Conflicts about reported defects in CPU design

[Geek-9pm]

Recently there have been reports of some kind of strange defects in almost all current CPU designs. These defects have been reported as 'Meltdown and Spectre exploits .'
Some reports claim the problem can be found in  ARM, AMD and Intell devices.
Below are some recent links:

https://www.theguardian.com/technology/2018/jan/04/meltdown-spectre-worst-cpu-bugs-ever-found-affect-computers-intel-processors-security-flaw

https://www.windowscentral.com/all-modern-processors-impacted-new-meltdown-and-spectre-exploits

https://www.theregister.co.uk/2018/01/02/intel_cpu_design_flaw/

http://appleinsider.com/articles/18/01/03/intel-claims-cpu-security-flaw-not-unique-to-its-chips-implies-arm-and-amd-chips-could-be-affected

Somehow this issue relates to an article by Forbes to the effect that AMD will benefit in this CPU defect debate. It was published on Jan 4 and You can read it below:
https://www.forbes.com/sites/kenkam/2018/01/04/amd-looks-poised-to-gain-at-intels-expense/#196ea84d74ec

The most recent was a 'good news for Linus' article form ZDNet.
http://www.zdnet.com/google-amp/article/linux-4-15-good-news-and-bad-news-about-meltdown-and-spectre/

Linus Torvalds, Linux's primary creator, had good and bad news about the chip security problems Meltdown and Spectre. The good news is the lead up to the Linux 4.15 was "quiet and small, and no last-minute panics, just small fixes for various issues". The bad news? "It's not like we're 'done' with Spectre/Meltdown."
On the Linux Kernel Mailing List (LKML), Torvalds explained, "The bulk of the 4.15 work is all the regular plodding 'boring' stuff. And I mean that in the best possible way. It may not be glamorous and get the headlines, but it's the bread and butter of kernel development, and is in many ways the really important stuff."

The above ZDNet story also has some more links and bout Intel and Linux and the reported CPU security issues.
In essence, they say it has not been fixed yet.   

[patio]

I don't consider even 1 of those sources credible...

I certainly don't respect their so-called Tech writers either.

[Geek-9pm]

I don't consider even 1 of those sources credible...

I certainly don't respect their so-called Tech writers either.

So noted. Here is where you  can get back to the first source:
https://meltdownattack.com/
They link to papers from  the University of Pennsylvania and University of Maryland.
Very bestirring. No pictures at all. 

A quote  with the tech stuff:

The issue affects Intel CPUs broadly, but also AMD and various ARM processors are suspect to a similar attack. Browser vendors have already started mitigating the issue with Microsoft, for example announcing improvements to Internet Explorer and Microsoft Edge browsers against Speculative Execution. Mozilla has also taken action against the new class of timing attacks and Chromium based browers a fix is scheduled for version 64. The WebKit team also did a writeup on the implications of Spectre and Meltdown in their blog.

https://react-etc.net/entry/exploiting-speculative-execution-meltdown-spectre-via-javascript

[patio]

You still don't get it...it is NOT a manuf. defect...it's a weakness that was being exploited.

BTW all the fixes to date have been more a nightmare than the exploit itself...

[Geek-9pm]

The Title is:
Re: Conflicts about reported defects in CPU design

In the reports there was no mention I saw about defects in the factory.
A very  weak over done design that makes people cry is a defect in design. The fact that it works good in an honest society does not justify its use. Does it? In a hostile world anything that has a weakness can be seen as a defect. One could argue that, but the fact remains that criminals take advantage of  whatever can be exploited. whether a law, a person or a machine.

The concept was to make a CPU faster by speculative pee fetch. The CPU does not really needed to per-fetch, but the engineers who design logic think the speed increase is worth the risk of something networking right.

If somehow I came across as saying there was a glitch in the factory, I Herby sate that it was not my intent to say the factories made the mistake. The defect or flaw is more at the abstract level. The CPU shall put unverified code or data into a cache if it might be needed.  The issue is when or how will the code or data be certified or tested for evil intentions of people.

Additionally, even if everybody in the world was honest, there is the question of data integrity.  Will corrupted data be allowed to enter the cache?

Presently there is no proof t his issue is causing any harm.  There are other things to worry about. I hope this post will help others locate the information they need. There is no  coming danger from the 'meltdown' thing..  The bad things that have happened were due to some plain simple  lazy security habits.

What we learn  is that a some thnings we  believe are fool proof and not.

[patio]

The Title is:
Re: Conflicts about reported defects in CPU design

Your words...not mine.

[Geek-9pm]

What's your definition of defect? 

fault, flaw, imperfection, deficiency, weakness, weak spot, inadequacy, shortcoming, limitation, failing; kink, deformity, blemish

In the above the abstract set comes comes  before the concrete set.

Yhe CPU has a design deflect. Or weakness. Or shortcoming.

[BC_Programmer]

I would say they are defects- similar to things like the FDIV bug, as they are issues with how the CPU operates that are part of it's design.

I think their realistic impact has been overstated. As Patio mentions the impact on people by the supposed fixes- such as for example making many AMD systems unbootable- is probably a greater impact than any security problems.

The Meltdown exploit which affects only Intel is the "easier" to exploit. It requires carefully crafted machine code and allows a process to read areas of memory in it's address space which normally are not accessible to it and are used by the OS Kernel. Spectre is a similar class of bug but requires even more carefully crafted machine code instructions and data is "guessed" based on specific timings of those instructions.

It doesn't seem like a vulnerability that is likely to allow particularly private information to be taken. Some sites claim it allows sensitive data to be stolen, but kernel memory doesn't have things like say passwords, and given it needs very careful ASM programming the exploit even the "easy" Meltdown exploit I'd argue that the common premise that this is something that "any website" can use to steal your passwords with javascript is just another act of the ongoing security circus.

[Geek-9pm]

Yes, the danger is overstated.

But what is worrisome is that for such a long time nobody ever nmotices.
One might wonder if there are other things the CPU can do that are not well known? Would there be some hidden danger?

Years ago some CPUs had undocumented instructions in the code set and hobbyists loved to speculate of what they might do. 

Modern CPU designs are so very complex that it would very very hard to verify that a design does not have a vulnerability that might show how be exploited.
Here is an article that explains this statement:
A CPU Researcher Explains Why It Took 22 Years to Discover Fundamental Chip Flaw
In this context the term 'clup flaw' does not mean something went bad in the factory that makes the chips. No, rather, it is behavior that is thought to be worrisome or unwanted.

To me, a layman, it’s odd that CPUs require so much research, since the architecture is designed by humans. Why do they require so much outside research to sort of understand what they’re doing?

Because CPUs are remarkably complex.
...
I’ve seen people writing that Spectre and Meltdown are a result of chip manufacturers favoring speed over security.

I’m not sure because I’m not really a CPU designer and the topic is ridiculously complex, but my guess is that they could have done something relatively easy. I’m not trying to say that Intel can do something about it in a short time frame because that’s an entirely different question, but I would guess they were taken by surprise by that.

You will find no simple answers.
Well, not simple answer of amazing value.

[BC_Programmer]

As I alluded, I think it's really just another symptom of the "Security Circus" that seems to pervade modern computing.

There is an excessive focus on exploits, software vulnerabilities, security updates, security flaws, and so forth and not nearly enough focus on educating users on applying basic common sense to their computing; to much presumption that missing security updates directly leads to malware infection when typically it is the user not being proficient in identifying threats or performing their own threat analysis of things they do or intend to do with their PC that causes problems.

Fact is that a diligent user isn't going to be  impacted by these CPU vulnerabilities because they will at some level require the exploitation of user vulnerabilities in order to be utilized.

[Geek-9pm]

I agree. This topic was blow out of proportion.

You remarks remind me that I need to get a better password manager.
The one I like to use has been downgraded by Mozilla, so I need to go over to a more secure add-on for Firefox.

Just yesterday I did a scan with  Malwarebytes  and was surprised about how many PUPs were found.  One has to be careful.

And yes, a complex exploit wooed have to have some kind of malware to let in into the system. So the solution is to maintain  and fortify the basic common sense security.