https://bugs.documentfoundation.org/show_bug.cgi?id=152943

Mike Kaganski <[email protected]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[email protected]

--- Comment #1 from Mike Kaganski <[email protected]> ---
https://grouper.ieee.org/groups/msc/ANSI_IEEE-Std-754-2019/background/nan-propagation.pdf

> The standard specifies that an operation with a quiet NaN input should 
> produce a
> quiet NaN result with the same payload.

This means that for a "0 + nan", executed in KahanSum::get -> KahanSum::add,
the result *must* be identical to the nan payload.

> Support for payload propagation is optional on RISC-V processors.

So likely, we just need to enable the propagation to force conformance to the
standard?

-- 
You are receiving this mail because:
You are the assignee for the bug.

Reply via email to