lkml.org 
[lkml]   [1999]   [Dec]   [17]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: [RFC] Giving capabilities to users and groups
Date
Sorry about my cryptic late-night post.
i really should have attached this to the 32bit-uid thread.

'scap' seems like a well thought user-level encapsulation
of the existing linux kernel capability mechanisms.
I didn't mean by injecting a discussion of future kernel extensions,
to dissipate your zeal for the project of making user tools
for handling existing kernel-supported mechanisms.

The parallel with 'su' is a good one.
Note also Jeremy Fitzhardinge's work on annotating ELF executables
with a capability set, as a much finer grained version of the setuid bits.
The two are complementary, 'scap' is to 'su' as ELF-caps is to SUID.
Well, sort of -- see <http://www.goop.org/~jeremy/caps>.

The worrying thing about making capabilities this accessible is that it
can lead to quick hacks using scap as a *tool*, which weaken system
integrity -- augmenting a process's capabilities make it a target
for cracking and spoofing from the unaugmented processes of the *same* uid.
I leapt upon the 32bit uid extension as a means to grant different
internal uids to different capability sets, thus breaking the relationship
of trust, and forcing same-public-uid processes with different
capability sets to go through some lengthier authentication process
before signalling/examining/meddling with each other.

But such complexity is against the spirit of linux...


> peter swain wrote:
> > i was about to reply to the author of the 32bit uid patch,
> > thinking that this should not change lightly, but should
> > wait until some discussion on extensible authorization spaces.

"Martijn van Oosterhout wrote:"
> Interesting idea, though I'm not sure if I'm understanding
> you right. Do you have anything places that describe this
> in more detail?

not yet.
i was just waffling on possibilities.
i'll drag my original outburst back to the 32bit-uid thread
when i've added a little more meat/sleep to it.

^..^
(oo) www.pobox.com/~swine

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:55    [W:1.468 / U:0.004 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site