Re: [BK PATCH] LSM changes for 2.5.27

[Greg KH <greg@kroah.com>]

On Tue, Jul 23, 2002 at 02:16:34PM +0100, Alan Cox wrote:
> On Tue, 2002-07-23 at 09:16, Roman Zippel wrote:
> > Hi,
> > 
> > On Mon, 22 Jul 2002, Greg KH wrote:
> > 
> > > +		error = security_ops->inode_setattr(dentry, attr);
> > 
> > Am I the only one who'd like to see this as an inline function?
> > 1. It can be optimized away.
> > 2. It's easier to read.

Yes, I've considered it.  I might still wrap them in a inline function
if people _really_ don't like the look of them.

> You are not the only one. At the kernel summit there were discussions
> about both wrapping the few performance impacting ones in ifdefs, and/or
> using dynamic patching.

Yes, for the hooks that might affect performance (like the network ones)
they will probably be wrapped in inline functions, and controlled by a
config option.

thanks,

greg k-h