On Mon, Oct 29, 2012 at 1:43 PM, Eli Friedman <[email protected]>wrote:

> On Mon, Oct 29, 2012 at 1:28 PM, Richard Trieu <[email protected]> wrote:
> > On Fri, Oct 26, 2012 at 5:37 PM, Eli Friedman <[email protected]>
> > wrote:
> >>
> >> On Fri, Oct 26, 2012 at 5:20 PM, Richard Trieu <[email protected]>
> wrote:
> >> > -Wbool-compare-tautological
> >> > Comparisons involving a boolean and an expression evaluating to 1, 0,
> >> > true,
> >> > or false such that the comparison always evaluate to true or false.
> >> > Also added to -Wtautological-compare group
> >> > High true positive rate.
> >> > Not previously caught by -Wtautological-compare
> >>
> >> Please just fix DiagnoseOutOfRangeComparison to do the right thing.
> >>
> >> > -Wbool-compare-tautological-out-of-range
> >> > Comparisons of a boolean and >1 or negative values.
> >> > Also added to -Wtautological-constant-out-of-range-compare group
> >> > High true positive rate.
> >> > Previously, only compares with values >1 were caught.
> >> > (x == 5)  currently caught
> >> > (x == -1)  not currently caught
> >>
> >> Same.
> >>
> >> -Eli
> >
> >
> > What do you think of the idea of separating out warnings on bool
> comparisons
> > into a sub group?
>
> I'm not exactly opposed... but why?  The tautological compare warning
> is on by default anyway.
>
I think that finer control over warnings is a good thing.  And that this
warning would fit in nicely if a new -Wbool-compare is created.

>
> (Either way, please fix DiagnoseOutOfRangeComparison instead of adding
> new code.)
>
Yes, I am looking at it now.

>
> -Eli
>
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

Reply via email to