On Thu, Mar 10, 2016 at 3:40 PM, Carsten Haitzler <ras...@rasterman.com> wrote: > On Thu, 10 Mar 2016 17:43:01 +0000 Tom Hacohen <t...@osg.samsung.com> said: >> On 10/03/16 17:02, Cedric BAIL wrote: >> > On Mar 10, 2016 8:10 AM, "Tom Hacohen" <t...@osg.samsung.com> wrote: >> >> Huh? It is valid and the code is correct. What seems to be the problem? >> >> Works here with no warnings (clang), but anyhow, the code is correct. >> > >> >>From gcc point of view gl was undefined when set to sd->genlist and that >> > did come with a big fat warning :-) >> > >> >> GCC is wrong, but I guess you had to silence it. :| > > gcc generates warnings because it sees this as a possible code bug - you pass > in a reference and fill it then also re-assign the value again as a return > from > the same func. in any sane world most people reading that code would go "wtf? > is that a bug? did they may to pass in &somethingelse instead of &gl because > this assigns it twice? are we losing a reference to something here?". gcc is > saying that. the code itself wasn't broken. it would work, BUT it LOOKS like a > bug to any "sane person". and gcc warns. the more warnings we let through, the > more real issues hide in warnings we now ignore in the sea of warnings > scrolling by. thus why it is being "shut up". > > but it is a point - it looks insane. imho eo_add given its current use of & > should not return anything and we assign return by reference. at least then it > doesn't look like a bug.
Agreed. It is surely very confusing to have something like obj = eo_add(&obj, ... -- Cedric BAIL ------------------------------------------------------------------------------ Transform Data into Opportunity. Accelerate data analysis in your applications with Intel Data Analytics Acceleration Library. Click to learn more. http://pubads.g.doubleclick.net/gampad/clk?id=278785111&iu=/4140 _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
