Add a new capable interface that will be used by systems that use audit to
commit06112163f5fd9e491a7f810443d81efa9d88e247
authorEric Paris <eparis@redhat.com>
Tue, 11 Nov 2008 11:02:50 +0000 (11 22:02 +1100)
committerJames Morris <jmorris@namei.org>
Tue, 11 Nov 2008 11:02:50 +0000 (11 22:02 +1100)
tree48039f7488abbec36c0982a57405b57d47311dd6
parent637d32dc720897616e8a1a4f9e9609e29d431800
Add a new capable interface that will be used by systems that use audit to
make an A or B type decision instead of a security decision.  Currently
this is the case at least for filesystems when deciding if a process can use
the reserved 'root' blocks and for the case of things like the oom
algorithm determining if processes are root processes and should be less
likely to be killed.  These types of security system requests should not be
audited or logged since they are not really security decisions.  It would be
possible to solve this problem like the vm_enough_memory security check did
by creating a new LSM interface and moving all of the policy into that
interface but proves the needlessly bloat the LSM and provide complex
indirection.

This merely allows those decisions to be made where they belong and to not
flood logs or printk with denials for thing that are not security decisions.

Signed-off-by: Eric Paris <eparis@redhat.com>
Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: James Morris <jmorris@namei.org>
include/linux/capability.h
include/linux/security.h
security/commoncap.c
security/security.c
security/selinux/hooks.c