News for hackers. Stuff that matters.
Re:Replacing x86 firmware with Linux and Go by mikeokner on Tuesday November 21, 2017 @12:26PM (#15749586)
You'd think if the ME truly wasn't nefarious that Intel would offer chips without it and capitalize on the extra features in the enterprise market. I've yet to encounter anyone who actually wants it.
Re:Replacing x86 firmware with Linux and Go by pgeorgi on Tuesday November 21, 2017 @01:00PM (#15749995)
It's also a convenient place to put in all the things they don't want to hard wire. Which gets more every day.

Need to maintain crypto keys for SGX enclave memory? Do it in the ME. Need to do some extra stuff on suspend/resume? Do it in the ME. Not sure if any other special handling might require updates at a later date? Do it in the ME. ...

There's no need for nefarious purposes to explain why the ME isn't optional anymore - it's just more convenient.

Re:Replacing x86 firmware with Linux and Go by srcmap on Tuesday November 21, 2017 @01:38PM (#15750367)
Need to add/remove/read system's crypto keys, Do it in ME.

Need to monitor/hack the computer when the users think it is "power off", Do it in ME.

Need to add other "features" to the system in the future, Do it in ME.

Re:Replacing x86 firmware with Linux and Go by pgeorgi on Thursday November 23, 2017 @11:39AM (#15766141)
Those are the concerns around the ME. But those motivations aren't necessary to explain the presence of the ME. Occam's razor and all that.
Re:Replacing x86 firmware with Linux and Go by tyingq on Tuesday November 21, 2017 @01:21PM (#15750232)
> I've yet to encounter anyone who actually wants it

Ignoring the security concerns, the remote access, imaging, etc, are actually pretty nice. Better done than most 3rd party IPMI implementations.

If it were open and documented , and able to be turned off, it has value.

Re:Replacing x86 firmware with Linux and Go by monocasa on Tuesday November 21, 2017 @12:44PM (#15749809)
It's in part used for DRM. You've probably used it if you've watched BluRays or Netflix over 720p.
Re:Replacing x86 firmware with Linux and Go by 0xFFC on Tuesday November 21, 2017 @12:53PM (#15749924)
I am genuinely asking, how intel ME is related to watching movies in Netflix? Would you mind to elaborate a little bit?
Re:Replacing x86 firmware with Linux and Go by monocasa on Tuesday November 21, 2017 @12:59PM (#15749990)
AFAIU, it's used in the HDCP encryption negotiation.
Re:Replacing x86 firmware with Linux and Go by sofaofthedamned on Tuesday November 21, 2017 @01:11PM (#15750132)
Yup. There's also rumours it holds the secure enclave equivalent, so getting root into this means you're double fucked:

https://twitter.com/mjg59/status/932730696614813696

Re:Replacing x86 firmware with Linux and Go by Groxx on Tuesday November 21, 2017 @02:48PM (#15750944)
Given Intel's announcement: https://security-center.intel.com/advisory.aspx?intelid=INTE...

>Based on the items identified through the comprehensive security review, an attacker could gain unauthorized access to platform, Intel® ME feature, and 3rd party secrets protected by the Intel® Management Engine (ME), Intel® Server Platform Service (SPS), or Intel® Trusted Execution Engine (TXE).

It seems like there's a reasonable chance of that being the case.

Re:Replacing x86 firmware with Linux and Go by freedomben on Tuesday November 21, 2017 @05:21PM (#15752166)
Link to Intel's announcement isn't working. Might be a server-side issue ¯\_(ツ)_/¯
Re:Replacing x86 firmware with Linux and Go by Fnoord on Tuesday November 21, 2017 @10:17PM (#15753949)
Apparently it needs to have appended &languageid=en-fr to work [1]

[1] https://security-center.intel.com/advisory.aspx?intelid=INTE...

Re:Replacing x86 firmware with Linux and Go by tscs37 on Tuesday November 21, 2017 @03:32PM (#15751271)
Several corporations use it for Lights-Out management or on laptops to ensure data security compliance.

The things for which you actually want a backdoor in your server to control it from. Maybe even in the face of an attacker who has gained full control of both software and hardware.

Re:Replacing x86 firmware with Linux and Go by freedomben on Tuesday November 21, 2017 @05:22PM (#15752179)
I would love a potential employer/recruiter to woo me with, "Your choice of non-backdoor-ed laptop that respects your privacy." I would at least give them a phone call for that line.
Re:Replacing x86 firmware with Linux and Go by mattnewton on Tuesday November 21, 2017 @06:56PM (#15752895)
What is private about a company laptop? Your own laptop shouldn’t be backdoored, but I think it’s irrational to expect the a laptop owned by the company won’t be managed by that company.
Re:Replacing x86 firmware with Linux and Go by freedomben on Tuesday November 21, 2017 @11:52PM (#15754347)
I don't expect it. That's why if a company offered it to me I would be impressed. It would speak volumes that they:

A) Trust their employees

B) Respect my privacy

C) Aren't paternalistic

Re:Replacing x86 firmware with Linux and Go by colejohnson66 on Tuesday November 21, 2017 @01:05PM (#15750037)
I say it also has to do with them just not caring about what their users want. You’re still gonna buy an x86 processor and AMD has their own ME-like tool too. What are you gonna do, run your desktop on ARM or RISC-V?
Re:Replacing x86 firmware with Linux and Go by cturner on Tuesday November 21, 2017 @01:47PM (#15750452)
Would be good to have a low performance riscv motherboard with something like a PCI bus. Then, run an x86 daughter card. Early arm systems (acorn RISC pc) could house a 486 daughter card like this, and you could run Windows on it in a box. Have one at home.
Re:Replacing x86 firmware with Linux and Go by prophesi on Tuesday November 21, 2017 @02:34PM (#15750808)
And even with ARM (I'm not familiar with RISC-V), you're likely going to have binary blobs for critical drivers.
Re:Replacing x86 firmware with Linux and Go by AstralStorm on Tuesday November 21, 2017 @02:43PM (#15750898)
Worse, in Qualcomm chips you have essentially the same OS as in AMD "Secure" Processor. Trustonic TEE OS. Handling ARM "Trust"Zone.
Re:Replacing x86 firmware with Linux and Go by ksk on Tuesday November 21, 2017 @02:37PM (#15750844)
>I say it also has to do with them just not caring about what their users want.

Why are people buying their products?

Re:Replacing x86 firmware with Linux and Go by colejohnson66 on Tuesday November 21, 2017 @03:00PM (#15751033)
Because the only alternative is AMD, who, until Ryzen, was lacking in performance quite a bit.
Re:Replacing x86 firmware with Linux and Go by ksk on Tuesday November 21, 2017 @04:34PM (#15751802)
In that case, it actually goes against your "they don't care about their users" narrative. Or maybe they partially care about their users.

Anyway, I see some value in the features that ME provides, and so I'm not as anti-ME as a lot of the commenters on here. But obviously, I want the security bugs to be fixed too.

Re:Replacing x86 firmware with Linux and Go by colejohnson66 on Tuesday November 21, 2017 @08:12PM (#15753347)
What I was trying to say is that Intel doesn’t really have a financial incentive to have non-ME SKUs because, besides the majority of users not caring, those who do care don’t really have any other options.
Re:Replacing x86 firmware with Linux and Go by drvdevd on Tuesday November 21, 2017 @01:59PM (#15750535)
Sadly, the main reason IMO this isn't possible is not just that desktop software is designed for the x86 instruction set, but that it's designed for lots of RAM and CPU usage, when it could be slimmer.
Re:Replacing x86 firmware with Linux and Go by ksk on Tuesday November 21, 2017 @02:35PM (#15750824)
I want it. Wearing multiple hats at a small company, I have to occasionally reimage machines and this would make it very useful for me.
Re:Replacing x86 firmware with Linux and Go by Oribi on Tuesday November 21, 2017 @09:43PM (#15753792)
You've already got it! If you are reimaging physical machines, most server class machines have IPMI features that allow reimaging.