Subject: Re: [PATCH] Re: Reduce XFS footprint (was Re: TAKE - remove a
	function xfs added to filemap.c
From: Eric Sandeen <sandeen@sgi.com>
To: utz lehmann <xfs@s2y4n2c.de>
Cc: Andi Kleen <ak@suse.de>, linux-xfs@oss.sgi.com
In-Reply-To: <20020129205653.A13502@s2y4n2c.de>
References: <200201291751.g0THp897004750@scare.vieo.com>
	<20020129194001.A16401@wotan.suse.de> <20020129202509.A31370@wotan.suse.de>
	 <20020129205653.A13502@s2y4n2c.de>
Content-Type: text/plain
Content-Transfer-Encoding: 7bit
Date: 29 Jan 2002 14:09:53 -0600
Message-Id: <1012334993.5905.30.camel@stout.americas.sgi.com>
Mime-Version: 1.0
Sender: owner-linux-xfs@oss.sgi.com
Precedence: bulk
Status: O
Content-Length: 675
Lines: 24

On Tue, 2002-01-29 at 13:56, utz lehmann wrote:
> Hi
> 
> I have a few questions.
> 
> Is xfs using less 64bit arithmetics when compiled without
> XFS_BIG_FILESYSTEMS and/or XFS_BIG_FILES? I think this will save some cpu
> cyles on x86.

I have this same question, I'll do some benchmarking to find out.
 
> What happends with old files larger than 2GB, when a kernel without
> XFS_BIG_FILES is used?

sb->s_maxbytes gets set to a smaller value, so all the normal kernel
size checks kick in, and you won't be able to seek/truncate/read/write
past the smaller value.

-Eric

-- 
Eric Sandeen      XFS for Linux     http://oss.sgi.com/projects/xfs
sandeen@sgi.com   SGI, Inc.