Re: [qemu-s390x] [Qemu-devel] [PATCH 2/3] qemu-error: make use of {error, warn}_report_once_cond

[Cornelia Huck]

On Wed, 29 Aug 2018 18:12:32 +0200
Markus Armbruster <address@hidden> wrote:

> Cornelia Huck <address@hidden> writes:
> 
> > {error,warn}_report_once() are a special case of the new functions
> > and can simply switch to them.
> >
> > Signed-off-by: Cornelia Huck <address@hidden>
> > ---
> >  include/qemu/error-report.h | 10 ++--------
> >  1 file changed, 2 insertions(+), 8 deletions(-)
> >
> > diff --git a/include/qemu/error-report.h b/include/qemu/error-report.h
> > index d2a6515e68..4e4c0e757c 100644
> > --- a/include/qemu/error-report.h
> > +++ b/include/qemu/error-report.h
> > @@ -58,10 +58,7 @@ void warn_report_once_cond(bool *printed, const char 
> > *fmt, ...)
> >          static bool print_once_;                \
> >          bool ret_print_once_ = !print_once_;    \
> >                                                  \
> > -        if (!print_once_) {                     \
> > -            print_once_ = true;                 \
> > -            error_report(fmt, ##__VA_ARGS__);   \
> > -        }                                       \
> > +        error_report_once_cond(&print_once_, fmt, ##__VA_ARGS__); \
> >          unlikely(ret_print_once_);              \
> >      })
> >  
> > @@ -74,10 +71,7 @@ void warn_report_once_cond(bool *printed, const char 
> > *fmt, ...)
> >          static bool print_once_;                \
> >          bool ret_print_once_ = !print_once_;    \
> >                                                  \
> > -        if (!print_once_) {                     \
> > -            print_once_ = true;                 \
> > -            warn_report(fmt, ##__VA_ARGS__);    \
> > -        }                                       \
> > +        warn_report_once_cond(&print_once_, fmt, ##__VA_ARGS__); \
> >          unlikely(ret_print_once_);              \
> >      })  
> 
> Hmm.  The macros return a value, while the functions do not.  Doesn't
> this contradict the commit message's claim the macros are ä special case
> of the new functions"?
> 
> If you make the functions return the value, too, the macros become
> simpler.  Moving complexity from macros to functions feels like a good
> deal, even when it's just a little bit of complexity like here.

Reducing macro complexity is a good idea, agreed.

Thanks for looking!