On Tue, Oct 15, 2002 at 05:07:06PM -0700, Greg KH wrote:
On Tue, Oct 15, 2002 at 01:28:28PM -0700, Greg KH wrote:
quoted
On Tue, Oct 15, 2002 at 01:10:37PM -0700, David S. Miller wrote:
quoted
I will not even look at the networking LSM bits until
CONFIG_SECURITY=n is available.
BTW, there's another big issues with LSM: so far all those hook
have no user in a mergeable shape. For all other additions
there is a strong need to present something mergable but LSM
doesn't. IMHO we should require a pointer to a module in mergaable
shape (i.e. certainly not selinux) for each new hook addition.
On Thu, Oct 17, 2002 at 02:21:49PM +0100, Christoph Hellwig wrote:
On Tue, Oct 15, 2002 at 05:07:06PM -0700, Greg KH wrote:
quoted
On Tue, Oct 15, 2002 at 01:28:28PM -0700, Greg KH wrote:
quoted
On Tue, Oct 15, 2002 at 01:10:37PM -0700, David S. Miller wrote:
quoted
I will not even look at the networking LSM bits until
CONFIG_SECURITY=n is available.
BTW, there's another big issues with LSM: so far all those hook
have no user in a mergeable shape. For all other additions
there is a strong need to present something mergable but LSM
doesn't. IMHO we should require a pointer to a module in mergaable
shape (i.e. certainly not selinux) for each new hook addition.
Heh, require this, and oops, all of the hooks disappear :)
Seriously, no, I don't agree with this. SELinux is currently being
audited by a number of different companies (include some Linux distros),
and after that happens, and the code is cleaned up, I think they will
probably want their module merged (but I don't speak for them at all.)
As for the other modules, I think the OWL-based one is good enough right
now, and I have a very simple module that is in the November issue of
Linux Journal that is probably clean enough to merge.
thanks,
greg k-h