All of lore.kernel.org
 help / color / mirror / Atom feed
From: Keir Fraser <keir@xen.org>
To: Ian Campbell <Ian.Campbell@citrix.com>,
	Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: George Dunlap <George.Dunlap@eu.citrix.com>,
	Windows Live <johneed@hotmail.com>,
	Tim Deegan <Tim.Deegan@eu.citrix.com>,
	"xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>
Subject: Re: acm
Date: Tue, 09 Aug 2011 15:41:11 +0100	[thread overview]
Message-ID: <CA670597.3025D%keir@xen.org> (raw)
In-Reply-To: <1312900480.26263.139.camel@zakaz.uk.xensource.com>

On 09/08/2011 15:34, "Ian Campbell" <Ian.Campbell@citrix.com> wrote:

> On Tue, 2011-08-09 at 15:29 +0100, Konrad Rzeszutek Wilk wrote:
>> Tim, George,
>> 
>> Any ideas who at might know more about the ACM flask?
> 
> Didn't Keir remove ACM support back in 23097:2aeebd5cbbad? Probably post
> Xen 4.1 but on that basis I wouldn't recommend enabling it on any recent
> release which happened to include it -- it was removed because it was
> unmaintained.

Yes, Flask and ACM are (were) two different security models implemented
under a common XSM security framework. Only Flask is being actively
maintained, so I deleted ACM. I would recommend Gentoo only enable xsm in
their build -- at least if that fails there are maintainers who should be
interested in helping to fix it. Noone will help you with acm.

 -- Keir

> Ian.
> 
>> 
>> On Sat, Aug 06, 2011 at 04:28:37PM +0800, Windows Live wrote:
>>> 
>>> ------- Comment
>>>           
>>>             #15 From
>>>           Konrad Rzeszutek Wilk
>>>           2011-08-05 19:11
>>>           
>>>           [reply]
>>>           -------
>>>         
>>>         
>>> 
>>> 
>>> Just copy-n-paste the bug and post it on xen-devel mailing list.
>>> 
>>> 
>>>  http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=1775
>>> 
>>> a reduced summary;
>>> 
>>> 
>>> This fault renders xen unworkable, only in the packaged install of distro
>>> gentoo/
>>> 
>>> gentoo64 grub # eix xen
>>> [I] app-emulation/xen
>>>      Available versions:  3.4.2-r1!t (~)4.1.0!t **9999!t {acm custom-cflags
>>> debug flask pae xsm}
>>>      Installed versions:  4.1.0!t(00:13:36 07/30/11)(acm debug flask pae xsm
>>> -custom-cflags)
>>>      Homepage:            http://xen.org/
>>>      Description:         The Xen virtual machine monitor
>>> 
>>> This make /boot/xen-4.1.0.gz.  On booting into this installed hypervisor, I
>>> can
>>> bring up virt-manager which happily lists the vms.
>>> 
>>> On attempting to boot a v the moment the xen equipped kernel attempts to
>>> boot
>>> the vm, the system is crashed.
>>> 
>>> That is, it seems to trigger reboot because that is what happens,
>>> Just reboot.
>>> I haven't bothered with any logs because this instantaneous crash should
>>> fail
>>> to write any events to logs, correct me if I'm wrong
>>> 
>>> ------- Comment
>>>           
>>>             #1 From
>>>           IAN DELANEY
>>>           2011-07-30 02:36
>>>           
>>>           [reply]
>>>           -------
>>>         
>>>         
>>> 
>>> 
>>> should have mentioned.
>>> 
>>> idella@gentoo64 ~/bin $ ls /boot/
>>> ...........
>>> xen-4.1.0-rc7-pre.gz
>>> xen-4.1.0.gz
>>> ........
>>> 
>>> The xen-4.1.0-rc7-pre.gz is fine
>>> The xen-4.1.0.gz hypervisor is broken.
>>> 
>>> idella@gentoo64 ~/bin $ emerge xen -pv
>>> 
>>> These are the packages that would be merged, in order:
>>> 
>>> Calculating dependencies... done!
>>> [ebuild   R    ] app-emulation/xen-4.1.0  USE="acm custom-cflags* debug
>>> flask
>>> pae xsm" 0 kB
>>> 
>>> shows the compiled in features.  What can trigger this type of system crash?
>>> 
>>>     
>>> 
>>>         
>>>           ------- Comment
>>>           
>>>             #3 From
>>>           IAN DELANEY
>>>           2011-07-30 07:31
>>>           
>>>           [reply]
>>>           -------
>>>         
>>>         
>>> 
>>> 
>>> It seems the cause of this is in gentoo terms the use of the use flags acm
>>> and
>>> xsm.  See https://bugs.gentoo.org/show_bug.cgi?id=361345.
>>> It seems disabling these flags allows the hypervisor to boot vms.
>>> What I need to know is where this flaw comes from.  There is a gentoo
>>> package
>>> xsm, but not asm.
>>> asm is a security portion of the source.
>>> Is the flaw in the gentoo package xsm, is it the xensource code, and can you
>>> list the step or steps to sensure xsm and acm directly in the xensource code
>>> so
>>> I can replicate it;  suspect they are options manually put after make on the
>>> command line.
>>> 
>>> ------- Comment #5 From IAN DELANEY 2011-07-30 08:01  [reply]  -------
>>>         
>>>         
>>> 
>>> 
>>> Aha
>>> 
>>> from Config.mk
>>> 
>>> # Enable XSM security module.  Enabling XSM requires selection of an
>>> # XSM security module (FLASK_ENABLE or ACM_SECURITY).
>>> XSM_ENABLE ?= n
>>> FLASK_ENABLE ?= n
>>> ACM_SECURITY ?= n
>>> 
>>> These are the default settings, so it seems the hypervisor I am using does
>>> in
>>> fact not have these set.
>>> So why are they turned off, and do they have an unresolved issue?
>>> 
>>>     
>>>   
>>>         
>>> 
>>>         
>>>           ------- Comment
>>>           
>>>             #6 From
>>>           Konrad Rzeszutek Wilk
>>>           2011-07-30 08:26
>>>           
>>>           [reply]
>>>           -------
>>>         
>>>         
>>> 
>>> 
>>> (In reply to comment #5)
>>>> Aha
>>>> 
>>>> from Config.mk
>>>> 
>>>> # Enable XSM security module.  Enabling XSM requires selection of an
>>>> # XSM security module (FLASK_ENABLE or ACM_SECURITY).
>>>> XSM_ENABLE ?= n
>>>> FLASK_ENABLE ?= n
>>>> ACM_SECURITY ?= n
>>>> 
>>>> These are the default settings, so it seems the hypervisor I am using does
>>>> in
>>>> fact not have these set.
>>>> So why are they turned off, and do they have an unresolved issue?
>>> 
>>> B/c the maintainer for them is .. gone? I suspect the code is bit-rotten.
>>> You
>>> might want to open a Gentoo bug to turn those off until somebody fixes
>>> whatever
>>> the issue you have with ACM.
>>> 
>>>> 
>>> 
>>>     
>>>   
>>>         
>>> 
>>>         
>>>           ------- Comment
>>>           
>>>             #7 From
>>>           IAN DELANEY
>>>           2011-07-30 08:40
>>>           
>>>           [reply]
>>>           -------
>>>         
>>>         
>>> 
>>> 
>>> Konrad;
>>> 
>>> 
>>> I can get the use flags withdrawn, but the task at hand is to get them to
>>> work.
>>> I can say that change to setting
>>> FLASK_ENABLE ?= y
>>> cause the emerge or build to fail in my gentoo
>>> 
>>> 
>>> 
>>>         
>>>           ------- Comment
>>>           
>>>             #8 From
>>>           IAN DELANEY
>>>           2011-07-30 10:04
>>>           
>>>           [reply]
>>>           -------
>>>         
>>>         
>>> 
>>> 
>>> ok I'm picking up the threads.
>>> 
>>> # make clean
>>> # nano Config.mk
>>> # emerge install-xen
>>> 
>>> for each alteration.
>>> 
>>> the build fails only for setting ACM_SECURITY ?= y.
>>> 
>>>     
>>>   
>>>         
>>> 
>>>         
>>>           ------- Comment
>>>           
>>>             #9 From
>>>           Konrad Rzeszutek Wilk
>>>           2011-07-31 08:14
>>>           
>>>           [reply]
>>>           -------
>>>         
>>>         
>>> 
>>> 
>>> (In reply to comment #7)
>>>> Konrad;
>>>> 
>>>> thanks.  When you say maintainer, do you mean a gentoo maintainer?
>>> 
>>> Cool. So in regards to disable ACM/FLask - that was to you. In regards to
>>> actually figuring why ACM/Flask does not seem to work - that was meant for
>>> the
>>> maintainer of that code in Xen - who seemed to have moved on to other
>>> things.
>>> 
>>> 
>>>> I can get the use flags withdrawn, but the task at hand is to get them to
>>>> work.
>>> 
>>> Ah, then you will need to figure out why FLASK does not work - I have no
>>> knowledge of that code so won't be much help. You might have better luck
>>> digging up the authors of said code and emailing them.
>>> 
>>>> I can say that change to setting
>>>> FLASK_ENABLE ?= y
>>>> cause the emerge or build to fail in my gentoo
>>>> 
>>>> Thanks for your important help.
>>>> 
>>> 
>>>     
>>>   
>>>         
>>> 
>>>         
>>>           ------- Comment
>>>           
>>>             #10 From
>>>           IAN DELANEY
>>>           2011-07-31 08:49
>>>           
>>>           [reply]
>>>           -------
>>>         
>>>         
>>> 
>>> 
>>> ok; to correct last entries.
>>> 
>>> It seems that flask is fine.  On repeating it a number of times, for
>>> whatever
>>> reason, on compiling the xensource package xen-4.1-testing.hg, it appears
>>> that
>>> entries flask and acm will in fact compile fine. It seems that setting flask
>>> on
>>> its own will fail, it depends upon xsm being set. Setting xsm om its own or
>>> with flask works.
>>> 
>>> In fact, setting all 3, the package builds. But like the gentoo emerge, the
>>> hypervisor breaks the system if ACM_SECURITY is set.  Setting use=ACM
>>> presumably leads to setting ACM_SECURITY ?= n in the Config.mk. simple.  So
>>> the
>>> flaw isn't in gentoo, it' in the source.           
>> 
>>> _______________________________________________
>>> Xen-devel mailing list
>>> Xen-devel@lists.xensource.com
>>> http://lists.xensource.com/xen-devel
>> 
>> 
>> _______________________________________________
>> Xen-devel mailing list
>> Xen-devel@lists.xensource.com
>> http://lists.xensource.com/xen-devel
> 
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel

      reply	other threads:[~2011-08-09 14:41 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-08-06  8:28 acm Windows Live
2011-08-09 14:29 ` acm Konrad Rzeszutek Wilk
2011-08-09 14:34   ` acm Ian Campbell
2011-08-09 14:41     ` Keir Fraser [this message]

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=CA670597.3025D%keir@xen.org \
    --to=keir@xen.org \
    --cc=George.Dunlap@eu.citrix.com \
    --cc=Ian.Campbell@citrix.com \
    --cc=Tim.Deegan@eu.citrix.com \
    --cc=johneed@hotmail.com \
    --cc=konrad.wilk@oracle.com \
    --cc=xen-devel@lists.xensource.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.