Re: [Patch] tree-parloops.c (eliminate_local_variables): Add braces to suppress warnings

[Magnus Fromreide <magfr at lysator dot liu dot se>]

On Fri, 2012-04-13 at 16:50 +0200, Bernd Schmidt wrote:
> On 04/13/2012 04:44 PM, Jakub Jelinek wrote:
> > On Fri, Apr 13, 2012 at 04:33:17PM +0200, Bernd Schmidt wrote:
> >> On 04/13/2012 04:20 PM, Mike Stump wrote:
> >>> On Apr 13, 2012, at 5:30 AM, NightStrike wrote:
> >>>>> no warning from trunk.  Which GCC version emits this warning?
> >>>
> >>>> Looks like cygwin gcc 3.4.4
> >>>
> >>> 3.4.4 is a little old now..  We'd encourage an upgrade to a fine new compiler...  :-)
> >>
> >> The thing is, the else really is ambiguous, so it looks like we have a
> >> warning regression.
> > 
> > To the compiler?  There is for in between, no need to put extra braces IMHO.
> > I believe it has been intentionally fixed for 4.0, but my bisect tree only
> > goes back to r90000.
> 
> Well, to the compiler even
> if ()
>   if ()
>     x
>   else
>     y
> 
> has an unambiguous meaning. The problem is that a human might be
> confused, for example due to bad indentation.

For me this warning is almost always a false positive since I am unable
to turn it of selectively.

I am programming in C++.
My main use of

if () ; else y

is to get a temporary scope that doesn't have any braces, e.g. in this
macro

#define DEBUG if (!enable_debugging) ; else debug_object()

where debug_object is an instance of std::ostream.
I then use it like

DEBUG << "fancy message goes here";

Now, writing

if (condition)
  DEBUG << "ooops, condition is true";

is not that far-fetched and obviously triggers the warning but I have a
hard time seeing how that is confusing to a casual reader of the code.


(In order to avoid this warning I am currently writing the above code
as

#define DEBUG switch (!enable_debugging) case false: debug_object()

but that makes the define harder to understand and so the net result of
the warning is to reduce the clarity of the code)

/MF