[Bug c/43128] [4.5 Regression] c-c++-common/pr41779.c doesn't work

["joseph at codesourcery dot com" <gcc-bugzilla@gcc.gnu.org>]

------- Comment #17 from joseph at codesourcery dot com  2010-02-21 18:32 -------
Subject: Re:  [4.5 Regression] c-c++-common/pr41779.c doesn't
 work

On Sun, 21 Feb 2010, manu at gcc dot gnu dot org wrote:

> ------- Comment #16 from manu at gcc dot gnu dot org  2010-02-21 18:25 -------
> (In reply to comment #15)
> > 
> > With the intended semantics, we should warn; there would be an actual 
> > conversion from integer to float there, that could change the value.
> 
> Great. Any hints where could be the problem or fix?

Where convert_and_check is called, for (a) conditional expressions and (b) 
binary operations, to convert two operands to a common type, the 
conversion is (correctly) to the type with excess precision (possibly long 
double where the semantic type is narrower, float or double).  I would 
suggest having a c_ep_convert_and_check or similar function that handles 
excess precision: it would take the result type, the semantic result type 
(the type that gets used eventually to build an EXCESS_PRECISION_EXPR) and 
the value to convert.  It would just call convert_and_check, ignoring the 
semantic type, *except* when the operand has integer type *and* the 
semantic type is non-NULL; in that case, it would first convert to the 
semantic type them to the result type.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43128