Grsecurity is about to be discontinued, unless...

Grsecurity is about to be discontinued, unless...

[Gabor MICSKO]

Hi,

As most of you probably know, a GPL licensed security solution called
grsecurity [1] has been available for the Linux kernel since a while. It
has a rather impressive list of features [2]. The lead developer has
been maintaining patches for the 2.4 and the 2.6 branch for many years.

According to their developers, the patch includes various advanced
security aspects which inspired several further projects [3].

A week ago, the latest - and probably the last - release was published.
The main developer lost its sole sponsor due to the financial crisis, so
the future of the project is in danger. As a result, the future
development of PaX [4], one of the definitive components of grsecurity
is also in deep trouble [5].

In the past, there have been several requests toward the Linux
developers to include grsecurity and PaX in the mainline kernel
[6][7][8], but in vain.

The common opinion of the developers of grsecurity, PaX and their users
is that acceptance of the code into the kernel would be the best
solution for saving the project, beside finding another long-term
sponsor.

Before the project would finally die, I would like to draw your
attention to the question of integration into the kernel again.

In short, I would like to know what is your answer to this request. And
in the case if you see no chance for the integration, I would like to
know what is the reason behind this decision.

[1] http://www.grsecurity.net
[2] http://www.grsecurity.net/features.php
[3] http://grsecurity.net/~spender/grsecurity_pax-influence.png
[4] http://pax.grsecurity.net/
[5] http://www.grsecurity.net/news.php#grsec2112
[6] http://lkml.org/lkml/2003/4/20/13
[7] http://lkml.indiana.edu/hypermail/linux/net/0602.0/0020.html
[8] http://lkml.org/lkml/2005/1/13/184


Thanks and best regards,


-- 
Gabor MICSKO - http://hup.hu

Re: Grsecurity is about to be discontinued, unless...

[Linus Torvalds]

On Sat, 3 Jan 2009, Gabor MICSKO wrote:
> 
> In short, I would like to know what is your answer to this request. And
> in the case if you see no chance for the integration, I would like to
> know what is the reason behind this decision.

Quite frankly, historically the patches have been a mis-match of

 - some sane stuff

 - lots of totally insane and very annoying and invasive code.

an example of the second class is the totally idiotic "emulate NX bit - 
badly - with code segment limits" kind of patches. These kinds of patches 
only add unmaintainable crap that isn't even relevant in the long run, and 
thus any sane maintainer (me) should refuse to take them.

The apparent inability (and perhaps more importantly - total 
unwillingless) from the PaX team to be able to see what makes sense in a 
long-term general kernel and what does not, and split things up and try to 
push the sensible things up (and know which things are too ugly or too 
specialized to make sense), caused many PaX features to never be merged.

Much of it did get merged over the years (mostly because some people spent 
the time to separate things out), but no, we're not going to suddenly 
start merging code like that just because the project is in trouble. None 
of the basic issues have been solved.

			Linus

Re: Grsecurity is about to be discontinued, unless...

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 819 bytes --]

On Sat, 03 Jan 2009 12:16:50 PST, Linus Torvalds said:
> On Sat, 3 Jan 2009, Gabor MICSKO wrote:
> > In short, I would like to know what is your answer to this request. And
> > in the case if you see no chance for the integration, I would like to
> > know what is the reason behind this decision.
> 
> Quite frankly, historically the patches have been a mis-match of
> 
>  - some sane stuff

Probably the best way to proceed would be for the stakeholders to come to some
agreement on which parts are the "sane stuff" (which could be an interesting
food fight), split those parts out, and submit them for inclusion as standalone
separate patches.

It's a shame that the grsecurity developers had issues with the LSM architecture,
as their RBAC would probably be a nice addition to the slowly-growing set of
LSM modules.

[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: Grsecurity is about to be discontinued, unless...

[Jiri Kosina]

On Sat, 3 Jan 2009, Gabor MICSKO wrote:

> In short, I would like to know what is your answer to this request. And 
> in the case if you see no chance for the integration, I would like to 
> know what is the reason behind this decision.

I think that the main question here is -- what exact features are you 
missing in the vanilla Linux kernel that grsecurity provides?

Over past months/years, vanilla kernel gained some of the features that 
were initially present only in the PaX/grsecurity patchsets (address space 
layout randomization for x86, to name the one that I have been involved 
in myself).
Other things (roles? && access control stuff ) could be achieved by code 
that is already present in the vanilla kernel (selinux, smack and 
friends).

Thanks,

-- 
Jiri Kosina
SUSE Labs