xfs
[Top] [All Lists]

Re: Linux page cache doubt

To: Steve Lord <lord@xxxxxxx>
Subject: Re: Linux page cache doubt
From: Andi Kleen <ak@xxxxxxx>
Date: Thu, 25 Oct 2001 01:19:51 +0200
Cc: Arun Ramakrishnan <ramakria@xxxxxxxxxxxxxxxxxx>, linux-xfs@xxxxxxxxxxx
In-reply-to: <200110242236.f9OMagj06647@xxxxxxxxxxxxxxxxxxxx>; from lord@xxxxxxx on Wed, Oct 24, 2001 at 05:36:41PM -0500
References: <ramakria@xxxxxxxxxxxxxxxxxx> <200110242236.f9OMagj06647@xxxxxxxxxxxxxxxxxxxx>
Sender: owner-linux-xfs@xxxxxxxxxxx
User-agent: Mutt/1.3.16i
On Wed, Oct 24, 2001 at 05:36:41PM -0500, Steve Lord wrote:
> various means depending on:
> 
>       o which kernel you are running
>       o which filesystem
> 
> All filesystems that I am aware of except for XFS use buffer heads and the
> block cache for metadata I/O. bdflush is responsible for flushing the data

(minor nit) 
Newer 2.4 ext2 uses pagecache metadata too (since somewhere in early 
2.5^W2.4)  


> Having said that, recent linux kernels have the buffer cache layered on top
> of an address space in a similar manner to how pagebuf does it. I am working
> on making the two address spaces the same one because 2.4.13 seems to have
> changed things sufficiently to break some tests on xfs - our user space
> utilities and the kernel are not agreeing on the state of the filesystem.

It'll only work for metadata of course because the inodes have their
own address space... so dump remains somewhat problematic :/

-Andi


<Prev in Thread] Current Thread [Next in Thread>