Re: svn commit: r510006 - /apr/apr/trunk/test/testpoll.c

[Ryan Phillips <ry...@trolocsis.com>]

The following #define returns in every instance even if pollcb is
different than NULL.  Was this meant to be this way?

I have included a small patch...

-Ryan

jorton@apache.org said:
> Author: jorton
> Date: Wed Feb 21 05:35:43 2007
> New Revision: 510006
> 
> URL: http://svn.apache.org/viewvc?view=rev&rev=510006
> Log:
> * test/testpoll.c (setup_pollcb, trigger_pollcb, timeout_pollcb,
> timeout_pollin_pollcb): Skip tests if pollcb interface is ENOTIMPL.
> 
> Modified:
>     apr/apr/trunk/test/testpoll.c
> 
> Modified: apr/apr/trunk/test/testpoll.c
> URL: http://svn.apache.org/viewvc/apr/apr/trunk/test/testpoll.c?view=diff&rev=510006&r1=510005&r2=510006
> ==============================================================================
> --- apr/apr/trunk/test/testpoll.c (original)
> +++ apr/apr/trunk/test/testpoll.c Wed Feb 21 05:35:43 2007
> @@ -553,11 +553,20 @@
>               (hot_files[1].client_data == (void *)1)));
>  }
>  
> +#define POLLCB_PREREQ do { if (pollcb == NULL) \
> +ABTS_NOT_IMPL(tc, "pollcb interface not supported"); return; } while (0)
> +

Re: svn commit: r510006 - /apr/apr/trunk/test/testpoll.c

[Joe Orton <jo...@redhat.com>]

On Wed, Feb 21, 2007 at 06:38:52PM -0800, Ryan Phillips wrote:
> The following #define returns in every instance even if pollcb is
> different than NULL.  Was this meant to be this way?

Oops, no.

> I have included a small patch...

Thanks Ryan; tweaked to still return in the NULL case, and applied
:)

joe