News for hackers. Stuff that matters.
Re:Native OpenBSD hypervisor hits -current by api on Sunday November 22, 2015 @08:02PM (#10612271)
Make it clean and fast and you'll have a compelling secure host OS for VMs.

One problem that would be good to address (and maybe it is) is making it really hard to dump the memory of a VM. We don't have homomorphic encryption yet, so right now if you dump a VM you own any secrets it holds. So at the very least this should be made as hard as possible on production VM hosts.

Re:Native OpenBSD hypervisor hits -current by x0x0 on Sunday November 22, 2015 @08:06PM (#10612287)
What does making it hard for the host to dump vm memory buy you? It seems like if the host wants, there's virtually endless things they can do. iow, you have to trust the host.
Re:Native OpenBSD hypervisor hits -current by doomrobo on Sunday November 22, 2015 @08:08PM (#10612292)
Agreed. If you own the host, you win. The point is to make it difficult to own the host from within the VM, not vice-versa.
Re:Native OpenBSD hypervisor hits -current by ams6110 on Sunday November 22, 2015 @08:35PM (#10612371)
If I'm running a guest on a hosting service, I'd like it if I didn't have to trust the host. Technically, that's maybe not possible right now, but it's not like it's a goal we should just shrug off.
Re:Native OpenBSD hypervisor hits -current by doomrobo on Sunday November 22, 2015 @08:43PM (#10612398)
I don't think that will ever be possible, barring some sort of quantum memory where reading clobbers data.
Re:Native OpenBSD hypervisor hits -current by toast0 on Sunday November 22, 2015 @08:55PM (#10612433)
To be pedantic, reading dram does clobber the data; but it gets re written as part of the read cycle.
Re:Native OpenBSD hypervisor hits -current by api on Sunday November 22, 2015 @08:17PM (#10612320)
Things are not insecure or secure. Things are secure to varying degrees; nothing is perfectly so and few things are totally not. (... and anyone who claims anything is totally secure is delusional.)

Making it hard buys you some resistance to less skilled attackers. Since skill tends to obey a power law distribution, this is most attackers. Make it as hard as possible and you get more and more protection from more and more attackers, and so on.

Your average attacker doesn't know enough to bust out x86 assembly language to inject root code into the kernel and dump a VM's RAM. So make it so you have to do this, and you've ruled out your average attacker. The average attacker will only 'dd' the VMs' disks, which if they've encrypted data at rest will not get them the crown jewels. But make it easy to dump VM RAM and you've exponentially increased the damage from the average attacker (due to power law distribution of skill).

Re:Native OpenBSD hypervisor hits -current by geofft on Sunday November 22, 2015 @09:06PM (#10612469)
It only takes one attacker (or curious college student) to figure it out and post it on GitHub, and then your average attacker can bust out all the assembly language required without actually knowing it.

For comparison -- does your average Windows attacker know how to cryptanalyze the LM hash function? No, but they can run the tool just the same.

There are absolutely places where your the general argument holds water: "no security by obscurity" is a simplistic framework, and ignores a much more reasonable and also simpler framework of costs and incentives. In several cases, the costs of figuring out something obscure outweigh the potential incentives of breaking it.

But in this particular case, there are no marginal costs to the attacker. Once someone somewhere invests the time to figure it out, for any motivation including curiosity, the defense is immediately worthless for everyone.

Re:Native OpenBSD hypervisor hits -current by hueving on Sunday November 22, 2015 @08:30PM (#10612351)
Whatever stupid barriers you put up will just be automated away in common tools. Then all you have is something more complex than it needs to be that probably has bugs and vulnerabilities because you made it convoluted.
Re:Native OpenBSD hypervisor hits -current by DasIch on Sunday November 22, 2015 @09:07PM (#10612472)
The name script kiddie isn't an accident. The average attacker uses an exploit developed by a more experienced one.

You're trying to protect yourself from an imaginary class of attacker that just doesn't exist.

Re:Native OpenBSD hypervisor hits -current by blibble on Sunday November 22, 2015 @08:39PM (#10612386)
every piece of vm software has a save/snapshot button which will save the state of the ram/cpu to disk.
Re:Native OpenBSD hypervisor hits -current by jodrellblank on Sunday November 22, 2015 @08:15PM (#10612316)
right now if you dump a VM you own any secrets it holds

Unrelated to OpenBSD but related to VM security from administrators - https://4sysops.com/archives/shielded-vms-in-windows-server-...

Re:Native OpenBSD hypervisor hits -current by tyho on Sunday November 22, 2015 @08:39PM (#10612385)
This is not very useful, it is a security policy change. The previous post mentioned homomorphic encryption. This is a method of performing computation on the cipher text. If this were implemented, then you could potentially create VM's where even in the event of a hypervisor compromise, dumping the plaintext memory of guests is not possible.

If you can take a memory dump of the host sytem, you can from that devise the encryption key for the disk encryption with the M$ solution to this issue.

Re:Native OpenBSD hypervisor hits -current by mkj on Sunday November 22, 2015 @09:29PM (#10612539)
There's Intel SGX which give secure enclaves through hardware, could probably do unreadable VMs.
Re:Native OpenBSD hypervisor hits -current by amluto on Monday November 23, 2015 @12:49AM (#10613091)
No, or at least not in the traditional sense. SGX is strictly CPL 3, i.e. user mode. You could plausibly stick a unikernel inside.
Re:Native OpenBSD hypervisor hits -current by mkj on Wednesday November 25, 2015 @06:41AM (#10626693)
Oh I'd missed that, thanks.

Maybe uclinux...? Or qemu - more nesting!

Re:Native OpenBSD hypervisor hits -current by alphonse23 on Sunday November 22, 2015 @08:25PM (#10612339)
homomorphic encryption exists. It's just slow, though it depends on what you consider slow: https://github.com/shaih/HElib