55 Commits

Author SHA1 Message Date
Konstantinos Margaritis
a837cf3bee [VSX] optimize shift operators 2022-09-07 12:16:14 +03:00
Konstantinos Margaritis
be20c2c519 [VSX] optimize shifting methods, replace template Unroller 2022-09-07 12:14:15 +03:00
Konstantinos Margaritis
dc6b8ae92d optimize comparemask implementation, clean up code, use union types instead of casts 2022-09-07 02:02:11 +03:00
Danila Kutenin
8a49e20bcd Fix formatting of a couple files 2022-06-26 22:59:58 +00:00
Danila Kutenin
49eb18ee4f Optimize vectorscan for aarch64 by using shrn instruction
This optimization is based on the thread
https://twitter.com/Danlark1/status/1539344279268691970 and uses
shift right and narrow by 4 instruction https://developer.arm.com/documentation/ddi0596/2020-12/SIMD-FP-Instructions/SHRN--SHRN2--Shift-Right-Narrow--immediate--

To achieve that, I needed to redesign a little movemask into comparemask
and have an additional step towards mask iteration. Our benchmarks
showed 10-15% improvement on average for long matches.
2022-06-26 22:55:45 +00:00
Konstantinos Margaritis
467db4a268 Minor changes to enable compilation on Mac M1 2021-12-11 15:43:55 +02:00
Konstantinos Margaritis
290eabbca0 fix compilation with clang and some incomplete/wrong implementations for arm this time 2021-12-06 18:22:58 +00:00
Konstantinos Margaritis
5aae719ecd fix build with clang, in particular VSX uses long long instead of int64_t, gcc allows this, clang does not 2021-12-02 18:01:00 +02:00
Konstantinos Margaritis
4aa32275f1 use same definition of the union for all types 2021-12-02 18:00:02 +02:00
Konstantinos Margaritis
0221dc1771 fix misompilations with clang++, as it is more strict 2021-12-01 23:22:15 +02:00
Konstantinos Margaritis
7ceca78db4 fix unit-internal release builds using __builtin_constant_p() as well 2021-11-25 15:09:01 +02:00
Konstantinos Margaritis
cd95b1a38c use __builtin_constant_p() instead for arm as well 2021-11-25 06:20:53 +00:00
Apostolos Tapsas
0287724413 WIP:tracking last bugs in failing tests for release build 2021-11-16 15:24:22 +00:00
apostolos
e09d8674b4 resolving conficts after merging 2021-11-13 18:58:22 +02:00
Apostolos Tapsas
ba90cdeb5a SuperVector constructors as well as andnot implementation fixed 2021-11-05 13:34:48 +00:00
Konstantinos Margaritis
713aaef799 move casemask helper functions to separate header 2021-11-01 16:05:43 +00:00
Konstantinos Margaritis
7b65b298c1 add arm vector types in union, avoid -flax-conversions, fix castings 2021-11-01 16:52:17 +02:00
apostolos
d9d39d48c5 prints commants and formating fixes 2021-11-01 10:09:15 +02:00
apostolos
bf54aae779 Special case for Shuffle test added as well as comments for respectives implementations 2021-10-26 11:48:33 +03:00
Apostolos Tapsas
1eb3b19f63 Shuffle simd and SuperVector implementetions as well as their test realy fixed 2021-10-25 09:19:30 +03:00
Apostolos Tapsas
d43d6733b6 SuperVector shuffle implementation and test function optimized 2021-10-22 11:55:39 +00:00
Apostolos Tapsas
558313a2c2 SuperVector operators fixes and simd_utils low/high64 functions implementations added 2021-10-18 12:26:38 +00:00
Apostolos Tapsas
e084c2d6e4 SuperVector vsh* implementations 2021-10-15 14:07:17 +00:00
apostolos
d0a41252c8 blockSigleMask implementations for ARCH_PPC64 added 2021-10-14 15:56:13 +03:00
apostolos
4d2acd59e2 Supervector vsh* added 2021-10-14 15:08:23 +03:00
Apostolos Tapsas
7888dd4418 WIP: Power VSX support almost completed 2021-10-14 13:53:55 +03:00
Vectorcamp
2231f7c024 compile fixes for vsc port 2021-10-14 13:53:55 +03:00
apostolos
90d3db1776 update powerpc simd util file functions 2021-10-14 13:53:55 +03:00
apostolos
0078c28ee6 implementations for powerpc64el architecture 2021-10-14 13:53:55 +03:00
Konstantinos Margaritis
c3baf3d296 fix multiple/undefined symbols when using fat runtimes 2021-10-12 11:51:35 +03:00
Konstantinos Margaritis
577e03e0c7 rearrange method declarations 2021-10-12 11:51:35 +03:00
Konstantinos Margaritis
8b7ba89cb5 add x86 vsh* implementations 2021-10-12 11:51:35 +03:00
Konstantinos Margaritis
67e0674df8 Changes/Additions to SuperVector class * added ==,!=,>=,>,<=,< operators * reworked shift operators to be more uniform and orthogonal, like Arm ISA * Added Unroller class to allow handling of multiple cases but avoid code duplication * pshufb method can now emulate Intel or not (avoids one instruction). 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
e7161fdfec initial SSE/AVX2 implementation 2021-10-12 11:51:34 +03:00
apostolos
67fa6d2738 alignr methods for avx2 and avx512 added 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
603bc14cdd fix failing corner case, add pshufb_maskz() 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
f2d9784979 fix loadu_maskz, add {l,r}shift128_var(), tab fixes 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
f8ce0bb922 minor fixes, add 2 constructors from half size vectors 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
ebb1b84ae3 provide an {l,r}shift128_var() to fix immediate value build failure in loadu_maskz 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
825460856f fix arm loadu_maskz() 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
86accf41a3 add arm rshift128/rshift128 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
6c51f7f591 add {l,r}shift128()+tests, rename printv_u64() to print64() 2021-10-12 11:51:34 +03:00
apostolos
6f88ecac44 Supervector test fixes 2021-10-12 11:51:34 +03:00
apostolos
ae6bc52076 SuperVector AVX512 implementations 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
7ae636dfe9 really fix lshift for avx2 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
dede600637 lots of fixes to AVX2 implementation 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
c45e72775f convert print helper functions to class methods 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
ec3f108d71 fix arm SuperVector implementation 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
f425951b49 fix x86 debug alignr 2021-10-12 11:51:34 +03:00
Konstantinos Margaritis
6d8f3b9ff8 compilation fixes for debug mode 2021-10-12 11:51:34 +03:00