Re: [rfc git pull] cpus4096 fixes, take 2

[Ingo Molnar <mingo@elte.hu>]

* Rusty Russell <rusty@rustcorp.com.au> wrote:

> On Tuesday 29 July 2008 06:57:00 Ingo Molnar wrote:
> > +/*
> > + * In cases where we take the address of the cpumask immediately,
> > + * gcc optimizes it out (it's a constant) and there's no huge stack
> > + * variable created:
> > + */
> > +#define cpumask_of_cpu(cpu) ({ *get_cpu_mask(cpu); })
> 
> Why use a statement expression here?  Isn't (*get_cpu_mask(cpu)) 
> sufficient?

Yeah, it's sufficient - no strong reasons - it felt a slightly bit more 
correct to isolate the read-only data structure. Not that it makes any 
real difference in practice - it's still possible to take the address of 
get_cpu_mask() and abuse that - gcc will only issue a warning.

But gcc 3.4.5 apparently craps out on this valid use of a gcc extension, 
see the report and the fix at:

  http://lkml.org/lkml/2008/7/29/154

	Ingo