On Mon, 2005-08-08 at 05:00 -0400, Rahul Iyer wrote:
> Bernd Petrovitsch wrote:
>
> >On Sun, 2005-08-07 at 15:57 +0200, Thomas Petazzoni wrote:
> >
> >
> >>Hi,
> >>
> >>Bob Smith a écrit :
> >>
> >>
> >>
> >>>Why is that ? I googled up reasons like minimizing return points,
> >>>but is that it ?
> >>>
> >>>
> >>Most of the time, 'goto' is used to minimize code duplication in error
> >>handling.
> >>
> >>Without 'goto':
> >>
> >>int myfunc (void)
> >>{
> >> ret = dosomething1();
> >> if (ret < 0)
> >> return ret;
> >>
> >> ret = dosomething2();
> >> if (ret < 0)
> >> {
> >> undosomething1();
> >> return ret;
> >> }
> >>
> >> ret = dosomething3();
> >> if (ret < 0)
> >> {
> >> undosomething2();
> >> undosomething1();
> >> return ret;
> >> }
> >>
> >> ret = dosomething4();
> >> if (ret < 0)
> >> {
> >> undosomething3();
> >> undosomething2();
> >> undosomething1();
> >> return ret;
> >> }
> >> return 0;
> >>}
> >>
> >>With 'goto':
> >>
> >>int myfunc(void)
> >>{
> >> ret = dosomething1()
> >> if (ret < 0)
> >> goto ret_dosomething1;
> >>
> >> ret = dosomething2()
> >> if (ret < 0)
> >> goto ret_dosomething2;
> >>
> >> ret = dosomething3()
> >> if (ret < 0)
> >> goto ret_dosomething3;
> >>
> >> ret = dosomething4()
> >> if (ret < 0)
> >> goto ret_dosomething4;
> >>
> >>ret_dosomething4:
> >> undosomething3();
> >>ret_dosomething3:
> >> undosomething2();
> >>ret_dosomething2:
> >> undosomething1();
> >>ret_dosomething1:
> >> return ret;
> >>}
> >>
> >>
> >
> >With 'if':
> >int myfunc(void)
> >{
> > ret = dosomething1()
> > if (ret >= 0) {
> > ret = dosomething2()
> > if (ret >= 0) {
> > ret = dosomething3()
> > if (ret >= 0) {
> > ret = dosomething4()
> > if (ret >= 0) {
> > ...
> > }
> > }
> > undosomething3();
> > }
> > undosomething2();
> > }
> > undosomething1();
> > return ret;
> >}
> >
> >Now remember that the kernel has 8 spaces indentation, not 3 as above.
> >Not that better, especially if your label names are quite descriptive.
[...]
> IMHO, there is a small problem with the code..
> what if ret is > 0? Then the code within the if is executed, only to be
> undone after the execution of the if.
>
> for example...
>
> ret = dosomething1()
> if (ret > 0) {
> /*do stuff */
> .......
> }
> undosomething1()
> return ret;
>
> if ret > 0, it will "do stuff" and then undo something1. That IIUC, is
> not the desired effect.
> Am i right, or is this me misinterpreting the code?
You are right. But I actually rewrote the example above which has teh
same behaviour.
And BTW I didn't assume any semantics so I don't want to say that the
above pattern is always correct or always wrong with respect to that
detail.
Bernd
--
Firmix Software GmbH http://www.firmix.at/
mobil: +43 664 4416156 fax: +43 1 7890849-55
Embedded Linux Development and Services
--
Kernelnewbies: Help each other learn about the Linux kernel.
Archive: http://mail.nl.linux.org/kernelnewbies/
FAQ: http://kernelnewbies.org/faq/
