Skip to content

Latest commit

 

History

History
84 lines (61 loc) · 3.61 KB

CHANGELOG.md

File metadata and controls

84 lines (61 loc) · 3.61 KB

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog and this project adheres to Semantic Versioning (as of v7.0.0).

Unreleased

[v8.0.2] - 2019-05-22

Added

  • .vscode/ directory added to .gitignore.

Fixed

  • Fixed bug where the FpF class was calculating incorrect results when peforming certain arithmetic (and when constructing), due to hard coded use of int32_t, where instead it should be using BaseClass (thanks TheZoq2 for finding this and sending in a MR). Added unit tests to prevent regresison.
  • Fixed URL to git repo in build section of README.
  • Updated dependencies in README.

[v8.0.1] - 2018-06-02

Fixed

  • Fixed bug where std::string and std::ostream headers were not included in FpF.hpp, closes #95.

v8.0.0 - 2018-02-12

Added

  • Added CLion project files, closes 92.
  • Added overload support for printing To ostream using '<<' (and a 'ToString()' method), closes #67.
  • Added unit tests for arithmetic between 'FpF' and 'int' objects.

Fixed

  • Fixed bug where code coverage library was looked for even if COVERAGE set to false, closes #91.
  • Removed un-used 'CountLeadingZeros()' and 'fixinv()' functions from 'FpF.hpp'.
  • Improved speed of FpF division and fixed incorrect result when BaseType was greater than 32 bits, closes #93.

v8.0.0-beta.2 - 2018-01-11

Added

  • Added compile time check to FpF class to make sure template parameters are the same before doing multiplication.
  • Added README code to a new example folder (to make sure it compiles correctly).
  • Added conversion methods, casting capabilities and associated unit tests for the FpF class.

Fixed

  • Removed reference to non-existant MFixedPoint library when linking the benchmark and unit test code.
  • Fixed failing sudo make install command.
  • Fixed failing FpF multiplication.
  • Fixed bug where FpF constructor accepting a double was first converting value to a float.
  • Example in README now compiles correctly.
  • Fixed bug where bit-shifting was overflowing in fixed-point constructors from doubles because '1' was not being cast to BaseType first.

v8.0.0-beta.1 - 2018-01-09

Added

  • Added Visual Studio Code project files, closes #89.
  • Added generic fixed-point classes (both slow and fast) with storage type for bits (aka base type) and overflow type templated, closes #90.

Changed

  • Improved the comments and code formatting of Fp32f.hpp.

Removed

  • All existing fixed-point classes! This is in favour of the generic fixed-point classes mentioned in the 'Added' section.

v7.0.1 - 2018-01-04

Added

  • Added MIT license as LICENSE.txt, closes #88.

[v7.0.0] - 2017-12-13

Added

  • Added enclosing mn namespace for all source code, closes #84.
  • Added this CHANGELOG.md file, closes #85.
  • Added documentation to 'docs/', closes #87.
  • Added code coverage support.

Changed

  • Changed the Fp namespace to MFixedPoint (to reflect actual project name), closes #86.
  • Modified header guard macros to include the text MN_MFIXEDPOINT_... (to include organization/project name).