Subject: Re: MAP_PRIVATE that stays private even on external write
From: Peter Zijlstra <peterz@infradead.org>
To: Miquel van Smoorenburg <miquels@cistron.nl>
Cc: linux-kernel@vger.kernel.org, Chris Mason <chris.mason@oracle.com>
In-Reply-To: <493508b5$0$200$e4fe514c@news.xs4all.nl>
References: <493508b5$0$200$e4fe514c@news.xs4all.nl>
Content-Type: text/plain
Content-Transfer-Encoding: 7bit
Date: Tue, 02 Dec 2008 11:27:31 +0100
Message-Id: <1228213651.7133.12.camel@twins>
Mime-Version: 1.0
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1449
Lines: 33

On Tue, 2008-12-02 at 10:06 +0000, Miquel van Smoorenburg wrote:
> I have an application that mmaps blocks from a file and then
> sends it out over a TCP socket.
> 
> The contents of the file may be overwritten while the blocks
> are sent out, but there is a very low chance of that.
> 
> To prevent this, I can ofcourse just use read() all blocks into
> memory in advance, check consistency, and go ahead .. but since
> there may be a lot of network connections active this might cost
> a lot of memory.
> 
> What I am looking for is a MAP_PRIVATE type flag that, when another
> process modifies pages of the file (through mmap() or write())
> makes sure that my mapping never sees that.
> 
> I found out that mmap(MAP_PRIVATE|MAP_POPULATE) does exactly what
> I want, but ofcourse it reads in the entire mapping all at once,
> which is equivalent to plain read().
> 
> There has been talk of a MAP_SNAPSHOT flag before, e.g.
> http://lkml.indiana.edu/hypermail/linux/kernel/0407.1/0416.htm
> 
> Has anyone ever looked at implementing something like this ?

I suppose that needs a snapshot filesystem for backing, and we don't
have such a creature (yet). I suppose BTRFS might be able to pull that
off.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/