Re: [PATCH 1/2] wait: introduce wait_event_common(wq, condition, state, timeout)

[Andrew Morton <akpm@linux-foundation.org>]

On Thu, 6 Jun 2013 22:03:16 +0200 Oleg Nesterov <oleg@redhat.com> wrote:

> 1. wait_event_timeout(wq, true, 0) returns zero, I think this is
>    wrong and should be fixed.
> 
>    __wait_event_timeout() was already changed by 4c663cfc but we
>    need the same logic in wait_event_timeout() if the fast-path
>    check succeeds.
> 
> 2. wait_event_timeout/__wait_event_timeout interface do not match
>    wait_event(), you can't use __wait_event_timeout() instead of
>    wait_event_timeout() if you do not need the fast-path check.
> 
> 3. wait_event_* macros duplicate the same code.
> 
> This patch adds a single helper wait_event_common() which hopefully
> does everything right. Compiler optimizes out the "dead" code when
> we do not need signal_pending/schedule_timeout.
> 
> With this patch "size vmlinux" reports that .text/data shrinks but
> I think this depends on gcc/config.

hm,

> -#define __wait_event_interruptible(wq, condition, ret)
> +#define __wait_event_interruptible(wq, condition)

net/irda/af_irda.c:2568:13: error: macro "__wait_event_interruptible" passed 3 arguments, but takes just 2

waddup with that?

__wait_event_interruptible() has several callsites.  I think I'll go
zap and await v2 ;)