[cfp-interest 3506] preferred quantum exponent

[Vincent Lefevre]

In C23 5.2.5.3.4p10 and the latest C2y draft (N3550) 5.3.5.3.4p10,
"how the preferred quantum exponents of the operands, Q(x), Q(y),
etc." does not make sense because the operands already have a
quantum exponent (there is nothing preferred for the operands).
The notion of *preferred* quantum exponent is just used for the
choice of the quantum exponent of a result. So this should be
"how the quantum exponents of the operands, Q(x), Q(y), etc."
(i.e. remove "preferred").

Moreover, the notion of preferred quantum exponent is used only
for exact results. In IEEE 754, this is perfectly specified thanks
to correct rounding. But C23/C2y does not require correct rounding
for all operations (F.3p20: "However, correct rounding, which
ISO/IEC 60559 specifies for its operations, is not required for
the C functions in the table."). So this is ambiguous for such
functions because an implementation does not need to determine
whether a result is exact or not (possibly except in special cases
given in Annex F, but the exactness is not explicit).

-- 
Vincent Lefèvre <vincent at vinc17.net> - Web: <https://www.vinc17.net/>
100% accessible validated (X)HTML - Blog: <https://www.vinc17.net/blog/>
Work: CR INRIA - computer arithmetic / Pascaline project (LIP, ENS-Lyon)