xfs
[Top] [All Lists]

Re: [Security] XFS swapext ioctl minor security issues

To: Dan Rosenberg <dan.j.rosenberg@xxxxxxxxx>
Subject: Re: [Security] XFS swapext ioctl minor security issues
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Thu, 17 Jun 2010 07:40:51 -0400
Cc: Christoph Hellwig <hch@xxxxxxxxxxxxx>, security@xxxxxxxxxx, aelder@xxxxxxx, Eugene Teo <eugeneteo@xxxxxxxxx>, xfs@xxxxxxxxxxx
In-reply-to: <AANLkTimJHo3zfi0t96vVlTYgDswh8QtsptoTNkVK7Eb9@xxxxxxxxxxxxxx>
References: <AANLkTilrwmh6n7yYkqyvy_y5-bgS-BEDept0WlLg5GE1@xxxxxxxxxxxxxx> <AANLkTikGFq8iv4S3QWp5ZCvXJsjuiP2tKweSl6QwHc6U@xxxxxxxxxxxxxx> <20100616121142.GA22317@xxxxxxxxxxxxx> <AANLkTimVtpQuRZpOJ4IGlFzRqWG6XaYGzvUnDuF06MyG@xxxxxxxxxxxxxx> <20100616133433.GA16437@xxxxxxxxxxxxx> <AANLkTimJHo3zfi0t96vVlTYgDswh8QtsptoTNkVK7Eb9@xxxxxxxxxxxxxx>
User-agent: Mutt/1.5.20 (2009-08-17)
On Wed, Jun 16, 2010 at 09:57:35AM -0400, Dan Rosenberg wrote:
> I removed the part of the patch dealing with suid/sgid bits - your
> reasoning seems good, we clearly don't want to just drop the suid/sgid
> bits. ?I was just trying to point out the case where the caller is not
> the owner and has write access to the file; since in the ordinary case
> writing to that file would result in dropping the suid bit, I thought
> this ioctl should try to replicate that behavior.

Looks good,

Reviewed-by: Christoph Hellwig <hch@xxxxxx>

Alex, can you push it to Linus ASAP?

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