Re: [PATCH v3] cleanup: adjust scoped_guard() macros to avoid potential warning

[Peter Zijlstra]

On Wed, Oct 23, 2024 at 03:43:22PM +0200, Przemek Kitszel wrote:
> On 10/18/24 12:50, Peter Zijlstra wrote:
> > On Sun, Oct 13, 2024 at 12:01:24PM +0800, kernel test robot wrote:

> > --- a/include/linux/cleanup.h
> > +++ b/include/linux/cleanup.h
> > @@ -323,7 +323,7 @@ static __maybe_unused const bool class_#
> >    */
> >   #define __scoped_guard(_name, _fail, _label, args...)				\
> >   	for (CLASS(_name, scope)(args);	true; ({ goto _label; }))		\
> > -		if (!__guard_ptr(_name)(&scope) && __is_cond_ptr(_name)) {	\
> > +		if (__is_cond_ptr(_name) && !__guard_ptr(_name)(&scope)) {	\
> 
> but this will purge the attempt to call __guard_ptr(), and thus newer
> lock ;) good that there is at least some comment above

No, __guard_ptr() will only return a pointer, it has no action. The lock
callback is in CLASS(_name, scope)(args).