Subject: Re: Lockless file reading
From: Timo Sirainen <tss@iki.fi>
To: Martin Konold <martin.konold@erfrakon.de>
Cc: Jamie Lokier <jamie@shareable.org>, root@chaos.analogic.com,
       linux-kernel@vger.kernel.org
In-Reply-To: <200308281113.59112.martin.konold@erfrakon.de>
References: <3217CEE6-D906-11D7-A165-000393CC2E90@iki.fi>
	 <200308281113.59112.martin.konold@erfrakon.de>
Content-Type: text/plain
Message-Id: <1062062858.1454.254.camel@hurina>
Mime-Version: 1.0
Date: Thu, 28 Aug 2003 12:27:38 +0300
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 861
Lines: 20

On Thu, 2003-08-28 at 12:13, Martin Konold wrote:
> > How about checksum[n] = data[n-1] ^ data[n]? That looks like it would
> 
> I propose you first make some benchmarks and try to figure out how big the 
> actual overhead of locking really is. I can easily assume that your 
> "solution" is actually slower than a simple mechanism/semaphore. 

It's not about CPU usage. It's mostly about being able to modify the
file even when there's thousands of simultaneous readers that could
otherwise keep the file locked almost constantly.

Also it'd be nice to support NFS with .lock files since no-one really
uses lockd.


-
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/