> On Nov. 10, 2012, 10:21 p.m., Rolf Eike Beer wrote:
> > A simple way to verify if this is correct is check if you and the compiler 
> > agree on the code. Run "make -n" in the build tree to get the full compiler 
> > command line, then insert a "-S" (assuming you are using gcc) and change 
> > the -o to point to a temporary file. This will output the assembler code. 
> > Do this with and without your modifications and look if the result is still 
> > the same.

yes, I've done it. They match.

bool kk1()
{
 int a = 1;
 int b = 2;
 return (!a ^ b);
}

bool kk2()
{
 int a = 1;
 int b = 2;
 return (!(a ^ b));
}

bool kk3()
{
 int a = 1;
 int b = 2;
 return (!(a) ^ b);
}

kk1 and kk3 produce the same assembler output.

But the problem is if the author really wanted the parenthesis that way or not.
for example, (!a) ^ b  (what gcc produces) vs. !(a ^ b) (what the author 
wanted?)


- Jaime Torres


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://git.reviewboard.kde.org/r/107213/#review21785
-----------------------------------------------------------


On Nov. 5, 2012, 3:13 p.m., Jaime Torres Amate wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://git.reviewboard.kde.org/r/107213/
> -----------------------------------------------------------
> 
> (Updated Nov. 5, 2012, 3:13 p.m.)
> 
> 
> Review request for kdelibs.
> 
> 
> Description
> -------
> 
> place some parenthesis around ! operators, with less priority than ^ 
> operators.
> place some parenthesis around = inside conditions 
> check for n not being null before using it
> simplify if (a==true) return true else return false;
> 
> 
> Diffs
> -----
> 
>   khtml/khtml_caret.cpp 45fd90c 
>   khtml/rendering/render_inline.cpp acfc1e4 
>   khtml/rendering/render_object.cpp f37627c 
>   solid/solid/backends/wmi/wmiopticaldisc.cpp c6e390f 
> 
> Diff: http://git.reviewboard.kde.org/r/107213/diff/
> 
> 
> Testing
> -------
> 
> I've been using this code several weeks.
> 
> 
> Thanks,
> 
> Jaime Torres Amate
> 
>

Reply via email to