|
|
@6143
|
7 months |
xwa163 |
lz4_grep: some bug fixing and refactor
|
|
|
@6142
|
7 months |
cameron |
Recursive application of NFD, NFKD
|
|
|
@6141
|
7 months |
cameron |
Support for searching grapheme clusters in all possible reorderings
|
|
|
@6140
|
7 months |
xwa163 |
UTF-8 support for Multiplexing LZ4 Grep
|
|
|
@6139
|
7 months |
cameron |
Adding Hangul decomposition algorithm
|
|
|
@6138
|
7 months |
cameron |
Unicode decomposition, casefolding: initial check-in
|
|
|
@6137
|
7 months |
xwa163 |
1. LZ4 ScanMatch? pipeline
2. Refactor LZ4 Generator
3. Adjust some naming
|
|
|
@6136
|
7 months |
xwa163 |
1. Cleanup legacy slow LZ4 related kernels
2. Rename lz4d_ext_dep to …
|
|
|
@6135
|
7 months |
xwa163 |
1. Implement twist_kernel and untwist_kernel by PEXT and PDEP
2. Use …
|
|
|
@6134
|
7 months |
cameron |
GetStringValue? method for string properties
|
|
|
@6133
|
7 months |
xwa163 |
1. Add sourceCC in multiplexed CC
2. Remove workaround FakeBasisBits? …
|
|
|
@6132
|
7 months |
xwa163 |
1. More experiment on lz4 grep
2. Improve performance of lzparabix grep
|
|
|
@6131
|
8 months |
xwa163 |
Minor bug fix for LZParabixAioKernel
|
|
|
@6130
|
8 months |
xwa163 |
Decompress u8NoFinalStream in lzparabix_grep
|
|
|
@6129
|
8 months |
cameron |
pablo.Infile(0) optimization
|
|
|
@6128
|
8 months |
xwa163 |
lzparabix_grep: Improve the performance by merging LineFeed? stream …
|
|
|
@6127
|
8 months |
xwa163 |
1. Improve RE_Compiler for multiplexing
2. Use faster approach for …
|
|
|
@6126
|
8 months |
cameron |
Big-endian numbering option for CharClasses? kernel
|
|
|
@6125
|
8 months |
cameron |
Eliminate code to reverse bit numbering
|
|
|
@6124
|
8 months |
xwa163 |
Minor bug fix for lzparabix grep and lz4 grep
|
|
|
@6123
|
8 months |
xwa163 |
Encode BitStream? directly in LZParabix compressed file
|
|
|
@6122
|
8 months |
cameron |
include <fstream>
|
|
|
@6121
|
8 months |
cameron |
Fixes for rand/srand signatures
|
|
|
@6120
|
8 months |
cameron |
Random value streams
|
|
|
@6119
|
8 months |
xwa163 |
1. Add some BasisSetNumbering? option to fix bug of multiplexing
2. Use …
|
|
|
@6118
|
8 months |
xwa163 |
Implement bitstream decompression version of LZ4 Aio Kernel
|
|
|
@6117
|
8 months |
xwa163 |
Fix bugs of LZParabixAioKernel
|
|
|
@6116
|
8 months |
xwa163 |
Small bug fix for LZParabix Compression kernel
|
|
|
@6115
|
8 months |
xwa163 |
fix bug of lzparabix compressoin kernel
|
|
|
@6114
|
8 months |
xwa163 |
Init check in for new compression format (lzparabix) related kernels …
|
|
|
@6113
|
8 months |
cameron |
hsimd_signmask support for various BlockSizes?
|
|
|
@6112
|
8 months |
xwa163 |
Remove S2PByPextKernel
|
|
|
@6111
|
8 months |
xwa163 |
1. Cleanup LZ4 AIO related kernels
2. Improve …
|
|
|
@6110
|
8 months |
cameron |
Prefer simd_select_hi/select_lo in place of simd_himask/lomask for …
|
|
|
@6109
|
8 months |
cameron |
IDISA progress
|
|
|
@6108
|
8 months |
cameron |
IDISA builder progress
|
|
|
@6107
|
8 months |
cameron |
Fixes for NativeBitBlockWidth?
|
|
|
@6106
|
8 months |
cameron |
mergeh, mergel for arbitrary BitBlockWidth?
|
|
|
@6105
|
8 months |
cameron |
NativeBitWidth? for idisa_builders, laneWidth replaces stride
|
|
|
@6104
|
8 months |
cameron |
getArgumentList() deprecated
|
|
|
@6103
|
8 months |
cameron |
AVX-512 pshufb has a new intrinsic as of LLVM 4.0
|
|
|
@6102
|
8 months |
cameron |
Use CreateZExt instead of AVX-512 for vector zeroextend
|
|
|
@6101
|
8 months |
cameron |
esimd_merge 8 for AVX-512; improves p2s performance
|
|
|
@6100
|
8 months |
cameron |
esimd bit merge optimizations for AVX-512
|
|
|
@6099
|
8 months |
cameron |
Optimization of simd_if for AVX-512 using ternary logic instruction …
|
|
|
@6098
|
8 months |
cameron |
esimd_mergel/h 1 for AVX-512
|
|
|
@6097
|
8 months |
cameron |
Small fix
|
|
|
@6096
|
8 months |
cameron |
simd_fill fix
|
|
|
@6095
|
8 months |
cameron |
u32u8 for 32-bit
|
|
|
@6094
|
8 months |
cameron |
mvmd_compress 32 for SSE2
|
|
|
@6093
|
8 months |
cameron |
u32u8_test target
|
|
|
@6092
|
8 months |
cameron |
Generic esimd_merge functions for fw = 1, 2
|
|
|
@6091
|
8 months |
cameron |
u32u8 test
|
|
|
@6090
|
8 months |
cameron |
Updated character_deletion
|
|
|
@6089
|
8 months |
cameron |
Little-endian/big-endian bit number options, default to little-endian
|
|
|
@6088
|
8 months |
cameron |
Allow a name suffix for FieldDepositKernel? to distinguish multiple …
|
|
|
@6087
|
8 months |
cameron |
IDISA builder updates for esimd_mergel/h
|
|
|
@6086
|
8 months |
cameron |
Bug fix for StreamCompress?
|
|
|
@6085
|
8 months |
cameron |
u32u8 progress
|
|
|
@6084
|
8 months |
xwa163 |
Use scatter for match copy in LZ4ParallelByteStreamAIOKernel
|
|
|
@6083
|
8 months |
cameron |
PrintRegister? goes to stderr
|
|
|
@6082
|
8 months |
xwa163 |
One missing line for AVX512 bug fixing
|
|
|
@6081
|
8 months |
xwa163 |
1. Add command line parameter -enable-gather and -enable-scatter in …
|
|
|
@6080
|
9 months |
xwa163 |
Use AVX512 gather and scatter intrinsic for literal copy in …
|
|
|
@6079
|
9 months |
cameron |
StreamExpandKernel?: eliminate kernel state: rely on PopcountOf? attribute
|
|
|
@6078
|
9 months |
cameron |
Interim bug fix for StreamExpandKernel?
|
|
|
@6077
|
9 months |
xwa163 |
Some bug fixing for LZ4 related kernels in AVX512 machine
|
|
|
@6076
|
9 months |
cameron |
IDISA_SSSE3 builder; other IDISA operations
|
|
|
@6075
|
9 months |
cameron |
Final block processing for u32u8 mask calculation
|
|
|
@6074
|
9 months |
xwa163 |
Bugs fix in LZ4ParallelByteStreamAIO for AVX512
|
|
|
@6073
|
9 months |
cameron |
Revert r6068: issues with setStride
|
|
|
@6072
|
9 months |
cameron |
character_deposit UseStreamDepositCompiler? option
|
|
|
@6071
|
9 months |
cameron |
u32u8.cpp initial check-in
|
|
|
@6070
|
9 months |
xwa163 |
Improve performance of literal copy and match copy in …
|
|
|
@6069
|
9 months |
cameron |
S2P_21 kernel for transposition of UTF-32; steps toward little-endian …
|
|
|
@6068
|
9 months |
cameron |
Stdout and Filesink kernels have improved performance with pagesize() …
|
|
|
@6067
|
9 months |
cameron |
Source kernels bug fix: size/items calculations when codeUnitSize != 8
|
|
|
@6066
|
9 months |
xwa163 |
fix some warning in lz4 related kernels
|
|
|
@6065
|
9 months |
xwa163 |
1. Fix some typo in LZ4 Grep Extract and Deposit pipeline
2. Small fix …
|
|
|
@6064
|
9 months |
xwa163 |
Init checkin for LZ4ParallelByteStreamAioKernel and related pipeline
|
|
|
@6063
|
9 months |
cameron |
bit interleave (esimd_merge<1>) for AVX2
|
|
|
@6062
|
9 months |
xwa163 |
fix output for lz4 count-only grep pipeline
|
|
|
@6061
|
9 months |
xwa163 |
Improve the output performance of LZ4ByteStreamAIO kernel by using …
|
|
|
@6060
|
9 months |
xwa163 |
Implement S2PByPextKernel
|
|
|
@6059
|
9 months |
xwa163 |
1. Enable swizzled match copy in multiplexing lz4_grep for some …
|
|
|
@6058
|
9 months |
cameron |
Small fixes
|
|
|
@6057
|
9 months |
cameron |
mvmd_shuffle<64> for SSE2
|
|
|
@6056
|
9 months |
cameron |
mvmd_dslli, mvmd_shuffle fixes for AVX-512
|
|
|
@6055
|
9 months |
cameron |
Various small fixes
|
|
|
@6054
|
9 months |
cameron |
IDISA testing: quiet mode, return code, pdep64 test case
|
|
|
@6053
|
9 months |
cameron |
esimd_mergel 4 argument order fix
|
|
|
@6052
|
9 months |
cameron |
IDISA test checking, more test operations, IDISA fixes
|
|
|
@6051
|
9 months |
cameron |
IDISA testing
|
|
|
@6050
|
9 months |
cameron |
IDISA testing
|
|
|
@6049
|
9 months |
nmedfort |
Bug fix for ReadSourceKernel?
|
|
|
@6048
|
9 months |
nmedfort |
Added USE_ADDRESS_SANITIZER compilation flag
|
|
|
@6047
|
9 months |
nmedfort |
Major refactoring of buffer types. Static buffers replace Circular and …
|
|
|
@6046
|
9 months |
cameron |
Some fixes
|
|
|
@6045
|
9 months |
cameron |
StreamExpand? kernel and compiler - initial check-in
|
|
|
@6044
|
9 months |
xwa163 |
1. Fix a bug of multiplexing lz4_grep in release build
2. Combine …
|
|
|