mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Kyle Moffett <mrmacman_g4@mac.com>
To: James Antill <james@and.org>
Cc: David Madore <david.madore@ens.fr>,
	Linux Kernel mailing-list <linux-kernel@vger.kernel.org>
Subject: Re: patch to make Linux capabilities into something useful (v 0.3.1)
Date: Thu, 7 Sep 2006 14:33:56 -0400	[thread overview]
Message-ID: <ED485CED-AE69-466E-876C-2CC9DADC5576@mac.com> (raw)
In-Reply-To: <m3r6yn4jxg.fsf@code.and.org>

On Sep 07, 2006, at 14:21:15, James Antill wrote:
> David Madore <david.madore@ens.fr> writes:
>
>> Hi.
>>
>> As we all know, capabilities under Linux are currently crippled to  
>> the point of being useless.  Attached is a patch (against 2.6.18- 
>> rc6) which attempts to make them work in a reasonably useful way  
>> and at the same time not break anything.  On top of the  
>> "additional" capabilities that lead up to root, it also adds  
>> "regular" capabilities which all processes have by default and  
>> which can be removed from specifically untrusted programs.
>
> Just a minor comment, can you break out the OPEN into at least  
> OPEN_R, OPEN_NONFILE_W and OPEN_W (possibly OPEN_A, but I don't  
> want that personally). The case I'm thinking about are network  
> daemons that need to open+write to the syslog socket but only have  
> read access elsewhere.
>
> Also there is much more than just bind9 using capabilities, the  
> obvious ones that come to mind are NOATIME and AUDIT.

To be honest, once you get to the point of having an OPEN_NONFILE_W  
capability you should really just be using SELinux.  Instead of  
spewing policy all over your filesystem xattrs you put it all into a  
single set of fairly admin-readable policy files, and it includes  
more fine-grained operations than you have space for in 32 bits of  
"un"-capabilities.  It also conveniently supports the existing pseudo- 
POSIX capabilities.  You can effectively say "root, netadmin, and  
netuser users are allowed to run ping programs which automatically  
transition into the ping domain and get CAP_NET_RAW privileges" in  
your binary policy, then just leave ping as chmod "755" and be done  
with it.

Cheers,
Kyle Moffett


  reply	other threads:[~2006-09-07 18:34 UTC|newest]

Thread overview: 52+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-09-05 21:26 David Madore
2006-09-06  0:27 ` Casey Schaufler
2006-09-06 10:06   ` David Madore
2006-09-06 13:26     ` David Madore
2006-09-07  0:11       ` Casey Schaufler
2006-09-07  0:32         ` David Madore
2006-09-07  1:01           ` Casey Schaufler
2006-09-07  1:29             ` David Wagner
2006-09-07 16:00               ` Casey Schaufler
2006-09-07 18:33                 ` David Wagner
2006-09-07 17:34             ` David Madore
2006-09-07 19:38               ` Bernd Eckenfels
2006-09-07 23:00                 ` Pavel Machek
2006-09-08  1:22                   ` Bernd Eckenfels
2006-09-08 10:45                     ` Pavel Machek
2006-09-08 16:08                       ` Casey Schaufler
2006-09-08 14:39                     ` Pavel Machek
2006-09-08 19:10                       ` Bernd Eckenfels
2006-09-07 22:54               ` Pavel Machek
2006-09-08  4:10                 ` David Madore
2006-09-08 10:52                   ` Pavel Machek
2006-09-08 22:51                     ` David Madore
2006-09-09  0:11                       ` Casey Schaufler
2006-09-09 11:59                         ` Pavel Machek
2006-09-09 11:40                       ` Pavel Machek
2006-09-10 10:41                         ` David Madore
2006-09-10 13:06                           ` Pavel Machek
2006-09-10 14:25                             ` capability inheritance (was: Re: patch to make Linux capabilities into something useful (v 0.3.1)) David Madore
2006-09-10 22:42                               ` Pavel Machek
2006-09-11 16:00                               ` Casey Schaufler
2006-09-11 17:39                                 ` David Madore
2006-09-09  0:59                   ` patch to make Linux capabilities into something useful (v 0.3.1) David Wagner
2006-09-09 12:49                     ` David Madore
2006-09-09 23:18       ` Theodore Tso
2006-09-10 10:13         ` David Madore
2006-09-10 12:36         ` Pavel Machek
2006-09-10 23:24           ` Theodore Tso
2006-09-11  8:09             ` Pavel Machek
2006-09-06 18:25 ` Serge E. Hallyn
2006-09-06 22:27   ` David Madore
2006-09-07  0:04     ` David Madore
2006-09-07 23:06       ` Serge E. Hallyn
2006-09-08  4:16         ` David Madore
2006-09-07  6:43     ` Jan Engelhardt
2006-09-07 23:02     ` Serge E. Hallyn
2006-09-08  1:08       ` David Madore
2006-09-08  1:31         ` Serge E. Hallyn
2006-09-08 21:45           ` David Madore
2006-09-07 18:21 ` James Antill
2006-09-07 18:33   ` Kyle Moffett [this message]
2006-09-07 20:05     ` James Antill
2006-09-08  4:00   ` David Madore

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=ED485CED-AE69-466E-876C-2CC9DADC5576@mac.com \
    --to=mrmacman_g4@mac.com \
    --cc=david.madore@ens.fr \
    --cc=james@and.org \
    --cc=linux-kernel@vger.kernel.org \
    /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

all inboxes | Powered by JetHome®