[libc-commits] [PATCH] D136799: [libc] Implement a high-precision floating point class.

[Tue Ly via Phabricator via libc-commits]

lntue added inline comments.


================
Comment at: libc/src/__support/FPUtil/dyadic_float.h:1
+//===-- A class to store high precision floating point numbers --*- C++ -*-===//
+//
----------------
sivachandra wrote:
> Will this class be ever used from outside of the `math` directory? If no, then it probably should live in `src/math`.
I plan to use this to replace `NormalFloat` and `XFloat` classes, along with their usages in `DivisionAndRemainderOperations` or `NearestIntegerOperations`.  Once the replacement completes, we will move them to `src/math` together.


================
Comment at: libc/src/__support/FPUtil/dyadic_float.h:34
+
+  template <typename T, cpp::enable_if_t<cpp::is_floating_point_v<T>, int> = 0>
+  DyadicFloat(T x) {
----------------
sivachandra wrote:
> This should like have additional constraints. For example, should this allow a quad-precision number to be converted to `DyadicFloat<64>`?
Added checks for mantissa lengths.


================
Comment at: libc/src/__support/FPUtil/dyadic_float.h:73
+
+  // Assume that it is already normalized and output is also normal.
+  // Output is rounded correctly with respect to the current rounding mode.
----------------
sivachandra wrote:
> Why is this assumption required?
Added explanation.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D136799/new/

https://reviews.llvm.org/D136799