On Thu, Feb 20, 2014 at 5:54 PM, Joel Sherrill <joel.sherr...@oarcorp.com> wrote: > > On 2/20/2014 4:45 PM, Peter Dufault wrote: >> On Feb 20, 2014, at 16:11 , Gedare Bloom <ged...@rtems.org> wrote: >> >>> +#if __RTEMS_HAVE_SYS_CPUSET_H__ >>> I think we decided to prefer the more explicit "#if defined(...)" for >>> checking for CPP defines. >>> Also, what is this define checking against / where is it defined? >> I think "#if defined()" isn't more explicit unless the test is "#if >> defined(__RTEMS_HAVE_SYS_CPUSET_H_DEFINED__)" because of the possibility of: >> >> #define __RTEMS_HAVE_SYS_CPUSET_H__ 0 >> >> I just hit this the other day. I personally don't like #ifdef. >> >> "#if defined(__RTEMS_HAVE_SYS_CPUSET_H__) && (__RTEMS_HAVE_SYS_CPUSET_H__ != >> 0)" addresses all possibilities and is kind of like wrapping macros in "do >> .. while (0)" guards, but it's verbose and odd. > I have nothing against paranoia but this is unfortunately a pattern > encouraged > by autoconf feature macros. If we can come to some agreement on the pattern, > then it should be possible to write a script to convert from old to new > pattern. > Peter is right. I looked back at the history I was thinking about, and we rely on the #if for value checking e.g. for BSPOPTS. While I prefer to use ifdef for feature flags, the historical use of 0/1 for the flag makes the ifdef a bug. We should drop this issue for now.
> But arbitrarily using a different pattern in new code isn't good either. > I would > rather see a massive sed script to fix them all in one sweep. >> Peter >> ----------------- >> Peter Dufault >> HD Associates, Inc. Software and System Engineering >> >> >> _______________________________________________ >> rtems-devel mailing list >> rtems-devel@rtems.org >> http://www.rtems.org/mailman/listinfo/rtems-devel > > -- > Joel Sherrill, Ph.D. Director of Research & Development > joel.sherr...@oarcorp.com On-Line Applications Research > Ask me about RTEMS: a free RTOS Huntsville AL 35805 > Support Available (256) 722-9985 > _______________________________________________ rtems-devel mailing list rtems-devel@rtems.org http://www.rtems.org/mailman/listinfo/rtems-devel
