On 2017-06-29 19:58, Steven Rostedt wrote:
On Thu, 29 Jun 2017 17:21:22 -0400
Richard Guy Briggs <rgb(a)redhat.com> wrote:
> > Looking at this again today, why would we want to clear name->dentry
> > in audit_copy_inode() if it is already set? Does that ever happen?
> > I'm not sure it does ...
>
> It has been nearly 3 months since I coded that, so I'll have to dive in
> and re-analyse what I was thinking at that time. I think that rationale
> was that if audit_copy_inode() is called again on that audit_name struct
> that it could be called by audit_log_link_denied() or __audit_inode()
> not needing the dentry reference or even by __audit_inode_child() and
> have it replaced, needing a reference count correction.
Just a note. If after 3 months you need to re-analyze, you either need
to design things simpler, or have better comments in the code.
Yep, both occurred to me. ;-) Thanks Steve.
-- Steve
- RGB
--
Richard Guy Briggs <rgb(a)redhat.com>
Sr. S/W Engineer, Kernel Security, Base Operating Systems
Remote, Ottawa, Red Hat Canada
IRC: rgb, SunRaycer
Voice: +1.647.777.2635, Internal: (81) 32635