[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Stds-754] motion for Aug: Relaxation of Invalid Operation for Convert Float to Fixed



Dan, I'd like to make the following motion at the August meeting.

Regards, Ron Smith

(The page numbers and text used in this motion are from
DRAFT Standard for Floating-Point Arithmetic P754 Draft 1.1.2,
dated July 28, 2006)
--------------------------------------------------
Motion: Relaxation of Invalid Operation for Convert Float to Fixed

Page 46, Section 9.2, "Invalid operation" items (h) and (l)

Current text:

h) conversion of an internal floating-point number to an integer or
external representation as a character sequence when overflow, infinity,
or NaN precludes a correctly-rounded representation in the destination
and this cannot otherwise be indicated;

l) conversion of an internal floating-point number, which would convert
to an integer less than zero under the prevailing rounding mode, to an
unsigned integer format and this cannot otherwise be indicated.

Change To:

h) conversion of an internal floating-point number to an external
representation as a character sequence when overflow, infinity, or NaN
precludes a correctly-rounded representation in the destination and this
cannot otherwise be indicated;

l) conversion of an internal floating-point number to an integer format,
when the source is NaN, infinity, or a value (which, under the
prevailing rounding mode) would convert to an integer outside the range
of the result format, it is language defined what results are delivered
and whether invalid operation is signaled.

Rationale

We see no reason to change the current definition in item (h) as it
applies to conversion to an external representation as a character
sequence.

For conversion to integer format, the term "overflow" should not be used
to describe the out-of-range condition as this is not consistent with
the use of the term "overflow" as used in the rest of the document.

Our interpretation of the term "otherwise indicated" as used in 754 and
in 754r is that for integer results "garbage in garbage out" is
acceptable as the only indication.  Thus, this proposal is not a change,
but only a clarification.

754 | revision | FAQ | references | list archive