linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Stephen Smalley <sds@tycho.nsa.gov>
To: Andreas Gruenbacher <agruen@suse.de>
Cc: James Morris <jmorris@namei.org>,
	John Johansen <jjohansen@suse.de>,
	Andrew Morton <akpm@linux-foundation.org>,
	linux-kernel@vger.kernel.org,
	linux-security-module@vger.kernel.org,
	Miklos Szeredi <mszeredi@suse.cz>
Subject: Re: [AppArmor 32/44] Enable LSM hooks to distinguish operations on file descriptors from operations on pathnames
Date: Tue, 03 Jul 2007 09:49:40 -0400	[thread overview]
Message-ID: <1183470580.12218.253.camel@moss-spartans.epoch.ncsc.mil> (raw)
In-Reply-To: <200706282015.51430.agruen@suse.de>

On Thu, 2007-06-28 at 20:15 +0200, Andreas Gruenbacher wrote:
> On Thursday 28 June 2007 18:12, James Morris wrote:
> > Are you trying to cater for the case where you're holding an open fd for a 
> > file which has been deleted, and thus has no pathname?
> 
> Yes, see the AA_CHECK_FD flag in security/apparmor/main.c:aa_perm_dentry(). We 
> want to distinguish between the following two cases:
> 
>  - process performs an operation on an open file descriptor,
> 
>  - process performs an operation on a pathname, and between the dentry
>    lookup and the LSM permission check, the file gets deleted.
> 
> In the former case, we obviously want to continue giving the process access to 
> his fd (the classical pattern: open temporary file; delete it so that it will 
> self-recycle, continue using the open file descriptor).
> 
> In the latter case, The file still existed at the time of the lookup but not 
> anymore at the time of the permission check. The file obviously doesn't have 
> a filename anymore, so we cannot check permissions. If we granted access in 
> that case, processes could bypass their profile permissions in that race 
> window. We close the race by returning -ENOENT in that case, the same result 
> as if the file had already been deleted before the lookup.

So you don't actually need/use the struct file pointer; you just need a
flag indicating whether or not access was by open file descriptor or by
pathname?

And what does this mean for a process that has "changed hats"?  Which
might not be authorized to access the file anymore, even via an already
opened descriptor.

-- 
Stephen Smalley
National Security Agency


  reply	other threads:[~2007-07-03 13:50 UTC|newest]

Thread overview: 76+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-06-26 23:07 [AppArmor 00/44] AppArmor security module overview jjohansen
2007-06-26 23:07 ` [AppArmor 01/44] Pass struct vfsmount to the inode_create LSM hook jjohansen
2007-06-30  9:29   ` Christoph Hellwig
2007-07-03 22:28     ` Andreas Gruenbacher
2007-06-26 23:07 ` [AppArmor 02/44] Pass struct path down to remove_suid and children jjohansen
2007-06-26 23:07 ` [AppArmor 03/44] Add a vfsmount parameter to notify_change() jjohansen
2007-06-26 23:08 ` [AppArmor 04/44] Pass struct vfsmount to the inode_setattr LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 05/44] Add struct vfsmount parameter to vfs_mkdir() jjohansen
2007-06-26 23:08 ` [AppArmor 06/44] Pass struct vfsmount to the inode_mkdir LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 07/44] Add a struct vfsmount parameter to vfs_mknod() jjohansen
2007-06-26 23:08 ` [AppArmor 08/44] Pass struct vfsmount to the inode_mknod LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 09/44] Add a struct vfsmount parameter to vfs_symlink() jjohansen
2007-06-26 23:08 ` [AppArmor 10/44] Pass struct vfsmount to the inode_symlink LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 11/44] Pass struct vfsmount to the inode_readlink " jjohansen
2007-06-26 23:08 ` [AppArmor 12/44] Add struct vfsmount parameters to vfs_link() jjohansen
2007-06-26 23:08 ` [AppArmor 13/44] Pass the struct vfsmounts to the inode_link LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 14/44] Add a struct vfsmount parameter to vfs_rmdir() jjohansen
2007-06-26 23:08 ` [AppArmor 15/44] Pass struct vfsmount to the inode_rmdir LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 16/44] Call lsm hook before unhashing dentry in vfs_rmdir() jjohansen
2007-06-26 23:08 ` [AppArmor 17/44] Add a struct vfsmount parameter to vfs_unlink() jjohansen
2007-06-26 23:08 ` [AppArmor 18/44] Pass struct vfsmount to the inode_unlink LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 19/44] Add struct vfsmount parameters to vfs_rename() jjohansen
2007-06-26 23:08 ` [AppArmor 20/44] Pass struct vfsmount to the inode_rename LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 21/44] Add a struct vfsmount parameter to vfs_setxattr() jjohansen
2007-06-26 23:08 ` [AppArmor 22/44] Pass struct vfsmount to the inode_setxattr LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 23/44] Add a struct vfsmount parameter to vfs_getxattr() jjohansen
2007-06-26 23:08 ` [AppArmor 25/44] Add a struct vfsmount parameter to vfs_listxattr() jjohansen
2007-06-26 23:08 ` [AppArmor 26/44] Pass struct vfsmount to the inode_listxattr LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 27/44] Add a struct vfsmount parameter to vfs_removexattr() jjohansen
2007-06-26 23:08 ` [AppArmor 28/44] Pass struct vfsmount to the inode_removexattr LSM hook jjohansen
2007-06-26 23:08 ` [AppArmor 29/44] Fix __d_path() for lazy unmounts and make it unambiguous jjohansen
2007-06-26 23:08 ` [AppArmor 30/44] Make d_path() consistent across mount operations jjohansen
2007-06-26 23:08 ` [AppArmor 32/44] Enable LSM hooks to distinguish operations on file descriptors from operations on pathnames jjohansen
2007-06-28 16:12   ` James Morris
2007-06-28 18:15     ` Andreas Gruenbacher
2007-07-03 13:49       ` Stephen Smalley [this message]
2007-07-03 20:01         ` Andreas Gruenbacher
2007-06-26 23:08 ` [AppArmor 33/44] Pass struct file down the inode_*xattr security LSM hooks jjohansen
2007-06-26 23:08 ` [AppArmor 34/44] Factor out sysctl pathname code jjohansen
2007-06-26 23:08 ` [AppArmor 35/44] Allow permission functions to tell between parent and leaf checks jjohansen
2007-06-26 23:08 ` [AppArmor 36/44] Export audit subsystem for use by modules jjohansen
2007-06-26 23:08 ` [AppArmor 37/44] AppArmor: Main Part jjohansen
2007-06-26 23:08 ` [AppArmor 38/44] AppArmor: Module and LSM hooks jjohansen
2007-06-26 23:08 ` [AppArmor 39/44] AppArmor: Profile loading and manipulation, pathname matching jjohansen
2007-06-26 23:08 ` [AppArmor 40/44] AppArmor: all the rest jjohansen
2007-06-26 23:08 ` [AppArmor 41/44] Add AppArmor LSM to security/Makefile jjohansen
2007-06-26 23:08 ` [AppArmor 42/44] Switch to vfs_permission() in do_path_lookup() jjohansen
2007-06-26 23:08 ` [AppArmor 43/44] Switch to vfs_permission() in sys_fchdir() jjohansen
2007-06-26 23:08 ` [AppArmor 44/44] Fix file_permission() jjohansen
2007-06-26 23:52 ` [AppArmor 00/44] AppArmor security module overview Andrew Morton
2007-06-27  2:24   ` John Johansen
2007-06-27  2:47     ` Andrew Morton
2007-06-27  6:43       ` John Johansen
2007-06-27 15:11       ` Adrian Bunk
2007-06-27 21:06         ` Crispin Cowan
2007-06-27 21:29           ` Sean
2007-06-27 22:46             ` Crispin Cowan
2007-06-27 23:05               ` David Miller
2007-06-28  0:27                 ` Casey Schaufler
2007-06-28  0:34                   ` David Miller
2007-06-28 10:23                   ` Alan Cox
2007-06-28 13:50                   ` Bill O'Donnell
2007-06-28 11:27                 ` Tilman Schmidt
2007-06-28 12:48                   ` Adrian Bunk
2007-06-27 22:41         ` Andreas Dilger
2007-07-02 16:45         ` Eric W. Biederman
2007-07-02 19:31           ` Casey Schaufler
2007-07-02 20:15             ` Christoph Hellwig
2007-07-02 21:09               ` Casey Schaufler
2007-07-03 16:33               ` Andreas Gruenbacher
2007-06-29 18:06       ` Pavel Machek
2007-07-03  6:20       ` Dave Jones
2007-06-27 10:58     ` Kyle Moffett
2007-06-27 13:37       ` Andreas Gruenbacher
2007-06-27  0:32 ` [AppArmor 24/44] Pass struct vfsmount to the inode_getxattr LSM hook jjohansen
2007-06-27  0:32 ` [AppArmor 31/44] Add d_namespace_path() to compute namespace relative pathnames jjohansen

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1183470580.12218.253.camel@moss-spartans.epoch.ncsc.mil \
    --to=sds@tycho.nsa.gov \
    --cc=agruen@suse.de \
    --cc=akpm@linux-foundation.org \
    --cc=jjohansen@suse.de \
    --cc=jmorris@namei.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-security-module@vger.kernel.org \
    --cc=mszeredi@suse.cz \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).