Martin Hořeňovský d99eb8bec8

Optimize 64x64 extended multiplication implementation ...

Now we use intrinsics when possible, and fallback to optimized
implementation in portable C++. The difference is about 4x when
we can use intrinsics and about 2x when we cannot.

This should speed up our Lemire's algorithm implementation nicely.

2024-04-03 13:28:25 +02:00

..

Baselines

less copies and allocations in replaceInPlace

2024-03-01 21:24:45 +01:00

helpers

Support literal-zero detectors using consteval int constructors

2024-02-12 00:52:53 +01:00

IntrospectiveTests

Optimize 64x64 extended multiplication implementation

2024-04-03 13:28:25 +02:00

Misc

Move tests from projects/ to tests/

2019-12-05 16:00:20 +01:00

TimingTests

Fix references to license file

2022-10-28 11:30:15 +02:00

UsageTests

Ignore C++98 related compiler warnings

2024-03-12 22:56:37 +01:00

TestRegistrations.cpp

Ignore C++98 related compiler warnings

2024-03-12 22:56:37 +01:00