-09-06 | 1 | -5/+5 |
* | x86/tx_float: implement inverse MDCT AVX2 assembly | Lynne | 2022-09-06 | 2 | -1/+216 |
* | x86/tx_float: add support for calling assembly functions from assembly | Lynne | 2022-09-06 | 2 | -47/+138 |
* | x86/tx_float: save a branch during coefficient deinterleaving | Lynne | 2022-08-09 | 1 | -4/+1 |
* | avutil/x86/float_dsp: Remove obsolete 3dnowext function | Andreas Rheinhardt | 2022-06-22 | 2 | -29/+1 |
* | avutil/x86/pixelutils: Remove obsolete MMX(EXT) functions | Andreas Rheinhardt | 2022-06-22 | 2 | -67/+0 |
* | x86/tx_float: replace fft_sr_avx with fft_sr_fma3 | Lynne | 2022-05-21 | 2 | -9/+9 |
* | x86/tx_float: improve temporary register allocation for loads | Lynne | 2022-05-21 | 1 | -24/+24 |
* | x86/tx_float: add AV_CPU_FLAG_AVXSLOW/SLOW_GATHER flags where appropriate | Lynne | 2022-05-21 | 1 | -14/+21 |
* | Revert "x86/tx_float: remove vgatherdpd usage" | Lynne | 2022-05-21 | 2 | -31/+43 |
* | x86/tx_float: remove vgatherdpd usage | Lynne | 2022-05-20 | 2 | -43/+31 |
* | avutil/cpu: add AVX512 Icelake flag | Wu Jianhua | 2022-03-10 | 3 | -28/+34 |
* | Remove unnecessary libavutil/(avutil|common|internal).h inclusions | Andreas Rheinhardt | 2022-02-24 | 1 | -2/+1 |
* | avutil/x86/emms: Don't unnecessarily include lavu/cpu.h | Andreas Rheinhardt | 2022-02-21 | 1 | -1/+4 |
* | libavutil: include assembly with full path from source root | Alexander Kanavin | 2022-02-08 | 7 | -7/+7 |
* | lavu/tx: refactor assembly codelet definition | Lynne | 2022-02-07 | 1 | -93/+47 |
* | x86/tx_float: avoid redefining macros | Lynne | 2022-02-02 | 1 | -6/+6 |
* | x86/tx_float: mark AVX2 functions as AVXSLOW | Lynne | 2022-01-29 | 1 | -2/+2 |
* | x86/tx_float: add missing FF_TX_OUT_OF_PLACE flag to functions | Lynne | 2022-01-27 | 1 | -2/+2 |