[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <alpine.LFD.1.10.0805302116440.3141@woody.linux-foundation.org>
Date: Fri, 30 May 2008 21:23:31 -0700 (PDT)
From: Linus Torvalds <torvalds@...ux-foundation.org>
To: Ulrich Drepper <drepper@...hat.com>
cc: linux-kernel@...r.kernel.org, akpm@...ux-foundation.org,
mtk.manpages@...il.com
Subject: Re: [PATCH 0/3] 64-bit futexes: Intro
On Fri, 30 May 2008, Linus Torvalds wrote:
>
> Also note how you can use 64-bit atomic ops to do that all in user space,
> without actually requiring 64-bit futex support - as long as the bits that
> matter for waking (like "was there more than one pending writer") fit in
> the one 32-bit word.
So here's an example of this:
- make all the readers/writers actually update a 64-bit word (by using
cmpxchg8b in user space to actually get the locks)
- but organize things so that a reader only needs to look at the high 32
bits to actually make it's wakeup-decision, and a sleeping writer only
needs to look at the low 32 bits. How? Make the low bits of the words
contain the "contention status".
Is it possible? I dunno. I personally suspect it is. I also suspect you
didn't even try.
Linus
--
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