apostolos
|
0dce69a6f1
|
Supervector Unit Tests
|
2021-07-26 00:10:54 +03:00 |
|
George Wort
|
530b33a309
|
Add SVE2 support for noodle
Change-Id: Iacb7d1f164bdd0ba50e2e13d26fe548cf9b45a6a
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
655c2eb87f
|
add missing ARM SuperVector methods, some tests still fail, WIP
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
799026e760
|
disable SuperVector unit tests for now, until ARM support is included
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
38600b717e
|
rename arm impl.hpp to impl.cpp, add operator|() to SuperVector class
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
b33b7a3ee4
|
refactor shufti algorithm to use SuperVector class, WIP
|
2021-07-26 00:10:54 +03:00 |
|
George Wort
|
27b1bea462
|
Add SVE, SVE2, and SVE2_BITPERM as targets
Change-Id: I5231e2eb0a31708a16c853dc83ea48db32e0b0a5
|
2021-07-26 00:10:54 +03:00 |
|
George Wort
|
2c350e7803
|
Enable cross compilation to aarch64
Change-Id: Iafc8ac60926f5286990ce63a4ff4f8b6a7c46bef
|
2021-07-26 00:10:54 +03:00 |
|
apostolos
|
32bd8b131c
|
SuperVector unit tests
|
2021-07-26 00:10:54 +03:00 |
|
apostolos
|
718641dfda
|
unit tests for supervector
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
cc26212560
|
add more functions, move defines here, enable inlining of template specializations only when running optimized code
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
74e460d89b
|
fix compilation on C++
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
492874db43
|
simplify function
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
c4685b14a6
|
move definitions elsewhere
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
faae8819f3
|
move SuperVector versions of noodleEngine scan functions to _simd.hpp file
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
d232a50d72
|
add arm support for the new SuperVector class
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
6e804d8b16
|
simplify scanSingleMain() and scanDoubleMain()
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
0ccf9280ff
|
delete separate implementations
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
73199585dd
|
add C++ template SIMD library (WIP)
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
dc519e22fc
|
add generic SIMD implementation
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
ca9d380a6f
|
rename project, change to noodle_engine.cpp
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
80a6b55bd4
|
convert to C++
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
61f68f9d3b
|
revert to push_back()
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
8d9dd3c286
|
add global definitions for CHUNKSIZE/VECTORSIZE, define HAVE_AVX512* only when BUILD_AVX512 is also enabled
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
c15018286a
|
fix typo
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
7c6e47c95b
|
replace push_back by emplace_back where possible
|
2021-07-26 00:10:54 +03:00 |
|
Konstantinos Margaritis
|
df769a9a9a
|
use -O3 for C++ code as well, makes a difference
|
2021-07-26 00:10:54 +03:00 |
|
Robbie Williamson
|
a280a5e253
|
Update README.md
Updated the Why the Fork section to be a bit more opensource friendly ;-).
|
2021-04-01 18:08:34 +03:00 |
|
Konstantinos Margaritis
|
8d464477ad
|
minor optimizations
|
2021-03-16 17:47:00 +02:00 |
|
Konstantinos Margaritis
|
40531ad6e5
|
prefetch works best when addresses are 64-byte aligned
|
2021-03-12 10:10:53 +02:00 |
|
Konstantinos Margaritis
|
53a3e33204
|
Revert "replace long macro and switch statement with function pointer array and branchless execution"
This reverts commit cc9dfed249 .
|
2021-02-26 16:40:58 +02:00 |
|
Konstantinos Margaritis
|
cc9dfed249
|
replace long macro and switch statement with function pointer array and branchless execution
|
2021-02-26 16:39:24 +02:00 |
|
Konstantinos Margaritis
|
87f3ab0463
|
optimise case handling
|
2021-02-22 13:59:05 +02:00 |
|
Konstantinos Margaritis
|
afe7061860
|
simplify and make scanSingle*()/scanDouble*() more uniform
|
2021-02-19 12:16:43 +02:00 |
|
Konstantinos Margaritis
|
d0c3764a4c
|
optimize caseMask handling
|
2021-02-16 22:10:42 +02:00 |
|
Konstantinos Margaritis
|
2eab2ec47b
|
bump version
|
2021-02-15 20:18:06 +02:00 |
|
Konstantinos Margaritis
|
27bd09454f
|
use correct function names for AVX512, fix build failure
v5.4.1+vectorscan
|
2021-02-15 13:54:19 +02:00 |
|
Konstantinos Margaritis
|
741d8246c5
|
fix some AVX512 function names, to fix AVX512 build failure, also rename the expand* functions to broadcast*() ones for consistency
|
2021-02-15 13:54:19 +02:00 |
|
Konstantinos Margaritis
|
c3c68b1c3f
|
fix x86 implementations for compress128/expand128
|
2021-02-15 13:54:19 +02:00 |
|
Konstantinos Margaritis
|
e21305aa23
|
align array
|
2021-02-15 13:54:19 +02:00 |
|
Konstantinos Margaritis
|
04567ab649
|
use correct include
|
2021-02-15 13:54:19 +02:00 |
|
Konstantinos Margaritis
|
814045201f
|
add BUILD_AVX2 definition, enable non-AVX2 building selectively
|
2021-02-15 13:54:19 +02:00 |
|
Konstantinos Margaritis
|
c078d355b6
|
Merge branch 'develop'
|
2021-02-15 13:44:30 +02:00 |
|
Konstantinos Margaritis
|
9fd94e0062
|
use unaligned loads for short scans
|
2021-02-11 14:21:57 +02:00 |
|
Konstantinos Margaritis
|
d3e03ed88a
|
optimize case mask AND out of the loop
|
2021-02-10 13:29:45 +02:00 |
|
Konstantinos Margaritis
|
be66cdb51d
|
fixes in shifting primitives
|
2021-02-08 19:38:20 +02:00 |
|
Konstantinos Margaritis
|
f541f75400
|
bugfix compress128/expand128, add unit tests
|
2021-02-08 19:20:37 +02:00 |
|
Konstantinos Margaritis
|
d9874898c7
|
make const
|
2021-02-08 19:19:52 +02:00 |
|
Konstantinos Margaritis
|
70c54ef144
|
Merge branch 'develop' of github.com:VectorCamp/vectorscan into develop
|
2021-01-26 18:22:28 +02:00 |
|
Konstantinos Margaritis
|
4cc93f5553
|
add necessary copyright info
|
2021-01-25 15:42:18 +02:00 |
|