xfs
[Top] [All Lists]

Re: O_DIRECT & rpm (for xfs)

To: Axel Thimm <Axel.Thimm@xxxxxxxxxxxxxxxxxxx>
Subject: Re: O_DIRECT & rpm (for xfs)
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Fri, 29 Aug 2003 18:12:44 +0100
Cc: Jeff Johnson <n3npq@xxxxxxxxx>, linux-xfs@xxxxxxxxxxx
In-reply-to: <20030829162928.GB2911@pua.nirvana>; from Axel.Thimm@physik.fu-berlin.de on Fri, Aug 29, 2003 at 07:29:28PM +0300
References: <20030829162928.GB2911@pua.nirvana>
Sender: linux-xfs-bounce@xxxxxxxxxxx
User-agent: Mutt/1.2.5.1i
On Fri, Aug 29, 2003 at 07:29:28PM +0300, Axel Thimm wrote:
> O_DIRECT -- as instantiated in linux -- is goofy and under development. 
> The behavior before was
> just to ignore O_DIRECT, it did not matter whether it was used or not. 
> Life was good.

Umm, that's wrong.  O_DIRECT went into Linux 2.4.10 (and was in the
XFS patches earlier) and never was ignored.  Someone during the last
year Red Hat decided to just clear the O_DIRECT flag in their kernels
instead of returning an error.  That's where this crap comes from.
It's a genuine Red Hat bug.

The alignment was blocksize previously and got down to sector size
these days - but that's lessing the requirement, nothing that can
harm an application.


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