Repository navigation

Implement the transcendental functions for the IEEE 754 decimal types - #131019

Merged

tannergooding merged 35 commits into

dotnet:mainfrom

tannergooding:tannergooding-decimal-transcendentals-binary-fp

Jul 20, 2026

Merged

Implement the transcendental functions for the IEEE 754 decimal types#131019
tannergooding merged 35 commits into
dotnet:mainfrom
tannergooding:tannergooding-decimal-transcendentals-binary-fp

Commits on Jul 18, 2026
- Add the Float128 software engine core for decimal transcendentals
  
  tannergooding
  and
  Copilot
  committed
  7f8b827
- Add the Float128 divide primitive and BID decimal conversions
  
  tannergooding
  and
  Copilot
  committed
  09d3188
- Add the exponential operation for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  844f8ff
- Add the base-2 and base-10 exponential operations for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  d439334
- Add the logarithm operations for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  f1e89d7
- Add the power operation for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  e856546
- Add the cube root operation for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  ca73a1f
- Add the hypotenuse operation for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  b248ca1
- Add the nth root operation for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  967e137
- Add the trigonometric operations for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  a5d6abb
- Add the inverse trigonometric operations for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  a100706
- Add the pi-based trigonometric operations for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  a47cbab
- Add the hyperbolic operations for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  7c75d57
- Add the inverse hyperbolic operations for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  0897926
- Declare the IFloatingPointIeee754 interface for the decimal types
  
  tannergooding
  and
  Copilot
  committed
  2576135
- Add the IDecimalFloatingPointIeee754 interface for the decimal types
  
  tannergooding
  and
  Copilot
  committed
  c938d3f
- Rename the internal Float128 working type to DiyFp128
  
  tannergooding
  and
  Copilot
  committed
  0601cd2
- Use ReadOnlySpan and decimal literals for the primitive coefficient tables
  
  tannergooding
  and
  Copilot
  committed
  b1be51e
- Clean up the internal helpers for the decimal transcendentals
  
  tannergooding
  and
  Copilot
  committed
  15f6639
- Use uppercase hex digits in the decimal transcendental tables
  
  tannergooding
  and
  Copilot
  committed
  a79eb79
- Route Decimal32 transcendentals through the binary128 engine
  
  tannergooding
  and
  Copilot
  committed
  15369f1
- Correct stale Decimal32 engine comments and strip the interface file BOM
  
  tannergooding
  and
  Copilot
  committed
  4244f53
- Fix RootN int.MinValue overflow and use the documented engine sign bit
  
  tannergooding
  and
  Copilot
  committed
  531f3b0
- Rename the decimal generic-surface test to match the interface it verifies
  
  tannergooding
  and
  Copilot
  committed
  95e3605
Commits on Jul 19, 2026
- Fix negative small-argument and half-integer sign errors in the decimal trig core
  
  tannergooding
  and
  Copilot
  committed
  434bdc9
- Reduce large trig arguments in the decimal domain
  
  tannergooding
  and
  Copilot
  committed
  ef88070
- Add the Intel transcendental reference-vector tests for the IEEE 754 decimal types
  
  tannergooding
  and
  Copilot
  committed
  3b14dbf
- Expose the UInt128 powers-of-ten table as a read-only span
  
  tannergooding
  and
  Copilot
  committed
  dd94989
- Bound the decimal trig reduction scratch buffer to the stack budget
  
  tannergooding
  and
  Copilot
  committed
  5ad4a1c
- Rename the decimal radian-reduction word buffer to integerWords
  
  tannergooding
  and
  Copilot
  committed
  ccfd4e0
- Strengthen the decimal pi-scaled trig accuracy tests
  
  tannergooding
  and
  Copilot
  committed
  ceac9bb
- Ensure that acosh of a zero decimal operand returns NaN
  
  tannergooding
  and
  Copilot
  committed
  aec8901
- Reject opposite-sign and wildly-off results in the decimal ULP comparator
  
  tannergooding
  and
  Copilot
  committed
  d9ccd9d
- Merge remote-tracking branch 'dotnet/main' into tannergooding-decimal-transcendentals-binary-fp
  tannergooding
  committed
  a337b38
- Ensure that the binary-to-decimal fast path preserves the preferred exponent
  
  tannergooding
  and
  Copilot
  committed
  6e3fabf

Commits

Commits on Jul 18, 2026

Commits on Jul 19, 2026