81 Commits (5ee52ef403cfcd4a9b70e014a3fa2ded1eecabf1)

Author SHA1 Message Date
Stefan Nikolei fc211a7aa7 Update Intrinsic usage 6 months ago
Stefan Nikolei ad4b3e90b2 Update ISA Groups 6 months ago
James Jackson-South 8a23d42bfd Port more V256 code 1 year ago
James Jackson-South 038f047a1c Initial fixes based on feedback 1 year ago
James Jackson-South 041e59dbce All Vector128 Load 1 year ago
James Jackson-South 29a56350ce Clean up and prep for Vector512 multiply 1 year ago
James Jackson-South 69caa490e0 Add Vector128 rounding 1 year ago
Günther Foidl a534328dc4 Optimized Block8x8F 3 years ago
James Jackson-South 0e9a066195 FS Namespaces and Implicit Global Usings 4 years ago
James Jackson-South eee64f25d6 Remove SUPPORTS conditions 4 years ago
James Jackson-South 97baeaf438 Update all files 4 years ago
Dmitry Pentin 219119aad4 Tests fix 4 years ago
Dmitry Pentin 7a2c22a966 Removed obsolete/redundant/unused code from Block8x8F 5 years ago
Dmitry Pentin 92445c3581 Added test for TransposeInplace 5 years ago
James Jackson-South 27e6b58643 Final final warnings 5 years ago
Dmitry Pentin 6532552b6b Naming fix & simd else if branch 5 years ago
Dmitry Pentin 4fd912b9dd Fixed Ssse3 zig-zag implementation 5 years ago
Dmitry Pentin 2f143bf9d3 New FDCT method, reciprocal quantization 5 years ago
Dmitry Pentin 6c5cf28ecd New zig-zag implementation 5 years ago
Dmitry Pentin 787ffa57ee Removed unused methods & constructor, fixed warnings 5 years ago
Dmitry Pentin e205bea702 Tests 5 years ago
Brian Popow 074d1a6a12 Add xunit trait attributes to tests 6 years ago
James Jackson-South 68d2406f4e Optimize low hanging fruit and fix naming 6 years ago
James Jackson-South 9380dd5b1e Use single Block8x8F.TransponseInto test 6 years ago
James Jackson-South a7626e16dc Revert to 8e5a59f 6 years ago
James Jackson-South 37b0585feb Use single test, enable runners 6 years ago
James Jackson-South 0c34ce36c6 First pass at HW feature tests 6 years ago
James Jackson-South a62347e761 Update Block8x8FTests.cs 6 years ago
James Jackson-South 3091072e38 Add AVX backed Block8x8F Transpose method 6 years ago
James Jackson-South f533726aef Remove last of pointer code and add benchmarks 6 years ago
James Jackson-South 411f474bfc Remove most pointer usage 6 years ago
James Jackson-South 874115bd0a Update license 6 years ago
James Jackson-South 6183dd8c13 Update source license info. 6 years ago
Anton Firszov 2e878694e3 CopyTo -> ScaledCopyTo 6 years ago
Anton Firszov eb76788232 IsAvx2CompatibleArchitecture => HasVector8 6 years ago
Brian Popow 7b052c8076 Fix stylecop warnings in the test project 6 years ago
Simon Cropp bcbd4726a6 use var where apparent (#972) 7 years ago
Simon Cropp 270a073fc6 Cleanup (#965) 7 years ago
Simon Cropp aa6d2ab997 fix some spelling (#957) 7 years ago
Stephan Vedder 84aef78c1e Work on 12-bit support 8 years ago
Anton Firszov b4c1f8bf37 AVX2 optimized Block8x8 -> Block8x8F conversion 8 years ago
James Jackson-South df6f544341 Move common components 8 years ago
Jason Nelson e118ee888c Cleanup tests 8 years ago
James Jackson-South 51e7f03393 Update tests 9 years ago
Anton Firszov 97dc26c01a additional Block8x8F cleanup 9 years ago
Anton Firszov 63e574cdb2 more cleanup + more tests 9 years ago
Anton Firszov 8b9a369504 cleanup and docs 9 years ago
Anton Firszov 57ccde4521 NormalizeColorsAndRoundAvx2() + JpegBlockPostProcessor cleanup 9 years ago
Anton Firszov 005a454b40 optimized quantization 9 years ago
Anton Firszov 2a7ce4cd90 fix rounding in YCbCr conversion, started refactoring dequantiziation logic 9 years ago