lazalong / CFixedPoint

C Fixed Point librarry

C Fixed Point [WIP]

C fixed point math library : 64bit Q48.18

• Type reference

• type min() max()

  • char -128 128
  • uchar 0 255
  • short -32,768 32,767
  • ushort 0 65535
  • wchar_t 0 65535
  • int -2147483648 2147483647
  • uint 0 4,294,967,295
  • long -2,147,483,648 2,147,483,647
  • ulong 0 4,294,967,295
  • longlong -9,223,372,036,854,775,808 9,223,372,036,854,775,807
  • ulonglong 0 18,446,744,073,709,551,615
  • float 1.17549e-38 3.40282e+38
  • double 2.22507e-308 1.79769e+308

  static const fp4816_t fix4816_one = { 65536 }; // 2^16 = 65536 = 0x10000000000000000
  static const float fp_rounding = 0.0005F;

  static const fp4816_t fp_maximum = { 0x7FFFFFFFFFFFFFFF }; // 9223372036854775807; float ???? static const fp4816_t fp_minumum = { 0xFFFFFFFFFFFFFFFF }; // -9223372036854775807 [note that sprintf(%lld) = -1]; aka float ??? static const fp4816_t fp_overflow = { 0x8000000000000000 }; // -9223372036854775808

  static const int64_t maxInteger = { 140737488355327 }; // { 0x7FFFFFFFFFFF } 140737488355327 = 140'737'488'355'327 static const int64_t minInteger = { -140737488355327 }; // -140737488355327; FFFFFFFFFFFF0000 static const fp4816_t fp_maxInteger = { 0x7FFFFFFFFFFF0000 }; // 9223372036854710272 = (longlong) 140737488355327 = 140'737'488'355'327 static const fp4816_t fp_minInteger = { 0xFFFFFFFFFFFF0000 }; // 9223372036854710272;

References

• +++ https://github.com/RomanZhu/FixedPoint-Sharp/blob/master/FixedPointSharp/fixnum/fp.cs

• https://en.wikipedia.org/wiki/Bitwise_operations_in_C

• https://docs.microsoft.com/en-us/dotnet/csharp/language-reference/operators/bitwise-and-shift-operators

• +++ https://github.com/RomanZhu/FixedPoint-Sharp

• +++ https://github.com/MikeLankamp/fpm

• +++ https://www.coranac.com/tonc/text/fixed.htm

• https://en.wikichip.org/w/images/d/db/Intrinsity_FixedPoint.pdf

• https://hal-lirmm.ccsd.cnrs.fr/lirmm-01277362/file/EJCIM2013.pdf

• ++ https://github.com/PetteriAimonen/libfixmath/tree/master/libfixmath https://github.com/PetteriAimonen/libfixmath/blob/master/libfixmath/fix16.h https://en.wikipedia.org/wiki/Libfixmath

• https://github.com/asik/FixedMath.Net/tree/master/src

• https://github.com/nxrighthere/RioSockets/blob/master/Source/riosockets.h

• https://www.allaboutcircuits.com/technical-articles/fixed-point-representation-the-q-format-and-addition-examples/

• Math and floating-point support https://docs.microsoft.com/en-us/cpp/c-runtime-library/floating-point-support?view=vs-2019

------------ TODOs ---------------------

• Consider using https://en.wikipedia.org/wiki/Rounding#Round_half_to_even