[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Pine.LNX.4.64.0704011029230.26331@alien.or.mcafeemobile.com>
Date: Sun, 1 Apr 2007 10:31:29 -0700 (PDT)
From: Davide Libenzi <davidel@...ilserver.org>
To: Linus Torvalds <torvalds@...ux-foundation.org>
cc: Avi Kivity <avi@...o.co.il>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Andrew Morton <akpm@...ux-foundation.org>,
Ingo Molnar <mingo@...e.hu>,
Suparna Bhattacharya <suparna@...ibm.com>,
Zach Brown <zach.brown@...cle.com>,
Benjamin LaHaise <bcrl@...ck.org>
Subject: Re: [patch 13/13] signal/timer/event fds v9 - KAIO eventfd support
example ...
On Sun, 1 Apr 2007, Linus Torvalds wrote:
> On Sun, 1 Apr 2007, Davide Libenzi wrote:
> >
> > > [also, minor nit: sys_eventfd() can legitimately return 0, but you're banning
> > > its use in aio. userspace could easily dup() and close(), but...]
> >
> > Noone said that binary compatibility comes free ;)
> > I'm truly open to other ways of integration, but I thought that binary
> > compatibility is the key here.
>
> I would suggest taking "aio_reserved3" and turning it into
>
> __u32 aio_flags;
> __u32 aio_resfd;
>
> if so. And then some bit in the "flags" field has to be set (say,
> "IOCB_FLAG_RESFB") to enable this codepath, instead of expecting that
> aio_resfd == 0 is special.
>
> It's generally always a good thing to have a "flags" field anyway, exactly
> because it allows for future expansion.
Makes sense. I was actually thinking of using the higher bit to signal
that, but a flags field makes more sense.
- Davide
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists