Processors

Update on Core 2 CPU bugs


It was just earlier this week when I posted a piece about a patch for Intel chips that was quietly released. If you followed the story, news of the bug was first posted by Intel in late April inside a Core 2 Specification Update PDF document. It was explained how internal testing caught an error that could potentially result in memory leaks and, ultimately, crashes.

Microsoft released a patch pertaining to this issue recently, but the download page was terse and the explanation was non-existent. Because of this dearth of information, there was speculation from some quarters that it could be a Windows bug as opposed to a hardware fault with the Core 2 processor.

Well, looks like The Inquirer has got the full scoop outlined here.

To summarize, it is indeed a CPU bug, and Intel partners were informed upon discovery of the issue. Later, Intel pushed out a microcode fix in the form of a BIOS update to its various hardware partners.

Most of the products in the channel already have the correct BIOSes right from the start. If not, downloading and applying the latest BIOS update should do the trick. On the OS front, other than Microsoft, it appears that Apple has already updated Mac OS X, while Linux -- according to Linus Torvalds -- is probably unaffected by the whole issue.

The eye-opener here is the assertion from Linus that "CPUs have always had bugs." For all you know, your next BSOD might not be due to poorly written code in Windows after all.  In fact, OpenBSD founder Theo de Raadt has a lot of not so kind words to say about the bugs in Core 2.

Do you think that faulty hardware -- motherboard, CPU, RAM, etc. -- could have caused its fair share of BSODs, as well as erratic behaviour? Join the discussion.

About

Paul Mah is a writer and blogger who lives in Singapore, where he has worked for a number of years in various capacities within the IT industry. Paul enjoys tinkering with tech gadgets, smartphones, and networking devices.

4 comments
fredeppy
fredeppy

I think that, like every other industry out there, the computer industry has gotten caught up in the rush to get newer and better tech out before the next guy and before it is thoroughly ready. All of our industial base should be looking at making what they have out there work better. Everyone is in such a rush to put out new products that do everything in one package, that we lose track of the initial objective, make it better. I do not want to watch tv on my computer, I do not want to make phone calls over it, and I do not wish to replace my home stereo system with it. I also do not wish to use it for gaming either, I have a SNES game system for that and it's good enough for my purposes. My computer is a tool and that is all it is. It is for online research, office type work and email (90% of which is sent to the trash can, unopened. I ocassionaly do some shopping and instant messaging also.

paulmah
paulmah

Do you think that faulty hardware - motherboard, CPU, RAM etc, could have caused its fair share of BSODs as well as erratic behaviour?

DanLM
DanLM

Is the bug that was found in the chips in the area of the x86 chip architecture? Haven't there been previous conversations here about how bloated this architecture is due to backward compatibilities? Would both the complexity and number of hidden bugs be reduced if some of these backward compatibilities were removed? I swear I seen this conversation here previously. Or was the bios on these boards only for 64 bit, which I just read does not use the x86 chip architecture. Dan

paulmah
paulmah

Hi Dan, It appears the the bug this time round involves the "TLB" or Translation Lookside Buffer. It appears that Intel modified how it operated, in an attempt to improve it, but without adequate notification to the software (OS) folks. In that case, it probably has nothing to do with legacy x86 architecture at all. Though of course, I'm sure there is a fairly extensive list of x86 bugs as well - and which is probably what you read about previously.

Editor's Picks