Re: NetBSD/amd64 current with HAVE_GCC=48

[Ryota Ozaki <ozaki-r%NetBSD.org@localhost>]

On Fri, Mar 7, 2014 at 10:17 AM, Matt Thomas <matt%3am-software.com@localhost> 
wrote:
>
> On Mar 6, 2014, at 4:35 PM, Ryota Ozaki <ozaki-r%netbsd.org@localhost> wrote:
>
>> The below patch should fix the error, but there may be another
>> better solution.
>>
>> Anyone any ideas?
>>
>>  ozaki-r
>>
>> Index: external/cddl/osnet/dist/uts/common/dtrace/dtrace.c
>> ===================================================================
>> RCS file: 
>> /cvs/cvsroot/src/external/cddl/osnet/dist/uts/common/dtrace/dtrace.c,v
>> retrieving revision 1.25
>> diff -u -r1.25 dtrace.c
>> --- external/cddl/osnet/dist/uts/common/dtrace/dtrace.c 5 Mar 2014
>> 20:14:15 -0000       1.25
>> +++ external/cddl/osnet/dist/uts/common/dtrace/dtrace.c 7 Mar 2014
>> 00:28:47 -0000
>> @@ -16682,6 +16682,7 @@
>>        cv_signal(&w->cv);
>>        mutex_exit(&w->lock);
>>        error = kthread_join(w->lwp);
>> +       (void) error; /* XXX gcc 4.8 */
>>        KASSERT(error == 0);
>>        cv_destroy(&w->cv);
>>        mutex_destroy(&w->lock);
>
> RCS file: /cvsroot/src/external/cddl/osnet/dist/uts/common/dtrace/dtrace.c,v
> retrieving revision 1.25
> diff -u -p -r1.25 dtrace.c
> --- external/cddl/osnet/dist/uts/common/dtrace/dtrace.c 5 Mar 2014 20:14:15 
> -0000       1.25
> +++ external/cddl/osnet/dist/uts/common/dtrace/dtrace.c 7 Mar 2014 01:17:11 
> -0000
> @@ -16674,14 +16674,13 @@ dtrace_state_worker_add(void (*fn)(dtrac
>  void
>  dtrace_state_worker_remove(struct dtrace_state_worker *w)
>  {
> -       int error;
>
>         KASSERT(!w->exiting);
>         mutex_enter(&w->lock);
>         w->exiting = true;
>         cv_signal(&w->cv);
>         mutex_exit(&w->lock);
> -       error = kthread_join(w->lwp);
> +       int error __diagused = kthread_join(w->lwp);
>         KASSERT(error == 0);
>         cv_destroy(&w->cv);
>         mutex_destroy(&w->lock);
>

Oh, I didn't know the attribute. I should read attribute(3)...

BTW I think we don't need to move the definition of error.

Thanks,
  ozaki-r