boostorg / boostorg/multiprecision
Tie-to-even rule is not followed for overload remainder(float128, int)
- Dominant language
- C++
- Stars
- 265
- Forks
- 128
- Avg merge
- 4h 48m
- Merged PRs (30d)
- 2
Description
Running the following code
```c++
#include
#include
int main() {
using std::remainder;
boost::multiprecision::float128 x = 180, d = 360;
std::cout << remainder(x, 360) << " " << remainder(x, d) << "\n";
}
```
prints out `-180 180`. I expected both numbers to be the same and the tie-to-even rule for `remainder` dictates that `180` is the correct result.
I'm on a Fedora 40 machine with boost version 18.3.0 installed.
Contributor guide
No contributing guide indexed for this repository
Research direction
Reproduce the mismatch with the example that includes boost/multiprecision/float128.hpp, comparing remainder(x, 360) with remainder(x, d). Inspect the overloads involved in that header and add regression coverage showing that the tie-to-even result is consistent for the integer and float128 divisor cases.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- backend
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Stale
- Clarity
- Clearly specified
- Newbie friendliness
- 42/100