Linus Torvalds doesn't agree >>I'd say: "Totally insignificant"<<

http://www.realworldtech.com/forums/index.cfm?action=detail&id=8055...

Theo at times gets himself all worked up for little reason:

http://thread.gmane.org/gmane.linux.kernel.wireless.general/1558

I'll bet AMD would like to have a chip as "buggy" as Conroe....
Mines been hammering at 3.2ghz since last November without missing a beat.
Let's hope Bugcelona proves to be as reliable...

However, Microsoft didn't release any information about what bugs the update actually fixes or what bugs may still remain.
Probably because Intel put in a veto...

Buggy as hell.......... hmm............ thanks for going into specifics there.....

Core 2 has been out for a year and has been running a huge variety of code, including two significantly different Windows kernel revisions and the entire galaxy of *nix. If there were any bugs that seriously affected even obscure running code, there would have been quite a stink. But there hasn't been a serious problem with any of it, and that includes BSD . Of course, there could be code that no sane person would write except to try to exploit one of these bugs, and that is always a concern, but that's the point of publishing the information so that the OS vendors can evaluate that possibility and work with Intel there's a potential problem.

All processors have bugs. Intel does a much better job of documenting them than most, and a better job of issuing microcode updates than many. Because their CPUs are used so widely, the worst of the bugs get found quickly (and indeed the worst are found before the design ever leaves simulation, or testing, so the "worst" that actually appear in shipping chips are generally insignificant). And because Intel is so large and their x86 CPUs so important, they can afford to produce a new processor stepping to fix the ones that must be fixed and can't be fixed in any other way.

But that's rare. Many of these "bugs" affect no working code in the real world; in some cases the errata is simply documenting behavior that was undefined previously, or warning you that if your code performs a sequence of operations that no code has ever performed (and no rational programmer would write, and/or no compiler would emit) then the results may be unexpected. In cases that do matter in the real world. Intel can often issue a microcode update, and the OS vendors typically work around the rest, just as they have for previous versions of Opterons and Pentiums.

And yes, the Opteron has errata and Windows and Linux and BSD have workarounds for some (but not all) of them. Like I said, all processors have bugs. But does is a bug that doesn't matter remain a bug?

I don't know enough to evaluate the claim that "20-30" could be exploited; that sounds high, but de Raadt does have the technical credentials to be taken seriously (even if he's prone to hyperbole at times). But this sounds to be as much about posturing to not "get left out in the cold" when Intel is providing information and fixes as it is about genuine concerns regarding exploitable bugs.

I work for Intel, and I thought I would give you our perspective. Months ago, we addressed a processor issue by providing a BIOS update for our customers that in no way affects system performance. We publicly documented this as an erratum in April. All processors from all companies have errata, and Intel has a well-known errata communication process to inform our customers and the public. Keep in mind the probability of encountering this issue is extremely low. Specification Updates for the affected processors are available at http://developer.intel.com. All errata are thoroughly investigated for issues and vulnerabilities, should they have any we fix them, usually through a microcode update. We feel we’ve resolved the issue and were open about it with customers and then publicly publishing it, but this is a good venue for ideas on how we could do better or more. I am interested in any constructive comments...

It updates the CPU's instruction set but doesn't fully "flash" the microcode on board with a complete new instruction set as is the case with a GPU BIOS flash.(I think) :-)

OMG, my CPU is full of bugs!!

Wait a sec, It's got a 17-day uptime and the last reboot was because I moved it, not because it crashed. Must be all those bugs saving up for a fatal exception to end all fatal exceptions...

OH NOES, We're all going to die!

Linus Torvalds posted on every day tech (or some website with a name similar to that, I forget) that these bugs are being blown out of proportion and went on to explain about bugs vs "features" of chips.

*edit* oops someone beat me to it. ANd provided links!

Are these security concerns, stability concerns, or computational accuracy concerns?

Some of this sounds less than unbiased: To paraphrase: "The Intel Core 2 processor has serious bugs, and Intel isn't sending us fixes very effectively. So boycott Core 2's. Oh, and AMD chips have bugs too, and AMD isn't being very forthcoming about helping us either, but I won't tell you not to buy their stuff."

maybe this is the stuff hellokitty was talking about :-P

Software guys hate it when the hardware guys make more work for them by forcing workarounds for buggy hardware. Hardware guys hate it when the software guys paint them into a corner by writing buggy code that happens to run that they then have to make sure continues to work on every subsequent chip.

The difference is that there aren't any significant open source microprocessors. So the Open Source software guys bitch in public, and the hardware guys bitch behind closed doors. You shouldn't make qualitative judgments based on the quantity of bitching you happen to hear.

Please, this is story is 95% FUD and 5% maybe it is a problem.

Every CPU architech has inherent problems that cannot be fixed. I still can remember the infamous "PIV" bug in P5 and some FFU issues in K6. Were these bugs show stoppers? Nope.

What did anyone expect? This is a new processor, and VERY complex. Issues listed are triggered in rare cases, but I am sure people will get right on it and exploit the problems. As stated: All processors have erratas, and for the complexity of this processor, it is quite short. Intel still amazes me.

Buggy as hell is not how I would rank the processor, I would rank it as: new. Not quite mature...

-LS

Actually, if you take a look at the link to his email rant he does go into specifics and provides a few links to go into further detail.

The summaries on TR aren't supposed to provide you will all the details, you need to read the material provided in the links.

I am dam sure Shintai would have had a solution for this ..