US-CERT Discloses Security Flaw in Intel Chips

CommanderFrank

Cat Can't Scratch It
Joined
May 9, 2000
Messages
75,399
According to The U.S. Computer Emergency Readiness Team, a flaw has been discovered in Intel chips that would allow hackers to execute malicious code on a variety of 64-bit operating systems. The flaw is limited to chips using SYSRET instruction. Intel has made no comment on the disclosure.

"If an operating system is written according to AMD's spec, but run on Intel hardware, the difference in implementation can be exploited by an attacker to write to arbitrary addresses in the operating system's memory."
 
... "but run on Intel hardware"

Did Intel just copy AMD's code and put it in their chips?
 
I guess that means that I'll be using my Pentium 3 & 4 laptops and the VIA C3 more.
 
Amd trojan horse in the form of 64 bit execution? would make a great conspiracy!! muahahah
 
Wasn't it really hard to get the chip to calculate something incorrectly or was that just Intel downplaying the problem. I didn't own a Pentium until they'd revised the design to correct the problem.
 
... "but run on Intel hardware"

Did Intel just copy AMD's code and put it in their chips?

AMD created "AMD64" which allowed 64 Bit processors to run 32 bit programs. This is how AMD came to be so popular with the Athlon64. Intel then licensed the technology from AMD so their processors could also be 64 bit, but still run 32 bit applications.

This is why when you boot up Windows in safe mode (or even when booting some Linux's) you will see an AMD64 process startup even though you have an Intel processor.

So AMD has to license the x86 instruction from Intel in order to make a processor, but Intel has to license AMD64 from AMD to make their processors 64 bit (and still work with everything). Its like a technological Mutual Assured Destruction!
 
AMD created "AMD64" which allowed 64 Bit processors to run 32 bit programs. This is how AMD came to be so popular with the Athlon64. Intel then licensed the technology from AMD so their processors could also be 64 bit, but still run 32 bit applications.

This is why when you boot up Windows in safe mode (or even when booting some Linux's) you will see an AMD64 process startup even though you have an Intel processor.

So AMD has to license the x86 instruction from Intel in order to make a processor, but Intel has to license AMD64 from AMD to make their processors 64 bit (and still work with everything). Its like a technological Mutual Assured Destruction!

I never knew that Intel ended up licensing their 64 bit tech from AMD. I always thought they'd done something with IA64 in their Itanium line to make it all work in x86. It's interesting how much technological incest there is going on out there.
 
The actual linked article does not say anything about code designed to AMD spec causing the problem on Intel processors. It merely states the problem is with Intel's implementation of the SYSRET instruction. The article title here on [H] implies something more.
 
You hafta kinda read between the lines in the article to draw that conclusion about AMD.
 
I never knew that Intel ended up licensing their 64 bit tech from AMD. I always thought they'd done something with IA64 in their Itanium line to make it all work in x86. It's interesting how much technological incest there is going on out there.

AMD and intel have had a cross licensing agreement for a long time and they share a lot of things for the x86
 
I never knew that Intel ended up licensing their 64 bit tech from AMD. I always thought they'd done something with IA64 in their Itanium line to make it all work in x86. It's interesting how much technological incest there is going on out there.

Duh, if they enforced patents like RIAA enforces copyrights, the entire industry would grind to a halt and pretty much all products embargoed.
 
Intel thought about trying this back in 2009. They gave AMD a 60 day ultimatum or else they would revoke the x86 license from AMD. Obviously in the end nothing happened since we still have both AMD and Intel. Like I said Assured Mutual Destruction :)
 
I never knew that Intel ended up licensing their 64 bit tech from AMD. I always thought they'd done something with IA64 in their Itanium line to make it all work in x86. It's interesting how much technological incest there is going on out there.

It can be generally said that AMD and Intel have open portfolios with the other due to an agreement in Intel's monopoly lawsuit settlements.

"In 1991, AMD filed an antitrust lawsuit against Intel claiming that they were trying to secure and maintain a monopoly,[citation needed] and one year later, a court ruled against Intel, awarding AMD US$10 million "plus a royalty-free license to any Intel patents used in AMD's own 386-style processor"."

"In November 2009, Intel agreed to pay AMD $1.25bn and renew a five-year patent cross-licensing agreement as part of a deal to settle all outstanding legal disputes between them."

Generally speaking, they use what the other gets. AMD got the upper hand with their 64bit x86 implementation and Intel had to follow suit because Itanium had no x86 backwards compatibility and was generally a miserable slog of a chip. Legacy won out.
 
I thought that address space layout randomization in memory was supposed to prevent these types of attacks from doing much of anything. I felt secure in my 16GB of random-izable space.
 
I never knew that Intel ended up licensing their 64 bit tech from AMD. I always thought they'd done something with IA64 in their Itanium line to make it all work in x86. It's interesting how much technological incest there is going on out there.

Cross-licensing so it isn't an issue. They do this with all of their ISAs derived under/included with x86. Intel's is never called AMD64, the actual name of the 64-bit x86 derivative but rather x86-64 or EM64T or Intel 64.

Intel's REAL 64-bit x86 doesn't exist because they opted to go with IA64, aka Itanium.
 
You have a math coprocessor! I've been looking for one to upgrade my 25 MHz 368 DX forever!

I thought I was sooo cool when I finally got a board with a numeric processor on it.

///making me all nostalgic
 
You have a 368! I've been trying to fine one of those obscure three-sixty-eights forever! ;)

Curse my stupid dyslexic brain! Notably, 386 DX 25 and the actual mathco is laying in a box someplace with my tons of other old processors like a Zilog Z-80, 8088, etc.
 
Wasn't it really hard to get the chip to calculate something incorrectly or was that just Intel downplaying the problem. I didn't own a Pentium until they'd revised the design to correct the problem.

it was apparently pretty hard to find, and was due to an error in the floating point lookup table, and would affect next to no one, other than people doing sciencey/maths stuff. But people freaked out and went normal. Article covers it (also this is relevant)!
 
Just another of many reasons that x86 CISC processors are a buggy architecture.
It's too bad Intel and AMD have to keep making x86 processors.
 
I love how you hear of an OS vulnerability only to find it was patched out last week.

They don't like announcing big vulnerabilities without fixes already out in case someone tries to take advantage of it. In some ways, it makes sense.
 
Back
Top